category.c 16.1 KB
Newer Older
Karl Heuer's avatar
Karl Heuer committed
1
/* GNU Emacs routines to deal with category tables.
2

3
Copyright (C) 1998, 2001-2014 Free Software Foundation, Inc.
4 5 6 7 8 9 10
Copyright (C) 1995, 1996, 1997, 1998, 1999, 2000, 2001, 2002, 2003, 2004,
  2005, 2006, 2007, 2008, 2009, 2010, 2011
  National Institute of Advanced Industrial Science and Technology (AIST)
  Registration Number H14PRO021
Copyright (C) 2003
  National Institute of Advanced Industrial Science and Technology (AIST)
  Registration Number H13PRO009
Karl Heuer's avatar
Karl Heuer committed
11 12 13

This file is part of GNU Emacs.

14
GNU Emacs is free software: you can redistribute it and/or modify
Karl Heuer's avatar
Karl Heuer committed
15
it under the terms of the GNU General Public License as published by
16 17
the Free Software Foundation, either version 3 of the License, or
(at your option) any later version.
Karl Heuer's avatar
Karl Heuer committed
18 19 20 21 22 23 24

GNU Emacs is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
GNU General Public License for more details.

You should have received a copy of the GNU General Public License
25
along with GNU Emacs.  If not, see <http://www.gnu.org/licenses/>.  */
Karl Heuer's avatar
Karl Heuer committed
26 27 28 29 30 31


/* Here we handle three objects: category, category set, and category
   table.  Read comments in the file category.h to understand them.  */

#include <config.h>
32

Karl Heuer's avatar
Karl Heuer committed
33
#include "lisp.h"
Kenichi Handa's avatar
Kenichi Handa committed
34
#include "character.h"
35
#include "buffer.h"
Karl Heuer's avatar
Karl Heuer committed
36 37
#include "charset.h"
#include "category.h"
Stefan Monnier's avatar
Stefan Monnier committed
38
#include "keymap.h"
Karl Heuer's avatar
Karl Heuer committed
39

Paul Eggert's avatar
Paul Eggert committed
40
/* This setter is used only in this file, so it can be private.  */
41
static void
Paul Eggert's avatar
Paul Eggert committed
42 43 44 45 46
bset_category_table (struct buffer *b, Lisp_Object val)
{
  b->INTERNAL_FIELD (category_table) = val;
}

Karl Heuer's avatar
Karl Heuer committed
47 48 49 50 51 52 53 54 55 56
/* The version number of the latest category table.  Each category
   table has a unique version number.  It is assigned a new number
   also when it is modified.  When a regular expression is compiled
   into the struct re_pattern_buffer, the version number of the
   category table (of the current buffer) at that moment is also
   embedded in the structure.

   For the moment, we are not using this feature.  */
static int category_table_version;

57
static Lisp_Object Qcategory_table, Qcategoryp, Qcategorysetp, Qcategory_table_p;
Karl Heuer's avatar
Karl Heuer committed
58 59 60

/* Category set staff.  */

61
static Lisp_Object
62
hash_get_category_set (Lisp_Object table, Lisp_Object category_set)
63 64
{
  struct Lisp_Hash_Table *h;
65
  ptrdiff_t i;
66
  EMACS_UINT hash;
67 68

  if (NILP (XCHAR_TABLE (table)->extras[1]))
69
    set_char_table_extras
70
      (table, 1,
71
       make_hash_table (hashtest_equal, make_number (DEFAULT_HASH_SIZE),
72 73
			make_float (DEFAULT_REHASH_SIZE),
			make_float (DEFAULT_REHASH_THRESHOLD),
74
			Qnil));
75 76 77 78 79 80 81 82
  h = XHASH_TABLE (XCHAR_TABLE (table)->extras[1]);
  i = hash_lookup (h, category_set, &hash);
  if (i >= 0)
    return HASH_KEY (h, i);
  hash_put (h, category_set, Qnil, hash);
  return category_set;
}

