Mercurial > emacs
comparison src/charset.c @ 89483:2f877ed80fa6
*** empty log message ***
author | Kenichi Handa <handa@m17n.org> |
---|---|
date | Mon, 08 Sep 2003 12:53:41 +0000 |
parents | 375f2633d815 31ac7465ef51 |
children | 36913565de8a |
comparison
equal
deleted
inserted
replaced
88123:375f2633d815 | 89483:2f877ed80fa6 |
---|---|
1 /* Basic multilingual character support. | 1 /* Basic character set support. |
2 Copyright (C) 1995, 1997, 1998 Electrotechnical Laboratory, JAPAN. | 2 Copyright (C) 1995, 97, 98, 2000, 2001 Electrotechnical Laboratory, JAPAN. |
3 Licensed to the Free Software Foundation. | 3 Licensed to the Free Software Foundation. |
4 Copyright (C) 2001 Free Software Foundation, Inc. | 4 Copyright (C) 2001 Free Software Foundation, Inc. |
5 Copyright (C) 2003 | |
6 National Institute of Advanced Industrial Science and Technology (AIST) | |
7 Registration Number H13PRO009 | |
5 | 8 |
6 This file is part of GNU Emacs. | 9 This file is part of GNU Emacs. |
7 | 10 |
8 GNU Emacs is free software; you can redistribute it and/or modify | 11 GNU Emacs is free software; you can redistribute it and/or modify |
9 it under the terms of the GNU General Public License as published by | 12 it under the terms of the GNU General Public License as published by |
18 You should have received a copy of the GNU General Public License | 21 You should have received a copy of the GNU General Public License |
19 along with GNU Emacs; see the file COPYING. If not, write to | 22 along with GNU Emacs; see the file COPYING. If not, write to |
20 the Free Software Foundation, Inc., 59 Temple Place - Suite 330, | 23 the Free Software Foundation, Inc., 59 Temple Place - Suite 330, |
21 Boston, MA 02111-1307, USA. */ | 24 Boston, MA 02111-1307, USA. */ |
22 | 25 |
23 /* At first, see the document in `charset.h' to understand the code in | |
24 this file. */ | |
25 | |
26 #ifdef emacs | |
27 #include <config.h> | 26 #include <config.h> |
28 #endif | |
29 | 27 |
30 #include <stdio.h> | 28 #include <stdio.h> |
31 | 29 #include <unistd.h> |
32 #ifdef emacs | 30 #include <ctype.h> |
33 | |
34 #include <sys/types.h> | 31 #include <sys/types.h> |
35 #include "lisp.h" | 32 #include "lisp.h" |
36 #include "buffer.h" | 33 #include "character.h" |
37 #include "charset.h" | 34 #include "charset.h" |
38 #include "composite.h" | |
39 #include "coding.h" | 35 #include "coding.h" |
40 #include "disptab.h" | 36 #include "disptab.h" |
41 | 37 #include "buffer.h" |
42 #else /* not emacs */ | 38 |
43 | 39 /*** GENERAL NOTES on CODED CHARACTER SETS (CHARSETS) *** |
44 #include "mulelib.h" | 40 |
45 | 41 A coded character set ("charset" hereafter) is a meaningful |
46 #endif /* emacs */ | 42 collection (i.e. language, culture, functionality, etc.) of |
47 | 43 characters. Emacs handles multiple charsets at once. In Emacs Lisp |
48 Lisp_Object Qcharset, Qascii, Qeight_bit_control, Qeight_bit_graphic; | 44 code, a charset is represented by a symbol. In C code, a charset is |
49 Lisp_Object Qunknown; | 45 represented by its ID number or by a pointer to a struct charset. |
50 | 46 |
51 /* Declaration of special leading-codes. */ | 47 The actual information about each charset is stored in two places. |
52 EMACS_INT leading_code_private_11; /* for private DIMENSION1 of 1-column */ | 48 Lispy information is stored in the hash table Vcharset_hash_table as |
53 EMACS_INT leading_code_private_12; /* for private DIMENSION1 of 2-column */ | 49 a vector (charset attributes). The other information is stored in |
54 EMACS_INT leading_code_private_21; /* for private DIMENSION2 of 1-column */ | 50 charset_table as a struct charset. |
55 EMACS_INT leading_code_private_22; /* for private DIMENSION2 of 2-column */ | 51 |
56 | 52 */ |
57 /* Declaration of special charsets. The values are set by | 53 |
58 Fsetup_special_charsets. */ | 54 /* List of all charsets. This variable is used only from Emacs |
59 int charset_latin_iso8859_1; /* ISO8859-1 (Latin-1) */ | 55 Lisp. */ |
60 int charset_jisx0208_1978; /* JISX0208.1978 (Japanese Kanji old set) */ | |
61 int charset_jisx0208; /* JISX0208.1983 (Japanese Kanji) */ | |
62 int charset_katakana_jisx0201; /* JISX0201.Kana (Japanese Katakana) */ | |
63 int charset_latin_jisx0201; /* JISX0201.Roman (Japanese Roman) */ | |
64 int charset_big5_1; /* Big5 Level 1 (Chinese Traditional) */ | |
65 int charset_big5_2; /* Big5 Level 2 (Chinese Traditional) */ | |
66 | |
67 Lisp_Object Qcharset_table; | |
68 | |
69 /* A char-table containing information of each character set. */ | |
70 Lisp_Object Vcharset_table; | |
71 | |
72 /* A vector of charset symbol indexed by charset-id. This is used | |
73 only for returning charset symbol from C functions. */ | |
74 Lisp_Object Vcharset_symbol_table; | |
75 | |
76 /* A list of charset symbols ever defined. */ | |
77 Lisp_Object Vcharset_list; | 56 Lisp_Object Vcharset_list; |
78 | 57 |
79 /* Vector of translation table ever defined. | 58 /* Hash table that contains attributes of each charset. Keys are |
80 ID of a translation table is used to index this vector. */ | 59 charset symbols, and values are vectors of charset attributes. */ |
81 Lisp_Object Vtranslation_table_vector; | 60 Lisp_Object Vcharset_hash_table; |
82 | 61 |
83 /* A char-table for characters which may invoke auto-filling. */ | 62 /* Table of struct charset. */ |
84 Lisp_Object Vauto_fill_chars; | 63 struct charset *charset_table; |
85 | 64 |
86 Lisp_Object Qauto_fill_chars; | 65 static int charset_table_size; |
87 | 66 int charset_table_used; |
88 /* Tables used by macros BYTES_BY_CHAR_HEAD and WIDTH_BY_CHAR_HEAD. */ | 67 |
89 int bytes_by_char_head[256]; | 68 Lisp_Object Qcharsetp; |
90 int width_by_char_head[256]; | 69 |
70 /* Special charset symbols. */ | |
71 Lisp_Object Qascii; | |
72 Lisp_Object Qeight_bit; | |
73 Lisp_Object Qiso_8859_1; | |
74 Lisp_Object Qunicode; | |
75 | |
76 /* The corresponding charsets. */ | |
77 int charset_ascii; | |
78 int charset_eight_bit; | |
79 int charset_iso_8859_1; | |
80 int charset_unicode; | |
81 | |
82 /* The other special charsets. */ | |
83 int charset_jisx0201_roman; | |
84 int charset_jisx0208_1978; | |
85 int charset_jisx0208; | |
86 | |
87 /* Value of charset attribute `charset-iso-plane'. */ | |
88 Lisp_Object Qgl, Qgr; | |
89 | |
90 /* Charset of unibyte characters. */ | |
91 int charset_unibyte; | |
92 | |
93 /* List of charsets ordered by the priority. */ | |
94 Lisp_Object Vcharset_ordered_list; | |
95 | |
96 /* Incremented everytime we change Vcharset_ordered_list. This is | |
97 unsigned short so that it fits in Lisp_Int and never matches | |
98 -1. */ | |
99 unsigned short charset_ordered_list_tick; | |
100 | |
101 /* List of iso-2022 charsets. */ | |
102 Lisp_Object Viso_2022_charset_list; | |
103 | |
104 /* List of emacs-mule charsets. */ | |
105 Lisp_Object Vemacs_mule_charset_list; | |
106 | |
107 struct charset *emacs_mule_charset[256]; | |
91 | 108 |
92 /* Mapping table from ISO2022's charset (specified by DIMENSION, | 109 /* Mapping table from ISO2022's charset (specified by DIMENSION, |
93 CHARS, and FINAL-CHAR) to Emacs' charset. */ | 110 CHARS, and FINAL-CHAR) to Emacs' charset. */ |
94 int iso_charset_table[2][2][128]; | 111 int iso_charset_table[ISO_MAX_DIMENSION][ISO_MAX_CHARS][ISO_MAX_FINAL]; |
95 | 112 |
96 /* Variables used locally in the macro FETCH_MULTIBYTE_CHAR. */ | 113 Lisp_Object Vcharset_map_directory; |
97 unsigned char *_fetch_multibyte_char_p; | 114 |
98 int _fetch_multibyte_char_len; | 115 Lisp_Object Vchar_unified_charset_table; |
99 | 116 |
100 /* Offset to add to a non-ASCII value when inserting it. */ | 117 /* Defined in chartab.c */ |
101 EMACS_INT nonascii_insert_offset; | 118 extern void |
102 | 119 map_char_table_for_charset P_ ((void (*c_function) (Lisp_Object, Lisp_Object), |
103 /* Translation table for converting non-ASCII unibyte characters | 120 Lisp_Object function, Lisp_Object table, |
104 to multibyte codes, or nil. */ | 121 Lisp_Object arg, struct charset *charset, |
105 Lisp_Object Vnonascii_translation_table; | 122 unsigned from, unsigned to)); |
106 | 123 |
107 /* List of all possible generic characters. */ | 124 #define CODE_POINT_TO_INDEX(charset, code) \ |
108 Lisp_Object Vgeneric_character_list; | 125 ((charset)->code_linear_p \ |
126 ? (code) - (charset)->min_code \ | |
127 : (((charset)->code_space_mask[(code) >> 24] & 0x8) \ | |
128 && ((charset)->code_space_mask[((code) >> 16) & 0xFF] & 0x4) \ | |
129 && ((charset)->code_space_mask[((code) >> 8) & 0xFF] & 0x2) \ | |
130 && ((charset)->code_space_mask[(code) & 0xFF] & 0x1)) \ | |
131 ? (((((code) >> 24) - (charset)->code_space[12]) \ | |
132 * (charset)->code_space[11]) \ | |
133 + (((((code) >> 16) & 0xFF) - (charset)->code_space[8]) \ | |
134 * (charset)->code_space[7]) \ | |
135 + (((((code) >> 8) & 0xFF) - (charset)->code_space[4]) \ | |
136 * (charset)->code_space[3]) \ | |
137 + (((code) & 0xFF) - (charset)->code_space[0]) \ | |
138 - ((charset)->char_index_offset)) \ | |
139 : -1) | |
140 | |
141 | |
142 /* Convert the character index IDX to code-point CODE for CHARSET. | |
143 It is assumed that IDX is in a valid range. */ | |
144 | |
145 #define INDEX_TO_CODE_POINT(charset, idx) \ | |
146 ((charset)->code_linear_p \ | |
147 ? (idx) + (charset)->min_code \ | |
148 : (idx += (charset)->char_index_offset, \ | |
149 (((charset)->code_space[0] + (idx) % (charset)->code_space[2]) \ | |
150 | (((charset)->code_space[4] \ | |
151 + ((idx) / (charset)->code_space[3] % (charset)->code_space[6])) \ | |
152 << 8) \ | |
153 | (((charset)->code_space[8] \ | |
154 + ((idx) / (charset)->code_space[7] % (charset)->code_space[10])) \ | |
155 << 16) \ | |
156 | (((charset)->code_space[12] + ((idx) / (charset)->code_space[11])) \ | |
157 << 24)))) | |
158 | |
109 | 159 |
110 | 160 |
161 | |
162 /* Set to 1 to warn that a charset map is loaded and thus a buffer | |
163 text and a string data may be relocated. */ | |
164 int charset_map_loaded; | |
165 | |
166 struct charset_map_entries | |
167 { | |
168 struct { | |
169 unsigned from, to; | |
170 int c; | |
171 } entry[0x10000]; | |
172 struct charset_map_entries *next; | |
173 }; | |
174 | |
175 /* Load the mapping information for CHARSET from ENTRIES. | |
176 | |
177 If CONTROL_FLAG is 0, setup CHARSET->min_char and CHARSET->max_char. | |
178 | |
179 If CONTROL_FLAG is 1, setup CHARSET->min_char, CHARSET->max_char, | |
180 CHARSET->decoder, and CHARSET->encoder. | |
181 | |
182 If CONTROL_FLAG is 2, setup CHARSET->deunifier and | |
183 Vchar_unify_table. If Vchar_unified_charset_table is non-nil, | |
184 setup it too. */ | |
185 | |
186 static void | |
187 load_charset_map (charset, entries, n_entries, control_flag) | |
188 struct charset *charset; | |
189 struct charset_map_entries *entries; | |
190 int n_entries; | |
191 int control_flag; | |
192 { | |
193 Lisp_Object vec, table; | |
194 unsigned max_code = CHARSET_MAX_CODE (charset); | |
195 int ascii_compatible_p = charset->ascii_compatible_p; | |
196 int min_char, max_char, nonascii_min_char; | |
197 int i; | |
198 unsigned char *fast_map = charset->fast_map; | |
199 | |
200 if (n_entries <= 0) | |
201 return; | |
202 | |
203 if (control_flag > 0) | |
204 { | |
205 int n = CODE_POINT_TO_INDEX (charset, max_code) + 1; | |
206 | |
207 table = Fmake_char_table (Qnil, Qnil); | |
208 if (control_flag == 1) | |
209 vec = Fmake_vector (make_number (n), make_number (-1)); | |
210 else if (! CHAR_TABLE_P (Vchar_unify_table)) | |
211 Vchar_unify_table = Fmake_char_table (Qnil, Qnil); | |
212 | |
213 charset_map_loaded = 1; | |
214 } | |
215 | |
216 min_char = max_char = entries->entry[0].c; | |
217 nonascii_min_char = MAX_CHAR; | |
218 for (i = 0; i < n_entries; i++) | |
219 { | |
220 unsigned from, to; | |
221 int from_index, to_index; | |
222 int from_c, to_c; | |
223 int idx = i % 0x10000; | |
224 | |
225 if (i > 0 && idx == 0) | |
226 entries = entries->next; | |
227 from = entries->entry[idx].from; | |
228 to = entries->entry[idx].to; | |
229 from_c = entries->entry[idx].c; | |
230 from_index = CODE_POINT_TO_INDEX (charset, from); | |
231 if (from == to) | |
232 { | |
233 to_index = from_index; | |
234 to_c = from_c; | |
235 } | |
236 else | |
237 { | |
238 to_index = CODE_POINT_TO_INDEX (charset, to); | |
239 to_c = from_c + (to_index - from_index); | |
240 } | |
241 if (from_index < 0 || to_index < 0) | |
242 continue; | |
243 | |
244 if (control_flag < 2) | |
245 { | |
246 int c; | |
247 | |
248 if (to_c > max_char) | |
249 max_char = to_c; | |
250 else if (from_c < min_char) | |
251 min_char = from_c; | |
252 if (ascii_compatible_p) | |
253 { | |
254 if (! ASCII_BYTE_P (from_c)) | |
255 { | |
256 if (from_c < nonascii_min_char) | |
257 nonascii_min_char = from_c; | |
258 } | |
259 else if (! ASCII_BYTE_P (to_c)) | |
260 { | |
261 nonascii_min_char = 0x80; | |
262 } | |
263 } | |
264 | |
265 for (c = from_c; c <= to_c; c++) | |
266 CHARSET_FAST_MAP_SET (c, fast_map); | |
267 | |
268 if (control_flag == 1) | |
269 { | |
270 unsigned code = from; | |
271 | |
272 if (CHARSET_COMPACT_CODES_P (charset)) | |
273 while (1) | |
274 { | |
275 ASET (vec, from_index, make_number (from_c)); | |
276 if (NILP (CHAR_TABLE_REF (table, from_c))) | |
277 CHAR_TABLE_SET (table, from_c, make_number (code)); | |
278 if (from_index == to_index) | |
279 break; | |
280 from_index++, from_c++; | |
281 code = INDEX_TO_CODE_POINT (charset, from_index); | |
282 } | |
283 else | |
284 for (; from_index <= to_index; from_index++, from_c++) | |
285 { | |
286 ASET (vec, from_index, make_number (from_c)); | |
287 if (NILP (CHAR_TABLE_REF (table, from_c))) | |
288 CHAR_TABLE_SET (table, from_c, make_number (from_index)); | |
289 } | |
290 } | |
291 } | |
292 else | |
293 { | |
294 unsigned code = from; | |
295 | |
296 while (1) | |
297 { | |
298 int c1 = DECODE_CHAR (charset, code); | |
299 | |
300 if (c1 >= 0) | |
301 { | |
302 CHAR_TABLE_SET (table, from_c, make_number (c1)); | |
303 CHAR_TABLE_SET (Vchar_unify_table, c1, make_number (from_c)); | |
304 if (CHAR_TABLE_P (Vchar_unified_charset_table)) | |
305 CHAR_TABLE_SET (Vchar_unified_charset_table, c1, | |
306 CHARSET_NAME (charset)); | |
307 } | |
308 if (from_index == to_index) | |
309 break; | |
310 from_index++, from_c++; | |
311 code = INDEX_TO_CODE_POINT (charset, from_index); | |
312 } | |
313 } | |
314 } | |
315 | |
316 if (control_flag < 2) | |
317 { | |
318 CHARSET_MIN_CHAR (charset) = (ascii_compatible_p | |
319 ? nonascii_min_char : min_char); | |
320 CHARSET_MAX_CHAR (charset) = max_char; | |
321 if (control_flag == 1) | |
322 { | |
323 CHARSET_DECODER (charset) = vec; | |
324 CHARSET_ENCODER (charset) = table; | |
325 } | |
326 } | |
327 else | |
328 CHARSET_DEUNIFIER (charset) = table; | |
329 } | |
330 | |
331 | |
332 /* Read a hexadecimal number (preceded by "0x") from the file FP while | |
333 paying attention to comment charcter '#'. */ | |
334 | |
335 static INLINE unsigned | |
336 read_hex (fp, eof) | |
337 FILE *fp; | |
338 int *eof; | |
339 { | |
340 int c; | |
341 unsigned n; | |
342 | |
343 while ((c = getc (fp)) != EOF) | |
344 { | |
345 if (c == '#') | |
346 { | |
347 while ((c = getc (fp)) != EOF && c != '\n'); | |
348 } | |
349 else if (c == '0') | |
350 { | |
351 if ((c = getc (fp)) == EOF || c == 'x') | |
352 break; | |
353 } | |
354 } | |
355 if (c == EOF) | |
356 { | |
357 *eof = 1; | |
358 return 0; | |
359 } | |
360 *eof = 0; | |
361 n = 0; | |
362 if (c == 'x') | |
363 while ((c = getc (fp)) != EOF && isxdigit (c)) | |
364 n = ((n << 4) | |
365 | (c <= '9' ? c - '0' : c <= 'F' ? c - 'A' + 10 : c - 'a' + 10)); | |
366 else | |
367 while ((c = getc (fp)) != EOF && isdigit (c)) | |
368 n = (n * 10) + c - '0'; | |
369 if (c != EOF) | |
370 ungetc (c, fp); | |
371 return n; | |
372 } | |
373 | |
374 | |
375 /* Return a mapping vector for CHARSET loaded from MAPFILE. | |
376 Each line of MAPFILE has this form | |
377 0xAAAA 0xCCCC | |
378 where 0xAAAA is a code-point and 0xCCCC is the corresponding | |
379 character code, or this form | |
380 0xAAAA-0xBBBB 0xCCCC | |
381 where 0xAAAA and 0xBBBB are code-points specifying a range, and | |
382 0xCCCC is the first character code of the range. | |
383 | |
384 The returned vector has this form: | |
385 [ CODE1 CHAR1 CODE2 CHAR2 .... ] | |
386 where CODE1 is a code-point or a cons of code-points specifying a | |
387 range. */ | |
388 | |
389 extern void add_to_log P_ ((char *, Lisp_Object, Lisp_Object)); | |
390 | |
391 static void | |
392 load_charset_map_from_file (charset, mapfile, control_flag) | |
393 struct charset *charset; | |
394 Lisp_Object mapfile; | |
395 int control_flag; | |
396 { | |
397 unsigned min_code = CHARSET_MIN_CODE (charset); | |
398 unsigned max_code = CHARSET_MAX_CODE (charset); | |
399 int fd; | |
400 FILE *fp; | |
401 int eof; | |
402 Lisp_Object suffixes; | |
403 struct charset_map_entries *head, *entries; | |
