Commit | Line | Data |
---|---|---|
3263d5a2 | 1 | /* Basic character set support. |
ab422c4d | 2 | Copyright (C) 2001-2013 Free Software Foundation, Inc. |
7976eda0 | 3 | Copyright (C) 1995, 1996, 1997, 1998, 1999, 2000, 2001, 2002, 2003, 2004, |
5df4f04c | 4 | 2005, 2006, 2007, 2008, 2009, 2010, 2011 |
ce03bf76 KH |
5 | National Institute of Advanced Industrial Science and Technology (AIST) |
6 | Registration Number H14PRO021 | |
4ed46869 | 7 | |
327719ee | 8 | Copyright (C) 2003, 2004 |
3263d5a2 KH |
9 | National Institute of Advanced Industrial Science and Technology (AIST) |
10 | Registration Number H13PRO009 | |
4ed46869 | 11 | |
369314dc KH |
12 | This file is part of GNU Emacs. |
13 | ||
9ec0b715 | 14 | GNU Emacs is free software: you can redistribute it and/or modify |
369314dc | 15 | it under the terms of the GNU General Public License as published by |
9ec0b715 GM |
16 | the Free Software Foundation, either version 3 of the License, or |
17 | (at your option) any later version. | |
4ed46869 | 18 | |
369314dc KH |
19 | GNU Emacs is distributed in the hope that it will be useful, |
20 | but WITHOUT ANY WARRANTY; without even the implied warranty of | |
21 | MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the | |
22 | GNU General Public License for more details. | |
4ed46869 | 23 | |
369314dc | 24 | You should have received a copy of the GNU General Public License |
9ec0b715 | 25 | along with GNU Emacs. If not, see <http://www.gnu.org/licenses/>. */ |
4ed46869 | 26 | |
68c45bf0 | 27 | #include <config.h> |
68c45bf0 | 28 | |
f162bcc3 PE |
29 | #define CHARSET_INLINE EXTERN_INLINE |
30 | ||
4ed46869 | 31 | #include <stdio.h> |
3263d5a2 | 32 | #include <unistd.h> |
726929c4 | 33 | #include <limits.h> |
4ed46869 | 34 | #include <sys/types.h> |
620f13b0 | 35 | #include <c-ctype.h> |
4ed46869 | 36 | #include "lisp.h" |
3263d5a2 | 37 | #include "character.h" |
4ed46869 KH |
38 | #include "charset.h" |
39 | #include "coding.h" | |
fc6b09bf | 40 | #include "disptab.h" |
3263d5a2 | 41 | #include "buffer.h" |
4ed46869 | 42 | |
04c2f2c5 | 43 | /*** GENERAL NOTES on CODED CHARACTER SETS (CHARSETS) *** |
4ed46869 | 44 | |
3263d5a2 | 45 | A coded character set ("charset" hereafter) is a meaningful |
04c2f2c5 | 46 | collection (i.e. language, culture, functionality, etc.) of |
3263d5a2 | 47 | characters. Emacs handles multiple charsets at once. In Emacs Lisp |
04c2f2c5 DL |
48 | code, a charset is represented by a symbol. In C code, a charset is |
49 | represented by its ID number or by a pointer to a struct charset. | |
4ed46869 | 50 | |
3263d5a2 KH |
51 | The actual information about each charset is stored in two places. |
52 | Lispy information is stored in the hash table Vcharset_hash_table as | |
53 | a vector (charset attributes). The other information is stored in | |
04c2f2c5 | 54 | charset_table as a struct charset. |
4ed46869 | 55 | |
3263d5a2 | 56 | */ |
4ed46869 | 57 | |
3263d5a2 KH |
58 | /* Hash table that contains attributes of each charset. Keys are |
59 | charset symbols, and values are vectors of charset attributes. */ | |
60 | Lisp_Object Vcharset_hash_table; | |
4ed46869 | 61 | |
3263d5a2 KH |
62 | /* Table of struct charset. */ |
63 | struct charset *charset_table; | |
4ed46869 | 64 | |
0065d054 | 65 | static ptrdiff_t charset_table_size; |
5af5dd92 | 66 | static int charset_table_used; |
4ed46869 | 67 | |
3263d5a2 | 68 | Lisp_Object Qcharsetp; |
4ed46869 | 69 | |
3263d5a2 KH |
70 | /* Special charset symbols. */ |
71 | Lisp_Object Qascii; | |
955cbe7b PE |
72 | static Lisp_Object Qeight_bit; |
73 | static Lisp_Object Qiso_8859_1; | |
74 | static Lisp_Object Qunicode; | |
75 | static Lisp_Object Qemacs; | |
4ed46869 | 76 | |
3263d5a2 KH |
77 | /* The corresponding charsets. */ |
78 | int charset_ascii; | |
2fe1edd1 | 79 | int charset_eight_bit; |
a2cb4e63 | 80 | static int charset_iso_8859_1; |
3263d5a2 | 81 | int charset_unicode; |
a2cb4e63 | 82 | static int charset_emacs; |
b0e3cf2b | 83 | |
7c7dceee KH |
84 | /* The other special charsets. */ |
85 | int charset_jisx0201_roman; | |
86 | int charset_jisx0208_1978; | |
87 | int charset_jisx0208; | |
d32320c4 | 88 | int charset_ksc5601; |
c1a08b4c | 89 | |
3263d5a2 | 90 | /* Value of charset attribute `charset-iso-plane'. */ |
955cbe7b | 91 | static Lisp_Object Qgl, Qgr; |
c1a08b4c | 92 | |
d1a04588 KH |
93 | /* Charset of unibyte characters. */ |
94 | int charset_unibyte; | |
4ed46869 | 95 | |
3263d5a2 KH |
96 | /* List of charsets ordered by the priority. */ |
97 | Lisp_Object Vcharset_ordered_list; | |
4ed46869 | 98 | |
6a9c90ec KH |
99 | /* Sub-list of Vcharset_ordered_list that contains all non-preferred |
100 | charsets. */ | |
101 | Lisp_Object Vcharset_non_preferred_head; | |
102 | ||
dbbb237d | 103 | /* Incremented everytime we change Vcharset_ordered_list. This is |
64165ae2 | 104 | unsigned short so that it fits in Lisp_Int and never matches |
dbbb237d KH |
105 | -1. */ |
106 | unsigned short charset_ordered_list_tick; | |
4ed46869 | 107 | |
3263d5a2 KH |
108 | /* List of iso-2022 charsets. */ |
109 | Lisp_Object Viso_2022_charset_list; | |
35e623fb | 110 | |
3263d5a2 KH |
111 | /* List of emacs-mule charsets. */ |
112 | Lisp_Object Vemacs_mule_charset_list; | |
113 | ||
b84ae584 | 114 | int emacs_mule_charset[256]; |
4ed46869 KH |
115 | |
116 | /* Mapping table from ISO2022's charset (specified by DIMENSION, | |
117 | CHARS, and FINAL-CHAR) to Emacs' charset. */ | |
3263d5a2 KH |
118 | int iso_charset_table[ISO_MAX_DIMENSION][ISO_MAX_CHARS][ISO_MAX_FINAL]; |
119 | ||
69f8de5b KH |
120 | #define CODE_POINT_TO_INDEX(charset, code) \ |
121 | ((charset)->code_linear_p \ | |
3c7649c1 | 122 | ? (int) ((code) - (charset)->min_code) \ |
69f8de5b KH |
123 | : (((charset)->code_space_mask[(code) >> 24] & 0x8) \ |
124 | && ((charset)->code_space_mask[((code) >> 16) & 0xFF] & 0x4) \ | |
125 | && ((charset)->code_space_mask[((code) >> 8) & 0xFF] & 0x2) \ | |
126 | && ((charset)->code_space_mask[(code) & 0xFF] & 0x1)) \ | |
3c7649c1 PE |
127 | ? (int) (((((code) >> 24) - (charset)->code_space[12]) \ |
128 | * (charset)->code_space[11]) \ | |
129 | + (((((code) >> 16) & 0xFF) - (charset)->code_space[8]) \ | |
130 | * (charset)->code_space[7]) \ | |
131 | + (((((code) >> 8) & 0xFF) - (charset)->code_space[4]) \ | |
132 | * (charset)->code_space[3]) \ | |
133 | + (((code) & 0xFF) - (charset)->code_space[0]) \ | |
134 | - ((charset)->char_index_offset)) \ | |
3263d5a2 KH |
135 | : -1) |
136 | ||
137 | ||
3c7649c1 PE |
138 | /* Return the code-point for the character index IDX in CHARSET. |
139 | IDX should be an unsigned int variable in a valid range (which is | |
140 | always in nonnegative int range too). IDX contains garbage afterwards. */ | |
3263d5a2 | 141 | |
820ee249 KH |
142 | #define INDEX_TO_CODE_POINT(charset, idx) \ |
143 | ((charset)->code_linear_p \ | |
144 | ? (idx) + (charset)->min_code \ | |
145 | : (idx += (charset)->char_index_offset, \ | |
146 | (((charset)->code_space[0] + (idx) % (charset)->code_space[2]) \ | |
147 | | (((charset)->code_space[4] \ | |
148 | + ((idx) / (charset)->code_space[3] % (charset)->code_space[6])) \ | |
149 | << 8) \ | |
150 | | (((charset)->code_space[8] \ | |
151 | + ((idx) / (charset)->code_space[7] % (charset)->code_space[10])) \ | |
152 | << 16) \ | |
153 | | (((charset)->code_space[12] + ((idx) / (charset)->code_space[11])) \ | |
154 | << 24)))) | |
4cf9710d | 155 | |
ecca2aad KH |
156 | /* Structure to hold mapping tables for a charset. Used by temacs |
157 | invoked for dumping. */ | |
8a73a704 | 158 | |
ecca2aad KH |
159 | static struct |
160 | { | |
161 | /* The current charset for which the following tables are setup. */ | |
162 | struct charset *current; | |
163 | ||
164 | /* 1 iff the following table is used for encoder. */ | |
165 | short for_encoder; | |
166 | ||
0b381c7e JB |
167 | /* When the following table is used for encoding, minimum and |
168 | maximum character of the current charset. */ | |
ecca2aad KH |
169 | int min_char, max_char; |
170 | ||
e4769531 | 171 | /* A Unicode character corresponding to the code index 0 (i.e. the |
ecca2aad | 172 | minimum code-point) of the current charset, or -1 if the code |
e4769531 | 173 | index 0 is not a Unicode character. This is checked when |
ecca2aad KH |
174 | table.encoder[CHAR] is zero. */ |
175 | int zero_index_char; | |
176 | ||
177 | union { | |
178 | /* Table mapping code-indices (not code-points) of the current | |
179 | charset to Unicode characters. If decoder[CHAR] is -1, CHAR | |
180 | doesn't belong to the current charset. */ | |
181 | int decoder[0x10000]; | |
182 | /* Table mapping Unicode characters to code-indices of the current | |
183 | charset. The first 0x10000 elements are for BMP (0..0xFFFF), | |
184 | and the last 0x10000 are for SMP (0x10000..0x1FFFF) or SIP | |
185 | (0x20000..0x2FFFF). Note that there is no charset map that | |
186 | uses both SMP and SIP. */ | |
187 | unsigned short encoder[0x20000]; | |
188 | } table; | |
189 | } *temp_charset_work; | |
190 | ||
191 | #define SET_TEMP_CHARSET_WORK_ENCODER(C, CODE) \ | |
192 | do { \ | |
193 | if ((CODE) == 0) \ | |
194 | temp_charset_work->zero_index_char = (C); \ | |
195 | else if ((C) < 0x20000) \ | |
196 | temp_charset_work->table.encoder[(C)] = (CODE); \ | |
197 | else \ | |
198 | temp_charset_work->table.encoder[(C) - 0x10000] = (CODE); \ | |
199 | } while (0) | |
200 | ||
201 | #define GET_TEMP_CHARSET_WORK_ENCODER(C) \ | |
202 | ((C) == temp_charset_work->zero_index_char ? 0 \ | |
203 | : (C) < 0x20000 ? (temp_charset_work->table.encoder[(C)] \ | |
204 | ? (int) temp_charset_work->table.encoder[(C)] : -1) \ | |
205 | : temp_charset_work->table.encoder[(C) - 0x10000] \ | |
206 | ? temp_charset_work->table.encoder[(C) - 0x10000] : -1) | |
207 | ||
208 | #define SET_TEMP_CHARSET_WORK_DECODER(C, CODE) \ | |
209 | (temp_charset_work->table.decoder[(CODE)] = (C)) | |
210 | ||
211 | #define GET_TEMP_CHARSET_WORK_DECODER(CODE) \ | |
212 | (temp_charset_work->table.decoder[(CODE)]) | |
046b1f03 | 213 | \f |
93bcb785 | 214 | |
e9ce014c KH |
215 | /* Set to 1 to warn that a charset map is loaded and thus a buffer |
216 | text and a string data may be relocated. */ | |
d5172d4f | 217 | bool charset_map_loaded; |
4ed46869 | 218 | |
e9ce014c | 219 | struct charset_map_entries |
4ed46869 | 220 | { |
e9ce014c KH |
221 | struct { |
222 | unsigned from, to; | |
223 | int c; | |
224 | } entry[0x10000]; | |
225 | struct charset_map_entries *next; | |
226 | }; | |
227 | ||
ecca2aad KH |
228 | /* Load the mapping information of CHARSET from ENTRIES for |
229 | initializing (CONTROL_FLAG == 0), decoding (CONTROL_FLAG == 1), and | |
230 | encoding (CONTROL_FLAG == 2). | |
231 | ||
232 | If CONTROL_FLAG is 0, setup CHARSET->min_char, CHARSET->max_char, | |
233 | and CHARSET->fast_map. | |
234 | ||
235 | If CONTROL_FLAG is 1, setup the following tables according to | |
236 | CHARSET->method and inhibit_load_charset_map. | |
4cf9710d | 237 | |
ecca2aad KH |
238 | CHARSET->method | inhibit_lcm == 0 | inhibit_lcm == 1 |
239 | ----------------------+--------------------+--------------------------- | |
240 | CHARSET_METHOD_MAP | CHARSET->decoder | temp_charset_work->decoder | |
241 | ----------------------+--------------------+--------------------------- | |
242 | CHARSET_METHOD_OFFSET | Vchar_unify_table | temp_charset_work->decoder | |
8a73a704 | 243 | |
ecca2aad | 244 | If CONTROL_FLAG is 2, setup the following tables. |
93bcb785 | 245 | |
ecca2aad KH |
246 | CHARSET->method | inhibit_lcm == 0 | inhibit_lcm == 1 |
247 | ----------------------+--------------------+--------------------------- | |
248 | CHARSET_METHOD_MAP | CHARSET->encoder | temp_charset_work->encoder | |
249 | ----------------------+--------------------+-------------------------- | |
250 | CHARSET_METHOD_OFFSET | CHARSET->deunifier | temp_charset_work->encoder | |
251 | */ | |
4ed46869 | 252 | |
3263d5a2 | 253 | static void |
971de7fb | 254 | load_charset_map (struct charset *charset, struct charset_map_entries *entries, int n_entries, int control_flag) |
4ed46869 | 255 | { |
89ef49df | 256 | Lisp_Object vec IF_LINT (= Qnil), table IF_LINT (= Qnil); |
3263d5a2 | 257 | unsigned max_code = CHARSET_MAX_CODE (charset); |
d5172d4f | 258 | bool ascii_compatible_p = charset->ascii_compatible_p; |
3263d5a2 | 259 | int min_char, max_char, nonascii_min_char; |
3263d5a2 | 260 | int i; |
3263d5a2 | 261 | unsigned char *fast_map = charset->fast_map; |
99529c2c | 262 | |
e9ce014c KH |
263 | if (n_entries <= 0) |
264 | return; | |
265 | ||
ecca2aad | 266 | if (control_flag) |
8ac5a9cc | 267 | { |
ecca2aad KH |
268 | if (! inhibit_load_charset_map) |
269 | { | |
270 | if (control_flag == 1) | |
271 | { | |
272 | if (charset->method == CHARSET_METHOD_MAP) | |
273 | { | |
274 | int n = CODE_POINT_TO_INDEX (charset, max_code) + 1; | |
