| 1 | \input texinfo @c -*-texinfo-*- |
| 2 | @comment %**start of header |
| 3 | @setfilename ../info/eintr |
| 4 | @c setfilename emacs-lisp-intro.info |
| 5 | @c sethtmlfilename emacs-lisp-intro.html |
| 6 | @settitle Programming in Emacs Lisp |
| 7 | @syncodeindex vr cp |
| 8 | @syncodeindex fn cp |
| 9 | @setchapternewpage odd |
| 10 | @finalout |
| 11 | |
| 12 | @c --------- |
| 13 | @c <<<< For hard copy printing, this file is now |
| 14 | @c set for smallbook, which works for all sizes |
| 15 | @c of paper, and with Postscript figures >>>> |
| 16 | @smallbook |
| 17 | @clear largebook |
| 18 | @set print-postscript-figures |
| 19 | @c set largebook |
| 20 | @c clear print-postscript-figures |
| 21 | @c --------- |
| 22 | |
| 23 | @comment %**end of header |
| 24 | |
| 25 | @set edition-number 2.14 |
| 26 | @set update-date 2004 Oct 12 |
| 27 | |
| 28 | @ignore |
| 29 | ## Summary of shell commands to create various output formats: |
| 30 | |
| 31 | pushd /usr/local/src/emacs/lispintro/ |
| 32 | |
| 33 | ## Info output |
| 34 | makeinfo --no-split --paragraph-indent=0 --verbose emacs-lisp-intro.texi |
| 35 | |
| 36 | ## DVI output |
| 37 | texi2dvi emacs-lisp-intro.texi |
| 38 | |
| 39 | ## HTML output |
| 40 | makeinfo --html --no-split --verbose emacs-lisp-intro.texi |
| 41 | |
| 42 | ## Plain text output |
| 43 | makeinfo --fill-column=70 --no-split --paragraph-indent=0 \ |
| 44 | --verbose --no-headers --output=emacs-lisp-intro.txt emacs-lisp-intro.texi |
| 45 | |
| 46 | ## DocBook output |
| 47 | makeinfo --docbook --no-split --paragraph-indent=0 \ |
| 48 | --verbose emacs-lisp-intro.texi |
| 49 | |
| 50 | ## XML output |
| 51 | makeinfo --xml --no-split --paragraph-indent=0 \ |
| 52 | --verbose emacs-lisp-intro.texi |
| 53 | |
| 54 | #### (You must be in the same directory as the viewed file.) |
| 55 | |
| 56 | ## View DVI output |
| 57 | xdvi emacs-lisp-intro.dvi & |
| 58 | |
| 59 | ## View HTML output |
| 60 | galeon emacs-lisp-intro.html |
| 61 | |
| 62 | ## View Info output with standalone reader |
| 63 | info emacs-lisp-intro.info |
| 64 | |
| 65 | ## popd |
| 66 | |
| 67 | @end ignore |
| 68 | |
| 69 | @c ================ Included Figures ================ |
| 70 | |
| 71 | @c Set print-postscript-figures if you print PostScript figures. |
| 72 | @c If you clear this, the ten figures will be printed as ASCII diagrams. |
| 73 | @c (This is not relevant to Info, since Info only handles ASCII.) |
| 74 | @c Your site may require editing changes to print PostScript; in this |
| 75 | @c case, search for `print-postscript-figures' and make appropriate changes. |
| 76 | |
| 77 | |
| 78 | @c ================ How to Create an Info file ================ |
| 79 | |
| 80 | @c If you have `makeinfo' installed, run the following command |
| 81 | |
| 82 | @c makeinfo emacs-lisp-intro.texi |
| 83 | |
| 84 | @c or, if you want a single, large Info file, and no paragraph indents: |
| 85 | @c makeinfo --no-split --paragraph-indent=0 --verbose emacs-lisp-intro.texi |
| 86 | |
| 87 | @c After creating the Info file, edit your Info `dir' file, if the |
| 88 | @c `dircategory' section below does not enable your system to |
| 89 | @c install the manual automatically. |
| 90 | @c (The `dir' file is often in the `/usr/local/info/' directory.) |
| 91 | |
| 92 | @c ================ How to Create an HTML file ================ |
| 93 | |
| 94 | @c To convert to HTML format |
| 95 | @c makeinfo --html --no-split --verbose emacs-lisp-intro.texi |
| 96 | |
| 97 | @c ================ How to Print a Book in Various Sizes ================ |
| 98 | |
| 99 | @c This book can be printed in any of three different sizes. |
| 100 | @c In the above header, set @-commands appropriately. |
| 101 | |
| 102 | @c 7 by 9.25 inches: |
| 103 | @c @smallbook |
| 104 | @c @clear largebook |
| 105 | |
| 106 | @c 8.5 by 11 inches: |
| 107 | @c @c smallbook |
| 108 | @c @set largebook |
| 109 | |
| 110 | @c European A4 size paper: |
| 111 | @c @c smallbook |
| 112 | @c @afourpaper |
| 113 | @c @set largebook |
| 114 | |
| 115 | @c ================ How to Typeset and Print ================ |
| 116 | |
| 117 | @c If you do not include PostScript figures, run either of the |
| 118 | @c following command sequences, or similar commands suited to your |
| 119 | @c system: |
| 120 | |
| 121 | @c texi2dvi emacs-lisp-intro.texi |
| 122 | @c lpr -d emacs-lisp-intro.dvi |
| 123 | |
| 124 | @c or else: |
| 125 | |
| 126 | @c tex emacs-lisp-intro.texi |
| 127 | @c texindex emacs-lisp-intro.?? |
| 128 | @c tex emacs-lisp-intro.texi |
| 129 | @c lpr -d emacs-lisp-intro.dvi |
| 130 | |
| 131 | @c If you include the PostScript figures, and you have old software, |
| 132 | @c you may need to convert the .dvi file to a .ps file before |
| 133 | @c printing. Run either of the following command sequences, or one |
| 134 | @c similar: |
| 135 | @c |
| 136 | @c dvips -f < emacs-lisp-intro.dvi > emacs-lisp-intro.ps |
| 137 | @c |
| 138 | @c or else: |
| 139 | @c |
| 140 | @c postscript -p < emacs-lisp-intro.dvi > emacs-lisp-intro.ps |
| 141 | @c |
| 142 | |
| 143 | @c (Note: if you edit the book so as to change the length of the |
| 144 | @c table of contents, you may have to change the value of `pageno' below.) |
| 145 | |
| 146 | @c ================ End of Formatting Sections ================ |
| 147 | |
| 148 | @c For next or subsequent edition: |
| 149 | @c create function using with-output-to-temp-buffer |
| 150 | @c create a major mode, with keymaps |
| 151 | @c run an asynchronous process, like grep or diff |
| 152 | |
| 153 | @c For 8.5 by 11 inch format: do not use such a small amount of |
| 154 | @c whitespace between paragraphs as smallbook format |
| 155 | @ifset largebook |
| 156 | @tex |
| 157 | \global\parskip 6pt plus 1pt |
| 158 | @end tex |
| 159 | @end ifset |
| 160 | |
| 161 | @c For all sized formats: print within-book cross |
| 162 | @c reference with ``...'' rather than [...] |
| 163 | |
| 164 | @c This works with the texinfo.tex file, version 2003-05-04.08, |
| 165 | @c in the Texinfo version 4.6 of the 2003 Jun 13 distribution. |
| 166 | |
| 167 | @tex |
| 168 | \global\def\xrefprintnodename#1{``#1''} |
| 169 | @end tex |
| 170 | |
| 171 | @c ---------------------------------------------------- |
| 172 | |
| 173 | @dircategory Emacs |
| 174 | @direntry |
| 175 | * Emacs Lisp Intro: (eintr). |
| 176 | A simple introduction to Emacs Lisp programming. |
| 177 | @end direntry |
| 178 | |
| 179 | @copying |
| 180 | This is an @cite{Introduction to Programming in Emacs Lisp}, for |
| 181 | people who are not programmers. |
| 182 | @sp 1 |
| 183 | Edition @value{edition-number}, @value{update-date} |
| 184 | @sp 1 |
| 185 | Copyright @copyright{} 1990, 1991, 1992, 1993, 1994, 1995, 1997, 2001, |
| 186 | 2002, 2003, 2004, 2005 Free Software Foundation, Inc. |
| 187 | @sp 1 |
| 188 | |
| 189 | @iftex |
| 190 | Published by the:@* |
| 191 | |
| 192 | GNU Press, @hfill @uref{http://www.gnupress.org}@* |
| 193 | a division of the @hfill General: @email{press@@gnu.org}@* |
| 194 | Free Software Foundation, Inc. @hfill Orders:@w{ } @email{sales@@gnu.org}@* |
| 195 | 51 Franklin Street, Fifth Floor @hfill Tel: +1 (617) 542-5942@* |
| 196 | Boston, MA 02110-1301 USA @hfill Fax: +1 (617) 542-2652@* |
| 197 | @end iftex |
| 198 | |
| 199 | @ifnottex |
| 200 | Published by the: |
| 201 | |
| 202 | @example |
| 203 | GNU Press, Website: http://www.gnupress.org |
| 204 | a division of the General: press@@gnu.org |
| 205 | Free Software Foundation, Inc. Orders: sales@@gnu.org |
| 206 | 51 Franklin Street, Fifth Floor Tel: +1 (617) 542-5942 |
| 207 | Boston, MA 02110-1301 USA Fax: +1 (617) 542-2652 |
| 208 | @end example |
| 209 | @end ifnottex |
| 210 | |
| 211 | @sp 1 |
| 212 | @c Printed copies are available for $30 each.@* |
| 213 | ISBN 1-882114-43-4 |
| 214 | |
| 215 | Permission is granted to copy, distribute and/or modify this document |
| 216 | under the terms of the GNU Free Documentation License, Version 1.2 or |
| 217 | any later version published by the Free Software Foundation; there |
| 218 | being no Invariant Section, with the Front-Cover Texts being ``A GNU |
| 219 | Manual'', and with the Back-Cover Texts as in (a) below. A copy of |
| 220 | the license is included in the section entitled ``GNU Free |
| 221 | Documentation License''. |
| 222 | |
| 223 | (a) The FSF's Back-Cover Text is: ``You have freedom to copy and |
| 224 | modify this GNU Manual, like GNU software. Copies published by the |
| 225 | Free Software Foundation raise funds for GNU development.'' |
| 226 | @end copying |
| 227 | |
| 228 | @c half title; two lines here, so do not use `shorttitlepage' |
| 229 | @tex |
| 230 | {\begingroup% |
| 231 | \hbox{}\vskip 1.5in \chaprm \centerline{An Introduction to}% |
| 232 | \endgroup}% |
| 233 | {\begingroup\hbox{}\vskip 0.25in \chaprm% |
| 234 | \centerline{Programming in Emacs Lisp}% |
| 235 | \endgroup\page\hbox{}\page} |
| 236 | @end tex |
| 237 | |
| 238 | @titlepage |
| 239 | @sp 6 |
| 240 | @center @titlefont{An Introduction to} |
| 241 | @sp 2 |
| 242 | @center @titlefont{Programming in Emacs Lisp} |
| 243 | @sp 2 |
| 244 | @center Revised Second Edition |
| 245 | @sp 4 |
| 246 | @center by Robert J. Chassell |
| 247 | |
| 248 | @page |
| 249 | @vskip 0pt plus 1filll |
| 250 | @insertcopying |
| 251 | @end titlepage |
| 252 | |
| 253 | @iftex |
| 254 | @headings off |
| 255 | @evenheading @thispage @| @| @thischapter |
| 256 | @oddheading @thissection @| @| @thispage |
| 257 | @end iftex |
| 258 | |
| 259 | @ifnothtml |
| 260 | @c Keep T.O.C. short by tightening up for largebook |
| 261 | @ifset largebook |
| 262 | @tex |
| 263 | \global\parskip 2pt plus 1pt |
| 264 | \global\advance\baselineskip by -1pt |
| 265 | @end tex |
| 266 | @end ifset |
| 267 | @end ifnothtml |
| 268 | |
| 269 | @shortcontents |
| 270 | @contents |
| 271 | |
| 272 | @ifnottex |
| 273 | @node Top, Preface, (dir), (dir) |
| 274 | @top An Introduction to Programming in Emacs Lisp |
| 275 | |
| 276 | @insertcopying |
| 277 | |
| 278 | This master menu first lists each chapter and index; then it lists |
| 279 | every node in every chapter. |
| 280 | @end ifnottex |
| 281 | |
| 282 | @menu |
| 283 | * Preface:: What to look for. |
| 284 | * List Processing:: What is Lisp? |
| 285 | * Practicing Evaluation:: Running several programs. |
| 286 | * Writing Defuns:: How to write function definitions. |
| 287 | * Buffer Walk Through:: Exploring a few buffer-related functions. |
| 288 | * More Complex:: A few, even more complex functions. |
| 289 | * Narrowing & Widening:: Restricting your and Emacs attention to |
| 290 | a region. |
| 291 | * car cdr & cons:: Fundamental functions in Lisp. |
| 292 | * Cutting & Storing Text:: Removing text and saving it. |
| 293 | * List Implementation:: How lists are implemented in the computer. |
| 294 | * Yanking:: Pasting stored text. |
| 295 | * Loops & Recursion:: How to repeat a process. |
| 296 | * Regexp Search:: Regular expression searches. |
| 297 | * Counting Words:: A review of repetition and regexps. |
| 298 | * Words in a defun:: Counting words in a @code{defun}. |
| 299 | * Readying a Graph:: A prototype graph printing function. |
| 300 | * Emacs Initialization:: How to write a @file{.emacs} file. |
| 301 | * Debugging:: How to run the Emacs Lisp debuggers. |
| 302 | * Conclusion:: Now you have the basics. |
| 303 | * the-the:: An appendix: how to find reduplicated words. |
| 304 | * Kill Ring:: An appendix: how the kill ring works. |
| 305 | * Full Graph:: How to create a graph with labelled axes. |
| 306 | * Free Software and Free Manuals:: |
| 307 | * GNU Free Documentation License:: |
| 308 | * Index:: |
| 309 | * About the Author:: |
| 310 | |
| 311 | @detailmenu |
| 312 | --- The Detailed Node Listing --- |
| 313 | |
| 314 | Preface |
| 315 | |
| 316 | * Why:: Why learn Emacs Lisp? |
| 317 | * On Reading this Text:: Read, gain familiarity, pick up habits.... |
| 318 | * Who You Are:: For whom this is written. |
| 319 | * Lisp History:: |
| 320 | * Note for Novices:: You can read this as a novice. |
| 321 | * Thank You:: |
| 322 | |
| 323 | List Processing |
| 324 | |
| 325 | * Lisp Lists:: What are lists? |
| 326 | * Run a Program:: Any list in Lisp is a program ready to run. |
| 327 | * Making Errors:: Generating an error message. |
| 328 | * Names & Definitions:: Names of symbols and function definitions. |
| 329 | * Lisp Interpreter:: What the Lisp interpreter does. |
| 330 | * Evaluation:: Running a program. |
| 331 | * Variables:: Returning a value from a variable. |
| 332 | * Arguments:: Passing information to a function. |
| 333 | * set & setq:: Setting the value of a variable. |
| 334 | * Summary:: The major points. |
| 335 | * Error Message Exercises:: |
| 336 | |
| 337 | Lisp Lists |
| 338 | |
| 339 | * Numbers Lists:: List have numbers, other lists, in them. |
| 340 | * Lisp Atoms:: Elemental entities. |
| 341 | * Whitespace in Lists:: Formating lists to be readable. |
| 342 | * Typing Lists:: How GNU Emacs helps you type lists. |
| 343 | |
| 344 | The Lisp Interpreter |
| 345 | |
| 346 | * Complications:: Variables, Special forms, Lists within. |
| 347 | * Byte Compiling:: Specially processing code for speed. |
| 348 | |
| 349 | Evaluation |
| 350 | |
| 351 | * Evaluating Inner Lists:: Lists within lists... |
| 352 | |
| 353 | Variables |
| 354 | |
| 355 | * fill-column Example:: |
| 356 | * Void Function:: The error message for a symbol |
| 357 | without a function. |
| 358 | * Void Variable:: The error message for a symbol without a value. |
| 359 | |
| 360 | Arguments |
| 361 | |
| 362 | * Data types:: Types of data passed to a function. |
| 363 | * Args as Variable or List:: An argument can be the value |
| 364 | of a variable or list. |
| 365 | * Variable Number of Arguments:: Some functions may take a |
| 366 | variable number of arguments. |
| 367 | * Wrong Type of Argument:: Passing an argument of the wrong type |
| 368 | to a function. |
| 369 | * message:: A useful function for sending messages. |
| 370 | |
| 371 | Setting the Value of a Variable |
| 372 | |
| 373 | * Using set:: Setting values. |
| 374 | * Using setq:: Setting a quoted value. |
| 375 | * Counting:: Using @code{setq} to count. |
| 376 | |
| 377 | Practicing Evaluation |
| 378 | |
| 379 | * How to Evaluate:: Typing editing commands or @kbd{C-x C-e} |
| 380 | causes evaluation. |
| 381 | * Buffer Names:: Buffers and files are different. |
| 382 | * Getting Buffers:: Getting a buffer itself, not merely its name. |
| 383 | * Switching Buffers:: How to change to another buffer. |
| 384 | * Buffer Size & Locations:: Where point is located and the size of |
| 385 | the buffer. |
| 386 | * Evaluation Exercise:: |
| 387 | |
| 388 | How To Write Function Definitions |
| 389 | |
| 390 | * Primitive Functions:: |
| 391 | * defun:: The @code{defun} special form. |
| 392 | * Install:: Install a function definition. |
| 393 | * Interactive:: Making a function interactive. |
| 394 | * Interactive Options:: Different options for @code{interactive}. |
| 395 | * Permanent Installation:: Installing code permanently. |
| 396 | * let:: Creating and initializing local variables. |
| 397 | * if:: What if? |
| 398 | * else:: If--then--else expressions. |
| 399 | * Truth & Falsehood:: What Lisp considers false and true. |
| 400 | * save-excursion:: Keeping track of point, mark, and buffer. |
| 401 | * Review:: |
| 402 | * defun Exercises:: |
| 403 | |
| 404 | Install a Function Definition |
| 405 | |
| 406 | * Effect of installation:: |
| 407 | * Change a defun:: How to change a function definition. |
| 408 | |
| 409 | Make a Function Interactive |
| 410 | |
| 411 | * Interactive multiply-by-seven:: An overview. |
| 412 | * multiply-by-seven in detail:: The interactive version. |
| 413 | |
| 414 | @code{let} |
| 415 | |
| 416 | * Prevent confusion:: |
| 417 | * Parts of let Expression:: |
| 418 | * Sample let Expression:: |
| 419 | * Uninitialized let Variables:: |
| 420 | |
| 421 | The @code{if} Special Form |
| 422 | |
| 423 | * if in more detail:: |
| 424 | * type-of-animal in detail:: An example of an @code{if} expression. |
| 425 | |
| 426 | Truth and Falsehood in Emacs Lisp |
| 427 | |
| 428 | * nil explained:: @code{nil} has two meanings. |
| 429 | |
| 430 | @code{save-excursion} |
| 431 | |
| 432 | * Point and mark:: A review of various locations. |
| 433 | * Template for save-excursion:: |
| 434 | |
| 435 | A Few Buffer--Related Functions |
| 436 | |
| 437 | * Finding More:: How to find more information. |
| 438 | * simplified-beginning-of-buffer:: Shows @code{goto-char}, |
| 439 | @code{point-min}, and @code{push-mark}. |
| 440 | * mark-whole-buffer:: Almost the same as @code{beginning-of-buffer}. |
| 441 | * append-to-buffer:: Uses @code{save-excursion} and |
| 442 | @code{insert-buffer-substring}. |
| 443 | * Buffer Related Review:: Review. |
| 444 | * Buffer Exercises:: |
| 445 | |
| 446 | The Definition of @code{mark-whole-buffer} |
| 447 | |
| 448 | * mark-whole-buffer overview:: |
| 449 | * Body of mark-whole-buffer:: Only three lines of code. |
| 450 | |
| 451 | The Definition of @code{append-to-buffer} |
| 452 | |
| 453 | * append-to-buffer overview:: |
| 454 | * append interactive:: A two part interactive expression. |
| 455 | * append-to-buffer body:: Incorporates a @code{let} expression. |
| 456 | * append save-excursion:: How the @code{save-excursion} works. |
| 457 | |
| 458 | A Few More Complex Functions |
| 459 | |
| 460 | * copy-to-buffer:: With @code{set-buffer}, @code{get-buffer-create}. |
| 461 | * insert-buffer:: Read-only, and with @code{or}. |
| 462 | * beginning-of-buffer:: Shows @code{goto-char}, |
| 463 | @code{point-min}, and @code{push-mark}. |
| 464 | * Second Buffer Related Review:: |
| 465 | * optional Exercise:: |
| 466 | |
| 467 | The Definition of @code{insert-buffer} |
| 468 | |
| 469 | * insert-buffer code:: |
| 470 | * insert-buffer interactive:: When you can read, but not write. |
| 471 | * insert-buffer body:: The body has an @code{or} and a @code{let}. |
| 472 | * if & or:: Using an @code{if} instead of an @code{or}. |
| 473 | * Insert or:: How the @code{or} expression works. |
| 474 | * Insert let:: Two @code{save-excursion} expressions. |
| 475 | |
| 476 | The Interactive Expression in @code{insert-buffer} |
| 477 | |
| 478 | * Read-only buffer:: When a buffer cannot be modified. |
| 479 | * b for interactive:: An existing buffer or else its name. |
| 480 | |
| 481 | Complete Definition of @code{beginning-of-buffer} |
| 482 | |
| 483 | * Optional Arguments:: |
| 484 | * beginning-of-buffer opt arg:: Example with optional argument. |
| 485 | * beginning-of-buffer complete:: |
| 486 | |
| 487 | @code{beginning-of-buffer} with an Argument |
| 488 | |
| 489 | * Disentangle beginning-of-buffer:: |
| 490 | * Large buffer case:: |
| 491 | * Small buffer case:: |
| 492 | |
| 493 | Narrowing and Widening |
| 494 | |
| 495 | * Narrowing advantages:: The advantages of narrowing |
| 496 | * save-restriction:: The @code{save-restriction} special form. |
| 497 | * what-line:: The number of the line that point is on. |
| 498 | * narrow Exercise:: |
| 499 | |
| 500 | @code{car}, @code{cdr}, @code{cons}: Fundamental Functions |
| 501 | |
| 502 | * Strange Names:: An historical aside: why the strange names? |
| 503 | * car & cdr:: Functions for extracting part of a list. |
| 504 | * cons:: Constructing a list. |
| 505 | * nthcdr:: Calling @code{cdr} repeatedly. |
| 506 | * nth:: |
| 507 | * setcar:: Changing the first element of a list. |
| 508 | * setcdr:: Changing the rest of a list. |
| 509 | * cons Exercise:: |
| 510 | |
| 511 | @code{cons} |
| 512 | |
| 513 | * Build a list:: |
| 514 | * length:: How to find the length of a list. |
| 515 | |
| 516 | Cutting and Storing Text |
| 517 | |
| 518 | * Storing Text:: Text is stored in a list. |
| 519 | * zap-to-char:: Cutting out text up to a character. |
| 520 | * kill-region:: Cutting text out of a region. |
| 521 | * Digression into C:: Minor note on C programming language macros. |
| 522 | * defvar:: How to give a variable an initial value. |
| 523 | * copy-region-as-kill:: A definition for copying text. |
| 524 | * cons & search-fwd Review:: |
| 525 | * search Exercises:: |
| 526 | |
| 527 | @code{zap-to-char} |
| 528 | |
| 529 | * Complete zap-to-char:: The complete implementation. |
| 530 | * zap-to-char interactive:: A three part interactive expression. |
| 531 | * zap-to-char body:: A short overview. |
| 532 | * search-forward:: How to search for a string. |
| 533 | * progn:: The @code{progn} special form. |
| 534 | * Summing up zap-to-char:: Using @code{point} and @code{search-forward}. |
| 535 | |
| 536 | @code{kill-region} |
| 537 | |
| 538 | * Complete kill-region:: The function definition. |
| 539 | * condition-case:: Dealing with a problem. |
| 540 | * delete-and-extract-region:: Doing the work. |
| 541 | |
| 542 | Initializing a Variable with @code{defvar} |
| 543 | |
| 544 | * See variable current value:: |
| 545 | * defvar and asterisk:: An old-time convention. |
| 546 | |
| 547 | @code{copy-region-as-kill} |
| 548 | |
| 549 | * Complete copy-region-as-kill:: The complete function definition. |
| 550 | * copy-region-as-kill body:: The body of @code{copy-region-as-kill}. |
| 551 | |
| 552 | The Body of @code{copy-region-as-kill} |
| 553 | |
| 554 | * last-command & this-command:: |
| 555 | * kill-append function:: |
| 556 | * kill-new function:: |
| 557 | |
| 558 | How Lists are Implemented |
| 559 | |
| 560 | * Lists diagrammed:: |
| 561 | * Symbols as Chest:: Exploring a powerful metaphor. |
| 562 | * List Exercise:: |
| 563 | |
| 564 | Yanking Text Back |
| 565 | |
| 566 | * Kill Ring Overview:: The kill ring is a list. |
| 567 | * kill-ring-yank-pointer:: The @code{kill-ring-yank-pointer} variable. |
| 568 | * yank nthcdr Exercises:: |
| 569 | |
| 570 | Loops and Recursion |
| 571 | |
| 572 | * while:: Causing a stretch of code to repeat. |
| 573 | * dolist dotimes:: |
| 574 | * Recursion:: Causing a function to call itself. |
| 575 | * Looping exercise:: |
| 576 | |
| 577 | @code{while} |
| 578 | |
| 579 | * Looping with while:: Repeat so long as test returns true. |
| 580 | * Loop Example:: A @code{while} loop that uses a list. |
| 581 | * print-elements-of-list:: Uses @code{while}, @code{car}, @code{cdr}. |
| 582 | * Incrementing Loop:: A loop with an incrementing counter. |
| 583 | * Decrementing Loop:: A loop with a decrementing counter. |
| 584 | |
| 585 | A Loop with an Incrementing Counter |
| 586 | |
| 587 | * Incrementing Example:: Counting pebbles in a triangle. |
| 588 | * Inc Example parts:: The parts of the function definition. |
| 589 | * Inc Example altogether:: Putting the function definition together. |
| 590 | |
| 591 | Loop with a Decrementing Counter |
| 592 | |
| 593 | * Decrementing Example:: More pebbles on the beach. |
| 594 | * Dec Example parts:: The parts of the function definition. |
| 595 | * Dec Example altogether:: Putting the function definition together. |
| 596 | |
| 597 | Save your time: @code{dolist} and @code{dotimes} |
| 598 | |
| 599 | * dolist:: |
| 600 | * dotimes:: |
| 601 | |
| 602 | Recursion |
| 603 | |
| 604 | * Building Robots:: Same model, different serial number ... |
| 605 | * Recursive Definition Parts:: Walk until you stop ... |
| 606 | * Recursion with list:: Using a list as the test whether to recurse. |
| 607 | * Recursive triangle function:: |
| 608 | * Recursion with cond:: |
| 609 | * Recursive Patterns:: Often used templates. |
| 610 | * No Deferment:: Don't store up work ... |
| 611 | * No deferment solution:: |
| 612 | |
| 613 | Recursion in Place of a Counter |
| 614 | |
| 615 | * Recursive Example arg of 1 or 2:: |
| 616 | * Recursive Example arg of 3 or 4:: |
| 617 | |
| 618 | Recursive Patterns |
| 619 | |
| 620 | * Every:: |
| 621 | * Accumulate:: |
| 622 | * Keep:: |
| 623 | |
| 624 | Regular Expression Searches |
| 625 | |
| 626 | * sentence-end:: The regular expression for @code{sentence-end}. |
| 627 | * re-search-forward:: Very similar to @code{search-forward}. |
| 628 | * forward-sentence:: A straightforward example of regexp search. |
| 629 | * forward-paragraph:: A somewhat complex example. |
| 630 | * etags:: How to create your own @file{TAGS} table. |
| 631 | * Regexp Review:: |
| 632 | * re-search Exercises:: |
| 633 | |
| 634 | @code{forward-sentence} |
| 635 | |
| 636 | * Complete forward-sentence:: |
| 637 | * fwd-sentence while loops:: Two @code{while} loops. |
| 638 | * fwd-sentence re-search:: A regular expression search. |
| 639 | |
| 640 | @code{forward-paragraph}: a Goldmine of Functions |
| 641 | |
| 642 | * forward-paragraph in brief:: Key parts of the function definition. |
| 643 | * fwd-para let:: The @code{let*} expression. |
| 644 | * fwd-para while:: The forward motion @code{while} loop. |
| 645 | * fwd-para between paragraphs:: Movement between paragraphs. |
| 646 | * fwd-para within paragraph:: Movement within paragraphs. |
| 647 | * fwd-para no fill prefix:: When there is no fill prefix. |
| 648 | * fwd-para with fill prefix:: When there is a fill prefix. |
| 649 | * fwd-para summary:: Summary of @code{forward-paragraph} code. |
| 650 | |
| 651 | Counting: Repetition and Regexps |
| 652 | |
| 653 | * Why Count Words:: |
| 654 | * count-words-region:: Use a regexp, but find a problem. |
| 655 | * recursive-count-words:: Start with case of no words in region. |
| 656 | * Counting Exercise:: |
| 657 | |
| 658 | The @code{count-words-region} Function |
| 659 | |
| 660 | * Design count-words-region:: The definition using a @code{while} loop. |
| 661 | * Whitespace Bug:: The Whitespace Bug in @code{count-words-region}. |
| 662 | |
| 663 | Counting Words in a @code{defun} |
| 664 | |
| 665 | * Divide and Conquer:: |
| 666 | * Words and Symbols:: What to count? |
| 667 | * Syntax:: What constitutes a word or symbol? |
| 668 | * count-words-in-defun:: Very like @code{count-words}. |
| 669 | * Several defuns:: Counting several defuns in a file. |
| 670 | * Find a File:: Do you want to look at a file? |
| 671 | * lengths-list-file:: A list of the lengths of many definitions. |
| 672 | * Several files:: Counting in definitions in different files. |
| 673 | * Several files recursively:: Recursively counting in different files. |
| 674 | * Prepare the data:: Prepare the data for display in a graph. |
| 675 | |
| 676 | Count Words in @code{defuns} in Different Files |
| 677 | |
| 678 | * lengths-list-many-files:: Return a list of the lengths of defuns. |
| 679 | * append:: Attach one list to another. |
| 680 | |
| 681 | Prepare the Data for Display in a Graph |
| 682 | |
| 683 | * Sorting:: Sorting lists. |
| 684 | * Files List:: Making a list of files. |
| 685 | * Counting function definitions:: |
| 686 | |
| 687 | Readying a Graph |
| 688 | |
| 689 | * Columns of a graph:: |
| 690 | * graph-body-print:: How to print the body of a graph. |
| 691 | * recursive-graph-body-print:: |
| 692 | * Printed Axes:: |
| 693 | * Line Graph Exercise:: |
| 694 | |
| 695 | Your @file{.emacs} File |
| 696 | |
| 697 | * Default Configuration:: |
| 698 | * Site-wide Init:: You can write site-wide init files. |
| 699 | * defcustom:: Emacs will write code for you. |
| 700 | * Beginning a .emacs File:: How to write a @code{.emacs file}. |
| 701 | * Text and Auto-fill:: Automatically wrap lines. |
| 702 | * Mail Aliases:: Use abbreviations for email addresses. |
| 703 | * Indent Tabs Mode:: Don't use tabs with @TeX{} |
| 704 | * Keybindings:: Create some personal keybindings. |
| 705 | * Keymaps:: More about key binding. |
| 706 | * Loading Files:: Load (i.e., evaluate) files automatically. |
| 707 | * Autoload:: Make functions available. |
| 708 | * Simple Extension:: Define a function; bind it to a key. |
| 709 | * X11 Colors:: Colors in version 19 in X. |
| 710 | * Miscellaneous:: |
| 711 | * Mode Line:: How to customize your mode line. |
| 712 | |
| 713 | Debugging |
| 714 | |
| 715 | * debug:: How to use the built-in debugger. |
| 716 | * debug-on-entry:: Start debugging when you call a function. |
| 717 | * debug-on-quit:: Start debugging when you quit with @kbd{C-g}. |
| 718 | * edebug:: How to use Edebug, a source level debugger. |
| 719 | * Debugging Exercises:: |
| 720 | |
| 721 | Handling the Kill Ring |
| 722 | |
| 723 | * rotate-yank-pointer:: Move a pointer along a list and around. |
| 724 | * yank:: Paste a copy of a clipped element. |
| 725 | * yank-pop:: Insert first element pointed to. |
| 726 | * ring file:: |
| 727 | |
| 728 | The @code{rotate-yank-pointer} Function |
| 729 | |
| 730 | * Understanding rotate-yk-ptr:: |
| 731 | * rotate-yk-ptr body:: The body of @code{rotate-yank-pointer}. |
| 732 | |
| 733 | The Body of @code{rotate-yank-pointer} |
| 734 | |
| 735 | * Digression concerning error:: How to mislead humans, but not computers. |
| 736 | * rotate-yk-ptr else-part:: The else-part of the @code{if} expression. |
| 737 | * Remainder Function:: The remainder, @code{%}, function. |
| 738 | * rotate-yk-ptr remainder:: Using @code{%} in @code{rotate-yank-pointer}. |
| 739 | * kill-rng-yk-ptr last elt:: Pointing to the last element. |
| 740 | |
| 741 | @code{yank} |
| 742 | |
| 743 | * rotate-yk-ptr arg:: Pass the argument to @code{rotate-yank-pointer}. |
| 744 | * rotate-yk-ptr negative arg:: Pass a negative argument. |
| 745 | |
| 746 | A Graph with Labelled Axes |
| 747 | |
| 748 | * Labelled Example:: |
| 749 | * print-graph Varlist:: @code{let} expression in @code{print-graph}. |
| 750 | * print-Y-axis:: Print a label for the vertical axis. |
| 751 | * print-X-axis:: Print a horizontal label. |
| 752 | * Print Whole Graph:: The function to print a complete graph. |
| 753 | |
| 754 | The @code{print-Y-axis} Function |
| 755 | |
| 756 | * Height of label:: What height for the Y axis? |
| 757 | * Compute a Remainder:: How to compute the remainder of a division. |
| 758 | * Y Axis Element:: Construct a line for the Y axis. |
| 759 | * Y-axis-column:: Generate a list of Y axis labels. |
| 760 | * print-Y-axis Penultimate:: A not quite final version. |
| 761 | |
| 762 | The @code{print-X-axis} Function |
| 763 | |
| 764 | * Similarities differences:: Much like @code{print-Y-axis}, but not exactly. |
| 765 | * X Axis Tic Marks:: Create tic marks for the horizontal axis. |
| 766 | |
| 767 | Printing the Whole Graph |
| 768 | |
| 769 | * The final version:: A few changes. |
| 770 | * Test print-graph:: Run a short test. |
| 771 | * Graphing words in defuns:: Executing the final code. |
| 772 | * lambda:: How to write an anonymous function. |
| 773 | * mapcar:: Apply a function to elements of a list. |
| 774 | * Another Bug:: Yet another bug @dots{} most insidious. |
| 775 | * Final printed graph:: The graph itself! |
| 776 | |
| 777 | @end detailmenu |
| 778 | @end menu |
| 779 | |
| 780 | @c >>>> Set pageno appropriately <<<< |
| 781 | |
| 782 | @c The first page of the Preface is a roman numeral; it is the first |
| 783 | @c right handed page after the Table of Contents; hence the following |
| 784 | @c setting must be for an odd negative number. |
| 785 | |
| 786 | @iftex |
| 787 | @global@pageno = -11 |
| 788 | @end iftex |
| 789 | |
| 790 | @node Preface, List Processing, Top, Top |
| 791 | @comment node-name, next, previous, up |
| 792 | @unnumbered Preface |
| 793 | |
| 794 | Most of the GNU Emacs integrated environment is written in the programming |
| 795 | language called Emacs Lisp. The code written in this programming |
| 796 | language is the software---the sets of instructions---that tell the |
| 797 | computer what to do when you give it commands. Emacs is designed so |
| 798 | that you can write new code in Emacs Lisp and easily install it as an |
| 799 | extension to the editor. |
| 800 | |
| 801 | (GNU Emacs is sometimes called an ``extensible editor'', but it does |
| 802 | much more than provide editing capabilities. It is better to refer to |
| 803 | Emacs as an ``extensible computing environment''. However, that |
| 804 | phrase is quite a mouthful. It is easier to refer to Emacs simply as |
| 805 | an editor. Moreover, everything you do in Emacs---find the Mayan date |
| 806 | and phases of the moon, simplify polynomials, debug code, manage |
| 807 | files, read letters, write books---all these activities are kinds of |
| 808 | editing in the most general sense of the word.) |
| 809 | |
| 810 | @menu |
| 811 | * Why:: Why learn Emacs Lisp? |
| 812 | * On Reading this Text:: Read, gain familiarity, pick up habits.... |
| 813 | * Who You Are:: For whom this is written. |
| 814 | * Lisp History:: |
| 815 | * Note for Novices:: You can read this as a novice. |
| 816 | * Thank You:: |
| 817 | @end menu |
| 818 | |
| 819 | @node Why, On Reading this Text, Preface, Preface |
| 820 | @ifnottex |
| 821 | @unnumberedsec Why Study Emacs Lisp? |
| 822 | @end ifnottex |
| 823 | |
| 824 | Although Emacs Lisp is usually thought of in association only with Emacs, |
| 825 | it is a full computer programming language. You can use Emacs Lisp as |
| 826 | you would any other programming language. |
| 827 | |
| 828 | Perhaps you want to understand programming; perhaps you want to extend |
| 829 | Emacs; or perhaps you want to become a programmer. This introduction to |
| 830 | Emacs Lisp is designed to get you started: to guide you in learning the |
| 831 | fundamentals of programming, and more importantly, to show you how you |
| 832 | can teach yourself to go further. |
| 833 | |
| 834 | @node On Reading this Text, Who You Are, Why, Preface |
| 835 | @comment node-name, next, previous, up |
| 836 | @unnumberedsec On Reading this Text |
| 837 | |
| 838 | All through this document, you will see little sample programs you can |
| 839 | run inside of Emacs. If you read this document in Info inside of GNU |
| 840 | Emacs, you can run the programs as they appear. (This is easy to do and |
| 841 | is explained when the examples are presented.) Alternatively, you can |
| 842 | read this introduction as a printed book while sitting beside a computer |
| 843 | running Emacs. (This is what I like to do; I like printed books.) If |
| 844 | you don't have a running Emacs beside you, you can still read this book, |
| 845 | but in this case, it is best to treat it as a novel or as a travel guide |
| 846 | to a country not yet visited: interesting, but not the same as being |
| 847 | there. |
| 848 | |
| 849 | Much of this introduction is dedicated to walk-throughs or guided tours |
| 850 | of code used in GNU Emacs. These tours are designed for two purposes: |
| 851 | first, to give you familiarity with real, working code (code you use |
| 852 | every day); and, second, to give you familiarity with the way Emacs |
| 853 | works. It is interesting to see how a working environment is |
| 854 | implemented. |
| 855 | Also, I |
| 856 | hope that you will pick up the habit of browsing through source code. |
| 857 | You can learn from it and mine it for ideas. Having GNU Emacs is like |
| 858 | having a dragon's cave of treasures. |
| 859 | |
| 860 | In addition to learning about Emacs as an editor and Emacs Lisp as a |
| 861 | programming language, the examples and guided tours will give you an |
| 862 | opportunity to get acquainted with Emacs as a Lisp programming |
| 863 | environment. GNU Emacs supports programming and provides tools that |
| 864 | you will want to become comfortable using, such as @kbd{M-.} (the key |
| 865 | which invokes the @code{find-tag} command). You will also learn about |
| 866 | buffers and other objects that are part of the environment. |
| 867 | Learning about these features of Emacs is like learning new routes |
| 868 | around your home town. |
| 869 | |
| 870 | @ignore |
| 871 | In addition, I have written several programs as extended examples. |
| 872 | Although these are examples, the programs are real. I use them. |
| 873 | Other people use them. You may use them. Beyond the fragments of |
| 874 | programs used for illustrations, there is very little in here that is |
| 875 | `just for teaching purposes'; what you see is used. This is a great |
| 876 | advantage of Emacs Lisp: it is easy to learn to use it for work. |
| 877 | @end ignore |
| 878 | |
| 879 | Finally, I hope to convey some of the skills for using Emacs to |
| 880 | learn aspects of programming that you don't know. You can often use |
| 881 | Emacs to help you understand what puzzles you or to find out how to do |
| 882 | something new. This self-reliance is not only a pleasure, but an |
| 883 | advantage. |
| 884 | |
| 885 | @node Who You Are, Lisp History, On Reading this Text, Preface |
| 886 | @comment node-name, next, previous, up |
| 887 | @unnumberedsec For Whom This is Written |
| 888 | |
| 889 | This text is written as an elementary introduction for people who are |
| 890 | not programmers. If you are a programmer, you may not be satisfied with |
| 891 | this primer. The reason is that you may have become expert at reading |
| 892 | reference manuals and be put off by the way this text is organized. |
| 893 | |
| 894 | An expert programmer who reviewed this text said to me: |
| 895 | |
| 896 | @quotation |
| 897 | @i{I prefer to learn from reference manuals. I ``dive into'' each |
| 898 | paragraph, and ``come up for air'' between paragraphs.} |
| 899 | |
| 900 | @i{When I get to the end of a paragraph, I assume that that subject is |
| 901 | done, finished, that I know everything I need (with the |
| 902 | possible exception of the case when the next paragraph starts talking |
| 903 | about it in more detail). I expect that a well written reference manual |
| 904 | will not have a lot of redundancy, and that it will have excellent |
| 905 | pointers to the (one) place where the information I want is.} |
| 906 | @end quotation |
| 907 | |
| 908 | This introduction is not written for this person! |
| 909 | |
| 910 | Firstly, I try to say everything at least three times: first, to |
| 911 | introduce it; second, to show it in context; and third, to show it in a |
| 912 | different context, or to review it. |
| 913 | |
| 914 | Secondly, I hardly ever put all the information about a subject in one |
| 915 | place, much less in one paragraph. To my way of thinking, that imposes |
| 916 | too heavy a burden on the reader. Instead I try to explain only what |
| 917 | you need to know at the time. (Sometimes I include a little extra |
| 918 | information so you won't be surprised later when the additional |
| 919 | information is formally introduced.) |
| 920 | |
| 921 | When you read this text, you are not expected to learn everything the |
| 922 | first time. Frequently, you need only make, as it were, a `nodding |
| 923 | acquaintance' with some of the items mentioned. My hope is that I have |
| 924 | structured the text and given you enough hints that you will be alert to |
| 925 | what is important, and concentrate on it. |
| 926 | |
| 927 | You will need to ``dive into'' some paragraphs; there is no other way |
| 928 | to read them. But I have tried to keep down the number of such |
| 929 | paragraphs. This book is intended as an approachable hill, rather than |
| 930 | as a daunting mountain. |
| 931 | |
| 932 | This introduction to @cite{Programming in Emacs Lisp} has a companion |
| 933 | document, |
| 934 | @iftex |
| 935 | @cite{The GNU Emacs Lisp Reference Manual}. |
| 936 | @end iftex |
| 937 | @ifnottex |
| 938 | @ref{Top, , The GNU Emacs Lisp Reference Manual, elisp, The GNU |
| 939 | Emacs Lisp Reference Manual}. |
| 940 | @end ifnottex |
| 941 | The reference manual has more detail than this introduction. In the |
| 942 | reference manual, all the information about one topic is concentrated |
| 943 | in one place. You should turn to it if you are like the programmer |
| 944 | quoted above. And, of course, after you have read this |
| 945 | @cite{Introduction}, you will find the @cite{Reference Manual} useful |
| 946 | when you are writing your own programs. |
| 947 | |
| 948 | @node Lisp History, Note for Novices, Who You Are, Preface |
| 949 | @unnumberedsec Lisp History |
| 950 | @cindex Lisp history |
| 951 | |
| 952 | Lisp was first developed in the late 1950s at the Massachusetts |
| 953 | Institute of Technology for research in artificial intelligence. The |
| 954 | great power of the Lisp language makes it superior for other purposes as |
| 955 | well, such as writing editor commands and integrated environments. |
| 956 | |
| 957 | @cindex Maclisp |
| 958 | @cindex Common Lisp |
| 959 | GNU Emacs Lisp is largely inspired by Maclisp, which was written at MIT |
| 960 | in the 1960s. It is somewhat inspired by Common Lisp, which became a |
| 961 | standard in the 1980s. However, Emacs Lisp is much simpler than Common |
| 962 | Lisp. (The standard Emacs distribution contains an optional extensions |
| 963 | file, @file{cl.el}, that adds many Common Lisp features to Emacs Lisp.) |
| 964 | |
| 965 | @node Note for Novices, Thank You, Lisp History, Preface |
| 966 | @comment node-name, next, previous, up |
| 967 | @unnumberedsec A Note for Novices |
| 968 | |
| 969 | If you don't know GNU Emacs, you can still read this document |
| 970 | profitably. However, I recommend you learn Emacs, if only to learn to |
| 971 | move around your computer screen. You can teach yourself how to use |
| 972 | Emacs with the on-line tutorial. To use it, type @kbd{C-h t}. (This |
| 973 | means you press and release the @key{CTRL} key and the @kbd{h} at the |
| 974 | same time, and then press and release @kbd{t}.) |
| 975 | |
| 976 | Also, I often refer to one of Emacs' standard commands by listing the |
| 977 | keys which you press to invoke the command and then giving the name of |
| 978 | the command in parentheses, like this: @kbd{M-C-\} |
| 979 | (@code{indent-region}). What this means is that the |
| 980 | @code{indent-region} command is customarily invoked by typing |
| 981 | @kbd{M-C-\}. (You can, if you wish, change the keys that are typed to |
| 982 | invoke the command; this is called @dfn{rebinding}. @xref{Keymaps, , |
| 983 | Keymaps}.) The abbreviation @kbd{M-C-\} means that you type your |
| 984 | @key{META} key, @key{CTRL} key and @key{\} key all at the same time. |
| 985 | (On many modern keyboards the @key{META} key is labelled |
| 986 | @key{ALT}.) |
| 987 | Sometimes a combination like this is called a keychord, since it is |
| 988 | similar to the way you play a chord on a piano. If your keyboard does |
| 989 | not have a @key{META} key, the @key{ESC} key prefix is used in place |
| 990 | of it. In this case, @kbd{M-C-\} means that you press and release your |
| 991 | @key{ESC} key and then type the @key{CTRL} key and the @key{\} key at |
| 992 | the same time. But usually @kbd{M-C-\} means press the @key{CTRL} key |
| 993 | along with the key that is labelled @key{ALT} and, at the same time, |
| 994 | press the @key{\} key. |
| 995 | |
| 996 | In addition to typing a lone keychord, you can prefix what you type |
| 997 | with @kbd{C-u}, which is called the `universal argument'. The |
| 998 | @kbd{C-u} keychord passes an argument to the subsequent command. |
| 999 | Thus, to indent a region of plain text by 6 spaces, mark the region, |
| 1000 | and then type @w{@kbd{C-u 6 M-C-\}}. (If you do not specify a number, |
| 1001 | Emacs either passes the number 4 to the command or otherwise runs the |
| 1002 | command differently than it would otherwise.) @xref{Arguments, , |
| 1003 | Numeric Arguments, emacs, The GNU Emacs Manual}. |
| 1004 | |
| 1005 | If you are reading this in Info using GNU Emacs, you can read through |
| 1006 | this whole document just by pressing the space bar, @key{SPC}. |
| 1007 | (To learn about Info, type @kbd{C-h i} and then select Info.) |
| 1008 | |
| 1009 | A note on terminology: when I use the word Lisp alone, I often am |
| 1010 | referring to the various dialects of Lisp in general, but when I speak |
| 1011 | of Emacs Lisp, I am referring to GNU Emacs Lisp in particular. |
| 1012 | |
| 1013 | @node Thank You, , Note for Novices, Preface |
| 1014 | @comment node-name, next, previous, up |
| 1015 | @unnumberedsec Thank You |
| 1016 | |
| 1017 | My thanks to all who helped me with this book. My especial thanks to |
| 1018 | @r{Jim Blandy}, @r{Noah Friedman}, @w{Jim Kingdon}, @r{Roland |
| 1019 | McGrath}, @w{Frank Ritter}, @w{Randy Smith}, @w{Richard M.@: |
| 1020 | Stallman}, and @w{Melissa Weisshaus}. My thanks also go to both |
| 1021 | @w{Philip Johnson} and @w{David Stampe} for their patient |
| 1022 | encouragement. My mistakes are my own. |
| 1023 | |
| 1024 | @flushright |
| 1025 | Robert J. Chassell |
| 1026 | @end flushright |
| 1027 | |
| 1028 | @c ================ Beginning of main text ================ |
| 1029 | |
| 1030 | @c Start main text on right-hand (verso) page |
| 1031 | |
| 1032 | @tex |
| 1033 | \par\vfill\supereject |
| 1034 | \headings off |
| 1035 | \ifodd\pageno |
| 1036 | \par\vfill\supereject |
| 1037 | \else |
| 1038 | \par\vfill\supereject |
| 1039 | \page\hbox{}\page |
| 1040 | \par\vfill\supereject |
| 1041 | \fi |
| 1042 | @end tex |
| 1043 | |
| 1044 | @iftex |
| 1045 | @headings off |
| 1046 | @evenheading @thispage @| @| @thischapter |
| 1047 | @oddheading @thissection @| @| @thispage |
| 1048 | @global@pageno = 1 |
| 1049 | @end iftex |
| 1050 | |
| 1051 | @node List Processing, Practicing Evaluation, Preface, Top |
| 1052 | @comment node-name, next, previous, up |
| 1053 | @chapter List Processing |
| 1054 | |
| 1055 | To the untutored eye, Lisp is a strange programming language. In Lisp |
| 1056 | code there are parentheses everywhere. Some people even claim that |
| 1057 | the name stands for `Lots of Isolated Silly Parentheses'. But the |
| 1058 | claim is unwarranted. Lisp stands for LISt Processing, and the |
| 1059 | programming language handles @emph{lists} (and lists of lists) by |
| 1060 | putting them between parentheses. The parentheses mark the boundaries |
| 1061 | of the list. Sometimes a list is preceded by a single apostrophe or |
| 1062 | quotation mark, @samp{'}@footnote{The single apostrophe or quotation |
| 1063 | mark is an abbreviation for the function @code{quote}; you need not |
| 1064 | think about functions now; functions are defined in @ref{Making |
| 1065 | Errors, , Generate an Error Message}.} Lists are the basis of Lisp. |
| 1066 | |
| 1067 | @menu |
| 1068 | * Lisp Lists:: What are lists? |
| 1069 | * Run a Program:: Any list in Lisp is a program ready to run. |
| 1070 | * Making Errors:: Generating an error message. |
| 1071 | * Names & Definitions:: Names of symbols and function definitions. |
| 1072 | * Lisp Interpreter:: What the Lisp interpreter does. |
| 1073 | * Evaluation:: Running a program. |
| 1074 | * Variables:: Returning a value from a variable. |
| 1075 | * Arguments:: Passing information to a function. |
| 1076 | * set & setq:: Setting the value of a variable. |
| 1077 | * Summary:: The major points. |
| 1078 | * Error Message Exercises:: |
| 1079 | @end menu |
| 1080 | |
| 1081 | @node Lisp Lists, Run a Program, List Processing, List Processing |
| 1082 | @comment node-name, next, previous, up |
| 1083 | @section Lisp Lists |
| 1084 | @cindex Lisp Lists |
| 1085 | |
| 1086 | In Lisp, a list looks like this: @code{'(rose violet daisy buttercup)}. |
| 1087 | This list is preceded by a single apostrophe. It could just as well be |
| 1088 | written as follows, which looks more like the kind of list you are likely |
| 1089 | to be familiar with: |
| 1090 | |
| 1091 | @smallexample |
| 1092 | @group |
| 1093 | '(rose |
| 1094 | violet |
| 1095 | daisy |
| 1096 | buttercup) |
| 1097 | @end group |
| 1098 | @end smallexample |
| 1099 | |
| 1100 | @noindent |
| 1101 | The elements of this list are the names of the four different flowers, |
| 1102 | separated from each other by whitespace and surrounded by parentheses, |
| 1103 | like flowers in a field with a stone wall around them. |
| 1104 | @cindex Flowers in a field |
| 1105 | |
| 1106 | @menu |
| 1107 | * Numbers Lists:: List have numbers, other lists, in them. |
| 1108 | * Lisp Atoms:: Elemental entities. |
| 1109 | * Whitespace in Lists:: Formating lists to be readable. |
| 1110 | * Typing Lists:: How GNU Emacs helps you type lists. |
| 1111 | @end menu |
| 1112 | |
| 1113 | @node Numbers Lists, Lisp Atoms, Lisp Lists, Lisp Lists |
| 1114 | @ifnottex |
| 1115 | @unnumberedsubsec Numbers, Lists inside of Lists |
| 1116 | @end ifnottex |
| 1117 | |
| 1118 | Lists can also have numbers in them, as in this list: @code{(+ 2 2)}. |
| 1119 | This list has a plus-sign, @samp{+}, followed by two @samp{2}s, each |
| 1120 | separated by whitespace. |
| 1121 | |
| 1122 | In Lisp, both data and programs are represented the same way; that is, |
| 1123 | they are both lists of words, numbers, or other lists, separated by |
| 1124 | whitespace and surrounded by parentheses. (Since a program looks like |
| 1125 | data, one program may easily serve as data for another; this is a very |
| 1126 | powerful feature of Lisp.) (Incidentally, these two parenthetical |
| 1127 | remarks are @emph{not} Lisp lists, because they contain @samp{;} and |
| 1128 | @samp{.} as punctuation marks.) |
| 1129 | |
| 1130 | @need 1200 |
| 1131 | Here is another list, this time with a list inside of it: |
| 1132 | |
| 1133 | @smallexample |
| 1134 | '(this list has (a list inside of it)) |
| 1135 | @end smallexample |
| 1136 | |
| 1137 | The components of this list are the words @samp{this}, @samp{list}, |
| 1138 | @samp{has}, and the list @samp{(a list inside of it)}. The interior |
| 1139 | list is made up of the words @samp{a}, @samp{list}, @samp{inside}, |
| 1140 | @samp{of}, @samp{it}. |
| 1141 | |
| 1142 | @node Lisp Atoms, Whitespace in Lists, Numbers Lists, Lisp Lists |
| 1143 | @comment node-name, next, previous, up |
| 1144 | @subsection Lisp Atoms |
| 1145 | @cindex Lisp Atoms |
| 1146 | |
| 1147 | In Lisp, what we have been calling words are called @dfn{atoms}. This |
| 1148 | term comes from the historical meaning of the word atom, which means |
| 1149 | `indivisible'. As far as Lisp is concerned, the words we have been |
| 1150 | using in the lists cannot be divided into any smaller parts and still |
| 1151 | mean the same thing as part of a program; likewise with numbers and |
| 1152 | single character symbols like @samp{+}. On the other hand, unlike an |
| 1153 | atom, a list can be split into parts. (@xref{car cdr & cons, , |
| 1154 | @code{car} @code{cdr} & @code{cons} Fundamental Functions}.) |
| 1155 | |
| 1156 | In a list, atoms are separated from each other by whitespace. They can be |
| 1157 | right next to a parenthesis. |
| 1158 | |
| 1159 | @cindex @samp{empty list} defined |
| 1160 | Technically speaking, a list in Lisp consists of parentheses surrounding |
| 1161 | atoms separated by whitespace or surrounding other lists or surrounding |
| 1162 | both atoms and other lists. A list can have just one atom in it or |
| 1163 | have nothing in it at all. A list with nothing in it looks like this: |
| 1164 | @code{()}, and is called the @dfn{empty list}. Unlike anything else, an |
| 1165 | empty list is considered both an atom and a list at the same time. |
| 1166 | |
| 1167 | @cindex Symbolic expressions, introduced |
| 1168 | @cindex @samp{expression} defined |
| 1169 | @cindex @samp{form} defined |
| 1170 | The printed representation of both atoms and lists are called |
| 1171 | @dfn{symbolic expressions} or, more concisely, @dfn{s-expressions}. |
| 1172 | The word @dfn{expression} by itself can refer to either the printed |
| 1173 | representation, or to the atom or list as it is held internally in the |
| 1174 | computer. Often, people use the term @dfn{expression} |
| 1175 | indiscriminately. (Also, in many texts, the word @dfn{form} is used |
| 1176 | as a synonym for expression.) |
| 1177 | |
| 1178 | Incidentally, the atoms that make up our universe were named such when |
| 1179 | they were thought to be indivisible; but it has been found that physical |
| 1180 | atoms are not indivisible. Parts can split off an atom or it can |
| 1181 | fission into two parts of roughly equal size. Physical atoms were named |
| 1182 | prematurely, before their truer nature was found. In Lisp, certain |
| 1183 | kinds of atom, such as an array, can be separated into parts; but the |
| 1184 | mechanism for doing this is different from the mechanism for splitting a |
| 1185 | list. As far as list operations are concerned, the atoms of a list are |
| 1186 | unsplittable. |
| 1187 | |
| 1188 | As in English, the meanings of the component letters of a Lisp atom |
| 1189 | are different from the meaning the letters make as a word. For |
| 1190 | example, the word for the South American sloth, the @samp{ai}, is |
| 1191 | completely different from the two words, @samp{a}, and @samp{i}. |
| 1192 | |
| 1193 | There are many kinds of atom in nature but only a few in Lisp: for |
| 1194 | example, @dfn{numbers}, such as 37, 511, or 1729, and @dfn{symbols}, such |
| 1195 | as @samp{+}, @samp{foo}, or @samp{forward-line}. The words we have |
| 1196 | listed in the examples above are all symbols. In everyday Lisp |
| 1197 | conversation, the word ``atom'' is not often used, because programmers |
| 1198 | usually try to be more specific about what kind of atom they are dealing |
| 1199 | with. Lisp programming is mostly about symbols (and sometimes numbers) |
| 1200 | within lists. (Incidentally, the preceding three word parenthetical |
| 1201 | remark is a proper list in Lisp, since it consists of atoms, which in |
| 1202 | this case are symbols, separated by whitespace and enclosed by |
| 1203 | parentheses, without any non-Lisp punctuation.) |
| 1204 | |
| 1205 | @need 1250 |
| 1206 | In addition, text between double quotation marks---even sentences or |
| 1207 | paragraphs---is an atom. Here is an example: |
| 1208 | @cindex Text between double quotation marks |
| 1209 | |
| 1210 | @smallexample |
| 1211 | '(this list includes "text between quotation marks.") |
| 1212 | @end smallexample |
| 1213 | |
| 1214 | @cindex @samp{string} defined |
| 1215 | @noindent |
| 1216 | In Lisp, all of the quoted text including the punctuation mark and the |
| 1217 | blank spaces is a single atom. This kind of atom is called a |
| 1218 | @dfn{string} (for `string of characters') and is the sort of thing that |
| 1219 | is used for messages that a computer can print for a human to read. |
| 1220 | Strings are a different kind of atom than numbers or symbols and are |
| 1221 | used differently. |
| 1222 | |
| 1223 | @node Whitespace in Lists, Typing Lists, Lisp Atoms, Lisp Lists |
| 1224 | @comment node-name, next, previous, up |
| 1225 | @subsection Whitespace in Lists |
| 1226 | @cindex Whitespace in lists |
| 1227 | |
| 1228 | @need 1200 |
| 1229 | The amount of whitespace in a list does not matter. From the point of view |
| 1230 | of the Lisp language, |
| 1231 | |
| 1232 | @smallexample |
| 1233 | @group |
| 1234 | '(this list |
| 1235 | looks like this) |
| 1236 | @end group |
| 1237 | @end smallexample |
| 1238 | |
| 1239 | @need 800 |
| 1240 | @noindent |
| 1241 | is exactly the same as this: |
| 1242 | |
| 1243 | @smallexample |
| 1244 | '(this list looks like this) |
| 1245 | @end smallexample |
| 1246 | |
| 1247 | Both examples show what to Lisp is the same list, the list made up of |
| 1248 | the symbols @samp{this}, @samp{list}, @samp{looks}, @samp{like}, and |
| 1249 | @samp{this} in that order. |
| 1250 | |
| 1251 | Extra whitespace and newlines are designed to make a list more readable |
| 1252 | by humans. When Lisp reads the expression, it gets rid of all the extra |
| 1253 | whitespace (but it needs to have at least one space between atoms in |
| 1254 | order to tell them apart.) |
| 1255 | |
| 1256 | Odd as it seems, the examples we have seen cover almost all of what Lisp |
| 1257 | lists look like! Every other list in Lisp looks more or less like one |
| 1258 | of these examples, except that the list may be longer and more complex. |
| 1259 | In brief, a list is between parentheses, a string is between quotation |
| 1260 | marks, a symbol looks like a word, and a number looks like a number. |
| 1261 | (For certain situations, square brackets, dots and a few other special |
| 1262 | characters may be used; however, we will go quite far without them.) |
| 1263 | |
| 1264 | @node Typing Lists, , Whitespace in Lists, Lisp Lists |
| 1265 | @comment node-name, next, previous, up |
| 1266 | @subsection GNU Emacs Helps You Type Lists |
| 1267 | @cindex Help typing lists |
| 1268 | @cindex Formatting help |
| 1269 | |
| 1270 | When you type a Lisp expression in GNU Emacs using either Lisp |
| 1271 | Interaction mode or Emacs Lisp mode, you have available to you several |
| 1272 | commands to format the Lisp expression so it is easy to read. For |
| 1273 | example, pressing the @key{TAB} key automatically indents the line the |
| 1274 | cursor is on by the right amount. A command to properly indent the |
| 1275 | code in a region is customarily bound to @kbd{M-C-\}. Indentation is |
| 1276 | designed so that you can see which elements of a list belong to which |
| 1277 | list---elements of a sub-list are indented more than the elements of |
| 1278 | the enclosing list. |
| 1279 | |
| 1280 | In addition, when you type a closing parenthesis, Emacs momentarily |
| 1281 | jumps the cursor back to the matching opening parenthesis, so you can |
| 1282 | see which one it is. This is very useful, since every list you type |
| 1283 | in Lisp must have its closing parenthesis match its opening |
| 1284 | parenthesis. (@xref{Major Modes, , Major Modes, emacs, The GNU Emacs |
| 1285 | Manual}, for more information about Emacs' modes.) |
| 1286 | |
| 1287 | @node Run a Program, Making Errors, Lisp Lists, List Processing |
| 1288 | @comment node-name, next, previous, up |
| 1289 | @section Run a Program |
| 1290 | @cindex Run a program |
| 1291 | @cindex Program, running one |
| 1292 | |
| 1293 | @cindex @samp{evaluate} defined |
| 1294 | A list in Lisp---any list---is a program ready to run. If you run it |
| 1295 | (for which the Lisp jargon is @dfn{evaluate}), the computer will do one |
| 1296 | of three things: do nothing except return to you the list itself; send |
| 1297 | you an error message; or, treat the first symbol in the list as a |
| 1298 | command to do something. (Usually, of course, it is the last of these |
| 1299 | three things that you really want!) |
| 1300 | |
| 1301 | @c use code for the single apostrophe, not samp. |
| 1302 | The single apostrophe, @code{'}, that I put in front of some of the |
| 1303 | example lists in preceding sections is called a @dfn{quote}; when it |
| 1304 | precedes a list, it tells Lisp to do nothing with the list, other than |
| 1305 | take it as it is written. But if there is no quote preceding a list, |
| 1306 | the first item of the list is special: it is a command for the computer |
| 1307 | to obey. (In Lisp, these commands are called @emph{functions}.) The list |
| 1308 | @code{(+ 2 2)} shown above did not have a quote in front of it, so Lisp |
| 1309 | understands that the @code{+} is an instruction to do something with the |
| 1310 | rest of the list: add the numbers that follow. |
| 1311 | |
| 1312 | @need 1250 |
| 1313 | If you are reading this inside of GNU Emacs in Info, here is how you can |
| 1314 | evaluate such a list: place your cursor immediately after the right |
| 1315 | hand parenthesis of the following list and then type @kbd{C-x C-e}: |
| 1316 | |
| 1317 | @smallexample |
| 1318 | (+ 2 2) |
| 1319 | @end smallexample |
| 1320 | |
| 1321 | @c use code for the number four, not samp. |
| 1322 | @noindent |
| 1323 | You will see the number @code{4} appear in the echo area. (In the |
| 1324 | jargon, what you have just done is ``evaluate the list.'' The echo area |
| 1325 | is the line at the bottom of the screen that displays or ``echoes'' |
| 1326 | text.) Now try the same thing with a quoted list: place the cursor |
| 1327 | right after the following list and type @kbd{C-x C-e}: |
| 1328 | |
| 1329 | @smallexample |
| 1330 | '(this is a quoted list) |
| 1331 | @end smallexample |
| 1332 | |
| 1333 | @noindent |
| 1334 | You will see @code{(this is a quoted list)} appear in the echo area. |
| 1335 | |
| 1336 | @cindex Lisp interpreter, explained |
| 1337 | @cindex Interpreter, Lisp, explained |
| 1338 | In both cases, what you are doing is giving a command to the program |
| 1339 | inside of GNU Emacs called the @dfn{Lisp interpreter}---giving the |
| 1340 | interpreter a command to evaluate the expression. The name of the Lisp |
| 1341 | interpreter comes from the word for the task done by a human who comes |
| 1342 | up with the meaning of an expression---who ``interprets'' it. |
| 1343 | |
| 1344 | You can also evaluate an atom that is not part of a list---one that is |
| 1345 | not surrounded by parentheses; again, the Lisp interpreter translates |
| 1346 | from the humanly readable expression to the language of the computer. |
| 1347 | But before discussing this (@pxref{Variables}), we will discuss what the |
| 1348 | Lisp interpreter does when you make an error. |
| 1349 | |
| 1350 | @node Making Errors, Names & Definitions, Run a Program, List Processing |
| 1351 | @comment node-name, next, previous, up |
| 1352 | @section Generate an Error Message |
| 1353 | @cindex Generate an error message |
| 1354 | @cindex Error message generation |
| 1355 | |
| 1356 | Partly so you won't worry if you do it accidentally, we will now give |
| 1357 | a command to the Lisp interpreter that generates an error message. |
| 1358 | This is a harmless activity; and indeed, we will often try to generate |
| 1359 | error messages intentionally. Once you understand the jargon, error |
| 1360 | messages can be informative. Instead of being called ``error'' |
| 1361 | messages, they should be called ``help'' messages. They are like |
| 1362 | signposts to a traveller in a strange country; deciphering them can be |
| 1363 | hard, but once understood, they can point the way. |
| 1364 | |
| 1365 | The error message is generated by a built-in GNU Emacs debugger. We |
| 1366 | will `enter the debugger'. You get out of the debugger by typing @code{q}. |
| 1367 | |
| 1368 | What we will do is evaluate a list that is not quoted and does not |
| 1369 | have a meaningful command as its first element. Here is a list almost |
| 1370 | exactly the same as the one we just used, but without the single-quote |
| 1371 | in front of it. Position the cursor right after it and type @kbd{C-x |
| 1372 | C-e}: |
| 1373 | |
| 1374 | @smallexample |
| 1375 | (this is an unquoted list) |
| 1376 | @end smallexample |
| 1377 | |
| 1378 | @noindent |
| 1379 | What you see depends on which version of Emacs you are running. GNU |
| 1380 | Emacs version 21 provides more information than version 20 and before. |
| 1381 | First, the more recent result of generating an error; then the |
| 1382 | earlier, version 20 result. |
| 1383 | |
| 1384 | @need 1250 |
| 1385 | @noindent |
| 1386 | In GNU Emacs version 21, a @file{*Backtrace*} window will open up and |
| 1387 | you will see the following in it: |
| 1388 | |
| 1389 | @smallexample |
| 1390 | @group |
| 1391 | ---------- Buffer: *Backtrace* ---------- |
| 1392 | Debugger entered--Lisp error: (void-function this) |
| 1393 | (this is an unquoted list) |
| 1394 | eval((this is an unquoted list)) |
| 1395 | eval-last-sexp-1(nil) |
| 1396 | eval-last-sexp(nil) |
| 1397 | call-interactively(eval-last-sexp) |
| 1398 | ---------- Buffer: *Backtrace* ---------- |
| 1399 | @end group |
| 1400 | @end smallexample |
| 1401 | |
| 1402 | @need 1200 |
| 1403 | @noindent |
| 1404 | Your cursor will be in this window (you may have to wait a few seconds |
| 1405 | before it becomes visible). To quit the debugger and make the |
| 1406 | debugger window go away, type: |
| 1407 | |
| 1408 | @smallexample |
| 1409 | q |
| 1410 | @end smallexample |
| 1411 | |
| 1412 | @noindent |
| 1413 | Please type @kbd{q} right now, so you become confident that you can |
| 1414 | get out of the debugger. Then, type @kbd{C-x C-e} again to re-enter |
| 1415 | it. |
| 1416 | |
| 1417 | @cindex @samp{function} defined |
| 1418 | Based on what we already know, we can almost read this error message. |
| 1419 | |
| 1420 | You read the @file{*Backtrace*} buffer from the bottom up; it tells |
| 1421 | you what Emacs did. When you typed @kbd{C-x C-e}, you made an |
| 1422 | interactive call to the command @code{eval-last-sexp}. @code{eval} is |
| 1423 | an abbreviation for `evaluate' and @code{sexp} is an abbreviation for |
| 1424 | `symbolic expression'. The command means `evaluate last symbolic |
| 1425 | expression', which is the expression just before your cursor. |
| 1426 | |
| 1427 | Each line above tells you what the Lisp interpreter evaluated next. |
| 1428 | The most recent action is at the top. The buffer is called the |
| 1429 | @file{*Backtrace*} buffer because it enables you to track Emacs |
| 1430 | backwards. |
| 1431 | |
| 1432 | @need 800 |
| 1433 | At the top of the @file{*Backtrace*} buffer, you see the line: |
| 1434 | |
| 1435 | @smallexample |
| 1436 | Debugger entered--Lisp error: (void-function this) |
| 1437 | @end smallexample |
| 1438 | |
| 1439 | @noindent |
| 1440 | The Lisp interpreter tried to evaluate the first atom of the list, the |
| 1441 | word @samp{this}. It is this action that generated the error message |
| 1442 | @samp{void-function this}. |
| 1443 | |
| 1444 | The message contains the words @samp{void-function} and @samp{this}. |
| 1445 | |
| 1446 | @cindex @samp{function} defined |
| 1447 | The word @samp{function} was mentioned once before. It is a very |
| 1448 | important word. For our purposes, we can define it by saying that a |
| 1449 | @dfn{function} is a set of instructions to the computer that tell the |
| 1450 | computer to do something. |
| 1451 | |
| 1452 | Now we can begin to understand the error message: @samp{void-function |
| 1453 | this}. The function (that is, the word @samp{this}) does not have a |
| 1454 | definition of any set of instructions for the computer to carry out. |
| 1455 | |
| 1456 | The slightly odd word, @samp{void-function}, is designed to cover the |
| 1457 | way Emacs Lisp is implemented, which is that when a symbol does not |
| 1458 | have a function definition attached to it, the place that should |
| 1459 | contain the instructions is `void'. |
| 1460 | |
| 1461 | On the other hand, since we were able to add 2 plus 2 successfully, by |
| 1462 | evaluating @code{(+ 2 2)}, we can infer that the symbol @code{+} must |
| 1463 | have a set of instructions for the computer to obey and those |
| 1464 | instructions must be to add the numbers that follow the @code{+}. |
| 1465 | |
| 1466 | @need 1250 |
| 1467 | In GNU Emacs version 20, and in earlier versions, you will see only |
| 1468 | one line of error message; it will appear in the echo area and look |
| 1469 | like this: |
| 1470 | |
| 1471 | @smallexample |
| 1472 | Symbol's function definition is void:@: this |
| 1473 | @end smallexample |
| 1474 | |
| 1475 | @noindent |
| 1476 | (Also, your terminal may beep at you---some do, some don't; and others |
| 1477 | blink. This is just a device to get your attention.) The message goes |
| 1478 | away as soon as you type another key, even just to move the cursor. |
| 1479 | |
| 1480 | We know the meaning of the word @samp{Symbol}. It refers to the first |
| 1481 | atom of the list, the word @samp{this}. The word @samp{function} |
| 1482 | refers to the instructions that tell the computer what to do. |
| 1483 | (Technically, the symbol tells the computer where to find the |
| 1484 | instructions, but this is a complication we can ignore for the |
| 1485 | moment.) |
| 1486 | |
| 1487 | The error message can be understood: @samp{Symbol's function |
| 1488 | definition is void:@: this}. The symbol (that is, the word |
| 1489 | @samp{this}) lacks instructions for the computer to carry out. |
| 1490 | |
| 1491 | @node Names & Definitions, Lisp Interpreter, Making Errors, List Processing |
| 1492 | @comment node-name, next, previous, up |
| 1493 | @section Symbol Names and Function Definitions |
| 1494 | @cindex Symbol names |
| 1495 | |
| 1496 | We can articulate another characteristic of Lisp based on what we have |
| 1497 | discussed so far---an important characteristic: a symbol, like |
| 1498 | @code{+}, is not itself the set of instructions for the computer to |
| 1499 | carry out. Instead, the symbol is used, perhaps temporarily, as a way |
| 1500 | of locating the definition or set of instructions. What we see is the |
| 1501 | name through which the instructions can be found. Names of people |
| 1502 | work the same way. I can be referred to as @samp{Bob}; however, I am |
| 1503 | not the letters @samp{B}, @samp{o}, @samp{b} but am, or were, the |
| 1504 | consciousness consistently associated with a particular life-form. |
| 1505 | The name is not me, but it can be used to refer to me. |
| 1506 | |
| 1507 | In Lisp, one set of instructions can be attached to several names. |
| 1508 | For example, the computer instructions for adding numbers can be |
| 1509 | linked to the symbol @code{plus} as well as to the symbol @code{+} |
| 1510 | (and are in some dialects of Lisp). Among humans, I can be referred |
| 1511 | to as @samp{Robert} as well as @samp{Bob} and by other words as well. |
| 1512 | |
| 1513 | On the other hand, a symbol can have only one function definition |
| 1514 | attached to it at a time. Otherwise, the computer would be confused as |
| 1515 | to which definition to use. If this were the case among people, only |
| 1516 | one person in the world could be named @samp{Bob}. However, the function |
| 1517 | definition to which the name refers can be changed readily. |
| 1518 | (@xref{Install, , Install a Function Definition}.) |
| 1519 | |
| 1520 | Since Emacs Lisp is large, it is customary to name symbols in a way |
| 1521 | that identifies the part of Emacs to which the function belongs. |
| 1522 | Thus, all the names for functions that deal with Texinfo start with |
| 1523 | @samp{texinfo-} and those for functions that deal with reading mail |
| 1524 | start with @samp{rmail-}. |
| 1525 | |
| 1526 | @node Lisp Interpreter, Evaluation, Names & Definitions, List Processing |
| 1527 | @comment node-name, next, previous, up |
| 1528 | @section The Lisp Interpreter |
| 1529 | @cindex Lisp interpreter, what it does |
| 1530 | @cindex Interpreter, what it does |
| 1531 | |
| 1532 | Based on what we have seen, we can now start to figure out what the |
| 1533 | Lisp interpreter does when we command it to evaluate a list. |
| 1534 | First, it looks to see whether there is a quote before the list; if |
| 1535 | there is, the interpreter just gives us the list. On the other |
| 1536 | hand, if there is no quote, the interpreter looks at the first element |
| 1537 | in the list and sees whether it has a function definition. If it does, |
| 1538 | the interpreter carries out the instructions in the function definition. |
| 1539 | Otherwise, the interpreter prints an error message. |
| 1540 | |
| 1541 | This is how Lisp works. Simple. There are added complications which we |
| 1542 | will get to in a minute, but these are the fundamentals. Of course, to |
| 1543 | write Lisp programs, you need to know how to write function definitions |
| 1544 | and attach them to names, and how to do this without confusing either |
| 1545 | yourself or the computer. |
| 1546 | |
| 1547 | @menu |
| 1548 | * Complications:: Variables, Special forms, Lists within. |
| 1549 | * Byte Compiling:: Specially processing code for speed. |
| 1550 | @end menu |
| 1551 | |
| 1552 | @node Complications, Byte Compiling, Lisp Interpreter, Lisp Interpreter |
| 1553 | @ifnottex |
| 1554 | @unnumberedsubsec Complications |
| 1555 | @end ifnottex |
| 1556 | |
| 1557 | Now, for the first complication. In addition to lists, the Lisp |
| 1558 | interpreter can evaluate a symbol that is not quoted and does not have |
| 1559 | parentheses around it. The Lisp interpreter will attempt to determine |
| 1560 | the symbol's value as a @dfn{variable}. This situation is described |
| 1561 | in the section on variables. (@xref{Variables}.) |
| 1562 | |
| 1563 | @cindex Special form |
| 1564 | The second complication occurs because some functions are unusual and do |
| 1565 | not work in the usual manner. Those that don't are called @dfn{special |
| 1566 | forms}. They are used for special jobs, like defining a function, and |
| 1567 | there are not many of them. In the next few chapters, you will be |
| 1568 | introduced to several of the more important special forms. |
| 1569 | |
| 1570 | The third and final complication is this: if the function that the |
| 1571 | Lisp interpreter is looking at is not a special form, and if it is part |
| 1572 | of a list, the Lisp interpreter looks to see whether the list has a list |
| 1573 | inside of it. If there is an inner list, the Lisp interpreter first |
| 1574 | figures out what it should do with the inside list, and then it works on |
| 1575 | the outside list. If there is yet another list embedded inside the |
| 1576 | inner list, it works on that one first, and so on. It always works on |
| 1577 | the innermost list first. The interpreter works on the innermost list |
| 1578 | first, to evaluate the result of that list. The result may be |
| 1579 | used by the enclosing expression. |
| 1580 | |
| 1581 | Otherwise, the interpreter works left to right, from one expression to |
| 1582 | the next. |
| 1583 | |
| 1584 | @node Byte Compiling, , Complications, Lisp Interpreter |
| 1585 | @subsection Byte Compiling |
| 1586 | @cindex Byte compiling |
| 1587 | |
| 1588 | One other aspect of interpreting: the Lisp interpreter is able to |
| 1589 | interpret two kinds of entity: humanly readable code, on which we will |
| 1590 | focus exclusively, and specially processed code, called @dfn{byte |
| 1591 | compiled} code, which is not humanly readable. Byte compiled code |
| 1592 | runs faster than humanly readable code. |
| 1593 | |
| 1594 | You can transform humanly readable code into byte compiled code by |
| 1595 | running one of the compile commands such as @code{byte-compile-file}. |
| 1596 | Byte compiled code is usually stored in a file that ends with a |
| 1597 | @file{.elc} extension rather than a @file{.el} extension. You will |
| 1598 | see both kinds of file in the @file{emacs/lisp} directory; the files |
| 1599 | to read are those with @file{.el} extensions. |
| 1600 | |
| 1601 | As a practical matter, for most things you might do to customize or |
| 1602 | extend Emacs, you do not need to byte compile; and I will not discuss |
| 1603 | the topic here. @xref{Byte Compilation, , Byte Compilation, elisp, |
| 1604 | The GNU Emacs Lisp Reference Manual}, for a full description of byte |
| 1605 | compilation. |
| 1606 | |
| 1607 | @node Evaluation, Variables, Lisp Interpreter, List Processing |
| 1608 | @comment node-name, next, previous, up |
| 1609 | @section Evaluation |
| 1610 | @cindex Evaluation |
| 1611 | |
| 1612 | When the Lisp interpreter works on an expression, the term for the |
| 1613 | activity is called @dfn{evaluation}. We say that the interpreter |
| 1614 | `evaluates the expression'. I've used this term several times before. |
| 1615 | The word comes from its use in everyday language, `to ascertain the |
| 1616 | value or amount of; to appraise', according to @cite{Webster's New |
| 1617 | Collegiate Dictionary}. |
| 1618 | |
| 1619 | After evaluating an expression, the Lisp interpreter will most likely |
| 1620 | @dfn{return} the value that the computer produces by carrying out the |
| 1621 | instructions it found in the function definition, or perhaps it will |
| 1622 | give up on that function and produce an error message. (The interpreter |
| 1623 | may also find itself tossed, so to speak, to a different function or it |
| 1624 | may attempt to repeat continually what it is doing for ever and ever in |
| 1625 | what is called an `infinite loop'. These actions are less common; and |
| 1626 | we can ignore them.) Most frequently, the interpreter returns a value. |
| 1627 | |
| 1628 | @cindex @samp{side effect} defined |
| 1629 | At the same time the interpreter returns a value, it may do something |
| 1630 | else as well, such as move a cursor or copy a file; this other kind of |
| 1631 | action is called a @dfn{side effect}. Actions that we humans think are |
| 1632 | important, such as printing results, are often ``side effects'' to the |
| 1633 | Lisp interpreter. The jargon can sound peculiar, but it turns out that |
| 1634 | it is fairly easy to learn to use side effects. |
| 1635 | |
| 1636 | In summary, evaluating a symbolic expression most commonly causes the |
| 1637 | Lisp interpreter to return a value and perhaps carry out a side effect; |
| 1638 | or else produce an error. |
| 1639 | |
| 1640 | @menu |
| 1641 | * Evaluating Inner Lists:: Lists within lists... |
| 1642 | @end menu |
| 1643 | |
| 1644 | @node Evaluating Inner Lists, , Evaluation, Evaluation |
| 1645 | @comment node-name, next, previous, up |
| 1646 | @subsection Evaluating Inner Lists |
| 1647 | @cindex Inner list evaluation |
| 1648 | @cindex Evaluating inner lists |
| 1649 | |
| 1650 | If evaluation applies to a list that is inside another list, the outer |
| 1651 | list may use the value returned by the first evaluation as information |
| 1652 | when the outer list is evaluated. This explains why inner expressions |
| 1653 | are evaluated first: the values they return are used by the outer |
| 1654 | expressions. |
| 1655 | |
| 1656 | @need 1250 |
| 1657 | We can investigate this process by evaluating another addition example. |
| 1658 | Place your cursor after the following expression and type @kbd{C-x C-e}: |
| 1659 | |
| 1660 | @smallexample |
| 1661 | (+ 2 (+ 3 3)) |
| 1662 | @end smallexample |
| 1663 | |
| 1664 | @noindent |
| 1665 | The number 8 will appear in the echo area. |
| 1666 | |
| 1667 | What happens is that the Lisp interpreter first evaluates the inner |
| 1668 | expression, @code{(+ 3 3)}, for which the value 6 is returned; then it |
| 1669 | evaluates the outer expression as if it were written @code{(+ 2 6)}, which |
| 1670 | returns the value 8. Since there are no more enclosing expressions to |
| 1671 | evaluate, the interpreter prints that value in the echo area. |
| 1672 | |
| 1673 | Now it is easy to understand the name of the command invoked by the |
| 1674 | keystrokes @kbd{C-x C-e}: the name is @code{eval-last-sexp}. The |
| 1675 | letters @code{sexp} are an abbreviation for `symbolic expression', and |
| 1676 | @code{eval} is an abbreviation for `evaluate'. The command means |
| 1677 | `evaluate last symbolic expression'. |
| 1678 | |
| 1679 | As an experiment, you can try evaluating the expression by putting the |
| 1680 | cursor at the beginning of the next line immediately following the |
| 1681 | expression, or inside the expression. |
| 1682 | |
| 1683 | @need 800 |
| 1684 | Here is another copy of the expression: |
| 1685 | |
| 1686 | @smallexample |
| 1687 | (+ 2 (+ 3 3)) |
| 1688 | @end smallexample |
| 1689 | |
| 1690 | @noindent |
| 1691 | If you place the cursor at the beginning of the blank line that |
| 1692 | immediately follows the expression and type @kbd{C-x C-e}, you will |
| 1693 | still get the value 8 printed in the echo area. Now try putting the |
| 1694 | cursor inside the expression. If you put it right after the next to |
| 1695 | last parenthesis (so it appears to sit on top of the last parenthesis), |
| 1696 | you will get a 6 printed in the echo area! This is because the command |
| 1697 | evaluates the expression @code{(+ 3 3)}. |
| 1698 | |
| 1699 | Now put the cursor immediately after a number. Type @kbd{C-x C-e} and |
| 1700 | you will get the number itself. In Lisp, if you evaluate a number, you |
| 1701 | get the number itself---this is how numbers differ from symbols. If you |
| 1702 | evaluate a list starting with a symbol like @code{+}, you will get a |
| 1703 | value returned that is the result of the computer carrying out the |
| 1704 | instructions in the function definition attached to that name. If a |
| 1705 | symbol by itself is evaluated, something different happens, as we will |
| 1706 | see in the next section. |
| 1707 | |
| 1708 | @node Variables, Arguments, Evaluation, List Processing |
| 1709 | @comment node-name, next, previous, up |
| 1710 | @section Variables |
| 1711 | @cindex Variables |
| 1712 | |
| 1713 | In Emacs Lisp, a symbol can have a value attached to it just as it can |
| 1714 | have a function definition attached to it. The two are different. |
| 1715 | The function definition is a set of instructions that a computer will |
| 1716 | obey. A value, on the other hand, is something, such as number or a |
| 1717 | name, that can vary (which is why such a symbol is called a variable). |
| 1718 | The value of a symbol can be any expression in Lisp, such as a symbol, |
| 1719 | number, list, or string. A symbol that has a value is often called a |
| 1720 | @dfn{variable}. |
| 1721 | |
| 1722 | A symbol can have both a function definition and a value attached to |
| 1723 | it at the same time. Or it can have just one or the other. |
| 1724 | The two are separate. This is somewhat similar |
| 1725 | to the way the name Cambridge can refer to the city in Massachusetts |
| 1726 | and have some information attached to the name as well, such as |
| 1727 | ``great programming center''. |
| 1728 | |
| 1729 | @ignore |
| 1730 | (Incidentally, in Emacs Lisp, a symbol can have two |
| 1731 | other things attached to it, too: a property list and a documentation |
| 1732 | string; these are discussed later.) |
| 1733 | @end ignore |
| 1734 | |
| 1735 | Another way to think about this is to imagine a symbol as being a chest |
| 1736 | of drawers. The function definition is put in one drawer, the value in |
| 1737 | another, and so on. What is put in the drawer holding the value can be |
| 1738 | changed without affecting the contents of the drawer holding the |
| 1739 | function definition, and vice-versa. |
| 1740 | |
| 1741 | @menu |
| 1742 | * fill-column Example:: |
| 1743 | * Void Function:: The error message for a symbol |
| 1744 | without a function. |
| 1745 | * Void Variable:: The error message for a symbol without a value. |
| 1746 | @end menu |
| 1747 | |
| 1748 | @node fill-column Example, Void Function, Variables, Variables |
| 1749 | @ifnottex |
| 1750 | @unnumberedsubsec @code{fill-column}, an Example Variable |
| 1751 | @end ifnottex |
| 1752 | |
| 1753 | @findex fill-column, @r{an example variable} |
| 1754 | @cindex Example variable, @code{fill-column} |
| 1755 | @cindex Variable, example of, @code{fill-column} |
| 1756 | The variable @code{fill-column} illustrates a symbol with a value |
| 1757 | attached to it: in every GNU Emacs buffer, this symbol is set to some |
| 1758 | value, usually 72 or 70, but sometimes to some other value. To find the |
| 1759 | value of this symbol, evaluate it by itself. If you are reading this in |
| 1760 | Info inside of GNU Emacs, you can do this by putting the cursor after |
| 1761 | the symbol and typing @kbd{C-x C-e}: |
| 1762 | |
| 1763 | @smallexample |
| 1764 | fill-column |
| 1765 | @end smallexample |
| 1766 | |
| 1767 | @noindent |
| 1768 | After I typed @kbd{C-x C-e}, Emacs printed the number 72 in my echo |
| 1769 | area. This is the value for which @code{fill-column} is set for me as I |
| 1770 | write this. It may be different for you in your Info buffer. Notice |
| 1771 | that the value returned as a variable is printed in exactly the same way |
| 1772 | as the value returned by a function carrying out its instructions. From |
| 1773 | the point of view of the Lisp interpreter, a value returned is a value |
| 1774 | returned. What kind of expression it came from ceases to matter once |
| 1775 | the value is known. |
| 1776 | |
| 1777 | A symbol can have any value attached to it or, to use the jargon, we can |
| 1778 | @dfn{bind} the variable to a value: to a number, such as 72; to a |
| 1779 | string, @code{"such as this"}; to a list, such as @code{(spruce pine |
| 1780 | oak)}; we can even bind a variable to a function definition. |
| 1781 | |
| 1782 | A symbol can be bound to a value in several ways. @xref{set & setq, , |
| 1783 | Setting the Value of a Variable}, for information about one way to do |
| 1784 | this. |
| 1785 | |
| 1786 | @node Void Function, Void Variable, fill-column Example, Variables |
| 1787 | @comment node-name, next, previous, up |
| 1788 | @subsection Error Message for a Symbol Without a Function |
| 1789 | @cindex Symbol without function error |
| 1790 | @cindex Error for symbol without function |
| 1791 | |
| 1792 | When we evaluated @code{fill-column} to find its value as a variable, |
| 1793 | we did not place parentheses around the word. This is because we did |
| 1794 | not intend to use it as a function name. |
| 1795 | |
| 1796 | If @code{fill-column} were the first or only element of a list, the |
| 1797 | Lisp interpreter would attempt to find the function definition |
| 1798 | attached to it. But @code{fill-column} has no function definition. |
| 1799 | Try evaluating this: |
| 1800 | |
| 1801 | @smallexample |
| 1802 | (fill-column) |
| 1803 | @end smallexample |
| 1804 | |
| 1805 | @need 1250 |
| 1806 | @noindent |
| 1807 | In GNU Emacs version 21, you will create a @file{*Backtrace*} buffer |
| 1808 | that says: |
| 1809 | |
| 1810 | @smallexample |
| 1811 | @group |
| 1812 | ---------- Buffer: *Backtrace* ---------- |
| 1813 | Debugger entered--Lisp error: (void-function fill-column) |
| 1814 | (fill-column) |
| 1815 | eval((fill-column)) |
| 1816 | eval-last-sexp-1(nil) |
| 1817 | eval-last-sexp(nil) |
| 1818 | call-interactively(eval-last-sexp) |
| 1819 | ---------- Buffer: *Backtrace* ---------- |
| 1820 | @end group |
| 1821 | @end smallexample |
| 1822 | |
| 1823 | @noindent |
| 1824 | (Remember, to quit the debugger and make the debugger window go away, |
| 1825 | type @kbd{q} in the @file{*Backtrace*} buffer.) |
| 1826 | |
| 1827 | @need 800 |
| 1828 | In GNU Emacs 20 and before, you will produce an error message that says: |
| 1829 | |
| 1830 | @smallexample |
| 1831 | Symbol's function definition is void:@: fill-column |
| 1832 | @end smallexample |
| 1833 | |
| 1834 | @noindent |
| 1835 | (The message will go away away as soon as you move the cursor or type |
| 1836 | another key.) |
| 1837 | |
| 1838 | @node Void Variable, , Void Function, Variables |
| 1839 | @comment node-name, next, previous, up |
| 1840 | @subsection Error Message for a Symbol Without a Value |
| 1841 | @cindex Symbol without value error |
| 1842 | @cindex Error for symbol without value |
| 1843 | |
| 1844 | If you attempt to evaluate a symbol that does not have a value bound to |
| 1845 | it, you will receive an error message. You can see this by |
| 1846 | experimenting with our 2 plus 2 addition. In the following expression, |
| 1847 | put your cursor right after the @code{+}, before the first number 2, |
| 1848 | type @kbd{C-x C-e}: |
| 1849 | |
| 1850 | @smallexample |
| 1851 | (+ 2 2) |
| 1852 | @end smallexample |
| 1853 | |
| 1854 | @need 1500 |
| 1855 | @noindent |
| 1856 | In GNU Emacs 21, you will create a @file{*Backtrace*} buffer that |
| 1857 | says: |
| 1858 | |
| 1859 | @smallexample |
| 1860 | @group |
| 1861 | ---------- Buffer: *Backtrace* ---------- |
| 1862 | Debugger entered--Lisp error: (void-variable +) |
| 1863 | eval(+) |
| 1864 | eval-last-sexp-1(nil) |
| 1865 | eval-last-sexp(nil) |
| 1866 | call-interactively(eval-last-sexp) |
| 1867 | ---------- Buffer: *Backtrace* ---------- |
| 1868 | @end group |
| 1869 | @end smallexample |
| 1870 | |
| 1871 | @noindent |
| 1872 | (As with the other times we entered the debugger, you can quit by |
| 1873 | typing @kbd{q} in the @file{*Backtrace*} buffer.) |
| 1874 | |
| 1875 | This backtrace is different from the very first error message we saw, |
| 1876 | which said, @samp{Debugger entered--Lisp error: (void-function this)}. |
| 1877 | In this case, the function does not have a value as a variable; while |
| 1878 | in the other error message, the function (the word `this') did not |
| 1879 | have a definition. |
| 1880 | |
| 1881 | In this experiment with the @code{+}, what we did was cause the Lisp |
| 1882 | interpreter to evaluate the @code{+} and look for the value of the |
| 1883 | variable instead of the function definition. We did this by placing the |
| 1884 | cursor right after the symbol rather than after the parenthesis of the |
| 1885 | enclosing list as we did before. As a consequence, the Lisp interpreter |
| 1886 | evaluated the preceding s-expression, which in this case was the |
| 1887 | @code{+} by itself. |
| 1888 | |
| 1889 | Since @code{+} does not have a value bound to it, just the function |
| 1890 | definition, the error message reported that the symbol's value as a |
| 1891 | variable was void. |
| 1892 | |
| 1893 | @need 800 |
| 1894 | In GNU Emacs version 20 and before, your error message will say: |
| 1895 | |
| 1896 | @example |
| 1897 | Symbol's value as variable is void:@: + |
| 1898 | @end example |
| 1899 | |
| 1900 | @noindent |
| 1901 | The meaning is the same as in GNU Emacs 21. |
| 1902 | |
| 1903 | @node Arguments, set & setq, Variables, List Processing |
| 1904 | @comment node-name, next, previous, up |
| 1905 | @section Arguments |
| 1906 | @cindex Arguments |
| 1907 | @cindex Passing information to functions |
| 1908 | |
| 1909 | To see how information is passed to functions, let's look again at |
| 1910 | our old standby, the addition of two plus two. In Lisp, this is written |
| 1911 | as follows: |
| 1912 | |
| 1913 | @smallexample |
| 1914 | (+ 2 2) |
| 1915 | @end smallexample |
| 1916 | |
| 1917 | If you evaluate this expression, the number 4 will appear in your echo |
| 1918 | area. What the Lisp interpreter does is add the numbers that follow |
| 1919 | the @code{+}. |
| 1920 | |
| 1921 | @cindex @samp{argument} defined |
| 1922 | The numbers added by @code{+} are called the @dfn{arguments} of the |
| 1923 | function @code{+}. These numbers are the information that is given to |
| 1924 | or @dfn{passed} to the function. |
| 1925 | |
| 1926 | The word `argument' comes from the way it is used in mathematics and |
| 1927 | does not refer to a disputation between two people; instead it refers to |
| 1928 | the information presented to the function, in this case, to the |
| 1929 | @code{+}. In Lisp, the arguments to a function are the atoms or lists |
| 1930 | that follow the function. The values returned by the evaluation of |
| 1931 | these atoms or lists are passed to the function. Different functions |
| 1932 | require different numbers of arguments; some functions require none at |
| 1933 | all.@footnote{It is curious to track the path by which the word `argument' |
| 1934 | came to have two different meanings, one in mathematics and the other in |
| 1935 | everyday English. According to the @cite{Oxford English Dictionary}, |
| 1936 | the word derives from the Latin for @samp{to make clear, prove}; thus it |
| 1937 | came to mean, by one thread of derivation, `the evidence offered as |
| 1938 | proof', which is to say, `the information offered', which led to its |
| 1939 | meaning in Lisp. But in the other thread of derivation, it came to mean |
| 1940 | `to assert in a manner against which others may make counter |
| 1941 | assertions', which led to the meaning of the word as a disputation. |
| 1942 | (Note here that the English word has two different definitions attached |
| 1943 | to it at the same time. By contrast, in Emacs Lisp, a symbol cannot |
| 1944 | have two different function definitions at the same time.)} |
| 1945 | |
| 1946 | @menu |
| 1947 | * Data types:: Types of data passed to a function. |
| 1948 | * Args as Variable or List:: An argument can be the value |
| 1949 | of a variable or list. |
| 1950 | * Variable Number of Arguments:: Some functions may take a |
| 1951 | variable number of arguments. |
| 1952 | * Wrong Type of Argument:: Passing an argument of the wrong type |
| 1953 | to a function. |
| 1954 | * message:: A useful function for sending messages. |
| 1955 | @end menu |
| 1956 | |
| 1957 | @node Data types, Args as Variable or List, Arguments, Arguments |
| 1958 | @comment node-name, next, previous, up |
| 1959 | @subsection Arguments' Data Types |
| 1960 | @cindex Data types |
| 1961 | @cindex Types of data |
| 1962 | @cindex Arguments' data types |
| 1963 | |
| 1964 | The type of data that should be passed to a function depends on what |
| 1965 | kind of information it uses. The arguments to a function such as |
| 1966 | @code{+} must have values that are numbers, since @code{+} adds numbers. |
| 1967 | Other functions use different kinds of data for their arguments. |
| 1968 | |
| 1969 | @need 1250 |
| 1970 | @findex concat |
| 1971 | For example, the @code{concat} function links together or unites two or |
| 1972 | more strings of text to produce a string. The arguments are strings. |
| 1973 | Concatenating the two character strings @code{abc}, @code{def} produces |
| 1974 | the single string @code{abcdef}. This can be seen by evaluating the |
| 1975 | following: |
| 1976 | |
| 1977 | @smallexample |
| 1978 | (concat "abc" "def") |
| 1979 | @end smallexample |
| 1980 | |
| 1981 | @noindent |
| 1982 | The value produced by evaluating this expression is @code{"abcdef"}. |
| 1983 | |
| 1984 | A function such as @code{substring} uses both a string and numbers as |
| 1985 | arguments. The function returns a part of the string, a substring of |
| 1986 | the first argument. This function takes three arguments. Its first |
| 1987 | argument is the string of characters, the second and third arguments are |
| 1988 | numbers that indicate the beginning and end of the substring. The |
| 1989 | numbers are a count of the number of characters (including spaces and |
| 1990 | punctuations) from the beginning of the string. |
| 1991 | |
| 1992 | @need 800 |
| 1993 | For example, if you evaluate the following: |
| 1994 | |
| 1995 | @smallexample |
| 1996 | (substring "The quick brown fox jumped." 16 19) |
| 1997 | @end smallexample |
| 1998 | |
| 1999 | @noindent |
| 2000 | you will see @code{"fox"} appear in the echo area. The arguments are the |
| 2001 | string and the two numbers. |
| 2002 | |
| 2003 | Note that the string passed to @code{substring} is a single atom even |
| 2004 | though it is made up of several words separated by spaces. Lisp counts |
| 2005 | everything between the two quotation marks as part of the string, |
| 2006 | including the spaces. You can think of the @code{substring} function as |
| 2007 | a kind of `atom smasher' since it takes an otherwise indivisible atom |
| 2008 | and extracts a part. However, @code{substring} is only able to extract |
| 2009 | a substring from an argument that is a string, not from another type of |
| 2010 | atom such as a number or symbol. |
| 2011 | |
| 2012 | @node Args as Variable or List, Variable Number of Arguments, Data types, Arguments |
| 2013 | @comment node-name, next, previous, up |
| 2014 | @subsection An Argument as the Value of a Variable or List |
| 2015 | |
| 2016 | An argument can be a symbol that returns a value when it is evaluated. |
| 2017 | For example, when the symbol @code{fill-column} by itself is evaluated, |
| 2018 | it returns a number. This number can be used in an addition. |
| 2019 | |
| 2020 | @need 1250 |
| 2021 | Position the cursor after the following expression and type @kbd{C-x |
| 2022 | C-e}: |
| 2023 | |
| 2024 | @smallexample |
| 2025 | (+ 2 fill-column) |
| 2026 | @end smallexample |
| 2027 | |
| 2028 | @noindent |
| 2029 | The value will be a number two more than what you get by evaluating |
| 2030 | @code{fill-column} alone. For me, this is 74, because the value of |
| 2031 | @code{fill-column} is 72. |
| 2032 | |
| 2033 | As we have just seen, an argument can be a symbol that returns a value |
| 2034 | when evaluated. In addition, an argument can be a list that returns a |
| 2035 | value when it is evaluated. For example, in the following expression, |
| 2036 | the arguments to the function @code{concat} are the strings |
| 2037 | @w{@code{"The "}} and @w{@code{" red foxes."}} and the list |
| 2038 | @code{(number-to-string (+ 2 fill-column))}. |
| 2039 | |
| 2040 | @c For Emacs 21, need number-to-string |
| 2041 | @smallexample |
| 2042 | (concat "The " (number-to-string (+ 2 fill-column)) " red foxes.") |
| 2043 | @end smallexample |
| 2044 | |
| 2045 | @noindent |
| 2046 | If you evaluate this expression---and if, as with my Emacs, |
| 2047 | @code{fill-column} evaluates to 72---@code{"The 74 red foxes."} will |
| 2048 | appear in the echo area. (Note that you must put spaces after the |
| 2049 | word @samp{The} and before the word @samp{red} so they will appear in |
| 2050 | the final string. The function @code{number-to-string} converts the |
| 2051 | integer that the addition function returns to a string. |
| 2052 | @code{number-to-string} is also known as @code{int-to-string}.) |
| 2053 | |
| 2054 | @node Variable Number of Arguments, Wrong Type of Argument, Args as Variable or List, Arguments |
| 2055 | @comment node-name, next, previous, up |
| 2056 | @subsection Variable Number of Arguments |
| 2057 | @cindex Variable number of arguments |
| 2058 | @cindex Arguments, variable number of |
| 2059 | |
| 2060 | Some functions, such as @code{concat}, @code{+} or @code{*}, take any |
| 2061 | number of arguments. (The @code{*} is the symbol for multiplication.) |
| 2062 | This can be seen by evaluating each of the following expressions in |
| 2063 | the usual way. What you will see in the echo area is printed in this |
| 2064 | text after @samp{@result{}}, which you may read as `evaluates to'. |
| 2065 | |
| 2066 | @need 1250 |
| 2067 | In the first set, the functions have no arguments: |
| 2068 | |
| 2069 | @smallexample |
| 2070 | @group |
| 2071 | (+) @result{} 0 |
| 2072 | |
| 2073 | (*) @result{} 1 |
| 2074 | @end group |
| 2075 | @end smallexample |
| 2076 | |
| 2077 | @need 1250 |
| 2078 | In this set, the functions have one argument each: |
| 2079 | |
| 2080 | @smallexample |
| 2081 | @group |
| 2082 | (+ 3) @result{} 3 |
| 2083 | |
| 2084 | (* 3) @result{} 3 |
| 2085 | @end group |
| 2086 | @end smallexample |
| 2087 | |
| 2088 | @need 1250 |
| 2089 | In this set, the functions have three arguments each: |
| 2090 | |
| 2091 | @smallexample |
| 2092 | @group |
| 2093 | (+ 3 4 5) @result{} 12 |
| 2094 | |
| 2095 | (* 3 4 5) @result{} 60 |
| 2096 | @end group |
| 2097 | @end smallexample |
| 2098 | |
| 2099 | @node Wrong Type of Argument, message, Variable Number of Arguments, Arguments |
| 2100 | @comment node-name, next, previous, up |
| 2101 | @subsection Using the Wrong Type Object as an Argument |
| 2102 | @cindex Wrong type of argument |
| 2103 | @cindex Argument, wrong type of |
| 2104 | |
| 2105 | When a function is passed an argument of the wrong type, the Lisp |
| 2106 | interpreter produces an error message. For example, the @code{+} |
| 2107 | function expects the values of its arguments to be numbers. As an |
| 2108 | experiment we can pass it the quoted symbol @code{hello} instead of a |
| 2109 | number. Position the cursor after the following expression and type |
| 2110 | @kbd{C-x C-e}: |
| 2111 | |
| 2112 | @smallexample |
| 2113 | (+ 2 'hello) |
| 2114 | @end smallexample |
| 2115 | |
| 2116 | @noindent |
| 2117 | When you do this you will generate an error message. What has happened |
| 2118 | is that @code{+} has tried to add the 2 to the value returned by |
| 2119 | @code{'hello}, but the value returned by @code{'hello} is the symbol |
| 2120 | @code{hello}, not a number. Only numbers can be added. So @code{+} |
| 2121 | could not carry out its addition. |
| 2122 | |
| 2123 | @need 1250 |
| 2124 | In GNU Emacs version 21, you will create and enter a |
| 2125 | @file{*Backtrace*} buffer that says: |
| 2126 | |
| 2127 | @noindent |
| 2128 | @smallexample |
| 2129 | @group |
| 2130 | ---------- Buffer: *Backtrace* ---------- |
| 2131 | Debugger entered--Lisp error: |
| 2132 | (wrong-type-argument number-or-marker-p hello) |
| 2133 | +(2 hello) |
| 2134 | eval((+ 2 (quote hello))) |
| 2135 | eval-last-sexp-1(nil) |
| 2136 | eval-last-sexp(nil) |
| 2137 | call-interactively(eval-last-sexp) |
| 2138 | ---------- Buffer: *Backtrace* ---------- |
| 2139 | @end group |
| 2140 | @end smallexample |
| 2141 | |
| 2142 | @need 1250 |
| 2143 | As usual, the error message tries to be helpful and makes sense after you |
| 2144 | learn how to read it.@footnote{@code{(quote hello)} is an expansion of |
| 2145 | the abbreviation @code{'hello}.} |
| 2146 | |
| 2147 | The first part of the error message is straightforward; it says |
| 2148 | @samp{wrong type argument}. Next comes the mysterious jargon word |
| 2149 | @w{@samp{number-or-marker-p}}. This word is trying to tell you what |
| 2150 | kind of argument the @code{+} expected. |
| 2151 | |
| 2152 | The symbol @code{number-or-marker-p} says that the Lisp interpreter is |
| 2153 | trying to determine whether the information presented it (the value of |
| 2154 | the argument) is a number or a marker (a special object representing a |
| 2155 | buffer position). What it does is test to see whether the @code{+} is |
| 2156 | being given numbers to add. It also tests to see whether the |
| 2157 | argument is something called a marker, which is a specific feature of |
| 2158 | Emacs Lisp. (In Emacs, locations in a buffer are recorded as markers. |
| 2159 | When the mark is set with the @kbd{C-@@} or @kbd{C-@key{SPC}} command, |
| 2160 | its position is kept as a marker. The mark can be considered a |
| 2161 | number---the number of characters the location is from the beginning |
| 2162 | of the buffer.) In Emacs Lisp, @code{+} can be used to add the |
| 2163 | numeric value of marker positions as numbers. |
| 2164 | |
| 2165 | The @samp{p} of @code{number-or-marker-p} is the embodiment of a |
| 2166 | practice started in the early days of Lisp programming. The @samp{p} |
| 2167 | stands for `predicate'. In the jargon used by the early Lisp |
| 2168 | researchers, a predicate refers to a function to determine whether some |
| 2169 | property is true or false. So the @samp{p} tells us that |
| 2170 | @code{number-or-marker-p} is the name of a function that determines |
| 2171 | whether it is true or false that the argument supplied is a number or |
| 2172 | a marker. Other Lisp symbols that end in @samp{p} include @code{zerop}, |
| 2173 | a function that tests whether its argument has the value of zero, and |
| 2174 | @code{listp}, a function that tests whether its argument is a list. |
| 2175 | |
| 2176 | Finally, the last part of the error message is the symbol @code{hello}. |
| 2177 | This is the value of the argument that was passed to @code{+}. If the |
| 2178 | addition had been passed the correct type of object, the value passed |
| 2179 | would have been a number, such as 37, rather than a symbol like |
| 2180 | @code{hello}. But then you would not have got the error message. |
| 2181 | |
| 2182 | @need 1250 |
| 2183 | In GNU Emacs version 20 and before, the echo area displays an error |
| 2184 | message that says: |
| 2185 | |
| 2186 | @smallexample |
| 2187 | Wrong type argument:@: number-or-marker-p, hello |
| 2188 | @end smallexample |
| 2189 | |
| 2190 | This says, in different words, the same as the top line of the |
| 2191 | @file{*Backtrace*} buffer. |
| 2192 | |
| 2193 | @node message, , Wrong Type of Argument, Arguments |
| 2194 | @comment node-name, next, previous, up |
| 2195 | @subsection The @code{message} Function |
| 2196 | @findex message |
| 2197 | |
| 2198 | Like @code{+}, the @code{message} function takes a variable number of |
| 2199 | arguments. It is used to send messages to the user and is so useful |
| 2200 | that we will describe it here. |
| 2201 | |
| 2202 | @need 1250 |
| 2203 | A message is printed in the echo area. For example, you can print a |
| 2204 | message in your echo area by evaluating the following list: |
| 2205 | |
| 2206 | @smallexample |
| 2207 | (message "This message appears in the echo area!") |
| 2208 | @end smallexample |
| 2209 | |
| 2210 | The whole string between double quotation marks is a single argument |
| 2211 | and is printed @i{in toto}. (Note that in this example, the message |
| 2212 | itself will appear in the echo area within double quotes; that is |
| 2213 | because you see the value returned by the @code{message} function. In |
| 2214 | most uses of @code{message} in programs that you write, the text will |
| 2215 | be printed in the echo area as a side-effect, without the quotes. |
| 2216 | @xref{multiply-by-seven in detail, , @code{multiply-by-seven} in |
| 2217 | detail}, for an example of this.) |
| 2218 | |
| 2219 | However, if there is a @samp{%s} in the quoted string of characters, the |
| 2220 | @code{message} function does not print the @samp{%s} as such, but looks |
| 2221 | to the argument that follows the string. It evaluates the second |
| 2222 | argument and prints the value at the location in the string where the |
| 2223 | @samp{%s} is. |
| 2224 | |
| 2225 | @need 1250 |
| 2226 | You can see this by positioning the cursor after the following |
| 2227 | expression and typing @kbd{C-x C-e}: |
| 2228 | |
| 2229 | @smallexample |
| 2230 | (message "The name of this buffer is: %s." (buffer-name)) |
| 2231 | @end smallexample |
| 2232 | |
| 2233 | @noindent |
| 2234 | In Info, @code{"The name of this buffer is: *info*."} will appear in the |
| 2235 | echo area. The function @code{buffer-name} returns the name of the |
| 2236 | buffer as a string, which the @code{message} function inserts in place |
| 2237 | of @code{%s}. |
| 2238 | |
| 2239 | To print a value as an integer, use @samp{%d} in the same way as |
| 2240 | @samp{%s}. For example, to print a message in the echo area that |
| 2241 | states the value of the @code{fill-column}, evaluate the following: |
| 2242 | |
| 2243 | @smallexample |
| 2244 | (message "The value of fill-column is %d." fill-column) |
| 2245 | @end smallexample |
| 2246 | |
| 2247 | @noindent |
| 2248 | On my system, when I evaluate this list, @code{"The value of |
| 2249 | fill-column is 72."} appears in my echo area@footnote{Actually, you |
| 2250 | can use @code{%s} to print a number. It is non-specific. @code{%d} |
| 2251 | prints only the part of a number left of a decimal point, and not |
| 2252 | anything that is not a number.}. |
| 2253 | |
| 2254 | If there is more than one @samp{%s} in the quoted string, the value of |
| 2255 | the first argument following the quoted string is printed at the |
| 2256 | location of the first @samp{%s} and the value of the second argument is |
| 2257 | printed at the location of the second @samp{%s}, and so on. |
| 2258 | |
| 2259 | @need 1250 |
| 2260 | For example, if you evaluate the following, |
| 2261 | |
| 2262 | @smallexample |
| 2263 | @group |
| 2264 | (message "There are %d %s in the office!" |
| 2265 | (- fill-column 14) "pink elephants") |
| 2266 | @end group |
| 2267 | @end smallexample |
| 2268 | |
| 2269 | @noindent |
| 2270 | a rather whimsical message will appear in your echo area. On my system |
| 2271 | it says, @code{"There are 58 pink elephants in the office!"}. |
| 2272 | |
| 2273 | The expression @code{(- fill-column 14)} is evaluated and the resulting |
| 2274 | number is inserted in place of the @samp{%d}; and the string in double |
| 2275 | quotes, @code{"pink elephants"}, is treated as a single argument and |
| 2276 | inserted in place of the @samp{%s}. (That is to say, a string between |
| 2277 | double quotes evaluates to itself, like a number.) |
| 2278 | |
| 2279 | Finally, here is a somewhat complex example that not only illustrates |
| 2280 | the computation of a number, but also shows how you can use an |
| 2281 | expression within an expression to generate the text that is substituted |
| 2282 | for @samp{%s}: |
| 2283 | |
| 2284 | @smallexample |
| 2285 | @group |
| 2286 | (message "He saw %d %s" |
| 2287 | (- fill-column 32) |
| 2288 | (concat "red " |
| 2289 | (substring |
| 2290 | "The quick brown foxes jumped." 16 21) |
| 2291 | " leaping.")) |
| 2292 | @end group |
| 2293 | @end smallexample |
| 2294 | |
| 2295 | In this example, @code{message} has three arguments: the string, |
| 2296 | @code{"He saw %d %s"}, the expression, @code{(- fill-column 32)}, and |
| 2297 | the expression beginning with the function @code{concat}. The value |
| 2298 | resulting from the evaluation of @code{(- fill-column 32)} is inserted |
| 2299 | in place of the @samp{%d}; and the value returned by the expression |
| 2300 | beginning with @code{concat} is inserted in place of the @samp{%s}. |
| 2301 | |
| 2302 | When I evaluate the expression, the message @code{"He saw 38 red |
| 2303 | foxes leaping."} appears in my echo area. |
| 2304 | |
| 2305 | @node set & setq, Summary, Arguments, List Processing |
| 2306 | @comment node-name, next, previous, up |
| 2307 | @section Setting the Value of a Variable |
| 2308 | @cindex Variable, setting value |
| 2309 | @cindex Setting value of variable |
| 2310 | |
| 2311 | @cindex @samp{bind} defined |
| 2312 | There are several ways by which a variable can be given a value. One of |
| 2313 | the ways is to use either the function @code{set} or the function |
| 2314 | @code{setq}. Another way is to use @code{let} (@pxref{let}). (The |
| 2315 | jargon for this process is to @dfn{bind} a variable to a value.) |
| 2316 | |
| 2317 | The following sections not only describe how @code{set} and @code{setq} |
| 2318 | work but also illustrate how arguments are passed. |
| 2319 | |
| 2320 | @menu |
| 2321 | * Using set:: Setting values. |
| 2322 | * Using setq:: Setting a quoted value. |
| 2323 | * Counting:: Using @code{setq} to count. |
| 2324 | @end menu |
| 2325 | |
| 2326 | @node Using set, Using setq, set & setq, set & setq |
| 2327 | @comment node-name, next, previous, up |
| 2328 | @subsection Using @code{set} |
| 2329 | @findex set |
| 2330 | |
| 2331 | To set the value of the symbol @code{flowers} to the list @code{'(rose |
| 2332 | violet daisy buttercup)}, evaluate the following expression by |
| 2333 | positioning the cursor after the expression and typing @kbd{C-x C-e}. |
| 2334 | |
| 2335 | @smallexample |
| 2336 | (set 'flowers '(rose violet daisy buttercup)) |
| 2337 | @end smallexample |
| 2338 | |
| 2339 | @noindent |
| 2340 | The list @code{(rose violet daisy buttercup)} will appear in the echo |
| 2341 | area. This is what is @emph{returned} by the @code{set} function. As a |
| 2342 | side effect, the symbol @code{flowers} is bound to the list ; that is, |
| 2343 | the symbol @code{flowers}, which can be viewed as a variable, is given |
| 2344 | the list as its value. (This process, by the way, illustrates how a |
| 2345 | side effect to the Lisp interpreter, setting the value, can be the |
| 2346 | primary effect that we humans are interested in. This is because every |
| 2347 | Lisp function must return a value if it does not get an error, but it |
| 2348 | will only have a side effect if it is designed to have one.) |
| 2349 | |
| 2350 | After evaluating the @code{set} expression, you can evaluate the symbol |
| 2351 | @code{flowers} and it will return the value you just set. Here is the |
| 2352 | symbol. Place your cursor after it and type @kbd{C-x C-e}. |
| 2353 | |
| 2354 | @smallexample |
| 2355 | flowers |
| 2356 | @end smallexample |
| 2357 | |
| 2358 | @noindent |
| 2359 | When you evaluate @code{flowers}, the list |
| 2360 | @code{(rose violet daisy buttercup)} appears in the echo area. |
| 2361 | |
| 2362 | Incidentally, if you evaluate @code{'flowers}, the variable with a quote |
| 2363 | in front of it, what you will see in the echo area is the symbol itself, |
| 2364 | @code{flowers}. Here is the quoted symbol, so you can try this: |
| 2365 | |
| 2366 | @smallexample |
| 2367 | 'flowers |
| 2368 | @end smallexample |
| 2369 | |
| 2370 | Note also, that when you use @code{set}, you need to quote both |
| 2371 | arguments to @code{set}, unless you want them evaluated. Since we do |
| 2372 | not want either argument evaluated, neither the variable |
| 2373 | @code{flowers} nor the list @code{(rose violet daisy buttercup)}, both |
| 2374 | are quoted. (When you use @code{set} without quoting its first |
| 2375 | argument, the first argument is evaluated before anything else is |
| 2376 | done. If you did this and @code{flowers} did not have a value |
| 2377 | already, you would get an error message that the @samp{Symbol's value |
| 2378 | as variable is void}; on the other hand, if @code{flowers} did return |
| 2379 | a value after it was evaluated, the @code{set} would attempt to set |
| 2380 | the value that was returned. There are situations where this is the |
| 2381 | right thing for the function to do; but such situations are rare.) |
| 2382 | |
| 2383 | @node Using setq, Counting, Using set, set & setq |
| 2384 | @comment node-name, next, previous, up |
| 2385 | @subsection Using @code{setq} |
| 2386 | @findex setq |
| 2387 | |
| 2388 | As a practical matter, you almost always quote the first argument to |
| 2389 | @code{set}. The combination of @code{set} and a quoted first argument |
| 2390 | is so common that it has its own name: the special form @code{setq}. |
| 2391 | This special form is just like @code{set} except that the first argument |
| 2392 | is quoted automatically, so you don't need to type the quote mark |
| 2393 | yourself. Also, as an added convenience, @code{setq} permits you to set |
| 2394 | several different variables to different values, all in one expression. |
| 2395 | |
| 2396 | To set the value of the variable @code{carnivores} to the list |
| 2397 | @code{'(lion tiger leopard)} using @code{setq}, the following expression |
| 2398 | is used: |
| 2399 | |
| 2400 | @smallexample |
| 2401 | (setq carnivores '(lion tiger leopard)) |
| 2402 | @end smallexample |
| 2403 | |
| 2404 | @noindent |
| 2405 | This is exactly the same as using @code{set} except the first argument |
| 2406 | is automatically quoted by @code{setq}. (The @samp{q} in @code{setq} |
| 2407 | means @code{quote}.) |
| 2408 | |
| 2409 | @need 1250 |
| 2410 | With @code{set}, the expression would look like this: |
| 2411 | |
| 2412 | @smallexample |
| 2413 | (set 'carnivores '(lion tiger leopard)) |
| 2414 | @end smallexample |
| 2415 | |
| 2416 | Also, @code{setq} can be used to assign different values to |
| 2417 | different variables. The first argument is bound to the value |
| 2418 | of the second argument, the third argument is bound to the value of the |
| 2419 | fourth argument, and so on. For example, you could use the following to |
| 2420 | assign a list of trees to the symbol @code{trees} and a list of herbivores |
| 2421 | to the symbol @code{herbivores}: |
| 2422 | |
| 2423 | @smallexample |
| 2424 | @group |
| 2425 | (setq trees '(pine fir oak maple) |
| 2426 | herbivores '(gazelle antelope zebra)) |
| 2427 | @end group |
| 2428 | @end smallexample |
| 2429 | |
| 2430 | @noindent |
| 2431 | (The expression could just as well have been on one line, but it might |
| 2432 | not have fit on a page; and humans find it easier to read nicely |
| 2433 | formatted lists.) |
| 2434 | |
| 2435 | Although I have been using the term `assign', there is another way of |
| 2436 | thinking about the workings of @code{set} and @code{setq}; and that is to |
| 2437 | say that @code{set} and @code{setq} make the symbol @emph{point} to the |
| 2438 | list. This latter way of thinking is very common and in forthcoming |
| 2439 | chapters we shall come upon at least one symbol that has `pointer' as |
| 2440 | part of its name. The name is chosen because the symbol has a value, |
| 2441 | specifically a list, attached to it; or, expressed another way, |
| 2442 | the symbol is set to ``point'' to the list. |
| 2443 | |
| 2444 | @node Counting, , Using setq, set & setq |
| 2445 | @comment node-name, next, previous, up |
| 2446 | @subsection Counting |
| 2447 | @cindex Counting |
| 2448 | |
| 2449 | Here is an example that shows how to use @code{setq} in a counter. You |
| 2450 | might use this to count how many times a part of your program repeats |
| 2451 | itself. First set a variable to zero; then add one to the number each |
| 2452 | time the program repeats itself. To do this, you need a variable that |
| 2453 | serves as a counter, and two expressions: an initial @code{setq} |
| 2454 | expression that sets the counter variable to zero; and a second |
| 2455 | @code{setq} expression that increments the counter each time it is |
| 2456 | evaluated. |
| 2457 | |
| 2458 | @smallexample |
| 2459 | @group |
| 2460 | (setq counter 0) ; @r{Let's call this the initializer.} |
| 2461 | |
| 2462 | (setq counter (+ counter 1)) ; @r{This is the incrementer.} |
| 2463 | |
| 2464 | counter ; @r{This is the counter.} |
| 2465 | @end group |
| 2466 | @end smallexample |
| 2467 | |
| 2468 | @noindent |
| 2469 | (The text following the @samp{;} are comments. @xref{Change a |
| 2470 | defun, , Change a Function Definition}.) |
| 2471 | |
| 2472 | If you evaluate the first of these expressions, the initializer, |
| 2473 | @code{(setq counter 0)}, and then evaluate the third expression, |
| 2474 | @code{counter}, the number @code{0} will appear in the echo area. If |
| 2475 | you then evaluate the second expression, the incrementer, @code{(setq |
| 2476 | counter (+ counter 1))}, the counter will get the value 1. So if you |
| 2477 | again evaluate @code{counter}, the number @code{1} will appear in the |
| 2478 | echo area. Each time you evaluate the second expression, the value of |
| 2479 | the counter will be incremented. |
| 2480 | |
| 2481 | When you evaluate the incrementer, @code{(setq counter (+ counter 1))}, |
| 2482 | the Lisp interpreter first evaluates the innermost list; this is the |
| 2483 | addition. In order to evaluate this list, it must evaluate the variable |
| 2484 | @code{counter} and the number @code{1}. When it evaluates the variable |
| 2485 | @code{counter}, it receives its current value. It passes this value and |
| 2486 | the number @code{1} to the @code{+} which adds them together. The sum |
| 2487 | is then returned as the value of the inner list and passed to the |
| 2488 | @code{setq} which sets the variable @code{counter} to this new value. |
| 2489 | Thus, the value of the variable, @code{counter}, is changed. |
| 2490 | |
| 2491 | @node Summary, Error Message Exercises, set & setq, List Processing |
| 2492 | @comment node-name, next, previous, up |
| 2493 | @section Summary |
| 2494 | |
| 2495 | Learning Lisp is like climbing a hill in which the first part is the |
| 2496 | steepest. You have now climbed the most difficult part; what remains |
| 2497 | becomes easier as you progress onwards. |
| 2498 | |
| 2499 | @need 1000 |
| 2500 | In summary, |
| 2501 | |
| 2502 | @itemize @bullet |
| 2503 | |
| 2504 | @item |
| 2505 | Lisp programs are made up of expressions, which are lists or single atoms. |
| 2506 | |
| 2507 | @item |
| 2508 | Lists are made up of zero or more atoms or inner lists, separated by whitespace and |
| 2509 | surrounded by parentheses. A list can be empty. |
| 2510 | |
| 2511 | @item |
| 2512 | Atoms are multi-character symbols, like @code{forward-paragraph}, single |
| 2513 | character symbols like @code{+}, strings of characters between double |
| 2514 | quotation marks, or numbers. |
| 2515 | |
| 2516 | @item |
| 2517 | A number evaluates to itself. |
| 2518 | |
| 2519 | @item |
| 2520 | A string between double quotes also evaluates to itself. |
| 2521 | |
| 2522 | @item |
| 2523 | When you evaluate a symbol by itself, its value is returned. |
| 2524 | |
| 2525 | @item |
| 2526 | When you evaluate a list, the Lisp interpreter looks at the first symbol |
| 2527 | in the list and then at the function definition bound to that symbol. |
| 2528 | Then the instructions in the function definition are carried out. |
| 2529 | |
| 2530 | @item |
| 2531 | A single-quote, @code{'}, tells the Lisp interpreter that it should |
| 2532 | return the following expression as written, and not evaluate it as it |
| 2533 | would if the quote were not there. |
| 2534 | |
| 2535 | @item |
| 2536 | Arguments are the information passed to a function. The arguments to a |
| 2537 | function are computed by evaluating the rest of the elements of the list |
| 2538 | of which the function is the first element. |
| 2539 | |
| 2540 | @item |
| 2541 | A function always returns a value when it is evaluated (unless it gets |
| 2542 | an error); in addition, it may also carry out some action called a |
| 2543 | ``side effect''. In many cases, a function's primary purpose is to |
| 2544 | create a side effect. |
| 2545 | @end itemize |
| 2546 | |
| 2547 | @node Error Message Exercises, , Summary, List Processing |
| 2548 | @comment node-name, next, previous, up |
| 2549 | @section Exercises |
| 2550 | |
| 2551 | A few simple exercises: |
| 2552 | |
| 2553 | @itemize @bullet |
| 2554 | @item |
| 2555 | Generate an error message by evaluating an appropriate symbol that is |
| 2556 | not within parentheses. |
| 2557 | |
| 2558 | @item |
| 2559 | Generate an error message by evaluating an appropriate symbol that is |
| 2560 | between parentheses. |
| 2561 | |
| 2562 | @item |
| 2563 | Create a counter that increments by two rather than one. |
| 2564 | |
| 2565 | @item |
| 2566 | Write an expression that prints a message in the echo area when |
| 2567 | evaluated. |
| 2568 | @end itemize |
| 2569 | |
| 2570 | @node Practicing Evaluation, Writing Defuns, List Processing, Top |
| 2571 | @comment node-name, next, previous, up |
| 2572 | @chapter Practicing Evaluation |
| 2573 | @cindex Practicing evaluation |
| 2574 | @cindex Evaluation practice |
| 2575 | |
| 2576 | Before learning how to write a function definition in Emacs Lisp, it is |
| 2577 | useful to spend a little time evaluating various expressions that have |
| 2578 | already been written. These expressions will be lists with the |
| 2579 | functions as their first (and often only) element. Since some of the |
| 2580 | functions associated with buffers are both simple and interesting, we |
| 2581 | will start with those. In this section, we will evaluate a few of |
| 2582 | these. In another section, we will study the code of several other |
| 2583 | buffer-related functions, to see how they were written. |
| 2584 | |
| 2585 | @menu |
| 2586 | * How to Evaluate:: Typing editing commands or @kbd{C-x C-e} |
| 2587 | causes evaluation. |
| 2588 | * Buffer Names:: Buffers and files are different. |
| 2589 | * Getting Buffers:: Getting a buffer itself, not merely its name. |
| 2590 | * Switching Buffers:: How to change to another buffer. |
| 2591 | * Buffer Size & Locations:: Where point is located and the size of |
| 2592 | the buffer. |
| 2593 | * Evaluation Exercise:: |
| 2594 | @end menu |
| 2595 | |
| 2596 | @node How to Evaluate, Buffer Names, Practicing Evaluation, Practicing Evaluation |
| 2597 | @ifnottex |
| 2598 | @unnumberedsec How to Evaluate |
| 2599 | @end ifnottex |
| 2600 | |
| 2601 | @i{Whenever you give an editing command} to Emacs Lisp, such as the |
| 2602 | command to move the cursor or to scroll the screen, @i{you are evaluating |
| 2603 | an expression,} the first element of which is a function. @i{This is |
| 2604 | how Emacs works.} |
| 2605 | |
| 2606 | @cindex @samp{interactive function} defined |
| 2607 | @cindex @samp{command} defined |
| 2608 | When you type keys, you cause the Lisp interpreter to evaluate an |
| 2609 | expression and that is how you get your results. Even typing plain text |
| 2610 | involves evaluating an Emacs Lisp function, in this case, one that uses |
| 2611 | @code{self-insert-command}, which simply inserts the character you |
| 2612 | typed. The functions you evaluate by typing keystrokes are called |
| 2613 | @dfn{interactive} functions, or @dfn{commands}; how you make a function |
| 2614 | interactive will be illustrated in the chapter on how to write function |
| 2615 | definitions. @xref{Interactive, , Making a Function Interactive}. |
| 2616 | |
| 2617 | In addition to typing keyboard commands, we have seen a second way to |
| 2618 | evaluate an expression: by positioning the cursor after a list and |
| 2619 | typing @kbd{C-x C-e}. This is what we will do in the rest of this |
| 2620 | section. There are other ways to evaluate an expression as well; these |
| 2621 | will be described as we come to them. |
| 2622 | |
| 2623 | Besides being used for practicing evaluation, the functions shown in the |
| 2624 | next few sections are important in their own right. A study of these |
| 2625 | functions makes clear the distinction between buffers and files, how to |
| 2626 | switch to a buffer, and how to determine a location within it. |
| 2627 | |
| 2628 | @node Buffer Names, Getting Buffers, How to Evaluate, Practicing Evaluation |
| 2629 | @comment node-name, next, previous, up |
| 2630 | @section Buffer Names |
| 2631 | @findex buffer-name |
| 2632 | @findex buffer-file-name |
| 2633 | |
| 2634 | The two functions, @code{buffer-name} and @code{buffer-file-name}, show |
| 2635 | the difference between a file and a buffer. When you evaluate the |
| 2636 | following expression, @code{(buffer-name)}, the name of the buffer |
