|
259
|
1 /* Code conversion functions.
|
|
|
2 Copyright (C) 1991, 1995 Free Software Foundation, Inc.
|
|
|
3 Copyright (C) 1995 Sun Microsystems, Inc.
|
|
|
4
|
|
|
5 This file is part of XEmacs.
|
|
|
6
|
|
|
7 XEmacs is free software; you can redistribute it and/or modify it
|
|
|
8 under the terms of the GNU General Public License as published by the
|
|
|
9 Free Software Foundation; either version 2, or (at your option) any
|
|
|
10 later version.
|
|
|
11
|
|
|
12 XEmacs is distributed in the hope that it will be useful, but WITHOUT
|
|
|
13 ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
|
|
|
14 FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
|
|
|
15 for more details.
|
|
|
16
|
|
|
17 You should have received a copy of the GNU General Public License
|
|
|
18 along with XEmacs; see the file COPYING. If not, write to
|
|
|
19 the Free Software Foundation, Inc., 59 Temple Place - Suite 330,
|
|
|
20 Boston, MA 02111-1307, USA. */
|
|
|
21
|
|
|
22 /* Synched up with: Mule 2.3. Not in FSF. */
|
|
|
23
|
|
298
|
24 /* Rewritten by Ben Wing <ben@xemacs.org>. */
|
|
259
|
25
|
|
|
26 #include <config.h>
|
|
|
27 #include "lisp.h"
|
|
398
|
28
|
|
259
|
29 #include "buffer.h"
|
|
|
30 #include "elhash.h"
|
|
|
31 #include "insdel.h"
|
|
|
32 #include "lstream.h"
|
|
398
|
33 #include "opaque.h"
|
|
259
|
34 #ifdef MULE
|
|
|
35 #include "mule-ccl.h"
|
|
396
|
36 #include "chartab.h"
|
|
259
|
37 #endif
|
|
|
38 #include "file-coding.h"
|
|
|
39
|
|
398
|
40 Lisp_Object Qcoding_system_error;
|
|
259
|
41
|
|
|
42 Lisp_Object Vkeyboard_coding_system;
|
|
|
43 Lisp_Object Vterminal_coding_system;
|
|
|
44 Lisp_Object Vcoding_system_for_read;
|
|
|
45 Lisp_Object Vcoding_system_for_write;
|
|
|
46 Lisp_Object Vfile_name_coding_system;
|
|
|
47
|
|
|
48 /* Table of symbols identifying each coding category. */
|
|
|
49 Lisp_Object coding_category_symbol[CODING_CATEGORY_LAST + 1];
|
|
|
50
|
|
398
|
51
|
|
|
52
|
|
|
53 struct file_coding_dump {
|
|
|
54 /* Coding system currently associated with each coding category. */
|
|
|
55 Lisp_Object coding_category_system[CODING_CATEGORY_LAST + 1];
|
|
|
56
|
|
|
57 /* Table of all coding categories in decreasing order of priority.
|
|
|
58 This describes a permutation of the possible coding categories. */
|
|
|
59 int coding_category_by_priority[CODING_CATEGORY_LAST + 1];
|
|
|
60
|
|
|
61 #ifdef MULE
|
|
|
62 Lisp_Object ucs_to_mule_table[65536];
|
|
|
63 #endif
|
|
|
64 } *fcd;
|
|
|
65
|
|
|
66 static const struct lrecord_description fcd_description_1[] = {
|
|
|
67 { XD_LISP_OBJECT_ARRAY, offsetof (struct file_coding_dump, coding_category_system), CODING_CATEGORY_LAST + 1 },
|
|
|
68 #ifdef MULE
|
|
|
69 { XD_LISP_OBJECT_ARRAY, offsetof (struct file_coding_dump, ucs_to_mule_table), 65536 },
|
|
|
70 #endif
|
|
|
71 { XD_END }
|
|
|
72 };
|
|
|
73
|
|
|
74 static const struct struct_description fcd_description = {
|
|
|
75 sizeof (struct file_coding_dump),
|
|
|
76 fcd_description_1
|
|
|
77 };
|
|
|
78
|
|
|
79 Lisp_Object mule_to_ucs_table;
|
|
|
80
|
|
|
81 Lisp_Object Qcoding_systemp;
|
|
|
82
|
|
|
83 Lisp_Object Qraw_text, Qno_conversion, Qccl, Qiso2022;
|
|
259
|
84 /* Qinternal in general.c */
|
|
|
85
|
|
|
86 Lisp_Object Qmnemonic, Qeol_type;
|
|
|
87 Lisp_Object Qcr, Qcrlf, Qlf;
|
|
|
88 Lisp_Object Qeol_cr, Qeol_crlf, Qeol_lf;
|
|
|
89 Lisp_Object Qpost_read_conversion;
|
|
|
90 Lisp_Object Qpre_write_conversion;
|
|
|
91
|
|
|
92 #ifdef MULE
|
|
396
|
93 Lisp_Object Qucs4, Qutf8;
|
|
259
|
94 Lisp_Object Qbig5, Qshift_jis;
|
|
|
95 Lisp_Object Qcharset_g0, Qcharset_g1, Qcharset_g2, Qcharset_g3;
|
|
|
96 Lisp_Object Qforce_g0_on_output, Qforce_g1_on_output;
|
|
|
97 Lisp_Object Qforce_g2_on_output, Qforce_g3_on_output;
|
|
|
98 Lisp_Object Qno_iso6429;
|
|
|
99 Lisp_Object Qinput_charset_conversion, Qoutput_charset_conversion;
|
|
398
|
100 Lisp_Object Qescape_quoted;
|
|
259
|
101 Lisp_Object Qshort, Qno_ascii_eol, Qno_ascii_cntl, Qseven, Qlock_shift;
|
|
263
|
102 #endif
|
|
|
103 Lisp_Object Qencode, Qdecode;
|
|
259
|
104
|
|
380
|
105 Lisp_Object Vcoding_system_hash_table;
|
|
259
|
106
|
|
|
107 int enable_multibyte_characters;
|
|
|
108
|
|
|
109 #ifdef MULE
|
|
|
110 /* Additional information used by the ISO2022 decoder and detector. */
|
|
|
111 struct iso2022_decoder
|
|
|
112 {
|
|
|
113 /* CHARSET holds the character sets currently assigned to the G0
|
|
|
114 through G3 variables. It is initialized from the array
|
|
|
115 INITIAL_CHARSET in CODESYS. */
|
|
|
116 Lisp_Object charset[4];
|
|
|
117
|
|
|
118 /* Which registers are currently invoked into the left (GL) and
|
|
|
119 right (GR) halves of the 8-bit encoding space? */
|
|
|
120 int register_left, register_right;
|
|
|
121
|
|
|
122 /* ISO_ESC holds a value indicating part of an escape sequence
|
|
|
123 that has already been seen. */
|
|
|
124 enum iso_esc_flag esc;
|
|
|
125
|
|
|
126 /* This records the bytes we've seen so far in an escape sequence,
|
|
|
127 in case the sequence is invalid (we spit out the bytes unchanged). */
|
|
|
128 unsigned char esc_bytes[8];
|
|
|
129
|
|
|
130 /* Index for next byte to store in ISO escape sequence. */
|
|
|
131 int esc_bytes_index;
|
|
|
132
|
|
396
|
133 #ifdef ENABLE_COMPOSITE_CHARS
|
|
259
|
134 /* Stuff seen so far when composing a string. */
|
|
|
135 unsigned_char_dynarr *composite_chars;
|
|
396
|
136 #endif
|
|
259
|
137
|
|
|
138 /* If we saw an invalid designation sequence for a particular
|
|
|
139 register, we flag it here and switch to ASCII. The next time we
|
|
|
140 see a valid designation for this register, we turn off the flag
|
|
|
141 and do the designation normally, but pretend the sequence was
|
|
|
142 invalid. The effect of all this is that (most of the time) the
|
|
|
143 escape sequences for both the switch to the unknown charset, and
|
|
|
144 the switch back to the known charset, get inserted literally into
|
|
|
145 the buffer and saved out as such. The hope is that we can
|
|
|
146 preserve the escape sequences so that the resulting written out
|
|
|
147 file makes sense. If we don't do any of this, the designation
|
|
|
148 to the invalid charset will be preserved but that switch back
|
|
|
149 to the known charset will probably get eaten because it was
|
|
|
150 the same charset that was already present in the register. */
|
|
|
151 unsigned char invalid_designated[4];
|
|
|
152
|
|
|
153 /* We try to do similar things as above for direction-switching
|
|
|
154 sequences. If we encountered a direction switch while an
|
|
|
155 invalid designation was present, or an invalid designation
|
|
|
156 just after a direction switch (i.e. no valid designation
|
|
|
157 encountered yet), we insert the direction-switch escape
|
|
|
158 sequence literally into the output stream, and later on
|
|
|
159 insert the corresponding direction-restoring escape sequence
|
|
|
160 literally also. */
|
|
|
161 unsigned int switched_dir_and_no_valid_charset_yet :1;
|
|
|
162 unsigned int invalid_switch_dir :1;
|
|
|
163
|
|
|
164 /* Tells the decoder to output the escape sequence literally
|
|
|
165 even though it was valid. Used in the games we play to
|
|
|
166 avoid lossage when we encounter invalid designations. */
|
|
|
167 unsigned int output_literally :1;
|
|
|
168 /* We encountered a direction switch followed by an invalid
|
|
|
169 designation. We didn't output the direction switch
|
|
|
170 literally because we didn't know about the invalid designation;
|
|
|
171 but we have to do so now. */
|
|
|
172 unsigned int output_direction_sequence :1;
|
|
|
173 };
|
|
272
|
174 #endif /* MULE */
|
|
|
175 EXFUN (Fcopy_coding_system, 2);
|
|
259
|
176 #ifdef MULE
|
|
|
177 struct detection_state;
|
|
|
178 static int detect_coding_sjis (struct detection_state *st,
|
|
398
|
179 const unsigned char *src,
|
|
259
|
180 unsigned int n);
|
|
|
181 static void decode_coding_sjis (Lstream *decoding,
|
|
398
|
182 const unsigned char *src,
|
|
259
|
183 unsigned_char_dynarr *dst,
|
|
|
184 unsigned int n);
|
|
|
185 static void encode_coding_sjis (Lstream *encoding,
|
|
398
|
186 const unsigned char *src,
|
|
259
|
187 unsigned_char_dynarr *dst,
|
|
|
188 unsigned int n);
|
|
|
189 static int detect_coding_big5 (struct detection_state *st,
|
|
398
|
190 const unsigned char *src,
|
|
259
|
191 unsigned int n);
|
|
|
192 static void decode_coding_big5 (Lstream *decoding,
|
|
398
|
193 const unsigned char *src,
|
|
259
|
194 unsigned_char_dynarr *dst, unsigned int n);
|
|
|
195 static void encode_coding_big5 (Lstream *encoding,
|
|
398
|
196 const unsigned char *src,
|
|
259
|
197 unsigned_char_dynarr *dst, unsigned int n);
|
|
396
|
198 static int detect_coding_ucs4 (struct detection_state *st,
|
|
398
|
199 const unsigned char *src,
|
|
396
|
200 unsigned int n);
|
|
|
201 static void decode_coding_ucs4 (Lstream *decoding,
|
|
398
|
202 const unsigned char *src,
|
|
396
|
203 unsigned_char_dynarr *dst, unsigned int n);
|
|
|
204 static void encode_coding_ucs4 (Lstream *encoding,
|
|
398
|
205 const unsigned char *src,
|
|
396
|
206 unsigned_char_dynarr *dst, unsigned int n);
|
|
|
207 static int detect_coding_utf8 (struct detection_state *st,
|
|
398
|
208 const unsigned char *src,
|
|
396
|
209 unsigned int n);
|
|
|
210 static void decode_coding_utf8 (Lstream *decoding,
|
|
398
|
211 const unsigned char *src,
|
|
396
|
212 unsigned_char_dynarr *dst, unsigned int n);
|
|
|
213 static void encode_coding_utf8 (Lstream *encoding,
|
|
398
|
214 const unsigned char *src,
|
|
396
|
215 unsigned_char_dynarr *dst, unsigned int n);
|
|
259
|
216 static int postprocess_iso2022_mask (int mask);
|
|
|
217 static void reset_iso2022 (Lisp_Object coding_system,
|
|
|
218 struct iso2022_decoder *iso);
|
|
|
219 static int detect_coding_iso2022 (struct detection_state *st,
|
|
398
|
220 const unsigned char *src,
|
|
259
|
221 unsigned int n);
|
|
|
222 static void decode_coding_iso2022 (Lstream *decoding,
|
|
398
|
223 const unsigned char *src,
|
|
259
|
224 unsigned_char_dynarr *dst, unsigned int n);
|
|
|
225 static void encode_coding_iso2022 (Lstream *encoding,
|
|
398
|
226 const unsigned char *src,
|
|
259
|
227 unsigned_char_dynarr *dst, unsigned int n);
|
|
|
228 #endif /* MULE */
|
|
|
229 static void decode_coding_no_conversion (Lstream *decoding,
|
|
398
|
230 const unsigned char *src,
|
|
259
|
231 unsigned_char_dynarr *dst,
|
|
|
232 unsigned int n);
|
|
|
233 static void encode_coding_no_conversion (Lstream *encoding,
|
|
398
|
234 const unsigned char *src,
|
|
259
|
235 unsigned_char_dynarr *dst,
|
|
|
236 unsigned int n);
|
|
398
|
237 static void mule_decode (Lstream *decoding, const unsigned char *src,
|
|
259
|
238 unsigned_char_dynarr *dst, unsigned int n);
|
|
398
|
239 static void mule_encode (Lstream *encoding, const unsigned char *src,
|
|
259
|
240 unsigned_char_dynarr *dst, unsigned int n);
|
|
|
241
|
|
|
242 typedef struct codesys_prop codesys_prop;
|
|
|
243 struct codesys_prop
|
|
|
244 {
|
|
|
245 Lisp_Object sym;
|
|
|
246 int prop_type;
|
|
|
247 };
|
|
|
248
|
|
|
249 typedef struct
|
|
|
250 {
|
|
|
251 Dynarr_declare (codesys_prop);
|
|
|
252 } codesys_prop_dynarr;
|
|
|
253
|
|
398
|
254 static const struct lrecord_description codesys_prop_description_1[] = {
|
|
|
255 { XD_LISP_OBJECT, offsetof (codesys_prop, sym) },
|
|
|
256 { XD_END }
|
|
|
257 };
|
|
|
258
|
|
|
259 static const struct struct_description codesys_prop_description = {
|
|
|
260 sizeof (codesys_prop),
|
|
|
261 codesys_prop_description_1
|
|
|
262 };
|
|
|
263
|
|
|
264 static const struct lrecord_description codesys_prop_dynarr_description_1[] = {
|
|
|
265 XD_DYNARR_DESC (codesys_prop_dynarr, &codesys_prop_description),
|
|
|
266 { XD_END }
|
|
|
267 };
|
|
|
268
|
|
|
269 static const struct struct_description codesys_prop_dynarr_description = {
|
|
|
270 sizeof (codesys_prop_dynarr),
|
|
|
271 codesys_prop_dynarr_description_1
|
|
|
272 };
|
|
|
273
|
|
259
|
274 codesys_prop_dynarr *the_codesys_prop_dynarr;
|
|
|
275
|
|
|
276 enum codesys_prop_enum
|
|
|
277 {
|
|
|
278 CODESYS_PROP_ALL_OK,
|
|
|
279 CODESYS_PROP_ISO2022,
|
|
|
280 CODESYS_PROP_CCL
|
|
|
281 };
|
|
|
282
|
|
|
283
|
|
|
284 /************************************************************************/
|
|
|
285 /* Coding system functions */
|
|
|
286 /************************************************************************/
|
|
|
287
|
|
398
|
288 static Lisp_Object mark_coding_system (Lisp_Object);
|
|
259
|
289 static void print_coding_system (Lisp_Object, Lisp_Object, int);
|
|
|
290 static void finalize_coding_system (void *header, int for_disksave);
|
|
|
291
|
|
398
|
292 #ifdef MULE
|
|
|
293 static const struct lrecord_description ccs_description_1[] = {
|
|
|
294 { XD_LISP_OBJECT, offsetof (charset_conversion_spec, from_charset) },
|
|
|
295 { XD_LISP_OBJECT, offsetof (charset_conversion_spec, to_charset) },
|
|
|
296 { XD_END }
|
|
|
297 };
|
|
|
298
|
|
|
299 static const struct struct_description ccs_description = {
|
|
|
300 sizeof (charset_conversion_spec),
|
|
|
301 ccs_description_1
|
|
|
302 };
|
|
|
303
|
|
|
304 static const struct lrecord_description ccsd_description_1[] = {
|
|
|
305 XD_DYNARR_DESC (charset_conversion_spec_dynarr, &ccs_description),
|
|
|
306 { XD_END }
|
|
|
307 };
|
|
|
308
|
|
|
309 static const struct struct_description ccsd_description = {
|
|
|
310 sizeof (charset_conversion_spec_dynarr),
|
|
|
311 ccsd_description_1
|
|
|
312 };
|
|
|
313 #endif
|
|
|
314
|
|
|
315 static const struct lrecord_description coding_system_description[] = {
|
|
|
316 { XD_LISP_OBJECT, offsetof (Lisp_Coding_System, name) },
|
|
|
317 { XD_LISP_OBJECT, offsetof (Lisp_Coding_System, doc_string) },
|
|
|
318 { XD_LISP_OBJECT, offsetof (Lisp_Coding_System, mnemonic) },
|
|
|
319 { XD_LISP_OBJECT, offsetof (Lisp_Coding_System, post_read_conversion) },
|
|
|
320 { XD_LISP_OBJECT, offsetof (Lisp_Coding_System, pre_write_conversion) },
|
|
|
321 { XD_LISP_OBJECT, offsetof (Lisp_Coding_System, eol_lf) },
|
|
|
322 { XD_LISP_OBJECT, offsetof (Lisp_Coding_System, eol_crlf) },
|
|
|
323 { XD_LISP_OBJECT, offsetof (Lisp_Coding_System, eol_cr) },
|
|
|
324 #ifdef MULE
|
|
|
325 { XD_LISP_OBJECT_ARRAY, offsetof (Lisp_Coding_System, iso2022.initial_charset), 4 },
|
|
|
326 { XD_STRUCT_PTR, offsetof (Lisp_Coding_System, iso2022.input_conv), 1, &ccsd_description },
|
|
|
327 { XD_STRUCT_PTR, offsetof (Lisp_Coding_System, iso2022.output_conv), 1, &ccsd_description },
|
|
|
328 { XD_LISP_OBJECT, offsetof (Lisp_Coding_System, ccl.decode) },
|
|
|
329 { XD_LISP_OBJECT, offsetof (Lisp_Coding_System, ccl.encode) },
|
|
|
330 #endif
|
|
|
331 { XD_END }
|
|
|
332 };
|
|
|
333
|
|
259
|
334 DEFINE_LRECORD_IMPLEMENTATION ("coding-system", coding_system,
|
|
|
335 mark_coding_system, print_coding_system,
|
|
|
336 finalize_coding_system,
|
|
398
|
337 0, 0, coding_system_description,
|
|
|
338 Lisp_Coding_System);
|
|
259
|
339
|
|
|
340 static Lisp_Object
|
|
398
|
341 mark_coding_system (Lisp_Object obj)
|
|
259
|
342 {
|
|
396
|
343 Lisp_Coding_System *codesys = XCODING_SYSTEM (obj);
|
|
259
|
344
|
|
398
|
345 mark_object (CODING_SYSTEM_NAME (codesys));
|
|
|
346 mark_object (CODING_SYSTEM_DOC_STRING (codesys));
|
|
|
347 mark_object (CODING_SYSTEM_MNEMONIC (codesys));
|
|
|
348 mark_object (CODING_SYSTEM_EOL_LF (codesys));
|
|
|
349 mark_object (CODING_SYSTEM_EOL_CRLF (codesys));
|
|
|
350 mark_object (CODING_SYSTEM_EOL_CR (codesys));
|
|
259
|
351
|
|
|
352 switch (CODING_SYSTEM_TYPE (codesys))
|
|
|
353 {
|
|
|
354 #ifdef MULE
|
|
|
355 int i;
|
|
|
356 case CODESYS_ISO2022:
|
|
|
357 for (i = 0; i < 4; i++)
|
|
398
|
358 mark_object (CODING_SYSTEM_ISO2022_INITIAL_CHARSET (codesys, i));
|
|
259
|
359 if (codesys->iso2022.input_conv)
|
|
|
360 {
|
|
|
361 for (i = 0; i < Dynarr_length (codesys->iso2022.input_conv); i++)
|
|
|
362 {
|
|
|
363 struct charset_conversion_spec *ccs =
|
|
|
364 Dynarr_atp (codesys->iso2022.input_conv, i);
|
|
398
|
365 mark_object (ccs->from_charset);
|
|
|
366 mark_object (ccs->to_charset);
|
|
259
|
367 }
|
|
|
368 }
|
|
|
369 if (codesys->iso2022.output_conv)
|
|
|
370 {
|
|
|
371 for (i = 0; i < Dynarr_length (codesys->iso2022.output_conv); i++)
|
|
|
372 {
|
|
|
373 struct charset_conversion_spec *ccs =
|
|
|
374 Dynarr_atp (codesys->iso2022.output_conv, i);
|
|
398
|
375 mark_object (ccs->from_charset);
|
|
|
376 mark_object (ccs->to_charset);
|
|
259
|
377 }
|
|
|
378 }
|
|
|
379 break;
|
|
|
380
|
|
|
381 case CODESYS_CCL:
|
|
398
|
382 mark_object (CODING_SYSTEM_CCL_DECODE (codesys));
|
|
|
383 mark_object (CODING_SYSTEM_CCL_ENCODE (codesys));
|
|
259
|
384 break;
|
|
272
|
385 #endif /* MULE */
|
|
259
|
386 default:
|
|
|
387 break;
|
|
|
388 }
|
|
|
389
|
|
398
|
390 mark_object (CODING_SYSTEM_PRE_WRITE_CONVERSION (codesys));
|
|
259
|
391 return CODING_SYSTEM_POST_READ_CONVERSION (codesys);
|
|
|
392 }
|
|
|
393
|
|
|
394 static void
|
|
|
395 print_coding_system (Lisp_Object obj, Lisp_Object printcharfun,
|
|
|
396 int escapeflag)
|
|
|
397 {
|
|
396
|
398 Lisp_Coding_System *c = XCODING_SYSTEM (obj);
|
|
259
|
399 if (print_readably)
|
|
|
400 error ("printing unreadable object #<coding_system 0x%x>",
|
|
|
401 c->header.uid);
|
|
|
402
|
|
|
403 write_c_string ("#<coding_system ", printcharfun);
|
|
|
404 print_internal (c->name, printcharfun, 1);
|
|
|
405 write_c_string (">", printcharfun);
|
|
|
406 }
|
|
|
407
|
|
|
408 static void
|
|
|
409 finalize_coding_system (void *header, int for_disksave)
|
|
|
410 {
|
|
396
|
411 Lisp_Coding_System *c = (Lisp_Coding_System *) header;
|
|
259
|
412 /* Since coding systems never go away, this function is not
|
|
|
413 necessary. But it would be necessary if we changed things
|
|
|
414 so that coding systems could go away. */
|
|
|
415 if (!for_disksave) /* see comment in lstream.c */
|
|
|
416 {
|
|
|
417 switch (CODING_SYSTEM_TYPE (c))
|
|
|
418 {
|
|
|
419 #ifdef MULE
|
|
|
420 case CODESYS_ISO2022:
|
|
|
421 if (c->iso2022.input_conv)
|
|
|
422 {
|
|
|
423 Dynarr_free (c->iso2022.input_conv);
|
|
|
424 c->iso2022.input_conv = 0;
|
|
|
425 }
|
|
|
426 if (c->iso2022.output_conv)
|
|
|
427 {
|
|
|
428 Dynarr_free (c->iso2022.output_conv);
|
|
|
429 c->iso2022.output_conv = 0;
|
|
|
430 }
|
|
|
431 break;
|
|
272
|
432 #endif /* MULE */
|
|
259
|
433 default:
|
|
|
434 break;
|
|
|
435 }
|
|
|
436 }
|
|
|
437 }
|
|
|
438
|
|
398
|
439 static eol_type_t
|
|
259
|
440 symbol_to_eol_type (Lisp_Object symbol)
|
|
|
441 {
|
|
|
442 CHECK_SYMBOL (symbol);
|
|
|
443 if (NILP (symbol)) return EOL_AUTODETECT;
|
|
|
444 if (EQ (symbol, Qlf)) return EOL_LF;
|
|
|
445 if (EQ (symbol, Qcrlf)) return EOL_CRLF;
|
|
|
446 if (EQ (symbol, Qcr)) return EOL_CR;
|
|
|
447
|
|
|
448 signal_simple_error ("Unrecognized eol type", symbol);
|
|
|
449 return EOL_AUTODETECT; /* not reached */
|
|
|
450 }
|
|
|
451
|
|
|
452 static Lisp_Object
|
|
398
|
453 eol_type_to_symbol (eol_type_t type)
|
|
259
|
454 {
|
|
|
455 switch (type)
|
|
|
456 {
|
|
380
|
457 default: abort ();
|
|
259
|
458 case EOL_LF: return Qlf;
|
|
|
459 case EOL_CRLF: return Qcrlf;
|
|
|
460 case EOL_CR: return Qcr;
|
|
|
461 case EOL_AUTODETECT: return Qnil;
|
|
|
462 }
|
|
|
463 }
|
|
|
464
|
|
|
465 static void
|
|
396
|
466 setup_eol_coding_systems (Lisp_Coding_System *codesys)
|
|
259
|
467 {
|
|
272
|
468 Lisp_Object codesys_obj;
|
|
259
|
469 int len = string_length (XSYMBOL (CODING_SYSTEM_NAME (codesys))->name);
|
|
|
470 char *codesys_name = (char *) alloca (len + 7);
|
|
261
|
471 int mlen = -1;
|
|
265
|
472 char *codesys_mnemonic=0;
|
|
259
|
473
|
|
|
474 Lisp_Object codesys_name_sym, sub_codesys_obj;
|
|
|
475
|
|
|
476 /* kludge */
|
|
|
477
|
|
|
478 XSETCODING_SYSTEM (codesys_obj, codesys);
|
|
|
479
|
|
|
480 memcpy (codesys_name,
|
|
|
481 string_data (XSYMBOL (CODING_SYSTEM_NAME (codesys))->name), len);
|
|
|
482
|
|
261
|
483 if (STRINGP (CODING_SYSTEM_MNEMONIC (codesys)))
|
|
|
484 {
|
|
272
|
485 mlen = XSTRING_LENGTH (CODING_SYSTEM_MNEMONIC (codesys));
|
|
261
|
486 codesys_mnemonic = (char *) alloca (mlen + 7);
|
|
|
487 memcpy (codesys_mnemonic,
|
|
|
488 XSTRING_DATA (CODING_SYSTEM_MNEMONIC (codesys)), mlen);
|
|
|
489 }
|
|
259
|
490
|
|
272
|
491 #define DEFINE_SUB_CODESYS(op_sys, op_sys_abbr, Type) do { \
|
|
|
492 strcpy (codesys_name + len, "-" op_sys); \
|
|
|
493 if (mlen != -1) \
|
|
|
494 strcpy (codesys_mnemonic + mlen, op_sys_abbr); \
|
|
|
495 codesys_name_sym = intern (codesys_name); \
|
|
|
496 sub_codesys_obj = Fcopy_coding_system (codesys_obj, codesys_name_sym); \
|
|
|
497 XCODING_SYSTEM_EOL_TYPE (sub_codesys_obj) = Type; \
|
|
|
498 if (mlen != -1) \
|
|
|
499 XCODING_SYSTEM_MNEMONIC(sub_codesys_obj) = \
|
|
|
500 build_string (codesys_mnemonic); \
|
|
|
501 CODING_SYSTEM_##Type (codesys) = sub_codesys_obj; \
|
|
259
|
502 } while (0)
|
|
|
503
|
|
|
504 DEFINE_SUB_CODESYS("unix", "", EOL_LF);
|
|
263
|
505 DEFINE_SUB_CODESYS("dos", ":T", EOL_CRLF);
|
|
|
506 DEFINE_SUB_CODESYS("mac", ":t", EOL_CR);
|
|
259
|
507 }
|
|
|
508
|
|
|
509 DEFUN ("coding-system-p", Fcoding_system_p, 1, 1, 0, /*
|
|
272
|
510 Return t if OBJECT is a coding system.
|
|
259
|
511 A coding system is an object that defines how text containing multiple
|
|
|
512 character sets is encoded into a stream of (typically 8-bit) bytes.
|
|
|
513 The coding system is used to decode the stream into a series of
|
|
|
514 characters (which may be from multiple charsets) when the text is read
|
|
|
515 from a file or process, and is used to encode the text back into the
|
|
|
516 same format when it is written out to a file or process.
|
|
|
517
|
|
|
518 For example, many ISO2022-compliant coding systems (such as Compound
|
|
|
519 Text, which is used for inter-client data under the X Window System)
|
|
|
520 use escape sequences to switch between different charsets -- Japanese
|
|
|
521 Kanji, for example, is invoked with "ESC $ ( B"; ASCII is invoked
|
|
|
522 with "ESC ( B"; and Cyrillic is invoked with "ESC - L". See
|
|
|
523 `make-coding-system' for more information.
|
|
|
524
|
|
|
525 Coding systems are normally identified using a symbol, and the
|
|
|
526 symbol is accepted in place of the actual coding system object whenever
|
|
|
527 a coding system is called for. (This is similar to how faces work.)
|
|
|
528 */
|
|
|
529 (object))
|
|
|
530 {
|
|
|
531 return CODING_SYSTEMP (object) ? Qt : Qnil;
|
|
|
532 }
|
|
|
533
|
|
|
534 DEFUN ("find-coding-system", Ffind_coding_system, 1, 1, 0, /*
|
|
|
535 Retrieve the coding system of the given name.
|
|
|
536
|
|
|
537 If CODING-SYSTEM-OR-NAME is a coding-system object, it is simply
|
|
|
538 returned. Otherwise, CODING-SYSTEM-OR-NAME should be a symbol.
|
|
|
539 If there is no such coding system, nil is returned. Otherwise the
|
|
|
540 associated coding system object is returned.
|
|
|
541 */
|
|
|
542 (coding_system_or_name))
|
|
|
543 {
|
|
|
544 if (NILP (coding_system_or_name))
|
|
|
545 coding_system_or_name = Qbinary;
|
|
398
|
546 else if (CODING_SYSTEMP (coding_system_or_name))
|
|
|
547 return coding_system_or_name;
|
|
272
|
548 else
|
|
|
549 CHECK_SYMBOL (coding_system_or_name);
|
|
259
|
550
|
|
398
|
551 while (1)
|
|
|
552 {
|
|
|
553 coding_system_or_name =
|
|
|
554 Fgethash (coding_system_or_name, Vcoding_system_hash_table, Qnil);
|
|
|
555
|
|
|
556 if (CODING_SYSTEMP (coding_system_or_name) || NILP (coding_system_or_name))
|
|
|
557 return coding_system_or_name;
|
|
|
558 }
|
|
259
|
559 }
|
|
|
560
|
|
|
561 DEFUN ("get-coding-system", Fget_coding_system, 1, 1, 0, /*
|
|
|
562 Retrieve the coding system of the given name.
|
|
|
563 Same as `find-coding-system' except that if there is no such
|
|
|
564 coding system, an error is signaled instead of returning nil.
|
|
|
565 */
|
|
|
566 (name))
|
|
|
567 {
|
|
|
568 Lisp_Object coding_system = Ffind_coding_system (name);
|
|
|
569
|
|
|
570 if (NILP (coding_system))
|
|
|
571 signal_simple_error ("No such coding system", name);
|
|
|
572 return coding_system;
|
|
|
573 }
|
|
|
574
|
|
|
575 /* We store the coding systems in hash tables with the names as the key and the
|
|
|
576 actual coding system object as the value. Occasionally we need to use them
|
|
|
577 in a list format. These routines provide us with that. */
|
|
|
578 struct coding_system_list_closure
|
|
|
579 {
|
|
|
580 Lisp_Object *coding_system_list;
|
|
|
581 };
|
|
|
582
|
|
|
583 static int
|
|
380
|
584 add_coding_system_to_list_mapper (Lisp_Object key, Lisp_Object value,
|
|
259
|
585 void *coding_system_list_closure)
|
|
|
586 {
|
|
|
587 /* This function can GC */
|
|
|
588 struct coding_system_list_closure *cscl =
|
|
|
589 (struct coding_system_list_closure *) coding_system_list_closure;
|
|
380
|
590 Lisp_Object *coding_system_list = cscl->coding_system_list;
|
|
|
591
|
|
398
|
592 *coding_system_list = Fcons (key, *coding_system_list);
|
|
259
|
593 return 0;
|
|
|
594 }
|
|
|
595
|
|
|
596 DEFUN ("coding-system-list", Fcoding_system_list, 0, 0, 0, /*
|
|
|
597 Return a list of the names of all defined coding systems.
|
|
|
598 */
|
|
|
599 ())
|
|
|
600 {
|
|
|
601 Lisp_Object coding_system_list = Qnil;
|
|
|
602 struct gcpro gcpro1;
|
|
|
603 struct coding_system_list_closure coding_system_list_closure;
|
|
|
604
|
|
|
605 GCPRO1 (coding_system_list);
|
|
|
606 coding_system_list_closure.coding_system_list = &coding_system_list;
|
|
380
|
607 elisp_maphash (add_coding_system_to_list_mapper, Vcoding_system_hash_table,
|
|
259
|
608 &coding_system_list_closure);
|
|
|
609 UNGCPRO;
|
|
|
610
|
|
|
611 return coding_system_list;
|
|
|
612 }
|
|
|
613
|
|
|
614 DEFUN ("coding-system-name", Fcoding_system_name, 1, 1, 0, /*
|
|
|
615 Return the name of the given coding system.
|
|
|
616 */
|
|
|
617 (coding_system))
|
|
|
618 {
|
|
|
619 coding_system = Fget_coding_system (coding_system);
|
|
|
620 return XCODING_SYSTEM_NAME (coding_system);
|
|
|
621 }
|
|
|
622
|
|
396
|
623 static Lisp_Coding_System *
|
|
259
|
624 allocate_coding_system (enum coding_system_type type, Lisp_Object name)
|
|
|
625 {
|
|
396
|
626 Lisp_Coding_System *codesys =
|
|
398
|
627 alloc_lcrecord_type (Lisp_Coding_System, &lrecord_coding_system);
|
|
259
|
628
|
|
|
629 zero_lcrecord (codesys);
|
|
|
630 CODING_SYSTEM_PRE_WRITE_CONVERSION (codesys) = Qnil;
|
|
|
631 CODING_SYSTEM_POST_READ_CONVERSION (codesys) = Qnil;
|
|
|
632 CODING_SYSTEM_EOL_TYPE (codesys) = EOL_AUTODETECT;
|
|
|
633 CODING_SYSTEM_EOL_CRLF (codesys) = Qnil;
|
|
|
634 CODING_SYSTEM_EOL_CR (codesys) = Qnil;
|
|
|
635 CODING_SYSTEM_EOL_LF (codesys) = Qnil;
|
|
|
636 CODING_SYSTEM_TYPE (codesys) = type;
|
|
261
|
637 CODING_SYSTEM_MNEMONIC (codesys) = Qnil;
|
|
259
|
638 #ifdef MULE
|
|
|
639 if (type == CODESYS_ISO2022)
|
|
|
640 {
|
|
|
641 int i;
|
|
|
642 for (i = 0; i < 4; i++)
|
|
|
643 CODING_SYSTEM_ISO2022_INITIAL_CHARSET (codesys, i) = Qnil;
|
|
|
644 }
|
|
|
645 else if (type == CODESYS_CCL)
|
|
|
646 {
|
|
|
647 CODING_SYSTEM_CCL_DECODE (codesys) = Qnil;
|
|
|
648 CODING_SYSTEM_CCL_ENCODE (codesys) = Qnil;
|
|
|
649 }
|
|
272
|
650 #endif /* MULE */
|
|
259
|
651 CODING_SYSTEM_NAME (codesys) = name;
|
|
|
652
|
|
|
653 return codesys;
|
|
|
654 }
|
|
|
655
|
|
|
656 #ifdef MULE
|
|
|
657 /* Given a list of charset conversion specs as specified in a Lisp
|
|
|
658 program, parse it into STORE_HERE. */
|
|
|
659
|
|
|
660 static void
|
|
|
661 parse_charset_conversion_specs (charset_conversion_spec_dynarr *store_here,
|
|
|
662 Lisp_Object spec_list)
|
|
|
663 {
|
|
|
664 Lisp_Object rest;
|
|
|
665
|
|
|
666 EXTERNAL_LIST_LOOP (rest, spec_list)
|
|
|
667 {
|
|
|
668 Lisp_Object car = XCAR (rest);
|
|
|
669 Lisp_Object from, to;
|
|
|
670 struct charset_conversion_spec spec;
|
|
|
671
|
|
|
672 if (!CONSP (car) || !CONSP (XCDR (car)) || !NILP (XCDR (XCDR (car))))
|
|
|
673 signal_simple_error ("Invalid charset conversion spec", car);
|
|
|
674 from = Fget_charset (XCAR (car));
|
|
|
675 to = Fget_charset (XCAR (XCDR (car)));
|
|
|
676 if (XCHARSET_TYPE (from) != XCHARSET_TYPE (to))
|
|
|
677 signal_simple_error_2
|
|
|
678 ("Attempted conversion between different charset types",
|
|
|
679 from, to);
|
|
|
680 spec.from_charset = from;
|
|
|
681 spec.to_charset = to;
|
|
|
682
|
|
|
683 Dynarr_add (store_here, spec);
|
|
|
684 }
|
|
|
685 }
|
|
|
686
|
|
|
687 /* Given a dynarr LOAD_HERE of internally-stored charset conversion
|
|
|
688 specs, return the equivalent as the Lisp programmer would see it.
|
|
|
689
|
|
|
690 If LOAD_HERE is 0, return Qnil. */
|
|
|
691
|
|
|
692 static Lisp_Object
|
|
|
693 unparse_charset_conversion_specs (charset_conversion_spec_dynarr *load_here)
|
|
|
694 {
|
|
|
695 int i;
|
|
272
|
696 Lisp_Object result;
|
|
259
|
697
|
|
|
698 if (!load_here)
|
|
|
699 return Qnil;
|
|
272
|
700 for (i = 0, result = Qnil; i < Dynarr_length (load_here); i++)
|
|
259
|
701 {
|
|
272
|
702 struct charset_conversion_spec *ccs = Dynarr_atp (load_here, i);
|
|
259
|
703 result = Fcons (list2 (ccs->from_charset, ccs->to_charset), result);
|
|
|
704 }
|
|
|
705
|
|
|
706 return Fnreverse (result);
|
|
|
707 }
|
|
|
708
|
|
272
|
709 #endif /* MULE */
|
|
259
|
710
|
|
|
711 DEFUN ("make-coding-system", Fmake_coding_system, 2, 4, 0, /*
|
|
|
712 Register symbol NAME as a coding system.
|
|
|
713
|
|
|
714 TYPE describes the conversion method used and should be one of
|
|
|
715
|
|
|
716 nil or 'undecided
|
|
|
717 Automatic conversion. XEmacs attempts to detect the coding system
|
|
|
718 used in the file.
