Mercurial > hg > xemacs-beta
annotate src/file-coding.c @ 4537:7ca6d57ce12d
Clarify syntax in configure.ac following Vladimir Ivanovic's change.
2008-12-27 Aidan Kehoe <kehoea@parhasard.net>
* configure.ac: Vladimir Ivanovic's change of 2008-09-17 was
effective, but not correct. The original issue was that m4
stripped [], so "char *v[]" in configure.ac became "char *v" in
configure. His change to configure.ac rendered the declaration
"char **v[]", which is effective but confusing; this change
renders it "char **v".
* configure: Regenerate.
author | Aidan Kehoe <kehoea@parhasard.net> |
---|---|
date | Sat, 27 Dec 2008 15:30:50 +0000 |
parents | 726060ee587c |
children | 80e0588fb42f |
rev | line source |
---|---|
771 | 1 /* Text encoding conversion functions; coding-system object. |
2 #### rename me to coding-system.c or coding.c | |
428 | 3 Copyright (C) 1991, 1995 Free Software Foundation, Inc. |
4 Copyright (C) 1995 Sun Microsystems, Inc. | |
3025 | 5 Copyright (C) 2000, 2001, 2002, 2003, 2005 Ben Wing. |
428 | 6 |
7 This file is part of XEmacs. | |
8 | |
9 XEmacs is free software; you can redistribute it and/or modify it | |
10 under the terms of the GNU General Public License as published by the | |
11 Free Software Foundation; either version 2, or (at your option) any | |
12 later version. | |
13 | |
14 XEmacs is distributed in the hope that it will be useful, but WITHOUT | |
15 ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or | |
16 FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License | |
17 for more details. | |
18 | |
19 You should have received a copy of the GNU General Public License | |
20 along with XEmacs; see the file COPYING. If not, write to | |
21 the Free Software Foundation, Inc., 59 Temple Place - Suite 330, | |
22 Boston, MA 02111-1307, USA. */ | |
23 | |
771 | 24 /* Synched up with: Not in FSF. */ |
25 | |
26 /* Authorship: | |
27 | |
28 Current primary author: Ben Wing <ben@xemacs.org> | |
29 | |
30 Rewritten by Ben Wing <ben@xemacs.org>, based originally on coding.c | |
31 from Mule 2.? but probably does not share one line of code with that | |
32 original source. Rewriting work started around Dec. 1994. or Jan. 1995. | |
33 Proceeded in earnest till Nov. 1995. | |
34 | |
35 Around Feb. 17, 1998, Andy Piper renamed what was then mule-coding.c to | |
36 file-coding.c, with the intention of using it to do end-of-line conversion | |
37 on non-MULE machines (specifically, on Windows machines). He separated | |
38 out the MULE stuff from non-MULE using ifdef's, and searched throughout | |
39 the rest of the source tree looking for coding-system-related code that | |
40 was ifdef MULE but should be ifdef HAVE_CODING_SYSTEMS. | |
41 | |
42 Sept. 4 - 8, 1998, Tomohiko Morioka added the UCS_4 and UTF_8 coding system | |
43 types, providing a primitive means of decoding and encoding externally- | |
44 formatted Unicode/UCS_4 and Unicode/UTF_8 data. | |
45 | |
46 January 25, 2000, Martin Buchholz redid and fleshed out the coding | |
47 system alias handling that was first added in prototype form by | |
48 Hrjove Niksic, April 15, 1999. | |
49 | |
50 April to May 2000, Ben Wing: More major reorganization. Adding features | |
51 needed for MS Windows (multibyte, unicode, unicode-to-multibyte), the | |
52 "chain" coding system for chaining two together, and doing a lot of | |
53 reorganization in preparation for properly abstracting out the different | |
54 coding system types. | |
55 | |
56 June 2001, Ben Wing: Added Unicode support. Eliminated previous | |
57 junky Unicode translation support. | |
58 | |
59 August 2001, Ben Wing: Moved Unicode support to unicode.c. Finished | |
60 abstracting everything except detection, which is hard to abstract (see | |
61 just below). | |
62 | |
63 September 2001, Ben Wing: Moved Mule code to mule-coding.c, Windows code | |
64 to intl-win32.c. Lots more rewriting; very little code is untouched | |
65 from before April 2000. Abstracted the detection code, added multiple | |
66 levels of likelihood to increase the reliability of the algorithm. | |
67 | |
68 October 2001, Ben Wing: HAVE_CODING_SYSTEMS is always now defined. | |
69 Removed the conditionals. | |
70 */ | |
71 | |
428 | 72 #include <config.h> |
73 #include "lisp.h" | |
74 | |
75 #include "buffer.h" | |
76 #include "elhash.h" | |
77 #include "insdel.h" | |
78 #include "lstream.h" | |
440 | 79 #include "opaque.h" |
771 | 80 #include "file-coding.h" |
81 | |
82 #ifdef HAVE_ZLIB | |
83 #include "zlib.h" | |
428 | 84 #endif |
85 | |
86 Lisp_Object Vkeyboard_coding_system; | |
87 Lisp_Object Vterminal_coding_system; | |
88 Lisp_Object Vcoding_system_for_read; | |
89 Lisp_Object Vcoding_system_for_write; | |
90 Lisp_Object Vfile_name_coding_system; | |
91 | |
771 | 92 #ifdef DEBUG_XEMACS |
93 Lisp_Object Vdebug_coding_detection; | |
440 | 94 #endif |
771 | 95 |
96 typedef struct coding_system_type_entry | |
97 { | |
98 struct coding_system_methods *meths; | |
99 } coding_system_type_entry; | |
100 | |
101 typedef struct | |
102 { | |
103 Dynarr_declare (coding_system_type_entry); | |
104 } coding_system_type_entry_dynarr; | |
105 | |
106 static coding_system_type_entry_dynarr *the_coding_system_type_entry_dynarr; | |
107 | |
1204 | 108 static const struct memory_description cste_description_1[] = { |
2551 | 109 { XD_BLOCK_PTR, offsetof (coding_system_type_entry, meths), 1, |
110 { &coding_system_methods_description } }, | |
771 | 111 { XD_END } |
112 }; | |
113 | |
1204 | 114 static const struct sized_memory_description cste_description = { |
771 | 115 sizeof (coding_system_type_entry), |
116 cste_description_1 | |
117 }; | |
118 | |
1204 | 119 static const struct memory_description csted_description_1[] = { |
771 | 120 XD_DYNARR_DESC (coding_system_type_entry_dynarr, &cste_description), |
428 | 121 { XD_END } |
122 }; | |
123 | |
1204 | 124 static const struct sized_memory_description csted_description = { |
771 | 125 sizeof (coding_system_type_entry_dynarr), |
126 csted_description_1 | |
127 }; | |
128 | |
129 static Lisp_Object Vcoding_system_type_list; | |
130 | |
131 /* Coding system currently associated with each coding category. */ | |
132 Lisp_Object coding_category_system[MAX_DETECTOR_CATEGORIES]; | |
133 | |
134 /* Table of all coding categories in decreasing order of priority. | |
135 This describes a permutation of the possible coding categories. */ | |
136 int coding_category_by_priority[MAX_DETECTOR_CATEGORIES]; | |
137 | |
138 /* Value used with to give a unique name to nameless coding systems */ | |
139 int coding_system_tick; | |
140 | |
141 int coding_detector_count; | |
142 int coding_detector_category_count; | |
143 | |
144 detector_dynarr *all_coding_detectors; | |
145 | |
1204 | 146 static const struct memory_description struct_detector_category_description_1[] |
771 | 147 = |
148 { | |
149 { XD_LISP_OBJECT, offsetof (struct detector_category, sym) }, | |
150 { XD_END } | |
151 }; | |
152 | |
1204 | 153 static const struct sized_memory_description struct_detector_category_description = |
771 | 154 { |
155 sizeof (struct detector_category), | |
156 struct_detector_category_description_1 | |
428 | 157 }; |
158 | |
1204 | 159 static const struct memory_description detector_category_dynarr_description_1[] = |
771 | 160 { |
161 XD_DYNARR_DESC (detector_category_dynarr, | |
162 &struct_detector_category_description), | |
163 { XD_END } | |
164 }; | |
165 | |
1204 | 166 static const struct sized_memory_description detector_category_dynarr_description = { |
771 | 167 sizeof (detector_category_dynarr), |
168 detector_category_dynarr_description_1 | |
169 }; | |
170 | |
1204 | 171 static const struct memory_description struct_detector_description_1[] |
771 | 172 = |
173 { | |
2367 | 174 { XD_BLOCK_PTR, offsetof (struct detector, cats), 1, |
2551 | 175 { &detector_category_dynarr_description } }, |
771 | 176 { XD_END } |
177 }; | |
178 | |
1204 | 179 static const struct sized_memory_description struct_detector_description = |
771 | 180 { |
181 sizeof (struct detector), | |
182 struct_detector_description_1 | |
183 }; | |
184 | |
1204 | 185 static const struct memory_description detector_dynarr_description_1[] = |
771 | 186 { |
187 XD_DYNARR_DESC (detector_dynarr, &struct_detector_description), | |
188 { XD_END } | |
189 }; | |
190 | |
1204 | 191 static const struct sized_memory_description detector_dynarr_description = { |
771 | 192 sizeof (detector_dynarr), |
193 detector_dynarr_description_1 | |
194 }; | |
428 | 195 |
196 Lisp_Object Qcoding_systemp; | |
197 | |
771 | 198 Lisp_Object Qraw_text; |
428 | 199 |
200 Lisp_Object Qmnemonic, Qeol_type; | |
201 Lisp_Object Qcr, Qcrlf, Qlf; | |
202 Lisp_Object Qeol_cr, Qeol_crlf, Qeol_lf; | |
203 Lisp_Object Qpost_read_conversion; | |
204 Lisp_Object Qpre_write_conversion; | |
205 | |
771 | 206 Lisp_Object Qtranslation_table_for_decode; |
207 Lisp_Object Qtranslation_table_for_encode; | |
208 Lisp_Object Qsafe_chars; | |
209 Lisp_Object Qsafe_charsets; | |
210 Lisp_Object Qmime_charset; | |
211 Lisp_Object Qvalid_codes; | |
212 | |
213 Lisp_Object Qno_conversion; | |
214 Lisp_Object Qconvert_eol; | |
440 | 215 Lisp_Object Qescape_quoted; |
771 | 216 Lisp_Object Qencode, Qdecode; |
217 | |
218 Lisp_Object Qconvert_eol_lf, Qconvert_eol_cr, Qconvert_eol_crlf; | |
219 Lisp_Object Qconvert_eol_autodetect; | |
220 | |
221 Lisp_Object Qnear_certainty, Qquite_probable, Qsomewhat_likely; | |
1494 | 222 Lisp_Object Qslightly_likely; |
771 | 223 Lisp_Object Qas_likely_as_unlikely, Qsomewhat_unlikely, Qquite_improbable; |
224 Lisp_Object Qnearly_impossible; | |
225 | |
226 Lisp_Object Qdo_eol, Qdo_coding; | |
227 | |
228 Lisp_Object Qcanonicalize_after_coding; | |
229 | |
1347 | 230 Lisp_Object QScoding_system_cookie; |
231 | |
4303 | 232 Lisp_Object Qposix_charset_to_coding_system_hash; |
233 | |
771 | 234 /* This is used to convert autodetected coding systems into existing |
235 systems. For example, the chain undecided->convert-eol-autodetect may | |
236 have its separate parts detected as mswindows-multibyte and | |
237 convert-eol-crlf, and the result needs to be mapped to | |
238 mswindows-multibyte-dos. */ | |
239 /* #### It's not clear we need this whole chain-canonicalize mechanism | |
240 any more. */ | |
241 static Lisp_Object Vchain_canonicalize_hash_table; | |
242 | |
243 #ifdef HAVE_ZLIB | |
244 Lisp_Object Qgzip; | |
428 | 245 #endif |
771 | 246 |
2297 | 247 /* Maps symbols (coding system names) to either coding system objects or |
248 (for aliases) other names. */ | |
771 | 249 static Lisp_Object Vcoding_system_hash_table; |
428 | 250 |
251 int enable_multibyte_characters; | |
252 | |
253 EXFUN (Fcopy_coding_system, 2); | |
254 | |
255 | |
256 /************************************************************************/ | |
771 | 257 /* Coding system object methods */ |
428 | 258 /************************************************************************/ |
259 | |
260 static Lisp_Object | |
261 mark_coding_system (Lisp_Object obj) | |
262 { | |
263 Lisp_Coding_System *codesys = XCODING_SYSTEM (obj); | |
264 | |
1204 | 265 #define MARKED_SLOT(x) mark_object (codesys->x); |
266 #include "coding-system-slots.h" | |
771 | 267 |
268 MAYBE_CODESYSMETH (codesys, mark, (obj)); | |
428 | 269 |
1204 | 270 return Qnil; |
428 | 271 } |
272 | |
273 static void | |
771 | 274 print_coding_system_properties (Lisp_Object obj, Lisp_Object printcharfun) |
275 { | |
276 Lisp_Coding_System *c = XCODING_SYSTEM (obj); | |
277 print_internal (c->methods->type, printcharfun, 1); | |
278 MAYBE_CODESYSMETH (c, print, (obj, printcharfun, 1)); | |
279 if (CODING_SYSTEM_EOL_TYPE (c) != EOL_AUTODETECT) | |
280 write_fmt_string_lisp (printcharfun, " eol-type=%s", | |
281 1, Fcoding_system_property (obj, Qeol_type)); | |
282 } | |
283 | |
284 static void | |
428 | 285 print_coding_system (Lisp_Object obj, Lisp_Object printcharfun, |
2286 | 286 int UNUSED (escapeflag)) |
428 | 287 { |
288 Lisp_Coding_System *c = XCODING_SYSTEM (obj); | |
289 if (print_readably) | |
771 | 290 printing_unreadable_object |
291 ("printing unreadable object #<coding-system 0x%x>", c->header.uid); | |
292 | |
293 write_fmt_string_lisp (printcharfun, "#<coding-system %s ", 1, c->name); | |
294 print_coding_system_properties (obj, printcharfun); | |
826 | 295 write_c_string (printcharfun, ">"); |
428 | 296 } |
297 | |
771 | 298 /* Print an abbreviated version of a coding system (but still containing |
299 all the information), for use within a coding system print method. */ | |
300 | |
301 static void | |
302 print_coding_system_in_print_method (Lisp_Object cs, Lisp_Object printcharfun, | |
2286 | 303 int UNUSED (escapeflag)) |
771 | 304 { |
800 | 305 write_fmt_string_lisp (printcharfun, "%s[", 1, XCODING_SYSTEM_NAME (cs)); |
771 | 306 print_coding_system_properties (cs, printcharfun); |
826 | 307 write_c_string (printcharfun, "]"); |
771 | 308 } |
309 | |
3263 | 310 #ifndef NEW_GC |
428 | 311 static void |
312 finalize_coding_system (void *header, int for_disksave) | |
313 { | |
771 | 314 Lisp_Object cs = wrap_coding_system ((Lisp_Coding_System *) header); |
428 | 315 /* Since coding systems never go away, this function is not |
316 necessary. But it would be necessary if we changed things | |
317 so that coding systems could go away. */ | |
318 if (!for_disksave) /* see comment in lstream.c */ | |
771 | 319 MAYBE_XCODESYSMETH (cs, finalize, (cs)); |
320 } | |
3263 | 321 #endif /* not NEW_GC */ |
771 | 322 |
323 static Bytecount | |
324 sizeof_coding_system (const void *header) | |
325 { | |
326 const Lisp_Coding_System *p = (const Lisp_Coding_System *) header; | |
327 return offsetof (Lisp_Coding_System, data) + p->methods->extra_data_size; | |
428 | 328 } |
329 | |
1204 | 330 static const struct memory_description coding_system_methods_description_1[] |
771 | 331 = { |
332 { XD_LISP_OBJECT, | |
333 offsetof (struct coding_system_methods, type) }, | |
334 { XD_LISP_OBJECT, | |
335 offsetof (struct coding_system_methods, predicate_symbol) }, | |
336 { XD_END } | |
337 }; | |
338 | |
1204 | 339 const struct sized_memory_description coding_system_methods_description = { |
771 | 340 sizeof (struct coding_system_methods), |
341 coding_system_methods_description_1 | |
342 }; | |
343 | |
1204 | 344 static const struct sized_memory_description coding_system_extra_description_map[] = |
345 { | |
346 { offsetof (Lisp_Coding_System, methods) }, | |
347 { offsetof (struct coding_system_methods, extra_description) }, | |
348 { -1 }, | |
771 | 349 }; |
350 | |
1204 | 351 static const struct memory_description coding_system_description[] = |
428 | 352 { |
2367 | 353 { XD_BLOCK_PTR, offsetof (Lisp_Coding_System, methods), 1, |
2551 | 354 { &coding_system_methods_description } }, |
1204 | 355 #define MARKED_SLOT(x) { XD_LISP_OBJECT, offsetof (Lisp_Coding_System, x) }, |
356 #define MARKED_SLOT_ARRAY(slot, size) \ | |
357 { XD_LISP_OBJECT_ARRAY, offsetof (Lisp_Coding_System, slot), size }, | |
358 #include "coding-system-slots.h" | |
2367 | 359 { XD_BLOCK_ARRAY, offsetof (Lisp_Coding_System, data), 1, |
2551 | 360 { coding_system_extra_description_map } }, |
1204 | 361 { XD_END } |
771 | 362 }; |
363 | |
1204 | 364 static const struct memory_description coding_system_empty_extra_description_1[] = |
365 { | |
366 { XD_END } | |
367 }; | |
368 | |
369 const struct sized_memory_description coding_system_empty_extra_description = { | |
370 0, coding_system_empty_extra_description_1 | |
371 }; | |
372 | |
3263 | 373 #ifdef NEW_GC |
374 DEFINE_LRECORD_SEQUENCE_IMPLEMENTATION ("coding-system", coding_system, | |
375 1, /*dumpable-flag*/ | |
376 mark_coding_system, | |
377 print_coding_system, | |
378 0, 0, 0, coding_system_description, | |
379 sizeof_coding_system, | |
380 Lisp_Coding_System); | |
381 #else /* not NEW_GC */ | |
934 | 382 DEFINE_LRECORD_SEQUENCE_IMPLEMENTATION ("coding-system", coding_system, |
383 1, /*dumpable-flag*/ | |
384 mark_coding_system, | |
385 print_coding_system, | |
386 finalize_coding_system, | |
387 0, 0, coding_system_description, | |
388 sizeof_coding_system, | |
389 Lisp_Coding_System); | |
3263 | 390 #endif /* not NEW_GC */ |
771 | 391 |
392 /************************************************************************/ | |
393 /* Creating coding systems */ | |
394 /************************************************************************/ | |
395 | |
396 static struct coding_system_methods * | |
397 decode_coding_system_type (Lisp_Object type, Error_Behavior errb) | |
428 | 398 { |
771 | 399 int i; |
400 | |
401 for (i = 0; i < Dynarr_length (the_coding_system_type_entry_dynarr); i++) | |
428 | 402 { |
771 | 403 if (EQ (type, |
404 Dynarr_at (the_coding_system_type_entry_dynarr, i).meths->type)) | |
405 return Dynarr_at (the_coding_system_type_entry_dynarr, i).meths; | |
428 | 406 } |
771 | 407 |
408 maybe_invalid_constant ("Invalid coding system type", type, | |
409 Qcoding_system, errb); | |
410 | |
411 return 0; | |
428 | 412 } |
413 | |
771 | 414 static int |
415 valid_coding_system_type_p (Lisp_Object type) | |
428 | 416 { |
771 | 417 return decode_coding_system_type (type, ERROR_ME_NOT) != 0; |
418 } | |
419 | |
420 DEFUN ("valid-coding-system-type-p", Fvalid_coding_system_type_p, 1, 1, 0, /* | |
421 Given a CODING-SYSTEM-TYPE, return non-nil if it is valid. | |
422 Valid types depend on how XEmacs was compiled but may include | |
3025 | 423 `undecided', `chain', `integer', `ccl', `iso2022', `big5', `shift-jis', |
424 `utf-16', `ucs-4', `utf-8', etc. | |
771 | 425 */ |
426 (coding_system_type)) | |
427 { | |
428 return valid_coding_system_type_p (coding_system_type) ? Qt : Qnil; | |
429 } | |
430 | |
431 DEFUN ("coding-system-type-list", Fcoding_system_type_list, 0, 0, 0, /* | |
432 Return a list of valid coding system types. | |
433 */ | |
434 ()) | |
435 { | |
436 return Fcopy_sequence (Vcoding_system_type_list); | |
437 } | |
438 | |
439 void | |
440 add_entry_to_coding_system_type_list (struct coding_system_methods *meths) | |
441 { | |
442 struct coding_system_type_entry entry; | |
443 | |
444 entry.meths = meths; | |
445 Dynarr_add (the_coding_system_type_entry_dynarr, entry); | |
446 Vcoding_system_type_list = Fcons (meths->type, Vcoding_system_type_list); | |
428 | 447 } |
448 | |
449 DEFUN ("coding-system-p", Fcoding_system_p, 1, 1, 0, /* | |
450 Return t if OBJECT is a coding system. | |
451 A coding system is an object that defines how text containing multiple | |
452 character sets is encoded into a stream of (typically 8-bit) bytes. | |
453 The coding system is used to decode the stream into a series of | |
454 characters (which may be from multiple charsets) when the text is read | |
455 from a file or process, and is used to encode the text back into the | |
456 same format when it is written out to a file or process. | |
457 | |
458 For example, many ISO2022-compliant coding systems (such as Compound | |
459 Text, which is used for inter-client data under the X Window System) | |
460 use escape sequences to switch between different charsets -- Japanese | |
461 Kanji, for example, is invoked with "ESC $ ( B"; ASCII is invoked | |
462 with "ESC ( B"; and Cyrillic is invoked with "ESC - L". See | |
463 `make-coding-system' for more information. | |
464 | |
465 Coding systems are normally identified using a symbol, and the | |
466 symbol is accepted in place of the actual coding system object whenever | |
467 a coding system is called for. (This is similar to how faces work.) | |
468 */ | |
469 (object)) | |
470 { | |
471 return CODING_SYSTEMP (object) ? Qt : Qnil; | |
472 } | |
473 | |
4303 | 474 static Lisp_Object |
475 find_coding_system (Lisp_Object coding_system_or_name, | |
476 int do_autoloads) | |
477 { | |
478 Lisp_Object lookup; | |
479 | |
480 if (NILP (coding_system_or_name)) | |
481 coding_system_or_name = Qbinary; | |
482 else if (CODING_SYSTEMP (coding_system_or_name)) | |
483 return coding_system_or_name; | |
484 else | |
485 CHECK_SYMBOL (coding_system_or_name); | |
486 | |
487 while (1) | |
488 { | |
489 lookup = | |
490 Fgethash (coding_system_or_name, Vcoding_system_hash_table, Qnil); | |
491 | |
492 if (CONSP (lookup) && do_autoloads) | |
493 { | |
494 struct gcpro gcpro1; | |
495 int length; | |
496 DECLARE_EISTRING (desired_base); | |
497 DECLARE_EISTRING (warning_info); | |
498 | |
499 eicpy_lstr (desired_base, XSYMBOL_NAME (coding_system_or_name)); | |
500 | |
501 /* Work out the name of the base coding system. */ | |
502 length = eilen (desired_base); | |
503 if (length > (int)(sizeof ("-unix") - 1)) | |
504 { | |
505 if (0 == qxestrcmp ((UAscbyte *)"-unix", (eidata (desired_base)) | |
506 + (length - (sizeof ("-unix") - 1)))) | |
507 { | |
508 eidel (desired_base, length - (sizeof ("-unix") - 1), | |
509 -1, 5, 5); | |
510 } | |
511 } | |
512 else if (length > (int)(sizeof ("-dos") - 1)) | |
513 { | |
514 if ((0 == qxestrcmp ((UAscbyte *)"-dos", (eidata (desired_base)) | |
515 + (length - (sizeof ("-dos") - 1)))) || | |
516 (0 == qxestrcmp ((UAscbyte *)"-mac", (eidata (desired_base)) | |
517 + (length - (sizeof ("-mac") - 1))))) | |
518 { | |
519 eidel (desired_base, length - (sizeof ("-dos") - 1), -1, | |
520 4, 4); | |
521 } | |
522 } | |
523 | |
524 coding_system_or_name = intern_int (eidata (desired_base)); | |
525 | |
526 /* Remove this coding system and its subsidiary coding | |
527 systems from the hash, to avoid calling this code recursively. */ | |
528 Fremhash (coding_system_or_name, Vcoding_system_hash_table); | |
529 Fremhash (add_suffix_to_symbol(coding_system_or_name, "-unix"), | |
530 Vcoding_system_hash_table); | |
531 Fremhash (add_suffix_to_symbol(coding_system_or_name, "-dos"), | |
532 Vcoding_system_hash_table); | |
533 Fremhash (add_suffix_to_symbol(coding_system_or_name, "-mac"), | |
534 Vcoding_system_hash_table); | |
535 | |
536 eicpy_ascii (warning_info, "Error autoloading coding system "); | |
537 eicat_lstr (warning_info, XSYMBOL_NAME (coding_system_or_name)); | |
538 | |
539 /* Keep around the form so it doesn't disappear from under | |
540 #'eval's feet. */ | |
541 GCPRO1 (lookup); | |
542 call1_trapping_problems ((const CIbyte *)eidata (warning_info), | |
543 Qeval, lookup, 0); | |
544 UNGCPRO; | |
545 | |
546 lookup = | |
547 Fgethash (coding_system_or_name, Vcoding_system_hash_table, Qnil); | |
548 } | |
549 | |
550 if (CODING_SYSTEMP (lookup) || NILP (lookup)) | |
551 return lookup; | |
552 | |
553 coding_system_or_name = lookup; | |
554 } | |
555 } | |
556 | |
428 | 557 DEFUN ("find-coding-system", Ffind_coding_system, 1, 1, 0, /* |
558 Retrieve the coding system of the given name. | |
559 | |
560 If CODING-SYSTEM-OR-NAME is a coding-system object, it is simply | |
561 returned. Otherwise, CODING-SYSTEM-OR-NAME should be a symbol. | |
562 If there is no such coding system, nil is returned. Otherwise the | |
563 associated coding system object is returned. | |
564 */ | |
565 (coding_system_or_name)) | |
566 { | |
4303 | 567 return find_coding_system(coding_system_or_name, 1); |
568 } | |
569 | |
570 DEFUN ("autoload-coding-system", Fautoload_coding_system, 2, 2, 0, /* | |
571 Define SYMBOL as a coding-system that is loaded on demand. | |
572 | |
573 FORM is a form to evaluate to define the coding-system. | |
574 */ | |
575 (symbol, form)) | |
576 { | |
577 Lisp_Object lookup; | |
578 | |
579 CHECK_SYMBOL (symbol); | |
580 CHECK_CONS (form); | |
581 | |
582 lookup = find_coding_system (symbol, 0); | |
583 | |
584 if (!NILP (lookup) && | |
585 /* Allow autoloads to be redefined. */ | |
586 !CONSP (lookup)) | |
440 | 587 { |
4303 | 588 invalid_operation ("Cannot redefine existing coding system", |
589 symbol); | |
440 | 590 } |
4303 | 591 |
592 Fputhash (symbol, form, Vcoding_system_hash_table); | |
593 Fputhash (add_suffix_to_symbol(symbol, "-unix"), form, | |
594 Vcoding_system_hash_table); | |
595 Fputhash (add_suffix_to_symbol(symbol, "-dos"), form, | |
596 Vcoding_system_hash_table); | |
597 Fputhash (add_suffix_to_symbol(symbol, "-mac"), form, | |
598 Vcoding_system_hash_table); | |
599 | |
600 /* Tell the POSIX locale infrastructure about this coding system (though | |
