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.
7 Copyright (C) 1999,2000,2001,2002,2003,2004 MORIOKA Tomohiko
9 This file is part of XEmacs.
11 XEmacs is free software; you can redistribute it and/or modify it
12 under the terms of the GNU General Public License as published by the
13 Free Software Foundation; either version 2, or (at your option) any
16 XEmacs is distributed in the hope that it will be useful, but WITHOUT
17 ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
18 FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
21 You should have received a copy of the GNU General Public License
22 along with XEmacs; see the file COPYING. If not, write to
23 the Free Software Foundation, Inc., 59 Temple Place - Suite 330,
24 Boston, MA 02111-1307, USA. */
26 /* Synched up with: Mule 2.3. Not synched with FSF.
28 This file was written independently of the FSF implementation,
29 and is not compatible. */
33 Ben Wing: wrote, for 19.13 (Mule). Some category table stuff
34 loosely based on the original Mule.
35 Jareth Hein: fixed a couple of bugs in the implementation, and
36 added regex support for categories with check_category_at
37 MORIOKA Tomohiko: Rewritten for XEmacs CHISE
50 Lisp_Object Qchar_tablep, Qchar_table;
52 Lisp_Object Vall_syntax_tables;
55 Lisp_Object Qcategory_table_p;
56 Lisp_Object Qcategory_designator_p;
57 Lisp_Object Qcategory_table_value_p;
59 Lisp_Object Vstandard_category_table;
61 /* Variables to determine word boundary. */
62 Lisp_Object Vword_combining_categories, Vword_separating_categories;
67 Lisp_Object Vchise_db_directory;
68 Lisp_Object Vchise_system_db_directory;
70 CHISE_DS *default_chise_data_source = NULL;
75 EXFUN (Fchar_refs_simplify_char_specs, 1);
76 extern Lisp_Object Qideographic_structure;
78 Lisp_Object Vnext_defined_char_id;
79 EXFUN (Fdefine_char, 1);
81 EXFUN (Fmap_char_attribute, 3);
84 EXFUN (Fmount_char_attribute_table, 1);
88 EXFUN (Fload_char_attribute_table, 1);
90 Lisp_Object Vchar_db_stingy_mode;
93 #define BT_UINT8_MIN 0
94 #define BT_UINT8_MAX (UCHAR_MAX - 4)
95 #define BT_UINT8_t (UCHAR_MAX - 3)
96 #define BT_UINT8_nil (UCHAR_MAX - 2)
97 #define BT_UINT8_unbound (UCHAR_MAX - 1)
98 #define BT_UINT8_unloaded UCHAR_MAX
100 INLINE_HEADER int INT_UINT8_P (Lisp_Object obj);
101 INLINE_HEADER int UINT8_VALUE_P (Lisp_Object obj);
102 INLINE_HEADER unsigned char UINT8_ENCODE (Lisp_Object obj);
103 INLINE_HEADER Lisp_Object UINT8_DECODE (unsigned char n);
104 INLINE_HEADER unsigned short UINT8_TO_UINT16 (unsigned char n);
107 INT_UINT8_P (Lisp_Object obj)
111 int num = XINT (obj);
113 return (BT_UINT8_MIN <= num) && (num <= BT_UINT8_MAX);
120 UINT8_VALUE_P (Lisp_Object obj)
122 return EQ (obj, Qunloaded) || EQ (obj, Qunbound)
123 || EQ (obj, Qnil) || EQ (obj, Qt) || INT_UINT8_P (obj);
126 INLINE_HEADER unsigned char
127 UINT8_ENCODE (Lisp_Object obj)
129 if (EQ (obj, Qunloaded))
130 return BT_UINT8_unloaded;
131 else if (EQ (obj, Qunbound))
132 return BT_UINT8_unbound;
133 else if (EQ (obj, Qnil))
135 else if (EQ (obj, Qt))
141 INLINE_HEADER Lisp_Object
142 UINT8_DECODE (unsigned char n)
144 if (n == BT_UINT8_unloaded)
146 else if (n == BT_UINT8_unbound)
148 else if (n == BT_UINT8_nil)
150 else if (n == BT_UINT8_t)
157 mark_uint8_byte_table (Lisp_Object obj)
163 print_uint8_byte_table (Lisp_Object obj,
164 Lisp_Object printcharfun, int escapeflag)
166 Lisp_Uint8_Byte_Table *bte = XUINT8_BYTE_TABLE (obj);
168 struct gcpro gcpro1, gcpro2;
169 GCPRO2 (obj, printcharfun);
171 write_c_string ("\n#<uint8-byte-table", printcharfun);
172 for (i = 0; i < 256; i++)
174 unsigned char n = bte->property[i];
176 write_c_string ("\n ", printcharfun);
177 write_c_string (" ", printcharfun);
178 if (n == BT_UINT8_unbound)
179 write_c_string ("void", printcharfun);
180 else if (n == BT_UINT8_nil)
181 write_c_string ("nil", printcharfun);
182 else if (n == BT_UINT8_t)
183 write_c_string ("t", printcharfun);
188 sprintf (buf, "%hd", n);
189 write_c_string (buf, printcharfun);
193 write_c_string (">", printcharfun);
197 uint8_byte_table_equal (Lisp_Object obj1, Lisp_Object obj2, int depth)
199 Lisp_Uint8_Byte_Table *te1 = XUINT8_BYTE_TABLE (obj1);
200 Lisp_Uint8_Byte_Table *te2 = XUINT8_BYTE_TABLE (obj2);
203 for (i = 0; i < 256; i++)
204 if (te1->property[i] != te2->property[i])
210 uint8_byte_table_hash (Lisp_Object obj, int depth)
212 Lisp_Uint8_Byte_Table *te = XUINT8_BYTE_TABLE (obj);
216 for (i = 0; i < 256; i++)
217 hash = HASH2 (hash, te->property[i]);
221 static const struct lrecord_description uint8_byte_table_description[] = {
225 DEFINE_LRECORD_IMPLEMENTATION ("uint8-byte-table", uint8_byte_table,
226 mark_uint8_byte_table,
227 print_uint8_byte_table,
228 0, uint8_byte_table_equal,
229 uint8_byte_table_hash,
230 uint8_byte_table_description,
231 Lisp_Uint8_Byte_Table);
234 make_uint8_byte_table (unsigned char initval)
238 Lisp_Uint8_Byte_Table *cte;
240 cte = alloc_lcrecord_type (Lisp_Uint8_Byte_Table,
241 &lrecord_uint8_byte_table);
243 for (i = 0; i < 256; i++)
244 cte->property[i] = initval;
246 XSETUINT8_BYTE_TABLE (obj, cte);
251 copy_uint8_byte_table (Lisp_Object entry)
253 Lisp_Uint8_Byte_Table *cte = XUINT8_BYTE_TABLE (entry);
256 Lisp_Uint8_Byte_Table *ctenew
257 = alloc_lcrecord_type (Lisp_Uint8_Byte_Table,
258 &lrecord_uint8_byte_table);
260 for (i = 0; i < 256; i++)
262 ctenew->property[i] = cte->property[i];
265 XSETUINT8_BYTE_TABLE (obj, ctenew);
270 uint8_byte_table_same_value_p (Lisp_Object obj)
272 Lisp_Uint8_Byte_Table *bte = XUINT8_BYTE_TABLE (obj);
273 unsigned char v0 = bte->property[0];
276 for (i = 1; i < 256; i++)
278 if (bte->property[i] != v0)
285 map_over_uint8_byte_table (Lisp_Uint8_Byte_Table *ct, Lisp_Char_Table* root,
286 Emchar ofs, int place,
287 int (*fn) (struct chartab_range *range,
288 Lisp_Object val, void *arg),
291 struct chartab_range rainj;
293 int unit = 1 << (8 * place);
297 rainj.type = CHARTAB_RANGE_CHAR;
299 for (i = 0, retval = 0; i < 256 && retval == 0; i++)
301 if (ct->property[i] == BT_UINT8_unloaded)
305 for (; c < c1 && retval == 0; c++)
307 Lisp_Object ret = get_char_id_table (root, c);
312 retval = (fn) (&rainj, ret, arg);
316 ct->property[i] = BT_UINT8_unbound;
320 else if (ct->property[i] != BT_UINT8_unbound)
323 for (; c < c1 && retval == 0; c++)
326 retval = (fn) (&rainj, UINT8_DECODE (ct->property[i]), arg);
337 save_uint8_byte_table (Lisp_Uint8_Byte_Table *ct, Lisp_Char_Table* root,
339 CHISE_Feature feature,
343 Emchar ofs, int place,
344 Lisp_Object (*filter)(Lisp_Object value))
346 struct chartab_range rainj;
348 int unit = 1 << (8 * place);
352 rainj.type = CHARTAB_RANGE_CHAR;
354 for (i = 0, retval = 0; i < 256 && retval == 0; i++)
356 if (ct->property[i] == BT_UINT8_unloaded)
360 else if (ct->property[i] != BT_UINT8_unbound)
363 for (; c < c1 && retval == 0; c++)
366 chise_char_set_feature_value
369 (Fprin1_to_string (UINT8_DECODE (ct->property[i]),
372 Fput_database (Fprin1_to_string (make_char (c), Qnil),
373 Fprin1_to_string (UINT8_DECODE (ct->property[i]),
385 #define BT_UINT16_MIN 0
386 #define BT_UINT16_MAX (USHRT_MAX - 4)
387 #define BT_UINT16_t (USHRT_MAX - 3)
388 #define BT_UINT16_nil (USHRT_MAX - 2)
389 #define BT_UINT16_unbound (USHRT_MAX - 1)
390 #define BT_UINT16_unloaded USHRT_MAX
392 INLINE_HEADER int INT_UINT16_P (Lisp_Object obj);
393 INLINE_HEADER int UINT16_VALUE_P (Lisp_Object obj);
394 INLINE_HEADER unsigned short UINT16_ENCODE (Lisp_Object obj);
395 INLINE_HEADER Lisp_Object UINT16_DECODE (unsigned short us);
398 INT_UINT16_P (Lisp_Object obj)
402 int num = XINT (obj);
404 return (BT_UINT16_MIN <= num) && (num <= BT_UINT16_MAX);
411 UINT16_VALUE_P (Lisp_Object obj)
413 return EQ (obj, Qunloaded) || EQ (obj, Qunbound)
414 || EQ (obj, Qnil) || EQ (obj, Qt) || INT_UINT16_P (obj);
417 INLINE_HEADER unsigned short
418 UINT16_ENCODE (Lisp_Object obj)
420 if (EQ (obj, Qunloaded))
421 return BT_UINT16_unloaded;
422 else if (EQ (obj, Qunbound))
423 return BT_UINT16_unbound;
424 else if (EQ (obj, Qnil))
425 return BT_UINT16_nil;
426 else if (EQ (obj, Qt))
432 INLINE_HEADER Lisp_Object
433 UINT16_DECODE (unsigned short n)
435 if (n == BT_UINT16_unloaded)
437 else if (n == BT_UINT16_unbound)
439 else if (n == BT_UINT16_nil)
441 else if (n == BT_UINT16_t)
447 INLINE_HEADER unsigned short
448 UINT8_TO_UINT16 (unsigned char n)
450 if (n == BT_UINT8_unloaded)
451 return BT_UINT16_unloaded;
452 else if (n == BT_UINT8_unbound)
453 return BT_UINT16_unbound;
454 else if (n == BT_UINT8_nil)
455 return BT_UINT16_nil;
456 else if (n == BT_UINT8_t)
463 mark_uint16_byte_table (Lisp_Object obj)
469 print_uint16_byte_table (Lisp_Object obj,
470 Lisp_Object printcharfun, int escapeflag)
472 Lisp_Uint16_Byte_Table *bte = XUINT16_BYTE_TABLE (obj);
474 struct gcpro gcpro1, gcpro2;
475 GCPRO2 (obj, printcharfun);
477 write_c_string ("\n#<uint16-byte-table", printcharfun);
478 for (i = 0; i < 256; i++)
480 unsigned short n = bte->property[i];
482 write_c_string ("\n ", printcharfun);
483 write_c_string (" ", printcharfun);
484 if (n == BT_UINT16_unbound)
485 write_c_string ("void", printcharfun);
486 else if (n == BT_UINT16_nil)
487 write_c_string ("nil", printcharfun);
488 else if (n == BT_UINT16_t)
489 write_c_string ("t", printcharfun);
494 sprintf (buf, "%hd", n);
495 write_c_string (buf, printcharfun);
499 write_c_string (">", printcharfun);
503 uint16_byte_table_equal (Lisp_Object obj1, Lisp_Object obj2, int depth)
505 Lisp_Uint16_Byte_Table *te1 = XUINT16_BYTE_TABLE (obj1);
506 Lisp_Uint16_Byte_Table *te2 = XUINT16_BYTE_TABLE (obj2);
509 for (i = 0; i < 256; i++)
510 if (te1->property[i] != te2->property[i])
516 uint16_byte_table_hash (Lisp_Object obj, int depth)
518 Lisp_Uint16_Byte_Table *te = XUINT16_BYTE_TABLE (obj);
522 for (i = 0; i < 256; i++)
523 hash = HASH2 (hash, te->property[i]);
527 static const struct lrecord_description uint16_byte_table_description[] = {
531 DEFINE_LRECORD_IMPLEMENTATION ("uint16-byte-table", uint16_byte_table,
532 mark_uint16_byte_table,
533 print_uint16_byte_table,
534 0, uint16_byte_table_equal,
535 uint16_byte_table_hash,
536 uint16_byte_table_description,
537 Lisp_Uint16_Byte_Table);
540 make_uint16_byte_table (unsigned short initval)
544 Lisp_Uint16_Byte_Table *cte;
546 cte = alloc_lcrecord_type (Lisp_Uint16_Byte_Table,
547 &lrecord_uint16_byte_table);
549 for (i = 0; i < 256; i++)
550 cte->property[i] = initval;
552 XSETUINT16_BYTE_TABLE (obj, cte);
557 copy_uint16_byte_table (Lisp_Object entry)
559 Lisp_Uint16_Byte_Table *cte = XUINT16_BYTE_TABLE (entry);
562 Lisp_Uint16_Byte_Table *ctenew
563 = alloc_lcrecord_type (Lisp_Uint16_Byte_Table,
564 &lrecord_uint16_byte_table);
566 for (i = 0; i < 256; i++)
568 ctenew->property[i] = cte->property[i];
571 XSETUINT16_BYTE_TABLE (obj, ctenew);
576 expand_uint8_byte_table_to_uint16 (Lisp_Object table)
580 Lisp_Uint8_Byte_Table* bte = XUINT8_BYTE_TABLE(table);
581 Lisp_Uint16_Byte_Table* cte;
583 cte = alloc_lcrecord_type (Lisp_Uint16_Byte_Table,
584 &lrecord_uint16_byte_table);
585 for (i = 0; i < 256; i++)
587 cte->property[i] = UINT8_TO_UINT16 (bte->property[i]);
589 XSETUINT16_BYTE_TABLE (obj, cte);
594 uint16_byte_table_same_value_p (Lisp_Object obj)
596 Lisp_Uint16_Byte_Table *bte = XUINT16_BYTE_TABLE (obj);
597 unsigned short v0 = bte->property[0];
600 for (i = 1; i < 256; i++)
602 if (bte->property[i] != v0)
