src/category.c

   1 /* GNU Emacs routines to deal with category tables.
   2    Copyright (C) 1995, 1997 Electrotechnical Laboratory, JAPAN.
   3      Licensed to the Free Software Foundation.
   4    Copyright (C) 2003
   5      National Institute of Advanced Industrial Science and Technology (AIST)
   6      Registration Number H13PRO009
   7
   8 This file is part of GNU Emacs.
   9
  10 GNU Emacs is free software; you can redistribute it and/or modify
  11 it under the terms of the GNU General Public License as published by
  12 the Free Software Foundation; either version 2, or (at your option)
  13 any later version.
  14
  15 GNU Emacs is distributed in the hope that it will be useful,
  16 but WITHOUT ANY WARRANTY; without even the implied warranty of
  17 MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  18 GNU General Public License for more details.
  19
  20 You should have received a copy of the GNU General Public License
  21 along with GNU Emacs; see the file COPYING.  If not, write to
  22 the Free Software Foundation, Inc., 59 Temple Place - Suite 330,
  23 Boston, MA 02111-1307, USA.  */
  24
  25
  26 /* Here we handle three objects: category, category set, and category
  27    table.  Read comments in the file category.h to understand them.  */
  28
  29 #include <config.h>
  30 #include <ctype.h>
  31 #include "lisp.h"
  32 #include "buffer.h"
  33 #include "character.h"
  34 #include "charset.h"
  35 #include "category.h"
  36 #include "keymap.h"
  37
  38 /* The version number of the latest category table.  Each category
  39    table has a unique version number.  It is assigned a new number
  40    also when it is modified.  When a regular expression is compiled
  41    into the struct re_pattern_buffer, the version number of the
  42    category table (of the current buffer) at that moment is also
  43    embedded in the structure.
  44
  45    For the moment, we are not using this feature.  */
  46 static int category_table_version;
  47
  48 Lisp_Object Qcategory_table, Qcategoryp, Qcategorysetp, Qcategory_table_p;
  49
  50 /* Variables to determine word boundary.  */
  51 Lisp_Object Vword_combining_categories, Vword_separating_categories;
  52
  53 /* Temporary internal variable used in macro CHAR_HAS_CATEGORY.  */
  54 Lisp_Object _temp_category_set;
  55
  56 \f
  57 /* Category set staff.  */
  58
  59 DEFUN ("make-category-set", Fmake_category_set, Smake_category_set, 1, 1, 0,
  60        doc: /* Return a newly created category-set which contains CATEGORIES.
  61 CATEGORIES is a string of category mnemonics.
  62 The value is a bool-vector which has t at the indices corresponding to
  63 those categories.  */)
  64      (categories)
  65      Lisp_Object categories;
  66 {
  67   Lisp_Object val;
  68   int len;
  69
  70   CHECK_STRING (categories);
  71   val = MAKE_CATEGORY_SET;
  72
  73   if (STRING_MULTIBYTE (categories))
  74     error ("Multibyte string in make-category-set");
  75
  76   len = SCHARS (categories);
  77   while (--len >= 0)
  78     {
  79       Lisp_Object category;
  80
  81       XSETFASTINT (category, SREF (categories, len));
  82       CHECK_CATEGORY (category);
  83       SET_CATEGORY_SET (val, category, Qt);
  84     }
  85   return val;
  86 }
  87
  88 \f
  89 /* Category staff.  */
  90
  91 Lisp_Object check_category_table ();
  92
  93 DEFUN ("define-category", Fdefine_category, Sdefine_category, 2, 3, 0,
  94        doc: /* Define CHAR as a category which is described by DOCSTRING.
