1 /* XEmacs routines to deal with char tables.
2 Copyright (C) 1992, 1995 Free Software Foundation, Inc.
3 Copyright (C) 1995 Sun Microsystems, Inc.
4 Copyright (C) 1995, 1996 Ben Wing.
5 Copyright (C) 1995, 1997, 1999 Electrotechnical Laboratory, JAPAN.
6 Licensed to the Free Software Foundation.
8 This file is part of XEmacs.
10 XEmacs is free software; you can redistribute it and/or modify it
11 under the terms of the GNU General Public License as published by the
12 Free Software Foundation; either version 2, or (at your option) any
15 XEmacs is distributed in the hope that it will be useful, but WITHOUT
16 ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
17 FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
20 You should have received a copy of the GNU General Public License
21 along with XEmacs; 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. */
25 /* Synched up with: Mule 2.3. Not synched with FSF.
27 This file was written independently of the FSF implementation,
28 and is not compatible. */
32 Ben Wing: wrote, for 19.13 (Mule). Some category table stuff
33 loosely based on the original Mule.
34 Jareth Hein: fixed a couple of bugs in the implementation, and
35 added regex support for categories with check_category_at
45 Lisp_Object Qchar_tablep, Qchar_table;
47 Lisp_Object Vall_syntax_tables;
50 Lisp_Object Qcategory_table_p;
51 Lisp_Object Qcategory_designator_p;
52 Lisp_Object Qcategory_table_value_p;
54 Lisp_Object Vstandard_category_table;
56 /* Variables to determine word boundary. */
57 Lisp_Object Vword_combining_categories, Vword_separating_categories;
61 /* A char table maps from ranges of characters to values.
63 Implementing a general data structure that maps from arbitrary
64 ranges of numbers to values is tricky to do efficiently. As it
65 happens, it should suffice (and is usually more convenient, anyway)
66 when dealing with characters to restrict the sorts of ranges that
67 can be assigned values, as follows:
70 2) All characters in a charset.
71 3) All characters in a particular row of a charset, where a "row"
72 means all characters with the same first byte.
73 4) A particular character in a charset.
75 We use char tables to generalize the 256-element vectors now
76 littering the Emacs code.
78 Possible uses (all should be converted at some point):
84 5) keyboard-translate-table?
87 abstract type to generalize the Emacs vectors and Mule
88 vectors-of-vectors goo.
91 /************************************************************************/
92 /* Char Table object */
93 /************************************************************************/
98 mark_char_table_entry (Lisp_Object obj)
100 Lisp_Char_Table_Entry *cte = XCHAR_TABLE_ENTRY (obj);
103 for (i = 0; i < 96; i++)
105 mark_object (cte->level2[i]);
111 char_table_entry_equal (Lisp_Object obj1, Lisp_Object obj2, int depth)
113 Lisp_Char_Table_Entry *cte1 = XCHAR_TABLE_ENTRY (obj1);
114 Lisp_Char_Table_Entry *cte2 = XCHAR_TABLE_ENTRY (obj2);
117 for (i = 0; i < 96; i++)
118 if (!internal_equal (cte1->level2[i], cte2->level2[i], depth + 1))
125 char_table_entry_hash (Lisp_Object obj, int depth)
127 Lisp_Char_Table_Entry *cte = XCHAR_TABLE_ENTRY (obj);
129 return internal_array_hash (cte->level2, 96, depth);
132 static const struct lrecord_description char_table_entry_description[] = {
133 { XD_LISP_OBJECT_ARRAY, offsetof (Lisp_Char_Table_Entry, level2), 96 },
137 DEFINE_LRECORD_IMPLEMENTATION ("char-table-entry", char_table_entry,
138 mark_char_table_entry, internal_object_printer,
139 0, char_table_entry_equal,
140 char_table_entry_hash,
141 char_table_entry_description,
142 Lisp_Char_Table_Entry);
146 mark_char_table (Lisp_Object obj)
148 Lisp_Char_Table *ct = XCHAR_TABLE (obj);
151 for (i = 0; i < NUM_ASCII_CHARS; i++)
152 mark_object (ct->ascii[i]);
154 for (i = 0; i < NUM_LEADING_BYTES; i++)
155 mark_object (ct->level1[i]);
157 return ct->mirror_table;
160 /* WARNING: All functions of this nature need to be written extremely
161 carefully to avoid crashes during GC. Cf. prune_specifiers()
162 and prune_weak_hash_tables(). */
165 prune_syntax_tables (void)
167 Lisp_Object rest, prev = Qnil;
169 for (rest = Vall_syntax_tables;
171 rest = XCHAR_TABLE (rest)->next_table)
173 if (! marked_p (rest))
175 /* This table is garbage. Remove it from the list. */
177 Vall_syntax_tables = XCHAR_TABLE (rest)->next_table;
179 XCHAR_TABLE (prev)->next_table =
180 XCHAR_TABLE (rest)->next_table;
186 char_table_type_to_symbol (enum char_table_type type)
191 case CHAR_TABLE_TYPE_GENERIC: return Qgeneric;
192 case CHAR_TABLE_TYPE_SYNTAX: return Qsyntax;
193 case CHAR_TABLE_TYPE_DISPLAY: return Qdisplay;
194 case CHAR_TABLE_TYPE_CHAR: return Qchar;
196 case CHAR_TABLE_TYPE_CATEGORY: return Qcategory;
201 static enum char_table_type
202 symbol_to_char_table_type (Lisp_Object symbol)
204 CHECK_SYMBOL (symbol);
206 if (EQ (symbol, Qgeneric)) return CHAR_TABLE_TYPE_GENERIC;
207 if (EQ (symbol, Qsyntax)) return CHAR_TABLE_TYPE_SYNTAX;
208 if (EQ (symbol, Qdisplay)) return CHAR_TABLE_TYPE_DISPLAY;
209 if (EQ (symbol, Qchar)) return CHAR_TABLE_TYPE_CHAR;
211 if (EQ (symbol, Qcategory)) return CHAR_TABLE_TYPE_CATEGORY;
214 signal_simple_error ("Unrecognized char table type", symbol);
215 return CHAR_TABLE_TYPE_GENERIC; /* not reached */
219 print_chartab_range (Emchar first, Emchar last, Lisp_Object val,
220 Lisp_Object printcharfun)
224 write_c_string (" (", printcharfun);
225 print_internal (make_char (first), printcharfun, 0);
226 write_c_string (" ", printcharfun);
227 print_internal (make_char (last), printcharfun, 0);
228 write_c_string (") ", printcharfun);
232 write_c_string (" ", printcharfun);
233 print_internal (make_char (first), printcharfun, 0);
234 write_c_string (" ", printcharfun);
236 print_internal (val, printcharfun, 1);