6662e69b | 275 | |
4939150c PE |
276 | vec = Fmake_vector (make_number (n), make_number (-1)); |
277 | set_charset_attr (charset, charset_decoder, vec); | |
ecca2aad KH |
278 | } |
279 | else | |
280 | { | |
281 | char_table_set_range (Vchar_unify_table, | |
282 | charset->min_char, charset->max_char, | |
283 | Qnil); | |
284 | } | |
285 | } | |
286 | else | |
287 | { | |
288 | table = Fmake_char_table (Qnil, Qnil); | |
4939150c PE |
289 | set_charset_attr (charset, |
290 | (charset->method == CHARSET_METHOD_MAP | |
291 | ? charset_encoder : charset_deunifier), | |
292 | table); | |
ecca2aad KH |
293 | } |
294 | } | |
295 | else | |
296 | { | |
297 | if (! temp_charset_work) | |
38182d90 | 298 | temp_charset_work = xmalloc (sizeof *temp_charset_work); |
ecca2aad KH |
299 | if (control_flag == 1) |
300 | { | |
301 | memset (temp_charset_work->table.decoder, -1, | |
302 | sizeof (int) * 0x10000); | |
ecca2aad KH |
303 | } |
304 | else | |
305 | { | |
306 | memset (temp_charset_work->table.encoder, 0, | |
307 | sizeof (unsigned short) * 0x20000); | |
308 | temp_charset_work->zero_index_char = -1; | |
309 | } | |
310 | temp_charset_work->current = charset; | |
311 | temp_charset_work->for_encoder = (control_flag == 2); | |
312 | control_flag += 2; | |
313 | } | |
3263d5a2 | 314 | charset_map_loaded = 1; |
2e344af3 | 315 | } |
6662e69b | 316 | |
e9ce014c | 317 | min_char = max_char = entries->entry[0].c; |
3263d5a2 | 318 | nonascii_min_char = MAX_CHAR; |
e9ce014c | 319 | for (i = 0; i < n_entries; i++) |
2e344af3 | 320 | { |
e9ce014c | 321 | unsigned from, to; |
3befa583 | 322 | int from_index, to_index, lim_index; |
3b4f4446 | 323 | int from_c, to_c; |
e9ce014c | 324 | int idx = i % 0x10000; |
3263d5a2 | 325 | |
e9ce014c KH |
326 | if (i > 0 && idx == 0) |
327 | entries = entries->next; | |
328 | from = entries->entry[idx].from; | |
329 | to = entries->entry[idx].to; | |
3b4f4446 KH |
330 | from_c = entries->entry[idx].c; |
331 | from_index = CODE_POINT_TO_INDEX (charset, from); | |
332 | if (from == to) | |
6662e69b | 333 | { |
3b4f4446 KH |
334 | to_index = from_index; |
335 | to_c = from_c; | |
6662e69b | 336 | } |
3b4f4446 | 337 | else |
6662e69b | 338 | { |
3b4f4446 KH |
339 | to_index = CODE_POINT_TO_INDEX (charset, to); |
340 | to_c = from_c + (to_index - from_index); | |
6662e69b | 341 | } |
3b4f4446 KH |
342 | if (from_index < 0 || to_index < 0) |
343 | continue; | |
3befa583 | 344 | lim_index = to_index + 1; |
3263d5a2 | 345 | |
ecca2aad KH |
346 | if (to_c > max_char) |
347 | max_char = to_c; | |
348 | else if (from_c < min_char) | |
349 | min_char = from_c; | |
6662e69b | 350 | |
ecca2aad KH |
351 | if (control_flag == 1) |
352 | { | |
353 | if (charset->method == CHARSET_METHOD_MAP) | |
3befa583 | 354 | for (; from_index < lim_index; from_index++, from_c++) |
ecca2aad KH |
355 | ASET (vec, from_index, make_number (from_c)); |
356 | else | |
3befa583 | 357 | for (; from_index < lim_index; from_index++, from_c++) |
ecca2aad KH |
358 | CHAR_TABLE_SET (Vchar_unify_table, |
359 | CHARSET_CODE_OFFSET (charset) + from_index, | |
360 | make_number (from_c)); | |
361 | } | |
362 | else if (control_flag == 2) | |
363 | { | |
364 | if (charset->method == CHARSET_METHOD_MAP | |
365 | && CHARSET_COMPACT_CODES_P (charset)) | |
3befa583 | 366 | for (; from_index < lim_index; from_index++, from_c++) |
ecca2aad | 367 | { |
3c7649c1 PE |
368 | unsigned code = from_index; |
369 | code = INDEX_TO_CODE_POINT (charset, code); | |
ecca2aad KH |
370 | |
371 | if (NILP (CHAR_TABLE_REF (table, from_c))) | |
372 | CHAR_TABLE_SET (table, from_c, make_number (code)); | |
373 | } | |
374 | else | |
3befa583 | 375 | for (; from_index < lim_index; from_index++, from_c++) |
ecca2aad KH |
376 | { |
377 | if (NILP (CHAR_TABLE_REF (table, from_c))) | |
378 | CHAR_TABLE_SET (table, from_c, make_number (from_index)); | |
379 | } | |
380 | } | |
381 | else if (control_flag == 3) | |
3befa583 | 382 | for (; from_index < lim_index; from_index++, from_c++) |
ecca2aad KH |
383 | SET_TEMP_CHARSET_WORK_DECODER (from_c, from_index); |
384 | else if (control_flag == 4) | |
3befa583 | 385 | for (; from_index < lim_index; from_index++, from_c++) |
ecca2aad KH |
386 | SET_TEMP_CHARSET_WORK_ENCODER (from_c, from_index); |
387 | else /* control_flag == 0 */ | |
388 | { | |
3b4f4446 KH |
389 | if (ascii_compatible_p) |
390 | { | |
391 | if (! ASCII_BYTE_P (from_c)) | |
392 | { | |
393 | if (from_c < nonascii_min_char) | |
394 | nonascii_min_char = from_c; | |
395 | } | |
396 | else if (! ASCII_BYTE_P (to_c)) | |
397 | { | |
398 | nonascii_min_char = 0x80; | |
399 | } | |
400 | } | |
177c0ea7 | 401 | |
ecca2aad KH |
402 | for (; from_c <= to_c; from_c++) |
403 | CHARSET_FAST_MAP_SET (from_c, fast_map); | |
2e344af3 | 404 | } |
8ac5a9cc | 405 | } |
3263d5a2 | 406 | |
ecca2aad | 407 | if (control_flag == 0) |
4ed46869 | 408 | { |
3263d5a2 KH |
409 | CHARSET_MIN_CHAR (charset) = (ascii_compatible_p |
410 | ? nonascii_min_char : min_char); | |
411 | CHARSET_MAX_CHAR (charset) = max_char; | |
4ed46869 | 412 | } |
ecca2aad KH |
413 | else if (control_flag == 4) |
414 | { | |
415 | temp_charset_work->min_char = min_char; | |
416 | temp_charset_work->max_char = max_char; | |
417 | } | |
4ed46869 KH |
418 | } |
419 | ||
12bcae05 | 420 | |
3263d5a2 | 421 | /* Read a hexadecimal number (preceded by "0x") from the file FP while |
3b59c351 | 422 | paying attention to comment character '#'. */ |
12bcae05 | 423 | |
b0ab8123 | 424 | static unsigned |
d5172d4f | 425 | read_hex (FILE *fp, bool *eof, bool *overflow) |
12bcae05 | 426 | { |
3263d5a2 KH |
427 | int c; |
428 | unsigned n; | |
12bcae05 | 429 | |
3263d5a2 KH |
430 | while ((c = getc (fp)) != EOF) |
431 | { | |
69f8de5b | 432 | if (c == '#') |
3263d5a2 KH |
433 | { |
434 | while ((c = getc (fp)) != EOF && c != '\n'); | |
435 | } | |
436 | else if (c == '0') | |
437 | { | |
438 | if ((c = getc (fp)) == EOF || c == 'x') | |
439 | break; | |
440 | } | |
8f924df7 | 441 | } |
3263d5a2 KH |
442 | if (c == EOF) |
443 | { | |
444 | *eof = 1; | |
445 | return 0; | |
446 | } | |
3263d5a2 | 447 | n = 0; |
620f13b0 | 448 | while (c_isxdigit (c = getc (fp))) |
caeeedc1 PE |
449 | { |
450 | if (UINT_MAX >> 4 < n) | |
451 | *overflow = 1; | |
3263d5a2 | 452 | n = ((n << 4) |
caeeedc1 PE |
453 | | (c - ('0' <= c && c <= '9' ? '0' |
454 | : 'A' <= c && c <= 'F' ? 'A' - 10 | |
455 | : 'a' - 10))); | |
456 | } | |
e9ce014c KH |
457 | if (c != EOF) |
458 | ungetc (c, fp); | |
3263d5a2 KH |
459 | return n; |
460 | } | |
12bcae05 | 461 | |
3263d5a2 | 462 | /* Return a mapping vector for CHARSET loaded from MAPFILE. |
e9ce014c KH |
463 | Each line of MAPFILE has this form |
464 | 0xAAAA 0xCCCC | |
465 | where 0xAAAA is a code-point and 0xCCCC is the corresponding | |
466 | character code, or this form | |
467 | 0xAAAA-0xBBBB 0xCCCC | |
468 | where 0xAAAA and 0xBBBB are code-points specifying a range, and | |
469 | 0xCCCC is the first character code of the range. | |
4ed46869 | 470 | |
3263d5a2 KH |
471 | The returned vector has this form: |
472 | [ CODE1 CHAR1 CODE2 CHAR2 .... ] | |
e9ce014c | 473 | where CODE1 is a code-point or a cons of code-points specifying a |
dde2559c KH |
474 | range. |
475 | ||
18a10a21 JB |
476 | Note that this function uses `openp' to open MAPFILE but ignores |
477 | `file-name-handler-alist' to avoid running any Lisp code. */ | |
4ed46869 | 478 | |
e9ce014c | 479 | static void |
971de7fb | 480 | load_charset_map_from_file (struct charset *charset, Lisp_Object mapfile, int control_flag) |
4ed46869 | 481 | { |
e9ce014c KH |
482 | unsigned min_code = CHARSET_MIN_CODE (charset); |
483 | unsigned max_code = CHARSET_MAX_CODE (charset); | |
3263d5a2 KH |
484 | int fd; |
485 | FILE *fp; | |
3263d5a2 | 486 | Lisp_Object suffixes; |
e9ce014c | 487 | struct charset_map_entries *head, *entries; |
d311d28c PE |
488 | int n_entries; |
489 | ptrdiff_t count; | |
a2f3eb19 | 490 | USE_SAFE_ALLOCA; |
4ed46869 | 491 | |
3263d5a2 KH |
492 | suffixes = Fcons (build_string (".map"), |
493 | Fcons (build_string (".TXT"), Qnil)); | |
4ed46869 | 494 | |
a2f3eb19 | 495 | count = SPECPDL_INDEX (); |
dde2559c | 496 | specbind (Qfile_name_handler_alist, Qnil); |
4beef065 | 497 | fd = openp (Vcharset_map_path, mapfile, suffixes, NULL, Qnil); |
dde2559c | 498 | unbind_to (count, Qnil); |
3263d5a2 KH |
499 | if (fd < 0 |
500 | || ! (fp = fdopen (fd, "r"))) | |
5fdb398c | 501 | error ("Failure in loading charset map: %s", SDATA (mapfile)); |
4ed46869 | 502 | |
a2f3eb19 CY |
503 | /* Use SAFE_ALLOCA instead of alloca, as `charset_map_entries' is |
504 | large (larger than MAX_ALLOCA). */ | |
98c6f1e3 | 505 | head = SAFE_ALLOCA (sizeof *head); |
a2f3eb19 | 506 | entries = head; |
72af86bd | 507 | memset (entries, 0, sizeof (struct charset_map_entries)); |
a2f3eb19 | 508 | |
e9ce014c | 509 | n_entries = 0; |
3263d5a2 KH |
510 | while (1) |
511 | { | |
caeeedc1 | 512 | unsigned from, to, c; |
e9ce014c | 513 | int idx; |
d5172d4f | 514 | bool eof = 0, overflow = 0; |
4ed46869 | 515 | |
caeeedc1 | 516 | from = read_hex (fp, &eof, &overflow); |
3263d5a2 KH |
517 | if (eof) |
518 | break; | |
e9ce014c | 519 | if (getc (fp) == '-') |
caeeedc1 | 520 | to = read_hex (fp, &eof, &overflow); |
e9ce014c KH |
521 | else |
522 | to = from; | |
caeeedc1 PE |
523 | if (eof) |
524 | break; | |
525 | c = read_hex (fp, &eof, &overflow); | |
526 | if (eof) | |
527 | break; | |
ac4137cc | 528 | |
caeeedc1 PE |
529 | if (overflow) |
530 | continue; | |
e9ce014c KH |
531 | if (from < min_code || to > max_code || from > to || c > MAX_CHAR) |
532 | continue; | |
ac4137cc | 533 | |
e9ce014c | 534 | if (n_entries > 0 && (n_entries % 0x10000) == 0) |
3263d5a2 | 535 | { |
98c6f1e3 | 536 | entries->next = SAFE_ALLOCA (sizeof *entries->next); |
e9ce014c | 537 | entries = entries->next; |
72af86bd | 538 | memset (entries, 0, sizeof (struct charset_map_entries)); |
d311d28c | 539 | n_entries = 0; |
3263d5a2 | 540 | } |
d311d28c | 541 | idx = n_entries; |
e9ce014c KH |
542 | entries->entry[idx].from = from; |
543 | entries->entry[idx].to = to; | |
544 | entries->entry[idx].c = c; | |
545 | n_entries++; | |
3263d5a2 KH |
546 | } |
547 | fclose (fp); | |
177c0ea7 | 548 | |
e9ce014c | 549 | load_charset_map (charset, head, n_entries, control_flag); |
a2f3eb19 | 550 | SAFE_FREE (); |
4ed46869 KH |
551 | } |
552 | ||
e9ce014c | 553 | static void |
971de7fb | 554 | load_charset_map_from_vector (struct charset *charset, Lisp_Object vec, int control_flag) |
23d2a7f1 | 555 | { |
e9ce014c KH |
556 | unsigned min_code = CHARSET_MIN_CODE (charset); |
557 | unsigned max_code = CHARSET_MAX_CODE (charset); | |
558 | struct charset_map_entries *head, *entries; | |
559 | int n_entries; | |
560 | int len = ASIZE (vec); | |
561 | int i; | |
a2f3eb19 | 562 | USE_SAFE_ALLOCA; |
23d2a7f1 | 563 | |
e9ce014c | 564 | if (len % 2 == 1) |
3263d5a2 | 565 | { |
e9ce014c KH |
566 | add_to_log ("Failure in loading charset map: %V", vec, Qnil); |
567 | return; | |
3263d5a2 | 568 | } |
35e623fb | 569 | |
a2f3eb19 CY |
570 | /* Use SAFE_ALLOCA instead of alloca, as `charset_map_entries' is |
571 | large (larger than MAX_ALLOCA). */ | |
98c6f1e3 | 572 | head = SAFE_ALLOCA (sizeof *head); |
a2f3eb19 | 573 | entries = head; |
72af86bd | 574 | memset (entries, 0, sizeof (struct charset_map_entries)); |
a2f3eb19 | 575 | |
e9ce014c KH |
576 | n_entries = 0; |
577 | for (i = 0; i < len; i += 2) | |
35e623fb | 578 | { |
e9ce014c KH |
579 | Lisp_Object val, val2; |
580 | unsigned from, to; | |
d311d28c | 581 | EMACS_INT c; |
e9ce014c | 582 | int idx; |
d2665018 | 583 | |
e9ce014c KH |
584 | val = AREF (vec, i); |
585 | if (CONSP (val)) | |
bbf12bb3 | 586 | { |
e9ce014c KH |
587 | val2 = XCDR (val); |
588 | val = XCAR (val); | |
e9ce014c KH |
589 | from = XFASTINT (val); |
590 | to = XFASTINT (val2); | |
bbf12bb3 | 591 | } |
e9ce014c | 592 | else |
d311d28c | 593 | from = to = XFASTINT (val); |
e9ce014c KH |
594 | val = AREF (vec, i + 1); |
595 | CHECK_NATNUM (val); | |
596 | c = XFASTINT (val); | |
76d7b829 | 597 | |
e9ce014c KH |
598 | if (from < min_code || to > max_code || from > to || c > MAX_CHAR) |
599 | continue; | |
76d7b829 | 600 | |
dcc694d7 | 601 | if (n_entries > 0 && (n_entries % 0x10000) == 0) |
e9ce014c | 602 | { |
98c6f1e3 | 603 | entries->next = SAFE_ALLOCA (sizeof *entries->next); |
e9ce014c | 604 | entries = entries->next; |
72af86bd | 605 | memset (entries, 0, sizeof (struct charset_map_entries)); |
e9ce014c KH |
606 | } |
607 | idx = n_entries % 0x10000; | |
608 | entries->entry[idx].from = from; | |
609 | entries->entry[idx].to = to; | |
610 | entries->entry[idx].c = c; | |
611 | n_entries++; | |
612 | } | |
76d7b829 | 613 | |
e9ce014c | 614 | load_charset_map (charset, head, n_entries, control_flag); |
a2f3eb19 | 615 | SAFE_FREE (); |
ac4137cc KH |
616 | } |
617 | ||