|
|
|
719 'no-conversion
|
|
|
720 No conversion. Use this for binary files and such. On output,
|
|
|
721 graphic characters that are not in ASCII or Latin-1 will be
|
|
|
722 replaced by a ?. (For a no-conversion-encoded buffer, these
|
|
|
723 characters will only be present if you explicitly insert them.)
|
|
|
724 'shift-jis
|
|
|
725 Shift-JIS (a Japanese encoding commonly used in PC operating systems).
|
|
396
|
726 'ucs-4
|
|
|
727 ISO 10646 UCS-4 encoding.
|
|
|
728 'utf-8
|
|
|
729 ISO 10646 UTF-8 encoding.
|
|
259
|
730 'iso2022
|
|
|
731 Any ISO2022-compliant encoding. Among other things, this includes
|
|
|
732 JIS (the Japanese encoding commonly used for e-mail), EUC (the
|
|
|
733 standard Unix encoding for Japanese and other languages), and
|
|
|
734 Compound Text (the encoding used in X11). You can specify more
|
|
371
|
735 specific information about the conversion with the FLAGS argument.
|
|
259
|
736 'big5
|
|
|
737 Big5 (the encoding commonly used for Taiwanese).
|
|
|
738 'ccl
|
|
|
739 The conversion is performed using a user-written pseudo-code
|
|
|
740 program. CCL (Code Conversion Language) is the name of this
|
|
|
741 pseudo-code.
|
|
|
742 'internal
|
|
|
743 Write out or read in the raw contents of the memory representing
|
|
|
744 the buffer's text. This is primarily useful for debugging
|
|
|
745 purposes, and is only enabled when XEmacs has been compiled with
|
|
|
746 DEBUG_XEMACS defined (via the --debug configure option).
|
|
|
747 WARNING: Reading in a file using 'internal conversion can result
|
|
|
748 in an internal inconsistency in the memory representing a
|
|
|
749 buffer's text, which will produce unpredictable results and may
|
|
|
750 cause XEmacs to crash. Under normal circumstances you should
|
|
|
751 never use 'internal conversion.
|
|
|
752
|
|
|
753 DOC-STRING is a string describing the coding system.
|
|
|
754
|
|
|
755 PROPS is a property list, describing the specific nature of the
|
|
|
756 character set. Recognized properties are:
|
|
|
757
|
|
|
758 'mnemonic
|
|
|
759 String to be displayed in the modeline when this coding system is
|
|
|
760 active.
|
|
|
761
|
|
|
762 'eol-type
|
|
|
763 End-of-line conversion to be used. It should be one of
|
|
|
764
|
|
|
765 nil
|
|
|
766 Automatically detect the end-of-line type (LF, CRLF,
|
|
|
767 or CR). Also generate subsidiary coding systems named
|
|
|
768 `NAME-unix', `NAME-dos', and `NAME-mac', that are
|
|
|
769 identical to this coding system but have an EOL-TYPE
|
|
|
770 value of 'lf, 'crlf, and 'cr, respectively.
|
|
|
771 'lf
|
|
|
772 The end of a line is marked externally using ASCII LF.
|
|
|
773 Since this is also the way that XEmacs represents an
|
|
|
774 end-of-line internally, specifying this option results
|
|
|
775 in no end-of-line conversion. This is the standard
|
|
|
776 format for Unix text files.
|
|
|
777 'crlf
|
|
|
778 The end of a line is marked externally using ASCII
|
|
|
779 CRLF. This is the standard format for MS-DOS text
|
|
|
780 files.
|
|
|
781 'cr
|
|
|
782 The end of a line is marked externally using ASCII CR.
|
|
|
783 This is the standard format for Macintosh text files.
|
|
|
784 t
|
|
|
785 Automatically detect the end-of-line type but do not
|
|
|
786 generate subsidiary coding systems. (This value is
|
|
|
787 converted to nil when stored internally, and
|
|
|
788 `coding-system-property' will return nil.)
|
|
|
789
|
|
|
790 'post-read-conversion
|
|
|
791 Function called after a file has been read in, to perform the
|
|
|
792 decoding. Called with two arguments, BEG and END, denoting
|
|
|
793 a region of the current buffer to be decoded.
|
|
|
794
|
|
|
795 'pre-write-conversion
|
|
|
796 Function called before a file is written out, to perform the
|
|
|
797 encoding. Called with two arguments, BEG and END, denoting
|
|
|
798 a region of the current buffer to be encoded.
|
|
|
799
|
|
|
800
|
|
|
801 The following additional properties are recognized if TYPE is 'iso2022:
|
|
|
802
|
|
|
803 'charset-g0
|
|
|
804 'charset-g1
|
|
|
805 'charset-g2
|
|
|
806 'charset-g3
|
|
|
807 The character set initially designated to the G0 - G3 registers.
|
|
|
808 The value should be one of
|
|
|
809
|
|
|
810 -- A charset object (designate that character set)
|
|
|
811 -- nil (do not ever use this register)
|
|
|
812 -- t (no character set is initially designated to
|
|
|
813 the register, but may be later on; this automatically
|
|
|
814 sets the corresponding `force-g*-on-output' property)
|
|
|
815
|
|
|
816 'force-g0-on-output
|
|
|
817 'force-g1-on-output
|
|
|
818 'force-g2-on-output
|
|
|
819 'force-g2-on-output
|
|
|
820 If non-nil, send an explicit designation sequence on output before
|
|
|
821 using the specified register.
|
|
|
822
|
|
|
823 'short
|
|
|
824 If non-nil, use the short forms "ESC $ @", "ESC $ A", and
|
|
|
825 "ESC $ B" on output in place of the full designation sequences
|
|
|
826 "ESC $ ( @", "ESC $ ( A", and "ESC $ ( B".
|
|
|
827
|
|
|
828 'no-ascii-eol
|
|
|
829 If non-nil, don't designate ASCII to G0 at each end of line on output.
|
|
|
830 Setting this to non-nil also suppresses other state-resetting that
|
|
|
831 normally happens at the end of a line.
|
|
|
832
|
|
|
833 'no-ascii-cntl
|
|
|
834 If non-nil, don't designate ASCII to G0 before control chars on output.
|
|
|
835
|
|
|
836 'seven
|
|
|
837 If non-nil, use 7-bit environment on output. Otherwise, use 8-bit
|
|
|
838 environment.
|
|
|
839
|
|
|
840 'lock-shift
|
|
|
841 If non-nil, use locking-shift (SO/SI) instead of single-shift
|
|
|
842 or designation by escape sequence.
|
|
|
843
|
|
|
844 'no-iso6429
|
|
|
845 If non-nil, don't use ISO6429's direction specification.
|
|
|
846
|
|
|
847 'escape-quoted
|
|
|
848 If non-nil, literal control characters that are the same as
|
|
|
849 the beginning of a recognized ISO2022 or ISO6429 escape sequence
|
|
|
850 (in particular, ESC (0x1B), SO (0x0E), SI (0x0F), SS2 (0x8E),
|
|
|
851 SS3 (0x8F), and CSI (0x9B)) are "quoted" with an escape character
|
|
|
852 so that they can be properly distinguished from an escape sequence.
|
|
|
853 (Note that doing this results in a non-portable encoding.) This
|
|
|
854 encoding flag is used for byte-compiled files. Note that ESC
|
|
|
855 is a good choice for a quoting character because there are no
|
|
|
856 escape sequences whose second byte is a character from the Control-0
|
|
|
857 or Control-1 character sets; this is explicitly disallowed by the
|
|
|
858 ISO2022 standard.
|
|
|
859
|
|
|
860 'input-charset-conversion
|
|
|
861 A list of conversion specifications, specifying conversion of
|
|
|
862 characters in one charset to another when decoding is performed.
|
|
|
863 Each specification is a list of two elements: the source charset,
|
|
|
864 and the destination charset.
|
|
|
865
|
|
|
866 'output-charset-conversion
|
|
|
867 A list of conversion specifications, specifying conversion of
|
|
|
868 characters in one charset to another when encoding is performed.
|
|
|
869 The form of each specification is the same as for
|
|
|
870 'input-charset-conversion.
|
|
|
871
|
|
|
872
|
|
|
873 The following additional properties are recognized (and required)
|
|
|
874 if TYPE is 'ccl:
|
|
|
875
|
|
|
876 'decode
|
|
|
877 CCL program used for decoding (converting to internal format).
|
|
|
878
|
|
|
879 'encode
|
|
|
880 CCL program used for encoding (converting to external format).
|
|
|
881 */
|
|
|
882 (name, type, doc_string, props))
|
|
|
883 {
|
|
396
|
884 Lisp_Coding_System *codesys;
|
|
259
|
885 Lisp_Object rest, key, value;
|
|
|
886 enum coding_system_type ty;
|
|
|
887 int need_to_setup_eol_systems = 1;
|
|
|
888
|
|
|
889 /* Convert type to constant */
|
|
|
890 if (NILP (type) || EQ (type, Qundecided))
|
|
|
891 { ty = CODESYS_AUTODETECT; }
|
|
|
892 #ifdef MULE
|
|
|
893 else if (EQ (type, Qshift_jis)) { ty = CODESYS_SHIFT_JIS; }
|
|
|
894 else if (EQ (type, Qiso2022)) { ty = CODESYS_ISO2022; }
|
|
|
895 else if (EQ (type, Qbig5)) { ty = CODESYS_BIG5; }
|
|
396
|
896 else if (EQ (type, Qucs4)) { ty = CODESYS_UCS4; }
|
|
|
897 else if (EQ (type, Qutf8)) { ty = CODESYS_UTF8; }
|
|
259
|
898 else if (EQ (type, Qccl)) { ty = CODESYS_CCL; }
|
|
|
899 #endif
|
|
|
900 else if (EQ (type, Qno_conversion)) { ty = CODESYS_NO_CONVERSION; }
|
|
|
901 #ifdef DEBUG_XEMACS
|
|
|
902 else if (EQ (type, Qinternal)) { ty = CODESYS_INTERNAL; }
|
|
|
903 #endif
|
|
|
904 else
|
|
|
905 signal_simple_error ("Invalid coding system type", type);
|
|
|
906
|
|
|
907 CHECK_SYMBOL (name);
|
|
|
908
|
|
|
909 codesys = allocate_coding_system (ty, name);
|
|
|
910
|
|
|
911 if (NILP (doc_string))
|
|
|
912 doc_string = build_string ("");
|
|
|
913 else
|
|
|
914 CHECK_STRING (doc_string);
|
|
|
915 CODING_SYSTEM_DOC_STRING (codesys) = doc_string;
|
|
|
916
|
|
|
917 EXTERNAL_PROPERTY_LIST_LOOP (rest, key, value, props)
|
|
|
918 {
|
|
|
919 if (EQ (key, Qmnemonic))
|
|
|
920 {
|
|
|
921 if (!NILP (value))
|
|
|
922 CHECK_STRING (value);
|
|
|
923 CODING_SYSTEM_MNEMONIC (codesys) = value;
|
|
|
924 }
|
|
|
925
|
|
|
926 else if (EQ (key, Qeol_type))
|
|
|
927 {
|
|
|
928 need_to_setup_eol_systems = NILP (value);
|
|
|
929 if (EQ (value, Qt))
|
|
|
930 value = Qnil;
|
|
|
931 CODING_SYSTEM_EOL_TYPE (codesys) = symbol_to_eol_type (value);
|
|
|
932 }
|
|
|
933
|
|
|
934 else if (EQ (key, Qpost_read_conversion)) CODING_SYSTEM_POST_READ_CONVERSION (codesys) = value;
|
|
|
935 else if (EQ (key, Qpre_write_conversion)) CODING_SYSTEM_PRE_WRITE_CONVERSION (codesys) = value;
|
|
|
936 #ifdef MULE
|
|
|
937 else if (ty == CODESYS_ISO2022)
|
|
|
938 {
|
|
|
939 #define FROB_INITIAL_CHARSET(charset_num) \
|
|
|
940 CODING_SYSTEM_ISO2022_INITIAL_CHARSET (codesys, charset_num) = \
|
|
|
941 ((EQ (value, Qt) || EQ (value, Qnil)) ? value : Fget_charset (value))
|
|
|
942
|
|
|
943 if (EQ (key, Qcharset_g0)) FROB_INITIAL_CHARSET (0);
|
|
|
944 else if (EQ (key, Qcharset_g1)) FROB_INITIAL_CHARSET (1);
|
|
|
945 else if (EQ (key, Qcharset_g2)) FROB_INITIAL_CHARSET (2);
|
|
|
946 else if (EQ (key, Qcharset_g3)) FROB_INITIAL_CHARSET (3);
|
|
|
947
|
|
|
948 #define FROB_FORCE_CHARSET(charset_num) \
|
|
|
949 CODING_SYSTEM_ISO2022_FORCE_CHARSET_ON_OUTPUT (codesys, charset_num) = !NILP (value)
|
|
|
950
|
|
|
951 else if (EQ (key, Qforce_g0_on_output)) FROB_FORCE_CHARSET (0);
|
|
|
952 else if (EQ (key, Qforce_g1_on_output)) FROB_FORCE_CHARSET (1);
|
|
|
953 else if (EQ (key, Qforce_g2_on_output)) FROB_FORCE_CHARSET (2);
|
|
|
954 else if (EQ (key, Qforce_g3_on_output)) FROB_FORCE_CHARSET (3);
|
|
|
955
|
|
|
956 #define FROB_BOOLEAN_PROPERTY(prop) \
|
|
|
957 CODING_SYSTEM_ISO2022_##prop (codesys) = !NILP (value)
|
|
|
958
|
|
|
959 else if (EQ (key, Qshort)) FROB_BOOLEAN_PROPERTY (SHORT);
|
|
|
960 else if (EQ (key, Qno_ascii_eol)) FROB_BOOLEAN_PROPERTY (NO_ASCII_EOL);
|
|
|
961 else if (EQ (key, Qno_ascii_cntl)) FROB_BOOLEAN_PROPERTY (NO_ASCII_CNTL);
|
|
|
962 else if (EQ (key, Qseven)) FROB_BOOLEAN_PROPERTY (SEVEN);
|
|
|
963 else if (EQ (key, Qlock_shift)) FROB_BOOLEAN_PROPERTY (LOCK_SHIFT);
|
|
|
964 else if (EQ (key, Qno_iso6429)) FROB_BOOLEAN_PROPERTY (NO_ISO6429);
|
|
|
965 else if (EQ (key, Qescape_quoted)) FROB_BOOLEAN_PROPERTY (ESCAPE_QUOTED);
|
|
|
966
|
|
|
967 else if (EQ (key, Qinput_charset_conversion))
|
|
|
968 {
|
|
|
969 codesys->iso2022.input_conv =
|
|
|
970 Dynarr_new (charset_conversion_spec);
|
|
|
971 parse_charset_conversion_specs (codesys->iso2022.input_conv,
|
|
|
972 value);
|
|
|
973 }
|
|
|
974 else if (EQ (key, Qoutput_charset_conversion))
|
|
|
975 {
|
|
|
976 codesys->iso2022.output_conv =
|
|
|
977 Dynarr_new (charset_conversion_spec);
|
|
|
978 parse_charset_conversion_specs (codesys->iso2022.output_conv,
|
|
|
979 value);
|
|
|
980 }
|
|
|
981 else
|
|
|
982 signal_simple_error ("Unrecognized property", key);
|
|
|
983 }
|
|
|
984 else if (EQ (type, Qccl))
|
|
|
985 {
|
|
|
986 if (EQ (key, Qdecode))
|
|
|
987 {
|
|
|
988 CHECK_VECTOR (value);
|
|
|
989 CODING_SYSTEM_CCL_DECODE (codesys) = value;
|
|
|
990 }
|
|
|
991 else if (EQ (key, Qencode))
|
|
|
992 {
|
|
|
993 CHECK_VECTOR (value);
|
|
|
994 CODING_SYSTEM_CCL_ENCODE (codesys) = value;
|
|
|
995 }
|
|
|
996 else
|
|
|
997 signal_simple_error ("Unrecognized property", key);
|
|
|
998 }
|
|
|
999 #endif /* MULE */
|
|
|
1000 else
|
|
|
1001 signal_simple_error ("Unrecognized property", key);
|
|
|
1002 }
|
|
|
1003
|
|
|
1004 if (need_to_setup_eol_systems)
|
|
|
1005 setup_eol_coding_systems (codesys);
|
|
|
1006
|
|
|
1007 {
|
|
|
1008 Lisp_Object codesys_obj;
|
|
|
1009 XSETCODING_SYSTEM (codesys_obj, codesys);
|
|
380
|
1010 Fputhash (name, codesys_obj, Vcoding_system_hash_table);
|
|
259
|
1011 return codesys_obj;
|
|
|
1012 }
|
|
|
1013 }
|
|
|
1014
|
|
|
1015 DEFUN ("copy-coding-system", Fcopy_coding_system, 2, 2, 0, /*
|
|
|
1016 Copy OLD-CODING-SYSTEM to NEW-NAME.
|
|
|
1017 If NEW-NAME does not name an existing coding system, a new one will
|
|
|
1018 be created.
|
|
|
1019 */
|
|
|
1020 (old_coding_system, new_name))
|
|
|
1021 {
|
|
|
1022 Lisp_Object new_coding_system;
|
|
|
1023 old_coding_system = Fget_coding_system (old_coding_system);
|
|
|
1024 new_coding_system = Ffind_coding_system (new_name);
|
|
|
1025 if (NILP (new_coding_system))
|
|
|
1026 {
|
|
|
1027 XSETCODING_SYSTEM (new_coding_system,
|
|
|
1028 allocate_coding_system
|
|
|
1029 (XCODING_SYSTEM_TYPE (old_coding_system),
|
|
|
1030 new_name));
|
|
380
|
1031 Fputhash (new_name, new_coding_system, Vcoding_system_hash_table);
|
|
259
|
1032 }
|
|
|
1033
|
|
|
1034 {
|
|
396
|
1035 Lisp_Coding_System *to = XCODING_SYSTEM (new_coding_system);
|
|
|
1036 Lisp_Coding_System *from = XCODING_SYSTEM (old_coding_system);
|
|
259
|
1037 memcpy (((char *) to ) + sizeof (to->header),
|
|
|
1038 ((char *) from) + sizeof (from->header),
|
|
|
1039 sizeof (*from) - sizeof (from->header));
|
|
|
1040 to->name = new_name;
|
|
|
1041 }
|
|
|
1042 return new_coding_system;
|
|
|
1043 }
|
|
|
1044
|
|
398
|
1045 DEFUN ("coding-system-canonical-name-p", Fcoding_system_canonical_name_p, 1, 1, 0, /*
|
|
|
1046 Return t if OBJECT names a coding system, and is not a coding system alias.
|
|
|
1047 */
|
|
|
1048 (object))
|
|
|
1049 {
|
|
|
1050 return CODING_SYSTEMP (Fgethash (object, Vcoding_system_hash_table, Qnil))
|
|
|
1051 ? Qt : Qnil;
|
|
|
1052 }
|
|
|
1053
|
|
|
1054 DEFUN ("coding-system-alias-p", Fcoding_system_alias_p, 1, 1, 0, /*
|
|
|
1055 Return t if OBJECT is a coding system alias.
|
|
|
1056 All coding system aliases are created by `define-coding-system-alias'.
|
|
|
1057 */
|
|
|
1058 (object))
|
|
|
1059 {
|
|
|
1060 return SYMBOLP (Fgethash (object, Vcoding_system_hash_table, Qzero))
|
|
|
1061 ? Qt : Qnil;
|
|
|
1062 }
|
|
|
1063
|
|
|
1064 DEFUN ("coding-system-aliasee", Fcoding_system_aliasee, 1, 1, 0, /*
|
|
|
1065 Return the coding-system symbol for which symbol ALIAS is an alias.
|
|
|
1066 */
|
|
|
1067 (alias))
|
|
|
1068 {
|
|
|
1069 Lisp_Object aliasee = Fgethash (alias, Vcoding_system_hash_table, Qnil);
|
|
|
1070 if (SYMBOLP (aliasee))
|
|
|
1071 return aliasee;
|
|
|
1072 else
|
|
|
1073 signal_simple_error ("Symbol is not a coding system alias", alias);
|
|
|
1074 }
|
|
|
1075
|
|
259
|
1076 static Lisp_Object
|
|
398
|
1077 append_suffix_to_symbol (Lisp_Object symbol, const char *ascii_string)
|
|
|
1078 {
|
|
|
1079 return Fintern (concat2 (Fsymbol_name (symbol), build_string (ascii_string)),
|
|
|
1080 Qnil);
|
|
|
1081 }
|
|
|
1082
|
|
|
1083 /* A maphash function, for removing dangling coding system aliases. */
|
|
|
1084 static int
|
|
|
1085 dangling_coding_system_alias_p (Lisp_Object alias,
|
|
|
1086 Lisp_Object aliasee,
|
|
|
1087 void *dangling_aliases)
|
|
|
1088 {
|
|
|
1089 if (SYMBOLP (aliasee)
|
|
|
1090 && NILP (Fgethash (aliasee, Vcoding_system_hash_table, Qnil)))
|
|
|
1091 {
|
|
|
1092 (*(int *) dangling_aliases)++;
|
|
|
1093 return 1;
|
|
|
1094 }
|
|
|
1095 else
|
|
|
1096 return 0;
|
|
|
1097 }
|
|
|
1098
|
|
|
1099 DEFUN ("define-coding-system-alias", Fdefine_coding_system_alias, 2, 2, 0, /*
|
|
|
1100 Define symbol ALIAS as an alias for coding system ALIASEE.
|
|
|
1101
|
|
|
1102 You can use this function to redefine an alias that has already been defined,
|
|
|
1103 but you cannot redefine a name which is the canonical name for a coding system.
|
|
|
1104 \(a canonical name of a coding system is what is returned when you call
|
|
|
1105 `coding-system-name' on a coding system).
|
|
|
1106
|
|
|
1107 ALIASEE itself can be an alias, which allows you to define nested aliases.
|
|
|
1108
|
|
|
1109 You are forbidden, however, from creating alias loops or `dangling' aliases.
|
|
|
1110 These will be detected, and an error will be signaled if you attempt to do so.
|
|
|
1111
|
|
|
1112 If ALIASEE is nil, then ALIAS will simply be undefined.
|
|
|
1113
|
|
|
1114 See also `coding-system-alias-p', `coding-system-aliasee',
|
|
|
1115 and `coding-system-canonical-name-p'.
|
|
|
1116 */
|
|
|
1117 (alias, aliasee))
|
|
|
1118 {
|
|
|
1119 Lisp_Object real_coding_system, probe;
|
|
|
1120
|
|
|
1121 CHECK_SYMBOL (alias);
|
|
|
1122
|
|
|
1123 if (!NILP (Fcoding_system_canonical_name_p (alias)))
|
|
|
1124 signal_simple_error
|
|
|
1125 ("Symbol is the canonical name of a coding system and cannot be redefined",
|
|
|
1126 alias);
|
|
|
1127
|
|
|
1128 if (NILP (aliasee))
|
|
|
1129 {
|
|
|
1130 Lisp_Object subsidiary_unix = append_suffix_to_symbol (alias, "-unix");
|
|
|
1131 Lisp_Object subsidiary_dos = append_suffix_to_symbol (alias, "-dos");
|
|
|
1132 Lisp_Object subsidiary_mac = append_suffix_to_symbol (alias, "-mac");
|
|
|
1133
|
|
|
1134 Fremhash (alias, Vcoding_system_hash_table);
|
|
|
1135
|
|
|
1136 /* Undefine subsidiary aliases,
|
|
|
1137 presumably created by a previous call to this function */
|
|
|
1138 if (! NILP (Fcoding_system_alias_p (subsidiary_unix)) &&
|
|
|
1139 ! NILP (Fcoding_system_alias_p (subsidiary_dos)) &&
|
|
|
1140 ! NILP (Fcoding_system_alias_p (subsidiary_mac)))
|
|
|
1141 {
|
|
|
1142 Fdefine_coding_system_alias (subsidiary_unix, Qnil);
|
|
|
1143 Fdefine_coding_system_alias (subsidiary_dos, Qnil);
|
|
|
1144 Fdefine_coding_system_alias (subsidiary_mac, Qnil);
|
|
|
1145 }
|
|
|
1146
|
|
|
1147 /* Undefine dangling coding system aliases. */
|
|
|
1148 {
|
|
|
1149 int dangling_aliases;
|
|
|
1150
|
|
|
1151 do {
|
|
|
1152 dangling_aliases = 0;
|
|
|
1153 elisp_map_remhash (dangling_coding_system_alias_p,
|
|
|
1154 Vcoding_system_hash_table,
|
|
|
1155 &dangling_aliases);
|
|
|
1156 } while (dangling_aliases > 0);
|
|
|
1157 }
|
|
|
1158
|
|
|
1159 return Qnil;
|
|
|
1160 }
|
|
|
1161
|
|
|
1162 if (CODING_SYSTEMP (aliasee))
|
|
|
1163 aliasee = XCODING_SYSTEM_NAME (aliasee);
|
|
|
1164
|
|
|
1165 /* Checks that aliasee names a coding-system */
|
|
|
1166 real_coding_system = Fget_coding_system (aliasee);
|
|
|
1167
|
|
|
1168 /* Check for coding system alias loops */
|
|
|
1169 if (EQ (alias, aliasee))
|
|
|
1170 alias_loop: signal_simple_error_2
|
|
|
1171 ("Attempt to create a coding system alias loop", alias, aliasee);
|
|
|
1172
|
|
|
1173 for (probe = aliasee;
|
|
|
1174 SYMBOLP (probe);
|
|
|
1175 probe = Fgethash (probe, Vcoding_system_hash_table, Qzero))
|
|
|
1176 {
|
|
|
1177 if (EQ (probe, alias))
|
|
|
1178 goto alias_loop;
|
|
|
1179 }
|
|
|
1180
|
|
|
1181 Fputhash (alias, aliasee, Vcoding_system_hash_table);
|
|
|
1182
|
|
|
1183 /* Set up aliases for subsidiaries.
|
|
|
1184 #### There must be a better way to handle subsidiary coding systems. */
|
|
|
1185 {
|
|
|
1186 static const char *suffixes[] = { "-unix", "-dos", "-mac" };
|
|
|
1187 int i;
|
|
|
1188 for (i = 0; i < countof (suffixes); i++)
|
|
|
1189 {
|
|
|
1190 Lisp_Object alias_subsidiary =
|
|
|
1191 append_suffix_to_symbol (alias, suffixes[i]);
|
|
|
1192 Lisp_Object aliasee_subsidiary =
|
|
|
1193 append_suffix_to_symbol (aliasee, suffixes[i]);
|
|
|
1194
|
|
|
1195 if (! NILP (Ffind_coding_system (aliasee_subsidiary)))
|
|
|
1196 Fdefine_coding_system_alias (alias_subsidiary, aliasee_subsidiary);
|
|
|
1197 }
|
|
|
1198 }
|
|
|
1199 /* FSF return value is a vector of [ALIAS-unix ALIAS-dos ALIAS-mac],
|
|
|
1200 but it doesn't look intentional, so I'd rather return something
|
|
|
1201 meaningful or nothing at all. */
|
|
|
1202 return Qnil;
|
|
|
1203 }
|
|
|
1204
|
|
|
1205 static Lisp_Object
|
|
|
1206 subsidiary_coding_system (Lisp_Object coding_system, eol_type_t type)
|
|
259
|
1207 {
|
|
396
|
1208 Lisp_Coding_System *cs = XCODING_SYSTEM (coding_system);
|
|
259
|
1209 Lisp_Object new_coding_system;
|
|
|
1210
|
|
|
1211 if (CODING_SYSTEM_EOL_TYPE (cs) != EOL_AUTODETECT)
|
|
|
1212 return coding_system;
|
|
|
1213
|
|
|
1214 switch (type)
|
|
|
1215 {
|
|
|
1216 case EOL_AUTODETECT: return coding_system;
|
|
|
1217 case EOL_LF: new_coding_system = CODING_SYSTEM_EOL_LF (cs); break;
|
|
|
1218 case EOL_CR: new_coding_system = CODING_SYSTEM_EOL_CR (cs); break;
|
|
|
1219 case EOL_CRLF: new_coding_system = CODING_SYSTEM_EOL_CRLF (cs); break;
|
|
|
1220 default: abort ();
|
|
|
1221 }
|
|
|
1222
|
|
|
1223 return NILP (new_coding_system) ? coding_system : new_coding_system;
|
|
|
1224 }
|
|
|
1225
|
|
|
1226 DEFUN ("subsidiary-coding-system", Fsubsidiary_coding_system, 2, 2, 0, /*
|
|
|
1227 Return the subsidiary coding system of CODING-SYSTEM with eol type EOL-TYPE.
|
|
|
1228 */
|
|
|
1229 (coding_system, eol_type))
|
|
|
1230 {
|
|
|
1231 coding_system = Fget_coding_system (coding_system);
|
|
|
1232
|
|
|
1233 return subsidiary_coding_system (coding_system,
|
|
|
1234 symbol_to_eol_type (eol_type));
|
|
|
1235 }
|
|
|
1236
|
|
|
1237
|
|
|
1238 /************************************************************************/
|
|
|
1239 /* Coding system accessors */
|
|
|
1240 /************************************************************************/
|
|
|
1241
|
|
|
1242 DEFUN ("coding-system-doc-string", Fcoding_system_doc_string, 1, 1, 0, /*
|
|
|
1243 Return the doc string for CODING-SYSTEM.
|
|
|
1244 */
|
|
|
1245 (coding_system))
|
|
|
1246 {
|
|
|
1247 coding_system = Fget_coding_system (coding_system);
|
|
|
1248 return XCODING_SYSTEM_DOC_STRING (coding_system);
|
|
|
1249 }
|
|
|
1250
|
|
|
1251 DEFUN ("coding-system-type", Fcoding_system_type, 1, 1, 0, /*
|
|
|
1252 Return the type of CODING-SYSTEM.
|
|
|
1253 */
|
|
|
1254 (coding_system))
|
|
|
1255 {
|
|
|
1256 switch (XCODING_SYSTEM_TYPE (Fget_coding_system (coding_system)))
|
|
|
1257 {
|
|
380
|
1258 default: abort ();
|
|
259
|
1259 case CODESYS_AUTODETECT: return Qundecided;
|
|
|
1260 #ifdef MULE
|
|
|
1261 case CODESYS_SHIFT_JIS: return Qshift_jis;
|
|
|
1262 case CODESYS_ISO2022: return Qiso2022;
|
|
|
1263 case CODESYS_BIG5: return Qbig5;
|
|
396
|
1264 case CODESYS_UCS4: return Qucs4;
|
|
|
1265 case CODESYS_UTF8: return Qutf8;
|
|
259
|
1266 case CODESYS_CCL: return Qccl;
|
|
|
1267 #endif
|
|
|
1268 case CODESYS_NO_CONVERSION: return Qno_conversion;
|
|
|
1269 #ifdef DEBUG_XEMACS
|
|
|
1270 case CODESYS_INTERNAL: return Qinternal;
|
|
|
1271 #endif
|
|
|
1272 }
|
|
|
1273 }
|
|
|
1274
|
|
|
1275 #ifdef MULE
|
|
|
1276 static
|
|
|
1277 Lisp_Object coding_system_charset (Lisp_Object coding_system, int gnum)
|
|
|
1278 {
|
|
|
1279 Lisp_Object cs
|
|
|
1280 = XCODING_SYSTEM_ISO2022_INITIAL_CHARSET (coding_system, gnum);
|
|
272
|
1281
|
|
|
1282 return CHARSETP (cs) ? XCHARSET_NAME (cs) : Qnil;
|
|
259
|
1283 }
|
|
|
1284
|
|
|
1285 DEFUN ("coding-system-charset", Fcoding_system_charset, 2, 2, 0, /*
|
|
|
1286 Return initial charset of CODING-SYSTEM designated to GNUM.
|
|
|
1287 GNUM allows 0 .. 3.
|
|
|
1288 */
|
|
|
1289 (coding_system, gnum))
|
|
|
1290 {
|
|
|
1291 coding_system = Fget_coding_system (coding_system);
|
|
|
1292 CHECK_INT (gnum);
|
|
|
1293
|
|
272
|
1294 return coding_system_charset (coding_system, XINT (gnum));
|
|
259
|
1295 }
|
|
272
|
1296 #endif /* MULE */
|
|
259
|
1297
|
|
|
1298 DEFUN ("coding-system-property", Fcoding_system_property, 2, 2, 0, /*
|
|
|
1299 Return the PROP property of CODING-SYSTEM.
|
|
|
1300 */
|
|
|
1301 (coding_system, prop))
|
|
|
1302 {
|
|
|
1303 int i, ok = 0;
|
|
|
1304 enum coding_system_type type;
|
|
|
1305
|
|
|
1306 coding_system = Fget_coding_system (coding_system);
|
|
|
1307 CHECK_SYMBOL (prop);
|
|
|
1308 type = XCODING_SYSTEM_TYPE (coding_system);
|
|
|
1309
|
|
|
1310 for (i = 0; !ok && i < Dynarr_length (the_codesys_prop_dynarr); i++)
|
|
|
1311 if (EQ (Dynarr_at (the_codesys_prop_dynarr, i).sym, prop))
|
|
|
1312 {
|
|
|
1313 ok = 1;
|
|
|
1314 switch (Dynarr_at (the_codesys_prop_dynarr, i).prop_type)
|
|
|
1315 {
|
|
|
1316 case CODESYS_PROP_ALL_OK:
|
|
|
1317 break;
|
|
|
1318 #ifdef MULE
|
|
|
1319 case CODESYS_PROP_ISO2022:
|
|
|
1320 if (type != CODESYS_ISO2022)
|
|
|
1321 signal_simple_error
|
|
|
1322 ("Property only valid in ISO2022 coding systems",
|
|
|
1323 prop);
|
|
|
1324 break;
|
|
|
1325
|
|
|
1326 case CODESYS_PROP_CCL:
|
|
|
1327 if (type != CODESYS_CCL)
|
|
|
1328 signal_simple_error
|
|
|
1329 ("Property only valid in CCL coding systems",
|
|
|
1330 prop);
|
|
|
1331 break;
|
|
|
1332 #endif /* MULE */
|
|
|
1333 default:
|
|
|
1334 abort ();
|
|
|
1335 }
|
|
|
1336 }
|
|
|
1337
|
|
|
1338 if (!ok)
|
|
|
1339 signal_simple_error ("Unrecognized property", prop);
|
|
|
1340
|
|
|
1341 if (EQ (prop, Qname))
|
|
|
1342 return XCODING_SYSTEM_NAME (coding_system);
|
|
|
1343 else if (EQ (prop, Qtype))
|
|
|
1344 return Fcoding_system_type (coding_system);
|
|
|
1345 else if (EQ (prop, Qdoc_string))
|
|
|
1346 return XCODING_SYSTEM_DOC_STRING (coding_system);
|
|
|
1347 else if (EQ (prop, Qmnemonic))
|
|
|
1348 return XCODING_SYSTEM_MNEMONIC (coding_system);
|
|
|
1349 else if (EQ (prop, Qeol_type))
|
|
|
1350 return eol_type_to_symbol (XCODING_SYSTEM_EOL_TYPE (coding_system));
|
|
|
1351 else if (EQ (prop, Qeol_lf))
|
|
|
1352 return XCODING_SYSTEM_EOL_LF (coding_system);
|
|
|
1353 else if (EQ (prop, Qeol_crlf))
|
|
|
1354 return XCODING_SYSTEM_EOL_CRLF (coding_system);
|
|
|
1355 else if (EQ (prop, Qeol_cr))
|
|
|
1356 return XCODING_SYSTEM_EOL_CR (coding_system);
|
|
|
1357 else if (EQ (prop, Qpost_read_conversion))
|
|
|
1358 return XCODING_SYSTEM_POST_READ_CONVERSION (coding_system);
|
|
|
1359 else if (EQ (prop, Qpre_write_conversion))
|
|
|
1360 return XCODING_SYSTEM_PRE_WRITE_CONVERSION (coding_system);
|
|
|
1361 #ifdef MULE
|
|
|
1362 else if (type == CODESYS_ISO2022)
|
|
|
1363 {
|
|
|
1364 if (EQ (prop, Qcharset_g0))
|
|
|
1365 return coding_system_charset (coding_system, 0);
|
|
|
1366 else if (EQ (prop, Qcharset_g1))
|
|
|
1367 return coding_system_charset (coding_system, 1);
|
|
|
1368 else if (EQ (prop, Qcharset_g2))
|
|
|
1369 return coding_system_charset (coding_system, 2);
|
|
|
1370 else if (EQ (prop, Qcharset_g3))
|
|
|
1371 return coding_system_charset (coding_system, 3);
|
|
|
1372
|
|
|
1373 #define FORCE_CHARSET(charset_num) \
|
|
|
1374 (XCODING_SYSTEM_ISO2022_FORCE_CHARSET_ON_OUTPUT \
|
|
|
1375 (coding_system, charset_num) ? Qt : Qnil)
|
|
|
1376
|
|
|
1377 else if (EQ (prop, Qforce_g0_on_output)) return FORCE_CHARSET (0);
|
|
|
1378 else if (EQ (prop, Qforce_g1_on_output)) return FORCE_CHARSET (1);
|
|
|
1379 else if (EQ (prop, Qforce_g2_on_output)) return FORCE_CHARSET (2);
|
|
|
1380 else if (EQ (prop, Qforce_g3_on_output)) return FORCE_CHARSET (3);
|
|
|
1381
|
|
|
1382 #define LISP_BOOLEAN(prop) \
|
|
|
1383 (XCODING_SYSTEM_ISO2022_##prop (coding_system) ? Qt : Qnil)
|
|
|
1384
|
|
|
1385 else if (EQ (prop, Qshort)) return LISP_BOOLEAN (SHORT);
|
|
|
1386 else if (EQ (prop, Qno_ascii_eol)) return LISP_BOOLEAN (NO_ASCII_EOL);
|
|
|
1387 else if (EQ (prop, Qno_ascii_cntl)) return LISP_BOOLEAN (NO_ASCII_CNTL);
|
|
|
1388 else if (EQ (prop, Qseven)) return LISP_BOOLEAN (SEVEN);
|
|
|
1389 else if (EQ (prop, Qlock_shift)) return LISP_BOOLEAN (LOCK_SHIFT);
|
|
|
1390 else if (EQ (prop, Qno_iso6429)) return LISP_BOOLEAN (NO_ISO6429);
|
|
|
1391 else if (EQ (prop, Qescape_quoted)) return LISP_BOOLEAN (ESCAPE_QUOTED);
|
|
|
1392
|
|
|
1393 else if (EQ (prop, Qinput_charset_conversion))
|
|
|
1394 return
|
|
|
1395 unparse_charset_conversion_specs
|
|
|
1396 (XCODING_SYSTEM (coding_system)->iso2022.input_conv);
|
|
|
1397 else if (EQ (prop, Qoutput_charset_conversion))
|
|
|
1398 return
|
|
|
1399 unparse_charset_conversion_specs
|
|
|
1400 (XCODING_SYSTEM (coding_system)->iso2022.output_conv);
|
|
|
1401 else
|
|
|
1402 abort ();
|
|
|
1403 }
|
|
|
1404 else if (type == CODESYS_CCL)
|
|
|
1405 {
|
|
|
1406 if (EQ (prop, Qdecode))
|
|
|
1407 return XCODING_SYSTEM_CCL_DECODE (coding_system);
|
|
|
1408 else if (EQ (prop, Qencode))
|
|
|
1409 return XCODING_SYSTEM_CCL_ENCODE (coding_system);
|
|
|
1410 else
|
|
|
1411 abort ();
|
|
|
1412 }
|
|
|
1413 #endif /* MULE */
|
|
|
1414 else
|
|
|
1415 abort ();
|
|
|
1416
|
|
|
1417 return Qnil; /* not reached */
|
|
|
1418 }
|
|
|
1419
|
|
|
1420
|
|
|
1421 /************************************************************************/
|
|
|
1422 /* Coding category functions */
|
|
|
1423 /************************************************************************/
|
|
|
1424
|
|
|
1425 static int
|
|
|
1426 decode_coding_category (Lisp_Object symbol)
|
|
|
1427 {
|
|
|
1428 int i;
|
|
|
1429
|
|
|
1430 CHECK_SYMBOL (symbol);
|
|
|
1431 for (i = 0; i <= CODING_CATEGORY_LAST; i++)
|
|
|
1432 if (EQ (coding_category_symbol[i], symbol))
|
|
|
1433 return i;
|
|
|
1434
|
|
|
1435 signal_simple_error ("Unrecognized coding category", symbol);
|
|
|
1436 return 0; /* not reached */
|
|
|
1437 }
|
|
|
1438
|
|
|
1439 DEFUN ("coding-category-list", Fcoding_category_list, 0, 0, 0, /*
|
|
|
1440 Return a list of all recognized coding categories.
|
|
|
1441 */
|
|
|
1442 ())
|
|
|
1443 {
|
|
|
1444 int i;
|
|
|
1445 Lisp_Object list = Qnil;
|
|
|
1446
|
|
|
1447 for (i = CODING_CATEGORY_LAST; i >= 0; i--)
|
|
|
1448 list = Fcons (coding_category_symbol[i], list);
|
|
|
1449 return list;
|
|
|
1450 }
|
|
|
1451
|
|
|
1452 DEFUN ("set-coding-priority-list", Fset_coding_priority_list, 1, 1, 0, /*
|
|
|
1453 Change the priority order of the coding categories.
|
|
|
1454 LIST should be list of coding categories, in descending order of
|
|
|
1455 priority. Unspecified coding categories will be lower in priority
|
|
|
1456 than all specified ones, in the same relative order they were in
|
|
|
1457 previously.
