1 /* SXEmacs 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 SXEmacs
10 SXEmacs is free software: you can redistribute it and/or modify
11 it under the terms of the GNU General Public License as published by
12 the Free Software Foundation, either version 3 of the License, or
13 (at your option) any later version.
15 SXEmacs is distributed in the hope that it will be useful,
16 but WITHOUT ANY WARRANTY; without even the implied warranty of
17 MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
18 GNU General Public License for more details.
20 You should have received a copy of the GNU General Public License
21 along with this program. If not, see <http://www.gnu.org/licenses/>. */
24 /* Synched up with: Mule 2.3. Not synched with FSF.
26 This file was written independently of the FSF implementation,
27 and is not compatible. */
31 Ben Wing: wrote, for 19.13 (Mule). Some category table stuff
32 loosely based on the original Mule.
33 Jareth Hein: fixed a couple of bugs in the implementation, and
34 added regex support for categories with check_category_at
44 Lisp_Object Qchar_tablep, Qchar_table;
46 Lisp_Object Vall_syntax_tables;
49 Lisp_Object Qcategory_table_p;
50 Lisp_Object Qcategory_designator_p;
51 Lisp_Object Qcategory_table_value_p;
53 Lisp_Object Vstandard_category_table;
55 /* Variables to determine word boundary. */
56 Lisp_Object Vword_combining_categories, Vword_separating_categories;
59 /* A char table maps from ranges of characters to values.
61 Implementing a general data structure that maps from arbitrary
62 ranges of numbers to values is tricky to do efficiently. As it
63 happens, it should suffice (and is usually more convenient, anyway)
64 when dealing with characters to restrict the sorts of ranges that
65 can be assigned values, as follows:
68 2) All characters in a charset.
69 3) All characters in a particular row of a charset, where a "row"
70 means all characters with the same first byte.
71 4) A particular character in a charset.
73 We use char tables to generalize the 256-element vectors now
74 littering the Emacs code.
76 Possible uses (all should be converted at some point):
82 5) keyboard-translate-table?
85 abstract type to generalize the Emacs vectors and Mule
86 vectors-of-vectors goo.
89 /************************************************************************/
90 /* Char Table object */
91 /************************************************************************/
95 static Lisp_Object mark_char_table_entry(Lisp_Object obj)
97 Lisp_Char_Table_Entry *cte = XCHAR_TABLE_ENTRY(obj);
100 for (i = 0; i < 96; i++) {
101 mark_object(cte->level2[i]);
106 static int char_table_entry_equal(Lisp_Object obj1, Lisp_Object obj2, int depth)
108 Lisp_Char_Table_Entry *cte1 = XCHAR_TABLE_ENTRY(obj1);
109 Lisp_Char_Table_Entry *cte2 = XCHAR_TABLE_ENTRY(obj2);
112 for (i = 0; i < 96; i++)
114 (cte1->level2[i], cte2->level2[i], depth + 1))
120 static unsigned long char_table_entry_hash(Lisp_Object obj, int depth)
122 Lisp_Char_Table_Entry *cte = XCHAR_TABLE_ENTRY(obj);
124 return internal_array_hash(cte->level2, 96, depth);
127 static const struct lrecord_description char_table_entry_description[] = {
128 {XD_LISP_OBJECT_ARRAY, offsetof(Lisp_Char_Table_Entry, level2), 96},
132 DEFINE_LRECORD_IMPLEMENTATION("char-table-entry", char_table_entry,
133 mark_char_table_entry, internal_object_printer,
134 0, char_table_entry_equal,
135 char_table_entry_hash,
136 char_table_entry_description,
137 Lisp_Char_Table_Entry);
140 static Lisp_Object mark_char_table(Lisp_Object obj)
142 Lisp_Char_Table *ct = XCHAR_TABLE(obj);
145 for (i = 0; i < NUM_ASCII_CHARS; i++)
146 mark_object(ct->ascii[i]);
148 for (i = 0; i < NUM_LEADING_BYTES; i++)
149 mark_object(ct->level1[i]);
151 return ct->mirror_table;
154 /* WARNING: All functions of this nature need to be written extremely
155 carefully to avoid crashes during GC. Cf. prune_specifiers()
156 and prune_weak_hash_tables(). */
158 void prune_syntax_tables(void)
160 Lisp_Object rest, prev = Qnil;
162 for (rest = Vall_syntax_tables;
163 !NILP(rest); rest = XCHAR_TABLE(rest)->next_table) {
164 if (!marked_p(rest)) {
165 /* This table is garbage. Remove it from the list. */
168 XCHAR_TABLE(rest)->next_table;
170 XCHAR_TABLE(prev)->next_table =
171 XCHAR_TABLE(rest)->next_table;
176 static Lisp_Object char_table_type_to_symbol(enum char_table_type type)
181 case CHAR_TABLE_TYPE_GENERIC:
183 case CHAR_TABLE_TYPE_SYNTAX:
185 case CHAR_TABLE_TYPE_DISPLAY:
187 case CHAR_TABLE_TYPE_CHAR:
190 case CHAR_TABLE_TYPE_CATEGORY:
196 static enum char_table_type symbol_to_char_table_type(Lisp_Object symbol)
198 CHECK_SYMBOL(symbol);
200 if (EQ(symbol, Qgeneric))
201 return CHAR_TABLE_TYPE_GENERIC;
202 if (EQ(symbol, Qsyntax))
203 return CHAR_TABLE_TYPE_SYNTAX;
204 if (EQ(symbol, Qdisplay))
205 return CHAR_TABLE_TYPE_DISPLAY;
206 if (EQ(symbol, Qchar))
207 return CHAR_TABLE_TYPE_CHAR;
209 if (EQ(symbol, Qcategory))
210 return CHAR_TABLE_TYPE_CATEGORY;
213 signal_simple_error("Unrecognized char table type", symbol);
214 return CHAR_TABLE_TYPE_GENERIC; /* not reached */
218 print_chartab_range(Emchar first, Emchar last, Lisp_Object val,
219 Lisp_Object printcharfun)
222 write_c_string(" (", printcharfun);
223 print_internal(make_char(first), printcharfun, 0);
224 write_c_string(" ", printcharfun);
225 print_internal(make_char(last), printcharfun, 0);
226 write_c_string(") ", printcharfun);
228 write_c_string(" ", printcharfun);
229 print_internal(make_char(first), printcharfun, 0);
230 write_c_string(" ", printcharfun);
232 print_internal(val, printcharfun, 1);
238 print_chartab_charset_row(Lisp_Object charset,
240 Lisp_Char_Table_Entry * cte, Lisp_Object printcharfun)
243 Lisp_Object cat = Qunbound;
