Commit | Line | Data |
---|---|---|
6bf7aab6 DL |
1 | \input texinfo @c -*-texinfo-*- |
2 | @setfilename ../info/cl | |
3 | @settitle Common Lisp Extensions | |
4 | ||
18f952d5 | 5 | @copying |
6bf7aab6 DL |
6 | This file documents the GNU Emacs Common Lisp emulation package. |
7 | ||
62eda0e2 GM |
8 | Copyright @copyright{} 1993, 2001, 2002, 2003, 2004, 2005, 2006, 2007 |
9 | Free Software Foundation, Inc. | |
308374ca | 10 | |
18f952d5 | 11 | @quotation |
308374ca | 12 | Permission is granted to copy, distribute and/or modify this document |
678e7c71 | 13 | under the terms of the GNU Free Documentation License, Version 1.2 or |
308374ca DL |
14 | any later version published by the Free Software Foundation; with no |
15 | Invariant Sections, with the Front-Cover texts being ``A GNU | |
16 | Manual'', and with the Back-Cover Texts as in (a) below. A copy of the | |
17 | license is included in the section entitled ``GNU Free Documentation | |
18 | License'' in the Emacs manual. | |
19 | ||
20 | (a) The FSF's Back-Cover Text is: ``You have freedom to copy and modify | |
21 | this GNU Manual, like GNU software. Copies published by the Free | |
22 | Software Foundation raise funds for GNU development.'' | |
23 | ||
24 | This document is part of a collection distributed under the GNU Free | |
25 | Documentation License. If you want to distribute this document | |
26 | separately from the collection, you can do so by adding a copy of the | |
27 | license to the document, as described in section 6 of the license. | |
18f952d5 KB |
28 | @end quotation |
29 | @end copying | |
30 | ||
31 | @dircategory Emacs | |
32 | @direntry | |
33 | * CL: (cl). Partial Common Lisp support for Emacs Lisp. | |
34 | @end direntry | |
35 | ||
36 | @finalout | |
6bf7aab6 DL |
37 | |
38 | @titlepage | |
39 | @sp 6 | |
40 | @center @titlefont{Common Lisp Extensions} | |
41 | @sp 4 | |
42 | @center For GNU Emacs Lisp | |
43 | @sp 1 | |
44 | @center Version 2.02 | |
45 | @sp 5 | |
46 | @center Dave Gillespie | |
47 | @center daveg@@synaptics.com | |
48 | @page | |
6bf7aab6 | 49 | @vskip 0pt plus 1filll |
18f952d5 | 50 | @insertcopying |
6bf7aab6 DL |
51 | @end titlepage |
52 | ||
c6ad01e8 | 53 | @node Top, Overview, (dir), (dir) |
3105a2cc | 54 | @chapter Introduction |
6bf7aab6 DL |
55 | |
56 | @noindent | |
57 | This document describes a set of Emacs Lisp facilities borrowed from | |
58 | Common Lisp. All the facilities are described here in detail. While | |
59 | this document does not assume any prior knowledge of Common Lisp, it | |
60 | does assume a basic familiarity with Emacs Lisp. | |
61 | ||
62 | @menu | |
63 | * Overview:: Installation, usage, etc. | |
64 | * Program Structure:: Arglists, `eval-when', `defalias' | |
65 | * Predicates:: `typep', `eql', and `equalp' | |
66 | * Control Structure:: `setf', `do', `loop', etc. | |
67 | * Macros:: Destructuring, `define-compiler-macro' | |
68 | * Declarations:: `proclaim', `declare', etc. | |
69 | * Symbols:: Property lists, `gensym' | |
70 | * Numbers:: Predicates, functions, random numbers | |
71 | * Sequences:: Mapping, functions, searching, sorting | |
72 | * Lists:: `cadr', `sublis', `member*', `assoc*', etc. | |
6bf7aab6 DL |
73 | * Structures:: `defstruct' |
74 | * Assertions:: `check-type', `assert', `ignore-errors'. | |
75 | ||
76 | * Efficiency Concerns:: Hints and techniques | |
77 | * Common Lisp Compatibility:: All known differences with Steele | |
78 | * Old CL Compatibility:: All known differences with old cl.el | |
79 | * Porting Common Lisp:: Hints for porting Common Lisp code | |
80 | ||
84247bb5 | 81 | * GNU Free Documentation License:: The license for this documentation. |
6bf7aab6 DL |
82 | * Function Index:: |
83 | * Variable Index:: | |
84 | @end menu | |
85 | ||
86 | @node Overview, Program Structure, Top, Top | |
3105a2cc | 87 | @ifnottex |
6bf7aab6 | 88 | @chapter Overview |
3105a2cc | 89 | @end ifnottex |
6bf7aab6 DL |
90 | |
91 | @noindent | |
92 | Common Lisp is a huge language, and Common Lisp systems tend to be | |
93 | massive and extremely complex. Emacs Lisp, by contrast, is rather | |
94 | minimalist in the choice of Lisp features it offers the programmer. | |
95 | As Emacs Lisp programmers have grown in number, and the applications | |
96 | they write have grown more ambitious, it has become clear that Emacs | |
97 | Lisp could benefit from many of the conveniences of Common Lisp. | |
98 | ||
99 | The @dfn{CL} package adds a number of Common Lisp functions and | |
100 | control structures to Emacs Lisp. While not a 100% complete | |
101 | implementation of Common Lisp, @dfn{CL} adds enough functionality | |
102 | to make Emacs Lisp programming significantly more convenient. | |
103 | ||
a0e204da RS |
104 | @strong{Please note:} the @dfn{CL} functions are not standard parts of |
105 | the Emacs Lisp name space, so it is legitimate for users to define | |
106 | them with other, conflicting meanings. To avoid conflicting with | |
107 | those user activities, we have a policy that packages installed in | |
108 | Emacs must not load @dfn{CL} at run time. (It is ok for them to load | |
109 | @dfn{CL} at compile time only, with @code{eval-when-compile}, and use | |
110 | the macros it provides.) If you are writing packages that you plan to | |
111 | distribute and invite widespread use for, you might want to observe | |
112 | the same rule. | |
113 | ||
6bf7aab6 DL |
114 | Some Common Lisp features have been omitted from this package |
115 | for various reasons: | |
116 | ||
117 | @itemize @bullet | |
118 | @item | |
119 | Some features are too complex or bulky relative to their benefit | |
120 | to Emacs Lisp programmers. CLOS and Common Lisp streams are fine | |
121 | examples of this group. | |
122 | ||
123 | @item | |
124 | Other features cannot be implemented without modification to the | |
125 | Emacs Lisp interpreter itself, such as multiple return values, | |
126 | lexical scoping, case-insensitive symbols, and complex numbers. | |
127 | The @dfn{CL} package generally makes no attempt to emulate these | |
128 | features. | |
129 | ||
130 | @item | |
131 | Some features conflict with existing things in Emacs Lisp. For | |
132 | example, Emacs' @code{assoc} function is incompatible with the | |
133 | Common Lisp @code{assoc}. In such cases, this package usually | |
134 | adds the suffix @samp{*} to the function name of the Common | |
135 | Lisp version of the function (e.g., @code{assoc*}). | |
136 | @end itemize | |
137 | ||
138 | The package described here was written by Dave Gillespie, | |
139 | @file{daveg@@synaptics.com}. It is a total rewrite of the original | |
140 | 1986 @file{cl.el} package by Cesar Quiroz. Most features of the | |
19188ec7 | 141 | Quiroz package have been retained; any incompatibilities are |
6bf7aab6 DL |
142 | noted in the descriptions below. Care has been taken in this |
143 | version to ensure that each function is defined efficiently, | |
144 | concisely, and with minimal impact on the rest of the Emacs | |
145 | environment. | |
146 | ||
147 | @menu | |
148 | * Usage:: How to use the CL package | |
149 | * Organization:: The package's five component files | |
150 | * Installation:: Compiling and installing CL | |
151 | * Naming Conventions:: Notes on CL function names | |
152 | @end menu | |
153 | ||
154 | @node Usage, Organization, Overview, Overview | |
155 | @section Usage | |
156 | ||
157 | @noindent | |
158 | Lisp code that uses features from the @dfn{CL} package should | |
159 | include at the beginning: | |
160 | ||
161 | @example | |
162 | (require 'cl) | |
163 | @end example | |
164 | ||
165 | @noindent | |
166 | If you want to ensure that the new (Gillespie) version of @dfn{CL} | |
167 | is the one that is present, add an additional @code{(require 'cl-19)} | |
168 | call: | |
169 | ||
170 | @example | |
171 | (require 'cl) | |
172 | (require 'cl-19) | |
173 | @end example | |
174 | ||
175 | @noindent | |
176 | The second call will fail (with ``@file{cl-19.el} not found'') if | |
177 | the old @file{cl.el} package was in use. | |
178 | ||
179 | It is safe to arrange to load @dfn{CL} at all times, e.g., | |
180 | in your @file{.emacs} file. But it's a good idea, for portability, | |
181 | to @code{(require 'cl)} in your code even if you do this. | |
182 | ||
183 | @node Organization, Installation, Usage, Overview | |
184 | @section Organization | |
185 | ||
186 | @noindent | |
187 | The Common Lisp package is organized into four files: | |
188 | ||
189 | @table @file | |
190 | @item cl.el | |
191 | This is the ``main'' file, which contains basic functions | |
192 | and information about the package. This file is relatively | |
193 | compact---about 700 lines. | |
194 | ||
195 | @item cl-extra.el | |
196 | This file contains the larger, more complex or unusual functions. | |
197 | It is kept separate so that packages which only want to use Common | |
198 | Lisp fundamentals like the @code{cadr} function won't need to pay | |
199 | the overhead of loading the more advanced functions. | |
200 | ||
201 | @item cl-seq.el | |
202 | This file contains most of the advanced functions for operating | |
203 | on sequences or lists, such as @code{delete-if} and @code{assoc*}. | |
204 | ||
205 | @item cl-macs.el | |
206 | This file contains the features of the packages which are macros | |
207 | instead of functions. Macros expand when the caller is compiled, | |
208 | not when it is run, so the macros generally only need to be | |
209 | present when the byte-compiler is running (or when the macros are | |
210 | used in uncompiled code such as a @file{.emacs} file). Most of | |
211 | the macros of this package are isolated in @file{cl-macs.el} so | |
212 | that they won't take up memory unless you are compiling. | |
213 | @end table | |
214 | ||
215 | The file @file{cl.el} includes all necessary @code{autoload} | |
216 | commands for the functions and macros in the other three files. | |
217 | All you have to do is @code{(require 'cl)}, and @file{cl.el} | |
218 | will take care of pulling in the other files when they are | |
219 | needed. | |
220 | ||
221 | There is another file, @file{cl-compat.el}, which defines some | |
222 | routines from the older @file{cl.el} package that are no longer | |
223 | present in the new package. This includes internal routines | |
224 | like @code{setelt} and @code{zip-lists}, deprecated features | |
225 | like @code{defkeyword}, and an emulation of the old-style | |
226 | multiple-values feature. @xref{Old CL Compatibility}. | |
227 | ||
228 | @node Installation, Naming Conventions, Organization, Overview | |
229 | @section Installation | |
230 | ||
231 | @noindent | |
232 | Installation of the @dfn{CL} package is simple: Just put the | |
233 | byte-compiled files @file{cl.elc}, @file{cl-extra.elc}, | |
234 | @file{cl-seq.elc}, @file{cl-macs.elc}, and @file{cl-compat.elc} | |
235 | into a directory on your @code{load-path}. | |
236 | ||
237 | There are no special requirements to compile this package: | |
238 | The files do not have to be loaded before they are compiled, | |
239 | nor do they need to be compiled in any particular order. | |
240 | ||
241 | You may choose to put the files into your main @file{lisp/} | |
242 | directory, replacing the original @file{cl.el} file there. Or, | |
243 | you could put them into a directory that comes before @file{lisp/} | |
244 | on your @code{load-path} so that the old @file{cl.el} is | |
245 | effectively hidden. | |
246 | ||
247 | Also, format the @file{cl.texinfo} file and put the resulting | |
248 | Info files in the @file{info/} directory or another suitable place. | |
249 | ||
250 | You may instead wish to leave this package's components all in | |
251 | their own directory, and then add this directory to your | |
c6ad01e8 | 252 | @code{load-path} and @code{Info-directory-list}. |
6bf7aab6 DL |
253 | Add the directory to the front of the list so the old @dfn{CL} |
254 | package and its documentation are hidden. | |
255 | ||
c6ad01e8 | 256 | @node Naming Conventions, , Installation, Overview |
6bf7aab6 DL |
257 | @section Naming Conventions |
258 | ||
259 | @noindent | |
260 | Except where noted, all functions defined by this package have the | |
261 | same names and calling conventions as their Common Lisp counterparts. | |
262 | ||
263 | Following is a complete list of functions whose names were changed | |
264 | from Common Lisp, usually to avoid conflicts with Emacs. In each | |
265 | case, a @samp{*} has been appended to the Common Lisp name to obtain | |
266 | the Emacs name: | |
267 | ||
268 | @example | |
269 | defun* defsubst* defmacro* function* | |
270 | member* assoc* rassoc* get* | |
271 | remove* delete* mapcar* sort* | |
272 | floor* ceiling* truncate* round* | |
b7a084f6 | 273 | mod* rem* random* |
6bf7aab6 DL |
274 | @end example |
275 | ||
276 | Internal function and variable names in the package are prefixed | |
277 | by @code{cl-}. Here is a complete list of functions @emph{not} | |
278 | prefixed by @code{cl-} which were not taken from Common Lisp: | |
279 | ||
280 | @example | |
76162e12 | 281 | floatp-safe lexical-let lexical-let* |
6bf7aab6 | 282 | callf callf2 letf letf* |
c6ad01e8 | 283 | defsubst* |
6bf7aab6 DL |
284 | @end example |
285 | ||
6bf7aab6 DL |
286 | The following simple functions and macros are defined in @file{cl.el}; |
287 | they do not cause other components like @file{cl-extra} to be loaded. | |
288 | ||
289 | @example | |
ed4d9494 | 290 | eql floatp-safe endp |
6bf7aab6 | 291 | evenp oddp plusp minusp |
2d4c9509 | 292 | caaar .. cddddr |
6bf7aab6 | 293 | list* ldiff rest first .. tenth |
ed4d9494 | 294 | copy-list subst mapcar* [2] |
6bf7aab6 DL |
295 | adjoin [3] acons pairlis pop [4] |
296 | push [4] pushnew [3,4] incf [4] decf [4] | |
297 | proclaim declaim | |
298 | @end example | |
299 | ||
6bf7aab6 DL |
300 | @noindent |
301 | [2] Only for one sequence argument or two list arguments. | |
302 | ||
303 | @noindent | |
304 | [3] Only if @code{:test} is @code{eq}, @code{equal}, or unspecified, | |
305 | and @code{:key} is not used. | |
306 | ||
307 | @noindent | |
308 | [4] Only when @var{place} is a plain variable name. | |
309 | ||
310 | @iftex | |
311 | @chapno=4 | |
312 | @end iftex | |
313 | ||
314 | @node Program Structure, Predicates, Overview, Top | |
315 | @chapter Program Structure | |
316 | ||
317 | @noindent | |
318 | This section describes features of the @dfn{CL} package which have to | |
319 | do with programs as a whole: advanced argument lists for functions, | |
320 | and the @code{eval-when} construct. | |
321 | ||
322 | @menu | |
323 | * Argument Lists:: `&key', `&aux', `defun*', `defmacro*'. | |
324 | * Time of Evaluation:: The `eval-when' construct. | |
6bf7aab6 DL |
325 | @end menu |
326 | ||
327 | @iftex | |
328 | @secno=1 | |
329 | @end iftex | |
330 | ||
331 | @node Argument Lists, Time of Evaluation, Program Structure, Program Structure | |
332 | @section Argument Lists | |
333 | ||
334 | @noindent | |
335 | Emacs Lisp's notation for argument lists of functions is a subset of | |
336 | the Common Lisp notation. As well as the familiar @code{&optional} | |
337 | and @code{&rest} markers, Common Lisp allows you to specify default | |
338 | values for optional arguments, and it provides the additional markers | |
339 | @code{&key} and @code{&aux}. | |
340 | ||
341 | Since argument parsing is built-in to Emacs, there is no way for | |
342 | this package to implement Common Lisp argument lists seamlessly. | |
343 | Instead, this package defines alternates for several Lisp forms | |
344 | which you must use if you need Common Lisp argument lists. | |
345 | ||
346 | @defspec defun* name arglist body... | |
347 | This form is identical to the regular @code{defun} form, except | |
348 | that @var{arglist} is allowed to be a full Common Lisp argument | |
349 | list. Also, the function body is enclosed in an implicit block | |
350 | called @var{name}; @pxref{Blocks and Exits}. | |
351 | @end defspec | |
352 | ||
353 | @defspec defsubst* name arglist body... | |
354 | This is just like @code{defun*}, except that the function that | |
355 | is defined is automatically proclaimed @code{inline}, i.e., | |
356 | calls to it may be expanded into in-line code by the byte compiler. | |
c6ad01e8 | 357 | This is analogous to the @code{defsubst} form; |
6bf7aab6 DL |
358 | @code{defsubst*} uses a different method (compiler macros) which |
359 | works in all version of Emacs, and also generates somewhat more | |
360 | efficient inline expansions. In particular, @code{defsubst*} | |
361 | arranges for the processing of keyword arguments, default values, | |
362 | etc., to be done at compile-time whenever possible. | |
363 | @end defspec | |
364 | ||
365 | @defspec defmacro* name arglist body... | |
366 | This is identical to the regular @code{defmacro} form, | |
367 | except that @var{arglist} is allowed to be a full Common Lisp | |
368 | argument list. The @code{&environment} keyword is supported as | |
369 | described in Steele. The @code{&whole} keyword is supported only | |
370 | within destructured lists (see below); top-level @code{&whole} | |
371 | cannot be implemented with the current Emacs Lisp interpreter. | |
372 | The macro expander body is enclosed in an implicit block called | |
373 | @var{name}. | |
374 | @end defspec | |
375 | ||
376 | @defspec function* symbol-or-lambda | |
377 | This is identical to the regular @code{function} form, | |
378 | except that if the argument is a @code{lambda} form then that | |
379 | form may use a full Common Lisp argument list. | |
380 | @end defspec | |
381 | ||
382 | Also, all forms (such as @code{defsetf} and @code{flet}) defined | |
383 | in this package that include @var{arglist}s in their syntax allow | |
384 | full Common Lisp argument lists. | |
385 | ||
386 | Note that it is @emph{not} necessary to use @code{defun*} in | |
387 | order to have access to most @dfn{CL} features in your function. | |
388 | These features are always present; @code{defun*}'s only | |
389 | difference from @code{defun} is its more flexible argument | |
390 | lists and its implicit block. | |
391 | ||
392 | The full form of a Common Lisp argument list is | |
393 | ||
394 | @example | |
395 | (@var{var}... | |
396 | &optional (@var{var} @var{initform} @var{svar})... | |
397 | &rest @var{var} | |
398 | &key ((@var{keyword} @var{var}) @var{initform} @var{svar})... | |
399 | &aux (@var{var} @var{initform})...) | |
400 | @end example | |
401 | ||
402 | Each of the five argument list sections is optional. The @var{svar}, | |
403 | @var{initform}, and @var{keyword} parts are optional; if they are | |
404 | omitted, then @samp{(@var{var})} may be written simply @samp{@var{var}}. | |
405 | ||
406 | The first section consists of zero or more @dfn{required} arguments. | |
407 | These arguments must always be specified in a call to the function; | |
408 | there is no difference between Emacs Lisp and Common Lisp as far as | |
409 | required arguments are concerned. | |
410 | ||
411 | The second section consists of @dfn{optional} arguments. These | |
412 | arguments may be specified in the function call; if they are not, | |
413 | @var{initform} specifies the default value used for the argument. | |
414 | (No @var{initform} means to use @code{nil} as the default.) The | |
415 | @var{initform} is evaluated with the bindings for the preceding | |
416 | arguments already established; @code{(a &optional (b (1+ a)))} | |
417 | matches one or two arguments, with the second argument defaulting | |
418 | to one plus the first argument. If the @var{svar} is specified, | |
419 | it is an auxiliary variable which is bound to @code{t} if the optional | |
420 | argument was specified, or to @code{nil} if the argument was omitted. | |
421 | If you don't use an @var{svar}, then there will be no way for your | |
422 | function to tell whether it was called with no argument, or with | |
423 | the default value passed explicitly as an argument. | |
424 | ||
425 | The third section consists of a single @dfn{rest} argument. If | |
426 | more arguments were passed to the function than are accounted for | |
427 | by the required and optional arguments, those extra arguments are | |
428 | collected into a list and bound to the ``rest'' argument variable. | |
429 | Common Lisp's @code{&rest} is equivalent to that of Emacs Lisp. | |
430 | Common Lisp accepts @code{&body} as a synonym for @code{&rest} in | |
431 | macro contexts; this package accepts it all the time. | |
432 | ||
433 | The fourth section consists of @dfn{keyword} arguments. These | |
434 | are optional arguments which are specified by name rather than | |
435 | positionally in the argument list. For example, | |
436 | ||
437 | @example | |
438 | (defun* foo (a &optional b &key c d (e 17))) | |
439 | @end example | |
440 | ||
441 | @noindent | |
442 | defines a function which may be called with one, two, or more | |
443 | arguments. The first two arguments are bound to @code{a} and | |
444 | @code{b} in the usual way. The remaining arguments must be | |
445 | pairs of the form @code{:c}, @code{:d}, or @code{:e} followed | |
446 | by the value to be bound to the corresponding argument variable. | |
447 | (Symbols whose names begin with a colon are called @dfn{keywords}, | |
448 | and they are self-quoting in the same way as @code{nil} and | |
449 | @code{t}.) | |
450 | ||
451 | For example, the call @code{(foo 1 2 :d 3 :c 4)} sets the five | |
452 | arguments to 1, 2, 4, 3, and 17, respectively. If the same keyword | |
453 | appears more than once in the function call, the first occurrence | |
454 | takes precedence over the later ones. Note that it is not possible | |
455 | to specify keyword arguments without specifying the optional | |
456 | argument @code{b} as well, since @code{(foo 1 :c 2)} would bind | |
457 | @code{b} to the keyword @code{:c}, then signal an error because | |
458 | @code{2} is not a valid keyword. | |
459 | ||
460 | If a @var{keyword} symbol is explicitly specified in the argument | |
461 | list as shown in the above diagram, then that keyword will be | |
462 | used instead of just the variable name prefixed with a colon. | |
463 | You can specify a @var{keyword} symbol which does not begin with | |
464 | a colon at all, but such symbols will not be self-quoting; you | |
465 | will have to quote them explicitly with an apostrophe in the | |
466 | function call. | |
467 | ||
468 | Ordinarily it is an error to pass an unrecognized keyword to | |
469 | a function, e.g., @code{(foo 1 2 :c 3 :goober 4)}. You can ask | |
470 | Lisp to ignore unrecognized keywords, either by adding the | |
471 | marker @code{&allow-other-keys} after the keyword section | |
472 | of the argument list, or by specifying an @code{:allow-other-keys} | |
473 | argument in the call whose value is non-@code{nil}. If the | |
474 | function uses both @code{&rest} and @code{&key} at the same time, | |
475 | the ``rest'' argument is bound to the keyword list as it appears | |
476 | in the call. For example: | |
477 | ||
478 | @smallexample | |
479 | (defun* find-thing (thing &rest rest &key need &allow-other-keys) | |
480 | (or (apply 'member* thing thing-list :allow-other-keys t rest) | |
481 | (if need (error "Thing not found")))) | |
482 | @end smallexample | |
483 | ||
484 | @noindent | |
485 | This function takes a @code{:need} keyword argument, but also | |
486 | accepts other keyword arguments which are passed on to the | |
487 | @code{member*} function. @code{allow-other-keys} is used to | |
488 | keep both @code{find-thing} and @code{member*} from complaining | |
489 | about each others' keywords in the arguments. | |
490 | ||
6bf7aab6 DL |
491 | The fifth section of the argument list consists of @dfn{auxiliary |
492 | variables}. These are not really arguments at all, but simply | |
493 | variables which are bound to @code{nil} or to the specified | |
494 | @var{initforms} during execution of the function. There is no | |
495 | difference between the following two functions, except for a | |
496 | matter of stylistic taste: | |
497 | ||
498 | @example | |
499 | (defun* foo (a b &aux (c (+ a b)) d) | |
500 | @var{body}) | |
501 | ||
502 | (defun* foo (a b) | |
503 | (let ((c (+ a b)) d) | |
504 | @var{body})) | |
505 | @end example | |
506 | ||
507 | Argument lists support @dfn{destructuring}. In Common Lisp, | |
508 | destructuring is only allowed with @code{defmacro}; this package | |
509 | allows it with @code{defun*} and other argument lists as well. | |
510 | In destructuring, any argument variable (@var{var} in the above | |
511 | diagram) can be replaced by a list of variables, or more generally, | |
512 | a recursive argument list. The corresponding argument value must | |
513 | be a list whose elements match this recursive argument list. | |
514 | For example: | |
515 | ||
516 | @example | |
517 | (defmacro* dolist ((var listform &optional resultform) | |
518 | &rest body) | |
519 | ...) | |
520 | @end example | |
521 | ||
522 | This says that the first argument of @code{dolist} must be a list | |
523 | of two or three items; if there are other arguments as well as this | |
524 | list, they are stored in @code{body}. All features allowed in | |
525 | regular argument lists are allowed in these recursive argument lists. | |
526 | In addition, the clause @samp{&whole @var{var}} is allowed at the | |
527 | front of a recursive argument list. It binds @var{var} to the | |
528 | whole list being matched; thus @code{(&whole all a b)} matches | |
529 | a list of two things, with @code{a} bound to the first thing, | |
530 | @code{b} bound to the second thing, and @code{all} bound to the | |
531 | list itself. (Common Lisp allows @code{&whole} in top-level | |
532 | @code{defmacro} argument lists as well, but Emacs Lisp does not | |
533 | support this usage.) | |
534 | ||
535 | One last feature of destructuring is that the argument list may be | |
536 | dotted, so that the argument list @code{(a b . c)} is functionally | |
537 | equivalent to @code{(a b &rest c)}. | |
538 | ||
539 | If the optimization quality @code{safety} is set to 0 | |
540 | (@pxref{Declarations}), error checking for wrong number of | |
541 | arguments and invalid keyword arguments is disabled. By default, | |
542 | argument lists are rigorously checked. | |
543 | ||
c6ad01e8 | 544 | @node Time of Evaluation, , Argument Lists, Program Structure |
6bf7aab6 DL |
545 | @section Time of Evaluation |
546 | ||
547 | @noindent | |
548 | Normally, the byte-compiler does not actually execute the forms in | |
549 | a file it compiles. For example, if a file contains @code{(setq foo t)}, | |
550 | the act of compiling it will not actually set @code{foo} to @code{t}. | |
551 | This is true even if the @code{setq} was a top-level form (i.e., not | |
552 | enclosed in a @code{defun} or other form). Sometimes, though, you | |
553 | would like to have certain top-level forms evaluated at compile-time. | |
554 | For example, the compiler effectively evaluates @code{defmacro} forms | |
555 | at compile-time so that later parts of the file can refer to the | |
556 | macros that are defined. | |
557 | ||
558 | @defspec eval-when (situations...) forms... | |
559 | This form controls when the body @var{forms} are evaluated. | |
560 | The @var{situations} list may contain any set of the symbols | |
561 | @code{compile}, @code{load}, and @code{eval} (or their long-winded | |
562 | ANSI equivalents, @code{:compile-toplevel}, @code{:load-toplevel}, | |
563 | and @code{:execute}). | |
564 | ||
565 | The @code{eval-when} form is handled differently depending on | |
566 | whether or not it is being compiled as a top-level form. | |
567 | Specifically, it gets special treatment if it is being compiled | |
568 | by a command such as @code{byte-compile-file} which compiles files | |
569 | or buffers of code, and it appears either literally at the | |
570 | top level of the file or inside a top-level @code{progn}. | |
571 | ||
572 | For compiled top-level @code{eval-when}s, the body @var{forms} are | |
573 | executed at compile-time if @code{compile} is in the @var{situations} | |
574 | list, and the @var{forms} are written out to the file (to be executed | |
575 | at load-time) if @code{load} is in the @var{situations} list. | |
576 | ||
577 | For non-compiled-top-level forms, only the @code{eval} situation is | |
578 | relevant. (This includes forms executed by the interpreter, forms | |
579 | compiled with @code{byte-compile} rather than @code{byte-compile-file}, | |
580 | and non-top-level forms.) The @code{eval-when} acts like a | |
581 | @code{progn} if @code{eval} is specified, and like @code{nil} | |
582 | (ignoring the body @var{forms}) if not. | |
583 | ||
584 | The rules become more subtle when @code{eval-when}s are nested; | |
585 | consult Steele (second edition) for the gruesome details (and | |
586 | some gruesome examples). | |
587 | ||
588 | Some simple examples: | |
589 | ||
590 | @example | |
591 | ;; Top-level forms in foo.el: | |
592 | (eval-when (compile) (setq foo1 'bar)) | |
593 | (eval-when (load) (setq foo2 'bar)) | |
594 | (eval-when (compile load) (setq foo3 'bar)) | |
595 | (eval-when (eval) (setq foo4 'bar)) | |
596 | (eval-when (eval compile) (setq foo5 'bar)) | |
597 | (eval-when (eval load) (setq foo6 'bar)) | |
598 | (eval-when (eval compile load) (setq foo7 'bar)) | |
599 | @end example | |
600 | ||
601 | When @file{foo.el} is compiled, these variables will be set during | |
602 | the compilation itself: | |
603 | ||
604 | @example | |
605 | foo1 foo3 foo5 foo7 ; `compile' | |
606 | @end example | |
607 | ||
608 | When @file{foo.elc} is loaded, these variables will be set: | |
609 | ||
610 | @example | |
611 | foo2 foo3 foo6 foo7 ; `load' | |
612 | @end example | |
613 | ||
614 | And if @file{foo.el} is loaded uncompiled, these variables will | |
615 | be set: | |
616 | ||
617 | @example | |
618 | foo4 foo5 foo6 foo7 ; `eval' | |
619 | @end example | |
620 | ||
621 | If these seven @code{eval-when}s had been, say, inside a @code{defun}, | |
622 | then the first three would have been equivalent to @code{nil} and the | |
623 | last four would have been equivalent to the corresponding @code{setq}s. | |
624 | ||
625 | Note that @code{(eval-when (load eval) @dots{})} is equivalent | |
626 | to @code{(progn @dots{})} in all contexts. The compiler treats | |
627 | certain top-level forms, like @code{defmacro} (sort-of) and | |
628 | @code{require}, as if they were wrapped in @code{(eval-when | |
629 | (compile load eval) @dots{})}. | |
630 | @end defspec | |
631 | ||
c6ad01e8 | 632 | Emacs includes two special forms related to @code{eval-when}. |
6bf7aab6 | 633 | One of these, @code{eval-when-compile}, is not quite equivalent to |
ed4d9494 | 634 | any @code{eval-when} construct and is described below. |
6bf7aab6 DL |
635 | |
636 | The other form, @code{(eval-and-compile @dots{})}, is exactly | |
637 | equivalent to @samp{(eval-when (compile load eval) @dots{})} and | |
638 | so is not itself defined by this package. | |
639 | ||
640 | @defspec eval-when-compile forms... | |
641 | The @var{forms} are evaluated at compile-time; at execution time, | |
642 | this form acts like a quoted constant of the resulting value. Used | |
643 | at top-level, @code{eval-when-compile} is just like @samp{eval-when | |
644 | (compile eval)}. In other contexts, @code{eval-when-compile} | |
645 | allows code to be evaluated once at compile-time for efficiency | |
646 | or other reasons. | |
647 | ||
648 | This form is similar to the @samp{#.} syntax of true Common Lisp. | |
649 | @end defspec | |
650 | ||
651 | @defspec load-time-value form | |
652 | The @var{form} is evaluated at load-time; at execution time, | |
653 | this form acts like a quoted constant of the resulting value. | |
654 | ||
655 | Early Common Lisp had a @samp{#,} syntax that was similar to | |
656 | this, but ANSI Common Lisp replaced it with @code{load-time-value} | |
657 | and gave it more well-defined semantics. | |
658 | ||
659 | In a compiled file, @code{load-time-value} arranges for @var{form} | |
660 | to be evaluated when the @file{.elc} file is loaded and then used | |
661 | as if it were a quoted constant. In code compiled by | |
662 | @code{byte-compile} rather than @code{byte-compile-file}, the | |
663 | effect is identical to @code{eval-when-compile}. In uncompiled | |
664 | code, both @code{eval-when-compile} and @code{load-time-value} | |
665 | act exactly like @code{progn}. | |
666 | ||
667 | @example | |
668 | (defun report () | |
669 | (insert "This function was executed on: " | |
670 | (current-time-string) | |
671 | ", compiled on: " | |
672 | (eval-when-compile (current-time-string)) | |
673 | ;; or '#.(current-time-string) in real Common Lisp | |
674 | ", and loaded on: " | |
675 | (load-time-value (current-time-string)))) | |
676 | @end example | |
677 | ||
678 | @noindent | |
679 | Byte-compiled, the above defun will result in the following code | |
680 | (or its compiled equivalent, of course) in the @file{.elc} file: | |
681 | ||
682 | @example | |
683 | (setq --temp-- (current-time-string)) | |
684 | (defun report () | |
685 | (insert "This function was executed on: " | |
686 | (current-time-string) | |
687 | ", compiled on: " | |
688 | '"Wed Jun 23 18:33:43 1993" | |
689 | ", and loaded on: " | |
690 | --temp--)) | |
691 | @end example | |
692 | @end defspec | |
693 | ||
6bf7aab6 DL |
694 | @node Predicates, Control Structure, Program Structure, Top |
695 | @chapter Predicates | |
696 | ||
697 | @noindent | |
698 | This section describes functions for testing whether various | |