ecca2aad KH |
618 | |
619 | /* Load a mapping table for CHARSET. CONTROL-FLAG tells what kind of | |
620 | map it is (see the comment of load_charset_map for the detail). */ | |
621 | ||
3263d5a2 | 622 | static void |
971de7fb | 623 | load_charset (struct charset *charset, int control_flag) |
76d7b829 | 624 | { |
ecca2aad | 625 | Lisp_Object map; |
76d7b829 | 626 | |
ecca2aad KH |
627 | if (inhibit_load_charset_map |
628 | && temp_charset_work | |
629 | && charset == temp_charset_work->current | |
56f00ed2 | 630 | && ((control_flag == 2) == temp_charset_work->for_encoder)) |
ecca2aad KH |
631 | return; |
632 | ||
633 | if (CHARSET_METHOD (charset) == CHARSET_METHOD_MAP) | |
634 | map = CHARSET_MAP (charset); | |
0ac2c299 PE |
635 | else |
636 | { | |
637 | if (! CHARSET_UNIFIED_P (charset)) | |
1088b922 | 638 | emacs_abort (); |
0ac2c299 PE |
639 | map = CHARSET_UNIFY_MAP (charset); |
640 | } | |
ecca2aad KH |
641 | if (STRINGP (map)) |
642 | load_charset_map_from_file (charset, map, control_flag); | |
643 | else | |
644 | load_charset_map_from_vector (charset, map, control_flag); | |
4ed46869 | 645 | } |
76d7b829 | 646 | |
3263d5a2 KH |
647 | |
648 | DEFUN ("charsetp", Fcharsetp, Scharsetp, 1, 1, 0, | |
649 | doc: /* Return non-nil if and only if OBJECT is a charset.*/) | |
5842a27b | 650 | (Lisp_Object object) |
23d2a7f1 | 651 | { |
3263d5a2 | 652 | return (CHARSETP (object) ? Qt : Qnil); |
76d7b829 KH |
653 | } |
654 | ||
4ed46869 | 655 | |
a2cb4e63 PE |
656 | static void |
657 | map_charset_for_dump (void (*c_function) (Lisp_Object, Lisp_Object), | |
658 | Lisp_Object function, Lisp_Object arg, | |
659 | unsigned int from, unsigned int to) | |
ecca2aad KH |
660 | { |
661 | int from_idx = CODE_POINT_TO_INDEX (temp_charset_work->current, from); | |
662 | int to_idx = CODE_POINT_TO_INDEX (temp_charset_work->current, to); | |
663 | Lisp_Object range; | |
664 | int c, stop; | |
665 | struct gcpro gcpro1; | |
666 | ||
667 | range = Fcons (Qnil, Qnil); | |
668 | GCPRO1 (range); | |
669 | ||
670 | c = temp_charset_work->min_char; | |
671 | stop = (temp_charset_work->max_char < 0x20000 | |
672 | ? temp_charset_work->max_char : 0xFFFF); | |
51b59d79 | 673 | |
ecca2aad KH |
674 | while (1) |
675 | { | |
f6095868 | 676 | int idx = GET_TEMP_CHARSET_WORK_ENCODER (c); |
ecca2aad | 677 | |
f6095868 | 678 | if (idx >= from_idx && idx <= to_idx) |
ecca2aad KH |
679 | { |
680 | if (NILP (XCAR (range))) | |
681 | XSETCAR (range, make_number (c)); | |
682 | } | |
683 | else if (! NILP (XCAR (range))) | |
684 | { | |
685 | XSETCDR (range, make_number (c - 1)); | |
686 | if (c_function) | |
687 | (*c_function) (arg, range); | |
688 | else | |
689 | call2 (function, range, arg); | |
690 | XSETCAR (range, Qnil); | |
691 | } | |
692 | if (c == stop) | |
693 | { | |
694 | if (c == temp_charset_work->max_char) | |
695 | { | |
696 | if (! NILP (XCAR (range))) | |
697 | { | |
698 | XSETCDR (range, make_number (c)); | |
699 | if (c_function) | |
700 | (*c_function) (arg, range); | |
701 | else | |
702 | call2 (function, range, arg); | |
703 | } | |
704 | break; | |
705 | } | |
706 | c = 0x1FFFF; | |
707 | stop = temp_charset_work->max_char; | |
708 | } | |
709 | c++; | |
710 | } | |
c542407d | 711 | UNGCPRO; |
ecca2aad KH |
712 | } |
713 | ||
4ed46869 | 714 | void |
6f704c76 DN |
715 | map_charset_chars (void (*c_function)(Lisp_Object, Lisp_Object), Lisp_Object function, |
716 | Lisp_Object arg, struct charset *charset, unsigned from, unsigned to) | |
4ed46869 | 717 | { |
3263d5a2 | 718 | Lisp_Object range; |
d5172d4f PE |
719 | bool partial = (from > CHARSET_MIN_CODE (charset) |
720 | || to < CHARSET_MAX_CODE (charset)); | |
374c5cfd | 721 | |
3263d5a2 | 722 | if (CHARSET_METHOD (charset) == CHARSET_METHOD_OFFSET) |
4ed46869 | 723 | { |
374c5cfd KH |
724 | int from_idx = CODE_POINT_TO_INDEX (charset, from); |
725 | int to_idx = CODE_POINT_TO_INDEX (charset, to); | |
726 | int from_c = from_idx + CHARSET_CODE_OFFSET (charset); | |
727 | int to_c = to_idx + CHARSET_CODE_OFFSET (charset); | |
728 | ||
ecca2aad KH |
729 | if (CHARSET_UNIFIED_P (charset)) |
730 | { | |
731 | if (! CHAR_TABLE_P (CHARSET_DEUNIFIER (charset))) | |
732 | load_charset (charset, 2); | |
733 | if (CHAR_TABLE_P (CHARSET_DEUNIFIER (charset))) | |
734 | map_char_table_for_charset (c_function, function, | |
735 | CHARSET_DEUNIFIER (charset), arg, | |
736 | partial ? charset : NULL, from, to); | |
737 | else | |
738 | map_charset_for_dump (c_function, function, arg, from, to); | |
739 | } | |
740 | ||
374c5cfd | 741 | range = Fcons (make_number (from_c), make_number (to_c)); |
3263d5a2 | 742 | if (NILP (function)) |
5af5dd92 | 743 | (*c_function) (arg, range); |
3263d5a2 KH |
744 | else |
745 | call2 (function, range, arg); | |
c83ef371 | 746 | } |
3263d5a2 KH |
747 | else if (CHARSET_METHOD (charset) == CHARSET_METHOD_MAP) |
748 | { | |
749 | if (! CHAR_TABLE_P (CHARSET_ENCODER (charset))) | |
ecca2aad KH |
750 | load_charset (charset, 2); |
751 | if (CHAR_TABLE_P (CHARSET_ENCODER (charset))) | |
752 | map_char_table_for_charset (c_function, function, | |
753 | CHARSET_ENCODER (charset), arg, | |
754 | partial ? charset : NULL, from, to); | |
755 | else | |
756 | map_charset_for_dump (c_function, function, arg, from, to); | |
3263d5a2 | 757 | } |
374c5cfd | 758 | else if (CHARSET_METHOD (charset) == CHARSET_METHOD_SUBSET) |
4ed46869 | 759 | { |
374c5cfd KH |
760 | Lisp_Object subset_info; |
761 | int offset; | |
762 | ||
763 | subset_info = CHARSET_SUBSET (charset); | |
764 | charset = CHARSET_FROM_ID (XFASTINT (AREF (subset_info, 0))); | |
765 | offset = XINT (AREF (subset_info, 3)); | |
766 | from -= offset; | |
767 | if (from < XFASTINT (AREF (subset_info, 1))) | |
768 | from = XFASTINT (AREF (subset_info, 1)); | |
769 | to -= offset; | |
770 | if (to > XFASTINT (AREF (subset_info, 2))) | |
771 | to = XFASTINT (AREF (subset_info, 2)); | |
772 | map_charset_chars (c_function, function, arg, charset, from, to); | |
4ed46869 | 773 | } |
374c5cfd KH |
774 | else /* i.e. CHARSET_METHOD_SUPERSET */ |
775 | { | |
776 | Lisp_Object parents; | |
4ed46869 | 777 | |
374c5cfd KH |
778 | for (parents = CHARSET_SUPERSET (charset); CONSP (parents); |
779 | parents = XCDR (parents)) | |
bbf12bb3 | 780 | { |
374c5cfd KH |
781 | int offset; |
782 | unsigned this_from, this_to; | |
783 | ||
784 | charset = CHARSET_FROM_ID (XFASTINT (XCAR (XCAR (parents)))); | |
785 | offset = XINT (XCDR (XCAR (parents))); | |
12b55765 KH |
786 | this_from = from > offset ? from - offset : 0; |
787 | this_to = to > offset ? to - offset : 0; | |
374c5cfd KH |
788 | if (this_from < CHARSET_MIN_CODE (charset)) |
789 | this_from = CHARSET_MIN_CODE (charset); | |
790 | if (this_to > CHARSET_MAX_CODE (charset)) | |
791 | this_to = CHARSET_MAX_CODE (charset); | |
111daccf KH |
792 | map_charset_chars (c_function, function, arg, charset, |
793 | this_from, this_to); | |
bbf12bb3 | 794 | } |
35e623fb | 795 | } |
4ed46869 KH |
796 | } |
797 | ||
374c5cfd | 798 | DEFUN ("map-charset-chars", Fmap_charset_chars, Smap_charset_chars, 2, 5, 0, |
04c2f2c5 | 799 | doc: /* Call FUNCTION for all characters in CHARSET. |
374c5cfd | 800 | FUNCTION is called with an argument RANGE and the optional 3rd |
3263d5a2 | 801 | argument ARG. |
4ed46869 | 802 | |
374c5cfd KH |
803 | RANGE is a cons (FROM . TO), where FROM and TO indicate a range of |
804 | characters contained in CHARSET. | |
4ed46869 | 805 | |
374c5cfd | 806 | The optional 4th and 5th arguments FROM-CODE and TO-CODE specify the |
72d51285 | 807 | range of code points (in CHARSET) of target characters. */) |
5842a27b | 808 | (Lisp_Object function, Lisp_Object charset, Lisp_Object arg, Lisp_Object from_code, Lisp_Object to_code) |
4ed46869 | 809 | { |
374c5cfd | 810 | struct charset *cs; |
16fed1fc | 811 | unsigned from, to; |
4ed46869 | 812 | |
374c5cfd KH |
813 | CHECK_CHARSET_GET_CHARSET (charset, cs); |
814 | if (NILP (from_code)) | |
16fed1fc | 815 | from = CHARSET_MIN_CODE (cs); |
970b7474 | 816 | else |
4ed46869 | 817 | { |
970b7474 KH |
818 | from = XINT (from_code); |
819 | if (from < CHARSET_MIN_CODE (cs)) | |
820 | from = CHARSET_MIN_CODE (cs); | |
4ed46869 | 821 | } |
374c5cfd | 822 | if (NILP (to_code)) |
970b7474 | 823 | to = CHARSET_MAX_CODE (cs); |
4ed46869 KH |
824 | else |
825 | { | |
970b7474 KH |
826 | to = XINT (to_code); |
827 | if (to > CHARSET_MAX_CODE (cs)) | |
828 | to = CHARSET_MAX_CODE (cs); | |
4ed46869 | 829 | } |
16fed1fc | 830 | map_charset_chars (NULL, function, arg, cs, from, to); |
3263d5a2 | 831 | return Qnil; |
35e623fb | 832 | } |
4ed46869 | 833 | |
4ed46869 | 834 | |
3263d5a2 KH |
835 | /* Define a charset according to the arguments. The Nth argument is |
836 | the Nth attribute of the charset (the last attribute `charset-id' | |
837 | is not included). See the docstring of `define-charset' for the | |
838 | detail. */ | |
4ed46869 | 839 | |
3263d5a2 KH |
840 | DEFUN ("define-charset-internal", Fdefine_charset_internal, |
841 | Sdefine_charset_internal, charset_arg_max, MANY, 0, | |
04c2f2c5 DL |
842 | doc: /* For internal use only. |
843 | usage: (define-charset-internal ...) */) | |
f66c7cf8 | 844 | (ptrdiff_t nargs, Lisp_Object *args) |
4ed46869 | 845 | { |
3263d5a2 KH |
846 | /* Charset attr vector. */ |
847 | Lisp_Object attrs; | |
848 | Lisp_Object val; | |
0de4bb68 | 849 | EMACS_UINT hash_code; |
3263d5a2 | 850 | struct Lisp_Hash_Table *hash_table = XHASH_TABLE (Vcharset_hash_table); |
69f8de5b | 851 | int i, j; |
3263d5a2 KH |
852 | struct charset charset; |
853 | int id; | |
854 | int dimension; | |
d5172d4f | 855 | bool new_definition_p; |
3263d5a2 KH |
856 | int nchars; |
857 | ||
858 | if (nargs != charset_arg_max) | |
859 | return Fsignal (Qwrong_number_of_arguments, | |
860 | Fcons (intern ("define-charset-internal"), | |
861 | make_number (nargs))); | |
862 | ||
863 | attrs = Fmake_vector (make_number (charset_attr_max), Qnil); | |
864 | ||
865 | CHECK_SYMBOL (args[charset_arg_name]); | |
866 | ASET (attrs, charset_name, args[charset_arg_name]); | |
867 | ||
868 | val = args[charset_arg_code_space]; | |
c032b5f8 | 869 | for (i = 0, dimension = 0, nchars = 1; ; i++) |
76d7b829 | 870 | { |
d311d28c | 871 | Lisp_Object min_byte_obj, max_byte_obj; |
3263d5a2 KH |
872 | int min_byte, max_byte; |
873 | ||
d311d28c PE |
874 | min_byte_obj = Faref (val, make_number (i * 2)); |
875 | max_byte_obj = Faref (val, make_number (i * 2 + 1)); | |
af5a5a98 | 876 | CHECK_RANGED_INTEGER (min_byte_obj, 0, 255); |
d311d28c | 877 | min_byte = XINT (min_byte_obj); |
af5a5a98 | 878 | CHECK_RANGED_INTEGER (max_byte_obj, min_byte, 255); |
d311d28c | 879 | max_byte = XINT (max_byte_obj); |
3263d5a2 KH |
880 | charset.code_space[i * 4] = min_byte; |
881 | charset.code_space[i * 4 + 1] = max_byte; | |
882 | charset.code_space[i * 4 + 2] = max_byte - min_byte + 1; | |
3263d5a2 KH |
883 | if (max_byte > 0) |
884 | dimension = i + 1; | |
c032b5f8 PE |
885 | if (i == 3) |
886 | break; | |
887 | nchars *= charset.code_space[i * 4 + 2]; | |
888 | charset.code_space[i * 4 + 3] = nchars; | |
3263d5a2 | 889 | } |
4ed46869 | 890 | |
3263d5a2 KH |
891 | val = args[charset_arg_dimension]; |
892 | if (NILP (val)) | |
893 | charset.dimension = dimension; | |
894 | else | |
4ed46869 | 895 | { |
af5a5a98 | 896 | CHECK_RANGED_INTEGER (val, 1, 4); |
3263d5a2 | 897 | charset.dimension = XINT (val); |
4ed46869 KH |
898 | } |
899 | ||
3263d5a2 KH |
900 | charset.code_linear_p |
901 | = (charset.dimension == 1 | |
902 | || (charset.code_space[2] == 256 | |
903 | && (charset.dimension == 2 | |
904 | || (charset.code_space[6] == 256 | |
905 | && (charset.dimension == 3 | |
906 | || charset.code_space[10] == 256))))); | |
907 | ||
69f8de5b | 908 | if (! charset.code_linear_p) |
4ed46869 | 909 | { |
23f86fce | 910 | charset.code_space_mask = xzalloc (256); |
69f8de5b KH |
911 | for (i = 0; i < 4; i++) |
912 | for (j = charset.code_space[i * 4]; j <= charset.code_space[i * 4 + 1]; | |
913 | j++) | |
914 | charset.code_space_mask[j] |= (1 << i); | |
4ed46869 KH |
915 | } |
916 | ||
3263d5a2 | 917 | charset.iso_chars_96 = charset.code_space[2] == 96; |
4ed46869 | 918 | |
3263d5a2 KH |
919 | charset.min_code = (charset.code_space[0] |
920 | | (charset.code_space[4] << 8) | |
921 | | (charset.code_space[8] << 16) | |
3c7649c1 | 922 | | ((unsigned) charset.code_space[12] << 24)); |
3263d5a2 KH |
923 | charset.max_code = (charset.code_space[1] |
924 | | (charset.code_space[5] << 8) | |
925 | | (charset.code_space[9] << 16) | |
3c7649c1 | 926 | | ((unsigned) charset.code_space[13] << 24)); |
820ee249 | 927 | charset.char_index_offset = 0; |
8a73a704 | 928 | |
820ee249 KH |
929 | val = args[charset_arg_min_code]; |
930 | if (! NILP (val)) | |
931 | { | |
be44ca6c | 932 | unsigned code = cons_to_unsigned (val, UINT_MAX); |
fdb82f93 | 933 | |
820ee249 KH |
