|
771
|
1 /* Conversion functions for I18N encodings, but not Unicode (in separate file).
|
|
|
2 Copyright (C) 1991, 1995 Free Software Foundation, Inc.
|
|
|
3 Copyright (C) 1995 Sun Microsystems, Inc.
|
|
|
4 Copyright (C) 2000, 2001, 2002 Ben Wing.
|
|
|
5
|
|
|
6 This file is part of XEmacs.
|
|
|
7
|
|
|
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
|
|
|
11 later version.
|
|
|
12
|
|
|
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
|
|
|
16 for more details.
|
|
|
17
|
|
|
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. */
|
|
|
22
|
|
|
23 /* Synched up with: Mule 2.3. Not in FSF. */
|
|
|
24
|
|
|
25 /* For previous history, see file-coding.c.
|
|
|
26
|
|
|
27 September 10, 2001: Extracted from file-coding.c by Ben Wing.
|
|
|
28
|
|
|
29 Later in September: Finished abstraction of detection system, rewrote
|
|
|
30 all the detectors to include multiple levels of likelihood.
|
|
|
31 */
|
|
|
32
|
|
|
33 #include <config.h>
|
|
|
34 #include "lisp.h"
|
|
|
35
|
|
|
36 #include "charset.h"
|
|
|
37 #include "mule-ccl.h"
|
|
|
38 #include "file-coding.h"
|
|
|
39
|
|
|
40 Lisp_Object Qshift_jis, Qiso2022, Qbig5, Qccl;
|
|
|
41
|
|
|
42 Lisp_Object Qcharset_g0, Qcharset_g1, Qcharset_g2, Qcharset_g3;
|
|
|
43 Lisp_Object Qforce_g0_on_output, Qforce_g1_on_output;
|
|
|
44 Lisp_Object Qforce_g2_on_output, Qforce_g3_on_output;
|
|
|
45 Lisp_Object Qno_iso6429;
|
|
|
46 Lisp_Object Qinput_charset_conversion, Qoutput_charset_conversion;
|
|
|
47 Lisp_Object Qshort, Qno_ascii_eol, Qno_ascii_cntl, Qseven, Qlock_shift;
|
|
|
48
|
|
|
49 Lisp_Object Qiso_7, Qiso_8_designate, Qiso_8_1, Qiso_8_2, Qiso_lock_shift;
|
|
|
50
|
|
|
51
|
|
|
52 /************************************************************************/
|
|
|
53 /* Shift-JIS methods */
|
|
|
54 /************************************************************************/
|
|
|
55
|
|
|
56 /* Shift-JIS; Hankaku (half-width) KANA is also supported. */
|
|
|
57 DEFINE_CODING_SYSTEM_TYPE (shift_jis);
|
|
|
58
|
|
|
59 /* Shift-JIS is a coding system encoding three character sets: ASCII, right
|
|
|
60 half of JISX0201-Kana, and JISX0208. An ASCII character is encoded
|
|
|
61 as is. A character of JISX0201-Kana (DIMENSION1_CHARS94 character set) is
|
|
|
62 encoded by "position-code + 0x80". A character of JISX0208
|
|
|
63 (DIMENSION2_CHARS94 character set) is encoded in 2-byte but two
|
|
|
64 position-codes are divided and shifted so that it fit in the range
|
|
|
65 below.
|
|
|
66
|
|
|
67 --- CODE RANGE of Shift-JIS ---
|
|
|
68 (character set) (range)
|
|
|
69 ASCII 0x00 .. 0x7F
|
|
|
70 JISX0201-Kana 0xA0 .. 0xDF
|
|
|
71 JISX0208 (1st byte) 0x80 .. 0x9F and 0xE0 .. 0xEF
|
|
|
72 (2nd byte) 0x40 .. 0x7E and 0x80 .. 0xFC
|
|
|
73 -------------------------------
|
|
|
74
|
|
|
75 */
|
|
|
76
|
|
|
77 /* Is this the first byte of a Shift-JIS two-byte char? */
|
|
|
78
|
|
826
|
79 inline static int
|
|
|
80 byte_shift_jis_two_byte_1_p (int c)
|
|
|
81 {
|
|
|
82 return (c >= 0x81 && c <= 0x9F) || (c >= 0xE0 && c <= 0xEF);
|
|
|
83 }
|
|
771
|
84
|
|
|
85 /* Is this the second byte of a Shift-JIS two-byte char? */
|
|
|
86
|
|
826
|
87 inline static int
|
|
|
88 byte_shift_jis_two_byte_2_p (int c)
|
|
|
89 {
|
|
|
90 return (c >= 0x40 && c <= 0x7E) || (c >= 0x80 && c <= 0xFC);
|
|
|
91 }
|
|
|
92
|
|
|
93 inline static int
|
|
|
94 byte_shift_jis_katakana_p (int c)
|
|
|
95 {
|
|
|
96 return c >= 0xA1 && c <= 0xDF;
|
|
|
97 }
|
|
771
|
98
|
|
|
99 /* Convert Shift-JIS data to internal format. */
|
|
|
100
|
|
|
101 static Bytecount
|
|
|
102 shift_jis_convert (struct coding_stream *str, const UExtbyte *src,
|
|
|
103 unsigned_char_dynarr *dst, Bytecount n)
|
|
|
104 {
|
|
|
105 unsigned int ch = str->ch;
|
|
|
106 Bytecount orign = n;
|
|
|
107
|
|
|
108 if (str->direction == CODING_DECODE)
|
|
|
109 {
|
|
|
110 while (n--)
|
|
|
111 {
|
|
|
112 UExtbyte c = *src++;
|
|
|
113
|
|
|
114 if (ch)
|
|
|
115 {
|
|
|
116 /* Previous character was first byte of Shift-JIS Kanji char. */
|
|
826
|
117 if (byte_shift_jis_two_byte_2_p (c))
|
|
771
|
118 {
|
|
867
|
119 Ibyte e1, e2;
|
|
771
|
120
|
|
|
121 Dynarr_add (dst, LEADING_BYTE_JAPANESE_JISX0208);
|
|
|
122 DECODE_SHIFT_JIS (ch, c, e1, e2);
|
|
|
123 Dynarr_add (dst, e1);
|
|
|
124 Dynarr_add (dst, e2);
|
|
|
125 }
|
|
|
126 else
|
|
|
127 {
|
|
|
128 DECODE_ADD_BINARY_CHAR (ch, dst);
|
|
|
129 DECODE_ADD_BINARY_CHAR (c, dst);
|
|
|
130 }
|
|
|
131 ch = 0;
|
|
|
132 }
|
|
|
133 else
|
|
|
134 {
|
|
826
|
135 if (byte_shift_jis_two_byte_1_p (c))
|
|
771
|
136 ch = c;
|
|
826
|
137 else if (byte_shift_jis_katakana_p (c))
|
|
771
|
138 {
|
|
|
139 Dynarr_add (dst, LEADING_BYTE_KATAKANA_JISX0201);
|
|
|
140 Dynarr_add (dst, c);
|
|
|
141 }
|
|
|
142 else
|
|
|
143 DECODE_ADD_BINARY_CHAR (c, dst);
|
|
|
144 }
|
|
|
145 }
|
|
|
146
|
|
|
147 if (str->eof)
|
|
|
148 DECODE_OUTPUT_PARTIAL_CHAR (ch, dst);
|
|
|
149 }
|
|
|
150 else
|
|
|
151 {
|
|
|
152 while (n--)
|
|
|
153 {
|
|
867
|
154 Ibyte c = *src++;
|
|
826
|
155 if (byte_ascii_p (c))
|
|
771
|
156 {
|
|
|
157 Dynarr_add (dst, c);
|
|
|
158 ch = 0;
|
|
|
159 }
|
|
867
|
160 else if (ibyte_leading_byte_p (c))
|
|
771
|
161 ch = (c == LEADING_BYTE_KATAKANA_JISX0201 ||
|
|
|
162 c == LEADING_BYTE_JAPANESE_JISX0208_1978 ||
|
|
|
163 c == LEADING_BYTE_JAPANESE_JISX0208) ? c : 0;
|
|
|
164 else if (ch)
|
|
|
165 {
|
|
|
166 if (ch == LEADING_BYTE_KATAKANA_JISX0201)
|
|
|
167 {
|
|
|
168 Dynarr_add (dst, c);
|
|
|
169 ch = 0;
|
|
|
170 }
|
|
|
171 else if (ch == LEADING_BYTE_JAPANESE_JISX0208_1978 ||
|
|
|
172 ch == LEADING_BYTE_JAPANESE_JISX0208)
|
|
|
173 ch = c;
|
|
|
174 else
|
|
|
175 {
|
|
|
176 UExtbyte j1, j2;
|
|
|
177 ENCODE_SHIFT_JIS (ch, c, j1, j2);
|
|
|
178 Dynarr_add (dst, j1);
|
|
|
179 Dynarr_add (dst, j2);
|
|
|
180 ch = 0;
|
|
|
181 }
|
|
|
182 }
|
|
|
183 }
|
|
|
184 }
|
|
|
185
|
|
|
186 str->ch = ch;
|
|
|
187
|
|
|
188 return orign;
|
|
|
189 }
|
|
|
190
|
|
|
191 DEFUN ("decode-shift-jis-char", Fdecode_shift_jis_char, 1, 1, 0, /*
|
|
|
192 Decode a JISX0208 character of Shift-JIS coding-system.
|
|
|
193 CODE is the character code in Shift-JIS as a cons of type bytes.
|
|
|
194 Return the corresponding character.
|
|
|
195 */
|
|
|
196 (code))
|
|
|
197 {
|
|
|
198 int c1, c2, s1, s2;
|
|
|
199
|
|
|
200 CHECK_CONS (code);
|
|
|
201 CHECK_INT (XCAR (code));
|
|
|
202 CHECK_INT (XCDR (code));
|
|
|
203 s1 = XINT (XCAR (code));
|
|
|
204 s2 = XINT (XCDR (code));
|
|
826
|
205 if (byte_shift_jis_two_byte_1_p (s1) &&
|
|
|
206 byte_shift_jis_two_byte_2_p (s2))
|
|
771
|
207 {
|
|
|
208 DECODE_SHIFT_JIS (s1, s2, c1, c2);
|
|
867
|
209 return make_char (make_ichar (Vcharset_japanese_jisx0208,
|
|
831
|
210 c1 & 0x7F, c2 & 0x7F));
|
|
771
|
211 }
|
|
|
212 else
|
|
|
213 return Qnil;
|
|
|
214 }
|
|
|
215
|
|
|
216 DEFUN ("encode-shift-jis-char", Fencode_shift_jis_char, 1, 1, 0, /*
|
|
|
217 Encode a JISX0208 character CHARACTER to SHIFT-JIS coding-system.
|
|
|
218 Return the corresponding character code in SHIFT-JIS as a cons of two bytes.
|
|
|
219 */
|
|
|
220 (character))
|
|
|
221 {
|
|
|
222 Lisp_Object charset;
|
|
|
223 int c1, c2, s1, s2;
|
|
|
224
|
|
|
225 CHECK_CHAR_COERCE_INT (character);
|
|
867
|
226 BREAKUP_ICHAR (XCHAR (character), charset, c1, c2);
|
|
771
|
227 if (EQ (charset, Vcharset_japanese_jisx0208))
|
|
|
228 {
|
|
|
229 ENCODE_SHIFT_JIS (c1 | 0x80, c2 | 0x80, s1, s2);
|
|
|
230 return Fcons (make_int (s1), make_int (s2));
|
|
|
231 }
|
|
|
232 else
|
|
|
233 return Qnil;
|
|
|
234 }
|
|
|
235
|
|
|
236
|
|
|
237 /************************************************************************/
|
|
|
238 /* Shift-JIS detector */
|
|
|
239 /************************************************************************/
|
|
|
240
|
|
|
241 DEFINE_DETECTOR (shift_jis);
|
|
|
242 DEFINE_DETECTOR_CATEGORY (shift_jis, shift_jis);
|
|
|
243
|
|
|
244 struct shift_jis_detector
|
|
|
245 {
|
|
|
246 int seen_jisx0208_char_in_c1;
|
|
|
247 int seen_jisx0208_char_in_upper;
|
|
|
248 int seen_jisx0201_char;
|
|
|
249 unsigned int seen_iso2022_esc:1;
|
|
|
250 unsigned int seen_bad_first_byte:1;
|
|
|
251 unsigned int seen_bad_second_byte:1;
|
|
|
252 /* temporary */
|
|
|
253 unsigned int in_second_byte:1;
|
|
|
254 unsigned int first_byte_was_c1:1;
|
|
|
255 };
|
|
|
256
|
|
|
257 static void
|
|
|
258 shift_jis_detect (struct detection_state *st, const UExtbyte *src,
|
|
|
259 Bytecount n)
|
|
|
260 {
|
|
|
261 struct shift_jis_detector *data = DETECTION_STATE_DATA (st, shift_jis);
|
|
|
262
|
|
|
263 while (n--)
|
|
|
264 {
|
|
|
265 UExtbyte c = *src++;
|
|
|
266 if (!data->in_second_byte)
|
|
|
267 {
|
|
|
268 if (c >= 0x80 && c <= 0x9F)
|
|
|
269 data->first_byte_was_c1 = 1;
|
|
|
270 if (c >= 0xA0 && c <= 0xDF)
|
|
|
271 data->seen_jisx0201_char++;
|
|
|
272 else if ((c >= 0x80 && c <= 0x9F) || (c >= 0xE0 && c <= 0xEF))
|
|
|
273 data->in_second_byte = 1;
|
|
|
274 else if (c == ISO_CODE_ESC || c == ISO_CODE_SI || c == ISO_CODE_SO)
|
|
|
275 data->seen_iso2022_esc = 1;
|
|
|
276 else if (c >= 0x80)
|
|
|
277 data->seen_bad_first_byte = 1;
|
|
|
278 }
|
|
|
279 else
|
|
|
280 {
|
|
|
281 if ((c >= 0x40 && c <= 0x7E) || (c >= 0x80 && c <= 0xFC))
|
|
|
282 {
|
|
|
283 if (data->first_byte_was_c1 || (c >= 0x80 && c <= 0x9F))
|
|
|
284 data->seen_jisx0208_char_in_c1++;
|
|
|
285 else
|
|
|
286 data->seen_jisx0208_char_in_upper++;
|
|
|
287 }
|
|
|
288 else
|
|
|
289 data->seen_bad_second_byte = 1;
|
|
|
290 data->in_second_byte = 0;
|
|
|
291 data->first_byte_was_c1 = 0;
|
|
|
292 }
|
|
|
293 }
|
|
|
294
|
|
|
295 if (data->seen_bad_second_byte)
|
|
|
296 DET_RESULT (st, shift_jis) = DET_NEARLY_IMPOSSIBLE;
|
|
|
297 else if (data->seen_bad_first_byte)
|
|
|
298 DET_RESULT (st, shift_jis) = DET_QUITE_IMPROBABLE;
|
|
|
299 else if (data->seen_iso2022_esc)
|
|
|
300 DET_RESULT (st, shift_jis) = DET_SOMEWHAT_UNLIKELY;
|
|
|
301 else if (data->seen_jisx0208_char_in_c1 >= 20 ||
|
|
|
302 (data->seen_jisx0208_char_in_c1 >= 10 &&
|
|
|
303 data->seen_jisx0208_char_in_upper >= 10))
|
|
|
304 DET_RESULT (st, shift_jis) = DET_QUITE_PROBABLE;
|
|
|
305 else if (data->seen_jisx0208_char_in_c1 > 3 ||
|
|
|
306 data->seen_jisx0208_char_in_upper >= 10 ||
|
|
|
307 /* Since the range is limited compared to what is often seen
|
|
|
308 is typical Latin-X charsets, the fact that we've seen a
|
|
|
309 bunch of them and none that are invalid is reasonably
|
|
|
310 strong statistical evidence of this encoding, or at least
|
|
|
311 not of the common Latin-X ones. */
|
|
|
312 data->seen_jisx0201_char >= 100)
|
|
|
313 DET_RESULT (st, shift_jis) = DET_SOMEWHAT_LIKELY;
|
|
|
314 else if (data->seen_jisx0208_char_in_c1 > 0 ||
|
|
|
315 data->seen_jisx0208_char_in_upper > 0 ||
|
|
|
316 data->seen_jisx0201_char > 0)
|
|
|
317 DET_RESULT (st, shift_jis) = DET_SLIGHTLY_LIKELY;
|
|
|
318 else
|
|
|
319 DET_RESULT (st, shift_jis) = DET_AS_LIKELY_AS_UNLIKELY;
|
|
|
320 }
|
|
|
321
|
|
|
322
|
|
|
323 /************************************************************************/
|
|
|
324 /* Big5 methods */
|
|
|
325 /************************************************************************/
|
|
|
326
|
|
|
327 /* BIG5 (used for Taiwanese). */
|
|
|
328 DEFINE_CODING_SYSTEM_TYPE (big5);
|
|
|
329
|
|
|
330 /* BIG5 is a coding system encoding two character sets: ASCII and
|
|
|
331 Big5. An ASCII character is encoded as is. Big5 is a two-byte
|
|
|
332 character set and is encoded in two-byte.
|
|
|
333
|
|
|
334 --- CODE RANGE of BIG5 ---
|
|
|
335 (character set) (range)
|
|
|
336 ASCII 0x00 .. 0x7F
|
|
|
337 Big5 (1st byte) 0xA1 .. 0xFE
|
|
|
338 (2nd byte) 0x40 .. 0x7E and 0xA1 .. 0xFE
|
|
|
339 --------------------------
|
|
|
340
|
|
|
341 Since the number of characters in Big5 is larger than maximum
|
|
|
342 characters in Emacs' charset (96x96), it can't be handled as one
|
|
|
343 charset. So, in XEmacs, Big5 is divided into two: `charset-big5-1'
|
|
|
344 and `charset-big5-2'. Both <type>s are DIMENSION2_CHARS94. The former
|
|
|
345 contains frequently used characters and the latter contains less
|
|
|
346 frequently used characters. */
|
|
|
347
|
|
826
|
348 inline static int
|
|
|
349 byte_big5_two_byte_1_p (int c)
|
|
|
350 {
|
|
|
351 return c >= 0xA1 && c <= 0xFE;
|
|
|
352 }
|
|
771
|
353
|
|
|
354 /* Is this the second byte of a Shift-JIS two-byte char? */
|
|
|
355
|
|
826
|
356 inline static int
|
|
|
357 byte_big5_two_byte_2_p (int c)
|
|
|
358 {
|
|
|
359 return (c >= 0x40 && c <= 0x7E) || (c >= 0xA1 && c <= 0xFE);
|
|
|
360 }
|
|
771
|
361
|
|
|
362 /* Number of Big5 characters which have the same code in 1st byte. */
|
|
|
363
|
|
|
364 #define BIG5_SAME_ROW (0xFF - 0xA1 + 0x7F - 0x40)
|
|
|
365
|
|
|
366 /* Code conversion macros. These are macros because they are used in
|
|
|
367 inner loops during code conversion.
|
|
|
368
|
|
|
369 Note that temporary variables in macros introduce the classic
|
|
|
370 dynamic-scoping problems with variable names. We use capital-
|
|
|
371 lettered variables in the assumption that XEmacs does not use
|
|
|
372 capital letters in variables except in a very formalized way
|
|
|
373 (e.g. Qstring). */
|
|
|
374
|
|
|
375 /* Convert Big5 code (b1, b2) into its internal string representation
|
|
|
376 (lb, c1, c2). */
|
|
|
377
|
|
|
378 /* There is a much simpler way to split the Big5 charset into two.
|
|
|
379 For the moment I'm going to leave the algorithm as-is because it
|
|
|
380 claims to separate out the most-used characters into a single
|
|
|
381 charset, which perhaps will lead to optimizations in various
|
|
|
382 places.
|
|
|
383
|
|
|
384 The way the algorithm works is something like this:
|
|
|
385
|
|
|
386 Big5 can be viewed as a 94x157 charset, where the row is
|
|
|
387 encoded into the bytes 0xA1 .. 0xFE and the column is encoded
|
|
|
388 into the bytes 0x40 .. 0x7E and 0xA1 .. 0xFE. As for frequency,
|
|
|
389 the split between low and high column numbers is apparently
|
|
|
390 meaningless; ascending rows produce less and less frequent chars.
|
|
|
391 Therefore, we assign the lower half of rows (0xA1 .. 0xC8) to
|
|
|
392 the first charset, and the upper half (0xC9 .. 0xFE) to the
|
|
|
393 second. To do the conversion, we convert the character into
|
|
|
394 a single number where 0 .. 156 is the first row, 157 .. 313
|
|
|
395 is the second, etc. That way, the characters are ordered by
|
|
|
396 decreasing frequency. Then we just chop the space in two
|
|
|
397 and coerce the result into a 94x94 space.