| 2637 | appears in the echo area. When you evaluate @code{(buffer-file-name)}, |
| 2638 | the name of the file to which the buffer refers appears in the echo |
| 2639 | area. Usually, the name returned by @code{(buffer-name)} is the same as |
| 2640 | the name of the file to which it refers, and the name returned by |
| 2641 | @code{(buffer-file-name)} is the full path-name of the file. |
| 2642 | |
| 2643 | A file and a buffer are two different entities. A file is information |
| 2644 | recorded permanently in the computer (unless you delete it). A buffer, |
| 2645 | on the other hand, is information inside of Emacs that will vanish at |
| 2646 | the end of the editing session (or when you kill the buffer). Usually, |
| 2647 | a buffer contains information that you have copied from a file; we say |
| 2648 | the buffer is @dfn{visiting} that file. This copy is what you work on |
| 2649 | and modify. Changes to the buffer do not change the file, until you |
| 2650 | save the buffer. When you save the buffer, the buffer is copied to the file |
| 2651 | and is thus saved permanently. |
| 2652 | |
| 2653 | @need 1250 |
| 2654 | If you are reading this in Info inside of GNU Emacs, you can evaluate |
| 2655 | each of the following expressions by positioning the cursor after it and |
| 2656 | typing @kbd{C-x C-e}. |
| 2657 | |
| 2658 | @smallexample |
| 2659 | @group |
| 2660 | (buffer-name) |
| 2661 | |
| 2662 | (buffer-file-name) |
| 2663 | @end group |
| 2664 | @end smallexample |
| 2665 | |
| 2666 | @noindent |
| 2667 | When I do this, @file{"introduction.texinfo"} is the value returned by |
| 2668 | evaluating @code{(buffer-name)}, and |
| 2669 | @file{"/gnu/work/intro/introduction.texinfo"} is the value returned by |
| 2670 | evaluating @code{(buffer-file-name)}. The former is the name of the |
| 2671 | buffer and the latter is the name of the file. (In the expressions, the |
| 2672 | parentheses tell the Lisp interpreter to treat @code{buffer-name} and |
| 2673 | @code{buffer-file-name} as functions; without the parentheses, the |
| 2674 | interpreter would attempt to evaluate the symbols as variables. |
| 2675 | @xref{Variables}.) |
| 2676 | |
| 2677 | In spite of the distinction between files and buffers, you will often |
| 2678 | find that people refer to a file when they mean a buffer and vice-versa. |
| 2679 | Indeed, most people say, ``I am editing a file,'' rather than saying, |
| 2680 | ``I am editing a buffer which I will soon save to a file.'' It is |
| 2681 | almost always clear from context what people mean. When dealing with |
| 2682 | computer programs, however, it is important to keep the distinction in mind, |
| 2683 | since the computer is not as smart as a person. |
| 2684 | |
| 2685 | @cindex Buffer, history of word |
| 2686 | The word `buffer', by the way, comes from the meaning of the word as a |
| 2687 | cushion that deadens the force of a collision. In early computers, a |
| 2688 | buffer cushioned the interaction between files and the computer's |
| 2689 | central processing unit. The drums or tapes that held a file and the |
| 2690 | central processing unit were pieces of equipment that were very |
| 2691 | different from each other, working at their own speeds, in spurts. The |
| 2692 | buffer made it possible for them to work together effectively. |
| 2693 | Eventually, the buffer grew from being an intermediary, a temporary |
| 2694 | holding place, to being the place where work is done. This |
| 2695 | transformation is rather like that of a small seaport that grew into a |
| 2696 | great city: once it was merely the place where cargo was warehoused |
| 2697 | temporarily before being loaded onto ships; then it became a business |
| 2698 | and cultural center in its own right. |
| 2699 | |
| 2700 | Not all buffers are associated with files. For example, when you start |
| 2701 | an Emacs session by typing the command @code{emacs} alone, without |
| 2702 | naming any files, Emacs will start with the @file{*scratch*} buffer on |
| 2703 | the screen. This buffer is not visiting any file. Similarly, a |
| 2704 | @file{*Help*} buffer is not associated with any file. |
| 2705 | |
| 2706 | @cindex @code{nil}, history of word |
| 2707 | If you switch to the @file{*scratch*} buffer, type @code{(buffer-name)}, |
| 2708 | position the cursor after it, and type @kbd{C-x C-e} to evaluate the |
| 2709 | expression, the name @code{"*scratch*"} is returned and will appear in |
| 2710 | the echo area. @code{"*scratch*"} is the name of the buffer. However, |
| 2711 | if you type @code{(buffer-file-name)} in the @file{*scratch*} buffer and |
| 2712 | evaluate that, @code{nil} will appear in the echo area. @code{nil} is |
| 2713 | from the Latin word for `nothing'; in this case, it means that the |
| 2714 | @file{*scratch*} buffer is not associated with any file. (In Lisp, |
| 2715 | @code{nil} is also used to mean `false' and is a synonym for the empty |
| 2716 | list, @code{()}.) |
| 2717 | |
| 2718 | Incidentally, if you are in the @file{*scratch*} buffer and want the |
| 2719 | value returned by an expression to appear in the @file{*scratch*} |
| 2720 | buffer itself rather than in the echo area, type @kbd{C-u C-x C-e} |
| 2721 | instead of @kbd{C-x C-e}. This causes the value returned to appear |
| 2722 | after the expression. The buffer will look like this: |
| 2723 | |
| 2724 | @smallexample |
| 2725 | (buffer-name)"*scratch*" |
| 2726 | @end smallexample |
| 2727 | |
| 2728 | @noindent |
| 2729 | You cannot do this in Info since Info is read-only and it will not allow |
| 2730 | you to change the contents of the buffer. But you can do this in any |
| 2731 | buffer you can edit; and when you write code or documentation (such as |
| 2732 | this book), this feature is very useful. |
| 2733 | |
| 2734 | @node Getting Buffers, Switching Buffers, Buffer Names, Practicing Evaluation |
| 2735 | @comment node-name, next, previous, up |
| 2736 | @section Getting Buffers |
| 2737 | @findex current-buffer |
| 2738 | @findex other-buffer |
| 2739 | @cindex Getting a buffer |
| 2740 | |
| 2741 | The @code{buffer-name} function returns the @emph{name} of the buffer; |
| 2742 | to get the buffer @emph{itself}, a different function is needed: the |
| 2743 | @code{current-buffer} function. If you use this function in code, what |
| 2744 | you get is the buffer itself. |
| 2745 | |
| 2746 | A name and the object or entity to which the name refers are different |
| 2747 | from each other. You are not your name. You are a person to whom |
| 2748 | others refer by name. If you ask to speak to George and someone hands you |
| 2749 | a card with the letters @samp{G}, @samp{e}, @samp{o}, @samp{r}, |
| 2750 | @samp{g}, and @samp{e} written on it, you might be amused, but you would |
| 2751 | not be satisfied. You do not want to speak to the name, but to the |
| 2752 | person to whom the name refers. A buffer is similar: the name of the |
| 2753 | scratch buffer is @file{*scratch*}, but the name is not the buffer. To |
| 2754 | get a buffer itself, you need to use a function such as |
| 2755 | @code{current-buffer}. |
| 2756 | |
| 2757 | However, there is a slight complication: if you evaluate |
| 2758 | @code{current-buffer} in an expression on its own, as we will do here, |
| 2759 | what you see is a printed representation of the name of the buffer |
| 2760 | without the contents of the buffer. Emacs works this way for two |
| 2761 | reasons: the buffer may be thousands of lines long---too long to be |
| 2762 | conveniently displayed; and, another buffer may have the same contents |
| 2763 | but a different name, and it is important to distinguish between them. |
| 2764 | |
| 2765 | @need 800 |
| 2766 | Here is an expression containing the function: |
| 2767 | |
| 2768 | @smallexample |
| 2769 | (current-buffer) |
| 2770 | @end smallexample |
| 2771 | |
| 2772 | @noindent |
| 2773 | If you evaluate the expression in the usual way, @file{#<buffer *info*>} |
| 2774 | appears in the echo area. The special format indicates that the |
| 2775 | buffer itself is being returned, rather than just its name. |
| 2776 | |
| 2777 | Incidentally, while you can type a number or symbol into a program, you |
| 2778 | cannot do that with the printed representation of a buffer: the only way |
| 2779 | to get a buffer itself is with a function such as @code{current-buffer}. |
| 2780 | |
| 2781 | A related function is @code{other-buffer}. This returns the most |
| 2782 | recently selected buffer other than the one you are in currently. If |
| 2783 | you have recently switched back and forth from the @file{*scratch*} |
| 2784 | buffer, @code{other-buffer} will return that buffer. |
| 2785 | |
| 2786 | @need 800 |
| 2787 | You can see this by evaluating the expression: |
| 2788 | |
| 2789 | @smallexample |
| 2790 | (other-buffer) |
| 2791 | @end smallexample |
| 2792 | |
| 2793 | @noindent |
| 2794 | You should see @file{#<buffer *scratch*>} appear in the echo area, or |
| 2795 | the name of whatever other buffer you switched back from most |
| 2796 | recently@footnote{Actually, by default, if the buffer from which you |
| 2797 | just switched is visible to you in another window, @code{other-buffer} |
| 2798 | will choose the most recent buffer that you cannot see; this is a |
| 2799 | subtlety that I often forget.}. |
| 2800 | |
| 2801 | @node Switching Buffers, Buffer Size & Locations, Getting Buffers, Practicing Evaluation |
| 2802 | @comment node-name, next, previous, up |
| 2803 | @section Switching Buffers |
| 2804 | @findex switch-to-buffer |
| 2805 | @findex set-buffer |
| 2806 | @cindex Switching to a buffer |
| 2807 | |
| 2808 | The @code{other-buffer} function actually provides a buffer when it is |
| 2809 | used as an argument to a function that requires one. We can see this |
| 2810 | by using @code{other-buffer} and @code{switch-to-buffer} to switch to a |
| 2811 | different buffer. |
| 2812 | |
| 2813 | But first, a brief introduction to the @code{switch-to-buffer} |
| 2814 | function. When you switched back and forth from Info to the |
| 2815 | @file{*scratch*} buffer to evaluate @code{(buffer-name)}, you most |
| 2816 | likely typed @kbd{C-x b} and then typed @file{*scratch*}@footnote{Or |
| 2817 | rather, to save typing, you probably typed just part of the name, such |
| 2818 | as @code{*sc}, and then pressed your @kbd{TAB} key to cause it to |
| 2819 | expand to the full name; and then typed your @kbd{RET} key.} when |
| 2820 | prompted in the minibuffer for the name of the buffer to which you |
| 2821 | wanted to switch. The keystrokes, @kbd{C-x b}, cause the Lisp |
| 2822 | interpreter to evaluate the interactive function |
| 2823 | @code{switch-to-buffer}. As we said before, this is how Emacs works: |
| 2824 | different keystrokes call or run different functions. For example, |
| 2825 | @kbd{C-f} calls @code{forward-char}, @kbd{M-e} calls |
| 2826 | @code{forward-sentence}, and so on. |
| 2827 | |
| 2828 | By writing @code{switch-to-buffer} in an expression, and giving it a |
| 2829 | buffer to switch to, we can switch buffers just the way @kbd{C-x b} |
| 2830 | does. |
| 2831 | |
| 2832 | @need 1000 |
| 2833 | Here is the Lisp expression: |
| 2834 | |
| 2835 | @smallexample |
| 2836 | (switch-to-buffer (other-buffer)) |
| 2837 | @end smallexample |
| 2838 | |
| 2839 | @noindent |
| 2840 | The symbol @code{switch-to-buffer} is the first element of the list, |
| 2841 | so the Lisp interpreter will treat it as a function and carry out the |
| 2842 | instructions that are attached to it. But before doing that, the |
| 2843 | interpreter will note that @code{other-buffer} is inside parentheses |
| 2844 | and work on that symbol first. @code{other-buffer} is the first (and |
| 2845 | in this case, the only) element of this list, so the Lisp interpreter |
| 2846 | calls or runs the function. It returns another buffer. Next, the |
| 2847 | interpreter runs @code{switch-to-buffer}, passing to it, as an |
| 2848 | argument, the other buffer, which is what Emacs will switch to. If |
| 2849 | you are reading this in Info, try this now. Evaluate the expression. |
| 2850 | (To get back, type @kbd{C-x b @key{RET}}.)@footnote{Remember, this |
| 2851 | expression will move you to your most recent other buffer that you |
| 2852 | cannot see. If you really want to go to your most recently selected |
| 2853 | buffer, even if you can still see it, you need to evaluate the |
| 2854 | following more complex expression: |
| 2855 | |
| 2856 | @smallexample |
| 2857 | (switch-to-buffer (other-buffer (current-buffer) t)) |
| 2858 | @end smallexample |
| 2859 | |
| 2860 | @c noindent |
| 2861 | In this case, the first argument to @code{other-buffer} tells it which |
| 2862 | buffer to skip---the current one---and the second argument tells |
| 2863 | @code{other-buffer} it is OK to switch to a visible buffer. |
| 2864 | In regular use, @code{switch-to-buffer} takes you to an invisible |
| 2865 | window since you would most likely use @kbd{C-x o} (@code{other-window}) |
| 2866 | to go to another visible buffer.} |
| 2867 | |
| 2868 | In the programming examples in later sections of this document, you will |
| 2869 | see the function @code{set-buffer} more often than |
| 2870 | @code{switch-to-buffer}. This is because of a difference between |
| 2871 | computer programs and humans: humans have eyes and expect to see the |
| 2872 | buffer on which they are working on their computer terminals. This is |
| 2873 | so obvious, it almost goes without saying. However, programs do not |
| 2874 | have eyes. When a computer program works on a buffer, that buffer does |
| 2875 | not need to be visible on the screen. |
| 2876 | |
| 2877 | @code{switch-to-buffer} is designed for humans and does two different |
| 2878 | things: it switches the buffer to which Emacs' attention is directed; and |
| 2879 | it switches the buffer displayed in the window to the new buffer. |
| 2880 | @code{set-buffer}, on the other hand, does only one thing: it switches |
| 2881 | the attention of the computer program to a different buffer. The buffer |
| 2882 | on the screen remains unchanged (of course, normally nothing happens |
| 2883 | there until the command finishes running). |
| 2884 | |
| 2885 | @cindex @samp{call} defined |
| 2886 | Also, we have just introduced another jargon term, the word @dfn{call}. |
| 2887 | When you evaluate a list in which the first symbol is a function, you |
| 2888 | are calling that function. The use of the term comes from the notion of |
| 2889 | the function as an entity that can do something for you if you `call' |
| 2890 | it---just as a plumber is an entity who can fix a leak if you call him |
| 2891 | or her. |
| 2892 | |
| 2893 | @node Buffer Size & Locations, Evaluation Exercise, Switching Buffers, Practicing Evaluation |
| 2894 | @comment node-name, next, previous, up |
| 2895 | @section Buffer Size and the Location of Point |
| 2896 | @cindex Size of buffer |
| 2897 | @cindex Buffer size |
| 2898 | @cindex Point location |
| 2899 | @cindex Location of point |
| 2900 | |
| 2901 | Finally, let's look at several rather simple functions, |
| 2902 | @code{buffer-size}, @code{point}, @code{point-min}, and |
| 2903 | @code{point-max}. These give information about the size of a buffer and |
| 2904 | the location of point within it. |
| 2905 | |
| 2906 | The function @code{buffer-size} tells you the size of the current |
| 2907 | buffer; that is, the function returns a count of the number of |
| 2908 | characters in the buffer. |
| 2909 | |
| 2910 | @smallexample |
| 2911 | (buffer-size) |
| 2912 | @end smallexample |
| 2913 | |
| 2914 | @noindent |
| 2915 | You can evaluate this in the usual way, by positioning the |
| 2916 | cursor after the expression and typing @kbd{C-x C-e}. |
| 2917 | |
| 2918 | @cindex @samp{point} defined |
| 2919 | In Emacs, the current position of the cursor is called @dfn{point}. |
| 2920 | The expression @code{(point)} returns a number that tells you where the |
| 2921 | cursor is located as a count of the number of characters from the |
| 2922 | beginning of the buffer up to point. |
| 2923 | |
| 2924 | @need 1250 |
| 2925 | You can see the character count for point in this buffer by evaluating |
| 2926 | the following expression in the usual way: |
| 2927 | |
| 2928 | @smallexample |
| 2929 | (point) |
| 2930 | @end smallexample |
| 2931 | |
| 2932 | @noindent |
| 2933 | As I write this, the value of @code{point} is 65724. The @code{point} |
| 2934 | function is frequently used in some of the examples later in this |
| 2935 | book. |
| 2936 | |
| 2937 | @need 1250 |
| 2938 | The value of point depends, of course, on its location within the |
| 2939 | buffer. If you evaluate point in this spot, the number will be larger: |
| 2940 | |
| 2941 | @smallexample |
| 2942 | (point) |
| 2943 | @end smallexample |
| 2944 | |
| 2945 | @noindent |
| 2946 | For me, the value of point in this location is 66043, which means that |
| 2947 | there are 319 characters (including spaces) between the two expressions. |
| 2948 | |
| 2949 | @cindex @samp{narrowing} defined |
| 2950 | The function @code{point-min} is somewhat similar to @code{point}, but |
| 2951 | it returns the value of the minimum permissible value of point in the |
| 2952 | current buffer. This is the number 1 unless @dfn{narrowing} is in |
| 2953 | effect. (Narrowing is a mechanism whereby you can restrict yourself, |
| 2954 | or a program, to operations on just a part of a buffer. |
| 2955 | @xref{Narrowing & Widening, , Narrowing and Widening}.) Likewise, the |
| 2956 | function @code{point-max} returns the value of the maximum permissible |
| 2957 | value of point in the current buffer. |
| 2958 | |
| 2959 | @node Evaluation Exercise, , Buffer Size & Locations, Practicing Evaluation |
| 2960 | @section Exercise |
| 2961 | |
| 2962 | Find a file with which you are working and move towards its middle. |
| 2963 | Find its buffer name, file name, length, and your position in the file. |
| 2964 | |
| 2965 | @node Writing Defuns, Buffer Walk Through, Practicing Evaluation, Top |
| 2966 | @comment node-name, next, previous, up |
| 2967 | @chapter How To Write Function Definitions |
| 2968 | @cindex Definition writing |
| 2969 | @cindex Function definition writing |
| 2970 | @cindex Writing a function definition |
| 2971 | |
| 2972 | When the Lisp interpreter evaluates a list, it looks to see whether the |
| 2973 | first symbol on the list has a function definition attached to it; or, |
| 2974 | put another way, whether the symbol points to a function definition. If |
| 2975 | it does, the computer carries out the instructions in the definition. A |
| 2976 | symbol that has a function definition is called, simply, a function |
| 2977 | (although, properly speaking, the definition is the function and the |
| 2978 | symbol refers to it.) |
| 2979 | |
| 2980 | @menu |
| 2981 | * Primitive Functions:: |
| 2982 | * defun:: The @code{defun} special form. |
| 2983 | * Install:: Install a function definition. |
| 2984 | * Interactive:: Making a function interactive. |
| 2985 | * Interactive Options:: Different options for @code{interactive}. |
| 2986 | * Permanent Installation:: Installing code permanently. |
| 2987 | * let:: Creating and initializing local variables. |
| 2988 | * if:: What if? |
| 2989 | * else:: If--then--else expressions. |
| 2990 | * Truth & Falsehood:: What Lisp considers false and true. |
| 2991 | * save-excursion:: Keeping track of point, mark, and buffer. |
| 2992 | * Review:: |
| 2993 | * defun Exercises:: |
| 2994 | @end menu |
| 2995 | |
| 2996 | @node Primitive Functions, defun, Writing Defuns, Writing Defuns |
| 2997 | @ifnottex |
| 2998 | @unnumberedsec An Aside about Primitive Functions |
| 2999 | @end ifnottex |
| 3000 | @cindex Primitive functions |
| 3001 | @cindex Functions, primitive |
| 3002 | |
| 3003 | @cindex C language primitives |
| 3004 | @cindex Primitives written in C |
| 3005 | All functions are defined in terms of other functions, except for a few |
| 3006 | @dfn{primitive} functions that are written in the C programming |
| 3007 | language. When you write functions' definitions, you will write them in |
| 3008 | Emacs Lisp and use other functions as your building blocks. Some of the |
| 3009 | functions you will use will themselves be written in Emacs Lisp (perhaps |
| 3010 | by you) and some will be primitives written in C. The primitive |
| 3011 | functions are used exactly like those written in Emacs Lisp and behave |
| 3012 | like them. They are written in C so we can easily run GNU Emacs on any |
| 3013 | computer that has sufficient power and can run C. |
| 3014 | |
| 3015 | Let me re-emphasize this: when you write code in Emacs Lisp, you do not |
| 3016 | distinguish between the use of functions written in C and the use of |
| 3017 | functions written in Emacs Lisp. The difference is irrelevant. I |
| 3018 | mention the distinction only because it is interesting to know. Indeed, |
| 3019 | unless you investigate, you won't know whether an already-written |
| 3020 | function is written in Emacs Lisp or C. |
| 3021 | |
| 3022 | @node defun, Install, Primitive Functions, Writing Defuns |
| 3023 | @comment node-name, next, previous, up |
| 3024 | @section The @code{defun} Special Form |
| 3025 | @findex defun |
| 3026 | @cindex Special form of @code{defun} |
| 3027 | |
| 3028 | @cindex @samp{function definition} defined |
| 3029 | In Lisp, a symbol such as @code{mark-whole-buffer} has code attached to |
| 3030 | it that tells the computer what to do when the function is called. |
| 3031 | This code is called the @dfn{function definition} and is created by |
| 3032 | evaluating a Lisp expression that starts with the symbol @code{defun} |
| 3033 | (which is an abbreviation for @emph{define function}). Because |
| 3034 | @code{defun} does not evaluate its arguments in the usual way, it is |
| 3035 | called a @dfn{special form}. |
| 3036 | |
| 3037 | In subsequent sections, we will look at function definitions from the |
| 3038 | Emacs source code, such as @code{mark-whole-buffer}. In this section, |
| 3039 | we will describe a simple function definition so you can see how it |
| 3040 | looks. This function definition uses arithmetic because it makes for a |
| 3041 | simple example. Some people dislike examples using arithmetic; however, |
| 3042 | if you are such a person, do not despair. Hardly any of the code we |
| 3043 | will study in the remainder of this introduction involves arithmetic or |
| 3044 | mathematics. The examples mostly involve text in one way or another. |
| 3045 | |
| 3046 | A function definition has up to five parts following the word |
| 3047 | @code{defun}: |
| 3048 | |
| 3049 | @enumerate |
| 3050 | @item |
| 3051 | The name of the symbol to which the function definition should be |
| 3052 | attached. |
| 3053 | |
| 3054 | @item |
| 3055 | A list of the arguments that will be passed to the function. If no |
| 3056 | arguments will be passed to the function, this is an empty list, |
| 3057 | @code{()}. |
| 3058 | |
| 3059 | @item |
| 3060 | Documentation describing the function. (Technically optional, but |
| 3061 | strongly recommended.) |
| 3062 | |
| 3063 | @item |
| 3064 | Optionally, an expression to make the function interactive so you can |
| 3065 | use it by typing @kbd{M-x} and then the name of the function; or by |
| 3066 | typing an appropriate key or keychord. |
| 3067 | |
| 3068 | @cindex @samp{body} defined |
| 3069 | @item |
| 3070 | The code that instructs the computer what to do: the @dfn{body} of the |
| 3071 | function definition. |
| 3072 | @end enumerate |
| 3073 | |
| 3074 | It is helpful to think of the five parts of a function definition as |
| 3075 | being organized in a template, with slots for each part: |
| 3076 | |
| 3077 | @smallexample |
| 3078 | @group |
| 3079 | (defun @var{function-name} (@var{arguments}@dots{}) |
| 3080 | "@var{optional-documentation}@dots{}" |
| 3081 | (interactive @var{argument-passing-info}) ; @r{optional} |
| 3082 | @var{body}@dots{}) |
| 3083 | @end group |
| 3084 | @end smallexample |
| 3085 | |
| 3086 | As an example, here is the code for a function that multiplies its |
| 3087 | argument by 7. (This example is not interactive. @xref{Interactive, |
| 3088 | , Making a Function Interactive}, for that information.) |
| 3089 | |
| 3090 | @smallexample |
| 3091 | @group |
| 3092 | (defun multiply-by-seven (number) |
| 3093 | "Multiply NUMBER by seven." |
| 3094 | (* 7 number)) |
| 3095 | @end group |
| 3096 | @end smallexample |
| 3097 | |
| 3098 | This definition begins with a parenthesis and the symbol @code{defun}, |
| 3099 | followed by the name of the function. |
| 3100 | |
| 3101 | @cindex @samp{argument list} defined |
| 3102 | The name of the function is followed by a list that contains the |
| 3103 | arguments that will be passed to the function. This list is called |
| 3104 | the @dfn{argument list}. In this example, the list has only one |
| 3105 | element, the symbol, @code{number}. When the function is used, the |
| 3106 | symbol will be bound to the value that is used as the argument to the |
| 3107 | function. |
| 3108 | |
| 3109 | Instead of choosing the word @code{number} for the name of the argument, |
| 3110 | I could have picked any other name. For example, I could have chosen |
| 3111 | the word @code{multiplicand}. I picked the word `number' because it |
| 3112 | tells what kind of value is intended for this slot; but I could just as |
| 3113 | well have chosen the word `multiplicand' to indicate the role that the |
| 3114 | value placed in this slot will play in the workings of the function. I |
| 3115 | could have called it @code{foogle}, but that would have been a bad |
| 3116 | choice because it would not tell humans what it means. The choice of |
| 3117 | name is up to the programmer and should be chosen to make the meaning of |
| 3118 | the function clear. |
| 3119 | |
| 3120 | Indeed, you can choose any name you wish for a symbol in an argument |
| 3121 | list, even the name of a symbol used in some other function: the name |
| 3122 | you use in an argument list is private to that particular definition. |
| 3123 | In that definition, the name refers to a different entity than any use |
| 3124 | of the same name outside the function definition. Suppose you have a |
| 3125 | nick-name `Shorty' in your family; when your family members refer to |
| 3126 | `Shorty', they mean you. But outside your family, in a movie, for |
| 3127 | example, the name `Shorty' refers to someone else. Because a name in an |
| 3128 | argument list is private to the function definition, you can change the |
| 3129 | value of such a symbol inside the body of a function without changing |
| 3130 | its value outside the function. The effect is similar to that produced |
| 3131 | by a @code{let} expression. (@xref{let, , @code{let}}.) |
| 3132 | |
| 3133 | @ignore |
| 3134 | Note also that we discuss the word `number' in two different ways: as a |
| 3135 | symbol that appears in the code, and as the name of something that will |
| 3136 | be replaced by a something else during the evaluation of the function. |
| 3137 | In the first case, @code{number} is a symbol, not a number; it happens |
| 3138 | that within the function, it is a variable who value is the number in |
| 3139 | question, but our primary interest in it is as a symbol. On the other |
| 3140 | hand, when we are talking about the function, our interest is that we |
| 3141 | will substitute a number for the word @var{number}. To keep this |
| 3142 | distinction clear, we use different typography for the two |
| 3143 | circumstances. When we talk about this function, or about how it works, |
| 3144 | we refer to this number by writing @var{number}. In the function |
| 3145 | itself, we refer to it by writing @code{number}. |
| 3146 | @end ignore |
| 3147 | |
| 3148 | The argument list is followed by the documentation string that |
| 3149 | describes the function. This is what you see when you type |
| 3150 | @w{@kbd{C-h f}} and the name of a function. Incidentally, when you |
| 3151 | write a documentation string like this, you should make the first line |
| 3152 | a complete sentence since some commands, such as @code{apropos}, print |
| 3153 | only the first line of a multi-line documentation string. Also, you |
| 3154 | should not indent the second line of a documentation string, if you |
| 3155 | have one, because that looks odd when you use @kbd{C-h f} |
| 3156 | (@code{describe-function}). The documentation string is optional, but |
| 3157 | it is so useful, it should be included in almost every function you |
| 3158 | write. |
| 3159 | |
| 3160 | @findex * @r{(multiplication)} |
| 3161 | The third line of the example consists of the body of the function |
| 3162 | definition. (Most functions' definitions, of course, are longer than |
| 3163 | this.) In this function, the body is the list, @code{(* 7 number)}, which |
| 3164 | says to multiply the value of @var{number} by 7. (In Emacs Lisp, |
| 3165 | @code{*} is the function for multiplication, just as @code{+} is the |
| 3166 | function for addition.) |
| 3167 | |
| 3168 | When you use the @code{multiply-by-seven} function, the argument |
| 3169 | @code{number} evaluates to the actual number you want used. Here is an |
| 3170 | example that shows how @code{multiply-by-seven} is used; but don't try |
| 3171 | to evaluate this yet! |
| 3172 | |
| 3173 | @smallexample |
| 3174 | (multiply-by-seven 3) |
| 3175 | @end smallexample |
| 3176 | |
| 3177 | @noindent |
| 3178 | The symbol @code{number}, specified in the function definition in the |
| 3179 | next section, is given or ``bound to'' the value 3 in the actual use of |
| 3180 | the function. Note that although @code{number} was inside parentheses |
| 3181 | in the function definition, the argument passed to the |
| 3182 | @code{multiply-by-seven} function is not in parentheses. The |
| 3183 | parentheses are written in the function definition so the computer can |
| 3184 | figure out where the argument list ends and the rest of the function |
| 3185 | definition begins. |
| 3186 | |
| 3187 | If you evaluate this example, you are likely to get an error message. |
| 3188 | (Go ahead, try it!) This is because we have written the function |
| 3189 | definition, but not yet told the computer about the definition---we have |
| 3190 | not yet installed (or `loaded') the function definition in Emacs. |
| 3191 | Installing a function is the process that tells the Lisp interpreter the |
| 3192 | definition of the function. Installation is described in the next |
| 3193 | section. |
| 3194 | |
| 3195 | @node Install, Interactive, defun, Writing Defuns |
| 3196 | @comment node-name, next, previous, up |
| 3197 | @section Install a Function Definition |
| 3198 | @cindex Install a Function Definition |
| 3199 | @cindex Definition installation |
| 3200 | @cindex Function definition installation |
| 3201 | |
| 3202 | If you are reading this inside of Info in Emacs, you can try out the |
| 3203 | @code{multiply-by-seven} function by first evaluating the function |
| 3204 | definition and then evaluating @code{(multiply-by-seven 3)}. A copy of |
| 3205 | the function definition follows. Place the cursor after the last |
| 3206 | parenthesis of the function definition and type @kbd{C-x C-e}. When you |
| 3207 | do this, @code{multiply-by-seven} will appear in the echo area. (What |
| 3208 | this means is that when a function definition is evaluated, the value it |
| 3209 | returns is the name of the defined function.) At the same time, this |
| 3210 | action installs the function definition. |
| 3211 | |
| 3212 | @smallexample |
| 3213 | @group |
| 3214 | (defun multiply-by-seven (number) |
| 3215 | "Multiply NUMBER by seven." |
| 3216 | (* 7 number)) |
| 3217 | @end group |
| 3218 | @end smallexample |
| 3219 | |
| 3220 | @noindent |
| 3221 | By evaluating this @code{defun}, you have just installed |
| 3222 | @code{multiply-by-seven} in Emacs. The function is now just as much a |
| 3223 | part of Emacs as @code{forward-word} or any other editing function you |
| 3224 | use. (@code{multiply-by-seven} will stay installed until you quit |
| 3225 | Emacs. To reload code automatically whenever you start Emacs, see |
| 3226 | @ref{Permanent Installation, , Installing Code Permanently}.) |
| 3227 | |
| 3228 | |
| 3229 | @menu |
| 3230 | * Effect of installation:: |
| 3231 | * Change a defun:: How to change a function definition. |
| 3232 | @end menu |
| 3233 | |
| 3234 | @node Effect of installation, Change a defun, Install, Install |
| 3235 | @ifnottex |
| 3236 | @unnumberedsubsec The effect of installation |
| 3237 | @end ifnottex |
| 3238 | |
| 3239 | |
| 3240 | You can see the effect of installing @code{multiply-by-seven} by |
| 3241 | evaluating the following sample. Place the cursor after the following |
| 3242 | expression and type @kbd{C-x C-e}. The number 21 will appear in the |
| 3243 | echo area. |
| 3244 | |
| 3245 | @smallexample |
| 3246 | (multiply-by-seven 3) |
| 3247 | @end smallexample |
| 3248 | |
| 3249 | If you wish, you can read the documentation for the function by typing |
| 3250 | @kbd{C-h f} (@code{describe-function}) and then the name of the |
| 3251 | function, @code{multiply-by-seven}. When you do this, a |
| 3252 | @file{*Help*} window will appear on your screen that says: |
| 3253 | |
| 3254 | @smallexample |
| 3255 | @group |
| 3256 | multiply-by-seven: |
| 3257 | Multiply NUMBER by seven. |
| 3258 | @end group |
| 3259 | @end smallexample |
| 3260 | |
| 3261 | @noindent |
| 3262 | (To return to a single window on your screen, type @kbd{C-x 1}.) |
| 3263 | |
| 3264 | @node Change a defun, , Effect of installation, Install |
| 3265 | @comment node-name, next, previous, up |
| 3266 | @subsection Change a Function Definition |
| 3267 | @cindex Changing a function definition |
| 3268 | @cindex Function definition, how to change |
| 3269 | @cindex Definition, how to change |
| 3270 | |
| 3271 | If you want to change the code in @code{multiply-by-seven}, just rewrite |
| 3272 | it. To install the new version in place of the old one, evaluate the |
| 3273 | function definition again. This is how you modify code in Emacs. It is |
| 3274 | very simple. |
| 3275 | |
| 3276 | As an example, you can change the @code{multiply-by-seven} function to |
| 3277 | add the number to itself seven times instead of multiplying the number |
| 3278 | by seven. It produces the same answer, but by a different path. At |
| 3279 | the same time, we will add a comment to the code; a comment is text |
| 3280 | that the Lisp interpreter ignores, but that a human reader may find |
| 3281 | useful or enlightening. The comment is that this is the ``second |
| 3282 | version''. |
| 3283 | |
| 3284 | @smallexample |
| 3285 | @group |
| 3286 | (defun multiply-by-seven (number) ; @r{Second version.} |
| 3287 | "Multiply NUMBER by seven." |
| 3288 | (+ number number number number number number number)) |
| 3289 | @end group |
| 3290 | @end smallexample |
| 3291 | |
| 3292 | @cindex Comments in Lisp code |
| 3293 | The comment follows a semicolon, @samp{;}. In Lisp, everything on a |
| 3294 | line that follows a semicolon is a comment. The end of the line is the |
| 3295 | end of the comment. To stretch a comment over two or more lines, begin |
| 3296 | each line with a semicolon. |
| 3297 | |
| 3298 | @xref{Beginning a .emacs File, , Beginning a @file{.emacs} |
| 3299 | File}, and @ref{Comments, , Comments, elisp, The GNU Emacs Lisp |
| 3300 | Reference Manual}, for more about comments. |
| 3301 | |
| 3302 | You can install this version of the @code{multiply-by-seven} function by |
| 3303 | evaluating it in the same way you evaluated the first function: place |
| 3304 | the cursor after the last parenthesis and type @kbd{C-x C-e}. |
| 3305 | |
| 3306 | In summary, this is how you write code in Emacs Lisp: you write a |
| 3307 | function; install it; test it; and then make fixes or enhancements and |
| 3308 | install it again. |
| 3309 | |
| 3310 | @node Interactive, Interactive Options, Install, Writing Defuns |
| 3311 | @comment node-name, next, previous, up |
| 3312 | @section Make a Function Interactive |
| 3313 | @cindex Interactive functions |
| 3314 | @findex interactive |
| 3315 | |
| 3316 | You make a function interactive by placing a list that begins with |
| 3317 | the special form @code{interactive} immediately after the |
| 3318 | documentation. A user can invoke an interactive function by typing |
| 3319 | @kbd{M-x} and then the name of the function; or by typing the keys to |
| 3320 | which it is bound, for example, by typing @kbd{C-n} for |
| 3321 | @code{next-line} or @kbd{C-x h} for @code{mark-whole-buffer}. |
| 3322 | |
| 3323 | Interestingly, when you call an interactive function interactively, |
| 3324 | the value returned is not automatically displayed in the echo area. |
| 3325 | This is because you often call an interactive function for its side |
| 3326 | effects, such as moving forward by a word or line, and not for the |
| 3327 | value returned. If the returned value were displayed in the echo area |
| 3328 | each time you typed a key, it would be very distracting. |
| 3329 | |
| 3330 | @menu |
| 3331 | * Interactive multiply-by-seven:: An overview. |
| 3332 | * multiply-by-seven in detail:: The interactive version. |
| 3333 | @end menu |
| 3334 | |
| 3335 | @node Interactive multiply-by-seven, multiply-by-seven in detail, Interactive, Interactive |
| 3336 | @ifnottex |
| 3337 | @unnumberedsubsec An Interactive @code{multiply-by-seven}, An Overview |
| 3338 | @end ifnottex |
| 3339 | |
| 3340 | Both the use of the special form @code{interactive} and one way to |
| 3341 | display a value in the echo area can be illustrated by creating an |
| 3342 | interactive version of @code{multiply-by-seven}. |
| 3343 | |
| 3344 | @need 1250 |
| 3345 | Here is the code: |
| 3346 | |
| 3347 | @smallexample |
| 3348 | @group |
| 3349 | (defun multiply-by-seven (number) ; @r{Interactive version.} |
| 3350 | "Multiply NUMBER by seven." |
| 3351 | (interactive "p") |
| 3352 | (message "The result is %d" (* 7 number))) |
| 3353 | @end group |
| 3354 | @end smallexample |
| 3355 | |
| 3356 | @noindent |
| 3357 | You can install this code by placing your cursor after it and typing |
| 3358 | @kbd{C-x C-e}. The name of the function will appear in your echo area. |
| 3359 | Then, you can use this code by typing @kbd{C-u} and a number and then |
| 3360 | typing @kbd{M-x multiply-by-seven} and pressing @key{RET}. The phrase |
| 3361 | @samp{The result is @dots{}} followed by the product will appear in the |
| 3362 | echo area. |
| 3363 | |
| 3364 | Speaking more generally, you invoke a function like this in either of two |
| 3365 | ways: |
| 3366 | |
| 3367 | @enumerate |
| 3368 | @item |
| 3369 | By typing a prefix argument that contains the number to be passed, and |
| 3370 | then typing @kbd{M-x} and the name of the function, as with |
| 3371 | @kbd{C-u 3 M-x forward-sentence}; or, |
| 3372 | |
| 3373 | @item |
| 3374 | By typing whatever key or keychord the function is bound to, as with |
| 3375 | @kbd{C-u 3 M-e}. |
| 3376 | @end enumerate |
| 3377 | |
| 3378 | @noindent |
| 3379 | Both the examples just mentioned work identically to move point forward |
| 3380 | three sentences. (Since @code{multiply-by-seven} is not bound to a key, |
| 3381 | it could not be used as an example of key binding.) |
| 3382 | |
| 3383 | (@xref{Keybindings, , Some Keybindings}, to learn how to bind a command |
| 3384 | to a key.) |
| 3385 | |
| 3386 | A prefix argument is passed to an interactive function by typing the |
| 3387 | @key{META} key followed by a number, for example, @kbd{M-3 M-e}, or by |
| 3388 | typing @kbd{C-u} and then a number, for example, @kbd{C-u 3 M-e} (if you |
| 3389 | type @kbd{C-u} without a number, it defaults to 4). |
| 3390 | |
| 3391 | @node multiply-by-seven in detail, , Interactive multiply-by-seven, Interactive |
| 3392 | @comment node-name, next, previous, up |
| 3393 | @subsection An Interactive @code{multiply-by-seven} |
| 3394 | |
| 3395 | Let's look at the use of the special form @code{interactive} and then at |
| 3396 | the function @code{message} in the interactive version of |
| 3397 | @code{multiply-by-seven}. You will recall that the function definition |
| 3398 | looks like this: |
| 3399 | |
| 3400 | @smallexample |
| 3401 | @group |
| 3402 | (defun multiply-by-seven (number) ; @r{Interactive version.} |
| 3403 | "Multiply NUMBER by seven." |
| 3404 | (interactive "p") |
| 3405 | (message "The result is %d" (* 7 number))) |
| 3406 | @end group |
| 3407 | @end smallexample |
| 3408 | |
| 3409 | In this function, the expression, @code{(interactive "p")}, is a list of |
| 3410 | two elements. The @code{"p"} tells Emacs to pass the prefix argument to |
| 3411 | the function and use its value for the argument of the function. |
| 3412 | |
| 3413 | @need 1000 |
| 3414 | The argument will be a number. This means that the symbol |
| 3415 | @code{number} will be bound to a number in the line: |
| 3416 | |
| 3417 | @smallexample |
| 3418 | (message "The result is %d" (* 7 number)) |
| 3419 | @end smallexample |
| 3420 | |
| 3421 | @need 1250 |
| 3422 | @noindent |
| 3423 | For example, if your prefix argument is 5, the Lisp interpreter will |
| 3424 | evaluate the line as if it were: |
| 3425 | |
| 3426 | @smallexample |
| 3427 | (message "The result is %d" (* 7 5)) |
| 3428 | @end smallexample |
| 3429 | |
| 3430 | @noindent |
| 3431 | (If you are reading this in GNU Emacs, you can evaluate this expression |
| 3432 | yourself.) First, the interpreter will evaluate the inner list, which |
| 3433 | is @code{(* 7 5)}. This returns a value of 35. Next, it |
| 3434 | will evaluate the outer list, passing the values of the second and |
| 3435 | subsequent elements of the list to the function @code{message}. |
| 3436 | |
| 3437 | As we have seen, @code{message} is an Emacs Lisp function especially |
| 3438 | designed for sending a one line message to a user. (@xref{message, , The |
| 3439 | @code{message} function}.) |
| 3440 | In summary, the @code{message} function prints its first argument in the |
| 3441 | echo area as is, except for occurrences of @samp{%d}, @samp{%s}, or |
| 3442 | @samp{%c}. When it sees one of these control sequences, the function |
| 3443 | looks to the second and subsequent arguments and prints the value of the |
| 3444 | argument in the location in the string where the control sequence is |
| 3445 | located. |
| 3446 | |
| 3447 | In the interactive @code{multiply-by-seven} function, the control string |
| 3448 | is @samp{%d}, which requires a number, and the value returned by |
| 3449 | evaluating @code{(* 7 5)} is the number 35. Consequently, the number 35 |
| 3450 | is printed in place of the @samp{%d} and the message is @samp{The result |
| 3451 | is 35}. |
| 3452 | |
| 3453 | (Note that when you call the function @code{multiply-by-seven}, the |
| 3454 | message is printed without quotes, but when you call @code{message}, the |
| 3455 | text is printed in double quotes. This is because the value returned by |
| 3456 | @code{message} is what appears in the echo area when you evaluate an |
| 3457 | expression whose first element is @code{message}; but when embedded in a |
| 3458 | function, @code{message} prints the text as a side effect without |
| 3459 | quotes.) |
| 3460 | |
| 3461 | @node Interactive Options, Permanent Installation, Interactive, Writing Defuns |
| 3462 | @comment node-name, next, previous, up |
| 3463 | @section Different Options for @code{interactive} |
| 3464 | @cindex Options for @code{interactive} |
| 3465 | @cindex Interactive options |
| 3466 | |
| 3467 | In the example, @code{multiply-by-seven} used @code{"p"} as the |
| 3468 | argument to @code{interactive}. This argument told Emacs to interpret |
| 3469 | your typing either @kbd{C-u} followed by a number or @key{META} |
| 3470 | followed by a number as a command to pass that number to the function |
| 3471 | as its argument. Emacs has more than twenty characters predefined for |
| 3472 | use with @code{interactive}. In almost every case, one of these |
| 3473 | options will enable you to pass the right information interactively to |
| 3474 | a function. (@xref{Interactive Codes, , Code Characters for |
| 3475 | @code{interactive}, elisp, The GNU Emacs Lisp Reference Manual}.) |
| 3476 | |
| 3477 | @need 1250 |
| 3478 | For example, the character @samp{r} causes Emacs to pass the beginning |
| 3479 | and end of the region (the current values of point and mark) to the |
| 3480 | function as two separate arguments. It is used as follows: |
| 3481 | |
| 3482 | @smallexample |
| 3483 | (interactive "r") |
| 3484 | @end smallexample |
| 3485 | |
| 3486 | On the other hand, a @samp{B} tells Emacs to ask for the name of a |
| 3487 | buffer that will be passed to the function. When it sees a @samp{B}, |
| 3488 | Emacs will ask for the name by prompting the user in the minibuffer, |
| 3489 | using a string that follows the @samp{B}, as in @code{"BAppend to |
| 3490 | buffer:@: "}. Not only will Emacs prompt for the name, but Emacs will |
| 3491 | complete the name if you type enough of it and press @key{TAB}. |
| 3492 | |
| 3493 | A function with two or more arguments can have information passed to |
| 3494 | each argument by adding parts to the string that follows |
| 3495 | @code{interactive}. When you do this, the information is passed to |
| 3496 | each argument in the same order it is specified in the |
| 3497 | @code{interactive} list. In the string, each part is separated from |
| 3498 | the next part by a @samp{\n}, which is a newline. For example, you |
| 3499 | could follow @code{"BAppend to buffer:@: "} with a @samp{\n} and an |
| 3500 | @samp{r}. This would cause Emacs to pass the values of point and mark |
| 3501 | to the function as well as prompt you for the buffer---three arguments |
| 3502 | in all. |
| 3503 | |
| 3504 | In this case, the function definition would look like the following, |
| 3505 | where @code{buffer}, @code{start}, and @code{end} are the symbols to |
| 3506 | which @code{interactive} binds the buffer and the current values of the |
| 3507 | beginning and ending of the region: |
| 3508 | |
| 3509 | @smallexample |
| 3510 | @group |
| 3511 | (defun @var{name-of-function} (buffer start end) |
| 3512 | "@var{documentation}@dots{}" |
| 3513 | (interactive "BAppend to buffer:@: \nr") |
| 3514 | @var{body-of-function}@dots{}) |
| 3515 | @end group |
| 3516 | @end smallexample |
| 3517 | |
| 3518 | @noindent |
| 3519 | (The space after the colon in the prompt makes it look better when you |
| 3520 | are prompted. The @code{append-to-buffer} function looks exactly like |
| 3521 | this. @xref{append-to-buffer, , The Definition of |
| 3522 | @code{append-to-buffer}}.) |
| 3523 | |
| 3524 | If a function does not have arguments, then @code{interactive} does not |
| 3525 | require any. Such a function contains the simple expression |
| 3526 | @code{(interactive)}. The @code{mark-whole-buffer} function is like |
| 3527 | this. |
| 3528 | |
| 3529 | Alternatively, if the special letter-codes are not right for your |
| 3530 | application, you can pass your own arguments to @code{interactive} as |
| 3531 | a list. @xref{Using Interactive, , Using @code{Interactive}, elisp, The |
| 3532 | GNU Emacs Lisp Reference Manual}, for more information about this advanced |
| 3533 | technique. |
| 3534 | |
| 3535 | @node Permanent Installation, let, Interactive Options, Writing Defuns |
| 3536 | @comment node-name, next, previous, up |
| 3537 | @section Install Code Permanently |
| 3538 | @cindex Install code permanently |
| 3539 | @cindex Permanent code installation |
| 3540 | @cindex Code installation |
| 3541 | |
| 3542 | When you install a function definition by evaluating it, it will stay |
| 3543 | installed until you quit Emacs. The next time you start a new session |
| 3544 | of Emacs, the function will not be installed unless you evaluate the |
| 3545 | function definition again. |
| 3546 | |
| 3547 | At some point, you may want to have code installed automatically |
| 3548 | whenever you start a new session of Emacs. There are several ways of |
| 3549 | doing this: |
| 3550 | |
| 3551 | @itemize @bullet |
| 3552 | @item |
| 3553 | If you have code that is just for yourself, you can put the code for the |
| 3554 | function definition in your @file{.emacs} initialization file. When you |
| 3555 | start Emacs, your @file{.emacs} file is automatically evaluated and all |
| 3556 | the function definitions within it are installed. |
| 3557 | @xref{Emacs Initialization, , Your @file{.emacs} File}. |
| 3558 | |
| 3559 | @item |
| 3560 | Alternatively, you can put the function definitions that you want |
| 3561 | installed in one or more files of their own and use the @code{load} |
| 3562 | function to cause Emacs to evaluate and thereby install each of the |
| 3563 | functions in the files. |
| 3564 | @xref{Loading Files, , Loading Files}. |
| 3565 | |
| 3566 | @item |
| 3567 | On the other hand, if you have code that your whole site will use, it |
| 3568 | is usual to put it in a file called @file{site-init.el} that is loaded |
| 3569 | when Emacs is built. This makes the code available to everyone who |
| 3570 | uses your machine. (See the @file{INSTALL} file that is part of the |
| 3571 | Emacs distribution.) |
| 3572 | @end itemize |
| 3573 | |
| 3574 | Finally, if you have code that everyone who uses Emacs may want, you |
| 3575 | can post it on a computer network or send a copy to the Free Software |
| 3576 | Foundation. (When you do this, please license the code and its |
| 3577 | documentation under a license that permits other people to run, copy, |
| 3578 | study, modify, and redistribute the code and which protects you from |
| 3579 | having your work taken from you.) If you send a copy of your code to |
| 3580 | the Free Software Foundation, and properly protect yourself and |
| 3581 | others, it may be included in the next release of Emacs. In large |
| 3582 | part, this is how Emacs has grown over the past years, by donations. |
| 3583 | |
| 3584 | @node let, if, Permanent Installation, Writing Defuns |
| 3585 | @comment node-name, next, previous, up |
| 3586 | @section @code{let} |
| 3587 | @findex let |
| 3588 | |
| 3589 | The @code{let} expression is a special form in Lisp that you will need |
| 3590 | to use in most function definitions. |
| 3591 | |
| 3592 | @code{let} is used to attach or bind a symbol to a value in such a way |
| 3593 | that the Lisp interpreter will not confuse the variable with a |
| 3594 | variable of the same name that is not part of the function. |
| 3595 | |
| 3596 | To understand why the @code{let} special form is necessary, consider |
| 3597 | the situation in which you own a home that you generally refer to as |
| 3598 | `the house', as in the sentence, ``The house needs painting.'' If you |
| 3599 | are visiting a friend and your host refers to `the house', he is |
| 3600 | likely to be referring to @emph{his} house, not yours, that is, to a |
| 3601 | different house. |
| 3602 | |
| 3603 | If your friend is referring to his house and you think he is referring |
| 3604 | to your house, you may be in for some confusion. The same thing could |
| 3605 | happen in Lisp if a variable that is used inside of one function has |
| 3606 | the same name as a variable that is used inside of another function, |
| 3607 | and the two are not intended to refer to the same value. The |
| 3608 | @code{let} special form prevents this kind of confusion. |
| 3609 | |
| 3610 | @menu |
| 3611 | * Prevent confusion:: |
| 3612 | * Parts of let Expression:: |
| 3613 | * Sample let Expression:: |
| 3614 | * Uninitialized let Variables:: |
| 3615 | @end menu |
| 3616 | |
| 3617 | @node Prevent confusion, Parts of let Expression, let, let |
| 3618 | @ifnottex |
| 3619 | @unnumberedsubsec @code{let} Prevents Confusion |
| 3620 | @end ifnottex |
| 3621 | |
| 3622 | @cindex @samp{local variable} defined |
| 3623 | The @code{let} special form prevents confusion. @code{let} creates a |
| 3624 | name for a @dfn{local variable} that overshadows any use of the same |
| 3625 | name outside the @code{let} expression. This is like understanding |
| 3626 | that whenever your host refers to `the house', he means his house, not |
| 3627 | yours. (Symbols used in argument lists work the same way. |
| 3628 | @xref{defun, , The @code{defun} Special Form}.) |
| 3629 | |
| 3630 | Local variables created by a @code{let} expression retain their value |
| 3631 | @emph{only} within the @code{let} expression itself (and within |
| 3632 | expressions called within the @code{let} expression); the local |
| 3633 | variables have no effect outside the @code{let} expression. |
| 3634 | |
| 3635 | Another way to think about @code{let} is that it is like a @code{setq} |
| 3636 | that is temporary and local. The values set by @code{let} are |
| 3637 | automatically undone when the @code{let} is finished. The setting |
| 3638 | only affects expressions that are inside the bounds of the @code{let} |
| 3639 | expression. In computer science jargon, we would say ``the binding of |
| 3640 | a symbol is visible only in functions called in the @code{let} form; |
| 3641 | in Emacs Lisp, scoping is dynamic, not lexical.'' |
| 3642 | |
| 3643 | @code{let} can create more than one variable at once. Also, |
| 3644 | @code{let} gives each variable it creates an initial value, either a |
| 3645 | value specified by you, or @code{nil}. (In the jargon, this is called |
| 3646 | `binding the variable to the value'.) After @code{let} has created |
| 3647 | and bound the variables, it executes the code in the body of the |
| 3648 | @code{let}, and returns the value of the last expression in the body, |
| 3649 | as the value of the whole @code{let} expression. (`Execute' is a jargon |
| 3650 | term that means to evaluate a list; it comes from the use of the word |
| 3651 | meaning `to give practical effect to' (@cite{Oxford English |
| 3652 | Dictionary}). Since you evaluate an expression to perform an action, |
| 3653 | `execute' has evolved as a synonym to `evaluate'.) |
| 3654 | |
| 3655 | @node Parts of let Expression, Sample let Expression, Prevent confusion, let |
| 3656 | @comment node-name, next, previous, up |
| 3657 | @subsection The Parts of a @code{let} Expression |
| 3658 | @cindex @code{let} expression, parts of |
| 3659 | @cindex Parts of @code{let} expression |
| 3660 | |
| 3661 | @cindex @samp{varlist} defined |
| 3662 | A @code{let} expression is a list of three parts. The first part is |
| 3663 | the symbol @code{let}. The second part is a list, called a |
| 3664 | @dfn{varlist}, each element of which is either a symbol by itself or a |
| 3665 | two-element list, the first element of which is a symbol. The third |
| 3666 | part of the @code{let} expression is the body of the @code{let}. The |
| 3667 | body usually consists of one or more lists. |
| 3668 | |
| 3669 | @need 800 |
| 3670 | A template for a @code{let} expression looks like this: |
| 3671 | |
| 3672 | @smallexample |
| 3673 | (let @var{varlist} @var{body}@dots{}) |
| 3674 | @end smallexample |
| 3675 | |
| 3676 | @noindent |
| 3677 | The symbols in the varlist are the variables that are given initial |
| 3678 | values by the @code{let} special form. Symbols by themselves are given |
| 3679 | the initial value of @code{nil}; and each symbol that is the first |
| 3680 | element of a two-element list is bound to the value that is returned |
| 3681 | when the Lisp interpreter evaluates the second element. |
| 3682 | |
| 3683 | Thus, a varlist might look like this: @code{(thread (needles 3))}. In |
| 3684 | this case, in a @code{let} expression, Emacs binds the symbol |
| 3685 | @code{thread} to an initial value of @code{nil}, and binds the symbol |
| 3686 | @code{needles} to an initial value of 3. |
| 3687 | |
| 3688 | When you write a @code{let} expression, what you do is put the |
| 3689 | appropriate expressions in the slots of the @code{let} expression |
| 3690 | template. |
| 3691 | |
| 3692 | If the varlist is composed of two-element lists, as is often the case, |
| 3693 | the template for the @code{let} expression looks like this: |
| 3694 | |
| 3695 | @smallexample |
| 3696 | @group |
| 3697 | (let ((@var{variable} @var{value}) |
| 3698 | (@var{variable} @var{value}) |
| 3699 | @dots{}) |
| 3700 | @var{body}@dots{}) |
| 3701 | @end group |
| 3702 | @end smallexample |
| 3703 | |
| 3704 | @node Sample let Expression, Uninitialized let Variables, Parts of let Expression, let |
| 3705 | @comment node-name, next, previous, up |
| 3706 | @subsection Sample @code{let} Expression |
| 3707 | @cindex Sample @code{let} expression |
| 3708 | @cindex @code{let} expression sample |
| 3709 | |
| 3710 | The following expression creates and gives initial values |
| 3711 | to the two variables @code{zebra} and @code{tiger}. The body of the |
| 3712 | @code{let} expression is a list which calls the @code{message} function. |
| 3713 | |
| 3714 | @smallexample |
| 3715 | @group |
| 3716 | (let ((zebra 'stripes) |
| 3717 | (tiger 'fierce)) |
| 3718 | (message "One kind of animal has %s and another is %s." |
| 3719 | zebra tiger)) |
| 3720 | @end group |
| 3721 | @end smallexample |
| 3722 | |
| 3723 | Here, the varlist is @code{((zebra 'stripes) (tiger 'fierce))}. |
| 3724 | |
| 3725 | The two variables are @code{zebra} and @code{tiger}. Each variable is |
| 3726 | the first element of a two-element list and each value is the second |
| 3727 | element of its two-element list. In the varlist, Emacs binds the |
| 3728 | variable @code{zebra} to the value @code{stripes}, and binds the |
| 3729 | variable @code{tiger} to the value @code{fierce}. In this example, |
| 3730 | both values are symbols preceded by a quote. The values could just as |
| 3731 | well have been another list or a string. The body of the @code{let} |
| 3732 | follows after the list holding the variables. In this example, the body |
| 3733 | is a list that uses the @code{message} function to print a string in |
| 3734 | the echo area. |
| 3735 | |
| 3736 | @need 1500 |
| 3737 | You may evaluate the example in the usual fashion, by placing the |
| 3738 | cursor after the last parenthesis and typing @kbd{C-x C-e}. When you do |
| 3739 | this, the following will appear in the echo area: |
| 3740 | |
| 3741 | @smallexample |
| 3742 | "One kind of animal has stripes and another is fierce." |
| 3743 | @end smallexample |
| 3744 | |
| 3745 | As we have seen before, the @code{message} function prints its first |
| 3746 | argument, except for @samp{%s}. In this example, the value of the variable |
| 3747 | @code{zebra} is printed at the location of the first @samp{%s} and the |
| 3748 | value of the variable @code{tiger} is printed at the location of the |
| 3749 | second @samp{%s}. |
| 3750 | |
| 3751 | @node Uninitialized let Variables, , Sample let Expression, let |
| 3752 | @comment node-name, next, previous, up |
| 3753 | @subsection Uninitialized Variables in a @code{let} Statement |
| 3754 | @cindex Uninitialized @code{let} variables |
| 3755 | @cindex @code{let} variables uninitialized |
| 3756 | |
| 3757 | If you do not bind the variables in a @code{let} statement to specific |
| 3758 | initial values, they will automatically be bound to an initial value of |
| 3759 | @code{nil}, as in the following expression: |
| 3760 | |
| 3761 | @smallexample |
| 3762 | @group |
| 3763 | (let ((birch 3) |
| 3764 | pine |
| 3765 | fir |
| 3766 | (oak 'some)) |
| 3767 | (message |
| 3768 | "Here are %d variables with %s, %s, and %s value." |
| 3769 | birch pine fir oak)) |
| 3770 | @end group |
| 3771 | @end smallexample |
| 3772 | |
| 3773 | @noindent |
| 3774 | Here, the varlist is @code{((birch 3) pine fir (oak 'some))}. |
| 3775 | |
| 3776 | @need 1250 |
| 3777 | If you evaluate this expression in the usual way, the following will |
| 3778 | appear in your echo area: |
| 3779 | |
| 3780 | @smallexample |
| 3781 | "Here are 3 variables with nil, nil, and some value." |
| 3782 | @end smallexample |
| 3783 | |
| 3784 | @noindent |
| 3785 | In this example, Emacs binds the symbol @code{birch} to the number 3, |
| 3786 | binds the symbols @code{pine} and @code{fir} to @code{nil}, and binds |
| 3787 | the symbol @code{oak} to the value @code{some}. |
| 3788 | |
| 3789 | Note that in the first part of the @code{let}, the variables @code{pine} |
| 3790 | and @code{fir} stand alone as atoms that are not surrounded by |
| 3791 | parentheses; this is because they are being bound to @code{nil}, the |
| 3792 | empty list. But @code{oak} is bound to @code{some} and so is a part of |
| 3793 | the list @code{(oak 'some)}. Similarly, @code{birch} is bound to the |
| 3794 | number 3 and so is in a list with that number. (Since a number |
| 3795 | evaluates to itself, the number does not need to be quoted. Also, the |
| 3796 | number is printed in the message using a @samp{%d} rather than a |
| 3797 | @samp{%s}.) The four variables as a group are put into a list to |
| 3798 | delimit them from the body of the @code{let}. |
| 3799 | |
| 3800 | @node if, else, let, Writing Defuns |
| 3801 | @comment node-name, next, previous, up |
| 3802 | @section The @code{if} Special Form |
| 3803 | @findex if |
| 3804 | @cindex Conditional with @code{if} |
| 3805 | |
| 3806 | A third special form, in addition to @code{defun} and @code{let}, is the |
| 3807 | conditional @code{if}. This form is used to instruct the computer to |
| 3808 | make decisions. You can write function definitions without using |
| 3809 | @code{if}, but it is used often enough, and is important enough, to be |
| 3810 | included here. It is used, for example, in the code for the |
| 3811 | function @code{beginning-of-buffer}. |
| 3812 | |
| 3813 | The basic idea behind an @code{if}, is that ``@emph{if} a test is true, |
| 3814 | @emph{then} an expression is evaluated.'' If the test is not true, the |
| 3815 | expression is not evaluated. For example, you might make a decision |
| 3816 | such as, ``if it is warm and sunny, then go to the beach!'' |
| 3817 | |
| 3818 | @menu |
| 3819 | * if in more detail:: |
| 3820 | * type-of-animal in detail:: An example of an @code{if} expression. |
| 3821 | @end menu |
| 3822 | |
| 3823 | @node if in more detail, type-of-animal in detail, if, if |
| 3824 | @ifnottex |
| 3825 | @unnumberedsubsec @code{if} in more detail |
| 3826 | @end ifnottex |
| 3827 | |
| 3828 | @cindex @samp{if-part} defined |
| 3829 | @cindex @samp{then-part} defined |
| 3830 | An @code{if} expression written in Lisp does not use the word `then'; |
| 3831 | the test and the action are the second and third elements of the list |
| 3832 | whose first element is @code{if}. Nonetheless, the test part of an |
| 3833 | @code{if} expression is often called the @dfn{if-part} and the second |
| 3834 | argument is often called the @dfn{then-part}. |
| 3835 | |
| 3836 | Also, when an @code{if} expression is written, the true-or-false-test |
| 3837 | is usually written on the same line as the symbol @code{if}, but the |
| 3838 | action to carry out if the test is true, the ``then-part'', is written |
| 3839 | on the second and subsequent lines. This makes the @code{if} |
| 3840 | expression easier to read. |
| 3841 | |
| 3842 | @smallexample |
| 3843 | @group |
| 3844 | (if @var{true-or-false-test} |
| 3845 | @var{action-to-carry-out-if-test-is-true}) |
| 3846 | @end group |
| 3847 | @end smallexample |
| 3848 | |
| 3849 | @noindent |
| 3850 | The true-or-false-test will be an expression that |
| 3851 | is evaluated by the Lisp interpreter. |
| 3852 | |
| 3853 | Here is an example that you can evaluate in the usual manner. The test |
| 3854 | is whether the number 5 is greater than the number 4. Since it is, the |
| 3855 | message @samp{5 is greater than 4!} will be printed. |
| 3856 | |
| 3857 | @smallexample |
| 3858 | @group |
| 3859 | (if (> 5 4) ; @r{if-part} |
| 3860 | (message "5 is greater than 4!")) ; @r{then-part} |
| 3861 | @end group |
| 3862 | @end smallexample |
| 3863 | |
| 3864 | @noindent |
| 3865 | (The function @code{>} tests whether its first argument is greater than |
| 3866 | its second argument and returns true if it is.) |
| 3867 | @findex > (greater than) |
| 3868 | |
| 3869 | Of course, in actual use, the test in an @code{if} expression will not |
| 3870 | be fixed for all time as it is by the expression @code{(> 5 4)}. |
| 3871 | Instead, at least one of the variables used in the test will be bound to |
| 3872 | a value that is not known ahead of time. (If the value were known ahead |
| 3873 | of time, we would not need to run the test!) |
| 3874 | |
| 3875 | For example, the value may be bound to an argument of a function |
| 3876 | definition. In the following function definition, the character of the |
| 3877 | animal is a value that is passed to the function. If the value bound to |
| 3878 | @code{characteristic} is @code{fierce}, then the message, @samp{It's a |
| 3879 | tiger!} will be printed; otherwise, @code{nil} will be returned. |
| 3880 | |
| 3881 | @smallexample |
| 3882 | @group |
| 3883 | (defun type-of-animal (characteristic) |
| 3884 | "Print message in echo area depending on CHARACTERISTIC. |
| 3885 | If the CHARACTERISTIC is the symbol `fierce', |
| 3886 | then warn of a tiger." |
| 3887 | (if (equal characteristic 'fierce) |
| 3888 | (message "It's a tiger!"))) |
| 3889 | @end group |
| 3890 | @end smallexample |
| 3891 | |
| 3892 | @need 1500 |
| 3893 | @noindent |
| 3894 | If you are reading this inside of GNU Emacs, you can evaluate the |
| 3895 | function definition in the usual way to install it in Emacs, and then you |
| 3896 | can evaluate the following two expressions to see the results: |
| 3897 | |
| 3898 | @smallexample |
| 3899 | @group |
| 3900 | (type-of-animal 'fierce) |
| 3901 | |
| 3902 | (type-of-animal 'zebra) |
| 3903 | |
| 3904 | @end group |
| 3905 | @end smallexample |
| 3906 | |
| 3907 | @c Following sentences rewritten to prevent overfull hbox. |
| 3908 | @noindent |
| 3909 | When you evaluate @code{(type-of-animal 'fierce)}, you will see the |
| 3910 | following message printed in the echo area: @code{"It's a tiger!"}; and |
| 3911 | when you evaluate @code{(type-of-animal 'zebra)} you will see @code{nil} |
| 3912 | printed in the echo area. |
| 3913 | |
| 3914 | @node type-of-animal in detail, , if in more detail, if |
| 3915 | @comment node-name, next, previous, up |
| 3916 | @subsection The @code{type-of-animal} Function in Detail |
| 3917 | |
| 3918 | Let's look at the @code{type-of-animal} function in detail. |
| 3919 | |
| 3920 | The function definition for @code{type-of-animal} was written by filling |
| 3921 | the slots of two templates, one for a function definition as a whole, and |
| 3922 | a second for an @code{if} expression. |
| 3923 | |
| 3924 | @need 1250 |
| 3925 | The template for every function that is not interactive is: |
| 3926 | |
| 3927 | @smallexample |
| 3928 | @group |
| 3929 | (defun @var{name-of-function} (@var{argument-list}) |
| 3930 | "@var{documentation}@dots{}" |
| 3931 | @var{body}@dots{}) |
| 3932 | @end group |
| 3933 | @end smallexample |
| 3934 | |
| 3935 | @need 800 |
| 3936 | The parts of the function that match this template look like this: |
| 3937 | |
| 3938 | @smallexample |
| 3939 | @group |
| 3940 | (defun type-of-animal (characteristic) |
| 3941 | "Print message in echo area depending on CHARACTERISTIC. |
| 3942 | If the CHARACTERISTIC is the symbol `fierce', |
| 3943 | then warn of a tiger." |
| 3944 | @var{body: the} @code{if} @var{expression}) |
| 3945 | @end group |
| 3946 | @end smallexample |
| 3947 | |
| 3948 | The name of function is @code{type-of-animal}; it is passed the value |
| 3949 | of one argument. The argument list is followed by a multi-line |
| 3950 | documentation string. The documentation string is included in the |
| 3951 | example because it is a good habit to write documentation string for |
| 3952 | every function definition. The body of the function definition |
| 3953 | consists of the @code{if} expression. |
| 3954 | |
| 3955 | @need 800 |
| 3956 | The template for an @code{if} expression looks like this: |
| 3957 | |
| 3958 | @smallexample |
| 3959 | @group |
| 3960 | (if @var{true-or-false-test} |
| 3961 | @var{action-to-carry-out-if-the-test-returns-true}) |
| 3962 | @end group |
| 3963 | @end smallexample |
| 3964 | |
| 3965 | @need 1250 |
| 3966 | In the @code{type-of-animal} function, the code for the @code{if} |
| 3967 | looks like this: |
| 3968 | |
| 3969 | @smallexample |
| 3970 | @group |
| 3971 | (if (equal characteristic 'fierce) |
| 3972 | (message "It's a tiger!"))) |
| 3973 | @end group |
| 3974 | @end smallexample |
| 3975 | |
| 3976 | @need 800 |
| 3977 | Here, the true-or-false-test is the expression: |
| 3978 | |
| 3979 | @smallexample |
| 3980 | (equal characteristic 'fierce) |
| 3981 | @end smallexample |
| 3982 | |
| 3983 | @noindent |
| 3984 | In Lisp, @code{equal} is a function that determines whether its first |
| 3985 | argument is equal to its second argument. The second argument is the |
| 3986 | quoted symbol @code{'fierce} and the first argument is the value of the |
| 3987 | symbol @code{characteristic}---in other words, the argument passed to |
| 3988 | this function. |
| 3989 | |
| 3990 | In the first exercise of @code{type-of-animal}, the argument |
| 3991 | @code{fierce} is passed to @code{type-of-animal}. Since @code{fierce} |
| 3992 | is equal to @code{fierce}, the expression, @code{(equal characteristic |
| 3993 | 'fierce)}, returns a value of true. When this happens, the @code{if} |
| 3994 | evaluates the second argument or then-part of the @code{if}: |
| 3995 | @code{(message "It's tiger!")}. |
| 3996 | |
| 3997 | On the other hand, in the second exercise of @code{type-of-animal}, the |
| 3998 | argument @code{zebra} is passed to @code{type-of-animal}. @code{zebra} |
| 3999 | is not equal to @code{fierce}, so the then-part is not evaluated and |
| 4000 | @code{nil} is returned by the @code{if} expression. |
| 4001 | |
| 4002 | @node else, Truth & Falsehood, if, Writing Defuns |
| 4003 | @comment node-name, next, previous, up |
| 4004 | @section If--then--else Expressions |
| 4005 | @cindex Else |
| 4006 | |
| 4007 | An @code{if} expression may have an optional third argument, called |
| 4008 | the @dfn{else-part}, for the case when the true-or-false-test returns |
| 4009 | false. When this happens, the second argument or then-part of the |
| 4010 | overall @code{if} expression is @emph{not} evaluated, but the third or |
| 4011 | else-part @emph{is} evaluated. You might think of this as the cloudy |
| 4012 | day alternative for the decision ``if it is warm and sunny, then go to |
| 4013 | the beach, else read a book!''. |
| 4014 | |
| 4015 | The word ``else'' is not written in the Lisp code; the else-part of an |
| 4016 | @code{if} expression comes after the then-part. In the written Lisp, the |
| 4017 | else-part is usually written to start on a line of its own and is |
| 4018 | indented less than the then-part: |
| 4019 | |
| 4020 | @smallexample |
| 4021 | @group |
| 4022 | (if @var{true-or-false-test} |
| 4023 | @var{action-to-carry-out-if-the-test-returns-true} |
| 4024 | @var{action-to-carry-out-if-the-test-returns-false}) |
| 4025 | @end group |
| 4026 | @end smallexample |
| 4027 | |
| 4028 | For example, the following @code{if} expression prints the message @samp{4 |
| 4029 | is not greater than 5!} when you evaluate it in the usual way: |
| 4030 | |
| 4031 | @smallexample |
| 4032 | @group |
| 4033 | (if (> 4 5) ; @r{if-part} |
| 4034 | (message "5 is greater than 4!") ; @r{then-part} |
| 4035 | (message "4 is not greater than 5!")) ; @r{else-part} |
| 4036 | @end group |
| 4037 | @end smallexample |
| 4038 | |
| 4039 | @noindent |
| 4040 | Note that the different levels of indentation make it easy to |
| 4041 | distinguish the then-part from the else-part. (GNU Emacs has several |
| 4042 | commands that automatically indent @code{if} expressions correctly. |
| 4043 | @xref{Typing Lists, , GNU Emacs Helps You Type Lists}.) |
| 4044 | |
| 4045 | We can extend the @code{type-of-animal} function to include an |
| 4046 | else-part by simply incorporating an additional part to the @code{if} |
| 4047 | expression. |
| 4048 | |
| 4049 | @need 1500 |
| 4050 | You can see the consequences of doing this if you evaluate the following |
| 4051 | version of the @code{type-of-animal} function definition to install it |
| 4052 | and then evaluate the two subsequent expressions to pass different |
| 4053 | arguments to the function. |
| 4054 | |
| 4055 | @smallexample |
| 4056 | @group |
| 4057 | (defun type-of-animal (characteristic) ; @r{Second version.} |
| 4058 | "Print message in echo area depending on CHARACTERISTIC. |
| 4059 | If the CHARACTERISTIC is the symbol `fierce', |
| 4060 | then warn of a tiger; |
| 4061 | else say it's not fierce." |
| 4062 | (if (equal characteristic 'fierce) |
| 4063 | (message "It's a tiger!") |
| 4064 | (message "It's not fierce!"))) |
| 4065 | @end group |
| 4066 | @end smallexample |
| 4067 | @sp 1 |
| 4068 | |
| 4069 | @smallexample |
| 4070 | @group |
| 4071 | (type-of-animal 'fierce) |
| 4072 | |
| 4073 | (type-of-animal 'zebra) |
| 4074 | |
| 4075 | @end group |
| 4076 | @end smallexample |
| 4077 | |
| 4078 | @c Following sentence rewritten to prevent overfull hbox. |
| 4079 | @noindent |
| 4080 | When you evaluate @code{(type-of-animal 'fierce)}, you will see the |
| 4081 | following message printed in the echo area: @code{"It's a tiger!"}; but |
| 4082 | when you evaluate @code{(type-of-animal 'zebra)}, you will see |
| 4083 | @code{"It's not fierce!"}. |
| 4084 | |
| 4085 | (Of course, if the @var{characteristic} were @code{ferocious}, the |
| 4086 | message @code{"It's not fierce!"} would be printed; and it would be |
| 4087 | misleading! When you write code, you need to take into account the |
| 4088 | possibility that some such argument will be tested by the @code{if} and |
| 4089 | write your program accordingly.) |
| 4090 | |
| 4091 | @node Truth & Falsehood, save-excursion, else, Writing Defuns |
| 4092 | @comment node-name, next, previous, up |
| 4093 | @section Truth and Falsehood in Emacs Lisp |
| 4094 | @cindex Truth and falsehood in Emacs Lisp |
| 4095 | @cindex Falsehood and truth in Emacs Lisp |
| 4096 | @findex nil |
| 4097 | |
| 4098 | There is an important aspect to the truth test in an @code{if} |
| 4099 | expression. So far, we have spoken of `true' and `false' as values of |
| 4100 | predicates as if they were new kinds of Emacs Lisp objects. In fact, |
| 4101 | `false' is just our old friend @code{nil}. Anything else---anything |
| 4102 | at all---is `true'. |
| 4103 | |
| 4104 | The expression that tests for truth is interpreted as @dfn{true} |
| 4105 | if the result of evaluating it is a value that is not @code{nil}. In |
| 4106 | other words, the result of the test is considered true if the value |
| 4107 | returned is a number such as 47, a string such as @code{"hello"}, or a |
| 4108 | symbol (other than @code{nil}) such as @code{flowers}, or a list, or |
| 4109 | even a buffer! |
| 4110 | |
| 4111 | @menu |
| 4112 | * nil explained:: @code{nil} has two meanings. |
| 4113 | @end menu |
| 4114 | |
| 4115 | @node nil explained, , Truth & Falsehood, Truth & Falsehood |
| 4116 | @ifnottex |
| 4117 | @unnumberedsubsec An explanation of @code{nil} |
| 4118 | @end ifnottex |
| 4119 | |
| 4120 | Before illustrating a test for truth, we need an explanation of @code{nil}. |
| 4121 | |
| 4122 | In Emacs Lisp, the symbol @code{nil} has two meanings. First, it means the |
| 4123 | empty list. Second, it means false and is the value returned when a |
| 4124 | true-or-false-test tests false. @code{nil} can be written as an empty |
| 4125 | list, @code{()}, or as @code{nil}. As far as the Lisp interpreter is |
| 4126 | concerned, @code{()} and @code{nil} are the same. Humans, however, tend |
| 4127 | to use @code{nil} for false and @code{()} for the empty list. |
| 4128 | |
| 4129 | In Emacs Lisp, any value that is not @code{nil}---is not the empty |
| 4130 | list---is considered true. This means that if an evaluation returns |
| 4131 | something that is not an empty list, an @code{if} expression will test |
| 4132 | true. For example, if a number is put in the slot for the test, it |
| 4133 | will be evaluated and will return itself, since that is what numbers |
| 4134 | do when evaluated. In this conditional, the @code{if} expression will |
| 4135 | test true. The expression tests false only when @code{nil}, an empty |
| 4136 | list, is returned by evaluating the expression. |
| 4137 | |
| 4138 | You can see this by evaluating the two expressions in the following examples. |
| 4139 | |
| 4140 | In the first example, the number 4 is evaluated as the test in the |
| 4141 | @code{if} expression and returns itself; consequently, the then-part |
| 4142 | of the expression is evaluated and returned: @samp{true} appears in |
| 4143 | the echo area. In the second example, the @code{nil} indicates false; |
| 4144 | consequently, the else-part of the expression is evaluated and |
| 4145 | returned: @samp{false} appears in the echo area. |
| 4146 | |
| 4147 | @smallexample |
| 4148 | @group |
| 4149 | (if 4 |
| 4150 | 'true |
| 4151 | 'false) |
| 4152 | @end group |
| 4153 | |
| 4154 | @group |
| 4155 | (if nil |
| 4156 | 'true |
| 4157 | 'false) |
| 4158 | @end group |
| 4159 | @end smallexample |
| 4160 | |
| 4161 | @need 1250 |
| 4162 | Incidentally, if some other useful value is not available for a test that |
| 4163 | returns true, then the Lisp interpreter will return the symbol @code{t} |
| 4164 | for true. For example, the expression @code{(> 5 4)} returns @code{t} |
| 4165 | when evaluated, as you can see by evaluating it in the usual way: |
| 4166 | |
| 4167 | @smallexample |
| 4168 | (> 5 4) |
| 4169 | @end smallexample |
| 4170 | |
| 4171 | @need 1250 |
| 4172 | @noindent |
| 4173 | On the other hand, this function returns @code{nil} if the test is false. |
| 4174 | |
| 4175 | @smallexample |
| 4176 | (> 4 5) |
| 4177 | @end smallexample |
| 4178 | |
| 4179 | @node save-excursion, Review, Truth & Falsehood, Writing Defuns |
| 4180 | @comment node-name, next, previous, up |
| 4181 | @section @code{save-excursion} |
| 4182 | @findex save-excursion |
| 4183 | @cindex Region, what it is |
| 4184 | @cindex Preserving point, mark, and buffer |
| 4185 | @cindex Point, mark, buffer preservation |
| 4186 | @findex point |
| 4187 | @findex mark |
| 4188 | |
| 4189 | The @code{save-excursion} function is the fourth and final special form |
| 4190 | that we will discuss in this chapter. |
| 4191 | |
| 4192 | In Emacs Lisp programs used for editing, the @code{save-excursion} |
| 4193 | function is very common. It saves the location of point and mark, |
| 4194 | executes the body of the function, and then restores point and mark to |
| 4195 | their previous positions if their locations were changed. Its primary |
| 4196 | purpose is to keep the user from being surprised and disturbed by |
| 4197 | unexpected movement of point or mark. |
| 4198 | |
| 4199 | @menu |
| 4200 | * Point and mark:: A review of various locations. |
| 4201 | * Template for save-excursion:: |
| 4202 | @end menu |
| 4203 | |
| 4204 | @node Point and mark, Template for save-excursion, save-excursion, save-excursion |
| 4205 | @ifnottex |
| 4206 | @unnumberedsubsec Point and Mark |
| 4207 | @end ifnottex |
| 4208 | |
| 4209 | Before discussing @code{save-excursion}, however, it may be useful |
| 4210 | first to review what point and mark are in GNU Emacs. @dfn{Point} is |
| 4211 | the current location of the cursor. Wherever the cursor |
| 4212 | is, that is point. More precisely, on terminals where the cursor |
| 4213 | appears to be on top of a character, point is immediately before the |
| 4214 | character. In Emacs Lisp, point is an integer. The first character in |
| 4215 | a buffer is number one, the second is number two, and so on. The |
| 4216 | function @code{point} returns the current position of the cursor as a |
| 4217 | number. Each buffer has its own value for point. |
| 4218 | |
| 4219 | The @dfn{mark} is another position in the buffer; its value can be set |
| 4220 | with a command such as @kbd{C-@key{SPC}} (@code{set-mark-command}). If |
| 4221 | a mark has been set, you can use the command @kbd{C-x C-x} |
| 4222 | (@code{exchange-point-and-mark}) to cause the cursor to jump to the mark |
| 4223 | and set the mark to be the previous position of point. In addition, if |
| 4224 | you set another mark, the position of the previous mark is saved in the |
| 4225 | mark ring. Many mark positions can be saved this way. You can jump the |
| 4226 | cursor to a saved mark by typing @kbd{C-u C-@key{SPC}} one or more |
| 4227 | times. |
| 4228 | |
| 4229 | The part of the buffer between point and mark is called @dfn{the |
| 4230 | region}. Numerous commands work on the region, including |
| 4231 | @code{center-region}, @code{count-lines-region}, @code{kill-region}, and |
| 4232 | @code{print-region}. |
| 4233 | |
| 4234 | The @code{save-excursion} special form saves the locations of point and |
| 4235 | mark and restores those positions after the code within the body of the |
| 4236 | special form is evaluated by the Lisp interpreter. Thus, if point were |
| 4237 | in the beginning of a piece of text and some code moved point to the end |
| 4238 | of the buffer, the @code{save-excursion} would put point back to where |
| 4239 | it was before, after the expressions in the body of the function were |
| 4240 | evaluated. |
| 4241 | |
| 4242 | In Emacs, a function frequently moves point as part of its internal |
| 4243 | workings even though a user would not expect this. For example, |
| 4244 | @code{count-lines-region} moves point. To prevent the user from being |
| 4245 | bothered by jumps that are both unexpected and (from the user's point of |
| 4246 | view) unnecessary, @code{save-excursion} is often used to keep point and |
| 4247 | mark in the location expected by the user. The use of |
| 4248 | @code{save-excursion} is good housekeeping. |
| 4249 | |
| 4250 | To make sure the house stays clean, @code{save-excursion} restores the |
| 4251 | values of point and mark even if something goes wrong in the code inside |
| 4252 | of it (or, to be more precise and to use the proper jargon, ``in case of |
| 4253 | abnormal exit''). This feature is very helpful. |
| 4254 | |
| 4255 | In addition to recording the values of point and mark, |
| 4256 | @code{save-excursion} keeps track of the current buffer, and restores |
| 4257 | it, too. This means you can write code that will change the buffer and |
| 4258 | have @code{save-excursion} switch you back to the original buffer. This |
| 4259 | is how @code{save-excursion} is used in @code{append-to-buffer}. |
| 4260 | (@xref{append-to-buffer, , The Definition of @code{append-to-buffer}}.) |
| 4261 | |
| 4262 | @node Template for save-excursion, , Point and mark, save-excursion |
| 4263 | @comment node-name, next, previous, up |
| 4264 | @subsection Template for a @code{save-excursion} Expression |
| 4265 | |
| 4266 | @need 800 |
| 4267 | The template for code using @code{save-excursion} is simple: |
| 4268 | |
| 4269 | @smallexample |
| 4270 | @group |
| 4271 | (save-excursion |
| 4272 | @var{body}@dots{}) |
| 4273 | @end group |
| 4274 | @end smallexample |
| 4275 | |
| 4276 | @noindent |
| 4277 | The body of the function is one or more expressions that will be |
| 4278 | evaluated in sequence by the Lisp interpreter. If there is more than |
| 4279 | one expression in the body, the value of the last one will be returned |
| 4280 | as the value of the @code{save-excursion} function. The other |
| 4281 | expressions in the body are evaluated only for their side effects; and |
| 4282 | @code{save-excursion} itself is used only for its side effect (which |
| 4283 | is restoring the positions of point and mark). |
| 4284 | |
| 4285 | @need 1250 |
| 4286 | In more detail, the template for a @code{save-excursion} expression |
| 4287 | looks like this: |
| 4288 | |
| 4289 | @smallexample |
| 4290 | @group |
| 4291 | (save-excursion |
| 4292 | @var{first-expression-in-body} |
| 4293 | @var{second-expression-in-body} |
| 4294 | @var{third-expression-in-body} |
| 4295 | @dots{} |
| 4296 | @var{last-expression-in-body}) |
| 4297 | @end group |
| 4298 | @end smallexample |
| 4299 | |
| 4300 | @noindent |
| 4301 | An expression, of course, may be a symbol on its own or a list. |
| 4302 | |
| 4303 | In Emacs Lisp code, a @code{save-excursion} expression often occurs |
| 4304 | within the body of a @code{let} expression. It looks like this: |
| 4305 | |
| 4306 | @smallexample |
| 4307 | @group |
| 4308 | (let @var{varlist} |
| 4309 | (save-excursion |
| 4310 | @var{body}@dots{})) |
| 4311 | @end group |
| 4312 | @end smallexample |
| 4313 | |
| 4314 | @node Review, defun Exercises, save-excursion, Writing Defuns |
| 4315 | @comment node-name, next, previous, up |
| 4316 | @section Review |
| 4317 | |
| 4318 | In the last few chapters we have introduced a fair number of functions |
| 4319 | and special forms. Here they are described in brief, along with a few |
| 4320 | similar functions that have not been mentioned yet. |
| 4321 | |
| 4322 | @table @code |
| 4323 | @item eval-last-sexp |
| 4324 | Evaluate the last symbolic expression before the current location of |
| 4325 | point. The value is printed in the echo area unless the function is |
| 4326 | invoked with an argument; in that case, the output is printed in the |
| 4327 | current buffer. This command is normally bound to @kbd{C-x C-e}. |
| 4328 | |
| 4329 | @item defun |
| 4330 | Define function. This special form has up to five parts: the name, |
| 4331 | a template for the arguments that will be passed to the function, |
| 4332 | documentation, an optional interactive declaration, and the body of the |
| 4333 | definition. |
| 4334 | |
| 4335 | @need 1250 |
| 4336 | For example: |
| 4337 | |
| 4338 | @smallexample |
| 4339 | @group |
| 4340 | (defun back-to-indentation () |
| 4341 | "Move point to first visible character on line." |
| 4342 | (interactive) |
| 4343 | (beginning-of-line 1) |
| 4344 | (skip-chars-forward " \t")) |
| 4345 | @end group |
| 4346 | @end smallexample |
| 4347 | |
| 4348 | @item interactive |
| 4349 | Declare to the interpreter that the function can be used |
| 4350 | interactively. This special form may be followed by a string with one |
| 4351 | or more parts that pass the information to the arguments of the |
| 4352 | function, in sequence. These parts may also tell the interpreter to |
| 4353 | prompt for information. Parts of the string are separated by |
| 4354 | newlines, @samp{\n}. |
| 4355 | |
| 4356 | @need 1000 |
| 4357 | Common code characters are: |
| 4358 | |
| 4359 | @table @code |
| 4360 | @item b |
| 4361 | The name of an existing buffer. |
| 4362 | |
| 4363 | @item f |
| 4364 | The name of an existing file. |
| 4365 | |
| 4366 | @item p |
| 4367 | The numeric prefix argument. (Note that this `p' is lower case.) |
| 4368 | |
| 4369 | @item r |
| 4370 | Point and the mark, as two numeric arguments, smallest first. This |
| 4371 | is the only code letter that specifies two successive arguments |
| 4372 | rather than one. |
| 4373 | @end table |
| 4374 | |
| 4375 | @xref{Interactive Codes, , Code Characters for @samp{interactive}, |
| 4376 | elisp, The GNU Emacs Lisp Reference Manual}, for a complete list of |
| 4377 | code characters. |
| 4378 | |
| 4379 | @item let |
| 4380 | Declare that a list of variables is for use within the body of the |
| 4381 | @code{let} and give them an initial value, either @code{nil} or a |
| 4382 | specified value; then evaluate the rest of the expressions in the body |
| 4383 | of the @code{let} and return the value of the last one. Inside the |
| 4384 | body of the @code{let}, the Lisp interpreter does not see the values of |
| 4385 | the variables of the same names that are bound outside of the |
| 4386 | @code{let}. |
| 4387 | |
| 4388 | @need 1250 |
| 4389 | For example, |
| 4390 | |
| 4391 | @smallexample |
| 4392 | @group |
| 4393 | (let ((foo (buffer-name)) |
| 4394 | (bar (buffer-size))) |
| 4395 | (message |
| 4396 | "This buffer is %s and has %d characters." |
| 4397 | foo bar)) |
| 4398 | @end group |
| 4399 | @end smallexample |
| 4400 | |
| 4401 | @item save-excursion |
| 4402 | Record the values of point and mark and the current buffer before |
| 4403 | evaluating the body of this special form. Restore the values of point |
| 4404 | and mark and buffer afterward. |
| 4405 | |
| 4406 | @need 1250 |
| 4407 | For example, |
| 4408 | |
| 4409 | @smallexample |
| 4410 | @group |
| 4411 | (message "We are %d characters into this buffer." |
| 4412 | (- (point) |
| 4413 | (save-excursion |
| 4414 | (goto-char (point-min)) (point)))) |
| 4415 | @end group |
| 4416 | @end smallexample |
| 4417 | |
| 4418 | @item if |
| 4419 | Evaluate the first argument to the function; if it is true, evaluate |
| 4420 | the second argument; else evaluate the third argument, if there is one. |
| 4421 | |
| 4422 | The @code{if} special form is called a @dfn{conditional}. There are |
| 4423 | other conditionals in Emacs Lisp, but @code{if} is perhaps the most |
| 4424 | commonly used. |
| 4425 | |
| 4426 | @need 1250 |
| 4427 | For example, |
| 4428 | |
| 4429 | @smallexample |
| 4430 | @group |
| 4431 | (if (string-equal |
| 4432 | (number-to-string 21) |
| 4433 | (substring (emacs-version) 10 12)) |
| 4434 | (message "This is version 21 Emacs") |
| 4435 | (message "This is not version 21 Emacs")) |
| 4436 | @end group |
| 4437 | @end smallexample |
| 4438 | |
| 4439 | @item equal |
| 4440 | @itemx eq |
| 4441 | Test whether two objects are the same. @code{equal} uses one meaning |
| 4442 | of the word `same' and @code{eq} uses another: @code{equal} returns |
| 4443 | true if the two objects have a similar structure and contents, such as |
| 4444 | two copies of the same book. On the other hand, @code{eq}, returns |
| 4445 | true if both arguments are actually the same object. |
| 4446 | @findex equal |
| 4447 | @findex eq |
| 4448 | |
| 4449 | @need 1250 |
| 4450 | @item < |
| 4451 | @itemx > |
| 4452 | @itemx <= |
| 4453 | @itemx >= |
| 4454 | The @code{<} function tests whether its first argument is smaller than |
| 4455 | its second argument. A corresponding function, @code{>}, tests whether |
| 4456 | the first argument is greater than the second. Likewise, @code{<=} |
| 4457 | tests whether the first argument is less than or equal to the second and |
| 4458 | @code{>=} tests whether the first argument is greater than or equal to |
| 4459 | the second. In all cases, both arguments must be numbers or markers |
| 4460 | (markers indicate positions in buffers). |
| 4461 | |
| 4462 | @item string< |
| 4463 | @itemx string-lessp |
| 4464 | @itemx string= |
| 4465 | @itemx string-equal |
| 4466 | The @code{string-lessp} function tests whether its first argument is |
| 4467 | smaller than the second argument. A shorter, alternative name for the |
| 4468 | same function (a @code{defalias}) is @code{string<}. |
| 4469 | |
| 4470 | The arguments to @code{string-lessp} must be strings or symbols; the |
| 4471 | ordering is lexicographic, so case is significant. The print names of |
| 4472 | symbols are used instead of the symbols themselves. |
| 4473 | |
| 4474 | @cindex @samp{empty string} defined |
| 4475 | An empty string, @samp{""}, a string with no characters in it, is |
| 4476 | smaller than any string of characters. |
| 4477 | |
| 4478 | @code{string-equal} provides the corresponding test for equality. Its |
| 4479 | shorter, alternative name is @code{string=}. There are no string test |
| 4480 | functions that correspond to @var{>}, @code{>=}, or @code{<=}. |
| 4481 | |
| 4482 | @item message |
| 4483 | Print a message in the echo area. The first argument is a string that |
| 4484 | can contain @samp{%s}, @samp{%d}, or @samp{%c} to print the value of |
| 4485 | arguments that follow the string. The argument used by @samp{%s} must |
| 4486 | be a string or a symbol; the argument used by @samp{%d} must be a |
| 4487 | number. The argument used by @samp{%c} must be an @sc{ascii} code |
| 4488 | number; it will be printed as the character with that @sc{ascii} code. |
| 4489 | |
| 4490 | @item setq |
| 4491 | @itemx set |
| 4492 | The @code{setq} function sets the value of its first argument to the |
| 4493 | value of the second argument. The first argument is automatically |
| 4494 | quoted by @code{setq}. It does the same for succeeding pairs of |
| 4495 | arguments. Another function, @code{set}, takes only two arguments and |
| 4496 | evaluates both of them before setting the value returned by its first |
| 4497 | argument to the value returned by its second argument. |
| 4498 | |
| 4499 | @item buffer-name |
| 4500 | Without an argument, return the name of the buffer, as a string. |
| 4501 | |
| 4502 | @itemx buffer-file-name |
| 4503 | Without an argument, return the name of the file the buffer is |
| 4504 | visiting. |
| 4505 | |
| 4506 | @item current-buffer |
| 4507 | Return the buffer in which Emacs is active; it may not be |
| 4508 | the buffer that is visible on the screen. |
| 4509 | |
| 4510 | @item other-buffer |
| 4511 | Return the most recently selected buffer (other than the buffer passed |
| 4512 | to @code{other-buffer} as an argument and other than the current |
| 4513 | buffer). |
| 4514 | |
| 4515 | @item switch-to-buffer |
| 4516 | Select a buffer for Emacs to be active in and display it in the current |
| 4517 | window so users can look at it. Usually bound to @kbd{C-x b}. |
| 4518 | |
| 4519 | @item set-buffer |
| 4520 | Switch Emacs' attention to a buffer on which programs will run. Don't |
| 4521 | alter what the window is showing. |
| 4522 | |
| 4523 | @item buffer-size |
| 4524 | Return the number of characters in the current buffer. |
| 4525 | |
| 4526 | @item point |
| 4527 | Return the value of the current position of the cursor, as an |
| 4528 | integer counting the number of characters from the beginning of the |
| 4529 | buffer. |
| 4530 | |
| 4531 | @item point-min |
| 4532 | Return the minimum permissible value of point in |
| 4533 | the current buffer. This is 1, unless narrowing is in effect. |
| 4534 | |
| 4535 | @item point-max |
| 4536 | Return the value of the maximum permissible value of point in the |
| 4537 | current buffer. This is the end of the buffer, unless narrowing is in |
| 4538 | effect. |
| 4539 | @end table |
| 4540 | |
| 4541 | @need 1500 |
| 4542 | @node defun Exercises, , Review, Writing Defuns |
| 4543 | @section Exercises |
| 4544 | |
| 4545 | @itemize @bullet |
| 4546 | @item |
| 4547 | Write a non-interactive function that doubles the value of its |
| 4548 | argument, a number. Make that function interactive. |
| 4549 | |
| 4550 | @item |
| 4551 | Write a function that tests whether the current value of |
| 4552 | @code{fill-column} is greater than the argument passed to the function, |
| 4553 | and if so, prints an appropriate message. |
| 4554 | @end itemize |
| 4555 | |
| 4556 | @node Buffer Walk Through, More Complex, Writing Defuns, Top |
| 4557 | @comment node-name, next, previous, up |
| 4558 | @chapter A Few Buffer--Related Functions |
| 4559 | |
| 4560 | In this chapter we study in detail several of the functions used in GNU |
| 4561 | Emacs. This is called a ``walk-through''. These functions are used as |
| 4562 | examples of Lisp code, but are not imaginary examples; with the |
| 4563 | exception of the first, simplified function definition, these functions |
| 4564 | show the actual code used in GNU Emacs. You can learn a great deal from |
| 4565 | these definitions. The functions described here are all related to |
| 4566 | buffers. Later, we will study other functions. |
| 4567 | |
| 4568 | @menu |
| 4569 | * Finding More:: How to find more information. |
| 4570 | * simplified-beginning-of-buffer:: Shows @code{goto-char}, |
| 4571 | @code{point-min}, and @code{push-mark}. |
| 4572 | * mark-whole-buffer:: Almost the same as @code{beginning-of-buffer}. |
| 4573 | * append-to-buffer:: Uses @code{save-excursion} and |
| 4574 | @code{insert-buffer-substring}. |
| 4575 | * Buffer Related Review:: Review. |
| 4576 | * Buffer Exercises:: |
| 4577 | @end menu |
| 4578 | |
| 4579 | @node Finding More, simplified-beginning-of-buffer, Buffer Walk Through, Buffer Walk Through |
| 4580 | @section Finding More Information |
| 4581 | |
| 4582 | @findex describe-function, @r{introduced} |
| 4583 | @cindex Find function documentation |
| 4584 | In this walk-through, I will describe each new function as we come to |
| 4585 | it, sometimes in detail and sometimes briefly. If you are interested, |
| 4586 | you can get the full documentation of any Emacs Lisp function at any |
| 4587 | time by typing @kbd{C-h f} and then the name of the function (and then |
| 4588 | @key{RET}). Similarly, you can get the full documentation for a |
| 4589 | variable by typing @kbd{C-h v} and then the name of the variable (and |
| 4590 | then @key{RET}). |
| 4591 | |
| 4592 | @cindex Find source of function |
| 4593 | In versions 20 and higher, when a function is written in Emacs Lisp, |
| 4594 | @code{describe-function} will also tell you the location of the |
| 4595 | function definition. If you move point over the file name and press |
| 4596 | the @key{RET} key, which in this case means @code{help-follow} rather |
| 4597 | than `return' or `enter', Emacs will take you directly to the function |
| 4598 | definition. |
| 4599 | |
| 4600 | More generally, if you want to see a function in its original source |
| 4601 | file, you can use the @code{find-tags} function to jump to it. |
| 4602 | @code{find-tags} works with a wide variety of languages, not just |
| 4603 | Lisp, and C, and it works with non-programming text as well. For |
| 4604 | example, @code{find-tags} will jump to the various nodes in the |
| 4605 | Texinfo source file of this document. |
| 4606 | |
| 4607 | The @code{find-tags} function depends on `tags tables' that record |
| 4608 | the locations of the functions, variables, and other items to which |
| 4609 | @code{find-tags} jumps. |
| 4610 | |
| 4611 | To use the @code{find-tags} command, type @kbd{M-.} (i.e., press the |
| 4612 | period key while holding down the @key{META} key, or else type the |
| 4613 | @key{ESC} key and then type the period key), and then, at the prompt, |
| 4614 | type in the name of the function whose source code you want to see, |
| 4615 | such as @code{mark-whole-buffer}, and then type @key{RET}. Emacs will |
| 4616 | switch buffers and display the source code for the function on your |
| 4617 | screen. To switch back to your current buffer, type @kbd{C-x b |
| 4618 | @key{RET}}. (On some keyboards, the @key{META} key is labelled |
| 4619 | @key{ALT}.) |
| 4620 | |
| 4621 | @c !!! 21.0.100 tags table location in this paragraph |
| 4622 | @cindex TAGS table, specifying |
| 4623 | @findex find-tags |
| 4624 | Depending on how the initial default values of your copy of Emacs are |
| 4625 | set, you may also need to specify the location of your `tags table', |
| 4626 | which is a file called @file{TAGS}. For example, if you are |
| 4627 | interested in Emacs sources, the tags table you will most likely want, |
| 4628 | if it has already been created for you, will be in a subdirectory of |
| 4629 | the @file{/usr/local/share/emacs/} directory; thus you would use the |
| 4630 | @code{M-x visit-tags-table} command and specify a pathname such as |
| 4631 | @file{/usr/local/share/emacs/21.0.100/lisp/TAGS} or |
| 4632 | @file{/usr/local/src/emacs/src/TAGS}. If the tags table has |
| 4633 | not already been created, you will have to create it yourself. |
| 4634 | |
| 4635 | @need 1250 |
| 4636 | To create a @file{TAGS} file in a specific directory, switch to that |
| 4637 | directory in Emacs using @kbd{M-x cd} command, or list the directory |
| 4638 | with @kbd{C-x d} (@code{dired}). Then run the compile command, with |
| 4639 | @w{@code{etags *.el}} as the command to execute |
| 4640 | |
| 4641 | @smallexample |
| 4642 | M-x compile RET etags *.el RET |
| 4643 | @end smallexample |
| 4644 | |
| 4645 | For more information, see @ref{etags, , Create Your Own @file{TAGS} File}. |
| 4646 | |
| 4647 | After you become more familiar with Emacs Lisp, you will find that you will |
| 4648 | frequently use @code{find-tags} to navigate your way around source code; |
| 4649 | and you will create your own @file{TAGS} tables. |
| 4650 | |
| 4651 | @cindex Library, as term for `file' |
| 4652 | Incidentally, the files that contain Lisp code are conventionally |
| 4653 | called @dfn{libraries}. The metaphor is derived from that of a |
| 4654 | specialized library, such as a law library or an engineering library, |
| 4655 | rather than a general library. Each library, or file, contains |
| 4656 | functions that relate to a particular topic or activity, such as |
| 4657 | @file{abbrev.el} for handling abbreviations and other typing |
| 4658 | shortcuts, and @file{help.el} for on-line help. (Sometimes several |
| 4659 | libraries provide code for a single activity, as the various |
| 4660 | @file{rmail@dots{}} files provide code for reading electronic mail.) |
| 4661 | In @cite{The GNU Emacs Manual}, you will see sentences such as ``The |
| 4662 | @kbd{C-h p} command lets you search the standard Emacs Lisp libraries |
| 4663 | by topic keywords.'' |
| 4664 | |
| 4665 | @node simplified-beginning-of-buffer, mark-whole-buffer, Finding More, Buffer Walk Through |
| 4666 | @comment node-name, next, previous, up |
| 4667 | @section A Simplified @code{beginning-of-buffer} Definition |
| 4668 | @findex simplified-beginning-of-buffer |
| 4669 | |
| 4670 | The @code{beginning-of-buffer} command is a good function to start with |
| 4671 | since you are likely to be familiar with it and it is easy to |
| 4672 | understand. Used as an interactive command, @code{beginning-of-buffer} |
| 4673 | moves the cursor to the beginning of the buffer, leaving the mark at the |
| 4674 | previous position. It is generally bound to @kbd{M-<}. |
| 4675 | |
| 4676 | In this section, we will discuss a shortened version of the function |
| 4677 | that shows how it is most frequently used. This shortened function |
| 4678 | works as written, but it does not contain the code for a complex option. |
| 4679 | In another section, we will describe the entire function. |
| 4680 | (@xref{beginning-of-buffer, , Complete Definition of |
| 4681 | @code{beginning-of-buffer}}.) |
| 4682 | |
| 4683 | Before looking at the code, let's consider what the function |
| 4684 | definition has to contain: it must include an expression that makes |
| 4685 | the function interactive so it can be called by typing @kbd{M-x |
| 4686 | beginning-of-buffer} or by typing a keychord such as @kbd{M-<}; it |
| 4687 | must include code to leave a mark at the original position in the |
| 4688 | buffer; and it must include code to move the cursor to the beginning |
| 4689 | of the buffer. |
| 4690 | |
| 4691 | @need 1250 |
| 4692 | Here is the complete text of the shortened version of the function: |
| 4693 | |
| 4694 | @smallexample |
| 4695 | @group |
| 4696 | (defun simplified-beginning-of-buffer () |
| 4697 | "Move point to the beginning of the buffer; |
| 4698 | leave mark at previous position." |
| 4699 | (interactive) |
| 4700 | (push-mark) |
| 4701 | (goto-char (point-min))) |
| 4702 | @end group |
| 4703 | @end smallexample |
| 4704 | |
| 4705 | Like all function definitions, this definition has five parts following |
| 4706 | the special form @code{defun}: |
| 4707 | |
| 4708 | @enumerate |
| 4709 | @item |
| 4710 | The name: in this example, @code{simplified-beginning-of-buffer}. |
| 4711 | |
| 4712 | @item |
| 4713 | A list of the arguments: in this example, an empty list, @code{()}, |
| 4714 | |
| 4715 | @item |
| 4716 | The documentation string. |
| 4717 | |
| 4718 | @item |
| 4719 | The interactive expression. |
| 4720 | |
| 4721 | @item |
| 4722 | The body. |
| 4723 | @end enumerate |
| 4724 | |
| 4725 | @noindent |
| 4726 | In this function definition, the argument list is empty; this means that |
| 4727 | this function does not require any arguments. (When we look at the |
| 4728 | definition for the complete function, we will see that it may be passed |
| 4729 | an optional argument.) |
| 4730 | |
| 4731 | The interactive expression tells Emacs that the function is intended to |
| 4732 | be used interactively. In this example, @code{interactive} does not have |
| 4733 | an argument because @code{simplified-beginning-of-buffer} does not |
| 4734 | require one. |
| 4735 | |
| 4736 | @need 800 |
| 4737 | The body of the function consists of the two lines: |
| 4738 | |
| 4739 | @smallexample |
| 4740 | @group |
| 4741 | (push-mark) |
| 4742 | (goto-char (point-min)) |
| 4743 | @end group |
| 4744 | @end smallexample |
| 4745 | |
| 4746 | The first of these lines is the expression, @code{(push-mark)}. When |
| 4747 | this expression is evaluated by the Lisp interpreter, it sets a mark at |
| 4748 | the current position of the cursor, wherever that may be. The position |
| 4749 | of this mark is saved in the mark ring. |
| 4750 | |
| 4751 | The next line is @code{(goto-char (point-min))}. This expression |
| 4752 | jumps the cursor to the minimum point in the buffer, that is, to the |
| 4753 | beginning of the buffer (or to the beginning of the accessible portion |
| 4754 | of the buffer if it is narrowed. @xref{Narrowing & Widening, , |
| 4755 | Narrowing and Widening}.) |
| 4756 | |
| 4757 | The @code{push-mark} command sets a mark at the place where the cursor |
| 4758 | was located before it was moved to the beginning of the buffer by the |
| 4759 | @code{(goto-char (point-min))} expression. Consequently, you can, if |
| 4760 | you wish, go back to where you were originally by typing @kbd{C-x C-x}. |
| 4761 | |
| 4762 | That is all there is to the function definition! |
| 4763 | |
| 4764 | @findex describe-function |
| 4765 | When you are reading code such as this and come upon an unfamiliar |
| 4766 | function, such as @code{goto-char}, you can find out what it does by |
| 4767 | using the @code{describe-function} command. To use this command, type |
| 4768 | @kbd{C-h f} and then type in the name of the function and press |
| 4769 | @key{RET}. The @code{describe-function} command will print the |
| 4770 | function's documentation string in a @file{*Help*} window. For |
| 4771 | example, the documentation for @code{goto-char} is: |
| 4772 | |
| 4773 | @smallexample |
| 4774 | @group |
| 4775 | One arg, a number. Set point to that number. |
| 4776 | Beginning of buffer is position (point-min), |
| 4777 | end is (point-max). |
| 4778 | @end group |
| 4779 | @end smallexample |
| 4780 | |
| 4781 | @noindent |
| 4782 | (The prompt for @code{describe-function} will offer you the symbol |
| 4783 | under or preceding the cursor, so you can save typing by positioning |
| 4784 | the cursor right over or after the function and then typing @kbd{C-h f |
| 4785 | @key{RET}}.) |
| 4786 | |
| 4787 | The @code{end-of-buffer} function definition is written in the same way as |
| 4788 | the @code{beginning-of-buffer} definition except that the body of the |
| 4789 | function contains the expression @code{(goto-char (point-max))} in place |
| 4790 | of @code{(goto-char (point-min))}. |
| 4791 | |
| 4792 | @node mark-whole-buffer, append-to-buffer, simplified-beginning-of-buffer, Buffer Walk Through |
| 4793 | @comment node-name, next, previous, up |
| 4794 | @section The Definition of @code{mark-whole-buffer} |
| 4795 | @findex mark-whole-buffer |
| 4796 | |
| 4797 | The @code{mark-whole-buffer} function is no harder to understand than the |
| 4798 | @code{simplified-beginning-of-buffer} function. In this case, however, |
| 4799 | we will look at the complete function, not a shortened version. |
| 4800 | |
| 4801 | The @code{mark-whole-buffer} function is not as commonly used as the |
| 4802 | @code{beginning-of-buffer} function, but is useful nonetheless: it |
| 4803 | marks a whole buffer as a region by putting point at the beginning and |
| 4804 | a mark at the end of the buffer. It is generally bound to @kbd{C-x |
| 4805 | h}. |
| 4806 | |
| 4807 | |
| 4808 | @menu |
| 4809 | * mark-whole-buffer overview:: |
| 4810 | * Body of mark-whole-buffer:: Only three lines of code. |
| 4811 | @end menu |
| 4812 | |
| 4813 | |
| 4814 | @node mark-whole-buffer overview, Body of mark-whole-buffer, mark-whole-buffer, mark-whole-buffer |
| 4815 | @ifnottex |
| 4816 | @unnumberedsubsec An overview of @code{mark-whole-buffer} |
| 4817 | @end ifnottex |
| 4818 | |
| 4819 | @need 1250 |
| 4820 | In GNU Emacs 20, the code for the complete function looks like this: |
| 4821 | |
| 4822 | @smallexample |
| 4823 | @group |
| 4824 | (defun mark-whole-buffer () |
| 4825 | "Put point at beginning and mark at end of buffer." |
| 4826 | (interactive) |
| 4827 | (push-mark (point)) |
| 4828 | (push-mark (point-max)) |
| 4829 | (goto-char (point-min))) |
| 4830 | @end group |
| 4831 | @end smallexample |
| 4832 | |
| 4833 | @need 1250 |
| 4834 | Like all other functions, the @code{mark-whole-buffer} function fits |
| 4835 | into the template for a function definition. The template looks like |
| 4836 | this: |
| 4837 | |
| 4838 | @smallexample |
| 4839 | @group |
| 4840 | (defun @var{name-of-function} (@var{argument-list}) |
| 4841 | "@var{documentation}@dots{}" |
| 4842 | (@var{interactive-expression}@dots{}) |
| 4843 | @var{body}@dots{}) |
| 4844 | @end group |
| 4845 | @end smallexample |
| 4846 | |
| 4847 | Here is how the function works: the name of the function is |
| 4848 | @code{mark-whole-buffer}; it is followed by an empty argument list, |
| 4849 | @samp{()}, which means that the function does not require arguments. |
| 4850 | The documentation comes next. |
| 4851 | |
| 4852 | The next line is an @code{(interactive)} expression that tells Emacs |
| 4853 | that the function will be used interactively. These details are similar |
| 4854 | to the @code{simplified-beginning-of-buffer} function described in the |
| 4855 | previous section. |
| 4856 | |
| 4857 | @need 1250 |
| 4858 | @node Body of mark-whole-buffer, , mark-whole-buffer overview, mark-whole-buffer |
| 4859 | @comment node-name, next, previous, up |
| 4860 | @subsection Body of @code{mark-whole-buffer} |
| 4861 | |
| 4862 | The body of the @code{mark-whole-buffer} function consists of three |
| 4863 | lines of code: |
| 4864 | |
| 4865 | @smallexample |
| 4866 | @group |
| 4867 | (push-mark (point)) |
| 4868 | (push-mark (point-max)) |
| 4869 | (goto-char (point-min)) |
| 4870 | @end group |
| 4871 | @end smallexample |
| 4872 | |
| 4873 | The first of these lines is the expression, @code{(push-mark (point))}. |
| 4874 | |
| 4875 | This line does exactly the same job as the first line of the body of |
| 4876 | the @code{simplified-beginning-of-buffer} function, which is written |
| 4877 | @code{(push-mark)}. In both cases, the Lisp interpreter sets a mark |
| 4878 | at the current position of the cursor. |
| 4879 | |
| 4880 | I don't know why the expression in @code{mark-whole-buffer} is written |
| 4881 | @code{(push-mark (point))} and the expression in |
| 4882 | @code{beginning-of-buffer} is written @code{(push-mark)}. Perhaps |
| 4883 | whoever wrote the code did not know that the arguments for |
| 4884 | @code{push-mark} are optional and that if @code{push-mark} is not |
| 4885 | passed an argument, the function automatically sets mark at the |
| 4886 | location of point by default. Or perhaps the expression was written |
| 4887 | so as to parallel the structure of the next line. In any case, the |
| 4888 | line causes Emacs to determine the position of point and set a mark |
| 4889 | there. |
| 4890 | |
| 4891 | The next line of @code{mark-whole-buffer} is @code{(push-mark (point-max)}. |
| 4892 | This expression sets a mark at the point in the buffer |
| 4893 | that has the highest number. This will be the end of the buffer (or, |
| 4894 | if the buffer is narrowed, the end of the accessible portion of the |
| 4895 | buffer. @xref{Narrowing & Widening, , Narrowing and Widening}, for |
| 4896 | more about narrowing.) After this mark has been set, the previous |
| 4897 | mark, the one set at point, is no longer set, but Emacs remembers its |
| 4898 | position, just as all other recent marks are always remembered. This |
| 4899 | means that you can, if you wish, go back to that position by typing |
| 4900 | @kbd{C-u C-@key{SPC}} twice. |
| 4901 | |
| 4902 | (In GNU Emacs 21, the @code{(push-mark (point-max)} is slightly more |
| 4903 | complicated than shown here. The line reads |
| 4904 | |
| 4905 | @smallexample |
| 4906 | (push-mark (point-max) nil t) |
| 4907 | @end smallexample |
| 4908 | |
| 4909 | @noindent |
| 4910 | (The expression works nearly the same as before. It sets a mark at |
| 4911 | the highest numbered place in the buffer that it can. However, in |
| 4912 | this version, @code{push-mark} has two additional arguments. The |
| 4913 | second argument to @code{push-mark} is @code{nil}. This tells the |
| 4914 | function it @emph{should} display a message that says `Mark set' when |
| 4915 | it pushes the mark. The third argument is @code{t}. This tells |
| 4916 | @code{push-mark} to activate the mark when Transient Mark mode is |
| 4917 | turned on. Transient Mark mode highlights the currently active |
| 4918 | region. It is usually turned off.) |
| 4919 | |
| 4920 | Finally, the last line of the function is @code{(goto-char |
| 4921 | (point-min)))}. This is written exactly the same way as it is written |
| 4922 | in @code{beginning-of-buffer}. The expression moves the cursor to |
| 4923 | the minimum point in the buffer, that is, to the beginning of the buffer |
| 4924 | (or to the beginning of the accessible portion of the buffer). As a |
| 4925 | result of this, point is placed at the beginning of the buffer and mark |
| 4926 | is set at the end of the buffer. The whole buffer is, therefore, the |
| 4927 | region. |
| 4928 | |
| 4929 | @node append-to-buffer, Buffer Related Review, mark-whole-buffer, Buffer Walk Through |
| 4930 | @comment node-name, next, previous, up |
| 4931 | @section The Definition of @code{append-to-buffer} |
| 4932 | @findex append-to-buffer |
| 4933 | |
| 4934 | The @code{append-to-buffer} command is very nearly as simple as the |
| 4935 | @code{mark-whole-buffer} command. What it does is copy the region (that |
| 4936 | is, the part of the buffer between point and mark) from the current |
| 4937 | buffer to a specified buffer. |
| 4938 | |
| 4939 | @menu |
| 4940 | * append-to-buffer overview:: |
| 4941 | * append interactive:: A two part interactive expression. |
| 4942 | * append-to-buffer body:: Incorporates a @code{let} expression. |
| 4943 | * append save-excursion:: How the @code{save-excursion} works. |
| 4944 | @end menu |
| 4945 | |
| 4946 | @node append-to-buffer overview, append interactive, append-to-buffer, append-to-buffer |
| 4947 | @ifnottex |
| 4948 | @unnumberedsubsec An Overview of @code{append-to-buffer} |
| 4949 | @end ifnottex |
| 4950 | |
| 4951 | @findex insert-buffer-substring |
| 4952 | The @code{append-to-buffer} command uses the |
| 4953 | @code{insert-buffer-substring} function to copy the region. |
| 4954 | @code{insert-buffer-substring} is described by its name: it takes a |
| 4955 | string of characters from part of a buffer, a ``substring'', and |
| 4956 | inserts them into another buffer. Most of @code{append-to-buffer} is |
| 4957 | concerned with setting up the conditions for |
| 4958 | @code{insert-buffer-substring} to work: the code must specify both the |
| 4959 | buffer to which the text will go and the region that will be copied. |
| 4960 | Here is the complete text of the function: |
| 4961 | |
| 4962 | @smallexample |
| 4963 | @group |
| 4964 | (defun append-to-buffer (buffer start end) |
| 4965 | "Append to specified buffer the text of the region. |
| 4966 | It is inserted into that buffer before its point. |
| 4967 | @end group |
| 4968 | |
| 4969 | @group |
| 4970 | When calling from a program, give three arguments: |
| 4971 | a buffer or the name of one, and two character numbers |
| 4972 | specifying the portion of the current buffer to be copied." |
| 4973 | (interactive "BAppend to buffer:@: \nr") |
| 4974 | (let ((oldbuf (current-buffer))) |
| 4975 | (save-excursion |
| 4976 | (set-buffer (get-buffer-create buffer)) |
| 4977 | (insert-buffer-substring oldbuf start end)))) |
| 4978 | @end group |
| 4979 | @end smallexample |
| 4980 | |
| 4981 | The function can be understood by looking at it as a series of |
| 4982 | filled-in templates. |
| 4983 | |
| 4984 | The outermost template is for the function definition. In this |
| 4985 | function, it looks like this (with several slots filled in): |
| 4986 | |
| 4987 | @smallexample |
| 4988 | @group |
| 4989 | (defun append-to-buffer (buffer start end) |
| 4990 | "@var{documentation}@dots{}" |
| 4991 | (interactive "BAppend to buffer:@: \nr") |
| 4992 | @var{body}@dots{}) |
| 4993 | @end group |
| 4994 | @end smallexample |
| 4995 | |
| 4996 | The first line of the function includes its name and three arguments. |
| 4997 | The arguments are the @code{buffer} to which the text will be copied, and |
| 4998 | the @code{start} and @code{end} of the region in the current buffer that |
| 4999 | will be copied. |
| 5000 | |
| 5001 | The next part of the function is the documentation, which is clear and |
| 5002 | complete. |
| 5003 | |
| 5004 | @node append interactive, append-to-buffer body, append-to-buffer overview, append-to-buffer |
| 5005 | @comment node-name, next, previous, up |
| 5006 | @subsection The @code{append-to-buffer} Interactive Expression |
| 5007 | |
| 5008 | Since the @code{append-to-buffer} function will be used interactively, |
| 5009 | the function must have an @code{interactive} expression. (For a |
| 5010 | review of @code{interactive}, see @ref{Interactive, , Making a |
| 5011 | Function Interactive}.) The expression reads as follows: |
| 5012 | |
| 5013 | @smallexample |
| 5014 | (interactive "BAppend to buffer:@: \nr") |
| 5015 | @end smallexample |
| 5016 | |
| 5017 | @noindent |
| 5018 | This expression has an argument inside of quotation marks and that |
| 5019 | argument has two parts, separated by @samp{\n}. |
| 5020 | |
| 5021 | The first part is @samp{BAppend to buffer:@: }. Here, the @samp{B} |
| 5022 | tells Emacs to ask for the name of the buffer that will be passed to the |
| 5023 | function. Emacs will ask for the name by prompting the user in the |
| 5024 | minibuffer, using the string following the @samp{B}, which is the string |
| 5025 | @samp{Append to buffer:@: }. Emacs then binds the variable @code{buffer} |
| 5026 | in the function's argument list to the specified buffer. |
| 5027 | |
| 5028 | The newline, @samp{\n}, separates the first part of the argument from |
| 5029 | the second part. It is followed by an @samp{r} that tells Emacs to bind |
| 5030 | the two arguments that follow the symbol @code{buffer} in the function's |
| 5031 | argument list (that is, @code{start} and @code{end}) to the values of |
| 5032 | point and mark. |
| 5033 | |
| 5034 | @node append-to-buffer body, append save-excursion, append interactive, append-to-buffer |
| 5035 | @comment node-name, next, previous, up |
| 5036 | @subsection The Body of @code{append-to-buffer} |
| 5037 | |
| 5038 | The body of the @code{append-to-buffer} function begins with @code{let}. |
| 5039 | |
| 5040 | As we have seen before (@pxref{let, , @code{let}}), the purpose of a |
| 5041 | @code{let} expression is to create and give initial values to one or |
| 5042 | more variables that will only be used within the body of the |
| 5043 | @code{let}. This means that such a variable will not be confused with |
| 5044 | any variable of the same name outside the @code{let} expression. |
| 5045 | |
| 5046 | We can see how the @code{let} expression fits into the function as a |
| 5047 | whole by showing a template for @code{append-to-buffer} with the |
| 5048 | @code{let} expression in outline: |
| 5049 | |
| 5050 | @smallexample |
| 5051 | @group |
| 5052 | (defun append-to-buffer (buffer start end) |
| 5053 | "@var{documentation}@dots{}" |
| 5054 | (interactive "BAppend to buffer:@: \nr") |
| 5055 | (let ((@var{variable} @var{value})) |
| 5056 | @var{body}@dots{}) |
| 5057 | @end group |
| 5058 | @end smallexample |
| 5059 | |
| 5060 | The @code{let} expression has three elements: |
| 5061 | |
| 5062 | @enumerate |
| 5063 | @item |
| 5064 | The symbol @code{let}; |
| 5065 | |
| 5066 | @item |
| 5067 | A varlist containing, in this case, a single two-element list, |
| 5068 | @code{(@var{variable} @var{value})}; |
| 5069 | |
| 5070 | @item |
| 5071 | The body of the @code{let} expression. |
| 5072 | @end enumerate |
| 5073 | |
| 5074 | @need 800 |
| 5075 | In the @code{append-to-buffer} function, the varlist looks like this: |
| 5076 | |
| 5077 | @smallexample |
| 5078 | (oldbuf (current-buffer)) |
| 5079 | @end smallexample |
| 5080 | |
| 5081 | @noindent |
| 5082 | In this part of the @code{let} expression, the one variable, |
| 5083 | @code{oldbuf}, is bound to the value returned by the |
| 5084 | @code{(current-buffer)} expression. The variable, @code{oldbuf}, is |
| 5085 | used to keep track of the buffer in which you are working and from |
| 5086 | which you will copy. |
| 5087 | |
| 5088 | The element or elements of a varlist are surrounded by a set of |
| 5089 | parentheses so the Lisp interpreter can distinguish the varlist from |
| 5090 | the body of the @code{let}. As a consequence, the two-element list |
| 5091 | within the varlist is surrounded by a circumscribing set of parentheses. |
| 5092 | The line looks like this: |
| 5093 | |
| 5094 | @smallexample |
| 5095 | @group |
| 5096 | (let ((oldbuf (current-buffer))) |
| 5097 | @dots{} ) |
| 5098 | @end group |
| 5099 | @end smallexample |
| 5100 | |
| 5101 | @noindent |
| 5102 | The two parentheses before @code{oldbuf} might surprise you if you did |
| 5103 | not realize that the first parenthesis before @code{oldbuf} marks the |
| 5104 | boundary of the varlist and the second parenthesis marks the beginning |
| 5105 | of the two-element list, @code{(oldbuf (current-buffer))}. |
| 5106 | |
| 5107 | @node append save-excursion, , append-to-buffer body, append-to-buffer |
| 5108 | @comment node-name, next, previous, up |
| 5109 | @subsection @code{save-excursion} in @code{append-to-buffer} |
| 5110 | |
| 5111 | The body of the @code{let} expression in @code{append-to-buffer} |
| 5112 | consists of a @code{save-excursion} expression. |
| 5113 | |
| 5114 | The @code{save-excursion} function saves the locations of point and |
| 5115 | mark, and restores them to those positions after the expressions in the |
| 5116 | body of the @code{save-excursion} complete execution. In addition, |
| 5117 | @code{save-excursion} keeps track of the original buffer, and |
| 5118 | restores it. This is how @code{save-excursion} is used in |
| 5119 | @code{append-to-buffer}. |
| 5120 | |
| 5121 | @need 1500 |
| 5122 | @cindex Indentation for formatting |
| 5123 | @cindex Formatting convention |
| 5124 | Incidentally, it is worth noting here that a Lisp function is normally |
| 5125 | formatted so that everything that is enclosed in a multi-line spread is |
| 5126 | indented more to the right than the first symbol. In this function |
| 5127 | definition, the @code{let} is indented more than the @code{defun}, and |
| 5128 | the @code{save-excursion} is indented more than the @code{let}, like |
| 5129 | this: |
| 5130 | |
| 5131 | @smallexample |
| 5132 | @group |
| 5133 | (defun @dots{} |
| 5134 | @dots{} |
| 5135 | @dots{} |
| 5136 | (let@dots{} |
| 5137 | (save-excursion |
| 5138 | @dots{} |
| 5139 | @end group |
| 5140 | @end smallexample |
| 5141 | |
| 5142 | @need 1500 |
| 5143 | @noindent |
| 5144 | This formatting convention makes it easy to see that the two lines in |
| 5145 | the body of the @code{save-excursion} are enclosed by the parentheses |
| 5146 | associated with @code{save-excursion}, just as the |
| 5147 | @code{save-excursion} itself is enclosed by the parentheses associated |
| 5148 | with the @code{let}: |
| 5149 | |
| 5150 | @smallexample |
| 5151 | @group |
| 5152 | (let ((oldbuf (current-buffer))) |
| 5153 | (save-excursion |
| 5154 | (set-buffer (get-buffer-create buffer)) |
| 5155 | (insert-buffer-substring oldbuf start end)))) |
| 5156 | @end group |
| 5157 | @end smallexample |
| 5158 | |
| 5159 | @need 1200 |
| 5160 | The use of the @code{save-excursion} function can be viewed as a process |
| 5161 | of filling in the slots of a template: |
| 5162 | |
| 5163 | @smallexample |
| 5164 | @group |
| 5165 | (save-excursion |
| 5166 | @var{first-expression-in-body} |
| 5167 | @var{second-expression-in-body} |
| 5168 | @dots{} |
| 5169 | @var{last-expression-in-body}) |
| 5170 | @end group |
| 5171 | @end smallexample |
| 5172 | |
| 5173 | @need 1200 |
| 5174 | @noindent |
| 5175 | In this function, the body of the @code{save-excursion} contains only |
| 5176 | two expressions. The body looks like this: |
| 5177 | |