934 | if (code < charset.min_code |
935 | || code > charset.max_code) | |
3c7649c1 PE |
936 | args_out_of_range_3 (make_fixnum_or_float (charset.min_code), |
937 | make_fixnum_or_float (charset.max_code), val); | |
820ee249 KH |
938 | charset.char_index_offset = CODE_POINT_TO_INDEX (&charset, code); |
939 | charset.min_code = code; | |
940 | } | |
3fac5a51 | 941 | |
820ee249 KH |
942 | val = args[charset_arg_max_code]; |
943 | if (! NILP (val)) | |
3fac5a51 | 944 | { |
be44ca6c | 945 | unsigned code = cons_to_unsigned (val, UINT_MAX); |
820ee249 | 946 | |
820ee249 KH |
947 | if (code < charset.min_code |
948 | || code > charset.max_code) | |
3c7649c1 PE |
949 | args_out_of_range_3 (make_fixnum_or_float (charset.min_code), |
950 | make_fixnum_or_float (charset.max_code), val); | |
820ee249 | 951 | charset.max_code = code; |
3fac5a51 | 952 | } |
3fac5a51 | 953 | |
ecca2aad | 954 | charset.compact_codes_p = charset.max_code < 0x10000; |
4ed46869 | 955 | |
3263d5a2 KH |
956 | val = args[charset_arg_invalid_code]; |
957 | if (NILP (val)) | |
958 | { | |
959 | if (charset.min_code > 0) | |
960 | charset.invalid_code = 0; | |
bbf12bb3 KH |
961 | else |
962 | { | |
3c7649c1 | 963 | if (charset.max_code < UINT_MAX) |
3263d5a2 KH |
964 | charset.invalid_code = charset.max_code + 1; |
965 | else | |
966 | error ("Attribute :invalid-code must be specified"); | |
76d7b829 | 967 | } |
76d7b829 | 968 | } |
3263d5a2 | 969 | else |
3c7649c1 | 970 | charset.invalid_code = cons_to_unsigned (val, UINT_MAX); |
4ed46869 | 971 | |
3263d5a2 KH |
972 | val = args[charset_arg_iso_final]; |
973 | if (NILP (val)) | |
974 | charset.iso_final = -1; | |
975 | else | |
976 | { | |
977 | CHECK_NUMBER (val); | |
978 | if (XINT (val) < '0' || XINT (val) > 127) | |
c2982e87 | 979 | error ("Invalid iso-final-char: %"pI"d", XINT (val)); |
3263d5a2 KH |
980 | charset.iso_final = XINT (val); |
981 | } | |
4ed46869 | 982 | |
3263d5a2 KH |
983 | val = args[charset_arg_iso_revision]; |
984 | if (NILP (val)) | |
985 | charset.iso_revision = -1; | |
986 | else | |
4ed46869 | 987 | { |
af5a5a98 | 988 | CHECK_RANGED_INTEGER (val, -1, 63); |
3263d5a2 | 989 | charset.iso_revision = XINT (val); |
4ed46869 | 990 | } |
4ed46869 | 991 | |
3263d5a2 KH |
992 | val = args[charset_arg_emacs_mule_id]; |
993 | if (NILP (val)) | |
994 | charset.emacs_mule_id = -1; | |
4ed46869 KH |
995 | else |
996 | { | |
3263d5a2 KH |
997 | CHECK_NATNUM (val); |
998 | if ((XINT (val) > 0 && XINT (val) <= 128) || XINT (val) >= 256) | |
c2982e87 | 999 | error ("Invalid emacs-mule-id: %"pI"d", XINT (val)); |
3263d5a2 | 1000 | charset.emacs_mule_id = XINT (val); |
c83ef371 | 1001 | } |
f6302ac9 | 1002 | |
3263d5a2 | 1003 | charset.ascii_compatible_p = ! NILP (args[charset_arg_ascii_compatible_p]); |
1d67c29b | 1004 | |
3263d5a2 | 1005 | charset.supplementary_p = ! NILP (args[charset_arg_supplementary_p]); |
4ed46869 | 1006 | |
3263d5a2 KH |
1007 | charset.unified_p = 0; |
1008 | ||
72af86bd | 1009 | memset (charset.fast_map, 0, sizeof (charset.fast_map)); |
3263d5a2 KH |
1010 | |
1011 | if (! NILP (args[charset_arg_code_offset])) | |
1012 | { | |
1013 | val = args[charset_arg_code_offset]; | |
3c7649c1 | 1014 | CHECK_CHARACTER (val); |
3263d5a2 KH |
1015 | |
1016 | charset.method = CHARSET_METHOD_OFFSET; | |
1017 | charset.code_offset = XINT (val); | |
1018 | ||
3263d5a2 | 1019 | i = CODE_POINT_TO_INDEX (&charset, charset.max_code); |
3c7649c1 | 1020 | if (MAX_CHAR - charset.code_offset < i) |
3263d5a2 | 1021 | error ("Unsupported max char: %d", charset.max_char); |
3c7649c1 PE |
1022 | charset.max_char = i + charset.code_offset; |
1023 | i = CODE_POINT_TO_INDEX (&charset, charset.min_code); | |
1024 | charset.min_char = i + charset.code_offset; | |
3263d5a2 | 1025 | |
f148205f KH |
1026 | i = (charset.min_char >> 7) << 7; |
1027 | for (; i < 0x10000 && i <= charset.max_char; i += 128) | |
3263d5a2 | 1028 | CHARSET_FAST_MAP_SET (i, charset.fast_map); |
f148205f | 1029 | i = (i >> 12) << 12; |
3263d5a2 KH |
1030 | for (; i <= charset.max_char; i += 0x1000) |
1031 | CHARSET_FAST_MAP_SET (i, charset.fast_map); | |
3620330b KH |
1032 | if (charset.code_offset == 0 && charset.max_char >= 0x80) |
1033 | charset.ascii_compatible_p = 1; | |
3263d5a2 KH |
1034 | } |
1035 | else if (! NILP (args[charset_arg_map])) | |
1036 | { | |
1037 | val = args[charset_arg_map]; | |
1038 | ASET (attrs, charset_map, val); | |
ecca2aad | 1039 | charset.method = CHARSET_METHOD_MAP; |
3263d5a2 | 1040 | } |
374c5cfd | 1041 | else if (! NILP (args[charset_arg_subset])) |
3263d5a2 | 1042 | { |
374c5cfd KH |
1043 | Lisp_Object parent; |
1044 | Lisp_Object parent_min_code, parent_max_code, parent_code_offset; | |
1045 | struct charset *parent_charset; | |
1046 | ||
1047 | val = args[charset_arg_subset]; | |
1048 | parent = Fcar (val); | |
1049 | CHECK_CHARSET_GET_CHARSET (parent, parent_charset); | |
1050 | parent_min_code = Fnth (make_number (1), val); | |
1051 | CHECK_NATNUM (parent_min_code); | |
1052 | parent_max_code = Fnth (make_number (2), val); | |
1053 | CHECK_NATNUM (parent_max_code); | |
1054 | parent_code_offset = Fnth (make_number (3), val); | |
1055 | CHECK_NUMBER (parent_code_offset); | |
25721f5b | 1056 | val = make_uninit_vector (4); |
374c5cfd KH |
1057 | ASET (val, 0, make_number (parent_charset->id)); |
1058 | ASET (val, 1, parent_min_code); | |
1059 | ASET (val, 2, parent_max_code); | |
1060 | ASET (val, 3, parent_code_offset); | |
1061 | ASET (attrs, charset_subset, val); | |
1062 | ||
1063 | charset.method = CHARSET_METHOD_SUBSET; | |
1064 | /* Here, we just copy the parent's fast_map. It's not accurate, | |
1065 | but at least it works for quickly detecting which character | |
1066 | DOESN'T belong to this charset. */ | |
1067 | for (i = 0; i < 190; i++) | |
1068 | charset.fast_map[i] = parent_charset->fast_map[i]; | |
1069 | ||
1070 | /* We also copy these for parents. */ | |
1071 | charset.min_char = parent_charset->min_char; | |
1072 | charset.max_char = parent_charset->max_char; | |
1073 | } | |
1074 | else if (! NILP (args[charset_arg_superset])) | |
0282eb69 | 1075 | { |
374c5cfd KH |
1076 | val = args[charset_arg_superset]; |
1077 | charset.method = CHARSET_METHOD_SUPERSET; | |
3263d5a2 | 1078 | val = Fcopy_sequence (val); |
374c5cfd | 1079 | ASET (attrs, charset_superset, val); |
3263d5a2 KH |
1080 | |
1081 | charset.min_char = MAX_CHAR; | |
1082 | charset.max_char = 0; | |
1083 | for (; ! NILP (val); val = Fcdr (val)) | |
0282eb69 | 1084 | { |
3263d5a2 KH |
1085 | Lisp_Object elt, car_part, cdr_part; |
1086 | int this_id, offset; | |
1087 | struct charset *this_charset; | |
2e344af3 | 1088 | |
3263d5a2 KH |
1089 | elt = Fcar (val); |
1090 | if (CONSP (elt)) | |
2e344af3 | 1091 | { |
3263d5a2 KH |
1092 | car_part = XCAR (elt); |
1093 | cdr_part = XCDR (elt); | |
1094 | CHECK_CHARSET_GET_ID (car_part, this_id); | |
d311d28c | 1095 | CHECK_TYPE_RANGED_INTEGER (int, cdr_part); |
3263d5a2 | 1096 | offset = XINT (cdr_part); |
177c0ea7 | 1097 | } |
3263d5a2 | 1098 | else |
4ed46869 | 1099 | { |
3263d5a2 KH |
1100 | CHECK_CHARSET_GET_ID (elt, this_id); |
1101 | offset = 0; | |
4ed46869 | 1102 | } |
3263d5a2 KH |
1103 | XSETCAR (val, Fcons (make_number (this_id), make_number (offset))); |
1104 | ||
1105 | this_charset = CHARSET_FROM_ID (this_id); | |
1106 | if (charset.min_char > this_charset->min_char) | |
1107 | charset.min_char = this_charset->min_char; | |
1108 | if (charset.max_char < this_charset->max_char) | |
1109 | charset.max_char = this_charset->max_char; | |
1110 | for (i = 0; i < 190; i++) | |
1111 | charset.fast_map[i] |= this_charset->fast_map[i]; | |
0282eb69 | 1112 | } |
0282eb69 | 1113 | } |
2e344af3 | 1114 | else |
3263d5a2 | 1115 | error ("None of :code-offset, :map, :parents are specified"); |
05505664 | 1116 | |
3263d5a2 KH |
1117 | val = args[charset_arg_unify_map]; |
1118 | if (! NILP (val) && !STRINGP (val)) | |
1119 | CHECK_VECTOR (val); | |
1120 | ASET (attrs, charset_unify_map, val); | |
05505664 | 1121 | |
3263d5a2 KH |
1122 | CHECK_LIST (args[charset_arg_plist]); |
1123 | ASET (attrs, charset_plist, args[charset_arg_plist]); | |
4ed46869 | 1124 | |
3263d5a2 KH |
1125 | charset.hash_index = hash_lookup (hash_table, args[charset_arg_name], |
1126 | &hash_code); | |
1127 | if (charset.hash_index >= 0) | |
1128 | { | |
1129 | new_definition_p = 0; | |
4f65af01 | 1130 | id = XFASTINT (CHARSET_SYMBOL_ID (args[charset_arg_name])); |
e83064be | 1131 | set_hash_value_slot (hash_table, charset.hash_index, attrs); |
3263d5a2 | 1132 | } |
1a45ff10 | 1133 | else |
3263d5a2 KH |
1134 | { |
1135 | charset.hash_index = hash_put (hash_table, args[charset_arg_name], attrs, | |
1136 | hash_code); | |
1137 | if (charset_table_used == charset_table_size) | |
1138 | { | |
e097a6fa | 1139 | /* Ensure that charset IDs fit into 'int' as well as into the |
0065d054 PE |
1140 | restriction imposed by fixnums. Although the 'int' restriction |
1141 | could be removed, too much other code would need altering; for | |
1142 | example, the IDs are stuffed into struct | |
1143 | coding_system.charbuf[i] entries, which are 'int'. */ | |
1144 | int old_size = charset_table_size; | |
2dd2e622 | 1145 | ptrdiff_t new_size = old_size; |
0065d054 | 1146 | struct charset *new_table = |
2dd2e622 | 1147 | xpalloc (0, &new_size, 1, |
0065d054 PE |
1148 | min (INT_MAX, MOST_POSITIVE_FIXNUM), |
1149 | sizeof *charset_table); | |
1150 | memcpy (new_table, charset_table, old_size * sizeof *new_table); | |
2fe1edd1 | 1151 | charset_table = new_table; |
2dd2e622 | 1152 | charset_table_size = new_size; |
f701dc2a PE |
1153 | /* FIXME: This leaks memory, as the old charset_table becomes |
1154 | unreachable. If the old charset table is charset_table_init | |
1155 | then this leak is intentional; otherwise, it's unclear. | |
1156 | If the latter memory leak is intentional, a | |
0065d054 PE |
1157 | comment should be added to explain this. If not, the old |
1158 | charset_table should be freed, by passing it as the 1st argument | |
1159 | to xpalloc and removing the memcpy. */ | |
3263d5a2 KH |
1160 | } |
1161 | id = charset_table_used++; | |
3263d5a2 KH |
1162 | new_definition_p = 1; |
1163 | } | |
2e344af3 | 1164 | |
4f65af01 | 1165 | ASET (attrs, charset_id, make_number (id)); |
3263d5a2 KH |
1166 | charset.id = id; |
1167 | charset_table[id] = charset; | |
2e344af3 | 1168 | |
ecca2aad | 1169 | if (charset.method == CHARSET_METHOD_MAP) |
b8ebe9dd KH |
1170 | { |
1171 | load_charset (&charset, 0); | |
1172 | charset_table[id] = charset; | |
1173 | } | |
ecca2aad | 1174 | |
3263d5a2 | 1175 | if (charset.iso_final >= 0) |
4ed46869 | 1176 | { |
3263d5a2 KH |
1177 | ISO_CHARSET_TABLE (charset.dimension, charset.iso_chars_96, |
1178 | charset.iso_final) = id; | |
1179 | if (new_definition_p) | |
1180 | Viso_2022_charset_list = nconc2 (Viso_2022_charset_list, | |
1181 | Fcons (make_number (id), Qnil)); | |
7c7dceee KH |
1182 | if (ISO_CHARSET_TABLE (1, 0, 'J') == id) |
1183 | charset_jisx0201_roman = id; | |
1184 | else if (ISO_CHARSET_TABLE (2, 0, '@') == id) | |
1185 | charset_jisx0208_1978 = id; | |
1186 | else if (ISO_CHARSET_TABLE (2, 0, 'B') == id) | |
1187 | charset_jisx0208 = id; | |
d32320c4 KH |
1188 | else if (ISO_CHARSET_TABLE (2, 0, 'C') == id) |
1189 | charset_ksc5601 = id; | |
4ed46869 | 1190 | } |
d0cf2d48 | 1191 | |
3263d5a2 | 1192 | if (charset.emacs_mule_id >= 0) |
4ed46869 | 1193 | { |
b84ae584 | 1194 | emacs_mule_charset[charset.emacs_mule_id] = id; |
4f65af01 KH |
1195 | if (charset.emacs_mule_id < 0xA0) |
1196 | emacs_mule_bytes[charset.emacs_mule_id] = charset.dimension + 1; | |
3b1ae89b KH |
1197 | else |
1198 | emacs_mule_bytes[charset.emacs_mule_id] = charset.dimension + 2; | |
3263d5a2 KH |
1199 | if (new_definition_p) |
1200 | Vemacs_mule_charset_list = nconc2 (Vemacs_mule_charset_list, | |
1201 | Fcons (make_number (id), Qnil)); | |
4ed46869 KH |
1202 | } |
1203 | ||
3263d5a2 KH |
1204 | if (new_definition_p) |
1205 | { | |
1206 | Vcharset_list = Fcons (args[charset_arg_name], Vcharset_list); | |
8055c66a KH |
1207 | if (charset.supplementary_p) |
1208 | Vcharset_ordered_list = nconc2 (Vcharset_ordered_list, | |
1209 | Fcons (make_number (id), Qnil)); | |
1210 | else | |
880820fe KH |
1211 | { |
1212 | Lisp_Object tail; | |
1213 | ||
1214 | for (tail = Vcharset_ordered_list; CONSP (tail); tail = XCDR (tail)) | |
1215 | { | |
1216 | struct charset *cs = CHARSET_FROM_ID (XINT (XCAR (tail))); | |
1217 | ||
1218 | if (cs->supplementary_p) | |
1219 | break; | |
1220 | } | |
1221 | if (EQ (tail, Vcharset_ordered_list)) | |
1222 | Vcharset_ordered_list = Fcons (make_number (id), | |
1223 | Vcharset_ordered_list); | |
1224 | else if (NILP (tail)) | |
1225 | Vcharset_ordered_list = nconc2 (Vcharset_ordered_list, | |
1226 | Fcons (make_number (id), Qnil)); | |
1227 | else | |
1228 | { | |