404 int n_entries; | |
405 | |
406 suffixes = Fcons (build_string (".map"), | |
407 Fcons (build_string (".TXT"), Qnil)); | |
408 | |
409 fd = openp (Fcons (Vcharset_map_directory, Qnil), mapfile, suffixes, | |
410 NULL, Qnil); | |
411 if (fd < 0 | |
412 || ! (fp = fdopen (fd, "r"))) | |
413 { | |
414 add_to_log ("Failure in loading charset map: %S", mapfile, Qnil); | |
415 return; | |
416 } | |
417 | |
418 head = entries = ((struct charset_map_entries *) | |
419 alloca (sizeof (struct charset_map_entries))); | |
420 n_entries = 0; | |
421 eof = 0; | |
422 while (1) | |
423 { | |
424 unsigned from, to; | |
425 int c; | |
426 int idx; | |
427 | |
428 from = read_hex (fp, &eof); | |
429 if (eof) | |
430 break; | |
431 if (getc (fp) == '-') | |
432 to = read_hex (fp, &eof); | |
433 else | |
434 to = from; | |
435 c = (int) read_hex (fp, &eof); | |
436 | |
437 if (from < min_code || to > max_code || from > to || c > MAX_CHAR) | |
438 continue; | |
439 | |
440 if (n_entries > 0 && (n_entries % 0x10000) == 0) | |
441 { | |
442 entries->next = ((struct charset_map_entries *) | |
443 alloca (sizeof (struct charset_map_entries))); | |
444 entries = entries->next; | |
445 } | |
446 idx = n_entries % 0x10000; | |
447 entries->entry[idx].from = from; | |
448 entries->entry[idx].to = to; | |
449 entries->entry[idx].c = c; | |
450 n_entries++; | |
451 } | |
452 fclose (fp); | |
453 close (fd); | |
454 | |
455 load_charset_map (charset, head, n_entries, control_flag); | |
456 } | |
457 | |
458 static void | |
459 load_charset_map_from_vector (charset, vec, control_flag) | |
460 struct charset *charset; | |
461 Lisp_Object vec; | |
462 int control_flag; | |
463 { | |
464 unsigned min_code = CHARSET_MIN_CODE (charset); | |
465 unsigned max_code = CHARSET_MAX_CODE (charset); | |
466 struct charset_map_entries *head, *entries; | |
467 int n_entries; | |
468 int len = ASIZE (vec); | |
469 int i; | |
470 | |
471 if (len % 2 == 1) | |
472 { | |
473 add_to_log ("Failure in loading charset map: %V", vec, Qnil); | |
474 return; | |
475 } | |
476 | |
477 head = entries = ((struct charset_map_entries *) | |
478 alloca (sizeof (struct charset_map_entries))); | |
479 n_entries = 0; | |
480 for (i = 0; i < len; i += 2) | |
481 { | |
482 Lisp_Object val, val2; | |
483 unsigned from, to; | |
484 int c; | |
485 int idx; | |
486 | |
487 val = AREF (vec, i); | |
488 if (CONSP (val)) | |
489 { | |
490 val2 = XCDR (val); | |
491 val = XCAR (val); | |
492 CHECK_NATNUM (val); | |
493 CHECK_NATNUM (val2); | |
494 from = XFASTINT (val); | |
495 to = XFASTINT (val2); | |
496 } | |
497 else | |
498 { | |
499 CHECK_NATNUM (val); | |
500 from = to = XFASTINT (val); | |
501 } | |
502 val = AREF (vec, i + 1); | |
503 CHECK_NATNUM (val); | |
504 c = XFASTINT (val); | |
505 | |
506 if (from < min_code || to > max_code || from > to || c > MAX_CHAR) | |
507 continue; | |
508 | |
509 if ((n_entries % 0x10000) == 0) | |
510 { | |
511 entries->next = ((struct charset_map_entries *) | |
512 alloca (sizeof (struct charset_map_entries))); | |
513 entries = entries->next; | |
514 } | |
515 idx = n_entries % 0x10000; | |
516 entries->entry[idx].from = from; | |
517 entries->entry[idx].to = to; | |
518 entries->entry[idx].c = c; | |
519 n_entries++; | |
520 } | |
521 | |
522 load_charset_map (charset, head, n_entries, control_flag); | |
523 } | |
524 | |
525 static void | |
526 load_charset (charset) | |
527 struct charset *charset; | |
528 { | |
529 if (CHARSET_METHOD (charset) == CHARSET_METHOD_MAP_DEFERRED) | |
530 { | |
531 Lisp_Object map; | |
532 | |
533 map = CHARSET_MAP (charset); | |
534 if (STRINGP (map)) | |
535 load_charset_map_from_file (charset, map, 1); | |
536 else | |
537 load_charset_map_from_vector (charset, map, 1); | |
538 CHARSET_METHOD (charset) = CHARSET_METHOD_MAP; | |
539 } | |
540 } | |
541 | |
542 | |
543 DEFUN ("charsetp", Fcharsetp, Scharsetp, 1, 1, 0, | |
544 doc: /* Return non-nil if and only if OBJECT is a charset.*/) | |
545 (object) | |
546 Lisp_Object object; | |
547 { | |
548 return (CHARSETP (object) ? Qt : Qnil); | |
549 } | |
550 | |
551 | |
111 void | 552 void |
112 invalid_character (c) | 553 map_charset_chars (c_function, function, arg, |
113 int c; | 554 charset, from, to) |
114 { | 555 void (*c_function) P_ ((Lisp_Object, Lisp_Object)); |
115 error ("Invalid character: 0%o, %d, 0x%x", c, c, c); | 556 Lisp_Object function, arg; |
116 } | 557 struct charset *charset; |
117 | 558 unsigned from, to; |
118 /* Parse string STR of length LENGTH and fetch information of a | 559 { |
119 character at STR. Set BYTES to the byte length the character | 560 Lisp_Object range; |
120 occupies, CHARSET, C1, C2 to proper values of the character. */ | 561 int partial; |
121 | 562 |
122 #define SPLIT_MULTIBYTE_SEQ(str, length, bytes, charset, c1, c2) \ | 563 if (CHARSET_METHOD (charset) == CHARSET_METHOD_MAP_DEFERRED) |
123 do { \ | 564 load_charset (charset); |
124 (c1) = *(str); \ | 565 |
125 (bytes) = BYTES_BY_CHAR_HEAD (c1); \ | 566 partial = (from > CHARSET_MIN_CODE (charset) |
126 if ((bytes) == 1) \ | 567 || to < CHARSET_MAX_CODE (charset)); |
127 (charset) = ASCII_BYTE_P (c1) ? CHARSET_ASCII : CHARSET_8_BIT_GRAPHIC; \ | 568 |
128 else if ((bytes) == 2) \ | 569 if (CHARSET_UNIFIED_P (charset) |
129 { \ | 570 && CHAR_TABLE_P (CHARSET_DEUNIFIER (charset))) |
130 if ((c1) == LEADING_CODE_8_BIT_CONTROL) \ | 571 { |
131 (charset) = CHARSET_8_BIT_CONTROL, (c1) = (str)[1] - 0x20; \ | 572 map_char_table_for_charset (c_function, function, |
132 else \ | 573 CHARSET_DEUNIFIER (charset), arg, |
133 (charset) = (c1), (c1) = (str)[1] & 0x7F; \ | 574 partial ? charset : NULL, from, to); |
134 } \ | 575 } |
135 else if ((bytes) == 3) \ | 576 |
136 { \ | 577 if (CHARSET_METHOD (charset) == CHARSET_METHOD_OFFSET) |
137 if ((c1) < LEADING_CODE_PRIVATE_11) \ | 578 { |
138 (charset) = (c1), (c1) = (str)[1] & 0x7F, (c2) = (str)[2] & 0x7F; \ | 579 int from_idx = CODE_POINT_TO_INDEX (charset, from); |
139 else \ | 580 int to_idx = CODE_POINT_TO_INDEX (charset, to); |
140 (charset) = (str)[1], (c1) = (str)[2] & 0x7F; \ | 581 int from_c = from_idx + CHARSET_CODE_OFFSET (charset); |
141 } \ | 582 int to_c = to_idx + CHARSET_CODE_OFFSET (charset); |
142 else \ | 583 |
143 (charset) = (str)[1], (c1) = (str)[2] & 0x7F, (c2) = (str)[3] & 0x7F; \ | 584 range = Fcons (make_number (from_c), make_number (to_c)); |
144 } while (0) | 585 if (NILP (function)) |
145 | 586 (*c_function) (range, arg); |
146 /* 1 if CHARSET, C1, and C2 compose a valid character, else 0. | |
147 Note that this intentionally allows invalid components, such | |
148 as 0xA0 0xA0, because there exist many files that contain | |
149 such invalid byte sequences, especially in EUC-GB. */ | |
150 #define CHAR_COMPONENTS_VALID_P(charset, c1, c2) \ | |
151 ((charset) == CHARSET_ASCII \ | |
152 ? ((c1) >= 0 && (c1) <= 0x7F) \ | |
153 : ((charset) == CHARSET_8_BIT_CONTROL \ | |
154 ? ((c1) >= 0x80 && (c1) <= 0x9F) \ | |
155 : ((charset) == CHARSET_8_BIT_GRAPHIC \ | |
156 ? ((c1) >= 0x80 && (c1) <= 0xFF) \ | |
157 : (CHARSET_DIMENSION (charset) == 1 \ | |
158 ? ((c1) >= 0x20 && (c1) <= 0x7F) \ | |
159 : ((c1) >= 0x20 && (c1) <= 0x7F \ | |
160 && (c2) >= 0x20 && (c2) <= 0x7F))))) | |
161 | |
162 /* Store multi-byte form of the character C in STR. The caller should | |
163 allocate at least 4-byte area at STR in advance. Returns the | |
164 length of the multi-byte form. If C is an invalid character code, | |
165 return -1. */ | |
166 | |
167 int | |
168 char_to_string_1 (c, str) | |
169 int c; | |
170 unsigned char *str; | |
171 { | |
172 unsigned char *p = str; | |
173 | |
174 if (c & CHAR_MODIFIER_MASK) /* This includes the case C is negative. */ | |
175 { | |
176 /* Multibyte character can't have a modifier bit. */ | |
177 if (! SINGLE_BYTE_CHAR_P ((c & ~CHAR_MODIFIER_MASK))) | |
178 return -1; | |
179 | |
180 /* For Meta, Shift, and Control modifiers, we need special care. */ | |
181 if (c & CHAR_META) | |
182 { | |
183 /* Move the meta bit to the right place for a string. */ | |
184 c = (c & ~CHAR_META) | 0x80; | |
185 } | |
186 if (c & CHAR_SHIFT) | |
187 { | |
188 /* Shift modifier is valid only with [A-Za-z]. */ | |
189 if ((c & 0377) >= 'A' && (c & 0377) <= 'Z') | |
190 c &= ~CHAR_SHIFT; | |
191 else if ((c & 0377) >= 'a' && (c & 0377) <= 'z') | |
192 c = (c & ~CHAR_SHIFT) - ('a' - 'A'); | |
193 } | |
194 if (c & CHAR_CTL) | |
195 { | |
196 /* Simulate the code in lread.c. */ | |
197 /* Allow `\C- ' and `\C-?'. */ | |
198 if (c == (CHAR_CTL | ' ')) | |
199 c = 0; | |
200 else if (c == (CHAR_CTL | '?')) | |
201 c = 127; | |
202 /* ASCII control chars are made from letters (both cases), | |
203 as well as the non-letters within 0100...0137. */ | |
204 else if ((c & 0137) >= 0101 && (c & 0137) <= 0132) | |
205 c &= (037 | (~0177 & ~CHAR_CTL)); | |
206 else if ((c & 0177) >= 0100 && (c & 0177) <= 0137) | |
207 c &= (037 | (~0177 & ~CHAR_CTL)); | |
208 } | |
209 | |
210 /* If C still has any modifier bits, just ignore it. */ | |
211 c &= ~CHAR_MODIFIER_MASK; | |
212 } | |
213 | |
214 if (SINGLE_BYTE_CHAR_P (c)) | |
215 { | |
216 if (ASCII_BYTE_P (c) || c >= 0xA0) | |
217 *p++ = c; | |
218 else | 587 else |
219 { | 588 call2 (function, range, arg); |
220 *p++ = LEADING_CODE_8_BIT_CONTROL; | 589 } |
221 *p++ = c + 0x20; | 590 else if (CHARSET_METHOD (charset) == CHARSET_METHOD_MAP) |
222 } | 591 { |
223 } | 592 if (! CHAR_TABLE_P (CHARSET_ENCODER (charset))) |
224 else if (CHAR_VALID_P (c, 0)) | 593 return; |
225 { | 594 if (CHARSET_ASCII_COMPATIBLE_P (charset) && from <= 127) |
226 int charset, c1, c2; | 595 { |
227 | 596 range = Fcons (make_number (from), make_number (to)); |
228 SPLIT_CHAR (c, charset, c1, c2); | 597 if (to >= 128) |
229 | 598 XSETCAR (range, make_number (127)); |
230 if (charset >= LEADING_CODE_EXT_11) | 599 |
231 *p++ = (charset < LEADING_CODE_EXT_12 | 600 if (NILP (function)) |
232 ? LEADING_CODE_PRIVATE_11 | 601 (*c_function) (range, arg); |
233 : (charset < LEADING_CODE_EXT_21 | 602 else |
234 ? LEADING_CODE_PRIVATE_12 | 603 call2 (function, range, arg); |
235 : (charset < LEADING_CODE_EXT_22 | 604 } |
236 ? LEADING_CODE_PRIVATE_21 | 605 map_char_table_for_charset (c_function, function, |
237 : LEADING_CODE_PRIVATE_22))); | 606 CHARSET_ENCODER (charset), arg, |
238 *p++ = charset; | 607 partial ? charset : NULL, from, to); |
239 if ((c1 > 0 && c1 < 32) || (c2 > 0 && c2 < 32)) | 608 } |
240 return -1; | 609 else if (CHARSET_METHOD (charset) == CHARSET_METHOD_SUBSET) |
241 if (c1) | 610 { |
242 { | 611 Lisp_Object subset_info; |
243 *p++ = c1 | 0x80; | 612 int offset; |
244 if (c2 > 0) | 613 |
245 *p++ = c2 | 0x80; | 614 subset_info = CHARSET_SUBSET (charset); |
246 } | 615 charset = CHARSET_FROM_ID (XFASTINT (AREF (subset_info, 0))); |
247 } | 616 offset = XINT (AREF (subset_info, 3)); |
617 from -= offset; | |
618 if (from < XFASTINT (AREF (subset_info, 1))) | |
619 from = XFASTINT (AREF (subset_info, 1)); | |
620 to -= offset; | |
621 if (to > XFASTINT (AREF (subset_info, 2))) | |
622 to = XFASTINT (AREF (subset_info, 2)); | |
623 map_charset_chars (c_function, function, arg, charset, from, to); | |
624 } | |
625 else /* i.e. CHARSET_METHOD_SUPERSET */ | |
626 { | |
627 Lisp_Object parents; | |
628 | |
629 for (parents = CHARSET_SUPERSET (charset); CONSP (parents); | |
630 parents = XCDR (parents)) | |
631 { | |
632 int offset; | |
633 unsigned this_from, this_to; | |
634 | |
635 charset = CHARSET_FROM_ID (XFASTINT (XCAR (XCAR (parents)))); | |
636 offset = XINT (XCDR (XCAR (parents))); | |
637 this_from = from - offset; | |
638 this_to = to - offset; | |
639 if (this_from < CHARSET_MIN_CODE (charset)) | |
640 this_from = CHARSET_MIN_CODE (charset); | |
641 if (this_to > CHARSET_MAX_CODE (charset)) | |
642 this_to = CHARSET_MAX_CODE (charset); | |
643 map_charset_chars (c_function, function, arg, charset, from, to); | |
644 } | |
645 } | |
646 } | |
647 | |
648 DEFUN ("map-charset-chars", Fmap_charset_chars, Smap_charset_chars, 2, 5, 0, | |
649 doc: /* Call FUNCTION for all characters in CHARSET. | |
650 FUNCTION is called with an argument RANGE and the optional 3rd | |
651 argument ARG. | |
652 | |
653 RANGE is a cons (FROM . TO), where FROM and TO indicate a range of | |
654 characters contained in CHARSET. | |
655 | |
656 The optional 4th and 5th arguments FROM-CODE and TO-CODE specify the | |
657 range of code points of target characters. */) | |
658 (function, charset, arg, from_code, to_code) | |
659 Lisp_Object function, charset, arg, from_code, to_code; | |
660 { | |
661 struct charset *cs; | |
662 unsigned from, to; | |
663 | |
664 CHECK_CHARSET_GET_CHARSET (charset, cs); | |
665 if (NILP (from_code)) | |
666 from = CHARSET_MIN_CODE (cs); | |
248 else | 667 else |
249 return -1; | 668 { |
250 | 669 CHECK_NATNUM (from_code); |
251 return (p - str); | 670 from = XINT (from_code); |
252 } | 671 if (from < CHARSET_MIN_CODE (cs)) |
253 | 672 from = CHARSET_MIN_CODE (cs); |
254 | 673 } |
255 /* Store multi-byte form of the character C in STR. The caller should | 674 if (NILP (to_code)) |
256 allocate at least 4-byte area at STR in advance. Returns the | 675 to = CHARSET_MAX_CODE (cs); |
257 length of the multi-byte form. If C is an invalid character code, | 676 else |
258 signal an error. | 677 { |
259 | 678 CHECK_NATNUM (to_code); |
260 Use macro `CHAR_STRING (C, STR)' instead of calling this function | 679 to = XINT (to_code); |
261 directly if C can be an ASCII character. */ | 680 if (to > CHARSET_MAX_CODE (cs)) |
262 | 681 to = CHARSET_MAX_CODE (cs); |
263 int | 682 } |
264 char_to_string (c, str) | 683 map_charset_chars (NULL, function, arg, cs, from, to); |
265 int c; | 684 return Qnil; |
266 unsigned char *str; | 685 } |
267 { | 686 |
268 int len; | 687 |
269 len = char_to_string_1 (c, str); | 688 /* Define a charset according to the arguments. The Nth argument is |
270 if (len == -1) | 689 the Nth attribute of the charset (the last attribute `charset-id' |
271 invalid_character (c); | 690 is not included). See the docstring of `define-charset' for the |
272 return len; | 691 detail. */ |
273 } | 692 |
274 | 693 DEFUN ("define-charset-internal", Fdefine_charset_internal, |
275 | 694 Sdefine_charset_internal, charset_arg_max, MANY, 0, |
276 /* Return the non-ASCII character corresponding to multi-byte form at | 695 doc: /* For internal use only. |
277 STR of length LEN. If ACTUAL_LEN is not NULL, store the byte | 696 usage: (define-charset-internal ...) */) |
278 length of the multibyte form in *ACTUAL_LEN. | 697 (nargs, args) |
279 | 698 int nargs; |
280 Use macros STRING_CHAR or STRING_CHAR_AND_LENGTH instead of calling | 699 Lisp_Object *args; |
281 this function directly if you want ot handle ASCII characters as | 700 { |
282 well. */ | 701 /* Charset attr vector. */ |
283 | 702 Lisp_Object attrs; |
284 int | 703 Lisp_Object val; |
285 string_to_char (str, len, actual_len) | 704 unsigned hash_code; |
286 const unsigned char *str; | 705 struct Lisp_Hash_Table *hash_table = XHASH_TABLE (Vcharset_hash_table); |
287 int len, *actual_len; | 706 int i, j; |
288 { | 707 struct charset charset; |
289 int c, bytes, charset, c1, c2; | 708 int id; |
290 | 709 int dimension; |
291 SPLIT_MULTIBYTE_SEQ (str, len, bytes, charset, c1, c2); | 710 int new_definition_p; |
292 c = MAKE_CHAR (charset, c1, c2); | 711 int nchars; |
293 if (actual_len) | 712 |
294 *actual_len = bytes; | 713 if (nargs != charset_arg_max) |
295 return c; | 714 return Fsignal (Qwrong_number_of_arguments, |
296 } | 715 Fcons (intern ("define-charset-internal"), |
297 | 716 make_number (nargs))); |
298 /* Return the length of the multi-byte form at string STR of length LEN. | 717 |
299 Use the macro MULTIBYTE_FORM_LENGTH instead. */ | 718 attrs = Fmake_vector (make_number (charset_attr_max), Qnil); |
300 int | 719 |
301 multibyte_form_length (str, len) | 720 CHECK_SYMBOL (args[charset_arg_name]); |
302 const unsigned char *str; | 721 ASET (attrs, charset_name, args[charset_arg_name]); |
303 int len; | 722 |
304 { | 723 val = args[charset_arg_code_space]; |
305 int bytes; | 724 for (i = 0, dimension = 0, nchars = 1; i < 4; i++) |
306 | 725 { |
307 PARSE_MULTIBYTE_SEQ (str, len, bytes); | 726 int min_byte, max_byte; |
308 return bytes; | 727 |
309 } | 728 min_byte = XINT (Faref (val, make_number (i * 2))); |
310 | 729 max_byte = XINT (Faref (val, make_number (i * 2 + 1))); |
311 /* Check multibyte form at string STR of length LEN and set variables | 730 if (min_byte < 0 || min_byte > max_byte || max_byte >= 256) |
312 pointed by CHARSET, C1, and C2 to charset and position codes of the | 731 error ("Invalid :code-space value"); |
313 character at STR, and return 0. If there's no multibyte character, | 732 charset.code_space[i * 4] = min_byte; |
314 return -1. This should be used only in the macro SPLIT_STRING | 733 charset.code_space[i * 4 + 1] = max_byte; |
