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
460 -- all characters (represented by t)
462 -- a single row in a two-octet charset (represented by a vector of two
463 elements: a two-octet charset and a row number; the row must be an
464 integer, not a character)
465 -- a single character
467 When Mule support is not present, the types of ranges that can be
470 -- all characters (represented by t)
471 -- a single character
473 To create a char table, use `make-char-table'.
474 To modify a char table, use `put-char-table' or `remove-char-table'.
475 To retrieve the value for a particular character, use `get-char-table'.
476 See also `map-char-table', `clear-char-table', `copy-char-table',
477 `valid-char-table-type-p', `char-table-type-list',
478 `valid-char-table-value-p', and `check-char-table-value'.
482 return CHAR_TABLEP (object) ? Qt : Qnil;
485 DEFUN ("char-table-type-list", Fchar_table_type_list, 0, 0, 0, /*
486 Return a list of the recognized char table types.
487 See `valid-char-table-type-p'.
492 return list5 (Qchar, Qcategory, Qdisplay, Qgeneric, Qsyntax);
494 return list4 (Qchar, Qdisplay, Qgeneric, Qsyntax);
498 DEFUN ("valid-char-table-type-p", Fvalid_char_table_type_p, 1, 1, 0, /*
499 Return t if TYPE if a recognized char table type.
501 Each char table type is used for a different purpose and allows different
502 sorts of values. The different char table types are
505 Used for category tables, which specify the regexp categories
506 that a character is in. The valid values are nil or a
507 bit vector of 95 elements. Higher-level Lisp functions are
508 provided for working with category tables. Currently categories
509 and category tables only exist when Mule support is present.
511 A generalized char table, for mapping from one character to
512 another. Used for case tables, syntax matching tables,
513 `keyboard-translate-table', etc. The valid values are characters.
515 An even more generalized char table, for mapping from a
516 character to anything.
518 Used for display tables, which specify how a particular character
519 is to appear when displayed. #### Not yet implemented.
521 Used for syntax tables, which specify the syntax of a particular
522 character. Higher-level Lisp functions are provided for
523 working with syntax tables. The valid values are integers.
528 return (EQ (type, Qchar) ||
530 EQ (type, Qcategory) ||
532 EQ (type, Qdisplay) ||
533 EQ (type, Qgeneric) ||
534 EQ (type, Qsyntax)) ? Qt : Qnil;
537 DEFUN ("char-table-type", Fchar_table_type, 1, 1, 0, /*
538 Return the type of CHAR-TABLE.
539 See `valid-char-table-type-p'.
543 CHECK_CHAR_TABLE (char_table);
544 return char_table_type_to_symbol (XCHAR_TABLE (char_table)->type);
548 fill_char_table (Lisp_Char_Table *ct, Lisp_Object value)
552 for (i = 0; i < NUM_ASCII_CHARS; i++)
553 ct->ascii[i] = value;
555 for (i = 0; i < NUM_LEADING_BYTES; i++)
556 ct->level1[i] = value;
559 if (ct->type == CHAR_TABLE_TYPE_SYNTAX)
560 update_syntax_table (ct);
563 DEFUN ("reset-char-table", Freset_char_table, 1, 1, 0, /*
564 Reset CHAR-TABLE to its default state.
570 CHECK_CHAR_TABLE (char_table);
571 ct = XCHAR_TABLE (char_table);
575 case CHAR_TABLE_TYPE_CHAR:
576 fill_char_table (ct, make_char (0));
578 case CHAR_TABLE_TYPE_DISPLAY:
579 case CHAR_TABLE_TYPE_GENERIC:
581 case CHAR_TABLE_TYPE_CATEGORY:
583 fill_char_table (ct, Qnil);
586 case CHAR_TABLE_TYPE_SYNTAX:
587 fill_char_table (ct, make_int (Sinherit));
597 DEFUN ("make-char-table", Fmake_char_table, 1, 1, 0, /*
598 Return a new, empty char table of type TYPE.
599 Currently recognized types are 'char, 'category, 'display, 'generic,
600 and 'syntax. See `valid-char-table-type-p'.
606 enum char_table_type ty = symbol_to_char_table_type (type);
608 ct = alloc_lcrecord_type (Lisp_Char_Table, &lrecord_char_table);
610 if (ty == CHAR_TABLE_TYPE_SYNTAX)
612 ct->mirror_table = Fmake_char_table (Qgeneric);
613 fill_char_table (XCHAR_TABLE (ct->mirror_table),
617 ct->mirror_table = Qnil;
618 ct->next_table = Qnil;
619 XSETCHAR_TABLE (obj, ct);
620 if (ty == CHAR_TABLE_TYPE_SYNTAX)
622 ct->next_table = Vall_syntax_tables;
623 Vall_syntax_tables = obj;
625 Freset_char_table (obj);
632 make_char_table_entry (Lisp_Object initval)
636 Lisp_Char_Table_Entry *cte =
637 alloc_lcrecord_type (Lisp_Char_Table_Entry, &lrecord_char_table_entry);
639 for (i = 0; i < 96; i++)
640 cte->level2[i] = initval;
642 XSETCHAR_TABLE_ENTRY (obj, cte);
647 copy_char_table_entry (Lisp_Object entry)
649 Lisp_Char_Table_Entry *cte = XCHAR_TABLE_ENTRY (entry);
652 Lisp_Char_Table_Entry *ctenew =
653 alloc_lcrecord_type (Lisp_Char_Table_Entry, &lrecord_char_table_entry);
655 for (i = 0; i < 96; i++)
657 Lisp_Object new = cte->level2[i];
658 if (CHAR_TABLE_ENTRYP (new))
659 ctenew->level2[i] = copy_char_table_entry (new);
661 ctenew->level2[i] = new;
664 XSETCHAR_TABLE_ENTRY (obj, ctenew);
670 DEFUN ("copy-char-table", Fcopy_char_table, 1, 1, 0, /*
671 Return a new char table which is a copy of CHAR-TABLE.
672 It will contain the same values for the same characters and ranges
673 as CHAR-TABLE. The values will not themselves be copied.