|
|
|
1458 */
|
|
|
1459 (list))
|
|
|
1460 {
|
|
|
1461 int category_to_priority[CODING_CATEGORY_LAST + 1];
|
|
|
1462 int i, j;
|
|
|
1463 Lisp_Object rest;
|
|
|
1464
|
|
|
1465 /* First generate a list that maps coding categories to priorities. */
|
|
|
1466
|
|
|
1467 for (i = 0; i <= CODING_CATEGORY_LAST; i++)
|
|
|
1468 category_to_priority[i] = -1;
|
|
|
1469
|
|
|
1470 /* Highest priority comes from the specified list. */
|
|
|
1471 i = 0;
|
|
|
1472 EXTERNAL_LIST_LOOP (rest, list)
|
|
|
1473 {
|
|
|
1474 int cat = decode_coding_category (XCAR (rest));
|
|
|
1475
|
|
|
1476 if (category_to_priority[cat] >= 0)
|
|
|
1477 signal_simple_error ("Duplicate coding category in list", XCAR (rest));
|
|
|
1478 category_to_priority[cat] = i++;
|
|
|
1479 }
|
|
|
1480
|
|
|
1481 /* Now go through the existing categories by priority to retrieve
|
|
|
1482 the categories not yet specified and preserve their priority
|
|
|
1483 order. */
|
|
|
1484 for (j = 0; j <= CODING_CATEGORY_LAST; j++)
|
|
|
1485 {
|
|
398
|
1486 int cat = fcd->coding_category_by_priority[j];
|
|
259
|
1487 if (category_to_priority[cat] < 0)
|
|
|
1488 category_to_priority[cat] = i++;
|
|
|
1489 }
|
|
|
1490
|
|
|
1491 /* Now we need to construct the inverse of the mapping we just
|
|
|
1492 constructed. */
|
|
|
1493
|
|
|
1494 for (i = 0; i <= CODING_CATEGORY_LAST; i++)
|
|
398
|
1495 fcd->coding_category_by_priority[category_to_priority[i]] = i;
|
|
259
|
1496
|
|
|
1497 /* Phew! That was confusing. */
|
|
|
1498 return Qnil;
|
|
|
1499 }
|
|
|
1500
|
|
|
1501 DEFUN ("coding-priority-list", Fcoding_priority_list, 0, 0, 0, /*
|
|
|
1502 Return a list of coding categories in descending order of priority.
|
|
|
1503 */
|
|
|
1504 ())
|
|
|
1505 {
|
|
|
1506 int i;
|
|
|
1507 Lisp_Object list = Qnil;
|
|
|
1508
|
|
|
1509 for (i = CODING_CATEGORY_LAST; i >= 0; i--)
|
|
398
|
1510 list = Fcons (coding_category_symbol[fcd->coding_category_by_priority[i]],
|
|
259
|
1511 list);
|
|
|
1512 return list;
|
|
|
1513 }
|
|
|
1514
|
|
|
1515 DEFUN ("set-coding-category-system", Fset_coding_category_system, 2, 2, 0, /*
|
|
|
1516 Change the coding system associated with a coding category.
|
|
|
1517 */
|
|
|
1518 (coding_category, coding_system))
|
|
|
1519 {
|
|
|
1520 int cat = decode_coding_category (coding_category);
|
|
|
1521
|
|
|
1522 coding_system = Fget_coding_system (coding_system);
|
|
398
|
1523 fcd->coding_category_system[cat] = coding_system;
|
|
259
|
1524 return Qnil;
|
|
|
1525 }
|
|
|
1526
|
|
|
1527 DEFUN ("coding-category-system", Fcoding_category_system, 1, 1, 0, /*
|
|
|
1528 Return the coding system associated with a coding category.
|
|
|
1529 */
|
|
|
1530 (coding_category))
|
|
|
1531 {
|
|
|
1532 int cat = decode_coding_category (coding_category);
|
|
398
|
1533 Lisp_Object sys = fcd->coding_category_system[cat];
|
|
259
|
1534
|
|
|
1535 if (!NILP (sys))
|
|
|
1536 return XCODING_SYSTEM_NAME (sys);
|
|
|
1537 return Qnil;
|
|
|
1538 }
|
|
|
1539
|
|
|
1540
|
|
|
1541 /************************************************************************/
|
|
|
1542 /* Detecting the encoding of data */
|
|
|
1543 /************************************************************************/
|
|
|
1544
|
|
|
1545 struct detection_state
|
|
|
1546 {
|
|
398
|
1547 eol_type_t eol_type;
|
|
259
|
1548 int seen_non_ascii;
|
|
|
1549 int mask;
|
|
|
1550 #ifdef MULE
|
|
|
1551 struct
|
|
|
1552 {
|
|
|
1553 int mask;
|
|
|
1554 int in_second_byte;
|
|
|
1555 }
|
|
|
1556 big5;
|
|
|
1557
|
|
|
1558 struct
|
|
|
1559 {
|
|
|
1560 int mask;
|
|
|
1561 int in_second_byte;
|
|
|
1562 }
|
|
|
1563 shift_jis;
|
|
|
1564
|
|
|
1565 struct
|
|
|
1566 {
|
|
|
1567 int mask;
|
|
396
|
1568 int in_byte;
|
|
|
1569 }
|
|
|
1570 ucs4;
|
|
|
1571
|
|
|
1572 struct
|
|
|
1573 {
|
|
|
1574 int mask;
|
|
|
1575 int in_byte;
|
|
|
1576 }
|
|
|
1577 utf8;
|
|
|
1578
|
|
|
1579 struct
|
|
|
1580 {
|
|
|
1581 int mask;
|
|
259
|
1582 int initted;
|
|
|
1583 struct iso2022_decoder iso;
|
|
|
1584 unsigned int flags;
|
|
|
1585 int high_byte_count;
|
|
|
1586 unsigned int saw_single_shift:1;
|
|
|
1587 }
|
|
|
1588 iso2022;
|
|
|
1589 #endif
|
|
|
1590 struct
|
|
|
1591 {
|
|
|
1592 int seen_anything;
|
|
|
1593 int just_saw_cr;
|
|
|
1594 }
|
|
|
1595 eol;
|
|
|
1596 };
|
|
|
1597
|
|
|
1598 static int
|
|
|
1599 acceptable_control_char_p (int c)
|
|
|
1600 {
|
|
|
1601 switch (c)
|
|
|
1602 {
|
|
|
1603 /* Allow and ignore control characters that you might
|
|
|
1604 reasonably see in a text file */
|
|
|
1605 case '\r':
|
|
|
1606 case '\n':
|
|
|
1607 case '\t':
|
|
|
1608 case 7: /* bell */
|
|
|
1609 case 8: /* backspace */
|
|
|
1610 case 11: /* vertical tab */
|
|
|
1611 case 12: /* form feed */
|
|
|
1612 case 26: /* MS-DOS C-z junk */
|
|
|
1613 case 31: /* '^_' -- for info */
|
|
|
1614 return 1;
|
|
|
1615 default:
|
|
|
1616 return 0;
|
|
|
1617 }
|
|
|
1618 }
|
|
|
1619
|
|
|
1620 static int
|
|
|
1621 mask_has_at_most_one_bit_p (int mask)
|
|
|
1622 {
|
|
|
1623 /* Perhaps the only thing useful you learn from intensive Microsoft
|
|
|
1624 technical interviews */
|
|
|
1625 return (mask & (mask - 1)) == 0;
|
|
|
1626 }
|
|
|
1627
|
|
398
|
1628 static eol_type_t
|
|
|
1629 detect_eol_type (struct detection_state *st, const unsigned char *src,
|
|
259
|
1630 unsigned int n)
|
|
|
1631 {
|
|
|
1632 int c;
|
|
|
1633
|
|
|
1634 while (n--)
|
|
|
1635 {
|
|
|
1636 c = *src++;
|
|
398
|
1637 if (c == '\n')
|
|
|
1638 {
|
|
|
1639 if (st->eol.just_saw_cr)
|
|
|
1640 return EOL_CRLF;
|
|
|
1641 else if (st->eol.seen_anything)
|
|
|
1642 return EOL_LF;
|
|
|
1643 }
|
|
|
1644 else if (st->eol.just_saw_cr)
|
|
|
1645 return EOL_CR;
|
|
|
1646 else if (c == '\r')
|
|
259
|
1647 st->eol.just_saw_cr = 1;
|
|
|
1648 else
|
|
398
|
1649 st->eol.just_saw_cr = 0;
|
|
259
|
1650 st->eol.seen_anything = 1;
|
|
|
1651 }
|
|
|
1652
|
|
|
1653 return EOL_AUTODETECT;
|
|
|
1654 }
|
|
|
1655
|
|
|
1656 /* Attempt to determine the encoding and EOL type of the given text.
|
|
|
1657 Before calling this function for the first type, you must initialize
|
|
|
1658 st->eol_type as appropriate and initialize st->mask to ~0.
|
|
|
1659
|
|
|
1660 st->eol_type holds the determined EOL type, or EOL_AUTODETECT if
|
|
|
1661 not yet known.
|
|
|
1662
|
|
|
1663 st->mask holds the determined coding category mask, or ~0 if only
|
|
|
1664 ASCII has been seen so far.
|
|
|
1665
|
|
|
1666 Returns:
|
|
|
1667
|
|
|
1668 0 == st->eol_type is EOL_AUTODETECT and/or more than coding category
|
|
|
1669 is present in st->mask
|
|
|
1670 1 == definitive answers are here for both st->eol_type and st->mask
|
|
|
1671 */
|
|
|
1672
|
|
|
1673 static int
|
|
398
|
1674 detect_coding_type (struct detection_state *st, const Extbyte *src,
|
|
259
|
1675 unsigned int n, int just_do_eol)
|
|
|
1676 {
|
|
|
1677 int c;
|
|
|
1678
|
|
|
1679 if (st->eol_type == EOL_AUTODETECT)
|
|
|
1680 st->eol_type = detect_eol_type (st, src, n);
|
|
|
1681
|
|
|
1682 if (just_do_eol)
|
|
|
1683 return st->eol_type != EOL_AUTODETECT;
|
|
|
1684
|
|
|
1685 if (!st->seen_non_ascii)
|
|
|
1686 {
|
|
|
1687 for (; n; n--, src++)
|
|
|
1688 {
|
|
|
1689 c = *src;
|
|
|
1690 if ((c < 0x20 && !acceptable_control_char_p (c)) || c >= 0x80)
|
|
|
1691 {
|
|
|
1692 st->seen_non_ascii = 1;
|
|
|
1693 #ifdef MULE
|
|
|
1694 st->shift_jis.mask = ~0;
|
|
|
1695 st->big5.mask = ~0;
|
|
396
|
1696 st->ucs4.mask = ~0;
|
|
|
1697 st->utf8.mask = ~0;
|
|
259
|
1698 st->iso2022.mask = ~0;
|
|
|
1699 #endif
|
|
|
1700 break;
|
|
|
1701 }
|
|
|
1702 }
|
|
|
1703 }
|
|
|
1704
|
|
|
1705 if (!n)
|
|
|
1706 return 0;
|
|
|
1707 #ifdef MULE
|
|
|
1708 if (!mask_has_at_most_one_bit_p (st->iso2022.mask))
|
|
|
1709 st->iso2022.mask = detect_coding_iso2022 (st, src, n);
|
|
|
1710 if (!mask_has_at_most_one_bit_p (st->shift_jis.mask))
|
|
|
1711 st->shift_jis.mask = detect_coding_sjis (st, src, n);
|
|
|
1712 if (!mask_has_at_most_one_bit_p (st->big5.mask))
|
|
|
1713 st->big5.mask = detect_coding_big5 (st, src, n);
|
|
396
|
1714 if (!mask_has_at_most_one_bit_p (st->utf8.mask))
|
|
|
1715 st->utf8.mask = detect_coding_utf8 (st, src, n);
|
|
|
1716 if (!mask_has_at_most_one_bit_p (st->ucs4.mask))
|
|
|
1717 st->ucs4.mask = detect_coding_ucs4 (st, src, n);
|
|
|
1718
|
|
|
1719 st->mask
|
|
|
1720 = st->iso2022.mask | st->shift_jis.mask | st->big5.mask
|
|
|
1721 | st->utf8.mask | st->ucs4.mask;
|
|
259
|
1722 #endif
|
|
|
1723 {
|
|
|
1724 int retval = mask_has_at_most_one_bit_p (st->mask);
|
|
|
1725 st->mask |= CODING_CATEGORY_NO_CONVERSION_MASK;
|
|
|
1726 return retval && st->eol_type != EOL_AUTODETECT;
|
|
|
1727 }
|
|
|
1728 }
|
|
|
1729
|
|
|
1730 static Lisp_Object
|
|
|
1731 coding_system_from_mask (int mask)
|
|
|
1732 {
|
|
|
1733 if (mask == ~0)
|
|
|
1734 {
|
|
|
1735 /* If the file was entirely or basically ASCII, use the
|
|
|
1736 default value of `buffer-file-coding-system'. */
|
|
|
1737 Lisp_Object retval =
|
|
|
1738 XBUFFER (Vbuffer_defaults)->buffer_file_coding_system;
|
|
|
1739 if (!NILP (retval))
|
|
|
1740 {
|
|
|
1741 retval = Ffind_coding_system (retval);
|
|
|
1742 if (NILP (retval))
|
|
|
1743 {
|
|
|
1744 warn_when_safe
|
|
|
1745 (Qbad_variable, Qwarning,
|
|
|
1746 "Invalid `default-buffer-file-coding-system', set to nil");
|
|
|
1747 XBUFFER (Vbuffer_defaults)->buffer_file_coding_system = Qnil;
|
|
|
1748 }
|
|
|
1749 }
|
|
|
1750 if (NILP (retval))
|
|
398
|
1751 retval = Fget_coding_system (Qraw_text);
|
|
259
|
1752 return retval;
|
|
|
1753 }
|
|
|
1754 else
|
|
|
1755 {
|
|
|
1756 int i;
|
|
|
1757 int cat = -1;
|
|
|
1758 #ifdef MULE
|
|
|
1759 mask = postprocess_iso2022_mask (mask);
|
|
|
1760 #endif
|
|
|
1761 /* Look through the coding categories by priority and find
|
|
|
1762 the first one that is allowed. */
|
|
|
1763 for (i = 0; i <= CODING_CATEGORY_LAST; i++)
|
|
|
1764 {
|
|
398
|
1765 cat = fcd->coding_category_by_priority[i];
|
|
259
|
1766 if ((mask & (1 << cat)) &&
|
|
398
|
1767 !NILP (fcd->coding_category_system[cat]))
|
|
259
|
1768 break;
|
|
|
1769 }
|
|
|
1770 if (cat >= 0)
|
|
398
|
1771 return fcd->coding_category_system[cat];
|
|
259
|
1772 else
|
|
398
|
1773 return Fget_coding_system (Qraw_text);
|
|
259
|
1774 }
|
|
|
1775 }
|
|
|
1776
|
|
|
1777 /* Given a seekable read stream and potential coding system and EOL type
|
|
|
1778 as specified, do any autodetection that is called for. If the
|
|
398
|
1779 coding system and/or EOL type are not `autodetect', they will be left
|
|
259
|
1780 alone; but this function will never return an autodetect coding system
|
|
|
1781 or EOL type.
|
|
|
1782
|
|
|
1783 This function does not automatically fetch subsidiary coding systems;
|
|
|
1784 that should be unnecessary with the explicit eol-type argument. */
|
|
|
1785
|
|
398
|
1786 #define LENGTH(string_constant) (sizeof (string_constant) - 1)
|
|
|
1787
|
|
259
|
1788 void
|
|
|
1789 determine_real_coding_system (Lstream *stream, Lisp_Object *codesys_in_out,
|
|
398
|
1790 eol_type_t *eol_type_in_out)
|
|
259
|
1791 {
|
|
|
1792 struct detection_state decst;
|
|
|
1793
|
|
|
1794 if (*eol_type_in_out == EOL_AUTODETECT)
|
|
|
1795 *eol_type_in_out = XCODING_SYSTEM_EOL_TYPE (*codesys_in_out);
|
|
|
1796
|
|
272
|
1797 xzero (decst);
|
|
259
|
1798 decst.eol_type = *eol_type_in_out;
|
|
|
1799 decst.mask = ~0;
|
|
|
1800
|
|
|
1801 /* If autodetection is called for, do it now. */
|
|
398
|
1802 if (XCODING_SYSTEM_TYPE (*codesys_in_out) == CODESYS_AUTODETECT
|
|
|
1803 || *eol_type_in_out == EOL_AUTODETECT)
|
|
259
|
1804 {
|
|
398
|
1805 Extbyte buf[4096];
|
|
|
1806 Lisp_Object coding_system = Qnil;
|
|
|
1807 Extbyte *p;
|
|
|
1808 ssize_t nread = Lstream_read (stream, buf, sizeof (buf));
|
|
|
1809 Extbyte *scan_end;
|
|
|
1810
|
|
|
1811 /* Look for initial "-*-"; mode line prefix */
|
|
|
1812 for (p = buf,
|
|
|
1813 scan_end = buf + nread - LENGTH ("-*-coding:?-*-");
|
|
|
1814 p <= scan_end
|
|
|
1815 && *p != '\n'
|
|
|
1816 && *p != '\r';
|
|
|
1817 p++)
|
|
|
1818 if (*p == '-' && *(p+1) == '*' && *(p+2) == '-')
|
|
|
1819 {
|
|
|
1820 Extbyte *local_vars_beg = p + 3;
|
|
|
1821 /* Look for final "-*-"; mode line suffix */
|
|
|
1822 for (p = local_vars_beg,
|
|
|
1823 scan_end = buf + nread - LENGTH ("-*-");
|
|
|
1824 p <= scan_end
|
|
|
1825 && *p != '\n'
|
|
|
1826 && *p != '\r';
|
|
|
1827 p++)
|
|
|
1828 if (*p == '-' && *(p+1) == '*' && *(p+2) == '-')
|
|
|
1829 {
|
|
|
1830 Extbyte *suffix = p;
|
|
|
1831 /* Look for "coding:" */
|
|
|
1832 for (p = local_vars_beg,
|
|
|
1833 scan_end = suffix - LENGTH ("coding:?");
|
|
|
1834 p <= scan_end;
|
|
|
1835 p++)
|
|
|
1836 if (memcmp ("coding:", p, LENGTH ("coding:")) == 0
|
|
|
1837 && (p == local_vars_beg
|
|
|
1838 || (*(p-1) == ' ' ||
|
|
|
1839 *(p-1) == '\t' ||
|
|
|
1840 *(p-1) == ';')))
|
|
|
1841 {
|
|
|
1842 Extbyte save;
|
|
|
1843 int n;
|
|
|
1844 p += LENGTH ("coding:");
|
|
|
1845 while (*p == ' ' || *p == '\t') p++;
|
|
|
1846
|
|
|
1847 /* Get coding system name */
|
|
|
1848 save = *suffix; *suffix = '\0';
|
|
|
1849 /* Characters valid in a MIME charset name (rfc 1521),
|
|
|
1850 and in a Lisp symbol name. */
|
|
|
1851 n = strspn ( (char *) p,
|
|
|
1852 "ABCDEFGHIJKLMNOPQRSTUVWXYZ"
|
|
|
1853 "abcdefghijklmnopqrstuvwxyz"
|
|
|
1854 "0123456789"
|
|
|
1855 "!$%&*+-.^_{|}~");
|
|
|
1856 *suffix = save;
|
|
|
1857 if (n > 0)
|
|
|
1858 {
|
|
|
1859 save = p[n]; p[n] = '\0';
|
|
|
1860 coding_system =
|
|
|
1861 Ffind_coding_system (intern ((char *) p));
|
|
|
1862 p[n] = save;
|
|
|
1863 }
|
|
|
1864 break;
|
|
|
1865 }
|
|
|
1866 break;
|
|
|
1867 }
|
|
259
|
1868 break;
|
|
398
|
1869 }
|
|
|
1870
|
|
|
1871 if (NILP (coding_system))
|
|
|
1872 do
|
|
|
1873 {
|
|
|
1874 if (detect_coding_type (&decst, buf, nread,
|
|
|
1875 XCODING_SYSTEM_TYPE (*codesys_in_out)
|
|
|
1876 != CODESYS_AUTODETECT))
|
|
|
1877 break;
|
|
|
1878 nread = Lstream_read (stream, buf, sizeof (buf));
|
|
|
1879 if (nread == 0)
|
|
|
1880 break;
|
|
|
1881 }
|
|
|
1882 while (1);
|
|
|
1883
|
|
|
1884 else if (XCODING_SYSTEM_TYPE (*codesys_in_out) == CODESYS_AUTODETECT
|
|
|
1885 && XCODING_SYSTEM_EOL_TYPE (coding_system) == EOL_AUTODETECT)
|
|
|
1886 do
|
|
|
1887 {
|
|
|
1888 if (detect_coding_type (&decst, buf, nread, 1))
|
|
|
1889 break;
|
|
|
1890 nread = Lstream_read (stream, buf, sizeof (buf));
|
|
|
1891 if (!nread)
|
|
|
1892 break;
|
|
|
1893 }
|
|
|
1894 while (1);
|
|
259
|
1895
|
|
|
1896 *eol_type_in_out = decst.eol_type;
|
|
|
1897 if (XCODING_SYSTEM_TYPE (*codesys_in_out) == CODESYS_AUTODETECT)
|
|
398
|
1898 {
|
|
|
1899 if (NILP (coding_system))
|
|
|
1900 *codesys_in_out = coding_system_from_mask (decst.mask);
|
|
|
1901 else
|
|
|
1902 *codesys_in_out = coding_system;
|
|
|
1903 }
|
|
259
|
1904 }
|
|
|
1905
|
|
|
1906 /* If we absolutely can't determine the EOL type, just assume LF. */
|
|
|
1907 if (*eol_type_in_out == EOL_AUTODETECT)
|
|
|
1908 *eol_type_in_out = EOL_LF;
|
|
|
1909
|
|
|
1910 Lstream_rewind (stream);
|
|
|
1911 }
|
|
|
1912
|
|
|
1913 DEFUN ("detect-coding-region", Fdetect_coding_region, 2, 3, 0, /*
|
|
|
1914 Detect coding system of the text in the region between START and END.
|
|
|
1915 Returned a list of possible coding systems ordered by priority.
|
|
|
1916 If only ASCII characters are found, it returns 'undecided or one of
|
|
|
1917 its subsidiary coding systems according to a detected end-of-line
|
|
|
1918 type. Optional arg BUFFER defaults to the current buffer.
|
|
|
1919 */
|
|
|
1920 (start, end, buffer))
|
|
|
1921 {
|
|
|
1922 Lisp_Object val = Qnil;
|
|
|
1923 struct buffer *buf = decode_buffer (buffer, 0);
|
|
|
1924 Bufpos b, e;
|
|
|
1925 Lisp_Object instream, lb_instream;
|
|
|
1926 Lstream *istr, *lb_istr;
|
|
|
1927 struct detection_state decst;
|
|
|
1928 struct gcpro gcpro1, gcpro2;
|
|
|
1929
|
|
|
1930 get_buffer_range_char (buf, start, end, &b, &e, 0);
|
|
|
1931 lb_instream = make_lisp_buffer_input_stream (buf, b, e, 0);
|
|
|
1932 lb_istr = XLSTREAM (lb_instream);
|
|
|
1933 instream = make_encoding_input_stream (lb_istr, Fget_coding_system (Qbinary));
|
|
|
1934 istr = XLSTREAM (instream);
|
|
|
1935 GCPRO2 (instream, lb_instream);
|
|
272
|
1936 xzero (decst);
|
|
259
|
1937 decst.eol_type = EOL_AUTODETECT;
|
|
|
1938 decst.mask = ~0;
|
|
|
1939 while (1)
|
|
|
1940 {
|
|
|
1941 unsigned char random_buffer[4096];
|
|
398
|
1942 ssize_t nread = Lstream_read (istr, random_buffer, sizeof (random_buffer));
|
|
259
|
1943
|
|
|
1944 if (!nread)
|
|
|
1945 break;
|
|
|
1946 if (detect_coding_type (&decst, random_buffer, nread, 0))
|
|
|
1947 break;
|
|
|
1948 }
|
|
|
1949
|
|
|
1950 if (decst.mask == ~0)
|
|
|
1951 val = subsidiary_coding_system (Fget_coding_system (Qundecided),
|
|
|
1952 decst.eol_type);
|
|
|
1953 else
|
|
|
1954 {
|
|
|
1955 int i;
|
|
|
1956
|
|
|
1957 val = Qnil;
|
|
|
1958 #ifdef MULE
|
|
|
1959 decst.mask = postprocess_iso2022_mask (decst.mask);
|
|
|
1960 #endif
|
|
|
1961 for (i = CODING_CATEGORY_LAST; i >= 0; i--)
|
|
|
1962 {
|
|
398
|
1963 int sys = fcd->coding_category_by_priority[i];
|
|
259
|
1964 if (decst.mask & (1 << sys))
|
|
|
1965 {
|
|
398
|
1966 Lisp_Object codesys = fcd->coding_category_system[sys];
|
|
259
|
1967 if (!NILP (codesys))
|
|
|
1968 codesys = subsidiary_coding_system (codesys, decst.eol_type);
|
|
|
1969 val = Fcons (codesys, val);
|
|
|
1970 }
|
|
|
1971 }
|
|
|
1972 }
|
|
|
1973 Lstream_close (istr);
|
|
|
1974 UNGCPRO;
|
|
|
1975 Lstream_delete (istr);
|
|
|
1976 Lstream_delete (lb_istr);
|
|
|
1977 return val;
|
|
|
1978 }
|
|
|
1979
|
|
|
1980
|
|
|
1981 /************************************************************************/
|
|
|
1982 /* Converting to internal Mule format ("decoding") */
|
|
|
1983 /************************************************************************/
|
|
|
1984
|
|
|
1985 /* A decoding stream is a stream used for decoding text (i.e.
|
|
|
1986 converting from some external format to internal format).
|
|
|
1987 The decoding-stream object keeps track of the actual coding
|
|
|
1988 stream, the stream that is at the other end, and data that
|
|
|
1989 needs to be persistent across the lifetime of the stream. */
|
|
|
1990
|
|
|
1991 /* Handle the EOL stuff related to just-read-in character C.
|
|
|
1992 EOL_TYPE is the EOL type of the coding stream.
|
|
|
1993 FLAGS is the current value of FLAGS in the coding stream, and may
|
|
|
1994 be modified by this macro. (The macro only looks at the
|
|
|
1995 CODING_STATE_CR flag.) DST is the Dynarr to which the decoded
|
|
|
1996 bytes are to be written. You need to also define a local goto
|
|
|
1997 label "label_continue_loop" that is at the end of the main
|
|
|
1998 character-reading loop.
|
|
|
1999
|
|
|
2000 If C is a CR character, then this macro handles it entirely and
|
|
|
2001 jumps to label_continue_loop. Otherwise, this macro does not add
|
|
|
2002 anything to DST, and continues normally. You should continue
|
|
|
2003 processing C normally after this macro. */
|
|
|
2004
|
|
|
2005 #define DECODE_HANDLE_EOL_TYPE(eol_type, c, flags, dst) \
|
|
|
2006 do { \
|
|
|
2007 if (c == '\r') \
|
|
|
2008 { \
|
|
|
2009 if (eol_type == EOL_CR) \
|
|
|
2010 Dynarr_add (dst, '\n'); \
|
|
|
2011 else if (eol_type != EOL_CRLF || flags & CODING_STATE_CR) \
|
|
|
2012 Dynarr_add (dst, c); \
|
|
|
2013 else \
|
|
|
2014 flags |= CODING_STATE_CR; \
|
|
|
2015 goto label_continue_loop; \
|
|
|
2016 } \
|
|
|
2017 else if (flags & CODING_STATE_CR) \
|
|
|
2018 { /* eol_type == CODING_SYSTEM_EOL_CRLF */ \
|
|
|
2019 if (c != '\n') \
|
|
|
2020 Dynarr_add (dst, '\r'); \
|
|
|
2021 flags &= ~CODING_STATE_CR; \
|
|
|
2022 } \
|
|
|
2023 } while (0)
|
|
|
2024
|
|
|
2025 /* C should be a binary character in the range 0 - 255; convert
|
|
|
2026 to internal format and add to Dynarr DST. */
|
|
|
2027
|
|
|
2028 #define DECODE_ADD_BINARY_CHAR(c, dst) \
|
|
|
2029 do { \
|
|
|
2030 if (BYTE_ASCII_P (c)) \
|
|
|
2031 Dynarr_add (dst, c); \
|
|
|
2032 else if (BYTE_C1_P (c)) \
|
|
|
2033 { \
|
|
|
2034 Dynarr_add (dst, LEADING_BYTE_CONTROL_1); \
|
|
|
2035 Dynarr_add (dst, c + 0x20); \
|
|
|
2036 } \
|
|
|
2037 else \
|
|
|
2038 { \
|
|
|
2039 Dynarr_add (dst, LEADING_BYTE_LATIN_ISO8859_1); \
|
|
|
2040 Dynarr_add (dst, c); \
|
|
|
2041 } \
|
|
|
2042 } while (0)
|
|
|
2043
|
|
|
2044 #define DECODE_OUTPUT_PARTIAL_CHAR(ch) \
|
|
|
2045 do { \
|
|
|
2046 if (ch) \
|
|
|
2047 { \
|
|
|
2048 DECODE_ADD_BINARY_CHAR (ch, dst); \
|
|
|
2049 ch = 0; \
|
|
|
2050 } \
|
|
|
2051 } while (0)
|
|
|
2052
|
|
|
2053 #define DECODE_HANDLE_END_OF_CONVERSION(flags, ch, dst) \
|
|
|
2054 do { \
|
|
398
|
2055 if (flags & CODING_STATE_END) \
|
|
|
2056 { \
|
|
|
2057 DECODE_OUTPUT_PARTIAL_CHAR (ch); \
|
|
|
2058 if (flags & CODING_STATE_CR) \
|
|
|
2059 Dynarr_add (dst, '\r'); \
|
|
|
2060 } \
|
|
259
|
2061 } while (0)
|
|
|
2062
|
|
|
2063 #define DECODING_STREAM_DATA(stream) LSTREAM_TYPE_DATA (stream, decoding)
|
|
|
2064
|
|
|
2065 struct decoding_stream
|
|
|
2066 {
|
|
|
2067 /* Coding system that governs the conversion. */
|
|
396
|
2068 Lisp_Coding_System *codesys;
|
|
259
|
2069
|
|
|
2070 /* Stream that we read the encoded data from or
|
|
|
2071 write the decoded data to. */
|
|
|
2072 Lstream *other_end;
|
|
|
2073
|
|
|
2074 /* If we are reading, then we can return only a fixed amount of
|
|
|
2075 data, so if the conversion resulted in too much data, we store it
|
|
|
2076 here for retrieval the next time around. */
|
|
|
2077 unsigned_char_dynarr *runoff;
|
|
|
2078
|
|
|
2079 /* FLAGS holds flags indicating the current state of the decoding.
|
|
|
2080 Some of these flags are dependent on the coding system. */
|
|
|
2081 unsigned int flags;
|
|
|
2082
|
|
|
2083 /* CH holds a partially built-up character. Since we only deal
|
|
|
2084 with one- and two-byte characters at the moment, we only use
|
|
|
2085 this to store the first byte of a two-byte character. */
|
|
|
2086 unsigned int ch;
|
|
|
2087
|
|
|
2088 /* EOL_TYPE specifies the type of end-of-line conversion that
|
|
|
2089 currently applies. We need to keep this separate from the
|
|
|
2090 EOL type stored in CODESYS because the latter might indicate
|
|
|
2091 automatic EOL-type detection while the former will always
|
|
|
2092 indicate a particular EOL type. */
|
|
398
|
2093 eol_type_t eol_type;
|
|
259
|
2094 #ifdef MULE
|
|
|
2095 /* Additional ISO2022 information. We define the structure above
|
|
|
2096 because it's also needed by the detection routines. */
|
|
|
2097 struct iso2022_decoder iso2022;
|
|
|
2098
|
|
|
2099 /* Additional information (the state of the running CCL program)
|
|
|
2100 used by the CCL decoder. */
|
|
|
2101 struct ccl_program ccl;
|
|
398
|
2102
|
|
|
2103 /* counter for UTF-8 or UCS-4 */
|
|
|
2104 unsigned char counter;
|
|
259
|
2105 #endif
|
|
|
2106 struct detection_state decst;
|
|
|
2107 };
|
|
|
2108
|
|
398
|
2109 static ssize_t decoding_reader (Lstream *stream,
|
|
|
2110 unsigned char *data, size_t size);
|
|
|
2111 static ssize_t decoding_writer (Lstream *stream,
|
|
|
2112 const unsigned char *data, size_t size);
|
|
259
|
2113 static int decoding_rewinder (Lstream *stream);
|
|
|
2114 static int decoding_seekable_p (Lstream *stream);
|
|
|
2115 static int decoding_flusher (Lstream *stream);
|
|
|
2116 static int decoding_closer (Lstream *stream);
|
|
|
2117
|
|
398
|
2118 static Lisp_Object decoding_marker (Lisp_Object stream);
|
|
259
|
2119
|
|
|
2120 DEFINE_LSTREAM_IMPLEMENTATION ("decoding", lstream_decoding,
|
|
|
2121 sizeof (struct decoding_stream));
|
|
|
2122
|
|
|
2123 static Lisp_Object
|
|
398
|
2124 decoding_marker (Lisp_Object stream)
|
|
259
|
2125 {
|
|
|
2126 Lstream *str = DECODING_STREAM_DATA (XLSTREAM (stream))->other_end;
|
|
|
2127 Lisp_Object str_obj;
|
|
|
2128
|
|
|
2129 /* We do not need to mark the coding systems or charsets stored
|
|
|
2130 within the stream because they are stored in a global list
|
|
|
2131 and automatically marked. */
|
|
|
2132
|
|
|
2133 XSETLSTREAM (str_obj, str);
|
|
398
|
2134 mark_object (str_obj);
|
|
259
|
2135 if (str->imp->marker)
|
|
398
|
2136 return (str->imp->marker) (str_obj);
|
|
259
|
2137 else
|
|
|
2138 return Qnil;
|
|
|
2139 }
|
|
|
2140
|
|
|
2141 /* Read SIZE bytes of data and store it into DATA. We are a decoding stream
|
|
|
2142 so we read data from the other end, decode it, and store it into DATA. */
|
|
|
2143
|
|
398
|
2144 static ssize_t
|
|
272
|
2145 decoding_reader (Lstream *stream, unsigned char *data, size_t size)
|
|
259
|
2146 {
|
|
|
2147 struct decoding_stream *str = DECODING_STREAM_DATA (stream);
|
|
|
2148 unsigned char *orig_data = data;
|
|
398
|
2149 ssize_t read_size;
|
|
259
|
2150 int error_occurred = 0;
|
|
|
2151
|
|
|
2152 /* We need to interface to mule_decode(), which expects to take some
|
|
|
2153 amount of data and store the result into a Dynarr. We have
|
|
|
2154 mule_decode() store into str->runoff, and take data from there
|
|
|
2155 as necessary. */
|
|
|
2156
|
|
|
2157 /* We loop until we have enough data, reading chunks from the other
|
|
|
2158 end and decoding it. */
|
|
|
2159 while (1)
|
|
|
2160 {
|
|
|
2161 /* Take data from the runoff if we can. Make sure to take at
|
|
|
2162 most SIZE bytes, and delete the data from the runoff. */
|
|
|
2163 if (Dynarr_length (str->runoff) > 0)
|
|
|
2164 {
|
|
272
|
2165 size_t chunk = min (size, (size_t) Dynarr_length (str->runoff));
|
|
259
|
2166 memcpy (data, Dynarr_atp (str->runoff, 0), chunk);
|
|
|
2167 Dynarr_delete_many (str->runoff, 0, chunk);
|
|
|
2168 data += chunk;
|
|
|
2169 size -= chunk;
|
|
|
2170 }
|
|
|
2171
|
|
|
2172 if (size == 0)
|
|
|
2173 break; /* No more room for data */
|
|
|
2174
|
|
|
2175 if (str->flags & CODING_STATE_END)
|
|
|
2176 /* This means that on the previous iteration, we hit the EOF on
|
|
|
2177 the other end. We loop once more so that mule_decode() can
|
|
|
2178 output any final stuff it may be holding, or any "go back
|
|
|
2179 to a sane state" escape sequences. (This latter makes sense
|
|
|
2180 during encoding.) */
|
|
|
2181 break;
|
|
|
2182
|
|
|
2183 /* Exhausted the runoff, so get some more. DATA has at least
|
|
|
2184 SIZE bytes left of storage in it, so it's OK to read directly
|
|
|
2185 into it. (We'll be overwriting above, after we've decoded it
|
|
|
2186 into the runoff.) */
|
|
|
2187 read_size = Lstream_read (str->other_end, data, size);
|
|
|
2188 if (read_size < 0)
|
|
|
2189 {
|
|
|
2190 error_occurred = 1;
|
|
|
2191 break;
|
|
|
2192 }
|
|
|
2193 if (read_size == 0)
|
|
|
2194 /* There might be some more end data produced in the translation.
