src/category.c

   1 /* GNU Emacs routines to deal with category tables.
   2    Copyright (C) 1998, 2001, 2002, 2003, 2004, 2005, 2006, 2007, 2008
   3      Free Software Foundation, Inc.
   4    Copyright (C) 1995, 1996, 1997, 1998, 1999, 2000, 2001, 2002, 2003, 2004,
   5      2005, 2006, 2007, 2008
   6      National Institute of Advanced Industrial Science and Technology (AIST)
   7      Registration Number H14PRO021
   8    Copyright (C) 2003
   9      National Institute of Advanced Industrial Science and Technology (AIST)
  10      Registration Number H13PRO009
  11
  12 This file is part of GNU Emacs.
  13
  14 GNU Emacs is free software: you can redistribute it and/or modify
  15 it under the terms of the GNU General Public License as published by
  16 the Free Software Foundation, either version 3 of the License, or
  17 (at your option) any later version.
  18
  19 GNU Emacs is distributed in the hope that it will be useful,
  20 but WITHOUT ANY WARRANTY; without even the implied warranty of
  21 MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  22 GNU General Public License for more details.
  23
  24 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/>.  */
  26
  27
  28 /* Here we handle three objects: category, category set, and category
  29    table.  Read comments in the file category.h to understand them.  */
  30
  31 #include <config.h>
  32 #include <ctype.h>
  33 #include "lisp.h"
  34 #include "buffer.h"
  35 #include "character.h"
  36 #include "charset.h"
  37 #include "category.h"
  38 #include "keymap.h"
  39
  40 /* The version number of the latest category table.  Each category
  41    table has a unique version number.  It is assigned a new number
  42    also when it is modified.  When a regular expression is compiled
  43    into the struct re_pattern_buffer, the version number of the
  44    category table (of the current buffer) at that moment is also
  45    embedded in the structure.
  46
  47    For the moment, we are not using this feature.  */
  48 static int category_table_version;
  49
  50 Lisp_Object Qcategory_table, Qcategoryp, Qcategorysetp, Qcategory_table_p;
  51
  52 /* Variables to determine word boundary.  */
  53 Lisp_Object Vword_combining_categories, Vword_separating_categories;
  54
  55 /* Temporary internal variable used in macro CHAR_HAS_CATEGORY.  */
  56 Lisp_Object _temp_category_set;
  57
  58 \f
  59 /* Category set staff.  */
  60
  61 DEFUN ("make-category-set", Fmake_category_set, Smake_category_set, 1, 1, 0,
  62        doc: /* Return a newly created category-set which contains CATEGORIES.
  63 CATEGORIES is a string of category mnemonics.
  64 The value is a bool-vector which has t at the indices corresponding to
  65 those categories.  */)
  66      (categories)
  67      Lisp_Object categories;
  68 {
  69   Lisp_Object val;
  70   int len;
  71
  72   CHECK_STRING (categories);
  73   val = MAKE_CATEGORY_SET;
  74
  75   if (STRING_MULTIBYTE (categories))
  76     error ("Multibyte string in `make-category-set'");
  77
  78   len = SCHARS (categories);
  79   while (--len >= 0)
  80     {
  81       Lisp_Object category;
  82
  83       XSETFASTINT (category, SREF (categories, len));
  84       CHECK_CATEGORY (category);
  85       SET_CATEGORY_SET (val, category, Qt);
  86     }
  87   return val;
  88 }
  89
  90 \f
  91 /* Category staff.  */
  92
  93 Lisp_Object check_category_table ();
  94
  95 DEFUN ("define-category", Fdefine_category, Sdefine_category, 2, 3, 0,
  96        doc: /* Define CATEGORY as a category which is described by DOCSTRING.