677 Lisp_Char_Table *ct, *ctnew;
681 CHECK_CHAR_TABLE (char_table);
682 ct = XCHAR_TABLE (char_table);
683 ctnew = alloc_lcrecord_type (Lisp_Char_Table, &lrecord_char_table);
684 ctnew->type = ct->type;
686 for (i = 0; i < NUM_ASCII_CHARS; i++)
688 Lisp_Object new = ct->ascii[i];
690 assert (! (CHAR_TABLE_ENTRYP (new)));
692 ctnew->ascii[i] = new;
697 for (i = 0; i < NUM_LEADING_BYTES; i++)
699 Lisp_Object new = ct->level1[i];
700 if (CHAR_TABLE_ENTRYP (new))
701 ctnew->level1[i] = copy_char_table_entry (new);
703 ctnew->level1[i] = new;
708 if (CHAR_TABLEP (ct->mirror_table))
709 ctnew->mirror_table = Fcopy_char_table (ct->mirror_table);
711 ctnew->mirror_table = ct->mirror_table;
712 ctnew->next_table = Qnil;
713 XSETCHAR_TABLE (obj, ctnew);
714 if (ctnew->type == CHAR_TABLE_TYPE_SYNTAX)
716 ctnew->next_table = Vall_syntax_tables;
717 Vall_syntax_tables = obj;
723 decode_char_table_range (Lisp_Object range, struct chartab_range *outrange)
726 outrange->type = CHARTAB_RANGE_ALL;
727 else if (CHAR_OR_CHAR_INTP (range))
729 outrange->type = CHARTAB_RANGE_CHAR;
730 outrange->ch = XCHAR_OR_CHAR_INT (range);
734 signal_simple_error ("Range must be t or a character", range);
736 else if (VECTORP (range))
738 Lisp_Vector *vec = XVECTOR (range);
739 Lisp_Object *elts = vector_data (vec);
740 if (vector_length (vec) != 2)
741 signal_simple_error ("Length of charset row vector must be 2",
743 outrange->type = CHARTAB_RANGE_ROW;
744 outrange->charset = Fget_charset (elts[0]);
746 outrange->row = XINT (elts[1]);
747 switch (XCHARSET_TYPE (outrange->charset))
749 case CHARSET_TYPE_94:
750 case CHARSET_TYPE_96:
751 signal_simple_error ("Charset in row vector must be multi-byte",
753 case CHARSET_TYPE_94X94:
754 check_int_range (outrange->row, 33, 126);
756 case CHARSET_TYPE_96X96:
757 check_int_range (outrange->row, 32, 127);
765 if (!CHARSETP (range) && !SYMBOLP (range))
767 ("Char table range must be t, charset, char, or vector", range);
768 outrange->type = CHARTAB_RANGE_CHARSET;
769 outrange->charset = Fget_charset (range);
776 /* called from CHAR_TABLE_VALUE(). */
778 get_non_ascii_char_table_value (Lisp_Char_Table *ct, int leading_byte,
782 Lisp_Object charset = CHARSET_BY_LEADING_BYTE (leading_byte);
785 BREAKUP_CHAR_1_UNSAFE (c, charset, byte1, byte2);
786 val = ct->level1[leading_byte - MIN_LEADING_BYTE];
787 if (CHAR_TABLE_ENTRYP (val))
789 Lisp_Char_Table_Entry *cte = XCHAR_TABLE_ENTRY (val);
790 val = cte->level2[byte1 - 32];
791 if (CHAR_TABLE_ENTRYP (val))
793 cte = XCHAR_TABLE_ENTRY (val);
794 assert (byte2 >= 32);
795 val = cte->level2[byte2 - 32];
796 assert (!CHAR_TABLE_ENTRYP (val));
806 get_char_table (Emchar ch, Lisp_Char_Table *ct)
814 BREAKUP_CHAR (ch, charset, byte1, byte2);
816 if (EQ (charset, Vcharset_ascii))
817 val = ct->ascii[byte1];
818 else if (EQ (charset, Vcharset_control_1))
819 val = ct->ascii[byte1 + 128];
822 int lb = XCHARSET_LEADING_BYTE (charset) - MIN_LEADING_BYTE;
823 val = ct->level1[lb];
824 if (CHAR_TABLE_ENTRYP (val))
826 Lisp_Char_Table_Entry *cte = XCHAR_TABLE_ENTRY (val);
827 val = cte->level2[byte1 - 32];
828 if (CHAR_TABLE_ENTRYP (val))
830 cte = XCHAR_TABLE_ENTRY (val);
831 assert (byte2 >= 32);
832 val = cte->level2[byte2 - 32];
833 assert (!CHAR_TABLE_ENTRYP (val));
841 return ct->ascii[(unsigned char)ch];
842 #endif /* not MULE */
846 DEFUN ("get-char-table", Fget_char_table, 2, 2, 0, /*
847 Find value for CHARACTER in CHAR-TABLE.
849 (character, char_table))
851 CHECK_CHAR_TABLE (char_table);
852 CHECK_CHAR_COERCE_INT (character);
854 return get_char_table (XCHAR (character), XCHAR_TABLE (char_table));
857 DEFUN ("get-range-char-table", Fget_range_char_table, 2, 3, 0, /*
858 Find value for RANGE in CHAR-TABLE.
859 If there is more than one value, return MULTI (defaults to nil).
861 Valid values for RANGE are single characters, charsets, a row in a
862 two-octet charset, and all characters. See `put-char-table'.