315 which checks range of STR in advance. */ | 734 charset.code_space[i * 4 + 2] = max_byte - min_byte + 1; |
316 | 735 nchars *= charset.code_space[i * 4 + 2]; |
317 int | 736 charset.code_space[i * 4 + 3] = nchars; |
318 split_string (str, len, charset, c1, c2) | 737 if (max_byte > 0) |
319 const unsigned char *str; | 738 dimension = i + 1; |
320 unsigned char *c1, *c2; | 739 } |
321 int len, *charset; | 740 |
322 { | 741 val = args[charset_arg_dimension]; |
323 register int bytes, cs, code1, code2 = -1; | 742 if (NILP (val)) |
324 | 743 charset.dimension = dimension; |
325 SPLIT_MULTIBYTE_SEQ (str, len, bytes, cs, code1, code2); | 744 else |
326 if (cs == CHARSET_ASCII) | 745 { |
327 return -1; | 746 CHECK_NATNUM (val); |
328 *charset = cs; | 747 charset.dimension = XINT (val); |
329 *c1 = code1; | 748 if (charset.dimension < 1 || charset.dimension > 4) |
330 *c2 = code2; | 749 args_out_of_range_3 (val, make_number (1), make_number (4)); |
331 return 0; | 750 } |
332 } | 751 |
333 | 752 charset.code_linear_p |
334 /* Return 1 iff character C has valid printable glyph. | 753 = (charset.dimension == 1 |
335 Use the macro CHAR_PRINTABLE_P instead. */ | 754 || (charset.code_space[2] == 256 |
336 int | 755 && (charset.dimension == 2 |
337 char_printable_p (c) | 756 || (charset.code_space[6] == 256 |
338 int c; | 757 && (charset.dimension == 3 |
339 { | 758 || charset.code_space[10] == 256))))); |
340 int charset, c1, c2; | 759 |
341 | 760 if (! charset.code_linear_p) |
342 if (ASCII_BYTE_P (c)) | 761 { |
343 return 1; | 762 charset.code_space_mask = (unsigned char *) xmalloc (256); |
344 else if (SINGLE_BYTE_CHAR_P (c)) | 763 bzero (charset.code_space_mask, 256); |
345 return 0; | 764 for (i = 0; i < 4; i++) |
346 else if (c >= MAX_CHAR) | 765 for (j = charset.code_space[i * 4]; j <= charset.code_space[i * 4 + 1]; |
347 return 0; | 766 j++) |
348 | 767 charset.code_space_mask[j] |= (1 << i); |
349 SPLIT_CHAR (c, charset, c1, c2); | 768 } |
350 if (! CHARSET_DEFINED_P (charset)) | 769 |
351 return 0; | 770 charset.iso_chars_96 = charset.code_space[2] == 96; |
352 if (CHARSET_CHARS (charset) == 94 | 771 |
353 ? c1 <= 32 || c1 >= 127 | 772 charset.min_code = (charset.code_space[0] |
354 : c1 < 32) | 773 | (charset.code_space[4] << 8) |
355 return 0; | 774 | (charset.code_space[8] << 16) |
356 if (CHARSET_DIMENSION (charset) == 2 | 775 | (charset.code_space[12] << 24)); |
357 && (CHARSET_CHARS (charset) == 94 | 776 charset.max_code = (charset.code_space[1] |
358 ? c2 <= 32 || c2 >= 127 | 777 | (charset.code_space[5] << 8) |
359 : c2 < 32)) | 778 | (charset.code_space[9] << 16) |
360 return 0; | 779 | (charset.code_space[13] << 24)); |
361 return 1; | 780 charset.char_index_offset = 0; |
362 } | 781 |
363 | 782 val = args[charset_arg_min_code]; |
364 /* Translate character C by translation table TABLE. If C | 783 if (! NILP (val)) |
365 is negative, translate a character specified by CHARSET, C1, and C2 | 784 { |
366 (C1 and C2 are code points of the character). If no translation is | 785 unsigned code; |
367 found in TABLE, return C. */ | 786 |
368 int | 787 if (INTEGERP (val)) |
369 translate_char (table, c, charset, c1, c2) | 788 code = XINT (val); |
370 Lisp_Object table; | |
371 int c, charset, c1, c2; | |
372 { | |
373 Lisp_Object ch; | |
374 int alt_charset, alt_c1, alt_c2, dimension; | |
375 | |
376 if (c < 0) c = MAKE_CHAR (charset, (c1 & 0x7F) , (c2 & 0x7F)); | |
377 if (!CHAR_TABLE_P (table) | |
378 || (ch = Faref (table, make_number (c)), !NATNUMP (ch))) | |
379 return c; | |
380 | |
381 SPLIT_CHAR (XFASTINT (ch), alt_charset, alt_c1, alt_c2); | |
382 dimension = CHARSET_DIMENSION (alt_charset); | |
383 if ((dimension == 1 && alt_c1 > 0) || (dimension == 2 && alt_c2 > 0)) | |
384 /* CH is not a generic character, just return it. */ | |
385 return XFASTINT (ch); | |
386 | |
387 /* Since CH is a generic character, we must return a specific | |
388 charater which has the same position codes as C from CH. */ | |
389 if (charset < 0) | |
390 SPLIT_CHAR (c, charset, c1, c2); | |
391 if (dimension != CHARSET_DIMENSION (charset)) | |
392 /* We can't make such a character because of dimension mismatch. */ | |
393 return c; | |
394 return MAKE_CHAR (alt_charset, c1, c2); | |
395 } | |
396 | |
397 /* Convert the unibyte character C to multibyte based on | |
398 Vnonascii_translation_table or nonascii_insert_offset. If they can't | |
399 convert C to a valid multibyte character, convert it based on | |
400 DEFAULT_NONASCII_INSERT_OFFSET which makes C a Latin-1 character. */ | |
401 | |
402 int | |
403 unibyte_char_to_multibyte (c) | |
404 int c; | |
405 { | |
406 if (c < 0400 && c >= 0200) | |
407 { | |
408 int c_save = c; | |
409 | |
410 if (! NILP (Vnonascii_translation_table)) | |
411 { | |
412 c = XINT (Faref (Vnonascii_translation_table, make_number (c))); | |
413 if (c >= 0400 && ! char_valid_p (c, 0)) | |
414 c = c_save + DEFAULT_NONASCII_INSERT_OFFSET; | |
415 } | |
416 else if (c >= 0240 && nonascii_insert_offset > 0) | |
417 { | |
418 c += nonascii_insert_offset; | |
419 if (c < 0400 || ! char_valid_p (c, 0)) | |
420 c = c_save + DEFAULT_NONASCII_INSERT_OFFSET; | |
421 } | |
422 else if (c >= 0240) | |
423 c = c_save + DEFAULT_NONASCII_INSERT_OFFSET; | |
424 } | |
425 return c; | |
426 } | |
427 | |
428 | |
429 /* Convert the multibyte character C to unibyte 8-bit character based | |
430 on Vnonascii_translation_table or nonascii_insert_offset. If | |
431 REV_TBL is non-nil, it should be a reverse table of | |
432 Vnonascii_translation_table, i.e. what given by: | |
433 Fchar_table_extra_slot (Vnonascii_translation_table, make_number (0)) */ | |
434 | |
435 int | |
436 multibyte_char_to_unibyte (c, rev_tbl) | |
437 int c; | |
438 Lisp_Object rev_tbl; | |
439 { | |
440 if (!SINGLE_BYTE_CHAR_P (c)) | |
441 { | |
442 int c_save = c; | |
443 | |
444 if (! CHAR_TABLE_P (rev_tbl) | |
445 && CHAR_TABLE_P (Vnonascii_translation_table)) | |
446 rev_tbl = Fchar_table_extra_slot (Vnonascii_translation_table, | |
447 make_number (0)); | |
448 if (CHAR_TABLE_P (rev_tbl)) | |
449 { | |
450 Lisp_Object temp; | |
451 temp = Faref (rev_tbl, make_number (c)); | |
452 if (INTEGERP (temp)) | |
453 c = XINT (temp); | |
454 if (c >= 256) | |
455 c = (c_save & 0177) + 0200; | |
456 } | |
457 else | 789 else |
458 { | 790 { |
459 if (nonascii_insert_offset > 0) | 791 CHECK_CONS (val); |
460 c -= nonascii_insert_offset; | 792 CHECK_NUMBER_CAR (val); |
461 if (c < 128 || c >= 256) | 793 CHECK_NUMBER_CDR (val); |
462 c = (c_save & 0177) + 0200; | 794 code = (XINT (XCAR (val)) << 16) | (XINT (XCDR (val))); |
463 } | 795 } |
464 } | 796 if (code < charset.min_code |
465 | 797 || code > charset.max_code) |
466 return c; | 798 args_out_of_range_3 (make_number (charset.min_code), |
467 } | 799 make_number (charset.max_code), val); |
468 | 800 charset.char_index_offset = CODE_POINT_TO_INDEX (&charset, code); |
469 | 801 charset.min_code = code; |
470 /* Update the table Vcharset_table with the given arguments (see the | 802 } |
471 document of `define-charset' for the meaning of each argument). | 803 |
472 Several other table contents are also updated. The caller should | 804 val = args[charset_arg_max_code]; |
473 check the validity of CHARSET-ID and the remaining arguments in | 805 if (! NILP (val)) |
474 advance. */ | 806 { |
475 | 807 unsigned code; |
476 void | 808 |
477 update_charset_table (charset_id, dimension, chars, width, direction, | 809 if (INTEGERP (val)) |
478 iso_final_char, iso_graphic_plane, | 810 code = XINT (val); |
479 short_name, long_name, description) | 811 else |
480 Lisp_Object charset_id, dimension, chars, width, direction; | 812 { |
481 Lisp_Object iso_final_char, iso_graphic_plane; | 813 CHECK_CONS (val); |
482 Lisp_Object short_name, long_name, description; | 814 CHECK_NUMBER_CAR (val); |
483 { | 815 CHECK_NUMBER_CDR (val); |
484 int charset = XINT (charset_id); | 816 code = (XINT (XCAR (val)) << 16) | (XINT (XCDR (val))); |
485 int bytes; | 817 } |
486 unsigned char leading_code_base, leading_code_ext; | 818 if (code < charset.min_code |
487 | 819 || code > charset.max_code) |
488 if (NILP (CHARSET_TABLE_ENTRY (charset))) | 820 args_out_of_range_3 (make_number (charset.min_code), |
489 CHARSET_TABLE_ENTRY (charset) | 821 make_number (charset.max_code), val); |
490 = Fmake_vector (make_number (CHARSET_MAX_IDX), Qnil); | 822 charset.max_code = code; |
491 | 823 } |
492 if (NILP (long_name)) | 824 |
493 long_name = short_name; | 825 charset.compact_codes_p = charset.max_code < 0x1000000; |
494 if (NILP (description)) | 826 |
495 description = long_name; | 827 val = args[charset_arg_invalid_code]; |
496 | 828 if (NILP (val)) |
497 /* Get byte length of multibyte form, base leading-code, and | 829 { |
498 extended leading-code of the charset. See the comment under the | 830 if (charset.min_code > 0) |
499 title "GENERAL NOTE on CHARACTER SET (CHARSET)" in charset.h. */ | 831 charset.invalid_code = 0; |
500 bytes = XINT (dimension); | 832 else |
501 if (charset < MIN_CHARSET_PRIVATE_DIMENSION1) | 833 { |
502 { | 834 XSETINT (val, charset.max_code + 1); |
503 /* Official charset, it doesn't have an extended leading-code. */ | 835 if (XINT (val) == charset.max_code + 1) |
504 if (charset != CHARSET_ASCII && charset != CHARSET_8_BIT_GRAPHIC) | 836 charset.invalid_code = charset.max_code + 1; |
505 bytes += 1; /* For a base leading-code. */ | 837 else |
506 leading_code_base = charset; | 838 error ("Attribute :invalid-code must be specified"); |
507 leading_code_ext = 0; | 839 } |
508 } | 840 } |
509 else | 841 else |
510 { | 842 { |
511 /* Private charset. */ | 843 CHECK_NATNUM (val); |
512 bytes += 2; /* For base and extended leading-codes. */ | 844 charset.invalid_code = XFASTINT (val); |
513 leading_code_base | 845 } |
514 = (charset < LEADING_CODE_EXT_12 | 846 |
515 ? LEADING_CODE_PRIVATE_11 | 847 val = args[charset_arg_iso_final]; |
516 : (charset < LEADING_CODE_EXT_21 | 848 if (NILP (val)) |
517 ? LEADING_CODE_PRIVATE_12 | 849 charset.iso_final = -1; |
518 : (charset < LEADING_CODE_EXT_22 | 850 else |
519 ? LEADING_CODE_PRIVATE_21 | 851 { |
520 : LEADING_CODE_PRIVATE_22))); | 852 CHECK_NUMBER (val); |
521 leading_code_ext = charset; | 853 if (XINT (val) < '0' || XINT (val) > 127) |
522 if (BYTES_BY_CHAR_HEAD (leading_code_base) != bytes) | 854 error ("Invalid iso-final-char: %d", XINT (val)); |
523 error ("Invalid dimension for the charset-ID %d", charset); | 855 charset.iso_final = XINT (val); |
524 } | 856 } |
525 | 857 |
526 CHARSET_TABLE_INFO (charset, CHARSET_ID_IDX) = charset_id; | 858 val = args[charset_arg_iso_revision]; |
527 CHARSET_TABLE_INFO (charset, CHARSET_BYTES_IDX) = make_number (bytes); | 859 if (NILP (val)) |
528 CHARSET_TABLE_INFO (charset, CHARSET_DIMENSION_IDX) = dimension; | 860 charset.iso_revision = -1; |
529 CHARSET_TABLE_INFO (charset, CHARSET_CHARS_IDX) = chars; | 861 else |
530 CHARSET_TABLE_INFO (charset, CHARSET_WIDTH_IDX) = width; | 862 { |
531 CHARSET_TABLE_INFO (charset, CHARSET_DIRECTION_IDX) = direction; | 863 CHECK_NUMBER (val); |
532 CHARSET_TABLE_INFO (charset, CHARSET_LEADING_CODE_BASE_IDX) | 864 if (XINT (val) > 63) |
533 = make_number (leading_code_base); | 865 args_out_of_range (make_number (63), val); |
534 CHARSET_TABLE_INFO (charset, CHARSET_LEADING_CODE_EXT_IDX) | 866 charset.iso_revision = XINT (val); |
535 = make_number (leading_code_ext); | 867 } |
536 CHARSET_TABLE_INFO (charset, CHARSET_ISO_FINAL_CHAR_IDX) = iso_final_char; | 868 |
537 CHARSET_TABLE_INFO (charset, CHARSET_ISO_GRAPHIC_PLANE_IDX) | 869 val = args[charset_arg_emacs_mule_id]; |
538 = iso_graphic_plane; | 870 if (NILP (val)) |
539 CHARSET_TABLE_INFO (charset, CHARSET_SHORT_NAME_IDX) = short_name; | 871 charset.emacs_mule_id = -1; |
540 CHARSET_TABLE_INFO (charset, CHARSET_LONG_NAME_IDX) = long_name; | 872 else |
541 CHARSET_TABLE_INFO (charset, CHARSET_DESCRIPTION_IDX) = description; | 873 { |
542 CHARSET_TABLE_INFO (charset, CHARSET_PLIST_IDX) = Qnil; | 874 CHECK_NATNUM (val); |
543 | 875 if ((XINT (val) > 0 && XINT (val) <= 128) || XINT (val) >= 256) |
544 { | 876 error ("Invalid emacs-mule-id: %d", XINT (val)); |
545 /* If we have already defined a charset which has the same | 877 charset.emacs_mule_id = XINT (val); |
546 DIMENSION, CHARS and ISO-FINAL-CHAR but the different | 878 } |
547 DIRECTION, we must update the entry REVERSE-CHARSET of both | 879 |
548 charsets. If there's no such charset, the value of the entry | 880 charset.ascii_compatible_p = ! NILP (args[charset_arg_ascii_compatible_p]); |
549 is set to nil. */ | 881 |
550 int i; | 882 charset.supplementary_p = ! NILP (args[charset_arg_supplementary_p]); |
551 | 883 |
552 for (i = 0; i <= MAX_CHARSET; i++) | 884 charset.unified_p = 0; |
553 if (!NILP (CHARSET_TABLE_ENTRY (i))) | 885 |
554 { | 886 bzero (charset.fast_map, sizeof (charset.fast_map)); |
555 if (CHARSET_DIMENSION (i) == XINT (dimension) | 887 |
556 && CHARSET_CHARS (i) == XINT (chars) | 888 if (! NILP (args[charset_arg_code_offset])) |
557 && CHARSET_ISO_FINAL_CHAR (i) == XINT (iso_final_char) | 889 { |
558 && CHARSET_DIRECTION (i) != XINT (direction)) | 890 val = args[charset_arg_code_offset]; |
891 CHECK_NUMBER (val); | |
892 | |
893 charset.method = CHARSET_METHOD_OFFSET; | |
894 charset.code_offset = XINT (val); | |
895 | |
896 i = CODE_POINT_TO_INDEX (&charset, charset.min_code); | |
897 charset.min_char = i + charset.code_offset; | |
898 i = CODE_POINT_TO_INDEX (&charset, charset.max_code); | |
899 charset.max_char = i + charset.code_offset; | |
900 if (charset.max_char > MAX_CHAR) | |
901 error ("Unsupported max char: %d", charset.max_char); | |
902 | |
903 i = (charset.min_char >> 7) << 7; | |
904 for (; i < 0x10000 && i <= charset.max_char; i += 128) | |
905 CHARSET_FAST_MAP_SET (i, charset.fast_map); | |
906 i = (i >> 12) << 12; | |
907 for (; i <= charset.max_char; i += 0x1000) | |
908 CHARSET_FAST_MAP_SET (i, charset.fast_map); | |
909 } | |
910 else if (! NILP (args[charset_arg_map])) | |
911 { | |
912 val = args[charset_arg_map]; | |
913 ASET (attrs, charset_map, val); | |
914 if (STRINGP (val)) | |
915 load_charset_map_from_file (&charset, val, 0); | |
916 else | |
917 load_charset_map_from_vector (&charset, val, 0); | |
918 charset.method = CHARSET_METHOD_MAP_DEFERRED; | |
919 } | |
920 else if (! NILP (args[charset_arg_subset])) | |
921 { | |
922 Lisp_Object parent; | |
923 Lisp_Object parent_min_code, parent_max_code, parent_code_offset; | |
924 struct charset *parent_charset; | |
925 | |
926 val = args[charset_arg_subset]; | |
927 parent = Fcar (val); | |
928 CHECK_CHARSET_GET_CHARSET (parent, parent_charset); | |
929 parent_min_code = Fnth (make_number (1), val); | |
930 CHECK_NATNUM (parent_min_code); | |
931 parent_max_code = Fnth (make_number (2), val); | |
932 CHECK_NATNUM (parent_max_code); | |
933 parent_code_offset = Fnth (make_number (3), val); | |
934 CHECK_NUMBER (parent_code_offset); | |
935 val = Fmake_vector (make_number (4), Qnil); | |
936 ASET (val, 0, make_number (parent_charset->id)); | |
937 ASET (val, 1, parent_min_code); | |
938 ASET (val, 2, parent_max_code); | |
939 ASET (val, 3, parent_code_offset); | |
940 ASET (attrs, charset_subset, val); | |
941 | |
942 charset.method = CHARSET_METHOD_SUBSET; | |
943 /* Here, we just copy the parent's fast_map. It's not accurate, | |
944 but at least it works for quickly detecting which character | |
945 DOESN'T belong to this charset. */ | |
946 for (i = 0; i < 190; i++) | |
947 charset.fast_map[i] = parent_charset->fast_map[i]; | |
948 | |
949 /* We also copy these for parents. */ | |
950 charset.min_char = parent_charset->min_char; | |
951 charset.max_char = parent_charset->max_char; | |
952 } | |
953 else if (! NILP (args[charset_arg_superset])) | |
954 { | |
955 val = args[charset_arg_superset]; | |
956 charset.method = CHARSET_METHOD_SUPERSET; | |
957 val = Fcopy_sequence (val); | |
958 ASET (attrs, charset_superset, val); | |
959 | |
960 charset.min_char = MAX_CHAR; | |
961 charset.max_char = 0; | |
962 for (; ! NILP (val); val = Fcdr (val)) | |
963 { | |
964 Lisp_Object elt, car_part, cdr_part; | |
965 int this_id, offset; | |
966 struct charset *this_charset; | |
967 | |
968 elt = Fcar (val); | |
969 if (CONSP (elt)) | |
559 { | 970 { |
560 CHARSET_TABLE_INFO (charset, CHARSET_REVERSE_CHARSET_IDX) | 971 car_part = XCAR (elt); |
561 = make_number (i); | 972 cdr_part = XCDR (elt); |
562 CHARSET_TABLE_INFO (i, CHARSET_REVERSE_CHARSET_IDX) = charset_id; | 973 CHECK_CHARSET_GET_ID (car_part, this_id); |
563 break; | 974 CHECK_NUMBER (cdr_part); |
975 offset = XINT (cdr_part); | |
564 } | 976 } |
565 } | 977 else |
566 if (i > MAX_CHARSET) | 978 { |
567 /* No such a charset. */ | 979 CHECK_CHARSET_GET_ID (elt, this_id); |
568 CHARSET_TABLE_INFO (charset, CHARSET_REVERSE_CHARSET_IDX) | 980 offset = 0; |
569 = make_number (-1); | 981 } |
570 } | 982 XSETCAR (val, Fcons (make_number (this_id), make_number (offset))); |