609 map_over_uint16_byte_table (Lisp_Uint16_Byte_Table *ct, Lisp_Char_Table* root,
610 Emchar ofs, int place,
611 int (*fn) (struct chartab_range *range,
612 Lisp_Object val, void *arg),
615 struct chartab_range rainj;
617 int unit = 1 << (8 * place);
621 rainj.type = CHARTAB_RANGE_CHAR;
623 for (i = 0, retval = 0; i < 256 && retval == 0; i++)
625 if (ct->property[i] == BT_UINT16_unloaded)
629 for (; c < c1 && retval == 0; c++)
631 Lisp_Object ret = get_char_id_table (root, c);
636 retval = (fn) (&rainj, ret, arg);
640 ct->property[i] = BT_UINT16_unbound;
644 else if (ct->property[i] != BT_UINT16_unbound)
647 for (; c < c1 && retval == 0; c++)
650 retval = (fn) (&rainj, UINT16_DECODE (ct->property[i]), arg);
661 save_uint16_byte_table (Lisp_Uint16_Byte_Table *ct, Lisp_Char_Table* root,
663 CHISE_Feature feature,
667 Emchar ofs, int place,
668 Lisp_Object (*filter)(Lisp_Object value))
670 struct chartab_range rainj;
672 int unit = 1 << (8 * place);
676 rainj.type = CHARTAB_RANGE_CHAR;
678 for (i = 0, retval = 0; i < 256 && retval == 0; i++)
680 if (ct->property[i] == BT_UINT16_unloaded)
684 else if (ct->property[i] != BT_UINT16_unbound)
687 for (; c < c1 && retval == 0; c++)
690 chise_char_set_feature_value
693 (Fprin1_to_string (UINT16_DECODE (ct->property[i]),
696 Fput_database (Fprin1_to_string (make_char (c), Qnil),
697 Fprin1_to_string (UINT16_DECODE (ct->property[i]),
711 mark_byte_table (Lisp_Object obj)
713 Lisp_Byte_Table *cte = XBYTE_TABLE (obj);
716 for (i = 0; i < 256; i++)
718 mark_object (cte->property[i]);
724 print_byte_table (Lisp_Object obj, Lisp_Object printcharfun, int escapeflag)
726 Lisp_Byte_Table *bte = XBYTE_TABLE (obj);
728 struct gcpro gcpro1, gcpro2;
729 GCPRO2 (obj, printcharfun);
731 write_c_string ("\n#<byte-table", printcharfun);
732 for (i = 0; i < 256; i++)
734 Lisp_Object elt = bte->property[i];
736 write_c_string ("\n ", printcharfun);
737 write_c_string (" ", printcharfun);
738 if (EQ (elt, Qunbound))
739 write_c_string ("void", printcharfun);
741 print_internal (elt, printcharfun, escapeflag);
744 write_c_string (">", printcharfun);
748 byte_table_equal (Lisp_Object obj1, Lisp_Object obj2, int depth)
750 Lisp_Byte_Table *cte1 = XBYTE_TABLE (obj1);
751 Lisp_Byte_Table *cte2 = XBYTE_TABLE (obj2);
754 for (i = 0; i < 256; i++)
755 if (BYTE_TABLE_P (cte1->property[i]))
757 if (BYTE_TABLE_P (cte2->property[i]))
759 if (!byte_table_equal (cte1->property[i],
760 cte2->property[i], depth + 1))
767 if (!internal_equal (cte1->property[i], cte2->property[i], depth + 1))
773 byte_table_hash (Lisp_Object obj, int depth)
775 Lisp_Byte_Table *cte = XBYTE_TABLE (obj);
777 return internal_array_hash (cte->property, 256, depth);
780 static const struct lrecord_description byte_table_description[] = {
781 { XD_LISP_OBJECT_ARRAY, offsetof(Lisp_Byte_Table, property), 256 },
785 DEFINE_LRECORD_IMPLEMENTATION ("byte-table", byte_table,
790 byte_table_description,
794 make_byte_table (Lisp_Object initval)
798 Lisp_Byte_Table *cte;
800 cte = alloc_lcrecord_type (Lisp_Byte_Table, &lrecord_byte_table);
802 for (i = 0; i < 256; i++)
803 cte->property[i] = initval;
805 XSETBYTE_TABLE (obj, cte);
810 copy_byte_table (Lisp_Object entry)
812 Lisp_Byte_Table *cte = XBYTE_TABLE (entry);
815 Lisp_Byte_Table *ctnew
816 = alloc_lcrecord_type (Lisp_Byte_Table, &lrecord_byte_table);
818 for (i = 0; i < 256; i++)
820 if (UINT8_BYTE_TABLE_P (cte->property[i]))
822 ctnew->property[i] = copy_uint8_byte_table (cte->property[i]);
824 else if (UINT16_BYTE_TABLE_P (cte->property[i]))
826 ctnew->property[i] = copy_uint16_byte_table (cte->property[i]);
828 else if (BYTE_TABLE_P (cte->property[i]))
830 ctnew->property[i] = copy_byte_table (cte->property[i]);
833 ctnew->property[i] = cte->property[i];
836 XSETBYTE_TABLE (obj, ctnew);
841 byte_table_same_value_p (Lisp_Object obj)
843 Lisp_Byte_Table *bte = XBYTE_TABLE (obj);
844 Lisp_Object v0 = bte->property[0];
847 for (i = 1; i < 256; i++)
849 if (!internal_equal (bte->property[i], v0, 0))
856 map_over_byte_table (Lisp_Byte_Table *ct, Lisp_Char_Table* root,
857 Emchar ofs, int place,
858 int (*fn) (struct chartab_range *range,
859 Lisp_Object val, void *arg),
864 int unit = 1 << (8 * place);
867 for (i = 0, retval = 0; i < 256 && retval == 0; i++)
870 if (UINT8_BYTE_TABLE_P (v))
873 = map_over_uint8_byte_table (XUINT8_BYTE_TABLE(v), root,
874 c, place - 1, fn, arg);
877 else if (UINT16_BYTE_TABLE_P (v))
880 = map_over_uint16_byte_table (XUINT16_BYTE_TABLE(v), root,
881 c, place - 1, fn, arg);
884 else if (BYTE_TABLE_P (v))
886 retval = map_over_byte_table (XBYTE_TABLE(v), root,
887 c, place - 1, fn, arg);
890 else if (EQ (v, Qunloaded))
893 struct chartab_range rainj;
894 Emchar c1 = c + unit;
896 rainj.type = CHARTAB_RANGE_CHAR;
898 for (; c < c1 && retval == 0; c++)
900 Lisp_Object ret = get_char_id_table (root, c);
905 retval = (fn) (&rainj, ret, arg);
909 ct->property[i] = Qunbound;
913 else if (!UNBOUNDP (v))
915 struct chartab_range rainj;
916 Emchar c1 = c + unit;
918 rainj.type = CHARTAB_RANGE_CHAR;
920 for (; c < c1 && retval == 0; c++)
923 retval = (fn) (&rainj, v, arg);
934 save_byte_table (Lisp_Byte_Table *ct, Lisp_Char_Table* root,
936 CHISE_Feature feature,
940 Emchar ofs, int place,
941 Lisp_Object (*filter)(Lisp_Object value))
945 int unit = 1 << (8 * place);
948 for (i = 0, retval = 0; i < 256 && retval == 0; i++)
951 if (UINT8_BYTE_TABLE_P (v))
953 save_uint8_byte_table (XUINT8_BYTE_TABLE(v), root,
959 c, place - 1, filter);
962 else if (UINT16_BYTE_TABLE_P (v))
964 save_uint16_byte_table (XUINT16_BYTE_TABLE(v), root,
970 c, place - 1, filter);
973 else if (BYTE_TABLE_P (v))
975 save_byte_table (XBYTE_TABLE(v), root,
981 c, place - 1, filter);
984 else if (EQ (v, Qunloaded))
988 else if (!UNBOUNDP (v))
990 struct chartab_range rainj;
991 Emchar c1 = c + unit;
996 rainj.type = CHARTAB_RANGE_CHAR;
998 for (; c < c1 && retval == 0; c++)
1000 #ifdef HAVE_LIBCHISE
1001 chise_char_set_feature_value
1002 (c, feature, XSTRING_DATA (Fprin1_to_string (v, Qnil)));
1004 Fput_database (Fprin1_to_string (make_char (c), Qnil),
1005 Fprin1_to_string (v, Qnil),
1017 get_byte_table (Lisp_Object table, unsigned char idx)
1019 if (UINT8_BYTE_TABLE_P (table))
1020 return UINT8_DECODE (XUINT8_BYTE_TABLE(table)->property[idx]);
1021 else if (UINT16_BYTE_TABLE_P (table))
1022 return UINT16_DECODE (XUINT16_BYTE_TABLE(table)->property[idx]);
1023 else if (BYTE_TABLE_P (table))
1024 return XBYTE_TABLE(table)->property[idx];
1030 put_byte_table (Lisp_Object table, unsigned char idx, Lisp_Object value)
1032 if (UINT8_BYTE_TABLE_P (table))
1034 if (UINT8_VALUE_P (value))
1036 XUINT8_BYTE_TABLE(table)->property[idx] = UINT8_ENCODE (value);
1037 if (!UINT8_BYTE_TABLE_P (value) &&
1038 !UINT16_BYTE_TABLE_P (value) && !BYTE_TABLE_P (value)
1039 && uint8_byte_table_same_value_p (table))
1044 else if (UINT16_VALUE_P (value))
1046 Lisp_Object new = expand_uint8_byte_table_to_uint16 (table);
1048 XUINT16_BYTE_TABLE(new)->property[idx] = UINT16_ENCODE (value);
1053 Lisp_Object new = make_byte_table (Qnil);
1056 for (i = 0; i < 256; i++)
1058 XBYTE_TABLE(new)->property[i]
1059 = UINT8_DECODE (XUINT8_BYTE_TABLE(table)->property[i]);
1061 XBYTE_TABLE(new)->property[idx] = value;
1065 else if (UINT16_BYTE_TABLE_P (table))
1067 if (UINT16_VALUE_P (value))
1069 XUINT16_BYTE_TABLE(table)->property[idx] = UINT16_ENCODE (value);
1070 if (!UINT8_BYTE_TABLE_P (value) &&
1071 !UINT16_BYTE_TABLE_P (value) && !BYTE_TABLE_P (value)
1072 && uint16_byte_table_same_value_p (table))
1079 Lisp_Object new = make_byte_table (Qnil);
1082 for (i = 0; i < 256; i++)
1084 XBYTE_TABLE(new)->property[i]
1085 = UINT16_DECODE (XUINT16_BYTE_TABLE(table)->property[i]);
1087 XBYTE_TABLE(new)->property[idx] = value;
1091 else if (BYTE_TABLE_P (table))
1093 XBYTE_TABLE(table)->property[idx] = value;
1094 if (!UINT8_BYTE_TABLE_P (value) &&
1095 !UINT16_BYTE_TABLE_P (value) && !BYTE_TABLE_P (value)
1096 && byte_table_same_value_p (table))
1101 else if (!internal_equal (table, value, 0))
1103 if (UINT8_VALUE_P (table) && UINT8_VALUE_P (value))
1105 table = make_uint8_byte_table (UINT8_ENCODE (table));
1106 XUINT8_BYTE_TABLE(table)->property[idx] = UINT8_ENCODE (value);
1108 else if (UINT16_VALUE_P (table) && UINT16_VALUE_P (value))
1110 table = make_uint16_byte_table (UINT16_ENCODE (table));
1111 XUINT16_BYTE_TABLE(table)->property[idx] = UINT16_ENCODE (value);
1115 table = make_byte_table (table);
1116 XBYTE_TABLE(table)->property[idx] = value;
1124 make_char_id_table (Lisp_Object initval)
1127 obj = Fmake_char_table (Qgeneric);
1128 fill_char_table (XCHAR_TABLE (obj), initval);
1133 #if defined(HAVE_CHISE) && !defined(HAVE_LIBCHISE_LIBCHISE)
1134 Lisp_Object Qsystem_char_id;
1137 Lisp_Object Qcomposition;
1138 Lisp_Object Q_decomposition;
1139 Lisp_Object Q_unified;
1140 Lisp_Object Q_unified_from;
1141 Lisp_Object Qto_ucs;
1142 Lisp_Object Q_ucs_unified;
1143 Lisp_Object Qcompat;
1144 Lisp_Object Qisolated;
1145 Lisp_Object Qinitial;
1146 Lisp_Object Qmedial;
1148 Lisp_Object Qvertical;
1149 Lisp_Object QnoBreak;
1150 Lisp_Object Qfraction;
1153 Lisp_Object Qcircle;
1154 Lisp_Object Qsquare;
1156 Lisp_Object Qnarrow;
1160 Emchar to_char_id (Lisp_Object v, char* err_msg, Lisp_Object err_arg);
1163 to_char_id (Lisp_Object v, char* err_msg, Lisp_Object err_arg)
1169 else if (EQ (v, Qcompat))
1171 else if (EQ (v, Qisolated))
1173 else if (EQ (v, Qinitial))
1175 else if (EQ (v, Qmedial))
1177 else if (EQ (v, Qfinal))
1179 else if (EQ (v, Qvertical))
1181 else if (EQ (v, QnoBreak))
1183 else if (EQ (v, Qfraction))
1185 else if (EQ (v, Qsuper))
1187 else if (EQ (v, Qsub))
1189 else if (EQ (v, Qcircle))
1191 else if (EQ (v, Qsquare))
1193 else if (EQ (v, Qwide))
1195 else if (EQ (v, Qnarrow))
1197 else if (EQ (v, Qsmall))
1199 else if (EQ (v, Qfont))
1202 signal_simple_error (err_msg, err_arg);
1205 DEFUN ("get-composite-char", Fget_composite_char, 1, 1, 0, /*
1206 Return character corresponding with list.
1210 Lisp_Object base, modifier;
1214 signal_simple_error ("Invalid value for composition", list);
1217 while (!NILP (rest))
1222 signal_simple_error ("Invalid value for composition", list);
1223 modifier = Fcar (rest);
1225 base = Fcdr (Fassq (modifier,
1226 Fget_char_attribute (base, Qcomposition, Qnil)));
1231 DEFUN ("char-variants", Fchar_variants, 1, 1, 0, /*
1232 Return variants of CHARACTER.
1238 CHECK_CHAR (character);
1239 ret = Fget_char_attribute (character, Q_ucs_unified, Qnil);
1241 return Fcopy_list (ret);
1249 /* A char table maps from ranges of characters to values.
1251 Implementing a general data structure that maps from arbitrary
1252 ranges of numbers to values is tricky to do efficiently. As it
1253 happens, it should suffice (and is usually more convenient, anyway)
1254 when dealing with characters to restrict the sorts of ranges that
1255 can be assigned values, as follows:
1258 2) All characters in a charset.
1259 3) All characters in a particular row of a charset, where a "row"
1260 means all characters with the same first byte.
1261 4) A particular character in a charset.
1263 We use char tables to generalize the 256-element vectors now
1264 littering the Emacs code.
1266 Possible uses (all should be converted at some point):
1272 5) keyboard-translate-table?
1275 abstract type to generalize the Emacs vectors and Mule
1276 vectors-of-vectors goo.