601 unfortunately it'll be too late for the startup locale sniffing. */ | |
602 if (!UNBOUNDP (Qposix_charset_to_coding_system_hash)) | |
603 { | |
604 Lisp_Object val = Fsymbol_value (Qposix_charset_to_coding_system_hash); | |
605 DECLARE_EISTRING (minimal_name); | |
606 Ibyte *full_name; | |
607 int len = XSTRING_LENGTH (XSYMBOL_NAME (symbol)), i; | |
608 | |
609 if (!NILP (val)) | |
610 { | |
611 full_name = XSTRING_DATA (XSYMBOL_NAME (symbol)); | |
612 for (i = 0; i < len; ++i) | |
613 { | |
614 if (full_name[i] >= '0' && full_name[i] <= '9') | |
615 { | |
616 eicat_ch (minimal_name, full_name[i]); | |
617 } | |
618 else if (full_name[i] >= 'a' && full_name[i] <= 'z') | |
619 { | |
620 eicat_ch (minimal_name, full_name[i]); | |
621 } | |
622 else if (full_name[i] >= 'A' && full_name[i] <= 'Z') | |
623 { | |
624 eicat_ch (minimal_name, full_name[i] + | |
625 ('a' - 'A')); | |
626 } | |
627 } | |
628 | |
629 if (eilen (minimal_name)) | |
630 { | |
631 CHECK_HASH_TABLE (val); | |
632 Fputhash (eimake_string(minimal_name), symbol, val); | |
633 } | |
634 } | |
635 } | |
636 | |
637 return Qt; | |
428 | 638 } |
639 | |
640 DEFUN ("get-coding-system", Fget_coding_system, 1, 1, 0, /* | |
641 Retrieve the coding system of the given name. | |
642 Same as `find-coding-system' except that if there is no such | |
643 coding system, an error is signaled instead of returning nil. | |
644 */ | |
645 (name)) | |
646 { | |
647 Lisp_Object coding_system = Ffind_coding_system (name); | |
648 | |
649 if (NILP (coding_system)) | |
563 | 650 invalid_argument ("No such coding system", name); |
428 | 651 return coding_system; |
652 } | |
653 | |
771 | 654 int |
655 coding_system_is_binary (Lisp_Object coding_system) | |
656 { | |
657 Lisp_Coding_System *cs = XCODING_SYSTEM (coding_system); | |
658 return | |
659 (EQ (CODING_SYSTEM_TYPE (cs), Qno_conversion) && | |
660 CODING_SYSTEM_EOL_TYPE (cs) == EOL_LF && | |
661 EQ (CODING_SYSTEM_POST_READ_CONVERSION (cs), Qnil) && | |
662 EQ (CODING_SYSTEM_PRE_WRITE_CONVERSION (cs), Qnil)); | |
663 } | |
664 | |
665 static Lisp_Object | |
666 coding_system_real_canonical (Lisp_Object cs) | |
667 { | |
668 if (!NILP (XCODING_SYSTEM_CANONICAL (cs))) | |
669 return XCODING_SYSTEM_CANONICAL (cs); | |
670 return cs; | |
671 } | |
672 | |
673 /* Return true if coding system is of the "standard" type that decodes | |
674 bytes into characters (suitable for decoding a text file). */ | |
675 int | |
676 coding_system_is_for_text_file (Lisp_Object coding_system) | |
677 { | |
678 return (XCODESYSMETH_OR_GIVEN | |
679 (coding_system, conversion_end_type, | |
680 (coding_system_real_canonical (coding_system)), | |
681 DECODES_BYTE_TO_CHARACTER) == | |
682 DECODES_BYTE_TO_CHARACTER); | |
683 } | |
684 | |
685 static int | |
686 decoding_source_sink_type_is_char (Lisp_Object cs, enum source_or_sink sex) | |
687 { | |
688 enum source_sink_type type = | |
689 XCODESYSMETH_OR_GIVEN (cs, conversion_end_type, | |
690 (coding_system_real_canonical (cs)), | |
691 DECODES_BYTE_TO_CHARACTER); | |
692 if (sex == CODING_SOURCE) | |
693 return (type == DECODES_CHARACTER_TO_CHARACTER || | |
694 type == DECODES_CHARACTER_TO_BYTE); | |
695 else | |
696 return (type == DECODES_CHARACTER_TO_CHARACTER || | |
697 type == DECODES_BYTE_TO_CHARACTER); | |
698 } | |
699 | |
700 static int | |
701 encoding_source_sink_type_is_char (Lisp_Object cs, enum source_or_sink sex) | |
702 { | |
703 return decoding_source_sink_type_is_char (cs, | |
704 /* Sex change */ | |
705 sex == CODING_SOURCE ? | |
706 CODING_SINK : CODING_SOURCE); | |
707 } | |
708 | |
709 /* Like Ffind_coding_system() but check that the coding system is of the | |
710 "standard" type that decodes bytes into characters (suitable for | |
711 decoding a text file), and if not, returns an appropriate wrapper that | |
712 does. Also, if EOL_WRAP is non-zero, check whether this coding system | |
713 wants EOL auto-detection, and if so, wrap with a convert-eol coding | |
714 system to do this. */ | |
715 | |
716 Lisp_Object | |
717 find_coding_system_for_text_file (Lisp_Object name, int eol_wrap) | |
718 { | |
719 Lisp_Object coding_system = Ffind_coding_system (name); | |
720 Lisp_Object wrapper = coding_system; | |
721 | |
722 if (NILP (coding_system)) | |
723 return Qnil; | |
724 if (!coding_system_is_for_text_file (coding_system)) | |
725 { | |
726 wrapper = XCODING_SYSTEM_TEXT_FILE_WRAPPER (coding_system); | |
727 if (NILP (wrapper)) | |
728 { | |
729 Lisp_Object chain; | |
730 if (!decoding_source_sink_type_is_char (coding_system, CODING_SINK)) | |
731 chain = list2 (coding_system, Qbinary); | |
732 else | |
733 chain = list1 (coding_system); | |
734 if (decoding_source_sink_type_is_char (coding_system, CODING_SOURCE)) | |
735 chain = Fcons (Qbinary, chain); | |
736 wrapper = | |
737 make_internal_coding_system | |
738 (coding_system, | |
739 "internal-text-file-wrapper", | |
740 Qchain, | |
741 Qunbound, list4 (Qchain, chain, | |
742 Qcanonicalize_after_coding, coding_system)); | |
743 XCODING_SYSTEM_TEXT_FILE_WRAPPER (coding_system) = wrapper; | |
744 } | |
745 } | |
746 | |
747 if (!eol_wrap || XCODING_SYSTEM_EOL_TYPE (coding_system) != EOL_AUTODETECT) | |
748 return wrapper; | |
749 | |
750 coding_system = wrapper; | |
751 wrapper = XCODING_SYSTEM_AUTO_EOL_WRAPPER (coding_system); | |
752 if (!NILP (wrapper)) | |
753 return wrapper; | |
754 wrapper = | |
755 make_internal_coding_system | |
756 (coding_system, | |
757 "internal-auto-eol-wrapper", | |
758 Qundecided, Qunbound, | |
759 list4 (Qcoding_system, coding_system, | |
760 Qdo_eol, Qt)); | |
761 XCODING_SYSTEM_AUTO_EOL_WRAPPER (coding_system) = wrapper; | |
762 return wrapper; | |
763 } | |
764 | |
765 /* Like Fget_coding_system() but verify that the coding system is of the | |
766 "standard" type that decodes bytes into characters (suitable for | |
767 decoding a text file), and if not, returns an appropriate wrapper that | |
768 does. Also, if EOL_WRAP is non-zero, check whether this coding system | |
769 wants EOL auto-detection, and if so, wrap with a convert-eol coding | |
770 system to do this. */ | |
771 | |
772 Lisp_Object | |
773 get_coding_system_for_text_file (Lisp_Object name, int eol_wrap) | |
774 { | |
775 Lisp_Object coding_system = find_coding_system_for_text_file (name, | |
776 eol_wrap); | |
777 if (NILP (coding_system)) | |
778 invalid_argument ("No such coding system", name); | |
779 return coding_system; | |
780 } | |
781 | |
782 /* We store the coding systems in hash tables with the names as the | |
783 key and the actual coding system object as the value. Occasionally | |
784 we need to use them in a list format. These routines provide us | |
785 with that. */ | |
428 | 786 struct coding_system_list_closure |
787 { | |
788 Lisp_Object *coding_system_list; | |
771 | 789 int normal; |
790 int internal; | |
428 | 791 }; |
792 | |
793 static int | |
4303 | 794 add_coding_system_to_list_mapper (Lisp_Object key, Lisp_Object value, |
428 | 795 void *coding_system_list_closure) |
796 { | |
797 /* This function can GC */ | |
798 struct coding_system_list_closure *cscl = | |
799 (struct coding_system_list_closure *) coding_system_list_closure; | |
800 Lisp_Object *coding_system_list = cscl->coding_system_list; | |
801 | |
771 | 802 /* We can't just use VALUE because KEY might be an alias, and we need |
4303 | 803 the real coding system object. |
804 | |
805 Autoloaded coding systems have conses for their values, and can't be | |
806 internal coding systems, or coding system aliases. */ | |
807 if (CONSP (value) || | |
808 (XCODING_SYSTEM (Ffind_coding_system (key))->internal_p ? | |
809 cscl->internal : cscl->normal)) | |
771 | 810 *coding_system_list = Fcons (key, *coding_system_list); |
428 | 811 return 0; |
812 } | |
813 | |
2297 | 814 /* #### should we specify a conventional for "all coding systems"? */ |
771 | 815 DEFUN ("coding-system-list", Fcoding_system_list, 0, 1, 0, /* |
428 | 816 Return a list of the names of all defined coding systems. |
771 | 817 If INTERNAL is nil, only the normal (non-internal) coding systems are |
818 included. (Internal coding systems are created for various internal | |
819 purposes, such as implementing EOL types of CRLF and CR; generally, you do | |
820 not want to see these.) If it is t, only the internal coding systems are | |
821 included. If it is any other non-nil value both normal and internal are | |
822 included. | |
428 | 823 */ |
771 | 824 (internal)) |
428 | 825 { |
826 Lisp_Object coding_system_list = Qnil; | |
827 struct gcpro gcpro1; | |
828 struct coding_system_list_closure coding_system_list_closure; | |
829 | |
830 GCPRO1 (coding_system_list); | |
831 coding_system_list_closure.coding_system_list = &coding_system_list; | |
771 | 832 coding_system_list_closure.normal = !EQ (internal, Qt); |
833 coding_system_list_closure.internal = !NILP (internal); | |
428 | 834 elisp_maphash (add_coding_system_to_list_mapper, Vcoding_system_hash_table, |
835 &coding_system_list_closure); | |
836 UNGCPRO; | |
837 | |
838 return coding_system_list; | |
839 } | |
840 | |
841 DEFUN ("coding-system-name", Fcoding_system_name, 1, 1, 0, /* | |
842 Return the name of the given coding system. | |
843 */ | |
844 (coding_system)) | |
845 { | |
846 coding_system = Fget_coding_system (coding_system); | |
847 return XCODING_SYSTEM_NAME (coding_system); | |
848 } | |
849 | |
850 static Lisp_Coding_System * | |
771 | 851 allocate_coding_system (struct coding_system_methods *codesys_meths, |
852 Bytecount data_size, | |
853 Lisp_Object name) | |
428 | 854 { |
771 | 855 Bytecount total_size = offsetof (Lisp_Coding_System, data) + data_size; |
428 | 856 Lisp_Coding_System *codesys = |
3024 | 857 (Lisp_Coding_System *) BASIC_ALLOC_LCRECORD (total_size, |
1204 | 858 &lrecord_coding_system); |
859 | |
771 | 860 codesys->methods = codesys_meths; |
1204 | 861 #define MARKED_SLOT(x) codesys->x = Qnil; |
862 #include "coding-system-slots.h" | |
863 | |
771 | 864 CODING_SYSTEM_EOL_TYPE (codesys) = EOL_LF; |
865 CODING_SYSTEM_NAME (codesys) = name; | |
866 | |
867 MAYBE_CODESYSMETH (codesys, init, (wrap_coding_system (codesys))); | |
428 | 868 |
869 return codesys; | |
870 } | |
871 | |
771 | 872 static enum eol_type |
873 symbol_to_eol_type (Lisp_Object symbol) | |
874 { | |
875 CHECK_SYMBOL (symbol); | |
876 if (NILP (symbol)) return EOL_AUTODETECT; | |
877 if (EQ (symbol, Qlf)) return EOL_LF; | |
878 if (EQ (symbol, Qcrlf)) return EOL_CRLF; | |
879 if (EQ (symbol, Qcr)) return EOL_CR; | |
880 | |
881 invalid_constant ("Unrecognized eol type", symbol); | |
1204 | 882 RETURN_NOT_REACHED (EOL_AUTODETECT); |
771 | 883 } |
884 | |
885 static Lisp_Object | |
886 eol_type_to_symbol (enum eol_type type) | |
887 { | |
888 switch (type) | |
889 { | |
2500 | 890 default: ABORT (); |
771 | 891 case EOL_LF: return Qlf; |
892 case EOL_CRLF: return Qcrlf; | |
893 case EOL_CR: return Qcr; | |
894 case EOL_AUTODETECT: return Qnil; | |
895 } | |
896 } | |
897 | |
898 struct subsidiary_type | |
899 { | |
4528
726060ee587c
First draft of g++ 4.3 warning removal patch. Builds. *Needs ChangeLogs.*
Stephen J. Turnbull <stephen@xemacs.org>
parents:
4522
diff
changeset
|
900 const Ascbyte *extension; |
726060ee587c
First draft of g++ 4.3 warning removal patch. Builds. *Needs ChangeLogs.*
Stephen J. Turnbull <stephen@xemacs.org>
parents:
4522
diff
changeset
|
901 const Ascbyte *mnemonic_ext; |
771 | 902 enum eol_type eol; |
903 }; | |
904 | |
905 static struct subsidiary_type coding_subsidiary_list[] = | |
906 { { "-unix", "", EOL_LF }, | |
907 { "-dos", ":T", EOL_CRLF }, | |
908 { "-mac", ":t", EOL_CR } }; | |
909 | |
910 /* kludge */ | |
428 | 911 static void |
771 | 912 setup_eol_coding_systems (Lisp_Object codesys) |
428 | 913 { |
793 | 914 int len = XSTRING_LENGTH (XSYMBOL (XCODING_SYSTEM_NAME (codesys))->name); |
2367 | 915 Ibyte *codesys_name = alloca_ibytes (len + 7); |
771 | 916 int mlen = -1; |
867 | 917 Ibyte *codesys_mnemonic = 0; |
771 | 918 Lisp_Object codesys_name_sym, sub_codesys; |
919 int i; | |
920 | |
921 memcpy (codesys_name, | |
793 | 922 XSTRING_DATA (XSYMBOL (XCODING_SYSTEM_NAME (codesys))->name), len); |
771 | 923 |
924 if (STRINGP (XCODING_SYSTEM_MNEMONIC (codesys))) | |
428 | 925 { |
771 | 926 mlen = XSTRING_LENGTH (XCODING_SYSTEM_MNEMONIC (codesys)); |
2367 | 927 codesys_mnemonic = alloca_ibytes (mlen + 7); |
771 | 928 memcpy (codesys_mnemonic, |
929 XSTRING_DATA (XCODING_SYSTEM_MNEMONIC (codesys)), mlen); | |
930 } | |
931 | |
932 /* Create three "subsidiary" coding systems, decoding data encoded using | |
933 each of the three EOL types. We do this for each subsidiary by | |
934 copying the original coding system, setting the EOL type | |
935 appropriately, and setting the CANONICAL member of the new coding | |
936 system to be a chain consisting of the original coding system followed | |
937 by a convert-eol coding system to do the EOL decoding. For EOL type | |
938 LF, however, we don't need any decoding, so we skip creating a | |
939 CANONICAL. | |
940 | |
941 If the original coding system is not a text-type coding system | |
942 (decodes byte->char), we need to coerce it to one by the appropriate | |
943 wrapping in CANONICAL. */ | |
944 | |
945 for (i = 0; i < countof (coding_subsidiary_list); i++) | |
946 { | |
4528
726060ee587c
First draft of g++ 4.3 warning removal patch. Builds. *Needs ChangeLogs.*
Stephen J. Turnbull <stephen@xemacs.org>
parents:
4522
diff
changeset
|
947 const Ascbyte *extension = coding_subsidiary_list[i].extension; |
726060ee587c
First draft of g++ 4.3 warning removal patch. Builds. *Needs ChangeLogs.*
Stephen J. Turnbull <stephen@xemacs.org>
parents:
4522
diff
changeset
|
948 const Ascbyte *mnemonic_ext = coding_subsidiary_list[i].mnemonic_ext; |
771 | 949 enum eol_type eol = coding_subsidiary_list[i].eol; |
950 | |
2367 | 951 qxestrcpy_ascii (codesys_name + len, extension); |
771 | 952 codesys_name_sym = intern_int (codesys_name); |
953 if (mlen != -1) | |
2367 | 954 qxestrcpy_ascii (codesys_mnemonic + mlen, mnemonic_ext); |
771 | 955 |
956 sub_codesys = Fcopy_coding_system (codesys, codesys_name_sym); | |
957 if (mlen != -1) | |
958 XCODING_SYSTEM_MNEMONIC (sub_codesys) = | |
959 build_intstring (codesys_mnemonic); | |
960 | |
961 if (eol != EOL_LF) | |
962 { | |
963 Lisp_Object chain = list2 (get_coding_system_for_text_file | |
964 (codesys, 0), | |
965 eol == EOL_CR ? Qconvert_eol_cr : | |
966 Qconvert_eol_crlf); | |
967 Lisp_Object canon = | |
968 make_internal_coding_system | |
969 (sub_codesys, "internal-subsidiary-eol-wrapper", | |
970 Qchain, Qunbound, | |
971 mlen != -1 ? | |
972 list6 (Qmnemonic, build_intstring (codesys_mnemonic), | |
973 Qchain, chain, | |
974 Qcanonicalize_after_coding, sub_codesys) : | |
975 list4 (Qchain, chain, | |
976 Qcanonicalize_after_coding, sub_codesys)); | |
977 XCODING_SYSTEM_CANONICAL (sub_codesys) = canon; | |
978 } | |
979 XCODING_SYSTEM_EOL_TYPE (sub_codesys) = eol; | |
980 XCODING_SYSTEM_SUBSIDIARY_PARENT (sub_codesys) = codesys; | |
981 XCODING_SYSTEM (codesys)->eol[eol] = sub_codesys; | |
428 | 982 } |
983 } | |
984 | |
771 | 985 /* Basic function to create new coding systems. For `make-coding-system', |
986 NAME-OR-EXISTING is the NAME argument, PREFIX is null, and TYPE, | |
987 DESCRIPTION, and PROPS are the same. All created coding systems are put | |
988 in a hash table indexed by NAME. | |
989 | |
990 If PREFIX is a string, NAME-OR-EXISTING should specify an existing | |
991 coding system (or nil), and an internal coding system will be created. | |
992 The name of the coding system will be constructed by combining PREFIX | |
993 with the name of the existing coding system (if given), and a number | |
994 will be appended to insure uniqueness. In such a case, if Qunbound is | |
995 given for DESCRIPTION, the description gets created based on the | |
996 generated name. Also, if no mnemonic is given in the properties list, a | |
997 mnemonic is created based on the generated name. | |
998 | |
999 For internal coding systems, the coding system is marked as internal | |
1000 (see `coding-system-list'), and no subsidiaries will be created or | |
1001 eol-wrapping will happen. Otherwise: | |
1002 | |
1003 -- if the eol-type property is `lf' or t, the coding system is merely | |
1004 created and returned. (For t, the coding system will be wrapped with | |
1005 an EOL autodetector when it's used to read a file.) | |
1006 | |
1007 -- if eol-type is `crlf' or `cr', after the coding system object is | |
1008 created, it will be wrapped in a chain with the appropriate | |
1009 convert-eol coding system (either `convert-eol-crlf' or | |
1010 `convert-eol-cr'), so that CRLF->LF or CR->LF conversion is done at | |
1011 decoding time, and the opposite at encoding time. The resulting | |
1012 chain becomes the CANONICAL field of the coding system object. | |
1013 | |
1014 -- if eol-type is nil or omitted, "subsidiaries" are generated: Three | |
1015 coding systems where the original coding system (before wrapping with | |
1016 convert-eol-autodetect) is either unwrapped or wrapped with | |
1017 convert-eol-crlf or convert-eol-cr, respectively, so that coding systems | |
1018 to handle LF, CRLF, and CR end-of-line indicators are created. (This | |
1019 crazy crap is based on existing behavior in other Mule versions, | |
1020 including FSF Emacs.) | |
1021 */ | |
428 | 1022 |
1023 static Lisp_Object | |
4528
726060ee587c
First draft of g++ 4.3 warning removal patch. Builds. *Needs ChangeLogs.*
Stephen J. Turnbull <stephen@xemacs.org>
parents:
4522
diff
changeset
|
1024 make_coding_system_1 (Lisp_Object name_or_existing, const Ascbyte *prefix, |
771 | 1025 Lisp_Object type, Lisp_Object description, |
1026 Lisp_Object props) | |
428 | 1027 { |
771 | 1028 Lisp_Coding_System *cs; |
1029 int need_to_setup_eol_systems = 1; | |
1030 enum eol_type eol_wrapper = EOL_AUTODETECT; | |
1031 struct coding_system_methods *meths; | |
1032 Lisp_Object csobj; | |
1033 Lisp_Object defmnem = Qnil; | |
1034 | |
1035 if (NILP (type)) | |
1036 type = Qundecided; | |
1037 meths = decode_coding_system_type (type, ERROR_ME); | |
1038 | |
1039 if (prefix) | |
428 | 1040 { |
867 | 1041 Ibyte *newname = |
771 | 1042 emacs_sprintf_malloc (NULL, "%s-%s-%d", |
1043 prefix, | |
867 | 1044 NILP (name_or_existing) ? (Ibyte *) "nil" : |
771 | 1045 XSTRING_DATA (Fsymbol_name (XCODING_SYSTEM_NAME |
1046 (name_or_existing))), | |
1047 ++coding_system_tick); | |
1048 name_or_existing = intern_int (newname); | |
1726 | 1049 xfree (newname, Ibyte *); |
771 | 1050 |
1051 if (UNBOUNDP (description)) | |
1052 { | |
1053 newname = | |
1054 emacs_sprintf_malloc | |
1055 (NULL, "For Internal Use (%s)", | |
1056 XSTRING_DATA (Fsymbol_name (name_or_existing))); | |
1057 description = build_intstring (newname); | |
1726 | 1058 xfree (newname, Ibyte *); |
771 | 1059 } |
1060 | |
1061 newname = emacs_sprintf_malloc (NULL, "Int%d", coding_system_tick); | |
1062 defmnem = build_intstring (newname); | |
1726 | 1063 xfree (newname, Ibyte *); |
428 | 1064 } |
771 | 1065 else |
1066 CHECK_SYMBOL (name_or_existing); | |
1067 | |
4303 | 1068 /* See is there an entry for name_or_existing in the defined coding system |
1069 hash table. */ | |
1070 csobj = find_coding_system (name_or_existing, 0); | |
1071 /* Error if it's there and not an autoload form. */ | |
1072 if (!NILP (csobj) && !CONSP (csobj)) | |
771 | 1073 invalid_operation ("Cannot redefine existing coding system", |
4303 | 1074 name_or_existing); |
771 | 1075 |
1076 cs = allocate_coding_system (meths, meths->extra_data_size, | |
1077 name_or_existing); | |
793 | 1078 csobj = wrap_coding_system (cs); |
771 | 1079 |
1080 cs->internal_p = !!prefix; | |
1081 | |
1082 if (NILP (description)) | |
1083 description = build_string (""); | |
1084 else | |
1085 CHECK_STRING (description); | |
1086 CODING_SYSTEM_DESCRIPTION (cs) = description; | |
1087 | |
1088 if (!NILP (defmnem)) | |
1089 CODING_SYSTEM_MNEMONIC (cs) = defmnem; | |
1090 | |
1091 { | |
1092 EXTERNAL_PROPERTY_LIST_LOOP_3 (key, value, props) | |
1093 { | |
1094 int recognized = 1; | |
1095 | |
1096 if (EQ (key, Qmnemonic)) | |
1097 { | |
1098 if (!NILP (value)) | |
1099 CHECK_STRING (value); | |
1100 CODING_SYSTEM_MNEMONIC (cs) = value; | |
1101 } | |
1102 | |
1103 else if (EQ (key, Qdocumentation)) | |
1104 { | |
1105 if (!NILP (value)) | |
1106 CHECK_STRING (value); | |
1107 CODING_SYSTEM_DOCUMENTATION (cs) = value; | |
1108 } | |
1109 | |
1110 else if (EQ (key, Qeol_type)) | |
1111 { | |
1112 need_to_setup_eol_systems = NILP (value); | |
1113 if (EQ (value, Qt)) | |
1114 value = Qnil; | |
1115 eol_wrapper = symbol_to_eol_type (value); | |
1116 } | |
1117 | |
1118 else if (EQ (key, Qpost_read_conversion)) | |
1119 CODING_SYSTEM_POST_READ_CONVERSION (cs) = value; | |
1120 else if (EQ (key, Qpre_write_conversion)) | |
1121 CODING_SYSTEM_PRE_WRITE_CONVERSION (cs) = value; | |
1122 /* FSF compatibility */ | |
1123 else if (EQ (key, Qtranslation_table_for_decode)) | |
1124 ; | |
1125 else if (EQ (key, Qtranslation_table_for_encode)) | |
1126 ; | |
1127 else if (EQ (key, Qsafe_chars)) | |
1128 ; | |
1129 else if (EQ (key, Qsafe_charsets)) | |
1130 ; | |
1131 else if (EQ (key, Qmime_charset)) | |
1132 ; | |
1133 else if (EQ (key, Qvalid_codes)) | |
1134 ; | |
1135 else | |
1136 recognized = CODESYSMETH_OR_GIVEN (cs, putprop, | |
1137 (csobj, key, value), 0); | |
1138 | |
1139 if (!recognized) | |
1140 invalid_constant ("Unrecognized property", key); | |
1141 } | |
1142 } | |
1143 | |
1144 { | |
1145 XCODING_SYSTEM_CANONICAL (csobj) = | |
1146 CODESYSMETH_OR_GIVEN (cs, canonicalize, (csobj), Qnil); | |
1147 XCODING_SYSTEM_EOL_TYPE (csobj) = EOL_AUTODETECT; /* for copy-coding-system | |
1148 below */ | |
1149 | |
4303 | 1150 Fputhash (name_or_existing, csobj, Vcoding_system_hash_table); |
1151 | |
771 | 1152 if (need_to_setup_eol_systems && !cs->internal_p) |
1153 setup_eol_coding_systems (csobj); | |
1154 else if (eol_wrapper == EOL_CR || eol_wrapper == EOL_CRLF) | |
1155 { | |
1156 /* If a specific eol-type (other than LF) was specified, we handle | |
1157 this by converting the coding system into a chain that wraps the | |
1158 coding system along with a convert-eol system after it, in | |
1159 exactly that same switcheroo fashion that the normal | |
1160 canonicalize method works -- BUT we will run into a problem if | |
1161 we do it the obvious way, because when `chain' creates its | |
1162 substreams, the substream containing the coding system we're | |
1163 creating will have canonicalization expansion done on it, | |
1164 leading to infinite recursion. So we have to generate a new, | |
1165 internal coding system with the previous value of CANONICAL. */ | |
867 | 1166 Ibyte *newname = |
771 | 1167 emacs_sprintf_malloc |
1168 (NULL, "internal-eol-copy-%s-%d", | |
1169 XSTRING_DATA (Fsymbol_name (name_or_existing)), | |
1170 ++coding_system_tick); | |
1171 Lisp_Object newnamesym = intern_int (newname); | |
1172 Lisp_Object copied = Fcopy_coding_system (csobj, newnamesym); | |
1726 | 1173 xfree (newname, Ibyte *); |
771 | 1174 |
1175 XCODING_SYSTEM_CANONICAL (csobj) = | |
1176 make_internal_coding_system | |
1177 (csobj, | |
1178 "internal-eol-wrapper", | |
1179 Qchain, Qunbound, | |
1180 list4 (Qchain, | |
1181 list2 (copied, | |
1182 eol_wrapper == EOL_CR ? | |
1183 Qconvert_eol_cr : | |
1184 Qconvert_eol_crlf), | |
1185 Qcanonicalize_after_coding, | |
1186 csobj)); | |
1187 } | |
1188 XCODING_SYSTEM_EOL_TYPE (csobj) = eol_wrapper; | |
1189 } | |
1190 | |
1191 return csobj; | |
428 | 1192 } |
1193 | |
771 | 1194 Lisp_Object |
4528
726060ee587c
First draft of g++ 4.3 warning removal patch. Builds. *Needs ChangeLogs.*
Stephen J. Turnbull <stephen@xemacs.org>
parents:
4522
diff
changeset
|
1195 make_internal_coding_system (Lisp_Object existing, const Ascbyte *prefix, |
771 | 1196 Lisp_Object type, Lisp_Object description, |
1197 Lisp_Object props) | |
1198 { | |
1199 return make_coding_system_1 (existing, prefix, type, description, props); | |
1200 } | |
428 | 1201 |