246 for (i = 32; i < 128; i++) {
247 Lisp_Object pam = cte->level2[i - 32];
257 print_chartab_range(MAKE_CHAR
259 MAKE_CHAR(charset, i - 1,
263 print_chartab_range(MAKE_CHAR
264 (charset, row, first),
265 MAKE_CHAR(charset, row,
275 print_chartab_range(MAKE_CHAR(charset, first, 0),
276 MAKE_CHAR(charset, i - 1, 0),
279 print_chartab_range(MAKE_CHAR(charset, row, first),
280 MAKE_CHAR(charset, row, i - 1),
286 print_chartab_two_byte_charset(Lisp_Object charset,
287 Lisp_Char_Table_Entry * cte,
288 Lisp_Object printcharfun)
292 for (i = 32; i < 128; i++) {
293 Lisp_Object jen = cte->level2[i - 32];
295 if (!CHAR_TABLE_ENTRYP(jen)) {
296 write_c_string(" [", printcharfun);
297 print_internal(XCHARSET_NAME(charset), printcharfun, 0);
298 write_fmt_str(printcharfun, " %d] ", i);
299 print_internal(jen, printcharfun, 0);
301 print_chartab_charset_row(charset, i,
302 XCHAR_TABLE_ENTRY(jen),
310 print_char_table(Lisp_Object obj, Lisp_Object printcharfun, int escapeflag)
312 Lisp_Char_Table *ct = XCHAR_TABLE(obj);
313 Lisp_Object tmp_sym = char_table_type_to_symbol(ct->type);
315 write_fmt_string(printcharfun, "#s(char-table type %s data (",
316 string_data(symbol_name(XSYMBOL(tmp_sym))));
318 /* Now write out the ASCII/Control-1 stuff. */
322 Lisp_Object val = Qunbound;
324 for (i = 0; i < NUM_ASCII_CHARS; i++) {
331 if (!EQ(ct->ascii[i], val)) {
332 print_chartab_range(first, i - 1, val,
340 print_chartab_range(first, i - 1, val, printcharfun);
347 for (i = MIN_LEADING_BYTE;
348 i < MIN_LEADING_BYTE + NUM_LEADING_BYTES; i++) {
349 Lisp_Object ann = ct->level1[i - MIN_LEADING_BYTE];
350 Lisp_Object charset = CHARSET_BY_LEADING_BYTE(i);
352 if (!CHARSETP(charset) || i == LEADING_BYTE_ASCII
353 || i == LEADING_BYTE_CONTROL_1)
355 if (!CHAR_TABLE_ENTRYP(ann)) {
356 write_c_string(" ", printcharfun);
357 print_internal(XCHARSET_NAME(charset),
359 write_c_string(" ", printcharfun);
360 print_internal(ann, printcharfun, 0);
362 Lisp_Char_Table_Entry *cte =
363 XCHAR_TABLE_ENTRY(ann);
364 if (XCHARSET_DIMENSION(charset) == 1)
365 print_chartab_charset_row(charset, -1,
369 print_chartab_two_byte_charset(charset,
377 write_c_string("))", printcharfun);
380 static int char_table_equal(Lisp_Object obj1, Lisp_Object obj2, int depth)
382 Lisp_Char_Table *ct1 = XCHAR_TABLE(obj1);
383 Lisp_Char_Table *ct2 = XCHAR_TABLE(obj2);
386 if (CHAR_TABLE_TYPE(ct1) != CHAR_TABLE_TYPE(ct2))
389 for (i = 0; i < NUM_ASCII_CHARS; i++)
390 if (!internal_equal(ct1->ascii[i], ct2->ascii[i], depth + 1))
394 for (i = 0; i < NUM_LEADING_BYTES; i++)
395 if (!internal_equal(ct1->level1[i], ct2->level1[i], depth + 1))
402 static unsigned long char_table_hash(Lisp_Object obj, int depth)
404 Lisp_Char_Table *ct = XCHAR_TABLE(obj);
405 unsigned long hashval = internal_array_hash(ct->ascii, NUM_ASCII_CHARS,
408 hashval = HASH2(hashval,
409 internal_array_hash(ct->level1, NUM_LEADING_BYTES,
415 static const struct lrecord_description char_table_description[] = {
416 {XD_LISP_OBJECT_ARRAY, offsetof(Lisp_Char_Table, ascii),
419 {XD_LISP_OBJECT_ARRAY, offsetof(Lisp_Char_Table, level1),
422 {XD_LISP_OBJECT, offsetof(Lisp_Char_Table, mirror_table)},
423 {XD_LO_LINK, offsetof(Lisp_Char_Table, next_table)},
427 DEFINE_LRECORD_IMPLEMENTATION("char-table", char_table,
428 mark_char_table, print_char_table, 0,
429 char_table_equal, char_table_hash,
430 char_table_description, Lisp_Char_Table);
432 DEFUN("char-table-p", Fchar_table_p, 1, 1, 0, /*
433 Return non-nil if OBJECT is a char table.
435 A char table is a table that maps characters (or ranges of characters)
436 to values. Char tables are specialized for characters, only allowing
437 particular sorts of ranges to be assigned values. Although this
438 loses in generality, it makes for extremely fast (constant-time)
439 lookups, and thus is feasible for applications that do an extremely
440 large number of lookups (e.g. scanning a buffer for a character in
441 a particular syntax, where a lookup in the syntax table must occur
444 When Mule support exists, the types of ranges that can be assigned
449 -- a single row in a two-octet charset
450 -- a single character
452 When Mule support is not present, the types of ranges that can be
456 -- a single character
458 To create a char table, use `make-char-table'.
459 To modify a char table, use `put-char-table' or `remove-char-table'.
460 To retrieve the value for a particular character, use `get-char-table'.
461 See also `map-char-table', `clear-char-table', `copy-char-table',
462 `valid-char-table-type-p', `char-table-type-list',
463 `valid-char-table-value-p', and `check-char-table-value'.
467 return CHAR_TABLEP(object) ? Qt : Qnil;
470 DEFUN("char-table-type-list", Fchar_table_type_list, 0, 0, 0, /*
471 Return a list of the recognized char table types.
472 See `valid-char-table-type-p'.
477 return list5(Qchar, Qcategory, Qdisplay, Qgeneric, Qsyntax);
479 return list4(Qchar, Qdisplay, Qgeneric, Qsyntax);
483 DEFUN("valid-char-table-type-p", Fvalid_char_table_type_p, 1, 1, 0, /*
484 Return t if TYPE if a recognized char table type.
486 Each char table type is used for a different purpose and allows different
487 sorts of values. The different char table types are
490 Used for category tables, which specify the regexp categories
491 that a character is in. The valid values are nil or a
492 bit vector of 95 elements. Higher-level Lisp functions are
493 provided for working with category tables. Currently categories
494 and category tables only exist when Mule support is present.
496 A generalized char table, for mapping from one character to
497 another. Used for case tables, syntax matching tables,
498 `keyboard-translate-table', etc. The valid values are characters.