  95 CHAR should be an ASCII printing character in the range ` ' to `~'.
  96 DOCSTRING is a documentation string of the category.
  97 The category is defined only in category table TABLE, which defaults to
  98  the current buffer's category table.  */)
  99      (category, docstring, table)
 100      Lisp_Object category, docstring, table;
 101 {
 102   CHECK_CATEGORY (category);
 103   CHECK_STRING (docstring);
 104   table = check_category_table (table);
 105
 106   if (!NILP (CATEGORY_DOCSTRING (table, XFASTINT (category))))
 107     error ("Category `%c' is already defined", XFASTINT (category));
 108   CATEGORY_DOCSTRING (table, XFASTINT (category)) = docstring;
 109
 110   return Qnil;
 111 }
 112
 113 DEFUN ("category-docstring", Fcategory_docstring, Scategory_docstring, 1, 2, 0,
 114        doc: /* Return the documentation string of CATEGORY, as defined in CATEGORY-TABLE.  */)
 115      (category, table)
 116      Lisp_Object category, table;
 117 {
 118   CHECK_CATEGORY (category);
 119   table = check_category_table (table);
 120
 121   return CATEGORY_DOCSTRING (table, XFASTINT (category));
 122 }
 123
 124 DEFUN ("get-unused-category", Fget_unused_category, Sget_unused_category,
 125        0, 1, 0,
 126        doc: /* Return a category which is not yet defined in CATEGORY-TABLE.
 127 If no category remains available, return nil.
 128 The optional argument CATEGORY-TABLE
 129 specifies which category table to modify;
 130 it defaults to the current buffer's category table.  */)
 131      (table)
 132      Lisp_Object table;
 133 {
 134   int i;
 135
 136   table = check_category_table (table);
 137
 138   for (i = ' '; i <= '~'; i++)
 139     if (NILP (CATEGORY_DOCSTRING (table, i)))
 140       return make_number (i);
 141
 142   return Qnil;
 143 }
 144
 145 \f
 146 /* Category-table staff.  */
 147
 148 DEFUN ("category-table-p", Fcategory_table_p, Scategory_table_p, 1, 1, 0,
 149        doc: /* Return t if ARG is a category table.  */)
 150      (arg)
 151      Lisp_Object arg;
 152 {
 153   if (CHAR_TABLE_P (arg)
 154       && EQ (XCHAR_TABLE (arg)->purpose, Qcategory_table))
 155     return Qt;
 156   return Qnil;
 157 }
 158
 159 /* If TABLE is nil, return the current category table.  If TABLE is
 160    not nil, check the validity of TABLE as a category table.  If
 161    valid, return TABLE itself, but if not valid, signal an error of
 162    wrong-type-argument.  */
 163
 164 Lisp_Object
 165 check_category_table (table)
 166      Lisp_Object table;
 167 {
 168   register Lisp_Object tem;
 169   if (NILP (table))
 170     return current_buffer->category_table;
 171   while (tem = Fcategory_table_p (table), NILP (tem))
 172     table = wrong_type_argument (Qcategory_table_p, table);
 173   return table;
 174 }
 175
 176 DEFUN ("category-table", Fcategory_table, Scategory_table, 0, 0, 0,
 177        doc: /* Return the current category table.
 178 This is the one specified by the current buffer.  */)
 179      ()
 180 {
 181   return current_buffer->category_table;
 182 }
 183
 184 DEFUN ("standard-category-table", Fstandard_category_table,
 185    Sstandard_category_table, 0, 0, 0,
 186        doc: /* Return the standard category table.
 187 This is the one used for new buffers.  */)
 188      ()
 189 {
 190   return Vstandard_category_table;
 191 }
 192
 193
 194 static void
 195 copy_category_entry (table, c, val)
 196      Lisp_Object table, c, val;
 197 {
 198   val = Fcopy_sequence (val);
 199   if (CONSP (c))
 200     char_table_set_range (table, XINT (XCAR (c)), XINT (XCDR (c)), val);
 201   else
 202     char_table_set (table, XINT (c), val);
 203 }
 204
 205 /* Return a copy of category table TABLE.  We can't simply use the
 206    function copy-sequence because no contents should be shared between
 207    the original and the copy.  This function is called recursively by
 208    binding TABLE to a sub char table.  */
 209
 210 Lisp_Object
 211 copy_category_table (table)
 212      Lisp_Object table;
 213 {
 214   table = copy_char_table (table);
 215
 216   if (! NILP (XCHAR_TABLE (table)->defalt))
 217     XCHAR_TABLE (table)->defalt
 218       = Fcopy_sequence (XCHAR_TABLE (table)->defalt);
 219   XCHAR_TABLE (table)->extras[0]
 220     = Fcopy_sequence (XCHAR_TABLE (table)->extras[0]);
 221   map_char_table (copy_category_entry, Qnil, table, table);
 222
 223   return table;
 224 }
 225
 226 DEFUN ("copy-category-table", Fcopy_category_table, Scopy_category_table,
 227        0, 1, 0,
 228        doc: /* Construct a new category table and return it.
