code.delx.au - gnu-emacs/blob - 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           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, tolen, j_byte = i;
 105               char *buf
 106                 = (char *) alloca ((len - i) * MAX_LENGTH_OF_MULTI_BYTE_FORM
 107                                    + i);
 108               unsigned char *str, workbuf[4];
 109
 110               /* Copy data already handled.  */
 111               bcopy (XSTRING (obj)->data, buf, i);
 112
 113               /* From now on, I counts bytes.  */
 114               while (i < len)
 115                 {
 116                   c = STRING_CHAR_AND_LENGTH (XSTRING (obj)->data + i,
 117                                               len - i, fromlen);
 118                   if (inword && flag != CASE_CAPITALIZE_UP)
 119                     c = DOWNCASE (c);
 120                   else if (!UPPERCASEP (c)
 121                            && (!inword || flag != CASE_CAPITALIZE_UP))
 122                     c = UPCASE1 (c);
 123                   tolen = CHAR_STRING (c, workbuf, str);
 124                   bcopy (str, buf + j_byte, tolen);
 125                   i += fromlen;
 126                   j_byte += tolen;
 127                   if ((int) flag >= (int) CASE_CAPITALIZE)
 128                     inword = SYNTAX (c) == Sword;
 129                 }
 130               obj = make_string (buf, j_byte);
 131             }
 132           return obj;
 133         }
 134       obj = wrong_type_argument (Qchar_or_string_p, obj);
 135     }
 136 }
 137
 138 DEFUN ("upcase", Fupcase, Supcase, 1, 1, 0,
 139   "Convert argument to upper case and return that.\n\
 140 The argument may be a character or string.  The result has the same type.\n\
 141 The argument object is not altered--the value is a copy.\n\
 142 See also `capitalize', `downcase' and `upcase-initials'.")
 143   (obj)
 144      Lisp_Object obj;
 145 {
 146   return casify_object (CASE_UP, obj);
 147 }
 148
 149 DEFUN ("downcase", Fdowncase, Sdowncase, 1, 1, 0,
 150   "Convert argument to lower case and return that.\n\
 151 The argument may be a character or string.  The result has the same type.\n\
 152 The argument object is not altered--the value is a copy.")
 153   (obj)
 154      Lisp_Object obj;
 155 {
 156   return casify_object (CASE_DOWN, obj);
 157 }
 158
 159 DEFUN ("capitalize", Fcapitalize, Scapitalize, 1, 1, 0,
 160   "Convert argument to capitalized form and return that.\n\
 161 This means that each word's first character is upper case\n\
 162 and the rest is lower case.\n\
 163 The argument may be a character or string.  The result has the same type.\n\
 164 The argument object is not altered--the value is a copy.")
 165   (obj)
 166      Lisp_Object obj;
 167 {
 168   return casify_object (CASE_CAPITALIZE, obj);
 169 }
 170
 171 /* Like Fcapitalize but change only the initials.  */
 172
 173 DEFUN ("upcase-initials", Fupcase_initials, Supcase_initials, 1, 1, 0,
 174   "Convert the initial of each word in the argument to upper case.\n\
 175 Do not change the other letters of each word.\n\
 176 The argument may be a character or string.  The result has the same type.\n\
 177 The argument object is not altered--the value is a copy.")
 178   (obj)
 179      Lisp_Object obj;
 180 {
 181   return casify_object (CASE_CAPITALIZE_UP, obj);
 182 }
 183 \f
 184 /* flag is CASE_UP, CASE_DOWN or CASE_CAPITALIZE or CASE_CAPITALIZE_UP.
