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 #include "keymap.h"
  30
  31 enum case_action {CASE_UP, CASE_DOWN, CASE_CAPITALIZE, CASE_CAPITALIZE_UP};
  32
  33 Lisp_Object Qidentity;
  34 \f
  35 Lisp_Object
  36 casify_object (flag, obj)
  37      enum case_action flag;
  38      Lisp_Object obj;
  39 {
  40   register int i, c, len;
  41   register int inword = flag == CASE_DOWN;
  42
  43   /* If the case table is flagged as modified, rescan it.  */
  44   if (NILP (XCHAR_TABLE (current_buffer->downcase_table)->extras[1]))
  45     Fset_case_table (current_buffer->downcase_table);
  46
  47   while (1)
  48     {
  49       if (INTEGERP (obj))
  50         {
  51           int flagbits = (CHAR_ALT | CHAR_SUPER | CHAR_HYPER
  52                           | CHAR_SHIFT | CHAR_CTL | CHAR_META);
  53           int flags = XINT (obj) & flagbits;
  54
  55           c = DOWNCASE (XFASTINT (obj) & ~flagbits);
  56           if (inword)
  57             XSETFASTINT (obj, c | flags);
  58           else if (c == (XFASTINT (obj) & ~flagbits))
  59             {
  60               c = UPCASE1 ((XFASTINT (obj) & ~flagbits));
  61               XSETFASTINT (obj, c | flags);
  62             }
  63           return obj;
  64         }
  65
  66       if (STRINGP (obj))
  67         {
  68           int multibyte = STRING_MULTIBYTE (obj);
  69
  70           obj = Fcopy_sequence (obj);
  71           len = SBYTES (obj);
  72
  73           /* Scan all single-byte characters from start of string.  */
  74           for (i = 0; i < len;)
  75             {
  76               c = SREF (obj, i);
  77
  78               if (multibyte && c >= 0x80)
  79                 /* A multibyte character can't be handled in this
  80                    simple loop.  */
  81                 break;
  82               if (inword && flag != CASE_CAPITALIZE_UP)
  83                 c = DOWNCASE (c);
  84               else if (!UPPERCASEP (c)
  85                        && (!inword || flag != CASE_CAPITALIZE_UP))
  86                 c = UPCASE1 (c);
  87               /* If this char won't fit in a single-byte string.
  88                  fall out to the multibyte case.  */
  89               if (multibyte ? ! ASCII_BYTE_P (c)
  90                   : ! SINGLE_BYTE_CHAR_P (c))
  91                 break;
  92
  93               SSET (obj, i, c);
  94               if ((int) flag >= (int) CASE_CAPITALIZE)
  95                 inword = SYNTAX (c) == Sword;
  96               i++;
  97             }
  98
  99           /* If we didn't do the whole string as single-byte,
 100              scan the rest in a more complex way.  */
 101           if (i < len)
 102             {
 103               /* The work is not yet finished because of a multibyte
 104                  character just encountered.  */
 105               int fromlen, j_byte = i;
 106               char *buf
 107                 = (char *) alloca ((len - i) * MAX_MULTIBYTE_LENGTH + i);
 108
 109               /* Copy data already handled.  */
 110               bcopy (SDATA (obj), buf, i);
 111
 112               /* From now on, I counts bytes.  */
 113               while (i < len)
 114                 {
 115                   c = STRING_CHAR_AND_LENGTH (SDATA (obj) + i,
 116                                               len - i, fromlen);
 117                   if (inword && flag != CASE_CAPITALIZE_UP)
 118                     c = DOWNCASE (c);
 119                   else if (!UPPERCASEP (c)
 120                            && (!inword || flag != CASE_CAPITALIZE_UP))
 121                     c = UPCASE1 (c);
 122                   i += fromlen;
 123                   j_byte += CHAR_STRING (c, buf + j_byte);
 124                   if ((int) flag >= (int) CASE_CAPITALIZE)
 125                     inword = SYNTAX (c) == Sword;
 126                 }
 127               obj = make_multibyte_string (buf, SCHARS (obj),
 128                                            j_byte);
 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        doc: /* Convert argument to upper case and return that.
 138 The argument may be a character or string.  The result has the same type.
 139 The argument object is not altered--the value is a copy.
 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        doc: /* Convert argument to lower case and return that.
 149 The argument may be a character or string.  The result has the same type.
 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        doc: /* Convert argument to capitalized form and return that.
 159 This means that each word's first character is upper case
 160 and the rest is lower case.
 161 The argument may be a character or string.  The result has the same type.
 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        doc: /* Convert the initial of each word in the argument to upper case.