| 5178 | @smallexample |
| 5179 | @group |
| 5180 | (set-buffer (get-buffer-create buffer)) |
| 5181 | (insert-buffer-substring oldbuf start end) |
| 5182 | @end group |
| 5183 | @end smallexample |
| 5184 | |
| 5185 | When the @code{append-to-buffer} function is evaluated, the two |
| 5186 | expressions in the body of the @code{save-excursion} are evaluated in |
| 5187 | sequence. The value of the last expression is returned as the value of |
| 5188 | the @code{save-excursion} function; the other expression is evaluated |
| 5189 | only for its side effects. |
| 5190 | |
| 5191 | The first line in the body of the @code{save-excursion} uses the |
| 5192 | @code{set-buffer} function to change the current buffer to the one |
| 5193 | specified in the first argument to @code{append-to-buffer}. (Changing |
| 5194 | the buffer is the side effect; as we have said before, in Lisp, a side |
| 5195 | effect is often the primary thing we want.) The second line does the |
| 5196 | primary work of the function. |
| 5197 | |
| 5198 | The @code{set-buffer} function changes Emacs' attention to the buffer to |
| 5199 | which the text will be copied and from which @code{save-excursion} will |
| 5200 | return. |
| 5201 | |
| 5202 | @need 800 |
| 5203 | The line looks like this: |
| 5204 | |
| 5205 | @smallexample |
| 5206 | (set-buffer (get-buffer-create buffer)) |
| 5207 | @end smallexample |
| 5208 | |
| 5209 | The innermost expression of this list is @code{(get-buffer-create |
| 5210 | buffer)}. This expression uses the @code{get-buffer-create} function, |
| 5211 | which either gets the named buffer, or if it does not exist, creates one |
| 5212 | with the given name. This means you can use @code{append-to-buffer} to |
| 5213 | put text into a buffer that did not previously exist. |
| 5214 | |
| 5215 | @code{get-buffer-create} also keeps @code{set-buffer} from getting an |
| 5216 | unnecessary error: @code{set-buffer} needs a buffer to go to; if you |
| 5217 | were to specify a buffer that does not exist, Emacs would baulk. |
| 5218 | Since @code{get-buffer-create} will create a buffer if none exists, |
| 5219 | @code{set-buffer} is always provided with a buffer. |
| 5220 | |
| 5221 | @need 1250 |
| 5222 | The last line of @code{append-to-buffer} does the work of appending |
| 5223 | the text: |
| 5224 | |
| 5225 | @smallexample |
| 5226 | (insert-buffer-substring oldbuf start end) |
| 5227 | @end smallexample |
| 5228 | |
| 5229 | @noindent |
| 5230 | The @code{insert-buffer-substring} function copies a string @emph{from} |
| 5231 | the buffer specified as its first argument and inserts the string into |
| 5232 | the present buffer. In this case, the argument to |
| 5233 | @code{insert-buffer-substring} is the value of the variable created and |
| 5234 | bound by the @code{let}, namely the value of @code{oldbuf}, which was |
| 5235 | the current buffer when you gave the @code{append-to-buffer} command. |
| 5236 | |
| 5237 | After @code{insert-buffer-substring} has done its work, |
| 5238 | @code{save-excursion} will restore the action to the original buffer and |
| 5239 | @code{append-to-buffer} will have done its job. |
| 5240 | |
| 5241 | @need 800 |
| 5242 | Written in skeletal form, the workings of the body look like this: |
| 5243 | |
| 5244 | @smallexample |
| 5245 | @group |
| 5246 | (let (@var{bind-}@code{oldbuf}@var{-to-value-of-}@code{current-buffer}) |
| 5247 | (save-excursion ; @r{Keep track of buffer.} |
| 5248 | @var{change-buffer} |
| 5249 | @var{insert-substring-from-}@code{oldbuf}@var{-into-buffer}) |
| 5250 | |
| 5251 | @var{change-back-to-original-buffer-when-finished} |
| 5252 | @var{let-the-local-meaning-of-}@code{oldbuf}@var{-disappear-when-finished} |
| 5253 | |
| 5254 | @end group |
| 5255 | @end smallexample |
| 5256 | |
| 5257 | In summary, @code{append-to-buffer} works as follows: it saves the value |
| 5258 | of the current buffer in the variable called @code{oldbuf}. It gets the |
| 5259 | new buffer, creating one if need be, and switches Emacs to it. Using |
| 5260 | the value of @code{oldbuf}, it inserts the region of text from the old |
| 5261 | buffer into the new buffer; and then using @code{save-excursion}, it |
| 5262 | brings you back to your original buffer. |
| 5263 | |
| 5264 | In looking at @code{append-to-buffer}, you have explored a fairly |
| 5265 | complex function. It shows how to use @code{let} and |
| 5266 | @code{save-excursion}, and how to change to and come back from another |
| 5267 | buffer. Many function definitions use @code{let}, |
| 5268 | @code{save-excursion}, and @code{set-buffer} this way. |
| 5269 | |
| 5270 | @node Buffer Related Review, Buffer Exercises, append-to-buffer, Buffer Walk Through |
| 5271 | @comment node-name, next, previous, up |
| 5272 | @section Review |
| 5273 | |
| 5274 | Here is a brief summary of the various functions discussed in this chapter. |
| 5275 | |
| 5276 | @table @code |
| 5277 | @item describe-function |
| 5278 | @itemx describe-variable |
| 5279 | Print the documentation for a function or variable. |
| 5280 | Conventionally bound to @kbd{C-h f} and @kbd{C-h v}. |
| 5281 | |
| 5282 | @item find-tag |
| 5283 | Find the file containing the source for a function or variable and |
| 5284 | switch buffers to it, positioning point at the beginning of the item. |
| 5285 | Conventionally bound to @kbd{M-.} (that's a period following the |
| 5286 | @key{META} key). |
| 5287 | |
| 5288 | @item save-excursion |
| 5289 | Save the location of point and mark and restore their values after the |
| 5290 | arguments to @code{save-excursion} have been evaluated. Also, remember |
| 5291 | the current buffer and return to it. |
| 5292 | |
| 5293 | @item push-mark |
| 5294 | Set mark at a location and record the value of the previous mark on the |
| 5295 | mark ring. The mark is a location in the buffer that will keep its |
| 5296 | relative position even if text is added to or removed from the buffer. |
| 5297 | |
| 5298 | @item goto-char |
| 5299 | Set point to the location specified by the value of the argument, which |
| 5300 | can be a number, a marker, or an expression that returns the number of |
| 5301 | a position, such as @code{(point-min)}. |
| 5302 | |
| 5303 | @item insert-buffer-substring |
| 5304 | Copy a region of text from a buffer that is passed to the function as |
| 5305 | an argument and insert the region into the current buffer. |
| 5306 | |
| 5307 | @item mark-whole-buffer |
| 5308 | Mark the whole buffer as a region. Normally bound to @kbd{C-x h}. |
| 5309 | |
| 5310 | @item set-buffer |
| 5311 | Switch the attention of Emacs to another buffer, but do not change the |
| 5312 | window being displayed. Used when the program rather than a human is |
| 5313 | to work on a different buffer. |
| 5314 | |
| 5315 | @item get-buffer-create |
| 5316 | @itemx get-buffer |
| 5317 | Find a named buffer or create one if a buffer of that name does not |
| 5318 | exist. The @code{get-buffer} function returns @code{nil} if the named |
| 5319 | buffer does not exist. |
| 5320 | @end table |
| 5321 | |
| 5322 | @need 1500 |
| 5323 | @node Buffer Exercises, , Buffer Related Review, Buffer Walk Through |
| 5324 | @section Exercises |
| 5325 | |
| 5326 | @itemize @bullet |
| 5327 | @item |
| 5328 | Write your own @code{simplified-end-of-buffer} function definition; |
| 5329 | then test it to see whether it works. |
| 5330 | |
| 5331 | @item |
| 5332 | Use @code{if} and @code{get-buffer} to write a function that prints a |
| 5333 | message telling you whether a buffer exists. |
| 5334 | |
| 5335 | @item |
| 5336 | Using @code{find-tag}, find the source for the @code{copy-to-buffer} |
| 5337 | function. |
| 5338 | @end itemize |
| 5339 | |
| 5340 | @node More Complex, Narrowing & Widening, Buffer Walk Through, Top |
| 5341 | @comment node-name, next, previous, up |
| 5342 | @chapter A Few More Complex Functions |
| 5343 | |
| 5344 | In this chapter, we build on what we have learned in previous chapters |
| 5345 | by looking at more complex functions. The @code{copy-to-buffer} |
| 5346 | function illustrates use of two @code{save-excursion} expressions in |
| 5347 | one definition, while the @code{insert-buffer} function illustrates |
| 5348 | use of an asterisk in an @code{interactive} expression, use of |
| 5349 | @code{or}, and the important distinction between a name and the object |
| 5350 | to which the name refers. |
| 5351 | |
| 5352 | @menu |
| 5353 | * copy-to-buffer:: With @code{set-buffer}, @code{get-buffer-create}. |
| 5354 | * insert-buffer:: Read-only, and with @code{or}. |
| 5355 | * beginning-of-buffer:: Shows @code{goto-char}, |
| 5356 | @code{point-min}, and @code{push-mark}. |
| 5357 | * Second Buffer Related Review:: |
| 5358 | * optional Exercise:: |
| 5359 | @end menu |
| 5360 | |
| 5361 | @node copy-to-buffer, insert-buffer, More Complex, More Complex |
| 5362 | @comment node-name, next, previous, up |
| 5363 | @section The Definition of @code{copy-to-buffer} |
| 5364 | @findex copy-to-buffer |
| 5365 | |
| 5366 | After understanding how @code{append-to-buffer} works, it is easy to |
| 5367 | understand @code{copy-to-buffer}. This function copies text into a |
| 5368 | buffer, but instead of adding to the second buffer, it replaces the |
| 5369 | previous text in the second buffer. The code for the |
| 5370 | @code{copy-to-buffer} function is almost the same as the code for |
| 5371 | @code{append-to-buffer}, except that @code{erase-buffer} and a second |
| 5372 | @code{save-excursion} are used. (@xref{append-to-buffer, , The |
| 5373 | Definition of @code{append-to-buffer}}, for the description of |
| 5374 | @code{append-to-buffer}.) |
| 5375 | |
| 5376 | @need 800 |
| 5377 | The body of @code{copy-to-buffer} looks like this |
| 5378 | |
| 5379 | @smallexample |
| 5380 | @group |
| 5381 | @dots{} |
| 5382 | (interactive "BCopy to buffer:@: \nr") |
| 5383 | (let ((oldbuf (current-buffer))) |
| 5384 | (save-excursion |
| 5385 | (set-buffer (get-buffer-create buffer)) |
| 5386 | (erase-buffer) |
| 5387 | (save-excursion |
| 5388 | (insert-buffer-substring oldbuf start end))))) |
| 5389 | @end group |
| 5390 | @end smallexample |
| 5391 | |
| 5392 | This code is similar to the code in @code{append-to-buffer}: it is |
| 5393 | only after changing to the buffer to which the text will be copied |
| 5394 | that the definition for this function diverges from the definition for |
| 5395 | @code{append-to-buffer}: the @code{copy-to-buffer} function erases the |
| 5396 | buffer's former contents. (This is what is meant by `replacement'; to |
| 5397 | replace text, Emacs erases the previous text and then inserts new |
| 5398 | text.) After erasing the previous contents of the buffer, |
| 5399 | @code{save-excursion} is used for a second time and the new text is |
| 5400 | inserted. |
| 5401 | |
| 5402 | Why is @code{save-excursion} used twice? Consider again what the |
| 5403 | function does. |
| 5404 | |
| 5405 | @need 1250 |
| 5406 | In outline, the body of @code{copy-to-buffer} looks like this: |
| 5407 | |
| 5408 | @smallexample |
| 5409 | @group |
| 5410 | (let (@var{bind-}@code{oldbuf}@var{-to-value-of-}@code{current-buffer}) |
| 5411 | (save-excursion ; @r{First use of @code{save-excursion}.} |
| 5412 | @var{change-buffer} |
| 5413 | (erase-buffer) |
| 5414 | (save-excursion ; @r{Second use of @code{save-excursion}.} |
| 5415 | @var{insert-substring-from-}@code{oldbuf}@var{-into-buffer}))) |
| 5416 | @end group |
| 5417 | @end smallexample |
| 5418 | |
| 5419 | The first use of @code{save-excursion} returns Emacs to the buffer from |
| 5420 | which the text is being copied. That is clear, and is just like its use |
| 5421 | in @code{append-to-buffer}. Why the second use? The reason is that |
| 5422 | @code{insert-buffer-substring} always leaves point at the @emph{end} of |
| 5423 | the region being inserted. The second @code{save-excursion} causes |
| 5424 | Emacs to leave point at the beginning of the text being inserted. In |
| 5425 | most circumstances, users prefer to find point at the beginning of |
| 5426 | inserted text. (Of course, the @code{copy-to-buffer} function returns |
| 5427 | the user to the original buffer when done---but if the user @emph{then} |
| 5428 | switches to the copied-to buffer, point will go to the beginning of the |
| 5429 | text. Thus, this use of a second @code{save-excursion} is a little |
| 5430 | nicety.) |
| 5431 | |
| 5432 | @node insert-buffer, beginning-of-buffer, copy-to-buffer, More Complex |
| 5433 | @comment node-name, next, previous, up |
| 5434 | @section The Definition of @code{insert-buffer} |
| 5435 | @findex insert-buffer |
| 5436 | |
| 5437 | @code{insert-buffer} is yet another buffer-related function. This |
| 5438 | command copies another buffer @emph{into} the current buffer. It is the |
| 5439 | reverse of @code{append-to-buffer} or @code{copy-to-buffer}, since they |
| 5440 | copy a region of text @emph{from} the current buffer to another buffer. |
| 5441 | |
| 5442 | Here is a discussion based on the original code. The code was |
| 5443 | simplified in 2003 and is harder to understand. |
| 5444 | |
| 5445 | In addition, this code illustrates the use of @code{interactive} with a |
| 5446 | buffer that might be @dfn{read-only} and the important distinction |
| 5447 | between the name of an object and the object actually referred to. |
| 5448 | |
| 5449 | @menu |
| 5450 | * insert-buffer code:: |
| 5451 | * insert-buffer interactive:: When you can read, but not write. |
| 5452 | * insert-buffer body:: The body has an @code{or} and a @code{let}. |
| 5453 | * if & or:: Using an @code{if} instead of an @code{or}. |
| 5454 | * Insert or:: How the @code{or} expression works. |
| 5455 | * Insert let:: Two @code{save-excursion} expressions. |
| 5456 | @end menu |
| 5457 | |
| 5458 | @node insert-buffer code, insert-buffer interactive, insert-buffer, insert-buffer |
| 5459 | @ifnottex |
| 5460 | @unnumberedsubsec The Code for @code{insert-buffer} |
| 5461 | @end ifnottex |
| 5462 | |
| 5463 | @need 800 |
| 5464 | Here is the code: |
| 5465 | |
| 5466 | @smallexample |
| 5467 | @group |
| 5468 | (defun insert-buffer (buffer) |
| 5469 | "Insert after point the contents of BUFFER. |
| 5470 | Puts mark after the inserted text. |
| 5471 | BUFFER may be a buffer or a buffer name." |
| 5472 | (interactive "*bInsert buffer:@: ") |
| 5473 | @end group |
| 5474 | @group |
| 5475 | (or (bufferp buffer) |
| 5476 | (setq buffer (get-buffer buffer))) |
| 5477 | (let (start end newmark) |
| 5478 | (save-excursion |
| 5479 | (save-excursion |
| 5480 | (set-buffer buffer) |
| 5481 | (setq start (point-min) end (point-max))) |
| 5482 | @end group |
| 5483 | @group |
| 5484 | (insert-buffer-substring buffer start end) |
| 5485 | (setq newmark (point))) |
| 5486 | (push-mark newmark))) |
| 5487 | @end group |
| 5488 | @end smallexample |
| 5489 | |
| 5490 | @need 1200 |
| 5491 | As with other function definitions, you can use a template to see an |
| 5492 | outline of the function: |
| 5493 | |
| 5494 | @smallexample |
| 5495 | @group |
| 5496 | (defun insert-buffer (buffer) |
| 5497 | "@var{documentation}@dots{}" |
| 5498 | (interactive "*bInsert buffer:@: ") |
| 5499 | @var{body}@dots{}) |
| 5500 | @end group |
| 5501 | @end smallexample |
| 5502 | |
| 5503 | @node insert-buffer interactive, insert-buffer body, insert-buffer code, insert-buffer |
| 5504 | @comment node-name, next, previous, up |
| 5505 | @subsection The Interactive Expression in @code{insert-buffer} |
| 5506 | @findex interactive, @r{example use of} |
| 5507 | |
| 5508 | In @code{insert-buffer}, the argument to the @code{interactive} |
| 5509 | declaration has two parts, an asterisk, @samp{*}, and @samp{bInsert |
| 5510 | buffer:@: }. |
| 5511 | |
| 5512 | @menu |
| 5513 | * Read-only buffer:: When a buffer cannot be modified. |
| 5514 | * b for interactive:: An existing buffer or else its name. |
| 5515 | @end menu |
| 5516 | |
| 5517 | @node Read-only buffer, b for interactive, insert-buffer interactive, insert-buffer interactive |
| 5518 | @comment node-name, next, previous, up |
| 5519 | @unnumberedsubsubsec A Read-only Buffer |
| 5520 | @cindex Read-only buffer |
| 5521 | @cindex Asterisk for read-only buffer |
| 5522 | @findex * @r{for read-only buffer} |
| 5523 | |
| 5524 | The asterisk is for the situation when the current buffer is a |
| 5525 | read-only buffer---a buffer that cannot be modified. If |
| 5526 | @code{insert-buffer} is called when the current buffer is read-only, a |
| 5527 | message to this effect is printed in the echo area and the terminal |
| 5528 | may beep or blink at you; you will not be permitted to insert anything |
| 5529 | into current buffer. The asterisk does not need to be followed by a |
| 5530 | newline to separate it from the next argument. |
| 5531 | |
| 5532 | @node b for interactive, , Read-only buffer, insert-buffer interactive |
| 5533 | @comment node-name, next, previous, up |
| 5534 | @unnumberedsubsubsec @samp{b} in an Interactive Expression |
| 5535 | |
| 5536 | The next argument in the interactive expression starts with a lower |
| 5537 | case @samp{b}. (This is different from the code for |
| 5538 | @code{append-to-buffer}, which uses an upper-case @samp{B}. |
| 5539 | @xref{append-to-buffer, , The Definition of @code{append-to-buffer}}.) |
| 5540 | The lower-case @samp{b} tells the Lisp interpreter that the argument |
| 5541 | for @code{insert-buffer} should be an existing buffer or else its |
| 5542 | name. (The upper-case @samp{B} option provides for the possibility |
| 5543 | that the buffer does not exist.) Emacs will prompt you for the name |
| 5544 | of the buffer, offering you a default buffer, with name completion |
| 5545 | enabled. If the buffer does not exist, you receive a message that |
| 5546 | says ``No match''; your terminal may beep at you as well. |
| 5547 | |
| 5548 | @node insert-buffer body, if & or, insert-buffer interactive, insert-buffer |
| 5549 | @comment node-name, next, previous, up |
| 5550 | @subsection The Body of the @code{insert-buffer} Function |
| 5551 | |
| 5552 | The body of the @code{insert-buffer} function has two major parts: an |
| 5553 | @code{or} expression and a @code{let} expression. The purpose of the |
| 5554 | @code{or} expression is to ensure that the argument @code{buffer} is |
| 5555 | bound to a buffer and not just the name of a buffer. The body of the |
| 5556 | @code{let} expression contains the code which copies the other buffer |
| 5557 | into the current buffer. |
| 5558 | |
| 5559 | @need 1250 |
| 5560 | In outline, the two expressions fit into the @code{insert-buffer} |
| 5561 | function like this: |
| 5562 | |
| 5563 | @smallexample |
| 5564 | @group |
| 5565 | (defun insert-buffer (buffer) |
| 5566 | "@var{documentation}@dots{}" |
| 5567 | (interactive "*bInsert buffer:@: ") |
| 5568 | (or @dots{} |
| 5569 | @dots{} |
| 5570 | @end group |
| 5571 | @group |
| 5572 | (let (@var{varlist}) |
| 5573 | @var{body-of-}@code{let}@dots{} ) |
| 5574 | @end group |
| 5575 | @end smallexample |
| 5576 | |
| 5577 | To understand how the @code{or} expression ensures that the argument |
| 5578 | @code{buffer} is bound to a buffer and not to the name of a buffer, it |
| 5579 | is first necessary to understand the @code{or} function. |
| 5580 | |
| 5581 | Before doing this, let me rewrite this part of the function using |
| 5582 | @code{if} so that you can see what is done in a manner that will be familiar. |
| 5583 | |
| 5584 | @node if & or, Insert or, insert-buffer body, insert-buffer |
| 5585 | @comment node-name, next, previous, up |
| 5586 | @subsection @code{insert-buffer} With an @code{if} Instead of an @code{or} |
| 5587 | |
| 5588 | The job to be done is to make sure the value of @code{buffer} is a |
| 5589 | buffer itself and not the name of a buffer. If the value is the name, |
| 5590 | then the buffer itself must be got. |
| 5591 | |
| 5592 | You can imagine yourself at a conference where an usher is wandering |
| 5593 | around holding a list with your name on it and looking for you: the |
| 5594 | usher is ``bound'' to your name, not to you; but when the usher finds |
| 5595 | you and takes your arm, the usher becomes ``bound'' to you. |
| 5596 | |
| 5597 | @need 800 |
| 5598 | In Lisp, you might describe this situation like this: |
| 5599 | |
| 5600 | @smallexample |
| 5601 | @group |
| 5602 | (if (not (holding-on-to-guest)) |
| 5603 | (find-and-take-arm-of-guest)) |
| 5604 | @end group |
| 5605 | @end smallexample |
| 5606 | |
| 5607 | We want to do the same thing with a buffer---if we do not have the |
| 5608 | buffer itself, we want to get it. |
| 5609 | |
| 5610 | @need 1200 |
| 5611 | Using a predicate called @code{bufferp} that tells us whether we have a |
| 5612 | buffer (rather than its name), we can write the code like this: |
| 5613 | |
| 5614 | @smallexample |
| 5615 | @group |
| 5616 | (if (not (bufferp buffer)) ; @r{if-part} |
| 5617 | (setq buffer (get-buffer buffer))) ; @r{then-part} |
| 5618 | @end group |
| 5619 | @end smallexample |
| 5620 | |
| 5621 | @noindent |
| 5622 | Here, the true-or-false-test of the @code{if} expression is |
| 5623 | @w{@code{(not (bufferp buffer))}}; and the then-part is the expression |
| 5624 | @w{@code{(setq buffer (get-buffer buffer))}}. |
| 5625 | |
| 5626 | In the test, the function @code{bufferp} returns true if its argument is |
| 5627 | a buffer---but false if its argument is the name of the buffer. (The |
| 5628 | last character of the function name @code{bufferp} is the character |
| 5629 | @samp{p}; as we saw earlier, such use of @samp{p} is a convention that |
| 5630 | indicates that the function is a predicate, which is a term that means |
| 5631 | that the function will determine whether some property is true or false. |
| 5632 | @xref{Wrong Type of Argument, , Using the Wrong Type Object as an |
| 5633 | Argument}.) |
| 5634 | |
| 5635 | @need 1200 |
| 5636 | The function @code{not} precedes the expression @code{(bufferp buffer)}, |
| 5637 | so the true-or-false-test looks like this: |
| 5638 | |
| 5639 | @smallexample |
| 5640 | (not (bufferp buffer)) |
| 5641 | @end smallexample |
| 5642 | |
| 5643 | @noindent |
| 5644 | @code{not} is a function that returns true if its argument is false |
| 5645 | and false if its argument is true. So if @code{(bufferp buffer)} |
| 5646 | returns true, the @code{not} expression returns false and vice-versa: |
| 5647 | what is ``not true'' is false and what is ``not false'' is true. |
| 5648 | |
| 5649 | Using this test, the @code{if} expression works as follows: when the |
| 5650 | value of the variable @code{buffer} is actually a buffer rather than |
| 5651 | its name, the true-or-false-test returns false and the @code{if} |
| 5652 | expression does not evaluate the then-part. This is fine, since we do |
| 5653 | not need to do anything to the variable @code{buffer} if it really is |
| 5654 | a buffer. |
| 5655 | |
| 5656 | On the other hand, when the value of @code{buffer} is not a buffer |
| 5657 | itself, but the name of a buffer, the true-or-false-test returns true |
| 5658 | and the then-part of the expression is evaluated. In this case, the |
| 5659 | then-part is @code{(setq buffer (get-buffer buffer))}. This |
| 5660 | expression uses the @code{get-buffer} function to return an actual |
| 5661 | buffer itself, given its name. The @code{setq} then sets the variable |
| 5662 | @code{buffer} to the value of the buffer itself, replacing its previous |
| 5663 | value (which was the name of the buffer). |
| 5664 | |
| 5665 | @node Insert or, Insert let, if & or, insert-buffer |
| 5666 | @comment node-name, next, previous, up |
| 5667 | @subsection The @code{or} in the Body |
| 5668 | |
| 5669 | The purpose of the @code{or} expression in the @code{insert-buffer} |
| 5670 | function is to ensure that the argument @code{buffer} is bound to a |
| 5671 | buffer and not just to the name of a buffer. The previous section shows |
| 5672 | how the job could have been done using an @code{if} expression. |
| 5673 | However, the @code{insert-buffer} function actually uses @code{or}. |
| 5674 | To understand this, it is necessary to understand how @code{or} works. |
| 5675 | |
| 5676 | @findex or |
| 5677 | An @code{or} function can have any number of arguments. It evaluates |
| 5678 | each argument in turn and returns the value of the first of its |
| 5679 | arguments that is not @code{nil}. Also, and this is a crucial feature |
| 5680 | of @code{or}, it does not evaluate any subsequent arguments after |
| 5681 | returning the first non-@code{nil} value. |
| 5682 | |
| 5683 | @need 800 |
| 5684 | The @code{or} expression looks like this: |
| 5685 | |
| 5686 | @smallexample |
| 5687 | @group |
| 5688 | (or (bufferp buffer) |
| 5689 | (setq buffer (get-buffer buffer))) |
| 5690 | @end group |
| 5691 | @end smallexample |
| 5692 | |
| 5693 | @noindent |
| 5694 | The first argument to @code{or} is the expression @code{(bufferp buffer)}. |
| 5695 | This expression returns true (a non-@code{nil} value) if the buffer is |
| 5696 | actually a buffer, and not just the name of a buffer. In the @code{or} |
| 5697 | expression, if this is the case, the @code{or} expression returns this |
| 5698 | true value and does not evaluate the next expression---and this is fine |
| 5699 | with us, since we do not want to do anything to the value of |
| 5700 | @code{buffer} if it really is a buffer. |
| 5701 | |
| 5702 | On the other hand, if the value of @code{(bufferp buffer)} is @code{nil}, |
| 5703 | which it will be if the value of @code{buffer} is the name of a buffer, |
| 5704 | the Lisp interpreter evaluates the next element of the @code{or} |
| 5705 | expression. This is the expression @code{(setq buffer (get-buffer |
| 5706 | buffer))}. This expression returns a non-@code{nil} value, which |
| 5707 | is the value to which it sets the variable @code{buffer}---and this |
| 5708 | value is a buffer itself, not the name of a buffer. |
| 5709 | |
| 5710 | The result of all this is that the symbol @code{buffer} is always |
| 5711 | bound to a buffer itself rather than to the name of a buffer. All |
| 5712 | this is necessary because the @code{set-buffer} function in a |
| 5713 | following line only works with a buffer itself, not with the name to a |
| 5714 | buffer. |
| 5715 | |
| 5716 | @need 1250 |
| 5717 | Incidentally, using @code{or}, the situation with the usher would be |
| 5718 | written like this: |
| 5719 | |
| 5720 | @smallexample |
| 5721 | (or (holding-on-to-guest) (find-and-take-arm-of-guest)) |
| 5722 | @end smallexample |
| 5723 | |
| 5724 | @node Insert let, , Insert or, insert-buffer |
| 5725 | @comment node-name, next, previous, up |
| 5726 | @subsection The @code{let} Expression in @code{insert-buffer} |
| 5727 | |
| 5728 | After ensuring that the variable @code{buffer} refers to a buffer itself |
| 5729 | and not just to the name of a buffer, the @code{insert-buffer function} |
| 5730 | continues with a @code{let} expression. This specifies three local |
| 5731 | variables, @code{start}, @code{end}, and @code{newmark} and binds them |
| 5732 | to the initial value @code{nil}. These variables are used inside the |
| 5733 | remainder of the @code{let} and temporarily hide any other occurrence of |
| 5734 | variables of the same name in Emacs until the end of the @code{let}. |
| 5735 | |
| 5736 | @need 1200 |
| 5737 | The body of the @code{let} contains two @code{save-excursion} |
| 5738 | expressions. First, we will look at the inner @code{save-excursion} |
| 5739 | expression in detail. The expression looks like this: |
| 5740 | |
| 5741 | @smallexample |
| 5742 | @group |
| 5743 | (save-excursion |
| 5744 | (set-buffer buffer) |
| 5745 | (setq start (point-min) end (point-max))) |
| 5746 | @end group |
| 5747 | @end smallexample |
| 5748 | |
| 5749 | @noindent |
| 5750 | The expression @code{(set-buffer buffer)} changes Emacs' attention |
| 5751 | from the current buffer to the one from which the text will copied. |
| 5752 | In that buffer, the variables @code{start} and @code{end} are set to |
| 5753 | the beginning and end of the buffer, using the commands |
| 5754 | @code{point-min} and @code{point-max}. Note that we have here an |
| 5755 | illustration of how @code{setq} is able to set two variables in the |
| 5756 | same expression. The first argument of @code{setq} is set to the |
| 5757 | value of its second, and its third argument is set to the value of its |
| 5758 | fourth. |
| 5759 | |
| 5760 | After the body of the inner @code{save-excursion} is evaluated, the |
| 5761 | @code{save-excursion} restores the original buffer, but @code{start} and |
| 5762 | @code{end} remain set to the values of the beginning and end of the |
| 5763 | buffer from which the text will be copied. |
| 5764 | |
| 5765 | @need 1250 |
| 5766 | The outer @code{save-excursion} expression looks like this: |
| 5767 | |
| 5768 | @smallexample |
| 5769 | @group |
| 5770 | (save-excursion |
| 5771 | (@var{inner-}@code{save-excursion}@var{-expression} |
| 5772 | (@var{go-to-new-buffer-and-set-}@code{start}@var{-and-}@code{end}) |
| 5773 | (insert-buffer-substring buffer start end) |
| 5774 | (setq newmark (point))) |
| 5775 | @end group |
| 5776 | @end smallexample |
| 5777 | |
| 5778 | @noindent |
| 5779 | The @code{insert-buffer-substring} function copies the text |
| 5780 | @emph{into} the current buffer @emph{from} the region indicated by |
| 5781 | @code{start} and @code{end} in @code{buffer}. Since the whole of the |
| 5782 | second buffer lies between @code{start} and @code{end}, the whole of |
| 5783 | the second buffer is copied into the buffer you are editing. Next, |
| 5784 | the value of point, which will be at the end of the inserted text, is |
| 5785 | recorded in the variable @code{newmark}. |
| 5786 | |
| 5787 | After the body of the outer @code{save-excursion} is evaluated, point |
| 5788 | and mark are relocated to their original places. |
| 5789 | |
| 5790 | However, it is convenient to locate a mark at the end of the newly |
| 5791 | inserted text and locate point at its beginning. The @code{newmark} |
| 5792 | variable records the end of the inserted text. In the last line of |
| 5793 | the @code{let} expression, the @code{(push-mark newmark)} expression |
| 5794 | function sets a mark to this location. (The previous location of the |
| 5795 | mark is still accessible; it is recorded on the mark ring and you can |
| 5796 | go back to it with @kbd{C-u C-@key{SPC}}.) Meanwhile, point is |
| 5797 | located at the beginning of the inserted text, which is where it was |
| 5798 | before you called the insert function, the position of which was saved |
| 5799 | by the first @code{save-excursion}. |
| 5800 | |
| 5801 | @need 1250 |
| 5802 | The whole @code{let} expression looks like this: |
| 5803 | |
| 5804 | @smallexample |
| 5805 | @group |
| 5806 | (let (start end newmark) |
| 5807 | (save-excursion |
| 5808 | (save-excursion |
| 5809 | (set-buffer buffer) |
| 5810 | (setq start (point-min) end (point-max))) |
| 5811 | (insert-buffer-substring buffer start end) |
| 5812 | (setq newmark (point))) |
| 5813 | (push-mark newmark)) |
| 5814 | @end group |
| 5815 | @end smallexample |
| 5816 | |
| 5817 | Like the @code{append-to-buffer} function, the @code{insert-buffer} |
| 5818 | function uses @code{let}, @code{save-excursion}, and |
| 5819 | @code{set-buffer}. In addition, the function illustrates one way to |
| 5820 | use @code{or}. All these functions are building blocks that we will |
| 5821 | find and use again and again. |
| 5822 | |
| 5823 | @node beginning-of-buffer, Second Buffer Related Review, insert-buffer, More Complex |
| 5824 | @comment node-name, next, previous, up |
| 5825 | @section Complete Definition of @code{beginning-of-buffer} |
| 5826 | @findex beginning-of-buffer |
| 5827 | |
| 5828 | The basic structure of the @code{beginning-of-buffer} function has |
| 5829 | already been discussed. (@xref{simplified-beginning-of-buffer, , A |
| 5830 | Simplified @code{beginning-of-buffer} Definition}.) |
| 5831 | This section describes the complex part of the definition. |
| 5832 | |
| 5833 | As previously described, when invoked without an argument, |
| 5834 | @code{beginning-of-buffer} moves the cursor to the beginning of the |
| 5835 | buffer, leaving the mark at the previous position. However, when the |
| 5836 | command is invoked with a number between one and ten, the function |
| 5837 | considers that number to be a fraction of the length of the buffer, |
| 5838 | measured in tenths, and Emacs moves the cursor that fraction of the way |
| 5839 | from the beginning of the buffer. Thus, you can either call this |
| 5840 | function with the key command @kbd{M-<}, which will move the cursor to |
| 5841 | the beginning of the buffer, or with a key command such as @kbd{C-u 7 |
| 5842 | M-<} which will move the cursor to a point 70% of the way through the |
| 5843 | buffer. If a number bigger than ten is used for the argument, it moves |
| 5844 | to the end of the buffer. |
| 5845 | |
| 5846 | The @code{beginning-of-buffer} function can be called with or without an |
| 5847 | argument. The use of the argument is optional. |
| 5848 | |
| 5849 | @menu |
| 5850 | * Optional Arguments:: |
| 5851 | * beginning-of-buffer opt arg:: Example with optional argument. |
| 5852 | * beginning-of-buffer complete:: |
| 5853 | @end menu |
| 5854 | |
| 5855 | @node Optional Arguments, beginning-of-buffer opt arg, beginning-of-buffer, beginning-of-buffer |
| 5856 | @subsection Optional Arguments |
| 5857 | |
| 5858 | Unless told otherwise, Lisp expects that a function with an argument in |
| 5859 | its function definition will be called with a value for that argument. |
| 5860 | If that does not happen, you get an error and a message that says |
| 5861 | @samp{Wrong number of arguments}. |
| 5862 | |
| 5863 | @cindex Optional arguments |
| 5864 | @cindex Keyword |
| 5865 | @findex optional |
| 5866 | However, optional arguments are a feature of Lisp: a @dfn{keyword} may |
| 5867 | be used to tell the Lisp interpreter that an argument is optional. |
| 5868 | The keyword is @code{&optional}. (The @samp{&} in front of |
| 5869 | @samp{optional} is part of the keyword.) In a function definition, if |
| 5870 | an argument follows the keyword @code{&optional}, a value does not |
| 5871 | need to be passed to that argument when the function is called. |
| 5872 | |
| 5873 | @need 1200 |
| 5874 | The first line of the function definition of @code{beginning-of-buffer} |
| 5875 | therefore looks like this: |
| 5876 | |
| 5877 | @smallexample |
| 5878 | (defun beginning-of-buffer (&optional arg) |
| 5879 | @end smallexample |
| 5880 | |
| 5881 | @need 1250 |
| 5882 | In outline, the whole function looks like this: |
| 5883 | |
| 5884 | @smallexample |
| 5885 | @group |
| 5886 | (defun beginning-of-buffer (&optional arg) |
| 5887 | "@var{documentation}@dots{}" |
| 5888 | (interactive "P") |
| 5889 | (push-mark) |
| 5890 | (goto-char |
| 5891 | (@var{if-there-is-an-argument} |
| 5892 | @var{figure-out-where-to-go} |
| 5893 | @var{else-go-to} |
| 5894 | (point-min)))) |
| 5895 | @end group |
| 5896 | @end smallexample |
| 5897 | |
| 5898 | The function is similar to the @code{simplified-beginning-of-buffer} |
| 5899 | function except that the @code{interactive} expression has @code{"P"} |
| 5900 | as an argument and the @code{goto-char} function is followed by an |
| 5901 | if-then-else expression that figures out where to put the cursor if |
| 5902 | there is an argument. |
| 5903 | |
| 5904 | The @code{"P"} in the @code{interactive} expression tells Emacs to pass |
| 5905 | a prefix argument, if there is one, to the function. A prefix argument |
| 5906 | is made by typing the @key{META} key followed by a number, or by typing |
| 5907 | @kbd{C-u} and then a number (if you don't type a number, @kbd{C-u} |
| 5908 | defaults to 4). |
| 5909 | |
| 5910 | The true-or-false-test of the @code{if} expression is simple: it is |
| 5911 | simply the argument @code{arg}. If @code{arg} has a value that is not |
| 5912 | @code{nil}, which will be the case if @code{beginning-of-buffer} is |
| 5913 | called with an argument, then this true-or-false-test will return true |
| 5914 | and the then-part of the @code{if} expression will be evaluated. On the |
| 5915 | other hand, if @code{beginning-of-buffer} is not called with an |
| 5916 | argument, the value of @code{arg} will be @code{nil} and the else-part |
| 5917 | of the @code{if} expression will be evaluated. The else-part is simply |
| 5918 | @code{point-min}, and when this is the outcome, the whole |
| 5919 | @code{goto-char} expression is @code{(goto-char (point-min))}, which is |
| 5920 | how we saw the @code{beginning-of-buffer} function in its simplified |
| 5921 | form. |
| 5922 | |
| 5923 | @node beginning-of-buffer opt arg, beginning-of-buffer complete, Optional Arguments, beginning-of-buffer |
| 5924 | @subsection @code{beginning-of-buffer} with an Argument |
| 5925 | |
| 5926 | When @code{beginning-of-buffer} is called with an argument, an |
| 5927 | expression is evaluated which calculates what value to pass to |
| 5928 | @code{goto-char}. This expression is rather complicated at first sight. |
| 5929 | It includes an inner @code{if} expression and much arithmetic. It looks |
| 5930 | like this: |
| 5931 | |
| 5932 | @smallexample |
| 5933 | @group |
| 5934 | (if (> (buffer-size) 10000) |
| 5935 | ;; @r{Avoid overflow for large buffer sizes!} |
| 5936 | (* (prefix-numeric-value arg) (/ (buffer-size) 10)) |
| 5937 | (/ |
| 5938 | (+ 10 |
| 5939 | (* |
| 5940 | (buffer-size) (prefix-numeric-value arg))) 10)) |
| 5941 | @end group |
| 5942 | @end smallexample |
| 5943 | |
| 5944 | @menu |
| 5945 | * Disentangle beginning-of-buffer:: |
| 5946 | * Large buffer case:: |
| 5947 | * Small buffer case:: |
| 5948 | @end menu |
| 5949 | |
| 5950 | @node Disentangle beginning-of-buffer, Large buffer case, beginning-of-buffer opt arg, beginning-of-buffer opt arg |
| 5951 | @ifnottex |
| 5952 | @unnumberedsubsubsec Disentangle @code{beginning-of-buffer} |
| 5953 | @end ifnottex |
| 5954 | |
| 5955 | Like other complex-looking expressions, the conditional expression |
| 5956 | within @code{beginning-of-buffer} can be disentangled by looking at it |
| 5957 | as parts of a template, in this case, the template for an if-then-else |
| 5958 | expression. In skeletal form, the expression looks like this: |
| 5959 | |
| 5960 | @smallexample |
| 5961 | @group |
| 5962 | (if (@var{buffer-is-large} |
| 5963 | @var{divide-buffer-size-by-10-and-multiply-by-arg} |
| 5964 | @var{else-use-alternate-calculation} |
| 5965 | @end group |
| 5966 | @end smallexample |
| 5967 | |
| 5968 | The true-or-false-test of this inner @code{if} expression checks the |
| 5969 | size of the buffer. The reason for this is that the old Version 18 |
| 5970 | Emacs used numbers that are no bigger than eight million or so |
| 5971 | and in the computation that followed, the programmer feared that Emacs |
| 5972 | might try to use over-large numbers if the buffer were large. The |
| 5973 | term `overflow', mentioned in the comment, means numbers that are over |
| 5974 | large. Version 21 Emacs uses larger numbers, but this code has not |
| 5975 | been touched, if only because people now look at buffers that are far, |
| 5976 | far larger than ever before. |
| 5977 | |
| 5978 | There are two cases: if the buffer is large and if it is not. |
| 5979 | |
| 5980 | @node Large buffer case, Small buffer case, Disentangle beginning-of-buffer, beginning-of-buffer opt arg |
| 5981 | @comment node-name, next, previous, up |
| 5982 | @unnumberedsubsubsec What happens in a large buffer |
| 5983 | |
| 5984 | In @code{beginning-of-buffer}, the inner @code{if} expression tests |
| 5985 | whether the size of the buffer is greater than 10,000 characters. To do |
| 5986 | this, it uses the @code{>} function and the @code{buffer-size} function. |
| 5987 | |
| 5988 | @need 800 |
| 5989 | The line looks like this: |
| 5990 | |
| 5991 | @smallexample |
| 5992 | (if (> (buffer-size) 10000) |
| 5993 | @end smallexample |
| 5994 | |
| 5995 | @need 1200 |
| 5996 | @noindent |
| 5997 | When the buffer is large, the then-part of the @code{if} expression is |
| 5998 | evaluated. It reads like this (after formatting for easy reading): |
| 5999 | |
| 6000 | @smallexample |
| 6001 | @group |
| 6002 | (* |
| 6003 | (prefix-numeric-value arg) |
| 6004 | (/ (buffer-size) 10)) |
| 6005 | @end group |
| 6006 | @end smallexample |
| 6007 | |
| 6008 | @noindent |
| 6009 | This expression is a multiplication, with two arguments to the function |
| 6010 | @code{*}. |
| 6011 | |
| 6012 | The first argument is @code{(prefix-numeric-value arg)}. When |
| 6013 | @code{"P"} is used as the argument for @code{interactive}, the value |
| 6014 | passed to the function as its argument is passed a ``raw prefix |
| 6015 | argument'', and not a number. (It is a number in a list.) To perform |
| 6016 | the arithmetic, a conversion is necessary, and |
| 6017 | @code{prefix-numeric-value} does the job. |
| 6018 | |
| 6019 | @findex / @r{(division)} |
| 6020 | @cindex Division |
| 6021 | The second argument is @code{(/ (buffer-size) 10)}. This expression |
| 6022 | divides the numeric value of the buffer by ten. This produces a number |
| 6023 | that tells how many characters make up one tenth of the buffer size. |
| 6024 | (In Lisp, @code{/} is used for division, just as @code{*} is |
| 6025 | used for multiplication.) |
| 6026 | |
| 6027 | @need 1200 |
| 6028 | In the multiplication expression as a whole, this amount is multiplied |
| 6029 | by the value of the prefix argument---the multiplication looks like this: |
| 6030 | |
| 6031 | @smallexample |
| 6032 | @group |
| 6033 | (* @var{numeric-value-of-prefix-arg} |
| 6034 | @var{number-of-characters-in-one-tenth-of-the-buffer}) |
| 6035 | @end group |
| 6036 | @end smallexample |
| 6037 | |
| 6038 | @noindent |
| 6039 | If, for example, the prefix argument is @samp{7}, the one-tenth value |
| 6040 | will be multiplied by 7 to give a position 70% of the way through the |
| 6041 | buffer. |
| 6042 | |
| 6043 | @need 1200 |
| 6044 | The result of all this is that if the buffer is large, the |
| 6045 | @code{goto-char} expression reads like this: |
| 6046 | |
| 6047 | @smallexample |
| 6048 | @group |
| 6049 | (goto-char (* (prefix-numeric-value arg) |
| 6050 | (/ (buffer-size) 10))) |
| 6051 | @end group |
| 6052 | @end smallexample |
| 6053 | |
| 6054 | This puts the cursor where we want it. |
| 6055 | |
| 6056 | @node Small buffer case, , Large buffer case, beginning-of-buffer opt arg |
| 6057 | @comment node-name, next, previous, up |
| 6058 | @unnumberedsubsubsec What happens in a small buffer |
| 6059 | |
| 6060 | If the buffer contains fewer than 10,000 characters, a slightly |
| 6061 | different computation is performed. You might think this is not |
| 6062 | necessary, since the first computation could do the job. However, in |
| 6063 | a small buffer, the first method may not put the cursor on exactly the |
| 6064 | desired line; the second method does a better job. |
| 6065 | |
| 6066 | @need 800 |
| 6067 | The code looks like this: |
| 6068 | |
| 6069 | @c Keep this on one line. |
| 6070 | @smallexample |
| 6071 | (/ (+ 10 (* (buffer-size) (prefix-numeric-value arg))) 10)) |
| 6072 | @end smallexample |
| 6073 | |
| 6074 | @need 1200 |
| 6075 | @noindent |
| 6076 | This is code in which you figure out what happens by discovering how the |
| 6077 | functions are embedded in parentheses. It is easier to read if you |
| 6078 | reformat it with each expression indented more deeply than its |
| 6079 | enclosing expression: |
| 6080 | |
| 6081 | @smallexample |
| 6082 | @group |
| 6083 | (/ |
| 6084 | (+ 10 |
| 6085 | (* |
| 6086 | (buffer-size) |
| 6087 | (prefix-numeric-value arg))) |
| 6088 | 10)) |
| 6089 | @end group |
| 6090 | @end smallexample |
| 6091 | |
| 6092 | @need 1200 |
| 6093 | @noindent |
| 6094 | Looking at parentheses, we see that the innermost operation is |
| 6095 | @code{(prefix-numeric-value arg)}, which converts the raw argument to a |
| 6096 | number. This number is multiplied by the buffer size in the following |
| 6097 | expression: |
| 6098 | |
| 6099 | @smallexample |
| 6100 | (* (buffer-size) (prefix-numeric-value arg)) |
| 6101 | @end smallexample |
| 6102 | |
| 6103 | @noindent |
| 6104 | This multiplication creates a number that may be larger than the size of |
| 6105 | the buffer---seven times larger if the argument is 7, for example. Ten |
| 6106 | is then added to this number and finally the large number is divided by |
| 6107 | ten to provide a value that is one character larger than the percentage |
| 6108 | position in the buffer. |
| 6109 | |
| 6110 | The number that results from all this is passed to @code{goto-char} and |
| 6111 | the cursor is moved to that point. |
| 6112 | |
| 6113 | @need 1500 |
| 6114 | @node beginning-of-buffer complete, , beginning-of-buffer opt arg, beginning-of-buffer |
| 6115 | @comment node-name, next, previous, up |
| 6116 | @subsection The Complete @code{beginning-of-buffer} |
| 6117 | |
| 6118 | @need 1000 |
| 6119 | Here is the complete text of the @code{beginning-of-buffer} function: |
| 6120 | @sp 1 |
| 6121 | |
| 6122 | @smallexample |
| 6123 | @group |
| 6124 | (defun beginning-of-buffer (&optional arg) |
| 6125 | "Move point to the beginning of the buffer; |
| 6126 | leave mark at previous position. |
| 6127 | With arg N, put point N/10 of the way |
| 6128 | from the true beginning. |
| 6129 | @end group |
| 6130 | @group |
| 6131 | Don't use this in Lisp programs! |
| 6132 | \(goto-char (point-min)) is faster |
| 6133 | and does not set the mark." |
| 6134 | (interactive "P") |
| 6135 | (push-mark) |
| 6136 | @end group |
| 6137 | @group |
| 6138 | (goto-char |
| 6139 | (if arg |
| 6140 | (if (> (buffer-size) 10000) |
| 6141 | ;; @r{Avoid overflow for large buffer sizes!} |
| 6142 | (* (prefix-numeric-value arg) |
| 6143 | (/ (buffer-size) 10)) |
| 6144 | @end group |
| 6145 | @group |
| 6146 | (/ (+ 10 (* (buffer-size) |
| 6147 | (prefix-numeric-value arg))) |
| 6148 | 10)) |
| 6149 | (point-min))) |
| 6150 | (if arg (forward-line 1))) |
| 6151 | @end group |
| 6152 | @end smallexample |
| 6153 | |
| 6154 | @noindent |
| 6155 | Except for two small points, the previous discussion shows how this |
| 6156 | function works. The first point deals with a detail in the |
| 6157 | documentation string, and the second point concerns the last line of |
| 6158 | the function. |
| 6159 | |
| 6160 | @need 800 |
| 6161 | In the documentation string, there is reference to an expression: |
| 6162 | |
| 6163 | @smallexample |
| 6164 | \(goto-char (point-min)) |
| 6165 | @end smallexample |
| 6166 | |
| 6167 | @noindent |
| 6168 | A @samp{\} is used before the first parenthesis of this expression. |
| 6169 | This @samp{\} tells the Lisp interpreter that the expression should be |
| 6170 | printed as shown in the documentation rather than evaluated as a |
| 6171 | symbolic expression, which is what it looks like. |
| 6172 | |
| 6173 | @need 1200 |
| 6174 | Finally, the last line of the @code{beginning-of-buffer} command says to |
| 6175 | move point to the beginning of the next line if the command is |
| 6176 | invoked with an argument: |
| 6177 | |
| 6178 | @smallexample |
| 6179 | (if arg (forward-line 1))) |
| 6180 | @end smallexample |
| 6181 | |
| 6182 | @noindent |
| 6183 | This puts the cursor at the beginning of the first line after the |
| 6184 | appropriate tenths position in the buffer. This is a flourish that |
| 6185 | means that the cursor is always located @emph{at least} the requested |
| 6186 | tenths of the way through the buffer, which is a nicety that is, |
| 6187 | perhaps, not necessary, but which, if it did not occur, would be sure to |
| 6188 | draw complaints. |
| 6189 | |
| 6190 | @node Second Buffer Related Review, optional Exercise, beginning-of-buffer, More Complex |
| 6191 | @comment node-name, next, previous, up |
| 6192 | @section Review |
| 6193 | |
| 6194 | Here is a brief summary of some of the topics covered in this chapter. |
| 6195 | |
| 6196 | @table @code |
| 6197 | @item or |
| 6198 | Evaluate each argument in sequence, and return the value of the first |
| 6199 | argument that is not @code{nil}; if none return a value that is not |
| 6200 | @code{nil}, return @code{nil}. In brief, return the first true value |
| 6201 | of the arguments; return a true value if one @emph{or} any of the |
| 6202 | other are true. |
| 6203 | |
| 6204 | @item and |
| 6205 | Evaluate each argument in sequence, and if any are @code{nil}, return |
| 6206 | @code{nil}; if none are @code{nil}, return the value of the last |
| 6207 | argument. In brief, return a true value only if all the arguments are |
| 6208 | true; return a true value if one @emph{and} each of the others is |
| 6209 | true. |
| 6210 | |
| 6211 | @item &optional |
| 6212 | A keyword used to indicate that an argument to a function definition |
| 6213 | is optional; this means that the function can be evaluated without the |
| 6214 | argument, if desired. |
| 6215 | |
| 6216 | @item prefix-numeric-value |
| 6217 | Convert the `raw prefix argument' produced by @code{(interactive |
| 6218 | "P")} to a numeric value. |
| 6219 | |
| 6220 | @item forward-line |
| 6221 | Move point forward to the beginning of the next line, or if the argument |
| 6222 | is greater than one, forward that many lines. If it can't move as far |
| 6223 | forward as it is supposed to, @code{forward-line} goes forward as far as |
| 6224 | it can and then returns a count of the number of additional lines it was |
| 6225 | supposed to move but couldn't. |
| 6226 | |
| 6227 | @item erase-buffer |
| 6228 | Delete the entire contents of the current buffer. |
| 6229 | |
| 6230 | @item bufferp |
| 6231 | Return @code{t} if its argument is a buffer; otherwise return @code{nil}. |
| 6232 | @end table |
| 6233 | |
| 6234 | @node optional Exercise, , Second Buffer Related Review, More Complex |
| 6235 | @section @code{optional} Argument Exercise |
| 6236 | |
| 6237 | Write an interactive function with an optional argument that tests |
| 6238 | whether its argument, a number, is greater than or equal to, or else, |
| 6239 | less than the value of @code{fill-column}, and tells you which, in a |
| 6240 | message. However, if you do not pass an argument to the function, use |
| 6241 | 56 as a default value. |
| 6242 | |
| 6243 | @node Narrowing & Widening, car cdr & cons, More Complex, Top |
| 6244 | @comment node-name, next, previous, up |
| 6245 | @chapter Narrowing and Widening |
| 6246 | @cindex Focusing attention (narrowing) |
| 6247 | @cindex Narrowing |
| 6248 | @cindex Widening |
| 6249 | |
| 6250 | Narrowing is a feature of Emacs that makes it possible for you to focus |
| 6251 | on a specific part of a buffer, and work without accidentally changing |
| 6252 | other parts. Narrowing is normally disabled since it can confuse |
| 6253 | novices. |
| 6254 | |
| 6255 | @menu |
| 6256 | * Narrowing advantages:: The advantages of narrowing |
| 6257 | * save-restriction:: The @code{save-restriction} special form. |
| 6258 | * what-line:: The number of the line that point is on. |
| 6259 | * narrow Exercise:: |
| 6260 | @end menu |
| 6261 | |
| 6262 | @node Narrowing advantages, save-restriction, Narrowing & Widening, Narrowing & Widening |
| 6263 | @ifnottex |
| 6264 | @unnumberedsec The Advantages of Narrowing |
| 6265 | @end ifnottex |
| 6266 | |
| 6267 | With narrowing, the rest of a buffer is made invisible, as if it weren't |
| 6268 | there. This is an advantage if, for example, you want to replace a word |
| 6269 | in one part of a buffer but not in another: you narrow to the part you want |
| 6270 | and the replacement is carried out only in that section, not in the rest |
| 6271 | of the buffer. Searches will only work within a narrowed region, not |
| 6272 | outside of one, so if you are fixing a part of a document, you can keep |
| 6273 | yourself from accidentally finding parts you do not need to fix by |
| 6274 | narrowing just to the region you want. |
| 6275 | (The key binding for @code{narrow-to-region} is @kbd{C-x n n}.) |
| 6276 | |
| 6277 | However, narrowing does make the rest of the buffer invisible, which |
| 6278 | can scare people who inadvertently invoke narrowing and think they |
| 6279 | have deleted a part of their file. Moreover, the @code{undo} command |
| 6280 | (which is usually bound to @kbd{C-x u}) does not turn off narrowing |
| 6281 | (nor should it), so people can become quite desperate if they do not |
| 6282 | know that they can return the rest of a buffer to visibility with the |
| 6283 | @code{widen} command. |
| 6284 | (The key binding for @code{widen} is @kbd{C-x n w}.) |
| 6285 | |
| 6286 | Narrowing is just as useful to the Lisp interpreter as to a human. |
| 6287 | Often, an Emacs Lisp function is designed to work on just part of a |
| 6288 | buffer; or conversely, an Emacs Lisp function needs to work on all of a |
| 6289 | buffer that has been narrowed. The @code{what-line} function, for |
| 6290 | example, removes the narrowing from a buffer, if it has any narrowing |
| 6291 | and when it has finished its job, restores the narrowing to what it was. |
| 6292 | On the other hand, the @code{count-lines} function, which is called by |
| 6293 | @code{what-line}, uses narrowing to restrict itself to just that portion |
| 6294 | of the buffer in which it is interested and then restores the previous |
| 6295 | situation. |
| 6296 | |
| 6297 | @node save-restriction, what-line, Narrowing advantages, Narrowing & Widening |
| 6298 | @comment node-name, next, previous, up |
| 6299 | @section The @code{save-restriction} Special Form |
| 6300 | @findex save-restriction |
| 6301 | |
| 6302 | In Emacs Lisp, you can use the @code{save-restriction} special form to |
| 6303 | keep track of whatever narrowing is in effect, if any. When the Lisp |
| 6304 | interpreter meets with @code{save-restriction}, it executes the code |
| 6305 | in the body of the @code{save-restriction} expression, and then undoes |
| 6306 | any changes to narrowing that the code caused. If, for example, the |
| 6307 | buffer is narrowed and the code that follows @code{save-restriction} |
| 6308 | gets rid of the narrowing, @code{save-restriction} returns the buffer |
| 6309 | to its narrowed region afterwards. In the @code{what-line} command, |
| 6310 | any narrowing the buffer may have is undone by the @code{widen} |
| 6311 | command that immediately follows the @code{save-restriction} command. |
| 6312 | Any original narrowing is restored just before the completion of the |
| 6313 | function. |
| 6314 | |
| 6315 | @need 1250 |
| 6316 | The template for a @code{save-restriction} expression is simple: |
| 6317 | |
| 6318 | @smallexample |
| 6319 | @group |
| 6320 | (save-restriction |
| 6321 | @var{body}@dots{} ) |
| 6322 | @end group |
| 6323 | @end smallexample |
| 6324 | |
| 6325 | @noindent |
| 6326 | The body of the @code{save-restriction} is one or more expressions that |
| 6327 | will be evaluated in sequence by the Lisp interpreter. |
| 6328 | |
| 6329 | Finally, a point to note: when you use both @code{save-excursion} and |
| 6330 | @code{save-restriction}, one right after the other, you should use |
| 6331 | @code{save-excursion} outermost. If you write them in reverse order, |
| 6332 | you may fail to record narrowing in the buffer to which Emacs switches |
| 6333 | after calling @code{save-excursion}. Thus, when written together, |
| 6334 | @code{save-excursion} and @code{save-restriction} should be written |
| 6335 | like this: |
| 6336 | |
| 6337 | @smallexample |
| 6338 | @group |
| 6339 | (save-excursion |
| 6340 | (save-restriction |
| 6341 | @var{body}@dots{})) |
| 6342 | @end group |
| 6343 | @end smallexample |
| 6344 | |
| 6345 | In other circumstances, when not written together, the |
| 6346 | @code{save-excursion} and @code{save-restriction} special forms must |
| 6347 | be written in the order appropriate to the function. |
| 6348 | |
| 6349 | @need 1250 |
| 6350 | For example, |
| 6351 | |
| 6352 | @smallexample |
| 6353 | @group |
| 6354 | (save-restriction |
| 6355 | (widen) |
| 6356 | (save-excursion |
| 6357 | @var{body}@dots{})) |
| 6358 | @end group |
| 6359 | @end smallexample |
| 6360 | |
| 6361 | @node what-line, narrow Exercise, save-restriction, Narrowing & Widening |
| 6362 | @comment node-name, next, previous, up |
| 6363 | @section @code{what-line} |
| 6364 | @findex what-line |
| 6365 | @cindex Widening, example of |
| 6366 | |
| 6367 | The @code{what-line} command tells you the number of the line in which |
| 6368 | the cursor is located. The function illustrates the use of the |
| 6369 | @code{save-restriction} and @code{save-excursion} commands. Here is the |
| 6370 | original text of the function: |
| 6371 | |
| 6372 | @smallexample |
| 6373 | @group |
| 6374 | (defun what-line () |
| 6375 | "Print the current line number (in the buffer) of point." |
| 6376 | (interactive) |
| 6377 | (save-restriction |
| 6378 | (widen) |
| 6379 | (save-excursion |
| 6380 | (beginning-of-line) |
| 6381 | (message "Line %d" |
| 6382 | (1+ (count-lines 1 (point))))))) |
| 6383 | @end group |
| 6384 | @end smallexample |
| 6385 | |
| 6386 | (In recent versions of GNU Emacs, the @code{what-line} function has |
| 6387 | been expanded to tell you your line number in a narrowed buffer as |
| 6388 | well as your line number in a widened buffer. The recent version is |
| 6389 | more complex than the version shown here. If you feel adventurous, |
| 6390 | you might want to look at it after figuring out how this version |
| 6391 | works. The newer version uses a conditional to determine whether the |
| 6392 | buffer has been narrowed, and rather than use @code{beginning-of-line} |
| 6393 | to move point to the beginning of the current line, if need be, the |
| 6394 | function uses @code{(forward-line 0)}.) |
| 6395 | |
| 6396 | The function as shown here has a documentation line and is |
| 6397 | interactive, as you would expect. The next two lines use the |
| 6398 | functions @code{save-restriction} and @code{widen}. |
| 6399 | |
| 6400 | The @code{save-restriction} special form notes whatever narrowing is in |
| 6401 | effect, if any, in the current buffer and restores that narrowing after |
| 6402 | the code in the body of the @code{save-restriction} has been evaluated. |
| 6403 | |
| 6404 | The @code{save-restriction} special form is followed by @code{widen}. |
| 6405 | This function undoes any narrowing the current buffer may have had |
| 6406 | when @code{what-line} was called. (The narrowing that was there is |
| 6407 | the narrowing that @code{save-restriction} remembers.) This widening |
| 6408 | makes it possible for the line counting commands to count from the |
| 6409 | beginning of the buffer. Otherwise, they would have been limited to |
| 6410 | counting within the accessible region. Any original narrowing is |
| 6411 | restored just before the completion of the function by the |
| 6412 | @code{save-restriction} special form. |
| 6413 | |
| 6414 | The call to @code{widen} is followed by @code{save-excursion}, which |
| 6415 | saves the location of the cursor (i.e., of point) and of the mark, and |
| 6416 | restores them after the code in the body of the @code{save-excursion} |
| 6417 | uses the @code{beginning-of-line} function to move point. |
| 6418 | |
| 6419 | (Note that the @code{(widen)} expression comes between the |
| 6420 | @code{save-restriction} and @code{save-excursion} special forms. When |
| 6421 | you write the two @code{save- @dots{}} expressions in sequence, write |
| 6422 | @code{save-excursion} outermost.) |
| 6423 | |
| 6424 | @need 1200 |
| 6425 | The last two lines of the @code{what-line} function are functions to |
| 6426 | count the number of lines in the buffer and then print the number in the |
| 6427 | echo area. |
| 6428 | |
| 6429 | @smallexample |
| 6430 | @group |
| 6431 | (message "Line %d" |
| 6432 | (1+ (count-lines 1 (point))))))) |
| 6433 | @end group |
| 6434 | @end smallexample |
| 6435 | |
| 6436 | The @code{message} function prints a one-line message at the bottom of the |
| 6437 | Emacs screen. The first argument is inside of quotation marks and is |
| 6438 | printed as a string of characters. However, it may contain @samp{%d}, |
| 6439 | @samp{%s}, or @samp{%c} to print arguments that follow the string. |
| 6440 | @samp{%d} prints the argument as a decimal, so the message will say |
| 6441 | something such as @samp{Line 243}. |
| 6442 | |
| 6443 | @need 1200 |
| 6444 | |
| 6445 | The number that is printed in place of the @samp{%d} is computed by the |
| 6446 | last line of the function: |
| 6447 | |
| 6448 | @smallexample |
| 6449 | (1+ (count-lines 1 (point))) |
| 6450 | @end smallexample |
| 6451 | |
| 6452 | @noindent |
| 6453 | What this does is count the lines from the first position of the |
| 6454 | buffer, indicated by the @code{1}, up to @code{(point)}, and then add |
| 6455 | one to that number. (The @code{1+} function adds one to its |
| 6456 | argument.) We add one to it because line 2 has only one line before |
| 6457 | it, and @code{count-lines} counts only the lines @emph{before} the |
| 6458 | current line. |
| 6459 | |
| 6460 | After @code{count-lines} has done its job, and the message has been |
| 6461 | printed in the echo area, the @code{save-excursion} restores point and |
| 6462 | mark to their original positions; and @code{save-restriction} restores |
| 6463 | the original narrowing, if any. |
| 6464 | |
| 6465 | @node narrow Exercise, , what-line, Narrowing & Widening |
| 6466 | @section Exercise with Narrowing |
| 6467 | |
| 6468 | Write a function that will display the first 60 characters of the |
| 6469 | current buffer, even if you have narrowed the buffer to its latter |
| 6470 | half so that the first line is inaccessible. Restore point, mark, and |
| 6471 | narrowing. For this exercise, you need to use a whole potpourri of |
| 6472 | functions, including @code{save-restriction}, @code{widen}, |
| 6473 | @code{goto-char}, @code{point-min}, @code{message}, and |
| 6474 | @code{buffer-substring}. |
| 6475 | |
| 6476 | @cindex Properties, mention of @code{buffer-substring-no-properties} |
| 6477 | (@code{buffer-substring} is a previously unmentioned function you will |
| 6478 | have to investigate yourself; or perhaps you will have to use |
| 6479 | @code{buffer-substring-no-properties} @dots{}, yet another function |
| 6480 | and one that introduces text properties, a feature otherwise not |
| 6481 | discussed here. @xref{Text Properties, , Text Properties, elisp, The |
| 6482 | GNU Emacs Lisp Reference Manual}. Additionally, do you really need |
| 6483 | @code{goto-char} or @code{point-min}? Or can you write the function |
| 6484 | without them?) |
| 6485 | |
| 6486 | @node car cdr & cons, Cutting & Storing Text, Narrowing & Widening, Top |
| 6487 | @comment node-name, next, previous, up |
| 6488 | @chapter @code{car}, @code{cdr}, @code{cons}: Fundamental Functions |
| 6489 | @findex car, @r{introduced} |
| 6490 | @findex cdr, @r{introduced} |
| 6491 | |
| 6492 | In Lisp, @code{car}, @code{cdr}, and @code{cons} are fundamental |
| 6493 | functions. The @code{cons} function is used to construct lists, and |
| 6494 | the @code{car} and @code{cdr} functions are used to take them apart. |
| 6495 | |
| 6496 | In the walk through of the @code{copy-region-as-kill} function, we |
| 6497 | will see @code{cons} as well as two variants on @code{cdr}, |
| 6498 | namely, @code{setcdr} and @code{nthcdr}. (@xref{copy-region-as-kill}.) |
| 6499 | |
| 6500 | @menu |
| 6501 | * Strange Names:: An historical aside: why the strange names? |
| 6502 | * car & cdr:: Functions for extracting part of a list. |
| 6503 | * cons:: Constructing a list. |
| 6504 | * nthcdr:: Calling @code{cdr} repeatedly. |
| 6505 | * nth:: |
| 6506 | * setcar:: Changing the first element of a list. |
| 6507 | * setcdr:: Changing the rest of a list. |
| 6508 | * cons Exercise:: |
| 6509 | @end menu |
| 6510 | |
| 6511 | @node Strange Names, car & cdr, car cdr & cons, car cdr & cons |
| 6512 | @ifnottex |
| 6513 | @unnumberedsec Strange Names |
| 6514 | @end ifnottex |
| 6515 | |
| 6516 | The name of the @code{cons} function is not unreasonable: it is an |
| 6517 | abbreviation of the word `construct'. The origins of the names for |
| 6518 | @code{car} and @code{cdr}, on the other hand, are esoteric: @code{car} |
| 6519 | is an acronym from the phrase `Contents of the Address part of the |
| 6520 | Register'; and @code{cdr} (pronounced `could-er') is an acronym from |
| 6521 | the phrase `Contents of the Decrement part of the Register'. These |
| 6522 | phrases refer to specific pieces of hardware on the very early |
| 6523 | computer on which the original Lisp was developed. Besides being |
| 6524 | obsolete, the phrases have been completely irrelevant for more than 25 |
| 6525 | years to anyone thinking about Lisp. Nonetheless, although a few |
| 6526 | brave scholars have begun to use more reasonable names for these |
| 6527 | functions, the old terms are still in use. In particular, since the |
| 6528 | terms are used in the Emacs Lisp source code, we will use them in this |
| 6529 | introduction. |
| 6530 | |
| 6531 | @node car & cdr, cons, Strange Names, car cdr & cons |
| 6532 | @comment node-name, next, previous, up |
| 6533 | @section @code{car} and @code{cdr} |
| 6534 | |
| 6535 | The @sc{car} of a list is, quite simply, the first item in the list. |
| 6536 | Thus the @sc{car} of the list @code{(rose violet daisy buttercup)} is |
| 6537 | @code{rose}. |
| 6538 | |
| 6539 | @need 1200 |
| 6540 | If you are reading this in Info in GNU Emacs, you can see this by |
| 6541 | evaluating the following: |
| 6542 | |
| 6543 | @smallexample |
| 6544 | (car '(rose violet daisy buttercup)) |
| 6545 | @end smallexample |
| 6546 | |
| 6547 | @noindent |
| 6548 | After evaluating the expression, @code{rose} will appear in the echo |
| 6549 | area. |
| 6550 | |
| 6551 | Clearly, a more reasonable name for the @code{car} function would be |
| 6552 | @code{first} and this is often suggested. |
| 6553 | |
| 6554 | @code{car} does not remove the first item from the list; it only reports |
| 6555 | what it is. After @code{car} has been applied to a list, the list is |
| 6556 | still the same as it was. In the jargon, @code{car} is |
| 6557 | `non-destructive'. This feature turns out to be important. |
| 6558 | |
| 6559 | The @sc{cdr} of a list is the rest of the list, that is, the |
| 6560 | @code{cdr} function returns the part of the list that follows the |
| 6561 | first item. Thus, while the @sc{car} of the list @code{'(rose violet |
| 6562 | daisy buttercup)} is @code{rose}, the rest of the list, the value |
| 6563 | returned by the @code{cdr} function, is @code{(violet daisy |
| 6564 | buttercup)}. |
| 6565 | |
| 6566 | @need 800 |
| 6567 | You can see this by evaluating the following in the usual way: |
| 6568 | |
| 6569 | @smallexample |
| 6570 | (cdr '(rose violet daisy buttercup)) |
| 6571 | @end smallexample |
| 6572 | |
| 6573 | @noindent |
| 6574 | When you evaluate this, @code{(violet daisy buttercup)} will appear in |
| 6575 | the echo area. |
| 6576 | |
| 6577 | Like @code{car}, @code{cdr} does not remove any elements from the |
| 6578 | list---it just returns a report of what the second and subsequent |
| 6579 | elements are. |
| 6580 | |
| 6581 | Incidentally, in the example, the list of flowers is quoted. If it were |
| 6582 | not, the Lisp interpreter would try to evaluate the list by calling |
| 6583 | @code{rose} as a function. In this example, we do not want to do that. |
| 6584 | |
| 6585 | Clearly, a more reasonable name for @code{cdr} would be @code{rest}. |
| 6586 | |
| 6587 | (There is a lesson here: when you name new functions, consider very |
| 6588 | carefully what you are doing, since you may be stuck with the names |
| 6589 | for far longer than you expect. The reason this document perpetuates |
| 6590 | these names is that the Emacs Lisp source code uses them, and if I did |
| 6591 | not use them, you would have a hard time reading the code; but do, |
| 6592 | please, try to avoid using these terms yourself. The people who come |
| 6593 | after you will be grateful to you.) |
| 6594 | |
| 6595 | When @code{car} and @code{cdr} are applied to a list made up of symbols, |
| 6596 | such as the list @code{(pine fir oak maple)}, the element of the list |
| 6597 | returned by the function @code{car} is the symbol @code{pine} without |
| 6598 | any parentheses around it. @code{pine} is the first element in the |
| 6599 | list. However, the @sc{cdr} of the list is a list itself, @code{(fir |
| 6600 | oak maple)}, as you can see by evaluating the following expressions in |
| 6601 | the usual way: |
| 6602 | |
| 6603 | @smallexample |
| 6604 | @group |
| 6605 | (car '(pine fir oak maple)) |
| 6606 | |
| 6607 | (cdr '(pine fir oak maple)) |
| 6608 | @end group |
| 6609 | @end smallexample |
| 6610 | |
| 6611 | On the other hand, in a list of lists, the first element is itself a |
| 6612 | list. @code{car} returns this first element as a list. For example, |
| 6613 | the following list contains three sub-lists, a list of carnivores, a |
| 6614 | list of herbivores and a list of sea mammals: |
| 6615 | |
| 6616 | @smallexample |
| 6617 | @group |
| 6618 | (car '((lion tiger cheetah) |
| 6619 | (gazelle antelope zebra) |
| 6620 | (whale dolphin seal))) |
| 6621 | @end group |
| 6622 | @end smallexample |
| 6623 | |
| 6624 | @noindent |
| 6625 | In this example, the first element or @sc{car} of the list is the list of |
| 6626 | carnivores, @code{(lion tiger cheetah)}, and the rest of the list is |
| 6627 | @code{((gazelle antelope zebra) (whale dolphin seal))}. |
| 6628 | |
| 6629 | @smallexample |
| 6630 | @group |
| 6631 | (cdr '((lion tiger cheetah) |
| 6632 | (gazelle antelope zebra) |
| 6633 | (whale dolphin seal))) |
| 6634 | @end group |
| 6635 | @end smallexample |
| 6636 | |
| 6637 | It is worth saying again that @code{car} and @code{cdr} are |
| 6638 | non-destructive---that is, they do not modify or change lists to which |
| 6639 | they are applied. This is very important for how they are used. |
| 6640 | |
| 6641 | Also, in the first chapter, in the discussion about atoms, I said that |
| 6642 | in Lisp, ``certain kinds of atom, such as an array, can be separated |
| 6643 | into parts; but the mechanism for doing this is different from the |
| 6644 | mechanism for splitting a list. As far as Lisp is concerned, the |
| 6645 | atoms of a list are unsplittable.'' (@xref{Lisp Atoms}.) The |
| 6646 | @code{car} and @code{cdr} functions are used for splitting lists and |
| 6647 | are considered fundamental to Lisp. Since they cannot split or gain |
| 6648 | access to the parts of an array, an array is considered an atom. |
| 6649 | Conversely, the other fundamental function, @code{cons}, can put |
| 6650 | together or construct a list, but not an array. (Arrays are handled |
| 6651 | by array-specific functions. @xref{Arrays, , Arrays, elisp, The GNU |
| 6652 | Emacs Lisp Reference Manual}.) |
| 6653 | |
| 6654 | @node cons, nthcdr, car & cdr, car cdr & cons |
| 6655 | @comment node-name, next, previous, up |
| 6656 | @section @code{cons} |
| 6657 | @findex cons, @r{introduced} |
| 6658 | |
| 6659 | The @code{cons} function constructs lists; it is the inverse of |
| 6660 | @code{car} and @code{cdr}. For example, @code{cons} can be used to make |
| 6661 | a four element list from the three element list, @code{(fir oak maple)}: |
| 6662 | |
| 6663 | @smallexample |
| 6664 | (cons 'pine '(fir oak maple)) |
| 6665 | @end smallexample |
| 6666 | |
| 6667 | @need 800 |
| 6668 | @noindent |
| 6669 | After evaluating this list, you will see |
| 6670 | |
| 6671 | @smallexample |
| 6672 | (pine fir oak maple) |
| 6673 | @end smallexample |
| 6674 | |
| 6675 | @noindent |
| 6676 | appear in the echo area. @code{cons} causes the creation of a new |
| 6677 | list in which the element is followed by the elements of the original |
| 6678 | list. |
| 6679 | |
| 6680 | We often say that `@code{cons} puts a new element at the beginning of |
| 6681 | a list; it attaches or pushes elements onto the list', but this |
| 6682 | phrasing can be misleading, since @code{cons} does not change an |
| 6683 | existing list, but creates a new one. |
| 6684 | |
| 6685 | Like @code{car} and @code{cdr}, @code{cons} is non-destructive. |
| 6686 | |
| 6687 | @menu |
| 6688 | * Build a list:: |
| 6689 | * length:: How to find the length of a list. |
| 6690 | @end menu |
| 6691 | |
| 6692 | @node Build a list, length, cons, cons |
| 6693 | @ifnottex |
| 6694 | @unnumberedsubsec Build a list |
| 6695 | @end ifnottex |
| 6696 | |
| 6697 | @code{cons} must have a list to attach to.@footnote{Actually, you can |
| 6698 | @code{cons} an element to an atom to produce a dotted pair. Dotted |
| 6699 | pairs are not discussed here; see @ref{Dotted Pair Notation, , Dotted |
| 6700 | Pair Notation, elisp, The GNU Emacs Lisp Reference Manual}.} You |
| 6701 | cannot start from absolutely nothing. If you are building a list, you |
| 6702 | need to provide at least an empty list at the beginning. Here is a |
| 6703 | series of @code{cons} expressions that build up a list of flowers. If |
| 6704 | you are reading this in Info in GNU Emacs, you can evaluate each of |
| 6705 | the expressions in the usual way; the value is printed in this text |
| 6706 | after @samp{@result{}}, which you may read as `evaluates to'. |
| 6707 | |
| 6708 | @smallexample |
| 6709 | @group |
| 6710 | (cons 'buttercup ()) |
| 6711 | @result{} (buttercup) |
| 6712 | @end group |
| 6713 | |
| 6714 | @group |
| 6715 | (cons 'daisy '(buttercup)) |
| 6716 | @result{} (daisy buttercup) |
| 6717 | @end group |
| 6718 | |
| 6719 | @group |
| 6720 | (cons 'violet '(daisy buttercup)) |
| 6721 | @result{} (violet daisy buttercup) |
| 6722 | @end group |
| 6723 | |
| 6724 | @group |
| 6725 | (cons 'rose '(violet daisy buttercup)) |
| 6726 | @result{} (rose violet daisy buttercup) |
| 6727 | @end group |
| 6728 | @end smallexample |
| 6729 | |
| 6730 | @noindent |
| 6731 | In the first example, the empty list is shown as @code{()} and a list |
| 6732 | made up of @code{buttercup} followed by the empty list is constructed. |
| 6733 | As you can see, the empty list is not shown in the list that was |
| 6734 | constructed. All that you see is @code{(buttercup)}. The empty list is |
| 6735 | not counted as an element of a list because there is nothing in an empty |
| 6736 | list. Generally speaking, an empty list is invisible. |
| 6737 | |
| 6738 | The second example, @code{(cons 'daisy '(buttercup))} constructs a new, |
| 6739 | two element list by putting @code{daisy} in front of @code{buttercup}; |
| 6740 | and the third example constructs a three element list by putting |
| 6741 | @code{violet} in front of @code{daisy} and @code{buttercup}. |
| 6742 | |
| 6743 | @node length, , Build a list, cons |
| 6744 | @comment node-name, next, previous, up |
| 6745 | @subsection Find the Length of a List: @code{length} |
| 6746 | @findex length |
| 6747 | |
| 6748 | You can find out how many elements there are in a list by using the Lisp |
| 6749 | function @code{length}, as in the following examples: |
| 6750 | |
| 6751 | @smallexample |
| 6752 | @group |
| 6753 | (length '(buttercup)) |
| 6754 | @result{} 1 |
| 6755 | @end group |
| 6756 | |
| 6757 | @group |
| 6758 | (length '(daisy buttercup)) |
| 6759 | @result{} 2 |
| 6760 | @end group |
| 6761 | |
| 6762 | @group |
| 6763 | (length (cons 'violet '(daisy buttercup))) |
| 6764 | @result{} 3 |
| 6765 | @end group |
| 6766 | @end smallexample |
| 6767 | |
| 6768 | @noindent |
| 6769 | In the third example, the @code{cons} function is used to construct a |
| 6770 | three element list which is then passed to the @code{length} function as |
| 6771 | its argument. |
| 6772 | |
| 6773 | @need 1200 |
| 6774 | We can also use @code{length} to count the number of elements in an |
| 6775 | empty list: |
| 6776 | |
| 6777 | @smallexample |
| 6778 | @group |
| 6779 | (length ()) |
| 6780 | @result{} 0 |
| 6781 | @end group |
| 6782 | @end smallexample |
| 6783 | |
| 6784 | @noindent |
| 6785 | As you would expect, the number of elements in an empty list is zero. |
| 6786 | |
| 6787 | An interesting experiment is to find out what happens if you try to find |
| 6788 | the length of no list at all; that is, if you try to call @code{length} |
| 6789 | without giving it an argument, not even an empty list: |
| 6790 | |
| 6791 | @smallexample |
| 6792 | (length ) |
| 6793 | @end smallexample |
| 6794 | |
| 6795 | @need 800 |
| 6796 | @noindent |
| 6797 | What you see, if you evaluate this, is the error message |
| 6798 | |
| 6799 | @smallexample |
| 6800 | Wrong number of arguments: #<subr length>, 0 |
| 6801 | @end smallexample |
| 6802 | |
| 6803 | @noindent |
| 6804 | This means that the function receives the wrong number of |
| 6805 | arguments, zero, when it expects some other number of arguments. In |
| 6806 | this case, one argument is expected, the argument being a list whose |
| 6807 | length the function is measuring. (Note that @emph{one} list is |
| 6808 | @emph{one} argument, even if the list has many elements inside it.) |
| 6809 | |
| 6810 | The part of the error message that says @samp{#<subr length>} is the |
| 6811 | name of the function. This is written with a special notation, |
| 6812 | @samp{#<subr}, that indicates that the function @code{length} is one |
| 6813 | of the primitive functions written in C rather than in Emacs Lisp. |
| 6814 | (@samp{subr} is an abbreviation for `subroutine'.) @xref{What Is a |
| 6815 | Function, , What Is a Function?, elisp , The GNU Emacs Lisp Reference |
| 6816 | Manual}, for more about subroutines. |
| 6817 | |
| 6818 | @node nthcdr, nth, cons, car cdr & cons |
| 6819 | @comment node-name, next, previous, up |
| 6820 | @section @code{nthcdr} |
| 6821 | @findex nthcdr |
| 6822 | |
| 6823 | The @code{nthcdr} function is associated with the @code{cdr} function. |
| 6824 | What it does is take the @sc{cdr} of a list repeatedly. |
| 6825 | |
| 6826 | If you take the @sc{cdr} of the list @code{(pine fir |
| 6827 | oak maple)}, you will be returned the list @code{(fir oak maple)}. If you |
| 6828 | repeat this on what was returned, you will be returned the list |
| 6829 | @code{(oak maple)}. (Of course, repeated @sc{cdr}ing on the original |
| 6830 | list will just give you the original @sc{cdr} since the function does |
| 6831 | not change the list. You need to evaluate the @sc{cdr} of the |
| 6832 | @sc{cdr} and so on.) If you continue this, eventually you will be |
| 6833 | returned an empty list, which in this case, instead of being shown as |
| 6834 | @code{()} is shown as @code{nil}. |
| 6835 | |
| 6836 | @need 1200 |
| 6837 | For review, here is a series of repeated @sc{cdr}s, the text following |
| 6838 | the @samp{@result{}} shows what is returned. |
| 6839 | |
| 6840 | @smallexample |
| 6841 | @group |
| 6842 | (cdr '(pine fir oak maple)) |
| 6843 | @result{}(fir oak maple) |
| 6844 | @end group |
| 6845 | |
| 6846 | @group |
| 6847 | (cdr '(fir oak maple)) |
| 6848 | @result{} (oak maple) |
| 6849 | @end group |
| 6850 | |
| 6851 | @group |
| 6852 | (cdr '(oak maple)) |
| 6853 | @result{}(maple) |
| 6854 | @end group |
| 6855 | |
| 6856 | @group |
| 6857 | (cdr '(maple)) |
| 6858 | @result{} nil |
| 6859 | @end group |
| 6860 | |
| 6861 | @group |
| 6862 | (cdr 'nil) |
| 6863 | @result{} nil |
| 6864 | @end group |
| 6865 | |
| 6866 | @group |
| 6867 | (cdr ()) |
| 6868 | @result{} nil |
| 6869 | @end group |
| 6870 | @end smallexample |
| 6871 | |
| 6872 | @need 1200 |
| 6873 | You can also do several @sc{cdr}s without printing the values in |
| 6874 | between, like this: |
| 6875 | |
| 6876 | @smallexample |
| 6877 | @group |
| 6878 | (cdr (cdr '(pine fir oak maple))) |
| 6879 | @result{} (oak maple) |
| 6880 | @end group |
| 6881 | @end smallexample |
| 6882 | |
| 6883 | @noindent |
| 6884 | In this example, the Lisp interpreter evaluates the innermost list first. |
| 6885 | The innermost list is quoted, so it just passes the list as it is to the |
| 6886 | innermost @code{cdr}. This @code{cdr} passes a list made up of the |
| 6887 | second and subsequent elements of the list to the outermost @code{cdr}, |
| 6888 | which produces a list composed of the third and subsequent elements of |
| 6889 | the original list. In this example, the @code{cdr} function is repeated |
| 6890 | and returns a list that consists of the original list without its |
| 6891 | first two elements. |
| 6892 | |
| 6893 | The @code{nthcdr} function does the same as repeating the call to |
| 6894 | @code{cdr}. In the following example, the argument 2 is passed to the |
| 6895 | function @code{nthcdr}, along with the list, and the value returned is |
| 6896 | the list without its first two items, which is exactly the same |
| 6897 | as repeating @code{cdr} twice on the list: |
| 6898 | |
| 6899 | @smallexample |
| 6900 | @group |
| 6901 | (nthcdr 2 '(pine fir oak maple)) |
| 6902 | @result{} (oak maple) |
| 6903 | @end group |
| 6904 | @end smallexample |
| 6905 | |
| 6906 | @need 1200 |
| 6907 | Using the original four element list, we can see what happens when |
| 6908 | various numeric arguments are passed to @code{nthcdr}, including 0, 1, |
| 6909 | and 5: |
| 6910 | |
| 6911 | @smallexample |
| 6912 | @group |
| 6913 | ;; @r{Leave the list as it was.} |
| 6914 | (nthcdr 0 '(pine fir oak maple)) |
| 6915 | @result{} (pine fir oak maple) |
| 6916 | @end group |
| 6917 | |
| 6918 | @group |
| 6919 | ;; @r{Return a copy without the first element.} |
| 6920 | (nthcdr 1 '(pine fir oak maple)) |
| 6921 | @result{} (fir oak maple) |
| 6922 | @end group |
| 6923 | |
| 6924 | @group |
| 6925 | ;; @r{Return a copy of the list without three elements.} |
| 6926 | (nthcdr 3 '(pine fir oak maple)) |
| 6927 | @result{} (maple) |
| 6928 | @end group |
| 6929 | |
| 6930 | @group |
| 6931 | ;; @r{Return a copy lacking all four elements.} |
| 6932 | (nthcdr 4 '(pine fir oak maple)) |
| 6933 | @result{} nil |
| 6934 | @end group |
| 6935 | |
| 6936 | @group |
| 6937 | ;; @r{Return a copy lacking all elements.} |
| 6938 | (nthcdr 5 '(pine fir oak maple)) |
| 6939 | @result{} nil |
| 6940 | @end group |
| 6941 | @end smallexample |
| 6942 | |
| 6943 | @node nth, setcar, nthcdr, car cdr & cons |
| 6944 | @comment node-name, next, previous, up |
| 6945 | @section @code{nth} |
| 6946 | @findex nth |
| 6947 | |
| 6948 | The @code{nthcdr} function takes the @sc{cdr} of a list repeatedly. |
| 6949 | The @code{nth} function takes the @sc{car} of the result returned by |
| 6950 | @code{nthcdr}. It returns the Nth element of the list. |
| 6951 | |
| 6952 | @need 1500 |
| 6953 | Thus, if it were not defined in C for speed, the definition of |
| 6954 | @code{nth} would be: |
| 6955 | |
| 6956 | @smallexample |
| 6957 | @group |
| 6958 | (defun nth (n list) |
| 6959 | "Returns the Nth element of LIST. |
| 6960 | N counts from zero. If LIST is not that long, nil is returned." |
| 6961 | (car (nthcdr n list))) |
| 6962 | @end group |
| 6963 | @end smallexample |
| 6964 | |
| 6965 | @noindent |
| 6966 | (Originally, @code{nth} was defined in Emacs Lisp in @file{subr.el}, |
| 6967 | but its definition was redone in C in the 1980s.) |
| 6968 | |
| 6969 | The @code{nth} function returns a single element of a list. |
| 6970 | This can be very convenient. |
| 6971 | |
| 6972 | Note that the elements are numbered from zero, not one. That is to |
| 6973 | say, the first element of a list, its @sc{car} is the zeroth element. |
| 6974 | This is called `zero-based' counting and often bothers people who |
| 6975 | are accustomed to the first element in a list being number one, which |
| 6976 | is `one-based'. |
| 6977 | |
| 6978 | @need 1250 |
| 6979 | For example: |
| 6980 | |
| 6981 | @smallexample |
| 6982 | @group |
| 6983 | (nth 0 '("one" "two" "three")) |
| 6984 | @result{} "one" |
| 6985 | |
| 6986 | (nth 1 '("one" "two" "three")) |
| 6987 | @result{} "two" |
| 6988 | @end group |
| 6989 | @end smallexample |
| 6990 | |
| 6991 | It is worth mentioning that @code{nth}, like @code{nthcdr} and |
| 6992 | @code{cdr}, does not change the original list---the function is |
| 6993 | non-destructive. This is in sharp contrast to the @code{setcar} and |
| 6994 | @code{setcdr} functions. |
| 6995 | |
| 6996 | @node setcar, setcdr, nth, car cdr & cons |
| 6997 | @comment node-name, next, previous, up |
| 6998 | @section @code{setcar} |
| 6999 | @findex setcar |
| 7000 | |
| 7001 | As you might guess from their names, the @code{setcar} and @code{setcdr} |
| 7002 | functions set the @sc{car} or the @sc{cdr} of a list to a new value. |
| 7003 | They actually change the original list, unlike @code{car} and @code{cdr} |
| 7004 | which leave the original list as it was. One way to find out how this |
| 7005 | works is to experiment. We will start with the @code{setcar} function. |
| 7006 | |
| 7007 | @need 1200 |
| 7008 | First, we can make a list and then set the value of a variable to the |
| 7009 | list, using the @code{setq} function. Here is a list of animals: |
| 7010 | |
| 7011 | @smallexample |
| 7012 | (setq animals '(antelope giraffe lion tiger)) |
| 7013 | @end smallexample |
| 7014 | |
| 7015 | @noindent |
| 7016 | If you are reading this in Info inside of GNU Emacs, you can evaluate |
| 7017 | this expression in the usual fashion, by positioning the cursor after |
| 7018 | the expression and typing @kbd{C-x C-e}. (I'm doing this right here as |
| 7019 | I write this. This is one of the advantages of having the interpreter |
| 7020 | built into the computing environment.) |
| 7021 | |
| 7022 | @need 1200 |
| 7023 | When we evaluate the variable @code{animals}, we see that it is bound to |
| 7024 | the list @code{(antelope giraffe lion tiger)}: |
| 7025 | |
| 7026 | @smallexample |
| 7027 | @group |
| 7028 | animals |
| 7029 | @result{} (antelope giraffe lion tiger) |
| 7030 | @end group |
| 7031 | @end smallexample |
| 7032 | |
| 7033 | @noindent |
| 7034 | Put another way, the variable @code{animals} points to the list |
| 7035 | @code{(antelope giraffe lion tiger)}. |
| 7036 | |
| 7037 | Next, evaluate the function @code{setcar} while passing it two |
| 7038 | arguments, the variable @code{animals} and the quoted symbol |
| 7039 | @code{hippopotamus}; this is done by writing the three element list |
| 7040 | @code{(setcar animals 'hippopotamus)} and then evaluating it in the |
| 7041 | usual fashion: |
| 7042 | |
| 7043 | @smallexample |
| 7044 | (setcar animals 'hippopotamus) |
| 7045 | @end smallexample |
| 7046 | |
| 7047 | @need 1200 |
| 7048 | @noindent |
| 7049 | After evaluating this expression, evaluate the variable @code{animals} |
| 7050 | again. You will see that the list of animals has changed: |
| 7051 | |
| 7052 | @smallexample |
| 7053 | @group |
| 7054 | animals |
| 7055 | @result{} (hippopotamus giraffe lion tiger) |
| 7056 | @end group |
| 7057 | @end smallexample |
| 7058 | |
| 7059 | @noindent |
| 7060 | The first element on the list, @code{antelope} is replaced by |
| 7061 | @code{hippopotamus}. |
| 7062 | |
| 7063 | So we can see that @code{setcar} did not add a new element to the list |
| 7064 | as @code{cons} would have; it replaced @code{giraffe} with |
| 7065 | @code{hippopotamus}; it @emph{changed} the list. |
| 7066 | |
| 7067 | @node setcdr, cons Exercise, setcar, car cdr & cons |
| 7068 | @comment node-name, next, previous, up |
| 7069 | @section @code{setcdr} |
| 7070 | @findex setcdr |
| 7071 | |
| 7072 | The @code{setcdr} function is similar to the @code{setcar} function, |
| 7073 | except that the function replaces the second and subsequent elements of |
| 7074 | a list rather than the first element. |
| 7075 | |
| 7076 | (To see how to change the last element of a list, look ahead to |
| 7077 | @ref{kill-new function, , The @code{kill-new} function}, which uses |
| 7078 | the @code{nthcdr} and @code{setcdr} functions.) |
| 7079 | |
| 7080 | @need 1200 |
| 7081 | To see how this works, set the value of the variable to a list of |
| 7082 | domesticated animals by evaluating the following expression: |
| 7083 | |
| 7084 | @smallexample |
| 7085 | (setq domesticated-animals '(horse cow sheep goat)) |
| 7086 | @end smallexample |
| 7087 | |
| 7088 | @need 1200 |
| 7089 | @noindent |
| 7090 | If you now evaluate the list, you will be returned the list |
| 7091 | @code{(horse cow sheep goat)}: |
| 7092 | |
| 7093 | @smallexample |
| 7094 | @group |
| 7095 | domesticated-animals |
| 7096 | @result{} (horse cow sheep goat) |
| 7097 | @end group |
| 7098 | @end smallexample |
| 7099 | |
| 7100 | @need 1200 |
| 7101 | Next, evaluate @code{setcdr} with two arguments, the name of the |
| 7102 | variable which has a list as its value, and the list to which the |
| 7103 | @sc{cdr} of the first list will be set; |
| 7104 | |
| 7105 | @smallexample |
| 7106 | (setcdr domesticated-animals '(cat dog)) |
| 7107 | @end smallexample |
| 7108 | |
| 7109 | @noindent |
| 7110 | If you evaluate this expression, the list @code{(cat dog)} will appear |
| 7111 | in the echo area. This is the value returned by the function. The |
| 7112 | result we are interested in is the ``side effect'', which we can see by |
| 7113 | evaluating the variable @code{domesticated-animals}: |
| 7114 | |
| 7115 | @smallexample |
| 7116 | @group |
| 7117 | domesticated-animals |
| 7118 | @result{} (horse cat dog) |
| 7119 | @end group |
| 7120 | @end smallexample |
| 7121 | |
| 7122 | @noindent |
| 7123 | Indeed, the list is changed from @code{(horse cow sheep goat)} to |
| 7124 | @code{(horse cat dog)}. The @sc{cdr} of the list is changed from |
| 7125 | @code{(cow sheep goat)} to @code{(cat dog)}. |
| 7126 | |
| 7127 | @node cons Exercise, , setcdr, car cdr & cons |
| 7128 | @section Exercise |
| 7129 | |
| 7130 | Construct a list of four birds by evaluating several expressions with |
| 7131 | @code{cons}. Find out what happens when you @code{cons} a list onto |
| 7132 | itself. Replace the first element of the list of four birds with a |
| 7133 | fish. Replace the rest of that list with a list of other fish. |
| 7134 | |
| 7135 | @node Cutting & Storing Text, List Implementation, car cdr & cons, Top |
| 7136 | @comment node-name, next, previous, up |
| 7137 | @chapter Cutting and Storing Text |
| 7138 | @cindex Cutting and storing text |
| 7139 | @cindex Storing and cutting text |
| 7140 | @cindex Killing text |
| 7141 | @cindex Clipping text |
| 7142 | @cindex Erasing text |
| 7143 | @cindex Deleting text |
| 7144 | |
| 7145 | Whenever you cut or clip text out of a buffer with a `kill' command in |
| 7146 | GNU Emacs, it is stored in a list and you can bring it back with a |
| 7147 | `yank' command. |
| 7148 | |
| 7149 | (The use of the word `kill' in Emacs for processes which specifically |
| 7150 | @emph{do not} destroy the values of the entities is an unfortunate |
| 7151 | historical accident. A much more appropriate word would be `clip' since |
| 7152 | that is what the kill commands do; they clip text out of a buffer and |
| 7153 | put it into storage from which it can be brought back. I have often |
| 7154 | been tempted to replace globally all occurrences of `kill' in the Emacs |
| 7155 | sources with `clip' and all occurrences of `killed' with `clipped'.) |
| 7156 | |
| 7157 | @menu |
| 7158 | * Storing Text:: Text is stored in a list. |
| 7159 | * zap-to-char:: Cutting out text up to a character. |
| 7160 | * kill-region:: Cutting text out of a region. |
| 7161 | * Digression into C:: Minor note on C programming language macros. |
| 7162 | * defvar:: How to give a variable an initial value. |
| 7163 | * copy-region-as-kill:: A definition for copying text. |
| 7164 | * cons & search-fwd Review:: |
| 7165 | * search Exercises:: |
| 7166 | @end menu |
| 7167 | |
| 7168 | @node Storing Text, zap-to-char, Cutting & Storing Text, Cutting & Storing Text |
| 7169 | @ifnottex |
| 7170 | @unnumberedsec Storing Text in a List |
| 7171 | @end ifnottex |
| 7172 | |
| 7173 | When text is cut out of a buffer, it is stored on a list. Successive |
| 7174 | pieces of text are stored on the list successively, so the list might |
| 7175 | look like this: |
| 7176 | |
| 7177 | @smallexample |
| 7178 | ("a piece of text" "previous piece") |
| 7179 | @end smallexample |
| 7180 | |
| 7181 | @need 1200 |
| 7182 | @noindent |
| 7183 | The function @code{cons} can be used to to create a new list from a |
| 7184 | piece of text (an `atom', to use the jargon) and an existing list, |
| 7185 | like this: |
| 7186 | |
| 7187 | @smallexample |
| 7188 | @group |
| 7189 | (cons "another piece" |
| 7190 | '("a piece of text" "previous piece")) |
| 7191 | @end group |
| 7192 | @end smallexample |
| 7193 | |
| 7194 | @need 1200 |
| 7195 | @noindent |
| 7196 | If you evaluate this expression, a list of three elements will appear in |
| 7197 | the echo area: |
| 7198 | |
| 7199 | @smallexample |
| 7200 | ("another piece" "a piece of text" "previous piece") |
| 7201 | @end smallexample |
| 7202 | |
| 7203 | With the @code{car} and @code{nthcdr} functions, you can retrieve |
| 7204 | whichever piece of text you want. For example, in the following code, |
| 7205 | @code{nthcdr 1 @dots{}} returns the list with the first item removed; |
| 7206 | and the @code{car} returns the first element of that remainder---the |
| 7207 | second element of the original list: |
| 7208 | |
| 7209 | @smallexample |
| 7210 | @group |
| 7211 | (car (nthcdr 1 '("another piece" |
| 7212 | "a piece of text" |
| 7213 | "previous piece"))) |
| 7214 | @result{} "a piece of text" |
| 7215 | @end group |
| 7216 | @end smallexample |
| 7217 | |
| 7218 | The actual functions in Emacs are more complex than this, of course. |
| 7219 | The code for cutting and retrieving text has to be written so that |
| 7220 | Emacs can figure out which element in the list you want---the first, |
| 7221 | second, third, or whatever. In addition, when you get to the end of |
| 7222 | the list, Emacs should give you the first element of the list, rather |
| 7223 | than nothing at all. |
| 7224 | |
| 7225 | The list that holds the pieces of text is called the @dfn{kill ring}. |
| 7226 | This chapter leads up to a description of the kill ring and how it is |
| 7227 | used by first tracing how the @code{zap-to-char} function works. This |
| 7228 | function uses (or `calls') a function that invokes a function that |
| 7229 | manipulates the kill ring. Thus, before reaching the mountains, we |
| 7230 | climb the foothills. |
| 7231 | |
| 7232 | A subsequent chapter describes how text that is cut from the buffer is |
| 7233 | retrieved. @xref{Yanking, , Yanking Text Back}. |
| 7234 | |
| 7235 | @node zap-to-char, kill-region, Storing Text, Cutting & Storing Text |
| 7236 | @comment node-name, next, previous, up |
| 7237 | @section @code{zap-to-char} |
| 7238 | @findex zap-to-char |
| 7239 | |
| 7240 | The @code{zap-to-char} function barely changed between GNU Emacs |
| 7241 | version 19 and GNU Emacs version 21. However, @code{zap-to-char} |
| 7242 | calls another function, @code{kill-region}, which enjoyed a major rewrite |
| 7243 | on the way to version 21. |
| 7244 | |
| 7245 | The @code{kill-region} function in Emacs 19 is complex, but does not |
| 7246 | use code that is important at this time. We will skip it. |
| 7247 | |
| 7248 | The @code{kill-region} function in Emacs 21 is easier to read than the |
| 7249 | same function in Emacs 19 and introduces a very important concept, |
| 7250 | that of error handling. We will walk through the function. |
| 7251 | |
| 7252 | But first, let us look at the interactive @code{zap-to-char} function. |
| 7253 | |
| 7254 | @menu |
| 7255 | * Complete zap-to-char:: The complete implementation. |
| 7256 | * zap-to-char interactive:: A three part interactive expression. |
| 7257 | * zap-to-char body:: A short overview. |
| 7258 | * search-forward:: How to search for a string. |
| 7259 | * progn:: The @code{progn} special form. |
| 7260 | * Summing up zap-to-char:: Using @code{point} and @code{search-forward}. |
| 7261 | @end menu |
| 7262 | |
| 7263 | @node Complete zap-to-char, zap-to-char interactive, zap-to-char, zap-to-char |
| 7264 | @ifnottex |
| 7265 | @unnumberedsubsec The Complete @code{zap-to-char} Implementation |
| 7266 | @end ifnottex |
| 7267 | |
| 7268 | The GNU Emacs version 19 and version 21 implementations of the |
| 7269 | @code{zap-to-char} function are nearly identical in form, and they |
| 7270 | work alike. The function removes the text in the region between the |
| 7271 | location of the cursor (i.e., of point) up to and including the next |
| 7272 | occurrence of a specified character. The text that @code{zap-to-char} |
| 7273 | removes is put in the kill ring; and it can be retrieved from the kill |
| 7274 | ring by typing @kbd{C-y} (@code{yank}). If the command is given an |
| 7275 | argument, it removes text through that number of occurrences. Thus, |
| 7276 | if the cursor were at the beginning of this sentence and the character |
| 7277 | were @samp{s}, @samp{Thus} would be removed. If the argument were |
| 7278 | two, @samp{Thus, if the curs} would be removed, up to and including |
| 7279 | the @samp{s} in @samp{cursor}. |
| 7280 | |
| 7281 | If the specified character is not found, @code{zap-to-char} will say |
| 7282 | ``Search failed'', tell you the character you typed, and not remove |
| 7283 | any text. |
| 7284 | |
| 7285 | In order to determine how much text to remove, @code{zap-to-char} uses |
| 7286 | a search function. Searches are used extensively in code that |
| 7287 | manipulates text, and we will focus attention on them as well as on the |
| 7288 | deletion command. |
| 7289 | |
| 7290 | @need 800 |
| 7291 | Here is the complete text of the version 19 implementation of the function: |
| 7292 | |
| 7293 | @c v 19 |
| 7294 | @smallexample |
| 7295 | @group |
| 7296 | (defun zap-to-char (arg char) ; version 19 implementation |
| 7297 | "Kill up to and including ARG'th occurrence of CHAR. |
| 7298 | Goes backward if ARG is negative; error if CHAR not found." |
| 7299 | (interactive "*p\ncZap to char: ") |
| 7300 | (kill-region (point) |
| 7301 | (progn |
| 7302 | (search-forward |
| 7303 | (char-to-string char) nil nil arg) |
| 7304 | (point)))) |
| 7305 | @end group |
| 7306 | @end smallexample |
| 7307 | |
| 7308 | @node zap-to-char interactive, zap-to-char body, Complete zap-to-char, zap-to-char |
| 7309 | @comment node-name, next, previous, up |
| 7310 | @subsection The @code{interactive} Expression |
| 7311 | |
| 7312 | @need 800 |
| 7313 | The interactive expression in the @code{zap-to-char} command looks like |
| 7314 | this: |
| 7315 | |
| 7316 | @smallexample |
| 7317 | (interactive "*p\ncZap to char: ") |
| 7318 | @end smallexample |
| 7319 | |
| 7320 | The part within quotation marks, @code{"*p\ncZap to char:@: "}, specifies |
| 7321 | three different things. First, and most simply, the asterisk, @samp{*}, |
| 7322 | causes an error to be signaled if the buffer is read-only. This means that |
| 7323 | if you try @code{zap-to-char} in a read-only buffer you will not be able to |
| 7324 | remove text, and you will receive a message that says ``Buffer is |
| 7325 | read-only''; your terminal may beep at you as well. |
| 7326 | |
| 7327 | The version 21 implementation does not have the asterisk, @samp{*}. The |
| 7328 | function works the same as in version 19: in both cases, it cannot |
| 7329 | remove text from a read-only buffer but the function does copy the |
| 7330 | text that would have been removed to the kill ring. Also, in both |
| 7331 | cases, you see an error message. |
| 7332 | |
| 7333 | However, the version 19 implementation copies text from a read-only |
| 7334 | buffer only because of a mistake in the implementation of |
| 7335 | @code{interactive}. According to the documentation for |
| 7336 | @code{interactive}, the asterisk, @samp{*}, should prevent the |
| 7337 | @code{zap-to-char} function from doing anything at all when the buffer |
| 7338 | is read only. In version 19, the function should not copy the text to |
| 7339 | the kill ring. It is a bug that it does. |
| 7340 | |
| 7341 | In version 21, the function is designed to copy the text to the kill |
| 7342 | ring; moreover, @code{interactive} is implemented correctly. So the |
| 7343 | asterisk, @samp{*}, had to be removed from the interactive |
| 7344 | specification. However, if you insert an @samp{*} yourself and |
| 7345 | evaluate the function definition, then the next time you run the |
| 7346 | @code{zap-to-char} function on a read-only buffer, you will not copy |
| 7347 | any text. |
| 7348 | |
| 7349 | That change aside, and a change to the documentation, the two versions |
| 7350 | of the @code{zap-to-char} function are identical. |
| 7351 | |
| 7352 | Let us continue with the interactive specification. |
| 7353 | |
| 7354 | The second part of @code{"*p\ncZap to char:@: "} is the @samp{p}. |
| 7355 | This part is separated from the next part by a newline, @samp{\n}. |
| 7356 | The @samp{p} means that the first argument to the function will be |
| 7357 | passed the value of a `processed prefix'. The prefix argument is |
| 7358 | passed by typing @kbd{C-u} and a number, or @kbd{M-} and a number. If |
| 7359 | the function is called interactively without a prefix, 1 is passed to |
| 7360 | this argument. |
| 7361 | |
| 7362 | The third part of @code{"*p\ncZap to char:@: "} is @samp{cZap to char:@: |
| 7363 | }. In this part, the lower case @samp{c} indicates that |
| 7364 | @code{interactive} expects a prompt and that the argument will be a |
| 7365 | character. The prompt follows the @samp{c} and is the string @samp{Zap |
| 7366 | to char:@: } (with a space after the colon to make it look good). |
| 7367 | |
| 7368 | What all this does is prepare the arguments to @code{zap-to-char} so they |
| 7369 | are of the right type, and give the user a prompt. |
| 7370 | |
| 7371 | @node zap-to-char body, search-forward, zap-to-char interactive, zap-to-char |
| 7372 | @comment node-name, next, previous, up |
| 7373 | @subsection The Body of @code{zap-to-char} |
| 7374 | |
| 7375 | The body of the @code{zap-to-char} function contains the code that |
| 7376 | kills (that is, removes) the text in the region from the current |
| 7377 | position of the cursor up to and including the specified character. |
| 7378 | The first part of the code looks like this: |
| 7379 | |
| 7380 | @smallexample |
| 7381 | (kill-region (point) @dots{} |
| 7382 | @end smallexample |
| 7383 | |
| 7384 | @noindent |
| 7385 | @code{(point)} is the current position of the cursor. |
| 7386 | |
| 7387 | The next part of the code is an expression using @code{progn}. The body |
| 7388 | of the @code{progn} consists of calls to @code{search-forward} and |
| 7389 | @code{point}. |
| 7390 | |
| 7391 | It is easier to understand how @code{progn} works after learning about |
| 7392 | @code{search-forward}, so we will look at @code{search-forward} and |
| 7393 | then at @code{progn}. |
| 7394 | |
| 7395 | @node search-forward, progn, zap-to-char body, zap-to-char |
| 7396 | @comment node-name, next, previous, up |
| 7397 | @subsection The @code{search-forward} Function |
| 7398 | @findex search-forward |
| 7399 | |
| 7400 | The @code{search-forward} function is used to locate the |
| 7401 | zapped-for-character in @code{zap-to-char}. If the search is |
| 7402 | successful, @code{search-forward} leaves point immediately after the |
| 7403 | last character in the target string. (In @code{zap-to-char}, the |
| 7404 | target string is just one character long.) If the search is |
| 7405 | backwards, @code{search-forward} leaves point just before the first |
| 7406 | character in the target. Also, @code{search-forward} returns @code{t} |
| 7407 | for true. (Moving point is therefore a `side effect'.) |
| 7408 | |
| 7409 | @need 1250 |
| 7410 | In @code{zap-to-char}, the @code{search-forward} function looks like this: |
| 7411 | |
| 7412 | @smallexample |
| 7413 | (search-forward (char-to-string char) nil nil arg) |
| 7414 | @end smallexample |
| 7415 | |
| 7416 | The @code{search-forward} function takes four arguments: |
| 7417 | |
| 7418 | @enumerate |
| 7419 | @item |
| 7420 | The first argument is the target, what is searched for. This must be a |
| 7421 | string, such as @samp{"z"}. |
| 7422 | |
| 7423 | As it happens, the argument passed to @code{zap-to-char} is a single |
| 7424 | character. Because of the way computers are built, the Lisp |
| 7425 | interpreter may treat a single character as being different from a |
| 7426 | string of characters. Inside the computer, a single character has a |
| 7427 | different electronic format than a string of one character. (A single |
| 7428 | character can often be recorded in the computer using exactly one |
| 7429 | byte; but a string may be longer, and the computer needs to be ready |
| 7430 | for this.) Since the @code{search-forward} function searches for a |
| 7431 | string, the character that the @code{zap-to-char} function receives as |
| 7432 | its argument must be converted inside the computer from one format to |
| 7433 | the other; otherwise the @code{search-forward} function will fail. |
| 7434 | The @code{char-to-string} function is used to make this conversion. |
| 7435 | |
| 7436 | @item |
| 7437 | The second argument bounds the search; it is specified as a position in |
| 7438 | the buffer. In this case, the search can go to the end of the buffer, |
| 7439 | so no bound is set and the second argument is @code{nil}. |
| 7440 | |
| 7441 | @item |
| 7442 | The third argument tells the function what it should do if the search |
| 7443 | fails---it can signal an error (and print a message) or it can return |
| 7444 | @code{nil}. A @code{nil} as the third argument causes the function to |
| 7445 | signal an error when the search fails. |
| 7446 | |
| 7447 | @item |
| 7448 | The fourth argument to @code{search-forward} is the repeat count---how |
| 7449 | many occurrences of the string to look for. This argument is optional |
| 7450 | and if the function is called without a repeat count, this argument is |
| 7451 | passed the value 1. If this argument is negative, the search goes |
| 7452 | backwards. |
| 7453 | @end enumerate |
| 7454 | |
| 7455 | @need 800 |
| 7456 | In template form, a @code{search-forward} expression looks like this: |
| 7457 | |
| 7458 | @smallexample |
| 7459 | @group |
| 7460 | (search-forward "@var{target-string}" |
| 7461 | @var{limit-of-search} |
| 7462 | @var{what-to-do-if-search-fails} |
| 7463 | @var{repeat-count}) |
| 7464 | @end group |
| 7465 | @end smallexample |
| 7466 | |
| 7467 | We will look at @code{progn} next. |
| 7468 | |
| 7469 | @node progn, Summing up zap-to-char, search-forward, zap-to-char |
| 7470 | @comment node-name, next, previous, up |
| 7471 | @subsection The @code{progn} Special Form |
| 7472 | @findex progn |
| 7473 | |
| 7474 | @code{progn} is a special form that causes each of its arguments to be |
| 7475 | evaluated in sequence and then returns the value of the last one. The |
| 7476 | preceding expressions are evaluated only for the side effects they |
| 7477 | perform. The values produced by them are discarded. |
| 7478 | |
| 7479 | @need 800 |
| 7480 | The template for a @code{progn} expression is very simple: |
| 7481 | |
| 7482 | @smallexample |
| 7483 | @group |
| 7484 | (progn |
| 7485 | @var{body}@dots{}) |
| 7486 | @end group |
| 7487 | @end smallexample |
| 7488 | |
| 7489 | In @code{zap-to-char}, the @code{progn} expression has to do two things: |
| 7490 | put point in exactly the right position; and return the location of |
| 7491 | point so that @code{kill-region} will know how far to kill to. |
| 7492 | |
| 7493 | The first argument to the @code{progn} is @code{search-forward}. When |
| 7494 | @code{search-forward} finds the string, the function leaves point |
| 7495 | immediately after the last character in the target string. (In this |
| 7496 | case the target string is just one character long.) If the search is |
| 7497 | backwards, @code{search-forward} leaves point just before the first |
| 7498 | character in the target. The movement of point is a side effect. |
| 7499 | |
| 7500 | The second and last argument to @code{progn} is the expression |
| 7501 | @code{(point)}. This expression returns the value of point, which in |
| 7502 | this case will be the location to which it has been moved by |
| 7503 | @code{search-forward}. This value is returned by the @code{progn} |
| 7504 | expression and is passed to @code{kill-region} as @code{kill-region}'s |
| 7505 | second argument. |
| 7506 | |
| 7507 | @node Summing up zap-to-char, , progn, zap-to-char |
| 7508 | @comment node-name, next, previous, up |
| 7509 | @subsection Summing up @code{zap-to-char} |
| 7510 | |
| 7511 | Now that we have seen how @code{search-forward} and @code{progn} work, |
| 7512 | we can see how the @code{zap-to-char} function works as a whole. |
| 7513 | |
| 7514 | The first argument to @code{kill-region} is the position of the cursor |
| 7515 | when the @code{zap-to-char} command is given---the value of point at |
| 7516 | that time. Within the @code{progn}, the search function then moves |
| 7517 | point to just after the zapped-to-character and @code{point} returns the |
| 7518 | value of this location. The @code{kill-region} function puts together |
| 7519 | these two values of point, the first one as the beginning of the region |
| 7520 | and the second one as the end of the region, and removes the region. |
| 7521 | |
| 7522 | The @code{progn} special form is necessary because the @code{kill-region} |
| 7523 | command takes two arguments; and it would fail if @code{search-forward} |
| 7524 | and @code{point} expressions were written in sequence as two |
| 7525 | additional arguments. The @code{progn} expression is a single argument |
| 7526 | to @code{kill-region} and returns the one value that @code{kill-region} |
| 7527 | needs for its second argument. |
| 7528 | |
| 7529 | @node kill-region, Digression into C, zap-to-char, Cutting & Storing Text |
| 7530 | @comment node-name, next, previous, up |
| 7531 | @section @code{kill-region} |
| 7532 | @findex kill-region |
| 7533 | |
| 7534 | The @code{zap-to-char} function uses the @code{kill-region} function. |
| 7535 | This function clips text from a region and copies that text to |
| 7536 | the kill ring, from which it may be retrieved. |
| 7537 | |
| 7538 | The Emacs 21 version of that function uses @code{condition-case} and |
| 7539 | @code{copy-region-as-kill}, both of which we will explain. |
| 7540 | @code{condition-case} is an important special form. |
| 7541 | |
| 7542 | In essence, the @code{kill-region} function calls |
| 7543 | @code{condition-case}, which takes three arguments. In this function, |
| 7544 | the first argument does nothing. The second argument contains the |
| 7545 | code that does the work when all goes well. The third argument |
| 7546 | contains the code that is called in the event of an error. |
| 7547 | |
| 7548 | @menu |
| 7549 | * Complete kill-region:: The function definition. |
| 7550 | * condition-case:: Dealing with a problem. |
| 7551 | * delete-and-extract-region:: Doing the work. |
| 7552 | @end menu |
| 7553 | |
| 7554 | @node Complete kill-region, condition-case, kill-region, kill-region |
| 7555 | @ifnottex |
| 7556 | @unnumberedsubsec The Complete @code{kill-region} Definition |
| 7557 | @end ifnottex |
| 7558 | |
| 7559 | @need 1200 |
| 7560 | We will go through the @code{condition-case} code in a moment. First, |
| 7561 | let us look at the original definition of @code{kill-region}, with |
| 7562 | comments added (the newer definition has an optional third argument |
| 7563 | and is more complex): |
| 7564 | |
| 7565 | @c v 21 |
| 7566 | @smallexample |
| 7567 | @group |
| 7568 | (defun kill-region (beg end) |
| 7569 | "Kill between point and mark. |
| 7570 | The text is deleted but saved in the kill ring." |
| 7571 | (interactive "r") |
| 7572 | @end group |
| 7573 | |
| 7574 | @group |
| 7575 | ;; 1. `condition-case' takes three arguments. |
| 7576 | ;; If the first argument is nil, as it is here, |
| 7577 | ;; information about the error signal is not |
| 7578 | ;; stored for use by another function. |
| 7579 | (condition-case nil |
| 7580 | @end group |
| 7581 | |
| 7582 | @group |
| 7583 | ;; 2. The second argument to `condition-case' |
| 7584 | ;; tells the Lisp interpreter what to do when all goes well. |
| 7585 | @end group |
| 7586 | |
| 7587 | @group |
| 7588 | ;; The `delete-and-extract-region' function usually does the |
| 7589 | ;; work. If the beginning and ending of the region are both |
| 7590 | ;; the same, then the variable `string' will be empty, or nil |
| 7591 | (let ((string (delete-and-extract-region beg end))) |
| 7592 | @end group |
| 7593 | |
| 7594 | @group |
| 7595 | ;; `when' is an `if' clause that cannot take an `else-part'. |
| 7596 | ;; Emacs normally sets the value of `last-command' to the |
| 7597 | ;; previous command. |
| 7598 | @end group |
| 7599 | @group |
| 7600 | ;; `kill-append' concatenates the new string and the old. |
| 7601 | ;; `kill-new' inserts text into a new item in the kill ring. |
| 7602 | (when string |
| 7603 | (if (eq last-command 'kill-region) |
| 7604 | ;; if true, prepend string |
| 7605 | (kill-append string (< end beg)) |
| 7606 | (kill-new string))) |
| 7607 | (setq this-command 'kill-region)) |
| 7608 | @end group |
| 7609 | |
| 7610 | @group |
| 7611 | ;; 3. The third argument to `condition-case' tells the interpreter |
| 7612 | ;; what to do with an error. |
| 7613 | @end group |
| 7614 | @group |
| 7615 | ;; The third argument has a conditions part and a body part. |
| 7616 | ;; If the conditions are met (in this case, |
| 7617 | ;; if text or buffer is read-only) |
| 7618 | ;; then the body is executed. |
| 7619 | @end group |
| 7620 | @group |
| 7621 | ((buffer-read-only text-read-only) ;; this is the if-part |
| 7622 | ;; then... |
| 7623 | (copy-region-as-kill beg end) |
| 7624 | @end group |
| 7625 | @group |
| 7626 | (if kill-read-only-ok ;; usually this variable is nil |
| 7627 | (message "Read only text copied to kill ring") |
| 7628 | ;; or else, signal an error if the buffer is read-only; |
| 7629 | (barf-if-buffer-read-only) |
| 7630 | ;; and, in any case, signal that the text is read-only. |
| 7631 | (signal 'text-read-only (list (current-buffer))))))) |
| 7632 | @end group |
| 7633 | @end smallexample |
| 7634 | |
| 7635 | @node condition-case, delete-and-extract-region, Complete kill-region, kill-region |
| 7636 | @comment node-name, next, previous, up |
| 7637 | @subsection @code{condition-case} |
| 7638 | @findex condition-case |
| 7639 | |
| 7640 | As we have seen earlier (@pxref{Making Errors, , Generate an Error |
| 7641 | Message}), when the Emacs Lisp interpreter has trouble evaluating an |
| 7642 | expression, it provides you with help; in the jargon, this is called |
| 7643 | ``signaling an error''. Usually, the computer stops the program and |
| 7644 | shows you a message. |
| 7645 | |
| 7646 | However, some programs undertake complicated actions. They should not |
| 7647 | simply stop on an error. In the @code{kill-region} function, the most |
| 7648 | likely error is that you will try to kill text that is read-only and |
| 7649 | cannot be removed. So the @code{kill-region} function contains code |
| 7650 | to handle this circumstance. This code, which makes up the body of |
| 7651 | the @code{kill-region} function, is inside of a @code{condition-case} |
| 7652 | special form. |
| 7653 | |
| 7654 | @need 800 |
| 7655 | The template for @code{condition-case} looks like this: |
| 7656 | |
| 7657 | @smallexample |
| 7658 | @group |
| 7659 | (condition-case |
| 7660 | @var{var} |
| 7661 | @var{bodyform} |
| 7662 | @var{error-handler}@dots{}) |
| 7663 | @end group |
| 7664 | @end smallexample |
| 7665 | |
| 7666 | The second argument, @var{bodyform}, is straightforward. The |
| 7667 | @code{condition-case} special form causes the Lisp interpreter to |
| 7668 | evaluate the code in @var{bodyform}. If no error occurs, the special |
| 7669 | form returns the code's value and produces the side-effects, if any. |
| 7670 | |
| 7671 | In short, the @var{bodyform} part of a @code{condition-case} |
| 7672 | expression determines what should happen when everything works |
| 7673 | correctly. |
| 7674 | |
| 7675 | However, if an error occurs, among its other actions, the function |
| 7676 | generating the error signal will define one or more error condition |
| 7677 | names. |
| 7678 | |
| 7679 | An error handler is the third argument to @code{condition case}. |
| 7680 | An error handler has two parts, a @var{condition-name} and a |
| 7681 | @var{body}. If the @var{condition-name} part of an error handler |
| 7682 | matches a condition name generated by an error, then the @var{body} |
| 7683 | part of the error handler is run. |
| 7684 | |
| 7685 | As you will expect, the @var{condition-name} part of an error handler |
| 7686 | may be either a single condition name or a list of condition names. |
| 7687 | |
| 7688 | Also, a complete @code{condition-case} expression may contain more |
| 7689 | than one error handler. When an error occurs, the first applicable |
| 7690 | handler is run. |
| 7691 | |
| 7692 | Lastly, the first argument to the @code{condition-case} expression, |
| 7693 | the @var{var} argument, is sometimes bound to a variable that |
| 7694 | contains information about the error. However, if that argument is |
| 7695 | nil, as is the case in @code{kill-region}, that information is |
| 7696 | discarded. |
| 7697 | |
| 7698 | @need 1200 |
| 7699 | In brief, in the @code{kill-region} function, the code |
| 7700 | @code{condition-case} works like this: |
| 7701 | |
| 7702 | @smallexample |
| 7703 | @group |
| 7704 | @var{If no errors}, @var{run only this code} |
| 7705 | @var{but}, @var{if errors}, @var{run this other code}. |
| 7706 | @end group |
| 7707 | @end smallexample |
| 7708 | |
| 7709 | @node delete-and-extract-region, , condition-case, kill-region |
| 7710 | @comment node-name, next, previous, up |
| 7711 | @subsection @code{delete-and-extract-region} |
| 7712 | @findex delete-and-extract-region |
| 7713 | |
| 7714 | A @code{condition-case} expression has two parts, a part that is |
| 7715 | evaluated in the expectation that all will go well, but which may |
| 7716 | generate an error; and a part that is evaluated when there is an |
| 7717 | error. |
| 7718 | |
| 7719 | First, let us look at the code in @code{kill-region} that is run in |
| 7720 | the expectation that all goes well. This is the core of the function. |
| 7721 | The code looks like this: |
| 7722 | |
| 7723 | @smallexample |
| 7724 | @group |
| 7725 | (let ((string (delete-and-extract-region beg end))) |
| 7726 | (when string |
| 7727 | (if (eq last-command 'kill-region) |
| 7728 | (kill-append string (< end beg)) |
| 7729 | (kill-new string))) |
| 7730 | (setq this-command 'kill-region)) |
| 7731 | @end group |
| 7732 | @end smallexample |
| 7733 | |
| 7734 | It looks complicated because we have the new functions |
| 7735 | @code{delete-and-extract-region}, @code{kill-append}, and |
| 7736 | @code{kill-new} as well as the new variables, |
| 7737 | @code{last-command} and @code{this-command}. |
| 7738 | |
| 7739 | The @code{delete-and-extract-region} function is straightforward. It |
| 7740 | is a built-in function that deletes the text in a region (a side |
| 7741 | effect) and also returns that text. This is the function that |
| 7742 | actually removes the text. (And if it cannot do that, it signals the |
| 7743 | error.) |
| 7744 | |
| 7745 | In this @code{let} expression, the text that |
| 7746 | @code{delete-and-extract-region} returns is placed in the local |
| 7747 | variable called @samp{string}. This is the text that is removed from |
| 7748 | the buffer. (To be more precise, the variable is set to point to the |
| 7749 | address of the extracted text; to say it is `placed in' the variable |
| 7750 | is simply a shorthand.) |
| 7751 | |
| 7752 | If the variable @samp{string} does point to text, that text is added |
| 7753 | to the kill ring. The variable will have a @code{nil} value if no |
| 7754 | text was removed. |
| 7755 | |
| 7756 | The code uses @code{when} to determine whether the variable |
| 7757 | @samp{string} points to text. A @code{when} statement is simply a |
| 7758 | programmers' convenience. A @code{when} statement is an @code{if} |
| 7759 | statement without the possibility of an else clause. In your mind, you |
| 7760 | can replace @code{when} with @code{if} and understand what goes on. |
| 7761 | That is what the Lisp interpreter does. |
| 7762 | |
| 7763 | @cindex Macro, lisp |
| 7764 | @cindex Lisp macro |
| 7765 | Technically speaking, @code{when} is a Lisp macro. A Lisp @dfn{macro} |
| 7766 | enables you to define new control constructs and other language |
| 7767 | features. It tells the interpreter how to compute another Lisp |
| 7768 | expression which will in turn compute the value. In this case, the |
| 7769 | `other expression' is an @code{if} expression. For more about Lisp |
| 7770 | macros, see @ref{Macros, , Macros, elisp, The GNU Emacs Lisp Reference |
| 7771 | Manual}. The C programming language also provides macros. These are |
| 7772 | different, but also useful. We will briefly look at C macros in |
| 7773 | @ref{Digression into C}. |
| 7774 | |
| 7775 | @need 1200 |
| 7776 | If the string has content, then another conditional expression is |
| 7777 | executed. This is an @code{if} with both a then-part and an else-part. |
| 7778 | |
| 7779 | @smallexample |
| 7780 | @group |
| 7781 | (if (eq last-command 'kill-region) |
| 7782 | (kill-append string (< end beg)) |
| 7783 | (kill-new string))) |
| 7784 | @end group |
| 7785 | @end smallexample |
| 7786 | |
| 7787 | The then-part is evaluated if the previous command was another call to |
| 7788 | @code{kill-region}; if not, the else-part is evaluated. |
| 7789 | |
| 7790 | @code{last-command} is a variable that comes with Emacs that we have |
| 7791 | not seen before. Normally, whenever a function is executed, Emacs |
| 7792 | sets the value of @code{last-command} to the previous command. |
| 7793 | |
| 7794 | @need 1200 |
| 7795 | In this segment of the definition, the @code{if} expression checks |
| 7796 | whether the previous command was @code{kill-region}. If it was, |
| 7797 | |
| 7798 | @smallexample |
| 7799 | (kill-append string (< end beg)) |
| 7800 | @end smallexample |
| 7801 | |
| 7802 | @noindent |
| 7803 | concatenates a copy of the newly clipped text to the just previously |
| 7804 | clipped text in the kill ring. (If the @w{@code{(< end beg))}} |
| 7805 | expression is true, @code{kill-append} prepends the string to the just |
| 7806 | previously clipped text. For a detailed discussion, see |
| 7807 | @ref{kill-append function, , The @code{kill-append} function}.) |
| 7808 | |
| 7809 | If you then yank back the text, i.e., `paste' it, you get both |
| 7810 | pieces of text at once. That way, if you delete two words in a row, |
| 7811 | and then yank them back, you get both words, in their proper order, |
| 7812 | with one yank. (The @w{@code{(< end beg))}} expression makes sure the |
| 7813 | order is correct.) |
| 7814 | |
| 7815 | On the other hand, if the previous command is not @code{kill-region}, |
| 7816 | then the @code{kill-new} function is called, which adds the text to |
| 7817 | the kill ring as the latest item, and sets the |
| 7818 | @code{kill-ring-yank-pointer} variable to point to it. |
| 7819 | |
| 7820 | @node Digression into C, defvar, kill-region, Cutting & Storing Text |
| 7821 | @comment node-name, next, previous, up |
| 7822 | @section Digression into C |
| 7823 | @findex delete-and-extract-region |
| 7824 | @cindex C, a digression into |
| 7825 | @cindex Digression into C |
| 7826 | |
| 7827 | The @code{zap-to-char} command uses the |
| 7828 | @code{delete-and-extract-region} function, which in turn uses two |
| 7829 | other functions, @code{copy-region-as-kill} and |
| 7830 | @code{del_range_1}. The @code{copy-region-as-kill} function will be |
| 7831 | described in a following section; it puts a copy of the region in the |
| 7832 | kill ring so it can be yanked back. (@xref{copy-region-as-kill, , |
| 7833 | @code{copy-region-as-kill}}.) |
| 7834 | |
| 7835 | The @code{delete-and-extract-region} function removes the contents of |
| 7836 | a region and you cannot get them back. |
| 7837 | |
| 7838 | Unlike the other code discussed here, @code{delete-and-extract-region} |
| 7839 | is not written in Emacs Lisp; it is written in C and is one of the |
| 7840 | primitives of the GNU Emacs system. Since it is very simple, I will |
| 7841 | digress briefly from Lisp and describe it here. |
| 7842 | |
| 7843 | @need 1500 |
| 7844 | Like many of the other Emacs primitives, |
| 7845 | @code{delete-and-extract-region} is written as an instance of a C |
| 7846 | macro, a macro being a template for code. The complete macro looks |
| 7847 | like this: |
| 7848 | |
| 7849 | @c /usr/local/src/emacs/src/editfns.c |
| 7850 | @smallexample |
| 7851 | @group |
| 7852 | DEFUN ("delete-and-extract-region", Fdelete_and_extract_region, |