699 | facts are true or false. | |
700 | ||
701 | @menu | |
702 | * Type Predicates:: `typep', `deftype', and `coerce' | |
703 | * Equality Predicates:: `eql' and `equalp' | |
704 | @end menu | |
705 | ||
706 | @node Type Predicates, Equality Predicates, Predicates, Predicates | |
707 | @section Type Predicates | |
708 | ||
709 | @noindent | |
710 | The @dfn{CL} package defines a version of the Common Lisp @code{typep} | |
711 | predicate. | |
712 | ||
713 | @defun typep object type | |
714 | Check if @var{object} is of type @var{type}, where @var{type} is a | |
715 | (quoted) type name of the sort used by Common Lisp. For example, | |
716 | @code{(typep foo 'integer)} is equivalent to @code{(integerp foo)}. | |
717 | @end defun | |
718 | ||
719 | The @var{type} argument to the above function is either a symbol | |
720 | or a list beginning with a symbol. | |
721 | ||
722 | @itemize @bullet | |
723 | @item | |
724 | If the type name is a symbol, Emacs appends @samp{-p} to the | |
725 | symbol name to form the name of a predicate function for testing | |
726 | the type. (Built-in predicates whose names end in @samp{p} rather | |
727 | than @samp{-p} are used when appropriate.) | |
728 | ||
729 | @item | |
730 | The type symbol @code{t} stands for the union of all types. | |
731 | @code{(typep @var{object} t)} is always true. Likewise, the | |
732 | type symbol @code{nil} stands for nothing at all, and | |
733 | @code{(typep @var{object} nil)} is always false. | |
734 | ||
735 | @item | |
736 | The type symbol @code{null} represents the symbol @code{nil}. | |
737 | Thus @code{(typep @var{object} 'null)} is equivalent to | |
738 | @code{(null @var{object})}. | |
739 | ||
eb383a92 EZ |
740 | @item |
741 | The type symbol @code{atom} represents all objects that are not cons | |
742 | cells. Thus @code{(typep @var{object} 'atom)} is equivalent to | |
743 | @code{(atom @var{object})}. | |
744 | ||
6bf7aab6 DL |
745 | @item |
746 | The type symbol @code{real} is a synonym for @code{number}, and | |
747 | @code{fixnum} is a synonym for @code{integer}. | |
748 | ||
749 | @item | |
750 | The type symbols @code{character} and @code{string-char} match | |
751 | integers in the range from 0 to 255. | |
752 | ||
753 | @item | |
754 | The type symbol @code{float} uses the @code{floatp-safe} predicate | |
755 | defined by this package rather than @code{floatp}, so it will work | |
756 | correctly even in Emacs versions without floating-point support. | |
757 | ||
758 | @item | |
759 | The type list @code{(integer @var{low} @var{high})} represents all | |
760 | integers between @var{low} and @var{high}, inclusive. Either bound | |
761 | may be a list of a single integer to specify an exclusive limit, | |
762 | or a @code{*} to specify no limit. The type @code{(integer * *)} | |
763 | is thus equivalent to @code{integer}. | |
764 | ||
765 | @item | |
766 | Likewise, lists beginning with @code{float}, @code{real}, or | |
767 | @code{number} represent numbers of that type falling in a particular | |
768 | range. | |
769 | ||
770 | @item | |
771 | Lists beginning with @code{and}, @code{or}, and @code{not} form | |
772 | combinations of types. For example, @code{(or integer (float 0 *))} | |
773 | represents all objects that are integers or non-negative floats. | |
774 | ||
775 | @item | |
776 | Lists beginning with @code{member} or @code{member*} represent | |
777 | objects @code{eql} to any of the following values. For example, | |
778 | @code{(member 1 2 3 4)} is equivalent to @code{(integer 1 4)}, | |
779 | and @code{(member nil)} is equivalent to @code{null}. | |
780 | ||
781 | @item | |
782 | Lists of the form @code{(satisfies @var{predicate})} represent | |
783 | all objects for which @var{predicate} returns true when called | |
784 | with that object as an argument. | |
785 | @end itemize | |
786 | ||
787 | The following function and macro (not technically predicates) are | |
788 | related to @code{typep}. | |
789 | ||
790 | @defun coerce object type | |
791 | This function attempts to convert @var{object} to the specified | |
792 | @var{type}. If @var{object} is already of that type as determined by | |
793 | @code{typep}, it is simply returned. Otherwise, certain types of | |
794 | conversions will be made: If @var{type} is any sequence type | |
795 | (@code{string}, @code{list}, etc.) then @var{object} will be | |
796 | converted to that type if possible. If @var{type} is | |
797 | @code{character}, then strings of length one and symbols with | |
798 | one-character names can be coerced. If @var{type} is @code{float}, | |
799 | then integers can be coerced in versions of Emacs that support | |
800 | floats. In all other circumstances, @code{coerce} signals an | |
801 | error. | |
802 | @end defun | |
803 | ||
804 | @defspec deftype name arglist forms... | |
805 | This macro defines a new type called @var{name}. It is similar | |
806 | to @code{defmacro} in many ways; when @var{name} is encountered | |
807 | as a type name, the body @var{forms} are evaluated and should | |
808 | return a type specifier that is equivalent to the type. The | |
809 | @var{arglist} is a Common Lisp argument list of the sort accepted | |
810 | by @code{defmacro*}. The type specifier @samp{(@var{name} @var{args}...)} | |
811 | is expanded by calling the expander with those arguments; the type | |
812 | symbol @samp{@var{name}} is expanded by calling the expander with | |
813 | no arguments. The @var{arglist} is processed the same as for | |
814 | @code{defmacro*} except that optional arguments without explicit | |
815 | defaults use @code{*} instead of @code{nil} as the ``default'' | |
816 | default. Some examples: | |
817 | ||
818 | @example | |
819 | (deftype null () '(satisfies null)) ; predefined | |
820 | (deftype list () '(or null cons)) ; predefined | |
821 | (deftype unsigned-byte (&optional bits) | |
822 | (list 'integer 0 (if (eq bits '*) bits (1- (lsh 1 bits))))) | |
823 | (unsigned-byte 8) @equiv{} (integer 0 255) | |
824 | (unsigned-byte) @equiv{} (integer 0 *) | |
825 | unsigned-byte @equiv{} (integer 0 *) | |
826 | @end example | |
827 | ||
828 | @noindent | |
829 | The last example shows how the Common Lisp @code{unsigned-byte} | |
830 | type specifier could be implemented if desired; this package does | |
831 | not implement @code{unsigned-byte} by default. | |
832 | @end defspec | |
833 | ||
834 | The @code{typecase} and @code{check-type} macros also use type | |
835 | names. @xref{Conditionals}. @xref{Assertions}. The @code{map}, | |
836 | @code{concatenate}, and @code{merge} functions take type-name | |
837 | arguments to specify the type of sequence to return. @xref{Sequences}. | |
838 | ||
c6ad01e8 | 839 | @node Equality Predicates, , Type Predicates, Predicates |
6bf7aab6 DL |
840 | @section Equality Predicates |
841 | ||
842 | @noindent | |
843 | This package defines two Common Lisp predicates, @code{eql} and | |
844 | @code{equalp}. | |
845 | ||
846 | @defun eql a b | |
847 | This function is almost the same as @code{eq}, except that if @var{a} | |
848 | and @var{b} are numbers of the same type, it compares them for numeric | |
849 | equality (as if by @code{equal} instead of @code{eq}). This makes a | |
850 | difference only for versions of Emacs that are compiled with | |
c6ad01e8 | 851 | floating-point support. Emacs floats are allocated |
6bf7aab6 DL |
852 | objects just like cons cells, which means that @code{(eq 3.0 3.0)} |
853 | will not necessarily be true---if the two @code{3.0}s were allocated | |
854 | separately, the pointers will be different even though the numbers are | |
855 | the same. But @code{(eql 3.0 3.0)} will always be true. | |
856 | ||
857 | The types of the arguments must match, so @code{(eql 3 3.0)} is | |
858 | still false. | |
859 | ||
860 | Note that Emacs integers are ``direct'' rather than allocated, which | |
861 | basically means @code{(eq 3 3)} will always be true. Thus @code{eq} | |
862 | and @code{eql} behave differently only if floating-point numbers are | |
863 | involved, and are indistinguishable on Emacs versions that don't | |
864 | support floats. | |
865 | ||
866 | There is a slight inconsistency with Common Lisp in the treatment of | |
867 | positive and negative zeros. Some machines, notably those with IEEE | |
868 | standard arithmetic, represent @code{+0} and @code{-0} as distinct | |
869 | values. Normally this doesn't matter because the standard specifies | |
870 | that @code{(= 0.0 -0.0)} should always be true, and this is indeed | |
871 | what Emacs Lisp and Common Lisp do. But the Common Lisp standard | |
872 | states that @code{(eql 0.0 -0.0)} and @code{(equal 0.0 -0.0)} should | |
873 | be false on IEEE-like machines; Emacs Lisp does not do this, and in | |
874 | fact the only known way to distinguish between the two zeros in Emacs | |
875 | Lisp is to @code{format} them and check for a minus sign. | |
876 | @end defun | |
877 | ||
878 | @defun equalp a b | |
879 | This function is a more flexible version of @code{equal}. In | |
880 | particular, it compares strings case-insensitively, and it compares | |
881 | numbers without regard to type (so that @code{(equalp 3 3.0)} is | |
882 | true). Vectors and conses are compared recursively. All other | |
883 | objects are compared as if by @code{equal}. | |
884 | ||
885 | This function differs from Common Lisp @code{equalp} in several | |
886 | respects. First, Common Lisp's @code{equalp} also compares | |
887 | @emph{characters} case-insensitively, which would be impractical | |
888 | in this package since Emacs does not distinguish between integers | |
889 | and characters. In keeping with the idea that strings are less | |
890 | vector-like in Emacs Lisp, this package's @code{equalp} also will | |
c6ad01e8 | 891 | not compare strings against vectors of integers. |
6bf7aab6 DL |
892 | @end defun |
893 | ||
894 | Also note that the Common Lisp functions @code{member} and @code{assoc} | |
895 | use @code{eql} to compare elements, whereas Emacs Lisp follows the | |
896 | MacLisp tradition and uses @code{equal} for these two functions. | |
897 | In Emacs, use @code{member*} and @code{assoc*} to get functions | |
898 | which use @code{eql} for comparisons. | |
899 | ||
900 | @node Control Structure, Macros, Predicates, Top | |
901 | @chapter Control Structure | |
902 | ||
903 | @noindent | |
904 | The features described in the following sections implement | |
905 | various advanced control structures, including the powerful | |
906 | @code{setf} facility and a number of looping and conditional | |
907 | constructs. | |
908 | ||
909 | @menu | |
910 | * Assignment:: The `psetq' form | |
911 | * Generalized Variables:: `setf', `incf', `push', etc. | |
912 | * Variable Bindings:: `progv', `lexical-let', `flet', `macrolet' | |
913 | * Conditionals:: `case', `typecase' | |
914 | * Blocks and Exits:: `block', `return', `return-from' | |
2995b042 | 915 | * Iteration:: `do', `dotimes', `dolist', `do-symbols' |
6bf7aab6 DL |
916 | * Loop Facility:: The Common Lisp `loop' macro |
917 | * Multiple Values:: `values', `multiple-value-bind', etc. | |
918 | @end menu | |
919 | ||
920 | @node Assignment, Generalized Variables, Control Structure, Control Structure | |
921 | @section Assignment | |
922 | ||
923 | @noindent | |
924 | The @code{psetq} form is just like @code{setq}, except that multiple | |
925 | assignments are done in parallel rather than sequentially. | |
926 | ||
927 | @defspec psetq [symbol form]@dots{} | |
928 | This special form (actually a macro) is used to assign to several | |
929 | variables simultaneously. Given only one @var{symbol} and @var{form}, | |
930 | it has the same effect as @code{setq}. Given several @var{symbol} | |
931 | and @var{form} pairs, it evaluates all the @var{form}s in advance | |
932 | and then stores the corresponding variables afterwards. | |
933 | ||
934 | @example | |
935 | (setq x 2 y 3) | |
936 | (setq x (+ x y) y (* x y)) | |
937 | x | |
938 | @result{} 5 | |
939 | y ; @r{@code{y} was computed after @code{x} was set.} | |
940 | @result{} 15 | |
941 | (setq x 2 y 3) | |
942 | (psetq x (+ x y) y (* x y)) | |
943 | x | |
944 | @result{} 5 | |
945 | y ; @r{@code{y} was computed before @code{x} was set.} | |
946 | @result{} 6 | |
947 | @end example | |
948 | ||
949 | The simplest use of @code{psetq} is @code{(psetq x y y x)}, which | |
950 | exchanges the values of two variables. (The @code{rotatef} form | |
951 | provides an even more convenient way to swap two variables; | |
952 | @pxref{Modify Macros}.) | |
953 | ||
954 | @code{psetq} always returns @code{nil}. | |
955 | @end defspec | |
956 | ||
957 | @node Generalized Variables, Variable Bindings, Assignment, Control Structure | |
958 | @section Generalized Variables | |
959 | ||
960 | @noindent | |
961 | A ``generalized variable'' or ``place form'' is one of the many places | |
962 | in Lisp memory where values can be stored. The simplest place form is | |
963 | a regular Lisp variable. But the cars and cdrs of lists, elements | |
964 | of arrays, properties of symbols, and many other locations are also | |
965 | places where Lisp values are stored. | |
966 | ||
967 | The @code{setf} form is like @code{setq}, except that it accepts | |
968 | arbitrary place forms on the left side rather than just | |
969 | symbols. For example, @code{(setf (car a) b)} sets the car of | |
970 | @code{a} to @code{b}, doing the same operation as @code{(setcar a b)} | |
971 | but without having to remember two separate functions for setting | |
972 | and accessing every type of place. | |
973 | ||
974 | Generalized variables are analogous to ``lvalues'' in the C | |
975 | language, where @samp{x = a[i]} gets an element from an array | |
976 | and @samp{a[i] = x} stores an element using the same notation. | |
977 | Just as certain forms like @code{a[i]} can be lvalues in C, there | |
978 | is a set of forms that can be generalized variables in Lisp. | |
979 | ||
980 | @menu | |
981 | * Basic Setf:: `setf' and place forms | |
982 | * Modify Macros:: `incf', `push', `rotatef', `letf', `callf', etc. | |
983 | * Customizing Setf:: `define-modify-macro', `defsetf', `define-setf-method' | |
984 | @end menu | |
985 | ||
986 | @node Basic Setf, Modify Macros, Generalized Variables, Generalized Variables | |
987 | @subsection Basic Setf | |
988 | ||
989 | @noindent | |
990 | The @code{setf} macro is the most basic way to operate on generalized | |
991 | variables. | |
992 | ||
993 | @defspec setf [place form]@dots{} | |
994 | This macro evaluates @var{form} and stores it in @var{place}, which | |
995 | must be a valid generalized variable form. If there are several | |
996 | @var{place} and @var{form} pairs, the assignments are done sequentially | |
997 | just as with @code{setq}. @code{setf} returns the value of the last | |
998 | @var{form}. | |
999 | ||
1000 | The following Lisp forms will work as generalized variables, and | |
2cbd16b9 | 1001 | so may appear in the @var{place} argument of @code{setf}: |
6bf7aab6 DL |
1002 | |
1003 | @itemize @bullet | |
1004 | @item | |
1005 | A symbol naming a variable. In other words, @code{(setf x y)} is | |
1006 | exactly equivalent to @code{(setq x y)}, and @code{setq} itself is | |
1007 | strictly speaking redundant now that @code{setf} exists. Many | |
1008 | programmers continue to prefer @code{setq} for setting simple | |
1009 | variables, though, purely for stylistic or historical reasons. | |
1010 | The macro @code{(setf x y)} actually expands to @code{(setq x y)}, | |
1011 | so there is no performance penalty for using it in compiled code. | |
1012 | ||
1013 | @item | |
1014 | A call to any of the following Lisp functions: | |
1015 | ||
1016 | @smallexample | |
1017 | car cdr caar .. cddddr | |
1018 | nth rest first .. tenth | |
1019 | aref elt nthcdr | |
1020 | symbol-function symbol-value symbol-plist | |
1021 | get get* getf | |
1022 | gethash subseq | |
1023 | @end smallexample | |
1024 | ||
1025 | @noindent | |
1026 | Note that for @code{nthcdr} and @code{getf}, the list argument | |
1027 | of the function must itself be a valid @var{place} form. For | |
1028 | example, @code{(setf (nthcdr 0 foo) 7)} will set @code{foo} itself | |
1029 | to 7. Note that @code{push} and @code{pop} on an @code{nthcdr} | |
1030 | place can be used to insert or delete at any position in a list. | |
1031 | The use of @code{nthcdr} as a @var{place} form is an extension | |
1032 | to standard Common Lisp. | |
1033 | ||
1034 | @item | |
1035 | The following Emacs-specific functions are also @code{setf}-able. | |
6bf7aab6 DL |
1036 | |
1037 | @smallexample | |
177c0ea7 JB |
1038 | buffer-file-name marker-position |
1039 | buffer-modified-p match-data | |
1040 | buffer-name mouse-position | |
1041 | buffer-string overlay-end | |
1042 | buffer-substring overlay-get | |
1043 | current-buffer overlay-start | |
1044 | current-case-table point | |
1045 | current-column point-marker | |
1046 | current-global-map point-max | |
1047 | current-input-mode point-min | |
1048 | current-local-map process-buffer | |
1049 | current-window-configuration process-filter | |
1050 | default-file-modes process-sentinel | |
1051 | default-value read-mouse-position | |
1052 | documentation-property screen-height | |
1053 | extent-data screen-menubar | |
1054 | extent-end-position screen-width | |
1055 | extent-start-position selected-window | |
1056 | face-background selected-screen | |
1057 | face-background-pixmap selected-frame | |
1058 | face-font standard-case-table | |
1059 | face-foreground syntax-table | |
1060 | face-underline-p window-buffer | |
1061 | file-modes window-dedicated-p | |
1062 | frame-height window-display-table | |
1063 | frame-parameters window-height | |
1064 | frame-visible-p window-hscroll | |
1065 | frame-width window-point | |
1066 | get-register window-start | |
1067 | getenv window-width | |
1068 | global-key-binding x-get-cut-buffer | |
1069 | keymap-parent x-get-cutbuffer | |
6bf7aab6 | 1070 | local-key-binding x-get-secondary-selection |
177c0ea7 JB |
1071 | mark x-get-selection |
1072 | mark-marker | |
6bf7aab6 DL |
1073 | @end smallexample |
1074 | ||
1075 | Most of these have directly corresponding ``set'' functions, like | |
1076 | @code{use-local-map} for @code{current-local-map}, or @code{goto-char} | |
1077 | for @code{point}. A few, like @code{point-min}, expand to longer | |
1078 | sequences of code when they are @code{setf}'d (@code{(narrow-to-region | |
1079 | x (point-max))} in this case). | |
1080 | ||
1081 | @item | |
1082 | A call of the form @code{(substring @var{subplace} @var{n} [@var{m}])}, | |
2cbd16b9 | 1083 | where @var{subplace} is itself a valid generalized variable whose |
6bf7aab6 DL |
1084 | current value is a string, and where the value stored is also a |
1085 | string. The new string is spliced into the specified part of the | |
1086 | destination string. For example: | |
1087 | ||
1088 | @example | |
1089 | (setq a (list "hello" "world")) | |
1090 | @result{} ("hello" "world") | |
1091 | (cadr a) | |
1092 | @result{} "world" | |
1093 | (substring (cadr a) 2 4) | |
1094 | @result{} "rl" | |
1095 | (setf (substring (cadr a) 2 4) "o") | |
1096 | @result{} "o" | |
1097 | (cadr a) | |
1098 | @result{} "wood" | |
1099 | a | |
1100 | @result{} ("hello" "wood") | |
1101 | @end example | |
1102 | ||
1103 | The generalized variable @code{buffer-substring}, listed above, | |
1104 | also works in this way by replacing a portion of the current buffer. | |
1105 | ||
1106 | @item | |
1107 | A call of the form @code{(apply '@var{func} @dots{})} or | |
1108 | @code{(apply (function @var{func}) @dots{})}, where @var{func} | |
1109 | is a @code{setf}-able function whose store function is ``suitable'' | |
1110 | in the sense described in Steele's book; since none of the standard | |
1111 | Emacs place functions are suitable in this sense, this feature is | |
1112 | only interesting when used with places you define yourself with | |
1113 | @code{define-setf-method} or the long form of @code{defsetf}. | |
1114 | ||
1115 | @item | |
1116 | A macro call, in which case the macro is expanded and @code{setf} | |
1117 | is applied to the resulting form. | |
1118 | ||
1119 | @item | |
1120 | Any form for which a @code{defsetf} or @code{define-setf-method} | |
1121 | has been made. | |
1122 | @end itemize | |
1123 | ||
1124 | Using any forms other than these in the @var{place} argument to | |
1125 | @code{setf} will signal an error. | |
1126 | ||
1127 | The @code{setf} macro takes care to evaluate all subforms in | |
1128 | the proper left-to-right order; for example, | |
1129 | ||
1130 | @example | |
1131 | (setf (aref vec (incf i)) i) | |
1132 | @end example | |
1133 | ||
1134 | @noindent | |
1135 | looks like it will evaluate @code{(incf i)} exactly once, before the | |
1136 | following access to @code{i}; the @code{setf} expander will insert | |
1137 | temporary variables as necessary to ensure that it does in fact work | |
1138 | this way no matter what setf-method is defined for @code{aref}. | |
1139 | (In this case, @code{aset} would be used and no such steps would | |
1140 | be necessary since @code{aset} takes its arguments in a convenient | |
1141 | order.) | |
1142 | ||
1143 | However, if the @var{place} form is a macro which explicitly | |
1144 | evaluates its arguments in an unusual order, this unusual order | |
1145 | will be preserved. Adapting an example from Steele, given | |
1146 | ||
1147 | @example | |
1148 | (defmacro wrong-order (x y) (list 'aref y x)) | |
1149 | @end example | |
1150 | ||
1151 | @noindent | |
1152 | the form @code{(setf (wrong-order @var{a} @var{b}) 17)} will | |
1153 | evaluate @var{b} first, then @var{a}, just as in an actual call | |
1154 | to @code{wrong-order}. | |
1155 | @end defspec | |
1156 | ||
1157 | @node Modify Macros, Customizing Setf, Basic Setf, Generalized Variables | |
1158 | @subsection Modify Macros | |
1159 | ||
1160 | @noindent | |
1161 | This package defines a number of other macros besides @code{setf} | |
1162 | that operate on generalized variables. Many are interesting and | |
1163 | useful even when the @var{place} is just a variable name. | |
1164 | ||
1165 | @defspec psetf [place form]@dots{} | |
1166 | This macro is to @code{setf} what @code{psetq} is to @code{setq}: | |
1167 | When several @var{place}s and @var{form}s are involved, the | |
1168 | assignments take place in parallel rather than sequentially. | |
1169 | Specifically, all subforms are evaluated from left to right, then | |
1170 | all the assignments are done (in an undefined order). | |
1171 | @end defspec | |
1172 | ||
1173 | @defspec incf place &optional x | |
1174 | This macro increments the number stored in @var{place} by one, or | |
1175 | by @var{x} if specified. The incremented value is returned. For | |
1176 | example, @code{(incf i)} is equivalent to @code{(setq i (1+ i))}, and | |
1177 | @code{(incf (car x) 2)} is equivalent to @code{(setcar x (+ (car x) 2))}. | |
1178 | ||
1179 | Once again, care is taken to preserve the ``apparent'' order of | |
1180 | evaluation. For example, | |
1181 | ||
1182 | @example | |
1183 | (incf (aref vec (incf i))) | |
1184 | @end example | |
1185 | ||
1186 | @noindent | |
1187 | appears to increment @code{i} once, then increment the element of | |
1188 | @code{vec} addressed by @code{i}; this is indeed exactly what it | |
1189 | does, which means the above form is @emph{not} equivalent to the | |
1190 | ``obvious'' expansion, | |
1191 | ||
1192 | @example | |
1193 | (setf (aref vec (incf i)) (1+ (aref vec (incf i)))) ; Wrong! | |
1194 | @end example | |
1195 | ||
1196 | @noindent | |
1197 | but rather to something more like | |
1198 | ||
1199 | @example | |
1200 | (let ((temp (incf i))) | |
1201 | (setf (aref vec temp) (1+ (aref vec temp)))) | |
1202 | @end example | |
1203 | ||
1204 | @noindent | |
1205 | Again, all of this is taken care of automatically by @code{incf} and | |
1206 | the other generalized-variable macros. | |
1207 | ||
1208 | As a more Emacs-specific example of @code{incf}, the expression | |
1209 | @code{(incf (point) @var{n})} is essentially equivalent to | |
1210 | @code{(forward-char @var{n})}. | |
1211 | @end defspec | |
1212 | ||
1213 | @defspec decf place &optional x | |
1214 | This macro decrements the number stored in @var{place} by one, or | |
1215 | by @var{x} if specified. | |
1216 | @end defspec | |
1217 | ||
1218 | @defspec pop place | |
1219 | This macro removes and returns the first element of the list stored | |
1220 | in @var{place}. It is analogous to @code{(prog1 (car @var{place}) | |
1221 | (setf @var{place} (cdr @var{place})))}, except that it takes care | |
1222 | to evaluate all subforms only once. | |
1223 | @end defspec | |
1224 | ||
1225 | @defspec push x place | |
1226 | This macro inserts @var{x} at the front of the list stored in | |
1227 | @var{place}. It is analogous to @code{(setf @var{place} (cons | |
1228 | @var{x} @var{place}))}, except for evaluation of the subforms. | |
1229 | @end defspec | |
1230 | ||
1231 | @defspec pushnew x place @t{&key :test :test-not :key} | |
1232 | This macro inserts @var{x} at the front of the list stored in | |
1233 | @var{place}, but only if @var{x} was not @code{eql} to any | |
1234 | existing element of the list. The optional keyword arguments | |
1235 | are interpreted in the same way as for @code{adjoin}. | |
1236 | @xref{Lists as Sets}. | |
1237 | @end defspec | |
1238 | ||
1239 | @defspec shiftf place@dots{} newvalue | |
1240 | This macro shifts the @var{place}s left by one, shifting in the | |
1241 | value of @var{newvalue} (which may be any Lisp expression, not just | |
1242 | a generalized variable), and returning the value shifted out of | |
1243 | the first @var{place}. Thus, @code{(shiftf @var{a} @var{b} @var{c} | |
1244 | @var{d})} is equivalent to | |
1245 | ||
1246 | @example | |
1247 | (prog1 | |
1248 | @var{a} | |
1249 | (psetf @var{a} @var{b} | |
1250 | @var{b} @var{c} | |
1251 | @var{c} @var{d})) | |
1252 | @end example | |
1253 | ||
1254 | @noindent | |
1255 | except that the subforms of @var{a}, @var{b}, and @var{c} are actually | |
1256 | evaluated only once each and in the apparent order. | |
1257 | @end defspec | |
1258 | ||
1259 | @defspec rotatef place@dots{} | |
1260 | This macro rotates the @var{place}s left by one in circular fashion. | |
1261 | Thus, @code{(rotatef @var{a} @var{b} @var{c} @var{d})} is equivalent to | |
1262 | ||
1263 | @example | |
1264 | (psetf @var{a} @var{b} | |
1265 | @var{b} @var{c} | |
1266 | @var{c} @var{d} | |
1267 | @var{d} @var{a}) | |
1268 | @end example | |
1269 | ||
1270 | @noindent | |
1271 | except for the evaluation of subforms. @code{rotatef} always | |
1272 | returns @code{nil}. Note that @code{(rotatef @var{a} @var{b})} | |
1273 | conveniently exchanges @var{a} and @var{b}. | |
1274 | @end defspec | |
1275 | ||
1276 | The following macros were invented for this package; they have no | |
1277 | analogues in Common Lisp. | |
1278 | ||
1279 | @defspec letf (bindings@dots{}) forms@dots{} | |
1280 | This macro is analogous to @code{let}, but for generalized variables | |
1281 | rather than just symbols. Each @var{binding} should be of the form | |
1282 | @code{(@var{place} @var{value})}; the original contents of the | |
1283 | @var{place}s are saved, the @var{value}s are stored in them, and | |
1284 | then the body @var{form}s are executed. Afterwards, the @var{places} | |
1285 | are set back to their original saved contents. This cleanup happens | |
1286 | even if the @var{form}s exit irregularly due to a @code{throw} or an | |
1287 | error. | |
1288 | ||
1289 | For example, | |
1290 | ||
1291 | @example | |
1292 | (letf (((point) (point-min)) | |
1293 | (a 17)) | |
1294 | ...) | |
1295 | @end example | |
1296 | ||
1297 | @noindent | |
1298 | moves ``point'' in the current buffer to the beginning of the buffer, | |
1299 | and also binds @code{a} to 17 (as if by a normal @code{let}, since | |
1300 | @code{a} is just a regular variable). After the body exits, @code{a} | |
1301 | is set back to its original value and point is moved back to its | |
1302 | original position. | |
1303 | ||
1304 | Note that @code{letf} on @code{(point)} is not quite like a | |
1305 | @code{save-excursion}, as the latter effectively saves a marker | |
1306 | which tracks insertions and deletions in the buffer. Actually, | |
1307 | a @code{letf} of @code{(point-marker)} is much closer to this | |
1308 | behavior. (@code{point} and @code{point-marker} are equivalent | |
1309 | as @code{setf} places; each will accept either an integer or a | |
1310 | marker as the stored value.) | |
1311 | ||
1312 | Since generalized variables look like lists, @code{let}'s shorthand | |
1313 | of using @samp{foo} for @samp{(foo nil)} as a @var{binding} would | |
1314 | be ambiguous in @code{letf} and is not allowed. | |
1315 | ||
1316 | However, a @var{binding} specifier may be a one-element list | |
1317 | @samp{(@var{place})}, which is similar to @samp{(@var{place} | |
1318 | @var{place})}. In other words, the @var{place} is not disturbed | |
1319 | on entry to the body, and the only effect of the @code{letf} is | |
1320 | to restore the original value of @var{place} afterwards. (The | |
1321 | redundant access-and-store suggested by the @code{(@var{place} | |
1322 | @var{place})} example does not actually occur.) | |
1323 | ||
1324 | In most cases, the @var{place} must have a well-defined value on | |
1325 | entry to the @code{letf} form. The only exceptions are plain | |
1326 | variables and calls to @code{symbol-value} and @code{symbol-function}. | |
1327 | If the symbol is not bound on entry, it is simply made unbound by | |
1328 | @code{makunbound} or @code{fmakunbound} on exit. | |
1329 | @end defspec | |
1330 | ||
1331 | @defspec letf* (bindings@dots{}) forms@dots{} | |
1332 | This macro is to @code{letf} what @code{let*} is to @code{let}: | |
1333 | It does the bindings in sequential rather than parallel order. | |
1334 | @end defspec | |
1335 | ||
1336 | @defspec callf @var{function} @var{place} @var{args}@dots{} | |
1337 | This is the ``generic'' modify macro. It calls @var{function}, | |
1338 | which should be an unquoted function name, macro name, or lambda. | |
1339 | It passes @var{place} and @var{args} as arguments, and assigns the | |
1340 | result back to @var{place}. For example, @code{(incf @var{place} | |
1341 | @var{n})} is the same as @code{(callf + @var{place} @var{n})}. | |
1342 | Some more examples: | |
1343 | ||
1344 | @example | |
1345 | (callf abs my-number) | |
1346 | (callf concat (buffer-name) "<" (int-to-string n) ">") | |
1347 | (callf union happy-people (list joe bob) :test 'same-person) | |
1348 | @end example | |
1349 | ||
1350 | @xref{Customizing Setf}, for @code{define-modify-macro}, a way | |
1351 | to create even more concise notations for modify macros. Note | |
1352 | again that @code{callf} is an extension to standard Common Lisp. | |
1353 | @end defspec | |
1354 | ||
1355 | @defspec callf2 @var{function} @var{arg1} @var{place} @var{args}@dots{} | |
1356 | This macro is like @code{callf}, except that @var{place} is | |
1357 | the @emph{second} argument of @var{function} rather than the | |
1358 | first. For example, @code{(push @var{x} @var{place})} is | |
1359 | equivalent to @code{(callf2 cons @var{x} @var{place})}. | |
1360 | @end defspec | |
1361 | ||
1362 | The @code{callf} and @code{callf2} macros serve as building | |
1363 | blocks for other macros like @code{incf}, @code{pushnew}, and | |
1364 | @code{define-modify-macro}. The @code{letf} and @code{letf*} | |
1365 | macros are used in the processing of symbol macros; | |
1366 | @pxref{Macro Bindings}. | |
1367 | ||
c6ad01e8 | 1368 | @node Customizing Setf, , Modify Macros, Generalized Variables |
6bf7aab6 DL |
1369 | @subsection Customizing Setf |
1370 | ||
1371 | @noindent | |
1372 | Common Lisp defines three macros, @code{define-modify-macro}, | |
1373 | @code{defsetf}, and @code{define-setf-method}, that allow the | |
1374 | user to extend generalized variables in various ways. | |
1375 | ||
1376 | @defspec define-modify-macro name arglist function [doc-string] | |
1377 | This macro defines a ``read-modify-write'' macro similar to | |
1378 | @code{incf} and @code{decf}. The macro @var{name} is defined | |
1379 | to take a @var{place} argument followed by additional arguments | |
1380 | described by @var{arglist}. The call | |
1381 | ||
1382 | @example | |
1383 | (@var{name} @var{place} @var{args}...) | |
1384 | @end example | |
1385 | ||
1386 | @noindent | |
1387 | will be expanded to | |
1388 | ||
1389 | @example | |
1390 | (callf @var{func} @var{place} @var{args}...) | |
1391 | @end example | |
1392 | ||
1393 | @noindent | |
1394 | which in turn is roughly equivalent to | |
1395 | ||
1396 | @example | |
1397 | (setf @var{place} (@var{func} @var{place} @var{args}...)) | |
1398 | @end example | |
1399 | ||
1400 | For example: | |
1401 | ||
1402 | @example | |
1403 | (define-modify-macro incf (&optional (n 1)) +) | |
1404 | (define-modify-macro concatf (&rest args) concat) | |
1405 | @end example | |
1406 | ||
1407 | Note that @code{&key} is not allowed in @var{arglist}, but | |
1408 | @code{&rest} is sufficient to pass keywords on to the function. | |
1409 | ||
1410 | Most of the modify macros defined by Common Lisp do not exactly | |
1411 | follow the pattern of @code{define-modify-macro}. For example, | |
1412 | @code{push} takes its arguments in the wrong order, and @code{pop} | |
1413 | is completely irregular. You can define these macros ``by hand'' | |
1414 | using @code{get-setf-method}, or consult the source file | |