 229 It is a copy of the TABLE, which defaults to the standard category table.  */)
 230      (table)
 231      Lisp_Object table;
 232 {
 233   if (!NILP (table))
 234     check_category_table (table);
 235   else
 236     table = Vstandard_category_table;
 237
 238   return copy_category_table (table);
 239 }
 240
 241 DEFUN ("make-category-table", Fmake_category_table, Smake_category_table,
 242        0, 0, 0,
 243        doc: /* Construct a new and empty category table and return it.  */)
 244      ()
 245 {
 246   Lisp_Object val;
 247   int i;
 248
 249   val = Fmake_char_table (Qcategory_table, Qnil);
 250   XCHAR_TABLE (val)->defalt = MAKE_CATEGORY_SET;
 251   for (i = 0; i < (1 << CHARTAB_SIZE_BITS_0); i++)
 252     XCHAR_TABLE (val)->contents[i] = MAKE_CATEGORY_SET;
 253   Fset_char_table_extra_slot (val, make_number (0),
 254                               Fmake_vector (make_number (95), Qnil));
 255   return val;
 256 }
 257
 258 DEFUN ("set-category-table", Fset_category_table, Sset_category_table, 1, 1, 0,
 259        doc: /* Specify TABLE as the category table for the current buffer.  */)
 260      (table)
 261      Lisp_Object table;
 262 {
 263   int idx;
 264   table = check_category_table (table);
 265   current_buffer->category_table = table;
 266   /* Indicate that this buffer now has a specified category table.  */
 267   idx = PER_BUFFER_VAR_IDX (category_table);
 268   SET_PER_BUFFER_VALUE_P (current_buffer, idx, 1);
 269   return table;
 270 }
 271
 272 \f
 273 Lisp_Object
 274 char_category_set (c)
 275      int c;
 276 {
 277   return CHAR_TABLE_REF (current_buffer->category_table, c);
 278 }
 279
 280 DEFUN ("char-category-set", Fchar_category_set, Schar_category_set, 1, 1, 0,
 281        doc: /* Return the category set of CHAR.  */)
 282      (ch)
 283      Lisp_Object ch;
 284 {
 285   CHECK_NUMBER (ch);
 286   return CATEGORY_SET (XFASTINT (ch));
 287 }
 288
 289 DEFUN ("category-set-mnemonics", Fcategory_set_mnemonics,
 290        Scategory_set_mnemonics, 1, 1, 0,
 291        doc: /* Return a string containing mnemonics of the categories in CATEGORY-SET.
 292 CATEGORY-SET is a bool-vector, and the categories \"in\" it are those
 293 that are indexes where t occurs the bool-vector.
 294 The return value is a string containing those same categories.  */)
 295      (category_set)
 296      Lisp_Object category_set;
 297 {
 298   int i, j;
 299   char str[96];
 300
 301   CHECK_CATEGORY_SET (category_set);
 302
 303   j = 0;
 304   for (i = 32; i < 127; i++)
 305     if (CATEGORY_MEMBER (i, category_set))
 306       str[j++] = i;
 307   str[j] = '\0';
 308
 309   return build_string (str);
 310 }
 311
 312 void
 313 set_category_set (category_set, category, val)
 314      Lisp_Object category_set, category, val;
 315 {
 316   do {
 317     int idx = XINT (category) / 8;
 318     unsigned char bits = 1 << (XINT (category) % 8);
 319
 320     if (NILP (val))
 321       XCATEGORY_SET (category_set)->data[idx] &= ~bits;
 322     else
 323       XCATEGORY_SET (category_set)->data[idx] |= bits;
 324   } while (0);
 325 }
 326
 327 DEFUN ("modify-category-entry", Fmodify_category_entry,
 328        Smodify_category_entry, 2, 4, 0,
 329        doc: /* Modify the category set of CHARACTER by adding CATEGORY to it.