864 (range, char_table, multi))
867 struct chartab_range rainj;
869 if (CHAR_OR_CHAR_INTP (range))
870 return Fget_char_table (range, char_table);
871 CHECK_CHAR_TABLE (char_table);
872 ct = XCHAR_TABLE (char_table);
874 decode_char_table_range (range, &rainj);
877 case CHARTAB_RANGE_ALL:
880 Lisp_Object first = ct->ascii[0];
882 for (i = 1; i < NUM_ASCII_CHARS; i++)
883 if (!EQ (first, ct->ascii[i]))
887 for (i = MIN_LEADING_BYTE; i < MIN_LEADING_BYTE + NUM_LEADING_BYTES;
890 if (!CHARSETP (CHARSET_BY_LEADING_BYTE (i))
891 || i == LEADING_BYTE_ASCII
892 || i == LEADING_BYTE_CONTROL_1)
894 if (!EQ (first, ct->level1[i - MIN_LEADING_BYTE]))
903 case CHARTAB_RANGE_CHARSET:
904 if (EQ (rainj.charset, Vcharset_ascii))
907 Lisp_Object first = ct->ascii[0];
909 for (i = 1; i < 128; i++)
910 if (!EQ (first, ct->ascii[i]))
915 if (EQ (rainj.charset, Vcharset_control_1))
918 Lisp_Object first = ct->ascii[128];
920 for (i = 129; i < 160; i++)
921 if (!EQ (first, ct->ascii[i]))
927 Lisp_Object val = ct->level1[XCHARSET_LEADING_BYTE (rainj.charset) -
929 if (CHAR_TABLE_ENTRYP (val))
934 case CHARTAB_RANGE_ROW:
936 Lisp_Object val = ct->level1[XCHARSET_LEADING_BYTE (rainj.charset) -
938 if (!CHAR_TABLE_ENTRYP (val))
940 val = XCHAR_TABLE_ENTRY (val)->level2[rainj.row - 32];
941 if (CHAR_TABLE_ENTRYP (val))
945 #endif /* not MULE */
951 return Qnil; /* not reached */
955 check_valid_char_table_value (Lisp_Object value, enum char_table_type type,
960 case CHAR_TABLE_TYPE_SYNTAX:
961 if (!ERRB_EQ (errb, ERROR_ME))
962 return INTP (value) || (CONSP (value) && INTP (XCAR (value))
963 && CHAR_OR_CHAR_INTP (XCDR (value)));
966 Lisp_Object cdr = XCDR (value);
967 CHECK_INT (XCAR (value));
968 CHECK_CHAR_COERCE_INT (cdr);
975 case CHAR_TABLE_TYPE_CATEGORY:
976 if (!ERRB_EQ (errb, ERROR_ME))
977 return CATEGORY_TABLE_VALUEP (value);
978 CHECK_CATEGORY_TABLE_VALUE (value);
982 case CHAR_TABLE_TYPE_GENERIC:
985 case CHAR_TABLE_TYPE_DISPLAY:
987 maybe_signal_simple_error ("Display char tables not yet implemented",
988 value, Qchar_table, errb);
991 case CHAR_TABLE_TYPE_CHAR:
992 if (!ERRB_EQ (errb, ERROR_ME))
993 return CHAR_OR_CHAR_INTP (value);
994 CHECK_CHAR_COERCE_INT (value);
1001 return 0; /* not reached */
1005 canonicalize_char_table_value (Lisp_Object value, enum char_table_type type)
1009 case CHAR_TABLE_TYPE_SYNTAX:
1012 Lisp_Object car = XCAR (value);
1013 Lisp_Object cdr = XCDR (value);
1014 CHECK_CHAR_COERCE_INT (cdr);
1015 return Fcons (car, cdr);
1018 case CHAR_TABLE_TYPE_CHAR:
1019 CHECK_CHAR_COERCE_INT (value);
1027 DEFUN ("valid-char-table-value-p", Fvalid_char_table_value_p, 2, 2, 0, /*
1028 Return non-nil if VALUE is a valid value for CHAR-TABLE-TYPE.
1030 (value, char_table_type))
1032 enum char_table_type type = symbol_to_char_table_type (char_table_type);
1034 return check_valid_char_table_value (value, type, ERROR_ME_NOT) ? Qt : Qnil;