| 7853 | Sdelete_and_extract_region, 2, 2, 0, |
| 7854 | "Delete the text between START and END and return it.") |
| 7855 | (start, end) |
| 7856 | Lisp_Object start, end; |
| 7857 | @{ |
| 7858 | validate_region (&start, &end); |
| 7859 | return del_range_1 (XINT (start), XINT (end), 1, 1); |
| 7860 | @} |
| 7861 | @end group |
| 7862 | @end smallexample |
| 7863 | |
| 7864 | Without going into the details of the macro writing process, let me |
| 7865 | point out that this macro starts with the word @code{DEFUN}. The word |
| 7866 | @code{DEFUN} was chosen since the code serves the same purpose as |
| 7867 | @code{defun} does in Lisp. The word @code{DEFUN} is followed by seven |
| 7868 | parts inside of parentheses: |
| 7869 | |
| 7870 | @itemize @bullet |
| 7871 | @item |
| 7872 | The first part is the name given to the function in Lisp, |
| 7873 | @code{delete-and-extract-region}. |
| 7874 | |
| 7875 | @item |
| 7876 | The second part is the name of the function in C, |
| 7877 | @code{Fdelete_and_extract_region}. By convention, it starts with |
| 7878 | @samp{F}. Since C does not use hyphens in names, underscores are used |
| 7879 | instead. |
| 7880 | |
| 7881 | @item |
| 7882 | The third part is the name for the C constant structure that records |
| 7883 | information on this function for internal use. It is the name of the |
| 7884 | function in C but begins with an @samp{S} instead of an @samp{F}. |
| 7885 | |
| 7886 | @item |
| 7887 | The fourth and fifth parts specify the minimum and maximum number of |
| 7888 | arguments the function can have. This function demands exactly 2 |
| 7889 | arguments. |
| 7890 | |
| 7891 | @item |
| 7892 | The sixth part is nearly like the argument that follows the |
| 7893 | @code{interactive} declaration in a function written in Lisp: a letter |
| 7894 | followed, perhaps, by a prompt. The only difference from the Lisp is |
| 7895 | when the macro is called with no arguments. Then you write a @code{0} |
| 7896 | (which is a `null string'), as in this macro. |
| 7897 | |
| 7898 | If you were to specify arguments, you would place them between |
| 7899 | quotation marks. The C macro for @code{goto-char} includes |
| 7900 | @code{"NGoto char: "} in this position to indicate that the function |
| 7901 | expects a raw prefix, in this case, a numerical location in a buffer, |
| 7902 | and provides a prompt. |
| 7903 | |
| 7904 | @item |
| 7905 | The seventh part is a documentation string, just like the one for a |
| 7906 | function written in Emacs Lisp, except that every newline must be |
| 7907 | written explicitly as @samp{\n} followed by a backslash and carriage |
| 7908 | return. |
| 7909 | |
| 7910 | @need 1000 |
| 7911 | Thus, the first two lines of documentation for @code{goto-char} are |
| 7912 | written like this: |
| 7913 | |
| 7914 | @smallexample |
| 7915 | @group |
| 7916 | "Set point to POSITION, a number or marker.\n\ |
| 7917 | Beginning of buffer is position (point-min), end is (point-max). |
| 7918 | @end group |
| 7919 | @end smallexample |
| 7920 | @end itemize |
| 7921 | |
| 7922 | @need 1200 |
| 7923 | In a C macro, the formal parameters come next, with a statement of |
| 7924 | what kind of object they are, followed by what might be called the `body' |
| 7925 | of the macro. For @code{delete-and-extract-region} the `body' |
| 7926 | consists of the following two lines: |
| 7927 | |
| 7928 | @smallexample |
| 7929 | @group |
| 7930 | validate_region (&start, &end); |
| 7931 | return del_range_1 (XINT (start), XINT (end), 1, 1); |
| 7932 | @end group |
| 7933 | @end smallexample |
| 7934 | |
| 7935 | The first function, @code{validate_region} checks whether the values |
| 7936 | passed as the beginning and end of the region are the proper type and |
| 7937 | are within range. The second function, @code{del_range_1}, actually |
| 7938 | deletes the text. |
| 7939 | |
| 7940 | @code{del_range_1} is a complex function we will not look into. It |
| 7941 | updates the buffer and does other things. |
| 7942 | |
| 7943 | However, it is worth looking at the two arguments passed to |
| 7944 | @code{del_range}. These are @w{@code{XINT (start)}} and @w{@code{XINT |
| 7945 | (end)}}. |
| 7946 | |
| 7947 | As far as the C language is concerned, @code{start} and @code{end} are |
| 7948 | two integers that mark the beginning and end of the region to be |
| 7949 | deleted@footnote{More precisely, and requiring more expert knowledge |
| 7950 | to understand, the two integers are of type `Lisp_Object', which can |
| 7951 | also be a C union instead of an integer type.}. |
| 7952 | |
| 7953 | In early versions of Emacs, these two numbers were thirty-two bits |
| 7954 | long, but the code is slowly being generalized to handle other |
| 7955 | lengths. Three of the available bits are used to specify the type of |
| 7956 | information and a fourth bit is used for handling the computer's |
| 7957 | memory; the remaining bits are used as `content'. |
| 7958 | |
| 7959 | @samp{XINT} is a C macro that extracts the relevant number from the |
| 7960 | longer collection of bits; the four other bits are discarded. |
| 7961 | |
| 7962 | @need 800 |
| 7963 | The command in @code{delete-and-extract-region} looks like this: |
| 7964 | |
| 7965 | @smallexample |
| 7966 | del_range_1 (XINT (start), XINT (end), 1, 1); |
| 7967 | @end smallexample |
| 7968 | |
| 7969 | @noindent |
| 7970 | It deletes the region between the beginning position, @code{start}, |
| 7971 | and the ending position, @code{end}. |
| 7972 | |
| 7973 | From the point of view of the person writing Lisp, Emacs is all very |
| 7974 | simple; but hidden underneath is a great deal of complexity to make it |
| 7975 | all work. |
| 7976 | |
| 7977 | @node defvar, copy-region-as-kill, Digression into C, Cutting & Storing Text |
| 7978 | @comment node-name, next, previous, up |
| 7979 | @section Initializing a Variable with @code{defvar} |
| 7980 | @findex defvar |
| 7981 | @cindex Initializing a variable |
| 7982 | @cindex Variable initialization |
| 7983 | |
| 7984 | Unlike the @code{delete-and-extract-region} function, the |
| 7985 | @code{copy-region-as-kill} function is written in Emacs Lisp. Two |
| 7986 | functions within it, @code{kill-append} and @code{kill-new}, copy a |
| 7987 | region in a buffer and save it in a variable called the |
| 7988 | @code{kill-ring}. This section describes how the @code{kill-ring} |
| 7989 | variable is created and initialized using the @code{defvar} special |
| 7990 | form. |
| 7991 | |
| 7992 | (Again we note that the term @code{kill-ring} is a misnomer. The text |
| 7993 | that is clipped out of the buffer can be brought back; it is not a ring |
| 7994 | of corpses, but a ring of resurrectable text.) |
| 7995 | |
| 7996 | In Emacs Lisp, a variable such as the @code{kill-ring} is created and |
| 7997 | given an initial value by using the @code{defvar} special form. The |
| 7998 | name comes from ``define variable''. |
| 7999 | |
| 8000 | The @code{defvar} special form is similar to @code{setq} in that it sets |
| 8001 | the value of a variable. It is unlike @code{setq} in two ways: first, |
| 8002 | it only sets the value of the variable if the variable does not already |
| 8003 | have a value. If the variable already has a value, @code{defvar} does |
| 8004 | not override the existing value. Second, @code{defvar} has a |
| 8005 | documentation string. |
| 8006 | |
| 8007 | (Another special form, @code{defcustom}, is designed for variables |
| 8008 | that people customize. It has more features than @code{defvar}. |
| 8009 | (@xref{defcustom, , Setting Variables with @code{defcustom}}.) |
| 8010 | |
| 8011 | @menu |
| 8012 | * See variable current value:: |
| 8013 | * defvar and asterisk:: An old-time convention. |
| 8014 | @end menu |
| 8015 | |
| 8016 | @node See variable current value, defvar and asterisk, defvar, defvar |
| 8017 | @ifnottex |
| 8018 | @unnumberedsubsec Seeing the Current Value of a Variable |
| 8019 | @end ifnottex |
| 8020 | |
| 8021 | You can see the current value of a variable, any variable, by using |
| 8022 | the @code{describe-variable} function, which is usually invoked by |
| 8023 | typing @kbd{C-h v}. If you type @kbd{C-h v} and then @code{kill-ring} |
| 8024 | (followed by @key{RET}) when prompted, you will see what is in your |
| 8025 | current kill ring---this may be quite a lot! Conversely, if you have |
| 8026 | been doing nothing this Emacs session except read this document, you |
| 8027 | may have nothing in it. Also, you will see the documentation for |
| 8028 | @code{kill-ring}: |
| 8029 | |
| 8030 | @smallexample |
| 8031 | @group |
| 8032 | Documentation: |
| 8033 | List of killed text sequences. |
| 8034 | Since the kill ring is supposed to interact nicely with cut-and-paste |
| 8035 | facilities offered by window systems, use of this variable should |
| 8036 | @end group |
| 8037 | @group |
| 8038 | interact nicely with `interprogram-cut-function' and |
| 8039 | `interprogram-paste-function'. The functions `kill-new', |
| 8040 | `kill-append', and `current-kill' are supposed to implement this |
| 8041 | interaction; you may want to use them instead of manipulating the kill |
| 8042 | ring directly. |
| 8043 | @end group |
| 8044 | @end smallexample |
| 8045 | |
| 8046 | @need 800 |
| 8047 | The kill ring is defined by a @code{defvar} in the following way: |
| 8048 | |
| 8049 | @smallexample |
| 8050 | @group |
| 8051 | (defvar kill-ring nil |
| 8052 | "List of killed text sequences. |
| 8053 | @dots{}") |
| 8054 | @end group |
| 8055 | @end smallexample |
| 8056 | |
| 8057 | @noindent |
| 8058 | In this variable definition, the variable is given an initial value of |
| 8059 | @code{nil}, which makes sense, since if you have saved nothing, you want |
| 8060 | nothing back if you give a @code{yank} command. The documentation |
| 8061 | string is written just like the documentation string of a @code{defun}. |
| 8062 | As with the documentation string of the @code{defun}, the first line of |
| 8063 | the documentation should be a complete sentence, since some commands, |
| 8064 | like @code{apropos}, print only the first line of documentation. |
| 8065 | Succeeding lines should not be indented; otherwise they look odd when |
| 8066 | you use @kbd{C-h v} (@code{describe-variable}). |
| 8067 | |
| 8068 | @node defvar and asterisk, , See variable current value, defvar |
| 8069 | @subsection @code{defvar} and an asterisk |
| 8070 | @findex defvar @r{for a user customizable variable} |
| 8071 | @findex defvar @r{with an asterisk} |
| 8072 | |
| 8073 | In the past, Emacs used the @code{defvar} special form both for |
| 8074 | internal variables that you would not expect a user to change and for |
| 8075 | variables that you do expect a user to change. Although you can still |
| 8076 | use @code{defvar} for user customizable variables, please use |
| 8077 | @code{defcustom} instead, since that special form provides a path into |
| 8078 | the Customization commands. (@xref{defcustom, , Specifying Variables |
| 8079 | using @code{defcustom}}.) |
| 8080 | |
| 8081 | When you specified a variable using the @code{defvar} special form, |
| 8082 | you could distinguish a readily settable variable from others by |
| 8083 | typing an asterisk, @samp{*}, in the first column of its documentation |
| 8084 | string. For example: |
| 8085 | |
| 8086 | @smallexample |
| 8087 | @group |
| 8088 | (defvar shell-command-default-error-buffer nil |
| 8089 | "*Buffer name for `shell-command' @dots{} error output. |
| 8090 | @dots{} ") |
| 8091 | @end group |
| 8092 | @end smallexample |
| 8093 | |
| 8094 | @findex set-variable |
| 8095 | @noindent |
| 8096 | You could (and still can) use the @code{set-variable} command to |
| 8097 | change the value of @code{shell-command-default-error-buffer} |
| 8098 | temporarily. However, options set using @code{set-variable} are set |
| 8099 | only for the duration of your editing session. The new values are not |
| 8100 | saved between sessions. Each time Emacs starts, it reads the original |
| 8101 | value, unless you change the value within your @file{.emacs} file, |
| 8102 | either by setting it manually or by using @code{customize}. |
| 8103 | @xref{Emacs Initialization, , Your @file{.emacs} File}. |
| 8104 | |
| 8105 | For me, the major use of the @code{set-variable} command is to suggest |
| 8106 | variables that I might want to set in my @file{.emacs} file. There |
| 8107 | are now more than 700 such variables --- far too many to remember |
| 8108 | readily. Fortunately, you can press @key{TAB} after calling the |
| 8109 | @code{M-x set-variable} command to see the list of variables. |
| 8110 | (@xref{Examining, , Examining and Setting Variables, emacs, |
| 8111 | The GNU Emacs Manual}.) |
| 8112 | |
| 8113 | @node copy-region-as-kill, cons & search-fwd Review, defvar, Cutting & Storing Text |
| 8114 | @comment node-name, next, previous, up |
| 8115 | @section @code{copy-region-as-kill} |
| 8116 | @findex copy-region-as-kill |
| 8117 | @findex nthcdr |
| 8118 | |
| 8119 | The @code{copy-region-as-kill} function copies a region of text from a |
| 8120 | buffer and (via either @code{kill-append} or @code{kill-new}) saves it |
| 8121 | in the @code{kill-ring}. |
| 8122 | |
| 8123 | If you call @code{copy-region-as-kill} immediately after a |
| 8124 | @code{kill-region} command, Emacs appends the newly copied text to the |
| 8125 | previously copied text. This means that if you yank back the text, you |
| 8126 | get it all, from both this and the previous operation. On the other |
| 8127 | hand, if some other command precedes the @code{copy-region-as-kill}, |
| 8128 | the function copies the text into a separate entry in the kill ring. |
| 8129 | |
| 8130 | @menu |
| 8131 | * Complete copy-region-as-kill:: The complete function definition. |
| 8132 | * copy-region-as-kill body:: The body of @code{copy-region-as-kill}. |
| 8133 | @end menu |
| 8134 | |
| 8135 | @node Complete copy-region-as-kill, copy-region-as-kill body, copy-region-as-kill, copy-region-as-kill |
| 8136 | @ifnottex |
| 8137 | @unnumberedsubsec The complete @code{copy-region-as-kill} function definition |
| 8138 | @end ifnottex |
| 8139 | |
| 8140 | @need 1200 |
| 8141 | Here is the complete text of the version 21 @code{copy-region-as-kill} |
| 8142 | function: |
| 8143 | |
| 8144 | @c !!! for no text properties, use buffer-substring-no-properties |
| 8145 | |
| 8146 | @smallexample |
| 8147 | @group |
| 8148 | (defun copy-region-as-kill (beg end) |
| 8149 | "Save the region as if killed, but don't kill it. |
| 8150 | In Transient Mark mode, deactivate the mark. |
| 8151 | If `interprogram-cut-function' is non-nil, also save |
| 8152 | the text for a window system cut and paste." |
| 8153 | (interactive "r") |
| 8154 | @end group |
| 8155 | @group |
| 8156 | (if (eq last-command 'kill-region) |
| 8157 | (kill-append (buffer-substring beg end) (< end beg)) |
| 8158 | (kill-new (buffer-substring beg end))) |
| 8159 | @end group |
| 8160 | @group |
| 8161 | (if transient-mark-mode |
| 8162 | (setq deactivate-mark t)) |
| 8163 | nil) |
| 8164 | @end group |
| 8165 | @end smallexample |
| 8166 | |
| 8167 | @need 800 |
| 8168 | As usual, this function can be divided into its component parts: |
| 8169 | |
| 8170 | @smallexample |
| 8171 | @group |
| 8172 | (defun copy-region-as-kill (@var{argument-list}) |
| 8173 | "@var{documentation}@dots{}" |
| 8174 | (interactive "r") |
| 8175 | @var{body}@dots{}) |
| 8176 | @end group |
| 8177 | @end smallexample |
| 8178 | |
| 8179 | The arguments are @code{beg} and @code{end} and the function is |
| 8180 | interactive with @code{"r"}, so the two arguments must refer to the |
| 8181 | beginning and end of the region. If you have been reading though this |
| 8182 | document from the beginning, understanding these parts of a function is |
| 8183 | almost becoming routine. |
| 8184 | |
| 8185 | The documentation is somewhat confusing unless you remember that the |
| 8186 | word `kill' has a meaning different from its usual meaning. The |
| 8187 | `Transient Mark' and @code{interprogram-cut-function} comments explain |
| 8188 | certain side-effects. |
| 8189 | |
| 8190 | After you once set a mark, a buffer always contains a region. If you |
| 8191 | wish, you can use Transient Mark mode to highlight the region |
| 8192 | temporarily. (No one wants to highlight the region all the time, so |
| 8193 | Transient Mark mode highlights it only at appropriate times. Many |
| 8194 | people turn off Transient Mark mode, so the region is never |
| 8195 | highlighted.) |
| 8196 | |
| 8197 | Also, a windowing system allows you to copy, cut, and paste among |
| 8198 | different programs. In the X windowing system, for example, the |
| 8199 | @code{interprogram-cut-function} function is @code{x-select-text}, |
| 8200 | which works with the windowing system's equivalent of the Emacs kill |
| 8201 | ring. |
| 8202 | |
| 8203 | The body of the @code{copy-region-as-kill} function starts with an |
| 8204 | @code{if} clause. What this clause does is distinguish between two |
| 8205 | different situations: whether or not this command is executed |
| 8206 | immediately after a previous @code{kill-region} command. In the first |
| 8207 | case, the new region is appended to the previously copied text. |
| 8208 | Otherwise, it is inserted into the beginning of the kill ring as a |
| 8209 | separate piece of text from the previous piece. |
| 8210 | |
| 8211 | The last two lines of the function prevent the region from lighting up |
| 8212 | if Transient Mark mode is turned on. |
| 8213 | |
| 8214 | The body of @code{copy-region-as-kill} merits discussion in detail. |
| 8215 | |
| 8216 | @node copy-region-as-kill body, , Complete copy-region-as-kill, copy-region-as-kill |
| 8217 | @comment node-name, next, previous, up |
| 8218 | @subsection The Body of @code{copy-region-as-kill} |
| 8219 | |
| 8220 | The @code{copy-region-as-kill} function works in much the same way as |
| 8221 | the @code{kill-region} function (@pxref{kill-region, |
| 8222 | ,@code{kill-region}}). Both are written so that two or more kills in |
| 8223 | a row combine their text into a single entry. If you yank back the |
| 8224 | text from the kill ring, you get it all in one piece. Moreover, kills |
| 8225 | that kill forward from the current position of the cursor are added to |
| 8226 | the end of the previously copied text and commands that copy text |
| 8227 | backwards add it to the beginning of the previously copied text. This |
| 8228 | way, the words in the text stay in the proper order. |
| 8229 | |
| 8230 | Like @code{kill-region}, the @code{copy-region-as-kill} function makes |
| 8231 | use of the @code{last-command} variable that keeps track of the |
| 8232 | previous Emacs command. |
| 8233 | |
| 8234 | @menu |
| 8235 | * last-command & this-command:: |
| 8236 | * kill-append function:: |
| 8237 | * kill-new function:: |
| 8238 | @end menu |
| 8239 | |
| 8240 | @node last-command & this-command, kill-append function, copy-region-as-kill body, copy-region-as-kill body |
| 8241 | @ifnottex |
| 8242 | @unnumberedsubsubsec @code{last-command} and @code{this-command} |
| 8243 | @end ifnottex |
| 8244 | |
| 8245 | Normally, whenever a function is executed, Emacs sets the value of |
| 8246 | @code{this-command} to the function being executed (which in this case |
| 8247 | would be @code{copy-region-as-kill}). At the same time, Emacs sets |
| 8248 | the value of @code{last-command} to the previous value of |
| 8249 | @code{this-command}. |
| 8250 | |
| 8251 | In the first part of the body of the @code{copy-region-as-kill} |
| 8252 | function, an @code{if} expression determines whether the value of |
| 8253 | @code{last-command} is @code{kill-region}. If so, the then-part of |
| 8254 | the @code{if} expression is evaluated; it uses the @code{kill-append} |
| 8255 | function to concatenate the text copied at this call to the function |
| 8256 | with the text already in the first element (the @sc{car}) of the kill |
| 8257 | ring. On the other hand, if the value of @code{last-command} is not |
| 8258 | @code{kill-region}, then the @code{copy-region-as-kill} function |
| 8259 | attaches a new element to the kill ring using the @code{kill-new} |
| 8260 | function. |
| 8261 | |
| 8262 | @need 1250 |
| 8263 | The @code{if} expression reads as follows; it uses @code{eq}, which is |
| 8264 | a function we have not yet seen: |
| 8265 | |
| 8266 | @smallexample |
| 8267 | @group |
| 8268 | (if (eq last-command 'kill-region) |
| 8269 | ;; @r{then-part} |
| 8270 | (kill-append (buffer-substring beg end) (< end beg)) |
| 8271 | ;; @r{else-part} |
| 8272 | (kill-new (buffer-substring beg end))) |
| 8273 | @end group |
| 8274 | @end smallexample |
| 8275 | |
| 8276 | @findex eq @r{(example of use)} |
| 8277 | @noindent |
| 8278 | The @code{eq} function tests whether its first argument is the same Lisp |
| 8279 | object as its second argument. The @code{eq} function is similar to the |
| 8280 | @code{equal} function in that it is used to test for equality, but |
| 8281 | differs in that it determines whether two representations are actually |
| 8282 | the same object inside the computer, but with different names. |
| 8283 | @code{equal} determines whether the structure and contents of two |
| 8284 | expressions are the same. |
| 8285 | |
| 8286 | If the previous command was @code{kill-region}, then the Emacs Lisp |
| 8287 | interpreter calls the @code{kill-append} function |
| 8288 | |
| 8289 | @node kill-append function, kill-new function, last-command & this-command, copy-region-as-kill body |
| 8290 | @unnumberedsubsubsec The @code{kill-append} function |
| 8291 | @findex kill-append |
| 8292 | |
| 8293 | @need 800 |
| 8294 | The @code{kill-append} function looks like this: |
| 8295 | |
| 8296 | @smallexample |
| 8297 | @group |
| 8298 | (defun kill-append (string before-p) |
| 8299 | "Append STRING to the end of the latest kill in the kill ring. |
| 8300 | If BEFORE-P is non-nil, prepend STRING to the kill. |
| 8301 | If `interprogram-cut-function' is set, pass the resulting kill to |
| 8302 | it." |
| 8303 | (kill-new (if before-p |
| 8304 | (concat string (car kill-ring)) |
| 8305 | (concat (car kill-ring) string)) |
| 8306 | t)) |
| 8307 | @end group |
| 8308 | @end smallexample |
| 8309 | |
| 8310 | @noindent |
| 8311 | The @code{kill-append} function is fairly straightforward. It uses |
| 8312 | the @code{kill-new} function, which we will discuss in more detail in |
| 8313 | a moment. |
| 8314 | |
| 8315 | First, let us look at the conditional that is one of the two arguments |
| 8316 | to @code{kill-new}. It uses @code{concat} to concatenate the new text |
| 8317 | to the @sc{car} of the kill ring. Whether it prepends or appends the |
| 8318 | text depends on the results of an @code{if} expression: |
| 8319 | |
| 8320 | @smallexample |
| 8321 | @group |
| 8322 | (if before-p ; @r{if-part} |
| 8323 | (concat string (car kill-ring)) ; @r{then-part} |
| 8324 | (concat (car kill-ring) string)) ; @r{else-part} |
| 8325 | @end group |
| 8326 | @end smallexample |
| 8327 | |
| 8328 | @noindent |
| 8329 | If the region being killed is before the region that was killed in the |
| 8330 | last command, then it should be prepended before the material that was |
| 8331 | saved in the previous kill; and conversely, if the killed text follows |
| 8332 | what was just killed, it should be appended after the previous text. |
| 8333 | The @code{if} expression depends on the predicate @code{before-p} to |
| 8334 | decide whether the newly saved text should be put before or after the |
| 8335 | previously saved text. |
| 8336 | |
| 8337 | The symbol @code{before-p} is the name of one of the arguments to |
| 8338 | @code{kill-append}. When the @code{kill-append} function is |
| 8339 | evaluated, it is bound to the value returned by evaluating the actual |
| 8340 | argument. In this case, this is the expression @code{(< end beg)}. |
| 8341 | This expression does not directly determine whether the killed text in |
| 8342 | this command is located before or after the kill text of the last |
| 8343 | command; what it does is determine whether the value of the variable |
| 8344 | @code{end} is less than the value of the variable @code{beg}. If it |
| 8345 | is, it means that the user is most likely heading towards the |
| 8346 | beginning of the buffer. Also, the result of evaluating the predicate |
| 8347 | expression, @code{(< end beg)}, will be true and the text will be |
| 8348 | prepended before the previous text. On the other hand, if the value of |
| 8349 | the variable @code{end} is greater than the value of the variable |
| 8350 | @code{beg}, the text will be appended after the previous text. |
| 8351 | |
| 8352 | @need 800 |
| 8353 | When the newly saved text will be prepended, then the string with the new |
| 8354 | text will be concatenated before the old text: |
| 8355 | |
| 8356 | @smallexample |
| 8357 | (concat string (car kill-ring)) |
| 8358 | @end smallexample |
| 8359 | |
| 8360 | @need 1200 |
| 8361 | @noindent |
| 8362 | But if the text will be appended, it will be concatenated |
| 8363 | after the old text: |
| 8364 | |
| 8365 | @smallexample |
| 8366 | (concat (car kill-ring) string)) |
| 8367 | @end smallexample |
| 8368 | |
| 8369 | To understand how this works, we first need to review the |
| 8370 | @code{concat} function. The @code{concat} function links together or |
| 8371 | unites two strings of text. The result is a string. For example: |
| 8372 | |
| 8373 | @smallexample |
| 8374 | @group |
| 8375 | (concat "abc" "def") |
| 8376 | @result{} "abcdef" |
| 8377 | @end group |
| 8378 | |
| 8379 | @group |
| 8380 | (concat "new " |
| 8381 | (car '("first element" "second element"))) |
| 8382 | @result{} "new first element" |
| 8383 | |
| 8384 | (concat (car |
| 8385 | '("first element" "second element")) " modified") |
| 8386 | @result{} "first element modified" |
| 8387 | @end group |
| 8388 | @end smallexample |
| 8389 | |
| 8390 | We can now make sense of @code{kill-append}: it modifies the contents |
| 8391 | of the kill ring. The kill ring is a list, each element of which is |
| 8392 | saved text. The @code{kill-append} function uses the @code{kill-new} |
| 8393 | function which in turn uses the @code{setcar} function. |
| 8394 | |
| 8395 | @node kill-new function, , kill-append function, copy-region-as-kill body |
| 8396 | @unnumberedsubsubsec The @code{kill-new} function |
| 8397 | @findex kill-new |
| 8398 | |
| 8399 | @need 1200 |
| 8400 | The @code{kill-new} function looks like this: |
| 8401 | |
| 8402 | @smallexample |
| 8403 | @group |
| 8404 | (defun kill-new (string &optional replace) |
| 8405 | "Make STRING the latest kill in the kill ring. |
| 8406 | Set the kill-ring-yank pointer to point to it. |
| 8407 | If `interprogram-cut-function' is non-nil, apply it to STRING. |
| 8408 | Optional second argument REPLACE non-nil means that STRING will replace |
| 8409 | the front of the kill ring, rather than being added to the list." |
| 8410 | @end group |
| 8411 | @group |
| 8412 | (and (fboundp 'menu-bar-update-yank-menu) |
| 8413 | (menu-bar-update-yank-menu string (and replace (car kill-ring)))) |
| 8414 | @end group |
| 8415 | @group |
| 8416 | (if (and replace kill-ring) |
| 8417 | (setcar kill-ring string) |
| 8418 | (setq kill-ring (cons string kill-ring)) |
| 8419 | (if (> (length kill-ring) kill-ring-max) |
| 8420 | (setcdr (nthcdr (1- kill-ring-max) kill-ring) nil))) |
| 8421 | @end group |
| 8422 | @group |
| 8423 | (setq kill-ring-yank-pointer kill-ring) |
| 8424 | (if interprogram-cut-function |
| 8425 | (funcall interprogram-cut-function string (not replace)))) |
| 8426 | @end group |
| 8427 | @end smallexample |
| 8428 | |
| 8429 | As usual, we can look at this function in parts. |
| 8430 | |
| 8431 | @need 1200 |
| 8432 | The first line of the documentation makes sense: |
| 8433 | |
| 8434 | @smallexample |
| 8435 | Make STRING the latest kill in the kill ring. |
| 8436 | @end smallexample |
| 8437 | |
| 8438 | @noindent |
| 8439 | Let's skip over the rest of the documentation for the moment. |
| 8440 | |
| 8441 | Also, let's skip over the first two lines of code, those involving |
| 8442 | @code{menu-bar-update-yank-menu}. We will explain them below. |
| 8443 | |
| 8444 | @need 1200 |
| 8445 | The critical lines are these: |
| 8446 | |
| 8447 | @smallexample |
| 8448 | @group |
| 8449 | (if (and replace kill-ring) |
| 8450 | ;; @r{then} |
| 8451 | (setcar kill-ring string) |
| 8452 | @end group |
| 8453 | @group |
| 8454 | ;; @r{else} |
| 8455 | (setq kill-ring (cons string kill-ring)) |
| 8456 | (if (> (length kill-ring) kill-ring-max) |
| 8457 | ;; @r{avoid overly long kill ring} |
| 8458 | (setcdr (nthcdr (1- kill-ring-max) kill-ring) nil))) |
| 8459 | @end group |
| 8460 | @group |
| 8461 | (setq kill-ring-yank-pointer kill-ring) |
| 8462 | (if interprogram-cut-function |
| 8463 | (funcall interprogram-cut-function string (not replace)))) |
| 8464 | @end group |
| 8465 | @end smallexample |
| 8466 | |
| 8467 | The conditional test is @w{@code{(and replace kill-ring)}}. |
| 8468 | This will be true when two conditions are met: the kill ring has |
| 8469 | something in it, and the @code{replace} variable is true. |
| 8470 | |
| 8471 | @need 1250 |
| 8472 | The @code{kill-append} function sets @code{replace} to be true; then, |
| 8473 | when the kill ring has at least one item in it, the @code{setcar} |
| 8474 | expression is executed: |
| 8475 | |
| 8476 | @smallexample |
| 8477 | (setcar kill-ring string) |
| 8478 | @end smallexample |
| 8479 | |
| 8480 | The @code{setcar} function actually changes the first element of the |
| 8481 | @code{kill-ring} list to the value of @code{string}. It replaces the |
| 8482 | first element. |
| 8483 | |
| 8484 | On the other hand, if the kill ring is empty, or replace is false, the |
| 8485 | else-part of the condition is executed: |
| 8486 | |
| 8487 | @smallexample |
| 8488 | @group |
| 8489 | (setq kill-ring (cons string kill-ring)) |
| 8490 | (if (> (length kill-ring) kill-ring-max) |
| 8491 | (setcdr (nthcdr (1- kill-ring-max) kill-ring) nil)) |
| 8492 | @end group |
| 8493 | @end smallexample |
| 8494 | |
| 8495 | @noindent |
| 8496 | This expression first constructs a new version of the kill ring by |
| 8497 | prepending @code{string} to the existing kill ring as a new element. |
| 8498 | Then it executes a second @code{if} clause. This second @code{if} |
| 8499 | clause keeps the kill ring from growing too long. |
| 8500 | |
| 8501 | Let's look at these two expressions in order. |
| 8502 | |
| 8503 | The @code{setq} line of the else-part sets the new value of the kill |
| 8504 | ring to what results from adding the string being killed to the old kill |
| 8505 | ring. |
| 8506 | |
| 8507 | @need 800 |
| 8508 | We can see how this works with an example: |
| 8509 | |
| 8510 | @smallexample |
| 8511 | (setq example-list '("here is a clause" "another clause")) |
| 8512 | @end smallexample |
| 8513 | |
| 8514 | @need 1200 |
| 8515 | @noindent |
| 8516 | After evaluating this expression with @kbd{C-x C-e}, you can evaluate |
| 8517 | @code{example-list} and see what it returns: |
| 8518 | |
| 8519 | @smallexample |
| 8520 | @group |
| 8521 | example-list |
| 8522 | @result{} ("here is a clause" "another clause") |
| 8523 | @end group |
| 8524 | @end smallexample |
| 8525 | |
| 8526 | @need 1200 |
| 8527 | @noindent |
| 8528 | Now, we can add a new element on to this list by evaluating the |
| 8529 | following expression: |
| 8530 | @findex cons, @r{example} |
| 8531 | |
| 8532 | @smallexample |
| 8533 | (setq example-list (cons "a third clause" example-list)) |
| 8534 | @end smallexample |
| 8535 | |
| 8536 | @need 800 |
| 8537 | @noindent |
| 8538 | When we evaluate @code{example-list}, we find its value is: |
| 8539 | |
| 8540 | @smallexample |
| 8541 | @group |
| 8542 | example-list |
| 8543 | @result{} ("a third clause" "here is a clause" "another clause") |
| 8544 | @end group |
| 8545 | @end smallexample |
| 8546 | |
| 8547 | @noindent |
| 8548 | Thus, the third clause was added to the list by @code{cons}. |
| 8549 | |
| 8550 | @need 1200 |
| 8551 | This is exactly similar to what the @code{setq} and @code{cons} do in |
| 8552 | the function. Here is the line again: |
| 8553 | |
| 8554 | @smallexample |
| 8555 | (setq kill-ring (cons string kill-ring)) |
| 8556 | @end smallexample |
| 8557 | |
| 8558 | @need 1200 |
| 8559 | Now for the second part of the @code{if} clause. This expression |
| 8560 | keeps the kill ring from growing too long. It looks like this: |
| 8561 | |
| 8562 | @smallexample |
| 8563 | @group |
| 8564 | (if (> (length kill-ring) kill-ring-max) |
| 8565 | (setcdr (nthcdr (1- kill-ring-max) kill-ring) nil)) |
| 8566 | @end group |
| 8567 | @end smallexample |
| 8568 | |
| 8569 | The code checks whether the length of the kill ring is greater than |
| 8570 | the maximum permitted length. This is the value of |
| 8571 | @code{kill-ring-max} (which is 60, by default). If the length of the |
| 8572 | kill ring is too long, then this code sets the last element of the |
| 8573 | kill ring to @code{nil}. It does this by using two functions, |
| 8574 | @code{nthcdr} and @code{setcdr}. |
| 8575 | |
| 8576 | We looked at @code{setcdr} earlier (@pxref{setcdr, , @code{setcdr}}). |
| 8577 | It sets the @sc{cdr} of a list, just as @code{setcar} sets the |
| 8578 | @sc{car} of a list. In this case, however, @code{setcdr} will not be |
| 8579 | setting the @sc{cdr} of the whole kill ring; the @code{nthcdr} |
| 8580 | function is used to cause it to set the @sc{cdr} of the next to last |
| 8581 | element of the kill ring---this means that since the @sc{cdr} of the |
| 8582 | next to last element is the last element of the kill ring, it will set |
| 8583 | the last element of the kill ring. |
| 8584 | |
| 8585 | @findex nthcdr, @r{example} |
| 8586 | The @code{nthcdr} function works by repeatedly taking the @sc{cdr} of a |
| 8587 | list---it takes the @sc{cdr} of the @sc{cdr} of the @sc{cdr} |
| 8588 | @dots{} It does this @var{N} times and returns the results. |
| 8589 | |
| 8590 | @findex setcdr, @r{example} |
| 8591 | Thus, if we had a four element list that was supposed to be three |
| 8592 | elements long, we could set the @sc{cdr} of the next to last element |
| 8593 | to @code{nil}, and thereby shorten the list. (If you sent the last |
| 8594 | element to some other value than @code{nil}, which you could do, then |
| 8595 | you would not have shortened the list.) |
| 8596 | |
| 8597 | You can see shortening by evaluating the following three expressions |
| 8598 | in turn. First set the value of @code{trees} to @code{(maple oak pine |
| 8599 | birch)}, then set the @sc{cdr} of its second @sc{cdr} to @code{nil} |
| 8600 | and then find the value of @code{trees}: |
| 8601 | |
| 8602 | @smallexample |
| 8603 | @group |
| 8604 | (setq trees '(maple oak pine birch)) |
| 8605 | @result{} (maple oak pine birch) |
| 8606 | @end group |
| 8607 | |
| 8608 | @group |
| 8609 | (setcdr (nthcdr 2 trees) nil) |
| 8610 | @result{} nil |
| 8611 | |
| 8612 | trees |
| 8613 | @result{} (maple oak pine) |
| 8614 | @end group |
| 8615 | @end smallexample |
| 8616 | |
| 8617 | @noindent |
| 8618 | (The value returned by the @code{setcdr} expression is @code{nil} since |
| 8619 | that is what the @sc{cdr} is set to.) |
| 8620 | |
| 8621 | To repeat, in @code{kill-new}, the @code{nthcdr} function takes the |
| 8622 | @sc{cdr} a number of times that is one less than the maximum permitted |
| 8623 | size of the kill ring and sets the @sc{cdr} of that element (which |
| 8624 | will be the rest of the elements in the kill ring) to @code{nil}. |
| 8625 | This prevents the kill ring from growing too long. |
| 8626 | |
| 8627 | @need 800 |
| 8628 | The next to last expression in the @code{kill-new} function is |
| 8629 | |
| 8630 | @smallexample |
| 8631 | (setq kill-ring-yank-pointer kill-ring) |
| 8632 | @end smallexample |
| 8633 | |
| 8634 | The @code{kill-ring-yank-pointer} is a global variable that is set to be |
| 8635 | the @code{kill-ring}. |
| 8636 | |
| 8637 | Even though the @code{kill-ring-yank-pointer} is called a |
| 8638 | @samp{pointer}, it is a variable just like the kill ring. However, the |
| 8639 | name has been chosen to help humans understand how the variable is used. |
| 8640 | The variable is used in functions such as @code{yank} and |
| 8641 | @code{yank-pop} (@pxref{Yanking, , Yanking Text Back}). |
| 8642 | |
| 8643 | @need 1200 |
| 8644 | Now, to return to the first two lines in the body of the function: |
| 8645 | |
| 8646 | @smallexample |
| 8647 | @group |
| 8648 | (and (fboundp 'menu-bar-update-yank-menu) |
| 8649 | (menu-bar-update-yank-menu string (and replace (car kill-ring)))) |
| 8650 | @end group |
| 8651 | @end smallexample |
| 8652 | |
| 8653 | @noindent |
| 8654 | This is an expression whose first element is the function @code{and}. |
| 8655 | |
| 8656 | @findex and, @r{introduced} |
| 8657 | The @code{and} special form evaluates each of its arguments until one of |
| 8658 | the arguments returns a value of @code{nil}, in which case the |
| 8659 | @code{and} expression returns @code{nil}; however, if none of the |
| 8660 | arguments returns a value of @code{nil}, the value resulting from |
| 8661 | evaluating the last argument is returned. (Since such a value is not |
| 8662 | @code{nil}, it is considered true in Emacs Lisp.) In other words, an |
| 8663 | @code{and} expression returns a true value only if all its arguments |
| 8664 | are true. |
| 8665 | @findex and |
| 8666 | |
| 8667 | In this case, the expression tests first to see whether |
| 8668 | @code{menu-bar-update-yank-menu} exists as a function, and if so, |
| 8669 | calls it. The @code{fboundp} function returns true if the symbol it |
| 8670 | is testing has a function definition that `is not void'. If the |
| 8671 | symbol's function definition were void, we would receive an error |
| 8672 | message, as we did when we created errors intentionally (@pxref{Making |
| 8673 | Errors, , Generate an Error Message}). |
| 8674 | |
| 8675 | @need 1200 |
| 8676 | Essentially, the @code{and} is an @code{if} expression that reads like |
| 8677 | this: |
| 8678 | |
| 8679 | @smallexample |
| 8680 | @group |
| 8681 | if @var{the-menu-bar-function-exists} |
| 8682 | then @var{execute-it} |
| 8683 | @end group |
| 8684 | @end smallexample |
| 8685 | |
| 8686 | @code{menu-bar-update-yank-menu} is one of the functions that make it |
| 8687 | possible to use the `Select and Paste' menu in the Edit item of a menu |
| 8688 | bar; using a mouse, you can look at the various pieces of text you |
| 8689 | have saved and select one piece to paste. |
| 8690 | |
| 8691 | Finally, the last expression in the @code{kill-new} function adds the |
| 8692 | newly copied string to whatever facility exists for copying and |
| 8693 | pasting among different programs running in a windowing system. In |
| 8694 | the X Windowing system, for example, the @code{x-select-text} function |
| 8695 | takes the string and stores it in memory operated by X. You can paste |
| 8696 | the string in another program, such as an Xterm. |
| 8697 | |
| 8698 | @need 1200 |
| 8699 | The expression looks like this: |
| 8700 | |
| 8701 | @smallexample |
| 8702 | @group |
| 8703 | (if interprogram-cut-function |
| 8704 | (funcall interprogram-cut-function string (not replace)))) |
| 8705 | @end group |
| 8706 | @end smallexample |
| 8707 | |
| 8708 | If an @code{interprogram-cut-function} exists, then Emacs executes |
| 8709 | @code{funcall}, which in turn calls its first argument as a function |
| 8710 | and passes the remaining arguments to it. (Incidentally, as far as I |
| 8711 | can see, this @code{if} expression could be replaced by an @code{and} |
| 8712 | expression similar to the one in the first part of the function.) |
| 8713 | |
| 8714 | We are not going to discuss windowing systems and other programs |
| 8715 | further, but merely note that this is a mechanism that enables GNU |
| 8716 | Emacs to work easily and well with other programs. |
| 8717 | |
| 8718 | This code for placing text in the kill ring, either concatenated with |
| 8719 | an existing element or as a new element, leads us to the code for |
| 8720 | bringing back text that has been cut out of the buffer---the yank |
| 8721 | commands. However, before discussing the yank commands, it is better |
| 8722 | to learn how lists are implemented in a computer. This will make |
| 8723 | clear such mysteries as the use of the term `pointer'. |
| 8724 | |
| 8725 | @need 1250 |
| 8726 | @node cons & search-fwd Review, search Exercises, copy-region-as-kill, Cutting & Storing Text |
| 8727 | @comment node-name, next, previous, up |
| 8728 | @section Review |
| 8729 | |
| 8730 | Here is a brief summary of some recently introduced functions. |
| 8731 | |
| 8732 | @table @code |
| 8733 | @item car |
| 8734 | @itemx cdr |
| 8735 | @code{car} returns the first element of a list; @code{cdr} returns the |
| 8736 | second and subsequent elements of a list. |
| 8737 | |
| 8738 | @need 1250 |
| 8739 | For example: |
| 8740 | |
| 8741 | @smallexample |
| 8742 | @group |
| 8743 | (car '(1 2 3 4 5 6 7)) |
| 8744 | @result{} 1 |
| 8745 | (cdr '(1 2 3 4 5 6 7)) |
| 8746 | @result{} (2 3 4 5 6 7) |
| 8747 | @end group |
| 8748 | @end smallexample |
| 8749 | |
| 8750 | @item cons |
| 8751 | @code{cons} constructs a list by prepending its first argument to its |
| 8752 | second argument. |
| 8753 | |
| 8754 | @need 1250 |
| 8755 | For example: |
| 8756 | |
| 8757 | @smallexample |
| 8758 | @group |
| 8759 | (cons 1 '(2 3 4)) |
| 8760 | @result{} (1 2 3 4) |
| 8761 | @end group |
| 8762 | @end smallexample |
| 8763 | |
| 8764 | @item nthcdr |
| 8765 | Return the result of taking @sc{cdr} `n' times on a list. |
| 8766 | @iftex |
| 8767 | The |
| 8768 | @tex |
| 8769 | $n^{th}$ |
| 8770 | @end tex |
| 8771 | @code{cdr}. |
| 8772 | @end iftex |
| 8773 | The `rest of the rest', as it were. |
| 8774 | |
| 8775 | @need 1250 |
| 8776 | For example: |
| 8777 | |
| 8778 | @smallexample |
| 8779 | @group |
| 8780 | (nthcdr 3 '(1 2 3 4 5 6 7)) |
| 8781 | @result{} (4 5 6 7) |
| 8782 | @end group |
| 8783 | @end smallexample |
| 8784 | |
| 8785 | @item setcar |
| 8786 | @itemx setcdr |
| 8787 | @code{setcar} changes the first element of a list; @code{setcdr} |
| 8788 | changes the second and subsequent elements of a list. |
| 8789 | |
| 8790 | @need 1250 |
| 8791 | For example: |
| 8792 | |
| 8793 | @smallexample |
| 8794 | @group |
| 8795 | (setq triple '(1 2 3)) |
| 8796 | |
| 8797 | (setcar triple '37) |
| 8798 | |
| 8799 | triple |
| 8800 | @result{} (37 2 3) |
| 8801 | |
| 8802 | (setcdr triple '("foo" "bar")) |
| 8803 | |
| 8804 | triple |
| 8805 | @result{} (37 "foo" "bar") |
| 8806 | @end group |
| 8807 | @end smallexample |
| 8808 | |
| 8809 | @item progn |
| 8810 | Evaluate each argument in sequence and then return the value of the |
| 8811 | last. |
| 8812 | |
| 8813 | @need 1250 |
| 8814 | For example: |
| 8815 | |
| 8816 | @smallexample |
| 8817 | @group |
| 8818 | (progn 1 2 3 4) |
| 8819 | @result{} 4 |
| 8820 | @end group |
| 8821 | @end smallexample |
| 8822 | |
| 8823 | @item save-restriction |
| 8824 | Record whatever narrowing is in effect in the current buffer, if any, |
| 8825 | and restore that narrowing after evaluating the arguments. |
| 8826 | |
| 8827 | @item search-forward |
| 8828 | Search for a string, and if the string is found, move point. |
| 8829 | |
| 8830 | @need 1250 |
| 8831 | @noindent |
| 8832 | Takes four arguments: |
| 8833 | |
| 8834 | @enumerate |
| 8835 | @item |
| 8836 | The string to search for. |
| 8837 | |
| 8838 | @item |
| 8839 | Optionally, the limit of the search. |
| 8840 | |
| 8841 | @item |
| 8842 | Optionally, what to do if the search fails, return @code{nil} or an |
| 8843 | error message. |
| 8844 | |
| 8845 | @item |
| 8846 | Optionally, how many times to repeat the search; if negative, the |
| 8847 | search goes backwards. |
| 8848 | @end enumerate |
| 8849 | |
| 8850 | @item kill-region |
| 8851 | @itemx delete-and-extract-region |
| 8852 | @itemx copy-region-as-kill |
| 8853 | |
| 8854 | @code{kill-region} cuts the text between point and mark from the |
| 8855 | buffer and stores that text in the kill ring, so you can get it back |
| 8856 | by yanking. |
| 8857 | |
| 8858 | @code{delete-and-extract-region} removes the text between point and |
| 8859 | mark from the buffer and throws it away. You cannot get it back. |
| 8860 | |
| 8861 | @code{copy-region-as-kill} copies the text between point and mark into |
| 8862 | the kill ring, from which you can get it by yanking. The function |
| 8863 | does not cut or remove the text from the buffer. |
| 8864 | @end table |
| 8865 | |
| 8866 | @need 1500 |
| 8867 | @node search Exercises, , cons & search-fwd Review, Cutting & Storing Text |
| 8868 | @section Searching Exercises |
| 8869 | |
| 8870 | @itemize @bullet |
| 8871 | @item |
| 8872 | Write an interactive function that searches for a string. If the |
| 8873 | search finds the string, leave point after it and display a message |
| 8874 | that says ``Found!''. (Do not use @code{search-forward} for the name |
| 8875 | of this function; if you do, you will overwrite the existing version of |
| 8876 | @code{search-forward} that comes with Emacs. Use a name such as |
| 8877 | @code{test-search} instead.) |
| 8878 | |
| 8879 | @item |
| 8880 | Write a function that prints the third element of the kill ring in the |
| 8881 | echo area, if any; if the kill ring does not contain a third element, |
| 8882 | print an appropriate message. |
| 8883 | @end itemize |
| 8884 | |
| 8885 | @node List Implementation, Yanking, Cutting & Storing Text, Top |
| 8886 | @comment node-name, next, previous, up |
| 8887 | @chapter How Lists are Implemented |
| 8888 | @cindex Lists in a computer |
| 8889 | |
| 8890 | In Lisp, atoms are recorded in a straightforward fashion; if the |
| 8891 | implementation is not straightforward in practice, it is, nonetheless, |
| 8892 | straightforward in theory. The atom @samp{rose}, for example, is |
| 8893 | recorded as the four contiguous letters @samp{r}, @samp{o}, @samp{s}, |
| 8894 | @samp{e}. A list, on the other hand, is kept differently. The mechanism |
| 8895 | is equally simple, but it takes a moment to get used to the idea. A |
| 8896 | list is kept using a series of pairs of pointers. In the series, the |
| 8897 | first pointer in each pair points to an atom or to another list, and the |
| 8898 | second pointer in each pair points to the next pair, or to the symbol |
| 8899 | @code{nil}, which marks the end of the list. |
| 8900 | |
| 8901 | A pointer itself is quite simply the electronic address of what is |
| 8902 | pointed to. Hence, a list is kept as a series of electronic addresses. |
| 8903 | |
| 8904 | @menu |
| 8905 | * Lists diagrammed:: |
| 8906 | * Symbols as Chest:: Exploring a powerful metaphor. |
| 8907 | * List Exercise:: |
| 8908 | @end menu |
| 8909 | |
| 8910 | @node Lists diagrammed, Symbols as Chest, List Implementation, List Implementation |
| 8911 | @ifnottex |
| 8912 | @unnumberedsec Lists diagrammed |
| 8913 | @end ifnottex |
| 8914 | |
| 8915 | For example, the list @code{(rose violet buttercup)} has three elements, |
| 8916 | @samp{rose}, @samp{violet}, and @samp{buttercup}. In the computer, the |
| 8917 | electronic address of @samp{rose} is recorded in a segment of computer |
| 8918 | memory along with the address that gives the electronic address of where |
| 8919 | the atom @samp{violet} is located; and that address (the one that tells |
| 8920 | where @samp{violet} is located) is kept along with an address that tells |
| 8921 | where the address for the atom @samp{buttercup} is located. |
| 8922 | |
| 8923 | @need 1200 |
| 8924 | This sounds more complicated than it is and is easier seen in a diagram: |
| 8925 | |
| 8926 | @c clear print-postscript-figures |
| 8927 | @c !!! cons-cell-diagram #1 |
| 8928 | @ifnottex |
| 8929 | @smallexample |
| 8930 | @group |
| 8931 | ___ ___ ___ ___ ___ ___ |
| 8932 | |___|___|--> |___|___|--> |___|___|--> nil |
| 8933 | | | | |
| 8934 | | | | |
| 8935 | --> rose --> violet --> buttercup |
| 8936 | @end group |
| 8937 | @end smallexample |
| 8938 | @end ifnottex |
| 8939 | @ifset print-postscript-figures |
| 8940 | @sp 1 |
| 8941 | @tex |
| 8942 | @image{cons-1} |
| 8943 | %%%% old method of including an image |
| 8944 | % \input /usr/local/lib/tex/inputs/psfig.tex |
| 8945 | % \centerline{\psfig{figure=/usr/local/lib/emacs/man/cons-1.eps}} |
| 8946 | % \catcode`\@=0 % |
| 8947 | @end tex |
| 8948 | @sp 1 |
| 8949 | @end ifset |
| 8950 | @ifclear print-postscript-figures |
| 8951 | @iftex |
| 8952 | @smallexample |
| 8953 | @group |
| 8954 | ___ ___ ___ ___ ___ ___ |
| 8955 | |___|___|--> |___|___|--> |___|___|--> nil |
| 8956 | | | | |
| 8957 | | | | |
| 8958 | --> rose --> violet --> buttercup |
| 8959 | @end group |
| 8960 | @end smallexample |
| 8961 | @end iftex |
| 8962 | @end ifclear |
| 8963 | |
| 8964 | @noindent |
| 8965 | In the diagram, each box represents a word of computer memory that |
| 8966 | holds a Lisp object, usually in the form of a memory address. The boxes, |
| 8967 | i.e.@: the addresses, are in pairs. Each arrow points to what the address |
| 8968 | is the address of, either an atom or another pair of addresses. The |
| 8969 | first box is the electronic address of @samp{rose} and the arrow points |
| 8970 | to @samp{rose}; the second box is the address of the next pair of boxes, |
| 8971 | the first part of which is the address of @samp{violet} and the second |
| 8972 | part of which is the address of the next pair. The very last box |
| 8973 | points to the symbol @code{nil}, which marks the end of the list. |
| 8974 | |
| 8975 | @need 1200 |
| 8976 | When a variable is set to a list with a function such as @code{setq}, |
| 8977 | it stores the address of the first box in the variable. Thus, |
| 8978 | evaluation of the expression |
| 8979 | |
| 8980 | @smallexample |
| 8981 | (setq bouquet '(rose violet buttercup)) |
| 8982 | @end smallexample |
| 8983 | |
| 8984 | @need 1250 |
| 8985 | @noindent |
| 8986 | creates a situation like this: |
| 8987 | |
| 8988 | @c cons-cell-diagram #2 |
| 8989 | @ifnottex |
| 8990 | @smallexample |
| 8991 | @group |
| 8992 | bouquet |
| 8993 | | |
| 8994 | | ___ ___ ___ ___ ___ ___ |
| 8995 | --> |___|___|--> |___|___|--> |___|___|--> nil |
| 8996 | | | | |
| 8997 | | | | |
| 8998 | --> rose --> violet --> buttercup |
| 8999 | @end group |
| 9000 | @end smallexample |
| 9001 | @end ifnottex |
| 9002 | @ifset print-postscript-figures |
| 9003 | @sp 1 |
| 9004 | @tex |
| 9005 | @image{cons-2} |
| 9006 | %%%% old method of including an image |
| 9007 | % \input /usr/local/lib/tex/inputs/psfig.tex |
| 9008 | % \centerline{\psfig{figure=/usr/local/lib/emacs/man/cons-2.eps}} |
| 9009 | % \catcode`\@=0 % |
| 9010 | @end tex |
| 9011 | @sp 1 |
| 9012 | @end ifset |
| 9013 | @ifclear print-postscript-figures |
| 9014 | @iftex |
| 9015 | @smallexample |
| 9016 | @group |
| 9017 | bouquet |
| 9018 | | |
| 9019 | | ___ ___ ___ ___ ___ ___ |
| 9020 | --> |___|___|--> |___|___|--> |___|___|--> nil |
| 9021 | | | | |
| 9022 | | | | |
| 9023 | --> rose --> violet --> buttercup |
| 9024 | @end group |
| 9025 | @end smallexample |
| 9026 | @end iftex |
| 9027 | @end ifclear |
| 9028 | |
| 9029 | @noindent |
| 9030 | In this example, the symbol @code{bouquet} holds the address of the first |
| 9031 | pair of boxes. |
| 9032 | |
| 9033 | @need 1200 |
| 9034 | This same list can be illustrated in a different sort of box notation |
| 9035 | like this: |
| 9036 | |
| 9037 | @c cons-cell-diagram #2a |
| 9038 | @ifnottex |
| 9039 | @smallexample |
| 9040 | @group |
| 9041 | bouquet |
| 9042 | | |
| 9043 | | -------------- --------------- ---------------- |
| 9044 | | | car | cdr | | car | cdr | | car | cdr | |
| 9045 | -->| rose | o------->| violet | o------->| butter- | nil | |
| 9046 | | | | | | | | cup | | |
| 9047 | -------------- --------------- ---------------- |
| 9048 | @end group |
| 9049 | @end smallexample |
| 9050 | @end ifnottex |
| 9051 | @ifset print-postscript-figures |
| 9052 | @sp 1 |
| 9053 | @tex |
| 9054 | @image{cons-2a} |
| 9055 | %%%% old method of including an image |
| 9056 | % \input /usr/local/lib/tex/inputs/psfig.tex |
| 9057 | % \centerline{\psfig{figure=/usr/local/lib/emacs/man/cons-2a.eps}} |
| 9058 | % \catcode`\@=0 % |
| 9059 | @end tex |
| 9060 | @sp 1 |
| 9061 | @end ifset |
| 9062 | @ifclear print-postscript-figures |
| 9063 | @iftex |
| 9064 | @smallexample |
| 9065 | @group |
| 9066 | bouquet |
| 9067 | | |
| 9068 | | -------------- --------------- ---------------- |
| 9069 | | | car | cdr | | car | cdr | | car | cdr | |
| 9070 | -->| rose | o------->| violet | o------->| butter- | nil | |
| 9071 | | | | | | | | cup | | |
| 9072 | -------------- --------------- ---------------- |
| 9073 | @end group |
| 9074 | @end smallexample |
| 9075 | @end iftex |
| 9076 | @end ifclear |
| 9077 | |
| 9078 | (Symbols consist of more than pairs of addresses, but the structure of |
| 9079 | a symbol is made up of addresses. Indeed, the symbol @code{bouquet} |
| 9080 | consists of a group of address-boxes, one of which is the address of |
| 9081 | the printed word @samp{bouquet}, a second of which is the address of a |
| 9082 | function definition attached to the symbol, if any, a third of which |
| 9083 | is the address of the first pair of address-boxes for the list |
| 9084 | @code{(rose violet buttercup)}, and so on. Here we are showing that |
| 9085 | the symbol's third address-box points to the first pair of |
| 9086 | address-boxes for the list.) |
| 9087 | |
| 9088 | If a symbol is set to the @sc{cdr} of a list, the list itself is not |
| 9089 | changed; the symbol simply has an address further down the list. (In |
| 9090 | the jargon, @sc{car} and @sc{cdr} are `non-destructive'.) Thus, |
| 9091 | evaluation of the following expression |
| 9092 | |
| 9093 | @smallexample |
| 9094 | (setq flowers (cdr bouquet)) |
| 9095 | @end smallexample |
| 9096 | |
| 9097 | @need 800 |
| 9098 | @noindent |
| 9099 | produces this: |
| 9100 | |
| 9101 | @c cons-cell-diagram #3 |
| 9102 | @ifnottex |
| 9103 | @sp 1 |
| 9104 | @smallexample |
| 9105 | @group |
| 9106 | bouquet flowers |
| 9107 | | | |
| 9108 | | ___ ___ | ___ ___ ___ ___ |
| 9109 | --> | | | --> | | | | | | |
| 9110 | |___|___|----> |___|___|--> |___|___|--> nil |
| 9111 | | | | |
| 9112 | | | | |
| 9113 | --> rose --> violet --> buttercup |
| 9114 | @end group |
| 9115 | @end smallexample |
| 9116 | @sp 1 |
| 9117 | @end ifnottex |
| 9118 | @ifset print-postscript-figures |
| 9119 | @sp 1 |
| 9120 | @tex |
| 9121 | @image{cons-3} |
| 9122 | %%%% old method of including an image |
| 9123 | % \input /usr/local/lib/tex/inputs/psfig.tex |
| 9124 | % \centerline{\psfig{figure=/usr/local/lib/emacs/man/cons-3.eps}} |
| 9125 | % \catcode`\@=0 % |
| 9126 | @end tex |
| 9127 | @sp 1 |
| 9128 | @end ifset |
| 9129 | @ifclear print-postscript-figures |
| 9130 | @iftex |
| 9131 | @sp 1 |
| 9132 | @smallexample |
| 9133 | @group |
| 9134 | bouquet flowers |
| 9135 | | | |
| 9136 | | ___ ___ | ___ ___ ___ ___ |
| 9137 | --> | | | --> | | | | | | |
| 9138 | |___|___|----> |___|___|--> |___|___|--> nil |
| 9139 | | | | |
| 9140 | | | | |
| 9141 | --> rose --> violet --> buttercup |
| 9142 | @end group |
| 9143 | @end smallexample |
| 9144 | @sp 1 |
| 9145 | @end iftex |
| 9146 | @end ifclear |
| 9147 | |
| 9148 | @noindent |
| 9149 | The value of @code{flowers} is @code{(violet buttercup)}, which is |
| 9150 | to say, the symbol @code{flowers} holds the address of the pair of |
| 9151 | address-boxes, the first of which holds the address of @code{violet}, |
| 9152 | and the second of which holds the address of @code{buttercup}. |
| 9153 | |
| 9154 | A pair of address-boxes is called a @dfn{cons cell} or @dfn{dotted |
| 9155 | pair}. @xref{Cons Cell Type, , Cons Cell and List Types, elisp, The GNU Emacs Lisp |
| 9156 | Reference Manual}, and @ref{Dotted Pair Notation, , Dotted Pair |
| 9157 | Notation, elisp, The GNU Emacs Lisp Reference Manual}, for more |
| 9158 | information about cons cells and dotted pairs. |
| 9159 | |
| 9160 | @need 1200 |
| 9161 | The function @code{cons} adds a new pair of addresses to the front of |
| 9162 | a series of addresses like that shown above. For example, evaluating |
| 9163 | the expression |
| 9164 | |
| 9165 | @smallexample |
| 9166 | (setq bouquet (cons 'lily bouquet)) |
| 9167 | @end smallexample |
| 9168 | |
| 9169 | @need 1500 |
| 9170 | @noindent |
| 9171 | produces: |
| 9172 | |
| 9173 | @c cons-cell-diagram #4 |
| 9174 | @ifnottex |
| 9175 | @sp 1 |
| 9176 | @smallexample |
| 9177 | @group |
| 9178 | bouquet flowers |
| 9179 | | | |
| 9180 | | ___ ___ ___ ___ | ___ ___ ___ ___ |
| 9181 | --> | | | | | | --> | | | | | | |
| 9182 | |___|___|----> |___|___|----> |___|___|---->|___|___|--> nil |
| 9183 | | | | | |
| 9184 | | | | | |
| 9185 | --> lily --> rose --> violet --> buttercup |
| 9186 | @end group |
| 9187 | @end smallexample |
| 9188 | @sp 1 |
| 9189 | @end ifnottex |
| 9190 | @ifset print-postscript-figures |
| 9191 | @sp 1 |
| 9192 | @tex |
| 9193 | @image{cons-4} |
| 9194 | %%%% old method of including an image |
| 9195 | % \input /usr/local/lib/tex/inputs/psfig.tex |
| 9196 | % \centerline{\psfig{figure=/usr/local/lib/emacs/man/cons-4.eps}} |
| 9197 | % \catcode`\@=0 % |
| 9198 | @end tex |
| 9199 | @sp 1 |
| 9200 | @end ifset |
| 9201 | @ifclear print-postscript-figures |
| 9202 | @iftex |
| 9203 | @sp 1 |
| 9204 | @smallexample |
| 9205 | @group |
| 9206 | bouquet flowers |
| 9207 | | | |
| 9208 | | ___ ___ ___ ___ | ___ ___ ___ ___ |
| 9209 | --> | | | | | | --> | | | | | | |
| 9210 | |___|___|----> |___|___|----> |___|___|---->|___|___|--> nil |
| 9211 | | | | | |
| 9212 | | | | | |
| 9213 | --> lily --> rose --> violet --> buttercup |
| 9214 | @end group |
| 9215 | @end smallexample |
| 9216 | @sp 1 |
| 9217 | @end iftex |
| 9218 | @end ifclear |
| 9219 | |
| 9220 | @need 1200 |
| 9221 | @noindent |
| 9222 | However, this does not change the value of the symbol |
| 9223 | @code{flowers}, as you can see by evaluating the following, |
| 9224 | |
| 9225 | @smallexample |
| 9226 | (eq (cdr (cdr bouquet)) flowers) |
| 9227 | @end smallexample |
| 9228 | |
| 9229 | @noindent |
| 9230 | which returns @code{t} for true. |
| 9231 | |
| 9232 | Until it is reset, @code{flowers} still has the value |
| 9233 | @code{(violet buttercup)}; that is, it has the address of the cons |
| 9234 | cell whose first address is of @code{violet}. Also, this does not |
| 9235 | alter any of the pre-existing cons cells; they are all still there. |
| 9236 | |
| 9237 | Thus, in Lisp, to get the @sc{cdr} of a list, you just get the address |
| 9238 | of the next cons cell in the series; to get the @sc{car} of a list, |
| 9239 | you get the address of the first element of the list; to @code{cons} a |
| 9240 | new element on a list, you add a new cons cell to the front of the list. |
| 9241 | That is all there is to it! The underlying structure of Lisp is |
| 9242 | brilliantly simple! |
| 9243 | |
| 9244 | And what does the last address in a series of cons cells refer to? It |
| 9245 | is the address of the empty list, of @code{nil}. |
| 9246 | |
| 9247 | In summary, when a Lisp variable is set to a value, it is provided with |
| 9248 | the address of the list to which the variable refers. |
| 9249 | |
| 9250 | @node Symbols as Chest, List Exercise, Lists diagrammed, List Implementation |
| 9251 | @section Symbols as a Chest of Drawers |
| 9252 | @cindex Symbols as a Chest of Drawers |
| 9253 | @cindex Chest of Drawers, metaphor for a symbol |
| 9254 | @cindex Drawers, Chest of, metaphor for a symbol |
| 9255 | |
| 9256 | In an earlier section, I suggested that you might imagine a symbol as |
| 9257 | being a chest of drawers. The function definition is put in one |
| 9258 | drawer, the value in another, and so on. What is put in the drawer |
| 9259 | holding the value can be changed without affecting the contents of the |
| 9260 | drawer holding the function definition, and vice-versa. |
| 9261 | |
| 9262 | Actually, what is put in each drawer is the address of the value or |
| 9263 | function definition. It is as if you found an old chest in the attic, |
| 9264 | and in one of its drawers you found a map giving you directions to |
| 9265 | where the buried treasure lies. |
| 9266 | |
| 9267 | (In addition to its name, symbol definition, and variable value, a |
| 9268 | symbol has a `drawer' for a @dfn{property list} which can be used to |
| 9269 | record other information. Property lists are not discussed here; see |
| 9270 | @ref{Property Lists, , Property Lists, elisp, The GNU Emacs Lisp |
| 9271 | Reference Manual}.) |
| 9272 | |
| 9273 | @need 1500 |
| 9274 | Here is a fanciful representation: |
| 9275 | |
| 9276 | @c chest-of-drawers diagram |
| 9277 | @ifnottex |
| 9278 | @sp 1 |
| 9279 | @smallexample |
| 9280 | @group |
| 9281 | Chest of Drawers Contents of Drawers |
| 9282 | |
| 9283 | __ o0O0o __ |
| 9284 | / \ |
| 9285 | --------------------- |
| 9286 | | directions to | [map to] |
| 9287 | | symbol name | bouquet |
| 9288 | | | |
| 9289 | +---------------------+ |
| 9290 | | directions to | |
| 9291 | | symbol definition | [none] |
| 9292 | | | |
| 9293 | +---------------------+ |
| 9294 | | directions to | [map to] |
| 9295 | | variable value | (rose violet buttercup) |
| 9296 | | | |
| 9297 | +---------------------+ |
| 9298 | | directions to | |
| 9299 | | property list | [not described here] |
| 9300 | | | |
| 9301 | +---------------------+ |
| 9302 | |/ \| |
| 9303 | @end group |
| 9304 | @end smallexample |
| 9305 | @sp 1 |
| 9306 | @end ifnottex |
| 9307 | @ifset print-postscript-figures |
| 9308 | @sp 1 |
| 9309 | @tex |
| 9310 | @image{drawers} |
| 9311 | %%%% old method of including an image |
| 9312 | % \input /usr/local/lib/tex/inputs/psfig.tex |
| 9313 | % \centerline{\psfig{figure=/usr/local/lib/emacs/man/drawers.eps}} |
| 9314 | % \catcode`\@=0 % |
| 9315 | @end tex |
| 9316 | @sp 1 |
| 9317 | @end ifset |
| 9318 | @ifclear print-postscript-figures |
| 9319 | @iftex |
| 9320 | @sp 1 |
| 9321 | @smallexample |
| 9322 | @group |
| 9323 | Chest of Drawers Contents of Drawers |
| 9324 | |
| 9325 | __ o0O0o __ |
| 9326 | / \ |
| 9327 | --------------------- |
| 9328 | | directions to | [map to] |
| 9329 | | symbol name | bouquet |
| 9330 | | | |
| 9331 | +---------------------+ |
| 9332 | | directions to | |
| 9333 | | symbol definition | [none] |
| 9334 | | | |
| 9335 | +---------------------+ |
| 9336 | | directions to | [map to] |
| 9337 | | variable value | (rose violet buttercup) |
| 9338 | | | |
| 9339 | +---------------------+ |
| 9340 | | directions to | |
| 9341 | | property list | [not described here] |
| 9342 | | | |
| 9343 | +---------------------+ |
| 9344 | |/ \| |
| 9345 | @end group |
| 9346 | @end smallexample |
| 9347 | @sp 1 |
| 9348 | @end iftex |
| 9349 | @end ifclear |
| 9350 | |
| 9351 | @node List Exercise, , Symbols as Chest, List Implementation |
| 9352 | @section Exercise |
| 9353 | |
| 9354 | Set @code{flowers} to @code{violet} and @code{buttercup}. Cons two |
| 9355 | more flowers on to this list and set this new list to |
| 9356 | @code{more-flowers}. Set the @sc{car} of @code{flowers} to a fish. |
| 9357 | What does the @code{more-flowers} list now contain? |
| 9358 | |
| 9359 | @node Yanking, Loops & Recursion, List Implementation, Top |
| 9360 | @comment node-name, next, previous, up |
| 9361 | @chapter Yanking Text Back |
| 9362 | @findex yank |
| 9363 | @findex rotate-yank-pointer |
| 9364 | @cindex Text retrieval |
| 9365 | @cindex Retrieving text |
| 9366 | @cindex Pasting text |
| 9367 | |
| 9368 | Whenever you cut text out of a buffer with a `kill' command in GNU Emacs, |
| 9369 | you can bring it back with a `yank' command. The text that is cut out of |
| 9370 | the buffer is put in the kill ring and the yank commands insert the |
| 9371 | appropriate contents of the kill ring back into a buffer (not necessarily |
| 9372 | the original buffer). |
| 9373 | |
| 9374 | A simple @kbd{C-y} (@code{yank}) command inserts the first item from |
| 9375 | the kill ring into the current buffer. If the @kbd{C-y} command is |
| 9376 | followed immediately by @kbd{M-y}, the first element is replaced by |
| 9377 | the second element. Successive @kbd{M-y} commands replace the second |
| 9378 | element with the third, fourth, or fifth element, and so on. When the |
| 9379 | last element in the kill ring is reached, it is replaced by the first |
| 9380 | element and the cycle is repeated. (Thus the kill ring is called a |
| 9381 | `ring' rather than just a `list'. However, the actual data structure |
| 9382 | that holds the text is a list. |
| 9383 | @xref{Kill Ring, , Handling the Kill Ring}, for the details of how the |
| 9384 | list is handled as a ring.) |
| 9385 | |
| 9386 | @menu |
| 9387 | * Kill Ring Overview:: The kill ring is a list. |
| 9388 | * kill-ring-yank-pointer:: The @code{kill-ring-yank-pointer} variable. |
| 9389 | * yank nthcdr Exercises:: |
| 9390 | @end menu |
| 9391 | |
| 9392 | @node Kill Ring Overview, kill-ring-yank-pointer, Yanking, Yanking |
| 9393 | @comment node-name, next, previous, up |
| 9394 | @section Kill Ring Overview |
| 9395 | @cindex Kill ring overview |
| 9396 | |
| 9397 | The kill ring is a list of textual strings. This is what it looks like: |
| 9398 | |
| 9399 | @smallexample |
| 9400 | ("some text" "a different piece of text" "yet more text") |
| 9401 | @end smallexample |
| 9402 | |
| 9403 | If this were the contents of my kill ring and I pressed @kbd{C-y}, the |
| 9404 | string of characters saying @samp{some text} would be inserted in this |
| 9405 | buffer where my cursor is located. |
| 9406 | |
| 9407 | The @code{yank} command is also used for duplicating text by copying it. |
| 9408 | The copied text is not cut from the buffer, but a copy of it is put on the |
| 9409 | kill ring and is inserted by yanking it back. |
| 9410 | |
| 9411 | Three functions are used for bringing text back from the kill ring: |
| 9412 | @code{yank}, which is usually bound to @kbd{C-y}; @code{yank-pop}, |
| 9413 | which is usually bound to @kbd{M-y}; and @code{rotate-yank-pointer}, |
| 9414 | which is used by the two other functions. |
| 9415 | |
| 9416 | These functions refer to the kill ring through a variable called the |
| 9417 | @code{kill-ring-yank-pointer}. Indeed, the insertion code for both the |
| 9418 | @code{yank} and @code{yank-pop} functions is: |
| 9419 | |
| 9420 | @smallexample |
| 9421 | (insert (car kill-ring-yank-pointer)) |
| 9422 | @end smallexample |
| 9423 | |
| 9424 | To begin to understand how @code{yank} and @code{yank-pop} work, it is |
| 9425 | first necessary to look at the @code{kill-ring-yank-pointer} variable |
| 9426 | and the @code{rotate-yank-pointer} function. |
| 9427 | |
| 9428 | @node kill-ring-yank-pointer, yank nthcdr Exercises, Kill Ring Overview, Yanking |
| 9429 | @comment node-name, next, previous, up |
| 9430 | @section The @code{kill-ring-yank-pointer} Variable |
| 9431 | |
| 9432 | @code{kill-ring-yank-pointer} is a variable, just as @code{kill-ring} is |
| 9433 | a variable. It points to something by being bound to the value of what |
| 9434 | it points to, like any other Lisp variable. |
| 9435 | |
| 9436 | @need 1000 |
| 9437 | Thus, if the value of the kill ring is: |
| 9438 | |
| 9439 | @smallexample |
| 9440 | ("some text" "a different piece of text" "yet more text") |
| 9441 | @end smallexample |
| 9442 | |
| 9443 | @need 1250 |
| 9444 | @noindent |
| 9445 | and the @code{kill-ring-yank-pointer} points to the second clause, the |
| 9446 | value of @code{kill-ring-yank-pointer} is: |
| 9447 | |
| 9448 | @smallexample |
| 9449 | ("a different piece of text" "yet more text") |
| 9450 | @end smallexample |
| 9451 | |
| 9452 | As explained in the previous chapter (@pxref{List Implementation}), the |
| 9453 | computer does not keep two different copies of the text being pointed to |
| 9454 | by both the @code{kill-ring} and the @code{kill-ring-yank-pointer}. The |
| 9455 | words ``a different piece of text'' and ``yet more text'' are not |
| 9456 | duplicated. Instead, the two Lisp variables point to the same pieces of |
| 9457 | text. Here is a diagram: |
| 9458 | |
| 9459 | @c cons-cell-diagram #5 |
| 9460 | @ifnottex |
| 9461 | @smallexample |
| 9462 | @group |
| 9463 | kill-ring kill-ring-yank-pointer |
| 9464 | | | |
| 9465 | | ___ ___ | ___ ___ ___ ___ |
| 9466 | ---> | | | --> | | | | | | |
| 9467 | |___|___|----> |___|___|--> |___|___|--> nil |
| 9468 | | | | |
| 9469 | | | | |
| 9470 | | | --> "yet more text" |
| 9471 | | | |
| 9472 | | --> "a different piece of text |
| 9473 | | |
| 9474 | --> "some text" |
| 9475 | @end group |
| 9476 | @end smallexample |
| 9477 | @sp 1 |
| 9478 | @end ifnottex |
| 9479 | @ifset print-postscript-figures |
| 9480 | @sp 1 |
| 9481 | @tex |
| 9482 | @image{cons-5} |
| 9483 | %%%% old method of including an image |
| 9484 | % \input /usr/local/lib/tex/inputs/psfig.tex |
| 9485 | % \centerline{\psfig{figure=/usr/local/lib/emacs/man/cons-5.eps}} |
| 9486 | % \catcode`\@=0 % |
| 9487 | @end tex |
| 9488 | @sp 1 |
| 9489 | @end ifset |
| 9490 | @ifclear print-postscript-figures |
| 9491 | @iftex |
| 9492 | @smallexample |
| 9493 | @group |
| 9494 | kill-ring kill-ring-yank-pointer |
| 9495 | | | |
| 9496 | | ___ ___ | ___ ___ ___ ___ |
| 9497 | ---> | | | --> | | | | | | |
| 9498 | |___|___|----> |___|___|--> |___|___|--> nil |
| 9499 | | | | |
| 9500 | | | | |
| 9501 | | | --> "yet more text" |
| 9502 | | | |
| 9503 | | --> "a different piece of text |
| 9504 | | |
| 9505 | --> "some text" |
| 9506 | @end group |
| 9507 | @end smallexample |
| 9508 | @sp 1 |
| 9509 | @end iftex |
| 9510 | @end ifclear |
| 9511 | |
| 9512 | Both the variable @code{kill-ring} and the variable |
| 9513 | @code{kill-ring-yank-pointer} are pointers. But the kill ring itself is |
| 9514 | usually described as if it were actually what it is composed of. The |
| 9515 | @code{kill-ring} is spoken of as if it were the list rather than that it |
| 9516 | points to the list. Conversely, the @code{kill-ring-yank-pointer} is |
| 9517 | spoken of as pointing to a list. |
| 9518 | |
| 9519 | These two ways of talking about the same thing sound confusing at first but |
| 9520 | make sense on reflection. The kill ring is generally thought of as the |
| 9521 | complete structure of data that holds the information of what has recently |
| 9522 | been cut out of the Emacs buffers. The @code{kill-ring-yank-pointer} |
| 9523 | on the other hand, serves to indicate---that is, to `point to'---that part |
| 9524 | of the kill ring of which the first element (the @sc{car}) will be |
| 9525 | inserted. |
| 9526 | |
| 9527 | The @code{rotate-yank-pointer} function changes the element in the |
| 9528 | kill ring to which the @code{kill-ring-yank-pointer} points; when the |
| 9529 | pointer is set to point to the next element beyond the end of the kill |
| 9530 | ring, it automatically sets it to point to the first element of the |
| 9531 | kill ring. This is how the list is transformed into a ring. The |
| 9532 | @code{rotate-yank-pointer} function itself is not difficult, but |
| 9533 | contains many details. It and the much simpler @code{yank} and |
| 9534 | @code{yank-pop} functions are described in an appendix. |
| 9535 | @xref{Kill Ring, , Handling the Kill Ring}. |
| 9536 | |
| 9537 | @need 1500 |
| 9538 | @node yank nthcdr Exercises, , kill-ring-yank-pointer, Yanking |
| 9539 | @section Exercises with @code{yank} and @code{nthcdr} |
| 9540 | |
| 9541 | @itemize @bullet |
| 9542 | @item |
| 9543 | Using @kbd{C-h v} (@code{describe-variable}), look at the value of |
| 9544 | your kill ring. Add several items to your kill ring; look at its |
| 9545 | value again. Using @kbd{M-y} (@code{yank-pop)}, move all the way |
| 9546 | around the kill ring. How many items were in your kill ring? Find |
| 9547 | the value of @code{kill-ring-max}. Was your kill ring full, or could |
| 9548 | you have kept more blocks of text within it? |
| 9549 | |
| 9550 | @item |
| 9551 | Using @code{nthcdr} and @code{car}, construct a series of expressions |
| 9552 | to return the first, second, third, and fourth elements of a list. |
| 9553 | @end itemize |
| 9554 | |
| 9555 | @node Loops & Recursion, Regexp Search, Yanking, Top |
| 9556 | @comment node-name, next, previous, up |
| 9557 | @chapter Loops and Recursion |
| 9558 | @cindex Loops and recursion |
| 9559 | @cindex Recursion and loops |
| 9560 | @cindex Repetition (loops) |
| 9561 | |
| 9562 | Emacs Lisp has two primary ways to cause an expression, or a series of |
| 9563 | expressions, to be evaluated repeatedly: one uses a @code{while} |
| 9564 | loop, and the other uses @dfn{recursion}. |
| 9565 | |
| 9566 | Repetition can be very valuable. For example, to move forward four |
| 9567 | sentences, you need only write a program that will move forward one |
| 9568 | sentence and then repeat the process four times. Since a computer does |
| 9569 | not get bored or tired, such repetitive action does not have the |
| 9570 | deleterious effects that excessive or the wrong kinds of repetition can |
| 9571 | have on humans. |
| 9572 | |
| 9573 | People mostly write Emacs Lisp functions using @code{while} loops and |
| 9574 | their kin; but you can use recursion, which provides a very powerful |
| 9575 | way to think about and then to solve problems@footnote{You can write |
| 9576 | recursive functions to be frugal or wasteful of mental or computer |
| 9577 | resources; as it happens, methods that people find easy---that are |
| 9578 | frugal of `mental resources'---sometimes use considerable computer |
| 9579 | resources. Emacs was designed to run on machines that we now consider |
| 9580 | limited and its default settings are conservative. You may want to |
| 9581 | increase the values of @code{max-specpdl-size} and |
| 9582 | @code{max-lisp-eval-depth}. In my @file{.emacs} file, I set them to |
| 9583 | 15 and 30 times their default value.}. |
| 9584 | |
| 9585 | @menu |
| 9586 | * while:: Causing a stretch of code to repeat. |
| 9587 | * dolist dotimes:: |
| 9588 | * Recursion:: Causing a function to call itself. |
| 9589 | * Looping exercise:: |
| 9590 | @end menu |
| 9591 | |
| 9592 | @node while, dolist dotimes, Loops & Recursion, Loops & Recursion |
| 9593 | @comment node-name, next, previous, up |
| 9594 | @section @code{while} |
| 9595 | @cindex Loops |
| 9596 | @findex while |
| 9597 | |
| 9598 | The @code{while} special form tests whether the value returned by |
| 9599 | evaluating its first argument is true or false. This is similar to what |
| 9600 | the Lisp interpreter does with an @code{if}; what the interpreter does |
| 9601 | next, however, is different. |
| 9602 | |
| 9603 | In a @code{while} expression, if the value returned by evaluating the |
| 9604 | first argument is false, the Lisp interpreter skips the rest of the |
| 9605 | expression (the @dfn{body} of the expression) and does not evaluate it. |
| 9606 | However, if the value is true, the Lisp interpreter evaluates the body |
| 9607 | of the expression and then again tests whether the first argument to |
| 9608 | @code{while} is true or false. If the value returned by evaluating the |
| 9609 | first argument is again true, the Lisp interpreter again evaluates the |
| 9610 | body of the expression. |
| 9611 | |
| 9612 | @need 1200 |
| 9613 | The template for a @code{while} expression looks like this: |
| 9614 | |
| 9615 | @smallexample |
| 9616 | @group |
| 9617 | (while @var{true-or-false-test} |
| 9618 | @var{body}@dots{}) |
| 9619 | @end group |
| 9620 | @end smallexample |
| 9621 | |
| 9622 | @menu |
| 9623 | * Looping with while:: Repeat so long as test returns true. |
| 9624 | * Loop Example:: A @code{while} loop that uses a list. |
| 9625 | * print-elements-of-list:: Uses @code{while}, @code{car}, @code{cdr}. |
| 9626 | * Incrementing Loop:: A loop with an incrementing counter. |
| 9627 | * Decrementing Loop:: A loop with a decrementing counter. |
| 9628 | @end menu |
| 9629 | |
| 9630 | @node Looping with while, Loop Example, while, while |
| 9631 | @ifnottex |
| 9632 | @unnumberedsubsec Looping with @code{while} |
| 9633 | @end ifnottex |
| 9634 | |
| 9635 | So long as the true-or-false-test of the @code{while} expression |
| 9636 | returns a true value when it is evaluated, the body is repeatedly |
| 9637 | evaluated. This process is called a loop since the Lisp interpreter |
| 9638 | repeats the same thing again and again, like an airplane doing a loop. |
| 9639 | When the result of evaluating the true-or-false-test is false, the |
| 9640 | Lisp interpreter does not evaluate the rest of the @code{while} |
| 9641 | expression and `exits the loop'. |
| 9642 | |
| 9643 | Clearly, if the value returned by evaluating the first argument to |
| 9644 | @code{while} is always true, the body following will be evaluated |
| 9645 | again and again @dots{} and again @dots{} forever. Conversely, if the |
| 9646 | value returned is never true, the expressions in the body will never |
| 9647 | be evaluated. The craft of writing a @code{while} loop consists of |
| 9648 | choosing a mechanism such that the true-or-false-test returns true |
| 9649 | just the number of times that you want the subsequent expressions to |
| 9650 | be evaluated, and then have the test return false. |
| 9651 | |
| 9652 | The value returned by evaluating a @code{while} is the value of the |
| 9653 | true-or-false-test. An interesting consequence of this is that a |
| 9654 | @code{while} loop that evaluates without error will return @code{nil} |
| 9655 | or false regardless of whether it has looped 1 or 100 times or none at |
| 9656 | all. A @code{while} expression that evaluates successfully never |
| 9657 | returns a true value! What this means is that @code{while} is always |
| 9658 | evaluated for its side effects, which is to say, the consequences of |
| 9659 | evaluating the expressions within the body of the @code{while} loop. |
| 9660 | This makes sense. It is not the mere act of looping that is desired, |
| 9661 | but the consequences of what happens when the expressions in the loop |
| 9662 | are repeatedly evaluated. |
| 9663 | |
| 9664 | @node Loop Example, print-elements-of-list, Looping with while, while |
| 9665 | @comment node-name, next, previous, up |
| 9666 | @subsection A @code{while} Loop and a List |
| 9667 | |
| 9668 | A common way to control a @code{while} loop is to test whether a list |
| 9669 | has any elements. If it does, the loop is repeated; but if it does not, |
| 9670 | the repetition is ended. Since this is an important technique, we will |
| 9671 | create a short example to illustrate it. |
| 9672 | |
| 9673 | A simple way to test whether a list has elements is to evaluate the |
| 9674 | list: if it has no elements, it is an empty list and will return the |
| 9675 | empty list, @code{()}, which is a synonym for @code{nil} or false. On |
| 9676 | the other hand, a list with elements will return those elements when it |
| 9677 | is evaluated. Since Emacs Lisp considers as true any value that is not |
| 9678 | @code{nil}, a list that returns elements will test true in a |
| 9679 | @code{while} loop. |
| 9680 | |
| 9681 | @need 1200 |
| 9682 | For example, you can set the variable @code{empty-list} to @code{nil} by |
| 9683 | evaluating the following @code{setq} expression: |
| 9684 | |
| 9685 | @smallexample |
| 9686 | (setq empty-list ()) |
| 9687 | @end smallexample |
| 9688 | |
| 9689 | @noindent |
| 9690 | After evaluating the @code{setq} expression, you can evaluate the |
| 9691 | variable @code{empty-list} in the usual way, by placing the cursor after |
| 9692 | the symbol and typing @kbd{C-x C-e}; @code{nil} will appear in your |
| 9693 | echo area: |
| 9694 | |
| 9695 | @smallexample |
| 9696 | empty-list |
| 9697 | @end smallexample |
| 9698 | |
| 9699 | On the other hand, if you set a variable to be a list with elements, the |
| 9700 | list will appear when you evaluate the variable, as you can see by |
| 9701 | evaluating the following two expressions: |
| 9702 | |
| 9703 | @smallexample |
| 9704 | @group |
| 9705 | (setq animals '(gazelle giraffe lion tiger)) |
| 9706 | |
| 9707 | animals |
| 9708 | @end group |
| 9709 | @end smallexample |
| 9710 | |
| 9711 | Thus, to create a @code{while} loop that tests whether there are any |
| 9712 | items in the list @code{animals}, the first part of the loop will be |
| 9713 | written like this: |
| 9714 | |
| 9715 | @smallexample |
| 9716 | @group |
| 9717 | (while animals |
| 9718 | @dots{} |
| 9719 | @end group |
| 9720 | @end smallexample |
| 9721 | |
| 9722 | @noindent |
| 9723 | When the @code{while} tests its first argument, the variable |
| 9724 | @code{animals} is evaluated. It returns a list. So long as the list |
| 9725 | has elements, the @code{while} considers the results of the test to be |
| 9726 | true; but when the list is empty, it considers the results of the test |
| 9727 | to be false. |
| 9728 | |
| 9729 | To prevent the @code{while} loop from running forever, some mechanism |
| 9730 | needs to be provided to empty the list eventually. An oft-used |
| 9731 | technique is to have one of the subsequent forms in the @code{while} |
| 9732 | expression set the value of the list to be the @sc{cdr} of the list. |
| 9733 | Each time the @code{cdr} function is evaluated, the list will be made |
| 9734 | shorter, until eventually only the empty list will be left. At this |
| 9735 | point, the test of the @code{while} loop will return false, and the |
| 9736 | arguments to the @code{while} will no longer be evaluated. |
| 9737 | |
| 9738 | For example, the list of animals bound to the variable @code{animals} |
| 9739 | can be set to be the @sc{cdr} of the original list with the |
| 9740 | following expression: |
| 9741 | |
| 9742 | @smallexample |
| 9743 | (setq animals (cdr animals)) |
| 9744 | @end smallexample |
| 9745 | |
| 9746 | @noindent |
| 9747 | If you have evaluated the previous expressions and then evaluate this |
| 9748 | expression, you will see @code{(giraffe lion tiger)} appear in the echo |
| 9749 | area. If you evaluate the expression again, @code{(lion tiger)} will |
| 9750 | appear in the echo area. If you evaluate it again and yet again, |
| 9751 | @code{(tiger)} appears and then the empty list, shown by @code{nil}. |
| 9752 | |
| 9753 | A template for a @code{while} loop that uses the @code{cdr} function |
| 9754 | repeatedly to cause the true-or-false-test eventually to test false |
| 9755 | looks like this: |
| 9756 | |
| 9757 | @smallexample |
| 9758 | @group |
| 9759 | (while @var{test-whether-list-is-empty} |
| 9760 | @var{body}@dots{} |
| 9761 | @var{set-list-to-cdr-of-list}) |
| 9762 | @end group |
| 9763 | @end smallexample |
| 9764 | |
| 9765 | This test and use of @code{cdr} can be put together in a function that |
| 9766 | goes through a list and prints each element of the list on a line of its |
| 9767 | own. |
| 9768 | |
| 9769 | @node print-elements-of-list, Incrementing Loop, Loop Example, while |
| 9770 | @subsection An Example: @code{print-elements-of-list} |
| 9771 | @findex print-elements-of-list |
| 9772 | |
| 9773 | The @code{print-elements-of-list} function illustrates a @code{while} |
| 9774 | loop with a list. |
| 9775 | |
| 9776 | @cindex @file{*scratch*} buffer |
| 9777 | The function requires several lines for its output. If you are |
| 9778 | reading this in Emacs 21 or a later version, you can evaluate the |
| 9779 | following expression inside of Info, as usual. |
| 9780 | |
| 9781 | If you are using an earlier version of Emacs, you need to copy the |
| 9782 | necessary expressions to your @file{*scratch*} buffer and evaluate |
| 9783 | them there. This is because the echo area had only one line in the |
| 9784 | earlier versions. |
| 9785 | |
| 9786 | You can copy the expressions by marking the beginning of the region |
| 9787 | with @kbd{C-@key{SPC}} (@code{set-mark-command}), moving the cursor to |
| 9788 | the end of the region and then copying the region using @kbd{M-w} |
| 9789 | (@code{copy-region-as-kill}). In the @file{*scratch*} buffer, you can |
| 9790 | yank the expressions back by typing @kbd{C-y} (@code{yank}). |
| 9791 | |
| 9792 | After you have copied the expressions to the @file{*scratch*} buffer, |
| 9793 | evaluate each expression in turn. Be sure to evaluate the last |
| 9794 | expression, @code{(print-elements-of-list animals)}, by typing |
| 9795 | @kbd{C-u C-x C-e}, that is, by giving an argument to |
| 9796 | @code{eval-last-sexp}. This will cause the result of the evaluation |
| 9797 | to be printed in the @file{*scratch*} buffer instead of being printed |
| 9798 | in the echo area. (Otherwise you will see something like this in your |
| 9799 | echo area: @code{^Jgazelle^J^Jgiraffe^J^Jlion^J^Jtiger^Jnil}, in which |
| 9800 | each @samp{^J} stands for a `newline'.) |
| 9801 | |
| 9802 | @need 1500 |
| 9803 | If you are using Emacs 21 or later, you can evaluate these expressions |
| 9804 | directly in the Info buffer, and the echo area will grow to show the |
| 9805 | results. |
| 9806 | |
| 9807 | @smallexample |
| 9808 | @group |
| 9809 | (setq animals '(gazelle giraffe lion tiger)) |
| 9810 | |
| 9811 | (defun print-elements-of-list (list) |
| 9812 | "Print each element of LIST on a line of its own." |
| 9813 | (while list |
| 9814 | (print (car list)) |
| 9815 | (setq list (cdr list)))) |
| 9816 | |
| 9817 | (print-elements-of-list animals) |
| 9818 | @end group |
| 9819 | @end smallexample |
| 9820 | |
| 9821 | @need 1200 |
| 9822 | @noindent |
| 9823 | When you evaluate the three expressions in sequence, you will see |
| 9824 | this: |
| 9825 | |
| 9826 | @smallexample |
| 9827 | @group |
| 9828 | gazelle |
| 9829 | |
| 9830 | giraffe |
| 9831 | |
| 9832 | lion |
| 9833 | |
| 9834 | tiger |
| 9835 | nil |
| 9836 | @end group |
| 9837 | @end smallexample |
| 9838 | |
| 9839 | Each element of the list is printed on a line of its own (that is what |
| 9840 | the function @code{print} does) and then the value returned by the |
| 9841 | function is printed. Since the last expression in the function is the |
| 9842 | @code{while} loop, and since @code{while} loops always return |
| 9843 | @code{nil}, a @code{nil} is printed after the last element of the list. |
| 9844 | |
| 9845 | @node Incrementing Loop, Decrementing Loop, print-elements-of-list, while |
| 9846 | @comment node-name, next, previous, up |
| 9847 | @subsection A Loop with an Incrementing Counter |
| 9848 | |
| 9849 | A loop is not useful unless it stops when it ought. Besides |
| 9850 | controlling a loop with a list, a common way of stopping a loop is to |
| 9851 | write the first argument as a test that returns false when the correct |
| 9852 | number of repetitions are complete. This means that the loop must |
| 9853 | have a counter---an expression that counts how many times the loop |
| 9854 | repeats itself. |
| 9855 | |
| 9856 | The test can be an expression such as @code{(< count desired-number)} |
| 9857 | which returns @code{t} for true if the value of @code{count} is less |
| 9858 | than the @code{desired-number} of repetitions and @code{nil} for false if |
| 9859 | the value of @code{count} is equal to or is greater than the |
| 9860 | @code{desired-number}. The expression that increments the count can be |
| 9861 | a simple @code{setq} such as @code{(setq count (1+ count))}, where |
| 9862 | @code{1+} is a built-in function in Emacs Lisp that adds 1 to its |
| 9863 | argument. (The expression @w{@code{(1+ count)}} has the same result as |
| 9864 | @w{@code{(+ count 1)}}, but is easier for a human to read.) |
| 9865 | |
| 9866 | @need 1250 |
| 9867 | The template for a @code{while} loop controlled by an incrementing |
| 9868 | counter looks like this: |
| 9869 | |
| 9870 | @smallexample |
| 9871 | @group |
| 9872 | @var{set-count-to-initial-value} |
| 9873 | (while (< count desired-number) ; @r{true-or-false-test} |
| 9874 | @var{body}@dots{} |
| 9875 | (setq count (1+ count))) ; @r{incrementer} |
| 9876 | @end group |
| 9877 | @end smallexample |
| 9878 | |
| 9879 | @noindent |
| 9880 | Note that you need to set the initial value of @code{count}; usually it |
| 9881 | is set to 1. |
| 9882 | |
| 9883 | @menu |
| 9884 | * Incrementing Example:: Counting pebbles in a triangle. |
| 9885 | * Inc Example parts:: The parts of the function definition. |
| 9886 | * Inc Example altogether:: Putting the function definition together. |
| 9887 | @end menu |
| 9888 | |
| 9889 | @node Incrementing Example, Inc Example parts, Incrementing Loop, Incrementing Loop |
| 9890 | @unnumberedsubsubsec Example with incrementing counter |
| 9891 | |
| 9892 | Suppose you are playing on the beach and decide to make a triangle of |
| 9893 | pebbles, putting one pebble in the first row, two in the second row, |
| 9894 | three in the third row and so on, like this: |
| 9895 | |
| 9896 | @sp 1 |
| 9897 | @c pebble diagram |
| 9898 | @ifnottex |
| 9899 | @smallexample |
| 9900 | @group |
| 9901 | * |
| 9902 | * * |
| 9903 | * * * |
| 9904 | * * * * |
| 9905 | @end group |
| 9906 | @end smallexample |
| 9907 | @end ifnottex |
| 9908 | @iftex |
| 9909 | @smallexample |
| 9910 | @group |
| 9911 | @bullet{} |
| 9912 | @bullet{} @bullet{} |
| 9913 | @bullet{} @bullet{} @bullet{} |
| 9914 | @bullet{} @bullet{} @bullet{} @bullet{} |
| 9915 | @end group |
| 9916 | @end smallexample |
| 9917 | @end iftex |
| 9918 | @sp 1 |
| 9919 | |
| 9920 | @noindent |
| 9921 | (About 2500 years ago, Pythagoras and others developed the beginnings of |
| 9922 | number theory by considering questions such as this.) |
| 9923 | |
| 9924 | Suppose you want to know how many pebbles you will need to make a |
| 9925 | triangle with 7 rows? |
| 9926 | |
| 9927 | Clearly, what you need to do is add up the numbers from 1 to 7. There |
| 9928 | are two ways to do this; start with the smallest number, one, and add up |
| 9929 | the list in sequence, 1, 2, 3, 4 and so on; or start with the largest |
| 9930 | number and add the list going down: 7, 6, 5, 4 and so on. Because both |
| 9931 | mechanisms illustrate common ways of writing @code{while} loops, we will |
| 9932 | create two examples, one counting up and the other counting down. In |
| 9933 | this first example, we will start with 1 and add 2, 3, 4 and so on. |
| 9934 | |
| 9935 | If you are just adding up a short list of numbers, the easiest way to do |
| 9936 | it is to add up all the numbers at once. However, if you do not know |
| 9937 | ahead of time how many numbers your list will have, or if you want to be |
| 9938 | prepared for a very long list, then you need to design your addition so |
| 9939 | that what you do is repeat a simple process many times instead of doing |
| 9940 | a more complex process once. |
| 9941 | |
| 9942 | For example, instead of adding up all the pebbles all at once, what you |
| 9943 | can do is add the number of pebbles in the first row, 1, to the number |
| 9944 | in the second row, 2, and then add the total of those two rows to the |
| 9945 | third row, 3. Then you can add the number in the fourth row, 4, to the |
| 9946 | total of the first three rows; and so on. |
| 9947 | |
| 9948 | The critical characteristic of the process is that each repetitive |
| 9949 | action is simple. In this case, at each step we add only two numbers, |
| 9950 | the number of pebbles in the row and the total already found. This |
| 9951 | process of adding two numbers is repeated again and again until the last |
| 9952 | row has been added to the total of all the preceding rows. In a more |
| 9953 | complex loop the repetitive action might not be so simple, but it will |
| 9954 | be simpler than doing everything all at once. |
| 9955 | |
| 9956 | @node Inc Example parts, Inc Example altogether, Incrementing Example, Incrementing Loop |
| 9957 | @unnumberedsubsubsec The parts of the function definition |
| 9958 | |
| 9959 | The preceding analysis gives us the bones of our function definition: |
| 9960 | first, we will need a variable that we can call @code{total} that will |
| 9961 | be the total number of pebbles. This will be the value returned by |
| 9962 | the function. |
| 9963 | |
| 9964 | Second, we know that the function will require an argument: this |
| 9965 | argument will be the total number of rows in the triangle. It can be |
| 9966 | called @code{number-of-rows}. |
| 9967 | |
| 9968 | Finally, we need a variable to use as a counter. We could call this |
| 9969 | variable @code{counter}, but a better name is @code{row-number}. |
| 9970 | That is because what the counter does is count rows, and a program |
| 9971 | should be written to be as understandable as possible. |
| 9972 | |
| 9973 | When the Lisp interpreter first starts evaluating the expressions in the |
| 9974 | function, the value of @code{total} should be set to zero, since we have |
| 9975 | not added anything to it. Then the function should add the number of |
| 9976 | pebbles in the first row to the total, and then add the number of |
| 9977 | pebbles in the second to the total, and then add the number of |
| 9978 | pebbles in the third row to the total, and so on, until there are no |
| 9979 | more rows left to add. |
| 9980 | |
| 9981 | Both @code{total} and @code{row-number} are used only inside the |
| 9982 | function, so they can be declared as local variables with @code{let} |
| 9983 | and given initial values. Clearly, the initial value for @code{total} |
| 9984 | should be 0. The initial value of @code{row-number} should be 1, |
| 9985 | since we start with the first row. This means that the @code{let} |
| 9986 | statement will look like this: |
| 9987 | |
| 9988 | @smallexample |
| 9989 | @group |
| 9990 | (let ((total 0) |
| 9991 | (row-number 1)) |
| 9992 | @var{body}@dots{}) |
| 9993 | @end group |
| 9994 | @end smallexample |
| 9995 | |
| 9996 | After the internal variables are declared and bound to their initial |
| 9997 | values, we can begin the @code{while} loop. The expression that serves |
| 9998 | as the test should return a value of @code{t} for true so long as the |
| 9999 | @code{row-number} is less than or equal to the @code{number-of-rows}. |
| 10000 | (If the expression tests true only so long as the row number is less |
| 10001 | than the number of rows in the triangle, the last row will never be |
| 10002 | added to the total; hence the row number has to be either less than or |
| 10003 | equal to the number of rows.) |
| 10004 | |
| 10005 | @need 1500 |
| 10006 | @findex <= @r{(less than or equal)} |
| 10007 | Lisp provides the @code{<=} function that returns true if the value of |
| 10008 | its first argument is less than or equal to the value of its second |
| 10009 | argument and false otherwise. So the expression that the @code{while} |
| 10010 | will evaluate as its test should look like this: |
| 10011 | |
| 10012 | @smallexample |
| 10013 | (<= row-number number-of-rows) |
| 10014 | @end smallexample |
| 10015 | |
| 10016 | The total number of pebbles can be found by repeatedly adding the number |
| 10017 | of pebbles in a row to the total already found. Since the number of |
| 10018 | pebbles in the row is equal to the row number, the total can be found by |
| 10019 | adding the row number to the total. (Clearly, in a more complex |
| 10020 | situation, the number of pebbles in the row might be related to the row |
| 10021 | number in a more complicated way; if this were the case, the row number |
| 10022 | would be replaced by the appropriate expression.) |
| 10023 | |
| 10024 | @smallexample |
| 10025 | (setq total (+ total row-number)) |
| 10026 | @end smallexample |
| 10027 | |
| 10028 | @noindent |
| 10029 | What this does is set the new value of @code{total} to be equal to the |
| 10030 | sum of adding the number of pebbles in the row to the previous total. |
| 10031 | |
| 10032 | After setting the value of @code{total}, the conditions need to be |
| 10033 | established for the next repetition of the loop, if there is one. This |
| 10034 | is done by incrementing the value of the @code{row-number} variable, |
| 10035 | which serves as a counter. After the @code{row-number} variable has |
| 10036 | been incremented, the true-or-false-test at the beginning of the |
| 10037 | @code{while} loop tests whether its value is still less than or equal to |
| 10038 | the value of the @code{number-of-rows} and if it is, adds the new value |
| 10039 | of the @code{row-number} variable to the @code{total} of the previous |
| 10040 | repetition of the loop. |
| 10041 | |
| 10042 | @need 1200 |
| 10043 | The built-in Emacs Lisp function @code{1+} adds 1 to a number, so the |
| 10044 | @code{row-number} variable can be incremented with this expression: |
| 10045 | |
| 10046 | @smallexample |
| 10047 | (setq row-number (1+ row-number)) |
| 10048 | @end smallexample |
| 10049 | |
| 10050 | @node Inc Example altogether, , Inc Example parts, Incrementing Loop |
| 10051 | @unnumberedsubsubsec Putting the function definition together |
| 10052 | |
| 10053 | We have created the parts for the function definition; now we need to |
| 10054 | put them together. |
| 10055 | |
| 10056 | @need 800 |
| 10057 | First, the contents of the @code{while} expression: |
| 10058 | |
| 10059 | @smallexample |
| 10060 | @group |
| 10061 | (while (<= row-number number-of-rows) ; @r{true-or-false-test} |
| 10062 | (setq total (+ total row-number)) |
| 10063 | (setq row-number (1+ row-number))) ; @r{incrementer} |
| 10064 | @end group |
| 10065 | @end smallexample |
| 10066 | |
| 10067 | Along with the @code{let} expression varlist, this very nearly |
| 10068 | completes the body of the function definition. However, it requires |
| 10069 | one final element, the need for which is somewhat subtle. |
| 10070 | |
| 10071 | The final touch is to place the variable @code{total} on a line by |
| 10072 | itself after the @code{while} expression. Otherwise, the value returned |
| 10073 | by the whole function is the value of the last expression that is |
| 10074 | evaluated in the body of the @code{let}, and this is the value |
| 10075 | returned by the @code{while}, which is always @code{nil}. |
| 10076 | |
| 10077 | This may not be evident at first sight. It almost looks as if the |
| 10078 | incrementing expression is the last expression of the whole function. |
| 10079 | But that expression is part of the body of the @code{while}; it is the |
| 10080 | last element of the list that starts with the symbol @code{while}. |
| 10081 | Moreover, the whole of the @code{while} loop is a list within the body |
| 10082 | of the @code{let}. |
| 10083 | |
| 10084 | @need 1250 |
| 10085 | In outline, the function will look like this: |
| 10086 | |
| 10087 | @smallexample |
| 10088 | @group |
| 10089 | (defun @var{name-of-function} (@var{argument-list}) |
| 10090 | "@var{documentation}@dots{}" |
| 10091 | (let (@var{varlist}) |
| 10092 | (while (@var{true-or-false-test}) |
| 10093 | @var{body-of-while}@dots{} ) |
| 10094 | @dots{} ) ; @r{Need final expression here.} |
| 10095 | @end group |
| 10096 | @end smallexample |
| 10097 | |
| 10098 | The result of evaluating the @code{let} is what is going to be returned |
| 10099 | by the @code{defun} since the @code{let} is not embedded within any |
| 10100 | containing list, except for the @code{defun} as a whole. However, if |
| 10101 | the @code{while} is the last element of the @code{let} expression, the |
| 10102 | function will always return @code{nil}. This is not what we want! |
| 10103 | Instead, what we want is the value of the variable @code{total}. This |
| 10104 | is returned by simply placing the symbol as the last element of the list |
| 10105 | starting with @code{let}. It gets evaluated after the preceding |
| 10106 | elements of the list are evaluated, which means it gets evaluated after |
| 10107 | it has been assigned the correct value for the total. |
| 10108 | |
| 10109 | It may be easier to see this by printing the list starting with |
| 10110 | @code{let} all on one line. This format makes it evident that the |
| 10111 | @var{varlist} and @code{while} expressions are the second and third |
| 10112 | elements of the list starting with @code{let}, and the @code{total} is |
| 10113 | the last element: |
| 10114 | |
| 10115 | @smallexample |
| 10116 | @group |
| 10117 | (let (@var{varlist}) (while (@var{true-or-false-test}) @var{body-of-while}@dots{} ) total) |
| 10118 | @end group |
| 10119 | @end smallexample |
| 10120 | |
| 10121 | @need 1200 |
| 10122 | Putting everything together, the @code{triangle} function definition |
| 10123 | looks like this: |
| 10124 | |
| 10125 | @smallexample |
| 10126 | @group |
| 10127 | (defun triangle (number-of-rows) ; @r{Version with} |
| 10128 | ; @r{ incrementing counter.} |
| 10129 | "Add up the number of pebbles in a triangle. |
| 10130 | The first row has one pebble, the second row two pebbles, |
| 10131 | the third row three pebbles, and so on. |
| 10132 | The argument is NUMBER-OF-ROWS." |
| 10133 | @end group |
| 10134 | @group |
| 10135 | (let ((total 0) |
| 10136 | (row-number 1)) |
| 10137 | (while (<= row-number number-of-rows) |
| 10138 | (setq total (+ total row-number)) |
| 10139 | (setq row-number (1+ row-number))) |
| 10140 | total)) |
| 10141 | @end group |
| 10142 | @end smallexample |
| 10143 | |
| 10144 | @need 1200 |
| 10145 | After you have installed @code{triangle} by evaluating the function, you |
| 10146 | can try it out. Here are two examples: |
| 10147 | |
| 10148 | @smallexample |
| 10149 | @group |
| 10150 | (triangle 4) |
| 10151 | |
| 10152 | (triangle 7) |
| 10153 | @end group |
| 10154 | @end smallexample |
| 10155 | |
| 10156 | @noindent |
| 10157 | The sum of the first four numbers is 10 and the sum of the first seven |
| 10158 | numbers is 28. |
| 10159 | |
| 10160 | @node Decrementing Loop, , Incrementing Loop, while |
| 10161 | @comment node-name, next, previous, up |
| 10162 | @subsection Loop with a Decrementing Counter |
| 10163 | |
| 10164 | Another common way to write a @code{while} loop is to write the test |
| 10165 | so that it determines whether a counter is greater than zero. So long |
| 10166 | as the counter is greater than zero, the loop is repeated. But when |
| 10167 | the counter is equal to or less than zero, the loop is stopped. For |
| 10168 | this to work, the counter has to start out greater than zero and then |
| 10169 | be made smaller and smaller by a form that is evaluated |
| 10170 | repeatedly. |
| 10171 | |
| 10172 | The test will be an expression such as @code{(> counter 0)} which |
| 10173 | returns @code{t} for true if the value of @code{counter} is greater |
| 10174 | than zero, and @code{nil} for false if the value of @code{counter} is |
| 10175 | equal to or less than zero. The expression that makes the number |
| 10176 | smaller and smaller can be a simple @code{setq} such as @code{(setq |
| 10177 | counter (1- counter))}, where @code{1-} is a built-in function in |
| 10178 | Emacs Lisp that subtracts 1 from its argument. |
| 10179 | |
| 10180 | @need 1250 |
| 10181 | The template for a decrementing @code{while} loop looks like this: |
| 10182 | |
| 10183 | @smallexample |
| 10184 | @group |
| 10185 | (while (> counter 0) ; @r{true-or-false-test} |
| 10186 | @var{body}@dots{} |
| 10187 | (setq counter (1- counter))) ; @r{decrementer} |
| 10188 | @end group |
| 10189 | @end smallexample |
| 10190 | |
| 10191 | @menu |
| 10192 | * Decrementing Example:: More pebbles on the beach. |
| 10193 | * Dec Example parts:: The parts of the function definition. |
| 10194 | * Dec Example altogether:: Putting the function definition together. |
| 10195 | @end menu |
| 10196 | |
| 10197 | @node Decrementing Example, Dec Example parts, Decrementing Loop, Decrementing Loop |
| 10198 | @unnumberedsubsubsec Example with decrementing counter |
| 10199 | |
| 10200 | To illustrate a loop with a decrementing counter, we will rewrite the |
| 10201 | @code{triangle} function so the counter decreases to zero. |
| 10202 | |
| 10203 | This is the reverse of the earlier version of the function. In this |
| 10204 | case, to find out how many pebbles are needed to make a triangle with |
| 10205 | 3 rows, add the number of pebbles in the third row, 3, to the number |
| 10206 | in the preceding row, 2, and then add the total of those two rows to |
| 10207 | the row that precedes them, which is 1. |
| 10208 | |
| 10209 | Likewise, to find the number of pebbles in a triangle with 7 rows, add |
| 10210 | the number of pebbles in the seventh row, 7, to the number in the |
| 10211 | preceding row, which is 6, and then add the total of those two rows to |
| 10212 | the row that precedes them, which is 5, and so on. As in the previous |
| 10213 | example, each addition only involves adding two numbers, the total of |
| 10214 | the rows already added up and the number of pebbles in the row that is |
| 10215 | being added to the total. This process of adding two numbers is |
| 10216 | repeated again and again until there are no more pebbles to add. |
| 10217 | |
| 10218 | We know how many pebbles to start with: the number of pebbles in the |
| 10219 | last row is equal to the number of rows. If the triangle has seven |
| 10220 | rows, the number of pebbles in the last row is 7. Likewise, we know how |
| 10221 | many pebbles are in the preceding row: it is one less than the number in |
| 10222 | the row. |
| 10223 | |
| 10224 | @node Dec Example parts, Dec Example altogether, Decrementing Example, Decrementing Loop |
| 10225 | @unnumberedsubsubsec The parts of the function definition |
| 10226 | |
| 10227 | We start with three variables: the total number of rows in the |
| 10228 | triangle; the number of pebbles in a row; and the total number of |
| 10229 | pebbles, which is what we want to calculate. These variables can be |
| 10230 | named @code{number-of-rows}, @code{number-of-pebbles-in-row}, and |
| 10231 | @code{total}, respectively. |
| 10232 | |
| 10233 | Both @code{total} and @code{number-of-pebbles-in-row} are used only |
| 10234 | inside the function and are declared with @code{let}. The initial |
| 10235 | value of @code{total} should, of course, be zero. However, the |
| 10236 | initial value of @code{number-of-pebbles-in-row} should be equal to |
| 10237 | the number of rows in the triangle, since the addition will start with |
| 10238 | the longest row. |
| 10239 | |
| 10240 | @need 1250 |
| 10241 | This means that the beginning of the @code{let} expression will look |
| 10242 | like this: |
| 10243 | |
| 10244 | @smallexample |
| 10245 | @group |
| 10246 | (let ((total 0) |
| 10247 | (number-of-pebbles-in-row number-of-rows)) |
| 10248 | @var{body}@dots{}) |
| 10249 | @end group |
| 10250 | @end smallexample |
| 10251 | |
| 10252 | The total number of pebbles can be found by repeatedly adding the number |
| 10253 | of pebbles in a row to the total already found, that is, by repeatedly |
| 10254 | evaluating the following expression: |
| 10255 | |
| 10256 | @smallexample |
| 10257 | (setq total (+ total number-of-pebbles-in-row)) |
| 10258 | @end smallexample |
| 10259 | |
| 10260 | @noindent |
| 10261 | After the @code{number-of-pebbles-in-row} is added to the @code{total}, |
| 10262 | the @code{number-of-pebbles-in-row} should be decremented by one, since |
| 10263 | the next time the loop repeats, the preceding row will be |
| 10264 | added to the total. |
| 10265 | |
| 10266 | The number of pebbles in a preceding row is one less than the number of |
| 10267 | pebbles in a row, so the built-in Emacs Lisp function @code{1-} can be |
| 10268 | used to compute the number of pebbles in the preceding row. This can be |
| 10269 | done with the following expression: |
| 10270 | |
| 10271 | @smallexample |
| 10272 | @group |
| 10273 | (setq number-of-pebbles-in-row |
| 10274 | (1- number-of-pebbles-in-row)) |
| 10275 | @end group |
| 10276 | @end smallexample |
| 10277 | |
| 10278 | Finally, we know that the @code{while} loop should stop making repeated |
| 10279 | additions when there are no pebbles in a row. So the test for |
| 10280 | the @code{while} loop is simply: |
| 10281 | |
| 10282 | @smallexample |
| 10283 | (while (> number-of-pebbles-in-row 0) |
| 10284 | @end smallexample |
| 10285 | |
| 10286 | @node Dec Example altogether, , Dec Example parts, Decrementing Loop |
| 10287 | @unnumberedsubsubsec Putting the function definition together |
| 10288 | |
| 10289 | We can put these expressions together to create a function definition |
| 10290 | that works. However, on examination, we find that one of the local |
| 10291 | variables is unneeded! |
| 10292 | |
| 10293 | @need 1250 |
| 10294 | The function definition looks like this: |
| 10295 | |
| 10296 | @smallexample |
| 10297 | @group |
| 10298 | ;;; @r{First subtractive version.} |
| 10299 | (defun triangle (number-of-rows) |
| 10300 | "Add up the number of pebbles in a triangle." |
| 10301 | (let ((total 0) |
| 10302 | (number-of-pebbles-in-row number-of-rows)) |
| 10303 | (while (> number-of-pebbles-in-row 0) |
| 10304 | (setq total (+ total number-of-pebbles-in-row)) |
| 10305 | (setq number-of-pebbles-in-row |
| 10306 | (1- number-of-pebbles-in-row))) |
| 10307 | total)) |
| 10308 | @end group |
| 10309 | @end smallexample |
| 10310 | |
| 10311 | As written, this function works. |
| 10312 | |
| 10313 | However, we do not need @code{number-of-pebbles-in-row}. |
| 10314 | |
| 10315 | @cindex Argument as local variable |
| 10316 | When the @code{triangle} function is evaluated, the symbol |
| 10317 | @code{number-of-rows} will be bound to a number, giving it an initial |
| 10318 | value. That number can be changed in the body of the function as if |
| 10319 | it were a local variable, without any fear that such a change will |
| 10320 | effect the value of the variable outside of the function. This is a |
| 10321 | very useful characteristic of Lisp; it means that the variable |
| 10322 | @code{number-of-rows} can be used anywhere in the function where |
| 10323 | @code{number-of-pebbles-in-row} is used. |
| 10324 | |
| 10325 | @need 800 |
| 10326 | Here is a second version of the function written a bit more cleanly: |
| 10327 | |
| 10328 | @smallexample |
| 10329 | @group |
| 10330 | (defun triangle (number) ; @r{Second version.} |
| 10331 | "Return sum of numbers 1 through NUMBER inclusive." |
| 10332 | (let ((total 0)) |
| 10333 | (while (> number 0) |
| 10334 | (setq total (+ total number)) |
| 10335 | (setq number (1- number))) |
| 10336 | total)) |
| 10337 | @end group |
| 10338 | @end smallexample |
| 10339 | |
| 10340 | In brief, a properly written @code{while} loop will consist of three parts: |
| 10341 | |
| 10342 | @enumerate |
| 10343 | @item |
| 10344 | A test that will return false after the loop has repeated itself the |
| 10345 | correct number of times. |
| 10346 | |
| 10347 | @item |
| 10348 | An expression the evaluation of which will return the value desired |
| 10349 | after being repeatedly evaluated. |
| 10350 | |
| 10351 | @item |
| 10352 | An expression to change the value passed to the true-or-false-test so |
| 10353 | that the test returns false after the loop has repeated itself the right |
| 10354 | number of times. |
| 10355 | @end enumerate |
| 10356 | |
| 10357 | @node dolist dotimes, Recursion, while, Loops & Recursion |
| 10358 | @comment node-name, next, previous, up |
| 10359 | @section Save your time: @code{dolist} and @code{dotimes} |
| 10360 | |
| 10361 | In addition to @code{while}, both @code{dolist} and @code{dotimes} |
| 10362 | provide for looping. Sometimes these are quicker to write than the |
| 10363 | equivalent @code{while} loop. Both are Lisp macros. (@xref{Macros, , |
| 10364 | Macros, elisp, The GNU Emacs Lisp Reference Manual}. ) |
| 10365 | |
| 10366 | @code{dolist} works like a @code{while} loop that `@sc{cdr}s down a |
| 10367 | list': @code{dolist} automatically shortens the list each time it |
| 10368 | loops---takes the @sc{cdr} of the list---and binds the @sc{car} of |
| 10369 | each shorter version of the list to the first of its arguments. |
| 10370 | |
| 10371 | @code{dotimes} loops a specific number of times: you specify the number. |
| 10372 | |
| 10373 | @menu |
| 10374 | * dolist:: |
| 10375 | * dotimes:: |
| 10376 | @end menu |
| 10377 | |
| 10378 | @node dolist, dotimes, dolist dotimes, dolist dotimes |
| 10379 | @unnumberedsubsubsec The @code{dolist} Macro |
| 10380 | @findex dolist |
| 10381 | |
| 10382 | Suppose, for example, you want to reverse a list, so that |
| 10383 | ``first'' ``second'' ``third'' becomes ``third'' ``second'' ``first''. |
| 10384 | |
| 10385 | @need 1250 |
| 10386 | In practice, you would use the @code{reverse} function, like this: |
| 10387 | |
| 10388 | @smallexample |
| 10389 | @group |
| 10390 | (setq animals '(gazelle giraffe lion tiger)) |
| 10391 | |
| 10392 | (reverse animals) |
| 10393 | @end group |
| 10394 | @end smallexample |
| 10395 | |
| 10396 | @need 800 |
| 10397 | @noindent |
| 10398 | Here is how you could reverse the list using a @code{while} loop: |
| 10399 | |
| 10400 | @smallexample |
| 10401 | @group |
| 10402 | (setq animals '(gazelle giraffe lion tiger)) |
| 10403 | |
| 10404 | (defun reverse-list-with-while (list) |
| 10405 | "Using while, reverse the order of LIST." |
| 10406 | (let (value) ; make sure list starts empty |
| 10407 | (while list |
| 10408 | (setq value (cons (car list) value)) |
| 10409 | (setq list (cdr list))) |
| 10410 | value)) |
| 10411 | |
| 10412 | (reverse-list-with-while animals) |
| 10413 | @end group |
| 10414 | @end smallexample |
| 10415 | |
| 10416 | @need 800 |
| 10417 | @noindent |
| 10418 | And here is how you could use the @code{dolist} macro: |
| 10419 | |
| 10420 | @smallexample |
| 10421 | @group |
| 10422 | (setq animals '(gazelle giraffe lion tiger)) |
| 10423 | |
| 10424 | (defun reverse-list-with-dolist (list) |
| 10425 | "Using dolist, reverse the order of LIST." |
| 10426 | (let (value) ; make sure list starts empty |
| 10427 | (dolist (element list value) |
| 10428 | (setq value (cons element value))))) |
| 10429 | |
| 10430 | (reverse-list-with-dolist animals) |
| 10431 | @end group |
| 10432 | @end smallexample |
| 10433 | |
| 10434 | @need 1250 |
| 10435 | @noindent |
| 10436 | In Info, you can place your cursor after the closing parenthesis of |
| 10437 | each expression and type @kbd{C-x C-e}; in each case, you should see |
| 10438 | |
| 10439 | @smallexample |
| 10440 | (tiger lion giraffe gazelle) |
| 10441 | @end smallexample |
| 10442 | |
| 10443 | @noindent |
| 10444 | in the echo area. |
| 10445 | |
| 10446 | For this example, the existing @code{reverse} function is obviously best. |
| 10447 | The @code{while} loop is just like our first example (@pxref{Loop |
| 10448 | Example, , A @code{while} Loop and a List}). The @code{while} first |
| 10449 | checks whether the list has elements; if so, it constructs a new list |
| 10450 | by adding the first element of the list to the existing list (which in |
| 10451 | the first iteration of the loop is @code{nil}). Since the second |
| 10452 | element is prepended in front of the first element, and the third |
| 10453 | element is prepended in front of the second element, the list is reversed. |
| 10454 | |
| 10455 | In the expression using a @code{while} loop, |
| 10456 | the @w{@code{(setq list (cdr list))}} |
| 10457 | expression shortens the list, so the @code{while} loop eventually |
| 10458 | stops. In addition, it provides the @code{cons} expression with a new |
| 10459 | first element by creating a new and shorter list at each repetition of |
| 10460 | the loop. |
| 10461 | |
| 10462 | The @code{dolist} expression does very much the same as the |
| 10463 | @code{while} expression, except that the @code{dolist} macro does some |
| 10464 | of the work you have to do when writing a @code{while} expression. |
| 10465 | |
| 10466 | Like a @code{while} loop, a @code{dolist} loops. What is different is |
| 10467 | that it automatically shortens the list each time it loops --- it |
| 10468 | `@sc{cdr}s down the list' on its own --- and it automatically binds |
| 10469 | the @sc{car} of each shorter version of the list to the first of its |
| 10470 | arguments. |
| 10471 | |
| 10472 | In the example, the @sc{car} of each shorter version of the list is |
| 10473 | referred to using the symbol @samp{element}, the list itself is called |
| 10474 | @samp{list}, and the value returned is called @samp{value}. The |
| 10475 | remainder of the @code{dolist} expression is the body. |
| 10476 | |
| 10477 | The @code{dolist} expression binds the @sc{car} of each shorter |
| 10478 | version of the list to @code{element} and then evaluates the body of |
| 10479 | the expression; and repeats the loop. The result is returned in |
| 10480 | @code{value}. |
| 10481 | |
| 10482 | @node dotimes, , dolist, dolist dotimes |
| 10483 | @unnumberedsubsubsec The @code{dotimes} Macro |
| 10484 | @findex dotimes |
| 10485 | |
| 10486 | The @code{dotimes} macro is similar to @code{dolist}, except that it |
| 10487 | loops a specific number of times. |
| 10488 | |
| 10489 | The first argument to @code{dotimes} is assigned the numbers 0, 1, 2 |
| 10490 | and so forth each time around the loop, and the value of the third |
| 10491 | argument is returned. You need to provide the value of the second |
| 10492 | argument, which is how many times the macro loops. |
| 10493 | |
| 10494 | @need 1250 |
| 10495 | For example, the following binds the numbers from 0 up to, but not |
| 10496 | including, the number 3 to the first argument, @var{number}, and then |
| 10497 | constructs a list of the three numbers. (The first number is 0, the |
| 10498 | second number is 1, and the third number is 2; this makes a total of |
| 10499 | three numbers in all, starting with zero as the first number.) |
| 10500 | |
| 10501 | @smallexample |
| 10502 | @group |
| 10503 | (let (value) ; otherwise a value is a void variable |
| 10504 | (dotimes (number 3 value) |
| 10505 | (setq value (cons number value)))) |
| 10506 | |
| 10507 | @result{} (2 1 0) |
| 10508 | @end group |
| 10509 | @end smallexample |
| 10510 | |
| 10511 | @noindent |
| 10512 | @code{dotimes} returns @code{value}, so the way to use |
| 10513 | @code{dotimes} is to operate on some expression @var{number} number of |
| 10514 | times and then return the result, either as a list or an atom. |
| 10515 | |
| 10516 | @need 1250 |
| 10517 | Here is an example of a @code{defun} that uses @code{dotimes} to add |
| 10518 | up the number of pebbles in a triangle. |
| 10519 | |
| 10520 | @smallexample |
| 10521 | @group |
| 10522 | (defun triangle-using-dotimes (number-of-rows) |
| 10523 | "Using dotimes, add up the number of pebbles in a triangle." |
| 10524 | (let ((total 0)) ; otherwise a total is a void variable |
| 10525 | (dotimes (number number-of-rows total) |
| 10526 | (setq total (+ total (1+ number)))))) |
| 10527 | |
| 10528 | (triangle-using-dotimes 4) |
| 10529 | @end group |
| 10530 | @end smallexample |
| 10531 | |
| 10532 | @node Recursion, Looping exercise, dolist dotimes, Loops & Recursion |
| 10533 | @comment node-name, next, previous, up |
| 10534 | @section Recursion |
| 10535 | @cindex Recursion |
| 10536 | |
| 10537 | A recursive function contains code that tells the Lisp interpreter to |
| 10538 | call a program that runs exactly like itself, but with slightly |
| 10539 | different arguments. The code runs exactly the same because it has |
| 10540 | the same name. However, even though the program has the same name, it |
| 10541 | is not the same entity. It is different. In the jargon, it is a |
| 10542 | different `instance'. |
| 10543 | |
| 10544 | Eventually, if the program is written correctly, the `slightly |
| 10545 | different arguments' will become sufficiently different from the first |
| 10546 | arguments that the final instance will stop. |
| 10547 | |
| 10548 | @menu |
| 10549 | * Building Robots:: Same model, different serial number ... |
| 10550 | * Recursive Definition Parts:: Walk until you stop ... |
| 10551 | * Recursion with list:: Using a list as the test whether to recurse. |
| 10552 | * Recursive triangle function:: |
| 10553 | * Recursion with cond:: |
| 10554 | * Recursive Patterns:: Often used templates. |
| 10555 | * No Deferment:: Don't store up work ... |
| 10556 | * No deferment solution:: |
| 10557 | @end menu |
| 10558 | |
| 10559 | @node Building Robots, Recursive Definition Parts, Recursion, Recursion |
| 10560 | @comment node-name, next, previous, up |
| 10561 | @subsection Building Robots: Extending the Metaphor |
| 10562 | @cindex Building robots |
| 10563 | @cindex Robots, building |
| 10564 | |
| 10565 | It is sometimes helpful to think of a running program as a robot that |
| 10566 | does a job. In doing its job, a recursive function calls on a second |
| 10567 | robot to help it. The second robot is identical to the first in every |
| 10568 | way, except that the second robot helps the first and has been |
| 10569 | passed different arguments than the first. |
| 10570 | |
| 10571 | In a recursive function, the second robot may call a third; and the |
| 10572 | third may call a fourth, and so on. Each of these is a different |
| 10573 | entity; but all are clones. |
| 10574 | |
| 10575 | Since each robot has slightly different instructions---the arguments |
| 10576 | will differ from one robot to the next---the last robot should know |
| 10577 | when to stop. |
| 10578 | |
| 10579 | Let's expand on the metaphor in which a computer program is a robot. |
| 10580 | |
| 10581 | A function definition provides the blueprints for a robot. When you |
| 10582 | install a function definition, that is, when you evaluate a |
| 10583 | @code{defun} special form, you install the necessary equipment to |
| 10584 | build robots. It is as if you were in a factory, setting up an |
| 10585 | assembly line. Robots with the same name are built according to the |
| 10586 | same blueprints. So they have, as it were, the same `model number', |
| 10587 | but a different `serial number'. |
| 10588 | |
| 10589 | We often say that a recursive function `calls itself'. What we mean |
| 10590 | is that the instructions in a recursive function cause the Lisp |
| 10591 | interpreter to run a different function that has the same name and |
| 10592 | does the same job as the first, but with different arguments. |
| 10593 | |
| 10594 | It is important that the arguments differ from one instance to the |
| 10595 | next; otherwise, the process will never stop. |
| 10596 | |
| 10597 | @node Recursive Definition Parts, Recursion with list, Building Robots, Recursion |
| 10598 | @comment node-name, next, previous, up |
| 10599 | @subsection The Parts of a Recursive Definition |
| 10600 | @cindex Parts of a Recursive Definition |
| 10601 | @cindex Recursive Definition Parts |
| 10602 | |
| 10603 | A recursive function typically contains a conditional expression which |
| 10604 | has three parts: |
| 10605 | |
| 10606 | @enumerate |
| 10607 | @item |
| 10608 | A true-or-false-test that determines whether the function is called |
| 10609 | again, here called the @dfn{do-again-test}. |
| 10610 | |
| 10611 | @item |
| 10612 | The name of the function. When this name is called, a new instance of |
| 10613 | the function---a new robot, as it were---is created and told what to do. |
| 10614 | |
| 10615 | @item |
| 10616 | An expression that returns a different value each time the function is |
| 10617 | called, here called the @dfn{next-step-expression}. Consequently, the |
| 10618 | argument (or arguments) passed to the new instance of the function |
| 10619 | will be different from that passed to the previous instance. This |
| 10620 | causes the conditional expression, the @dfn{do-again-test}, to test |
| 10621 | false after the correct number of repetitions. |
| 10622 | @end enumerate |
| 10623 | |
| 10624 | Recursive functions can be much simpler than any other kind of |
| 10625 | function. Indeed, when people first start to use them, they often look |
| 10626 | so mysteriously simple as to be incomprehensible. Like riding a |
| 10627 | bicycle, reading a recursive function definition takes a certain knack |
| 10628 | which is hard at first but then seems simple. |
| 10629 | |
| 10630 | @need 1200 |
| 10631 | There are several different common recursive patterns. A very simple |
| 10632 | pattern looks like this: |
| 10633 | |
| 10634 | @smallexample |
| 10635 | @group |
| 10636 | (defun @var{name-of-recursive-function} (@var{argument-list}) |
| 10637 | "@var{documentation}@dots{}" |
| 10638 | (if @var{do-again-test} |
| 10639 | @var{body}@dots{} |
| 10640 | (@var{name-of-recursive-function} |
| 10641 | @var{next-step-expression}))) |
| 10642 | @end group |
| 10643 | @end smallexample |
| 10644 | |
| 10645 | Each time a recursive function is evaluated, a new instance of it is |
| 10646 | created and told what to do. The arguments tell the instance what to do. |
| 10647 | |
| 10648 | An argument is bound to the value of the next-step-expression. Each |
| 10649 | instance runs with a different value of the next-step-expression. |
| 10650 | |