|
|
|
398 */
|
|
|
399
|
|
|
400 #define DECODE_BIG5(b1, b2, lb, c1, c2) do \
|
|
|
401 { \
|
|
|
402 int B1 = b1, B2 = b2; \
|
|
|
403 int I \
|
|
|
404 = (B1 - 0xA1) * BIG5_SAME_ROW + B2 - (B2 < 0x7F ? 0x40 : 0x62); \
|
|
|
405 \
|
|
|
406 if (B1 < 0xC9) \
|
|
|
407 { \
|
|
|
408 lb = LEADING_BYTE_CHINESE_BIG5_1; \
|
|
|
409 } \
|
|
|
410 else \
|
|
|
411 { \
|
|
|
412 lb = LEADING_BYTE_CHINESE_BIG5_2; \
|
|
|
413 I -= (BIG5_SAME_ROW) * (0xC9 - 0xA1); \
|
|
|
414 } \
|
|
|
415 c1 = I / (0xFF - 0xA1) + 0xA1; \
|
|
|
416 c2 = I % (0xFF - 0xA1) + 0xA1; \
|
|
|
417 } while (0)
|
|
|
418
|
|
|
419 /* Convert the internal string representation of a Big5 character
|
|
|
420 (lb, c1, c2) into Big5 code (b1, b2). */
|
|
|
421
|
|
|
422 #define ENCODE_BIG5(lb, c1, c2, b1, b2) do \
|
|
|
423 { \
|
|
|
424 int I = ((c1) - 0xA1) * (0xFF - 0xA1) + ((c2) - 0xA1); \
|
|
|
425 \
|
|
|
426 if (lb == LEADING_BYTE_CHINESE_BIG5_2) \
|
|
|
427 { \
|
|
|
428 I += BIG5_SAME_ROW * (0xC9 - 0xA1); \
|
|
|
429 } \
|
|
|
430 b1 = I / BIG5_SAME_ROW + 0xA1; \
|
|
|
431 b2 = I % BIG5_SAME_ROW; \
|
|
|
432 b2 += b2 < 0x3F ? 0x40 : 0x62; \
|
|
|
433 } while (0)
|
|
|
434
|
|
|
435 /* Convert Big5 data to internal format. */
|
|
|
436
|
|
|
437 static Bytecount
|
|
|
438 big5_convert (struct coding_stream *str, const UExtbyte *src,
|
|
|
439 unsigned_char_dynarr *dst, Bytecount n)
|
|
|
440 {
|
|
|
441 unsigned int ch = str->ch;
|
|
|
442 Bytecount orign = n;
|
|
|
443
|
|
|
444 if (str->direction == CODING_DECODE)
|
|
|
445 {
|
|
|
446 while (n--)
|
|
|
447 {
|
|
|
448 UExtbyte c = *src++;
|
|
|
449 if (ch)
|
|
|
450 {
|
|
|
451 /* Previous character was first byte of Big5 char. */
|
|
826
|
452 if (byte_big5_two_byte_2_p (c))
|
|
771
|
453 {
|
|
867
|
454 Ibyte b1, b2, b3;
|
|
771
|
455 DECODE_BIG5 (ch, c, b1, b2, b3);
|
|
|
456 Dynarr_add (dst, b1);
|
|
|
457 Dynarr_add (dst, b2);
|
|
|
458 Dynarr_add (dst, b3);
|
|
|
459 }
|
|
|
460 else
|
|
|
461 {
|
|
|
462 DECODE_ADD_BINARY_CHAR (ch, dst);
|
|
|
463 DECODE_ADD_BINARY_CHAR (c, dst);
|
|
|
464 }
|
|
|
465 ch = 0;
|
|
|
466 }
|
|
|
467 else
|
|
|
468 {
|
|
826
|
469 if (byte_big5_two_byte_1_p (c))
|
|
771
|
470 ch = c;
|
|
|
471 else
|
|
|
472 DECODE_ADD_BINARY_CHAR (c, dst);
|
|
|
473 }
|
|
|
474 }
|
|
|
475
|
|
|
476 if (str->eof)
|
|
|
477 DECODE_OUTPUT_PARTIAL_CHAR (ch, dst);
|
|
|
478 }
|
|
|
479 else
|
|
|
480 {
|
|
|
481 while (n--)
|
|
|
482 {
|
|
867
|
483 Ibyte c = *src++;
|
|
826
|
484 if (byte_ascii_p (c))
|
|
771
|
485 {
|
|
|
486 /* ASCII. */
|
|
|
487 Dynarr_add (dst, c);
|
|
|
488 }
|
|
867
|
489 else if (ibyte_leading_byte_p (c))
|
|
771
|
490 {
|
|
|
491 if (c == LEADING_BYTE_CHINESE_BIG5_1 ||
|
|
|
492 c == LEADING_BYTE_CHINESE_BIG5_2)
|
|
|
493 {
|
|
|
494 /* A recognized leading byte. */
|
|
|
495 ch = c;
|
|
|
496 continue; /* not done with this character. */
|
|
|
497 }
|
|
|
498 /* otherwise just ignore this character. */
|
|
|
499 }
|
|
|
500 else if (ch == LEADING_BYTE_CHINESE_BIG5_1 ||
|
|
|
501 ch == LEADING_BYTE_CHINESE_BIG5_2)
|
|
|
502 {
|
|
|
503 /* Previous char was a recognized leading byte. */
|
|
|
504 ch = (ch << 8) | c;
|
|
|
505 continue; /* not done with this character. */
|
|
|
506 }
|
|
|
507 else if (ch)
|
|
|
508 {
|
|
|
509 /* Encountering second byte of a Big5 character. */
|
|
|
510 UExtbyte b1, b2;
|
|
|
511
|
|
|
512 ENCODE_BIG5 (ch >> 8, ch & 0xFF, c, b1, b2);
|
|
|
513 Dynarr_add (dst, b1);
|
|
|
514 Dynarr_add (dst, b2);
|
|
|
515 }
|
|
|
516
|
|
|
517 ch = 0;
|
|
|
518 }
|
|
|
519 }
|
|
|
520
|
|
|
521 str->ch = ch;
|
|
|
522 return orign;
|
|
|
523 }
|
|
|
524
|
|
867
|
525 Ichar
|
|
771
|
526 decode_big5_char (int b1, int b2)
|
|
|
527 {
|
|
826
|
528 if (byte_big5_two_byte_1_p (b1) &&
|
|
|
529 byte_big5_two_byte_2_p (b2))
|
|
771
|
530 {
|
|
|
531 int leading_byte;
|
|
|
532 Lisp_Object charset;
|
|
|
533 int c1, c2;
|
|
|
534
|
|
|
535 DECODE_BIG5 (b1, b2, leading_byte, c1, c2);
|
|
826
|
536 charset = charset_by_leading_byte (leading_byte);
|
|
867
|
537 return make_ichar (charset, c1 & 0x7F, c2 & 0x7F);
|
|
771
|
538 }
|
|
|
539 else
|
|
|
540 return -1;
|
|
|
541 }
|
|
|
542
|
|
|
543 DEFUN ("decode-big5-char", Fdecode_big5_char, 1, 1, 0, /*
|
|
|
544 Convert Big Five character codes in CODE into a character.
|
|
|
545 CODE is a cons of two integers specifying the codepoints in Big Five.
|
|
|
546 Return the corresponding character, or nil if the codepoints are out of range.
|
|
|
547
|
|
|
548 The term `decode' is used because the codepoints can be viewed as the
|
|
|
549 representation of the character in the external Big Five encoding, and thus
|
|
|
550 converting them to a character is analogous to any other operation that
|
|
|
551 decodes an external representation.
|
|
|
552 */
|
|
|
553 (code))
|
|
|
554 {
|
|
867
|
555 Ichar ch;
|
|
771
|
556
|
|
|
557 CHECK_CONS (code);
|
|
|
558 CHECK_INT (XCAR (code));
|
|
|
559 CHECK_INT (XCDR (code));
|
|
|
560 ch = decode_big5_char (XINT (XCAR (code)), XINT (XCDR (code)));
|
|
|
561 if (ch == -1)
|
|
|
562 return Qnil;
|
|
|
563 else
|
|
|
564 return make_char (ch);
|
|
|
565 }
|
|
|
566
|
|
|
567 DEFUN ("encode-big5-char", Fencode_big5_char, 1, 1, 0, /*
|
|
|
568 Convert the specified Big Five character into its codepoints.
|
|
|
569 The codepoints are returned as a cons of two integers, specifying the
|
|
|
570 Big Five codepoints. See `decode-big5-char' for the reason why the
|
|
|
571 term `encode' is used for this operation.
|
|
|
572 */
|
|
|
573 (character))
|
|
|
574 {
|
|
|
575 Lisp_Object charset;
|
|
|
576 int c1, c2, b1, b2;
|
|
|
577
|
|
|
578 CHECK_CHAR_COERCE_INT (character);
|
|
867
|
579 BREAKUP_ICHAR (XCHAR (character), charset, c1, c2);
|
|
771
|
580 if (EQ (charset, Vcharset_chinese_big5_1) ||
|
|
|
581 EQ (charset, Vcharset_chinese_big5_2))
|
|
|
582 {
|
|
|
583 ENCODE_BIG5 (XCHARSET_LEADING_BYTE (charset), c1 | 0x80, c2 | 0x80,
|
|
|
584 b1, b2);
|
|
|
585 return Fcons (make_int (b1), make_int (b2));
|
|
|
586 }
|
|
|
587 else
|
|
|
588 return Qnil;
|
|
|
589 }
|
|
|
590
|
|
|
591
|
|
|
592 /************************************************************************/
|
|
|
593 /* Big5 detector */
|
|
|
594 /************************************************************************/
|
|
|
595
|
|
|
596 DEFINE_DETECTOR (big5);
|
|
|
597 DEFINE_DETECTOR_CATEGORY (big5, big5);
|
|
|
598
|
|
|
599 struct big5_detector
|
|
|
600 {
|
|
|
601 int seen_big5_char;
|
|
|
602 unsigned int seen_iso2022_esc:1;
|
|
|
603 unsigned int seen_bad_first_byte:1;
|
|
|
604 unsigned int seen_bad_second_byte:1;
|
|
|
605
|
|
|
606 /* temporary */
|
|
|
607 unsigned int in_second_byte:1;
|
|
|
608 };
|
|
|
609
|
|
|
610 static void
|
|
|
611 big5_detect (struct detection_state *st, const UExtbyte *src,
|
|
|
612 Bytecount n)
|
|
|
613 {
|
|
|
614 struct big5_detector *data = DETECTION_STATE_DATA (st, big5);
|
|
|
615
|
|
|
616 while (n--)
|
|
|
617 {
|
|
|
618 UExtbyte c = *src++;
|
|
|
619 if (!data->in_second_byte)
|
|
|
620 {
|
|
|
621 if (c >= 0xA1 && c <= 0xFE)
|
|
|
622 data->in_second_byte = 1;
|
|
|
623 else if (c == ISO_CODE_ESC || c == ISO_CODE_SI || c == ISO_CODE_SO)
|
|
|
624 data->seen_iso2022_esc = 1;
|
|
|
625 else if (c >= 0x80)
|
|
|
626 data->seen_bad_first_byte = 1;
|
|
|
627 }
|
|
|
628 else
|
|
|
629 {
|
|
|
630 data->in_second_byte = 0;
|
|
|
631 if ((c >= 0x40 && c <= 0x7E) || (c >= 0xA1 && c <= 0xFE))
|
|
|
632 data->seen_big5_char++;
|
|
|
633 else
|
|
|
634 data->seen_bad_second_byte = 1;
|
|
|
635 }
|
|
|
636 }
|
|
|
637
|
|
|
638 if (data->seen_bad_second_byte)
|
|
|
639 DET_RESULT (st, big5) = DET_NEARLY_IMPOSSIBLE;
|
|
|
640 else if (data->seen_bad_first_byte)
|
|
|
641 DET_RESULT (st, big5) = DET_QUITE_IMPROBABLE;
|
|
|
642 else if (data->seen_iso2022_esc)
|
|
|
643 DET_RESULT (st, big5) = DET_SOMEWHAT_UNLIKELY;
|
|
|
644 else if (data->seen_big5_char >= 4)
|
|
|
645 DET_RESULT (st, big5) = DET_SOMEWHAT_LIKELY;
|
|
|
646 else
|
|
|
647 DET_RESULT (st, big5) = DET_AS_LIKELY_AS_UNLIKELY;
|
|
|
648 }
|
|
|
649
|
|
|
650
|
|
|
651 /************************************************************************/
|
|
|
652 /* ISO2022 methods */
|
|
|
653 /************************************************************************/
|
|
|
654
|
|
|
655 /* Any ISO-2022-compliant coding system. Includes JIS, EUC, CTEXT
|
|
|
656 (Compound Text, the encoding of selections in X Windows). See below for
|
|
|
657 a complete description of ISO-2022. */
|
|
|
658 DEFINE_CODING_SYSTEM_TYPE (iso2022);
|
|
|
659
|
|
|
660 /* Flags indicating what we've seen so far when parsing an
|
|
|
661 ISO2022 escape sequence. */
|
|
|
662 enum iso_esc_flag
|
|
|
663 {
|
|
|
664 /* Partial sequences */
|
|
|
665 ISO_ESC_NOTHING, /* Nothing has been seen. */
|
|
|
666 ISO_ESC, /* We've seen ESC. */
|
|
|
667 ISO_ESC_2_4, /* We've seen ESC $. This indicates
|
|
|
668 that we're designating a multi-byte, rather
|
|
|
669 than a single-byte, character set. */
|
|
|
670 ISO_ESC_2_8, /* We've seen ESC 0x28, i.e. ESC (.
|
|
|
671 This means designate a 94-character
|
|
|
672 character set into G0. */
|
|
|
673 ISO_ESC_2_9, /* We've seen ESC 0x29 -- designate a
|
|
|
674 94-character character set into G1. */
|
|
|
675 ISO_ESC_2_10, /* We've seen ESC 0x2A. */
|
|
|
676 ISO_ESC_2_11, /* We've seen ESC 0x2B. */
|
|
|
677 ISO_ESC_2_12, /* We've seen ESC 0x2C -- designate a
|
|
|
678 96-character character set into G0.
|
|
|
679 (This is not ISO2022-standard.
|
|
|
680 The following 96-character
|
|
|
681 control sequences are standard,
|
|
|
682 though.) */
|
|
|
683 ISO_ESC_2_13, /* We've seen ESC 0x2D -- designate a
|
|
|
684 96-character character set into G1.
|
|
|
685 */
|
|
|
686 ISO_ESC_2_14, /* We've seen ESC 0x2E. */
|
|
|
687 ISO_ESC_2_15, /* We've seen ESC 0x2F. */
|
|
|
688 ISO_ESC_2_4_8, /* We've seen ESC $ 0x28 -- designate
|
|
|
689 a 94^N character set into G0. */
|
|
|
690 ISO_ESC_2_4_9, /* We've seen ESC $ 0x29. */
|
|
|
691 ISO_ESC_2_4_10, /* We've seen ESC $ 0x2A. */
|
|
|
692 ISO_ESC_2_4_11, /* We've seen ESC $ 0x2B. */
|
|
|
693 ISO_ESC_2_4_12, /* We've seen ESC $ 0x2C. */
|
|
|
694 ISO_ESC_2_4_13, /* We've seen ESC $ 0x2D. */
|
|
|
695 ISO_ESC_2_4_14, /* We've seen ESC $ 0x2E. */
|
|
|
696 ISO_ESC_2_4_15, /* We've seen ESC $ 0x2F. */
|
|
|
697 ISO_ESC_5_11, /* We've seen ESC [ or 0x9B. This
|
|
|
698 starts a directionality-control
|
|
|
699 sequence. The next character
|
|
|
700 must be 0, 1, 2, or ]. */
|
|
|
701 ISO_ESC_5_11_0, /* We've seen 0x9B 0. The next character must be ]. */
|
|
|
702 ISO_ESC_5_11_1, /* We've seen 0x9B 1. The next character must be ]. */
|
|
|
703 ISO_ESC_5_11_2, /* We've seen 0x9B 2. The next character must be ]. */
|
|
|
704
|
|
|
705 /* Full sequences. */
|
|
|
706 ISO_ESC_START_COMPOSITE, /* Private usage for START COMPOSING */
|
|
|
707 ISO_ESC_END_COMPOSITE, /* Private usage for END COMPOSING */
|
|
|
708 ISO_ESC_SINGLE_SHIFT, /* We've seen a complete single-shift sequence. */
|
|
|
709 ISO_ESC_LOCKING_SHIFT,/* We've seen a complete locking-shift sequence. */
|
|
|
710 ISO_ESC_DESIGNATE, /* We've seen a complete designation sequence. */
|
|
|
711 ISO_ESC_DIRECTIONALITY,/* We've seen a complete ISO6429 directionality
|
|
|
712 sequence. */
|
|
|
713 ISO_ESC_LITERAL /* We've seen a literal character ala
|
|
|
714 escape-quoting. */
|
|
|
715 };
|
|
|
716
|
|
|
717 enum iso_error
|
|
|
718 {
|
|
|
719 ISO_ERROR_BAD_FINAL,
|
|
|
720 ISO_ERROR_UNKWOWN_ESC_SEQUENCE,
|
|
|
721 ISO_ERROR_INVALID_CODE_POINT_CHARACTER,
|
|
|
722 };
|
|
|
723
|
|
|
724
|
|
|
725 /* Flags indicating current state while converting code. */
|
|
|
726
|
|
|
727 /************ Used during encoding and decoding: ************/
|
|
|
728 /* If set, the current directionality is right-to-left. Otherwise, it's
|
|
|
729 left-to-right. */
|
|
|
730 #define ISO_STATE_R2L (1 << 0)
|
|
|
731
|
|
|
732 /************ Used during encoding: ************/
|
|
|
733 /* If set, we just saw a CR. */
|
|
|
734 #define ISO_STATE_CR (1 << 1)
|
|
|
735
|
|
|
736 /************ Used during decoding: ************/
|
|
|
737 /* If set, we're currently parsing an escape sequence and the upper 16 bits
|
|
|
738 should be looked at to indicate what partial escape sequence we've seen
|
|
|
739 so far. Otherwise, we're running through actual text. */
|
|
|
740 #define ISO_STATE_ESCAPE (1 << 2)
|
|
|
741 /* If set, G2 is invoked into GL, but only for the next character. */
|
|
|
742 #define ISO_STATE_SS2 (1 << 3)
|
|
|
743 /* If set, G3 is invoked into GL, but only for the next character. If both
|
|
|
744 ISO_STATE_SS2 and ISO_STATE_SS3 are set, ISO_STATE_SS2 overrides; but
|
|
|
745 this probably indicates an error in the text encoding. */
|
|
|
746 #define ISO_STATE_SS3 (1 << 4)
|
|
|
747 /* If set, we're currently processing a composite character (i.e. a
|
|
|
748 character constructed by overstriking two or more characters). */
|
|
|
749 #define ISO_STATE_COMPOSITE (1 << 5)
|
|
|
750
|
|
|
751 /* ISO_STATE_LOCK is the mask of flags that remain on until explicitly
|
|
|
752 turned off when in the ISO2022 encoder/decoder. Other flags are turned
|
|
|
753 off at the end of processing each character or escape sequence. */
|
|
|
754 # define ISO_STATE_LOCK \
|
|
|
755 (ISO_STATE_COMPOSITE | ISO_STATE_R2L)
|
|
|
756
|
|
|
757 typedef struct charset_conversion_spec
|
|
|
758 {
|
|
|
759 Lisp_Object from_charset;
|
|
|
760 Lisp_Object to_charset;
|
|
|
761 } charset_conversion_spec;
|
|
|
762
|
|
|
763 typedef struct
|
|
|
764 {
|
|
|
765 Dynarr_declare (charset_conversion_spec);
|
|
|
766 } charset_conversion_spec_dynarr;
|
|
|
767
|
|
|
768 struct iso2022_coding_system
|
|
|
769 {
|
|
|
770 /* What are the charsets to be initially designated to G0, G1,
|
|
|
771 G2, G3? If t, no charset is initially designated. If nil,
|
|
|
772 no charset is initially designated and no charset is allowed
|
|
|
773 to be designated. */
|
|
|
774 Lisp_Object initial_charset[4];
|
|
|
775
|
|
|
776 /* If true, a designation escape sequence needs to be sent on output
|
|
|
777 for the charset in G[0-3] before that charset is used. */
|
|
|
778 unsigned char force_charset_on_output[4];
|
|
|
779
|
|
|
780 charset_conversion_spec_dynarr *input_conv;
|
|
|
781 charset_conversion_spec_dynarr *output_conv;
|
|
|
782
|
|
|
783 unsigned int shoort :1; /* C makes you speak Dutch */
|
|
|
784 unsigned int no_ascii_eol :1;
|
|
|
785 unsigned int no_ascii_cntl :1;
|
|
|
786 unsigned int seven :1;
|
|
|
787 unsigned int lock_shift :1;
|
|
|
788 unsigned int no_iso6429 :1;
|
|
|
789 unsigned int escape_quoted :1;
|
|
|
790 };
|
|
|
791
|
|
|
792 #define CODING_SYSTEM_ISO2022_INITIAL_CHARSET(codesys, g) \
|
|
|
793 (CODING_SYSTEM_TYPE_DATA (codesys, iso2022)->initial_charset[g])
|
|
|
794 #define CODING_SYSTEM_ISO2022_FORCE_CHARSET_ON_OUTPUT(codesys, g) \
|
|
|
795 (CODING_SYSTEM_TYPE_DATA (codesys, iso2022)->force_charset_on_output[g])
|
|
|
796 #define CODING_SYSTEM_ISO2022_SHORT(codesys) \
|
|
|
797 (CODING_SYSTEM_TYPE_DATA (codesys, iso2022)->shoort)
|
|
|
798 #define CODING_SYSTEM_ISO2022_NO_ASCII_EOL(codesys) \
|
|
|
799 (CODING_SYSTEM_TYPE_DATA (codesys, iso2022)->no_ascii_eol)
|
|
|
800 #define CODING_SYSTEM_ISO2022_NO_ASCII_CNTL(codesys) \
|
|
|
801 (CODING_SYSTEM_TYPE_DATA (codesys, iso2022)->no_ascii_cntl)
|
|
|
802 #define CODING_SYSTEM_ISO2022_SEVEN(codesys) \