|
|
|
2195 See the comment above. */
|
|
|
2196 str->flags |= CODING_STATE_END;
|
|
|
2197 mule_decode (stream, data, str->runoff, read_size);
|
|
|
2198 }
|
|
|
2199
|
|
|
2200 if (data - orig_data == 0)
|
|
|
2201 return error_occurred ? -1 : 0;
|
|
|
2202 else
|
|
|
2203 return data - orig_data;
|
|
|
2204 }
|
|
|
2205
|
|
398
|
2206 static ssize_t
|
|
|
2207 decoding_writer (Lstream *stream, const unsigned char *data, size_t size)
|
|
259
|
2208 {
|
|
|
2209 struct decoding_stream *str = DECODING_STREAM_DATA (stream);
|
|
398
|
2210 ssize_t retval;
|
|
259
|
2211
|
|
|
2212 /* Decode all our data into the runoff, and then attempt to write
|
|
|
2213 it all out to the other end. Remove whatever chunk we succeeded
|
|
|
2214 in writing. */
|
|
|
2215 mule_decode (stream, data, str->runoff, size);
|
|
|
2216 retval = Lstream_write (str->other_end, Dynarr_atp (str->runoff, 0),
|
|
|
2217 Dynarr_length (str->runoff));
|
|
|
2218 if (retval > 0)
|
|
|
2219 Dynarr_delete_many (str->runoff, 0, retval);
|
|
|
2220 /* Do NOT return retval. The return value indicates how much
|
|
|
2221 of the incoming data was written, not how many bytes were
|
|
|
2222 written. */
|
|
|
2223 return size;
|
|
|
2224 }
|
|
|
2225
|
|
|
2226 static void
|
|
|
2227 reset_decoding_stream (struct decoding_stream *str)
|
|
|
2228 {
|
|
|
2229 #ifdef MULE
|
|
|
2230 if (CODING_SYSTEM_TYPE (str->codesys) == CODESYS_ISO2022)
|
|
|
2231 {
|
|
272
|
2232 Lisp_Object coding_system;
|
|
259
|
2233 XSETCODING_SYSTEM (coding_system, str->codesys);
|
|
|
2234 reset_iso2022 (coding_system, &str->iso2022);
|
|
|
2235 }
|
|
|
2236 else if (CODING_SYSTEM_TYPE (str->codesys) == CODESYS_CCL)
|
|
|
2237 {
|
|
|
2238 setup_ccl_program (&str->ccl, CODING_SYSTEM_CCL_DECODE (str->codesys));
|
|
|
2239 }
|
|
398
|
2240 str->counter = 0;
|
|
272
|
2241 #endif /* MULE */
|
|
259
|
2242 str->flags = str->ch = 0;
|
|
|
2243 }
|
|
|
2244
|
|
|
2245 static int
|
|
|
2246 decoding_rewinder (Lstream *stream)
|
|
|
2247 {
|
|
|
2248 struct decoding_stream *str = DECODING_STREAM_DATA (stream);
|
|
|
2249 reset_decoding_stream (str);
|
|
|
2250 Dynarr_reset (str->runoff);
|
|
|
2251 return Lstream_rewind (str->other_end);
|
|
|
2252 }
|
|
|
2253
|
|
|
2254 static int
|
|
|
2255 decoding_seekable_p (Lstream *stream)
|
|
|
2256 {
|
|
|
2257 struct decoding_stream *str = DECODING_STREAM_DATA (stream);
|
|
|
2258 return Lstream_seekable_p (str->other_end);
|
|
|
2259 }
|
|
|
2260
|
|
|
2261 static int
|
|
|
2262 decoding_flusher (Lstream *stream)
|
|
|
2263 {
|
|
|
2264 struct decoding_stream *str = DECODING_STREAM_DATA (stream);
|
|
|
2265 return Lstream_flush (str->other_end);
|
|
|
2266 }
|
|
|
2267
|
|
|
2268 static int
|
|
|
2269 decoding_closer (Lstream *stream)
|
|
|
2270 {
|
|
|
2271 struct decoding_stream *str = DECODING_STREAM_DATA (stream);
|
|
|
2272 if (stream->flags & LSTREAM_FL_WRITE)
|
|
|
2273 {
|
|
|
2274 str->flags |= CODING_STATE_END;
|
|
|
2275 decoding_writer (stream, 0, 0);
|
|
|
2276 }
|
|
|
2277 Dynarr_free (str->runoff);
|
|
|
2278 #ifdef MULE
|
|
396
|
2279 #ifdef ENABLE_COMPOSITE_CHARS
|
|
259
|
2280 if (str->iso2022.composite_chars)
|
|
|
2281 Dynarr_free (str->iso2022.composite_chars);
|
|
|
2282 #endif
|
|
396
|
2283 #endif
|
|
259
|
2284 return Lstream_close (str->other_end);
|
|
|
2285 }
|
|
|
2286
|
|
|
2287 Lisp_Object
|
|
|
2288 decoding_stream_coding_system (Lstream *stream)
|
|
|
2289 {
|
|
272
|
2290 Lisp_Object coding_system;
|
|
259
|
2291 struct decoding_stream *str = DECODING_STREAM_DATA (stream);
|
|
|
2292
|
|
|
2293 XSETCODING_SYSTEM (coding_system, str->codesys);
|
|
|
2294 return subsidiary_coding_system (coding_system, str->eol_type);
|
|
|
2295 }
|
|
|
2296
|
|
|
2297 void
|
|
|
2298 set_decoding_stream_coding_system (Lstream *lstr, Lisp_Object codesys)
|
|
|
2299 {
|
|
396
|
2300 Lisp_Coding_System *cs = XCODING_SYSTEM (codesys);
|
|
259
|
2301 struct decoding_stream *str = DECODING_STREAM_DATA (lstr);
|
|
|
2302 str->codesys = cs;
|
|
|
2303 if (CODING_SYSTEM_EOL_TYPE (cs) != EOL_AUTODETECT)
|
|
|
2304 str->eol_type = CODING_SYSTEM_EOL_TYPE (cs);
|
|
|
2305 reset_decoding_stream (str);
|
|
|
2306 }
|
|
|
2307
|
|
|
2308 /* WARNING WARNING WARNING WARNING!!!!! If you open up a decoding
|
|
|
2309 stream for writing, no automatic code detection will be performed.
|
|
|
2310 The reason for this is that automatic code detection requires a
|
|
|
2311 seekable input. Things will also fail if you open a decoding
|
|
|
2312 stream for reading using a non-fully-specified coding system and
|
|
|
2313 a non-seekable input stream. */
|
|
|
2314
|
|
|
2315 static Lisp_Object
|
|
|
2316 make_decoding_stream_1 (Lstream *stream, Lisp_Object codesys,
|
|
398
|
2317 const char *mode)
|
|
259
|
2318 {
|
|
|
2319 Lstream *lstr = Lstream_new (lstream_decoding, mode);
|
|
|
2320 struct decoding_stream *str = DECODING_STREAM_DATA (lstr);
|
|
|
2321 Lisp_Object obj;
|
|
|
2322
|
|
272
|
2323 xzero (*str);
|
|
259
|
2324 str->other_end = stream;
|
|
|
2325 str->runoff = (unsigned_char_dynarr *) Dynarr_new (unsigned_char);
|
|
|
2326 str->eol_type = EOL_AUTODETECT;
|
|
|
2327 if (!strcmp (mode, "r")
|
|
|
2328 && Lstream_seekable_p (stream))
|
|
|
2329 /* We can determine the coding system now. */
|
|
|
2330 determine_real_coding_system (stream, &codesys, &str->eol_type);
|
|
|
2331 set_decoding_stream_coding_system (lstr, codesys);
|
|
|
2332 str->decst.eol_type = str->eol_type;
|
|
|
2333 str->decst.mask = ~0;
|
|
|
2334 XSETLSTREAM (obj, lstr);
|
|
|
2335 return obj;
|
|
|
2336 }
|
|
|
2337
|
|
|
2338 Lisp_Object
|
|
|
2339 make_decoding_input_stream (Lstream *stream, Lisp_Object codesys)
|
|
|
2340 {
|
|
|
2341 return make_decoding_stream_1 (stream, codesys, "r");
|
|
|
2342 }
|
|
|
2343
|
|
|
2344 Lisp_Object
|
|
|
2345 make_decoding_output_stream (Lstream *stream, Lisp_Object codesys)
|
|
|
2346 {
|
|
|
2347 return make_decoding_stream_1 (stream, codesys, "w");
|
|
|
2348 }
|
|
|
2349
|
|
|
2350 /* Note: the decode_coding_* functions all take the same
|
|
|
2351 arguments as mule_decode(), which is to say some SRC data of
|
|
|
2352 size N, which is to be stored into dynamic array DST.
|
|
|
2353 DECODING is the stream within which the decoding is
|
|
|
2354 taking place, but no data is actually read from or
|
|
|
2355 written to that stream; that is handled in decoding_reader()
|
|
|
2356 or decoding_writer(). This allows the same functions to
|
|
|
2357 be used for both reading and writing. */
|
|
|
2358
|
|
|
2359 static void
|
|
398
|
2360 mule_decode (Lstream *decoding, const unsigned char *src,
|
|
259
|
2361 unsigned_char_dynarr *dst, unsigned int n)
|
|
|
2362 {
|
|
|
2363 struct decoding_stream *str = DECODING_STREAM_DATA (decoding);
|
|
|
2364
|
|
|
2365 /* If necessary, do encoding-detection now. We do this when
|
|
|
2366 we're a writing stream or a non-seekable reading stream,
|
|
|
2367 meaning that we can't just process the whole input,
|
|
|
2368 rewind, and start over. */
|
|
|
2369
|
|
|
2370 if (CODING_SYSTEM_TYPE (str->codesys) == CODESYS_AUTODETECT ||
|
|
|
2371 str->eol_type == EOL_AUTODETECT)
|
|
|
2372 {
|
|
272
|
2373 Lisp_Object codesys;
|
|
259
|
2374
|
|
|
2375 XSETCODING_SYSTEM (codesys, str->codesys);
|
|
|
2376 detect_coding_type (&str->decst, src, n,
|
|
|
2377 CODING_SYSTEM_TYPE (str->codesys) !=
|
|
|
2378 CODESYS_AUTODETECT);
|
|
|
2379 if (CODING_SYSTEM_TYPE (str->codesys) == CODESYS_AUTODETECT &&
|
|
|
2380 str->decst.mask != ~0)
|
|
|
2381 /* #### This is cheesy. What we really ought to do is
|
|
|
2382 buffer up a certain amount of data so as to get a
|
|
|
2383 less random result. */
|
|
|
2384 codesys = coding_system_from_mask (str->decst.mask);
|
|
|
2385 str->eol_type = str->decst.eol_type;
|
|
|
2386 if (XCODING_SYSTEM (codesys) != str->codesys)
|
|
|
2387 {
|
|
|
2388 /* Preserve the CODING_STATE_END flag in case it was set.
|
|
|
2389 If we erase it, bad things might happen. */
|
|
|
2390 int was_end = str->flags & CODING_STATE_END;
|
|
|
2391 set_decoding_stream_coding_system (decoding, codesys);
|
|
|
2392 if (was_end)
|
|
|
2393 str->flags |= CODING_STATE_END;
|
|
|
2394 }
|
|
|
2395 }
|
|
|
2396
|
|
|
2397 switch (CODING_SYSTEM_TYPE (str->codesys))
|
|
|
2398 {
|
|
|
2399 #ifdef DEBUG_XEMACS
|
|
|
2400 case CODESYS_INTERNAL:
|
|
|
2401 Dynarr_add_many (dst, src, n);
|
|
|
2402 break;
|
|
|
2403 #endif
|
|
|
2404 case CODESYS_AUTODETECT:
|
|
|
2405 /* If we got this far and still haven't decided on the coding
|
|
|
2406 system, then do no conversion. */
|
|
|
2407 case CODESYS_NO_CONVERSION:
|
|
|
2408 decode_coding_no_conversion (decoding, src, dst, n);
|
|
|
2409 break;
|
|
|
2410 #ifdef MULE
|
|
|
2411 case CODESYS_SHIFT_JIS:
|
|
|
2412 decode_coding_sjis (decoding, src, dst, n);
|
|
|
2413 break;
|
|
|
2414 case CODESYS_BIG5:
|
|
|
2415 decode_coding_big5 (decoding, src, dst, n);
|
|
|
2416 break;
|
|
396
|
2417 case CODESYS_UCS4:
|
|
|
2418 decode_coding_ucs4 (decoding, src, dst, n);
|
|
|
2419 break;
|
|
|
2420 case CODESYS_UTF8:
|
|
|
2421 decode_coding_utf8 (decoding, src, dst, n);
|
|
|
2422 break;
|
|
259
|
2423 case CODESYS_CCL:
|
|
398
|
2424 str->ccl.last_block = str->flags & CODING_STATE_END;
|
|
|
2425 ccl_driver (&str->ccl, src, dst, n, 0, CCL_MODE_DECODING);
|
|
259
|
2426 break;
|
|
|
2427 case CODESYS_ISO2022:
|
|
|
2428 decode_coding_iso2022 (decoding, src, dst, n);
|
|
|
2429 break;
|
|
272
|
2430 #endif /* MULE */
|
|
259
|
2431 default:
|
|
|
2432 abort ();
|
|
|
2433 }
|
|
|
2434 }
|
|
|
2435
|
|
|
2436 DEFUN ("decode-coding-region", Fdecode_coding_region, 3, 4, 0, /*
|
|
|
2437 Decode the text between START and END which is encoded in CODING-SYSTEM.
|
|
|
2438 This is useful if you've read in encoded text from a file without decoding
|
|
|
2439 it (e.g. you read in a JIS-formatted file but used the `binary' or
|
|
|
2440 `no-conversion' coding system, so that it shows up as "^[$B!<!+^[(B").
|
|
|
2441 Return length of decoded text.
|
|
|
2442 BUFFER defaults to the current buffer if unspecified.
|
|
|
2443 */
|
|
|
2444 (start, end, coding_system, buffer))
|
|
|
2445 {
|
|
|
2446 Bufpos b, e;
|
|
|
2447 struct buffer *buf = decode_buffer (buffer, 0);
|
|
|
2448 Lisp_Object instream, lb_outstream, de_outstream, outstream;
|
|
|
2449 Lstream *istr, *ostr;
|
|
|
2450 struct gcpro gcpro1, gcpro2, gcpro3, gcpro4;
|
|
|
2451
|
|
|
2452 get_buffer_range_char (buf, start, end, &b, &e, 0);
|
|
|
2453
|
|
|
2454 barf_if_buffer_read_only (buf, b, e);
|
|
|
2455
|
|
|
2456 coding_system = Fget_coding_system (coding_system);
|
|
|
2457 instream = make_lisp_buffer_input_stream (buf, b, e, 0);
|
|
|
2458 lb_outstream = make_lisp_buffer_output_stream (buf, b, 0);
|
|
|
2459 de_outstream = make_decoding_output_stream (XLSTREAM (lb_outstream),
|
|
|
2460 coding_system);
|
|
|
2461 outstream = make_encoding_output_stream (XLSTREAM (de_outstream),
|
|
|
2462 Fget_coding_system (Qbinary));
|
|
|
2463 istr = XLSTREAM (instream);
|
|
|
2464 ostr = XLSTREAM (outstream);
|
|
|
2465 GCPRO4 (instream, lb_outstream, de_outstream, outstream);
|
|
|
2466
|
|
|
2467 /* The chain of streams looks like this:
|
|
|
2468
|
|
|
2469 [BUFFER] <----- send through
|
|
|
2470 ------> [ENCODE AS BINARY]
|
|
|
2471 ------> [DECODE AS SPECIFIED]
|
|
|
2472 ------> [BUFFER]
|
|
|
2473 */
|
|
|
2474
|
|
|
2475 while (1)
|
|
|
2476 {
|
|
|
2477 char tempbuf[1024]; /* some random amount */
|
|
|
2478 Bufpos newpos, even_newer_pos;
|
|
|
2479 Bufpos oldpos = lisp_buffer_stream_startpos (istr);
|
|
398
|
2480 ssize_t size_in_bytes = Lstream_read (istr, tempbuf, sizeof (tempbuf));
|
|
259
|
2481
|
|
|
2482 if (!size_in_bytes)
|
|
|
2483 break;
|
|
|
2484 newpos = lisp_buffer_stream_startpos (istr);
|
|
|
2485 Lstream_write (ostr, tempbuf, size_in_bytes);
|
|
|
2486 even_newer_pos = lisp_buffer_stream_startpos (istr);
|
|
|
2487 buffer_delete_range (buf, even_newer_pos - (newpos - oldpos),
|
|
|
2488 even_newer_pos, 0);
|
|
|
2489 }
|
|
|
2490 Lstream_close (istr);
|
|
|
2491 Lstream_close (ostr);
|
|
|
2492 UNGCPRO;
|
|
|
2493 Lstream_delete (istr);
|
|
|
2494 Lstream_delete (ostr);
|
|
|
2495 Lstream_delete (XLSTREAM (de_outstream));
|
|
|
2496 Lstream_delete (XLSTREAM (lb_outstream));
|
|
|
2497 return Qnil;
|
|
|
2498 }
|
|
|
2499
|
|
|
2500
|
|
|
2501 /************************************************************************/
|
|
|
2502 /* Converting to an external encoding ("encoding") */
|
|
|
2503 /************************************************************************/
|
|
|
2504
|
|
|
2505 /* An encoding stream is an output stream. When you create the
|
|
|
2506 stream, you specify the coding system that governs the encoding
|
|
|
2507 and another stream that the resulting encoded data is to be
|
|
|
2508 sent to, and then start sending data to it. */
|
|
|
2509
|
|
|
2510 #define ENCODING_STREAM_DATA(stream) LSTREAM_TYPE_DATA (stream, encoding)
|
|
|
2511
|
|
|
2512 struct encoding_stream
|
|
|
2513 {
|
|
|
2514 /* Coding system that governs the conversion. */
|
|
396
|
2515 Lisp_Coding_System *codesys;
|
|
259
|
2516
|
|
|
2517 /* Stream that we read the encoded data from or
|
|
|
2518 write the decoded data to. */
|
|
|
2519 Lstream *other_end;
|
|
|
2520
|
|
|
2521 /* If we are reading, then we can return only a fixed amount of
|
|
|
2522 data, so if the conversion resulted in too much data, we store it
|
|
|
2523 here for retrieval the next time around. */
|
|
|
2524 unsigned_char_dynarr *runoff;
|
|
|
2525
|
|
|
2526 /* FLAGS holds flags indicating the current state of the encoding.
|
|
|
2527 Some of these flags are dependent on the coding system. */
|
|
|
2528 unsigned int flags;
|
|
|
2529
|
|
|
2530 /* CH holds a partially built-up character. Since we only deal
|
|
|
2531 with one- and two-byte characters at the moment, we only use
|
|
|
2532 this to store the first byte of a two-byte character. */
|
|
|
2533 unsigned int ch;
|
|
|
2534 #ifdef MULE
|
|
|
2535 /* Additional information used by the ISO2022 encoder. */
|
|
|
2536 struct
|
|
|
2537 {
|
|
|
2538 /* CHARSET holds the character sets currently assigned to the G0
|
|
|
2539 through G3 registers. It is initialized from the array
|
|
|
2540 INITIAL_CHARSET in CODESYS. */
|
|
|
2541 Lisp_Object charset[4];
|
|
|
2542
|
|
|
2543 /* Which registers are currently invoked into the left (GL) and
|
|
|
2544 right (GR) halves of the 8-bit encoding space? */
|
|
|
2545 int register_left, register_right;
|
|
|
2546
|
|
|
2547 /* Whether we need to explicitly designate the charset in the
|
|
|
2548 G? register before using it. It is initialized from the
|
|
|
2549 array FORCE_CHARSET_ON_OUTPUT in CODESYS. */
|
|
|
2550 unsigned char force_charset_on_output[4];
|
|
|
2551
|
|
|
2552 /* Other state variables that need to be preserved across
|
|
|
2553 invocations. */
|
|
|
2554 Lisp_Object current_charset;
|
|
|
2555 int current_half;
|
|
|
2556 int current_char_boundary;
|
|
|
2557 } iso2022;
|
|
|
2558
|
|
|
2559 /* Additional information (the state of the running CCL program)
|
|
|
2560 used by the CCL encoder. */
|
|
|
2561 struct ccl_program ccl;
|
|
272
|
2562 #endif /* MULE */
|
|
259
|
2563 };
|
|
|
2564
|
|
398
|
2565 static ssize_t encoding_reader (Lstream *stream, unsigned char *data, size_t size);
|
|
|
2566 static ssize_t encoding_writer (Lstream *stream, const unsigned char *data,
|
|
|
2567 size_t size);
|
|
259
|
2568 static int encoding_rewinder (Lstream *stream);
|
|
|
2569 static int encoding_seekable_p (Lstream *stream);
|
|
|
2570 static int encoding_flusher (Lstream *stream);
|
|
|
2571 static int encoding_closer (Lstream *stream);
|
|
|
2572
|
|
398
|
2573 static Lisp_Object encoding_marker (Lisp_Object stream);
|
|
259
|
2574
|
|
|
2575 DEFINE_LSTREAM_IMPLEMENTATION ("encoding", lstream_encoding,
|
|
|
2576 sizeof (struct encoding_stream));
|
|
|
2577
|
|
|
2578 static Lisp_Object
|
|
398
|
2579 encoding_marker (Lisp_Object stream)
|
|
259
|
2580 {
|
|
|
2581 Lstream *str = ENCODING_STREAM_DATA (XLSTREAM (stream))->other_end;
|
|
|
2582 Lisp_Object str_obj;
|
|
|
2583
|
|
|
2584 /* We do not need to mark the coding systems or charsets stored
|
|
|
2585 within the stream because they are stored in a global list
|
|
|
2586 and automatically marked. */
|
|
|
2587
|
|
|
2588 XSETLSTREAM (str_obj, str);
|
|
398
|
2589 mark_object (str_obj);
|
|
259
|
2590 if (str->imp->marker)
|
|
398
|
2591 return (str->imp->marker) (str_obj);
|
|
259
|
2592 else
|
|
|
2593 return Qnil;
|
|
|
2594 }
|
|
|
2595
|
|
|
2596 /* Read SIZE bytes of data and store it into DATA. We are a encoding stream
|
|
|
2597 so we read data from the other end, encode it, and store it into DATA. */
|
|
|
2598
|
|
398
|
2599 static ssize_t
|
|
272
|
2600 encoding_reader (Lstream *stream, unsigned char *data, size_t size)
|
|
259
|
2601 {
|
|
|
2602 struct encoding_stream *str = ENCODING_STREAM_DATA (stream);
|
|
|
2603 unsigned char *orig_data = data;
|
|
398
|
2604 ssize_t read_size;
|
|
259
|
2605 int error_occurred = 0;
|
|
|
2606
|
|
|
2607 /* We need to interface to mule_encode(), which expects to take some
|
|
|
2608 amount of data and store the result into a Dynarr. We have
|
|
|
2609 mule_encode() store into str->runoff, and take data from there
|
|
|
2610 as necessary. */
|
|
|
2611
|
|
|
2612 /* We loop until we have enough data, reading chunks from the other
|
|
|
2613 end and encoding it. */
|
|
|
2614 while (1)
|
|
|
2615 {
|
|
|
2616 /* Take data from the runoff if we can. Make sure to take at
|
|
|
2617 most SIZE bytes, and delete the data from the runoff. */
|
|
|
2618 if (Dynarr_length (str->runoff) > 0)
|
|
|
2619 {
|
|
272
|
2620 int chunk = min ((int) size, Dynarr_length (str->runoff));
|
|
259
|
2621 memcpy (data, Dynarr_atp (str->runoff, 0), chunk);
|
|
|
2622 Dynarr_delete_many (str->runoff, 0, chunk);
|
|
|
2623 data += chunk;
|
|
|
2624 size -= chunk;
|
|
|
2625 }
|
|
|
2626
|
|
|
2627 if (size == 0)
|
|
|
2628 break; /* No more room for data */
|
|
|
2629
|
|
|
2630 if (str->flags & CODING_STATE_END)
|
|
|
2631 /* This means that on the previous iteration, we hit the EOF on
|
|
|
2632 the other end. We loop once more so that mule_encode() can
|
|
|
2633 output any final stuff it may be holding, or any "go back
|
|
|
2634 to a sane state" escape sequences. (This latter makes sense
|
|
|
2635 during encoding.) */
|
|
|
2636 break;
|
|
|
2637
|
|
|
2638 /* Exhausted the runoff, so get some more. DATA at least SIZE bytes
|
|
|
2639 left of storage in it, so it's OK to read directly into it.
|
|
|
2640 (We'll be overwriting above, after we've encoded it into the
|
|
|
2641 runoff.) */
|
|
|
2642 read_size = Lstream_read (str->other_end, data, size);
|
|
|
2643 if (read_size < 0)
|
|
|
2644 {
|
|
|
2645 error_occurred = 1;
|
|
|
2646 break;
|
|
|
2647 }
|
|
|
2648 if (read_size == 0)
|
|
|
2649 /* There might be some more end data produced in the translation.
|
|
|
2650 See the comment above. */
|
|
|
2651 str->flags |= CODING_STATE_END;
|
|
|
2652 mule_encode (stream, data, str->runoff, read_size);
|
|
|
2653 }
|
|
|
2654
|
|
|
2655 if (data == orig_data)
|
|
|
2656 return error_occurred ? -1 : 0;
|
|
|
2657 else
|
|
|
2658 return data - orig_data;
|
|
|
2659 }
|
|
|
2660
|
|
398
|
2661 static ssize_t
|
|
|
2662 encoding_writer (Lstream *stream, const unsigned char *data, size_t size)
|
|
259
|
2663 {
|
|
|
2664 struct encoding_stream *str = ENCODING_STREAM_DATA (stream);
|
|
398
|
2665 ssize_t retval;
|
|
259
|
2666
|
|
|
2667 /* Encode all our data into the runoff, and then attempt to write
|
|
|
2668 it all out to the other end. Remove whatever chunk we succeeded
|
|
|
2669 in writing. */
|
|
|
2670 mule_encode (stream, data, str->runoff, size);
|
|
|
2671 retval = Lstream_write (str->other_end, Dynarr_atp (str->runoff, 0),
|
|
|
2672 Dynarr_length (str->runoff));
|
|
|
2673 if (retval > 0)
|
|
|
2674 Dynarr_delete_many (str->runoff, 0, retval);
|
|
|
2675 /* Do NOT return retval. The return value indicates how much
|
|
|
2676 of the incoming data was written, not how many bytes were
|
|
|
2677 written. */
|
|
|
2678 return size;
|
|
|
2679 }
|
|
|
2680
|
|
|
2681 static void
|
|
|
2682 reset_encoding_stream (struct encoding_stream *str)
|
|
|
2683 {
|
|
272
|
2684 #ifdef MULE
|
|
259
|
2685 switch (CODING_SYSTEM_TYPE (str->codesys))
|
|
|
2686 {
|
|
|
2687 case CODESYS_ISO2022:
|
|
|
2688 {
|
|
|
2689 int i;
|
|
|
2690
|
|
|
2691 for (i = 0; i < 4; i++)
|
|
|
2692 {
|
|
|
2693 str->iso2022.charset[i] =
|
|
|
2694 CODING_SYSTEM_ISO2022_INITIAL_CHARSET (str->codesys, i);
|
|
|
2695 str->iso2022.force_charset_on_output[i] =
|
|
|
2696 CODING_SYSTEM_ISO2022_FORCE_CHARSET_ON_OUTPUT (str->codesys, i);
|
|
|
2697 }
|
|
|
2698 str->iso2022.register_left = 0;
|
|
|
2699 str->iso2022.register_right = 1;
|
|
|
2700 str->iso2022.current_charset = Qnil;
|
|
|
2701 str->iso2022.current_half = 0;
|
|
|
2702 str->iso2022.current_char_boundary = 1;
|
|
|
2703 break;
|
|
|
2704 }
|
|
|
2705 case CODESYS_CCL:
|
|
|
2706 setup_ccl_program (&str->ccl, CODING_SYSTEM_CCL_ENCODE (str->codesys));
|
|
|
2707 break;
|
|
|
2708 default:
|
|
|
2709 break;
|
|
|
2710 }
|
|
272
|
2711 #endif /* MULE */
|
|
259
|
2712
|
|
|
2713 str->flags = str->ch = 0;
|
|
|
2714 }
|
|
|
2715
|
|
|
2716 static int
|
|
|
2717 encoding_rewinder (Lstream *stream)
|
|
|
2718 {
|
|
|
2719 struct encoding_stream *str = ENCODING_STREAM_DATA (stream);
|
|
|
2720 reset_encoding_stream (str);
|
|
|
2721 Dynarr_reset (str->runoff);
|
|
|
2722 return Lstream_rewind (str->other_end);
|
|
|
2723 }
|
|
|
2724
|
|
|
2725 static int
|
|
|
2726 encoding_seekable_p (Lstream *stream)
|
|
|
2727 {
|
|
|
2728 struct encoding_stream *str = ENCODING_STREAM_DATA (stream);
|
|
|
2729 return Lstream_seekable_p (str->other_end);
|
|
|
2730 }
|
|
|
2731
|
|
|
2732 static int
|
|
|
2733 encoding_flusher (Lstream *stream)
|
|
|
2734 {
|
|
|
2735 struct encoding_stream *str = ENCODING_STREAM_DATA (stream);
|
|
|
2736 return Lstream_flush (str->other_end);
|
|
|
2737 }
|
|
|
2738
|
|
|
2739 static int
|
|
|
2740 encoding_closer (Lstream *stream)
|
|
|
2741 {
|
|
|
2742 struct encoding_stream *str = ENCODING_STREAM_DATA (stream);
|
|
|
2743 if (stream->flags & LSTREAM_FL_WRITE)
|
|
|
2744 {
|
|
|
2745 str->flags |= CODING_STATE_END;
|
|
|
2746 encoding_writer (stream, 0, 0);
|
|
|
2747 }
|
|
|
2748 Dynarr_free (str->runoff);
|
|
|
2749 return Lstream_close (str->other_end);
|
|
|
2750 }
|
|
|
2751
|
|
|
2752 Lisp_Object
|
|
|
2753 encoding_stream_coding_system (Lstream *stream)
|
|
|
2754 {
|
|
272
|
2755 Lisp_Object coding_system;
|
|
259
|
2756 struct encoding_stream *str = ENCODING_STREAM_DATA (stream);
|
|
|
2757
|
|
|
2758 XSETCODING_SYSTEM (coding_system, str->codesys);
|
|
|
2759 return coding_system;
|
|
|
2760 }
|
|
|
2761
|
|
|
2762 void
|
|
|
2763 set_encoding_stream_coding_system (Lstream *lstr, Lisp_Object codesys)
|
|
|
2764 {
|
|
396
|
2765 Lisp_Coding_System *cs = XCODING_SYSTEM (codesys);
|
|
259
|
2766 struct encoding_stream *str = ENCODING_STREAM_DATA (lstr);
|
|
|
2767 str->codesys = cs;
|
|
|
2768 reset_encoding_stream (str);
|
|
|
2769 }
|
|
|
2770
|
|
|
2771 static Lisp_Object
|
|
|
2772 make_encoding_stream_1 (Lstream *stream, Lisp_Object codesys,
|
|
398
|
2773 const char *mode)
|
|
259
|
2774 {
|
|
|
2775 Lstream *lstr = Lstream_new (lstream_encoding, mode);
|
|
|
2776 struct encoding_stream *str = ENCODING_STREAM_DATA (lstr);
|
|
|
2777 Lisp_Object obj;
|
|
|
2778
|
|
272
|
2779 xzero (*str);
|
|
259
|
2780 str->runoff = Dynarr_new (unsigned_char);
|
|
|
2781 str->other_end = stream;
|
|
|
2782 set_encoding_stream_coding_system (lstr, codesys);
|
|
|
2783 XSETLSTREAM (obj, lstr);
|
|
|
2784 return obj;
|
|
|
2785 }
|
|
|
2786
|
|
|
2787 Lisp_Object
|
|
|
2788 make_encoding_input_stream (Lstream *stream, Lisp_Object codesys)
|
|
|
2789 {
|
|
|
2790 return make_encoding_stream_1 (stream, codesys, "r");
|
|
|
2791 }
|
|
|
2792
|
|
|
2793 Lisp_Object
|
|
|
2794 make_encoding_output_stream (Lstream *stream, Lisp_Object codesys)
|
|
|
2795 {
|
|
|
2796 return make_encoding_stream_1 (stream, codesys, "w");
|
|
|
2797 }
|
|
|
2798
|
|
|
2799 /* Convert N bytes of internally-formatted data stored in SRC to an
|
|
|
2800 external format, according to the encoding stream ENCODING.
|
|
|
2801 Store the encoded data into DST. */
|
|
|
2802
|
|
|
2803 static void
|
|
398
|
2804 mule_encode (Lstream *encoding, const unsigned char *src,
|
|
259
|
2805 unsigned_char_dynarr *dst, unsigned int n)
|
|
|
2806 {
|
|
|
2807 struct encoding_stream *str = ENCODING_STREAM_DATA (encoding);
|
|
|
2808
|
|
|
2809 switch (CODING_SYSTEM_TYPE (str->codesys))
|
|
|
2810 {
|
|
|
2811 #ifdef DEBUG_XEMACS
|
|
|
2812 case CODESYS_INTERNAL:
|
|
|
2813 Dynarr_add_many (dst, src, n);
|
|
|
2814 break;
|
|
|
2815 #endif
|
|
|
2816 case CODESYS_AUTODETECT:
|
|
|
2817 /* If we got this far and still haven't decided on the coding
|
|
|
2818 system, then do no conversion. */
|
|
|
2819 case CODESYS_NO_CONVERSION:
|
|
|
2820 encode_coding_no_conversion (encoding, src, dst, n);
|
|
|
2821 break;
|
|
|
2822 #ifdef MULE
|
|
|
2823 case CODESYS_SHIFT_JIS:
|
|
|
2824 encode_coding_sjis (encoding, src, dst, n);
|
|
|
2825 break;
|
|
|
2826 case CODESYS_BIG5:
|
|
|
2827 encode_coding_big5 (encoding, src, dst, n);
|
|
|
2828 break;
|
|
396
|
2829 case CODESYS_UCS4:
|
|
|
2830 encode_coding_ucs4 (encoding, src, dst, n);
|
|
|
2831 break;
|
|
|
2832 case CODESYS_UTF8:
|
|
|
2833 encode_coding_utf8 (encoding, src, dst, n);
|
|
|
2834 break;
|
|
259
|
2835 case CODESYS_CCL:
|
|
398
|
2836 str->ccl.last_block = str->flags & CODING_STATE_END;
|
|
|
2837 ccl_driver (&str->ccl, src, dst, n, 0, CCL_MODE_ENCODING);
|
|
259
|
2838 break;
|
|
|
2839 case CODESYS_ISO2022:
|
|
|
2840 encode_coding_iso2022 (encoding, src, dst, n);
|
|
|
2841 break;
|
|
|
2842 #endif /* MULE */
|
|
|
2843 default:
|
|
|
2844 abort ();
|
|
|
2845 }
|
|
|
2846 }
|
|
|
2847
|
|
|
2848 DEFUN ("encode-coding-region", Fencode_coding_region, 3, 4, 0, /*
|
|
|
2849 Encode the text between START and END using CODING-SYSTEM.
|
|
|
2850 This will, for example, convert Japanese characters into stuff such as
|
|
|
2851 "^[$B!<!+^[(B" if you use the JIS encoding. Return length of encoded
|
|
|
2852 text. BUFFER defaults to the current buffer if unspecified.
|
|
|
2853 */
|
|
|
2854 (start, end, coding_system, buffer))
|
|
|
2855 {
|
|
|
2856 Bufpos b, e;
|
|
|
2857 struct buffer *buf = decode_buffer (buffer, 0);
|
|
|
2858 Lisp_Object instream, lb_outstream, de_outstream, outstream;
|
|
|
2859 Lstream *istr, *ostr;
|
|
|
2860 struct gcpro gcpro1, gcpro2, gcpro3, gcpro4;
|
|
|
2861
|
|
|
2862 get_buffer_range_char (buf, start, end, &b, &e, 0);
|
|
|
2863
|
|
|
2864 barf_if_buffer_read_only (buf, b, e);
|
|
|
2865
|
|
|
2866 coding_system = Fget_coding_system (coding_system);
|
|
|
2867 instream = make_lisp_buffer_input_stream (buf, b, e, 0);
|
|
|
2868 lb_outstream = make_lisp_buffer_output_stream (buf, b, 0);
|
|
|
2869 de_outstream = make_decoding_output_stream (XLSTREAM (lb_outstream),
|
|
|
2870 Fget_coding_system (Qbinary));
|
|
|
2871 outstream = make_encoding_output_stream (XLSTREAM (de_outstream),
|
|
|
2872 coding_system);
|
|
|
2873 istr = XLSTREAM (instream);
|
|
|
2874 ostr = XLSTREAM (outstream);
|
|
|
2875 GCPRO4 (instream, outstream, de_outstream, lb_outstream);
|
|
|
2876 /* The chain of streams looks like this:
|
|
|
2877
|
|
|
2878 [BUFFER] <----- send through
|
|
|
2879 ------> [ENCODE AS SPECIFIED]
|
|
|
2880 ------> [DECODE AS BINARY]
|
|
|
2881 ------> [BUFFER]
|
|
|
2882 */
|
|
|
2883 while (1)
|
|
|
2884 {
|
|
|
2885 char tempbuf[1024]; /* some random amount */
|
|
|
2886 Bufpos newpos, even_newer_pos;
|
|
|
2887 Bufpos oldpos = lisp_buffer_stream_startpos (istr);
|
|
398
|
2888 ssize_t size_in_bytes = Lstream_read (istr, tempbuf, sizeof (tempbuf));
|
|
259
|
2889
|
|
|
2890 if (!size_in_bytes)
|
|
|
2891 break;
|
|
|
2892 newpos = lisp_buffer_stream_startpos (istr);
|
|
|
2893 Lstream_write (ostr, tempbuf, size_in_bytes);
|
|
|
2894 even_newer_pos = lisp_buffer_stream_startpos (istr);
|
|
|
2895 buffer_delete_range (buf, even_newer_pos - (newpos - oldpos),
|
|
|
2896 even_newer_pos, 0);
|
|
|
2897 }
|
|
|
2898
|
|
|
2899 {
|
|
|
2900 Charcount retlen =
|
|
|
2901 lisp_buffer_stream_startpos (XLSTREAM (instream)) - b;
|
|
|
2902 Lstream_close (istr);
|
|
|
2903 Lstream_close (ostr);
|
|
|
2904 UNGCPRO;
|
|
|
2905 Lstream_delete (istr);
|
|
|
2906 Lstream_delete (ostr);
|
|
|
2907 Lstream_delete (XLSTREAM (de_outstream));
|
|
|
2908 Lstream_delete (XLSTREAM (lb_outstream));
|
|
|
2909 return make_int (retlen);
|
|
|
2910 }
|
|
|
2911 }
|
|
|
2912
|
|
|
2913 #ifdef MULE
|
|
|
2914
|
|
|
2915 /************************************************************************/
|
|
|
2916 /* Shift-JIS methods */
|
|
|
2917 /************************************************************************/
|
|
|
2918
|
|
|
2919 /* Shift-JIS is a coding system encoding three character sets: ASCII, right
|
|
|
2920 half of JISX0201-Kana, and JISX0208. An ASCII character is encoded
|
|
396
|
2921 as is. A character of JISX0201-Kana (DIMENSION1_CHARS94 character set) is
|
|
259
|
2922 encoded by "position-code + 0x80". A character of JISX0208
|
|
396
|
2923 (DIMENSION2_CHARS94 character set) is encoded in 2-byte but two
|
|
259
|
2924 position-codes are divided and shifted so that it fit in the range
|
|
|
2925 below.