| 10651 | The value in the next-step-expression is used in the do-again-test. |
| 10652 | |
| 10653 | The value returned by the next-step-expression is passed to the new |
| 10654 | instance of the function, which evaluates it (or some |
| 10655 | transmogrification of it) to determine whether to continue or stop. |
| 10656 | The next-step-expression is designed so that the do-again-test returns |
| 10657 | false when the function should no longer be repeated. |
| 10658 | |
| 10659 | The do-again-test is sometimes called the @dfn{stop condition}, |
| 10660 | since it stops the repetitions when it tests false. |
| 10661 | |
| 10662 | @node Recursion with list, Recursive triangle function, Recursive Definition Parts, Recursion |
| 10663 | @comment node-name, next, previous, up |
| 10664 | @subsection Recursion with a List |
| 10665 | |
| 10666 | The example of a @code{while} loop that printed the elements of a list |
| 10667 | of numbers can be written recursively. Here is the code, including |
| 10668 | an expression to set the value of the variable @code{animals} to a list. |
| 10669 | |
| 10670 | If you are using Emacs 20 or before, this example must be copied to |
| 10671 | the @file{*scratch*} buffer and each expression must be evaluated |
| 10672 | there. Use @kbd{C-u C-x C-e} to evaluate the |
| 10673 | @code{(print-elements-recursively animals)} expression so that the |
| 10674 | results are printed in the buffer; otherwise the Lisp interpreter will |
| 10675 | try to squeeze the results into the one line of the echo area. |
| 10676 | |
| 10677 | Also, place your cursor immediately after the last closing parenthesis |
| 10678 | of the @code{print-elements-recursively} function, before the comment. |
| 10679 | Otherwise, the Lisp interpreter will try to evaluate the comment. |
| 10680 | |
| 10681 | If you are using Emacs 21 or later, you can evaluate this expression |
| 10682 | directly in Info. |
| 10683 | |
| 10684 | @findex print-elements-recursively |
| 10685 | @smallexample |
| 10686 | @group |
| 10687 | (setq animals '(gazelle giraffe lion tiger)) |
| 10688 | |
| 10689 | (defun print-elements-recursively (list) |
| 10690 | "Print each element of LIST on a line of its own. |
| 10691 | Uses recursion." |
| 10692 | (if list ; @r{do-again-test} |
| 10693 | (progn |
| 10694 | (print (car list)) ; @r{body} |
| 10695 | (print-elements-recursively ; @r{recursive call} |
| 10696 | (cdr list))))) ; @r{next-step-expression} |
| 10697 | |
| 10698 | (print-elements-recursively animals) |
| 10699 | @end group |
| 10700 | @end smallexample |
| 10701 | |
| 10702 | The @code{print-elements-recursively} function first tests whether |
| 10703 | there is any content in the list; if there is, the function prints the |
| 10704 | first element of the list, the @sc{car} of the list. Then the |
| 10705 | function `invokes itself', but gives itself as its argument, not the |
| 10706 | whole list, but the second and subsequent elements of the list, the |
| 10707 | @sc{cdr} of the list. |
| 10708 | |
| 10709 | Put another way, if the list is not empty, the function invokes |
| 10710 | another instance of code that is similar to the initial code, but is a |
| 10711 | different thread of execution, with different arguments than the first |
| 10712 | instance. |
| 10713 | |
| 10714 | Put in yet another way, if the list is not empty, the first robot |
| 10715 | assemblies a second robot and tells it what to do; the second robot is |
| 10716 | a different individual from the first, but is the same model. |
| 10717 | |
| 10718 | When the second evaluation occurs, the @code{if} expression is |
| 10719 | evaluated and if true, prints the first element of the list it |
| 10720 | receives as its argument (which is the second element of the original |
| 10721 | list). Then the function `calls itself' with the @sc{cdr} of the list |
| 10722 | it is invoked with, which (the second time around) is the @sc{cdr} of |
| 10723 | the @sc{cdr} of the original list. |
| 10724 | |
| 10725 | Note that although we say that the function `calls itself', what we |
| 10726 | mean is that the Lisp interpreter assembles and instructs a new |
| 10727 | instance of the program. The new instance is a clone of the first, |
| 10728 | but is a separate individual. |
| 10729 | |
| 10730 | Each time the function `invokes itself', it invokes itself on a |
| 10731 | shorter version of the original list. It creates a new instance that |
| 10732 | works on a shorter list. |
| 10733 | |
| 10734 | Eventually, the function invokes itself on an empty list. It creates |
| 10735 | a new instance whose argument is @code{nil}. The conditional expression |
| 10736 | tests the value of @code{list}. Since the value of @code{list} is |
| 10737 | @code{nil}, the @code{if} expression tests false so the then-part is |
| 10738 | not evaluated. The function as a whole then returns @code{nil}. |
| 10739 | |
| 10740 | @need 1200 |
| 10741 | When you evaluate @code{(print-elements-recursively animals)} in the |
| 10742 | @file{*scratch*} buffer, you see this result: |
| 10743 | |
| 10744 | @smallexample |
| 10745 | @group |
| 10746 | gazelle |
| 10747 | |
| 10748 | giraffe |
| 10749 | |
| 10750 | lion |
| 10751 | |
| 10752 | tiger |
| 10753 | nil |
| 10754 | @end group |
| 10755 | @end smallexample |
| 10756 | |
| 10757 | @node Recursive triangle function, Recursion with cond, Recursion with list, Recursion |
| 10758 | @comment node-name, next, previous, up |
| 10759 | @subsection Recursion in Place of a Counter |
| 10760 | @findex triangle-recursively |
| 10761 | |
| 10762 | @need 1200 |
| 10763 | The @code{triangle} function described in a previous section can also |
| 10764 | be written recursively. It looks like this: |
| 10765 | |
| 10766 | @smallexample |
| 10767 | @group |
| 10768 | (defun triangle-recursively (number) |
| 10769 | "Return the sum of the numbers 1 through NUMBER inclusive. |
| 10770 | Uses recursion." |
| 10771 | (if (= number 1) ; @r{do-again-test} |
| 10772 | 1 ; @r{then-part} |
| 10773 | (+ number ; @r{else-part} |
| 10774 | (triangle-recursively ; @r{recursive call} |
| 10775 | (1- number))))) ; @r{next-step-expression} |
| 10776 | |
| 10777 | (triangle-recursively 7) |
| 10778 | @end group |
| 10779 | @end smallexample |
| 10780 | |
| 10781 | @noindent |
| 10782 | You can install this function by evaluating it and then try it by |
| 10783 | evaluating @code{(triangle-recursively 7)}. (Remember to put your |
| 10784 | cursor immediately after the last parenthesis of the function |
| 10785 | definition, before the comment.) The function evaluates to 28. |
| 10786 | |
| 10787 | To understand how this function works, let's consider what happens in the |
| 10788 | various cases when the function is passed 1, 2, 3, or 4 as the value of |
| 10789 | its argument. |
| 10790 | |
| 10791 | @menu |
| 10792 | * Recursive Example arg of 1 or 2:: |
| 10793 | * Recursive Example arg of 3 or 4:: |
| 10794 | @end menu |
| 10795 | |
| 10796 | @node Recursive Example arg of 1 or 2, Recursive Example arg of 3 or 4, Recursive triangle function, Recursive triangle function |
| 10797 | @ifnottex |
| 10798 | @unnumberedsubsubsec An argument of 1 or 2 |
| 10799 | @end ifnottex |
| 10800 | |
| 10801 | First, what happens if the value of the argument is 1? |
| 10802 | |
| 10803 | The function has an @code{if} expression after the documentation |
| 10804 | string. It tests whether the value of @code{number} is equal to 1; if |
| 10805 | so, Emacs evaluates the then-part of the @code{if} expression, which |
| 10806 | returns the number 1 as the value of the function. (A triangle with |
| 10807 | one row has one pebble in it.) |
| 10808 | |
| 10809 | Suppose, however, that the value of the argument is 2. In this case, |
| 10810 | Emacs evaluates the else-part of the @code{if} expression. |
| 10811 | |
| 10812 | @need 1200 |
| 10813 | The else-part consists of an addition, the recursive call to |
| 10814 | @code{triangle-recursively} and a decrementing action; and it looks like |
| 10815 | this: |
| 10816 | |
| 10817 | @smallexample |
| 10818 | (+ number (triangle-recursively (1- number))) |
| 10819 | @end smallexample |
| 10820 | |
| 10821 | When Emacs evaluates this expression, the innermost expression is |
| 10822 | evaluated first; then the other parts in sequence. Here are the steps |
| 10823 | in detail: |
| 10824 | |
| 10825 | @table @i |
| 10826 | @item Step 1 @w{ } Evaluate the innermost expression. |
| 10827 | |
| 10828 | The innermost expression is @code{(1- number)} so Emacs decrements the |
| 10829 | value of @code{number} from 2 to 1. |
| 10830 | |
| 10831 | @item Step 2 @w{ } Evaluate the @code{triangle-recursively} function. |
| 10832 | |
| 10833 | The Lisp interpreter creates an individual instance of |
| 10834 | @code{triangle-recursively}. It does not matter that this function is |
| 10835 | contained within itself. Emacs passes the result Step 1 as the |
| 10836 | argument used by this instance of the @code{triangle-recursively} |
| 10837 | function |
| 10838 | |
| 10839 | In this case, Emacs evaluates @code{triangle-recursively} with an |
| 10840 | argument of 1. This means that this evaluation of |
| 10841 | @code{triangle-recursively} returns 1. |
| 10842 | |
| 10843 | @item Step 3 @w{ } Evaluate the value of @code{number}. |
| 10844 | |
| 10845 | The variable @code{number} is the second element of the list that |
| 10846 | starts with @code{+}; its value is 2. |
| 10847 | |
| 10848 | @item Step 4 @w{ } Evaluate the @code{+} expression. |
| 10849 | |
| 10850 | The @code{+} expression receives two arguments, the first |
| 10851 | from the evaluation of @code{number} (Step 3) and the second from the |
| 10852 | evaluation of @code{triangle-recursively} (Step 2). |
| 10853 | |
| 10854 | The result of the addition is the sum of 2 plus 1, and the number 3 is |
| 10855 | returned, which is correct. A triangle with two rows has three |
| 10856 | pebbles in it. |
| 10857 | @end table |
| 10858 | |
| 10859 | @node Recursive Example arg of 3 or 4, , Recursive Example arg of 1 or 2, Recursive triangle function |
| 10860 | @unnumberedsubsubsec An argument of 3 or 4 |
| 10861 | |
| 10862 | Suppose that @code{triangle-recursively} is called with an argument of |
| 10863 | 3. |
| 10864 | |
| 10865 | @table @i |
| 10866 | @item Step 1 @w{ } Evaluate the do-again-test. |
| 10867 | |
| 10868 | The @code{if} expression is evaluated first. This is the do-again |
| 10869 | test and returns false, so the else-part of the @code{if} expression |
| 10870 | is evaluated. (Note that in this example, the do-again-test causes |
| 10871 | the function to call itself when it tests false, not when it tests |
| 10872 | true.) |
| 10873 | |
| 10874 | @item Step 2 @w{ } Evaluate the innermost expression of the else-part. |
| 10875 | |
| 10876 | The innermost expression of the else-part is evaluated, which decrements |
| 10877 | 3 to 2. This is the next-step-expression. |
| 10878 | |
| 10879 | @item Step 3 @w{ } Evaluate the @code{triangle-recursively} function. |
| 10880 | |
| 10881 | The number 2 is passed to the @code{triangle-recursively} function. |
| 10882 | |
| 10883 | We know what happens when Emacs evaluates @code{triangle-recursively} with |
| 10884 | an argument of 2. After going through the sequence of actions described |
| 10885 | earlier, it returns a value of 3. So that is what will happen here. |
| 10886 | |
| 10887 | @item Step 4 @w{ } Evaluate the addition. |
| 10888 | |
| 10889 | 3 will be passed as an argument to the addition and will be added to the |
| 10890 | number with which the function was called, which is 3. |
| 10891 | @end table |
| 10892 | |
| 10893 | @noindent |
| 10894 | The value returned by the function as a whole will be 6. |
| 10895 | |
| 10896 | Now that we know what will happen when @code{triangle-recursively} is |
| 10897 | called with an argument of 3, it is evident what will happen if it is |
| 10898 | called with an argument of 4: |
| 10899 | |
| 10900 | @quotation |
| 10901 | @need 800 |
| 10902 | In the recursive call, the evaluation of |
| 10903 | |
| 10904 | @smallexample |
| 10905 | (triangle-recursively (1- 4)) |
| 10906 | @end smallexample |
| 10907 | |
| 10908 | @need 800 |
| 10909 | @noindent |
| 10910 | will return the value of evaluating |
| 10911 | |
| 10912 | @smallexample |
| 10913 | (triangle-recursively 3) |
| 10914 | @end smallexample |
| 10915 | |
| 10916 | @noindent |
| 10917 | which is 6 and this value will be added to 4 by the addition in the |
| 10918 | third line. |
| 10919 | @end quotation |
| 10920 | |
| 10921 | @noindent |
| 10922 | The value returned by the function as a whole will be 10. |
| 10923 | |
| 10924 | Each time @code{triangle-recursively} is evaluated, it evaluates a |
| 10925 | version of itself---a different instance of itself---with a smaller |
| 10926 | argument, until the argument is small enough so that it does not |
| 10927 | evaluate itself. |
| 10928 | |
| 10929 | Note that this particular design for a recursive function |
| 10930 | requires that operations be deferred. |
| 10931 | |
| 10932 | Before @code{(triangle-recursively 7)} can calculate its answer, it |
| 10933 | must call @code{(triangle-recursively 6)}; and before |
| 10934 | @code{(triangle-recursively 6)} can calculate its answer, it must call |
| 10935 | @code{(triangle-recursively 5)}; and so on. That is to say, the |
| 10936 | calculation that @code{(triangle-recursively 7)} makes must be |
| 10937 | deferred until @code{(triangle-recursively 6)} makes its calculation; |
| 10938 | and @code{(triangle-recursively 6)} must defer until |
| 10939 | @code{(triangle-recursively 5)} completes; and so on. |
| 10940 | |
| 10941 | If each of these instances of @code{triangle-recursively} are thought |
| 10942 | of as different robots, the first robot must wait for the second to |
| 10943 | complete its job, which must wait until the third completes, and so |
| 10944 | on. |
| 10945 | |
| 10946 | There is a way around this kind of waiting, which we will discuss in |
| 10947 | @ref{No Deferment, , Recursion without Deferments}. |
| 10948 | |
| 10949 | @node Recursion with cond, Recursive Patterns, Recursive triangle function, Recursion |
| 10950 | @comment node-name, next, previous, up |
| 10951 | @subsection Recursion Example Using @code{cond} |
| 10952 | @findex cond |
| 10953 | |
| 10954 | The version of @code{triangle-recursively} described earlier is written |
| 10955 | with the @code{if} special form. It can also be written using another |
| 10956 | special form called @code{cond}. The name of the special form |
| 10957 | @code{cond} is an abbreviation of the word @samp{conditional}. |
| 10958 | |
| 10959 | Although the @code{cond} special form is not used as often in the |
| 10960 | Emacs Lisp sources as @code{if}, it is used often enough to justify |
| 10961 | explaining it. |
| 10962 | |
| 10963 | @need 800 |
| 10964 | The template for a @code{cond} expression looks like this: |
| 10965 | |
| 10966 | @smallexample |
| 10967 | @group |
| 10968 | (cond |
| 10969 | @var{body}@dots{}) |
| 10970 | @end group |
| 10971 | @end smallexample |
| 10972 | |
| 10973 | @noindent |
| 10974 | where the @var{body} is a series of lists. |
| 10975 | |
| 10976 | @need 800 |
| 10977 | Written out more fully, the template looks like this: |
| 10978 | |
| 10979 | @smallexample |
| 10980 | @group |
| 10981 | (cond |
| 10982 | (@var{first-true-or-false-test} @var{first-consequent}) |
| 10983 | (@var{second-true-or-false-test} @var{second-consequent}) |
| 10984 | (@var{third-true-or-false-test} @var{third-consequent}) |
| 10985 | @dots{}) |
| 10986 | @end group |
| 10987 | @end smallexample |
| 10988 | |
| 10989 | When the Lisp interpreter evaluates the @code{cond} expression, it |
| 10990 | evaluates the first element (the @sc{car} or true-or-false-test) of |
| 10991 | the first expression in a series of expressions within the body of the |
| 10992 | @code{cond}. |
| 10993 | |
| 10994 | If the true-or-false-test returns @code{nil} the rest of that |
| 10995 | expression, the consequent, is skipped and the true-or-false-test of the |
| 10996 | next expression is evaluated. When an expression is found whose |
| 10997 | true-or-false-test returns a value that is not @code{nil}, the |
| 10998 | consequent of that expression is evaluated. The consequent can be one |
| 10999 | or more expressions. If the consequent consists of more than one |
| 11000 | expression, the expressions are evaluated in sequence and the value of |
| 11001 | the last one is returned. If the expression does not have a consequent, |
| 11002 | the value of the true-or-false-test is returned. |
| 11003 | |
| 11004 | If none of the true-or-false-tests test true, the @code{cond} expression |
| 11005 | returns @code{nil}. |
| 11006 | |
| 11007 | @need 1250 |
| 11008 | Written using @code{cond}, the @code{triangle} function looks like this: |
| 11009 | |
| 11010 | @smallexample |
| 11011 | @group |
| 11012 | (defun triangle-using-cond (number) |
| 11013 | (cond ((<= number 0) 0) |
| 11014 | ((= number 1) 1) |
| 11015 | ((> number 1) |
| 11016 | (+ number (triangle-using-cond (1- number)))))) |
| 11017 | @end group |
| 11018 | @end smallexample |
| 11019 | |
| 11020 | @noindent |
| 11021 | In this example, the @code{cond} returns 0 if the number is less than or |
| 11022 | equal to 0, it returns 1 if the number is 1 and it evaluates @code{(+ |
| 11023 | number (triangle-using-cond (1- number)))} if the number is greater than |
| 11024 | 1. |
| 11025 | |
| 11026 | @node Recursive Patterns, No Deferment, Recursion with cond, Recursion |
| 11027 | @comment node-name, next, previous, up |
| 11028 | @subsection Recursive Patterns |
| 11029 | @cindex Recursive Patterns |
| 11030 | |
| 11031 | Here are three common recursive patterns. Each involves a list. |
| 11032 | Recursion does not need to involve lists, but Lisp is designed for lists |
| 11033 | and this provides a sense of its primal capabilities. |
| 11034 | |
| 11035 | @menu |
| 11036 | * Every:: |
| 11037 | * Accumulate:: |
| 11038 | * Keep:: |
| 11039 | @end menu |
| 11040 | |
| 11041 | @node Every, Accumulate, Recursive Patterns, Recursive Patterns |
| 11042 | @comment node-name, next, previous, up |
| 11043 | @unnumberedsubsubsec Recursive Pattern: @emph{every} |
| 11044 | @cindex Every, type of recursive pattern |
| 11045 | @cindex Recursive pattern: every |
| 11046 | |
| 11047 | In the @code{every} recursive pattern, an action is performed on every |
| 11048 | element of a list. |
| 11049 | |
| 11050 | @need 1500 |
| 11051 | The basic pattern is: |
| 11052 | |
| 11053 | @itemize @bullet |
| 11054 | @item |
| 11055 | If a list be empty, return @code{nil}. |
| 11056 | @item |
| 11057 | Else, act on the beginning of the list (the @sc{car} of the list) |
| 11058 | @itemize @minus |
| 11059 | @item |
| 11060 | through a recursive call by the function on the rest (the |
| 11061 | @sc{cdr}) of the list, |
| 11062 | @item |
| 11063 | and, optionally, combine the acted-on element, using @code{cons}, |
| 11064 | with the results of acting on the rest. |
| 11065 | @end itemize |
| 11066 | @end itemize |
| 11067 | |
| 11068 | @need 1500 |
| 11069 | Here is example: |
| 11070 | |
| 11071 | @smallexample |
| 11072 | @group |
| 11073 | (defun square-each (numbers-list) |
| 11074 | "Square each of a NUMBERS LIST, recursively." |
| 11075 | (if (not numbers-list) ; do-again-test |
| 11076 | nil |
| 11077 | (cons |
| 11078 | (* (car numbers-list) (car numbers-list)) |
| 11079 | (square-each (cdr numbers-list))))) ; next-step-expression |
| 11080 | @end group |
| 11081 | |
| 11082 | @group |
| 11083 | (square-each '(1 2 3)) |
| 11084 | @result{} (1 4 9) |
| 11085 | @end group |
| 11086 | @end smallexample |
| 11087 | |
| 11088 | @need 1200 |
| 11089 | @noindent |
| 11090 | If @code{numbers-list} is empty, do nothing. But if it has content, |
| 11091 | construct a list combining the square of the first number in the list |
| 11092 | with the result of the recursive call. |
| 11093 | |
| 11094 | (The example follows the pattern exactly: @code{nil} is returned if |
| 11095 | the numbers' list is empty. In practice, you would write the |
| 11096 | conditional so it carries out the action when the numbers' list is not |
| 11097 | empty.) |
| 11098 | |
| 11099 | The @code{print-elements-recursively} function (@pxref{Recursion with |
| 11100 | list, , Recursion with a List}) is another example of an @code{every} |
| 11101 | pattern, except in this case, rather than bring the results together |
| 11102 | using @code{cons}, we print each element of output. |
| 11103 | |
| 11104 | @need 1250 |
| 11105 | The @code{print-elements-recursively} function looks like this: |
| 11106 | |
| 11107 | @smallexample |
| 11108 | @group |
| 11109 | (setq animals '(gazelle giraffe lion tiger)) |
| 11110 | @end group |
| 11111 | |
| 11112 | @group |
| 11113 | (defun print-elements-recursively (list) |
| 11114 | "Print each element of LIST on a line of its own. |
| 11115 | Uses recursion." |
| 11116 | (if list ; @r{do-again-test} |
| 11117 | (progn |
| 11118 | (print (car list)) ; @r{body} |
| 11119 | (print-elements-recursively ; @r{recursive call} |
| 11120 | (cdr list))))) ; @r{next-step-expression} |
| 11121 | |
| 11122 | (print-elements-recursively animals) |
| 11123 | @end group |
| 11124 | @end smallexample |
| 11125 | |
| 11126 | @need 1500 |
| 11127 | The pattern for @code{print-elements-recursively} is: |
| 11128 | |
| 11129 | @itemize @bullet |
| 11130 | @item |
| 11131 | If the list be empty, do nothing. |
| 11132 | @item |
| 11133 | But if the list has at least one element, |
| 11134 | @itemize @minus |
| 11135 | @item |
| 11136 | act on the beginning of the list (the @sc{car} of the list), |
| 11137 | @item |
| 11138 | and make a recursive call on the rest (the @sc{cdr}) of the list. |
| 11139 | @end itemize |
| 11140 | @end itemize |
| 11141 | |
| 11142 | @node Accumulate, Keep, Every, Recursive Patterns |
| 11143 | @comment node-name, next, previous, up |
| 11144 | @unnumberedsubsubsec Recursive Pattern: @emph{accumulate} |
| 11145 | @cindex Accumulate, type of recursive pattern |
| 11146 | @cindex Recursive pattern: accumulate |
| 11147 | |
| 11148 | Another recursive pattern is called the @code{accumulate} pattern. In |
| 11149 | the @code{accumulate} recursive pattern, an action is performed on |
| 11150 | every element of a list and the result of that action is accumulated |
| 11151 | with the results of performing the action on the other elements. |
| 11152 | |
| 11153 | This is very like the `every' pattern using @code{cons}, except that |
| 11154 | @code{cons} is not used, but some other combiner. |
| 11155 | |
| 11156 | @need 1500 |
| 11157 | The pattern is: |
| 11158 | |
| 11159 | @itemize @bullet |
| 11160 | @item |
| 11161 | If a list be empty, return zero or some other constant. |
| 11162 | @item |
| 11163 | Else, act on the beginning of the list (the @sc{car} of the list), |
| 11164 | @itemize @minus |
| 11165 | @item |
| 11166 | and combine that acted-on element, using @code{+} or |
| 11167 | some other combining function, with |
| 11168 | @item |
| 11169 | a recursive call by the function on the rest (the @sc{cdr}) of the list. |
| 11170 | @end itemize |
| 11171 | @end itemize |
| 11172 | |
| 11173 | @need 1500 |
| 11174 | Here is an example: |
| 11175 | |
| 11176 | @smallexample |
| 11177 | @group |
| 11178 | (defun add-elements (numbers-list) |
| 11179 | "Add the elements of NUMBERS-LIST together." |
| 11180 | (if (not numbers-list) |
| 11181 | 0 |
| 11182 | (+ (car numbers-list) (add-elements (cdr numbers-list))))) |
| 11183 | @end group |
| 11184 | |
| 11185 | @group |
| 11186 | (add-elements '(1 2 3 4)) |
| 11187 | @result{} 10 |
| 11188 | @end group |
| 11189 | @end smallexample |
| 11190 | |
| 11191 | @xref{Files List, , Making a List of Files}, for an example of the |
| 11192 | accumulate pattern. |
| 11193 | |
| 11194 | @node Keep, , Accumulate, Recursive Patterns |
| 11195 | @comment node-name, next, previous, up |
| 11196 | @unnumberedsubsubsec Recursive Pattern: @emph{keep} |
| 11197 | @cindex Keep, type of recursive pattern |
| 11198 | @cindex Recursive pattern: keep |
| 11199 | |
| 11200 | A third recursive pattern is called the @code{keep} pattern. |
| 11201 | In the @code{keep} recursive pattern, each element of a list is tested; |
| 11202 | the element is acted on and the results are kept only if the element |
| 11203 | meets a criterion. |
| 11204 | |
| 11205 | Again, this is very like the `every' pattern, except the element is |
| 11206 | skipped unless it meets a criterion. |
| 11207 | |
| 11208 | @need 1500 |
| 11209 | The pattern has three parts: |
| 11210 | |
| 11211 | @itemize @bullet |
| 11212 | @item |
| 11213 | If a list be empty, return @code{nil}. |
| 11214 | @item |
| 11215 | Else, if the beginning of the list (the @sc{car} of the list) passes |
| 11216 | a test |
| 11217 | @itemize @minus |
| 11218 | @item |
| 11219 | act on that element and combine it, using @code{cons} with |
| 11220 | @item |
| 11221 | a recursive call by the function on the rest (the @sc{cdr}) of the list. |
| 11222 | @end itemize |
| 11223 | @item |
| 11224 | Otherwise, if the beginning of the list (the @sc{car} of the list) fails |
| 11225 | the test |
| 11226 | @itemize @minus |
| 11227 | @item |
| 11228 | skip on that element, |
| 11229 | @item |
| 11230 | and, recursively call the function on the rest (the @sc{cdr}) of the list. |
| 11231 | @end itemize |
| 11232 | @end itemize |
| 11233 | |
| 11234 | @need 1500 |
| 11235 | Here is an example that uses @code{cond}: |
| 11236 | |
| 11237 | @smallexample |
| 11238 | @group |
| 11239 | (defun keep-three-letter-words (word-list) |
| 11240 | "Keep three letter words in WORD-LIST." |
| 11241 | (cond |
| 11242 | ;; First do-again-test: stop-condition |
| 11243 | ((not word-list) nil) |
| 11244 | |
| 11245 | ;; Second do-again-test: when to act |
| 11246 | ((eq 3 (length (symbol-name (car word-list)))) |
| 11247 | ;; combine acted-on element with recursive call on shorter list |
| 11248 | (cons (car word-list) (keep-three-letter-words (cdr word-list)))) |
| 11249 | |
| 11250 | ;; Third do-again-test: when to skip element; |
| 11251 | ;; recursively call shorter list with next-step expression |
| 11252 | (t (keep-three-letter-words (cdr word-list))))) |
| 11253 | @end group |
| 11254 | |
| 11255 | @group |
| 11256 | (keep-three-letter-words '(one two three four five six)) |
| 11257 | @result{} (one two six) |
| 11258 | @end group |
| 11259 | @end smallexample |
| 11260 | |
| 11261 | It goes without saying that you need not use @code{nil} as the test for |
| 11262 | when to stop; and you can, of course, combine these patterns. |
| 11263 | |
| 11264 | @node No Deferment, No deferment solution, Recursive Patterns, Recursion |
| 11265 | @subsection Recursion without Deferments |
| 11266 | @cindex Deferment in recursion |
| 11267 | @cindex Recursion without Deferments |
| 11268 | |
| 11269 | Let's consider again what happens with the @code{triangle-recursively} |
| 11270 | function. We will find that the intermediate calculations are |
| 11271 | deferred until all can be done. |
| 11272 | |
| 11273 | @need 800 |
| 11274 | Here is the function definition: |
| 11275 | |
| 11276 | @smallexample |
| 11277 | @group |
| 11278 | (defun triangle-recursively (number) |
| 11279 | "Return the sum of the numbers 1 through NUMBER inclusive. |
| 11280 | Uses recursion." |
| 11281 | (if (= number 1) ; @r{do-again-test} |
| 11282 | 1 ; @r{then-part} |
| 11283 | (+ number ; @r{else-part} |
| 11284 | (triangle-recursively ; @r{recursive call} |
| 11285 | (1- number))))) ; @r{next-step-expression} |
| 11286 | @end group |
| 11287 | @end smallexample |
| 11288 | |
| 11289 | What happens when we call this function with a argument of 7? |
| 11290 | |
| 11291 | The first instance of the @code{triangle-recursively} function adds |
| 11292 | the number 7 to the value returned by a second instance of |
| 11293 | @code{triangle-recursively}, an instance that has been passed an |
| 11294 | argument of 6. That is to say, the first calculation is: |
| 11295 | |
| 11296 | @smallexample |
| 11297 | (+ 7 (triangle-recursively 6)) |
| 11298 | @end smallexample |
| 11299 | |
| 11300 | @noindent |
| 11301 | The first instance of @code{triangle-recursively}---you may want to |
| 11302 | think of it as a little robot---cannot complete its job. It must hand |
| 11303 | off the calculation for @code{(triangle-recursively 6)} to a second |
| 11304 | instance of the program, to a second robot. This second individual is |
| 11305 | completely different from the first one; it is, in the jargon, a |
| 11306 | `different instantiation'. Or, put another way, it is a different |
| 11307 | robot. It is the same model as the first; it calculates triangle |
| 11308 | numbers recursively; but it has a different serial number. |
| 11309 | |
| 11310 | And what does @code{(triangle-recursively 6)} return? It returns the |
| 11311 | number 6 added to the value returned by evaluating |
| 11312 | @code{triangle-recursively} with an argument of 5. Using the robot |
| 11313 | metaphor, it asks yet another robot to help it. |
| 11314 | |
| 11315 | @need 800 |
| 11316 | Now the total is: |
| 11317 | |
| 11318 | @smallexample |
| 11319 | (+ 7 6 (triangle-recursively 5)) |
| 11320 | @end smallexample |
| 11321 | |
| 11322 | @need 800 |
| 11323 | And what happens next? |
| 11324 | |
| 11325 | @smallexample |
| 11326 | (+ 7 6 5 (triangle-recursively 4)) |
| 11327 | @end smallexample |
| 11328 | |
| 11329 | Each time @code{triangle-recursively} is called, except for the last |
| 11330 | time, it creates another instance of the program---another robot---and |
| 11331 | asks it to make a calculation. |
| 11332 | |
| 11333 | @need 800 |
| 11334 | Eventually, the full addition is set up and performed: |
| 11335 | |
| 11336 | @smallexample |
| 11337 | (+ 7 6 5 4 3 2 1) |
| 11338 | @end smallexample |
| 11339 | |
| 11340 | This design for the function defers the calculation of the first step |
| 11341 | until the second can be done, and defers that until the third can be |
| 11342 | done, and so on. Each deferment means the computer must remember what |
| 11343 | is being waited on. This is not a problem when there are only a few |
| 11344 | steps, as in this example. But it can be a problem when there are |
| 11345 | more steps. |
| 11346 | |
| 11347 | @node No deferment solution, , No Deferment, Recursion |
| 11348 | @subsection No Deferment Solution |
| 11349 | @cindex No deferment solution |
| 11350 | @cindex Defermentless solution |
| 11351 | @cindex Solution without deferment |
| 11352 | |
| 11353 | The solution to the problem of deferred operations is to write in a |
| 11354 | manner that does not defer operations@footnote{The phrase @dfn{tail |
| 11355 | recursive} is used to describe such a process, one that uses |
| 11356 | `constant space'.}. This requires |
| 11357 | writing to a different pattern, often one that involves writing two |
| 11358 | function definitions, an `initialization' function and a `helper' |
| 11359 | function. |
| 11360 | |
| 11361 | The `initialization' function sets up the job; the `helper' function |
| 11362 | does the work. |
| 11363 | |
| 11364 | @need 1200 |
| 11365 | Here are the two function definitions for adding up numbers. They are |
| 11366 | so simple, I find them hard to understand. |
| 11367 | |
| 11368 | @smallexample |
| 11369 | @group |
| 11370 | (defun triangle-initialization (number) |
| 11371 | "Return the sum of the numbers 1 through NUMBER inclusive. |
| 11372 | This is the `initialization' component of a two function |
| 11373 | duo that uses recursion." |
| 11374 | (triangle-recursive-helper 0 0 number)) |
| 11375 | @end group |
| 11376 | @end smallexample |
| 11377 | |
| 11378 | @smallexample |
| 11379 | @group |
| 11380 | (defun triangle-recursive-helper (sum counter number) |
| 11381 | "Return SUM, using COUNTER, through NUMBER inclusive. |
| 11382 | This is the `helper' component of a two function duo |
| 11383 | that uses recursion." |
| 11384 | (if (> counter number) |
| 11385 | sum |
| 11386 | (triangle-recursive-helper (+ sum counter) ; @r{sum} |
| 11387 | (1+ counter) ; @r{counter} |
| 11388 | number))) ; @r{number} |
| 11389 | @end group |
| 11390 | @end smallexample |
| 11391 | |
| 11392 | @need 1250 |
| 11393 | Install both function definitions by evaluating them, then call |
| 11394 | @code{triangle-initialization} with 2 rows: |
| 11395 | |
| 11396 | @smallexample |
| 11397 | @group |
| 11398 | (triangle-initialization 2) |
| 11399 | @result{} 3 |
| 11400 | @end group |
| 11401 | @end smallexample |
| 11402 | |
| 11403 | The `initialization' function calls the first instance of the `helper' |
| 11404 | function with three arguments: zero, zero, and a number which is the |
| 11405 | number of rows in the triangle. |
| 11406 | |
| 11407 | The first two arguments passed to the `helper' function are |
| 11408 | initialization values. These values are changed when |
| 11409 | @code{triangle-recursive-helper} invokes new instances.@footnote{The |
| 11410 | jargon is mildly confusing: @code{triangle-recursive-helper} uses a |
| 11411 | process that is iterative in a procedure that is recursive. The |
| 11412 | process is called iterative because the computer need only record the |
| 11413 | three values, @code{sum}, @code{counter}, and @code{number}; the |
| 11414 | procedure is recursive because the function `calls itself'. On the |
| 11415 | other hand, both the process and the procedure used by |
| 11416 | @code{triangle-recursively} are called recursive. The word |
| 11417 | `recursive' has different meanings in the two contexts.} |
| 11418 | |
| 11419 | Let's see what happens when we have a triangle that has one row. (This |
| 11420 | triangle will have one pebble in it!) |
| 11421 | |
| 11422 | @need 1200 |
| 11423 | @code{triangle-initialization} will call its helper with |
| 11424 | the arguments @w{@code{0 0 1}}. That function will run the conditional |
| 11425 | test whether @code{(> counter number)}: |
| 11426 | |
| 11427 | @smallexample |
| 11428 | (> 0 1) |
| 11429 | @end smallexample |
| 11430 | |
| 11431 | @need 1200 |
| 11432 | @noindent |
| 11433 | and find that the result is false, so it will invoke |
| 11434 | the then-part of the @code{if} clause: |
| 11435 | |
| 11436 | @smallexample |
| 11437 | @group |
| 11438 | (triangle-recursive-helper |
| 11439 | (+ sum counter) ; @r{sum plus counter} @result{} @r{sum} |
| 11440 | (1+ counter) ; @r{increment counter} @result{} @r{counter} |
| 11441 | number) ; @r{number stays the same} |
| 11442 | @end group |
| 11443 | @end smallexample |
| 11444 | |
| 11445 | @need 800 |
| 11446 | @noindent |
| 11447 | which will first compute: |
| 11448 | |
| 11449 | @smallexample |
| 11450 | @group |
| 11451 | (triangle-recursive-helper (+ 0 0) ; @r{sum} |
| 11452 | (1+ 0) ; @r{counter} |
| 11453 | 1) ; @r{number} |
| 11454 | @exdent which is: |
| 11455 | |
| 11456 | (triangle-recursive-helper 0 1 1) |
| 11457 | @end group |
| 11458 | @end smallexample |
| 11459 | |
| 11460 | Again, @code{(> counter number)} will be false, so again, the Lisp |
| 11461 | interpreter will evaluate @code{triangle-recursive-helper}, creating a |
| 11462 | new instance with new arguments. |
| 11463 | |
| 11464 | @need 800 |
| 11465 | This new instance will be; |
| 11466 | |
| 11467 | @smallexample |
| 11468 | @group |
| 11469 | (triangle-recursive-helper |
| 11470 | (+ sum counter) ; @r{sum plus counter} @result{} @r{sum} |
| 11471 | (1+ counter) ; @r{increment counter} @result{} @r{counter} |
| 11472 | number) ; @r{number stays the same} |
| 11473 | |
| 11474 | @exdent which is: |
| 11475 | |
| 11476 | (triangle-recursive-helper 1 2 1) |
| 11477 | @end group |
| 11478 | @end smallexample |
| 11479 | |
| 11480 | In this case, the @code{(> counter number)} test will be true! So the |
| 11481 | instance will return the value of the sum, which will be 1, as |
| 11482 | expected. |
| 11483 | |
| 11484 | Now, let's pass @code{triangle-initialization} an argument |
| 11485 | of 2, to find out how many pebbles there are in a triangle with two rows. |
| 11486 | |
| 11487 | That function calls @code{(triangle-recursive-helper 0 0 2)}. |
| 11488 | |
| 11489 | @need 800 |
| 11490 | In stages, the instances called will be: |
| 11491 | |
| 11492 | @smallexample |
| 11493 | @group |
| 11494 | @r{sum counter number} |
| 11495 | (triangle-recursive-helper 0 1 2) |
| 11496 | |
| 11497 | (triangle-recursive-helper 1 2 2) |
| 11498 | |
| 11499 | (triangle-recursive-helper 3 3 2) |
| 11500 | @end group |
| 11501 | @end smallexample |
| 11502 | |
| 11503 | When the last instance is called, the @code{(> counter number)} test |
| 11504 | will be true, so the instance will return the value of @code{sum}, |
| 11505 | which will be 3. |
| 11506 | |
| 11507 | This kind of pattern helps when you are writing functions that can use |
| 11508 | many resources in a computer. |
| 11509 | |
| 11510 | @need 1500 |
| 11511 | @node Looping exercise, , Recursion, Loops & Recursion |
| 11512 | @section Looping Exercise |
| 11513 | |
| 11514 | @itemize @bullet |
| 11515 | @item |
| 11516 | Write a function similar to @code{triangle} in which each row has a |
| 11517 | value which is the square of the row number. Use a @code{while} loop. |
| 11518 | |
| 11519 | @item |
| 11520 | Write a function similar to @code{triangle} that multiplies instead of |
| 11521 | adds the values. |
| 11522 | |
| 11523 | @item |
| 11524 | Rewrite these two functions recursively. Rewrite these functions |
| 11525 | using @code{cond}. |
| 11526 | |
| 11527 | @c comma in printed title causes problem in Info cross reference |
| 11528 | @item |
| 11529 | Write a function for Texinfo mode that creates an index entry at the |
| 11530 | beginning of a paragraph for every @samp{@@dfn} within the paragraph. |
| 11531 | (In a Texinfo file, @samp{@@dfn} marks a definition. This book is |
| 11532 | written in Texinfo.) |
| 11533 | |
| 11534 | Many of the functions you will need are described in two of the |
| 11535 | previous chapters, @ref{Cutting & Storing Text, , Cutting and Storing |
| 11536 | Text}, and @ref{Yanking, , Yanking Text Back}. If you use |
| 11537 | @code{forward-paragraph} to put the index entry at the beginning of |
| 11538 | the paragraph, you will have to use @w{@kbd{C-h f}} |
| 11539 | (@code{describe-function}) to find out how to make the command go |
| 11540 | backwards. |
| 11541 | |
| 11542 | For more information, see |
| 11543 | @ifinfo |
| 11544 | @ref{Indicating, , Indicating Definitions, texinfo}. |
| 11545 | @end ifinfo |
| 11546 | @ifhtml |
| 11547 | @ref{Indicating, , Indicating, texinfo, Texinfo Manual}, which goes to |
| 11548 | a Texinfo manual in the current directory. Or, if you are on the |
| 11549 | Internet, see |
| 11550 | @uref{http://www.gnu.org/manual/texinfo-4.6/html_node/Indicating.html} |
| 11551 | @end ifhtml |
| 11552 | @iftex |
| 11553 | ``Indicating Definitions, Commands, etc.'' in @cite{Texinfo, The GNU |
| 11554 | Documentation Format}. |
| 11555 | @end iftex |
| 11556 | @end itemize |
| 11557 | |
| 11558 | @node Regexp Search, Counting Words, Loops & Recursion, Top |
| 11559 | @comment node-name, next, previous, up |
| 11560 | @chapter Regular Expression Searches |
| 11561 | @cindex Searches, illustrating |
| 11562 | @cindex Regular expression searches |
| 11563 | @cindex Patterns, searching for |
| 11564 | @cindex Motion by sentence and paragraph |
| 11565 | @cindex Sentences, movement by |
| 11566 | @cindex Paragraphs, movement by |
| 11567 | |
| 11568 | Regular expression searches are used extensively in GNU Emacs. The |
| 11569 | two functions, @code{forward-sentence} and @code{forward-paragraph}, |
| 11570 | illustrate these searches well. They use regular expressions to find |
| 11571 | where to move point. The phrase `regular expression' is often written |
| 11572 | as `regexp'. |
| 11573 | |
| 11574 | Regular expression searches are described in @ref{Regexp Search, , |
| 11575 | Regular Expression Search, emacs, The GNU Emacs Manual}, as well as in |
| 11576 | @ref{Regular Expressions, , , elisp, The GNU Emacs Lisp Reference |
| 11577 | Manual}. In writing this chapter, I am presuming that you have at |
| 11578 | least a mild acquaintance with them. The major point to remember is |
| 11579 | that regular expressions permit you to search for patterns as well as |
| 11580 | for literal strings of characters. For example, the code in |
| 11581 | @code{forward-sentence} searches for the pattern of possible |
| 11582 | characters that could mark the end of a sentence, and moves point to |
| 11583 | that spot. |
| 11584 | |
| 11585 | Before looking at the code for the @code{forward-sentence} function, it |
| 11586 | is worth considering what the pattern that marks the end of a sentence |
| 11587 | must be. The pattern is discussed in the next section; following that |
| 11588 | is a description of the regular expression search function, |
| 11589 | @code{re-search-forward}. The @code{forward-sentence} function |
| 11590 | is described in the section following. Finally, the |
| 11591 | @code{forward-paragraph} function is described in the last section of |
| 11592 | this chapter. @code{forward-paragraph} is a complex function that |
| 11593 | introduces several new features. |
| 11594 | |
| 11595 | @menu |
| 11596 | * sentence-end:: The regular expression for @code{sentence-end}. |
| 11597 | * re-search-forward:: Very similar to @code{search-forward}. |
| 11598 | * forward-sentence:: A straightforward example of regexp search. |
| 11599 | * forward-paragraph:: A somewhat complex example. |
| 11600 | * etags:: How to create your own @file{TAGS} table. |
| 11601 | * Regexp Review:: |
| 11602 | * re-search Exercises:: |
| 11603 | @end menu |
| 11604 | |
| 11605 | @node sentence-end, re-search-forward, Regexp Search, Regexp Search |
| 11606 | @comment node-name, next, previous, up |
| 11607 | @section The Regular Expression for @code{sentence-end} |
| 11608 | @findex sentence-end |
| 11609 | |
| 11610 | The symbol @code{sentence-end} is bound to the pattern that marks the |
| 11611 | end of a sentence. What should this regular expression be? |
| 11612 | |
| 11613 | Clearly, a sentence may be ended by a period, a question mark, or an |
| 11614 | exclamation mark. Indeed, only clauses that end with one of those three |
| 11615 | characters should be considered the end of a sentence. This means that |
| 11616 | the pattern should include the character set: |
| 11617 | |
| 11618 | @smallexample |
| 11619 | [.?!] |
| 11620 | @end smallexample |
| 11621 | |
| 11622 | However, we do not want @code{forward-sentence} merely to jump to a |
| 11623 | period, a question mark, or an exclamation mark, because such a character |
| 11624 | might be used in the middle of a sentence. A period, for example, is |
| 11625 | used after abbreviations. So other information is needed. |
| 11626 | |
| 11627 | According to convention, you type two spaces after every sentence, but |
| 11628 | only one space after a period, a question mark, or an exclamation mark in |
| 11629 | the body of a sentence. So a period, a question mark, or an exclamation |
| 11630 | mark followed by two spaces is a good indicator of an end of sentence. |
| 11631 | However, in a file, the two spaces may instead be a tab or the end of a |
| 11632 | line. This means that the regular expression should include these three |
| 11633 | items as alternatives. |
| 11634 | |
| 11635 | @need 800 |
| 11636 | This group of alternatives will look like this: |
| 11637 | |
| 11638 | @smallexample |
| 11639 | @group |
| 11640 | \\($\\| \\| \\) |
| 11641 | ^ ^^ |
| 11642 | TAB SPC |
| 11643 | @end group |
| 11644 | @end smallexample |
| 11645 | |
| 11646 | @noindent |
| 11647 | Here, @samp{$} indicates the end of the line, and I have pointed out |
| 11648 | where the tab and two spaces are inserted in the expression. Both are |
| 11649 | inserted by putting the actual characters into the expression. |
| 11650 | |
| 11651 | Two backslashes, @samp{\\}, are required before the parentheses and |
| 11652 | vertical bars: the first backslash quotes the following backslash in |
| 11653 | Emacs; and the second indicates that the following character, the |
| 11654 | parenthesis or the vertical bar, is special. |
| 11655 | |
| 11656 | @need 1000 |
| 11657 | Also, a sentence may be followed by one or more carriage returns, like |
| 11658 | this: |
| 11659 | |
| 11660 | @smallexample |
| 11661 | @group |
| 11662 | [ |
| 11663 | ]* |
| 11664 | @end group |
| 11665 | @end smallexample |
| 11666 | |
| 11667 | @noindent |
| 11668 | Like tabs and spaces, a carriage return is inserted into a regular |
| 11669 | expression by inserting it literally. The asterisk indicates that the |
| 11670 | @key{RET} is repeated zero or more times. |
| 11671 | |
| 11672 | But a sentence end does not consist only of a period, a question mark or |
| 11673 | an exclamation mark followed by appropriate space: a closing quotation |
| 11674 | mark or a closing brace of some kind may precede the space. Indeed more |
| 11675 | than one such mark or brace may precede the space. These require a |
| 11676 | expression that looks like this: |
| 11677 | |
| 11678 | @smallexample |
| 11679 | []\"')@}]* |
| 11680 | @end smallexample |
| 11681 | |
| 11682 | In this expression, the first @samp{]} is the first character in the |
| 11683 | expression; the second character is @samp{"}, which is preceded by a |
| 11684 | @samp{\} to tell Emacs the @samp{"} is @emph{not} special. The last |
| 11685 | three characters are @samp{'}, @samp{)}, and @samp{@}}. |
| 11686 | |
| 11687 | All this suggests what the regular expression pattern for matching the |
| 11688 | end of a sentence should be; and, indeed, if we evaluate |
| 11689 | @code{sentence-end} we find that it returns the following value: |
| 11690 | |
| 11691 | @smallexample |
| 11692 | @group |
| 11693 | sentence-end |
| 11694 | @result{} "[.?!][]\"')@}]*\\($\\| \\| \\)[ |
| 11695 | ]*" |
| 11696 | @end group |
| 11697 | @end smallexample |
| 11698 | |
| 11699 | @ignore |
| 11700 | |
| 11701 | @noindent |
| 11702 | (Note that here the @key{TAB}, two spaces, and @key{RET} are shown |
| 11703 | literally in the pattern.) |
| 11704 | |
| 11705 | This regular expression can be decyphered as follows: |
| 11706 | |
| 11707 | @table @code |
| 11708 | @item [.?!] |
| 11709 | The first part of the pattern is the three characters, a period, a question |
| 11710 | mark and an exclamation mark, within square brackets. The pattern must |
| 11711 | begin with one or other of these characters. |
| 11712 | |
| 11713 | @item []\"')@}]* |
| 11714 | The second part of the pattern is the group of closing braces and |
| 11715 | quotation marks, which can appear zero or more times. These may follow |
| 11716 | the period, question mark or exclamation mark. In a regular expression, |
| 11717 | the backslash, @samp{\}, followed by the double quotation mark, |
| 11718 | @samp{"}, indicates the class of string-quote characters. Usually, the |
| 11719 | double quotation mark is the only character in this class. The |
| 11720 | asterisk, @samp{*}, indicates that the items in the previous group (the |
| 11721 | group surrounded by square brackets, @samp{[]}) may be repeated zero or |
| 11722 | more times. |
| 11723 | |
| 11724 | @item \\($\\| \\| \\) |
| 11725 | The third part of the pattern is one or other of: either the end of a |
| 11726 | line, or two blank spaces, or a tab. The double back-slashes are used |
| 11727 | to prevent Emacs from reading the parentheses and vertical bars as part |
| 11728 | of the search pattern; the parentheses are used to mark the group and |
| 11729 | the vertical bars are used to indicated that the patterns to either side |
| 11730 | of them are alternatives. The dollar sign is used to indicate the end |
| 11731 | of a line and both the two spaces and the tab are each inserted as is to |
| 11732 | indicate what they are. |
| 11733 | |
| 11734 | @item [@key{RET}]* |
| 11735 | Finally, the last part of the pattern indicates that the end of the line |
| 11736 | or the whitespace following the period, question mark or exclamation |
| 11737 | mark may, but need not, be followed by one or more carriage returns. In |
| 11738 | the pattern, the carriage return is inserted as an actual carriage |
| 11739 | return between square brackets but here it is shown as @key{RET}. |
| 11740 | @end table |
| 11741 | |
| 11742 | @end ignore |
| 11743 | |
| 11744 | @node re-search-forward, forward-sentence, sentence-end, Regexp Search |
| 11745 | @comment node-name, next, previous, up |
| 11746 | @section The @code{re-search-forward} Function |
| 11747 | @findex re-search-forward |
| 11748 | |
| 11749 | The @code{re-search-forward} function is very like the |
| 11750 | @code{search-forward} function. (@xref{search-forward, , The |
| 11751 | @code{search-forward} Function}.) |
| 11752 | |
| 11753 | @code{re-search-forward} searches for a regular expression. If the |
| 11754 | search is successful, it leaves point immediately after the last |
| 11755 | character in the target. If the search is backwards, it leaves point |
| 11756 | just before the first character in the target. You may tell |
| 11757 | @code{re-search-forward} to return @code{t} for true. (Moving point |
| 11758 | is therefore a `side effect'.) |
| 11759 | |
| 11760 | Like @code{search-forward}, the @code{re-search-forward} function takes |
| 11761 | four arguments: |
| 11762 | |
| 11763 | @enumerate |
| 11764 | @item |
| 11765 | The first argument is the regular expression that the function searches |
| 11766 | for. The regular expression will be a string between quotations marks. |
| 11767 | |
| 11768 | @item |
| 11769 | The optional second argument limits how far the function will search; it is a |
| 11770 | bound, which is specified as a position in the buffer. |
| 11771 | |
| 11772 | @item |
| 11773 | The optional third argument specifies how the function responds to |
| 11774 | failure: @code{nil} as the third argument causes the function to |
| 11775 | signal an error (and print a message) when the search fails; any other |
| 11776 | value causes it to return @code{nil} if the search fails and @code{t} |
| 11777 | if the search succeeds. |
| 11778 | |
| 11779 | @item |
| 11780 | The optional fourth argument is the repeat count. A negative repeat |
| 11781 | count causes @code{re-search-forward} to search backwards. |
| 11782 | @end enumerate |
| 11783 | |
| 11784 | @need 800 |
| 11785 | The template for @code{re-search-forward} looks like this: |
| 11786 | |
| 11787 | @smallexample |
| 11788 | @group |
| 11789 | (re-search-forward "@var{regular-expression}" |
| 11790 | @var{limit-of-search} |
| 11791 | @var{what-to-do-if-search-fails} |
| 11792 | @var{repeat-count}) |
| 11793 | @end group |
| 11794 | @end smallexample |
| 11795 | |
| 11796 | The second, third, and fourth arguments are optional. However, if you |
| 11797 | want to pass a value to either or both of the last two arguments, you |
| 11798 | must also pass a value to all the preceding arguments. Otherwise, the |
| 11799 | Lisp interpreter will mistake which argument you are passing the value |
| 11800 | to. |
| 11801 | |
| 11802 | @need 1200 |
| 11803 | In the @code{forward-sentence} function, the regular expression will be |
| 11804 | the value of the variable @code{sentence-end}, namely: |
| 11805 | |
| 11806 | @smallexample |
| 11807 | @group |
| 11808 | "[.?!][]\"')@}]*\\($\\| \\| \\)[ |
| 11809 | ]*" |
| 11810 | @end group |
| 11811 | @end smallexample |
| 11812 | |
| 11813 | @noindent |
| 11814 | The limit of the search will be the end of the paragraph (since a |
| 11815 | sentence cannot go beyond a paragraph). If the search fails, the |
| 11816 | function will return @code{nil}; and the repeat count will be provided |
| 11817 | by the argument to the @code{forward-sentence} function. |
| 11818 | |
| 11819 | @node forward-sentence, forward-paragraph, re-search-forward, Regexp Search |
| 11820 | @comment node-name, next, previous, up |
| 11821 | @section @code{forward-sentence} |
| 11822 | @findex forward-sentence |
| 11823 | |
| 11824 | The command to move the cursor forward a sentence is a straightforward |
| 11825 | illustration of how to use regular expression searches in Emacs Lisp. |
| 11826 | Indeed, the function looks longer and more complicated than it is; this |
| 11827 | is because the function is designed to go backwards as well as forwards; |
| 11828 | and, optionally, over more than one sentence. The function is usually |
| 11829 | bound to the key command @kbd{M-e}. |
| 11830 | |
| 11831 | @menu |
| 11832 | * Complete forward-sentence:: |
| 11833 | * fwd-sentence while loops:: Two @code{while} loops. |
| 11834 | * fwd-sentence re-search:: A regular expression search. |
| 11835 | @end menu |
| 11836 | |
| 11837 | @node Complete forward-sentence, fwd-sentence while loops, forward-sentence, forward-sentence |
| 11838 | @ifnottex |
| 11839 | @unnumberedsubsec Complete @code{forward-sentence} function definition |
| 11840 | @end ifnottex |
| 11841 | |
| 11842 | @need 1250 |
| 11843 | Here is the code for @code{forward-sentence}: |
| 11844 | |
| 11845 | @smallexample |
| 11846 | @group |
| 11847 | (defun forward-sentence (&optional arg) |
| 11848 | "Move forward to next sentence-end. With argument, repeat. |
| 11849 | With negative argument, move backward repeatedly to sentence-beginning. |
| 11850 | Sentence ends are identified by the value of sentence-end |
| 11851 | treated as a regular expression. Also, every paragraph boundary |
| 11852 | terminates sentences as well." |
| 11853 | @end group |
| 11854 | @group |
| 11855 | (interactive "p") |
| 11856 | (or arg (setq arg 1)) |
| 11857 | (while (< arg 0) |
| 11858 | (let ((par-beg |
| 11859 | (save-excursion (start-of-paragraph-text) (point)))) |
| 11860 | (if (re-search-backward |
| 11861 | (concat sentence-end "[^ \t\n]") par-beg t) |
| 11862 | (goto-char (1- (match-end 0))) |
| 11863 | (goto-char par-beg))) |
| 11864 | (setq arg (1+ arg))) |
| 11865 | (while (> arg 0) |
| 11866 | (let ((par-end |
| 11867 | (save-excursion (end-of-paragraph-text) (point)))) |
| 11868 | (if (re-search-forward sentence-end par-end t) |
| 11869 | (skip-chars-backward " \t\n") |
| 11870 | (goto-char par-end))) |
| 11871 | (setq arg (1- arg)))) |
| 11872 | @end group |
| 11873 | @end smallexample |
| 11874 | |
| 11875 | The function looks long at first sight and it is best to look at its |
| 11876 | skeleton first, and then its muscle. The way to see the skeleton is to |
| 11877 | look at the expressions that start in the left-most columns: |
| 11878 | |
| 11879 | @smallexample |
| 11880 | @group |
| 11881 | (defun forward-sentence (&optional arg) |
| 11882 | "@var{documentation}@dots{}" |
| 11883 | (interactive "p") |
| 11884 | (or arg (setq arg 1)) |
| 11885 | (while (< arg 0) |
| 11886 | @var{body-of-while-loop} |
| 11887 | (while (> arg 0) |
| 11888 | @var{body-of-while-loop} |
| 11889 | @end group |
| 11890 | @end smallexample |
| 11891 | |
| 11892 | This looks much simpler! The function definition consists of |
| 11893 | documentation, an @code{interactive} expression, an @code{or} |
| 11894 | expression, and @code{while} loops. |
| 11895 | |
| 11896 | Let's look at each of these parts in turn. |
| 11897 | |
| 11898 | We note that the documentation is thorough and understandable. |
| 11899 | |
| 11900 | The function has an @code{interactive "p"} declaration. This means |
| 11901 | that the processed prefix argument, if any, is passed to the |
| 11902 | function as its argument. (This will be a number.) If the function |
| 11903 | is not passed an argument (it is optional) then the argument |
| 11904 | @code{arg} will be bound to 1. When @code{forward-sentence} is called |
| 11905 | non-interactively without an argument, @code{arg} is bound to |
| 11906 | @code{nil}. |
| 11907 | |
| 11908 | The @code{or} expression handles the prefix argument. What it does is |
| 11909 | either leave the value of @code{arg} as it is, but only if @code{arg} |
| 11910 | is bound to a value; or it sets the value of @code{arg} to 1, in the |
| 11911 | case when @code{arg} is bound to @code{nil}. |
| 11912 | |
| 11913 | @node fwd-sentence while loops, fwd-sentence re-search, Complete forward-sentence, forward-sentence |
| 11914 | @unnumberedsubsec The @code{while} loops |
| 11915 | |
| 11916 | Two @code{while} loops follow the @code{or} expression. The first |
| 11917 | @code{while} has a true-or-false-test that tests true if the prefix |
| 11918 | argument for @code{forward-sentence} is a negative number. This is for |
| 11919 | going backwards. The body of this loop is similar to the body of the |
| 11920 | second @code{while} clause, but it is not exactly the same. We will |
| 11921 | skip this @code{while} loop and concentrate on the second @code{while} |
| 11922 | loop. |
| 11923 | |
| 11924 | @need 1500 |
| 11925 | The second @code{while} loop is for moving point forward. Its skeleton |
| 11926 | looks like this: |
| 11927 | |
| 11928 | @smallexample |
| 11929 | @group |
| 11930 | (while (> arg 0) ; @r{true-or-false-test} |
| 11931 | (let @var{varlist} |
| 11932 | (if (@var{true-or-false-test}) |
| 11933 | @var{then-part} |
| 11934 | @var{else-part} |
| 11935 | (setq arg (1- arg)))) ; @code{while} @r{loop decrementer} |
| 11936 | @end group |
| 11937 | @end smallexample |
| 11938 | |
| 11939 | The @code{while} loop is of the decrementing kind. |
| 11940 | (@xref{Decrementing Loop, , A Loop with a Decrementing Counter}.) It |
| 11941 | has a true-or-false-test that tests true so long as the counter (in |
| 11942 | this case, the variable @code{arg}) is greater than zero; and it has a |
| 11943 | decrementer that subtracts 1 from the value of the counter every time |
| 11944 | the loop repeats. |
| 11945 | |
| 11946 | If no prefix argument is given to @code{forward-sentence}, which is |
| 11947 | the most common way the command is used, this @code{while} loop will |
| 11948 | run once, since the value of @code{arg} will be 1. |
| 11949 | |
| 11950 | The body of the @code{while} loop consists of a @code{let} expression, |
| 11951 | which creates and binds a local variable, and has, as its body, an |
| 11952 | @code{if} expression. |
| 11953 | |
| 11954 | @need 1250 |
| 11955 | The body of the @code{while} loop looks like this: |
| 11956 | |
| 11957 | @smallexample |
| 11958 | @group |
| 11959 | (let ((par-end |
| 11960 | (save-excursion (end-of-paragraph-text) (point)))) |
| 11961 | (if (re-search-forward sentence-end par-end t) |
| 11962 | (skip-chars-backward " \t\n") |
| 11963 | (goto-char par-end))) |
| 11964 | @end group |
| 11965 | @end smallexample |
| 11966 | |
| 11967 | The @code{let} expression creates and binds the local variable |
| 11968 | @code{par-end}. As we shall see, this local variable is designed to |
| 11969 | provide a bound or limit to the regular expression search. If the |
| 11970 | search fails to find a proper sentence ending in the paragraph, it will |
| 11971 | stop on reaching the end of the paragraph. |
| 11972 | |
| 11973 | But first, let us examine how @code{par-end} is bound to the value of |
| 11974 | the end of the paragraph. What happens is that the @code{let} sets the |
| 11975 | value of @code{par-end} to the value returned when the Lisp interpreter |
| 11976 | evaluates the expression |
| 11977 | |
| 11978 | @smallexample |
| 11979 | @group |
| 11980 | (save-excursion (end-of-paragraph-text) (point)) |
| 11981 | @end group |
| 11982 | @end smallexample |
| 11983 | |
| 11984 | @noindent |
| 11985 | In this expression, @code{(end-of-paragraph-text)} moves point to the |
| 11986 | end of the paragraph, @code{(point)} returns the value of point, and then |
| 11987 | @code{save-excursion} restores point to its original position. Thus, |
| 11988 | the @code{let} binds @code{par-end} to the value returned by the |
| 11989 | @code{save-excursion} expression, which is the position of the end of |
| 11990 | the paragraph. (The @code{(end-of-paragraph-text)} function uses |
| 11991 | @code{forward-paragraph}, which we will discuss shortly.) |
| 11992 | |
| 11993 | @need 1200 |
| 11994 | Emacs next evaluates the body of the @code{let}, which is an @code{if} |
| 11995 | expression that looks like this: |
| 11996 | |
| 11997 | @smallexample |
| 11998 | @group |
| 11999 | (if (re-search-forward sentence-end par-end t) ; @r{if-part} |
| 12000 | (skip-chars-backward " \t\n") ; @r{then-part} |
| 12001 | (goto-char par-end))) ; @r{else-part} |
| 12002 | @end group |
| 12003 | @end smallexample |
| 12004 | |
| 12005 | The @code{if} tests whether its first argument is true and if so, |
| 12006 | evaluates its then-part; otherwise, the Emacs Lisp interpreter |
| 12007 | evaluates the else-part. The true-or-false-test of the @code{if} |
| 12008 | expression is the regular expression search. |
| 12009 | |
| 12010 | It may seem odd to have what looks like the `real work' of |
| 12011 | the @code{forward-sentence} function buried here, but this is a common |
| 12012 | way this kind of operation is carried out in Lisp. |
| 12013 | |
| 12014 | @node fwd-sentence re-search, , fwd-sentence while loops, forward-sentence |
| 12015 | @unnumberedsubsec The regular expression search |
| 12016 | |
| 12017 | The @code{re-search-forward} function searches for the end of the |
| 12018 | sentence, that is, for the pattern defined by the @code{sentence-end} |
| 12019 | regular expression. If the pattern is found---if the end of the sentence is |
| 12020 | found---then the @code{re-search-forward} function does two things: |
| 12021 | |
| 12022 | @enumerate |
| 12023 | @item |
| 12024 | The @code{re-search-forward} function carries out a side effect, which |
| 12025 | is to move point to the end of the occurrence found. |
| 12026 | |
| 12027 | @item |
| 12028 | The @code{re-search-forward} function returns a value of true. This is |
| 12029 | the value received by the @code{if}, and means that the search was |
| 12030 | successful. |
| 12031 | @end enumerate |
| 12032 | |
| 12033 | @noindent |
| 12034 | The side effect, the movement of point, is completed before the |
| 12035 | @code{if} function is handed the value returned by the successful |
| 12036 | conclusion of the search. |
| 12037 | |
| 12038 | When the @code{if} function receives the value of true from a successful |
| 12039 | call to @code{re-search-forward}, the @code{if} evaluates the then-part, |
| 12040 | which is the expression @code{(skip-chars-backward " \t\n")}. This |
| 12041 | expression moves backwards over any blank spaces, tabs or carriage |
| 12042 | returns until a printed character is found and then leaves point after |
| 12043 | the character. Since point has already been moved to the end of the |
| 12044 | pattern that marks the end of the sentence, this action leaves point |
| 12045 | right after the closing printed character of the sentence, which is |
| 12046 | usually a period. |
| 12047 | |
| 12048 | On the other hand, if the @code{re-search-forward} function fails to |
| 12049 | find a pattern marking the end of the sentence, the function returns |
| 12050 | false. The false then causes the @code{if} to evaluate its third |
| 12051 | argument, which is @code{(goto-char par-end)}: it moves point to the |
| 12052 | end of the paragraph. |
| 12053 | |
| 12054 | Regular expression searches are exceptionally useful and the pattern |
| 12055 | illustrated by @code{re-search-forward}, in which the search is the |
| 12056 | test of an @code{if} expression, is handy. You will see or write code |
| 12057 | incorporating this pattern often. |
| 12058 | |
| 12059 | @node forward-paragraph, etags, forward-sentence, Regexp Search |
| 12060 | @comment node-name, next, previous, up |
| 12061 | @section @code{forward-paragraph}: a Goldmine of Functions |
| 12062 | @findex forward-paragraph |
| 12063 | |
| 12064 | The @code{forward-paragraph} function moves point forward to the end |
| 12065 | of the paragraph. It is usually bound to @kbd{M-@}} and makes use of a |
| 12066 | number of functions that are important in themselves, including |
| 12067 | @code{let*}, @code{match-beginning}, and @code{looking-at}. |
| 12068 | |
| 12069 | The function definition for @code{forward-paragraph} is considerably |
| 12070 | longer than the function definition for @code{forward-sentence} |
| 12071 | because it works with a paragraph, each line of which may begin with a |
| 12072 | fill prefix. |
| 12073 | |
| 12074 | A fill prefix consists of a string of characters that are repeated at |
| 12075 | the beginning of each line. For example, in Lisp code, it is a |
| 12076 | convention to start each line of a paragraph-long comment with |
| 12077 | @samp{;;; }. In Text mode, four blank spaces make up another common |
| 12078 | fill prefix, creating an indented paragraph. (@xref{Fill Prefix, , , |
| 12079 | emacs, The GNU Emacs Manual}, for more information about fill |
| 12080 | prefixes.) |
| 12081 | |
| 12082 | The existence of a fill prefix means that in addition to being able to |
| 12083 | find the end of a paragraph whose lines begin on the left-most |
| 12084 | column, the @code{forward-paragraph} function must be able to find the |
| 12085 | end of a paragraph when all or many of the lines in the buffer begin |
| 12086 | with the fill prefix. |
| 12087 | |
| 12088 | Moreover, it is sometimes practical to ignore a fill prefix that |
| 12089 | exists, especially when blank lines separate paragraphs. |
| 12090 | This is an added complication. |
| 12091 | |
| 12092 | @menu |
| 12093 | * forward-paragraph in brief:: Key parts of the function definition. |
| 12094 | * fwd-para let:: The @code{let*} expression. |
| 12095 | * fwd-para while:: The forward motion @code{while} loop. |
| 12096 | * fwd-para between paragraphs:: Movement between paragraphs. |
| 12097 | * fwd-para within paragraph:: Movement within paragraphs. |
| 12098 | * fwd-para no fill prefix:: When there is no fill prefix. |
| 12099 | * fwd-para with fill prefix:: When there is a fill prefix. |
| 12100 | * fwd-para summary:: Summary of @code{forward-paragraph} code. |
| 12101 | @end menu |
| 12102 | |
| 12103 | @node forward-paragraph in brief, fwd-para let, forward-paragraph, forward-paragraph |
| 12104 | @ifnottex |
| 12105 | @unnumberedsubsec Shortened @code{forward-paragraph} function definition |
| 12106 | @end ifnottex |
| 12107 | |
| 12108 | Rather than print all of the @code{forward-paragraph} function, we |
| 12109 | will only print parts of it. Read without preparation, the function |
| 12110 | can be daunting! |
| 12111 | |
| 12112 | @need 800 |
| 12113 | In outline, the function looks like this: |
| 12114 | |
| 12115 | @smallexample |
| 12116 | @group |
| 12117 | (defun forward-paragraph (&optional arg) |
| 12118 | "@var{documentation}@dots{}" |
| 12119 | (interactive "p") |
| 12120 | (or arg (setq arg 1)) |
| 12121 | (let* |
| 12122 | @var{varlist} |
| 12123 | (while (< arg 0) ; @r{backward-moving-code} |
| 12124 | @dots{} |
| 12125 | (setq arg (1+ arg))) |
| 12126 | (while (> arg 0) ; @r{forward-moving-code} |
| 12127 | @dots{} |
| 12128 | (setq arg (1- arg))))) |
| 12129 | @end group |
| 12130 | @end smallexample |
| 12131 | |
| 12132 | The first parts of the function are routine: the function's argument |
| 12133 | list consists of one optional argument. Documentation follows. |
| 12134 | |
| 12135 | The lower case @samp{p} in the @code{interactive} declaration means |
| 12136 | that the processed prefix argument, if any, is passed to the function. |
| 12137 | This will be a number, and is the repeat count of how many paragraphs |
| 12138 | point will move. The @code{or} expression in the next line handles |
| 12139 | the common case when no argument is passed to the function, which occurs |
| 12140 | if the function is called from other code rather than interactively. |
| 12141 | This case was described earlier. (@xref{forward-sentence, The |
| 12142 | @code{forward-sentence} function}.) Now we reach the end of the |
| 12143 | familiar part of this function. |
| 12144 | |
| 12145 | @node fwd-para let, fwd-para while, forward-paragraph in brief, forward-paragraph |
| 12146 | @unnumberedsubsec The @code{let*} expression |
| 12147 | |
| 12148 | The next line of the @code{forward-paragraph} function begins a |
| 12149 | @code{let*} expression. This is a different kind of expression than |
| 12150 | we have seen so far. The symbol is @code{let*} not @code{let}. |
| 12151 | |
| 12152 | The @code{let*} special form is like @code{let} except that Emacs sets |
| 12153 | each variable in sequence, one after another, and variables in the |
| 12154 | latter part of the varlist can make use of the values to which Emacs |
| 12155 | set variables in the earlier part of the varlist. |
| 12156 | |
| 12157 | In the @code{let*} expression in this function, Emacs binds two |
| 12158 | variables: @code{fill-prefix-regexp} and @code{paragraph-separate}. |
| 12159 | The value to which @code{paragraph-separate} is bound depends on the |
| 12160 | value of @code{fill-prefix-regexp}. |
| 12161 | |
| 12162 | @need 1200 |
| 12163 | Let's look at each in turn. The symbol @code{fill-prefix-regexp} is |
| 12164 | set to the value returned by evaluating the following list: |
| 12165 | |
| 12166 | @smallexample |
| 12167 | @group |
| 12168 | (and fill-prefix |
| 12169 | (not (equal fill-prefix "")) |
| 12170 | (not paragraph-ignore-fill-prefix) |
| 12171 | (regexp-quote fill-prefix)) |
| 12172 | @end group |
| 12173 | @end smallexample |
| 12174 | |
| 12175 | @noindent |
| 12176 | This is an expression whose first element is the @code{and} special form. |
| 12177 | |
| 12178 | As we learned earlier (@pxref{kill-new function, , The @code{kill-new} |
| 12179 | function}), the @code{and} special form evaluates each of its |
| 12180 | arguments until one of the arguments returns a value of @code{nil}, in |
| 12181 | which case the @code{and} expression returns @code{nil}; however, if |
| 12182 | none of the arguments returns a value of @code{nil}, the value |
| 12183 | resulting from evaluating the last argument is returned. (Since such |
| 12184 | a value is not @code{nil}, it is considered true in Lisp.) In other |
| 12185 | words, an @code{and} expression returns a true value only if all its |
| 12186 | arguments are true. |
| 12187 | @findex and |
| 12188 | |
| 12189 | In this case, the variable @code{fill-prefix-regexp} is bound to a |
| 12190 | non-@code{nil} value only if the following four expressions produce a |
| 12191 | true (i.e., a non-@code{nil}) value when they are evaluated; otherwise, |
| 12192 | @code{fill-prefix-regexp} is bound to @code{nil}. |
| 12193 | |
| 12194 | @table @code |
| 12195 | @item fill-prefix |
| 12196 | When this variable is evaluated, the value of the fill prefix, if any, |
| 12197 | is returned. If there is no fill prefix, this variable returns |
| 12198 | @code{nil}. |
| 12199 | |
| 12200 | @item (not (equal fill-prefix "") |
| 12201 | This expression checks whether an existing fill prefix is an empty |
| 12202 | string, that is, a string with no characters in it. An empty string is |
| 12203 | not a useful fill prefix. |
| 12204 | |
| 12205 | @item (not paragraph-ignore-fill-prefix) |
| 12206 | This expression returns @code{nil} if the variable |
| 12207 | @code{paragraph-ignore-fill-prefix} has been turned on by being set to a |
| 12208 | true value such as @code{t}. |
| 12209 | |
| 12210 | @item (regexp-quote fill-prefix) |
| 12211 | This is the last argument to the @code{and} special form. If all the |
| 12212 | arguments to the @code{and} are true, the value resulting from |
| 12213 | evaluating this expression will be returned by the @code{and} expression |
| 12214 | and bound to the variable @code{fill-prefix-regexp}, |
| 12215 | @end table |
| 12216 | |
| 12217 | @findex regexp-quote |
| 12218 | @noindent |
| 12219 | The result of evaluating this @code{and} expression successfully is that |
| 12220 | @code{fill-prefix-regexp} will be bound to the value of |
| 12221 | @code{fill-prefix} as modified by the @code{regexp-quote} function. |
| 12222 | What @code{regexp-quote} does is read a string and return a regular |
| 12223 | expression that will exactly match the string and match nothing else. |
| 12224 | This means that @code{fill-prefix-regexp} will be set to a value that |
| 12225 | will exactly match the fill prefix if the fill prefix exists. |
| 12226 | Otherwise, the variable will be set to @code{nil}. |
| 12227 | |
| 12228 | The second local variable in the @code{let*} expression is |
| 12229 | @code{paragraph-separate}. It is bound to the value returned by |
| 12230 | evaluating the expression: |
| 12231 | |
| 12232 | @smallexample |
| 12233 | @group |
| 12234 | (if fill-prefix-regexp |
| 12235 | (concat paragraph-separate |
| 12236 | "\\|^" fill-prefix-regexp "[ \t]*$") |
| 12237 | paragraph-separate))) |
| 12238 | @end group |
| 12239 | @end smallexample |
| 12240 | |
| 12241 | This expression shows why @code{let*} rather than @code{let} was used. |
| 12242 | The true-or-false-test for the @code{if} depends on whether the variable |
| 12243 | @code{fill-prefix-regexp} evaluates to @code{nil} or some other value. |
| 12244 | |
| 12245 | If @code{fill-prefix-regexp} does not have a value, Emacs evaluates |
| 12246 | the else-part of the @code{if} expression and binds |
| 12247 | @code{paragraph-separate} to its local value. |
| 12248 | (@code{paragraph-separate} is a regular expression that matches what |
| 12249 | separates paragraphs.) |
| 12250 | |
| 12251 | But if @code{fill-prefix-regexp} does have a value, Emacs evaluates |
| 12252 | the then-part of the @code{if} expression and binds |
| 12253 | @code{paragraph-separate} to a regular expression that includes the |
| 12254 | @code{fill-prefix-regexp} as part of the pattern. |
| 12255 | |
| 12256 | Specifically, @code{paragraph-separate} is set to the original value |
| 12257 | of the paragraph separate regular expression concatenated with an |
| 12258 | alternative expression that consists of the @code{fill-prefix-regexp} |
| 12259 | followed by a blank line. The @samp{^} indicates that the |
| 12260 | @code{fill-prefix-regexp} must begin a line, and the optional |
| 12261 | whitespace to the end of the line is defined by @w{@code{"[ \t]*$"}}.) |
| 12262 | The @samp{\\|} defines this portion of the regexp as an alternative to |
| 12263 | @code{paragraph-separate}. |
| 12264 | |
| 12265 | Now we get into the body of the @code{let*}. The first part of the body |
| 12266 | of the @code{let*} deals with the case when the function is given a |
| 12267 | negative argument and is therefore moving backwards. We will skip this |
| 12268 | section. |
| 12269 | |
| 12270 | @node fwd-para while, fwd-para between paragraphs, fwd-para let, forward-paragraph |
| 12271 | @unnumberedsubsec The forward motion @code{while} loop |
| 12272 | |
| 12273 | The second part of the body of the @code{let*} deals with forward |
| 12274 | motion. It is a @code{while} loop that repeats itself so long as the |
| 12275 | value of @code{arg} is greater than zero. In the most common use of |
| 12276 | the function, the value of the argument is 1, so the body of the |
| 12277 | @code{while} loop is evaluated exactly once, and the cursor moves |
| 12278 | forward one paragraph. |
| 12279 | |
| 12280 | This part handles three situations: when point is between paragraphs, |
| 12281 | when point is within a paragraph and there is a fill prefix, and |
| 12282 | when point is within a paragraph and there is no fill prefix. |
| 12283 | |
| 12284 | @need 800 |
| 12285 | The @code{while} loop looks like this: |
| 12286 | |
| 12287 | @smallexample |
| 12288 | @group |
| 12289 | (while (> arg 0) |
| 12290 | (beginning-of-line) |
| 12291 | |
| 12292 | ;; @r{between paragraphs} |
| 12293 | (while (prog1 (and (not (eobp)) |
| 12294 | (looking-at paragraph-separate)) |
| 12295 | (forward-line 1))) |
| 12296 | @end group |
| 12297 | |
| 12298 | @group |
| 12299 | ;; @r{within paragraphs, with a fill prefix} |
| 12300 | (if fill-prefix-regexp |
| 12301 | ;; @r{There is a fill prefix; it overrides paragraph-start.} |
| 12302 | (while (and (not (eobp)) |
| 12303 | (not (looking-at paragraph-separate)) |
| 12304 | (looking-at fill-prefix-regexp)) |
| 12305 | (forward-line 1)) |
| 12306 | @end group |
| 12307 | |
| 12308 | @group |
| 12309 | ;; @r{within paragraphs, no fill prefix} |
| 12310 | (if (re-search-forward paragraph-start nil t) |
| 12311 | (goto-char (match-beginning 0)) |
| 12312 | (goto-char (point-max)))) |
| 12313 | |
| 12314 | (setq arg (1- arg))) |
| 12315 | @end group |
| 12316 | @end smallexample |
| 12317 | |
| 12318 | We can see immediately that this is a decrementing counter @code{while} |
| 12319 | loop, using the expression @code{(setq arg (1- arg))} as the decrementer. |
| 12320 | |
| 12321 | @need 800 |
| 12322 | The body of the loop consists of three expressions: |
| 12323 | |
| 12324 | @smallexample |
| 12325 | @group |
| 12326 | ;; @r{between paragraphs} |
| 12327 | (beginning-of-line) |
| 12328 | (while |
| 12329 | @var{body-of-while}) |
| 12330 | @end group |
| 12331 | |
| 12332 | @group |
| 12333 | ;; @r{within paragraphs, with fill prefix} |
| 12334 | (if @var{true-or-false-test} |
| 12335 | @var{then-part} |
| 12336 | @end group |
| 12337 | |
| 12338 | @group |
| 12339 | ;; @r{within paragraphs, no fill prefix} |
| 12340 | @var{else-part} |
| 12341 | @end group |
| 12342 | @end smallexample |
| 12343 | |
| 12344 | @noindent |
| 12345 | When the Emacs Lisp interpreter evaluates the body of the |
| 12346 | @code{while} loop, the first thing it does is evaluate the |
| 12347 | @code{(beginning-of-line)} expression and move point to the beginning |
| 12348 | of the line. Then there is an inner @code{while} loop. This |
| 12349 | @code{while} loop is designed to move the cursor out of the blank |
| 12350 | space between paragraphs, if it should happen to be there. Finally, |
| 12351 | there is an @code{if} expression that actually moves point to the end |
| 12352 | of the paragraph. |
| 12353 | |
| 12354 | @node fwd-para between paragraphs, fwd-para within paragraph, fwd-para while, forward-paragraph |
| 12355 | @unnumberedsubsec Between paragraphs |
| 12356 | |
| 12357 | First, let us look at the inner @code{while} loop. This loop handles |
| 12358 | the case when point is between paragraphs; it uses three functions |
| 12359 | that are new to us: @code{prog1}, @code{eobp} and @code{looking-at}. |
| 12360 | @findex prog1 |
| 12361 | @findex eobp |
| 12362 | @findex looking-at |
| 12363 | |
| 12364 | @itemize @bullet |
| 12365 | @item |
| 12366 | @code{prog1} is similar to the @code{progn} special form, |
| 12367 | except that @code{prog1} evaluates its arguments in sequence and then |
| 12368 | returns the value of its first argument as the value of the whole |
| 12369 | expression. (@code{progn} returns the value of its last argument as the |
| 12370 | value of the expression.) The second and subsequent arguments to |
| 12371 | @code{prog1} are evaluated only for their side effects. |
| 12372 | |
| 12373 | @item |
| 12374 | @code{eobp} is an abbreviation of @samp{End Of Buffer P} and is a |
| 12375 | function that returns true if point is at the end of the buffer. |
| 12376 | |
| 12377 | @item |
| 12378 | @code{looking-at} is a function that returns true if the text following |
| 12379 | point matches the regular expression passed @code{looking-at} as its |
| 12380 | argument. |
| 12381 | @end itemize |
| 12382 | |
| 12383 | @need 800 |
| 12384 | The @code{while} loop we are studying looks like this: |
| 12385 | |
| 12386 | @smallexample |
| 12387 | @group |
| 12388 | (while (prog1 (and (not (eobp)) |
| 12389 | (looking-at paragraph-separate)) |
| 12390 | (forward-line 1))) |
| 12391 | @end group |
| 12392 | @end smallexample |
| 12393 | |
| 12394 | @need 1200 |
| 12395 | @noindent |
| 12396 | This is a @code{while} loop with no body! The true-or-false-test of the |
| 12397 | loop is the expression: |
| 12398 | |
| 12399 | @smallexample |
| 12400 | @group |
| 12401 | (prog1 (and (not (eobp)) |
| 12402 | (looking-at paragraph-separate)) |
| 12403 | (forward-line 1)) |
| 12404 | @end group |
| 12405 | @end smallexample |
| 12406 | |
| 12407 | @noindent |
| 12408 | The first argument to the @code{prog1} is the @code{and} expression. It |
| 12409 | has within in it a test of whether point is at the end of the buffer and |
| 12410 | also a test of whether the pattern following point matches the regular |
| 12411 | expression for separating paragraphs. |
| 12412 | |
| 12413 | If the cursor is not at the end of the buffer and if the characters |
| 12414 | following the cursor mark the separation between two paragraphs, then |
| 12415 | the @code{and} expression is true. After evaluating the @code{and} |
| 12416 | expression, the Lisp interpreter evaluates the second argument to |
| 12417 | @code{prog1}, which is @code{forward-line}. This moves point forward |
| 12418 | one line. The value returned by the @code{prog1} however, is the |
| 12419 | value of its first argument, so the @code{while} loop continues so |
| 12420 | long as point is not at the end of the buffer and is between |
| 12421 | paragraphs. When, finally, point is moved to a paragraph, the |
| 12422 | @code{and} expression tests false. Note however, that the |
| 12423 | @code{forward-line} command is carried out anyhow. This means that |
| 12424 | when point is moved from between paragraphs to a paragraph, it is left |
| 12425 | at the beginning of the second line of the paragraph. |
| 12426 | |
| 12427 | @node fwd-para within paragraph, fwd-para no fill prefix, fwd-para between paragraphs, forward-paragraph |
| 12428 | @unnumberedsubsec Within paragraphs |
| 12429 | |
| 12430 | The next expression in the outer @code{while} loop is an @code{if} |
| 12431 | expression. The Lisp interpreter evaluates the then-part of the |
| 12432 | @code{if} when the @code{fill-prefix-regexp} variable has a value other |
| 12433 | than @code{nil}, and it evaluates the else-part when the value of |
| 12434 | @code{if fill-prefix-regexp} is @code{nil}, that is, when there is no |
| 12435 | fill prefix. |
| 12436 | |
| 12437 | @node fwd-para no fill prefix, fwd-para with fill prefix, fwd-para within paragraph, forward-paragraph |
| 12438 | @unnumberedsubsec No fill prefix |
| 12439 | |
| 12440 | It is simplest to look at the code for the case when there is no fill |
| 12441 | prefix first. This code consists of yet another inner @code{if} |
| 12442 | expression, and reads as follows: |
| 12443 | |
| 12444 | @smallexample |
| 12445 | @group |
| 12446 | (if (re-search-forward paragraph-start nil t) |
| 12447 | (goto-char (match-beginning 0)) |
| 12448 | (goto-char (point-max))) |
| 12449 | @end group |
| 12450 | @end smallexample |
| 12451 | |
| 12452 | @noindent |
| 12453 | This expression actually does the work that most people think of as |
| 12454 | the primary purpose of the @code{forward-paragraph} command: it causes |
| 12455 | a regular expression search to occur that searches forward to the |
| 12456 | start of the next paragraph and if it is found, moves point there; but |
| 12457 | if the start of another paragraph if not found, it moves point to the |
| 12458 | end of the accessible region of the buffer. |
| 12459 | |
| 12460 | The only unfamiliar part of this is the use of @code{match-beginning}. |
| 12461 | This is another function that is new to us. The |
| 12462 | @code{match-beginning} function returns a number specifying the |
| 12463 | location of the start of the text that was matched by the last regular |
| 12464 | expression search. |
| 12465 | |
| 12466 | The @code{match-beginning} function is used here because of a |
| 12467 | characteristic of a forward search: a successful forward search, |
| 12468 | regardless of whether it is a plain search or a regular expression |
| 12469 | search, will move point to the end of the text that is found. In this |
| 12470 | case, a successful search will move point to the end of the pattern for |
| 12471 | @code{paragraph-start}, which will be the beginning of the next |
| 12472 | paragraph rather than the end of the current one. |
| 12473 | |
| 12474 | However, we want to put point at the end of the current paragraph, not at |
| 12475 | the beginning of the next one. The two positions may be different, |
| 12476 | because there may be several blank lines between paragraphs. |
| 12477 | |
| 12478 | @findex match-beginning |
| 12479 | When given an argument of 0, @code{match-beginning} returns the position |
| 12480 | that is the start of the text that the most recent regular |
| 12481 | expression search matched. In this case, the most recent regular |
| 12482 | expression search is the one looking for @code{paragraph-start}, so |
| 12483 | @code{match-beginning} returns the beginning position of the pattern, |
| 12484 | rather than the end of the pattern. The beginning position is the end |
| 12485 | of the paragraph. |
| 12486 | |
| 12487 | (Incidentally, when passed a positive number as an argument, the |
| 12488 | @code{match-beginning} function will place point at that parenthesized |
| 12489 | expression in the last regular expression. It is a useful function.) |
| 12490 | |
| 12491 | @node fwd-para with fill prefix, fwd-para summary, fwd-para no fill prefix, forward-paragraph |
| 12492 | @unnumberedsubsec With a fill prefix |
| 12493 | |
| 12494 | The inner @code{if} expression just discussed is the else-part of an enclosing |
| 12495 | @code{if} expression which tests whether there is a fill prefix. If |
| 12496 | there is a fill prefix, the then-part of this @code{if} is evaluated. |
| 12497 | It looks like this: |
| 12498 | |
| 12499 | @smallexample |
| 12500 | @group |
| 12501 | (while (and (not (eobp)) |
| 12502 | (not (looking-at paragraph-separate)) |
| 12503 | (looking-at fill-prefix-regexp)) |
| 12504 | (forward-line 1)) |
| 12505 | @end group |
| 12506 | @end smallexample |
| 12507 | |
| 12508 | @noindent |
| 12509 | What this expression does is move point forward line by line so long |
| 12510 | as three conditions are true: |
| 12511 | |
| 12512 | @enumerate |
| 12513 | @item |
| 12514 | Point is not at the end of the buffer. |
| 12515 | |
| 12516 | @item |
| 12517 | The text following point does not separate paragraphs. |
| 12518 | |
| 12519 | @item |
| 12520 | The pattern following point is the fill prefix regular expression. |
| 12521 | @end enumerate |
| 12522 | |
| 12523 | The last condition may be puzzling, until you remember that point was |
| 12524 | moved to the beginning of the line early in the @code{forward-paragraph} |
| 12525 | function. This means that if the text has a fill prefix, the |
| 12526 | @code{looking-at} function will see it. |
| 12527 | |
| 12528 | @node fwd-para summary, , fwd-para with fill prefix, forward-paragraph |
| 12529 | @unnumberedsubsec Summary |
| 12530 | |
| 12531 | In summary, when moving forward, the @code{forward-paragraph} function |
| 12532 | does the following: |
| 12533 | |
| 12534 | @itemize @bullet |
| 12535 | @item |
| 12536 | Move point to the beginning of the line. |
| 12537 | |
| 12538 | @item |
| 12539 | Skip over lines between paragraphs. |
| 12540 | |
| 12541 | @item |
| 12542 | Check whether there is a fill prefix, and if there is: |
| 12543 | |
| 12544 | @itemize --- |
| 12545 | |
| 12546 | @item |
| 12547 | Go forward line by line so long as the line is not a paragraph |
| 12548 | separating line. |
| 12549 | @end itemize |
| 12550 | |
| 12551 | @item |
| 12552 | But if there is no fill prefix, |
| 12553 | |
| 12554 | @itemize --- |
| 12555 | |
| 12556 | @item |
| 12557 | Search for the next paragraph start pattern. |
| 12558 | |
| 12559 | @item |
| 12560 | Go to the beginning of the paragraph start pattern, which will be the |
| 12561 | end of the previous paragraph. |
| 12562 | |
| 12563 | @item |
| 12564 | Or else go to the end of the accessible portion of the buffer. |
| 12565 | @end itemize |
| 12566 | @end itemize |
| 12567 | |
| 12568 | @need 1200 |
| 12569 | For review, here is the code we have just been discussing, formatted |
| 12570 | for clarity: |
| 12571 | |
| 12572 | @smallexample |
| 12573 | @group |
| 12574 | (interactive "p") |
| 12575 | (or arg (setq arg 1)) |
| 12576 | (let* ( |
| 12577 | (fill-prefix-regexp |
| 12578 | (and fill-prefix (not (equal fill-prefix "")) |
| 12579 | (not paragraph-ignore-fill-prefix) |
| 12580 | (regexp-quote fill-prefix))) |
| 12581 | @end group |
| 12582 | |
| 12583 | @group |
| 12584 | (paragraph-separate |
| 12585 | (if fill-prefix-regexp |
| 12586 | (concat paragraph-separate |
| 12587 | "\\|^" |
| 12588 | fill-prefix-regexp |
| 12589 | "[ \t]*$") |
| 12590 | paragraph-separate))) |
| 12591 | |
| 12592 | @var{omitted-backward-moving-code} @dots{} |
| 12593 | @end group |
| 12594 | |
| 12595 | @group |
| 12596 | (while (> arg 0) ; @r{forward-moving-code} |
| 12597 | (beginning-of-line) |
| 12598 | |
| 12599 | (while (prog1 (and (not (eobp)) |
| 12600 | (looking-at paragraph-separate)) |
| 12601 | (forward-line 1))) |
| 12602 | @end group |
| 12603 | |
| 12604 | @group |
| 12605 | (if fill-prefix-regexp |
| 12606 | (while (and (not (eobp)) ; @r{then-part} |
| 12607 | (not (looking-at paragraph-separate)) |
| 12608 | (looking-at fill-prefix-regexp)) |
| 12609 | (forward-line 1)) |
| 12610 | @end group |
| 12611 | @group |
| 12612 | ; @r{else-part: the inner-if} |
| 12613 | (if (re-search-forward paragraph-start nil t) |
| 12614 | (goto-char (match-beginning 0)) |
| 12615 | (goto-char (point-max)))) |
| 12616 | |
| 12617 | (setq arg (1- arg))))) ; @r{decrementer} |
| 12618 | @end group |
| 12619 | @end smallexample |
| 12620 | |
| 12621 | The full definition for the @code{forward-paragraph} function not only |
| 12622 | includes this code for going forwards, but also code for going backwards. |
| 12623 | |
| 12624 | If you are reading this inside of GNU Emacs and you want to see the |
| 12625 | whole function, you can type @kbd{C-h f} (@code{describe-function}) |
| 12626 | and the name of the function. This gives you the function |
| 12627 | documentation and the name of the library containing the function's |
| 12628 | source. Place point over the name of the library and press the RET |
| 12629 | key; you will be taken directly to the source. (Be sure to install |
| 12630 | your sources! Without them, you are like a person who tries to drive |
| 12631 | a car with his eyes shut!) |
| 12632 | |
| 12633 | @c !!! again, 21.0.100 tags table location in this paragraph |
| 12634 | Or -- a good habit to get into -- you can type @kbd{M-.} |
| 12635 | (@code{find-tag}) and the name of the function when prompted for it. |
| 12636 | This will take you directly to the source. If the @code{find-tag} |
| 12637 | function first asks you for the name of a @file{TAGS} table, give it |
| 12638 | the name of the @file{TAGS} file such as |
| 12639 | @file{/usr/local/share/emacs/21.0.100/lisp/TAGS}. (The exact path to your |
| 12640 | @file{TAGS} file depends on how your copy of Emacs was installed.) |
| 12641 | |
| 12642 | You can also create your own @file{TAGS} file for directories that |
| 12643 | lack one. |
| 12644 | @ifnottex |
| 12645 | @xref{etags, , Create Your Own @file{TAGS} File}. |
| 12646 | @end ifnottex |
| 12647 | |
| 12648 | @node etags, Regexp Review, forward-paragraph, Regexp Search |
| 12649 | @section Create Your Own @file{TAGS} File |
| 12650 | @findex etags |
| 12651 | @cindex @file{TAGS} file, create own |
| 12652 | |
| 12653 | The @kbd{M-.} (@code{find-tag}) command takes you directly to the |
| 12654 | source for a function, variable, node, or other source. The function |
| 12655 | depends on tags tables to tell it where to go. |
| 12656 | |
| 12657 | You often need to build and install tags tables yourself. They are |
| 12658 | not built automatically. A tags table is called a @file{TAGS} file; |
| 12659 | the name is in upper case letters. |
| 12660 | |
| 12661 | You can create a @file{TAGS} file by calling the @code{etags} program |
| 12662 | that comes as a part of the Emacs distribution. Usually, @code{etags} |
| 12663 | is compiled and installed when Emacs is built. (@code{etags} is not |
| 12664 | an Emacs Lisp function or a part of Emacs; it is a C program.) |
| 12665 | |
| 12666 | @need 1250 |
| 12667 | To create a @file{TAGS} file, first switch to the directory in which |
| 12668 | you want to create the file. In Emacs you can do this with the |
| 12669 | @kbd{M-x cd} command, or by visiting a file in the directory, or by |
| 12670 | listing the directory with @kbd{C-x d} (@code{dired}). Then run the |
| 12671 | compile command, with @w{@code{etags *.el}} as the command to execute |
| 12672 | |
| 12673 | @smallexample |
| 12674 | M-x compile RET etags *.el RET |
| 12675 | @end smallexample |
| 12676 | |
| 12677 | @noindent |
| 12678 | to create a @file{TAGS} file. |
| 12679 | |
| 12680 | For example, if you have a large number of files in your |
| 12681 | @file{~/emacs} directory, as I do---I have 137 @file{.el} files in it, |
| 12682 | of which I load 12---you can create a @file{TAGS} file for the Emacs |
| 12683 | Lisp files in that directory. |
| 12684 | |
| 12685 | @need 1250 |
| 12686 | The @code{etags} program takes all the |
| 12687 | usual shell `wildcards'. For example, if you have two directories for |
| 12688 | which you want a single @file{TAGS file}, type |
| 12689 | @w{@code{etags *.el ../elisp/*.el}}, |
| 12690 | where @file{../elisp/} is the second directory: |
| 12691 | |
| 12692 | @smallexample |
| 12693 | M-x compile RET etags *.el ../elisp/*.el RET |
| 12694 | @end smallexample |
| 12695 | |
| 12696 | @need 1250 |
| 12697 | Type |
| 12698 | |
| 12699 | @smallexample |
| 12700 | M-x compile RET etags --help RET |
| 12701 | @end smallexample |
| 12702 | |
| 12703 | @noindent |
| 12704 | to see a list of the options accepted by @code{etags} as well as a |
| 12705 | list of supported languages. |
| 12706 | |
| 12707 | The @code{etags} program handles more than 20 languages, including |
| 12708 | Emacs Lisp, Common Lisp, Scheme, C, C++, Ada, Fortran, Java, LaTeX, |
| 12709 | Pascal, Perl, Python, Texinfo, makefiles, and most assemblers. The |
| 12710 | program has no switches for specifying the language; it recognizes the |
| 12711 | language in an input file according to its file name and contents. |
| 12712 | |
| 12713 | @file{etags} is very helpful when you are writing code yourself and |
| 12714 | want to refer back to functions you have already written. Just run |
| 12715 | @code{etags} again at intervals as you write new functions, so they |
| 12716 | become part of the @file{TAGS} file. |
| 12717 | |
| 12718 | If you think an appropriate @file{TAGS} file already exists for what |
| 12719 | you want, but do not know where it is, you can use the @code{locate} |
| 12720 | program to attempt to find it. |
| 12721 | |
| 12722 | Type @w{@kbd{M-x locate RET TAGS RET}} and Emacs will list for you the |
| 12723 | full path names of all your @file{TAGS} files. On my system, this |
| 12724 | command lists 34 @file{TAGS} files. On the other hand, a `plain |
| 12725 | vanilla' system I recently installed did not contain any @file{TAGS} |
| 12726 | files. |
| 12727 | |
| 12728 | If the tags table you want has been created, you can use the @code{M-x |
| 12729 | visit-tags-table} command to specify it. Otherwise, you will need to |
| 12730 | create the tag table yourself and then use @code{M-x |
| 12731 | visit-tags-table}. |
| 12732 | |
| 12733 | @subsubheading Building Tags in the Emacs sources |
| 12734 | @cindex Building Tags in the Emacs sources |
| 12735 | @cindex Tags in the Emacs sources |
| 12736 | @findex make tags |
| 12737 | |
| 12738 | The GNU Emacs sources come with a @file{Makefile} that contains a |
| 12739 | sophisticated @code{etags} command that creates, collects, and merges |
| 12740 | tags tables from all over the Emacs sources and puts the information |
| 12741 | into one @file{TAGS} file in the @file{src/} directory below the top |
| 12742 | level of your Emacs source directory. |
| 12743 | |
| 12744 | @need 1250 |
| 12745 | To build this @file{TAGS} file, go to the top level of your Emacs |
| 12746 | source directory and run the compile command @code{make tags}: |
| 12747 | |
| 12748 | @smallexample |
| 12749 | M-x compile RET make tags RET |
| 12750 | @end smallexample |
| 12751 | |
| 12752 | @noindent |
| 12753 | (The @code{make tags} command works well with the GNU Emacs sources, |
| 12754 | as well as with some other source packages.) |
| 12755 | |
| 12756 | For more information, see @ref{Tags, , Tag Tables, emacs, The GNU Emacs |
| 12757 | Manual}. |
| 12758 | |
| 12759 | @node Regexp Review, re-search Exercises, etags, Regexp Search |
| 12760 | @comment node-name, next, previous, up |
| 12761 | @section Review |
| 12762 | |
| 12763 | Here is a brief summary of some recently introduced functions. |
| 12764 | |
| 12765 | @table @code |
| 12766 | @item while |
| 12767 | Repeatedly evaluate the body of the expression so long as the first |
| 12768 | element of the body tests true. Then return @code{nil}. (The |
| 12769 | expression is evaluated only for its side effects.) |
| 12770 | |
| 12771 | @need 1250 |
| 12772 | For example: |
| 12773 | |
| 12774 | @smallexample |
| 12775 | @group |
| 12776 | (let ((foo 2)) |
| 12777 | (while (> foo 0) |
| 12778 | (insert (format "foo is %d.\n" foo)) |
| 12779 | (setq foo (1- foo)))) |
| 12780 | |
| 12781 | @result{} foo is 2. |
| 12782 | foo is 1. |
| 12783 | nil |
| 12784 | @end group |
| 12785 | @end smallexample |
| 12786 | @noindent |
| 12787 | (The @code{insert} function inserts its arguments at point; the |
| 12788 | @code{format} function returns a string formatted from its arguments |
| 12789 | the way @code{message} formats its arguments; @code{\n} produces a new |
| 12790 | line.) |
| 12791 | |
| 12792 | @item re-search-forward |
| 12793 | Search for a pattern, and if the pattern is found, move point to rest |
| 12794 | just after it. |
| 12795 | |
| 12796 | @noindent |
| 12797 | Takes four arguments, like @code{search-forward}: |
| 12798 | |
| 12799 | @enumerate |
| 12800 | @item |
| 12801 | A regular expression that specifies the pattern to search for. |
| 12802 | |
| 12803 | @item |
| 12804 | Optionally, the limit of the search. |
| 12805 | |
| 12806 | @item |
| 12807 | Optionally, what to do if the search fails, return @code{nil} or an |
| 12808 | error message. |
| 12809 | |
| 12810 | @item |
| 12811 | Optionally, how many times to repeat the search; if negative, the |
| 12812 | search goes backwards. |
| 12813 | @end enumerate |
| 12814 | |
| 12815 | @item let* |
| 12816 | Bind some variables locally to particular values, |
| 12817 | and then evaluate the remaining arguments, returning the value of the |
| 12818 | last one. While binding the local variables, use the local values of |
| 12819 | variables bound earlier, if any. |
| 12820 | |
| 12821 | @need 1250 |
| 12822 | For example: |
| 12823 | |
| 12824 | @smallexample |
| 12825 | @group |
| 12826 | (let* ((foo 7) |
| 12827 | (bar (* 3 foo))) |
| 12828 | (message "`bar' is %d." bar)) |
| 12829 | @result{} `bar' is 21. |
| 12830 | @end group |
| 12831 | @end smallexample |
| 12832 | |
| 12833 | @item match-beginning |
| 12834 | Return the position of the start of the text found by the last regular |
| 12835 | expression search. |
| 12836 | |
| 12837 | @item looking-at |
| 12838 | Return @code{t} for true if the text after point matches the argument, |
| 12839 | which should be a regular expression. |
| 12840 | |
| 12841 | @item eobp |
| 12842 | Return @code{t} for true if point is at the end of the accessible part |
| 12843 | of a buffer. The end of the accessible part is the end of the buffer |
| 12844 | if the buffer is not narrowed; it is the end of the narrowed part if |
| 12845 | the buffer is narrowed. |
| 12846 | |
| 12847 | @item prog1 |
| 12848 | Evaluate each argument in sequence and then return the value of the |
| 12849 | @emph{first}. |
| 12850 | |
| 12851 | @need 1250 |
| 12852 | For example: |
| 12853 | |
| 12854 | @smallexample |
| 12855 | @group |
| 12856 | (prog1 1 2 3 4) |
| 12857 | @result{} 1 |
| 12858 | @end group |
| 12859 | @end smallexample |
| 12860 | @end table |
| 12861 | |
| 12862 | @need 1500 |
| 12863 | @node re-search Exercises, , Regexp Review, Regexp Search |
| 12864 | @section Exercises with @code{re-search-forward} |
| 12865 | |
| 12866 | @itemize @bullet |
| 12867 | @item |
| 12868 | Write a function to search for a regular expression that matches two |
| 12869 | or more blank lines in sequence. |
| 12870 | |
| 12871 | @item |
| 12872 | Write a function to search for duplicated words, such as `the the'. |
| 12873 | @xref{Regexps, , Syntax of Regular Expressions, emacs, The GNU Emacs |
| 12874 | Manual}, for information on how to write a regexp (a regular |
| 12875 | expression) to match a string that is composed of two identical |
| 12876 | halves. You can devise several regexps; some are better than others. |
| 12877 | The function I use is described in an appendix, along with several |
| 12878 | regexps. @xref{the-the, , @code{the-the} Duplicated Words Function}. |
| 12879 | @end itemize |
| 12880 | |
| 12881 | @node Counting Words, Words in a defun, Regexp Search, Top |
| 12882 | @chapter Counting: Repetition and Regexps |
| 12883 | @cindex Repetition for word counting |
| 12884 | @cindex Regular expressions for word counting |
| 12885 | |
| 12886 | Repetition and regular expression searches are powerful tools that you |
| 12887 | often use when you write code in Emacs Lisp. This chapter illustrates |
| 12888 | the use of regular expression searches through the construction of |
| 12889 | word count commands using @code{while} loops and recursion. |
| 12890 | |
| 12891 | @menu |
| 12892 | * Why Count Words:: |
| 12893 | * count-words-region:: Use a regexp, but find a problem. |
| 12894 | * recursive-count-words:: Start with case of no words in region. |
| 12895 | * Counting Exercise:: |
| 12896 | @end menu |
| 12897 | |
| 12898 | @node Why Count Words, count-words-region, Counting Words, Counting Words |
| 12899 | @ifnottex |
| 12900 | @unnumberedsec Counting words |
| 12901 | @end ifnottex |
| 12902 | |
| 12903 | The standard Emacs distribution contains a function for counting the |
| 12904 | number of lines within a region. However, there is no corresponding |
| 12905 | function for counting words. |
| 12906 | |
| 12907 | Certain types of writing ask you to count words. Thus, if you write |
| 12908 | an essay, you may be limited to 800 words; if you write a novel, you |
| 12909 | may discipline yourself to write 1000 words a day. It seems odd to me |
| 12910 | that Emacs lacks a word count command. Perhaps people use Emacs |
| 12911 | mostly for code or types of documentation that do not require word |
| 12912 | counts; or perhaps they restrict themselves to the operating system |
| 12913 | word count command, @code{wc}. Alternatively, people may follow |
| 12914 | the publishers' convention and compute a word count by dividing the |
| 12915 | number of characters in a document by five. In any event, here are |
| 12916 | commands to count words. |
| 12917 | |
| 12918 | @node count-words-region, recursive-count-words, Why Count Words, Counting Words |
| 12919 | @comment node-name, next, previous, up |
| 12920 | @section The @code{count-words-region} Function |
| 12921 | @findex count-words-region |
| 12922 | |
| 12923 | A word count command could count words in a line, paragraph, region, |
| 12924 | or buffer. What should the command cover? You could design the |
| 12925 | command to count the number of words in a complete buffer. However, |
| 12926 | the Emacs tradition encourages flexibility---you may want to count |
| 12927 | words in just a section, rather than all of a buffer. So it makes |
| 12928 | more sense to design the command to count the number of words in a |
| 12929 | region. Once you have a @code{count-words-region} command, you can, |
| 12930 | if you wish, count words in a whole buffer by marking it with @kbd{C-x |
| 12931 | h} (@code{mark-whole-buffer}). |
| 12932 | |
| 12933 | Clearly, counting words is a repetitive act: starting from the |
| 12934 | beginning of the region, you count the first word, then the second |
| 12935 | word, then the third word, and so on, until you reach the end of the |
| 12936 | region. This means that word counting is ideally suited to recursion |
| 12937 | or to a @code{while} loop. |
| 12938 | |
| 12939 | @menu |
| 12940 | * Design count-words-region:: The definition using a @code{while} loop. |
| 12941 | * Whitespace Bug:: The Whitespace Bug in @code{count-words-region}. |
| 12942 | @end menu |
| 12943 | |
| 12944 | @node Design count-words-region, Whitespace Bug, count-words-region, count-words-region |
| 12945 | @ifnottex |
| 12946 | @unnumberedsubsec Designing @code{count-words-region} |
| 12947 | @end ifnottex |
| 12948 | |
| 12949 | First, we will implement the word count command with a @code{while} |
| 12950 | loop, then with recursion. The command will, of course, be |
| 12951 | interactive. |
| 12952 | |
| 12953 | @need 800 |
| 12954 | The template for an interactive function definition is, as always: |
| 12955 | |
| 12956 | @smallexample |
| 12957 | @group |
| 12958 | (defun @var{name-of-function} (@var{argument-list}) |
| 12959 | "@var{documentation}@dots{}" |
| 12960 | (@var{interactive-expression}@dots{}) |
| 12961 | @var{body}@dots{}) |
| 12962 | @end group |
| 12963 | @end smallexample |
| 12964 | |
| 12965 | What we need to do is fill in the slots. |
| 12966 | |
| 12967 | The name of the function should be self-explanatory and similar to the |
| 12968 | existing @code{count-lines-region} name. This makes the name easier |
| 12969 | to remember. @code{count-words-region} is a good choice. |
| 12970 | |
| 12971 | The function counts words within a region. This means that the |
| 12972 | argument list must contain symbols that are bound to the two |
| 12973 | positions, the beginning and end of the region. These two positions |
| 12974 | can be called @samp{beginning} and @samp{end} respectively. The first |
| 12975 | line of the documentation should be a single sentence, since that is |
| 12976 | all that is printed as documentation by a command such as |
| 12977 | @code{apropos}. The interactive expression will be of the form |
| 12978 | @samp{(interactive "r")}, since that will cause Emacs to pass the |
| 12979 | beginning and end of the region to the function's argument list. All |
| 12980 | this is routine. |
| 12981 | |
| 12982 | The body of the function needs to be written to do three tasks: |
| 12983 | first, to set up conditions under which the @code{while} loop can |
| 12984 | count words, second, to run the @code{while} loop, and third, to send |
| 12985 | a message to the user. |
| 12986 | |
| 12987 | When a user calls @code{count-words-region}, point may be at the |
| 12988 | beginning or the end of the region. However, the counting process |
| 12989 | must start at the beginning of the region. This means we will want |
| 12990 | to put point there if it is not already there. Executing |
| 12991 | @code{(goto-char beginning)} ensures this. Of course, we will want to |
| 12992 | return point to its expected position when the function finishes its |
| 12993 | work. For this reason, the body must be enclosed in a |
| 12994 | @code{save-excursion} expression. |
| 12995 | |
| 12996 | The central part of the body of the function consists of a |
| 12997 | @code{while} loop in which one expression jumps point forward word by |
| 12998 | word, and another expression counts those jumps. The true-or-false-test |
| 12999 | of the @code{while} loop should test true so long as point should jump |
| 13000 | forward, and false when point is at the end of the region. |
| 13001 | |
| 13002 | We could use @code{(forward-word 1)} as the expression for moving point |
| 13003 | forward word by word, but it is easier to see what Emacs identifies as a |
| 13004 | `word' if we use a regular expression search. |
| 13005 | |
| 13006 | A regular expression search that finds the pattern for which it is |
| 13007 | searching leaves point after the last character matched. This means |
| 13008 | that a succession of successful word searches will move point forward |
| 13009 | word by word. |
| 13010 | |
| 13011 | As a practical matter, we want the regular expression search to jump |
| 13012 | over whitespace and punctuation between words as well as over the |
| 13013 | words themselves. A regexp that refuses to jump over interword |
| 13014 | whitespace would never jump more than one word! This means that |
| 13015 | the regexp should include the whitespace and punctuation that follows |
| 13016 | a word, if any, as well as the word itself. (A word may end a buffer |
| 13017 | and not have any following whitespace or punctuation, so that part of |
| 13018 | the regexp must be optional.) |
| 13019 | |
| 13020 | Thus, what we want for the regexp is a pattern defining one or more |
| 13021 | word constituent characters followed, optionally, by one or more |
| 13022 | characters that are not word constituents. The regular expression for |
| 13023 | this is: |
| 13024 | |
| 13025 | @smallexample |
| 13026 | \w+\W* |
| 13027 | @end smallexample |
| 13028 | |
| 13029 | @noindent |
| 13030 | The buffer's syntax table determines which characters are and are not |
| 13031 | word constituents. (@xref{Syntax, , What Constitutes a Word or |
| 13032 | Symbol?}, for more about syntax. Also, see @ref{Syntax, Syntax, The |
| 13033 | Syntax Table, emacs, The GNU Emacs Manual}, and @ref{Syntax Tables, , |
| 13034 | Syntax Tables, elisp, The GNU Emacs Lisp Reference Manual}.) |
| 13035 | |
| 13036 | @need 800 |
| 13037 | The search expression looks like this: |
| 13038 | |
| 13039 | @smallexample |
| 13040 | (re-search-forward "\\w+\\W*") |
| 13041 | @end smallexample |
| 13042 | |
| 13043 | @noindent |
| 13044 | (Note that paired backslashes precede the @samp{w} and @samp{W}. A |
| 13045 | single backslash has special meaning to the Emacs Lisp interpreter. It |
| 13046 | indicates that the following character is interpreted differently than |
| 13047 | usual. For example, the two characters, @samp{\n}, stand for |
| 13048 | @samp{newline}, rather than for a backslash followed by @samp{n}. Two |
| 13049 | backslashes in a row stand for an ordinary, `unspecial' backslash.) |
| 13050 | |
| 13051 | We need a counter to count how many words there are; this variable |
| 13052 | must first be set to 0 and then incremented each time Emacs goes |
| 13053 | around the @code{while} loop. The incrementing expression is simply: |
| 13054 | |
| 13055 | @smallexample |
| 13056 | (setq count (1+ count)) |
| 13057 | @end smallexample |
| 13058 | |
| 13059 | Finally, we want to tell the user how many words there are in the |
| 13060 | region. The @code{message} function is intended for presenting this |
| 13061 | kind of information to the user. The message has to be phrased so |
| 13062 | that it reads properly regardless of how many words there are in the |
| 13063 | region: we don't want to say that ``there are 1 words in the region''. |
| 13064 | The conflict between singular and plural is ungrammatical. We can |
| 13065 | solve this problem by using a conditional expression that evaluates |
| 13066 | different messages depending on the number of words in the region. |
| 13067 | There are three possibilities: no words in the region, one word in the |
| 13068 | region, and more than one word. This means that the @code{cond} |
| 13069 | special form is appropriate. |
| 13070 | |
| 13071 | @need 1500 |
| 13072 | All this leads to the following function definition: |
| 13073 | |
| 13074 | @smallexample |
| 13075 | @group |
| 13076 | ;;; @r{First version; has bugs!} |
| 13077 | (defun count-words-region (beginning end) |
| 13078 | "Print number of words in the region. |
| 13079 | Words are defined as at least one word-constituent |
| 13080 | character followed by at least one character that |
| 13081 | is not a word-constituent. The buffer's syntax |
| 13082 | table determines which characters these are." |
| 13083 | (interactive "r") |
| 13084 | (message "Counting words in region ... ") |
| 13085 | @end group |
| 13086 | |
| 13087 | @group |
| 13088 | ;;; @r{1. Set up appropriate conditions.} |
| 13089 | (save-excursion |
| 13090 | (goto-char beginning) |
| 13091 | (let ((count 0)) |
| 13092 | @end group |
| 13093 | |
| 13094 | @group |
| 13095 | ;;; @r{2. Run the} while @r{loop.} |
| 13096 | (while (< (point) end) |
| 13097 | (re-search-forward "\\w+\\W*") |
| 13098 | (setq count (1+ count))) |
| 13099 | @end group |
| 13100 | |
| 13101 | @group |
| 13102 | ;;; @r{3. Send a message to the user.} |
| 13103 | (cond ((zerop count) |
| 13104 | (message |
| 13105 | "The region does NOT have any words.")) |
| 13106 | ((= 1 count) |
| 13107 | (message |
| 13108 | "The region has 1 word.")) |
| 13109 | (t |
| 13110 | (message |
| 13111 | "The region has %d words." count)))))) |
| 13112 | @end group |
| 13113 | @end smallexample |
| 13114 | |
| 13115 | @noindent |
| 13116 | As written, the function works, but not in all circumstances. |
| 13117 | |
| 13118 | @node Whitespace Bug, , Design count-words-region, count-words-region |
| 13119 | @comment node-name, next, previous, up |
| 13120 | @subsection The Whitespace Bug in @code{count-words-region} |
| 13121 | |
| 13122 | The @code{count-words-region} command described in the preceding |
| 13123 | section has two bugs, or rather, one bug with two manifestations. |
| 13124 | First, if you mark a region containing only whitespace in the middle |
| 13125 | of some text, the @code{count-words-region} command tells you that the |
| 13126 | region contains one word! Second, if you mark a region containing |
| 13127 | only whitespace at the end of the buffer or the accessible portion of |
| 13128 | a narrowed buffer, the command displays an error message that looks |
| 13129 | like this: |
| 13130 | |
| 13131 | @smallexample |
| 13132 | Search failed: "\\w+\\W*" |
| 13133 | @end smallexample |
| 13134 | |
| 13135 | If you are reading this in Info in GNU Emacs, you can test for these |
| 13136 | bugs yourself. |
| 13137 | |
| 13138 | First, evaluate the function in the usual manner to install it. |
| 13139 | @ifinfo |
| 13140 | Here is a copy of the definition. Place your cursor after the closing |
| 13141 | parenthesis and type @kbd{C-x C-e} to install it. |
| 13142 | |
| 13143 | @smallexample |
| 13144 | @group |
| 13145 | ;; @r{First version; has bugs!} |
| 13146 | (defun count-words-region (beginning end) |
| 13147 | "Print number of words in the region. |
| 13148 | Words are defined as at least one word-constituent character followed |
| 13149 | by at least one character that is not a word-constituent. The buffer's |
| 13150 | syntax table determines which characters these are." |
| 13151 | @end group |
| 13152 | @group |
| 13153 | (interactive "r") |
| 13154 | (message "Counting words in region ... ") |
| 13155 | @end group |
| 13156 | |
| 13157 | @group |
| 13158 | ;;; @r{1. Set up appropriate conditions.} |
| 13159 | (save-excursion |
| 13160 | (goto-char beginning) |
| 13161 | (let ((count 0)) |
| 13162 | @end group |
| 13163 | |
| 13164 | @group |
| 13165 | ;;; @r{2. Run the} while @r{loop.} |
| 13166 | (while (< (point) end) |
| 13167 | (re-search-forward "\\w+\\W*") |
| 13168 | (setq count (1+ count))) |
| 13169 | @end group |
| 13170 | |
| 13171 | @group |
| 13172 | ;;; @r{3. Send a message to the user.} |
| 13173 | (cond ((zerop count) |
| 13174 | (message "The region does NOT have any words.")) |
| 13175 | ((= 1 count) (message "The region has 1 word.")) |
| 13176 | (t (message "The region has %d words." count)))))) |
| 13177 | @end group |
| 13178 | @end smallexample |
| 13179 | @end ifinfo |
| 13180 | |
| 13181 | @need 1000 |
| 13182 | If you wish, you can also install this keybinding by evaluating it: |
| 13183 | |
| 13184 | @smallexample |
| 13185 | (global-set-key "\C-c=" 'count-words-region) |
| 13186 | @end smallexample |
| 13187 | |
| 13188 | To conduct the first test, set mark and point to the beginning and end |
| 13189 | of the following line and then type @kbd{C-c =} (or @kbd{M-x |
| 13190 | count-words-region} if you have not bound @kbd{C-c =}): |
| 13191 | |
| 13192 | @smallexample |
| 13193 | one two three |
| 13194 | @end smallexample |
| 13195 | |
| 13196 | @noindent |
| 13197 | Emacs will tell you, correctly, that the region has three words. |
| 13198 | |
| 13199 | Repeat the test, but place mark at the beginning of the line and place |
| 13200 | point just @emph{before} the word @samp{one}. Again type the command |
| 13201 | @kbd{C-c =} (or @kbd{M-x count-words-region}). Emacs should tell you |
| 13202 | that the region has no words, since it is composed only of the |
| 13203 | whitespace at the beginning of the line. But instead Emacs tells you |
| 13204 | that the region has one word! |
| 13205 | |
| 13206 | For the third test, copy the sample line to the end of the |
| 13207 | @file{*scratch*} buffer and then type several spaces at the end of the |
| 13208 | line. Place mark right after the word @samp{three} and point at the |
| 13209 | end of line. (The end of the line will be the end of the buffer.) |
| 13210 | Type @kbd{C-c =} (or @kbd{M-x count-words-region}) as you did before. |
| 13211 | Again, Emacs should tell you that the region has no words, since it is |
| 13212 | composed only of the whitespace at the end of the line. Instead, |
| 13213 | Emacs displays an error message saying @samp{Search failed}. |
| 13214 | |
| 13215 | The two bugs stem from the same problem. |
| 13216 | |
| 13217 | Consider the first manifestation of the bug, in which the command |
| 13218 | tells you that the whitespace at the beginning of the line contains |
| 13219 | one word. What happens is this: The @code{M-x count-words-region} |
| 13220 | command moves point to the beginning of the region. The @code{while} |
| 13221 | tests whether the value of point is smaller than the value of |
| 13222 | @code{end}, which it is. Consequently, the regular expression search |
| 13223 | looks for and finds the first word. It leaves point after the word. |
| 13224 | @code{count} is set to one. The @code{while} loop repeats; but this |
| 13225 | time the value of point is larger than the value of @code{end}, the |
| 13226 | loop is exited; and the function displays a message saying the number |
| 13227 | of words in the region is one. In brief, the regular expression |
| 13228 | search looks for and finds the word even though it is outside |
| 13229 | the marked region. |
| 13230 | |
| 13231 | In the second manifestation of the bug, the region is whitespace at |
| 13232 | the end of the buffer. Emacs says @samp{Search failed}. What happens |
| 13233 | is that the true-or-false-test in the @code{while} loop tests true, so |
| 13234 | the search expression is executed. But since there are no more words |
| 13235 | in the buffer, the search fails. |
| 13236 | |
| 13237 | In both manifestations of the bug, the search extends or attempts to |
| 13238 | extend outside of the region. |
| 13239 | |
| 13240 | The solution is to limit the search to the region---this is a fairly |
| 13241 | simple action, but as you may have come to expect, it is not quite as |
| 13242 | simple as you might think. |
| 13243 | |
| 13244 | As we have seen, the @code{re-search-forward} function takes a search |
| 13245 | pattern as its first argument. But in addition to this first, |
| 13246 | mandatory argument, it accepts three optional arguments. The optional |
| 13247 | second argument bounds the search. The optional third argument, if |
| 13248 | @code{t}, causes the function to return @code{nil} rather than signal |
| 13249 | an error if the search fails. The optional fourth argument is a |
| 13250 | repeat count. (In Emacs, you can see a function's documentation by |
| 13251 | typing @kbd{C-h f}, the name of the function, and then @key{RET}.) |
| 13252 | |
| 13253 | In the @code{count-words-region} definition, the value of the end of |
| 13254 | the region is held by the variable @code{end} which is passed as an |
| 13255 | argument to the function. Thus, we can add @code{end} as an argument |
| 13256 | to the regular expression search expression: |
| 13257 | |
| 13258 | @smallexample |
| 13259 | (re-search-forward "\\w+\\W*" end) |
| 13260 | @end smallexample |
| 13261 | |
| 13262 | However, if you make only this change to the @code{count-words-region} |
| 13263 | definition and then test the new version of the definition on a |
| 13264 | stretch of whitespace, you will receive an error message saying |
| 13265 | @samp{Search failed}. |
| 13266 | |
| 13267 | What happens is this: the search is limited to the region, and fails |
| 13268 | as you expect because there are no word-constituent characters in the |
| 13269 | region. Since it fails, we receive an error message. But we do not |
| 13270 | want to receive an error message in this case; we want to receive the |
| 13271 | message that "The region does NOT have any words." |
| 13272 | |
| 13273 | The solution to this problem is to provide @code{re-search-forward} |
| 13274 | with a third argument of @code{t}, which causes the function to return |
| 13275 | @code{nil} rather than signal an error if the search fails. |
| 13276 | |
| 13277 | However, if you make this change and try it, you will see the message |
| 13278 | ``Counting words in region ... '' and @dots{} you will keep on seeing |
| 13279 | that message @dots{}, until you type @kbd{C-g} (@code{keyboard-quit}). |
| 13280 | |
| 13281 | Here is what happens: the search is limited to the region, as before, |
| 13282 | and it fails because there are no word-constituent characters in the |
| 13283 | region, as expected. Consequently, the @code{re-search-forward} |
| 13284 | expression returns @code{nil}. It does nothing else. In particular, |
| 13285 | it does not move point, which it does as a side effect if it finds the |
| 13286 | search target. After the @code{re-search-forward} expression returns |
| 13287 | @code{nil}, the next expression in the @code{while} loop is evaluated. |
| 13288 | This expression increments the count. Then the loop repeats. The |
| 13289 | true-or-false-test tests true because the value of point is still less |
| 13290 | than the value of end, since the @code{re-search-forward} expression |
| 13291 | did not move point. @dots{} and the cycle repeats @dots{} |
| 13292 | |
| 13293 | The @code{count-words-region} definition requires yet another |
| 13294 | modification, to cause the true-or-false-test of the @code{while} loop |
| 13295 | to test false if the search fails. Put another way, there are two |
| 13296 | conditions that must be satisfied in the true-or-false-test before the |
| 13297 | word count variable is incremented: point must still be within the |
| 13298 | region and the search expression must have found a word to count. |
| 13299 | |
| 13300 | Since both the first condition and the second condition must be true |
| 13301 | together, the two expressions, the region test and the search |
| 13302 | expression, can be joined with an @code{and} special form and embedded in |
| 13303 | the @code{while} loop as the true-or-false-test, like this: |
| 13304 | |
| 13305 | @smallexample |
| 13306 | (and (< (point) end) (re-search-forward "\\w+\\W*" end t)) |
| 13307 | @end smallexample |
| 13308 | |
| 13309 | @c colon in printed section title causes problem in Info cross reference |
| 13310 | @c also trouble with an overfull hbox |
| 13311 | @iftex |
| 13312 | @noindent |
| 13313 | (For information about @code{and}, see |
| 13314 | @ref{forward-paragraph, , @code{forward-paragraph}: a Goldmine of |
| 13315 | Functions}.) |
| 13316 | @end iftex |
| 13317 | @ifinfo |
| 13318 | @noindent |
| 13319 | (@xref{forward-paragraph}, for information about @code{and}.) |
| 13320 | @end ifinfo |
| 13321 | |
| 13322 | The @code{re-search-forward} expression returns @code{t} if the search |
| 13323 | succeeds and as a side effect moves point. Consequently, as words are |
| 13324 | found, point is moved through the region. When the search |
| 13325 | expression fails to find another word, or when point reaches the end |
| 13326 | of the region, the true-or-false-test tests false, the @code{while} |
| 13327 | loop exists, and the @code{count-words-region} function displays one |
| 13328 | or other of its messages. |
| 13329 | |
| 13330 | After incorporating these final changes, the @code{count-words-region} |
| 13331 | works without bugs (or at least, without bugs that I have found!). |
| 13332 | Here is what it looks like: |
| 13333 | |
| 13334 | @smallexample |
| 13335 | @group |
| 13336 | ;;; @r{Final version:} @code{while} |
| 13337 | (defun count-words-region (beginning end) |
| 13338 | "Print number of words in the region." |
| 13339 | (interactive "r") |
| 13340 | (message "Counting words in region ... ") |
| 13341 | @end group |
| 13342 | |
| 13343 | @group |
| 13344 | ;;; @r{1. Set up appropriate conditions.} |
| 13345 | (save-excursion |
| 13346 | (let ((count 0)) |
| 13347 | (goto-char beginning) |
| 13348 | @end group |
| 13349 | |
| 13350 | @group |
| 13351 | ;;; @r{2. Run the} while @r{loop.} |
| 13352 | (while (and (< (point) end) |