1202 DEFUN ("make-coding-system", Fmake_coding_system, 2, 4, 0, /* | |
1203 Register symbol NAME as a coding system. | |
1204 | |
1205 TYPE describes the conversion method used and should be one of | |
1206 | |
3025 | 1207 nil or `undecided' |
428 | 1208 Automatic conversion. XEmacs attempts to detect the coding system |
1209 used in the file. | |
3025 | 1210 `chain' |
771 | 1211 Chain two or more coding systems together to make a combination coding |
1212 system. | |
3025 | 1213 `no-conversion' |
428 | 1214 No conversion. Use this for binary files and such. On output, |
1215 graphic characters that are not in ASCII or Latin-1 will be | |
1216 replaced by a ?. (For a no-conversion-encoded buffer, these | |
1217 characters will only be present if you explicitly insert them.) | |
3025 | 1218 `convert-eol' |
771 | 1219 Convert CRLF sequences or CR to LF. |
3025 | 1220 `shift-jis' |
428 | 1221 Shift-JIS (a Japanese encoding commonly used in PC operating systems). |
3025 | 1222 `unicode' |
771 | 1223 Any Unicode encoding (UCS-4, UTF-8, UTF-16, etc.). |
3025 | 1224 `mswindows-unicode-to-multibyte' |
771 | 1225 (MS Windows only) Converts from Windows Unicode to Windows Multibyte |
1226 (any code page encoding) upon encoding, and the other way upon decoding. | |
3025 | 1227 `mswindows-multibyte' |
771 | 1228 Converts to or from Windows Multibyte (any code page encoding). |
1229 This is resolved into a chain of `mswindows-unicode' and | |
1230 `mswindows-unicode-to-multibyte'. | |
3025 | 1231 `iso2022' |
428 | 1232 Any ISO2022-compliant encoding. Among other things, this includes |
1233 JIS (the Japanese encoding commonly used for e-mail), EUC (the | |
1234 standard Unix encoding for Japanese and other languages), and | |
1235 Compound Text (the encoding used in X11). You can specify more | |
442 | 1236 specific information about the conversion with the PROPS argument. |
3025 | 1237 `big5' |
2819 | 1238 Big5 (the encoding commonly used for Mandarin Chinese in Taiwan). |
3025 | 1239 `ccl' |
428 | 1240 The conversion is performed using a user-written pseudo-code |
1241 program. CCL (Code Conversion Language) is the name of this | |
1242 pseudo-code. | |
3025 | 1243 `gzip' |
771 | 1244 GZIP compression format. |
3025 | 1245 `internal' |
428 | 1246 Write out or read in the raw contents of the memory representing |
1247 the buffer's text. This is primarily useful for debugging | |
1248 purposes, and is only enabled when XEmacs has been compiled with | |
1249 DEBUG_XEMACS defined (via the --debug configure option). | |
3025 | 1250 WARNING: Reading in a file using `internal' conversion can result |
428 | 1251 in an internal inconsistency in the memory representing a |
1252 buffer's text, which will produce unpredictable results and may | |
1253 cause XEmacs to crash. Under normal circumstances you should | |
3025 | 1254 never use `internal' conversion. |
428 | 1255 |
771 | 1256 DESCRIPTION is a short English phrase describing the coding system, |
1257 suitable for use as a menu item. (See also the `documentation' property | |
1258 below.) | |
428 | 1259 |
1260 PROPS is a property list, describing the specific nature of the | |
1261 character set. Recognized properties are: | |
1262 | |
3025 | 1263 `mnemonic' |
428 | 1264 String to be displayed in the modeline when this coding system is |
1265 active. | |
1266 | |
3025 | 1267 `documentation' |
771 | 1268 Detailed documentation on the coding system. |
1269 | |
3025 | 1270 `eol-type' |
428 | 1271 End-of-line conversion to be used. It should be one of |
1272 | |
1273 nil | |
1274 Automatically detect the end-of-line type (LF, CRLF, | |
1275 or CR). Also generate subsidiary coding systems named | |
1276 `NAME-unix', `NAME-dos', and `NAME-mac', that are | |
1277 identical to this coding system but have an EOL-TYPE | |
3025 | 1278 value of `lf', `crlf', and `cr', respectively. |
1279 `lf' | |
428 | 1280 The end of a line is marked externally using ASCII LF. |
1281 Since this is also the way that XEmacs represents an | |
1282 end-of-line internally, specifying this option results | |
1283 in no end-of-line conversion. This is the standard | |
1284 format for Unix text files. | |
3025 | 1285 `crlf' |
428 | 1286 The end of a line is marked externally using ASCII |
1287 CRLF. This is the standard format for MS-DOS text | |
1288 files. | |
3025 | 1289 `cr' |
428 | 1290 The end of a line is marked externally using ASCII CR. |
1291 This is the standard format for Macintosh text files. | |
1292 t | |
1293 Automatically detect the end-of-line type but do not | |
1294 generate subsidiary coding systems. (This value is | |
1295 converted to nil when stored internally, and | |
1296 `coding-system-property' will return nil.) | |
1297 | |
3025 | 1298 `post-read-conversion' |
771 | 1299 The value is a function to call after some text is inserted and |
1300 decoded by the coding system itself and before any functions in | |
1301 `after-change-functions' are called. (#### Not actually true in | |
1302 XEmacs. `after-change-functions' will be called twice if | |
1303 `post-read-conversion' changes something.) The argument of this | |
1304 function is the same as for a function in | |
1305 `after-insert-file-functions', i.e. LENGTH of the text inserted, | |
1306 with point at the head of the text to be decoded. | |
428 | 1307 |
3025 | 1308 `pre-write-conversion' |
771 | 1309 The value is a function to call after all functions in |
1310 `write-region-annotate-functions' and `buffer-file-format' are | |
1311 called, and before the text is encoded by the coding system itself. | |
1312 The arguments to this function are the same as those of a function | |
1313 in `write-region-annotate-functions', i.e. FROM and TO, specifying | |
1314 a region of text. | |
1315 | |
1316 | |
1317 | |
1318 The following properties are allowed for FSF compatibility but currently | |
1319 ignored: | |
1320 | |
3025 | 1321 `translation-table-for-decode' |
771 | 1322 The value is a translation table to be applied on decoding. See |
1323 the function `make-translation-table' for the format of translation | |
1324 table. This is not applicable to CCL-based coding systems. | |
1325 | |
3025 | 1326 `translation-table-for-encode' |
771 | 1327 The value is a translation table to be applied on encoding. This is |
1328 not applicable to CCL-based coding systems. | |
1329 | |
3025 | 1330 `safe-chars' |
771 | 1331 The value is a char table. If a character has non-nil value in it, |
1332 the character is safely supported by the coding system. This | |
1333 overrides the specification of safe-charsets. | |
1334 | |
3025 | 1335 `safe-charsets' |
771 | 1336 The value is a list of charsets safely supported by the coding |
1337 system. The value t means that all charsets Emacs handles are | |
1338 supported. Even if some charset is not in this list, it doesn't | |
1339 mean that the charset can't be encoded in the coding system; | |
1340 it just means that some other receiver of text encoded | |
1341 in the coding system won't be able to handle that charset. | |
1342 | |
3025 | 1343 `mime-charset' |
771 | 1344 The value is a symbol of which name is `MIME-charset' parameter of |
1345 the coding system. | |
1346 | |
3025 | 1347 `valid-codes' (meaningful only for a coding system based on CCL) |
771 | 1348 The value is a list to indicate valid byte ranges of the encoded |
1349 file. Each element of the list is an integer or a cons of integer. | |
1350 In the former case, the integer value is a valid byte code. In the | |
1351 latter case, the integers specifies the range of valid byte codes. | |
1352 | |
1353 | |
1354 | |
3025 | 1355 The following additional property is recognized if TYPE is `convert-eol': |
1356 | |
1357 `subtype' | |
793 | 1358 One of `lf', `crlf', `cr' or nil (for autodetection). When decoding, |
1359 the corresponding sequence will be converted to LF. When encoding, | |
1360 the opposite happens. This coding system converts characters to | |
771 | 1361 characters. |
1362 | |
428 | 1363 |
1364 | |
3025 | 1365 The following additional properties are recognized if TYPE is `iso2022': |
1366 | |
1367 `charset-g0' | |
1368 `charset-g1' | |
1369 `charset-g2' | |
1370 `charset-g3' | |
428 | 1371 The character set initially designated to the G0 - G3 registers. |
1372 The value should be one of | |
1373 | |
1374 -- A charset object (designate that character set) | |
1375 -- nil (do not ever use this register) | |
1376 -- t (no character set is initially designated to | |
1377 the register, but may be later on; this automatically | |
1378 sets the corresponding `force-g*-on-output' property) | |
1379 | |
3025 | 1380 `force-g0-on-output' |
1381 `force-g1-on-output' | |
1382 `force-g2-on-output' | |
1383 `force-g2-on-output' | |
428 | 1384 If non-nil, send an explicit designation sequence on output before |
1385 using the specified register. | |
1386 | |
3025 | 1387 `short' |
428 | 1388 If non-nil, use the short forms "ESC $ @", "ESC $ A", and |
1389 "ESC $ B" on output in place of the full designation sequences | |
1390 "ESC $ ( @", "ESC $ ( A", and "ESC $ ( B". | |
1391 | |
3025 | 1392 `no-ascii-eol' |
428 | 1393 If non-nil, don't designate ASCII to G0 at each end of line on output. |
1394 Setting this to non-nil also suppresses other state-resetting that | |
1395 normally happens at the end of a line. | |
1396 | |
3025 | 1397 `no-ascii-cntl' |
428 | 1398 If non-nil, don't designate ASCII to G0 before control chars on output. |
1399 | |
3025 | 1400 `seven' |
428 | 1401 If non-nil, use 7-bit environment on output. Otherwise, use 8-bit |
1402 environment. | |
1403 | |
3025 | 1404 `lock-shift' |
428 | 1405 If non-nil, use locking-shift (SO/SI) instead of single-shift |
1406 or designation by escape sequence. | |
1407 | |
3025 | 1408 `no-iso6429' |
428 | 1409 If non-nil, don't use ISO6429's direction specification. |
1410 | |
3025 | 1411 `escape-quoted' |
428 | 1412 If non-nil, literal control characters that are the same as |
1413 the beginning of a recognized ISO2022 or ISO6429 escape sequence | |
1414 (in particular, ESC (0x1B), SO (0x0E), SI (0x0F), SS2 (0x8E), | |
1415 SS3 (0x8F), and CSI (0x9B)) are "quoted" with an escape character | |
1416 so that they can be properly distinguished from an escape sequence. | |
1417 (Note that doing this results in a non-portable encoding.) This | |
1418 encoding flag is used for byte-compiled files. Note that ESC | |
1419 is a good choice for a quoting character because there are no | |
1420 escape sequences whose second byte is a character from the Control-0 | |
1421 or Control-1 character sets; this is explicitly disallowed by the | |
1422 ISO2022 standard. | |
1423 | |
3025 | 1424 `input-charset-conversion' |
428 | 1425 A list of conversion specifications, specifying conversion of |
1426 characters in one charset to another when decoding is performed. | |
1427 Each specification is a list of two elements: the source charset, | |
1428 and the destination charset. | |
1429 | |
3025 | 1430 `output-charset-conversion' |
428 | 1431 A list of conversion specifications, specifying conversion of |
1432 characters in one charset to another when encoding is performed. | |
1433 The form of each specification is the same as for | |
3025 | 1434 `input-charset-conversion'. |
428 | 1435 |
1436 | |
771 | 1437 |
428 | 1438 The following additional properties are recognized (and required) |
3025 | 1439 if TYPE is `ccl': |
1440 | |
1441 `decode' | |
428 | 1442 CCL program used for decoding (converting to internal format). |
1443 | |
3025 | 1444 `encode' |
428 | 1445 CCL program used for encoding (converting to external format). |
771 | 1446 |
1447 | |
3025 | 1448 The following additional properties are recognized if TYPE is `chain': |
1449 | |
1450 `chain' | |
771 | 1451 List of coding systems to be chained together, in decoding order. |
1452 | |
3025 | 1453 `canonicalize-after-coding' |
771 | 1454 Coding system to be returned by the detector routines in place of |
1455 this coding system. | |
1456 | |
1457 | |
1458 | |
3025 | 1459 The following additional properties are recognized if TYPE is `unicode': |
1460 | |
3767 | 1461 `unicode-type' |
771 | 1462 One of `utf-16', `utf-8', `ucs-4', or `utf-7' (the latter is not |
1463 yet implemented). `utf-16' is the basic two-byte encoding; | |
1464 `ucs-4' is the four-byte encoding; `utf-8' is an ASCII-compatible | |
1465 variable-width 8-bit encoding; `utf-7' is a 7-bit encoding using | |
1466 only characters that will safely pass through all mail gateways. | |
2297 | 1467 [[ This should be \"transformation format\". There should also be |
1468 `ucs-2' (or `bmp' -- no surrogates) and `utf-32' (range checked). ]] | |
771 | 1469 |
3025 | 1470 `little-endian' |
771 | 1471 If non-nil, `utf-16' and `ucs-4' will write out the groups of two |
1472 or four bytes little-endian instead of big-endian. This is required, | |
1473 for example, under Windows. | |
1474 | |
3025 | 1475 `need-bom' |
771 | 1476 If non-nil, a byte order mark (BOM, or Unicode FFFE) should be |
1477 written out at the beginning of the data. This serves both to | |
1478 identify the endianness of the following data and to mark the | |
1479 data as Unicode (at least, this is how Windows uses it). | |
2297 | 1480 [[ The correct term is \"signature\", since this technique may also |
1481 be used with UTF-8. That is the term used in the standard. ]] | |
771 | 1482 |
1483 | |
1484 The following additional properties are recognized if TYPE is | |
3025 | 1485 `mswindows-multibyte': |
1486 | |
1487 `code-page' | |
771 | 1488 Either a number (specifying a particular code page) or one of the |
1489 symbols `ansi', `oem', `mac', or `ebcdic', specifying the ANSI, | |
1490 OEM, Macintosh, or EBCDIC code page associated with a particular | |
1491 locale (given by the `locale' property). NOTE: EBCDIC code pages | |
1492 only exist in Windows 2000 and later. | |
1493 | |
3025 | 1494 `locale' |
771 | 1495 If `code-page' is a symbol, this specifies the locale whose code |
1496 page of the corresponding type should be used. This should be | |
1497 one of the following: A cons of two strings, (LANGUAGE | |
1498 . SUBLANGUAGE) (see `mswindows-set-current-locale'); a string (a | |
1499 language; SUBLANG_DEFAULT, i.e. the default sublanguage, is | |
1500 used); or one of the symbols `current', `user-default', or | |
1501 `system-default', corresponding to the values of | |
1502 `mswindows-current-locale', `mswindows-user-default-locale', or | |
1503 `mswindows-system-default-locale', respectively. | |
1504 | |
1505 | |
1506 | |
3025 | 1507 The following additional properties are recognized if TYPE is `undecided': |
4072 | 1508 \[[ Doesn't GNU use \"detect-*\" for the following two? ]] |
771 | 1509 |
3025 | 1510 `do-eol' |
771 | 1511 Do EOL detection. |
1512 | |
3025 | 1513 `do-coding' |
771 | 1514 Do encoding detection. |
1515 | |
3025 | 1516 `coding-system' |
771 | 1517 If encoding detection is not done, use the specified coding system |
1518 to do decoding. This is used internally when implementing coding | |
1519 systems with an EOL type that specifies autodetection (the default), | |
1520 so that the detector routines return the proper subsidiary. | |
1521 | |
1522 | |
1523 | |
3025 | 1524 The following additional property is recognized if TYPE is `gzip': |
1525 | |
1526 `level' | |
771 | 1527 Compression level: 0 through 9, or `default' (currently 6). |
1528 | |
428 | 1529 */ |
771 | 1530 (name, type, description, props)) |
428 | 1531 { |
771 | 1532 return make_coding_system_1 (name, 0, type, description, props); |
428 | 1533 } |
1534 | |
1535 DEFUN ("copy-coding-system", Fcopy_coding_system, 2, 2, 0, /* | |
1536 Copy OLD-CODING-SYSTEM to NEW-NAME. | |
1537 If NEW-NAME does not name an existing coding system, a new one will | |
1538 be created. | |
771 | 1539 If you are using this function to create an alias, think again: |
1540 Use `define-coding-system-alias' instead. | |
428 | 1541 */ |
1542 (old_coding_system, new_name)) | |
1543 { | |
1544 Lisp_Object new_coding_system; | |
1545 old_coding_system = Fget_coding_system (old_coding_system); | |
771 | 1546 new_coding_system = |
4303 | 1547 UNBOUNDP (new_name) ? Qnil : find_coding_system (new_name, 0); |
428 | 1548 if (NILP (new_coding_system)) |
1549 { | |
793 | 1550 new_coding_system = |
1551 wrap_coding_system | |
1552 (allocate_coding_system | |
1553 (XCODING_SYSTEM (old_coding_system)->methods, | |
1554 XCODING_SYSTEM (old_coding_system)->methods->extra_data_size, | |
1555 new_name)); | |
771 | 1556 if (!UNBOUNDP (new_name)) |
1557 Fputhash (new_name, new_coding_system, Vcoding_system_hash_table); | |
428 | 1558 } |
771 | 1559 else if (XCODING_SYSTEM (old_coding_system)->methods != |
1560 XCODING_SYSTEM (new_coding_system)->methods) | |
1561 invalid_operation_2 ("Coding systems not same type", | |
1562 old_coding_system, new_coding_system); | |
428 | 1563 |
1564 { | |
1565 Lisp_Coding_System *to = XCODING_SYSTEM (new_coding_system); | |
1566 Lisp_Coding_System *from = XCODING_SYSTEM (old_coding_system); | |
3017 | 1567 COPY_SIZED_LCRECORD (to, from, sizeof_coding_system (from)); |
428 | 1568 to->name = new_name; |
1569 } | |
1570 return new_coding_system; | |
1571 } | |
1572 | |
771 | 1573 DEFUN ("coding-system-canonical-name-p", Fcoding_system_canonical_name_p, |
1574 1, 1, 0, /* | |
440 | 1575 Return t if OBJECT names a coding system, and is not a coding system alias. |
428 | 1576 */ |
440 | 1577 (object)) |
1578 { | |
1579 return CODING_SYSTEMP (Fgethash (object, Vcoding_system_hash_table, Qnil)) | |
1580 ? Qt : Qnil; | |
1581 } | |
1582 | |
2297 | 1583 /* #### Shouldn't this really be a find/get pair? */ |
1584 | |
440 | 1585 DEFUN ("coding-system-alias-p", Fcoding_system_alias_p, 1, 1, 0, /* |
1586 Return t if OBJECT is a coding system alias. | |
1587 All coding system aliases are created by `define-coding-system-alias'. | |
1588 */ | |
1589 (object)) | |
428 | 1590 { |
440 | 1591 return SYMBOLP (Fgethash (object, Vcoding_system_hash_table, Qzero)) |
1592 ? Qt : Qnil; | |
1593 } | |
1594 | |
1595 DEFUN ("coding-system-aliasee", Fcoding_system_aliasee, 1, 1, 0, /* | |
1596 Return the coding-system symbol for which symbol ALIAS is an alias. | |
1597 */ | |
1598 (alias)) | |
1599 { | |
1600 Lisp_Object aliasee = Fgethash (alias, Vcoding_system_hash_table, Qnil); | |
1601 if (SYMBOLP (aliasee)) | |
1602 return aliasee; | |
1603 else | |
563 | 1604 invalid_argument ("Symbol is not a coding system alias", alias); |
1204 | 1605 RETURN_NOT_REACHED (Qnil); |
440 | 1606 } |
1607 | |
1608 /* A maphash function, for removing dangling coding system aliases. */ | |
1609 static int | |
2286 | 1610 dangling_coding_system_alias_p (Lisp_Object UNUSED (alias), |
440 | 1611 Lisp_Object aliasee, |
1612 void *dangling_aliases) | |
1613 { | |
1614 if (SYMBOLP (aliasee) | |
1615 && NILP (Fgethash (aliasee, Vcoding_system_hash_table, Qnil))) | |
428 | 1616 { |
440 | 1617 (*(int *) dangling_aliases)++; |
1618 return 1; | |
428 | 1619 } |
440 | 1620 else |
1621 return 0; | |
1622 } | |
1623 | |
1624 DEFUN ("define-coding-system-alias", Fdefine_coding_system_alias, 2, 2, 0, /* | |
1625 Define symbol ALIAS as an alias for coding system ALIASEE. | |
1626 | |
1627 You can use this function to redefine an alias that has already been defined, | |
1628 but you cannot redefine a name which is the canonical name for a coding system. | |
1629 \(a canonical name of a coding system is what is returned when you call | |
1630 `coding-system-name' on a coding system). | |
1631 | |
1632 ALIASEE itself can be an alias, which allows you to define nested aliases. | |
1633 | |
1634 You are forbidden, however, from creating alias loops or `dangling' aliases. | |
1635 These will be detected, and an error will be signaled if you attempt to do so. | |
1636 | |
1637 If ALIASEE is nil, then ALIAS will simply be undefined. | |
1638 | |
1639 See also `coding-system-alias-p', `coding-system-aliasee', | |
1640 and `coding-system-canonical-name-p'. | |
1641 */ | |
1642 (alias, aliasee)) | |
1643 { | |
2286 | 1644 Lisp_Object probe; |
440 | 1645 |
1646 CHECK_SYMBOL (alias); | |
1647 | |
1648 if (!NILP (Fcoding_system_canonical_name_p (alias))) | |
563 | 1649 invalid_change |
440 | 1650 ("Symbol is the canonical name of a coding system and cannot be redefined", |
1651 alias); | |
1652 | |
1653 if (NILP (aliasee)) | |
1654 { | |
771 | 1655 Lisp_Object subsidiary_unix = add_suffix_to_symbol (alias, "-unix"); |
1656 Lisp_Object subsidiary_dos = add_suffix_to_symbol (alias, "-dos"); | |
1657 Lisp_Object subsidiary_mac = add_suffix_to_symbol (alias, "-mac"); | |
440 | 1658 |
1659 Fremhash (alias, Vcoding_system_hash_table); | |
1660 | |
1661 /* Undefine subsidiary aliases, | |
1662 presumably created by a previous call to this function */ | |
1663 if (! NILP (Fcoding_system_alias_p (subsidiary_unix)) && | |
1664 ! NILP (Fcoding_system_alias_p (subsidiary_dos)) && | |
1665 ! NILP (Fcoding_system_alias_p (subsidiary_mac))) | |
1666 { | |
1667 Fdefine_coding_system_alias (subsidiary_unix, Qnil); | |
1668 Fdefine_coding_system_alias (subsidiary_dos, Qnil); | |
1669 Fdefine_coding_system_alias (subsidiary_mac, Qnil); | |
1670 } | |
1671 | |
1672 /* Undefine dangling coding system aliases. */ | |
1673 { | |
1674 int dangling_aliases; | |
1675 | |
1676 do { | |
1677 dangling_aliases = 0; | |
1678 elisp_map_remhash (dangling_coding_system_alias_p, | |
1679 Vcoding_system_hash_table, | |
1680 &dangling_aliases); | |
1681 } while (dangling_aliases > 0); | |
1682 } | |
1683 | |
1684 return Qnil; | |
1685 } | |
1686 | |
1687 if (CODING_SYSTEMP (aliasee)) | |
1688 aliasee = XCODING_SYSTEM_NAME (aliasee); | |
1689 | |
1690 /* Checks that aliasee names a coding-system */ | |
2286 | 1691 (void) Fget_coding_system (aliasee); |
440 | 1692 |
1693 /* Check for coding system alias loops */ | |
1694 if (EQ (alias, aliasee)) | |
563 | 1695 alias_loop: invalid_operation_2 |
440 | 1696 ("Attempt to create a coding system alias loop", alias, aliasee); |
1697 | |
1698 for (probe = aliasee; | |
1699 SYMBOLP (probe); | |
1700 probe = Fgethash (probe, Vcoding_system_hash_table, Qzero)) | |
1701 { | |
1702 if (EQ (probe, alias)) | |
1703 goto alias_loop; | |
1704 } | |
1705 | |
1706 Fputhash (alias, aliasee, Vcoding_system_hash_table); | |
1707 | |
1708 /* Set up aliases for subsidiaries. | |
2297 | 1709 #### There must be a better way to handle subsidiary coding systems. |
1710 Inquiring Minds Want To Know: shouldn't they always be chains? */ | |
440 | 1711 { |
1712 static const char *suffixes[] = { "-unix", "-dos", "-mac" }; | |
1713 int i; | |
1714 for (i = 0; i < countof (suffixes); i++) | |
1715 { | |
1716 Lisp_Object alias_subsidiary = | |
771 | 1717 add_suffix_to_symbol (alias, suffixes[i]); |
440 | 1718 Lisp_Object aliasee_subsidiary = |
771 | 1719 add_suffix_to_symbol (aliasee, suffixes[i]); |
440 | 1720 |
1721 if (! NILP (Ffind_coding_system (aliasee_subsidiary))) | |
1722 Fdefine_coding_system_alias (alias_subsidiary, aliasee_subsidiary); | |
1723 } | |
1724 } | |
428 | 1725 /* FSF return value is a vector of [ALIAS-unix ALIAS-dos ALIAS-mac], |
1726 but it doesn't look intentional, so I'd rather return something | |
1727 meaningful or nothing at all. */ | |
1728 return Qnil; | |
1729 } | |
1730 | |
1731 static Lisp_Object | |
771 | 1732 subsidiary_coding_system (Lisp_Object coding_system, enum eol_type type) |
428 | 1733 { |
1734 Lisp_Coding_System *cs = XCODING_SYSTEM (coding_system); | |
1735 Lisp_Object new_coding_system; | |
1736 | |
1737 switch (type) | |
1738 { | |
1739 case EOL_AUTODETECT: return coding_system; | |
1740 case EOL_LF: new_coding_system = CODING_SYSTEM_EOL_LF (cs); break; | |
1741 case EOL_CR: new_coding_system = CODING_SYSTEM_EOL_CR (cs); break; | |
1742 case EOL_CRLF: new_coding_system = CODING_SYSTEM_EOL_CRLF (cs); break; | |
2500 | 1743 default: ABORT (); return Qnil; |
428 | 1744 } |
1745 | |
1746 return NILP (new_coding_system) ? coding_system : new_coding_system; | |
1747 } | |
1748 | |
1749 DEFUN ("subsidiary-coding-system", Fsubsidiary_coding_system, 2, 2, 0, /* | |