|
|
|
2926
|
|
|
2927 --- CODE RANGE of Shift-JIS ---
|
|
|
2928 (character set) (range)
|
|
|
2929 ASCII 0x00 .. 0x7F
|
|
|
2930 JISX0201-Kana 0xA0 .. 0xDF
|
|
|
2931 JISX0208 (1st byte) 0x80 .. 0x9F and 0xE0 .. 0xEF
|
|
|
2932 (2nd byte) 0x40 .. 0x7E and 0x80 .. 0xFC
|
|
|
2933 -------------------------------
|
|
|
2934
|
|
|
2935 */
|
|
|
2936
|
|
|
2937 /* Is this the first byte of a Shift-JIS two-byte char? */
|
|
|
2938
|
|
|
2939 #define BYTE_SJIS_TWO_BYTE_1_P(c) \
|
|
|
2940 (((c) >= 0x81 && (c) <= 0x9F) || ((c) >= 0xE0 && (c) <= 0xEF))
|
|
|
2941
|
|
|
2942 /* Is this the second byte of a Shift-JIS two-byte char? */
|
|
|
2943
|
|
|
2944 #define BYTE_SJIS_TWO_BYTE_2_P(c) \
|
|
|
2945 (((c) >= 0x40 && (c) <= 0x7E) || ((c) >= 0x80 && (c) <= 0xFC))
|
|
|
2946
|
|
|
2947 #define BYTE_SJIS_KATAKANA_P(c) \
|
|
|
2948 ((c) >= 0xA1 && (c) <= 0xDF)
|
|
|
2949
|
|
|
2950 static int
|
|
398
|
2951 detect_coding_sjis (struct detection_state *st, const unsigned char *src,
|
|
259
|
2952 unsigned int n)
|
|
|
2953 {
|
|
|
2954 int c;
|
|
|
2955
|
|
|
2956 while (n--)
|
|
|
2957 {
|
|
|
2958 c = *src++;
|
|
|
2959 if (c == ISO_CODE_ESC || c == ISO_CODE_SI || c == ISO_CODE_SO)
|
|
|
2960 return 0;
|
|
|
2961 if (st->shift_jis.in_second_byte)
|
|
|
2962 {
|
|
|
2963 st->shift_jis.in_second_byte = 0;
|
|
|
2964 if (c < 0x40)
|
|
|
2965 return 0;
|
|
|
2966 }
|
|
|
2967 else if ((c >= 0x80 && c < 0xA0) || c >= 0xE0)
|
|
|
2968 st->shift_jis.in_second_byte = 1;
|
|
|
2969 }
|
|
|
2970 return CODING_CATEGORY_SHIFT_JIS_MASK;
|
|
|
2971 }
|
|
|
2972
|
|
|
2973 /* Convert Shift-JIS data to internal format. */
|
|
|
2974
|
|
|
2975 static void
|
|
398
|
2976 decode_coding_sjis (Lstream *decoding, const unsigned char *src,
|
|
259
|
2977 unsigned_char_dynarr *dst, unsigned int n)
|
|
|
2978 {
|
|
|
2979 unsigned char c;
|
|
|
2980 struct decoding_stream *str = DECODING_STREAM_DATA (decoding);
|
|
396
|
2981 unsigned int flags = str->flags;
|
|
|
2982 unsigned int ch = str->ch;
|
|
|
2983 eol_type_t eol_type = str->eol_type;
|
|
259
|
2984
|
|
|
2985 while (n--)
|
|
|
2986 {
|
|
|
2987 c = *src++;
|
|
|
2988
|
|
|
2989 if (ch)
|
|
|
2990 {
|
|
|
2991 /* Previous character was first byte of Shift-JIS Kanji char. */
|
|
|
2992 if (BYTE_SJIS_TWO_BYTE_2_P (c))
|
|
|
2993 {
|
|
|
2994 unsigned char e1, e2;
|
|
|
2995
|
|
|
2996 Dynarr_add (dst, LEADING_BYTE_JAPANESE_JISX0208);
|
|
|
2997 DECODE_SJIS (ch, c, e1, e2);
|
|
|
2998 Dynarr_add (dst, e1);
|
|
|
2999 Dynarr_add (dst, e2);
|
|
|
3000 }
|
|
|
3001 else
|
|
|
3002 {
|
|
|
3003 DECODE_ADD_BINARY_CHAR (ch, dst);
|
|
|
3004 DECODE_ADD_BINARY_CHAR (c, dst);
|
|
|
3005 }
|
|
|
3006 ch = 0;
|
|
|
3007 }
|
|
|
3008 else
|
|
|
3009 {
|
|
|
3010 DECODE_HANDLE_EOL_TYPE (eol_type, c, flags, dst);
|
|
|
3011 if (BYTE_SJIS_TWO_BYTE_1_P (c))
|
|
|
3012 ch = c;
|
|
|
3013 else if (BYTE_SJIS_KATAKANA_P (c))
|
|
|
3014 {
|
|
|
3015 Dynarr_add (dst, LEADING_BYTE_KATAKANA_JISX0201);
|
|
|
3016 Dynarr_add (dst, c);
|
|
|
3017 }
|
|
|
3018 else
|
|
|
3019 DECODE_ADD_BINARY_CHAR (c, dst);
|
|
|
3020 }
|
|
|
3021 label_continue_loop:;
|
|
|
3022 }
|
|
|
3023
|
|
|
3024 DECODE_HANDLE_END_OF_CONVERSION (flags, ch, dst);
|
|
|
3025
|
|
396
|
3026 str->flags = flags;
|
|
|
3027 str->ch = ch;
|
|
259
|
3028 }
|
|
|
3029
|
|
|
3030 /* Convert internally-formatted data to Shift-JIS. */
|
|
|
3031
|
|
|
3032 static void
|
|
398
|
3033 encode_coding_sjis (Lstream *encoding, const unsigned char *src,
|
|
259
|
3034 unsigned_char_dynarr *dst, unsigned int n)
|
|
|
3035 {
|
|
|
3036 unsigned char c;
|
|
|
3037 struct encoding_stream *str = ENCODING_STREAM_DATA (encoding);
|
|
396
|
3038 unsigned int flags = str->flags;
|
|
|
3039 unsigned int ch = str->ch;
|
|
|
3040 eol_type_t eol_type = CODING_SYSTEM_EOL_TYPE (str->codesys);
|
|
259
|
3041
|
|
|
3042 while (n--)
|
|
|
3043 {
|
|
|
3044 c = *src++;
|
|
|
3045 if (c == '\n')
|
|
|
3046 {
|
|
|
3047 if (eol_type != EOL_LF && eol_type != EOL_AUTODETECT)
|
|
|
3048 Dynarr_add (dst, '\r');
|
|
|
3049 if (eol_type != EOL_CR)
|
|
|
3050 Dynarr_add (dst, '\n');
|
|
|
3051 ch = 0;
|
|
|
3052 }
|
|
|
3053 else if (BYTE_ASCII_P (c))
|
|
|
3054 {
|
|
|
3055 Dynarr_add (dst, c);
|
|
|
3056 ch = 0;
|
|
|
3057 }
|
|
|
3058 else if (BUFBYTE_LEADING_BYTE_P (c))
|
|
|
3059 ch = (c == LEADING_BYTE_KATAKANA_JISX0201 ||
|
|
|
3060 c == LEADING_BYTE_JAPANESE_JISX0208_1978 ||
|
|
|
3061 c == LEADING_BYTE_JAPANESE_JISX0208) ? c : 0;
|
|
|
3062 else if (ch)
|
|
|
3063 {
|
|
|
3064 if (ch == LEADING_BYTE_KATAKANA_JISX0201)
|
|
|
3065 {
|
|
|
3066 Dynarr_add (dst, c);
|
|
|
3067 ch = 0;
|
|
|
3068 }
|
|
|
3069 else if (ch == LEADING_BYTE_JAPANESE_JISX0208_1978 ||
|
|
|
3070 ch == LEADING_BYTE_JAPANESE_JISX0208)
|
|
|
3071 ch = c;
|
|
|
3072 else
|
|
|
3073 {
|
|
|
3074 unsigned char j1, j2;
|
|
|
3075 ENCODE_SJIS (ch, c, j1, j2);
|
|
|
3076 Dynarr_add (dst, j1);
|
|
|
3077 Dynarr_add (dst, j2);
|
|
|
3078 ch = 0;
|
|
|
3079 }
|
|
|
3080 }
|
|
|
3081 }
|
|
|
3082
|
|
396
|
3083 str->flags = flags;
|
|
|
3084 str->ch = ch;
|
|
259
|
3085 }
|
|
|
3086
|
|
|
3087 DEFUN ("decode-shift-jis-char", Fdecode_shift_jis_char, 1, 1, 0, /*
|
|
|
3088 Decode a JISX0208 character of Shift-JIS coding-system.
|
|
|
3089 CODE is the character code in Shift-JIS as a cons of type bytes.
|
|
|
3090 Return the corresponding character.
|
|
|
3091 */
|
|
|
3092 (code))
|
|
|
3093 {
|
|
|
3094 unsigned char c1, c2, s1, s2;
|
|
|
3095
|
|
|
3096 CHECK_CONS (code);
|
|
|
3097 CHECK_INT (XCAR (code));
|
|
|
3098 CHECK_INT (XCDR (code));
|
|
|
3099 s1 = XINT (XCAR (code));
|
|
|
3100 s2 = XINT (XCDR (code));
|
|
|
3101 if (BYTE_SJIS_TWO_BYTE_1_P (s1) &&
|
|
|
3102 BYTE_SJIS_TWO_BYTE_2_P (s2))
|
|
|
3103 {
|
|
|
3104 DECODE_SJIS (s1, s2, c1, c2);
|
|
|
3105 return make_char (MAKE_CHAR (Vcharset_japanese_jisx0208,
|
|
|
3106 c1 & 0x7F, c2 & 0x7F));
|
|
|
3107 }
|
|
|
3108 else
|
|
|
3109 return Qnil;
|
|
|
3110 }
|
|
|
3111
|
|
|
3112 DEFUN ("encode-shift-jis-char", Fencode_shift_jis_char, 1, 1, 0, /*
|
|
|
3113 Encode a JISX0208 character CHAR to SHIFT-JIS coding-system.
|
|
|
3114 Return the corresponding character code in SHIFT-JIS as a cons of two bytes.
|
|
|
3115 */
|
|
|
3116 (ch))
|
|
|
3117 {
|
|
|
3118 Lisp_Object charset;
|
|
|
3119 int c1, c2, s1, s2;
|
|
|
3120
|
|
|
3121 CHECK_CHAR_COERCE_INT (ch);
|
|
|
3122 BREAKUP_CHAR (XCHAR (ch), charset, c1, c2);
|
|
|
3123 if (EQ (charset, Vcharset_japanese_jisx0208))
|
|
|
3124 {
|
|
|
3125 ENCODE_SJIS (c1 | 0x80, c2 | 0x80, s1, s2);
|
|
|
3126 return Fcons (make_int (s1), make_int (s2));
|
|
|
3127 }
|
|
|
3128 else
|
|
|
3129 return Qnil;
|
|
|
3130 }
|
|
|
3131
|
|
|
3132
|
|
|
3133 /************************************************************************/
|
|
|
3134 /* Big5 methods */
|
|
|
3135 /************************************************************************/
|
|
|
3136
|
|
|
3137 /* BIG5 is a coding system encoding two character sets: ASCII and
|
|
|
3138 Big5. An ASCII character is encoded as is. Big5 is a two-byte
|
|
|
3139 character set and is encoded in two-byte.
|
|
|
3140
|
|
|
3141 --- CODE RANGE of BIG5 ---
|
|
|
3142 (character set) (range)
|
|
|
3143 ASCII 0x00 .. 0x7F
|
|
|
3144 Big5 (1st byte) 0xA1 .. 0xFE
|
|
|
3145 (2nd byte) 0x40 .. 0x7E and 0xA1 .. 0xFE
|
|
|
3146 --------------------------
|
|
|
3147
|
|
|
3148 Since the number of characters in Big5 is larger than maximum
|
|
|
3149 characters in Emacs' charset (96x96), it can't be handled as one
|
|
380
|
3150 charset. So, in Emacs, Big5 is divided into two: `charset-big5-1'
|
|
396
|
3151 and `charset-big5-2'. Both <type>s are DIMENSION2_CHARS94. The former
|
|
259
|
3152 contains frequently used characters and the latter contains less
|
|
|
3153 frequently used characters. */
|
|
|
3154
|
|
|
3155 #define BYTE_BIG5_TWO_BYTE_1_P(c) \
|
|
|
3156 ((c) >= 0xA1 && (c) <= 0xFE)
|
|
|
3157
|
|
|
3158 /* Is this the second byte of a Shift-JIS two-byte char? */
|
|
|
3159
|
|
|
3160 #define BYTE_BIG5_TWO_BYTE_2_P(c) \
|
|
|
3161 (((c) >= 0x40 && (c) <= 0x7E) || ((c) >= 0xA1 && (c) <= 0xFE))
|
|
|
3162
|
|
|
3163 /* Number of Big5 characters which have the same code in 1st byte. */
|
|
|
3164
|
|
|
3165 #define BIG5_SAME_ROW (0xFF - 0xA1 + 0x7F - 0x40)
|
|
|
3166
|
|
|
3167 /* Code conversion macros. These are macros because they are used in
|
|
|
3168 inner loops during code conversion.
|
|
|
3169
|
|
|
3170 Note that temporary variables in macros introduce the classic
|
|
|
3171 dynamic-scoping problems with variable names. We use capital-
|
|
|
3172 lettered variables in the assumption that XEmacs does not use
|
|
|
3173 capital letters in variables except in a very formalized way
|
|
|
3174 (e.g. Qstring). */
|
|
|
3175
|
|
|
3176 /* Convert Big5 code (b1, b2) into its internal string representation
|
|
|
3177 (lb, c1, c2). */
|
|
|
3178
|
|
|
3179 /* There is a much simpler way to split the Big5 charset into two.
|
|
|
3180 For the moment I'm going to leave the algorithm as-is because it
|
|
|
3181 claims to separate out the most-used characters into a single
|
|
|
3182 charset, which perhaps will lead to optimizations in various
|
|
|
3183 places.
|
|
|
3184
|
|
|
3185 The way the algorithm works is something like this:
|
|
|
3186
|
|
|
3187 Big5 can be viewed as a 94x157 charset, where the row is
|
|
|
3188 encoded into the bytes 0xA1 .. 0xFE and the column is encoded
|
|
|
3189 into the bytes 0x40 .. 0x7E and 0xA1 .. 0xFE. As for frequency,
|
|
|
3190 the split between low and high column numbers is apparently
|
|
|
3191 meaningless; ascending rows produce less and less frequent chars.
|
|
|
3192 Therefore, we assign the lower half of rows (0xA1 .. 0xC8) to
|
|
|
3193 the first charset, and the upper half (0xC9 .. 0xFE) to the
|
|
|
3194 second. To do the conversion, we convert the character into
|
|
|
3195 a single number where 0 .. 156 is the first row, 157 .. 313
|
|
|
3196 is the second, etc. That way, the characters are ordered by
|
|
|
3197 decreasing frequency. Then we just chop the space in two
|
|
|
3198 and coerce the result into a 94x94 space.
|
|
|
3199 */
|
|
|
3200
|
|
|
3201 #define DECODE_BIG5(b1, b2, lb, c1, c2) do \
|
|
|
3202 { \
|
|
|
3203 int B1 = b1, B2 = b2; \
|
|
|
3204 unsigned int I \
|
|
|
3205 = (B1 - 0xA1) * BIG5_SAME_ROW + B2 - (B2 < 0x7F ? 0x40 : 0x62); \
|
|
|
3206 \
|
|
|
3207 if (B1 < 0xC9) \
|
|
|
3208 { \
|
|
|
3209 lb = LEADING_BYTE_CHINESE_BIG5_1; \
|
|
|
3210 } \
|
|
|
3211 else \
|
|
|
3212 { \
|
|
|
3213 lb = LEADING_BYTE_CHINESE_BIG5_2; \
|
|
|
3214 I -= (BIG5_SAME_ROW) * (0xC9 - 0xA1); \
|
|
|
3215 } \
|
|
|
3216 c1 = I / (0xFF - 0xA1) + 0xA1; \
|
|
|
3217 c2 = I % (0xFF - 0xA1) + 0xA1; \
|
|
|
3218 } while (0)
|
|
|
3219
|
|
|
3220 /* Convert the internal string representation of a Big5 character
|
|
|
3221 (lb, c1, c2) into Big5 code (b1, b2). */
|
|
|
3222
|
|
|
3223 #define ENCODE_BIG5(lb, c1, c2, b1, b2) do \
|
|
|
3224 { \
|
|
|
3225 unsigned int I = ((c1) - 0xA1) * (0xFF - 0xA1) + ((c2) - 0xA1); \
|
|
|
3226 \
|
|
|
3227 if (lb == LEADING_BYTE_CHINESE_BIG5_2) \
|
|
|
3228 { \
|
|
|
3229 I += BIG5_SAME_ROW * (0xC9 - 0xA1); \
|
|
|
3230 } \
|
|
|
3231 b1 = I / BIG5_SAME_ROW + 0xA1; \
|
|
|
3232 b2 = I % BIG5_SAME_ROW; \
|
|
|
3233 b2 += b2 < 0x3F ? 0x40 : 0x62; \
|
|
|
3234 } while (0)
|
|
|
3235
|
|
|
3236 static int
|
|
398
|
3237 detect_coding_big5 (struct detection_state *st, const unsigned char *src,
|
|
259
|
3238 unsigned int n)
|
|
|
3239 {
|
|
|
3240 int c;
|
|
|
3241
|
|
|
3242 while (n--)
|
|
|
3243 {
|
|
|
3244 c = *src++;
|
|
|
3245 if (c == ISO_CODE_ESC || c == ISO_CODE_SI || c == ISO_CODE_SO ||
|
|
|
3246 (c >= 0x80 && c <= 0xA0))
|
|
|
3247 return 0;
|
|
|
3248 if (st->big5.in_second_byte)
|
|
|
3249 {
|
|
|
3250 st->big5.in_second_byte = 0;
|
|
|
3251 if (c < 0x40 || (c >= 0x80 && c <= 0xA0))
|
|
|
3252 return 0;
|
|
|
3253 }
|
|
|
3254 else if (c >= 0xA1)
|
|
|
3255 st->big5.in_second_byte = 1;
|
|
|
3256 }
|
|
|
3257 return CODING_CATEGORY_BIG5_MASK;
|
|
|
3258 }
|
|
|
3259
|
|
|
3260 /* Convert Big5 data to internal format. */
|
|
|
3261
|
|
|
3262 static void
|
|
398
|
3263 decode_coding_big5 (Lstream *decoding, const unsigned char *src,
|
|
259
|
3264 unsigned_char_dynarr *dst, unsigned int n)
|
|
|
3265 {
|
|
|
3266 unsigned char c;
|
|
|
3267 struct decoding_stream *str = DECODING_STREAM_DATA (decoding);
|
|
396
|
3268 unsigned int flags = str->flags;
|
|
|
3269 unsigned int ch = str->ch;
|
|
|
3270 eol_type_t eol_type = str->eol_type;
|
|
259
|
3271
|
|
|
3272 while (n--)
|
|
|
3273 {
|
|
|
3274 c = *src++;
|
|
|
3275 if (ch)
|
|
|
3276 {
|
|
|
3277 /* Previous character was first byte of Big5 char. */
|
|
|
3278 if (BYTE_BIG5_TWO_BYTE_2_P (c))
|
|
|
3279 {
|
|
|
3280 unsigned char b1, b2, b3;
|
|
|
3281 DECODE_BIG5 (ch, c, b1, b2, b3);
|
|
|
3282 Dynarr_add (dst, b1);
|
|
|
3283 Dynarr_add (dst, b2);
|
|
|
3284 Dynarr_add (dst, b3);
|
|
|
3285 }
|
|
|
3286 else
|
|
|
3287 {
|
|
|
3288 DECODE_ADD_BINARY_CHAR (ch, dst);
|
|
|
3289 DECODE_ADD_BINARY_CHAR (c, dst);
|
|
|
3290 }
|
|
|
3291 ch = 0;
|
|
|
3292 }
|
|
|
3293 else
|
|
|
3294 {
|
|
|
3295 DECODE_HANDLE_EOL_TYPE (eol_type, c, flags, dst);
|
|
|
3296 if (BYTE_BIG5_TWO_BYTE_1_P (c))
|
|
|
3297 ch = c;
|
|
|
3298 else
|
|
|
3299 DECODE_ADD_BINARY_CHAR (c, dst);
|
|
|
3300 }
|
|
|
3301 label_continue_loop:;
|
|
|
3302 }
|
|
|
3303
|
|
|
3304 DECODE_HANDLE_END_OF_CONVERSION (flags, ch, dst);
|
|
|
3305
|
|
396
|
3306 str->flags = flags;
|
|
|
3307 str->ch = ch;
|
|
259
|
3308 }
|
|
|
3309
|
|
|
3310 /* Convert internally-formatted data to Big5. */
|
|
|
3311
|
|
|
3312 static void
|
|
398
|
3313 encode_coding_big5 (Lstream *encoding, const unsigned char *src,
|
|
259
|
3314 unsigned_char_dynarr *dst, unsigned int n)
|
|
|
3315 {
|
|
|
3316 unsigned char c;
|
|
|
3317 struct encoding_stream *str = ENCODING_STREAM_DATA (encoding);
|
|
396
|
3318 unsigned int flags = str->flags;
|
|
|
3319 unsigned int ch = str->ch;
|
|
|
3320 eol_type_t eol_type = CODING_SYSTEM_EOL_TYPE (str->codesys);
|
|
259
|
3321
|
|
|
3322 while (n--)
|
|
|
3323 {
|
|
|
3324 c = *src++;
|
|
|
3325 if (c == '\n')
|
|
|
3326 {
|
|
|
3327 if (eol_type != EOL_LF && eol_type != EOL_AUTODETECT)
|
|
|
3328 Dynarr_add (dst, '\r');
|
|
|
3329 if (eol_type != EOL_CR)
|
|
|
3330 Dynarr_add (dst, '\n');
|
|
|
3331 }
|
|
|
3332 else if (BYTE_ASCII_P (c))
|
|
|
3333 {
|
|
|
3334 /* ASCII. */
|
|
|
3335 Dynarr_add (dst, c);
|
|
|
3336 }
|
|
|
3337 else if (BUFBYTE_LEADING_BYTE_P (c))
|
|
|
3338 {
|
|
|
3339 if (c == LEADING_BYTE_CHINESE_BIG5_1 ||
|
|
|
3340 c == LEADING_BYTE_CHINESE_BIG5_2)
|
|
|
3341 {
|
|
|
3342 /* A recognized leading byte. */
|
|
|
3343 ch = c;
|
|
|
3344 continue; /* not done with this character. */
|
|
|
3345 }
|
|
|
3346 /* otherwise just ignore this character. */
|
|
|
3347 }
|
|
|
3348 else if (ch == LEADING_BYTE_CHINESE_BIG5_1 ||
|
|
|
3349 ch == LEADING_BYTE_CHINESE_BIG5_2)
|
|
|
3350 {
|
|
|
3351 /* Previous char was a recognized leading byte. */
|
|
|
3352 ch = (ch << 8) | c;
|
|
|
3353 continue; /* not done with this character. */
|
|
|
3354 }
|
|
|
3355 else if (ch)
|
|
|
3356 {
|
|
|
3357 /* Encountering second byte of a Big5 character. */
|
|
|
3358 unsigned char b1, b2;
|
|
|
3359
|
|
|
3360 ENCODE_BIG5 (ch >> 8, ch & 0xFF, c, b1, b2);
|
|
|
3361 Dynarr_add (dst, b1);
|
|
|
3362 Dynarr_add (dst, b2);
|
|
|
3363 }
|
|
|
3364
|
|
|
3365 ch = 0;
|
|
|
3366 }
|
|
|
3367
|
|
396
|
3368 str->flags = flags;
|
|
|
3369 str->ch = ch;
|
|
259
|
3370 }
|
|
|
3371
|
|
|
3372
|
|
|
3373 DEFUN ("decode-big5-char", Fdecode_big5_char, 1, 1, 0, /*
|
|
|
3374 Decode a Big5 character CODE of BIG5 coding-system.
|
|
|
3375 CODE is the character code in BIG5, a cons of two integers.
|
|
|
3376 Return the corresponding character.
|
|
|
3377 */
|
|
|
3378 (code))
|
|
|
3379 {
|
|
|
3380 unsigned char c1, c2, b1, b2;
|
|
|
3381
|
|
|
3382 CHECK_CONS (code);
|
|
|
3383 CHECK_INT (XCAR (code));
|
|
|
3384 CHECK_INT (XCDR (code));
|
|
|
3385 b1 = XINT (XCAR (code));
|
|
|
3386 b2 = XINT (XCDR (code));
|
|
|
3387 if (BYTE_BIG5_TWO_BYTE_1_P (b1) &&
|
|
|
3388 BYTE_BIG5_TWO_BYTE_2_P (b2))
|
|
|
3389 {
|
|
|
3390 int leading_byte;
|
|
|
3391 Lisp_Object charset;
|
|
|
3392 DECODE_BIG5 (b1, b2, leading_byte, c1, c2);
|
|
|
3393 charset = CHARSET_BY_LEADING_BYTE (leading_byte);
|
|
|
3394 return make_char (MAKE_CHAR (charset, c1 & 0x7F, c2 & 0x7F));
|
|
|
3395 }
|
|
|
3396 else
|
|
|
3397 return Qnil;
|
|
|
3398 }
|
|
|
3399
|
|
|
3400 DEFUN ("encode-big5-char", Fencode_big5_char, 1, 1, 0, /*
|
|
|
3401 Encode the Big5 character CH to BIG5 coding-system.
|
|
|
3402 Return the corresponding character code in Big5.
|
|
|
3403 */
|
|
|
3404 (ch))
|
|
|
3405 {
|
|
|
3406 Lisp_Object charset;
|
|
|
3407 int c1, c2, b1, b2;
|
|
|
3408
|
|
|
3409 CHECK_CHAR_COERCE_INT (ch);
|
|
|
3410 BREAKUP_CHAR (XCHAR (ch), charset, c1, c2);
|
|
|
3411 if (EQ (charset, Vcharset_chinese_big5_1) ||
|
|
|
3412 EQ (charset, Vcharset_chinese_big5_2))
|
|
|
3413 {
|
|
|
3414 ENCODE_BIG5 (XCHARSET_LEADING_BYTE (charset), c1 | 0x80, c2 | 0x80,
|
|
|
3415 b1, b2);
|
|
|
3416 return Fcons (make_int (b1), make_int (b2));
|
|
|
3417 }
|
|
|
3418 else
|
|
|
3419 return Qnil;
|
|
|
3420 }
|
|
|
3421
|
|
|
3422
|
|
|
3423 /************************************************************************/
|
|
396
|
3424 /* UCS-4 methods */
|
|
|
3425 /* */
|
|
|
3426 /* UCS-4 character codes are implemented as nonnegative integers. */
|
|
|
3427 /* */
|
|
|
3428 /************************************************************************/
|
|
|
3429
|
|
|
3430
|
|
|
3431 DEFUN ("set-ucs-char", Fset_ucs_char, 2, 2, 0, /*
|
|
|
3432 Map UCS-4 code CODE to Mule character CHARACTER.
|
|
|
3433
|
|
|
3434 Return T on success, NIL on failure.
|
|
|
3435 */
|
|
|
3436 (code, character))
|
|
|
3437 {
|
|
|
3438 unsigned int c;
|
|
|
3439
|
|
|
3440 CHECK_CHAR (character);
|
|
|
3441 CHECK_INT (code);
|
|
|
3442 c = XINT (code);
|
|
|
3443
|
|
398
|
3444 if (c < sizeof (fcd->ucs_to_mule_table))
|
|
396
|
3445 {
|
|
398
|
3446 fcd->ucs_to_mule_table[c] = character;
|
|
396
|
3447 return Qt;
|
|
|
3448 }
|
|
|
3449 else
|
|
|
3450 return Qnil;
|
|
|
3451 }
|
|
|
3452
|
|
|
3453 static Lisp_Object
|
|
|
3454 ucs_to_char (unsigned long code)
|
|
|
3455 {
|
|
398
|
3456 if (code < sizeof (fcd->ucs_to_mule_table))
|
|
396
|
3457 {
|
|
398
|
3458 return fcd->ucs_to_mule_table[code];
|
|
396
|
3459 }
|
|
|
3460 else if ((0xe00000 <= code) && (code <= 0xe00000 + 94 * 94 * 14))
|
|
|
3461 {
|
|
|
3462 unsigned int c;
|
|
|
3463
|
|
|
3464 code -= 0xe00000;
|
|
|
3465 c = code % (94 * 94);
|
|
|
3466 return make_char
|
|
|
3467 (MAKE_CHAR (CHARSET_BY_ATTRIBUTES
|
|
|
3468 (CHARSET_TYPE_94X94, code / (94 * 94) + '@',
|
|
|
3469 CHARSET_LEFT_TO_RIGHT),
|
|
|
3470 c / 94 + 33, c % 94 + 33));
|
|
|
3471 }
|
|
|
3472 else
|
|
|
3473 return Qnil;
|
|
|
3474 }
|
|
|
3475
|
|
|
3476 DEFUN ("ucs-char", Fucs_char, 1, 1, 0, /*
|
|
|
3477 Return Mule character corresponding to UCS code CODE (a positive integer).
|
|
|
3478 */
|
|
|
3479 (code))
|
|
|
3480 {
|
|
|
3481 CHECK_NATNUM (code);
|
|
|
3482 return ucs_to_char (XINT (code));
|
|
|
3483 }
|
|
|
3484
|
|
|
3485 DEFUN ("set-char-ucs", Fset_char_ucs, 2, 2, 0, /*
|
|
|
3486 Map Mule character CHARACTER to UCS code CODE (a positive integer).
|
|
|
3487 */
|
|
|
3488 (character, code))
|
|
|
3489 {
|
|
|
3490 /* #### Isn't this gilding the lily? Fput_char_table checks its args.
|
|
|
3491 Fset_char_ucs is more restrictive on index arg, but should
|
|
|
3492 check code arg in a char_table method. */
|
|
|
3493 CHECK_CHAR (character);
|
|
|
3494 CHECK_NATNUM (code);
|
|
|
3495 return Fput_char_table (character, code, mule_to_ucs_table);
|
|
|
3496 }
|
|
|
3497
|
|
|
3498 DEFUN ("char-ucs", Fchar_ucs, 1, 1, 0, /*
|
|
|
3499 Return the UCS code (a positive integer) corresponding to CHARACTER.
|
|
|
3500 */
|
|
|
3501 (character))
|
|
|
3502 {
|
|
|
3503 return Fget_char_table (character, mule_to_ucs_table);
|
|
|
3504 }
|
|
|
3505
|
|
|
3506 /* Decode a UCS-4 character into a buffer. If the lookup fails, use
|
|
398
|
3507 <GETA MARK> (U+3013) of JIS X 0208, which means correct character
|
|
|
3508 is not found, instead.
|
|
396
|
3509 #### do something more appropriate (use blob?)
|
|
|
3510 Danger, Will Robinson! Data loss. Should we signal user? */
|
|
|
3511 static void
|
|
|
3512 decode_ucs4 (unsigned long ch, unsigned_char_dynarr *dst)
|
|
|
3513 {
|
|
|
3514 Lisp_Object chr = ucs_to_char (ch);
|
|
|
3515
|
|
|
3516 if (! NILP (chr))
|
|
|
3517 {
|
|
|
3518 Bufbyte work[MAX_EMCHAR_LEN];
|
|
|
3519 int len;
|
|
|
3520
|
|
|
3521 ch = XCHAR (chr);
|
|
|
3522 len = (ch < 128) ?