|
|
|
803 (CODING_SYSTEM_TYPE_DATA (codesys, iso2022)->seven)
|
|
|
804 #define CODING_SYSTEM_ISO2022_LOCK_SHIFT(codesys) \
|
|
|
805 (CODING_SYSTEM_TYPE_DATA (codesys, iso2022)->lock_shift)
|
|
|
806 #define CODING_SYSTEM_ISO2022_NO_ISO6429(codesys) \
|
|
|
807 (CODING_SYSTEM_TYPE_DATA (codesys, iso2022)->no_iso6429)
|
|
|
808 #define CODING_SYSTEM_ISO2022_ESCAPE_QUOTED(codesys) \
|
|
|
809 (CODING_SYSTEM_TYPE_DATA (codesys, iso2022)->escape_quoted)
|
|
|
810 #define CODING_SYSTEM_ISO2022_INPUT_CONV(codesys) \
|
|
|
811 (CODING_SYSTEM_TYPE_DATA (codesys, iso2022)->input_conv)
|
|
|
812 #define CODING_SYSTEM_ISO2022_OUTPUT_CONV(codesys) \
|
|
|
813 (CODING_SYSTEM_TYPE_DATA (codesys, iso2022)->output_conv)
|
|
|
814
|
|
|
815 #define XCODING_SYSTEM_ISO2022_INITIAL_CHARSET(codesys, g) \
|
|
|
816 CODING_SYSTEM_ISO2022_INITIAL_CHARSET (XCODING_SYSTEM (codesys), g)
|
|
|
817 #define XCODING_SYSTEM_ISO2022_FORCE_CHARSET_ON_OUTPUT(codesys, g) \
|
|
|
818 CODING_SYSTEM_ISO2022_FORCE_CHARSET_ON_OUTPUT (XCODING_SYSTEM (codesys), g)
|
|
|
819 #define XCODING_SYSTEM_ISO2022_SHORT(codesys) \
|
|
|
820 CODING_SYSTEM_ISO2022_SHORT (XCODING_SYSTEM (codesys))
|
|
|
821 #define XCODING_SYSTEM_ISO2022_NO_ASCII_EOL(codesys) \
|
|
|
822 CODING_SYSTEM_ISO2022_NO_ASCII_EOL (XCODING_SYSTEM (codesys))
|
|
|
823 #define XCODING_SYSTEM_ISO2022_NO_ASCII_CNTL(codesys) \
|
|
|
824 CODING_SYSTEM_ISO2022_NO_ASCII_CNTL (XCODING_SYSTEM (codesys))
|
|
|
825 #define XCODING_SYSTEM_ISO2022_SEVEN(codesys) \
|
|
|
826 CODING_SYSTEM_ISO2022_SEVEN (XCODING_SYSTEM (codesys))
|
|
|
827 #define XCODING_SYSTEM_ISO2022_LOCK_SHIFT(codesys) \
|
|
|
828 CODING_SYSTEM_ISO2022_LOCK_SHIFT (XCODING_SYSTEM (codesys))
|
|
|
829 #define XCODING_SYSTEM_ISO2022_NO_ISO6429(codesys) \
|
|
|
830 CODING_SYSTEM_ISO2022_NO_ISO6429 (XCODING_SYSTEM (codesys))
|
|
|
831 #define XCODING_SYSTEM_ISO2022_ESCAPE_QUOTED(codesys) \
|
|
|
832 CODING_SYSTEM_ISO2022_ESCAPE_QUOTED (XCODING_SYSTEM (codesys))
|
|
|
833 #define XCODING_SYSTEM_ISO2022_INPUT_CONV(codesys) \
|
|
|
834 CODING_SYSTEM_ISO2022_INPUT_CONV (XCODING_SYSTEM (codesys))
|
|
|
835 #define XCODING_SYSTEM_ISO2022_OUTPUT_CONV(codesys) \
|
|
|
836 CODING_SYSTEM_ISO2022_OUTPUT_CONV (XCODING_SYSTEM (codesys))
|
|
|
837
|
|
|
838 /* Additional information used by the ISO2022 decoder and detector. */
|
|
|
839 struct iso2022_coding_stream
|
|
|
840 {
|
|
|
841 /* CHARSET holds the character sets currently assigned to the G0
|
|
|
842 through G3 variables. It is initialized from the array
|
|
|
843 INITIAL_CHARSET in CODESYS. */
|
|
|
844 Lisp_Object charset[4];
|
|
|
845
|
|
|
846 /* Which registers are currently invoked into the left (GL) and
|
|
|
847 right (GR) halves of the 8-bit encoding space? */
|
|
|
848 int register_left, register_right;
|
|
|
849
|
|
|
850 /* FLAGS holds flags indicating the current state of the encoding. Some of
|
|
|
851 these flags are actually part of the state-dependent data and should be
|
|
|
852 moved there. */
|
|
|
853 unsigned int flags;
|
|
|
854
|
|
|
855 /**************** for decoding ****************/
|
|
|
856
|
|
|
857 /* ISO_ESC holds a value indicating part of an escape sequence
|
|
|
858 that has already been seen. */
|
|
|
859 enum iso_esc_flag esc;
|
|
|
860
|
|
|
861 /* This records the bytes we've seen so far in an escape sequence,
|
|
|
862 in case the sequence is invalid (we spit out the bytes unchanged). */
|
|
|
863 unsigned char esc_bytes[8];
|
|
|
864
|
|
|
865 /* Index for next byte to store in ISO escape sequence. */
|
|
|
866 int esc_bytes_index;
|
|
|
867
|
|
|
868 #ifdef ENABLE_COMPOSITE_CHARS
|
|
|
869 /* Stuff seen so far when composing a string. */
|
|
|
870 unsigned_char_dynarr *composite_chars;
|
|
|
871 #endif
|
|
|
872
|
|
|
873 /* If we saw an invalid designation sequence for a particular
|
|
|
874 register, we flag it here and switch to ASCII. The next time we
|
|
|
875 see a valid designation for this register, we turn off the flag
|
|
|
876 and do the designation normally, but pretend the sequence was
|
|
|
877 invalid. The effect of all this is that (most of the time) the
|
|
|
878 escape sequences for both the switch to the unknown charset, and
|
|
|
879 the switch back to the known charset, get inserted literally into
|
|
|
880 the buffer and saved out as such. The hope is that we can
|
|
|
881 preserve the escape sequences so that the resulting written out
|
|
|
882 file makes sense. If we don't do any of this, the designation
|
|
|
883 to the invalid charset will be preserved but that switch back
|
|
|
884 to the known charset will probably get eaten because it was
|
|
|
885 the same charset that was already present in the register. */
|
|
|
886 unsigned char invalid_designated[4];
|
|
|
887
|
|
|
888 /* We try to do similar things as above for direction-switching
|
|
|
889 sequences. If we encountered a direction switch while an
|
|
|
890 invalid designation was present, or an invalid designation
|
|
|
891 just after a direction switch (i.e. no valid designation
|
|
|
892 encountered yet), we insert the direction-switch escape
|
|
|
893 sequence literally into the output stream, and later on
|
|
|
894 insert the corresponding direction-restoring escape sequence
|
|
|
895 literally also. */
|
|
|
896 unsigned int switched_dir_and_no_valid_charset_yet :1;
|
|
|
897 unsigned int invalid_switch_dir :1;
|
|
|
898
|
|
|
899 /* Tells the decoder to output the escape sequence literally
|
|
|
900 even though it was valid. Used in the games we play to
|
|
|
901 avoid lossage when we encounter invalid designations. */
|
|
|
902 unsigned int output_literally :1;
|
|
|
903 /* We encountered a direction switch followed by an invalid
|
|
|
904 designation. We didn't output the direction switch
|
|
|
905 literally because we didn't know about the invalid designation;
|
|
|
906 but we have to do so now. */
|
|
|
907 unsigned int output_direction_sequence :1;
|
|
|
908
|
|
|
909 /**************** for encoding ****************/
|
|
|
910
|
|
|
911 /* Whether we need to explicitly designate the charset in the
|
|
|
912 G? register before using it. It is initialized from the
|
|
|
913 array FORCE_CHARSET_ON_OUTPUT in CODESYS. */
|
|
|
914 unsigned char force_charset_on_output[4];
|
|
|
915
|
|
|
916 /* Other state variables that need to be preserved across
|
|
|
917 invocations. */
|
|
|
918 Lisp_Object current_charset;
|
|
|
919 int current_half;
|
|
|
920 int current_char_boundary;
|
|
|
921 };
|
|
|
922
|
|
|
923 static const struct lrecord_description ccs_description_1[] =
|
|
|
924 {
|
|
|
925 { XD_LISP_OBJECT, offsetof (charset_conversion_spec, from_charset) },
|
|
|
926 { XD_LISP_OBJECT, offsetof (charset_conversion_spec, to_charset) },
|
|
|
927 { XD_END }
|
|
|
928 };
|
|
|
929
|
|
|
930 static const struct struct_description ccs_description =
|
|
|
931 {
|
|
|
932 sizeof (charset_conversion_spec),
|
|
|
933 ccs_description_1
|
|
|
934 };
|
|
|
935
|
|
|
936 static const struct lrecord_description ccsd_description_1[] =
|
|
|
937 {
|
|
|
938 XD_DYNARR_DESC (charset_conversion_spec_dynarr, &ccs_description),
|
|
|
939 { XD_END }
|
|
|
940 };
|
|
|
941
|
|
|
942 static const struct struct_description ccsd_description =
|
|
|
943 {
|
|
|
944 sizeof (charset_conversion_spec_dynarr),
|
|
|
945 ccsd_description_1
|
|
|
946 };
|
|
|
947
|
|
|
948 static const struct lrecord_description iso2022_coding_system_description[] = {
|
|
|
949 { XD_LISP_OBJECT_ARRAY,
|
|
|
950 coding_system_data_offset + offsetof (struct iso2022_coding_system,
|
|
|
951 initial_charset), 4 },
|
|
|
952 { XD_STRUCT_PTR,
|
|
|
953 coding_system_data_offset + offsetof (struct iso2022_coding_system,
|
|
|
954 input_conv),
|
|
|
955 1, &ccsd_description },
|
|
|
956 { XD_STRUCT_PTR,
|
|
|
957 coding_system_data_offset + offsetof (struct iso2022_coding_system,
|
|
|
958 output_conv),
|
|
|
959 1, &ccsd_description },
|
|
|
960 { XD_END }
|
|
|
961 };
|
|
|
962
|
|
|
963 /* The following note taken directly from FSF 21.0.103. */
|
|
|
964
|
|
|
965 /* The following note describes the coding system ISO2022 briefly.
|
|
|
966 Since the intention of this note is to help understand the
|
|
|
967 functions in this file, some parts are NOT ACCURATE or are OVERLY
|
|
|
968 SIMPLIFIED. For thorough understanding, please refer to the
|
|
|
969 original document of ISO2022. This is equivalent to the standard
|
|
|
970 ECMA-35, obtainable from <URL:http://www.ecma.ch/> (*).
|
|
|
971
|
|
|
972 ISO2022 provides many mechanisms to encode several character sets
|
|
|
973 in 7-bit and 8-bit environments. For 7-bit environments, all text
|
|
|
974 is encoded using bytes less than 128. This may make the encoded
|
|
|
975 text a little bit longer, but the text passes more easily through
|
|
|
976 several types of gateway, some of which strip off the MSB (Most
|
|
|
977 Significant Bit).
|
|
|
978
|
|
|
979 There are two kinds of character sets: control character sets and
|
|
|
980 graphic character sets. The former contain control characters such
|
|
|
981 as `newline' and `escape' to provide control functions (control
|
|
|
982 functions are also provided by escape sequences). The latter
|
|
|
983 contain graphic characters such as 'A' and '-'. Emacs recognizes
|
|
|
984 two control character sets and many graphic character sets.
|
|
|
985
|
|
|
986 Graphic character sets are classified into one of the following
|
|
|
987 four classes, according to the number of bytes (DIMENSION) and
|
|
|
988 number of characters in one dimension (CHARS) of the set:
|
|
|
989 - DIMENSION1_CHARS94
|
|
|
990 - DIMENSION1_CHARS96
|
|
|
991 - DIMENSION2_CHARS94
|
|
|
992 - DIMENSION2_CHARS96
|
|
|
993
|
|
|
994 In addition, each character set is assigned an identification tag,
|
|
|
995 unique for each set, called the "final character" (denoted as <F>
|
|
|
996 hereafter). The <F> of each character set is decided by ECMA(*)
|
|
|
997 when it is registered in ISO. The code range of <F> is 0x30..0x7F
|
|
|
998 (0x30..0x3F are for private use only).
|
|
|
999
|
|
|
1000 Note (*): ECMA = European Computer Manufacturers Association
|
|
|
1001
|
|
|
1002 Here are examples of graphic character sets [NAME(<F>)]:
|
|
|
1003 o DIMENSION1_CHARS94 -- ASCII('B'), right-half-of-JISX0201('I'), ...
|
|
|
1004 o DIMENSION1_CHARS96 -- right-half-of-ISO8859-1('A'), ...
|
|
|
1005 o DIMENSION2_CHARS94 -- GB2312('A'), JISX0208('B'), ...
|
|
|
1006 o DIMENSION2_CHARS96 -- none for the moment
|
|
|
1007
|
|
|
1008 A code area (1 byte=8 bits) is divided into 4 areas, C0, GL, C1, and GR.
|
|
|
1009 C0 [0x00..0x1F] -- control character plane 0
|
|
|
1010 GL [0x20..0x7F] -- graphic character plane 0
|
|
|
1011 C1 [0x80..0x9F] -- control character plane 1
|
|
|
1012 GR [0xA0..0xFF] -- graphic character plane 1
|
|
|
1013
|
|
|
1014 A control character set is directly designated and invoked to C0 or
|
|
|
1015 C1 by an escape sequence. The most common case is that:
|
|
|
1016 - ISO646's control character set is designated/invoked to C0, and
|
|
|
1017 - ISO6429's control character set is designated/invoked to C1,
|
|
|
1018 and usually these designations/invocations are omitted in encoded
|
|
|
1019 text. In a 7-bit environment, only C0 can be used, and a control
|
|
|
1020 character for C1 is encoded by an appropriate escape sequence to
|
|
|
1021 fit into the environment. All control characters for C1 are
|
|
|
1022 defined to have corresponding escape sequences.
|
|
|
1023
|
|
|
1024 A graphic character set is at first designated to one of four
|
|
|
1025 graphic registers (G0 through G3), then these graphic registers are
|
|
|
1026 invoked to GL or GR. These designations and invocations can be
|
|
|
1027 done independently. The most common case is that G0 is invoked to
|
|
|
1028 GL, G1 is invoked to GR, and ASCII is designated to G0. Usually
|
|
|
1029 these invocations and designations are omitted in encoded text.
|
|
|
1030 In a 7-bit environment, only GL can be used.
|
|
|
1031
|
|
|
1032 When a graphic character set of CHARS94 is invoked to GL, codes
|
|
|
1033 0x20 and 0x7F of the GL area work as control characters SPACE and
|
|
|
1034 DEL respectively, and codes 0xA0 and 0xFF of the GR area should not
|
|
|
1035 be used.
|
|
|
1036
|
|
|
1037 There are two ways of invocation: locking-shift and single-shift.
|
|
|
1038 With locking-shift, the invocation lasts until the next different
|
|
|
1039 invocation, whereas with single-shift, the invocation affects the
|
|
|
1040 following character only and doesn't affect the locking-shift
|
|
|
1041 state. Invocations are done by the following control characters or
|
|
|
1042 escape sequences:
|
|
|
1043
|
|
|
1044 ----------------------------------------------------------------------
|
|
|
1045 abbrev function cntrl escape seq description
|
|
|
1046 ----------------------------------------------------------------------
|
|
|
1047 SI/LS0 (shift-in) 0x0F none invoke G0 into GL
|
|
|
1048 SO/LS1 (shift-out) 0x0E none invoke G1 into GL
|
|
|
1049 LS2 (locking-shift-2) none ESC 'n' invoke G2 into GL
|
|
|
1050 LS3 (locking-shift-3) none ESC 'o' invoke G3 into GL
|
|
|
1051 LS1R (locking-shift-1 right) none ESC '~' invoke G1 into GR (*)
|
|
|
1052 LS2R (locking-shift-2 right) none ESC '}' invoke G2 into GR (*)
|
|
|
1053 LS3R (locking-shift 3 right) none ESC '|' invoke G3 into GR (*)
|
|
|
1054 SS2 (single-shift-2) 0x8E ESC 'N' invoke G2 for one char
|
|
|
1055 SS3 (single-shift-3) 0x8F ESC 'O' invoke G3 for one char
|
|
|
1056 ----------------------------------------------------------------------
|
|
|
1057 (*) These are not used by any known coding system.
|
|
|
1058
|
|
|
1059 Control characters for these functions are defined by macros
|
|
|
1060 ISO_CODE_XXX in `coding.h'.
|
|
|
1061
|
|
|
1062 Designations are done by the following escape sequences:
|
|
|
1063 ----------------------------------------------------------------------
|
|
|
1064 escape sequence description
|
|
|
1065 ----------------------------------------------------------------------
|
|
|
1066 ESC '(' <F> designate DIMENSION1_CHARS94<F> to G0
|
|
|
1067 ESC ')' <F> designate DIMENSION1_CHARS94<F> to G1
|
|
|
1068 ESC '*' <F> designate DIMENSION1_CHARS94<F> to G2
|
|
|
1069 ESC '+' <F> designate DIMENSION1_CHARS94<F> to G3
|
|
|
1070 ESC ',' <F> designate DIMENSION1_CHARS96<F> to G0 (*)
|
|
|
1071 ESC '-' <F> designate DIMENSION1_CHARS96<F> to G1
|
|
|
1072 ESC '.' <F> designate DIMENSION1_CHARS96<F> to G2
|
|
|
1073 ESC '/' <F> designate DIMENSION1_CHARS96<F> to G3
|
|
|
1074 ESC '$' '(' <F> designate DIMENSION2_CHARS94<F> to G0 (**)
|
|
|
1075 ESC '$' ')' <F> designate DIMENSION2_CHARS94<F> to G1
|
|
|
1076 ESC '$' '*' <F> designate DIMENSION2_CHARS94<F> to G2
|
|
|
1077 ESC '$' '+' <F> designate DIMENSION2_CHARS94<F> to G3
|
|
|
1078 ESC '$' ',' <F> designate DIMENSION2_CHARS96<F> to G0 (*)
|
|
|
1079 ESC '$' '-' <F> designate DIMENSION2_CHARS96<F> to G1
|
|
|
1080 ESC '$' '.' <F> designate DIMENSION2_CHARS96<F> to G2
|
|
|
1081 ESC '$' '/' <F> designate DIMENSION2_CHARS96<F> to G3
|
|
|
1082 ----------------------------------------------------------------------
|
|
|
1083
|
|
|
1084 In this list, "DIMENSION1_CHARS94<F>" means a graphic character set
|
|
|
1085 of dimension 1, chars 94, and final character <F>, etc...