242 print_chartab_charset_row (Lisp_Object charset,
244 Lisp_Char_Table_Entry *cte,
245 Lisp_Object printcharfun)
248 Lisp_Object cat = Qunbound;
251 for (i = 32; i < 128; i++)
253 Lisp_Object pam = cte->level2[i - 32];
265 print_chartab_range (MAKE_CHAR (charset, first, 0),
266 MAKE_CHAR (charset, i - 1, 0),
269 print_chartab_range (MAKE_CHAR (charset, row, first),
270 MAKE_CHAR (charset, row, i - 1),
280 print_chartab_range (MAKE_CHAR (charset, first, 0),
281 MAKE_CHAR (charset, i - 1, 0),
284 print_chartab_range (MAKE_CHAR (charset, row, first),
285 MAKE_CHAR (charset, row, i - 1),
291 print_chartab_two_byte_charset (Lisp_Object charset,
292 Lisp_Char_Table_Entry *cte,
293 Lisp_Object printcharfun)
297 for (i = 32; i < 128; i++)
299 Lisp_Object jen = cte->level2[i - 32];
301 if (!CHAR_TABLE_ENTRYP (jen))
305 write_c_string (" [", printcharfun);
306 print_internal (XCHARSET_NAME (charset), printcharfun, 0);
307 sprintf (buf, " %d] ", i);
308 write_c_string (buf, printcharfun);
309 print_internal (jen, printcharfun, 0);
312 print_chartab_charset_row (charset, i, XCHAR_TABLE_ENTRY (jen),
320 print_char_table (Lisp_Object obj, Lisp_Object printcharfun, int escapeflag)
322 Lisp_Char_Table *ct = XCHAR_TABLE (obj);
325 sprintf (buf, "#s(char-table type %s data (",
326 string_data (symbol_name (XSYMBOL
327 (char_table_type_to_symbol (ct->type)))));
328 write_c_string (buf, printcharfun);
330 /* Now write out the ASCII/Control-1 stuff. */
334 Lisp_Object val = Qunbound;
336 for (i = 0; i < NUM_ASCII_CHARS; i++)
345 if (!EQ (ct->ascii[i], val))
347 print_chartab_range (first, i - 1, val, printcharfun);
354 print_chartab_range (first, i - 1, val, printcharfun);
361 for (i = MIN_LEADING_BYTE; i < MIN_LEADING_BYTE + NUM_LEADING_BYTES;
364 Lisp_Object ann = ct->level1[i - MIN_LEADING_BYTE];
365 Lisp_Object charset = CHARSET_BY_LEADING_BYTE (i);
367 if (!CHARSETP (charset) || i == LEADING_BYTE_ASCII
368 || i == LEADING_BYTE_CONTROL_1)
370 if (!CHAR_TABLE_ENTRYP (ann))
372 write_c_string (" ", printcharfun);
373 print_internal (XCHARSET_NAME (charset),
375 write_c_string (" ", printcharfun);
376 print_internal (ann, printcharfun, 0);
380 Lisp_Char_Table_Entry *cte = XCHAR_TABLE_ENTRY (ann);
381 if (XCHARSET_DIMENSION (charset) == 1)
382 print_chartab_charset_row (charset, -1, cte, printcharfun);
384 print_chartab_two_byte_charset (charset, cte, printcharfun);
390 write_c_string ("))", printcharfun);
394 char_table_equal (Lisp_Object obj1, Lisp_Object obj2, int depth)
396 Lisp_Char_Table *ct1 = XCHAR_TABLE (obj1);
397 Lisp_Char_Table *ct2 = XCHAR_TABLE (obj2);
400 if (CHAR_TABLE_TYPE (ct1) != CHAR_TABLE_TYPE (ct2))
403 for (i = 0; i < NUM_ASCII_CHARS; i++)
404 if (!internal_equal (ct1->ascii[i], ct2->ascii[i], depth + 1))
408 for (i = 0; i < NUM_LEADING_BYTES; i++)
409 if (!internal_equal (ct1->level1[i], ct2->level1[i], depth + 1))
417 char_table_hash (Lisp_Object obj, int depth)
419 Lisp_Char_Table *ct = XCHAR_TABLE (obj);
420 unsigned long hashval = internal_array_hash (ct->ascii, NUM_ASCII_CHARS,
423 hashval = HASH2 (hashval,
424 internal_array_hash (ct->level1, NUM_LEADING_BYTES, depth));
429 static const struct lrecord_description char_table_description[] = {
430 { XD_LISP_OBJECT_ARRAY, offsetof (Lisp_Char_Table, ascii), NUM_ASCII_CHARS },
432 { XD_LISP_OBJECT_ARRAY, offsetof (Lisp_Char_Table, level1), NUM_LEADING_BYTES },
434 { XD_LISP_OBJECT, offsetof (Lisp_Char_Table, mirror_table) },
435 { XD_LO_LINK, offsetof (Lisp_Char_Table, next_table) },
439 DEFINE_LRECORD_IMPLEMENTATION ("char-table", char_table,
440 mark_char_table, print_char_table, 0,
441 char_table_equal, char_table_hash,
442 char_table_description,
445 DEFUN ("char-table-p", Fchar_table_p, 1, 1, 0, /*
446 Return non-nil if OBJECT is a char table.
448 A char table is a table that maps characters (or ranges of characters)
449 to values. Char tables are specialized for characters, only allowing
450 particular sorts of ranges to be assigned values. Although this
451 loses in generality, it makes for extremely fast (constant-time)
452 lookups, and thus is feasible for applications that do an extremely
453 large number of lookups (e.g. scanning a buffer for a character in
454 a particular syntax, where a lookup in the syntax table must occur
457 When Mule support exists, the types of ranges that can be assigned
462 -- a single row in a two-octet charset
463 -- a single character
465 When Mule support is not present, the types of ranges that can be
469 -- a single character
471 To create a char table, use `make-char-table'. To modify a char
472 table, use `put-char-table' or `remove-char-table'. To retrieve the
473 value for a particular character, use `get-char-table'. See also
474 `map-char-table', `clear-char-table', `copy-char-table',
475 `valid-char-table-type-p', `char-table-type-list', `valid-char-table-value-p',
476 and `check-char-table-value'.
480 return CHAR_TABLEP (object) ? Qt : Qnil;
483 DEFUN ("char-table-type-list", Fchar_table_type_list, 0, 0, 0, /*
484 Return a list of the recognized char table types.
485 See `valid-char-table-type-p'.
490 return list5 (Qchar, Qcategory, Qdisplay, Qgeneric, Qsyntax);
492 return list4 (Qchar, Qdisplay, Qgeneric, Qsyntax);
496 DEFUN ("valid-char-table-type-p", Fvalid_char_table_type_p, 1, 1, 0, /*
497 Return t if TYPE if a recognized char table type.
499 Each char table type is used for a different purpose and allows different
500 sorts of values. The different char table types are
503 Used for category tables, which specify the regexp categories
504 that a character is in. The valid values are nil or a
505 bit vector of 95 elements. Higher-level Lisp functions are
506 provided for working with category tables. Currently categories
507 and category tables only exist when Mule support is present.