500 An even more generalized char table, for mapping from a
501 character to anything.
503 Used for display tables, which specify how a particular character
504 is to appear when displayed. #### Not yet implemented.
506 Used for syntax tables, which specify the syntax of a particular
507 character. Higher-level Lisp functions are provided for
508 working with syntax tables. The valid values are integers.
513 return (EQ(type, Qchar) ||
515 EQ(type, Qcategory) ||
517 EQ(type, Qdisplay) ||
518 EQ(type, Qgeneric) || EQ(type, Qsyntax)) ? Qt : Qnil;
521 DEFUN("char-table-type", Fchar_table_type, 1, 1, 0, /*
522 Return the type of CHAR-TABLE.
523 See `valid-char-table-type-p'.
527 CHECK_CHAR_TABLE(char_table);
528 return char_table_type_to_symbol(XCHAR_TABLE(char_table)->type);
531 void fill_char_table(Lisp_Char_Table * ct, Lisp_Object value)
535 for (i = 0; i < NUM_ASCII_CHARS; i++)
536 ct->ascii[i] = value;
538 for (i = 0; i < NUM_LEADING_BYTES; i++)
539 ct->level1[i] = value;
542 if (ct->type == CHAR_TABLE_TYPE_SYNTAX)
543 update_syntax_table(ct);
546 DEFUN("reset-char-table", Freset_char_table, 1, 1, 0, /*
547 Reset CHAR-TABLE to its default state.
553 CHECK_CHAR_TABLE(char_table);
554 ct = XCHAR_TABLE(char_table);
557 case CHAR_TABLE_TYPE_CHAR:
558 fill_char_table(ct, make_char(0));
560 case CHAR_TABLE_TYPE_DISPLAY:
561 case CHAR_TABLE_TYPE_GENERIC:
563 case CHAR_TABLE_TYPE_CATEGORY:
565 fill_char_table(ct, Qnil);
568 case CHAR_TABLE_TYPE_SYNTAX:
569 fill_char_table(ct, make_int(Sinherit));
579 DEFUN("make-char-table", Fmake_char_table, 1, 1, 0, /*
580 Return a new, empty char table of type TYPE.
581 Currently recognized types are 'char, 'category, 'display, 'generic,
582 and 'syntax. See `valid-char-table-type-p'.
588 enum char_table_type ty = symbol_to_char_table_type(type);
590 ct = alloc_lcrecord_type(Lisp_Char_Table, &lrecord_char_table);
592 if (ty == CHAR_TABLE_TYPE_SYNTAX) {
593 ct->mirror_table = Fmake_char_table(Qgeneric);
594 fill_char_table(XCHAR_TABLE(ct->mirror_table),
597 ct->mirror_table = Qnil;
598 ct->next_table = Qnil;
599 XSETCHAR_TABLE(obj, ct);
600 if (ty == CHAR_TABLE_TYPE_SYNTAX) {
601 ct->next_table = Vall_syntax_tables;
602 Vall_syntax_tables = obj;
604 Freset_char_table(obj);
610 static Lisp_Object make_char_table_entry(Lisp_Object initval)
614 Lisp_Char_Table_Entry *cte =
615 alloc_lcrecord_type(Lisp_Char_Table_Entry,
616 &lrecord_char_table_entry);
618 for (i = 0; i < 96; i++)
619 cte->level2[i] = initval;
621 XSETCHAR_TABLE_ENTRY(obj, cte);
625 static Lisp_Object copy_char_table_entry(Lisp_Object entry)
627 Lisp_Char_Table_Entry *cte = XCHAR_TABLE_ENTRY(entry);
630 Lisp_Char_Table_Entry *ctenew =
631 alloc_lcrecord_type(Lisp_Char_Table_Entry,
632 &lrecord_char_table_entry);
634 for (i = 0; i < 96; i++) {
635 Lisp_Object new = cte->level2[i];
636 if (CHAR_TABLE_ENTRYP(new))
637 ctenew->level2[i] = copy_char_table_entry(new);
639 ctenew->level2[i] = new;
642 XSETCHAR_TABLE_ENTRY(obj, ctenew);
648 DEFUN("copy-char-table", Fcopy_char_table, 1, 1, 0, /*
649 Return a new char table which is a copy of CHAR-TABLE.
650 It will contain the same values for the same characters and ranges
651 as CHAR-TABLE. The values will not themselves be copied.
655 Lisp_Char_Table *ct, *ctnew;
659 CHECK_CHAR_TABLE(char_table);
660 ct = XCHAR_TABLE(char_table);
661 ctnew = alloc_lcrecord_type(Lisp_Char_Table, &lrecord_char_table);
662 ctnew->type = ct->type;
664 for (i = 0; i < NUM_ASCII_CHARS; i++) {
665 Lisp_Object new = ct->ascii[i];
667 assert(!(CHAR_TABLE_ENTRYP(new)));
669 ctnew->ascii[i] = new;
674 for (i = 0; i < NUM_LEADING_BYTES; i++) {
675 Lisp_Object new = ct->level1[i];
676 if (CHAR_TABLE_ENTRYP(new))
677 ctnew->level1[i] = copy_char_table_entry(new);
679 ctnew->level1[i] = new;
684 if (CHAR_TABLEP(ct->mirror_table))
685 ctnew->mirror_table = Fcopy_char_table(ct->mirror_table);
687 ctnew->mirror_table = ct->mirror_table;
688 ctnew->next_table = Qnil;
689 XSETCHAR_TABLE(obj, ctnew);
690 if (ctnew->type == CHAR_TABLE_TYPE_SYNTAX) {
691 ctnew->next_table = Vall_syntax_tables;
692 Vall_syntax_tables = obj;
698 decode_char_table_range(Lisp_Object range, struct chartab_range *outrange)
701 outrange->type = CHARTAB_RANGE_ALL;
702 else if (CHAR_OR_CHAR_INTP(range)) {
703 outrange->type = CHARTAB_RANGE_CHAR;
704 outrange->ch = XCHAR_OR_CHAR_INT(range);
708 signal_simple_error("Range must be t or a character", range);
710 else if (VECTORP(range)) {
711 Lisp_Vector *vec = XVECTOR(range);
712 Lisp_Object *elts = vector_data(vec);
713 if (vector_length(vec) != 2)
715 ("Length of charset row vector must be 2", range);
716 outrange->type = CHARTAB_RANGE_ROW;
717 outrange->charset = Fget_charset(elts[0]);
719 outrange->row = XINT(elts[1]);
720 switch (XCHARSET_TYPE(outrange->charset)) {