1750 Return the subsidiary coding system of CODING-SYSTEM with eol type EOL-TYPE. | |
771 | 1751 The logically opposite operation is `coding-system-base'. |
428 | 1752 */ |
1753 (coding_system, eol_type)) | |
1754 { | |
771 | 1755 coding_system = get_coding_system_for_text_file (coding_system, 0); |
428 | 1756 |
1757 return subsidiary_coding_system (coding_system, | |
1758 symbol_to_eol_type (eol_type)); | |
1759 } | |
1760 | |
771 | 1761 DEFUN ("coding-system-base", Fcoding_system_base, |
1762 1, 1, 0, /* | |
1763 Return the base coding system of CODING-SYSTEM. | |
1764 If CODING-SYSTEM is a subsidiary, this returns its parent; otherwise, it | |
1765 returns CODING-SYSTEM. | |
1766 The logically opposite operation is `subsidiary-coding-system'. | |
1767 */ | |
1768 (coding_system)) | |
1769 { | |
1770 Lisp_Object base; | |
1771 | |
1772 coding_system = Fget_coding_system (coding_system); | |
1773 if (EQ (XCODING_SYSTEM_NAME (coding_system), Qbinary)) | |
1774 return Fget_coding_system (Qraw_text); /* hack! */ | |
1775 base = XCODING_SYSTEM_SUBSIDIARY_PARENT (coding_system); | |
1776 if (!NILP (base)) | |
1777 return base; | |
1778 return coding_system; | |
1779 } | |
1780 | |
1781 DEFUN ("coding-system-used-for-io", Fcoding_system_used_for_io, | |
1782 1, 1, 0, /* | |
1783 Return the coding system actually used for I/O. | |
1784 In some cases (e.g. when a particular EOL type is specified) this won't be | |
2297 | 1785 the coding system itself. This can be useful when trying to determine |
1786 precisely how data was decoded. | |
771 | 1787 */ |
1788 (coding_system)) | |
1789 { | |
1790 Lisp_Object canon; | |
1791 | |
1792 coding_system = Fget_coding_system (coding_system); | |
1793 canon = XCODING_SYSTEM_CANONICAL (coding_system); | |
1794 if (!NILP (canon)) | |
1795 return canon; | |
1796 return coding_system; | |
1797 } | |
1798 | |
428 | 1799 |
1800 /************************************************************************/ | |
1801 /* Coding system accessors */ | |
1802 /************************************************************************/ | |
1803 | |
771 | 1804 DEFUN ("coding-system-description", Fcoding_system_description, 1, 1, 0, /* |
1805 Return the description for CODING-SYSTEM. | |
1806 The `description' of a coding system is a short English phrase giving the | |
1807 name rendered according to English punctuation rules, plus possibly some | |
1808 explanatory text (typically in the form of a parenthetical phrase). The | |
1809 description is intended to be short enough that it can appear as a menu item, | |
1810 and clear enough to be recognizable even to someone who is assumed to have | |
1811 some basic familiarity with different encodings but may not know all the | |
1812 technical names; thus, for `cn-gb-2312' is described as "Chinese EUC" and | |
1813 `hz-gb-2312' is described as "Hz/ZW (Chinese)", where the actual name of | |
1814 the encoding is given, followed by a note that this is a Chinese encoding, | |
1815 because the great majority of people encountering this would have no idea | |
1816 what it is, and giving the language indicates whether the encoding should | |
1817 just be ignored or (conceivably) investigated more thoroughly. | |
428 | 1818 */ |
1819 (coding_system)) | |
1820 { | |
1821 coding_system = Fget_coding_system (coding_system); | |
771 | 1822 return XCODING_SYSTEM_DESCRIPTION (coding_system); |
428 | 1823 } |
1824 | |
1825 DEFUN ("coding-system-type", Fcoding_system_type, 1, 1, 0, /* | |
1826 Return the type of CODING-SYSTEM. | |
1827 */ | |
1828 (coding_system)) | |
1829 { | |
771 | 1830 coding_system = Fget_coding_system (coding_system); |
1831 return XCODING_SYSTEM_TYPE (coding_system); | |
428 | 1832 } |
1833 | |
1834 DEFUN ("coding-system-property", Fcoding_system_property, 2, 2, 0, /* | |
1835 Return the PROP property of CODING-SYSTEM. | |
1836 */ | |
1837 (coding_system, prop)) | |
1838 { | |
1839 coding_system = Fget_coding_system (coding_system); | |
1840 CHECK_SYMBOL (prop); | |
1841 | |
1842 if (EQ (prop, Qname)) | |
1843 return XCODING_SYSTEM_NAME (coding_system); | |
1844 else if (EQ (prop, Qtype)) | |
1845 return Fcoding_system_type (coding_system); | |
771 | 1846 else if (EQ (prop, Qdescription)) |
1847 return XCODING_SYSTEM_DESCRIPTION (coding_system); | |
428 | 1848 else if (EQ (prop, Qmnemonic)) |
1849 return XCODING_SYSTEM_MNEMONIC (coding_system); | |
771 | 1850 else if (EQ (prop, Qdocumentation)) |
1851 return XCODING_SYSTEM_DOCUMENTATION (coding_system); | |
428 | 1852 else if (EQ (prop, Qeol_type)) |
771 | 1853 return eol_type_to_symbol (XCODING_SYSTEM_EOL_TYPE |
1854 (coding_system)); | |
428 | 1855 else if (EQ (prop, Qeol_lf)) |
1856 return XCODING_SYSTEM_EOL_LF (coding_system); | |
1857 else if (EQ (prop, Qeol_crlf)) | |
1858 return XCODING_SYSTEM_EOL_CRLF (coding_system); | |
1859 else if (EQ (prop, Qeol_cr)) | |
1860 return XCODING_SYSTEM_EOL_CR (coding_system); | |
1861 else if (EQ (prop, Qpost_read_conversion)) | |
1862 return XCODING_SYSTEM_POST_READ_CONVERSION (coding_system); | |
1863 else if (EQ (prop, Qpre_write_conversion)) | |
1864 return XCODING_SYSTEM_PRE_WRITE_CONVERSION (coding_system); | |
771 | 1865 else |
1866 { | |
1867 Lisp_Object value = CODESYSMETH_OR_GIVEN (XCODING_SYSTEM (coding_system), | |
1868 getprop, | |
1869 (coding_system, prop), | |
1870 Qunbound); | |
1871 if (UNBOUNDP (value)) | |
1872 invalid_constant ("Unrecognized property", prop); | |
1873 return value; | |
1874 } | |
1875 } | |
1876 | |
1877 | |
1878 /************************************************************************/ | |
1879 /* Coding stream functions */ | |
1880 /************************************************************************/ | |
1881 | |
1882 /* A coding stream is a stream used for encoding or decoding text. The | |
1883 coding-stream object keeps track of the actual coding system, the stream | |
1884 that is at the other end, and data that needs to be persistent across | |
1885 the lifetime of the stream. */ | |
1886 | |
1204 | 1887 extern const struct sized_memory_description chain_coding_stream_description; |
1888 extern const struct sized_memory_description undecided_coding_stream_description; | |
1889 | |
1890 static const struct memory_description coding_stream_data_description_1 []= { | |
2551 | 1891 { XD_BLOCK_PTR, chain_coding_system, 1, |
1892 { &chain_coding_stream_description } }, | |
1893 { XD_BLOCK_PTR, undecided_coding_system, 1, | |
1894 { &undecided_coding_stream_description } }, | |
1204 | 1895 { XD_END } |
1896 }; | |
1897 | |
1898 static const struct sized_memory_description coding_stream_data_description = { | |
1899 sizeof (void *), coding_stream_data_description_1 | |
1900 }; | |
1901 | |
1902 static const struct memory_description coding_lstream_description[] = { | |
1903 { XD_INT, offsetof (struct coding_stream, type) }, | |
1904 { XD_LISP_OBJECT, offsetof (struct coding_stream, orig_codesys) }, | |
1905 { XD_LISP_OBJECT, offsetof (struct coding_stream, codesys) }, | |
1906 { XD_LISP_OBJECT, offsetof (struct coding_stream, other_end) }, | |
1907 { XD_UNION, offsetof (struct coding_stream, data), | |
2551 | 1908 XD_INDIRECT (0, 0), { &coding_stream_data_description } }, |
1204 | 1909 { XD_END } |
1910 }; | |
1911 | |
1912 DEFINE_LSTREAM_IMPLEMENTATION_WITH_DATA ("coding", coding); | |
771 | 1913 |
1914 /* Encoding and decoding are parallel operations, so we create just one | |
1915 stream for both. "Decoding" may involve the extra step of autodetection | |
1916 of the data format, but that's only because of the conventional | |
1917 definition of decoding as converting from external- to | |
1918 internal-formatted data. | |
1919 | |
2297 | 1920 [[ REWRITE ME! ]] |
1921 | |
771 | 1922 #### We really need to abstract out the concept of "data formats" and |
1923 define "converters" that convert from and to specified formats, | |
1924 eliminating the idea of decoding and encoding. When specifying a | |
1925 conversion process, we need to give the data formats themselves, not the | |
1926 conversion processes -- e.g. a coding system called "Unicode->multibyte" | |
1927 converts in both directions, and we could auto-detect the format of data | |
1928 at either end. */ | |
1929 | |
1930 static Bytecount | |
1931 coding_reader (Lstream *stream, unsigned char *data, Bytecount size) | |
1932 { | |
1933 unsigned char *orig_data = data; | |
1934 Bytecount read_size; | |
1935 int error_occurred = 0; | |
1936 struct coding_stream *str = CODING_STREAM_DATA (stream); | |
1937 | |
1938 /* We need to interface to coding_{de,en}code_1(), which expects to take | |
1939 some amount of data and store the result into a Dynarr. We have | |
1940 coding_{de,en}code_1() store into c->runoff, and take data from there | |
1941 as necessary. */ | |
1942 | |
1943 /* We loop until we have enough data, reading chunks from the other | |
1944 end and converting it. */ | |
1945 while (1) | |
1946 { | |
1947 /* Take data from convert_to if we can. Make sure to take at | |
1948 most SIZE bytes, and delete the data from convert_to. */ | |
1949 if (Dynarr_length (str->convert_to) > 0) | |
1950 { | |
1951 Bytecount chunk = | |
1952 min (size, (Bytecount) Dynarr_length (str->convert_to)); | |
1953 memcpy (data, Dynarr_atp (str->convert_to, 0), chunk); | |
1954 Dynarr_delete_many (str->convert_to, 0, chunk); | |
1955 data += chunk; | |
1956 size -= chunk; | |
1957 } | |
1958 | |
1959 if (size == 0) | |
1960 break; /* No more room for data */ | |
1961 | |
1962 if (str->eof) | |
1963 break; | |
1964 | |
1965 { | |
1966 /* Exhausted convert_to, so get some more. Read into convert_from, | |
1967 after existing "rejected" data from the last conversion. */ | |
1968 Bytecount rejected = Dynarr_length (str->convert_from); | |
1969 /* #### 1024 is arbitrary; we really need to separate 0 from EOF, | |
1970 and when we get 0, keep taking more data until we don't get 0 -- | |
1971 we don't know how much data the conversion routine might need | |
2297 | 1972 before it can generate any data of its own (eg, bzip2). */ |
814 | 1973 Bytecount readmore = |
1974 str->one_byte_at_a_time ? (Bytecount) 1 : | |
1975 max (size, (Bytecount) 1024); | |
771 | 1976 |
1977 Dynarr_add_many (str->convert_from, 0, readmore); | |
1978 read_size = Lstream_read (str->other_end, | |
1979 Dynarr_atp (str->convert_from, rejected), | |
1980 readmore); | |
1981 /* Trim size down to how much we actually got */ | |
1982 Dynarr_set_size (str->convert_from, rejected + max (0, read_size)); | |
1983 } | |
1984 | |
1985 if (read_size < 0) /* LSTREAM_ERROR */ | |
1986 { | |
1987 error_occurred = 1; | |
1988 break; | |
1989 } | |
1990 if (read_size == 0) /* LSTREAM_EOF */ | |
1991 /* There might be some more end data produced in the translation, | |
1992 so we set a flag and call the conversion method once more to | |
1993 output any final stuff it may be holding, any "go back to a sane | |
1994 state" escape sequences, etc. The conversion method is free to | |
1995 look at this flag, and we use it above to stop looping. */ | |
1996 str->eof = 1; | |
1997 { | |
1998 Bytecount processed; | |
1999 Bytecount to_process = Dynarr_length (str->convert_from); | |
2000 | |
2001 /* Convert the data, and save any rejected data in convert_from */ | |
2002 processed = | |
2003 XCODESYSMETH (str->codesys, convert, | |
2004 (str, Dynarr_atp (str->convert_from, 0), | |
2005 str->convert_to, to_process)); | |
2006 if (processed < 0) | |
2007 { | |
2008 error_occurred = 1; | |
2009 break; | |
2010 } | |
2011 assert (processed <= to_process); | |
2012 if (processed < to_process) | |
2013 memmove (Dynarr_atp (str->convert_from, 0), | |
2014 Dynarr_atp (str->convert_from, processed), | |
2015 to_process - processed); | |
2016 Dynarr_set_size (str->convert_from, to_process - processed); | |
2017 } | |
2018 } | |
2019 | |
2020 if (data - orig_data == 0) | |
2021 return error_occurred ? -1 : 0; | |
2022 else | |
2023 return data - orig_data; | |
2024 } | |
2025 | |
2026 static Bytecount | |
2027 coding_writer (Lstream *stream, const unsigned char *data, Bytecount size) | |
2028 { | |
2029 struct coding_stream *str = CODING_STREAM_DATA (stream); | |
2030 | |
2031 /* Convert all our data into convert_to, and then attempt to write | |
2032 it all out to the other end. */ | |
2033 Dynarr_reset (str->convert_to); | |
2034 size = XCODESYSMETH (str->codesys, convert, | |
2035 (str, data, str->convert_to, size)); | |
2036 if (Lstream_write (str->other_end, Dynarr_atp (str->convert_to, 0), | |
2037 Dynarr_length (str->convert_to)) < 0) | |
2038 return -1; | |
2039 else | |
2040 /* The return value indicates how much of the incoming data was | |
2041 processed, not how many bytes were written. */ | |
2042 return size; | |
2043 } | |
2044 | |
2045 static int | |
2046 encode_decode_source_sink_type_is_char (Lisp_Object cs, | |
2047 enum source_or_sink sex, | |
2048 enum encode_decode direction) | |
2049 { | |
2050 return (direction == CODING_DECODE ? | |
2051 decoding_source_sink_type_is_char (cs, sex) : | |
2052 encoding_source_sink_type_is_char (cs, sex)); | |
2053 } | |
2054 | |
2055 /* Ensure that the convert methods only get full characters sent to them to | |
2056 convert if the source of that conversion is characters; and that no such | |
2057 full-character checking happens when the source is bytes. Keep in mind | |
2058 that (1) the conversion_end_type return values take the perspective of | |
2059 encoding; (2) the source for decoding is the same as the sink for | |
2060 encoding; (3) when writing, the data is given to us, and we set our own | |
2061 stream to be character mode or not; (4) when reading, the data comes | |
2062 from the other_end stream, and we set that one to be character mode or | |
2063 not. This is consistent with the comment above the prototype for | |
2064 Lstream_set_character_mode(), which lays out rules for who is allowed to | |
2065 modify the character type mode on a stream. | |
2066 | |
814 | 2067 If we're a read stream, we're always setting character mode on the |
2068 source, but we also set it on ourselves consistent with the flag that | |
2069 can disable this (see again the comment above | |
2070 Lstream_set_character_mode()). | |
2071 */ | |
771 | 2072 |
2073 static void | |
2074 set_coding_character_mode (Lstream *stream) | |
2075 { | |
2076 struct coding_stream *str = CODING_STREAM_DATA (stream); | |
2077 Lstream *stream_to_set = | |
2078 stream->flags & LSTREAM_FL_WRITE ? stream : str->other_end; | |
2079 if (encode_decode_source_sink_type_is_char | |
2080 (str->codesys, CODING_SOURCE, str->direction)) | |
2081 Lstream_set_character_mode (stream_to_set); | |
2082 else | |
2083 Lstream_unset_character_mode (stream_to_set); | |
814 | 2084 if (str->set_char_mode_on_us_when_reading && |
2085 (stream->flags & LSTREAM_FL_READ)) | |
2086 { | |
2087 if (encode_decode_source_sink_type_is_char | |
2088 (str->codesys, CODING_SINK, str->direction)) | |
2089 Lstream_set_character_mode (stream); | |
2090 else | |
2091 Lstream_unset_character_mode (stream); | |
2092 } | |
771 | 2093 } |
2094 | |
2095 static Lisp_Object | |
2096 coding_marker (Lisp_Object stream) | |
2097 { | |
2098 struct coding_stream *str = CODING_STREAM_DATA (XLSTREAM (stream)); | |
2099 | |
2100 mark_object (str->orig_codesys); | |
2101 mark_object (str->codesys); | |
2102 MAYBE_XCODESYSMETH (str->codesys, mark_coding_stream, (str)); | |
2103 return wrap_lstream (str->other_end); | |
2104 } | |
2105 | |
2106 static int | |
2107 coding_rewinder (Lstream *stream) | |
2108 { | |
2109 struct coding_stream *str = CODING_STREAM_DATA (stream); | |
2110 MAYBE_XCODESYSMETH (str->codesys, rewind_coding_stream, (str)); | |
2111 | |
2112 str->ch = 0; | |
2113 Dynarr_reset (str->convert_to); | |
2114 Dynarr_reset (str->convert_from); | |
2115 return Lstream_rewind (str->other_end); | |
2116 } | |
2117 | |
2118 static int | |
2119 coding_seekable_p (Lstream *stream) | |
2120 { | |
2121 struct coding_stream *str = CODING_STREAM_DATA (stream); | |
2122 return Lstream_seekable_p (str->other_end); | |
2123 } | |
2124 | |
2125 static int | |
2126 coding_flusher (Lstream *stream) | |
2127 { | |
2128 struct coding_stream *str = CODING_STREAM_DATA (stream); | |
2129 return Lstream_flush (str->other_end); | |
2130 } | |
2131 | |
2132 static int | |
2133 coding_closer (Lstream *stream) | |
2134 { | |
2135 struct coding_stream *str = CODING_STREAM_DATA (stream); | |
2136 if (stream->flags & LSTREAM_FL_WRITE) | |
2137 { | |
2138 str->eof = 1; | |
2139 coding_writer (stream, 0, 0); | |
2140 str->eof = 0; | |
2141 } | |
2142 /* It's safe to free the runoff dynarrs now because they are used only | |
2143 during conversion. We need to keep the type-specific data around, | |
2144 though, because of canonicalize_after_coding. */ | |
2145 if (str->convert_to) | |
2146 { | |
2147 Dynarr_free (str->convert_to); | |
2148 str->convert_to = 0; | |
2149 } | |
2150 if (str->convert_from) | |
428 | 2151 { |
771 | 2152 Dynarr_free (str->convert_from); |
2153 str->convert_from = 0; | |
2154 } | |
2155 | |
800 | 2156 if (str->no_close_other) |
2157 return Lstream_flush (str->other_end); | |
2158 else | |
2159 return Lstream_close (str->other_end); | |
771 | 2160 } |
2161 | |
2162 static void | |
2163 coding_finalizer (Lstream *stream) | |
2164 { | |
2165 struct coding_stream *str = CODING_STREAM_DATA (stream); | |
2166 | |
2167 assert (!str->finalized); | |
2168 MAYBE_XCODESYSMETH (str->codesys, finalize_coding_stream, (str)); | |
2169 if (str->data) | |
2170 { | |
1726 | 2171 xfree (str->data, void *); |
771 | 2172 str->data = 0; |
2173 } | |
2174 str->finalized = 1; | |
2175 } | |
2176 | |
2177 static Lisp_Object | |
2178 coding_stream_canonicalize_after_coding (Lstream *stream) | |
2179 { | |
2180 struct coding_stream *str = CODING_STREAM_DATA (stream); | |
2181 | |
2182 return XCODESYSMETH_OR_GIVEN (str->codesys, canonicalize_after_coding, | |
2183 (str), str->codesys); | |
2184 } | |
2185 | |
2186 Lisp_Object | |
2187 coding_stream_detected_coding_system (Lstream *stream) | |
2188 { | |
2189 Lisp_Object codesys = | |
2190 coding_stream_canonicalize_after_coding (stream); | |
2191 if (NILP (codesys)) | |
2192 return Fget_coding_system (Qidentity); | |
2193 return codesys; | |
2194 } | |
2195 | |
2196 Lisp_Object | |
2197 coding_stream_coding_system (Lstream *stream) | |
2198 { | |
2199 return CODING_STREAM_DATA (stream)->codesys; | |
2200 } | |
2201 | |
2202 /* Change the coding system associated with a stream. */ | |
2203 | |
2204 void | |
2205 set_coding_stream_coding_system (Lstream *lstr, Lisp_Object codesys) | |
2206 { | |
2207 struct coding_stream *str = CODING_STREAM_DATA (lstr); | |
2208 if (EQ (str->orig_codesys, codesys)) | |
2209 return; | |
2210 /* We do the equivalent of closing the stream, destroying it, and | |
2211 reinitializing it. This includes flushing out the data and signalling | |
2212 EOF, if we're a writing stream; we also replace the type-specific data | |
2213 with the data appropriate for the new coding system. */ | |
2214 if (!NILP (str->codesys)) | |
2215 { | |
2216 if (lstr->flags & LSTREAM_FL_WRITE) | |
2217 { | |
2218 Lstream_flush (lstr); | |
2219 str->eof = 1; | |
2220 coding_writer (lstr, 0, 0); | |
2221 str->eof = 0; | |
2222 } | |
2223 MAYBE_XCODESYSMETH (str->codesys, finalize_coding_stream, (str)); | |
2224 } | |
2225 str->orig_codesys = codesys; | |
2226 str->codesys = coding_system_real_canonical (codesys); | |
2227 | |
2228 if (str->data) | |
2229 { | |
1726 | 2230 xfree (str->data, void *); |
771 | 2231 str->data = 0; |
428 | 2232 } |
771 | 2233 if (XCODING_SYSTEM_METHODS (str->codesys)->coding_data_size) |
1204 | 2234 { |
2235 str->data = | |
2236 xmalloc_and_zero (XCODING_SYSTEM_METHODS (str->codesys)-> | |
2237 coding_data_size); | |
2238 str->type = XCODING_SYSTEM_METHODS (str->codesys)->enumtype; | |
2239 } | |
771 | 2240 MAYBE_XCODESYSMETH (str->codesys, init_coding_stream, (str)); |
2241 /* The new coding system may have different ideas regarding whether its | |
2242 ends are characters or bytes. */ | |
2243 set_coding_character_mode (lstr); | |
2244 } | |
2245 | |
2246 /* WARNING WARNING WARNING WARNING!!!!! If you open up a coding | |
2247 stream for writing, no automatic code detection will be performed. | |
2248 The reason for this is that automatic code detection requires a | |
2249 seekable input. Things will also fail if you open a coding | |
2250 stream for reading using a non-fully-specified coding system and | |
2251 a non-seekable input stream. */ | |
2252 | |
2253 static Lisp_Object | |
2254 make_coding_stream_1 (Lstream *stream, Lisp_Object codesys, | |
800 | 2255 const char *mode, enum encode_decode direction, |
802 | 2256 int flags) |
771 | 2257 { |
2258 Lstream *lstr = Lstream_new (lstream_coding, mode); | |
2259 struct coding_stream *str = CODING_STREAM_DATA (lstr); | |
2260 | |
2261 codesys = Fget_coding_system (codesys); | |
2262 xzero (*str); | |
2263 str->codesys = Qnil; | |
2264 str->orig_codesys = Qnil; | |
2265 str->us = lstr; | |
2266 str->other_end = stream; | |
2267 str->convert_to = Dynarr_new (unsigned_char); | |
2268 str->convert_from = Dynarr_new (unsigned_char); | |
2269 str->direction = direction; | |
814 | 2270 if (flags & LSTREAM_FL_NO_CLOSE_OTHER) |
802 | 2271 str->no_close_other = 1; |
814 | 2272 if (flags & LSTREAM_FL_READ_ONE_BYTE_AT_A_TIME) |
802 | 2273 str->one_byte_at_a_time = 1; |
814 | 2274 if (!(flags & LSTREAM_FL_NO_INIT_CHAR_MODE_WHEN_READING)) |
2275 str->set_char_mode_on_us_when_reading = 1; | |
802 | 2276 |
771 | 2277 set_coding_stream_coding_system (lstr, codesys); |
793 | 2278 return wrap_lstream (lstr); |
771 | 2279 } |
2280 | |
814 | 2281 /* FLAGS: |
2282 | |
2283 LSTREAM_FL_NO_CLOSE_OTHER | |
2284 Don't close STREAM (the stream at the other end) when this stream is | |
2285 closed. | |
2286 | |
2287 LSTREAM_FL_READ_ONE_BYTE_AT_A_TIME | |
2288 When reading from STREAM, read and process one byte at a time rather | |
2289 than in large chunks. This is for reading from TTY's, so we don't | |
2290 block. #### We should instead create a non-blocking filedesc stream | |
2291 that emulates the behavior as necessary using select(), when the | |
2292 fcntls don't work. (As seems to be the case on Cygwin.) | |
2293 | |
2294 LSTREAM_FL_NO_INIT_CHAR_MODE_WHEN_READING | |
2295 When reading from STREAM, read and process one byte at a time rather | |
2296 than in large chunks. This is for reading from TTY's, so we don't | |
2297 block. #### We should instead create a non-blocking filedesc stream | |
2298 that emulates the behavior as necessary using select(), when the | |
2299 fcntls don't work. (As seems to be the case on Cygwin.) | |
2300 */ | |
771 | 2301 Lisp_Object |
2302 make_coding_input_stream (Lstream *stream, Lisp_Object codesys, | |
802 | 2303 enum encode_decode direction, int flags) |
771 | 2304 { |
800 | 2305 return make_coding_stream_1 (stream, codesys, "r", direction, |
802 | 2306 flags); |
771 | 2307 } |
2308 | |
814 | 2309 /* FLAGS: |
2310 | |
2311 LSTREAM_FL_NO_CLOSE_OTHER | |
2312 Don't close STREAM (the stream at the other end) when this stream is | |
2313 closed. | |
2314 */ | |
771 | 2315 Lisp_Object |
2316 make_coding_output_stream (Lstream *stream, Lisp_Object codesys, | |
802 | 2317 enum encode_decode direction, int flags) |
771 | 2318 { |
800 | 2319 return make_coding_stream_1 (stream, codesys, "w", direction, |
802 | 2320 flags); |
771 | 2321 } |
2322 | |
2323 static Lisp_Object | |
2324 encode_decode_coding_region (Lisp_Object start, Lisp_Object end, | |
2325 Lisp_Object coding_system, Lisp_Object buffer, | |
2326 enum encode_decode direction) | |
2327 { | |
2328 Charbpos b, e; | |
2329 struct buffer *buf = decode_buffer (buffer, 0); | |
2330 Lisp_Object instream = Qnil, to_outstream = Qnil, outstream = Qnil; | |
2331 Lisp_Object from_outstream = Qnil, auto_outstream = Qnil; | |
2332 Lisp_Object lb_outstream = Qnil; | |
2333 Lisp_Object next; | |
2334 Lstream *istr, *ostr; | |
2335 struct gcpro gcpro1, gcpro2, gcpro3, gcpro4, gcpro5; | |
2336 struct gcpro ngcpro1; | |