1037 DEFUN ("check-valid-char-table-value", Fcheck_valid_char_table_value, 2, 2, 0, /*
1038 Signal an error if VALUE is not a valid value for CHAR-TABLE-TYPE.
1040 (value, char_table_type))
1042 enum char_table_type type = symbol_to_char_table_type (char_table_type);
1044 check_valid_char_table_value (value, type, ERROR_ME);
1048 /* Assign VAL to all characters in RANGE in char table CT. */
1051 put_char_table (Lisp_Char_Table *ct, struct chartab_range *range,
1054 switch (range->type)
1056 case CHARTAB_RANGE_ALL:
1057 fill_char_table (ct, val);
1058 return; /* avoid the duplicate call to update_syntax_table() below,
1059 since fill_char_table() also did that. */
1062 case CHARTAB_RANGE_CHARSET:
1063 if (EQ (range->charset, Vcharset_ascii))
1066 for (i = 0; i < 128; i++)
1069 else if (EQ (range->charset, Vcharset_control_1))
1072 for (i = 128; i < 160; i++)
1077 int lb = XCHARSET_LEADING_BYTE (range->charset) - MIN_LEADING_BYTE;
1078 ct->level1[lb] = val;
1082 case CHARTAB_RANGE_ROW:
1084 Lisp_Char_Table_Entry *cte;
1085 int lb = XCHARSET_LEADING_BYTE (range->charset) - MIN_LEADING_BYTE;
1086 /* make sure that there is a separate entry for the row. */
1087 if (!CHAR_TABLE_ENTRYP (ct->level1[lb]))
1088 ct->level1[lb] = make_char_table_entry (ct->level1[lb]);
1089 cte = XCHAR_TABLE_ENTRY (ct->level1[lb]);
1090 cte->level2[range->row - 32] = val;
1095 case CHARTAB_RANGE_CHAR:
1098 Lisp_Object charset;
1101 BREAKUP_CHAR (range->ch, charset, byte1, byte2);
1102 if (EQ (charset, Vcharset_ascii))
1103 ct->ascii[byte1] = val;
1104 else if (EQ (charset, Vcharset_control_1))
1105 ct->ascii[byte1 + 128] = val;
1108 Lisp_Char_Table_Entry *cte;
1109 int lb = XCHARSET_LEADING_BYTE (charset) - MIN_LEADING_BYTE;
1110 /* make sure that there is a separate entry for the row. */
1111 if (!CHAR_TABLE_ENTRYP (ct->level1[lb]))
1112 ct->level1[lb] = make_char_table_entry (ct->level1[lb]);
1113 cte = XCHAR_TABLE_ENTRY (ct->level1[lb]);
1114 /* now CTE is a char table entry for the charset;
1115 each entry is for a single row (or character of
1116 a one-octet charset). */
1117 if (XCHARSET_DIMENSION (charset) == 1)
1118 cte->level2[byte1 - 32] = val;
1121 /* assigning to one character in a two-octet charset. */
1122 /* make sure that the charset row contains a separate
1123 entry for each character. */
1124 if (!CHAR_TABLE_ENTRYP (cte->level2[byte1 - 32]))
1125 cte->level2[byte1 - 32] =
1126 make_char_table_entry (cte->level2[byte1 - 32]);
1127 cte = XCHAR_TABLE_ENTRY (cte->level2[byte1 - 32]);
1128 cte->level2[byte2 - 32] = val;
1132 #else /* not MULE */
1133 ct->ascii[(unsigned char) (range->ch)] = val;
1135 #endif /* not MULE */
1138 if (ct->type == CHAR_TABLE_TYPE_SYNTAX)
1139 update_syntax_table (ct);
1142 DEFUN ("put-char-table", Fput_char_table, 3, 3, 0, /*
1143 Set the value for chars in RANGE to be VALUE in CHAR-TABLE.
1145 RANGE specifies one or more characters to be affected and should be
1146 one of the following:
1148 -- t (all characters are affected)
1149 -- A charset (only allowed when Mule support is present)
1150 -- A vector of two elements: a two-octet charset and a row number; the row
1151 must be an integer, not a character (only allowed when Mule support is
1153 -- A single character
1155 VALUE must be a value appropriate for the type of CHAR-TABLE.
1156 See `valid-char-table-type-p'.
1158 (range, value, char_table))
1160 Lisp_Char_Table *ct;
1161 struct chartab_range rainj;
1163 CHECK_CHAR_TABLE (char_table);
1164 ct = XCHAR_TABLE (char_table);
1165 check_valid_char_table_value (value, ct->type, ERROR_ME);
1166 decode_char_table_range (range, &rainj);
1167 value = canonicalize_char_table_value (value, ct->type);
1168 put_char_table (ct, &rainj, value);
1172 /* Map FN over the ASCII chars in CT. */
1175 map_over_charset_ascii (Lisp_Char_Table *ct,
1176 int (*fn) (struct chartab_range *range,
1177 Lisp_Object val, void *arg),
1180 struct chartab_range rainj;
1189 rainj.type = CHARTAB_RANGE_CHAR;
1191 for (i = start, retval = 0; i < stop && retval == 0; i++)
1193 rainj.ch = (Emchar) i;
1194 retval = (fn) (&rainj, ct->ascii[i], arg);
1202 /* Map FN over the Control-1 chars in CT. */
1205 map_over_charset_control_1 (Lisp_Char_Table *ct,
1206 int (*fn) (struct chartab_range *range,
1207 Lisp_Object val, void *arg),
1210 struct chartab_range rainj;
1213 int stop = start + 32;
1215 rainj.type = CHARTAB_RANGE_CHAR;
1217 for (i = start, retval = 0; i < stop && retval == 0; i++)
1219 rainj.ch = (Emchar) (i);
1220 retval = (fn) (&rainj, ct->ascii[i], arg);
1226 /* Map FN over the row ROW of two-byte charset CHARSET.