1415 | @file{cl-macs.el} to see how to use the internal @code{setf} | |
1416 | building blocks. | |
1417 | @end defspec | |
1418 | ||
1419 | @defspec defsetf access-fn update-fn | |
1420 | This is the simpler of two @code{defsetf} forms. Where | |
1421 | @var{access-fn} is the name of a function which accesses a place, | |
1422 | this declares @var{update-fn} to be the corresponding store | |
1423 | function. From now on, | |
1424 | ||
1425 | @example | |
1426 | (setf (@var{access-fn} @var{arg1} @var{arg2} @var{arg3}) @var{value}) | |
1427 | @end example | |
1428 | ||
1429 | @noindent | |
1430 | will be expanded to | |
1431 | ||
1432 | @example | |
1433 | (@var{update-fn} @var{arg1} @var{arg2} @var{arg3} @var{value}) | |
1434 | @end example | |
1435 | ||
1436 | @noindent | |
1437 | The @var{update-fn} is required to be either a true function, or | |
1438 | a macro which evaluates its arguments in a function-like way. Also, | |
1439 | the @var{update-fn} is expected to return @var{value} as its result. | |
1440 | Otherwise, the above expansion would not obey the rules for the way | |
1441 | @code{setf} is supposed to behave. | |
1442 | ||
1443 | As a special (non-Common-Lisp) extension, a third argument of @code{t} | |
1444 | to @code{defsetf} says that the @code{update-fn}'s return value is | |
1445 | not suitable, so that the above @code{setf} should be expanded to | |
1446 | something more like | |
1447 | ||
1448 | @example | |
1449 | (let ((temp @var{value})) | |
1450 | (@var{update-fn} @var{arg1} @var{arg2} @var{arg3} temp) | |
1451 | temp) | |
1452 | @end example | |
1453 | ||
1454 | Some examples of the use of @code{defsetf}, drawn from the standard | |
1455 | suite of setf methods, are: | |
1456 | ||
1457 | @example | |
1458 | (defsetf car setcar) | |
1459 | (defsetf symbol-value set) | |
1460 | (defsetf buffer-name rename-buffer t) | |
1461 | @end example | |
1462 | @end defspec | |
1463 | ||
1464 | @defspec defsetf access-fn arglist (store-var) forms@dots{} | |
1465 | This is the second, more complex, form of @code{defsetf}. It is | |
1466 | rather like @code{defmacro} except for the additional @var{store-var} | |
1467 | argument. The @var{forms} should return a Lisp form which stores | |
1468 | the value of @var{store-var} into the generalized variable formed | |
1469 | by a call to @var{access-fn} with arguments described by @var{arglist}. | |
1470 | The @var{forms} may begin with a string which documents the @code{setf} | |
1471 | method (analogous to the doc string that appears at the front of a | |
1472 | function). | |
1473 | ||
1474 | For example, the simple form of @code{defsetf} is shorthand for | |
1475 | ||
1476 | @example | |
1477 | (defsetf @var{access-fn} (&rest args) (store) | |
1478 | (append '(@var{update-fn}) args (list store))) | |
1479 | @end example | |
1480 | ||
1481 | The Lisp form that is returned can access the arguments from | |
1482 | @var{arglist} and @var{store-var} in an unrestricted fashion; | |
1483 | macros like @code{setf} and @code{incf} which invoke this | |
1484 | setf-method will insert temporary variables as needed to make | |
1485 | sure the apparent order of evaluation is preserved. | |
1486 | ||
1487 | Another example drawn from the standard package: | |
1488 | ||
1489 | @example | |
1490 | (defsetf nth (n x) (store) | |
1491 | (list 'setcar (list 'nthcdr n x) store)) | |
1492 | @end example | |
1493 | @end defspec | |
1494 | ||
1495 | @defspec define-setf-method access-fn arglist forms@dots{} | |
1496 | This is the most general way to create new place forms. When | |
1497 | a @code{setf} to @var{access-fn} with arguments described by | |
1498 | @var{arglist} is expanded, the @var{forms} are evaluated and | |
1499 | must return a list of five items: | |
1500 | ||
1501 | @enumerate | |
1502 | @item | |
1503 | A list of @dfn{temporary variables}. | |
1504 | ||
1505 | @item | |
1506 | A list of @dfn{value forms} corresponding to the temporary variables | |
1507 | above. The temporary variables will be bound to these value forms | |
1508 | as the first step of any operation on the generalized variable. | |
1509 | ||
1510 | @item | |
1511 | A list of exactly one @dfn{store variable} (generally obtained | |
1512 | from a call to @code{gensym}). | |
1513 | ||
1514 | @item | |
1515 | A Lisp form which stores the contents of the store variable into | |
1516 | the generalized variable, assuming the temporaries have been | |
1517 | bound as described above. | |
1518 | ||
1519 | @item | |
1520 | A Lisp form which accesses the contents of the generalized variable, | |
1521 | assuming the temporaries have been bound. | |
1522 | @end enumerate | |
1523 | ||
1524 | This is exactly like the Common Lisp macro of the same name, | |
1525 | except that the method returns a list of five values rather | |
1526 | than the five values themselves, since Emacs Lisp does not | |
1527 | support Common Lisp's notion of multiple return values. | |
1528 | ||
1529 | Once again, the @var{forms} may begin with a documentation string. | |
1530 | ||
1531 | A setf-method should be maximally conservative with regard to | |
1532 | temporary variables. In the setf-methods generated by | |
1533 | @code{defsetf}, the second return value is simply the list of | |
1534 | arguments in the place form, and the first return value is a | |
1535 | list of a corresponding number of temporary variables generated | |
1536 | by @code{gensym}. Macros like @code{setf} and @code{incf} which | |
1537 | use this setf-method will optimize away most temporaries that | |
1538 | turn out to be unnecessary, so there is little reason for the | |
1539 | setf-method itself to optimize. | |
1540 | @end defspec | |
1541 | ||
1542 | @defun get-setf-method place &optional env | |
1543 | This function returns the setf-method for @var{place}, by | |
1544 | invoking the definition previously recorded by @code{defsetf} | |
1545 | or @code{define-setf-method}. The result is a list of five | |
1546 | values as described above. You can use this function to build | |
1547 | your own @code{incf}-like modify macros. (Actually, it is | |
1548 | better to use the internal functions @code{cl-setf-do-modify} | |
1549 | and @code{cl-setf-do-store}, which are a bit easier to use and | |
1550 | which also do a number of optimizations; consult the source | |
1551 | code for the @code{incf} function for a simple example.) | |
1552 | ||
1553 | The argument @var{env} specifies the ``environment'' to be | |
1554 | passed on to @code{macroexpand} if @code{get-setf-method} should | |
1555 | need to expand a macro in @var{place}. It should come from | |
1556 | an @code{&environment} argument to the macro or setf-method | |
1557 | that called @code{get-setf-method}. | |
1558 | ||
1559 | See also the source code for the setf-methods for @code{apply} | |
1560 | and @code{substring}, each of which works by calling | |
1561 | @code{get-setf-method} on a simpler case, then massaging | |
1562 | the result in various ways. | |
1563 | @end defun | |
1564 | ||
1565 | Modern Common Lisp defines a second, independent way to specify | |
1566 | the @code{setf} behavior of a function, namely ``@code{setf} | |
1567 | functions'' whose names are lists @code{(setf @var{name})} | |
1568 | rather than symbols. For example, @code{(defun (setf foo) @dots{})} | |
1569 | defines the function that is used when @code{setf} is applied to | |
1570 | @code{foo}. This package does not currently support @code{setf} | |
1571 | functions. In particular, it is a compile-time error to use | |
1572 | @code{setf} on a form which has not already been @code{defsetf}'d | |
1573 | or otherwise declared; in newer Common Lisps, this would not be | |
1574 | an error since the function @code{(setf @var{func})} might be | |
1575 | defined later. | |
1576 | ||
1577 | @iftex | |
1578 | @secno=4 | |
1579 | @end iftex | |
1580 | ||
1581 | @node Variable Bindings, Conditionals, Generalized Variables, Control Structure | |
1582 | @section Variable Bindings | |
1583 | ||
1584 | @noindent | |
1585 | These Lisp forms make bindings to variables and function names, | |
1586 | analogous to Lisp's built-in @code{let} form. | |
1587 | ||
1588 | @xref{Modify Macros}, for the @code{letf} and @code{letf*} forms which | |
1589 | are also related to variable bindings. | |
1590 | ||
1591 | @menu | |
1592 | * Dynamic Bindings:: The `progv' form | |
1593 | * Lexical Bindings:: `lexical-let' and lexical closures | |
1594 | * Function Bindings:: `flet' and `labels' | |
1595 | * Macro Bindings:: `macrolet' and `symbol-macrolet' | |
1596 | @end menu | |
1597 | ||
1598 | @node Dynamic Bindings, Lexical Bindings, Variable Bindings, Variable Bindings | |
1599 | @subsection Dynamic Bindings | |
1600 | ||
1601 | @noindent | |
1602 | The standard @code{let} form binds variables whose names are known | |
1603 | at compile-time. The @code{progv} form provides an easy way to | |
1604 | bind variables whose names are computed at run-time. | |
1605 | ||
1606 | @defspec progv symbols values forms@dots{} | |
1607 | This form establishes @code{let}-style variable bindings on a | |
1608 | set of variables computed at run-time. The expressions | |
1609 | @var{symbols} and @var{values} are evaluated, and must return lists | |
1610 | of symbols and values, respectively. The symbols are bound to the | |
1611 | corresponding values for the duration of the body @var{form}s. | |
1612 | If @var{values} is shorter than @var{symbols}, the last few symbols | |
1613 | are made unbound (as if by @code{makunbound}) inside the body. | |
1614 | If @var{symbols} is shorter than @var{values}, the excess values | |
1615 | are ignored. | |
1616 | @end defspec | |
1617 | ||
1618 | @node Lexical Bindings, Function Bindings, Dynamic Bindings, Variable Bindings | |
1619 | @subsection Lexical Bindings | |
1620 | ||
1621 | @noindent | |
1622 | The @dfn{CL} package defines the following macro which | |
1623 | more closely follows the Common Lisp @code{let} form: | |
1624 | ||
1625 | @defspec lexical-let (bindings@dots{}) forms@dots{} | |
1626 | This form is exactly like @code{let} except that the bindings it | |
1627 | establishes are purely lexical. Lexical bindings are similar to | |
1628 | local variables in a language like C: Only the code physically | |
1629 | within the body of the @code{lexical-let} (after macro expansion) | |
1630 | may refer to the bound variables. | |
1631 | ||
1632 | @example | |
1633 | (setq a 5) | |
1634 | (defun foo (b) (+ a b)) | |
1635 | (let ((a 2)) (foo a)) | |
1636 | @result{} 4 | |
1637 | (lexical-let ((a 2)) (foo a)) | |
1638 | @result{} 7 | |
1639 | @end example | |
1640 | ||
1641 | @noindent | |
1642 | In this example, a regular @code{let} binding of @code{a} actually | |
1643 | makes a temporary change to the global variable @code{a}, so @code{foo} | |
1644 | is able to see the binding of @code{a} to 2. But @code{lexical-let} | |
1645 | actually creates a distinct local variable @code{a} for use within its | |
1646 | body, without any effect on the global variable of the same name. | |
1647 | ||
1648 | The most important use of lexical bindings is to create @dfn{closures}. | |
1649 | A closure is a function object that refers to an outside lexical | |
1650 | variable. For example: | |
1651 | ||
1652 | @example | |
1653 | (defun make-adder (n) | |
1654 | (lexical-let ((n n)) | |
1655 | (function (lambda (m) (+ n m))))) | |
1656 | (setq add17 (make-adder 17)) | |
1657 | (funcall add17 4) | |
1658 | @result{} 21 | |
1659 | @end example | |
1660 | ||
1661 | @noindent | |
1662 | The call @code{(make-adder 17)} returns a function object which adds | |
1663 | 17 to its argument. If @code{let} had been used instead of | |
1664 | @code{lexical-let}, the function object would have referred to the | |
1665 | global @code{n}, which would have been bound to 17 only during the | |
1666 | call to @code{make-adder} itself. | |
1667 | ||
1668 | @example | |
1669 | (defun make-counter () | |
1670 | (lexical-let ((n 0)) | |
1671 | (function* (lambda (&optional (m 1)) (incf n m))))) | |
1672 | (setq count-1 (make-counter)) | |
1673 | (funcall count-1 3) | |
1674 | @result{} 3 | |
1675 | (funcall count-1 14) | |
1676 | @result{} 17 | |
1677 | (setq count-2 (make-counter)) | |
1678 | (funcall count-2 5) | |
1679 | @result{} 5 | |
1680 | (funcall count-1 2) | |
1681 | @result{} 19 | |
1682 | (funcall count-2) | |
1683 | @result{} 6 | |
1684 | @end example | |
1685 | ||
1686 | @noindent | |
1687 | Here we see that each call to @code{make-counter} creates a distinct | |
1688 | local variable @code{n}, which serves as a private counter for the | |
1689 | function object that is returned. | |
1690 | ||
1691 | Closed-over lexical variables persist until the last reference to | |
1692 | them goes away, just like all other Lisp objects. For example, | |
1693 | @code{count-2} refers to a function object which refers to an | |
1694 | instance of the variable @code{n}; this is the only reference | |
1695 | to that variable, so after @code{(setq count-2 nil)} the garbage | |
1696 | collector would be able to delete this instance of @code{n}. | |
1697 | Of course, if a @code{lexical-let} does not actually create any | |
1698 | closures, then the lexical variables are free as soon as the | |
1699 | @code{lexical-let} returns. | |
1700 | ||
1701 | Many closures are used only during the extent of the bindings they | |
1702 | refer to; these are known as ``downward funargs'' in Lisp parlance. | |
1703 | When a closure is used in this way, regular Emacs Lisp dynamic | |
1704 | bindings suffice and will be more efficient than @code{lexical-let} | |
1705 | closures: | |
1706 | ||
1707 | @example | |
1708 | (defun add-to-list (x list) | |
bed44076 | 1709 | (mapcar (lambda (y) (+ x y))) list) |
6bf7aab6 DL |
1710 | (add-to-list 7 '(1 2 5)) |
1711 | @result{} (8 9 12) | |
1712 | @end example | |
1713 | ||
1714 | @noindent | |
1715 | Since this lambda is only used while @code{x} is still bound, | |
1716 | it is not necessary to make a true closure out of it. | |
1717 | ||
1718 | You can use @code{defun} or @code{flet} inside a @code{lexical-let} | |
1719 | to create a named closure. If several closures are created in the | |
1720 | body of a single @code{lexical-let}, they all close over the same | |
1721 | instance of the lexical variable. | |
1722 | ||
1723 | The @code{lexical-let} form is an extension to Common Lisp. In | |
1724 | true Common Lisp, all bindings are lexical unless declared otherwise. | |
1725 | @end defspec | |
1726 | ||
1727 | @defspec lexical-let* (bindings@dots{}) forms@dots{} | |
1728 | This form is just like @code{lexical-let}, except that the bindings | |
1729 | are made sequentially in the manner of @code{let*}. | |
1730 | @end defspec | |
1731 | ||
1732 | @node Function Bindings, Macro Bindings, Lexical Bindings, Variable Bindings | |
1733 | @subsection Function Bindings | |
1734 | ||
1735 | @noindent | |
1736 | These forms make @code{let}-like bindings to functions instead | |
1737 | of variables. | |
1738 | ||
1739 | @defspec flet (bindings@dots{}) forms@dots{} | |
1740 | This form establishes @code{let}-style bindings on the function | |
1741 | cells of symbols rather than on the value cells. Each @var{binding} | |
1742 | must be a list of the form @samp{(@var{name} @var{arglist} | |
1743 | @var{forms}@dots{})}, which defines a function exactly as if | |
1744 | it were a @code{defun*} form. The function @var{name} is defined | |
1745 | accordingly for the duration of the body of the @code{flet}; then | |
1746 | the old function definition, or lack thereof, is restored. | |
1747 | ||
1748 | While @code{flet} in Common Lisp establishes a lexical binding of | |
1749 | @var{name}, Emacs Lisp @code{flet} makes a dynamic binding. The | |
1750 | result is that @code{flet} affects indirect calls to a function as | |
1751 | well as calls directly inside the @code{flet} form itself. | |
1752 | ||
1753 | You can use @code{flet} to disable or modify the behavior of a | |
1754 | function in a temporary fashion. This will even work on Emacs | |
1755 | primitives, although note that some calls to primitive functions | |
1756 | internal to Emacs are made without going through the symbol's | |
1757 | function cell, and so will not be affected by @code{flet}. For | |
1758 | example, | |
1759 | ||
1760 | @example | |
1761 | (flet ((message (&rest args) (push args saved-msgs))) | |
1762 | (do-something)) | |
1763 | @end example | |
1764 | ||
1765 | This code attempts to replace the built-in function @code{message} | |
1766 | with a function that simply saves the messages in a list rather | |
1767 | than displaying them. The original definition of @code{message} | |
1768 | will be restored after @code{do-something} exits. This code will | |
1769 | work fine on messages generated by other Lisp code, but messages | |
1770 | generated directly inside Emacs will not be caught since they make | |
1771 | direct C-language calls to the message routines rather than going | |
1772 | through the Lisp @code{message} function. | |
1773 | ||
1774 | Functions defined by @code{flet} may use the full Common Lisp | |
1775 | argument notation supported by @code{defun*}; also, the function | |
1776 | body is enclosed in an implicit block as if by @code{defun*}. | |
1777 | @xref{Program Structure}. | |
1778 | @end defspec | |
1779 | ||
1780 | @defspec labels (bindings@dots{}) forms@dots{} | |
1781 | The @code{labels} form is like @code{flet}, except that it | |
1782 | makes lexical bindings of the function names rather than | |
1783 | dynamic bindings. (In true Common Lisp, both @code{flet} and | |
1784 | @code{labels} make lexical bindings of slightly different sorts; | |
1785 | since Emacs Lisp is dynamically bound by default, it seemed | |
1786 | more appropriate for @code{flet} also to use dynamic binding. | |
1787 | The @code{labels} form, with its lexical binding, is fully | |
1788 | compatible with Common Lisp.) | |
1789 | ||
1790 | Lexical scoping means that all references to the named | |
1791 | functions must appear physically within the body of the | |
1792 | @code{labels} form. References may appear both in the body | |
1793 | @var{forms} of @code{labels} itself, and in the bodies of | |
1794 | the functions themselves. Thus, @code{labels} can define | |
1795 | local recursive functions, or mutually-recursive sets of | |
1796 | functions. | |
1797 | ||
1798 | A ``reference'' to a function name is either a call to that | |
1799 | function, or a use of its name quoted by @code{quote} or | |
1800 | @code{function} to be passed on to, say, @code{mapcar}. | |
1801 | @end defspec | |
1802 | ||
c6ad01e8 | 1803 | @node Macro Bindings, , Function Bindings, Variable Bindings |
6bf7aab6 DL |
1804 | @subsection Macro Bindings |
1805 | ||
1806 | @noindent | |
1807 | These forms create local macros and ``symbol macros.'' | |
1808 | ||
1809 | @defspec macrolet (bindings@dots{}) forms@dots{} | |
1810 | This form is analogous to @code{flet}, but for macros instead of | |
1811 | functions. Each @var{binding} is a list of the same form as the | |
1812 | arguments to @code{defmacro*} (i.e., a macro name, argument list, | |
1813 | and macro-expander forms). The macro is defined accordingly for | |
1814 | use within the body of the @code{macrolet}. | |
1815 | ||
1816 | Because of the nature of macros, @code{macrolet} is lexically | |
1817 | scoped even in Emacs Lisp: The @code{macrolet} binding will | |
1818 | affect only calls that appear physically within the body | |
1819 | @var{forms}, possibly after expansion of other macros in the | |
1820 | body. | |
1821 | @end defspec | |
1822 | ||
1823 | @defspec symbol-macrolet (bindings@dots{}) forms@dots{} | |
1824 | This form creates @dfn{symbol macros}, which are macros that look | |
1825 | like variable references rather than function calls. Each | |
1826 | @var{binding} is a list @samp{(@var{var} @var{expansion})}; | |
1827 | any reference to @var{var} within the body @var{forms} is | |
1828 | replaced by @var{expansion}. | |
1829 | ||
1830 | @example | |
1831 | (setq bar '(5 . 9)) | |
1832 | (symbol-macrolet ((foo (car bar))) | |
1833 | (incf foo)) | |
1834 | bar | |
1835 | @result{} (6 . 9) | |
1836 | @end example | |
1837 | ||
1838 | A @code{setq} of a symbol macro is treated the same as a @code{setf}. | |
1839 | I.e., @code{(setq foo 4)} in the above would be equivalent to | |
1840 | @code{(setf foo 4)}, which in turn expands to @code{(setf (car bar) 4)}. | |
1841 | ||
1842 | Likewise, a @code{let} or @code{let*} binding a symbol macro is | |
1843 | treated like a @code{letf} or @code{letf*}. This differs from true | |
1844 | Common Lisp, where the rules of lexical scoping cause a @code{let} | |
1845 | binding to shadow a @code{symbol-macrolet} binding. In this package, | |
1846 | only @code{lexical-let} and @code{lexical-let*} will shadow a symbol | |
1847 | macro. | |
1848 | ||
1849 | There is no analogue of @code{defmacro} for symbol macros; all symbol | |
1850 | macros are local. A typical use of @code{symbol-macrolet} is in the | |
1851 | expansion of another macro: | |
1852 | ||
1853 | @example | |
1854 | (defmacro* my-dolist ((x list) &rest body) | |
1855 | (let ((var (gensym))) | |
1856 | (list 'loop 'for var 'on list 'do | |
1857 | (list* 'symbol-macrolet (list (list x (list 'car var))) | |
1858 | body)))) | |
1859 | ||
1860 | (setq mylist '(1 2 3 4)) | |
1861 | (my-dolist (x mylist) (incf x)) | |
1862 | mylist | |
1863 | @result{} (2 3 4 5) | |
1864 | @end example | |
1865 | ||
1866 | @noindent | |
1867 | In this example, the @code{my-dolist} macro is similar to @code{dolist} | |
2995b042 | 1868 | (@pxref{Iteration}) except that the variable @code{x} becomes a true |
6bf7aab6 DL |
1869 | reference onto the elements of the list. The @code{my-dolist} call |
1870 | shown here expands to | |
1871 | ||
1872 | @example | |
1873 | (loop for G1234 on mylist do | |
1874 | (symbol-macrolet ((x (car G1234))) | |
1875 | (incf x))) | |
1876 | @end example | |
1877 | ||
1878 | @noindent | |
1879 | which in turn expands to | |
1880 | ||
1881 | @example | |
1882 | (loop for G1234 on mylist do (incf (car G1234))) | |
1883 | @end example | |
1884 | ||
1885 | @xref{Loop Facility}, for a description of the @code{loop} macro. | |
1886 | This package defines a nonstandard @code{in-ref} loop clause that | |
1887 | works much like @code{my-dolist}. | |
1888 | @end defspec | |
1889 | ||
1890 | @node Conditionals, Blocks and Exits, Variable Bindings, Control Structure | |
1891 | @section Conditionals | |
1892 | ||
1893 | @noindent | |
1894 | These conditional forms augment Emacs Lisp's simple @code{if}, | |
1895 | @code{and}, @code{or}, and @code{cond} forms. | |
1896 | ||
1897 | @defspec case keyform clause@dots{} | |
1898 | This macro evaluates @var{keyform}, then compares it with the key | |
1899 | values listed in the various @var{clause}s. Whichever clause matches | |
1900 | the key is executed; comparison is done by @code{eql}. If no clause | |
1901 | matches, the @code{case} form returns @code{nil}. The clauses are | |
1902 | of the form | |
1903 | ||
1904 | @example | |
1905 | (@var{keylist} @var{body-forms}@dots{}) | |
1906 | @end example | |
1907 | ||
1908 | @noindent | |
1909 | where @var{keylist} is a list of key values. If there is exactly | |
1910 | one value, and it is not a cons cell or the symbol @code{nil} or | |
1911 | @code{t}, then it can be used by itself as a @var{keylist} without | |
1912 | being enclosed in a list. All key values in the @code{case} form | |
1913 | must be distinct. The final clauses may use @code{t} in place of | |
1914 | a @var{keylist} to indicate a default clause that should be taken | |
1915 | if none of the other clauses match. (The symbol @code{otherwise} | |
1916 | is also recognized in place of @code{t}. To make a clause that | |
1917 | matches the actual symbol @code{t}, @code{nil}, or @code{otherwise}, | |
1918 | enclose the symbol in a list.) | |
1919 | ||
1920 | For example, this expression reads a keystroke, then does one of | |
1921 | four things depending on whether it is an @samp{a}, a @samp{b}, | |
1922 | a @key{RET} or @kbd{C-j}, or anything else. | |
1923 | ||
1924 | @example | |
1925 | (case (read-char) | |
1926 | (?a (do-a-thing)) | |
1927 | (?b (do-b-thing)) | |
1928 | ((?\r ?\n) (do-ret-thing)) | |
1929 | (t (do-other-thing))) | |
1930 | @end example | |
1931 | @end defspec | |
1932 | ||
1933 | @defspec ecase keyform clause@dots{} | |
1934 | This macro is just like @code{case}, except that if the key does | |
1935 | not match any of the clauses, an error is signaled rather than | |
1936 | simply returning @code{nil}. | |
1937 | @end defspec | |
1938 | ||
1939 | @defspec typecase keyform clause@dots{} | |
1940 | This macro is a version of @code{case} that checks for types | |
1941 | rather than values. Each @var{clause} is of the form | |
1942 | @samp{(@var{type} @var{body}...)}. @xref{Type Predicates}, | |
1943 | for a description of type specifiers. For example, | |
1944 | ||
1945 | @example | |
1946 | (typecase x | |
1947 | (integer (munch-integer x)) | |
1948 | (float (munch-float x)) | |
1949 | (string (munch-integer (string-to-int x))) | |
1950 | (t (munch-anything x))) | |
1951 | @end example | |
1952 | ||
1953 | The type specifier @code{t} matches any type of object; the word | |
1954 | @code{otherwise} is also allowed. To make one clause match any of | |
1955 | several types, use an @code{(or ...)} type specifier. | |
1956 | @end defspec | |
1957 | ||
1958 | @defspec etypecase keyform clause@dots{} | |
1959 | This macro is just like @code{typecase}, except that if the key does | |
1960 | not match any of the clauses, an error is signaled rather than | |
1961 | simply returning @code{nil}. | |
1962 | @end defspec | |
1963 | ||
1964 | @node Blocks and Exits, Iteration, Conditionals, Control Structure | |
1965 | @section Blocks and Exits | |
1966 | ||
1967 | @noindent | |
1968 | Common Lisp @dfn{blocks} provide a non-local exit mechanism very | |
1969 | similar to @code{catch} and @code{throw}, but lexically rather than | |
1970 | dynamically scoped. This package actually implements @code{block} | |
1971 | in terms of @code{catch}; however, the lexical scoping allows the | |
1972 | optimizing byte-compiler to omit the costly @code{catch} step if the | |
1973 | body of the block does not actually @code{return-from} the block. | |
1974 | ||
1975 | @defspec block name forms@dots{} | |
1976 | The @var{forms} are evaluated as if by a @code{progn}. However, | |
1977 | if any of the @var{forms} execute @code{(return-from @var{name})}, | |
1978 | they will jump out and return directly from the @code{block} form. | |
1979 | The @code{block} returns the result of the last @var{form} unless | |
1980 | a @code{return-from} occurs. | |
1981 | ||
1982 | The @code{block}/@code{return-from} mechanism is quite similar to | |
1983 | the @code{catch}/@code{throw} mechanism. The main differences are | |
1984 | that block @var{name}s are unevaluated symbols, rather than forms | |
1985 | (such as quoted symbols) which evaluate to a tag at run-time; and | |
1986 | also that blocks are lexically scoped whereas @code{catch}/@code{throw} | |
1987 | are dynamically scoped. This means that functions called from the | |
1988 | body of a @code{catch} can also @code{throw} to the @code{catch}, | |
1989 | but the @code{return-from} referring to a block name must appear | |
1990 | physically within the @var{forms} that make up the body of the block. | |
1991 | They may not appear within other called functions, although they may | |
1992 | appear within macro expansions or @code{lambda}s in the body. Block | |
1993 | names and @code{catch} names form independent name-spaces. | |
1994 | ||
1995 | In true Common Lisp, @code{defun} and @code{defmacro} surround | |
1996 | the function or expander bodies with implicit blocks with the | |
1997 | same name as the function or macro. This does not occur in Emacs | |
1998 | Lisp, but this package provides @code{defun*} and @code{defmacro*} | |
1999 | forms which do create the implicit block. | |
2000 | ||
2001 | The Common Lisp looping constructs defined by this package, | |
2995b042 | 2002 | such as @code{loop} and @code{dolist}, also create implicit blocks |
6bf7aab6 DL |
2003 | just as in Common Lisp. |
2004 | ||
2005 | Because they are implemented in terms of Emacs Lisp @code{catch} | |
2006 | and @code{throw}, blocks have the same overhead as actual | |
2007 | @code{catch} constructs (roughly two function calls). However, | |
177c0ea7 | 2008 | the optimizing byte compiler will optimize away the @code{catch} |
c6ad01e8 | 2009 | if the block does |
6bf7aab6 DL |
2010 | not in fact contain any @code{return} or @code{return-from} calls |
2011 | that jump to it. This means that @code{do} loops and @code{defun*} | |
2012 | functions which don't use @code{return} don't pay the overhead to | |
2013 | support it. | |
2014 | @end defspec | |
2015 | ||
2016 | @defspec return-from name [result] | |
2017 | This macro returns from the block named @var{name}, which must be | |
2018 | an (unevaluated) symbol. If a @var{result} form is specified, it | |
2019 | is evaluated to produce the result returned from the @code{block}. | |
2020 | Otherwise, @code{nil} is returned. | |
2021 | @end defspec | |
2022 | ||
2023 | @defspec return [result] | |
2024 | This macro is exactly like @code{(return-from nil @var{result})}. | |
2995b042 | 2025 | Common Lisp loops like @code{do} and @code{dolist} implicitly enclose |
6bf7aab6 DL |
2026 | themselves in @code{nil} blocks. |
2027 | @end defspec | |
2028 | ||
2029 | @node Iteration, Loop Facility, Blocks and Exits, Control Structure | |
2030 | @section Iteration | |
2031 | ||
2032 | @noindent | |
2033 | The macros described here provide more sophisticated, high-level | |
2034 | looping constructs to complement Emacs Lisp's basic @code{while} | |
2035 | loop. | |
2036 | ||
2037 | @defspec loop forms@dots{} | |
2038 | The @dfn{CL} package supports both the simple, old-style meaning of | |
2039 | @code{loop} and the extremely powerful and flexible feature known as | |
2040 | the @dfn{Loop Facility} or @dfn{Loop Macro}. This more advanced | |
2041 | facility is discussed in the following section; @pxref{Loop Facility}. | |
2042 | The simple form of @code{loop} is described here. | |
2043 | ||
2044 | If @code{loop} is followed by zero or more Lisp expressions, | |
2045 | then @code{(loop @var{exprs}@dots{})} simply creates an infinite | |
2046 | loop executing the expressions over and over. The loop is | |
2047 | enclosed in an implicit @code{nil} block. Thus, | |
2048 | ||
2049 | @example | |
2050 | (loop (foo) (if (no-more) (return 72)) (bar)) | |
2051 | @end example | |
2052 | ||
2053 | @noindent | |
2054 | is exactly equivalent to | |
2055 | ||
2056 | @example | |
2057 | (block nil (while t (foo) (if (no-more) (return 72)) (bar))) | |
2058 | @end example | |
2059 | ||
2060 | If any of the expressions are plain symbols, the loop is instead | |
2061 | interpreted as a Loop Macro specification as described later. | |
2062 | (This is not a restriction in practice, since a plain symbol | |
2063 | in the above notation would simply access and throw away the | |
2064 | value of a variable.) | |
2065 | @end defspec | |
2066 | ||
2067 | @defspec do (spec@dots{}) (end-test [result@dots{}]) forms@dots{} | |
2068 | This macro creates a general iterative loop. Each @var{spec} is | |
2069 | of the form | |
2070 | ||
2071 | @example | |
2072 | (@var{var} [@var{init} [@var{step}]]) | |
2073 | @end example | |
2074 | ||
2075 | The loop works as follows: First, each @var{var} is bound to the | |
2076 | associated @var{init} value as if by a @code{let} form. Then, in | |
2077 | each iteration of the loop, the @var{end-test} is evaluated; if | |
2078 | true, the loop is finished. Otherwise, the body @var{forms} are | |
2079 | evaluated, then each @var{var} is set to the associated @var{step} | |
2080 | expression (as if by a @code{psetq} form) and the next iteration | |
2081 | begins. Once the @var{end-test} becomes true, the @var{result} | |
2082 | forms are evaluated (with the @var{var}s still bound to their | |
2083 | values) to produce the result returned by @code{do}. | |
2084 | ||
2085 | The entire @code{do} loop is enclosed in an implicit @code{nil} | |
2086 | block, so that you can use @code{(return)} to break out of the | |
2087 | loop at any time. | |
2088 | ||
2089 | If there are no @var{result} forms, the loop returns @code{nil}. | |
2090 | If a given @var{var} has no @var{step} form, it is bound to its | |
2091 | @var{init} value but not otherwise modified during the @code{do} | |
2092 | loop (unless the code explicitly modifies it); this case is just | |
2093 | a shorthand for putting a @code{(let ((@var{var} @var{init})) @dots{})} | |
2094 | around the loop. If @var{init} is also omitted it defaults to | |
2095 | @code{nil}, and in this case a plain @samp{@var{var}} can be used | |
2096 | in place of @samp{(@var{var})}, again following the analogy with | |
2097 | @code{let}. | |
2098 | ||
2099 | This example (from Steele) illustrates a loop which applies the | |
2100 | function @code{f} to successive pairs of values from the lists | |
2101 | @code{foo} and @code{bar}; it is equivalent to the call | |
2102 | @code{(mapcar* 'f foo bar)}. Note that this loop has no body | |
2103 | @var{forms} at all, performing all its work as side effects of | |
2104 | the rest of the loop. | |
2105 | ||
2106 | @example | |
2107 | (do ((x foo (cdr x)) | |
2108 | (y bar (cdr y)) | |
2109 | (z nil (cons (f (car x) (car y)) z))) | |