721 case CHARSET_TYPE_94:
722 case CHARSET_TYPE_96:
724 ("Charset in row vector must be multi-byte",
726 case CHARSET_TYPE_94X94:
727 check_int_range(outrange->row, 33, 126);
729 case CHARSET_TYPE_96X96:
730 check_int_range(outrange->row, 32, 127);
736 if (!CHARSETP(range) && !SYMBOLP(range))
738 ("Char table range must be t, charset, char, or vector",
740 outrange->type = CHARTAB_RANGE_CHARSET;
741 outrange->charset = Fget_charset(range);
748 /* called from CHAR_TABLE_VALUE(). */
750 get_non_ascii_char_table_value(Lisp_Char_Table * ct, int leading_byte, Emchar c)
753 Lisp_Object charset = CHARSET_BY_LEADING_BYTE(leading_byte);
756 BREAKUP_CHAR_1_UNSAFE(c, charset, byte1, byte2);
757 val = ct->level1[leading_byte - MIN_LEADING_BYTE];
758 if (CHAR_TABLE_ENTRYP(val)) {
759 Lisp_Char_Table_Entry *cte = XCHAR_TABLE_ENTRY(val);
760 val = cte->level2[byte1 - 32];
761 if (CHAR_TABLE_ENTRYP(val)) {
762 cte = XCHAR_TABLE_ENTRY(val);
764 val = cte->level2[byte2 - 32];
765 assert(!CHAR_TABLE_ENTRYP(val));
774 Lisp_Object get_char_table(Emchar ch, Lisp_Char_Table * ct)
782 BREAKUP_CHAR(ch, charset, byte1, byte2);
784 if (EQ(charset, Vcharset_ascii))
785 val = ct->ascii[byte1];
786 else if (EQ(charset, Vcharset_control_1))
787 val = ct->ascii[byte1 + 128];
790 XCHARSET_LEADING_BYTE(charset) - MIN_LEADING_BYTE;
791 val = ct->level1[lb];
792 if (CHAR_TABLE_ENTRYP(val)) {
793 Lisp_Char_Table_Entry *cte =
794 XCHAR_TABLE_ENTRY(val);
795 val = cte->level2[byte1 - 32];
796 if (CHAR_TABLE_ENTRYP(val)) {
797 cte = XCHAR_TABLE_ENTRY(val);
799 val = cte->level2[byte2 - 32];
800 assert(!CHAR_TABLE_ENTRYP(val));
808 return ct->ascii[(unsigned char)ch];
809 #endif /* not MULE */
812 DEFUN("get-char-table", Fget_char_table, 2, 2, 0, /*
813 Find value for CHARACTER in CHAR-TABLE.
815 (character, char_table))
817 CHECK_CHAR_TABLE(char_table);
818 CHECK_CHAR_COERCE_INT(character);
820 return get_char_table(XCHAR(character), XCHAR_TABLE(char_table));
823 DEFUN("get-range-char-table", Fget_range_char_table, 2, 3, 0, /*
824 Find value for a range in CHAR-TABLE.
825 If there is more than one value, return MULTI (defaults to nil).
827 (range, char_table, multi))
830 struct chartab_range rainj;
832 if (CHAR_OR_CHAR_INTP(range))
833 return Fget_char_table(range, char_table);
834 CHECK_CHAR_TABLE(char_table);
835 ct = XCHAR_TABLE(char_table);
837 decode_char_table_range(range, &rainj);
838 switch (rainj.type) {
839 case CHARTAB_RANGE_ALL:
842 Lisp_Object first = ct->ascii[0];
844 for (i = 1; i < NUM_ASCII_CHARS; i++)
845 if (!EQ(first, ct->ascii[i]))
849 for (i = MIN_LEADING_BYTE;
850 i < MIN_LEADING_BYTE + NUM_LEADING_BYTES; i++) {
851 Lisp_Object foo = CHARSET_BY_LEADING_BYTE(i);
853 || i == LEADING_BYTE_ASCII
854 || i == LEADING_BYTE_CONTROL_1) {
858 (first, ct->level1[i - MIN_LEADING_BYTE])) {
868 case CHARTAB_RANGE_CHARSET:
869 if (EQ(rainj.charset, Vcharset_ascii)) {
871 Lisp_Object first = ct->ascii[0];
873 for (i = 1; i < 128; i++)
874 if (!EQ(first, ct->ascii[i]))
879 if (EQ(rainj.charset, Vcharset_control_1)) {
881 Lisp_Object first = ct->ascii[128];
883 for (i = 129; i < 160; i++)
884 if (!EQ(first, ct->ascii[i]))
891 ct->level1[XCHARSET_LEADING_BYTE(rainj.charset) -
893 if (CHAR_TABLE_ENTRYP(val))
898 case CHARTAB_RANGE_ROW:
901 ct->level1[XCHARSET_LEADING_BYTE(rainj.charset) -
903 if (!CHAR_TABLE_ENTRYP(val))
905 val = XCHAR_TABLE_ENTRY(val)->level2[rainj.row - 32];
906 if (CHAR_TABLE_ENTRYP(val))
912 case CHARTAB_RANGE_CHAR:
913 #endif /* not MULE */
919 return Qnil; /* not reached */
923 check_valid_char_table_value(Lisp_Object value, enum char_table_type type,
927 case CHAR_TABLE_TYPE_SYNTAX:
928 if (!ERRB_EQ(errb, ERROR_ME))
929 return INTP(value) || (CONSP(value) && INTP(XCAR(value))
931 CHAR_OR_CHAR_INTP(XCDR(value)));
933 Lisp_Object cdr = XCDR(value);
934 CHECK_INT(XCAR(value));
935 CHECK_CHAR_COERCE_INT(cdr);
941 case CHAR_TABLE_TYPE_CATEGORY:
942 if (!ERRB_EQ(errb, ERROR_ME))
943 return CATEGORY_TABLE_VALUEP(value);
944 CHECK_CATEGORY_TABLE_VALUE(value);
948 case CHAR_TABLE_TYPE_GENERIC:
951 case CHAR_TABLE_TYPE_DISPLAY:
953 maybe_signal_simple_error
954 ("Display char tables not yet implemented", value,
958 case CHAR_TABLE_TYPE_CHAR:
959 if (!ERRB_EQ(errb, ERROR_ME))
960 return CHAR_OR_CHAR_INTP(value);
961 CHECK_CHAR_COERCE_INT(value);
968 return 0; /* not reached */
972 canonicalize_char_table_value(Lisp_Object value, enum char_table_type type)
975 case CHAR_TABLE_TYPE_SYNTAX:
977 Lisp_Object car = XCAR(value);
978 Lisp_Object cdr = XCDR(value);
979 CHECK_CHAR_COERCE_INT(cdr);
980 return Fcons(car, cdr);
983 case CHAR_TABLE_TYPE_CHAR:
984 CHECK_CHAR_COERCE_INT(value);
987 case CHAR_TABLE_TYPE_GENERIC:
989 case CHAR_TABLE_TYPE_CATEGORY:
991 case CHAR_TABLE_TYPE_DISPLAY:
998 DEFUN("valid-char-table-value-p", Fvalid_char_table_value_p, 2, 2, 0, /*
999 Return non-nil if VALUE is a valid value for CHAR-TABLE-TYPE.
1001 (value, char_table_type))
1003 enum char_table_type type = symbol_to_char_table_type(char_table_type);
1005 return check_valid_char_table_value(value, type,
1006 ERROR_ME_NOT) ? Qt : Qnil;
1009 DEFUN("check-valid-char-table-value", Fcheck_valid_char_table_value, 2, 2, 0, /*
1010 Signal an error if VALUE is not a valid value for CHAR-TABLE-TYPE.
1012 (value, char_table_type))
1014 enum char_table_type type = symbol_to_char_table_type(char_table_type);
1016 check_valid_char_table_value(value, type, ERROR_ME);
1020 /* Assign VAL to all characters in RANGE in char table CT. */
1023 put_char_table(Lisp_Char_Table * ct, struct chartab_range *range,
1026 switch (range->type) {
1027 case CHARTAB_RANGE_ALL:
1028 fill_char_table(ct, val);
1029 /* avoid the duplicate call to update_syntax_table() below,
1030 since fill_char_table() also did that. */
1034 case CHARTAB_RANGE_CHARSET:
1035 if (EQ(range->charset, Vcharset_ascii)) {
1037 for (i = 0; i < 128; i++)
1039 } else if (EQ(range->charset, Vcharset_control_1)) {
1041 for (i = 128; i < 160; i++)
1045 XCHARSET_LEADING_BYTE(range->charset) -
1047 ct->level1[lb] = val;
1051 case CHARTAB_RANGE_ROW:
1053 Lisp_Char_Table_Entry *cte;
1055 XCHARSET_LEADING_BYTE(range->charset) -
1057 /* make sure that there is a separate entry for the
1059 if (!CHAR_TABLE_ENTRYP(ct->level1[lb]))
1061 make_char_table_entry(ct->level1[lb]);
1062 cte = XCHAR_TABLE_ENTRY(ct->level1[lb]);
1063 cte->level2[range->row - 32] = val;
1068 case CHARTAB_RANGE_CHAR: {
1070 Lisp_Object charset;
1073 BREAKUP_CHAR(range->ch, charset, byte1, byte2);
1074 if (EQ(charset, Vcharset_ascii))
1075 ct->ascii[byte1] = val;
1076 else if (EQ(charset, Vcharset_control_1))
1077 ct->ascii[byte1 + 128] = val;
1079 Lisp_Char_Table_Entry *cte;
1081 XCHARSET_LEADING_BYTE(charset) -
1083 /* make sure that there is a separate entry for the
1085 if (!CHAR_TABLE_ENTRYP(ct->level1[lb]))
1087 make_char_table_entry(ct->
1089 cte = XCHAR_TABLE_ENTRY(ct->level1[lb]);
1090 /* now CTE is a char table entry for the charset;
1091 each entry is for a single row (or character of
1092 a one-octet charset). */
1093 if (XCHARSET_DIMENSION(charset) == 1)
1094 cte->level2[byte1 - 32] = val;
1096 /* assigning to one character in a two-octet
1098 /* make sure that the charset row contains a
1099 separate entry for each character. */
1100 if (!CHAR_TABLE_ENTRYP
1101 (cte->level2[byte1 - 32]))
1102 cte->level2[byte1 - 32] =
1103 make_char_table_entry(cte->
1108 cte = XCHAR_TABLE_ENTRY(cte->
1109 level2[byte1 - 32]);
1110 cte->level2[byte2 - 32] = val;
1113 #else /* not MULE */
1114 ct->ascii[(unsigned char)(range->ch)] = val;
1116 #endif /* not MULE */
1119 /* shouldnt happen should it? */
1123 if (ct->type == CHAR_TABLE_TYPE_SYNTAX) {
1124 update_syntax_table(ct);
1129 DEFUN("put-char-table", Fput_char_table, 3, 3, 0, /*
1130 Set the value for chars in RANGE to be VALUE in CHAR-TABLE.
1132 RANGE specifies one or more characters to be affected and should be
1133 one of the following:
1135 -- t (all characters are affected)
1136 -- A charset (only allowed when Mule support is present)
1137 -- A vector of two elements: a two-octet charset and a row number
1138 (only allowed when Mule support is present)
1139 -- A single character
1141 VALUE must be a value appropriate for the type of CHAR-TABLE.
1142 See `valid-char-table-type-p'.
1144 (range, value, char_table))
1146 Lisp_Char_Table *ct;
1147 struct chartab_range rainj;
1149 CHECK_CHAR_TABLE(char_table);
1150 ct = XCHAR_TABLE(char_table);
1151 check_valid_char_table_value(value, ct->type, ERROR_ME);
1152 decode_char_table_range(range, &rainj);
1153 value = canonicalize_char_table_value(value, ct->type);
1154 put_char_table(ct, &rainj, value);
1158 /* Map FN over the ASCII chars in CT. */
1161 map_over_charset_ascii(Lisp_Char_Table * ct,
1162 int (*fn) (struct chartab_range * range,
1163 Lisp_Object val, void *arg), void *arg)
1165 struct chartab_range rainj;
1174 rainj.type = CHARTAB_RANGE_CHAR;
1176 for (i = start, retval = 0; i < stop && retval == 0; i++) {
1177 rainj.ch = (Emchar) i;
1178 retval = (fn) (&rainj, ct->ascii[i], arg);
1186 /* Map FN over the Control-1 chars in CT. */
1189 map_over_charset_control_1(Lisp_Char_Table * ct,
1190 int (*fn) (struct chartab_range * range,
1191 Lisp_Object val, void *arg), void *arg)
1193 struct chartab_range rainj;
1196 int stop = start + 32;
1198 rainj.type = CHARTAB_RANGE_CHAR;
1200 for (i = start, retval = 0; i < stop && retval == 0; i++) {
1201 rainj.ch = (Emchar) (i);
1202 retval = (fn) (&rainj, ct->ascii[i], arg);
1208 /* Map FN over the row ROW of two-byte charset CHARSET.