2337 int source_char, sink_char; | |
2338 | |
2339 get_buffer_range_char (buf, start, end, &b, &e, 0); | |
2340 barf_if_buffer_read_only (buf, b, e); | |
2341 | |
2342 GCPRO5 (instream, to_outstream, outstream, from_outstream, lb_outstream); | |
2343 NGCPRO1 (auto_outstream); | |
2344 | |
2345 coding_system = Fget_coding_system (coding_system); | |
2346 source_char = encode_decode_source_sink_type_is_char (coding_system, | |
2347 CODING_SOURCE, | |
2348 direction); | |
2349 sink_char = encode_decode_source_sink_type_is_char (coding_system, | |
2350 CODING_SINK, | |
2351 direction); | |
2352 | |
2353 /* Order is IN <---> [TO] -> OUT -> [FROM] -> [AUTODETECT-EOL] -> LB */ | |
2354 instream = make_lisp_buffer_input_stream (buf, b, e, 0); | |
2355 next = lb_outstream = make_lisp_buffer_output_stream (buf, b, 0); | |
2356 | |
2357 if (direction == CODING_DECODE && | |
2358 XCODING_SYSTEM_EOL_TYPE (coding_system) == EOL_AUTODETECT) | |
2359 next = auto_outstream = | |
2360 make_coding_output_stream | |
800 | 2361 (XLSTREAM (next), Fget_coding_system (Qconvert_eol_autodetect), |
2362 CODING_DECODE, 0); | |
771 | 2363 |
2364 if (!sink_char) | |
2365 next = from_outstream = | |
800 | 2366 make_coding_output_stream (XLSTREAM (next), Qbinary, CODING_DECODE, 0); |
771 | 2367 outstream = make_coding_output_stream (XLSTREAM (next), coding_system, |
800 | 2368 direction, 0); |
771 | 2369 if (!source_char) |
428 | 2370 { |
771 | 2371 to_outstream = |
2372 make_coding_output_stream (XLSTREAM (outstream), | |
800 | 2373 Qbinary, CODING_ENCODE, 0); |
771 | 2374 ostr = XLSTREAM (to_outstream); |
2375 } | |
2376 else | |
2377 ostr = XLSTREAM (outstream); | |
2378 istr = XLSTREAM (instream); | |
2379 | |
2380 /* The chain of streams looks like this: | |
2381 | |
2297 | 2382 [BUFFER] <----- (( read from/send to loop )) |
771 | 2383 ------> [CHAR->BYTE i.e. ENCODE AS BINARY if source is |
2384 in bytes] | |
2385 ------> [ENCODE/DECODE AS SPECIFIED] | |
2386 ------> [BYTE->CHAR i.e. DECODE AS BINARY | |
2387 if sink is in bytes] | |
2388 ------> [AUTODETECT EOL if | |
2389 we're decoding and | |
2390 coding system calls | |
2391 for this] | |
2392 ------> [BUFFER] | |
2393 */ | |
2367 | 2394 |
2395 /* #### See comment | |
2396 | |
2397 EFFICIENCY OF CODING CONVERSION WITH MULTIPLE COPIES/CHAINS | |
2398 | |
2399 in text.c. | |
2400 */ | |
2401 | |
771 | 2402 while (1) |
2403 { | |
2404 char tempbuf[1024]; /* some random amount */ | |
2405 Charbpos newpos, even_newer_pos; | |
2406 Charbpos oldpos = lisp_buffer_stream_startpos (istr); | |
2407 Bytecount size_in_bytes = | |
2408 Lstream_read (istr, tempbuf, sizeof (tempbuf)); | |
2409 | |
2410 if (!size_in_bytes) | |
2411 break; | |
2412 newpos = lisp_buffer_stream_startpos (istr); | |
2413 Lstream_write (ostr, tempbuf, size_in_bytes); | |
2414 even_newer_pos = lisp_buffer_stream_startpos (istr); | |
2415 buffer_delete_range (buf, even_newer_pos - (newpos - oldpos), | |
2416 even_newer_pos, 0); | |
428 | 2417 } |
771 | 2418 |
2419 { | |
2420 Charcount retlen = | |
2421 lisp_buffer_stream_startpos (XLSTREAM (instream)) - b; | |
2422 Lstream_close (istr); | |
2423 Lstream_close (ostr); | |
2424 NUNGCPRO; | |
2425 UNGCPRO; | |
2426 Lstream_delete (istr); | |
2427 if (!NILP (from_outstream)) | |
2428 Lstream_delete (XLSTREAM (from_outstream)); | |
2429 Lstream_delete (XLSTREAM (outstream)); | |
2430 if (!NILP (to_outstream)) | |
2431 Lstream_delete (XLSTREAM (to_outstream)); | |
2432 if (!NILP (auto_outstream)) | |
2433 Lstream_delete (XLSTREAM (auto_outstream)); | |
2434 Lstream_delete (XLSTREAM (lb_outstream)); | |
2435 return make_int (retlen); | |
2436 } | |
2437 } | |
2438 | |
3302 | 2439 DEFUN ("decode-coding-region", Fdecode_coding_region, 3, 4, |
2440 "*r\nzDecode from coding system: \ni", /* | |
771 | 2441 Decode the text between START and END which is encoded in CODING-SYSTEM. |
2442 This is useful if you've read in encoded text from a file without decoding | |
2443 it (e.g. you read in a JIS-formatted file but used the `binary' or | |
2444 `no-conversion' coding system, so that it shows up as "^[$B!<!+^[(B"). | |
2445 Return length of decoded text. | |
3302 | 2446 BUFFER defaults to the current buffer if unspecified, and when interactive. |
771 | 2447 */ |
2448 (start, end, coding_system, buffer)) | |
2449 { | |
2450 return encode_decode_coding_region (start, end, coding_system, buffer, | |
2451 CODING_DECODE); | |
2452 } | |
2453 | |
3302 | 2454 DEFUN ("encode-coding-region", Fencode_coding_region, 3, 4, |
2455 "*r\nzEncode to coding system: \ni", /* | |
771 | 2456 Encode the text between START and END using CODING-SYSTEM. |
2457 This will, for example, convert Japanese characters into stuff such as | |
3302 | 2458 "^[$B!<!+^[(B" if you use the JIS encoding. Return length of encoded text. |
2459 BUFFER defaults to the current buffer if unspecified, and when interactive. | |
771 | 2460 */ |
2461 (start, end, coding_system, buffer)) | |
2462 { | |
2463 return encode_decode_coding_region (start, end, coding_system, buffer, | |
2464 CODING_ENCODE); | |
428 | 2465 } |
2466 | |
2467 | |
2468 /************************************************************************/ | |
771 | 2469 /* Chain methods */ |
428 | 2470 /************************************************************************/ |
2471 | |
771 | 2472 /* #### Need a way to create "opposite-direction" coding systems. */ |
2473 | |
2474 /* Chain two or more coding systems together to make a combination coding | |
2475 system. */ | |
2476 | |
2477 struct chain_coding_system | |
2478 { | |
2479 /* List of coding systems, in decode order */ | |
2480 Lisp_Object *chain; | |
2481 /* Number of coding systems in list */ | |
2482 int count; | |
2483 /* Coding system to return as a result of canonicalize-after-coding */ | |
2484 Lisp_Object canonicalize_after_coding; | |
2485 }; | |
2486 | |
2487 struct chain_coding_stream | |
2488 { | |
2489 int initted; | |
2490 /* Lstreams for chain coding system */ | |
2491 Lisp_Object *lstreams; | |
2492 int lstream_count; | |
2493 }; | |
2494 | |
1204 | 2495 static const struct memory_description chain_coding_system_description[] = { |
2496 { XD_INT, offsetof (struct chain_coding_system, count) }, | |
2367 | 2497 { XD_BLOCK_PTR, offsetof (struct chain_coding_system, chain), |
2551 | 2498 XD_INDIRECT (0, 0), { &lisp_object_description } }, |
1204 | 2499 { XD_LISP_OBJECT, offsetof (struct chain_coding_system, |
2500 canonicalize_after_coding) }, | |
771 | 2501 { XD_END } |
2502 }; | |
2503 | |
1204 | 2504 static const struct memory_description chain_coding_stream_description_1 [] = { |
2505 { XD_INT, offsetof (struct chain_coding_stream, lstream_count) }, | |
2367 | 2506 { XD_BLOCK_PTR, offsetof (struct chain_coding_stream, lstreams), |
2551 | 2507 XD_INDIRECT (0, 0), { &lisp_object_description } }, |
771 | 2508 { XD_END } |
2509 }; | |
2510 | |
1204 | 2511 const struct sized_memory_description chain_coding_stream_description = { |
2512 sizeof (struct chain_coding_stream), chain_coding_stream_description_1 | |
2513 }; | |
2514 | |
2515 DEFINE_CODING_SYSTEM_TYPE_WITH_DATA (chain); | |
2516 | |
771 | 2517 static Lisp_Object |
2518 chain_canonicalize (Lisp_Object codesys) | |
2519 { | |
2520 /* We make use of the fact that this method is called at init time, after | |
2521 properties have been parsed. init_method is called too early. */ | |
2522 /* #### It's not clear we need this whole chain-canonicalize mechanism | |
2523 any more. */ | |
2524 Lisp_Object chain = Flist (XCODING_SYSTEM_CHAIN_COUNT (codesys), | |
2525 XCODING_SYSTEM_CHAIN_CHAIN (codesys)); | |
2526 chain = Fcons (XCODING_SYSTEM_PRE_WRITE_CONVERSION (codesys), | |
2527 Fcons (XCODING_SYSTEM_POST_READ_CONVERSION (codesys), | |
2528 chain)); | |
2529 Fputhash (chain, codesys, Vchain_canonicalize_hash_table); | |
2530 return codesys; | |
2531 } | |
2532 | |
2533 static Lisp_Object | |
2534 chain_canonicalize_after_coding (struct coding_stream *str) | |
2535 { | |
2536 Lisp_Object cac = | |
2537 XCODING_SYSTEM_CHAIN_CANONICALIZE_AFTER_CODING (str->codesys); | |
2538 if (!NILP (cac)) | |
2539 return cac; | |
2540 return str->codesys; | |
2541 #if 0 | |
2542 struct chain_coding_stream *data = CODING_STREAM_TYPE_DATA (str, chain); | |
2543 Lisp_Object us = str->codesys, codesys; | |
2544 int i; | |
2545 Lisp_Object chain; | |
2546 Lisp_Object tail; | |
2547 int changed = 0; | |
2548 | |
2549 /* #### It's not clear we need this whole chain-canonicalize mechanism | |
2550 any more. */ | |
2551 if (str->direction == CODING_ENCODE || !data->initted) | |
2552 return us; | |
2553 | |
2554 chain = Flist (XCODING_SYSTEM_CHAIN_COUNT (us), | |
2555 XCODING_SYSTEM_CHAIN_CHAIN (us)); | |
2556 | |
2557 tail = chain; | |
2558 for (i = 0; i < XCODING_SYSTEM_CHAIN_COUNT (us); i++) | |
2559 { | |
2560 codesys = (coding_stream_canonicalize_after_coding | |
2561 (XLSTREAM (data->lstreams[i]))); | |
2562 if (!EQ (codesys, XCAR (tail))) | |
2563 changed = 1; | |
2564 XCAR (tail) = codesys; | |
2565 tail = XCDR (tail); | |
2566 } | |
2567 | |
2568 if (!changed) | |
2569 return us; | |
2570 | |
2571 chain = delq_no_quit (Qnil, chain); | |
2572 | |
2573 if (NILP (XCODING_SYSTEM_PRE_WRITE_CONVERSION (us)) && | |
2574 NILP (XCODING_SYSTEM_POST_READ_CONVERSION (us))) | |
2575 { | |
2576 if (NILP (chain)) | |
2577 return Qnil; | |
2578 if (NILP (XCDR (chain))) | |
2579 return XCAR (chain); | |
2580 } | |
2581 | |
2582 codesys = Fgethash (Fcons (XCODING_SYSTEM_PRE_WRITE_CONVERSION (us), | |
2583 Fcons (XCODING_SYSTEM_POST_READ_CONVERSION (us), | |
2584 chain)), Vchain_canonicalize_hash_table, | |
2585 Qnil); | |
2586 if (!NILP (codesys)) | |
2587 return codesys; | |
2588 return make_internal_coding_system | |
2589 (us, "internal-chain-canonicalizer-wrapper", | |
2590 Qchain, Qunbound, list2 (Qchain, chain)); | |
2591 #endif /* 0 */ | |
2592 } | |
2593 | |
2594 static void | |
2595 chain_init (Lisp_Object codesys) | |
2596 { | |
2597 XCODING_SYSTEM_CHAIN_CANONICALIZE_AFTER_CODING (codesys) = Qnil; | |
2598 } | |
2599 | |
2600 static void | |
2601 chain_mark (Lisp_Object codesys) | |
2602 { | |
2603 int i; | |
2604 | |
2605 for (i = 0; i < XCODING_SYSTEM_CHAIN_COUNT (codesys); i++) | |
2606 mark_object (XCODING_SYSTEM_CHAIN_CHAIN (codesys)[i]); | |
2607 mark_object (XCODING_SYSTEM_CHAIN_CANONICALIZE_AFTER_CODING (codesys)); | |
2608 } | |
2609 | |
2610 static void | |
2611 chain_mark_coding_stream_1 (struct chain_coding_stream *data) | |
2612 { | |
2613 int i; | |
2614 | |
2615 for (i = 0; i < data->lstream_count; i++) | |
2616 mark_object (data->lstreams[i]); | |
2617 } | |
2618 | |
2619 static void | |
2620 chain_mark_coding_stream (struct coding_stream *str) | |
2621 { | |
2622 chain_mark_coding_stream_1 (CODING_STREAM_TYPE_DATA (str, chain)); | |
2623 } | |
2624 | |
2625 static void | |
2626 chain_print (Lisp_Object cs, Lisp_Object printcharfun, int escapeflag) | |
2627 { | |
2628 int i; | |
2629 | |
826 | 2630 write_c_string (printcharfun, "("); |
771 | 2631 for (i = 0; i < XCODING_SYSTEM_CHAIN_COUNT (cs); i++) |
2632 { | |
826 | 2633 write_c_string (printcharfun, i == 0 ? "" : "->"); |
771 | 2634 print_coding_system_in_print_method (XCODING_SYSTEM_CHAIN_CHAIN (cs)[i], |
2635 printcharfun, escapeflag); | |
2636 } | |
2637 { | |
2638 Lisp_Object cac = XCODING_SYSTEM_CHAIN_CANONICALIZE_AFTER_CODING (cs); | |
2639 if (!NILP (cac)) | |
2640 { | |
2641 if (i > 0) | |
826 | 2642 write_c_string (printcharfun, " "); |
2643 write_c_string (printcharfun, "canonicalize-after-coding="); | |
771 | 2644 print_coding_system_in_print_method (cac, printcharfun, escapeflag); |
2645 } | |
2646 } | |
2647 | |
826 | 2648 write_c_string (printcharfun, ")"); |
771 | 2649 } |
2650 | |
2651 static void | |
2652 chain_rewind_coding_stream_1 (struct chain_coding_stream *data) | |
2653 { | |
2654 /* Each will rewind the next; there is always at least one stream (the | |
2655 dynarr stream at the end) if we're initted */ | |
2656 if (data->initted) | |
2657 Lstream_rewind (XLSTREAM (data->lstreams[0])); | |
2658 } | |
2659 | |
2660 static void | |
2661 chain_rewind_coding_stream (struct coding_stream *str) | |
2662 { | |
2663 chain_rewind_coding_stream_1 (CODING_STREAM_TYPE_DATA (str, chain)); | |
2664 } | |
2665 | |
2666 static void | |
2667 chain_init_coding_streams_1 (struct chain_coding_stream *data, | |
2668 unsigned_char_dynarr *dst, | |
2669 int ncodesys, Lisp_Object *codesys, | |
2670 enum encode_decode direction) | |
2671 { | |
2672 int i; | |
2673 Lisp_Object lstream_out; | |
2674 | |
2675 data->lstream_count = ncodesys + 1; | |
2676 data->lstreams = xnew_array (Lisp_Object, data->lstream_count); | |
2677 | |
2678 lstream_out = make_dynarr_output_stream (dst); | |
2679 Lstream_set_buffering (XLSTREAM (lstream_out), LSTREAM_UNBUFFERED, 0); | |
2680 data->lstreams[data->lstream_count - 1] = lstream_out; | |
2681 | |
2682 for (i = ncodesys - 1; i >= 0; i--) | |
2683 { | |
2684 data->lstreams[i] = | |
2685 make_coding_output_stream | |
2686 (XLSTREAM (lstream_out), | |
2687 codesys[direction == CODING_ENCODE ? ncodesys - (i + 1) : i], | |
800 | 2688 direction, 0); |
771 | 2689 lstream_out = data->lstreams[i]; |
2690 Lstream_set_buffering (XLSTREAM (lstream_out), LSTREAM_UNBUFFERED, | |
2691 0); | |
2692 } | |
2693 data->initted = 1; | |
2694 } | |
2695 | |
2696 static Bytecount | |
2697 chain_convert (struct coding_stream *str, const UExtbyte *src, | |
2698 unsigned_char_dynarr *dst, Bytecount n) | |
2699 { | |
2700 struct chain_coding_stream *data = CODING_STREAM_TYPE_DATA (str, chain); | |
2701 | |
2702 if (str->eof) | |
2703 { | |
2704 /* Each will close the next; there is always at least one stream (the | |
2705 dynarr stream at the end) if we're initted. We need to close now | |
2706 because more data may be generated. */ | |
2707 if (data->initted) | |
2708 Lstream_close (XLSTREAM (data->lstreams[0])); | |
2709 return n; | |
2710 } | |
2711 | |
2712 if (!data->initted) | |
2713 chain_init_coding_streams_1 | |
2714 (data, dst, XCODING_SYSTEM_CHAIN_COUNT (str->codesys), | |
2715 XCODING_SYSTEM_CHAIN_CHAIN (str->codesys), str->direction); | |
2716 | |
2717 if (Lstream_write (XLSTREAM (data->lstreams[0]), src, n) < 0) | |
2718 return -1; | |
2719 return n; | |
2720 } | |
2721 | |
2722 static void | |
2723 chain_finalize_coding_stream_1 (struct chain_coding_stream *data) | |
2724 { | |
2725 if (data->lstreams) | |
2726 { | |
2297 | 2727 /* During GC, these objects are unmarked, and are about to be freed. |
2728 We do NOT want them on the free list, and that will cause lots of | |
2729 nastiness including crashes. Just let them be freed normally. */ | |
771 | 2730 if (!gc_in_progress) |
2731 { | |
2732 int i; | |
2297 | 2733 /* Order of deletion is important here! Delete from the head of |
2734 the chain and work your way towards the tail. In general, | |
2735 when you delete an object, there should be *NO* pointers to it | |
2736 anywhere. Deleting back-to-front would be a problem because | |
2737 there are pointers going forward. If there were pointers in | |
2738 both directions, you'd have to disconnect the pointers to a | |
2739 particular object before deleting it. */ | |
771 | 2740 for (i = 0; i < data->lstream_count; i++) |
2741 Lstream_delete (XLSTREAM ((data->lstreams)[i])); | |
2742 } | |
1726 | 2743 xfree (data->lstreams, Lisp_Object *); |
771 | 2744 } |
2745 } | |
2746 | |
2747 static void | |
2748 chain_finalize_coding_stream (struct coding_stream *str) | |
2749 { | |
2750 chain_finalize_coding_stream_1 (CODING_STREAM_TYPE_DATA (str, chain)); | |
2751 } | |
2752 | |
2753 static void | |
2754 chain_finalize (Lisp_Object c) | |
2755 { | |
2756 if (XCODING_SYSTEM_CHAIN_CHAIN (c)) | |
1726 | 2757 xfree (XCODING_SYSTEM_CHAIN_CHAIN (c), Lisp_Object *); |
771 | 2758 } |
2759 | |
428 | 2760 static int |
771 | 2761 chain_putprop (Lisp_Object codesys, Lisp_Object key, Lisp_Object value) |
2762 { | |
2763 if (EQ (key, Qchain)) | |
2764 { | |
2765 Lisp_Object *cslist; | |
2766 int count = 0; | |
2767 int i; | |
2768 | |
2367 | 2769 { |
2770 EXTERNAL_LIST_LOOP_2 (elt, value) | |
2771 { | |
2772 Fget_coding_system (elt); | |
2773 count++; | |
2774 } | |
2775 } | |
771 | 2776 |
2777 cslist = xnew_array (Lisp_Object, count); | |
2778 XCODING_SYSTEM_CHAIN_CHAIN (codesys) = cslist; | |
2779 | |
2780 count = 0; | |
2367 | 2781 { |
2782 EXTERNAL_LIST_LOOP_2 (elt, value) | |
2783 { | |
2784 cslist[count] = Fget_coding_system (elt); | |
2785 count++; | |
2786 } | |
2787 } | |
771 | 2788 |
2789 XCODING_SYSTEM_CHAIN_COUNT (codesys) = count; | |
2790 | |
2791 for (i = 0; i < count - 1; i++) | |
2792 { | |
2793 if (decoding_source_sink_type_is_char (cslist[i], CODING_SINK) != | |
2794 decoding_source_sink_type_is_char (cslist[i + 1], CODING_SOURCE)) | |
2795 invalid_argument_2 ("Sink of first must match source of second", | |
2796 cslist[i], cslist[i + 1]); | |
2797 } | |
2798 } | |
2799 else if (EQ (key, Qcanonicalize_after_coding)) | |
2800 XCODING_SYSTEM_CHAIN_CANONICALIZE_AFTER_CODING (codesys) = | |
2801 Fget_coding_system (value); | |
2802 else | |
2803 return 0; | |
2804 return 1; | |
2805 } | |
2806 | |
2807 static Lisp_Object | |
2808 chain_getprop (Lisp_Object coding_system, Lisp_Object prop) | |
2809 { | |
2810 if (EQ (prop, Qchain)) | |
2811 { | |
2812 Lisp_Object result = Qnil; | |
2813 int i; | |
2814 | |
2815 for (i = 0; i < XCODING_SYSTEM_CHAIN_COUNT (coding_system); i++) | |
2816 result = Fcons (XCODING_SYSTEM_CHAIN_CHAIN (coding_system)[i], | |
2817 result); | |
2818 | |
2819 return Fnreverse (result); | |
2820 } | |
2821 else if (EQ (prop, Qcanonicalize_after_coding)) | |
2822 return XCODING_SYSTEM_CHAIN_CANONICALIZE_AFTER_CODING (coding_system); | |
2823 else | |
2824 return Qunbound; | |
2825 } | |
2826 | |
2827 static enum source_sink_type | |
2828 chain_conversion_end_type (Lisp_Object codesys) | |
2829 { | |
2830 Lisp_Object *cslist = XCODING_SYSTEM_CHAIN_CHAIN (codesys); | |
2831 int n = XCODING_SYSTEM_CHAIN_COUNT (codesys); | |
2832 int charp_source, charp_sink; | |
2833 | |
2834 if (n == 0) | |
2835 return DECODES_BYTE_TO_BYTE; /* arbitrary */ | |
2836 charp_source = decoding_source_sink_type_is_char (cslist[0], CODING_SOURCE); | |
2837 charp_sink = decoding_source_sink_type_is_char (cslist[n - 1], CODING_SINK); | |
2838 | |
2839 switch (charp_source * 2 + charp_sink) | |
2840 { | |
2841 case 0: return DECODES_BYTE_TO_BYTE; | |
2842 case 1: return DECODES_BYTE_TO_CHARACTER; | |
2843 case 2: return DECODES_CHARACTER_TO_BYTE; | |
2844 case 3: return DECODES_CHARACTER_TO_CHARACTER; | |
2845 } | |
2846 | |
2500 | 2847 ABORT (); |
771 | 2848 return DECODES_BYTE_TO_BYTE; |
2849 } | |
2850 | |
2851 | |
2852 /************************************************************************/ | |
2853 /* No-conversion methods */ | |
2854 /************************************************************************/ | |
2855 | |
2856 /* "No conversion"; used for binary files. We use quotes because there | |
2857 really is some conversion being applied (it does byte<->char | |
2858 conversion), but it appears to the user as if the text is read in | |
2297 | 2859 without conversion. |
2860 | |
2861 #### Shouldn't we _call_ it that, then? And while we're at it, | |
2862 separate it into "to_internal" and "to_external"? */ | |
771 | 2863 DEFINE_CODING_SYSTEM_TYPE (no_conversion); |
2864 | |
2865 /* This is used when reading in "binary" files -- i.e. files that may | |
2866 contain all 256 possible byte values and that are not to be | |
2867 interpreted as being in any particular encoding. */ | |
2868 static Bytecount | |
2869 no_conversion_convert (struct coding_stream *str, | |
2870 const UExtbyte *src, | |
2871 unsigned_char_dynarr *dst, Bytecount n) | |
2872 { | |
2873 UExtbyte c; | |
2874 unsigned int ch = str->ch; | |
2875 Bytecount orign = n; | |
2876 | |
2877 if (str->direction == CODING_DECODE) | |
2878 { | |
2879 while (n--) | |
2880 { | |
2881 c = *src++; | |
2882 | |
2883 DECODE_ADD_BINARY_CHAR (c, dst); | |
2884 } | |
2885 | |
2886 if (str->eof) | |
2887 DECODE_OUTPUT_PARTIAL_CHAR (ch, dst); | |
2888 } | |
2889 else | |
2890 { | |
2891 | |
2892 while (n--) | |
2893 { | |
2894 c = *src++; | |
826 | 2895 if (byte_ascii_p (c)) |
771 | 2896 { |
2897 assert (ch == 0); | |
2898 Dynarr_add (dst, c); | |
2899 } | |
2900 #ifdef MULE | |
867 | 2901 else if (ibyte_leading_byte_p (c)) |
771 | 2902 { |
2903 assert (ch == 0); | |
2904 if (c == LEADING_BYTE_LATIN_ISO8859_1 || | |
2905 c == LEADING_BYTE_CONTROL_1) | |
2906 ch = c; | |
2907 else | |
2297 | 2908 /* #### This is just plain unacceptable. */ |
771 | 2909 Dynarr_add (dst, '~'); /* untranslatable character */ |
2910 } | |
2911 else | |
2912 { | |
2913 if (ch == LEADING_BYTE_LATIN_ISO8859_1) | |
2914 Dynarr_add (dst, c); | |
2915 else if (ch == LEADING_BYTE_CONTROL_1) | |
2916 { | |
2917 assert (c < 0xC0); | |
2918 Dynarr_add (dst, c - 0x20); | |
2919 } | |
2920 /* else it should be the second or third byte of an | |
2921 untranslatable character, so ignore it */ | |
2922 ch = 0; | |
2923 } | |
2924 #endif /* MULE */ | |
2925 | |
2926 } | |
2927 } | |
2928 | |
2929 str->ch = ch; | |
2930 return orign; | |
2931 } | |
2932 | |
2933 DEFINE_DETECTOR (no_conversion); | |
2934 DEFINE_DETECTOR_CATEGORY (no_conversion, no_conversion); | |
2935 | |
2936 struct no_conversion_detector | |
2937 { | |
2938 int dummy; | |
2939 }; | |
2940 | |
2941 static void | |
2286 | 2942 no_conversion_detect (struct detection_state *st, const UExtbyte *UNUSED (src), |
2943 Bytecount UNUSED (n)) | |
771 | 2944 { |
2945 /* Hack until we get better handling of this stuff! */ | |
2946 DET_RESULT (st, no_conversion) = DET_SLIGHTLY_LIKELY; | |
2947 } | |
2948 | |
2949 | |
2950 /************************************************************************/ | |
2951 /* Convert-eol methods */ | |
2952 /************************************************************************/ | |
2953 | |
2954 /* This is used to handle end-of-line (EOL) differences. It is | |
2819 | 2955 character-to-character, and works (when encoding) *BEFORE* sending data to |
2956 the main encoding routine -- thus, that routine must handle different EOL | |
2957 types itself if it does line-oriented type processing. This is unavoidable | |
2958 because we don't know whether the output of the main encoding routine is | |
2959 ASCII compatible (UTF-16 is definitely not, for example). [[ sjt sez this | |
2960 is bogus. There should be _no_ EOL processing (or processing of any kind) | |
2961 after conversion to external. ]] | |
771 | 2962 |
793 | 2963 There is one parameter: `subtype', either `cr', `lf', `crlf', or nil. |
771 | 2964 */ |
2965 | |
2966 struct convert_eol_coding_system | |
2967 { | |
2968 enum eol_type subtype; | |
2132 | 2969 int dummy; /* On some architectures (eg ia64) the portable dumper can |
2970 produce unaligned access errors without this field. Probably | |
2971 because the combined structure of this structure and | |
2972 Lisp_Coding_System is not properly aligned. */ | |
771 | 2973 }; |
2974 | |
2975 #define CODING_SYSTEM_CONVERT_EOL_SUBTYPE(codesys) \ | |
2976 (CODING_SYSTEM_TYPE_DATA (codesys, convert_eol)->subtype) | |
2977 #define XCODING_SYSTEM_CONVERT_EOL_SUBTYPE(codesys) \ | |
2978 (XCODING_SYSTEM_TYPE_DATA (codesys, convert_eol)->subtype) | |
2979 | |
2980 struct convert_eol_coding_stream | |
2981 { | |
2982 enum eol_type actual; | |
2983 }; | |
2984 | |
1204 | 2985 static const struct memory_description |
771 | 2986 convert_eol_coding_system_description[] = { |
2987 { XD_END } | |
2988 }; | |
2989 | |
1204 | 2990 DEFINE_CODING_SYSTEM_TYPE_WITH_DATA (convert_eol); |
2991 | |
771 | 2992 static void |
2286 | 2993 convert_eol_print (Lisp_Object cs, Lisp_Object printcharfun, |
2994 int UNUSED (escapeflag)) | |
771 | 2995 { |
2996 struct convert_eol_coding_system *data = | |
2997 XCODING_SYSTEM_TYPE_DATA (cs, convert_eol); | |
2998 | |
2999 write_fmt_string (printcharfun, "(%s)", | |
3000 data->subtype == EOL_LF ? "lf" : | |
3001 data->subtype == EOL_CRLF ? "crlf" : | |
3002 data->subtype == EOL_CR ? "cr" : | |