|
|
|
1086
|
|
|
1087 Note (*): Although these designations are not allowed in ISO2022,
|
|
|
1088 Emacs accepts them on decoding, and produces them on encoding
|
|
|
1089 CHARS96 character sets in a coding system which is characterized as
|
|
|
1090 7-bit environment, non-locking-shift, and non-single-shift.
|
|
|
1091
|
|
|
1092 Note (**): If <F> is '@', 'A', or 'B', the intermediate character
|
|
|
1093 '(' can be omitted. We refer to this as "short-form" hereafter.
|
|
|
1094
|
|
|
1095 Now you may notice that there are a lot of ways of encoding the
|
|
|
1096 same multilingual text in ISO2022. Actually, there exist many
|
|
|
1097 coding systems such as Compound Text (used in X11's inter client
|
|
|
1098 communication, ISO-2022-JP (used in Japanese Internet), ISO-2022-KR
|
|
|
1099 (used in Korean Internet), EUC (Extended UNIX Code, used in Asian
|
|
|
1100 localized platforms), and all of these are variants of ISO2022.
|
|
|
1101
|
|
|
1102 In addition to the above, Emacs handles two more kinds of escape
|
|
|
1103 sequences: ISO6429's direction specification and Emacs' private
|
|
|
1104 sequence for specifying character composition.
|
|
|
1105
|
|
|
1106 ISO6429's direction specification takes the following form:
|
|
|
1107 o CSI ']' -- end of the current direction
|
|
|
1108 o CSI '0' ']' -- end of the current direction
|
|
|
1109 o CSI '1' ']' -- start of left-to-right text
|
|
|
1110 o CSI '2' ']' -- start of right-to-left text
|
|
|
1111 The control character CSI (0x9B: control sequence introducer) is
|
|
|
1112 abbreviated to the escape sequence ESC '[' in a 7-bit environment.
|
|
|
1113
|
|
|
1114 Character composition specification takes the following form:
|
|
|
1115 o ESC '0' -- start relative composition
|
|
|
1116 o ESC '1' -- end composition
|
|
|
1117 o ESC '2' -- start rule-base composition (*)
|
|
|
1118 o ESC '3' -- start relative composition with alternate chars (**)
|
|
|
1119 o ESC '4' -- start rule-base composition with alternate chars (**)
|
|
|
1120 Since these are not standard escape sequences of any ISO standard,
|
|
|
1121 the use of them with these meanings is restricted to Emacs only.
|
|
|
1122
|
|
|
1123 (*) This form is used only in Emacs 20.5 and older versions,
|
|
|
1124 but the newer versions can safely decode it.
|
|
|
1125 (**) This form is used only in Emacs 21.1 and newer versions,
|
|
|
1126 and the older versions can't decode it.
|
|
|
1127
|
|
|
1128 Here's a list of example usages of these composition escape
|
|
|
1129 sequences (categorized by `enum composition_method').
|
|
|
1130
|
|
|
1131 COMPOSITION_RELATIVE:
|
|
|
1132 ESC 0 CHAR [ CHAR ] ESC 1
|
|
|
1133 COMPOSITION_WITH_RULE:
|
|
|
1134 ESC 2 CHAR [ RULE CHAR ] ESC 1
|
|
|
1135 COMPOSITION_WITH_ALTCHARS:
|
|
|
1136 ESC 3 ALTCHAR [ ALTCHAR ] ESC 0 CHAR [ CHAR ] ESC 1
|
|
|
1137 COMPOSITION_WITH_RULE_ALTCHARS:
|
|
|
1138 ESC 4 ALTCHAR [ RULE ALTCHAR ] ESC 0 CHAR [ CHAR ] ESC 1 */
|
|
|
1139
|
|
|
1140 static void
|
|
|
1141 reset_iso2022_decode (Lisp_Object coding_system,
|
|
|
1142 struct iso2022_coding_stream *data)
|
|
|
1143 {
|
|
|
1144 int i;
|
|
|
1145 #ifdef ENABLE_COMPOSITE_CHARS
|
|
|
1146 unsigned_char_dynarr *old_composite_chars = data->composite_chars;
|
|
|
1147 #endif
|
|
|
1148
|
|
|
1149 xzero (*data);
|
|
|
1150
|
|
|
1151 for (i = 0; i < 4; i++)
|
|
|
1152 {
|
|
|
1153 if (!NILP (coding_system))
|
|
|
1154 data->charset[i] =
|
|
|
1155 XCODING_SYSTEM_ISO2022_INITIAL_CHARSET (coding_system, i);
|
|
|
1156 else
|
|
|
1157 data->charset[i] = Qt;
|
|
|
1158 }
|
|
|
1159 data->esc = ISO_ESC_NOTHING;
|
|
|
1160 data->register_right = 1;
|
|
|
1161 #ifdef ENABLE_COMPOSITE_CHARS
|
|
|
1162 if (old_composite_chars)
|
|
|
1163 {
|
|
|
1164 data->composite_chars = old_composite_chars;
|
|
|
1165 Dynarr_reset (data->composite_chars);
|
|
|
1166 }
|
|
|
1167 #endif
|
|
|
1168 }
|
|
|
1169
|
|
|
1170 static void
|
|
|
1171 reset_iso2022_encode (Lisp_Object coding_system,
|
|
|
1172 struct iso2022_coding_stream *data)
|
|
|
1173 {
|
|
|
1174 int i;
|
|
|
1175
|
|
|
1176 xzero (*data);
|
|
|
1177
|
|
|
1178 for (i = 0; i < 4; i++)
|
|
|
1179 {
|
|
|
1180 data->charset[i] =
|
|
|
1181 XCODING_SYSTEM_ISO2022_INITIAL_CHARSET (coding_system, i);
|
|
|
1182 data->force_charset_on_output[i] =
|
|
|
1183 XCODING_SYSTEM_ISO2022_FORCE_CHARSET_ON_OUTPUT (coding_system, i);
|
|
|
1184 }
|
|
|
1185 data->register_right = 1;
|
|
|
1186 data->current_charset = Qnil;
|
|
|
1187 data->current_char_boundary = 1;
|
|
|
1188 }
|
|
|
1189
|
|
|
1190 static void
|
|
|
1191 iso2022_init_coding_stream (struct coding_stream *str)
|
|
|
1192 {
|
|
|
1193 if (str->direction == CODING_DECODE)
|
|
|
1194 reset_iso2022_decode (str->codesys,
|
|
|
1195 CODING_STREAM_TYPE_DATA (str, iso2022));
|
|
|
1196 else
|
|
|
1197 reset_iso2022_encode (str->codesys,
|
|
|
1198 CODING_STREAM_TYPE_DATA (str, iso2022));
|
|
|
1199 }
|
|
|
1200
|
|
|
1201 static void
|
|
|
1202 iso2022_rewind_coding_stream (struct coding_stream *str)
|
|
|
1203 {
|
|
|
1204 iso2022_init_coding_stream (str);
|
|
|
1205 }
|
|
|
1206
|
|
|
1207 static int
|
|
|
1208 fit_to_be_escape_quoted (unsigned char c)
|
|
|
1209 {
|
|
|
1210 switch (c)
|
|
|
1211 {
|
|
|
1212 case ISO_CODE_ESC:
|
|
|
1213 case ISO_CODE_CSI:
|
|
|
1214 case ISO_CODE_SS2:
|
|
|
1215 case ISO_CODE_SS3:
|
|
|
1216 case ISO_CODE_SO:
|
|
|
1217 case ISO_CODE_SI:
|
|
|
1218 return 1;
|
|
|
1219
|
|
|
1220 default:
|
|
|
1221 return 0;
|
|
|
1222 }
|
|
|
1223 }
|
|
|
1224
|
|
|
1225 static Lisp_Object
|
|
867
|
1226 charset_by_attributes_or_create_one (int type, Ibyte final, int dir)
|
|
771
|
1227 {
|
|
826
|
1228 Lisp_Object charset = charset_by_attributes (type, final, dir);
|
|
771
|
1229
|
|
|
1230 if (NILP (charset))
|
|
|
1231 {
|
|
|
1232 int chars, dim;
|
|
|
1233
|
|
|
1234 switch (type)
|
|
|
1235 {
|
|
|
1236 case CHARSET_TYPE_94:
|
|
|
1237 chars = 94; dim = 1;
|
|
|
1238 break;
|
|
|
1239 case CHARSET_TYPE_96:
|
|
|
1240 chars = 96; dim = 1;
|
|
|
1241 break;
|
|
|
1242 case CHARSET_TYPE_94X94:
|
|
|
1243 chars = 94; dim = 2;
|
|
|
1244 break;
|
|
|
1245 case CHARSET_TYPE_96X96:
|
|
|
1246 chars = 96; dim = 2;
|
|
|
1247 break;
|
|
|
1248 default:
|
|
|
1249 abort (); chars = 0; dim = 0;
|
|
|
1250 }
|
|
|
1251
|
|
|
1252 charset = Fmake_charset (Qunbound, Qnil,
|
|
|
1253 nconc2 (list6 (Qfinal, make_char (final),
|
|
|
1254 Qchars, make_int (chars),
|
|
|
1255 Qdimension, make_int (dim)),
|
|
|
1256 list2 (Qdirection,
|
|
|
1257 dir == CHARSET_LEFT_TO_RIGHT ?
|
|
|
1258 Ql2r : Qr2l)));
|
|
|
1259 }
|
|
|
1260
|
|
|
1261 return charset;
|
|
|
1262 }
|
|
|
1263
|
|
|
1264 /* Parse one byte of an ISO2022 escape sequence.
|
|
|
1265 If the result is an invalid escape sequence, return 0 and
|
|
|
1266 do not change anything in STR. Otherwise, if the result is
|
|
|
1267 an incomplete escape sequence, update ISO2022.ESC and
|
|
|
1268 ISO2022.ESC_BYTES and return -1. Otherwise, update
|
|
|
1269 all the state variables (but not ISO2022.ESC_BYTES) and
|
|
|
1270 return 1.
|
|
|
1271
|
|
|
1272 If CHECK_INVALID_CHARSETS is non-zero, check for designation
|
|
|
1273 or invocation of an invalid character set and treat that as
|
|
|
1274 an unrecognized escape sequence.
|
|
|
1275
|
|
|
1276 ********************************************************************
|
|
|
1277
|
|
|
1278 #### Strategies for error annotation and coding orthogonalization
|
|
|
1279
|
|
|
1280 We really want to separate out a number of things. Conceptually,
|
|
|
1281 there is a nested syntax.
|
|
|
1282
|
|
|
1283 At the top level is the ISO 2022 extension syntax, including charset
|
|
|
1284 designation and invocation, and certain auxiliary controls such as the
|
|
|
1285 ISO 6429 direction specification. These are octet-oriented, with the
|
|
|
1286 single exception (AFAIK) of the "exit Unicode" sequence which uses the
|
|
|
1287 UTF's natural width (1 byte for UTF-7 and UTF-8, 2 bytes for UCS-2 and
|
|
|
1288 UTF-16, and 4 bytes for UCS-4 and UTF-32). This will be treated as a
|
|
|
1289 (deprecated) special case in Unicode processing.
|
|
|
1290
|
|
|
1291 The middle layer is ISO 2022 character interpretation. This will depend
|
|
|
1292 on the current state of the ISO 2022 registers, and assembles octets
|
|
|
1293 into the character's internal representation.
|
|
|
1294
|
|
|
1295 The lowest level is translating system control conventions. At present
|
|
|
1296 this is restricted to newline translation, but one could imagine doing
|
|
|
1297 tab conversion or line wrapping here. "Escape from Unicode" processing
|
|
|
1298 would be done at this level.
|
|
|
1299
|
|
|
1300 At each level the parser will verify the syntax. In the case of a
|
|
|
1301 syntax error or warning (such as a redundant escape sequence that affects
|
|
|
1302 no characters), the parser will take some action, typically inserting the
|
|
|
1303 erroneous octets directly into the output and creating an annotation
|
|
|
1304 which can be used by higher level I/O to mark the affected region.
|
|
|
1305
|
|
|
1306 This should make it possible to do something sensible about separating
|
|
|
1307 newline convention processing from character construction, and about
|
|
|
1308 preventing ISO 2022 escape sequences from being recognized
|
|
|
1309 inappropriately.
|
|
|
1310
|
|
|
1311 The basic strategy will be to have octet classification tables, and
|
|
|
1312 switch processing according to the table entry.
|
|
|
1313
|
|
|
1314 It's possible that, by doing the processing with tables of functions or
|
|
|
1315 the like, the parser can be used for both detection and translation. */
|
|
|
1316
|
|
|
1317 static int
|
|
|
1318 parse_iso2022_esc (Lisp_Object codesys, struct iso2022_coding_stream *iso,
|
|
|
1319 unsigned char c, unsigned int *flags,
|
|
|
1320 int check_invalid_charsets)
|
|
|
1321 {
|
|
|
1322 /* (1) If we're at the end of a designation sequence, CS is the
|
|
|
1323 charset being designated and REG is the register to designate
|
|
|
1324 it to.
|
|
|
1325
|
|
|
1326 (2) If we're at the end of a locking-shift sequence, REG is
|
|
|
1327 the register to invoke and HALF (0 == left, 1 == right) is
|
|
|
1328 the half to invoke it into.
|
|
|
1329
|
|
|
1330 (3) If we're at the end of a single-shift sequence, REG is
|
|
|
1331 the register to invoke. */
|
|
|
1332 Lisp_Object cs = Qnil;
|
|
|
1333 int reg, half;
|
|
|
1334
|
|
|
1335 /* NOTE: This code does goto's all over the fucking place.
|
|
|
1336 The reason for this is that we're basically implementing
|
|
|
1337 a state machine here, and hierarchical languages like C
|
|
|
1338 don't really provide a clean way of doing this. */
|
|
|
1339
|
|
|
1340 if (! (*flags & ISO_STATE_ESCAPE))
|
|
|
1341 /* At beginning of escape sequence; we need to reset our
|
|
|
1342 escape-state variables. */
|
|
|
1343 iso->esc = ISO_ESC_NOTHING;
|
|
|
1344
|
|
|
1345 iso->output_literally = 0;
|
|
|
1346 iso->output_direction_sequence = 0;
|
|
|
1347
|
|
|
1348 switch (iso->esc)
|
|
|
1349 {
|
|
|
1350 case ISO_ESC_NOTHING:
|
|
|
1351 iso->esc_bytes_index = 0;
|
|
|
1352 switch (c)
|
|
|
1353 {
|
|
|
1354 case ISO_CODE_ESC: /* Start escape sequence */
|
|
|
1355 *flags |= ISO_STATE_ESCAPE;
|
|
|
1356 iso->esc = ISO_ESC;
|
|
|
1357 goto not_done;
|
|
|
1358
|
|
|
1359 case ISO_CODE_CSI: /* ISO6429 (specifying directionality) */
|
|
|
1360 *flags |= ISO_STATE_ESCAPE;
|
|
|
1361 iso->esc = ISO_ESC_5_11;
|
|
|
1362 goto not_done;
|
|
|
1363
|
|
|
1364 case ISO_CODE_SO: /* locking shift 1 */
|
|
|
1365 reg = 1; half = 0;
|
|
|
1366 goto locking_shift;
|
|
|
1367 case ISO_CODE_SI: /* locking shift 0 */
|
|
|
1368 reg = 0; half = 0;
|
|
|
1369 goto locking_shift;
|
|
|
1370
|
|
|
1371 case ISO_CODE_SS2: /* single shift */
|
|
|
1372 reg = 2;
|
|
|
1373 goto single_shift;
|
|
|
1374 case ISO_CODE_SS3: /* single shift */
|
|
|
1375 reg = 3;
|
|
|
1376 goto single_shift;
|
|
|
1377
|
|
|
1378 default: /* Other control characters */
|
|
|
1379 error:
|
|
|
1380 *flags &= ISO_STATE_LOCK;
|
|
|
1381 return 0;
|
|
|
1382 }
|
|
|
1383
|
|
|
1384 case ISO_ESC:
|
|
|
1385 switch (c)
|
|
|
1386 {
|
|
|
1387 /**** single shift ****/
|
|
|
1388
|
|
|
1389 case 'N': /* single shift 2 */
|
|
|
1390 reg = 2;
|
|
|
1391 goto single_shift;
|
|
|
1392 case 'O': /* single shift 3 */
|
|
|
1393 reg = 3;
|
|
|
1394 goto single_shift;
|
|
|
1395
|
|
|
1396 /**** locking shift ****/
|
|
|
1397
|
|
|
1398 case '~': /* locking shift 1 right */
|
|
|
1399 reg = 1; half = 1;
|
|
|
1400 goto locking_shift;
|
|
|
1401 case 'n': /* locking shift 2 */
|
|
|
1402 reg = 2; half = 0;
|
|
|
1403 goto locking_shift;
|
|
|
1404 case '}': /* locking shift 2 right */
|
|
|
1405 reg = 2; half = 1;
|
|
|
1406 goto locking_shift;
|
|
|
1407 case 'o': /* locking shift 3 */
|
|
|
1408 reg = 3; half = 0;
|
|
|
1409 goto locking_shift;
|
|
|
1410 case '|': /* locking shift 3 right */
|
|
|
1411 reg = 3; half = 1;
|
|
|
1412 goto locking_shift;
|
|
|
1413
|
|
|
1414 /**** composite ****/
|
|
|
1415
|
|
|
1416 #ifdef ENABLE_COMPOSITE_CHARS
|
|
|
1417 case '0':
|
|
|
1418 iso->esc = ISO_ESC_START_COMPOSITE;
|
|
|
1419 *flags = (*flags & ISO_STATE_LOCK) |
|
|
|
1420 ISO_STATE_COMPOSITE;
|
|
|
1421 return 1;
|
|
|
1422
|
|
|
1423 case '1':
|
|
|
1424 iso->esc = ISO_ESC_END_COMPOSITE;
|
|
|
1425 *flags = (*flags & ISO_STATE_LOCK) &
|
|
|
1426 ~ISO_STATE_COMPOSITE;
|
|
|
1427 return 1;
|
|
|
1428 #else
|
|
|
1429 case '0': case '1': case '2': case '3': case '4':
|
|
|
1430 /* We simply return a flag indicating that some composite
|
|
|
1431 escape was seen. The caller will use the particular
|
|
|
1432 character to encode the appropriate "composite hack"
|
|
|
1433 character out of Vcharset_composite, so that we will
|
|
|
1434 preserve these values on output. */
|
|
|
1435 iso->esc = ISO_ESC_START_COMPOSITE;
|
|
|
1436 *flags &= ISO_STATE_LOCK;
|
|
|
1437 return 1;
|
|
|
1438 #endif /* ENABLE_COMPOSITE_CHARS */
|
|
|
1439
|
|
|
1440 /**** directionality ****/
|
|
|
1441
|
|
|
1442 case '[':
|
|
|
1443 iso->esc = ISO_ESC_5_11;
|
|
|
1444 goto not_done;
|
|
|
1445
|
|
|
1446 /**** designation ****/
|
|
|
1447
|
|
|
1448 case '$': /* multibyte charset prefix */
|
|
|
1449 iso->esc = ISO_ESC_2_4;
|
|
|
1450 goto not_done;
|
|
|
1451
|
|
|
1452 default:
|
|
|
1453 if (0x28 <= c && c <= 0x2F)
|
|
|
1454 {
|
|
|
1455 iso->esc = (enum iso_esc_flag) (c - 0x28 + ISO_ESC_2_8);
|
|
|
1456 goto not_done;
|
|
|
1457 }
|
|
|
1458
|
|
|
1459 /* This function is called with CODESYS equal to nil when
|
|
|
1460 doing coding-system detection. */
|
|
|
1461 if (!NILP (codesys)
|
|
|
1462 && XCODING_SYSTEM_ISO2022_ESCAPE_QUOTED (codesys)
|
|
|
1463 && fit_to_be_escape_quoted (c))
|
|
|
1464 {
|
|
|
1465 iso->esc = ISO_ESC_LITERAL;
|
|
|
1466 *flags &= ISO_STATE_LOCK;
|
|
|
1467 return 1;
|
|
|
1468 }
|
|
|
1469
|
|
|
1470 /* bzzzt! */
|
|
|
1471 goto error;
|
|
|
1472 }
|
|
|
1473
|
|
|
1474
|
|
|
1475
|
|
|
1476 /**** directionality ****/
|
|
|
1477
|
|
|
1478 case ISO_ESC_5_11: /* ISO6429 direction control */
|
|
|
1479 if (c == ']')
|
|
|
1480 {
|
|
|
1481 *flags &= (ISO_STATE_LOCK & ~ISO_STATE_R2L);
|
|
|
1482 goto directionality;
|
|
|
1483 }
|
|
|
1484 if (c == '0') iso->esc = ISO_ESC_5_11_0;
|
|
|
1485 else if (c == '1') iso->esc = ISO_ESC_5_11_1;
|
|
|
1486 else if (c == '2') iso->esc = ISO_ESC_5_11_2;
|
|
|
1487 else goto error;
|
|
|
1488 goto not_done;
|
|
|
1489
|
|
|
1490 case ISO_ESC_5_11_0:
|
|
|
1491 if (c == ']')
|
|
|
1492 {
|
|
|
1493 *flags &= (ISO_STATE_LOCK & ~ISO_STATE_R2L);
|
|
|
1494 goto directionality;
|
|
|
1495 }
|
|
|
1496 goto error;
|
|
|
1497
|
|
|
1498 case ISO_ESC_5_11_1:
|
|
|
1499 if (c == ']')
|
|
|
1500 {
|
|
|
1501 *flags = (ISO_STATE_LOCK & ~ISO_STATE_R2L);
|
|
|
1502 goto directionality;
|
|
|
1503 }
|
|
|
1504 goto error;
|
|
|
1505
|
|
|
1506 case ISO_ESC_5_11_2:
|
|
|
1507 if (c == ']')
|
|
|
1508 {
|
|
|
1509 *flags = (*flags & ISO_STATE_LOCK) | ISO_STATE_R2L;
|
|
|
1510 goto directionality;
|
|
|
1511 }
|
|
|
1512 goto error;
|
|
|
1513
|
|
|
1514 directionality:
|
|
|
1515 iso->esc = ISO_ESC_DIRECTIONALITY;
|
|
|
1516 /* Various junk here to attempt to preserve the direction sequences
|
|
|
1517 literally in the text if they would otherwise be swallowed due
|
|
|
1518 to invalid designations that don't show up as actual charset
|
|
|
1519 changes in the text. */
|
|
|
1520 if (iso->invalid_switch_dir)
|
|
|
1521 {
|
|
|
1522 /* We already inserted a direction switch literally into the
|
|
|
1523 text. We assume (#### this may not be right) that the
|
|
|
1524 next direction switch is the one going the other way,
|
|
|
1525 and we need to output that literally as well. */
|
|
|
1526 iso->output_literally = 1;
|
|
|
1527 iso->invalid_switch_dir = 0;
|
|
|
1528 }
|
|
|
1529 else
|
|
|
1530 {
|
|
|
1531 int jj;
|
|
|
1532
|
|
|
1533 /* If we are in the thrall of an invalid designation,
|
|
|
1534 then stick the directionality sequence literally into the
|
|
|
1535 output stream so it ends up in the original text again. */
|
|
|
1536 for (jj = 0; jj < 4; jj++)
|
|
|
1537 if (iso->invalid_designated[jj])
|
|
|
1538 break;
|
|
|
1539 if (jj < 4)
|
|
|
1540 {
|
|
|
1541 iso->output_literally = 1;
|
|
|
1542 iso->invalid_switch_dir = 1;
|
|
|
1543 }
|
|
|
1544 else
|
|
|
1545 /* Indicate that we haven't yet seen a valid designation,
|
|
|
1546 so that if a switch-dir is directly followed by an
|
|
|
1547 invalid designation, both get inserted literally. */
|
|
|
1548 iso->switched_dir_and_no_valid_charset_yet = 1;
|
|
|
1549 }
|
|
|
1550 return 1;
|
|
|
1551
|
|
|
1552
|
|
|
1553 /**** designation ****/
|
|
|
1554
|
|
|
1555 case ISO_ESC_2_4:
|
|
|
1556 if (0x28 <= c && c <= 0x2F)
|
|
|
1557 {
|
|
|
1558 iso->esc = (enum iso_esc_flag) (c - 0x28 + ISO_ESC_2_4_8);
|
|
|
1559 goto not_done;
|
|
|
1560 }
|
|
|
1561 if (0x40 <= c && c <= 0x42)
|
|
|
1562 {
|
|
|
1563 cs = charset_by_attributes_or_create_one (CHARSET_TYPE_94X94, c,
|
|
|
1564 *flags & ISO_STATE_R2L ?