571 | 983 |
572 if (charset != CHARSET_ASCII && charset != CHARSET_8_BIT_GRAPHIC | 984 this_charset = CHARSET_FROM_ID (this_id); |
573 && charset < MIN_CHARSET_PRIVATE_DIMENSION1) | 985 if (charset.min_char > this_charset->min_char) |
574 { | 986 charset.min_char = this_charset->min_char; |
575 bytes_by_char_head[leading_code_base] = bytes; | 987 if (charset.max_char < this_charset->max_char) |
576 width_by_char_head[leading_code_base] = XINT (width); | 988 charset.max_char = this_charset->max_char; |
577 | 989 for (i = 0; i < 190; i++) |
578 /* Update table emacs_code_class. */ | 990 charset.fast_map[i] |= this_charset->fast_map[i]; |
579 emacs_code_class[charset] = (bytes == 2 | 991 } |
580 ? EMACS_leading_code_2 | 992 } |
581 : (bytes == 3 | 993 else |
582 ? EMACS_leading_code_3 | 994 error ("None of :code-offset, :map, :parents are specified"); |
583 : EMACS_leading_code_4)); | 995 |
584 } | 996 val = args[charset_arg_unify_map]; |
585 | 997 if (! NILP (val) && !STRINGP (val)) |
586 /* Update table iso_charset_table. */ | 998 CHECK_VECTOR (val); |
587 if (XINT (iso_final_char) >= 0 | 999 ASET (attrs, charset_unify_map, val); |
588 && ISO_CHARSET_TABLE (dimension, chars, iso_final_char) < 0) | 1000 |
589 ISO_CHARSET_TABLE (dimension, chars, iso_final_char) = charset; | 1001 CHECK_LIST (args[charset_arg_plist]); |
590 } | 1002 ASET (attrs, charset_plist, args[charset_arg_plist]); |
591 | 1003 |
592 #ifdef emacs | 1004 charset.hash_index = hash_lookup (hash_table, args[charset_arg_name], |
593 | 1005 &hash_code); |
594 /* Return charset id of CHARSET_SYMBOL, or return -1 if CHARSET_SYMBOL | 1006 if (charset.hash_index >= 0) |
595 is invalid. */ | 1007 { |
596 int | 1008 new_definition_p = 0; |
597 get_charset_id (charset_symbol) | 1009 id = XFASTINT (CHARSET_SYMBOL_ID (args[charset_arg_name])); |
598 Lisp_Object charset_symbol; | 1010 HASH_VALUE (hash_table, charset.hash_index) = attrs; |
599 { | 1011 } |
1012 else | |
1013 { | |
1014 charset.hash_index = hash_put (hash_table, args[charset_arg_name], attrs, | |
1015 hash_code); | |
1016 if (charset_table_used == charset_table_size) | |
1017 { | |
1018 struct charset *new_table | |
1019 = (struct charset *) xmalloc (sizeof (struct charset) | |
1020 * (charset_table_size + 16)); | |
1021 bcopy (charset_table, new_table, | |
1022 sizeof (struct charset) * charset_table_size); | |
1023 charset_table_size += 16; | |
1024 charset_table = new_table; | |
1025 } | |
1026 id = charset_table_used++; | |
1027 new_definition_p = 1; | |
1028 } | |
1029 | |
1030 ASET (attrs, charset_id, make_number (id)); | |
1031 charset.id = id; | |
1032 charset_table[id] = charset; | |
1033 | |
1034 if (charset.iso_final >= 0) | |
1035 { | |
1036 ISO_CHARSET_TABLE (charset.dimension, charset.iso_chars_96, | |
1037 charset.iso_final) = id; | |
1038 if (new_definition_p) | |
1039 Viso_2022_charset_list = nconc2 (Viso_2022_charset_list, | |
1040 Fcons (make_number (id), Qnil)); | |
1041 if (ISO_CHARSET_TABLE (1, 0, 'J') == id) | |
1042 charset_jisx0201_roman = id; | |
1043 else if (ISO_CHARSET_TABLE (2, 0, '@') == id) | |
1044 charset_jisx0208_1978 = id; | |
1045 else if (ISO_CHARSET_TABLE (2, 0, 'B') == id) | |
1046 charset_jisx0208 = id; | |
1047 } | |
1048 | |
1049 if (charset.emacs_mule_id >= 0) | |
1050 { | |
1051 emacs_mule_charset[charset.emacs_mule_id] = CHARSET_FROM_ID (id); | |
1052 if (charset.emacs_mule_id < 0xA0) | |
1053 emacs_mule_bytes[charset.emacs_mule_id] = charset.dimension + 1; | |
1054 if (new_definition_p) | |
1055 Vemacs_mule_charset_list = nconc2 (Vemacs_mule_charset_list, | |
1056 Fcons (make_number (id), Qnil)); | |
1057 } | |
1058 | |
1059 if (new_definition_p) | |
1060 { | |
1061 Vcharset_list = Fcons (args[charset_arg_name], Vcharset_list); | |
1062 Vcharset_ordered_list = nconc2 (Vcharset_ordered_list, | |
1063 Fcons (make_number (id), Qnil)); | |
1064 charset_ordered_list_tick++; | |
1065 } | |
1066 | |
1067 return Qnil; | |
1068 } | |
1069 | |
1070 | |
1071 /* Same as Fdefine_charset_internal but arguments are more convenient | |
1072 to call from C (typically in syms_of_charset). This can define a | |
1073 charset of `offset' method only. Return the ID of the new | |
1074 charset. */ | |
1075 | |
1076 static int | |
1077 define_charset_internal (name, dimension, code_space, min_code, max_code, | |
1078 iso_final, iso_revision, emacs_mule_id, | |
1079 ascii_compatible, supplementary, | |
1080 code_offset) | |
1081 Lisp_Object name; | |
1082 int dimension; | |
1083 unsigned char *code_space; | |
1084 unsigned min_code, max_code; | |
1085 int iso_final, iso_revision, emacs_mule_id; | |
1086 int ascii_compatible, supplementary; | |
1087 int code_offset; | |
1088 { | |
1089 Lisp_Object args[charset_arg_max]; | |
1090 Lisp_Object plist[14]; | |
600 Lisp_Object val; | 1091 Lisp_Object val; |
601 int charset; | 1092 int i; |
602 | 1093 |
603 /* This originally used a ?: operator, but reportedly the HP-UX | 1094 args[charset_arg_name] = name; |
604 compiler version HP92453-01 A.10.32.22 miscompiles that. */ | 1095 args[charset_arg_dimension] = make_number (dimension); |
605 if (SYMBOLP (charset_symbol) | 1096 val = Fmake_vector (make_number (8), make_number (0)); |
606 && VECTORP (val = Fget (charset_symbol, Qcharset)) | 1097 for (i = 0; i < 8; i++) |
607 && CHARSET_VALID_P (charset = | 1098 ASET (val, i, make_number (code_space[i])); |
608 XINT (XVECTOR (val)->contents[CHARSET_ID_IDX]))) | 1099 args[charset_arg_code_space] = val; |
609 return charset; | 1100 args[charset_arg_min_code] = make_number (min_code); |
610 else | 1101 args[charset_arg_max_code] = make_number (max_code); |
611 return -1; | 1102 args[charset_arg_iso_final] |
612 } | 1103 = (iso_final < 0 ? Qnil : make_number (iso_final)); |
613 | 1104 args[charset_arg_iso_revision] = make_number (iso_revision); |
614 /* Return an identification number for a new private charset of | 1105 args[charset_arg_emacs_mule_id] |
615 DIMENSION and WIDTH. If there's no more room for the new charset, | 1106 = (emacs_mule_id < 0 ? Qnil : make_number (emacs_mule_id)); |
616 return 0. */ | 1107 args[charset_arg_ascii_compatible_p] = ascii_compatible ? Qt : Qnil; |
617 Lisp_Object | 1108 args[charset_arg_supplementary_p] = supplementary ? Qt : Qnil; |
618 get_new_private_charset_id (dimension, width) | 1109 args[charset_arg_invalid_code] = Qnil; |
619 int dimension, width; | 1110 args[charset_arg_code_offset] = make_number (code_offset); |
620 { | 1111 args[charset_arg_map] = Qnil; |
621 int charset, from, to; | 1112 args[charset_arg_subset] = Qnil; |
622 | 1113 args[charset_arg_superset] = Qnil; |
623 if (dimension == 1) | 1114 args[charset_arg_unify_map] = Qnil; |
624 { | 1115 |
625 from = LEADING_CODE_EXT_11; | 1116 plist[0] = intern (":name"); |
626 to = LEADING_CODE_EXT_21; | 1117 plist[1] = args[charset_arg_name]; |
627 } | 1118 plist[2] = intern (":dimension"); |
628 else | 1119 plist[3] = args[charset_arg_dimension]; |
629 { | 1120 plist[4] = intern (":code-space"); |
630 from = LEADING_CODE_EXT_21; | 1121 plist[5] = args[charset_arg_code_space]; |
631 to = LEADING_CODE_EXT_MAX + 1; | 1122 plist[6] = intern (":iso-final-char"); |
632 } | 1123 plist[7] = args[charset_arg_iso_final]; |
633 | 1124 plist[8] = intern (":emacs-mule-id"); |
634 for (charset = from; charset < to; charset++) | 1125 plist[9] = args[charset_arg_emacs_mule_id]; |
635 if (!CHARSET_DEFINED_P (charset)) break; | 1126 plist[10] = intern (":ascii-compatible-p"); |
636 | 1127 plist[11] = args[charset_arg_ascii_compatible_p]; |
637 return make_number (charset < to ? charset : 0); | 1128 plist[12] = intern (":code-offset"); |
638 } | 1129 plist[13] = args[charset_arg_code_offset]; |
639 | 1130 |
640 DEFUN ("define-charset", Fdefine_charset, Sdefine_charset, 3, 3, 0, | 1131 args[charset_arg_plist] = Flist (14, plist); |
641 doc: /* Define CHARSET-ID as the identification number of CHARSET with INFO-VECTOR. | 1132 Fdefine_charset_internal (charset_arg_max, args); |
642 If CHARSET-ID is nil, it is decided automatically, which means CHARSET is | 1133 |
643 treated as a private charset. | 1134 return XINT (CHARSET_SYMBOL_ID (name)); |
644 INFO-VECTOR is a vector of the format: | 1135 } |
645 [DIMENSION CHARS WIDTH DIRECTION ISO-FINAL-CHAR ISO-GRAPHIC-PLANE | 1136 |
646 SHORT-NAME LONG-NAME DESCRIPTION] | 1137 |
647 The meanings of each elements is as follows: | 1138 DEFUN ("define-charset-alias", Fdefine_charset_alias, |
648 DIMENSION (integer) is the number of bytes to represent a character: 1 or 2. | 1139 Sdefine_charset_alias, 2, 2, 0, |
649 CHARS (integer) is the number of characters in a dimension: 94 or 96. | 1140 doc: /* Define ALIAS as an alias for charset CHARSET. */) |
650 WIDTH (integer) is the number of columns a character in the charset | 1141 (alias, charset) |
651 occupies on the screen: one of 0, 1, and 2. | 1142 Lisp_Object alias, charset; |
652 | 1143 { |
653 DIRECTION (integer) is the rendering direction of characters in the | 1144 Lisp_Object attr; |
654 charset when rendering. If 0, render from left to right, else | 1145 |
655 render from right to left. | 1146 CHECK_CHARSET_GET_ATTR (charset, attr); |
656 | 1147 Fputhash (alias, attr, Vcharset_hash_table); |
657 ISO-FINAL-CHAR (character) is the final character of the | 1148 Vcharset_list = Fcons (alias, Vcharset_list); |
658 corresponding ISO 2022 charset. | |
659 It may be -1 if the charset is internal use only. | |
660 | |
661 ISO-GRAPHIC-PLANE (integer) is the graphic plane to be invoked | |
662 while encoding to variants of ISO 2022 coding system, one of the | |
663 following: 0/graphic-plane-left(GL), 1/graphic-plane-right(GR). | |
664 It may be -1 if the charset is internal use only. | |
665 | |
666 SHORT-NAME (string) is the short name to refer to the charset. | |
667 | |
668 LONG-NAME (string) is the long name to refer to the charset. | |
669 | |
670 DESCRIPTION (string) is the description string of the charset. */) | |
671 (charset_id, charset_symbol, info_vector) | |
672 Lisp_Object charset_id, charset_symbol, info_vector; | |
673 { | |
674 Lisp_Object *vec; | |
675 | |
676 if (!NILP (charset_id)) | |
677 CHECK_NUMBER (charset_id); | |
678 CHECK_SYMBOL (charset_symbol); | |
679 CHECK_VECTOR (info_vector); | |
680 | |
681 if (! NILP (charset_id)) | |
682 { | |
683 if (! CHARSET_VALID_P (XINT (charset_id))) | |
684 error ("Invalid CHARSET: %d", XINT (charset_id)); | |
685 else if (CHARSET_DEFINED_P (XINT (charset_id))) | |
686 error ("Already defined charset: %d", XINT (charset_id)); | |
687 } | |
688 | |
689 vec = XVECTOR (info_vector)->contents; | |
690 if (XVECTOR (info_vector)->size != 9 | |
691 || !INTEGERP (vec[0]) || !(XINT (vec[0]) == 1 || XINT (vec[0]) == 2) | |
692 || !INTEGERP (vec[1]) || !(XINT (vec[1]) == 94 || XINT (vec[1]) == 96) | |
693 || !INTEGERP (vec[2]) || !(XINT (vec[2]) == 1 || XINT (vec[2]) == 2) | |
694 || !INTEGERP (vec[3]) || !(XINT (vec[3]) == 0 || XINT (vec[3]) == 1) | |
695 || !INTEGERP (vec[4]) | |
696 || !(XINT (vec[4]) == -1 || (XINT (vec[4]) >= '0' && XINT (vec[4]) <= '~')) | |
697 || !INTEGERP (vec[5]) | |
698 || !(XINT (vec[5]) == -1 || XINT (vec[5]) == 0 || XINT (vec[5]) == 1) | |
699 || !STRINGP (vec[6]) | |
700 || !STRINGP (vec[7]) | |
701 || !STRINGP (vec[8])) | |
702 error ("Invalid info-vector argument for defining charset %s", | |
703 SDATA (SYMBOL_NAME (charset_symbol))); | |
704 | |
705 if (NILP (charset_id)) | |
706 { | |
707 charset_id = get_new_private_charset_id (XINT (vec[0]), XINT (vec[2])); | |
708 if (XINT (charset_id) == 0) | |
709 error ("There's no room for a new private charset %s", | |
710 SDATA (SYMBOL_NAME (charset_symbol))); | |
711 } | |
712 | |
713 update_charset_table (charset_id, vec[0], vec[1], vec[2], vec[3], | |
714 vec[4], vec[5], vec[6], vec[7], vec[8]); | |
715 Fput (charset_symbol, Qcharset, CHARSET_TABLE_ENTRY (XINT (charset_id))); | |
716 CHARSET_SYMBOL (XINT (charset_id)) = charset_symbol; | |
717 Vcharset_list = Fcons (charset_symbol, Vcharset_list); | |
718 Fupdate_coding_systems_internal (); | |
719 return Qnil; | 1149 return Qnil; |
720 } | 1150 } |
721 | 1151 |
722 DEFUN ("generic-character-list", Fgeneric_character_list, | 1152 |
723 Sgeneric_character_list, 0, 0, 0, | 1153 DEFUN ("unibyte-charset", Funibyte_charset, Sunibyte_charset, 0, 0, 0, |
724 doc: /* Return a list of all possible generic characters. | 1154 doc: /* Return the unibyte charset (set by `set-unibyte-charset'). */) |
725 It includes a generic character for a charset not yet defined. */) | |
726 () | 1155 () |
727 { | 1156 { |
728 return Vgeneric_character_list; | 1157 return CHARSET_NAME (CHARSET_FROM_ID (charset_unibyte)); |
1158 } | |
1159 | |
1160 | |
1161 DEFUN ("set-unibyte-charset", Fset_unibyte_charset, Sset_unibyte_charset, | |
1162 1, 1, 0, | |
1163 doc: /* Set the unibyte charset to CHARSET. | |
1164 This determines how unibyte/multibyte conversion is done. See also | |
1165 function `unibyte-charset'. */) | |
1166 (charset) | |
1167 Lisp_Object charset; | |
1168 { | |
1169 struct charset *cs; | |
1170 int i, c; | |
1171 | |
1172 CHECK_CHARSET_GET_CHARSET (charset, cs); | |
1173 if (! cs->ascii_compatible_p | |
1174 || cs->dimension != 1) | |
1175 error ("Inappropriate unibyte charset: %s", SDATA (SYMBOL_NAME (charset))); | |
1176 charset_unibyte = cs->id; | |
1177 for (i = 128; i < 256; i++) | |
1178 { | |
1179 c = DECODE_CHAR (cs, i); | |
1180 unibyte_to_multibyte_table[i] = (c < 0 ? BYTE8_TO_CHAR (i) : c); | |
1181 } | |
1182 | |
1183 return Qnil; | |
1184 } | |
1185 | |
1186 | |
1187 DEFUN ("charset-plist", Fcharset_plist, Scharset_plist, 1, 1, 0, | |
1188 doc: /* Return the property list of CHARSET. */) | |
1189 (charset) | |
1190 Lisp_Object charset; | |
1191 { | |
1192 Lisp_Object attrs; | |
1193 | |
1194 CHECK_CHARSET_GET_ATTR (charset, attrs); | |
1195 return CHARSET_ATTR_PLIST (attrs); | |
1196 } | |
1197 | |
1198 | |
1199 DEFUN ("set-charset-plist", Fset_charset_plist, Sset_charset_plist, 2, 2, 0, | |
1200 doc: /* Set CHARSET's property list to PLIST. */) | |
1201 (charset, plist) | |
1202 Lisp_Object charset, plist; | |
1203 { | |
1204 Lisp_Object attrs; | |
1205 | |
1206 CHECK_CHARSET_GET_ATTR (charset, attrs); | |
1207 CHARSET_ATTR_PLIST (attrs) = plist; | |
1208 return plist; | |
1209 } | |
1210 | |
1211 | |
1212 DEFUN ("unify-charset", Funify_charset, Sunify_charset, 1, 3, 0, | |
1213 doc: /* Unify characters of CHARSET with Unicode. | |
1214 This means reading the relevant file and installing the table defined | |
1215 by CHARSET's `:unify-map' property. | |
1216 | |
1217 Optional second arg UNIFY-MAP is a file name string or a vector. It has | |
1218 the same meaning as the `:unify-map' attribute in the function | |
1219 `define-charset' (which see). | |
1220 | |
1221 Optional third argument DEUNIFY, if non-nil, means to de-unify CHARSET. */) | |
1222 (charset, unify_map, deunify) | |
1223 Lisp_Object charset, unify_map, deunify; | |
1224 { | |
1225 int id; | |
1226 struct charset *cs; | |
1227 | |
1228 CHECK_CHARSET_GET_ID (charset, id); | |
1229 cs = CHARSET_FROM_ID (id); | |
1230 if (CHARSET_METHOD (cs) == CHARSET_METHOD_MAP_DEFERRED) | |
1231 load_charset (cs); | |
1232 if (NILP (deunify) | |
1233 ? CHARSET_UNIFIED_P (cs) && ! NILP (CHARSET_DEUNIFIER (cs)) | |
1234 : ! CHARSET_UNIFIED_P (cs)) | |
1235 return Qnil; | |
1236 | |
1237 CHARSET_UNIFIED_P (cs) = 0; | |
1238 if (NILP (deunify)) | |
1239 { | |
1240 if (CHARSET_METHOD (cs) != CHARSET_METHOD_OFFSET) | |
1241 error ("Can't unify charset: %s", SDATA (SYMBOL_NAME (charset))); | |
1242 if (NILP (unify_map)) | |
1243 unify_map = CHARSET_UNIFY_MAP (cs); | |
1244 if (STRINGP (unify_map)) | |
1245 load_charset_map_from_file (cs, unify_map, 2); | |
1246 else if (VECTORP (unify_map)) | |
1247 load_charset_map_from_vector (cs, unify_map, 2); | |
1248 else if (NILP (unify_map)) | |
1249 error ("No unify-map for charset"); | |
1250 else | |
1251 error ("Bad unify-map arg"); | |
1252 CHARSET_UNIFIED_P (cs) = 1; | |
1253 } | |
1254 else if (CHAR_TABLE_P (Vchar_unify_table)) | |
1255 { | |
1256 int min_code = CHARSET_MIN_CODE (cs); | |
1257 int max_code = CHARSET_MAX_CODE (cs); | |
1258 int min_char = DECODE_CHAR (cs, min_code); | |
1259 int max_char = DECODE_CHAR (cs, max_code); | |
1260 | |
1261 char_table_set_range (Vchar_unify_table, min_char, max_char, Qnil); | |
1262 } | |
1263 | |
1264 return Qnil; | |
729 } | 1265 } |
730 | 1266 |
731 DEFUN ("get-unused-iso-final-char", Fget_unused_iso_final_char, | 1267 DEFUN ("get-unused-iso-final-char", Fget_unused_iso_final_char, |
732 Sget_unused_iso_final_char, 2, 2, 0, | 1268 Sget_unused_iso_final_char, 2, 2, 0, |
733 doc: /* Return an unsed ISO's final char for a charset of DIMENISION and CHARS. | 1269 doc: /* |
1270 Return an unsed ISO final char for a charset of DIMENISION and CHARS. | |