2110 | ((or (null x) (null y)) | |
2111 | (nreverse z))) | |
2112 | @end example | |
2113 | @end defspec | |
2114 | ||
2115 | @defspec do* (spec@dots{}) (end-test [result@dots{}]) forms@dots{} | |
2116 | This is to @code{do} what @code{let*} is to @code{let}. In | |
2117 | particular, the initial values are bound as if by @code{let*} | |
2118 | rather than @code{let}, and the steps are assigned as if by | |
2119 | @code{setq} rather than @code{psetq}. | |
2120 | ||
2121 | Here is another way to write the above loop: | |
2122 | ||
2123 | @example | |
2124 | (do* ((xp foo (cdr xp)) | |
2125 | (yp bar (cdr yp)) | |
2126 | (x (car xp) (car xp)) | |
2127 | (y (car yp) (car yp)) | |
2128 | z) | |
2129 | ((or (null xp) (null yp)) | |
2130 | (nreverse z)) | |
2131 | (push (f x y) z)) | |
2132 | @end example | |
2133 | @end defspec | |
2134 | ||
2995b042 DL |
2135 | @defspec dolist (var list [result]) forms@dots{} |
2136 | This is a more specialized loop which iterates across the elements | |
2137 | of a list. @var{list} should evaluate to a list; the body @var{forms} | |
2138 | are executed with @var{var} bound to each element of the list in | |
2139 | turn. Finally, the @var{result} form (or @code{nil}) is evaluated | |
2140 | with @var{var} bound to @code{nil} to produce the result returned by | |
2141 | the loop. Unlike with Emacs's built in @code{dolist}, the loop is | |
2142 | surrounded by an implicit @code{nil} block. | |
2143 | @end defspec | |
2144 | ||
2145 | @defspec dotimes (var count [result]) forms@dots{} | |
2146 | This is a more specialized loop which iterates a specified number | |
2147 | of times. The body is executed with @var{var} bound to the integers | |
2148 | from zero (inclusive) to @var{count} (exclusive), in turn. Then | |
2149 | the @code{result} form is evaluated with @var{var} bound to the total | |
2150 | number of iterations that were done (i.e., @code{(max 0 @var{count})}) | |
2151 | to get the return value for the loop form. Unlike with Emacs's built in | |
2152 | @code{dolist}, the loop is surrounded by an implicit @code{nil} block. | |
2153 | @end defspec | |
2154 | ||
6bf7aab6 DL |
2155 | @defspec do-symbols (var [obarray [result]]) forms@dots{} |
2156 | This loop iterates over all interned symbols. If @var{obarray} | |
2157 | is specified and is not @code{nil}, it loops over all symbols in | |
2158 | that obarray. For each symbol, the body @var{forms} are evaluated | |
2159 | with @var{var} bound to that symbol. The symbols are visited in | |
2160 | an unspecified order. Afterward the @var{result} form, if any, | |
2161 | is evaluated (with @var{var} bound to @code{nil}) to get the return | |
2162 | value. The loop is surrounded by an implicit @code{nil} block. | |
2163 | @end defspec | |
2164 | ||
2165 | @defspec do-all-symbols (var [result]) forms@dots{} | |
2166 | This is identical to @code{do-symbols} except that the @var{obarray} | |
2167 | argument is omitted; it always iterates over the default obarray. | |
2168 | @end defspec | |
2169 | ||
2170 | @xref{Mapping over Sequences}, for some more functions for | |
2171 | iterating over vectors or lists. | |
2172 | ||
2173 | @node Loop Facility, Multiple Values, Iteration, Control Structure | |
2174 | @section Loop Facility | |
2175 | ||
2176 | @noindent | |
2177 | A common complaint with Lisp's traditional looping constructs is | |
2178 | that they are either too simple and limited, such as Common Lisp's | |
2179 | @code{dotimes} or Emacs Lisp's @code{while}, or too unreadable and | |
2180 | obscure, like Common Lisp's @code{do} loop. | |
2181 | ||
2182 | To remedy this, recent versions of Common Lisp have added a new | |
2183 | construct called the ``Loop Facility'' or ``@code{loop} macro,'' | |
2184 | with an easy-to-use but very powerful and expressive syntax. | |
2185 | ||
2186 | @menu | |
2187 | * Loop Basics:: `loop' macro, basic clause structure | |
2188 | * Loop Examples:: Working examples of `loop' macro | |
2189 | * For Clauses:: Clauses introduced by `for' or `as' | |
2190 | * Iteration Clauses:: `repeat', `while', `thereis', etc. | |
2191 | * Accumulation Clauses:: `collect', `sum', `maximize', etc. | |
2192 | * Other Clauses:: `with', `if', `initially', `finally' | |
2193 | @end menu | |
2194 | ||
2195 | @node Loop Basics, Loop Examples, Loop Facility, Loop Facility | |
2196 | @subsection Loop Basics | |
2197 | ||
2198 | @noindent | |
2199 | The @code{loop} macro essentially creates a mini-language within | |
2200 | Lisp that is specially tailored for describing loops. While this | |
2201 | language is a little strange-looking by the standards of regular Lisp, | |
2202 | it turns out to be very easy to learn and well-suited to its purpose. | |
2203 | ||
2204 | Since @code{loop} is a macro, all parsing of the loop language | |
2205 | takes place at byte-compile time; compiled @code{loop}s are just | |
2206 | as efficient as the equivalent @code{while} loops written longhand. | |
2207 | ||
2208 | @defspec loop clauses@dots{} | |
2209 | A loop construct consists of a series of @var{clause}s, each | |
2210 | introduced by a symbol like @code{for} or @code{do}. Clauses | |
2211 | are simply strung together in the argument list of @code{loop}, | |
2212 | with minimal extra parentheses. The various types of clauses | |
2213 | specify initializations, such as the binding of temporary | |
2214 | variables, actions to be taken in the loop, stepping actions, | |
2215 | and final cleanup. | |
2216 | ||
2217 | Common Lisp specifies a certain general order of clauses in a | |
2218 | loop: | |
2219 | ||
2220 | @example | |
2221 | (loop @var{name-clause} | |
2222 | @var{var-clauses}@dots{} | |
2223 | @var{action-clauses}@dots{}) | |
2224 | @end example | |
2225 | ||
2226 | The @var{name-clause} optionally gives a name to the implicit | |
2227 | block that surrounds the loop. By default, the implicit block | |
2228 | is named @code{nil}. The @var{var-clauses} specify what | |
2229 | variables should be bound during the loop, and how they should | |
2230 | be modified or iterated throughout the course of the loop. The | |
2231 | @var{action-clauses} are things to be done during the loop, such | |
2232 | as computing, collecting, and returning values. | |
2233 | ||
2234 | The Emacs version of the @code{loop} macro is less restrictive about | |
2235 | the order of clauses, but things will behave most predictably if | |
2236 | you put the variable-binding clauses @code{with}, @code{for}, and | |
2237 | @code{repeat} before the action clauses. As in Common Lisp, | |
2238 | @code{initially} and @code{finally} clauses can go anywhere. | |
2239 | ||
2240 | Loops generally return @code{nil} by default, but you can cause | |
2241 | them to return a value by using an accumulation clause like | |
2242 | @code{collect}, an end-test clause like @code{always}, or an | |
2243 | explicit @code{return} clause to jump out of the implicit block. | |
2244 | (Because the loop body is enclosed in an implicit block, you can | |
2245 | also use regular Lisp @code{return} or @code{return-from} to | |
2246 | break out of the loop.) | |
2247 | @end defspec | |
2248 | ||
2249 | The following sections give some examples of the Loop Macro in | |
2250 | action, and describe the particular loop clauses in great detail. | |
2251 | Consult the second edition of Steele's @dfn{Common Lisp, the Language}, | |
2252 | for additional discussion and examples of the @code{loop} macro. | |
2253 | ||
2254 | @node Loop Examples, For Clauses, Loop Basics, Loop Facility | |
2255 | @subsection Loop Examples | |
2256 | ||
2257 | @noindent | |
2258 | Before listing the full set of clauses that are allowed, let's | |
2259 | look at a few example loops just to get a feel for the @code{loop} | |
2260 | language. | |
2261 | ||
2262 | @example | |
2263 | (loop for buf in (buffer-list) | |
2264 | collect (buffer-file-name buf)) | |
2265 | @end example | |
2266 | ||
2267 | @noindent | |
2268 | This loop iterates over all Emacs buffers, using the list | |
2269 | returned by @code{buffer-list}. For each buffer @code{buf}, | |
2270 | it calls @code{buffer-file-name} and collects the results into | |
2271 | a list, which is then returned from the @code{loop} construct. | |
2272 | The result is a list of the file names of all the buffers in | |
2273 | Emacs' memory. The words @code{for}, @code{in}, and @code{collect} | |
2274 | are reserved words in the @code{loop} language. | |
2275 | ||
2276 | @example | |
2277 | (loop repeat 20 do (insert "Yowsa\n")) | |
2278 | @end example | |
2279 | ||
2280 | @noindent | |
2281 | This loop inserts the phrase ``Yowsa'' twenty times in the | |
2282 | current buffer. | |
2283 | ||
2284 | @example | |
2285 | (loop until (eobp) do (munch-line) (forward-line 1)) | |
2286 | @end example | |
2287 | ||
2288 | @noindent | |
2289 | This loop calls @code{munch-line} on every line until the end | |
2290 | of the buffer. If point is already at the end of the buffer, | |
2291 | the loop exits immediately. | |
2292 | ||
2293 | @example | |
2294 | (loop do (munch-line) until (eobp) do (forward-line 1)) | |
2295 | @end example | |
2296 | ||
2297 | @noindent | |
2298 | This loop is similar to the above one, except that @code{munch-line} | |
2299 | is always called at least once. | |
2300 | ||
2301 | @example | |
2302 | (loop for x from 1 to 100 | |
2303 | for y = (* x x) | |
2304 | until (>= y 729) | |
2305 | finally return (list x (= y 729))) | |
2306 | @end example | |
2307 | ||
2308 | @noindent | |
2309 | This more complicated loop searches for a number @code{x} whose | |
2310 | square is 729. For safety's sake it only examines @code{x} | |
2311 | values up to 100; dropping the phrase @samp{to 100} would | |
2312 | cause the loop to count upwards with no limit. The second | |
2313 | @code{for} clause defines @code{y} to be the square of @code{x} | |
2314 | within the loop; the expression after the @code{=} sign is | |
2315 | reevaluated each time through the loop. The @code{until} | |
2316 | clause gives a condition for terminating the loop, and the | |
2317 | @code{finally} clause says what to do when the loop finishes. | |
2318 | (This particular example was written less concisely than it | |
2319 | could have been, just for the sake of illustration.) | |
2320 | ||
2321 | Note that even though this loop contains three clauses (two | |
2322 | @code{for}s and an @code{until}) that would have been enough to | |
2323 | define loops all by themselves, it still creates a single loop | |
2324 | rather than some sort of triple-nested loop. You must explicitly | |
2325 | nest your @code{loop} constructs if you want nested loops. | |
2326 | ||
2327 | @node For Clauses, Iteration Clauses, Loop Examples, Loop Facility | |
2328 | @subsection For Clauses | |
2329 | ||
2330 | @noindent | |
2331 | Most loops are governed by one or more @code{for} clauses. | |
2332 | A @code{for} clause simultaneously describes variables to be | |
2333 | bound, how those variables are to be stepped during the loop, | |
2334 | and usually an end condition based on those variables. | |
2335 | ||
2336 | The word @code{as} is a synonym for the word @code{for}. This | |
2337 | word is followed by a variable name, then a word like @code{from} | |
2338 | or @code{across} that describes the kind of iteration desired. | |
2339 | In Common Lisp, the phrase @code{being the} sometimes precedes | |
2340 | the type of iteration; in this package both @code{being} and | |
2341 | @code{the} are optional. The word @code{each} is a synonym | |
2342 | for @code{the}, and the word that follows it may be singular | |
2343 | or plural: @samp{for x being the elements of y} or | |
2344 | @samp{for x being each element of y}. Which form you use | |
2345 | is purely a matter of style. | |
2346 | ||
2347 | The variable is bound around the loop as if by @code{let}: | |
2348 | ||
2349 | @example | |
2350 | (setq i 'happy) | |
2351 | (loop for i from 1 to 10 do (do-something-with i)) | |
2352 | i | |
2353 | @result{} happy | |
2354 | @end example | |
2355 | ||
2356 | @table @code | |
2357 | @item for @var{var} from @var{expr1} to @var{expr2} by @var{expr3} | |
2358 | This type of @code{for} clause creates a counting loop. Each of | |
2359 | the three sub-terms is optional, though there must be at least one | |
2360 | term so that the clause is marked as a counting clause. | |
2361 | ||
2362 | The three expressions are the starting value, the ending value, and | |
2363 | the step value, respectively, of the variable. The loop counts | |
2364 | upwards by default (@var{expr3} must be positive), from @var{expr1} | |
2365 | to @var{expr2} inclusively. If you omit the @code{from} term, the | |
2366 | loop counts from zero; if you omit the @code{to} term, the loop | |
2367 | counts forever without stopping (unless stopped by some other | |
2368 | loop clause, of course); if you omit the @code{by} term, the loop | |
2369 | counts in steps of one. | |
2370 | ||
2371 | You can replace the word @code{from} with @code{upfrom} or | |
2372 | @code{downfrom} to indicate the direction of the loop. Likewise, | |
2373 | you can replace @code{to} with @code{upto} or @code{downto}. | |
2374 | For example, @samp{for x from 5 downto 1} executes five times | |
2375 | with @code{x} taking on the integers from 5 down to 1 in turn. | |
2376 | Also, you can replace @code{to} with @code{below} or @code{above}, | |
2377 | which are like @code{upto} and @code{downto} respectively except | |
2378 | that they are exclusive rather than inclusive limits: | |
2379 | ||
2380 | @example | |
2381 | (loop for x to 10 collect x) | |
2382 | @result{} (0 1 2 3 4 5 6 7 8 9 10) | |
2383 | (loop for x below 10 collect x) | |
2384 | @result{} (0 1 2 3 4 5 6 7 8 9) | |
2385 | @end example | |
2386 | ||
2387 | The @code{by} value is always positive, even for downward-counting | |
2388 | loops. Some sort of @code{from} value is required for downward | |
2cbd16b9 | 2389 | loops; @samp{for x downto 5} is not a valid loop clause all by |
6bf7aab6 DL |
2390 | itself. |
2391 | ||
2392 | @item for @var{var} in @var{list} by @var{function} | |
2393 | This clause iterates @var{var} over all the elements of @var{list}, | |
2394 | in turn. If you specify the @code{by} term, then @var{function} | |
2395 | is used to traverse the list instead of @code{cdr}; it must be a | |
2396 | function taking one argument. For example: | |
2397 | ||
2398 | @example | |
2399 | (loop for x in '(1 2 3 4 5 6) collect (* x x)) | |
2400 | @result{} (1 4 9 16 25 36) | |
2401 | (loop for x in '(1 2 3 4 5 6) by 'cddr collect (* x x)) | |
2402 | @result{} (1 9 25) | |
2403 | @end example | |
2404 | ||
2405 | @item for @var{var} on @var{list} by @var{function} | |
2406 | This clause iterates @var{var} over all the cons cells of @var{list}. | |
2407 | ||
2408 | @example | |
2409 | (loop for x on '(1 2 3 4) collect x) | |
2410 | @result{} ((1 2 3 4) (2 3 4) (3 4) (4)) | |
2411 | @end example | |
2412 | ||
2413 | With @code{by}, there is no real reason that the @code{on} expression | |
2414 | must be a list. For example: | |
2415 | ||
2416 | @example | |
2417 | (loop for x on first-animal by 'next-animal collect x) | |
2418 | @end example | |
2419 | ||
2420 | @noindent | |
2421 | where @code{(next-animal x)} takes an ``animal'' @var{x} and returns | |
2422 | the next in the (assumed) sequence of animals, or @code{nil} if | |
2423 | @var{x} was the last animal in the sequence. | |
2424 | ||
2425 | @item for @var{var} in-ref @var{list} by @var{function} | |
2426 | This is like a regular @code{in} clause, but @var{var} becomes | |
2427 | a @code{setf}-able ``reference'' onto the elements of the list | |
2428 | rather than just a temporary variable. For example, | |
2429 | ||
2430 | @example | |
2431 | (loop for x in-ref my-list do (incf x)) | |
2432 | @end example | |
2433 | ||
2434 | @noindent | |
2435 | increments every element of @code{my-list} in place. This clause | |
2436 | is an extension to standard Common Lisp. | |
2437 | ||
2438 | @item for @var{var} across @var{array} | |
2439 | This clause iterates @var{var} over all the elements of @var{array}, | |
2440 | which may be a vector or a string. | |
2441 | ||
2442 | @example | |
2443 | (loop for x across "aeiou" | |
2444 | do (use-vowel (char-to-string x))) | |
2445 | @end example | |
2446 | ||
2447 | @item for @var{var} across-ref @var{array} | |
2448 | This clause iterates over an array, with @var{var} a @code{setf}-able | |
2449 | reference onto the elements; see @code{in-ref} above. | |
2450 | ||
2451 | @item for @var{var} being the elements of @var{sequence} | |
2452 | This clause iterates over the elements of @var{sequence}, which may | |
2453 | be a list, vector, or string. Since the type must be determined | |
2454 | at run-time, this is somewhat less efficient than @code{in} or | |
2455 | @code{across}. The clause may be followed by the additional term | |
2456 | @samp{using (index @var{var2})} to cause @var{var2} to be bound to | |
2457 | the successive indices (starting at 0) of the elements. | |
2458 | ||
2459 | This clause type is taken from older versions of the @code{loop} macro, | |
2460 | and is not present in modern Common Lisp. The @samp{using (sequence ...)} | |
2461 | term of the older macros is not supported. | |
2462 | ||
2463 | @item for @var{var} being the elements of-ref @var{sequence} | |
2464 | This clause iterates over a sequence, with @var{var} a @code{setf}-able | |
2465 | reference onto the elements; see @code{in-ref} above. | |
2466 | ||
2467 | @item for @var{var} being the symbols [of @var{obarray}] | |
2468 | This clause iterates over symbols, either over all interned symbols | |
2469 | or over all symbols in @var{obarray}. The loop is executed with | |
2470 | @var{var} bound to each symbol in turn. The symbols are visited in | |
2471 | an unspecified order. | |
2472 | ||
2473 | As an example, | |
2474 | ||
2475 | @example | |
2476 | (loop for sym being the symbols | |
2477 | when (fboundp sym) | |
2478 | when (string-match "^map" (symbol-name sym)) | |
2479 | collect sym) | |
2480 | @end example | |
2481 | ||
2482 | @noindent | |
2483 | returns a list of all the functions whose names begin with @samp{map}. | |
2484 | ||
2485 | The Common Lisp words @code{external-symbols} and @code{present-symbols} | |
2486 | are also recognized but are equivalent to @code{symbols} in Emacs Lisp. | |
2487 | ||
2488 | Due to a minor implementation restriction, it will not work to have | |
2489 | more than one @code{for} clause iterating over symbols, hash tables, | |
2490 | keymaps, overlays, or intervals in a given @code{loop}. Fortunately, | |
2cbd16b9 | 2491 | it would rarely if ever be useful to do so. It @emph{is} valid to mix |
6bf7aab6 DL |
2492 | one of these types of clauses with other clauses like @code{for ... to} |
2493 | or @code{while}. | |
2494 | ||
2495 | @item for @var{var} being the hash-keys of @var{hash-table} | |
2496 | This clause iterates over the entries in @var{hash-table}. For each | |
2497 | hash table entry, @var{var} is bound to the entry's key. If you write | |
2498 | @samp{the hash-values} instead, @var{var} is bound to the values | |
2499 | of the entries. The clause may be followed by the additional | |
2500 | term @samp{using (hash-values @var{var2})} (where @code{hash-values} | |
2501 | is the opposite word of the word following @code{the}) to cause | |
2502 | @var{var} and @var{var2} to be bound to the two parts of each | |
2503 | hash table entry. | |
2504 | ||
2505 | @item for @var{var} being the key-codes of @var{keymap} | |
c6ad01e8 DL |
2506 | This clause iterates over the entries in @var{keymap}. |
2507 | The iteration does not enter nested keymaps or inherited (parent) keymaps. | |
6bf7aab6 DL |
2508 | You can use @samp{the key-bindings} to access the commands bound to |
2509 | the keys rather than the key codes, and you can add a @code{using} | |
2510 | clause to access both the codes and the bindings together. | |
2511 | ||
2512 | @item for @var{var} being the key-seqs of @var{keymap} | |
2513 | This clause iterates over all key sequences defined by @var{keymap} | |
2514 | and its nested keymaps, where @var{var} takes on values which are | |
ed4d9494 | 2515 | vectors. The strings or vectors |
6bf7aab6 DL |
2516 | are reused for each iteration, so you must copy them if you wish to keep |
2517 | them permanently. You can add a @samp{using (key-bindings ...)} | |
2518 | clause to get the command bindings as well. | |
2519 | ||
2520 | @item for @var{var} being the overlays [of @var{buffer}] @dots{} | |
c6ad01e8 DL |
2521 | This clause iterates over the ``overlays'' of a buffer |
2522 | (the clause @code{extents} is synonymous | |
ed4d9494 DL |
2523 | with @code{overlays}). If the @code{of} term is omitted, the current |
2524 | buffer is used. | |
6bf7aab6 DL |
2525 | This clause also accepts optional @samp{from @var{pos}} and |
2526 | @samp{to @var{pos}} terms, limiting the clause to overlays which | |
2527 | overlap the specified region. | |
2528 | ||
2529 | @item for @var{var} being the intervals [of @var{buffer}] @dots{} | |
2530 | This clause iterates over all intervals of a buffer with constant | |
2531 | text properties. The variable @var{var} will be bound to conses | |
2532 | of start and end positions, where one start position is always equal | |
2533 | to the previous end position. The clause allows @code{of}, | |
2534 | @code{from}, @code{to}, and @code{property} terms, where the latter | |
2535 | term restricts the search to just the specified property. The | |
c6ad01e8 | 2536 | @code{of} term may specify either a buffer or a string. |
6bf7aab6 DL |
2537 | |
2538 | @item for @var{var} being the frames | |
2539 | This clause iterates over all frames, i.e., X window system windows | |
c6ad01e8 | 2540 | open on Emacs files. The |
6bf7aab6 DL |
2541 | clause @code{screens} is a synonym for @code{frames}. The frames |
2542 | are visited in @code{next-frame} order starting from | |
2543 | @code{selected-frame}. | |
2544 | ||
2545 | @item for @var{var} being the windows [of @var{frame}] | |
2546 | This clause iterates over the windows (in the Emacs sense) of | |
ed4d9494 | 2547 | the current frame, or of the specified @var{frame}. |
6bf7aab6 DL |
2548 | |
2549 | @item for @var{var} being the buffers | |
2550 | This clause iterates over all buffers in Emacs. It is equivalent | |
2551 | to @samp{for @var{var} in (buffer-list)}. | |
2552 | ||
2553 | @item for @var{var} = @var{expr1} then @var{expr2} | |
2554 | This clause does a general iteration. The first time through | |
2555 | the loop, @var{var} will be bound to @var{expr1}. On the second | |
2556 | and successive iterations it will be set by evaluating @var{expr2} | |
2557 | (which may refer to the old value of @var{var}). For example, | |
2558 | these two loops are effectively the same: | |
2559 | ||
2560 | @example | |
2561 | (loop for x on my-list by 'cddr do ...) | |
2562 | (loop for x = my-list then (cddr x) while x do ...) | |
2563 | @end example | |
2564 | ||
2565 | Note that this type of @code{for} clause does not imply any sort | |
2566 | of terminating condition; the above example combines it with a | |
2567 | @code{while} clause to tell when to end the loop. | |
2568 | ||
2569 | If you omit the @code{then} term, @var{expr1} is used both for | |
2570 | the initial setting and for successive settings: | |
2571 | ||
2572 | @example | |
2573 | (loop for x = (random) when (> x 0) return x) | |
2574 | @end example | |
2575 | ||
2576 | @noindent | |
2577 | This loop keeps taking random numbers from the @code{(random)} | |
2578 | function until it gets a positive one, which it then returns. | |
2579 | @end table | |
2580 | ||
2581 | If you include several @code{for} clauses in a row, they are | |
2582 | treated sequentially (as if by @code{let*} and @code{setq}). | |
2583 | You can instead use the word @code{and} to link the clauses, | |
2584 | in which case they are processed in parallel (as if by @code{let} | |
2585 | and @code{psetq}). | |
2586 | ||
2587 | @example | |
2588 | (loop for x below 5 for y = nil then x collect (list x y)) | |
2589 | @result{} ((0 nil) (1 1) (2 2) (3 3) (4 4)) | |
2590 | (loop for x below 5 and y = nil then x collect (list x y)) | |
2591 | @result{} ((0 nil) (1 0) (2 1) (3 2) (4 3)) | |
2592 | @end example | |
2593 | ||
2594 | @noindent | |
2595 | In the first loop, @code{y} is set based on the value of @code{x} | |
2596 | that was just set by the previous clause; in the second loop, | |
2597 | @code{x} and @code{y} are set simultaneously so @code{y} is set | |
2598 | based on the value of @code{x} left over from the previous time | |
2599 | through the loop. | |
2600 | ||
2601 | Another feature of the @code{loop} macro is @dfn{destructuring}, | |
2602 | similar in concept to the destructuring provided by @code{defmacro}. | |
2603 | The @var{var} part of any @code{for} clause can be given as a list | |
2604 | of variables instead of a single variable. The values produced | |
2605 | during loop execution must be lists; the values in the lists are | |
2606 | stored in the corresponding variables. | |
2607 | ||
2608 | @example | |
2609 | (loop for (x y) in '((2 3) (4 5) (6 7)) collect (+ x y)) | |
2610 | @result{} (5 9 13) | |
2611 | @end example | |
2612 | ||
2613 | In loop destructuring, if there are more values than variables | |
2614 | the trailing values are ignored, and if there are more variables | |
2615 | than values the trailing variables get the value @code{nil}. | |
2616 | If @code{nil} is used as a variable name, the corresponding | |
2617 | values are ignored. Destructuring may be nested, and dotted | |
2618 | lists of variables like @code{(x . y)} are allowed. | |
2619 | ||
2620 | @node Iteration Clauses, Accumulation Clauses, For Clauses, Loop Facility | |
2621 | @subsection Iteration Clauses | |
2622 | ||
2623 | @noindent | |
2624 | Aside from @code{for} clauses, there are several other loop clauses | |
2625 | that control the way the loop operates. They might be used by | |
2626 | themselves, or in conjunction with one or more @code{for} clauses. | |
2627 | ||
2628 | @table @code | |
2629 | @item repeat @var{integer} | |
2630 | This clause simply counts up to the specified number using an | |
2631 | internal temporary variable. The loops | |
2632 | ||
2633 | @example | |
2634 | (loop repeat n do ...) | |
2635 | (loop for temp to n do ...) | |
2636 | @end example | |
2637 | ||
2638 | @noindent | |
2639 | are identical except that the second one forces you to choose | |
2640 | a name for a variable you aren't actually going to use. | |
2641 | ||
2642 | @item while @var{condition} | |
2643 | This clause stops the loop when the specified condition (any Lisp | |
2644 | expression) becomes @code{nil}. For example, the following two | |
2645 | loops are equivalent, except for the implicit @code{nil} block | |
2646 | that surrounds the second one: | |
2647 | ||
2648 | @example | |
2649 | (while @var{cond} @var{forms}@dots{}) | |
2650 | (loop while @var{cond} do @var{forms}@dots{}) | |
2651 | @end example | |
2652 | ||
2653 | @item until @var{condition} | |
2654 | This clause stops the loop when the specified condition is true, | |
2655 | i.e., non-@code{nil}. | |
2656 | ||
2657 | @item always @var{condition} | |
2658 | This clause stops the loop when the specified condition is @code{nil}. | |
2659 | Unlike @code{while}, it stops the loop using @code{return nil} so that | |
2660 | the @code{finally} clauses are not executed. If all the conditions | |
2661 | were non-@code{nil}, the loop returns @code{t}: | |
2662 | ||
2663 | @example | |
2664 | (if (loop for size in size-list always (> size 10)) | |
2665 | (some-big-sizes) | |
2666 | (no-big-sizes)) | |
2667 | @end example | |
2668 | ||
2669 | @item never @var{condition} | |
2670 | This clause is like @code{always}, except that the loop returns | |
2671 | @code{t} if any conditions were false, or @code{nil} otherwise. | |
2672 | ||
2673 | @item thereis @var{condition} | |
2674 | This clause stops the loop when the specified form is non-@code{nil}; | |
2675 | in this case, it returns that non-@code{nil} value. If all the | |
2676 | values were @code{nil}, the loop returns @code{nil}. | |
2677 | @end table | |
2678 | ||
2679 | @node Accumulation Clauses, Other Clauses, Iteration Clauses, Loop Facility | |
2680 | @subsection Accumulation Clauses | |
2681 | ||
2682 | @noindent | |
2683 | These clauses cause the loop to accumulate information about the | |
2684 | specified Lisp @var{form}. The accumulated result is returned | |
2685 | from the loop unless overridden, say, by a @code{return} clause. | |
2686 | ||
2687 | @table @code | |
2688 | @item collect @var{form} | |
2689 | This clause collects the values of @var{form} into a list. Several | |
2690 | examples of @code{collect} appear elsewhere in this manual. | |
2691 | ||
2692 | The word @code{collecting} is a synonym for @code{collect}, and | |
2693 | likewise for the other accumulation clauses. | |
2694 | ||
2695 | @item append @var{form} | |
2696 | This clause collects lists of values into a result list using | |
2697 | @code{append}. | |
2698 | ||
2699 | @item nconc @var{form} | |
2700 | This clause collects lists of values into a result list by | |
2701 | destructively modifying the lists rather than copying them. | |
2702 | ||
2703 | @item concat @var{form} | |
2704 | This clause concatenates the values of the specified @var{form} | |
2705 | into a string. (It and the following clause are extensions to | |
2706 | standard Common Lisp.) | |
2707 | ||
2708 | @item vconcat @var{form} | |
2709 | This clause concatenates the values of the specified @var{form} | |
2710 | into a vector. | |
2711 | ||
2712 | @item count @var{form} | |
2713 | This clause counts the number of times the specified @var{form} | |
2714 | evaluates to a non-@code{nil} value. | |
2715 | ||
2716 | @item sum @var{form} | |
2717 | This clause accumulates the sum of the values of the specified | |
2718 | @var{form}, which must evaluate to a number. | |
2719 | ||
2720 | @item maximize @var{form} | |
2721 | This clause accumulates the maximum value of the specified @var{form}, | |
2722 | which must evaluate to a number. The return value is undefined if | |
2723 | @code{maximize} is executed zero times. | |
2724 | ||
2725 | @item minimize @var{form} | |
2726 | This clause accumulates the minimum value of the specified @var{form}. | |
2727 | @end table | |
2728 | ||
2729 | Accumulation clauses can be followed by @samp{into @var{var}} to | |
2730 | cause the data to be collected into variable @var{var} (which is | |
2731 | automatically @code{let}-bound during the loop) rather than an | |
2732 | unnamed temporary variable. Also, @code{into} accumulations do | |
2733 | not automatically imply a return value. The loop must use some | |
2734 | explicit mechanism, such as @code{finally return}, to return | |
2735 | the accumulated result. | |
2736 | ||
2cbd16b9 | 2737 | It is valid for several accumulation clauses of the same type to |
6bf7aab6 DL |
2738 | accumulate into the same place. From Steele: |
2739 | ||
2740 | @example | |
2741 | (loop for name in '(fred sue alice joe june) | |
2742 | for kids in '((bob ken) () () (kris sunshine) ()) | |
2743 | collect name | |
2744 | append kids) | |
2745 | @result{} (fred bob ken sue alice joe kris sunshine june) | |
2746 | @end example | |
2747 | ||
c6ad01e8 | 2748 | @node Other Clauses, , Accumulation Clauses, Loop Facility |
6bf7aab6 DL |
2749 | @subsection Other Clauses |
2750 | ||
2751 | @noindent | |
2752 | This section describes the remaining loop clauses. | |
2753 | ||
2754 | @table @code | |
2755 | @item with @var{var} = @var{value} | |
2756 | This clause binds a variable to a value around the loop, but | |
2757 | otherwise leaves the variable alone during the loop. The following | |
2758 | loops are basically equivalent: | |
2759 | ||
2760 | @example | |
2761 | (loop with x = 17 do ...) | |
2762 | (let ((x 17)) (loop do ...)) | |
2763 | (loop for x = 17 then x do ...) | |
2764 | @end example | |
2765 | ||
2766 | Naturally, the variable @var{var} might be used for some purpose | |
2767 | in the rest of the loop. For example: | |
2768 | ||
2769 | @example | |
2770 | (loop for x in my-list with res = nil do (push x res) | |
2771 | finally return res) | |
2772 | @end example | |
2773 | ||
2774 | This loop inserts the elements of @code{my-list} at the front of | |
2775 | a new list being accumulated in @code{res}, then returns the | |
2776 | list @code{res} at the end of the loop. The effect is similar | |
2777 | to that of a @code{collect} clause, but the list gets reversed | |
2778 | by virtue of the fact that elements are being pushed onto the | |
2779 | front of @code{res} rather than the end. | |
2780 | ||
2781 | If you omit the @code{=} term, the variable is initialized to | |
2782 | @code{nil}. (Thus the @samp{= nil} in the above example is | |
2783 | unnecessary.) | |
2784 | ||
2785 | Bindings made by @code{with} are sequential by default, as if | |
2786 | by @code{let*}. Just like @code{for} clauses, @code{with} clauses | |
2787 | can be linked with @code{and} to cause the bindings to be made by | |
2788 | @code{let} instead. | |
2789 | ||
2790 | @item if @var{condition} @var{clause} | |
2791 | This clause executes the following loop clause only if the specified | |
2792 | condition is true. The following @var{clause} should be an accumulation, | |
2793 | @code{do}, @code{return}, @code{if}, or @code{unless} clause. | |
2794 | Several clauses may be linked by separating them with @code{and}. | |
2795 | These clauses may be followed by @code{else} and a clause or clauses | |