793 | 3003 data->subtype == EOL_AUTODETECT ? "nil" : |
2500 | 3004 (ABORT(), "")); |
771 | 3005 } |
3006 | |
3007 static enum source_sink_type | |
2286 | 3008 convert_eol_conversion_end_type (Lisp_Object UNUSED (codesys)) |
771 | 3009 { |
3010 return DECODES_CHARACTER_TO_CHARACTER; | |
3011 } | |
3012 | |
3013 static int | |
3014 convert_eol_putprop (Lisp_Object codesys, | |
3015 Lisp_Object key, | |
3016 Lisp_Object value) | |
3017 { | |
3018 struct convert_eol_coding_system *data = | |
3019 XCODING_SYSTEM_TYPE_DATA (codesys, convert_eol); | |
3020 | |
3021 if (EQ (key, Qsubtype)) | |
3022 { | |
3023 if (EQ (value, Qlf) /* || EQ (value, Qunix) */) | |
3024 data->subtype = EOL_LF; | |
3025 else if (EQ (value, Qcrlf) /* || EQ (value, Qdos) */) | |
3026 data->subtype = EOL_CRLF; | |
3027 else if (EQ (value, Qcr) /* || EQ (value, Qmac) */) | |
3028 data->subtype = EOL_CR; | |
793 | 3029 else if (EQ (value, Qnil)) |
771 | 3030 data->subtype = EOL_AUTODETECT; |
3031 else invalid_constant ("Unrecognized eol type", value); | |
3032 } | |
3033 else | |
3034 return 0; | |
3035 return 1; | |
3036 } | |
3037 | |
3038 static Lisp_Object | |
3039 convert_eol_getprop (Lisp_Object coding_system, Lisp_Object prop) | |
3040 { | |
3041 struct convert_eol_coding_system *data = | |
3042 XCODING_SYSTEM_TYPE_DATA (coding_system, convert_eol); | |
3043 | |
3044 if (EQ (prop, Qsubtype)) | |
3045 { | |
3046 switch (data->subtype) | |
3047 { | |
3048 case EOL_LF: return Qlf; | |
3049 case EOL_CRLF: return Qcrlf; | |
3050 case EOL_CR: return Qcr; | |
793 | 3051 case EOL_AUTODETECT: return Qnil; |
2500 | 3052 default: ABORT (); |
771 | 3053 } |
3054 } | |
3055 | |
3056 return Qunbound; | |
3057 } | |
3058 | |
3059 static void | |
3060 convert_eol_init_coding_stream (struct coding_stream *str) | |
3061 { | |
3062 struct convert_eol_coding_stream *data = | |
3063 CODING_STREAM_TYPE_DATA (str, convert_eol); | |
3064 data->actual = XCODING_SYSTEM_CONVERT_EOL_SUBTYPE (str->codesys); | |
3065 } | |
3066 | |
3067 static Bytecount | |
867 | 3068 convert_eol_convert (struct coding_stream *str, const Ibyte *src, |
771 | 3069 unsigned_char_dynarr *dst, Bytecount n) |
3070 { | |
3071 if (str->direction == CODING_DECODE) | |
3072 { | |
3073 struct convert_eol_coding_stream *data = | |
3074 CODING_STREAM_TYPE_DATA (str, convert_eol); | |
3075 | |
3076 if (data->actual == EOL_AUTODETECT) | |
3077 { | |
3078 Bytecount n2 = n; | |
867 | 3079 const Ibyte *src2 = src; |
771 | 3080 |
3081 for (; n2; n2--) | |
3082 { | |
867 | 3083 Ibyte c = *src2++; |
771 | 3084 if (c == '\n') |
3085 { | |
3086 data->actual = EOL_LF; | |
3087 break; | |
3088 } | |
3089 else if (c == '\r') | |
3090 { | |
3091 if (n2 == 1) | |
3092 { | |
3093 /* If we're seeing a '\r' at the end of the data, then | |
3094 reject the '\r' right now so it doesn't become an | |
3095 issue in the code below -- unless we're at the end of | |
3096 the stream, in which case we can't do that (because | |
3097 then the '\r' will never get written out), and in any | |
3098 case we should be recognizing it at EOL_CR format. */ | |
3099 if (str->eof) | |
3100 data->actual = EOL_CR; | |
3101 else | |
3102 n--; | |
3103 break; | |
3104 } | |
3105 else if (*src2 == '\n') | |
3106 data->actual = EOL_CRLF; | |
3107 else | |
3108 data->actual = EOL_CR; | |
3109 break; | |
3110 } | |
3111 } | |
3112 } | |
3113 | |
3114 /* str->eof is set, the caller reached EOF on the other end and has | |
3115 no new data to give us. The only data we get is the data we | |
3116 rejected from last time. */ | |
3117 if (data->actual == EOL_LF || data->actual == EOL_AUTODETECT || | |
3118 (str->eof)) | |
3119 Dynarr_add_many (dst, src, n); | |
3120 else | |
3121 { | |
867 | 3122 const Ibyte *end = src + n; |
771 | 3123 while (1) |
3124 { | |
3125 /* Find the next section with no \r and add it. */ | |
867 | 3126 const Ibyte *runstart = src; |
3127 src = (Ibyte *) memchr (src, '\r', end - src); | |
771 | 3128 if (!src) |
3129 src = end; | |
3130 Dynarr_add_many (dst, runstart, src - runstart); | |
3131 /* Stop if at end ... */ | |
3132 if (src == end) | |
3133 break; | |
3134 /* ... else, translate as necessary. */ | |
3135 src++; | |
3136 if (data->actual == EOL_CR) | |
3137 Dynarr_add (dst, '\n'); | |
3138 /* We need to be careful here with CRLF. If we see a CR at the | |
3139 end of the data, we don't know if it's part of a CRLF, so we | |
3140 reject it. Otherwise: If it's part of a CRLF, eat it and | |
3141 loop; the following LF gets added next time around. If it's | |
3142 not part of a CRLF, add the CR and loop. The following | |
3143 character will be processed in the next loop iteration. This | |
3144 correctly handles a sequence like CR+CR+LF. */ | |
3145 else if (src == end) | |
3146 return n - 1; /* reject the CR at the end; we'll get it again | |
3147 next time the convert method is called */ | |
3148 else if (*src != '\n') | |
3149 Dynarr_add (dst, '\r'); | |
3150 } | |
3151 } | |
3152 | |
3153 return n; | |
3154 } | |
3155 else | |
3156 { | |
3157 enum eol_type subtype = | |
3158 XCODING_SYSTEM_CONVERT_EOL_SUBTYPE (str->codesys); | |
867 | 3159 const Ibyte *end = src + n; |
771 | 3160 |
3161 /* We try to be relatively efficient here. */ | |
3162 if (subtype == EOL_LF) | |
3163 Dynarr_add_many (dst, src, n); | |
3164 else | |
3165 { | |
3166 while (1) | |
3167 { | |
3168 /* Find the next section with no \n and add it. */ | |
867 | 3169 const Ibyte *runstart = src; |
3170 src = (Ibyte *) memchr (src, '\n', end - src); | |
771 | 3171 if (!src) |
3172 src = end; | |
3173 Dynarr_add_many (dst, runstart, src - runstart); | |
3174 /* Stop if at end ... */ | |
3175 if (src == end) | |
3176 break; | |
3177 /* ... else, skip over \n and add its translation. */ | |
3178 src++; | |
3179 Dynarr_add (dst, '\r'); | |
3180 if (subtype == EOL_CRLF) | |
3181 Dynarr_add (dst, '\n'); | |
3182 } | |
3183 } | |
3184 | |
3185 return n; | |
3186 } | |
3187 } | |
3188 | |
3189 static Lisp_Object | |
3190 convert_eol_canonicalize_after_coding (struct coding_stream *str) | |
3191 { | |
3192 struct convert_eol_coding_stream *data = | |
3193 CODING_STREAM_TYPE_DATA (str, convert_eol); | |
3194 | |
3195 if (str->direction == CODING_ENCODE) | |
3196 return str->codesys; | |
3197 | |
3198 switch (data->actual) | |
3199 { | |
3200 case EOL_LF: return Fget_coding_system (Qconvert_eol_lf); | |
3201 case EOL_CRLF: return Fget_coding_system (Qconvert_eol_crlf); | |
3202 case EOL_CR: return Fget_coding_system (Qconvert_eol_cr); | |
3203 case EOL_AUTODETECT: return str->codesys; | |
2500 | 3204 default: ABORT (); return Qnil; |
771 | 3205 } |
3206 } | |
3207 | |
3208 | |
3209 /************************************************************************/ | |
3210 /* Undecided methods */ | |
3211 /************************************************************************/ | |
3212 | |
3213 /* Do autodetection. We can autodetect the EOL type only, the coding | |
3214 system only, or both. We only do autodetection when decoding; when | |
3215 encoding, we just pass the data through. | |
3216 | |
3217 When doing just EOL detection, a coding system can be specified; if so, | |
3218 we will decode this data through the coding system before doing EOL | |
3219 detection. The reason for specifying this is so that | |
3220 canonicalize-after-coding works: We will canonicalize the specified | |
3221 coding system into the appropriate EOL type. When doing both coding and | |
3222 EOL detection, we do similar canonicalization, and also catch situations | |
3223 where the EOL type is overspecified, i.e. the detected coding system | |
3224 specifies an EOL type, and either switch to the equivalent | |
3225 non-EOL-processing coding system (if possible), or terminate EOL | |
3226 detection and use the specified EOL type. This prevents data from being | |
3227 EOL-processed twice. | |
3228 */ | |
3229 | |
3230 struct undecided_coding_system | |
3231 { | |
3232 int do_eol, do_coding; | |
3233 Lisp_Object cs; | |
3234 }; | |
3235 | |
3236 struct undecided_coding_stream | |
3237 { | |
3238 Lisp_Object actual; | |
3239 /* Either 2 or 3 lstreams here; see undecided_convert */ | |
3240 struct chain_coding_stream c; | |
3241 | |
3242 struct detection_state *st; | |
3243 }; | |
3244 | |
1204 | 3245 static const struct memory_description undecided_coding_system_description[] = { |
3246 { XD_LISP_OBJECT, offsetof (struct undecided_coding_system, cs) }, | |
771 | 3247 { XD_END } |
3248 }; | |
3249 | |
1204 | 3250 static const struct memory_description undecided_coding_stream_description_1 [] = { |
3251 { XD_LISP_OBJECT, offsetof (struct undecided_coding_stream, actual) }, | |
2367 | 3252 { XD_BLOCK_ARRAY, offsetof (struct undecided_coding_stream, c), |
2551 | 3253 1, { &chain_coding_stream_description } }, |
1204 | 3254 { XD_END } |
3255 }; | |
3256 | |
3257 const struct sized_memory_description undecided_coding_stream_description = { | |
3258 sizeof (struct undecided_coding_stream), undecided_coding_stream_description_1 | |
3259 }; | |
3260 | |
3261 DEFINE_CODING_SYSTEM_TYPE_WITH_DATA (undecided); | |
3262 | |
771 | 3263 static void |
3264 undecided_init (Lisp_Object codesys) | |
3265 { | |
3266 struct undecided_coding_system *data = | |
3267 XCODING_SYSTEM_TYPE_DATA (codesys, undecided); | |
3268 | |
3269 data->cs = Qnil; | |
3270 } | |
3271 | |
3272 static void | |
3273 undecided_mark (Lisp_Object codesys) | |
3274 { | |
3275 struct undecided_coding_system *data = | |
3276 XCODING_SYSTEM_TYPE_DATA (codesys, undecided); | |
3277 | |
3278 mark_object (data->cs); | |
3279 } | |
3280 | |
3281 static void | |
3282 undecided_print (Lisp_Object cs, Lisp_Object printcharfun, int escapeflag) | |
3283 { | |
3284 struct undecided_coding_system *data = | |
3285 XCODING_SYSTEM_TYPE_DATA (cs, undecided); | |
3286 int need_space = 0; | |
3287 | |
826 | 3288 write_c_string (printcharfun, "("); |
771 | 3289 if (data->do_eol) |
3290 { | |
826 | 3291 write_c_string (printcharfun, "do-eol"); |
771 | 3292 need_space = 1; |
3293 } | |
3294 if (data->do_coding) | |
3295 { | |
3296 if (need_space) | |
826 | 3297 write_c_string (printcharfun, " "); |
3298 write_c_string (printcharfun, "do-coding"); | |
771 | 3299 need_space = 1; |
3300 } | |
3301 if (!NILP (data->cs)) | |
3302 { | |
3303 if (need_space) | |
826 | 3304 write_c_string (printcharfun, " "); |
3305 write_c_string (printcharfun, "coding-system="); | |
771 | 3306 print_coding_system_in_print_method (data->cs, printcharfun, escapeflag); |
3307 } | |
826 | 3308 write_c_string (printcharfun, ")"); |
771 | 3309 } |
3310 | |
3311 static void | |
3312 undecided_mark_coding_stream (struct coding_stream *str) | |
3313 { | |
1204 | 3314 mark_object (CODING_STREAM_TYPE_DATA (str, undecided)->actual); |
771 | 3315 chain_mark_coding_stream_1 (&CODING_STREAM_TYPE_DATA (str, undecided)->c); |
3316 } | |
3317 | |
3318 static int | |
3319 undecided_putprop (Lisp_Object codesys, Lisp_Object key, Lisp_Object value) | |
3320 { | |
3321 struct undecided_coding_system *data = | |
3322 XCODING_SYSTEM_TYPE_DATA (codesys, undecided); | |
3323 | |
3324 if (EQ (key, Qdo_eol)) | |
3325 data->do_eol = 1; | |
3326 else if (EQ (key, Qdo_coding)) | |
3327 data->do_coding = 1; | |
3328 else if (EQ (key, Qcoding_system)) | |
3329 data->cs = get_coding_system_for_text_file (value, 0); | |
3330 else | |
3331 return 0; | |
3332 return 1; | |
3333 } | |
3334 | |
3335 static Lisp_Object | |
3336 undecided_getprop (Lisp_Object codesys, Lisp_Object prop) | |
3337 { | |
3338 struct undecided_coding_system *data = | |
3339 XCODING_SYSTEM_TYPE_DATA (codesys, undecided); | |
3340 | |
3341 if (EQ (prop, Qdo_eol)) | |
3342 return data->do_eol ? Qt : Qnil; | |
3343 if (EQ (prop, Qdo_coding)) | |
3344 return data->do_coding ? Qt : Qnil; | |
3345 if (EQ (prop, Qcoding_system)) | |
3346 return data->cs; | |
3347 return Qunbound; | |
3348 } | |
3349 | |
3350 static struct detection_state * | |
3351 allocate_detection_state (void) | |
3352 { | |
3353 int i; | |
3354 Bytecount size = MAX_ALIGN_SIZE (sizeof (struct detection_state)); | |
3355 struct detection_state *block; | |
3356 | |
3357 for (i = 0; i < coding_detector_count; i++) | |
3358 size += MAX_ALIGN_SIZE (Dynarr_at (all_coding_detectors, i).data_size); | |
3359 | |
3360 block = (struct detection_state *) xmalloc_and_zero (size); | |
3361 | |
3362 size = MAX_ALIGN_SIZE (sizeof (struct detection_state)); | |
3363 for (i = 0; i < coding_detector_count; i++) | |
3364 { | |
3365 block->data_offset[i] = size; | |
3366 size += MAX_ALIGN_SIZE (Dynarr_at (all_coding_detectors, i).data_size); | |
3367 } | |
3368 | |
3369 return block; | |
3370 } | |
3371 | |
3372 static void | |
3373 free_detection_state (struct detection_state *st) | |
3374 { | |
3375 int i; | |
3376 | |
3377 for (i = 0; i < coding_detector_count; i++) | |
3378 { | |
3379 if (Dynarr_at (all_coding_detectors, i).finalize_detection_state_method) | |
3380 Dynarr_at (all_coding_detectors, i).finalize_detection_state_method | |
3381 (st); | |
3382 } | |
3383 | |
1726 | 3384 xfree (st, struct detection_state *); |
771 | 3385 } |
3386 | |
3387 static int | |
3388 coding_category_symbol_to_id (Lisp_Object symbol) | |
428 | 3389 { |
3390 int i; | |
3391 | |
3392 CHECK_SYMBOL (symbol); | |
771 | 3393 for (i = 0; i < coding_detector_count; i++) |
3394 { | |
3395 detector_category_dynarr *cats = | |
3396 Dynarr_at (all_coding_detectors, i).cats; | |
3397 int j; | |
3398 | |
3399 for (j = 0; j < Dynarr_length (cats); j++) | |
3400 if (EQ (Dynarr_at (cats, j).sym, symbol)) | |
3401 return Dynarr_at (cats, j).id; | |
3402 } | |
3403 | |
563 | 3404 invalid_constant ("Unrecognized coding category", symbol); |
1204 | 3405 RETURN_NOT_REACHED (0); |
428 | 3406 } |
3407 | |
771 | 3408 static Lisp_Object |
3409 coding_category_id_to_symbol (int id) | |
428 | 3410 { |
3411 int i; | |
771 | 3412 |
3413 for (i = 0; i < coding_detector_count; i++) | |
3414 { | |
3415 detector_category_dynarr *cats = | |
3416 Dynarr_at (all_coding_detectors, i).cats; | |
3417 int j; | |
3418 | |
3419 for (j = 0; j < Dynarr_length (cats); j++) | |
3420 if (id == Dynarr_at (cats, j).id) | |
3421 return Dynarr_at (cats, j).sym; | |
3422 } | |
3423 | |
2500 | 3424 ABORT (); |
771 | 3425 return Qnil; /* (usually) not reached */ |
428 | 3426 } |
3427 | |
771 | 3428 static Lisp_Object |
3429 detection_result_number_to_symbol (enum detection_result result) | |
428 | 3430 { |
1494 | 3431 /* let compiler warn if not all enumerators are handled */ |
3432 switch (result) { | |
3433 #define FROB(sym, num) case num: return (sym) | |
771 | 3434 FROB (Qnear_certainty, DET_NEAR_CERTAINTY); |
3435 FROB (Qquite_probable, DET_QUITE_PROBABLE); | |
3436 FROB (Qsomewhat_likely, DET_SOMEWHAT_LIKELY); | |
1494 | 3437 FROB (Qslightly_likely, DET_SLIGHTLY_LIKELY); |
771 | 3438 FROB (Qas_likely_as_unlikely, DET_AS_LIKELY_AS_UNLIKELY); |
3439 FROB (Qsomewhat_unlikely, DET_SOMEWHAT_UNLIKELY); | |
3440 FROB (Qquite_improbable, DET_QUITE_IMPROBABLE); | |
3441 FROB (Qnearly_impossible, DET_NEARLY_IMPOSSIBLE); | |
3442 #undef FROB | |
1494 | 3443 } |
771 | 3444 |
2500 | 3445 ABORT (); |
771 | 3446 return Qnil; /* (usually) not reached */ |
3447 } | |
3448 | |
778 | 3449 #if 0 /* not used */ |
771 | 3450 static enum detection_result |
3451 detection_result_symbol_to_number (Lisp_Object symbol) | |
3452 { | |
1494 | 3453 /* using switch here would be bad style, and doesn't help */ |
771 | 3454 #define FROB(sym, num) if (EQ (symbol, sym)) return (num) |
3455 FROB (Qnear_certainty, DET_NEAR_CERTAINTY); | |
3456 FROB (Qquite_probable, DET_QUITE_PROBABLE); | |
3457 FROB (Qsomewhat_likely, DET_SOMEWHAT_LIKELY); | |
1494 | 3458 FROB (Qslightly_likely, DET_SLIGHTLY_LIKELY); |
771 | 3459 FROB (Qas_likely_as_unlikely, DET_AS_LIKELY_AS_UNLIKELY); |
3460 FROB (Qsomewhat_unlikely, DET_SOMEWHAT_UNLIKELY); | |
3461 FROB (Qquite_improbable, DET_QUITE_IMPROBABLE); | |
3462 FROB (Qnearly_impossible, DET_NEARLY_IMPOSSIBLE); | |
3463 #undef FROB | |
3464 | |
3465 invalid_constant ("Unrecognized detection result", symbol); | |
3466 return ((enum detection_result) 0); /* not reached */ | |
3467 } | |
778 | 3468 #endif /* 0 */ |
771 | 3469 |
3470 /* Set all detection results for a given detector to a specified value. */ | |
3471 void | |
3472 set_detection_results (struct detection_state *st, int detector, int given) | |
3473 { | |
3474 detector_category_dynarr *cats = | |
3475 Dynarr_at (all_coding_detectors, detector).cats; | |
3476 int i; | |
3477 | |
3478 for (i = 0; i < Dynarr_length (cats); i++) | |
3479 st->categories[Dynarr_at (cats, i).id] = given; | |
3480 } | |
428 | 3481 |
3482 static int | |
3483 acceptable_control_char_p (int c) | |
3484 { | |
3485 switch (c) | |
3486 { | |
3487 /* Allow and ignore control characters that you might | |
3488 reasonably see in a text file */ | |
3489 case '\r': | |
3490 case '\n': | |
3491 case '\t': | |
3492 case 7: /* bell */ | |
3493 case 8: /* backspace */ | |
3494 case 11: /* vertical tab */ | |
3495 case 12: /* form feed */ | |
3496 case 26: /* MS-DOS C-z junk */ | |
3497 case 31: /* '^_' -- for info */ | |
3498 return 1; | |
3499 default: | |
3500 return 0; | |
3501 } | |
3502 } | |
3503 | |
771 | 3504 #ifdef DEBUG_XEMACS |
3505 | |
3506 static UExtbyte | |
3507 hex_digit_to_char (int digit) | |
428 | 3508 { |
771 | 3509 if (digit < 10) |
3510 return digit + '0'; | |
3511 else | |
3512 return digit - 10 + 'A'; | |
428 | 3513 } |
3514 | |
771 | 3515 static void |
3516 output_bytes_in_ascii_and_hex (const UExtbyte *src, Bytecount n) | |
428 | 3517 { |
3425 | 3518 Extbyte *ascii = alloca_array (Extbyte, n + 1); |
3519 Extbyte *hex = alloca_array (Extbyte, 3 * n + 1); | |
771 | 3520 int i; |
3413 | 3521 DECLARE_EISTRING (eistr_ascii); |
3522 DECLARE_EISTRING (eistr_hex); | |
771 | 3523 |
3524 for (i = 0; i < n; i++) | |
428 | 3525 { |
3425 | 3526 Extbyte c = src[i]; |
771 | 3527 if (c < 0x20) |
3528 ascii[i] = '.'; | |
428 | 3529 else |
771 | 3530 ascii[i] = c; |
3531 hex[3 * i] = hex_digit_to_char (c >> 4); | |
3532 hex[3 * i + 1] = hex_digit_to_char (c & 0xF); | |
3533 hex[3 * i + 2] = ' '; | |
428 | 3534 } |
771 | 3535 ascii[i] = '\0'; |
3536 hex[3 * i - 1] = '\0'; | |
3413 | 3537 |
3538 eicpy_ext(eistr_hex, hex, Qbinary); | |
3539 eicpy_ext(eistr_ascii, ascii, Qbinary); | |
3540 | |
3425 | 3541 stderr_out ("%s %s", eidata(eistr_ascii), eidata(eistr_hex)); |
428 | 3542 } |
3543 | |
771 | 3544 #endif /* DEBUG_XEMACS */ |
3545 | |
3546 /* Attempt to determine the encoding of the given text. Before calling | |
3547 this function for the first time, you must zero out the detection state. | |
428 | 3548 |
3549 Returns: | |
3550 | |
771 | 3551 0 == keep going |
3552 1 == stop | |
428 | 3553 */ |
3554 | |
3555 static int | |
771 | 3556 detect_coding_type (struct detection_state *st, const UExtbyte *src, |
3557 Bytecount n) | |
428 | 3558 { |
771 | 3559 Bytecount n2 = n; |
3560 const UExtbyte *src2 = src; | |
3561 int i; | |
3562 | |
3563 #ifdef DEBUG_XEMACS | |
3564 if (!NILP (Vdebug_coding_detection)) | |
3565 { | |
3566 int bytes = min (16, n); | |
3567 stderr_out ("detect_coding_type: processing %ld bytes\n", n); | |
3568 stderr_out ("First %d: ", bytes); | |
3569 output_bytes_in_ascii_and_hex (src, bytes); | |
3570 stderr_out ("\nLast %d: ", bytes); | |
3571 output_bytes_in_ascii_and_hex (src + n - bytes, bytes); | |
3572 stderr_out ("\n"); | |
3573 } | |
3574 #endif /* DEBUG_XEMACS */ | |
428 | 3575 if (!st->seen_non_ascii) |
3576 { | |
771 | 3577 for (; n2; n2--, src2++) |
428 | 3578 { |
771 | 3579 UExtbyte c = *src2; |
428 | 3580 if ((c < 0x20 && !acceptable_control_char_p (c)) || c >= 0x80) |
3581 { | |
3582 st->seen_non_ascii = 1; | |
3583 break; | |
3584 } | |
3585 } | |
3586 } | |
3587 | |
771 | 3588 for (i = 0; i < coding_detector_count; i++) |
3589 Dynarr_at (all_coding_detectors, i).detect_method (st, src, n); | |
3590 | |
3591 st->bytes_seen += n; | |
3592 | |
3593 #ifdef DEBUG_XEMACS | |
3594 if (!NILP (Vdebug_coding_detection)) | |
3595 { | |
3596 stderr_out ("seen_non_ascii: %d\n", st->seen_non_ascii); | |
1494 | 3597 if (coding_detector_category_count <= 0) |
3598 stderr_out ("found %d detector categories\n", | |
3599 coding_detector_category_count); | |
771 | 3600 for (i = 0; i < coding_detector_category_count; i++) |
3601 stderr_out_lisp | |
3602 ("%s: %s\n", | |
3603 2, | |
3604 coding_category_id_to_symbol (i), | |
3605 detection_result_number_to_symbol ((enum detection_result) | |
3606 st->categories[i])); | |
3607 } | |
3608 #endif /* DEBUG_XEMACS */ | |
3609 | |
3610 { | |
3611 int not_unlikely = 0; | |
3612 int retval; | |
3613 | |
3614 for (i = 0; i < coding_detector_category_count; i++) | |
3615 if (st->categories[i] >= 0) | |
3616 not_unlikely++; | |
3617 | |
3618 retval = (not_unlikely <= 1 | |
3619 #if 0 /* this is bogus */ | |
3620 || st->bytes_seen >= MAX_BYTES_PROCESSED_FOR_DETECTION | |
428 | 3621 #endif |
771 | 3622 ); |
3623 | |
3624 #ifdef DEBUG_XEMACS | |
3625 if (!NILP (Vdebug_coding_detection)) | |
3626 stderr_out ("detect_coding_type: returning %d (%s)\n", | |
3627 retval, retval ? "stop" : "keep going"); | |
3628 #endif /* DEBUG_XEMACS */ | |
3629 | |
3630 return retval; | |
428 | 3631 } |
3632 } | |
3633 | |
3634 static Lisp_Object | |
771 | 3635 detected_coding_system (struct detection_state *st) |
428 | 3636 { |
771 | 3637 int i; |
3638 int even = 1; | |
3639 | |
3640 if (st->seen_non_ascii) | |
3641 { | |
3642 for (i = 0; i < coding_detector_category_count; i++) | |
3643 if (st->categories[i] != DET_AS_LIKELY_AS_UNLIKELY) | |
3644 { | |
3645 even = 0; | |
3646 break; | |
3647 } | |
3648 } | |
3649 | |
3650 /* #### Here we are ignoring the results of detection when it's all | |
3651 ASCII. This is obviously a bad thing. But we need to fix up the | |
3652 existing detection methods somewhat before we can switch. */ | |
3653 if (even) | |
428 | 3654 { |
3655 /* If the file was entirely or basically ASCII, use the | |
3656 default value of `buffer-file-coding-system'. */ | |
3657 Lisp_Object retval = | |
3658 XBUFFER (Vbuffer_defaults)->buffer_file_coding_system; | |
3659 if (!NILP (retval)) | |
3660 { | |
771 | 3661 retval = find_coding_system_for_text_file (retval, 0); |
428 | 3662 if (NILP (retval)) |
3663 { | |
3664 warn_when_safe | |
3665 (Qbad_variable, Qwarning, | |
3666 "Invalid `default-buffer-file-coding-system', set to nil"); | |
3667 XBUFFER (Vbuffer_defaults)->buffer_file_coding_system = Qnil; | |
3668 } | |
3669 } | |
3670 if (NILP (retval)) | |
4100 | 3671 retval = Fget_coding_system (Qbinary); |
428 | 3672 return retval; |
3673 } | |
3674 else | |
3675 { | |
771 | 3676 int likelihood; |
3677 Lisp_Object retval = Qnil; | |
3678 | |
3679 /* Look through the coding categories first by likelihood and then by | |
3680 priority and find the first one that is allowed. */ | |
3681 | |
3682 for (likelihood = DET_HIGHEST; likelihood >= DET_LOWEST; likelihood--) | |
428 | 3683 { |
771 | 3684 for (i = 0; i < coding_detector_category_count; i++) |
3685 { | |
3686 int cat = coding_category_by_priority[i]; | |
3687 if (st->categories[cat] == likelihood && | |
3688 !NILP (coding_category_system[cat])) | |
3689 { | |
3690 retval = (get_coding_system_for_text_file | |
3691 (coding_category_system[cat], 0)); | |
3692 if (likelihood < DET_AS_LIKELY_AS_UNLIKELY) | |
3693 warn_when_safe_lispobj | |
3694 (intern ("detection"), | |
793 | 3695 Qwarning, |
771 | 3696 emacs_sprintf_string_lisp |
3697 ( | |
3698 "Detected coding %s is unlikely to be correct (likelihood == `%s')", | |
3699 Qnil, 2, XCODING_SYSTEM_NAME (retval), | |
3700 detection_result_number_to_symbol | |
3701 ((enum detection_result) likelihood))); | |
3702 return retval; | |
3703 } | |
3704 } | |
428 | 3705 } |
771 | 3706 |
3707 return Fget_coding_system (Qraw_text); | |
428 | 3708 } |
3709 } | |
3710 | |
1347 | 3711 /* Look for a coding system in the string (skipping over leading |
3712 blanks). If found, return it, otherwise nil. */ | |
3713 | |
3714 static Lisp_Object | |
2531 | 3715 snarf_coding_system (const UExtbyte *p, Bytecount len) |
1347 | 3716 { |
3717 Bytecount n; | |
