src/casefiddle.c

   1 /* GNU Emacs case conversion functions.
   2    Copyright (C) 1985, 1994, 1997 Free Software Foundation, Inc.
   3
   4 This file is part of GNU Emacs.
   5
   6 GNU Emacs is free software; you can redistribute it and/or modify
   7 it under the terms of the GNU General Public License as published by
   8 the Free Software Foundation; either version 2, or (at your option)
   9 any later version.
  10
  11 GNU Emacs is distributed in the hope that it will be useful,
  12 but WITHOUT ANY WARRANTY; without even the implied warranty of
  13 MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  14 GNU General Public License for more details.
  15
  16 You should have received a copy of the GNU General Public License
  17 along with GNU Emacs; see the file COPYING.  If not, write to
  18 the Free Software Foundation, Inc., 59 Temple Place - Suite 330,
  19 Boston, MA 02111-1307, USA.  */
  20
  21
  22 #include <config.h>
  23 #include "lisp.h"
  24 #include "buffer.h"
  25 #include "charset.h"
  26 #include "commands.h"
  27 #include "syntax.h"
  28
  29 enum case_action {CASE_UP, CASE_DOWN, CASE_CAPITALIZE, CASE_CAPITALIZE_UP};
  30
  31 Lisp_Object Qidentity;
  32 \f
  33 Lisp_Object
  34 casify_object (flag, obj)
  35      enum case_action flag;
  36      Lisp_Object obj;
  37 {
  38   register int i, c, len;
  39   register int inword = flag == CASE_DOWN;
  40   Lisp_Object tem;
  41
  42   /* If the case table is flagged as modified, rescan it.  */
  43   if (NILP (XCHAR_TABLE (current_buffer->downcase_table)->extras[1]))
  44     Fset_case_table (current_buffer->downcase_table);
  45
  46   while (1)
  47     {
  48       if (INTEGERP (obj))
  49         {
  50           c = DOWNCASE (XFASTINT (obj));
  51           if (inword)
  52             XSETFASTINT (obj, c);
  53           else if (c == XFASTINT (obj))
  54             {
  55               c = UPCASE1 (XFASTINT (obj));
  56               XSETFASTINT (obj, c);
  57             }
  58           return obj;
  59         }
  60
  61       if (STRINGP (obj))
  62         {
  63           int multibyte = STRING_MULTIBYTE (obj);
  64
  65           obj = Fcopy_sequence (obj);
  66           len = STRING_BYTES (XSTRING (obj));
  67
  68           /* Scan all single-byte characters from start of string.  */
  69           for (i = 0; i < len;)
  70             {
  71               c = XSTRING (obj)->data[i];
  72
  73               if (multibyte && c >= 0x80)
  74                 /* A multibyte character can't be handled in this
  75                    simple loop.  */
  76                 break;
  77               if (inword && flag != CASE_CAPITALIZE_UP)
  78                 c = DOWNCASE (c);
  79               else if (!UPPERCASEP (c)
  80                        && (!inword || flag != CASE_CAPITALIZE_UP))
  81                 c = UPCASE1 (c);
  82               /* If this char won't fit in a single-byte string.
  83                  fall out to the multibyte case.  */
  84               if (multibyte ? ! ASCII_BYTE_P (c)
  85                   : ! SINGLE_BYTE_CHAR_P (c))
  86                 break;
  87
  88               XSTRING (obj)->data[i] = c;
  89               if ((int) flag >= (int) CASE_CAPITALIZE)
  90                 inword = SYNTAX (c) == Sword;
  91               i++;
  92             }
  93
  94           /* If we didn't do the whole string as single-byte,
  95              scan the rest in a more complex way.  */
  96           if (i < len)
  97             {
  98               /* The work is not yet finished because of a multibyte
  99                  character just encountered.  */
 100               int fromlen, tolen, j = i, j_byte = i;
 101               char *buf
 102                 = (char *) alloca ((len - i) * MAX_LENGTH_OF_MULTI_BYTE_FORM
 103                                    + i);
 104               unsigned char *str, workbuf[4];
 105
 106               /* Copy data already handled.  */
 107               bcopy (XSTRING (obj)->data, buf, i);
 108
 109               /* From now on, I counts bytes.  */
 110               while (i < len)
 111                 {
 112                   c = STRING_CHAR_AND_LENGTH (XSTRING (obj)->data + i,
 113                                               len - i, fromlen);
 114                   if (inword && flag != CASE_CAPITALIZE_UP)
 115                     c = DOWNCASE (c);
 116                   else if (!UPPERCASEP (c)
 117                            && (!inword || flag != CASE_CAPITALIZE_UP))
 118                     c = UPCASE1 (c);
 119                   tolen = CHAR_STRING (c, workbuf, str);
 120                   bcopy (str, buf + j_byte, tolen);
 121                   i += fromlen;
 122                   j++;
 123                   j_byte += tolen;
 124                   if ((int) flag >= (int) CASE_CAPITALIZE)
 125                     inword = SYNTAX (c) == Sword;
 126                 }
 127               obj = make_specified_string (buf, j, j_byte,
 128                                            STRING_MULTIBYTE (obj));
 129             }
 130           return obj;
 131         }
 132       obj = wrong_type_argument (Qchar_or_string_p, obj);