1279 /************************************************************************/
1280 /* Char Table object */
1281 /************************************************************************/
1283 #if defined(MULE)&&!defined(UTF2000)
1286 mark_char_table_entry (Lisp_Object obj)
1288 Lisp_Char_Table_Entry *cte = XCHAR_TABLE_ENTRY (obj);
1291 for (i = 0; i < 96; i++)
1293 mark_object (cte->level2[i]);
1299 char_table_entry_equal (Lisp_Object obj1, Lisp_Object obj2, int depth)
1301 Lisp_Char_Table_Entry *cte1 = XCHAR_TABLE_ENTRY (obj1);
1302 Lisp_Char_Table_Entry *cte2 = XCHAR_TABLE_ENTRY (obj2);
1305 for (i = 0; i < 96; i++)
1306 if (!internal_equal (cte1->level2[i], cte2->level2[i], depth + 1))
1312 static unsigned long
1313 char_table_entry_hash (Lisp_Object obj, int depth)
1315 Lisp_Char_Table_Entry *cte = XCHAR_TABLE_ENTRY (obj);
1317 return internal_array_hash (cte->level2, 96, depth);
1320 static const struct lrecord_description char_table_entry_description[] = {
1321 { XD_LISP_OBJECT_ARRAY, offsetof (Lisp_Char_Table_Entry, level2), 96 },
1325 DEFINE_LRECORD_IMPLEMENTATION ("char-table-entry", char_table_entry,
1326 mark_char_table_entry, internal_object_printer,
1327 0, char_table_entry_equal,
1328 char_table_entry_hash,
1329 char_table_entry_description,
1330 Lisp_Char_Table_Entry);
1334 mark_char_table (Lisp_Object obj)
1336 Lisp_Char_Table *ct = XCHAR_TABLE (obj);
1339 mark_object (ct->table);
1340 mark_object (ct->name);
1341 #ifndef HAVE_LIBCHISE
1342 mark_object (ct->db);
1347 for (i = 0; i < NUM_ASCII_CHARS; i++)
1348 mark_object (ct->ascii[i]);
1350 for (i = 0; i < NUM_LEADING_BYTES; i++)
1351 mark_object (ct->level1[i]);
1355 return ct->default_value;
1357 return ct->mirror_table;
1361 /* WARNING: All functions of this nature need to be written extremely
1362 carefully to avoid crashes during GC. Cf. prune_specifiers()
1363 and prune_weak_hash_tables(). */
1366 prune_syntax_tables (void)
1368 Lisp_Object rest, prev = Qnil;
1370 for (rest = Vall_syntax_tables;
1372 rest = XCHAR_TABLE (rest)->next_table)
1374 if (! marked_p (rest))
1376 /* This table is garbage. Remove it from the list. */
1378 Vall_syntax_tables = XCHAR_TABLE (rest)->next_table;
1380 XCHAR_TABLE (prev)->next_table =
1381 XCHAR_TABLE (rest)->next_table;
1387 char_table_type_to_symbol (enum char_table_type type)
1392 case CHAR_TABLE_TYPE_GENERIC: return Qgeneric;
1393 case CHAR_TABLE_TYPE_SYNTAX: return Qsyntax;
1394 case CHAR_TABLE_TYPE_DISPLAY: return Qdisplay;
1395 case CHAR_TABLE_TYPE_CHAR: return Qchar;
1397 case CHAR_TABLE_TYPE_CATEGORY: return Qcategory;
1402 static enum char_table_type
1403 symbol_to_char_table_type (Lisp_Object symbol)
1405 CHECK_SYMBOL (symbol);
1407 if (EQ (symbol, Qgeneric)) return CHAR_TABLE_TYPE_GENERIC;
1408 if (EQ (symbol, Qsyntax)) return CHAR_TABLE_TYPE_SYNTAX;
1409 if (EQ (symbol, Qdisplay)) return CHAR_TABLE_TYPE_DISPLAY;
1410 if (EQ (symbol, Qchar)) return CHAR_TABLE_TYPE_CHAR;
1412 if (EQ (symbol, Qcategory)) return CHAR_TABLE_TYPE_CATEGORY;
1415 signal_simple_error ("Unrecognized char table type", symbol);
1416 return CHAR_TABLE_TYPE_GENERIC; /* not reached */
1421 print_chartab_range (Emchar first, Emchar last, Lisp_Object val,
1422 Lisp_Object printcharfun)
1426 write_c_string (" (", printcharfun);
1427 print_internal (make_char (first), printcharfun, 0);
1428 write_c_string (" ", printcharfun);
1429 print_internal (make_char (last), printcharfun, 0);
1430 write_c_string (") ", printcharfun);
1434 write_c_string (" ", printcharfun);
1435 print_internal (make_char (first), printcharfun, 0);
1436 write_c_string (" ", printcharfun);
1438 print_internal (val, printcharfun, 1);
1442 #if defined(MULE)&&!defined(UTF2000)
1445 print_chartab_charset_row (Lisp_Object charset,
1447 Lisp_Char_Table_Entry *cte,
1448 Lisp_Object printcharfun)
1451 Lisp_Object cat = Qunbound;
1454 for (i = 32; i < 128; i++)
1456 Lisp_Object pam = cte->level2[i - 32];
1468 print_chartab_range (MAKE_CHAR (charset, first, 0),
1469 MAKE_CHAR (charset, i - 1, 0),
1472 print_chartab_range (MAKE_CHAR (charset, row, first),
1473 MAKE_CHAR (charset, row, i - 1),
1483 print_chartab_range (MAKE_CHAR (charset, first, 0),
1484 MAKE_CHAR (charset, i - 1, 0),
1487 print_chartab_range (MAKE_CHAR (charset, row, first),
1488 MAKE_CHAR (charset, row, i - 1),
1494 print_chartab_two_byte_charset (Lisp_Object charset,
1495 Lisp_Char_Table_Entry *cte,
1496 Lisp_Object printcharfun)
1500 for (i = 32; i < 128; i++)
1502 Lisp_Object jen = cte->level2[i - 32];
1504 if (!CHAR_TABLE_ENTRYP (jen))
1508 write_c_string (" [", printcharfun);
1509 print_internal (XCHARSET_NAME (charset), printcharfun, 0);
1510 sprintf (buf, " %d] ", i);
1511 write_c_string (buf, printcharfun);
1512 print_internal (jen, printcharfun, 0);
1515 print_chartab_charset_row (charset, i, XCHAR_TABLE_ENTRY (jen),
1523 print_char_table (Lisp_Object obj, Lisp_Object printcharfun, int escapeflag)
1525 Lisp_Char_Table *ct = XCHAR_TABLE (obj);
1528 struct gcpro gcpro1, gcpro2;
1529 GCPRO2 (obj, printcharfun);
1531 write_c_string ("#s(char-table ", printcharfun);
1532 write_c_string (" ", printcharfun);
1533 write_c_string (string_data
1535 (XSYMBOL (char_table_type_to_symbol (ct->type)))),
1537 write_c_string ("\n ", printcharfun);
1538 print_internal (ct->default_value, printcharfun, escapeflag);
1539 for (i = 0; i < 256; i++)
1541 Lisp_Object elt = get_byte_table (ct->table, i);
1542 if (i != 0) write_c_string ("\n ", printcharfun);
1543 if (EQ (elt, Qunbound))
1544 write_c_string ("void", printcharfun);
1546 print_internal (elt, printcharfun, escapeflag);
1549 #else /* non UTF2000 */
1552 sprintf (buf, "#s(char-table type %s data (",
1553 string_data (symbol_name (XSYMBOL
1554 (char_table_type_to_symbol (ct->type)))));
1555 write_c_string (buf, printcharfun);
1557 /* Now write out the ASCII/Control-1 stuff. */
1561 Lisp_Object val = Qunbound;
1563 for (i = 0; i < NUM_ASCII_CHARS; i++)
1572 if (!EQ (ct->ascii[i], val))
1574 print_chartab_range (first, i - 1, val, printcharfun);
1581 print_chartab_range (first, i - 1, val, printcharfun);
1588 for (i = MIN_LEADING_BYTE; i < MIN_LEADING_BYTE + NUM_LEADING_BYTES;
1591 Lisp_Object ann = ct->level1[i - MIN_LEADING_BYTE];
1592 Lisp_Object charset = CHARSET_BY_LEADING_BYTE (i);
1594 if (!CHARSETP (charset) || i == LEADING_BYTE_ASCII
1595 || i == LEADING_BYTE_CONTROL_1)
1597 if (!CHAR_TABLE_ENTRYP (ann))
1599 write_c_string (" ", printcharfun);
1600 print_internal (XCHARSET_NAME (charset),
1602 write_c_string (" ", printcharfun);
1603 print_internal (ann, printcharfun, 0);
1607 Lisp_Char_Table_Entry *cte = XCHAR_TABLE_ENTRY (ann);
1608 if (XCHARSET_DIMENSION (charset) == 1)
1609 print_chartab_charset_row (charset, -1, cte, printcharfun);
1611 print_chartab_two_byte_charset (charset, cte, printcharfun);
1616 #endif /* non UTF2000 */
1618 write_c_string ("))", printcharfun);
1622 char_table_equal (Lisp_Object obj1, Lisp_Object obj2, int depth)
1624 Lisp_Char_Table *ct1 = XCHAR_TABLE (obj1);
1625 Lisp_Char_Table *ct2 = XCHAR_TABLE (obj2);
1628 if (CHAR_TABLE_TYPE (ct1) != CHAR_TABLE_TYPE (ct2))
1632 for (i = 0; i < 256; i++)
1634 if (!internal_equal (get_byte_table (ct1->table, i),
1635 get_byte_table (ct2->table, i), 0))
1639 for (i = 0; i < NUM_ASCII_CHARS; i++)
1640 if (!internal_equal (ct1->ascii[i], ct2->ascii[i], depth + 1))
1644 for (i = 0; i < NUM_LEADING_BYTES; i++)
1645 if (!internal_equal (ct1->level1[i], ct2->level1[i], depth + 1))
1648 #endif /* non UTF2000 */
1653 static unsigned long
1654 char_table_hash (Lisp_Object obj, int depth)
1656 Lisp_Char_Table *ct = XCHAR_TABLE (obj);
1658 return byte_table_hash (ct->table, depth + 1);
1660 unsigned long hashval = internal_array_hash (ct->ascii, NUM_ASCII_CHARS,
1663 hashval = HASH2 (hashval,
1664 internal_array_hash (ct->level1, NUM_LEADING_BYTES, depth));
1670 static const struct lrecord_description char_table_description[] = {
1672 { XD_LISP_OBJECT, offsetof(Lisp_Char_Table, table) },
1673 { XD_LISP_OBJECT, offsetof(Lisp_Char_Table, default_value) },
1674 { XD_LISP_OBJECT, offsetof(Lisp_Char_Table, name) },
1675 #ifndef HAVE_LIBCHISE
1676 { XD_LISP_OBJECT, offsetof(Lisp_Char_Table, db) },
1679 { XD_LISP_OBJECT_ARRAY, offsetof (Lisp_Char_Table, ascii), NUM_ASCII_CHARS },
1681 { XD_LISP_OBJECT_ARRAY, offsetof (Lisp_Char_Table, level1), NUM_LEADING_BYTES },
1685 { XD_LISP_OBJECT, offsetof (Lisp_Char_Table, mirror_table) },
1687 { XD_LO_LINK, offsetof (Lisp_Char_Table, next_table) },
1691 DEFINE_LRECORD_IMPLEMENTATION ("char-table", char_table,
1692 mark_char_table, print_char_table, 0,
1693 char_table_equal, char_table_hash,
1694 char_table_description,
1697 DEFUN ("char-table-p", Fchar_table_p, 1, 1, 0, /*
1698 Return non-nil if OBJECT is a char table.
1700 A char table is a table that maps characters (or ranges of characters)
1701 to values. Char tables are specialized for characters, only allowing
1702 particular sorts of ranges to be assigned values. Although this
1703 loses in generality, it makes for extremely fast (constant-time)
1704 lookups, and thus is feasible for applications that do an extremely
1705 large number of lookups (e.g. scanning a buffer for a character in
1706 a particular syntax, where a lookup in the syntax table must occur
1707 once per character).
1709 When Mule support exists, the types of ranges that can be assigned
1713 -- an entire charset
1714 -- a single row in a two-octet charset
1715 -- a single character
1717 When Mule support is not present, the types of ranges that can be
1721 -- a single character
1723 To create a char table, use `make-char-table'.
1724 To modify a char table, use `put-char-table' or `remove-char-table'.
1725 To retrieve the value for a particular character, use `get-char-table'.
1726 See also `map-char-table', `clear-char-table', `copy-char-table',
1727 `valid-char-table-type-p', `char-table-type-list',
1728 `valid-char-table-value-p', and `check-char-table-value'.
1732 return CHAR_TABLEP (object) ? Qt : Qnil;
1735 DEFUN ("char-table-type-list", Fchar_table_type_list, 0, 0, 0, /*
1736 Return a list of the recognized char table types.
1737 See `valid-char-table-type-p'.
1742 return list5 (Qchar, Qcategory, Qdisplay, Qgeneric, Qsyntax);
1744 return list4 (Qchar, Qdisplay, Qgeneric, Qsyntax);
1748 DEFUN ("valid-char-table-type-p", Fvalid_char_table_type_p, 1, 1, 0, /*
1749 Return t if TYPE if a recognized char table type.
1751 Each char table type is used for a different purpose and allows different
1752 sorts of values. The different char table types are
1755 Used for category tables, which specify the regexp categories
1756 that a character is in. The valid values are nil or a
1757 bit vector of 95 elements. Higher-level Lisp functions are
1758 provided for working with category tables. Currently categories
1759 and category tables only exist when Mule support is present.
1761 A generalized char table, for mapping from one character to
1762 another. Used for case tables, syntax matching tables,
1763 `keyboard-translate-table', etc. The valid values are characters.
1765 An even more generalized char table, for mapping from a
1766 character to anything.
1768 Used for display tables, which specify how a particular character
1769 is to appear when displayed. #### Not yet implemented.
1771 Used for syntax tables, which specify the syntax of a particular
1772 character. Higher-level Lisp functions are provided for
1773 working with syntax tables. The valid values are integers.
1778 return (EQ (type, Qchar) ||
1780 EQ (type, Qcategory) ||
1782 EQ (type, Qdisplay) ||
1783 EQ (type, Qgeneric) ||
1784 EQ (type, Qsyntax)) ? Qt : Qnil;
1787 DEFUN ("char-table-type", Fchar_table_type, 1, 1, 0, /*
1788 Return the type of CHAR-TABLE.
1789 See `valid-char-table-type-p'.
1793 CHECK_CHAR_TABLE (char_table);
1794 return char_table_type_to_symbol (XCHAR_TABLE (char_table)->type);
1798 fill_char_table (Lisp_Char_Table *ct, Lisp_Object value)
1801 ct->table = Qunbound;
1802 ct->default_value = value;
1807 for (i = 0; i < NUM_ASCII_CHARS; i++)
1808 ct->ascii[i] = value;
1810 for (i = 0; i < NUM_LEADING_BYTES; i++)
1811 ct->level1[i] = value;
1816 if (ct->type == CHAR_TABLE_TYPE_SYNTAX)
1817 update_syntax_table (ct);
1821 DEFUN ("reset-char-table", Freset_char_table, 1, 1, 0, /*
1822 Reset CHAR-TABLE to its default state.
1826 Lisp_Char_Table *ct;
1828 CHECK_CHAR_TABLE (char_table);
1829 ct = XCHAR_TABLE (char_table);
1833 case CHAR_TABLE_TYPE_CHAR:
1834 fill_char_table (ct, make_char (0));
1836 case CHAR_TABLE_TYPE_DISPLAY:
1837 case CHAR_TABLE_TYPE_GENERIC:
1839 case CHAR_TABLE_TYPE_CATEGORY:
1841 fill_char_table (ct, Qnil);
1844 case CHAR_TABLE_TYPE_SYNTAX:
1845 fill_char_table (ct, make_int (Sinherit));
1855 DEFUN ("make-char-table", Fmake_char_table, 1, 1, 0, /*
1856 Return a new, empty char table of type TYPE.
1857 Currently recognized types are 'char, 'category, 'display, 'generic,
1858 and 'syntax. See `valid-char-table-type-p'.