83 84 85 86 87 88 89
/* Make CATEGORY_SET include (if VAL) or exclude (if !VAL) CATEGORY.  */

static void
set_category_set (Lisp_Object category_set, EMACS_INT category, bool val)
{
  bool_vector_set (category_set, category, val);
}
90

Karl Heuer's avatar
Karl Heuer committed
91
DEFUN ("make-category-set", Fmake_category_set, Smake_category_set, 1, 1, 0,
92 93 94 95
       doc: /* Return a newly created category-set which contains CATEGORIES.
CATEGORIES is a string of category mnemonics.
The value is a bool-vector which has t at the indices corresponding to
those categories.  */)
Dan Nicolaescu's avatar
Dan Nicolaescu committed
96
  (Lisp_Object categories)
Karl Heuer's avatar
Karl Heuer committed
97 98
{
  Lisp_Object val;
99
  ptrdiff_t len;
Karl Heuer's avatar
Karl Heuer committed
100

101
  CHECK_STRING (categories);
Karl Heuer's avatar
Karl Heuer committed
102 103
  val = MAKE_CATEGORY_SET;

104
  if (STRING_MULTIBYTE (categories))
105
    error ("Multibyte string in `make-category-set'");
106

107
  len = SCHARS (categories);
Karl Heuer's avatar
Karl Heuer committed
108 109
  while (--len >= 0)
    {
110 111
      unsigned char cat = SREF (categories, len);
      Lisp_Object category = make_number (cat);
Karl Heuer's avatar
Karl Heuer committed
112

113
      CHECK_CATEGORY (category);
114
      set_category_set (val, cat, 1);
Karl Heuer's avatar
Karl Heuer committed
115 116 117 118 119 120 121
    }
  return val;
}


/* Category staff.  */

122
static Lisp_Object check_category_table (Lisp_Object table);
Karl Heuer's avatar
Karl Heuer committed
123 124

DEFUN ("define-category", Fdefine_category, Sdefine_category, 2, 3, 0,
Kenichi Handa's avatar
Kenichi Handa committed
125 126
       doc: /* Define CATEGORY as a category which is described by DOCSTRING.
CATEGORY should be an ASCII printing character in the range ` ' to `~'.
127 128 129
DOCSTRING is the documentation string of the category.  The first line
should be a terse text (preferably less than 16 characters),
and the rest lines should be the full description.
130
The category is defined only in category table TABLE, which defaults to
Kenichi Handa's avatar
Kenichi Handa committed
131
the current buffer's category table.  */)
Dan Nicolaescu's avatar
Dan Nicolaescu committed
132
  (Lisp_Object category, Lisp_Object docstring, Lisp_Object table)
Karl Heuer's avatar
Karl Heuer committed
133
{
134 135
  CHECK_CATEGORY (category);
  CHECK_STRING (docstring);
Karl Heuer's avatar
Karl Heuer committed
136 137 138
  table = check_category_table (table);

  if (!NILP (CATEGORY_DOCSTRING (table, XFASTINT (category))))
139
    error ("Category `%c' is already defined", (int) XFASTINT (category));
140 141
  if (!NILP (Vpurify_flag))
    docstring = Fpurecopy (docstring);
142
  SET_CATEGORY_DOCSTRING (table, XFASTINT (category), docstring);
Karl Heuer's avatar
Karl Heuer committed
143 144 145 146 147

  return Qnil;
}

DEFUN ("category-docstring", Fcategory_docstring, Scategory_docstring, 1, 2, 0,
Kenichi Handa's avatar
Kenichi Handa committed
148 149 150
       doc: /* Return the documentation string of CATEGORY, as defined in TABLE.
TABLE should be a category table and defaults to the current buffer's
category table.  */)
Dan Nicolaescu's avatar
Dan Nicolaescu committed
151
  (Lisp_Object category, Lisp_Object table)
Karl Heuer's avatar
Karl Heuer committed
152
{
153
  CHECK_CATEGORY (category);
Karl Heuer's avatar
Karl Heuer committed
154 155 156 157 158 159 160
  table = check_category_table (table);