509 A generalized char table, for mapping from one character to
510 another. Used for case tables, syntax matching tables,
511 `keyboard-translate-table', etc. The valid values are characters.
513 An even more generalized char table, for mapping from a
514 character to anything.
516 Used for display tables, which specify how a particular character
517 is to appear when displayed. #### Not yet implemented.
519 Used for syntax tables, which specify the syntax of a particular
520 character. Higher-level Lisp functions are provided for
521 working with syntax tables. The valid values are integers.
526 return (EQ (type, Qchar) ||
528 EQ (type, Qcategory) ||
530 EQ (type, Qdisplay) ||
531 EQ (type, Qgeneric) ||
532 EQ (type, Qsyntax)) ? Qt : Qnil;
535 DEFUN ("char-table-type", Fchar_table_type, 1, 1, 0, /*
536 Return the type of char table TABLE.
537 See `valid-char-table-type-p'.
541 CHECK_CHAR_TABLE (table);
542 return char_table_type_to_symbol (XCHAR_TABLE (table)->type);
546 fill_char_table (Lisp_Char_Table *ct, Lisp_Object value)
550 for (i = 0; i < NUM_ASCII_CHARS; i++)
551 ct->ascii[i] = value;
553 for (i = 0; i < NUM_LEADING_BYTES; i++)
554 ct->level1[i] = value;
557 if (ct->type == CHAR_TABLE_TYPE_SYNTAX)
558 update_syntax_table (ct);
561 DEFUN ("reset-char-table", Freset_char_table, 1, 1, 0, /*
562 Reset a char table to its default state.
568 CHECK_CHAR_TABLE (table);
569 ct = XCHAR_TABLE (table);
573 case CHAR_TABLE_TYPE_CHAR:
574 fill_char_table (ct, make_char (0));
576 case CHAR_TABLE_TYPE_DISPLAY:
577 case CHAR_TABLE_TYPE_GENERIC:
579 case CHAR_TABLE_TYPE_CATEGORY:
581 fill_char_table (ct, Qnil);
584 case CHAR_TABLE_TYPE_SYNTAX:
585 fill_char_table (ct, make_int (Sinherit));
595 DEFUN ("make-char-table", Fmake_char_table, 1, 1, 0, /*
596 Return a new, empty char table of type TYPE.
597 Currently recognized types are 'char, 'category, 'display, 'generic,
598 and 'syntax. See `valid-char-table-type-p'.
604 enum char_table_type ty = symbol_to_char_table_type (type);
606 ct = alloc_lcrecord_type (Lisp_Char_Table, &lrecord_char_table);
608 if (ty == CHAR_TABLE_TYPE_SYNTAX)
610 ct->mirror_table = Fmake_char_table (Qgeneric);
611 fill_char_table (XCHAR_TABLE (ct->mirror_table),
615 ct->mirror_table = Qnil;
616 ct->next_table = Qnil;
617 XSETCHAR_TABLE (obj, ct);
618 if (ty == CHAR_TABLE_TYPE_SYNTAX)
620 ct->next_table = Vall_syntax_tables;
621 Vall_syntax_tables = obj;
623 Freset_char_table (obj);
630 make_char_table_entry (Lisp_Object initval)
634 Lisp_Char_Table_Entry *cte =
635 alloc_lcrecord_type (Lisp_Char_Table_Entry, &lrecord_char_table_entry);
637 for (i = 0; i < 96; i++)
638 cte->level2[i] = initval;
640 XSETCHAR_TABLE_ENTRY (obj, cte);
645 copy_char_table_entry (Lisp_Object entry)
647 Lisp_Char_Table_Entry *cte = XCHAR_TABLE_ENTRY (entry);
650 Lisp_Char_Table_Entry *ctenew =
651 alloc_lcrecord_type (Lisp_Char_Table_Entry, &lrecord_char_table_entry);
653 for (i = 0; i < 96; i++)
655 Lisp_Object new = cte->level2[i];
656 if (CHAR_TABLE_ENTRYP (new))
657 ctenew->level2[i] = copy_char_table_entry (new);
659 ctenew->level2[i] = new;
662 XSETCHAR_TABLE_ENTRY (obj, ctenew);
668 DEFUN ("copy-char-table", Fcopy_char_table, 1, 1, 0, /*
669 Make a new char table which is a copy of OLD-TABLE.
670 It will contain the same values for the same characters and ranges
671 as OLD-TABLE. The values will not themselves be copied.
675 Lisp_Char_Table *ct, *ctnew;
679 CHECK_CHAR_TABLE (old_table);
680 ct = XCHAR_TABLE (old_table);
681 ctnew = alloc_lcrecord_type (Lisp_Char_Table, &lrecord_char_table);
682 ctnew->type = ct->type;
684 for (i = 0; i < NUM_ASCII_CHARS; i++)
686 Lisp_Object new = ct->ascii[i];
688 assert (! (CHAR_TABLE_ENTRYP (new)));
690 ctnew->ascii[i] = new;
695 for (i = 0; i < NUM_LEADING_BYTES; i++)
697 Lisp_Object new = ct->level1[i];
698 if (CHAR_TABLE_ENTRYP (new))
699 ctnew->level1[i] = copy_char_table_entry (new);
701 ctnew->level1[i] = new;
706 if (CHAR_TABLEP (ct->mirror_table))
707 ctnew->mirror_table = Fcopy_char_table (ct->mirror_table);
709 ctnew->mirror_table = ct->mirror_table;
710 ctnew->next_table = Qnil;
711 XSETCHAR_TABLE (obj, ctnew);
712 if (ctnew->type == CHAR_TABLE_TYPE_SYNTAX)
714 ctnew->next_table = Vall_syntax_tables;
715 Vall_syntax_tables = obj;
721 decode_char_table_range (Lisp_Object range, struct chartab_range *outrange)
724 outrange->type = CHARTAB_RANGE_ALL;
725 else if (CHAR_OR_CHAR_INTP (range))
727 outrange->type = CHARTAB_RANGE_CHAR;
728 outrange->ch = XCHAR_OR_CHAR_INT (range);
732 signal_simple_error ("Range must be t or a character", range);
734 else if (VECTORP (range))
736 Lisp_Vector *vec = XVECTOR (range);
737 Lisp_Object *elts = vector_data (vec);
738 if (vector_length (vec) != 2)
739 signal_simple_error ("Length of charset row vector must be 2",
741 outrange->type = CHARTAB_RANGE_ROW;