 133     }
 134 }
 135
 136 DEFUN ("upcase", Fupcase, Supcase, 1, 1, 0,
 137   "Convert argument to upper case and return that.\n\
 138 The argument may be a character or string.  The result has the same type.\n\
 139 The argument object is not altered--the value is a copy.\n\
 140 See also `capitalize', `downcase' and `upcase-initials'.")
 141   (obj)
 142      Lisp_Object obj;
 143 {
 144   return casify_object (CASE_UP, obj);
 145 }
 146
 147 DEFUN ("downcase", Fdowncase, Sdowncase, 1, 1, 0,
 148   "Convert argument to lower case and return that.\n\
 149 The argument may be a character or string.  The result has the same type.\n\
 150 The argument object is not altered--the value is a copy.")
 151   (obj)
 152      Lisp_Object obj;
 153 {
 154   return casify_object (CASE_DOWN, obj);
 155 }
 156
 157 DEFUN ("capitalize", Fcapitalize, Scapitalize, 1, 1, 0,
 158   "Convert argument to capitalized form and return that.\n\
 159 This means that each word's first character is upper case\n\
 160 and the rest is lower case.\n\
 161 The argument may be a character or string.  The result has the same type.\n\
 162 The argument object is not altered--the value is a copy.")
 163   (obj)
 164      Lisp_Object obj;
 165 {
 166   return casify_object (CASE_CAPITALIZE, obj);
 167 }
 168
 169 /* Like Fcapitalize but change only the initials.  */
 170
 171 DEFUN ("upcase-initials", Fupcase_initials, Supcase_initials, 1, 1, 0,
 172   "Convert the initial of each word in the argument to upper case.\n\
 173 Do not change the other letters of each word.\n\
 174 The argument may be a character or string.  The result has the same type.\n\
 175 The argument object is not altered--the value is a copy.")
 176   (obj)
 177      Lisp_Object obj;
 178 {
 179   return casify_object (CASE_CAPITALIZE_UP, obj);
 180 }
 181 \f
 182 /* flag is CASE_UP, CASE_DOWN or CASE_CAPITALIZE or CASE_CAPITALIZE_UP.
 183    b and e specify range of buffer to operate on. */
 184
 185 casify_region (flag, b, e)
 186      enum case_action flag;
 187      Lisp_Object b, e;
 188 {
 189   register int i;
 190   register int c;
 191   register int inword = flag == CASE_DOWN;
 192   register int multibyte = !NILP (current_buffer->enable_multibyte_characters);
 193   int start, end;
 194   int start_byte, end_byte;
 195   Lisp_Object ch, downch, val;
 196
 197   if (EQ (b, e))
 198     /* Not modifying because nothing marked */
 199     return;
 200
 201   /* If the case table is flagged as modified, rescan it.  */
 202   if (NILP (XCHAR_TABLE (current_buffer->downcase_table)->extras[1]))
 203     Fset_case_table (current_buffer->downcase_table);
 204
 205   validate_region (&b, &e);
 206   start = XFASTINT (b);
 207   end = XFASTINT (e);
 208   modify_region (current_buffer, start, end);
 209   record_change (start, end - start);
 210   start_byte = CHAR_TO_BYTE (start);
 211   end_byte = CHAR_TO_BYTE (end);
 212
 213   for (i = start_byte; i < end_byte; i++)
 214     {
 215       c = FETCH_BYTE (i);
 216       if (multibyte && c >= 0x80)
 217         /* A multibyte character can't be handled in this simple loop.  */
 218         break;
 219       if (inword && flag != CASE_CAPITALIZE_UP)
 220         c = DOWNCASE (c);
 221       else if (!UPPERCASEP (c)
 222                && (!inword || flag != CASE_CAPITALIZE_UP))
 223         c = UPCASE1 (c);
 224       FETCH_BYTE (i) = c;
 225       if ((int) flag >= (int) CASE_CAPITALIZE)
 226         inword = SYNTAX (c) == Sword;
 227     }