2531 | 3718 UExtbyte *name; |
1347 | 3719 |
3720 while (*p == ' ' || *p == '\t') p++, len--; | |
3721 len = min (len, 1000); | |
3722 name = alloca_ibytes (len + 1); | |
3723 memcpy (name, p, len); | |
3724 name[len] = '\0'; | |
3725 | |
3726 /* Get coding system name */ | |
3727 /* Characters valid in a MIME charset name (rfc 1521), | |
3728 and in a Lisp symbol name. */ | |
3729 n = qxestrspn (name, | |
3730 "ABCDEFGHIJKLMNOPQRSTUVWXYZ" | |
3731 "abcdefghijklmnopqrstuvwxyz" | |
3732 "0123456789" | |
3733 "!$%&*+-.^_{|}~"); | |
3734 if (n > 0) | |
3735 { | |
3736 name[n] = '\0'; | |
2531 | 3737 /* This call to intern_int() is OK because we already verified that |
3738 there are only ASCII characters in the string */ | |
3739 return find_coding_system_for_text_file (intern_int ((Ibyte *) name), 0); | |
1347 | 3740 } |
3741 | |
3742 return Qnil; | |
3743 } | |
3744 | |
428 | 3745 /* Given a seekable read stream and potential coding system and EOL type |
3746 as specified, do any autodetection that is called for. If the | |
3747 coding system and/or EOL type are not `autodetect', they will be left | |
3748 alone; but this function will never return an autodetect coding system | |
3749 or EOL type. | |
3750 | |
3751 This function does not automatically fetch subsidiary coding systems; | |
3752 that should be unnecessary with the explicit eol-type argument. */ | |
3753 | |
3754 #define LENGTH(string_constant) (sizeof (string_constant) - 1) | |
3755 | |
771 | 3756 static Lisp_Object |
3757 unwind_free_detection_state (Lisp_Object opaque) | |
3758 { | |
3759 struct detection_state *st = | |
3760 (struct detection_state *) get_opaque_ptr (opaque); | |
3761 free_detection_state (st); | |
3762 free_opaque_ptr (opaque); | |
3763 return Qnil; | |
3764 } | |
3765 | |
1347 | 3766 /* #### This duplicates code in `find-coding-system-magic-cookie-in-file' |
3767 in files.el. Look into combining them. */ | |
3768 | |
771 | 3769 static Lisp_Object |
3770 look_for_coding_system_magic_cookie (const UExtbyte *data, Bytecount len) | |
428 | 3771 { |
771 | 3772 const UExtbyte *p; |
3773 const UExtbyte *scan_end; | |
2531 | 3774 Bytecount cookie_len; |
771 | 3775 |
3776 /* Look for initial "-*-"; mode line prefix */ | |
3777 for (p = data, | |
3778 scan_end = data + len - LENGTH ("-*-coding:?-*-"); | |
3779 p <= scan_end | |
3780 && *p != '\n' | |
3781 && *p != '\r'; | |
3782 p++) | |
3783 if (*p == '-' && *(p+1) == '*' && *(p+2) == '-') | |
3784 { | |
3785 const UExtbyte *local_vars_beg = p + 3; | |
3786 /* Look for final "-*-"; mode line suffix */ | |
3787 for (p = local_vars_beg, | |
3788 scan_end = data + len - LENGTH ("-*-"); | |
3789 p <= scan_end | |
428 | 3790 && *p != '\n' |
3791 && *p != '\r'; | |
771 | 3792 p++) |
3793 if (*p == '-' && *(p+1) == '*' && *(p+2) == '-') | |
3794 { | |
3795 const UExtbyte *suffix = p; | |
3796 /* Look for "coding:" */ | |
3797 for (p = local_vars_beg, | |
3798 scan_end = suffix - LENGTH ("coding:?"); | |
3799 p <= scan_end; | |
3800 p++) | |
3801 if (memcmp ("coding:", p, LENGTH ("coding:")) == 0 | |
3802 && (p == local_vars_beg | |
3803 || (*(p-1) == ' ' || | |
3804 *(p-1) == '\t' || | |
3805 *(p-1) == ';'))) | |
3806 { | |
3807 p += LENGTH ("coding:"); | |
1347 | 3808 return snarf_coding_system (p, suffix - p); |
771 | 3809 break; |
3810 } | |
3811 break; | |
3812 } | |
3813 break; | |
3814 } | |
3815 | |
2531 | 3816 /* Look for ;;;###coding system */ |
3817 | |
3818 cookie_len = LENGTH (";;;###coding system: "); | |
3819 | |
3820 for (p = data, | |
3821 scan_end = data + len - cookie_len; | |
3822 p <= scan_end; | |
3823 p++) | |
1347 | 3824 { |
2531 | 3825 if (*p == ';' && !memcmp (p, ";;;###coding system: ", cookie_len)) |
3826 { | |
3827 const UExtbyte *suffix; | |
3828 | |
3829 p += cookie_len; | |
3830 suffix = p; | |
3831 while (suffix < scan_end && !isspace (*suffix)) | |
3832 suffix++; | |
3833 return snarf_coding_system (p, suffix - p); | |
3834 } | |
1347 | 3835 } |
3836 | |
3837 return Qnil; | |
771 | 3838 } |
3839 | |
3840 static Lisp_Object | |
3841 determine_real_coding_system (Lstream *stream) | |
3842 { | |
3843 struct detection_state *st = allocate_detection_state (); | |
3844 int depth = record_unwind_protect (unwind_free_detection_state, | |
3845 make_opaque_ptr (st)); | |
3846 UExtbyte buf[4096]; | |
3847 Bytecount nread = Lstream_read (stream, buf, sizeof (buf)); | |
3848 Lisp_Object coding_system = look_for_coding_system_magic_cookie (buf, nread); | |
3849 | |
3850 if (NILP (coding_system)) | |
3851 { | |
3852 while (1) | |
3853 { | |
3854 if (detect_coding_type (st, buf, nread)) | |
428 | 3855 break; |
771 | 3856 nread = Lstream_read (stream, buf, sizeof (buf)); |
3857 if (nread == 0) | |
3858 break; | |
428 | 3859 } |
771 | 3860 |
3861 coding_system = detected_coding_system (st); | |
428 | 3862 } |
3863 | |
3864 Lstream_rewind (stream); | |
771 | 3865 |
3866 unbind_to (depth); | |
3867 return coding_system; | |
3868 } | |
3869 | |
3870 static void | |
3871 undecided_init_coding_stream (struct coding_stream *str) | |
3872 { | |
3873 struct undecided_coding_stream *data = | |
3874 CODING_STREAM_TYPE_DATA (str, undecided); | |
3875 struct undecided_coding_system *csdata = | |
3876 XCODING_SYSTEM_TYPE_DATA (str->codesys, undecided); | |
3877 | |
3878 data->actual = Qnil; | |
3879 | |
3880 if (str->direction == CODING_DECODE) | |
3881 { | |
3882 Lstream *lst = str->other_end; | |
3883 | |
3884 if ((lst->flags & LSTREAM_FL_READ) && | |
3885 Lstream_seekable_p (lst) && | |
3886 csdata->do_coding) | |
3887 /* We can determine the coding system now. */ | |
3888 data->actual = determine_real_coding_system (lst); | |
3889 } | |
1494 | 3890 |
3891 #ifdef DEBUG_XEMACS | |
3892 if (!NILP (Vdebug_coding_detection)) | |
3893 stderr_out_lisp ("detected coding system: %s\n", 1, data->actual); | |
3894 #endif /* DEBUG_XEMACS */ | |
771 | 3895 } |
3896 | |
3897 static void | |
3898 undecided_rewind_coding_stream (struct coding_stream *str) | |
3899 { | |
3900 chain_rewind_coding_stream_1 (&CODING_STREAM_TYPE_DATA (str, undecided)->c); | |
3901 } | |
3902 | |
3903 static void | |
3904 undecided_finalize_coding_stream (struct coding_stream *str) | |
3905 { | |
3906 struct undecided_coding_stream *data = | |
3907 CODING_STREAM_TYPE_DATA (str, undecided); | |
3908 | |
3909 chain_finalize_coding_stream_1 | |
3910 (&CODING_STREAM_TYPE_DATA (str, undecided)->c); | |
3911 if (data->st) | |
3912 free_detection_state (data->st); | |
3913 } | |
3914 | |
3915 static Lisp_Object | |
3916 undecided_canonicalize (Lisp_Object codesys) | |
3917 { | |
3918 struct undecided_coding_system *csdata = | |
3919 XCODING_SYSTEM_TYPE_DATA (codesys, undecided); | |
3920 if (!csdata->do_eol && !csdata->do_coding) | |
3921 return NILP (csdata->cs) ? Fget_coding_system (Qbinary) : csdata->cs; | |
3922 if (csdata->do_eol && !csdata->do_coding && NILP (csdata->cs)) | |
3923 return Fget_coding_system (Qconvert_eol_autodetect); | |
3924 return codesys; | |
3925 } | |
3926 | |
3927 static Bytecount | |
3928 undecided_convert (struct coding_stream *str, const UExtbyte *src, | |
3929 unsigned_char_dynarr *dst, Bytecount n) | |
3930 { | |
3931 int first_time = 0; | |
3932 | |
3933 if (str->direction == CODING_DECODE) | |
3934 { | |
3935 /* At this point, we have only the following possibilities: | |
3936 | |
3937 do_eol && do_coding | |
3938 do_coding only | |
3939 do_eol only and a coding system was specified | |
3940 | |
3941 Other possibilities are removed during undecided_canonicalize. | |
3942 | |
3943 Therefore, our substreams are either | |
3944 | |
3945 lstream_coding -> lstream_dynarr, or | |
3946 lstream_coding -> lstream_eol -> lstream_dynarr. | |
3947 */ | |
3948 struct undecided_coding_system *csdata = | |
3949 XCODING_SYSTEM_TYPE_DATA (str->codesys, undecided); | |
3950 struct undecided_coding_stream *data = | |
3951 CODING_STREAM_TYPE_DATA (str, undecided); | |
3952 | |
3953 if (str->eof) | |
3954 { | |
3955 /* Each will close the next. We need to close now because more | |
3956 data may be generated. */ | |
3957 if (data->c.initted) | |
3958 Lstream_close (XLSTREAM (data->c.lstreams[0])); | |
3959 return n; | |
3960 } | |
3961 | |
3962 if (!data->c.initted) | |
3963 { | |
3964 data->c.lstream_count = csdata->do_eol ? 3 : 2; | |
3965 data->c.lstreams = xnew_array (Lisp_Object, data->c.lstream_count); | |
3966 | |
3967 data->c.lstreams[data->c.lstream_count - 1] = | |
3968 make_dynarr_output_stream (dst); | |
3969 Lstream_set_buffering | |
3970 (XLSTREAM (data->c.lstreams[data->c.lstream_count - 1]), | |
3971 LSTREAM_UNBUFFERED, 0); | |
3972 if (csdata->do_eol) | |
3973 { | |
3974 data->c.lstreams[1] = | |
3975 make_coding_output_stream | |
3976 (XLSTREAM (data->c.lstreams[data->c.lstream_count - 1]), | |
3977 Fget_coding_system (Qconvert_eol_autodetect), | |
800 | 3978 CODING_DECODE, 0); |
771 | 3979 Lstream_set_buffering |
3980 (XLSTREAM (data->c.lstreams[1]), | |
3981 LSTREAM_UNBUFFERED, 0); | |
3982 } | |
3983 | |
3984 data->c.lstreams[0] = | |
3985 make_coding_output_stream | |
3986 (XLSTREAM (data->c.lstreams[1]), | |
3987 /* Substitute binary if we need to detect the encoding */ | |
3988 csdata->do_coding ? Qbinary : csdata->cs, | |
800 | 3989 CODING_DECODE, 0); |
771 | 3990 Lstream_set_buffering (XLSTREAM (data->c.lstreams[0]), |
3991 LSTREAM_UNBUFFERED, 0); | |
3992 | |
3993 first_time = 1; | |
3994 data->c.initted = 1; | |
3995 } | |
3996 | |
3997 /* If necessary, do encoding-detection now. We do this when we're a | |
3998 writing stream or a non-seekable reading stream, meaning that we | |
3999 can't just process the whole input, rewind, and start over. */ | |
4000 | |
4001 if (csdata->do_coding) | |
4002 { | |
4003 int actual_was_nil = NILP (data->actual); | |
4004 if (NILP (data->actual)) | |
4005 { | |
4006 if (!data->st) | |
4007 data->st = allocate_detection_state (); | |
4008 if (first_time) | |
4009 /* #### This is cheesy. What we really ought to do is buffer | |
4010 up a certain minimum amount of data to get a better result. | |
4011 */ | |
4012 data->actual = look_for_coding_system_magic_cookie (src, n); | |
4013 if (NILP (data->actual)) | |
4014 { | |
4015 /* #### This is cheesy. What we really ought to do is buffer | |
4016 up a certain minimum amount of data so as to get a less | |
4017 random result when doing subprocess detection. */ | |
4018 detect_coding_type (data->st, src, n); | |
4019 data->actual = detected_coding_system (data->st); | |
4100 | 4020 /* kludge to prevent infinite recursion */ |
4021 if (XCODING_SYSTEM(data->actual)->methods->enumtype == undecided_coding_system) | |
4022 data->actual = Fget_coding_system (Qbinary); | |
771 | 4023 } |
4024 } | |
4025 /* We need to set the detected coding system if we actually have | |
4026 such a coding system but didn't before. That is the case | |
4027 either when we just detected it in the previous code or when | |
4028 it was detected during undecided_init_coding_stream(). We | |
4029 can check for that using first_time. */ | |
4030 if (!NILP (data->actual) && (actual_was_nil || first_time)) | |
4031 { | |
4032 /* If the detected coding system doesn't allow for EOL | |
4033 autodetection, try to get the equivalent that does; | |
4034 otherwise, disable EOL detection (overriding whatever | |
4035 may already have been detected). */ | |
4036 if (XCODING_SYSTEM_EOL_TYPE (data->actual) != EOL_AUTODETECT) | |
4037 { | |
4038 if (!NILP (XCODING_SYSTEM_SUBSIDIARY_PARENT (data->actual))) | |
4039 data->actual = | |
4040 XCODING_SYSTEM_SUBSIDIARY_PARENT (data->actual); | |
4041 else if (data->c.lstream_count == 3) | |
4042 set_coding_stream_coding_system | |
4043 (XLSTREAM (data->c.lstreams[1]), | |
4044 Fget_coding_system (Qidentity)); | |
4045 } | |
4046 set_coding_stream_coding_system | |
4047 (XLSTREAM (data->c.lstreams[0]), data->actual); | |
4048 } | |
4049 } | |
4050 | |
4051 if (Lstream_write (XLSTREAM (data->c.lstreams[0]), src, n) < 0) | |
4052 return -1; | |
4053 return n; | |
4054 } | |
4055 else | |
4056 return no_conversion_convert (str, src, dst, n); | |
4057 } | |
4058 | |
4059 static Lisp_Object | |
4060 undecided_canonicalize_after_coding (struct coding_stream *str) | |
4061 { | |
4062 struct undecided_coding_stream *data = | |
4063 CODING_STREAM_TYPE_DATA (str, undecided); | |
4064 Lisp_Object ret, eolret; | |
4065 | |
4066 if (str->direction == CODING_ENCODE) | |
4067 return str->codesys; | |
4068 | |
4069 if (!data->c.initted) | |
4070 return Fget_coding_system (Qundecided); | |
4071 | |
4072 ret = coding_stream_canonicalize_after_coding | |
4073 (XLSTREAM (data->c.lstreams[0])); | |
4074 if (NILP (ret)) | |
4075 ret = Fget_coding_system (Qundecided); | |
4076 if (XCODING_SYSTEM_EOL_TYPE (ret) != EOL_AUTODETECT) | |
4077 return ret; | |
4078 eolret = coding_stream_canonicalize_after_coding | |
4079 (XLSTREAM (data->c.lstreams[1])); | |
4080 if (!EQ (XCODING_SYSTEM_TYPE (eolret), Qconvert_eol)) | |
4081 return ret; | |
4082 return | |
4083 Fsubsidiary_coding_system (ret, Fcoding_system_property (eolret, | |
4084 Qsubtype)); | |
4085 } | |
4086 | |
4087 | |
4088 /************************************************************************/ | |
4089 /* Lisp interface: Coding category functions and detection */ | |
4090 /************************************************************************/ | |
4091 | |
4092 DEFUN ("coding-category-list", Fcoding_category_list, 0, 0, 0, /* | |
4093 Return a list of all recognized coding categories. | |
4094 */ | |
4095 ()) | |
4096 { | |
4097 int i; | |
4098 Lisp_Object list = Qnil; | |
4099 | |
4100 for (i = 0; i < coding_detector_count; i++) | |
4101 { | |
4102 detector_category_dynarr *cats = | |
4103 Dynarr_at (all_coding_detectors, i).cats; | |
4104 int j; | |
4105 | |
4106 for (j = 0; j < Dynarr_length (cats); j++) | |
4107 list = Fcons (Dynarr_at (cats, j).sym, list); | |
4108 } | |
4109 | |
4110 return Fnreverse (list); | |
4111 } | |
4112 | |
4113 DEFUN ("set-coding-priority-list", Fset_coding_priority_list, 1, 1, 0, /* | |
4114 Change the priority order of the coding categories. | |
4115 LIST should be list of coding categories, in descending order of | |
4116 priority. Unspecified coding categories will be lower in priority | |
4117 than all specified ones, in the same relative order they were in | |
4118 previously. | |
4119 */ | |
4120 (list)) | |
4121 { | |
4122 int *category_to_priority = | |
4123 alloca_array (int, coding_detector_category_count); | |
4124 int i, j; | |
4125 | |
4126 /* First generate a list that maps coding categories to priorities. */ | |
4127 | |
4128 for (i = 0; i < coding_detector_category_count; i++) | |
4129 category_to_priority[i] = -1; | |
4130 | |
4131 /* Highest priority comes from the specified list. */ | |
4132 i = 0; | |
2367 | 4133 { |
4134 EXTERNAL_LIST_LOOP_2 (elt, list) | |
4135 { | |
4136 int cat = coding_category_symbol_to_id (elt); | |
4137 | |
4138 if (category_to_priority[cat] >= 0) | |
4139 sferror ("Duplicate coding category in list", elt); | |
4140 category_to_priority[cat] = i++; | |
4141 } | |
4142 } | |
771 | 4143 |
4144 /* Now go through the existing categories by priority to retrieve | |
4145 the categories not yet specified and preserve their priority | |
4146 order. */ | |
4147 for (j = 0; j < coding_detector_category_count; j++) | |
4148 { | |
4149 int cat = coding_category_by_priority[j]; | |
4150 if (category_to_priority[cat] < 0) | |
4151 category_to_priority[cat] = i++; | |
4152 } | |
4153 | |
4154 /* Now we need to construct the inverse of the mapping we just | |
4155 constructed. */ | |
4156 | |
4157 for (i = 0; i < coding_detector_category_count; i++) | |
4158 coding_category_by_priority[category_to_priority[i]] = i; | |
4159 | |
4160 /* Phew! That was confusing. */ | |
4161 return Qnil; | |
4162 } | |
4163 | |
4164 DEFUN ("coding-priority-list", Fcoding_priority_list, 0, 0, 0, /* | |
4165 Return a list of coding categories in descending order of priority. | |
4166 */ | |
4167 ()) | |
4168 { | |
4169 int i; | |
4170 Lisp_Object list = Qnil; | |
4171 | |
4172 for (i = 0; i < coding_detector_category_count; i++) | |
4173 list = | |
4174 Fcons (coding_category_id_to_symbol (coding_category_by_priority[i]), | |
4175 list); | |
4176 return Fnreverse (list); | |
4177 } | |
4178 | |
4179 DEFUN ("set-coding-category-system", Fset_coding_category_system, 2, 2, 0, /* | |
4180 Change the coding system associated with a coding category. | |
4181 */ | |
4182 (coding_category, coding_system)) | |
4183 { | |
4184 coding_category_system[coding_category_symbol_to_id (coding_category)] = | |
4185 Fget_coding_system (coding_system); | |
4186 return Qnil; | |
4187 } | |
4188 | |
4189 DEFUN ("coding-category-system", Fcoding_category_system, 1, 1, 0, /* | |
4190 Return the coding system associated with a coding category. | |
4191 */ | |
4192 (coding_category)) | |
4193 { | |
4194 Lisp_Object sys = | |
4195 coding_category_system[coding_category_symbol_to_id (coding_category)]; | |
4196 | |
4197 if (!NILP (sys)) | |
4198 return XCODING_SYSTEM_NAME (sys); | |
4199 return Qnil; | |
4200 } | |
4201 | |
800 | 4202 /* Detect the encoding of STREAM. Assumes stream is at the begnning and will |
4203 read through to the end of STREAM, leaving it there but open. */ | |
4204 | |
771 | 4205 Lisp_Object |
4206 detect_coding_stream (Lisp_Object stream) | |
4207 { | |
4208 Lisp_Object val = Qnil; | |
4209 struct gcpro gcpro1, gcpro2, gcpro3; | |
4210 UExtbyte random_buffer[65536]; | |
4211 Lisp_Object binary_instream = | |
4212 make_coding_input_stream | |
4213 (XLSTREAM (stream), Qbinary, | |
814 | 4214 CODING_ENCODE, LSTREAM_FL_NO_CLOSE_OTHER); |
771 | 4215 Lisp_Object decstream = |
4216 make_coding_input_stream | |
4217 (XLSTREAM (binary_instream), | |
800 | 4218 Qundecided, CODING_DECODE, 0); |
771 | 4219 Lstream *decstr = XLSTREAM (decstream); |
4220 | |
4221 GCPRO3 (decstream, stream, binary_instream); | |
4222 /* Read and discard all data; detection happens as a side effect of this, | |
4223 and we examine what was detected afterwards. */ | |
4224 while (Lstream_read (decstr, random_buffer, sizeof (random_buffer)) > 0) | |
4225 ; | |
4226 | |
4227 val = coding_stream_detected_coding_system (decstr); | |
4228 Lstream_close (decstr); | |
4229 Lstream_delete (decstr); | |
4230 Lstream_delete (XLSTREAM (binary_instream)); | |
4231 UNGCPRO; | |
4232 return val; | |
428 | 4233 } |
4234 | |
4235 DEFUN ("detect-coding-region", Fdetect_coding_region, 2, 3, 0, /* | |
4236 Detect coding system of the text in the region between START and END. | |
444 | 4237 Return a list of possible coding systems ordered by priority. |
3025 | 4238 If only ASCII characters are found, return `undecided' or one of |
428 | 4239 its subsidiary coding systems according to a detected end-of-line |
4240 type. Optional arg BUFFER defaults to the current buffer. | |
4241 */ | |
4242 (start, end, buffer)) | |
4243 { | |
4244 Lisp_Object val = Qnil; | |
4245 struct buffer *buf = decode_buffer (buffer, 0); | |
665 | 4246 Charbpos b, e; |
771 | 4247 Lisp_Object lb_instream; |
428 | 4248 |
4249 get_buffer_range_char (buf, start, end, &b, &e, 0); | |
4250 lb_instream = make_lisp_buffer_input_stream (buf, b, e, 0); | |
771 | 4251 |
4252 val = detect_coding_stream (lb_instream); | |
4253 Lstream_delete (XLSTREAM (lb_instream)); | |
428 | 4254 return val; |
4255 } | |
4256 | |
4257 | |
771 | 4258 |
4259 #ifdef DEBUG_XEMACS | |
4260 | |
428 | 4261 /************************************************************************/ |
771 | 4262 /* Internal methods */ |
4263 /************************************************************************/ | |
4264 | |
4265 /* Raw (internally-formatted) data. */ | |
4266 DEFINE_CODING_SYSTEM_TYPE (internal); | |
428 | 4267 |
665 | 4268 static Bytecount |
2286 | 4269 internal_convert (struct coding_stream *UNUSED (str), const UExtbyte *src, |
771 | 4270 unsigned_char_dynarr *dst, Bytecount n) |
4271 { | |
4272 Bytecount orign = n; | |
4273 Dynarr_add_many (dst, src, n); | |
4274 return orign; | |
4275 } | |
4276 | |
4277 #endif /* DEBUG_XEMACS */ | |
4278 | |
4279 | |
4280 | |
4281 #ifdef HAVE_ZLIB | |
4282 | |
4283 /************************************************************************/ | |
4284 /* Gzip methods */ | |
4285 /************************************************************************/ | |
4286 | |
4287 struct gzip_coding_system | |
428 | 4288 { |
771 | 4289 int level; /* 0 through 9, or -1 for default */ |
4290 }; | |
4291 | |
4292 #define CODING_SYSTEM_GZIP_LEVEL(codesys) \ | |
4293 (CODING_SYSTEM_TYPE_DATA (codesys, gzip)->level) | |
4294 #define XCODING_SYSTEM_GZIP_LEVEL(codesys) \ | |
4295 (XCODING_SYSTEM_TYPE_DATA (codesys, gzip)->level) | |
4296 | |
4297 struct gzip_coding_stream | |
428 | 4298 { |
771 | 4299 z_stream stream; |
4300 int stream_initted; | |
4301 int reached_eof; /* #### this should be handled by the caller, once we | |
4302 return LSTREAM_EOF */ | |
4303 }; | |
4304 | |
1204 | 4305 static const struct memory_description |
771 | 4306 gzip_coding_system_description[] = { |
4307 { XD_END } | |
4308 }; | |
4309 | |
1204 | 4310 DEFINE_CODING_SYSTEM_TYPE_WITH_DATA (gzip); |
4311 | |
771 | 4312 enum source_sink_type |
4313 gzip_conversion_end_type (Lisp_Object codesys) | |
4314 { | |
4315 return DECODES_BYTE_TO_BYTE; | |
428 | 4316 } |
4317 | |
4318 static void | |
771 | 4319 gzip_init (Lisp_Object codesys) |
4320 { | |
4321 struct gzip_coding_system *data = XCODING_SYSTEM_TYPE_DATA (codesys, gzip); | |
4322 data->level = -1; | |
4323 } | |
4324 | |
4325 static void | |
4326 gzip_print (Lisp_Object cs, Lisp_Object printcharfun, int escapeflag) | |
428 | 4327 { |
771 | 4328 struct gzip_coding_system *data = XCODING_SYSTEM_TYPE_DATA (cs, gzip); |
4329 | |
826 | 4330 write_c_string (printcharfun, "("); |
771 | 4331 if (data->level == -1) |
826 | 4332 write_c_string (printcharfun, "default"); |
771 | 4333 else |
4334 print_internal (make_int (data->level), printcharfun, 0); | |
826 | 4335 write_c_string (printcharfun, ")"); |
428 | 4336 } |
4337 | |
4338 static int | |
771 | 4339 gzip_putprop (Lisp_Object codesys, Lisp_Object key, Lisp_Object value) |
428 | 4340 { |
771 | 4341 struct gzip_coding_system *data = XCODING_SYSTEM_TYPE_DATA (codesys, gzip); |
4342 | |
4343 if (EQ (key, Qlevel)) | |
428 | 4344 { |
771 | 4345 if (EQ (value, Qdefault)) |
4346 data->level = -1; | |
4347 else | |
428 | 4348 { |
771 | 4349 CHECK_INT (value); |
4350 check_int_range (XINT (value), 0, 9); | |
4351 data->level = XINT (value); | |
428 | 4352 } |
4353 } | |
4354 else | |
771 | 4355 return 0; |
4356 return 1; | |
428 | 4357 } |
4358 | |
4359 static Lisp_Object | |
771 | 4360 gzip_getprop (Lisp_Object coding_system, Lisp_Object prop) |
428 | 4361 { |
771 | 4362 struct gzip_coding_system *data = |
4363 XCODING_SYSTEM_TYPE_DATA (coding_system, gzip); | |
4364 | |
4365 if (EQ (prop, Qlevel)) | |
428 | 4366 { |
771 | 4367 if (data->level == -1) |
4368 return Qdefault; | |
4369 return make_int (data->level); | |
428 | 4370 } |
771 | 4371 |
4372 return Qunbound; | |
428 | 4373 } |
4374 | |
4375 static void | |
771 | 4376 gzip_init_coding_stream (struct coding_stream *str) |
428 | 4377 { |
771 | 4378 struct gzip_coding_stream *data = CODING_STREAM_TYPE_DATA (str, gzip); |
4379 if (data->stream_initted) | |
428 | 4380 { |
771 | 4381 if (str->direction == CODING_DECODE) |
4382 inflateEnd (&data->stream); | |
4383 else | |
4384 deflateEnd (&data->stream); | |
4385 data->stream_initted = 0; | |
428 | 4386 } |
771 | 4387 data->reached_eof = 0; |
428 | 4388 } |
4389 | |
4390 static void | |
771 | 4391 gzip_rewind_coding_stream (struct coding_stream *str) |
428 | 4392 { |
771 | 4393 gzip_init_coding_stream (str); |
428 | 4394 } |
4395 | |
771 | 4396 static Bytecount |
4397 gzip_convert (struct coding_stream *str, | |
4398 const UExtbyte *src, | |
4399 unsigned_char_dynarr *dst, Bytecount n) | |
428 | 4400 { |
771 | 4401 struct gzip_coding_stream *data = CODING_STREAM_TYPE_DATA (str, gzip); |
4402 int zerr; | |
4403 if (str->direction == CODING_DECODE) | |
428 | 4404 { |
771 | 4405 if (data->reached_eof) |