1862 Lisp_Char_Table *ct;
1864 enum char_table_type ty = symbol_to_char_table_type (type);
1866 ct = alloc_lcrecord_type (Lisp_Char_Table, &lrecord_char_table);
1869 if (ty == CHAR_TABLE_TYPE_SYNTAX)
1871 ct->mirror_table = Fmake_char_table (Qgeneric);
1872 fill_char_table (XCHAR_TABLE (ct->mirror_table),
1876 ct->mirror_table = Qnil;
1879 #ifndef HAVE_LIBCHISE
1883 ct->next_table = Qnil;
1884 XSETCHAR_TABLE (obj, ct);
1885 if (ty == CHAR_TABLE_TYPE_SYNTAX)
1887 ct->next_table = Vall_syntax_tables;
1888 Vall_syntax_tables = obj;
1890 Freset_char_table (obj);
1894 #if defined(MULE)&&!defined(UTF2000)
1897 make_char_table_entry (Lisp_Object initval)
1901 Lisp_Char_Table_Entry *cte =
1902 alloc_lcrecord_type (Lisp_Char_Table_Entry, &lrecord_char_table_entry);
1904 for (i = 0; i < 96; i++)
1905 cte->level2[i] = initval;
1907 XSETCHAR_TABLE_ENTRY (obj, cte);
1912 copy_char_table_entry (Lisp_Object entry)
1914 Lisp_Char_Table_Entry *cte = XCHAR_TABLE_ENTRY (entry);
1917 Lisp_Char_Table_Entry *ctenew =
1918 alloc_lcrecord_type (Lisp_Char_Table_Entry, &lrecord_char_table_entry);
1920 for (i = 0; i < 96; i++)
1922 Lisp_Object new = cte->level2[i];
1923 if (CHAR_TABLE_ENTRYP (new))
1924 ctenew->level2[i] = copy_char_table_entry (new);
1926 ctenew->level2[i] = new;
1929 XSETCHAR_TABLE_ENTRY (obj, ctenew);
1935 DEFUN ("copy-char-table", Fcopy_char_table, 1, 1, 0, /*
1936 Return a new char table which is a copy of CHAR-TABLE.
1937 It will contain the same values for the same characters and ranges
1938 as CHAR-TABLE. The values will not themselves be copied.
1942 Lisp_Char_Table *ct, *ctnew;
1948 CHECK_CHAR_TABLE (char_table);
1949 ct = XCHAR_TABLE (char_table);
1950 ctnew = alloc_lcrecord_type (Lisp_Char_Table, &lrecord_char_table);
1951 ctnew->type = ct->type;
1953 ctnew->default_value = ct->default_value;
1954 /* [tomo:2002-01-21] Perhaps this code seems wrong */
1955 ctnew->name = ct->name;
1956 #ifndef HAVE_LIBCHISE
1960 if (UINT8_BYTE_TABLE_P (ct->table))
1962 ctnew->table = copy_uint8_byte_table (ct->table);
1964 else if (UINT16_BYTE_TABLE_P (ct->table))
1966 ctnew->table = copy_uint16_byte_table (ct->table);
1968 else if (BYTE_TABLE_P (ct->table))
1970 ctnew->table = copy_byte_table (ct->table);
1972 else if (!UNBOUNDP (ct->table))
1973 ctnew->table = ct->table;
1974 #else /* non UTF2000 */
1976 for (i = 0; i < NUM_ASCII_CHARS; i++)
1978 Lisp_Object new = ct->ascii[i];
1980 assert (! (CHAR_TABLE_ENTRYP (new)));
1982 ctnew->ascii[i] = new;
1987 for (i = 0; i < NUM_LEADING_BYTES; i++)
1989 Lisp_Object new = ct->level1[i];
1990 if (CHAR_TABLE_ENTRYP (new))
1991 ctnew->level1[i] = copy_char_table_entry (new);
1993 ctnew->level1[i] = new;
1997 #endif /* non UTF2000 */
2000 if (CHAR_TABLEP (ct->mirror_table))
2001 ctnew->mirror_table = Fcopy_char_table (ct->mirror_table);
2003 ctnew->mirror_table = ct->mirror_table;
2005 ctnew->next_table = Qnil;
2006 XSETCHAR_TABLE (obj, ctnew);
2007 if (ctnew->type == CHAR_TABLE_TYPE_SYNTAX)
2009 ctnew->next_table = Vall_syntax_tables;
2010 Vall_syntax_tables = obj;
2015 INLINE_HEADER int XCHARSET_CELL_RANGE (Lisp_Object ccs);
2017 XCHARSET_CELL_RANGE (Lisp_Object ccs)
2019 switch (XCHARSET_CHARS (ccs))
2022 return (33 << 8) | 126;
2024 return (32 << 8) | 127;
2027 return (0 << 8) | 127;
2029 return (0 << 8) | 255;
2041 decode_char_table_range (Lisp_Object range, struct chartab_range *outrange)
2044 outrange->type = CHARTAB_RANGE_ALL;
2046 else if (EQ (range, Qnil))
2047 outrange->type = CHARTAB_RANGE_DEFAULT;
2049 else if (CHAR_OR_CHAR_INTP (range))
2051 outrange->type = CHARTAB_RANGE_CHAR;
2052 outrange->ch = XCHAR_OR_CHAR_INT (range);
2056 signal_simple_error ("Range must be t or a character", range);
2058 else if (VECTORP (range))
2060 Lisp_Vector *vec = XVECTOR (range);
2061 Lisp_Object *elts = vector_data (vec);
2062 int cell_min, cell_max;
2064 outrange->type = CHARTAB_RANGE_ROW;
2065 outrange->charset = Fget_charset (elts[0]);
2066 CHECK_INT (elts[1]);
2067 outrange->row = XINT (elts[1]);
2068 if (XCHARSET_DIMENSION (outrange->charset) < 2)
2069 signal_simple_error ("Charset in row vector must be multi-byte",
2073 int ret = XCHARSET_CELL_RANGE (outrange->charset);
2075 cell_min = ret >> 8;
2076 cell_max = ret & 0xFF;
2078 if (XCHARSET_DIMENSION (outrange->charset) == 2)
2079 check_int_range (outrange->row, cell_min, cell_max);
2081 else if (XCHARSET_DIMENSION (outrange->charset) == 3)
2083 check_int_range (outrange->row >> 8 , cell_min, cell_max);
2084 check_int_range (outrange->row & 0xFF, cell_min, cell_max);
2086 else if (XCHARSET_DIMENSION (outrange->charset) == 4)
2088 check_int_range ( outrange->row >> 16 , cell_min, cell_max);
2089 check_int_range ((outrange->row >> 8) & 0xFF, cell_min, cell_max);
2090 check_int_range ( outrange->row & 0xFF, cell_min, cell_max);
2098 if (!CHARSETP (range) && !SYMBOLP (range))
2100 ("Char table range must be t, charset, char, or vector", range);
2101 outrange->type = CHARTAB_RANGE_CHARSET;
2102 outrange->charset = Fget_charset (range);
2107 #if defined(MULE)&&!defined(UTF2000)
2109 /* called from CHAR_TABLE_VALUE(). */
2111 get_non_ascii_char_table_value (Lisp_Char_Table *ct, Charset_ID leading_byte,
2116 Lisp_Object charset;
2118 Lisp_Object charset = CHARSET_BY_LEADING_BYTE (leading_byte);
2123 BREAKUP_CHAR (c, charset, byte1, byte2);
2125 BREAKUP_CHAR_1_UNSAFE (c, charset, byte1, byte2);
2127 val = ct->level1[leading_byte - MIN_LEADING_BYTE];
2128 if (CHAR_TABLE_ENTRYP (val))
2130 Lisp_Char_Table_Entry *cte = XCHAR_TABLE_ENTRY (val);
2131 val = cte->level2[byte1 - 32];
2132 if (CHAR_TABLE_ENTRYP (val))
2134 cte = XCHAR_TABLE_ENTRY (val);
2135 assert (byte2 >= 32);
2136 val = cte->level2[byte2 - 32];
2137 assert (!CHAR_TABLE_ENTRYP (val));
2147 get_char_table (Emchar ch, Lisp_Char_Table *ct)
2151 Lisp_Object ret = get_char_id_table (ct, ch);
2156 if (EQ (CHAR_TABLE_NAME (ct), Qdowncase))
2157 ret = Fget_char_attribute (make_char (ch), Q_lowercase, Qnil);
2158 else if (EQ (CHAR_TABLE_NAME (ct), Qflippedcase))
2159 ret = Fget_char_attribute (make_char (ch), Q_uppercase, Qnil);
2164 ret = Ffind_char (ret);
2172 Lisp_Object charset;
2176 BREAKUP_CHAR (ch, charset, byte1, byte2);
2178 if (EQ (charset, Vcharset_ascii))
2179 val = ct->ascii[byte1];
2180 else if (EQ (charset, Vcharset_control_1))
2181 val = ct->ascii[byte1 + 128];
2184 int lb = XCHARSET_LEADING_BYTE (charset) - MIN_LEADING_BYTE;
2185 val = ct->level1[lb];
2186 if (CHAR_TABLE_ENTRYP (val))
2188 Lisp_Char_Table_Entry *cte = XCHAR_TABLE_ENTRY (val);
2189 val = cte->level2[byte1 - 32];
2190 if (CHAR_TABLE_ENTRYP (val))
2192 cte = XCHAR_TABLE_ENTRY (val);
2193 assert (byte2 >= 32);
2194 val = cte->level2[byte2 - 32];
2195 assert (!CHAR_TABLE_ENTRYP (val));
2202 #else /* not MULE */
2203 return ct->ascii[(unsigned char)ch];
2204 #endif /* not MULE */
2208 DEFUN ("get-char-table", Fget_char_table, 2, 2, 0, /*
2209 Find value for CHARACTER in CHAR-TABLE.
2211 (character, char_table))
2213 CHECK_CHAR_TABLE (char_table);
2214 CHECK_CHAR_COERCE_INT (character);
2216 return get_char_table (XCHAR (character), XCHAR_TABLE (char_table));
2219 DEFUN ("get-range-char-table", Fget_range_char_table, 2, 3, 0, /*
2220 Find value for a range in CHAR-TABLE.
2221 If there is more than one value, return MULTI (defaults to nil).
2223 (range, char_table, multi))
2225 Lisp_Char_Table *ct;
2226 struct chartab_range rainj;
2228 if (CHAR_OR_CHAR_INTP (range))
2229 return Fget_char_table (range, char_table);
2230 CHECK_CHAR_TABLE (char_table);
2231 ct = XCHAR_TABLE (char_table);
2233 decode_char_table_range (range, &rainj);
2236 case CHARTAB_RANGE_ALL:
2239 if (UINT8_BYTE_TABLE_P (ct->table))
2241 else if (UINT16_BYTE_TABLE_P (ct->table))
2243 else if (BYTE_TABLE_P (ct->table))
2247 #else /* non UTF2000 */
2249 Lisp_Object first = ct->ascii[0];
2251 for (i = 1; i < NUM_ASCII_CHARS; i++)
2252 if (!EQ (first, ct->ascii[i]))
2256 for (i = MIN_LEADING_BYTE; i < MIN_LEADING_BYTE + NUM_LEADING_BYTES;
2259 if (!CHARSETP (CHARSET_BY_LEADING_BYTE (i))
2260 || i == LEADING_BYTE_ASCII
2261 || i == LEADING_BYTE_CONTROL_1)
2263 if (!EQ (first, ct->level1[i - MIN_LEADING_BYTE]))
2269 #endif /* non UTF2000 */
2273 case CHARTAB_RANGE_CHARSET:
2277 if (EQ (rainj.charset, Vcharset_ascii))
2280 Lisp_Object first = ct->ascii[0];
2282 for (i = 1; i < 128; i++)
2283 if (!EQ (first, ct->ascii[i]))
2288 if (EQ (rainj.charset, Vcharset_control_1))
2291 Lisp_Object first = ct->ascii[128];
2293 for (i = 129; i < 160; i++)
2294 if (!EQ (first, ct->ascii[i]))
2300 Lisp_Object val = ct->level1[XCHARSET_LEADING_BYTE (rainj.charset) -
2302 if (CHAR_TABLE_ENTRYP (val))
2308 case CHARTAB_RANGE_ROW:
2313 Lisp_Object val = ct->level1[XCHARSET_LEADING_BYTE (rainj.charset) -
2315 if (!CHAR_TABLE_ENTRYP (val))
2317 val = XCHAR_TABLE_ENTRY (val)->level2[rainj.row - 32];
2318 if (CHAR_TABLE_ENTRYP (val))
2322 #endif /* not UTF2000 */
2323 #endif /* not MULE */
2329 return Qnil; /* not reached */
2333 check_valid_char_table_value (Lisp_Object value, enum char_table_type type,
2334 Error_behavior errb)
2338 case CHAR_TABLE_TYPE_SYNTAX:
2339 if (!ERRB_EQ (errb, ERROR_ME))
2340 return INTP (value) || (CONSP (value) && INTP (XCAR (value))
2341 && CHAR_OR_CHAR_INTP (XCDR (value)));
2344 Lisp_Object cdr = XCDR (value);
2345 CHECK_INT (XCAR (value));
2346 CHECK_CHAR_COERCE_INT (cdr);
2353 case CHAR_TABLE_TYPE_CATEGORY:
2354 if (!ERRB_EQ (errb, ERROR_ME))
2355 return CATEGORY_TABLE_VALUEP (value);
2356 CHECK_CATEGORY_TABLE_VALUE (value);
2360 case CHAR_TABLE_TYPE_GENERIC:
2363 case CHAR_TABLE_TYPE_DISPLAY:
2365 maybe_signal_simple_error ("Display char tables not yet implemented",
2366 value, Qchar_table, errb);
2369 case CHAR_TABLE_TYPE_CHAR:
2370 if (!ERRB_EQ (errb, ERROR_ME))
2371 return CHAR_OR_CHAR_INTP (value);
2372 CHECK_CHAR_COERCE_INT (value);
2379 return 0; /* not reached */
2383 canonicalize_char_table_value (Lisp_Object value, enum char_table_type type)
2387 case CHAR_TABLE_TYPE_SYNTAX:
2390 Lisp_Object car = XCAR (value);
2391 Lisp_Object cdr = XCDR (value);
2392 CHECK_CHAR_COERCE_INT (cdr);
2393 return Fcons (car, cdr);
2396 case CHAR_TABLE_TYPE_CHAR:
2397 CHECK_CHAR_COERCE_INT (value);
2405 DEFUN ("valid-char-table-value-p", Fvalid_char_table_value_p, 2, 2, 0, /*
2406 Return non-nil if VALUE is a valid value for CHAR-TABLE-TYPE.
2408 (value, char_table_type))
2410 enum char_table_type type = symbol_to_char_table_type (char_table_type);
2412 return check_valid_char_table_value (value, type, ERROR_ME_NOT) ? Qt : Qnil;