 185    b and e specify range of buffer to operate on. */
 186
 187 void
 188 casify_region (flag, b, e)
 189      enum case_action flag;
 190      Lisp_Object b, e;
 191 {
 192   register int i;
 193   register int c;
 194   register int inword = flag == CASE_DOWN;
 195   register int multibyte = !NILP (current_buffer->enable_multibyte_characters);
 196   int start, end;
 197   int start_byte, end_byte;
 198   Lisp_Object ch, downch, val;
 199
 200   if (EQ (b, e))
 201     /* Not modifying because nothing marked */
 202     return;
 203
 204   /* If the case table is flagged as modified, rescan it.  */
 205   if (NILP (XCHAR_TABLE (current_buffer->downcase_table)->extras[1]))
 206     Fset_case_table (current_buffer->downcase_table);
 207
 208   validate_region (&b, &e);
 209   start = XFASTINT (b);
 210   end = XFASTINT (e);
 211   modify_region (current_buffer, start, end);
 212   record_change (start, end - start);
 213   start_byte = CHAR_TO_BYTE (start);
 214   end_byte = CHAR_TO_BYTE (end);
 215
 216   for (i = start_byte; i < end_byte; i++)
 217     {
 218       c = FETCH_BYTE (i);
 219       if (multibyte && c >= 0x80)
 220         /* A multibyte character can't be handled in this simple loop.  */
 221         break;
 222       if (inword && flag != CASE_CAPITALIZE_UP)
 223         c = DOWNCASE (c);
 224       else if (!UPPERCASEP (c)
 225                && (!inword || flag != CASE_CAPITALIZE_UP))
 226         c = UPCASE1 (c);
 227       FETCH_BYTE (i) = c;
 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 workbuf[4], *str;
 253
 254               /* Handle the most likely case */
 255               if (c < 0400 && c2 < 0400)
 256                 FETCH_BYTE (i) = c2;
 257               else if (fromlen = CHAR_STRING (c, workbuf, str),
 258                        tolen = CHAR_STRING (c2, workbuf, str),
 259                        fromlen == tolen)
 260                 {
 261                   for (j = 0; j < tolen; ++j)
 262                     FETCH_BYTE (i + j) = str[j];
 263                 }
 264               else
 265                 {
 266                   error ("Can't casify letters that change length");
 267 #if 0 /* This is approximately what we'd like to be able to do here */
 268                   if (tolen < fromlen)
 269                     del_range_1 (i + tolen, i + fromlen, 0);
 270                   else if (tolen > fromlen)
 271                     {
 272                       TEMP_SET_PT (i + fromlen);
 273                       insert_1 (str + fromlen, tolen - fromlen, 1, 0, 0);
 274                     }
 275 #endif
 276                 }
 277             }
 278           if ((int) flag >= (int) CASE_CAPITALIZE)
 279             inword = SYNTAX (c2) == Sword;
 280           INC_POS (i);
 281         }
 282       TEMP_SET_PT_BOTH (opoint, opoint_byte);
 283     }
 284
 285   signal_after_change (start, end - start, end - start);
 286 }
 287
 288 DEFUN ("upcase-region", Fupcase_region, Supcase_region, 2, 2, "r",
 289   "Convert the region to upper case.  In programs, wants two arguments.\n\
 290 These arguments specify the starting and ending character numbers of\n\
 291 the region to operate on.  When used as a command, the text between\n\
 292 point and the mark is operated on.\n\
 293 See also `capitalize-region'.")
 294   (beg, end)
 295      Lisp_Object beg, end;
 296 {
 297   casify_region (CASE_UP, beg, end);
 298   return Qnil;
 299 }
 300
 301 DEFUN ("downcase-region", Fdowncase_region, Sdowncase_region, 2, 2, "r",
 302   "Convert the region to lower case.  In programs, wants two arguments.\n\
 303 These arguments specify the starting and ending character numbers of\n\
 304 the region to operate on.  When used as a command, the text between\n\
 305 point and the mark is operated on.")
 306   (beg, end)
 307      Lisp_Object beg, end;
 308 {
 309   casify_region (CASE_DOWN, beg, end);
 310   return Qnil;
 311 }
 312
 313 DEFUN ("capitalize-region", Fcapitalize_region, Scapitalize_region, 2, 2, "r",
 314   "Convert the region to capitalized form.\n\
 315 Capitalized form means each word's first character is upper case\n\
 316 and the rest of it is lower case.\n\
 317 In programs, give two arguments, the starting and ending\n\
 318 character positions to operate on.")
 319   (beg, end)
 320      Lisp_Object beg, end;
 321 {
 322   casify_region (CASE_CAPITALIZE, beg, end);
 323   return Qnil;
 324 }
 325
 326 /* Like Fcapitalize_region but change only the initials.  */
 327
 328 DEFUN ("upcase-initials-region", Fupcase_initials_region,
 329        Supcase_initials_region, 2, 2, "r",
 330   "Upcase the initial of each word in the region.\n\
 331 Subsequent letters of each word are not changed.\n\
 332 In programs, give two arguments, the starting and ending\n\
 333 character positions to operate on.")