 228   if (i < end_byte)
 229     {
 230       /* The work is not yet finished because of a multibyte character
 231          just encountered.  */
 232       int opoint = PT;
 233       int opoint_byte = PT_BYTE;
 234       int c2;
 235
 236       while (i < end_byte)
 237         {
 238           if ((c = FETCH_BYTE (i)) >= 0x80)
 239             c = FETCH_MULTIBYTE_CHAR (i);
 240           c2 = c;
 241           if (inword && flag != CASE_CAPITALIZE_UP)
 242             c2 = DOWNCASE (c);
 243           else if (!UPPERCASEP (c)
 244                    && (!inword || flag != CASE_CAPITALIZE_UP))
 245             c2 = UPCASE1 (c);
 246           if (c != c2)
 247             {
 248               int fromlen, tolen, j;
 249               unsigned char workbuf[4], *str;
 250
 251               /* Handle the most likely case */
 252               if (c < 0400 && c2 < 0400)
 253                 FETCH_BYTE (i) = c2;
 254               else if (fromlen = CHAR_STRING (c, workbuf, str),
 255                        tolen = CHAR_STRING (c2, workbuf, str),
 256                        fromlen == tolen)
 257                 {
 258                   for (j = 0; j < tolen; ++j)
 259                     FETCH_BYTE (i + j) = str[j];
 260                 }
 261               else
 262                 {
 263                   error ("Can't casify letters that change length");
 264 #if 0 /* This is approximately what we'd like to be able to do here */
 265                   if (tolen < fromlen)
 266                     del_range_1 (i + tolen, i + fromlen, 0);
 267                   else if (tolen > fromlen)
 268                     {
 269                       TEMP_SET_PT (i + fromlen);
 270                       insert_1 (str + fromlen, tolen - fromlen, 1, 0, 0);
 271                     }
 272 #endif
 273                 }
 274             }
 275           if ((int) flag >= (int) CASE_CAPITALIZE)
 276             inword = SYNTAX (c2) == Sword;
 277           INC_POS (i);
 278         }
 279       TEMP_SET_PT_BOTH (opoint, opoint_byte);
 280     }
 281
 282   signal_after_change (start, end - start, end - start);
 283 }
 284
 285 DEFUN ("upcase-region", Fupcase_region, Supcase_region, 2, 2, "r",
 286   "Convert the region to upper case.  In programs, wants two arguments.\n\
 287 These arguments specify the starting and ending character numbers of\n\
 288 the region to operate on.  When used as a command, the text between\n\
 289 point and the mark is operated on.\n\
 290 See also `capitalize-region'.")
 291   (beg, end)
 292      Lisp_Object beg, end;
 293 {
 294   casify_region (CASE_UP, beg, end);
 295   return Qnil;
 296 }
 297
 298 DEFUN ("downcase-region", Fdowncase_region, Sdowncase_region, 2, 2, "r",
 299   "Convert the region to lower case.  In programs, wants two arguments.\n\
 300 These arguments specify the starting and ending character numbers of\n\
 301 the region to operate on.  When used as a command, the text between\n\
 302 point and the mark is operated on.")
 303   (beg, end)
 304      Lisp_Object beg, end;
 305 {
 306   casify_region (CASE_DOWN, beg, end);
 307   return Qnil;
 308 }
 309
 310 DEFUN ("capitalize-region", Fcapitalize_region, Scapitalize_region, 2, 2, "r",
 311   "Convert the region to capitalized form.\n\
 312 Capitalized form means each word's first character is upper case\n\
 313 and the rest of it is lower case.\n\
 314 In programs, give two arguments, the starting and ending\n\
 315 character positions to operate on.")