734 DIMENSION is the number of bytes to represent a character: 1 or 2. | 1271 DIMENSION is the number of bytes to represent a character: 1 or 2. |
735 CHARS is the number of characters in a dimension: 94 or 96. | 1272 CHARS is the number of characters in a dimension: 94 or 96. |
736 | 1273 |
737 This final char is for private use, thus the range is `0' (48) .. `?' (63). | 1274 This final char is for private use, thus the range is `0' (48) .. `?' (63). |
738 If there's no unused final char for the specified kind of charset, | 1275 If there's no unused final char for the specified kind of charset, |
742 { | 1279 { |
743 int final_char; | 1280 int final_char; |
744 | 1281 |
745 CHECK_NUMBER (dimension); | 1282 CHECK_NUMBER (dimension); |
746 CHECK_NUMBER (chars); | 1283 CHECK_NUMBER (chars); |
747 if (XINT (dimension) != 1 && XINT (dimension) != 2) | 1284 if (XINT (dimension) != 1 && XINT (dimension) != 2 && XINT (dimension) != 3) |
748 error ("Invalid charset dimension %d, it should be 1 or 2", | 1285 args_out_of_range_3 (dimension, make_number (1), make_number (3)); |
749 XINT (dimension)); | |
750 if (XINT (chars) != 94 && XINT (chars) != 96) | 1286 if (XINT (chars) != 94 && XINT (chars) != 96) |
751 error ("Invalid charset chars %d, it should be 94 or 96", | 1287 args_out_of_range_3 (chars, make_number (94), make_number (96)); |
752 XINT (chars)); | |
753 for (final_char = '0'; final_char <= '?'; final_char++) | 1288 for (final_char = '0'; final_char <= '?'; final_char++) |
754 { | 1289 if (ISO_CHARSET_TABLE (XINT (dimension), XINT (chars), final_char) < 0) |
755 if (ISO_CHARSET_TABLE (dimension, chars, make_number (final_char)) < 0) | 1290 break; |
756 break; | |
757 } | |
758 return (final_char <= '?' ? make_number (final_char) : Qnil); | 1291 return (final_char <= '?' ? make_number (final_char) : Qnil); |
759 } | 1292 } |
1293 | |
1294 static void | |
1295 check_iso_charset_parameter (dimension, chars, final_char) | |
1296 Lisp_Object dimension, chars, final_char; | |
1297 { | |
1298 CHECK_NATNUM (dimension); | |
1299 CHECK_NATNUM (chars); | |
1300 CHECK_NATNUM (final_char); | |
1301 | |
1302 if (XINT (dimension) > 3) | |
1303 error ("Invalid DIMENSION %d, it should be 1, 2, or 3", XINT (dimension)); | |
1304 if (XINT (chars) != 94 && XINT (chars) != 96) | |
1305 error ("Invalid CHARS %d, it should be 94 or 96", XINT (chars)); | |
1306 if (XINT (final_char) < '0' || XINT (final_char) > '~') | |
1307 error ("Invalid FINAL-CHAR %c, it should be `0'..`~'", XINT (chars)); | |
1308 } | |
1309 | |
760 | 1310 |
761 DEFUN ("declare-equiv-charset", Fdeclare_equiv_charset, Sdeclare_equiv_charset, | 1311 DEFUN ("declare-equiv-charset", Fdeclare_equiv_charset, Sdeclare_equiv_charset, |
762 4, 4, 0, | 1312 4, 4, 0, |
763 doc: /* Declare a charset of DIMENSION, CHARS, FINAL-CHAR is the same as CHARSET. | 1313 doc: /* |
764 CHARSET should be defined by `defined-charset' in advance. */) | 1314 Declare a charset of DIMENSION, CHARS, FINAL-CHAR is the same as CHARSET. |
765 (dimension, chars, final_char, charset_symbol) | 1315 CHARSET should be defined by `define-charset' in advance. */) |
766 Lisp_Object dimension, chars, final_char, charset_symbol; | 1316 (dimension, chars, final_char, charset) |
767 { | 1317 Lisp_Object dimension, chars, final_char, charset; |
768 int charset; | 1318 { |
769 | 1319 int id; |
770 CHECK_NUMBER (dimension); | 1320 |
771 CHECK_NUMBER (chars); | 1321 CHECK_CHARSET_GET_ID (charset, id); |
772 CHECK_NUMBER (final_char); | 1322 check_iso_charset_parameter (dimension, chars, final_char); |
773 CHECK_SYMBOL (charset_symbol); | 1323 |
774 | 1324 ISO_CHARSET_TABLE (XINT (dimension), XINT (chars), XINT (final_char)) = id; |
775 if (XINT (dimension) != 1 && XINT (dimension) != 2) | |
776 error ("Invalid DIMENSION %d, it should be 1 or 2", XINT (dimension)); | |
777 if (XINT (chars) != 94 && XINT (chars) != 96) | |
778 error ("Invalid CHARS %d, it should be 94 or 96", XINT (chars)); | |
779 if (XINT (final_char) < '0' || XFASTINT (final_char) > '~') | |
780 error ("Invalid FINAL-CHAR %c, it should be `0'..`~'", XINT (chars)); | |
781 if ((charset = get_charset_id (charset_symbol)) < 0) | |
782 error ("Invalid charset %s", SDATA (SYMBOL_NAME (charset_symbol))); | |
783 | |
784 ISO_CHARSET_TABLE (dimension, chars, final_char) = charset; | |
785 return Qnil; | 1325 return Qnil; |
786 } | 1326 } |
1327 | |
787 | 1328 |
788 /* Return information about charsets in the text at PTR of NBYTES | 1329 /* Return information about charsets in the text at PTR of NBYTES |
789 bytes, which are NCHARS characters. The value is: | 1330 bytes, which are NCHARS characters. The value is: |
790 | 1331 |
791 0: Each character is represented by one byte. This is always | 1332 0: Each character is represented by one byte. This is always |
792 true for unibyte text. | 1333 true for a unibyte string. For a multibyte string, true if |
793 1: No charsets other than ascii eight-bit-control, | 1334 it contains only ASCII characters. |
794 eight-bit-graphic, and latin-1 are found. | 1335 |
1336 1: No charsets other than ascii, control-1, and latin-1 are | |
1337 found. | |
1338 | |
795 2: Otherwise. | 1339 2: Otherwise. |
796 | 1340 */ |
797 In addition, if CHARSETS is nonzero, for each found charset N, set | |
798 CHARSETS[N] to 1. For that, callers should allocate CHARSETS | |
799 (MAX_CHARSET + 1 elements) in advance. It may lookup a translation | |
800 table TABLE if supplied. For invalid charsets, set CHARSETS[1] to | |
801 1 (note that there's no charset whose ID is 1). */ | |
802 | 1341 |
803 int | 1342 int |
804 find_charset_in_text (ptr, nchars, nbytes, charsets, table) | 1343 string_xstring_p (string) |
1344 Lisp_Object string; | |
1345 { | |
1346 const unsigned char *p = SDATA (string); | |
1347 const unsigned char *endp = p + SBYTES (string); | |
1348 struct charset *charset; | |
1349 | |
1350 if (SCHARS (string) == SBYTES (string)) | |
1351 return 0; | |
1352 | |
1353 charset = CHARSET_FROM_ID (charset_iso_8859_1); | |
1354 while (p < endp) | |
1355 { | |
1356 int c = STRING_CHAR_ADVANCE (p); | |
1357 | |
1358 /* Fixme: comparison of unsigned expression < 0 is always false */ | |
1359 if (ENCODE_CHAR (charset, c) < 0) | |
1360 return 2; | |
1361 } | |
1362 return 1; | |
1363 } | |
1364 | |
1365 | |
1366 /* Find charsets in the string at PTR of NCHARS and NBYTES. | |
1367 | |
1368 CHARSETS is a vector. Each element is a cons of CHARSET and | |
1369 FOUND-FLAG. CHARSET is a charset id, and FOUND-FLAG is nil or t. | |
1370 FOUND-FLAG t (or nil) means that the corresponding charset is | |
1371 already found (or not yet found). | |
1372 | |
1373 It may lookup a translation table TABLE if supplied. */ | |
1374 | |
1375 static void | |
1376 find_charsets_in_text (ptr, nchars, nbytes, charsets, table) | |
805 const unsigned char *ptr; | 1377 const unsigned char *ptr; |
806 int nchars, nbytes, *charsets; | 1378 int nchars, nbytes; |
807 Lisp_Object table; | 1379 Lisp_Object charsets, table; |
808 { | 1380 { |
1381 const unsigned char *pend = ptr + nbytes; | |
1382 int ncharsets = ASIZE (charsets); | |
1383 | |
809 if (nchars == nbytes) | 1384 if (nchars == nbytes) |
810 { | 1385 return; |
811 if (charsets && nbytes > 0) | 1386 |
812 { | 1387 while (ptr < pend) |
813 const unsigned char *endp = ptr + nbytes; | 1388 { |
814 int maskbits = 0; | 1389 int c = STRING_CHAR_ADVANCE (ptr); |
815 | 1390 int i; |
816 while (ptr < endp && maskbits != 7) | 1391 int all_found = 1; |
1392 Lisp_Object elt; | |
1393 | |
1394 if (!NILP (table)) | |
1395 c = translate_char (table, c); | |
1396 for (i = 0; i < ncharsets; i++) | |
1397 { | |
1398 elt = AREF (charsets, i); | |
1399 if (NILP (XCDR (elt))) | |
817 { | 1400 { |
818 maskbits |= (*ptr < 0x80 ? 1 : *ptr < 0xA0 ? 2 : 4); | 1401 struct charset *charset = CHARSET_FROM_ID (XINT (XCAR (elt))); |
819 ptr++; | 1402 |
1403 if (ENCODE_CHAR (charset, c) != CHARSET_INVALID_CODE (charset)) | |
1404 XSETCDR (elt, Qt); | |
1405 else | |
1406 all_found = 0; | |
820 } | 1407 } |
821 | 1408 } |
822 if (maskbits & 1) | 1409 if (all_found) |
823 charsets[CHARSET_ASCII] = 1; | 1410 break; |
824 if (maskbits & 2) | 1411 } |
825 charsets[CHARSET_8_BIT_CONTROL] = 1; | 1412 } |
826 if (maskbits & 4) | 1413 |
827 charsets[CHARSET_8_BIT_GRAPHIC] = 1; | 1414 /* Fixme: returns nil for unibyte. */ |
828 } | |
829 return 0; | |
830 } | |
831 else | |
832 { | |
833 int return_val = 1; | |
834 int bytes, charset, c1, c2; | |
835 | |
836 if (! CHAR_TABLE_P (table)) | |
837 table = Qnil; | |
838 | |
839 while (nchars-- > 0) | |
840 { | |
841 SPLIT_MULTIBYTE_SEQ (ptr, len, bytes, charset, c1, c2); | |
842 ptr += bytes; | |
843 | |
844 if (!CHARSET_DEFINED_P (charset)) | |
845 charset = 1; | |
846 else if (! NILP (table)) | |
847 { | |
848 int c = translate_char (table, -1, charset, c1, c2); | |
849 if (c >= 0) | |
850 charset = CHAR_CHARSET (c); | |
851 } | |
852 | |
853 if (return_val == 1 | |
854 && charset != CHARSET_ASCII | |
855 && charset != CHARSET_8_BIT_CONTROL | |
856 && charset != CHARSET_8_BIT_GRAPHIC | |
857 && charset != charset_latin_iso8859_1) | |
858 return_val = 2; | |
859 | |
860 if (charsets) | |
861 charsets[charset] = 1; | |
862 else if (return_val == 2) | |
863 break; | |
864 } | |
865 return return_val; | |
866 } | |
867 } | |
868 | |
869 DEFUN ("find-charset-region", Ffind_charset_region, Sfind_charset_region, | 1415 DEFUN ("find-charset-region", Ffind_charset_region, Sfind_charset_region, |
870 2, 3, 0, | 1416 2, 3, 0, |
871 doc: /* Return a list of charsets in the region between BEG and END. | 1417 doc: /* Return a list of charsets in the region between BEG and END. |
872 BEG and END are buffer positions. | 1418 BEG and END are buffer positions. |
873 Optional arg TABLE if non-nil is a translation table to look up. | 1419 Optional arg TABLE if non-nil is a translation table to look up. |
874 | 1420 |
875 If the region contains invalid multibyte characters, | |
876 `unknown' is included in the returned list. | |
877 | |
878 If the current buffer is unibyte, the returned list may contain | 1421 If the current buffer is unibyte, the returned list may contain |
879 only `ascii', `eight-bit-control', and `eight-bit-graphic'. */) | 1422 only `ascii', `eight-bit-control', and `eight-bit-graphic'. */) |
880 (beg, end, table) | 1423 (beg, end, table) |
881 Lisp_Object beg, end, table; | 1424 Lisp_Object beg, end, table; |
882 { | 1425 { |
883 int charsets[MAX_CHARSET + 1]; | 1426 Lisp_Object charsets; |
884 int from, from_byte, to, stop, stop_byte, i; | 1427 int from, from_byte, to, stop, stop_byte, i; |
885 Lisp_Object val; | 1428 Lisp_Object val; |
886 | 1429 |
887 validate_region (&beg, &end); | 1430 validate_region (&beg, &end); |
888 from = XFASTINT (beg); | 1431 from = XFASTINT (beg); |
896 else | 1439 else |
897 stop_byte = CHAR_TO_BYTE (stop); | 1440 stop_byte = CHAR_TO_BYTE (stop); |
898 | 1441 |
899 from_byte = CHAR_TO_BYTE (from); | 1442 from_byte = CHAR_TO_BYTE (from); |
900 | 1443 |
901 bzero (charsets, (MAX_CHARSET + 1) * sizeof (int)); | 1444 charsets = Fmake_vector (make_number (charset_table_used), Qnil); |
1445 for (i = 0; i < charset_table_used; i++) | |
1446 ASET (charsets, i, Fcons (make_number (i), Qnil)); | |
1447 | |
902 while (1) | 1448 while (1) |
903 { | 1449 { |
904 find_charset_in_text (BYTE_POS_ADDR (from_byte), stop - from, | 1450 find_charsets_in_text (BYTE_POS_ADDR (from_byte), stop - from, |
905 stop_byte - from_byte, charsets, table); | 1451 stop_byte - from_byte, charsets, table); |
906 if (stop < to) | 1452 if (stop < to) |
907 { | 1453 { |
908 from = stop, from_byte = stop_byte; | 1454 from = stop, from_byte = stop_byte; |
909 stop = to, stop_byte = CHAR_TO_BYTE (stop); | 1455 stop = to, stop_byte = CHAR_TO_BYTE (stop); |
910 } | 1456 } |
911 else | 1457 else |
912 break; | 1458 break; |
913 } | 1459 } |
914 | 1460 |
915 val = Qnil; | 1461 val = Qnil; |
916 if (charsets[1]) | 1462 for (i = charset_table_used - 1; i >= 0; i--) |
917 val = Fcons (Qunknown, val); | 1463 if (!NILP (XCDR (AREF (charsets, i)))) |
918 for (i = MAX_CHARSET; i >= MIN_CHARSET_OFFICIAL_DIMENSION1; i--) | 1464 val = Fcons (CHARSET_NAME (charset_table + i), val); |
919 if (charsets[i]) | |
920 val = Fcons (CHARSET_SYMBOL (i), val); | |
921 if (charsets[0]) | |
922 val = Fcons (Qascii, val); | |
923 return val; | 1465 return val; |
924 } | 1466 } |
925 | 1467 |
1468 /* Fixme: returns nil for unibyte. */ | |
926 DEFUN ("find-charset-string", Ffind_charset_string, Sfind_charset_string, | 1469 DEFUN ("find-charset-string", Ffind_charset_string, Sfind_charset_string, |
927 1, 2, 0, | 1470 1, 2, 0, |
928 doc: /* Return a list of charsets in STR. | 1471 doc: /* Return a list of charsets in STR. |
929 Optional arg TABLE if non-nil is a translation table to look up. | 1472 Optional arg TABLE if non-nil is a translation table to look up. |
930 | 1473 |
931 If the string contains invalid multibyte characters, | |
932 `unknown' is included in the returned list. | |
933 | |
934 If STR is unibyte, the returned list may contain | 1474 If STR is unibyte, the returned list may contain |
935 only `ascii', `eight-bit-control', and `eight-bit-graphic'. */) | 1475 only `ascii', `eight-bit-control', and `eight-bit-graphic'. */) |
936 (str, table) | 1476 (str, table) |
937 Lisp_Object str, table; | 1477 Lisp_Object str, table; |
938 { | 1478 { |
939 int charsets[MAX_CHARSET + 1]; | 1479 Lisp_Object charsets; |
940 int i; | 1480 int i; |
941 Lisp_Object val; | 1481 Lisp_Object val; |
942 | 1482 |
943 CHECK_STRING (str); | 1483 CHECK_STRING (str); |
944 | 1484 |
945 bzero (charsets, (MAX_CHARSET + 1) * sizeof (int)); | 1485 charsets = Fmake_vector (make_number (charset_table_used), Qnil); |
946 find_charset_in_text (SDATA (str), SCHARS (str), | 1486 for (i = 0; i < charset_table_used; i++) |
947 SBYTES (str), charsets, table); | 1487 ASET (charsets, i, Fcons (make_number (i), Qnil)); |
1488 find_charsets_in_text (SDATA (str), SCHARS (str), SBYTES (str), | |
1489 charsets, table); | |
948 | 1490 |
949 val = Qnil; | 1491 val = Qnil; |
950 if (charsets[1]) | 1492 for (i = charset_table_used - 1; i >= 0; i--) |
951 val = Fcons (Qunknown, val); | 1493 if (!NILP (XCDR (AREF (charsets, i)))) |
952 for (i = MAX_CHARSET; i >= MIN_CHARSET_OFFICIAL_DIMENSION1; i--) | 1494 val = Fcons (CHARSET_NAME (charset_table + i), val); |
953 if (charsets[i]) | |
954 val = Fcons (CHARSET_SYMBOL (i), val); | |
955 if (charsets[0]) | |
956 val = Fcons (Qascii, val); | |
957 return val; | 1495 return val; |
958 } | 1496 } |
959 | 1497 |
960 | 1498 |
961 DEFUN ("make-char-internal", Fmake_char_internal, Smake_char_internal, 1, 3, 0, | 1499 |
962 doc: /* Return a character made from arguments. | 1500 /* Return a character correponding to the code-point CODE of |
963 Internal use only. */) | 1501 CHARSET. */ |
964 (charset, code1, code2) | 1502 |
965 Lisp_Object charset, code1, code2; | 1503 int |
966 { | 1504 decode_char (charset, code) |
967 int charset_id, c1, c2; | 1505 struct charset *charset; |
968 | 1506 unsigned code; |
969 CHECK_NUMBER (charset); | 1507 { |
970 charset_id = XINT (charset); | 1508 int c, char_index; |
971 if (!CHARSET_DEFINED_P (charset_id)) | 1509 enum charset_method method = CHARSET_METHOD (charset); |
972 error ("Invalid charset ID: %d", XINT (charset)); | 1510 |
973 | 1511 if (code < CHARSET_MIN_CODE (charset) || code > CHARSET_MAX_CODE (charset)) |
1512 return -1; | |
1513 | |
1514 if (method == CHARSET_METHOD_MAP_DEFERRED) | |
1515 { | |
1516 load_charset (charset); | |
1517 method = CHARSET_METHOD (charset); | |
1518 } | |
1519 | |
1520 if (method == CHARSET_METHOD_SUBSET) | |
1521 { | |
1522 Lisp_Object subset_info; | |
1523 | |
1524 subset_info = CHARSET_SUBSET (charset); | |
1525 charset = CHARSET_FROM_ID (XFASTINT (AREF (subset_info, 0))); | |
1526 code -= XINT (AREF (subset_info, 3)); | |
1527 if (code < XFASTINT (AREF (subset_info, 1)) | |
1528 || code > XFASTINT (AREF (subset_info, 2))) | |
1529 c = -1; | |
1530 else | |
1531 c = DECODE_CHAR (charset, code); | |
1532 } | |
1533 else if (method == CHARSET_METHOD_SUPERSET) | |
1534 { | |
1535 Lisp_Object parents; | |
1536 | |
1537 parents = CHARSET_SUPERSET (charset); | |
1538 c = -1; | |
1539 for (; CONSP (parents); parents = XCDR (parents)) | |
1540 { | |
1541 int id = XINT (XCAR (XCAR (parents))); | |
1542 int code_offset = XINT (XCDR (XCAR (parents))); | |
1543 unsigned this_code = code - code_offset; | |
1544 | |
1545 charset = CHARSET_FROM_ID (id); | |
1546 if ((c = DECODE_CHAR (charset, this_code)) >= 0) | |
1547 break; | |
1548 } | |
1549 } | |
1550 else | |
1551 { | |
1552 char_index = CODE_POINT_TO_INDEX (charset, code); | |
1553 if (char_index < 0) | |
1554 return -1; | |
1555 | |
1556 if (method == CHARSET_METHOD_MAP) | |
1557 { | |
1558 Lisp_Object decoder; | |
1559 | |
1560 decoder = CHARSET_DECODER (charset); | |
1561 if (! VECTORP (decoder)) | |
1562 return -1; | |
1563 c = XINT (AREF (decoder, char_index)); | |