  return CATEGORY_DOCSTRING (table, XFASTINT (category));
}

DEFUN ("get-unused-category", Fget_unused_category, Sget_unused_category,
       0, 1, 0,
Kenichi Handa's avatar
Kenichi Handa committed
161
       doc: /* Return a category which is not yet defined in TABLE.
162
If no category remains available, return nil.
Kenichi Handa's avatar
Kenichi Handa committed
163
The optional argument TABLE specifies which category table to modify;
164
it defaults to the current buffer's category table.  */)
Dan Nicolaescu's avatar
Dan Nicolaescu committed
165
  (Lisp_Object table)
Karl Heuer's avatar
Karl Heuer committed
166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181
{
  int i;

  table = check_category_table (table);

  for (i = ' '; i <= '~'; i++)
    if (NILP (CATEGORY_DOCSTRING (table, i)))
      return make_number (i);

  return Qnil;
}


/* Category-table staff.  */

DEFUN ("category-table-p", Fcategory_table_p, Scategory_table_p, 1, 1, 0,
182
       doc: /* Return t if ARG is a category table.  */)
Dan Nicolaescu's avatar
Dan Nicolaescu committed
183
  (Lisp_Object arg)
Karl Heuer's avatar
Karl Heuer committed
184 185
{
  if (CHAR_TABLE_P (arg)
186
      && EQ (XCHAR_TABLE (arg)->purpose, Qcategory_table))
Karl Heuer's avatar
Karl Heuer committed
187 188 189 190 191 192 193 194 195
    return Qt;
  return Qnil;
}

/* If TABLE is nil, return the current category table.  If TABLE is
   not nil, check the validity of TABLE as a category table.  If
   valid, return TABLE itself, but if not valid, signal an error of
   wrong-type-argument.  */

196
static Lisp_Object
197
check_category_table (Lisp_Object table)
Karl Heuer's avatar
Karl Heuer committed
198 199
{
  if (NILP (table))
Tom Tromey's avatar
Tom Tromey committed
200
    return BVAR (current_buffer, category_table);
201
  CHECK_TYPE (!NILP (Fcategory_table_p (table)), Qcategory_table_p, table);
Karl Heuer's avatar
Karl Heuer committed
202
  return table;
203
}
Karl Heuer's avatar
Karl Heuer committed
204 205

DEFUN ("category-table", Fcategory_table, Scategory_table, 0, 0, 0,
206 207
       doc: /* Return the current category table.
This is the one specified by the current buffer.  */)
Dan Nicolaescu's avatar
Dan Nicolaescu committed
208
  (void)
Karl Heuer's avatar
Karl Heuer committed
209
{
Tom Tromey's avatar
Tom Tromey committed
210
  return BVAR (current_buffer, category_table);
Karl Heuer's avatar
Karl Heuer committed
211 212 213 214
}

DEFUN ("standard-category-table", Fstandard_category_table,
   Sstandard_category_table, 0, 0, 0,
215 216
       doc: /* Return the standard category table.
This is the one used for new buffers.  */)
Dan Nicolaescu's avatar
Dan Nicolaescu committed
217
  (void)
Karl Heuer's avatar
Karl Heuer committed
218 219 220 221
{
  return Vstandard_category_table;
}

Kenichi Handa's avatar
Kenichi Handa committed
222 223

static void
224
copy_category_entry (Lisp_Object table, Lisp_Object c, Lisp_Object val)
Kenichi Handa's avatar
Kenichi Handa committed
225
{
226
  val = Fcopy_sequence (val);
Kenichi Handa's avatar
Kenichi Handa committed
227 228
  if (CONSP (c))
    char_table_set_range (table, XINT (XCAR (c)), XINT (XCDR (c)), val);
229
  else
Kenichi Handa's avatar
Kenichi Handa committed
230
    char_table_set (table, XINT (c), val);
Kenichi Handa's avatar
Kenichi Handa committed
231 232
}