2415 DEFUN ("check-valid-char-table-value", Fcheck_valid_char_table_value, 2, 2, 0, /*
2416 Signal an error if VALUE is not a valid value for CHAR-TABLE-TYPE.
2418 (value, char_table_type))
2420 enum char_table_type type = symbol_to_char_table_type (char_table_type);
2422 check_valid_char_table_value (value, type, ERROR_ME);
2427 Lisp_Char_Table* char_attribute_table_to_put;
2428 Lisp_Object Qput_char_table_map_function;
2429 Lisp_Object value_to_put;
2431 DEFUN ("put-char-table-map-function",
2432 Fput_char_table_map_function, 2, 2, 0, /*
2433 For internal use. Don't use it.
2437 put_char_id_table_0 (char_attribute_table_to_put,
2438 XCHAR (c), value_to_put);
2443 /* Assign VAL to all characters in RANGE in char table CT. */
2446 put_char_table (Lisp_Char_Table *ct, struct chartab_range *range,
2449 switch (range->type)
2451 case CHARTAB_RANGE_ALL:
2452 /* printf ("put-char-table: range = all\n"); */
2453 fill_char_table (ct, val);
2454 return; /* avoid the duplicate call to update_syntax_table() below,
2455 since fill_char_table() also did that. */
2458 case CHARTAB_RANGE_DEFAULT:
2459 ct->default_value = val;
2464 case CHARTAB_RANGE_CHARSET:
2467 Lisp_Object encoding_table = XCHARSET_ENCODING_TABLE (range->charset);
2469 /* printf ("put-char-table: range = charset: %d\n",
2470 XCHARSET_LEADING_BYTE (range->charset));
2472 if ( CHAR_TABLEP (encoding_table) )
2474 Lisp_Object mother = XCHARSET_MOTHER (range->charset);
2476 char_attribute_table_to_put = ct;
2478 Fmap_char_attribute (Qput_char_table_map_function,
2479 XCHAR_TABLE_NAME (encoding_table),
2481 if ( CHARSETP (mother) )
2483 struct chartab_range r;
2485 r.type = CHARTAB_RANGE_CHARSET;
2487 put_char_table (ct, &r, val);
2495 for (c = 0; c < 1 << 24; c++)
2497 if ( charset_code_point (range->charset, c) >= 0 )
2498 put_char_id_table_0 (ct, c, val);
2504 if (EQ (range->charset, Vcharset_ascii))
2507 for (i = 0; i < 128; i++)
2510 else if (EQ (range->charset, Vcharset_control_1))
2513 for (i = 128; i < 160; i++)
2518 int lb = XCHARSET_LEADING_BYTE (range->charset) - MIN_LEADING_BYTE;
2519 ct->level1[lb] = val;
2524 case CHARTAB_RANGE_ROW:
2527 int cell_min, cell_max, i;
2529 i = XCHARSET_CELL_RANGE (range->charset);
2531 cell_max = i & 0xFF;
2532 for (i = cell_min; i <= cell_max; i++)
2535 = DECODE_CHAR (range->charset, (range->row << 8) | i, 0);
2537 if ( charset_code_point (range->charset, ch, 0) >= 0 )
2538 put_char_id_table_0 (ct, ch, val);
2543 Lisp_Char_Table_Entry *cte;
2544 int lb = XCHARSET_LEADING_BYTE (range->charset) - MIN_LEADING_BYTE;
2545 /* make sure that there is a separate entry for the row. */
2546 if (!CHAR_TABLE_ENTRYP (ct->level1[lb]))
2547 ct->level1[lb] = make_char_table_entry (ct->level1[lb]);
2548 cte = XCHAR_TABLE_ENTRY (ct->level1[lb]);
2549 cte->level2[range->row - 32] = val;
2551 #endif /* not UTF2000 */
2555 case CHARTAB_RANGE_CHAR:
2557 /* printf ("put-char-table: range = char: 0x%x\n", range->ch); */
2558 put_char_id_table_0 (ct, range->ch, val);
2562 Lisp_Object charset;
2565 BREAKUP_CHAR (range->ch, charset, byte1, byte2);
2566 if (EQ (charset, Vcharset_ascii))
2567 ct->ascii[byte1] = val;
2568 else if (EQ (charset, Vcharset_control_1))
2569 ct->ascii[byte1 + 128] = val;
2572 Lisp_Char_Table_Entry *cte;
2573 int lb = XCHARSET_LEADING_BYTE (charset) - MIN_LEADING_BYTE;
2574 /* make sure that there is a separate entry for the row. */
2575 if (!CHAR_TABLE_ENTRYP (ct->level1[lb]))
2576 ct->level1[lb] = make_char_table_entry (ct->level1[lb]);
2577 cte = XCHAR_TABLE_ENTRY (ct->level1[lb]);
2578 /* now CTE is a char table entry for the charset;
2579 each entry is for a single row (or character of
2580 a one-octet charset). */
2581 if (XCHARSET_DIMENSION (charset) == 1)
2582 cte->level2[byte1 - 32] = val;
2585 /* assigning to one character in a two-octet charset. */
2586 /* make sure that the charset row contains a separate
2587 entry for each character. */
2588 if (!CHAR_TABLE_ENTRYP (cte->level2[byte1 - 32]))
2589 cte->level2[byte1 - 32] =
2590 make_char_table_entry (cte->level2[byte1 - 32]);
2591 cte = XCHAR_TABLE_ENTRY (cte->level2[byte1 - 32]);
2592 cte->level2[byte2 - 32] = val;
2596 #else /* not MULE */
2597 ct->ascii[(unsigned char) (range->ch)] = val;
2599 #endif /* not MULE */
2603 if (ct->type == CHAR_TABLE_TYPE_SYNTAX)
2604 update_syntax_table (ct);
2608 DEFUN ("put-char-table", Fput_char_table, 3, 3, 0, /*
2609 Set the value for chars in RANGE to be VALUE in CHAR-TABLE.
2611 RANGE specifies one or more characters to be affected and should be
2612 one of the following:
2614 -- t (all characters are affected)
2615 -- A charset (only allowed when Mule support is present)
2616 -- A vector of two elements: a two-octet charset and a row number
2617 (only allowed when Mule support is present)
2618 -- A single character
2620 VALUE must be a value appropriate for the type of CHAR-TABLE.
2621 See `valid-char-table-type-p'.
2623 (range, value, char_table))
2625 Lisp_Char_Table *ct;
2626 struct chartab_range rainj;
2628 CHECK_CHAR_TABLE (char_table);
2629 ct = XCHAR_TABLE (char_table);
2630 check_valid_char_table_value (value, ct->type, ERROR_ME);
2631 decode_char_table_range (range, &rainj);
2632 value = canonicalize_char_table_value (value, ct->type);
2633 put_char_table (ct, &rainj, value);
2638 /* Map FN over the ASCII chars in CT. */
2641 map_over_charset_ascii (Lisp_Char_Table *ct,
2642 int (*fn) (struct chartab_range *range,
2643 Lisp_Object val, void *arg),
2646 struct chartab_range rainj;
2655 rainj.type = CHARTAB_RANGE_CHAR;
2657 for (i = start, retval = 0; i < stop && retval == 0; i++)
2659 rainj.ch = (Emchar) i;
2660 retval = (fn) (&rainj, ct->ascii[i], arg);
2668 /* Map FN over the Control-1 chars in CT. */
2671 map_over_charset_control_1 (Lisp_Char_Table *ct,
2672 int (*fn) (struct chartab_range *range,
2673 Lisp_Object val, void *arg),
2676 struct chartab_range rainj;
2679 int stop = start + 32;
2681 rainj.type = CHARTAB_RANGE_CHAR;
2683 for (i = start, retval = 0; i < stop && retval == 0; i++)
2685 rainj.ch = (Emchar) (i);
2686 retval = (fn) (&rainj, ct->ascii[i], arg);
2692 /* Map FN over the row ROW of two-byte charset CHARSET.
2693 There must be a separate value for that row in the char table.
2694 CTE specifies the char table entry for CHARSET. */
2697 map_over_charset_row (Lisp_Char_Table_Entry *cte,
2698 Lisp_Object charset, int row,
2699 int (*fn) (struct chartab_range *range,
2700 Lisp_Object val, void *arg),
2703 Lisp_Object val = cte->level2[row - 32];
2705 if (!CHAR_TABLE_ENTRYP (val))
2707 struct chartab_range rainj;
2709 rainj.type = CHARTAB_RANGE_ROW;
2710 rainj.charset = charset;
2712 return (fn) (&rainj, val, arg);
2716 struct chartab_range rainj;
2718 int charset94_p = (XCHARSET_CHARS (charset) == 94);
2719 int start = charset94_p ? 33 : 32;
2720 int stop = charset94_p ? 127 : 128;
2722 cte = XCHAR_TABLE_ENTRY (val);
2724 rainj.type = CHARTAB_RANGE_CHAR;
2726 for (i = start, retval = 0; i < stop && retval == 0; i++)
2728 rainj.ch = MAKE_CHAR (charset, row, i);
2729 retval = (fn) (&rainj, cte->level2[i - 32], arg);
2737 map_over_other_charset (Lisp_Char_Table *ct, Charset_ID lb,
2738 int (*fn) (struct chartab_range *range,
2739 Lisp_Object val, void *arg),
2742 Lisp_Object val = ct->level1[lb - MIN_LEADING_BYTE];
2743 Lisp_Object charset = CHARSET_BY_LEADING_BYTE (lb);
2745 if (!CHARSETP (charset)
2746 || lb == LEADING_BYTE_ASCII
2747 || lb == LEADING_BYTE_CONTROL_1)
2750 if (!CHAR_TABLE_ENTRYP (val))
2752 struct chartab_range rainj;
2754 rainj.type = CHARTAB_RANGE_CHARSET;
2755 rainj.charset = charset;
2756 return (fn) (&rainj, val, arg);
2760 Lisp_Char_Table_Entry *cte = XCHAR_TABLE_ENTRY (val);
2761 int charset94_p = (XCHARSET_CHARS (charset) == 94);
2762 int start = charset94_p ? 33 : 32;
2763 int stop = charset94_p ? 127 : 128;
2766 if (XCHARSET_DIMENSION (charset) == 1)
2768 struct chartab_range rainj;
2769 rainj.type = CHARTAB_RANGE_CHAR;
2771 for (i = start, retval = 0; i < stop && retval == 0; i++)
2773 rainj.ch = MAKE_CHAR (charset, i, 0);
2774 retval = (fn) (&rainj, cte->level2[i - 32], arg);
2779 for (i = start, retval = 0; i < stop && retval == 0; i++)
2780 retval = map_over_charset_row (cte, charset, i, fn, arg);
2788 #endif /* not UTF2000 */
2791 struct map_char_table_for_charset_arg
2793 int (*fn) (struct chartab_range *range, Lisp_Object val, void *arg);
2794 Lisp_Char_Table *ct;
2799 map_char_table_for_charset_fun (struct chartab_range *range,
2800 Lisp_Object val, void *arg)
2802 struct map_char_table_for_charset_arg *closure =
2803 (struct map_char_table_for_charset_arg *) arg;
2806 switch (range->type)
2808 case CHARTAB_RANGE_ALL:
2811 case CHARTAB_RANGE_DEFAULT:
2814 case CHARTAB_RANGE_CHARSET:
2817 case CHARTAB_RANGE_ROW:
2820 case CHARTAB_RANGE_CHAR:
2821 ret = get_char_table (range->ch, closure->ct);
2822 if (!UNBOUNDP (ret))
2823 return (closure->fn) (range, ret, closure->arg);
2835 /* Map FN (with client data ARG) over range RANGE in char table CT.
2836 Mapping stops the first time FN returns non-zero, and that value
2837 becomes the return value of map_char_table(). */
2840 map_char_table (Lisp_Char_Table *ct,
2841 struct chartab_range *range,
2842 int (*fn) (struct chartab_range *range,
2843 Lisp_Object val, void *arg),
2846 switch (range->type)
2848 case CHARTAB_RANGE_ALL:
2850 if (!UNBOUNDP (ct->default_value))
2852 struct chartab_range rainj;
2855 rainj.type = CHARTAB_RANGE_DEFAULT;
2856 retval = (fn) (&rainj, ct->default_value, arg);
2860 if (UINT8_BYTE_TABLE_P (ct->table))
2861 return map_over_uint8_byte_table (XUINT8_BYTE_TABLE(ct->table), ct,
2863 else if (UINT16_BYTE_TABLE_P (ct->table))
2864 return map_over_uint16_byte_table (XUINT16_BYTE_TABLE(ct->table), ct,
2866 else if (BYTE_TABLE_P (ct->table))
2867 return map_over_byte_table (XBYTE_TABLE(ct->table), ct,
2869 else if (EQ (ct->table, Qunloaded))
2872 struct chartab_range rainj;
2875 Emchar c1 = c + unit;
2878 rainj.type = CHARTAB_RANGE_CHAR;
2880 for (retval = 0; c < c1 && retval == 0; c++)
2882 Lisp_Object ret = get_char_id_table (ct, c);
2884 if (!UNBOUNDP (ret))
2887 retval = (fn) (&rainj, ct->table, arg);
2892 ct->table = Qunbound;
2895 else if (!UNBOUNDP (ct->table))
2896 return (fn) (range, ct->table, arg);
2902 retval = map_over_charset_ascii (ct, fn, arg);
2906 retval = map_over_charset_control_1 (ct, fn, arg);
2911 Charset_ID start = MIN_LEADING_BYTE;
2912 Charset_ID stop = start + NUM_LEADING_BYTES;
2914 for (i = start, retval = 0; i < stop && retval == 0; i++)
2916 retval = map_over_other_charset (ct, i, fn, arg);
2925 case CHARTAB_RANGE_DEFAULT:
2926 if (!UNBOUNDP (ct->default_value))
2927 return (fn) (range, ct->default_value, arg);
2932 case CHARTAB_RANGE_CHARSET:
2935 Lisp_Object encoding_table
2936 = XCHARSET_ENCODING_TABLE (range->charset);
2938 if (!NILP (encoding_table))
2940 struct chartab_range rainj;
2941 struct map_char_table_for_charset_arg mcarg;
2944 if (XCHAR_TABLE_UNLOADED(encoding_table))
2945 Fload_char_attribute_table (XCHAR_TABLE_NAME (encoding_table));
2950 rainj.type = CHARTAB_RANGE_ALL;
2951 return map_char_table (XCHAR_TABLE(encoding_table),
2953 &map_char_table_for_charset_fun,
2959 return map_over_other_charset (ct,
2960 XCHARSET_LEADING_BYTE (range->charset),
2964 case CHARTAB_RANGE_ROW:
2967 int cell_min, cell_max, i;
2969 struct chartab_range rainj;
2971 i = XCHARSET_CELL_RANGE (range->charset);
2973 cell_max = i & 0xFF;
2974 rainj.type = CHARTAB_RANGE_CHAR;
2975 for (retval =0, i = cell_min; i <= cell_max && retval == 0; i++)
2978 = DECODE_CHAR (range->charset, (range->row << 8) | i, 0);
2980 if ( charset_code_point (range->charset, ch, 0) >= 0 )
2983 = get_byte_table (get_byte_table
2987 (unsigned char)(ch >> 24)),
2988 (unsigned char) (ch >> 16)),
2989 (unsigned char) (ch >> 8)),
2990 (unsigned char) ch);
2993 val = ct->default_value;
2995 retval = (fn) (&rainj, val, arg);
3002 Lisp_Object val = ct->level1[XCHARSET_LEADING_BYTE (range->charset)
3003 - MIN_LEADING_BYTE];
3004 if (!CHAR_TABLE_ENTRYP (val))
3006 struct chartab_range rainj;
3008 rainj.type = CHARTAB_RANGE_ROW;
3009 rainj.charset = range->charset;
3010 rainj.row = range->row;
3011 return (fn) (&rainj, val, arg);
3014 return map_over_charset_row (XCHAR_TABLE_ENTRY (val),
3015 range->charset, range->row,
3018 #endif /* not UTF2000 */
3021 case CHARTAB_RANGE_CHAR:
3023 Emchar ch = range->ch;
3024 Lisp_Object val = CHAR_TABLE_VALUE_UNSAFE (ct, ch);
3026 if (!UNBOUNDP (val))
3028 struct chartab_range rainj;
3030 rainj.type = CHARTAB_RANGE_CHAR;
3032 return (fn) (&rainj, val, arg);
3044 struct slow_map_char_table_arg
3046 Lisp_Object function;
3051 slow_map_char_table_fun (struct chartab_range *range,
3052 Lisp_Object val, void *arg)
3054 Lisp_Object ranjarg = Qnil;
3055 struct slow_map_char_table_arg *closure =
3056 (struct slow_map_char_table_arg *) arg;
3058 switch (range->type)
3060 case CHARTAB_RANGE_ALL:
3065 case CHARTAB_RANGE_DEFAULT:
3071 case CHARTAB_RANGE_CHARSET:
3072 ranjarg = XCHARSET_NAME (range->charset);
3075 case CHARTAB_RANGE_ROW:
3076 ranjarg = vector2 (XCHARSET_NAME (range->charset),
3077 make_int (range->row));
3080 case CHARTAB_RANGE_CHAR:
3081 ranjarg = make_char (range->ch);
3087 closure->retval = call2 (closure->function, ranjarg, val);
3088 return !NILP (closure->retval);
3091 DEFUN ("map-char-table", Fmap_char_table, 2, 3, 0, /*
3092 Map FUNCTION over entries in CHAR-TABLE, calling it with two args,
3093 each key and value in the table.