|
|
|
1565 CHARSET_RIGHT_TO_LEFT :
|
|
|
1566 CHARSET_LEFT_TO_RIGHT);
|
|
|
1567 reg = 0;
|
|
|
1568 goto designated;
|
|
|
1569 }
|
|
|
1570 goto error;
|
|
|
1571
|
|
|
1572 default:
|
|
|
1573 {
|
|
|
1574 int type = -1;
|
|
|
1575
|
|
|
1576 if (iso->esc >= ISO_ESC_2_8 &&
|
|
|
1577 iso->esc <= ISO_ESC_2_15)
|
|
|
1578 {
|
|
|
1579 type = ((iso->esc >= ISO_ESC_2_12) ?
|
|
|
1580 CHARSET_TYPE_96 : CHARSET_TYPE_94);
|
|
|
1581 reg = (iso->esc - ISO_ESC_2_8) & 3;
|
|
|
1582 }
|
|
|
1583 else if (iso->esc >= ISO_ESC_2_4_8 &&
|
|
|
1584 iso->esc <= ISO_ESC_2_4_15)
|
|
|
1585 {
|
|
|
1586 type = ((iso->esc >= ISO_ESC_2_4_12) ?
|
|
|
1587 CHARSET_TYPE_96X96 : CHARSET_TYPE_94X94);
|
|
|
1588 reg = (iso->esc - ISO_ESC_2_4_8) & 3;
|
|
|
1589 }
|
|
|
1590 else
|
|
|
1591 {
|
|
|
1592 /* Can this ever be reached? -slb */
|
|
|
1593 abort ();
|
|
|
1594 goto error;
|
|
|
1595 }
|
|
|
1596
|
|
|
1597 if (c < '0' || c > '~' ||
|
|
|
1598 (c > 0x5F && (type == CHARSET_TYPE_94X94 ||
|
|
|
1599 type == CHARSET_TYPE_96X96)))
|
|
|
1600 goto error; /* bad final byte */
|
|
|
1601
|
|
|
1602 cs = charset_by_attributes_or_create_one (type, c,
|
|
|
1603 *flags & ISO_STATE_R2L ?
|
|
|
1604 CHARSET_RIGHT_TO_LEFT :
|
|
|
1605 CHARSET_LEFT_TO_RIGHT);
|
|
|
1606 goto designated;
|
|
|
1607 }
|
|
|
1608 }
|
|
|
1609
|
|
|
1610 not_done:
|
|
|
1611 iso->esc_bytes[iso->esc_bytes_index++] = (unsigned char) c;
|
|
|
1612 return -1;
|
|
|
1613
|
|
|
1614 single_shift:
|
|
|
1615 if (check_invalid_charsets && !CHARSETP (iso->charset[reg]))
|
|
|
1616 /* can't invoke something that ain't there. */
|
|
|
1617 goto error;
|
|
|
1618 iso->esc = ISO_ESC_SINGLE_SHIFT;
|
|
|
1619 *flags &= ISO_STATE_LOCK;
|
|
|
1620 if (reg == 2)
|
|
|
1621 *flags |= ISO_STATE_SS2;
|
|
|
1622 else
|
|
|
1623 *flags |= ISO_STATE_SS3;
|
|
|
1624 return 1;
|
|
|
1625
|
|
|
1626 locking_shift:
|
|
|
1627 if (check_invalid_charsets &&
|
|
|
1628 !CHARSETP (iso->charset[reg]))
|
|
|
1629 /* can't invoke something that ain't there. */
|
|
|
1630 goto error;
|
|
|
1631 if (half)
|
|
|
1632 iso->register_right = reg;
|
|
|
1633 else
|
|
|
1634 iso->register_left = reg;
|
|
|
1635 *flags &= ISO_STATE_LOCK;
|
|
|
1636 iso->esc = ISO_ESC_LOCKING_SHIFT;
|
|
|
1637 return 1;
|
|
|
1638
|
|
|
1639 designated:
|
|
|
1640 if (NILP (cs) && check_invalid_charsets)
|
|
|
1641 {
|
|
|
1642 abort ();
|
|
|
1643 /* #### This should never happen now that we automatically create
|
|
|
1644 temporary charsets as necessary. We should probably remove
|
|
|
1645 this code. --ben */
|
|
|
1646 iso->invalid_designated[reg] = 1;
|
|
|
1647 iso->charset[reg] = Vcharset_ascii;
|
|
|
1648 iso->esc = ISO_ESC_DESIGNATE;
|
|
|
1649 *flags &= ISO_STATE_LOCK;
|
|
|
1650 iso->output_literally = 1;
|
|
|
1651 if (iso->switched_dir_and_no_valid_charset_yet)
|
|
|
1652 {
|
|
|
1653 /* We encountered a switch-direction followed by an
|
|
|
1654 invalid designation. Ensure that the switch-direction
|
|
|
1655 gets outputted; otherwise it will probably get eaten
|
|
|
1656 when the text is written out again. */
|
|
|
1657 iso->switched_dir_and_no_valid_charset_yet = 0;
|
|
|
1658 iso->output_direction_sequence = 1;
|
|
|
1659 /* And make sure that the switch-dir going the other
|
|
|
1660 way gets outputted, as well. */
|
|
|
1661 iso->invalid_switch_dir = 1;
|
|
|
1662 }
|
|
|
1663 return 1;
|
|
|
1664 }
|
|
|
1665 /* This function is called with CODESYS equal to nil when
|
|
|
1666 doing coding-system detection. */
|
|
|
1667 if (!NILP (codesys))
|
|
|
1668 {
|
|
|
1669 charset_conversion_spec_dynarr *dyn =
|
|
|
1670 XCODING_SYSTEM_ISO2022_INPUT_CONV (codesys);
|
|
|
1671
|
|
|
1672 if (dyn)
|
|
|
1673 {
|
|
|
1674 int i;
|
|
|
1675
|
|
|
1676 for (i = 0; i < Dynarr_length (dyn); i++)
|
|
|
1677 {
|
|
|
1678 struct charset_conversion_spec *spec = Dynarr_atp (dyn, i);
|
|
|
1679 if (EQ (cs, spec->from_charset))
|
|
|
1680 cs = spec->to_charset;
|
|
|
1681 }
|
|
|
1682 }
|
|
|
1683 }
|
|
|
1684
|
|
|
1685 iso->charset[reg] = cs;
|
|
|
1686 iso->esc = ISO_ESC_DESIGNATE;
|
|
|
1687 *flags &= ISO_STATE_LOCK;
|
|
|
1688 if (iso->invalid_designated[reg])
|
|
|
1689 {
|
|
|
1690 iso->invalid_designated[reg] = 0;
|
|
|
1691 iso->output_literally = 1;
|
|
|
1692 }
|
|
|
1693 if (iso->switched_dir_and_no_valid_charset_yet)
|
|
|
1694 iso->switched_dir_and_no_valid_charset_yet = 0;
|
|
|
1695 return 1;
|
|
|
1696 }
|
|
|
1697
|
|
|
1698 /* If FLAGS is a null pointer or specifies right-to-left motion,
|
|
|
1699 output a switch-dir-to-left-to-right sequence to DST.
|
|
|
1700 Also update FLAGS if it is not a null pointer.
|
|
|
1701 If INTERNAL_P is set, we are outputting in internal format and
|
|
|
1702 need to handle the CSI differently. */
|
|
|
1703
|
|
|
1704 static void
|
|
|
1705 restore_left_to_right_direction (Lisp_Object codesys,
|
|
|
1706 unsigned_char_dynarr *dst,
|
|
|
1707 unsigned int *flags,
|
|
|
1708 int internal_p)
|
|
|
1709 {
|
|
|
1710 if (!flags || (*flags & ISO_STATE_R2L))
|
|
|
1711 {
|
|
|
1712 if (XCODING_SYSTEM_ISO2022_SEVEN (codesys))
|
|
|
1713 {
|
|
|
1714 Dynarr_add (dst, ISO_CODE_ESC);
|
|
|
1715 Dynarr_add (dst, '[');
|
|
|
1716 }
|
|
|
1717 else if (internal_p)
|
|
|
1718 DECODE_ADD_BINARY_CHAR (ISO_CODE_CSI, dst);
|
|
|
1719 else
|
|
|
1720 Dynarr_add (dst, ISO_CODE_CSI);
|
|
|
1721 Dynarr_add (dst, '0');
|
|
|
1722 Dynarr_add (dst, ']');
|
|
|
1723 if (flags)
|
|
|
1724 *flags &= ~ISO_STATE_R2L;
|
|
|
1725 }
|
|
|
1726 }
|
|
|
1727
|
|
|
1728 /* If FLAGS is a null pointer or specifies a direction different from
|
|
|
1729 DIRECTION (which should be either CHARSET_RIGHT_TO_LEFT or
|
|
|
1730 CHARSET_LEFT_TO_RIGHT), output the appropriate switch-dir escape
|
|
|
1731 sequence to DST. Also update FLAGS if it is not a null pointer.
|
|
|
1732 If INTERNAL_P is set, we are outputting in internal format and
|
|
|
1733 need to handle the CSI differently. */
|
|
|
1734
|
|
|
1735 static void
|
|
|
1736 ensure_correct_direction (int direction, Lisp_Object codesys,
|
|
|
1737 unsigned_char_dynarr *dst, unsigned int *flags,
|
|
|
1738 int internal_p)
|
|
|
1739 {
|
|
|
1740 if ((!flags || (*flags & ISO_STATE_R2L)) &&
|
|
|
1741 direction == CHARSET_LEFT_TO_RIGHT)
|
|
|
1742 restore_left_to_right_direction (codesys, dst, flags, internal_p);
|
|
|
1743 else if (!XCODING_SYSTEM_ISO2022_NO_ISO6429 (codesys)
|
|
|
1744 && (!flags || !(*flags & ISO_STATE_R2L)) &&
|
|
|
1745 direction == CHARSET_RIGHT_TO_LEFT)
|
|
|
1746 {
|
|
|
1747 if (XCODING_SYSTEM_ISO2022_SEVEN (codesys))
|
|
|
1748 {
|
|
|
1749 Dynarr_add (dst, ISO_CODE_ESC);
|
|
|
1750 Dynarr_add (dst, '[');
|
|
|
1751 }
|
|
|
1752 else if (internal_p)
|
|
|
1753 DECODE_ADD_BINARY_CHAR (ISO_CODE_CSI, dst);
|
|
|
1754 else
|
|
|
1755 Dynarr_add (dst, ISO_CODE_CSI);
|
|
|
1756 Dynarr_add (dst, '2');
|
|
|
1757 Dynarr_add (dst, ']');
|
|
|
1758 if (flags)
|
|
|
1759 *flags |= ISO_STATE_R2L;
|
|
|
1760 }
|
|
|
1761 }
|
|
|
1762
|
|
|
1763 /* Convert ISO2022-format data to internal format. */
|
|
|
1764
|
|
|
1765 static Bytecount
|
|
|
1766 iso2022_decode (struct coding_stream *str, const UExtbyte *src,
|
|
|
1767 unsigned_char_dynarr *dst, Bytecount n)
|
|
|
1768 {
|
|
|
1769 unsigned int ch = str->ch;
|
|
|
1770 #ifdef ENABLE_COMPOSITE_CHARS
|
|
|
1771 unsigned_char_dynarr *real_dst = dst;
|
|
|
1772 #endif
|
|
|
1773 struct iso2022_coding_stream *data =
|
|
|
1774 CODING_STREAM_TYPE_DATA (str, iso2022);
|
|
|
1775 unsigned int flags = data->flags;
|
|
|
1776 Bytecount orign = n;
|
|
|
1777
|
|
|
1778 #ifdef ENABLE_COMPOSITE_CHARS
|
|
|
1779 if (flags & ISO_STATE_COMPOSITE)
|
|
|
1780 dst = data->composite_chars;
|
|
|
1781 #endif /* ENABLE_COMPOSITE_CHARS */
|
|
|
1782
|
|
|
1783 while (n--)
|
|
|
1784 {
|
|
|
1785 UExtbyte c = *src++;
|
|
|
1786 if (flags & ISO_STATE_ESCAPE)
|
|
|
1787 { /* Within ESC sequence */
|
|
|
1788 int retval = parse_iso2022_esc (str->codesys, data,
|
|
|
1789 c, &flags, 1);
|
|
|
1790
|
|
|
1791 if (retval)
|
|
|
1792 {
|
|
|
1793 switch (data->esc)
|
|
|
1794 {
|
|
|
1795 #ifdef ENABLE_COMPOSITE_CHARS
|
|
|
1796 case ISO_ESC_START_COMPOSITE:
|
|
|
1797 if (data->composite_chars)
|
|
|
1798 Dynarr_reset (data->composite_chars);
|
|
|
1799 else
|
|
|
1800 data->composite_chars = Dynarr_new (unsigned_char);
|
|
|
1801 dst = data->composite_chars;
|
|
|
1802 break;
|
|
|
1803 case ISO_ESC_END_COMPOSITE:
|
|
|
1804 {
|
|
867
|
1805 Ibyte comstr[MAX_ICHAR_LEN];
|
|
771
|
1806 Bytecount len;
|
|
867
|
1807 Ichar emch = lookup_composite_char (Dynarr_atp (dst, 0),
|
|
771
|
1808 Dynarr_length (dst));
|
|
|
1809 dst = real_dst;
|
|
867
|
1810 len = set_itext_ichar (comstr, emch);
|
|
771
|
1811 Dynarr_add_many (dst, comstr, len);
|
|
|
1812 break;
|
|
|
1813 }
|
|
|
1814 #else
|
|
|
1815 case ISO_ESC_START_COMPOSITE:
|
|
|
1816 {
|
|
867
|
1817 Ibyte comstr[MAX_ICHAR_LEN];
|
|
771
|
1818 Bytecount len;
|
|
867
|
1819 Ichar emch = make_ichar (Vcharset_composite, c - '0' + ' ',
|
|
771
|
1820 0);
|
|
867
|
1821 len = set_itext_ichar (comstr, emch);
|
|
771
|
1822 Dynarr_add_many (dst, comstr, len);
|
|
|
1823 break;
|
|
|
1824 }
|
|
|
1825 #endif /* ENABLE_COMPOSITE_CHARS */
|
|
|
1826
|
|
|
1827 case ISO_ESC_LITERAL:
|
|
|
1828 DECODE_ADD_BINARY_CHAR (c, dst);
|
|
|
1829 break;
|
|
|
1830
|
|
|
1831 default:
|
|
|
1832 /* Everything else handled already */
|
|
|
1833 break;
|
|
|
1834 }
|
|
|
1835 }
|
|
|
1836
|
|
|
1837 /* Attempted error recovery. */
|
|
|
1838 if (data->output_direction_sequence)
|
|
|
1839 ensure_correct_direction (flags & ISO_STATE_R2L ?
|
|
|
1840 CHARSET_RIGHT_TO_LEFT :
|
|
|
1841 CHARSET_LEFT_TO_RIGHT,
|
|
|
1842 str->codesys, dst, 0, 1);
|
|
|
1843 /* More error recovery. */
|
|
|
1844 if (!retval || data->output_literally)
|
|
|
1845 {
|
|
|
1846 /* Output the (possibly invalid) sequence */
|
|
|
1847 int i;
|
|
|
1848 for (i = 0; i < data->esc_bytes_index; i++)
|
|
|
1849 DECODE_ADD_BINARY_CHAR (data->esc_bytes[i], dst);
|
|
|
1850 flags &= ISO_STATE_LOCK;
|
|
|
1851 if (!retval)
|
|
|
1852 n++, src--;/* Repeat the loop with the same character. */
|
|
|
1853 else
|
|
|
1854 {
|
|
|
1855 /* No sense in reprocessing the final byte of the
|
|
|
1856 escape sequence; it could mess things up anyway.
|
|
|
1857 Just add it now. */
|
|
|
1858 DECODE_ADD_BINARY_CHAR (c, dst);
|
|
|
1859 }
|
|
|
1860 }
|
|
|
1861 ch = 0;
|
|
|
1862 }
|
|
826
|
1863 else if (byte_c0_p (c) || byte_c1_p (c))
|
|
771
|
1864 { /* Control characters */
|
|
|
1865
|
|
|
1866 /***** Error-handling *****/
|
|
|
1867
|
|
|
1868 /* If we were in the middle of a character, dump out the
|
|
|
1869 partial character. */
|
|
|
1870 DECODE_OUTPUT_PARTIAL_CHAR (ch, dst);
|
|
|
1871
|
|
|
1872 /* If we just saw a single-shift character, dump it out.