Karl Heuer's avatar
Karl Heuer committed
233 234
/* Return a copy of category table TABLE.  We can't simply use the
   function copy-sequence because no contents should be shared between
235
   the original and the copy.  This function is called recursively by
236
   binding TABLE to a sub char table.  */
Karl Heuer's avatar
Karl Heuer committed
237

238
static Lisp_Object
239
copy_category_table (Lisp_Object table)
Karl Heuer's avatar
Karl Heuer committed
240
{
Kenichi Handa's avatar
Kenichi Handa committed
241
  table = copy_char_table (table);
Karl Heuer's avatar
Karl Heuer committed
242

Kenichi Handa's avatar
Kenichi Handa committed
243
  if (! NILP (XCHAR_TABLE (table)->defalt))
Paul Eggert's avatar
Paul Eggert committed
244 245
    set_char_table_defalt (table,
			   Fcopy_sequence (XCHAR_TABLE (table)->defalt));
246
  set_char_table_extras
247
    (table, 0, Fcopy_sequence (XCHAR_TABLE (table)->extras[0]));
Kenichi Handa's avatar
Kenichi Handa committed
248
  map_char_table (copy_category_entry, Qnil, table, table);
249

Karl Heuer's avatar
Karl Heuer committed
250 251 252 253 254
  return table;
}

DEFUN ("copy-category-table", Fcopy_category_table, Scopy_category_table,
       0, 1, 0,
255 256
       doc: /* Construct a new category table and return it.
It is a copy of the TABLE, which defaults to the standard category table.  */)
Dan Nicolaescu's avatar
Dan Nicolaescu committed
257
  (Lisp_Object table)
Karl Heuer's avatar
Karl Heuer committed
258 259 260 261 262 263
{
  if (!NILP (table))
    check_category_table (table);
  else
    table = Vstandard_category_table;

264
  return copy_category_table (table);
Karl Heuer's avatar
Karl Heuer committed
265 266
}

267 268
DEFUN ("make-category-table", Fmake_category_table, Smake_category_table,
       0, 0, 0,
269
       doc: /* Construct a new and empty category table and return it.  */)
Dan Nicolaescu's avatar
Dan Nicolaescu committed
270
  (void)
271 272
{
  Lisp_Object val;
Kenichi Handa's avatar
Kenichi Handa committed
273
  int i;
274 275

  val = Fmake_char_table (Qcategory_table, Qnil);
Paul Eggert's avatar
Paul Eggert committed
276
  set_char_table_defalt (val, MAKE_CATEGORY_SET);
Kenichi Handa's avatar
Kenichi Handa committed
277
  for (i = 0; i < (1 << CHARTAB_SIZE_BITS_0); i++)
278
    set_char_table_contents (val, i, MAKE_CATEGORY_SET);
279 280 281 282 283
  Fset_char_table_extra_slot (val, make_number (0),
			      Fmake_vector (make_number (95), Qnil));
  return val;
}

Karl Heuer's avatar
Karl Heuer committed
284
DEFUN ("set-category-table", Fset_category_table, Sset_category_table, 1, 1, 0,
Kenichi Handa's avatar
Kenichi Handa committed
285 286
       doc: /* Specify TABLE as the category table for the current buffer.
Return TABLE.  */)
Dan Nicolaescu's avatar
Dan Nicolaescu committed
287
  (Lisp_Object table)
Karl Heuer's avatar
Karl Heuer committed
288
{
289
  int idx;
Karl Heuer's avatar
Karl Heuer committed
290
  table = check_category_table (table);
Paul Eggert's avatar
Paul Eggert committed
291
  bset_category_table (current_buffer, table);
Karl Heuer's avatar
Karl Heuer committed
292
  /* Indicate that this buffer now has a specified category table.  */
Gerd Moellmann's avatar
Gerd Moellmann committed
293 294
  idx = PER_BUFFER_VAR_IDX (category_table);
  SET_PER_BUFFER_VALUE_P (current_buffer, idx, 1);
Karl Heuer's avatar
Karl Heuer committed
295 296 297 298
  return table;
}


