1 /* Code conversion functions.
2 Copyright (C) 1991, 1995 Free Software Foundation, Inc.
3 Copyright (C) 1995 Sun Microsystems, Inc.
4 Copyright (C) 1999,2000,2001 MORIOKA Tomohiko
6 This file is part of XEmacs.
8 XEmacs is free software; you can redistribute it and/or modify it
9 under the terms of the GNU General Public License as published by the
10 Free Software Foundation; either version 2, or (at your option) any
13 XEmacs is distributed in the hope that it will be useful, but WITHOUT
14 ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
15 FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
18 You should have received a copy of the GNU General Public License
19 along with XEmacs; see the file COPYING. If not, write to
20 the Free Software Foundation, Inc., 59 Temple Place - Suite 330,
21 Boston, MA 02111-1307, USA. */
23 /* Synched up with: Mule 2.3. Not in FSF. */
25 /* Rewritten by Ben Wing <ben@xemacs.org>. */
39 #include "file-coding.h"
41 Lisp_Object Qcoding_system_error;
43 Lisp_Object Vkeyboard_coding_system;
44 Lisp_Object Vterminal_coding_system;
45 Lisp_Object Vcoding_system_for_read;
46 Lisp_Object Vcoding_system_for_write;
47 Lisp_Object Vfile_name_coding_system;
49 /* Table of symbols identifying each coding category. */
50 Lisp_Object coding_category_symbol[CODING_CATEGORY_LAST];
54 struct file_coding_dump {
55 /* Coding system currently associated with each coding category. */
56 Lisp_Object coding_category_system[CODING_CATEGORY_LAST];
58 /* Table of all coding categories in decreasing order of priority.
59 This describes a permutation of the possible coding categories. */
60 int coding_category_by_priority[CODING_CATEGORY_LAST];
62 #if defined(MULE) && !defined(UTF2000)
63 Lisp_Object ucs_to_mule_table[65536];
67 static const struct lrecord_description fcd_description_1[] = {
68 { XD_LISP_OBJECT_ARRAY, offsetof (struct file_coding_dump, coding_category_system), CODING_CATEGORY_LAST },
69 #if defined(MULE) && !defined(UTF2000)
70 { XD_LISP_OBJECT_ARRAY, offsetof (struct file_coding_dump, ucs_to_mule_table), countof (fcd->ucs_to_mule_table) },
75 static const struct struct_description fcd_description = {
76 sizeof (struct file_coding_dump),
80 Lisp_Object mule_to_ucs_table;
82 Lisp_Object Qcoding_systemp;
84 Lisp_Object Qraw_text, Qno_conversion, Qccl, Qiso2022;
85 /* Qinternal in general.c */
87 Lisp_Object Qmnemonic, Qeol_type;
88 Lisp_Object Qcr, Qcrlf, Qlf;
89 Lisp_Object Qeol_cr, Qeol_crlf, Qeol_lf;
90 Lisp_Object Qpost_read_conversion;
91 Lisp_Object Qpre_write_conversion;
94 Lisp_Object Qucs4, Qutf8;
95 Lisp_Object Qbig5, Qshift_jis;
96 Lisp_Object Qcharset_g0, Qcharset_g1, Qcharset_g2, Qcharset_g3;
97 Lisp_Object Qforce_g0_on_output, Qforce_g1_on_output;
98 Lisp_Object Qforce_g2_on_output, Qforce_g3_on_output;
99 Lisp_Object Qno_iso6429;
100 Lisp_Object Qinput_charset_conversion, Qoutput_charset_conversion;
101 Lisp_Object Qescape_quoted;
102 Lisp_Object Qshort, Qno_ascii_eol, Qno_ascii_cntl, Qseven, Qlock_shift;
105 Lisp_Object Qdisable_composition;
107 Lisp_Object Qencode, Qdecode;
109 Lisp_Object Vcoding_system_hash_table;
111 int enable_multibyte_characters;
114 /* Additional information used by the ISO2022 decoder and detector. */
115 struct iso2022_decoder
117 /* CHARSET holds the character sets currently assigned to the G0
118 through G3 variables. It is initialized from the array
119 INITIAL_CHARSET in CODESYS. */
120 Lisp_Object charset[4];
122 /* Which registers are currently invoked into the left (GL) and
123 right (GR) halves of the 8-bit encoding space? */
124 int register_left, register_right;
126 /* ISO_ESC holds a value indicating part of an escape sequence
127 that has already been seen. */
128 enum iso_esc_flag esc;
130 /* This records the bytes we've seen so far in an escape sequence,
131 in case the sequence is invalid (we spit out the bytes unchanged). */
132 unsigned char esc_bytes[8];
134 /* Index for next byte to store in ISO escape sequence. */
137 #ifdef ENABLE_COMPOSITE_CHARS
138 /* Stuff seen so far when composing a string. */
139 unsigned_char_dynarr *composite_chars;
142 /* If we saw an invalid designation sequence for a particular
143 register, we flag it here and switch to ASCII. The next time we
144 see a valid designation for this register, we turn off the flag
145 and do the designation normally, but pretend the sequence was
146 invalid. The effect of all this is that (most of the time) the
147 escape sequences for both the switch to the unknown charset, and
148 the switch back to the known charset, get inserted literally into
149 the buffer and saved out as such. The hope is that we can
150 preserve the escape sequences so that the resulting written out
151 file makes sense. If we don't do any of this, the designation
152 to the invalid charset will be preserved but that switch back
153 to the known charset will probably get eaten because it was
154 the same charset that was already present in the register. */
155 unsigned char invalid_designated[4];
157 /* We try to do similar things as above for direction-switching
158 sequences. If we encountered a direction switch while an
159 invalid designation was present, or an invalid designation
160 just after a direction switch (i.e. no valid designation
161 encountered yet), we insert the direction-switch escape
162 sequence literally into the output stream, and later on
163 insert the corresponding direction-restoring escape sequence
165 unsigned int switched_dir_and_no_valid_charset_yet :1;
166 unsigned int invalid_switch_dir :1;
168 /* Tells the decoder to output the escape sequence literally
169 even though it was valid. Used in the games we play to
170 avoid lossage when we encounter invalid designations. */
171 unsigned int output_literally :1;
172 /* We encountered a direction switch followed by an invalid
173 designation. We didn't output the direction switch
174 literally because we didn't know about the invalid designation;
175 but we have to do so now. */
176 unsigned int output_direction_sequence :1;
179 EXFUN (Fcopy_coding_system, 2);
181 struct detection_state;
184 text_encode_generic (Lstream *encoding, const Bufbyte *src,
185 unsigned_char_dynarr *dst, size_t n);
187 static int detect_coding_sjis (struct detection_state *st,
188 const Extbyte *src, size_t n);
189 static void decode_coding_sjis (Lstream *decoding, const Extbyte *src,
190 unsigned_char_dynarr *dst, size_t n);
191 void char_encode_shift_jis (struct encoding_stream *str, Emchar c,
192 unsigned_char_dynarr *dst, unsigned int *flags);
193 void char_finish_shift_jis (struct encoding_stream *str,
194 unsigned_char_dynarr *dst, unsigned int *flags);
196 static int detect_coding_big5 (struct detection_state *st,
197 const Extbyte *src, size_t n);
198 static void decode_coding_big5 (Lstream *decoding, const Extbyte *src,
199 unsigned_char_dynarr *dst, size_t n);
200 void char_encode_big5 (struct encoding_stream *str, Emchar c,
201 unsigned_char_dynarr *dst, unsigned int *flags);
202 void char_finish_big5 (struct encoding_stream *str,
203 unsigned_char_dynarr *dst, unsigned int *flags);
205 static int detect_coding_ucs4 (struct detection_state *st,
206 const Extbyte *src, size_t n);
207 static void decode_coding_ucs4 (Lstream *decoding, const Extbyte *src,
208 unsigned_char_dynarr *dst, size_t n);
209 void char_encode_ucs4 (struct encoding_stream *str, Emchar c,
210 unsigned_char_dynarr *dst, unsigned int *flags);
211 void char_finish_ucs4 (struct encoding_stream *str,
212 unsigned_char_dynarr *dst, unsigned int *flags);
214 static int detect_coding_utf8 (struct detection_state *st,
215 const Extbyte *src, size_t n);
216 static void decode_coding_utf8 (Lstream *decoding, const Extbyte *src,
217 unsigned_char_dynarr *dst, size_t n);
218 void char_encode_utf8 (struct encoding_stream *str, Emchar c,
219 unsigned_char_dynarr *dst, unsigned int *flags);
220 void char_finish_utf8 (struct encoding_stream *str,
221 unsigned_char_dynarr *dst, unsigned int *flags);
223 static int postprocess_iso2022_mask (int mask);
224 static void reset_iso2022 (Lisp_Object coding_system,
225 struct iso2022_decoder *iso);
226 static int detect_coding_iso2022 (struct detection_state *st,
227 const Extbyte *src, size_t n);
228 static void decode_coding_iso2022 (Lstream *decoding, const Extbyte *src,
229 unsigned_char_dynarr *dst, size_t n);
230 void char_encode_iso2022 (struct encoding_stream *str, Emchar c,
231 unsigned_char_dynarr *dst, unsigned int *flags);
232 void char_finish_iso2022 (struct encoding_stream *str,
233 unsigned_char_dynarr *dst, unsigned int *flags);
235 static void decode_coding_no_conversion (Lstream *decoding, const Extbyte *src,
236 unsigned_char_dynarr *dst, size_t n);
237 static void encode_coding_no_conversion (Lstream *encoding, const Bufbyte *src,
238 unsigned_char_dynarr *dst, size_t n);
239 static void mule_decode (Lstream *decoding, const Extbyte *src,
240 unsigned_char_dynarr *dst, size_t n);
241 static void mule_encode (Lstream *encoding, const Bufbyte *src,
242 unsigned_char_dynarr *dst, size_t n);
244 typedef struct codesys_prop codesys_prop;
253 Dynarr_declare (codesys_prop);
254 } codesys_prop_dynarr;
256 static const struct lrecord_description codesys_prop_description_1[] = {
257 { XD_LISP_OBJECT, offsetof (codesys_prop, sym) },
261 static const struct struct_description codesys_prop_description = {
262 sizeof (codesys_prop),
263 codesys_prop_description_1
266 static const struct lrecord_description codesys_prop_dynarr_description_1[] = {
267 XD_DYNARR_DESC (codesys_prop_dynarr, &codesys_prop_description),
271 static const struct struct_description codesys_prop_dynarr_description = {
272 sizeof (codesys_prop_dynarr),
273 codesys_prop_dynarr_description_1
276 codesys_prop_dynarr *the_codesys_prop_dynarr;
278 enum codesys_prop_enum
281 CODESYS_PROP_ISO2022,
286 /************************************************************************/
287 /* Coding system functions */
288 /************************************************************************/
290 static Lisp_Object mark_coding_system (Lisp_Object);
291 static void print_coding_system (Lisp_Object, Lisp_Object, int);
292 static void finalize_coding_system (void *header, int for_disksave);
295 static const struct lrecord_description ccs_description_1[] = {
296 { XD_LISP_OBJECT, offsetof (charset_conversion_spec, from_charset) },
297 { XD_LISP_OBJECT, offsetof (charset_conversion_spec, to_charset) },
301 static const struct struct_description ccs_description = {
302 sizeof (charset_conversion_spec),
306 static const struct lrecord_description ccsd_description_1[] = {
307 XD_DYNARR_DESC (charset_conversion_spec_dynarr, &ccs_description),
311 static const struct struct_description ccsd_description = {
312 sizeof (charset_conversion_spec_dynarr),
317 static const struct lrecord_description coding_system_description[] = {
318 { XD_LISP_OBJECT, offsetof (Lisp_Coding_System, name) },
319 { XD_LISP_OBJECT, offsetof (Lisp_Coding_System, doc_string) },
320 { XD_LISP_OBJECT, offsetof (Lisp_Coding_System, mnemonic) },
321 { XD_LISP_OBJECT, offsetof (Lisp_Coding_System, post_read_conversion) },
322 { XD_LISP_OBJECT, offsetof (Lisp_Coding_System, pre_write_conversion) },
323 { XD_LISP_OBJECT, offsetof (Lisp_Coding_System, eol_lf) },
324 { XD_LISP_OBJECT, offsetof (Lisp_Coding_System, eol_crlf) },
325 { XD_LISP_OBJECT, offsetof (Lisp_Coding_System, eol_cr) },
327 { XD_LISP_OBJECT_ARRAY, offsetof (Lisp_Coding_System, iso2022.initial_charset), 4 },
328 { XD_STRUCT_PTR, offsetof (Lisp_Coding_System, iso2022.input_conv), 1, &ccsd_description },
329 { XD_STRUCT_PTR, offsetof (Lisp_Coding_System, iso2022.output_conv), 1, &ccsd_description },
330 { XD_LISP_OBJECT, offsetof (Lisp_Coding_System, ccl.decode) },
331 { XD_LISP_OBJECT, offsetof (Lisp_Coding_System, ccl.encode) },
336 DEFINE_LRECORD_IMPLEMENTATION ("coding-system", coding_system,
337 mark_coding_system, print_coding_system,
338 finalize_coding_system,
339 0, 0, coding_system_description,
343 mark_coding_system (Lisp_Object obj)
345 Lisp_Coding_System *codesys = XCODING_SYSTEM (obj);
347 mark_object (CODING_SYSTEM_NAME (codesys));
348 mark_object (CODING_SYSTEM_DOC_STRING (codesys));
349 mark_object (CODING_SYSTEM_MNEMONIC (codesys));
350 mark_object (CODING_SYSTEM_EOL_LF (codesys));
351 mark_object (CODING_SYSTEM_EOL_CRLF (codesys));
352 mark_object (CODING_SYSTEM_EOL_CR (codesys));
354 switch (CODING_SYSTEM_TYPE (codesys))
358 case CODESYS_ISO2022:
359 for (i = 0; i < 4; i++)
360 mark_object (CODING_SYSTEM_ISO2022_INITIAL_CHARSET (codesys, i));
361 if (codesys->iso2022.input_conv)
363 for (i = 0; i < Dynarr_length (codesys->iso2022.input_conv); i++)
365 struct charset_conversion_spec *ccs =
366 Dynarr_atp (codesys->iso2022.input_conv, i);
367 mark_object (ccs->from_charset);
368 mark_object (ccs->to_charset);
371 if (codesys->iso2022.output_conv)
373 for (i = 0; i < Dynarr_length (codesys->iso2022.output_conv); i++)
375 struct charset_conversion_spec *ccs =
376 Dynarr_atp (codesys->iso2022.output_conv, i);
377 mark_object (ccs->from_charset);
378 mark_object (ccs->to_charset);
384 mark_object (CODING_SYSTEM_CCL_DECODE (codesys));
385 mark_object (CODING_SYSTEM_CCL_ENCODE (codesys));
392 mark_object (CODING_SYSTEM_PRE_WRITE_CONVERSION (codesys));
393 return CODING_SYSTEM_POST_READ_CONVERSION (codesys);
397 print_coding_system (Lisp_Object obj, Lisp_Object printcharfun,
400 Lisp_Coding_System *c = XCODING_SYSTEM (obj);
402 error ("printing unreadable object #<coding_system 0x%x>",
405 write_c_string ("#<coding_system ", printcharfun);
406 print_internal (c->name, printcharfun, 1);
407 write_c_string (">", printcharfun);
411 finalize_coding_system (void *header, int for_disksave)
413 Lisp_Coding_System *c = (Lisp_Coding_System *) header;
414 /* Since coding systems never go away, this function is not
415 necessary. But it would be necessary if we changed things
416 so that coding systems could go away. */
417 if (!for_disksave) /* see comment in lstream.c */
419 switch (CODING_SYSTEM_TYPE (c))
422 case CODESYS_ISO2022:
423 if (c->iso2022.input_conv)
425 Dynarr_free (c->iso2022.input_conv);
426 c->iso2022.input_conv = 0;
428 if (c->iso2022.output_conv)
430 Dynarr_free (c->iso2022.output_conv);
431 c->iso2022.output_conv = 0;
442 symbol_to_eol_type (Lisp_Object symbol)
444 CHECK_SYMBOL (symbol);
445 if (NILP (symbol)) return EOL_AUTODETECT;
446 if (EQ (symbol, Qlf)) return EOL_LF;
447 if (EQ (symbol, Qcrlf)) return EOL_CRLF;
448 if (EQ (symbol, Qcr)) return EOL_CR;
450 signal_simple_error ("Unrecognized eol type", symbol);
451 return EOL_AUTODETECT; /* not reached */
455 eol_type_to_symbol (eol_type_t type)
460 case EOL_LF: return Qlf;
461 case EOL_CRLF: return Qcrlf;
462 case EOL_CR: return Qcr;
463 case EOL_AUTODETECT: return Qnil;
468 setup_eol_coding_systems (Lisp_Coding_System *codesys)
470 Lisp_Object codesys_obj;
471 int len = string_length (XSYMBOL (CODING_SYSTEM_NAME (codesys))->name);
472 char *codesys_name = (char *) alloca (len + 7);
474 char *codesys_mnemonic=0;
476 Lisp_Object codesys_name_sym, sub_codesys_obj;
480 XSETCODING_SYSTEM (codesys_obj, codesys);
482 memcpy (codesys_name,
483 string_data (XSYMBOL (CODING_SYSTEM_NAME (codesys))->name), len);
485 if (STRINGP (CODING_SYSTEM_MNEMONIC (codesys)))
487 mlen = XSTRING_LENGTH (CODING_SYSTEM_MNEMONIC (codesys));
488 codesys_mnemonic = (char *) alloca (mlen + 7);
489 memcpy (codesys_mnemonic,
490 XSTRING_DATA (CODING_SYSTEM_MNEMONIC (codesys)), mlen);
493 #define DEFINE_SUB_CODESYS(op_sys, op_sys_abbr, Type) do { \
494 strcpy (codesys_name + len, "-" op_sys); \
496 strcpy (codesys_mnemonic + mlen, op_sys_abbr); \
497 codesys_name_sym = intern (codesys_name); \
498 sub_codesys_obj = Fcopy_coding_system (codesys_obj, codesys_name_sym); \
499 XCODING_SYSTEM_EOL_TYPE (sub_codesys_obj) = Type; \
501 XCODING_SYSTEM_MNEMONIC(sub_codesys_obj) = \
502 build_string (codesys_mnemonic); \
503 CODING_SYSTEM_##Type (codesys) = sub_codesys_obj; \
506 DEFINE_SUB_CODESYS("unix", "", EOL_LF);
507 DEFINE_SUB_CODESYS("dos", ":T", EOL_CRLF);
508 DEFINE_SUB_CODESYS("mac", ":t", EOL_CR);
511 DEFUN ("coding-system-p", Fcoding_system_p, 1, 1, 0, /*
512 Return t if OBJECT is a coding system.
513 A coding system is an object that defines how text containing multiple
514 character sets is encoded into a stream of (typically 8-bit) bytes.
515 The coding system is used to decode the stream into a series of
516 characters (which may be from multiple charsets) when the text is read
517 from a file or process, and is used to encode the text back into the
518 same format when it is written out to a file or process.
520 For example, many ISO2022-compliant coding systems (such as Compound
521 Text, which is used for inter-client data under the X Window System)
522 use escape sequences to switch between different charsets -- Japanese
523 Kanji, for example, is invoked with "ESC $ ( B"; ASCII is invoked
524 with "ESC ( B"; and Cyrillic is invoked with "ESC - L". See
525 `make-coding-system' for more information.
527 Coding systems are normally identified using a symbol, and the
528 symbol is accepted in place of the actual coding system object whenever
529 a coding system is called for. (This is similar to how faces work.)
533 return CODING_SYSTEMP (object) ? Qt : Qnil;
536 DEFUN ("find-coding-system", Ffind_coding_system, 1, 1, 0, /*
537 Retrieve the coding system of the given name.
539 If CODING-SYSTEM-OR-NAME is a coding-system object, it is simply
540 returned. Otherwise, CODING-SYSTEM-OR-NAME should be a symbol.
541 If there is no such coding system, nil is returned. Otherwise the
542 associated coding system object is returned.
544 (coding_system_or_name))
546 if (NILP (coding_system_or_name))
547 coding_system_or_name = Qbinary;
548 else if (CODING_SYSTEMP (coding_system_or_name))
549 return coding_system_or_name;
551 CHECK_SYMBOL (coding_system_or_name);
555 coding_system_or_name =
556 Fgethash (coding_system_or_name, Vcoding_system_hash_table, Qnil);
558 if (CODING_SYSTEMP (coding_system_or_name) || NILP (coding_system_or_name))
559 return coding_system_or_name;
563 DEFUN ("get-coding-system", Fget_coding_system, 1, 1, 0, /*
564 Retrieve the coding system of the given name.
565 Same as `find-coding-system' except that if there is no such
566 coding system, an error is signaled instead of returning nil.
570 Lisp_Object coding_system = Ffind_coding_system (name);
572 if (NILP (coding_system))
573 signal_simple_error ("No such coding system", name);
574 return coding_system;
577 /* We store the coding systems in hash tables with the names as the key and the
578 actual coding system object as the value. Occasionally we need to use them
579 in a list format. These routines provide us with that. */
580 struct coding_system_list_closure
582 Lisp_Object *coding_system_list;
586 add_coding_system_to_list_mapper (Lisp_Object key, Lisp_Object value,
587 void *coding_system_list_closure)
589 /* This function can GC */
590 struct coding_system_list_closure *cscl =
591 (struct coding_system_list_closure *) coding_system_list_closure;
592 Lisp_Object *coding_system_list = cscl->coding_system_list;
594 *coding_system_list = Fcons (key, *coding_system_list);
598 DEFUN ("coding-system-list", Fcoding_system_list, 0, 0, 0, /*
599 Return a list of the names of all defined coding systems.
603 Lisp_Object coding_system_list = Qnil;
605 struct coding_system_list_closure coding_system_list_closure;
607 GCPRO1 (coding_system_list);
608 coding_system_list_closure.coding_system_list = &coding_system_list;
609 elisp_maphash (add_coding_system_to_list_mapper, Vcoding_system_hash_table,
610 &coding_system_list_closure);
613 return coding_system_list;
616 DEFUN ("coding-system-name", Fcoding_system_name, 1, 1, 0, /*
617 Return the name of the given coding system.
621 coding_system = Fget_coding_system (coding_system);
622 return XCODING_SYSTEM_NAME (coding_system);
625 static Lisp_Coding_System *
626 allocate_coding_system (enum coding_system_type type, Lisp_Object name)
628 Lisp_Coding_System *codesys =
629 alloc_lcrecord_type (Lisp_Coding_System, &lrecord_coding_system);
631 zero_lcrecord (codesys);
632 CODING_SYSTEM_PRE_WRITE_CONVERSION (codesys) = Qnil;
633 CODING_SYSTEM_POST_READ_CONVERSION (codesys) = Qnil;
634 CODING_SYSTEM_EOL_TYPE (codesys) = EOL_AUTODETECT;
635 CODING_SYSTEM_EOL_CRLF (codesys) = Qnil;
636 CODING_SYSTEM_EOL_CR (codesys) = Qnil;
637 CODING_SYSTEM_EOL_LF (codesys) = Qnil;
638 CODING_SYSTEM_TYPE (codesys) = type;
639 CODING_SYSTEM_MNEMONIC (codesys) = Qnil;
641 if (type == CODESYS_ISO2022)
644 for (i = 0; i < 4; i++)
645 CODING_SYSTEM_ISO2022_INITIAL_CHARSET (codesys, i) = Qnil;
647 else if (type == CODESYS_CCL)
649 CODING_SYSTEM_CCL_DECODE (codesys) = Qnil;
650 CODING_SYSTEM_CCL_ENCODE (codesys) = Qnil;
653 CODING_SYSTEM_NAME (codesys) = name;
659 /* Given a list of charset conversion specs as specified in a Lisp
660 program, parse it into STORE_HERE. */
663 parse_charset_conversion_specs (charset_conversion_spec_dynarr *store_here,
664 Lisp_Object spec_list)
668 EXTERNAL_LIST_LOOP (rest, spec_list)
670 Lisp_Object car = XCAR (rest);
671 Lisp_Object from, to;
672 struct charset_conversion_spec spec;
674 if (!CONSP (car) || !CONSP (XCDR (car)) || !NILP (XCDR (XCDR (car))))
675 signal_simple_error ("Invalid charset conversion spec", car);
676 from = Fget_charset (XCAR (car));
677 to = Fget_charset (XCAR (XCDR (car)));
678 if ( (XCHARSET_CHARS (from) != XCHARSET_CHARS (to)) ||
679 (XCHARSET_DIMENSION (from) != XCHARSET_DIMENSION (to)) )
680 signal_simple_error_2
681 ("Attempted conversion between different charset types",
683 spec.from_charset = from;
684 spec.to_charset = to;
686 Dynarr_add (store_here, spec);
690 /* Given a dynarr LOAD_HERE of internally-stored charset conversion
691 specs, return the equivalent as the Lisp programmer would see it.