|
|
|
1873 This may dump out the wrong sort of single-shift character,
|
|
|
1874 but least it will give an indication that something went
|
|
|
1875 wrong. */
|
|
|
1876 if (flags & ISO_STATE_SS2)
|
|
|
1877 {
|
|
|
1878 DECODE_ADD_BINARY_CHAR (ISO_CODE_SS2, dst);
|
|
|
1879 flags &= ~ISO_STATE_SS2;
|
|
|
1880 }
|
|
|
1881 if (flags & ISO_STATE_SS3)
|
|
|
1882 {
|
|
|
1883 DECODE_ADD_BINARY_CHAR (ISO_CODE_SS3, dst);
|
|
|
1884 flags &= ~ISO_STATE_SS3;
|
|
|
1885 }
|
|
|
1886
|
|
|
1887 /***** Now handle the control characters. *****/
|
|
|
1888
|
|
|
1889 flags &= ISO_STATE_LOCK;
|
|
|
1890
|
|
|
1891 if (!parse_iso2022_esc (str->codesys, data, c, &flags, 1))
|
|
|
1892 DECODE_ADD_BINARY_CHAR (c, dst);
|
|
|
1893 }
|
|
|
1894 else
|
|
|
1895 { /* Graphic characters */
|
|
|
1896 Lisp_Object charset;
|
|
|
1897 int lb;
|
|
|
1898 int reg;
|
|
|
1899
|
|
|
1900 /* Now determine the charset. */
|
|
|
1901 reg = ((flags & ISO_STATE_SS2) ? 2
|
|
|
1902 : (flags & ISO_STATE_SS3) ? 3
|
|
826
|
1903 : !byte_ascii_p (c) ? data->register_right
|
|
771
|
1904 : data->register_left);
|
|
|
1905 charset = data->charset[reg];
|
|
|
1906
|
|
|
1907 /* Error checking: */
|
|
|
1908 if (! CHARSETP (charset)
|
|
|
1909 || data->invalid_designated[reg]
|
|
|
1910 || (((c & 0x7F) == ' ' || (c & 0x7F) == ISO_CODE_DEL)
|
|
|
1911 && XCHARSET_CHARS (charset) == 94))
|
|
|
1912 /* Mrmph. We are trying to invoke a register that has no
|
|
|
1913 or an invalid charset in it, or trying to add a character
|
|
|
1914 outside the range of the charset. Insert that char literally
|
|
|
1915 to preserve it for the output. */
|
|
|
1916 {
|
|
|
1917 DECODE_OUTPUT_PARTIAL_CHAR (ch, dst);
|
|
|
1918 DECODE_ADD_BINARY_CHAR (c, dst);
|
|
|
1919 }
|
|
|
1920
|
|
|
1921 else
|
|
|
1922 {
|
|
|
1923 /* Things are probably hunky-dorey. */
|
|
|
1924
|
|
|
1925 /* Fetch reverse charset, maybe. */
|
|
|
1926 if (((flags & ISO_STATE_R2L) &&
|
|
|
1927 XCHARSET_DIRECTION (charset) == CHARSET_LEFT_TO_RIGHT)
|
|
|
1928 ||
|
|
|
1929 (!(flags & ISO_STATE_R2L) &&
|
|
|
1930 XCHARSET_DIRECTION (charset) == CHARSET_RIGHT_TO_LEFT))
|
|
|
1931 {
|
|
|
1932 Lisp_Object new_charset =
|
|
|
1933 XCHARSET_REVERSE_DIRECTION_CHARSET (charset);
|
|
|
1934 if (!NILP (new_charset))
|
|
|
1935 charset = new_charset;
|
|
|
1936 }
|
|
|
1937
|
|
|
1938 lb = XCHARSET_LEADING_BYTE (charset);
|
|
|
1939 switch (XCHARSET_REP_BYTES (charset))
|
|
|
1940 {
|
|
|
1941 case 1: /* ASCII */
|
|
|
1942 DECODE_OUTPUT_PARTIAL_CHAR (ch, dst);
|
|
|
1943 Dynarr_add (dst, c & 0x7F);
|
|
|
1944 break;
|
|
|
1945
|
|
|
1946 case 2: /* one-byte official */
|
|
|
1947 DECODE_OUTPUT_PARTIAL_CHAR (ch, dst);
|
|
|
1948 Dynarr_add (dst, lb);
|
|
|
1949 Dynarr_add (dst, c | 0x80);
|
|
|
1950 break;
|
|
|
1951
|
|
|
1952 case 3: /* one-byte private or two-byte official */
|
|
|
1953 if (XCHARSET_PRIVATE_P (charset))
|
|
|
1954 {
|
|
|
1955 DECODE_OUTPUT_PARTIAL_CHAR (ch, dst);
|
|
|
1956 Dynarr_add (dst, PRE_LEADING_BYTE_PRIVATE_1);
|
|
|
1957 Dynarr_add (dst, lb);
|
|
|
1958 Dynarr_add (dst, c | 0x80);
|
|
|
1959 }
|
|
|
1960 else
|
|
|
1961 {
|
|
|
1962 if (ch)
|
|
|
1963 {
|
|
|
1964 Dynarr_add (dst, lb);
|
|
|
1965 Dynarr_add (dst, ch | 0x80);
|
|
|
1966 Dynarr_add (dst, c | 0x80);
|
|
|
1967 ch = 0;
|
|
|
1968 }
|
|
|
1969 else
|
|
|
1970 ch = c;
|
|
|
1971 }
|
|
|
1972 break;
|
|
|
1973
|
|
|
1974 default: /* two-byte private */
|
|
|
1975 if (ch)
|
|
|
1976 {
|
|
|
1977 Dynarr_add (dst, PRE_LEADING_BYTE_PRIVATE_2);
|
|
|
1978 Dynarr_add (dst, lb);
|
|
|
1979 Dynarr_add (dst, ch | 0x80);
|
|
|
1980 Dynarr_add (dst, c | 0x80);
|
|
|
1981 ch = 0;
|
|
|
1982 }
|
|
|
1983 else
|
|
|
1984 ch = c;
|
|
|
1985 }
|
|
|
1986 }
|
|
|
1987
|
|
|
1988 if (!ch)
|
|
|
1989 flags &= ISO_STATE_LOCK;
|
|
|
1990 }
|
|
|
1991
|
|
|
1992 }
|
|
|
1993
|
|
|
1994 if (str->eof)
|
|
|
1995 DECODE_OUTPUT_PARTIAL_CHAR (ch, dst);
|
|
|
1996
|
|
|
1997 data->flags = flags;
|
|
|
1998 str->ch = ch;
|
|
|
1999 return orign;
|
|
|
2000 }
|
|
|
2001
|
|
|
2002
|
|
|
2003 /***** ISO2022 encoder *****/
|
|
|
2004
|
|
|
2005 /* Designate CHARSET into register REG. */
|
|
|
2006
|
|
|
2007 static void
|
|
|
2008 iso2022_designate (Lisp_Object charset, int reg,
|
|
|
2009 struct coding_stream *str, unsigned_char_dynarr *dst)
|
|
|
2010 {
|
|
|
2011 static const char inter94[] = "()*+";
|
|
|
2012 static const char inter96[] = ",-./";
|
|
|
2013 int type;
|
|
|
2014 unsigned char final;
|
|
|
2015 struct iso2022_coding_stream *data =
|
|
|
2016 CODING_STREAM_TYPE_DATA (str, iso2022);
|
|
|
2017 Lisp_Object old_charset = data->charset[reg];
|
|
|
2018
|
|
|
2019 data->charset[reg] = charset;
|
|
|
2020 if (!CHARSETP (charset))
|
|
|
2021 /* charset might be an initial nil or t. */
|
|
|
2022 return;
|
|
|
2023 type = XCHARSET_TYPE (charset);
|
|
|
2024 final = XCHARSET_FINAL (charset);
|
|
|
2025 if (!data->force_charset_on_output[reg] &&
|
|
|
2026 CHARSETP (old_charset) &&
|
|
|
2027 XCHARSET_TYPE (old_charset) == type &&
|
|
|
2028 XCHARSET_FINAL (old_charset) == final)
|
|
|
2029 return;
|
|
|
2030
|
|
|
2031 data->force_charset_on_output[reg] = 0;
|
|
|
2032
|
|
|
2033 {
|
|
|
2034 charset_conversion_spec_dynarr *dyn =
|
|
|
2035 XCODING_SYSTEM_ISO2022_OUTPUT_CONV (str->codesys);
|
|
|
2036
|
|
|
2037 if (dyn)
|
|
|
2038 {
|
|
|
2039 int i;
|
|
|
2040
|
|
|
2041 for (i = 0; i < Dynarr_length (dyn); i++)
|
|
|
2042 {
|
|
|
2043 struct charset_conversion_spec *spec = Dynarr_atp (dyn, i);
|
|
|
2044 if (EQ (charset, spec->from_charset))
|
|
|
2045 charset = spec->to_charset;
|
|
|
2046 }
|
|
|
2047 }
|
|
|
2048 }
|
|
|
2049
|
|
|
2050 Dynarr_add (dst, ISO_CODE_ESC);
|
|
|
2051 switch (type)
|
|
|
2052 {
|
|
|
2053 case CHARSET_TYPE_94:
|
|
|
2054 Dynarr_add (dst, inter94[reg]);
|
|
|
2055 break;
|
|
|
2056 case CHARSET_TYPE_96:
|
|
|
2057 Dynarr_add (dst, inter96[reg]);
|
|
|
2058 break;
|
|
|
2059 case CHARSET_TYPE_94X94:
|
|
|
2060 Dynarr_add (dst, '$');
|
|
|
2061 if (reg != 0
|
|
|
2062 || !(XCODING_SYSTEM_ISO2022_SHORT (str->codesys))
|
|
|
2063 || final < '@'
|
|
|
2064 || final > 'B')
|
|
|
2065 Dynarr_add (dst, inter94[reg]);
|
|
|
2066 break;
|
|
|
2067 case CHARSET_TYPE_96X96:
|
|
|
2068 Dynarr_add (dst, '$');
|
|
|
2069 Dynarr_add (dst, inter96[reg]);
|
|
|
2070 break;
|
|
|
2071 }
|
|
|
2072 Dynarr_add (dst, final);
|
|
|
2073 }
|
|
|
2074
|
|
|
2075 static void
|
|
|
2076 ensure_normal_shift (struct coding_stream *str, unsigned_char_dynarr *dst)
|
|
|
2077 {
|
|
|
2078 struct iso2022_coding_stream *data =
|
|
|
2079 CODING_STREAM_TYPE_DATA (str, iso2022);
|
|
|
2080
|
|
|
2081 if (data->register_left != 0)
|
|
|
2082 {
|
|
|
2083 Dynarr_add (dst, ISO_CODE_SI);
|
|
|
2084 data->register_left = 0;
|
|
|
2085 }
|
|
|
2086 }
|
|
|
2087
|
|
|
2088 static void
|
|
|
2089 ensure_shift_out (struct coding_stream *str, unsigned_char_dynarr *dst)
|
|
|
2090 {
|
|
|
2091 struct iso2022_coding_stream *data =
|
|
|
2092 CODING_STREAM_TYPE_DATA (str, iso2022);
|
|
|
2093
|
|
|
2094 if (data->register_left != 1)
|
|
|
2095 {
|
|
|
2096 Dynarr_add (dst, ISO_CODE_SO);
|
|
|
2097 data->register_left = 1;
|
|
|
2098 }
|
|
|
2099 }
|
|
|
2100
|
|
|
2101 /* Convert internally-formatted data to ISO2022 format. */
|
|
|
2102
|
|
|
2103 static Bytecount
|
|
867
|
2104 iso2022_encode (struct coding_stream *str, const Ibyte *src,
|
|
771
|
2105 unsigned_char_dynarr *dst, Bytecount n)
|
|
|
2106 {
|
|
|
2107 unsigned char charmask;
|
|
867
|
2108 Ibyte c;
|
|
771
|
2109 unsigned char char_boundary;
|
|
|
2110 unsigned int ch = str->ch;
|
|
|
2111 Lisp_Object codesys = str->codesys;
|
|
|
2112 int i;
|
|
|
2113 Lisp_Object charset;
|
|
|
2114 int half;
|
|
|
2115 struct iso2022_coding_stream *data =
|
|
|
2116 CODING_STREAM_TYPE_DATA (str, iso2022);
|
|
|
2117 unsigned int flags = data->flags;
|
|
|
2118 Bytecount orign = n;
|
|
|
2119
|
|
|
2120 #ifdef ENABLE_COMPOSITE_CHARS
|
|
|
2121 /* flags for handling composite chars. We do a little switcheroo
|
|
|
2122 on the source while we're outputting the composite char. */
|
|
|
2123 Bytecount saved_n = 0;
|
|
867
|
2124 const Ibyte *saved_src = NULL;
|
|
771
|
2125 int in_composite = 0;
|
|
|
2126 #endif /* ENABLE_COMPOSITE_CHARS */
|
|
|
2127
|
|
|
2128 char_boundary = data->current_char_boundary;
|
|
|
2129 charset = data->current_charset;
|
|
|
2130 half = data->current_half;
|
|
|
2131
|
|
|
2132 #ifdef ENABLE_COMPOSITE_CHARS
|
|
|
2133 back_to_square_n:
|
|
|
2134 #endif
|
|
|
2135 while (n--)
|
|
|
2136 {
|
|
|
2137 c = *src++;
|
|
|
2138
|
|
826
|
2139 if (byte_ascii_p (c))
|
|
771
|
2140 { /* Processing ASCII character */
|
|
|
2141 ch = 0;
|
|
|
2142
|
|
|
2143 restore_left_to_right_direction (codesys, dst, &flags, 0);
|
|
|
2144
|
|
|
2145 /* Make sure G0 contains ASCII */
|
|
|
2146 if ((c > ' ' && c < ISO_CODE_DEL) ||
|
|
|
2147 !XCODING_SYSTEM_ISO2022_NO_ASCII_CNTL (codesys))
|
|
|
2148 {
|
|
|
2149 ensure_normal_shift (str, dst);
|
|
|
2150 iso2022_designate (Vcharset_ascii, 0, str, dst);
|
|
|
2151 }
|
|
|
2152
|
|
|
2153 /* If necessary, restore everything to the default state
|
|
|
2154 at end-of-line */
|
|
|
2155 if (!(XCODING_SYSTEM_ISO2022_NO_ASCII_EOL (codesys)))
|
|
|
2156 {
|
|
|
2157 /* NOTE: CRLF encoding happens *BEFORE* other encoding.
|
|
|
2158 Thus, even though we're working with internal-format
|
|
|
2159 data, there may be CR's or CRLF sequences representing
|
|
|
2160 newlines. */
|
|
|
2161 if (c == '\r' || (c == '\n' && !(flags & ISO_STATE_CR)))
|
|
|
2162 {
|
|
|
2163 restore_left_to_right_direction (codesys, dst, &flags, 0);
|
|
|
2164
|
|
|
2165 ensure_normal_shift (str, dst);
|
|
|
2166
|
|
|
2167 for (i = 0; i < 4; i++)
|
|
|
2168 {
|
|
|
2169 Lisp_Object initial_charset =
|
|
|
2170 XCODING_SYSTEM_ISO2022_INITIAL_CHARSET (codesys, i);
|
|
|
2171 iso2022_designate (initial_charset, i, str, dst);
|
|
|
2172 }
|
|
|
2173 }
|
|
|
2174 if (c == '\r')
|
|
|
2175 flags |= ISO_STATE_CR;
|
|
|
2176 else
|
|
|
2177 flags &= ~ISO_STATE_CR;
|
|
|
2178 }
|
|
|
2179
|
|
|
2180 if (XCODING_SYSTEM_ISO2022_ESCAPE_QUOTED (codesys)
|
|
|
2181 && fit_to_be_escape_quoted (c))
|
|
|
2182 Dynarr_add (dst, ISO_CODE_ESC);
|
|
|
2183 Dynarr_add (dst, c);
|
|
|
2184 char_boundary = 1;
|
|
|
2185 }
|
|
|
2186
|
|
867
|
2187 else if (ibyte_leading_byte_p (c) || ibyte_leading_byte_p (ch))
|
|
771
|
2188 { /* Processing Leading Byte */
|
|
|
2189 ch = 0;
|
|
826
|
2190 charset = charset_by_leading_byte (c);
|
|
|
2191 if (leading_byte_prefix_p (c))
|
|
771
|
2192 ch = c;
|
|
|
2193 else if (!EQ (charset, Vcharset_control_1)
|
|
|
2194 && !EQ (charset, Vcharset_composite))
|
|
|
2195 {
|
|
|
2196 int reg;
|
|
|
2197
|
|
|
2198 ensure_correct_direction (XCHARSET_DIRECTION (charset),
|
|
|
2199 codesys, dst, &flags, 0);
|
|
|
2200
|
|
|
2201 /* Now determine which register to use. */
|
|
|
2202 reg = -1;
|
|
|
2203 for (i = 0; i < 4; i++)
|
|
|
2204 {
|
|
|
2205 if (EQ (charset, data->charset[i]) ||
|
|
|
2206 EQ (charset,
|
|
|
2207 XCODING_SYSTEM_ISO2022_INITIAL_CHARSET (codesys, i)))
|
|
|
2208 {
|
|
|
2209 reg = i;
|
|
|
2210 break;
|
|
|
2211 }
|
|
|
2212 }
|
|
|
2213
|
|
|
2214 if (reg == -1)
|
|
|
2215 {
|
|
|
2216 if (XCHARSET_GRAPHIC (charset) != 0)
|
|
|
2217 {
|
|
|
2218 if (!NILP (data->charset[1]) &&
|
|
|
2219 (!XCODING_SYSTEM_ISO2022_SEVEN (codesys) ||
|
|
|
2220 XCODING_SYSTEM_ISO2022_LOCK_SHIFT (codesys)))
|
|
|
2221 reg = 1;
|
|
|
2222 else if (!NILP (data->charset[2]))
|
|
|
2223 reg = 2;
|
|
|
2224 else if (!NILP (data->charset[3]))
|
|
|
2225 reg = 3;
|
|
|
2226 else
|
|
|
2227 reg = 0;
|
|
|
2228 }
|
|
|
2229 else
|
|
|
2230 reg = 0;
|
|
|
2231 }
|
|
|
2232
|
|
|
2233 iso2022_designate (charset, reg, str, dst);
|
|
|
2234
|
|
|
2235 /* Now invoke that register. */
|
|
|
2236 switch (reg)
|
|
|
2237 {
|
|
|
2238 case 0:
|
|
|
2239 ensure_normal_shift (str, dst);
|
|
|
2240 half = 0;
|
|
|
2241 break;
|
|
|
2242
|
|
|
2243 case 1:
|
|
|
2244 if (XCODING_SYSTEM_ISO2022_SEVEN (codesys))
|
|
|
2245 {
|
|
|
2246 ensure_shift_out (str, dst);
|
|
|
2247 half = 0;
|
|
|
2248 }
|
|
|
2249 else
|
|
|
2250 half = 1;
|
|
|
2251 break;
|
|
|
2252
|
|
|
2253 case 2:
|
|
|
2254 if (XCODING_SYSTEM_ISO2022_SEVEN (str->codesys))
|
|
|
2255 {
|
|
|
2256 Dynarr_add (dst, ISO_CODE_ESC);
|
|
|
2257 Dynarr_add (dst, 'N');
|
|
|
2258 half = 0;
|
|
|
2259 }
|
|
|
2260 else
|
|
|
2261 {
|
|
|
2262 Dynarr_add (dst, ISO_CODE_SS2);
|
|
|
2263 half = 1;
|
|
|
2264 }
|
|
|
2265 break;
|
|
|
2266
|
|
|
2267 case 3:
|
|
|
2268 if (XCODING_SYSTEM_ISO2022_SEVEN (str->codesys))
|
|
|
2269 {
|
|
|
2270 Dynarr_add (dst, ISO_CODE_ESC);
|
|
|
2271 Dynarr_add (dst, 'O');
|
|
|
2272 half = 0;
|
|
|
2273 }
|
|
|
2274 else
|
|
|
2275 {
|
|
|
2276 Dynarr_add (dst, ISO_CODE_SS3);
|
|
|
2277 half = 1;
|
|
|
2278 }
|
|
|
2279 break;
|
|
|
2280
|
|
|
2281 default:
|
|
|
2282 abort ();
|
|
|
2283 }
|
|
|
2284 }
|
|
|
2285 char_boundary = 0;
|
|
|
2286 }
|
|
|
2287 else
|
|
|
2288 { /* Processing Non-ASCII character */
|
|
|
2289 charmask = (half == 0 ? 0x7F : 0xFF);
|
|
|
2290 char_boundary = 1;
|
|
|
2291 if (EQ (charset, Vcharset_control_1))
|
|
|
2292 {
|
|
|
2293 if (XCODING_SYSTEM_ISO2022_ESCAPE_QUOTED (codesys)
|
|
|
2294 && fit_to_be_escape_quoted (c))
|
|
|
2295 Dynarr_add (dst, ISO_CODE_ESC);
|
|
|
2296 /* you asked for it ... */
|
|
|
2297 Dynarr_add (dst, c - 0x20);
|
|
|
2298 }
|
|
|
2299 #ifndef ENABLE_COMPOSITE_CHARS
|
|
|
2300 else if (EQ (charset, Vcharset_composite))
|
|
|
2301 {
|
|
|
2302 if (c >= 160 || c <= 164) /* Someone might have stuck in
|
|
|
2303 something else */
|
|
|
2304 {
|
|
|
2305 Dynarr_add (dst, ISO_CODE_ESC);
|
|
|
2306 Dynarr_add (dst, c - 160 + '0');
|
|
|
2307 }
|
|
|
2308 }
|
|
|
2309 #endif
|
|
|
2310 else
|
|
|
2311 {
|
|
|
2312 switch (XCHARSET_REP_BYTES (charset))
|
|
|
2313 {
|
|
|
2314 case 2:
|
|
|
2315 Dynarr_add (dst, c & charmask);
|
|
|
2316 break;
|
|
|
2317 case 3:
|
|
|
2318 if (XCHARSET_PRIVATE_P (charset))
|
|
|
2319 {
|
|
|
2320 Dynarr_add (dst, c & charmask);
|
|
|
2321 ch = 0;
|
|
|
2322 }
|
|
|
2323 else if (ch)
|
|
|
2324 {
|
|
|
2325 #ifdef ENABLE_COMPOSITE_CHARS
|
|
|
2326 if (EQ (charset, Vcharset_composite))
|
|
|
2327 {
|
|
|
2328 if (in_composite)
|
|
|
2329 {
|
|
|
2330 /* #### Bother! We don't know how to
|
|
|
2331 handle this yet. */
|
|
|
2332 Dynarr_add (dst, '~');
|
|
|
2333 }
|
|
|
2334 else
|
|
|
2335 {
|
|
867
|
2336 Ichar emch = make_ichar (Vcharset_composite,
|
|
771
|
2337 ch & 0x7F, c & 0x7F);
|
|
|
2338 Lisp_Object lstr = composite_char_string (emch);
|
|
|
2339 saved_n = n;
|
|
|
2340 saved_src = src;
|
|
|
2341 in_composite = 1;
|
|
|
2342 src = XSTRING_DATA (lstr);
|
|
|
2343 n = XSTRING_LENGTH (lstr);
|
|
|
2344 Dynarr_add (dst, ISO_CODE_ESC);
|
|
|
2345 Dynarr_add (dst, '0'); /* start composing */
|
|
|
2346 }
|
|
|
2347 }
|
|
|
2348 else
|
|
|
2349 #endif /* ENABLE_COMPOSITE_CHARS */
|
|
|
2350 {
|
|
|
2351 Dynarr_add (dst, ch & charmask);
|
|
|
2352 Dynarr_add (dst, c & charmask);
|
|
|
2353 }
|
|
|
2354 ch = 0;
|
|
|
2355 }
|
|
|
2356 else
|
|
|
2357 {
|
|
|
2358 ch = c;
|
|
|
2359 char_boundary = 0;
|
|
|
2360 }
|
|
|
2361 break;
|
|
|
2362 case 4:
|
|
|
2363 if (ch)
|
|
|
2364 {
|
|
|
2365 Dynarr_add (dst, ch & charmask);
|
|
|
2366 Dynarr_add (dst, c & charmask);
|