 334   (beg, end)
 335      Lisp_Object beg, end;
 336 {
 337   casify_region (CASE_CAPITALIZE_UP, beg, end);
 338   return Qnil;
 339 }
 340 \f
 341 Lisp_Object
 342 operate_on_word (arg, newpoint)
 343      Lisp_Object arg;
 344      int *newpoint;
 345 {
 346   Lisp_Object val;
 347   int farend;
 348   int iarg;
 349
 350   CHECK_NUMBER (arg, 0);
 351   iarg = XINT (arg);
 352   farend = scan_words (PT, iarg);
 353   if (!farend)
 354     farend = iarg > 0 ? ZV : BEGV;
 355
 356   *newpoint = PT > farend ? PT : farend;
 357   XSETFASTINT (val, farend);
 358
 359   return val;
 360 }
 361
 362 DEFUN ("upcase-word", Fupcase_word, Supcase_word, 1, 1, "p",
 363   "Convert following word (or ARG words) to upper case, moving over.\n\
 364 With negative argument, convert previous words but do not move.\n\
 365 See also `capitalize-word'.")
 366   (arg)
 367      Lisp_Object arg;
 368 {
 369   Lisp_Object beg, end;
 370   int newpoint;
 371   XSETFASTINT (beg, PT);
 372   end = operate_on_word (arg, &newpoint);
 373   casify_region (CASE_UP, beg, end);
 374   SET_PT (newpoint);
 375   return Qnil;
 376 }
 377
 378 DEFUN ("downcase-word", Fdowncase_word, Sdowncase_word, 1, 1, "p",
 379   "Convert following word (or ARG words) to lower case, moving over.\n\
 380 With negative argument, convert previous words but do not move.")
 381   (arg)
 382      Lisp_Object arg;
 383 {
 384   Lisp_Object beg, end;
 385   int newpoint;
 386   XSETFASTINT (beg, PT);
 387   end = operate_on_word (arg, &newpoint);
 388   casify_region (CASE_DOWN, beg, end);
 389   SET_PT (newpoint);
 390   return Qnil;
 391 }
 392
 393 DEFUN ("capitalize-word", Fcapitalize_word, Scapitalize_word, 1, 1, "p",
 394   "Capitalize the following word (or ARG words), moving over.\n\
 395 This gives the word(s) a first character in upper case\n\
 396 and the rest lower case.\n\
 397 With negative argument, capitalize previous words but do not move.")
 398   (arg)
 399      Lisp_Object arg;
 400 {
 401   Lisp_Object beg, end;
 402   int newpoint;
 403   XSETFASTINT (beg, PT);
 404   end = operate_on_word (arg, &newpoint);
 405   casify_region (CASE_CAPITALIZE, beg, end);
 406   SET_PT (newpoint);
 407   return Qnil;
 408 }
 409 \f
 410 void
 411 syms_of_casefiddle ()
 412 {
 413   Qidentity = intern ("identity");
 414   staticpro (&Qidentity);
 415   defsubr (&Supcase);
 416   defsubr (&Sdowncase);
 417   defsubr (&Scapitalize);
 418   defsubr (&Supcase_initials);
 419   defsubr (&Supcase_region);
 420   defsubr (&Sdowncase_region);
 421   defsubr (&Scapitalize_region);
 422   defsubr (&Supcase_initials_region);
 423   defsubr (&Supcase_word);
 424   defsubr (&Sdowncase_word);
 425   defsubr (&Scapitalize_word);
 426 }
 427
 428 void
 429 keys_of_casefiddle ()
 430 {
 431   initial_define_key (control_x_map, Ctl('U'), "upcase-region");
 432   Fput (intern ("upcase-region"), Qdisabled, Qt);
 433   initial_define_key (control_x_map, Ctl('L'), "downcase-region");
 434   Fput (intern ("downcase-region"), Qdisabled, Qt);
 435
 436   initial_define_key (meta_map, 'u', "upcase-word");
 437   initial_define_key (meta_map, 'l', "downcase-word");
 438   initial_define_key (meta_map, 'c', "capitalize-word");
 439 }