693 If LOAD_HERE is 0, return Qnil. */
696 unparse_charset_conversion_specs (charset_conversion_spec_dynarr *load_here)
703 for (i = 0, result = Qnil; i < Dynarr_length (load_here); i++)
705 struct charset_conversion_spec *ccs = Dynarr_atp (load_here, i);
706 result = Fcons (list2 (ccs->from_charset, ccs->to_charset), result);
709 return Fnreverse (result);
714 DEFUN ("make-coding-system", Fmake_coding_system, 2, 4, 0, /*
715 Register symbol NAME as a coding system.
717 TYPE describes the conversion method used and should be one of
720 Automatic conversion. XEmacs attempts to detect the coding system
723 No conversion. Use this for binary files and such. On output,
724 graphic characters that are not in ASCII or Latin-1 will be
725 replaced by a ?. (For a no-conversion-encoded buffer, these
726 characters will only be present if you explicitly insert them.)
728 Shift-JIS (a Japanese encoding commonly used in PC operating systems).
730 ISO 10646 UCS-4 encoding.
732 ISO 10646 UTF-8 encoding.
734 Any ISO2022-compliant encoding. Among other things, this includes
735 JIS (the Japanese encoding commonly used for e-mail), EUC (the
736 standard Unix encoding for Japanese and other languages), and
737 Compound Text (the encoding used in X11). You can specify more
738 specific information about the conversion with the PROPS argument.
740 Big5 (the encoding commonly used for Taiwanese).
742 The conversion is performed using a user-written pseudo-code
743 program. CCL (Code Conversion Language) is the name of this
746 Write out or read in the raw contents of the memory representing
747 the buffer's text. This is primarily useful for debugging
748 purposes, and is only enabled when XEmacs has been compiled with
749 DEBUG_XEMACS defined (via the --debug configure option).
750 WARNING: Reading in a file using 'internal conversion can result
751 in an internal inconsistency in the memory representing a
752 buffer's text, which will produce unpredictable results and may
753 cause XEmacs to crash. Under normal circumstances you should
754 never use 'internal conversion.
756 DOC-STRING is a string describing the coding system.
758 PROPS is a property list, describing the specific nature of the
759 character set. Recognized properties are:
762 String to be displayed in the modeline when this coding system is
766 End-of-line conversion to be used. It should be one of
769 Automatically detect the end-of-line type (LF, CRLF,
770 or CR). Also generate subsidiary coding systems named
771 `NAME-unix', `NAME-dos', and `NAME-mac', that are
772 identical to this coding system but have an EOL-TYPE
773 value of 'lf, 'crlf, and 'cr, respectively.
775 The end of a line is marked externally using ASCII LF.
776 Since this is also the way that XEmacs represents an
777 end-of-line internally, specifying this option results
778 in no end-of-line conversion. This is the standard
779 format for Unix text files.
781 The end of a line is marked externally using ASCII
782 CRLF. This is the standard format for MS-DOS text
785 The end of a line is marked externally using ASCII CR.
786 This is the standard format for Macintosh text files.
788 Automatically detect the end-of-line type but do not
789 generate subsidiary coding systems. (This value is
790 converted to nil when stored internally, and
791 `coding-system-property' will return nil.)
794 If non-nil, composition/decomposition for combining characters
797 'post-read-conversion
798 Function called after a file has been read in, to perform the
799 decoding. Called with two arguments, START and END, denoting
800 a region of the current buffer to be decoded.
802 'pre-write-conversion
803 Function called before a file is written out, to perform the
804 encoding. Called with two arguments, START and END, denoting
805 a region of the current buffer to be encoded.
808 The following additional properties are recognized if TYPE is 'iso2022:
814 The character set initially designated to the G0 - G3 registers.
815 The value should be one of
817 -- A charset object (designate that character set)
818 -- nil (do not ever use this register)
819 -- t (no character set is initially designated to
820 the register, but may be later on; this automatically
821 sets the corresponding `force-g*-on-output' property)
827 If non-nil, send an explicit designation sequence on output before
828 using the specified register.
831 If non-nil, use the short forms "ESC $ @", "ESC $ A", and
832 "ESC $ B" on output in place of the full designation sequences
833 "ESC $ ( @", "ESC $ ( A", and "ESC $ ( B".
836 If non-nil, don't designate ASCII to G0 at each end of line on output.
837 Setting this to non-nil also suppresses other state-resetting that
838 normally happens at the end of a line.
841 If non-nil, don't designate ASCII to G0 before control chars on output.
844 If non-nil, use 7-bit environment on output. Otherwise, use 8-bit
848 If non-nil, use locking-shift (SO/SI) instead of single-shift
849 or designation by escape sequence.
852 If non-nil, don't use ISO6429's direction specification.
855 If non-nil, literal control characters that are the same as
856 the beginning of a recognized ISO2022 or ISO6429 escape sequence
857 (in particular, ESC (0x1B), SO (0x0E), SI (0x0F), SS2 (0x8E),
858 SS3 (0x8F), and CSI (0x9B)) are "quoted" with an escape character
859 so that they can be properly distinguished from an escape sequence.
860 (Note that doing this results in a non-portable encoding.) This
861 encoding flag is used for byte-compiled files. Note that ESC
862 is a good choice for a quoting character because there are no
863 escape sequences whose second byte is a character from the Control-0
864 or Control-1 character sets; this is explicitly disallowed by the
867 'input-charset-conversion
868 A list of conversion specifications, specifying conversion of
869 characters in one charset to another when decoding is performed.
870 Each specification is a list of two elements: the source charset,
871 and the destination charset.
873 'output-charset-conversion
874 A list of conversion specifications, specifying conversion of
875 characters in one charset to another when encoding is performed.
876 The form of each specification is the same as for
877 'input-charset-conversion.
880 The following additional properties are recognized (and required)
884 CCL program used for decoding (converting to internal format).
887 CCL program used for encoding (converting to external format).
889 (name, type, doc_string, props))
891 Lisp_Coding_System *codesys;
892 enum coding_system_type ty;
893 int need_to_setup_eol_systems = 1;
895 /* Convert type to constant */
896 if (NILP (type) || EQ (type, Qundecided))
897 { ty = CODESYS_AUTODETECT; }
899 else if (EQ (type, Qshift_jis)) { ty = CODESYS_SHIFT_JIS; }
900 else if (EQ (type, Qiso2022)) { ty = CODESYS_ISO2022; }
901 else if (EQ (type, Qbig5)) { ty = CODESYS_BIG5; }
902 else if (EQ (type, Qucs4)) { ty = CODESYS_UCS4; }
903 else if (EQ (type, Qutf8)) { ty = CODESYS_UTF8; }
904 else if (EQ (type, Qccl)) { ty = CODESYS_CCL; }
906 else if (EQ (type, Qno_conversion)) { ty = CODESYS_NO_CONVERSION; }
908 else if (EQ (type, Qinternal)) { ty = CODESYS_INTERNAL; }
911 signal_simple_error ("Invalid coding system type", type);
915 codesys = allocate_coding_system (ty, name);
917 if (NILP (doc_string))
918 doc_string = build_string ("");
920 CHECK_STRING (doc_string);
921 CODING_SYSTEM_DOC_STRING (codesys) = doc_string;
924 EXTERNAL_PROPERTY_LIST_LOOP_3 (key, value, props)
926 if (EQ (key, Qmnemonic))
929 CHECK_STRING (value);
930 CODING_SYSTEM_MNEMONIC (codesys) = value;
933 else if (EQ (key, Qeol_type))
935 need_to_setup_eol_systems = NILP (value);
938 CODING_SYSTEM_EOL_TYPE (codesys) = symbol_to_eol_type (value);
941 else if (EQ (key, Qpost_read_conversion))
942 CODING_SYSTEM_POST_READ_CONVERSION (codesys) = value;
943 else if (EQ (key, Qpre_write_conversion))
944 CODING_SYSTEM_PRE_WRITE_CONVERSION (codesys) = value;
946 else if (EQ (key, Qdisable_composition))
947 CODING_SYSTEM_DISABLE_COMPOSITION (codesys) = !NILP (value);
950 else if (ty == CODESYS_ISO2022)
952 #define FROB_INITIAL_CHARSET(charset_num) \
953 CODING_SYSTEM_ISO2022_INITIAL_CHARSET (codesys, charset_num) = \
954 ((EQ (value, Qt) || EQ (value, Qnil)) ? value : Fget_charset (value))
956 if (EQ (key, Qcharset_g0)) FROB_INITIAL_CHARSET (0);
957 else if (EQ (key, Qcharset_g1)) FROB_INITIAL_CHARSET (1);
958 else if (EQ (key, Qcharset_g2)) FROB_INITIAL_CHARSET (2);
959 else if (EQ (key, Qcharset_g3)) FROB_INITIAL_CHARSET (3);
961 #define FROB_FORCE_CHARSET(charset_num) \
962 CODING_SYSTEM_ISO2022_FORCE_CHARSET_ON_OUTPUT (codesys, charset_num) = !NILP (value)
964 else if (EQ (key, Qforce_g0_on_output)) FROB_FORCE_CHARSET (0);
965 else if (EQ (key, Qforce_g1_on_output)) FROB_FORCE_CHARSET (1);
966 else if (EQ (key, Qforce_g2_on_output)) FROB_FORCE_CHARSET (2);
967 else if (EQ (key, Qforce_g3_on_output)) FROB_FORCE_CHARSET (3);
969 #define FROB_BOOLEAN_PROPERTY(prop) \
970 CODING_SYSTEM_ISO2022_##prop (codesys) = !NILP (value)
972 else if (EQ (key, Qshort)) FROB_BOOLEAN_PROPERTY (SHORT);
973 else if (EQ (key, Qno_ascii_eol)) FROB_BOOLEAN_PROPERTY (NO_ASCII_EOL);
974 else if (EQ (key, Qno_ascii_cntl)) FROB_BOOLEAN_PROPERTY (NO_ASCII_CNTL);
975 else if (EQ (key, Qseven)) FROB_BOOLEAN_PROPERTY (SEVEN);
976 else if (EQ (key, Qlock_shift)) FROB_BOOLEAN_PROPERTY (LOCK_SHIFT);
977 else if (EQ (key, Qno_iso6429)) FROB_BOOLEAN_PROPERTY (NO_ISO6429);
978 else if (EQ (key, Qescape_quoted)) FROB_BOOLEAN_PROPERTY (ESCAPE_QUOTED);
980 else if (EQ (key, Qinput_charset_conversion))
982 codesys->iso2022.input_conv =
983 Dynarr_new (charset_conversion_spec);
984 parse_charset_conversion_specs (codesys->iso2022.input_conv,
987 else if (EQ (key, Qoutput_charset_conversion))
989 codesys->iso2022.output_conv =
990 Dynarr_new (charset_conversion_spec);
991 parse_charset_conversion_specs (codesys->iso2022.output_conv,
995 signal_simple_error ("Unrecognized property", key);
997 else if (EQ (type, Qccl))
1000 struct ccl_program test_ccl;
1003 /* Check key first. */
1004 if (EQ (key, Qdecode))
1005 suffix = "-ccl-decode";
1006 else if (EQ (key, Qencode))
1007 suffix = "-ccl-encode";
1009 signal_simple_error ("Unrecognized property", key);
1011 /* If value is vector, register it as a ccl program
1012 associated with an newly created symbol for
1013 backward compatibility. */
1014 if (VECTORP (value))
1016 sym = Fintern (concat2 (Fsymbol_name (name),
1017 build_string (suffix)),
1019 Fregister_ccl_program (sym, value);
1023 CHECK_SYMBOL (value);
1026 /* check if the given ccl programs are valid. */
1027 if (setup_ccl_program (&test_ccl, sym) < 0)
1028 signal_simple_error ("Invalid CCL program", value);
1030 if (EQ (key, Qdecode))
1031 CODING_SYSTEM_CCL_DECODE (codesys) = sym;
1032 else if (EQ (key, Qencode))
1033 CODING_SYSTEM_CCL_ENCODE (codesys) = sym;
1038 signal_simple_error ("Unrecognized property", key);
1042 if (need_to_setup_eol_systems)
1043 setup_eol_coding_systems (codesys);
1046 Lisp_Object codesys_obj;
1047 XSETCODING_SYSTEM (codesys_obj, codesys);
1048 Fputhash (name, codesys_obj, Vcoding_system_hash_table);
1053 DEFUN ("copy-coding-system", Fcopy_coding_system, 2, 2, 0, /*
1054 Copy OLD-CODING-SYSTEM to NEW-NAME.
1055 If NEW-NAME does not name an existing coding system, a new one will
1058 (old_coding_system, new_name))
1060 Lisp_Object new_coding_system;
1061 old_coding_system = Fget_coding_system (old_coding_system);
1062 new_coding_system = Ffind_coding_system (new_name);
1063 if (NILP (new_coding_system))
1065 XSETCODING_SYSTEM (new_coding_system,
1066 allocate_coding_system
1067 (XCODING_SYSTEM_TYPE (old_coding_system),
1069 Fputhash (new_name, new_coding_system, Vcoding_system_hash_table);
1073 Lisp_Coding_System *to = XCODING_SYSTEM (new_coding_system);
1074 Lisp_Coding_System *from = XCODING_SYSTEM (old_coding_system);
1075 memcpy (((char *) to ) + sizeof (to->header),
1076 ((char *) from) + sizeof (from->header),
1077 sizeof (*from) - sizeof (from->header));
1078 to->name = new_name;
1080 return new_coding_system;
1083 DEFUN ("coding-system-canonical-name-p", Fcoding_system_canonical_name_p, 1, 1, 0, /*
1084 Return t if OBJECT names a coding system, and is not a coding system alias.
1088 return CODING_SYSTEMP (Fgethash (object, Vcoding_system_hash_table, Qnil))
1092 DEFUN ("coding-system-alias-p", Fcoding_system_alias_p, 1, 1, 0, /*
1093 Return t if OBJECT is a coding system alias.
1094 All coding system aliases are created by `define-coding-system-alias'.
1098 return SYMBOLP (Fgethash (object, Vcoding_system_hash_table, Qzero))
1102 DEFUN ("coding-system-aliasee", Fcoding_system_aliasee, 1, 1, 0, /*
1103 Return the coding-system symbol for which symbol ALIAS is an alias.
1107 Lisp_Object aliasee = Fgethash (alias, Vcoding_system_hash_table, Qnil);
1108 if (SYMBOLP (aliasee))
1111 signal_simple_error ("Symbol is not a coding system alias", alias);
1112 return Qnil; /* To keep the compiler happy */
1116 append_suffix_to_symbol (Lisp_Object symbol, const char *ascii_string)
1118 return Fintern (concat2 (Fsymbol_name (symbol), build_string (ascii_string)),
1122 /* A maphash function, for removing dangling coding system aliases. */
1124 dangling_coding_system_alias_p (Lisp_Object alias,
1125 Lisp_Object aliasee,
1126 void *dangling_aliases)
1128 if (SYMBOLP (aliasee)
1129 && NILP (Fgethash (aliasee, Vcoding_system_hash_table, Qnil)))
1131 (*(int *) dangling_aliases)++;
1138 DEFUN ("define-coding-system-alias", Fdefine_coding_system_alias, 2, 2, 0, /*
1139 Define symbol ALIAS as an alias for coding system ALIASEE.
1141 You can use this function to redefine an alias that has already been defined,
1142 but you cannot redefine a name which is the canonical name for a coding system.
1143 \(a canonical name of a coding system is what is returned when you call
1144 `coding-system-name' on a coding system).
1146 ALIASEE itself can be an alias, which allows you to define nested aliases.
1148 You are forbidden, however, from creating alias loops or `dangling' aliases.
1149 These will be detected, and an error will be signaled if you attempt to do so.
1151 If ALIASEE is nil, then ALIAS will simply be undefined.
1153 See also `coding-system-alias-p', `coding-system-aliasee',
1154 and `coding-system-canonical-name-p'.
1158 Lisp_Object real_coding_system, probe;
1160 CHECK_SYMBOL (alias);
1162 if (!NILP (Fcoding_system_canonical_name_p (alias)))
1164 ("Symbol is the canonical name of a coding system and cannot be redefined",
1169 Lisp_Object subsidiary_unix = append_suffix_to_symbol (alias, "-unix");
1170 Lisp_Object subsidiary_dos = append_suffix_to_symbol (alias, "-dos");
1171 Lisp_Object subsidiary_mac = append_suffix_to_symbol (alias, "-mac");
1173 Fremhash (alias, Vcoding_system_hash_table);
1175 /* Undefine subsidiary aliases,
1176 presumably created by a previous call to this function */
1177 if (! NILP (Fcoding_system_alias_p (subsidiary_unix)) &&
1178 ! NILP (Fcoding_system_alias_p (subsidiary_dos)) &&
1179 ! NILP (Fcoding_system_alias_p (subsidiary_mac)))
1181 Fdefine_coding_system_alias (subsidiary_unix, Qnil);
1182 Fdefine_coding_system_alias (subsidiary_dos, Qnil);
1183 Fdefine_coding_system_alias (subsidiary_mac, Qnil);
1186 /* Undefine dangling coding system aliases. */
1188 int dangling_aliases;
1191 dangling_aliases = 0;
1192 elisp_map_remhash (dangling_coding_system_alias_p,
1193 Vcoding_system_hash_table,
1195 } while (dangling_aliases > 0);
1201 if (CODING_SYSTEMP (aliasee))
1202 aliasee = XCODING_SYSTEM_NAME (aliasee);
1204 /* Checks that aliasee names a coding-system */
1205 real_coding_system = Fget_coding_system (aliasee);
1207 /* Check for coding system alias loops */
1208 if (EQ (alias, aliasee))
1209 alias_loop: signal_simple_error_2
1210 ("Attempt to create a coding system alias loop", alias, aliasee);
1212 for (probe = aliasee;
1214 probe = Fgethash (probe, Vcoding_system_hash_table, Qzero))
1216 if (EQ (probe, alias))
1220 Fputhash (alias, aliasee, Vcoding_system_hash_table);
1222 /* Set up aliases for subsidiaries.
1223 #### There must be a better way to handle subsidiary coding systems. */
1225 static const char *suffixes[] = { "-unix", "-dos", "-mac" };
1227 for (i = 0; i < countof (suffixes); i++)
1229 Lisp_Object alias_subsidiary =
1230 append_suffix_to_symbol (alias, suffixes[i]);
1231 Lisp_Object aliasee_subsidiary =
1232 append_suffix_to_symbol (aliasee, suffixes[i]);
1234 if (! NILP (Ffind_coding_system (aliasee_subsidiary)))
1235 Fdefine_coding_system_alias (alias_subsidiary, aliasee_subsidiary);
1238 /* FSF return value is a vector of [ALIAS-unix ALIAS-dos ALIAS-mac],
1239 but it doesn't look intentional, so I'd rather return something
1240 meaningful or nothing at all. */
1245 subsidiary_coding_system (Lisp_Object coding_system, eol_type_t type)
1247 Lisp_Coding_System *cs = XCODING_SYSTEM (coding_system);
1248 Lisp_Object new_coding_system;
1250 if (CODING_SYSTEM_EOL_TYPE (cs) != EOL_AUTODETECT)
1251 return coding_system;
1255 case EOL_AUTODETECT: return coding_system;
1256 case EOL_LF: new_coding_system = CODING_SYSTEM_EOL_LF (cs); break;
1257 case EOL_CR: new_coding_system = CODING_SYSTEM_EOL_CR (cs); break;
1258 case EOL_CRLF: new_coding_system = CODING_SYSTEM_EOL_CRLF (cs); break;
1259 default: abort (); return Qnil;
1262 return NILP (new_coding_system) ? coding_system : new_coding_system;
1265 DEFUN ("subsidiary-coding-system", Fsubsidiary_coding_system, 2, 2, 0, /*
1266 Return the subsidiary coding system of CODING-SYSTEM with eol type EOL-TYPE.
1268 (coding_system, eol_type))
1270 coding_system = Fget_coding_system (coding_system);
1272 return subsidiary_coding_system (coding_system,
1273 symbol_to_eol_type (eol_type));
1277 /************************************************************************/
1278 /* Coding system accessors */
1279 /************************************************************************/
1281 DEFUN ("coding-system-doc-string", Fcoding_system_doc_string, 1, 1, 0, /*
1282 Return the doc string for CODING-SYSTEM.
1286 coding_system = Fget_coding_system (coding_system);
1287 return XCODING_SYSTEM_DOC_STRING (coding_system);
1290 DEFUN ("coding-system-type", Fcoding_system_type, 1, 1, 0, /*
1291 Return the type of CODING-SYSTEM.
1295 switch (XCODING_SYSTEM_TYPE (Fget_coding_system (coding_system)))
1298 case CODESYS_AUTODETECT: return Qundecided;
1300 case CODESYS_SHIFT_JIS: return Qshift_jis;
1301 case CODESYS_ISO2022: return Qiso2022;
1302 case CODESYS_BIG5: return Qbig5;
1303 case CODESYS_UCS4: return Qucs4;
1304 case CODESYS_UTF8: return Qutf8;
1305 case CODESYS_CCL: return Qccl;
1307 case CODESYS_NO_CONVERSION: return Qno_conversion;
1309 case CODESYS_INTERNAL: return Qinternal;
1316 Lisp_Object coding_system_charset (Lisp_Object coding_system, int gnum)
1319 = XCODING_SYSTEM_ISO2022_INITIAL_CHARSET (coding_system, gnum);
1321 return CHARSETP (cs) ? XCHARSET_NAME (cs) : Qnil;
1324 DEFUN ("coding-system-charset", Fcoding_system_charset, 2, 2, 0, /*
1325 Return initial charset of CODING-SYSTEM designated to GNUM.
1328 (coding_system, gnum))
1330 coding_system = Fget_coding_system (coding_system);
1333 return coding_system_charset (coding_system, XINT (gnum));