4406 return n; /* eat the data */ | |
4407 | |
4408 if (!data->stream_initted) | |
428 | 4409 { |
771 | 4410 xzero (data->stream); |
4411 if (inflateInit (&data->stream) != Z_OK) | |
4412 return LSTREAM_ERROR; | |
4413 data->stream_initted = 1; | |
428 | 4414 } |
771 | 4415 |
4416 data->stream.next_in = (Bytef *) src; | |
4417 data->stream.avail_in = n; | |
4418 | |
4419 /* Normally we stop when we've fed all data to the decompressor; but | |
4420 if we're at the end of the input, and the decompressor hasn't | |
4421 reported EOF, we need to keep going, as there might be more output | |
4422 to generate. Z_OK from the decompressor means input was processed | |
4423 or output was generated; if neither, we break out of the loop. | |
4424 Other return values are: | |
4425 | |
4426 Z_STREAM_END EOF from decompressor | |
4427 Z_DATA_ERROR Corrupted data | |
4428 Z_BUF_ERROR No progress possible (this should happen if | |
4429 we try to feed it an incomplete file) | |
4430 Z_MEM_ERROR Out of memory | |
4431 Z_STREAM_ERROR (should never happen) | |
4432 Z_NEED_DICT (#### when will this happen?) | |
4433 */ | |
4434 while (data->stream.avail_in > 0 || str->eof) | |
4435 { | |
4436 /* Reserve an output buffer of the same size as the input buffer; | |
4437 if that's not enough, we keep reserving the same size. */ | |
4438 Bytecount reserved = n; | |
4439 Dynarr_add_many (dst, 0, reserved); | |
4440 /* Careful here! Don't retrieve the pointer until after | |
4441 reserving the space, or it might be bogus */ | |
4442 data->stream.next_out = | |
4443 Dynarr_atp (dst, Dynarr_length (dst) - reserved); | |
4444 data->stream.avail_out = reserved; | |
4445 zerr = inflate (&data->stream, Z_NO_FLUSH); | |
4446 /* Lop off the unused portion */ | |
4447 Dynarr_set_size (dst, Dynarr_length (dst) - data->stream.avail_out); | |
4448 if (zerr != Z_OK) | |
4449 break; | |
4450 } | |
4451 | |
4452 if (zerr == Z_STREAM_END) | |
4453 data->reached_eof = 1; | |
4454 | |
4455 if ((Bytecount) data->stream.avail_in < n) | |
4456 return n - data->stream.avail_in; | |
4457 | |
4458 if (zerr == Z_OK || zerr == Z_STREAM_END) | |
4459 return 0; | |
4460 | |
4461 return LSTREAM_ERROR; | |
428 | 4462 } |
4463 else | |
4464 { | |
771 | 4465 if (!data->stream_initted) |
4466 { | |
4467 int level = XCODING_SYSTEM_GZIP_LEVEL (str->codesys); | |
4468 xzero (data->stream); | |
4469 if (deflateInit (&data->stream, | |
4470 level == -1 ? Z_DEFAULT_COMPRESSION : level) != | |
4471 Z_OK) | |
4472 return LSTREAM_ERROR; | |
4473 data->stream_initted = 1; | |
428 | 4474 } |
771 | 4475 |
4476 data->stream.next_in = (Bytef *) src; | |
4477 data->stream.avail_in = n; | |
4478 | |
4479 /* Normally we stop when we've fed all data to the compressor; but if | |
4480 we're at the end of the input, and the compressor hasn't reported | |
4481 EOF, we need to keep going, as there might be more output to | |
4482 generate. (To signal EOF on our end, we set the FLUSH parameter | |
4483 to Z_FINISH; when all data is output, Z_STREAM_END will be | |
4484 returned.) Z_OK from the compressor means input was processed or | |
4485 output was generated; if neither, we break out of the loop. Other | |
4486 return values are: | |
4487 | |
4488 Z_STREAM_END EOF from compressor | |
4489 Z_BUF_ERROR No progress possible (should never happen) | |
4490 Z_STREAM_ERROR (should never happen) | |
4491 */ | |
4492 while (data->stream.avail_in > 0 || str->eof) | |
4493 { | |
4494 /* Reserve an output buffer of the same size as the input buffer; | |
4495 if that's not enough, we keep reserving the same size. */ | |
4496 Bytecount reserved = n; | |
4497 Dynarr_add_many (dst, 0, reserved); | |
4498 /* Careful here! Don't retrieve the pointer until after | |
4499 reserving the space, or it might be bogus */ | |
4500 data->stream.next_out = | |
4501 Dynarr_atp (dst, Dynarr_length (dst) - reserved); | |
4502 data->stream.avail_out = reserved; | |
4503 zerr = | |
4504 deflate (&data->stream, | |
4505 str->eof ? Z_FINISH : Z_NO_FLUSH); | |
4506 /* Lop off the unused portion */ | |
4507 Dynarr_set_size (dst, Dynarr_length (dst) - data->stream.avail_out); | |
4508 if (zerr != Z_OK) | |
4509 break; | |
4510 } | |
4511 | |
4512 if ((Bytecount) data->stream.avail_in < n) | |
4513 return n - data->stream.avail_in; | |
4514 | |
4515 if (zerr == Z_OK || zerr == Z_STREAM_END) | |
4516 return 0; | |
4517 | |
4518 return LSTREAM_ERROR; | |
428 | 4519 } |
4520 } | |
4521 | |
771 | 4522 #endif /* HAVE_ZLIB */ |
428 | 4523 |
4524 | |
4525 /************************************************************************/ | |
4526 /* Initialization */ | |
4527 /************************************************************************/ | |
4528 | |
4529 void | |
4530 syms_of_file_coding (void) | |
4531 { | |
442 | 4532 INIT_LRECORD_IMPLEMENTATION (coding_system); |
4533 | |
771 | 4534 DEFSUBR (Fvalid_coding_system_type_p); |
4535 DEFSUBR (Fcoding_system_type_list); | |
428 | 4536 DEFSUBR (Fcoding_system_p); |
4303 | 4537 DEFSUBR (Fautoload_coding_system); |
428 | 4538 DEFSUBR (Ffind_coding_system); |
4539 DEFSUBR (Fget_coding_system); | |
4540 DEFSUBR (Fcoding_system_list); | |
4541 DEFSUBR (Fcoding_system_name); | |
4542 DEFSUBR (Fmake_coding_system); | |
4543 DEFSUBR (Fcopy_coding_system); | |
440 | 4544 DEFSUBR (Fcoding_system_canonical_name_p); |
4545 DEFSUBR (Fcoding_system_alias_p); | |
4546 DEFSUBR (Fcoding_system_aliasee); | |
428 | 4547 DEFSUBR (Fdefine_coding_system_alias); |
4548 DEFSUBR (Fsubsidiary_coding_system); | |
771 | 4549 DEFSUBR (Fcoding_system_base); |
4550 DEFSUBR (Fcoding_system_used_for_io); | |
428 | 4551 |
4552 DEFSUBR (Fcoding_system_type); | |
771 | 4553 DEFSUBR (Fcoding_system_description); |
428 | 4554 DEFSUBR (Fcoding_system_property); |
4555 | |
4556 DEFSUBR (Fcoding_category_list); | |
4557 DEFSUBR (Fset_coding_priority_list); | |
4558 DEFSUBR (Fcoding_priority_list); | |
4559 DEFSUBR (Fset_coding_category_system); | |
4560 DEFSUBR (Fcoding_category_system); | |
4561 | |
4562 DEFSUBR (Fdetect_coding_region); | |
4563 DEFSUBR (Fdecode_coding_region); | |
4564 DEFSUBR (Fencode_coding_region); | |
563 | 4565 DEFSYMBOL_MULTIWORD_PREDICATE (Qcoding_systemp); |
4566 DEFSYMBOL (Qno_conversion); | |
771 | 4567 DEFSYMBOL (Qconvert_eol); |
4568 DEFSYMBOL (Qconvert_eol_autodetect); | |
4569 DEFSYMBOL (Qconvert_eol_lf); | |
4570 DEFSYMBOL (Qconvert_eol_cr); | |
4571 DEFSYMBOL (Qconvert_eol_crlf); | |
563 | 4572 DEFSYMBOL (Qraw_text); |
771 | 4573 |
563 | 4574 DEFSYMBOL (Qmnemonic); |
4575 DEFSYMBOL (Qeol_type); | |
4576 DEFSYMBOL (Qpost_read_conversion); | |
4577 DEFSYMBOL (Qpre_write_conversion); | |
4578 | |
771 | 4579 DEFSYMBOL (Qtranslation_table_for_decode); |
4580 DEFSYMBOL (Qtranslation_table_for_encode); | |
4581 DEFSYMBOL (Qsafe_chars); | |
4582 DEFSYMBOL (Qsafe_charsets); | |
4583 DEFSYMBOL (Qmime_charset); | |
4584 DEFSYMBOL (Qvalid_codes); | |
4585 | |
563 | 4586 DEFSYMBOL (Qcr); |
4587 DEFSYMBOL (Qlf); | |
4588 DEFSYMBOL (Qcrlf); | |
4589 DEFSYMBOL (Qeol_cr); | |
4590 DEFSYMBOL (Qeol_lf); | |
4591 DEFSYMBOL (Qeol_crlf); | |
4592 DEFSYMBOL (Qencode); | |
4593 DEFSYMBOL (Qdecode); | |
428 | 4594 |
771 | 4595 DEFSYMBOL (Qnear_certainty); |
4596 DEFSYMBOL (Qquite_probable); | |
4597 DEFSYMBOL (Qsomewhat_likely); | |
1494 | 4598 DEFSYMBOL (Qslightly_likely); |
771 | 4599 DEFSYMBOL (Qas_likely_as_unlikely); |
4600 DEFSYMBOL (Qsomewhat_unlikely); | |
4601 DEFSYMBOL (Qquite_improbable); | |
4602 DEFSYMBOL (Qnearly_impossible); | |
4603 | |
4604 DEFSYMBOL (Qdo_eol); | |
4605 DEFSYMBOL (Qdo_coding); | |
4606 | |
4607 DEFSYMBOL (Qcanonicalize_after_coding); | |
4608 | |
4303 | 4609 DEFSYMBOL (Qposix_charset_to_coding_system_hash); |
4610 | |
771 | 4611 DEFSYMBOL (Qescape_quoted); |
4612 | |
4613 #ifdef HAVE_ZLIB | |
4614 DEFSYMBOL (Qgzip); | |
4615 #endif | |
4616 | |
428 | 4617 } |
4618 | |
4619 void | |
4620 lstream_type_create_file_coding (void) | |
4621 { | |
771 | 4622 LSTREAM_HAS_METHOD (coding, reader); |
4623 LSTREAM_HAS_METHOD (coding, writer); | |
4624 LSTREAM_HAS_METHOD (coding, rewinder); | |
4625 LSTREAM_HAS_METHOD (coding, seekable_p); | |
4626 LSTREAM_HAS_METHOD (coding, marker); | |
4627 LSTREAM_HAS_METHOD (coding, flusher); | |
4628 LSTREAM_HAS_METHOD (coding, closer); | |
4629 LSTREAM_HAS_METHOD (coding, finalizer); | |
4630 } | |
4631 | |
4632 void | |
4633 coding_system_type_create (void) | |
4634 { | |
4635 int i; | |
4636 | |
4637 staticpro (&Vcoding_system_hash_table); | |
4638 Vcoding_system_hash_table = | |
4639 make_lisp_hash_table (50, HASH_TABLE_NON_WEAK, HASH_TABLE_EQ); | |
4640 | |
4641 the_coding_system_type_entry_dynarr = Dynarr_new (coding_system_type_entry); | |
2367 | 4642 dump_add_root_block_ptr (&the_coding_system_type_entry_dynarr, |
771 | 4643 &csted_description); |
4644 | |
4645 Vcoding_system_type_list = Qnil; | |
4646 staticpro (&Vcoding_system_type_list); | |
4647 | |
4648 /* Initialize to something reasonable ... */ | |
4649 for (i = 0; i < MAX_DETECTOR_CATEGORIES; i++) | |
4650 { | |
4651 coding_category_system[i] = Qnil; | |
1204 | 4652 dump_add_root_lisp_object (&coding_category_system[i]); |
771 | 4653 coding_category_by_priority[i] = i; |
4654 } | |
4655 | |
4656 dump_add_opaque (coding_category_by_priority, | |
4657 sizeof (coding_category_by_priority)); | |
4658 | |
4659 all_coding_detectors = Dynarr_new2 (detector_dynarr, struct detector); | |
2367 | 4660 dump_add_root_block_ptr (&all_coding_detectors, |
771 | 4661 &detector_dynarr_description); |
4662 | |
4663 dump_add_opaque_int (&coding_system_tick); | |
4664 dump_add_opaque_int (&coding_detector_count); | |
4665 dump_add_opaque_int (&coding_detector_category_count); | |
4666 | |
4667 INITIALIZE_CODING_SYSTEM_TYPE (no_conversion, | |
4668 "no-conversion-coding-system-p"); | |
4669 CODING_SYSTEM_HAS_METHOD (no_conversion, convert); | |
4670 | |
4671 INITIALIZE_DETECTOR (no_conversion); | |
4672 DETECTOR_HAS_METHOD (no_conversion, detect); | |
4673 INITIALIZE_DETECTOR_CATEGORY (no_conversion, no_conversion); | |
4674 | |
4675 INITIALIZE_CODING_SYSTEM_TYPE_WITH_DATA (convert_eol, | |
4676 "convert-eol-coding-system-p"); | |
4677 CODING_SYSTEM_HAS_METHOD (convert_eol, print); | |
4678 CODING_SYSTEM_HAS_METHOD (convert_eol, convert); | |
4679 CODING_SYSTEM_HAS_METHOD (convert_eol, getprop); | |
4680 CODING_SYSTEM_HAS_METHOD (convert_eol, putprop); | |
4681 CODING_SYSTEM_HAS_METHOD (convert_eol, conversion_end_type); | |
4682 CODING_SYSTEM_HAS_METHOD (convert_eol, canonicalize_after_coding); | |
4683 CODING_SYSTEM_HAS_METHOD (convert_eol, init_coding_stream); | |
4684 | |
4685 INITIALIZE_CODING_SYSTEM_TYPE_WITH_DATA (undecided, | |
4686 "undecided-coding-system-p"); | |
4687 CODING_SYSTEM_HAS_METHOD (undecided, init); | |
4688 CODING_SYSTEM_HAS_METHOD (undecided, mark); | |
4689 CODING_SYSTEM_HAS_METHOD (undecided, print); | |
4690 CODING_SYSTEM_HAS_METHOD (undecided, convert); | |
4691 CODING_SYSTEM_HAS_METHOD (undecided, putprop); | |
4692 CODING_SYSTEM_HAS_METHOD (undecided, getprop); | |
4693 CODING_SYSTEM_HAS_METHOD (undecided, init_coding_stream); | |
4694 CODING_SYSTEM_HAS_METHOD (undecided, rewind_coding_stream); | |
4695 CODING_SYSTEM_HAS_METHOD (undecided, finalize_coding_stream); | |
4696 CODING_SYSTEM_HAS_METHOD (undecided, mark_coding_stream); | |
4697 CODING_SYSTEM_HAS_METHOD (undecided, canonicalize); | |
4698 CODING_SYSTEM_HAS_METHOD (undecided, canonicalize_after_coding); | |
4699 | |
4700 INITIALIZE_CODING_SYSTEM_TYPE_WITH_DATA (chain, "chain-coding-system-p"); | |
4701 | |
4702 CODING_SYSTEM_HAS_METHOD (chain, print); | |
4703 CODING_SYSTEM_HAS_METHOD (chain, canonicalize); | |
4704 CODING_SYSTEM_HAS_METHOD (chain, init); | |
4705 CODING_SYSTEM_HAS_METHOD (chain, mark); | |
4706 CODING_SYSTEM_HAS_METHOD (chain, mark_coding_stream); | |
4707 CODING_SYSTEM_HAS_METHOD (chain, convert); | |
4708 CODING_SYSTEM_HAS_METHOD (chain, rewind_coding_stream); | |
4709 CODING_SYSTEM_HAS_METHOD (chain, finalize_coding_stream); | |
4710 CODING_SYSTEM_HAS_METHOD (chain, finalize); | |
4711 CODING_SYSTEM_HAS_METHOD (chain, putprop); | |
4712 CODING_SYSTEM_HAS_METHOD (chain, getprop); | |
4713 CODING_SYSTEM_HAS_METHOD (chain, conversion_end_type); | |
4714 CODING_SYSTEM_HAS_METHOD (chain, canonicalize_after_coding); | |
4715 | |
4716 #ifdef DEBUG_XEMACS | |
4717 INITIALIZE_CODING_SYSTEM_TYPE (internal, "internal-coding-system-p"); | |
4718 CODING_SYSTEM_HAS_METHOD (internal, convert); | |
4719 #endif | |
4720 | |
4721 #ifdef HAVE_ZLIB | |
4722 INITIALIZE_CODING_SYSTEM_TYPE_WITH_DATA (gzip, "gzip-coding-system-p"); | |
4723 CODING_SYSTEM_HAS_METHOD (gzip, conversion_end_type); | |
4724 CODING_SYSTEM_HAS_METHOD (gzip, convert); | |
4725 CODING_SYSTEM_HAS_METHOD (gzip, init); | |
4726 CODING_SYSTEM_HAS_METHOD (gzip, print); | |
4727 CODING_SYSTEM_HAS_METHOD (gzip, init_coding_stream); | |
4728 CODING_SYSTEM_HAS_METHOD (gzip, rewind_coding_stream); | |
4729 CODING_SYSTEM_HAS_METHOD (gzip, putprop); | |
4730 CODING_SYSTEM_HAS_METHOD (gzip, getprop); | |
4731 #endif | |
4732 } | |
4733 | |
4734 void | |
4735 reinit_coding_system_type_create (void) | |
4736 { | |
4737 REINITIALIZE_CODING_SYSTEM_TYPE (no_conversion); | |
4738 REINITIALIZE_CODING_SYSTEM_TYPE (convert_eol); | |
4739 REINITIALIZE_CODING_SYSTEM_TYPE (undecided); | |
4740 REINITIALIZE_CODING_SYSTEM_TYPE (chain); | |
4741 #if 0 | |
4742 REINITIALIZE_CODING_SYSTEM_TYPE (text_file_wrapper); | |
4743 #endif /* 0 */ | |
4744 #ifdef DEBUG_XEMACS | |
4745 REINITIALIZE_CODING_SYSTEM_TYPE (internal); | |
4746 #endif | |
4747 #ifdef HAVE_ZLIB | |
4748 REINITIALIZE_CODING_SYSTEM_TYPE (gzip); | |
4749 #endif | |
4750 } | |
4751 | |
4752 void | |
4753 reinit_vars_of_file_coding (void) | |
4754 { | |
428 | 4755 } |
4756 | |
4757 void | |
4758 vars_of_file_coding (void) | |
4759 { | |
771 | 4760 /* We always have file-coding support */ |
428 | 4761 Fprovide (intern ("file-coding")); |
4762 | |
1347 | 4763 QScoding_system_cookie = build_string (";;;###coding system: "); |
4764 staticpro (&QScoding_system_cookie); | |
4765 | |
1242 | 4766 #ifdef HAVE_DEFAULT_EOL_DETECTION |
2297 | 4767 /* #### Find a more appropriate place for this comment. |
4768 WARNING: The existing categories are intimately tied to the function | |
1242 | 4769 `coding-system-category' in coding.el. If you change a category, or |
4770 change the layout of any coding system associated with a category, you | |
4771 need to check that function and make sure it's written properly. */ | |
4772 | |
4773 Fprovide (intern ("unix-default-eol-detection")); | |
4774 #endif | |
4775 | |
428 | 4776 DEFVAR_LISP ("keyboard-coding-system", &Vkeyboard_coding_system /* |
3142 | 4777 Default coding system used for TTY and X11 keyboard input. |
4778 Under X11, used only to interpet the character for a key event when that | |
4779 event has a KeySym of NoSymbol but does have an associated string keysym, | |
4780 something that's seen with input methods. | |
4781 | |
4782 If you need to set these things to different coding systems, call the | |
4783 function `set-console-tty-coding-system' for the TTY and use this variable | |
4784 for X11. | |
428 | 4785 */ ); |
4786 Vkeyboard_coding_system = Qnil; | |
4787 | |
4788 DEFVAR_LISP ("terminal-coding-system", &Vterminal_coding_system /* | |
4789 Coding system used for TTY display output. | |
4790 Not used under a windowing system. | |
4791 */ ); | |
4792 Vterminal_coding_system = Qnil; | |
4793 | |
4794 DEFVAR_LISP ("coding-system-for-read", &Vcoding_system_for_read /* | |
440 | 4795 Overriding coding system used when reading from a file or process. |
4796 You should bind this variable with `let', but do not set it globally. | |
4797 If this is non-nil, it specifies the coding system that will be used | |
4798 to decode input on read operations, such as from a file or process. | |
4799 It overrides `buffer-file-coding-system-for-read', | |
428 | 4800 `insert-file-contents-pre-hook', etc. Use those variables instead of |
440 | 4801 this one for permanent changes to the environment. */ ); |
428 | 4802 Vcoding_system_for_read = Qnil; |
4803 | |
4804 DEFVAR_LISP ("coding-system-for-write", | |
4805 &Vcoding_system_for_write /* | |
440 | 4806 Overriding coding system used when writing to a file or process. |
4807 You should bind this variable with `let', but do not set it globally. | |
4808 If this is non-nil, it specifies the coding system that will be used | |
4809 to encode output for write operations, such as to a file or process. | |
4810 It overrides `buffer-file-coding-system', `write-region-pre-hook', etc. | |
4811 Use those variables instead of this one for permanent changes to the | |
4812 environment. */ ); | |
428 | 4813 Vcoding_system_for_write = Qnil; |
4814 | |
4815 DEFVAR_LISP ("file-name-coding-system", &Vfile_name_coding_system /* | |
4816 Coding system used to convert pathnames when accessing files. | |
4817 */ ); | |
4818 Vfile_name_coding_system = Qnil; | |
4819 | |
4820 DEFVAR_BOOL ("enable-multibyte-characters", &enable_multibyte_characters /* | |
771 | 4821 Setting this has no effect. It is purely for FSF compatibility. |
428 | 4822 */ ); |
4823 enable_multibyte_characters = 1; | |
771 | 4824 |
4825 Vchain_canonicalize_hash_table = | |
4826 make_lisp_hash_table (50, HASH_TABLE_NON_WEAK, HASH_TABLE_EQUAL); | |
4827 staticpro (&Vchain_canonicalize_hash_table); | |
4828 | |
4829 #ifdef DEBUG_XEMACS | |
4830 DEFVAR_LISP ("debug-coding-detection", &Vdebug_coding_detection /* | |
4831 If non-nil, display debug information about detection operations in progress. | |
4832 Information is displayed on stderr. | |
4833 */ ); | |
4834 Vdebug_coding_detection = Qnil; | |
4835 #endif | |
428 | 4836 } |
4837 | |
2297 | 4838 /* #### reformat this for consistent appearance? */ |
4839 | |
428 | 4840 void |
4841 complex_vars_of_file_coding (void) | |
4842 { | |
771 | 4843 Fmake_coding_system |
4844 (Qconvert_eol_cr, Qconvert_eol, | |
4845 build_msg_string ("Convert CR to LF"), | |
4846 nconc2 (list6 (Qdocumentation, | |
4847 build_msg_string ( | |
4848 "Converts CR (used to mark the end of a line on Macintosh systems) to LF\n" | |
4849 "(used internally and under Unix to mark the end of a line)."), | |
4850 Qmnemonic, build_string ("CR->LF"), | |
4851 Qsubtype, Qcr), | |
4852 /* VERY IMPORTANT! Tell make-coding-system not to generate | |
4853 subsidiaries -- it needs the coding systems we're creating | |
4854 to do so! */ | |
4855 list2 (Qeol_type, Qlf))); | |
4856 | |
4857 Fmake_coding_system | |
4858 (Qconvert_eol_lf, Qconvert_eol, | |
4859 build_msg_string ("Convert LF to LF (do nothing)"), | |
4860 nconc2 (list6 (Qdocumentation, | |
4861 build_msg_string ( | |
4862 "Do nothing."), | |
4863 Qmnemonic, build_string ("LF->LF"), | |
4864 Qsubtype, Qlf), | |
4865 /* VERY IMPORTANT! Tell make-coding-system not to generate | |
4866 subsidiaries -- it needs the coding systems we're creating | |
4867 to do so! */ | |
4868 list2 (Qeol_type, Qlf))); | |
4869 | |
4870 Fmake_coding_system | |
4871 (Qconvert_eol_crlf, Qconvert_eol, | |
4872 build_msg_string ("Convert CRLF to LF"), | |
4873 nconc2 (list6 (Qdocumentation, | |
4874 build_msg_string ( | |
4875 "Converts CR+LF (used to mark the end of a line on Macintosh systems) to LF\n" | |
4876 "(used internally and under Unix to mark the end of a line)."), | |
4877 Qmnemonic, build_string ("CRLF->LF"), | |
4878 Qsubtype, Qcrlf), | |
4879 /* VERY IMPORTANT! Tell make-coding-system not to generate | |
4880 subsidiaries -- it needs the coding systems we're creating | |
4881 to do so! */ | |
4882 list2 (Qeol_type, Qlf))); | |
4883 | |
4884 Fmake_coding_system | |
4885 (Qconvert_eol_autodetect, Qconvert_eol, | |
4886 build_msg_string ("Autodetect EOL type"), | |
4887 nconc2 (list6 (Qdocumentation, | |
4888 build_msg_string ( | |
4889 "Autodetect the end-of-line type."), | |
4890 Qmnemonic, build_string ("Auto-EOL"), | |
793 | 4891 Qsubtype, Qnil), |
771 | 4892 /* VERY IMPORTANT! Tell make-coding-system not to generate |
4893 subsidiaries -- it needs the coding systems we're creating | |
4894 to do so! */ | |
4895 list2 (Qeol_type, Qlf))); | |
4896 | |
4897 Fmake_coding_system | |
4898 (Qundecided, Qundecided, | |
4899 build_msg_string ("Undecided (auto-detect)"), | |
4900 nconc2 (list4 (Qdocumentation, | |
4901 build_msg_string | |
4902 ("Automatically detects the correct encoding."), | |
4903 Qmnemonic, build_string ("Auto")), | |
4904 list6 (Qdo_eol, Qt, Qdo_coding, Qt, | |
4905 /* We do EOL detection ourselves so we don't need to be | |
4906 wrapped in an EOL detector. (It doesn't actually hurt, | |
4907 though, I don't think.) */ | |
4908 Qeol_type, Qlf))); | |
4909 | |
4910 Fmake_coding_system | |
4911 (intern ("undecided-dos"), Qundecided, | |
4912 build_msg_string ("Undecided (auto-detect) (CRLF)"), | |
4913 nconc2 (list4 (Qdocumentation, | |
4914 build_msg_string | |
4915 ("Automatically detects the correct encoding; EOL type of CRLF forced."), | |
4916 Qmnemonic, build_string ("Auto")), | |
4917 list4 (Qdo_coding, Qt, | |
4918 Qeol_type, Qcrlf))); | |
4919 | |
4920 Fmake_coding_system | |
4921 (intern ("undecided-unix"), Qundecided, | |
4922 build_msg_string ("Undecided (auto-detect) (LF)"), | |
4923 nconc2 (list4 (Qdocumentation, | |
4924 build_msg_string | |
4925 ("Automatically detects the correct encoding; EOL type of LF forced."), | |
4926 Qmnemonic, build_string ("Auto")), | |
4927 list4 (Qdo_coding, Qt, | |
4928 Qeol_type, Qlf))); | |
4929 | |
4930 Fmake_coding_system | |
4931 (intern ("undecided-mac"), Qundecided, | |
4932 build_msg_string ("Undecided (auto-detect) (CR)"), | |
4933 nconc2 (list4 (Qdocumentation, | |
4934 build_msg_string | |
4935 ("Automatically detects the correct encoding; EOL type of CR forced."), | |
4936 Qmnemonic, build_string ("Auto")), | |
4937 list4 (Qdo_coding, Qt, | |
4938 Qeol_type, Qcr))); | |
4939 | |
428 | 4940 /* Need to create this here or we're really screwed. */ |
4941 Fmake_coding_system | |
4942 (Qraw_text, Qno_conversion, | |
771 | 4943 build_msg_string ("Raw Text"), |
4944 list4 (Qdocumentation, | |
4945 build_msg_string ("Raw text converts only line-break codes, and acts otherwise like `binary'."), | |
4946 Qmnemonic, build_string ("Raw"))); | |
428 | 4947 |
4948 Fmake_coding_system | |
4949 (Qbinary, Qno_conversion, | |
771 | 4950 build_msg_string ("Binary"), |
4951 list6 (Qdocumentation, | |
4952 build_msg_string ( | |
4953 "This coding system is as close as it comes to doing no conversion.\n" | |
4954 "On input, each byte is converted directly into the character\n" | |
4955 "with the corresponding code -- i.e. from the `ascii', `control-1',\n" | |
4956 "or `latin-1' character sets. On output, these characters are\n" | |
4957 "converted back to the corresponding bytes, and other characters\n" | |
4958 "are converted to the default character, i.e. `~'."), | |
4959 Qeol_type, Qlf, | |
428 | 4960 Qmnemonic, build_string ("Binary"))); |
4961 | |
771 | 4962 /* Formerly aliased to raw-text! Completely bogus and not even the same |
4963 as FSF Emacs. */ | |
4964 Fdefine_coding_system_alias (Qno_conversion, Qbinary); | |
4965 Fdefine_coding_system_alias (intern ("no-conversion-unix"), | |
4966 intern ("raw-text-unix")); | |
4967 Fdefine_coding_system_alias (intern ("no-conversion-dos"), | |
4968 intern ("raw-text-dos")); | |
4969 Fdefine_coding_system_alias (intern ("no-conversion-mac"), | |
4970 intern ("raw-text-mac")); | |
4971 | |
1318 | 4972 /* These three below will get their defaults set correctly |
4973 in code-init.el. We init them now so we can handle stuff at dump | |
771 | 4974 time before we get to code-init.el. */ |
1318 | 4975 Fdefine_coding_system_alias (Qnative, Qbinary); |
440 | 4976 Fdefine_coding_system_alias (Qterminal, Qbinary); |
4977 Fdefine_coding_system_alias (Qkeyboard, Qbinary); | |
4978 | |
1318 | 4979 Fdefine_coding_system_alias (Qfile_name, Qnative); |
771 | 4980 Fdefine_coding_system_alias (Qidentity, Qconvert_eol_lf); |
4981 | |
428 | 4982 /* Need this for bootstrapping */ |
771 | 4983 coding_category_system[detector_category_no_conversion] = |
428 | 4984 Fget_coding_system (Qraw_text); |
4985 } |