 330 The category is changed only for table TABLE, which defaults to
 331 the current buffer's category table.
 332 CHARACTER can be either a single character or a cons representing the
 333 lower and upper ends of an inclusive character range to modify.
 334 If optional fourth argument RESET is non-nil,
 335 then delete CATEGORY from the category set instead of adding it.  */)
 336      (character, category, table, reset)
 337      Lisp_Object character, category, table, reset;
 338 {
 339   Lisp_Object set_value;        /* Actual value to be set in category sets.  */
 340   Lisp_Object category_set;
 341   int start, end;
 342   int from, to;
 343
 344   if (INTEGERP (character))
 345     {
 346       CHECK_CHARACTER (character);
 347       start = end = XFASTINT (character);
 348     }
 349   else
 350     {
 351       CHECK_CONS (character);
 352       CHECK_CHARACTER_CAR (character);
 353       CHECK_CHARACTER_CDR (character);
 354       start = XFASTINT (XCAR (character));
 355       end = XFASTINT (XCDR (character));
 356     }
 357
 358   CHECK_CATEGORY (category);
 359   table = check_category_table (table);
 360
 361   if (NILP (CATEGORY_DOCSTRING (table, XFASTINT (category))))
 362     error ("Undefined category: %c", XFASTINT (category));
 363
 364   set_value = NILP (reset) ? Qt : Qnil;
 365
 366   while (start <= end)
 367     {
 368       category_set = char_table_ref_and_range (table, start, &from, &to);
 369       if (CATEGORY_MEMBER (XFASTINT (category), category_set) != NILP (reset))
 370         {
 371           category_set = Fcopy_sequence (category_set);
 372           SET_CATEGORY_SET (category_set, category, set_value);
 373           if (to > end)
 374             char_table_set_range (table, start, end, category_set);
 375           else
 376             char_table_set_range (table, start, to, category_set);
 377         }
 378       start = to + 1;
 379     }
 380
 381   return Qnil;
 382 }
 383 \f
 384 /* Return 1 if there is a word boundary between two word-constituent
 385    characters C1 and C2 if they appear in this order, else return 0.
 386    Use the macro WORD_BOUNDARY_P instead of calling this function
 387    directly.  */
 388
 389 int
 390 word_boundary_p (c1, c2)
 391      int c1, c2;
 392 {
 393   Lisp_Object category_set1, category_set2;
 394   Lisp_Object tail;
 395   int default_result;
 396
 397   if (CHAR_CHARSET (c1) == CHAR_CHARSET (c2))
 398     {
 399       tail = Vword_separating_categories;
 400       default_result = 0;
 401     }
 402   else
 403     {
 404       tail = Vword_combining_categories;
 405       default_result = 1;
 406     }
 407
 408   category_set1 = CATEGORY_SET (c1);
 409   if (NILP (category_set1))
 410     return default_result;
 411   category_set2 = CATEGORY_SET (c2);
 412   if (NILP (category_set2))
 413     return default_result;
 414
 415   for (; CONSP (tail); tail = XCDR (tail))
 416     {
 417       Lisp_Object elt = XCAR (tail);
 418
 419       if (CONSP (elt)
 420           && CATEGORYP (XCAR (elt))
 421           && CATEGORYP (XCDR (elt))
 422           && CATEGORY_MEMBER (XFASTINT (XCAR (elt)), category_set1)
 423           && CATEGORY_MEMBER (XFASTINT (XCDR (elt)), category_set2))
 424         return !default_result;
 425     }
 426   return default_result;
 427 }
 428
 429 \f
 430 void
 431 init_category_once ()
 432 {
 433   /* This has to be done here, before we call Fmake_char_table.  */
 434   Qcategory_table = intern ("category-table");
 435   staticpro (&Qcategory_table);
 436
 437   /* Intern this now in case it isn't already done.