1337 DEFUN ("coding-system-property", Fcoding_system_property, 2, 2, 0, /*
1338 Return the PROP property of CODING-SYSTEM.
1340 (coding_system, prop))
1343 enum coding_system_type type;
1345 coding_system = Fget_coding_system (coding_system);
1346 CHECK_SYMBOL (prop);
1347 type = XCODING_SYSTEM_TYPE (coding_system);
1349 for (i = 0; !ok && i < Dynarr_length (the_codesys_prop_dynarr); i++)
1350 if (EQ (Dynarr_at (the_codesys_prop_dynarr, i).sym, prop))
1353 switch (Dynarr_at (the_codesys_prop_dynarr, i).prop_type)
1355 case CODESYS_PROP_ALL_OK:
1358 case CODESYS_PROP_ISO2022:
1359 if (type != CODESYS_ISO2022)
1361 ("Property only valid in ISO2022 coding systems",
1365 case CODESYS_PROP_CCL:
1366 if (type != CODESYS_CCL)
1368 ("Property only valid in CCL coding systems",
1378 signal_simple_error ("Unrecognized property", prop);
1380 if (EQ (prop, Qname))
1381 return XCODING_SYSTEM_NAME (coding_system);
1382 else if (EQ (prop, Qtype))
1383 return Fcoding_system_type (coding_system);
1384 else if (EQ (prop, Qdoc_string))
1385 return XCODING_SYSTEM_DOC_STRING (coding_system);
1386 else if (EQ (prop, Qmnemonic))
1387 return XCODING_SYSTEM_MNEMONIC (coding_system);
1388 else if (EQ (prop, Qeol_type))
1389 return eol_type_to_symbol (XCODING_SYSTEM_EOL_TYPE (coding_system));
1390 else if (EQ (prop, Qeol_lf))
1391 return XCODING_SYSTEM_EOL_LF (coding_system);
1392 else if (EQ (prop, Qeol_crlf))
1393 return XCODING_SYSTEM_EOL_CRLF (coding_system);
1394 else if (EQ (prop, Qeol_cr))
1395 return XCODING_SYSTEM_EOL_CR (coding_system);
1396 else if (EQ (prop, Qpost_read_conversion))
1397 return XCODING_SYSTEM_POST_READ_CONVERSION (coding_system);
1398 else if (EQ (prop, Qpre_write_conversion))
1399 return XCODING_SYSTEM_PRE_WRITE_CONVERSION (coding_system);
1401 else if (type == CODESYS_ISO2022)
1403 if (EQ (prop, Qcharset_g0))
1404 return coding_system_charset (coding_system, 0);
1405 else if (EQ (prop, Qcharset_g1))
1406 return coding_system_charset (coding_system, 1);
1407 else if (EQ (prop, Qcharset_g2))
1408 return coding_system_charset (coding_system, 2);
1409 else if (EQ (prop, Qcharset_g3))
1410 return coding_system_charset (coding_system, 3);
1412 #define FORCE_CHARSET(charset_num) \
1413 (XCODING_SYSTEM_ISO2022_FORCE_CHARSET_ON_OUTPUT \
1414 (coding_system, charset_num) ? Qt : Qnil)
1416 else if (EQ (prop, Qforce_g0_on_output)) return FORCE_CHARSET (0);
1417 else if (EQ (prop, Qforce_g1_on_output)) return FORCE_CHARSET (1);
1418 else if (EQ (prop, Qforce_g2_on_output)) return FORCE_CHARSET (2);
1419 else if (EQ (prop, Qforce_g3_on_output)) return FORCE_CHARSET (3);
1421 #define LISP_BOOLEAN(prop) \
1422 (XCODING_SYSTEM_ISO2022_##prop (coding_system) ? Qt : Qnil)
1424 else if (EQ (prop, Qshort)) return LISP_BOOLEAN (SHORT);
1425 else if (EQ (prop, Qno_ascii_eol)) return LISP_BOOLEAN (NO_ASCII_EOL);
1426 else if (EQ (prop, Qno_ascii_cntl)) return LISP_BOOLEAN (NO_ASCII_CNTL);
1427 else if (EQ (prop, Qseven)) return LISP_BOOLEAN (SEVEN);
1428 else if (EQ (prop, Qlock_shift)) return LISP_BOOLEAN (LOCK_SHIFT);
1429 else if (EQ (prop, Qno_iso6429)) return LISP_BOOLEAN (NO_ISO6429);
1430 else if (EQ (prop, Qescape_quoted)) return LISP_BOOLEAN (ESCAPE_QUOTED);
1432 else if (EQ (prop, Qinput_charset_conversion))
1434 unparse_charset_conversion_specs
1435 (XCODING_SYSTEM (coding_system)->iso2022.input_conv);
1436 else if (EQ (prop, Qoutput_charset_conversion))
1438 unparse_charset_conversion_specs
1439 (XCODING_SYSTEM (coding_system)->iso2022.output_conv);
1443 else if (type == CODESYS_CCL)
1445 if (EQ (prop, Qdecode))
1446 return XCODING_SYSTEM_CCL_DECODE (coding_system);
1447 else if (EQ (prop, Qencode))
1448 return XCODING_SYSTEM_CCL_ENCODE (coding_system);
1456 return Qnil; /* not reached */
1460 /************************************************************************/
1461 /* Coding category functions */
1462 /************************************************************************/
1465 decode_coding_category (Lisp_Object symbol)
1469 CHECK_SYMBOL (symbol);
1470 for (i = 0; i < CODING_CATEGORY_LAST; i++)
1471 if (EQ (coding_category_symbol[i], symbol))
1474 signal_simple_error ("Unrecognized coding category", symbol);
1475 return 0; /* not reached */
1478 DEFUN ("coding-category-list", Fcoding_category_list, 0, 0, 0, /*
1479 Return a list of all recognized coding categories.
1484 Lisp_Object list = Qnil;
1486 for (i = CODING_CATEGORY_LAST - 1; i >= 0; i--)
1487 list = Fcons (coding_category_symbol[i], list);
1491 DEFUN ("set-coding-priority-list", Fset_coding_priority_list, 1, 1, 0, /*
1492 Change the priority order of the coding categories.
1493 LIST should be list of coding categories, in descending order of
1494 priority. Unspecified coding categories will be lower in priority
1495 than all specified ones, in the same relative order they were in
1500 int category_to_priority[CODING_CATEGORY_LAST];
1504 /* First generate a list that maps coding categories to priorities. */
1506 for (i = 0; i < CODING_CATEGORY_LAST; i++)
1507 category_to_priority[i] = -1;
1509 /* Highest priority comes from the specified list. */
1511 EXTERNAL_LIST_LOOP (rest, list)
1513 int cat = decode_coding_category (XCAR (rest));
1515 if (category_to_priority[cat] >= 0)
1516 signal_simple_error ("Duplicate coding category in list", XCAR (rest));
1517 category_to_priority[cat] = i++;
1520 /* Now go through the existing categories by priority to retrieve
1521 the categories not yet specified and preserve their priority
1523 for (j = 0; j < CODING_CATEGORY_LAST; j++)
1525 int cat = fcd->coding_category_by_priority[j];
1526 if (category_to_priority[cat] < 0)
1527 category_to_priority[cat] = i++;
1530 /* Now we need to construct the inverse of the mapping we just
1533 for (i = 0; i < CODING_CATEGORY_LAST; i++)
1534 fcd->coding_category_by_priority[category_to_priority[i]] = i;
1536 /* Phew! That was confusing. */
1540 DEFUN ("coding-priority-list", Fcoding_priority_list, 0, 0, 0, /*
1541 Return a list of coding categories in descending order of priority.
1546 Lisp_Object list = Qnil;
1548 for (i = CODING_CATEGORY_LAST - 1; i >= 0; i--)
1549 list = Fcons (coding_category_symbol[fcd->coding_category_by_priority[i]],
1554 DEFUN ("set-coding-category-system", Fset_coding_category_system, 2, 2, 0, /*
1555 Change the coding system associated with a coding category.
1557 (coding_category, coding_system))
1559 int cat = decode_coding_category (coding_category);
1561 coding_system = Fget_coding_system (coding_system);
1562 fcd->coding_category_system[cat] = coding_system;
1566 DEFUN ("coding-category-system", Fcoding_category_system, 1, 1, 0, /*
1567 Return the coding system associated with a coding category.
1571 int cat = decode_coding_category (coding_category);
1572 Lisp_Object sys = fcd->coding_category_system[cat];
1575 return XCODING_SYSTEM_NAME (sys);
1580 /************************************************************************/
1581 /* Detecting the encoding of data */
1582 /************************************************************************/
1584 struct detection_state
1586 eol_type_t eol_type;
1622 struct iso2022_decoder iso;
1624 int high_byte_count;
1625 unsigned int saw_single_shift:1;
1638 acceptable_control_char_p (int c)
1642 /* Allow and ignore control characters that you might
1643 reasonably see in a text file */
1648 case 8: /* backspace */
1649 case 11: /* vertical tab */
1650 case 12: /* form feed */
1651 case 26: /* MS-DOS C-z junk */
1652 case 31: /* '^_' -- for info */
1660 mask_has_at_most_one_bit_p (int mask)
1662 /* Perhaps the only thing useful you learn from intensive Microsoft
1663 technical interviews */
1664 return (mask & (mask - 1)) == 0;
1668 detect_eol_type (struct detection_state *st, const Extbyte *src,
1673 unsigned char c = *(unsigned char *)src++;
1676 if (st->eol.just_saw_cr)
1678 else if (st->eol.seen_anything)
1681 else if (st->eol.just_saw_cr)
1684 st->eol.just_saw_cr = 1;
1686 st->eol.just_saw_cr = 0;
1687 st->eol.seen_anything = 1;
1690 return EOL_AUTODETECT;
1693 /* Attempt to determine the encoding and EOL type of the given text.
1694 Before calling this function for the first type, you must initialize
1695 st->eol_type as appropriate and initialize st->mask to ~0.
1697 st->eol_type holds the determined EOL type, or EOL_AUTODETECT if
1700 st->mask holds the determined coding category mask, or ~0 if only
1701 ASCII has been seen so far.
1705 0 == st->eol_type is EOL_AUTODETECT and/or more than coding category
1706 is present in st->mask
1707 1 == definitive answers are here for both st->eol_type and st->mask
1711 detect_coding_type (struct detection_state *st, const Extbyte *src,
1712 size_t n, int just_do_eol)
1714 if (st->eol_type == EOL_AUTODETECT)
1715 st->eol_type = detect_eol_type (st, src, n);
1718 return st->eol_type != EOL_AUTODETECT;
1720 if (!st->seen_non_ascii)
1722 for (; n; n--, src++)
1724 unsigned char c = *(unsigned char *) src;
1725 if ((c < 0x20 && !acceptable_control_char_p (c)) || c >= 0x80)
1727 st->seen_non_ascii = 1;
1729 st->shift_jis.mask = ~0;
1733 st->iso2022.mask = ~0;
1743 if (!mask_has_at_most_one_bit_p (st->iso2022.mask))
1744 st->iso2022.mask = detect_coding_iso2022 (st, src, n);
1745 if (!mask_has_at_most_one_bit_p (st->shift_jis.mask))
1746 st->shift_jis.mask = detect_coding_sjis (st, src, n);
1747 if (!mask_has_at_most_one_bit_p (st->big5.mask))
1748 st->big5.mask = detect_coding_big5 (st, src, n);
1749 if (!mask_has_at_most_one_bit_p (st->utf8.mask))
1750 st->utf8.mask = detect_coding_utf8 (st, src, n);
1751 if (!mask_has_at_most_one_bit_p (st->ucs4.mask))
1752 st->ucs4.mask = detect_coding_ucs4 (st, src, n);
1755 = st->iso2022.mask | st->shift_jis.mask | st->big5.mask
1756 | st->utf8.mask | st->ucs4.mask;
1759 int retval = mask_has_at_most_one_bit_p (st->mask);
1760 st->mask |= CODING_CATEGORY_NO_CONVERSION_MASK;
1761 return retval && st->eol_type != EOL_AUTODETECT;
1766 coding_system_from_mask (int mask)
1770 /* If the file was entirely or basically ASCII, use the
1771 default value of `buffer-file-coding-system'. */
1772 Lisp_Object retval =
1773 XBUFFER (Vbuffer_defaults)->buffer_file_coding_system;
1776 retval = Ffind_coding_system (retval);
1780 (Qbad_variable, Qwarning,
1781 "Invalid `default-buffer-file-coding-system', set to nil");
1782 XBUFFER (Vbuffer_defaults)->buffer_file_coding_system = Qnil;
1786 retval = Fget_coding_system (Qraw_text);
1794 mask = postprocess_iso2022_mask (mask);
1796 /* Look through the coding categories by priority and find
1797 the first one that is allowed. */
1798 for (i = 0; i < CODING_CATEGORY_LAST; i++)
1800 cat = fcd->coding_category_by_priority[i];
1801 if ((mask & (1 << cat)) &&
1802 !NILP (fcd->coding_category_system[cat]))
1806 return fcd->coding_category_system[cat];
1808 return Fget_coding_system (Qraw_text);
1812 /* Given a seekable read stream and potential coding system and EOL type
1813 as specified, do any autodetection that is called for. If the
1814 coding system and/or EOL type are not `autodetect', they will be left
1815 alone; but this function will never return an autodetect coding system
1818 This function does not automatically fetch subsidiary coding systems;
1819 that should be unnecessary with the explicit eol-type argument. */
1821 #define LENGTH(string_constant) (sizeof (string_constant) - 1)
1824 determine_real_coding_system (Lstream *stream, Lisp_Object *codesys_in_out,
1825 eol_type_t *eol_type_in_out)
1827 struct detection_state decst;
1829 if (*eol_type_in_out == EOL_AUTODETECT)
1830 *eol_type_in_out = XCODING_SYSTEM_EOL_TYPE (*codesys_in_out);
1833 decst.eol_type = *eol_type_in_out;
1836 /* If autodetection is called for, do it now. */
1837 if (XCODING_SYSTEM_TYPE (*codesys_in_out) == CODESYS_AUTODETECT
1838 || *eol_type_in_out == EOL_AUTODETECT)
1841 Lisp_Object coding_system = Qnil;
1843 ssize_t nread = Lstream_read (stream, buf, sizeof (buf));
1846 /* Look for initial "-*-"; mode line prefix */
1848 scan_end = buf + nread - LENGTH ("-*-coding:?-*-");
1853 if (*p == '-' && *(p+1) == '*' && *(p+2) == '-')
1855 Extbyte *local_vars_beg = p + 3;
1856 /* Look for final "-*-"; mode line suffix */
1857 for (p = local_vars_beg,
1858 scan_end = buf + nread - LENGTH ("-*-");
1863 if (*p == '-' && *(p+1) == '*' && *(p+2) == '-')
1865 Extbyte *suffix = p;
1866 /* Look for "coding:" */
1867 for (p = local_vars_beg,
1868 scan_end = suffix - LENGTH ("coding:?");
1871 if (memcmp ("coding:", p, LENGTH ("coding:")) == 0
1872 && (p == local_vars_beg
1873 || (*(p-1) == ' ' ||
1879 p += LENGTH ("coding:");
1880 while (*p == ' ' || *p == '\t') p++;
1882 /* Get coding system name */
1883 save = *suffix; *suffix = '\0';
1884 /* Characters valid in a MIME charset name (rfc 1521),
1885 and in a Lisp symbol name. */
1886 n = strspn ( (char *) p,
1887 "ABCDEFGHIJKLMNOPQRSTUVWXYZ"
1888 "abcdefghijklmnopqrstuvwxyz"
1894 save = p[n]; p[n] = '\0';
1896 Ffind_coding_system (intern ((char *) p));
1906 if (NILP (coding_system))
1909 if (detect_coding_type (&decst, buf, nread,
1910 XCODING_SYSTEM_TYPE (*codesys_in_out)
1911 != CODESYS_AUTODETECT))
1913 nread = Lstream_read (stream, buf, sizeof (buf));
1919 else if (XCODING_SYSTEM_TYPE (*codesys_in_out) == CODESYS_AUTODETECT
1920 && XCODING_SYSTEM_EOL_TYPE (coding_system) == EOL_AUTODETECT)
1923 if (detect_coding_type (&decst, buf, nread, 1))
1925 nread = Lstream_read (stream, buf, sizeof (buf));
1931 *eol_type_in_out = decst.eol_type;
1932 if (XCODING_SYSTEM_TYPE (*codesys_in_out) == CODESYS_AUTODETECT)
1934 if (NILP (coding_system))
1935 *codesys_in_out = coding_system_from_mask (decst.mask);
1937 *codesys_in_out = coding_system;
1941 /* If we absolutely can't determine the EOL type, just assume LF. */
1942 if (*eol_type_in_out == EOL_AUTODETECT)
1943 *eol_type_in_out = EOL_LF;
1945 Lstream_rewind (stream);
1948 DEFUN ("detect-coding-region", Fdetect_coding_region, 2, 3, 0, /*
1949 Detect coding system of the text in the region between START and END.
1950 Return a list of possible coding systems ordered by priority.
1951 If only ASCII characters are found, return 'undecided or one of
1952 its subsidiary coding systems according to a detected end-of-line
1953 type. Optional arg BUFFER defaults to the current buffer.
1955 (start, end, buffer))
1957 Lisp_Object val = Qnil;
1958 struct buffer *buf = decode_buffer (buffer, 0);
1960 Lisp_Object instream, lb_instream;
1961 Lstream *istr, *lb_istr;
1962 struct detection_state decst;
1963 struct gcpro gcpro1, gcpro2;
1965 get_buffer_range_char (buf, start, end, &b, &e, 0);
1966 lb_instream = make_lisp_buffer_input_stream (buf, b, e, 0);
1967 lb_istr = XLSTREAM (lb_instream);
1968 instream = make_encoding_input_stream (lb_istr, Fget_coding_system (Qbinary));
1969 istr = XLSTREAM (instream);
1970 GCPRO2 (instream, lb_instream);
1972 decst.eol_type = EOL_AUTODETECT;
1976 Extbyte random_buffer[4096];
1977 ssize_t nread = Lstream_read (istr, random_buffer, sizeof (random_buffer));
1981 if (detect_coding_type (&decst, random_buffer, nread, 0))
1985 if (decst.mask == ~0)
1986 val = subsidiary_coding_system (Fget_coding_system (Qundecided),
1994 decst.mask = postprocess_iso2022_mask (decst.mask);
1996 for (i = CODING_CATEGORY_LAST - 1; i >= 0; i--)
1998 int sys = fcd->coding_category_by_priority[i];
1999 if (decst.mask & (1 << sys))
2001 Lisp_Object codesys = fcd->coding_category_system[sys];
2002 if (!NILP (codesys))
2003 codesys = subsidiary_coding_system (codesys, decst.eol_type);
2004 val = Fcons (codesys, val);
2008 Lstream_close (istr);
2010 Lstream_delete (istr);
2011 Lstream_delete (lb_istr);
2016 /************************************************************************/
2017 /* Converting to internal Mule format ("decoding") */
2018 /************************************************************************/
2020 /* A decoding stream is a stream used for decoding text (i.e.
2021 converting from some external format to internal format).
2022 The decoding-stream object keeps track of the actual coding
2023 stream, the stream that is at the other end, and data that
2024 needs to be persistent across the lifetime of the stream. */
2026 /* Handle the EOL stuff related to just-read-in character C.
2027 EOL_TYPE is the EOL type of the coding stream.
2028 FLAGS is the current value of FLAGS in the coding stream, and may
2029 be modified by this macro. (The macro only looks at the
2030 CODING_STATE_CR flag.) DST is the Dynarr to which the decoded
2031 bytes are to be written. You need to also define a local goto
2032 label "label_continue_loop" that is at the end of the main
2033 character-reading loop.
2035 If C is a CR character, then this macro handles it entirely and
2036 jumps to label_continue_loop. Otherwise, this macro does not add
2037 anything to DST, and continues normally. You should continue
2038 processing C normally after this macro. */
2040 #define DECODE_HANDLE_EOL_TYPE(eol_type, c, flags, dst) \
2044 if (eol_type == EOL_CR) \
2045 Dynarr_add (dst, '\n'); \
2046 else if (eol_type != EOL_CRLF || flags & CODING_STATE_CR) \
2047 Dynarr_add (dst, c); \
2049 flags |= CODING_STATE_CR; \
2050 goto label_continue_loop; \
2052 else if (flags & CODING_STATE_CR) \
2053 { /* eol_type == CODING_SYSTEM_EOL_CRLF */ \
2055 Dynarr_add (dst, '\r'); \
2056 flags &= ~CODING_STATE_CR; \
2060 /* C should be a binary character in the range 0 - 255; convert
2061 to internal format and add to Dynarr DST. */
2064 #define DECODE_ADD_BINARY_CHAR(c, dst) \
2066 if (BYTE_ASCII_P (c)) \
2067 Dynarr_add (dst, c); \
2070 Dynarr_add (dst, (c >> 6) | 0xc0); \
2071 Dynarr_add (dst, (c & 0x3f) | 0x80); \
2075 INLINE_HEADER void DECODE_ADD_UCS_CHAR(Emchar c, unsigned_char_dynarr* dst);
2077 DECODE_ADD_UCS_CHAR(Emchar c, unsigned_char_dynarr* dst)
2081 Dynarr_add (dst, c);
2083 else if ( c <= 0x7ff )
2085 Dynarr_add (dst, (c >> 6) | 0xc0);
2086 Dynarr_add (dst, (c & 0x3f) | 0x80);
2088 else if ( c <= 0xffff )
2090 Dynarr_add (dst, (c >> 12) | 0xe0);
2091 Dynarr_add (dst, ((c >> 6) & 0x3f) | 0x80);
2092 Dynarr_add (dst, (c & 0x3f) | 0x80);
2094 else if ( c <= 0x1fffff )
2096 Dynarr_add (dst, (c >> 18) | 0xf0);
2097 Dynarr_add (dst, ((c >> 12) & 0x3f) | 0x80);
2098 Dynarr_add (dst, ((c >> 6) & 0x3f) | 0x80);
2099 Dynarr_add (dst, (c & 0x3f) | 0x80);
2101 else if ( c <= 0x3ffffff )
2103 Dynarr_add (dst, (c >> 24) | 0xf8);
2104 Dynarr_add (dst, ((c >> 18) & 0x3f) | 0x80);
2105 Dynarr_add (dst, ((c >> 12) & 0x3f) | 0x80);
2106 Dynarr_add (dst, ((c >> 6) & 0x3f) | 0x80);
2107 Dynarr_add (dst, (c & 0x3f) | 0x80);
2111 Dynarr_add (dst, (c >> 30) | 0xfc);
2112 Dynarr_add (dst, ((c >> 24) & 0x3f) | 0x80);
2113 Dynarr_add (dst, ((c >> 18) & 0x3f) | 0x80);
2114 Dynarr_add (dst, ((c >> 12) & 0x3f) | 0x80);
2115 Dynarr_add (dst, ((c >> 6) & 0x3f) | 0x80);
2116 Dynarr_add (dst, (c & 0x3f) | 0x80);
2120 #define DECODE_ADD_BINARY_CHAR(c, dst) \
2122 if (BYTE_ASCII_P (c)) \
2123 Dynarr_add (dst, c); \
2124 else if (BYTE_C1_P (c)) \
2126 Dynarr_add (dst, LEADING_BYTE_CONTROL_1); \
2127 Dynarr_add (dst, c + 0x20); \
2131 Dynarr_add (dst, LEADING_BYTE_LATIN_ISO8859_1); \
2132 Dynarr_add (dst, c); \
2137 #define DECODE_OUTPUT_PARTIAL_CHAR(ch) \
2141 DECODE_ADD_BINARY_CHAR (ch, dst); \
2146 #define DECODE_HANDLE_END_OF_CONVERSION(flags, ch, dst) \
2148 if (flags & CODING_STATE_END) \
2150 DECODE_OUTPUT_PARTIAL_CHAR (ch); \
2151 if (flags & CODING_STATE_CR) \
2152 Dynarr_add (dst, '\r'); \
2156 #define DECODING_STREAM_DATA(stream) LSTREAM_TYPE_DATA (stream, decoding)
2158 struct decoding_stream
2160 /* Coding system that governs the conversion. */
2161 Lisp_Coding_System *codesys;
2163 /* Stream that we read the encoded data from or
2164 write the decoded data to. */
2167 /* If we are reading, then we can return only a fixed amount of
2168 data, so if the conversion resulted in too much data, we store it
2169 here for retrieval the next time around. */
2170 unsigned_char_dynarr *runoff;
2172 /* FLAGS holds flags indicating the current state of the decoding.