742 outrange->charset = Fget_charset (elts[0]);
744 outrange->row = XINT (elts[1]);
745 switch (XCHARSET_TYPE (outrange->charset))
747 case CHARSET_TYPE_94:
748 case CHARSET_TYPE_96:
749 signal_simple_error ("Charset in row vector must be multi-byte",
751 case CHARSET_TYPE_94X94:
752 check_int_range (outrange->row, 33, 126);
754 case CHARSET_TYPE_96X96:
755 check_int_range (outrange->row, 32, 127);
763 if (!CHARSETP (range) && !SYMBOLP (range))
765 ("Char table range must be t, charset, char, or vector", range);
766 outrange->type = CHARTAB_RANGE_CHARSET;
767 outrange->charset = Fget_charset (range);
774 /* called from CHAR_TABLE_VALUE(). */
776 get_non_ascii_char_table_value (Lisp_Char_Table *ct, int leading_byte,
780 Lisp_Object charset = CHARSET_BY_LEADING_BYTE (leading_byte);
783 BREAKUP_CHAR_1_UNSAFE (c, charset, byte1, byte2);
784 val = ct->level1[leading_byte - MIN_LEADING_BYTE];
785 if (CHAR_TABLE_ENTRYP (val))
787 Lisp_Char_Table_Entry *cte = XCHAR_TABLE_ENTRY (val);
788 val = cte->level2[byte1 - 32];
789 if (CHAR_TABLE_ENTRYP (val))
791 cte = XCHAR_TABLE_ENTRY (val);
792 assert (byte2 >= 32);
793 val = cte->level2[byte2 - 32];
794 assert (!CHAR_TABLE_ENTRYP (val));
804 get_char_table (Emchar ch, Lisp_Char_Table *ct)
812 BREAKUP_CHAR (ch, charset, byte1, byte2);
814 if (EQ (charset, Vcharset_ascii))
815 val = ct->ascii[byte1];
816 else if (EQ (charset, Vcharset_control_1))
817 val = ct->ascii[byte1 + 128];
820 int lb = XCHARSET_LEADING_BYTE (charset) - MIN_LEADING_BYTE;
821 val = ct->level1[lb];
822 if (CHAR_TABLE_ENTRYP (val))
824 Lisp_Char_Table_Entry *cte = XCHAR_TABLE_ENTRY (val);
825 val = cte->level2[byte1 - 32];
826 if (CHAR_TABLE_ENTRYP (val))
828 cte = XCHAR_TABLE_ENTRY (val);
829 assert (byte2 >= 32);
830 val = cte->level2[byte2 - 32];
831 assert (!CHAR_TABLE_ENTRYP (val));
839 return ct->ascii[(unsigned char)ch];
840 #endif /* not MULE */
844 DEFUN ("get-char-table", Fget_char_table, 2, 2, 0, /*
845 Find value for char CH in TABLE.
851 CHECK_CHAR_TABLE (table);
852 ct = XCHAR_TABLE (table);
853 CHECK_CHAR_COERCE_INT (ch);
855 return get_char_table (XCHAR (ch), ct);
858 DEFUN ("get-range-char-table", Fget_range_char_table, 2, 3, 0, /*
859 Find value for a range in TABLE.
860 If there is more than one value, return MULTI (defaults to nil).
862 (range, table, multi))
865 struct chartab_range rainj;
867 if (CHAR_OR_CHAR_INTP (range))
868 return Fget_char_table (range, table);
869 CHECK_CHAR_TABLE (table);
870 ct = XCHAR_TABLE (table);
872 decode_char_table_range (range, &rainj);
875 case CHARTAB_RANGE_ALL:
878 Lisp_Object first = ct->ascii[0];
880 for (i = 1; i < NUM_ASCII_CHARS; i++)
881 if (!EQ (first, ct->ascii[i]))
885 for (i = MIN_LEADING_BYTE; i < MIN_LEADING_BYTE + NUM_LEADING_BYTES;
888 if (!CHARSETP (CHARSET_BY_LEADING_BYTE (i))
889 || i == LEADING_BYTE_ASCII
890 || i == LEADING_BYTE_CONTROL_1)
892 if (!EQ (first, ct->level1[i - MIN_LEADING_BYTE]))
901 case CHARTAB_RANGE_CHARSET:
902 if (EQ (rainj.charset, Vcharset_ascii))
905 Lisp_Object first = ct->ascii[0];
907 for (i = 1; i < 128; i++)
908 if (!EQ (first, ct->ascii[i]))
913 if (EQ (rainj.charset, Vcharset_control_1))
916 Lisp_Object first = ct->ascii[128];
918 for (i = 129; i < 160; i++)
919 if (!EQ (first, ct->ascii[i]))
925 Lisp_Object val = ct->level1[XCHARSET_LEADING_BYTE (rainj.charset) -
927 if (CHAR_TABLE_ENTRYP (val))
932 case CHARTAB_RANGE_ROW:
934 Lisp_Object val = ct->level1[XCHARSET_LEADING_BYTE (rainj.charset) -
936 if (!CHAR_TABLE_ENTRYP (val))
938 val = XCHAR_TABLE_ENTRY (val)->level2[rainj.row - 32];
939 if (CHAR_TABLE_ENTRYP (val))
943 #endif /* not MULE */
949 return Qnil; /* not reached */
953 check_valid_char_table_value (Lisp_Object value, enum char_table_type type,
958 case CHAR_TABLE_TYPE_SYNTAX:
959 if (!ERRB_EQ (errb, ERROR_ME))
960 return INTP (value) || (CONSP (value) && INTP (XCAR (value))
961 && CHAR_OR_CHAR_INTP (XCDR (value)));
964 Lisp_Object cdr = XCDR (value);
965 CHECK_INT (XCAR (value));
966 CHECK_CHAR_COERCE_INT (cdr);
973 case CHAR_TABLE_TYPE_CATEGORY:
974 if (!ERRB_EQ (errb, ERROR_ME))
975 return CATEGORY_TABLE_VALUEP (value);
976 CHECK_CATEGORY_TABLE_VALUE (value);
980 case CHAR_TABLE_TYPE_GENERIC:
983 case CHAR_TABLE_TYPE_DISPLAY:
985 maybe_signal_simple_error ("Display char tables not yet implemented",
986 value, Qchar_table, errb);
989 case CHAR_TABLE_TYPE_CHAR:
990 if (!ERRB_EQ (errb, ERROR_ME))
991 return CHAR_OR_CHAR_INTP (value);
992 CHECK_CHAR_COERCE_INT (value);
999 return 0; /* not reached */
1003 canonicalize_char_table_value (Lisp_Object value, enum char_table_type type)
1007 case CHAR_TABLE_TYPE_SYNTAX:
1010 Lisp_Object car = XCAR (value);
1011 Lisp_Object cdr = XCDR (value);
1012 CHECK_CHAR_COERCE_INT (cdr);
1013 return Fcons (car, cdr);
1016 case CHAR_TABLE_TYPE_CHAR:
1017 CHECK_CHAR_COERCE_INT (value);
1025 DEFUN ("valid-char-table-value-p", Fvalid_char_table_value_p, 2, 2, 0, /*
1026 Return non-nil if VALUE is a valid value for CHAR-TABLE-TYPE.
1028 (value, char_table_type))
1030 enum char_table_type type = symbol_to_char_table_type (char_table_type);
1032 return check_valid_char_table_value (value, type, ERROR_ME_NOT) ? Qt : Qnil;
1035 DEFUN ("check-valid-char-table-value", Fcheck_valid_char_table_value, 2, 2, 0, /*
1036 Signal an error if VALUE is not a valid value for CHAR-TABLE-TYPE.
1038 (value, char_table_type))
1040 enum char_table_type type = symbol_to_char_table_type (char_table_type);
1042 check_valid_char_table_value (value, type, ERROR_ME);
1046 /* Assign VAL to all characters in RANGE in char table CT. */
1049 put_char_table (Lisp_Char_Table *ct, struct chartab_range *range,
1052 switch (range->type)
1054 case CHARTAB_RANGE_ALL:
1055 fill_char_table (ct, val);
1056 return; /* avoid the duplicate call to update_syntax_table() below,
1057 since fill_char_table() also did that. */
1060 case CHARTAB_RANGE_CHARSET:
1061 if (EQ (range->charset, Vcharset_ascii))
1064 for (i = 0; i < 128; i++)
1067 else if (EQ (range->charset, Vcharset_control_1))
1070 for (i = 128; i < 160; i++)
1075 int lb = XCHARSET_LEADING_BYTE (range->charset) - MIN_LEADING_BYTE;
1076 ct->level1[lb] = val;
1080 case CHARTAB_RANGE_ROW:
1082 Lisp_Char_Table_Entry *cte;
1083 int lb = XCHARSET_LEADING_BYTE (range->charset) - MIN_LEADING_BYTE;
1084 /* make sure that there is a separate entry for the row. */
1085 if (!CHAR_TABLE_ENTRYP (ct->level1[lb]))
1086 ct->level1[lb] = make_char_table_entry (ct->level1[lb]);
1087 cte = XCHAR_TABLE_ENTRY (ct->level1[lb]);
1088 cte->level2[range->row - 32] = val;
1093 case CHARTAB_RANGE_CHAR:
1096 Lisp_Object charset;
1099 BREAKUP_CHAR (range->ch, charset, byte1, byte2);
1100 if (EQ (charset, Vcharset_ascii))
1101 ct->ascii[byte1] = val;
1102 else if (EQ (charset, Vcharset_control_1))
1103 ct->ascii[byte1 + 128] = val;
1106 Lisp_Char_Table_Entry *cte;
1107 int lb = XCHARSET_LEADING_BYTE (charset) - MIN_LEADING_BYTE;
1108 /* make sure that there is a separate entry for the row. */
1109 if (!CHAR_TABLE_ENTRYP (ct->level1[lb]))
1110 ct->level1[lb] = make_char_table_entry (ct->level1[lb]);
1111 cte = XCHAR_TABLE_ENTRY (ct->level1[lb]);
1112 /* now CTE is a char table entry for the charset;
1113 each entry is for a single row (or character of
1114 a one-octet charset). */
1115 if (XCHARSET_DIMENSION (charset) == 1)
1116 cte->level2[byte1 - 32] = val;
1119 /* assigning to one character in a two-octet charset. */
1120 /* make sure that the charset row contains a separate
1121 entry for each character. */
1122 if (!CHAR_TABLE_ENTRYP (cte->level2[byte1 - 32]))
1123 cte->level2[byte1 - 32] =
1124 make_char_table_entry (cte->level2[byte1 - 32]);
1125 cte = XCHAR_TABLE_ENTRY (cte->level2[byte1 - 32]);
1126 cte->level2[byte2 - 32] = val;
1130 #else /* not MULE */
1131 ct->ascii[(unsigned char) (range->ch)] = val;
1133 #endif /* not MULE */
1136 if (ct->type == CHAR_TABLE_TYPE_SYNTAX)
1137 update_syntax_table (ct);
1140 DEFUN ("put-char-table", Fput_char_table, 3, 3, 0, /*
1141 Set the value for chars in RANGE to be VAL in TABLE.
1143 RANGE specifies one or more characters to be affected and should be
1144 one of the following:
1146 -- t (all characters are affected)
1147 -- A charset (only allowed when Mule support is present)
1148 -- A vector of two elements: a two-octet charset and a row number
1149 (only allowed when Mule support is present)
1150 -- A single character
1152 VAL must be a value appropriate for the type of TABLE.
1153 See `valid-char-table-type-p'.
1155 (range, val, table))
1157 Lisp_Char_Table *ct;
1158 struct chartab_range rainj;
1160 CHECK_CHAR_TABLE (table);
1161 ct = XCHAR_TABLE (table);
1162 check_valid_char_table_value (val, ct->type, ERROR_ME);
1163 decode_char_table_range (range, &rainj);
1164 val = canonicalize_char_table_value (val, ct->type);
1165 put_char_table (ct, &rainj, val);
1169 /* Map FN over the ASCII chars in CT. */
1172 map_over_charset_ascii (Lisp_Char_Table *ct,
1173 int (*fn) (struct chartab_range *range,
1174 Lisp_Object val, void *arg),
1177 struct chartab_range rainj;
1186 rainj.type = CHARTAB_RANGE_CHAR;
1188 for (i = start, retval = 0; i < stop && retval == 0; i++)
1190 rainj.ch = (Emchar) i;
1191 retval = (fn) (&rainj, ct->ascii[i], arg);
1199 /* Map FN over the Control-1 chars in CT. */
1202 map_over_charset_control_1 (Lisp_Char_Table *ct,
1203 int (*fn) (struct chartab_range *range,
1204 Lisp_Object val, void *arg),
1207 struct chartab_range rainj;
1210 int stop = start + 32;
1212 rainj.type = CHARTAB_RANGE_CHAR;
1214 for (i = start, retval = 0; i < stop && retval == 0; i++)
1216 rainj.ch = (Emchar) (i);
1217 retval = (fn) (&rainj, ct->ascii[i], arg);
1223 /* Map FN over the row ROW of two-byte charset CHARSET.