1227 There must be a separate value for that row in the char table.
1228 CTE specifies the char table entry for CHARSET. */
1231 map_over_charset_row (Lisp_Char_Table_Entry *cte,
1232 Lisp_Object charset, int row,
1233 int (*fn) (struct chartab_range *range,
1234 Lisp_Object val, void *arg),
1237 Lisp_Object val = cte->level2[row - 32];
1239 if (!CHAR_TABLE_ENTRYP (val))
1241 struct chartab_range rainj;
1243 rainj.type = CHARTAB_RANGE_ROW;
1244 rainj.charset = charset;
1246 return (fn) (&rainj, val, arg);
1250 struct chartab_range rainj;
1252 int charset94_p = (XCHARSET_CHARS (charset) == 94);
1253 int start = charset94_p ? 33 : 32;
1254 int stop = charset94_p ? 127 : 128;
1256 cte = XCHAR_TABLE_ENTRY (val);
1258 rainj.type = CHARTAB_RANGE_CHAR;
1260 for (i = start, retval = 0; i < stop && retval == 0; i++)
1262 rainj.ch = MAKE_CHAR (charset, row, i);
1263 retval = (fn) (&rainj, cte->level2[i - 32], arg);
1271 map_over_other_charset (Lisp_Char_Table *ct, int lb,
1272 int (*fn) (struct chartab_range *range,
1273 Lisp_Object val, void *arg),
1276 Lisp_Object val = ct->level1[lb - MIN_LEADING_BYTE];
1277 Lisp_Object charset = CHARSET_BY_LEADING_BYTE (lb);
1279 if (!CHARSETP (charset)
1280 || lb == LEADING_BYTE_ASCII
1281 || lb == LEADING_BYTE_CONTROL_1)
1284 if (!CHAR_TABLE_ENTRYP (val))
1286 struct chartab_range rainj;
1288 rainj.type = CHARTAB_RANGE_CHARSET;
1289 rainj.charset = charset;
1290 return (fn) (&rainj, val, arg);
1294 Lisp_Char_Table_Entry *cte = XCHAR_TABLE_ENTRY (val);
1295 int charset94_p = (XCHARSET_CHARS (charset) == 94);
1296 int start = charset94_p ? 33 : 32;
1297 int stop = charset94_p ? 127 : 128;
1300 if (XCHARSET_DIMENSION (charset) == 1)
1302 struct chartab_range rainj;
1303 rainj.type = CHARTAB_RANGE_CHAR;
1305 for (i = start, retval = 0; i < stop && retval == 0; i++)
1307 rainj.ch = MAKE_CHAR (charset, i, 0);
1308 retval = (fn) (&rainj, cte->level2[i - 32], arg);
1313 for (i = start, retval = 0; i < stop && retval == 0; i++)
1314 retval = map_over_charset_row (cte, charset, i, fn, arg);
1323 /* Map FN (with client data ARG) over range RANGE in char table CT.
1324 Mapping stops the first time FN returns non-zero, and that value
1325 becomes the return value of map_char_table(). */
1328 map_char_table (Lisp_Char_Table *ct,
1329 struct chartab_range *range,
1330 int (*fn) (struct chartab_range *range,
1331 Lisp_Object val, void *arg),
1334 switch (range->type)
1336 case CHARTAB_RANGE_ALL:
1340 retval = map_over_charset_ascii (ct, fn, arg);
1344 retval = map_over_charset_control_1 (ct, fn, arg);
1349 int start = MIN_LEADING_BYTE;
1350 int stop = start + NUM_LEADING_BYTES;
1352 for (i = start, retval = 0; i < stop && retval == 0; i++)
1354 retval = map_over_other_charset (ct, i, fn, arg);
1362 case CHARTAB_RANGE_CHARSET:
1363 return map_over_other_charset (ct,
1364 XCHARSET_LEADING_BYTE (range->charset),
1367 case CHARTAB_RANGE_ROW:
1369 Lisp_Object val = ct->level1[XCHARSET_LEADING_BYTE (range->charset) - MIN_LEADING_BYTE];
1370 if (!CHAR_TABLE_ENTRYP (val))
1372 struct chartab_range rainj;
1374 rainj.type = CHARTAB_RANGE_ROW;
1375 rainj.charset = range->charset;
1376 rainj.row = range->row;
1377 return (fn) (&rainj, val, arg);
1380 return map_over_charset_row (XCHAR_TABLE_ENTRY (val),
1381 range->charset, range->row,
1386 case CHARTAB_RANGE_CHAR:
1388 Emchar ch = range->ch;
1389 Lisp_Object val = CHAR_TABLE_VALUE_UNSAFE (ct, ch);
1390 struct chartab_range rainj;
1392 rainj.type = CHARTAB_RANGE_CHAR;
1394 return (fn) (&rainj, val, arg);
1404 struct slow_map_char_table_arg
1406 Lisp_Object function;
1411 slow_map_char_table_fun (struct chartab_range *range,
1412 Lisp_Object val, void *arg)
1414 Lisp_Object ranjarg = Qnil;
1415 struct slow_map_char_table_arg *closure =
1416 (struct slow_map_char_table_arg *) arg;
1418 switch (range->type)
1420 case CHARTAB_RANGE_ALL:
1425 case CHARTAB_RANGE_CHARSET:
1426 ranjarg = XCHARSET_NAME (range->charset);
1429 case CHARTAB_RANGE_ROW:
1430 ranjarg = vector2 (XCHARSET_NAME (range->charset),
1431 make_int (range->row));
1434 case CHARTAB_RANGE_CHAR:
1435 ranjarg = make_char (range->ch);
1441 closure->retval = call2 (closure->function, ranjarg, val);
1442 return !NILP (closure->retval);
