src/category.c

   1 /* GNU Emacs routines to deal with category tables.
   2    Copyright (C) 1998, 2001, 2004, 2005, 2006 Free Software Foundation, Inc.
   3    Copyright (C) 1995, 1997, 1998, 1999
   4      National Institute of Advanced Industrial Science and Technology (AIST)
   5      Registration Number H14PRO021
   6    Copyright (C) 2003
   7      National Institute of Advanced Industrial Science and Technology (AIST)
   8      Registration Number H13PRO009
   9
  10 This file is part of GNU Emacs.
  11
  12 GNU Emacs is free software; you can redistribute it and/or modify
  13 it under the terms of the GNU General Public License as published by
  14 the Free Software Foundation; either version 2, or (at your option)
  15 any later version.
  16
  17 GNU Emacs is distributed in the hope that it will be useful,
  18 but WITHOUT ANY WARRANTY; without even the implied warranty of
  19 MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  20 GNU General Public License for more details.
  21
  22 You should have received a copy of the GNU General Public License
  23 along with GNU Emacs; see the file COPYING.  If not, write to
  24 the Free Software Foundation, Inc., 51 Franklin Street, Fifth Floor,
  25 Boston, MA 02110-1301, USA.  */
  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   register Lisp_Object tem;
 172   if (NILP (table))
 173     return current_buffer->category_table;
 174   while (tem = Fcategory_table_p (table), NILP (tem))
 175     table = wrong_type_argument (Qcategory_table_p, table);
 176   return table;
 177 }
 178
 179 DEFUN ("category-table", Fcategory_table, Scategory_table, 0, 0, 0,
 180        doc: /* Return the current category table.
 181 This is the one specified by the current buffer.  */)
 182      ()
 183 {
 184   return current_buffer->category_table;
 185 }
 186
 187 DEFUN ("standard-category-table", Fstandard_category_table,
 188    Sstandard_category_table, 0, 0, 0,
 189        doc: /* Return the standard category table.
 190 This is the one used for new buffers.  */)
 191      ()
 192 {
 193   return Vstandard_category_table;
 194 }
 195
 196
 197 static void
 198 copy_category_entry (table, c, val)
 199      Lisp_Object table, c, val;
 200 {
 201   val = Fcopy_sequence (val);
 202   if (CONSP (c))
 203     char_table_set_range (table, XINT (XCAR (c)), XINT (XCDR (c)), val);
 204   else
 205     char_table_set (table, XINT (c), val);
 206 }
 207
 208 /* Return a copy of category table TABLE.  We can't simply use the
 209    function copy-sequence because no contents should be shared between
 210    the original and the copy.  This function is called recursively by
 211    binding TABLE to a sub char table.  */
 212
 213 Lisp_Object
 214 copy_category_table (table)
 215      Lisp_Object table;
 216 {
 217   table = copy_char_table (table);
 218
 219   if (! NILP (XCHAR_TABLE (table)->defalt))
 220     XCHAR_TABLE (table)->defalt
 221       = Fcopy_sequence (XCHAR_TABLE (table)->defalt);
 222   XCHAR_TABLE (table)->extras[0]
 223     = Fcopy_sequence (XCHAR_TABLE (table)->extras[0]);
 224   map_char_table (copy_category_entry, Qnil, table, table);
 225
 226   return table;
 227 }
 228
 229 DEFUN ("copy-category-table", Fcopy_category_table, Scopy_category_table,
 230        0, 1, 0,
 231        doc: /* Construct a new category table and return it.
 232 It is a copy of the TABLE, which defaults to the standard category table.  */)
 233      (table)
 234      Lisp_Object table;
 235 {
 236   if (!NILP (table))
 237     check_category_table (table);
 238   else
 239     table = Vstandard_category_table;
 240
 241   return copy_category_table (table);
 242 }
 243
 244 DEFUN ("make-category-table", Fmake_category_table, Smake_category_table,
 245        0, 0, 0,
 246        doc: /* Construct a new and empty category table and return it.  */)
 247      ()
 248 {
 249   Lisp_Object val;
 250   int i;
 251
 252   val = Fmake_char_table (Qcategory_table, Qnil);
 253   XCHAR_TABLE (val)->defalt = MAKE_CATEGORY_SET;
 254   for (i = 0; i < (1 << CHARTAB_SIZE_BITS_0); i++)
 255     XCHAR_TABLE (val)->contents[i] = MAKE_CATEGORY_SET;
 256   Fset_char_table_extra_slot (val, make_number (0),
 257                               Fmake_vector (make_number (95), Qnil));
 258   return val;
 259 }
 260
 261 DEFUN ("set-category-table", Fset_category_table, Sset_category_table, 1, 1, 0,
 262        doc: /* Specify TABLE as the category table for the current buffer.