1224 There must be a separate value for that row in the char table.
1225 CTE specifies the char table entry for CHARSET. */
1228 map_over_charset_row (Lisp_Char_Table_Entry *cte,
1229 Lisp_Object charset, int row,
1230 int (*fn) (struct chartab_range *range,
1231 Lisp_Object val, void *arg),
1234 Lisp_Object val = cte->level2[row - 32];
1236 if (!CHAR_TABLE_ENTRYP (val))
1238 struct chartab_range rainj;
1240 rainj.type = CHARTAB_RANGE_ROW;
1241 rainj.charset = charset;
1243 return (fn) (&rainj, val, arg);
1247 struct chartab_range rainj;
1249 int charset94_p = (XCHARSET_CHARS (charset) == 94);
1250 int start = charset94_p ? 33 : 32;
1251 int stop = charset94_p ? 127 : 128;
1253 cte = XCHAR_TABLE_ENTRY (val);
1255 rainj.type = CHARTAB_RANGE_CHAR;
1257 for (i = start, retval = 0; i < stop && retval == 0; i++)
1259 rainj.ch = MAKE_CHAR (charset, row, i);
1260 retval = (fn) (&rainj, cte->level2[i - 32], arg);
1268 map_over_other_charset (Lisp_Char_Table *ct, int lb,
1269 int (*fn) (struct chartab_range *range,
1270 Lisp_Object val, void *arg),
1273 Lisp_Object val = ct->level1[lb - MIN_LEADING_BYTE];
1274 Lisp_Object charset = CHARSET_BY_LEADING_BYTE (lb);
1276 if (!CHARSETP (charset)
1277 || lb == LEADING_BYTE_ASCII
1278 || lb == LEADING_BYTE_CONTROL_1)
1281 if (!CHAR_TABLE_ENTRYP (val))
1283 struct chartab_range rainj;
1285 rainj.type = CHARTAB_RANGE_CHARSET;
1286 rainj.charset = charset;
1287 return (fn) (&rainj, val, arg);
1291 Lisp_Char_Table_Entry *cte = XCHAR_TABLE_ENTRY (val);
1292 int charset94_p = (XCHARSET_CHARS (charset) == 94);
1293 int start = charset94_p ? 33 : 32;
1294 int stop = charset94_p ? 127 : 128;
1297 if (XCHARSET_DIMENSION (charset) == 1)
1299 struct chartab_range rainj;
1300 rainj.type = CHARTAB_RANGE_CHAR;
1302 for (i = start, retval = 0; i < stop && retval == 0; i++)
1304 rainj.ch = MAKE_CHAR (charset, i, 0);
1305 retval = (fn) (&rainj, cte->level2[i - 32], arg);
1310 for (i = start, retval = 0; i < stop && retval == 0; i++)
1311 retval = map_over_charset_row (cte, charset, i, fn, arg);
1320 /* Map FN (with client data ARG) over range RANGE in char table CT.
1321 Mapping stops the first time FN returns non-zero, and that value
1322 becomes the return value of map_char_table(). */
1325 map_char_table (Lisp_Char_Table *ct,
1326 struct chartab_range *range,
1327 int (*fn) (struct chartab_range *range,
1328 Lisp_Object val, void *arg),
1331 switch (range->type)
1333 case CHARTAB_RANGE_ALL:
1337 retval = map_over_charset_ascii (ct, fn, arg);
1341 retval = map_over_charset_control_1 (ct, fn, arg);
1346 int start = MIN_LEADING_BYTE;
1347 int stop = start + NUM_LEADING_BYTES;
1349 for (i = start, retval = 0; i < stop && retval == 0; i++)
1351 retval = map_over_other_charset (ct, i, fn, arg);
1359 case CHARTAB_RANGE_CHARSET:
1360 return map_over_other_charset (ct,
1361 XCHARSET_LEADING_BYTE (range->charset),
1364 case CHARTAB_RANGE_ROW:
1366 Lisp_Object val = ct->level1[XCHARSET_LEADING_BYTE (range->charset) - MIN_LEADING_BYTE];
1367 if (!CHAR_TABLE_ENTRYP (val))
1369 struct chartab_range rainj;
1371 rainj.type = CHARTAB_RANGE_ROW;
1372 rainj.charset = range->charset;
1373 rainj.row = range->row;
1374 return (fn) (&rainj, val, arg);
1377 return map_over_charset_row (XCHAR_TABLE_ENTRY (val),
1378 range->charset, range->row,
1383 case CHARTAB_RANGE_CHAR:
1385 Emchar ch = range->ch;
1386 Lisp_Object val = CHAR_TABLE_VALUE_UNSAFE (ct, ch);
1387 struct chartab_range rainj;
1389 rainj.type = CHARTAB_RANGE_CHAR;
1391 return (fn) (&rainj, val, arg);
1401 struct slow_map_char_table_arg
1403 Lisp_Object function;
1408 slow_map_char_table_fun (struct chartab_range *range,
1409 Lisp_Object val, void *arg)
1411 Lisp_Object ranjarg = Qnil;
1412 struct slow_map_char_table_arg *closure =
1413 (struct slow_map_char_table_arg *) arg;
1415 switch (range->type)
1417 case CHARTAB_RANGE_ALL:
1422 case CHARTAB_RANGE_CHARSET:
1423 ranjarg = XCHARSET_NAME (range->charset);
1426 case CHARTAB_RANGE_ROW:
1427 ranjarg = vector2 (XCHARSET_NAME (range->charset),
1428 make_int (range->row));
1431 case CHARTAB_RANGE_CHAR:
1432 ranjarg = make_char (range->ch);
1438 closure->retval = call2 (closure->function, ranjarg, val);
1439 return !NILP (closure->retval);
1442 DEFUN ("map-char-table", Fmap_char_table, 2, 3, 0, /*
1443 Map FUNCTION over entries in TABLE, calling it with two args,
1444 each key and value in the table.