3095 RANGE specifies a subrange to map over and is in the same format as
3096 the RANGE argument to `put-range-table'. If omitted or t, it defaults to
3099 (function, char_table, range))
3101 Lisp_Char_Table *ct;
3102 struct slow_map_char_table_arg slarg;
3103 struct gcpro gcpro1, gcpro2;
3104 struct chartab_range rainj;
3106 CHECK_CHAR_TABLE (char_table);
3107 ct = XCHAR_TABLE (char_table);
3110 decode_char_table_range (range, &rainj);
3111 slarg.function = function;
3112 slarg.retval = Qnil;
3113 GCPRO2 (slarg.function, slarg.retval);
3114 map_char_table (ct, &rainj, slow_map_char_table_fun, &slarg);
3117 return slarg.retval;
3121 /************************************************************************/
3122 /* Character Attributes */
3123 /************************************************************************/
3127 Lisp_Object Vchar_attribute_hash_table;
3129 /* We store the char-attributes in hash tables with the names as the
3130 key and the actual char-id-table object as the value. Occasionally
3131 we need to use them in a list format. These routines provide us
3133 struct char_attribute_list_closure
3135 Lisp_Object *char_attribute_list;
3139 add_char_attribute_to_list_mapper (Lisp_Object key, Lisp_Object value,
3140 void *char_attribute_list_closure)
3142 /* This function can GC */
3143 struct char_attribute_list_closure *calcl
3144 = (struct char_attribute_list_closure*) char_attribute_list_closure;
3145 Lisp_Object *char_attribute_list = calcl->char_attribute_list;
3147 *char_attribute_list = Fcons (key, *char_attribute_list);
3151 #ifdef HAVE_LIBCHISE
3153 char_attribute_list_reset_map_func (CHISE_DS *ds, unsigned char *name)
3155 Fmount_char_attribute_table (intern (name));
3159 DEFUN ("char-attribute-list", Fchar_attribute_list, 0, 1, 0, /*
3160 Return the list of all existing character attributes except coded-charsets.
3164 DEFUN ("char-attribute-list", Fchar_attribute_list, 0, 0, 0, /*
3165 Return the list of all existing character attributes except coded-charsets.
3170 Lisp_Object char_attribute_list = Qnil;
3171 struct gcpro gcpro1;
3172 struct char_attribute_list_closure char_attribute_list_closure;
3174 #ifdef HAVE_LIBCHISE
3177 open_chise_data_source_maybe ();
3178 chise_ds_foreach_char_feature_name
3179 (default_chise_data_source, &char_attribute_list_reset_map_func);
3182 GCPRO1 (char_attribute_list);
3183 char_attribute_list_closure.char_attribute_list = &char_attribute_list;
3184 elisp_maphash (add_char_attribute_to_list_mapper,
3185 Vchar_attribute_hash_table,
3186 &char_attribute_list_closure);
3188 return char_attribute_list;
3191 DEFUN ("find-char-attribute-table", Ffind_char_attribute_table, 1, 1, 0, /*
3192 Return char-id-table corresponding to ATTRIBUTE.
3196 return Fgethash (attribute, Vchar_attribute_hash_table, Qnil);
3200 /* We store the char-id-tables in hash tables with the attributes as
3201 the key and the actual char-id-table object as the value. Each
3202 char-id-table stores values of an attribute corresponding with
3203 characters. Occasionally we need to get attributes of a character
3204 in a association-list format. These routines provide us with
3206 struct char_attribute_alist_closure
3209 Lisp_Object *char_attribute_alist;
3213 add_char_attribute_alist_mapper (Lisp_Object key, Lisp_Object value,
3214 void *char_attribute_alist_closure)
3216 /* This function can GC */
3217 struct char_attribute_alist_closure *caacl =
3218 (struct char_attribute_alist_closure*) char_attribute_alist_closure;
3220 = get_char_id_table (XCHAR_TABLE(value), caacl->char_id);
3221 if (!UNBOUNDP (ret))
3223 Lisp_Object *char_attribute_alist = caacl->char_attribute_alist;
3224 *char_attribute_alist
3225 = Fcons (Fcons (key, ret), *char_attribute_alist);
3230 DEFUN ("char-attribute-alist", Fchar_attribute_alist, 1, 1, 0, /*
3231 Return the alist of attributes of CHARACTER.
3235 struct gcpro gcpro1;
3236 struct char_attribute_alist_closure char_attribute_alist_closure;
3237 Lisp_Object alist = Qnil;
3239 CHECK_CHAR (character);
3242 char_attribute_alist_closure.char_id = XCHAR (character);
3243 char_attribute_alist_closure.char_attribute_alist = &alist;
3244 elisp_maphash (add_char_attribute_alist_mapper,
3245 Vchar_attribute_hash_table,
3246 &char_attribute_alist_closure);
3252 DEFUN ("get-char-attribute", Fget_char_attribute, 2, 3, 0, /*
3253 Return the value of CHARACTER's ATTRIBUTE.
3254 Return DEFAULT-VALUE if the value is not exist.
3256 (character, attribute, default_value))
3260 CHECK_CHAR (character);
3262 if (CHARSETP (attribute))
3263 attribute = XCHARSET_NAME (attribute);
3265 table = Fgethash (attribute, Vchar_attribute_hash_table,
3267 if (!UNBOUNDP (table))
3269 Lisp_Object ret = get_char_id_table (XCHAR_TABLE(table),
3271 if (!UNBOUNDP (ret))
3274 return default_value;
3277 void put_char_composition (Lisp_Object character, Lisp_Object value);
3279 put_char_composition (Lisp_Object character, Lisp_Object value)
3282 signal_simple_error ("Invalid value for ->decomposition",
3285 if (CONSP (Fcdr (value)))
3287 if (NILP (Fcdr (Fcdr (value))))
3289 Lisp_Object base = Fcar (value);
3290 Lisp_Object modifier = Fcar (Fcdr (value));
3294 base = make_char (XINT (base));
3295 Fsetcar (value, base);
3297 if (INTP (modifier))
3299 modifier = make_char (XINT (modifier));
3300 Fsetcar (Fcdr (value), modifier);
3305 = Fget_char_attribute (base, Qcomposition, Qnil);
3306 Lisp_Object ret = Fassq (modifier, alist);
3309 Fput_char_attribute (base, Qcomposition,
3310 Fcons (Fcons (modifier, character),
3313 Fsetcdr (ret, character);
3319 Lisp_Object v = Fcar (value);
3323 Emchar c = XINT (v);
3325 = Fget_char_attribute (make_char (c), Q_ucs_unified, Qnil);
3329 Fput_char_attribute (make_char (c), Q_ucs_unified,
3330 Fcons (character, Qnil));
3332 else if (NILP (Fmemq (character, ret)))
3334 Fput_char_attribute (make_char (c), Q_ucs_unified,
3335 Fcons (character, ret));
3341 DEFUN ("put-char-attribute", Fput_char_attribute, 3, 3, 0, /*
3342 Store CHARACTER's ATTRIBUTE with VALUE.
3344 (character, attribute, value))
3346 Lisp_Object ccs = Ffind_charset (attribute);
3348 CHECK_CHAR (character);
3352 value = put_char_ccs_code_point (character, ccs, value);
3353 attribute = XCHARSET_NAME (ccs);
3355 else if (EQ (attribute, Q_decomposition))
3356 put_char_composition (character, value);
3357 else if (EQ (attribute, Qto_ucs))
3363 signal_simple_error ("Invalid value for =>ucs", value);
3367 ret = Fget_char_attribute (make_char (c), Q_ucs_unified, Qnil);
3370 Fput_char_attribute (make_char (c), Q_ucs_unified,
3371 Fcons (character, Qnil));
3373 else if (NILP (Fmemq (character, ret)))
3375 Fput_char_attribute (make_char (c), Q_ucs_unified,
3376 Fcons (character, ret));
3379 else if (EQ (attribute, Q_unified))
3381 Lisp_Object rest = value;
3384 while (CONSP (rest))
3389 ret = Fdefine_char (ret);
3393 Fput_char_attribute (ret, Q_unified_from, list1 (character));
3394 Fsetcar (rest, ret);
3400 else if (EQ (attribute, Qideographic_structure))
3401 value = Fcopy_sequence (Fchar_refs_simplify_char_specs (value));
3404 Lisp_Object table = Fgethash (attribute,
3405 Vchar_attribute_hash_table,
3410 table = make_char_id_table (Qunbound);
3411 Fputhash (attribute, table, Vchar_attribute_hash_table);
3413 XCHAR_TABLE_NAME (table) = attribute;
3416 put_char_id_table (XCHAR_TABLE(table), character, value);
3421 DEFUN ("remove-char-attribute", Fremove_char_attribute, 2, 2, 0, /*
3422 Remove CHARACTER's ATTRIBUTE.
3424 (character, attribute))
3428 CHECK_CHAR (character);
3429 ccs = Ffind_charset (attribute);
3432 return remove_char_ccs (character, ccs);
3436 Lisp_Object table = Fgethash (attribute,
3437 Vchar_attribute_hash_table,
3439 if (!UNBOUNDP (table))
3441 put_char_id_table (XCHAR_TABLE(table), character, Qunbound);
3450 int char_table_open_db_maybe (Lisp_Char_Table* cit);
3451 void char_table_close_db_maybe (Lisp_Char_Table* cit);
3452 Lisp_Object char_table_get_db (Lisp_Char_Table* cit, Emchar ch);
3454 #ifdef HAVE_LIBCHISE
3456 open_chise_data_source_maybe ()
3458 if (default_chise_data_source == NULL)
3460 Lisp_Object db_dir = Vexec_directory;
3461 int modemask = 0755; /* rwxr-xr-x */
3464 db_dir = build_string ("../lib-src");
3465 db_dir = Fexpand_file_name (build_string ("chise-db"), db_dir);
3467 default_chise_data_source
3468 = CHISE_DS_open (CHISE_DS_Berkeley_DB, XSTRING_DATA (db_dir),
3469 0 /* DB_HASH */, modemask);
3470 if (default_chise_data_source == NULL)
3475 #endif /* HAVE_LIBCHISE */
3477 DEFUN ("close-char-data-source", Fclose_char_data_source, 0, 0, 0, /*
3478 Close data-source of CHISE.
3482 #ifdef HAVE_LIBCHISE
3483 int status = CHISE_DS_close (default_chise_data_source);
3485 default_chise_data_source = NULL;
3488 #endif /* HAVE_LIBCHISE */
3493 char_table_open_db_maybe (Lisp_Char_Table* cit)
3495 Lisp_Object attribute = CHAR_TABLE_NAME (cit);
3497 if (!NILP (attribute))
3499 #ifdef HAVE_LIBCHISE
3500 if ( open_chise_data_source_maybe () )
3502 #else /* HAVE_LIBCHISE */
3503 if (NILP (Fdatabase_live_p (cit->db)))
3506 = char_attribute_system_db_file (Qsystem_char_id, attribute, 0);
3508 cit->db = Fopen_database (db_file, Qnil, Qnil,
3509 build_string ("r"), Qnil);
3513 #endif /* not HAVE_LIBCHISE */
3521 char_table_close_db_maybe (Lisp_Char_Table* cit)
3523 #ifndef HAVE_LIBCHISE
3524 if (!NILP (cit->db))
3526 if (!NILP (Fdatabase_live_p (cit->db)))
3527 Fclose_database (cit->db);
3530 #endif /* not HAVE_LIBCHISE */
3534 char_table_get_db (Lisp_Char_Table* cit, Emchar ch)
3537 #ifdef HAVE_LIBCHISE
3540 = chise_ds_load_char_feature_value (default_chise_data_source, ch,
3541 XSTRING_DATA(Fsymbol_name
3547 val = Fread (make_string (chise_value_data (&value),
3548 chise_value_size (&value) ));
3552 #else /* HAVE_LIBCHISE */
3553 val = Fget_database (Fprin1_to_string (make_char (ch), Qnil),
3555 if (!UNBOUNDP (val))
3559 #endif /* not HAVE_LIBCHISE */
3563 #ifndef HAVE_LIBCHISE
3565 char_attribute_system_db_file (Lisp_Object key_type, Lisp_Object attribute,
3568 Lisp_Object db_dir = Vexec_directory;
3571 db_dir = build_string ("../lib-src");
3573 db_dir = Fexpand_file_name (build_string ("chise-db"), db_dir);
3574 if (writing_mode && NILP (Ffile_exists_p (db_dir)))
3575 Fmake_directory_internal (db_dir);
3577 db_dir = Fexpand_file_name (Fsymbol_name (key_type), db_dir);
3578 if (writing_mode && NILP (Ffile_exists_p (db_dir)))
3579 Fmake_directory_internal (db_dir);
3582 Lisp_Object attribute_name = Fsymbol_name (attribute);
3583 Lisp_Object dest = Qnil, ret;
3585 struct gcpro gcpro1, gcpro2;
3586 int len = XSTRING_CHAR_LENGTH (attribute_name);
3590 for (i = 0; i < len; i++)
3592 Emchar c = string_char (XSTRING (attribute_name), i);
3594 if ( (c == '/') || (c == '%') )
3598 sprintf (str, "%%%02X", c);
3599 dest = concat3 (dest,
3600 Fsubstring (attribute_name,
3601 make_int (base), make_int (i)),
3602 build_string (str));
3606 ret = Fsubstring (attribute_name, make_int (base), make_int (len));
3607 dest = concat2 (dest, ret);
3609 return Fexpand_file_name (dest, db_dir);
