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 #include "composite.h"
  29
  30 enum case_action {CASE_UP, CASE_DOWN, CASE_CAPITALIZE, CASE_CAPITALIZE_UP};
  31
  32 Lisp_Object Qidentity;
  33 \f
  34 Lisp_Object
  35 casify_object (flag, obj)
  36      enum case_action flag;
  37      Lisp_Object obj;
  38 {
  39   register int i, c, len;
  40   register int inword = flag == CASE_DOWN;
  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           int flagbits = (CHAR_ALT | CHAR_SUPER | CHAR_HYPER
  51                           | CHAR_SHIFT | CHAR_CTL | CHAR_META);
  52           int flags = XINT (obj) & flagbits;
  53
  54           c = DOWNCASE (XFASTINT (obj) & ~flagbits);
  55           if (inword)
  56             XSETFASTINT (obj, c | flags);
  57           else if (c == (XFASTINT (obj) & ~flagbits))
  58             {
  59               c = UPCASE1 ((XFASTINT (obj) & ~flagbits));
  60               XSETFASTINT (obj, c | flags);
  61             }
  62           return obj;
  63         }
  64
  65       if (STRINGP (obj))
  66         {
  67           int multibyte = STRING_MULTIBYTE (obj);
  68
  69           obj = Fcopy_sequence (obj);
  70           len = STRING_BYTES (XSTRING (obj));
  71
  72           /* Scan all single-byte characters from start of string.  */
  73           for (i = 0; i < len;)
  74             {
  75               c = XSTRING (obj)->data[i];
  76
  77               if (multibyte && c >= 0x80)
  78                 /* A multibyte character can't be handled in this
  79                    simple loop.  */
  80                 break;
  81               if (inword && flag != CASE_CAPITALIZE_UP)
  82                 c = DOWNCASE (c);
  83               else if (!UPPERCASEP (c)
  84                        && (!inword || flag != CASE_CAPITALIZE_UP))
  85                 c = UPCASE1 (c);
  86               /* If this char won't fit in a single-byte string.
  87                  fall out to the multibyte case.  */
  88               if (multibyte ? ! ASCII_BYTE_P (c)
  89                   : ! SINGLE_BYTE_CHAR_P (c))
  90                 break;
  91
  92               XSTRING (obj)->data[i] = c;
  93               if ((int) flag >= (int) CASE_CAPITALIZE)
  94                 inword = SYNTAX (c) == Sword;
  95               i++;
  96             }
  97
  98           /* If we didn't do the whole string as single-byte,
  99              scan the rest in a more complex way.  */
 100           if (i < len)
 101             {
 102               /* The work is not yet finished because of a multibyte
 103                  character just encountered.  */
 104               int fromlen, j_byte = i;
 105               char *buf
 106                 = (char *) alloca ((len - i) * MAX_MULTIBYTE_LENGTH + i);
 107
 108               /* Copy data already handled.  */
 109               bcopy (XSTRING (obj)->data, buf, i);
 110
 111               /* From now on, I counts bytes.  */
 112               while (i < len)
 113                 {
 114                   c = STRING_CHAR_AND_LENGTH (XSTRING (obj)->data + i,
 115                                               len - i, fromlen);
 116                   if (inword && flag != CASE_CAPITALIZE_UP)
 117                     c = DOWNCASE (c);
 118                   else if (!UPPERCASEP (c)
 119                            && (!inword || flag != CASE_CAPITALIZE_UP))
 120                     c = UPCASE1 (c);
 121                   i += fromlen;
 122                   j_byte += CHAR_STRING (c, buf + j_byte);
 123                   if ((int) flag >= (int) CASE_CAPITALIZE)
 124                     inword = SYNTAX (c) == Sword;
 125                 }
 126               obj = make_multibyte_string (buf, XSTRING (obj)->size,
 127                                            j_byte);
 128             }
 129           return obj;
 130         }
 131       obj = wrong_type_argument (Qchar_or_string_p, obj);
 132     }
 133 }
 134
 135 DEFUN ("upcase", Fupcase, Supcase, 1, 1, 0,
 136   "Convert argument to upper case and return that.\n\
 137 The argument may be a character or string.  The result has the same type.\n\
 138 The argument object is not altered--the value is a copy.\n\
 139 See also `capitalize', `downcase' and `upcase-initials'.")
 140   (obj)
 141      Lisp_Object obj;
 142 {
 143   return casify_object (CASE_UP, obj);
 144 }
 145
 146 DEFUN ("downcase", Fdowncase, Sdowncase, 1, 1, 0,
 147   "Convert argument to lower case and return that.\n\
 148 The argument may be a character or string.  The result has the same type.\n\
 149 The argument object is not altered--the value is a copy.")
 150   (obj)
 151      Lisp_Object obj;
 152 {
 153   return casify_object (CASE_DOWN, obj);