1209 There must be a separate value for that row in the char table.
1210 CTE specifies the char table entry for CHARSET. */
1213 map_over_charset_row(Lisp_Char_Table_Entry * cte,
1214 Lisp_Object charset, int row,
1215 int (*fn) (struct chartab_range * range,
1216 Lisp_Object val, void *arg), void *arg)
1218 Lisp_Object val = cte->level2[row - 32];
1220 if (!CHAR_TABLE_ENTRYP(val)) {
1221 struct chartab_range rainj;
1223 rainj.type = CHARTAB_RANGE_ROW;
1224 rainj.charset = charset;
1226 return (fn) (&rainj, val, arg);
1228 struct chartab_range rainj;
1230 int charset94_p = (XCHARSET_CHARS(charset) == 94);
1231 int start = charset94_p ? 33 : 32;
1232 int stop = charset94_p ? 127 : 128;
1234 cte = XCHAR_TABLE_ENTRY(val);
1236 rainj.type = CHARTAB_RANGE_CHAR;
1238 for (i = start, retval = 0; i < stop && retval == 0; i++) {
1239 rainj.ch = MAKE_CHAR(charset, row, i);
1240 retval = (fn) (&rainj, cte->level2[i - 32], arg);
1247 map_over_other_charset(Lisp_Char_Table * ct, int lb,
1248 int (*fn) (struct chartab_range * range,
1249 Lisp_Object val, void *arg), void *arg)
1251 Lisp_Object val = ct->level1[lb - MIN_LEADING_BYTE];
1252 Lisp_Object charset = CHARSET_BY_LEADING_BYTE(lb);
1254 if (!CHARSETP(charset)
1255 || lb == LEADING_BYTE_ASCII || lb == LEADING_BYTE_CONTROL_1)
1258 if (!CHAR_TABLE_ENTRYP(val)) {
1259 struct chartab_range rainj;
1261 rainj.type = CHARTAB_RANGE_CHARSET;
1262 rainj.charset = charset;
1263 return (fn) (&rainj, val, arg);
1267 Lisp_Char_Table_Entry *cte = XCHAR_TABLE_ENTRY(val);
1268 int charset94_p = (XCHARSET_CHARS(charset) == 94);
1269 int start = charset94_p ? 33 : 32;
1270 int stop = charset94_p ? 127 : 128;
1273 if (XCHARSET_DIMENSION(charset) == 1) {
1274 struct chartab_range rainj;
1275 rainj.type = CHARTAB_RANGE_CHAR;
1277 for (i = start, retval = 0; i < stop && retval == 0;
1279 rainj.ch = MAKE_CHAR(charset, i, 0);
1281 (fn) (&rainj, cte->level2[i - 32], arg);
1284 for (i = start, retval = 0; i < stop && retval == 0;
1287 map_over_charset_row(cte, charset, i, fn,
1297 /* Map FN (with client data ARG) over range RANGE in char table CT.
1298 Mapping stops the first time FN returns non-zero, and that value
1299 becomes the return value of map_char_table(). */
1302 map_char_table(Lisp_Char_Table * ct,
1303 struct chartab_range *range,
1304 int (*fn) (struct chartab_range * range,
1305 Lisp_Object val, void *arg), void *arg)
1307 switch (range->type) {
1308 case CHARTAB_RANGE_ALL:
1312 retval = map_over_charset_ascii(ct, fn, arg);
1316 retval = map_over_charset_control_1(ct, fn, arg);
1321 int start = MIN_LEADING_BYTE;
1322 int stop = start + NUM_LEADING_BYTES;
1324 for (i = start, retval = 0;
1325 i < stop && retval == 0; i++) {
1327 map_over_other_charset(ct, i, fn,
1336 case CHARTAB_RANGE_CHARSET:
1337 return map_over_other_charset(ct,
1338 XCHARSET_LEADING_BYTE(range->
1342 case CHARTAB_RANGE_ROW:
1345 ct->level1[XCHARSET_LEADING_BYTE(range->charset) -
1347 if (!CHAR_TABLE_ENTRYP(val)) {
1348 struct chartab_range rainj;
1350 rainj.type = CHARTAB_RANGE_ROW;
1351 rainj.charset = range->charset;
1352 rainj.row = range->row;
1353 return (fn) (&rainj, val, arg);
1356 map_over_charset_row(XCHAR_TABLE_ENTRY(val),
1358 range->row, fn, arg);
1362 case CHARTAB_RANGE_CHAR:
1364 Emchar ch = range->ch;
1365 Lisp_Object val = CHAR_TABLE_VALUE_UNSAFE(ct, ch);
1366 struct chartab_range rainj;
1368 rainj.type = CHARTAB_RANGE_CHAR;
1370 return (fn) (&rainj, val, arg);
1380 struct slow_map_char_table_arg {
1381 Lisp_Object function;
1386 slow_map_char_table_fun(struct chartab_range *range, Lisp_Object val, void *arg)
1388 Lisp_Object ranjarg = Qnil;
1389 struct slow_map_char_table_arg *closure =
1390 (struct slow_map_char_table_arg *)arg;
1392 switch (range->type) {
1393 case CHARTAB_RANGE_ALL:
1398 case CHARTAB_RANGE_CHARSET:
1399 ranjarg = XCHARSET_NAME(range->charset);
1402 case CHARTAB_RANGE_ROW:
1403 ranjarg = vector2(XCHARSET_NAME(range->charset),
1404 make_int(range->row));
1407 case CHARTAB_RANGE_CHAR:
1408 ranjarg = make_char(range->ch);
1414 closure->retval = call2(closure->function, ranjarg, val);
1415 return !NILP(closure->retval);
1418 DEFUN("map-char-table", Fmap_char_table, 2, 3, 0, /*
1419 Map FUNCTION over entries in CHAR-TABLE, calling it with two args,
1420 each key and value in the table.