3612 #endif /* not HAVE_LIBCHISE */
3614 DEFUN ("save-char-attribute-table", Fsave_char_attribute_table, 1, 1, 0, /*
3615 Save values of ATTRIBUTE into database file.
3619 Lisp_Object table = Fgethash (attribute,
3620 Vchar_attribute_hash_table, Qunbound);
3621 Lisp_Char_Table *ct;
3622 #ifdef HAVE_LIBCHISE
3623 CHISE_Feature feature;
3624 #else /* HAVE_LIBCHISE */
3625 Lisp_Object db_file;
3627 #endif /* not HAVE_LIBCHISE */
3629 if (CHAR_TABLEP (table))
3630 ct = XCHAR_TABLE (table);
3634 #ifdef HAVE_LIBCHISE
3635 if ( open_chise_data_source_maybe () )
3638 = chise_ds_get_feature (default_chise_data_source,
3639 XSTRING_DATA (Fsymbol_name (attribute)));
3640 #else /* HAVE_LIBCHISE */
3641 db_file = char_attribute_system_db_file (Qsystem_char_id, attribute, 1);
3642 db = Fopen_database (db_file, Qnil, Qnil, build_string ("w+"), Qnil);
3643 #endif /* not HAVE_LIBCHISE */
3645 #ifdef HAVE_LIBCHISE
3647 #else /* HAVE_LIBCHISE */
3649 #endif /* not HAVE_LIBCHISE */
3652 Lisp_Object (*filter)(Lisp_Object value);
3654 if (EQ (attribute, Qideographic_structure))
3655 filter = &Fchar_refs_simplify_char_specs;
3659 if (UINT8_BYTE_TABLE_P (ct->table))
3660 save_uint8_byte_table (XUINT8_BYTE_TABLE(ct->table), ct,
3661 #ifdef HAVE_LIBCHISE
3663 #else /* HAVE_LIBCHISE */
3665 #endif /* not HAVE_LIBCHISE */
3667 else if (UINT16_BYTE_TABLE_P (ct->table))
3668 save_uint16_byte_table (XUINT16_BYTE_TABLE(ct->table), ct,
3669 #ifdef HAVE_LIBCHISE
3671 #else /* HAVE_LIBCHISE */
3673 #endif /* not HAVE_LIBCHISE */
3675 else if (BYTE_TABLE_P (ct->table))
3676 save_byte_table (XBYTE_TABLE(ct->table), ct,
3677 #ifdef HAVE_LIBCHISE
3679 #else /* HAVE_LIBCHISE */
3681 #endif /* not HAVE_LIBCHISE */
3683 #ifdef HAVE_LIBCHISE
3684 chise_feature_sync (feature);
3685 #else /* HAVE_LIBCHISE */
3686 Fclose_database (db);
3687 #endif /* not HAVE_LIBCHISE */
3694 DEFUN ("mount-char-attribute-table", Fmount_char_attribute_table, 1, 1, 0, /*
3695 Mount database file on char-attribute-table ATTRIBUTE.
3699 Lisp_Object table = Fgethash (attribute,
3700 Vchar_attribute_hash_table, Qunbound);
3702 if (UNBOUNDP (table))
3704 Lisp_Char_Table *ct;
3706 table = make_char_id_table (Qunbound);
3707 Fputhash (attribute, table, Vchar_attribute_hash_table);
3708 XCHAR_TABLE_NAME(table) = attribute;
3709 ct = XCHAR_TABLE (table);
3710 ct->table = Qunloaded;
3711 XCHAR_TABLE_UNLOADED(table) = 1;
3712 #ifndef HAVE_LIBCHISE
3714 #endif /* not HAVE_LIBCHISE */
3720 DEFUN ("close-char-attribute-table", Fclose_char_attribute_table, 1, 1, 0, /*
3721 Close database of ATTRIBUTE.
3725 Lisp_Object table = Fgethash (attribute,
3726 Vchar_attribute_hash_table, Qunbound);
3727 Lisp_Char_Table *ct;
3729 if (CHAR_TABLEP (table))
3730 ct = XCHAR_TABLE (table);
3733 char_table_close_db_maybe (ct);
3737 DEFUN ("reset-char-attribute-table", Freset_char_attribute_table, 1, 1, 0, /*
3738 Reset values of ATTRIBUTE with database file.
3742 #ifdef HAVE_LIBCHISE
3743 CHISE_Feature feature
3744 = chise_ds_get_feature (default_chise_data_source,
3745 XSTRING_DATA (Fsymbol_name
3748 if (feature == NULL)
3751 if (chise_feature_setup_db (feature, 0) == 0)
3753 Lisp_Object table = Fgethash (attribute,
3754 Vchar_attribute_hash_table, Qunbound);
3755 Lisp_Char_Table *ct;
3757 chise_feature_sync (feature);
3758 if (UNBOUNDP (table))
3760 table = make_char_id_table (Qunbound);
3761 Fputhash (attribute, table, Vchar_attribute_hash_table);
3762 XCHAR_TABLE_NAME(table) = attribute;
3764 ct = XCHAR_TABLE (table);
3765 ct->table = Qunloaded;
3766 char_table_close_db_maybe (ct);
3767 XCHAR_TABLE_UNLOADED(table) = 1;
3771 Lisp_Object table = Fgethash (attribute,
3772 Vchar_attribute_hash_table, Qunbound);
3773 Lisp_Char_Table *ct;
3775 = char_attribute_system_db_file (Qsystem_char_id, attribute, 0);
3777 if (!NILP (Ffile_exists_p (db_file)))
3779 if (UNBOUNDP (table))
3781 table = make_char_id_table (Qunbound);
3782 Fputhash (attribute, table, Vchar_attribute_hash_table);
3783 XCHAR_TABLE_NAME(table) = attribute;
3785 ct = XCHAR_TABLE (table);
3786 ct->table = Qunloaded;
3787 char_table_close_db_maybe (ct);
3788 XCHAR_TABLE_UNLOADED(table) = 1;
3796 load_char_attribute_maybe (Lisp_Char_Table* cit, Emchar ch)
3798 Lisp_Object attribute = CHAR_TABLE_NAME (cit);
3800 if (!NILP (attribute))
3804 if (char_table_open_db_maybe (cit))
3807 val = char_table_get_db (cit, ch);
3809 if (!NILP (Vchar_db_stingy_mode))
3810 char_table_close_db_maybe (cit);
3817 Lisp_Char_Table* char_attribute_table_to_load;
3819 #ifdef HAVE_LIBCHISE
3821 load_char_attribute_table_map_func (CHISE_Char_ID cid,
3822 CHISE_Feature feature,
3823 CHISE_Value *value);
3825 load_char_attribute_table_map_func (CHISE_Char_ID cid,
3826 CHISE_Feature feature,
3830 Lisp_Object ret = get_char_id_table_0 (char_attribute_table_to_load, code);
3832 if (EQ (ret, Qunloaded))
3833 put_char_id_table_0 (char_attribute_table_to_load, code,
3834 Fread (make_string ((Bufbyte *) value->data,
3838 #else /* HAVE_LIBCHISE */
3839 Lisp_Object Qload_char_attribute_table_map_function;
3841 DEFUN ("load-char-attribute-table-map-function",
3842 Fload_char_attribute_table_map_function, 2, 2, 0, /*
3843 For internal use. Don't use it.
3847 Lisp_Object c = Fread (key);
3848 Emchar code = XCHAR (c);
3849 Lisp_Object ret = get_char_id_table_0 (char_attribute_table_to_load, code);
3851 if (EQ (ret, Qunloaded))
3852 put_char_id_table_0 (char_attribute_table_to_load, code, Fread (value));
3855 #endif /* not HAVE_LIBCHISE */
3857 DEFUN ("load-char-attribute-table", Fload_char_attribute_table, 1, 1, 0, /*
3858 Load values of ATTRIBUTE into database file.
3862 Lisp_Object table = Fgethash (attribute,
3863 Vchar_attribute_hash_table,
3865 if (CHAR_TABLEP (table))
3867 Lisp_Char_Table *cit = XCHAR_TABLE (table);
3869 if (char_table_open_db_maybe (cit))
3872 char_attribute_table_to_load = XCHAR_TABLE (table);
3874 struct gcpro gcpro1;
3877 #ifdef HAVE_LIBCHISE
3878 chise_feature_foreach_char_with_value
3879 (chise_ds_get_feature (default_chise_data_source,
3880 XSTRING_DATA (Fsymbol_name (cit->name))),
3881 &load_char_attribute_table_map_func);
3882 #else /* HAVE_LIBCHISE */
3883 Fmap_database (Qload_char_attribute_table_map_function, cit->db);
3884 #endif /* not HAVE_LIBCHISE */
3887 char_table_close_db_maybe (cit);
3888 XCHAR_TABLE_UNLOADED(table) = 0;
3893 #endif /* HAVE_CHISE */
3895 DEFUN ("map-char-attribute", Fmap_char_attribute, 2, 3, 0, /*
3896 Map FUNCTION over entries in ATTRIBUTE, calling it with two args,
3897 each key and value in the table.
3899 RANGE specifies a subrange to map over and is in the same format as
3900 the RANGE argument to `put-range-table'. If omitted or t, it defaults to
3903 (function, attribute, range))
3906 Lisp_Char_Table *ct;
3907 struct slow_map_char_table_arg slarg;
3908 struct gcpro gcpro1, gcpro2;
3909 struct chartab_range rainj;
3911 if (!NILP (ccs = Ffind_charset (attribute)))
3913 Lisp_Object encoding_table = XCHARSET_ENCODING_TABLE (ccs);
3915 if (CHAR_TABLEP (encoding_table))
3916 ct = XCHAR_TABLE (encoding_table);
3922 Lisp_Object table = Fgethash (attribute,
3923 Vchar_attribute_hash_table,
3925 if (CHAR_TABLEP (table))
3926 ct = XCHAR_TABLE (table);
3932 decode_char_table_range (range, &rainj);
3934 if (CHAR_TABLE_UNLOADED(ct))
3935 Fload_char_attribute_table (attribute);
3937 slarg.function = function;
3938 slarg.retval = Qnil;
3939 GCPRO2 (slarg.function, slarg.retval);
3940 map_char_table (ct, &rainj, slow_map_char_table_fun, &slarg);
3943 return slarg.retval;
3946 DEFUN ("define-char", Fdefine_char, 1, 1, 0, /*
3947 Store character's ATTRIBUTES.
3951 Lisp_Object rest = attributes;
3952 Lisp_Object code = Fcdr (Fassq (Qmap_ucs, attributes));
3953 Lisp_Object character;
3956 code = Fcdr (Fassq (Qucs, attributes));
3959 while (CONSP (rest))
3961 Lisp_Object cell = Fcar (rest);
3965 signal_simple_error ("Invalid argument", attributes);
3966 if (!NILP (ccs = Ffind_charset (Fcar (cell)))
3967 && ((XCHARSET_FINAL (ccs) != 0) ||
3968 (XCHARSET_MAX_CODE (ccs) > 0) ||
3969 (EQ (ccs, Vcharset_chinese_big5))) )
3973 character = Fmake_char (ccs, Fcar (cell), Fcar (Fcdr (cell)));
3975 character = Fdecode_char (ccs, cell, Qnil, Qt);
3976 if (!NILP (character))
3977 goto setup_attributes;
3983 int cid = XINT (Vnext_defined_char_id);
3985 if (cid <= 0xE00000)
3987 character = make_char (cid);
3988 Vnext_defined_char_id = make_int (cid + 1);
3989 goto setup_attributes;
3993 if ( (!NILP (code = Fcdr (Fassq (Qto_ucs, attributes)))) )
3996 signal_simple_error ("Invalid argument", attributes);
3998 character = make_char (XINT (code) + 0x100000);
3999 goto setup_attributes;
4004 else if (!INTP (code))
4005 signal_simple_error ("Invalid argument", attributes);
4007 character = make_char (XINT (code));
4011 while (CONSP (rest))
4013 Lisp_Object cell = Fcar (rest);
4016 signal_simple_error ("Invalid argument", attributes);
4018 Fput_char_attribute (character, Fcar (cell), Fcdr (cell));
4024 DEFUN ("find-char", Ffind_char, 1, 1, 0, /*
4025 Retrieve the character of the given ATTRIBUTES.
4029 Lisp_Object rest = attributes;
4032 while (CONSP (rest))
4034 Lisp_Object cell = Fcar (rest);
4038 signal_simple_error ("Invalid argument", attributes);
4039 if (!NILP (ccs = Ffind_charset (Fcar (cell))))
4043 return Fmake_char (ccs, Fcar (cell), Fcar (Fcdr (cell)));
4045 return Fdecode_char (ccs, cell, Qnil, Qnil);
4049 if ( (!NILP (code = Fcdr (Fassq (Qto_ucs, attributes)))) )
4052 signal_simple_error ("Invalid argument", attributes);
4054 return make_char (XINT (code) + 0x100000);
4062 /************************************************************************/
4063 /* Char table read syntax */
4064 /************************************************************************/
4067 chartab_type_validate (Lisp_Object keyword, Lisp_Object value,
4068 Error_behavior errb)
4070 /* #### should deal with ERRB */
4071 symbol_to_char_table_type (value);
4076 chartab_data_validate (Lisp_Object keyword, Lisp_Object value,
4077 Error_behavior errb)
4081 /* #### should deal with ERRB */
4082 EXTERNAL_LIST_LOOP (rest, value)
4084 Lisp_Object range = XCAR (rest);
4085 struct chartab_range dummy;
4089 signal_simple_error ("Invalid list format", value);
4092 if (!CONSP (XCDR (range))
4093 || !NILP (XCDR (XCDR (range))))
4094 signal_simple_error ("Invalid range format", range);
4095 decode_char_table_range (XCAR (range), &dummy);
4096 decode_char_table_range (XCAR (XCDR (range)), &dummy);
4099 decode_char_table_range (range, &dummy);
4106 chartab_instantiate (Lisp_Object data)
4108 Lisp_Object chartab;
4109 Lisp_Object type = Qgeneric;
4110 Lisp_Object dataval = Qnil;
4112 while (!NILP (data))
4114 Lisp_Object keyw = Fcar (data);
4120 if (EQ (keyw, Qtype))
4122 else if (EQ (keyw, Qdata))
4126 chartab = Fmake_char_table (type);
4129 while (!NILP (data))
4131 Lisp_Object range = Fcar (data);
4132 Lisp_Object val = Fcar (Fcdr (data));
4134 data = Fcdr (Fcdr (data));
4137 if (CHAR_OR_CHAR_INTP (XCAR (range)))
4139 Emchar first = XCHAR_OR_CHAR_INT (Fcar (range));
4140 Emchar last = XCHAR_OR_CHAR_INT (Fcar (Fcdr (range)));
4143 for (i = first; i <= last; i++)
4144 Fput_char_table (make_char (i), val, chartab);
4150 Fput_char_table (range, val, chartab);
4159 /************************************************************************/
4160 /* Category Tables, specifically */
4161 /************************************************************************/
4163 DEFUN ("category-table-p", Fcategory_table_p, 1, 1, 0, /*
4164 Return t if OBJECT is a category table.
4165 A category table is a type of char table used for keeping track of
4166 categories. Categories are used for classifying characters for use
4167 in regexps -- you can refer to a category rather than having to use
4168 a complicated [] expression (and category lookups are significantly
4171 There are 95 different categories available, one for each printable
4172 character (including space) in the ASCII charset. Each category
4173 is designated by one such character, called a "category designator".