1564 } | |
1565 else | |
1566 { | |
1567 c = char_index + CHARSET_CODE_OFFSET (charset); | |
1568 } | |
1569 } | |
1570 | |
1571 if (CHARSET_UNIFIED_P (charset) | |
1572 && c >= 0) | |
1573 { | |
1574 MAYBE_UNIFY_CHAR (c); | |
1575 } | |
1576 | |
1577 return c; | |
1578 } | |
1579 | |
1580 /* Variable used temporarily by the macro ENCODE_CHAR. */ | |
1581 Lisp_Object charset_work; | |
1582 | |
1583 /* Return a code-point of CHAR in CHARSET. If CHAR doesn't belong to | |
1584 CHARSET, return CHARSET_INVALID_CODE (CHARSET). If STRICT is true, | |
1585 use CHARSET's strict_max_char instead of max_char. */ | |
1586 | |
1587 unsigned | |
1588 encode_char (charset, c) | |
1589 struct charset *charset; | |
1590 int c; | |
1591 { | |
1592 unsigned code; | |
1593 enum charset_method method = CHARSET_METHOD (charset); | |
1594 | |
1595 if (CHARSET_UNIFIED_P (charset)) | |
1596 { | |
1597 Lisp_Object deunifier, deunified; | |
1598 | |
1599 deunifier = CHARSET_DEUNIFIER (charset); | |
1600 if (! CHAR_TABLE_P (deunifier)) | |
1601 { | |
1602 Funify_charset (CHARSET_NAME (charset), Qnil, Qnil); | |
1603 deunifier = CHARSET_DEUNIFIER (charset); | |
1604 } | |
1605 deunified = CHAR_TABLE_REF (deunifier, c); | |
1606 if (! NILP (deunified)) | |
1607 c = XINT (deunified); | |
1608 } | |
1609 | |
1610 if (method == CHARSET_METHOD_SUBSET) | |
1611 { | |
1612 Lisp_Object subset_info; | |
1613 struct charset *this_charset; | |
1614 | |
1615 subset_info = CHARSET_SUBSET (charset); | |
1616 this_charset = CHARSET_FROM_ID (XFASTINT (AREF (subset_info, 0))); | |
1617 code = ENCODE_CHAR (this_charset, c); | |
1618 if (code == CHARSET_INVALID_CODE (this_charset) | |
1619 || code < XFASTINT (AREF (subset_info, 1)) | |
1620 || code > XFASTINT (AREF (subset_info, 2))) | |
1621 return CHARSET_INVALID_CODE (charset); | |
1622 code += XINT (AREF (subset_info, 3)); | |
1623 return code; | |
1624 } | |
1625 | |
1626 if (method == CHARSET_METHOD_SUPERSET) | |
1627 { | |
1628 Lisp_Object parents; | |
1629 | |
1630 parents = CHARSET_SUPERSET (charset); | |
1631 for (; CONSP (parents); parents = XCDR (parents)) | |
1632 { | |
1633 int id = XINT (XCAR (XCAR (parents))); | |
1634 int code_offset = XINT (XCDR (XCAR (parents))); | |
1635 struct charset *this_charset = CHARSET_FROM_ID (id); | |
1636 | |
1637 code = ENCODE_CHAR (this_charset, c); | |
1638 if (code != CHARSET_INVALID_CODE (this_charset)) | |
1639 return code + code_offset; | |
1640 } | |
1641 return CHARSET_INVALID_CODE (charset); | |
1642 } | |
1643 | |
1644 if (! CHARSET_FAST_MAP_REF ((c), charset->fast_map) | |
1645 || c < CHARSET_MIN_CHAR (charset) || c > CHARSET_MAX_CHAR (charset)) | |
1646 return CHARSET_INVALID_CODE (charset); | |
1647 | |
1648 if (method == CHARSET_METHOD_MAP_DEFERRED) | |
1649 { | |
1650 load_charset (charset); | |
1651 method = CHARSET_METHOD (charset); | |
1652 } | |
1653 | |
1654 if (method == CHARSET_METHOD_MAP) | |
1655 { | |
1656 Lisp_Object encoder; | |
1657 Lisp_Object val; | |
1658 | |
1659 encoder = CHARSET_ENCODER (charset); | |
1660 if (! CHAR_TABLE_P (CHARSET_ENCODER (charset))) | |
1661 return CHARSET_INVALID_CODE (charset); | |
1662 val = CHAR_TABLE_REF (encoder, c); | |
1663 if (NILP (val)) | |
1664 return CHARSET_INVALID_CODE (charset); | |
1665 code = XINT (val); | |
1666 if (! CHARSET_COMPACT_CODES_P (charset)) | |
1667 code = INDEX_TO_CODE_POINT (charset, code); | |
1668 } | |
1669 else /* method == CHARSET_METHOD_OFFSET */ | |
1670 { | |
1671 code = c - CHARSET_CODE_OFFSET (charset); | |
1672 code = INDEX_TO_CODE_POINT (charset, code); | |
1673 } | |
1674 | |
1675 return code; | |
1676 } | |
1677 | |
1678 | |
1679 DEFUN ("decode-char", Fdecode_char, Sdecode_char, 2, 3, 0, | |
1680 doc: /* Decode the pair of CHARSET and CODE-POINT into a character. | |
1681 Return nil if CODE-POINT is not valid in CHARSET. | |
1682 | |
1683 CODE-POINT may be a cons (HIGHER-16-BIT-VALUE . LOWER-16-BIT-VALUE). | |
1684 | |
1685 Optional argument RESTRICTION specifies a way to map the pair of CCS | |
1686 and CODE-POINT to a chracter. Currently not supported and just ignored. */) | |
1687 (charset, code_point, restriction) | |
1688 Lisp_Object charset, code_point, restriction; | |
1689 { | |
1690 int c, id; | |
1691 unsigned code; | |
1692 struct charset *charsetp; | |
1693 | |
1694 CHECK_CHARSET_GET_ID (charset, id); | |
1695 if (CONSP (code_point)) | |
1696 { | |
1697 CHECK_NATNUM_CAR (code_point); | |
1698 CHECK_NATNUM_CDR (code_point); | |
1699 code = (XINT (XCAR (code_point)) << 16) | (XINT (XCDR (code_point))); | |
1700 } | |
1701 else | |
1702 { | |
1703 CHECK_NATNUM (code_point); | |
1704 code = XINT (code_point); | |
1705 } | |
1706 charsetp = CHARSET_FROM_ID (id); | |
1707 c = DECODE_CHAR (charsetp, code); | |
1708 return (c >= 0 ? make_number (c) : Qnil); | |
1709 } | |
1710 | |
1711 | |
1712 DEFUN ("encode-char", Fencode_char, Sencode_char, 2, 3, 0, | |
1713 doc: /* Encode the character CH into a code-point of CHARSET. | |
1714 Return nil if CHARSET doesn't include CH. | |
1715 | |
1716 Optional argument RESTRICTION specifies a way to map CHAR to a | |
1717 code-point in CCS. Currently not supported and just ignored. */) | |
1718 (ch, charset, restriction) | |
1719 Lisp_Object ch, charset, restriction; | |
1720 { | |
1721 int id; | |
1722 unsigned code; | |
1723 struct charset *charsetp; | |
1724 | |
1725 CHECK_CHARSET_GET_ID (charset, id); | |
1726 CHECK_NATNUM (ch); | |
1727 charsetp = CHARSET_FROM_ID (id); | |
1728 code = ENCODE_CHAR (charsetp, XINT (ch)); | |
1729 if (code == CHARSET_INVALID_CODE (charsetp)) | |
1730 return Qnil; | |
1731 if (code > 0x7FFFFFF) | |
1732 return Fcons (make_number (code >> 16), make_number (code & 0xFFFF)); | |
1733 return make_number (code); | |
1734 } | |
1735 | |
1736 | |
1737 DEFUN ("make-char", Fmake_char, Smake_char, 1, 5, 0, | |
1738 doc: | |
1739 /* Return a character of CHARSET whose position codes are CODEn. | |
1740 | |
1741 CODE1 through CODE4 are optional, but if you don't supply sufficient | |
1742 position codes, it is assumed that the minimum code in each dimension | |
1743 is specified. */) | |
1744 (charset, code1, code2, code3, code4) | |
1745 Lisp_Object charset, code1, code2, code3, code4; | |
1746 { | |
1747 int id, dimension; | |
1748 struct charset *charsetp; | |
1749 unsigned code; | |
1750 int c; | |
1751 | |
1752 CHECK_CHARSET_GET_ID (charset, id); | |
1753 charsetp = CHARSET_FROM_ID (id); | |
1754 | |
1755 dimension = CHARSET_DIMENSION (charsetp); | |
974 if (NILP (code1)) | 1756 if (NILP (code1)) |
975 c1 = 0; | 1757 code = (CHARSET_ASCII_COMPATIBLE_P (charsetp) |
1758 ? 0 : CHARSET_MIN_CODE (charsetp)); | |
976 else | 1759 else |
977 { | 1760 { |
978 CHECK_NUMBER (code1); | 1761 CHECK_NATNUM (code1); |
979 c1 = XINT (code1); | 1762 if (XFASTINT (code1) >= 0x100) |
980 } | 1763 args_out_of_range (make_number (0xFF), code1); |
981 if (NILP (code2)) | 1764 code = XFASTINT (code1); |
982 c2 = 0; | 1765 |
983 else | 1766 if (dimension > 1) |
984 { | 1767 { |
985 CHECK_NUMBER (code2); | 1768 code <<= 8; |
986 c2 = XINT (code2); | 1769 if (NILP (code2)) |
987 } | 1770 code |= charsetp->code_space[(dimension - 2) * 4]; |
988 | 1771 else |
989 if (charset_id == CHARSET_ASCII) | 1772 { |
990 { | 1773 CHECK_NATNUM (code2); |
991 if (c1 < 0 || c1 > 0x7F) | 1774 if (XFASTINT (code2) >= 0x100) |
992 goto invalid_code_posints; | 1775 args_out_of_range (make_number (0xFF), code2); |
993 return make_number (c1); | 1776 code |= XFASTINT (code2); |
994 } | 1777 } |
995 else if (charset_id == CHARSET_8_BIT_CONTROL) | 1778 |
996 { | 1779 if (dimension > 2) |
997 if (NILP (code1)) | 1780 { |
998 c1 = 0x80; | 1781 code <<= 8; |
999 else if (c1 < 0x80 || c1 > 0x9F) | 1782 if (NILP (code3)) |
1000 goto invalid_code_posints; | 1783 code |= charsetp->code_space[(dimension - 3) * 4]; |
1001 return make_number (c1); | 1784 else |
1002 } | 1785 { |
1003 else if (charset_id == CHARSET_8_BIT_GRAPHIC) | 1786 CHECK_NATNUM (code3); |
1004 { | 1787 if (XFASTINT (code3) >= 0x100) |
1005 if (NILP (code1)) | 1788 args_out_of_range (make_number (0xFF), code3); |
1006 c1 = 0xA0; | 1789 code |= XFASTINT (code3); |
1007 else if (c1 < 0xA0 || c1 > 0xFF) | 1790 } |
1008 goto invalid_code_posints; | 1791 |
1009 return make_number (c1); | 1792 if (dimension > 3) |
1010 } | 1793 { |
1011 else if (c1 < 0 || c1 > 0xFF || c2 < 0 || c2 > 0xFF) | 1794 code <<= 8; |
1012 goto invalid_code_posints; | 1795 if (NILP (code4)) |
1013 c1 &= 0x7F; | 1796 code |= charsetp->code_space[0]; |
1014 c2 &= 0x7F; | 1797 else |
1015 if (c1 == 0 | 1798 { |
1016 ? c2 != 0 | 1799 CHECK_NATNUM (code4); |
1017 : (c2 == 0 | 1800 if (XFASTINT (code4) >= 0x100) |
1018 ? !CHAR_COMPONENTS_VALID_P (charset_id, c1, 0x20) | 1801 args_out_of_range (make_number (0xFF), code4); |
1019 : !CHAR_COMPONENTS_VALID_P (charset_id, c1, c2))) | 1802 code |= XFASTINT (code4); |
1020 goto invalid_code_posints; | 1803 } |
1021 return make_number (MAKE_CHAR (charset_id, c1, c2)); | 1804 } |
1022 | 1805 } |
1023 invalid_code_posints: | 1806 } |
1024 error ("Invalid code points for charset ID %d: %d %d", charset_id, c1, c2); | 1807 } |
1025 } | 1808 |
1026 | 1809 if (CHARSET_ISO_FINAL (charsetp) >= 0) |
1810 code &= 0x7F7F7F7F; | |
1811 c = DECODE_CHAR (charsetp, code); | |
1812 if (c < 0) | |
1813 error ("Invalid code(s)"); | |
1814 return make_number (c); | |
1815 } | |
1816 | |
1817 | |
1818 /* Return the first charset in CHARSET_LIST that contains C. | |
1819 CHARSET_LIST is a list of charset IDs. If it is nil, use | |
1820 Vcharset_ordered_list. */ | |
1821 | |
1822 struct charset * | |
1823 char_charset (c, charset_list, code_return) | |
1824 int c; | |
1825 Lisp_Object charset_list; | |
1826 unsigned *code_return; | |
1827 { | |
1828 if (NILP (charset_list)) | |
1829 charset_list = Vcharset_ordered_list; | |
1830 | |
1831 while (CONSP (charset_list)) | |
1832 { | |
1833 struct charset *charset = CHARSET_FROM_ID (XINT (XCAR (charset_list))); | |
1834 unsigned code = ENCODE_CHAR (charset, c); | |
1835 | |
1836 if (code != CHARSET_INVALID_CODE (charset)) | |
1837 { | |
1838 if (code_return) | |
1839 *code_return = code; | |
1840 return charset; | |
1841 } | |
1842 charset_list = XCDR (charset_list); | |
1843 } | |
1844 return NULL; | |
1845 } | |
1846 | |
1847 | |
1848 /* Fixme: `unknown' can't happen now? */ | |
1027 DEFUN ("split-char", Fsplit_char, Ssplit_char, 1, 1, 0, | 1849 DEFUN ("split-char", Fsplit_char, Ssplit_char, 1, 1, 0, |
1028 doc: /* Return list of charset and one or two position-codes of CHAR. | 1850 doc: /*Return list of charset and one to three position-codes of CHAR. |
1029 If CHAR is invalid as a character code, | 1851 If CHAR is invalid as a character code, return a list `(unknown CHAR)'. */) |
1030 return a list of symbol `unknown' and CHAR. */) | |
1031 (ch) | 1852 (ch) |
1032 Lisp_Object ch; | 1853 Lisp_Object ch; |
1033 { | 1854 { |
1034 int c, charset, c1, c2; | 1855 struct charset *charset; |
1035 | 1856 int c, dimension; |
1036 CHECK_NUMBER (ch); | 1857 unsigned code; |
1858 Lisp_Object val; | |
1859 | |
1860 CHECK_CHARACTER (ch); | |
1037 c = XFASTINT (ch); | 1861 c = XFASTINT (ch); |
1038 if (!CHAR_VALID_P (c, 1)) | 1862 charset = CHAR_CHARSET (c); |
1039 return Fcons (Qunknown, Fcons (ch, Qnil)); | 1863 if (! charset) |
1040 SPLIT_CHAR (XFASTINT (ch), charset, c1, c2); | 1864 return Fcons (intern ("unknown"), Fcons (ch, Qnil)); |
1041 return (c2 >= 0 | 1865 |
1042 ? Fcons (CHARSET_SYMBOL (charset), | 1866 code = ENCODE_CHAR (charset, c); |
1043 Fcons (make_number (c1), Fcons (make_number (c2), Qnil))) | 1867 if (code == CHARSET_INVALID_CODE (charset)) |
1044 : Fcons (CHARSET_SYMBOL (charset), Fcons (make_number (c1), Qnil))); | 1868 abort (); |
1045 } | 1869 dimension = CHARSET_DIMENSION (charset); |
1870 val = (dimension == 1 ? Fcons (make_number (code), Qnil) | |
1871 : dimension == 2 ? Fcons (make_number (code >> 8), | |
1872 Fcons (make_number (code & 0xFF), Qnil)) | |
1873 : Fcons (make_number (code >> 16), | |
1874 Fcons (make_number ((code >> 8) & 0xFF), | |
1875 Fcons (make_number (code & 0xFF), Qnil)))); | |
1876 return Fcons (CHARSET_NAME (charset), val); | |
1877 } | |
1878 | |
1046 | 1879 |
1047 DEFUN ("char-charset", Fchar_charset, Schar_charset, 1, 1, 0, | 1880 DEFUN ("char-charset", Fchar_charset, Schar_charset, 1, 1, 0, |
1048 doc: /* Return charset of CHAR. */) | 1881 doc: /* Return the charset of highest priority that contains CHAR. */) |
1049 (ch) | 1882 (ch) |
1050 Lisp_Object ch; | 1883 Lisp_Object ch; |
1051 { | 1884 { |
1052 CHECK_NUMBER (ch); | 1885 struct charset *charset; |
1053 | 1886 |
1054 return CHARSET_SYMBOL (CHAR_CHARSET (XINT (ch))); | 1887 CHECK_CHARACTER (ch); |
1055 } | 1888 charset = CHAR_CHARSET (XINT (ch)); |
1889 return (CHARSET_NAME (charset)); | |
1890 } | |
1891 | |
1056 | 1892 |
1057 DEFUN ("charset-after", Fcharset_after, Scharset_after, 0, 1, 0, | 1893 DEFUN ("charset-after", Fcharset_after, Scharset_after, 0, 1, 0, |
1058 doc: /* Return charset of a character in the current buffer at position POS. | 1894 doc: /* |
1895 Return charset of a character in the current buffer at position POS. | |
1059 If POS is nil, it defauls to the current point. | 1896 If POS is nil, it defauls to the current point. |
1060 If POS is out of range, the value is nil. */) | 1897 If POS is out of range, the value is nil. */) |
1061 (pos) | 1898 (pos) |
1062 Lisp_Object pos; | 1899 Lisp_Object pos; |
1063 { | 1900 { |
1064 Lisp_Object ch; | 1901 Lisp_Object ch; |
1065 int charset; | 1902 struct charset *charset; |
1066 | 1903 |
1067 ch = Fchar_after (pos); | 1904 ch = Fchar_after (pos); |
1068 if (! INTEGERP (ch)) | 1905 if (! INTEGERP (ch)) |
1069 return ch; | 1906 return ch; |
1070 charset = CHAR_CHARSET (XINT (ch)); | 1907 charset = CHAR_CHARSET (XINT (ch)); |
1071 return CHARSET_SYMBOL (charset); | 1908 return (CHARSET_NAME (charset)); |
1072 } | 1909 } |
1910 | |
1073 | 1911 |
1074 DEFUN ("iso-charset", Fiso_charset, Siso_charset, 3, 3, 0, | 1912 DEFUN ("iso-charset", Fiso_charset, Siso_charset, 3, 3, 0, |
1075 doc: /* Return charset of ISO's specification DIMENSION, CHARS, and FINAL-CHAR. | 1913 doc: /* |
1914 Return charset of ISO's specification DIMENSION, CHARS, and FINAL-CHAR. | |
1076 | 1915 |
1077 ISO 2022's designation sequence (escape sequence) distinguishes charsets | 1916 ISO 2022's designation sequence (escape sequence) distinguishes charsets |
1078 by their DIMENSION, CHARS, and FINAL-CHAR, | 1917 by their DIMENSION, CHARS, and FINAL-CHAR, |
1079 where as Emacs distinguishes them by charset symbol. | 1918 where as Emacs distinguishes them by charset symbol. |
1080 See the documentation of the function `charset-info' for the meanings of | 1919 See the documentation of the function `charset-info' for the meanings of |
1081 DIMENSION, CHARS, and FINAL-CHAR. */) | 1920 DIMENSION, CHARS, and FINAL-CHAR. */) |
1082 (dimension, chars, final_char) | 1921 (dimension, chars, final_char) |
1083 Lisp_Object dimension, chars, final_char; | 1922 Lisp_Object dimension, chars, final_char; |
1084 { | 1923 { |
1085 int charset; | 1924 int id; |
1086 | 1925 |
1087 CHECK_NUMBER (dimension); | 1926 check_iso_charset_parameter (dimension, chars, final_char); |
1088 CHECK_NUMBER (chars); | 1927 id = ISO_CHARSET_TABLE (XFASTINT (dimension), XFASTINT (chars), |
1089 CHECK_NUMBER (final_char); | 1928 XFASTINT (final_char)); |
1090 | 1929 return (id >= 0 ? CHARSET_NAME (CHARSET_FROM_ID (id)) : Qnil); |
1091 if ((charset = ISO_CHARSET_TABLE (dimension, chars, final_char)) < 0) | 1930 } |
1092 return Qnil; | 1931 |
1093 return CHARSET_SYMBOL (charset); | 1932 |
1094 } | 1933 DEFUN ("clear-charset-maps", Fclear_charset_maps, Sclear_charset_maps, |
1095 | 1934 0, 0, 0, |
1096 /* If GENERICP is nonzero, return nonzero iff C is a valid normal or | 1935 doc: /* |
1097 generic character. If GENERICP is zero, return nonzero iff C is a | 1936 Clear encoder and decoder of charsets that are loaded from mapfiles. */) |
1098 valid normal character. Do not call this function directly, | 1937 () |
1099 instead use macro CHAR_VALID_P. */ | 1938 { |
1100 int | 1939 int i; |
1101 char_valid_p (c, genericp) | 1940 struct charset *charset; |
1102 int c, genericp; | 1941 Lisp_Object attrs; |
1103 { | 1942 |
1104 int charset, c1, c2; | 1943 for (i = 0; i < charset_table_used; i++) |
1105 | 1944 { |
1106 if (c < 0 || c >= MAX_CHAR) | 1945 charset = CHARSET_FROM_ID (i); |
1107 return 0; | 1946 attrs = CHARSET_ATTRIBUTES (charset); |
1108 if (SINGLE_BYTE_CHAR_P (c)) | 1947 |
1109 return 1; | 1948 if (CHARSET_METHOD (charset) == CHARSET_METHOD_MAP) |
1110 SPLIT_CHAR (c, charset, c1, c2); | 1949 { |
1111 if (genericp) | 1950 CHARSET_ATTR_DECODER (attrs) = Qnil; |
1112 { | 1951 CHARSET_ATTR_ENCODER (attrs) = Qnil; |
1113 if (c1) | 1952 CHARSET_METHOD (charset) = CHARSET_METHOD_MAP_DEFERRED; |
1114 { | 1953 } |