2173 Some of these flags are dependent on the coding system. */
2176 /* CPOS holds a partially built-up code-point of character. */
2179 /* EOL_TYPE specifies the type of end-of-line conversion that
2180 currently applies. We need to keep this separate from the
2181 EOL type stored in CODESYS because the latter might indicate
2182 automatic EOL-type detection while the former will always
2183 indicate a particular EOL type. */
2184 eol_type_t eol_type;
2186 /* Additional ISO2022 information. We define the structure above
2187 because it's also needed by the detection routines. */
2188 struct iso2022_decoder iso2022;
2190 /* Additional information (the state of the running CCL program)
2191 used by the CCL decoder. */
2192 struct ccl_program ccl;
2194 /* counter for UTF-8 or UCS-4 */
2195 unsigned char counter;
2198 unsigned combined_char_count;
2199 Emchar combined_chars[16];
2200 Lisp_Object combining_table;
2202 struct detection_state decst;
2206 extern Lisp_Object Vcharacter_composition_table;
2209 COMPOSE_FLUSH_CHARS (struct decoding_stream *str, unsigned_char_dynarr* dst);
2211 COMPOSE_FLUSH_CHARS (struct decoding_stream *str, unsigned_char_dynarr* dst)
2215 for (i = 0; i < str->combined_char_count; i++)
2216 DECODE_ADD_UCS_CHAR (str->combined_chars[i], dst);
2217 str->combined_char_count = 0;
2218 str->combining_table = Qnil;
2221 void COMPOSE_ADD_CHAR(struct decoding_stream *str, Emchar character,
2222 unsigned_char_dynarr* dst);
2224 COMPOSE_ADD_CHAR(struct decoding_stream *str,
2225 Emchar character, unsigned_char_dynarr* dst)
2227 if (CODING_SYSTEM_DISABLE_COMPOSITION (str->codesys))
2228 DECODE_ADD_UCS_CHAR (character, dst);
2229 else if (!CHAR_ID_TABLE_P (str->combining_table))
2232 = get_char_id_table (character, Vcharacter_composition_table);
2235 DECODE_ADD_UCS_CHAR (character, dst);
2238 str->combined_chars[0] = character;
2239 str->combined_char_count = 1;
2240 str->combining_table = ret;
2246 = get_char_id_table (character, str->combining_table);
2250 Emchar char2 = XCHARVAL (ret);
2251 ret = get_char_id_table (char2, Vcharacter_composition_table);
2254 DECODE_ADD_UCS_CHAR (char2, dst);
2255 str->combined_char_count = 0;
2256 str->combining_table = Qnil;
2260 str->combined_chars[0] = char2;
2261 str->combined_char_count = 1;
2262 str->combining_table = ret;
2265 else if (CHAR_ID_TABLE_P (ret))
2267 str->combined_chars[str->combined_char_count++] = character;
2268 str->combining_table = ret;
2272 COMPOSE_FLUSH_CHARS (str, dst);
2273 DECODE_ADD_UCS_CHAR (character, dst);
2277 #else /* not UTF2000 */
2278 #define COMPOSE_FLUSH_CHARS(str, dst)
2279 #define COMPOSE_ADD_CHAR(str, ch, dst) DECODE_ADD_UCS_CHAR (ch, dst)
2280 #endif /* UTF2000 */
2282 static ssize_t decoding_reader (Lstream *stream,
2283 unsigned char *data, size_t size);
2284 static ssize_t decoding_writer (Lstream *stream,
2285 const unsigned char *data, size_t size);
2286 static int decoding_rewinder (Lstream *stream);
2287 static int decoding_seekable_p (Lstream *stream);
2288 static int decoding_flusher (Lstream *stream);
2289 static int decoding_closer (Lstream *stream);
2291 static Lisp_Object decoding_marker (Lisp_Object stream);
2293 DEFINE_LSTREAM_IMPLEMENTATION ("decoding", lstream_decoding,
2294 sizeof (struct decoding_stream));
2297 decoding_marker (Lisp_Object stream)
2299 Lstream *str = DECODING_STREAM_DATA (XLSTREAM (stream))->other_end;
2300 Lisp_Object str_obj;
2302 /* We do not need to mark the coding systems or charsets stored
2303 within the stream because they are stored in a global list
2304 and automatically marked. */
2306 XSETLSTREAM (str_obj, str);
2307 mark_object (str_obj);
2308 if (str->imp->marker)
2309 return (str->imp->marker) (str_obj);
2314 /* Read SIZE bytes of data and store it into DATA. We are a decoding stream
2315 so we read data from the other end, decode it, and store it into DATA. */
2318 decoding_reader (Lstream *stream, unsigned char *data, size_t size)
2320 struct decoding_stream *str = DECODING_STREAM_DATA (stream);
2321 unsigned char *orig_data = data;
2323 int error_occurred = 0;
2325 /* We need to interface to mule_decode(), which expects to take some
2326 amount of data and store the result into a Dynarr. We have
2327 mule_decode() store into str->runoff, and take data from there
2330 /* We loop until we have enough data, reading chunks from the other
2331 end and decoding it. */
2334 /* Take data from the runoff if we can. Make sure to take at
2335 most SIZE bytes, and delete the data from the runoff. */
2336 if (Dynarr_length (str->runoff) > 0)
2338 size_t chunk = min (size, (size_t) Dynarr_length (str->runoff));
2339 memcpy (data, Dynarr_atp (str->runoff, 0), chunk);
2340 Dynarr_delete_many (str->runoff, 0, chunk);
2346 break; /* No more room for data */
2348 if (str->flags & CODING_STATE_END)
2349 /* This means that on the previous iteration, we hit the EOF on
2350 the other end. We loop once more so that mule_decode() can
2351 output any final stuff it may be holding, or any "go back
2352 to a sane state" escape sequences. (This latter makes sense
2353 during encoding.) */
2356 /* Exhausted the runoff, so get some more. DATA has at least
2357 SIZE bytes left of storage in it, so it's OK to read directly
2358 into it. (We'll be overwriting above, after we've decoded it
2359 into the runoff.) */
2360 read_size = Lstream_read (str->other_end, data, size);
2367 /* There might be some more end data produced in the translation.
2368 See the comment above. */
2369 str->flags |= CODING_STATE_END;
2370 mule_decode (stream, (Extbyte *) data, str->runoff, read_size);
2373 if (data - orig_data == 0)
2374 return error_occurred ? -1 : 0;
2376 return data - orig_data;
2380 decoding_writer (Lstream *stream, const unsigned char *data, size_t size)
2382 struct decoding_stream *str = DECODING_STREAM_DATA (stream);
2385 /* Decode all our data into the runoff, and then attempt to write
2386 it all out to the other end. Remove whatever chunk we succeeded
2388 mule_decode (stream, (Extbyte *) data, str->runoff, size);
2389 retval = Lstream_write (str->other_end, Dynarr_atp (str->runoff, 0),
2390 Dynarr_length (str->runoff));
2392 Dynarr_delete_many (str->runoff, 0, retval);
2393 /* Do NOT return retval. The return value indicates how much
2394 of the incoming data was written, not how many bytes were
2400 reset_decoding_stream (struct decoding_stream *str)
2403 if (CODING_SYSTEM_TYPE (str->codesys) == CODESYS_ISO2022)
2405 Lisp_Object coding_system;
2406 XSETCODING_SYSTEM (coding_system, str->codesys);
2407 reset_iso2022 (coding_system, &str->iso2022);
2409 else if (CODING_SYSTEM_TYPE (str->codesys) == CODESYS_CCL)
2411 setup_ccl_program (&str->ccl, CODING_SYSTEM_CCL_DECODE (str->codesys));
2416 str->combined_char_count = 0;
2417 str->combining_table = Qnil;
2419 str->flags = str->cpos = 0;
2423 decoding_rewinder (Lstream *stream)
2425 struct decoding_stream *str = DECODING_STREAM_DATA (stream);
2426 reset_decoding_stream (str);
2427 Dynarr_reset (str->runoff);
2428 return Lstream_rewind (str->other_end);
2432 decoding_seekable_p (Lstream *stream)
2434 struct decoding_stream *str = DECODING_STREAM_DATA (stream);
2435 return Lstream_seekable_p (str->other_end);
2439 decoding_flusher (Lstream *stream)
2441 struct decoding_stream *str = DECODING_STREAM_DATA (stream);
2442 return Lstream_flush (str->other_end);
2446 decoding_closer (Lstream *stream)
2448 struct decoding_stream *str = DECODING_STREAM_DATA (stream);
2449 if (stream->flags & LSTREAM_FL_WRITE)
2451 str->flags |= CODING_STATE_END;
2452 decoding_writer (stream, 0, 0);
2454 Dynarr_free (str->runoff);
2456 #ifdef ENABLE_COMPOSITE_CHARS
2457 if (str->iso2022.composite_chars)
2458 Dynarr_free (str->iso2022.composite_chars);
2461 return Lstream_close (str->other_end);
2465 decoding_stream_coding_system (Lstream *stream)
2467 Lisp_Object coding_system;
2468 struct decoding_stream *str = DECODING_STREAM_DATA (stream);
2470 XSETCODING_SYSTEM (coding_system, str->codesys);
2471 return subsidiary_coding_system (coding_system, str->eol_type);
2475 set_decoding_stream_coding_system (Lstream *lstr, Lisp_Object codesys)
2477 Lisp_Coding_System *cs = XCODING_SYSTEM (codesys);
2478 struct decoding_stream *str = DECODING_STREAM_DATA (lstr);
2480 if (CODING_SYSTEM_EOL_TYPE (cs) != EOL_AUTODETECT)
2481 str->eol_type = CODING_SYSTEM_EOL_TYPE (cs);
2482 reset_decoding_stream (str);
2485 /* WARNING WARNING WARNING WARNING!!!!! If you open up a decoding
2486 stream for writing, no automatic code detection will be performed.
2487 The reason for this is that automatic code detection requires a
2488 seekable input. Things will also fail if you open a decoding
2489 stream for reading using a non-fully-specified coding system and
2490 a non-seekable input stream. */
2493 make_decoding_stream_1 (Lstream *stream, Lisp_Object codesys,
2496 Lstream *lstr = Lstream_new (lstream_decoding, mode);
2497 struct decoding_stream *str = DECODING_STREAM_DATA (lstr);
2501 str->other_end = stream;
2502 str->runoff = (unsigned_char_dynarr *) Dynarr_new (unsigned_char);
2503 str->eol_type = EOL_AUTODETECT;
2504 if (!strcmp (mode, "r")
2505 && Lstream_seekable_p (stream))
2506 /* We can determine the coding system now. */
2507 determine_real_coding_system (stream, &codesys, &str->eol_type);
2508 set_decoding_stream_coding_system (lstr, codesys);
2509 str->decst.eol_type = str->eol_type;
2510 str->decst.mask = ~0;
2511 XSETLSTREAM (obj, lstr);
2516 make_decoding_input_stream (Lstream *stream, Lisp_Object codesys)
2518 return make_decoding_stream_1 (stream, codesys, "r");
2522 make_decoding_output_stream (Lstream *stream, Lisp_Object codesys)
2524 return make_decoding_stream_1 (stream, codesys, "w");
2527 /* Note: the decode_coding_* functions all take the same
2528 arguments as mule_decode(), which is to say some SRC data of
2529 size N, which is to be stored into dynamic array DST.
2530 DECODING is the stream within which the decoding is
2531 taking place, but no data is actually read from or
2532 written to that stream; that is handled in decoding_reader()
2533 or decoding_writer(). This allows the same functions to
2534 be used for both reading and writing. */
2537 mule_decode (Lstream *decoding, const Extbyte *src,
2538 unsigned_char_dynarr *dst, size_t n)
2540 struct decoding_stream *str = DECODING_STREAM_DATA (decoding);
2542 /* If necessary, do encoding-detection now. We do this when
2543 we're a writing stream or a non-seekable reading stream,
2544 meaning that we can't just process the whole input,
2545 rewind, and start over. */
2547 if (CODING_SYSTEM_TYPE (str->codesys) == CODESYS_AUTODETECT ||
2548 str->eol_type == EOL_AUTODETECT)
2550 Lisp_Object codesys;
2552 XSETCODING_SYSTEM (codesys, str->codesys);
2553 detect_coding_type (&str->decst, src, n,
2554 CODING_SYSTEM_TYPE (str->codesys) !=
2555 CODESYS_AUTODETECT);
2556 if (CODING_SYSTEM_TYPE (str->codesys) == CODESYS_AUTODETECT &&
2557 str->decst.mask != ~0)
2558 /* #### This is cheesy. What we really ought to do is
2559 buffer up a certain amount of data so as to get a
2560 less random result. */
2561 codesys = coding_system_from_mask (str->decst.mask);
2562 str->eol_type = str->decst.eol_type;
2563 if (XCODING_SYSTEM (codesys) != str->codesys)
2565 /* Preserve the CODING_STATE_END flag in case it was set.
2566 If we erase it, bad things might happen. */
2567 int was_end = str->flags & CODING_STATE_END;
2568 set_decoding_stream_coding_system (decoding, codesys);
2570 str->flags |= CODING_STATE_END;
2574 switch (CODING_SYSTEM_TYPE (str->codesys))
2577 case CODESYS_INTERNAL:
2578 Dynarr_add_many (dst, src, n);
2581 case CODESYS_AUTODETECT:
2582 /* If we got this far and still haven't decided on the coding
2583 system, then do no conversion. */
2584 case CODESYS_NO_CONVERSION:
2585 decode_coding_no_conversion (decoding, src, dst, n);
2588 case CODESYS_SHIFT_JIS:
2589 decode_coding_sjis (decoding, src, dst, n);
2592 decode_coding_big5 (decoding, src, dst, n);
2595 decode_coding_ucs4 (decoding, src, dst, n);
2598 decode_coding_utf8 (decoding, src, dst, n);
2601 str->ccl.last_block = str->flags & CODING_STATE_END;
2602 /* When applying ccl program to stream, MUST NOT set NULL
2604 ccl_driver (&str->ccl, (src ? (unsigned char *)src : (unsigned char*)""),
2605 dst, n, 0, CCL_MODE_DECODING);
2607 case CODESYS_ISO2022:
2608 decode_coding_iso2022 (decoding, src, dst, n);
2616 DEFUN ("decode-coding-region", Fdecode_coding_region, 3, 4, 0, /*
2617 Decode the text between START and END which is encoded in CODING-SYSTEM.
2618 This is useful if you've read in encoded text from a file without decoding
2619 it (e.g. you read in a JIS-formatted file but used the `binary' or
2620 `no-conversion' coding system, so that it shows up as "^[$B!<!+^[(B").
2621 Return length of decoded text.
2622 BUFFER defaults to the current buffer if unspecified.
2624 (start, end, coding_system, buffer))
2627 struct buffer *buf = decode_buffer (buffer, 0);
2628 Lisp_Object instream, lb_outstream, de_outstream, outstream;
2629 Lstream *istr, *ostr;
2630 struct gcpro gcpro1, gcpro2, gcpro3, gcpro4;
2632 get_buffer_range_char (buf, start, end, &b, &e, 0);
2634 barf_if_buffer_read_only (buf, b, e);
2636 coding_system = Fget_coding_system (coding_system);
2637 instream = make_lisp_buffer_input_stream (buf, b, e, 0);
2638 lb_outstream = make_lisp_buffer_output_stream (buf, b, 0);
2639 de_outstream = make_decoding_output_stream (XLSTREAM (lb_outstream),
2641 outstream = make_encoding_output_stream (XLSTREAM (de_outstream),
2642 Fget_coding_system (Qbinary));
2643 istr = XLSTREAM (instream);
2644 ostr = XLSTREAM (outstream);
2645 GCPRO4 (instream, lb_outstream, de_outstream, outstream);
2647 /* The chain of streams looks like this:
2649 [BUFFER] <----- send through
2650 ------> [ENCODE AS BINARY]
2651 ------> [DECODE AS SPECIFIED]
2657 char tempbuf[1024]; /* some random amount */
2658 Bufpos newpos, even_newer_pos;
2659 Bufpos oldpos = lisp_buffer_stream_startpos (istr);
2660 ssize_t size_in_bytes = Lstream_read (istr, tempbuf, sizeof (tempbuf));
2664 newpos = lisp_buffer_stream_startpos (istr);
2665 Lstream_write (ostr, tempbuf, size_in_bytes);
2666 even_newer_pos = lisp_buffer_stream_startpos (istr);
2667 buffer_delete_range (buf, even_newer_pos - (newpos - oldpos),
2670 Lstream_close (istr);
2671 Lstream_close (ostr);
2673 Lstream_delete (istr);
2674 Lstream_delete (ostr);
2675 Lstream_delete (XLSTREAM (de_outstream));
2676 Lstream_delete (XLSTREAM (lb_outstream));
2681 /************************************************************************/
2682 /* Converting to an external encoding ("encoding") */
2683 /************************************************************************/
2685 /* An encoding stream is an output stream. When you create the
2686 stream, you specify the coding system that governs the encoding
2687 and another stream that the resulting encoded data is to be
2688 sent to, and then start sending data to it. */
2690 #define ENCODING_STREAM_DATA(stream) LSTREAM_TYPE_DATA (stream, encoding)
2692 struct encoding_stream
2694 /* Coding system that governs the conversion. */
2695 Lisp_Coding_System *codesys;
2697 /* Stream that we read the encoded data from or
2698 write the decoded data to. */
2701 /* If we are reading, then we can return only a fixed amount of
2702 data, so if the conversion resulted in too much data, we store it
2703 here for retrieval the next time around. */
2704 unsigned_char_dynarr *runoff;
2706 /* FLAGS holds flags indicating the current state of the encoding.
2707 Some of these flags are dependent on the coding system. */
2710 /* CH holds a partially built-up character. Since we only deal
2711 with one- and two-byte characters at the moment, we only use
2712 this to store the first byte of a two-byte character. */
2715 /* Additional information used by the ISO2022 encoder. */
2718 /* CHARSET holds the character sets currently assigned to the G0
2719 through G3 registers. It is initialized from the array
2720 INITIAL_CHARSET in CODESYS. */
2721 Lisp_Object charset[4];
2723 /* Which registers are currently invoked into the left (GL) and
2724 right (GR) halves of the 8-bit encoding space? */
2725 int register_left, register_right;
2727 /* Whether we need to explicitly designate the charset in the
2728 G? register before using it. It is initialized from the
2729 array FORCE_CHARSET_ON_OUTPUT in CODESYS. */
2730 unsigned char force_charset_on_output[4];
2732 /* Other state variables that need to be preserved across
2734 Lisp_Object current_charset;
2736 int current_char_boundary;
2739 void (*encode_char) (struct encoding_stream *str, Emchar c,
2740 unsigned_char_dynarr *dst, unsigned int *flags);
2741 void (*finish) (struct encoding_stream *str,
2742 unsigned_char_dynarr *dst, unsigned int *flags);
2744 /* Additional information (the state of the running CCL program)
2745 used by the CCL encoder. */
2746 struct ccl_program ccl;
2750 static ssize_t encoding_reader (Lstream *stream, unsigned char *data, size_t size);
2751 static ssize_t encoding_writer (Lstream *stream, const unsigned char *data,
2753 static int encoding_rewinder (Lstream *stream);
2754 static int encoding_seekable_p (Lstream *stream);
2755 static int encoding_flusher (Lstream *stream);
2756 static int encoding_closer (Lstream *stream);
2758 static Lisp_Object encoding_marker (Lisp_Object stream);
2760 DEFINE_LSTREAM_IMPLEMENTATION ("encoding", lstream_encoding,
2761 sizeof (struct encoding_stream));
2764 encoding_marker (Lisp_Object stream)
2766 Lstream *str = ENCODING_STREAM_DATA (XLSTREAM (stream))->other_end;
2767 Lisp_Object str_obj;
2769 /* We do not need to mark the coding systems or charsets stored
2770 within the stream because they are stored in a global list
2771 and automatically marked. */
2773 XSETLSTREAM (str_obj, str);
2774 mark_object (str_obj);
2775 if (str->imp->marker)
2776 return (str->imp->marker) (str_obj);
2781 /* Read SIZE bytes of data and store it into DATA. We are a encoding stream
2782 so we read data from the other end, encode it, and store it into DATA. */
2785 encoding_reader (Lstream *stream, unsigned char *data, size_t size)
2787 struct encoding_stream *str = ENCODING_STREAM_DATA (stream);
2788 unsigned char *orig_data = data;
2790 int error_occurred = 0;
2792 /* We need to interface to mule_encode(), which expects to take some
2793 amount of data and store the result into a Dynarr. We have
2794 mule_encode() store into str->runoff, and take data from there
2797 /* We loop until we have enough data, reading chunks from the other
2798 end and encoding it. */
2801 /* Take data from the runoff if we can. Make sure to take at
2802 most SIZE bytes, and delete the data from the runoff. */
2803 if (Dynarr_length (str->runoff) > 0)
2805 int chunk = min ((int) size, Dynarr_length (str->runoff));
2806 memcpy (data, Dynarr_atp (str->runoff, 0), chunk);
2807 Dynarr_delete_many (str->runoff, 0, chunk);
2813 break; /* No more room for data */
2815 if (str->flags & CODING_STATE_END)
2816 /* This means that on the previous iteration, we hit the EOF on
2817 the other end. We loop once more so that mule_encode() can
2818 output any final stuff it may be holding, or any "go back
2819 to a sane state" escape sequences. (This latter makes sense
2820 during encoding.) */
2823 /* Exhausted the runoff, so get some more. DATA at least SIZE bytes
2824 left of storage in it, so it's OK to read directly into it.
2825 (We'll be overwriting above, after we've encoded it into the
2827 read_size = Lstream_read (str->other_end, data, size);
2834 /* There might be some more end data produced in the translation.
2835 See the comment above. */
2836 str->flags |= CODING_STATE_END;
2837 mule_encode (stream, data, str->runoff, read_size);
2840 if (data == orig_data)
2841 return error_occurred ? -1 : 0;
2843 return data - orig_data;
2847 encoding_writer (Lstream *stream, const unsigned char *data, size_t size)
2849 struct encoding_stream *str = ENCODING_STREAM_DATA (stream);
2852 /* Encode all our data into the runoff, and then attempt to write
2853 it all out to the other end. Remove whatever chunk we succeeded
2855 mule_encode (stream, data, str->runoff, size);
2856 retval = Lstream_write (str->other_end, Dynarr_atp (str->runoff, 0),
2857 Dynarr_length (str->runoff));
2859 Dynarr_delete_many (str->runoff, 0, retval);
2860 /* Do NOT return retval. The return value indicates how much
2861 of the incoming data was written, not how many bytes were
2867 reset_encoding_stream (struct encoding_stream *str)
2870 switch (CODING_SYSTEM_TYPE (str->codesys))
2872 case CODESYS_ISO2022:
2876 str->encode_char = &char_encode_iso2022;
2877 str->finish = &char_finish_iso2022;
2878 for (i = 0; i < 4; i++)
2880 str->iso2022.charset[i] =
2881 CODING_SYSTEM_ISO2022_INITIAL_CHARSET (str->codesys, i);
2882 str->iso2022.force_charset_on_output[i] =
2883 CODING_SYSTEM_ISO2022_FORCE_CHARSET_ON_OUTPUT (str->codesys, i);
2885 str->iso2022.register_left = 0;
2886 str->iso2022.register_right = 1;
2887 str->iso2022.current_charset = Qnil;
2888 str->iso2022.current_half = 0;
2892 setup_ccl_program (&str->ccl, CODING_SYSTEM_CCL_ENCODE (str->codesys));
2895 str->encode_char = &char_encode_utf8;
2896 str->finish = &char_finish_utf8;
2899 str->encode_char = &char_encode_ucs4;
2900 str->finish = &char_finish_ucs4;
2902 case CODESYS_SHIFT_JIS:
2903 str->encode_char = &char_encode_shift_jis;
2904 str->finish = &char_finish_shift_jis;
2907 str->encode_char = &char_encode_big5;
2908 str->finish = &char_finish_big5;
2914 str->iso2022.current_char_boundary = 0;
2915 str->flags = str->ch = 0;
2919 encoding_rewinder (Lstream *stream)
2921 struct encoding_stream *str = ENCODING_STREAM_DATA (stream);
2922 reset_encoding_stream (str);
2923 Dynarr_reset (str->runoff);
2924 return Lstream_rewind (str->other_end);
2928 encoding_seekable_p (Lstream *stream)
2930 struct encoding_stream *str = ENCODING_STREAM_DATA (stream);
2931 return Lstream_seekable_p (str->other_end);
2935 encoding_flusher (Lstream *stream)
2937 struct encoding_stream *str = ENCODING_STREAM_DATA (stream);
2938 return Lstream_flush (str->other_end);
2942 encoding_closer (Lstream *stream)
2944 struct encoding_stream *str = ENCODING_STREAM_DATA (stream);
2945 if (stream->flags & LSTREAM_FL_WRITE)
2947 str->flags |= CODING_STATE_END;
2948 encoding_writer (stream, 0, 0);
2950 Dynarr_free (str->runoff);
2951 return Lstream_close (str->other_end);
2955 encoding_stream_coding_system (Lstream *stream)
2957 Lisp_Object coding_system;
2958 struct encoding_stream *str = ENCODING_STREAM_DATA (stream);
2960 XSETCODING_SYSTEM (coding_system, str->codesys);
2961 return coding_system;
2965 set_encoding_stream_coding_system (Lstream *lstr, Lisp_Object codesys)
2967 Lisp_Coding_System *cs = XCODING_SYSTEM (codesys);
2968 struct encoding_stream *str = ENCODING_STREAM_DATA (lstr);
2970 reset_encoding_stream (str);
2974 make_encoding_stream_1 (Lstream *stream, Lisp_Object codesys,
2977 Lstream *lstr = Lstream_new (lstream_encoding, mode);
2978 struct encoding_stream *str = ENCODING_STREAM_DATA (lstr);
2982 str->runoff = Dynarr_new (unsigned_char);
2983 str->other_end = stream;
2984 set_encoding_stream_coding_system (lstr, codesys);
2985 XSETLSTREAM (obj, lstr);
2990 make_encoding_input_stream (Lstream *stream, Lisp_Object codesys)
2992 return make_encoding_stream_1 (stream, codesys, "r");
2996 make_encoding_output_stream (Lstream *stream, Lisp_Object codesys)
2998 return make_encoding_stream_1 (stream, codesys, "w");
3001 /* Convert N bytes of internally-formatted data stored in SRC to an
3002 external format, according to the encoding stream ENCODING.