1445 DEFUN ("map-char-table", Fmap_char_table, 2, 3, 0, /*
1446 Map FUNCTION over CHAR-TABLE until it returns non-nil; return that value.
1447 FUNCTION is called with two arguments, each key and entry in the table.
1449 RANGE specifies a subrange to map over. If omitted or t, it defaults to
1452 Both RANGE and the keys passed to FUNCTION are in the same format as the
1453 RANGE argument to `put-char-table'. N.B. This function does NOT map over
1454 all characters in RANGE, but over the subranges that have been assigned to.
1455 Thus this function is most suitable for searching a char-table, or for
1456 populating one char-table based on the contents of another. The current
1457 implementation does not coalesce ranges all of whose values are the same.
1459 (function, char_table, range))
1461 Lisp_Char_Table *ct;
1462 struct slow_map_char_table_arg slarg;
1463 struct gcpro gcpro1, gcpro2;
1464 struct chartab_range rainj;
1466 CHECK_CHAR_TABLE (char_table);
1467 ct = XCHAR_TABLE (char_table);
1470 decode_char_table_range (range, &rainj);
1471 slarg.function = function;
1472 slarg.retval = Qnil;
1473 GCPRO2 (slarg.function, slarg.retval);
1474 map_char_table (ct, &rainj, slow_map_char_table_fun, &slarg);
1477 return slarg.retval;
1482 /************************************************************************/
1483 /* Char table read syntax */
1484 /************************************************************************/
1487 chartab_type_validate (Lisp_Object keyword, Lisp_Object value,
1488 Error_behavior errb)
1490 /* #### should deal with ERRB */
1491 symbol_to_char_table_type (value);
1496 chartab_data_validate (Lisp_Object keyword, Lisp_Object value,
1497 Error_behavior errb)
1501 /* #### should deal with ERRB */
1502 EXTERNAL_LIST_LOOP (rest, value)
1504 Lisp_Object range = XCAR (rest);
1505 struct chartab_range dummy;
1509 signal_simple_error ("Invalid list format", value);
1512 if (!CONSP (XCDR (range))
1513 || !NILP (XCDR (XCDR (range))))
1514 signal_simple_error ("Invalid range format", range);
1515 decode_char_table_range (XCAR (range), &dummy);
1516 decode_char_table_range (XCAR (XCDR (range)), &dummy);
1519 decode_char_table_range (range, &dummy);
1526 chartab_instantiate (Lisp_Object data)
1528 Lisp_Object chartab;
1529 Lisp_Object type = Qgeneric;
1530 Lisp_Object dataval = Qnil;
1532 while (!NILP (data))
1534 Lisp_Object keyw = Fcar (data);
1540 if (EQ (keyw, Qtype))
1542 else if (EQ (keyw, Qdata))
1546 chartab = Fmake_char_table (type);
1549 while (!NILP (data))
1551 Lisp_Object range = Fcar (data);
1552 Lisp_Object val = Fcar (Fcdr (data));
1554 data = Fcdr (Fcdr (data));
1557 if (CHAR_OR_CHAR_INTP (XCAR (range)))
1559 Emchar first = XCHAR_OR_CHAR_INT (Fcar (range));
1560 Emchar last = XCHAR_OR_CHAR_INT (Fcar (Fcdr (range)));
1563 for (i = first; i <= last; i++)
1564 Fput_char_table (make_char (i), val, chartab);
1570 Fput_char_table (range, val, chartab);
1579 /************************************************************************/
1580 /* Category Tables, specifically */
1581 /************************************************************************/
1583 DEFUN ("category-table-p", Fcategory_table_p, 1, 1, 0, /*
1584 Return t if OBJECT is a category table.
1585 A category table is a type of char table used for keeping track of
1586 categories. Categories are used for classifying characters for use
1587 in regexps -- you can refer to a category rather than having to use
1588 a complicated [] expression (and category lookups are significantly
1591 There are 95 different categories available, one for each printable
1592 character (including space) in the ASCII charset. Each category
1593 is designated by one such character, called a "category designator".
1594 They are specified in a regexp using the syntax "\\cX", where X is
1595 a category designator.
1597 A category table specifies, for each character, the categories that
1598 the character is in. Note that a character can be in more than one
1599 category. More specifically, a category table maps from a character
1600 to either the value nil (meaning the character is in no categories)
1601 or a 95-element bit vector, specifying for each of the 95 categories
1602 whether the character is in that category.
1604 Special Lisp functions are provided that abstract this, so you do not
1605 have to directly manipulate bit vectors.