Kenichi Handa's avatar
Kenichi Handa committed
299
Lisp_Object
300
char_category_set (int c)
Kenichi Handa's avatar
Kenichi Handa committed
301
{
Tom Tromey's avatar
Tom Tromey committed
302
  return CHAR_TABLE_REF (BVAR (current_buffer, category_table), c);
Kenichi Handa's avatar
Kenichi Handa committed
303 304
}

Karl Heuer's avatar
Karl Heuer committed
305
DEFUN ("char-category-set", Fchar_category_set, Schar_category_set, 1, 1, 0,
306 307
       doc: /* Return the category set of CHAR.
usage: (char-category-set CHAR)  */)
Dan Nicolaescu's avatar
Dan Nicolaescu committed
308
  (Lisp_Object ch)
Karl Heuer's avatar
Karl Heuer committed
309
{
310
  CHECK_CHARACTER (ch);
Karl Heuer's avatar
Karl Heuer committed
311 312 313 314 315
  return CATEGORY_SET (XFASTINT (ch));
}

DEFUN ("category-set-mnemonics", Fcategory_set_mnemonics,
       Scategory_set_mnemonics, 1, 1, 0,
316 317
       doc: /* Return a string containing mnemonics of the categories in CATEGORY-SET.
CATEGORY-SET is a bool-vector, and the categories \"in\" it are those
Kenichi Handa's avatar
Kenichi Handa committed
318
that are indexes where t occurs in the bool-vector.
319
The return value is a string containing those same categories.  */)
Dan Nicolaescu's avatar
Dan Nicolaescu committed
320
  (Lisp_Object category_set)
Karl Heuer's avatar
Karl Heuer committed
321 322 323 324
{
  int i, j;
  char str[96];

325
  CHECK_CATEGORY_SET (category_set);
Karl Heuer's avatar
Karl Heuer committed
326 327 328 329 330 331 332 333 334 335 336 337

  j = 0;
  for (i = 32; i < 127; i++)
    if (CATEGORY_MEMBER (i, category_set))
      str[j++] = i;
  str[j] = '\0';

  return build_string (str);
}

DEFUN ("modify-category-entry", Fmodify_category_entry,
       Smodify_category_entry, 2, 4, 0,
338 339
       doc: /* Modify the category set of CHARACTER by adding CATEGORY to it.
The category is changed only for table TABLE, which defaults to
340 341 342
the current buffer's category table.
CHARACTER can be either a single character or a cons representing the
lower and upper ends of an inclusive character range to modify.
343 344
If optional fourth argument RESET is non-nil,
then delete CATEGORY from the category set instead of adding it.  */)
Dan Nicolaescu's avatar
Dan Nicolaescu committed
345
  (Lisp_Object character, Lisp_Object category, Lisp_Object table, Lisp_Object reset)
Karl Heuer's avatar
Karl Heuer committed
346
{
347
  bool set_value;	/* Actual value to be set in category sets.  */
348
  Lisp_Object category_set;
Kenichi Handa's avatar
Kenichi Handa committed
349 350
  int start, end;
  int from, to;
Karl Heuer's avatar
Karl Heuer committed
351

Kenichi Handa's avatar
Kenichi Handa committed
352
  if (INTEGERP (character))
Karl Heuer's avatar
Karl Heuer committed
353
    {
Kenichi Handa's avatar
Kenichi Handa committed
354 355
      CHECK_CHARACTER (character);
      start = end = XFASTINT (character);
Karl Heuer's avatar
Karl Heuer committed
356
    }
Kenichi Handa's avatar
Kenichi Handa committed
357
  else
Karl Heuer's avatar
Karl Heuer committed
358
    {
Kenichi Handa's avatar
Kenichi Handa committed
359
      CHECK_CONS (character);
Kenichi Handa's avatar
Kenichi Handa committed
360 361
      CHECK_CHARACTER_CAR (character);
      CHECK_CHARACTER_CDR (character);
Kenichi Handa's avatar
Kenichi Handa committed
362 363
      start = XFASTINT (XCAR (character));
      end = XFASTINT (XCDR (character));
Karl Heuer's avatar
Karl Heuer committed
364 365
    }