3003 Store the encoded data into DST. */
3006 mule_encode (Lstream *encoding, const Bufbyte *src,
3007 unsigned_char_dynarr *dst, size_t n)
3009 struct encoding_stream *str = ENCODING_STREAM_DATA (encoding);
3011 switch (CODING_SYSTEM_TYPE (str->codesys))
3014 case CODESYS_INTERNAL:
3015 Dynarr_add_many (dst, src, n);
3018 case CODESYS_AUTODETECT:
3019 /* If we got this far and still haven't decided on the coding
3020 system, then do no conversion. */
3021 case CODESYS_NO_CONVERSION:
3022 encode_coding_no_conversion (encoding, src, dst, n);
3026 str->ccl.last_block = str->flags & CODING_STATE_END;
3027 /* When applying ccl program to stream, MUST NOT set NULL
3029 ccl_driver (&str->ccl, ((src) ? src : (unsigned char*)""),
3030 dst, n, 0, CCL_MODE_ENCODING);
3034 text_encode_generic (encoding, src, dst, n);
3038 DEFUN ("encode-coding-region", Fencode_coding_region, 3, 4, 0, /*
3039 Encode the text between START and END using CODING-SYSTEM.
3040 This will, for example, convert Japanese characters into stuff such as
3041 "^[$B!<!+^[(B" if you use the JIS encoding. Return length of encoded
3042 text. BUFFER defaults to the current buffer if unspecified.
3044 (start, end, coding_system, buffer))
3047 struct buffer *buf = decode_buffer (buffer, 0);
3048 Lisp_Object instream, lb_outstream, de_outstream, outstream;
3049 Lstream *istr, *ostr;
3050 struct gcpro gcpro1, gcpro2, gcpro3, gcpro4;
3052 get_buffer_range_char (buf, start, end, &b, &e, 0);
3054 barf_if_buffer_read_only (buf, b, e);
3056 coding_system = Fget_coding_system (coding_system);
3057 instream = make_lisp_buffer_input_stream (buf, b, e, 0);
3058 lb_outstream = make_lisp_buffer_output_stream (buf, b, 0);
3059 de_outstream = make_decoding_output_stream (XLSTREAM (lb_outstream),
3060 Fget_coding_system (Qbinary));
3061 outstream = make_encoding_output_stream (XLSTREAM (de_outstream),
3063 istr = XLSTREAM (instream);
3064 ostr = XLSTREAM (outstream);
3065 GCPRO4 (instream, outstream, de_outstream, lb_outstream);
3066 /* The chain of streams looks like this:
3068 [BUFFER] <----- send through
3069 ------> [ENCODE AS SPECIFIED]
3070 ------> [DECODE AS BINARY]
3075 char tempbuf[1024]; /* some random amount */
3076 Bufpos newpos, even_newer_pos;
3077 Bufpos oldpos = lisp_buffer_stream_startpos (istr);
3078 ssize_t size_in_bytes = Lstream_read (istr, tempbuf, sizeof (tempbuf));
3082 newpos = lisp_buffer_stream_startpos (istr);
3083 Lstream_write (ostr, tempbuf, size_in_bytes);
3084 even_newer_pos = lisp_buffer_stream_startpos (istr);
3085 buffer_delete_range (buf, even_newer_pos - (newpos - oldpos),
3091 lisp_buffer_stream_startpos (XLSTREAM (instream)) - b;
3092 Lstream_close (istr);
3093 Lstream_close (ostr);
3095 Lstream_delete (istr);
3096 Lstream_delete (ostr);
3097 Lstream_delete (XLSTREAM (de_outstream));
3098 Lstream_delete (XLSTREAM (lb_outstream));
3099 return make_int (retlen);
3106 text_encode_generic (Lstream *encoding, const Bufbyte *src,
3107 unsigned_char_dynarr *dst, size_t n)
3110 unsigned char char_boundary;
3111 struct encoding_stream *str = ENCODING_STREAM_DATA (encoding);
3112 unsigned int flags = str->flags;
3113 Emchar ch = str->ch;
3115 char_boundary = str->iso2022.current_char_boundary;
3121 if (char_boundary == 0)
3149 (*str->encode_char) (str, c, dst, &flags);
3151 else if (char_boundary == 1)
3153 (*str->encode_char) (str, (ch << 6) | (c & 0x3f), dst, &flags);
3159 ch = (ch << 6) | (c & 0x3f);
3164 if ((char_boundary == 0) && (flags & CODING_STATE_END))
3166 (*str->finish) (str, dst, &flags);
3171 str->iso2022.current_char_boundary = char_boundary;
3175 /************************************************************************/
3176 /* Shift-JIS methods */
3177 /************************************************************************/
3179 /* Shift-JIS is a coding system encoding three character sets: ASCII, right
3180 half of JISX0201-Kana, and JISX0208. An ASCII character is encoded
3181 as is. A character of JISX0201-Kana (DIMENSION1_CHARS94 character set) is
3182 encoded by "position-code + 0x80". A character of JISX0208
3183 (DIMENSION2_CHARS94 character set) is encoded in 2-byte but two
3184 position-codes are divided and shifted so that it fit in the range
3187 --- CODE RANGE of Shift-JIS ---
3188 (character set) (range)
3190 JISX0201-Kana 0xA0 .. 0xDF
3191 JISX0208 (1st byte) 0x80 .. 0x9F and 0xE0 .. 0xEF
3192 (2nd byte) 0x40 .. 0x7E and 0x80 .. 0xFC
3193 -------------------------------
3197 /* Is this the first byte of a Shift-JIS two-byte char? */
3199 #define BYTE_SJIS_TWO_BYTE_1_P(c) \
3200 (((c) >= 0x81 && (c) <= 0x9F) || ((c) >= 0xE0 && (c) <= 0xEF))
3202 /* Is this the second byte of a Shift-JIS two-byte char? */
3204 #define BYTE_SJIS_TWO_BYTE_2_P(c) \
3205 (((c) >= 0x40 && (c) <= 0x7E) || ((c) >= 0x80 && (c) <= 0xFC))
3207 #define BYTE_SJIS_KATAKANA_P(c) \
3208 ((c) >= 0xA1 && (c) <= 0xDF)
3211 detect_coding_sjis (struct detection_state *st, const Extbyte *src, size_t n)
3215 unsigned char c = *(unsigned char *)src++;
3216 if (c == ISO_CODE_ESC || c == ISO_CODE_SI || c == ISO_CODE_SO)
3218 if (st->shift_jis.in_second_byte)
3220 st->shift_jis.in_second_byte = 0;
3224 else if ((c >= 0x80 && c < 0xA0) || c >= 0xE0)
3225 st->shift_jis.in_second_byte = 1;
3227 return CODING_CATEGORY_SHIFT_JIS_MASK;
3230 /* Convert Shift-JIS data to internal format. */
3233 decode_coding_sjis (Lstream *decoding, const Extbyte *src,
3234 unsigned_char_dynarr *dst, size_t n)
3236 struct decoding_stream *str = DECODING_STREAM_DATA (decoding);
3237 unsigned int flags = str->flags;
3238 unsigned int cpos = str->cpos;
3239 eol_type_t eol_type = str->eol_type;
3243 unsigned char c = *(unsigned char *)src++;
3247 /* Previous character was first byte of Shift-JIS Kanji char. */
3248 if (BYTE_SJIS_TWO_BYTE_2_P (c))
3250 unsigned char e1, e2;
3252 DECODE_SJIS (cpos, c, e1, e2);
3254 DECODE_ADD_UCS_CHAR(MAKE_CHAR(Vcharset_japanese_jisx0208,
3258 Dynarr_add (dst, LEADING_BYTE_JAPANESE_JISX0208);
3259 Dynarr_add (dst, e1);
3260 Dynarr_add (dst, e2);
3265 DECODE_ADD_BINARY_CHAR (cpos, dst);
3266 DECODE_ADD_BINARY_CHAR (c, dst);
3272 DECODE_HANDLE_EOL_TYPE (eol_type, c, flags, dst);
3273 if (BYTE_SJIS_TWO_BYTE_1_P (c))
3275 else if (BYTE_SJIS_KATAKANA_P (c))
3278 DECODE_ADD_UCS_CHAR(MAKE_CHAR(Vcharset_katakana_jisx0201,
3281 Dynarr_add (dst, LEADING_BYTE_KATAKANA_JISX0201);
3282 Dynarr_add (dst, c);
3287 DECODE_ADD_UCS_CHAR(MAKE_CHAR(Vcharset_latin_jisx0201,
3291 DECODE_ADD_BINARY_CHAR (c, dst);
3293 label_continue_loop:;
3296 DECODE_HANDLE_END_OF_CONVERSION (flags, cpos, dst);
3302 /* Convert internal character representation to Shift_JIS. */
3305 char_encode_shift_jis (struct encoding_stream *str, Emchar ch,
3306 unsigned_char_dynarr *dst, unsigned int *flags)
3308 eol_type_t eol_type = CODING_SYSTEM_EOL_TYPE (str->codesys);
3312 if (eol_type != EOL_LF && eol_type != EOL_AUTODETECT)
3313 Dynarr_add (dst, '\r');
3314 if (eol_type != EOL_CR)
3315 Dynarr_add (dst, ch);
3319 unsigned int s1, s2;
3321 int code_point = charset_code_point (Vcharset_latin_jisx0201, ch);
3323 if (code_point >= 0)
3324 Dynarr_add (dst, code_point);
3325 else if ((code_point
3326 = charset_code_point (Vcharset_japanese_jisx0208_1990, ch))
3329 ENCODE_SJIS ((code_point >> 8) | 0x80,
3330 (code_point & 0xFF) | 0x80, s1, s2);
3331 Dynarr_add (dst, s1);
3332 Dynarr_add (dst, s2);
3334 else if ((code_point
3335 = charset_code_point (Vcharset_katakana_jisx0201, ch))
3337 Dynarr_add (dst, code_point | 0x80);
3338 else if ((code_point
3339 = charset_code_point (Vcharset_japanese_jisx0208, ch))
3342 ENCODE_SJIS ((code_point >> 8) | 0x80,
3343 (code_point & 0xFF) | 0x80, s1, s2);
3344 Dynarr_add (dst, s1);
3345 Dynarr_add (dst, s2);
3347 else if ((code_point = charset_code_point (Vcharset_ascii, ch))
3349 Dynarr_add (dst, code_point);
3351 Dynarr_add (dst, '?');
3353 Lisp_Object charset;
3354 unsigned int c1, c2;
3356 BREAKUP_CHAR (ch, charset, c1, c2);
3358 if (EQ(charset, Vcharset_katakana_jisx0201))
3360 Dynarr_add (dst, c1 | 0x80);
3364 Dynarr_add (dst, c1);
3366 else if (EQ(charset, Vcharset_japanese_jisx0208))
3368 ENCODE_SJIS (c1 | 0x80, c2 | 0x80, s1, s2);
3369 Dynarr_add (dst, s1);
3370 Dynarr_add (dst, s2);
3373 Dynarr_add (dst, '?');
3379 char_finish_shift_jis (struct encoding_stream *str, unsigned_char_dynarr *dst,
3380 unsigned int *flags)
3384 DEFUN ("decode-shift-jis-char", Fdecode_shift_jis_char, 1, 1, 0, /*
3385 Decode a JISX0208 character of Shift-JIS coding-system.
3386 CODE is the character code in Shift-JIS as a cons of type bytes.
3387 Return the corresponding character.
3391 unsigned char c1, c2, s1, s2;
3394 CHECK_INT (XCAR (code));
3395 CHECK_INT (XCDR (code));
3396 s1 = XINT (XCAR (code));
3397 s2 = XINT (XCDR (code));
3398 if (BYTE_SJIS_TWO_BYTE_1_P (s1) &&
3399 BYTE_SJIS_TWO_BYTE_2_P (s2))
3401 DECODE_SJIS (s1, s2, c1, c2);
3402 return make_char (MAKE_CHAR (Vcharset_japanese_jisx0208,
3403 c1 & 0x7F, c2 & 0x7F));
3409 DEFUN ("encode-shift-jis-char", Fencode_shift_jis_char, 1, 1, 0, /*
3410 Encode a JISX0208 character CHARACTER to SHIFT-JIS coding-system.
3411 Return the corresponding character code in SHIFT-JIS as a cons of two bytes.
3415 Lisp_Object charset;
3418 CHECK_CHAR_COERCE_INT (character);
3419 BREAKUP_CHAR (XCHAR (character), charset, c1, c2);
3420 if (EQ (charset, Vcharset_japanese_jisx0208))
3422 ENCODE_SJIS (c1 | 0x80, c2 | 0x80, s1, s2);
3423 return Fcons (make_int (s1), make_int (s2));
3430 /************************************************************************/
3432 /************************************************************************/
3434 /* BIG5 is a coding system encoding two character sets: ASCII and
3435 Big5. An ASCII character is encoded as is. Big5 is a two-byte
3436 character set and is encoded in two-byte.
3438 --- CODE RANGE of BIG5 ---
3439 (character set) (range)
3441 Big5 (1st byte) 0xA1 .. 0xFE
3442 (2nd byte) 0x40 .. 0x7E and 0xA1 .. 0xFE
3443 --------------------------
3445 Since the number of characters in Big5 is larger than maximum
3446 characters in Emacs' charset (96x96), it can't be handled as one
3447 charset. So, in Emacs, Big5 is divided into two: `charset-big5-1'
3448 and `charset-big5-2'. Both <type>s are DIMENSION2_CHARS94. The former
3449 contains frequently used characters and the latter contains less
3450 frequently used characters. */
3453 #define BYTE_BIG5_TWO_BYTE_1_P(c) \
3454 ((c) >= 0x81 && (c) <= 0xFE)
3456 #define BYTE_BIG5_TWO_BYTE_1_P(c) \
3457 ((c) >= 0xA1 && (c) <= 0xFE)
3460 /* Is this the second byte of a Shift-JIS two-byte char? */
3462 #define BYTE_BIG5_TWO_BYTE_2_P(c) \
3463 (((c) >= 0x40 && (c) <= 0x7E) || ((c) >= 0xA1 && (c) <= 0xFE))
3465 /* Number of Big5 characters which have the same code in 1st byte. */
3467 #define BIG5_SAME_ROW (0xFF - 0xA1 + 0x7F - 0x40)
3469 /* Code conversion macros. These are macros because they are used in
3470 inner loops during code conversion.
3472 Note that temporary variables in macros introduce the classic
3473 dynamic-scoping problems with variable names. We use capital-
3474 lettered variables in the assumption that XEmacs does not use
3475 capital letters in variables except in a very formalized way
3478 /* Convert Big5 code (b1, b2) into its internal string representation
3481 /* There is a much simpler way to split the Big5 charset into two.
3482 For the moment I'm going to leave the algorithm as-is because it
3483 claims to separate out the most-used characters into a single
3484 charset, which perhaps will lead to optimizations in various
3487 The way the algorithm works is something like this:
3489 Big5 can be viewed as a 94x157 charset, where the row is
3490 encoded into the bytes 0xA1 .. 0xFE and the column is encoded
3491 into the bytes 0x40 .. 0x7E and 0xA1 .. 0xFE. As for frequency,
3492 the split between low and high column numbers is apparently
3493 meaningless; ascending rows produce less and less frequent chars.
3494 Therefore, we assign the lower half of rows (0xA1 .. 0xC8) to
3495 the first charset, and the upper half (0xC9 .. 0xFE) to the
3496 second. To do the conversion, we convert the character into
3497 a single number where 0 .. 156 is the first row, 157 .. 313
3498 is the second, etc. That way, the characters are ordered by
3499 decreasing frequency. Then we just chop the space in two
3500 and coerce the result into a 94x94 space.
3503 #define DECODE_BIG5(b1, b2, lb, c1, c2) do \
3505 int B1 = b1, B2 = b2; \
3507 = (B1 - 0xA1) * BIG5_SAME_ROW + B2 - (B2 < 0x7F ? 0x40 : 0x62); \
3511 lb = LEADING_BYTE_CHINESE_BIG5_1; \
3515 lb = LEADING_BYTE_CHINESE_BIG5_2; \
3516 I -= (BIG5_SAME_ROW) * (0xC9 - 0xA1); \
3518 c1 = I / (0xFF - 0xA1) + 0xA1; \
3519 c2 = I % (0xFF - 0xA1) + 0xA1; \
3522 /* Convert the internal string representation of a Big5 character
3523 (lb, c1, c2) into Big5 code (b1, b2). */
3525 #define ENCODE_BIG5(lb, c1, c2, b1, b2) do \
3527 unsigned int I = ((c1) - 0xA1) * (0xFF - 0xA1) + ((c2) - 0xA1); \
3529 if (lb == LEADING_BYTE_CHINESE_BIG5_2) \
3531 I += BIG5_SAME_ROW * (0xC9 - 0xA1); \
3533 b1 = I / BIG5_SAME_ROW + 0xA1; \
3534 b2 = I % BIG5_SAME_ROW; \
3535 b2 += b2 < 0x3F ? 0x40 : 0x62; \
3539 detect_coding_big5 (struct detection_state *st, const Extbyte *src, size_t n)
3543 unsigned char c = *(unsigned char *)src++;
3544 if (c == ISO_CODE_ESC || c == ISO_CODE_SI || c == ISO_CODE_SO
3546 || (c >= 0x80 && c <= 0xA0)
3550 if (st->big5.in_second_byte)
3552 st->big5.in_second_byte = 0;
3553 if (c < 0x40 || (c >= 0x80 && c <= 0xA0))
3563 st->big5.in_second_byte = 1;
3565 return CODING_CATEGORY_BIG5_MASK;
3568 /* Convert Big5 data to internal format. */
3571 decode_coding_big5 (Lstream *decoding, const Extbyte *src,
3572 unsigned_char_dynarr *dst, size_t n)
3574 struct decoding_stream *str = DECODING_STREAM_DATA (decoding);
3575 unsigned int flags = str->flags;
3576 unsigned int cpos = str->cpos;
3577 eol_type_t eol_type = str->eol_type;
3581 unsigned char c = *(unsigned char *)src++;
3584 /* Previous character was first byte of Big5 char. */
3585 if (BYTE_BIG5_TWO_BYTE_2_P (c))
3589 (DECODE_CHAR (Vcharset_chinese_big5, (cpos << 8) | c),
3592 unsigned char b1, b2, b3;
3593 DECODE_BIG5 (cpos, c, b1, b2, b3);
3594 Dynarr_add (dst, b1);
3595 Dynarr_add (dst, b2);
3596 Dynarr_add (dst, b3);
3601 DECODE_ADD_BINARY_CHAR (cpos, dst);
3602 DECODE_ADD_BINARY_CHAR (c, dst);
3608 DECODE_HANDLE_EOL_TYPE (eol_type, c, flags, dst);
3609 if (BYTE_BIG5_TWO_BYTE_1_P (c))
3612 DECODE_ADD_BINARY_CHAR (c, dst);
3614 label_continue_loop:;
3617 DECODE_HANDLE_END_OF_CONVERSION (flags, cpos, dst);
3623 /* Convert internally-formatted data to Big5. */
3626 char_encode_big5 (struct encoding_stream *str, Emchar ch,
3627 unsigned_char_dynarr *dst, unsigned int *flags)
3629 eol_type_t eol_type = CODING_SYSTEM_EOL_TYPE (str->codesys);
3633 if (eol_type != EOL_LF && eol_type != EOL_AUTODETECT)
3634 Dynarr_add (dst, '\r');
3635 if (eol_type != EOL_CR)
3636 Dynarr_add (dst, ch);
3643 if ((code_point = charset_code_point (Vcharset_ascii, ch)) >= 0)
3644 Dynarr_add (dst, code_point);
3645 else if ((code_point
3646 = charset_code_point (Vcharset_chinese_big5, ch)) >= 0)
3648 Dynarr_add (dst, code_point >> 8);
3649 Dynarr_add (dst, code_point & 0xFF);
3651 else if ((code_point
3652 = charset_code_point (Vcharset_chinese_big5_1, ch)) >= 0)
3655 = ((code_point >> 8) - 33) * (0xFF - 0xA1)
3656 + ((code_point & 0xFF) - 33);
3657 unsigned char b1 = I / BIG5_SAME_ROW + 0xA1;
3658 unsigned char b2 = I % BIG5_SAME_ROW;
3660 b2 += b2 < 0x3F ? 0x40 : 0x62;
3661 Dynarr_add (dst, b1);
3662 Dynarr_add (dst, b2);
3664 else if ((code_point
3665 = charset_code_point (Vcharset_chinese_big5_2, ch)) >= 0)
3668 = ((code_point >> 8) - 33) * (0xFF - 0xA1)
3669 + ((code_point & 0xFF) - 33);
3670 unsigned char b1, b2;
3672 I += BIG5_SAME_ROW * (0xC9 - 0xA1);
3673 b1 = I / BIG5_SAME_ROW + 0xA1;
3674 b2 = I % BIG5_SAME_ROW;
3675 b2 += b2 < 0x3F ? 0x40 : 0x62;
3676 Dynarr_add (dst, b1);
3677 Dynarr_add (dst, b2);
3680 Dynarr_add (dst, '?');
3687 char_finish_big5 (struct encoding_stream *str, unsigned_char_dynarr *dst,
3688 unsigned int *flags)
3693 DEFUN ("decode-big5-char", Fdecode_big5_char, 1, 1, 0, /*
3694 Decode a Big5 character CODE of BIG5 coding-system.
3695 CODE is the character code in BIG5, a cons of two integers.
3696 Return the corresponding character.
3700 unsigned char c1, c2, b1, b2;
3703 CHECK_INT (XCAR (code));
3704 CHECK_INT (XCDR (code));
3705 b1 = XINT (XCAR (code));
3706 b2 = XINT (XCDR (code));
3707 if (BYTE_BIG5_TWO_BYTE_1_P (b1) &&
3708 BYTE_BIG5_TWO_BYTE_2_P (b2))
3710 Charset_ID leading_byte;
3711 Lisp_Object charset;
3712 DECODE_BIG5 (b1, b2, leading_byte, c1, c2);
3713 charset = CHARSET_BY_LEADING_BYTE (leading_byte);
3714 return make_char (MAKE_CHAR (charset, c1 & 0x7F, c2 & 0x7F));
