src/category.c

   1 /* GNU Emacs routines to deal with category tables.
   2
   3 Copyright (C) 1998, 2001-2016 Free Software Foundation, Inc.
   4 Copyright (C) 1995, 1996, 1997, 1998, 1999, 2000, 2001, 2002, 2003, 2004,
   5   2005, 2006, 2007, 2008, 2009, 2010, 2011
   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 (at
  17 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
  33 #include "lisp.h"
  34 #include "character.h"
  35 #include "buffer.h"
  36 #include "category.h"
  37
  38 /* This setter is used only in this file, so it can be private.  */
  39 static void
  40 bset_category_table (struct buffer *b, Lisp_Object val)
  41 {
  42   b->category_table_ = val;
  43 }
  44
  45 /* The version number of the latest category table.  Each category
  46    table has a unique version number.  It is assigned a new number
  47    also when it is modified.  When a regular expression is compiled
  48    into the struct re_pattern_buffer, the version number of the
  49    category table (of the current buffer) at that moment is also
  50    embedded in the structure.
  51
  52    For the moment, we are not using this feature.  */
  53 static int category_table_version;
  54 \f
  55 /* Category set staff.  */
  56
  57 static Lisp_Object
  58 hash_get_category_set (Lisp_Object table, Lisp_Object category_set)
  59 {
  60   struct Lisp_Hash_Table *h;
  61   ptrdiff_t i;
  62   EMACS_UINT hash;
  63
  64   if (NILP (XCHAR_TABLE (table)->extras[1]))
  65     set_char_table_extras
  66       (table, 1,
  67        make_hash_table (hashtest_equal, make_number (DEFAULT_HASH_SIZE),
  68                         make_float (DEFAULT_REHASH_SIZE),
  69                         make_float (DEFAULT_REHASH_THRESHOLD),
  70                         Qnil));
  71   h = XHASH_TABLE (XCHAR_TABLE (table)->extras[1]);
  72   i = hash_lookup (h, category_set, &hash);
  73   if (i >= 0)
  74     return HASH_KEY (h, i);
  75   hash_put (h, category_set, Qnil, hash);
  76   return category_set;
  77 }
  78
  79 /* Make CATEGORY_SET include (if VAL) or exclude (if !VAL) CATEGORY.  */
  80
  81 static void
  82 set_category_set (Lisp_Object category_set, EMACS_INT category, bool val)
  83 {
  84   bool_vector_set (category_set, category, val);
  85 }
  86
  87 DEFUN ("make-category-set", Fmake_category_set, Smake_category_set, 1, 1, 0,
  88        doc: /* Return a newly created category-set which contains CATEGORIES.
  89 CATEGORIES is a string of category mnemonics.
  90 The value is a bool-vector which has t at the indices corresponding to
  91 those categories.  */)
  92   (Lisp_Object categories)
  93 {
  94   Lisp_Object val;
  95   ptrdiff_t len;
  96
  97   CHECK_STRING (categories);
  98   val = MAKE_CATEGORY_SET;
  99
 100   if (STRING_MULTIBYTE (categories))
 101     error ("Multibyte string in `make-category-set'");
 102
 103   len = SCHARS (categories);
 104   while (--len >= 0)
 105     {
 106       unsigned char cat = SREF (categories, len);
 107       Lisp_Object category = make_number (cat);
 108
 109       CHECK_CATEGORY (category);
 110       set_category_set (val, cat, 1);
 111     }
 112   return val;
 113 }
 114
 115 \f
 116 /* Category staff.  */
 117
 118 static Lisp_Object check_category_table (Lisp_Object table);
 119
 120 DEFUN ("define-category", Fdefine_category, Sdefine_category, 2, 3, 0,
 121        doc: /* Define CATEGORY as a category which is described by DOCSTRING.