366
  CHECK_CATEGORY (category);
Karl Heuer's avatar
Karl Heuer committed
367 368 369
  table = check_category_table (table);

  if (NILP (CATEGORY_DOCSTRING (table, XFASTINT (category))))
370
    error ("Undefined category: %c", (int) XFASTINT (category));
Karl Heuer's avatar
Karl Heuer committed
371

372
  set_value = NILP (reset);
Karl Heuer's avatar
Karl Heuer committed
373

Kenichi Handa's avatar
Kenichi Handa committed
374
  while (start <= end)
Karl Heuer's avatar
Karl Heuer committed
375
    {
376
      from = start, to = end;
Kenichi Handa's avatar
Kenichi Handa committed
377
      category_set = char_table_ref_and_range (table, start, &from, &to);
378
      if (CATEGORY_MEMBER (XFASTINT (category), category_set) != NILP (reset))
Kenichi Handa's avatar
Kenichi Handa committed
379
	{
380
	  category_set = Fcopy_sequence (category_set);
381
	  set_category_set (category_set, XFASTINT (category), set_value);
382 383
	  category_set = hash_get_category_set (table, category_set);
	  char_table_set_range (table, start, to, category_set);
Kenichi Handa's avatar
Kenichi Handa committed
384 385
	}
      start = to + 1;
Karl Heuer's avatar
Karl Heuer committed
386 387 388 389 390
    }

  return Qnil;
}

391 392
/* Return true if there is a word boundary between two word-constituent
   characters C1 and C2 if they appear in this order.
Karl Heuer's avatar
Karl Heuer committed
393 394 395
   Use the macro WORD_BOUNDARY_P instead of calling this function
   directly.  */

396
bool
397
word_boundary_p (int c1, int c2)
Karl Heuer's avatar
Karl Heuer committed
398 399 400
{
  Lisp_Object category_set1, category_set2;
  Lisp_Object tail;
401
  bool default_result;
Karl Heuer's avatar
Karl Heuer committed
402

403 404
  if (EQ (CHAR_TABLE_REF (Vchar_script_table, c1),
	  CHAR_TABLE_REF (Vchar_script_table, c2)))
Karl Heuer's avatar
Karl Heuer committed
405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421
    {
      tail = Vword_separating_categories;
      default_result = 0;
    }
  else
    {
      tail = Vword_combining_categories;
      default_result = 1;
    }

  category_set1 = CATEGORY_SET (c1);
  if (NILP (category_set1))
    return default_result;
  category_set2 = CATEGORY_SET (c2);
  if (NILP (category_set2))
    return default_result;

422
  for (; CONSP (tail); tail = XCDR (tail))
Karl Heuer's avatar
Karl Heuer committed
423
    {
424
      Lisp_Object elt = XCAR (tail);
Karl Heuer's avatar
Karl Heuer committed
425 426

      if (CONSP (elt)
427 428
	  && (NILP (XCAR (elt))
	      || (CATEGORYP (XCAR (elt))
429 430
		  && CATEGORY_MEMBER (XFASTINT (XCAR (elt)), category_set1)
		  && ! CATEGORY_MEMBER (XFASTINT (XCAR (elt)), category_set2)))
431 432
	  && (NILP (XCDR (elt))
	      || (CATEGORYP (XCDR (elt))
433
		  && ! CATEGORY_MEMBER (XFASTINT (XCDR (elt)), category_set1)
434
		  && CATEGORY_MEMBER (XFASTINT (XCDR (elt)), category_set2))))
Karl Heuer's avatar
Karl Heuer committed
435 436 437 438 439 440
	return !default_result;
    }
  return default_result;
}