3720 DEFUN ("encode-big5-char", Fencode_big5_char, 1, 1, 0, /*
3721 Encode the Big5 character CHARACTER in the BIG5 coding-system.
3722 Return the corresponding character code in Big5.
3726 Lisp_Object charset;
3729 CHECK_CHAR_COERCE_INT (character);
3730 BREAKUP_CHAR (XCHAR (character), charset, c1, c2);
3731 if (EQ (charset, Vcharset_chinese_big5_1) ||
3732 EQ (charset, Vcharset_chinese_big5_2))
3734 ENCODE_BIG5 (XCHARSET_LEADING_BYTE (charset), c1 | 0x80, c2 | 0x80,
3736 return Fcons (make_int (b1), make_int (b2));
3743 /************************************************************************/
3745 /************************************************************************/
3748 detect_coding_ucs4 (struct detection_state *st, const Extbyte *src, size_t n)
3752 unsigned char c = *(unsigned char *)src++;
3753 switch (st->ucs4.in_byte)
3762 st->ucs4.in_byte = 0;
3768 return CODING_CATEGORY_UCS4_MASK;
3772 decode_coding_ucs4 (Lstream *decoding, const Extbyte *src,
3773 unsigned_char_dynarr *dst, size_t n)
3775 struct decoding_stream *str = DECODING_STREAM_DATA (decoding);
3776 unsigned int flags = str->flags;
3777 unsigned int cpos = str->cpos;
3778 unsigned char counter = str->counter;
3782 unsigned char c = *(unsigned char *)src++;
3790 DECODE_ADD_UCS_CHAR ((cpos << 8) | c, dst);
3795 cpos = ( cpos << 8 ) | c;
3799 if (counter & CODING_STATE_END)
3800 DECODE_OUTPUT_PARTIAL_CHAR (cpos);
3804 str->counter = counter;
3808 char_encode_ucs4 (struct encoding_stream *str, Emchar ch,
3809 unsigned_char_dynarr *dst, unsigned int *flags)
3811 Dynarr_add (dst, ch >> 24);
3812 Dynarr_add (dst, ch >> 16);
3813 Dynarr_add (dst, ch >> 8);
3814 Dynarr_add (dst, ch );
3818 char_finish_ucs4 (struct encoding_stream *str, unsigned_char_dynarr *dst,
3819 unsigned int *flags)
3824 /************************************************************************/
3826 /************************************************************************/
3829 detect_coding_utf8 (struct detection_state *st, const Extbyte *src, size_t n)
3833 unsigned char c = *(unsigned char *)src++;
3834 switch (st->utf8.in_byte)
3837 if (c == ISO_CODE_ESC || c == ISO_CODE_SI || c == ISO_CODE_SO)
3840 st->utf8.in_byte = 5;
3842 st->utf8.in_byte = 4;
3844 st->utf8.in_byte = 3;
3846 st->utf8.in_byte = 2;
3848 st->utf8.in_byte = 1;
3853 if ((c & 0xc0) != 0x80)
3859 return CODING_CATEGORY_UTF8_MASK;
3863 decode_output_utf8_partial_char (unsigned char counter,
3865 unsigned_char_dynarr *dst)
3868 DECODE_ADD_BINARY_CHAR ( (cpos|0xFC), dst);
3869 else if (counter == 4)
3871 if (cpos < (1 << 6))
3872 DECODE_ADD_BINARY_CHAR ( (cpos|0xF8), dst);
3875 DECODE_ADD_BINARY_CHAR ( ((cpos >> 6)|0xFC), dst);
3876 DECODE_ADD_BINARY_CHAR ( ((cpos&0x3F)|0x80), dst);
3879 else if (counter == 3)
3881 if (cpos < (1 << 6))
3882 DECODE_ADD_BINARY_CHAR ( (cpos|0xF0), dst);
3883 else if (cpos < (1 << 12))
3885 DECODE_ADD_BINARY_CHAR ( ((cpos >> 6)|0xF8), dst);
3886 DECODE_ADD_BINARY_CHAR ( ((cpos&0x3F)|0x80), dst);
3890 DECODE_ADD_BINARY_CHAR ( ( (cpos >> 12)|0xFC), dst);
3891 DECODE_ADD_BINARY_CHAR ( (((cpos >> 6)&0x3F)|0x80), dst);
3892 DECODE_ADD_BINARY_CHAR ( ( (cpos &0x3F)|0x80), dst);
3895 else if (counter == 2)
3897 if (cpos < (1 << 6))
3898 DECODE_ADD_BINARY_CHAR ( (cpos|0xE0), dst);
3899 else if (cpos < (1 << 12))
3901 DECODE_ADD_BINARY_CHAR ( ((cpos >> 6)|0xF0), dst);
3902 DECODE_ADD_BINARY_CHAR ( ((cpos&0x3F)|0x80), dst);
3904 else if (cpos < (1 << 18))
3906 DECODE_ADD_BINARY_CHAR ( ( (cpos >> 12)|0xF8), dst);
3907 DECODE_ADD_BINARY_CHAR ( (((cpos >> 6)&0x3F)|0x80), dst);
3908 DECODE_ADD_BINARY_CHAR ( ( (cpos &0x3F)|0x80), dst);
3912 DECODE_ADD_BINARY_CHAR ( ( (cpos >> 18)|0xFC), dst);
3913 DECODE_ADD_BINARY_CHAR ( (((cpos >> 12)&0x3F)|0x80), dst);
3914 DECODE_ADD_BINARY_CHAR ( (((cpos >> 6)&0x3F)|0x80), dst);
3915 DECODE_ADD_BINARY_CHAR ( ( (cpos &0x3F)|0x80), dst);
3920 if (cpos < (1 << 6))
3921 DECODE_ADD_BINARY_CHAR ( (cpos|0xC0), dst);
3922 else if (cpos < (1 << 12))
3924 DECODE_ADD_BINARY_CHAR ( ((cpos >> 6)|0xE0), dst);
3925 DECODE_ADD_BINARY_CHAR ( ((cpos&0x3F)|0x80), dst);
3927 else if (cpos < (1 << 18))
3929 DECODE_ADD_BINARY_CHAR ( ( (cpos >> 12)|0xF0), dst);
3930 DECODE_ADD_BINARY_CHAR ( (((cpos >> 6)&0x3F)|0x80), dst);
3931 DECODE_ADD_BINARY_CHAR ( ( (cpos &0x3F)|0x80), dst);
3933 else if (cpos < (1 << 24))
3935 DECODE_ADD_BINARY_CHAR ( ( (cpos >> 18)|0xF8), dst);
3936 DECODE_ADD_BINARY_CHAR ( (((cpos >> 12)&0x3F)|0x80), dst);
3937 DECODE_ADD_BINARY_CHAR ( (((cpos >> 6)&0x3F)|0x80), dst);
3938 DECODE_ADD_BINARY_CHAR ( ( (cpos &0x3F)|0x80), dst);
3942 DECODE_ADD_BINARY_CHAR ( ( (cpos >> 24)|0xFC), dst);
3943 DECODE_ADD_BINARY_CHAR ( (((cpos >> 18)&0x3F)|0x80), dst);
3944 DECODE_ADD_BINARY_CHAR ( (((cpos >> 12)&0x3F)|0x80), dst);
3945 DECODE_ADD_BINARY_CHAR ( (((cpos >> 6)&0x3F)|0x80), dst);
3946 DECODE_ADD_BINARY_CHAR ( ( (cpos &0x3F)|0x80), dst);
3952 decode_coding_utf8 (Lstream *decoding, const Extbyte *src,
3953 unsigned_char_dynarr *dst, size_t n)
3955 struct decoding_stream *str = DECODING_STREAM_DATA (decoding);
3956 unsigned int flags = str->flags;
3957 unsigned int cpos = str->cpos;
3958 eol_type_t eol_type = str->eol_type;
3959 unsigned char counter = str->counter;
3963 unsigned char c = *(unsigned char *)src++;
3968 DECODE_HANDLE_EOL_TYPE (eol_type, c, flags, dst);
3969 DECODE_ADD_UCS_CHAR (c, dst);
3971 else if ( c < 0xE0 )
3976 else if ( c < 0xF0 )
3981 else if ( c < 0xF8 )
3986 else if ( c < 0xFC )
3997 else if ( (c & 0xC0) == 0x80 )
3999 cpos = ( cpos << 6 ) | ( c & 0x3f );
4002 DECODE_ADD_UCS_CHAR (cpos, dst);
4011 decode_output_utf8_partial_char (counter, cpos, dst);
4012 DECODE_ADD_BINARY_CHAR (c, dst);
4016 label_continue_loop:;
4019 if (flags & CODING_STATE_END)
4022 decode_output_utf8_partial_char (counter, cpos, dst);
4028 str->counter = counter;
4032 char_encode_utf8 (struct encoding_stream *str, Emchar ch,
4033 unsigned_char_dynarr *dst, unsigned int *flags)
4035 eol_type_t eol_type = CODING_SYSTEM_EOL_TYPE (str->codesys);
4039 if (eol_type != EOL_LF && eol_type != EOL_AUTODETECT)
4040 Dynarr_add (dst, '\r');
4041 if (eol_type != EOL_CR)
4042 Dynarr_add (dst, ch);
4044 else if (ch <= 0x7f)
4046 Dynarr_add (dst, ch);
4048 else if (ch <= 0x7ff)
4050 Dynarr_add (dst, (ch >> 6) | 0xc0);
4051 Dynarr_add (dst, (ch & 0x3f) | 0x80);
4053 else if (ch <= 0xffff)
4055 Dynarr_add (dst, (ch >> 12) | 0xe0);
4056 Dynarr_add (dst, ((ch >> 6) & 0x3f) | 0x80);
4057 Dynarr_add (dst, (ch & 0x3f) | 0x80);
4059 else if (ch <= 0x1fffff)
4061 Dynarr_add (dst, (ch >> 18) | 0xf0);
4062 Dynarr_add (dst, ((ch >> 12) & 0x3f) | 0x80);
4063 Dynarr_add (dst, ((ch >> 6) & 0x3f) | 0x80);
4064 Dynarr_add (dst, (ch & 0x3f) | 0x80);
4066 else if (ch <= 0x3ffffff)
4068 Dynarr_add (dst, (ch >> 24) | 0xf8);
4069 Dynarr_add (dst, ((ch >> 18) & 0x3f) | 0x80);
4070 Dynarr_add (dst, ((ch >> 12) & 0x3f) | 0x80);
4071 Dynarr_add (dst, ((ch >> 6) & 0x3f) | 0x80);
4072 Dynarr_add (dst, (ch & 0x3f) | 0x80);
4076 Dynarr_add (dst, (ch >> 30) | 0xfc);
4077 Dynarr_add (dst, ((ch >> 24) & 0x3f) | 0x80);
4078 Dynarr_add (dst, ((ch >> 18) & 0x3f) | 0x80);
4079 Dynarr_add (dst, ((ch >> 12) & 0x3f) | 0x80);
4080 Dynarr_add (dst, ((ch >> 6) & 0x3f) | 0x80);
4081 Dynarr_add (dst, (ch & 0x3f) | 0x80);
4086 char_finish_utf8 (struct encoding_stream *str, unsigned_char_dynarr *dst,
4087 unsigned int *flags)
4092 /************************************************************************/
4093 /* ISO2022 methods */
4094 /************************************************************************/
4096 /* The following note describes the coding system ISO2022 briefly.
4097 Since the intention of this note is to help understand the
4098 functions in this file, some parts are NOT ACCURATE or OVERLY
4099 SIMPLIFIED. For thorough understanding, please refer to the
4100 original document of ISO2022.
4102 ISO2022 provides many mechanisms to encode several character sets
4103 in 7-bit and 8-bit environments. For 7-bit environments, all text
4104 is encoded using bytes less than 128. This may make the encoded
4105 text a little bit longer, but the text passes more easily through
4106 several gateways, some of which strip off MSB (Most Signigant Bit).
4108 There are two kinds of character sets: control character set and
4109 graphic character set. The former contains control characters such
4110 as `newline' and `escape' to provide control functions (control
4111 functions are also provided by escape sequences). The latter
4112 contains graphic characters such as 'A' and '-'. Emacs recognizes
4113 two control character sets and many graphic character sets.
4115 Graphic character sets are classified into one of the following
4116 four classes, according to the number of bytes (DIMENSION) and
4117 number of characters in one dimension (CHARS) of the set:
4118 - DIMENSION1_CHARS94
4119 - DIMENSION1_CHARS96
4120 - DIMENSION2_CHARS94
4121 - DIMENSION2_CHARS96
4123 In addition, each character set is assigned an identification tag,
4124 unique for each set, called "final character" (denoted as <F>
4125 hereafter). The <F> of each character set is decided by ECMA(*)
4126 when it is registered in ISO. The code range of <F> is 0x30..0x7F
4127 (0x30..0x3F are for private use only).
4129 Note (*): ECMA = European Computer Manufacturers Association
4131 Here are examples of graphic character set [NAME(<F>)]:
4132 o DIMENSION1_CHARS94 -- ASCII('B'), right-half-of-JISX0201('I'), ...
4133 o DIMENSION1_CHARS96 -- right-half-of-ISO8859-1('A'), ...
4134 o DIMENSION2_CHARS94 -- GB2312('A'), JISX0208('B'), ...
4135 o DIMENSION2_CHARS96 -- none for the moment
4137 A code area (1 byte = 8 bits) is divided into 4 areas, C0, GL, C1, and GR.
4138 C0 [0x00..0x1F] -- control character plane 0
4139 GL [0x20..0x7F] -- graphic character plane 0
4140 C1 [0x80..0x9F] -- control character plane 1
4141 GR [0xA0..0xFF] -- graphic character plane 1
4143 A control character set is directly designated and invoked to C0 or
4144 C1 by an escape sequence. The most common case is that:
4145 - ISO646's control character set is designated/invoked to C0, and
4146 - ISO6429's control character set is designated/invoked to C1,
4147 and usually these designations/invocations are omitted in encoded
4148 text. In a 7-bit environment, only C0 can be used, and a control
4149 character for C1 is encoded by an appropriate escape sequence to
4150 fit into the environment. All control characters for C1 are
4151 defined to have corresponding escape sequences.
4153 A graphic character set is at first designated to one of four
4154 graphic registers (G0 through G3), then these graphic registers are
4155 invoked to GL or GR. These designations and invocations can be
4156 done independently. The most common case is that G0 is invoked to
4157 GL, G1 is invoked to GR, and ASCII is designated to G0. Usually
4158 these invocations and designations are omitted in encoded text.
4159 In a 7-bit environment, only GL can be used.
4161 When a graphic character set of CHARS94 is invoked to GL, codes
4162 0x20 and 0x7F of the GL area work as control characters SPACE and
4163 DEL respectively, and codes 0xA0 and 0xFF of the GR area should not
4166 There are two ways of invocation: locking-shift and single-shift.
4167 With locking-shift, the invocation lasts until the next different
4168 invocation, whereas with single-shift, the invocation affects the
4169 following character only and doesn't affect the locking-shift
4170 state. Invocations are done by the following control characters or
4173 ----------------------------------------------------------------------
4174 abbrev function cntrl escape seq description
4175 ----------------------------------------------------------------------
4176 SI/LS0 (shift-in) 0x0F none invoke G0 into GL
4177 SO/LS1 (shift-out) 0x0E none invoke G1 into GL
4178 LS2 (locking-shift-2) none ESC 'n' invoke G2 into GL
4179 LS3 (locking-shift-3) none ESC 'o' invoke G3 into GL
4180 LS1R (locking-shift-1 right) none ESC '~' invoke G1 into GR (*)
4181 LS2R (locking-shift-2 right) none ESC '}' invoke G2 into GR (*)
4182 LS3R (locking-shift 3 right) none ESC '|' invoke G3 into GR (*)
4183 SS2 (single-shift-2) 0x8E ESC 'N' invoke G2 for one char
4184 SS3 (single-shift-3) 0x8F ESC 'O' invoke G3 for one char
4185 ----------------------------------------------------------------------
4186 (*) These are not used by any known coding system.
4188 Control characters for these functions are defined by macros
4189 ISO_CODE_XXX in `coding.h'.
4191 Designations are done by the following escape sequences:
4192 ----------------------------------------------------------------------
4193 escape sequence description
4194 ----------------------------------------------------------------------
4195 ESC '(' <F> designate DIMENSION1_CHARS94<F> to G0
4196 ESC ')' <F> designate DIMENSION1_CHARS94<F> to G1
4197 ESC '*' <F> designate DIMENSION1_CHARS94<F> to G2
4198 ESC '+' <F> designate DIMENSION1_CHARS94<F> to G3
4199 ESC ',' <F> designate DIMENSION1_CHARS96<F> to G0 (*)
4200 ESC '-' <F> designate DIMENSION1_CHARS96<F> to G1
4201 ESC '.' <F> designate DIMENSION1_CHARS96<F> to G2
4202 ESC '/' <F> designate DIMENSION1_CHARS96<F> to G3
4203 ESC '$' '(' <F> designate DIMENSION2_CHARS94<F> to G0 (**)
4204 ESC '$' ')' <F> designate DIMENSION2_CHARS94<F> to G1
4205 ESC '$' '*' <F> designate DIMENSION2_CHARS94<F> to G2
4206 ESC '$' '+' <F> designate DIMENSION2_CHARS94<F> to G3
4207 ESC '$' ',' <F> designate DIMENSION2_CHARS96<F> to G0 (*)
4208 ESC '$' '-' <F> designate DIMENSION2_CHARS96<F> to G1
4209 ESC '$' '.' <F> designate DIMENSION2_CHARS96<F> to G2
4210 ESC '$' '/' <F> designate DIMENSION2_CHARS96<F> to G3
4211 ----------------------------------------------------------------------
4213 In this list, "DIMENSION1_CHARS94<F>" means a graphic character set
4214 of dimension 1, chars 94, and final character <F>, etc...
4216 Note (*): Although these designations are not allowed in ISO2022,
4217 Emacs accepts them on decoding, and produces them on encoding
4218 CHARS96 character sets in a coding system which is characterized as
4219 7-bit environment, non-locking-shift, and non-single-shift.
4221 Note (**): If <F> is '@', 'A', or 'B', the intermediate character
4222 '(' can be omitted. We refer to this as "short-form" hereafter.
4224 Now you may notice that there are a lot of ways for encoding the
4225 same multilingual text in ISO2022. Actually, there exist many
4226 coding systems such as Compound Text (used in X11's inter client
4227 communication, ISO-2022-JP (used in Japanese internet), ISO-2022-KR
4228 (used in Korean internet), EUC (Extended UNIX Code, used in Asian
4229 localized platforms), and all of these are variants of ISO2022.
4231 In addition to the above, Emacs handles two more kinds of escape
4232 sequences: ISO6429's direction specification and Emacs' private
4233 sequence for specifying character composition.
4235 ISO6429's direction specification takes the following form:
4236 o CSI ']' -- end of the current direction
4237 o CSI '0' ']' -- end of the current direction
4238 o CSI '1' ']' -- start of left-to-right text
4239 o CSI '2' ']' -- start of right-to-left text
4240 The control character CSI (0x9B: control sequence introducer) is
4241 abbreviated to the escape sequence ESC '[' in a 7-bit environment.
4243 Character composition specification takes the following form:
4244 o ESC '0' -- start character composition
4245 o ESC '1' -- end character composition
4246 Since these are not standard escape sequences of any ISO standard,
4247 their use with these meanings is restricted to Emacs only. */
4250 reset_iso2022 (Lisp_Object coding_system, struct iso2022_decoder *iso)
4254 for (i = 0; i < 4; i++)
4256 if (!NILP (coding_system))
4258 XCODING_SYSTEM_ISO2022_INITIAL_CHARSET (coding_system, i);
4260 iso->charset[i] = Qt;
4261 iso->invalid_designated[i] = 0;
4263 iso->esc = ISO_ESC_NOTHING;
4264 iso->esc_bytes_index = 0;
4265 iso->register_left = 0;
4266 iso->register_right = 1;
4267 iso->switched_dir_and_no_valid_charset_yet = 0;
4268 iso->invalid_switch_dir = 0;
4269 iso->output_direction_sequence = 0;
4270 iso->output_literally = 0;
4271 #ifdef ENABLE_COMPOSITE_CHARS
4272 if (iso->composite_chars)
4273 Dynarr_reset (iso->composite_chars);
4278 fit_to_be_escape_quoted (unsigned char c)
4295 /* Parse one byte of an ISO2022 escape sequence.
4296 If the result is an invalid escape sequence, return 0 and
4297 do not change anything in STR. Otherwise, if the result is
4298 an incomplete escape sequence, update ISO2022.ESC and
4299 ISO2022.ESC_BYTES and return -1. Otherwise, update
4300 all the state variables (but not ISO2022.ESC_BYTES) and
4303 If CHECK_INVALID_CHARSETS is non-zero, check for designation
4304 or invocation of an invalid character set and treat that as
4305 an unrecognized escape sequence. */
4308 parse_iso2022_esc (Lisp_Object codesys, struct iso2022_decoder *iso,
4309 unsigned char c, unsigned int *flags,
4310 int check_invalid_charsets)
4312 /* (1) If we're at the end of a designation sequence, CS is the
4313 charset being designated and REG is the register to designate
4316 (2) If we're at the end of a locking-shift sequence, REG is
4317 the register to invoke and HALF (0 == left, 1 == right) is
4318 the half to invoke it into.
4320 (3) If we're at the end of a single-shift sequence, REG is
4321 the register to invoke. */
4322 Lisp_Object cs = Qnil;
4325 /* NOTE: This code does goto's all over the fucking place.
4326 The reason for this is that we're basically implementing
4327 a state machine here, and hierarchical languages like C
4328 don't really provide a clean way of doing this. */
4330 if (! (*flags & CODING_STATE_ESCAPE))
4331 /* At beginning of escape sequence; we need to reset our
4332 escape-state variables. */
4333 iso->esc = ISO_ESC_NOTHING;
4335 iso->output_literally = 0;
4336 iso->output_direction_sequence = 0;
4340 case ISO_ESC_NOTHING:
4341 iso->esc_bytes_index = 0;
4344 case ISO_CODE_ESC: /* Start escape sequence */
4345 *flags |= CODING_STATE_ESCAPE;
4349 case ISO_CODE_CSI: /* ISO6429 (specifying directionality) */
4350 *flags |= CODING_STATE_ESCAPE;
4351 iso->esc = ISO_ESC_5_11;
4354 case ISO_CODE_SO: /* locking shift 1 */
4357 case ISO_CODE_SI: /* locking shift 0 */
4361 case ISO_CODE_SS2: /* single shift */
4364 case ISO_CODE_SS3: /* single shift */
4368 default: /* Other control characters */
4375 /**** single shift ****/
4377 case 'N': /* single shift 2 */
4380 case 'O': /* single shift 3 */
4384 /**** locking shift ****/
4386 case '~': /* locking shift 1 right */
4389 case 'n': /* locking shift 2 */
4392 case '}': /* locking shift 2 right */
4395 case 'o': /* locking shift 3 */
4398 case '|': /* locking shift 3 right */
4402 #ifdef ENABLE_COMPOSITE_CHARS
4403 /**** composite ****/
4406 iso->esc = ISO_ESC_START_COMPOSITE;
4407 *flags = (*flags & CODING_STATE_ISO2022_LOCK) |
4408 CODING_STATE_COMPOSITE;
4412 iso->esc = ISO_ESC_END_COMPOSITE;
4413 *flags = (*flags & CODING_STATE_ISO2022_LOCK) &
4414 ~CODING_STATE_COMPOSITE;
4416 #endif /* ENABLE_COMPOSITE_CHARS */
4418 /**** directionality ****/
4421 iso->esc = ISO_ESC_5_11;
4424 /**** designation ****/
4426 case '$': /* multibyte charset prefix */
4427 iso->esc = ISO_ESC_2_4;
4431 if (0x28 <= c && c <= 0x2F)
4433 iso->esc = (enum iso_esc_flag) (c - 0x28 + ISO_ESC_2_8);
4437 /* This function is called with CODESYS equal to nil when
4438 doing coding-system detection. */
4440 && XCODING_SYSTEM_ISO2022_ESCAPE_QUOTED (codesys)
4441 && fit_to_be_escape_quoted (c))
4443 iso->esc = ISO_ESC_LITERAL;
4444 *flags &= CODING_STATE_ISO2022_LOCK;
4454 /**** directionality ****/
4456 case ISO_ESC_5_11: /* ISO6429 direction control */
4459 *flags &= (CODING_STATE_ISO2022_LOCK & ~CODING_STATE_R2L);
4460 goto directionality;
4462 if (c == '0') iso->esc = ISO_ESC_5_11_0;
4463 else if (c == '1') iso->esc = ISO_ESC_5_11_1;
4464 else if (c == '2') iso->esc = ISO_ESC_5_11_2;
4468 case ISO_ESC_5_11_0:
4471 *flags &= (CODING_STATE_ISO2022_LOCK & ~CODING_STATE_R2L);
4472 goto directionality;
4476 case ISO_ESC_5_11_1:
4479 *flags = (CODING_STATE_ISO2022_LOCK & ~CODING_STATE_R2L);
4480 goto directionality;
4484 case ISO_ESC_5_11_2:
4487 *flags = (*flags & CODING_STATE_ISO2022_LOCK) | CODING_STATE_R2L;
4488 goto directionality;
4493 iso->esc = ISO_ESC_DIRECTIONALITY;
4494 /* Various junk here to attempt to preserve the direction sequences
4495 literally in the text if they would otherwise be swallowed due
4496 to invalid designations that don't show up as actual charset
4497 changes in the text. */
4498 if (iso->invalid_switch_dir)
4500 /* We already inserted a direction switch literally into the
4501 text. We assume (#### this may not be right) that the
4502 next direction switch is the one going the other way,
4503 and we need to output that literally as well. */
4504 iso->output_literally = 1;
4505 iso->invalid_switch_dir = 0;
4511 /* If we are in the thrall of an invalid designation,
4512 then stick the directionality sequence literally into the
4513 output stream so it ends up in the original text again. */
4514 for (jj = 0; jj < 4; jj++)
4515 if (iso->invalid_designated[jj])
4519 iso->output_literally = 1;
4520 iso->invalid_switch_dir = 1;
4523 /* Indicate that we haven't yet seen a valid designation,
4524 so that if a switch-dir is directly followed by an
4525 invalid designation, both get inserted literally. */
4526 iso->switched_dir_and_no_valid_charset_yet = 1;
4531 /**** designation ****/
4534 if (0x28 <= c && c <= 0x2F)
4536 iso->esc = (enum iso_esc_flag) (c - 0x28 + ISO_ESC_2_4_8);
4539 if (0x40 <= c && c <= 0x42)
4542 cs = CHARSET_BY_ATTRIBUTES (94, -1, c,
4543 *flags & CODING_STATE_R2L ?