 122 CATEGORY should be an ASCII printing character in the range ` ' to `~'.
 123 DOCSTRING is the documentation string of the category.  The first line
 124 should be a terse text (preferably less than 16 characters),
 125 and the rest lines should be the full description.
 126 The category is defined only in category table TABLE, which defaults to
 127 the current buffer's category table.  */)
 128   (Lisp_Object category, Lisp_Object docstring, Lisp_Object table)
 129 {
 130   CHECK_CATEGORY (category);
 131   CHECK_STRING (docstring);
 132   table = check_category_table (table);
 133
 134   if (!NILP (CATEGORY_DOCSTRING (table, XFASTINT (category))))
 135     error ("Category `%c' is already defined", (int) XFASTINT (category));
 136   if (!NILP (Vpurify_flag))
 137     docstring = Fpurecopy (docstring);
 138   SET_CATEGORY_DOCSTRING (table, XFASTINT (category), docstring);
 139
 140   return Qnil;
 141 }
 142
 143 DEFUN ("category-docstring", Fcategory_docstring, Scategory_docstring, 1, 2, 0,
 144        doc: /* Return the documentation string of CATEGORY, as defined in TABLE.
 145 TABLE should be a category table and defaults to the current buffer's
 146 category table.  */)
 147   (Lisp_Object category, Lisp_Object table)
 148 {
 149   CHECK_CATEGORY (category);
 150   table = check_category_table (table);
 151
 152   return CATEGORY_DOCSTRING (table, XFASTINT (category));
 153 }
 154
 155 DEFUN ("get-unused-category", Fget_unused_category, Sget_unused_category,
 156        0, 1, 0,
 157        doc: /* Return a category which is not yet defined in TABLE.
 158 If no category remains available, return nil.
 159 The optional argument TABLE specifies which category table to modify;
 160 it defaults to the current buffer's category table.  */)
 161   (Lisp_Object table)
 162 {
 163   int i;
 164
 165   table = check_category_table (table);
 166
 167   for (i = ' '; i <= '~'; i++)
 168     if (NILP (CATEGORY_DOCSTRING (table, i)))
 169       return make_number (i);
 170
 171   return Qnil;
 172 }
 173
 174 \f
 175 /* Category-table staff.  */
 176
 177 DEFUN ("category-table-p", Fcategory_table_p, Scategory_table_p, 1, 1, 0,
 178        doc: /* Return t if ARG is a category table.  */)
 179   (Lisp_Object arg)
 180 {
 181   if (CHAR_TABLE_P (arg)
 182       && EQ (XCHAR_TABLE (arg)->purpose, Qcategory_table))
 183     return Qt;
 184   return Qnil;
 185 }
 186
 187 /* If TABLE is nil, return the current category table.  If TABLE is
 188    not nil, check the validity of TABLE as a category table.  If
 189    valid, return TABLE itself, but if not valid, signal an error of
 190    wrong-type-argument.  */
 191
 192 static Lisp_Object
 193 check_category_table (Lisp_Object table)
 194 {
 195   if (NILP (table))
 196     return BVAR (current_buffer, category_table);
 197   CHECK_TYPE (!NILP (Fcategory_table_p (table)), Qcategory_table_p, table);
 198   return table;
 199 }
 200
 201 DEFUN ("category-table", Fcategory_table, Scategory_table, 0, 0, 0,
 202        doc: /* Return the current category table.
 203 This is the one specified by the current buffer.  */)
 204   (void)
 205 {
 206   return BVAR (current_buffer, category_table);
 207 }
 208
 209 DEFUN ("standard-category-table", Fstandard_category_table,
 210    Sstandard_category_table, 0, 0, 0,
 211        doc: /* Return the standard category table.