  97 CATEGORY should be an ASCII printing character in the range ` ' to `~'.
  98 DOCSTRING is the documentation string of the category.
  99 The category is defined only in category table TABLE, which defaults to
 100 the current buffer's category table.  */)
 101      (category, docstring, table)
 102      Lisp_Object category, docstring, table;
 103 {
 104   CHECK_CATEGORY (category);
 105   CHECK_STRING (docstring);
 106   table = check_category_table (table);
 107
 108   if (!NILP (CATEGORY_DOCSTRING (table, XFASTINT (category))))
 109     error ("Category `%c' is already defined", XFASTINT (category));
 110   CATEGORY_DOCSTRING (table, XFASTINT (category)) = docstring;
 111
 112   return Qnil;
 113 }
 114
 115 DEFUN ("category-docstring", Fcategory_docstring, Scategory_docstring, 1, 2, 0,
 116        doc: /* Return the documentation string of CATEGORY, as defined in TABLE.
 117 TABLE should be a category table and defaults to the current buffer's
 118 category table.  */)
 119      (category, table)
 120      Lisp_Object category, table;
 121 {
 122   CHECK_CATEGORY (category);
 123   table = check_category_table (table);
 124
 125   return CATEGORY_DOCSTRING (table, XFASTINT (category));
 126 }
 127
 128 DEFUN ("get-unused-category", Fget_unused_category, Sget_unused_category,
 129        0, 1, 0,
 130        doc: /* Return a category which is not yet defined in TABLE.
 131 If no category remains available, return nil.
 132 The optional argument TABLE specifies which category table to modify;
 133 it defaults to the current buffer's category table.  */)
 134      (table)
 135      Lisp_Object table;
 136 {
 137   int i;
 138
 139   table = check_category_table (table);
 140
 141   for (i = ' '; i <= '~'; i++)
 142     if (NILP (CATEGORY_DOCSTRING (table, i)))
 143       return make_number (i);
 144
 145   return Qnil;
 146 }
 147
 148 \f
 149 /* Category-table staff.  */
 150
 151 DEFUN ("category-table-p", Fcategory_table_p, Scategory_table_p, 1, 1, 0,
 152        doc: /* Return t if ARG is a category table.  */)
 153      (arg)
 154      Lisp_Object arg;
 155 {
 156   if (CHAR_TABLE_P (arg)
 157       && EQ (XCHAR_TABLE (arg)->purpose, Qcategory_table))
 158     return Qt;
 159   return Qnil;
 160 }
 161
 162 /* If TABLE is nil, return the current category table.  If TABLE is
 163    not nil, check the validity of TABLE as a category table.  If
 164    valid, return TABLE itself, but if not valid, signal an error of
 165    wrong-type-argument.  */
 166
 167 Lisp_Object
 168 check_category_table (table)
 169      Lisp_Object table;
 170 {
 171   if (NILP (table))
 172     return current_buffer->category_table;
 173   CHECK_TYPE (!NILP (Fcategory_table_p (table)), Qcategory_table_p, table);
 174   return table;
 175 }
 176
 177 DEFUN ("category-table", Fcategory_table, Scategory_table, 0, 0, 0,
 178        doc: /* Return the current category table.
 179 This is the one specified by the current buffer.  */)
 180      ()
 181 {
 182   return current_buffer->category_table;
 183 }
 184
 185 DEFUN ("standard-category-table", Fstandard_category_table,
 186    Sstandard_category_table, 0, 0, 0,
 187        doc: /* Return the standard category table.
 188 This is the one used for new buffers.  */)
 189      ()
 190 {
 191   return Vstandard_category_table;
 192 }
 193
 194
 195 static void
 196 copy_category_entry (table, c, val)
 197      Lisp_Object table, c, val;
 198 {
 199   val = Fcopy_sequence (val);
 200   if (CONSP (c))
 201     char_table_set_range (table, XINT (XCAR (c)), XINT (XCDR (c)), val);
 202   else
 203     char_table_set (table, XINT (c), val);
 204 }
 205
 206 /* Return a copy of category table TABLE.  We can't simply use the
 207    function copy-sequence because no contents should be shared between
 208    the original and the copy.  This function is called recursively by
 209    binding TABLE to a sub char table.  */
 210
 211 Lisp_Object
 212 copy_category_table (table)
 213      Lisp_Object table;
 214 {
 215   table = copy_char_table (table);
 216
 217   if (! NILP (XCHAR_TABLE (table)->defalt))
 218     XCHAR_TABLE (table)->defalt
 219       = Fcopy_sequence (XCHAR_TABLE (table)->defalt);
 220   XCHAR_TABLE (table)->extras[0]
 221     = Fcopy_sequence (XCHAR_TABLE (table)->extras[0]);
 222   map_char_table (copy_category_entry, Qnil, table, table);
 223
 224   return table;
 225 }
 226
 227 DEFUN ("copy-category-table", Fcopy_category_table, Scopy_category_table,
 228        0, 1, 0,
 229        doc: /* Construct a new category table and return it.