2796 | to execute if the condition was false. The whole construct may | |
2797 | optionally be followed by the word @code{end} (which may be used to | |
2798 | disambiguate an @code{else} or @code{and} in a nested @code{if}). | |
2799 | ||
2800 | The actual non-@code{nil} value of the condition form is available | |
2801 | by the name @code{it} in the ``then'' part. For example: | |
2802 | ||
2803 | @example | |
2804 | (setq funny-numbers '(6 13 -1)) | |
2805 | @result{} (6 13 -1) | |
2806 | (loop for x below 10 | |
2807 | if (oddp x) | |
2808 | collect x into odds | |
2809 | and if (memq x funny-numbers) return (cdr it) end | |
2810 | else | |
2811 | collect x into evens | |
2812 | finally return (vector odds evens)) | |
2813 | @result{} [(1 3 5 7 9) (0 2 4 6 8)] | |
2814 | (setq funny-numbers '(6 7 13 -1)) | |
2815 | @result{} (6 7 13 -1) | |
2816 | (loop <@r{same thing again}>) | |
2817 | @result{} (13 -1) | |
2818 | @end example | |
2819 | ||
2820 | Note the use of @code{and} to put two clauses into the ``then'' | |
2821 | part, one of which is itself an @code{if} clause. Note also that | |
2822 | @code{end}, while normally optional, was necessary here to make | |
2823 | it clear that the @code{else} refers to the outermost @code{if} | |
2824 | clause. In the first case, the loop returns a vector of lists | |
2825 | of the odd and even values of @var{x}. In the second case, the | |
2826 | odd number 7 is one of the @code{funny-numbers} so the loop | |
2827 | returns early; the actual returned value is based on the result | |
2828 | of the @code{memq} call. | |
2829 | ||
2830 | @item when @var{condition} @var{clause} | |
2831 | This clause is just a synonym for @code{if}. | |
2832 | ||
2833 | @item unless @var{condition} @var{clause} | |
2834 | The @code{unless} clause is just like @code{if} except that the | |
2835 | sense of the condition is reversed. | |
2836 | ||
2837 | @item named @var{name} | |
2838 | This clause gives a name other than @code{nil} to the implicit | |
2839 | block surrounding the loop. The @var{name} is the symbol to be | |
2840 | used as the block name. | |
2841 | ||
2842 | @item initially [do] @var{forms}... | |
2843 | This keyword introduces one or more Lisp forms which will be | |
2844 | executed before the loop itself begins (but after any variables | |
2845 | requested by @code{for} or @code{with} have been bound to their | |
2846 | initial values). @code{initially} clauses can appear anywhere; | |
2847 | if there are several, they are executed in the order they appear | |
2848 | in the loop. The keyword @code{do} is optional. | |
2849 | ||
2850 | @item finally [do] @var{forms}... | |
2851 | This introduces Lisp forms which will be executed after the loop | |
2852 | finishes (say, on request of a @code{for} or @code{while}). | |
2853 | @code{initially} and @code{finally} clauses may appear anywhere | |
2854 | in the loop construct, but they are executed (in the specified | |
2855 | order) at the beginning or end, respectively, of the loop. | |
2856 | ||
2857 | @item finally return @var{form} | |
2858 | This says that @var{form} should be executed after the loop | |
2859 | is done to obtain a return value. (Without this, or some other | |
2860 | clause like @code{collect} or @code{return}, the loop will simply | |
2861 | return @code{nil}.) Variables bound by @code{for}, @code{with}, | |
2862 | or @code{into} will still contain their final values when @var{form} | |
2863 | is executed. | |
2864 | ||
2865 | @item do @var{forms}... | |
2866 | The word @code{do} may be followed by any number of Lisp expressions | |
2867 | which are executed as an implicit @code{progn} in the body of the | |
2868 | loop. Many of the examples in this section illustrate the use of | |
2869 | @code{do}. | |
2870 | ||
2871 | @item return @var{form} | |
2872 | This clause causes the loop to return immediately. The following | |
2873 | Lisp form is evaluated to give the return value of the @code{loop} | |
2874 | form. The @code{finally} clauses, if any, are not executed. | |
2875 | Of course, @code{return} is generally used inside an @code{if} or | |
2876 | @code{unless}, as its use in a top-level loop clause would mean | |
2877 | the loop would never get to ``loop'' more than once. | |
2878 | ||
2879 | The clause @samp{return @var{form}} is equivalent to | |
2880 | @samp{do (return @var{form})} (or @code{return-from} if the loop | |
2881 | was named). The @code{return} clause is implemented a bit more | |
2882 | efficiently, though. | |
2883 | @end table | |
2884 | ||
2885 | While there is no high-level way to add user extensions to @code{loop} | |
2886 | (comparable to @code{defsetf} for @code{setf}, say), this package | |
2887 | does offer two properties called @code{cl-loop-handler} and | |
2888 | @code{cl-loop-for-handler} which are functions to be called when | |
2889 | a given symbol is encountered as a top-level loop clause or | |
2890 | @code{for} clause, respectively. Consult the source code in | |
2891 | file @file{cl-macs.el} for details. | |
2892 | ||
2893 | This package's @code{loop} macro is compatible with that of Common | |
2894 | Lisp, except that a few features are not implemented: @code{loop-finish} | |
2895 | and data-type specifiers. Naturally, the @code{for} clauses which | |
2896 | iterate over keymaps, overlays, intervals, frames, windows, and | |
2897 | buffers are Emacs-specific extensions. | |
2898 | ||
c6ad01e8 | 2899 | @node Multiple Values, , Loop Facility, Control Structure |
6bf7aab6 DL |
2900 | @section Multiple Values |
2901 | ||
2902 | @noindent | |
2903 | Common Lisp functions can return zero or more results. Emacs Lisp | |
2904 | functions, by contrast, always return exactly one result. This | |
2905 | package makes no attempt to emulate Common Lisp multiple return | |
2906 | values; Emacs versions of Common Lisp functions that return more | |
2907 | than one value either return just the first value (as in | |
2908 | @code{compiler-macroexpand}) or return a list of values (as in | |
2909 | @code{get-setf-method}). This package @emph{does} define placeholders | |
2910 | for the Common Lisp functions that work with multiple values, but | |
2911 | in Emacs Lisp these functions simply operate on lists instead. | |
2912 | The @code{values} form, for example, is a synonym for @code{list} | |
2913 | in Emacs. | |
2914 | ||
2915 | @defspec multiple-value-bind (var@dots{}) values-form forms@dots{} | |
2916 | This form evaluates @var{values-form}, which must return a list of | |
2917 | values. It then binds the @var{var}s to these respective values, | |
2918 | as if by @code{let}, and then executes the body @var{forms}. | |
2919 | If there are more @var{var}s than values, the extra @var{var}s | |
2920 | are bound to @code{nil}. If there are fewer @var{var}s than | |
2921 | values, the excess values are ignored. | |
2922 | @end defspec | |
2923 | ||
2924 | @defspec multiple-value-setq (var@dots{}) form | |
2925 | This form evaluates @var{form}, which must return a list of values. | |
2926 | It then sets the @var{var}s to these respective values, as if by | |
2927 | @code{setq}. Extra @var{var}s or values are treated the same as | |
2928 | in @code{multiple-value-bind}. | |
2929 | @end defspec | |
2930 | ||
2931 | The older Quiroz package attempted a more faithful (but still | |
2932 | imperfect) emulation of Common Lisp multiple values. The old | |
2933 | method ``usually'' simulated true multiple values quite well, | |
2934 | but under certain circumstances would leave spurious return | |
2935 | values in memory where a later, unrelated @code{multiple-value-bind} | |
2936 | form would see them. | |
2937 | ||
2938 | Since a perfect emulation is not feasible in Emacs Lisp, this | |
2939 | package opts to keep it as simple and predictable as possible. | |
2940 | ||
2941 | @node Macros, Declarations, Control Structure, Top | |
2942 | @chapter Macros | |
2943 | ||
2944 | @noindent | |
2945 | This package implements the various Common Lisp features of | |
2946 | @code{defmacro}, such as destructuring, @code{&environment}, | |
2947 | and @code{&body}. Top-level @code{&whole} is not implemented | |
2948 | for @code{defmacro} due to technical difficulties. | |
2949 | @xref{Argument Lists}. | |
2950 | ||
2951 | Destructuring is made available to the user by way of the | |
2952 | following macro: | |
2953 | ||
2954 | @defspec destructuring-bind arglist expr forms@dots{} | |
2955 | This macro expands to code which executes @var{forms}, with | |
2956 | the variables in @var{arglist} bound to the list of values | |
2957 | returned by @var{expr}. The @var{arglist} can include all | |
2958 | the features allowed for @code{defmacro} argument lists, | |
2959 | including destructuring. (The @code{&environment} keyword | |
2960 | is not allowed.) The macro expansion will signal an error | |
2961 | if @var{expr} returns a list of the wrong number of arguments | |
2962 | or with incorrect keyword arguments. | |
2963 | @end defspec | |
2964 | ||
2965 | This package also includes the Common Lisp @code{define-compiler-macro} | |
2966 | facility, which allows you to define compile-time expansions and | |
2967 | optimizations for your functions. | |
2968 | ||
2969 | @defspec define-compiler-macro name arglist forms@dots{} | |
2970 | This form is similar to @code{defmacro}, except that it only expands | |
2971 | calls to @var{name} at compile-time; calls processed by the Lisp | |
2972 | interpreter are not expanded, nor are they expanded by the | |
2973 | @code{macroexpand} function. | |
2974 | ||
2975 | The argument list may begin with a @code{&whole} keyword and a | |
2976 | variable. This variable is bound to the macro-call form itself, | |
2977 | i.e., to a list of the form @samp{(@var{name} @var{args}@dots{})}. | |
2978 | If the macro expander returns this form unchanged, then the | |
2979 | compiler treats it as a normal function call. This allows | |
2980 | compiler macros to work as optimizers for special cases of a | |
2981 | function, leaving complicated cases alone. | |
2982 | ||
2983 | For example, here is a simplified version of a definition that | |
2984 | appears as a standard part of this package: | |
2985 | ||
2986 | @example | |
2987 | (define-compiler-macro member* (&whole form a list &rest keys) | |
2988 | (if (and (null keys) | |
2989 | (eq (car-safe a) 'quote) | |
2990 | (not (floatp-safe (cadr a)))) | |
2991 | (list 'memq a list) | |
2992 | form)) | |
2993 | @end example | |
2994 | ||
2995 | @noindent | |
2996 | This definition causes @code{(member* @var{a} @var{list})} to change | |
2997 | to a call to the faster @code{memq} in the common case where @var{a} | |
2998 | is a non-floating-point constant; if @var{a} is anything else, or | |
2999 | if there are any keyword arguments in the call, then the original | |
3000 | @code{member*} call is left intact. (The actual compiler macro | |
3001 | for @code{member*} optimizes a number of other cases, including | |
3002 | common @code{:test} predicates.) | |
3003 | @end defspec | |
3004 | ||
3005 | @defun compiler-macroexpand form | |
3006 | This function is analogous to @code{macroexpand}, except that it | |
3007 | expands compiler macros rather than regular macros. It returns | |
3008 | @var{form} unchanged if it is not a call to a function for which | |
3009 | a compiler macro has been defined, or if that compiler macro | |
3010 | decided to punt by returning its @code{&whole} argument. Like | |
3011 | @code{macroexpand}, it expands repeatedly until it reaches a form | |
3012 | for which no further expansion is possible. | |
3013 | @end defun | |
3014 | ||
3015 | @xref{Macro Bindings}, for descriptions of the @code{macrolet} | |
3016 | and @code{symbol-macrolet} forms for making ``local'' macro | |
3017 | definitions. | |
3018 | ||
3019 | @node Declarations, Symbols, Macros, Top | |
3020 | @chapter Declarations | |
3021 | ||
3022 | @noindent | |
3023 | Common Lisp includes a complex and powerful ``declaration'' | |
3024 | mechanism that allows you to give the compiler special hints | |
3025 | about the types of data that will be stored in particular variables, | |
3026 | and about the ways those variables and functions will be used. This | |
3027 | package defines versions of all the Common Lisp declaration forms: | |
3028 | @code{declare}, @code{locally}, @code{proclaim}, @code{declaim}, | |
3029 | and @code{the}. | |
3030 | ||
3031 | Most of the Common Lisp declarations are not currently useful in | |
3032 | Emacs Lisp, as the byte-code system provides little opportunity | |
3033 | to benefit from type information, and @code{special} declarations | |
3034 | are redundant in a fully dynamically-scoped Lisp. A few | |
c6ad01e8 | 3035 | declarations are meaningful when the optimizing byte |
6bf7aab6 DL |
3036 | compiler is being used, however. Under the earlier non-optimizing |
3037 | compiler, these declarations will effectively be ignored. | |
3038 | ||
3039 | @defun proclaim decl-spec | |
3040 | This function records a ``global'' declaration specified by | |
3041 | @var{decl-spec}. Since @code{proclaim} is a function, @var{decl-spec} | |
3042 | is evaluated and thus should normally be quoted. | |
3043 | @end defun | |
3044 | ||
3045 | @defspec declaim decl-specs@dots{} | |
3046 | This macro is like @code{proclaim}, except that it takes any number | |
3047 | of @var{decl-spec} arguments, and the arguments are unevaluated and | |
3048 | unquoted. The @code{declaim} macro also puts an @code{(eval-when | |
3049 | (compile load eval) ...)} around the declarations so that they will | |
3050 | be registered at compile-time as well as at run-time. (This is vital, | |
3051 | since normally the declarations are meant to influence the way the | |
3052 | compiler treats the rest of the file that contains the @code{declaim} | |
3053 | form.) | |
3054 | @end defspec | |
3055 | ||
3056 | @defspec declare decl-specs@dots{} | |
3057 | This macro is used to make declarations within functions and other | |
3058 | code. Common Lisp allows declarations in various locations, generally | |
3059 | at the beginning of any of the many ``implicit @code{progn}s'' | |
3060 | throughout Lisp syntax, such as function bodies, @code{let} bodies, | |
3061 | etc. Currently the only declaration understood by @code{declare} | |
3062 | is @code{special}. | |
3063 | @end defspec | |
3064 | ||
3065 | @defspec locally declarations@dots{} forms@dots{} | |
3066 | In this package, @code{locally} is no different from @code{progn}. | |
3067 | @end defspec | |
3068 | ||
3069 | @defspec the type form | |
3070 | Type information provided by @code{the} is ignored in this package; | |
3071 | in other words, @code{(the @var{type} @var{form})} is equivalent | |
3072 | to @var{form}. Future versions of the optimizing byte-compiler may | |
3073 | make use of this information. | |
3074 | ||
3075 | For example, @code{mapcar} can map over both lists and arrays. It is | |
3076 | hard for the compiler to expand @code{mapcar} into an in-line loop | |
3077 | unless it knows whether the sequence will be a list or an array ahead | |
3078 | of time. With @code{(mapcar 'car (the vector foo))}, a future | |
3079 | compiler would have enough information to expand the loop in-line. | |
3080 | For now, Emacs Lisp will treat the above code as exactly equivalent | |
3081 | to @code{(mapcar 'car foo)}. | |
3082 | @end defspec | |
3083 | ||
3084 | Each @var{decl-spec} in a @code{proclaim}, @code{declaim}, or | |
3085 | @code{declare} should be a list beginning with a symbol that says | |
3086 | what kind of declaration it is. This package currently understands | |
3087 | @code{special}, @code{inline}, @code{notinline}, @code{optimize}, | |
3088 | and @code{warn} declarations. (The @code{warn} declaration is an | |
3089 | extension of standard Common Lisp.) Other Common Lisp declarations, | |
3090 | such as @code{type} and @code{ftype}, are silently ignored. | |
3091 | ||
3092 | @table @code | |
3093 | @item special | |
3094 | Since all variables in Emacs Lisp are ``special'' (in the Common | |
3095 | Lisp sense), @code{special} declarations are only advisory. They | |
3096 | simply tell the optimizing byte compiler that the specified | |
3097 | variables are intentionally being referred to without being | |
3098 | bound in the body of the function. The compiler normally emits | |
3099 | warnings for such references, since they could be typographical | |
3100 | errors for references to local variables. | |
3101 | ||
3102 | The declaration @code{(declare (special @var{var1} @var{var2}))} is | |
3103 | equivalent to @code{(defvar @var{var1}) (defvar @var{var2})} in the | |
3104 | optimizing compiler, or to nothing at all in older compilers (which | |
3105 | do not warn for non-local references). | |
3106 | ||
3107 | In top-level contexts, it is generally better to write | |
3108 | @code{(defvar @var{var})} than @code{(declaim (special @var{var}))}, | |
3109 | since @code{defvar} makes your intentions clearer. But the older | |
3110 | byte compilers can not handle @code{defvar}s appearing inside of | |
3111 | functions, while @code{(declare (special @var{var}))} takes care | |
3112 | to work correctly with all compilers. | |
3113 | ||
3114 | @item inline | |
3115 | The @code{inline} @var{decl-spec} lists one or more functions | |
3116 | whose bodies should be expanded ``in-line'' into calling functions | |
3117 | whenever the compiler is able to arrange for it. For example, | |
3118 | the Common Lisp function @code{cadr} is declared @code{inline} | |
3119 | by this package so that the form @code{(cadr @var{x})} will | |
3120 | expand directly into @code{(car (cdr @var{x}))} when it is called | |
3121 | in user functions, for a savings of one (relatively expensive) | |
3122 | function call. | |
3123 | ||
3124 | The following declarations are all equivalent. Note that the | |
3125 | @code{defsubst} form is a convenient way to define a function | |
c6ad01e8 | 3126 | and declare it inline all at once. |
6bf7aab6 DL |
3127 | |
3128 | @example | |
3129 | (declaim (inline foo bar)) | |
3130 | (eval-when (compile load eval) (proclaim '(inline foo bar))) | |
c6ad01e8 | 3131 | (defsubst foo (...) ...) ; instead of defun |
6bf7aab6 DL |
3132 | @end example |
3133 | ||
bab5541b | 3134 | @strong{Please note:} this declaration remains in effect after the |
6bf7aab6 DL |
3135 | containing source file is done. It is correct to use it to |
3136 | request that a function you have defined should be inlined, | |
3137 | but it is impolite to use it to request inlining of an external | |
3138 | function. | |
3139 | ||
3140 | In Common Lisp, it is possible to use @code{(declare (inline @dots{}))} | |
3141 | before a particular call to a function to cause just that call to | |
3142 | be inlined; the current byte compilers provide no way to implement | |
3143 | this, so @code{(declare (inline @dots{}))} is currently ignored by | |
3144 | this package. | |
3145 | ||
3146 | @item notinline | |
3147 | The @code{notinline} declaration lists functions which should | |
3148 | not be inlined after all; it cancels a previous @code{inline} | |
3149 | declaration. | |
3150 | ||
3151 | @item optimize | |
3152 | This declaration controls how much optimization is performed by | |
3153 | the compiler. Naturally, it is ignored by the earlier non-optimizing | |
3154 | compilers. | |
3155 | ||
3156 | The word @code{optimize} is followed by any number of lists like | |
3157 | @code{(speed 3)} or @code{(safety 2)}. Common Lisp defines several | |
3158 | optimization ``qualities''; this package ignores all but @code{speed} | |
3159 | and @code{safety}. The value of a quality should be an integer from | |
3160 | 0 to 3, with 0 meaning ``unimportant'' and 3 meaning ``very important.'' | |
3161 | The default level for both qualities is 1. | |
3162 | ||
c6ad01e8 | 3163 | In this package, with the optimizing compiler, the |
6bf7aab6 DL |
3164 | @code{speed} quality is tied to the @code{byte-compile-optimize} |
3165 | flag, which is set to @code{nil} for @code{(speed 0)} and to | |
3166 | @code{t} for higher settings; and the @code{safety} quality is | |
3167 | tied to the @code{byte-compile-delete-errors} flag, which is | |
3168 | set to @code{t} for @code{(safety 3)} and to @code{nil} for all | |
3169 | lower settings. (The latter flag controls whether the compiler | |
3170 | is allowed to optimize out code whose only side-effect could | |
3171 | be to signal an error, e.g., rewriting @code{(progn foo bar)} to | |
3172 | @code{bar} when it is not known whether @code{foo} will be bound | |
3173 | at run-time.) | |
3174 | ||
3175 | Note that even compiling with @code{(safety 0)}, the Emacs | |
3176 | byte-code system provides sufficient checking to prevent real | |
3177 | harm from being done. For example, barring serious bugs in | |
3178 | Emacs itself, Emacs will not crash with a segmentation fault | |
3179 | just because of an error in a fully-optimized Lisp program. | |
3180 | ||
3181 | The @code{optimize} declaration is normally used in a top-level | |
3182 | @code{proclaim} or @code{declaim} in a file; Common Lisp allows | |
3183 | it to be used with @code{declare} to set the level of optimization | |
3184 | locally for a given form, but this will not work correctly with the | |
3185 | current version of the optimizing compiler. (The @code{declare} | |
3186 | will set the new optimization level, but that level will not | |
3187 | automatically be unset after the enclosing form is done.) | |
3188 | ||
3189 | @item warn | |
3190 | This declaration controls what sorts of warnings are generated | |
3191 | by the byte compiler. Again, only the optimizing compiler | |
3192 | generates warnings. The word @code{warn} is followed by any | |
3193 | number of ``warning qualities,'' similar in form to optimization | |
3194 | qualities. The currently supported warning types are | |
3195 | @code{redefine}, @code{callargs}, @code{unresolved}, and | |
3196 | @code{free-vars}; in the current system, a value of 0 will | |
3197 | disable these warnings and any higher value will enable them. | |
3198 | See the documentation for the optimizing byte compiler for details. | |
3199 | @end table | |
3200 | ||
3201 | @node Symbols, Numbers, Declarations, Top | |
3202 | @chapter Symbols | |
3203 | ||
3204 | @noindent | |
3205 | This package defines several symbol-related features that were | |
3206 | missing from Emacs Lisp. | |
3207 | ||
3208 | @menu | |
3209 | * Property Lists:: `get*', `remprop', `getf', `remf' | |
3210 | * Creating Symbols:: `gensym', `gentemp' | |
3211 | @end menu | |
3212 | ||
3213 | @node Property Lists, Creating Symbols, Symbols, Symbols | |
3214 | @section Property Lists | |
3215 | ||
3216 | @noindent | |
3217 | These functions augment the standard Emacs Lisp functions @code{get} | |
3218 | and @code{put} for operating on properties attached to symbols. | |
3219 | There are also functions for working with property lists as | |
3220 | first-class data structures not attached to particular symbols. | |
3221 | ||
3222 | @defun get* symbol property &optional default | |
3223 | This function is like @code{get}, except that if the property is | |
3224 | not found, the @var{default} argument provides the return value. | |
3225 | (The Emacs Lisp @code{get} function always uses @code{nil} as | |
3226 | the default; this package's @code{get*} is equivalent to Common | |
3227 | Lisp's @code{get}.) | |
3228 | ||
3229 | The @code{get*} function is @code{setf}-able; when used in this | |
3230 | fashion, the @var{default} argument is allowed but ignored. | |
3231 | @end defun | |
3232 | ||
3233 | @defun remprop symbol property | |
3234 | This function removes the entry for @var{property} from the property | |
3235 | list of @var{symbol}. It returns a true value if the property was | |
3236 | indeed found and removed, or @code{nil} if there was no such property. | |
3237 | (This function was probably omitted from Emacs originally because, | |
3238 | since @code{get} did not allow a @var{default}, it was very difficult | |
3239 | to distinguish between a missing property and a property whose value | |
3240 | was @code{nil}; thus, setting a property to @code{nil} was close | |
3241 | enough to @code{remprop} for most purposes.) | |
3242 | @end defun | |
3243 | ||
3244 | @defun getf place property &optional default | |
3245 | This function scans the list @var{place} as if it were a property | |
3246 | list, i.e., a list of alternating property names and values. If | |
3247 | an even-numbered element of @var{place} is found which is @code{eq} | |
3248 | to @var{property}, the following odd-numbered element is returned. | |
3249 | Otherwise, @var{default} is returned (or @code{nil} if no default | |
3250 | is given). | |
3251 | ||
3252 | In particular, | |
3253 | ||
3254 | @example | |
3255 | (get sym prop) @equiv{} (getf (symbol-plist sym) prop) | |
3256 | @end example | |
3257 | ||
2cbd16b9 WL |
3258 | It is valid to use @code{getf} as a @code{setf} place, in which case |
3259 | its @var{place} argument must itself be a valid @code{setf} place. | |
6bf7aab6 DL |
3260 | The @var{default} argument, if any, is ignored in this context. |
3261 | The effect is to change (via @code{setcar}) the value cell in the | |
3262 | list that corresponds to @var{property}, or to cons a new property-value | |
3263 | pair onto the list if the property is not yet present. | |
3264 | ||
3265 | @example | |
3266 | (put sym prop val) @equiv{} (setf (getf (symbol-plist sym) prop) val) | |
3267 | @end example | |
3268 | ||
3269 | The @code{get} and @code{get*} functions are also @code{setf}-able. | |
3270 | The fact that @code{default} is ignored can sometimes be useful: | |
3271 | ||
3272 | @example | |
3273 | (incf (get* 'foo 'usage-count 0)) | |
3274 | @end example | |
3275 | ||
3276 | Here, symbol @code{foo}'s @code{usage-count} property is incremented | |
3277 | if it exists, or set to 1 (an incremented 0) otherwise. | |
3278 | ||
3279 | When not used as a @code{setf} form, @code{getf} is just a regular | |
3280 | function and its @var{place} argument can actually be any Lisp | |
3281 | expression. | |
3282 | @end defun | |
3283 | ||
3284 | @defspec remf place property | |
3285 | This macro removes the property-value pair for @var{property} from | |
3286 | the property list stored at @var{place}, which is any @code{setf}-able | |
3287 | place expression. It returns true if the property was found. Note | |
3288 | that if @var{property} happens to be first on the list, this will | |
3289 | effectively do a @code{(setf @var{place} (cddr @var{place}))}, | |
3290 | whereas if it occurs later, this simply uses @code{setcdr} to splice | |
3291 | out the property and value cells. | |
3292 | @end defspec | |
3293 | ||
3294 | @iftex | |
3295 | @secno=2 | |
3296 | @end iftex | |
3297 | ||
c6ad01e8 | 3298 | @node Creating Symbols, , Property Lists, Symbols |
6bf7aab6 DL |
3299 | @section Creating Symbols |
3300 | ||
3301 | @noindent | |
3302 | These functions create unique symbols, typically for use as | |
3303 | temporary variables. | |
3304 | ||
3305 | @defun gensym &optional x | |
3306 | This function creates a new, uninterned symbol (using @code{make-symbol}) | |
3307 | with a unique name. (The name of an uninterned symbol is relevant | |
3308 | only if the symbol is printed.) By default, the name is generated | |
3309 | from an increasing sequence of numbers, @samp{G1000}, @samp{G1001}, | |
3310 | @samp{G1002}, etc. If the optional argument @var{x} is a string, that | |
3311 | string is used as a prefix instead of @samp{G}. Uninterned symbols | |
3312 | are used in macro expansions for temporary variables, to ensure that | |
3313 | their names will not conflict with ``real'' variables in the user's | |
3314 | code. | |
3315 | @end defun | |
3316 | ||
3317 | @defvar *gensym-counter* | |
3318 | This variable holds the counter used to generate @code{gensym} names. | |
3319 | It is incremented after each use by @code{gensym}. In Common Lisp | |
3320 | this is initialized with 0, but this package initializes it with a | |
3321 | random (time-dependent) value to avoid trouble when two files that | |
3322 | each used @code{gensym} in their compilation are loaded together. | |
3323 | (Uninterned symbols become interned when the compiler writes them | |
3324 | out to a file and the Emacs loader loads them, so their names have to | |
3325 | be treated a bit more carefully than in Common Lisp where uninterned | |
3326 | symbols remain uninterned after loading.) | |
3327 | @end defvar | |
3328 | ||
3329 | @defun gentemp &optional x | |
3330 | This function is like @code{gensym}, except that it produces a new | |
3331 | @emph{interned} symbol. If the symbol that is generated already | |
3332 | exists, the function keeps incrementing the counter and trying | |
3333 | again until a new symbol is generated. | |
3334 | @end defun | |
3335 | ||
3336 | The Quiroz @file{cl.el} package also defined a @code{defkeyword} | |
3337 | form for creating self-quoting keyword symbols. This package | |
3338 | automatically creates all keywords that are called for by | |
3339 | @code{&key} argument specifiers, and discourages the use of | |
3340 | keywords as data unrelated to keyword arguments, so the | |
3341 | @code{defkeyword} form has been discontinued. | |
3342 | ||
3343 | @iftex | |
3344 | @chapno=11 | |
3345 | @end iftex | |
3346 | ||
3347 | @node Numbers, Sequences, Symbols, Top | |
3348 | @chapter Numbers | |
3349 | ||
3350 | @noindent | |
3351 | This section defines a few simple Common Lisp operations on numbers | |
3352 | which were left out of Emacs Lisp. | |
3353 | ||
3354 | @menu | |
3355 | * Predicates on Numbers:: `plusp', `oddp', `floatp-safe', etc. | |
76162e12 | 3356 | * Numerical Functions:: `abs', `floor*', etc. |
6bf7aab6 | 3357 | * Random Numbers:: `random*', `make-random-state' |
3e16e035 | 3358 | * Implementation Parameters:: `most-positive-float' |
6bf7aab6 DL |
3359 | @end menu |
3360 | ||
3361 | @iftex | |
3362 | @secno=1 | |
3363 | @end iftex | |
3364 | ||
3365 | @node Predicates on Numbers, Numerical Functions, Numbers, Numbers | |
3366 | @section Predicates on Numbers | |
3367 | ||
3368 | @noindent | |
3369 | These functions return @code{t} if the specified condition is | |
3370 | true of the numerical argument, or @code{nil} otherwise. | |
3371 | ||
3372 | @defun plusp number | |
3373 | This predicate tests whether @var{number} is positive. It is an | |
3374 | error if the argument is not a number. | |
3375 | @end defun | |
3376 | ||
3377 | @defun minusp number | |
3378 | This predicate tests whether @var{number} is negative. It is an | |
3379 | error if the argument is not a number. | |
3380 | @end defun | |
3381 | ||
3382 | @defun oddp integer | |
3383 | This predicate tests whether @var{integer} is odd. It is an | |
3384 | error if the argument is not an integer. | |
3385 | @end defun | |
3386 | ||
3387 | @defun evenp integer | |
3388 | This predicate tests whether @var{integer} is even. It is an | |
3389 | error if the argument is not an integer. | |
3390 | @end defun | |
3391 | ||
3392 | @defun floatp-safe object | |
3393 | This predicate tests whether @var{object} is a floating-point | |
3394 | number. On systems that support floating-point, this is equivalent | |
3395 | to @code{floatp}. On other systems, this always returns @code{nil}. | |
3396 | @end defun | |
3397 | ||
3398 | @iftex | |
3399 | @secno=3 | |
3400 | @end iftex | |
3401 | ||
3402 | @node Numerical Functions, Random Numbers, Predicates on Numbers, Numbers | |
3403 | @section Numerical Functions | |
3404 | ||
3405 | @noindent | |
3406 | These functions perform various arithmetic operations on numbers. | |
3407 | ||
6bf7aab6 DL |
3408 | @defun gcd &rest integers |
3409 | This function returns the Greatest Common Divisor of the arguments. | |
3410 | For one argument, it returns the absolute value of that argument. | |
3411 | For zero arguments, it returns zero. | |
3412 | @end defun | |
3413 | ||
3414 | @defun lcm &rest integers | |
3415 | This function returns the Least Common Multiple of the arguments. | |
3416 | For one argument, it returns the absolute value of that argument. | |
3417 | For zero arguments, it returns one. | |
3418 | @end defun | |
3419 | ||
3420 | @defun isqrt integer | |
3421 | This function computes the ``integer square root'' of its integer | |
3422 | argument, i.e., the greatest integer less than or equal to the true | |
3423 | square root of the argument. | |
3424 | @end defun | |
3425 | ||
3426 | @defun floor* number &optional divisor | |
3427 | This function implements the Common Lisp @code{floor} function. | |
3428 | It is called @code{floor*} to avoid name conflicts with the | |
c6ad01e8 | 3429 | simpler @code{floor} function built-in to Emacs. |
6bf7aab6 DL |
3430 | |
3431 | With one argument, @code{floor*} returns a list of two numbers: | |
3432 | The argument rounded down (toward minus infinity) to an integer, | |
3433 | and the ``remainder'' which would have to be added back to the | |
3434 | first return value to yield the argument again. If the argument | |
3435 | is an integer @var{x}, the result is always the list @code{(@var{x} 0)}. | |
c6ad01e8 | 3436 | If the argument is a floating-point number, the first |
6bf7aab6 DL |
3437 | result is a Lisp integer and the second is a Lisp float between |
3438 | 0 (inclusive) and 1 (exclusive). | |
3439 | ||
3440 | With two arguments, @code{floor*} divides @var{number} by | |
3441 | @var{divisor}, and returns the floor of the quotient and the | |
3442 | corresponding remainder as a list of two numbers. If | |