1422 RANGE specifies a subrange to map over and is in the same format as
1423 the RANGE argument to `put-range-table'. If omitted or t, it defaults to
1426 (function, char_table, range))
1428 Lisp_Char_Table *ct;
1429 struct slow_map_char_table_arg slarg;
1430 struct gcpro gcpro1, gcpro2;
1431 struct chartab_range rainj;
1433 CHECK_CHAR_TABLE(char_table);
1434 ct = XCHAR_TABLE(char_table);
1437 decode_char_table_range(range, &rainj);
1438 slarg.function = function;
1439 slarg.retval = Qnil;
1440 GCPRO2(slarg.function, slarg.retval);
1441 map_char_table(ct, &rainj, slow_map_char_table_fun, &slarg);
1444 return slarg.retval;
1447 /************************************************************************/
1448 /* Char table read syntax */
1449 /************************************************************************/
1452 chartab_type_validate(Lisp_Object keyword, Lisp_Object value,
1453 Error_behavior errb)
1455 /* #### should deal with ERRB */
1456 symbol_to_char_table_type(value);
1461 chartab_data_validate(Lisp_Object keyword, Lisp_Object value,
1462 Error_behavior errb)
1466 /* #### should deal with ERRB */
1467 EXTERNAL_LIST_LOOP(rest, value) {
1468 Lisp_Object range = XCAR(rest);
1469 struct chartab_range dummy;
1473 signal_simple_error("Invalid list format", value);
1475 if (!CONSP(XCDR(range))
1476 || !NILP(XCDR(XCDR(range))))
1477 signal_simple_error("Invalid range format",
1479 decode_char_table_range(XCAR(range), &dummy);
1480 decode_char_table_range(XCAR(XCDR(range)), &dummy);
1482 decode_char_table_range(range, &dummy);
1488 static Lisp_Object chartab_instantiate(Lisp_Object data)
1490 Lisp_Object chartab;
1491 Lisp_Object type = Qgeneric;
1492 Lisp_Object dataval = Qnil;
1494 while (!NILP(data)) {
1495 Lisp_Object keyw = Fcar(data);
1501 if (EQ(keyw, Qtype))
1503 else if (EQ(keyw, Qdata))
1507 chartab = Fmake_char_table(type);
1510 while (!NILP(data)) {
1511 Lisp_Object range = Fcar(data);
1512 Lisp_Object val = Fcar(Fcdr(data));
1514 data = Fcdr(Fcdr(data));
1516 if (CHAR_OR_CHAR_INTP(XCAR(range))) {
1517 Emchar first = XCHAR_OR_CHAR_INT(Fcar(range));
1519 XCHAR_OR_CHAR_INT(Fcar(Fcdr(range)));
1522 for (i = first; i <= last; i++)
1523 Fput_char_table(make_char(i), val,
1528 Fput_char_table(range, val, chartab);
1536 /************************************************************************/
1537 /* Category Tables, specifically */
1538 /************************************************************************/
1540 DEFUN("category-table-p", Fcategory_table_p, 1, 1, 0, /*
1541 Return t if OBJECT is a category table.
1542 A category table is a type of char table used for keeping track of
1543 categories. Categories are used for classifying characters for use
1544 in regexps -- you can refer to a category rather than having to use
1545 a complicated [] expression (and category lookups are significantly
1548 There are 95 different categories available, one for each printable
1549 character (including space) in the ASCII charset. Each category
1550 is designated by one such character, called a "category designator".
1551 They are specified in a regexp using the syntax "\\cX", where X is
1552 a category designator.
1554 A category table specifies, for each character, the categories that
1555 the character is in. Note that a character can be in more than one
1556 category. More specifically, a category table maps from a character
1557 to either the value nil (meaning the character is in no categories)
1558 or a 95-element bit vector, specifying for each of the 95 categories
1559 whether the character is in that category.
1561 Special Lisp functions are provided that abstract this, so you do not
1562 have to directly manipulate bit vectors.
1566 return (CHAR_TABLEP(object) &&
1567 XCHAR_TABLE_TYPE(object) == CHAR_TABLE_TYPE_CATEGORY) ?
1572 check_category_table(Lisp_Object object, Lisp_Object default_)
1576 while (NILP(Fcategory_table_p(object)))
1577 object = wrong_type_argument(Qcategory_table_p, object);
1582 check_category_char(Emchar ch, Lisp_Object table,
1583 unsigned int designator, unsigned int not_p)
1585 REGISTER Lisp_Object temp;
1586 Lisp_Char_Table *ctbl;
1587 #ifdef ERROR_CHECK_TYPECHECK
1588 if (NILP(Fcategory_table_p(table)))
1589 signal_simple_error("Expected category table", table);
1591 ctbl = XCHAR_TABLE(table);
1592 temp = get_char_table(ch, ctbl);
1597 return bit_vector_bit(XBIT_VECTOR(temp), designator) ? !not_p : not_p;
1600 DEFUN("check-category-at", Fcheck_category_at, 2, 4, 0, /*
1601 Return t if category of the character at POSITION includes DESIGNATOR.
1602 Optional third arg BUFFER specifies which buffer to use, and defaults
1603 to the current buffer.
1604 Optional fourth arg CATEGORY-TABLE specifies the category table to
1605 use, and defaults to BUFFER's category table.
1607 (position, designator, buffer, category_table))
1612 struct buffer *buf = decode_buffer(buffer, 0);
1614 CHECK_INT(position);
1615 CHECK_CATEGORY_DESIGNATOR(designator);
1616 des = XCHAR(designator);
1617 ctbl = check_category_table(category_table, Vstandard_category_table);
1618 ch = BUF_FETCH_CHAR(buf, XINT(position));
1619 return check_category_char(ch, ctbl, des, 0) ? Qt : Qnil;
1622 DEFUN("char-in-category-p", Fchar_in_category_p, 2, 3, 0, /*
1623 Return t if category of CHARACTER includes DESIGNATOR, else nil.
1624 Optional third arg CATEGORY-TABLE specifies the category table to use,
1625 and defaults to the standard category table.
1627 (character, designator, category_table))
1633 CHECK_CATEGORY_DESIGNATOR(designator);
1634 des = XCHAR(designator);
1635 CHECK_CHAR(character);
1636 ch = XCHAR(character);
1637 ctbl = check_category_table(category_table, Vstandard_category_table);
1638 return check_category_char(ch, ctbl, des, 0) ? Qt : Qnil;
1641 DEFUN("category-table", Fcategory_table, 0, 1, 0, /*
1642 Return BUFFER's current category table.
1643 BUFFER defaults to the current buffer.
1647 return decode_buffer(buffer, 0)->category_table;
1650 DEFUN("standard-category-table", Fstandard_category_table, 0, 0, 0, /*
1651 Return the standard category table.
1652 This is the one used for new buffers.
1656 return Vstandard_category_table;
1659 DEFUN("copy-category-table", Fcopy_category_table, 0, 1, 0, /*
1660 Return a new category table which is a copy of CATEGORY-TABLE.
1661 CATEGORY-TABLE defaults to the standard category table.
1665 if (NILP(Vstandard_category_table))
1666 return Fmake_char_table(Qcategory);
1669 check_category_table(category_table, Vstandard_category_table);
1670 return Fcopy_char_table(category_table);
1673 DEFUN("set-category-table", Fset_category_table, 1, 2, 0, /*
1674 Select CATEGORY-TABLE as the new category table for BUFFER.
1675 BUFFER defaults to the current buffer if omitted.
1677 (category_table, buffer))
1679 struct buffer *buf = decode_buffer(buffer, 0);
1680 category_table = check_category_table(category_table, Qnil);
1681 buf->category_table = category_table;
1682 /* Indicate that this buffer now has a specified category table. */
1683 buf->local_var_flags |= XINT(buffer_local_flags.category_table);
1684 return category_table;
1687 DEFUN("category-designator-p", Fcategory_designator_p, 1, 1, 0, /*
1688 Return t if OBJECT is a category designator (a char in the range ' ' to '~').
1692 return CATEGORY_DESIGNATORP(object) ? Qt : Qnil;
1695 DEFUN("category-table-value-p", Fcategory_table_value_p, 1, 1, 0, /*
1696 Return t if OBJECT is a category table value.
1697 Valid values are nil or a bit vector of size 95.
1701 return CATEGORY_TABLE_VALUEP(object) ? Qt : Qnil;
1704 #define CATEGORYP(x) \
1705 (CHARP (x) && XCHAR (x) >= 0x20 && XCHAR (x) <= 0x7E)
1707 #define CATEGORY_SET(c) \
1708 (get_char_table(c, XCHAR_TABLE(current_buffer->category_table)))
1710 /* Return 1 if CATEGORY_SET contains CATEGORY, else return 0.