 230 It is a copy of the TABLE, which defaults to the standard category table.  */)
 231      (table)
 232      Lisp_Object table;
 233 {
 234   if (!NILP (table))
 235     check_category_table (table);
 236   else
 237     table = Vstandard_category_table;
 238
 239   return copy_category_table (table);
 240 }
 241
 242 DEFUN ("make-category-table", Fmake_category_table, Smake_category_table,
 243        0, 0, 0,
 244        doc: /* Construct a new and empty category table and return it.  */)
 245      ()
 246 {
 247   Lisp_Object val;
 248   int i;
 249
 250   val = Fmake_char_table (Qcategory_table, Qnil);
 251   XCHAR_TABLE (val)->defalt = MAKE_CATEGORY_SET;
 252   for (i = 0; i < (1 << CHARTAB_SIZE_BITS_0); i++)
 253     XCHAR_TABLE (val)->contents[i] = MAKE_CATEGORY_SET;
 254   Fset_char_table_extra_slot (val, make_number (0),
 255                               Fmake_vector (make_number (95), Qnil));
 256   return val;
 257 }
 258
 259 DEFUN ("set-category-table", Fset_category_table, Sset_category_table, 1, 1, 0,
 260        doc: /* Specify TABLE as the category table for the current buffer.
 261 Return TABLE.  */)
 262      (table)
 263      Lisp_Object table;
 264 {
 265   int idx;
 266   table = check_category_table (table);
 267   current_buffer->category_table = table;
 268   /* Indicate that this buffer now has a specified category table.  */
 269   idx = PER_BUFFER_VAR_IDX (category_table);
 270   SET_PER_BUFFER_VALUE_P (current_buffer, idx, 1);
 271   return table;
 272 }
 273
 274 \f
 275 Lisp_Object
 276 char_category_set (c)
 277      int c;
 278 {
 279   return CHAR_TABLE_REF (current_buffer->category_table, c);
 280 }
 281
 282 DEFUN ("char-category-set", Fchar_category_set, Schar_category_set, 1, 1, 0,
 283        doc: /* Return the category set of CHAR.
 284 usage: (char-category-set CHAR)  */)
 285      (ch)
 286      Lisp_Object ch;
 287 {
 288   CHECK_NUMBER (ch);
 289   return CATEGORY_SET (XFASTINT (ch));
 290 }
 291
 292 DEFUN ("category-set-mnemonics", Fcategory_set_mnemonics,
 293        Scategory_set_mnemonics, 1, 1, 0,
 294        doc: /* Return a string containing mnemonics of the categories in CATEGORY-SET.
 295 CATEGORY-SET is a bool-vector, and the categories \"in\" it are those
 296 that are indexes where t occurs in the bool-vector.
 297 The return value is a string containing those same categories.  */)
 298      (category_set)
 299      Lisp_Object category_set;
 300 {
 301   int i, j;
 302   char str[96];
 303
 304   CHECK_CATEGORY_SET (category_set);
 305
 306   j = 0;
 307   for (i = 32; i < 127; i++)
 308     if (CATEGORY_MEMBER (i, category_set))
 309       str[j++] = i;
 310   str[j] = '\0';
 311
 312   return build_string (str);
 313 }
 314
 315 void
 316 set_category_set (category_set, category, val)
 317      Lisp_Object category_set, category, val;
 318 {
 319   do {
 320     int idx = XINT (category) / 8;
 321     unsigned char bits = 1 << (XINT (category) % 8);
 322
 323     if (NILP (val))
 324       XCATEGORY_SET (category_set)->data[idx] &= ~bits;
 325     else
 326       XCATEGORY_SET (category_set)->data[idx] |= bits;
 327   } while (0);
 328 }
 329
 330 DEFUN ("modify-category-entry", Fmodify_category_entry,
 331        Smodify_category_entry, 2, 4, 0,
 332        doc: /* Modify the category set of CHARACTER by adding CATEGORY to it.