 154 }
 155
 156 DEFUN ("capitalize", Fcapitalize, Scapitalize, 1, 1, 0,
 157   "Convert argument to capitalized form and return that.\n\
 158 This means that each word's first character is upper case\n\
 159 and the rest is lower case.\n\
 160 The argument may be a character or string.  The result has the same type.\n\
 161 The argument object is not altered--the value is a copy.")
 162   (obj)
 163      Lisp_Object obj;
 164 {
 165   return casify_object (CASE_CAPITALIZE, obj);
 166 }
 167
 168 /* Like Fcapitalize but change only the initials.  */
 169
 170 DEFUN ("upcase-initials", Fupcase_initials, Supcase_initials, 1, 1, 0,
 171   "Convert the initial of each word in the argument to upper case.\n\
 172 Do not change the other letters of each word.\n\
 173 The argument may be a character or string.  The result has the same type.\n\
 174 The argument object is not altered--the value is a copy.")
 175   (obj)
 176      Lisp_Object obj;
 177 {
 178   return casify_object (CASE_CAPITALIZE_UP, obj);
 179 }
 180 \f
 181 /* flag is CASE_UP, CASE_DOWN or CASE_CAPITALIZE or CASE_CAPITALIZE_UP.
 182    b and e specify range of buffer to operate on. */
 183
 184 void
 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   int changed = 0;
 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++, start++)
 214     {
 215       int c2;
 216       c = c2 = FETCH_BYTE (i);
 217       if (multibyte && c >= 0x80)
 218         /* A multibyte character can't be handled in this simple loop.  */
 219         break;
 220       if (inword && flag != CASE_CAPITALIZE_UP)
 221         c = DOWNCASE (c);
 222       else if (!UPPERCASEP (c)
 223                && (!inword || flag != CASE_CAPITALIZE_UP))
 224         c = UPCASE1 (c);
 225       FETCH_BYTE (i) = c;
 226       if (c != c2)
 227         changed = 1;
 228       if ((int) flag >= (int) CASE_CAPITALIZE)
 229         inword = SYNTAX (c) == Sword;
 230     }
 231   if (i < end_byte)
 232     {
 233       /* The work is not yet finished because of a multibyte character
 234          just encountered.  */
 235       int opoint = PT;
 236       int opoint_byte = PT_BYTE;
 237       int c2;
 238
 239       while (i < end_byte)
 240         {
 241           if ((c = FETCH_BYTE (i)) >= 0x80)
 242             c = FETCH_MULTIBYTE_CHAR (i);
 243           c2 = c;
 244           if (inword && flag != CASE_CAPITALIZE_UP)
 245             c2 = DOWNCASE (c);
 246           else if (!UPPERCASEP (c)
 247                    && (!inword || flag != CASE_CAPITALIZE_UP))
 248             c2 = UPCASE1 (c);
 249           if (c != c2)
 250             {
 251               int fromlen, tolen, j;
 252               unsigned char str[MAX_MULTIBYTE_LENGTH];
 253
 254               changed = 1;
 255               /* Handle the most likely case */
 256               if (c < 0400 && c2 < 0400)
 257                 FETCH_BYTE (i) = c2;
 258               else if (fromlen = CHAR_STRING (c, str),
 259                        tolen = CHAR_STRING (c2, str),
 260                        fromlen == tolen)
 261                 {
 262                   for (j = 0; j < tolen; ++j)
 263                     FETCH_BYTE (i + j) = str[j];
 264                 }
 265               else
 266                 {
 267                   error ("Can't casify letters that change length");
 268 #if 0 /* This is approximately what we'd like to be able to do here */
 269                   if (tolen < fromlen)
 270                     del_range_1 (i + tolen, i + fromlen, 0, 0);
 271                   else if (tolen > fromlen)
 272                     {
 273                       TEMP_SET_PT (i + fromlen);
 274                       insert_1 (str + fromlen, tolen - fromlen, 1, 0, 0);
 275                     }
 276 #endif
 277                 }
 278             }
 279           if ((int) flag >= (int) CASE_CAPITALIZE)
 280             inword = SYNTAX (c2) == Sword;
 281           INC_BOTH (start, i);
 282         }
 283       TEMP_SET_PT_BOTH (opoint, opoint_byte);
 284     }
 285
 286   start = XFASTINT (b);
 287   if (changed)
 288     {
 289       signal_after_change (start, end - start, end - start);
 290       update_compositions (start, end, CHECK_ALL);
 291     }
 292 }
 293
 294 DEFUN ("upcase-region", Fupcase_region, Supcase_region, 2, 2, "r",
 295   "Convert the region to upper case.  In programs, wants two arguments.\n\
 296 These arguments specify the starting and ending character numbers of\n\
 297 the region to operate on.  When used as a command, the text between\n\
 298 point and the mark is operated on.\n\
 299 See also `capitalize-region'.")
 300   (beg, end)
 301      Lisp_Object beg, end;
 302 {
 303   casify_region (CASE_UP, beg, end);
 304   return Qnil;
 305 }
 306
 307 DEFUN ("downcase-region", Fdowncase_region, Sdowncase_region, 2, 2, "r",
 308   "Convert the region to lower case.  In programs, wants two arguments.\n\
 309 These arguments specify the starting and ending character numbers of\n\
 310 the region to operate on.  When used as a command, the text between\n\
 311 point and the mark is operated on.")