1229 | val = Fcons (XCAR (tail), XCDR (tail)); | |
1230 | XSETCDR (tail, val); | |
1231 | XSETCAR (tail, make_number (id)); | |
1232 | } | |
1233 | } | |
dbbb237d | 1234 | charset_ordered_list_tick++; |
3263d5a2 | 1235 | } |
4ed46869 | 1236 | |
3263d5a2 | 1237 | return Qnil; |
4ed46869 KH |
1238 | } |
1239 | ||
2fe1edd1 KH |
1240 | |
1241 | /* Same as Fdefine_charset_internal but arguments are more convenient | |
1242 | to call from C (typically in syms_of_charset). This can define a | |
1243 | charset of `offset' method only. Return the ID of the new | |
1244 | charset. */ | |
1245 | ||
1246 | static int | |
dd4c5104 DN |
1247 | define_charset_internal (Lisp_Object name, |
1248 | int dimension, | |
dfb6afda | 1249 | const char *code_space_chars, |
dd4c5104 DN |
1250 | unsigned min_code, unsigned max_code, |
1251 | int iso_final, int iso_revision, int emacs_mule_id, | |
d5172d4f | 1252 | bool ascii_compatible, bool supplementary, |
dd4c5104 | 1253 | int code_offset) |
2fe1edd1 | 1254 | { |
dfb6afda | 1255 | const unsigned char *code_space = (const unsigned char *) code_space_chars; |
2fe1edd1 | 1256 | Lisp_Object args[charset_arg_max]; |
2fe1edd1 KH |
1257 | Lisp_Object val; |
1258 | int i; | |
1259 | ||
1260 | args[charset_arg_name] = name; | |
1261 | args[charset_arg_dimension] = make_number (dimension); | |
25721f5b | 1262 | val = make_uninit_vector (8); |
2fe1edd1 KH |
1263 | for (i = 0; i < 8; i++) |
1264 | ASET (val, i, make_number (code_space[i])); | |
1265 | args[charset_arg_code_space] = val; | |
1266 | args[charset_arg_min_code] = make_number (min_code); | |
1267 | args[charset_arg_max_code] = make_number (max_code); | |
1268 | args[charset_arg_iso_final] | |
1269 | = (iso_final < 0 ? Qnil : make_number (iso_final)); | |
1270 | args[charset_arg_iso_revision] = make_number (iso_revision); | |
1271 | args[charset_arg_emacs_mule_id] | |
1272 | = (emacs_mule_id < 0 ? Qnil : make_number (emacs_mule_id)); | |
1273 | args[charset_arg_ascii_compatible_p] = ascii_compatible ? Qt : Qnil; | |
7acf89e6 | 1274 | args[charset_arg_supplementary_p] = supplementary ? Qt : Qnil; |
2fe1edd1 KH |
1275 | args[charset_arg_invalid_code] = Qnil; |
1276 | args[charset_arg_code_offset] = make_number (code_offset); | |
1277 | args[charset_arg_map] = Qnil; | |
1278 | args[charset_arg_subset] = Qnil; | |
1279 | args[charset_arg_superset] = Qnil; | |
1280 | args[charset_arg_unify_map] = Qnil; | |
1281 | ||
694b6c97 | 1282 | args[charset_arg_plist] = |
3438fe21 | 1283 | listn (CONSTYPE_HEAP, 14, |
694b6c97 DA |
1284 | intern_c_string (":name"), |
1285 | args[charset_arg_name], | |
1286 | intern_c_string (":dimension"), | |
1287 | args[charset_arg_dimension], | |
1288 | intern_c_string (":code-space"), | |
1289 | args[charset_arg_code_space], | |
1290 | intern_c_string (":iso-final-char"), | |
1291 | args[charset_arg_iso_final], | |
1292 | intern_c_string (":emacs-mule-id"), | |
1293 | args[charset_arg_emacs_mule_id], | |
1294 | intern_c_string (":ascii-compatible-p"), | |
1295 | args[charset_arg_ascii_compatible_p], | |
1296 | intern_c_string (":code-offset"), | |
1297 | args[charset_arg_code_offset]); | |
2fe1edd1 KH |
1298 | Fdefine_charset_internal (charset_arg_max, args); |
1299 | ||
1300 | return XINT (CHARSET_SYMBOL_ID (name)); | |
1301 | } | |
1302 | ||
1303 | ||
3263d5a2 KH |
1304 | DEFUN ("define-charset-alias", Fdefine_charset_alias, |
1305 | Sdefine_charset_alias, 2, 2, 0, | |
1306 | doc: /* Define ALIAS as an alias for charset CHARSET. */) | |
5842a27b | 1307 | (Lisp_Object alias, Lisp_Object charset) |
4ed46869 | 1308 | { |
3263d5a2 KH |
1309 | Lisp_Object attr; |
1310 | ||
1311 | CHECK_CHARSET_GET_ATTR (charset, attr); | |
1312 | Fputhash (alias, attr, Vcharset_hash_table); | |
528623a0 | 1313 | Vcharset_list = Fcons (alias, Vcharset_list); |
3263d5a2 KH |
1314 | return Qnil; |
1315 | } | |
4ed46869 | 1316 | |
4ed46869 | 1317 | |
3263d5a2 | 1318 | DEFUN ("charset-plist", Fcharset_plist, Scharset_plist, 1, 1, 0, |
56a46d1d | 1319 | doc: /* Return the property list of CHARSET. */) |
5842a27b | 1320 | (Lisp_Object charset) |
3263d5a2 KH |
1321 | { |
1322 | Lisp_Object attrs; | |
1323 | ||
1324 | CHECK_CHARSET_GET_ATTR (charset, attrs); | |
1325 | return CHARSET_ATTR_PLIST (attrs); | |
1326 | } | |
1327 | ||
1328 | ||
1329 | DEFUN ("set-charset-plist", Fset_charset_plist, Sset_charset_plist, 2, 2, 0, | |
1330 | doc: /* Set CHARSET's property list to PLIST. */) | |
5842a27b | 1331 | (Lisp_Object charset, Lisp_Object plist) |
3263d5a2 KH |
1332 | { |
1333 | Lisp_Object attrs; | |
1334 | ||
1335 | CHECK_CHARSET_GET_ATTR (charset, attrs); | |
4939150c | 1336 | ASET (attrs, charset_plist, plist); |
3263d5a2 KH |
1337 | return plist; |
1338 | } | |
1339 | ||
1340 | ||
dbbb237d | 1341 | DEFUN ("unify-charset", Funify_charset, Sunify_charset, 1, 3, 0, |
56a46d1d DL |
1342 | doc: /* Unify characters of CHARSET with Unicode. |
1343 | This means reading the relevant file and installing the table defined | |
dbbb237d KH |
1344 | by CHARSET's `:unify-map' property. |
1345 | ||
64165ae2 DL |
1346 | Optional second arg UNIFY-MAP is a file name string or a vector. It has |
1347 | the same meaning as the `:unify-map' attribute in the function | |
dbbb237d KH |
1348 | `define-charset' (which see). |
1349 | ||
1350 | Optional third argument DEUNIFY, if non-nil, means to de-unify CHARSET. */) | |
5842a27b | 1351 | (Lisp_Object charset, Lisp_Object unify_map, Lisp_Object deunify) |
8a73a704 | 1352 | { |
3263d5a2 KH |
1353 | int id; |
1354 | struct charset *cs; | |
8f924df7 | 1355 | |
3263d5a2 KH |
1356 | CHECK_CHARSET_GET_ID (charset, id); |
1357 | cs = CHARSET_FROM_ID (id); | |
dbbb237d KH |
1358 | if (NILP (deunify) |
1359 | ? CHARSET_UNIFIED_P (cs) && ! NILP (CHARSET_DEUNIFIER (cs)) | |
1360 | : ! CHARSET_UNIFIED_P (cs)) | |
3263d5a2 | 1361 | return Qnil; |
dbbb237d | 1362 | |
3263d5a2 | 1363 | CHARSET_UNIFIED_P (cs) = 0; |
dbbb237d KH |
1364 | if (NILP (deunify)) |
1365 | { | |
ecca2aad KH |
1366 | if (CHARSET_METHOD (cs) != CHARSET_METHOD_OFFSET |
1367 | || CHARSET_CODE_OFFSET (cs) < 0x110000) | |
8f924df7 | 1368 | error ("Can't unify charset: %s", SDATA (SYMBOL_NAME (charset))); |
dbbb237d KH |
1369 | if (NILP (unify_map)) |
1370 | unify_map = CHARSET_UNIFY_MAP (cs); | |
dbbb237d | 1371 | else |
ecca2aad KH |
1372 | { |
1373 | if (! STRINGP (unify_map) && ! VECTORP (unify_map)) | |
1374 | signal_error ("Bad unify-map", unify_map); | |
4939150c | 1375 | set_charset_attr (cs, charset_unify_map, unify_map); |
ecca2aad KH |
1376 | } |
1377 | if (NILP (Vchar_unify_table)) | |
1378 | Vchar_unify_table = Fmake_char_table (Qnil, Qnil); | |
1379 | char_table_set_range (Vchar_unify_table, | |
1380 | cs->min_char, cs->max_char, charset); | |
dbbb237d KH |
1381 | CHARSET_UNIFIED_P (cs) = 1; |
1382 | } | |
1383 | else if (CHAR_TABLE_P (Vchar_unify_table)) | |
1384 | { | |
3c7649c1 PE |
1385 | unsigned min_code = CHARSET_MIN_CODE (cs); |
1386 | unsigned max_code = CHARSET_MAX_CODE (cs); | |
dbbb237d KH |
1387 | int min_char = DECODE_CHAR (cs, min_code); |
1388 | int max_char = DECODE_CHAR (cs, max_code); | |
8f924df7 | 1389 | |
dbbb237d KH |
1390 | char_table_set_range (Vchar_unify_table, min_char, max_char, Qnil); |
1391 | } | |
8f924df7 | 1392 | |
3263d5a2 | 1393 | return Qnil; |
8a73a704 KH |
1394 | } |
1395 | ||
3fac5a51 KH |
1396 | DEFUN ("get-unused-iso-final-char", Fget_unused_iso_final_char, |
1397 | Sget_unused_iso_final_char, 2, 2, 0, | |
3263d5a2 | 1398 | doc: /* |
d0cf2d48 | 1399 | Return an unused ISO final char for a charset of DIMENSION and CHARS. |
fdb82f93 PJ |
1400 | DIMENSION is the number of bytes to represent a character: 1 or 2. |
1401 | CHARS is the number of characters in a dimension: 94 or 96. | |
1402 | ||
1403 | This final char is for private use, thus the range is `0' (48) .. `?' (63). | |
1721b6af | 1404 | If there's no unused final char for the specified kind of charset, |
fdb82f93 | 1405 | return nil. */) |
5842a27b | 1406 | (Lisp_Object dimension, Lisp_Object chars) |
3fac5a51 KH |
1407 | { |
1408 | int final_char; | |
1409 | ||
b7826503 PJ |
1410 | CHECK_NUMBER (dimension); |
1411 | CHECK_NUMBER (chars); | |
3263d5a2 KH |
1412 | if (XINT (dimension) != 1 && XINT (dimension) != 2 && XINT (dimension) != 3) |
1413 | args_out_of_range_3 (dimension, make_number (1), make_number (3)); | |
3fac5a51 | 1414 | if (XINT (chars) != 94 && XINT (chars) != 96) |
3263d5a2 | 1415 | args_out_of_range_3 (chars, make_number (94), make_number (96)); |
3fac5a51 | 1416 | for (final_char = '0'; final_char <= '?'; final_char++) |
3263d5a2 KH |
1417 | if (ISO_CHARSET_TABLE (XINT (dimension), XINT (chars), final_char) < 0) |
1418 | break; | |
3fac5a51 KH |
1419 | return (final_char <= '?' ? make_number (final_char) : Qnil); |
1420 | } | |
1421 | ||
3263d5a2 | 1422 | static void |
971de7fb | 1423 | check_iso_charset_parameter (Lisp_Object dimension, Lisp_Object chars, Lisp_Object final_char) |
4ed46869 | 1424 | { |
3263d5a2 KH |
1425 | CHECK_NATNUM (dimension); |
1426 | CHECK_NATNUM (chars); | |
e6c3da20 | 1427 | CHECK_CHARACTER (final_char); |
4ed46869 | 1428 | |
3263d5a2 | 1429 | if (XINT (dimension) > 3) |
c2982e87 | 1430 | error ("Invalid DIMENSION %"pI"d, it should be 1, 2, or 3", |
5fdb398c | 1431 | XINT (dimension)); |
4ed46869 | 1432 | if (XINT (chars) != 94 && XINT (chars) != 96) |
c2982e87 | 1433 | error ("Invalid CHARS %"pI"d, it should be 94 or 96", XINT (chars)); |
3263d5a2 | 1434 | if (XINT (final_char) < '0' || XINT (final_char) > '~') |
e6c3da20 EZ |
1435 | error ("Invalid FINAL-CHAR %c, it should be `0'..`~'", |
1436 | (int)XINT (final_char)); | |
3263d5a2 KH |
1437 | } |
1438 | ||
1439 | ||
1440 | DEFUN ("declare-equiv-charset", Fdeclare_equiv_charset, Sdeclare_equiv_charset, | |
1441 | 4, 4, 0, | |
cefd8c4f KH |
1442 | doc: /* Declare an equivalent charset for ISO-2022 decoding. |
1443 | ||
1444 | On decoding by an ISO-2022 base coding system, when a charset | |
1445 | specified by DIMENSION, CHARS, and FINAL-CHAR is designated, behave as | |
1446 | if CHARSET is designated instead. */) | |
5842a27b | 1447 | (Lisp_Object dimension, Lisp_Object chars, Lisp_Object final_char, Lisp_Object charset) |
3263d5a2 KH |
1448 | { |
1449 | int id; | |
d5172d4f | 1450 | bool chars_flag; |
4ed46869 | 1451 | |
3263d5a2 KH |
1452 | CHECK_CHARSET_GET_ID (charset, id); |
1453 | check_iso_charset_parameter (dimension, chars, final_char); | |
82215ce9 KH |
1454 | chars_flag = XINT (chars) == 96; |
1455 | ISO_CHARSET_TABLE (XINT (dimension), chars_flag, XINT (final_char)) = id; | |
4ed46869 KH |
1456 | return Qnil; |
1457 | } | |
1458 | ||
3263d5a2 | 1459 | |
2e344af3 KH |
1460 | /* Return information about charsets in the text at PTR of NBYTES |
1461 | bytes, which are NCHARS characters. The value is: | |
f6302ac9 | 1462 | |
cfe34140 | 1463 | 0: Each character is represented by one byte. This is always |
3263d5a2 KH |
1464 | true for a unibyte string. For a multibyte string, true if |
1465 | it contains only ASCII characters. | |
1466 | ||
28c026cd DL |
1467 | 1: No charsets other than ascii, control-1, and latin-1 are |
1468 | found. | |
1d67c29b | 1469 | |
3263d5a2 KH |
1470 | 2: Otherwise. |
1471 | */ | |
4ed46869 KH |
1472 | |
1473 | int | |
971de7fb | 1474 | string_xstring_p (Lisp_Object string) |
4ed46869 | 1475 | { |
8f924df7 KH |
1476 | const unsigned char *p = SDATA (string); |
1477 | const unsigned char *endp = p + SBYTES (string); | |
3263d5a2 | 1478 | |
8f924df7 | 1479 | if (SCHARS (string) == SBYTES (string)) |
3263d5a2 KH |
1480 | return 0; |
1481 | ||
3263d5a2 | 1482 | while (p < endp) |
0282eb69 | 1483 | { |
3263d5a2 | 1484 | int c = STRING_CHAR_ADVANCE (p); |
2e344af3 | 1485 | |
3cc67a4d | 1486 | if (c >= 0x100) |
3263d5a2 | 1487 | return 2; |
0282eb69 | 1488 | } |
3263d5a2 KH |
1489 | return 1; |
1490 | } | |
05505664 | 1491 | |
05505664 | 1492 | |
3263d5a2 | 1493 | /* Find charsets in the string at PTR of NCHARS and NBYTES. |
4ed46869 | 1494 | |
3cc67a4d KH |
1495 | CHARSETS is a vector. If Nth element is non-nil, it means the |
1496 | charset whose id is N is already found. | |
2e344af3 | 1497 | |
3263d5a2 | 1498 | It may lookup a translation table TABLE if supplied. */ |
2e344af3 | 1499 | |
3263d5a2 | 1500 | static void |
d5172d4f PE |
1501 | find_charsets_in_text (const unsigned char *ptr, ptrdiff_t nchars, |
1502 | ptrdiff_t nbytes, Lisp_Object charsets, | |
1503 | Lisp_Object table, bool multibyte) | |
3263d5a2 | 1504 | { |
dbbb237d | 1505 | const unsigned char *pend = ptr + nbytes; |
3263d5a2 KH |
1506 | |
1507 | if (nchars == nbytes) | |
3263d5a2 | 1508 | { |
3cc67a4d KH |
1509 | if (multibyte) |
1510 | ASET (charsets, charset_ascii, Qt); | |
1511 | else | |
1512 | while (ptr < pend) | |
1513 | { | |
1514 | int c = *ptr++; | |
1515 | ||
1516 | if (!NILP (table)) | |
1517 | c = translate_char (table, c); | |