 333 The category is changed only for table TABLE, which defaults to
 334 the current buffer's category table.
 335 CHARACTER can be either a single character or a cons representing the
 336 lower and upper ends of an inclusive character range to modify.
 337 If optional fourth argument RESET is non-nil,
 338 then delete CATEGORY from the category set instead of adding it.  */)
 339      (character, category, table, reset)
 340      Lisp_Object character, category, table, reset;
 341 {
 342   Lisp_Object set_value;        /* Actual value to be set in category sets.  */
 343   Lisp_Object category_set;
 344   int start, end;
 345   int from, to;
 346
 347   if (INTEGERP (character))
 348     {
 349       CHECK_CHARACTER (character);
 350       start = end = XFASTINT (character);
 351     }
 352   else
 353     {
 354       CHECK_CONS (character);
 355       CHECK_CHARACTER_CAR (character);
 356       CHECK_CHARACTER_CDR (character);
 357       start = XFASTINT (XCAR (character));
 358       end = XFASTINT (XCDR (character));
 359     }
 360
 361   CHECK_CATEGORY (category);
 362   table = check_category_table (table);
 363
 364   if (NILP (CATEGORY_DOCSTRING (table, XFASTINT (category))))
 365     error ("Undefined category: %c", XFASTINT (category));
 366
 367   set_value = NILP (reset) ? Qt : Qnil;
 368
 369   while (start <= end)
 370     {
 371       category_set = char_table_ref_and_range (table, start, &from, &to);
 372       if (CATEGORY_MEMBER (XFASTINT (category), category_set) != NILP (reset))
 373         {
 374           category_set = Fcopy_sequence (category_set);
 375           SET_CATEGORY_SET (category_set, category, set_value);
 376           if (to > end)
 377             char_table_set_range (table, start, end, category_set);
 378           else
 379             char_table_set_range (table, start, to, category_set);
 380         }
 381       start = to + 1;
 382     }
 383
 384   return Qnil;
 385 }
 386 \f
 387 /* Return 1 if there is a word boundary between two word-constituent
 388    characters C1 and C2 if they appear in this order, else return 0.
 389    Use the macro WORD_BOUNDARY_P instead of calling this function
 390    directly.  */
 391
 392 int
 393 word_boundary_p (c1, c2)
 394      int c1, c2;
 395 {
 396   Lisp_Object category_set1, category_set2;
 397   Lisp_Object tail;
 398   int default_result;
 399
 400   if (EQ (CHAR_TABLE_REF (Vchar_script_table, c1),
 401           CHAR_TABLE_REF (Vchar_script_table, c2)))
 402     {
 403       tail = Vword_separating_categories;
 404       default_result = 0;
 405     }
 406   else
 407     {
 408       tail = Vword_combining_categories;
 409       default_result = 1;
 410     }
 411
 412   category_set1 = CATEGORY_SET (c1);
 413   if (NILP (category_set1))
 414     return default_result;
 415   category_set2 = CATEGORY_SET (c2);
 416   if (NILP (category_set2))
 417     return default_result;
 418
 419   for (; CONSP (tail); tail = XCDR (tail))
 420     {
 421       Lisp_Object elt = XCAR (tail);
 422
 423       if (CONSP (elt)
 424           && (NILP (XCAR (elt))
 425               || (CATEGORYP (XCAR (elt))
 426                   && CATEGORY_MEMBER (XFASTINT (XCAR (elt)), category_set1)))
 427           && (NILP (XCDR (elt))
 428               || (CATEGORYP (XCDR (elt))
 429                   && CATEGORY_MEMBER (XFASTINT (XCDR (elt)), category_set2))))
 430         return !default_result;
 431     }
 432   return default_result;
 433 }
 434
 435 \f
 436 void
 437 init_category_once ()
 438 {
 439   /* This has to be done here, before we call Fmake_char_table.  */
 440   Qcategory_table = intern ("category-table");
 441   staticpro (&Qcategory_table);
 442
 443   /* Intern this now in case it isn't already done.