 263 Return TABLE.  */)
 264      (table)
 265      Lisp_Object table;
 266 {
 267   int idx;
 268   table = check_category_table (table);
 269   current_buffer->category_table = table;
 270   /* Indicate that this buffer now has a specified category table.  */
 271   idx = PER_BUFFER_VAR_IDX (category_table);
 272   SET_PER_BUFFER_VALUE_P (current_buffer, idx, 1);
 273   return table;
 274 }
 275
 276 \f
 277 Lisp_Object
 278 char_category_set (c)
 279      int c;
 280 {
 281   return CHAR_TABLE_REF (current_buffer->category_table, c);
 282 }
 283
 284 DEFUN ("char-category-set", Fchar_category_set, Schar_category_set, 1, 1, 0,
 285        doc: /* Return the category set of CHAR.  */)
 286      (ch)
 287      Lisp_Object ch;
 288 {
 289   CHECK_NUMBER (ch);
 290   return CATEGORY_SET (XFASTINT (ch));
 291 }
 292
 293 DEFUN ("category-set-mnemonics", Fcategory_set_mnemonics,
 294        Scategory_set_mnemonics, 1, 1, 0,
 295        doc: /* Return a string containing mnemonics of the categories in CATEGORY-SET.
 296 CATEGORY-SET is a bool-vector, and the categories \"in\" it are those
 297 that are indexes where t occurs in the bool-vector.
 298 The return value is a string containing those same categories.  */)
 299      (category_set)
 300      Lisp_Object category_set;
 301 {
 302   int i, j;
 303   char str[96];
 304
 305   CHECK_CATEGORY_SET (category_set);
 306
 307   j = 0;
 308   for (i = 32; i < 127; i++)
 309     if (CATEGORY_MEMBER (i, category_set))
 310       str[j++] = i;
 311   str[j] = '\0';
 312
 313   return build_string (str);
 314 }
 315
 316 void
 317 set_category_set (category_set, category, val)
 318      Lisp_Object category_set, category, val;
 319 {
 320   do {
 321     int idx = XINT (category) / 8;
 322     unsigned char bits = 1 << (XINT (category) % 8);
 323
 324     if (NILP (val))
 325       XCATEGORY_SET (category_set)->data[idx] &= ~bits;
 326     else
 327       XCATEGORY_SET (category_set)->data[idx] |= bits;
 328   } while (0);
 329 }
 330
 331 DEFUN ("modify-category-entry", Fmodify_category_entry,
 332        Smodify_category_entry, 2, 4, 0,
 333        doc: /* Modify the category set of CHARACTER by adding CATEGORY to it.
 334 The category is changed only for table TABLE, which defaults to
 335 the current buffer's category table.
 336 CHARACTER can be either a single character or a cons representing the
 337 lower and upper ends of an inclusive character range to modify.
 338 If optional fourth argument RESET is non-nil,
 339 then delete CATEGORY from the category set instead of adding it.  */)
 340      (character, category, table, reset)
 341      Lisp_Object character, category, table, reset;
 342 {
 343   Lisp_Object set_value;        /* Actual value to be set in category sets.  */
 344   Lisp_Object category_set;
 345   int start, end;
 346   int from, to;
 347
 348   if (INTEGERP (character))
 349     {
 350       CHECK_CHARACTER (character);
 351       start = end = XFASTINT (character);
 352     }
 353   else
 354     {
 355       CHECK_CONS (character);
 356       CHECK_CHARACTER_CAR (character);
 357       CHECK_CHARACTER_CDR (character);
 358       start = XFASTINT (XCAR (character));
 359       end = XFASTINT (XCDR (character));
 360     }
 361
 362   CHECK_CATEGORY (category);
 363   table = check_category_table (table);
 364
 365   if (NILP (CATEGORY_DOCSTRING (table, XFASTINT (category))))
 366     error ("Undefined category: %c", XFASTINT (category));
 367
 368   set_value = NILP (reset) ? Qt : Qnil;
 369
 370   while (start <= end)
 371     {
 372       category_set = char_table_ref_and_range (table, start, &from, &to);
 373       if (CATEGORY_MEMBER (XFASTINT (category), category_set) != NILP (reset))
 374         {
 375           category_set = Fcopy_sequence (category_set);
 376           SET_CATEGORY_SET (category_set, category, set_value);
 377           if (to > end)
 378             char_table_set_range (table, start, end, category_set);
 379           else
 380             char_table_set_range (table, start, to, category_set);
 381         }
 382       start = to + 1;
 383     }
 384
 385   return Qnil;
 386 }
 387 \f
 388 /* Return 1 if there is a word boundary between two word-constituent
 389    characters C1 and C2 if they appear in this order, else return 0.