3443 | @code{(floor* @var{x} @var{y})} returns @code{(@var{q} @var{r})}, | |
3444 | then @code{@var{q}*@var{y} + @var{r} = @var{x}}, with @var{r} | |
3445 | between 0 (inclusive) and @var{r} (exclusive). Also, note | |
3446 | that @code{(floor* @var{x})} is exactly equivalent to | |
3447 | @code{(floor* @var{x} 1)}. | |
3448 | ||
3449 | This function is entirely compatible with Common Lisp's @code{floor} | |
3450 | function, except that it returns the two results in a list since | |
3451 | Emacs Lisp does not support multiple-valued functions. | |
3452 | @end defun | |
3453 | ||
3454 | @defun ceiling* number &optional divisor | |
3455 | This function implements the Common Lisp @code{ceiling} function, | |
3456 | which is analogous to @code{floor} except that it rounds the | |
3457 | argument or quotient of the arguments up toward plus infinity. | |
3458 | The remainder will be between 0 and minus @var{r}. | |
3459 | @end defun | |
3460 | ||
3461 | @defun truncate* number &optional divisor | |
3462 | This function implements the Common Lisp @code{truncate} function, | |
3463 | which is analogous to @code{floor} except that it rounds the | |
3464 | argument or quotient of the arguments toward zero. Thus it is | |
3465 | equivalent to @code{floor*} if the argument or quotient is | |
3466 | positive, or to @code{ceiling*} otherwise. The remainder has | |
3467 | the same sign as @var{number}. | |
3468 | @end defun | |
3469 | ||
3470 | @defun round* number &optional divisor | |
3471 | This function implements the Common Lisp @code{round} function, | |
3472 | which is analogous to @code{floor} except that it rounds the | |
3473 | argument or quotient of the arguments to the nearest integer. | |
3474 | In the case of a tie (the argument or quotient is exactly | |
3475 | halfway between two integers), it rounds to the even integer. | |
3476 | @end defun | |
3477 | ||
3478 | @defun mod* number divisor | |
3479 | This function returns the same value as the second return value | |
3480 | of @code{floor}. | |
3481 | @end defun | |
3482 | ||
3483 | @defun rem* number divisor | |
3484 | This function returns the same value as the second return value | |
3485 | of @code{truncate}. | |
3486 | @end defun | |
3487 | ||
3488 | These definitions are compatible with those in the Quiroz | |
3489 | @file{cl.el} package, except that this package appends @samp{*} | |
3490 | to certain function names to avoid conflicts with existing | |
c6ad01e8 | 3491 | Emacs functions, and that the mechanism for returning |
6bf7aab6 DL |
3492 | multiple values is different. |
3493 | ||
3494 | @iftex | |
3495 | @secno=8 | |
3496 | @end iftex | |
3497 | ||
3498 | @node Random Numbers, Implementation Parameters, Numerical Functions, Numbers | |
3499 | @section Random Numbers | |
3500 | ||
3501 | @noindent | |
3502 | This package also provides an implementation of the Common Lisp | |
3503 | random number generator. It uses its own additive-congruential | |
3504 | algorithm, which is much more likely to give statistically clean | |
3505 | random numbers than the simple generators supplied by many | |
3506 | operating systems. | |
3507 | ||
3508 | @defun random* number &optional state | |
3509 | This function returns a random nonnegative number less than | |
3510 | @var{number}, and of the same type (either integer or floating-point). | |
3511 | The @var{state} argument should be a @code{random-state} object | |
3512 | which holds the state of the random number generator. The | |
3513 | function modifies this state object as a side effect. If | |
3514 | @var{state} is omitted, it defaults to the variable | |
3515 | @code{*random-state*}, which contains a pre-initialized | |
3516 | @code{random-state} object. | |
3517 | @end defun | |
3518 | ||
3519 | @defvar *random-state* | |
3520 | This variable contains the system ``default'' @code{random-state} | |
3521 | object, used for calls to @code{random*} that do not specify an | |
3522 | alternative state object. Since any number of programs in the | |
3523 | Emacs process may be accessing @code{*random-state*} in interleaved | |
3524 | fashion, the sequence generated from this variable will be | |
3525 | irreproducible for all intents and purposes. | |
3526 | @end defvar | |
3527 | ||
3528 | @defun make-random-state &optional state | |
3529 | This function creates or copies a @code{random-state} object. | |
3530 | If @var{state} is omitted or @code{nil}, it returns a new copy of | |
3531 | @code{*random-state*}. This is a copy in the sense that future | |
3532 | sequences of calls to @code{(random* @var{n})} and | |
3533 | @code{(random* @var{n} @var{s})} (where @var{s} is the new | |
3534 | random-state object) will return identical sequences of random | |
3535 | numbers. | |
3536 | ||
3537 | If @var{state} is a @code{random-state} object, this function | |
3538 | returns a copy of that object. If @var{state} is @code{t}, this | |
3539 | function returns a new @code{random-state} object seeded from the | |
3540 | date and time. As an extension to Common Lisp, @var{state} may also | |
3541 | be an integer in which case the new object is seeded from that | |
3542 | integer; each different integer seed will result in a completely | |
3543 | different sequence of random numbers. | |
3544 | ||
2cbd16b9 | 3545 | It is valid to print a @code{random-state} object to a buffer or |
6bf7aab6 DL |
3546 | file and later read it back with @code{read}. If a program wishes |
3547 | to use a sequence of pseudo-random numbers which can be reproduced | |
3548 | later for debugging, it can call @code{(make-random-state t)} to | |
3549 | get a new sequence, then print this sequence to a file. When the | |
3550 | program is later rerun, it can read the original run's random-state | |
3551 | from the file. | |
3552 | @end defun | |
3553 | ||
3554 | @defun random-state-p object | |
3555 | This predicate returns @code{t} if @var{object} is a | |
3556 | @code{random-state} object, or @code{nil} otherwise. | |
3557 | @end defun | |
3558 | ||
c6ad01e8 | 3559 | @node Implementation Parameters, , Random Numbers, Numbers |
6bf7aab6 DL |
3560 | @section Implementation Parameters |
3561 | ||
3562 | @noindent | |
3563 | This package defines several useful constants having to with numbers. | |
3564 | ||
6bf7aab6 DL |
3565 | The following parameters have to do with floating-point numbers. |
3566 | This package determines their values by exercising the computer's | |
3567 | floating-point arithmetic in various ways. Because this operation | |
3568 | might be slow, the code for initializing them is kept in a separate | |
3569 | function that must be called before the parameters can be used. | |
3570 | ||
3571 | @defun cl-float-limits | |
c6ad01e8 DL |
3572 | This function makes sure that the Common Lisp floating-point parameters |
3573 | like @code{most-positive-float} have been initialized. Until it is | |
3574 | called, these parameters will be @code{nil}. If this version of Emacs | |
3575 | does not support floats, the parameters will remain @code{nil}. If the | |
3576 | parameters have already been initialized, the function returns | |
3577 | immediately. | |
6bf7aab6 DL |
3578 | |
3579 | The algorithm makes assumptions that will be valid for most modern | |
3580 | machines, but will fail if the machine's arithmetic is extremely | |
3581 | unusual, e.g., decimal. | |
3582 | @end defun | |
3583 | ||
3584 | Since true Common Lisp supports up to four different floating-point | |
3585 | precisions, it has families of constants like | |
3586 | @code{most-positive-single-float}, @code{most-positive-double-float}, | |
3587 | @code{most-positive-long-float}, and so on. Emacs has only one | |
3588 | floating-point precision, so this package omits the precision word | |
3589 | from the constants' names. | |
3590 | ||
3591 | @defvar most-positive-float | |
3592 | This constant equals the largest value a Lisp float can hold. | |
3593 | For those systems whose arithmetic supports infinities, this is | |
3594 | the largest @emph{finite} value. For IEEE machines, the value | |
3595 | is approximately @code{1.79e+308}. | |
3596 | @end defvar | |
3597 | ||
3598 | @defvar most-negative-float | |
3599 | This constant equals the most-negative value a Lisp float can hold. | |
3600 | (It is assumed to be equal to @code{(- most-positive-float)}.) | |
3601 | @end defvar | |
3602 | ||
3603 | @defvar least-positive-float | |
3604 | This constant equals the smallest Lisp float value greater than zero. | |
3605 | For IEEE machines, it is about @code{4.94e-324} if denormals are | |
3606 | supported or @code{2.22e-308} if not. | |
3607 | @end defvar | |
3608 | ||
3609 | @defvar least-positive-normalized-float | |
3610 | This constant equals the smallest @emph{normalized} Lisp float greater | |
3611 | than zero, i.e., the smallest value for which IEEE denormalization | |
3612 | will not result in a loss of precision. For IEEE machines, this | |
3613 | value is about @code{2.22e-308}. For machines that do not support | |
3614 | the concept of denormalization and gradual underflow, this constant | |
3615 | will always equal @code{least-positive-float}. | |
3616 | @end defvar | |
3617 | ||
3618 | @defvar least-negative-float | |
3619 | This constant is the negative counterpart of @code{least-positive-float}. | |
3620 | @end defvar | |
3621 | ||
3622 | @defvar least-negative-normalized-float | |
3623 | This constant is the negative counterpart of | |
3624 | @code{least-positive-normalized-float}. | |
3625 | @end defvar | |
3626 | ||
3627 | @defvar float-epsilon | |
3628 | This constant is the smallest positive Lisp float that can be added | |
3629 | to 1.0 to produce a distinct value. Adding a smaller number to 1.0 | |
3630 | will yield 1.0 again due to roundoff. For IEEE machines, epsilon | |
3631 | is about @code{2.22e-16}. | |
3632 | @end defvar | |
3633 | ||
3634 | @defvar float-negative-epsilon | |
3635 | This is the smallest positive value that can be subtracted from | |
3636 | 1.0 to produce a distinct value. For IEEE machines, it is about | |
3637 | @code{1.11e-16}. | |
3638 | @end defvar | |
3639 | ||
3640 | @iftex | |
3641 | @chapno=13 | |
3642 | @end iftex | |
3643 | ||
3644 | @node Sequences, Lists, Numbers, Top | |
3645 | @chapter Sequences | |
3646 | ||
3647 | @noindent | |
3648 | Common Lisp defines a number of functions that operate on | |
3649 | @dfn{sequences}, which are either lists, strings, or vectors. | |
3650 | Emacs Lisp includes a few of these, notably @code{elt} and | |
3651 | @code{length}; this package defines most of the rest. | |
3652 | ||
3653 | @menu | |
3654 | * Sequence Basics:: Arguments shared by all sequence functions | |
3655 | * Mapping over Sequences:: `mapcar*', `mapcan', `map', `every', etc. | |
3656 | * Sequence Functions:: `subseq', `remove*', `substitute', etc. | |
3657 | * Searching Sequences:: `find', `position', `count', `search', etc. | |
3658 | * Sorting Sequences:: `sort*', `stable-sort', `merge' | |
3659 | @end menu | |
3660 | ||
3661 | @node Sequence Basics, Mapping over Sequences, Sequences, Sequences | |
3662 | @section Sequence Basics | |
3663 | ||
3664 | @noindent | |
3665 | Many of the sequence functions take keyword arguments; @pxref{Argument | |
3666 | Lists}. All keyword arguments are optional and, if specified, | |
3667 | may appear in any order. | |
3668 | ||
3669 | The @code{:key} argument should be passed either @code{nil}, or a | |
3670 | function of one argument. This key function is used as a filter | |
3671 | through which the elements of the sequence are seen; for example, | |
3672 | @code{(find x y :key 'car)} is similar to @code{(assoc* x y)}: | |
3673 | It searches for an element of the list whose @code{car} equals | |
3674 | @code{x}, rather than for an element which equals @code{x} itself. | |
3675 | If @code{:key} is omitted or @code{nil}, the filter is effectively | |
3676 | the identity function. | |
3677 | ||
3678 | The @code{:test} and @code{:test-not} arguments should be either | |
3679 | @code{nil}, or functions of two arguments. The test function is | |
3680 | used to compare two sequence elements, or to compare a search value | |
3681 | with sequence elements. (The two values are passed to the test | |
3682 | function in the same order as the original sequence function | |
3683 | arguments from which they are derived, or, if they both come from | |
3684 | the same sequence, in the same order as they appear in that sequence.) | |
3685 | The @code{:test} argument specifies a function which must return | |
3686 | true (non-@code{nil}) to indicate a match; instead, you may use | |
3687 | @code{:test-not} to give a function which returns @emph{false} to | |
3688 | indicate a match. The default test function is @code{:test 'eql}. | |
3689 | ||
3690 | Many functions which take @var{item} and @code{:test} or @code{:test-not} | |
3691 | arguments also come in @code{-if} and @code{-if-not} varieties, | |
3692 | where a @var{predicate} function is passed instead of @var{item}, | |
3693 | and sequence elements match if the predicate returns true on them | |
3694 | (or false in the case of @code{-if-not}). For example: | |
3695 | ||
3696 | @example | |
3697 | (remove* 0 seq :test '=) @equiv{} (remove-if 'zerop seq) | |
3698 | @end example | |
3699 | ||
3700 | @noindent | |
3701 | to remove all zeros from sequence @code{seq}. | |
3702 | ||
3703 | Some operations can work on a subsequence of the argument sequence; | |
3704 | these function take @code{:start} and @code{:end} arguments which | |
3705 | default to zero and the length of the sequence, respectively. | |
3706 | Only elements between @var{start} (inclusive) and @var{end} | |
3707 | (exclusive) are affected by the operation. The @var{end} argument | |
3708 | may be passed @code{nil} to signify the length of the sequence; | |
3709 | otherwise, both @var{start} and @var{end} must be integers, with | |
3710 | @code{0 <= @var{start} <= @var{end} <= (length @var{seq})}. | |
3711 | If the function takes two sequence arguments, the limits are | |
3712 | defined by keywords @code{:start1} and @code{:end1} for the first, | |
3713 | and @code{:start2} and @code{:end2} for the second. | |
3714 | ||
3715 | A few functions accept a @code{:from-end} argument, which, if | |
3716 | non-@code{nil}, causes the operation to go from right-to-left | |
3717 | through the sequence instead of left-to-right, and a @code{:count} | |
3718 | argument, which specifies an integer maximum number of elements | |
3719 | to be removed or otherwise processed. | |
3720 | ||
3721 | The sequence functions make no guarantees about the order in | |
3722 | which the @code{:test}, @code{:test-not}, and @code{:key} functions | |
3723 | are called on various elements. Therefore, it is a bad idea to depend | |
3724 | on side effects of these functions. For example, @code{:from-end} | |
3725 | may cause the sequence to be scanned actually in reverse, or it may | |
3726 | be scanned forwards but computing a result ``as if'' it were scanned | |
3727 | backwards. (Some functions, like @code{mapcar*} and @code{every}, | |
3728 | @emph{do} specify exactly the order in which the function is called | |
3729 | so side effects are perfectly acceptable in those cases.) | |
3730 | ||
c6ad01e8 | 3731 | Strings may contain ``text properties'' as well |
6bf7aab6 DL |
3732 | as character data. Except as noted, it is undefined whether or |
3733 | not text properties are preserved by sequence functions. For | |
3734 | example, @code{(remove* ?A @var{str})} may or may not preserve | |
3735 | the properties of the characters copied from @var{str} into the | |
3736 | result. | |
3737 | ||
3738 | @node Mapping over Sequences, Sequence Functions, Sequence Basics, Sequences | |
3739 | @section Mapping over Sequences | |
3740 | ||
3741 | @noindent | |
3742 | These functions ``map'' the function you specify over the elements | |
3743 | of lists or arrays. They are all variations on the theme of the | |
3744 | built-in function @code{mapcar}. | |
3745 | ||
3746 | @defun mapcar* function seq &rest more-seqs | |
3747 | This function calls @var{function} on successive parallel sets of | |
3748 | elements from its argument sequences. Given a single @var{seq} | |
3749 | argument it is equivalent to @code{mapcar}; given @var{n} sequences, | |
3750 | it calls the function with the first elements of each of the sequences | |
3751 | as the @var{n} arguments to yield the first element of the result | |
3752 | list, then with the second elements, and so on. The mapping stops as | |
3753 | soon as the shortest sequence runs out. The argument sequences may | |
3754 | be any mixture of lists, strings, and vectors; the return sequence | |
3755 | is always a list. | |
3756 | ||
3757 | Common Lisp's @code{mapcar} accepts multiple arguments but works | |
3758 | only on lists; Emacs Lisp's @code{mapcar} accepts a single sequence | |
3759 | argument. This package's @code{mapcar*} works as a compatible | |
3760 | superset of both. | |
3761 | @end defun | |
3762 | ||
3763 | @defun map result-type function seq &rest more-seqs | |
3764 | This function maps @var{function} over the argument sequences, | |
3765 | just like @code{mapcar*}, but it returns a sequence of type | |
3766 | @var{result-type} rather than a list. @var{result-type} must | |
3767 | be one of the following symbols: @code{vector}, @code{string}, | |
3768 | @code{list} (in which case the effect is the same as for | |
3769 | @code{mapcar*}), or @code{nil} (in which case the results are | |
3770 | thrown away and @code{map} returns @code{nil}). | |
3771 | @end defun | |
3772 | ||
3773 | @defun maplist function list &rest more-lists | |
3774 | This function calls @var{function} on each of its argument lists, | |
3775 | then on the @code{cdr}s of those lists, and so on, until the | |
3776 | shortest list runs out. The results are returned in the form | |
3777 | of a list. Thus, @code{maplist} is like @code{mapcar*} except | |
3778 | that it passes in the list pointers themselves rather than the | |
3779 | @code{car}s of the advancing pointers. | |
3780 | @end defun | |
3781 | ||
3782 | @defun mapc function seq &rest more-seqs | |
b7a084f6 DL |
3783 | This function is like @code{mapcar*}, except that the values returned |
3784 | by @var{function} are ignored and thrown away rather than being | |
3785 | collected into a list. The return value of @code{mapc} is @var{seq}, | |
3786 | the first sequence. This function is more general than the Emacs | |
3787 | primitive @code{mapc}. | |
6bf7aab6 DL |
3788 | @end defun |
3789 | ||
3790 | @defun mapl function list &rest more-lists | |
3791 | This function is like @code{maplist}, except that it throws away | |
3792 | the values returned by @var{function}. | |
3793 | @end defun | |
3794 | ||
3795 | @defun mapcan function seq &rest more-seqs | |
3796 | This function is like @code{mapcar*}, except that it concatenates | |
3797 | the return values (which must be lists) using @code{nconc}, | |
3798 | rather than simply collecting them into a list. | |
3799 | @end defun | |
3800 | ||
3801 | @defun mapcon function list &rest more-lists | |
3802 | This function is like @code{maplist}, except that it concatenates | |
3803 | the return values using @code{nconc}. | |
3804 | @end defun | |
3805 | ||
3806 | @defun some predicate seq &rest more-seqs | |
3807 | This function calls @var{predicate} on each element of @var{seq} | |
3808 | in turn; if @var{predicate} returns a non-@code{nil} value, | |
3809 | @code{some} returns that value, otherwise it returns @code{nil}. | |
3810 | Given several sequence arguments, it steps through the sequences | |
3811 | in parallel until the shortest one runs out, just as in | |
3812 | @code{mapcar*}. You can rely on the left-to-right order in which | |
3813 | the elements are visited, and on the fact that mapping stops | |
3814 | immediately as soon as @var{predicate} returns non-@code{nil}. | |
3815 | @end defun | |
3816 | ||
3817 | @defun every predicate seq &rest more-seqs | |
3818 | This function calls @var{predicate} on each element of the sequence(s) | |
3819 | in turn; it returns @code{nil} as soon as @var{predicate} returns | |
3820 | @code{nil} for any element, or @code{t} if the predicate was true | |
3821 | for all elements. | |
3822 | @end defun | |
3823 | ||
3824 | @defun notany predicate seq &rest more-seqs | |
3825 | This function calls @var{predicate} on each element of the sequence(s) | |
3826 | in turn; it returns @code{nil} as soon as @var{predicate} returns | |
3827 | a non-@code{nil} value for any element, or @code{t} if the predicate | |
3828 | was @code{nil} for all elements. | |
3829 | @end defun | |
3830 | ||
3831 | @defun notevery predicate seq &rest more-seqs | |
3832 | This function calls @var{predicate} on each element of the sequence(s) | |
3833 | in turn; it returns a non-@code{nil} value as soon as @var{predicate} | |
3834 | returns @code{nil} for any element, or @code{t} if the predicate was | |
3835 | true for all elements. | |
3836 | @end defun | |
3837 | ||
3838 | @defun reduce function seq @t{&key :from-end :start :end :initial-value :key} | |
3839 | This function combines the elements of @var{seq} using an associative | |
3840 | binary operation. Suppose @var{function} is @code{*} and @var{seq} is | |
3841 | the list @code{(2 3 4 5)}. The first two elements of the list are | |
3842 | combined with @code{(* 2 3) = 6}; this is combined with the next | |
3843 | element, @code{(* 6 4) = 24}, and that is combined with the final | |
3844 | element: @code{(* 24 5) = 120}. Note that the @code{*} function happens | |
3845 | to be self-reducing, so that @code{(* 2 3 4 5)} has the same effect as | |
3846 | an explicit call to @code{reduce}. | |
3847 | ||
3848 | If @code{:from-end} is true, the reduction is right-associative instead | |
3849 | of left-associative: | |
3850 | ||
3851 | @example | |
3852 | (reduce '- '(1 2 3 4)) | |
3853 | @equiv{} (- (- (- 1 2) 3) 4) @result{} -8 | |
3854 | (reduce '- '(1 2 3 4) :from-end t) | |
3855 | @equiv{} (- 1 (- 2 (- 3 4))) @result{} -2 | |
3856 | @end example | |
3857 | ||
3858 | If @code{:key} is specified, it is a function of one argument which | |
3859 | is called on each of the sequence elements in turn. | |
3860 | ||
3861 | If @code{:initial-value} is specified, it is effectively added to the | |
3862 | front (or rear in the case of @code{:from-end}) of the sequence. | |
3863 | The @code{:key} function is @emph{not} applied to the initial value. | |
3864 | ||
3865 | If the sequence, including the initial value, has exactly one element | |
3866 | then that element is returned without ever calling @var{function}. | |
3867 | If the sequence is empty (and there is no initial value), then | |
3868 | @var{function} is called with no arguments to obtain the return value. | |
3869 | @end defun | |
3870 | ||
3871 | All of these mapping operations can be expressed conveniently in | |
3872 | terms of the @code{loop} macro. In compiled code, @code{loop} will | |
3873 | be faster since it generates the loop as in-line code with no | |
3874 | function calls. | |
3875 | ||
3876 | @node Sequence Functions, Searching Sequences, Mapping over Sequences, Sequences | |
3877 | @section Sequence Functions | |
3878 | ||
3879 | @noindent | |
3880 | This section describes a number of Common Lisp functions for | |
3881 | operating on sequences. | |
3882 | ||
3883 | @defun subseq sequence start &optional end | |
3884 | This function returns a given subsequence of the argument | |
3885 | @var{sequence}, which may be a list, string, or vector. | |
3886 | The indices @var{start} and @var{end} must be in range, and | |
3887 | @var{start} must be no greater than @var{end}. If @var{end} | |
3888 | is omitted, it defaults to the length of the sequence. The | |
3889 | return value is always a copy; it does not share structure | |
3890 | with @var{sequence}. | |
3891 | ||
3892 | As an extension to Common Lisp, @var{start} and/or @var{end} | |
3893 | may be negative, in which case they represent a distance back | |
3894 | from the end of the sequence. This is for compatibility with | |
3895 | Emacs' @code{substring} function. Note that @code{subseq} is | |
3896 | the @emph{only} sequence function that allows negative | |
3897 | @var{start} and @var{end}. | |
3898 | ||
3899 | You can use @code{setf} on a @code{subseq} form to replace a | |
3900 | specified range of elements with elements from another sequence. | |
3901 | The replacement is done as if by @code{replace}, described below. | |
3902 | @end defun | |
3903 | ||
3904 | @defun concatenate result-type &rest seqs | |
3905 | This function concatenates the argument sequences together to | |
3906 | form a result sequence of type @var{result-type}, one of the | |
3907 | symbols @code{vector}, @code{string}, or @code{list}. The | |
3908 | arguments are always copied, even in cases such as | |
3909 | @code{(concatenate 'list '(1 2 3))} where the result is | |
3910 | identical to an argument. | |
3911 | @end defun | |
3912 | ||
3913 | @defun fill seq item @t{&key :start :end} | |
3914 | This function fills the elements of the sequence (or the specified | |
3915 | part of the sequence) with the value @var{item}. | |
3916 | @end defun | |
3917 | ||
3918 | @defun replace seq1 seq2 @t{&key :start1 :end1 :start2 :end2} | |
3919 | This function copies part of @var{seq2} into part of @var{seq1}. | |
3920 | The sequence @var{seq1} is not stretched or resized; the amount | |
3921 | of data copied is simply the shorter of the source and destination | |
3922 | (sub)sequences. The function returns @var{seq1}. | |
3923 | ||
3924 | If @var{seq1} and @var{seq2} are @code{eq}, then the replacement | |
3925 | will work correctly even if the regions indicated by the start | |
3926 | and end arguments overlap. However, if @var{seq1} and @var{seq2} | |
3927 | are lists which share storage but are not @code{eq}, and the | |
3928 | start and end arguments specify overlapping regions, the effect | |
3929 | is undefined. | |
3930 | @end defun | |
3931 | ||
3932 | @defun remove* item seq @t{&key :test :test-not :key :count :start :end :from-end} | |
3933 | This returns a copy of @var{seq} with all elements matching | |
3934 | @var{item} removed. The result may share storage with or be | |
3935 | @code{eq} to @var{seq} in some circumstances, but the original | |
3936 | @var{seq} will not be modified. The @code{:test}, @code{:test-not}, | |
3937 | and @code{:key} arguments define the matching test that is used; | |
3938 | by default, elements @code{eql} to @var{item} are removed. The | |
3939 | @code{:count} argument specifies the maximum number of matching | |
3940 | elements that can be removed (only the leftmost @var{count} matches | |
3941 | are removed). The @code{:start} and @code{:end} arguments specify | |
3942 | a region in @var{seq} in which elements will be removed; elements | |
3943 | outside that region are not matched or removed. The @code{:from-end} | |
3944 | argument, if true, says that elements should be deleted from the | |
3945 | end of the sequence rather than the beginning (this matters only | |
3946 | if @var{count} was also specified). | |
3947 | @end defun | |
3948 | ||
3949 | @defun delete* item seq @t{&key :test :test-not :key :count :start :end :from-end} | |
3950 | This deletes all elements of @var{seq} which match @var{item}. | |
3951 | It is a destructive operation. Since Emacs Lisp does not support | |
3952 | stretchable strings or vectors, this is the same as @code{remove*} | |
3953 | for those sequence types. On lists, @code{remove*} will copy the | |
3954 | list if necessary to preserve the original list, whereas | |
3955 | @code{delete*} will splice out parts of the argument list. | |
3956 | Compare @code{append} and @code{nconc}, which are analogous | |
3957 | non-destructive and destructive list operations in Emacs Lisp. | |
3958 | @end defun | |
3959 | ||
3960 | @findex remove-if | |
3961 | @findex remove-if-not | |
3962 | @findex delete-if | |
3963 | @findex delete-if-not | |
3964 | The predicate-oriented functions @code{remove-if}, @code{remove-if-not}, | |
3965 | @code{delete-if}, and @code{delete-if-not} are defined similarly. | |
3966 | ||
6bf7aab6 DL |
3967 | @defun remove-duplicates seq @t{&key :test :test-not :key :start :end :from-end} |
3968 | This function returns a copy of @var{seq} with duplicate elements | |
3969 | removed. Specifically, if two elements from the sequence match | |
3970 | according to the @code{:test}, @code{:test-not}, and @code{:key} | |
3971 | arguments, only the rightmost one is retained. If @code{:from-end} | |
3972 | is true, the leftmost one is retained instead. If @code{:start} or | |
3973 | @code{:end} is specified, only elements within that subsequence are | |
3974 | examined or removed. | |
3975 | @end defun | |
3976 | ||
3977 | @defun delete-duplicates seq @t{&key :test :test-not :key :start :end :from-end} | |
3978 | This function deletes duplicate elements from @var{seq}. It is | |
3979 | a destructive version of @code{remove-duplicates}. | |
3980 | @end defun | |
3981 | ||
3982 | @defun substitute new old seq @t{&key :test :test-not :key :count :start :end :from-end} | |
3983 | This function returns a copy of @var{seq}, with all elements | |
3984 | matching @var{old} replaced with @var{new}. The @code{:count}, | |
3985 | @code{:start}, @code{:end}, and @code{:from-end} arguments may be | |
3986 | used to limit the number of substitutions made. | |
3987 | @end defun | |
3988 | ||
3989 | @defun nsubstitute new old seq @t{&key :test :test-not :key :count :start :end :from-end} | |
3990 | This is a destructive version of @code{substitute}; it performs | |
3991 | the substitution using @code{setcar} or @code{aset} rather than | |
3992 | by returning a changed copy of the sequence. | |
3993 | @end defun | |
3994 | ||
3995 | @findex substitute-if | |
3996 | @findex substitute-if-not | |
3997 | @findex nsubstitute-if | |
3998 | @findex nsubstitute-if-not | |
3999 | The @code{substitute-if}, @code{substitute-if-not}, @code{nsubstitute-if}, | |
4000 | and @code{nsubstitute-if-not} functions are defined similarly. For | |
4001 | these, a @var{predicate} is given in place of the @var{old} argument. | |
4002 | ||
4003 | @node Searching Sequences, Sorting Sequences, Sequence Functions, Sequences | |
4004 | @section Searching Sequences | |
4005 | ||
4006 | @noindent | |
4007 | These functions search for elements or subsequences in a sequence. | |
4008 | (See also @code{member*} and @code{assoc*}; @pxref{Lists}.) | |
4009 | ||
4010 | @defun find item seq @t{&key :test :test-not :key :start :end :from-end} | |
4011 | This function searches @var{seq} for an element matching @var{item}. | |
4012 | If it finds a match, it returns the matching element. Otherwise, | |
4013 | it returns @code{nil}. It returns the leftmost match, unless | |
4014 | @code{:from-end} is true, in which case it returns the rightmost | |
4015 | match. The @code{:start} and @code{:end} arguments may be used to | |
4016 | limit the range of elements that are searched. | |
4017 | @end defun | |
4018 | ||
4019 | @defun position item seq @t{&key :test :test-not :key :start :end :from-end} | |
4020 | This function is like @code{find}, except that it returns the | |
4021 | integer position in the sequence of the matching item rather than | |
4022 | the item itself. The position is relative to the start of the | |
4023 | sequence as a whole, even if @code{:start} is non-zero. The function | |
4024 | returns @code{nil} if no matching element was found. | |
4025 | @end defun | |
4026 | ||
4027 | @defun count item seq @t{&key :test :test-not :key :start :end} | |
4028 | This function returns the number of elements of @var{seq} which | |
4029 | match @var{item}. The result is always a nonnegative integer. | |
4030 | @end defun | |
4031 | ||
4032 | @findex find-if | |
4033 | @findex find-if-not | |
4034 | @findex position-if | |
4035 | @findex position-if-not | |
4036 | @findex count-if | |
4037 | @findex count-if-not | |
4038 | The @code{find-if}, @code{find-if-not}, @code{position-if}, | |
4039 | @code{position-if-not}, @code{count-if}, and @code{count-if-not} | |
4040 | functions are defined similarly. | |
4041 | ||
4042 | @defun mismatch seq1 seq2 @t{&key :test :test-not :key :start1 :end1 :start2 :end2 :from-end} | |
4043 | This function compares the specified parts of @var{seq1} and | |
4044 | @var{seq2}. If they are the same length and the corresponding | |
4045 | elements match (according to @code{:test}, @code{:test-not}, | |
4046 | and @code{:key}), the function returns @code{nil}. If there is | |
4047 | a mismatch, the function returns the index (relative to @var{seq1}) | |
4048 | of the first mismatching element. This will be the leftmost pair of | |
4049 | elements which do not match, or the position at which the shorter of | |
4050 | the two otherwise-matching sequences runs out. | |
4051 | ||
4052 | If @code{:from-end} is true, then the elements are compared from right | |
4053 | to left starting at @code{(1- @var{end1})} and @code{(1- @var{end2})}. | |
4054 | If the sequences differ, then one plus the index of the rightmost | |
4055 | difference (relative to @var{seq1}) is returned. | |
4056 | ||
4057 | An interesting example is @code{(mismatch str1 str2 :key 'upcase)}, | |
4058 | which compares two strings case-insensitively. | |
4059 | @end defun | |
4060 | ||
4061 | @defun search seq1 seq2 @t{&key :test :test-not :key :from-end :start1 :end1 :start2 :end2} | |
4062 | This function searches @var{seq2} for a subsequence that matches | |
4063 | @var{seq1} (or part of it specified by @code{:start1} and | |