4544 CHARSET_RIGHT_TO_LEFT :
4545 CHARSET_LEFT_TO_RIGHT);
4556 if (c < '0' || c > '~')
4557 return 0; /* bad final byte */
4559 if (iso->esc >= ISO_ESC_2_8 &&
4560 iso->esc <= ISO_ESC_2_15)
4562 chars = (iso->esc >= ISO_ESC_2_12) ? 96 : 94;
4563 single = 1; /* single-byte */
4564 reg = (iso->esc - ISO_ESC_2_8) & 3;
4566 else if (iso->esc >= ISO_ESC_2_4_8 &&
4567 iso->esc <= ISO_ESC_2_4_15)
4569 chars = (iso->esc >= ISO_ESC_2_4_12) ? 96 : 94;
4570 single = -1; /* multi-byte */
4571 reg = (iso->esc - ISO_ESC_2_4_8) & 3;
4575 /* Can this ever be reached? -slb */
4579 cs = CHARSET_BY_ATTRIBUTES (chars, single, c,
4580 *flags & CODING_STATE_R2L ?
4581 CHARSET_RIGHT_TO_LEFT :
4582 CHARSET_LEFT_TO_RIGHT);
4588 iso->esc_bytes[iso->esc_bytes_index++] = (unsigned char) c;
4592 if (check_invalid_charsets && !CHARSETP (iso->charset[reg]))
4593 /* can't invoke something that ain't there. */
4595 iso->esc = ISO_ESC_SINGLE_SHIFT;
4596 *flags &= CODING_STATE_ISO2022_LOCK;
4598 *flags |= CODING_STATE_SS2;
4600 *flags |= CODING_STATE_SS3;
4604 if (check_invalid_charsets &&
4605 !CHARSETP (iso->charset[reg]))
4606 /* can't invoke something that ain't there. */
4609 iso->register_right = reg;
4611 iso->register_left = reg;
4612 *flags &= CODING_STATE_ISO2022_LOCK;
4613 iso->esc = ISO_ESC_LOCKING_SHIFT;
4617 if (NILP (cs) && check_invalid_charsets)
4619 iso->invalid_designated[reg] = 1;
4620 iso->charset[reg] = Vcharset_ascii;
4621 iso->esc = ISO_ESC_DESIGNATE;
4622 *flags &= CODING_STATE_ISO2022_LOCK;
4623 iso->output_literally = 1;
4624 if (iso->switched_dir_and_no_valid_charset_yet)
4626 /* We encountered a switch-direction followed by an
4627 invalid designation. Ensure that the switch-direction
4628 gets outputted; otherwise it will probably get eaten
4629 when the text is written out again. */
4630 iso->switched_dir_and_no_valid_charset_yet = 0;
4631 iso->output_direction_sequence = 1;
4632 /* And make sure that the switch-dir going the other
4633 way gets outputted, as well. */
4634 iso->invalid_switch_dir = 1;
4638 /* This function is called with CODESYS equal to nil when
4639 doing coding-system detection. */
4640 if (!NILP (codesys))
4642 charset_conversion_spec_dynarr *dyn =
4643 XCODING_SYSTEM (codesys)->iso2022.input_conv;
4649 for (i = 0; i < Dynarr_length (dyn); i++)
4651 struct charset_conversion_spec *spec = Dynarr_atp (dyn, i);
4652 if (EQ (cs, spec->from_charset))
4653 cs = spec->to_charset;
4658 iso->charset[reg] = cs;
4659 iso->esc = ISO_ESC_DESIGNATE;
4660 *flags &= CODING_STATE_ISO2022_LOCK;
4661 if (iso->invalid_designated[reg])
4663 iso->invalid_designated[reg] = 0;
4664 iso->output_literally = 1;
4666 if (iso->switched_dir_and_no_valid_charset_yet)
4667 iso->switched_dir_and_no_valid_charset_yet = 0;
4672 detect_coding_iso2022 (struct detection_state *st, const Extbyte *src, size_t n)
4676 /* #### There are serious deficiencies in the recognition mechanism
4677 here. This needs to be much smarter if it's going to cut it.
4678 The sequence "\xff\x0f" is currently detected as LOCK_SHIFT while
4679 it should be detected as Latin-1.
4680 All the ISO2022 stuff in this file should be synced up with the
4681 code from FSF Emacs-20.4, in which Mule should be more or less stable.
4682 Perhaps we should wait till R2L works in FSF Emacs? */
4684 if (!st->iso2022.initted)
4686 reset_iso2022 (Qnil, &st->iso2022.iso);
4687 st->iso2022.mask = (CODING_CATEGORY_ISO_7_MASK |
4688 CODING_CATEGORY_ISO_8_DESIGNATE_MASK |
4689 CODING_CATEGORY_ISO_8_1_MASK |
4690 CODING_CATEGORY_ISO_8_2_MASK |
4691 CODING_CATEGORY_ISO_LOCK_SHIFT_MASK);
4692 st->iso2022.flags = 0;
4693 st->iso2022.high_byte_count = 0;
4694 st->iso2022.saw_single_shift = 0;
4695 st->iso2022.initted = 1;
4698 mask = st->iso2022.mask;
4702 unsigned char c = *(unsigned char *)src++;
4705 mask &= ~CODING_CATEGORY_ISO_7_MASK;
4706 st->iso2022.high_byte_count++;
4710 if (st->iso2022.high_byte_count && !st->iso2022.saw_single_shift)
4712 if (st->iso2022.high_byte_count & 1)
4713 /* odd number of high bytes; assume not iso-8-2 */
4714 mask &= ~CODING_CATEGORY_ISO_8_2_MASK;
4716 st->iso2022.high_byte_count = 0;
4717 st->iso2022.saw_single_shift = 0;
4719 mask &= ~CODING_CATEGORY_ISO_7_MASK;
4721 if (!(st->iso2022.flags & CODING_STATE_ESCAPE)
4722 && (BYTE_C0_P (c) || BYTE_C1_P (c)))
4723 { /* control chars */
4726 /* Allow and ignore control characters that you might
4727 reasonably see in a text file */
4732 case 8: /* backspace */
4733 case 11: /* vertical tab */
4734 case 12: /* form feed */
4735 case 26: /* MS-DOS C-z junk */
4736 case 31: /* '^_' -- for info */
4737 goto label_continue_loop;
4744 if ((st->iso2022.flags & CODING_STATE_ESCAPE) || BYTE_C0_P (c)
4747 if (parse_iso2022_esc (Qnil, &st->iso2022.iso, c,
4748 &st->iso2022.flags, 0))
4750 switch (st->iso2022.iso.esc)
4752 case ISO_ESC_DESIGNATE:
4753 mask &= ~CODING_CATEGORY_ISO_8_1_MASK;
4754 mask &= ~CODING_CATEGORY_ISO_8_2_MASK;
4756 case ISO_ESC_LOCKING_SHIFT:
4757 mask = CODING_CATEGORY_ISO_LOCK_SHIFT_MASK;
4758 goto ran_out_of_chars;
4759 case ISO_ESC_SINGLE_SHIFT:
4760 mask &= ~CODING_CATEGORY_ISO_8_DESIGNATE_MASK;
4761 st->iso2022.saw_single_shift = 1;
4770 goto ran_out_of_chars;
4773 label_continue_loop:;
4782 postprocess_iso2022_mask (int mask)
4784 /* #### kind of cheesy */
4785 /* If seven-bit ISO is allowed, then assume that the encoding is
4786 entirely seven-bit and turn off the eight-bit ones. */
4787 if (mask & CODING_CATEGORY_ISO_7_MASK)
4788 mask &= ~ (CODING_CATEGORY_ISO_8_DESIGNATE_MASK |
4789 CODING_CATEGORY_ISO_8_1_MASK |
4790 CODING_CATEGORY_ISO_8_2_MASK);
4794 /* If FLAGS is a null pointer or specifies right-to-left motion,
4795 output a switch-dir-to-left-to-right sequence to DST.
4796 Also update FLAGS if it is not a null pointer.
4797 If INTERNAL_P is set, we are outputting in internal format and
4798 need to handle the CSI differently. */
4801 restore_left_to_right_direction (Lisp_Coding_System *codesys,
4802 unsigned_char_dynarr *dst,
4803 unsigned int *flags,
4806 if (!flags || (*flags & CODING_STATE_R2L))
4808 if (CODING_SYSTEM_ISO2022_SEVEN (codesys))
4810 Dynarr_add (dst, ISO_CODE_ESC);
4811 Dynarr_add (dst, '[');
4813 else if (internal_p)
4814 DECODE_ADD_BINARY_CHAR (ISO_CODE_CSI, dst);
4816 Dynarr_add (dst, ISO_CODE_CSI);
4817 Dynarr_add (dst, '0');
4818 Dynarr_add (dst, ']');
4820 *flags &= ~CODING_STATE_R2L;
4824 /* If FLAGS is a null pointer or specifies a direction different from
4825 DIRECTION (which should be either CHARSET_RIGHT_TO_LEFT or
4826 CHARSET_LEFT_TO_RIGHT), output the appropriate switch-dir escape
4827 sequence to DST. Also update FLAGS if it is not a null pointer.
4828 If INTERNAL_P is set, we are outputting in internal format and
4829 need to handle the CSI differently. */
4832 ensure_correct_direction (int direction, Lisp_Coding_System *codesys,
4833 unsigned_char_dynarr *dst, unsigned int *flags,
4836 if ((!flags || (*flags & CODING_STATE_R2L)) &&
4837 direction == CHARSET_LEFT_TO_RIGHT)
4838 restore_left_to_right_direction (codesys, dst, flags, internal_p);
4839 else if (!CODING_SYSTEM_ISO2022_NO_ISO6429 (codesys)
4840 && (!flags || !(*flags & CODING_STATE_R2L)) &&
4841 direction == CHARSET_RIGHT_TO_LEFT)
4843 if (CODING_SYSTEM_ISO2022_SEVEN (codesys))
4845 Dynarr_add (dst, ISO_CODE_ESC);
4846 Dynarr_add (dst, '[');
4848 else if (internal_p)
4849 DECODE_ADD_BINARY_CHAR (ISO_CODE_CSI, dst);
4851 Dynarr_add (dst, ISO_CODE_CSI);
4852 Dynarr_add (dst, '2');
4853 Dynarr_add (dst, ']');
4855 *flags |= CODING_STATE_R2L;
4859 /* Convert ISO2022-format data to internal format. */
4862 decode_coding_iso2022 (Lstream *decoding, const Extbyte *src,
4863 unsigned_char_dynarr *dst, size_t n)
4865 struct decoding_stream *str = DECODING_STREAM_DATA (decoding);
4866 unsigned int flags = str->flags;
4867 unsigned int cpos = str->cpos;
4868 unsigned char counter = str->counter;
4869 eol_type_t eol_type = str->eol_type;
4870 #ifdef ENABLE_COMPOSITE_CHARS
4871 unsigned_char_dynarr *real_dst = dst;
4873 Lisp_Object coding_system;
4875 XSETCODING_SYSTEM (coding_system, str->codesys);
4877 #ifdef ENABLE_COMPOSITE_CHARS
4878 if (flags & CODING_STATE_COMPOSITE)
4879 dst = str->iso2022.composite_chars;
4880 #endif /* ENABLE_COMPOSITE_CHARS */
4884 unsigned char c = *(unsigned char *)src++;
4885 if (flags & CODING_STATE_ESCAPE)
4886 { /* Within ESC sequence */
4887 int retval = parse_iso2022_esc (coding_system, &str->iso2022,
4892 switch (str->iso2022.esc)
4894 #ifdef ENABLE_COMPOSITE_CHARS
4895 case ISO_ESC_START_COMPOSITE:
4896 if (str->iso2022.composite_chars)
4897 Dynarr_reset (str->iso2022.composite_chars);
4899 str->iso2022.composite_chars = Dynarr_new (unsigned_char);
4900 dst = str->iso2022.composite_chars;
4902 case ISO_ESC_END_COMPOSITE:
4904 Bufbyte comstr[MAX_EMCHAR_LEN];
4906 Emchar emch = lookup_composite_char (Dynarr_atp (dst, 0),
4907 Dynarr_length (dst));
4909 len = set_charptr_emchar (comstr, emch);
4910 Dynarr_add_many (dst, comstr, len);
4913 #endif /* ENABLE_COMPOSITE_CHARS */
4915 case ISO_ESC_LITERAL:
4916 COMPOSE_FLUSH_CHARS (str, dst);
4917 DECODE_ADD_BINARY_CHAR (c, dst);
4921 /* Everything else handled already */
4926 /* Attempted error recovery. */
4927 if (str->iso2022.output_direction_sequence)
4928 ensure_correct_direction (flags & CODING_STATE_R2L ?
4929 CHARSET_RIGHT_TO_LEFT :
4930 CHARSET_LEFT_TO_RIGHT,
4931 str->codesys, dst, 0, 1);
4932 /* More error recovery. */
4933 if (!retval || str->iso2022.output_literally)
4935 /* Output the (possibly invalid) sequence */
4937 COMPOSE_FLUSH_CHARS (str, dst);
4938 for (i = 0; i < str->iso2022.esc_bytes_index; i++)
4939 DECODE_ADD_BINARY_CHAR (str->iso2022.esc_bytes[i], dst);
4940 flags &= CODING_STATE_ISO2022_LOCK;
4942 n++, src--;/* Repeat the loop with the same character. */
4945 /* No sense in reprocessing the final byte of the
4946 escape sequence; it could mess things up anyway.
4948 COMPOSE_FLUSH_CHARS (str, dst);
4949 DECODE_ADD_BINARY_CHAR (c, dst);
4955 else if (BYTE_C0_P (c) || BYTE_C1_P (c))
4956 { /* Control characters */
4958 /***** Error-handling *****/
4960 /* If we were in the middle of a character, dump out the
4961 partial character. */
4964 COMPOSE_FLUSH_CHARS (str, dst);
4968 DECODE_ADD_BINARY_CHAR
4969 ((unsigned char)(cpos >> (counter * 8)), dst);
4974 /* If we just saw a single-shift character, dump it out.
4975 This may dump out the wrong sort of single-shift character,
4976 but least it will give an indication that something went
4978 if (flags & CODING_STATE_SS2)
4980 COMPOSE_FLUSH_CHARS (str, dst);
4981 DECODE_ADD_BINARY_CHAR (ISO_CODE_SS2, dst);
4982 flags &= ~CODING_STATE_SS2;
4984 if (flags & CODING_STATE_SS3)
4986 COMPOSE_FLUSH_CHARS (str, dst);
4987 DECODE_ADD_BINARY_CHAR (ISO_CODE_SS3, dst);
4988 flags &= ~CODING_STATE_SS3;
4991 /***** Now handle the control characters. *****/
4997 COMPOSE_FLUSH_CHARS (str, dst);
4998 if (eol_type == EOL_CR)
4999 Dynarr_add (dst, '\n');
5000 else if (eol_type != EOL_CRLF || flags & CODING_STATE_CR)
5001 Dynarr_add (dst, c);
5003 flags |= CODING_STATE_CR;
5004 goto label_continue_loop;
5006 else if (flags & CODING_STATE_CR)
5007 { /* eol_type == CODING_SYSTEM_EOL_CRLF */
5009 Dynarr_add (dst, '\r');
5010 flags &= ~CODING_STATE_CR;
5013 DECODE_HANDLE_EOL_TYPE (eol_type, c, flags, dst);
5016 flags &= CODING_STATE_ISO2022_LOCK;
5018 if (!parse_iso2022_esc (coding_system, &str->iso2022, c, &flags, 1))
5020 COMPOSE_FLUSH_CHARS (str, dst);
5021 DECODE_ADD_BINARY_CHAR (c, dst);
5025 { /* Graphic characters */
5026 Lisp_Object charset;
5035 COMPOSE_FLUSH_CHARS (str, dst);
5036 if (eol_type == EOL_CR)
5037 Dynarr_add (dst, '\n');
5038 else if (eol_type != EOL_CRLF || flags & CODING_STATE_CR)
5039 Dynarr_add (dst, c);
5041 flags |= CODING_STATE_CR;
5042 goto label_continue_loop;
5044 else if (flags & CODING_STATE_CR)
5045 { /* eol_type == CODING_SYSTEM_EOL_CRLF */
5047 Dynarr_add (dst, '\r');
5048 flags &= ~CODING_STATE_CR;
5051 DECODE_HANDLE_EOL_TYPE (eol_type, c, flags, dst);
5054 /* Now determine the charset. */
5055 reg = ((flags & CODING_STATE_SS2) ? 2
5056 : (flags & CODING_STATE_SS3) ? 3
5057 : !BYTE_ASCII_P (c) ? str->iso2022.register_right
5058 : str->iso2022.register_left);
5059 charset = str->iso2022.charset[reg];
5061 /* Error checking: */
5062 if (! CHARSETP (charset)
5063 || str->iso2022.invalid_designated[reg]
5064 || (((c & 0x7F) == ' ' || (c & 0x7F) == ISO_CODE_DEL)
5065 && XCHARSET_CHARS (charset) == 94))
5066 /* Mrmph. We are trying to invoke a register that has no
5067 or an invalid charset in it, or trying to add a character
5068 outside the range of the charset. Insert that char literally
5069 to preserve it for the output. */
5071 COMPOSE_FLUSH_CHARS (str, dst);
5075 DECODE_ADD_BINARY_CHAR
5076 ((unsigned char)(cpos >> (counter * 8)), dst);
5079 DECODE_ADD_BINARY_CHAR (c, dst);
5084 /* Things are probably hunky-dorey. */
5086 /* Fetch reverse charset, maybe. */
5087 if (((flags & CODING_STATE_R2L) &&
5088 XCHARSET_DIRECTION (charset) == CHARSET_LEFT_TO_RIGHT)
5090 (!(flags & CODING_STATE_R2L) &&
5091 XCHARSET_DIRECTION (charset) == CHARSET_RIGHT_TO_LEFT))
5093 Lisp_Object new_charset =
5094 XCHARSET_REVERSE_DIRECTION_CHARSET (charset);
5095 if (!NILP (new_charset))
5096 charset = new_charset;
5101 if (XCHARSET_DIMENSION (charset) == counter)
5103 COMPOSE_ADD_CHAR (str,
5104 DECODE_CHAR (charset,
5105 ((cpos & 0x7F7F7F) << 8)
5112 cpos = (cpos << 8) | c;
5114 lb = XCHARSET_LEADING_BYTE (charset);
5115 switch (XCHARSET_REP_BYTES (charset))
5118 DECODE_OUTPUT_PARTIAL_CHAR (ch);
5119 Dynarr_add (dst, c & 0x7F);
5122 case 2: /* one-byte official */
5123 DECODE_OUTPUT_PARTIAL_CHAR (ch);
5124 Dynarr_add (dst, lb);
5125 Dynarr_add (dst, c | 0x80);
5128 case 3: /* one-byte private or two-byte official */
5129 if (XCHARSET_PRIVATE_P (charset))
5131 DECODE_OUTPUT_PARTIAL_CHAR (ch);
5132 Dynarr_add (dst, PRE_LEADING_BYTE_PRIVATE_1);
5133 Dynarr_add (dst, lb);
5134 Dynarr_add (dst, c | 0x80);
5140 Dynarr_add (dst, lb);
5141 Dynarr_add (dst, ch | 0x80);
5142 Dynarr_add (dst, c | 0x80);
5150 default: /* two-byte private */
5153 Dynarr_add (dst, PRE_LEADING_BYTE_PRIVATE_2);
5154 Dynarr_add (dst, lb);
5155 Dynarr_add (dst, ch | 0x80);
5156 Dynarr_add (dst, c | 0x80);
5166 flags &= CODING_STATE_ISO2022_LOCK;
5169 label_continue_loop:;
5172 if (flags & CODING_STATE_END)
5174 COMPOSE_FLUSH_CHARS (str, dst);
5175 DECODE_OUTPUT_PARTIAL_CHAR (cpos);
5179 str->counter = counter;
5183 /***** ISO2022 encoder *****/
5185 /* Designate CHARSET into register REG. */
5188 iso2022_designate (Lisp_Object charset, unsigned char reg,
5189 struct encoding_stream *str, unsigned_char_dynarr *dst)
5191 static const char inter94[] = "()*+";
5192 static const char inter96[] = ",-./";
5193 unsigned short chars;
5194 unsigned char dimension;
5195 unsigned char final;
5196 Lisp_Object old_charset = str->iso2022.charset[reg];
5198 str->iso2022.charset[reg] = charset;
5199 if (!CHARSETP (charset))
5200 /* charset might be an initial nil or t. */
5202 chars = XCHARSET_CHARS (charset);
5203 dimension = XCHARSET_DIMENSION (charset);
5204 final = XCHARSET_FINAL (charset);
5205 if (!str->iso2022.force_charset_on_output[reg] &&
5206 CHARSETP (old_charset) &&
5207 XCHARSET_CHARS (old_charset) == chars &&
5208 XCHARSET_DIMENSION (old_charset) == dimension &&
5209 XCHARSET_FINAL (old_charset) == final)
5212 str->iso2022.force_charset_on_output[reg] = 0;
5215 charset_conversion_spec_dynarr *dyn =
5216 str->codesys->iso2022.output_conv;
5222 for (i = 0; i < Dynarr_length (dyn); i++)
5224 struct charset_conversion_spec *spec = Dynarr_atp (dyn, i);
5225 if (EQ (charset, spec->from_charset))
5226 charset = spec->to_charset;
5231 Dynarr_add (dst, ISO_CODE_ESC);
5236 Dynarr_add (dst, inter94[reg]);
5239 Dynarr_add (dst, '$');
5241 || !(CODING_SYSTEM_ISO2022_SHORT (str->codesys))
5244 Dynarr_add (dst, inter94[reg]);
5249 Dynarr_add (dst, inter96[reg]);
5252 Dynarr_add (dst, '$');
5253 Dynarr_add (dst, inter96[reg]);
5257 Dynarr_add (dst, final);
5261 ensure_normal_shift (struct encoding_stream *str, unsigned_char_dynarr *dst)
5263 if (str->iso2022.register_left != 0)
5265 Dynarr_add (dst, ISO_CODE_SI);
5266 str->iso2022.register_left = 0;
5271 ensure_shift_out (struct encoding_stream *str, unsigned_char_dynarr *dst)
5273 if (str->iso2022.register_left != 1)
5275 Dynarr_add (dst, ISO_CODE_SO);
5276 str->iso2022.register_left = 1;
5281 char_encode_iso2022 (struct encoding_stream *str, Emchar ch,
5282 unsigned_char_dynarr *dst, unsigned int *flags)
5284 unsigned char charmask;
5285 Lisp_Coding_System* codesys = str->codesys;
5286 eol_type_t eol_type = CODING_SYSTEM_EOL_TYPE (str->codesys);
5288 Lisp_Object charset = str->iso2022.current_charset;
5289 int half = str->iso2022.current_half;
5290 int code_point = -1;
5294 restore_left_to_right_direction (codesys, dst, flags, 0);
5296 /* Make sure G0 contains ASCII */
5297 if ((ch > ' ' && ch < ISO_CODE_DEL)
5298 || !CODING_SYSTEM_ISO2022_NO_ASCII_CNTL (codesys))
5300 ensure_normal_shift (str, dst);