 390    Use the macro WORD_BOUNDARY_P instead of calling this function
 391    directly.  */
 392
 393 int
 394 word_boundary_p (c1, c2)
 395      int c1, c2;
 396 {
 397   Lisp_Object category_set1, category_set2;
 398   Lisp_Object tail;
 399   int default_result;
 400
 401   if (CHAR_CHARSET (c1) == CHAR_CHARSET (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           && CATEGORYP (XCAR (elt))
 425           && CATEGORYP (XCDR (elt))
 426           && CATEGORY_MEMBER (XFASTINT (XCAR (elt)), category_set1)
 427           && CATEGORY_MEMBER (XFASTINT (XCDR (elt)), category_set2))
 428         return !default_result;
 429     }
 430   return default_result;
 431 }
 432
 433 \f
 434 void
 435 init_category_once ()
 436 {
 437   /* This has to be done here, before we call Fmake_char_table.  */
 438   Qcategory_table = intern ("category-table");
 439   staticpro (&Qcategory_table);
 440
 441   /* Intern this now in case it isn't already done.
 442      Setting this variable twice is harmless.
 443      But don't staticpro it here--that is done in alloc.c.  */
 444   Qchar_table_extra_slots = intern ("char-table-extra-slots");
 445
 446   /* Now we are ready to set up this property, so we can
 447      create category tables.  */
 448   Fput (Qcategory_table, Qchar_table_extra_slots, make_number (2));
 449
 450   Vstandard_category_table = Fmake_char_table (Qcategory_table, Qnil);
 451   /* Set a category set which contains nothing to the default.  */
 452   XCHAR_TABLE (Vstandard_category_table)->defalt = MAKE_CATEGORY_SET;
 453   Fset_char_table_extra_slot (Vstandard_category_table, make_number (0),
 454                               Fmake_vector (make_number (95), Qnil));
 455 }
 456
 457 void
 458 syms_of_category ()
 459 {
 460   Qcategoryp = intern ("categoryp");
 461   staticpro (&Qcategoryp);
 462   Qcategorysetp = intern ("categorysetp");
 463   staticpro (&Qcategorysetp);
 464   Qcategory_table_p = intern ("category-table-p");
 465   staticpro (&Qcategory_table_p);
 466
 467   DEFVAR_LISP ("word-combining-categories", &Vword_combining_categories,
 468                doc: /* List of pair (cons) of categories to determine word boundary.
 469
 470 Emacs treats a sequence of word constituent characters as a single
 471 word (i.e. finds no word boundary between them) iff they belongs to
 472 the same charset.  But, exceptions are allowed in the following cases.
 473
 474 \(1) The case that characters are in different charsets is controlled
 475 by the variable `word-combining-categories'.
 476
 477 Emacs finds no word boundary between characters of different charsets
 478 if they have categories matching some element of this list.
 479
 480 More precisely, if an element of this list is a cons of category CAT1
 481 and CAT2, and a multibyte character C1 which has CAT1 is followed by
 482 C2 which has CAT2, there's no word boundary between C1 and C2.
 483
 484 For instance, to tell that ASCII characters and Latin-1 characters can
 485 form a single word, the element `(?l . ?l)' should be in this list
 486 because both characters have the category `l' (Latin characters).
 487
 488 \(2) The case that character are in the same charset is controlled by
 489 the variable `word-separating-categories'.
 490
 491 Emacs find a word boundary between characters of the same charset
 492 if they have categories matching some element of this list.
 493
 494 More precisely, if an element of this list is a cons of category CAT1
 495 and CAT2, and a multibyte character C1 which has CAT1 is followed by
 496 C2 which has CAT2, there's a word boundary between C1 and C2.
 497
 498 For instance, to tell that there's a word boundary between Japanese
 499 Hiragana and Japanese Kanji (both are in the same charset), the
 500 element `(?H . ?C) should be in this list.  */);
 501
 502   Vword_combining_categories = Qnil;
 503
 504   DEFVAR_LISP ("word-separating-categories", &Vword_separating_categories,
 505                doc: /* List of pair (cons) of categories to determine word boundary.
 506 See the documentation of the variable `word-combining-categories'.  */);
 507
 508   Vword_separating_categories = Qnil;
 509
 510   defsubr (&Smake_category_set);
 511   defsubr (&Sdefine_category);
 512   defsubr (&Scategory_docstring);
 513   defsubr (&Sget_unused_category);
 514   defsubr (&Scategory_table_p);
 515   defsubr (&Scategory_table);
 516   defsubr (&Sstandard_category_table);
 517   defsubr (&Scopy_category_table);
 518   defsubr (&Smake_category_table);
 519   defsubr (&Sset_category_table);
 520   defsubr (&Schar_category_set);
 521   defsubr (&Scategory_set_mnemonics);
 522   defsubr (&Smodify_category_entry);
 523
 524   category_table_version = 0;
 525 }
 526
 527 /* arch-tag: 74ebf524-121b-4d9c-bd68-07f8d708b211
 528    (do not change this comment) */