1115 if (c2 <= 0) c2 = 0x20; | 1954 |
1116 } | 1955 if (CHARSET_UNIFIED_P (charset)) |
1117 else | 1956 CHARSET_ATTR_DEUNIFIER (attrs) = Qnil; |
1118 { | 1957 } |
1119 if (c2 <= 0) c1 = c2 = 0x20; | 1958 |
1120 } | 1959 if (CHAR_TABLE_P (Vchar_unified_charset_table)) |
1121 } | 1960 { |
1122 return (CHARSET_DEFINED_P (charset) | 1961 Foptimize_char_table (Vchar_unified_charset_table); |
1123 && CHAR_COMPONENTS_VALID_P (charset, c1, c2)); | 1962 Vchar_unify_table = Vchar_unified_charset_table; |
1124 } | 1963 Vchar_unified_charset_table = Qnil; |
1125 | 1964 } |
1126 DEFUN ("char-valid-p", Fchar_valid_p, Schar_valid_p, 1, 2, 0, | 1965 |
1127 doc: /* Return t if OBJECT is a valid normal character. | 1966 return Qnil; |
1128 If optional arg GENERICP is non-nil, also return t if OBJECT is | 1967 } |
1129 a valid generic character. */) | 1968 |
1130 (object, genericp) | 1969 DEFUN ("charset-priority-list", Fcharset_priority_list, |
1131 Lisp_Object object, genericp; | 1970 Scharset_priority_list, 0, 1, 0, |
1132 { | 1971 doc: /* Return the list of charsets ordered by priority. |
1133 if (! NATNUMP (object)) | 1972 HIGHESTP non-nil means just return the highest priority one. */) |
1134 return Qnil; | 1973 (highestp) |
1135 return (CHAR_VALID_P (XFASTINT (object), !NILP (genericp)) ? Qt : Qnil); | 1974 Lisp_Object highestp; |
1136 } | 1975 { |
1137 | 1976 Lisp_Object val = Qnil, list = Vcharset_ordered_list; |
1138 DEFUN ("unibyte-char-to-multibyte", Funibyte_char_to_multibyte, | 1977 |
1139 Sunibyte_char_to_multibyte, 1, 1, 0, | 1978 if (!NILP (highestp)) |
1140 doc: /* Convert the unibyte character CH to multibyte character. | 1979 return CHARSET_NAME (CHARSET_FROM_ID (XINT (Fcar (list)))); |
1141 The conversion is done based on `nonascii-translation-table' (which see) | 1980 |
1142 or `nonascii-insert-offset' (which see). */) | 1981 while (!NILP (list)) |
1143 (ch) | 1982 { |
1144 Lisp_Object ch; | 1983 val = Fcons (CHARSET_NAME (CHARSET_FROM_ID (XINT (XCAR (list)))), val); |
1145 { | 1984 list = XCDR (list); |
1146 int c; | 1985 } |
1147 | 1986 return Fnreverse (val); |
1148 CHECK_NUMBER (ch); | 1987 } |
1149 c = XINT (ch); | 1988 |
1150 if (c < 0 || c >= 0400) | 1989 DEFUN ("set-charset-priority", Fset_charset_priority, Sset_charset_priority, |
1151 error ("Invalid unibyte character: %d", c); | 1990 1, MANY, 0, |
1152 c = unibyte_char_to_multibyte (c); | 1991 doc: /* Assign higher priority to the charsets given as arguments. |
1153 if (c < 0) | 1992 usage: (set-charset-priority &rest charsets) */) |
1154 error ("Can't convert to multibyte character: %d", XINT (ch)); | 1993 (nargs, args) |
1155 return make_number (c); | 1994 int nargs; |
1156 } | 1995 Lisp_Object *args; |
1157 | 1996 { |
1158 DEFUN ("multibyte-char-to-unibyte", Fmultibyte_char_to_unibyte, | 1997 Lisp_Object new_head, old_list, arglist[2]; |
1159 Smultibyte_char_to_unibyte, 1, 1, 0, | 1998 int i, id; |
1160 doc: /* Convert the multibyte character CH to unibyte character. | 1999 |
1161 The conversion is done based on `nonascii-translation-table' (which see) | 2000 old_list = Fcopy_sequence (Vcharset_ordered_list); |
1162 or `nonascii-insert-offset' (which see). */) | 2001 new_head = Qnil; |
1163 (ch) | 2002 for (i = 0; i < nargs; i++) |
1164 Lisp_Object ch; | 2003 { |
1165 { | 2004 CHECK_CHARSET_GET_ID (args[i], id); |
1166 int c; | 2005 if (! NILP (Fmemq (make_number (id), old_list))) |
1167 | 2006 { |
1168 CHECK_NUMBER (ch); | 2007 old_list = Fdelq (make_number (id), old_list); |
1169 c = XINT (ch); | 2008 new_head = Fcons (make_number (id), new_head); |
1170 if (! CHAR_VALID_P (c, 0)) | 2009 } |
1171 error ("Invalid multibyte character: %d", c); | 2010 } |
1172 c = multibyte_char_to_unibyte (c, Qnil); | 2011 arglist[0] = Fnreverse (new_head); |
1173 if (c < 0) | 2012 arglist[1] = old_list; |
1174 error ("Can't convert to unibyte character: %d", XINT (ch)); | 2013 Vcharset_ordered_list = Fnconc (2, arglist); |
1175 return make_number (c); | 2014 charset_ordered_list_tick++; |
1176 } | 2015 |
1177 | 2016 for (old_list = Vcharset_ordered_list, new_head = Qnil; |
1178 DEFUN ("char-bytes", Fchar_bytes, Schar_bytes, 1, 1, 0, | 2017 CONSP (old_list); old_list = XCDR (old_list)) |
1179 doc: /* Return 1 regardless of the argument CHAR. */) | 2018 { |
1180 (ch) | 2019 if (! NILP (Fmemq (XCAR (old_list), Viso_2022_charset_list))) |
1181 Lisp_Object ch; | 2020 new_head = Fcons (XCAR (old_list), new_head); |
1182 { | 2021 } |
1183 CHECK_NUMBER (ch); | 2022 Viso_2022_charset_list = Fnreverse (new_head); |
1184 return make_number (1); | 2023 |
1185 } | 2024 return Qnil; |
1186 | 2025 } |
1187 /* Return how many bytes C will occupy in a multibyte buffer. | 2026 |
1188 Don't call this function directly, instead use macro CHAR_BYTES. */ | 2027 DEFUN ("charset-id-internal", Fcharset_id_internal, Scharset_id_internal, |
1189 int | 2028 0, 1, 0, |
1190 char_bytes (c) | 2029 doc: /* Internal use only. |
1191 int c; | 2030 Return charset identification number of CHARSET. */) |
1192 { | 2031 (charset) |
1193 int charset; | 2032 Lisp_Object charset; |
1194 | 2033 { |
1195 if (ASCII_BYTE_P (c) || (c & ~((1 << CHARACTERBITS) -1))) | 2034 int id; |
1196 return 1; | 2035 |
1197 if (SINGLE_BYTE_CHAR_P (c) && c >= 0xA0) | 2036 CHECK_CHARSET_GET_ID (charset, id); |
1198 return 1; | 2037 return make_number (id); |
1199 | 2038 } |
1200 charset = CHAR_CHARSET (c); | 2039 |
1201 return (CHARSET_DEFINED_P (charset) ? CHARSET_BYTES (charset) : 1); | 2040 |
1202 } | |
1203 | |
1204 /* Return the width of character of which multi-byte form starts with | |
1205 C. The width is measured by how many columns occupied on the | |
1206 screen when displayed in the current buffer. */ | |
1207 | |
1208 #define ONE_BYTE_CHAR_WIDTH(c) \ | |
1209 (c < 0x20 \ | |
1210 ? (c == '\t' \ | |
1211 ? XFASTINT (current_buffer->tab_width) \ | |
1212 : (c == '\n' ? 0 : (NILP (current_buffer->ctl_arrow) ? 4 : 2))) \ | |
1213 : (c < 0x7f \ | |
1214 ? 1 \ | |
1215 : (c == 0x7F \ | |
1216 ? (NILP (current_buffer->ctl_arrow) ? 4 : 2) \ | |
1217 : ((! NILP (current_buffer->enable_multibyte_characters) \ | |
1218 && BASE_LEADING_CODE_P (c)) \ | |
1219 ? WIDTH_BY_CHAR_HEAD (c) \ | |
1220 : 4)))) | |
1221 | |
1222 DEFUN ("char-width", Fchar_width, Schar_width, 1, 1, 0, | |
1223 doc: /* Return width of CHAR when displayed in the current buffer. | |
1224 The width is measured by how many columns it occupies on the screen. | |
1225 Tab is taken to occupy `tab-width' columns. */) | |
1226 (ch) | |
1227 Lisp_Object ch; | |
1228 { | |
1229 Lisp_Object val, disp; | |
1230 int c; | |
1231 struct Lisp_Char_Table *dp = buffer_display_table (); | |
1232 | |
1233 CHECK_NUMBER (ch); | |
1234 | |
1235 c = XINT (ch); | |
1236 | |
1237 /* Get the way the display table would display it. */ | |
1238 disp = dp ? DISP_CHAR_VECTOR (dp, c) : Qnil; | |
1239 | |
1240 if (VECTORP (disp)) | |
1241 XSETINT (val, XVECTOR (disp)->size); | |
1242 else if (SINGLE_BYTE_CHAR_P (c)) | |
1243 XSETINT (val, ONE_BYTE_CHAR_WIDTH (c)); | |
1244 else | |
1245 { | |
1246 int charset = CHAR_CHARSET (c); | |
1247 | |
1248 XSETFASTINT (val, CHARSET_WIDTH (charset)); | |
1249 } | |
1250 return val; | |
1251 } | |
1252 | |
1253 /* Return width of string STR of length LEN when displayed in the | |
1254 current buffer. The width is measured by how many columns it | |
1255 occupies on the screen. */ | |
1256 | |
1257 int | |
1258 strwidth (str, len) | |
1259 unsigned char *str; | |
1260 int len; | |
1261 { | |
1262 return c_string_width (str, len, -1, NULL, NULL); | |
1263 } | |
1264 | |
1265 /* Return width of string STR of length LEN when displayed in the | |
1266 current buffer. The width is measured by how many columns it | |
1267 occupies on the screen. If PRECISION > 0, return the width of | |
1268 longest substring that doesn't exceed PRECISION, and set number of | |
1269 characters and bytes of the substring in *NCHARS and *NBYTES | |
1270 respectively. */ | |
1271 | |
1272 int | |
1273 c_string_width (str, len, precision, nchars, nbytes) | |
1274 const unsigned char *str; | |
1275 int len, precision, *nchars, *nbytes; | |
1276 { | |
1277 int i = 0, i_byte = 0; | |
1278 int width = 0; | |
1279 int chars; | |
1280 struct Lisp_Char_Table *dp = buffer_display_table (); | |
1281 | |
1282 while (i_byte < len) | |
1283 { | |
1284 int bytes, thiswidth; | |
1285 Lisp_Object val; | |
1286 | |
1287 if (dp) | |
1288 { | |
1289 int c = STRING_CHAR_AND_LENGTH (str + i_byte, len - i_byte, bytes); | |
1290 | |
1291 chars = 1; | |
1292 val = DISP_CHAR_VECTOR (dp, c); | |
1293 if (VECTORP (val)) | |
1294 thiswidth = XVECTOR (val)->size; | |
1295 else | |
1296 thiswidth = ONE_BYTE_CHAR_WIDTH (str[i_byte]); | |
1297 } | |
1298 else | |
1299 { | |
1300 chars = 1; | |
1301 PARSE_MULTIBYTE_SEQ (str + i_byte, len - i_byte, bytes); | |
1302 thiswidth = ONE_BYTE_CHAR_WIDTH (str[i_byte]); | |
1303 } | |
1304 | |
1305 if (precision > 0 | |
1306 && (width + thiswidth > precision)) | |
1307 { | |
1308 *nchars = i; | |
1309 *nbytes = i_byte; | |
1310 return width; | |
1311 } | |
1312 i++; | |
1313 i_byte += bytes; | |
1314 width += thiswidth; | |
1315 } | |
1316 | |
1317 if (precision > 0) | |
1318 { | |
1319 *nchars = i; | |
1320 *nbytes = i_byte; | |
1321 } | |
1322 | |
1323 return width; | |
1324 } | |
1325 | |
1326 /* Return width of Lisp string STRING when displayed in the current | |
1327 buffer. The width is measured by how many columns it occupies on | |
1328 the screen while paying attention to compositions. If PRECISION > | |
1329 0, return the width of longest substring that doesn't exceed | |
1330 PRECISION, and set number of characters and bytes of the substring | |
1331 in *NCHARS and *NBYTES respectively. */ | |
1332 | |
1333 int | |
1334 lisp_string_width (string, precision, nchars, nbytes) | |
1335 Lisp_Object string; | |
1336 int precision, *nchars, *nbytes; | |
1337 { | |
1338 int len = SCHARS (string); | |
1339 int len_byte = SBYTES (string); | |
1340 const unsigned char *str = SDATA (string); | |
1341 int i = 0, i_byte = 0; | |
1342 int width = 0; | |
1343 struct Lisp_Char_Table *dp = buffer_display_table (); | |
1344 | |
1345 while (i < len) | |
1346 { | |
1347 int chars, bytes, thiswidth; | |
1348 Lisp_Object val; | |
1349 int cmp_id; | |
1350 int ignore, end; | |
1351 | |
1352 if (find_composition (i, -1, &ignore, &end, &val, string) | |
1353 && ((cmp_id = get_composition_id (i, i_byte, end - i, val, string)) | |
1354 >= 0)) | |
1355 { | |
1356 thiswidth = composition_table[cmp_id]->width; | |
1357 chars = end - i; | |
1358 bytes = string_char_to_byte (string, end) - i_byte; | |
1359 } | |
1360 else if (dp) | |
1361 { | |
1362 int c = STRING_CHAR_AND_LENGTH (str + i_byte, len - i_byte, bytes); | |
1363 | |
1364 chars = 1; | |
1365 val = DISP_CHAR_VECTOR (dp, c); | |
1366 if (VECTORP (val)) | |
1367 thiswidth = XVECTOR (val)->size; | |
1368 else | |
1369 thiswidth = ONE_BYTE_CHAR_WIDTH (str[i_byte]); | |
1370 } | |
1371 else | |
1372 { | |
1373 chars = 1; | |
1374 PARSE_MULTIBYTE_SEQ (str + i_byte, len_byte - i_byte, bytes); | |
1375 thiswidth = ONE_BYTE_CHAR_WIDTH (str[i_byte]); | |
1376 } | |
1377 | |
1378 if (precision > 0 | |
1379 && (width + thiswidth > precision)) | |
1380 { | |
1381 *nchars = i; | |
1382 *nbytes = i_byte; | |
1383 return width; | |
1384 } | |
1385 i += chars; | |
1386 i_byte += bytes; | |
1387 width += thiswidth; | |
1388 } | |
1389 | |
1390 if (precision > 0) | |
1391 { | |
1392 *nchars = i; | |
1393 *nbytes = i_byte; | |
1394 } | |
1395 | |
1396 return width; | |
1397 } | |
1398 | |
1399 DEFUN ("string-width", Fstring_width, Sstring_width, 1, 1, 0, | |
1400 doc: /* Return width of STRING when displayed in the current buffer. | |
1401 Width is measured by how many columns it occupies on the screen. | |
1402 When calculating width of a multibyte character in STRING, | |
1403 only the base leading-code is considered; the validity of | |
1404 the following bytes is not checked. Tabs in STRING are always | |
1405 taken to occupy `tab-width' columns. */) | |
1406 (str) | |
1407 Lisp_Object str; | |
1408 { | |
1409 Lisp_Object val; | |
1410 | |
1411 CHECK_STRING (str); | |
1412 XSETFASTINT (val, lisp_string_width (str, -1, NULL, NULL)); | |
1413 return val; | |
1414 } | |
1415 | |
1416 DEFUN ("char-direction", Fchar_direction, Schar_direction, 1, 1, 0, | |
1417 doc: /* Return the direction of CHAR. | |
1418 The returned value is 0 for left-to-right and 1 for right-to-left. */) | |
1419 (ch) | |
1420 Lisp_Object ch; | |
1421 { | |
1422 int charset; | |
1423 | |
1424 CHECK_NUMBER (ch); | |
1425 charset = CHAR_CHARSET (XFASTINT (ch)); | |
1426 if (!CHARSET_DEFINED_P (charset)) | |
1427 invalid_character (XINT (ch)); | |
1428 return CHARSET_TABLE_INFO (charset, CHARSET_DIRECTION_IDX); | |
1429 } | |
1430 | |
1431 DEFUN ("chars-in-region", Fchars_in_region, Schars_in_region, 2, 2, 0, | |
1432 doc: /* Return number of characters between BEG and END. */) | |
1433 (beg, end) | |
1434 Lisp_Object beg, end; | |
1435 { | |
1436 int from, to; | |
1437 | |
1438 CHECK_NUMBER_COERCE_MARKER (beg); | |
1439 CHECK_NUMBER_COERCE_MARKER (end); | |
1440 | |
1441 from = min (XFASTINT (beg), XFASTINT (end)); | |
1442 to = max (XFASTINT (beg), XFASTINT (end)); | |
1443 | |
1444 return make_number (to - from); | |
1445 } | |
1446 | |
1447 /* Return the number of characters in the NBYTES bytes at PTR. | |
1448 This works by looking at the contents and checking for multibyte sequences. | |
1449 However, if the current buffer has enable-multibyte-characters = nil, | |
1450 we treat each byte as a character. */ | |
1451 | |
1452 int | |
1453 chars_in_text (ptr, nbytes) | |
1454 const unsigned char *ptr; | |
1455 int nbytes; | |
1456 { | |
1457 /* current_buffer is null at early stages of Emacs initialization. */ | |
1458 if (current_buffer == 0 | |
1459 || NILP (current_buffer->enable_multibyte_characters)) | |
1460 return nbytes; | |
1461 | |
1462 return multibyte_chars_in_text (ptr, nbytes); | |
1463 } | |
1464 | |
1465 /* Return the number of characters in the NBYTES bytes at PTR. | |
1466 This works by looking at the contents and checking for multibyte sequences. | |
1467 It ignores enable-multibyte-characters. */ | |
1468 | |
1469 int | |
1470 multibyte_chars_in_text (ptr, nbytes) | |
1471 const unsigned char *ptr; | |
1472 int nbytes; | |
1473 { | |
1474 const unsigned char *endp; | |
1475 int chars, bytes; | |
1476 | |
1477 endp = ptr + nbytes; | |
1478 chars = 0; | |
1479 | |
1480 while (ptr < endp) | |
1481 { | |
1482 PARSE_MULTIBYTE_SEQ (ptr, endp - ptr, bytes); | |
1483 ptr += bytes; | |
1484 chars++; | |
1485 } | |
1486 | |
1487 return chars; | |
1488 } | |
1489 | |
1490 /* Parse unibyte text at STR of LEN bytes as multibyte text, and | |
1491 count the numbers of characters and bytes in it. On counting | |
1492 bytes, pay attention to the fact that 8-bit characters in the range | |
1493 0x80..0x9F are represented by 2 bytes in multibyte text. */ | |
1494 void | 2041 void |
1495 parse_str_as_multibyte (str, len, nchars, nbytes) | 2042 init_charset () |
1496 const unsigned char *str; | 2043 { |
1497 int len, *nchars, *nbytes; | 2044 |
1498 { | 2045 } |
1499 const unsigned char *endp = str + len; | 2046 |
1500 int n, chars = 0, bytes = 0; | |
1501 | |
1502 while (str < endp) | |
1503 { | |
1504 if (UNIBYTE_STR_AS_MULTIBYTE_P (str, endp - str, n)) | |
1505 str += n, bytes += n; | |
1506 else | |
1507 str++, bytes += 2; | |
1508 chars++; | |
1509 } | |
1510 *nchars = chars; | |
1511 *nbytes = bytes; | |
1512 return; | |
1513 } | |
1514 | |
1515 /* Arrange unibyte text at STR of NBYTES bytes as multibyte text. | |
1516 It actually converts only 8-bit characters in the range 0x80..0x9F | |
1517 that don't contruct multibyte characters to multibyte forms. If | |
1518 NCHARS is nonzero, set *NCHARS to the number of characters in the | |
1519 text. It is assured that we can use LEN bytes at STR as a work | |
1520 area and that is enough. Return the number of bytes of the | |
1521 resulting text. */ | |
1522 | |
1523 int | |
1524 str_as_multibyte (str, len, nbytes, nchars) | |
1525 unsigned char *str; | |
1526 int len, nbytes, *nchars; | |
1527 { | |
1528 unsigned char *p = str, *endp = str + nbytes; | |
1529 unsigned char *to; | |
1530 int chars = 0; | |
1531 int n; | |
1532 | |
1533 while (p < endp && UNIBYTE_STR_AS_MULTIBYTE_P (p, endp - p, n)) | |
1534 p += n, chars++; | |
1535 if (nchars) | |
1536 *nchars = chars; | |
1537 if (p == endp) | |
1538 return nbytes; | |
1539 | |
1540 to = p; | |
1541 nbytes = endp - p; | |
1542 endp = str + len; | |
1543 safe_bcopy (p, endp - nbytes, nbytes); | |
1544 p = endp - nbytes; | |
1545 while (p < endp) | |
1546 { | |
1547 if (UNIBYTE_STR_AS_MULTIBYTE_P (p, endp - p, n)) | |
1548 { | |
1549 while (n--) | |
1550 *to++ = *p++; | |
1551 } | |
1552 else | |
1553 { | |
1554 *to++ = LEADING_CODE_8_BIT_CONTROL; | |