 212 This is the one used for new buffers.  */)
 213   (void)
 214 {
 215   return Vstandard_category_table;
 216 }
 217
 218
 219 static void
 220 copy_category_entry (Lisp_Object table, Lisp_Object c, Lisp_Object val)
 221 {
 222   val = Fcopy_sequence (val);
 223   if (CONSP (c))
 224     char_table_set_range (table, XINT (XCAR (c)), XINT (XCDR (c)), val);
 225   else
 226     char_table_set (table, XINT (c), val);
 227 }
 228
 229 /* Return a copy of category table TABLE.  We can't simply use the
 230    function copy-sequence because no contents should be shared between
 231    the original and the copy.  This function is called recursively by
 232    binding TABLE to a sub char table.  */
 233
 234 static Lisp_Object
 235 copy_category_table (Lisp_Object table)
 236 {
 237   table = copy_char_table (table);
 238
 239   if (! NILP (XCHAR_TABLE (table)->defalt))
 240     set_char_table_defalt (table,
 241                            Fcopy_sequence (XCHAR_TABLE (table)->defalt));
 242   set_char_table_extras
 243     (table, 0, Fcopy_sequence (XCHAR_TABLE (table)->extras[0]));
 244   map_char_table (copy_category_entry, Qnil, table, table);
 245
 246   return table;
 247 }
 248
 249 DEFUN ("copy-category-table", Fcopy_category_table, Scopy_category_table,
 250        0, 1, 0,
 251        doc: /* Construct a new category table and return it.
 252 It is a copy of the TABLE, which defaults to the standard category table.  */)
 253   (Lisp_Object table)
 254 {
 255   if (!NILP (table))
 256     check_category_table (table);
 257   else
 258     table = Vstandard_category_table;
 259
 260   return copy_category_table (table);
 261 }
 262
 263 DEFUN ("make-category-table", Fmake_category_table, Smake_category_table,
 264        0, 0, 0,
 265        doc: /* Construct a new and empty category table and return it.  */)
 266   (void)
 267 {
 268   Lisp_Object val;
 269   int i;
 270
 271   val = Fmake_char_table (Qcategory_table, Qnil);
 272   set_char_table_defalt (val, MAKE_CATEGORY_SET);
 273   for (i = 0; i < (1 << CHARTAB_SIZE_BITS_0); i++)
 274     set_char_table_contents (val, i, MAKE_CATEGORY_SET);
 275   Fset_char_table_extra_slot (val, make_number (0),
 276                               Fmake_vector (make_number (95), Qnil));
 277   return val;
 278 }
 279
 280 DEFUN ("set-category-table", Fset_category_table, Sset_category_table, 1, 1, 0,
 281        doc: /* Specify TABLE as the category table for the current buffer.
 282 Return TABLE.  */)
 283   (Lisp_Object table)
 284 {
 285   int idx;
 286   table = check_category_table (table);
 287   bset_category_table (current_buffer, table);
 288   /* Indicate that this buffer now has a specified category table.  */
 289   idx = PER_BUFFER_VAR_IDX (category_table);
 290   SET_PER_BUFFER_VALUE_P (current_buffer, idx, 1);
 291   return table;
 292 }
 293
 294 \f
 295 Lisp_Object
 296 char_category_set (int c)
 297 {
 298   return CHAR_TABLE_REF (BVAR (current_buffer, category_table), c);
 299 }
 300
 301 DEFUN ("char-category-set", Fchar_category_set, Schar_category_set, 1, 1, 0,
 302        doc: /* Return the category set of CHAR.
 303 usage: (char-category-set CHAR)  */)
 304   (Lisp_Object ch)
 305 {
 306   CHECK_CHARACTER (ch);
 307   return CATEGORY_SET (XFASTINT (ch));
 308 }
 309
 310 DEFUN ("category-set-mnemonics", Fcategory_set_mnemonics,
 311        Scategory_set_mnemonics, 1, 1, 0,
 312        doc: /* Return a string containing mnemonics of the categories in CATEGORY-SET.
 313 CATEGORY-SET is a bool-vector, and the categories \"in\" it are those
 314 that are indexes where t occurs in the bool-vector.