|
|
2367 ch = 0;
|
|
|
2368 }
|
|
|
2369 else
|
|
|
2370 {
|
|
|
2371 ch = c;
|
|
|
2372 char_boundary = 0;
|
|
|
2373 }
|
|
|
2374 break;
|
|
|
2375 default:
|
|
|
2376 abort ();
|
|
|
2377 }
|
|
|
2378 }
|
|
|
2379 }
|
|
|
2380 }
|
|
|
2381
|
|
|
2382 #ifdef ENABLE_COMPOSITE_CHARS
|
|
|
2383 if (in_composite)
|
|
|
2384 {
|
|
|
2385 n = saved_n;
|
|
|
2386 src = saved_src;
|
|
|
2387 in_composite = 0;
|
|
|
2388 Dynarr_add (dst, ISO_CODE_ESC);
|
|
|
2389 Dynarr_add (dst, '1'); /* end composing */
|
|
|
2390 goto back_to_square_n; /* Wheeeeeeeee ..... */
|
|
|
2391 }
|
|
|
2392 #endif /* ENABLE_COMPOSITE_CHARS */
|
|
|
2393
|
|
|
2394 if (char_boundary && str->eof)
|
|
|
2395 {
|
|
|
2396 restore_left_to_right_direction (codesys, dst, &flags, 0);
|
|
|
2397 ensure_normal_shift (str, dst);
|
|
|
2398 for (i = 0; i < 4; i++)
|
|
|
2399 {
|
|
|
2400 Lisp_Object initial_charset =
|
|
|
2401 XCODING_SYSTEM_ISO2022_INITIAL_CHARSET (codesys, i);
|
|
|
2402 iso2022_designate (initial_charset, i, str, dst);
|
|
|
2403 }
|
|
|
2404 }
|
|
|
2405
|
|
|
2406 data->flags = flags;
|
|
|
2407 str->ch = ch;
|
|
|
2408 data->current_char_boundary = char_boundary;
|
|
|
2409 data->current_charset = charset;
|
|
|
2410 data->current_half = half;
|
|
|
2411
|
|
|
2412 /* Verbum caro factum est! */
|
|
|
2413 return orign;
|
|
|
2414 }
|
|
|
2415
|
|
|
2416 static Bytecount
|
|
|
2417 iso2022_convert (struct coding_stream *str,
|
|
|
2418 const UExtbyte *src,
|
|
|
2419 unsigned_char_dynarr *dst, Bytecount n)
|
|
|
2420 {
|
|
|
2421 if (str->direction == CODING_DECODE)
|
|
|
2422 return iso2022_decode (str, src, dst, n);
|
|
|
2423 else
|
|
|
2424 return iso2022_encode (str, src, dst, n);
|
|
|
2425 }
|
|
|
2426
|
|
|
2427 static void
|
|
|
2428 iso2022_mark (Lisp_Object codesys)
|
|
|
2429 {
|
|
|
2430 int i;
|
|
|
2431
|
|
|
2432 for (i = 0; i < 4; i++)
|
|
|
2433 mark_object (XCODING_SYSTEM_ISO2022_INITIAL_CHARSET (codesys, i));
|
|
|
2434 if (XCODING_SYSTEM_ISO2022_INPUT_CONV (codesys))
|
|
|
2435 {
|
|
|
2436 for (i = 0;
|
|
|
2437 i < Dynarr_length (XCODING_SYSTEM_ISO2022_INPUT_CONV (codesys));
|
|
|
2438 i++)
|
|
|
2439 {
|
|
|
2440 struct charset_conversion_spec *ccs =
|
|
|
2441 Dynarr_atp (XCODING_SYSTEM_ISO2022_INPUT_CONV (codesys), i);
|
|
|
2442 mark_object (ccs->from_charset);
|
|
|
2443 mark_object (ccs->to_charset);
|
|
|
2444 }
|
|
|
2445 }
|
|
|
2446 if (XCODING_SYSTEM_ISO2022_OUTPUT_CONV (codesys))
|
|
|
2447 {
|
|
|
2448 for (i = 0;
|
|
|
2449 i < Dynarr_length (XCODING_SYSTEM_ISO2022_OUTPUT_CONV (codesys));
|
|
|
2450 i++)
|
|
|
2451 {
|
|
|
2452 struct charset_conversion_spec *ccs =
|
|
|
2453 Dynarr_atp (XCODING_SYSTEM_ISO2022_OUTPUT_CONV (codesys), i);
|
|
|
2454 mark_object (ccs->from_charset);
|
|
|
2455 mark_object (ccs->to_charset);
|
|
|
2456 }
|
|
|
2457 }
|
|
|
2458 }
|
|
|
2459
|
|
|
2460 static void
|
|
|
2461 iso2022_finalize (Lisp_Object cs)
|
|
|
2462 {
|
|
|
2463 if (XCODING_SYSTEM_ISO2022_INPUT_CONV (cs))
|
|
|
2464 {
|
|
|
2465 Dynarr_free (XCODING_SYSTEM_ISO2022_INPUT_CONV (cs));
|
|
|
2466 XCODING_SYSTEM_ISO2022_INPUT_CONV (cs) = 0;
|
|
|
2467 }
|
|
|
2468 if (XCODING_SYSTEM_ISO2022_OUTPUT_CONV (cs))
|
|
|
2469 {
|
|
|
2470 Dynarr_free (XCODING_SYSTEM_ISO2022_OUTPUT_CONV (cs));
|
|
|
2471 XCODING_SYSTEM_ISO2022_OUTPUT_CONV (cs) = 0;
|
|
|
2472 }
|
|
|
2473 }
|
|
|
2474
|
|
|
2475 /* Given a list of charset conversion specs as specified in a Lisp
|
|
|
2476 program, parse it into STORE_HERE. */
|
|
|
2477
|
|
|
2478 static void
|
|
|
2479 parse_charset_conversion_specs (charset_conversion_spec_dynarr *store_here,
|
|
|
2480 Lisp_Object spec_list)
|
|
|
2481 {
|
|
|
2482 Lisp_Object rest;
|
|
|
2483
|
|
|
2484 EXTERNAL_LIST_LOOP (rest, spec_list)
|
|
|
2485 {
|
|
|
2486 Lisp_Object car = XCAR (rest);
|
|
|
2487 Lisp_Object from, to;
|
|
|
2488 struct charset_conversion_spec spec;
|
|
|
2489
|
|
|
2490 if (!CONSP (car) || !CONSP (XCDR (car)) || !NILP (XCDR (XCDR (car))))
|
|
|
2491 invalid_argument ("Invalid charset conversion spec", car);
|
|
|
2492 from = Fget_charset (XCAR (car));
|
|
|
2493 to = Fget_charset (XCAR (XCDR (car)));
|
|
|
2494 if (XCHARSET_TYPE (from) != XCHARSET_TYPE (to))
|
|
|
2495 invalid_operation_2
|
|
|
2496 ("Attempted conversion between different charset types",
|
|
|
2497 from, to);
|
|
|
2498 spec.from_charset = from;
|
|
|
2499 spec.to_charset = to;
|
|
|
2500
|
|
|
2501 Dynarr_add (store_here, spec);
|
|
|
2502 }
|
|
|
2503 }
|
|
|
2504
|
|
|
2505 /* Given a dynarr LOAD_HERE of internally-stored charset conversion
|
|
|
2506 specs, return the equivalent as the Lisp programmer would see it.
|
|
|
2507
|
|
|
2508 If LOAD_HERE is 0, return Qnil. */
|
|
|
2509
|
|
|
2510 static Lisp_Object
|
|
|
2511 unparse_charset_conversion_specs (charset_conversion_spec_dynarr *load_here,
|
|
|
2512 int names)
|
|
|
2513 {
|
|
|
2514 int i;
|
|
|
2515 Lisp_Object result;
|
|
|
2516
|
|
|
2517 if (!load_here)
|
|
|
2518 return Qnil;
|
|
|
2519 for (i = 0, result = Qnil; i < Dynarr_length (load_here); i++)
|
|
|
2520 {
|
|
|
2521 struct charset_conversion_spec *ccs = Dynarr_atp (load_here, i);
|
|
|
2522 if (names)
|
|
|
2523 result = Fcons (list2 (XCHARSET_NAME (ccs->from_charset),
|
|
|
2524 XCHARSET_NAME (ccs->to_charset)), result);
|
|
|
2525 else
|
|
|
2526 result = Fcons (list2 (ccs->from_charset, ccs->to_charset), result);
|
|
|
2527 }
|
|
|
2528
|
|
|
2529 return Fnreverse (result);
|
|
|
2530 }
|
|
|
2531
|
|
|
2532 static int
|
|
|
2533 iso2022_putprop (Lisp_Object codesys,
|
|
|
2534 Lisp_Object key,
|
|
|
2535 Lisp_Object value)
|
|
|
2536 {
|
|
|
2537 #define FROB_INITIAL_CHARSET(charset_num) \
|
|
|
2538 XCODING_SYSTEM_ISO2022_INITIAL_CHARSET (codesys, charset_num) = \
|
|
|
2539 ((EQ (value, Qt) || EQ (value, Qnil)) ? value : Fget_charset (value))
|
|
|
2540
|
|
|
2541 if (EQ (key, Qcharset_g0)) FROB_INITIAL_CHARSET (0);
|
|
|
2542 else if (EQ (key, Qcharset_g1)) FROB_INITIAL_CHARSET (1);
|
|
|
2543 else if (EQ (key, Qcharset_g2)) FROB_INITIAL_CHARSET (2);
|
|
|
2544 else if (EQ (key, Qcharset_g3)) FROB_INITIAL_CHARSET (3);
|
|
|
2545
|
|
|
2546 #define FROB_FORCE_CHARSET(charset_num) \
|
|
|
2547 XCODING_SYSTEM_ISO2022_FORCE_CHARSET_ON_OUTPUT (codesys, charset_num) = \
|
|
|
2548 !NILP (value)
|
|
|
2549
|
|
|
2550 else if (EQ (key, Qforce_g0_on_output)) FROB_FORCE_CHARSET (0);
|
|
|
2551 else if (EQ (key, Qforce_g1_on_output)) FROB_FORCE_CHARSET (1);
|
|
|
2552 else if (EQ (key, Qforce_g2_on_output)) FROB_FORCE_CHARSET (2);
|
|
|
2553 else if (EQ (key, Qforce_g3_on_output)) FROB_FORCE_CHARSET (3);
|
|
|
2554
|
|
|
2555 #define FROB_BOOLEAN_PROPERTY(prop) \
|
|
|
2556 XCODING_SYSTEM_ISO2022_##prop (codesys) = !NILP (value)
|
|
|
2557
|
|
|
2558 else if (EQ (key, Qshort)) FROB_BOOLEAN_PROPERTY (SHORT);
|
|
|
2559 else if (EQ (key, Qno_ascii_eol)) FROB_BOOLEAN_PROPERTY (NO_ASCII_EOL);
|
|
|
2560 else if (EQ (key, Qno_ascii_cntl)) FROB_BOOLEAN_PROPERTY (NO_ASCII_CNTL);
|
|
|
2561 else if (EQ (key, Qseven)) FROB_BOOLEAN_PROPERTY (SEVEN);
|
|
|
2562 else if (EQ (key, Qlock_shift)) FROB_BOOLEAN_PROPERTY (LOCK_SHIFT);
|
|
|
2563 else if (EQ (key, Qno_iso6429)) FROB_BOOLEAN_PROPERTY (NO_ISO6429);
|
|
|
2564 else if (EQ (key, Qescape_quoted)) FROB_BOOLEAN_PROPERTY (ESCAPE_QUOTED);
|
|
|
2565
|
|
|
2566 else if (EQ (key, Qinput_charset_conversion))
|
|
|
2567 {
|
|
|
2568 XCODING_SYSTEM_ISO2022_INPUT_CONV (codesys) =
|
|
|
2569 Dynarr_new (charset_conversion_spec);
|
|
|
2570 parse_charset_conversion_specs
|
|
|
2571 (XCODING_SYSTEM_ISO2022_INPUT_CONV (codesys), value);
|
|
|
2572 }
|
|
|
2573 else if (EQ (key, Qoutput_charset_conversion))
|
|
|
2574 {
|
|
|
2575 XCODING_SYSTEM_ISO2022_OUTPUT_CONV (codesys) =
|
|
|
2576 Dynarr_new (charset_conversion_spec);
|
|
|
2577 parse_charset_conversion_specs
|
|
|
2578 (XCODING_SYSTEM_ISO2022_OUTPUT_CONV (codesys), value);
|
|
|
2579 }
|
|
|
2580 else
|
|
|
2581 return 0;
|
|
|
2582
|
|
|
2583 return 1;
|
|
|
2584 }
|
|
|
2585
|
|
|
2586 static void
|
|
|
2587 iso2022_finalize_coding_stream (struct coding_stream *str)
|
|
|
2588 {
|
|
|
2589 #ifdef ENABLE_COMPOSITE_CHARS
|
|
|
2590 struct iso2022_coding_stream *data =
|
|
|
2591 CODING_STREAM_TYPE_DATA (str, iso2022);
|
|
|
2592
|
|
|
2593 if (data->composite_chars)
|
|
|
2594 Dynarr_free (data->composite_chars);
|
|
|
2595 #endif
|
|
|
2596 }
|
|
|
2597
|
|
|
2598 static void
|
|
|
2599 iso2022_init (Lisp_Object codesys)
|
|
|
2600 {
|
|
|
2601 int i;
|
|
|
2602 for (i = 0; i < 4; i++)
|
|
|
2603 XCODING_SYSTEM_ISO2022_INITIAL_CHARSET (codesys, i) = Qnil;
|
|
|
2604 }
|
|
|
2605
|
|
|
2606 static Lisp_Object
|
|
|
2607 coding_system_charset (Lisp_Object coding_system, int gnum)
|
|
|
2608 {
|
|
|
2609 Lisp_Object cs
|
|
|
2610 = XCODING_SYSTEM_ISO2022_INITIAL_CHARSET (coding_system, gnum);
|
|
|
2611
|
|
|
2612 return CHARSETP (cs) ? XCHARSET_NAME (cs) : Qnil;
|
|
|
2613 }
|
|
|
2614
|
|
|
2615 static Lisp_Object
|
|
|
2616 iso2022_getprop (Lisp_Object coding_system, Lisp_Object prop)
|
|
|
2617 {
|
|
|
2618 if (EQ (prop, Qcharset_g0))
|
|
|
2619 return coding_system_charset (coding_system, 0);
|
|
|
2620 else if (EQ (prop, Qcharset_g1))
|
|
|
2621 return coding_system_charset (coding_system, 1);
|
|
|
2622 else if (EQ (prop, Qcharset_g2))
|
|
|
2623 return coding_system_charset (coding_system, 2);
|
|
|
2624 else if (EQ (prop, Qcharset_g3))
|
|
|
2625 return coding_system_charset (coding_system, 3);
|
|
|
2626
|
|
|
2627 #define FORCE_CHARSET(charset_num) \
|
|
|
2628 (XCODING_SYSTEM_ISO2022_FORCE_CHARSET_ON_OUTPUT \
|
|
|
2629 (coding_system, charset_num) ? Qt : Qnil)
|
|
|
2630
|
|
|
2631 else if (EQ (prop, Qforce_g0_on_output))
|
|
|
2632 return FORCE_CHARSET (0);
|
|
|
2633 else if (EQ (prop, Qforce_g1_on_output))
|
|
|
2634 return FORCE_CHARSET (1);
|
|
|
2635 else if (EQ (prop, Qforce_g2_on_output))
|
|
|
2636 return FORCE_CHARSET (2);
|
|
|
2637 else if (EQ (prop, Qforce_g3_on_output))
|
|
|
2638 return FORCE_CHARSET (3);
|
|
|
2639
|
|
|
2640 #define LISP_BOOLEAN(prop) \
|
|
|
2641 (XCODING_SYSTEM_ISO2022_##prop (coding_system) ? Qt : Qnil)
|
|
|
2642
|
|
|
2643 else if (EQ (prop, Qshort)) return LISP_BOOLEAN (SHORT);
|
|
|
2644 else if (EQ (prop, Qno_ascii_eol)) return LISP_BOOLEAN (NO_ASCII_EOL);
|
|
|
2645 else if (EQ (prop, Qno_ascii_cntl)) return LISP_BOOLEAN (NO_ASCII_CNTL);
|
|
|
2646 else if (EQ (prop, Qseven)) return LISP_BOOLEAN (SEVEN);
|
|
|
2647 else if (EQ (prop, Qlock_shift)) return LISP_BOOLEAN (LOCK_SHIFT);
|
|
|
2648 else if (EQ (prop, Qno_iso6429)) return LISP_BOOLEAN (NO_ISO6429);
|
|
|
2649 else if (EQ (prop, Qescape_quoted)) return LISP_BOOLEAN (ESCAPE_QUOTED);
|
|
|
2650
|
|
|
2651 else if (EQ (prop, Qinput_charset_conversion))
|
|
|
2652 return
|
|
|
2653 unparse_charset_conversion_specs
|
|
|
2654 (XCODING_SYSTEM_ISO2022_INPUT_CONV (coding_system), 0);
|
|
|
2655 else if (EQ (prop, Qoutput_charset_conversion))
|
|
|
2656 return
|
|
|
2657 unparse_charset_conversion_specs
|
|
|
2658 (XCODING_SYSTEM_ISO2022_OUTPUT_CONV (coding_system), 0);
|
|
|
2659 else
|
|
|
2660 return Qunbound;
|
|
|
2661 }
|
|
|
2662
|
|
|
2663 static void
|
|
|
2664 iso2022_print (Lisp_Object cs, Lisp_Object printcharfun, int escapeflag)
|
|
|
2665 {
|
|
|
2666 int i;
|
|
|
2667
|
|
826
|
2668 write_c_string (printcharfun, "(");
|
|
771
|
2669 for (i = 0; i < 4; i++)
|
|
|
2670 {
|
|
|
2671 Lisp_Object charset = coding_system_charset (cs, i);
|
|
|
2672 if (i > 0)
|
|
826
|
2673 write_c_string (printcharfun, ", ");
|
|
771
|
2674 write_fmt_string (printcharfun, "g%d=", i);
|
|
800
|
2675 print_internal (CHARSETP (charset) ? XCHARSET_NAME (charset) : charset, printcharfun, 0);
|
|
771
|
2676 if (XCODING_SYSTEM_ISO2022_FORCE_CHARSET_ON_OUTPUT (cs, i))
|
|
826
|
2677 write_c_string (printcharfun, "(force)");
|
|
771
|
2678 }
|
|
|
2679
|
|
800
|
2680 #define FROB(prop) \
|
|
|
2681 if (!NILP (iso2022_getprop (cs, prop))) \
|
|
|
2682 { \
|
|
|
2683 write_fmt_string (printcharfun, ", %s", prop); \
|
|
771
|
2684 }
|
|
|
2685
|
|
|
2686 FROB (Qshort);
|
|
|
2687 FROB (Qno_ascii_eol);
|
|
|
2688 FROB (Qno_ascii_cntl);
|
|
|
2689 FROB (Qseven);
|
|
|
2690 FROB (Qlock_shift);
|
|
|
2691 FROB (Qno_iso6429);
|
|
|
2692 FROB (Qescape_quoted);
|
|
|
2693
|
|
|
2694 {
|
|
|
2695 Lisp_Object val =
|
|
|
2696 unparse_charset_conversion_specs
|
|
|
2697 (XCODING_SYSTEM_ISO2022_INPUT_CONV (cs), 1);
|
|
|
2698 if (!NILP (val))
|
|
|
2699 {
|
|
800
|
2700 write_fmt_string_lisp (printcharfun, ", input-charset-conversion=%s", 1, val);
|
|
771
|
2701 }
|
|
|
2702 val =
|
|
|
2703 unparse_charset_conversion_specs
|
|
|
2704 (XCODING_SYSTEM_ISO2022_OUTPUT_CONV (cs), 1);
|
|
|
2705 if (!NILP (val))
|
|
|
2706 {
|
|
800
|
2707 write_fmt_string_lisp (printcharfun, ", output-charset-conversion=%s", 1, val);
|
|
771
|
2708 }
|
|
826
|
2709 write_c_string (printcharfun, ")");
|
|
771
|
2710 }
|
|
|
2711 }
|
|
|
2712
|
|
|
2713
|
|
|
2714 /************************************************************************/
|
|
|
2715 /* ISO2022 detector */
|
|
|
2716 /************************************************************************/
|
|
|
2717
|
|
|
2718 DEFINE_DETECTOR (iso2022);
|
|
|
2719 /* ISO2022 system using only seven-bit bytes, no locking shift */
|
|
|
2720 DEFINE_DETECTOR_CATEGORY (iso2022, iso_7);
|
|
|
2721 /* ISO2022 system using eight-bit bytes, no locking shift, no single shift,
|
|
|
2722 using designation to switch charsets */
|
|
|
2723 DEFINE_DETECTOR_CATEGORY (iso2022, iso_8_designate);
|
|
|
2724 /* ISO2022 system using eight-bit bytes, no locking shift, no designation
|
|
|
2725 sequences, one-dimension characters in the upper half. */
|
|
|
2726 DEFINE_DETECTOR_CATEGORY (iso2022, iso_8_1);
|
|
|
2727 /* ISO2022 system using eight-bit bytes, no locking shift, no designation
|
|
|
2728 sequences, two-dimension characters in the upper half. */
|
|
|
2729 DEFINE_DETECTOR_CATEGORY (iso2022, iso_8_2);
|
|
|
2730 /* ISO2022 system using locking shift */
|
|
|
2731 DEFINE_DETECTOR_CATEGORY (iso2022, iso_lock_shift);
|
|
|
2732
|
|
|
2733 struct iso2022_detector
|
|
|
2734 {
|
|
|
2735 int initted;
|
|
|
2736 struct iso2022_coding_stream *iso;
|
|
|
2737 unsigned int flags;
|
|
|
2738
|
|
|
2739 /* for keeping temporary track of high-byte groups */
|
|
|
2740 int high_byte_count;
|
|
|
2741 unsigned int saw_single_shift_just_now:1;
|
|
|
2742
|
|
|
2743 /* running state; we set the likelihoods at the end */
|
|
|
2744 unsigned int seen_high_byte:1;
|
|
|
2745 unsigned int seen_single_shift:1;
|
|
|
2746 unsigned int seen_locking_shift:1;
|
|
|
2747 unsigned int seen_designate:1;
|
|
|
2748 unsigned int bad_single_byte_sequences;
|
|
|
2749 unsigned int bad_multibyte_escape_sequences;
|
|
|
2750 unsigned int good_multibyte_escape_sequences;
|
|
|
2751 int even_high_byte_groups;
|
|
|
2752 int odd_high_byte_groups;
|
|
|
2753 };
|
|
|
2754
|
|
|
2755 static void
|
|
|
2756 iso2022_detect (struct detection_state *st, const UExtbyte *src,
|
|
|
2757 Bytecount n)
|
|
|
2758 {
|
|
|
2759 Bytecount orign = n;
|
|
|
2760 struct iso2022_detector *data = DETECTION_STATE_DATA (st, iso2022);
|
|
|
2761
|
|
|
2762 /* #### There are serious deficiencies in the recognition mechanism
|
|
|
2763 here. This needs to be much smarter if it's going to cut it.