4064 | @code{:end1}.) Only matches which fall entirely within the region | |
4065 | defined by @code{:start2} and @code{:end2} will be considered. | |
4066 | The return value is the index of the leftmost element of the | |
4067 | leftmost match, relative to the start of @var{seq2}, or @code{nil} | |
4068 | if no matches were found. If @code{:from-end} is true, the | |
4069 | function finds the @emph{rightmost} matching subsequence. | |
4070 | @end defun | |
4071 | ||
c6ad01e8 | 4072 | @node Sorting Sequences, , Searching Sequences, Sequences |
6bf7aab6 DL |
4073 | @section Sorting Sequences |
4074 | ||
4075 | @defun sort* seq predicate @t{&key :key} | |
4076 | This function sorts @var{seq} into increasing order as determined | |
4077 | by using @var{predicate} to compare pairs of elements. @var{predicate} | |
4078 | should return true (non-@code{nil}) if and only if its first argument | |
4079 | is less than (not equal to) its second argument. For example, | |
4080 | @code{<} and @code{string-lessp} are suitable predicate functions | |
4081 | for sorting numbers and strings, respectively; @code{>} would sort | |
4082 | numbers into decreasing rather than increasing order. | |
4083 | ||
4084 | This function differs from Emacs' built-in @code{sort} in that it | |
4085 | can operate on any type of sequence, not just lists. Also, it | |
4086 | accepts a @code{:key} argument which is used to preprocess data | |
4087 | fed to the @var{predicate} function. For example, | |
4088 | ||
4089 | @example | |
8f918489 | 4090 | (setq data (sort* data 'string-lessp :key 'downcase)) |
6bf7aab6 DL |
4091 | @end example |
4092 | ||
4093 | @noindent | |
4094 | sorts @var{data}, a sequence of strings, into increasing alphabetical | |
4095 | order without regard to case. A @code{:key} function of @code{car} | |
92ad3e9f JB |
4096 | would be useful for sorting association lists. It should only be a |
4097 | simple accessor though, it's used heavily in the current | |
4098 | implementation. | |
6bf7aab6 DL |
4099 | |
4100 | The @code{sort*} function is destructive; it sorts lists by actually | |
4101 | rearranging the @code{cdr} pointers in suitable fashion. | |
4102 | @end defun | |
4103 | ||
4104 | @defun stable-sort seq predicate @t{&key :key} | |
4105 | This function sorts @var{seq} @dfn{stably}, meaning two elements | |
4106 | which are equal in terms of @var{predicate} are guaranteed not to | |
4107 | be rearranged out of their original order by the sort. | |
4108 | ||
4109 | In practice, @code{sort*} and @code{stable-sort} are equivalent | |
4110 | in Emacs Lisp because the underlying @code{sort} function is | |
4111 | stable by default. However, this package reserves the right to | |
4112 | use non-stable methods for @code{sort*} in the future. | |
4113 | @end defun | |
4114 | ||
4115 | @defun merge type seq1 seq2 predicate @t{&key :key} | |
4116 | This function merges two sequences @var{seq1} and @var{seq2} by | |
4117 | interleaving their elements. The result sequence, of type @var{type} | |
4118 | (in the sense of @code{concatenate}), has length equal to the sum | |
4119 | of the lengths of the two input sequences. The sequences may be | |
4120 | modified destructively. Order of elements within @var{seq1} and | |
4121 | @var{seq2} is preserved in the interleaving; elements of the two | |
4122 | sequences are compared by @var{predicate} (in the sense of | |
4123 | @code{sort}) and the lesser element goes first in the result. | |
4124 | When elements are equal, those from @var{seq1} precede those from | |
4125 | @var{seq2} in the result. Thus, if @var{seq1} and @var{seq2} are | |
4126 | both sorted according to @var{predicate}, then the result will be | |
4127 | a merged sequence which is (stably) sorted according to | |
4128 | @var{predicate}. | |
4129 | @end defun | |
4130 | ||
c6ad01e8 | 4131 | @node Lists, Structures, Sequences, Top |
6bf7aab6 DL |
4132 | @chapter Lists |
4133 | ||
4134 | @noindent | |
4135 | The functions described here operate on lists. | |
4136 | ||
4137 | @menu | |
b7a084f6 | 4138 | * List Functions:: `caddr', `first', `list*', etc. |
6bf7aab6 DL |
4139 | * Substitution of Expressions:: `subst', `sublis', etc. |
4140 | * Lists as Sets:: `member*', `adjoin', `union', etc. | |
4141 | * Association Lists:: `assoc*', `rassoc*', `acons', `pairlis' | |
4142 | @end menu | |
4143 | ||
4144 | @node List Functions, Substitution of Expressions, Lists, Lists | |
4145 | @section List Functions | |
4146 | ||
4147 | @noindent | |
4148 | This section describes a number of simple operations on lists, | |
4149 | i.e., chains of cons cells. | |
4150 | ||
4151 | @defun caddr x | |
4152 | This function is equivalent to @code{(car (cdr (cdr @var{x})))}. | |
4153 | Likewise, this package defines all 28 @code{c@var{xxx}r} functions | |
4154 | where @var{xxx} is up to four @samp{a}s and/or @samp{d}s. | |
4155 | All of these functions are @code{setf}-able, and calls to them | |
4156 | are expanded inline by the byte-compiler for maximum efficiency. | |
4157 | @end defun | |
4158 | ||
4159 | @defun first x | |
4160 | This function is a synonym for @code{(car @var{x})}. Likewise, | |
4161 | the functions @code{second}, @code{third}, @dots{}, through | |
4162 | @code{tenth} return the given element of the list @var{x}. | |
4163 | @end defun | |
4164 | ||
4165 | @defun rest x | |
4166 | This function is a synonym for @code{(cdr @var{x})}. | |
4167 | @end defun | |
4168 | ||
4169 | @defun endp x | |
4170 | Common Lisp defines this function to act like @code{null}, but | |
4171 | signaling an error if @code{x} is neither a @code{nil} nor a | |
4172 | cons cell. This package simply defines @code{endp} as a synonym | |
4173 | for @code{null}. | |
4174 | @end defun | |
4175 | ||
4176 | @defun list-length x | |
4177 | This function returns the length of list @var{x}, exactly like | |
4178 | @code{(length @var{x})}, except that if @var{x} is a circular | |
4179 | list (where the cdr-chain forms a loop rather than terminating | |
4180 | with @code{nil}), this function returns @code{nil}. (The regular | |
4181 | @code{length} function would get stuck if given a circular list.) | |
4182 | @end defun | |
4183 | ||
6bf7aab6 DL |
4184 | @defun list* arg &rest others |
4185 | This function constructs a list of its arguments. The final | |
4186 | argument becomes the @code{cdr} of the last cell constructed. | |
4187 | Thus, @code{(list* @var{a} @var{b} @var{c})} is equivalent to | |
4188 | @code{(cons @var{a} (cons @var{b} @var{c}))}, and | |
4189 | @code{(list* @var{a} @var{b} nil)} is equivalent to | |
4190 | @code{(list @var{a} @var{b})}. | |
4191 | ||
4192 | (Note that this function really is called @code{list*} in Common | |
4193 | Lisp; it is not a name invented for this package like @code{member*} | |
4194 | or @code{defun*}.) | |
4195 | @end defun | |
4196 | ||
4197 | @defun ldiff list sublist | |
4198 | If @var{sublist} is a sublist of @var{list}, i.e., is @code{eq} to | |
4199 | one of the cons cells of @var{list}, then this function returns | |
4200 | a copy of the part of @var{list} up to but not including | |
4201 | @var{sublist}. For example, @code{(ldiff x (cddr x))} returns | |
4202 | the first two elements of the list @code{x}. The result is a | |
4203 | copy; the original @var{list} is not modified. If @var{sublist} | |
4204 | is not a sublist of @var{list}, a copy of the entire @var{list} | |
4205 | is returned. | |
4206 | @end defun | |
4207 | ||
4208 | @defun copy-list list | |
4209 | This function returns a copy of the list @var{list}. It copies | |
4210 | dotted lists like @code{(1 2 . 3)} correctly. | |
4211 | @end defun | |
4212 | ||
4213 | @defun copy-tree x &optional vecp | |
4214 | This function returns a copy of the tree of cons cells @var{x}. | |
4215 | Unlike @code{copy-sequence} (and its alias @code{copy-list}), | |
4216 | which copies only along the @code{cdr} direction, this function | |
4217 | copies (recursively) along both the @code{car} and the @code{cdr} | |
4218 | directions. If @var{x} is not a cons cell, the function simply | |
4219 | returns @var{x} unchanged. If the optional @var{vecp} argument | |
4220 | is true, this function copies vectors (recursively) as well as | |
4221 | cons cells. | |
4222 | @end defun | |
4223 | ||
4224 | @defun tree-equal x y @t{&key :test :test-not :key} | |
4225 | This function compares two trees of cons cells. If @var{x} and | |
4226 | @var{y} are both cons cells, their @code{car}s and @code{cdr}s are | |
4227 | compared recursively. If neither @var{x} nor @var{y} is a cons | |
4228 | cell, they are compared by @code{eql}, or according to the | |
4229 | specified test. The @code{:key} function, if specified, is | |
4230 | applied to the elements of both trees. @xref{Sequences}. | |
4231 | @end defun | |
4232 | ||
4233 | @iftex | |
4234 | @secno=3 | |
4235 | @end iftex | |
4236 | ||
4237 | @node Substitution of Expressions, Lists as Sets, List Functions, Lists | |
4238 | @section Substitution of Expressions | |
4239 | ||
4240 | @noindent | |
4241 | These functions substitute elements throughout a tree of cons | |
4242 | cells. (@xref{Sequence Functions}, for the @code{substitute} | |
4243 | function, which works on just the top-level elements of a list.) | |
4244 | ||
4245 | @defun subst new old tree @t{&key :test :test-not :key} | |
4246 | This function substitutes occurrences of @var{old} with @var{new} | |
4247 | in @var{tree}, a tree of cons cells. It returns a substituted | |
4248 | tree, which will be a copy except that it may share storage with | |
4249 | the argument @var{tree} in parts where no substitutions occurred. | |
4250 | The original @var{tree} is not modified. This function recurses | |
4251 | on, and compares against @var{old}, both @code{car}s and @code{cdr}s | |
4252 | of the component cons cells. If @var{old} is itself a cons cell, | |
4253 | then matching cells in the tree are substituted as usual without | |
4254 | recursively substituting in that cell. Comparisons with @var{old} | |
4255 | are done according to the specified test (@code{eql} by default). | |
4256 | The @code{:key} function is applied to the elements of the tree | |
4257 | but not to @var{old}. | |
4258 | @end defun | |
4259 | ||
4260 | @defun nsubst new old tree @t{&key :test :test-not :key} | |
4261 | This function is like @code{subst}, except that it works by | |
4262 | destructive modification (by @code{setcar} or @code{setcdr}) | |
4263 | rather than copying. | |
4264 | @end defun | |
4265 | ||
4266 | @findex subst-if | |
4267 | @findex subst-if-not | |
4268 | @findex nsubst-if | |
4269 | @findex nsubst-if-not | |
4270 | The @code{subst-if}, @code{subst-if-not}, @code{nsubst-if}, and | |
4271 | @code{nsubst-if-not} functions are defined similarly. | |
4272 | ||
4273 | @defun sublis alist tree @t{&key :test :test-not :key} | |
4274 | This function is like @code{subst}, except that it takes an | |
4275 | association list @var{alist} of @var{old}-@var{new} pairs. | |
4276 | Each element of the tree (after applying the @code{:key} | |
4277 | function, if any), is compared with the @code{car}s of | |
4278 | @var{alist}; if it matches, it is replaced by the corresponding | |
4279 | @code{cdr}. | |
4280 | @end defun | |
4281 | ||
4282 | @defun nsublis alist tree @t{&key :test :test-not :key} | |
4283 | This is a destructive version of @code{sublis}. | |
4284 | @end defun | |
4285 | ||
4286 | @node Lists as Sets, Association Lists, Substitution of Expressions, Lists | |
4287 | @section Lists as Sets | |
4288 | ||
4289 | @noindent | |
4290 | These functions perform operations on lists which represent sets | |
4291 | of elements. | |
4292 | ||
6bf7aab6 DL |
4293 | @defun member* item list @t{&key :test :test-not :key} |
4294 | This function searches @var{list} for an element matching @var{item}. | |
4295 | If a match is found, it returns the cons cell whose @code{car} was | |
4296 | the matching element. Otherwise, it returns @code{nil}. Elements | |
4297 | are compared by @code{eql} by default; you can use the @code{:test}, | |
4298 | @code{:test-not}, and @code{:key} arguments to modify this behavior. | |
4299 | @xref{Sequences}. | |
4300 | ||
4301 | Note that this function's name is suffixed by @samp{*} to avoid | |
c6ad01e8 | 4302 | the incompatible @code{member} function defined in Emacs. |
6bf7aab6 DL |
4303 | (That function uses @code{equal} for comparisons; it is equivalent |
4304 | to @code{(member* @var{item} @var{list} :test 'equal)}.) | |
4305 | @end defun | |
4306 | ||
4307 | @findex member-if | |
4308 | @findex member-if-not | |
4309 | The @code{member-if} and @code{member-if-not} functions | |
4310 | analogously search for elements which satisfy a given predicate. | |
4311 | ||
4312 | @defun tailp sublist list | |
4313 | This function returns @code{t} if @var{sublist} is a sublist of | |
4314 | @var{list}, i.e., if @var{sublist} is @code{eql} to @var{list} or to | |
4315 | any of its @code{cdr}s. | |
4316 | @end defun | |
4317 | ||
4318 | @defun adjoin item list @t{&key :test :test-not :key} | |
4319 | This function conses @var{item} onto the front of @var{list}, | |
4320 | like @code{(cons @var{item} @var{list})}, but only if @var{item} | |
4321 | is not already present on the list (as determined by @code{member*}). | |
4322 | If a @code{:key} argument is specified, it is applied to | |
4323 | @var{item} as well as to the elements of @var{list} during | |
4324 | the search, on the reasoning that @var{item} is ``about'' to | |
4325 | become part of the list. | |
4326 | @end defun | |
4327 | ||
4328 | @defun union list1 list2 @t{&key :test :test-not :key} | |
4329 | This function combines two lists which represent sets of items, | |
4330 | returning a list that represents the union of those two sets. | |
4331 | The result list will contain all items which appear in @var{list1} | |
4332 | or @var{list2}, and no others. If an item appears in both | |
4333 | @var{list1} and @var{list2} it will be copied only once. If | |
4334 | an item is duplicated in @var{list1} or @var{list2}, it is | |
4335 | undefined whether or not that duplication will survive in the | |
4336 | result list. The order of elements in the result list is also | |
4337 | undefined. | |
4338 | @end defun | |
4339 | ||
4340 | @defun nunion list1 list2 @t{&key :test :test-not :key} | |
4341 | This is a destructive version of @code{union}; rather than copying, | |
4342 | it tries to reuse the storage of the argument lists if possible. | |
4343 | @end defun | |
4344 | ||
4345 | @defun intersection list1 list2 @t{&key :test :test-not :key} | |
4346 | This function computes the intersection of the sets represented | |
4347 | by @var{list1} and @var{list2}. It returns the list of items | |
4348 | which appear in both @var{list1} and @var{list2}. | |
4349 | @end defun | |
4350 | ||
4351 | @defun nintersection list1 list2 @t{&key :test :test-not :key} | |
4352 | This is a destructive version of @code{intersection}. It | |
4353 | tries to reuse storage of @var{list1} rather than copying. | |
4354 | It does @emph{not} reuse the storage of @var{list2}. | |
4355 | @end defun | |
4356 | ||
4357 | @defun set-difference list1 list2 @t{&key :test :test-not :key} | |
4358 | This function computes the ``set difference'' of @var{list1} | |
4359 | and @var{list2}, i.e., the set of elements that appear in | |
4360 | @var{list1} but @emph{not} in @var{list2}. | |
4361 | @end defun | |
4362 | ||
4363 | @defun nset-difference list1 list2 @t{&key :test :test-not :key} | |
4364 | This is a destructive @code{set-difference}, which will try | |
4365 | to reuse @var{list1} if possible. | |
4366 | @end defun | |
4367 | ||
4368 | @defun set-exclusive-or list1 list2 @t{&key :test :test-not :key} | |
4369 | This function computes the ``set exclusive or'' of @var{list1} | |
4370 | and @var{list2}, i.e., the set of elements that appear in | |
4371 | exactly one of @var{list1} and @var{list2}. | |
4372 | @end defun | |
4373 | ||
4374 | @defun nset-exclusive-or list1 list2 @t{&key :test :test-not :key} | |
4375 | This is a destructive @code{set-exclusive-or}, which will try | |
4376 | to reuse @var{list1} and @var{list2} if possible. | |
4377 | @end defun | |
4378 | ||
4379 | @defun subsetp list1 list2 @t{&key :test :test-not :key} | |
4380 | This function checks whether @var{list1} represents a subset | |
4381 | of @var{list2}, i.e., whether every element of @var{list1} | |
4382 | also appears in @var{list2}. | |
4383 | @end defun | |
4384 | ||
c6ad01e8 | 4385 | @node Association Lists, , Lists as Sets, Lists |
6bf7aab6 DL |
4386 | @section Association Lists |
4387 | ||
4388 | @noindent | |
4389 | An @dfn{association list} is a list representing a mapping from | |
4390 | one set of values to another; any list whose elements are cons | |
4391 | cells is an association list. | |
4392 | ||
4393 | @defun assoc* item a-list @t{&key :test :test-not :key} | |
4394 | This function searches the association list @var{a-list} for an | |
4395 | element whose @code{car} matches (in the sense of @code{:test}, | |
4396 | @code{:test-not}, and @code{:key}, or by comparison with @code{eql}) | |
4397 | a given @var{item}. It returns the matching element, if any, | |
4398 | otherwise @code{nil}. It ignores elements of @var{a-list} which | |
4399 | are not cons cells. (This corresponds to the behavior of | |
4400 | @code{assq} and @code{assoc} in Emacs Lisp; Common Lisp's | |
4401 | @code{assoc} ignores @code{nil}s but considers any other non-cons | |
4402 | elements of @var{a-list} to be an error.) | |
4403 | @end defun | |
4404 | ||
4405 | @defun rassoc* item a-list @t{&key :test :test-not :key} | |
4406 | This function searches for an element whose @code{cdr} matches | |
4407 | @var{item}. If @var{a-list} represents a mapping, this applies | |
4408 | the inverse of the mapping to @var{item}. | |
4409 | @end defun | |
4410 | ||
6bf7aab6 DL |
4411 | @findex assoc-if |
4412 | @findex assoc-if-not | |
4413 | @findex rassoc-if | |
4414 | @findex rassoc-if-not | |
4415 | The @code{assoc-if}, @code{assoc-if-not}, @code{rassoc-if}, | |
4416 | and @code{rassoc-if-not} functions are defined similarly. | |
4417 | ||
4418 | Two simple functions for constructing association lists are: | |
4419 | ||
4420 | @defun acons key value alist | |
4421 | This is equivalent to @code{(cons (cons @var{key} @var{value}) @var{alist})}. | |
4422 | @end defun | |
4423 | ||
4424 | @defun pairlis keys values &optional alist | |
4425 | This is equivalent to @code{(nconc (mapcar* 'cons @var{keys} @var{values}) | |
4426 | @var{alist})}. | |
4427 | @end defun | |
4428 | ||
6bf7aab6 DL |
4429 | @iftex |
4430 | @chapno=18 | |
4431 | @end iftex | |
4432 | ||
c6ad01e8 | 4433 | @node Structures, Assertions, Lists, Top |
6bf7aab6 DL |
4434 | @chapter Structures |
4435 | ||
4436 | @noindent | |
4437 | The Common Lisp @dfn{structure} mechanism provides a general way | |
4438 | to define data types similar to C's @code{struct} types. A | |
4439 | structure is a Lisp object containing some number of @dfn{slots}, | |
4440 | each of which can hold any Lisp data object. Functions are | |
4441 | provided for accessing and setting the slots, creating or copying | |
4442 | structure objects, and recognizing objects of a particular structure | |
4443 | type. | |
4444 | ||
4445 | In true Common Lisp, each structure type is a new type distinct | |
4446 | from all existing Lisp types. Since the underlying Emacs Lisp | |
4447 | system provides no way to create new distinct types, this package | |
4448 | implements structures as vectors (or lists upon request) with a | |
4449 | special ``tag'' symbol to identify them. | |
4450 | ||
4451 | @defspec defstruct name slots@dots{} | |
4452 | The @code{defstruct} form defines a new structure type called | |
4453 | @var{name}, with the specified @var{slots}. (The @var{slots} | |
4454 | may begin with a string which documents the structure type.) | |
4455 | In the simplest case, @var{name} and each of the @var{slots} | |
4456 | are symbols. For example, | |
4457 | ||
4458 | @example | |
4459 | (defstruct person name age sex) | |
4460 | @end example | |
4461 | ||
4462 | @noindent | |
4463 | defines a struct type called @code{person} which contains three | |
4464 | slots. Given a @code{person} object @var{p}, you can access those | |
4465 | slots by calling @code{(person-name @var{p})}, @code{(person-age @var{p})}, | |
4466 | and @code{(person-sex @var{p})}. You can also change these slots by | |
4467 | using @code{setf} on any of these place forms: | |
4468 | ||
4469 | @example | |
4470 | (incf (person-age birthday-boy)) | |
4471 | @end example | |
4472 | ||
4473 | You can create a new @code{person} by calling @code{make-person}, | |
4474 | which takes keyword arguments @code{:name}, @code{:age}, and | |
4475 | @code{:sex} to specify the initial values of these slots in the | |
4476 | new object. (Omitting any of these arguments leaves the corresponding | |
4477 | slot ``undefined,'' according to the Common Lisp standard; in Emacs | |
4478 | Lisp, such uninitialized slots are filled with @code{nil}.) | |
4479 | ||
4480 | Given a @code{person}, @code{(copy-person @var{p})} makes a new | |
4481 | object of the same type whose slots are @code{eq} to those of @var{p}. | |
4482 | ||
4483 | Given any Lisp object @var{x}, @code{(person-p @var{x})} returns | |
4484 | true if @var{x} looks like a @code{person}, false otherwise. (Again, | |
4485 | in Common Lisp this predicate would be exact; in Emacs Lisp the | |
4486 | best it can do is verify that @var{x} is a vector of the correct | |
4487 | length which starts with the correct tag symbol.) | |
4488 | ||
4489 | Accessors like @code{person-name} normally check their arguments | |
4490 | (effectively using @code{person-p}) and signal an error if the | |
4491 | argument is the wrong type. This check is affected by | |
4492 | @code{(optimize (safety @dots{}))} declarations. Safety level 1, | |
4493 | the default, uses a somewhat optimized check that will detect all | |
4494 | incorrect arguments, but may use an uninformative error message | |
4495 | (e.g., ``expected a vector'' instead of ``expected a @code{person}''). | |
4496 | Safety level 0 omits all checks except as provided by the underlying | |
4497 | @code{aref} call; safety levels 2 and 3 do rigorous checking that will | |
4498 | always print a descriptive error message for incorrect inputs. | |
4499 | @xref{Declarations}. | |
4500 | ||
4501 | @example | |
4502 | (setq dave (make-person :name "Dave" :sex 'male)) | |
4503 | @result{} [cl-struct-person "Dave" nil male] | |
4504 | (setq other (copy-person dave)) | |
4505 | @result{} [cl-struct-person "Dave" nil male] | |
4506 | (eq dave other) | |
4507 | @result{} nil | |
4508 | (eq (person-name dave) (person-name other)) | |
4509 | @result{} t | |
4510 | (person-p dave) | |
4511 | @result{} t | |
4512 | (person-p [1 2 3 4]) | |
4513 | @result{} nil | |
4514 | (person-p "Bogus") | |
4515 | @result{} nil | |
4516 | (person-p '[cl-struct-person counterfeit person object]) | |
4517 | @result{} t | |
4518 | @end example | |
4519 | ||
4520 | In general, @var{name} is either a name symbol or a list of a name | |
4521 | symbol followed by any number of @dfn{struct options}; each @var{slot} | |
4522 | is either a slot symbol or a list of the form @samp{(@var{slot-name} | |
4523 | @var{default-value} @var{slot-options}@dots{})}. The @var{default-value} | |
4524 | is a Lisp form which is evaluated any time an instance of the | |
4525 | structure type is created without specifying that slot's value. | |
4526 | ||
4527 | Common Lisp defines several slot options, but the only one | |
4528 | implemented in this package is @code{:read-only}. A non-@code{nil} | |
4529 | value for this option means the slot should not be @code{setf}-able; | |
4530 | the slot's value is determined when the object is created and does | |
4531 | not change afterward. | |
4532 | ||
4533 | @example | |
4534 | (defstruct person | |
4535 | (name nil :read-only t) | |
4536 | age | |
4537 | (sex 'unknown)) | |
4538 | @end example | |
4539 | ||
4540 | Any slot options other than @code{:read-only} are ignored. | |
4541 | ||
4542 | For obscure historical reasons, structure options take a different | |
4543 | form than slot options. A structure option is either a keyword | |
4544 | symbol, or a list beginning with a keyword symbol possibly followed | |
4545 | by arguments. (By contrast, slot options are key-value pairs not | |
4546 | enclosed in lists.) | |
4547 | ||
4548 | @example | |
4549 | (defstruct (person (:constructor create-person) | |
4550 | (:type list) | |
4551 | :named) | |
4552 | name age sex) | |
4553 | @end example | |
4554 | ||
4555 | The following structure options are recognized. | |
4556 | ||
4557 | @table @code | |
4558 | @iftex | |
4559 | @itemmax=0 in | |
4560 | @advance@leftskip-.5@tableindent | |
4561 | @end iftex | |
4562 | @item :conc-name | |
4563 | The argument is a symbol whose print name is used as the prefix for | |
4564 | the names of slot accessor functions. The default is the name of | |
4565 | the struct type followed by a hyphen. The option @code{(:conc-name p-)} | |
4566 | would change this prefix to @code{p-}. Specifying @code{nil} as an | |
4567 | argument means no prefix, so that the slot names themselves are used | |
4568 | to name the accessor functions. | |
4569 | ||
4570 | @item :constructor | |
4571 | In the simple case, this option takes one argument which is an | |
4572 | alternate name to use for the constructor function. The default | |
4573 | is @code{make-@var{name}}, e.g., @code{make-person}. The above | |
4574 | example changes this to @code{create-person}. Specifying @code{nil} | |
4575 | as an argument means that no standard constructor should be | |
4576 | generated at all. | |
4577 | ||
4578 | In the full form of this option, the constructor name is followed | |
4579 | by an arbitrary argument list. @xref{Program Structure}, for a | |
4580 | description of the format of Common Lisp argument lists. All | |
4581 | options, such as @code{&rest} and @code{&key}, are supported. | |
4582 | The argument names should match the slot names; each slot is | |
4583 | initialized from the corresponding argument. Slots whose names | |
4584 | do not appear in the argument list are initialized based on the | |
4585 | @var{default-value} in their slot descriptor. Also, @code{&optional} | |
4586 | and @code{&key} arguments which don't specify defaults take their | |
2cbd16b9 | 4587 | defaults from the slot descriptor. It is valid to include arguments |
6bf7aab6 DL |
4588 | which don't correspond to slot names; these are useful if they are |
4589 | referred to in the defaults for optional, keyword, or @code{&aux} | |
4590 | arguments which @emph{do} correspond to slots. | |
4591 | ||
4592 | You can specify any number of full-format @code{:constructor} | |
4593 | options on a structure. The default constructor is still generated | |
4594 | as well unless you disable it with a simple-format @code{:constructor} | |
4595 | option. | |
4596 | ||
4597 | @example | |
4598 | (defstruct | |
4599 | (person | |
4600 | (:constructor nil) ; no default constructor | |
4601 | (:constructor new-person (name sex &optional (age 0))) | |
4602 | (:constructor new-hound (&key (name "Rover") | |
4603 | (dog-years 0) | |
4604 | &aux (age (* 7 dog-years)) | |
4605 | (sex 'canine)))) | |
4606 | name age sex) | |
4607 | @end example | |
4608 | ||
4609 | The first constructor here takes its arguments positionally rather | |
4610 | than by keyword. (In official Common Lisp terminology, constructors | |
4611 | that work By Order of Arguments instead of by keyword are called | |
4612 | ``BOA constructors.'' No, I'm not making this up.) For example, | |
4613 | @code{(new-person "Jane" 'female)} generates a person whose slots | |
4614 | are @code{"Jane"}, 0, and @code{female}, respectively. | |
4615 | ||
4616 | The second constructor takes two keyword arguments, @code{:name}, | |
4617 | which initializes the @code{name} slot and defaults to @code{"Rover"}, | |
4618 | and @code{:dog-years}, which does not itself correspond to a slot | |
4619 | but which is used to initialize the @code{age} slot. The @code{sex} | |
4620 | slot is forced to the symbol @code{canine} with no syntax for | |
4621 | overriding it. | |
4622 | ||
4623 | @item :copier | |
4624 | The argument is an alternate name for the copier function for | |
4625 | this type. The default is @code{copy-@var{name}}. @code{nil} | |
4626 | means not to generate a copier function. (In this implementation, | |
4627 | all copier functions are simply synonyms for @code{copy-sequence}.) | |
4628 | ||
4629 | @item :predicate | |
4630 | The argument is an alternate name for the predicate which recognizes | |
4631 | objects of this type. The default is @code{@var{name}-p}. @code{nil} | |
4632 | means not to generate a predicate function. (If the @code{:type} | |
4633 | option is used without the @code{:named} option, no predicate is | |
4634 | ever generated.) | |
4635 | ||
4636 | In true Common Lisp, @code{typep} is always able to recognize a | |
4637 | structure object even if @code{:predicate} was used. In this | |
4638 | package, @code{typep} simply looks for a function called | |
4639 | @code{@var{typename}-p}, so it will work for structure types | |
4640 | only if they used the default predicate name. | |
4641 | ||
4642 | @item :include | |
4643 | This option implements a very limited form of C++-style inheritance. | |
4644 | The argument is the name of another structure type previously | |
4645 | created with @code{defstruct}. The effect is to cause the new | |
4646 | structure type to inherit all of the included structure's slots | |
4647 | (plus, of course, any new slots described by this struct's slot | |
4648 | descriptors). The new structure is considered a ``specialization'' | |
4649 | of the included one. In fact, the predicate and slot accessors | |
4650 | for the included type will also accept objects of the new type. | |
4651 | ||
4652 | If there are extra arguments to the @code{:include} option after | |
4653 | the included-structure name, these options are treated as replacement | |
4654 | slot descriptors for slots in the included structure, possibly with | |
4655 | modified default values. Borrowing an example from Steele: | |
4656 | ||
4657 | @example | |
4658 | (defstruct person name (age 0) sex) | |
4659 | @result{} person | |
4660 | (defstruct (astronaut (:include person (age 45))) | |
4661 | helmet-size | |
4662 | (favorite-beverage 'tang)) | |
4663 | @result{} astronaut | |
4664 | ||
4665 | (setq joe (make-person :name "Joe")) | |
4666 | @result{} [cl-struct-person "Joe" 0 nil] | |
4667 | (setq buzz (make-astronaut :name "Buzz")) | |
4668 | @result{} [cl-struct-astronaut "Buzz" 45 nil nil tang] | |
4669 | ||
4670 | (list (person-p joe) (person-p buzz)) | |
4671 | @result{} (t t) | |
4672 | (list (astronaut-p joe) (astronaut-p buzz)) | |
4673 | @result{} (nil t) | |
4674 | ||
4675 | (person-name buzz) | |
4676 | @result{} "Buzz" | |
4677 | (astronaut-name joe) | |
4678 | @result{} error: "astronaut-name accessing a non-astronaut" | |
4679 | @end example | |
4680 | ||
4681 | Thus, if @code{astronaut} is a specialization of @code{person}, | |
4682 | then every @code{astronaut} is also a @code{person} (but not the | |
4683 | other way around). Every @code{astronaut} includes all the slots | |
4684 | of a @code{person}, plus extra slots that are specific to | |
4685 | astronauts. Operations that work on people (like @code{person-name}) | |
4686 | work on astronauts just like other people. | |
4687 | ||
4688 | @item :print-function | |
4689 | In full Common Lisp, this option allows you to specify a function | |
4690 | which is called to print an instance of the structure type. The | |
4691 | Emacs Lisp system offers no hooks into the Lisp printer which would | |
4692 | allow for such a feature, so this package simply ignores | |
4693 | @code{:print-function}. | |
4694 | ||
4695 | @item :type | |
4696 | The argument should be one of the symbols @code{vector} or @code{list}. | |
4697 | This tells which underlying Lisp data type should be used to implement | |
4698 | the new structure type. Vectors are used by default, but | |
4699 | @code{(:type list)} will cause structure objects to be stored as | |
4700 | lists instead. | |
4701 | ||
4702 | The vector representation for structure objects has the advantage | |
4703 | that all structure slots can be accessed quickly, although creating | |
4704 | vectors is a bit slower in Emacs Lisp. Lists are easier to create, | |
4705 | but take a relatively long time accessing the later slots. | |
4706 | ||
4707 | @item :named | |
4708 | This option, which takes no arguments, causes a characteristic ``tag'' | |
4709 | symbol to be stored at the front of the structure object. Using | |
4710 | @code{:type} without also using @code{:named} will result in a | |
4711 | structure type stored as plain vectors or lists with no identifying | |
4712 | features. | |
4713 | ||
4714 | The default, if you don't specify @code{:type} explicitly, is to | |
4715 | use named vectors. Therefore, @code{:named} is only useful in | |
4716 | conjunction with @code{:type}. | |
4717 | ||
4718 | @example | |
4719 | (defstruct (person1) name age sex) | |
4720 | (defstruct (person2 (:type list) :named) name age sex) | |
4721 | (defstruct (person3 (:type list)) name age sex) | |
4722 | ||
4723 | (setq p1 (make-person1)) | |
4724 | @result{} [cl-struct-person1 nil nil nil] | |
4725 | (setq p2 (make-person2)) | |
4726 | @result{} (person2 nil nil nil) | |
4727 | (setq p3 (make-person3)) | |