1609 return (CHAR_TABLEP (object) &&
1610 XCHAR_TABLE_TYPE (object) == CHAR_TABLE_TYPE_CATEGORY) ?
1615 check_category_table (Lisp_Object object, Lisp_Object default_)
1619 while (NILP (Fcategory_table_p (object)))
1620 object = wrong_type_argument (Qcategory_table_p, object);
1625 check_category_char (Emchar ch, Lisp_Object table,
1626 unsigned int designator, unsigned int not_p)
1628 REGISTER Lisp_Object temp;
1629 Lisp_Char_Table *ctbl;
1630 #ifdef ERROR_CHECK_TYPECHECK
1631 if (NILP (Fcategory_table_p (table)))
1632 signal_simple_error ("Expected category table", table);
1634 ctbl = XCHAR_TABLE (table);
1635 temp = get_char_table (ch, ctbl);
1640 return bit_vector_bit (XBIT_VECTOR (temp), designator) ? !not_p : not_p;
1643 DEFUN ("check-category-at", Fcheck_category_at, 2, 4, 0, /*
1644 Return t if category of the character at POSITION includes DESIGNATOR.
1645 Optional third arg BUFFER specifies which buffer to use, and defaults
1646 to the current buffer.
1647 Optional fourth arg CATEGORY-TABLE specifies the category table to
1648 use, and defaults to BUFFER's category table.
1650 (position, designator, buffer, category_table))
1655 struct buffer *buf = decode_buffer (buffer, 0);
1657 CHECK_INT (position);
1658 CHECK_CATEGORY_DESIGNATOR (designator);
1659 des = XCHAR (designator);
1660 ctbl = check_category_table (category_table, Vstandard_category_table);
1661 ch = BUF_FETCH_CHAR (buf, XINT (position));
1662 return check_category_char (ch, ctbl, des, 0) ? Qt : Qnil;
1665 DEFUN ("char-in-category-p", Fchar_in_category_p, 2, 3, 0, /*
1666 Return t if category of CHARACTER includes DESIGNATOR, else nil.
1667 Optional third arg CATEGORY-TABLE specifies the category table to use,
1668 and defaults to the standard category table.
1670 (character, designator, category_table))
1676 CHECK_CATEGORY_DESIGNATOR (designator);
1677 des = XCHAR (designator);
1678 CHECK_CHAR (character);
1679 ch = XCHAR (character);
1680 ctbl = check_category_table (category_table, Vstandard_category_table);
1681 return check_category_char (ch, ctbl, des, 0) ? Qt : Qnil;
1684 DEFUN ("category-table", Fcategory_table, 0, 1, 0, /*
1685 Return BUFFER's current category table.
1686 BUFFER defaults to the current buffer.
1690 return decode_buffer (buffer, 0)->category_table;
1693 DEFUN ("standard-category-table", Fstandard_category_table, 0, 0, 0, /*
1694 Return the standard category table.
1695 This is the one used for new buffers.
1699 return Vstandard_category_table;
1702 DEFUN ("copy-category-table", Fcopy_category_table, 0, 1, 0, /*
1703 Return a new category table which is a copy of CATEGORY-TABLE.
1704 CATEGORY-TABLE defaults to the standard category table.
1708 if (NILP (Vstandard_category_table))
1709 return Fmake_char_table (Qcategory);
1712 check_category_table (category_table, Vstandard_category_table);
1713 return Fcopy_char_table (category_table);
1716 DEFUN ("set-category-table", Fset_category_table, 1, 2, 0, /*
1717 Select CATEGORY-TABLE as the new category table for BUFFER.
1718 BUFFER defaults to the current buffer if omitted.
1720 (category_table, buffer))
1722 struct buffer *buf = decode_buffer (buffer, 0);
1723 category_table = check_category_table (category_table, Qnil);
1724 buf->category_table = category_table;
1725 /* Indicate that this buffer now has a specified category table. */
1726 buf->local_var_flags |= XINT (buffer_local_flags.category_table);
1727 return category_table;
1730 DEFUN ("category-designator-p", Fcategory_designator_p, 1, 1, 0, /*
1731 Return t if OBJECT is a category designator (a char in the range ' ' to '~').
1735 return CATEGORY_DESIGNATORP (object) ? Qt : Qnil;
1738 DEFUN ("category-table-value-p", Fcategory_table_value_p, 1, 1, 0, /*
1739 Return t if OBJECT is a category table value.
1740 Valid values are nil or a bit vector of size 95.
1744 return CATEGORY_TABLE_VALUEP (object) ? Qt : Qnil;
1748 #define CATEGORYP(x) \
1749 (CHARP (x) && XCHAR (x) >= 0x20 && XCHAR (x) <= 0x7E)
1751 #define CATEGORY_SET(c) \
1752 (get_char_table(c, XCHAR_TABLE(current_buffer->category_table)))
1754 /* Return 1 if CATEGORY_SET contains CATEGORY, else return 0.
1755 The faster version of `!NILP (Faref (category_set, category))'. */
1756 #define CATEGORY_MEMBER(category, category_set) \
1757 (bit_vector_bit(XBIT_VECTOR (category_set), category - 32))
1759 /* Return 1 if there is a word boundary between two word-constituent
1760 characters C1 and C2 if they appear in this order, else return 0.