|
|
|
3523 simple_set_charptr_emchar (work, ch) :
|
|
|
3524 non_ascii_set_charptr_emchar (work, ch);
|
|
|
3525 Dynarr_add_many (dst, work, len);
|
|
|
3526 }
|
|
|
3527 else
|
|
|
3528 {
|
|
|
3529 Dynarr_add (dst, LEADING_BYTE_JAPANESE_JISX0208);
|
|
|
3530 Dynarr_add (dst, 34 + 128);
|
|
|
3531 Dynarr_add (dst, 46 + 128);
|
|
|
3532 }
|
|
|
3533 }
|
|
|
3534
|
|
|
3535 static unsigned long
|
|
|
3536 mule_char_to_ucs4 (Lisp_Object charset,
|
|
|
3537 unsigned char h, unsigned char l)
|
|
|
3538 {
|
|
|
3539 Lisp_Object code
|
|
|
3540 = Fget_char_table (make_char (MAKE_CHAR (charset, h & 127, l & 127)),
|
|
|
3541 mule_to_ucs_table);
|
|
|
3542
|
|
|
3543 if (INTP (code))
|
|
|
3544 {
|
|
|
3545 return XINT (code);
|
|
|
3546 }
|
|
|
3547 else if ( (XCHARSET_DIMENSION (charset) == 2) &&
|
|
|
3548 (XCHARSET_CHARS (charset) == 94) )
|
|
|
3549 {
|
|
|
3550 unsigned char final = XCHARSET_FINAL (charset);
|
|
|
3551
|
|
|
3552 if ( ('@' <= final) && (final < 0x7f) )
|
|
|
3553 {
|
|
|
3554 return 0xe00000 + (final - '@') * 94 * 94
|
|
|
3555 + ((h & 127) - 33) * 94 + (l & 127) - 33;
|
|
|
3556 }
|
|
|
3557 else
|
|
|
3558 {
|
|
|
3559 return '?';
|
|
|
3560 }
|
|
|
3561 }
|
|
|
3562 else
|
|
|
3563 {
|
|
|
3564 return '?';
|
|
|
3565 }
|
|
|
3566 }
|
|
|
3567
|
|
|
3568 static void
|
|
|
3569 encode_ucs4 (Lisp_Object charset,
|
|
|
3570 unsigned char h, unsigned char l, unsigned_char_dynarr *dst)
|
|
|
3571 {
|
|
|
3572 unsigned long code = mule_char_to_ucs4 (charset, h, l);
|
|
|
3573 Dynarr_add (dst, code >> 24);
|
|
|
3574 Dynarr_add (dst, (code >> 16) & 255);
|
|
|
3575 Dynarr_add (dst, (code >> 8) & 255);
|
|
|
3576 Dynarr_add (dst, code & 255);
|
|
|
3577 }
|
|
|
3578
|
|
|
3579 static int
|
|
398
|
3580 detect_coding_ucs4 (struct detection_state *st, const unsigned char *src,
|
|
396
|
3581 unsigned int n)
|
|
|
3582 {
|
|
|
3583 while (n--)
|
|
|
3584 {
|
|
|
3585 int c = *src++;
|
|
|
3586 switch (st->ucs4.in_byte)
|
|
|
3587 {
|
|
|
3588 case 0:
|
|
|
3589 if (c >= 128)
|
|
|
3590 return 0;
|
|
|
3591 else
|
|
|
3592 st->ucs4.in_byte++;
|
|
|
3593 break;
|
|
|
3594 case 3:
|
|
|
3595 st->ucs4.in_byte = 0;
|
|
|
3596 break;
|
|
|
3597 default:
|
|
|
3598 st->ucs4.in_byte++;
|
|
|
3599 }
|
|
|
3600 }
|
|
|
3601 return CODING_CATEGORY_UCS4_MASK;
|
|
|
3602 }
|
|
|
3603
|
|
|
3604 static void
|
|
398
|
3605 decode_coding_ucs4 (Lstream *decoding, const unsigned char *src,
|
|
396
|
3606 unsigned_char_dynarr *dst, unsigned int n)
|
|
|
3607 {
|
|
|
3608 struct decoding_stream *str = DECODING_STREAM_DATA (decoding);
|
|
|
3609 unsigned int flags = str->flags;
|
|
|
3610 unsigned int ch = str->ch;
|
|
398
|
3611 unsigned char counter = str->counter;
|
|
396
|
3612
|
|
|
3613 while (n--)
|
|
|
3614 {
|
|
|
3615 unsigned char c = *src++;
|
|
398
|
3616 switch (counter)
|
|
396
|
3617 {
|
|
|
3618 case 0:
|
|
|
3619 ch = c;
|
|
398
|
3620 counter = 3;
|
|
396
|
3621 break;
|
|
|
3622 case 1:
|
|
|
3623 decode_ucs4 ( ( ch << 8 ) | c, dst);
|
|
|
3624 ch = 0;
|
|
398
|
3625 counter = 0;
|
|
396
|
3626 break;
|
|
|
3627 default:
|
|
|
3628 ch = ( ch << 8 ) | c;
|
|
398
|
3629 counter--;
|
|
396
|
3630 }
|
|
|
3631 }
|
|
398
|
3632 if (counter & CODING_STATE_END)
|
|
396
|
3633 DECODE_OUTPUT_PARTIAL_CHAR (ch);
|
|
|
3634
|
|
|
3635 str->flags = flags;
|
|
|
3636 str->ch = ch;
|
|
398
|
3637 str->counter = counter;
|
|
396
|
3638 }
|
|
|
3639
|
|
|
3640 static void
|
|
398
|
3641 encode_coding_ucs4 (Lstream *encoding, const unsigned char *src,
|
|
396
|
3642 unsigned_char_dynarr *dst, unsigned int n)
|
|
|
3643 {
|
|
|
3644 struct encoding_stream *str = ENCODING_STREAM_DATA (encoding);
|
|
|
3645 unsigned int flags = str->flags;
|
|
|
3646 unsigned int ch = str->ch;
|
|
|
3647 unsigned char char_boundary = str->iso2022.current_char_boundary;
|
|
|
3648 Lisp_Object charset = str->iso2022.current_charset;
|
|
|
3649
|
|
|
3650 #ifdef ENABLE_COMPOSITE_CHARS
|
|
|
3651 /* flags for handling composite chars. We do a little switcharoo
|
|
|
3652 on the source while we're outputting the composite char. */
|
|
|
3653 unsigned int saved_n = 0;
|
|
398
|
3654 const unsigned char *saved_src = NULL;
|
|
396
|
3655 int in_composite = 0;
|
|
|
3656
|
|
|
3657 back_to_square_n:
|
|
|
3658 #endif
|
|
|
3659
|
|
|
3660 while (n--)
|
|
|
3661 {
|
|
|
3662 unsigned char c = *src++;
|
|
|
3663
|
|
|
3664 if (BYTE_ASCII_P (c))
|
|
|
3665 { /* Processing ASCII character */
|
|
|
3666 ch = 0;
|
|
|
3667 encode_ucs4 (Vcharset_ascii, c, 0, dst);
|
|
|
3668 char_boundary = 1;
|
|
|
3669 }
|
|
|
3670 else if (BUFBYTE_LEADING_BYTE_P (c) || BUFBYTE_LEADING_BYTE_P (ch))
|
|
|
3671 { /* Processing Leading Byte */
|
|
|
3672 ch = 0;
|
|
|
3673 charset = CHARSET_BY_LEADING_BYTE (c);
|
|
|
3674 if (LEADING_BYTE_PREFIX_P(c))
|
|
|
3675 ch = c;
|
|
|
3676 char_boundary = 0;
|
|
|
3677 }
|
|
|
3678 else
|
|
|
3679 { /* Processing Non-ASCII character */
|
|
|
3680 char_boundary = 1;
|
|
|
3681 if (EQ (charset, Vcharset_control_1))
|
|
|
3682 {
|
|
|
3683 encode_ucs4 (Vcharset_control_1, c, 0, dst);
|
|
|
3684 }
|
|
|
3685 else
|
|
|
3686 {
|
|
|
3687 switch (XCHARSET_REP_BYTES (charset))
|
|
|
3688 {
|
|
|
3689 case 2:
|
|
|
3690 encode_ucs4 (charset, c, 0, dst);
|
|
|
3691 break;
|
|
|
3692 case 3:
|
|
|
3693 if (XCHARSET_PRIVATE_P (charset))
|
|
|
3694 {
|
|
|
3695 encode_ucs4 (charset, c, 0, dst);
|
|
|
3696 ch = 0;
|
|
|
3697 }
|
|
|
3698 else if (ch)
|
|
|
3699 {
|
|
|
3700 #ifdef ENABLE_COMPOSITE_CHARS
|
|
|
3701 if (EQ (charset, Vcharset_composite))
|
|
|
3702 {
|
|
|
3703 if (in_composite)
|
|
|
3704 {
|
|
|
3705 /* #### Bother! We don't know how to
|
|
|
3706 handle this yet. */
|
|
|
3707 Dynarr_add (dst, 0);
|
|
|
3708 Dynarr_add (dst, 0);
|
|
|
3709 Dynarr_add (dst, 0);
|
|
|
3710 Dynarr_add (dst, '~');
|
|
|
3711 }
|
|
|
3712 else
|
|
|
3713 {
|
|
|
3714 Emchar emch = MAKE_CHAR (Vcharset_composite,
|
|
|
3715 ch & 0x7F, c & 0x7F);
|
|
|
3716 Lisp_Object lstr = composite_char_string (emch);
|
|
|
3717 saved_n = n;
|
|
|
3718 saved_src = src;
|
|
|
3719 in_composite = 1;
|
|
|
3720 src = XSTRING_DATA (lstr);
|
|
|
3721 n = XSTRING_LENGTH (lstr);
|
|
|
3722 }
|
|
|
3723 }
|
|
|
3724 else
|
|
|
3725 #endif /* ENABLE_COMPOSITE_CHARS */
|
|
|
3726 {
|
|
|
3727 encode_ucs4(charset, ch, c, dst);
|
|
|
3728 }
|
|
|
3729 ch = 0;
|
|
|
3730 }
|
|
|
3731 else
|
|
|
3732 {
|
|
|
3733 ch = c;
|
|
|
3734 char_boundary = 0;
|
|
|
3735 }
|
|
|
3736 break;
|
|
|
3737 case 4:
|
|
|
3738 if (ch)
|
|
|
3739 {
|
|
|
3740 encode_ucs4 (charset, ch, c, dst);
|
|
|
3741 ch = 0;
|
|
|
3742 }
|
|
|
3743 else
|
|
|
3744 {
|
|
|
3745 ch = c;
|
|
|
3746 char_boundary = 0;
|
|
|
3747 }
|
|
|
3748 break;
|
|
|
3749 default:
|
|
|
3750 abort ();
|
|
|
3751 }
|
|
|
3752 }
|
|
|
3753 }
|
|
|
3754 }
|
|
|
3755
|
|
|
3756 #ifdef ENABLE_COMPOSITE_CHARS
|
|
|
3757 if (in_composite)
|
|
|
3758 {
|
|
|
3759 n = saved_n;
|
|
|
3760 src = saved_src;
|
|
|
3761 in_composite = 0;
|
|
|
3762 goto back_to_square_n; /* Wheeeeeeeee ..... */
|
|
|
3763 }
|
|
|
3764 #endif /* ENABLE_COMPOSITE_CHARS */
|
|
|
3765
|
|
|
3766 str->flags = flags;
|
|
|
3767 str->ch = ch;
|
|
|
3768 str->iso2022.current_char_boundary = char_boundary;
|
|
|
3769 str->iso2022.current_charset = charset;
|
|
|
3770
|
|
|
3771 /* Verbum caro factum est! */
|
|
|
3772 }
|
|
|
3773
|
|
|
3774
|
|
|
3775 /************************************************************************/
|
|
|
3776 /* UTF-8 methods */
|
|
|
3777 /************************************************************************/
|
|
|
3778
|
|
|
3779 static int
|
|
398
|
3780 detect_coding_utf8 (struct detection_state *st, const unsigned char *src,
|
|
396
|
3781 unsigned int n)
|
|
|
3782 {
|
|
|
3783 while (n--)
|
|
|
3784 {
|
|
|
3785 unsigned char c = *src++;
|
|
|
3786 switch (st->utf8.in_byte)
|
|
|
3787 {
|
|
|
3788 case 0:
|
|
|
3789 if (c == ISO_CODE_ESC || c == ISO_CODE_SI || c == ISO_CODE_SO)
|
|
|
3790 return 0;
|
|
|
3791 else if (c >= 0xfc)
|
|
|
3792 st->utf8.in_byte = 5;
|
|
|
3793 else if (c >= 0xf8)
|
|
|
3794 st->utf8.in_byte = 4;
|
|
|
3795 else if (c >= 0xf0)
|
|
|
3796 st->utf8.in_byte = 3;
|
|
|
3797 else if (c >= 0xe0)
|
|
|
3798 st->utf8.in_byte = 2;
|
|
|
3799 else if (c >= 0xc0)
|
|
|
3800 st->utf8.in_byte = 1;
|
|
|
3801 else if (c >= 0x80)
|
|
|
3802 return 0;
|
|
|
3803 break;
|
|
|
3804 default:
|
|
|
3805 if ((c & 0xc0) != 0x80)
|
|
|
3806 return 0;
|
|
|
3807 else
|
|
|
3808 st->utf8.in_byte--;
|
|
|
3809 }
|
|
|
3810 }
|
|
|
3811 return CODING_CATEGORY_UTF8_MASK;
|
|
|
3812 }
|
|
|
3813
|
|
|
3814 static void
|
|
398
|
3815 decode_coding_utf8 (Lstream *decoding, const unsigned char *src,
|
|
396
|
3816 unsigned_char_dynarr *dst, unsigned int n)
|
|
|
3817 {
|
|
|
3818 struct decoding_stream *str = DECODING_STREAM_DATA (decoding);
|
|
|
3819 unsigned int flags = str->flags;
|
|
|
3820 unsigned int ch = str->ch;
|
|
|
3821 eol_type_t eol_type = str->eol_type;
|
|
398
|
3822 unsigned char counter = str->counter;
|
|
396
|
3823
|
|
|
3824 while (n--)
|
|
|
3825 {
|
|
|
3826 unsigned char c = *src++;
|
|
398
|
3827 switch (counter)
|
|
396
|
3828 {
|
|
|
3829 case 0:
|
|
|
3830 if ( c >= 0xfc )
|
|
|
3831 {
|
|
|
3832 ch = c & 0x01;
|
|
398
|
3833 counter = 5;
|
|
396
|
3834 }
|
|
|
3835 else if ( c >= 0xf8 )
|
|
|
3836 {
|
|
|
3837 ch = c & 0x03;
|
|
398
|
3838 counter = 4;
|
|
396
|
3839 }
|
|
|
3840 else if ( c >= 0xf0 )
|
|
|
3841 {
|
|
|
3842 ch = c & 0x07;
|
|
398
|
3843 counter = 3;
|
|
396
|
3844 }
|
|
|
3845 else if ( c >= 0xe0 )
|
|
|
3846 {
|
|
|
3847 ch = c & 0x0f;
|
|
398
|
3848 counter = 2;
|
|
396
|
3849 }
|
|
|
3850 else if ( c >= 0xc0 )
|
|
|
3851 {
|
|
|
3852 ch = c & 0x1f;
|
|
398
|
3853 counter = 1;
|
|
396
|
3854 }
|
|
|
3855 else
|
|
|
3856 {
|
|
|
3857 DECODE_HANDLE_EOL_TYPE (eol_type, c, flags, dst);
|
|
|
3858 decode_ucs4 (c, dst);
|
|
|
3859 }
|
|
|
3860 break;
|
|
|
3861 case 1:
|
|
|
3862 ch = ( ch << 6 ) | ( c & 0x3f );
|
|
|
3863 decode_ucs4 (ch, dst);
|
|
|
3864 ch = 0;
|
|
398
|
3865 counter = 0;
|
|
396
|
3866 break;
|
|
|
3867 default:
|
|
|
3868 ch = ( ch << 6 ) | ( c & 0x3f );
|
|
398
|
3869 counter--;
|
|
396
|
3870 }
|
|
|
3871 label_continue_loop:;
|
|
|
3872 }
|
|
|
3873
|
|
|
3874 if (flags & CODING_STATE_END)
|
|
|
3875 DECODE_OUTPUT_PARTIAL_CHAR (ch);
|
|
|
3876
|
|
|
3877 str->flags = flags;
|
|
|
3878 str->ch = ch;
|
|
398
|
3879 str->counter = counter;
|
|
396
|
3880 }
|
|
|
3881
|
|
|
3882 static void
|
|
|
3883 encode_utf8 (Lisp_Object charset,
|
|
|
3884 unsigned char h, unsigned char l, unsigned_char_dynarr *dst)
|
|
|
3885 {
|
|
|
3886 unsigned long code = mule_char_to_ucs4 (charset, h, l);
|
|
|
3887 if ( code <= 0x7f )
|
|
|
3888 {
|
|
|
3889 Dynarr_add (dst, code);
|
|
|
3890 }
|
|
|
3891 else if ( code <= 0x7ff )
|
|
|
3892 {
|
|
|
3893 Dynarr_add (dst, (code >> 6) | 0xc0);
|
|
|
3894 Dynarr_add (dst, (code & 0x3f) | 0x80);
|
|
|
3895 }
|
|
|
3896 else if ( code <= 0xffff )
|
|
|
3897 {
|
|
|
3898 Dynarr_add (dst, (code >> 12) | 0xe0);
|
|
|
3899 Dynarr_add (dst, ((code >> 6) & 0x3f) | 0x80);
|
|
|
3900 Dynarr_add (dst, (code & 0x3f) | 0x80);
|
|
|
3901 }
|
|
|
3902 else if ( code <= 0x1fffff )
|
|
|
3903 {
|
|
|
3904 Dynarr_add (dst, (code >> 18) | 0xf0);
|
|
|
3905 Dynarr_add (dst, ((code >> 12) & 0x3f) | 0x80);
|
|
|
3906 Dynarr_add (dst, ((code >> 6) & 0x3f) | 0x80);
|
|
|
3907 Dynarr_add (dst, (code & 0x3f) | 0x80);
|
|
|
3908 }
|
|
|
3909 else if ( code <= 0x3ffffff )
|
|
|
3910 {
|
|
|
3911 Dynarr_add (dst, (code >> 24) | 0xf8);
|
|
|
3912 Dynarr_add (dst, ((code >> 18) & 0x3f) | 0x80);
|
|
|
3913 Dynarr_add (dst, ((code >> 12) & 0x3f) | 0x80);
|
|
|
3914 Dynarr_add (dst, ((code >> 6) & 0x3f) | 0x80);
|
|
|
3915 Dynarr_add (dst, (code & 0x3f) | 0x80);
|
|
|
3916 }
|
|
|
3917 else
|
|
|
3918 {
|
|
|
3919 Dynarr_add (dst, (code >> 30) | 0xfc);
|
|
|
3920 Dynarr_add (dst, ((code >> 24) & 0x3f) | 0x80);
|
|
|
3921 Dynarr_add (dst, ((code >> 18) & 0x3f) | 0x80);
|
|
|
3922 Dynarr_add (dst, ((code >> 12) & 0x3f) | 0x80);
|
|
|
3923 Dynarr_add (dst, ((code >> 6) & 0x3f) | 0x80);
|
|
|
3924 Dynarr_add (dst, (code & 0x3f) | 0x80);
|
|
|
3925 }
|
|
|
3926 }
|
|
|
3927
|
|
|
3928 static void
|
|
398
|
3929 encode_coding_utf8 (Lstream *encoding, const unsigned char *src,
|
|
396
|
3930 unsigned_char_dynarr *dst, unsigned int n)
|
|
|
3931 {
|
|
|
3932 struct encoding_stream *str = ENCODING_STREAM_DATA (encoding);
|
|
|
3933 unsigned int flags = str->flags;
|
|
|
3934 unsigned int ch = str->ch;
|
|
|
3935 eol_type_t eol_type = CODING_SYSTEM_EOL_TYPE (str->codesys);
|
|
|
3936 unsigned char char_boundary = str->iso2022.current_char_boundary;
|
|
|
3937 Lisp_Object charset = str->iso2022.current_charset;
|
|
|
3938
|
|
|
3939 #ifdef ENABLE_COMPOSITE_CHARS
|
|
|
3940 /* flags for handling composite chars. We do a little switcharoo
|
|
|
3941 on the source while we're outputting the composite char. */
|
|
|
3942 unsigned int saved_n = 0;
|
|
398
|
3943 const unsigned char *saved_src = NULL;
|
|
396
|
3944 int in_composite = 0;
|
|
|
3945
|
|
|
3946 back_to_square_n:
|
|
|
3947 #endif /* ENABLE_COMPOSITE_CHARS */
|
|
398
|
3948
|
|
396
|
3949 while (n--)
|
|
|
3950 {
|
|
|
3951 unsigned char c = *src++;
|
|
|
3952
|
|
|
3953 if (BYTE_ASCII_P (c))
|
|
|
3954 { /* Processing ASCII character */
|
|
|
3955 ch = 0;
|
|
|
3956 if (c == '\n')
|
|
|
3957 {
|
|
|
3958 if (eol_type != EOL_LF && eol_type != EOL_AUTODETECT)
|
|
|
3959 Dynarr_add (dst, '\r');
|
|
|
3960 if (eol_type != EOL_CR)
|
|
|
3961 Dynarr_add (dst, c);
|
|
|
3962 }
|
|
|
3963 else
|
|
|
3964 encode_utf8 (Vcharset_ascii, c, 0, dst);
|
|
|
3965 char_boundary = 1;
|
|
|
3966 }
|
|
|
3967 else if (BUFBYTE_LEADING_BYTE_P (c) || BUFBYTE_LEADING_BYTE_P (ch))
|
|
|
3968 { /* Processing Leading Byte */
|
|
|
3969 ch = 0;
|
|
|
3970 charset = CHARSET_BY_LEADING_BYTE (c);
|
|
|
3971 if (LEADING_BYTE_PREFIX_P(c))
|
|
|
3972 ch = c;
|
|
|
3973 char_boundary = 0;
|
|
|
3974 }
|
|
|
3975 else
|
|
|
3976 { /* Processing Non-ASCII character */
|
|
|
3977 char_boundary = 1;
|
|
|
3978 if (EQ (charset, Vcharset_control_1))
|
|
|
3979 {
|
|
|
3980 encode_utf8 (Vcharset_control_1, c, 0, dst);
|
|
|
3981 }
|
|
|
3982 else
|
|
|
3983 {
|
|
|
3984 switch (XCHARSET_REP_BYTES (charset))
|
|
|
3985 {
|
|
|
3986 case 2:
|
|
|
3987 encode_utf8 (charset, c, 0, dst);
|
|
|
3988 break;
|
|
|
3989 case 3:
|
|
|
3990 if (XCHARSET_PRIVATE_P (charset))
|
|
|
3991 {
|
|
|
3992 encode_utf8 (charset, c, 0, dst);
|
|
|
3993 ch = 0;
|
|
|
3994 }
|
|
|
3995 else if (ch)
|
|
|
3996 {
|
|
|
3997 #ifdef ENABLE_COMPOSITE_CHARS
|
|
|
3998 if (EQ (charset, Vcharset_composite))
|
|
|
3999 {
|
|
|
4000 if (in_composite)
|
|
|
4001 {
|
|
|
4002 /* #### Bother! We don't know how to
|
|
|
4003 handle this yet. */
|
|
|
4004 encode_utf8 (Vcharset_ascii, '~', 0, dst);
|
|
|
4005 }
|
|
|
4006 else
|
|
|
4007 {
|
|
|
4008 Emchar emch = MAKE_CHAR (Vcharset_composite,
|
|
|
4009 ch & 0x7F, c & 0x7F);
|
|
|
4010 Lisp_Object lstr = composite_char_string (emch);
|
|
|
4011 saved_n = n;
|
|
|
4012 saved_src = src;
|
|
|
4013 in_composite = 1;
|
|
|
4014 src = XSTRING_DATA (lstr);
|
|
|
4015 n = XSTRING_LENGTH (lstr);
|
|
|
4016 }
|
|
|
4017 }
|
|
|
4018 else
|
|
|
4019 #endif /* ENABLE_COMPOSITE_CHARS */
|
|
|
4020 {
|
|
|
4021 encode_utf8 (charset, ch, c, dst);
|
|
|
4022 }
|
|
|
4023 ch = 0;
|
|
|
4024 }
|
|
|
4025 else
|
|
|
4026 {
|
|
|
4027 ch = c;
|
|
|
4028 char_boundary = 0;
|
|
|
4029 }
|
|
|
4030 break;
|
|
|
4031 case 4:
|
|
|
4032 if (ch)
|
|
|
4033 {
|
|
|
4034 encode_utf8 (charset, ch, c, dst);
|
|
|
4035 ch = 0;
|
|
|
4036 }
|
|
|
4037 else
|
|
|
4038 {
|
|
|
4039 ch = c;
|
|
|
4040 char_boundary = 0;
|
|
|
4041 }
|
|
|
4042 break;
|
|
|
4043 default:
|
|
|
4044 abort ();
|
|
|
4045 }
|
|
|
4046 }
|
|
|
4047 }
|
|
|
4048 }
|
|
|
4049
|
|
|
4050 #ifdef ENABLE_COMPOSITE_CHARS
|
|
|
4051 if (in_composite)
|
|
|
4052 {
|
|
|
4053 n = saved_n;
|
|
|
4054 src = saved_src;
|
|
|
4055 in_composite = 0;
|
|
|
4056 goto back_to_square_n; /* Wheeeeeeeee ..... */
|
|
|
4057 }
|
|
|
4058 #endif
|
|
|
4059
|
|
|
4060 str->flags = flags;
|
|
|
4061 str->ch = ch;
|
|
|
4062 str->iso2022.current_char_boundary = char_boundary;
|
|
|
4063 str->iso2022.current_charset = charset;
|
|
|
4064
|
|
|
4065 /* Verbum caro factum est! */
|
|
|
4066 }
|
|
|
4067
|
|
|
4068
|
|
|
4069 /************************************************************************/
|
|
259
|
4070 /* ISO2022 methods */
|
|
|
4071 /************************************************************************/
|
|
|
4072
|
|
|
4073 /* The following note describes the coding system ISO2022 briefly.
|
|
396
|
4074 Since the intention of this note is to help understand the
|
|
|
4075 functions in this file, some parts are NOT ACCURATE or OVERLY
|
|
259
|
4076 SIMPLIFIED. For thorough understanding, please refer to the
|
|
|
4077 original document of ISO2022.
|
|
|
4078
|
|
|
4079 ISO2022 provides many mechanisms to encode several character sets
|
|
396
|
4080 in 7-bit and 8-bit environments. For 7-bit environments, all text
|
|
|
4081 is encoded using bytes less than 128. This may make the encoded
|
|
|
4082 text a little bit longer, but the text passes more easily through
|
|
|
4083 several gateways, some of which strip off MSB (Most Signigant Bit).
|
|
|
4084
|
|
|
4085 There are two kinds of character sets: control character set and
|
|
259
|
4086 graphic character set. The former contains control characters such
|
|
|
4087 as `newline' and `escape' to provide control functions (control
|
|
396
|
4088 functions are also provided by escape sequences). The latter
|
|
259
|
4089 contains graphic characters such as 'A' and '-'. Emacs recognizes
|
|
|
4090 two control character sets and many graphic character sets.
|
|
|
4091
|
|
396
|
4092 Graphic character sets are classified into one of the following
|
|
|
4093 four classes, according to the number of bytes (DIMENSION) and
|
|
|
4094 number of characters in one dimension (CHARS) of the set:
|
|
|
4095 - DIMENSION1_CHARS94
|
|
|
4096 - DIMENSION1_CHARS96
|
|
|
4097 - DIMENSION2_CHARS94
|
|
|
4098 - DIMENSION2_CHARS96
|
|
|
4099
|
|
|
4100 In addition, each character set is assigned an identification tag,
|
|
|
4101 unique for each set, called "final character" (denoted as <F>
|
|
|
4102 hereafter). The <F> of each character set is decided by ECMA(*)
|
|
|
4103 when it is registered in ISO. The code range of <F> is 0x30..0x7F
|
|
|
4104 (0x30..0x3F are for private use only).
|
|
259
|
4105
|
|
|
4106 Note (*): ECMA = European Computer Manufacturers Association
|
|
|
4107
|
|
|
4108 Here are examples of graphic character set [NAME(<F>)]:
|
|
396
|
4109 o DIMENSION1_CHARS94 -- ASCII('B'), right-half-of-JISX0201('I'), ...
|
|
|
4110 o DIMENSION1_CHARS96 -- right-half-of-ISO8859-1('A'), ...
|
|
|
4111 o DIMENSION2_CHARS94 -- GB2312('A'), JISX0208('B'), ...
|
|
|
4112 o DIMENSION2_CHARS96 -- none for the moment
|
|
|
4113
|
|
|
4114 A code area (1 byte = 8 bits) is divided into 4 areas, C0, GL, C1, and GR.
|
|
259
|
4115 C0 [0x00..0x1F] -- control character plane 0
|
|
|
4116 GL [0x20..0x7F] -- graphic character plane 0
|
|
|
4117 C1 [0x80..0x9F] -- control character plane 1
|
|
|
4118 GR [0xA0..0xFF] -- graphic character plane 1
|
|
|
4119
|
|
|
4120 A control character set is directly designated and invoked to C0 or
|
|
|
4121 C1 by an escape sequence. The most common case is that:
|
|
|
4122 - ISO646's control character set is designated/invoked to C0, and
|
|
|
4123 - ISO6429's control character set is designated/invoked to C1,
|
|
|
4124 and usually these designations/invocations are omitted in encoded
|
|
|
4125 text. In a 7-bit environment, only C0 can be used, and a control
|
|
|
4126 character for C1 is encoded by an appropriate escape sequence to
|
|
|
4127 fit into the environment. All control characters for C1 are
|
|
|
4128 defined to have corresponding escape sequences.
|
|
|
4129
|
|
|
4130 A graphic character set is at first designated to one of four
|
|
|
4131 graphic registers (G0 through G3), then these graphic registers are
|
|
|
4132 invoked to GL or GR. These designations and invocations can be
|
|
|
4133 done independently. The most common case is that G0 is invoked to
|
|
|
4134 GL, G1 is invoked to GR, and ASCII is designated to G0. Usually
|
|
|
4135 these invocations and designations are omitted in encoded text.
|
|
|
4136 In a 7-bit environment, only GL can be used.
|
|
|
4137
|
|
396
|
4138 When a graphic character set of CHARS94 is invoked to GL, codes
|
|
|
4139 0x20 and 0x7F of the GL area work as control characters SPACE and
|
|
|
4140 DEL respectively, and codes 0xA0 and 0xFF of the GR area should not
|
|
|
4141 be used.
|
|
259
|
4142
|
|
|
4143 There are two ways of invocation: locking-shift and single-shift.
|
|
|
4144 With locking-shift, the invocation lasts until the next different
|
|
396
|
4145 invocation, whereas with single-shift, the invocation affects the
|
|
|
4146 following character only and doesn't affect the locking-shift
|
|
|
4147 state. Invocations are done by the following control characters or
|
|
|
4148 escape sequences:
|
|
259
|
4149
|
|
|
4150 ----------------------------------------------------------------------
|
|
|
4151 abbrev function cntrl escape seq description
|
|
|
4152 ----------------------------------------------------------------------
|
|
|
4153 SI/LS0 (shift-in) 0x0F none invoke G0 into GL
|
|
|
4154 SO/LS1 (shift-out) 0x0E none invoke G1 into GL
|
|
|
4155 LS2 (locking-shift-2) none ESC 'n' invoke G2 into GL
|
|
|
4156 LS3 (locking-shift-3) none ESC 'o' invoke G3 into GL
|
|
396
|
4157 LS1R (locking-shift-1 right) none ESC '~' invoke G1 into GR (*)
|
|
|
4158 LS2R (locking-shift-2 right) none ESC '}' invoke G2 into GR (*)
|
|
|
4159 LS3R (locking-shift 3 right) none ESC '|' invoke G3 into GR (*)
|
|
259
|
4160 SS2 (single-shift-2) 0x8E ESC 'N' invoke G2 for one char
|
|
|
4161 SS3 (single-shift-3) 0x8F ESC 'O' invoke G3 for one char
|
|
|
4162 ----------------------------------------------------------------------
|
|
396
|
4163 (*) These are not used by any known coding system.
|
|
|
4164
|
|
|
4165 Control characters for these functions are defined by macros
|
|
|
4166 ISO_CODE_XXX in `coding.h'.
|
|
|
4167
|
|
|
4168 Designations are done by the following escape sequences:
|
|
259
|
4169 ----------------------------------------------------------------------
|
|
|
4170 escape sequence description
|
|
|
4171 ----------------------------------------------------------------------
|
|
396
|
4172 ESC '(' <F> designate DIMENSION1_CHARS94<F> to G0
|
|
|
4173 ESC ')' <F> designate DIMENSION1_CHARS94<F> to G1
|
|
|
4174 ESC '*' <F> designate DIMENSION1_CHARS94<F> to G2
|
|
|
4175 ESC '+' <F> designate DIMENSION1_CHARS94<F> to G3
|
|
|
4176 ESC ',' <F> designate DIMENSION1_CHARS96<F> to G0 (*)
|
|
|
4177 ESC '-' <F> designate DIMENSION1_CHARS96<F> to G1
|
|
|
4178 ESC '.' <F> designate DIMENSION1_CHARS96<F> to G2
|
|
|
4179 ESC '/' <F> designate DIMENSION1_CHARS96<F> to G3
|
|
|
4180 ESC '$' '(' <F> designate DIMENSION2_CHARS94<F> to G0 (**)
|
|
|
4181 ESC '$' ')' <F> designate DIMENSION2_CHARS94<F> to G1
|
|
|
4182 ESC '$' '*' <F> designate DIMENSION2_CHARS94<F> to G2
|
|
|
4183 ESC '$' '+' <F> designate DIMENSION2_CHARS94<F> to G3
|
|
|
4184 ESC '$' ',' <F> designate DIMENSION2_CHARS96<F> to G0 (*)
|
|
|
4185 ESC '$' '-' <F> designate DIMENSION2_CHARS96<F> to G1
|
|
|
4186 ESC '$' '.' <F> designate DIMENSION2_CHARS96<F> to G2
|
|
|
4187 ESC '$' '/' <F> designate DIMENSION2_CHARS96<F> to G3
|
|
259
|
4188 ----------------------------------------------------------------------
|
|
396
|
4189
|
|
|
4190 In this list, "DIMENSION1_CHARS94<F>" means a graphic character set
|
|
|
4191 of dimension 1, chars 94, and final character <F>, etc...
|
|
259
|
4192
|
|
|
4193 Note (*): Although these designations are not allowed in ISO2022,
|
|
|
4194 Emacs accepts them on decoding, and produces them on encoding
|
|
396
|
4195 CHARS96 character sets in a coding system which is characterized as
|
|
|
4196 7-bit environment, non-locking-shift, and non-single-shift.
|
|
259
|
4197
|
|
|
4198 Note (**): If <F> is '@', 'A', or 'B', the intermediate character
|
|
396
|
4199 '(' can be omitted. We refer to this as "short-form" hereafter.
|
|
259
|
4200
|
|
|
4201 Now you may notice that there are a lot of ways for encoding the
|
|
|
4202 same multilingual text in ISO2022. Actually, there exist many
|
|
396
|
4203 coding systems such as Compound Text (used in X11's inter client
|
|
259
|
4204 communication, ISO-2022-JP (used in Japanese internet), ISO-2022-KR
|
|
|
4205 (used in Korean internet), EUC (Extended UNIX Code, used in Asian
|
|
|
4206 localized platforms), and all of these are variants of ISO2022.
|
|
|
4207
|
|
|
4208 In addition to the above, Emacs handles two more kinds of escape
|
|
|
4209 sequences: ISO6429's direction specification and Emacs' private
|
|
|
4210 sequence for specifying character composition.
|
|
|
4211
|
|
396
|
4212 ISO6429's direction specification takes the following form:
|
|
259
|
4213 o CSI ']' -- end of the current direction
|
|
|
4214 o CSI '0' ']' -- end of the current direction
|
|
|
4215 o CSI '1' ']' -- start of left-to-right text
|
|
|
4216 o CSI '2' ']' -- start of right-to-left text
|
|
|
4217 The control character CSI (0x9B: control sequence introducer) is
|
|
396
|
4218 abbreviated to the escape sequence ESC '[' in a 7-bit environment.
|
|
|
4219
|
|
|
4220 Character composition specification takes the following form:
|
|
259
|
4221 o ESC '0' -- start character composition
|
|
|
4222 o ESC '1' -- end character composition
|
|
396
|
4223 Since these are not standard escape sequences of any ISO standard,
|
|
|
4224 their use with these meanings is restricted to Emacs only. */
|
|
259
|
4225
|
|
|
4226 static void
|
|
|
4227 reset_iso2022 (Lisp_Object coding_system, struct iso2022_decoder *iso)
|
|
|
4228 {
|
|
|
4229 int i;
|
|
|
4230
|
|
|
4231 for (i = 0; i < 4; i++)
|
|
|
4232 {
|
|
|
4233 if (!NILP (coding_system))
|
|
|
4234 iso->charset[i] =
|
|
|
4235 XCODING_SYSTEM_ISO2022_INITIAL_CHARSET (coding_system, i);
|
|
|
4236 else
|
|
|
4237 iso->charset[i] = Qt;
|
|
|
4238 iso->invalid_designated[i] = 0;
|
|
|
4239 }
|
|
|
4240 iso->esc = ISO_ESC_NOTHING;
|
|
|
4241 iso->esc_bytes_index = 0;
|
|
|
4242 iso->register_left = 0;
|
|
|
4243 iso->register_right = 1;
|
|
|
4244 iso->switched_dir_and_no_valid_charset_yet = 0;
|
|
|
4245 iso->invalid_switch_dir = 0;
|
|
|
4246 iso->output_direction_sequence = 0;
|
|
|
4247 iso->output_literally = 0;
|
|
396
|
4248 #ifdef ENABLE_COMPOSITE_CHARS
|
|
259
|
4249 if (iso->composite_chars)
|
|
|
4250 Dynarr_reset (iso->composite_chars);
|
|
396
|
4251 #endif
|
|
259
|
4252 }
|
|
|
4253
|
|
|
4254 static int
|
|
|
4255 fit_to_be_escape_quoted (unsigned char c)
|
|
|
4256 {
|
|
|
4257 switch (c)
|
|
|
4258 {
|
|
|
4259 case ISO_CODE_ESC:
|
|
|
4260 case ISO_CODE_CSI:
|
|
|
4261 case ISO_CODE_SS2:
|
|
|
4262 case ISO_CODE_SS3:
|
|
|
4263 case ISO_CODE_SO:
|
|
|
4264 case ISO_CODE_SI:
|
|
|
4265 return 1;
|
|
|
4266
|
|
|
4267 default:
|
|
|
4268 return 0;
|
|
|
4269 }
|
|
|
4270 }
|
|
|
4271
|
|
|
4272 /* Parse one byte of an ISO2022 escape sequence.
|
|
|
4273 If the result is an invalid escape sequence, return 0 and
|
|
|
4274 do not change anything in STR. Otherwise, if the result is
|
|
|
4275 an incomplete escape sequence, update ISO2022.ESC and
|
|
|
4276 ISO2022.ESC_BYTES and return -1. Otherwise, update
|
|
|
4277 all the state variables (but not ISO2022.ESC_BYTES) and
|
|
|
4278 return 1.
|
|
|
4279
|
|
|
4280 If CHECK_INVALID_CHARSETS is non-zero, check for designation
|
|
|
4281 or invocation of an invalid character set and treat that as
|
|
|
4282 an unrecognized escape sequence. */
|
|
|
4283
|
|
|
4284 static int
|
|
|
4285 parse_iso2022_esc (Lisp_Object codesys, struct iso2022_decoder *iso,
|
|
|
4286 unsigned char c, unsigned int *flags,
|
|
|
4287 int check_invalid_charsets)
|
|
|
4288 {
|
|
|
4289 /* (1) If we're at the end of a designation sequence, CS is the
|
|
|
4290 charset being designated and REG is the register to designate
|
|
|
4291 it to.
|
|
|
4292
|
|
|
4293 (2) If we're at the end of a locking-shift sequence, REG is
|
|
|
4294 the register to invoke and HALF (0 == left, 1 == right) is
|
|
|
4295 the half to invoke it into.
|
|
|
4296
|
|
|
4297 (3) If we're at the end of a single-shift sequence, REG is
|
|
|
4298 the register to invoke. */
|
|
|
4299 Lisp_Object cs = Qnil;
|
|
|
4300 int reg, half;
|
|
|
4301
|
|
|
4302 /* NOTE: This code does goto's all over the fucking place.