| 13353 | (re-search-forward "\\w+\\W*" end t)) |
| 13354 | (setq count (1+ count))) |
| 13355 | @end group |
| 13356 | |
| 13357 | @group |
| 13358 | ;;; @r{3. Send a message to the user.} |
| 13359 | (cond ((zerop count) |
| 13360 | (message |
| 13361 | "The region does NOT have any words.")) |
| 13362 | ((= 1 count) |
| 13363 | (message |
| 13364 | "The region has 1 word.")) |
| 13365 | (t |
| 13366 | (message |
| 13367 | "The region has %d words." count)))))) |
| 13368 | @end group |
| 13369 | @end smallexample |
| 13370 | |
| 13371 | @node recursive-count-words, Counting Exercise, count-words-region, Counting Words |
| 13372 | @comment node-name, next, previous, up |
| 13373 | @section Count Words Recursively |
| 13374 | @cindex Count words recursively |
| 13375 | @cindex Recursively counting words |
| 13376 | @cindex Words, counted recursively |
| 13377 | |
| 13378 | You can write the function for counting words recursively as well as |
| 13379 | with a @code{while} loop. Let's see how this is done. |
| 13380 | |
| 13381 | First, we need to recognize that the @code{count-words-region} |
| 13382 | function has three jobs: it sets up the appropriate conditions for |
| 13383 | counting to occur; it counts the words in the region; and it sends a |
| 13384 | message to the user telling how many words there are. |
| 13385 | |
| 13386 | If we write a single recursive function to do everything, we will |
| 13387 | receive a message for every recursive call. If the region contains 13 |
| 13388 | words, we will receive thirteen messages, one right after the other. |
| 13389 | We don't want this! Instead, we must write two functions to do the |
| 13390 | job, one of which (the recursive function) will be used inside of the |
| 13391 | other. One function will set up the conditions and display the |
| 13392 | message; the other will return the word count. |
| 13393 | |
| 13394 | Let us start with the function that causes the message to be displayed. |
| 13395 | We can continue to call this @code{count-words-region}. |
| 13396 | |
| 13397 | This is the function that the user will call. It will be interactive. |
| 13398 | Indeed, it will be similar to our previous versions of this |
| 13399 | function, except that it will call @code{recursive-count-words} to |
| 13400 | determine how many words are in the region. |
| 13401 | |
| 13402 | @need 1250 |
| 13403 | We can readily construct a template for this function, based on our |
| 13404 | previous versions: |
| 13405 | |
| 13406 | @smallexample |
| 13407 | @group |
| 13408 | ;; @r{Recursive version; uses regular expression search} |
| 13409 | (defun count-words-region (beginning end) |
| 13410 | "@var{documentation}@dots{}" |
| 13411 | (@var{interactive-expression}@dots{}) |
| 13412 | @end group |
| 13413 | @group |
| 13414 | |
| 13415 | ;;; @r{1. Set up appropriate conditions.} |
| 13416 | (@var{explanatory message}) |
| 13417 | (@var{set-up functions}@dots{} |
| 13418 | @end group |
| 13419 | @group |
| 13420 | |
| 13421 | ;;; @r{2. Count the words.} |
| 13422 | @var{recursive call} |
| 13423 | @end group |
| 13424 | @group |
| 13425 | |
| 13426 | ;;; @r{3. Send a message to the user.} |
| 13427 | @var{message providing word count})) |
| 13428 | @end group |
| 13429 | @end smallexample |
| 13430 | |
| 13431 | The definition looks straightforward, except that somehow the count |
| 13432 | returned by the recursive call must be passed to the message |
| 13433 | displaying the word count. A little thought suggests that this can be |
| 13434 | done by making use of a @code{let} expression: we can bind a variable |
| 13435 | in the varlist of a @code{let} expression to the number of words in |
| 13436 | the region, as returned by the recursive call; and then the |
| 13437 | @code{cond} expression, using binding, can display the value to the |
| 13438 | user. |
| 13439 | |
| 13440 | Often, one thinks of the binding within a @code{let} expression as |
| 13441 | somehow secondary to the `primary' work of a function. But in this |
| 13442 | case, what you might consider the `primary' job of the function, |
| 13443 | counting words, is done within the @code{let} expression. |
| 13444 | |
| 13445 | @need 1250 |
| 13446 | Using @code{let}, the function definition looks like this: |
| 13447 | |
| 13448 | @smallexample |
| 13449 | @group |
| 13450 | (defun count-words-region (beginning end) |
| 13451 | "Print number of words in the region." |
| 13452 | (interactive "r") |
| 13453 | @end group |
| 13454 | |
| 13455 | @group |
| 13456 | ;;; @r{1. Set up appropriate conditions.} |
| 13457 | (message "Counting words in region ... ") |
| 13458 | (save-excursion |
| 13459 | (goto-char beginning) |
| 13460 | @end group |
| 13461 | |
| 13462 | @group |
| 13463 | ;;; @r{2. Count the words.} |
| 13464 | (let ((count (recursive-count-words end))) |
| 13465 | @end group |
| 13466 | |
| 13467 | @group |
| 13468 | ;;; @r{3. Send a message to the user.} |
| 13469 | (cond ((zerop count) |
| 13470 | (message |
| 13471 | "The region does NOT have any words.")) |
| 13472 | ((= 1 count) |
| 13473 | (message |
| 13474 | "The region has 1 word.")) |
| 13475 | (t |
| 13476 | (message |
| 13477 | "The region has %d words." count)))))) |
| 13478 | @end group |
| 13479 | @end smallexample |
| 13480 | |
| 13481 | Next, we need to write the recursive counting function. |
| 13482 | |
| 13483 | A recursive function has at least three parts: the `do-again-test', the |
| 13484 | `next-step-expression', and the recursive call. |
| 13485 | |
| 13486 | The do-again-test determines whether the function will or will not be |
| 13487 | called again. Since we are counting words in a region and can use a |
| 13488 | function that moves point forward for every word, the do-again-test |
| 13489 | can check whether point is still within the region. The do-again-test |
| 13490 | should find the value of point and determine whether point is before, |
| 13491 | at, or after the value of the end of the region. We can use the |
| 13492 | @code{point} function to locate point. Clearly, we must pass the |
| 13493 | value of the end of the region to the recursive counting function as an |
| 13494 | argument. |
| 13495 | |
| 13496 | In addition, the do-again-test should also test whether the search finds a |
| 13497 | word. If it does not, the function should not call itself again. |
| 13498 | |
| 13499 | The next-step-expression changes a value so that when the recursive |
| 13500 | function is supposed to stop calling itself, it stops. More |
| 13501 | precisely, the next-step-expression changes a value so that at the |
| 13502 | right time, the do-again-test stops the recursive function from |
| 13503 | calling itself again. In this case, the next-step-expression can be |
| 13504 | the expression that moves point forward, word by word. |
| 13505 | |
| 13506 | The third part of a recursive function is the recursive call. |
| 13507 | |
| 13508 | Somewhere, also, we also need a part that does the `work' of the |
| 13509 | function, a part that does the counting. A vital part! |
| 13510 | |
| 13511 | @need 1250 |
| 13512 | But already, we have an outline of the recursive counting function: |
| 13513 | |
| 13514 | @smallexample |
| 13515 | @group |
| 13516 | (defun recursive-count-words (region-end) |
| 13517 | "@var{documentation}@dots{}" |
| 13518 | @var{do-again-test} |
| 13519 | @var{next-step-expression} |
| 13520 | @var{recursive call}) |
| 13521 | @end group |
| 13522 | @end smallexample |
| 13523 | |
| 13524 | Now we need to fill in the slots. Let's start with the simplest cases |
| 13525 | first: if point is at or beyond the end of the region, there cannot |
| 13526 | be any words in the region, so the function should return zero. |
| 13527 | Likewise, if the search fails, there are no words to count, so the |
| 13528 | function should return zero. |
| 13529 | |
| 13530 | On the other hand, if point is within the region and the search |
| 13531 | succeeds, the function should call itself again. |
| 13532 | |
| 13533 | @need 800 |
| 13534 | Thus, the do-again-test should look like this: |
| 13535 | |
| 13536 | @smallexample |
| 13537 | @group |
| 13538 | (and (< (point) region-end) |
| 13539 | (re-search-forward "\\w+\\W*" region-end t)) |
| 13540 | @end group |
| 13541 | @end smallexample |
| 13542 | |
| 13543 | Note that the search expression is part of the do-again-test---the |
| 13544 | function returns @code{t} if its search succeeds and @code{nil} if it |
| 13545 | fails. (@xref{Whitespace Bug, , The Whitespace Bug in |
| 13546 | @code{count-words-region}}, for an explanation of how |
| 13547 | @code{re-search-forward} works.) |
| 13548 | |
| 13549 | The do-again-test is the true-or-false test of an @code{if} clause. |
| 13550 | Clearly, if the do-again-test succeeds, the then-part of the @code{if} |
| 13551 | clause should call the function again; but if it fails, the else-part |
| 13552 | should return zero since either point is outside the region or the |
| 13553 | search failed because there were no words to find. |
| 13554 | |
| 13555 | But before considering the recursive call, we need to consider the |
| 13556 | next-step-expression. What is it? Interestingly, it is the search |
| 13557 | part of the do-again-test. |
| 13558 | |
| 13559 | In addition to returning @code{t} or @code{nil} for the |
| 13560 | do-again-test, @code{re-search-forward} moves point forward as a side |
| 13561 | effect of a successful search. This is the action that changes the |
| 13562 | value of point so that the recursive function stops calling itself |
| 13563 | when point completes its movement through the region. Consequently, |
| 13564 | the @code{re-search-forward} expression is the next-step-expression. |
| 13565 | |
| 13566 | @need 1200 |
| 13567 | In outline, then, the body of the @code{recursive-count-words} |
| 13568 | function looks like this: |
| 13569 | |
| 13570 | @smallexample |
| 13571 | @group |
| 13572 | (if @var{do-again-test-and-next-step-combined} |
| 13573 | ;; @r{then} |
| 13574 | @var{recursive-call-returning-count} |
| 13575 | ;; @r{else} |
| 13576 | @var{return-zero}) |
| 13577 | @end group |
| 13578 | @end smallexample |
| 13579 | |
| 13580 | How to incorporate the mechanism that counts? |
| 13581 | |
| 13582 | If you are not used to writing recursive functions, a question like |
| 13583 | this can be troublesome. But it can and should be approached |
| 13584 | systematically. |
| 13585 | |
| 13586 | We know that the counting mechanism should be associated in some way |
| 13587 | with the recursive call. Indeed, since the next-step-expression moves |
| 13588 | point forward by one word, and since a recursive call is made for |
| 13589 | each word, the counting mechanism must be an expression that adds one |
| 13590 | to the value returned by a call to @code{recursive-count-words}. |
| 13591 | |
| 13592 | Consider several cases: |
| 13593 | |
| 13594 | @itemize @bullet |
| 13595 | @item |
| 13596 | If there are two words in the region, the function should return |
| 13597 | a value resulting from adding one to the value returned when it counts |
| 13598 | the first word, plus the number returned when it counts the remaining |
| 13599 | words in the region, which in this case is one. |
| 13600 | |
| 13601 | @item |
| 13602 | If there is one word in the region, the function should return |
| 13603 | a value resulting from adding one to the value returned when it counts |
| 13604 | that word, plus the number returned when it counts the remaining |
| 13605 | words in the region, which in this case is zero. |
| 13606 | |
| 13607 | @item |
| 13608 | If there are no words in the region, the function should return zero. |
| 13609 | @end itemize |
| 13610 | |
| 13611 | From the sketch we can see that the else-part of the @code{if} returns |
| 13612 | zero for the case of no words. This means that the then-part of the |
| 13613 | @code{if} must return a value resulting from adding one to the value |
| 13614 | returned from a count of the remaining words. |
| 13615 | |
| 13616 | @need 1200 |
| 13617 | The expression will look like this, where @code{1+} is a function that |
| 13618 | adds one to its argument. |
| 13619 | |
| 13620 | @smallexample |
| 13621 | (1+ (recursive-count-words region-end)) |
| 13622 | @end smallexample |
| 13623 | |
| 13624 | @need 1200 |
| 13625 | The whole @code{recursive-count-words} function will then look like |
| 13626 | this: |
| 13627 | |
| 13628 | @smallexample |
| 13629 | @group |
| 13630 | (defun recursive-count-words (region-end) |
| 13631 | "@var{documentation}@dots{}" |
| 13632 | |
| 13633 | ;;; @r{1. do-again-test} |
| 13634 | (if (and (< (point) region-end) |
| 13635 | (re-search-forward "\\w+\\W*" region-end t)) |
| 13636 | @end group |
| 13637 | |
| 13638 | @group |
| 13639 | ;;; @r{2. then-part: the recursive call} |
| 13640 | (1+ (recursive-count-words region-end)) |
| 13641 | |
| 13642 | ;;; @r{3. else-part} |
| 13643 | 0)) |
| 13644 | @end group |
| 13645 | @end smallexample |
| 13646 | |
| 13647 | @need 1250 |
| 13648 | Let's examine how this works: |
| 13649 | |
| 13650 | If there are no words in the region, the else part of the @code{if} |
| 13651 | expression is evaluated and consequently the function returns zero. |
| 13652 | |
| 13653 | If there is one word in the region, the value of point is less than |
| 13654 | the value of @code{region-end} and the search succeeds. In this case, |
| 13655 | the true-or-false-test of the @code{if} expression tests true, and the |
| 13656 | then-part of the @code{if} expression is evaluated. The counting |
| 13657 | expression is evaluated. This expression returns a value (which will |
| 13658 | be the value returned by the whole function) that is the sum of one |
| 13659 | added to the value returned by a recursive call. |
| 13660 | |
| 13661 | Meanwhile, the next-step-expression has caused point to jump over the |
| 13662 | first (and in this case only) word in the region. This means that |
| 13663 | when @code{(recursive-count-words region-end)} is evaluated a second |
| 13664 | time, as a result of the recursive call, the value of point will be |
| 13665 | equal to or greater than the value of region end. So this time, |
| 13666 | @code{recursive-count-words} will return zero. The zero will be added |
| 13667 | to one, and the original evaluation of @code{recursive-count-words} |
| 13668 | will return one plus zero, which is one, which is the correct amount. |
| 13669 | |
| 13670 | Clearly, if there are two words in the region, the first call to |
| 13671 | @code{recursive-count-words} returns one added to the value returned |
| 13672 | by calling @code{recursive-count-words} on a region containing the |
| 13673 | remaining word---that is, it adds one to one, producing two, which is |
| 13674 | the correct amount. |
| 13675 | |
| 13676 | Similarly, if there are three words in the region, the first call to |
| 13677 | @code{recursive-count-words} returns one added to the value returned |
| 13678 | by calling @code{recursive-count-words} on a region containing the |
| 13679 | remaining two words---and so on and so on. |
| 13680 | |
| 13681 | @need 1250 |
| 13682 | @noindent |
| 13683 | With full documentation the two functions look like this: |
| 13684 | |
| 13685 | @need 1250 |
| 13686 | @noindent |
| 13687 | The recursive function: |
| 13688 | |
| 13689 | @findex recursive-count-words |
| 13690 | @smallexample |
| 13691 | @group |
| 13692 | (defun recursive-count-words (region-end) |
| 13693 | "Number of words between point and REGION-END." |
| 13694 | @end group |
| 13695 | |
| 13696 | @group |
| 13697 | ;;; @r{1. do-again-test} |
| 13698 | (if (and (< (point) region-end) |
| 13699 | (re-search-forward "\\w+\\W*" region-end t)) |
| 13700 | @end group |
| 13701 | |
| 13702 | @group |
| 13703 | ;;; @r{2. then-part: the recursive call} |
| 13704 | (1+ (recursive-count-words region-end)) |
| 13705 | |
| 13706 | ;;; @r{3. else-part} |
| 13707 | 0)) |
| 13708 | @end group |
| 13709 | @end smallexample |
| 13710 | |
| 13711 | @need 800 |
| 13712 | @noindent |
| 13713 | The wrapper: |
| 13714 | |
| 13715 | @smallexample |
| 13716 | @group |
| 13717 | ;;; @r{Recursive version} |
| 13718 | (defun count-words-region (beginning end) |
| 13719 | "Print number of words in the region. |
| 13720 | @end group |
| 13721 | |
| 13722 | @group |
| 13723 | Words are defined as at least one word-constituent |
| 13724 | character followed by at least one character that is |
| 13725 | not a word-constituent. The buffer's syntax table |
| 13726 | determines which characters these are." |
| 13727 | @end group |
| 13728 | @group |
| 13729 | (interactive "r") |
| 13730 | (message "Counting words in region ... ") |
| 13731 | (save-excursion |
| 13732 | (goto-char beginning) |
| 13733 | (let ((count (recursive-count-words end))) |
| 13734 | @end group |
| 13735 | @group |
| 13736 | (cond ((zerop count) |
| 13737 | (message |
| 13738 | "The region does NOT have any words.")) |
| 13739 | @end group |
| 13740 | @group |
| 13741 | ((= 1 count) |
| 13742 | (message "The region has 1 word.")) |
| 13743 | (t |
| 13744 | (message |
| 13745 | "The region has %d words." count)))))) |
| 13746 | @end group |
| 13747 | @end smallexample |
| 13748 | |
| 13749 | @node Counting Exercise, , recursive-count-words, Counting Words |
| 13750 | @section Exercise: Counting Punctuation |
| 13751 | |
| 13752 | Using a @code{while} loop, write a function to count the number of |
| 13753 | punctuation marks in a region---period, comma, semicolon, colon, |
| 13754 | exclamation mark, and question mark. Do the same using recursion. |
| 13755 | |
| 13756 | @node Words in a defun, Readying a Graph, Counting Words, Top |
| 13757 | @chapter Counting Words in a @code{defun} |
| 13758 | @cindex Counting words in a @code{defun} |
| 13759 | @cindex Word counting in a @code{defun} |
| 13760 | |
| 13761 | Our next project is to count the number of words in a function |
| 13762 | definition. Clearly, this can be done using some variant of |
| 13763 | @code{count-word-region}. @xref{Counting Words, , Counting Words: |
| 13764 | Repetition and Regexps}. If we are just going to count the words in |
| 13765 | one definition, it is easy enough to mark the definition with the |
| 13766 | @kbd{C-M-h} (@code{mark-defun}) command, and then call |
| 13767 | @code{count-word-region}. |
| 13768 | |
| 13769 | However, I am more ambitious: I want to count the words and symbols in |
| 13770 | every definition in the Emacs sources and then print a graph that |
| 13771 | shows how many functions there are of each length: how many contain 40 |
| 13772 | to 49 words or symbols, how many contain 50 to 59 words or symbols, |
| 13773 | and so on. I have often been curious how long a typical function is, |
| 13774 | and this will tell. |
| 13775 | |
| 13776 | @menu |
| 13777 | * Divide and Conquer:: |
| 13778 | * Words and Symbols:: What to count? |
| 13779 | * Syntax:: What constitutes a word or symbol? |
| 13780 | * count-words-in-defun:: Very like @code{count-words}. |
| 13781 | * Several defuns:: Counting several defuns in a file. |
| 13782 | * Find a File:: Do you want to look at a file? |
| 13783 | * lengths-list-file:: A list of the lengths of many definitions. |
| 13784 | * Several files:: Counting in definitions in different files. |
| 13785 | * Several files recursively:: Recursively counting in different files. |
| 13786 | * Prepare the data:: Prepare the data for display in a graph. |
| 13787 | @end menu |
| 13788 | |
| 13789 | @node Divide and Conquer, Words and Symbols, Words in a defun, Words in a defun |
| 13790 | @ifnottex |
| 13791 | @unnumberedsec Divide and Conquer |
| 13792 | @end ifnottex |
| 13793 | |
| 13794 | Described in one phrase, the histogram project is daunting; but |
| 13795 | divided into numerous small steps, each of which we can take one at a |
| 13796 | time, the project becomes less fearsome. Let us consider what the |
| 13797 | steps must be: |
| 13798 | |
| 13799 | @itemize @bullet |
| 13800 | @item |
| 13801 | First, write a function to count the words in one definition. This |
| 13802 | includes the problem of handling symbols as well as words. |
| 13803 | |
| 13804 | @item |
| 13805 | Second, write a function to list the numbers of words in each function |
| 13806 | in a file. This function can use the @code{count-words-in-defun} |
| 13807 | function. |
| 13808 | |
| 13809 | @item |
| 13810 | Third, write a function to list the numbers of words in each function |
| 13811 | in each of several files. This entails automatically finding the |
| 13812 | various files, switching to them, and counting the words in the |
| 13813 | definitions within them. |
| 13814 | |
| 13815 | @item |
| 13816 | Fourth, write a function to convert the list of numbers that we |
| 13817 | created in step three to a form that will be suitable for printing as |
| 13818 | a graph. |
| 13819 | |
| 13820 | @item |
| 13821 | Fifth, write a function to print the results as a graph. |
| 13822 | @end itemize |
| 13823 | |
| 13824 | This is quite a project! But if we take each step slowly, it will not |
| 13825 | be difficult. |
| 13826 | |
| 13827 | @node Words and Symbols, Syntax, Divide and Conquer, Words in a defun |
| 13828 | @section What to Count? |
| 13829 | @cindex Words and symbols in defun |
| 13830 | |
| 13831 | When we first start thinking about how to count the words in a |
| 13832 | function definition, the first question is (or ought to be) what are |
| 13833 | we going to count? When we speak of `words' with respect to a Lisp |
| 13834 | function definition, we are actually speaking, in large part, of |
| 13835 | `symbols'. For example, the following @code{multiply-by-seven} |
| 13836 | function contains the five symbols @code{defun}, |
| 13837 | @code{multiply-by-seven}, @code{number}, @code{*}, and @code{7}. In |
| 13838 | addition, in the documentation string, it contains the four words |
| 13839 | @samp{Multiply}, @samp{NUMBER}, @samp{by}, and @samp{seven}. The |
| 13840 | symbol @samp{number} is repeated, so the definition contains a total |
| 13841 | of ten words and symbols. |
| 13842 | |
| 13843 | @smallexample |
| 13844 | @group |
| 13845 | (defun multiply-by-seven (number) |
| 13846 | "Multiply NUMBER by seven." |
| 13847 | (* 7 number)) |
| 13848 | @end group |
| 13849 | @end smallexample |
| 13850 | |
| 13851 | @noindent |
| 13852 | However, if we mark the @code{multiply-by-seven} definition with |
| 13853 | @kbd{C-M-h} (@code{mark-defun}), and then call |
| 13854 | @code{count-words-region} on it, we will find that |
| 13855 | @code{count-words-region} claims the definition has eleven words, not |
| 13856 | ten! Something is wrong! |
| 13857 | |
| 13858 | The problem is twofold: @code{count-words-region} does not count the |
| 13859 | @samp{*} as a word, and it counts the single symbol, |
| 13860 | @code{multiply-by-seven}, as containing three words. The hyphens are |
| 13861 | treated as if they were interword spaces rather than intraword |
| 13862 | connectors: @samp{multiply-by-seven} is counted as if it were written |
| 13863 | @samp{multiply by seven}. |
| 13864 | |
| 13865 | The cause of this confusion is the regular expression search within |
| 13866 | the @code{count-words-region} definition that moves point forward word |
| 13867 | by word. In the canonical version of @code{count-words-region}, the |
| 13868 | regexp is: |
| 13869 | |
| 13870 | @smallexample |
| 13871 | "\\w+\\W*" |
| 13872 | @end smallexample |
| 13873 | |
| 13874 | @noindent |
| 13875 | This regular expression is a pattern defining one or more word |
| 13876 | constituent characters possibly followed by one or more characters |
| 13877 | that are not word constituents. What is meant by `word constituent |
| 13878 | characters' brings us to the issue of syntax, which is worth a section |
| 13879 | of its own. |
| 13880 | |
| 13881 | @node Syntax, count-words-in-defun, Words and Symbols, Words in a defun |
| 13882 | @section What Constitutes a Word or Symbol? |
| 13883 | @cindex Syntax categories and tables |
| 13884 | |
| 13885 | Emacs treats different characters as belonging to different |
| 13886 | @dfn{syntax categories}. For example, the regular expression, |
| 13887 | @samp{\\w+}, is a pattern specifying one or more @emph{word |
| 13888 | constituent} characters. Word constituent characters are members of |
| 13889 | one syntax category. Other syntax categories include the class of |
| 13890 | punctuation characters, such as the period and the comma, and the |
| 13891 | class of whitespace characters, such as the blank space and the tab |
| 13892 | character. (For more information, see @ref{Syntax, Syntax, The Syntax |
| 13893 | Table, emacs, The GNU Emacs Manual}, and @ref{Syntax Tables, , Syntax |
| 13894 | Tables, elisp, The GNU Emacs Lisp Reference Manual}.) |
| 13895 | |
| 13896 | Syntax tables specify which characters belong to which categories. |
| 13897 | Usually, a hyphen is not specified as a `word constituent character'. |
| 13898 | Instead, it is specified as being in the `class of characters that are |
| 13899 | part of symbol names but not words.' This means that the |
| 13900 | @code{count-words-region} function treats it in the same way it treats |
| 13901 | an interword white space, which is why @code{count-words-region} |
| 13902 | counts @samp{multiply-by-seven} as three words. |
| 13903 | |
| 13904 | There are two ways to cause Emacs to count @samp{multiply-by-seven} as |
| 13905 | one symbol: modify the syntax table or modify the regular expression. |
| 13906 | |
| 13907 | We could redefine a hyphen as a word constituent character by |
| 13908 | modifying the syntax table that Emacs keeps for each mode. This |
| 13909 | action would serve our purpose, except that a hyphen is merely the |
| 13910 | most common character within symbols that is not typically a word |
| 13911 | constituent character; there are others, too. |
| 13912 | |
| 13913 | Alternatively, we can redefine the regular expression used in the |
| 13914 | @code{count-words} definition so as to include symbols. This |
| 13915 | procedure has the merit of clarity, but the task is a little tricky. |
| 13916 | |
| 13917 | @need 1200 |
| 13918 | The first part is simple enough: the pattern must match ``at least one |
| 13919 | character that is a word or symbol constituent''. Thus: |
| 13920 | |
| 13921 | @smallexample |
| 13922 | "\\(\\w\\|\\s_\\)+" |
| 13923 | @end smallexample |
| 13924 | |
| 13925 | @noindent |
| 13926 | The @samp{\\(} is the first part of the grouping construct that |
| 13927 | includes the @samp{\\w} and the @samp{\\s_} as alternatives, separated |
| 13928 | by the @samp{\\|}. The @samp{\\w} matches any word-constituent |
| 13929 | character and the @samp{\\s_} matches any character that is part of a |
| 13930 | symbol name but not a word-constituent character. The @samp{+} |
| 13931 | following the group indicates that the word or symbol constituent |
| 13932 | characters must be matched at least once. |
| 13933 | |
| 13934 | However, the second part of the regexp is more difficult to design. |
| 13935 | What we want is to follow the first part with ``optionally one or more |
| 13936 | characters that are not constituents of a word or symbol''. At first, |
| 13937 | I thought I could define this with the following: |
| 13938 | |
| 13939 | @smallexample |
| 13940 | "\\(\\W\\|\\S_\\)*" |
| 13941 | @end smallexample |
| 13942 | |
| 13943 | @noindent |
| 13944 | The upper case @samp{W} and @samp{S} match characters that are |
| 13945 | @emph{not} word or symbol constituents. Unfortunately, this |
| 13946 | expression matches any character that is either not a word constituent |
| 13947 | or not a symbol constituent. This matches any character! |
| 13948 | |
| 13949 | I then noticed that every word or symbol in my test region was |
| 13950 | followed by white space (blank space, tab, or newline). So I tried |
| 13951 | placing a pattern to match one or more blank spaces after the pattern |
| 13952 | for one or more word or symbol constituents. This failed, too. Words |
| 13953 | and symbols are often separated by whitespace, but in actual code |
| 13954 | parentheses may follow symbols and punctuation may follow words. So |
| 13955 | finally, I designed a pattern in which the word or symbol constituents |
| 13956 | are followed optionally by characters that are not white space and |
| 13957 | then followed optionally by white space. |
| 13958 | |
| 13959 | @need 800 |
| 13960 | Here is the full regular expression: |
| 13961 | |
| 13962 | @smallexample |
| 13963 | "\\(\\w\\|\\s_\\)+[^ \t\n]*[ \t\n]*" |
| 13964 | @end smallexample |
| 13965 | |
| 13966 | @node count-words-in-defun, Several defuns, Syntax, Words in a defun |
| 13967 | @section The @code{count-words-in-defun} Function |
| 13968 | @cindex Counting words in a @code{defun} |
| 13969 | |
| 13970 | We have seen that there are several ways to write a |
| 13971 | @code{count-word-region} function. To write a |
| 13972 | @code{count-words-in-defun}, we need merely adapt one of these |
| 13973 | versions. |
| 13974 | |
| 13975 | The version that uses a @code{while} loop is easy to understand, so I |
| 13976 | am going to adapt that. Because @code{count-words-in-defun} will be |
| 13977 | part of a more complex program, it need not be interactive and it need |
| 13978 | not display a message but just return the count. These considerations |
| 13979 | simplify the definition a little. |
| 13980 | |
| 13981 | On the other hand, @code{count-words-in-defun} will be used within a |
| 13982 | buffer that contains function definitions. Consequently, it is |
| 13983 | reasonable to ask that the function determine whether it is called |
| 13984 | when point is within a function definition, and if it is, to return |
| 13985 | the count for that definition. This adds complexity to the |
| 13986 | definition, but saves us from needing to pass arguments to the |
| 13987 | function. |
| 13988 | |
| 13989 | @need 1250 |
| 13990 | These considerations lead us to prepare the following template: |
| 13991 | |
| 13992 | @smallexample |
| 13993 | @group |
| 13994 | (defun count-words-in-defun () |
| 13995 | "@var{documentation}@dots{}" |
| 13996 | (@var{set up}@dots{} |
| 13997 | (@var{while loop}@dots{}) |
| 13998 | @var{return count}) |
| 13999 | @end group |
| 14000 | @end smallexample |
| 14001 | |
| 14002 | @noindent |
| 14003 | As usual, our job is to fill in the slots. |
| 14004 | |
| 14005 | First, the set up. |
| 14006 | |
| 14007 | We are presuming that this function will be called within a buffer |
| 14008 | containing function definitions. Point will either be within a |
| 14009 | function definition or not. For @code{count-words-in-defun} to work, |
| 14010 | point must move to the beginning of the definition, a counter must |
| 14011 | start at zero, and the counting loop must stop when point reaches the |
| 14012 | end of the definition. |
| 14013 | |
| 14014 | The @code{beginning-of-defun} function searches backwards for an |
| 14015 | opening delimiter such as a @samp{(} at the beginning of a line, and |
| 14016 | moves point to that position, or else to the limit of the search. In |
| 14017 | practice, this means that @code{beginning-of-defun} moves point to the |
| 14018 | beginning of an enclosing or preceding function definition, or else to |
| 14019 | the beginning of the buffer. We can use @code{beginning-of-defun} to |
| 14020 | place point where we wish to start. |
| 14021 | |
| 14022 | The @code{while} loop requires a counter to keep track of the words or |
| 14023 | symbols being counted. A @code{let} expression can be used to create |
| 14024 | a local variable for this purpose, and bind it to an initial value of zero. |
| 14025 | |
| 14026 | The @code{end-of-defun} function works like @code{beginning-of-defun} |
| 14027 | except that it moves point to the end of the definition. |
| 14028 | @code{end-of-defun} can be used as part of an expression that |
| 14029 | determines the position of the end of the definition. |
| 14030 | |
| 14031 | The set up for @code{count-words-in-defun} takes shape rapidly: first |
| 14032 | we move point to the beginning of the definition, then we create a |
| 14033 | local variable to hold the count, and finally, we record the position |
| 14034 | of the end of the definition so the @code{while} loop will know when to stop |
| 14035 | looping. |
| 14036 | |
| 14037 | @need 1250 |
| 14038 | The code looks like this: |
| 14039 | |
| 14040 | @smallexample |
| 14041 | @group |
| 14042 | (beginning-of-defun) |
| 14043 | (let ((count 0) |
| 14044 | (end (save-excursion (end-of-defun) (point)))) |
| 14045 | @end group |
| 14046 | @end smallexample |
| 14047 | |
| 14048 | @noindent |
| 14049 | The code is simple. The only slight complication is likely to concern |
| 14050 | @code{end}: it is bound to the position of the end of the definition |
| 14051 | by a @code{save-excursion} expression that returns the value of point |
| 14052 | after @code{end-of-defun} temporarily moves it to the end of the |
| 14053 | definition. |
| 14054 | |
| 14055 | The second part of the @code{count-words-in-defun}, after the set up, |
| 14056 | is the @code{while} loop. |
| 14057 | |
| 14058 | The loop must contain an expression that jumps point forward word by |
| 14059 | word and symbol by symbol, and another expression that counts the |
| 14060 | jumps. The true-or-false-test for the @code{while} loop should test |
| 14061 | true so long as point should jump forward, and false when point is at |
| 14062 | the end of the definition. We have already redefined the regular |
| 14063 | expression for this (@pxref{Syntax}), so the loop is straightforward: |
| 14064 | |
| 14065 | @smallexample |
| 14066 | @group |
| 14067 | (while (and (< (point) end) |
| 14068 | (re-search-forward |
| 14069 | "\\(\\w\\|\\s_\\)+[^ \t\n]*[ \t\n]*" end t) |
| 14070 | (setq count (1+ count))) |
| 14071 | @end group |
| 14072 | @end smallexample |
| 14073 | |
| 14074 | The third part of the function definition returns the count of words |
| 14075 | and symbols. This part is the last expression within the body of the |
| 14076 | @code{let} expression, and can be, very simply, the local variable |
| 14077 | @code{count}, which when evaluated returns the count. |
| 14078 | |
| 14079 | @need 1250 |
| 14080 | Put together, the @code{count-words-in-defun} definition looks like this: |
| 14081 | |
| 14082 | @findex count-words-in-defun |
| 14083 | @smallexample |
| 14084 | @group |
| 14085 | (defun count-words-in-defun () |
| 14086 | "Return the number of words and symbols in a defun." |
| 14087 | (beginning-of-defun) |
| 14088 | (let ((count 0) |
| 14089 | (end (save-excursion (end-of-defun) (point)))) |
| 14090 | @end group |
| 14091 | @group |
| 14092 | (while |
| 14093 | (and (< (point) end) |
| 14094 | (re-search-forward |
| 14095 | "\\(\\w\\|\\s_\\)+[^ \t\n]*[ \t\n]*" |
| 14096 | end t)) |
| 14097 | (setq count (1+ count))) |
| 14098 | count)) |
| 14099 | @end group |
| 14100 | @end smallexample |
| 14101 | |
| 14102 | How to test this? The function is not interactive, but it is easy to |
| 14103 | put a wrapper around the function to make it interactive; we can use |
| 14104 | almost the same code as for the recursive version of |
| 14105 | @code{count-words-region}: |
| 14106 | |
| 14107 | @smallexample |
| 14108 | @group |
| 14109 | ;;; @r{Interactive version.} |
| 14110 | (defun count-words-defun () |
| 14111 | "Number of words and symbols in a function definition." |
| 14112 | (interactive) |
| 14113 | (message |
| 14114 | "Counting words and symbols in function definition ... ") |
| 14115 | @end group |
| 14116 | @group |
| 14117 | (let ((count (count-words-in-defun))) |
| 14118 | (cond |
| 14119 | ((zerop count) |
| 14120 | (message |
| 14121 | "The definition does NOT have any words or symbols.")) |
| 14122 | @end group |
| 14123 | @group |
| 14124 | ((= 1 count) |
| 14125 | (message |
| 14126 | "The definition has 1 word or symbol.")) |
| 14127 | (t |
| 14128 | (message |
| 14129 | "The definition has %d words or symbols." count))))) |
| 14130 | @end group |
| 14131 | @end smallexample |
| 14132 | |
| 14133 | @need 800 |
| 14134 | @noindent |
| 14135 | Let's re-use @kbd{C-c =} as a convenient keybinding: |
| 14136 | |
| 14137 | @smallexample |
| 14138 | (global-set-key "\C-c=" 'count-words-defun) |
| 14139 | @end smallexample |
| 14140 | |
| 14141 | Now we can try out @code{count-words-defun}: install both |
| 14142 | @code{count-words-in-defun} and @code{count-words-defun}, and set the |
| 14143 | keybinding, and then place the cursor within the following definition: |
| 14144 | |
| 14145 | @smallexample |
| 14146 | @group |
| 14147 | (defun multiply-by-seven (number) |
| 14148 | "Multiply NUMBER by seven." |
| 14149 | (* 7 number)) |
| 14150 | @result{} 10 |
| 14151 | @end group |
| 14152 | @end smallexample |
| 14153 | |
| 14154 | @noindent |
| 14155 | Success! The definition has 10 words and symbols. |
| 14156 | |
| 14157 | The next problem is to count the numbers of words and symbols in |
| 14158 | several definitions within a single file. |
| 14159 | |
| 14160 | @node Several defuns, Find a File, count-words-in-defun, Words in a defun |
| 14161 | @section Count Several @code{defuns} Within a File |
| 14162 | |
| 14163 | A file such as @file{simple.el} may have 80 or more function |
| 14164 | definitions within it. Our long term goal is to collect statistics on |
| 14165 | many files, but as a first step, our immediate goal is to collect |
| 14166 | statistics on one file. |
| 14167 | |
| 14168 | The information will be a series of numbers, each number being the |
| 14169 | length of a function definition. We can store the numbers in a list. |
| 14170 | |
| 14171 | We know that we will want to incorporate the information regarding one |
| 14172 | file with information about many other files; this means that the |
| 14173 | function for counting definition lengths within one file need only |
| 14174 | return the list of lengths. It need not and should not display any |
| 14175 | messages. |
| 14176 | |
| 14177 | The word count commands contain one expression to jump point forward |
| 14178 | word by word and another expression to count the jumps. The function |
| 14179 | to return the lengths of definitions can be designed to work the same |
| 14180 | way, with one expression to jump point forward definition by |
| 14181 | definition and another expression to construct the lengths' list. |
| 14182 | |
| 14183 | This statement of the problem makes it elementary to write the |
| 14184 | function definition. Clearly, we will start the count at the |
| 14185 | beginning of the file, so the first command will be @code{(goto-char |
| 14186 | (point-min))}. Next, we start the @code{while} loop; and the |
| 14187 | true-or-false test of the loop can be a regular expression search for |
| 14188 | the next function definition---so long as the search succeeds, point |
| 14189 | is moved forward and then the body of the loop is evaluated. The body |
| 14190 | needs an expression that constructs the lengths' list. @code{cons}, |
| 14191 | the list construction command, can be used to create the list. That |
| 14192 | is almost all there is to it. |
| 14193 | |
| 14194 | @need 800 |
| 14195 | Here is what this fragment of code looks like: |
| 14196 | |
| 14197 | @smallexample |
| 14198 | @group |
| 14199 | (goto-char (point-min)) |
| 14200 | (while (re-search-forward "^(defun" nil t) |
| 14201 | (setq lengths-list |
| 14202 | (cons (count-words-in-defun) lengths-list))) |
| 14203 | @end group |
| 14204 | @end smallexample |
| 14205 | |
| 14206 | What we have left out is the mechanism for finding the file that |
| 14207 | contains the function definitions. |
| 14208 | |
| 14209 | In previous examples, we either used this, the Info file, or we |
| 14210 | switched back and forth to some other buffer, such as the |
| 14211 | @file{*scratch*} buffer. |
| 14212 | |
| 14213 | Finding a file is a new process that we have not yet discussed. |
| 14214 | |
| 14215 | @node Find a File, lengths-list-file, Several defuns, Words in a defun |
| 14216 | @comment node-name, next, previous, up |
| 14217 | @section Find a File |
| 14218 | @cindex Find a File |
| 14219 | |
| 14220 | To find a file in Emacs, you use the @kbd{C-x C-f} (@code{find-file}) |
| 14221 | command. This command is almost, but not quite right for the lengths |
| 14222 | problem. |
| 14223 | |
| 14224 | @need 1200 |
| 14225 | Let's look at the source for @code{find-file} (you can use the |
| 14226 | @code{find-tag} command or @kbd{C-h f} (@code{describe-function}) to |
| 14227 | find the source of a function): |
| 14228 | |
| 14229 | @smallexample |
| 14230 | @group |
| 14231 | (defun find-file (filename) |
| 14232 | "Edit file FILENAME. |
| 14233 | Switch to a buffer visiting file FILENAME, |
| 14234 | creating one if none already exists." |
| 14235 | (interactive "FFind file: ") |
| 14236 | (switch-to-buffer (find-file-noselect filename))) |
| 14237 | @end group |
| 14238 | @end smallexample |
| 14239 | |
| 14240 | The definition possesses short but complete documentation and an |
| 14241 | interactive specification that prompts you for a file name when you |
| 14242 | use the command interactively. The body of the definition contains |
| 14243 | two functions, @code{find-file-noselect} and @code{switch-to-buffer}. |
| 14244 | |
| 14245 | According to its documentation as shown by @kbd{C-h f} (the |
| 14246 | @code{describe-function} command), the @code{find-file-noselect} |
| 14247 | function reads the named file into a buffer and returns the buffer. |
| 14248 | However, the buffer is not selected. Emacs does not switch its |
| 14249 | attention (or yours if you are using @code{find-file-noselect}) to the |
| 14250 | named buffer. That is what @code{switch-to-buffer} does: it switches |
| 14251 | the buffer to which Emacs attention is directed; and it switches the |
| 14252 | buffer displayed in the window to the new buffer. We have discussed |
| 14253 | buffer switching elsewhere. (@xref{Switching Buffers}.) |
| 14254 | |
| 14255 | In this histogram project, we do not need to display each file on the |
| 14256 | screen as the program determines the length of each definition within |
| 14257 | it. Instead of employing @code{switch-to-buffer}, we can work with |
| 14258 | @code{set-buffer}, which redirects the attention of the computer |
| 14259 | program to a different buffer but does not redisplay it on the screen. |
| 14260 | So instead of calling on @code{find-file} to do the job, we must write |
| 14261 | our own expression. |
| 14262 | |
| 14263 | The task is easy: use @code{find-file-noselect} and @code{set-buffer}. |
| 14264 | |
| 14265 | @node lengths-list-file, Several files, Find a File, Words in a defun |
| 14266 | @section @code{lengths-list-file} in Detail |
| 14267 | |
| 14268 | The core of the @code{lengths-list-file} function is a @code{while} |
| 14269 | loop containing a function to move point forward `defun by defun' and |
| 14270 | a function to count the number of words and symbols in each defun. |
| 14271 | This core must be surrounded by functions that do various other tasks, |
| 14272 | including finding the file, and ensuring that point starts out at the |
| 14273 | beginning of the file. The function definition looks like this: |
| 14274 | @findex lengths-list-file |
| 14275 | |
| 14276 | @smallexample |
| 14277 | @group |
| 14278 | (defun lengths-list-file (filename) |
| 14279 | "Return list of definitions' lengths within FILE. |
| 14280 | The returned list is a list of numbers. |
| 14281 | Each number is the number of words or |
| 14282 | symbols in one function definition." |
| 14283 | @end group |
| 14284 | @group |
| 14285 | (message "Working on `%s' ... " filename) |
| 14286 | (save-excursion |
| 14287 | (let ((buffer (find-file-noselect filename)) |
| 14288 | (lengths-list)) |
| 14289 | (set-buffer buffer) |
| 14290 | (setq buffer-read-only t) |
| 14291 | (widen) |
| 14292 | (goto-char (point-min)) |
| 14293 | (while (re-search-forward "^(defun" nil t) |
| 14294 | (setq lengths-list |
| 14295 | (cons (count-words-in-defun) lengths-list))) |
| 14296 | (kill-buffer buffer) |
| 14297 | lengths-list))) |
| 14298 | @end group |
| 14299 | @end smallexample |
| 14300 | |
| 14301 | @noindent |
| 14302 | The function is passed one argument, the name of the file on which it |
| 14303 | will work. It has four lines of documentation, but no interactive |
| 14304 | specification. Since people worry that a computer is broken if they |
| 14305 | don't see anything going on, the first line of the body is a |
| 14306 | message. |
| 14307 | |
| 14308 | The next line contains a @code{save-excursion} that returns Emacs' |
| 14309 | attention to the current buffer when the function completes. This is |
| 14310 | useful in case you embed this function in another function that |
| 14311 | presumes point is restored to the original buffer. |
| 14312 | |
| 14313 | In the varlist of the @code{let} expression, Emacs finds the file and |
| 14314 | binds the local variable @code{buffer} to the buffer containing the |
| 14315 | file. At the same time, Emacs creates @code{lengths-list} as a local |
| 14316 | variable. |
| 14317 | |
| 14318 | Next, Emacs switches its attention to the buffer. |
| 14319 | |
| 14320 | In the following line, Emacs makes the buffer read-only. Ideally, |
| 14321 | this line is not necessary. None of the functions for counting words |
| 14322 | and symbols in a function definition should change the buffer. |
| 14323 | Besides, the buffer is not going to be saved, even if it were changed. |
| 14324 | This line is entirely the consequence of great, perhaps excessive, |
| 14325 | caution. The reason for the caution is that this function and those |
| 14326 | it calls work on the sources for Emacs and it is very inconvenient if |
| 14327 | they are inadvertently modified. It goes without saying that I did |
| 14328 | not realize a need for this line until an experiment went awry and |
| 14329 | started to modify my Emacs source files @dots{} |
| 14330 | |
| 14331 | Next comes a call to widen the buffer if it is narrowed. This |
| 14332 | function is usually not needed---Emacs creates a fresh buffer if none |
| 14333 | already exists; but if a buffer visiting the file already exists Emacs |
| 14334 | returns that one. In this case, the buffer may be narrowed and must |
| 14335 | be widened. If we wanted to be fully `user-friendly', we would |
| 14336 | arrange to save the restriction and the location of point, but we |
| 14337 | won't. |
| 14338 | |
| 14339 | The @code{(goto-char (point-min))} expression moves point to the |
| 14340 | beginning of the buffer. |
| 14341 | |
| 14342 | Then comes a @code{while} loop in which the `work' of the function is |
| 14343 | carried out. In the loop, Emacs determines the length of each |
| 14344 | definition and constructs a lengths' list containing the information. |
| 14345 | |
| 14346 | Emacs kills the buffer after working through it. This is to save |
| 14347 | space inside of Emacs. My version of Emacs 19 contained over 300 |
| 14348 | source files of interest; Emacs 21 contains over 800 source files. |
| 14349 | Another function will apply @code{lengths-list-file} to each of the |
| 14350 | files. |
| 14351 | |
| 14352 | Finally, the last expression within the @code{let} expression is the |
| 14353 | @code{lengths-list} variable; its value is returned as the value of |
| 14354 | the whole function. |
| 14355 | |
| 14356 | You can try this function by installing it in the usual fashion. Then |
| 14357 | place your cursor after the following expression and type @kbd{C-x |
| 14358 | C-e} (@code{eval-last-sexp}). |
| 14359 | |
| 14360 | @c !!! 21.0.100 lisp sources location here |
| 14361 | @smallexample |
| 14362 | (lengths-list-file |
| 14363 | "/usr/local/share/emacs/21.0.100/lisp/emacs-lisp/debug.el") |
| 14364 | @end smallexample |
| 14365 | |
| 14366 | @c was: (lengths-list-file "../lisp/debug.el") |
| 14367 | @c !!! as of 21, Info file is in |
| 14368 | @c /usr/share/info/emacs-lisp-intro.info.gz |
| 14369 | @c but debug.el is in /usr/local/share/emacs/21.0.100/lisp/emacs-lisp/debug.el |
| 14370 | |
| 14371 | @noindent |
| 14372 | (You may need to change the pathname of the file; the one here worked |
| 14373 | with GNU Emacs version 21.0.100. To change the expression, copy it to |
| 14374 | the @file{*scratch*} buffer and edit it. |
| 14375 | |
| 14376 | @need 1200 |
| 14377 | @noindent |
| 14378 | (Also, to see the full length of the list, rather than a truncated |
| 14379 | version, you may have to evaluate the following: |
| 14380 | |
| 14381 | @smallexample |
| 14382 | (custom-set-variables '(eval-expression-print-length nil)) |
| 14383 | @end smallexample |
| 14384 | |
| 14385 | @noindent |
| 14386 | (@xref{defcustom, , Specifying Variables using @code{defcustom}}.) |
| 14387 | Then evaluate the @code{lengths-list-file} expression.) |
| 14388 | |
| 14389 | @need 1200 |
| 14390 | The lengths' list for @file{debug.el} takes less than a second to |
| 14391 | produce and looks like this: |
| 14392 | |
| 14393 | @smallexample |
| 14394 | (77 95 85 87 131 89 50 25 44 44 68 35 64 45 17 34 167 457) |
| 14395 | @end smallexample |
| 14396 | |
| 14397 | @need 1500 |
| 14398 | (Using my old machine, the version 19 lengths' list for @file{debug.el} |
| 14399 | took seven seconds to produce and looked like this: |
| 14400 | |
| 14401 | @smallexample |
| 14402 | (75 41 80 62 20 45 44 68 45 12 34 235) |
| 14403 | @end smallexample |
| 14404 | |
| 14405 | (The newer version of @file{debug.el} contains more defuns than the |
| 14406 | earlier one; and my new machine is much faster than the old one.) |
| 14407 | |
| 14408 | Note that the length of the last definition in the file is first in |
| 14409 | the list. |
| 14410 | |
| 14411 | @node Several files, Several files recursively, lengths-list-file, Words in a defun |
| 14412 | @section Count Words in @code{defuns} in Different Files |
| 14413 | |
| 14414 | In the previous section, we created a function that returns a list of |
| 14415 | the lengths of each definition in a file. Now, we want to define a |
| 14416 | function to return a master list of the lengths of the definitions in |
| 14417 | a list of files. |
| 14418 | |
| 14419 | Working on each of a list of files is a repetitious act, so we can use |
| 14420 | either a @code{while} loop or recursion. |
| 14421 | |
| 14422 | @menu |
| 14423 | * lengths-list-many-files:: Return a list of the lengths of defuns. |
| 14424 | * append:: Attach one list to another. |
| 14425 | @end menu |
| 14426 | |
| 14427 | @node lengths-list-many-files, append, Several files, Several files |
| 14428 | @ifnottex |
| 14429 | @unnumberedsubsec Determine the lengths of @code{defuns} |
| 14430 | @end ifnottex |
| 14431 | |
| 14432 | The design using a @code{while} loop is routine. The argument passed |
| 14433 | the function is a list of files. As we saw earlier (@pxref{Loop |
| 14434 | Example}), you can write a @code{while} loop so that the body of the |
| 14435 | loop is evaluated if such a list contains elements, but to exit the |
| 14436 | loop if the list is empty. For this design to work, the body of the |
| 14437 | loop must contain an expression that shortens the list each time the |
| 14438 | body is evaluated, so that eventually the list is empty. The usual |
| 14439 | technique is to set the value of the list to the value of the @sc{cdr} |
| 14440 | of the list each time the body is evaluated. |
| 14441 | |
| 14442 | @need 800 |
| 14443 | The template looks like this: |
| 14444 | |
| 14445 | @smallexample |
| 14446 | @group |
| 14447 | (while @var{test-whether-list-is-empty} |
| 14448 | @var{body}@dots{} |
| 14449 | @var{set-list-to-cdr-of-list}) |
| 14450 | @end group |
| 14451 | @end smallexample |
| 14452 | |
| 14453 | Also, we remember that a @code{while} loop returns @code{nil} (the |
| 14454 | result of evaluating the true-or-false-test), not the result of any |
| 14455 | evaluation within its body. (The evaluations within the body of the |
| 14456 | loop are done for their side effects.) However, the expression that |
| 14457 | sets the lengths' list is part of the body---and that is the value |
| 14458 | that we want returned by the function as a whole. To do this, we |
| 14459 | enclose the @code{while} loop within a @code{let} expression, and |
| 14460 | arrange that the last element of the @code{let} expression contains |
| 14461 | the value of the lengths' list. (@xref{Incrementing Example, , Loop |
| 14462 | Example with an Incrementing Counter}.) |
| 14463 | |
| 14464 | @findex lengths-list-many-files |
| 14465 | @need 1250 |
| 14466 | These considerations lead us directly to the function itself: |
| 14467 | |
| 14468 | @smallexample |
| 14469 | @group |
| 14470 | ;;; @r{Use @code{while} loop.} |
| 14471 | (defun lengths-list-many-files (list-of-files) |
| 14472 | "Return list of lengths of defuns in LIST-OF-FILES." |
| 14473 | @end group |
| 14474 | @group |
| 14475 | (let (lengths-list) |
| 14476 | |
| 14477 | ;;; @r{true-or-false-test} |
| 14478 | (while list-of-files |
| 14479 | (setq lengths-list |
| 14480 | (append |
| 14481 | lengths-list |
| 14482 | |
| 14483 | ;;; @r{Generate a lengths' list.} |
| 14484 | (lengths-list-file |
| 14485 | (expand-file-name (car list-of-files))))) |
| 14486 | @end group |
| 14487 | |
| 14488 | @group |
| 14489 | ;;; @r{Make files' list shorter.} |
| 14490 | (setq list-of-files (cdr list-of-files))) |
| 14491 | |
| 14492 | ;;; @r{Return final value of lengths' list.} |
| 14493 | lengths-list)) |
| 14494 | @end group |
| 14495 | @end smallexample |
| 14496 | |
| 14497 | @code{expand-file-name} is a built-in function that converts a file |
| 14498 | name to the absolute, long, path name form of the directory in which |
| 14499 | the function is called. |
| 14500 | |
| 14501 | @c !!! 21.0.100 lisp sources location here |
| 14502 | @need 1500 |
| 14503 | Thus, if @code{expand-file-name} is called on @code{debug.el} when |
| 14504 | Emacs is visiting the |
| 14505 | @file{/usr/local/share/emacs/21.0.100/lisp/emacs-lisp/} directory, |
| 14506 | |
| 14507 | @smallexample |
| 14508 | debug.el |
| 14509 | @end smallexample |
| 14510 | |
| 14511 | @need 800 |
| 14512 | @noindent |
| 14513 | becomes |
| 14514 | |
| 14515 | @c !!! 21.0.100 lisp sources location here |
| 14516 | @smallexample |
| 14517 | /usr/local/share/emacs/21.0.100/lisp/emacs-lisp/debug.el |
| 14518 | @end smallexample |
| 14519 | |
| 14520 | The only other new element of this function definition is the as yet |
| 14521 | unstudied function @code{append}, which merits a short section for |
| 14522 | itself. |
| 14523 | |
| 14524 | @node append, , lengths-list-many-files, Several files |
| 14525 | @subsection The @code{append} Function |
| 14526 | |
| 14527 | @need 800 |
| 14528 | The @code{append} function attaches one list to another. Thus, |
| 14529 | |
| 14530 | @smallexample |
| 14531 | (append '(1 2 3 4) '(5 6 7 8)) |
| 14532 | @end smallexample |
| 14533 | |
| 14534 | @need 800 |
| 14535 | @noindent |
| 14536 | produces the list |
| 14537 | |
| 14538 | @smallexample |
| 14539 | (1 2 3 4 5 6 7 8) |
| 14540 | @end smallexample |
| 14541 | |
| 14542 | This is exactly how we want to attach two lengths' lists produced by |
| 14543 | @code{lengths-list-file} to each other. The results contrast with |
| 14544 | @code{cons}, |
| 14545 | |
| 14546 | @smallexample |
| 14547 | (cons '(1 2 3 4) '(5 6 7 8)) |
| 14548 | @end smallexample |
| 14549 | |
| 14550 | @need 1250 |
| 14551 | @noindent |
| 14552 | which constructs a new list in which the first argument to @code{cons} |
| 14553 | becomes the first element of the new list: |
| 14554 | |
| 14555 | @smallexample |
| 14556 | ((1 2 3 4) 5 6 7 8) |
| 14557 | @end smallexample |
| 14558 | |
| 14559 | @node Several files recursively, Prepare the data, Several files, Words in a defun |
| 14560 | @section Recursively Count Words in Different Files |
| 14561 | |
| 14562 | Besides a @code{while} loop, you can work on each of a list of files |
| 14563 | with recursion. A recursive version of @code{lengths-list-many-files} |
| 14564 | is short and simple. |
| 14565 | |
| 14566 | The recursive function has the usual parts: the `do-again-test', the |
| 14567 | `next-step-expression', and the recursive call. The `do-again-test' |
| 14568 | determines whether the function should call itself again, which it |
| 14569 | will do if the @code{list-of-files} contains any remaining elements; |
| 14570 | the `next-step-expression' resets the @code{list-of-files} to the |
| 14571 | @sc{cdr} of itself, so eventually the list will be empty; and the |
| 14572 | recursive call calls itself on the shorter list. The complete |
| 14573 | function is shorter than this description! |
| 14574 | @findex recursive-lengths-list-many-files |
| 14575 | |
| 14576 | @smallexample |
| 14577 | @group |
| 14578 | (defun recursive-lengths-list-many-files (list-of-files) |
| 14579 | "Return list of lengths of each defun in LIST-OF-FILES." |
| 14580 | (if list-of-files ; @r{do-again-test} |
| 14581 | (append |
| 14582 | (lengths-list-file |
| 14583 | (expand-file-name (car list-of-files))) |
| 14584 | (recursive-lengths-list-many-files |
| 14585 | (cdr list-of-files))))) |
| 14586 | @end group |
| 14587 | @end smallexample |
| 14588 | |
| 14589 | @noindent |
| 14590 | In a sentence, the function returns the lengths' list for the first of |
| 14591 | the @code{list-of-files} appended to the result of calling itself on |
| 14592 | the rest of the @code{list-of-files}. |
| 14593 | |
| 14594 | Here is a test of @code{recursive-lengths-list-many-files}, along with |
| 14595 | the results of running @code{lengths-list-file} on each of the files |
| 14596 | individually. |
| 14597 | |
| 14598 | Install @code{recursive-lengths-list-many-files} and |
| 14599 | @code{lengths-list-file}, if necessary, and then evaluate the |
| 14600 | following expressions. You may need to change the files' pathnames; |
| 14601 | those here work when this Info file and the Emacs sources are located |
| 14602 | in their customary places. To change the expressions, copy them to |
| 14603 | the @file{*scratch*} buffer, edit them, and then evaluate them. |
| 14604 | |
| 14605 | The results are shown after the @samp{@result{}}. (These results are |
| 14606 | for files from Emacs Version 21.0.100; files from other versions of |
| 14607 | Emacs may produce different results.) |
| 14608 | |
| 14609 | @c !!! 21.0.100 lisp sources location here |
| 14610 | @smallexample |
| 14611 | @group |
| 14612 | (cd "/usr/local/share/emacs/21.0.100/") |
| 14613 | |
| 14614 | (lengths-list-file "./lisp/macros.el") |
| 14615 | @result{} (273 263 456 90) |
| 14616 | @end group |
| 14617 | |
| 14618 | @group |
| 14619 | (lengths-list-file "./lisp/mail/mailalias.el") |
| 14620 | @result{} (38 32 26 77 174 180 321 198 324) |
| 14621 | @end group |
| 14622 | |
| 14623 | @group |
| 14624 | (lengths-list-file "./lisp/makesum.el") |
| 14625 | @result{} (85 181) |
| 14626 | @end group |
| 14627 | |
| 14628 | @group |
| 14629 | (recursive-lengths-list-many-files |
| 14630 | '("./lisp/macros.el" |
| 14631 | "./lisp/mail/mailalias.el" |
| 14632 | "./lisp/makesum.el")) |
| 14633 | @result{} (273 263 456 90 38 32 26 77 174 180 321 198 324 85 181) |
| 14634 | @end group |
| 14635 | @end smallexample |
| 14636 | |
| 14637 | The @code{recursive-lengths-list-many-files} function produces the |
| 14638 | output we want. |
| 14639 | |
| 14640 | The next step is to prepare the data in the list for display in a graph. |
| 14641 | |
| 14642 | @node Prepare the data, , Several files recursively, Words in a defun |
| 14643 | @section Prepare the Data for Display in a Graph |
| 14644 | |
| 14645 | The @code{recursive-lengths-list-many-files} function returns a list |
| 14646 | of numbers. Each number records the length of a function definition. |
| 14647 | What we need to do now is transform this data into a list of numbers |
| 14648 | suitable for generating a graph. The new list will tell how many |
| 14649 | functions definitions contain less than 10 words and |
| 14650 | symbols, how many contain between 10 and 19 words and symbols, how |
| 14651 | many contain between 20 and 29 words and symbols, and so on. |
| 14652 | |
| 14653 | In brief, we need to go through the lengths' list produced by the |
| 14654 | @code{recursive-lengths-list-many-files} function and count the number |
| 14655 | of defuns within each range of lengths, and produce a list of those |
| 14656 | numbers. |
| 14657 | |
| 14658 | Based on what we have done before, we can readily foresee that it |
| 14659 | should not be too hard to write a function that `@sc{cdr}s' down the |
| 14660 | lengths' list, looks at each element, determines which length range it |
| 14661 | is in, and increments a counter for that range. |
| 14662 | |
| 14663 | However, before beginning to write such a function, we should consider |
| 14664 | the advantages of sorting the lengths' list first, so the numbers are |
| 14665 | ordered from smallest to largest. First, sorting will make it easier |
| 14666 | to count the numbers in each range, since two adjacent numbers will |
| 14667 | either be in the same length range or in adjacent ranges. Second, by |
| 14668 | inspecting a sorted list, we can discover the highest and lowest |
| 14669 | number, and thereby determine the largest and smallest length range |
| 14670 | that we will need. |
| 14671 | |
| 14672 | @menu |
| 14673 | * Sorting:: Sorting lists. |
| 14674 | * Files List:: Making a list of files. |
| 14675 | * Counting function definitions:: |
| 14676 | @end menu |
| 14677 | |
| 14678 | @node Sorting, Files List, Prepare the data, Prepare the data |
| 14679 | @subsection Sorting Lists |
| 14680 | @findex sort |
| 14681 | |
| 14682 | Emacs contains a function to sort lists, called (as you might guess) |
| 14683 | @code{sort}. The @code{sort} function takes two arguments, the list |
| 14684 | to be sorted, and a predicate that determines whether the first of |
| 14685 | two list elements is ``less'' than the second. |
| 14686 | |
| 14687 | As we saw earlier (@pxref{Wrong Type of Argument, , Using the Wrong |
| 14688 | Type Object as an Argument}), a predicate is a function that |
| 14689 | determines whether some property is true or false. The @code{sort} |
| 14690 | function will reorder a list according to whatever property the |
| 14691 | predicate uses; this means that @code{sort} can be used to sort |
| 14692 | non-numeric lists by non-numeric criteria---it can, for example, |
| 14693 | alphabetize a list. |
| 14694 | |
| 14695 | @need 1250 |
| 14696 | The @code{<} function is used when sorting a numeric list. For example, |
| 14697 | |
| 14698 | @smallexample |
| 14699 | (sort '(4 8 21 17 33 7 21 7) '<) |
| 14700 | @end smallexample |
| 14701 | |
| 14702 | @need 800 |
| 14703 | @noindent |
| 14704 | produces this: |
| 14705 | |
| 14706 | @smallexample |
| 14707 | (4 7 7 8 17 21 21 33) |
| 14708 | @end smallexample |
| 14709 | |
| 14710 | @noindent |
| 14711 | (Note that in this example, both the arguments are quoted so that the |
| 14712 | symbols are not evaluated before being passed to @code{sort} as |
| 14713 | arguments.) |
| 14714 | |
| 14715 | Sorting the list returned by the |
| 14716 | @code{recursive-lengths-list-many-files} function is straightforward; |
| 14717 | it uses the @code{<} function: |
| 14718 | |
| 14719 | @smallexample |
| 14720 | @group |
| 14721 | (sort |
| 14722 | (recursive-lengths-list-many-files |
| 14723 | '("../lisp/macros.el" |
| 14724 | "../lisp/mailalias.el" |
| 14725 | "../lisp/makesum.el")) |
| 14726 | '<) |
| 14727 | @end group |
| 14728 | @end smallexample |
| 14729 | |
| 14730 | @need 800 |
| 14731 | @noindent |
| 14732 | which produces: |
| 14733 | |
| 14734 | @smallexample |
| 14735 | (85 86 116 122 154 176 179 265) |
| 14736 | @end smallexample |
| 14737 | |
| 14738 | @noindent |
| 14739 | (Note that in this example, the first argument to @code{sort} is not |
| 14740 | quoted, since the expression must be evaluated so as to produce the |
| 14741 | list that is passed to @code{sort}.) |
| 14742 | |
| 14743 | @node Files List, Counting function definitions, Sorting, Prepare the data |
| 14744 | @subsection Making a List of Files |
| 14745 | |
| 14746 | The @code{recursive-lengths-list-many-files} function requires a list |
| 14747 | of files as its argument. For our test examples, we constructed such |
| 14748 | a list by hand; but the Emacs Lisp source directory is too large for |
| 14749 | us to do for that. Instead, we will write a function to do the job |
| 14750 | for us. In this function, we will use both a @code{while} loop and a |
| 14751 | recursive call. |
| 14752 | |
| 14753 | @findex directory-files |
| 14754 | We did not have to write a function like this for older versions of |
| 14755 | GNU Emacs, since they placed all the @samp{.el} files in one |
| 14756 | directory. Instead, we were able to use the @code{directory-files} |
| 14757 | function, which lists the names of files that match a specified |
| 14758 | pattern within a single directory. |
| 14759 | |
| 14760 | However, recent versions of Emacs place Emacs Lisp files in |
| 14761 | sub-directories of the top level @file{lisp} directory. This |
| 14762 | re-arrangement eases navigation. For example, all the mail related |
| 14763 | files are in a @file{lisp} sub-directory called @file{mail}. But at |
| 14764 | the same time, this arrangement forces us to create a file listing |
| 14765 | function that descends into the sub-directories. |
| 14766 | |
| 14767 | @findex files-in-below-directory |
| 14768 | We can create this function, called @code{files-in-below-directory}, |
| 14769 | using familiar functions such as @code{car}, @code{nthcdr}, and |
| 14770 | @code{substring} in conjunction with an existing function called |
| 14771 | @code{directory-files-and-attributes}. This latter function not only |
| 14772 | lists all the filenames in a directory, including the names |
| 14773 | of sub-directories, but also their attributes. |
| 14774 | |
| 14775 | To restate our goal: to create a function that will enable us |
| 14776 | to feed filenames to @code{recursive-lengths-list-many-files} |
| 14777 | as a list that looks like this (but with more elements): |
| 14778 | |
| 14779 | @smallexample |
| 14780 | @group |
| 14781 | ("../lisp/macros.el" |
| 14782 | "../lisp/mail/rmail.el" |
| 14783 | "../lisp/makesum.el") |
| 14784 | @end group |
| 14785 | @end smallexample |
| 14786 | |
| 14787 | The @code{directory-files-and-attributes} function returns a list of |
| 14788 | lists. Each of the lists within the main list consists of 13 |
| 14789 | elements. The first element is a string that contains the name of the |
| 14790 | file -- which, in GNU/Linux, may be a `directory file', that is to |
| 14791 | say, a file with the special attributes of a directory. The second |
| 14792 | element of the list is @code{t} for a directory, a string |
| 14793 | for symbolic link (the string is the name linked to), or @code{nil}. |
| 14794 | |
| 14795 | For example, the first @samp{.el} file in the @file{lisp/} directory |
| 14796 | is @file{abbrev.el}. Its name is |
| 14797 | @file{/usr/local/share/emacs/21.0.100/lisp/abbrev.el} and it is not a |
| 14798 | directory or a symbolic link. |
| 14799 | |
| 14800 | @need 1000 |
| 14801 | This is how @code{directory-files-and-attributes} lists that file and |
| 14802 | its attributes: |
| 14803 | |
| 14804 | @smallexample |
| 14805 | @group |
| 14806 | ("/usr/local/share/emacs/21.0.100/lisp/abbrev.el" |
| 14807 | nil |
| 14808 | 1 |
| 14809 | 1000 |
| 14810 | 100 |
| 14811 | @end group |
| 14812 | @group |
| 14813 | (15019 32380) |
| 14814 | (14883 48041) |
| 14815 | (15214 49336) |
| 14816 | 11583 |
| 14817 | "-rw-rw-r--" |
| 14818 | @end group |
| 14819 | @group |
| 14820 | t |
| 14821 | 341385 |
| 14822 | 776) |
| 14823 | @end group |
| 14824 | @end smallexample |
| 14825 | |
| 14826 | @need 1200 |
| 14827 | On the other hand, @file{mail/} is a directory within the @file{lisp/} |
| 14828 | directory. The beginning of its listing looks like this: |
| 14829 | |
| 14830 | @smallexample |
| 14831 | @group |
| 14832 | ("/usr/local/share/emacs/21.0.100/lisp/mail" |
| 14833 | t |
| 14834 | @dots{} |
| 14835 | ) |
| 14836 | @end group |
| 14837 | @end smallexample |
| 14838 | |
| 14839 | (Look at the documentation of @code{file-attributes} to learn about |
| 14840 | the different attributes. Bear in mind that the |
| 14841 | @code{file-attributes} function does not list the filename, so its |
| 14842 | first element is @code{directory-files-and-attributes}'s second |
| 14843 | element.) |
| 14844 | |
| 14845 | We will want our new function, @code{files-in-below-directory}, to |
| 14846 | list the @samp{.el} files in the directory it is told to check, and in |
| 14847 | any directories below that directory. |
| 14848 | |
| 14849 | This gives us a hint on how to construct |
| 14850 | @code{files-in-below-directory}: within a directory, the function |
| 14851 | should add @samp{.el} filenames to a list; and if, within a directory, |
| 14852 | the function comes upon a sub-directory, it should go into that |
| 14853 | sub-directory and repeat its actions. |
| 14854 | |
| 14855 | However, we should note that every directory contains a name that |
| 14856 | refers to itself, called @file{.}, (``dot'') and a name that refers to |
| 14857 | its parent directory, called @file{..} (``double dot''). (In |
| 14858 | @file{/}, the root directory, @file{..} refers to itself, since |
| 14859 | @file{/} has no parent.) Clearly, we do not want our |
| 14860 | @code{files-in-below-directory} function to enter those directories, |
| 14861 | since they always lead us, directly or indirectly, to the current |
| 14862 | directory. |
| 14863 | |
| 14864 | Consequently, our @code{files-in-below-directory} function must do |
| 14865 | several tasks: |
| 14866 | |
| 14867 | @itemize @bullet |
| 14868 | @item |
| 14869 | Check to see whether it is looking at a filename that ends in |
| 14870 | @samp{.el}; and if so, add its name to a list. |
| 14871 | |
| 14872 | @item |
| 14873 | Check to see whether it is looking at a filename that is the name of a |
| 14874 | directory; and if so, |
| 14875 | |
| 14876 | @itemize @minus |
| 14877 | @item |
| 14878 | Check to see whether it is looking at @file{.} or @file{..}; and if |
| 14879 | so skip it. |
| 14880 | |
| 14881 | @item |
| 14882 | Or else, go into that directory and repeat the process. |
| 14883 | @end itemize |
| 14884 | @end itemize |
| 14885 | |
| 14886 | Let's write a function definition to do these tasks. We will use a |
| 14887 | @code{while} loop to move from one filename to another within a |
| 14888 | directory, checking what needs to be done; and we will use a recursive |
| 14889 | call to repeat the actions on each sub-directory. The recursive |
| 14890 | pattern is `accumulate' |
| 14891 | (@pxref{Accumulate, , Recursive Pattern: @emph{accumulate}}), |
| 14892 | using @code{append} as the combiner. |
| 14893 | |
| 14894 | @ignore |
| 14895 | (directory-files "/usr/local/share/emacs/21.0.100/lisp/" t "\\.el$") |
| 14896 | (shell-command "find /usr/local/share/emacs/21.0.100/lisp/ -name '*.el'") |
| 14897 | @end ignore |
| 14898 | |
| 14899 | @c /usr/local/share/emacs/21.0.100/lisp/ |
| 14900 | |
| 14901 | @need 800 |
| 14902 | Here is the function: |
| 14903 | |
| 14904 | @smallexample |
| 14905 | @group |
| 14906 | (defun files-in-below-directory (directory) |
| 14907 | "List the .el files in DIRECTORY and in its sub-directories." |
| 14908 | ;; Although the function will be used non-interactively, |
| 14909 | ;; it will be easier to test if we make it interactive. |
| 14910 | ;; The directory will have a name such as |
| 14911 | ;; "/usr/local/share/emacs/21.0.100/lisp/" |
| 14912 | (interactive "DDirectory name: ") |
| 14913 | @end group |
| 14914 | @group |
| 14915 | (let (el-files-list |
| 14916 | (current-directory-list |
| 14917 | (directory-files-and-attributes directory t))) |
| 14918 | ;; while we are in the current directory |
| 14919 | (while current-directory-list |
| 14920 | @end group |
| 14921 | @group |
| 14922 | (cond |
| 14923 | ;; check to see whether filename ends in `.el' |
| 14924 | ;; and if so, append its name to a list. |
| 14925 | ((equal ".el" (substring (car (car current-directory-list)) -3)) |
| 14926 | (setq el-files-list |
| 14927 | (cons (car (car current-directory-list)) el-files-list))) |
| 14928 | @end group |
| 14929 | @group |
| 14930 | ;; check whether filename is that of a directory |
| 14931 | ((eq t (car (cdr (car current-directory-list)))) |
| 14932 | ;; decide whether to skip or recurse |
| 14933 | (if |
| 14934 | (equal "." |
| 14935 | (substring (car (car current-directory-list)) -1)) |
| 14936 | ;; then do nothing since filename is that of |
| 14937 | ;; current directory or parent, "." or ".." |
| 14938 | () |
| 14939 | @end group |
| 14940 | @group |
| 14941 | ;; else descend into the directory and repeat the process |
| 14942 | (setq el-files-list |
| 14943 | (append |
| 14944 | (files-in-below-directory |
| 14945 | (car (car current-directory-list))) |
| 14946 | el-files-list))))) |
| 14947 | ;; move to the next filename in the list; this also |
| 14948 | ;; shortens the list so the while loop eventually comes to an end |
| 14949 | (setq current-directory-list (cdr current-directory-list))) |
| 14950 | ;; return the filenames |
| 14951 | el-files-list)) |
| 14952 | @end group |
| 14953 | @end smallexample |
| 14954 | |
| 14955 | @c (files-in-below-directory "/usr/local/share/emacs/21.0.100/lisp/") |
| 14956 | |
| 14957 | The @code{files-in-below-directory} @code{directory-files} function |
| 14958 | takes one argument, the name of a directory. |
| 14959 | |
| 14960 | @need 1250 |
| 14961 | Thus, on my system, |
| 14962 | |
| 14963 | @c !!! 21.0.100 lisp sources location here |
| 14964 | @smallexample |
| 14965 | @group |
| 14966 | (length |
| 14967 | (files-in-below-directory "/usr/local/share/emacs/21.0.100/lisp/")) |
| 14968 | @end group |
| 14969 | @end smallexample |
| 14970 | |
| 14971 | @noindent |
| 14972 | tells me that my version 21.0.100 Lisp sources directory contains 754 |
| 14973 | @samp{.el} files. |
| 14974 | |
| 14975 | @code{files-in-below-directory} returns a list in reverse alphabetical |
| 14976 | order. An expression to sort the list in alphabetical order looks |
| 14977 | like this: |
| 14978 | |
| 14979 | @smallexample |
| 14980 | @group |
| 14981 | (sort |
| 14982 | (files-in-below-directory "/usr/local/share/emacs/21.0.100/lisp/") |
| 14983 | 'string-lessp) |
| 14984 | @end group |
| 14985 | @end smallexample |
| 14986 | |
| 14987 | @ignore |
| 14988 | (defun test () |
| 14989 | "Test how long it takes to find lengths of all elisp defuns." |
| 14990 | (insert "\n" (current-time-string) "\n") |
| 14991 | (sit-for 0) |
| 14992 | (sort |
| 14993 | (recursive-lengths-list-many-files |
| 14994 | '("../lisp/macros.el" |
| 14995 | "../lisp/mailalias.el" |
| 14996 | "../lisp/makesum.el")) |
| 14997 | '<) |
| 14998 | (insert (format "%s" (current-time-string)))) |
| 14999 | |
| 15000 | @end ignore |
| 15001 | |
| 15002 | @node Counting function definitions, , Files List, Prepare the data |
| 15003 | @subsection Counting function definitions |
| 15004 | |
| 15005 | Our immediate goal is to generate a list that tells us how many |
| 15006 | function definitions contain fewer than 10 words and symbols, how many |
| 15007 | contain between 10 and 19 words and symbols, how many contain between |
| 15008 | 20 and 29 words and symbols, and so on. |
| 15009 | |
| 15010 | With a sorted list of numbers, this is easy: count how many elements |
| 15011 | of the list are smaller than 10, then, after moving past the numbers |
| 15012 | just counted, count how many are smaller than 20, then, after moving |
| 15013 | past the numbers just counted, count how many are smaller than 30, and |
| 15014 | so on. Each of the numbers, 10, 20, 30, 40, and the like, is one |
| 15015 | larger than the top of that range. We can call the list of such |
| 15016 | numbers the @code{top-of-ranges} list. |
| 15017 | |
| 15018 | @need 1200 |
| 15019 | If we wished, we could generate this list automatically, but it is |
| 15020 | simpler to write a list manually. Here it is: |
| 15021 | @vindex top-of-ranges |
| 15022 | |
| 15023 | @smallexample |
| 15024 | @group |
| 15025 | (defvar top-of-ranges |
| 15026 | '(10 20 30 40 50 |
| 15027 | 60 70 80 90 100 |
| 15028 | 110 120 130 140 150 |
| 15029 | 160 170 180 190 200 |
| 15030 | 210 220 230 240 250 |
| 15031 | 260 270 280 290 300) |
| 15032 | "List specifying ranges for `defuns-per-range'.") |
| 15033 | @end group |
| 15034 | @end smallexample |
| 15035 | |
| 15036 | To change the ranges, we edit this list. |
| 15037 | |
| 15038 | Next, we need to write the function that creates the list of the |
| 15039 | number of definitions within each range. Clearly, this function must |
| 15040 | take the @code{sorted-lengths} and the @code{top-of-ranges} lists |
| 15041 | as arguments. |
| 15042 | |
| 15043 | The @code{defuns-per-range} function must do two things again and |
| 15044 | again: it must count the number of definitions within a range |
| 15045 | specified by the current top-of-range value; and it must shift to the |
| 15046 | next higher value in the @code{top-of-ranges} list after counting the |
| 15047 | number of definitions in the current range. Since each of these |
| 15048 | actions is repetitive, we can use @code{while} loops for the job. |
| 15049 | One loop counts the number of definitions in the range defined by the |
| 15050 | current top-of-range value, and the other loop selects each of the |
| 15051 | top-of-range values in turn. |
| 15052 | |
| 15053 | Several entries of the @code{sorted-lengths} list are counted for each |
| 15054 | range; this means that the loop for the @code{sorted-lengths} list |
| 15055 | will be inside the loop for the @code{top-of-ranges} list, like a |
| 15056 | small gear inside a big gear. |
| 15057 | |
| 15058 | The inner loop counts the number of definitions within the range. It |
| 15059 | is a simple counting loop of the type we have seen before. |
| 15060 | (@xref{Incrementing Loop, , A loop with an incrementing counter}.) |
| 15061 | The true-or-false test of the loop tests whether the value from the |
| 15062 | @code{sorted-lengths} list is smaller than the current value of the |
| 15063 | top of the range. If it is, the function increments the counter and |
| 15064 | tests the next value from the @code{sorted-lengths} list. |
| 15065 | |
| 15066 | @need 1250 |
| 15067 | The inner loop looks like this: |
| 15068 | |
| 15069 | @smallexample |
| 15070 | @group |
| 15071 | (while @var{length-element-smaller-than-top-of-range} |
| 15072 | (setq number-within-range (1+ number-within-range)) |
| 15073 | (setq sorted-lengths (cdr sorted-lengths))) |
| 15074 | @end group |
| 15075 | @end smallexample |
| 15076 | |
| 15077 | The outer loop must start with the lowest value of the |
| 15078 | @code{top-of-ranges} list, and then be set to each of the succeeding |
| 15079 | higher values in turn. This can be done with a loop like this: |
| 15080 | |
| 15081 | @smallexample |
| 15082 | @group |
| 15083 | (while top-of-ranges |
| 15084 | @var{body-of-loop}@dots{} |
| 15085 | (setq top-of-ranges (cdr top-of-ranges))) |
| 15086 | @end group |
| 15087 | @end smallexample |
| 15088 | |
| 15089 | @need 1200 |
| 15090 | Put together, the two loops look like this: |
| 15091 | |
| 15092 | @smallexample |
| 15093 | @group |
| 15094 | (while top-of-ranges |
| 15095 | |
| 15096 | ;; @r{Count the number of elements within the current range.} |
| 15097 | (while @var{length-element-smaller-than-top-of-range} |
| 15098 | (setq number-within-range (1+ number-within-range)) |
| 15099 | (setq sorted-lengths (cdr sorted-lengths))) |
| 15100 | |
| 15101 | ;; @r{Move to next range.} |
| 15102 | (setq top-of-ranges (cdr top-of-ranges))) |
| 15103 | @end group |
| 15104 | @end smallexample |
| 15105 | |
| 15106 | In addition, in each circuit of the outer loop, Emacs should record |
| 15107 | the number of definitions within that range (the value of |
| 15108 | @code{number-within-range}) in a list. We can use @code{cons} for |
| 15109 | this purpose. (@xref{cons, , @code{cons}}.) |
| 15110 | |
| 15111 | The @code{cons} function works fine, except that the list it |
| 15112 | constructs will contain the number of definitions for the highest |
| 15113 | range at its beginning and the number of definitions for the lowest |
| 15114 | range at its end. This is because @code{cons} attaches new elements |
| 15115 | of the list to the beginning of the list, and since the two loops are |
| 15116 | working their way through the lengths' list from the lower end first, |
| 15117 | the @code{defuns-per-range-list} will end up largest number first. |
| 15118 | But we will want to print our graph with smallest values first and the |
| 15119 | larger later. The solution is to reverse the order of the |
| 15120 | @code{defuns-per-range-list}. We can do this using the |
| 15121 | @code{nreverse} function, which reverses the order of a list. |
| 15122 | @findex nreverse |
| 15123 | |
| 15124 | @need 800 |
| 15125 | For example, |
| 15126 | |
| 15127 | @smallexample |
| 15128 | (nreverse '(1 2 3 4)) |
| 15129 | @end smallexample |
| 15130 | |
| 15131 | @need 800 |
| 15132 | @noindent |
| 15133 | produces: |
| 15134 | |
| 15135 | @smallexample |
| 15136 | (4 3 2 1) |
| 15137 | @end smallexample |
| 15138 | |
| 15139 | Note that the @code{nreverse} function is ``destructive''---that is, |
| 15140 | it changes the list to which it is applied; this contrasts with the |
| 15141 | @code{car} and @code{cdr} functions, which are non-destructive. In |
| 15142 | this case, we do not want the original @code{defuns-per-range-list}, |
| 15143 | so it does not matter that it is destroyed. (The @code{reverse} |
| 15144 | function provides a reversed copy of a list, leaving the original list |
| 15145 | as is.) |
| 15146 | @findex reverse |
| 15147 | |
| 15148 | @need 1250 |
| 15149 | Put all together, the @code{defuns-per-range} looks like this: |
| 15150 | |
| 15151 | @smallexample |
| 15152 | @group |
| 15153 | (defun defuns-per-range (sorted-lengths top-of-ranges) |
| 15154 | "SORTED-LENGTHS defuns in each TOP-OF-RANGES range." |
| 15155 | (let ((top-of-range (car top-of-ranges)) |
| 15156 | (number-within-range 0) |
| 15157 | defuns-per-range-list) |
| 15158 | @end group |
| 15159 | |
| 15160 | @group |
| 15161 | ;; @r{Outer loop.} |
| 15162 | (while top-of-ranges |
| 15163 | @end group |
| 15164 | |
| 15165 | @group |
| 15166 | ;; @r{Inner loop.} |
| 15167 | (while (and |
| 15168 | ;; @r{Need number for numeric test.} |
| 15169 | (car sorted-lengths) |
| 15170 | (< (car sorted-lengths) top-of-range)) |
| 15171 | @end group |
| 15172 | |
| 15173 | @group |
| 15174 | ;; @r{Count number of definitions within current range.} |
| 15175 | (setq number-within-range (1+ number-within-range)) |
| 15176 | (setq sorted-lengths (cdr sorted-lengths))) |
| 15177 | |
| 15178 | ;; @r{Exit inner loop but remain within outer loop.} |
| 15179 | @end group |
| 15180 | |
| 15181 | @group |
| 15182 | (setq defuns-per-range-list |
| 15183 | (cons number-within-range defuns-per-range-list)) |
| 15184 | (setq number-within-range 0) ; @r{Reset count to zero.} |
| 15185 | @end group |
| 15186 | |
| 15187 | @group |
| 15188 | ;; @r{Move to next range.} |
| 15189 | (setq top-of-ranges (cdr top-of-ranges)) |
| 15190 | ;; @r{Specify next top of range value.} |
| 15191 | (setq top-of-range (car top-of-ranges))) |
| 15192 | @end group |
| 15193 | |
| 15194 | @group |
| 15195 | ;; @r{Exit outer loop and count the number of defuns larger than} |
| 15196 | ;; @r{ the largest top-of-range value.} |
| 15197 | (setq defuns-per-range-list |
| 15198 | (cons |
| 15199 | (length sorted-lengths) |
| 15200 | defuns-per-range-list)) |
| 15201 | @end group |
| 15202 | |
| 15203 | @group |
| 15204 | ;; @r{Return a list of the number of definitions within each range,} |
| 15205 | ;; @r{ smallest to largest.} |
| 15206 | (nreverse defuns-per-range-list))) |
| 15207 | @end group |
| 15208 | @end smallexample |
| 15209 | |
| 15210 | @need 1200 |
| 15211 | @noindent |
| 15212 | The function is straightforward except for one subtle feature. The |
| 15213 | true-or-false test of the inner loop looks like this: |
| 15214 | |
| 15215 | @smallexample |
| 15216 | @group |
| 15217 | (and (car sorted-lengths) |
| 15218 | (< (car sorted-lengths) top-of-range)) |
| 15219 | @end group |
| 15220 | @end smallexample |
| 15221 | |
| 15222 | @need 800 |
| 15223 | @noindent |
| 15224 | instead of like this: |
| 15225 | |
| 15226 | @smallexample |
| 15227 | (< (car sorted-lengths) top-of-range) |
| 15228 | @end smallexample |
| 15229 | |
| 15230 | The purpose of the test is to determine whether the first item in the |
| 15231 | @code{sorted-lengths} list is less than the value of the top of the |
| 15232 | range. |
| 15233 | |
| 15234 | The simple version of the test works fine unless the |
| 15235 | @code{sorted-lengths} list has a @code{nil} value. In that case, the |
| 15236 | @code{(car sorted-lengths)} expression function returns |
| 15237 | @code{nil}. The @code{<} function cannot compare a number to |
| 15238 | @code{nil}, which is an empty list, so Emacs signals an error and |
| 15239 | stops the function from attempting to continue to execute. |
| 15240 | |
| 15241 | The @code{sorted-lengths} list always becomes @code{nil} when the |
| 15242 | counter reaches the end of the list. This means that any attempt to |
| 15243 | use the @code{defuns-per-range} function with the simple version of |
| 15244 | the test will fail. |
| 15245 | |
| 15246 | We solve the problem by using the @code{(car sorted-lengths)} |
| 15247 | expression in conjunction with the @code{and} expression. The |
| 15248 | @code{(car sorted-lengths)} expression returns a non-@code{nil} |
| 15249 | value so long as the list has at least one number within it, but |
| 15250 | returns @code{nil} if the list is empty. The @code{and} expression |
| 15251 | first evaluates the @code{(car sorted-lengths)} expression, and |
| 15252 | if it is @code{nil}, returns false @emph{without} evaluating the |
| 15253 | @code{<} expression. But if the @code{(car sorted-lengths)} |
| 15254 | expression returns a non-@code{nil} value, the @code{and} expression |
| 15255 | evaluates the @code{<} expression, and returns that value as the value |
| 15256 | of the @code{and} expression. |
| 15257 | |
| 15258 | @c colon in printed section title causes problem in Info cross reference |
| 15259 | This way, we avoid an error. |
| 15260 | @iftex |
| 15261 | @xref{forward-paragraph, , @code{forward-paragraph}: a Goldmine of |
| 15262 | Functions}, for more information about @code{and}. |
| 15263 | @end iftex |
| 15264 | @ifinfo |
| 15265 | @xref{forward-paragraph}, for more information about @code{and}. |
| 15266 | @end ifinfo |
| 15267 | |
| 15268 | Here is a short test of the @code{defuns-per-range} function. First, |
| 15269 | evaluate the expression that binds (a shortened) |
| 15270 | @code{top-of-ranges} list to the list of values, then evaluate the |
| 15271 | expression for binding the @code{sorted-lengths} list, and then |
| 15272 | evaluate the @code{defuns-per-range} function. |
| 15273 | |
| 15274 | @smallexample |
| 15275 | @group |
| 15276 | ;; @r{(Shorter list than we will use later.)} |
| 15277 | (setq top-of-ranges |
| 15278 | '(110 120 130 140 150 |
| 15279 | 160 170 180 190 200)) |
| 15280 | |
| 15281 | (setq sorted-lengths |
| 15282 | '(85 86 110 116 122 129 154 176 179 200 265 300 300)) |
| 15283 | |
| 15284 | (defuns-per-range sorted-lengths top-of-ranges) |
| 15285 | @end group |
| 15286 | @end smallexample |
| 15287 | |
| 15288 | @need 800 |
| 15289 | @noindent |
| 15290 | The list returned looks like this: |
| 15291 | |
| 15292 | @smallexample |
| 15293 | (2 2 2 0 0 1 0 2 0 0 4) |
| 15294 | @end smallexample |
| 15295 | |
| 15296 | @noindent |
| 15297 | Indeed, there are two elements of the @code{sorted-lengths} list |
| 15298 | smaller than 110, two elements between 110 and 119, two elements |
| 15299 | between 120 and 129, and so on. There are four elements with a value |
| 15300 | of 200 or larger. |
| 15301 | |
| 15302 | @c The next step is to turn this numbers' list into a graph. |
| 15303 | |
| 15304 | @node Readying a Graph, Emacs Initialization, Words in a defun, Top |
| 15305 | @chapter Readying a Graph |
| 15306 | @cindex Readying a graph |
| 15307 | @cindex Graph prototype |
| 15308 | @cindex Prototype graph |
| 15309 | @cindex Body of graph |
| 15310 | |
| 15311 | Our goal is to construct a graph showing the numbers of function |
| 15312 | definitions of various lengths in the Emacs lisp sources. |
| 15313 | |
| 15314 | As a practical matter, if you were creating a graph, you would |
| 15315 | probably use a program such as @code{gnuplot} to do the job. |
| 15316 | (@code{gnuplot} is nicely integrated into GNU Emacs.) In this case, |
| 15317 | however, we create one from scratch, and in the process we will |
| 15318 | re-acquaint ourselves with some of what we learned before and learn |
| 15319 | more. |
| 15320 | |
| 15321 | In this chapter, we will first write a simple graph printing function. |
| 15322 | This first definition will be a @dfn{prototype}, a rapidly written |
| 15323 | function that enables us to reconnoiter this unknown graph-making |
| 15324 | territory. We will discover dragons, or find that they are myth. |
| 15325 | After scouting the terrain, we will feel more confident and enhance |
| 15326 | the function to label the axes automatically. |
| 15327 | |
| 15328 | @menu |
| 15329 | * Columns of a graph:: |
| 15330 | * graph-body-print:: How to print the body of a graph. |
| 15331 | * recursive-graph-body-print:: |
| 15332 | * Printed Axes:: |
| 15333 | * Line Graph Exercise:: |
| 15334 | @end menu |
| 15335 | |
| 15336 | @node Columns of a graph, graph-body-print, Readying a Graph, Readying a Graph |
| 15337 | @ifnottex |
| 15338 | @unnumberedsec Printing the Columns of a Graph |
| 15339 | @end ifnottex |
| 15340 | |
| 15341 | Since Emacs is designed to be flexible and work with all kinds of |
| 15342 | terminals, including character-only terminals, the graph will need to |
| 15343 | be made from one of the `typewriter' symbols. An asterisk will do; as |
| 15344 | we enhance the graph-printing function, we can make the choice of |
| 15345 | symbol a user option. |
| 15346 | |
| 15347 | We can call this function @code{graph-body-print}; it will take a |
| 15348 | @code{numbers-list} as its only argument. At this stage, we will not |
| 15349 | label the graph, but only print its body. |
| 15350 | |
| 15351 | The @code{graph-body-print} function inserts a vertical column of |
| 15352 | asterisks for each element in the @code{numbers-list}. The height of |
| 15353 | each line is determined by the value of that element of the |
| 15354 | @code{numbers-list}. |
| 15355 | |
| 15356 | Inserting columns is a repetitive act; that means that this function can |
| 15357 | be written either with a @code{while} loop or recursively. |
| 15358 | |
| 15359 | Our first challenge is to discover how to print a column of asterisks. |
| 15360 | Usually, in Emacs, we print characters onto a screen horizontally, |
| 15361 | line by line, by typing. We have two routes we can follow: write our |
| 15362 | own column-insertion function or discover whether one exists in Emacs. |
| 15363 | |
| 15364 | To see whether there is one in Emacs, we can use the @kbd{M-x apropos} |
| 15365 | command. This command is like the @kbd{C-h a} (command-apropos) |
| 15366 | command, except that the latter finds only those functions that are |
| 15367 | commands. The @kbd{M-x apropos} command lists all symbols that match |
| 15368 | a regular expression, including functions that are not interactive. |
| 15369 | @findex apropos |
| 15370 | |
| 15371 | What we want to look for is some command that prints or inserts |
| 15372 | columns. Very likely, the name of the function will contain either |
| 15373 | the word `print' or the word `insert' or the word `column'. |
| 15374 | Therefore, we can simply type @kbd{M-x apropos RET |
| 15375 | print\|insert\|column RET} and look at the result. On my system, this |
| 15376 | command takes quite some time, and then produces a list of 79 |
| 15377 | functions and variables. Scanning down the list, the only function |
| 15378 | that looks as if it might do the job is @code{insert-rectangle}. |
| 15379 | |
| 15380 | @need 1200 |
| 15381 | Indeed, this is the function we want; its documentation says: |
| 15382 | |
| 15383 | @smallexample |
| 15384 | @group |
| 15385 | insert-rectangle: |
| 15386 | Insert text of RECTANGLE with upper left corner at point. |
| 15387 | RECTANGLE's first line is inserted at point, |
| 15388 | its second line is inserted at a point vertically under point, etc. |
| 15389 | RECTANGLE should be a list of strings. |
| 15390 | @end group |
| 15391 | @end smallexample |
| 15392 | |
| 15393 | We can run a quick test, to make sure it does what we expect of it. |
| 15394 | |
| 15395 | Here is the result of placing the cursor after the |
| 15396 | @code{insert-rectangle} expression and typing @kbd{C-u C-x C-e} |
| 15397 | (@code{eval-last-sexp}). The function inserts the strings |
| 15398 | @samp{"first"}, @samp{"second"}, and @samp{"third"} at and below |
| 15399 | point. Also the function returns @code{nil}. |
| 15400 | |
| 15401 | @smallexample |
| 15402 | @group |
| 15403 | (insert-rectangle '("first" "second" "third"))first |
| 15404 | second |
| 15405 | third |
| 15406 | nil |
| 15407 | @end group |
| 15408 | @end smallexample |
| 15409 | |
| 15410 | @noindent |
| 15411 | Of course, we won't be inserting the text of the |
| 15412 | @code{insert-rectangle} expression itself into the buffer in which we |
| 15413 | are making the graph, but will call the function from our program. We |
| 15414 | shall, however, have to make sure that point is in the buffer at the |
| 15415 | place where the @code{insert-rectangle} function will insert its |
| 15416 | column of strings. |
| 15417 | |
| 15418 | If you are reading this in Info, you can see how this works by |
| 15419 | switching to another buffer, such as the @file{*scratch*} buffer, |
| 15420 | placing point somewhere in the buffer, typing @kbd{M-:}, |
| 15421 | typing the @code{insert-rectangle} expression into the minibuffer at |
| 15422 | the prompt, and then typing @key{RET}. This causes Emacs to evaluate |
| 15423 | the expression in the minibuffer, but to use as the value of point the |
| 15424 | position of point in the @file{*scratch*} buffer. (@kbd{M-:} |
| 15425 | is the keybinding for @code{eval-expression}.) |
| 15426 | |
| 15427 | We find when we do this that point ends up at the end of the last |
| 15428 | inserted line---that is to say, this function moves point as a |
| 15429 | side-effect. If we were to repeat the command, with point at this |
| 15430 | position, the next insertion would be below and to the right of the |
| 15431 | previous insertion. We don't want this! If we are going to make a |
| 15432 | bar graph, the columns need to be beside each other. |
| 15433 | |
| 15434 | So we discover that each cycle of the column-inserting @code{while} |
| 15435 | loop must reposition point to the place we want it, and that place |
| 15436 | will be at the top, not the bottom, of the column. Moreover, we |
| 15437 | remember that when we print a graph, we do not expect all the columns |
| 15438 | to be the same height. This means that the top of each column may be |
| 15439 | at a different height from the previous one. We cannot simply |
| 15440 | reposition point to the same line each time, but moved over to the |
| 15441 | right---or perhaps we can@dots{} |
| 15442 | |
| 15443 | We are planning to make the columns of the bar graph out of asterisks. |
| 15444 | The number of asterisks in the column is the number specified by the |
| 15445 | current element of the @code{numbers-list}. We need to construct a |
| 15446 | list of asterisks of the right length for each call to |
| 15447 | @code{insert-rectangle}. If this list consists solely of the requisite |
| 15448 | number of asterisks, then we will have position point the right number |
| 15449 | of lines above the base for the graph to print correctly. This could |
| 15450 | be difficult. |
| 15451 | |
| 15452 | Alternatively, if we can figure out some way to pass |
| 15453 | @code{insert-rectangle} a list of the same length each time, then we |
| 15454 | can place point on the same line each time, but move it over one |
| 15455 | column to the right for each new column. If we do this, however, some |
| 15456 | of the entries in the list passed to @code{insert-rectangle} must be |
| 15457 | blanks rather than asterisks. For example, if the maximum height of |
| 15458 | the graph is 5, but the height of the column is 3, then |
| 15459 | @code{insert-rectangle} requires an argument that looks like this: |
| 15460 | |
| 15461 | @smallexample |
| 15462 | (" " " " "*" "*" "*") |
| 15463 | @end smallexample |
| 15464 | |
| 15465 | This last proposal is not so difficult, so long as we can determine |
| 15466 | the column height. There are two ways for us to specify the column |
| 15467 | height: we can arbitrarily state what it will be, which would work |
| 15468 | fine for graphs of that height; or we can search through the list of |
| 15469 | numbers and use the maximum height of the list as the maximum height |
| 15470 | of the graph. If the latter operation were difficult, then the former |
| 15471 | procedure would be easiest, but there is a function built into Emacs |
| 15472 | that determines the maximum of its arguments. We can use that |
| 15473 | function. The function is called @code{max} and it returns the |
| 15474 | largest of all its arguments, which must be numbers. Thus, for |
| 15475 | example, |
| 15476 | |
| 15477 | @smallexample |
| 15478 | (max 3 4 6 5 7 3) |
| 15479 | @end smallexample |
| 15480 | |
| 15481 | @noindent |
| 15482 | returns 7. (A corresponding function called @code{min} returns the |
| 15483 | smallest of all its arguments.) |
| 15484 | @findex max |
| 15485 | @findex min |
| 15486 | |
| 15487 | However, we cannot simply call @code{max} on the @code{numbers-list}; |
| 15488 | the @code{max} function expects numbers as its argument, not a list of |
| 15489 | numbers. Thus, the following expression, |
| 15490 | |
| 15491 | @smallexample |
| 15492 | (max '(3 4 6 5 7 3)) |
| 15493 | @end smallexample |
| 15494 | |
| 15495 | @need 800 |
| 15496 | @noindent |
| 15497 | produces the following error message; |
| 15498 | |
| 15499 | @smallexample |
| 15500 | Wrong type of argument: number-or-marker-p, (3 4 6 5 7 3) |
| 15501 | @end smallexample |
| 15502 | |
| 15503 | @findex apply |
| 15504 | We need a function that passes a list of arguments to a function. |
| 15505 | This function is @code{apply}. This function `applies' its first |
| 15506 | argument (a function) to its remaining arguments, the last of which |
| 15507 | may be a list. |
| 15508 | |
| 15509 | @need 1250 |
| 15510 | For example, |
| 15511 | |
| 15512 | @smallexample |
| 15513 | (apply 'max 3 4 7 3 '(4 8 5)) |
| 15514 | @end smallexample |
| 15515 | |
| 15516 | @noindent |
| 15517 | returns 8. |
| 15518 | |
| 15519 | (Incidentally, I don't know how you would learn of this function |
| 15520 | without a book such as this. It is possible to discover other |
| 15521 | functions, like @code{search-forward} or @code{insert-rectangle}, by |
| 15522 | guessing at a part of their names and then using @code{apropos}. Even |
| 15523 | though its base in metaphor is clear---`apply' its first argument to |
| 15524 | the rest---I doubt a novice would come up with that particular word |
| 15525 | when using @code{apropos} or other aid. Of course, I could be wrong; |
| 15526 | after all, the function was first named by someone who had to invent |
| 15527 | it.) |
| 15528 | |
| 15529 | The second and subsequent arguments to @code{apply} are optional, so |
| 15530 | we can use @code{apply} to call a function and pass the elements of a |
| 15531 | list to it, like this, which also returns 8: |
| 15532 | |
| 15533 | @smallexample |
| 15534 | (apply 'max '(4 8 5)) |
| 15535 | @end smallexample |
| 15536 | |
| 15537 | This latter way is how we will use @code{apply}. The |
| 15538 | @code{recursive-lengths-list-many-files} function returns a numbers' |
| 15539 | list to which we can apply @code{max} (we could also apply @code{max} to |
| 15540 | the sorted numbers' list; it does not matter whether the list is |
| 15541 | sorted or not.) |
| 15542 | |
| 15543 | @need 800 |
| 15544 | Hence, the operation for finding the maximum height of the graph is this: |
| 15545 | |
| 15546 | @smallexample |
| 15547 | (setq max-graph-height (apply 'max numbers-list)) |
| 15548 | @end smallexample |
| 15549 | |
| 15550 | Now we can return to the question of how to create a list of strings |
| 15551 | for a column of the graph. Told the maximum height of the graph |
| 15552 | and the number of asterisks that should appear in the column, the |
| 15553 | function should return a list of strings for the |
| 15554 | @code{insert-rectangle} command to insert. |
| 15555 | |
| 15556 | Each column is made up of asterisks or blanks. Since the function is |
| 15557 | passed the value of the height of the column and the number of |
| 15558 | asterisks in the column, the number of blanks can be found by |
| 15559 | subtracting the number of asterisks from the height of the column. |
| 15560 | Given the number of blanks and the number of asterisks, two |
| 15561 | @code{while} loops can be used to construct the list: |
| 15562 | |
| 15563 | @smallexample |
| 15564 | @group |
| 15565 | ;;; @r{First version.} |
| 15566 | (defun column-of-graph (max-graph-height actual-height) |
| 15567 | "Return list of strings that is one column of a graph." |
| 15568 | (let ((insert-list nil) |
| 15569 | (number-of-top-blanks |
| 15570 | (- max-graph-height actual-height))) |
| 15571 | @end group |
| 15572 | |
| 15573 | @group |
| 15574 | ;; @r{Fill in asterisks.} |
| 15575 | (while (> actual-height 0) |
| 15576 | (setq insert-list (cons "*" insert-list)) |
| 15577 | (setq actual-height (1- actual-height))) |
| 15578 | @end group |
| 15579 | |
| 15580 | @group |
| 15581 | ;; @r{Fill in blanks.} |
| 15582 | (while (> number-of-top-blanks 0) |
| 15583 | (setq insert-list (cons " " insert-list)) |
| 15584 | (setq number-of-top-blanks |
| 15585 | (1- number-of-top-blanks))) |
| 15586 | @end group |
| 15587 | |
| 15588 | @group |
| 15589 | ;; @r{Return whole list.} |
| 15590 | insert-list)) |
| 15591 | @end group |
| 15592 | @end smallexample |
| 15593 | |
| 15594 | If you install this function and then evaluate the following |
| 15595 | expression you will see that it returns the list as desired: |
| 15596 | |
| 15597 | @smallexample |
| 15598 | (column-of-graph 5 3) |
| 15599 | @end smallexample |
| 15600 | |
| 15601 | @need 800 |
| 15602 | @noindent |
| 15603 | returns |
| 15604 | |
| 15605 | @smallexample |
| 15606 | (" " " " "*" "*" "*") |
| 15607 | @end smallexample |
| 15608 | |
| 15609 | As written, @code{column-of-graph} contains a major flaw: the symbols |
| 15610 | used for the blank and for the marked entries in the column are |
| 15611 | `hard-coded' as a space and asterisk. This is fine for a prototype, |
| 15612 | but you, or another user, may wish to use other symbols. For example, |
| 15613 | in testing the graph function, you many want to use a period in place |
| 15614 | of the space, to make sure the point is being repositioned properly |
| 15615 | each time the @code{insert-rectangle} function is called; or you might |
| 15616 | want to substitute a @samp{+} sign or other symbol for the asterisk. |
| 15617 | You might even want to make a graph-column that is more than one |
| 15618 | display column wide. The program should be more flexible. The way to |
| 15619 | do that is to replace the blank and the asterisk with two variables |
| 15620 | that we can call @code{graph-blank} and @code{graph-symbol} and define |
| 15621 | those variables separately. |
| 15622 | |
| 15623 | Also, the documentation is not well written. These considerations |
| 15624 | lead us to the second version of the function: |
| 15625 | |
| 15626 | @smallexample |
| 15627 | @group |
| 15628 | (defvar graph-symbol "*" |
| 15629 | "String used as symbol in graph, usually an asterisk.") |
| 15630 | @end group |
| 15631 | |
| 15632 | @group |
| 15633 | (defvar graph-blank " " |
| 15634 | "String used as blank in graph, usually a blank space. |
| 15635 | graph-blank must be the same number of columns wide |
| 15636 | as graph-symbol.") |
| 15637 | @end group |
| 15638 | @end smallexample |
| 15639 | |
| 15640 | @noindent |
| 15641 | (For an explanation of @code{defvar}, see |
| 15642 | @ref{defvar, , Initializing a Variable with @code{defvar}}.) |
| 15643 | |
| 15644 | @smallexample |
| 15645 | @group |
| 15646 | ;;; @r{Second version.} |
| 15647 | (defun column-of-graph (max-graph-height actual-height) |
| 15648 | "Return MAX-GRAPH-HEIGHT strings; ACTUAL-HEIGHT are graph-symbols. |
| 15649 | |
| 15650 | @end group |
| 15651 | @group |
| 15652 | The graph-symbols are contiguous entries at the end |
| 15653 | of the list. |
| 15654 | The list will be inserted as one column of a graph. |
| 15655 | The strings are either graph-blank or graph-symbol." |
| 15656 | @end group |
| 15657 | |
| 15658 | @group |
| 15659 | (let ((insert-list nil) |
| 15660 | (number-of-top-blanks |
| 15661 | (- max-graph-height actual-height))) |
| 15662 | @end group |
| 15663 | |
| 15664 | @group |
| 15665 | ;; @r{Fill in @code{graph-symbols}.} |
| 15666 | (while (> actual-height 0) |
| 15667 | (setq insert-list (cons graph-symbol insert-list)) |
| 15668 | (setq actual-height (1- actual-height))) |
| 15669 | @end group |
| 15670 | |
| 15671 | @group |
| 15672 | ;; @r{Fill in @code{graph-blanks}.} |
| 15673 | (while (> number-of-top-blanks 0) |
| 15674 | (setq insert-list (cons graph-blank insert-list)) |
| 15675 | (setq number-of-top-blanks |
| 15676 | (1- number-of-top-blanks))) |
| 15677 | |
| 15678 | ;; @r{Return whole list.} |
| 15679 | insert-list)) |
| 15680 | @end group |
| 15681 | @end smallexample |
| 15682 | |
| 15683 | If we wished, we could rewrite @code{column-of-graph} a third time to |
| 15684 | provide optionally for a line graph as well as for a bar graph. This |
| 15685 | would not be hard to do. One way to think of a line graph is that it |
| 15686 | is no more than a bar graph in which the part of each bar that is |
| 15687 | below the top is blank. To construct a column for a line graph, the |
| 15688 | function first constructs a list of blanks that is one shorter than |
| 15689 | the value, then it uses @code{cons} to attach a graph symbol to the |
| 15690 | list; then it uses @code{cons} again to attach the `top blanks' to |
| 15691 | the list. |
| 15692 | |
| 15693 | It is easy to see how to write such a function, but since we don't |
| 15694 | need it, we will not do it. But the job could be done, and if it were |
| 15695 | done, it would be done with @code{column-of-graph}. Even more |
| 15696 | important, it is worth noting that few changes would have to be made |
| 15697 | anywhere else. The enhancement, if we ever wish to make it, is |
| 15698 | simple. |
| 15699 | |
| 15700 | Now, finally, we come to our first actual graph printing function. |
| 15701 | This prints the body of a graph, not the labels for the vertical and |
| 15702 | horizontal axes, so we can call this @code{graph-body-print}. |
| 15703 | |
| 15704 | @node graph-body-print, recursive-graph-body-print, Columns of a graph, Readying a Graph |
| 15705 | @section The @code{graph-body-print} Function |
| 15706 | @findex graph-body-print |
| 15707 | |
| 15708 | After our preparation in the preceding section, the |
| 15709 | @code{graph-body-print} function is straightforward. The function |
| 15710 | will print column after column of asterisks and blanks, using the |
| 15711 | elements of a numbers' list to specify the number of asterisks in each |
| 15712 | column. This is a repetitive act, which means we can use a |
| 15713 | decrementing @code{while} loop or recursive function for the job. In |
| 15714 | this section, we will write the definition using a @code{while} loop. |
| 15715 | |
| 15716 | The @code{column-of-graph} function requires the height of the graph |
| 15717 | as an argument, so we should determine and record that as a local variable. |
| 15718 | |
| 15719 | This leads us to the following template for the @code{while} loop |
| 15720 | version of this function: |
| 15721 | |
| 15722 | @smallexample |
| 15723 | @group |
| 15724 | (defun graph-body-print (numbers-list) |
| 15725 | "@var{documentation}@dots{}" |
| 15726 | (let ((height @dots{} |
| 15727 | @dots{})) |
| 15728 | @end group |
| 15729 | |
| 15730 | @group |
| 15731 | (while numbers-list |
| 15732 | @var{insert-columns-and-reposition-point} |
| 15733 | (setq numbers-list (cdr numbers-list))))) |
| 15734 | @end group |
| 15735 | @end smallexample |
| 15736 | |
| 15737 | @noindent |
| 15738 | We need to fill in the slots of the template. |
| 15739 | |
| 15740 | Clearly, we can use the @code{(apply 'max numbers-list)} expression to |
| 15741 | determine the height of the graph. |
| 15742 | |
| 15743 | The @code{while} loop will cycle through the @code{numbers-list} one |
| 15744 | element at a time. As it is shortened by the @code{(setq numbers-list |
| 15745 | (cdr numbers-list))} expression, the @sc{car} of each instance of the |
| 15746 | list is the value of the argument for @code{column-of-graph}. |
| 15747 | |
| 15748 | At each cycle of the @code{while} loop, the @code{insert-rectangle} |
| 15749 | function inserts the list returned by @code{column-of-graph}. Since |
| 15750 | the @code{insert-rectangle} function moves point to the lower right of |
| 15751 | the inserted rectangle, we need to save the location of point at the |
| 15752 | time the rectangle is inserted, move back to that position after the |
| 15753 | rectangle is inserted, and then move horizontally to the next place |
| 15754 | from which @code{insert-rectangle} is called. |
| 15755 | |
| 15756 | If the inserted columns are one character wide, as they will be if |
| 15757 | single blanks and asterisks are used, the repositioning command is |
| 15758 | simply @code{(forward-char 1)}; however, the width of a column may be |
| 15759 | greater than one. This means that the repositioning command should be |
| 15760 | written @code{(forward-char symbol-width)}. The @code{symbol-width} |
| 15761 | itself is the length of a @code{graph-blank} and can be found using |
| 15762 | the expression @code{(length graph-blank)}. The best place to bind |
| 15763 | the @code{symbol-width} variable to the value of the width of graph |
| 15764 | column is in the varlist of the @code{let} expression. |
| 15765 | |
| 15766 | @need 1250 |
| 15767 | These considerations lead to the following function definition: |
| 15768 | |
| 15769 | @smallexample |
| 15770 | @group |
| 15771 | (defun graph-body-print (numbers-list) |
| 15772 | "Print a bar graph of the NUMBERS-LIST. |
| 15773 | The numbers-list consists of the Y-axis values." |
| 15774 | |
| 15775 | (let ((height (apply 'max numbers-list)) |
| 15776 | (symbol-width (length graph-blank)) |
| 15777 | from-position) |
| 15778 | @end group |
| 15779 | |
| 15780 | @group |
| 15781 | (while numbers-list |
| 15782 | (setq from-position (point)) |
| 15783 | (insert-rectangle |
| 15784 | (column-of-graph height (car numbers-list))) |
| 15785 | (goto-char from-position) |
| 15786 | (forward-char symbol-width) |
| 15787 | @end group |
| 15788 | @group |
| 15789 | ;; @r{Draw graph column by column.} |
| 15790 | (sit-for 0) |
| 15791 | (setq numbers-list (cdr numbers-list))) |
| 15792 | @end group |
| 15793 | @group |
| 15794 | ;; @r{Place point for X axis labels.} |
| 15795 | (forward-line height) |
| 15796 | (insert "\n") |
| 15797 | )) |
| 15798 | @end group |
| 15799 | @end smallexample |
| 15800 | |
| 15801 | @noindent |
| 15802 | The one unexpected expression in this function is the |
| 15803 | @w{@code{(sit-for 0)}} expression in the @code{while} loop. This |
| 15804 | expression makes the graph printing operation more interesting to |
| 15805 | watch than it would be otherwise. The expression causes Emacs to |
| 15806 | `sit' or do nothing for a zero length of time and then redraw the |
| 15807 | screen. Placed here, it causes Emacs to redraw the screen column by |
| 15808 | column. Without it, Emacs would not redraw the screen until the |
| 15809 | function exits. |
| 15810 | |
| 15811 | We can test @code{graph-body-print} with a short list of numbers. |
| 15812 | |
| 15813 | @enumerate |
| 15814 | @item |
| 15815 | Install @code{graph-symbol}, @code{graph-blank}, |
| 15816 | @code{column-of-graph}, which are in |
| 15817 | @iftex |
| 15818 | @ref{Readying a Graph, , Readying a Graph}, |
| 15819 | @end iftex |
| 15820 | @ifinfo |
| 15821 | @ref{Columns of a graph}, |
| 15822 | @end ifinfo |
| 15823 | and @code{graph-body-print}. |
| 15824 | |
| 15825 | @need 800 |
| 15826 | @item |
| 15827 | Copy the following expression: |
| 15828 | |
| 15829 | @smallexample |
| 15830 | (graph-body-print '(1 2 3 4 6 4 3 5 7 6 5 2 3)) |
| 15831 | @end smallexample |
| 15832 | |
| 15833 | @item |
| 15834 | Switch to the @file{*scratch*} buffer and place the cursor where you |
| 15835 | want the graph to start. |
| 15836 | |
| 15837 | @item |
| 15838 | Type @kbd{M-:} (@code{eval-expression}). |
| 15839 | |
| 15840 | @item |
| 15841 | Yank the @code{graph-body-print} expression into the minibuffer |
| 15842 | with @kbd{C-y} (@code{yank)}. |
| 15843 | |
| 15844 | @item |
| 15845 | Press @key{RET} to evaluate the @code{graph-body-print} expression. |
| 15846 | @end enumerate |
| 15847 | |
| 15848 | @need 800 |
| 15849 | Emacs will print a graph like this: |
| 15850 | |
| 15851 | @smallexample |
| 15852 | @group |
| 15853 | * |
| 15854 | * ** |
| 15855 | * **** |
| 15856 | *** **** |
| 15857 | ********* * |
| 15858 | ************ |
| 15859 | ************* |
| 15860 | @end group |
| 15861 | @end smallexample |
| 15862 | |
| 15863 | @node recursive-graph-body-print, Printed Axes, graph-body-print, Readying a Graph |
| 15864 | @section The @code{recursive-graph-body-print} Function |
| 15865 | @findex recursive-graph-body-print |
| 15866 | |
| 15867 | The @code{graph-body-print} function may also be written recursively. |
| 15868 | The recursive solution is divided into two parts: an outside `wrapper' |
| 15869 | that uses a @code{let} expression to determine the values of several |
| 15870 | variables that need only be found once, such as the maximum height of |
| 15871 | the graph, and an inside function that is called recursively to print |
| 15872 | the graph. |
| 15873 | |
| 15874 | @need 1250 |
| 15875 | The `wrapper' is uncomplicated: |
| 15876 | |
| 15877 | @smallexample |
| 15878 | @group |
| 15879 | (defun recursive-graph-body-print (numbers-list) |
| 15880 | "Print a bar graph of the NUMBERS-LIST. |
| 15881 | The numbers-list consists of the Y-axis values." |
| 15882 | (let ((height (apply 'max numbers-list)) |
| 15883 | (symbol-width (length graph-blank)) |
| 15884 | from-position) |
| 15885 | (recursive-graph-body-print-internal |
| 15886 | numbers-list |
| 15887 | height |
| 15888 | symbol-width))) |
| 15889 | @end group |
| 15890 | @end smallexample |
| 15891 | |
| 15892 | The recursive function is a little more difficult. It has four parts: |
| 15893 | the `do-again-test', the printing code, the recursive call, and the |
| 15894 | `next-step-expression'. The `do-again-test' is an @code{if} |
| 15895 | expression that determines whether the @code{numbers-list} contains |
| 15896 | any remaining elements; if it does, the function prints one column of |
| 15897 | the graph using the printing code and calls itself again. The |
| 15898 | function calls itself again according to the value produced by the |
| 15899 | `next-step-expression' which causes the call to act on a shorter |
| 15900 | version of the @code{numbers-list}. |
| 15901 | |
| 15902 | @smallexample |
| 15903 | @group |
| 15904 | (defun recursive-graph-body-print-internal |
| 15905 | (numbers-list height symbol-width) |
| 15906 | "Print a bar graph. |
| 15907 | Used within recursive-graph-body-print function." |
| 15908 | @end group |
| 15909 | |
| 15910 | @group |
| 15911 | (if numbers-list |
| 15912 | (progn |
| 15913 | (setq from-position (point)) |
| 15914 | (insert-rectangle |
| 15915 | (column-of-graph height (car numbers-list))) |
| 15916 | @end group |
| 15917 | @group |
| 15918 | (goto-char from-position) |
| 15919 | (forward-char symbol-width) |
| 15920 | (sit-for 0) ; @r{Draw graph column by column.} |
| 15921 | (recursive-graph-body-print-internal |
| 15922 | (cdr numbers-list) height symbol-width)))) |
| 15923 | @end group |
| 15924 | @end smallexample |
| 15925 | |
| 15926 | @need 1250 |
| 15927 | After installation, this expression can be tested; here is a sample: |
| 15928 | |
| 15929 | @smallexample |
| 15930 | (recursive-graph-body-print '(3 2 5 6 7 5 3 4 6 4 3 2 1)) |
| 15931 | @end smallexample |
| 15932 | |
| 15933 | @need 800 |
| 15934 | Here is what @code{recursive-graph-body-print} produces: |
| 15935 | |
| 15936 | @smallexample |
| 15937 | @group |
| 15938 | * |
| 15939 | ** * |
| 15940 | **** * |
| 15941 | **** *** |
| 15942 | * ********* |
| 15943 | ************ |
| 15944 | ************* |
| 15945 | @end group |
| 15946 | @end smallexample |
| 15947 | |
| 15948 | Either of these two functions, @code{graph-body-print} or |
| 15949 | @code{recursive-graph-body-print}, create the body of a graph. |
| 15950 | |
| 15951 | @node Printed Axes, Line Graph Exercise, recursive-graph-body-print, Readying a Graph |
| 15952 | @section Need for Printed Axes |
| 15953 | |
| 15954 | A graph needs printed axes, so you can orient yourself. For a do-once |
| 15955 | project, it may be reasonable to draw the axes by hand using Emacs' |
| 15956 | Picture mode; but a graph drawing function may be used more than once. |
| 15957 | |
| 15958 | For this reason, I have written enhancements to the basic |
| 15959 | @code{print-graph-body} function that automatically print labels for |
| 15960 | the horizontal and vertical axes. Since the label printing functions |
| 15961 | do not contain much new material, I have placed their description in |
| 15962 | an appendix. @xref{Full Graph, , A Graph with Labelled Axes}. |
| 15963 | |
| 15964 | @node Line Graph Exercise, , Printed Axes, Readying a Graph |
| 15965 | @section Exercise |
| 15966 | |
| 15967 | Write a line graph version of the graph printing functions. |
| 15968 | |
| 15969 | @node Emacs Initialization, Debugging, Readying a Graph, Top |
| 15970 | @chapter Your @file{.emacs} File |
| 15971 | @cindex @file{.emacs} file |
| 15972 | @cindex Customizing your @file{.emacs} file |
| 15973 | @cindex Initialization file |
| 15974 | |
| 15975 | ``You don't have to like Emacs to like it'' -- this seemingly |
| 15976 | paradoxical statement is the secret of GNU Emacs. The plain, `out of |
| 15977 | the box' Emacs is a generic tool. Most people who use it, customize |
| 15978 | it to suit themselves. |
| 15979 | |
| 15980 | GNU Emacs is mostly written in Emacs Lisp; this means that by writing |
| 15981 | expressions in Emacs Lisp you can change or extend Emacs. |
| 15982 | |
| 15983 | @menu |
| 15984 | * Default Configuration:: |
| 15985 | * Site-wide Init:: You can write site-wide init files. |
| 15986 | * defcustom:: Emacs will write code for you. |
| 15987 | * Beginning a .emacs File:: How to write a @code{.emacs file}. |
| 15988 | * Text and Auto-fill:: Automatically wrap lines. |
| 15989 | * Mail Aliases:: Use abbreviations for email addresses. |
| 15990 | * Indent Tabs Mode:: Don't use tabs with @TeX{} |
| 15991 | * Keybindings:: Create some personal keybindings. |
| 15992 | * Keymaps:: More about key binding. |
| 15993 | * Loading Files:: Load (i.e., evaluate) files automatically. |
| 15994 | * Autoload:: Make functions available. |
| 15995 | * Simple Extension:: Define a function; bind it to a key. |
| 15996 | * X11 Colors:: Colors in version 19 in X. |
| 15997 | * Miscellaneous:: |
| 15998 | * Mode Line:: How to customize your mode line. |
| 15999 | @end menu |
| 16000 | |
| 16001 | @node Default Configuration, Site-wide Init, Emacs Initialization, Emacs Initialization |
| 16002 | @ifnottex |
| 16003 | @unnumberedsec Emacs' Default Configuration |
| 16004 | @end ifnottex |
| 16005 | |
| 16006 | There are those who appreciate Emacs' default configuration. After |
| 16007 | all, Emacs starts you in C mode when you edit a C file, starts you in |
| 16008 | Fortran mode when you edit a Fortran file, and starts you in |
| 16009 | Fundamental mode when you edit an unadorned file. This all makes |
| 16010 | sense, if you do not know who is going to use Emacs. Who knows what a |
| 16011 | person hopes to do with an unadorned file? Fundamental mode is the |
| 16012 | right default for such a file, just as C mode is the right default for |
| 16013 | editing C code. But when you do know who is going to use Emacs---you, |
| 16014 | yourself---then it makes sense to customize Emacs. |
| 16015 | |
| 16016 | For example, I seldom want Fundamental mode when I edit an |
| 16017 | otherwise undistinguished file; I want Text mode. This is why I |
| 16018 | customize Emacs: so it suits me. |
| 16019 | |
| 16020 | You can customize and extend Emacs by writing or adapting a |
| 16021 | @file{~/.emacs} file. This is your personal initialization file; its |
| 16022 | contents, written in Emacs Lisp, tell Emacs what to do.@footnote{You |
| 16023 | may also add @file{.el} to @file{~/.emacs} and call it a |
| 16024 | @file{~/.emacs.el} file. In the past, you were forbidden to type the |
| 16025 | extra keystrokes that the name @file{~/.emacs.el} requires, but now |
| 16026 | you may. The new format is consistent with the Emacs Lisp file |
| 16027 | naming conventions; the old format saves typing.} |
| 16028 | |
| 16029 | A @file{~/.emacs} file contains Emacs Lisp code. You can write this |
| 16030 | code yourself; or you can use Emacs' @code{customize} feature to write |
| 16031 | the code for you. You can combine your own expressions and |
| 16032 | auto-written Customize expressions in your @file{.emacs} file. |
| 16033 | |
| 16034 | (I myself prefer to write my own expressions, except for those, |
| 16035 | particularly fonts, that I find easier to manipulate using the |
| 16036 | @code{customize} command. I combine the two methods.) |
| 16037 | |
| 16038 | Most of this chapter is about writing expressions yourself. It |
| 16039 | describes a simple @file{.emacs} file; for more information, see |
| 16040 | @ref{Init File, , The Init File, emacs, The GNU Emacs Manual}, and |
| 16041 | @ref{Init File, , The Init File, elisp, The GNU Emacs Lisp Reference |
| 16042 | Manual}. |
| 16043 | |
| 16044 | @node Site-wide Init, defcustom, Default Configuration, Emacs Initialization |
| 16045 | @section Site-wide Initialization Files |
| 16046 | |
| 16047 | @cindex @file{default.el} init file |
| 16048 | @cindex @file{site-init.el} init file |
| 16049 | @cindex @file{site-load.el} init file |
| 16050 | In addition to your personal initialization file, Emacs automatically |
| 16051 | loads various site-wide initialization files, if they exist. These |
| 16052 | have the same form as your @file{.emacs} file, but are loaded by |
| 16053 | everyone. |
| 16054 | |
| 16055 | Two site-wide initialization files, @file{site-load.el} and |
| 16056 | @file{site-init.el}, are loaded into Emacs and then `dumped' if a |
| 16057 | `dumped' version of Emacs is created, as is most common. (Dumped |
| 16058 | copies of Emacs load more quickly. However, once a file is loaded and |
| 16059 | dumped, a change to it does not lead to a change in Emacs unless you |
| 16060 | load it yourself or re-dump Emacs. @xref{Building Emacs, , Building |
| 16061 | Emacs, elisp, The GNU Emacs Lisp Reference Manual}, and the |
| 16062 | @file{INSTALL} file.) |
| 16063 | |
| 16064 | Three other site-wide initialization files are loaded automatically |
| 16065 | each time you start Emacs, if they exist. These are |
| 16066 | @file{site-start.el}, which is loaded @emph{before} your @file{.emacs} |
| 16067 | file, and @file{default.el}, and the terminal type file, which are both |
| 16068 | loaded @emph{after} your @file{.emacs} file. |
| 16069 | |
| 16070 | Settings and definitions in your @file{.emacs} file will overwrite |
| 16071 | conflicting settings and definitions in a @file{site-start.el} file, |
| 16072 | if it exists; but the settings and definitions in a @file{default.el} |
| 16073 | or terminal type file will overwrite those in your @file{.emacs} file. |
| 16074 | (You can prevent interference from a terminal type file by setting |
| 16075 | @code{term-file-prefix} to @code{nil}. @xref{Simple Extension, , A |
| 16076 | Simple Extension}.) |
| 16077 | |