 315 The return value is a string containing those same categories.  */)
 316   (Lisp_Object category_set)
 317 {
 318   int i, j;
 319   char str[96];
 320
 321   CHECK_CATEGORY_SET (category_set);
 322
 323   j = 0;
 324   for (i = 32; i < 127; i++)
 325     if (CATEGORY_MEMBER (i, category_set))
 326       str[j++] = i;
 327   str[j] = '\0';
 328
 329   return build_string (str);
 330 }
 331
 332 DEFUN ("modify-category-entry", Fmodify_category_entry,
 333        Smodify_category_entry, 2, 4, 0,
 334        doc: /* Modify the category set of CHARACTER by adding CATEGORY to it.
 335 The category is changed only for table TABLE, which defaults to
 336 the current buffer's category table.
 337 CHARACTER can be either a single character or a cons representing the
 338 lower and upper ends of an inclusive character range to modify.
 339 CATEGORY must be a category name (a character between ` ' and `~').
 340 Use `describe-categories' to see existing category names.
 341 If optional fourth argument RESET is non-nil,
 342 then delete CATEGORY from the category set instead of adding it.  */)
 343   (Lisp_Object character, Lisp_Object category, Lisp_Object table, Lisp_Object reset)
 344 {
 345   bool set_value;       /* Actual value to be set in category sets.  */
 346   Lisp_Object category_set;
 347   int start, end;
 348   int from, to;
 349
 350   if (INTEGERP (character))
 351     {
 352       CHECK_CHARACTER (character);
 353       start = end = XFASTINT (character);
 354     }
 355   else
 356     {
 357       CHECK_CONS (character);
 358       CHECK_CHARACTER_CAR (character);
 359       CHECK_CHARACTER_CDR (character);
 360       start = XFASTINT (XCAR (character));
 361       end = XFASTINT (XCDR (character));
 362     }
 363
 364   CHECK_CATEGORY (category);
 365   table = check_category_table (table);
 366
 367   if (NILP (CATEGORY_DOCSTRING (table, XFASTINT (category))))
 368     error ("Undefined category: %c", (int) XFASTINT (category));
 369
 370   set_value = NILP (reset);
 371
 372   while (start <= end)
 373     {
 374       from = start, to = end;
 375       category_set = char_table_ref_and_range (table, start, &from, &to);
 376       if (CATEGORY_MEMBER (XFASTINT (category), category_set) != NILP (reset))
 377         {
 378           category_set = Fcopy_sequence (category_set);
 379           set_category_set (category_set, XFASTINT (category), set_value);
 380           category_set = hash_get_category_set (table, category_set);
 381           char_table_set_range (table, start, to, category_set);
 382         }
 383       start = to + 1;
 384     }
 385
 386   return Qnil;
 387 }
 388 \f
 389 /* Return true if there is a word boundary between two word-constituent
 390    characters C1 and C2 if they appear in this order.