|
|
|
4303 The reason for this is that we're basically implementing
|
|
|
4304 a state machine here, and hierarchical languages like C
|
|
|
4305 don't really provide a clean way of doing this. */
|
|
|
4306
|
|
|
4307 if (! (*flags & CODING_STATE_ESCAPE))
|
|
|
4308 /* At beginning of escape sequence; we need to reset our
|
|
|
4309 escape-state variables. */
|
|
|
4310 iso->esc = ISO_ESC_NOTHING;
|
|
|
4311
|
|
|
4312 iso->output_literally = 0;
|
|
|
4313 iso->output_direction_sequence = 0;
|
|
|
4314
|
|
|
4315 switch (iso->esc)
|
|
|
4316 {
|
|
|
4317 case ISO_ESC_NOTHING:
|
|
|
4318 iso->esc_bytes_index = 0;
|
|
|
4319 switch (c)
|
|
|
4320 {
|
|
|
4321 case ISO_CODE_ESC: /* Start escape sequence */
|
|
|
4322 *flags |= CODING_STATE_ESCAPE;
|
|
|
4323 iso->esc = ISO_ESC;
|
|
|
4324 goto not_done;
|
|
|
4325
|
|
|
4326 case ISO_CODE_CSI: /* ISO6429 (specifying directionality) */
|
|
|
4327 *flags |= CODING_STATE_ESCAPE;
|
|
|
4328 iso->esc = ISO_ESC_5_11;
|
|
|
4329 goto not_done;
|
|
|
4330
|
|
|
4331 case ISO_CODE_SO: /* locking shift 1 */
|
|
|
4332 reg = 1; half = 0;
|
|
|
4333 goto locking_shift;
|
|
|
4334 case ISO_CODE_SI: /* locking shift 0 */
|
|
|
4335 reg = 0; half = 0;
|
|
|
4336 goto locking_shift;
|
|
|
4337
|
|
|
4338 case ISO_CODE_SS2: /* single shift */
|
|
|
4339 reg = 2;
|
|
|
4340 goto single_shift;
|
|
|
4341 case ISO_CODE_SS3: /* single shift */
|
|
|
4342 reg = 3;
|
|
|
4343 goto single_shift;
|
|
|
4344
|
|
|
4345 default: /* Other control characters */
|
|
|
4346 return 0;
|
|
|
4347 }
|
|
|
4348
|
|
|
4349 case ISO_ESC:
|
|
|
4350 switch (c)
|
|
|
4351 {
|
|
|
4352 /**** single shift ****/
|
|
|
4353
|
|
|
4354 case 'N': /* single shift 2 */
|
|
|
4355 reg = 2;
|
|
|
4356 goto single_shift;
|
|
|
4357 case 'O': /* single shift 3 */
|
|
|
4358 reg = 3;
|
|
|
4359 goto single_shift;
|
|
|
4360
|
|
|
4361 /**** locking shift ****/
|
|
|
4362
|
|
|
4363 case '~': /* locking shift 1 right */
|
|
|
4364 reg = 1; half = 1;
|
|
|
4365 goto locking_shift;
|
|
|
4366 case 'n': /* locking shift 2 */
|
|
|
4367 reg = 2; half = 0;
|
|
|
4368 goto locking_shift;
|
|
|
4369 case '}': /* locking shift 2 right */
|
|
|
4370 reg = 2; half = 1;
|
|
|
4371 goto locking_shift;
|
|
|
4372 case 'o': /* locking shift 3 */
|
|
|
4373 reg = 3; half = 0;
|
|
|
4374 goto locking_shift;
|
|
|
4375 case '|': /* locking shift 3 right */
|
|
|
4376 reg = 3; half = 1;
|
|
|
4377 goto locking_shift;
|
|
|
4378
|
|
396
|
4379 #ifdef ENABLE_COMPOSITE_CHARS
|
|
259
|
4380 /**** composite ****/
|
|
|
4381
|
|
|
4382 case '0':
|
|
|
4383 iso->esc = ISO_ESC_START_COMPOSITE;
|
|
|
4384 *flags = (*flags & CODING_STATE_ISO2022_LOCK) |
|
|
|
4385 CODING_STATE_COMPOSITE;
|
|
|
4386 return 1;
|
|
|
4387
|
|
|
4388 case '1':
|
|
|
4389 iso->esc = ISO_ESC_END_COMPOSITE;
|
|
|
4390 *flags = (*flags & CODING_STATE_ISO2022_LOCK) &
|
|
|
4391 ~CODING_STATE_COMPOSITE;
|
|
|
4392 return 1;
|
|
396
|
4393 #endif /* ENABLE_COMPOSITE_CHARS */
|
|
259
|
4394
|
|
|
4395 /**** directionality ****/
|
|
|
4396
|
|
|
4397 case '[':
|
|
|
4398 iso->esc = ISO_ESC_5_11;
|
|
|
4399 goto not_done;
|
|
|
4400
|
|
|
4401 /**** designation ****/
|
|
|
4402
|
|
|
4403 case '$': /* multibyte charset prefix */
|
|
|
4404 iso->esc = ISO_ESC_2_4;
|
|
|
4405 goto not_done;
|
|
|
4406
|
|
|
4407 default:
|
|
|
4408 if (0x28 <= c && c <= 0x2F)
|
|
|
4409 {
|
|
|
4410 iso->esc = (enum iso_esc_flag) (c - 0x28 + ISO_ESC_2_8);
|
|
|
4411 goto not_done;
|
|
|
4412 }
|
|
|
4413
|
|
|
4414 /* This function is called with CODESYS equal to nil when
|
|
|
4415 doing coding-system detection. */
|
|
|
4416 if (!NILP (codesys)
|
|
|
4417 && XCODING_SYSTEM_ISO2022_ESCAPE_QUOTED (codesys)
|
|
|
4418 && fit_to_be_escape_quoted (c))
|
|
|
4419 {
|
|
|
4420 iso->esc = ISO_ESC_LITERAL;
|
|
|
4421 *flags &= CODING_STATE_ISO2022_LOCK;
|
|
|
4422 return 1;
|
|
|
4423 }
|
|
|
4424
|
|
|
4425 /* bzzzt! */
|
|
|
4426 return 0;
|
|
|
4427 }
|
|
|
4428
|
|
|
4429
|
|
|
4430
|
|
|
4431 /**** directionality ****/
|
|
|
4432
|
|
|
4433 case ISO_ESC_5_11: /* ISO6429 direction control */
|
|
|
4434 if (c == ']')
|
|
|
4435 {
|
|
|
4436 *flags &= (CODING_STATE_ISO2022_LOCK & ~CODING_STATE_R2L);
|
|
|
4437 goto directionality;
|
|
|
4438 }
|
|
|
4439 if (c == '0') iso->esc = ISO_ESC_5_11_0;
|
|
|
4440 else if (c == '1') iso->esc = ISO_ESC_5_11_1;
|
|
|
4441 else if (c == '2') iso->esc = ISO_ESC_5_11_2;
|
|
|
4442 else return 0;
|
|
|
4443 goto not_done;
|
|
|
4444
|
|
|
4445 case ISO_ESC_5_11_0:
|
|
|
4446 if (c == ']')
|
|
|
4447 {
|
|
|
4448 *flags &= (CODING_STATE_ISO2022_LOCK & ~CODING_STATE_R2L);
|
|
|
4449 goto directionality;
|
|
|
4450 }
|
|
|
4451 return 0;
|
|
|
4452
|
|
|
4453 case ISO_ESC_5_11_1:
|
|
|
4454 if (c == ']')
|
|
|
4455 {
|
|
|
4456 *flags = (CODING_STATE_ISO2022_LOCK & ~CODING_STATE_R2L);
|
|
|
4457 goto directionality;
|
|
|
4458 }
|
|
|
4459 return 0;
|
|
|
4460
|
|
|
4461 case ISO_ESC_5_11_2:
|
|
|
4462 if (c == ']')
|
|
|
4463 {
|
|
|
4464 *flags = (*flags & CODING_STATE_ISO2022_LOCK) | CODING_STATE_R2L;
|
|
|
4465 goto directionality;
|
|
|
4466 }
|
|
|
4467 return 0;
|
|
|
4468
|
|
|
4469 directionality:
|
|
|
4470 iso->esc = ISO_ESC_DIRECTIONALITY;
|
|
|
4471 /* Various junk here to attempt to preserve the direction sequences
|
|
|
4472 literally in the text if they would otherwise be swallowed due
|
|
|
4473 to invalid designations that don't show up as actual charset
|
|
|
4474 changes in the text. */
|
|
|
4475 if (iso->invalid_switch_dir)
|
|
|
4476 {
|
|
|
4477 /* We already inserted a direction switch literally into the
|
|
|
4478 text. We assume (#### this may not be right) that the
|
|
|
4479 next direction switch is the one going the other way,
|
|
|
4480 and we need to output that literally as well. */
|
|
|
4481 iso->output_literally = 1;
|
|
|
4482 iso->invalid_switch_dir = 0;
|
|
|
4483 }
|
|
|
4484 else
|
|
|
4485 {
|
|
|
4486 int jj;
|
|
|
4487
|
|
|
4488 /* If we are in the thrall of an invalid designation,
|
|
|
4489 then stick the directionality sequence literally into the
|
|
|
4490 output stream so it ends up in the original text again. */
|
|
|
4491 for (jj = 0; jj < 4; jj++)
|
|
|
4492 if (iso->invalid_designated[jj])
|
|
|
4493 break;
|
|
|
4494 if (jj < 4)
|
|
|
4495 {
|
|
|
4496 iso->output_literally = 1;
|
|
|
4497 iso->invalid_switch_dir = 1;
|
|
|
4498 }
|
|
|
4499 else
|
|
|
4500 /* Indicate that we haven't yet seen a valid designation,
|
|
|
4501 so that if a switch-dir is directly followed by an
|
|
|
4502 invalid designation, both get inserted literally. */
|
|
|
4503 iso->switched_dir_and_no_valid_charset_yet = 1;
|
|
|
4504 }
|
|
|
4505 return 1;
|
|
|
4506
|
|
|
4507
|
|
|
4508 /**** designation ****/
|
|
|
4509
|
|
|
4510 case ISO_ESC_2_4:
|
|
|
4511 if (0x28 <= c && c <= 0x2F)
|
|
|
4512 {
|
|
|
4513 iso->esc = (enum iso_esc_flag) (c - 0x28 + ISO_ESC_2_4_8);
|
|
|
4514 goto not_done;
|
|
|
4515 }
|
|
|
4516 if (0x40 <= c && c <= 0x42)
|
|
|
4517 {
|
|
|
4518 cs = CHARSET_BY_ATTRIBUTES (CHARSET_TYPE_94X94, c,
|
|
|
4519 *flags & CODING_STATE_R2L ?
|
|
|
4520 CHARSET_RIGHT_TO_LEFT :
|
|
|
4521 CHARSET_LEFT_TO_RIGHT);
|
|
|
4522 reg = 0;
|
|
|
4523 goto designated;
|
|
|
4524 }
|
|
|
4525 return 0;
|
|
|
4526
|
|
|
4527 default:
|
|
|
4528 {
|
|
|
4529 int type =-1;
|
|
|
4530
|
|
|
4531 if (c < '0' || c > '~')
|
|
|
4532 return 0; /* bad final byte */
|
|
|
4533
|
|
|
4534 if (iso->esc >= ISO_ESC_2_8 &&
|
|
|
4535 iso->esc <= ISO_ESC_2_15)
|
|
|
4536 {
|
|
|
4537 type = ((iso->esc >= ISO_ESC_2_12) ?
|
|
|
4538 CHARSET_TYPE_96 : CHARSET_TYPE_94);
|
|
|
4539 reg = (iso->esc - ISO_ESC_2_8) & 3;
|
|
|
4540 }
|
|
|
4541 else if (iso->esc >= ISO_ESC_2_4_8 &&
|
|
|
4542 iso->esc <= ISO_ESC_2_4_15)
|
|
|
4543 {
|
|
|
4544 type = ((iso->esc >= ISO_ESC_2_4_12) ?
|
|
|
4545 CHARSET_TYPE_96X96 : CHARSET_TYPE_94X94);
|
|
|
4546 reg = (iso->esc - ISO_ESC_2_4_8) & 3;
|
|
|
4547 }
|
|
|
4548 else
|
|
|
4549 {
|
|
|
4550 /* Can this ever be reached? -slb */
|
|
|
4551 abort();
|
|
|
4552 }
|
|
|
4553
|
|
|
4554 cs = CHARSET_BY_ATTRIBUTES (type, c,
|
|
|
4555 *flags & CODING_STATE_R2L ?
|
|
|
4556 CHARSET_RIGHT_TO_LEFT :
|
|
|
4557 CHARSET_LEFT_TO_RIGHT);
|
|
|
4558 goto designated;
|
|
|
4559 }
|
|
|
4560 }
|
|
|
4561
|
|
|
4562 not_done:
|
|
|
4563 iso->esc_bytes[iso->esc_bytes_index++] = (unsigned char) c;
|
|
|
4564 return -1;
|
|
|
4565
|
|
|
4566 single_shift:
|
|
|
4567 if (check_invalid_charsets && !CHARSETP (iso->charset[reg]))
|
|
|
4568 /* can't invoke something that ain't there. */
|
|
|
4569 return 0;
|
|
|
4570 iso->esc = ISO_ESC_SINGLE_SHIFT;
|
|
|
4571 *flags &= CODING_STATE_ISO2022_LOCK;
|
|
|
4572 if (reg == 2)
|
|
|
4573 *flags |= CODING_STATE_SS2;
|
|
|
4574 else
|
|
|
4575 *flags |= CODING_STATE_SS3;
|
|
|
4576 return 1;
|
|
|
4577
|
|
|
4578 locking_shift:
|
|
|
4579 if (check_invalid_charsets &&
|
|
|
4580 !CHARSETP (iso->charset[reg]))
|
|
|
4581 /* can't invoke something that ain't there. */
|
|
|
4582 return 0;
|
|
|
4583 if (half)
|
|
|
4584 iso->register_right = reg;
|
|
|
4585 else
|
|
|
4586 iso->register_left = reg;
|
|
|
4587 *flags &= CODING_STATE_ISO2022_LOCK;
|
|
|
4588 iso->esc = ISO_ESC_LOCKING_SHIFT;
|
|
|
4589 return 1;
|
|
|
4590
|
|
|
4591 designated:
|
|
|
4592 if (NILP (cs) && check_invalid_charsets)
|
|
|
4593 {
|
|
|
4594 iso->invalid_designated[reg] = 1;
|
|
|
4595 iso->charset[reg] = Vcharset_ascii;
|
|
|
4596 iso->esc = ISO_ESC_DESIGNATE;
|
|
|
4597 *flags &= CODING_STATE_ISO2022_LOCK;
|
|
|
4598 iso->output_literally = 1;
|
|
|
4599 if (iso->switched_dir_and_no_valid_charset_yet)
|
|
|
4600 {
|
|
|
4601 /* We encountered a switch-direction followed by an
|
|
|
4602 invalid designation. Ensure that the switch-direction
|
|
|
4603 gets outputted; otherwise it will probably get eaten
|
|
|
4604 when the text is written out again. */
|
|
|
4605 iso->switched_dir_and_no_valid_charset_yet = 0;
|
|
|
4606 iso->output_direction_sequence = 1;
|
|
|
4607 /* And make sure that the switch-dir going the other
|
|
|
4608 way gets outputted, as well. */
|
|
|
4609 iso->invalid_switch_dir = 1;
|
|
|
4610 }
|
|
|
4611 return 1;
|
|
|
4612 }
|
|
|
4613 /* This function is called with CODESYS equal to nil when
|
|
|
4614 doing coding-system detection. */
|
|
|
4615 if (!NILP (codesys))
|
|
|
4616 {
|
|
|
4617 charset_conversion_spec_dynarr *dyn =
|
|
|
4618 XCODING_SYSTEM (codesys)->iso2022.input_conv;
|
|
|
4619
|
|
|
4620 if (dyn)
|
|
|
4621 {
|
|
|
4622 int i;
|
|
|
4623
|
|
|
4624 for (i = 0; i < Dynarr_length (dyn); i++)
|
|
|
4625 {
|
|
|
4626 struct charset_conversion_spec *spec = Dynarr_atp (dyn, i);
|
|
|
4627 if (EQ (cs, spec->from_charset))
|
|
|
4628 cs = spec->to_charset;
|
|
|
4629 }
|
|
|
4630 }
|
|
|
4631 }
|
|
|
4632
|
|
|
4633 iso->charset[reg] = cs;
|
|
|
4634 iso->esc = ISO_ESC_DESIGNATE;
|
|
|
4635 *flags &= CODING_STATE_ISO2022_LOCK;
|
|
|
4636 if (iso->invalid_designated[reg])
|
|
|
4637 {
|
|
|
4638 iso->invalid_designated[reg] = 0;
|
|
|
4639 iso->output_literally = 1;
|
|
|
4640 }
|
|
|
4641 if (iso->switched_dir_and_no_valid_charset_yet)
|
|
|
4642 iso->switched_dir_and_no_valid_charset_yet = 0;
|
|
|
4643 return 1;
|
|
|
4644 }
|
|
|
4645
|
|
|
4646 static int
|
|
398
|
4647 detect_coding_iso2022 (struct detection_state *st, const unsigned char *src,
|
|
259
|
4648 unsigned int n)
|
|
|
4649 {
|
|
|
4650 int mask;
|
|
|
4651
|
|
|
4652 /* #### There are serious deficiencies in the recognition mechanism
|
|
388
|
4653 here. This needs to be much smarter if it's going to cut it.
|
|
|
4654 The sequence "\xff\x0f" is currently detected as LOCK_SHIFT while
|
|
|
4655 it should be detected as Latin-1.
|
|
|
4656 All the ISO2022 stuff in this file should be synced up with the
|
|
|
4657 code from FSF Emacs-20.4, in which Mule should be more or less stable.
|
|
|
4658 Perhaps we should wait till R2L works in FSF Emacs? */
|
|
259
|
4659
|
|
|
4660 if (!st->iso2022.initted)
|
|
|
4661 {
|
|
|
4662 reset_iso2022 (Qnil, &st->iso2022.iso);
|
|
|
4663 st->iso2022.mask = (CODING_CATEGORY_ISO_7_MASK |
|
|
|
4664 CODING_CATEGORY_ISO_8_DESIGNATE_MASK |
|
|
|
4665 CODING_CATEGORY_ISO_8_1_MASK |
|
|
|
4666 CODING_CATEGORY_ISO_8_2_MASK |
|
|
|
4667 CODING_CATEGORY_ISO_LOCK_SHIFT_MASK);
|
|
|
4668 st->iso2022.flags = 0;
|
|
|
4669 st->iso2022.high_byte_count = 0;
|
|
|
4670 st->iso2022.saw_single_shift = 0;
|
|
|
4671 st->iso2022.initted = 1;
|
|
|
4672 }
|
|
|
4673
|
|
|
4674 mask = st->iso2022.mask;
|
|
|
4675
|
|
|
4676 while (n--)
|
|
|
4677 {
|
|
388
|
4678 int c = *src++;
|
|
259
|
4679 if (c >= 0xA0)
|
|
|
4680 {
|
|
|
4681 mask &= ~CODING_CATEGORY_ISO_7_MASK;
|
|
|
4682 st->iso2022.high_byte_count++;
|
|
|
4683 }
|
|
|
4684 else
|
|
|
4685 {
|
|
|
4686 if (st->iso2022.high_byte_count && !st->iso2022.saw_single_shift)
|
|
|
4687 {
|
|
|
4688 if (st->iso2022.high_byte_count & 1)
|
|
|
4689 /* odd number of high bytes; assume not iso-8-2 */
|
|
|
4690 mask &= ~CODING_CATEGORY_ISO_8_2_MASK;
|
|
|
4691 }
|
|
|
4692 st->iso2022.high_byte_count = 0;
|
|
|
4693 st->iso2022.saw_single_shift = 0;
|
|
|
4694 if (c > 0x80)
|
|
|
4695 mask &= ~CODING_CATEGORY_ISO_7_MASK;
|
|
|
4696 }
|
|
|
4697 if (!(st->iso2022.flags & CODING_STATE_ESCAPE)
|
|
|
4698 && (BYTE_C0_P (c) || BYTE_C1_P (c)))
|
|
|
4699 { /* control chars */
|
|
|
4700 switch (c)
|
|
|
4701 {
|
|
|
4702 /* Allow and ignore control characters that you might
|
|
|
4703 reasonably see in a text file */
|
|
|
4704 case '\r':
|
|
|
4705 case '\n':
|
|
|
4706 case '\t':
|
|
|
4707 case 7: /* bell */
|
|
|
4708 case 8: /* backspace */
|
|
|
4709 case 11: /* vertical tab */
|
|
|
4710 case 12: /* form feed */
|
|
|
4711 case 26: /* MS-DOS C-z junk */
|
|
|
4712 case 31: /* '^_' -- for info */
|
|
|
4713 goto label_continue_loop;
|
|
|
4714
|
|
|
4715 default:
|
|
|
4716 break;
|
|
|
4717 }
|
|
|
4718 }
|
|
|
4719
|
|
|
4720 if ((st->iso2022.flags & CODING_STATE_ESCAPE) || BYTE_C0_P (c)
|
|
|
4721 || BYTE_C1_P (c))
|
|
|
4722 {
|
|
|
4723 if (parse_iso2022_esc (Qnil, &st->iso2022.iso, c,
|
|
|
4724 &st->iso2022.flags, 0))
|
|
|
4725 {
|
|
|
4726 switch (st->iso2022.iso.esc)
|
|
|
4727 {
|
|
|
4728 case ISO_ESC_DESIGNATE:
|
|
|
4729 mask &= ~CODING_CATEGORY_ISO_8_1_MASK;
|
|
|
4730 mask &= ~CODING_CATEGORY_ISO_8_2_MASK;
|
|
|
4731 break;
|
|
|
4732 case ISO_ESC_LOCKING_SHIFT:
|
|
|
4733 mask = CODING_CATEGORY_ISO_LOCK_SHIFT_MASK;
|
|
|
4734 goto ran_out_of_chars;
|
|
|
4735 case ISO_ESC_SINGLE_SHIFT:
|
|
|
4736 mask &= ~CODING_CATEGORY_ISO_8_DESIGNATE_MASK;
|
|
|
4737 st->iso2022.saw_single_shift = 1;
|
|
|
4738 break;
|
|
|
4739 default:
|
|
|
4740 break;
|
|
|
4741 }
|
|
|
4742 }
|
|
|
4743 else
|
|
|
4744 {
|
|
|
4745 mask = 0;
|
|
|
4746 goto ran_out_of_chars;
|
|
|
4747 }
|
|
|
4748 }
|
|
|
4749 label_continue_loop:;
|
|
|
4750 }
|
|
|
4751
|
|
|
4752 ran_out_of_chars:
|
|
|
4753
|
|
|
4754 return mask;
|
|
|
4755 }
|
|
|
4756
|
|
|
4757 static int
|
|
|
4758 postprocess_iso2022_mask (int mask)
|
|
|
4759 {
|
|
|
4760 /* #### kind of cheesy */
|
|
|
4761 /* If seven-bit ISO is allowed, then assume that the encoding is
|
|
|
4762 entirely seven-bit and turn off the eight-bit ones. */
|
|
|
4763 if (mask & CODING_CATEGORY_ISO_7_MASK)
|
|
|
4764 mask &= ~ (CODING_CATEGORY_ISO_8_DESIGNATE_MASK |
|
|
|
4765 CODING_CATEGORY_ISO_8_1_MASK |
|
|
|
4766 CODING_CATEGORY_ISO_8_2_MASK);
|
|
|
4767 return mask;
|
|
|
4768 }
|
|
|
4769
|
|
|
4770 /* If FLAGS is a null pointer or specifies right-to-left motion,
|
|
|
4771 output a switch-dir-to-left-to-right sequence to DST.
|
|
|
4772 Also update FLAGS if it is not a null pointer.
|
|
|
4773 If INTERNAL_P is set, we are outputting in internal format and
|
|
|
4774 need to handle the CSI differently. */
|
|
|
4775
|
|
|
4776 static void
|
|
396
|
4777 restore_left_to_right_direction (Lisp_Coding_System *codesys,
|
|
259
|
4778 unsigned_char_dynarr *dst,
|
|
|
4779 unsigned int *flags,
|
|
|
4780 int internal_p)
|
|
|
4781 {
|
|
|
4782 if (!flags || (*flags & CODING_STATE_R2L))
|
|
|
4783 {
|
|
|
4784 if (CODING_SYSTEM_ISO2022_SEVEN (codesys))
|
|
|
4785 {
|
|
|
4786 Dynarr_add (dst, ISO_CODE_ESC);
|
|
|
4787 Dynarr_add (dst, '[');
|
|
|
4788 }
|
|
|
4789 else if (internal_p)
|
|
|
4790 DECODE_ADD_BINARY_CHAR (ISO_CODE_CSI, dst);
|
|
|
4791 else
|
|
|
4792 Dynarr_add (dst, ISO_CODE_CSI);
|
|
|
4793 Dynarr_add (dst, '0');
|
|
|
4794 Dynarr_add (dst, ']');
|
|
|
4795 if (flags)
|
|
|
4796 *flags &= ~CODING_STATE_R2L;
|
|
|
4797 }
|
|
|
4798 }
|
|
|
4799
|
|
|
4800 /* If FLAGS is a null pointer or specifies a direction different from
|
|
|
4801 DIRECTION (which should be either CHARSET_RIGHT_TO_LEFT or
|
|
|
4802 CHARSET_LEFT_TO_RIGHT), output the appropriate switch-dir escape
|
|
|
4803 sequence to DST. Also update FLAGS if it is not a null pointer.
|
|
|
4804 If INTERNAL_P is set, we are outputting in internal format and
|
|
|
4805 need to handle the CSI differently. */
|
|
|
4806
|
|
|
4807 static void
|
|
396
|
4808 ensure_correct_direction (int direction, Lisp_Coding_System *codesys,
|
|
259
|
4809 unsigned_char_dynarr *dst, unsigned int *flags,
|
|
|
4810 int internal_p)
|
|
|
4811 {
|
|
|
4812 if ((!flags || (*flags & CODING_STATE_R2L)) &&
|
|
|
4813 direction == CHARSET_LEFT_TO_RIGHT)
|
|
|
4814 restore_left_to_right_direction (codesys, dst, flags, internal_p);
|
|
|
4815 else if (!CODING_SYSTEM_ISO2022_NO_ISO6429 (codesys)
|
|
|
4816 && (!flags || !(*flags & CODING_STATE_R2L)) &&
|
|
|
4817 direction == CHARSET_RIGHT_TO_LEFT)
|
|
|
4818 {
|
|
|
4819 if (CODING_SYSTEM_ISO2022_SEVEN (codesys))
|
|
|
4820 {
|
|
|
4821 Dynarr_add (dst, ISO_CODE_ESC);
|
|
|
4822 Dynarr_add (dst, '[');
|
|
|
4823 }
|
|
|
4824 else if (internal_p)
|
|
|
4825 DECODE_ADD_BINARY_CHAR (ISO_CODE_CSI, dst);
|
|
|
4826 else
|
|
|
4827 Dynarr_add (dst, ISO_CODE_CSI);
|
|
|
4828 Dynarr_add (dst, '2');
|
|
|
4829 Dynarr_add (dst, ']');
|
|
|
4830 if (flags)
|
|
|
4831 *flags |= CODING_STATE_R2L;
|
|
|
4832 }
|
|
|
4833 }
|
|
|
4834
|
|
|
4835 /* Convert ISO2022-format data to internal format. */
|
|
|
4836
|
|
|
4837 static void
|
|
398
|
4838 decode_coding_iso2022 (Lstream *decoding, const unsigned char *src,
|
|
259
|
4839 unsigned_char_dynarr *dst, unsigned int n)
|
|
|
4840 {
|
|
|
4841 struct decoding_stream *str = DECODING_STREAM_DATA (decoding);
|
|
396
|
4842 unsigned int flags = str->flags;
|
|
|
4843 unsigned int ch = str->ch;
|
|
|
4844 eol_type_t eol_type = str->eol_type;
|
|
|
4845 #ifdef ENABLE_COMPOSITE_CHARS
|
|
259
|
4846 unsigned_char_dynarr *real_dst = dst;
|
|
396
|
4847 #endif
|
|
|
4848 Lisp_Object coding_system;
|
|
|
4849
|
|
259
|
4850 XSETCODING_SYSTEM (coding_system, str->codesys);
|
|
|
4851
|
|
396
|
4852 #ifdef ENABLE_COMPOSITE_CHARS
|
|
259
|
4853 if (flags & CODING_STATE_COMPOSITE)
|
|
|
4854 dst = str->iso2022.composite_chars;
|
|
396
|
4855 #endif /* ENABLE_COMPOSITE_CHARS */
|
|
259
|
4856
|
|
|
4857 while (n--)
|
|
|
4858 {
|
|
388
|
4859 unsigned char c = *src++;
|
|
259
|
4860 if (flags & CODING_STATE_ESCAPE)
|
|
|
4861 { /* Within ESC sequence */
|
|
|
4862 int retval = parse_iso2022_esc (coding_system, &str->iso2022,
|
|
|
4863 c, &flags, 1);
|
|
|
4864
|
|
|
4865 if (retval)
|
|
|
4866 {
|
|
|
4867 switch (str->iso2022.esc)
|
|
|
4868 {
|
|
396
|
4869 #ifdef ENABLE_COMPOSITE_CHARS
|
|
259
|
4870 case ISO_ESC_START_COMPOSITE:
|
|
|
4871 if (str->iso2022.composite_chars)
|
|
|
4872 Dynarr_reset (str->iso2022.composite_chars);
|
|
|
4873 else
|
|
|
4874 str->iso2022.composite_chars = Dynarr_new (unsigned_char);
|
|
|
4875 dst = str->iso2022.composite_chars;
|
|
|
4876 break;
|
|
|
4877 case ISO_ESC_END_COMPOSITE:
|
|
|
4878 {
|
|
|
4879 Bufbyte comstr[MAX_EMCHAR_LEN];
|
|
|
4880 Bytecount len;
|
|
|
4881 Emchar emch = lookup_composite_char (Dynarr_atp (dst, 0),
|
|
|
4882 Dynarr_length (dst));
|
|
|
4883 dst = real_dst;
|
|
|
4884 len = set_charptr_emchar (comstr, emch);
|
|
|
4885 Dynarr_add_many (dst, comstr, len);
|
|
|
4886 break;
|
|
|
4887 }
|
|
396
|
4888 #endif /* ENABLE_COMPOSITE_CHARS */
|
|
259
|
4889
|
|
|
4890 case ISO_ESC_LITERAL:
|
|
|
4891 DECODE_ADD_BINARY_CHAR (c, dst);
|
|
|
4892 break;
|
|
|
4893
|
|
|
4894 default:
|
|
|
4895 /* Everything else handled already */
|
|
|
4896 break;
|
|
|
4897 }
|
|
|
4898 }
|
|
|
4899
|
|
|
4900 /* Attempted error recovery. */
|
|
|
4901 if (str->iso2022.output_direction_sequence)
|
|
|
4902 ensure_correct_direction (flags & CODING_STATE_R2L ?
|
|
|
4903 CHARSET_RIGHT_TO_LEFT :
|
|
|
4904 CHARSET_LEFT_TO_RIGHT,
|
|
|
4905 str->codesys, dst, 0, 1);
|
|
|
4906 /* More error recovery. */
|
|
|
4907 if (!retval || str->iso2022.output_literally)
|
|
|
4908 {
|
|
|
4909 /* Output the (possibly invalid) sequence */
|
|
|
4910 int i;
|
|
|
4911 for (i = 0; i < str->iso2022.esc_bytes_index; i++)
|
|
|
4912 DECODE_ADD_BINARY_CHAR (str->iso2022.esc_bytes[i], dst);
|
|
|
4913 flags &= CODING_STATE_ISO2022_LOCK;
|
|
|
4914 if (!retval)
|
|
|
4915 n++, src--;/* Repeat the loop with the same character. */
|
|
|
4916 else
|
|
|
4917 {
|
|
|
4918 /* No sense in reprocessing the final byte of the
|
|
|
4919 escape sequence; it could mess things up anyway.
|
|
|
4920 Just add it now. */
|
|
|
4921 DECODE_ADD_BINARY_CHAR (c, dst);
|
|
|
4922 }
|
|
|
4923 }
|
|
|
4924 ch = 0;
|
|
|
4925 }
|
|
|
4926 else if (BYTE_C0_P (c) || BYTE_C1_P (c))
|
|
|
4927 { /* Control characters */
|
|
|
4928
|
|
|
4929 /***** Error-handling *****/
|
|
|
4930
|
|
|
4931 /* If we were in the middle of a character, dump out the
|
|
|
4932 partial character. */
|
|
|
4933 DECODE_OUTPUT_PARTIAL_CHAR (ch);
|
|
|
4934
|
|
|
4935 /* If we just saw a single-shift character, dump it out.