4174 They are specified in a regexp using the syntax "\\cX", where X is
4175 a category designator.
4177 A category table specifies, for each character, the categories that
4178 the character is in. Note that a character can be in more than one
4179 category. More specifically, a category table maps from a character
4180 to either the value nil (meaning the character is in no categories)
4181 or a 95-element bit vector, specifying for each of the 95 categories
4182 whether the character is in that category.
4184 Special Lisp functions are provided that abstract this, so you do not
4185 have to directly manipulate bit vectors.
4189 return (CHAR_TABLEP (object) &&
4190 XCHAR_TABLE_TYPE (object) == CHAR_TABLE_TYPE_CATEGORY) ?
4195 check_category_table (Lisp_Object object, Lisp_Object default_)
4199 while (NILP (Fcategory_table_p (object)))
4200 object = wrong_type_argument (Qcategory_table_p, object);
4205 check_category_char (Emchar ch, Lisp_Object table,
4206 unsigned int designator, unsigned int not_p)
4208 REGISTER Lisp_Object temp;
4209 Lisp_Char_Table *ctbl;
4210 #ifdef ERROR_CHECK_TYPECHECK
4211 if (NILP (Fcategory_table_p (table)))
4212 signal_simple_error ("Expected category table", table);
4214 ctbl = XCHAR_TABLE (table);
4215 temp = get_char_table (ch, ctbl);
4220 return bit_vector_bit (XBIT_VECTOR (temp), designator) ? !not_p : not_p;
4223 DEFUN ("check-category-at", Fcheck_category_at, 2, 4, 0, /*
4224 Return t if category of the character at POSITION includes DESIGNATOR.
4225 Optional third arg BUFFER specifies which buffer to use, and defaults
4226 to the current buffer.
4227 Optional fourth arg CATEGORY-TABLE specifies the category table to
4228 use, and defaults to BUFFER's category table.
4230 (position, designator, buffer, category_table))
4235 struct buffer *buf = decode_buffer (buffer, 0);
4237 CHECK_INT (position);
4238 CHECK_CATEGORY_DESIGNATOR (designator);
4239 des = XCHAR (designator);
4240 ctbl = check_category_table (category_table, Vstandard_category_table);
4241 ch = BUF_FETCH_CHAR (buf, XINT (position));
4242 return check_category_char (ch, ctbl, des, 0) ? Qt : Qnil;
4245 DEFUN ("char-in-category-p", Fchar_in_category_p, 2, 3, 0, /*
4246 Return t if category of CHARACTER includes DESIGNATOR, else nil.
4247 Optional third arg CATEGORY-TABLE specifies the category table to use,
4248 and defaults to the standard category table.
4250 (character, designator, category_table))
4256 CHECK_CATEGORY_DESIGNATOR (designator);
4257 des = XCHAR (designator);
4258 CHECK_CHAR (character);
4259 ch = XCHAR (character);
4260 ctbl = check_category_table (category_table, Vstandard_category_table);
4261 return check_category_char (ch, ctbl, des, 0) ? Qt : Qnil;
4264 DEFUN ("category-table", Fcategory_table, 0, 1, 0, /*
4265 Return BUFFER's current category table.
4266 BUFFER defaults to the current buffer.
4270 return decode_buffer (buffer, 0)->category_table;
4273 DEFUN ("standard-category-table", Fstandard_category_table, 0, 0, 0, /*
4274 Return the standard category table.
4275 This is the one used for new buffers.
4279 return Vstandard_category_table;
4282 DEFUN ("copy-category-table", Fcopy_category_table, 0, 1, 0, /*
4283 Return a new category table which is a copy of CATEGORY-TABLE.
4284 CATEGORY-TABLE defaults to the standard category table.
4288 if (NILP (Vstandard_category_table))
4289 return Fmake_char_table (Qcategory);
4292 check_category_table (category_table, Vstandard_category_table);
4293 return Fcopy_char_table (category_table);
4296 DEFUN ("set-category-table", Fset_category_table, 1, 2, 0, /*
4297 Select CATEGORY-TABLE as the new category table for BUFFER.
4298 BUFFER defaults to the current buffer if omitted.
4300 (category_table, buffer))
4302 struct buffer *buf = decode_buffer (buffer, 0);
4303 category_table = check_category_table (category_table, Qnil);
4304 buf->category_table = category_table;
4305 /* Indicate that this buffer now has a specified category table. */
4306 buf->local_var_flags |= XINT (buffer_local_flags.category_table);
4307 return category_table;
4310 DEFUN ("category-designator-p", Fcategory_designator_p, 1, 1, 0, /*
4311 Return t if OBJECT is a category designator (a char in the range ' ' to '~').
4315 return CATEGORY_DESIGNATORP (object) ? Qt : Qnil;
4318 DEFUN ("category-table-value-p", Fcategory_table_value_p, 1, 1, 0, /*
4319 Return t if OBJECT is a category table value.
4320 Valid values are nil or a bit vector of size 95.
4324 return CATEGORY_TABLE_VALUEP (object) ? Qt : Qnil;
4328 #define CATEGORYP(x) \
4329 (CHARP (x) && XCHAR (x) >= 0x20 && XCHAR (x) <= 0x7E)
4331 #define CATEGORY_SET(c) \
4332 (get_char_table(c, XCHAR_TABLE(current_buffer->category_table)))
4334 /* Return 1 if CATEGORY_SET contains CATEGORY, else return 0.
4335 The faster version of `!NILP (Faref (category_set, category))'. */
4336 #define CATEGORY_MEMBER(category, category_set) \
4337 (bit_vector_bit(XBIT_VECTOR (category_set), category - 32))
4339 /* Return 1 if there is a word boundary between two word-constituent
4340 characters C1 and C2 if they appear in this order, else return 0.
4341 Use the macro WORD_BOUNDARY_P instead of calling this function
4344 int word_boundary_p (Emchar c1, Emchar c2);
4346 word_boundary_p (Emchar c1, Emchar c2)
4348 Lisp_Object category_set1, category_set2;
4353 if (COMPOSITE_CHAR_P (c1))
4354 c1 = cmpchar_component (c1, 0, 1);
4355 if (COMPOSITE_CHAR_P (c2))
4356 c2 = cmpchar_component (c2, 0, 1);
4360 if (EQ (CHAR_CHARSET (c1), CHAR_CHARSET (c2)))
4363 tail = Vword_separating_categories;
4369 tail = Vword_combining_categories;
4374 category_set1 = CATEGORY_SET (c1);
4375 if (NILP (category_set1))
4376 return default_result;
4377 category_set2 = CATEGORY_SET (c2);
4378 if (NILP (category_set2))
4379 return default_result;
4381 for (; CONSP (tail); tail = XCONS (tail)->cdr)
4383 Lisp_Object elt = XCONS(tail)->car;
4386 && CATEGORYP (XCONS (elt)->car)
4387 && CATEGORYP (XCONS (elt)->cdr)
4388 && CATEGORY_MEMBER (XCHAR (XCONS (elt)->car), category_set1)
4389 && CATEGORY_MEMBER (XCHAR (XCONS (elt)->cdr), category_set2))
4390 return !default_result;
4392 return default_result;
4398 syms_of_chartab (void)
4401 INIT_LRECORD_IMPLEMENTATION (uint8_byte_table);
4402 INIT_LRECORD_IMPLEMENTATION (uint16_byte_table);
4403 INIT_LRECORD_IMPLEMENTATION (byte_table);
4405 #if defined(HAVE_CHISE) && !defined(HAVE_LIBCHISE_LIBCHISE)
4406 defsymbol (&Qsystem_char_id, "system-char-id");
4409 defsymbol (&Qto_ucs, "=>ucs");
4410 defsymbol (&Q_ucs_unified, "->ucs-unified");
4411 defsymbol (&Q_unified, "->unified");
4412 defsymbol (&Q_unified_from, "<-unified");
4413 defsymbol (&Qcomposition, "composition");
4414 defsymbol (&Q_decomposition, "->decomposition");
4415 defsymbol (&Qcompat, "compat");
4416 defsymbol (&Qisolated, "isolated");
4417 defsymbol (&Qinitial, "initial");
4418 defsymbol (&Qmedial, "medial");
4419 defsymbol (&Qfinal, "final");
4420 defsymbol (&Qvertical, "vertical");
4421 defsymbol (&QnoBreak, "noBreak");
4422 defsymbol (&Qfraction, "fraction");
4423 defsymbol (&Qsuper, "super");
4424 defsymbol (&Qsub, "sub");
4425 defsymbol (&Qcircle, "circle");
4426 defsymbol (&Qsquare, "square");
4427 defsymbol (&Qwide, "wide");
4428 defsymbol (&Qnarrow, "narrow");
4429 defsymbol (&Qsmall, "small");
4430 defsymbol (&Qfont, "font");
4432 DEFSUBR (Fchar_attribute_list);
4433 DEFSUBR (Ffind_char_attribute_table);
4434 defsymbol (&Qput_char_table_map_function, "put-char-table-map-function");
4435 DEFSUBR (Fput_char_table_map_function);
4437 DEFSUBR (Fsave_char_attribute_table);
4438 DEFSUBR (Fmount_char_attribute_table);
4439 DEFSUBR (Freset_char_attribute_table);
4440 DEFSUBR (Fclose_char_attribute_table);
4441 DEFSUBR (Fclose_char_data_source);
4442 #ifndef HAVE_LIBCHISE
4443 defsymbol (&Qload_char_attribute_table_map_function,
4444 "load-char-attribute-table-map-function");
4445 DEFSUBR (Fload_char_attribute_table_map_function);
4447 DEFSUBR (Fload_char_attribute_table);
4449 DEFSUBR (Fchar_attribute_alist);
4450 DEFSUBR (Fget_char_attribute);
4451 DEFSUBR (Fput_char_attribute);
4452 DEFSUBR (Fremove_char_attribute);
4453 DEFSUBR (Fmap_char_attribute);
4454 DEFSUBR (Fdefine_char);
4455 DEFSUBR (Ffind_char);
4456 DEFSUBR (Fchar_variants);
4458 DEFSUBR (Fget_composite_char);
4461 INIT_LRECORD_IMPLEMENTATION (char_table);
4465 INIT_LRECORD_IMPLEMENTATION (char_table_entry);
4468 defsymbol (&Qcategory_table_p, "category-table-p");
4469 defsymbol (&Qcategory_designator_p, "category-designator-p");
4470 defsymbol (&Qcategory_table_value_p, "category-table-value-p");
4473 defsymbol (&Qchar_table, "char-table");
4474 defsymbol (&Qchar_tablep, "char-table-p");
4476 DEFSUBR (Fchar_table_p);
4477 DEFSUBR (Fchar_table_type_list);
4478 DEFSUBR (Fvalid_char_table_type_p);
4479 DEFSUBR (Fchar_table_type);
4480 DEFSUBR (Freset_char_table);
4481 DEFSUBR (Fmake_char_table);
4482 DEFSUBR (Fcopy_char_table);
4483 DEFSUBR (Fget_char_table);
4484 DEFSUBR (Fget_range_char_table);
4485 DEFSUBR (Fvalid_char_table_value_p);
4486 DEFSUBR (Fcheck_valid_char_table_value);
4487 DEFSUBR (Fput_char_table);
4488 DEFSUBR (Fmap_char_table);
4491 DEFSUBR (Fcategory_table_p);
4492 DEFSUBR (Fcategory_table);
4493 DEFSUBR (Fstandard_category_table);
4494 DEFSUBR (Fcopy_category_table);
4495 DEFSUBR (Fset_category_table);
4496 DEFSUBR (Fcheck_category_at);
4497 DEFSUBR (Fchar_in_category_p);
4498 DEFSUBR (Fcategory_designator_p);
4499 DEFSUBR (Fcategory_table_value_p);
4505 vars_of_chartab (void)
4508 DEFVAR_LISP ("next-defined-char-id", &Vnext_defined_char_id /*
4510 Vnext_defined_char_id = make_int (0x0F0000);
4514 DEFVAR_LISP ("char-db-stingy-mode", &Vchar_db_stingy_mode /*
4516 Vchar_db_stingy_mode = Qt;
4518 #ifdef HAVE_LIBCHISE
4519 Vchise_db_directory = build_string(chise_db_dir);
4520 DEFVAR_LISP ("chise-db-directory", &Vchise_db_directory /*
4521 Directory of CHISE character databases.
4524 Vchise_system_db_directory = build_string(chise_system_db_dir);
4525 DEFVAR_LISP ("chise-system-db-directory", &Vchise_system_db_directory /*
4526 Directory of system character database of CHISE.
4530 #endif /* HAVE_CHISE */
4531 /* DO NOT staticpro this. It works just like Vweak_hash_tables. */
4532 Vall_syntax_tables = Qnil;
4533 dump_add_weak_object_chain (&Vall_syntax_tables);
4537 structure_type_create_chartab (void)
4539 struct structure_type *st;
4541 st = define_structure_type (Qchar_table, 0, chartab_instantiate);
4543 define_structure_type_keyword (st, Qtype, chartab_type_validate);
4544 define_structure_type_keyword (st, Qdata, chartab_data_validate);
4548 complex_vars_of_chartab (void)
4551 staticpro (&Vchar_attribute_hash_table);
4552 Vchar_attribute_hash_table
4553 = make_lisp_hash_table (16, HASH_TABLE_NON_WEAK, HASH_TABLE_EQ);
4554 #endif /* UTF2000 */
4556 /* Set this now, so first buffer creation can refer to it. */
4557 /* Make it nil before calling copy-category-table
4558 so that copy-category-table will know not to try to copy from garbage */
4559 Vstandard_category_table = Qnil;
4560 Vstandard_category_table = Fcopy_category_table (Qnil);
4561 staticpro (&Vstandard_category_table);
4563 DEFVAR_LISP ("word-combining-categories", &Vword_combining_categories /*
4564 List of pair (cons) of categories to determine word boundary.
4566 Emacs treats a sequence of word constituent characters as a single
4567 word (i.e. finds no word boundary between them) iff they belongs to
4568 the same charset. But, exceptions are allowed in the following cases.
4570 \(1) The case that characters are in different charsets is controlled
4571 by the variable `word-combining-categories'.
4573 Emacs finds no word boundary between characters of different charsets
4574 if they have categories matching some element of this list.
4576 More precisely, if an element of this list is a cons of category CAT1
4577 and CAT2, and a multibyte character C1 which has CAT1 is followed by
4578 C2 which has CAT2, there's no word boundary between C1 and C2.
4580 For instance, to tell that ASCII characters and Latin-1 characters can
4581 form a single word, the element `(?l . ?l)' should be in this list
4582 because both characters have the category `l' (Latin characters).
4584 \(2) The case that character are in the same charset is controlled by
4585 the variable `word-separating-categories'.
4587 Emacs find a word boundary between characters of the same charset
4588 if they have categories matching some element of this list.
4590 More precisely, if an element of this list is a cons of category CAT1
4591 and CAT2, and a multibyte character C1 which has CAT1 is followed by
4592 C2 which has CAT2, there's a word boundary between C1 and C2.
4594 For instance, to tell that there's a word boundary between Japanese
4595 Hiragana and Japanese Kanji (both are in the same charset), the
4596 element `(?H . ?C) should be in this list.
4599 Vword_combining_categories = Qnil;
4601 DEFVAR_LISP ("word-separating-categories", &Vword_separating_categories /*
4602 List of pair (cons) of categories to determine word boundary.
4603 See the documentation of the variable `word-combining-categories'.
4606 Vword_separating_categories = Qnil;