4728 | @result{} (nil nil nil) | |
4729 | ||
4730 | (person1-p p1) | |
4731 | @result{} t | |
4732 | (person2-p p2) | |
4733 | @result{} t | |
4734 | (person3-p p3) | |
4735 | @result{} error: function person3-p undefined | |
4736 | @end example | |
4737 | ||
4738 | Since unnamed structures don't have tags, @code{defstruct} is not | |
4739 | able to make a useful predicate for recognizing them. Also, | |
4740 | accessors like @code{person3-name} will be generated but they | |
4741 | will not be able to do any type checking. The @code{person3-name} | |
4742 | function, for example, will simply be a synonym for @code{car} in | |
4743 | this case. By contrast, @code{person2-name} is able to verify | |
4744 | that its argument is indeed a @code{person2} object before | |
4745 | proceeding. | |
4746 | ||
4747 | @item :initial-offset | |
4748 | The argument must be a nonnegative integer. It specifies a | |
4749 | number of slots to be left ``empty'' at the front of the | |
4750 | structure. If the structure is named, the tag appears at the | |
4751 | specified position in the list or vector; otherwise, the first | |
4752 | slot appears at that position. Earlier positions are filled | |
4753 | with @code{nil} by the constructors and ignored otherwise. If | |
4754 | the type @code{:include}s another type, then @code{:initial-offset} | |
4755 | specifies a number of slots to be skipped between the last slot | |
4756 | of the included type and the first new slot. | |
4757 | @end table | |
4758 | @end defspec | |
4759 | ||
4760 | Except as noted, the @code{defstruct} facility of this package is | |
4761 | entirely compatible with that of Common Lisp. | |
4762 | ||
4763 | @iftex | |
4764 | @chapno=23 | |
4765 | @end iftex | |
4766 | ||
4767 | @node Assertions, Efficiency Concerns, Structures, Top | |
4768 | @chapter Assertions and Errors | |
4769 | ||
4770 | @noindent | |
4771 | This section describes two macros that test @dfn{assertions}, i.e., | |
4772 | conditions which must be true if the program is operating correctly. | |
4773 | Assertions never add to the behavior of a Lisp program; they simply | |
4774 | make ``sanity checks'' to make sure everything is as it should be. | |
4775 | ||
4776 | If the optimization property @code{speed} has been set to 3, and | |
4777 | @code{safety} is less than 3, then the byte-compiler will optimize | |
4778 | away the following assertions. Because assertions might be optimized | |
4779 | away, it is a bad idea for them to include side-effects. | |
4780 | ||
4781 | @defspec assert test-form [show-args string args@dots{}] | |
4782 | This form verifies that @var{test-form} is true (i.e., evaluates to | |
4783 | a non-@code{nil} value). If so, it returns @code{nil}. If the test | |
4784 | is not satisfied, @code{assert} signals an error. | |
4785 | ||
4786 | A default error message will be supplied which includes @var{test-form}. | |
4787 | You can specify a different error message by including a @var{string} | |
4788 | argument plus optional extra arguments. Those arguments are simply | |
4789 | passed to @code{error} to signal the error. | |
4790 | ||
4791 | If the optional second argument @var{show-args} is @code{t} instead | |
4792 | of @code{nil}, then the error message (with or without @var{string}) | |
4793 | will also include all non-constant arguments of the top-level | |
4794 | @var{form}. For example: | |
4795 | ||
4796 | @example | |
4797 | (assert (> x 10) t "x is too small: %d") | |
4798 | @end example | |
4799 | ||
4800 | This usage of @var{show-args} is an extension to Common Lisp. In | |
4801 | true Common Lisp, the second argument gives a list of @var{places} | |
4802 | which can be @code{setf}'d by the user before continuing from the | |
4803 | error. Since Emacs Lisp does not support continuable errors, it | |
4804 | makes no sense to specify @var{places}. | |
4805 | @end defspec | |
4806 | ||
4807 | @defspec check-type form type [string] | |
4808 | This form verifies that @var{form} evaluates to a value of type | |
4809 | @var{type}. If so, it returns @code{nil}. If not, @code{check-type} | |
4810 | signals a @code{wrong-type-argument} error. The default error message | |
4811 | lists the erroneous value along with @var{type} and @var{form} | |
4812 | themselves. If @var{string} is specified, it is included in the | |
4813 | error message in place of @var{type}. For example: | |
4814 | ||
4815 | @example | |
4816 | (check-type x (integer 1 *) "a positive integer") | |
4817 | @end example | |
4818 | ||
4819 | @xref{Type Predicates}, for a description of the type specifiers | |
4820 | that may be used for @var{type}. | |
4821 | ||
4822 | Note that in Common Lisp, the first argument to @code{check-type} | |
4823 | must be a @var{place} suitable for use by @code{setf}, because | |
4824 | @code{check-type} signals a continuable error that allows the | |
4825 | user to modify @var{place}. | |
4826 | @end defspec | |
4827 | ||
4828 | The following error-related macro is also defined: | |
4829 | ||
4830 | @defspec ignore-errors forms@dots{} | |
4831 | This executes @var{forms} exactly like a @code{progn}, except that | |
4832 | errors are ignored during the @var{forms}. More precisely, if | |
4833 | an error is signaled then @code{ignore-errors} immediately | |
4834 | aborts execution of the @var{forms} and returns @code{nil}. | |
4835 | If the @var{forms} complete successfully, @code{ignore-errors} | |
4836 | returns the result of the last @var{form}. | |
4837 | @end defspec | |
4838 | ||
4839 | @node Efficiency Concerns, Common Lisp Compatibility, Assertions, Top | |
4840 | @appendix Efficiency Concerns | |
4841 | ||
4842 | @appendixsec Macros | |
4843 | ||
4844 | @noindent | |
4845 | Many of the advanced features of this package, such as @code{defun*}, | |
4846 | @code{loop}, and @code{setf}, are implemented as Lisp macros. In | |
4847 | byte-compiled code, these complex notations will be expanded into | |
4848 | equivalent Lisp code which is simple and efficient. For example, | |
4849 | the forms | |
4850 | ||
4851 | @example | |
4852 | (incf i n) | |
4853 | (push x (car p)) | |
4854 | @end example | |
4855 | ||
4856 | @noindent | |
4857 | are expanded at compile-time to the Lisp forms | |
4858 | ||
4859 | @example | |
4860 | (setq i (+ i n)) | |
4861 | (setcar p (cons x (car p))) | |
4862 | @end example | |
4863 | ||
4864 | @noindent | |
4865 | which are the most efficient ways of doing these respective operations | |
4866 | in Lisp. Thus, there is no performance penalty for using the more | |
4867 | readable @code{incf} and @code{push} forms in your compiled code. | |
4868 | ||
4869 | @emph{Interpreted} code, on the other hand, must expand these macros | |
4870 | every time they are executed. For this reason it is strongly | |
4871 | recommended that code making heavy use of macros be compiled. | |
4872 | (The features labeled ``Special Form'' instead of ``Function'' in | |
4873 | this manual are macros.) A loop using @code{incf} a hundred times | |
4874 | will execute considerably faster if compiled, and will also | |
4875 | garbage-collect less because the macro expansion will not have | |
4876 | to be generated, used, and thrown away a hundred times. | |
4877 | ||
4878 | You can find out how a macro expands by using the | |
4879 | @code{cl-prettyexpand} function. | |
4880 | ||
4881 | @defun cl-prettyexpand form &optional full | |
4882 | This function takes a single Lisp form as an argument and inserts | |
4883 | a nicely formatted copy of it in the current buffer (which must be | |
4884 | in Lisp mode so that indentation works properly). It also expands | |
4885 | all Lisp macros which appear in the form. The easiest way to use | |
4886 | this function is to go to the @code{*scratch*} buffer and type, say, | |
4887 | ||
4888 | @example | |
4889 | (cl-prettyexpand '(loop for x below 10 collect x)) | |
4890 | @end example | |
4891 | ||
4892 | @noindent | |
4893 | and type @kbd{C-x C-e} immediately after the closing parenthesis; | |
4894 | the expansion | |
4895 | ||
4896 | @example | |
4897 | (block nil | |
4898 | (let* ((x 0) | |
4899 | (G1004 nil)) | |
4900 | (while (< x 10) | |
4901 | (setq G1004 (cons x G1004)) | |
4902 | (setq x (+ x 1))) | |
4903 | (nreverse G1004))) | |
4904 | @end example | |
4905 | ||
4906 | @noindent | |
4907 | will be inserted into the buffer. (The @code{block} macro is | |
4908 | expanded differently in the interpreter and compiler, so | |
4909 | @code{cl-prettyexpand} just leaves it alone. The temporary | |
4910 | variable @code{G1004} was created by @code{gensym}.) | |
4911 | ||
4912 | If the optional argument @var{full} is true, then @emph{all} | |
4913 | macros are expanded, including @code{block}, @code{eval-when}, | |
4914 | and compiler macros. Expansion is done as if @var{form} were | |
4915 | a top-level form in a file being compiled. For example, | |
4916 | ||
4917 | @example | |
4918 | (cl-prettyexpand '(pushnew 'x list)) | |
4919 | @print{} (setq list (adjoin 'x list)) | |
4920 | (cl-prettyexpand '(pushnew 'x list) t) | |
4921 | @print{} (setq list (if (memq 'x list) list (cons 'x list))) | |
4922 | (cl-prettyexpand '(caddr (member* 'a list)) t) | |
4923 | @print{} (car (cdr (cdr (memq 'a list)))) | |
4924 | @end example | |
4925 | ||
4926 | Note that @code{adjoin}, @code{caddr}, and @code{member*} all | |
4927 | have built-in compiler macros to optimize them in common cases. | |
4928 | @end defun | |
4929 | ||
4930 | @ifinfo | |
4931 | @example | |
4932 | ||
4933 | @end example | |
4934 | @end ifinfo | |
4935 | @appendixsec Error Checking | |
4936 | ||
4937 | @noindent | |
4938 | Common Lisp compliance has in general not been sacrificed for the | |
4939 | sake of efficiency. A few exceptions have been made for cases | |
4940 | where substantial gains were possible at the expense of marginal | |
a9c498a5 | 4941 | incompatibility. |
6bf7aab6 DL |
4942 | |
4943 | The Common Lisp standard (as embodied in Steele's book) uses the | |
4944 | phrase ``it is an error if'' to indicate a situation which is not | |
4945 | supposed to arise in complying programs; implementations are strongly | |
4946 | encouraged but not required to signal an error in these situations. | |
4947 | This package sometimes omits such error checking in the interest of | |
4948 | compactness and efficiency. For example, @code{do} variable | |
4949 | specifiers are supposed to be lists of one, two, or three forms; | |
4950 | extra forms are ignored by this package rather than signaling a | |
4951 | syntax error. The @code{endp} function is simply a synonym for | |
4952 | @code{null} in this package. Functions taking keyword arguments | |
4953 | will accept an odd number of arguments, treating the trailing | |
4954 | keyword as if it were followed by the value @code{nil}. | |
4955 | ||
4956 | Argument lists (as processed by @code{defun*} and friends) | |
4957 | @emph{are} checked rigorously except for the minor point just | |
4958 | mentioned; in particular, keyword arguments are checked for | |
4959 | validity, and @code{&allow-other-keys} and @code{:allow-other-keys} | |
4960 | are fully implemented. Keyword validity checking is slightly | |
4961 | time consuming (though not too bad in byte-compiled code); | |
4962 | you can use @code{&allow-other-keys} to omit this check. Functions | |
4963 | defined in this package such as @code{find} and @code{member*} | |
4964 | do check their keyword arguments for validity. | |
4965 | ||
4966 | @ifinfo | |
4967 | @example | |
4968 | ||
4969 | @end example | |
4970 | @end ifinfo | |
4971 | @appendixsec Optimizing Compiler | |
4972 | ||
4973 | @noindent | |
c6ad01e8 | 4974 | Use of the optimizing Emacs compiler is highly recommended; many of the Common |
ed4d9494 | 4975 | Lisp macros emit |
6bf7aab6 DL |
4976 | code which can be improved by optimization. In particular, |
4977 | @code{block}s (whether explicit or implicit in constructs like | |
4978 | @code{defun*} and @code{loop}) carry a fair run-time penalty; the | |
4979 | optimizing compiler removes @code{block}s which are not actually | |
4980 | referenced by @code{return} or @code{return-from} inside the block. | |
4981 | ||
4982 | @node Common Lisp Compatibility, Old CL Compatibility, Efficiency Concerns, Top | |
4983 | @appendix Common Lisp Compatibility | |
4984 | ||
4985 | @noindent | |
4986 | Following is a list of all known incompatibilities between this | |
4987 | package and Common Lisp as documented in Steele (2nd edition). | |
4988 | ||
4989 | Certain function names, such as @code{member}, @code{assoc}, and | |
4990 | @code{floor}, were already taken by (incompatible) Emacs Lisp | |
4991 | functions; this package appends @samp{*} to the names of its | |
4992 | Common Lisp versions of these functions. | |
4993 | ||
4994 | The word @code{defun*} is required instead of @code{defun} in order | |
4995 | to use extended Common Lisp argument lists in a function. Likewise, | |
4996 | @code{defmacro*} and @code{function*} are versions of those forms | |
4997 | which understand full-featured argument lists. The @code{&whole} | |
4998 | keyword does not work in @code{defmacro} argument lists (except | |
4999 | inside recursive argument lists). | |
5000 | ||
6bf7aab6 DL |
5001 | The @code{eql} and @code{equal} predicates do not distinguish |
5002 | between IEEE floating-point plus and minus zero. The @code{equalp} | |
5003 | predicate has several differences with Common Lisp; @pxref{Predicates}. | |
5004 | ||
5005 | The @code{setf} mechanism is entirely compatible, except that | |
5006 | setf-methods return a list of five values rather than five | |
5007 | values directly. Also, the new ``@code{setf} function'' concept | |
5008 | (typified by @code{(defun (setf foo) @dots{})}) is not implemented. | |
5009 | ||
5010 | The @code{do-all-symbols} form is the same as @code{do-symbols} | |
5011 | with no @var{obarray} argument. In Common Lisp, this form would | |
5012 | iterate over all symbols in all packages. Since Emacs obarrays | |
5013 | are not a first-class package mechanism, there is no way for | |
5014 | @code{do-all-symbols} to locate any but the default obarray. | |
5015 | ||
5016 | The @code{loop} macro is complete except that @code{loop-finish} | |
5017 | and type specifiers are unimplemented. | |
5018 | ||
5019 | The multiple-value return facility treats lists as multiple | |
5020 | values, since Emacs Lisp cannot support multiple return values | |
5021 | directly. The macros will be compatible with Common Lisp if | |
5022 | @code{values} or @code{values-list} is always used to return to | |
5023 | a @code{multiple-value-bind} or other multiple-value receiver; | |
5024 | if @code{values} is used without @code{multiple-value-@dots{}} | |
5025 | or vice-versa the effect will be different from Common Lisp. | |
5026 | ||
5027 | Many Common Lisp declarations are ignored, and others match | |
5028 | the Common Lisp standard in concept but not in detail. For | |
5029 | example, local @code{special} declarations, which are purely | |
5030 | advisory in Emacs Lisp, do not rigorously obey the scoping rules | |
5031 | set down in Steele's book. | |
5032 | ||
5033 | The variable @code{*gensym-counter*} starts out with a pseudo-random | |
5034 | value rather than with zero. This is to cope with the fact that | |
5035 | generated symbols become interned when they are written to and | |
5036 | loaded back from a file. | |
5037 | ||
5038 | The @code{defstruct} facility is compatible, except that structures | |
5039 | are of type @code{:type vector :named} by default rather than some | |
5040 | special, distinct type. Also, the @code{:type} slot option is ignored. | |
5041 | ||
5042 | The second argument of @code{check-type} is treated differently. | |
5043 | ||
5044 | @node Old CL Compatibility, Porting Common Lisp, Common Lisp Compatibility, Top | |
5045 | @appendix Old CL Compatibility | |
5046 | ||
5047 | @noindent | |
5048 | Following is a list of all known incompatibilities between this package | |
5049 | and the older Quiroz @file{cl.el} package. | |
5050 | ||
5051 | This package's emulation of multiple return values in functions is | |
5052 | incompatible with that of the older package. That package attempted | |
5053 | to come as close as possible to true Common Lisp multiple return | |
5054 | values; unfortunately, it could not be 100% reliable and so was prone | |
5055 | to occasional surprises if used freely. This package uses a simpler | |
5056 | method, namely replacing multiple values with lists of values, which | |
5057 | is more predictable though more noticeably different from Common Lisp. | |
5058 | ||
5059 | The @code{defkeyword} form and @code{keywordp} function are not | |
5060 | implemented in this package. | |
5061 | ||
5062 | The @code{member}, @code{floor}, @code{ceiling}, @code{truncate}, | |
5063 | @code{round}, @code{mod}, and @code{rem} functions are suffixed | |
5064 | by @samp{*} in this package to avoid collision with existing | |
ed4d9494 | 5065 | functions in Emacs. The older package simply |
6bf7aab6 | 5066 | redefined these functions, overwriting the built-in meanings and |
c6ad01e8 | 5067 | causing serious portability problems. (Some more |
6bf7aab6 DL |
5068 | recent versions of the Quiroz package changed the names to |
5069 | @code{cl-member}, etc.; this package defines the latter names as | |
5070 | aliases for @code{member*}, etc.) | |
5071 | ||
5072 | Certain functions in the old package which were buggy or inconsistent | |
5073 | with the Common Lisp standard are incompatible with the conforming | |
5074 | versions in this package. For example, @code{eql} and @code{member} | |
5075 | were synonyms for @code{eq} and @code{memq} in that package, @code{setf} | |
5076 | failed to preserve correct order of evaluation of its arguments, etc. | |
5077 | ||
5078 | Finally, unlike the older package, this package is careful to | |
5079 | prefix all of its internal names with @code{cl-}. Except for a | |
5080 | few functions which are explicitly defined as additional features | |
5081 | (such as @code{floatp-safe} and @code{letf}), this package does not | |
5082 | export any non-@samp{cl-} symbols which are not also part of Common | |
5083 | Lisp. | |
5084 | ||
5085 | @ifinfo | |
5086 | @example | |
5087 | ||
5088 | @end example | |
5089 | @end ifinfo | |
5090 | @appendixsec The @code{cl-compat} package | |
5091 | ||
5092 | @noindent | |
5093 | The @dfn{CL} package includes emulations of some features of the | |
5094 | old @file{cl.el}, in the form of a compatibility package | |
5095 | @code{cl-compat}. To use it, put @code{(require 'cl-compat)} in | |
5096 | your program. | |
5097 | ||
5098 | The old package defined a number of internal routines without | |
5099 | @code{cl-} prefixes or other annotations. Call to these routines | |
5100 | may have crept into existing Lisp code. @code{cl-compat} | |
5101 | provides emulations of the following internal routines: | |
5102 | @code{pair-with-newsyms}, @code{zip-lists}, @code{unzip-lists}, | |
5103 | @code{reassemble-arglists}, @code{duplicate-symbols-p}, | |
5104 | @code{safe-idiv}. | |
5105 | ||
5106 | Some @code{setf} forms translated into calls to internal | |
5107 | functions that user code might call directly. The functions | |
5108 | @code{setnth}, @code{setnthcdr}, and @code{setelt} fall in | |
5109 | this category; they are defined by @code{cl-compat}, but the | |
5110 | best fix is to change to use @code{setf} properly. | |
5111 | ||
5112 | The @code{cl-compat} file defines the keyword functions | |
5113 | @code{keywordp}, @code{keyword-of}, and @code{defkeyword}, | |
5114 | which are not defined by the new @dfn{CL} package because the | |
5115 | use of keywords as data is discouraged. | |
5116 | ||
5117 | The @code{build-klist} mechanism for parsing keyword arguments | |
5118 | is emulated by @code{cl-compat}; the @code{with-keyword-args} | |
5119 | macro is not, however, and in any case it's best to change to | |
5120 | use the more natural keyword argument processing offered by | |
5121 | @code{defun*}. | |
5122 | ||
5123 | Multiple return values are treated differently by the two | |
5124 | Common Lisp packages. The old package's method was more | |
5125 | compatible with true Common Lisp, though it used heuristics | |
5126 | that caused it to report spurious multiple return values in | |
5127 | certain cases. The @code{cl-compat} package defines a set | |
5128 | of multiple-value macros that are compatible with the old | |
5129 | CL package; again, they are heuristic in nature, but they | |
5130 | are guaranteed to work in any case where the old package's | |
5131 | macros worked. To avoid name collision with the ``official'' | |
5132 | multiple-value facilities, the ones in @code{cl-compat} have | |
5133 | capitalized names: @code{Values}, @code{Values-list}, | |
5134 | @code{Multiple-value-bind}, etc. | |
5135 | ||
5136 | The functions @code{cl-floor}, @code{cl-ceiling}, @code{cl-truncate}, | |
5137 | and @code{cl-round} are defined by @code{cl-compat} to use the | |
5138 | old-style multiple-value mechanism, just as they did in the old | |
5139 | package. The newer @code{floor*} and friends return their two | |
5140 | results in a list rather than as multiple values. Note that | |
5141 | older versions of the old package used the unadorned names | |
5142 | @code{floor}, @code{ceiling}, etc.; @code{cl-compat} cannot use | |
c6ad01e8 | 5143 | these names because they conflict with Emacs built-ins. |
6bf7aab6 | 5144 | |
84247bb5 | 5145 | @node Porting Common Lisp, GNU Free Documentation License, Old CL Compatibility, Top |
6bf7aab6 DL |
5146 | @appendix Porting Common Lisp |
5147 | ||
5148 | @noindent | |
5149 | This package is meant to be used as an extension to Emacs Lisp, | |
5150 | not as an Emacs implementation of true Common Lisp. Some of the | |
5151 | remaining differences between Emacs Lisp and Common Lisp make it | |
5152 | difficult to port large Common Lisp applications to Emacs. For | |
5153 | one, some of the features in this package are not fully compliant | |
5154 | with ANSI or Steele; @pxref{Common Lisp Compatibility}. But there | |
5155 | are also quite a few features that this package does not provide | |
4f79343e | 5156 | at all. Here are some major omissions that you will want to watch out |
6bf7aab6 DL |
5157 | for when bringing Common Lisp code into Emacs. |
5158 | ||
5159 | @itemize @bullet | |
5160 | @item | |
5161 | Case-insensitivity. Symbols in Common Lisp are case-insensitive | |
5162 | by default. Some programs refer to a function or variable as | |
5163 | @code{foo} in one place and @code{Foo} or @code{FOO} in another. | |
5164 | Emacs Lisp will treat these as three distinct symbols. | |
5165 | ||
5166 | Some Common Lisp code is written entirely in upper case. While Emacs | |
5167 | is happy to let the program's own functions and variables use | |
5168 | this convention, calls to Lisp builtins like @code{if} and | |
5169 | @code{defun} will have to be changed to lower case. | |
5170 | ||
5171 | @item | |
5172 | Lexical scoping. In Common Lisp, function arguments and @code{let} | |
5173 | bindings apply only to references physically within their bodies | |
5174 | (or within macro expansions in their bodies). Emacs Lisp, by | |
5175 | contrast, uses @dfn{dynamic scoping} wherein a binding to a | |
5176 | variable is visible even inside functions called from the body. | |
5177 | ||
5178 | Variables in Common Lisp can be made dynamically scoped by | |
5179 | declaring them @code{special} or using @code{defvar}. In Emacs | |
5180 | Lisp it is as if all variables were declared @code{special}. | |
5181 | ||
5182 | Often you can use code that was written for lexical scoping | |
5183 | even in a dynamically scoped Lisp, but not always. Here is | |
5184 | an example of a Common Lisp code fragment that would fail in | |
5185 | Emacs Lisp: | |
5186 | ||
5187 | @example | |
5188 | (defun map-odd-elements (func list) | |
5189 | (loop for x in list | |
5190 | for flag = t then (not flag) | |
5191 | collect (if flag x (funcall func x)))) | |
5192 | ||
5193 | (defun add-odd-elements (list x) | |
bed44076 | 5194 | (map-odd-elements (lambda (a) (+ a x))) list) |
6bf7aab6 DL |
5195 | @end example |
5196 | ||
5197 | @noindent | |
5198 | In Common Lisp, the two functions' usages of @code{x} are completely | |
5199 | independent. In Emacs Lisp, the binding to @code{x} made by | |
5200 | @code{add-odd-elements} will have been hidden by the binding | |
5201 | in @code{map-odd-elements} by the time the @code{(+ a x)} function | |
5202 | is called. | |
5203 | ||
5204 | (This package avoids such problems in its own mapping functions | |
5205 | by using names like @code{cl-x} instead of @code{x} internally; | |
5206 | as long as you don't use the @code{cl-} prefix for your own | |
5207 | variables no collision can occur.) | |
5208 | ||
5209 | @xref{Lexical Bindings}, for a description of the @code{lexical-let} | |
5210 | form which establishes a Common Lisp-style lexical binding, and some | |
5211 | examples of how it differs from Emacs' regular @code{let}. | |
5212 | ||
5213 | @item | |
5214 | Reader macros. Common Lisp includes a second type of macro that | |
5215 | works at the level of individual characters. For example, Common | |
5216 | Lisp implements the quote notation by a reader macro called @code{'}, | |
5217 | whereas Emacs Lisp's parser just treats quote as a special case. | |
5218 | Some Lisp packages use reader macros to create special syntaxes | |
5219 | for themselves, which the Emacs parser is incapable of reading. | |
5220 | ||
5221 | The lack of reader macros, incidentally, is the reason behind | |
5222 | Emacs Lisp's unusual backquote syntax. Since backquotes are | |
5223 | implemented as a Lisp package and not built-in to the Emacs | |
5224 | parser, they are forced to use a regular macro named @code{`} | |
5225 | which is used with the standard function/macro call notation. | |
5226 | ||
5227 | @item | |
5228 | Other syntactic features. Common Lisp provides a number of | |
5229 | notations beginning with @code{#} that the Emacs Lisp parser | |
5230 | won't understand. For example, @samp{#| ... |#} is an | |
5231 | alternate comment notation, and @samp{#+lucid (foo)} tells | |
5232 | the parser to ignore the @code{(foo)} except in Lucid Common | |
5233 | Lisp. | |
5234 | ||
5235 | @item | |
5236 | Packages. In Common Lisp, symbols are divided into @dfn{packages}. | |
5237 | Symbols that are Lisp built-ins are typically stored in one package; | |
5238 | symbols that are vendor extensions are put in another, and each | |
5239 | application program would have a package for its own symbols. | |
5240 | Certain symbols are ``exported'' by a package and others are | |
5241 | internal; certain packages ``use'' or import the exported symbols | |
5242 | of other packages. To access symbols that would not normally be | |
5243 | visible due to this importing and exporting, Common Lisp provides | |
5244 | a syntax like @code{package:symbol} or @code{package::symbol}. | |
5245 | ||
5246 | Emacs Lisp has a single namespace for all interned symbols, and | |
5247 | then uses a naming convention of putting a prefix like @code{cl-} | |
5248 | in front of the name. Some Emacs packages adopt the Common Lisp-like | |
5249 | convention of using @code{cl:} or @code{cl::} as the prefix. | |
5250 | However, the Emacs parser does not understand colons and just | |
5251 | treats them as part of the symbol name. Thus, while @code{mapcar} | |
5252 | and @code{lisp:mapcar} may refer to the same symbol in Common | |
5253 | Lisp, they are totally distinct in Emacs Lisp. Common Lisp | |
5254 | programs which refer to a symbol by the full name sometimes | |
5255 | and the short name other times will not port cleanly to Emacs. | |
5256 | ||
5257 | Emacs Lisp does have a concept of ``obarrays,'' which are | |
5258 | package-like collections of symbols, but this feature is not | |
5259 | strong enough to be used as a true package mechanism. | |
5260 | ||
5261 | @item | |
5262 | The @code{format} function is quite different between Common | |
5263 | Lisp and Emacs Lisp. It takes an additional ``destination'' | |
5264 | argument before the format string. A destination of @code{nil} | |
5265 | means to format to a string as in Emacs Lisp; a destination | |
5266 | of @code{t} means to write to the terminal (similar to | |
5267 | @code{message} in Emacs). Also, format control strings are | |
5268 | utterly different; @code{~} is used instead of @code{%} to | |
5269 | introduce format codes, and the set of available codes is | |
5270 | much richer. There are no notations like @code{\n} for | |
5271 | string literals; instead, @code{format} is used with the | |
5272 | ``newline'' format code, @code{~%}. More advanced formatting | |
5273 | codes provide such features as paragraph filling, case | |
5274 | conversion, and even loops and conditionals. | |
5275 | ||
5276 | While it would have been possible to implement most of Common | |
5277 | Lisp @code{format} in this package (under the name @code{format*}, | |
5278 | of course), it was not deemed worthwhile. It would have required | |
5279 | a huge amount of code to implement even a decent subset of | |
5280 | @code{format*}, yet the functionality it would provide over | |
5281 | Emacs Lisp's @code{format} would rarely be useful. | |
5282 | ||
5283 | @item | |
5284 | Vector constants use square brackets in Emacs Lisp, but | |
5285 | @code{#(a b c)} notation in Common Lisp. To further complicate | |
c6ad01e8 | 5286 | matters, Emacs has its own @code{#(} notation for |
6bf7aab6 DL |
5287 | something entirely different---strings with properties. |
5288 | ||
5289 | @item | |
5290 | Characters are distinct from integers in Common Lisp. The | |
5291 | notation for character constants is also different: @code{#\A} | |
5292 | instead of @code{?A}. Also, @code{string=} and @code{string-equal} | |
5293 | are synonyms in Emacs Lisp whereas the latter is case-insensitive | |
5294 | in Common Lisp. | |
5295 | ||
5296 | @item | |
5297 | Data types. Some Common Lisp data types do not exist in Emacs | |
5298 | Lisp. Rational numbers and complex numbers are not present, | |
5299 | nor are large integers (all integers are ``fixnums''). All | |
5300 | arrays are one-dimensional. There are no readtables or pathnames; | |
5301 | streams are a set of existing data types rather than a new data | |
5302 | type of their own. Hash tables, random-states, structures, and | |
5303 | packages (obarrays) are built from Lisp vectors or lists rather | |
5304 | than being distinct types. | |
5305 | ||
5306 | @item | |
5307 | The Common Lisp Object System (CLOS) is not implemented, | |
5308 | nor is the Common Lisp Condition System. However, the EIEIO package | |
5309 | from @uref{ftp://ftp.ultranet.com/pub/zappo} does implement some | |
5310 | CLOS functionality. | |
5311 | ||
5312 | @item | |
5313 | Common Lisp features that are completely redundant with Emacs | |
5314 | Lisp features of a different name generally have not been | |
5315 | implemented. For example, Common Lisp writes @code{defconstant} | |
5316 | where Emacs Lisp uses @code{defconst}. Similarly, @code{make-list} | |
5317 | takes its arguments in different ways in the two Lisps but does | |
5318 | exactly the same thing, so this package has not bothered to | |
5319 | implement a Common Lisp-style @code{make-list}. | |
5320 | ||
5321 | @item | |
5322 | A few more notable Common Lisp features not included in this | |
5323 | package: @code{compiler-let}, @code{tagbody}, @code{prog}, | |
5324 | @code{ldb/dpb}, @code{parse-integer}, @code{cerror}. | |
5325 | ||
5326 | @item | |
5327 | Recursion. While recursion works in Emacs Lisp just like it | |
5328 | does in Common Lisp, various details of the Emacs Lisp system | |
5329 | and compiler make recursion much less efficient than it is in | |
5330 | most Lisps. Some schools of thought prefer to use recursion | |
5331 | in Lisp over other techniques; they would sum a list of | |
5332 | numbers using something like | |
5333 | ||
5334 | @example | |
5335 | (defun sum-list (list) | |
5336 | (if list | |
5337 | (+ (car list) (sum-list (cdr list))) | |
5338 | 0)) | |
5339 | @end example | |
5340 | ||
5341 | @noindent | |
5342 | where a more iteratively-minded programmer might write one of | |
5343 | these forms: | |
5344 | ||
5345 | @example | |
5346 | (let ((total 0)) (dolist (x my-list) (incf total x)) total) | |
5347 | (loop for x in my-list sum x) | |
5348 | @end example | |
5349 | ||
5350 | While this would be mainly a stylistic choice in most Common Lisps, | |
5351 | in Emacs Lisp you should be aware that the iterative forms are | |
5352 | much faster than recursion. Also, Lisp programmers will want to | |
5353 | note that the current Emacs Lisp compiler does not optimize tail | |
5354 | recursion. | |
5355 | @end itemize | |
5356 | ||
84247bb5 CY |
5357 | @node GNU Free Documentation License, Function Index, Porting Common Lisp, Top |
5358 | @appendix GNU Free Documentation License | |
5359 | @include doclicense.texi | |
5360 | ||
5361 | @node Function Index, Variable Index, GNU Free Documentation License, Top | |
6bf7aab6 DL |
5362 | @unnumbered Function Index |
5363 | ||
5364 | @printindex fn | |
5365 | ||
c6ad01e8 | 5366 | @node Variable Index, , Function Index, Top |
6bf7aab6 DL |
5367 | @unnumbered Variable Index |
5368 | ||
5369 | @printindex vr | |
5370 | ||
56f7c94a | 5371 | @setchapternewpage odd |
6bf7aab6 DL |
5372 | @contents |
5373 | @bye | |
ab5796a9 MB |
5374 | |
5375 | @ignore | |
5376 | arch-tag: b61e7200-3bfa-4a70-a9d3-095e152696f8 | |
5377 | @end ignore |