 173 Do not change the other letters of each word.
 174 The argument may be a character or string.  The result has the same type.
 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 void
 186 casify_region (flag, b, e)
 187      enum case_action flag;
 188      Lisp_Object b, e;
 189 {
 190   register int i;
 191   register int c;
 192   register int inword = flag == CASE_DOWN;
 193   register int multibyte = !NILP (current_buffer->enable_multibyte_characters);
 194   int start, end;
 195   int start_byte, end_byte;
 196   int changed = 0;
 197
 198   if (EQ (b, e))
 199     /* Not modifying because nothing marked */
 200     return;
 201
 202   /* If the case table is flagged as modified, rescan it.  */
 203   if (NILP (XCHAR_TABLE (current_buffer->downcase_table)->extras[1]))
 204     Fset_case_table (current_buffer->downcase_table);
 205
 206   validate_region (&b, &e);
 207   start = XFASTINT (b);
 208   end = XFASTINT (e);
 209   modify_region (current_buffer, start, end);
 210   record_change (start, end - start);
 211   start_byte = CHAR_TO_BYTE (start);
 212   end_byte = CHAR_TO_BYTE (end);
 213
 214   for (i = start_byte; i < end_byte; i++, start++)
 215     {
 216       int c2;
 217       c = c2 = FETCH_BYTE (i);
 218       if (multibyte && c >= 0x80)
 219         /* A multibyte character can't be handled in this simple loop.  */
 220         break;
 221       if (inword && flag != CASE_CAPITALIZE_UP)
 222         c = DOWNCASE (c);
 223       else if (!UPPERCASEP (c)
 224                && (!inword || flag != CASE_CAPITALIZE_UP))
 225         c = UPCASE1 (c);
 226       FETCH_BYTE (i) = c;
 227       if (c != c2)
 228         changed = 1;
 229       if ((int) flag >= (int) CASE_CAPITALIZE)
 230         inword = SYNTAX (c) == Sword && (inword || !SYNTAX_PREFIX (c));
 231     }
 232   if (i < end_byte)
 233     {
 234       /* The work is not yet finished because of a multibyte character
 235          just encountered.  */
 236       int opoint = PT;
 237       int opoint_byte = PT_BYTE;
 238       int c2;
 239
 240       while (i < end_byte)
 241         {
 242           if ((c = FETCH_BYTE (i)) >= 0x80)
 243             c = FETCH_MULTIBYTE_CHAR (i);
 244           c2 = c;
 245           if (inword && flag != CASE_CAPITALIZE_UP)
 246             c2 = DOWNCASE (c);
 247           else if (!UPPERCASEP (c)
 248                    && (!inword || flag != CASE_CAPITALIZE_UP))
 249             c2 = UPCASE1 (c);
 250           if (c != c2)
 251             {
 252               int fromlen, tolen, j;
 253               unsigned char str[MAX_MULTIBYTE_LENGTH];
 254
 255               changed = 1;
 256               /* Handle the most likely case */
 257               if (c < 0400 && c2 < 0400)
 258                 FETCH_BYTE (i) = c2;
 259               else if (fromlen = CHAR_STRING (c, str),
 260                        tolen = CHAR_STRING (c2, str),
 261                        fromlen == tolen)
 262                 {
 263                   for (j = 0; j < tolen; ++j)
 264                     FETCH_BYTE (i + j) = str[j];
 265                 }
 266               else
 267                 {
 268                   error ("Can't casify letters that change length");
 269 #if 0 /* This is approximately what we'd like to be able to do here */
 270                   if (tolen < fromlen)
 271                     del_range_1 (i + tolen, i + fromlen, 0, 0);
 272                   else if (tolen > fromlen)
 273                     {
 274                       TEMP_SET_PT (i + fromlen);
 275                       insert_1 (str + fromlen, tolen - fromlen, 1, 0, 0);
 276                     }
 277 #endif
 278                 }
 279             }
 280           if ((int) flag >= (int) CASE_CAPITALIZE)
 281             inword = SYNTAX (c2) == Sword;
 282           INC_BOTH (start, i);
 283         }
 284       TEMP_SET_PT_BOTH (opoint, opoint_byte);
 285     }
 286
 287   start = XFASTINT (b);
 288   if (changed)
 289     {
 290       signal_after_change (start, end - start, end - start);
 291       update_compositions (start, end, CHECK_ALL);
 292     }
 293 }
 294
 295 DEFUN ("upcase-region", Fupcase_region, Supcase_region, 2, 2, "r",
 296        doc: /* Convert the region to upper case.  In programs, wants two arguments.
 297 These arguments specify the starting and ending character numbers of
 298 the region to operate on.  When used as a command, the text between
 299 point and the mark is operated on.
 300 See also `capitalize-region'.  */)
 301      (beg, end)
 302      Lisp_Object beg, end;
 303 {
 304   casify_region (CASE_UP, beg, end);
 305   return Qnil;
 306 }
 307
 308 DEFUN ("downcase-region", Fdowncase_region, Sdowncase_region, 2, 2, "r",
 309        doc: /* Convert the region to lower case.  In programs, wants two arguments.
 310 These arguments specify the starting and ending character numbers of
 311 the region to operate on.  When used as a command, the text between
 312 point and the mark is operated on.  */)
 313      (beg, end)
 314      Lisp_Object beg, end;
 315 {
 316   casify_region (CASE_DOWN, beg, end);
 317   return Qnil;
 318 }
 319
 320 DEFUN ("capitalize-region", Fcapitalize_region, Scapitalize_region, 2, 2, "r",
 321        doc: /* Convert the region to capitalized form.