1555 *to++ = *p++ + 0x20; | |
1556 } | |
1557 chars++; | |
1558 } | |
1559 if (nchars) | |
1560 *nchars = chars; | |
1561 return (to - str); | |
1562 } | |
1563 | |
1564 /* Parse unibyte string at STR of LEN bytes, and return the number of | |
1565 bytes it may ocupy when converted to multibyte string by | |
1566 `str_to_multibyte'. */ | |
1567 | |
1568 int | |
1569 parse_str_to_multibyte (str, len) | |
1570 unsigned char *str; | |
1571 int len; | |
1572 { | |
1573 unsigned char *endp = str + len; | |
1574 int bytes; | |
1575 | |
1576 for (bytes = 0; str < endp; str++) | |
1577 bytes += (*str < 0x80 || *str >= 0xA0) ? 1 : 2; | |
1578 return bytes; | |
1579 } | |
1580 | |
1581 /* Convert unibyte text at STR of NBYTES bytes to multibyte text | |
1582 that contains the same single-byte characters. It actually | |
1583 converts all 8-bit characters to multibyte forms. It is assured | |
1584 that we can use LEN bytes at STR as a work area and that is | |
1585 enough. */ | |
1586 | |
1587 int | |
1588 str_to_multibyte (str, len, bytes) | |
1589 unsigned char *str; | |
1590 int len, bytes; | |
1591 { | |
1592 unsigned char *p = str, *endp = str + bytes; | |
1593 unsigned char *to; | |
1594 | |
1595 while (p < endp && (*p < 0x80 || *p >= 0xA0)) p++; | |
1596 if (p == endp) | |
1597 return bytes; | |
1598 to = p; | |
1599 bytes = endp - p; | |
1600 endp = str + len; | |
1601 safe_bcopy (p, endp - bytes, bytes); | |
1602 p = endp - bytes; | |
1603 while (p < endp) | |
1604 { | |
1605 if (*p < 0x80 || *p >= 0xA0) | |
1606 *to++ = *p++; | |
1607 else | |
1608 *to++ = LEADING_CODE_8_BIT_CONTROL, *to++ = *p++ + 0x20; | |
1609 } | |
1610 return (to - str); | |
1611 } | |
1612 | |
1613 /* Arrange multibyte text at STR of LEN bytes as a unibyte text. It | |
1614 actually converts only 8-bit characters in the range 0x80..0x9F to | |
1615 unibyte forms. */ | |
1616 | |
1617 int | |
1618 str_as_unibyte (str, bytes) | |
1619 unsigned char *str; | |
1620 int bytes; | |
1621 { | |
1622 unsigned char *p = str, *endp = str + bytes; | |
1623 unsigned char *to = str; | |
1624 | |
1625 while (p < endp && *p != LEADING_CODE_8_BIT_CONTROL) p++; | |
1626 to = p; | |
1627 while (p < endp) | |
1628 { | |
1629 if (*p == LEADING_CODE_8_BIT_CONTROL) | |
1630 *to++ = *(p + 1) - 0x20, p += 2; | |
1631 else | |
1632 *to++ = *p++; | |
1633 } | |
1634 return (to - str); | |
1635 } | |
1636 | |
1637 | |
1638 DEFUN ("string", Fstring, Sstring, 0, MANY, 0, | |
1639 doc: /* Concatenate all the argument characters and make the result a string. | |
1640 usage: (string &rest CHARACTERS) */) | |
1641 (n, args) | |
1642 int n; | |
1643 Lisp_Object *args; | |
1644 { | |
1645 int i; | |
1646 unsigned char *buf = (unsigned char *) alloca (MAX_MULTIBYTE_LENGTH * n); | |
1647 unsigned char *p = buf; | |
1648 int c; | |
1649 int multibyte = 0; | |
1650 | |
1651 for (i = 0; i < n; i++) | |
1652 { | |
1653 CHECK_NUMBER (args[i]); | |
1654 if (!multibyte && !SINGLE_BYTE_CHAR_P (XFASTINT (args[i]))) | |
1655 multibyte = 1; | |
1656 } | |
1657 | |
1658 for (i = 0; i < n; i++) | |
1659 { | |
1660 c = XINT (args[i]); | |
1661 if (multibyte) | |
1662 p += CHAR_STRING (c, p); | |
1663 else | |
1664 *p++ = c; | |
1665 } | |
1666 | |
1667 return make_string_from_bytes (buf, n, p - buf); | |
1668 } | |
1669 | |
1670 #endif /* emacs */ | |
1671 | |
1672 int | |
1673 charset_id_internal (charset_name) | |
1674 char *charset_name; | |
1675 { | |
1676 Lisp_Object val; | |
1677 | |
1678 val= Fget (intern (charset_name), Qcharset); | |
1679 if (!VECTORP (val)) | |
1680 error ("Charset %s is not defined", charset_name); | |
1681 | |
1682 return (XINT (XVECTOR (val)->contents[0])); | |
1683 } | |
1684 | |
1685 DEFUN ("setup-special-charsets", Fsetup_special_charsets, | |
1686 Ssetup_special_charsets, 0, 0, 0, doc: /* Internal use only. */) | |
1687 () | |
1688 { | |
1689 charset_latin_iso8859_1 = charset_id_internal ("latin-iso8859-1"); | |
1690 charset_jisx0208_1978 = charset_id_internal ("japanese-jisx0208-1978"); | |
1691 charset_jisx0208 = charset_id_internal ("japanese-jisx0208"); | |
1692 charset_katakana_jisx0201 = charset_id_internal ("katakana-jisx0201"); | |
1693 charset_latin_jisx0201 = charset_id_internal ("latin-jisx0201"); | |
1694 charset_big5_1 = charset_id_internal ("chinese-big5-1"); | |
1695 charset_big5_2 = charset_id_internal ("chinese-big5-2"); | |
1696 return Qnil; | |
1697 } | |
1698 | 2047 |
1699 void | 2048 void |
1700 init_charset_once () | 2049 init_charset_once () |
1701 { | 2050 { |
1702 int i, j, k; | 2051 int i, j, k; |
1703 | 2052 |
1704 staticpro (&Vcharset_table); | 2053 for (i = 0; i < ISO_MAX_DIMENSION; i++) |
1705 staticpro (&Vcharset_symbol_table); | 2054 for (j = 0; j < ISO_MAX_CHARS; j++) |
1706 staticpro (&Vgeneric_character_list); | 2055 for (k = 0; k < ISO_MAX_FINAL; k++) |
1707 | 2056 iso_charset_table[i][j][k] = -1; |
1708 /* This has to be done here, before we call Fmake_char_table. */ | |
1709 Qcharset_table = intern ("charset-table"); | |
1710 staticpro (&Qcharset_table); | |
1711 | |
1712 /* Intern this now in case it isn't already done. | |
1713 Setting this variable twice is harmless. | |
1714 But don't staticpro it here--that is done in alloc.c. */ | |
1715 Qchar_table_extra_slots = intern ("char-table-extra-slots"); | |
1716 | |
1717 /* Now we are ready to set up this property, so we can | |
1718 create the charset table. */ | |
1719 Fput (Qcharset_table, Qchar_table_extra_slots, make_number (0)); | |
1720 Vcharset_table = Fmake_char_table (Qcharset_table, Qnil); | |
1721 | |
1722 Qunknown = intern ("unknown"); | |
1723 staticpro (&Qunknown); | |
1724 Vcharset_symbol_table = Fmake_vector (make_number (MAX_CHARSET + 1), | |
1725 Qunknown); | |
1726 | |
1727 /* Setup tables. */ | |
1728 for (i = 0; i < 2; i++) | |
1729 for (j = 0; j < 2; j++) | |
1730 for (k = 0; k < 128; k++) | |
1731 iso_charset_table [i][j][k] = -1; | |
1732 | 2057 |
1733 for (i = 0; i < 256; i++) | 2058 for (i = 0; i < 256; i++) |
1734 bytes_by_char_head[i] = 1; | 2059 emacs_mule_charset[i] = NULL; |
1735 bytes_by_char_head[LEADING_CODE_PRIVATE_11] = 3; | 2060 |
1736 bytes_by_char_head[LEADING_CODE_PRIVATE_12] = 3; | 2061 charset_jisx0201_roman = -1; |
1737 bytes_by_char_head[LEADING_CODE_PRIVATE_21] = 4; | 2062 charset_jisx0208_1978 = -1; |
1738 bytes_by_char_head[LEADING_CODE_PRIVATE_22] = 4; | 2063 charset_jisx0208 = -1; |
1739 | 2064 |
1740 for (i = 0; i < 128; i++) | 2065 for (i = 0; i < 128; i++) |
1741 width_by_char_head[i] = 1; | 2066 unibyte_to_multibyte_table[i] = i; |
1742 for (; i < 256; i++) | 2067 for (; i < 256; i++) |
1743 width_by_char_head[i] = 4; | 2068 unibyte_to_multibyte_table[i] = BYTE8_TO_CHAR (i); |
1744 width_by_char_head[LEADING_CODE_PRIVATE_11] = 1; | |
1745 width_by_char_head[LEADING_CODE_PRIVATE_12] = 2; | |
1746 width_by_char_head[LEADING_CODE_PRIVATE_21] = 1; | |
1747 width_by_char_head[LEADING_CODE_PRIVATE_22] = 2; | |
1748 | |
1749 { | |
1750 Lisp_Object val; | |
1751 | |
1752 val = Qnil; | |
1753 for (i = 0x81; i < 0x90; i++) | |
1754 val = Fcons (make_number ((i - 0x70) << 7), val); | |
1755 for (; i < 0x9A; i++) | |
1756 val = Fcons (make_number ((i - 0x8F) << 14), val); | |
1757 for (i = 0xA0; i < 0xF0; i++) | |
1758 val = Fcons (make_number ((i - 0x70) << 7), val); | |
1759 for (; i < 0xFF; i++) | |
1760 val = Fcons (make_number ((i - 0xE0) << 14), val); | |
1761 Vgeneric_character_list = Fnreverse (val); | |
1762 } | |
1763 | |
1764 nonascii_insert_offset = 0; | |
1765 Vnonascii_translation_table = Qnil; | |
1766 } | 2069 } |
1767 | 2070 |
1768 #ifdef emacs | 2071 #ifdef emacs |
1769 | 2072 |
1770 void | 2073 void |
1771 syms_of_charset () | 2074 syms_of_charset () |
1772 { | 2075 { |
1773 Qcharset = intern ("charset"); | 2076 char *p; |
1774 staticpro (&Qcharset); | 2077 |
1775 | 2078 DEFSYM (Qcharsetp, "charsetp"); |
1776 Qascii = intern ("ascii"); | 2079 |
1777 staticpro (&Qascii); | 2080 DEFSYM (Qascii, "ascii"); |
1778 | 2081 DEFSYM (Qunicode, "unicode"); |
1779 Qeight_bit_control = intern ("eight-bit-control"); | 2082 DEFSYM (Qeight_bit, "eight-bit"); |
1780 staticpro (&Qeight_bit_control); | 2083 DEFSYM (Qiso_8859_1, "iso-8859-1"); |
1781 | 2084 |
1782 Qeight_bit_graphic = intern ("eight-bit-graphic"); | 2085 DEFSYM (Qgl, "gl"); |
1783 staticpro (&Qeight_bit_graphic); | 2086 DEFSYM (Qgr, "gr"); |
1784 | 2087 |
1785 /* Define special charsets ascii, eight-bit-control, and | 2088 p = (char *) xmalloc (30000); |
1786 eight-bit-graphic. */ | 2089 |
1787 update_charset_table (make_number (CHARSET_ASCII), | 2090 staticpro (&Vcharset_ordered_list); |
1788 make_number (1), make_number (94), | 2091 Vcharset_ordered_list = Qnil; |
1789 make_number (1), | 2092 |
1790 make_number (0), | 2093 staticpro (&Viso_2022_charset_list); |
1791 make_number ('B'), | 2094 Viso_2022_charset_list = Qnil; |
1792 make_number (0), | 2095 |
1793 build_string ("ASCII"), | 2096 staticpro (&Vemacs_mule_charset_list); |
1794 Qnil, /* same as above */ | 2097 Vemacs_mule_charset_list = Qnil; |
1795 build_string ("ASCII (ISO646 IRV)")); | 2098 |
1796 CHARSET_SYMBOL (CHARSET_ASCII) = Qascii; | 2099 staticpro (&Vcharset_hash_table); |
1797 Fput (Qascii, Qcharset, CHARSET_TABLE_ENTRY (CHARSET_ASCII)); | 2100 { |
1798 | 2101 Lisp_Object args[2]; |
1799 update_charset_table (make_number (CHARSET_8_BIT_CONTROL), | 2102 args[0] = QCtest; |
1800 make_number (1), make_number (96), | 2103 args[1] = Qeq; |
1801 make_number (4), | 2104 Vcharset_hash_table = Fmake_hash_table (2, args); |
1802 make_number (0), | 2105 } |
1803 make_number (-1), | 2106 |
1804 make_number (-1), | 2107 charset_table_size = 128; |
1805 build_string ("8-bit control code (0x80..0x9F)"), | 2108 charset_table = ((struct charset *) |
1806 Qnil, /* same as above */ | 2109 xmalloc (sizeof (struct charset) * charset_table_size)); |
1807 Qnil); /* same as above */ | 2110 charset_table_used = 0; |
1808 CHARSET_SYMBOL (CHARSET_8_BIT_CONTROL) = Qeight_bit_control; | 2111 |
1809 Fput (Qeight_bit_control, Qcharset, | 2112 staticpro (&Vchar_unified_charset_table); |
1810 CHARSET_TABLE_ENTRY (CHARSET_8_BIT_CONTROL)); | 2113 Vchar_unified_charset_table = Fmake_char_table (Qnil, make_number (-1)); |
1811 | 2114 |
1812 update_charset_table (make_number (CHARSET_8_BIT_GRAPHIC), | 2115 defsubr (&Scharsetp); |
1813 make_number (1), make_number (96), | 2116 defsubr (&Smap_charset_chars); |
1814 make_number (4), | 2117 defsubr (&Sdefine_charset_internal); |
1815 make_number (0), | 2118 defsubr (&Sdefine_charset_alias); |
1816 make_number (-1), | 2119 defsubr (&Sunibyte_charset); |
1817 make_number (-1), | 2120 defsubr (&Sset_unibyte_charset); |
1818 build_string ("8-bit graphic char (0xA0..0xFF)"), | 2121 defsubr (&Scharset_plist); |
1819 Qnil, /* same as above */ | 2122 defsubr (&Sset_charset_plist); |
1820 Qnil); /* same as above */ | 2123 defsubr (&Sunify_charset); |
1821 CHARSET_SYMBOL (CHARSET_8_BIT_GRAPHIC) = Qeight_bit_graphic; | |
1822 Fput (Qeight_bit_graphic, Qcharset, | |
1823 CHARSET_TABLE_ENTRY (CHARSET_8_BIT_GRAPHIC)); | |
1824 | |
1825 Qauto_fill_chars = intern ("auto-fill-chars"); | |
1826 staticpro (&Qauto_fill_chars); | |
1827 Fput (Qauto_fill_chars, Qchar_table_extra_slots, make_number (0)); | |
1828 | |
1829 defsubr (&Sdefine_charset); | |
1830 defsubr (&Sgeneric_character_list); | |
1831 defsubr (&Sget_unused_iso_final_char); | 2124 defsubr (&Sget_unused_iso_final_char); |
1832 defsubr (&Sdeclare_equiv_charset); | 2125 defsubr (&Sdeclare_equiv_charset); |
1833 defsubr (&Sfind_charset_region); | 2126 defsubr (&Sfind_charset_region); |
1834 defsubr (&Sfind_charset_string); | 2127 defsubr (&Sfind_charset_string); |
1835 defsubr (&Smake_char_internal); | 2128 defsubr (&Sdecode_char); |
2129 defsubr (&Sencode_char); | |
1836 defsubr (&Ssplit_char); | 2130 defsubr (&Ssplit_char); |
2131 defsubr (&Smake_char); | |
1837 defsubr (&Schar_charset); | 2132 defsubr (&Schar_charset); |
1838 defsubr (&Scharset_after); | 2133 defsubr (&Scharset_after); |
1839 defsubr (&Siso_charset); | 2134 defsubr (&Siso_charset); |
1840 defsubr (&Schar_valid_p); | 2135 defsubr (&Sclear_charset_maps); |
1841 defsubr (&Sunibyte_char_to_multibyte); | 2136 defsubr (&Scharset_priority_list); |
1842 defsubr (&Smultibyte_char_to_unibyte); | 2137 defsubr (&Sset_charset_priority); |
1843 defsubr (&Schar_bytes); | 2138 defsubr (&Scharset_id_internal); |
1844 defsubr (&Schar_width); | 2139 |
1845 defsubr (&Sstring_width); | 2140 DEFVAR_LISP ("charset-map-directory", &Vcharset_map_directory, |
1846 defsubr (&Schar_direction); | 2141 doc: /* Directory of charset map files that come with GNU Emacs. |
1847 defsubr (&Schars_in_region); | 2142 The default value is sub-directory "charsets" of `data-directory'. */); |
1848 defsubr (&Sstring); | 2143 Vcharset_map_directory = Fexpand_file_name (build_string ("charsets"), |
1849 defsubr (&Ssetup_special_charsets); | 2144 Vdata_directory); |
1850 | 2145 |
1851 DEFVAR_LISP ("charset-list", &Vcharset_list, | 2146 DEFVAR_LISP ("charset-list", &Vcharset_list, |
1852 doc: /* List of charsets ever defined. */); | 2147 doc: /* List of all charsets ever defined. */); |
1853 Vcharset_list = Fcons (Qascii, Fcons (Qeight_bit_control, | 2148 Vcharset_list = Qnil; |
1854 Fcons (Qeight_bit_graphic, Qnil))); | 2149 |
1855 | 2150 charset_ascii |
1856 DEFVAR_LISP ("translation-table-vector", &Vtranslation_table_vector, | 2151 = define_charset_internal (Qascii, 1, "\x00\x7F\x00\x00\x00\x00", |
1857 doc: /* Vector of cons cell of a symbol and translation table ever defined. | 2152 0, 127, 'B', -1, 0, 1, 0, 0); |
1858 An ID of a translation table is an index of this vector. */); | 2153 charset_iso_8859_1 |
1859 Vtranslation_table_vector = Fmake_vector (make_number (16), Qnil); | 2154 = define_charset_internal (Qiso_8859_1, 1, "\x00\xFF\x00\x00\x00\x00", |
1860 | 2155 0, 255, -1, -1, -1, 1, 0, 0); |
1861 DEFVAR_INT ("leading-code-private-11", &leading_code_private_11, | 2156 charset_unicode |
1862 doc: /* Leading-code of private TYPE9N charset of column-width 1. */); | 2157 = define_charset_internal (Qunicode, 3, "\x00\xFF\x00\xFF\x00\x10", |
1863 leading_code_private_11 = LEADING_CODE_PRIVATE_11; | 2158 0, MAX_UNICODE_CHAR, -1, 0, -1, 1, 0, 0); |
1864 | 2159 charset_eight_bit |
1865 DEFVAR_INT ("leading-code-private-12", &leading_code_private_12, | 2160 = define_charset_internal (Qeight_bit, 1, "\x80\xFF\x00\x00\x00\x00", |
1866 doc: /* Leading-code of private TYPE9N charset of column-width 2. */); | 2161 128, 255, -1, 0, -1, 0, 0, |
1867 leading_code_private_12 = LEADING_CODE_PRIVATE_12; | 2162 MAX_5_BYTE_CHAR + 1); |
1868 | |
1869 DEFVAR_INT ("leading-code-private-21", &leading_code_private_21, | |
1870 doc: /* Leading-code of private TYPE9Nx9N charset of column-width 1. */); | |
1871 leading_code_private_21 = LEADING_CODE_PRIVATE_21; | |
1872 | |
1873 DEFVAR_INT ("leading-code-private-22", &leading_code_private_22, | |
1874 doc: /* Leading-code of private TYPE9Nx9N charset of column-width 2. */); | |
1875 leading_code_private_22 = LEADING_CODE_PRIVATE_22; | |
1876 | |
1877 DEFVAR_INT ("nonascii-insert-offset", &nonascii_insert_offset, | |
1878 doc: /* Offset for converting non-ASCII unibyte codes 0240...0377 to multibyte. | |
1879 This is used for converting unibyte text to multibyte, | |
1880 and for inserting character codes specified by number. | |
1881 | |
1882 This serves to convert a Latin-1 or similar 8-bit character code | |
1883 to the corresponding Emacs multibyte character code. | |
1884 Typically the value should be (- (make-char CHARSET 0) 128), | |
1885 for your choice of character set. | |
1886 If `nonascii-translation-table' is non-nil, it overrides this variable. */); | |
1887 nonascii_insert_offset = 0; | |
1888 | |
1889 DEFVAR_LISP ("nonascii-translation-table", &Vnonascii_translation_table, | |
1890 doc: /* Translation table to convert non-ASCII unibyte codes to multibyte. | |
1891 This is used for converting unibyte text to multibyte, | |
1892 and for inserting character codes specified by number. | |
1893 | |
1894 Conversion is performed only when multibyte characters are enabled, | |
1895 and it serves to convert a Latin-1 or similar 8-bit character code | |
1896 to the corresponding Emacs character code. | |
1897 | |
1898 If this is nil, `nonascii-insert-offset' is used instead. | |
1899 See also the docstring of `make-translation-table'. */); | |
1900 Vnonascii_translation_table = Qnil; | |
1901 | |
1902 DEFVAR_LISP ("auto-fill-chars", &Vauto_fill_chars, | |
1903 doc: /* A char-table for characters which invoke auto-filling. | |
1904 Such characters have value t in this table. */); | |
1905 Vauto_fill_chars = Fmake_char_table (Qauto_fill_chars, Qnil); | |
1906 CHAR_TABLE_SET (Vauto_fill_chars, make_number (' '), Qt); | |
1907 CHAR_TABLE_SET (Vauto_fill_chars, make_number ('\n'), Qt); | |
1908 } | 2163 } |
1909 | 2164 |
1910 #endif /* emacs */ | 2165 #endif /* emacs */ |