1518 | if (ASCII_BYTE_P (c)) | |
1519 | ASET (charsets, charset_ascii, Qt); | |
1520 | else | |
1521 | ASET (charsets, charset_eight_bit, Qt); | |
1522 | } | |
1523 | } | |
1524 | else | |
1525 | { | |
1526 | while (ptr < pend) | |
3263d5a2 | 1527 | { |
3cc67a4d KH |
1528 | int c = STRING_CHAR_ADVANCE (ptr); |
1529 | struct charset *charset; | |
3263d5a2 | 1530 | |
3cc67a4d KH |
1531 | if (!NILP (table)) |
1532 | c = translate_char (table, c); | |
1533 | charset = CHAR_CHARSET (c); | |
1534 | ASET (charsets, CHARSET_ID (charset), Qt); | |
4ed46869 | 1535 | } |
4ed46869 | 1536 | } |
4ed46869 KH |
1537 | } |
1538 | ||
1539 | DEFUN ("find-charset-region", Ffind_charset_region, Sfind_charset_region, | |
23d2a7f1 | 1540 | 2, 3, 0, |
fdb82f93 PJ |
1541 | doc: /* Return a list of charsets in the region between BEG and END. |
1542 | BEG and END are buffer positions. | |
1543 | Optional arg TABLE if non-nil is a translation table to look up. | |
1544 | ||
fdb82f93 PJ |
1545 | If the current buffer is unibyte, the returned list may contain |
1546 | only `ascii', `eight-bit-control', and `eight-bit-graphic'. */) | |
5842a27b | 1547 | (Lisp_Object beg, Lisp_Object end, Lisp_Object table) |
4ed46869 | 1548 | { |
3263d5a2 | 1549 | Lisp_Object charsets; |
d311d28c | 1550 | ptrdiff_t from, from_byte, to, stop, stop_byte; |
42ca828e | 1551 | int i; |
4ed46869 | 1552 | Lisp_Object val; |
d5172d4f | 1553 | bool multibyte = ! NILP (BVAR (current_buffer, enable_multibyte_characters)); |
4ed46869 KH |
1554 | |
1555 | validate_region (&beg, &end); | |
1556 | from = XFASTINT (beg); | |
1557 | stop = to = XFASTINT (end); | |
6ae1f27e | 1558 | |
4ed46869 | 1559 | if (from < GPT && GPT < to) |
6ae1f27e RS |
1560 | { |
1561 | stop = GPT; | |
1562 | stop_byte = GPT_BYTE; | |
1563 | } | |
1564 | else | |
1565 | stop_byte = CHAR_TO_BYTE (stop); | |
1566 | ||
1567 | from_byte = CHAR_TO_BYTE (from); | |
1568 | ||
3263d5a2 | 1569 | charsets = Fmake_vector (make_number (charset_table_used), Qnil); |
4ed46869 KH |
1570 | while (1) |
1571 | { | |
3263d5a2 | 1572 | find_charsets_in_text (BYTE_POS_ADDR (from_byte), stop - from, |
3cc67a4d KH |
1573 | stop_byte - from_byte, charsets, table, |
1574 | multibyte); | |
4ed46869 | 1575 | if (stop < to) |
6ae1f27e RS |
1576 | { |
1577 | from = stop, from_byte = stop_byte; | |
1578 | stop = to, stop_byte = CHAR_TO_BYTE (stop); | |
1579 | } | |
4ed46869 KH |
1580 | else |
1581 | break; | |
1582 | } | |
6ae1f27e | 1583 | |
4ed46869 | 1584 | val = Qnil; |
3263d5a2 | 1585 | for (i = charset_table_used - 1; i >= 0; i--) |
3cc67a4d | 1586 | if (!NILP (AREF (charsets, i))) |
3263d5a2 | 1587 | val = Fcons (CHARSET_NAME (charset_table + i), val); |
4ed46869 KH |
1588 | return val; |
1589 | } | |
1590 | ||
1591 | DEFUN ("find-charset-string", Ffind_charset_string, Sfind_charset_string, | |
23d2a7f1 | 1592 | 1, 2, 0, |
fdb82f93 PJ |
1593 | doc: /* Return a list of charsets in STR. |
1594 | Optional arg TABLE if non-nil is a translation table to look up. | |
1595 | ||
fdb82f93 | 1596 | If STR is unibyte, the returned list may contain |
3263d5a2 | 1597 | only `ascii', `eight-bit-control', and `eight-bit-graphic'. */) |
5842a27b | 1598 | (Lisp_Object str, Lisp_Object table) |
4ed46869 | 1599 | { |
3263d5a2 | 1600 | Lisp_Object charsets; |
4ed46869 KH |
1601 | int i; |
1602 | Lisp_Object val; | |
1603 | ||
b7826503 | 1604 | CHECK_STRING (str); |
87b089ad | 1605 | |
3263d5a2 | 1606 | charsets = Fmake_vector (make_number (charset_table_used), Qnil); |
8f924df7 | 1607 | find_charsets_in_text (SDATA (str), SCHARS (str), SBYTES (str), |
3cc67a4d KH |
1608 | charsets, table, |
1609 | STRING_MULTIBYTE (str)); | |
4ed46869 | 1610 | val = Qnil; |
3263d5a2 | 1611 | for (i = charset_table_used - 1; i >= 0; i--) |
3cc67a4d | 1612 | if (!NILP (AREF (charsets, i))) |
3263d5a2 | 1613 | val = Fcons (CHARSET_NAME (charset_table + i), val); |
4ed46869 KH |
1614 | return val; |
1615 | } | |
2e344af3 | 1616 | |
4ed46869 | 1617 | \f |
3263d5a2 | 1618 | |
ecca2aad KH |
1619 | /* Return a unified character code for C (>= 0x110000). VAL is a |
1620 | value of Vchar_unify_table for C; i.e. it is nil, an integer, or a | |
1621 | charset symbol. */ | |
41c8bfcf | 1622 | static int |
971de7fb | 1623 | maybe_unify_char (int c, Lisp_Object val) |
ecca2aad KH |
1624 | { |
1625 | struct charset *charset; | |
1626 | ||
1627 | if (INTEGERP (val)) | |
ccd9a01a | 1628 | return XFASTINT (val); |
ecca2aad KH |
1629 | if (NILP (val)) |
1630 | return c; | |
1631 | ||
1632 | CHECK_CHARSET_GET_CHARSET (val, charset); | |
291d430f | 1633 | #ifdef REL_ALLOC |
efc00ab1 | 1634 | /* The call to load_charset below can allocate memory, which screws |
291d430f EZ |
1635 | callers of this function through STRING_CHAR_* macros that hold C |
1636 | pointers to buffer text, if REL_ALLOC is used. */ | |
1637 | r_alloc_inhibit_buffer_relocation (1); | |
1638 | #endif | |
ecca2aad KH |
1639 | load_charset (charset, 1); |
1640 | if (! inhibit_load_charset_map) | |
1641 | { | |
1642 | val = CHAR_TABLE_REF (Vchar_unify_table, c); | |
1643 | if (! NILP (val)) | |
ccd9a01a | 1644 | c = XFASTINT (val); |
ecca2aad KH |
1645 | } |
1646 | else | |
1647 | { | |
1648 | int code_index = c - CHARSET_CODE_OFFSET (charset); | |
1649 | int unified = GET_TEMP_CHARSET_WORK_DECODER (code_index); | |
1650 | ||
1651 | if (unified > 0) | |
1652 | c = unified; | |
1653 | } | |
291d430f EZ |
1654 | #ifdef REL_ALLOC |
1655 | r_alloc_inhibit_buffer_relocation (0); | |
1656 | #endif | |
ecca2aad KH |
1657 | return c; |
1658 | } | |
1659 | ||
1660 | ||
bbd240ce | 1661 | /* Return a character corresponding to the code-point CODE of |
3263d5a2 KH |
1662 | CHARSET. */ |
1663 | ||
1664 | int | |
971de7fb | 1665 | decode_char (struct charset *charset, unsigned int code) |
4ed46869 | 1666 | { |
3263d5a2 KH |
1667 | int c, char_index; |
1668 | enum charset_method method = CHARSET_METHOD (charset); | |
ac4137cc | 1669 | |
3263d5a2 KH |
1670 | if (code < CHARSET_MIN_CODE (charset) || code > CHARSET_MAX_CODE (charset)) |
1671 | return -1; | |
4ed46869 | 1672 | |
374c5cfd | 1673 | if (method == CHARSET_METHOD_SUBSET) |
2e344af3 | 1674 | { |
374c5cfd KH |
1675 | Lisp_Object subset_info; |
1676 | ||
1677 | subset_info = CHARSET_SUBSET (charset); | |
1678 | charset = CHARSET_FROM_ID (XFASTINT (AREF (subset_info, 0))); | |
1679 | code -= XINT (AREF (subset_info, 3)); | |
1680 | if (code < XFASTINT (AREF (subset_info, 1)) | |
1681 | || code > XFASTINT (AREF (subset_info, 2))) | |
1682 | c = -1; | |
1683 | else | |
1684 | c = DECODE_CHAR (charset, code); | |
2e344af3 | 1685 | } |
374c5cfd | 1686 | else if (method == CHARSET_METHOD_SUPERSET) |
2e344af3 | 1687 | { |
3263d5a2 | 1688 | Lisp_Object parents; |
4ed46869 | 1689 | |
374c5cfd | 1690 | parents = CHARSET_SUPERSET (charset); |
3263d5a2 KH |
1691 | c = -1; |
1692 | for (; CONSP (parents); parents = XCDR (parents)) | |
1693 | { | |
1694 | int id = XINT (XCAR (XCAR (parents))); | |
1695 | int code_offset = XINT (XCDR (XCAR (parents))); | |
374c5cfd | 1696 | unsigned this_code = code - code_offset; |
4ed46869 | 1697 | |
3263d5a2 KH |
1698 | charset = CHARSET_FROM_ID (id); |
1699 | if ((c = DECODE_CHAR (charset, this_code)) >= 0) | |
1700 | break; | |
1701 | } | |
1702 | } | |
1703 | else | |
ac4137cc | 1704 | { |
3263d5a2 | 1705 | char_index = CODE_POINT_TO_INDEX (charset, code); |
69f8de5b KH |
1706 | if (char_index < 0) |
1707 | return -1; | |
4ed46869 | 1708 | |
3263d5a2 | 1709 | if (method == CHARSET_METHOD_MAP) |
ac4137cc | 1710 | { |
3263d5a2 | 1711 | Lisp_Object decoder; |
4ed46869 | 1712 | |
3263d5a2 KH |
1713 | decoder = CHARSET_DECODER (charset); |
1714 | if (! VECTORP (decoder)) | |
ecca2aad KH |
1715 | { |
1716 | load_charset (charset, 1); | |
1717 | decoder = CHARSET_DECODER (charset); | |
1718 | } | |
1719 | if (VECTORP (decoder)) | |
1720 | c = XINT (AREF (decoder, char_index)); | |
1721 | else | |
1722 | c = GET_TEMP_CHARSET_WORK_DECODER (char_index); | |
ac4137cc | 1723 | } |
ecca2aad | 1724 | else /* method == CHARSET_METHOD_OFFSET */ |
ac4137cc | 1725 | { |
3263d5a2 | 1726 | c = char_index + CHARSET_CODE_OFFSET (charset); |
ecca2aad | 1727 | if (CHARSET_UNIFIED_P (charset) |
41c8bfcf PE |
1728 | && MAX_UNICODE_CHAR < c && c <= MAX_5_BYTE_CHAR) |
1729 | { | |
1730 | /* Unify C with a Unicode character if possible. */ | |
1731 | Lisp_Object val = CHAR_TABLE_REF (Vchar_unify_table, c); | |
1732 | c = maybe_unify_char (c, val); | |
1733 | } | |
ac4137cc KH |
1734 | } |
1735 | } | |
4ed46869 | 1736 | |
3263d5a2 | 1737 | return c; |
90d7b74e KH |
1738 | } |
1739 | ||
374c5cfd KH |
1740 | /* Variable used temporarily by the macro ENCODE_CHAR. */ |
1741 | Lisp_Object charset_work; | |
4ed46869 | 1742 | |
d5172d4f | 1743 | /* Return a code-point of C in CHARSET. If C doesn't belong to |
28c026cd DL |
1744 | CHARSET, return CHARSET_INVALID_CODE (CHARSET). If STRICT is true, |
1745 | use CHARSET's strict_max_char instead of max_char. */ | |
4ed46869 | 1746 | |
3263d5a2 | 1747 | unsigned |
971de7fb | 1748 | encode_char (struct charset *charset, int c) |
9d3d8cba | 1749 | { |
3263d5a2 KH |
1750 | unsigned code; |
1751 | enum charset_method method = CHARSET_METHOD (charset); | |
9d3d8cba | 1752 | |
3263d5a2 | 1753 | if (CHARSET_UNIFIED_P (charset)) |
ac4137cc | 1754 | { |
6809ca75 | 1755 | Lisp_Object deunifier; |
ecca2aad | 1756 | int code_index = -1; |
4ed46869 | 1757 | |
3263d5a2 KH |
1758 | deunifier = CHARSET_DEUNIFIER (charset); |
1759 | if (! CHAR_TABLE_P (deunifier)) | |
ac4137cc | 1760 | { |
ecca2aad | 1761 | load_charset (charset, 2); |
3263d5a2 | 1762 | deunifier = CHARSET_DEUNIFIER (charset); |
ac4137cc | 1763 | } |
ecca2aad KH |
1764 | if (CHAR_TABLE_P (deunifier)) |
1765 | { | |
1766 | Lisp_Object deunified = CHAR_TABLE_REF (deunifier, c); | |
1767 | ||
1768 | if (INTEGERP (deunified)) | |
1769 | code_index = XINT (deunified); | |
1770 | } | |
1771 | else | |
1772 | { | |
1773 | code_index = GET_TEMP_CHARSET_WORK_ENCODER (c); | |
1774 | } | |
1775 | if (code_index >= 0) | |
1776 | c = CHARSET_CODE_OFFSET (charset) + code_index; | |
ac4137cc | 1777 | } |
9d3d8cba | 1778 | |
374c5cfd KH |
1779 | if (method == CHARSET_METHOD_SUBSET) |
1780 | { | |
1781 | Lisp_Object subset_info; | |
1782 | struct charset *this_charset; | |
1783 | ||
1784 | subset_info = CHARSET_SUBSET (charset); | |
1785 | this_charset = CHARSET_FROM_ID (XFASTINT (AREF (subset_info, 0))); | |
1786 | code = ENCODE_CHAR (this_charset, c); | |
1787 | if (code == CHARSET_INVALID_CODE (this_charset) | |
1788 | || code < XFASTINT (AREF (subset_info, 1)) | |
1789 | || code > XFASTINT (AREF (subset_info, 2))) | |
1790 | return CHARSET_INVALID_CODE (charset); | |
1791 | code += XINT (AREF (subset_info, 3)); | |
1792 | return code; | |
1793 | } | |
9d3d8cba | 1794 | |
374c5cfd | 1795 | if (method == CHARSET_METHOD_SUPERSET) |
859f2b3c | 1796 | { |
3263d5a2 | 1797 | Lisp_Object parents; |
d2665018 | 1798 | |
374c5cfd | 1799 | parents = CHARSET_SUPERSET (charset); |
3263d5a2 | 1800 | for (; CONSP (parents); parents = XCDR (parents)) |
beeedaad | 1801 | { |
3263d5a2 KH |
1802 | int id = XINT (XCAR (XCAR (parents))); |
1803 | int code_offset = XINT (XCDR (XCAR (parents))); | |
1804 | struct charset *this_charset = CHARSET_FROM_ID (id); | |
d2665018 | 1805 | |
3263d5a2 | 1806 | code = ENCODE_CHAR (this_charset, c); |
dbbb237d KH |
1807 | if (code != CHARSET_INVALID_CODE (this_charset)) |
1808 | return code + code_offset; | |
beeedaad | 1809 | } |
3263d5a2 KH |
1810 | return CHARSET_INVALID_CODE (charset); |
1811 | } | |
1bcc1567 | 1812 | |
15c85a88 KH |
1813 | if (! CHARSET_FAST_MAP_REF ((c), charset->fast_map) |
1814 | || c < CHARSET_MIN_CHAR (charset) || c > CHARSET_MAX_CHAR (charset)) | |
1815 | return CHARSET_INVALID_CODE (charset); | |
1bcc1567 | 1816 | |
3263d5a2 | 1817 | if (method == CHARSET_METHOD_MAP) |
3f62427c | 1818 | { |
3263d5a2 | 1819 | Lisp_Object encoder; |
beeedaad | 1820 | Lisp_Object val; |
9b6a601f | 1821 | |
3263d5a2 KH |
1822 | encoder = CHARSET_ENCODER (charset); |
1823 | if (! CHAR_TABLE_P (CHARSET_ENCODER (charset))) | |
b8ebe9dd KH |
1824 | { |
1825 | load_charset (charset, 2); | |
1826 | encoder = CHARSET_ENCODER (charset); | |
1827 | } | |
1828 | if (CHAR_TABLE_P (encoder)) | |
ecca2aad KH |
1829 | { |
1830 | val = CHAR_TABLE_REF (encoder, c); | |
1831 | if (NILP (val)) | |
1832 | return CHARSET_INVALID_CODE (charset); | |
1833 | code = XINT (val); | |
1834 | if (! CHARSET_COMPACT_CODES_P (charset)) | |
1835 | code = INDEX_TO_CODE_POINT (charset, code); | |
1836 | } | |
1837 | else | |
1838 | { | |
1839 | code = GET_TEMP_CHARSET_WORK_ENCODER (c); | |
51b59d79 | 1840 | code = INDEX_TO_CODE_POINT (charset, code); |
ecca2aad | 1841 | } |
3263d5a2 | 1842 | } |
820ee249 | 1843 | else /* method == CHARSET_METHOD_OFFSET */ |
beeedaad | 1844 | { |
58c8a77d PE |