1446 RANGE specifies a subrange to map over and is in the same format as
1447 the RANGE argument to `put-range-table'. If omitted or t, it defaults to
1450 (function, table, range))
1452 Lisp_Char_Table *ct;
1453 struct slow_map_char_table_arg slarg;
1454 struct gcpro gcpro1, gcpro2;
1455 struct chartab_range rainj;
1457 CHECK_CHAR_TABLE (table);
1458 ct = XCHAR_TABLE (table);
1461 decode_char_table_range (range, &rainj);
1462 slarg.function = function;
1463 slarg.retval = Qnil;
1464 GCPRO2 (slarg.function, slarg.retval);
1465 map_char_table (ct, &rainj, slow_map_char_table_fun, &slarg);
1468 return slarg.retval;
1473 /************************************************************************/
1474 /* Char table read syntax */
1475 /************************************************************************/
1478 chartab_type_validate (Lisp_Object keyword, Lisp_Object value,
1479 Error_behavior errb)
1481 /* #### should deal with ERRB */
1482 symbol_to_char_table_type (value);
1487 chartab_data_validate (Lisp_Object keyword, Lisp_Object value,
1488 Error_behavior errb)
1492 /* #### should deal with ERRB */
1493 EXTERNAL_LIST_LOOP (rest, value)
1495 Lisp_Object range = XCAR (rest);
1496 struct chartab_range dummy;
1500 signal_simple_error ("Invalid list format", value);
1503 if (!CONSP (XCDR (range))
1504 || !NILP (XCDR (XCDR (range))))
1505 signal_simple_error ("Invalid range format", range);
1506 decode_char_table_range (XCAR (range), &dummy);
1507 decode_char_table_range (XCAR (XCDR (range)), &dummy);
1510 decode_char_table_range (range, &dummy);
1517 chartab_instantiate (Lisp_Object data)
1519 Lisp_Object chartab;
1520 Lisp_Object type = Qgeneric;
1521 Lisp_Object dataval = Qnil;
1523 while (!NILP (data))
1525 Lisp_Object keyw = Fcar (data);
1531 if (EQ (keyw, Qtype))
1533 else if (EQ (keyw, Qdata))
1537 chartab = Fmake_char_table (type);
1540 while (!NILP (data))
1542 Lisp_Object range = Fcar (data);
1543 Lisp_Object val = Fcar (Fcdr (data));
1545 data = Fcdr (Fcdr (data));
1548 if (CHAR_OR_CHAR_INTP (XCAR (range)))
1550 Emchar first = XCHAR_OR_CHAR_INT (Fcar (range));
1551 Emchar last = XCHAR_OR_CHAR_INT (Fcar (Fcdr (range)));
1554 for (i = first; i <= last; i++)
1555 Fput_char_table (make_char (i), val, chartab);
1561 Fput_char_table (range, val, chartab);
1570 /************************************************************************/
1571 /* Category Tables, specifically */
1572 /************************************************************************/
1574 DEFUN ("category-table-p", Fcategory_table_p, 1, 1, 0, /*
1575 Return t if ARG is a category table.
1576 A category table is a type of char table used for keeping track of
1577 categories. Categories are used for classifying characters for use
1578 in regexps -- you can refer to a category rather than having to use
1579 a complicated [] expression (and category lookups are significantly
1582 There are 95 different categories available, one for each printable
1583 character (including space) in the ASCII charset. Each category
1584 is designated by one such character, called a "category designator".
1585 They are specified in a regexp using the syntax "\\cX", where X is
1586 a category designator.
1588 A category table specifies, for each character, the categories that
1589 the character is in. Note that a character can be in more than one
1590 category. More specifically, a category table maps from a character
1591 to either the value nil (meaning the character is in no categories)
1592 or a 95-element bit vector, specifying for each of the 95 categories
1593 whether the character is in that category.
1595 Special Lisp functions are provided that abstract this, so you do not
1596 have to directly manipulate bit vectors.
1600 return (CHAR_TABLEP (obj) &&
1601 XCHAR_TABLE_TYPE (obj) == CHAR_TABLE_TYPE_CATEGORY) ?
1606 check_category_table (Lisp_Object obj, Lisp_Object def)
1610 while (NILP (Fcategory_table_p (obj)))
1611 obj = wrong_type_argument (Qcategory_table_p, obj);
1616 check_category_char (Emchar ch, Lisp_Object table,
1617 unsigned int designator, unsigned int not)
1619 REGISTER Lisp_Object temp;
1620 Lisp_Char_Table *ctbl;
1621 #ifdef ERROR_CHECK_TYPECHECK
1622 if (NILP (Fcategory_table_p (table)))
1623 signal_simple_error ("Expected category table", table);
1625 ctbl = XCHAR_TABLE (table);
1626 temp = get_char_table (ch, ctbl);
1631 return bit_vector_bit (XBIT_VECTOR (temp), designator) ? !not : not;
1634 DEFUN ("check-category-at", Fcheck_category_at, 2, 4, 0, /*
1635 Return t if category of a character at POS includes DESIGNATOR,
1636 else return nil. Optional third arg specifies which buffer
1637 \(defaulting to current), and fourth specifies the CATEGORY-TABLE,
1638 \(defaulting to the buffer's category table).
1640 (pos, designator, buffer, category_table))
1645 struct buffer *buf = decode_buffer (buffer, 0);
1648 CHECK_CATEGORY_DESIGNATOR (designator);
1649 des = XCHAR (designator);
1650 ctbl = check_category_table (category_table, Vstandard_category_table);
1651 ch = BUF_FETCH_CHAR (buf, XINT (pos));
1652 return check_category_char (ch, ctbl, des, 0) ? Qt : Qnil;
1655 DEFUN ("char-in-category-p", Fchar_in_category_p, 2, 3, 0, /*
1656 Return t if category of character CHR includes DESIGNATOR, else nil.
1657 Optional third arg specifies the CATEGORY-TABLE to use,
1658 which defaults to the system default table.
1660 (chr, designator, category_table))
1666 CHECK_CATEGORY_DESIGNATOR (designator);
1667 des = XCHAR (designator);
1670 ctbl = check_category_table (category_table, Vstandard_category_table);
1671 return check_category_char (ch, ctbl, des, 0) ? Qt : Qnil;
1674 DEFUN ("category-table", Fcategory_table, 0, 1, 0, /*
1675 Return the current category table.
1676 This is the one specified by the current buffer, or by BUFFER if it
1681 return decode_buffer (buffer, 0)->category_table;
1684 DEFUN ("standard-category-table", Fstandard_category_table, 0, 0, 0, /*
1685 Return the standard category table.
1686 This is the one used for new buffers.
1690 return Vstandard_category_table;
1693 DEFUN ("copy-category-table", Fcopy_category_table, 0, 1, 0, /*
1694 Construct a new category table and return it.
1695 It is a copy of the TABLE, which defaults to the standard category table.
1699 if (NILP (Vstandard_category_table))
1700 return Fmake_char_table (Qcategory);
1702 table = check_category_table (table, Vstandard_category_table);
1703 return Fcopy_char_table (table);
1706 DEFUN ("set-category-table", Fset_category_table, 1, 2, 0, /*
1707 Select a new category table for BUFFER.
1708 One argument, a category table.
1709 BUFFER defaults to the current buffer if omitted.
1713 struct buffer *buf = decode_buffer (buffer, 0);
1714 table = check_category_table (table, Qnil);
1715 buf->category_table = table;
1716 /* Indicate that this buffer now has a specified category table. */
1717 buf->local_var_flags |= XINT (buffer_local_flags.category_table);
1721 DEFUN ("category-designator-p", Fcategory_designator_p, 1, 1, 0, /*
1722 Return t if ARG is a category designator (a char in the range ' ' to '~').
1726 return CATEGORY_DESIGNATORP (obj) ? Qt : Qnil;
1729 DEFUN ("category-table-value-p", Fcategory_table_value_p, 1, 1, 0, /*
1730 Return t if ARG is a category table value.
1731 Valid values are nil or a bit vector of size 95.
1735 return CATEGORY_TABLE_VALUEP (obj) ? Qt : Qnil;
1739 #define CATEGORYP(x) \
1740 (CHARP (x) && XCHAR (x) >= 0x20 && XCHAR (x) <= 0x7E)
1742 #define CATEGORY_SET(c) \
1743 (get_char_table(c, XCHAR_TABLE(current_buffer->category_table)))
1745 /* Return 1 if CATEGORY_SET contains CATEGORY, else return 0.