1711 The faster version of `!NILP (Faref (category_set, category))'. */
1712 #define CATEGORY_MEMBER(category, category_set) \
1713 (bit_vector_bit(XBIT_VECTOR (category_set), category - 32))
1715 /* Return 1 if there is a word boundary between two word-constituent
1716 characters C1 and C2 if they appear in this order, else return 0.
1717 Use the macro WORD_BOUNDARY_P instead of calling this function
1720 int word_boundary_p(Emchar c1, Emchar c2);
1721 int word_boundary_p(Emchar c1, Emchar c2)
1723 Lisp_Object category_set1, category_set2;
1728 if (COMPOSITE_CHAR_P(c1))
1729 c1 = cmpchar_component(c1, 0, 1);
1730 if (COMPOSITE_CHAR_P(c2))
1731 c2 = cmpchar_component(c2, 0, 1);
1734 if (EQ(CHAR_CHARSET(c1), CHAR_CHARSET(c2))) {
1735 tail = Vword_separating_categories;
1738 tail = Vword_combining_categories;
1742 category_set1 = CATEGORY_SET(c1);
1743 if (NILP(category_set1))
1744 return default_result;
1745 category_set2 = CATEGORY_SET(c2);
1746 if (NILP(category_set2))
1747 return default_result;
1749 for (; CONSP(tail); tail = XCONS(tail)->cdr) {
1750 Lisp_Object elt = XCONS(tail)->car;
1753 && CATEGORYP(XCONS(elt)->car)
1754 && CATEGORYP(XCONS(elt)->cdr)
1755 && CATEGORY_MEMBER(XCHAR(XCONS(elt)->car), category_set1)
1756 && CATEGORY_MEMBER(XCHAR(XCONS(elt)->cdr), category_set2))
1757 return !default_result;
1759 return default_result;
1763 void syms_of_chartab(void)
1765 INIT_LRECORD_IMPLEMENTATION(char_table);
1768 INIT_LRECORD_IMPLEMENTATION(char_table_entry);
1770 defsymbol(&Qcategory_table_p, "category-table-p");
1771 defsymbol(&Qcategory_designator_p, "category-designator-p");
1772 defsymbol(&Qcategory_table_value_p, "category-table-value-p");
1775 defsymbol(&Qchar_table, "char-table");
1776 defsymbol(&Qchar_tablep, "char-table-p");
1778 DEFSUBR(Fchar_table_p);
1779 DEFSUBR(Fchar_table_type_list);
1780 DEFSUBR(Fvalid_char_table_type_p);
1781 DEFSUBR(Fchar_table_type);
1782 DEFSUBR(Freset_char_table);
1783 DEFSUBR(Fmake_char_table);
1784 DEFSUBR(Fcopy_char_table);
1785 DEFSUBR(Fget_char_table);
1786 DEFSUBR(Fget_range_char_table);
1787 DEFSUBR(Fvalid_char_table_value_p);
1788 DEFSUBR(Fcheck_valid_char_table_value);
1789 DEFSUBR(Fput_char_table);
1790 DEFSUBR(Fmap_char_table);
1793 DEFSUBR(Fcategory_table_p);
1794 DEFSUBR(Fcategory_table);
1795 DEFSUBR(Fstandard_category_table);
1796 DEFSUBR(Fcopy_category_table);
1797 DEFSUBR(Fset_category_table);
1798 DEFSUBR(Fcheck_category_at);
1799 DEFSUBR(Fchar_in_category_p);
1800 DEFSUBR(Fcategory_designator_p);
1801 DEFSUBR(Fcategory_table_value_p);
1806 void vars_of_chartab(void)
1808 /* DO NOT staticpro this. It works just like Vweak_hash_tables. */
1809 Vall_syntax_tables = Qnil;
1810 dump_add_weak_object_chain(&Vall_syntax_tables);
1813 void structure_type_create_chartab(void)
1815 struct structure_type *st;
1817 st = define_structure_type(Qchar_table, 0, chartab_instantiate);
1819 define_structure_type_keyword(st, Qtype, chartab_type_validate);
1820 define_structure_type_keyword(st, Qdata, chartab_data_validate);
1823 void complex_vars_of_chartab(void)
1826 /* Set this now, so first buffer creation can refer to it. */
1827 /* Make it nil before calling copy-category-table
1828 so that copy-category-table will know not to try to copy from garbage */
1829 Vstandard_category_table = Qnil;
1830 Vstandard_category_table = Fcopy_category_table(Qnil);
1831 staticpro(&Vstandard_category_table);
1833 DEFVAR_LISP("word-combining-categories", &Vword_combining_categories /*
1834 List of pair (cons) of categories to determine word boundary.
1836 Emacs treats a sequence of word constituent characters as a single
1837 word (i.e. finds no word boundary between them) iff they belongs to
1838 the same charset. But, exceptions are allowed in the following cases.
1840 \(1) The case that characters are in different charsets is controlled
1841 by the variable `word-combining-categories'.
1843 Emacs finds no word boundary between characters of different charsets
1844 if they have categories matching some element of this list.
1846 More precisely, if an element of this list is a cons of category CAT1
1847 and CAT2, and a multibyte character C1 which has CAT1 is followed by
1848 C2 which has CAT2, there's no word boundary between C1 and C2.
1850 For instance, to tell that ASCII characters and Latin-1 characters can
1851 form a single word, the element `(?l . ?l)' should be in this list
1852 because both characters have the category `l' (Latin characters).
1854 \(2) The case that character are in the same charset is controlled by
1855 the variable `word-separating-categories'.
1857 Emacs find a word boundary between characters of the same charset
1858 if they have categories matching some element of this list.
1860 More precisely, if an element of this list is a cons of category CAT1
1861 and CAT2, and a multibyte character C1 which has CAT1 is followed by
1862 C2 which has CAT2, there's a word boundary between C1 and C2.
1864 For instance, to tell that there's a word boundary between Japanese
1865 Hiragana and Japanese Kanji (both are in the same charset), the
1866 element `(?H . ?C) should be in this list.
1869 Vword_combining_categories = Qnil;
1871 DEFVAR_LISP("word-separating-categories", &Vword_separating_categories /*
1872 List of pair (cons) of categories to determine word boundary.
1873 See the documentation of the variable `word-combining-categories'.
1876 Vword_separating_categories = Qnil;