1845 | unsigned code_index = c - CHARSET_CODE_OFFSET (charset); |
1846 | ||
1847 | code = INDEX_TO_CODE_POINT (charset, code_index); | |
3f62427c | 1848 | } |
8ac5a9cc | 1849 | |
3263d5a2 | 1850 | return code; |
4ed46869 KH |
1851 | } |
1852 | ||
4ed46869 | 1853 | |
3263d5a2 KH |
1854 | DEFUN ("decode-char", Fdecode_char, Sdecode_char, 2, 3, 0, |
1855 | doc: /* Decode the pair of CHARSET and CODE-POINT into a character. | |
1856 | Return nil if CODE-POINT is not valid in CHARSET. | |
4ed46869 | 1857 | |
3263d5a2 | 1858 | CODE-POINT may be a cons (HIGHER-16-BIT-VALUE . LOWER-16-BIT-VALUE). |
859f2b3c | 1859 | |
3263d5a2 | 1860 | Optional argument RESTRICTION specifies a way to map the pair of CCS |
d0cf2d48 | 1861 | and CODE-POINT to a character. Currently not supported and just ignored. */) |
5842a27b | 1862 | (Lisp_Object charset, Lisp_Object code_point, Lisp_Object restriction) |
4ed46869 | 1863 | { |
3263d5a2 KH |
1864 | int c, id; |
1865 | unsigned code; | |
1866 | struct charset *charsetp; | |
859f2b3c | 1867 | |
3263d5a2 | 1868 | CHECK_CHARSET_GET_ID (charset, id); |
be44ca6c | 1869 | code = cons_to_unsigned (code_point, UINT_MAX); |
3263d5a2 KH |
1870 | charsetp = CHARSET_FROM_ID (id); |
1871 | c = DECODE_CHAR (charsetp, code); | |
1872 | return (c >= 0 ? make_number (c) : Qnil); | |
4ed46869 KH |
1873 | } |
1874 | ||
859f2b3c | 1875 | |
3263d5a2 KH |
1876 | DEFUN ("encode-char", Fencode_char, Sencode_char, 2, 3, 0, |
1877 | doc: /* Encode the character CH into a code-point of CHARSET. | |
1878 | Return nil if CHARSET doesn't include CH. | |
17e7ef1b | 1879 | |
d0cf2d48 | 1880 | Optional argument RESTRICTION specifies a way to map CH to a |
3263d5a2 | 1881 | code-point in CCS. Currently not supported and just ignored. */) |
5842a27b | 1882 | (Lisp_Object ch, Lisp_Object charset, Lisp_Object restriction) |
4ed46869 | 1883 | { |
34206dd2 | 1884 | int c, id; |
3263d5a2 KH |
1885 | unsigned code; |
1886 | struct charset *charsetp; | |
046b1f03 | 1887 | |
3263d5a2 | 1888 | CHECK_CHARSET_GET_ID (charset, id); |
34206dd2 PE |
1889 | CHECK_CHARACTER (ch); |
1890 | c = XFASTINT (ch); | |
3263d5a2 | 1891 | charsetp = CHARSET_FROM_ID (id); |
34206dd2 | 1892 | code = ENCODE_CHAR (charsetp, c); |
3263d5a2 KH |
1893 | if (code == CHARSET_INVALID_CODE (charsetp)) |
1894 | return Qnil; | |
be44ca6c | 1895 | return INTEGER_TO_CONS (code); |
beeedaad KH |
1896 | } |
1897 | ||
beeedaad | 1898 | |
b121a744 KH |
1899 | DEFUN ("make-char", Fmake_char, Smake_char, 1, 5, 0, |
1900 | doc: | |
1901 | /* Return a character of CHARSET whose position codes are CODEn. | |
1902 | ||
1903 | CODE1 through CODE4 are optional, but if you don't supply sufficient | |
1904 | position codes, it is assumed that the minimum code in each dimension | |
04c2f2c5 | 1905 | is specified. */) |
5842a27b | 1906 | (Lisp_Object charset, Lisp_Object code1, Lisp_Object code2, Lisp_Object code3, Lisp_Object code4) |
beeedaad | 1907 | { |
3263d5a2 KH |
1908 | int id, dimension; |
1909 | struct charset *charsetp; | |
b121a744 KH |
1910 | unsigned code; |
1911 | int c; | |
87b089ad | 1912 | |
3263d5a2 KH |
1913 | CHECK_CHARSET_GET_ID (charset, id); |
1914 | charsetp = CHARSET_FROM_ID (id); | |
4ed46869 | 1915 | |
b121a744 KH |
1916 | dimension = CHARSET_DIMENSION (charsetp); |
1917 | if (NILP (code1)) | |
d47073ca KH |
1918 | code = (CHARSET_ASCII_COMPATIBLE_P (charsetp) |
1919 | ? 0 : CHARSET_MIN_CODE (charsetp)); | |
3263d5a2 | 1920 | else |
859f2b3c | 1921 | { |
b121a744 KH |
1922 | CHECK_NATNUM (code1); |
1923 | if (XFASTINT (code1) >= 0x100) | |
1924 | args_out_of_range (make_number (0xFF), code1); | |
1925 | code = XFASTINT (code1); | |
859f2b3c | 1926 | |
b0a1e45e | 1927 | if (dimension > 1) |
beeedaad | 1928 | { |
b121a744 | 1929 | code <<= 8; |
b0a1e45e KH |
1930 | if (NILP (code2)) |
1931 | code |= charsetp->code_space[(dimension - 2) * 4]; | |
beeedaad | 1932 | else |
b121a744 | 1933 | { |
b0a1e45e KH |
1934 | CHECK_NATNUM (code2); |
1935 | if (XFASTINT (code2) >= 0x100) | |
1936 | args_out_of_range (make_number (0xFF), code2); | |
1937 | code |= XFASTINT (code2); | |
b121a744 | 1938 | } |
99529c2c | 1939 | |
b0a1e45e | 1940 | if (dimension > 2) |
b121a744 KH |
1941 | { |
1942 | code <<= 8; | |
b0a1e45e KH |
1943 | if (NILP (code3)) |
1944 | code |= charsetp->code_space[(dimension - 3) * 4]; | |
b121a744 KH |
1945 | else |
1946 | { | |
b0a1e45e KH |
1947 | CHECK_NATNUM (code3); |
1948 | if (XFASTINT (code3) >= 0x100) | |
1949 | args_out_of_range (make_number (0xFF), code3); | |
1950 | code |= XFASTINT (code3); | |
1951 | } | |
1952 | ||
1953 | if (dimension > 3) | |
1954 | { | |
1955 | code <<= 8; | |
1956 | if (NILP (code4)) | |
1957 | code |= charsetp->code_space[0]; | |
1958 | else | |
1959 | { | |
1960 | CHECK_NATNUM (code4); | |
1961 | if (XFASTINT (code4) >= 0x100) | |
1962 | args_out_of_range (make_number (0xFF), code4); | |
1963 | code |= XFASTINT (code4); | |
1964 | } | |
b121a744 KH |
1965 | } |
1966 | } | |
beeedaad | 1967 | } |
859f2b3c | 1968 | } |
beeedaad | 1969 | |
b121a744 KH |
1970 | if (CHARSET_ISO_FINAL (charsetp) >= 0) |
1971 | code &= 0x7F7F7F7F; | |
1972 | c = DECODE_CHAR (charsetp, code); | |
1973 | if (c < 0) | |
1974 | error ("Invalid code(s)"); | |
3263d5a2 | 1975 | return make_number (c); |
4ed46869 KH |
1976 | } |
1977 | ||
beeedaad | 1978 | |
3263d5a2 KH |
1979 | /* Return the first charset in CHARSET_LIST that contains C. |
1980 | CHARSET_LIST is a list of charset IDs. If it is nil, use | |
1981 | Vcharset_ordered_list. */ | |
beeedaad | 1982 | |
3263d5a2 | 1983 | struct charset * |
971de7fb | 1984 | char_charset (int c, Lisp_Object charset_list, unsigned int *code_return) |
2e344af3 | 1985 | { |
d5172d4f | 1986 | bool maybe_null = 0; |
8a1816bb | 1987 | |
3263d5a2 KH |
1988 | if (NILP (charset_list)) |
1989 | charset_list = Vcharset_ordered_list; | |
8a1816bb KH |
1990 | else |
1991 | maybe_null = 1; | |
beeedaad | 1992 | |
6c652beb | 1993 | while (CONSP (charset_list)) |
2e344af3 | 1994 | { |
3263d5a2 KH |
1995 | struct charset *charset = CHARSET_FROM_ID (XINT (XCAR (charset_list))); |
1996 | unsigned code = ENCODE_CHAR (charset, c); | |
beeedaad | 1997 | |
3263d5a2 | 1998 | if (code != CHARSET_INVALID_CODE (charset)) |
beeedaad | 1999 | { |
3263d5a2 KH |
2000 | if (code_return) |
2001 | *code_return = code; | |
2002 | return charset; | |
3f62427c | 2003 | } |
3263d5a2 | 2004 | charset_list = XCDR (charset_list); |
c0be27fd KH |
2005 | if (! maybe_null |
2006 | && c <= MAX_UNICODE_CHAR | |
2007 | && EQ (charset_list, Vcharset_non_preferred_head)) | |
6c652beb | 2008 | return CHARSET_FROM_ID (charset_unicode); |
3f62427c | 2009 | } |
8a1816bb KH |
2010 | return (maybe_null ? NULL |
2011 | : c <= MAX_5_BYTE_CHAR ? CHARSET_FROM_ID (charset_emacs) | |
6a9c90ec | 2012 | : CHARSET_FROM_ID (charset_eight_bit)); |
3f62427c KH |
2013 | } |
2014 | ||
2e344af3 | 2015 | |
3263d5a2 | 2016 | DEFUN ("split-char", Fsplit_char, Ssplit_char, 1, 1, 0, |
3cc67a4d | 2017 | doc: |
d0cf2d48 | 2018 | /*Return list of charset and one to four position-codes of CH. |
3cc67a4d KH |
2019 | The charset is decided by the current priority order of charsets. |
2020 | A position-code is a byte value of each dimension of the code-point of | |
d0cf2d48 | 2021 | CH in the charset. */) |
5842a27b | 2022 | (Lisp_Object ch) |
4ed46869 | 2023 | { |
3263d5a2 KH |
2024 | struct charset *charset; |
2025 | int c, dimension; | |
2026 | unsigned code; | |
4ed46869 KH |
2027 | Lisp_Object val; |
2028 | ||
3263d5a2 KH |
2029 | CHECK_CHARACTER (ch); |
2030 | c = XFASTINT (ch); | |
2031 | charset = CHAR_CHARSET (c); | |
2032 | if (! charset) | |
1088b922 | 2033 | emacs_abort (); |
3263d5a2 KH |
2034 | code = ENCODE_CHAR (charset, c); |
2035 | if (code == CHARSET_INVALID_CODE (charset)) | |
1088b922 | 2036 | emacs_abort (); |
3263d5a2 | 2037 | dimension = CHARSET_DIMENSION (charset); |
3cc67a4d KH |
2038 | for (val = Qnil; dimension > 0; dimension--) |
2039 | { | |
2040 | val = Fcons (make_number (code & 0xFF), val); | |
2041 | code >>= 8; | |
2042 | } | |
3263d5a2 | 2043 | return Fcons (CHARSET_NAME (charset), val); |
4ed46869 KH |
2044 | } |
2045 | ||
740f080d | 2046 | |
4cb75c4b KH |
2047 | DEFUN ("char-charset", Fchar_charset, Schar_charset, 1, 2, 0, |
2048 | doc: /* Return the charset of highest priority that contains CH. | |
2049 | If optional 2nd arg RESTRICTION is non-nil, it is a list of charsets | |
2050 | from which to find the charset. It may also be a coding system. In | |
2051 | that case, find the charset from what supported by that coding system. */) | |
5842a27b | 2052 | (Lisp_Object ch, Lisp_Object restriction) |
4ed46869 | 2053 | { |
3263d5a2 | 2054 | struct charset *charset; |
4ed46869 | 2055 | |
3263d5a2 | 2056 | CHECK_CHARACTER (ch); |
4cb75c4b KH |
2057 | if (NILP (restriction)) |
2058 | charset = CHAR_CHARSET (XINT (ch)); | |
2059 | else | |
2060 | { | |
4cb75c4b KH |
2061 | if (CONSP (restriction)) |
2062 | { | |
7a84eee5 KH |
2063 | int c = XFASTINT (ch); |
2064 | ||
2065 | for (; CONSP (restriction); restriction = XCDR (restriction)) | |
4cb75c4b | 2066 | { |
f6095868 | 2067 | struct charset *rcharset; |
4cb75c4b | 2068 | |
f6095868 PE |
2069 | CHECK_CHARSET_GET_CHARSET (XCAR (restriction), rcharset); |
2070 | if (ENCODE_CHAR (rcharset, c) != CHARSET_INVALID_CODE (rcharset)) | |
7a84eee5 | 2071 | return XCAR (restriction); |
4cb75c4b | 2072 | } |
7a84eee5 | 2073 | return Qnil; |
4cb75c4b | 2074 | } |
7a84eee5 KH |
2075 | restriction = coding_system_charset_list (restriction); |
2076 | charset = char_charset (XINT (ch), restriction, NULL); | |
4cb75c4b KH |
2077 | if (! charset) |
2078 | return Qnil; | |
2079 | } | |
3263d5a2 | 2080 | return (CHARSET_NAME (charset)); |
4ed46869 KH |
2081 | } |
2082 | ||
17e7ef1b | 2083 | |
3263d5a2 KH |
2084 | DEFUN ("charset-after", Fcharset_after, Scharset_after, 0, 1, 0, |
2085 | doc: /* | |
2086 | Return charset of a character in the current buffer at position POS. | |
cd1181db | 2087 | If POS is nil, it defaults to the current point. |
3263d5a2 | 2088 | If POS is out of range, the value is nil. */) |
5842a27b | 2089 | (Lisp_Object pos) |
2e344af3 | 2090 | { |
3263d5a2 KH |
2091 | Lisp_Object ch; |
2092 | struct charset *charset; | |
046b1f03 | 2093 | |
3263d5a2 KH |
2094 | ch = Fchar_after (pos); |
2095 | if (! INTEGERP (ch)) | |
2096 | return ch; | |
2097 | charset = CHAR_CHARSET (XINT (ch)); | |
2098 | return (CHARSET_NAME (charset)); | |
6ae1f27e | 2099 | } |
9036eb45 | 2100 | |
87b089ad | 2101 | |
3263d5a2 KH |
2102 | DEFUN ("iso-charset", Fiso_charset, Siso_charset, 3, 3, 0, |
2103 | doc: /* | |
2104 | Return charset of ISO's specification DIMENSION, CHARS, and FINAL-CHAR. | |
2105 | ||
2106 | ISO 2022's designation sequence (escape sequence) distinguishes charsets | |
2107 | by their DIMENSION, CHARS, and FINAL-CHAR, | |
d0cf2d48 | 2108 | whereas Emacs distinguishes them by charset symbol. |
3263d5a2 KH |
2109 | See the documentation of the function `charset-info' for the meanings of |
2110 | DIMENSION, CHARS, and FINAL-CHAR. */) | |
5842a27b | 2111 | (Lisp_Object dimension, Lisp_Object chars, Lisp_Object final_char) |
6ae1f27e | 2112 | { |
3263d5a2 | 2113 | int id; |
d5172d4f | 2114 | bool chars_flag; |
a8a35e61 | 2115 | |
3263d5a2 | 2116 | check_iso_charset_parameter (dimension, chars, final_char); |
82215ce9 KH |
2117 | chars_flag = XFASTINT (chars) == 96; |
2118 | id = ISO_CHARSET_TABLE (XFASTINT (dimension), chars_flag, | |
3263d5a2 KH |
2119 | XFASTINT (final_char)); |
2120 | return (id >= 0 ? CHARSET_NAME (CHARSET_FROM_ID (id)) : Qnil); | |
046b1f03 RS |
2121 | } |
2122 | ||
87b089ad | 2123 | |
3263d5a2 KH |
2124 | DEFUN ("clear-charset-maps", Fclear_charset_maps, Sclear_charset_maps, |
2125 | 0, 0, 0, | |
2126 | doc: /* | |
ecca2aad KH |
2127 | Internal use only. |
2128 | Clear temporary charset mapping tables. | |
2129 | It should be called only from temacs invoked for dumping. */) | |
5842a27b | 2130 | (void) |
87b089ad | 2131 | { |
ecca2aad | 2132 | if (temp_charset_work) |
87b089ad | 2133 | { |
baad03f0 | 2134 | xfree (temp_charset_work); |
ecca2aad | 2135 | temp_charset_work = NULL; |
2e344af3 | 2136 | } |
2e344af3 | 2137 | |
ecca2aad KH |
2138 | if (CHAR_TABLE_P (Vchar_unify_table)) |
2139 | Foptimize_char_table (Vchar_unify_table, Qnil); | |
740f080d | 2140 | |
3263d5a2 | 2141 | return Qnil; |
740f080d KH |
2142 | } |
2143 | ||
8ddf5e57 DL |
2144 | DEFUN ("charset-priority-list", Fcharset_priority_list, |
2145 | Scharset_priority_list, 0, 1, 0, | |
2146 | doc: /* Return the list of charsets ordered by priority. | |
2147 | HIGHESTP non-nil means just return the highest priority one. */) | |
5842a27b | 2148 | (Lisp_Object highestp) |
2e344af3 | 2149 | { |
8ddf5e57 | 2150 | Lisp_Object val = Qnil, list = Vcharset_ordered_list; |
2e344af3 | 2151 | |
8ddf5e57 | 2152 | if (!NILP (highestp)) |