1746 The faster version of `!NILP (Faref (category_set, category))'. */
1747 #define CATEGORY_MEMBER(category, category_set) \
1748 (bit_vector_bit(XBIT_VECTOR (category_set), category - 32))
1750 /* Return 1 if there is a word boundary between two word-constituent
1751 characters C1 and C2 if they appear in this order, else return 0.
1752 Use the macro WORD_BOUNDARY_P instead of calling this function
1755 int word_boundary_p (Emchar c1, Emchar c2);
1757 word_boundary_p (Emchar c1, Emchar c2)
1759 Lisp_Object category_set1, category_set2;
1764 if (COMPOSITE_CHAR_P (c1))
1765 c1 = cmpchar_component (c1, 0, 1);
1766 if (COMPOSITE_CHAR_P (c2))
1767 c2 = cmpchar_component (c2, 0, 1);
1770 if (EQ (CHAR_CHARSET (c1), CHAR_CHARSET (c2)))
1772 tail = Vword_separating_categories;
1777 tail = Vword_combining_categories;
1781 category_set1 = CATEGORY_SET (c1);
1782 if (NILP (category_set1))
1783 return default_result;
1784 category_set2 = CATEGORY_SET (c2);
1785 if (NILP (category_set2))
1786 return default_result;
1788 for (; CONSP (tail); tail = XCONS (tail)->cdr)
1790 Lisp_Object elt = XCONS(tail)->car;
1793 && CATEGORYP (XCONS (elt)->car)
1794 && CATEGORYP (XCONS (elt)->cdr)
1795 && CATEGORY_MEMBER (XCHAR (XCONS (elt)->car), category_set1)
1796 && CATEGORY_MEMBER (XCHAR (XCONS (elt)->cdr), category_set2))
1797 return !default_result;
1799 return default_result;
1805 syms_of_chartab (void)
1807 INIT_LRECORD_IMPLEMENTATION (char_table);
1810 INIT_LRECORD_IMPLEMENTATION (char_table_entry);
1812 defsymbol (&Qcategory_table_p, "category-table-p");
1813 defsymbol (&Qcategory_designator_p, "category-designator-p");
1814 defsymbol (&Qcategory_table_value_p, "category-table-value-p");
1817 defsymbol (&Qchar_table, "char-table");
1818 defsymbol (&Qchar_tablep, "char-table-p");
1820 DEFSUBR (Fchar_table_p);
1821 DEFSUBR (Fchar_table_type_list);
1822 DEFSUBR (Fvalid_char_table_type_p);
1823 DEFSUBR (Fchar_table_type);
1824 DEFSUBR (Freset_char_table);
1825 DEFSUBR (Fmake_char_table);
1826 DEFSUBR (Fcopy_char_table);
1827 DEFSUBR (Fget_char_table);
1828 DEFSUBR (Fget_range_char_table);
1829 DEFSUBR (Fvalid_char_table_value_p);
1830 DEFSUBR (Fcheck_valid_char_table_value);
1831 DEFSUBR (Fput_char_table);
1832 DEFSUBR (Fmap_char_table);
1835 DEFSUBR (Fcategory_table_p);
1836 DEFSUBR (Fcategory_table);
1837 DEFSUBR (Fstandard_category_table);
1838 DEFSUBR (Fcopy_category_table);
1839 DEFSUBR (Fset_category_table);
1840 DEFSUBR (Fcheck_category_at);
1841 DEFSUBR (Fchar_in_category_p);
1842 DEFSUBR (Fcategory_designator_p);
1843 DEFSUBR (Fcategory_table_value_p);
1849 vars_of_chartab (void)
1851 /* DO NOT staticpro this. It works just like Vweak_hash_tables. */
1852 Vall_syntax_tables = Qnil;
1853 pdump_wire_list (&Vall_syntax_tables);
1857 structure_type_create_chartab (void)
1859 struct structure_type *st;
1861 st = define_structure_type (Qchar_table, 0, chartab_instantiate);
1863 define_structure_type_keyword (st, Qtype, chartab_type_validate);
1864 define_structure_type_keyword (st, Qdata, chartab_data_validate);
1868 complex_vars_of_chartab (void)
1871 /* Set this now, so first buffer creation can refer to it. */
1872 /* Make it nil before calling copy-category-table
1873 so that copy-category-table will know not to try to copy from garbage */
1874 Vstandard_category_table = Qnil;
1875 Vstandard_category_table = Fcopy_category_table (Qnil);
1876 staticpro (&Vstandard_category_table);
1878 DEFVAR_LISP ("word-combining-categories", &Vword_combining_categories /*
1879 List of pair (cons) of categories to determine word boundary.
1881 Emacs treats a sequence of word constituent characters as a single
1882 word (i.e. finds no word boundary between them) iff they belongs to
1883 the same charset. But, exceptions are allowed in the following cases.
1885 (1) The case that characters are in different charsets is controlled
1886 by the variable `word-combining-categories'.
1888 Emacs finds no word boundary between characters of different charsets
1889 if they have categories matching some element of this list.
1891 More precisely, if an element of this list is a cons of category CAT1
1892 and CAT2, and a multibyte character C1 which has CAT1 is followed by
1893 C2 which has CAT2, there's no word boundary between C1 and C2.
1895 For instance, to tell that ASCII characters and Latin-1 characters can
1896 form a single word, the element `(?l . ?l)' should be in this list
1897 because both characters have the category `l' (Latin characters).
1899 (2) The case that character are in the same charset is controlled by
1900 the variable `word-separating-categories'.
1902 Emacs find a word boundary between characters of the same charset
1903 if they have categories matching some element of this list.
1905 More precisely, if an element of this list is a cons of category CAT1
1906 and CAT2, and a multibyte character C1 which has CAT1 is followed by
1907 C2 which has CAT2, there's a word boundary between C1 and C2.
1909 For instance, to tell that there's a word boundary between Japanese
1910 Hiragana and Japanese Kanji (both are in the same charset), the
1911 element `(?H . ?C) should be in this list.
1914 Vword_combining_categories = Qnil;
1916 DEFVAR_LISP ("word-separating-categories", &Vword_separating_categories /*
1917 List of pair (cons) of categories to determine word boundary.
1918 See the documentation of the variable `word-combining-categories'.
1921 Vword_separating_categories = Qnil;