 391    Use the macro WORD_BOUNDARY_P instead of calling this function
 392    directly.  */
 393
 394 bool
 395 word_boundary_p (int c1, int c2)
 396 {
 397   Lisp_Object category_set1, category_set2;
 398   Lisp_Object tail;
 399   bool default_result;
 400
 401   if (EQ (CHAR_TABLE_REF (Vchar_script_table, c1),
 402           CHAR_TABLE_REF (Vchar_script_table, c2)))
 403     {
 404       tail = Vword_separating_categories;
 405       default_result = 0;
 406     }
 407   else
 408     {
 409       tail = Vword_combining_categories;
 410       default_result = 1;
 411     }
 412
 413   category_set1 = CATEGORY_SET (c1);
 414   if (NILP (category_set1))
 415     return default_result;
 416   category_set2 = CATEGORY_SET (c2);
 417   if (NILP (category_set2))
 418     return default_result;
 419
 420   for (; CONSP (tail); tail = XCDR (tail))
 421     {
 422       Lisp_Object elt = XCAR (tail);
 423
 424       if (CONSP (elt)
 425           && (NILP (XCAR (elt))
 426               || (CATEGORYP (XCAR (elt))
 427                   && CATEGORY_MEMBER (XFASTINT (XCAR (elt)), category_set1)
 428                   && ! CATEGORY_MEMBER (XFASTINT (XCAR (elt)), category_set2)))
 429           && (NILP (XCDR (elt))
 430               || (CATEGORYP (XCDR (elt))
 431                   && ! CATEGORY_MEMBER (XFASTINT (XCDR (elt)), category_set1)
 432                   && CATEGORY_MEMBER (XFASTINT (XCDR (elt)), category_set2))))
 433         return !default_result;
 434     }
 435   return default_result;
 436 }
 437
 438 \f
 439 void
 440 init_category_once (void)
 441 {
 442   /* This has to be done here, before we call Fmake_char_table.  */
 443   DEFSYM (Qcategory_table, "category-table");
 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   set_char_table_defalt (Vstandard_category_table, 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 (void)
 455 {
 456   DEFSYM (Qcategoryp, "categoryp");
 457   DEFSYM (Qcategorysetp, "categorysetp");
 458   DEFSYM (Qcategory_table_p, "category-table-p");
 459
 460   DEFVAR_LISP ("word-combining-categories", Vword_combining_categories,
 461                doc: /* List of pair (cons) of categories to determine word boundary.
 462
 463 Emacs treats a sequence of word constituent characters as a single
 464 word (i.e. finds no word boundary between them) only if they belong to
 465 the same script.  But, exceptions are allowed in the following cases.
 466
 467 \(1) The case that characters are in different scripts is controlled
 468 by the variable `word-combining-categories'.
 469
 470 Emacs finds no word boundary between characters of different scripts
 471 if they have categories matching some element of this list.
 472
 473 More precisely, if an element of this list is a cons of category CAT1
 474 and CAT2, and a multibyte character C1 which has CAT1 is followed by
 475 C2 which has CAT2, there's no word boundary between C1 and C2.
 476
 477 For instance, to tell that Han characters followed by Hiragana
 478 characters can form a single word, the element `(?C . ?H)' should be
 479 in this list.
 480
 481 \(2) The case that character are in the same script is controlled by
 482 the variable `word-separating-categories'.
 483
 484 Emacs finds a word boundary between characters of the same script
 485 if they have categories matching some element of this list.
 486
 487 More precisely, if an element of this list is a cons of category CAT1
 488 and CAT2, and a multibyte character C1 which has CAT1 but not CAT2 is
 489 followed by C2 which has CAT2 but not CAT1, there's a word boundary
 490 between C1 and C2.
 491
 492 For instance, to tell that there's a word boundary between Hiragana
 493 and Katakana (both are in the same script `kana'),
 494 the element `(?H . ?K)' should be in this list.  */);
 495
 496   Vword_combining_categories = Qnil;
 497
 498   DEFVAR_LISP ("word-separating-categories", Vword_separating_categories,
 499                doc: /* List of pair (cons) of categories to determine word boundary.
 500 See the documentation of the variable `word-combining-categories'.  */);
 501
 502   Vword_separating_categories = Qnil;
 503
 504   defsubr (&Smake_category_set);
 505   defsubr (&Sdefine_category);
 506   defsubr (&Scategory_docstring);
 507   defsubr (&Sget_unused_category);
 508   defsubr (&Scategory_table_p);
 509   defsubr (&Scategory_table);
 510   defsubr (&Sstandard_category_table);
 511   defsubr (&Scopy_category_table);
 512   defsubr (&Smake_category_table);
 513   defsubr (&Sset_category_table);
 514   defsubr (&Schar_category_set);
 515   defsubr (&Scategory_set_mnemonics);
 516   defsubr (&Smodify_category_entry);
 517
 518   category_table_version = 0;
 519 }