 316   (beg, end)
 317      Lisp_Object beg, end;
 318 {
 319   casify_region (CASE_CAPITALIZE, beg, end);
 320   return Qnil;
 321 }
 322
 323 /* Like Fcapitalize_region but change only the initials.  */
 324
 325 DEFUN ("upcase-initials-region", Fupcase_initials_region,
 326        Supcase_initials_region, 2, 2, "r",
 327   "Upcase the initial of each word in the region.\n\
 328 Subsequent letters of each word are not changed.\n\
 329 In programs, give two arguments, the starting and ending\n\
 330 character positions to operate on.")
 331   (beg, end)
 332      Lisp_Object beg, end;
 333 {
 334   casify_region (CASE_CAPITALIZE_UP, beg, end);
 335   return Qnil;
 336 }
 337 \f
 338 Lisp_Object
 339 operate_on_word (arg, newpoint)
 340      Lisp_Object arg;
 341      int *newpoint;
 342 {
 343   Lisp_Object val;
 344   int farend;
 345   int iarg;
 346
 347   CHECK_NUMBER (arg, 0);
 348   iarg = XINT (arg);
 349   farend = scan_words (PT, iarg);
 350   if (!farend)
 351     farend = iarg > 0 ? ZV : BEGV;
 352
 353   *newpoint = PT > farend ? PT : farend;
 354   XSETFASTINT (val, farend);
 355
 356   return val;
 357 }
 358
 359 DEFUN ("upcase-word", Fupcase_word, Supcase_word, 1, 1, "p",
 360   "Convert following word (or ARG words) to upper case, moving over.\n\
 361 With negative argument, convert previous words but do not move.\n\
 362 See also `capitalize-word'.")
 363   (arg)
 364      Lisp_Object arg;
 365 {
 366   Lisp_Object beg, end;
 367   int newpoint;
 368   XSETFASTINT (beg, PT);
 369   end = operate_on_word (arg, &newpoint);
 370   casify_region (CASE_UP, beg, end);
 371   SET_PT (newpoint);
 372   return Qnil;
 373 }
 374
 375 DEFUN ("downcase-word", Fdowncase_word, Sdowncase_word, 1, 1, "p",
 376   "Convert following word (or ARG words) to lower case, moving over.\n\
 377 With negative argument, convert previous words but do not move.")
 378   (arg)
 379      Lisp_Object arg;
 380 {
 381   Lisp_Object beg, end;
 382   int newpoint;
 383   XSETFASTINT (beg, PT);
 384   end = operate_on_word (arg, &newpoint);
 385   casify_region (CASE_DOWN, beg, end);
 386   SET_PT (newpoint);
 387   return Qnil;
 388 }
 389
 390 DEFUN ("capitalize-word", Fcapitalize_word, Scapitalize_word, 1, 1, "p",
 391   "Capitalize the following word (or ARG words), moving over.\n\
 392 This gives the word(s) a first character in upper case\n\
 393 and the rest lower case.\n\
 394 With negative argument, capitalize previous words but do not move.")
 395   (arg)
 396      Lisp_Object arg;
 397 {
 398   Lisp_Object beg, end;
 399   int newpoint;
 400   XSETFASTINT (beg, PT);
 401   end = operate_on_word (arg, &newpoint);
 402   casify_region (CASE_CAPITALIZE, beg, end);
 403   SET_PT (newpoint);
 404   return Qnil;
 405 }
 406 \f
 407 syms_of_casefiddle ()
 408 {
 409   Qidentity = intern ("identity");
 410   staticpro (&Qidentity);
 411   defsubr (&Supcase);
 412   defsubr (&Sdowncase);
 413   defsubr (&Scapitalize);
 414   defsubr (&Supcase_initials);
 415   defsubr (&Supcase_region);
 416   defsubr (&Sdowncase_region);
 417   defsubr (&Scapitalize_region);
 418   defsubr (&Supcase_initials_region);
 419   defsubr (&Supcase_word);
 420   defsubr (&Sdowncase_word);
 421   defsubr (&Scapitalize_word);
 422 }
 423
 424 keys_of_casefiddle ()
 425 {
 426   initial_define_key (control_x_map, Ctl('U'), "upcase-region");
 427   Fput (intern ("upcase-region"), Qdisabled, Qt);
 428   initial_define_key (control_x_map, Ctl('L'), "downcase-region");
 429   Fput (intern ("downcase-region"), Qdisabled, Qt);
 430
 431   initial_define_key (meta_map, 'u', "upcase-word");
 432   initial_define_key (meta_map, 'l', "downcase-word");
 433   initial_define_key (meta_map, 'c', "capitalize-word");
 434 }