 444      Setting this variable twice is harmless.
 445      But don't staticpro it here--that is done in alloc.c.  */
 446   Qchar_table_extra_slots = intern ("char-table-extra-slots");
 447
 448   /* Now we are ready to set up this property, so we can
 449      create category tables.  */
 450   Fput (Qcategory_table, Qchar_table_extra_slots, make_number (2));
 451
 452   Vstandard_category_table = Fmake_char_table (Qcategory_table, Qnil);
 453   /* Set a category set which contains nothing to the default.  */
 454   XCHAR_TABLE (Vstandard_category_table)->defalt = MAKE_CATEGORY_SET;
 455   Fset_char_table_extra_slot (Vstandard_category_table, make_number (0),
 456                               Fmake_vector (make_number (95), Qnil));
 457 }
 458
 459 void
 460 syms_of_category ()
 461 {
 462   Qcategoryp = intern ("categoryp");
 463   staticpro (&Qcategoryp);
 464   Qcategorysetp = intern ("categorysetp");
 465   staticpro (&Qcategorysetp);
 466   Qcategory_table_p = intern ("category-table-p");
 467   staticpro (&Qcategory_table_p);
 468
 469   DEFVAR_LISP ("word-combining-categories", &Vword_combining_categories,
 470                doc: /* List of pair (cons) of categories to determine word boundary.
 471
 472 Emacs treats a sequence of word constituent characters as a single
 473 word (i.e. finds no word boundary between them) only if they belong to
 474 the same script.  But, exceptions are allowed in the following cases.
 475
 476 \(1) The case that characters are in different scripts is controlled
 477 by the variable `word-combining-categories'.
 478
 479 Emacs finds no word boundary between characters of different scripts
 480 if they have categories matching some element of this list.
 481
 482 More precisely, if an element of this list is a cons of category CAT1
 483 and CAT2, and a multibyte character C1 which has CAT1 is followed by
 484 C2 which has CAT2, there's no word boundary between C1 and C2.
 485
 486 For instance, to tell that Han characters followed by Hiragana
 487 characters can form a single word, the element `(?C . ?H)' should be
 488 in this list.
 489
 490 \(2) The case that character are in the same script is controlled by
 491 the variable `word-separating-categories'.
 492
 493 Emacs find a word boundary between characters of the same script
 494 if they have categories matching some element of this list.
 495
 496 More precisely, if an element of this list is a cons of category CAT1
 497 and CAT2, and a multibyte character C1 which has CAT1 is followed by
 498 C2 which has CAT2, there's a word boundary between C1 and C2.
 499
 500 For instance, to tell that there's a word boundary between Hiragana
 501 and Katakana (both are in the same script `kana'),
 502 the element `(?H . ?K) should be in this list.  */);
 503
 504   Vword_combining_categories = Qnil;
 505
 506   DEFVAR_LISP ("word-separating-categories", &Vword_separating_categories,
 507                doc: /* List of pair (cons) of categories to determine word boundary.
 508 See the documentation of the variable `word-combining-categories'.  */);
 509
 510   Vword_separating_categories = Qnil;
 511
 512   defsubr (&Smake_category_set);
 513   defsubr (&Sdefine_category);
 514   defsubr (&Scategory_docstring);
 515   defsubr (&Sget_unused_category);
 516   defsubr (&Scategory_table_p);
 517   defsubr (&Scategory_table);
 518   defsubr (&Sstandard_category_table);
 519   defsubr (&Scopy_category_table);
 520   defsubr (&Smake_category_table);
 521   defsubr (&Sset_category_table);
 522   defsubr (&Schar_category_set);
 523   defsubr (&Scategory_set_mnemonics);
 524   defsubr (&Smodify_category_entry);
 525
 526   category_table_version = 0;
 527 }
 528
 529 /* arch-tag: 74ebf524-121b-4d9c-bd68-07f8d708b211
 530    (do not change this comment) */