| 16078 | @c Rewritten to avoid overfull hbox. |
| 16079 | The @file{INSTALL} file that comes in the distribution contains |
| 16080 | descriptions of the @file{site-init.el} and @file{site-load.el} files. |
| 16081 | |
| 16082 | The @file{loadup.el}, @file{startup.el}, and @file{loaddefs.el} files |
| 16083 | control loading. These files are in the @file{lisp} directory of the |
| 16084 | Emacs distribution and are worth perusing. |
| 16085 | |
| 16086 | The @file{loaddefs.el} file contains a good many suggestions as to |
| 16087 | what to put into your own @file{.emacs} file, or into a site-wide |
| 16088 | initialization file. |
| 16089 | |
| 16090 | @node defcustom, Beginning a .emacs File, Site-wide Init, Emacs Initialization |
| 16091 | @section Specifying Variables using @code{defcustom} |
| 16092 | @findex defcustom |
| 16093 | |
| 16094 | You can specify variables using @code{defcustom} so that you and |
| 16095 | others can then use Emacs' @code{customize} feature to set their |
| 16096 | values. (You cannot use @code{customize} to write function |
| 16097 | definitions; but you can write @code{defuns} in your @file{.emacs} |
| 16098 | file. Indeed, you can write any Lisp expression in your @file{.emacs} |
| 16099 | file.) |
| 16100 | |
| 16101 | The @code{customize} feature depends on the @code{defcustom} special |
| 16102 | form. Although you can use @code{defvar} or @code{setq} for variables |
| 16103 | that users set, the @code{defcustom} special form is designed for the |
| 16104 | job. |
| 16105 | |
| 16106 | You can use your knowledge of @code{defvar} for writing the |
| 16107 | first three arguments for @code{defcustom}. The first argument to |
| 16108 | @code{defcustom} is the name of the variable. The second argument is |
| 16109 | the variable's initial value, if any; and this value is set only if |
| 16110 | the value has not already been set. The third argument is the |
| 16111 | documentation. |
| 16112 | |
| 16113 | The fourth and subsequent arguments to @code{defcustom} specify types |
| 16114 | and options; these are not featured in @code{defvar}. (These |
| 16115 | arguments are optional.) |
| 16116 | |
| 16117 | Each of these arguments consists of a keyword followed by a value. |
| 16118 | Each keyword starts with the colon character @samp{:}. |
| 16119 | |
| 16120 | @need 1250 |
| 16121 | For example, the customizable user option variable |
| 16122 | @code{text-mode-hook} looks like this: |
| 16123 | |
| 16124 | @smallexample |
| 16125 | @group |
| 16126 | (defcustom text-mode-hook nil |
| 16127 | "Normal hook run when entering Text mode and many related modes." |
| 16128 | :type 'hook |
| 16129 | :options '(turn-on-auto-fill flyspell-mode) |
| 16130 | :group 'data) |
| 16131 | @end group |
| 16132 | @end smallexample |
| 16133 | |
| 16134 | @noindent |
| 16135 | The name of the variable is @code{text-mode-hook}; it has no default |
| 16136 | value; and its documentation string tells you what it does. |
| 16137 | |
| 16138 | The @code{:type} keyword tells Emacs the kind of data to which |
| 16139 | @code{text-mode-hook} should be set and how to display the value in a |
| 16140 | Customization buffer. |
| 16141 | |
| 16142 | The @code{:options} keyword specifies a suggested list of values for |
| 16143 | the variable. Currently, you can use @code{:options} only for a hook. |
| 16144 | The list is only a suggestion; it is not exclusive; a person who sets |
| 16145 | the variable may set it to other values; the list shown following the |
| 16146 | @code{:options} keyword is intended to offer convenient choices to a |
| 16147 | user. |
| 16148 | |
| 16149 | Finally, the @code{:group} keyword tells the Emacs Customization |
| 16150 | command in which group the variable is located. This tells where to |
| 16151 | find it. |
| 16152 | |
| 16153 | For more information, see @ref{Customization, , Writing Customization |
| 16154 | Definitions, elisp, The GNU Emacs Lisp Reference Manual}. |
| 16155 | |
| 16156 | Consider @code{text-mode-hook} as an example. |
| 16157 | |
| 16158 | There are two ways to customize this variable. You can use the |
| 16159 | customization command or write the appropriate expressions yourself. |
| 16160 | |
| 16161 | @need 800 |
| 16162 | Using the customization command, you can type: |
| 16163 | |
| 16164 | @smallexample |
| 16165 | M-x customize |
| 16166 | @end smallexample |
| 16167 | |
| 16168 | @noindent |
| 16169 | and find that the group for editing files of data is called `data'. |
| 16170 | Enter that group. Text Mode Hook is the first member. You can click |
| 16171 | on its various options to set the values. After you click on the |
| 16172 | button to |
| 16173 | |
| 16174 | @smallexample |
| 16175 | Save for Future Sessions |
| 16176 | @end smallexample |
| 16177 | |
| 16178 | @noindent |
| 16179 | Emacs will write an expression into your @file{.emacs} file. |
| 16180 | It will look like this: |
| 16181 | |
| 16182 | @smallexample |
| 16183 | @group |
| 16184 | (custom-set-variables |
| 16185 | ;; custom-set-variables was added by Custom -- |
| 16186 | ;; don't edit or cut/paste it! |
| 16187 | ;; Your init file should contain only one such instance. |
| 16188 | '(text-mode-hook (quote (turn-on-auto-fill text-mode-hook-identify)))) |
| 16189 | @end group |
| 16190 | @end smallexample |
| 16191 | |
| 16192 | @noindent |
| 16193 | (The @code{text-mode-hook-identify} function tells |
| 16194 | @code{toggle-text-mode-auto-fill} which buffers are in Text mode.) |
| 16195 | |
| 16196 | In spite of the warning, you certainly may edit, cut, and paste the |
| 16197 | expression! I do all time. The purpose of the warning is to scare |
| 16198 | those who do not know what they are doing, so they do not |
| 16199 | inadvertently generate an error. |
| 16200 | |
| 16201 | The @code{custom-set-variables} function works somewhat differently |
| 16202 | than a @code{setq}. While I have never learned the differences, I do |
| 16203 | modify the @code{custom-set-variables} expressions in my @file{.emacs} |
| 16204 | file by hand: I make the changes in what appears to me to be a |
| 16205 | reasonable manner and have not had any problems. Others prefer to use |
| 16206 | the Customization command and let Emacs do the work for them. |
| 16207 | |
| 16208 | Another @code{custom-set-@dots{}} function is @code{custom-set-faces}. |
| 16209 | This function sets the various font faces. Over time, I have set a |
| 16210 | considerable number of faces. Some of the time, I re-set them using |
| 16211 | @code{customize}; other times, I simply edit the |
| 16212 | @code{custom-set-faces} expression in my @file{.emacs} file itself. |
| 16213 | |
| 16214 | The second way to customize your @code{text-mode-hook} is to set it |
| 16215 | yourself in your @file{.emacs} file using code that has nothing to do |
| 16216 | with the @code{custom-set-@dots{}} functions. |
| 16217 | |
| 16218 | @need 800 |
| 16219 | When you do this, and later use @code{customize}, you will see a |
| 16220 | message that says |
| 16221 | |
| 16222 | @smallexample |
| 16223 | this option has been changed outside the customize buffer. |
| 16224 | @end smallexample |
| 16225 | |
| 16226 | @need 800 |
| 16227 | This message is only a warning. If you click on the button to |
| 16228 | |
| 16229 | @smallexample |
| 16230 | Save for Future Sessions |
| 16231 | @end smallexample |
| 16232 | |
| 16233 | @noindent |
| 16234 | Emacs will write a @code{custom-set-@dots{}} expression near the end |
| 16235 | of your @file{.emacs} file that will be evaluated after your |
| 16236 | hand-written expression. It will, therefore, overrule your |
| 16237 | hand-written expression. No harm will be done. When you do this, |
| 16238 | however, be careful to remember which expression is active; if you |
| 16239 | forget, you may confuse yourself. |
| 16240 | |
| 16241 | So long as you remember where the values are set, you will have no |
| 16242 | trouble. In any event, the values are always set in your |
| 16243 | initialization file, which is usually called @file{.emacs}. |
| 16244 | |
| 16245 | I myself use @code{customize} for hardly anything. Mostly, I write |
| 16246 | expressions myself. |
| 16247 | |
| 16248 | @node Beginning a .emacs File, Text and Auto-fill, defcustom, Emacs Initialization |
| 16249 | @section Beginning a @file{.emacs} File |
| 16250 | @cindex @file{.emacs} file, beginning of |
| 16251 | |
| 16252 | When you start Emacs, it loads your @file{.emacs} file unless you tell |
| 16253 | it not to by specifying @samp{-q} on the command line. (The |
| 16254 | @code{emacs -q} command gives you a plain, out-of-the-box Emacs.) |
| 16255 | |
| 16256 | A @file{.emacs} file contains Lisp expressions. Often, these are no |
| 16257 | more than expressions to set values; sometimes they are function |
| 16258 | definitions. |
| 16259 | |
| 16260 | @xref{Init File, , The Init File @file{~/.emacs}, emacs, The GNU Emacs |
| 16261 | Manual}, for a short description of initialization files. |
| 16262 | |
| 16263 | This chapter goes over some of the same ground, but is a walk among |
| 16264 | extracts from a complete, long-used @file{.emacs} file---my own. |
| 16265 | |
| 16266 | The first part of the file consists of comments: reminders to myself. |
| 16267 | By now, of course, I remember these things, but when I started, I did |
| 16268 | not. |
| 16269 | |
| 16270 | @need 1200 |
| 16271 | @smallexample |
| 16272 | @group |
| 16273 | ;;;; Bob's .emacs file |
| 16274 | ; Robert J. Chassell |
| 16275 | ; 26 September 1985 |
| 16276 | @end group |
| 16277 | @end smallexample |
| 16278 | |
| 16279 | @noindent |
| 16280 | Look at that date! I started this file a long time ago. I have been |
| 16281 | adding to it ever since. |
| 16282 | |
| 16283 | @smallexample |
| 16284 | @group |
| 16285 | ; Each section in this file is introduced by a |
| 16286 | ; line beginning with four semicolons; and each |
| 16287 | ; entry is introduced by a line beginning with |
| 16288 | ; three semicolons. |
| 16289 | @end group |
| 16290 | @end smallexample |
| 16291 | |
| 16292 | @noindent |
| 16293 | This describes the usual conventions for comments in Emacs Lisp. |
| 16294 | Everything on a line that follows a semicolon is a comment. Two, |
| 16295 | three, and four semicolons are used as section and subsection |
| 16296 | markers. (@xref{Comments, ,, elisp, The GNU Emacs Lisp Reference |
| 16297 | Manual}, for more about comments.) |
| 16298 | |
| 16299 | @smallexample |
| 16300 | @group |
| 16301 | ;;;; The Help Key |
| 16302 | ; Control-h is the help key; |
| 16303 | ; after typing control-h, type a letter to |
| 16304 | ; indicate the subject about which you want help. |
| 16305 | ; For an explanation of the help facility, |
| 16306 | ; type control-h two times in a row. |
| 16307 | @end group |
| 16308 | @end smallexample |
| 16309 | |
| 16310 | @noindent |
| 16311 | Just remember: type @kbd{C-h} two times for help. |
| 16312 | |
| 16313 | @smallexample |
| 16314 | @group |
| 16315 | ; To find out about any mode, type control-h m |
| 16316 | ; while in that mode. For example, to find out |
| 16317 | ; about mail mode, enter mail mode and then type |
| 16318 | ; control-h m. |
| 16319 | @end group |
| 16320 | @end smallexample |
| 16321 | |
| 16322 | @noindent |
| 16323 | `Mode help', as I call this, is very helpful. Usually, it tells you |
| 16324 | all you need to know. |
| 16325 | |
| 16326 | Of course, you don't need to include comments like these in your |
| 16327 | @file{.emacs} file. I included them in mine because I kept forgetting |
| 16328 | about Mode help or the conventions for comments---but I was able to |
| 16329 | remember to look here to remind myself. |
| 16330 | |
| 16331 | @node Text and Auto-fill, Mail Aliases, Beginning a .emacs File, Emacs Initialization |
| 16332 | @section Text and Auto Fill Mode |
| 16333 | |
| 16334 | Now we come to the part that `turns on' Text mode and |
| 16335 | Auto Fill mode. |
| 16336 | |
| 16337 | @smallexample |
| 16338 | @group |
| 16339 | ;;; Text mode and Auto Fill mode |
| 16340 | ; The next three lines put Emacs into Text mode |
| 16341 | ; and Auto Fill mode, and are for writers who |
| 16342 | ; want to start writing prose rather than code. |
| 16343 | |
| 16344 | (setq default-major-mode 'text-mode) |
| 16345 | (add-hook 'text-mode-hook 'text-mode-hook-identify) |
| 16346 | (add-hook 'text-mode-hook 'turn-on-auto-fill) |
| 16347 | @end group |
| 16348 | @end smallexample |
| 16349 | |
| 16350 | Here is the first part of this @file{.emacs} file that does something |
| 16351 | besides remind a forgetful human! |
| 16352 | |
| 16353 | The first of the two lines in parentheses tells Emacs to turn on Text |
| 16354 | mode when you find a file, @emph{unless} that file should go into some |
| 16355 | other mode, such as C mode. |
| 16356 | |
| 16357 | @cindex Per-buffer, local variables list |
| 16358 | @cindex Local variables list, per-buffer, |
| 16359 | @cindex Automatic mode selection |
| 16360 | @cindex Mode selection, automatic |
| 16361 | When Emacs reads a file, it looks at the extension to the file name, |
| 16362 | if any. (The extension is the part that comes after a @samp{.}.) If |
| 16363 | the file ends with a @samp{.c} or @samp{.h} extension then Emacs turns |
| 16364 | on C mode. Also, Emacs looks at first nonblank line of the file; if |
| 16365 | the line says @w{@samp{-*- C -*-}}, Emacs turns on C mode. Emacs |
| 16366 | possesses a list of extensions and specifications that it uses |
| 16367 | automatically. In addition, Emacs looks near the last page for a |
| 16368 | per-buffer, ``local variables list'', if any. |
| 16369 | |
| 16370 | @ifinfo |
| 16371 | @xref{Choosing Modes, , How Major Modes are Chosen, emacs, The GNU |
| 16372 | Emacs Manual}. |
| 16373 | |
| 16374 | @xref{File Variables, , Local Variables in Files, emacs, The GNU Emacs |
| 16375 | Manual}. |
| 16376 | @end ifinfo |
| 16377 | @iftex |
| 16378 | See sections ``How Major Modes are Chosen'' and ``Local Variables in |
| 16379 | Files'' in @cite{The GNU Emacs Manual}. |
| 16380 | @end iftex |
| 16381 | |
| 16382 | Now, back to the @file{.emacs} file. |
| 16383 | |
| 16384 | @need 800 |
| 16385 | Here is the line again; how does it work? |
| 16386 | |
| 16387 | @cindex Text Mode turned on |
| 16388 | @smallexample |
| 16389 | (setq default-major-mode 'text-mode) |
| 16390 | @end smallexample |
| 16391 | |
| 16392 | @noindent |
| 16393 | This line is a short, but complete Emacs Lisp expression. |
| 16394 | |
| 16395 | We are already familiar with @code{setq}. It sets the following variable, |
| 16396 | @code{default-major-mode}, to the subsequent value, which is |
| 16397 | @code{text-mode}. The single quote mark before @code{text-mode} tells |
| 16398 | Emacs to deal directly with the @code{text-mode} variable, not with |
| 16399 | whatever it might stand for. @xref{set & setq, , Setting the Value of |
| 16400 | a Variable}, for a reminder of how @code{setq} works. The main point |
| 16401 | is that there is no difference between the procedure you use to set |
| 16402 | a value in your @file{.emacs} file and the procedure you use anywhere |
| 16403 | else in Emacs. |
| 16404 | |
| 16405 | @need 800 |
| 16406 | Here are the next two lines: |
| 16407 | |
| 16408 | @cindex Auto Fill mode turned on |
| 16409 | @findex add-hook |
| 16410 | @smallexample |
| 16411 | (add-hook 'text-mode-hook 'text-mode-hook-identify) |
| 16412 | (add-hook 'text-mode-hook 'turn-on-auto-fill) |
| 16413 | @end smallexample |
| 16414 | |
| 16415 | @noindent |
| 16416 | In these two lines, the @code{add-hook} command first adds |
| 16417 | @code{text-mode-hook-identify} to the variable called |
| 16418 | @code{text-mode-hook} and then adds @code{turn-on-auto-fill} to the |
| 16419 | variable. |
| 16420 | |
| 16421 | @code{turn-on-auto-fill} is the name of a program, that, you guessed |
| 16422 | it!, turns on Auto Fill mode. @code{text-mode-hook-identify} is a |
| 16423 | function that tells @code{toggle-text-mode-auto-fill} which buffers |
| 16424 | are in Text mode. |
| 16425 | |
| 16426 | Every time Emacs turns on Text mode, Emacs runs the commands `hooked' |
| 16427 | onto Text mode. So every time Emacs turns on Text mode, Emacs also |
| 16428 | turns on Auto Fill mode. |
| 16429 | |
| 16430 | In brief, the first line causes Emacs to enter Text mode when you edit |
| 16431 | a file, unless the file name extension, first non-blank line, or local |
| 16432 | variables tell Emacs otherwise. |
| 16433 | |
| 16434 | Text mode among other actions, sets the syntax table to work |
| 16435 | conveniently for writers. In Text mode, Emacs considers an apostrophe |
| 16436 | as part of a word like a letter; but Emacs does not consider a period |
| 16437 | or a space as part of a word. Thus, @kbd{M-f} moves you over |
| 16438 | @samp{it's}. On the other hand, in C mode, @kbd{M-f} stops just after |
| 16439 | the @samp{t} of @samp{it's}. |
| 16440 | |
| 16441 | The second and third lines causes Emacs to turn on Auto Fill mode when |
| 16442 | it turns on Text mode. In Auto Fill mode, Emacs automatically breaks |
| 16443 | a line that is too wide and brings the excessively wide part of the |
| 16444 | line down to the next line. Emacs breaks lines between words, not |
| 16445 | within them. |
| 16446 | |
| 16447 | When Auto Fill mode is turned off, lines continue to the right as you |
| 16448 | type them. Depending on how you set the value of |
| 16449 | @code{truncate-lines}, the words you type either disappear off the |
| 16450 | right side of the screen, or else are shown, in a rather ugly and |
| 16451 | unreadable manner, as a continuation line on the screen. |
| 16452 | |
| 16453 | @need 1250 |
| 16454 | In addition, in this part of my @file{.emacs} file, I tell the Emacs |
| 16455 | fill commands to insert two spaces after a colon: |
| 16456 | |
| 16457 | @smallexample |
| 16458 | (setq colon-double-space t) |
| 16459 | @end smallexample |
| 16460 | |
| 16461 | @node Mail Aliases, Indent Tabs Mode, Text and Auto-fill, Emacs Initialization |
| 16462 | @section Mail Aliases |
| 16463 | |
| 16464 | Here is a @code{setq} that `turns on' mail aliases, along with more |
| 16465 | reminders. |
| 16466 | |
| 16467 | @smallexample |
| 16468 | @group |
| 16469 | ;;; Mail mode |
| 16470 | ; To enter mail mode, type `C-x m' |
| 16471 | ; To enter RMAIL (for reading mail), |
| 16472 | ; type `M-x rmail' |
| 16473 | |
| 16474 | (setq mail-aliases t) |
| 16475 | @end group |
| 16476 | @end smallexample |
| 16477 | |
| 16478 | @cindex Mail aliases |
| 16479 | @noindent |
| 16480 | This @code{setq} command sets the value of the variable |
| 16481 | @code{mail-aliases} to @code{t}. Since @code{t} means true, the line |
| 16482 | says, in effect, ``Yes, use mail aliases.'' |
| 16483 | |
| 16484 | Mail aliases are convenient short names for long email addresses or |
| 16485 | for lists of email addresses. The file where you keep your `aliases' |
| 16486 | is @file{~/.mailrc}. You write an alias like this: |
| 16487 | |
| 16488 | @smallexample |
| 16489 | alias geo george@@foobar.wiz.edu |
| 16490 | @end smallexample |
| 16491 | |
| 16492 | @noindent |
| 16493 | When you write a message to George, address it to @samp{geo}; the |
| 16494 | mailer will automatically expand @samp{geo} to the full address. |
| 16495 | |
| 16496 | @node Indent Tabs Mode, Keybindings, Mail Aliases, Emacs Initialization |
| 16497 | @section Indent Tabs Mode |
| 16498 | @cindex Tabs, preventing |
| 16499 | @findex indent-tabs-mode |
| 16500 | |
| 16501 | By default, Emacs inserts tabs in place of multiple spaces when it |
| 16502 | formats a region. (For example, you might indent many lines of text |
| 16503 | all at once with the @code{indent-region} command.) Tabs look fine on |
| 16504 | a terminal or with ordinary printing, but they produce badly indented |
| 16505 | output when you use @TeX{} or Texinfo since @TeX{} ignores tabs. |
| 16506 | |
| 16507 | @need 1250 |
| 16508 | The following turns off Indent Tabs mode: |
| 16509 | |
| 16510 | @smallexample |
| 16511 | @group |
| 16512 | ;;; Prevent Extraneous Tabs |
| 16513 | (setq-default indent-tabs-mode nil) |
| 16514 | @end group |
| 16515 | @end smallexample |
| 16516 | |
| 16517 | Note that this line uses @code{setq-default} rather than the |
| 16518 | @code{setq} command that we have seen before. The @code{setq-default} |
| 16519 | command sets values only in buffers that do not have their own local |
| 16520 | values for the variable. |
| 16521 | |
| 16522 | @ifinfo |
| 16523 | @xref{Just Spaces, , Tabs vs. Spaces, emacs, The GNU Emacs Manual}. |
| 16524 | |
| 16525 | @xref{File Variables, , Local Variables in Files, emacs, The GNU Emacs |
| 16526 | Manual}. |
| 16527 | @end ifinfo |
| 16528 | @iftex |
| 16529 | See sections ``Tabs vs.@: Spaces'' and ``Local Variables in |
| 16530 | Files'' in @cite{The GNU Emacs Manual}. |
| 16531 | @end iftex |
| 16532 | |
| 16533 | @node Keybindings, Keymaps, Indent Tabs Mode, Emacs Initialization |
| 16534 | @section Some Keybindings |
| 16535 | |
| 16536 | Now for some personal keybindings: |
| 16537 | |
| 16538 | @smallexample |
| 16539 | @group |
| 16540 | ;;; Compare windows |
| 16541 | (global-set-key "\C-cw" 'compare-windows) |
| 16542 | @end group |
| 16543 | @end smallexample |
| 16544 | |
| 16545 | @findex compare-windows |
| 16546 | @code{compare-windows} is a nifty command that compares the text in |
| 16547 | your current window with text in the next window. It makes the |
| 16548 | comparison by starting at point in each window, moving over text in |
| 16549 | each window as far as they match. I use this command all the time. |
| 16550 | |
| 16551 | This also shows how to set a key globally, for all modes. |
| 16552 | |
| 16553 | @cindex Setting a key globally |
| 16554 | @cindex Global set key |
| 16555 | @cindex Key setting globally |
| 16556 | @findex global-set-key |
| 16557 | The command is @code{global-set-key}. It is followed by the |
| 16558 | keybinding. In a @file{.emacs} file, the keybinding is written as |
| 16559 | shown: @code{\C-c} stands for `control-c', which means `press the |
| 16560 | control key and the @kbd{c} key at the same time'. The @code{w} means |
| 16561 | `press the @kbd{w} key'. The keybinding is surrounded by double |
| 16562 | quotation marks. In documentation, you would write this as @kbd{C-c |
| 16563 | w}. (If you were binding a @key{META} key, such as @kbd{M-c}, rather |
| 16564 | than a @key{CTL} key, you would write @code{\M-c}. @xref{Init |
| 16565 | Rebinding, , Rebinding Keys in Your Init File, emacs, The GNU Emacs |
| 16566 | Manual}, for details.) |
| 16567 | |
| 16568 | The command invoked by the keys is @code{compare-windows}. Note that |
| 16569 | @code{compare-windows} is preceded by a single quote; otherwise, Emacs |
| 16570 | would first try to evaluate the symbol to determine its value. |
| 16571 | |
| 16572 | These three things, the double quotation marks, the backslash before |
| 16573 | the @samp{C}, and the single quote mark are necessary parts of |
| 16574 | keybinding that I tend to forget. Fortunately, I have come to |
| 16575 | remember that I should look at my existing @file{.emacs} file, and |
| 16576 | adapt what is there. |
| 16577 | |
| 16578 | As for the keybinding itself: @kbd{C-c w}. This combines the prefix |
| 16579 | key, @kbd{C-c}, with a single character, in this case, @kbd{w}. This |
| 16580 | set of keys, @kbd{C-c} followed by a single character, is strictly |
| 16581 | reserved for individuals' own use. (I call these `own' keys, since |
| 16582 | these are for my own use.) You should always be able to create such a |
| 16583 | keybinding for your own use without stomping on someone else's |
| 16584 | keybinding. If you ever write an extension to Emacs, please avoid |
| 16585 | taking any of these keys for public use. Create a key like @kbd{C-c |
| 16586 | C-w} instead. Otherwise, we will run out of `own' keys. |
| 16587 | |
| 16588 | @need 1250 |
| 16589 | Here is another keybinding, with a comment: |
| 16590 | |
| 16591 | @smallexample |
| 16592 | @group |
| 16593 | ;;; Keybinding for `occur' |
| 16594 | ; I use occur a lot, so let's bind it to a key: |
| 16595 | (global-set-key "\C-co" 'occur) |
| 16596 | @end group |
| 16597 | @end smallexample |
| 16598 | |
| 16599 | @findex occur |
| 16600 | The @code{occur} command shows all the lines in the current buffer |
| 16601 | that contain a match for a regular expression. Matching lines are |
| 16602 | shown in a buffer called @file{*Occur*}. That buffer serves as a menu |
| 16603 | to jump to occurrences. |
| 16604 | |
| 16605 | @findex global-unset-key |
| 16606 | @cindex Unbinding key |
| 16607 | @cindex Key unbinding |
| 16608 | @need 1250 |
| 16609 | Here is how to unbind a key, so it does not |
| 16610 | work: |
| 16611 | |
| 16612 | @smallexample |
| 16613 | @group |
| 16614 | ;;; Unbind `C-x f' |
| 16615 | (global-unset-key "\C-xf") |
| 16616 | @end group |
| 16617 | @end smallexample |
| 16618 | |
| 16619 | There is a reason for this unbinding: I found I inadvertently typed |
| 16620 | @w{@kbd{C-x f}} when I meant to type @kbd{C-x C-f}. Rather than find a |
| 16621 | file, as I intended, I accidentally set the width for filled text, |
| 16622 | almost always to a width I did not want. Since I hardly ever reset my |
| 16623 | default width, I simply unbound the key. |
| 16624 | |
| 16625 | @findex list-buffers, @r{rebound} |
| 16626 | @findex buffer-menu, @r{bound to key} |
| 16627 | @need 1250 |
| 16628 | The following rebinds an existing key: |
| 16629 | |
| 16630 | @smallexample |
| 16631 | @group |
| 16632 | ;;; Rebind `C-x C-b' for `buffer-menu' |
| 16633 | (global-set-key "\C-x\C-b" 'buffer-menu) |
| 16634 | @end group |
| 16635 | @end smallexample |
| 16636 | |
| 16637 | By default, @kbd{C-x C-b} runs the |
| 16638 | @code{list-buffers} command. This command lists |
| 16639 | your buffers in @emph{another} window. Since I |
| 16640 | almost always want to do something in that |
| 16641 | window, I prefer the @code{buffer-menu} |
| 16642 | command, which not only lists the buffers, |
| 16643 | but moves point into that window. |
| 16644 | |
| 16645 | @node Keymaps, Loading Files, Keybindings, Emacs Initialization |
| 16646 | @section Keymaps |
| 16647 | @cindex Keymaps |
| 16648 | @cindex Rebinding keys |
| 16649 | |
| 16650 | Emacs uses @dfn{keymaps} to record which keys call which commands. |
| 16651 | When you use @code{global-set-key} to set the keybinding for a single |
| 16652 | command in all parts of Emacs, you are specifying the keybinding in |
| 16653 | @code{current-global-map}. |
| 16654 | |
| 16655 | Specific modes, such as C mode or Text mode, have their own keymaps; |
| 16656 | the mode-specific keymaps override the global map that is shared by |
| 16657 | all buffers. |
| 16658 | |
| 16659 | The @code{global-set-key} function binds, or rebinds, the global |
| 16660 | keymap. For example, the following binds the key @kbd{C-x C-b} to the |
| 16661 | function @code{buffer-menu}: |
| 16662 | |
| 16663 | @smallexample |
| 16664 | (global-set-key "\C-x\C-b" 'buffer-menu) |
| 16665 | @end smallexample |
| 16666 | |
| 16667 | Mode-specific keymaps are bound using the @code{define-key} function, |
| 16668 | which takes a specific keymap as an argument, as well as the key and |
| 16669 | the command. For example, my @file{.emacs} file contains the |
| 16670 | following expression to bind the @code{texinfo-insert-@@group} command |
| 16671 | to @kbd{C-c C-c g}: |
| 16672 | |
| 16673 | @smallexample |
| 16674 | @group |
| 16675 | (define-key texinfo-mode-map "\C-c\C-cg" 'texinfo-insert-@@group) |
| 16676 | @end group |
| 16677 | @end smallexample |
| 16678 | |
| 16679 | @noindent |
| 16680 | The @code{texinfo-insert-@@group} function itself is a little extension |
| 16681 | to Texinfo mode that inserts @samp{@@group} into a Texinfo file. I |
| 16682 | use this command all the time and prefer to type the three strokes |
| 16683 | @kbd{C-c C-c g} rather than the six strokes @kbd{@@ g r o u p}. |
| 16684 | (@samp{@@group} and its matching @samp{@@end group} are commands that |
| 16685 | keep all enclosed text together on one page; many multi-line examples |
| 16686 | in this book are surrounded by @samp{@@group @dots{} @@end group}.) |
| 16687 | |
| 16688 | @need 1250 |
| 16689 | Here is the @code{texinfo-insert-@@group} function definition: |
| 16690 | |
| 16691 | @smallexample |
| 16692 | @group |
| 16693 | (defun texinfo-insert-@@group () |
| 16694 | "Insert the string @@group in a Texinfo buffer." |
| 16695 | (interactive) |
| 16696 | (beginning-of-line) |
| 16697 | (insert "@@group\n")) |
| 16698 | @end group |
| 16699 | @end smallexample |
| 16700 | |
| 16701 | (Of course, I could have used Abbrev mode to save typing, rather than |
| 16702 | write a function to insert a word; but I prefer key strokes consistent |
| 16703 | with other Texinfo mode key bindings.) |
| 16704 | |
| 16705 | You will see numerous @code{define-key} expressions in |
| 16706 | @file{loaddefs.el} as well as in the various mode libraries, such as |
| 16707 | @file{cc-mode.el} and @file{lisp-mode.el}. |
| 16708 | |
| 16709 | @xref{Key Bindings, , Customizing Key Bindings, emacs, The GNU Emacs |
| 16710 | Manual}, and @ref{Keymaps, , Keymaps, elisp, The GNU Emacs Lisp |
| 16711 | Reference Manual}, for more information about keymaps. |
| 16712 | |
| 16713 | @node Loading Files, Autoload, Keymaps, Emacs Initialization |
| 16714 | @section Loading Files |
| 16715 | @cindex Loading files |
| 16716 | @c findex load |
| 16717 | |
| 16718 | Many people in the GNU Emacs community have written extensions to |
| 16719 | Emacs. As time goes by, these extensions are often included in new |
| 16720 | releases. For example, the Calendar and Diary packages are now part |
| 16721 | of the standard GNU Emacs, as is Calc. |
| 16722 | |
| 16723 | You can use a @code{load} command to evaluate a complete file and |
| 16724 | thereby install all the functions and variables in the file into Emacs. |
| 16725 | For example: |
| 16726 | |
| 16727 | @c (auto-compression-mode t) |
| 16728 | |
| 16729 | @smallexample |
| 16730 | (load "~/emacs/slowsplit") |
| 16731 | @end smallexample |
| 16732 | |
| 16733 | This evaluates, i.e.@: loads, the @file{slowsplit.el} file or if it |
| 16734 | exists, the faster, byte compiled @file{slowsplit.elc} file from the |
| 16735 | @file{emacs} sub-directory of your home directory. The file contains |
| 16736 | the function @code{split-window-quietly}, which John Robinson wrote in |
| 16737 | 1989. |
| 16738 | |
| 16739 | The @code{split-window-quietly} function splits a window with the |
| 16740 | minimum of redisplay. I installed it in 1989 because it worked well |
| 16741 | with the slow 1200 baud terminals I was then using. Nowadays, I only |
| 16742 | occasionally come across such a slow connection, but I continue to use |
| 16743 | the function because I like the way it leaves the bottom half of a |
| 16744 | buffer in the lower of the new windows and the top half in the upper |
| 16745 | window. |
| 16746 | |
| 16747 | @need 1250 |
| 16748 | To replace the key binding for the default |
| 16749 | @code{split-window-vertically}, you must also unset that key and bind |
| 16750 | the keys to @code{split-window-quietly}, like this: |
| 16751 | |
| 16752 | @smallexample |
| 16753 | @group |
| 16754 | (global-unset-key "\C-x2") |
| 16755 | (global-set-key "\C-x2" 'split-window-quietly) |
| 16756 | @end group |
| 16757 | @end smallexample |
| 16758 | |
| 16759 | @vindex load-path |
| 16760 | If you load many extensions, as I do, then instead of specifying the |
| 16761 | exact location of the extension file, as shown above, you can specify |
| 16762 | that directory as part of Emacs' @code{load-path}. Then, when Emacs |
| 16763 | loads a file, it will search that directory as well as its default |
| 16764 | list of directories. (The default list is specified in @file{paths.h} |
| 16765 | when Emacs is built.) |
| 16766 | |
| 16767 | @need 1250 |
| 16768 | The following command adds your @file{~/emacs} directory to the |
| 16769 | existing load path: |
| 16770 | |
| 16771 | @smallexample |
| 16772 | @group |
| 16773 | ;;; Emacs Load Path |
| 16774 | (setq load-path (cons "~/emacs" load-path)) |
| 16775 | @end group |
| 16776 | @end smallexample |
| 16777 | |
| 16778 | Incidentally, @code{load-library} is an interactive interface to the |
| 16779 | @code{load} function. The complete function looks like this: |
| 16780 | |
| 16781 | @findex load-library |
| 16782 | @smallexample |
| 16783 | @group |
| 16784 | (defun load-library (library) |
| 16785 | "Load the library named LIBRARY. |
| 16786 | This is an interface to the function `load'." |
| 16787 | (interactive "sLoad library: ") |
| 16788 | (load library)) |
| 16789 | @end group |
| 16790 | @end smallexample |
| 16791 | |
| 16792 | The name of the function, @code{load-library}, comes from the use of |
| 16793 | `library' as a conventional synonym for `file'. The source for the |
| 16794 | @code{load-library} command is in the @file{files.el} library. |
| 16795 | |
| 16796 | Another interactive command that does a slightly different job is |
| 16797 | @code{load-file}. @xref{Lisp Libraries, , Libraries of Lisp Code for |
| 16798 | Emacs, emacs, The GNU Emacs Manual}, for information on the |
| 16799 | distinction between @code{load-library} and this command. |
| 16800 | |
| 16801 | @node Autoload, Simple Extension, Loading Files, Emacs Initialization |
| 16802 | @section Autoloading |
| 16803 | @findex autoload |
| 16804 | |
| 16805 | Instead of installing a function by loading the file that contains it, |
| 16806 | or by evaluating the function definition, you can make the function |
| 16807 | available but not actually install it until it is first called. This |
| 16808 | is called @dfn{autoloading}. |
| 16809 | |
| 16810 | When you execute an autoloaded function, Emacs automatically evaluates |
| 16811 | the file that contains the definition, and then calls the function. |
| 16812 | |
| 16813 | Emacs starts quicker with autoloaded functions, since their libraries |
| 16814 | are not loaded right away; but you need to wait a moment when you |
| 16815 | first use such a function, while its containing file is evaluated. |
| 16816 | |
| 16817 | Rarely used functions are frequently autoloaded. The |
| 16818 | @file{loaddefs.el} library contains hundreds of autoloaded functions, |
| 16819 | from @code{bookmark-set} to @code{wordstar-mode}. Of course, you may |
| 16820 | come to use a `rare' function frequently. When you do, you should |
| 16821 | load that function's file with a @code{load} expression in your |
| 16822 | @file{.emacs} file. |
| 16823 | |
| 16824 | In my @file{.emacs} file for Emacs version 21, I load 12 libraries |
| 16825 | that contain functions that would otherwise be autoloaded. (Actually, |
| 16826 | it would have been better to include these files in my `dumped' Emacs |
| 16827 | when I built it, but I forgot. @xref{Building Emacs, , Building |
| 16828 | Emacs, elisp, The GNU Emacs Lisp Reference Manual}, and the @file{INSTALL} |
| 16829 | file for more about dumping.) |
| 16830 | |
| 16831 | You may also want to include autoloaded expressions in your @file{.emacs} |
| 16832 | file. @code{autoload} is a built-in function that takes up to five |
| 16833 | arguments, the final three of which are optional. The first argument |
| 16834 | is the name of the function to be autoloaded; the second is the name |
| 16835 | of the file to be loaded. The third argument is documentation for the |
| 16836 | function, and the fourth tells whether the function can be called |
| 16837 | interactively. The fifth argument tells what type of |
| 16838 | object---@code{autoload} can handle a keymap or macro as well as a |
| 16839 | function (the default is a function). |
| 16840 | |
| 16841 | @need 800 |
| 16842 | Here is a typical example: |
| 16843 | |
| 16844 | @smallexample |
| 16845 | @group |
| 16846 | (autoload 'html-helper-mode |
| 16847 | "html-helper-mode" "Edit HTML documents" t) |
| 16848 | @end group |
| 16849 | @end smallexample |
| 16850 | |
| 16851 | @noindent |
| 16852 | (@code{html-helper-mode} is an alternative to @code{html-mode}, which |
| 16853 | is a standard part of the distribution). |
| 16854 | |
| 16855 | @noindent |
| 16856 | This expression autoloads the @code{html-helper-mode} function. It |
| 16857 | takes it from the @file{html-helper-mode.el} file (or from the byte |
| 16858 | compiled file @file{html-helper-mode.elc}, if it exists.) The file |
| 16859 | must be located in a directory specified by @code{load-path}. The |
| 16860 | documentation says that this is a mode to help you edit documents |
| 16861 | written in the HyperText Markup Language. You can call this mode |
| 16862 | interactively by typing @kbd{M-x html-helper-mode}. (You need to |
| 16863 | duplicate the function's regular documentation in the autoload |
| 16864 | expression because the regular function is not yet loaded, so its |
| 16865 | documentation is not available.) |
| 16866 | |
| 16867 | @xref{Autoload, , Autoload, elisp, The GNU Emacs Lisp Reference |
| 16868 | Manual}, for more information. |
| 16869 | |
| 16870 | @node Simple Extension, X11 Colors, Autoload, Emacs Initialization |
| 16871 | @section A Simple Extension: @code{line-to-top-of-window} |
| 16872 | @findex line-to-top-of-window |
| 16873 | @cindex Simple extension in @file{.emacs} file |
| 16874 | |
| 16875 | Here is a simple extension to Emacs that moves the line point is on to |
| 16876 | the top of the window. I use this all the time, to make text easier |
| 16877 | to read. |
| 16878 | |
| 16879 | You can put the following code into a separate file and then load it |
| 16880 | from your @file{.emacs} file, or you can include it within your |
| 16881 | @file{.emacs} file. |
| 16882 | |
| 16883 | @need 1250 |
| 16884 | Here is the definition: |
| 16885 | |
| 16886 | @smallexample |
| 16887 | @group |
| 16888 | ;;; Line to top of window; |
| 16889 | ;;; replace three keystroke sequence C-u 0 C-l |
| 16890 | (defun line-to-top-of-window () |
| 16891 | "Move the line point is on to top of window." |
| 16892 | (interactive) |
| 16893 | (recenter 0)) |
| 16894 | @end group |
| 16895 | @end smallexample |
| 16896 | |
| 16897 | @need 1250 |
| 16898 | Now for the keybinding. |
| 16899 | |
| 16900 | Nowadays, function keys as well as mouse button events and |
| 16901 | non-@sc{ascii} characters are written within square brackets, without |
| 16902 | quotation marks. (In Emacs version 18 and before, you had to write |
| 16903 | different function key bindings for each different make of terminal.) |
| 16904 | |
| 16905 | I bind @code{line-to-top-of-window} to my @key{F6} function key like |
| 16906 | this: |
| 16907 | |
| 16908 | @smallexample |
| 16909 | (global-set-key [f6] 'line-to-top-of-window) |
| 16910 | @end smallexample |
| 16911 | |
| 16912 | For more information, see @ref{Init Rebinding, , Rebinding Keys in |
| 16913 | Your Init File, emacs, The GNU Emacs Manual}. |
| 16914 | |
| 16915 | @cindex Conditional 'twixt two versions of Emacs |
| 16916 | @cindex Version of Emacs, choosing |
| 16917 | @cindex Emacs version, choosing |
| 16918 | If you run two versions of GNU Emacs, such as versions 20 and 21, and |
| 16919 | use one @file{.emacs} file, you can select which code to evaluate with |
| 16920 | the following conditional: |
| 16921 | |
| 16922 | @smallexample |
| 16923 | @group |
| 16924 | (cond |
| 16925 | ((string-equal (number-to-string 20) (substring (emacs-version) 10 12)) |
| 16926 | ;; evaluate version 20 code |
| 16927 | ( @dots{} )) |
| 16928 | ((string-equal (number-to-string 21) (substring (emacs-version) 10 12)) |
| 16929 | ;; evaluate version 21 code |
| 16930 | ( @dots{} ))) |
| 16931 | @end group |
| 16932 | @end smallexample |
| 16933 | |
| 16934 | For example, in contrast to version 20, version 21 blinks its cursor |
| 16935 | by default. I hate such blinking, as well as some other features in |
| 16936 | version 21, so I placed the following in my @file{.emacs} |
| 16937 | file@footnote{When I start instances of Emacs that do not load my |
| 16938 | @file{.emacs} file or any site file, I also turn off blinking: |
| 16939 | |
| 16940 | @smallexample |
| 16941 | emacs -q --no-site-file -eval '(blink-cursor-mode nil)' |
| 16942 | @end smallexample |
| 16943 | }: |
| 16944 | |
| 16945 | @smallexample |
| 16946 | @group |
| 16947 | (if (string-equal "21" (substring (emacs-version) 10 12)) |
| 16948 | (progn |
| 16949 | (blink-cursor-mode 0) |
| 16950 | ;; Insert newline when you press `C-n' (next-line) |
| 16951 | ;; at the end of the buffer |
| 16952 | (setq next-line-add-newlines t) |
| 16953 | @end group |
| 16954 | @group |
| 16955 | ;; Turn on image viewing |
| 16956 | (auto-image-file-mode t) |
| 16957 | @end group |
| 16958 | @group |
| 16959 | ;; Turn on menu bar (this bar has text) |
| 16960 | ;; (Use numeric argument to turn on) |
| 16961 | (menu-bar-mode 1) |
| 16962 | @end group |
| 16963 | @group |
| 16964 | ;; Turn off tool bar (this bar has icons) |
| 16965 | ;; (Use numeric argument to turn on) |
| 16966 | (tool-bar-mode nil) |
| 16967 | @end group |
| 16968 | @group |
| 16969 | ;; Turn off tooltip mode for tool bar |
| 16970 | ;; (This mode causes icon explanations to pop up) |
| 16971 | ;; (Use numeric argument to turn on) |
| 16972 | (tooltip-mode nil) |
| 16973 | ;; If tooltips turned on, make tips appear promptly |
| 16974 | (setq tooltip-delay 0.1) ; default is one second |
| 16975 | )) |
| 16976 | @end group |
| 16977 | @end smallexample |
| 16978 | |
| 16979 | @noindent |
| 16980 | (You will note that instead of typing @code{(number-to-string 21)}, I |
| 16981 | decided to save typing and wrote `21' as a string, @code{"21"}, rather |
| 16982 | than convert it from an integer to a string. In this instance, this |
| 16983 | expression is better than the longer, but more general |
| 16984 | @code{(number-to-string 21)}. However, if you do not know ahead of |
| 16985 | time what type of information will be returned, then the |
| 16986 | @code{number-to-string} function will be needed.) |
| 16987 | |
| 16988 | @node X11 Colors, Miscellaneous, Simple Extension, Emacs Initialization |
| 16989 | @section X11 Colors |
| 16990 | |
| 16991 | You can specify colors when you use Emacs with the MIT X Windowing |
| 16992 | system. |
| 16993 | |
| 16994 | I dislike the default colors and specify my own. |
| 16995 | |
| 16996 | @need 1250 |
| 16997 | Here are the expressions in my @file{.emacs} |
| 16998 | file that set values: |
| 16999 | |
| 17000 | @smallexample |
| 17001 | @group |
| 17002 | ;; Set cursor color |
| 17003 | (set-cursor-color "white") |
| 17004 | |
| 17005 | ;; Set mouse color |
| 17006 | (set-mouse-color "white") |
| 17007 | |
| 17008 | ;; Set foreground and background |
| 17009 | (set-foreground-color "white") |
| 17010 | (set-background-color "darkblue") |
| 17011 | @end group |
| 17012 | |
| 17013 | @group |
| 17014 | ;;; Set highlighting colors for isearch and drag |
| 17015 | (set-face-foreground 'highlight "white") |
| 17016 | (set-face-background 'highlight "blue") |
| 17017 | @end group |
| 17018 | |
| 17019 | @group |
| 17020 | (set-face-foreground 'region "cyan") |
| 17021 | (set-face-background 'region "blue") |
| 17022 | @end group |
| 17023 | |
| 17024 | @group |
| 17025 | (set-face-foreground 'secondary-selection "skyblue") |
| 17026 | (set-face-background 'secondary-selection "darkblue") |
| 17027 | @end group |
| 17028 | |
| 17029 | @group |
| 17030 | ;; Set calendar highlighting colors |
| 17031 | (setq calendar-load-hook |
| 17032 | '(lambda () |
| 17033 | (set-face-foreground 'diary-face "skyblue") |
| 17034 | (set-face-background 'holiday-face "slate blue") |
| 17035 | (set-face-foreground 'holiday-face "white"))) |
| 17036 | @end group |
| 17037 | @end smallexample |
| 17038 | |
| 17039 | The various shades of blue soothe my eye and prevent me from seeing |
| 17040 | the screen flicker. |
| 17041 | |
| 17042 | Alternatively, I could have set my specifications in various X |
| 17043 | initialization files. For example, I could set the foreground, |
| 17044 | background, cursor, and pointer (i.e., mouse) colors in my |
| 17045 | @file{~/.Xresources} file like this: |
| 17046 | |
| 17047 | @smallexample |
| 17048 | @group |
| 17049 | Emacs*foreground: white |
| 17050 | Emacs*background: darkblue |
| 17051 | Emacs*cursorColor: white |
| 17052 | Emacs*pointerColor: white |
| 17053 | @end group |
| 17054 | @end smallexample |
| 17055 | |
| 17056 | In any event, since it is not part of Emacs, I set the root color of |
| 17057 | my X window in my @file{~/.xinitrc} file, like this@footnote{I |
| 17058 | occasionally run more modern window managers, such as Sawfish with |
| 17059 | GNOME, Enlightenment, SCWM, or KDE; in those cases, I often specify an |
| 17060 | image rather than a plain color.}: |
| 17061 | |
| 17062 | @smallexample |
| 17063 | @group |
| 17064 | # I use TWM for window manager. |
| 17065 | xsetroot -solid Navy -fg white & |
| 17066 | @end group |
| 17067 | @end smallexample |
| 17068 | |
| 17069 | @node Miscellaneous, Mode Line, X11 Colors, Emacs Initialization |
| 17070 | @section Miscellaneous Settings for a @file{.emacs} File |
| 17071 | |
| 17072 | Here are a few miscellaneous settings: |
| 17073 | @sp 1 |
| 17074 | |
| 17075 | @itemize @minus |
| 17076 | @item |
| 17077 | Set the shape and color of the mouse cursor: |
| 17078 | @smallexample |
| 17079 | @group |
| 17080 | ; Cursor shapes are defined in |
| 17081 | ; `/usr/include/X11/cursorfont.h'; |
| 17082 | ; for example, the `target' cursor is number 128; |
| 17083 | ; the `top_left_arrow' cursor is number 132. |
| 17084 | @end group |
| 17085 | |
| 17086 | @group |
| 17087 | (let ((mpointer (x-get-resource "*mpointer" |
| 17088 | "*emacs*mpointer"))) |
| 17089 | ;; If you have not set your mouse pointer |
| 17090 | ;; then set it, otherwise leave as is: |
| 17091 | (if (eq mpointer nil) |
| 17092 | (setq mpointer "132")) ; top_left_arrow |
| 17093 | @end group |
| 17094 | @group |
| 17095 | (setq x-pointer-shape (string-to-int mpointer)) |
| 17096 | (set-mouse-color "white")) |
| 17097 | @end group |
| 17098 | @end smallexample |
| 17099 | |
| 17100 | @item |
| 17101 | Convert @kbd{@key{CTL}-h} into @key{DEL} and @key{DEL} |
| 17102 | into @kbd{@key{CTL}-h}.@* |
| 17103 | (Some olders keyboards needed this, although I have not seen the |
| 17104 | problem recently.) |
| 17105 | |
| 17106 | @smallexample |
| 17107 | @group |
| 17108 | ;; Translate `C-h' to <DEL>. |
| 17109 | ; (keyboard-translate ?\C-h ?\C-?) |
| 17110 | |
| 17111 | ;; Translate <DEL> to `C-h'. |
| 17112 | (keyboard-translate ?\C-? ?\C-h) |
| 17113 | @end group |
| 17114 | @end smallexample |
| 17115 | |
| 17116 | @item Turn off a blinking cursor! |
| 17117 | |
| 17118 | @smallexample |
| 17119 | @group |
| 17120 | (if (fboundp 'blink-cursor-mode) |
| 17121 | (blink-cursor-mode -1)) |
| 17122 | @end group |
| 17123 | @end smallexample |
| 17124 | |
| 17125 | @item Ignore case when using `grep'@* |
| 17126 | @samp{-n}@w{ } Prefix each line of output with line number@* |
| 17127 | @samp{-i}@w{ } Ignore case distinctions@* |
| 17128 | @samp{-e}@w{ } Protect patterns beginning with a hyphen character, @samp{-} |
| 17129 | |
| 17130 | @smallexample |
| 17131 | (setq grep-command "grep -n -i -e ") |
| 17132 | @end smallexample |
| 17133 | |
| 17134 | @item Automatically uncompress compressed files when visiting them |
| 17135 | |
| 17136 | @smallexample |
| 17137 | (load "uncompress") |
| 17138 | @end smallexample |
| 17139 | |
| 17140 | @item Find an existing buffer, even if it has a different name@* |
| 17141 | This avoids problems with symbolic links. |
| 17142 | |
| 17143 | @smallexample |
| 17144 | (setq find-file-existing-other-name t) |
| 17145 | @end smallexample |
| 17146 | |
| 17147 | @item Set your language environment and default input method |
| 17148 | |
| 17149 | @smallexample |
| 17150 | @group |
| 17151 | (set-language-environment "latin-1") |
| 17152 | ;; Remember you can enable or disable multilingual text input |
| 17153 | ;; with the @code{toggle-input-method'} (@kbd{C-\}) command |
| 17154 | (setq default-input-method "latin-1-prefix") |
| 17155 | @end group |
| 17156 | @end smallexample |
| 17157 | |
| 17158 | If you want to write with Chinese `GB' characters, set this instead: |
| 17159 | |
| 17160 | @smallexample |
| 17161 | @group |
| 17162 | (set-language-environment "Chinese-GB") |
| 17163 | (setq default-input-method "chinese-tonepy") |
| 17164 | @end group |
| 17165 | @end smallexample |
| 17166 | @end itemize |
| 17167 | |
| 17168 | @subsubheading Fixing Unpleasant Key Bindings |
| 17169 | @cindex Key bindings, fixing |
| 17170 | @cindex Bindings, key, fixing unpleasant |
| 17171 | |
| 17172 | Some systems bind keys unpleasantly. Sometimes, for example, the |
| 17173 | @key{CTL} key appears in an awkward spot rather than at the far left |
| 17174 | of the home row. |
| 17175 | |
| 17176 | Usually, when people fix these sorts of keybindings, they do not |
| 17177 | change their @file{~/.emacs} file. Instead, they bind the proper keys |
| 17178 | on their consoles with the @code{loadkeys} or @code{install-keymap} |
| 17179 | commands in their boot script and then include @code{xmodmap} commands |
| 17180 | in their @file{.xinitrc} or @file{.Xsession} file for X Windows. |
| 17181 | |
| 17182 | @need 1250 |
| 17183 | @noindent |
| 17184 | For a boot script: |
| 17185 | |
| 17186 | @smallexample |
| 17187 | @group |
| 17188 | loadkeys /usr/share/keymaps/i386/qwerty/emacs2.kmap.gz |
| 17189 | |
| 17190 | @exdent or |
| 17191 | |
| 17192 | install-keymap emacs2 |
| 17193 | @end group |
| 17194 | @end smallexample |
| 17195 | |
| 17196 | @need 1250 |
| 17197 | @noindent |
| 17198 | For a @file{.xinitrc} or @file{.Xsession} file when the @key{Caps |
| 17199 | Lock} key is at the far left of the home row: |
| 17200 | |
| 17201 | @smallexample |
| 17202 | @group |
| 17203 | # Bind the key labeled `Caps Lock' to `Control' |
| 17204 | # (Such a broken user interface suggests that keyboard manufacturers |
| 17205 | # think that computers are typewriters from 1885.) |
| 17206 | |
| 17207 | xmodmap -e "clear Lock" |
| 17208 | xmodmap -e "add Control = Caps_Lock" |
| 17209 | @end group |
| 17210 | @end smallexample |
| 17211 | |
| 17212 | @need 1250 |
| 17213 | @noindent |
| 17214 | In a @file{.xinitrc} or @file{.Xsession} file, to convert an @key{ALT} |
| 17215 | key to a @key{META} key: |
| 17216 | |
| 17217 | @smallexample |
| 17218 | @group |
| 17219 | # Some ill designed keyboards have a key labeled ALT and no Meta |
| 17220 | xmodmap -e "keysym Alt_L = Meta_L Alt_L" |
| 17221 | @end group |
| 17222 | @end smallexample |
| 17223 | |
| 17224 | @node Mode Line, , Miscellaneous, Emacs Initialization |
| 17225 | @section A Modified Mode Line |
| 17226 | @vindex default-mode-line-format |
| 17227 | @cindex Mode line format |
| 17228 | |
| 17229 | Finally, a feature I really like: a modified mode line. |
| 17230 | |
| 17231 | When I work over a network, I forget which machine I am using. Also, |
| 17232 | I tend to I lose track of where I am, and which line point is on. |
| 17233 | |
| 17234 | So I reset my mode line to look like this: |
| 17235 | |
| 17236 | @smallexample |
| 17237 | -:-- foo.texi rattlesnake:/home/bob/ Line 1 (Texinfo Fill) Top |
| 17238 | @end smallexample |
| 17239 | |
| 17240 | I am visiting a file called @file{foo.texi}, on my machine |
| 17241 | @file{rattlesnake} in my @file{/home/bob} buffer. I am on line 1, in |
| 17242 | Texinfo mode, and am at the top of the buffer. |
| 17243 | |
| 17244 | @need 1200 |
| 17245 | My @file{.emacs} file has a section that looks like this: |
| 17246 | |
| 17247 | @smallexample |
| 17248 | @group |
| 17249 | ;; Set a Mode Line that tells me which machine, which directory, |
| 17250 | ;; and which line I am on, plus the other customary information. |
| 17251 | (setq default-mode-line-format |
| 17252 | (quote |
| 17253 | (#("-" 0 1 |
| 17254 | (help-echo |
| 17255 | "mouse-1: select window, mouse-2: delete others ...")) |
| 17256 | mode-line-mule-info |
| 17257 | mode-line-modified |
| 17258 | mode-line-frame-identification |
| 17259 | " " |
| 17260 | @end group |
| 17261 | @group |
| 17262 | mode-line-buffer-identification |
| 17263 | " " |
| 17264 | (:eval (substring |
| 17265 | (system-name) 0 (string-match "\\..+" (system-name)))) |
| 17266 | ":" |
| 17267 | default-directory |
| 17268 | #(" " 0 1 |
| 17269 | (help-echo |
| 17270 | "mouse-1: select window, mouse-2: delete others ...")) |
| 17271 | (line-number-mode " Line %l ") |
| 17272 | global-mode-string |
| 17273 | @end group |
| 17274 | @group |
| 17275 | #(" %[(" 0 6 |
| 17276 | (help-echo |
| 17277 | "mouse-1: select window, mouse-2: delete others ...")) |
| 17278 | (:eval (mode-line-mode-name)) |
| 17279 | mode-line-process |
| 17280 | minor-mode-alist |
| 17281 | #("%n" 0 2 (help-echo "mouse-2: widen" local-map (keymap ...))) |
| 17282 | ")%] " |
| 17283 | (-3 . "%P") |
| 17284 | ;; "-%-" |
| 17285 | ))) |
| 17286 | @end group |
| 17287 | @end smallexample |
| 17288 | |
| 17289 | @noindent |
| 17290 | Here, I redefine the default mode line. Most of the parts are from |
| 17291 | the original; but I make a few changes. I set the @emph{default} mode |
| 17292 | line format so as to permit various modes, such as Info, to override |
| 17293 | it. |
| 17294 | |
| 17295 | Many elements in the list are self-explanatory: |
| 17296 | @code{mode-line-modified} is a variable that tells whether the buffer |
| 17297 | has been modified, @code{mode-name} tells the name of the mode, and so |
| 17298 | on. However, the format looks complicated because of two features we |
| 17299 | have not discussed. |
| 17300 | |
| 17301 | @cindex Properties, in mode line example |
| 17302 | The first string in the mode line is a dash or hyphen, @samp{-}. In |
| 17303 | the old days, it would have been specified simply as @code{"-"}. But |
| 17304 | nowadays, Emacs can add properties to a string, such as highlighting |
| 17305 | or, as in this case, a help feature. If you place your mouse cursor |
| 17306 | over the hyphen, some help information appears (By default, you must |
| 17307 | wait one second before the information appears. You can change that |
| 17308 | timing by changing the value of @code{tooltip-delay}.) |
| 17309 | |
| 17310 | @need 1000 |
| 17311 | The new string format has a special syntax: |
| 17312 | |
| 17313 | @smallexample |
| 17314 | #("-" 0 1 (help-echo "mouse-1: select window, ...")) |
| 17315 | @end smallexample |
| 17316 | |
| 17317 | @noindent |
| 17318 | The @code{#(} begins a list. The first element of the list is the |
| 17319 | string itself, just one @samp{-}. The second and third |
| 17320 | elements specify the range over which the fourth element applies. A |
| 17321 | range starts @emph{after} a character, so a zero means the range |
| 17322 | starts just before the first character; a 1 means that the range ends |
| 17323 | just after the first character. The third element is the property for |
| 17324 | the range. It consists of a property list, a |
| 17325 | property name, in this case, @samp{help-echo}, followed by a value, in this |
| 17326 | case, a string. The second, third, and fourth elements of this new |
| 17327 | string format can be repeated. |
| 17328 | |
| 17329 | @xref{Text Properties, , Text Properties, elisp, The GNU Emacs Lisp |
| 17330 | Reference Manual}, and see @ref{Mode Line Format, , Mode Line Format, |
| 17331 | elisp, The GNU Emacs Lisp Reference Manual}, for more information. |
| 17332 | |
| 17333 | @code{mode-line-buffer-identification} |
| 17334 | displays the current buffer name. It is a list |
| 17335 | beginning @code{(#("%12b" 0 4 @dots{}}. |
| 17336 | The @code{#(} begins the list. |
| 17337 | |
| 17338 | The @samp{"%12b"} displays the current buffer name, using the |
| 17339 | @code{buffer-name} function with which we are familiar; the `12' |
| 17340 | specifies the maximum number of characters that will be displayed. |
| 17341 | When a name has fewer characters, whitespace is added to fill out to |
| 17342 | this number. (Buffer names can and often should be longer than 12 |
| 17343 | characters; this length works well in a typical 80 column wide |
| 17344 | window.) |
| 17345 | |
| 17346 | @code{:eval} is a new feature in GNU Emacs version 21. It says to |
| 17347 | evaluate the following form and use the result as a string to display. |
| 17348 | In this case, the expression displays the first component of the full |
| 17349 | system name. The end of the first component is a @samp{.} (`period'), |
| 17350 | so I use the @code{string-match} function to tell me the length of the |
| 17351 | first component. The substring from the zeroth character to that |
| 17352 | length is the name of the machine. |
| 17353 | |
| 17354 | @need 1250 |
| 17355 | This is the expression: |
| 17356 | |
| 17357 | @smallexample |
| 17358 | @group |
| 17359 | (:eval (substring |
| 17360 | (system-name) 0 (string-match "\\..+" (system-name)))) |
| 17361 | @end group |
| 17362 | @end smallexample |
| 17363 | |
| 17364 | @samp{%[} and @samp{%]} cause a pair of square brackets |
| 17365 | to appear for each recursive editing level. @samp{%n} says `Narrow' |
| 17366 | when narrowing is in effect. @samp{%P} tells you the percentage of |
| 17367 | the buffer that is above the bottom of the window, or `Top', `Bottom', |
| 17368 | or `All'. (A lower case @samp{p} tell you the percentage above the |
| 17369 | @emph{top} of the window.) @samp{%-} inserts enough dashes to fill |
| 17370 | out the line. |
| 17371 | |
| 17372 | Remember, ``You don't have to like Emacs to like it'' --- your own |
| 17373 | Emacs can have different colors, different commands, and different |
| 17374 | keys than a default Emacs. |
| 17375 | |
| 17376 | On the other hand, if you want to bring up a plain `out of the box' |
| 17377 | Emacs, with no customization, type: |
| 17378 | |
| 17379 | @smallexample |
| 17380 | emacs -q |
| 17381 | @end smallexample |
| 17382 | |
| 17383 | @noindent |
| 17384 | This will start an Emacs that does @emph{not} load your |
| 17385 | @file{~/.emacs} initialization file. A plain, default Emacs. Nothing |
| 17386 | more. |
| 17387 | |
| 17388 | @node Debugging, Conclusion, Emacs Initialization, Top |
| 17389 | @chapter Debugging |
| 17390 | @cindex debugging |
| 17391 | |
| 17392 | GNU Emacs has two debuggers, @code{debug} and @code{edebug}. The |
| 17393 | first is built into the internals of Emacs and is always with you; |
| 17394 | the second requires that you instrument a function before you can use it. |
| 17395 | |
| 17396 | Both debuggers are described extensively in @ref{Debugging, , |
| 17397 | Debugging Lisp Programs, elisp, The GNU Emacs Lisp Reference Manual}. |
| 17398 | In this chapter, I will walk through a short example of each. |
| 17399 | |
| 17400 | @menu |
| 17401 | * debug:: How to use the built-in debugger. |
| 17402 | * debug-on-entry:: Start debugging when you call a function. |
| 17403 | * debug-on-quit:: Start debugging when you quit with @kbd{C-g}. |
| 17404 | * edebug:: How to use Edebug, a source level debugger. |
| 17405 | * Debugging Exercises:: |
| 17406 | @end menu |
| 17407 | |
| 17408 | @node debug, debug-on-entry, Debugging, Debugging |
| 17409 | @section @code{debug} |
| 17410 | @findex debug |
| 17411 | |
| 17412 | Suppose you have written a function definition that is intended to |
| 17413 | return the sum of the numbers 1 through a given number. (This is the |
| 17414 | @code{triangle} function discussed earlier. @xref{Decrementing |
| 17415 | Example, , Example with Decrementing Counter}, for a discussion.) |
| 17416 | @c xref{Decrementing Loop,, Loop with a Decrementing Counter}, for a discussion.) |
| 17417 | |
| 17418 | However, your function definition has a bug. You have mistyped |
| 17419 | @samp{1=} for @samp{1-}. Here is the broken definition: |
| 17420 | |
| 17421 | @findex triangle-bugged |
| 17422 | @smallexample |
| 17423 | @group |
| 17424 | (defun triangle-bugged (number) |
| 17425 | "Return sum of numbers 1 through NUMBER inclusive." |
| 17426 | (let ((total 0)) |
| 17427 | (while (> number 0) |
| 17428 | (setq total (+ total number)) |
| 17429 | (setq number (1= number))) ; @r{Error here.} |
| 17430 | total)) |
| 17431 | @end group |
| 17432 | @end smallexample |
| 17433 | |
| 17434 | If you are reading this in Info, you can evaluate this definition in |
| 17435 | the normal fashion. You will see @code{triangle-bugged} appear in the |
| 17436 | echo area. |
| 17437 | |
| 17438 | @need 1250 |
| 17439 | Now evaluate the @code{triangle-bugged} function with an |
| 17440 | argument of 4: |
| 17441 | |
| 17442 | @smallexample |
| 17443 | (triangle-bugged 4) |
| 17444 | @end smallexample |
| 17445 | |
| 17446 | @noindent |
| 17447 | In GNU Emacs version 21, you will create and enter a |
| 17448 | @file{*Backtrace*} buffer that says: |
| 17449 | |
| 17450 | @noindent |
| 17451 | @smallexample |
| 17452 | @group |
| 17453 | ---------- Buffer: *Backtrace* ---------- |
| 17454 | Debugger entered--Lisp error: (void-function 1=) |
| 17455 | (1= number) |
| 17456 | (setq number (1= number)) |
| 17457 | (while (> number 0) (setq total (+ total number)) |
| 17458 | (setq number (1= number))) |
| 17459 | (let ((total 0)) (while (> number 0) (setq total ...) |
| 17460 | (setq number ...)) total) |
| 17461 | triangle-bugged(4) |
| 17462 | @end group |
| 17463 | @group |
| 17464 | eval((triangle-bugged 4)) |
| 17465 | eval-last-sexp-1(nil) |
| 17466 | eval-last-sexp(nil) |
| 17467 | call-interactively(eval-last-sexp) |
| 17468 | ---------- Buffer: *Backtrace* ---------- |
| 17469 | @end group |
| 17470 | @end smallexample |
| 17471 | |
| 17472 | @noindent |
| 17473 | (I have reformatted this example slightly; the debugger does not fold |
| 17474 | long lines. As usual, you can quit the debugger by typing @kbd{q} in |
| 17475 | the @file{*Backtrace*} buffer.) |
| 17476 | |
| 17477 | In practice, for a bug as simple as this, the `Lisp error' line will |
| 17478 | tell you what you need to know to correct the definition. The |
| 17479 | function @code{1=} is `void'. |
| 17480 | |
| 17481 | @need 800 |
| 17482 | In GNU Emacs 20 and before, you will see: |
| 17483 | |
| 17484 | @smallexample |
| 17485 | Symbol's function definition is void:@: 1= |
| 17486 | @end smallexample |
| 17487 | |
| 17488 | @noindent |
| 17489 | which has the same meaning as the @file{*Backtrace*} buffer line in |
| 17490 | version 21. |
| 17491 | |
| 17492 | However, suppose you are not quite certain what is going on? |
| 17493 | You can read the complete backtrace. |
| 17494 | |
| 17495 | In this case, you need to run GNU Emacs 21, which automatically starts |
| 17496 | the debugger that puts you in the @file{*Backtrace*} buffer; or else, |
| 17497 | you need to start the debugger manually as described below. |
| 17498 | |
| 17499 | Read the @file{*Backtrace*} buffer from the bottom up; it tells you |
| 17500 | what Emacs did that led to the error. Emacs made an interactive call |
| 17501 | to @kbd{C-x C-e} (@code{eval-last-sexp}), which led to the evaluation |
| 17502 | of the @code{triangle-bugged} expression. Each line above tells you |
| 17503 | what the Lisp interpreter evaluated next. |
| 17504 | |
| 17505 | @need 1250 |
| 17506 | The third line from the top of the buffer is |
| 17507 | |
| 17508 | @smallexample |
| 17509 | (setq number (1= number)) |
| 17510 | @end smallexample |
| 17511 | |
| 17512 | @noindent |
| 17513 | Emacs tried to evaluate this expression; in order to do so, it tried |
| 17514 | to evaluate the inner expression shown on the second line from the |
| 17515 | top: |
| 17516 | |
| 17517 | @smallexample |
| 17518 | (1= number) |
| 17519 | @end smallexample |
| 17520 | |
| 17521 | @need 1250 |
| 17522 | @noindent |
| 17523 | This is where the error occurred; as the top line says: |
| 17524 | |
| 17525 | @smallexample |
| 17526 | Debugger entered--Lisp error: (void-function 1=) |
| 17527 | @end smallexample |
| 17528 | |
| 17529 | @noindent |
| 17530 | You can correct the mistake, re-evaluate the function definition, and |
| 17531 | then run your test again. |
| 17532 | |
| 17533 | @node debug-on-entry, debug-on-quit, debug, Debugging |
| 17534 | @section @code{debug-on-entry} |
| 17535 | @findex debug-on-entry |
| 17536 | |
| 17537 | GNU Emacs 21 starts the debugger automatically when your function has |
| 17538 | an error. GNU Emacs version 20 and before did not; it simply |
| 17539 | presented you with an error message. You had to start the debugger |
| 17540 | manually. |
| 17541 | |
| 17542 | You can start the debugger manually for all versions of Emacs; the |
| 17543 | advantage is that the debugger runs even if you do not have a bug in |
| 17544 | your code. Sometimes your code will be free of bugs! |
| 17545 | |
| 17546 | You can enter the debugger when you call the function by calling |
| 17547 | @code{debug-on-entry}. |
| 17548 | |
| 17549 | @need 1250 |
| 17550 | @noindent |
| 17551 | Type: |
| 17552 | |
| 17553 | @smallexample |
| 17554 | M-x debug-on-entry RET triangle-bugged RET |
| 17555 | @end smallexample |
| 17556 | |
| 17557 | @need 1250 |
| 17558 | @noindent |
| 17559 | Now, evaluate the following: |
| 17560 | |
| 17561 | @smallexample |
| 17562 | (triangle-bugged 5) |
| 17563 | @end smallexample |
| 17564 | |
| 17565 | @noindent |
| 17566 | All versions of Emacs will create a @file{*Backtrace*} buffer and tell |
| 17567 | you that it is beginning to evaluate the @code{triangle-bugged} |
| 17568 | function: |
| 17569 | |
| 17570 | @smallexample |
| 17571 | @group |
| 17572 | ---------- Buffer: *Backtrace* ---------- |
| 17573 | Debugger entered--entering a function: |
| 17574 | * triangle-bugged(5) |
| 17575 | eval((triangle-bugged 5)) |
| 17576 | @end group |
| 17577 | @group |
| 17578 | eval-last-sexp-1(nil) |
| 17579 | eval-last-sexp(nil) |
| 17580 | call-interactively(eval-last-sexp) |
| 17581 | ---------- Buffer: *Backtrace* ---------- |
| 17582 | @end group |
| 17583 | @end smallexample |
| 17584 | |
| 17585 | In the @file{*Backtrace*} buffer, type @kbd{d}. Emacs will evaluate |
| 17586 | the first expression in @code{triangle-bugged}; the buffer will look |
| 17587 | like this: |
| 17588 | |
| 17589 | @smallexample |
| 17590 | @group |
| 17591 | ---------- Buffer: *Backtrace* ---------- |
| 17592 | Debugger entered--beginning evaluation of function call form: |
| 17593 | * (let ((total 0)) (while (> number 0) (setq total ...) |
| 17594 | (setq number ...)) total) |
| 17595 | * triangle-bugged(5) |
| 17596 | eval((triangle-bugged 5)) |
| 17597 | @end group |
| 17598 | @group |
| 17599 | eval-last-sexp-1(nil) |
| 17600 | eval-last-sexp(nil) |
| 17601 | call-interactively(eval-last-sexp) |
| 17602 | ---------- Buffer: *Backtrace* ---------- |
| 17603 | @end group |
| 17604 | @end smallexample |
| 17605 | |
| 17606 | @noindent |
| 17607 | Now, type @kbd{d} again, eight times, slowly. Each time you type |
| 17608 | @kbd{d}, Emacs will evaluate another expression in the function |
| 17609 | definition. |
| 17610 | |
| 17611 | @need 1750 |
| 17612 | Eventually, the buffer will look like this: |
| 17613 | |
| 17614 | @smallexample |
| 17615 | @group |
| 17616 | ---------- Buffer: *Backtrace* ---------- |
| 17617 | Debugger entered--beginning evaluation of function call form: |
| 17618 | * (setq number (1= number)) |
| 17619 | * (while (> number 0) (setq total (+ total number)) |
| 17620 | (setq number (1= number))) |
| 17621 | @group |
| 17622 | @end group |
| 17623 | * (let ((total 0)) (while (> number 0) (setq total ...) |
| 17624 | (setq number ...)) total) |
| 17625 | * triangle-bugged(5) |
| 17626 | eval((triangle-bugged 5)) |
| 17627 | @group |
| 17628 | @end group |
| 17629 | eval-last-sexp-1(nil) |
| 17630 | eval-last-sexp(nil) |
| 17631 | call-interactively(eval-last-sexp) |
| 17632 | ---------- Buffer: *Backtrace* ---------- |
| 17633 | @end group |
| 17634 | @end smallexample |
| 17635 | |
| 17636 | @need 1500 |
| 17637 | @noindent |
| 17638 | Finally, after you type @kbd{d} two more times, Emacs will reach the |
| 17639 | error, and the top two lines of the @file{*Backtrace*} buffer will look |
| 17640 | like this: |
| 17641 | |
| 17642 | @smallexample |
| 17643 | @group |
| 17644 | ---------- Buffer: *Backtrace* ---------- |
| 17645 | Debugger entered--Lisp error: (void-function 1=) |
| 17646 | * (1= number) |
| 17647 | @dots{} |
| 17648 | ---------- Buffer: *Backtrace* ---------- |
| 17649 | @end group |
| 17650 | @end smallexample |
| 17651 | |
| 17652 | By typing @kbd{d}, you were able to step through the function. |
| 17653 | |
| 17654 | You can quit a @file{*Backtrace*} buffer by typing @kbd{q} in it; this |
| 17655 | quits the trace, but does not cancel @code{debug-on-entry}. |
| 17656 | |
| 17657 | @findex cancel-debug-on-entry |
| 17658 | To cancel the effect of @code{debug-on-entry}, call |
| 17659 | @code{cancel-debug-on-entry} and the name of the function, like this: |
| 17660 | |
| 17661 | @smallexample |
| 17662 | M-x cancel-debug-on-entry RET triangle-bugged RET |
| 17663 | @end smallexample |
| 17664 | |
| 17665 | @noindent |
| 17666 | (If you are reading this in Info, cancel @code{debug-on-entry} now.) |
| 17667 | |
| 17668 | @node debug-on-quit, edebug, debug-on-entry, Debugging |
| 17669 | @section @code{debug-on-quit} and @code{(debug)} |
| 17670 | |
| 17671 | In addition to setting @code{debug-on-error} or calling @code{debug-on-entry}, |
| 17672 | there are two other ways to start @code{debug}. |
| 17673 | |
| 17674 | @findex debug-on-quit |
| 17675 | You can start @code{debug} whenever you type @kbd{C-g} |
| 17676 | (@code{keyboard-quit}) by setting the variable @code{debug-on-quit} to |
| 17677 | @code{t}. This is useful for debugging infinite loops. |
| 17678 | |
| 17679 | @need 1500 |
| 17680 | @cindex @code{(debug)} in code |
| 17681 | Or, you can insert a line that says @code{(debug)} into your code |
| 17682 | where you want the debugger to start, like this: |
| 17683 | |
| 17684 | @smallexample |
| 17685 | @group |
| 17686 | (defun triangle-bugged (number) |
| 17687 | "Return sum of numbers 1 through NUMBER inclusive." |
| 17688 | (let ((total 0)) |
| 17689 | (while (> number 0) |
| 17690 | (setq total (+ total number)) |
| 17691 | (debug) ; @r{Start debugger.} |
| 17692 | (setq number (1= number))) ; @r{Error here.} |
| 17693 | total)) |
| 17694 | @end group |
| 17695 | @end smallexample |
| 17696 | |
| 17697 | The @code{debug} function is described in detail in @ref{Debugger, , |
| 17698 | The Lisp Debugger, elisp, The GNU Emacs Lisp Reference Manual}. |
| 17699 | |
| 17700 | @node edebug, Debugging Exercises, debug-on-quit, Debugging |
| 17701 | @section The @code{edebug} Source Level Debugger |
| 17702 | @cindex Source level debugger |
| 17703 | @findex edebug |
| 17704 | |
| 17705 | Edebug is a source level debugger. Edebug normally displays the |
| 17706 | source of the code you are debugging, with an arrow at the left that |
| 17707 | shows which line you are currently executing. |
| 17708 | |
| 17709 | You can walk through the execution of a function, line by line, or run |
| 17710 | quickly until reaching a @dfn{breakpoint} where execution stops. |
| 17711 | |
| 17712 | Edebug is described in @ref{edebug, , Edebug, elisp, The GNU Emacs |
| 17713 | Lisp Reference Manual}. |
| 17714 | |
| 17715 | @need 1250 |
| 17716 | Here is a bugged function definition for @code{triangle-recursively}. |
| 17717 | @xref{Recursive triangle function, , Recursion in place of a counter}, |
| 17718 | for a review of it. |
| 17719 | |
| 17720 | @smallexample |
| 17721 | @group |
| 17722 | (defun triangle-recursively-bugged (number) |
| 17723 | "Return sum of numbers 1 through NUMBER inclusive. |
| 17724 | Uses recursion." |
| 17725 | (if (= number 1) |
| 17726 | 1 |
| 17727 | (+ number |
| 17728 | (triangle-recursively-bugged |
| 17729 | (1= number))))) ; @r{Error here.} |
| 17730 | @end group |
| 17731 | @end smallexample |
| 17732 | |
| 17733 | @noindent |
| 17734 | Normally, you would install this definition by positioning your cursor |
| 17735 | after the function's closing parenthesis and typing @kbd{C-x C-e} |
| 17736 | (@code{eval-last-sexp}) or else by positioning your cursor within the |
| 17737 | definition and typing @kbd{C-M-x} (@code{eval-defun}). (By default, |
| 17738 | the @code{eval-defun} command works only in Emacs Lisp mode or in Lisp |
| 17739 | Interactive mode.) |
| 17740 | |
| 17741 | @need 1500 |
| 17742 | However, to prepare this function definition for Edebug, you must |
| 17743 | first @dfn{instrument} the code using a different command. You can do |
| 17744 | this by positioning your cursor within the definition and typing |
| 17745 | |
| 17746 | @smallexample |
| 17747 | M-x edebug-defun RET |
| 17748 | @end smallexample |
| 17749 | |
| 17750 | @noindent |
| 17751 | This will cause Emacs to load Edebug automatically if it is not |
| 17752 | already loaded, and properly instrument the function. |
| 17753 | |
| 17754 | After instrumenting the function, place your cursor after the |
| 17755 | following expression and type @kbd{C-x C-e} (@code{eval-last-sexp}): |
| 17756 | |
| 17757 | @smallexample |
| 17758 | (triangle-recursively-bugged 3) |
| 17759 | @end smallexample |
| 17760 | |
| 17761 | @noindent |
| 17762 | You will be jumped back to the source for |
| 17763 | @code{triangle-recursively-bugged} and the cursor positioned at the |
| 17764 | beginning of the @code{if} line of the function. Also, you will see |
| 17765 | an arrowhead at the left hand side of that line. The arrowhead marks |
| 17766 | the line where the function is executing. (In the following examples, |
| 17767 | we show the arrowhead with @samp{=>}; in a windowing system, you may |
| 17768 | see the arrowhead as a solid triangle in the window `fringe'.) |
| 17769 | |
| 17770 | @smallexample |
| 17771 | =>@point{}(if (= number 1) |
| 17772 | @end smallexample |
| 17773 | |
| 17774 | @noindent |
| 17775 | @iftex |
| 17776 | In the example, the location of point is displayed with a star, |
| 17777 | @samp{@point{}} (in Info, it is displayed as @samp{-!-}). |
| 17778 | @end iftex |
| 17779 | @ifnottex |
| 17780 | In the example, the location of point is displayed as @samp{@point{}} |
| 17781 | (in a printed book, it is displayed with a five pointed star). |
| 17782 | @end ifnottex |
| 17783 | |
| 17784 | If you now press @key{SPC}, point will move to the next expression to |
| 17785 | be executed; the line will look like this: |
| 17786 | |
| 17787 | @smallexample |
| 17788 | =>(if @point{}(= number 1) |
| 17789 | @end smallexample |
| 17790 | |
| 17791 | @noindent |
| 17792 | As you continue to press @key{SPC}, point will move from expression to |
| 17793 | expression. At the same time, whenever an expression returns a value, |
| 17794 | that value will be displayed in the echo area. For example, after you |
| 17795 | move point past @code{number}, you will see the following: |
| 17796 | |
| 17797 | @smallexample |
| 17798 | Result: 3 = C-c |
| 17799 | @end smallexample |
| 17800 | |
| 17801 | @noindent |
| 17802 | This means the value of @code{number} is 3, which is @sc{ascii} |
| 17803 | `control-c' (the third letter of the alphabet, in case you need to |
| 17804 | know this information). |
| 17805 | |
| 17806 | You can continue moving through the code until you reach the line with |
| 17807 | the error. Before evaluation, that line looks like this: |
| 17808 | |
| 17809 | @smallexample |
| 17810 | => @point{}(1= number))))) ; @r{Error here.} |
| 17811 | @end smallexample |
| 17812 | |
| 17813 | @need 1250 |
| 17814 | @noindent |
| 17815 | When you press @key{SPC} once again, you will produce an error message |
| 17816 | that says: |
| 17817 | |
| 17818 | @smallexample |
| 17819 | Symbol's function definition is void:@: 1= |
| 17820 | @end smallexample |
| 17821 | |
| 17822 | @noindent |
| 17823 | This is the bug. |
| 17824 | |
| 17825 | Press @kbd{q} to quit Edebug. |
| 17826 | |
| 17827 | To remove instrumentation from a function definition, simply |
| 17828 | re-evaluate it with a command that does not instrument it. |
| 17829 | For example, you could place your cursor after the definition's |
| 17830 | closing parenthesis and type @kbd{C-x C-e}. |
| 17831 | |
| 17832 | Edebug does a great deal more than walk with you through a function. |
| 17833 | You can set it so it races through on its own, stopping only at an |
| 17834 | error or at specified stopping points; you can cause it to display the |
| 17835 | changing values of various expressions; you can find out how many |
| 17836 | times a function is called, and more. |
| 17837 | |
| 17838 | Edebug is described in @ref{edebug, , Edebug, elisp, The GNU Emacs |
| 17839 | Lisp Reference Manual}. |
| 17840 | |
| 17841 | @need 1500 |
| 17842 | @node Debugging Exercises, , edebug, Debugging |
| 17843 | @section Debugging Exercises |
| 17844 | |
| 17845 | @itemize @bullet |
| 17846 | @item |
| 17847 | Install the @code{count-words-region} function and then cause it to |
| 17848 | enter the built-in debugger when you call it. Run the command on a |
| 17849 | region containing two words. You will need to press @kbd{d} a |
| 17850 | remarkable number of times. On your system, is a `hook' called after |
| 17851 | the command finishes? (For information on hooks, see @ref{Command |
| 17852 | Overview, , Command Loop Overview, elisp, The GNU Emacs Lisp Reference |
| 17853 | Manual}.) |
| 17854 | |
| 17855 | @item |
| 17856 | Copy @code{count-words-region} into the @file{*scratch*} buffer, |
| 17857 | instrument the function for Edebug, and walk through its execution. |
| 17858 | The function does not need to have a bug, although you can introduce |
| 17859 | one if you wish. If the function lacks a bug, the walk-through |
| 17860 | completes without problems. |
| 17861 | |
| 17862 | @item |
| 17863 | While running Edebug, type @kbd{?} to see a list of all the Edebug commands. |
| 17864 | (The @code{global-edebug-prefix} is usually @kbd{C-x X}, i.e.@: |
| 17865 | @kbd{@key{CTL}-x} followed by an upper case @kbd{X}; use this prefix |
| 17866 | for commands made outside of the Edebug debugging buffer.) |
| 17867 | |
| 17868 | @item |
| 17869 | In the Edebug debugging buffer, use the @kbd{p} |
| 17870 | (@code{edebug-bounce-point}) command to see where in the region the |
| 17871 | @code{count-words-region} is working. |
| 17872 | |
| 17873 | @item |
| 17874 | Move point to some spot further down function and then type the |
| 17875 | @kbd{h} (@code{edebug-goto-here}) command to jump to that location. |
| 17876 | |
| 17877 | @item |
| 17878 | Use the @kbd{t} (@code{edebug-trace-mode}) command to cause Edebug to |
| 17879 | walk through the function on its own; use an upper case @kbd{T} for |
| 17880 | @code{edebug-Trace-fast-mode}. |
| 17881 | |
| 17882 | @item |
| 17883 | Set a breakpoint, then run Edebug in Trace mode until it reaches the |
| 17884 | stopping point. |
| 17885 | @end itemize |
| 17886 | |
| 17887 | @node Conclusion, the-the, Debugging, Top |
| 17888 | @chapter Conclusion |
| 17889 | |
| 17890 | We have now reached the end of this Introduction. You have now |
| 17891 | learned enough about programming in Emacs Lisp to set values, to write |
| 17892 | simple @file{.emacs} files for yourself and your friends, and write |
| 17893 | simple customizations and extensions to Emacs. |
| 17894 | |
| 17895 | This is a place to stop. Or, if you wish, you can now go onward, and |
| 17896 | teach yourself. |
| 17897 | |
| 17898 | You have learned some of the basic nuts and bolts of programming. But |
| 17899 | only some. There are a great many more brackets and hinges that are |
| 17900 | easy to use that we have not touched. |
| 17901 | |
| 17902 | A path you can follow right now lies among the sources to GNU Emacs |
| 17903 | and in |
| 17904 | @ifnotinfo |
| 17905 | @cite{The GNU Emacs Lisp Reference Manual}. |
| 17906 | @end ifnotinfo |
| 17907 | @ifinfo |
| 17908 | @ref{Top, , The GNU Emacs Lisp Reference Manual, elisp, The GNU |
| 17909 | Emacs Lisp Reference Manual}. |
| 17910 | @end ifinfo |
| 17911 | |
| 17912 | The Emacs Lisp sources are an adventure. When you read the sources and |
| 17913 | come across a function or expression that is unfamiliar, you need to |
| 17914 | figure out or find out what it does. |
| 17915 | |
| 17916 | Go to the Reference Manual. It is a thorough, complete, and fairly |
| 17917 | easy-to-read description of Emacs Lisp. It is written not only for |
| 17918 | experts, but for people who know what you know. (The @cite{Reference |
| 17919 | Manual} comes with the standard GNU Emacs distribution. Like this |
| 17920 | introduction, it comes as a Texinfo source file, so you can read it |
| 17921 | on-line and as a typeset, printed book.) |
| 17922 | |
| 17923 | Go to the other on-line help that is part of GNU Emacs: the on-line |
| 17924 | documentation for all functions, and @code{find-tags}, the program |
| 17925 | that takes you to sources. |
| 17926 | |
| 17927 | Here is an example of how I explore the sources. Because of its name, |
| 17928 | @file{simple.el} is the file I looked at first, a long time ago. As |
| 17929 | it happens some of the functions in @file{simple.el} are complicated, |
| 17930 | or at least look complicated at first sight. The @code{open-line} |
| 17931 | function, for example, looks complicated. |
| 17932 | |
| 17933 | You may want to walk through this function slowly, as we did with the |
| 17934 | @code{forward-sentence} function. |
| 17935 | @ifnottex |
| 17936 | (@xref{forward-sentence}.) |
| 17937 | @end ifnottex |
| 17938 | @iftex |
| 17939 | (@xref{forward-sentence, , @code{forward-sentence}}.) |
| 17940 | @end iftex |
| 17941 | Or you may want to skip that function and look at another, such as |
| 17942 | @code{split-line}. You don't need to read all the functions. |
| 17943 | According to @code{count-words-in-defun}, the @code{split-line} |
| 17944 | function contains 27 words and symbols. |
| 17945 | |
| 17946 | Even though it is short, @code{split-line} contains four expressions |
| 17947 | we have not studied: @code{skip-chars-forward}, @code{indent-to}, |
| 17948 | @code{current-column} and @samp{?\n}. |
| 17949 | |
| 17950 | Consider the @code{skip-chars-forward} function. (It is part of the |
| 17951 | function definition for @code{back-to-indentation}, which is shown in |
| 17952 | @ref{Review, , Review}.) |
| 17953 | |
| 17954 | In GNU Emacs, you can find out more about @code{skip-chars-forward} by |
| 17955 | typing @kbd{C-h f} (@code{describe-function}) and the name of the |
| 17956 | function. This gives you the function documentation. |
| 17957 | |
| 17958 | You may be able to guess what is done by a well named function such as |
| 17959 | @code{indent-to}; or you can look it up, too. Incidentally, the |
| 17960 | @code{describe-function} function itself is in @file{help.el}; it is |
| 17961 | one of those long, but decipherable functions. You can look up |
| 17962 | @code{describe-function} using the @kbd{C-h f} command! |
| 17963 | |
| 17964 | In this instance, since the code is Lisp, the @file{*Help*} buffer |
| 17965 | contains the name of the library containing the function's source. |
| 17966 | You can put point over the name of the library and press the RET key, |
| 17967 | which in this situation is bound to @code{help-follow}, and be taken |
| 17968 | directly to the source, in the same way as @kbd{M-.} |
| 17969 | (@code{find-tag}). |
| 17970 | |
| 17971 | The definition for @code{describe-function} illustrates how to |
| 17972 | customize the @code{interactive} expression without using the standard |
| 17973 | character codes; and it shows how to create a temporary buffer. |
| 17974 | |
| 17975 | (The @code{indent-to} function is written in C rather than Emacs Lisp; |
| 17976 | it is a `built-in' function. @code{help-follow} only provides you |
| 17977 | with the documentation of a built-in function; it does not take you to |
| 17978 | the source. But @code{find-tag} will take you to the source, if |
| 17979 | properly set up.) |
| 17980 | |
| 17981 | You can look at a function's source using @code{find-tag}, which is |
| 17982 | bound to @kbd{M-.} Finally, you can find out what the Reference |
| 17983 | Manual has to say by visiting the manual in Info, and typing @kbd{i} |
| 17984 | (@code{Info-index}) and the name of the function, or by looking up |
| 17985 | @code{skip-chars-forward} in the index to a printed copy of the |
| 17986 | manual. |
| 17987 | |
| 17988 | Similarly, you can find out what is meant by @samp{?\n}. You can try |
| 17989 | using @code{Info-index} with @samp{?\n}. It turns out that this |
| 17990 | action won't help; but don't give up. If you search the index for |
| 17991 | @samp{\n} without the @samp{?}, you will be taken directly to the |
| 17992 | relevant section of the manual. (@xref{Character Type, , Character |
| 17993 | Type, elisp, The GNU Emacs Lisp Reference Manual}. @samp{?\n} stands |
| 17994 | for the newline character.) |
| 17995 | |
| 17996 | Other interesting source files include @file{paragraphs.el}, |
| 17997 | @file{loaddefs.el}, and @file{loadup.el}. The @file{paragraphs.el} |
| 17998 | file includes short, easily understood functions as well as longer |
| 17999 | ones. The @file{loaddefs.el} file contains the many standard |
| 18000 | autoloads and many keymaps. I have never looked at it all; only at |
| 18001 | parts. @file{loadup.el} is the file that loads the standard parts of |
| 18002 | Emacs; it tells you a great deal about how Emacs is built. |
| 18003 | (@xref{Building Emacs, , Building Emacs, elisp, The GNU Emacs Lisp |
| 18004 | Reference Manual}, for more about building.) |
| 18005 | |
| 18006 | As I said, you have learned some nuts and bolts; however, and very |
| 18007 | importantly, we have hardly touched major aspects of programming; I |
| 18008 | have said nothing about how to sort information, except to use the |
| 18009 | predefined @code{sort} function; I have said nothing about how to store |
| 18010 | information, except to use variables and lists; I have said nothing |
| 18011 | about how to write programs that write programs. These are topics for |
| 18012 | another, and different kind of book, a different kind of learning. |
| 18013 | |
| 18014 | What you have done is learn enough for much practical work with GNU |
| 18015 | Emacs. What you have done is get started. This is the end of a |
| 18016 | beginning. |
| 18017 | |
| 18018 | @c ================ Appendix ================ |
| 18019 | |
| 18020 | @node the-the, Kill Ring, Conclusion, Top |
| 18021 | @appendix The @code{the-the} Function |
| 18022 | @findex the-the |
| 18023 | @cindex Duplicated words function |
| 18024 | @cindex Words, duplicated |
| 18025 | |
| 18026 | Sometimes when you you write text, you duplicate words---as with ``you |
| 18027 | you'' near the beginning of this sentence. I find that most |
| 18028 | frequently, I duplicate ``the''; hence, I call the function for |
| 18029 | detecting duplicated words, @code{the-the}. |
| 18030 | |
| 18031 | @need 1250 |
| 18032 | As a first step, you could use the following regular expression to |
| 18033 | search for duplicates: |
| 18034 | |
| 18035 | @smallexample |
| 18036 | \\(\\w+[ \t\n]+\\)\\1 |
| 18037 | @end smallexample |
| 18038 | |
| 18039 | @noindent |
| 18040 | This regexp matches one or more word-constituent characters followed |
| 18041 | by one or more spaces, tabs, or newlines. However, it does not detect |
| 18042 | duplicated words on different lines, since the ending of the first |
| 18043 | word, the end of the line, is different from the ending of the second |
| 18044 | word, a space. (For more information about regular expressions, see |
| 18045 | @ref{Regexp Search, , Regular Expression Searches}, as well as |
| 18046 | @ref{Regexps, , Syntax of Regular Expressions, emacs, The GNU Emacs |
| 18047 | Manual}, and @ref{Regular Expressions, , Regular Expressions, elisp, |
| 18048 | The GNU Emacs Lisp Reference Manual}.) |
| 18049 | |
| 18050 | You might try searching just for duplicated word-constituent |
| 18051 | characters but that does not work since the pattern detects doubles |
| 18052 | such as the two occurrences of `th' in `with the'. |
| 18053 | |
| 18054 | Another possible regexp searches for word-constituent characters |
| 18055 | followed by non-word-constituent characters, reduplicated. Here, |
| 18056 | @w{@samp{\\w+}} matches one or more word-constituent characters and |
| 18057 | @w{@samp{\\W*}} matches zero or more non-word-constituent characters. |
| 18058 | |
| 18059 | @smallexample |
| 18060 | \\(\\(\\w+\\)\\W*\\)\\1 |
| 18061 | @end smallexample |
| 18062 | |
| 18063 | @noindent |
| 18064 | Again, not useful. |
| 18065 | |
| 18066 | Here is the pattern that I use. It is not perfect, but good enough. |
| 18067 | @w{@samp{\\b}} matches the empty string, provided it is at the beginning |
| 18068 | or end of a word; @w{@samp{[^@@ \n\t]+}} matches one or more occurrences of |
| 18069 | any characters that are @emph{not} an @@-sign, space, newline, or tab. |
| 18070 | |
| 18071 | @smallexample |
| 18072 | \\b\\([^@@ \n\t]+\\)[ \n\t]+\\1\\b |
| 18073 | @end smallexample |
| 18074 | |
| 18075 | One can write more complicated expressions, but I found that this |
| 18076 | expression is good enough, so I use it. |
| 18077 | |
| 18078 | Here is the @code{the-the} function, as I include it in my |
| 18079 | @file{.emacs} file, along with a handy global key binding: |
| 18080 | |
| 18081 | @smallexample |
| 18082 | @group |
| 18083 | (defun the-the () |
| 18084 | "Search forward for for a duplicated word." |
| 18085 | (interactive) |
| 18086 | (message "Searching for for duplicated words ...") |
| 18087 | (push-mark) |
| 18088 | @end group |
| 18089 | @group |
| 18090 | ;; This regexp is not perfect |
| 18091 | ;; but is fairly good over all: |
| 18092 | (if (re-search-forward |
| 18093 | "\\b\\([^@@ \n\t]+\\)[ \n\t]+\\1\\b" nil 'move) |
| 18094 | (message "Found duplicated word.") |
| 18095 | (message "End of buffer"))) |
| 18096 | @end group |
| 18097 | |
| 18098 | @group |
| 18099 | ;; Bind `the-the' to C-c \ |
| 18100 | (global-set-key "\C-c\\" 'the-the) |
| 18101 | @end group |
| 18102 | @end smallexample |
| 18103 | |
| 18104 | @sp 1 |
| 18105 | Here is test text: |
| 18106 | |
| 18107 | @smallexample |
| 18108 | @group |
| 18109 | one two two three four five |
| 18110 | five six seven |
| 18111 | @end group |
| 18112 | @end smallexample |
| 18113 | |
| 18114 | You can substitute the other regular expressions shown above in the |
| 18115 | function definition and try each of them on this list. |
| 18116 | |
| 18117 | @node Kill Ring, Full Graph, the-the, Top |
| 18118 | @appendix Handling the Kill Ring |
| 18119 | @cindex Kill ring handling |
| 18120 | @cindex Handling the kill ring |
| 18121 | @cindex Ring, making a list like a |
| 18122 | |
| 18123 | The kill ring is a list that is transformed into a ring by the |
| 18124 | workings of the @code{rotate-yank-pointer} function. The @code{yank} |
| 18125 | and @code{yank-pop} commands use the @code{rotate-yank-pointer} |
| 18126 | function. This appendix describes the @code{rotate-yank-pointer} |
| 18127 | function as well as both the @code{yank} and the @code{yank-pop} |
| 18128 | commands. |
| 18129 | |
| 18130 | @menu |
| 18131 | * rotate-yank-pointer:: Move a pointer along a list and around. |
| 18132 | * yank:: Paste a copy of a clipped element. |
| 18133 | * yank-pop:: Insert first element pointed to. |
| 18134 | * ring file:: |
| 18135 | @end menu |
| 18136 | |
| 18137 | @node rotate-yank-pointer, yank, Kill Ring, Kill Ring |
| 18138 | @comment node-name, next, previous, up |
| 18139 | @appendixsec The @code{rotate-yank-pointer} Function |
| 18140 | @findex rotate-yank-pointer |
| 18141 | |
| 18142 | The @code{rotate-yank-pointer} function changes the element in the kill |
| 18143 | ring to which @code{kill-ring-yank-pointer} points. For example, it can |
| 18144 | change @code{kill-ring-yank-pointer} from pointing to the second |
| 18145 | element to point to the third element. |
| 18146 | |
| 18147 | @need 800 |
| 18148 | Here is the code for @code{rotate-yank-pointer}: |
| 18149 | |
| 18150 | @smallexample |
| 18151 | @group |
| 18152 | (defun rotate-yank-pointer (arg) |
| 18153 | "Rotate the yanking point in the kill ring." |
| 18154 | (interactive "p") |
| 18155 | (let ((length (length kill-ring))) |
| 18156 | @end group |
| 18157 | @group |
| 18158 | (if (zerop length) |
| 18159 | ;; @r{then-part} |
| 18160 | (error "Kill ring is empty") |
| 18161 | @end group |
| 18162 | @group |
| 18163 | ;; @r{else-part} |
| 18164 | (setq kill-ring-yank-pointer |
| 18165 | (nthcdr (% (+ arg |
| 18166 | (- length |
| 18167 | (length |
| 18168 | kill-ring-yank-pointer))) |
| 18169 | length) |
| 18170 | kill-ring))))) |
| 18171 | @end group |
| 18172 | @end smallexample |
| 18173 | |
| 18174 | @menu |
| 18175 | * Understanding rotate-yk-ptr:: |
| 18176 | * rotate-yk-ptr body:: The body of @code{rotate-yank-pointer}. |
| 18177 | @end menu |
| 18178 | |
| 18179 | @node Understanding rotate-yk-ptr, rotate-yk-ptr body, rotate-yank-pointer, rotate-yank-pointer |
| 18180 | @ifnottex |
| 18181 | @unnumberedsubsec @code{rotate-yank-pointer} in Outline |
| 18182 | @end ifnottex |
| 18183 | |
| 18184 | The @code{rotate-yank-pointer} function looks complex, but as usual, |
| 18185 | it can be understood by taking it apart piece by piece. First look at |
| 18186 | it in skeletal form: |
| 18187 | |
| 18188 | @smallexample |
| 18189 | @group |
| 18190 | (defun rotate-yank-pointer (arg) |
| 18191 | "Rotate the yanking point in the kill ring." |
| 18192 | (interactive "p") |
| 18193 | (let @var{varlist} |
| 18194 | @var{body}@dots{}) |
| 18195 | @end group |
| 18196 | @end smallexample |
| 18197 | |
| 18198 | This function takes one argument, called @code{arg}. It has a brief |
| 18199 | documentation string; and it is interactive with a small @samp{p}, which |
| 18200 | means that the argument must be a processed prefix passed to the |
| 18201 | function as a number. |
| 18202 | |
| 18203 | The body of the function definition is a @code{let} expression, which |
| 18204 | itself has a body as well as a @var{varlist}. |
| 18205 | |
| 18206 | The @code{let} expression declares a variable that will be only usable |
| 18207 | within the bounds of this function. This variable is called |
| 18208 | @code{length} and is bound to a value that is equal to the number of |
| 18209 | items in the kill ring. This is done by using the function called |
| 18210 | @code{length}. (Note that this function has the same name as the |
| 18211 | variable called @code{length}; but one use of the word is to name the |
| 18212 | function and the other is to name the variable. The two are quite |
| 18213 | distinct. Similarly, an English speaker will distinguish between the |
| 18214 | meanings of the word @samp{ship} when he says: "I must ship this package |
| 18215 | immediately." and "I must get aboard the ship immediately.") |
| 18216 | |
| 18217 | The function @code{length} tells the number of items there are in a list, |
| 18218 | so @code{(length kill-ring)} returns the number of items there are in the |
| 18219 | kill ring. |
| 18220 | |
| 18221 | @node rotate-yk-ptr body, , Understanding rotate-yk-ptr, rotate-yank-pointer |
| 18222 | @comment node-name, next, previous, up |
| 18223 | @appendixsubsec The Body of @code{rotate-yank-pointer} |
| 18224 | |
| 18225 | The body of @code{rotate-yank-pointer} is a @code{let} expression and |
| 18226 | the body of the @code{let} expression is an @code{if} expression. |
| 18227 | |
| 18228 | The purpose of the @code{if} expression is to find out whether there is |
| 18229 | anything in the kill ring. If the kill ring is empty, the @code{error} |
| 18230 | function stops evaluation of the function and prints a message in the |
| 18231 | echo area. On the other hand, if the kill ring has something in it, the |
| 18232 | work of the function is done. |
| 18233 | |
| 18234 | Here is the if-part and then-part of the @code{if} expression: |
| 18235 | |
| 18236 | @findex zerop |
| 18237 | @findex error |
| 18238 | @smallexample |
| 18239 | @group |
| 18240 | (if (zerop length) ; @r{if-part} |
| 18241 | (error "Kill ring is empty") ; @r{then-part} |
| 18242 | @dots{} |
| 18243 | @end group |
| 18244 | @end smallexample |
| 18245 | |
| 18246 | @noindent |
| 18247 | If there is not anything in the kill ring, its length must be zero and |
| 18248 | an error message sent to the user: @samp{Kill ring is empty}. The |
| 18249 | @code{if} expression uses the function @code{zerop} which returns true |
| 18250 | if the value it is testing is zero. When @code{zerop} tests true, the |
| 18251 | then-part of the @code{if} is evaluated. The then-part is a list |
| 18252 | starting with the function @code{error}, which is a function that is |
| 18253 | similar to the @code{message} function (@pxref{message}), in that it |
| 18254 | prints a one-line message in the echo area. However, in addition to |
| 18255 | printing a message, @code{error} also stops evaluation of the function |
| 18256 | within which it is embedded. This means that the rest of the function |
| 18257 | will not be evaluated if the length of the kill ring is zero. |
| 18258 | |
| 18259 | @menu |
| 18260 | * Digression concerning error:: How to mislead humans, but not computers. |
| 18261 | * rotate-yk-ptr else-part:: The else-part of the @code{if} expression. |
| 18262 | * Remainder Function:: The remainder, @code{%}, function. |
| 18263 | * rotate-yk-ptr remainder:: Using @code{%} in @code{rotate-yank-pointer}. |
| 18264 | * kill-rng-yk-ptr last elt:: Pointing to the last element. |
| 18265 | @end menu |
| 18266 | |
| 18267 | @node Digression concerning error, rotate-yk-ptr else-part, rotate-yk-ptr body, rotate-yk-ptr body |
| 18268 | @ifnottex |
| 18269 | @unnumberedsubsubsec Digression about the word `error' |
| 18270 | @end ifnottex |
| 18271 | |
| 18272 | (In my opinion, it is slightly misleading, at least to humans, to use |
| 18273 | the term `error' as the name of the @code{error} function. A better |
| 18274 | term would be `cancel'. Strictly speaking, of course, you cannot |
| 18275 | point to, much less rotate a pointer to a list that has no length, so |
| 18276 | from the point of view of the computer, the word `error' is correct. |
| 18277 | But a human expects to attempt this sort of thing, if only to find out |
| 18278 | whether the kill ring is full or empty. This is an act of |
| 18279 | exploration. |
| 18280 | |
| 18281 | (From the human point of view, the act of exploration and discovery is |
| 18282 | not necessarily an error, and therefore should not be labelled as one, |
| 18283 | even in the bowels of a computer. As it is, the code in Emacs implies |
| 18284 | that a human who is acting virtuously, by exploring his or her |
| 18285 | environment, is making an error. This is bad. Even though the computer |
| 18286 | takes the same steps as it does when there is an `error', a term such as |
| 18287 | `cancel' would have a clearer connotation.) |
| 18288 | |
| 18289 | @node rotate-yk-ptr else-part, Remainder Function, Digression concerning error, rotate-yk-ptr body |
| 18290 | @unnumberedsubsubsec The else-part of the @code{if} expression |
| 18291 | |
| 18292 | The else-part of the @code{if} expression is dedicated to setting the |
| 18293 | value of @code{kill-ring-yank-pointer} when the kill ring has something |
| 18294 | in it. The code looks like this: |
| 18295 | |
| 18296 | @smallexample |
| 18297 | @group |
| 18298 | (setq kill-ring-yank-pointer |
| 18299 | (nthcdr (% (+ arg |
| 18300 | (- length |
| 18301 | (length kill-ring-yank-pointer))) |
| 18302 | length) |
| 18303 | kill-ring))))) |
| 18304 | @end group |
| 18305 | @end smallexample |
| 18306 | |
| 18307 | This needs some examination. Clearly, @code{kill-ring-yank-pointer} |
| 18308 | is being set to be equal to some @sc{cdr} of the kill ring, using the |
| 18309 | @code{nthcdr} function that is described in an earlier section. |
| 18310 | (@xref{copy-region-as-kill}.) But exactly how does it do this? |
| 18311 | |
| 18312 | Before looking at the details of the code let's first consider the |
| 18313 | purpose of the @code{rotate-yank-pointer} function. |
| 18314 | |
| 18315 | The @code{rotate-yank-pointer} function changes what |
| 18316 | @code{kill-ring-yank-pointer} points to. If |
| 18317 | @code{kill-ring-yank-pointer} starts by pointing to the first element |
| 18318 | of a list, a call to @code{rotate-yank-pointer} causes it to point to |
| 18319 | the second element; and if @code{kill-ring-yank-pointer} points to the |
| 18320 | second element, a call to @code{rotate-yank-pointer} causes it to |
| 18321 | point to the third element. (And if @code{rotate-yank-pointer} is |
| 18322 | given an argument greater than 1, it jumps the pointer that many |
| 18323 | elements.) |
| 18324 | |
| 18325 | The @code{rotate-yank-pointer} function uses @code{setq} to reset what |
| 18326 | the @code{kill-ring-yank-pointer} points to. If |
| 18327 | @code{kill-ring-yank-pointer} points to the first element of the kill |
| 18328 | ring, then, in the simplest case, the @code{rotate-yank-pointer} |
| 18329 | function must cause it to point to the second element. Put another |
| 18330 | way, @code{kill-ring-yank-pointer} must be reset to have a value equal |
| 18331 | to the @sc{cdr} of the kill ring. |
| 18332 | |
| 18333 | @need 1250 |
| 18334 | That is, under these circumstances, |
| 18335 | |
| 18336 | @smallexample |
| 18337 | @group |
| 18338 | (setq kill-ring-yank-pointer |
| 18339 | ("some text" "a different piece of text" "yet more text")) |
| 18340 | |
| 18341 | (setq kill-ring |
| 18342 | ("some text" "a different piece of text" "yet more text")) |
| 18343 | @end group |
| 18344 | @end smallexample |
| 18345 | |
| 18346 | @need 800 |
| 18347 | @noindent |
| 18348 | the code should do this: |
| 18349 | |
| 18350 | @smallexample |
| 18351 | (setq kill-ring-yank-pointer (cdr kill-ring)) |
| 18352 | @end smallexample |
| 18353 | |
| 18354 | @need 1000 |
| 18355 | @noindent |
| 18356 | As a result, the @code{kill-ring-yank-pointer} will look like this: |
| 18357 | |
| 18358 | @smallexample |
| 18359 | @group |
| 18360 | kill-ring-yank-pointer |
| 18361 | @result{} ("a different piece of text" "yet more text")) |
| 18362 | @end group |
| 18363 | @end smallexample |
| 18364 | |
| 18365 | The actual @code{setq} expression uses the @code{nthcdr} function to do |
| 18366 | the job. |
| 18367 | |
| 18368 | As we have seen before (@pxref{nthcdr}), the @code{nthcdr} function |
| 18369 | works by repeatedly taking the @sc{cdr} of a list---it takes the |
| 18370 | @sc{cdr} of the @sc{cdr} of the @sc{cdr} @dots{} |
| 18371 | |
| 18372 | @need 800 |
| 18373 | The two following expressions produce the same result: |
| 18374 | |
| 18375 | @smallexample |
| 18376 | @group |
| 18377 | (setq kill-ring-yank-pointer (cdr kill-ring)) |
| 18378 | |
| 18379 | (setq kill-ring-yank-pointer (nthcdr 1 kill-ring)) |
| 18380 | @end group |
| 18381 | @end smallexample |
| 18382 | |
| 18383 | In the @code{rotate-yank-pointer} function, however, the first |
| 18384 | argument to @code{nthcdr} is a rather complex looking expression with |
| 18385 | lots of arithmetic inside of it: |
| 18386 | |
| 18387 | @smallexample |
| 18388 | @group |
| 18389 | (% (+ arg |
| 18390 | (- length |
| 18391 | (length kill-ring-yank-pointer))) |
| 18392 | length) |
| 18393 | @end group |
| 18394 | @end smallexample |
| 18395 | |
| 18396 | As usual, we need to look at the most deeply embedded expression first |
| 18397 | and then work our way towards the light. |
| 18398 | |
| 18399 | The most deeply embedded expression is @code{(length |
| 18400 | kill-ring-yank-pointer)}. This finds the length of the current value of |
| 18401 | the @code{kill-ring-yank-pointer}. (Remember that the |
| 18402 | @code{kill-ring-yank-pointer} is the name of a variable whose value is a |
| 18403 | list.) |
| 18404 | |
| 18405 | @need 800 |
| 18406 | The measurement of the length is inside the expression: |
| 18407 | |
| 18408 | @smallexample |
| 18409 | (- length (length kill-ring-yank-pointer)) |
| 18410 | @end smallexample |
| 18411 | |
| 18412 | @noindent |
| 18413 | In this expression, the first @code{length} is the variable that was |
| 18414 | assigned the length of the kill ring in the @code{let} statement at the |
| 18415 | beginning of the function. (One might think this function would be |
| 18416 | clearer if the variable @code{length} were named |
| 18417 | @code{length-of-kill-ring} instead; but if you look at the text of the |
| 18418 | whole function, you will see that it is so short that naming this |
| 18419 | variable @code{length} is not a bother, unless you are pulling the |
| 18420 | function apart into very tiny pieces as we are doing here.) |
| 18421 | |
| 18422 | So the line @code{(- length (length kill-ring-yank-pointer))} tells the |
| 18423 | difference between the length of the kill ring and the length of the list |
| 18424 | whose name is @code{kill-ring-yank-pointer}. |
| 18425 | |
| 18426 | To see how all this fits into the @code{rotate-yank-pointer} |
| 18427 | function, let's begin by analyzing the case where |
| 18428 | @code{kill-ring-yank-pointer} points to the first element of the kill |
| 18429 | ring, just as @code{kill-ring} does, and see what happens when |
| 18430 | @code{rotate-yank-pointer} is called with an argument of 1. |
| 18431 | |
| 18432 | The variable @code{length} and the value of the expression |
| 18433 | @code{(length kill-ring-yank-pointer)} will be the same since the |
| 18434 | variable @code{length} is the length of the kill ring and the |
| 18435 | @code{kill-ring-yank-pointer} is pointing to the whole kill ring. |
| 18436 | Consequently, the value of |
| 18437 | |
| 18438 | @smallexample |
| 18439 | (- length (length kill-ring-yank-pointer)) |
| 18440 | @end smallexample |
| 18441 | |
| 18442 | @noindent |
| 18443 | will be zero. Since the value of @code{arg} will be 1, this will mean |
| 18444 | that the value of the whole expression |
| 18445 | |
| 18446 | @smallexample |
| 18447 | (+ arg (- length (length kill-ring-yank-pointer))) |
| 18448 | @end smallexample |
| 18449 | |
| 18450 | @noindent |
| 18451 | will be 1. |
| 18452 | |
| 18453 | @need 1200 |
| 18454 | Consequently, the argument to @code{nthcdr} will be found as the result of |
| 18455 | the expression |
| 18456 | |
| 18457 | @smallexample |
| 18458 | (% 1 length) |
| 18459 | @end smallexample |
| 18460 | |
| 18461 | @node Remainder Function, rotate-yk-ptr remainder, rotate-yk-ptr else-part, rotate-yk-ptr body |
| 18462 | @unnumberedsubsubsec The @code{%} remainder function |
| 18463 | |
| 18464 | To understand @code{(% 1 length)}, we need to understand @code{%}. |
| 18465 | According to its documentation (which I just found by typing @kbd{C-h |
| 18466 | f @kbd{%} @key{RET}}), the @code{%} function returns the remainder of |
| 18467 | its first argument divided by its second argument. For example, the |
| 18468 | remainder of 5 divided by 2 is 1. (2 goes into 5 twice with a |
| 18469 | remainder of 1.) |
| 18470 | |
| 18471 | What surprises people who don't often do arithmetic is that a smaller |
| 18472 | number can be divided by a larger number and have a remainder. In the |
| 18473 | example we just used, 5 was divided by 2. We can reverse that and ask, |
| 18474 | what is the result of dividing 2 by 5? If you can use fractions, the |
| 18475 | answer is obviously 2/5 or .4; but if, as here, you can only use whole |
| 18476 | numbers, the result has to be something different. Clearly, 5 can go into |
| 18477 | 2 zero times, but what of the remainder? To see what the answer is, |
| 18478 | consider a case that has to be familiar from childhood: |
| 18479 | |
| 18480 | @itemize @bullet |
| 18481 | @item |
| 18482 | 5 divided by 5 is 1 with a remainder of 0; |
| 18483 | |
| 18484 | @item |
| 18485 | 6 divided by 5 is 1 with a remainder of 1; |
| 18486 | |
| 18487 | @item |
| 18488 | 7 divided by 5 is 1 with a remainder of 2. |
| 18489 | |
| 18490 | @item |
| 18491 | Similarly, 10 divided by 5 is 2 with a remainder of 0; |
| 18492 | |
| 18493 | @item |
| 18494 | 11 divided by 5 is 2 with a remainder of 1; |
| 18495 | |
| 18496 | @item |
| 18497 | 12 divided by 5 is 1 with a remainder of 2. |
| 18498 | @end itemize |
| 18499 | |
| 18500 | @need 1250 |
| 18501 | @noindent |
| 18502 | By considering the cases as parallel, we can see that |
| 18503 | |
| 18504 | @itemize @bullet |
| 18505 | @item |
| 18506 | zero divided by 5 must be zero with a remainder of zero; |
| 18507 | |
| 18508 | @item |
| 18509 | 1 divided by 5 must be zero with a remainder of 1; |
| 18510 | |
| 18511 | @item |
| 18512 | 2 divided by 5 must be zero with a remainder of 2; |
| 18513 | @end itemize |
| 18514 | |
| 18515 | @noindent |
| 18516 | and so on. |
| 18517 | |
| 18518 | @need 1250 |
| 18519 | So, in this code, if the value of @code{length} is 5, then the result of |
| 18520 | evaluating |
| 18521 | |
| 18522 | @smallexample |
| 18523 | (% 1 5) |
| 18524 | @end smallexample |
| 18525 | |
| 18526 | @noindent |
| 18527 | is 1. (I just checked this by placing the cursor after the expression |
| 18528 | and typing @kbd{C-x C-e}. Indeed, 1 is printed in the echo area.) |
| 18529 | |
| 18530 | @need 2000 |
| 18531 | @node rotate-yk-ptr remainder, kill-rng-yk-ptr last elt, Remainder Function, rotate-yk-ptr body |
| 18532 | @unnumberedsubsubsec Using @code{%} in @code{rotate-yank-pointer} |
| 18533 | |
| 18534 | When the @code{kill-ring-yank-pointer} points to the |
| 18535 | beginning of the kill ring, and the argument passed to |
| 18536 | @code{rotate-yank-pointer} is 1, the @code{%} expression returns 1: |
| 18537 | |
| 18538 | @smallexample |
| 18539 | @group |
| 18540 | (- length (length kill-ring-yank-pointer)) |
| 18541 | @result{} 0 |
| 18542 | @end group |
| 18543 | @end smallexample |
| 18544 | |
| 18545 | @need 1250 |
| 18546 | @noindent |
| 18547 | therefore, |
| 18548 | |
| 18549 | @smallexample |
| 18550 | @group |
| 18551 | (+ arg (- length (length kill-ring-yank-pointer))) |
| 18552 | @result{} 1 |
| 18553 | @end group |
| 18554 | @end smallexample |
| 18555 | |
| 18556 | @need 1250 |
| 18557 | @noindent |
| 18558 | and consequently: |
| 18559 | |
| 18560 | @smallexample |
| 18561 | @group |
| 18562 | (% (+ arg (- length (length kill-ring-yank-pointer))) |
| 18563 | length) |
| 18564 | @result{} 1 |
| 18565 | @end group |
| 18566 | @end smallexample |
| 18567 | |
| 18568 | @noindent |
| 18569 | regardless of the value of @code{length}. |
| 18570 | |
| 18571 | @need 1250 |
| 18572 | @noindent |
| 18573 | As a result of this, the @code{setq kill-ring-yank-pointer} expression |
| 18574 | simplifies to: |
| 18575 | |
| 18576 | @smallexample |
| 18577 | (setq kill-ring-yank-pointer (nthcdr 1 kill-ring)) |
| 18578 | @end smallexample |
| 18579 | |
| 18580 | @noindent |
| 18581 | What it does is now easy to understand. Instead of pointing as it did |
| 18582 | to the first element of the kill ring, the |
| 18583 | @code{kill-ring-yank-pointer} is set to point to the second element. |
| 18584 | |
| 18585 | Clearly, if the argument passed to @code{rotate-yank-pointer} is two, then |
| 18586 | the @code{kill-ring-yank-pointer} is set to @code{(nthcdr 2 kill-ring)}; |
| 18587 | and so on for different values of the argument. |
| 18588 | |
| 18589 | Similarly, if the @code{kill-ring-yank-pointer} starts out pointing to |
| 18590 | the second element of the kill ring, its length is shorter than the |
| 18591 | length of the kill ring by 1, so the computation of the remainder is |
| 18592 | based on the expression @code{(% (+ arg 1) length)}. This means that |
| 18593 | the @code{kill-ring-yank-pointer} is moved from the second element of |
| 18594 | the kill ring to the third element if the argument passed to |
| 18595 | @code{rotate-yank-pointer} is 1. |
| 18596 | |
| 18597 | @node kill-rng-yk-ptr last elt, , rotate-yk-ptr remainder, rotate-yk-ptr body |
| 18598 | @unnumberedsubsubsec Pointing to the last element |
| 18599 | |
| 18600 | The final question is, what happens if the @code{kill-ring-yank-pointer} |
| 18601 | is set to the @emph{last} element of the kill ring? Will a call to |
| 18602 | @code{rotate-yank-pointer} mean that nothing more can be taken from the |
| 18603 | kill ring? The answer is no. What happens is different and useful. |
| 18604 | The @code{kill-ring-yank-pointer} is set to point to the beginning of |
| 18605 | the kill ring instead. |
| 18606 | |
| 18607 | Let's see how this works by looking at the code, assuming the length of the |
| 18608 | kill ring is 5 and the argument passed to @code{rotate-yank-pointer} is 1. |
| 18609 | When the @code{kill-ring-yank-pointer} points to the last element of |
| 18610 | the kill ring, its length is 1. The code looks like this: |
| 18611 | |
| 18612 | @smallexample |
| 18613 | (% (+ arg (- length (length kill-ring-yank-pointer))) length) |
| 18614 | @end smallexample |
| 18615 | |
| 18616 | @need 1250 |
| 18617 | When the variables are replaced by their numeric values, the expression |
| 18618 | looks like this: |
| 18619 | |
| 18620 | @smallexample |
| 18621 | (% (+ 1 (- 5 1)) 5) |
| 18622 | @end smallexample |
| 18623 | |
| 18624 | @noindent |
| 18625 | This expression can be evaluated by looking at the most embedded inner |
| 18626 | expression first and working outwards: The value of @code{(- 5 1)} is 4; |
| 18627 | the sum of @code{(+ 1 4)} is 5; and the remainder of dividing 5 by 5 is |
| 18628 | zero. So what @code{rotate-yank-pointer} will do is |
| 18629 | |
| 18630 | @smallexample |
| 18631 | (setq kill-ring-yank-pointer (nthcdr 0 kill-ring)) |
| 18632 | @end smallexample |
| 18633 | |
| 18634 | @noindent |
| 18635 | which will set the @code{kill-ring-yank-pointer} to point to the beginning |
| 18636 | of the kill ring. |
| 18637 | |
| 18638 | So what happens with successive calls to @code{rotate-yank-pointer} is that |
| 18639 | it moves the @code{kill-ring-yank-pointer} from element to element in the |
| 18640 | kill ring until it reaches the end; then it jumps back to the beginning. |
| 18641 | And this is why the kill ring is called a ring, since by jumping back to |
| 18642 | the beginning, it is as if the list has no end! (And what is a ring, but |
| 18643 | an entity with no end?) |
| 18644 | |
| 18645 | @node yank, yank-pop, rotate-yank-pointer, Kill Ring |
| 18646 | @comment node-name, next, previous, up |
| 18647 | @appendixsec @code{yank} |
| 18648 | @findex yank |
| 18649 | |
| 18650 | After learning about @code{rotate-yank-pointer}, the code for the |
| 18651 | @code{yank} function is almost easy. It has only one tricky part, which is |
| 18652 | the computation of the argument to be passed to @code{rotate-yank-pointer}. |
| 18653 | |
| 18654 | @need 1250 |
| 18655 | The code looks like this: |
| 18656 | |
| 18657 | @smallexample |
| 18658 | @group |
| 18659 | (defun yank (&optional arg) |
| 18660 | "Reinsert the last stretch of killed text. |
| 18661 | More precisely, reinsert the stretch of killed text most |
| 18662 | recently killed OR yanked. |
| 18663 | With just C-U as argument, same but put point in front |
| 18664 | (and mark at end). With argument n, reinsert the nth |
| 18665 | most recently killed stretch of killed text. |
| 18666 | See also the command \\[yank-pop]." |
| 18667 | @end group |
| 18668 | @group |
| 18669 | |
| 18670 | (interactive "*P") |
| 18671 | (rotate-yank-pointer (if (listp arg) 0 |
| 18672 | (if (eq arg '-) -1 |
| 18673 | (1- arg)))) |
| 18674 | (push-mark (point)) |
| 18675 | (insert (car kill-ring-yank-pointer)) |
| 18676 | (if (consp arg) |
| 18677 | (exchange-point-and-mark))) |
| 18678 | @end group |
| 18679 | @end smallexample |
| 18680 | |
| 18681 | Glancing over this code, we can understand the last few lines readily |
| 18682 | enough. The mark is pushed, that is, remembered; then the first element |
| 18683 | (the @sc{car}) of what the @code{kill-ring-yank-pointer} points to is |
| 18684 | inserted; and then, if the argument passed the function is a |
| 18685 | @code{cons}, point and mark are exchanged so the point is put in the |
| 18686 | front of the inserted text rather than at the end. This option is |
| 18687 | explained in the documentation. The function itself is interactive with |
| 18688 | @code{"*P"}. This means it will not work on a read-only buffer, and that |
| 18689 | the unprocessed prefix argument is passed to the function. |
| 18690 | |
| 18691 | @menu |
| 18692 | * rotate-yk-ptr arg:: Pass the argument to @code{rotate-yank-pointer}. |
| 18693 | * rotate-yk-ptr negative arg:: Pass a negative argument. |
| 18694 | @end menu |
| 18695 | |
| 18696 | @node rotate-yk-ptr arg, rotate-yk-ptr negative arg, yank, yank |
| 18697 | @unnumberedsubsubsec Passing the argument |
| 18698 | |
| 18699 | The hard part of @code{yank} is understanding the computation that |
| 18700 | determines the value of the argument passed to |
| 18701 | @code{rotate-yank-pointer}. Fortunately, it is not so difficult as it |
| 18702 | looks at first sight. |
| 18703 | |
| 18704 | What happens is that the result of evaluating one or both of the |
| 18705 | @code{if} expressions will be a number and that number will be the |
| 18706 | argument passed to @code{rotate-yank-pointer}. |
| 18707 | |
| 18708 | @need 1250 |
| 18709 | Laid out with comments, the code looks like this: |
| 18710 | |
| 18711 | @smallexample |
| 18712 | @group |
| 18713 | (if (listp arg) ; @r{if-part} |
| 18714 | 0 ; @r{then-part} |
| 18715 | (if (eq arg '-) ; @r{else-part, inner if} |
| 18716 | -1 ; @r{inner if's then-part} |
| 18717 | (1- arg)))) ; @r{inner if's else-part} |
| 18718 | @end group |
| 18719 | @end smallexample |
| 18720 | |
| 18721 | @noindent |
| 18722 | This code consists of two @code{if} expression, one the else-part of |
| 18723 | the other. |
| 18724 | |
| 18725 | The first or outer @code{if} expression tests whether the argument |
| 18726 | passed to @code{yank} is a list. Oddly enough, this will be true if |
| 18727 | @code{yank} is called without an argument---because then it will be |
| 18728 | passed the value of @code{nil} for the optional argument and an |
| 18729 | evaluation of @code{(listp nil)} returns true! So, if no argument is |
| 18730 | passed to @code{yank}, the argument passed to |
| 18731 | @code{rotate-yank-pointer} inside of @code{yank} is zero. This means |
| 18732 | the pointer is not moved and the first element to which |
| 18733 | @code{kill-ring-yank-pointer} points is inserted, as we expect. |
| 18734 | Similarly, if the argument for @code{yank} is @kbd{C-u}, this will be |
| 18735 | read as a list, so again, a zero will be passed to |
| 18736 | @code{rotate-yank-pointer}. (@kbd{C-u} produces an unprocessed prefix |
| 18737 | argument of @code{(4)}, which is a list of one element.) At the same |
| 18738 | time, later in the function, this argument will be read as a |
| 18739 | @code{cons} so point will be put in the front and mark at the end of |
| 18740 | the insertion. (The @code{P} argument to @code{interactive} is |
| 18741 | designed to provide these values for the case when an optional |
| 18742 | argument is not provided or when it is @kbd{C-u}.) |
| 18743 | |
| 18744 | The then-part of the outer @code{if} expression handles the case when |
| 18745 | there is no argument or when it is @kbd{C-u}. The else-part handles the |
| 18746 | other situations. The else-part is itself another @code{if} expression. |
| 18747 | |
| 18748 | The inner @code{if} expression tests whether the argument is a minus |
| 18749 | sign. (This is done by pressing the @key{META} and @kbd{-} keys at the |
| 18750 | same time, or the @key{ESC} key and then the @kbd{-} key). In this |
| 18751 | case, the @code{rotate-yank-pointer} function is passed @kbd{-1} as an |
| 18752 | argument. This moves the @code{kill-ring-yank-pointer} backwards, which |
| 18753 | is what is desired. |
| 18754 | |
| 18755 | If the true-or-false-test of the inner @code{if} expression is false |
| 18756 | (that is, if the argument is not a minus sign), the else-part of the |
| 18757 | expression is evaluated. This is the expression @code{(1- arg)}. |
| 18758 | Because of the two @code{if} expressions, it will only occur when the |
| 18759 | argument is a positive number or when it is a negative number (not |
| 18760 | just a minus sign on its own). What @code{(1- arg)} does is decrement |
| 18761 | the number and return it. (The @code{1-} function subtracts one from |
| 18762 | its argument.) This means that if the argument to |
| 18763 | @code{rotate-yank-pointer} is 1, it is reduced to zero, which means |
| 18764 | the first element to which @code{kill-ring-yank-pointer} points is |
| 18765 | yanked back, as you would expect. |
| 18766 | |
| 18767 | @node rotate-yk-ptr negative arg, , rotate-yk-ptr arg, yank |
| 18768 | @unnumberedsubsubsec Passing a negative argument |
| 18769 | |
| 18770 | Finally, the question arises, what happens if either the remainder |
| 18771 | function, @code{%}, or the @code{nthcdr} function is passed a negative |
| 18772 | argument, as they quite well may? |
| 18773 | |
| 18774 | The answers can be found by a quick test. When @code{(% -1 5)} is |
| 18775 | evaluated, a negative number is returned; and if @code{nthcdr} is |
| 18776 | called with a negative number, it returns the same value as if it were |
| 18777 | called with a first argument of zero. This can be seen by evaluating |
| 18778 | the following code. |
| 18779 | |
| 18780 | Here the @samp{@result{}} points to the result of evaluating the code |
| 18781 | preceding it. This was done by positioning the cursor after the code |
| 18782 | and typing @kbd{C-x C-e} (@code{eval-last-sexp}) in the usual fashion. |
| 18783 | You can do this if you are reading this in Info inside of GNU Emacs. |
| 18784 | |
| 18785 | @smallexample |
| 18786 | @group |
| 18787 | (% -1 5) |
| 18788 | @result{} -1 |
| 18789 | @end group |
| 18790 | |
| 18791 | @group |
| 18792 | (setq animals '(cats dogs elephants)) |
| 18793 | @result{} (cats dogs elephants) |
| 18794 | @end group |
| 18795 | |
| 18796 | @group |
| 18797 | (nthcdr 1 animals) |
| 18798 | @result{} (dogs elephants) |
| 18799 | @end group |
| 18800 | |
| 18801 | @group |
| 18802 | (nthcdr 0 animals) |
| 18803 | @result{} (cats dogs elephants) |
| 18804 | @end group |
| 18805 | |
| 18806 | @group |
| 18807 | (nthcdr -1 animals) |
| 18808 | @result{} (cats dogs elephants) |
| 18809 | @end group |
| 18810 | @end smallexample |
| 18811 | |
| 18812 | So, if a minus sign or a negative number is passed to @code{yank}, the |
| 18813 | @code{kill-ring-yank-point} is rotated backwards until it reaches the |
| 18814 | beginning of the list. Then it stays there. Unlike the other case, |
| 18815 | when it jumps from the end of the list to the beginning of the list, |
| 18816 | making a ring, it stops. This makes sense. You often want to get back |
| 18817 | to the most recently clipped out piece of text, but you don't usually |
| 18818 | want to insert text from as many as thirty kill commands ago. So you |
| 18819 | need to work through the ring to get to the end, but won't cycle around |
| 18820 | it inadvertently if you are trying to come back to the beginning. |
| 18821 | |
| 18822 | Incidentally, any number passed to @code{yank} with a minus sign |
| 18823 | preceding it will be treated as @minus{}1. This is evidently a |
| 18824 | simplification for writing the program. You don't need to jump back |
| 18825 | towards the beginning of the kill ring more than one place at a time |
| 18826 | and doing this is easier than writing a function to determine the |
| 18827 | magnitude of the number that follows the minus sign. |
| 18828 | |
| 18829 | @node yank-pop, ring file, yank, Kill Ring |
| 18830 | @comment node-name, next, previous, up |
| 18831 | @appendixsec @code{yank-pop} |
| 18832 | @findex yank-pop |
| 18833 | |
| 18834 | After understanding @code{yank}, the @code{yank-pop} function is easy. |
| 18835 | Leaving out the documentation to save space, it looks like this: |
| 18836 | |
| 18837 | @smallexample |
| 18838 | @group |
| 18839 | (defun yank-pop (arg) |
| 18840 | (interactive "*p") |
| 18841 | (if (not (eq last-command 'yank)) |
| 18842 | (error "Previous command was not a yank")) |
| 18843 | @end group |
| 18844 | @group |
| 18845 | (setq this-command 'yank) |
| 18846 | (let ((before (< (point) (mark)))) |
| 18847 | (delete-region (point) (mark)) |
| 18848 | (rotate-yank-pointer arg) |
| 18849 | @end group |
| 18850 | @group |
| 18851 | (set-mark (point)) |
| 18852 | (insert (car kill-ring-yank-pointer)) |
| 18853 | (if before (exchange-point-and-mark)))) |
| 18854 | @end group |
| 18855 | @end smallexample |
| 18856 | |
| 18857 | The function is interactive with a small @samp{p} so the prefix |
| 18858 | argument is processed and passed to the function. The command can |
| 18859 | only be used after a previous yank; otherwise an error message is |
| 18860 | sent. This check uses the variable @code{last-command} which is |
| 18861 | discussed elsewhere. (@xref{copy-region-as-kill}.) |