16fed1fc | 2153 | return CHARSET_NAME (CHARSET_FROM_ID (XINT (Fcar (list)))); |
2e344af3 | 2154 | |
8ddf5e57 | 2155 | while (!NILP (list)) |
2e344af3 | 2156 | { |
16fed1fc | 2157 | val = Fcons (CHARSET_NAME (CHARSET_FROM_ID (XINT (XCAR (list)))), val); |
8ddf5e57 | 2158 | list = XCDR (list); |
2e344af3 | 2159 | } |
8ddf5e57 | 2160 | return Fnreverse (val); |
2e344af3 KH |
2161 | } |
2162 | ||
8ddf5e57 DL |
2163 | DEFUN ("set-charset-priority", Fset_charset_priority, Sset_charset_priority, |
2164 | 1, MANY, 0, | |
2165 | doc: /* Assign higher priority to the charsets given as arguments. | |
2166 | usage: (set-charset-priority &rest charsets) */) | |
f66c7cf8 | 2167 | (ptrdiff_t nargs, Lisp_Object *args) |
4ed46869 | 2168 | { |
af7c60ca | 2169 | Lisp_Object new_head, old_list, arglist[2]; |
321c819c | 2170 | Lisp_Object list_2022, list_emacs_mule; |
f66c7cf8 | 2171 | ptrdiff_t i; |
c5101a77 | 2172 | int id; |
4ed46869 | 2173 | |
8ddf5e57 | 2174 | old_list = Fcopy_sequence (Vcharset_ordered_list); |
af7c60ca | 2175 | new_head = Qnil; |
8ddf5e57 | 2176 | for (i = 0; i < nargs; i++) |
4ed46869 | 2177 | { |
8ddf5e57 | 2178 | CHECK_CHARSET_GET_ID (args[i], id); |
af7c60ca KH |
2179 | if (! NILP (Fmemq (make_number (id), old_list))) |
2180 | { | |
2181 | old_list = Fdelq (make_number (id), old_list); | |
2182 | new_head = Fcons (make_number (id), new_head); | |
2183 | } | |
5729c92f | 2184 | } |
8ddf5e57 | 2185 | arglist[0] = Fnreverse (new_head); |
6a9c90ec | 2186 | arglist[1] = Vcharset_non_preferred_head = old_list; |
8ddf5e57 | 2187 | Vcharset_ordered_list = Fnconc (2, arglist); |
dbbb237d | 2188 | charset_ordered_list_tick++; |
5729c92f | 2189 | |
6809ca75 | 2190 | charset_unibyte = -1; |
321c819c | 2191 | for (old_list = Vcharset_ordered_list, list_2022 = list_emacs_mule = Qnil; |
d017b41e | 2192 | CONSP (old_list); old_list = XCDR (old_list)) |
5729c92f | 2193 | { |
e77415b0 | 2194 | if (! NILP (Fmemq (XCAR (old_list), Viso_2022_charset_list))) |
321c819c KH |
2195 | list_2022 = Fcons (XCAR (old_list), list_2022); |
2196 | if (! NILP (Fmemq (XCAR (old_list), Vemacs_mule_charset_list))) | |
2197 | list_emacs_mule = Fcons (XCAR (old_list), list_emacs_mule); | |
6809ca75 KH |
2198 | if (charset_unibyte < 0) |
2199 | { | |
2200 | struct charset *charset = CHARSET_FROM_ID (XINT (XCAR (old_list))); | |
2201 | ||
2202 | if (CHARSET_DIMENSION (charset) == 1 | |
2203 | && CHARSET_ASCII_COMPATIBLE_P (charset) | |
2204 | && CHARSET_MAX_CHAR (charset) >= 0x80) | |
2205 | charset_unibyte = CHARSET_ID (charset); | |
2206 | } | |
4ed46869 | 2207 | } |
321c819c KH |
2208 | Viso_2022_charset_list = Fnreverse (list_2022); |
2209 | Vemacs_mule_charset_list = Fnreverse (list_emacs_mule); | |
6809ca75 KH |
2210 | if (charset_unibyte < 0) |
2211 | charset_unibyte = charset_iso_8859_1; | |
4ed46869 | 2212 | |
8ddf5e57 | 2213 | return Qnil; |
4ed46869 KH |
2214 | } |
2215 | ||
d5b33309 KH |
2216 | DEFUN ("charset-id-internal", Fcharset_id_internal, Scharset_id_internal, |
2217 | 0, 1, 0, | |
2218 | doc: /* Internal use only. | |
2219 | Return charset identification number of CHARSET. */) | |
5842a27b | 2220 | (Lisp_Object charset) |
4ed46869 | 2221 | { |
d5b33309 | 2222 | int id; |
4ed46869 | 2223 | |
d5b33309 KH |
2224 | CHECK_CHARSET_GET_ID (charset, id); |
2225 | return make_number (id); | |
4ed46869 KH |
2226 | } |
2227 | ||
7a84eee5 KH |
2228 | struct charset_sort_data |
2229 | { | |
2230 | Lisp_Object charset; | |
2231 | int id; | |
0065d054 | 2232 | ptrdiff_t priority; |
7a84eee5 KH |
2233 | }; |
2234 | ||
2235 | static int | |
2236 | charset_compare (const void *d1, const void *d2) | |
2237 | { | |
2238 | const struct charset_sort_data *data1 = d1, *data2 = d2; | |
0065d054 PE |
2239 | if (data1->priority != data2->priority) |
2240 | return data1->priority < data2->priority ? -1 : 1; | |
2241 | return 0; | |
7a84eee5 KH |
2242 | } |
2243 | ||
2244 | DEFUN ("sort-charsets", Fsort_charsets, Ssort_charsets, 1, 1, 0, | |
2245 | doc: /* Sort charset list CHARSETS by a priority of each charset. | |
2246 | Return the sorted list. CHARSETS is modified by side effects. | |
2247 | See also `charset-priority-list' and `set-charset-priority'. */) | |
2248 | (Lisp_Object charsets) | |
2249 | { | |
2250 | Lisp_Object len = Flength (charsets); | |
0065d054 PE |
2251 | ptrdiff_t n = XFASTINT (len), i, j; |
2252 | int done; | |
7a84eee5 KH |
2253 | Lisp_Object tail, elt, attrs; |
2254 | struct charset_sort_data *sort_data; | |
726929c4 | 2255 | int id, min_id = INT_MAX, max_id = INT_MIN; |
7a84eee5 KH |
2256 | USE_SAFE_ALLOCA; |
2257 | ||
2258 | if (n == 0) | |
2259 | return Qnil; | |
0065d054 | 2260 | SAFE_NALLOCA (sort_data, 1, n); |
7a84eee5 KH |
2261 | for (tail = charsets, i = 0; CONSP (tail); tail = XCDR (tail), i++) |
2262 | { | |
2263 | elt = XCAR (tail); | |
2264 | CHECK_CHARSET_GET_ATTR (elt, attrs); | |
2265 | sort_data[i].charset = elt; | |
2266 | sort_data[i].id = id = XINT (CHARSET_ATTR_ID (attrs)); | |
726929c4 | 2267 | if (id < min_id) |
7a84eee5 | 2268 | min_id = id; |
726929c4 | 2269 | if (id > max_id) |
7a84eee5 KH |
2270 | max_id = id; |
2271 | } | |
2272 | for (done = 0, tail = Vcharset_ordered_list, i = 0; | |
2273 | done < n && CONSP (tail); tail = XCDR (tail), i++) | |
2274 | { | |
2275 | elt = XCAR (tail); | |
2276 | id = XFASTINT (elt); | |
2277 | if (id >= min_id && id <= max_id) | |
2278 | for (j = 0; j < n; j++) | |
2279 | if (sort_data[j].id == id) | |
2280 | { | |
2281 | sort_data[j].priority = i; | |
2282 | done++; | |
2283 | } | |
2284 | } | |
2285 | qsort (sort_data, n, sizeof *sort_data, charset_compare); | |
2286 | for (i = 0, tail = charsets; CONSP (tail); tail = XCDR (tail), i++) | |
2287 | XSETCAR (tail, sort_data[i].charset); | |
2288 | SAFE_FREE (); | |
2289 | return charsets; | |
2290 | } | |
2291 | ||
4ed46869 | 2292 | \f |
3263d5a2 | 2293 | void |
971de7fb | 2294 | init_charset (void) |
4ed46869 | 2295 | { |
c8f94403 GM |
2296 | Lisp_Object tempdir; |
2297 | tempdir = Fexpand_file_name (build_string ("charsets"), Vdata_directory); | |
73dcdb9f | 2298 | if (! file_accessible_directory_p (SSDATA (tempdir))) |
c8f94403 | 2299 | { |
63541ed8 GM |
2300 | /* This used to be non-fatal (dir_warning), but it should not |
2301 | happen, and if it does sooner or later it will cause some | |
16905bec | 2302 | obscure problem (eg bug#6401), so better abort. */ |
63541ed8 GM |
2303 | fprintf (stderr, "Error: charsets directory not found:\n\ |
2304 | %s\n\ | |
c8f94403 GM |
2305 | Emacs will not function correctly without the character map files.\n\ |
2306 | Please check your installation!\n", | |
63541ed8 GM |
2307 | SDATA (tempdir)); |
2308 | exit (1); | |
c8f94403 GM |
2309 | } |
2310 | ||
2311 | Vcharset_map_path = Fcons (tempdir, Qnil); | |
4ed46869 KH |
2312 | } |
2313 | ||
4ed46869 | 2314 | |
dfcf069d | 2315 | void |
971de7fb | 2316 | init_charset_once (void) |
4ed46869 KH |
2317 | { |
2318 | int i, j, k; | |
2319 | ||
3263d5a2 KH |
2320 | for (i = 0; i < ISO_MAX_DIMENSION; i++) |
2321 | for (j = 0; j < ISO_MAX_CHARS; j++) | |
2322 | for (k = 0; k < ISO_MAX_FINAL; k++) | |
2323 | iso_charset_table[i][j][k] = -1; | |
4ed46869 | 2324 | |
60383934 | 2325 | for (i = 0; i < 256; i++) |
b84ae584 | 2326 | emacs_mule_charset[i] = -1; |
4ed46869 | 2327 | |
7c7dceee KH |
2328 | charset_jisx0201_roman = -1; |
2329 | charset_jisx0208_1978 = -1; | |
2330 | charset_jisx0208 = -1; | |
d32320c4 | 2331 | charset_ksc5601 = -1; |
4ed46869 KH |
2332 | } |
2333 | ||
2334 | #ifdef emacs | |
2335 | ||
f701dc2a PE |
2336 | /* Allocate an initial charset table that is large enough to handle |
2337 | Emacs while it is bootstrapping. As of September 2011, the size | |
2338 | needs to be at least 166; make it a bit bigger to allow for future | |
2339 | expansion. | |
2340 | ||
2341 | Don't make the value so small that the table is reallocated during | |
2342 | bootstrapping, as glibc malloc calls larger than just under 64 KiB | |
2343 | during an initial bootstrap wreak havoc after dumping; see the | |
2344 | M_MMAP_THRESHOLD value in alloc.c, plus there is a extra overhead | |
2345 | internal to glibc malloc and perhaps to Emacs malloc debugging. */ | |
2346 | static struct charset charset_table_init[180]; | |
2347 | ||
dfcf069d | 2348 | void |
971de7fb | 2349 | syms_of_charset (void) |
4ed46869 | 2350 | { |
3263d5a2 KH |
2351 | DEFSYM (Qcharsetp, "charsetp"); |
2352 | ||
2353 | DEFSYM (Qascii, "ascii"); | |
2354 | DEFSYM (Qunicode, "unicode"); | |
6c652beb | 2355 | DEFSYM (Qemacs, "emacs"); |
2fe1edd1 | 2356 | DEFSYM (Qeight_bit, "eight-bit"); |
3263d5a2 KH |
2357 | DEFSYM (Qiso_8859_1, "iso-8859-1"); |
2358 | ||
2359 | DEFSYM (Qgl, "gl"); | |
2360 | DEFSYM (Qgr, "gr"); | |
2361 | ||
3263d5a2 KH |
2362 | staticpro (&Vcharset_ordered_list); |
2363 | Vcharset_ordered_list = Qnil; | |
2364 | ||
2365 | staticpro (&Viso_2022_charset_list); | |
2366 | Viso_2022_charset_list = Qnil; | |
2367 | ||
2368 | staticpro (&Vemacs_mule_charset_list); | |
2369 | Vemacs_mule_charset_list = Qnil; | |
2370 | ||
3943ed76 | 2371 | /* Don't staticpro them here. It's done in syms_of_fns. */ |
088dcc3e DN |
2372 | QCtest = intern_c_string (":test"); |
2373 | Qeq = intern_c_string ("eq"); | |
3943ed76 | 2374 | |
3263d5a2 | 2375 | staticpro (&Vcharset_hash_table); |
8f924df7 KH |
2376 | { |
2377 | Lisp_Object args[2]; | |
2378 | args[0] = QCtest; | |
2379 | args[1] = Qeq; | |
2380 | Vcharset_hash_table = Fmake_hash_table (2, args); | |
2381 | } | |
3263d5a2 | 2382 | |
f701dc2a PE |
2383 | charset_table = charset_table_init; |
2384 | charset_table_size = sizeof charset_table_init / sizeof *charset_table_init; | |
3263d5a2 KH |
2385 | charset_table_used = 0; |
2386 | ||
3263d5a2 KH |
2387 | defsubr (&Scharsetp); |
2388 | defsubr (&Smap_charset_chars); | |
2389 | defsubr (&Sdefine_charset_internal); | |
2390 | defsubr (&Sdefine_charset_alias); | |
3263d5a2 KH |
2391 | defsubr (&Scharset_plist); |
2392 | defsubr (&Sset_charset_plist); | |
2393 | defsubr (&Sunify_charset); | |
3fac5a51 | 2394 | defsubr (&Sget_unused_iso_final_char); |
4ed46869 KH |
2395 | defsubr (&Sdeclare_equiv_charset); |
2396 | defsubr (&Sfind_charset_region); | |
2397 | defsubr (&Sfind_charset_string); | |
3263d5a2 KH |
2398 | defsubr (&Sdecode_char); |
2399 | defsubr (&Sencode_char); | |
4ed46869 | 2400 | defsubr (&Ssplit_char); |
3263d5a2 | 2401 | defsubr (&Smake_char); |
4ed46869 | 2402 | defsubr (&Schar_charset); |
90d7b74e | 2403 | defsubr (&Scharset_after); |
4ed46869 | 2404 | defsubr (&Siso_charset); |
3263d5a2 | 2405 | defsubr (&Sclear_charset_maps); |
8ddf5e57 DL |
2406 | defsubr (&Scharset_priority_list); |
2407 | defsubr (&Sset_charset_priority); | |
d5b33309 | 2408 | defsubr (&Scharset_id_internal); |
7a84eee5 | 2409 | defsubr (&Ssort_charsets); |
3263d5a2 | 2410 | |
29208e82 | 2411 | DEFVAR_LISP ("charset-map-path", Vcharset_map_path, |
fb7ada5f | 2412 | doc: /* List of directories to search for charset map files. */); |
4beef065 | 2413 | Vcharset_map_path = Qnil; |
4ed46869 | 2414 | |
29208e82 | 2415 | DEFVAR_BOOL ("inhibit-load-charset-map", inhibit_load_charset_map, |
ecca2aad KH |
2416 | doc: /* Inhibit loading of charset maps. Used when dumping Emacs. */); |
2417 | inhibit_load_charset_map = 0; | |
2418 | ||
29208e82 | 2419 | DEFVAR_LISP ("charset-list", Vcharset_list, |
528623a0 | 2420 | doc: /* List of all charsets ever defined. */); |
3263d5a2 KH |
2421 | Vcharset_list = Qnil; |
2422 | ||
29208e82 | 2423 | DEFVAR_LISP ("current-iso639-language", Vcurrent_iso639_language, |
6a9c90ec KH |
2424 | doc: /* ISO639 language mnemonic symbol for the current language environment. |
2425 | If the current language environment is for multiple languages (e.g. "Latin-1"), | |
2426 | the value may be a list of mnemonics. */); | |
2427 | Vcurrent_iso639_language = Qnil; | |
2428 | ||
2fe1edd1 KH |
2429 | charset_ascii |
2430 | = define_charset_internal (Qascii, 1, "\x00\x7F\x00\x00\x00\x00", | |
2431 | 0, 127, 'B', -1, 0, 1, 0, 0); | |
14e3d523 KH |
2432 | charset_iso_8859_1 |
2433 | = define_charset_internal (Qiso_8859_1, 1, "\x00\xFF\x00\x00\x00\x00", | |
2434 | 0, 255, -1, -1, -1, 1, 0, 0); | |
2fe1edd1 | 2435 | charset_unicode |
73fbf2d9 | 2436 | = define_charset_internal (Qunicode, 3, "\x00\xFF\x00\xFF\x00\x10", |
2fe1edd1 | 2437 | 0, MAX_UNICODE_CHAR, -1, 0, -1, 1, 0, 0); |
6c652beb KH |
2438 | charset_emacs |
2439 | = define_charset_internal (Qemacs, 3, "\x00\xFF\x00\xFF\x00\x3F", | |
2440 | 0, MAX_5_BYTE_CHAR, -1, 0, -1, 1, 1, 0); | |
2fe1edd1 KH |
2441 | charset_eight_bit |
2442 | = define_charset_internal (Qeight_bit, 1, "\x80\xFF\x00\x00\x00\x00", | |
880820fe | 2443 | 128, 255, -1, 0, -1, 0, 1, |
2fe1edd1 | 2444 | MAX_5_BYTE_CHAR + 1); |
6809ca75 | 2445 | charset_unibyte = charset_iso_8859_1; |
4ed46869 KH |
2446 | } |
2447 | ||
2448 | #endif /* emacs */ |