 438      Setting this variable twice is harmless.
 439      But don't staticpro it here--that is done in alloc.c.  */
 440   Qchar_table_extra_slots = intern ("char-table-extra-slots");
 441
 442   /* Now we are ready to set up this property, so we can
 443      create category tables.  */
 444   Fput (Qcategory_table, Qchar_table_extra_slots, make_number (2));
 445
 446   Vstandard_category_table = Fmake_char_table (Qcategory_table, Qnil);
 447   /* Set a category set which contains nothing to the default.  */
 448   XCHAR_TABLE (Vstandard_category_table)->defalt = MAKE_CATEGORY_SET;
 449   Fset_char_table_extra_slot (Vstandard_category_table, make_number (0),
 450                               Fmake_vector (make_number (95), Qnil));
 451 }
 452
 453 void
 454 syms_of_category ()
 455 {
 456   Qcategoryp = intern ("categoryp");
 457   staticpro (&Qcategoryp);
 458   Qcategorysetp = intern ("categorysetp");
 459   staticpro (&Qcategorysetp);
 460   Qcategory_table_p = intern ("category-table-p");
 461   staticpro (&Qcategory_table_p);
 462
 463   DEFVAR_LISP ("word-combining-categories", &Vword_combining_categories,
 464                doc: /* List of pair (cons) of categories to determine word boundary.
 465
 466 Emacs treats a sequence of word constituent characters as a single
 467 word (i.e. finds no word boundary between them) iff they belongs to
 468 the same charset.  But, exceptions are allowed in the following cases.
 469
 470 \(1) The case that characters are in different charsets is controlled
 471 by the variable `word-combining-categories'.
 472
 473 Emacs finds no word boundary between characters of different charsets
 474 if they have categories matching some element of this list.
 475
 476 More precisely, if an element of this list is a cons of category CAT1
 477 and CAT2, and a multibyte character C1 which has CAT1 is followed by
 478 C2 which has CAT2, there's no word boundary between C1 and C2.
 479
 480 For instance, to tell that ASCII characters and Latin-1 characters can
 481 form a single word, the element `(?l . ?l)' should be in this list
 482 because both characters have the category `l' (Latin characters).
 483
 484 \(2) The case that character are in the same charset is controlled by
 485 the variable `word-separating-categories'.
 486
 487 Emacs find a word boundary between characters of the same charset
 488 if they have categories matching some element of this list.
 489
 490 More precisely, if an element of this list is a cons of category CAT1
 491 and CAT2, and a multibyte character C1 which has CAT1 is followed by
 492 C2 which has CAT2, there's a word boundary between C1 and C2.
 493
 494 For instance, to tell that there's a word boundary between Japanese
 495 Hiragana and Japanese Kanji (both are in the same charset), the
 496 element `(?H . ?C) should be in this list.  */);
 497
 498   Vword_combining_categories = Qnil;
 499
 500   DEFVAR_LISP ("word-separating-categories", &Vword_separating_categories,
 501                doc: /* List of pair (cons) of categories to determine word boundary.
 502 See the documentation of the variable `word-combining-categories'.  */);
 503
 504   Vword_separating_categories = Qnil;
 505
 506   defsubr (&Smake_category_set);
 507   defsubr (&Sdefine_category);
 508   defsubr (&Scategory_docstring);
 509   defsubr (&Sget_unused_category);
 510   defsubr (&Scategory_table_p);
 511   defsubr (&Scategory_table);
 512   defsubr (&Sstandard_category_table);
 513   defsubr (&Scopy_category_table);
 514   defsubr (&Smake_category_table);
 515   defsubr (&Sset_category_table);
 516   defsubr (&Schar_category_set);
 517   defsubr (&Scategory_set_mnemonics);
 518   defsubr (&Smodify_category_entry);
 519
 520   category_table_version = 0;
 521 }