| 18862 | |
| 18863 | The @code{let} clause sets the variable @code{before} to true or false |
| 18864 | depending whether point is before or after mark and then the region |
| 18865 | between point and mark is deleted. This is the region that was just |
| 18866 | inserted by the previous yank and it is this text that will be |
| 18867 | replaced. Next the @code{kill-ring-yank-pointer} is rotated so that |
| 18868 | the previously inserted text is not reinserted yet again. Mark is set |
| 18869 | at the beginning of the place the new text will be inserted and then |
| 18870 | the first element to which @code{kill-ring-yank-pointer} points is |
| 18871 | inserted. This leaves point after the new text. If in the previous |
| 18872 | yank, point was left before the inserted text, point and mark are now |
| 18873 | exchanged so point is again left in front of the newly inserted text. |
| 18874 | That is all there is to it! |
| 18875 | |
| 18876 | @node ring file, , yank-pop, Kill Ring |
| 18877 | @comment node-name, next, previous, up |
| 18878 | @appendixsec The @file{ring.el} File |
| 18879 | @cindex @file{ring.el} file |
| 18880 | |
| 18881 | Interestingly, GNU Emacs posses a file called @file{ring.el} that |
| 18882 | provides many of the features we just discussed. But functions such |
| 18883 | as @code{kill-ring-yank-pointer} do not use this library, possibly |
| 18884 | because they were written earlier. |
| 18885 | |
| 18886 | @node Full Graph, Free Software and Free Manuals, Kill Ring, Top |
| 18887 | @appendix A Graph with Labelled Axes |
| 18888 | |
| 18889 | Printed axes help you understand a graph. They convey scale. In an |
| 18890 | earlier chapter (@pxref{Readying a Graph, , Readying a Graph}), we |
| 18891 | wrote the code to print the body of a graph. Here we write the code |
| 18892 | for printing and labelling vertical and horizontal axes, along with the |
| 18893 | body itself. |
| 18894 | |
| 18895 | @menu |
| 18896 | * Labelled Example:: |
| 18897 | * print-graph Varlist:: @code{let} expression in @code{print-graph}. |
| 18898 | * print-Y-axis:: Print a label for the vertical axis. |
| 18899 | * print-X-axis:: Print a horizontal label. |
| 18900 | * Print Whole Graph:: The function to print a complete graph. |
| 18901 | @end menu |
| 18902 | |
| 18903 | @node Labelled Example, print-graph Varlist, Full Graph, Full Graph |
| 18904 | @ifnottex |
| 18905 | @unnumberedsec Labelled Example Graph |
| 18906 | @end ifnottex |
| 18907 | |
| 18908 | Since insertions fill a buffer to the right and below point, the new |
| 18909 | graph printing function should first print the Y or vertical axis, |
| 18910 | then the body of the graph, and finally the X or horizontal axis. |
| 18911 | This sequence lays out for us the contents of the function: |
| 18912 | |
| 18913 | @enumerate |
| 18914 | @item |
| 18915 | Set up code. |
| 18916 | |
| 18917 | @item |
| 18918 | Print Y axis. |
| 18919 | |
| 18920 | @item |
| 18921 | Print body of graph. |
| 18922 | |
| 18923 | @item |
| 18924 | Print X axis. |
| 18925 | @end enumerate |
| 18926 | |
| 18927 | @need 800 |
| 18928 | Here is an example of how a finished graph should look: |
| 18929 | |
| 18930 | @smallexample |
| 18931 | @group |
| 18932 | 10 - |
| 18933 | * |
| 18934 | * * |
| 18935 | * ** |
| 18936 | * *** |
| 18937 | 5 - * ******* |
| 18938 | * *** ******* |
| 18939 | ************* |
| 18940 | *************** |
| 18941 | 1 - **************** |
| 18942 | | | | | |
| 18943 | 1 5 10 15 |
| 18944 | @end group |
| 18945 | @end smallexample |
| 18946 | |
| 18947 | @noindent |
| 18948 | In this graph, both the vertical and the horizontal axes are labelled |
| 18949 | with numbers. However, in some graphs, the horizontal axis is time |
| 18950 | and would be better labelled with months, like this: |
| 18951 | |
| 18952 | @smallexample |
| 18953 | @group |
| 18954 | 5 - * |
| 18955 | * ** * |
| 18956 | ******* |
| 18957 | ********** ** |
| 18958 | 1 - ************** |
| 18959 | | ^ | |
| 18960 | Jan June Jan |
| 18961 | @end group |
| 18962 | @end smallexample |
| 18963 | |
| 18964 | Indeed, with a little thought, we can easily come up with a variety of |
| 18965 | vertical and horizontal labelling schemes. Our task could become |
| 18966 | complicated. But complications breed confusion. Rather than permit |
| 18967 | this, it is better choose a simple labelling scheme for our first |
| 18968 | effort, and to modify or replace it later. |
| 18969 | |
| 18970 | @need 1200 |
| 18971 | These considerations suggest the following outline for the |
| 18972 | @code{print-graph} function: |
| 18973 | |
| 18974 | @smallexample |
| 18975 | @group |
| 18976 | (defun print-graph (numbers-list) |
| 18977 | "@var{documentation}@dots{}" |
| 18978 | (let ((height @dots{} |
| 18979 | @dots{})) |
| 18980 | @end group |
| 18981 | @group |
| 18982 | (print-Y-axis height @dots{} ) |
| 18983 | (graph-body-print numbers-list) |
| 18984 | (print-X-axis @dots{} ))) |
| 18985 | @end group |
| 18986 | @end smallexample |
| 18987 | |
| 18988 | We can work on each part of the @code{print-graph} function definition |
| 18989 | in turn. |
| 18990 | |
| 18991 | @node print-graph Varlist, print-Y-axis, Labelled Example, Full Graph |
| 18992 | @comment node-name, next, previous, up |
| 18993 | @appendixsec The @code{print-graph} Varlist |
| 18994 | @cindex @code{print-graph} varlist |
| 18995 | |
| 18996 | In writing the @code{print-graph} function, the first task is to write |
| 18997 | the varlist in the @code{let} expression. (We will leave aside for the |
| 18998 | moment any thoughts about making the function interactive or about the |
| 18999 | contents of its documentation string.) |
| 19000 | |
| 19001 | The varlist should set several values. Clearly, the top of the label |
| 19002 | for the vertical axis must be at least the height of the graph, which |
| 19003 | means that we must obtain this information here. Note that the |
| 19004 | @code{print-graph-body} function also requires this information. There |
| 19005 | is no reason to calculate the height of the graph in two different |
| 19006 | places, so we should change @code{print-graph-body} from the way we |
| 19007 | defined it earlier to take advantage of the calculation. |
| 19008 | |
| 19009 | Similarly, both the function for printing the X axis labels and the |
| 19010 | @code{print-graph-body} function need to learn the value of the width of |
| 19011 | each symbol. We can perform the calculation here and change the |
| 19012 | definition for @code{print-graph-body} from the way we defined it in the |
| 19013 | previous chapter. |
| 19014 | |
| 19015 | The length of the label for the horizontal axis must be at least as long |
| 19016 | as the graph. However, this information is used only in the function |
| 19017 | that prints the horizontal axis, so it does not need to be calculated here. |
| 19018 | |
| 19019 | These thoughts lead us directly to the following form for the varlist |
| 19020 | in the @code{let} for @code{print-graph}: |
| 19021 | |
| 19022 | @smallexample |
| 19023 | @group |
| 19024 | (let ((height (apply 'max numbers-list)) ; @r{First version.} |
| 19025 | (symbol-width (length graph-blank))) |
| 19026 | @end group |
| 19027 | @end smallexample |
| 19028 | |
| 19029 | @noindent |
| 19030 | As we shall see, this expression is not quite right. |
| 19031 | |
| 19032 | @need 2000 |
| 19033 | @node print-Y-axis, print-X-axis, print-graph Varlist, Full Graph |
| 19034 | @comment node-name, next, previous, up |
| 19035 | @appendixsec The @code{print-Y-axis} Function |
| 19036 | @cindex Axis, print vertical |
| 19037 | @cindex Y axis printing |
| 19038 | @cindex Vertical axis printing |
| 19039 | @cindex Print vertical axis |
| 19040 | |
| 19041 | The job of the @code{print-Y-axis} function is to print a label for |
| 19042 | the vertical axis that looks like this: |
| 19043 | |
| 19044 | @smallexample |
| 19045 | @group |
| 19046 | 10 - |
| 19047 | |
| 19048 | |
| 19049 | |
| 19050 | |
| 19051 | 5 - |
| 19052 | |
| 19053 | |
| 19054 | |
| 19055 | 1 - |
| 19056 | @end group |
| 19057 | @end smallexample |
| 19058 | |
| 19059 | @noindent |
| 19060 | The function should be passed the height of the graph, and then should |
| 19061 | construct and insert the appropriate numbers and marks. |
| 19062 | |
| 19063 | It is easy enough to see in the figure what the Y axis label should |
| 19064 | look like; but to say in words, and then to write a function |
| 19065 | definition to do the job is another matter. It is not quite true to |
| 19066 | say that we want a number and a tic every five lines: there are only |
| 19067 | three lines between the @samp{1} and the @samp{5} (lines 2, 3, and 4), |
| 19068 | but four lines between the @samp{5} and the @samp{10} (lines 6, 7, 8, |
| 19069 | and 9). It is better to say that we want a number and a tic mark on |
| 19070 | the base line (number 1) and then that we want a number and a tic on |
| 19071 | the fifth line from the bottom and on every line that is a multiple of |
| 19072 | five. |
| 19073 | |
| 19074 | @menu |
| 19075 | * Height of label:: What height for the Y axis? |
| 19076 | * Compute a Remainder:: How to compute the remainder of a division. |
| 19077 | * Y Axis Element:: Construct a line for the Y axis. |
| 19078 | * Y-axis-column:: Generate a list of Y axis labels. |
| 19079 | * print-Y-axis Penultimate:: A not quite final version. |
| 19080 | @end menu |
| 19081 | |
| 19082 | @node Height of label, Compute a Remainder, print-Y-axis, print-Y-axis |
| 19083 | @ifnottex |
| 19084 | @unnumberedsubsec What height should the label be? |
| 19085 | @end ifnottex |
| 19086 | |
| 19087 | The next issue is what height the label should be? Suppose the maximum |
| 19088 | height of tallest column of the graph is seven. Should the highest |
| 19089 | label on the Y axis be @samp{5 -}, and should the graph stick up above |
| 19090 | the label? Or should the highest label be @samp{7 -}, and mark the peak |
| 19091 | of the graph? Or should the highest label be @code{10 -}, which is a |
| 19092 | multiple of five, and be higher than the topmost value of the graph? |
| 19093 | |
| 19094 | The latter form is preferred. Most graphs are drawn within rectangles |
| 19095 | whose sides are an integral number of steps long---5, 10, 15, and so |
| 19096 | on for a step distance of five. But as soon as we decide to use a |
| 19097 | step height for the vertical axis, we discover that the simple |
| 19098 | expression in the varlist for computing the height is wrong. The |
| 19099 | expression is @code{(apply 'max numbers-list)}. This returns the |
| 19100 | precise height, not the maximum height plus whatever is necessary to |
| 19101 | round up to the nearest multiple of five. A more complex expression |
| 19102 | is required. |
| 19103 | |
| 19104 | As usual in cases like this, a complex problem becomes simpler if it is |
| 19105 | divided into several smaller problems. |
| 19106 | |
| 19107 | First, consider the case when the highest value of the graph is an |
| 19108 | integral multiple of five---when it is 5, 10, 15, or some higher |
| 19109 | multiple of five. We can use this value as the Y axis height. |
| 19110 | |
| 19111 | A fairly simply way to determine whether a number is a multiple of |
| 19112 | five is to divide it by five and see if the division results in a |
| 19113 | remainder. If there is no remainder, the number is a multiple of |
| 19114 | five. Thus, seven divided by five has a remainder of two, and seven |
| 19115 | is not an integral multiple of five. Put in slightly different |
| 19116 | language, more reminiscent of the classroom, five goes into seven |
| 19117 | once, with a remainder of two. However, five goes into ten twice, |
| 19118 | with no remainder: ten is an integral multiple of five. |
| 19119 | |
| 19120 | @node Compute a Remainder, Y Axis Element, Height of label, print-Y-axis |
| 19121 | @appendixsubsec Side Trip: Compute a Remainder |
| 19122 | |
| 19123 | @findex % @r{(remainder function)} |
| 19124 | @cindex Remainder function, @code{%} |
| 19125 | In Lisp, the function for computing a remainder is @code{%}. The |
| 19126 | function returns the remainder of its first argument divided by its |
| 19127 | second argument. As it happens, @code{%} is a function in Emacs Lisp |
| 19128 | that you cannot discover using @code{apropos}: you find nothing if you |
| 19129 | type @kbd{M-x apropos @key{RET} remainder @key{RET}}. The only way to |
| 19130 | learn of the existence of @code{%} is to read about it in a book such |
| 19131 | as this or in the Emacs Lisp sources. The @code{%} function is used |
| 19132 | in the code for @code{rotate-yank-pointer}, which is described in an |
| 19133 | appendix. (@xref{rotate-yk-ptr body, , The Body of |
| 19134 | @code{rotate-yank-pointer}}.) |
| 19135 | |
| 19136 | You can try the @code{%} function by evaluating the following two |
| 19137 | expressions: |
| 19138 | |
| 19139 | @smallexample |
| 19140 | @group |
| 19141 | (% 7 5) |
| 19142 | |
| 19143 | (% 10 5) |
| 19144 | @end group |
| 19145 | @end smallexample |
| 19146 | |
| 19147 | @noindent |
| 19148 | The first expression returns 2 and the second expression returns 0. |
| 19149 | |
| 19150 | To test whether the returned value is zero or some other number, we |
| 19151 | can use the @code{zerop} function. This function returns @code{t} if |
| 19152 | its argument, which must be a number, is zero. |
| 19153 | |
| 19154 | @smallexample |
| 19155 | @group |
| 19156 | (zerop (% 7 5)) |
| 19157 | @result{} nil |
| 19158 | |
| 19159 | (zerop (% 10 5)) |
| 19160 | @result{} t |
| 19161 | @end group |
| 19162 | @end smallexample |
| 19163 | |
| 19164 | Thus, the following expression will return @code{t} if the height |
| 19165 | of the graph is evenly divisible by five: |
| 19166 | |
| 19167 | @smallexample |
| 19168 | (zerop (% height 5)) |
| 19169 | @end smallexample |
| 19170 | |
| 19171 | @noindent |
| 19172 | (The value of @code{height}, of course, can be found from @code{(apply |
| 19173 | 'max numbers-list)}.) |
| 19174 | |
| 19175 | On the other hand, if the value of @code{height} is not a multiple of |
| 19176 | five, we want to reset the value to the next higher multiple of five. |
| 19177 | This is straightforward arithmetic using functions with which we are |
| 19178 | already familiar. First, we divide the value of @code{height} by five |
| 19179 | to determine how many times five goes into the number. Thus, five |
| 19180 | goes into twelve twice. If we add one to this quotient and multiply by |
| 19181 | five, we will obtain the value of the next multiple of five that is |
| 19182 | larger than the height. Five goes into twelve twice. Add one to two, |
| 19183 | and multiply by five; the result is fifteen, which is the next multiple |
| 19184 | of five that is higher than twelve. The Lisp expression for this is: |
| 19185 | |
| 19186 | @smallexample |
| 19187 | (* (1+ (/ height 5)) 5) |
| 19188 | @end smallexample |
| 19189 | |
| 19190 | @noindent |
| 19191 | For example, if you evaluate the following, the result is 15: |
| 19192 | |
| 19193 | @smallexample |
| 19194 | (* (1+ (/ 12 5)) 5) |
| 19195 | @end smallexample |
| 19196 | |
| 19197 | All through this discussion, we have been using `five' as the value |
| 19198 | for spacing labels on the Y axis; but we may want to use some other |
| 19199 | value. For generality, we should replace `five' with a variable to |
| 19200 | which we can assign a value. The best name I can think of for this |
| 19201 | variable is @code{Y-axis-label-spacing}. |
| 19202 | |
| 19203 | @need 1250 |
| 19204 | Using this term, and an @code{if} expression, we produce the |
| 19205 | following: |
| 19206 | |
| 19207 | @smallexample |
| 19208 | @group |
| 19209 | (if (zerop (% height Y-axis-label-spacing)) |
| 19210 | height |
| 19211 | ;; @r{else} |
| 19212 | (* (1+ (/ height Y-axis-label-spacing)) |
| 19213 | Y-axis-label-spacing)) |
| 19214 | @end group |
| 19215 | @end smallexample |
| 19216 | |
| 19217 | @noindent |
| 19218 | This expression returns the value of @code{height} itself if the height |
| 19219 | is an even multiple of the value of the @code{Y-axis-label-spacing} or |
| 19220 | else it computes and returns a value of @code{height} that is equal to |
| 19221 | the next higher multiple of the value of the @code{Y-axis-label-spacing}. |
| 19222 | |
| 19223 | We can now include this expression in the @code{let} expression of the |
| 19224 | @code{print-graph} function (after first setting the value of |
| 19225 | @code{Y-axis-label-spacing}): |
| 19226 | @vindex Y-axis-label-spacing |
| 19227 | |
| 19228 | @smallexample |
| 19229 | @group |
| 19230 | (defvar Y-axis-label-spacing 5 |
| 19231 | "Number of lines from one Y axis label to next.") |
| 19232 | @end group |
| 19233 | |
| 19234 | @group |
| 19235 | @dots{} |
| 19236 | (let* ((height (apply 'max numbers-list)) |
| 19237 | (height-of-top-line |
| 19238 | (if (zerop (% height Y-axis-label-spacing)) |
| 19239 | height |
| 19240 | @end group |
| 19241 | @group |
| 19242 | ;; @r{else} |
| 19243 | (* (1+ (/ height Y-axis-label-spacing)) |
| 19244 | Y-axis-label-spacing))) |
| 19245 | (symbol-width (length graph-blank)))) |
| 19246 | @dots{} |
| 19247 | @end group |
| 19248 | @end smallexample |
| 19249 | |
| 19250 | @noindent |
| 19251 | (Note use of the @code{let*} function: the initial value of height is |
| 19252 | computed once by the @code{(apply 'max numbers-list)} expression and |
| 19253 | then the resulting value of @code{height} is used to compute its |
| 19254 | final value. @xref{fwd-para let, , The @code{let*} expression}, for |
| 19255 | more about @code{let*}.) |
| 19256 | |
| 19257 | @node Y Axis Element, Y-axis-column, Compute a Remainder, print-Y-axis |
| 19258 | @appendixsubsec Construct a Y Axis Element |
| 19259 | |
| 19260 | When we print the vertical axis, we want to insert strings such as |
| 19261 | @w{@samp{5 -}} and @w{@samp{10 - }} every five lines. |
| 19262 | Moreover, we want the numbers and dashes to line up, so shorter |
| 19263 | numbers must be padded with leading spaces. If some of the strings |
| 19264 | use two digit numbers, the strings with single digit numbers must |
| 19265 | include a leading blank space before the number. |
| 19266 | |
| 19267 | @findex number-to-string |
| 19268 | To figure out the length of the number, the @code{length} function is |
| 19269 | used. But the @code{length} function works only with a string, not with |
| 19270 | a number. So the number has to be converted from being a number to |
| 19271 | being a string. This is done with the @code{number-to-string} function. |
| 19272 | For example, |
| 19273 | |
| 19274 | @smallexample |
| 19275 | @group |
| 19276 | (length (number-to-string 35)) |
| 19277 | @result{} 2 |
| 19278 | |
| 19279 | (length (number-to-string 100)) |
| 19280 | @result{} 3 |
| 19281 | @end group |
| 19282 | @end smallexample |
| 19283 | |
| 19284 | @noindent |
| 19285 | (@code{number-to-string} is also called @code{int-to-string}; you will |
| 19286 | see this alternative name in various sources.) |
| 19287 | |
| 19288 | In addition, in each label, each number is followed by a string such |
| 19289 | as @w{@samp{ - }}, which we will call the @code{Y-axis-tic} marker. |
| 19290 | This variable is defined with @code{defvar}: |
| 19291 | |
| 19292 | @vindex Y-axis-tic |
| 19293 | @smallexample |
| 19294 | @group |
| 19295 | (defvar Y-axis-tic " - " |
| 19296 | "String that follows number in a Y axis label.") |
| 19297 | @end group |
| 19298 | @end smallexample |
| 19299 | |
| 19300 | The length of the Y label is the sum of the length of the Y axis tic |
| 19301 | mark and the length of the number of the top of the graph. |
| 19302 | |
| 19303 | @smallexample |
| 19304 | (length (concat (number-to-string height) Y-axis-tic))) |
| 19305 | @end smallexample |
| 19306 | |
| 19307 | This value will be calculated by the @code{print-graph} function in |
| 19308 | its varlist as @code{full-Y-label-width} and passed on. (Note that we |
| 19309 | did not think to include this in the varlist when we first proposed it.) |
| 19310 | |
| 19311 | To make a complete vertical axis label, a tic mark is concatenated |
| 19312 | with a number; and the two together may be preceded by one or more |
| 19313 | spaces depending on how long the number is. The label consists of |
| 19314 | three parts: the (optional) leading spaces, the number, and the tic |
| 19315 | mark. The function is passed the value of the number for the specific |
| 19316 | row, and the value of the width of the top line, which is calculated |
| 19317 | (just once) by @code{print-graph}. |
| 19318 | |
| 19319 | @smallexample |
| 19320 | @group |
| 19321 | (defun Y-axis-element (number full-Y-label-width) |
| 19322 | "Construct a NUMBERed label element. |
| 19323 | A numbered element looks like this ` 5 - ', |
| 19324 | and is padded as needed so all line up with |
| 19325 | the element for the largest number." |
| 19326 | @end group |
| 19327 | @group |
| 19328 | (let* ((leading-spaces |
| 19329 | (- full-Y-label-width |
| 19330 | (length |
| 19331 | (concat (number-to-string number) |
| 19332 | Y-axis-tic))))) |
| 19333 | @end group |
| 19334 | @group |
| 19335 | (concat |
| 19336 | (make-string leading-spaces ? ) |
| 19337 | (number-to-string number) |
| 19338 | Y-axis-tic))) |
| 19339 | @end group |
| 19340 | @end smallexample |
| 19341 | |
| 19342 | The @code{Y-axis-element} function concatenates together the leading |
| 19343 | spaces, if any; the number, as a string; and the tic mark. |
| 19344 | |
| 19345 | To figure out how many leading spaces the label will need, the |
| 19346 | function subtracts the actual length of the label---the length of the |
| 19347 | number plus the length of the tic mark---from the desired label width. |
| 19348 | |
| 19349 | @findex make-string |
| 19350 | Blank spaces are inserted using the @code{make-string} function. This |
| 19351 | function takes two arguments: the first tells it how long the string |
| 19352 | will be and the second is a symbol for the character to insert, in a |
| 19353 | special format. The format is a question mark followed by a blank |
| 19354 | space, like this, @samp{? }. @xref{Character Type, , Character Type, |
| 19355 | elisp, The GNU Emacs Lisp Reference Manual}, for a description of the |
| 19356 | syntax for characters. |
| 19357 | |
| 19358 | The @code{number-to-string} function is used in the concatenation |
| 19359 | expression, to convert the number to a string that is concatenated |
| 19360 | with the leading spaces and the tic mark. |
| 19361 | |
| 19362 | @node Y-axis-column, print-Y-axis Penultimate, Y Axis Element, print-Y-axis |
| 19363 | @appendixsubsec Create a Y Axis Column |
| 19364 | |
| 19365 | The preceding functions provide all the tools needed to construct a |
| 19366 | function that generates a list of numbered and blank strings to insert |
| 19367 | as the label for the vertical axis: |
| 19368 | |
| 19369 | @findex Y-axis-column |
| 19370 | @smallexample |
| 19371 | @group |
| 19372 | (defun Y-axis-column (height width-of-label) |
| 19373 | "Construct list of Y axis labels and blank strings. |
| 19374 | For HEIGHT of line above base and WIDTH-OF-LABEL." |
| 19375 | (let (Y-axis) |
| 19376 | @group |
| 19377 | @end group |
| 19378 | (while (> height 1) |
| 19379 | (if (zerop (% height Y-axis-label-spacing)) |
| 19380 | ;; @r{Insert label.} |
| 19381 | (setq Y-axis |
| 19382 | (cons |
| 19383 | (Y-axis-element height width-of-label) |
| 19384 | Y-axis)) |
| 19385 | @group |
| 19386 | @end group |
| 19387 | ;; @r{Else, insert blanks.} |
| 19388 | (setq Y-axis |
| 19389 | (cons |
| 19390 | (make-string width-of-label ? ) |
| 19391 | Y-axis))) |
| 19392 | (setq height (1- height))) |
| 19393 | ;; @r{Insert base line.} |
| 19394 | (setq Y-axis |
| 19395 | (cons (Y-axis-element 1 width-of-label) Y-axis)) |
| 19396 | (nreverse Y-axis))) |
| 19397 | @end group |
| 19398 | @end smallexample |
| 19399 | |
| 19400 | In this function, we start with the value of @code{height} and |
| 19401 | repetitively subtract one from its value. After each subtraction, we |
| 19402 | test to see whether the value is an integral multiple of the |
| 19403 | @code{Y-axis-label-spacing}. If it is, we construct a numbered label |
| 19404 | using the @code{Y-axis-element} function; if not, we construct a |
| 19405 | blank label using the @code{make-string} function. The base line |
| 19406 | consists of the number one followed by a tic mark. |
| 19407 | |
| 19408 | @need 2000 |
| 19409 | @node print-Y-axis Penultimate, , Y-axis-column, print-Y-axis |
| 19410 | @appendixsubsec The Not Quite Final Version of @code{print-Y-axis} |
| 19411 | |
| 19412 | The list constructed by the @code{Y-axis-column} function is passed to |
| 19413 | the @code{print-Y-axis} function, which inserts the list as a column. |
| 19414 | |
| 19415 | @findex print-Y-axis |
| 19416 | @smallexample |
| 19417 | @group |
| 19418 | (defun print-Y-axis (height full-Y-label-width) |
| 19419 | "Insert Y axis using HEIGHT and FULL-Y-LABEL-WIDTH. |
| 19420 | Height must be the maximum height of the graph. |
| 19421 | Full width is the width of the highest label element." |
| 19422 | ;; Value of height and full-Y-label-width |
| 19423 | ;; are passed by `print-graph'. |
| 19424 | @end group |
| 19425 | @group |
| 19426 | (let ((start (point))) |
| 19427 | (insert-rectangle |
| 19428 | (Y-axis-column height full-Y-label-width)) |
| 19429 | ;; @r{Place point ready for inserting graph.} |
| 19430 | (goto-char start) |
| 19431 | ;; @r{Move point forward by value of} full-Y-label-width |
| 19432 | (forward-char full-Y-label-width))) |
| 19433 | @end group |
| 19434 | @end smallexample |
| 19435 | |
| 19436 | The @code{print-Y-axis} uses the @code{insert-rectangle} function to |
| 19437 | insert the Y axis labels created by the @code{Y-axis-column} function. |
| 19438 | In addition, it places point at the correct position for printing the body of |
| 19439 | the graph. |
| 19440 | |
| 19441 | You can test @code{print-Y-axis}: |
| 19442 | |
| 19443 | @enumerate |
| 19444 | @item |
| 19445 | Install |
| 19446 | |
| 19447 | @smallexample |
| 19448 | @group |
| 19449 | Y-axis-label-spacing |
| 19450 | Y-axis-tic |
| 19451 | Y-axis-element |
| 19452 | Y-axis-column |
| 19453 | print-Y-axis |
| 19454 | @end group |
| 19455 | @end smallexample |
| 19456 | |
| 19457 | @item |
| 19458 | Copy the following expression: |
| 19459 | |
| 19460 | @smallexample |
| 19461 | (print-Y-axis 12 5) |
| 19462 | @end smallexample |
| 19463 | |
| 19464 | @item |
| 19465 | Switch to the @file{*scratch*} buffer and place the cursor where you |
| 19466 | want the axis labels to start. |
| 19467 | |
| 19468 | @item |
| 19469 | Type @kbd{M-:} (@code{eval-expression}). |
| 19470 | |
| 19471 | @item |
| 19472 | Yank the @code{graph-body-print} expression into the minibuffer |
| 19473 | with @kbd{C-y} (@code{yank)}. |
| 19474 | |
| 19475 | @item |
| 19476 | Press @key{RET} to evaluate the expression. |
| 19477 | @end enumerate |
| 19478 | |
| 19479 | Emacs will print labels vertically, the top one being |
| 19480 | @w{@samp{10 -@w{ }}}. (The @code{print-graph} function |
| 19481 | will pass the value of @code{height-of-top-line}, which |
| 19482 | in this case would end up as 15.) |
| 19483 | |
| 19484 | @need 2000 |
| 19485 | @node print-X-axis, Print Whole Graph, print-Y-axis, Full Graph |
| 19486 | @appendixsec The @code{print-X-axis} Function |
| 19487 | @cindex Axis, print horizontal |
| 19488 | @cindex X axis printing |
| 19489 | @cindex Print horizontal axis |
| 19490 | @cindex Horizontal axis printing |
| 19491 | |
| 19492 | X axis labels are much like Y axis labels, except that the tics are on a |
| 19493 | line above the numbers. Labels should look like this: |
| 19494 | |
| 19495 | @smallexample |
| 19496 | @group |
| 19497 | | | | | |
| 19498 | 1 5 10 15 |
| 19499 | @end group |
| 19500 | @end smallexample |
| 19501 | |
| 19502 | The first tic is under the first column of the graph and is preceded by |
| 19503 | several blank spaces. These spaces provide room in rows above for the Y |
| 19504 | axis labels. The second, third, fourth, and subsequent tics are all |
| 19505 | spaced equally, according to the value of @code{X-axis-label-spacing}. |
| 19506 | |
| 19507 | The second row of the X axis consists of numbers, preceded by several |
| 19508 | blank spaces and also separated according to the value of the variable |
| 19509 | @code{X-axis-label-spacing}. |
| 19510 | |
| 19511 | The value of the variable @code{X-axis-label-spacing} should itself be |
| 19512 | measured in units of @code{symbol-width}, since you may want to change |
| 19513 | the width of the symbols that you are using to print the body of the |
| 19514 | graph without changing the ways the graph is labelled. |
| 19515 | |
| 19516 | @menu |
| 19517 | * Similarities differences:: Much like @code{print-Y-axis}, but not exactly. |
| 19518 | * X Axis Tic Marks:: Create tic marks for the horizontal axis. |
| 19519 | @end menu |
| 19520 | |
| 19521 | @node Similarities differences, X Axis Tic Marks, print-X-axis, print-X-axis |
| 19522 | @ifnottex |
| 19523 | @unnumberedsubsec Similarities and differences |
| 19524 | @end ifnottex |
| 19525 | |
| 19526 | The @code{print-X-axis} function is constructed in more or less the |
| 19527 | same fashion as the @code{print-Y-axis} function except that it has |
| 19528 | two lines: the line of tic marks and the numbers. We will write a |
| 19529 | separate function to print each line and then combine them within the |
| 19530 | @code{print-X-axis} function. |
| 19531 | |
| 19532 | This is a three step process: |
| 19533 | |
| 19534 | @enumerate |
| 19535 | @item |
| 19536 | Write a function to print the X axis tic marks, @code{print-X-axis-tic-line}. |
| 19537 | |
| 19538 | @item |
| 19539 | Write a function to print the X numbers, @code{print-X-axis-numbered-line}. |
| 19540 | |
| 19541 | @item |
| 19542 | Write a function to print both lines, the @code{print-X-axis} function, |
| 19543 | using @code{print-X-axis-tic-line} and |
| 19544 | @code{print-X-axis-numbered-line}. |
| 19545 | @end enumerate |
| 19546 | |
| 19547 | @node X Axis Tic Marks, , Similarities differences, print-X-axis |
| 19548 | @appendixsubsec X Axis Tic Marks |
| 19549 | |
| 19550 | The first function should print the X axis tic marks. We must specify |
| 19551 | the tic marks themselves and their spacing: |
| 19552 | |
| 19553 | @smallexample |
| 19554 | @group |
| 19555 | (defvar X-axis-label-spacing |
| 19556 | (if (boundp 'graph-blank) |
| 19557 | (* 5 (length graph-blank)) 5) |
| 19558 | "Number of units from one X axis label to next.") |
| 19559 | @end group |
| 19560 | @end smallexample |
| 19561 | |
| 19562 | @noindent |
| 19563 | (Note that the value of @code{graph-blank} is set by another |
| 19564 | @code{defvar}. The @code{boundp} predicate checks whether it has |
| 19565 | already been set; @code{boundp} returns @code{nil} if it has not. |
| 19566 | If @code{graph-blank} were unbound and we did not use this conditional |
| 19567 | construction, in GNU Emacs 21, we would enter the debugger and see an |
| 19568 | error message saying |
| 19569 | @samp{@w{Debugger entered--Lisp error:} @w{(void-variable graph-blank)}}.) |
| 19570 | |
| 19571 | @need 1200 |
| 19572 | Here is the @code{defvar} for @code{X-axis-tic-symbol}: |
| 19573 | |
| 19574 | @smallexample |
| 19575 | @group |
| 19576 | (defvar X-axis-tic-symbol "|" |
| 19577 | "String to insert to point to a column in X axis.") |
| 19578 | @end group |
| 19579 | @end smallexample |
| 19580 | |
| 19581 | @need 1250 |
| 19582 | The goal is to make a line that looks like this: |
| 19583 | |
| 19584 | @smallexample |
| 19585 | | | | | |
| 19586 | @end smallexample |
| 19587 | |
| 19588 | The first tic is indented so that it is under the first column, which is |
| 19589 | indented to provide space for the Y axis labels. |
| 19590 | |
| 19591 | A tic element consists of the blank spaces that stretch from one tic to |
| 19592 | the next plus a tic symbol. The number of blanks is determined by the |
| 19593 | width of the tic symbol and the @code{X-axis-label-spacing}. |
| 19594 | |
| 19595 | @need 1250 |
| 19596 | The code looks like this: |
| 19597 | |
| 19598 | @smallexample |
| 19599 | @group |
| 19600 | ;;; X-axis-tic-element |
| 19601 | @dots{} |
| 19602 | (concat |
| 19603 | (make-string |
| 19604 | ;; @r{Make a string of blanks.} |
| 19605 | (- (* symbol-width X-axis-label-spacing) |
| 19606 | (length X-axis-tic-symbol)) |
| 19607 | ? ) |
| 19608 | ;; @r{Concatenate blanks with tic symbol.} |
| 19609 | X-axis-tic-symbol) |
| 19610 | @dots{} |
| 19611 | @end group |
| 19612 | @end smallexample |
| 19613 | |
| 19614 | Next, we determine how many blanks are needed to indent the first tic |
| 19615 | mark to the first column of the graph. This uses the value of |
| 19616 | @code{full-Y-label-width} passed it by the @code{print-graph} function. |
| 19617 | |
| 19618 | @need 1250 |
| 19619 | The code to make @code{X-axis-leading-spaces} |
| 19620 | looks like this: |
| 19621 | |
| 19622 | @smallexample |
| 19623 | @group |
| 19624 | ;; X-axis-leading-spaces |
| 19625 | @dots{} |
| 19626 | (make-string full-Y-label-width ? ) |
| 19627 | @dots{} |
| 19628 | @end group |
| 19629 | @end smallexample |
| 19630 | |
| 19631 | We also need to determine the length of the horizontal axis, which is |
| 19632 | the length of the numbers list, and the number of tics in the horizontal |
| 19633 | axis: |
| 19634 | |
| 19635 | @smallexample |
| 19636 | @group |
| 19637 | ;; X-length |
| 19638 | @dots{} |
| 19639 | (length numbers-list) |
| 19640 | @end group |
| 19641 | |
| 19642 | @group |
| 19643 | ;; tic-width |
| 19644 | @dots{} |
| 19645 | (* symbol-width X-axis-label-spacing) |
| 19646 | @end group |
| 19647 | |
| 19648 | @group |
| 19649 | ;; number-of-X-tics |
| 19650 | (if (zerop (% (X-length tic-width))) |
| 19651 | (/ (X-length tic-width)) |
| 19652 | (1+ (/ (X-length tic-width)))) |
| 19653 | @end group |
| 19654 | @end smallexample |
| 19655 | |
| 19656 | @need 1250 |
| 19657 | All this leads us directly to the function for printing the X axis tic line: |
| 19658 | |
| 19659 | @findex print-X-axis-tic-line |
| 19660 | @smallexample |
| 19661 | @group |
| 19662 | (defun print-X-axis-tic-line |
| 19663 | (number-of-X-tics X-axis-leading-spaces X-axis-tic-element) |
| 19664 | "Print tics for X axis." |
| 19665 | (insert X-axis-leading-spaces) |
| 19666 | (insert X-axis-tic-symbol) ; @r{Under first column.} |
| 19667 | @end group |
| 19668 | @group |
| 19669 | ;; @r{Insert second tic in the right spot.} |
| 19670 | (insert (concat |
| 19671 | (make-string |
| 19672 | (- (* symbol-width X-axis-label-spacing) |
| 19673 | ;; @r{Insert white space up to second tic symbol.} |
| 19674 | (* 2 (length X-axis-tic-symbol))) |
| 19675 | ? ) |
| 19676 | X-axis-tic-symbol)) |
| 19677 | @end group |
| 19678 | @group |
| 19679 | ;; @r{Insert remaining tics.} |
| 19680 | (while (> number-of-X-tics 1) |
| 19681 | (insert X-axis-tic-element) |
| 19682 | (setq number-of-X-tics (1- number-of-X-tics)))) |
| 19683 | @end group |
| 19684 | @end smallexample |
| 19685 | |
| 19686 | The line of numbers is equally straightforward: |
| 19687 | |
| 19688 | @need 1250 |
| 19689 | First, we create a numbered element with blank spaces before each number: |
| 19690 | |
| 19691 | @findex X-axis-element |
| 19692 | @smallexample |
| 19693 | @group |
| 19694 | (defun X-axis-element (number) |
| 19695 | "Construct a numbered X axis element." |
| 19696 | (let ((leading-spaces |
| 19697 | (- (* symbol-width X-axis-label-spacing) |
| 19698 | (length (number-to-string number))))) |
| 19699 | (concat (make-string leading-spaces ? ) |
| 19700 | (number-to-string number)))) |
| 19701 | @end group |
| 19702 | @end smallexample |
| 19703 | |
| 19704 | Next, we create the function to print the numbered line, starting with |
| 19705 | the number ``1'' under the first column: |
| 19706 | |
| 19707 | @findex print-X-axis-numbered-line |
| 19708 | @smallexample |
| 19709 | @group |
| 19710 | (defun print-X-axis-numbered-line |
| 19711 | (number-of-X-tics X-axis-leading-spaces) |
| 19712 | "Print line of X-axis numbers" |
| 19713 | (let ((number X-axis-label-spacing)) |
| 19714 | (insert X-axis-leading-spaces) |
| 19715 | (insert "1") |
| 19716 | @end group |
| 19717 | @group |
| 19718 | (insert (concat |
| 19719 | (make-string |
| 19720 | ;; @r{Insert white space up to next number.} |
| 19721 | (- (* symbol-width X-axis-label-spacing) 2) |
| 19722 | ? ) |
| 19723 | (number-to-string number))) |
| 19724 | @end group |
| 19725 | @group |
| 19726 | ;; @r{Insert remaining numbers.} |
| 19727 | (setq number (+ number X-axis-label-spacing)) |
| 19728 | (while (> number-of-X-tics 1) |
| 19729 | (insert (X-axis-element number)) |
| 19730 | (setq number (+ number X-axis-label-spacing)) |
| 19731 | (setq number-of-X-tics (1- number-of-X-tics))))) |
| 19732 | @end group |
| 19733 | @end smallexample |
| 19734 | |
| 19735 | Finally, we need to write the @code{print-X-axis} that uses |
| 19736 | @code{print-X-axis-tic-line} and |
| 19737 | @code{print-X-axis-numbered-line}. |
| 19738 | |
| 19739 | The function must determine the local values of the variables used by both |
| 19740 | @code{print-X-axis-tic-line} and @code{print-X-axis-numbered-line}, and |
| 19741 | then it must call them. Also, it must print the carriage return that |
| 19742 | separates the two lines. |
| 19743 | |
| 19744 | The function consists of a varlist that specifies five local variables, |
| 19745 | and calls to each of the two line printing functions: |
| 19746 | |
| 19747 | @findex print-X-axis |
| 19748 | @smallexample |
| 19749 | @group |
| 19750 | (defun print-X-axis (numbers-list) |
| 19751 | "Print X axis labels to length of NUMBERS-LIST." |
| 19752 | (let* ((leading-spaces |
| 19753 | (make-string full-Y-label-width ? )) |
| 19754 | @end group |
| 19755 | @group |
| 19756 | ;; symbol-width @r{is provided by} graph-body-print |
| 19757 | (tic-width (* symbol-width X-axis-label-spacing)) |
| 19758 | (X-length (length numbers-list)) |
| 19759 | @end group |
| 19760 | @group |
| 19761 | (X-tic |
| 19762 | (concat |
| 19763 | (make-string |
| 19764 | @end group |
| 19765 | @group |
| 19766 | ;; @r{Make a string of blanks.} |
| 19767 | (- (* symbol-width X-axis-label-spacing) |
| 19768 | (length X-axis-tic-symbol)) |
| 19769 | ? ) |
| 19770 | @end group |
| 19771 | @group |
| 19772 | ;; @r{Concatenate blanks with tic symbol.} |
| 19773 | X-axis-tic-symbol)) |
| 19774 | @end group |
| 19775 | @group |
| 19776 | (tic-number |
| 19777 | (if (zerop (% X-length tic-width)) |
| 19778 | (/ X-length tic-width) |
| 19779 | (1+ (/ X-length tic-width))))) |
| 19780 | @end group |
| 19781 | @group |
| 19782 | (print-X-axis-tic-line tic-number leading-spaces X-tic) |
| 19783 | (insert "\n") |
| 19784 | (print-X-axis-numbered-line tic-number leading-spaces))) |
| 19785 | @end group |
| 19786 | @end smallexample |
| 19787 | |
| 19788 | @need 1250 |
| 19789 | You can test @code{print-X-axis}: |
| 19790 | |
| 19791 | @enumerate |
| 19792 | @item |
| 19793 | Install @code{X-axis-tic-symbol}, @code{X-axis-label-spacing}, |
| 19794 | @code{print-X-axis-tic-line}, as well as @code{X-axis-element}, |
| 19795 | @code{print-X-axis-numbered-line}, and @code{print-X-axis}. |
| 19796 | |
| 19797 | @item |
| 19798 | Copy the following expression: |
| 19799 | |
| 19800 | @smallexample |
| 19801 | @group |
| 19802 | (progn |
| 19803 | (let ((full-Y-label-width 5) |
| 19804 | (symbol-width 1)) |
| 19805 | (print-X-axis |
| 19806 | '(1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16)))) |
| 19807 | @end group |
| 19808 | @end smallexample |
| 19809 | |
| 19810 | @item |
| 19811 | Switch to the @file{*scratch*} buffer and place the cursor where you |
| 19812 | want the axis labels to start. |
| 19813 | |
| 19814 | @item |
| 19815 | Type @kbd{M-:} (@code{eval-expression}). |
| 19816 | |
| 19817 | @item |
| 19818 | Yank the test expression into the minibuffer |
| 19819 | with @kbd{C-y} (@code{yank)}. |
| 19820 | |
| 19821 | @item |
| 19822 | Press @key{RET} to evaluate the expression. |
| 19823 | @end enumerate |
| 19824 | |
| 19825 | @need 1250 |
| 19826 | Emacs will print the horizontal axis like this: |
| 19827 | @sp 1 |
| 19828 | |
| 19829 | @smallexample |
| 19830 | @group |
| 19831 | | | | | | |
| 19832 | 1 5 10 15 20 |
| 19833 | @end group |
| 19834 | @end smallexample |
| 19835 | |
| 19836 | @node Print Whole Graph, , print-X-axis, Full Graph |
| 19837 | @appendixsec Printing the Whole Graph |
| 19838 | @cindex Printing the whole graph |
| 19839 | @cindex Whole graph printing |
| 19840 | @cindex Graph, printing all |
| 19841 | |
| 19842 | Now we are nearly ready to print the whole graph. |
| 19843 | |
| 19844 | The function to print the graph with the proper labels follows the |
| 19845 | outline we created earlier (@pxref{Full Graph, , A Graph with Labelled |
| 19846 | Axes}), but with additions. |
| 19847 | |
| 19848 | @need 1250 |
| 19849 | Here is the outline: |
| 19850 | |
| 19851 | @smallexample |
| 19852 | @group |
| 19853 | (defun print-graph (numbers-list) |
| 19854 | "@var{documentation}@dots{}" |
| 19855 | (let ((height @dots{} |
| 19856 | @dots{})) |
| 19857 | @end group |
| 19858 | @group |
| 19859 | (print-Y-axis height @dots{} ) |
| 19860 | (graph-body-print numbers-list) |
| 19861 | (print-X-axis @dots{} ))) |
| 19862 | @end group |
| 19863 | @end smallexample |
| 19864 | |
| 19865 | @menu |
| 19866 | * The final version:: A few changes. |
| 19867 | * Test print-graph:: Run a short test. |
| 19868 | * Graphing words in defuns:: Executing the final code. |
| 19869 | * lambda:: How to write an anonymous function. |
| 19870 | * mapcar:: Apply a function to elements of a list. |
| 19871 | * Another Bug:: Yet another bug @dots{} most insidious. |
| 19872 | * Final printed graph:: The graph itself! |
| 19873 | @end menu |
| 19874 | |
| 19875 | @node The final version, Test print-graph, Print Whole Graph, Print Whole Graph |
| 19876 | @ifnottex |
| 19877 | @unnumberedsubsec Changes for the Final Version |
| 19878 | @end ifnottex |
| 19879 | |
| 19880 | The final version is different from what we planned in two ways: |
| 19881 | first, it contains additional values calculated once in the varlist; |
| 19882 | second, it carries an option to specify the labels' increment per row. |
| 19883 | This latter feature turns out to be essential; otherwise, a graph may |
| 19884 | have more rows than fit on a display or on a sheet of paper. |
| 19885 | |
| 19886 | @need 1500 |
| 19887 | This new feature requires a change to the @code{Y-axis-column} |
| 19888 | function, to add @code{vertical-step} to it. The function looks like |
| 19889 | this: |
| 19890 | |
| 19891 | @findex Y-axis-column @r{Final version.} |
| 19892 | @smallexample |
| 19893 | @group |
| 19894 | ;;; @r{Final version.} |
| 19895 | (defun Y-axis-column |
| 19896 | (height width-of-label &optional vertical-step) |
| 19897 | "Construct list of labels for Y axis. |
| 19898 | HEIGHT is maximum height of graph. |
| 19899 | WIDTH-OF-LABEL is maximum width of label. |
| 19900 | VERTICAL-STEP, an option, is a positive integer |
| 19901 | that specifies how much a Y axis label increments |
| 19902 | for each line. For example, a step of 5 means |
| 19903 | that each line is five units of the graph." |
| 19904 | @end group |
| 19905 | @group |
| 19906 | (let (Y-axis |
| 19907 | (number-per-line (or vertical-step 1))) |
| 19908 | (while (> height 1) |
| 19909 | (if (zerop (% height Y-axis-label-spacing)) |
| 19910 | @end group |
| 19911 | @group |
| 19912 | ;; @r{Insert label.} |
| 19913 | (setq Y-axis |
| 19914 | (cons |
| 19915 | (Y-axis-element |
| 19916 | (* height number-per-line) |
| 19917 | width-of-label) |
| 19918 | Y-axis)) |
| 19919 | @end group |
| 19920 | @group |
| 19921 | ;; @r{Else, insert blanks.} |
| 19922 | (setq Y-axis |
| 19923 | (cons |
| 19924 | (make-string width-of-label ? ) |
| 19925 | Y-axis))) |
| 19926 | (setq height (1- height))) |
| 19927 | @end group |
| 19928 | @group |
| 19929 | ;; @r{Insert base line.} |
| 19930 | (setq Y-axis (cons (Y-axis-element |
| 19931 | (or vertical-step 1) |
| 19932 | width-of-label) |
| 19933 | Y-axis)) |
| 19934 | (nreverse Y-axis))) |
| 19935 | @end group |
| 19936 | @end smallexample |
| 19937 | |
| 19938 | The values for the maximum height of graph and the width of a symbol |
| 19939 | are computed by @code{print-graph} in its @code{let} expression; so |
| 19940 | @code{graph-body-print} must be changed to accept them. |
| 19941 | |
| 19942 | @findex graph-body-print @r{Final version.} |
| 19943 | @smallexample |
| 19944 | @group |
| 19945 | ;;; @r{Final version.} |
| 19946 | (defun graph-body-print (numbers-list height symbol-width) |
| 19947 | "Print a bar graph of the NUMBERS-LIST. |
| 19948 | The numbers-list consists of the Y-axis values. |
| 19949 | HEIGHT is maximum height of graph. |
| 19950 | SYMBOL-WIDTH is number of each column." |
| 19951 | @end group |
| 19952 | @group |
| 19953 | (let (from-position) |
| 19954 | (while numbers-list |
| 19955 | (setq from-position (point)) |
| 19956 | (insert-rectangle |
| 19957 | (column-of-graph height (car numbers-list))) |
| 19958 | (goto-char from-position) |
| 19959 | (forward-char symbol-width) |
| 19960 | @end group |
| 19961 | @group |
| 19962 | ;; @r{Draw graph column by column.} |
| 19963 | (sit-for 0) |
| 19964 | (setq numbers-list (cdr numbers-list))) |
| 19965 | ;; @r{Place point for X axis labels.} |
| 19966 | (forward-line height) |
| 19967 | (insert "\n"))) |
| 19968 | @end group |
| 19969 | @end smallexample |
| 19970 | |
| 19971 | @need 1250 |
| 19972 | Finally, the code for the @code{print-graph} function: |
| 19973 | |
| 19974 | @findex print-graph @r{Final version.} |
| 19975 | @smallexample |
| 19976 | @group |
| 19977 | ;;; @r{Final version.} |
| 19978 | (defun print-graph |
| 19979 | (numbers-list &optional vertical-step) |
| 19980 | "Print labelled bar graph of the NUMBERS-LIST. |
| 19981 | The numbers-list consists of the Y-axis values. |
| 19982 | @end group |
| 19983 | |
| 19984 | @group |
| 19985 | Optionally, VERTICAL-STEP, a positive integer, |
| 19986 | specifies how much a Y axis label increments for |
| 19987 | each line. For example, a step of 5 means that |
| 19988 | each row is five units." |
| 19989 | @end group |
| 19990 | @group |
| 19991 | (let* ((symbol-width (length graph-blank)) |
| 19992 | ;; @code{height} @r{is both the largest number} |
| 19993 | ;; @r{and the number with the most digits.} |
| 19994 | (height (apply 'max numbers-list)) |
| 19995 | @end group |
| 19996 | @group |
| 19997 | (height-of-top-line |
| 19998 | (if (zerop (% height Y-axis-label-spacing)) |
| 19999 | height |
| 20000 | ;; @r{else} |
| 20001 | (* (1+ (/ height Y-axis-label-spacing)) |
| 20002 | Y-axis-label-spacing))) |
| 20003 | @end group |
| 20004 | @group |
| 20005 | (vertical-step (or vertical-step 1)) |
| 20006 | (full-Y-label-width |
| 20007 | (length |
| 20008 | @end group |
| 20009 | @group |
| 20010 | (concat |
| 20011 | (number-to-string |
| 20012 | (* height-of-top-line vertical-step)) |
| 20013 | Y-axis-tic)))) |
| 20014 | @end group |
| 20015 | |
| 20016 | @group |
| 20017 | (print-Y-axis |
| 20018 | height-of-top-line full-Y-label-width vertical-step) |
| 20019 | @end group |
| 20020 | @group |
| 20021 | (graph-body-print |
| 20022 | numbers-list height-of-top-line symbol-width) |
| 20023 | (print-X-axis numbers-list))) |
| 20024 | @end group |
| 20025 | @end smallexample |
| 20026 | |
| 20027 | @node Test print-graph, Graphing words in defuns, The final version, Print Whole Graph |
| 20028 | @appendixsubsec Testing @code{print-graph} |
| 20029 | |
| 20030 | @need 1250 |
| 20031 | We can test the @code{print-graph} function with a short list of numbers: |
| 20032 | |
| 20033 | @enumerate |
| 20034 | @item |
| 20035 | Install the final versions of @code{Y-axis-column}, |
| 20036 | @code{graph-body-print}, and @code{print-graph} (in addition to the |
| 20037 | rest of the code.) |
| 20038 | |
| 20039 | @item |
| 20040 | Copy the following expression: |
| 20041 | |
| 20042 | @smallexample |
| 20043 | (print-graph '(3 2 5 6 7 5 3 4 6 4 3 2 1)) |
| 20044 | @end smallexample |
| 20045 | |
| 20046 | @item |
| 20047 | Switch to the @file{*scratch*} buffer and place the cursor where you |
| 20048 | want the axis labels to start. |
| 20049 | |
| 20050 | @item |
| 20051 | Type @kbd{M-:} (@code{eval-expression}). |
| 20052 | |
| 20053 | @item |
| 20054 | Yank the test expression into the minibuffer |
| 20055 | with @kbd{C-y} (@code{yank)}. |
| 20056 | |
| 20057 | @item |
| 20058 | Press @key{RET} to evaluate the expression. |
| 20059 | @end enumerate |
| 20060 | |
| 20061 | @need 1250 |
| 20062 | Emacs will print a graph that looks like this: |
| 20063 | |
| 20064 | @smallexample |
| 20065 | @group |
| 20066 | 10 - |
| 20067 | |
| 20068 | |
| 20069 | * |
| 20070 | ** * |
| 20071 | 5 - **** * |
| 20072 | **** *** |
| 20073 | * ********* |
| 20074 | ************ |
| 20075 | 1 - ************* |
| 20076 | |
| 20077 | | | | | |
| 20078 | 1 5 10 15 |
| 20079 | @end group |
| 20080 | @end smallexample |
| 20081 | |
| 20082 | @need 1200 |
| 20083 | On the other hand, if you pass @code{print-graph} a |
| 20084 | @code{vertical-step} value of 2, by evaluating this expression: |
| 20085 | |
| 20086 | @smallexample |
| 20087 | (print-graph '(3 2 5 6 7 5 3 4 6 4 3 2 1) 2) |
| 20088 | @end smallexample |
| 20089 | |
| 20090 | @need 1250 |
| 20091 | @noindent |
| 20092 | The graph looks like this: |
| 20093 | |
| 20094 | @smallexample |
| 20095 | @group |
| 20096 | 20 - |
| 20097 | |
| 20098 | |
| 20099 | * |
| 20100 | ** * |
| 20101 | 10 - **** * |
| 20102 | **** *** |
| 20103 | * ********* |
| 20104 | ************ |
| 20105 | 2 - ************* |
| 20106 | |
| 20107 | | | | | |
| 20108 | 1 5 10 15 |
| 20109 | @end group |
| 20110 | @end smallexample |
| 20111 | |
| 20112 | @noindent |
| 20113 | (A question: is the `2' on the bottom of the vertical axis a bug or a |
| 20114 | feature? If you think it is a bug, and should be a `1' instead, (or |
| 20115 | even a `0'), you can modify the sources.) |
| 20116 | |
| 20117 | @node Graphing words in defuns, lambda, Test print-graph, Print Whole Graph |
| 20118 | @appendixsubsec Graphing Numbers of Words and Symbols |
| 20119 | |
| 20120 | Now for the graph for which all this code was written: a graph that |
| 20121 | shows how many function definitions contain fewer than 10 words and |
| 20122 | symbols, how many contain between 10 and 19 words and symbols, how |
| 20123 | many contain between 20 and 29 words and symbols, and so on. |
| 20124 | |
| 20125 | This is a multi-step process. First make sure you have loaded all the |
| 20126 | requisite code. |
| 20127 | |
| 20128 | @need 1500 |
| 20129 | It is a good idea to reset the value of @code{top-of-ranges} in case |
| 20130 | you have set it to some different value. You can evaluate the |
| 20131 | following: |
| 20132 | |
| 20133 | @smallexample |
| 20134 | @group |
| 20135 | (setq top-of-ranges |
| 20136 | '(10 20 30 40 50 |
| 20137 | 60 70 80 90 100 |
| 20138 | 110 120 130 140 150 |
| 20139 | 160 170 180 190 200 |
| 20140 | 210 220 230 240 250 |
| 20141 | 260 270 280 290 300) |
| 20142 | @end group |
| 20143 | @end smallexample |
| 20144 | |
| 20145 | @noindent |
| 20146 | Next create a list of the number of words and symbols in each range. |
| 20147 | |
| 20148 | @need 1500 |
| 20149 | @noindent |
| 20150 | Evaluate the following: |
| 20151 | |
| 20152 | @smallexample |
| 20153 | @group |
| 20154 | (setq list-for-graph |
| 20155 | (defuns-per-range |
| 20156 | (sort |
| 20157 | (recursive-lengths-list-many-files |
| 20158 | (directory-files "/usr/local/emacs/lisp" |
| 20159 | t ".+el$")) |
| 20160 | '<) |
| 20161 | top-of-ranges)) |
| 20162 | @end group |
| 20163 | @end smallexample |
| 20164 | |
| 20165 | @noindent |
| 20166 | On my old machine, this took about an hour. It looked though 303 Lisp |
| 20167 | files in my copy of Emacs version 19.23. After all that computing, |
| 20168 | the @code{list-for-graph} had this value: |
| 20169 | |
| 20170 | @smallexample |
| 20171 | @group |
| 20172 | (537 1027 955 785 594 483 349 292 224 199 166 120 116 99 |
| 20173 | 90 80 67 48 52 45 41 33 28 26 25 20 12 28 11 13 220) |
| 20174 | @end group |
| 20175 | @end smallexample |
| 20176 | |
| 20177 | @noindent |
| 20178 | This means that my copy of Emacs had 537 function definitions with |
| 20179 | fewer than 10 words or symbols in them, 1,027 function definitions |
| 20180 | with 10 to 19 words or symbols in them, 955 function definitions with |
| 20181 | 20 to 29 words or symbols in them, and so on. |
| 20182 | |
| 20183 | Clearly, just by looking at this list we can see that most function |
| 20184 | definitions contain ten to thirty words and symbols. |
| 20185 | |
| 20186 | Now for printing. We do @emph{not} want to print a graph that is |
| 20187 | 1,030 lines high @dots{} Instead, we should print a graph that is |
| 20188 | fewer than twenty-five lines high. A graph that height can be |
| 20189 | displayed on almost any monitor, and easily printed on a sheet of paper. |
| 20190 | |
| 20191 | This means that each value in @code{list-for-graph} must be reduced to |
| 20192 | one-fiftieth its present value. |
| 20193 | |
| 20194 | Here is a short function to do just that, using two functions we have |
| 20195 | not yet seen, @code{mapcar} and @code{lambda}. |
| 20196 | |
| 20197 | @smallexample |
| 20198 | @group |
| 20199 | (defun one-fiftieth (full-range) |
| 20200 | "Return list, each number one-fiftieth of previous." |
| 20201 | (mapcar '(lambda (arg) (/ arg 50)) full-range)) |
| 20202 | @end group |
| 20203 | @end smallexample |
| 20204 | |
| 20205 | @node lambda, mapcar, Graphing words in defuns, Print Whole Graph |
| 20206 | @appendixsubsec A @code{lambda} Expression: Useful Anonymity |
| 20207 | @cindex Anonymous function |
| 20208 | @findex lambda |
| 20209 | |
| 20210 | @code{lambda} is the symbol for an anonymous function, a function |
| 20211 | without a name. Every time you use an anonymous function, you need to |
| 20212 | include its whole body. |
| 20213 | |
| 20214 | @need 1250 |
| 20215 | @noindent |
| 20216 | Thus, |
| 20217 | |
| 20218 | @smallexample |
| 20219 | (lambda (arg) (/ arg 50)) |
| 20220 | @end smallexample |
| 20221 | |
| 20222 | @noindent |
| 20223 | is a function definition that says `return the value resulting from |
| 20224 | dividing whatever is passed to me as @code{arg} by 50'. |
| 20225 | |
| 20226 | @need 1200 |
| 20227 | Earlier, for example, we had a function @code{multiply-by-seven}; it |
| 20228 | multiplied its argument by 7. This function is similar, except it |
| 20229 | divides its argument by 50; and, it has no name. The anonymous |
| 20230 | equivalent of @code{multiply-by-seven} is: |
| 20231 | |
| 20232 | @smallexample |
| 20233 | (lambda (number) (* 7 number)) |
| 20234 | @end smallexample |
| 20235 | |
| 20236 | @noindent |
| 20237 | (@xref{defun, , The @code{defun} Special Form}.) |
| 20238 | |
| 20239 | @need 1250 |
| 20240 | @noindent |
| 20241 | If we want to multiply 3 by 7, we can write: |
| 20242 | |
| 20243 | @c !!! Clear print-postscript-figures if the computer formatting this |
| 20244 | @c document is too small and cannot handle all the diagrams and figures. |
| 20245 | @c clear print-postscript-figures |
| 20246 | @c set print-postscript-figures |
| 20247 | @c lambda example diagram #1 |
| 20248 | @ifnottex |
| 20249 | @smallexample |
| 20250 | @group |
| 20251 | (multiply-by-seven 3) |
| 20252 | \_______________/ ^ |
| 20253 | | | |
| 20254 | function argument |
| 20255 | @end group |
| 20256 | @end smallexample |
| 20257 | @end ifnottex |
| 20258 | @ifset print-postscript-figures |
| 20259 | @sp 1 |
| 20260 | @tex |
| 20261 | @image{lambda-1} |
| 20262 | %%%% old method of including an image |
| 20263 | % \input /usr/local/lib/tex/inputs/psfig.tex |
| 20264 | % \centerline{\psfig{figure=/usr/local/lib/emacs/man/lambda-1.eps}} |
| 20265 | % \catcode`\@=0 % |
| 20266 | @end tex |
| 20267 | @sp 1 |
| 20268 | @end ifset |
| 20269 | @ifclear print-postscript-figures |
| 20270 | @iftex |
| 20271 | @smallexample |
| 20272 | @group |
| 20273 | (multiply-by-seven 3) |
| 20274 | \_______________/ ^ |
| 20275 | | | |
| 20276 | function argument |
| 20277 | @end group |
| 20278 | @end smallexample |
| 20279 | @end iftex |
| 20280 | @end ifclear |
| 20281 | |
| 20282 | @noindent |
| 20283 | This expression returns 21. |
| 20284 | |
| 20285 | @need 1250 |
| 20286 | @noindent |
| 20287 | Similarly, we can write: |
| 20288 | |
| 20289 | @c lambda example diagram #2 |
| 20290 | @ifnottex |
| 20291 | @smallexample |
| 20292 | @group |
| 20293 | ((lambda (number) (* 7 number)) 3) |
| 20294 | \____________________________/ ^ |
| 20295 | | | |
| 20296 | anonymous function argument |
| 20297 | @end group |
| 20298 | @end smallexample |
| 20299 | @end ifnottex |
| 20300 | @ifset print-postscript-figures |
| 20301 | @sp 1 |
| 20302 | @tex |
| 20303 | @image{lambda-2} |
| 20304 | %%%% old method of including an image |
| 20305 | % \input /usr/local/lib/tex/inputs/psfig.tex |
| 20306 | % \centerline{\psfig{figure=/usr/local/lib/emacs/man/lambda-2.eps}} |
| 20307 | % \catcode`\@=0 % |
| 20308 | @end tex |
| 20309 | @sp 1 |
| 20310 | @end ifset |
| 20311 | @ifclear print-postscript-figures |
| 20312 | @iftex |
| 20313 | @smallexample |
| 20314 | @group |
| 20315 | ((lambda (number) (* 7 number)) 3) |
| 20316 | \____________________________/ ^ |
| 20317 | | | |
| 20318 | anonymous function argument |
| 20319 | @end group |
| 20320 | @end smallexample |
| 20321 | @end iftex |
| 20322 | @end ifclear |
| 20323 | |
| 20324 | @need 1250 |
| 20325 | @noindent |
| 20326 | If we want to divide 100 by 50, we can write: |
| 20327 | |
| 20328 | @c lambda example diagram #3 |
| 20329 | @ifnottex |
| 20330 | @smallexample |
| 20331 | @group |
| 20332 | ((lambda (arg) (/ arg 50)) 100) |
| 20333 | \______________________/ \_/ |
| 20334 | | | |
| 20335 | anonymous function argument |
| 20336 | @end group |
| 20337 | @end smallexample |
| 20338 | @end ifnottex |
| 20339 | @ifset print-postscript-figures |
| 20340 | @sp 1 |
| 20341 | @tex |
| 20342 | @image{lambda-3} |
| 20343 | %%%% old method of including an image |
| 20344 | % \input /usr/local/lib/tex/inputs/psfig.tex |
| 20345 | % \centerline{\psfig{figure=/usr/local/lib/emacs/man/lambda-3.eps}} |
| 20346 | % \catcode`\@=0 % |
| 20347 | @end tex |
| 20348 | @sp 1 |
| 20349 | @end ifset |
| 20350 | @ifclear print-postscript-figures |
| 20351 | @iftex |
| 20352 | @smallexample |
| 20353 | @group |
| 20354 | ((lambda (arg) (/ arg 50)) 100) |
| 20355 | \______________________/ \_/ |
| 20356 | | | |
| 20357 | anonymous function argument |
| 20358 | @end group |
| 20359 | @end smallexample |
| 20360 | @end iftex |
| 20361 | @end ifclear |
| 20362 | |
| 20363 | @noindent |
| 20364 | This expression returns 2. The 100 is passed to the function, which |
| 20365 | divides that number by 50. |
| 20366 | |
| 20367 | @xref{Lambda Expressions, , Lambda Expressions, elisp, The GNU Emacs |
| 20368 | Lisp Reference Manual}, for more about @code{lambda}. Lisp and lambda |
| 20369 | expressions derive from the Lambda Calculus. |
| 20370 | |
| 20371 | @node mapcar, Another Bug, lambda, Print Whole Graph |
| 20372 | @appendixsubsec The @code{mapcar} Function |
| 20373 | @findex mapcar |
| 20374 | |
| 20375 | @code{mapcar} is a function that calls its first argument with each |
| 20376 | element of its second argument, in turn. The second argument must be |
| 20377 | a sequence. |
| 20378 | |
| 20379 | The @samp{map} part of the name comes from the mathematical phrase, |
| 20380 | `mapping over a domain', meaning to apply a function to each of the |
| 20381 | elements in a domain. The mathematical phrase is based on the |
| 20382 | metaphor of a surveyor walking, one step at a time, over an area he is |
| 20383 | mapping. And @samp{car}, of course, comes from the Lisp notion of the |
| 20384 | first of a list. |
| 20385 | |
| 20386 | @need 1250 |
| 20387 | @noindent |
| 20388 | For example, |
| 20389 | |
| 20390 | @smallexample |
| 20391 | @group |
| 20392 | (mapcar '1+ '(2 4 6)) |
| 20393 | @result{} (3 5 7) |
| 20394 | @end group |
| 20395 | @end smallexample |
| 20396 | |
| 20397 | @noindent |
| 20398 | The function @code{1+} which adds one to its argument, is executed on |
| 20399 | @emph{each} element of the list, and a new list is returned. |
| 20400 | |
| 20401 | Contrast this with @code{apply}, which applies its first argument to |
| 20402 | all the remaining. |
| 20403 | (@xref{Readying a Graph, , Readying a Graph}, for a explanation of |
| 20404 | @code{apply}.) |
| 20405 | |
| 20406 | @need 1250 |
| 20407 | In the definition of @code{one-fiftieth}, the first argument is the |
| 20408 | anonymous function: |
| 20409 | |
| 20410 | @smallexample |
| 20411 | (lambda (arg) (/ arg 50)) |
| 20412 | @end smallexample |
| 20413 | |
| 20414 | @noindent |
| 20415 | and the second argument is @code{full-range}, which will be bound to |
| 20416 | @code{list-for-graph}. |
| 20417 | |
| 20418 | @need 1250 |
| 20419 | The whole expression looks like this: |
| 20420 | |
| 20421 | @smallexample |
| 20422 | (mapcar '(lambda (arg) (/ arg 50)) full-range)) |
| 20423 | @end smallexample |
| 20424 | |
| 20425 | @xref{Mapping Functions, , Mapping Functions, elisp, The GNU Emacs |
| 20426 | Lisp Reference Manual}, for more about @code{mapcar}. |
| 20427 | |
| 20428 | Using the @code{one-fiftieth} function, we can generate a list in |
| 20429 | which each element is one-fiftieth the size of the corresponding |
| 20430 | element in @code{list-for-graph}. |
| 20431 | |
| 20432 | @smallexample |
| 20433 | @group |
| 20434 | (setq fiftieth-list-for-graph |
| 20435 | (one-fiftieth list-for-graph)) |
| 20436 | @end group |
| 20437 | @end smallexample |
| 20438 | |
| 20439 | @need 1250 |
| 20440 | The resulting list looks like this: |
| 20441 | |
| 20442 | @smallexample |
| 20443 | @group |
| 20444 | (10 20 19 15 11 9 6 5 4 3 3 2 2 |
| 20445 | 1 1 1 1 0 1 0 0 0 0 0 0 0 0 0 0 0 4) |
| 20446 | @end group |
| 20447 | @end smallexample |
| 20448 | |
| 20449 | @noindent |
| 20450 | This, we are almost ready to print! (We also notice the loss of |
| 20451 | information: many of the higher ranges are 0, meaning that fewer than |
| 20452 | 50 defuns had that many words or symbols---but not necessarily meaning |
| 20453 | that none had that many words or symbols.) |
| 20454 | |
| 20455 | @node Another Bug, Final printed graph, mapcar, Print Whole Graph |
| 20456 | @appendixsubsec Another Bug @dots{} Most Insidious |
| 20457 | @cindex Bug, most insidious type |
| 20458 | @cindex Insidious type of bug |
| 20459 | |
| 20460 | I said `almost ready to print'! Of course, there is a bug in the |
| 20461 | @code{print-graph} function @dots{} It has a @code{vertical-step} |
| 20462 | option, but not a @code{horizontal-step} option. The |
| 20463 | @code{top-of-range} scale goes from 10 to 300 by tens. But the |
| 20464 | @code{print-graph} function will print only by ones. |
| 20465 | |
| 20466 | This is a classic example of what some consider the most insidious |
| 20467 | type of bug, the bug of omission. This is not the kind of bug you can |
| 20468 | find by studying the code, for it is not in the code; it is an omitted |
| 20469 | feature. Your best actions are to try your program early and often; |
| 20470 | and try to arrange, as much as you can, to write code that is easy to |
| 20471 | understand and easy to change. Try to be aware, whenever you can, |
| 20472 | that whatever you have written, @emph{will} be rewritten, if not soon, |
| 20473 | eventually. A hard maxim to follow. |
| 20474 | |
| 20475 | It is the @code{print-X-axis-numbered-line} function that needs the |
| 20476 | work; and then the @code{print-X-axis} and the @code{print-graph} |
| 20477 | functions need to be adapted. Not much needs to be done; there is one |
| 20478 | nicety: the numbers ought to line up under the tic marks. This takes |
| 20479 | a little thought. |
| 20480 | |
| 20481 | @need 1250 |
| 20482 | Here is the corrected @code{print-X-axis-numbered-line}: |
| 20483 | |
| 20484 | @smallexample |
| 20485 | @group |
| 20486 | (defun print-X-axis-numbered-line |
| 20487 | (number-of-X-tics X-axis-leading-spaces |
| 20488 | &optional horizontal-step) |
| 20489 | "Print line of X-axis numbers" |
| 20490 | (let ((number X-axis-label-spacing) |
| 20491 | (horizontal-step (or horizontal-step 1))) |
| 20492 | @end group |
| 20493 | @group |
| 20494 | (insert X-axis-leading-spaces) |
| 20495 | ;; @r{Delete extra leading spaces.} |
| 20496 | (delete-char |
| 20497 | (- (1- |
| 20498 | (length (number-to-string horizontal-step))))) |
| 20499 | (insert (concat |
| 20500 | (make-string |
| 20501 | @end group |
| 20502 | @group |
| 20503 | ;; @r{Insert white space.} |
| 20504 | (- (* symbol-width |
| 20505 | X-axis-label-spacing) |
| 20506 | (1- |
| 20507 | (length |
| 20508 | (number-to-string horizontal-step))) |
| 20509 | 2) |
| 20510 | ? ) |
| 20511 | (number-to-string |
| 20512 | (* number horizontal-step)))) |
| 20513 | @end group |
| 20514 | @group |
| 20515 | ;; @r{Insert remaining numbers.} |
| 20516 | (setq number (+ number X-axis-label-spacing)) |
| 20517 | (while (> number-of-X-tics 1) |
| 20518 | (insert (X-axis-element |
| 20519 | (* number horizontal-step))) |
| 20520 | (setq number (+ number X-axis-label-spacing)) |
| 20521 | (setq number-of-X-tics (1- number-of-X-tics))))) |
| 20522 | @end group |
| 20523 | @end smallexample |
| 20524 | |
| 20525 | @need 1500 |
| 20526 | If you are reading this in Info, you can see the new versions of |
| 20527 | @code{print-X-axis} @code{print-graph} and evaluate them. If you are |
| 20528 | reading this in a printed book, you can see the changed lines here |
| 20529 | (the full text is too much to print). |
| 20530 | |
| 20531 | @iftex |
| 20532 | @smallexample |
| 20533 | @group |
| 20534 | (defun print-X-axis (numbers-list horizontal-step) |
| 20535 | @dots{} |
| 20536 | (print-X-axis-numbered-line |
| 20537 | tic-number leading-spaces horizontal-step)) |
| 20538 | @end group |
| 20539 | @end smallexample |
| 20540 | |
| 20541 | @smallexample |
| 20542 | @group |
| 20543 | (defun print-graph |
| 20544 | (numbers-list |
| 20545 | &optional vertical-step horizontal-step) |
| 20546 | @dots{} |
| 20547 | (print-X-axis numbers-list horizontal-step)) |
| 20548 | @end group |
| 20549 | @end smallexample |
| 20550 | @end iftex |
| 20551 | |
| 20552 | @ifnottex |
| 20553 | @smallexample |
| 20554 | @group |
| 20555 | (defun print-X-axis (numbers-list horizontal-step) |
| 20556 | "Print X axis labels to length of NUMBERS-LIST. |
| 20557 | Optionally, HORIZONTAL-STEP, a positive integer, |
| 20558 | specifies how much an X axis label increments for |
| 20559 | each column." |
| 20560 | @end group |
| 20561 | @group |
| 20562 | ;; Value of symbol-width and full-Y-label-width |
| 20563 | ;; are passed by `print-graph'. |
| 20564 | (let* ((leading-spaces |
| 20565 | (make-string full-Y-label-width ? )) |
| 20566 | ;; symbol-width @r{is provided by} graph-body-print |
| 20567 | (tic-width (* symbol-width X-axis-label-spacing)) |
| 20568 | (X-length (length numbers-list)) |
| 20569 | @end group |
| 20570 | @group |
| 20571 | (X-tic |
| 20572 | (concat |
| 20573 | (make-string |
| 20574 | ;; @r{Make a string of blanks.} |
| 20575 | (- (* symbol-width X-axis-label-spacing) |
| 20576 | (length X-axis-tic-symbol)) |
| 20577 | ? ) |
| 20578 | @end group |
| 20579 | @group |
| 20580 | ;; @r{Concatenate blanks with tic symbol.} |
| 20581 | X-axis-tic-symbol)) |
| 20582 | (tic-number |
| 20583 | (if (zerop (% X-length tic-width)) |
| 20584 | (/ X-length tic-width) |
| 20585 | (1+ (/ X-length tic-width))))) |
| 20586 | @end group |
| 20587 | |
| 20588 | @group |
| 20589 | (print-X-axis-tic-line |
| 20590 | tic-number leading-spaces X-tic) |
| 20591 | (insert "\n") |
| 20592 | (print-X-axis-numbered-line |
| 20593 | tic-number leading-spaces horizontal-step))) |
| 20594 | @end group |
| 20595 | @end smallexample |
| 20596 | |
| 20597 | @smallexample |
| 20598 | @group |
| 20599 | (defun print-graph |
| 20600 | (numbers-list &optional vertical-step horizontal-step) |
| 20601 | "Print labelled bar graph of the NUMBERS-LIST. |
| 20602 | The numbers-list consists of the Y-axis values. |
| 20603 | @end group |
| 20604 | |
| 20605 | @group |
| 20606 | Optionally, VERTICAL-STEP, a positive integer, |
| 20607 | specifies how much a Y axis label increments for |
| 20608 | each line. For example, a step of 5 means that |
| 20609 | each row is five units. |
| 20610 | @end group |
| 20611 | |
| 20612 | @group |
| 20613 | Optionally, HORIZONTAL-STEP, a positive integer, |
| 20614 | specifies how much an X axis label increments for |
| 20615 | each column." |
| 20616 | (let* ((symbol-width (length graph-blank)) |
| 20617 | ;; @code{height} @r{is both the largest number} |
| 20618 | ;; @r{and the number with the most digits.} |
| 20619 | (height (apply 'max numbers-list)) |
| 20620 | @end group |
| 20621 | @group |
| 20622 | (height-of-top-line |
| 20623 | (if (zerop (% height Y-axis-label-spacing)) |
| 20624 | height |
| 20625 | ;; @r{else} |
| 20626 | (* (1+ (/ height Y-axis-label-spacing)) |
| 20627 | Y-axis-label-spacing))) |
| 20628 | @end group |
| 20629 | @group |
| 20630 | (vertical-step (or vertical-step 1)) |
| 20631 | (full-Y-label-width |
| 20632 | (length |
| 20633 | (concat |
| 20634 | (number-to-string |
| 20635 | (* height-of-top-line vertical-step)) |
| 20636 | Y-axis-tic)))) |
| 20637 | @end group |
| 20638 | @group |
| 20639 | (print-Y-axis |
| 20640 | height-of-top-line full-Y-label-width vertical-step) |
| 20641 | (graph-body-print |
| 20642 | numbers-list height-of-top-line symbol-width) |
| 20643 | (print-X-axis numbers-list horizontal-step))) |
| 20644 | @end group |
| 20645 | @end smallexample |
| 20646 | @end ifnottex |
| 20647 | |
| 20648 | @c qqq |
| 20649 | @ignore |
| 20650 | Graphing Definitions Re-listed |
| 20651 | |
| 20652 | @need 1250 |
| 20653 | Here are all the graphing definitions in their final form: |
| 20654 | |
| 20655 | @smallexample |
| 20656 | @group |
| 20657 | (defvar top-of-ranges |
| 20658 | '(10 20 30 40 50 |
| 20659 | 60 70 80 90 100 |
| 20660 | 110 120 130 140 150 |
| 20661 | 160 170 180 190 200 |
| 20662 | 210 220 230 240 250) |
| 20663 | "List specifying ranges for `defuns-per-range'.") |
| 20664 | @end group |
| 20665 | |
| 20666 | @group |
| 20667 | (defvar graph-symbol "*" |
| 20668 | "String used as symbol in graph, usually an asterisk.") |
| 20669 | @end group |
| 20670 | |
| 20671 | @group |
| 20672 | (defvar graph-blank " " |
| 20673 | "String used as blank in graph, usually a blank space. |
| 20674 | graph-blank must be the same number of columns wide |
| 20675 | as graph-symbol.") |
| 20676 | @end group |
| 20677 | |
| 20678 | @group |
| 20679 | (defvar Y-axis-tic " - " |
| 20680 | "String that follows number in a Y axis label.") |
| 20681 | @end group |
| 20682 | |
| 20683 | @group |
| 20684 | (defvar Y-axis-label-spacing 5 |
| 20685 | "Number of lines from one Y axis label to next.") |
| 20686 | @end group |
| 20687 | |
| 20688 | @group |
| 20689 | (defvar X-axis-tic-symbol "|" |
| 20690 | "String to insert to point to a column in X axis.") |
| 20691 | @end group |
| 20692 | |
| 20693 | @group |
| 20694 | (defvar X-axis-label-spacing |
| 20695 | (if (boundp 'graph-blank) |
| 20696 | (* 5 (length graph-blank)) 5) |
| 20697 | "Number of units from one X axis label to next.") |
| 20698 | @end group |
| 20699 | @end smallexample |
| 20700 | |
| 20701 | @smallexample |
| 20702 | @group |
| 20703 | (defun count-words-in-defun () |
| 20704 | "Return the number of words and symbols in a defun." |
| 20705 | (beginning-of-defun) |
| 20706 | (let ((count 0) |
| 20707 | (end (save-excursion (end-of-defun) (point)))) |
| 20708 | @end group |
| 20709 | |
| 20710 | @group |
| 20711 | (while |
| 20712 | (and (< (point) end) |
| 20713 | (re-search-forward |
| 20714 | "\\(\\w\\|\\s_\\)+[^ \t\n]*[ \t\n]*" |
| 20715 | end t)) |
| 20716 | (setq count (1+ count))) |
| 20717 | count)) |
| 20718 | @end group |
| 20719 | @end smallexample |
| 20720 | |
| 20721 | @smallexample |
| 20722 | @group |
| 20723 | (defun lengths-list-file (filename) |
| 20724 | "Return list of definitions' lengths within FILE. |
| 20725 | The returned list is a list of numbers. |
| 20726 | Each number is the number of words or |
| 20727 | symbols in one function definition." |
| 20728 | @end group |
| 20729 | |
| 20730 | @group |
| 20731 | (message "Working on `%s' ... " filename) |
| 20732 | (save-excursion |
| 20733 | (let ((buffer (find-file-noselect filename)) |
| 20734 | (lengths-list)) |
| 20735 | (set-buffer buffer) |
| 20736 | (setq buffer-read-only t) |
| 20737 | (widen) |
| 20738 | (goto-char (point-min)) |
| 20739 | @end group |
| 20740 | |
| 20741 | @group |
| 20742 | (while (re-search-forward "^(defun" nil t) |
| 20743 | (setq lengths-list |
| 20744 | (cons (count-words-in-defun) lengths-list))) |
| 20745 | (kill-buffer buffer) |
| 20746 | lengths-list))) |
| 20747 | @end group |
| 20748 | @end smallexample |
| 20749 | |
| 20750 | @smallexample |
| 20751 | @group |
| 20752 | (defun lengths-list-many-files (list-of-files) |
| 20753 | "Return list of lengths of defuns in LIST-OF-FILES." |
| 20754 | (let (lengths-list) |
| 20755 | ;;; @r{true-or-false-test} |
| 20756 | (while list-of-files |
| 20757 | (setq lengths-list |
| 20758 | (append |
| 20759 | lengths-list |
| 20760 | @end group |
| 20761 | @group |
| 20762 | ;;; @r{Generate a lengths' list.} |
| 20763 | (lengths-list-file |
| 20764 | (expand-file-name (car list-of-files))))) |
| 20765 | ;;; @r{Make files' list shorter.} |
| 20766 | (setq list-of-files (cdr list-of-files))) |
| 20767 | ;;; @r{Return final value of lengths' list.} |
| 20768 | lengths-list)) |
| 20769 | @end group |
| 20770 | @end smallexample |
| 20771 | |
| 20772 | @smallexample |
| 20773 | @group |
| 20774 | (defun defuns-per-range (sorted-lengths top-of-ranges) |
| 20775 | "SORTED-LENGTHS defuns in each TOP-OF-RANGES range." |
| 20776 | (let ((top-of-range (car top-of-ranges)) |
| 20777 | (number-within-range 0) |
| 20778 | defuns-per-range-list) |
| 20779 | @end group |
| 20780 | |
| 20781 | @group |
| 20782 | ;; @r{Outer loop.} |
| 20783 | (while top-of-ranges |
| 20784 | |
| 20785 | ;; @r{Inner loop.} |
| 20786 | (while (and |
| 20787 | ;; @r{Need number for numeric test.} |
| 20788 | (car sorted-lengths) |
| 20789 | (< (car sorted-lengths) top-of-range)) |
| 20790 | |
| 20791 | ;; @r{Count number of definitions within current range.} |
| 20792 | (setq number-within-range (1+ number-within-range)) |
| 20793 | (setq sorted-lengths (cdr sorted-lengths))) |
| 20794 | @end group |
| 20795 | |
| 20796 | @group |
| 20797 | ;; @r{Exit inner loop but remain within outer loop.} |
| 20798 | |
| 20799 | (setq defuns-per-range-list |
| 20800 | (cons number-within-range defuns-per-range-list)) |
| 20801 | (setq number-within-range 0) ; @r{Reset count to zero.} |
| 20802 | |
| 20803 | ;; @r{Move to next range.} |
| 20804 | (setq top-of-ranges (cdr top-of-ranges)) |
| 20805 | ;; @r{Specify next top of range value.} |
| 20806 | (setq top-of-range (car top-of-ranges))) |
| 20807 | @end group |
| 20808 | |
| 20809 | @group |
| 20810 | ;; @r{Exit outer loop and count the number of defuns larger than} |
| 20811 | ;; @r{ the largest top-of-range value.} |
| 20812 | (setq defuns-per-range-list |
| 20813 | (cons |
| 20814 | (length sorted-lengths) |
| 20815 | defuns-per-range-list)) |
| 20816 | |
| 20817 | ;; @r{Return a list of the number of definitions within each range,} |
| 20818 | ;; @r{ smallest to largest.} |
| 20819 | (nreverse defuns-per-range-list))) |
| 20820 | @end group |
| 20821 | @end smallexample |
| 20822 | |
| 20823 | @smallexample |
| 20824 | @group |
| 20825 | (defun column-of-graph (max-graph-height actual-height) |
| 20826 | "Return list of MAX-GRAPH-HEIGHT strings; |
| 20827 | ACTUAL-HEIGHT are graph-symbols. |
| 20828 | The graph-symbols are contiguous entries at the end |
| 20829 | of the list. |
| 20830 | The list will be inserted as one column of a graph. |
| 20831 | The strings are either graph-blank or graph-symbol." |
| 20832 | @end group |
| 20833 | |
| 20834 | @group |
| 20835 | (let ((insert-list nil) |
| 20836 | (number-of-top-blanks |
| 20837 | (- max-graph-height actual-height))) |
| 20838 | |
| 20839 | ;; @r{Fill in @code{graph-symbols}.} |
| 20840 | (while (> actual-height 0) |
| 20841 | (setq insert-list (cons graph-symbol insert-list)) |
| 20842 | (setq actual-height (1- actual-height))) |
| 20843 | @end group |
| 20844 | |
| 20845 | @group |
| 20846 | ;; @r{Fill in @code{graph-blanks}.} |
| 20847 | (while (> number-of-top-blanks 0) |
| 20848 | (setq insert-list (cons graph-blank insert-list)) |
| 20849 | (setq number-of-top-blanks |
| 20850 | (1- number-of-top-blanks))) |
| 20851 | |
| 20852 | ;; @r{Return whole list.} |
| 20853 | insert-list)) |
| 20854 | @end group |
| 20855 | @end smallexample |
| 20856 | |
| 20857 | @smallexample |
| 20858 | @group |
| 20859 | (defun Y-axis-element (number full-Y-label-width) |
| 20860 | "Construct a NUMBERed label element. |
| 20861 | A numbered element looks like this ` 5 - ', |
| 20862 | and is padded as needed so all line up with |
| 20863 | the element for the largest number." |
| 20864 | @end group |
| 20865 | @group |
| 20866 | (let* ((leading-spaces |
| 20867 | (- full-Y-label-width |
| 20868 | (length |
| 20869 | (concat (number-to-string number) |
| 20870 | Y-axis-tic))))) |
| 20871 | @end group |
| 20872 | @group |
| 20873 | (concat |
| 20874 | (make-string leading-spaces ? ) |
| 20875 | (number-to-string number) |
| 20876 | Y-axis-tic))) |
| 20877 | @end group |
| 20878 | @end smallexample |
| 20879 | |
| 20880 | @smallexample |
| 20881 | @group |
| 20882 | (defun print-Y-axis |
| 20883 | (height full-Y-label-width &optional vertical-step) |
| 20884 | "Insert Y axis by HEIGHT and FULL-Y-LABEL-WIDTH. |
| 20885 | Height must be the maximum height of the graph. |
| 20886 | Full width is the width of the highest label element. |
| 20887 | Optionally, print according to VERTICAL-STEP." |
| 20888 | @end group |
| 20889 | @group |
| 20890 | ;; Value of height and full-Y-label-width |
| 20891 | ;; are passed by `print-graph'. |
| 20892 | (let ((start (point))) |
| 20893 | (insert-rectangle |
| 20894 | (Y-axis-column height full-Y-label-width vertical-step)) |
| 20895 | @end group |
| 20896 | @group |
| 20897 | ;; @r{Place point ready for inserting graph.} |
| 20898 | (goto-char start) |
| 20899 | ;; @r{Move point forward by value of} full-Y-label-width |
| 20900 | (forward-char full-Y-label-width))) |
| 20901 | @end group |
| 20902 | @end smallexample |
| 20903 | |
| 20904 | @smallexample |
| 20905 | @group |
| 20906 | (defun print-X-axis-tic-line |
| 20907 | (number-of-X-tics X-axis-leading-spaces X-axis-tic-element) |
| 20908 | "Print tics for X axis." |
| 20909 | (insert X-axis-leading-spaces) |
| 20910 | (insert X-axis-tic-symbol) ; @r{Under first column.} |
| 20911 | @end group |
| 20912 | @group |
| 20913 | ;; @r{Insert second tic in the right spot.} |
| 20914 | (insert (concat |
| 20915 | (make-string |
| 20916 | (- (* symbol-width X-axis-label-spacing) |
| 20917 | ;; @r{Insert white space up to second tic symbol.} |
| 20918 | (* 2 (length X-axis-tic-symbol))) |
| 20919 | ? ) |
| 20920 | X-axis-tic-symbol)) |
| 20921 | @end group |
| 20922 | @group |
| 20923 | ;; @r{Insert remaining tics.} |
| 20924 | (while (> number-of-X-tics 1) |
| 20925 | (insert X-axis-tic-element) |
| 20926 | (setq number-of-X-tics (1- number-of-X-tics)))) |
| 20927 | @end group |
| 20928 | @end smallexample |
| 20929 | |
| 20930 | @smallexample |
| 20931 | @group |
| 20932 | (defun X-axis-element (number) |
| 20933 | "Construct a numbered X axis element." |
| 20934 | (let ((leading-spaces |
| 20935 | (- (* symbol-width X-axis-label-spacing) |
| 20936 | (length (number-to-string number))))) |
| 20937 | (concat (make-string leading-spaces ? ) |
| 20938 | (number-to-string number)))) |
| 20939 | @end group |
| 20940 | @end smallexample |
| 20941 | |
| 20942 | @smallexample |
| 20943 | @group |
| 20944 | (defun graph-body-print (numbers-list height symbol-width) |
| 20945 | "Print a bar graph of the NUMBERS-LIST. |
| 20946 | The numbers-list consists of the Y-axis values. |
| 20947 | HEIGHT is maximum height of graph. |
| 20948 | SYMBOL-WIDTH is number of each column." |
| 20949 | @end group |
| 20950 | @group |
| 20951 | (let (from-position) |
| 20952 | (while numbers-list |
| 20953 | (setq from-position (point)) |
| 20954 | (insert-rectangle |
| 20955 | (column-of-graph height (car numbers-list))) |
| 20956 | (goto-char from-position) |
| 20957 | (forward-char symbol-width) |
| 20958 | @end group |
| 20959 | @group |
| 20960 | ;; @r{Draw graph column by column.} |
| 20961 | (sit-for 0) |
| 20962 | (setq numbers-list (cdr numbers-list))) |
| 20963 | ;; @r{Place point for X axis labels.} |
| 20964 | (forward-line height) |
| 20965 | (insert "\n"))) |
| 20966 | @end group |
| 20967 | @end smallexample |
| 20968 | |
| 20969 | @smallexample |
| 20970 | @group |
| 20971 | (defun Y-axis-column |
| 20972 | (height width-of-label &optional vertical-step) |
| 20973 | "Construct list of labels for Y axis. |
| 20974 | HEIGHT is maximum height of graph. |
| 20975 | WIDTH-OF-LABEL is maximum width of label. |
| 20976 | @end group |
| 20977 | @group |
| 20978 | VERTICAL-STEP, an option, is a positive integer |
| 20979 | that specifies how much a Y axis label increments |
| 20980 | for each line. For example, a step of 5 means |
| 20981 | that each line is five units of the graph." |
| 20982 | (let (Y-axis |
| 20983 | (number-per-line (or vertical-step 1))) |
| 20984 | @end group |
| 20985 | @group |
| 20986 | (while (> height 1) |
| 20987 | (if (zerop (% height Y-axis-label-spacing)) |
| 20988 | ;; @r{Insert label.} |
| 20989 | (setq Y-axis |
| 20990 | (cons |
| 20991 | (Y-axis-element |
| 20992 | (* height number-per-line) |
| 20993 | width-of-label) |
| 20994 | Y-axis)) |
| 20995 | @end group |
| 20996 | @group |
| 20997 | ;; @r{Else, insert blanks.} |
| 20998 | (setq Y-axis |
| 20999 | (cons |
| 21000 | (make-string width-of-label ? ) |
| 21001 | Y-axis))) |
| 21002 | (setq height (1- height))) |
| 21003 | @end group |
| 21004 | @group |
| 21005 | ;; @r{Insert base line.} |
| 21006 | (setq Y-axis (cons (Y-axis-element |
| 21007 | (or vertical-step 1) |
| 21008 | width-of-label) |
| 21009 | Y-axis)) |
| 21010 | (nreverse Y-axis))) |
| 21011 | @end group |
| 21012 | @end smallexample |
| 21013 | |
| 21014 | @smallexample |
| 21015 | @group |
| 21016 | (defun print-X-axis-numbered-line |
| 21017 | (number-of-X-tics X-axis-leading-spaces |
| 21018 | &optional horizontal-step) |
| 21019 | "Print line of X-axis numbers" |
| 21020 | (let ((number X-axis-label-spacing) |
| 21021 | (horizontal-step (or horizontal-step 1))) |
| 21022 | @end group |
| 21023 | @group |
| 21024 | (insert X-axis-leading-spaces) |
| 21025 | ;; line up number |
| 21026 | (delete-char (- (1- (length (number-to-string horizontal-step))))) |
| 21027 | (insert (concat |
| 21028 | (make-string |
| 21029 | ;; @r{Insert white space up to next number.} |
| 21030 | (- (* symbol-width X-axis-label-spacing) |
| 21031 | (1- (length (number-to-string horizontal-step))) |
| 21032 | 2) |
| 21033 | ? ) |
| 21034 | (number-to-string (* number horizontal-step)))) |
| 21035 | @end group |
| 21036 | @group |
| 21037 | ;; @r{Insert remaining numbers.} |
| 21038 | (setq number (+ number X-axis-label-spacing)) |
| 21039 | (while (> number-of-X-tics 1) |
| 21040 | (insert (X-axis-element (* number horizontal-step))) |
| 21041 | (setq number (+ number X-axis-label-spacing)) |
| 21042 | (setq number-of-X-tics (1- number-of-X-tics))))) |
| 21043 | @end group |
| 21044 | @end smallexample |
| 21045 | |
| 21046 | @smallexample |
| 21047 | @group |
| 21048 | (defun print-X-axis (numbers-list horizontal-step) |
| 21049 | "Print X axis labels to length of NUMBERS-LIST. |
| 21050 | Optionally, HORIZONTAL-STEP, a positive integer, |
| 21051 | specifies how much an X axis label increments for |
| 21052 | each column." |
| 21053 | @end group |
| 21054 | @group |
| 21055 | ;; Value of symbol-width and full-Y-label-width |
| 21056 | ;; are passed by `print-graph'. |
| 21057 | (let* ((leading-spaces |
| 21058 | (make-string full-Y-label-width ? )) |
| 21059 | ;; symbol-width @r{is provided by} graph-body-print |
| 21060 | (tic-width (* symbol-width X-axis-label-spacing)) |
| 21061 | (X-length (length numbers-list)) |
| 21062 | @end group |
| 21063 | @group |
| 21064 | (X-tic |
| 21065 | (concat |
| 21066 | (make-string |
| 21067 | ;; @r{Make a string of blanks.} |
| 21068 | (- (* symbol-width X-axis-label-spacing) |
| 21069 | (length X-axis-tic-symbol)) |
| 21070 | ? ) |
| 21071 | @end group |
| 21072 | @group |
| 21073 | ;; @r{Concatenate blanks with tic symbol.} |
| 21074 | X-axis-tic-symbol)) |
| 21075 | (tic-number |
| 21076 | (if (zerop (% X-length tic-width)) |
| 21077 | (/ X-length tic-width) |
| 21078 | (1+ (/ X-length tic-width))))) |
| 21079 | @end group |
| 21080 | |
| 21081 | @group |
| 21082 | (print-X-axis-tic-line |
| 21083 | tic-number leading-spaces X-tic) |
| 21084 | (insert "\n") |
| 21085 | (print-X-axis-numbered-line |
| 21086 | tic-number leading-spaces horizontal-step))) |
| 21087 | @end group |
| 21088 | @end smallexample |
| 21089 | |
| 21090 | @smallexample |
| 21091 | @group |
| 21092 | (defun one-fiftieth (full-range) |
| 21093 | "Return list, each number of which is 1/50th previous." |
| 21094 | (mapcar '(lambda (arg) (/ arg 50)) full-range)) |
| 21095 | @end group |
| 21096 | @end smallexample |
| 21097 | |
| 21098 | @smallexample |
| 21099 | @group |
| 21100 | (defun print-graph |
| 21101 | (numbers-list &optional vertical-step horizontal-step) |
| 21102 | "Print labelled bar graph of the NUMBERS-LIST. |
| 21103 | The numbers-list consists of the Y-axis values. |
| 21104 | @end group |
| 21105 | |
| 21106 | @group |
| 21107 | Optionally, VERTICAL-STEP, a positive integer, |
| 21108 | specifies how much a Y axis label increments for |
| 21109 | each line. For example, a step of 5 means that |
| 21110 | each row is five units. |
| 21111 | @end group |
| 21112 | |
| 21113 | @group |
| 21114 | Optionally, HORIZONTAL-STEP, a positive integer, |
| 21115 | specifies how much an X axis label increments for |
| 21116 | each column." |
| 21117 | (let* ((symbol-width (length graph-blank)) |
| 21118 | ;; @code{height} @r{is both the largest number} |
| 21119 | ;; @r{and the number with the most digits.} |
| 21120 | (height (apply 'max numbers-list)) |
| 21121 | @end group |
| 21122 | @group |
| 21123 | (height-of-top-line |
| 21124 | (if (zerop (% height Y-axis-label-spacing)) |
| 21125 | height |
| 21126 | ;; @r{else} |
| 21127 | (* (1+ (/ height Y-axis-label-spacing)) |
| 21128 | Y-axis-label-spacing))) |
| 21129 | @end group |
| 21130 | @group |
| 21131 | (vertical-step (or vertical-step 1)) |
| 21132 | (full-Y-label-width |
| 21133 | (length |
| 21134 | (concat |
| 21135 | (number-to-string |
| 21136 | (* height-of-top-line vertical-step)) |
| 21137 | Y-axis-tic)))) |
| 21138 | @end group |
| 21139 | @group |
| 21140 | |
| 21141 | (print-Y-axis |
| 21142 | height-of-top-line full-Y-label-width vertical-step) |
| 21143 | (graph-body-print |
| 21144 | numbers-list height-of-top-line symbol-width) |
| 21145 | (print-X-axis numbers-list horizontal-step))) |
| 21146 | @end group |
| 21147 | @end smallexample |
| 21148 | @c qqq |
| 21149 | @end ignore |
| 21150 | |
| 21151 | @page |
| 21152 | @node Final printed graph, , Another Bug, Print Whole Graph |
| 21153 | @appendixsubsec The Printed Graph |
| 21154 | |
| 21155 | When made and installed, you can call the @code{print-graph} command |
| 21156 | like this: |
| 21157 | @sp 1 |
| 21158 | |
| 21159 | @smallexample |
| 21160 | @group |
| 21161 | (print-graph fiftieth-list-for-graph 50 10) |
| 21162 | @end group |
| 21163 | @end smallexample |
| 21164 | @sp 1 |
| 21165 | |
| 21166 | @noindent |
| 21167 | Here is the graph: |
| 21168 | @sp 2 |
| 21169 | |
| 21170 | @smallexample |
| 21171 | @group |
| 21172 | 1000 - * |
| 21173 | ** |
| 21174 | ** |
| 21175 | ** |
| 21176 | ** |
| 21177 | 750 - *** |
| 21178 | *** |
| 21179 | *** |
| 21180 | *** |
| 21181 | **** |
| 21182 | 500 - ***** |
| 21183 | ****** |
| 21184 | ****** |
| 21185 | ****** |
| 21186 | ******* |
| 21187 | 250 - ******** |
| 21188 | ********* * |
| 21189 | *********** * |
| 21190 | ************* * |
| 21191 | 50 - ***************** * * |
| 21192 | | | | | | | | | |
| 21193 | 10 50 100 150 200 250 300 350 |
| 21194 | @end group |
| 21195 | @end smallexample |
| 21196 | |
| 21197 | @sp 2 |
| 21198 | |
| 21199 | @noindent |
| 21200 | The largest group of functions contain 10 -- 19 words and symbols each. |
| 21201 | |
| 21202 | @node Free Software and Free Manuals, GNU Free Documentation License, Full Graph, Top |
| 21203 | @appendix Free Software and Free Manuals |
| 21204 | |
| 21205 | @strong{by Richard M. Stallman} |
| 21206 | @sp 1 |
| 21207 | |
| 21208 | The biggest deficiency in free operating systems is not in the |
| 21209 | software---it is the lack of good free manuals that we can include in |
| 21210 | these systems. Many of our most important programs do not come with |
| 21211 | full manuals. Documentation is an essential part of any software |
| 21212 | package; when an important free software package does not come with a |
| 21213 | free manual, that is a major gap. We have many such gaps today. |
| 21214 | |
| 21215 | Once upon a time, many years ago, I thought I would learn Perl. I got |
| 21216 | a copy of a free manual, but I found it hard to read. When I asked |
| 21217 | Perl users about alternatives, they told me that there were better |
| 21218 | introductory manuals---but those were not free. |
| 21219 | |
| 21220 | Why was this? The authors of the good manuals had written them for |
| 21221 | O'Reilly Associates, which published them with restrictive terms---no |
| 21222 | copying, no modification, source files not available---which exclude |
| 21223 | them from the free software community. |
| 21224 | |
| 21225 | That wasn't the first time this sort of thing has happened, and (to |
| 21226 | our community's great loss) it was far from the last. Proprietary |
| 21227 | manual publishers have enticed a great many authors to restrict their |
| 21228 | manuals since then. Many times I have heard a GNU user eagerly tell me |
| 21229 | about a manual that he is writing, with which he expects to help the |
| 21230 | GNU project---and then had my hopes dashed, as he proceeded to explain |
| 21231 | that he had signed a contract with a publisher that would restrict it |
| 21232 | so that we cannot use it. |
| 21233 | |
| 21234 | Given that writing good English is a rare skill among programmers, we |
| 21235 | can ill afford to lose manuals this way. |
| 21236 | |
| 21237 | @c (texinfo)uref |
| 21238 | (The Free Software Foundation |
| 21239 | @uref{http://www.gnu.org/doc/doc.html#DescriptionsOfGNUDocumentation, , |
| 21240 | sells printed copies} of free @uref{http://www.gnu.org/doc/doc.html, |
| 21241 | GNU manuals}, too.) |
| 21242 | |
| 21243 | Free documentation, like free software, is a matter of freedom, not |
| 21244 | price. The problem with these manuals was not that O'Reilly Associates |
| 21245 | charged a price for printed copies---that in itself is fine. (The Free |
| 21246 | Software Foundation sells printed copies of free GNU manuals, too.) |
| 21247 | But GNU manuals are available in source code form, while these manuals |
| 21248 | are available only on paper. GNU manuals come with permission to copy |
| 21249 | and modify; the Perl manuals do not. These restrictions are the |
| 21250 | problems. |
| 21251 | |
| 21252 | The criterion for a free manual is pretty much the same as for free |
| 21253 | software: it is a matter of giving all users certain |
| 21254 | freedoms. Redistribution (including commercial redistribution) must be |
| 21255 | permitted, so that the manual can accompany every copy of the program, |
| 21256 | on-line or on paper. Permission for modification is crucial too. |
| 21257 | |
| 21258 | As a general rule, I don't believe that it is essential for people to |
| 21259 | have permission to modify all sorts of articles and books. The issues |
| 21260 | for writings are not necessarily the same as those for software. For |
| 21261 | example, I don't think you or I are obliged to give permission to |
| 21262 | modify articles like this one, which describe our actions and our |
| 21263 | views. |
| 21264 | |
| 21265 | But there is a particular reason why the freedom to modify is crucial |
| 21266 | for documentation for free software. When people exercise their right |
| 21267 | to modify the software, and add or change its features, if they are |
| 21268 | conscientious they will change the manual too---so they can provide |
| 21269 | accurate and usable documentation with the modified program. A manual |
| 21270 | which forbids programmers to be conscientious and finish the job, or |
| 21271 | more precisely requires them to write a new manual from scratch if |
| 21272 | they change the program, does not fill our community's needs. |
| 21273 | |
| 21274 | While a blanket prohibition on modification is unacceptable, some |
| 21275 | kinds of limits on the method of modification pose no problem. For |
| 21276 | example, requirements to preserve the original author's copyright |
| 21277 | notice, the distribution terms, or the list of authors, are ok. It is |
| 21278 | also no problem to require modified versions to include notice that |
| 21279 | they were modified, even to have entire sections that may not be |
| 21280 | deleted or changed, as long as these sections deal with nontechnical |
| 21281 | topics. (Some GNU manuals have them.) |
| 21282 | |
| 21283 | These kinds of restrictions are not a problem because, as a practical |
| 21284 | matter, they don't stop the conscientious programmer from adapting the |
| 21285 | manual to fit the modified program. In other words, they don't block |
| 21286 | the free software community from making full use of the manual. |
| 21287 | |
| 21288 | However, it must be possible to modify all the technical content of |
| 21289 | the manual, and then distribute the result in all the usual media, |
| 21290 | through all the usual channels; otherwise, the restrictions do block |
| 21291 | the community, the manual is not free, and so we need another manual. |
| 21292 | |
| 21293 | Unfortunately, it is often hard to find someone to write another |
| 21294 | manual when a proprietary manual exists. The obstacle is that many |
| 21295 | users think that a proprietary manual is good enough---so they don't |
| 21296 | see the need to write a free manual. They do not see that the free |
| 21297 | operating system has a gap that needs filling. |
| 21298 | |
| 21299 | Why do users think that proprietary manuals are good enough? Some have |
| 21300 | not considered the issue. I hope this article will do something to |
| 21301 | change that. |
| 21302 | |
| 21303 | Other users consider proprietary manuals acceptable for the same |
| 21304 | reason so many people consider proprietary software acceptable: they |
| 21305 | judge in purely practical terms, not using freedom as a |
| 21306 | criterion. These people are entitled to their opinions, but since |
| 21307 | those opinions spring from values which do not include freedom, they |
| 21308 | are no guide for those of us who do value freedom. |
| 21309 | |
| 21310 | Please spread the word about this issue. We continue to lose manuals |
| 21311 | to proprietary publishing. If we spread the word that proprietary |
| 21312 | manuals are not sufficient, perhaps the next person who wants to help |
| 21313 | GNU by writing documentation will realize, before it is too late, that |
| 21314 | he must above all make it free. |
| 21315 | |
| 21316 | We can also encourage commercial publishers to sell free, copylefted |
| 21317 | manuals instead of proprietary ones. One way you can help this is to |
| 21318 | check the distribution terms of a manual before you buy it, and prefer |
| 21319 | copylefted manuals to non-copylefted ones. |
| 21320 | |
| 21321 | @sp 2 |
| 21322 | @noindent |
| 21323 | Note: The Free Software Foundation maintains a page on its Web site |
| 21324 | that lists free books available from other publishers:@* |
| 21325 | @uref{http://www.gnu.org/doc/other-free-books.html} |
| 21326 | |
| 21327 | |
| 21328 | @node GNU Free Documentation License, Index, Free Software and Free Manuals, Top |
| 21329 | @appendix GNU Free Documentation License |
| 21330 | |
| 21331 | @cindex FDL, GNU Free Documentation License |
| 21332 | @center Version 1.2, November 2002 |
| 21333 | |
| 21334 | @display |
| 21335 | Copyright @copyright{} 2000,2001,2002 Free Software Foundation, Inc. |
| 21336 | 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA |
| 21337 | |
| 21338 | Everyone is permitted to copy and distribute verbatim copies |
| 21339 | of this license document, but changing it is not allowed. |
| 21340 | @end display |
| 21341 | |
| 21342 | @enumerate 0 |
| 21343 | @item |
| 21344 | PREAMBLE |
| 21345 | |
| 21346 | The purpose of this License is to make a manual, textbook, or other |
| 21347 | functional and useful document @dfn{free} in the sense of freedom: to |
| 21348 | assure everyone the effective freedom to copy and redistribute it, |
| 21349 | with or without modifying it, either commercially or noncommercially. |
| 21350 | Secondarily, this License preserves for the author and publisher a way |
| 21351 | to get credit for their work, while not being considered responsible |
| 21352 | for modifications made by others. |
| 21353 | |
| 21354 | This License is a kind of ``copyleft'', which means that derivative |
| 21355 | works of the document must themselves be free in the same sense. It |
| 21356 | complements the GNU General Public License, which is a copyleft |
| 21357 | license designed for free software. |
| 21358 | |
| 21359 | We have designed this License in order to use it for manuals for free |
| 21360 | software, because free software needs free documentation: a free |
| 21361 | program should come with manuals providing the same freedoms that the |
| 21362 | software does. But this License is not limited to software manuals; |
| 21363 | it can be used for any textual work, regardless of subject matter or |
| 21364 | whether it is published as a printed book. We recommend this License |
| 21365 | principally for works whose purpose is instruction or reference. |
| 21366 | |
| 21367 | @item |
| 21368 | APPLICABILITY AND DEFINITIONS |
| 21369 | |
| 21370 | This License applies to any manual or other work, in any medium, that |
| 21371 | contains a notice placed by the copyright holder saying it can be |
| 21372 | distributed under the terms of this License. Such a notice grants a |
| 21373 | world-wide, royalty-free license, unlimited in duration, to use that |
| 21374 | work under the conditions stated herein. The ``Document'', below, |
| 21375 | refers to any such manual or work. Any member of the public is a |
| 21376 | licensee, and is addressed as ``you''. You accept the license if you |
| 21377 | copy, modify or distribute the work in a way requiring permission |
| 21378 | under copyright law. |
| 21379 | |
| 21380 | A ``Modified Version'' of the Document means any work containing the |
| 21381 | Document or a portion of it, either copied verbatim, or with |
| 21382 | modifications and/or translated into another language. |
| 21383 | |
| 21384 | A ``Secondary Section'' is a named appendix or a front-matter section |
| 21385 | of the Document that deals exclusively with the relationship of the |
| 21386 | publishers or authors of the Document to the Document's overall |
| 21387 | subject (or to related matters) and contains nothing that could fall |
| 21388 | directly within that overall subject. (Thus, if the Document is in |
| 21389 | part a textbook of mathematics, a Secondary Section may not explain |
| 21390 | any mathematics.) The relationship could be a matter of historical |
| 21391 | connection with the subject or with related matters, or of legal, |
| 21392 | commercial, philosophical, ethical or political position regarding |
| 21393 | them. |
| 21394 | |
| 21395 | The ``Invariant Sections'' are certain Secondary Sections whose titles |
| 21396 | are designated, as being those of Invariant Sections, in the notice |
| 21397 | that says that the Document is released under this License. If a |
| 21398 | section does not fit the above definition of Secondary then it is not |
| 21399 | allowed to be designated as Invariant. The Document may contain zero |
| 21400 | Invariant Sections. If the Document does not identify any Invariant |
| 21401 | Sections then there are none. |
| 21402 | |
| 21403 | The ``Cover Texts'' are certain short passages of text that are listed, |
| 21404 | as Front-Cover Texts or Back-Cover Texts, in the notice that says that |
| 21405 | the Document is released under this License. A Front-Cover Text may |
| 21406 | be at most 5 words, and a Back-Cover Text may be at most 25 words. |
| 21407 | |
| 21408 | A ``Transparent'' copy of the Document means a machine-readable copy, |
| 21409 | represented in a format whose specification is available to the |
| 21410 | general public, that is suitable for revising the document |
| 21411 | straightforwardly with generic text editors or (for images composed of |
| 21412 | pixels) generic paint programs or (for drawings) some widely available |
| 21413 | drawing editor, and that is suitable for input to text formatters or |
| 21414 | for automatic translation to a variety of formats suitable for input |
| 21415 | to text formatters. A copy made in an otherwise Transparent file |
| 21416 | format whose markup, or absence of markup, has been arranged to thwart |
| 21417 | or discourage subsequent modification by readers is not Transparent. |
| 21418 | An image format is not Transparent if used for any substantial amount |
| 21419 | of text. A copy that is not ``Transparent'' is called ``Opaque''. |
| 21420 | |
| 21421 | Examples of suitable formats for Transparent copies include plain |
| 21422 | @sc{ascii} without markup, Texinfo input format, La@TeX{} input |
| 21423 | format, @acronym{SGML} or @acronym{XML} using a publicly available |
| 21424 | @acronym{DTD}, and standard-conforming simple @acronym{HTML}, |
| 21425 | PostScript or @acronym{PDF} designed for human modification. Examples |
| 21426 | of transparent image formats include @acronym{PNG}, @acronym{XCF} and |
| 21427 | @acronym{JPG}. Opaque formats include proprietary formats that can be |
| 21428 | read and edited only by proprietary word processors, @acronym{SGML} or |
| 21429 | @acronym{XML} for which the @acronym{DTD} and/or processing tools are |
| 21430 | not generally available, and the machine-generated @acronym{HTML}, |
| 21431 | PostScript or @acronym{PDF} produced by some word processors for |
| 21432 | output purposes only. |
| 21433 | |
| 21434 | The ``Title Page'' means, for a printed book, the title page itself, |
| 21435 | plus such following pages as are needed to hold, legibly, the material |
| 21436 | this License requires to appear in the title page. For works in |
| 21437 | formats which do not have any title page as such, ``Title Page'' means |
| 21438 | the text near the most prominent appearance of the work's title, |
| 21439 | preceding the beginning of the body of the text. |
| 21440 | |
| 21441 | A section ``Entitled XYZ'' means a named subunit of the Document whose |
| 21442 | title either is precisely XYZ or contains XYZ in parentheses following |
| 21443 | text that translates XYZ in another language. (Here XYZ stands for a |
| 21444 | specific section name mentioned below, such as ``Acknowledgements'', |
| 21445 | ``Dedications'', ``Endorsements'', or ``History''.) To ``Preserve the Title'' |
| 21446 | of such a section when you modify the Document means that it remains a |
| 21447 | section ``Entitled XYZ'' according to this definition. |
| 21448 | |
| 21449 | The Document may include Warranty Disclaimers next to the notice which |
| 21450 | states that this License applies to the Document. These Warranty |
| 21451 | Disclaimers are considered to be included by reference in this |
| 21452 | License, but only as regards disclaiming warranties: any other |
| 21453 | implication that these Warranty Disclaimers may have is void and has |
| 21454 | no effect on the meaning of this License. |
| 21455 | |
| 21456 | @item |
| 21457 | VERBATIM COPYING |
| 21458 | |
| 21459 | You may copy and distribute the Document in any medium, either |
| 21460 | commercially or noncommercially, provided that this License, the |
| 21461 | copyright notices, and the license notice saying this License applies |
| 21462 | to the Document are reproduced in all copies, and that you add no other |
| 21463 | conditions whatsoever to those of this License. You may not use |
| 21464 | technical measures to obstruct or control the reading or further |
| 21465 | copying of the copies you make or distribute. However, you may accept |
| 21466 | compensation in exchange for copies. If you distribute a large enough |
| 21467 | number of copies you must also follow the conditions in section 3. |
| 21468 | |
| 21469 | You may also lend copies, under the same conditions stated above, and |
| 21470 | you may publicly display copies. |
| 21471 | |
| 21472 | @item |
| 21473 | COPYING IN QUANTITY |
| 21474 | |
| 21475 | If you publish printed copies (or copies in media that commonly have |
| 21476 | printed covers) of the Document, numbering more than 100, and the |
| 21477 | Document's license notice requires Cover Texts, you must enclose the |
| 21478 | copies in covers that carry, clearly and legibly, all these Cover |
| 21479 | Texts: Front-Cover Texts on the front cover, and Back-Cover Texts on |
| 21480 | the back cover. Both covers must also clearly and legibly identify |
| 21481 | you as the publisher of these copies. The front cover must present |
| 21482 | the full title with all words of the title equally prominent and |
| 21483 | visible. You may add other material on the covers in addition. |
| 21484 | Copying with changes limited to the covers, as long as they preserve |
| 21485 | the title of the Document and satisfy these conditions, can be treated |
| 21486 | as verbatim copying in other respects. |
| 21487 | |
| 21488 | If the required texts for either cover are too voluminous to fit |
| 21489 | legibly, you should put the first ones listed (as many as fit |
| 21490 | reasonably) on the actual cover, and continue the rest onto adjacent |
| 21491 | pages. |
| 21492 | |
| 21493 | If you publish or distribute Opaque copies of the Document numbering |
| 21494 | more than 100, you must either include a machine-readable Transparent |
| 21495 | copy along with each Opaque copy, or state in or with each Opaque copy |
| 21496 | a computer-network location from which the general network-using |
| 21497 | public has access to download using public-standard network protocols |
| 21498 | a complete Transparent copy of the Document, free of added material. |
| 21499 | If you use the latter option, you must take reasonably prudent steps, |
| 21500 | when you begin distribution of Opaque copies in quantity, to ensure |
| 21501 | that this Transparent copy will remain thus accessible at the stated |
| 21502 | location until at least one year after the last time you distribute an |
| 21503 | Opaque copy (directly or through your agents or retailers) of that |
| 21504 | edition to the public. |
| 21505 | |
| 21506 | It is requested, but not required, that you contact the authors of the |
| 21507 | Document well before redistributing any large number of copies, to give |
| 21508 | them a chance to provide you with an updated version of the Document. |
| 21509 | |
| 21510 | @item |
| 21511 | MODIFICATIONS |
| 21512 | |
| 21513 | You may copy and distribute a Modified Version of the Document under |
| 21514 | the conditions of sections 2 and 3 above, provided that you release |
| 21515 | the Modified Version under precisely this License, with the Modified |
| 21516 | Version filling the role of the Document, thus licensing distribution |
| 21517 | and modification of the Modified Version to whoever possesses a copy |
| 21518 | of it. In addition, you must do these things in the Modified Version: |
| 21519 | |
| 21520 | @enumerate A |
| 21521 | @item |
| 21522 | Use in the Title Page (and on the covers, if any) a title distinct |
| 21523 | from that of the Document, and from those of previous versions |
| 21524 | (which should, if there were any, be listed in the History section |
| 21525 | of the Document). You may use the same title as a previous version |
| 21526 | if the original publisher of that version gives permission. |
| 21527 | |
| 21528 | @item |
| 21529 | List on the Title Page, as authors, one or more persons or entities |
| 21530 | responsible for authorship of the modifications in the Modified |
| 21531 | Version, together with at least five of the principal authors of the |
| 21532 | Document (all of its principal authors, if it has fewer than five), |
| 21533 | unless they release you from this requirement. |
| 21534 | |
| 21535 | @item |
| 21536 | State on the Title page the name of the publisher of the |
| 21537 | Modified Version, as the publisher. |
| 21538 | |
| 21539 | @item |
| 21540 | Preserve all the copyright notices of the Document. |
| 21541 | |
| 21542 | @item |
| 21543 | Add an appropriate copyright notice for your modifications |
| 21544 | adjacent to the other copyright notices. |
| 21545 | |
| 21546 | @item |
| 21547 | Include, immediately after the copyright notices, a license notice |
| 21548 | giving the public permission to use the Modified Version under the |
| 21549 | terms of this License, in the form shown in the Addendum below. |
| 21550 | |
| 21551 | @item |
| 21552 | Preserve in that license notice the full lists of Invariant Sections |
| 21553 | and required Cover Texts given in the Document's license notice. |
| 21554 | |
| 21555 | @item |
| 21556 | Include an unaltered copy of this License. |
| 21557 | |
| 21558 | @item |
| 21559 | Preserve the section Entitled ``History'', Preserve its Title, and add |
| 21560 | to it an item stating at least the title, year, new authors, and |
| 21561 | publisher of the Modified Version as given on the Title Page. If |
| 21562 | there is no section Entitled ``History'' in the Document, create one |
| 21563 | stating the title, year, authors, and publisher of the Document as |
| 21564 | given on its Title Page, then add an item describing the Modified |
| 21565 | Version as stated in the previous sentence. |
| 21566 | |
| 21567 | @item |
| 21568 | Preserve the network location, if any, given in the Document for |
| 21569 | public access to a Transparent copy of the Document, and likewise |
| 21570 | the network locations given in the Document for previous versions |
| 21571 | it was based on. These may be placed in the ``History'' section. |
| 21572 | You may omit a network location for a work that was published at |
| 21573 | least four years before the Document itself, or if the original |
| 21574 | publisher of the version it refers to gives permission. |
| 21575 | |
| 21576 | @item |
| 21577 | For any section Entitled ``Acknowledgements'' or ``Dedications'', Preserve |
| 21578 | the Title of the section, and preserve in the section all the |
| 21579 | substance and tone of each of the contributor acknowledgements and/or |
| 21580 | dedications given therein. |
| 21581 | |
| 21582 | @item |
| 21583 | Preserve all the Invariant Sections of the Document, |
| 21584 | unaltered in their text and in their titles. Section numbers |
| 21585 | or the equivalent are not considered part of the section titles. |
| 21586 | |
| 21587 | @item |
| 21588 | Delete any section Entitled ``Endorsements''. Such a section |
| 21589 | may not be included in the Modified Version. |
| 21590 | |
| 21591 | @item |
| 21592 | Do not retitle any existing section to be Entitled ``Endorsements'' or |
| 21593 | to conflict in title with any Invariant Section. |
| 21594 | |
| 21595 | @item |
| 21596 | Preserve any Warranty Disclaimers. |
| 21597 | @end enumerate |
| 21598 | |
| 21599 | If the Modified Version includes new front-matter sections or |
| 21600 | appendices that qualify as Secondary Sections and contain no material |
| 21601 | copied from the Document, you may at your option designate some or all |
| 21602 | of these sections as invariant. To do this, add their titles to the |
| 21603 | list of Invariant Sections in the Modified Version's license notice. |
| 21604 | These titles must be distinct from any other section titles. |
| 21605 | |
| 21606 | You may add a section Entitled ``Endorsements'', provided it contains |
| 21607 | nothing but endorsements of your Modified Version by various |
| 21608 | parties---for example, statements of peer review or that the text has |
| 21609 | been approved by an organization as the authoritative definition of a |
| 21610 | standard. |
| 21611 | |
| 21612 | You may add a passage of up to five words as a Front-Cover Text, and a |
| 21613 | passage of up to 25 words as a Back-Cover Text, to the end of the list |
| 21614 | of Cover Texts in the Modified Version. Only one passage of |
| 21615 | Front-Cover Text and one of Back-Cover Text may be added by (or |
| 21616 | through arrangements made by) any one entity. If the Document already |
| 21617 | includes a cover text for the same cover, previously added by you or |
| 21618 | by arrangement made by the same entity you are acting on behalf of, |
| 21619 | you may not add another; but you may replace the old one, on explicit |
| 21620 | permission from the previous publisher that added the old one. |
| 21621 | |
| 21622 | The author(s) and publisher(s) of the Document do not by this License |
| 21623 | give permission to use their names for publicity for or to assert or |
| 21624 | imply endorsement of any Modified Version. |
| 21625 | |
| 21626 | @item |
| 21627 | COMBINING DOCUMENTS |
| 21628 | |
| 21629 | You may combine the Document with other documents released under this |
| 21630 | License, under the terms defined in section 4 above for modified |
| 21631 | versions, provided that you include in the combination all of the |
| 21632 | Invariant Sections of all of the original documents, unmodified, and |
| 21633 | list them all as Invariant Sections of your combined work in its |
| 21634 | license notice, and that you preserve all their Warranty Disclaimers. |
| 21635 | |
| 21636 | The combined work need only contain one copy of this License, and |
| 21637 | multiple identical Invariant Sections may be replaced with a single |
| 21638 | copy. If there are multiple Invariant Sections with the same name but |
| 21639 | different contents, make the title of each such section unique by |
| 21640 | adding at the end of it, in parentheses, the name of the original |
| 21641 | author or publisher of that section if known, or else a unique number. |
| 21642 | Make the same adjustment to the section titles in the list of |
| 21643 | Invariant Sections in the license notice of the combined work. |
| 21644 | |
| 21645 | In the combination, you must combine any sections Entitled ``History'' |
| 21646 | in the various original documents, forming one section Entitled |
| 21647 | ``History''; likewise combine any sections Entitled ``Acknowledgements'', |
| 21648 | and any sections Entitled ``Dedications''. You must delete all |
| 21649 | sections Entitled ``Endorsements.'' |
| 21650 | |
| 21651 | @item |
| 21652 | COLLECTIONS OF DOCUMENTS |
| 21653 | |
| 21654 | You may make a collection consisting of the Document and other documents |
| 21655 | released under this License, and replace the individual copies of this |
| 21656 | License in the various documents with a single copy that is included in |
| 21657 | the collection, provided that you follow the rules of this License for |
| 21658 | verbatim copying of each of the documents in all other respects. |
| 21659 | |
| 21660 | You may extract a single document from such a collection, and distribute |
| 21661 | it individually under this License, provided you insert a copy of this |
| 21662 | License into the extracted document, and follow this License in all |
| 21663 | other respects regarding verbatim copying of that document. |
| 21664 | |
| 21665 | @item |
| 21666 | AGGREGATION WITH INDEPENDENT WORKS |
| 21667 | |
| 21668 | A compilation of the Document or its derivatives with other separate |
| 21669 | and independent documents or works, in or on a volume of a storage or |
| 21670 | distribution medium, is called an ``aggregate'' if the copyright |
| 21671 | resulting from the compilation is not used to limit the legal rights |
| 21672 | of the compilation's users beyond what the individual works permit. |
| 21673 | When the Document is included in an aggregate, this License does not |
| 21674 | apply to the other works in the aggregate which are not themselves |
| 21675 | derivative works of the Document. |
| 21676 | |
| 21677 | If the Cover Text requirement of section 3 is applicable to these |
| 21678 | copies of the Document, then if the Document is less than one half of |
| 21679 | the entire aggregate, the Document's Cover Texts may be placed on |
| 21680 | covers that bracket the Document within the aggregate, or the |
| 21681 | electronic equivalent of covers if the Document is in electronic form. |
| 21682 | Otherwise they must appear on printed covers that bracket the whole |
| 21683 | aggregate. |
| 21684 | |
| 21685 | @item |
| 21686 | TRANSLATION |
| 21687 | |
| 21688 | Translation is considered a kind of modification, so you may |
| 21689 | distribute translations of the Document under the terms of section 4. |
| 21690 | Replacing Invariant Sections with translations requires special |
| 21691 | permission from their copyright holders, but you may include |
| 21692 | translations of some or all Invariant Sections in addition to the |
| 21693 | original versions of these Invariant Sections. You may include a |
| 21694 | translation of this License, and all the license notices in the |
| 21695 | Document, and any Warranty Disclaimers, provided that you also include |
| 21696 | the original English version of this License and the original versions |
| 21697 | of those notices and disclaimers. In case of a disagreement between |
| 21698 | the translation and the original version of this License or a notice |
| 21699 | or disclaimer, the original version will prevail. |
| 21700 | |
| 21701 | If a section in the Document is Entitled ``Acknowledgements'', |
| 21702 | ``Dedications'', or ``History'', the requirement (section 4) to Preserve |
| 21703 | its Title (section 1) will typically require changing the actual |
| 21704 | title. |
| 21705 | |
| 21706 | @item |
| 21707 | TERMINATION |
| 21708 | |
| 21709 | You may not copy, modify, sublicense, or distribute the Document except |
| 21710 | as expressly provided for under this License. Any other attempt to |
| 21711 | copy, modify, sublicense or distribute the Document is void, and will |
| 21712 | automatically terminate your rights under this License. However, |
| 21713 | parties who have received copies, or rights, from you under this |
| 21714 | License will not have their licenses terminated so long as such |
| 21715 | parties remain in full compliance. |
| 21716 | |
| 21717 | @item |
| 21718 | FUTURE REVISIONS OF THIS LICENSE |
| 21719 | |
| 21720 | The Free Software Foundation may publish new, revised versions |
| 21721 | of the GNU Free Documentation License from time to time. Such new |
| 21722 | versions will be similar in spirit to the present version, but may |
| 21723 | differ in detail to address new problems or concerns. See |
| 21724 | @uref{http://www.gnu.org/copyleft/}. |
| 21725 | |
| 21726 | Each version of the License is given a distinguishing version number. |
| 21727 | If the Document specifies that a particular numbered version of this |
| 21728 | License ``or any later version'' applies to it, you have the option of |
| 21729 | following the terms and conditions either of that specified version or |
| 21730 | of any later version that has been published (not as a draft) by the |
| 21731 | Free Software Foundation. If the Document does not specify a version |
| 21732 | number of this License, you may choose any version ever published (not |
| 21733 | as a draft) by the Free Software Foundation. |
| 21734 | @end enumerate |
| 21735 | |
| 21736 | @page |
| 21737 | @appendixsubsec ADDENDUM: How to use this License for your documents |
| 21738 | |
| 21739 | To use this License in a document you have written, include a copy of |
| 21740 | the License in the document and put the following copyright and |
| 21741 | license notices just after the title page: |
| 21742 | |
| 21743 | @smallexample |
| 21744 | @group |
| 21745 | Copyright (C) @var{year} @var{your name}. |
| 21746 | Permission is granted to copy, distribute and/or modify this document |
| 21747 | under the terms of the GNU Free Documentation License, Version 1.2 |
| 21748 | or any later version published by the Free Software Foundation; |
| 21749 | with no Invariant Sections, no Front-Cover Texts, and no Back-Cover Texts. |
| 21750 | A copy of the license is included in the section entitled ``GNU |
| 21751 | Free Documentation License''. |
| 21752 | @end group |
| 21753 | @end smallexample |
| 21754 | |
| 21755 | If you have Invariant Sections, Front-Cover Texts and Back-Cover Texts, |
| 21756 | replace the ``with...Texts.'' line with this: |
| 21757 | |
| 21758 | @smallexample |
| 21759 | @group |
| 21760 | with the Invariant Sections being @var{list their titles}, with |
| 21761 | the Front-Cover Texts being @var{list}, and with the Back-Cover Texts |
| 21762 | being @var{list}. |
| 21763 | @end group |
| 21764 | @end smallexample |
| 21765 | |
| 21766 | If you have Invariant Sections without Cover Texts, or some other |
| 21767 | combination of the three, merge those two alternatives to suit the |
| 21768 | situation. |
| 21769 | |
| 21770 | If your document contains nontrivial examples of program code, we |
| 21771 | recommend releasing these examples in parallel under your choice of |
| 21772 | free software license, such as the GNU General Public License, |
| 21773 | to permit their use in free software. |
| 21774 | |
| 21775 | @node Index, About the Author, GNU Free Documentation License, Top |
| 21776 | @comment node-name, next, previous, up |
| 21777 | @unnumbered Index |
| 21778 | |
| 21779 | @ignore |
| 21780 | MENU ENTRY: NODE NAME. |
| 21781 | @end ignore |
| 21782 | |
| 21783 | @printindex cp |
| 21784 | |
| 21785 | @iftex |
| 21786 | @c Place biographical information on right-hand (verso) page |
| 21787 | |
| 21788 | @tex |
| 21789 | \ifodd\pageno |
| 21790 | \par\vfill\supereject |
| 21791 | \global\evenheadline={\hfil} \global\evenfootline={\hfil} |
| 21792 | \global\oddheadline={\hfil} \global\oddfootline={\hfil} |
| 21793 | \page\hbox{}\page |
| 21794 | \else |
| 21795 | \par\vfill\supereject |
| 21796 | \par\vfill\supereject |
| 21797 | \global\evenheadline={\hfil} \global\evenfootline={\hfil} |
| 21798 | \global\oddheadline={\hfil} \global\oddfootline={\hfil} |
| 21799 | \page\hbox{}\page |
| 21800 | \page\hbox{}\page |
| 21801 | \fi |
| 21802 | @end tex |
| 21803 | |
| 21804 | @page |
| 21805 | @w{ } |
| 21806 | |
| 21807 | @c ================ Biographical information ================ |
| 21808 | |
| 21809 | @w{ } |
| 21810 | @sp 8 |
| 21811 | @center About the Author |
| 21812 | @sp 1 |
| 21813 | @end iftex |
| 21814 | |
| 21815 | @ifnottex |
| 21816 | @node About the Author, , Index, Top |
| 21817 | @unnumbered About the Author |
| 21818 | @end ifnottex |
| 21819 | |
| 21820 | @quotation |
| 21821 | Robert J. Chassell has worked with GNU Emacs since 1985. He writes |
| 21822 | and edits, teaches Emacs and Emacs Lisp, and speaks throughout the |
| 21823 | world on software freedom. Chassell was a founding Director and |
| 21824 | Treasurer of the Free Software Foundation, Inc. He is co-author of |
| 21825 | the @cite{Texinfo} manual, and has edited more than a dozen other |
| 21826 | books. He graduated from Cambridge University, in England. He has an |
| 21827 | abiding interest in social and economic history and flies his own |
| 21828 | airplane. |
| 21829 | @end quotation |
| 21830 | |
| 21831 | @page |
| 21832 | @w{ } |
| 21833 | |
| 21834 | @c Prevent page number on blank verso, so eject it first. |
| 21835 | @tex |
| 21836 | \par\vfill\supereject |
| 21837 | @end tex |
| 21838 | |
| 21839 | @iftex |
| 21840 | @headings off |
| 21841 | @evenheading @thispage @| @| @thistitle |
| 21842 | @oddheading @| @| @thispage |
| 21843 | @end iftex |
| 21844 | |
| 21845 | @bye |
| 21846 | |
| 21847 | @ignore |
| 21848 | arch-tag: da1a2154-531f-43a8-8e33-fc7faad10acf |
| 21849 | @end ignore |