Andreas Schwab's avatar
Andreas Schwab committed
441
void
442
init_category_once (void)
Karl Heuer's avatar
Karl Heuer committed
443 444
{
  /* This has to be done here, before we call Fmake_char_table.  */
445
  DEFSYM (Qcategory_table, "category-table");
Karl Heuer's avatar
Karl Heuer committed
446 447 448
  Fput (Qcategory_table, Qchar_table_extra_slots, make_number (2));

  Vstandard_category_table = Fmake_char_table (Qcategory_table, Qnil);
449
  /* Set a category set which contains nothing to the default.  */
Paul Eggert's avatar
Paul Eggert committed
450
  set_char_table_defalt (Vstandard_category_table, MAKE_CATEGORY_SET);
451
  Fset_char_table_extra_slot (Vstandard_category_table, make_number (0),
Karl Heuer's avatar
Karl Heuer committed
452 453 454
			      Fmake_vector (make_number (95), Qnil));
}

Andreas Schwab's avatar
Andreas Schwab committed
455
void
456
syms_of_category (void)
Karl Heuer's avatar
Karl Heuer committed
457
{
458 459 460
  DEFSYM (Qcategoryp, "categoryp");
  DEFSYM (Qcategorysetp, "categorysetp");
  DEFSYM (Qcategory_table_p, "category-table-p");
Karl Heuer's avatar
Karl Heuer committed
461

462
  DEFVAR_LISP ("word-combining-categories", Vword_combining_categories,
463 464 465
	       doc: /* List of pair (cons) of categories to determine word boundary.

Emacs treats a sequence of word constituent characters as a single
Glenn Morris's avatar
Glenn Morris committed
466
word (i.e. finds no word boundary between them) only if they belong to
467
the same script.  But, exceptions are allowed in the following cases.
468

469
\(1) The case that characters are in different scripts is controlled
470 471
by the variable `word-combining-categories'.

472
Emacs finds no word boundary between characters of different scripts
473 474 475 476 477 478
if they have categories matching some element of this list.

More precisely, if an element of this list is a cons of category CAT1
and CAT2, and a multibyte character C1 which has CAT1 is followed by
C2 which has CAT2, there's no word boundary between C1 and C2.

479 480 481
For instance, to tell that Han characters followed by Hiragana
characters can form a single word, the element `(?C . ?H)' should be
in this list.
482

483
\(2) The case that character are in the same script is controlled by
484 485
the variable `word-separating-categories'.

486
Emacs finds a word boundary between characters of the same script
487 488 489
if they have categories matching some element of this list.

More precisely, if an element of this list is a cons of category CAT1
490 491 492
and CAT2, and a multibyte character C1 which has CAT1 but not CAT2 is
followed by C2 which has CAT2 but not CAT1, there's a word boundary
between C1 and C2.
493

494 495 496
For instance, to tell that there's a word boundary between Hiragana
and Katakana (both are in the same script `kana'),
the element `(?H . ?K) should be in this list.  */);
Karl Heuer's avatar
Karl Heuer committed
497 498 499

  Vword_combining_categories = Qnil;

500
  DEFVAR_LISP ("word-separating-categories", Vword_separating_categories,
501 502
	       doc: /* List of pair (cons) of categories to determine word boundary.
See the documentation of the variable `word-combining-categories'.  */);
Karl Heuer's avatar
Karl Heuer committed
503 504 505 506 507 508 509 510 511 512 513

  Vword_separating_categories = Qnil;

  defsubr (&Smake_category_set);
  defsubr (&Sdefine_category);
  defsubr (&Scategory_docstring);
  defsubr (&Sget_unused_category);
  defsubr (&Scategory_table_p);
  defsubr (&Scategory_table);
  defsubr (&Sstandard_category_table);
  defsubr (&Scopy_category_table);
514
  defsubr (&Smake_category_table);
Karl Heuer's avatar
Karl Heuer committed
515 516 517 518 519 520 521
  defsubr (&Sset_category_table);
  defsubr (&Schar_category_set);
  defsubr (&Scategory_set_mnemonics);
  defsubr (&Smodify_category_entry);

  category_table_version = 0;
}