1761 Use the macro WORD_BOUNDARY_P instead of calling this function
1764 int word_boundary_p (Emchar c1, Emchar c2);
1766 word_boundary_p (Emchar c1, Emchar c2)
1768 Lisp_Object category_set1, category_set2;
1773 if (COMPOSITE_CHAR_P (c1))
1774 c1 = cmpchar_component (c1, 0, 1);
1775 if (COMPOSITE_CHAR_P (c2))
1776 c2 = cmpchar_component (c2, 0, 1);
1779 if (EQ (CHAR_CHARSET (c1), CHAR_CHARSET (c2)))
1781 tail = Vword_separating_categories;
1786 tail = Vword_combining_categories;
1790 category_set1 = CATEGORY_SET (c1);
1791 if (NILP (category_set1))
1792 return default_result;
1793 category_set2 = CATEGORY_SET (c2);
1794 if (NILP (category_set2))
1795 return default_result;
1797 for (; CONSP (tail); tail = XCONS (tail)->cdr)
1799 Lisp_Object elt = XCONS(tail)->car;
1802 && CATEGORYP (XCONS (elt)->car)
1803 && CATEGORYP (XCONS (elt)->cdr)
1804 && CATEGORY_MEMBER (XCHAR (XCONS (elt)->car), category_set1)
1805 && CATEGORY_MEMBER (XCHAR (XCONS (elt)->cdr), category_set2))
1806 return !default_result;
1808 return default_result;
1814 syms_of_chartab (void)
1816 INIT_LRECORD_IMPLEMENTATION (char_table);
1819 INIT_LRECORD_IMPLEMENTATION (char_table_entry);
1821 defsymbol (&Qcategory_table_p, "category-table-p");
1822 defsymbol (&Qcategory_designator_p, "category-designator-p");
1823 defsymbol (&Qcategory_table_value_p, "category-table-value-p");
1826 defsymbol (&Qchar_table, "char-table");
1827 defsymbol (&Qchar_tablep, "char-table-p");
1829 DEFSUBR (Fchar_table_p);
1830 DEFSUBR (Fchar_table_type_list);
1831 DEFSUBR (Fvalid_char_table_type_p);
1832 DEFSUBR (Fchar_table_type);
1833 DEFSUBR (Freset_char_table);
1834 DEFSUBR (Fmake_char_table);
1835 DEFSUBR (Fcopy_char_table);
1836 DEFSUBR (Fget_char_table);
1837 DEFSUBR (Fget_range_char_table);
1838 DEFSUBR (Fvalid_char_table_value_p);
1839 DEFSUBR (Fcheck_valid_char_table_value);
1840 DEFSUBR (Fput_char_table);
1841 DEFSUBR (Fmap_char_table);
1844 DEFSUBR (Fcategory_table_p);
1845 DEFSUBR (Fcategory_table);
1846 DEFSUBR (Fstandard_category_table);
1847 DEFSUBR (Fcopy_category_table);
1848 DEFSUBR (Fset_category_table);
1849 DEFSUBR (Fcheck_category_at);
1850 DEFSUBR (Fchar_in_category_p);
1851 DEFSUBR (Fcategory_designator_p);
1852 DEFSUBR (Fcategory_table_value_p);
1858 vars_of_chartab (void)
1860 /* DO NOT staticpro this. It works just like Vweak_hash_tables. */
1861 Vall_syntax_tables = Qnil;
1862 dump_add_weak_object_chain (&Vall_syntax_tables);
1866 structure_type_create_chartab (void)
1868 struct structure_type *st;
1870 st = define_structure_type (Qchar_table, 0, chartab_instantiate);
1872 define_structure_type_keyword (st, Qtype, chartab_type_validate);
1873 define_structure_type_keyword (st, Qdata, chartab_data_validate);
1877 complex_vars_of_chartab (void)
1880 /* Set this now, so first buffer creation can refer to it. */
1881 /* Make it nil before calling copy-category-table
1882 so that copy-category-table will know not to try to copy from garbage */
1883 Vstandard_category_table = Qnil;
1884 Vstandard_category_table = Fcopy_category_table (Qnil);
1885 staticpro (&Vstandard_category_table);
1887 DEFVAR_LISP ("word-combining-categories", &Vword_combining_categories /*
1888 List of pair (cons) of categories to determine word boundary.
1890 Emacs treats a sequence of word constituent characters as a single
1891 word (i.e. finds no word boundary between them) iff they belongs to
1892 the same charset. But, exceptions are allowed in the following cases.
1894 \(1) The case that characters are in different charsets is controlled
1895 by the variable `word-combining-categories'.
1897 Emacs finds no word boundary between characters of different charsets
1898 if they have categories matching some element of this list.
1900 More precisely, if an element of this list is a cons of category CAT1
1901 and CAT2, and a multibyte character C1 which has CAT1 is followed by
1902 C2 which has CAT2, there's no word boundary between C1 and C2.
1904 For instance, to tell that ASCII characters and Latin-1 characters can
1905 form a single word, the element `(?l . ?l)' should be in this list
1906 because both characters have the category `l' (Latin characters).
1908 \(2) The case that character are in the same charset is controlled by
1909 the variable `word-separating-categories'.
1911 Emacs find a word boundary between characters of the same charset
1912 if they have categories matching some element of this list.
1914 More precisely, if an element of this list is a cons of category CAT1
1915 and CAT2, and a multibyte character C1 which has CAT1 is followed by
1916 C2 which has CAT2, there's a word boundary between C1 and C2.
1918 For instance, to tell that there's a word boundary between Japanese
1919 Hiragana and Japanese Kanji (both are in the same charset), the
1920 element `(?H . ?C) should be in this list.
1923 Vword_combining_categories = Qnil;
1925 DEFVAR_LISP ("word-separating-categories", &Vword_separating_categories /*
1926 List of pair (cons) of categories to determine word boundary.
1927 See the documentation of the variable `word-combining-categories'.
1930 Vword_separating_categories = Qnil;