 312   (beg, end)
 313      Lisp_Object beg, end;
 314 {
 315   casify_region (CASE_DOWN, beg, end);
 316   return Qnil;
 317 }
 318
 319 DEFUN ("capitalize-region", Fcapitalize_region, Scapitalize_region, 2, 2, "r",
 320   "Convert the region to capitalized form.\n\
 321 Capitalized form means each word's first character is upper case\n\
 322 and the rest of it is lower case.\n\
 323 In programs, give two arguments, the starting and ending\n\
 324 character positions to operate on.")
 325   (beg, end)
 326      Lisp_Object beg, end;
 327 {
 328   casify_region (CASE_CAPITALIZE, beg, end);
 329   return Qnil;
 330 }
 331
 332 /* Like Fcapitalize_region but change only the initials.  */
 333
 334 DEFUN ("upcase-initials-region", Fupcase_initials_region,
 335        Supcase_initials_region, 2, 2, "r",
 336   "Upcase the initial of each word in the region.\n\
 337 Subsequent letters of each word are not changed.\n\
 338 In programs, give two arguments, the starting and ending\n\
 339 character positions to operate on.")
 340   (beg, end)
 341      Lisp_Object beg, end;
 342 {
 343   casify_region (CASE_CAPITALIZE_UP, beg, end);
 344   return Qnil;
 345 }
 346 \f
 347 Lisp_Object
 348 operate_on_word (arg, newpoint)
 349      Lisp_Object arg;
 350      int *newpoint;
 351 {
 352   Lisp_Object val;
 353   int farend;
 354   int iarg;
 355
 356   CHECK_NUMBER (arg, 0);
 357   iarg = XINT (arg);
 358   farend = scan_words (PT, iarg);
 359   if (!farend)
 360     farend = iarg > 0 ? ZV : BEGV;
 361
 362   *newpoint = PT > farend ? PT : farend;
 363   XSETFASTINT (val, farend);
 364
 365   return val;
 366 }
 367
 368 DEFUN ("upcase-word", Fupcase_word, Supcase_word, 1, 1, "p",
 369   "Convert following word (or ARG words) to upper case, moving over.\n\
 370 With negative argument, convert previous words but do not move.\n\
 371 See also `capitalize-word'.")
 372   (arg)
 373      Lisp_Object arg;
 374 {
 375   Lisp_Object beg, end;
 376   int newpoint;
 377   XSETFASTINT (beg, PT);
 378   end = operate_on_word (arg, &newpoint);
 379   casify_region (CASE_UP, beg, end);
 380   SET_PT (newpoint);
 381   return Qnil;
 382 }
 383
 384 DEFUN ("downcase-word", Fdowncase_word, Sdowncase_word, 1, 1, "p",
 385   "Convert following word (or ARG words) to lower case, moving over.\n\
 386 With negative argument, convert previous words but do not move.")
 387   (arg)
 388      Lisp_Object arg;
 389 {
 390   Lisp_Object beg, end;
 391   int newpoint;
 392   XSETFASTINT (beg, PT);
 393   end = operate_on_word (arg, &newpoint);
 394   casify_region (CASE_DOWN, beg, end);
 395   SET_PT (newpoint);
 396   return Qnil;
 397 }
 398
 399 DEFUN ("capitalize-word", Fcapitalize_word, Scapitalize_word, 1, 1, "p",
 400   "Capitalize the following word (or ARG words), moving over.\n\
 401 This gives the word(s) a first character in upper case\n\
 402 and the rest lower case.\n\
 403 With negative argument, capitalize previous words but do not move.")
 404   (arg)
 405      Lisp_Object arg;
 406 {
 407   Lisp_Object beg, end;
 408   int newpoint;
 409   XSETFASTINT (beg, PT);
 410   end = operate_on_word (arg, &newpoint);
 411   casify_region (CASE_CAPITALIZE, beg, end);
 412   SET_PT (newpoint);
 413   return Qnil;
 414 }
 415 \f
 416 void
 417 syms_of_casefiddle ()
 418 {
 419   Qidentity = intern ("identity");
 420   staticpro (&Qidentity);
 421   defsubr (&Supcase);
 422   defsubr (&Sdowncase);
 423   defsubr (&Scapitalize);
 424   defsubr (&Supcase_initials);
 425   defsubr (&Supcase_region);
 426   defsubr (&Sdowncase_region);
 427   defsubr (&Scapitalize_region);
 428   defsubr (&Supcase_initials_region);
 429   defsubr (&Supcase_word);
 430   defsubr (&Sdowncase_word);
 431   defsubr (&Scapitalize_word);
 432 }
 433
 434 void
 435 keys_of_casefiddle ()
 436 {
 437   initial_define_key (control_x_map, Ctl('U'), "upcase-region");
 438   Fput (intern ("upcase-region"), Qdisabled, Qt);
 439   initial_define_key (control_x_map, Ctl('L'), "downcase-region");
 440   Fput (intern ("downcase-region"), Qdisabled, Qt);
 441
 442   initial_define_key (meta_map, 'u', "upcase-word");
 443   initial_define_key (meta_map, 'l', "downcase-word");
 444   initial_define_key (meta_map, 'c', "capitalize-word");
 445 }