5301 iso2022_designate (Vcharset_ascii, 0, str, dst);
5304 /* If necessary, restore everything to the default state
5306 if (ch == '\n' && !(CODING_SYSTEM_ISO2022_NO_ASCII_EOL (codesys)))
5308 restore_left_to_right_direction (codesys, dst, flags, 0);
5310 ensure_normal_shift (str, dst);
5312 for (i = 0; i < 4; i++)
5314 Lisp_Object initial_charset =
5315 CODING_SYSTEM_ISO2022_INITIAL_CHARSET (codesys, i);
5316 iso2022_designate (initial_charset, i, str, dst);
5321 if (eol_type != EOL_LF && eol_type != EOL_AUTODETECT)
5322 Dynarr_add (dst, '\r');
5323 if (eol_type != EOL_CR)
5324 Dynarr_add (dst, ch);
5328 if (CODING_SYSTEM_ISO2022_ESCAPE_QUOTED (codesys)
5329 && fit_to_be_escape_quoted (ch))
5330 Dynarr_add (dst, ISO_CODE_ESC);
5331 Dynarr_add (dst, ch);
5334 else if ( (0x80 <= ch) && (ch <= 0x9f) )
5336 charmask = (half == 0 ? 0x00 : 0x80);
5338 if (CODING_SYSTEM_ISO2022_ESCAPE_QUOTED (codesys)
5339 && fit_to_be_escape_quoted (ch))
5340 Dynarr_add (dst, ISO_CODE_ESC);
5341 /* you asked for it ... */
5342 Dynarr_add (dst, ch);
5348 /* Now determine which register to use. */
5350 for (i = 0; i < 4; i++)
5352 if ((CHARSETP (charset = str->iso2022.charset[i])
5353 && ((code_point = charset_code_point (charset, ch)) >= 0))
5357 = CODING_SYSTEM_ISO2022_INITIAL_CHARSET (codesys, i))
5358 && ((code_point = charset_code_point (charset, ch)) >= 0)))
5366 Lisp_Object original_default_coded_charset_priority_list
5367 = Vdefault_coded_charset_priority_list;
5369 while (!EQ (Vdefault_coded_charset_priority_list, Qnil))
5371 code_point = ENCODE_CHAR (ch, charset);
5372 if (XCHARSET_FINAL (charset))
5374 Vdefault_coded_charset_priority_list
5375 = Fcdr (Fmemq (XCHARSET_NAME (charset),
5376 Vdefault_coded_charset_priority_list));
5378 code_point = ENCODE_CHAR (ch, charset);
5379 if (!XCHARSET_FINAL (charset))
5381 charset = Vcharset_ascii;
5385 Vdefault_coded_charset_priority_list
5386 = original_default_coded_charset_priority_list;
5388 ensure_correct_direction (XCHARSET_DIRECTION (charset),
5389 codesys, dst, flags, 0);
5393 if (XCHARSET_GRAPHIC (charset) != 0)
5395 if (!NILP (str->iso2022.charset[1]) &&
5396 (!CODING_SYSTEM_ISO2022_SEVEN (codesys)
5397 || CODING_SYSTEM_ISO2022_LOCK_SHIFT (codesys)))
5399 else if (!NILP (str->iso2022.charset[2]))
5401 else if (!NILP (str->iso2022.charset[3]))
5410 iso2022_designate (charset, reg, str, dst);
5412 /* Now invoke that register. */
5416 ensure_normal_shift (str, dst);
5420 if (CODING_SYSTEM_ISO2022_SEVEN (codesys))
5422 ensure_shift_out (str, dst);
5429 if (CODING_SYSTEM_ISO2022_SEVEN (str->codesys))
5431 Dynarr_add (dst, ISO_CODE_ESC);
5432 Dynarr_add (dst, 'N');
5437 Dynarr_add (dst, ISO_CODE_SS2);
5442 if (CODING_SYSTEM_ISO2022_SEVEN (str->codesys))
5444 Dynarr_add (dst, ISO_CODE_ESC);
5445 Dynarr_add (dst, 'O');
5450 Dynarr_add (dst, ISO_CODE_SS3);
5458 charmask = (half == 0 ? 0x00 : 0x80);
5460 switch (XCHARSET_DIMENSION (charset))
5463 Dynarr_add (dst, (code_point & 0xFF) | charmask);
5466 Dynarr_add (dst, ((code_point >> 8) & 0xFF) | charmask);
5467 Dynarr_add (dst, ( code_point & 0xFF) | charmask);
5470 Dynarr_add (dst, ((code_point >> 16) & 0xFF) | charmask);
5471 Dynarr_add (dst, ((code_point >> 8) & 0xFF) | charmask);
5472 Dynarr_add (dst, ( code_point & 0xFF) | charmask);
5475 Dynarr_add (dst, ((code_point >> 24) & 0xFF) | charmask);
5476 Dynarr_add (dst, ((code_point >> 16) & 0xFF) | charmask);
5477 Dynarr_add (dst, ((code_point >> 8) & 0xFF) | charmask);
5478 Dynarr_add (dst, ( code_point & 0xFF) | charmask);
5484 str->iso2022.current_charset = charset;
5485 str->iso2022.current_half = half;
5489 char_finish_iso2022 (struct encoding_stream *str, unsigned_char_dynarr *dst,
5490 unsigned int *flags)
5492 Lisp_Coding_System* codesys = str->codesys;
5495 restore_left_to_right_direction (codesys, dst, flags, 0);
5496 ensure_normal_shift (str, dst);
5497 for (i = 0; i < 4; i++)
5499 Lisp_Object initial_charset
5500 = CODING_SYSTEM_ISO2022_INITIAL_CHARSET (codesys, i);
5501 iso2022_designate (initial_charset, i, str, dst);
5506 /************************************************************************/
5507 /* No-conversion methods */
5508 /************************************************************************/
5510 /* This is used when reading in "binary" files -- i.e. files that may
5511 contain all 256 possible byte values and that are not to be
5512 interpreted as being in any particular decoding. */
5514 decode_coding_no_conversion (Lstream *decoding, const Extbyte *src,
5515 unsigned_char_dynarr *dst, size_t n)
5517 struct decoding_stream *str = DECODING_STREAM_DATA (decoding);
5518 unsigned int flags = str->flags;
5519 unsigned int cpos = str->cpos;
5520 eol_type_t eol_type = str->eol_type;
5524 unsigned char c = *(unsigned char *)src++;
5526 DECODE_HANDLE_EOL_TYPE (eol_type, c, flags, dst);
5527 DECODE_ADD_BINARY_CHAR (c, dst);
5528 label_continue_loop:;
5531 DECODE_HANDLE_END_OF_CONVERSION (flags, cpos, dst);
5538 encode_coding_no_conversion (Lstream *encoding, const Bufbyte *src,
5539 unsigned_char_dynarr *dst, size_t n)
5542 struct encoding_stream *str = ENCODING_STREAM_DATA (encoding);
5543 unsigned int flags = str->flags;
5544 unsigned int ch = str->ch;
5545 eol_type_t eol_type = CODING_SYSTEM_EOL_TYPE (str->codesys);
5547 unsigned char char_boundary = str->iso2022.current_char_boundary;
5554 if (char_boundary == 0)
5560 else if ( c >= 0xf8 )
5565 else if ( c >= 0xf0 )
5570 else if ( c >= 0xe0 )
5575 else if ( c >= 0xc0 )
5585 if (eol_type != EOL_LF && eol_type != EOL_AUTODETECT)
5586 Dynarr_add (dst, '\r');
5587 if (eol_type != EOL_CR)
5588 Dynarr_add (dst, c);
5591 Dynarr_add (dst, c);
5594 else if (char_boundary == 1)
5596 ch = ( ch << 6 ) | ( c & 0x3f );
5597 Dynarr_add (dst, ch & 0xff);
5602 ch = ( ch << 6 ) | ( c & 0x3f );
5605 #else /* not UTF2000 */
5608 if (eol_type != EOL_LF && eol_type != EOL_AUTODETECT)
5609 Dynarr_add (dst, '\r');
5610 if (eol_type != EOL_CR)
5611 Dynarr_add (dst, '\n');
5614 else if (BYTE_ASCII_P (c))
5617 Dynarr_add (dst, c);
5619 else if (BUFBYTE_LEADING_BYTE_P (c))
5622 if (c == LEADING_BYTE_LATIN_ISO8859_1 ||
5623 c == LEADING_BYTE_CONTROL_1)
5626 Dynarr_add (dst, '~'); /* untranslatable character */
5630 if (ch == LEADING_BYTE_LATIN_ISO8859_1)
5631 Dynarr_add (dst, c);
5632 else if (ch == LEADING_BYTE_CONTROL_1)
5635 Dynarr_add (dst, c - 0x20);
5637 /* else it should be the second or third byte of an
5638 untranslatable character, so ignore it */
5641 #endif /* not UTF2000 */
5647 str->iso2022.current_char_boundary = char_boundary;
5653 /************************************************************************/
5654 /* Initialization */
5655 /************************************************************************/
5658 syms_of_file_coding (void)
5660 INIT_LRECORD_IMPLEMENTATION (coding_system);
5662 deferror (&Qcoding_system_error, "coding-system-error",
5663 "Coding-system error", Qio_error);
5665 DEFSUBR (Fcoding_system_p);
5666 DEFSUBR (Ffind_coding_system);
5667 DEFSUBR (Fget_coding_system);
5668 DEFSUBR (Fcoding_system_list);
5669 DEFSUBR (Fcoding_system_name);
5670 DEFSUBR (Fmake_coding_system);
5671 DEFSUBR (Fcopy_coding_system);
5672 DEFSUBR (Fcoding_system_canonical_name_p);
5673 DEFSUBR (Fcoding_system_alias_p);
5674 DEFSUBR (Fcoding_system_aliasee);
5675 DEFSUBR (Fdefine_coding_system_alias);
5676 DEFSUBR (Fsubsidiary_coding_system);
5678 DEFSUBR (Fcoding_system_type);
5679 DEFSUBR (Fcoding_system_doc_string);
5681 DEFSUBR (Fcoding_system_charset);
5683 DEFSUBR (Fcoding_system_property);
5685 DEFSUBR (Fcoding_category_list);
5686 DEFSUBR (Fset_coding_priority_list);
5687 DEFSUBR (Fcoding_priority_list);
5688 DEFSUBR (Fset_coding_category_system);
5689 DEFSUBR (Fcoding_category_system);
5691 DEFSUBR (Fdetect_coding_region);
5692 DEFSUBR (Fdecode_coding_region);
5693 DEFSUBR (Fencode_coding_region);
5695 DEFSUBR (Fdecode_shift_jis_char);
5696 DEFSUBR (Fencode_shift_jis_char);
5697 DEFSUBR (Fdecode_big5_char);
5698 DEFSUBR (Fencode_big5_char);
5700 defsymbol (&Qcoding_systemp, "coding-system-p");
5701 defsymbol (&Qno_conversion, "no-conversion");
5702 defsymbol (&Qraw_text, "raw-text");
5704 defsymbol (&Qbig5, "big5");
5705 defsymbol (&Qshift_jis, "shift-jis");
5706 defsymbol (&Qucs4, "ucs-4");
5707 defsymbol (&Qutf8, "utf-8");
5708 defsymbol (&Qccl, "ccl");
5709 defsymbol (&Qiso2022, "iso2022");
5711 defsymbol (&Qmnemonic, "mnemonic");
5712 defsymbol (&Qeol_type, "eol-type");
5713 defsymbol (&Qpost_read_conversion, "post-read-conversion");
5714 defsymbol (&Qpre_write_conversion, "pre-write-conversion");
5716 defsymbol (&Qcr, "cr");
5717 defsymbol (&Qlf, "lf");
5718 defsymbol (&Qcrlf, "crlf");
5719 defsymbol (&Qeol_cr, "eol-cr");
5720 defsymbol (&Qeol_lf, "eol-lf");
5721 defsymbol (&Qeol_crlf, "eol-crlf");
5723 defsymbol (&Qcharset_g0, "charset-g0");
5724 defsymbol (&Qcharset_g1, "charset-g1");
5725 defsymbol (&Qcharset_g2, "charset-g2");
5726 defsymbol (&Qcharset_g3, "charset-g3");
5727 defsymbol (&Qforce_g0_on_output, "force-g0-on-output");
5728 defsymbol (&Qforce_g1_on_output, "force-g1-on-output");
5729 defsymbol (&Qforce_g2_on_output, "force-g2-on-output");
5730 defsymbol (&Qforce_g3_on_output, "force-g3-on-output");
5731 defsymbol (&Qno_iso6429, "no-iso6429");
5732 defsymbol (&Qinput_charset_conversion, "input-charset-conversion");
5733 defsymbol (&Qoutput_charset_conversion, "output-charset-conversion");
5735 defsymbol (&Qshort, "short");
5736 defsymbol (&Qno_ascii_eol, "no-ascii-eol");
5737 defsymbol (&Qno_ascii_cntl, "no-ascii-cntl");
5738 defsymbol (&Qseven, "seven");
5739 defsymbol (&Qlock_shift, "lock-shift");
5740 defsymbol (&Qescape_quoted, "escape-quoted");
5743 defsymbol (&Qdisable_composition, "disable-composition");
5745 defsymbol (&Qencode, "encode");
5746 defsymbol (&Qdecode, "decode");
5749 defsymbol (&coding_category_symbol[CODING_CATEGORY_SHIFT_JIS],
5751 defsymbol (&coding_category_symbol[CODING_CATEGORY_BIG5],
5753 defsymbol (&coding_category_symbol[CODING_CATEGORY_UCS4],
5755 defsymbol (&coding_category_symbol[CODING_CATEGORY_UTF8],
5757 defsymbol (&coding_category_symbol[CODING_CATEGORY_ISO_7],
5759 defsymbol (&coding_category_symbol[CODING_CATEGORY_ISO_8_DESIGNATE],
5761 defsymbol (&coding_category_symbol[CODING_CATEGORY_ISO_8_1],
5763 defsymbol (&coding_category_symbol[CODING_CATEGORY_ISO_8_2],
5765 defsymbol (&coding_category_symbol[CODING_CATEGORY_ISO_LOCK_SHIFT],
5768 defsymbol (&coding_category_symbol[CODING_CATEGORY_NO_CONVERSION],
5773 lstream_type_create_file_coding (void)
5775 LSTREAM_HAS_METHOD (decoding, reader);
5776 LSTREAM_HAS_METHOD (decoding, writer);
5777 LSTREAM_HAS_METHOD (decoding, rewinder);
5778 LSTREAM_HAS_METHOD (decoding, seekable_p);
5779 LSTREAM_HAS_METHOD (decoding, flusher);
5780 LSTREAM_HAS_METHOD (decoding, closer);
5781 LSTREAM_HAS_METHOD (decoding, marker);
5783 LSTREAM_HAS_METHOD (encoding, reader);
5784 LSTREAM_HAS_METHOD (encoding, writer);
5785 LSTREAM_HAS_METHOD (encoding, rewinder);
5786 LSTREAM_HAS_METHOD (encoding, seekable_p);
5787 LSTREAM_HAS_METHOD (encoding, flusher);
5788 LSTREAM_HAS_METHOD (encoding, closer);
5789 LSTREAM_HAS_METHOD (encoding, marker);
5793 vars_of_file_coding (void)
5797 fcd = xnew (struct file_coding_dump);
5798 dump_add_root_struct_ptr (&fcd, &fcd_description);
5800 /* Initialize to something reasonable ... */
5801 for (i = 0; i < CODING_CATEGORY_LAST; i++)
5803 fcd->coding_category_system[i] = Qnil;
5804 fcd->coding_category_by_priority[i] = i;
5807 Fprovide (intern ("file-coding"));
5809 DEFVAR_LISP ("keyboard-coding-system", &Vkeyboard_coding_system /*
5810 Coding system used for TTY keyboard input.
5811 Not used under a windowing system.
5813 Vkeyboard_coding_system = Qnil;
5815 DEFVAR_LISP ("terminal-coding-system", &Vterminal_coding_system /*
5816 Coding system used for TTY display output.
5817 Not used under a windowing system.
5819 Vterminal_coding_system = Qnil;
5821 DEFVAR_LISP ("coding-system-for-read", &Vcoding_system_for_read /*
5822 Overriding coding system used when reading from a file or process.
5823 You should bind this variable with `let', but do not set it globally.
5824 If this is non-nil, it specifies the coding system that will be used
5825 to decode input on read operations, such as from a file or process.
5826 It overrides `buffer-file-coding-system-for-read',
5827 `insert-file-contents-pre-hook', etc. Use those variables instead of
5828 this one for permanent changes to the environment. */ );
5829 Vcoding_system_for_read = Qnil;
5831 DEFVAR_LISP ("coding-system-for-write",
5832 &Vcoding_system_for_write /*
5833 Overriding coding system used when writing to a file or process.
5834 You should bind this variable with `let', but do not set it globally.
5835 If this is non-nil, it specifies the coding system that will be used
5836 to encode output for write operations, such as to a file or process.
5837 It overrides `buffer-file-coding-system', `write-region-pre-hook', etc.
5838 Use those variables instead of this one for permanent changes to the
5840 Vcoding_system_for_write = Qnil;
5842 DEFVAR_LISP ("file-name-coding-system", &Vfile_name_coding_system /*
5843 Coding system used to convert pathnames when accessing files.
5845 Vfile_name_coding_system = Qnil;
5847 DEFVAR_BOOL ("enable-multibyte-characters", &enable_multibyte_characters /*
5848 Non-nil means the buffer contents are regarded as multi-byte form
5849 of characters, not a binary code. This affects the display, file I/O,
5850 and behaviors of various editing commands.
5852 Setting this to nil does not do anything.
5854 enable_multibyte_characters = 1;
5858 complex_vars_of_file_coding (void)
5860 staticpro (&Vcoding_system_hash_table);
5861 Vcoding_system_hash_table =
5862 make_lisp_hash_table (50, HASH_TABLE_NON_WEAK, HASH_TABLE_EQ);
5864 the_codesys_prop_dynarr = Dynarr_new (codesys_prop);
5865 dump_add_root_struct_ptr (&the_codesys_prop_dynarr, &codesys_prop_dynarr_description);
5867 #define DEFINE_CODESYS_PROP(Prop_Type, Sym) do \
5869 struct codesys_prop csp; \
5871 csp.prop_type = (Prop_Type); \
5872 Dynarr_add (the_codesys_prop_dynarr, csp); \
5875 DEFINE_CODESYS_PROP (CODESYS_PROP_ALL_OK, Qmnemonic);
5876 DEFINE_CODESYS_PROP (CODESYS_PROP_ALL_OK, Qeol_type);
5877 DEFINE_CODESYS_PROP (CODESYS_PROP_ALL_OK, Qeol_cr);
5878 DEFINE_CODESYS_PROP (CODESYS_PROP_ALL_OK, Qeol_crlf);
5879 DEFINE_CODESYS_PROP (CODESYS_PROP_ALL_OK, Qeol_lf);
5880 DEFINE_CODESYS_PROP (CODESYS_PROP_ALL_OK, Qpost_read_conversion);
5881 DEFINE_CODESYS_PROP (CODESYS_PROP_ALL_OK, Qpre_write_conversion);
5883 DEFINE_CODESYS_PROP (CODESYS_PROP_ISO2022, Qcharset_g0);
5884 DEFINE_CODESYS_PROP (CODESYS_PROP_ISO2022, Qcharset_g1);
5885 DEFINE_CODESYS_PROP (CODESYS_PROP_ISO2022, Qcharset_g2);
5886 DEFINE_CODESYS_PROP (CODESYS_PROP_ISO2022, Qcharset_g3);
5887 DEFINE_CODESYS_PROP (CODESYS_PROP_ISO2022, Qforce_g0_on_output);
5888 DEFINE_CODESYS_PROP (CODESYS_PROP_ISO2022, Qforce_g1_on_output);
5889 DEFINE_CODESYS_PROP (CODESYS_PROP_ISO2022, Qforce_g2_on_output);
5890 DEFINE_CODESYS_PROP (CODESYS_PROP_ISO2022, Qforce_g3_on_output);
5891 DEFINE_CODESYS_PROP (CODESYS_PROP_ISO2022, Qshort);
5892 DEFINE_CODESYS_PROP (CODESYS_PROP_ISO2022, Qno_ascii_eol);
5893 DEFINE_CODESYS_PROP (CODESYS_PROP_ISO2022, Qno_ascii_cntl);
5894 DEFINE_CODESYS_PROP (CODESYS_PROP_ISO2022, Qseven);
5895 DEFINE_CODESYS_PROP (CODESYS_PROP_ISO2022, Qlock_shift);
5896 DEFINE_CODESYS_PROP (CODESYS_PROP_ISO2022, Qno_iso6429);
5897 DEFINE_CODESYS_PROP (CODESYS_PROP_ISO2022, Qescape_quoted);
5898 DEFINE_CODESYS_PROP (CODESYS_PROP_ISO2022, Qinput_charset_conversion);
5899 DEFINE_CODESYS_PROP (CODESYS_PROP_ISO2022, Qoutput_charset_conversion);
5901 DEFINE_CODESYS_PROP (CODESYS_PROP_CCL, Qencode);
5902 DEFINE_CODESYS_PROP (CODESYS_PROP_CCL, Qdecode);
5904 /* Need to create this here or we're really screwed. */
5906 (Qraw_text, Qno_conversion,
5907 build_string ("Raw text, which means it converts only line-break-codes."),
5908 list2 (Qmnemonic, build_string ("Raw")));
5911 (Qbinary, Qno_conversion,
5912 build_string ("Binary, which means it does not convert anything."),
5913 list4 (Qeol_type, Qlf,
5914 Qmnemonic, build_string ("Binary")));
5919 build_string ("Coding-system of ISO/IEC 10646 UTF-8."),
5920 list2 (Qmnemonic, build_string ("UTF8")));
5923 Fdefine_coding_system_alias (Qno_conversion, Qraw_text);
5925 Fdefine_coding_system_alias (Qfile_name, Qbinary);
5927 Fdefine_coding_system_alias (Qterminal, Qbinary);
5928 Fdefine_coding_system_alias (Qkeyboard, Qbinary);
5930 /* Need this for bootstrapping */
5931 fcd->coding_category_system[CODING_CATEGORY_NO_CONVERSION] =
5932 Fget_coding_system (Qraw_text);
5935 fcd->coding_category_system[CODING_CATEGORY_UTF8]
5936 = Fget_coding_system (Qutf8);
5939 #if defined(MULE) && !defined(UTF2000)
5943 for (i = 0; i < countof (fcd->ucs_to_mule_table); i++)
5944 fcd->ucs_to_mule_table[i] = Qnil;
5946 staticpro (&mule_to_ucs_table);
5947 mule_to_ucs_table = Fmake_char_table(Qgeneric);
5948 #endif /* defined(MULE) && !defined(UTF2000) */