 322 Capitalized form means each word's first character is upper case
 323 and the rest of it is lower case.
 324 In programs, give two arguments, the starting and ending
 325 character positions to operate on.  */)
 326      (beg, end)
 327      Lisp_Object beg, end;
 328 {
 329   casify_region (CASE_CAPITALIZE, beg, end);
 330   return Qnil;
 331 }
 332
 333 /* Like Fcapitalize_region but change only the initials.  */
 334
 335 DEFUN ("upcase-initials-region", Fupcase_initials_region,
 336        Supcase_initials_region, 2, 2, "r",
 337        doc: /* Upcase the initial of each word in the region.
 338 Subsequent letters of each word are not changed.
 339 In programs, give two arguments, the starting and ending
 340 character positions to operate on.  */)
 341      (beg, end)
 342      Lisp_Object beg, end;
 343 {
 344   casify_region (CASE_CAPITALIZE_UP, beg, end);
 345   return Qnil;
 346 }
 347 \f
 348 Lisp_Object
 349 operate_on_word (arg, newpoint)
 350      Lisp_Object arg;
 351      int *newpoint;
 352 {
 353   Lisp_Object val;
 354   int farend;
 355   int iarg;
 356
 357   CHECK_NUMBER (arg);
 358   iarg = XINT (arg);
 359   farend = scan_words (PT, iarg);
 360   if (!farend)
 361     farend = iarg > 0 ? ZV : BEGV;
 362
 363   *newpoint = PT > farend ? PT : farend;
 364   XSETFASTINT (val, farend);
 365
 366   return val;
 367 }
 368
 369 DEFUN ("upcase-word", Fupcase_word, Supcase_word, 1, 1, "p",
 370        doc: /* Convert following word (or ARG words) to upper case, moving over.
 371 With negative argument, convert previous words but do not move.
 372 See also `capitalize-word'.  */)
 373      (arg)
 374      Lisp_Object arg;
 375 {
 376   Lisp_Object beg, end;
 377   int newpoint;
 378   XSETFASTINT (beg, PT);
 379   end = operate_on_word (arg, &newpoint);
 380   casify_region (CASE_UP, beg, end);
 381   SET_PT (newpoint);
 382   return Qnil;
 383 }
 384
 385 DEFUN ("downcase-word", Fdowncase_word, Sdowncase_word, 1, 1, "p",
 386        doc: /* Convert following word (or ARG words) to lower case, moving over.
 387 With negative argument, convert previous words but do not move.  */)
 388      (arg)
 389      Lisp_Object arg;
 390 {
 391   Lisp_Object beg, end;
 392   int newpoint;
 393   XSETFASTINT (beg, PT);
 394   end = operate_on_word (arg, &newpoint);
 395   casify_region (CASE_DOWN, beg, end);
 396   SET_PT (newpoint);
 397   return Qnil;
 398 }
 399
 400 DEFUN ("capitalize-word", Fcapitalize_word, Scapitalize_word, 1, 1, "p",
 401        doc: /* Capitalize the following word (or ARG words), moving over.
 402 This gives the word(s) a first character in upper case
 403 and the rest lower case.
 404 With negative argument, capitalize previous words but do not move.  */)
 405      (arg)
 406      Lisp_Object arg;
 407 {
 408   Lisp_Object beg, end;
 409   int newpoint;
 410   XSETFASTINT (beg, PT);
 411   end = operate_on_word (arg, &newpoint);
 412   casify_region (CASE_CAPITALIZE, beg, end);
 413   SET_PT (newpoint);
 414   return Qnil;
 415 }
 416 \f
 417 void
 418 syms_of_casefiddle ()
 419 {
 420   Qidentity = intern ("identity");
 421   staticpro (&Qidentity);
 422   defsubr (&Supcase);
 423   defsubr (&Sdowncase);
 424   defsubr (&Scapitalize);
 425   defsubr (&Supcase_initials);
 426   defsubr (&Supcase_region);
 427   defsubr (&Sdowncase_region);
 428   defsubr (&Scapitalize_region);
 429   defsubr (&Supcase_initials_region);
 430   defsubr (&Supcase_word);
 431   defsubr (&Sdowncase_word);
 432   defsubr (&Scapitalize_word);
 433 }
 434
 435 void
 436 keys_of_casefiddle ()
 437 {
 438   initial_define_key (control_x_map, Ctl('U'), "upcase-region");
 439   Fput (intern ("upcase-region"), Qdisabled, Qt);
 440   initial_define_key (control_x_map, Ctl('L'), "downcase-region");
 441   Fput (intern ("downcase-region"), Qdisabled, Qt);
 442
 443   initial_define_key (meta_map, 'u', "upcase-word");
 444   initial_define_key (meta_map, 'l', "downcase-word");
 445   initial_define_key (meta_map, 'c', "capitalize-word");
 446 }