|
|
|
4936 This may dump out the wrong sort of single-shift character,
|
|
|
4937 but least it will give an indication that something went
|
|
|
4938 wrong. */
|
|
|
4939 if (flags & CODING_STATE_SS2)
|
|
|
4940 {
|
|
|
4941 DECODE_ADD_BINARY_CHAR (ISO_CODE_SS2, dst);
|
|
|
4942 flags &= ~CODING_STATE_SS2;
|
|
|
4943 }
|
|
|
4944 if (flags & CODING_STATE_SS3)
|
|
|
4945 {
|
|
|
4946 DECODE_ADD_BINARY_CHAR (ISO_CODE_SS3, dst);
|
|
|
4947 flags &= ~CODING_STATE_SS3;
|
|
|
4948 }
|
|
|
4949
|
|
|
4950 /***** Now handle the control characters. *****/
|
|
|
4951
|
|
|
4952 /* Handle CR/LF */
|
|
|
4953 DECODE_HANDLE_EOL_TYPE (eol_type, c, flags, dst);
|
|
|
4954
|
|
|
4955 flags &= CODING_STATE_ISO2022_LOCK;
|
|
|
4956
|
|
|
4957 if (!parse_iso2022_esc (coding_system, &str->iso2022, c, &flags, 1))
|
|
|
4958 DECODE_ADD_BINARY_CHAR (c, dst);
|
|
|
4959 }
|
|
|
4960 else
|
|
|
4961 { /* Graphic characters */
|
|
|
4962 Lisp_Object charset;
|
|
|
4963 int lb;
|
|
|
4964 int reg;
|
|
|
4965
|
|
|
4966 DECODE_HANDLE_EOL_TYPE (eol_type, c, flags, dst);
|
|
|
4967
|
|
|
4968 /* Now determine the charset. */
|
|
|
4969 reg = ((flags & CODING_STATE_SS2) ? 2
|
|
|
4970 : (flags & CODING_STATE_SS3) ? 3
|
|
|
4971 : !BYTE_ASCII_P (c) ? str->iso2022.register_right
|
|
|
4972 : str->iso2022.register_left);
|
|
|
4973 charset = str->iso2022.charset[reg];
|
|
|
4974
|
|
|
4975 /* Error checking: */
|
|
388
|
4976 if (! CHARSETP (charset)
|
|
|
4977 || str->iso2022.invalid_designated[reg]
|
|
259
|
4978 || (((c & 0x7F) == ' ' || (c & 0x7F) == ISO_CODE_DEL)
|
|
|
4979 && XCHARSET_CHARS (charset) == 94))
|
|
|
4980 /* Mrmph. We are trying to invoke a register that has no
|
|
|
4981 or an invalid charset in it, or trying to add a character
|
|
|
4982 outside the range of the charset. Insert that char literally
|
|
|
4983 to preserve it for the output. */
|
|
|
4984 {
|
|
|
4985 DECODE_OUTPUT_PARTIAL_CHAR (ch);
|
|
|
4986 DECODE_ADD_BINARY_CHAR (c, dst);
|
|
|
4987 }
|
|
|
4988
|
|
|
4989 else
|
|
|
4990 {
|
|
|
4991 /* Things are probably hunky-dorey. */
|
|
|
4992
|
|
|
4993 /* Fetch reverse charset, maybe. */
|
|
|
4994 if (((flags & CODING_STATE_R2L) &&
|
|
|
4995 XCHARSET_DIRECTION (charset) == CHARSET_LEFT_TO_RIGHT)
|
|
|
4996 ||
|
|
|
4997 (!(flags & CODING_STATE_R2L) &&
|
|
|
4998 XCHARSET_DIRECTION (charset) == CHARSET_RIGHT_TO_LEFT))
|
|
|
4999 {
|
|
|
5000 Lisp_Object new_charset =
|
|
|
5001 XCHARSET_REVERSE_DIRECTION_CHARSET (charset);
|
|
|
5002 if (!NILP (new_charset))
|
|
|
5003 charset = new_charset;
|
|
|
5004 }
|
|
|
5005
|
|
|
5006 lb = XCHARSET_LEADING_BYTE (charset);
|
|
|
5007 switch (XCHARSET_REP_BYTES (charset))
|
|
|
5008 {
|
|
|
5009 case 1: /* ASCII */
|
|
|
5010 DECODE_OUTPUT_PARTIAL_CHAR (ch);
|
|
|
5011 Dynarr_add (dst, c & 0x7F);
|
|
|
5012 break;
|
|
|
5013
|
|
|
5014 case 2: /* one-byte official */
|
|
|
5015 DECODE_OUTPUT_PARTIAL_CHAR (ch);
|
|
|
5016 Dynarr_add (dst, lb);
|
|
|
5017 Dynarr_add (dst, c | 0x80);
|
|
|
5018 break;
|
|
|
5019
|
|
|
5020 case 3: /* one-byte private or two-byte official */
|
|
|
5021 if (XCHARSET_PRIVATE_P (charset))
|
|
|
5022 {
|
|
|
5023 DECODE_OUTPUT_PARTIAL_CHAR (ch);
|
|
|
5024 Dynarr_add (dst, PRE_LEADING_BYTE_PRIVATE_1);
|
|
|
5025 Dynarr_add (dst, lb);
|
|
|
5026 Dynarr_add (dst, c | 0x80);
|
|
|
5027 }
|
|
|
5028 else
|
|
|
5029 {
|
|
|
5030 if (ch)
|
|
|
5031 {
|
|
|
5032 Dynarr_add (dst, lb);
|
|
|
5033 Dynarr_add (dst, ch | 0x80);
|
|
|
5034 Dynarr_add (dst, c | 0x80);
|
|
|
5035 ch = 0;
|
|
|
5036 }
|
|
|
5037 else
|
|
|
5038 ch = c;
|
|
|
5039 }
|
|
|
5040 break;
|
|
|
5041
|
|
|
5042 default: /* two-byte private */
|
|
|
5043 if (ch)
|
|
|
5044 {
|
|
|
5045 Dynarr_add (dst, PRE_LEADING_BYTE_PRIVATE_2);
|
|
|
5046 Dynarr_add (dst, lb);
|
|
|
5047 Dynarr_add (dst, ch | 0x80);
|
|
|
5048 Dynarr_add (dst, c | 0x80);
|
|
|
5049 ch = 0;
|
|
|
5050 }
|
|
|
5051 else
|
|
|
5052 ch = c;
|
|
|
5053 }
|
|
|
5054 }
|
|
|
5055
|
|
|
5056 if (!ch)
|
|
|
5057 flags &= CODING_STATE_ISO2022_LOCK;
|
|
|
5058 }
|
|
|
5059
|
|
|
5060 label_continue_loop:;
|
|
|
5061 }
|
|
|
5062
|
|
|
5063 if (flags & CODING_STATE_END)
|
|
|
5064 DECODE_OUTPUT_PARTIAL_CHAR (ch);
|
|
|
5065
|
|
396
|
5066 str->flags = flags;
|
|
|
5067 str->ch = ch;
|
|
259
|
5068 }
|
|
|
5069
|
|
|
5070
|
|
|
5071 /***** ISO2022 encoder *****/
|
|
|
5072
|
|
|
5073 /* Designate CHARSET into register REG. */
|
|
|
5074
|
|
|
5075 static void
|
|
|
5076 iso2022_designate (Lisp_Object charset, unsigned char reg,
|
|
|
5077 struct encoding_stream *str, unsigned_char_dynarr *dst)
|
|
|
5078 {
|
|
398
|
5079 static const char inter94[] = "()*+";
|
|
|
5080 static const char inter96[] = ",-./";
|
|
272
|
5081 unsigned int type;
|
|
259
|
5082 unsigned char final;
|
|
|
5083 Lisp_Object old_charset = str->iso2022.charset[reg];
|
|
|
5084
|
|
|
5085 str->iso2022.charset[reg] = charset;
|
|
|
5086 if (!CHARSETP (charset))
|
|
|
5087 /* charset might be an initial nil or t. */
|
|
|
5088 return;
|
|
|
5089 type = XCHARSET_TYPE (charset);
|
|
|
5090 final = XCHARSET_FINAL (charset);
|
|
|
5091 if (!str->iso2022.force_charset_on_output[reg] &&
|
|
|
5092 CHARSETP (old_charset) &&
|
|
|
5093 XCHARSET_TYPE (old_charset) == type &&
|
|
|
5094 XCHARSET_FINAL (old_charset) == final)
|
|
|
5095 return;
|
|
|
5096
|
|
|
5097 str->iso2022.force_charset_on_output[reg] = 0;
|
|
|
5098
|
|
|
5099 {
|
|
|
5100 charset_conversion_spec_dynarr *dyn =
|
|
|
5101 str->codesys->iso2022.output_conv;
|
|
|
5102
|
|
|
5103 if (dyn)
|
|
|
5104 {
|
|
|
5105 int i;
|
|
|
5106
|
|
|
5107 for (i = 0; i < Dynarr_length (dyn); i++)
|
|
|
5108 {
|
|
|
5109 struct charset_conversion_spec *spec = Dynarr_atp (dyn, i);
|
|
|
5110 if (EQ (charset, spec->from_charset))
|
|
|
5111 charset = spec->to_charset;
|
|
|
5112 }
|
|
|
5113 }
|
|
|
5114 }
|
|
|
5115
|
|
|
5116 Dynarr_add (dst, ISO_CODE_ESC);
|
|
|
5117 switch (type)
|
|
|
5118 {
|
|
|
5119 case CHARSET_TYPE_94:
|
|
|
5120 Dynarr_add (dst, inter94[reg]);
|
|
|
5121 break;
|
|
|
5122 case CHARSET_TYPE_96:
|
|
|
5123 Dynarr_add (dst, inter96[reg]);
|
|
|
5124 break;
|
|
|
5125 case CHARSET_TYPE_94X94:
|
|
|
5126 Dynarr_add (dst, '$');
|
|
|
5127 if (reg != 0
|
|
|
5128 || !(CODING_SYSTEM_ISO2022_SHORT (str->codesys))
|
|
|
5129 || final < '@'
|
|
|
5130 || final > 'B')
|
|
|
5131 Dynarr_add (dst, inter94[reg]);
|
|
|
5132 break;
|
|
|
5133 case CHARSET_TYPE_96X96:
|
|
|
5134 Dynarr_add (dst, '$');
|
|
|
5135 Dynarr_add (dst, inter96[reg]);
|
|
|
5136 break;
|
|
|
5137 }
|
|
|
5138 Dynarr_add (dst, final);
|
|
|
5139 }
|
|
|
5140
|
|
|
5141 static void
|
|
|
5142 ensure_normal_shift (struct encoding_stream *str, unsigned_char_dynarr *dst)
|
|
|
5143 {
|
|
|
5144 if (str->iso2022.register_left != 0)
|
|
|
5145 {
|
|
|
5146 Dynarr_add (dst, ISO_CODE_SI);
|
|
|
5147 str->iso2022.register_left = 0;
|
|
|
5148 }
|
|
|
5149 }
|
|
|
5150
|
|
|
5151 static void
|
|
|
5152 ensure_shift_out (struct encoding_stream *str, unsigned_char_dynarr *dst)
|
|
|
5153 {
|
|
|
5154 if (str->iso2022.register_left != 1)
|
|
|
5155 {
|
|
|
5156 Dynarr_add (dst, ISO_CODE_SO);
|
|
|
5157 str->iso2022.register_left = 1;
|
|
|
5158 }
|
|
|
5159 }
|
|
|
5160
|
|
|
5161 /* Convert internally-formatted data to ISO2022 format. */
|
|
|
5162
|
|
|
5163 static void
|
|
398
|
5164 encode_coding_iso2022 (Lstream *encoding, const unsigned char *src,
|
|
259
|
5165 unsigned_char_dynarr *dst, unsigned int n)
|
|
|
5166 {
|
|
|
5167 unsigned char charmask, c;
|
|
|
5168 unsigned char char_boundary;
|
|
|
5169 struct encoding_stream *str = ENCODING_STREAM_DATA (encoding);
|
|
396
|
5170 unsigned int flags = str->flags;
|
|
|
5171 unsigned int ch = str->ch;
|
|
|
5172 Lisp_Coding_System *codesys = str->codesys;
|
|
|
5173 eol_type_t eol_type = CODING_SYSTEM_EOL_TYPE (str->codesys);
|
|
259
|
5174 int i;
|
|
|
5175 Lisp_Object charset;
|
|
|
5176 int half;
|
|
|
5177
|
|
396
|
5178 #ifdef ENABLE_COMPOSITE_CHARS
|
|
259
|
5179 /* flags for handling composite chars. We do a little switcharoo
|
|
|
5180 on the source while we're outputting the composite char. */
|
|
|
5181 unsigned int saved_n = 0;
|
|
398
|
5182 const unsigned char *saved_src = NULL;
|
|
259
|
5183 int in_composite = 0;
|
|
396
|
5184 #endif /* ENABLE_COMPOSITE_CHARS */
|
|
|
5185
|
|
259
|
5186 char_boundary = str->iso2022.current_char_boundary;
|
|
|
5187 charset = str->iso2022.current_charset;
|
|
|
5188 half = str->iso2022.current_half;
|
|
|
5189
|
|
396
|
5190 #ifdef ENABLE_COMPOSITE_CHARS
|
|
259
|
5191 back_to_square_n:
|
|
396
|
5192 #endif
|
|
259
|
5193 while (n--)
|
|
|
5194 {
|
|
|
5195 c = *src++;
|
|
|
5196
|
|
|
5197 if (BYTE_ASCII_P (c))
|
|
|
5198 { /* Processing ASCII character */
|
|
|
5199 ch = 0;
|
|
|
5200
|
|
|
5201 restore_left_to_right_direction (codesys, dst, &flags, 0);
|
|
|
5202
|
|
|
5203 /* Make sure G0 contains ASCII */
|
|
|
5204 if ((c > ' ' && c < ISO_CODE_DEL) ||
|
|
|
5205 !CODING_SYSTEM_ISO2022_NO_ASCII_CNTL (codesys))
|
|
|
5206 {
|
|
|
5207 ensure_normal_shift (str, dst);
|
|
|
5208 iso2022_designate (Vcharset_ascii, 0, str, dst);
|
|
|
5209 }
|
|
|
5210
|
|
|
5211 /* If necessary, restore everything to the default state
|
|
|
5212 at end-of-line */
|
|
|
5213 if (c == '\n' &&
|
|
|
5214 !(CODING_SYSTEM_ISO2022_NO_ASCII_EOL (codesys)))
|
|
|
5215 {
|
|
|
5216 restore_left_to_right_direction (codesys, dst, &flags, 0);
|
|
|
5217
|
|
|
5218 ensure_normal_shift (str, dst);
|
|
|
5219
|
|
|
5220 for (i = 0; i < 4; i++)
|
|
|
5221 {
|
|
|
5222 Lisp_Object initial_charset =
|
|
|
5223 CODING_SYSTEM_ISO2022_INITIAL_CHARSET (codesys, i);
|
|
|
5224 iso2022_designate (initial_charset, i, str, dst);
|
|
|
5225 }
|
|
|
5226 }
|
|
|
5227 if (c == '\n')
|
|
|
5228 {
|
|
|
5229 if (eol_type != EOL_LF && eol_type != EOL_AUTODETECT)
|
|
|
5230 Dynarr_add (dst, '\r');
|
|
|
5231 if (eol_type != EOL_CR)
|
|
|
5232 Dynarr_add (dst, c);
|
|
|
5233 }
|
|
|
5234 else
|
|
|
5235 {
|
|
|
5236 if (CODING_SYSTEM_ISO2022_ESCAPE_QUOTED (codesys)
|
|
|
5237 && fit_to_be_escape_quoted (c))
|
|
|
5238 Dynarr_add (dst, ISO_CODE_ESC);
|
|
|
5239 Dynarr_add (dst, c);
|
|
|
5240 }
|
|
|
5241 char_boundary = 1;
|
|
|
5242 }
|
|
|
5243
|
|
|
5244 else if (BUFBYTE_LEADING_BYTE_P (c) || BUFBYTE_LEADING_BYTE_P (ch))
|
|
|
5245 { /* Processing Leading Byte */
|
|
|
5246 ch = 0;
|
|
|
5247 charset = CHARSET_BY_LEADING_BYTE (c);
|
|
|
5248 if (LEADING_BYTE_PREFIX_P(c))
|
|
|
5249 ch = c;
|
|
|
5250 else if (!EQ (charset, Vcharset_control_1)
|
|
396
|
5251 #ifdef ENABLE_COMPOSITE_CHARS
|
|
|
5252 && !EQ (charset, Vcharset_composite)
|
|
|
5253 #endif
|
|
|
5254 )
|
|
259
|
5255 {
|
|
|
5256 int reg;
|
|
|
5257
|
|
|
5258 ensure_correct_direction (XCHARSET_DIRECTION (charset),
|
|
|
5259 codesys, dst, &flags, 0);
|
|
|
5260
|
|
|
5261 /* Now determine which register to use. */
|
|
|
5262 reg = -1;
|
|
|
5263 for (i = 0; i < 4; i++)
|
|
|
5264 {
|
|
|
5265 if (EQ (charset, str->iso2022.charset[i]) ||
|
|
|
5266 EQ (charset,
|
|
|
5267 CODING_SYSTEM_ISO2022_INITIAL_CHARSET (codesys, i)))
|
|
|
5268 {
|
|
|
5269 reg = i;
|
|
|
5270 break;
|
|
|
5271 }
|
|
|
5272 }
|
|
|
5273
|
|
|
5274 if (reg == -1)
|
|
|
5275 {
|
|
|
5276 if (XCHARSET_GRAPHIC (charset) != 0)
|
|
|
5277 {
|
|
|
5278 if (!NILP (str->iso2022.charset[1]) &&
|
|
|
5279 (!CODING_SYSTEM_ISO2022_SEVEN (codesys) ||
|
|
|
5280 CODING_SYSTEM_ISO2022_LOCK_SHIFT (codesys)))
|
|
|
5281 reg = 1;
|
|
|
5282 else if (!NILP (str->iso2022.charset[2]))
|
|
|
5283 reg = 2;
|
|
|
5284 else if (!NILP (str->iso2022.charset[3]))
|
|
|
5285 reg = 3;
|
|
|
5286 else
|
|
|
5287 reg = 0;
|
|
|
5288 }
|
|
|
5289 else
|
|
|
5290 reg = 0;
|
|
|
5291 }
|
|
|
5292
|
|
|
5293 iso2022_designate (charset, reg, str, dst);
|
|
|
5294
|
|
|
5295 /* Now invoke that register. */
|
|
|
5296 switch (reg)
|
|
|
5297 {
|
|
|
5298 case 0:
|
|
|
5299 ensure_normal_shift (str, dst);
|
|
|
5300 half = 0;
|
|
|
5301 break;
|
|
|
5302
|
|
|
5303 case 1:
|
|
|
5304 if (CODING_SYSTEM_ISO2022_SEVEN (codesys))
|
|
|
5305 {
|
|
|
5306 ensure_shift_out (str, dst);
|
|
|
5307 half = 0;
|
|
|
5308 }
|
|
|
5309 else
|
|
|
5310 half = 1;
|
|
|
5311 break;
|
|
|
5312
|
|
|
5313 case 2:
|
|
|
5314 if (CODING_SYSTEM_ISO2022_SEVEN (str->codesys))
|
|
|
5315 {
|
|
|
5316 Dynarr_add (dst, ISO_CODE_ESC);
|
|
|
5317 Dynarr_add (dst, 'N');
|
|
|
5318 half = 0;
|
|
|
5319 }
|
|
|
5320 else
|
|
|
5321 {
|
|
|
5322 Dynarr_add (dst, ISO_CODE_SS2);
|
|
|
5323 half = 1;
|
|
|
5324 }
|
|
|
5325 break;
|
|
|
5326
|
|
|
5327 case 3:
|
|
|
5328 if (CODING_SYSTEM_ISO2022_SEVEN (str->codesys))
|
|
|
5329 {
|
|
|
5330 Dynarr_add (dst, ISO_CODE_ESC);
|
|
|
5331 Dynarr_add (dst, 'O');
|
|
|
5332 half = 0;
|
|
|
5333 }
|
|
|
5334 else
|
|
|
5335 {
|
|
|
5336 Dynarr_add (dst, ISO_CODE_SS3);
|
|
|
5337 half = 1;
|
|
|
5338 }
|
|
|
5339 break;
|
|
|
5340
|
|
|
5341 default:
|
|
|
5342 abort ();
|
|
|
5343 }
|
|
|
5344 }
|
|
|
5345 char_boundary = 0;
|
|
|
5346 }
|
|
|
5347 else
|
|
|
5348 { /* Processing Non-ASCII character */
|
|
|
5349 charmask = (half == 0 ? 0x7F : 0xFF);
|
|
|
5350 char_boundary = 1;
|
|
|
5351 if (EQ (charset, Vcharset_control_1))
|
|
|
5352 {
|
|
|
5353 if (CODING_SYSTEM_ISO2022_ESCAPE_QUOTED (codesys)
|
|
|
5354 && fit_to_be_escape_quoted (c))
|
|
|
5355 Dynarr_add (dst, ISO_CODE_ESC);
|
|
|
5356 /* you asked for it ... */
|
|
|
5357 Dynarr_add (dst, c - 0x20);
|
|
|
5358 }
|
|
|
5359 else
|
|
|
5360 {
|
|
|
5361 switch (XCHARSET_REP_BYTES (charset))
|
|
|
5362 {
|
|
|
5363 case 2:
|
|
|
5364 Dynarr_add (dst, c & charmask);
|
|
|
5365 break;
|
|
|
5366 case 3:
|
|
|
5367 if (XCHARSET_PRIVATE_P (charset))
|
|
|
5368 {
|
|
|
5369 Dynarr_add (dst, c & charmask);
|
|
|
5370 ch = 0;
|
|
|
5371 }
|
|
|
5372 else if (ch)
|
|
|
5373 {
|
|
396
|
5374 #ifdef ENABLE_COMPOSITE_CHARS
|
|
259
|
5375 if (EQ (charset, Vcharset_composite))
|
|
|
5376 {
|
|
|
5377 if (in_composite)
|
|
|
5378 {
|
|
|
5379 /* #### Bother! We don't know how to
|
|
|
5380 handle this yet. */
|
|
|
5381 Dynarr_add (dst, '~');
|
|
|
5382 }
|
|
|
5383 else
|
|
|
5384 {
|
|
|
5385 Emchar emch = MAKE_CHAR (Vcharset_composite,
|
|
|
5386 ch & 0x7F, c & 0x7F);
|
|
|
5387 Lisp_Object lstr = composite_char_string (emch);
|
|
|
5388 saved_n = n;
|
|
|
5389 saved_src = src;
|
|
|
5390 in_composite = 1;
|
|
|
5391 src = XSTRING_DATA (lstr);
|
|
|
5392 n = XSTRING_LENGTH (lstr);
|
|
|
5393 Dynarr_add (dst, ISO_CODE_ESC);
|
|
|
5394 Dynarr_add (dst, '0'); /* start composing */
|
|
|
5395 }
|
|
|
5396 }
|
|
|
5397 else
|
|
396
|
5398 #endif /* ENABLE_COMPOSITE_CHARS */
|
|
259
|
5399 {
|
|
|
5400 Dynarr_add (dst, ch & charmask);
|
|
|
5401 Dynarr_add (dst, c & charmask);
|
|
|
5402 }
|
|
|
5403 ch = 0;
|
|
|
5404 }
|
|
|
5405 else
|
|
|
5406 {
|
|
|
5407 ch = c;
|
|
|
5408 char_boundary = 0;
|
|
|
5409 }
|
|
|
5410 break;
|
|
|
5411 case 4:
|
|
|
5412 if (ch)
|
|
|
5413 {
|
|
|
5414 Dynarr_add (dst, ch & charmask);
|
|
|
5415 Dynarr_add (dst, c & charmask);
|
|
|
5416 ch = 0;
|
|
|
5417 }
|
|
|
5418 else
|
|
|
5419 {
|
|
|
5420 ch = c;
|
|
|
5421 char_boundary = 0;
|
|
|
5422 }
|
|
|
5423 break;
|
|
|
5424 default:
|
|
|
5425 abort ();
|
|
|
5426 }
|
|
|
5427 }
|
|
|
5428 }
|
|
|
5429 }
|
|
|
5430
|
|
396
|
5431 #ifdef ENABLE_COMPOSITE_CHARS
|
|
259
|
5432 if (in_composite)
|
|
|
5433 {
|
|
|
5434 n = saved_n;
|
|
|
5435 src = saved_src;
|
|
|
5436 in_composite = 0;
|
|
|
5437 Dynarr_add (dst, ISO_CODE_ESC);
|
|
|
5438 Dynarr_add (dst, '1'); /* end composing */
|
|
|
5439 goto back_to_square_n; /* Wheeeeeeeee ..... */
|
|
|
5440 }
|
|
396
|
5441 #endif /* ENABLE_COMPOSITE_CHARS */
|
|
259
|
5442
|
|
|
5443 if (char_boundary && flags & CODING_STATE_END)
|
|
|
5444 {
|
|
|
5445 restore_left_to_right_direction (codesys, dst, &flags, 0);
|
|
|
5446 ensure_normal_shift (str, dst);
|
|
|
5447 for (i = 0; i < 4; i++)
|
|
|
5448 {
|
|
|
5449 Lisp_Object initial_charset =
|
|
|
5450 CODING_SYSTEM_ISO2022_INITIAL_CHARSET (codesys, i);
|
|
|
5451 iso2022_designate (initial_charset, i, str, dst);
|
|
|
5452 }
|
|
|
5453 }
|
|
|
5454
|
|
396
|
5455 str->flags = flags;
|
|
|
5456 str->ch = ch;
|
|
259
|
5457 str->iso2022.current_char_boundary = char_boundary;
|
|
|
5458 str->iso2022.current_charset = charset;
|
|
|
5459 str->iso2022.current_half = half;
|
|
|
5460
|
|
|
5461 /* Verbum caro factum est! */
|
|
|
5462 }
|
|
|
5463 #endif /* MULE */
|
|
|
5464
|
|
|
5465 /************************************************************************/
|
|
|
5466 /* No-conversion methods */
|
|
|
5467 /************************************************************************/
|
|
|
5468
|
|
|
5469 /* This is used when reading in "binary" files -- i.e. files that may
|
|
|
5470 contain all 256 possible byte values and that are not to be
|
|
|
5471 interpreted as being in any particular decoding. */
|
|
|
5472 static void
|
|
398
|
5473 decode_coding_no_conversion (Lstream *decoding, const unsigned char *src,
|
|
259
|
5474 unsigned_char_dynarr *dst, unsigned int n)
|
|
|
5475 {
|
|
|
5476 unsigned char c;
|
|
|
5477 struct decoding_stream *str = DECODING_STREAM_DATA (decoding);
|
|
396
|
5478 unsigned int flags = str->flags;
|
|
|
5479 unsigned int ch = str->ch;
|
|
|
5480 eol_type_t eol_type = str->eol_type;
|
|
259
|
5481
|
|
|
5482 while (n--)
|
|
|
5483 {
|
|
|
5484 c = *src++;
|
|
|
5485
|
|
|
5486 DECODE_HANDLE_EOL_TYPE (eol_type, c, flags, dst);
|
|
|
5487 DECODE_ADD_BINARY_CHAR (c, dst);
|
|
|
5488 label_continue_loop:;
|
|
|
5489 }
|
|
|
5490
|
|
|
5491 DECODE_HANDLE_END_OF_CONVERSION (flags, ch, dst);
|
|
|
5492
|
|
396
|
5493 str->flags = flags;
|
|
|
5494 str->ch = ch;
|
|
259
|
5495 }
|
|
|
5496
|
|
|
5497 static void
|
|
398
|
5498 encode_coding_no_conversion (Lstream *encoding, const unsigned char *src,
|
|
259
|
5499 unsigned_char_dynarr *dst, unsigned int n)
|
|
|
5500 {
|
|
|
5501 unsigned char c;
|
|
|
5502 struct encoding_stream *str = ENCODING_STREAM_DATA (encoding);
|
|
396
|
5503 unsigned int flags = str->flags;
|
|
|
5504 unsigned int ch = str->ch;
|
|
|
5505 eol_type_t eol_type = CODING_SYSTEM_EOL_TYPE (str->codesys);
|
|
259
|
5506
|
|
|
5507 while (n--)
|
|
|
5508 {
|
|
|
5509 c = *src++;
|
|
|
5510 if (c == '\n')
|
|
|
5511 {
|
|
|
5512 if (eol_type != EOL_LF && eol_type != EOL_AUTODETECT)
|
|
|
5513 Dynarr_add (dst, '\r');
|
|
|
5514 if (eol_type != EOL_CR)
|
|
|
5515 Dynarr_add (dst, '\n');
|
|
|
5516 ch = 0;
|
|
|
5517 }
|
|
|
5518 else if (BYTE_ASCII_P (c))
|
|
|
5519 {
|
|
|
5520 assert (ch == 0);
|
|
|
5521 Dynarr_add (dst, c);
|
|
|
5522 }
|
|
|
5523 else if (BUFBYTE_LEADING_BYTE_P (c))
|
|
|
5524 {
|
|
|
5525 assert (ch == 0);
|
|
|
5526 if (c == LEADING_BYTE_LATIN_ISO8859_1 ||
|
|
|
5527 c == LEADING_BYTE_CONTROL_1)
|
|
|
5528 ch = c;
|
|
|
5529 else
|
|
|
5530 Dynarr_add (dst, '~'); /* untranslatable character */
|
|
|
5531 }
|
|
|
5532 else
|
|
|
5533 {
|
|
|
5534 if (ch == LEADING_BYTE_LATIN_ISO8859_1)
|
|
|
5535 Dynarr_add (dst, c);
|
|
|
5536 else if (ch == LEADING_BYTE_CONTROL_1)
|
|
|
5537 {
|
|
|
5538 assert (c < 0xC0);
|
|
|
5539 Dynarr_add (dst, c - 0x20);
|
|
|
5540 }
|
|
|
5541 /* else it should be the second or third byte of an
|
|
|
5542 untranslatable character, so ignore it */
|
|
|
5543 ch = 0;
|
|
|
5544 }
|
|
|
5545 }
|
|
|
5546
|
|
396
|
5547 str->flags = flags;
|
|
|
5548 str->ch = ch;
|
|
259
|
5549 }
|
|
|
5550
|
|
|
5551
|
|
398
|
5552
|
|
259
|
5553 /************************************************************************/
|
|
|
5554 /* Initialization */
|
|
|
5555 /************************************************************************/
|
|
|
5556
|
|
|
5557 void
|
|
398
|
5558 syms_of_file_coding (void)
|
|
259
|
5559 {
|
|
400
|
5560 INIT_LRECORD_IMPLEMENTATION (coding_system);
|
|
|
5561
|
|
259
|
5562 deferror (&Qcoding_system_error, "coding-system-error",
|
|
|
5563 "Coding-system error", Qio_error);
|
|
|
5564
|
|
|
5565 DEFSUBR (Fcoding_system_p);
|
|
|
5566 DEFSUBR (Ffind_coding_system);
|
|
|
5567 DEFSUBR (Fget_coding_system);
|
|
|
5568 DEFSUBR (Fcoding_system_list);
|
|
|
5569 DEFSUBR (Fcoding_system_name);
|
|
|
5570 DEFSUBR (Fmake_coding_system);
|
|
|
5571 DEFSUBR (Fcopy_coding_system);
|
|
398
|
5572 DEFSUBR (Fcoding_system_canonical_name_p);
|
|
|
5573 DEFSUBR (Fcoding_system_alias_p);
|
|
|
5574 DEFSUBR (Fcoding_system_aliasee);
|
|
|
5575 DEFSUBR (Fdefine_coding_system_alias);
|
|
259
|
5576 DEFSUBR (Fsubsidiary_coding_system);
|
|
|
5577
|
|
|
5578 DEFSUBR (Fcoding_system_type);
|
|
|
5579 DEFSUBR (Fcoding_system_doc_string);
|
|
|
5580 #ifdef MULE
|
|
|
5581 DEFSUBR (Fcoding_system_charset);
|
|
|
5582 #endif
|
|
|
5583 DEFSUBR (Fcoding_system_property);
|
|
|
5584
|
|
|
5585 DEFSUBR (Fcoding_category_list);
|
|
|
5586 DEFSUBR (Fset_coding_priority_list);
|
|
|
5587 DEFSUBR (Fcoding_priority_list);
|
|
|
5588 DEFSUBR (Fset_coding_category_system);
|
|
|
5589 DEFSUBR (Fcoding_category_system);
|
|
|
5590
|
|
|
5591 DEFSUBR (Fdetect_coding_region);
|
|
|
5592 DEFSUBR (Fdecode_coding_region);
|
|
|
5593 DEFSUBR (Fencode_coding_region);
|
|
|
5594 #ifdef MULE
|
|
|
5595 DEFSUBR (Fdecode_shift_jis_char);
|
|
|
5596 DEFSUBR (Fencode_shift_jis_char);
|
|
|
5597 DEFSUBR (Fdecode_big5_char);
|
|
|
5598 DEFSUBR (Fencode_big5_char);
|
|
396
|
5599 DEFSUBR (Fset_ucs_char);
|
|
|
5600 DEFSUBR (Fucs_char);
|
|
|
5601 DEFSUBR (Fset_char_ucs);
|
|
|
5602 DEFSUBR (Fchar_ucs);
|
|
259
|
5603 #endif /* MULE */
|
|
398
|
5604 defsymbol (&Qcoding_systemp, "coding-system-p");
|
|
259
|
5605 defsymbol (&Qno_conversion, "no-conversion");
|
|
398
|
5606 defsymbol (&Qraw_text, "raw-text");
|
|
259
|
5607 #ifdef MULE
|
|
|
5608 defsymbol (&Qbig5, "big5");
|
|
|
5609 defsymbol (&Qshift_jis, "shift-jis");
|
|
396
|
5610 defsymbol (&Qucs4, "ucs-4");
|
|
|
5611 defsymbol (&Qutf8, "utf-8");
|
|
259
|
5612 defsymbol (&Qccl, "ccl");
|
|
|
5613 defsymbol (&Qiso2022, "iso2022");
|
|
|
5614 #endif /* MULE */
|
|
|
5615 defsymbol (&Qmnemonic, "mnemonic");
|
|
|
5616 defsymbol (&Qeol_type, "eol-type");
|
|
|
5617 defsymbol (&Qpost_read_conversion, "post-read-conversion");
|
|
|
5618 defsymbol (&Qpre_write_conversion, "pre-write-conversion");
|
|
|
5619
|
|
|
5620 defsymbol (&Qcr, "cr");
|
|
|
5621 defsymbol (&Qlf, "lf");
|
|
|
5622 defsymbol (&Qcrlf, "crlf");
|
|
|
5623 defsymbol (&Qeol_cr, "eol-cr");
|
|
|
5624 defsymbol (&Qeol_lf, "eol-lf");
|
|
|
5625 defsymbol (&Qeol_crlf, "eol-crlf");
|
|
|
5626 #ifdef MULE
|
|
|
5627 defsymbol (&Qcharset_g0, "charset-g0");
|
|
|
5628 defsymbol (&Qcharset_g1, "charset-g1");
|
|
|
5629 defsymbol (&Qcharset_g2, "charset-g2");
|
|
|
5630 defsymbol (&Qcharset_g3, "charset-g3");
|
|
|
5631 defsymbol (&Qforce_g0_on_output, "force-g0-on-output");
|
|
|
5632 defsymbol (&Qforce_g1_on_output, "force-g1-on-output");
|
|
|
5633 defsymbol (&Qforce_g2_on_output, "force-g2-on-output");
|
|
|
5634 defsymbol (&Qforce_g3_on_output, "force-g3-on-output");
|
|
|
5635 defsymbol (&Qno_iso6429, "no-iso6429");
|
|
|
5636 defsymbol (&Qinput_charset_conversion, "input-charset-conversion");
|
|
|
5637 defsymbol (&Qoutput_charset_conversion, "output-charset-conversion");
|
|
272
|
5638
|
|
259
|
5639 defsymbol (&Qshort, "short");
|
|
|
5640 defsymbol (&Qno_ascii_eol, "no-ascii-eol");
|
|
|
5641 defsymbol (&Qno_ascii_cntl, "no-ascii-cntl");
|
|
|
5642 defsymbol (&Qseven, "seven");
|
|
|
5643 defsymbol (&Qlock_shift, "lock-shift");
|
|
|
5644 defsymbol (&Qescape_quoted, "escape-quoted");
|
|
272
|
5645 #endif /* MULE */
|
|
259
|
5646 defsymbol (&Qencode, "encode");
|
|
|
5647 defsymbol (&Qdecode, "decode");
|
|
|
5648
|
|
|
5649 #ifdef MULE
|
|
|
5650 defsymbol (&coding_category_symbol[CODING_CATEGORY_SHIFT_JIS],
|
|
|
5651 "shift-jis");
|
|
|
5652 defsymbol (&coding_category_symbol[CODING_CATEGORY_BIG5],
|
|
|
5653 "big5");
|
|
396
|
5654 defsymbol (&coding_category_symbol[CODING_CATEGORY_UCS4],
|
|
|
5655 "ucs-4");
|
|
|
5656 defsymbol (&coding_category_symbol[CODING_CATEGORY_UTF8],
|
|
|
5657 "utf-8");
|
|
259
|
5658 defsymbol (&coding_category_symbol[CODING_CATEGORY_ISO_7],
|
|
|
5659 "iso-7");
|
|
|
5660 defsymbol (&coding_category_symbol[CODING_CATEGORY_ISO_8_DESIGNATE],
|
|
|
5661 "iso-8-designate");
|
|
|
5662 defsymbol (&coding_category_symbol[CODING_CATEGORY_ISO_8_1],
|
|
|
5663 "iso-8-1");
|
|
|
5664 defsymbol (&coding_category_symbol[CODING_CATEGORY_ISO_8_2],
|
|
|
5665 "iso-8-2");
|
|
|
5666 defsymbol (&coding_category_symbol[CODING_CATEGORY_ISO_LOCK_SHIFT],
|
|
|
5667 "iso-lock-shift");
|
|
272
|
5668 #endif /* MULE */
|
|
259
|
5669 defsymbol (&coding_category_symbol[CODING_CATEGORY_NO_CONVERSION],
|
|
|
5670 "no-conversion");
|
|
|
5671 }
|
|
|
5672
|
|
|
5673 void
|
|
398
|
5674 lstream_type_create_file_coding (void)
|
|
259
|
5675 {
|
|
|
5676 LSTREAM_HAS_METHOD (decoding, reader);
|
|
|
5677 LSTREAM_HAS_METHOD (decoding, writer);
|
|
|
5678 LSTREAM_HAS_METHOD (decoding, rewinder);
|
|
|
5679 LSTREAM_HAS_METHOD (decoding, seekable_p);
|
|
|
5680 LSTREAM_HAS_METHOD (decoding, flusher);
|
|
|
5681 LSTREAM_HAS_METHOD (decoding, closer);
|
|
|
5682 LSTREAM_HAS_METHOD (decoding, marker);
|
|
|
5683
|
|
|
5684 LSTREAM_HAS_METHOD (encoding, reader);
|
|
|
5685 LSTREAM_HAS_METHOD (encoding, writer);
|
|
|
5686 LSTREAM_HAS_METHOD (encoding, rewinder);
|
|
|
5687 LSTREAM_HAS_METHOD (encoding, seekable_p);
|
|
|
5688 LSTREAM_HAS_METHOD (encoding, flusher);
|
|
|
5689 LSTREAM_HAS_METHOD (encoding, closer);
|
|
|
5690 LSTREAM_HAS_METHOD (encoding, marker);
|
|
|
5691 }
|
|
|
5692
|
|
|
5693 void
|
|
398
|
5694 vars_of_file_coding (void)
|
|
259
|
5695 {
|
|
|
5696 int i;
|
|
|
5697
|
|
398
|
5698 fcd = xnew (struct file_coding_dump);
|
|
|
5699 dumpstruct (&fcd, &fcd_description);
|
|
|
5700
|
|
259
|
5701 /* Initialize to something reasonable ... */
|
|
|
5702 for (i = 0; i <= CODING_CATEGORY_LAST; i++)
|
|
|
5703 {
|
|
398
|
5704 fcd->coding_category_system[i] = Qnil;
|
|
|
5705 fcd->coding_category_by_priority[i] = i;
|
|
259
|
5706 }
|
|
|
5707
|
|
|
5708 Fprovide (intern ("file-coding"));
|
|
|
5709
|
|
|
5710 DEFVAR_LISP ("keyboard-coding-system", &Vkeyboard_coding_system /*
|
|
|
5711 Coding system used for TTY keyboard input.
|
|
|
5712 Not used under a windowing system.
|
|
|
5713 */ );
|
|
|
5714 Vkeyboard_coding_system = Qnil;
|
|
|
5715
|
|
|
5716 DEFVAR_LISP ("terminal-coding-system", &Vterminal_coding_system /*
|
|
|
5717 Coding system used for TTY display output.
|
|
|
5718 Not used under a windowing system.
|
|
|
5719 */ );
|
|
|
5720 Vterminal_coding_system = Qnil;
|
|
|
5721
|
|
|
5722 DEFVAR_LISP ("coding-system-for-read", &Vcoding_system_for_read /*
|
|
398
|
5723 Overriding coding system used when reading from a file or process.
|
|
|
5724 You should bind this variable with `let', but do not set it globally.
|
|
|
5725 If this is non-nil, it specifies the coding system that will be used
|
|
|
5726 to decode input on read operations, such as from a file or process.
|
|
|
5727 It overrides `buffer-file-coding-system-for-read',
|
|
259
|
5728 `insert-file-contents-pre-hook', etc. Use those variables instead of
|
|
398
|
5729 this one for permanent changes to the environment. */ );
|
|
259
|
5730 Vcoding_system_for_read = Qnil;
|
|
|
5731
|
|
|
5732 DEFVAR_LISP ("coding-system-for-write",
|
|
|
5733 &Vcoding_system_for_write /*
|
|
398
|
5734 Overriding coding system used when writing to a file or process.
|
|
|
5735 You should bind this variable with `let', but do not set it globally.
|
|
|
5736 If this is non-nil, it specifies the coding system that will be used
|
|
|
5737 to encode output for write operations, such as to a file or process.
|
|
|
5738 It overrides `buffer-file-coding-system', `write-region-pre-hook', etc.
|
|
|
5739 Use those variables instead of this one for permanent changes to the
|
|
|
5740 environment. */ );
|
|
259
|
5741 Vcoding_system_for_write = Qnil;
|
|
|
5742
|
|
|
5743 DEFVAR_LISP ("file-name-coding-system", &Vfile_name_coding_system /*
|
|
|
5744 Coding system used to convert pathnames when accessing files.
|
|
|
5745 */ );
|
|
|
5746 Vfile_name_coding_system = Qnil;
|
|
|
5747
|
|
|
5748 DEFVAR_BOOL ("enable-multibyte-characters", &enable_multibyte_characters /*
|
|
|
5749 Non-nil means the buffer contents are regarded as multi-byte form
|
|
|
5750 of characters, not a binary code. This affects the display, file I/O,
|
|
|
5751 and behaviors of various editing commands.
|
|
|
5752
|
|
|
5753 Setting this to nil does not do anything.
|
|
|
5754 */ );
|
|
|
5755 enable_multibyte_characters = 1;
|
|
|
5756 }
|
|
|
5757
|
|
|
5758 void
|
|
398
|
5759 complex_vars_of_file_coding (void)
|
|
259
|
5760 {
|
|
380
|
5761 staticpro (&Vcoding_system_hash_table);
|
|
|
5762 Vcoding_system_hash_table =
|
|
|
5763 make_lisp_hash_table (50, HASH_TABLE_NON_WEAK, HASH_TABLE_EQ);
|
|
259
|
5764
|
|
|
5765 the_codesys_prop_dynarr = Dynarr_new (codesys_prop);
|
|
398
|
5766 dumpstruct (&the_codesys_prop_dynarr, &codesys_prop_dynarr_description);
|
|
259
|
5767
|
|
|
5768 #define DEFINE_CODESYS_PROP(Prop_Type, Sym) do \
|
|
|
5769 { \
|
|
|
5770 struct codesys_prop csp; \
|
|
|
5771 csp.sym = (Sym); \
|
|
|
5772 csp.prop_type = (Prop_Type); \
|
|
|
5773 Dynarr_add (the_codesys_prop_dynarr, csp); \
|
|
|
5774 } while (0)
|
|
|
5775
|
|
|
5776 DEFINE_CODESYS_PROP (CODESYS_PROP_ALL_OK, Qmnemonic);
|
|
|
5777 DEFINE_CODESYS_PROP (CODESYS_PROP_ALL_OK, Qeol_type);
|
|
|
5778 DEFINE_CODESYS_PROP (CODESYS_PROP_ALL_OK, Qeol_cr);
|
|
|
5779 DEFINE_CODESYS_PROP (CODESYS_PROP_ALL_OK, Qeol_crlf);
|
|
|
5780 DEFINE_CODESYS_PROP (CODESYS_PROP_ALL_OK, Qeol_lf);
|
|
|
5781 DEFINE_CODESYS_PROP (CODESYS_PROP_ALL_OK, Qpost_read_conversion);
|
|
|
5782 DEFINE_CODESYS_PROP (CODESYS_PROP_ALL_OK, Qpre_write_conversion);
|
|
|
5783 #ifdef MULE
|
|
|
5784 DEFINE_CODESYS_PROP (CODESYS_PROP_ISO2022, Qcharset_g0);
|
|
|
5785 DEFINE_CODESYS_PROP (CODESYS_PROP_ISO2022, Qcharset_g1);
|
|
|
5786 DEFINE_CODESYS_PROP (CODESYS_PROP_ISO2022, Qcharset_g2);
|
|
|
5787 DEFINE_CODESYS_PROP (CODESYS_PROP_ISO2022, Qcharset_g3);
|
|
|
5788 DEFINE_CODESYS_PROP (CODESYS_PROP_ISO2022, Qforce_g0_on_output);
|
|
|
5789 DEFINE_CODESYS_PROP (CODESYS_PROP_ISO2022, Qforce_g1_on_output);
|
|
|
5790 DEFINE_CODESYS_PROP (CODESYS_PROP_ISO2022, Qforce_g2_on_output);
|
|
|
5791 DEFINE_CODESYS_PROP (CODESYS_PROP_ISO2022, Qforce_g3_on_output);
|
|
|
5792 DEFINE_CODESYS_PROP (CODESYS_PROP_ISO2022, Qshort);
|
|
|
5793 DEFINE_CODESYS_PROP (CODESYS_PROP_ISO2022, Qno_ascii_eol);
|
|
|
5794 DEFINE_CODESYS_PROP (CODESYS_PROP_ISO2022, Qno_ascii_cntl);
|
|
|
5795 DEFINE_CODESYS_PROP (CODESYS_PROP_ISO2022, Qseven);
|
|
|
5796 DEFINE_CODESYS_PROP (CODESYS_PROP_ISO2022, Qlock_shift);
|
|
|
5797 DEFINE_CODESYS_PROP (CODESYS_PROP_ISO2022, Qno_iso6429);
|
|
|
5798 DEFINE_CODESYS_PROP (CODESYS_PROP_ISO2022, Qescape_quoted);
|
|
|
5799 DEFINE_CODESYS_PROP (CODESYS_PROP_ISO2022, Qinput_charset_conversion);
|
|
|
5800 DEFINE_CODESYS_PROP (CODESYS_PROP_ISO2022, Qoutput_charset_conversion);
|
|
|
5801
|
|
|
5802 DEFINE_CODESYS_PROP (CODESYS_PROP_CCL, Qencode);
|
|
|
5803 DEFINE_CODESYS_PROP (CODESYS_PROP_CCL, Qdecode);
|
|
|
5804 #endif /* MULE */
|
|
|
5805 /* Need to create this here or we're really screwed. */
|
|
398
|
5806 Fmake_coding_system
|
|
|
5807 (Qraw_text, Qno_conversion,
|
|
|
5808 build_string ("Raw text, which means it converts only line-break-codes."),
|
|
|
5809 list2 (Qmnemonic, build_string ("Raw")));
|
|
|
5810
|
|
|
5811 Fmake_coding_system
|
|
|
5812 (Qbinary, Qno_conversion,
|
|
|
5813 build_string ("Binary, which means it does not convert anything."),
|
|
|
5814 list4 (Qeol_type, Qlf,
|
|
|
5815 Qmnemonic, build_string ("Binary")));
|
|
|
5816
|
|
|
5817 Fdefine_coding_system_alias (Qno_conversion, Qraw_text);
|
|
|
5818
|
|
|
5819 Fdefine_coding_system_alias (Qfile_name, Qbinary);
|
|
|
5820
|
|
|
5821 Fdefine_coding_system_alias (Qterminal, Qbinary);
|
|
|
5822 Fdefine_coding_system_alias (Qkeyboard, Qbinary);
|
|
259
|
5823
|
|
|
5824 /* Need this for bootstrapping */
|
|
398
|
5825 fcd->coding_category_system[CODING_CATEGORY_NO_CONVERSION] =
|
|
|
5826 Fget_coding_system (Qraw_text);
|
|
396
|
5827
|
|
|
5828 #ifdef MULE
|
|
|
5829 {
|
|
|
5830 unsigned int i;
|
|
|
5831
|
|
|
5832 for (i = 0; i < 65536; i++)
|
|
398
|
5833 fcd->ucs_to_mule_table[i] = Qnil;
|
|
396
|
5834 }
|
|
|
5835 staticpro (&mule_to_ucs_table);
|
|
|
5836 mule_to_ucs_table = Fmake_char_table(Qgeneric);
|
|
|
5837 #endif /* MULE */
|
|
259
|
5838 }
|