|
|
|
2764 The sequence "\xff\x0f" is currently detected as LOCK_SHIFT while
|
|
|
2765 it should be detected as Latin-1.
|
|
|
2766 All the ISO2022 stuff in this file should be synced up with the
|
|
|
2767 code from FSF Emacs-21.0, in which Mule should be more or less stable.
|
|
|
2768 Perhaps we should wait till R2L works in FSF Emacs? */
|
|
|
2769
|
|
|
2770 /* We keep track of running state on our own, and set the categories at the
|
|
|
2771 end; that way we can reflect the correct state each time we finish, but
|
|
|
2772 not get confused by those results the next time around. */
|
|
|
2773
|
|
|
2774 if (!data->initted)
|
|
|
2775 {
|
|
|
2776 xzero (*data);
|
|
|
2777 data->iso = xnew_and_zero (struct iso2022_coding_stream);
|
|
|
2778 reset_iso2022_decode (Qnil, data->iso);
|
|
|
2779 data->initted = 1;
|
|
|
2780 }
|
|
|
2781
|
|
|
2782 while (n--)
|
|
|
2783 {
|
|
|
2784 UExtbyte c = *src++;
|
|
|
2785 if (c >= 0x80)
|
|
|
2786 data->seen_high_byte = 1;
|
|
|
2787 if (c >= 0xA0)
|
|
|
2788 data->high_byte_count++;
|
|
|
2789 else
|
|
|
2790 {
|
|
|
2791 if (data->high_byte_count &&
|
|
|
2792 !data->saw_single_shift_just_now)
|
|
|
2793 {
|
|
|
2794 if (data->high_byte_count & 1)
|
|
|
2795 data->odd_high_byte_groups++;
|
|
|
2796 else
|
|
|
2797 data->even_high_byte_groups++;
|
|
|
2798 }
|
|
|
2799 data->high_byte_count = 0;
|
|
|
2800 data->saw_single_shift_just_now = 0;
|
|
|
2801 }
|
|
|
2802 if (!(data->flags & ISO_STATE_ESCAPE)
|
|
826
|
2803 && (byte_c0_p (c) || byte_c1_p (c)))
|
|
771
|
2804 { /* control chars */
|
|
|
2805 switch (c)
|
|
|
2806 {
|
|
|
2807 /* Allow and ignore control characters that you might
|
|
|
2808 reasonably see in a text file */
|
|
|
2809 case '\r':
|
|
|
2810 case '\n':
|
|
|
2811 case '\t':
|
|
|
2812 case 7: /* bell */
|
|
|
2813 case 8: /* backspace */
|
|
|
2814 case 11: /* vertical tab */
|
|
|
2815 case 12: /* form feed */
|
|
|
2816 case 26: /* MS-DOS C-z junk */
|
|
|
2817 case 31: /* '^_' -- for info */
|
|
|
2818 goto label_continue_loop;
|
|
|
2819
|
|
|
2820 default:
|
|
|
2821 break;
|
|
|
2822 }
|
|
|
2823 }
|
|
|
2824
|
|
826
|
2825 if ((data->flags & ISO_STATE_ESCAPE) || byte_c0_p (c)
|
|
|
2826 || byte_c1_p (c))
|
|
771
|
2827 {
|
|
|
2828 switch (parse_iso2022_esc (Qnil, data->iso, c,
|
|
|
2829 &data->flags, 0))
|
|
|
2830 {
|
|
|
2831 case 1: /* done */
|
|
|
2832 if (data->iso->esc_bytes_index > 0)
|
|
|
2833 data->good_multibyte_escape_sequences++;
|
|
|
2834 switch (data->iso->esc)
|
|
|
2835 {
|
|
|
2836 case ISO_ESC_DESIGNATE:
|
|
|
2837 data->seen_designate = 1;
|
|
|
2838 break;
|
|
|
2839 case ISO_ESC_LOCKING_SHIFT:
|
|
|
2840 data->seen_locking_shift = 1;
|
|
|
2841 break;
|
|
|
2842 case ISO_ESC_SINGLE_SHIFT:
|
|
|
2843 data->saw_single_shift_just_now = 1;
|
|
|
2844 data->seen_single_shift = 1;
|
|
|
2845 break;
|
|
|
2846 default:
|
|
|
2847 break;
|
|
|
2848 }
|
|
|
2849 break;
|
|
|
2850
|
|
|
2851 case -1: /* not done */
|
|
|
2852 break;
|
|
|
2853
|
|
|
2854 case 0: /* error */
|
|
|
2855 if (data->iso->esc == ISO_ESC_NOTHING)
|
|
|
2856 data->bad_single_byte_sequences++;
|
|
|
2857 else
|
|
|
2858 data->bad_multibyte_escape_sequences++;
|
|
|
2859 }
|
|
|
2860 }
|
|
|
2861 label_continue_loop:;
|
|
|
2862 }
|
|
|
2863
|
|
|
2864 if (data->bad_multibyte_escape_sequences > 2 ||
|
|
|
2865 (data->bad_multibyte_escape_sequences > 0 &&
|
|
|
2866 data->good_multibyte_escape_sequences /
|
|
|
2867 data->bad_multibyte_escape_sequences < 10))
|
|
|
2868 /* Just making it up ... */
|
|
|
2869 SET_DET_RESULTS (st, iso2022, DET_NEARLY_IMPOSSIBLE);
|
|
|
2870 else if (data->bad_single_byte_sequences > 5 ||
|
|
|
2871 (data->bad_single_byte_sequences > 0 &&
|
|
|
2872 (data->good_multibyte_escape_sequences +
|
|
|
2873 data->even_high_byte_groups +
|
|
|
2874 data->odd_high_byte_groups) /
|
|
|
2875 data->bad_single_byte_sequences < 10))
|
|
|
2876 SET_DET_RESULTS (st, iso2022, DET_SOMEWHAT_UNLIKELY);
|
|
|
2877 else if (data->seen_locking_shift)
|
|
|
2878 {
|
|
|
2879 SET_DET_RESULTS (st, iso2022, DET_QUITE_IMPROBABLE);
|
|
|
2880 DET_RESULT (st, iso_lock_shift) = DET_QUITE_PROBABLE;
|
|
|
2881 }
|
|
|
2882 else if (!data->seen_high_byte)
|
|
|
2883 {
|
|
|
2884 SET_DET_RESULTS (st, iso2022, DET_SOMEWHAT_UNLIKELY);
|
|
|
2885 if (data->good_multibyte_escape_sequences)
|
|
|
2886 DET_RESULT (st, iso_7) = DET_QUITE_PROBABLE;
|
|
|
2887 else if (data->seen_single_shift)
|
|
|
2888 DET_RESULT (st, iso_7) = DET_SOMEWHAT_LIKELY;
|
|
|
2889 else
|
|
|
2890 {
|
|
|
2891 /* If we've just seen pure 7-bit data, no escape sequences,
|
|
|
2892 then we can't give much likelihood; but if we've seen enough
|
|
|
2893 of this data, we can assume some unlikelihood of any 8-bit
|
|
|
2894 encoding */
|
|
|
2895 if (orign + st->bytes_seen >= 1000)
|
|
|
2896 DET_RESULT (st, iso_7) = DET_AS_LIKELY_AS_UNLIKELY;
|
|
|
2897 else
|
|
|
2898 SET_DET_RESULTS (st, iso2022, DET_AS_LIKELY_AS_UNLIKELY);
|
|
|
2899 }
|
|
|
2900 }
|
|
|
2901 else if (data->seen_designate)
|
|
|
2902 {
|
|
|
2903 SET_DET_RESULTS (st, iso2022, DET_QUITE_IMPROBABLE);
|
|
|
2904 if (data->seen_single_shift)
|
|
|
2905 /* #### Does this really make sense? */
|
|
|
2906 DET_RESULT (st, iso_8_designate) = DET_SOMEWHAT_UNLIKELY;
|
|
|
2907 else
|
|
|
2908 DET_RESULT (st, iso_8_designate) = DET_QUITE_PROBABLE;
|
|
|
2909 }
|
|
|
2910 else if (data->odd_high_byte_groups > 0 &&
|
|
|
2911 data->even_high_byte_groups == 0)
|
|
|
2912 {
|
|
|
2913 SET_DET_RESULTS (st, iso2022, DET_SOMEWHAT_UNLIKELY);
|
|
|
2914 if (data->seen_single_shift)
|
|
|
2915 DET_RESULT (st, iso_8_1) = DET_QUITE_PROBABLE;
|
|
|
2916 else
|
|
|
2917 DET_RESULT (st, iso_8_1) = DET_SOMEWHAT_LIKELY;
|
|
|
2918 }
|
|
|
2919 else if (data->odd_high_byte_groups == 0 &&
|
|
|
2920 data->even_high_byte_groups > 0)
|
|
|
2921 {
|
|
|
2922 SET_DET_RESULTS (st, iso2022, DET_SOMEWHAT_UNLIKELY);
|
|
|
2923 if (data->even_high_byte_groups > 10)
|
|
|
2924 {
|
|
|
2925 if (data->seen_single_shift)
|
|
|
2926 DET_RESULT (st, iso_8_2) = DET_QUITE_PROBABLE;
|
|
|
2927 else
|
|
|
2928 DET_RESULT (st, iso_8_2) = DET_SOMEWHAT_LIKELY;
|
|
|
2929 if (data->even_high_byte_groups < 50)
|
|
|
2930 DET_RESULT (st, iso_8_1) = DET_SOMEWHAT_UNLIKELY;
|
|
|
2931 /* else it stays at quite improbable */
|
|
|
2932 }
|
|
|
2933 }
|
|
|
2934 else if (data->odd_high_byte_groups > 0 &&
|
|
|
2935 data->even_high_byte_groups > 0)
|
|
|
2936 SET_DET_RESULTS (st, iso2022, DET_SOMEWHAT_UNLIKELY);
|
|
|
2937 else
|
|
|
2938 SET_DET_RESULTS (st, iso2022, DET_AS_LIKELY_AS_UNLIKELY);
|
|
|
2939 }
|
|
|
2940
|
|
|
2941 static void
|
|
|
2942 iso2022_finalize_detection_state (struct detection_state *st)
|
|
|
2943 {
|
|
|
2944 struct iso2022_detector *data = DETECTION_STATE_DATA (st, iso2022);
|
|
|
2945 if (data->iso)
|
|
|
2946 xfree (data->iso);
|
|
|
2947 }
|
|
|
2948
|
|
|
2949
|
|
|
2950 /************************************************************************/
|
|
|
2951 /* CCL methods */
|
|
|
2952 /************************************************************************/
|
|
|
2953
|
|
|
2954 /* Converter written in CCL. */
|
|
|
2955 DEFINE_CODING_SYSTEM_TYPE (ccl);
|
|
|
2956
|
|
|
2957 struct ccl_coding_system
|
|
|
2958 {
|
|
|
2959 /* For a CCL coding system, these specify the CCL programs used for
|
|
|
2960 decoding (input) and encoding (output). */
|
|
|
2961 Lisp_Object decode;
|
|
|
2962 Lisp_Object encode;
|
|
|
2963 };
|
|
|
2964
|
|
|
2965 #define CODING_SYSTEM_CCL_DECODE(codesys) \
|
|
|
2966 (CODING_SYSTEM_TYPE_DATA (codesys, ccl)->decode)
|
|
|
2967 #define CODING_SYSTEM_CCL_ENCODE(codesys) \
|
|
|
2968 (CODING_SYSTEM_TYPE_DATA (codesys, ccl)->encode)
|
|
|
2969 #define XCODING_SYSTEM_CCL_DECODE(codesys) \
|
|
|
2970 CODING_SYSTEM_CCL_DECODE (XCODING_SYSTEM (codesys))
|
|
|
2971 #define XCODING_SYSTEM_CCL_ENCODE(codesys) \
|
|
|
2972 CODING_SYSTEM_CCL_ENCODE (XCODING_SYSTEM (codesys))
|
|
|
2973
|
|
|
2974 struct ccl_coding_stream
|
|
|
2975 {
|
|
|
2976 /* state of the running CCL program */
|
|
|
2977 struct ccl_program ccl;
|
|
|
2978 };
|
|
|
2979
|
|
|
2980 static const struct lrecord_description ccl_coding_system_description[] = {
|
|
|
2981 { XD_LISP_OBJECT,
|
|
|
2982 coding_system_data_offset + offsetof (struct ccl_coding_system,
|
|
|
2983 decode) },
|
|
|
2984 { XD_LISP_OBJECT,
|
|
|
2985 coding_system_data_offset + offsetof (struct ccl_coding_system,
|
|
|
2986 encode) },
|
|
|
2987 { XD_END }
|
|
|
2988 };
|
|
|
2989
|
|
|
2990 static void
|
|
|
2991 ccl_mark (Lisp_Object codesys)
|
|
|
2992 {
|
|
|
2993 mark_object (XCODING_SYSTEM_CCL_DECODE (codesys));
|
|
|
2994 mark_object (XCODING_SYSTEM_CCL_ENCODE (codesys));
|
|
|
2995 }
|
|
|
2996
|
|
|
2997 static Bytecount
|
|
|
2998 ccl_convert (struct coding_stream *str, const UExtbyte *src,
|
|
|
2999 unsigned_char_dynarr *dst, Bytecount n)
|
|
|
3000 {
|
|
|
3001 struct ccl_coding_stream *data =
|
|
|
3002 CODING_STREAM_TYPE_DATA (str, ccl);
|
|
|
3003 Bytecount orign = n;
|
|
|
3004
|
|
|
3005 data->ccl.last_block = str->eof;
|
|
|
3006 /* When applying a CCL program to a stream, SRC must not be NULL -- this
|
|
|
3007 is a special signal to the driver that read and write operations are
|
|
|
3008 not allowed. The code does not actually look at what SRC points to if
|
|
|
3009 N == 0.
|
|
|
3010 */
|
|
|
3011 ccl_driver (&data->ccl, src ? src : (const unsigned char *) "",
|
|
|
3012 dst, n, 0,
|
|
|
3013 str->direction == CODING_DECODE ? CCL_MODE_DECODING :
|
|
|
3014 CCL_MODE_ENCODING);
|
|
|
3015 return orign;
|
|
|
3016 }
|
|
|
3017
|
|
|
3018 static void
|
|
|
3019 ccl_init_coding_stream (struct coding_stream *str)
|
|
|
3020 {
|
|
|
3021 struct ccl_coding_stream *data =
|
|
|
3022 CODING_STREAM_TYPE_DATA (str, ccl);
|
|
|
3023
|
|
|
3024 setup_ccl_program (&data->ccl,
|
|
|
3025 str->direction == CODING_DECODE ?
|
|
|
3026 XCODING_SYSTEM_CCL_DECODE (str->codesys) :
|
|
|
3027 XCODING_SYSTEM_CCL_ENCODE (str->codesys));
|
|
|
3028 }
|
|
|
3029
|
|
|
3030 static void
|
|
|
3031 ccl_rewind_coding_stream (struct coding_stream *str)
|
|
|
3032 {
|
|
|
3033 ccl_init_coding_stream (str);
|
|
|
3034 }
|
|
|
3035
|
|
|
3036 static void
|
|
|
3037 ccl_init (Lisp_Object codesys)
|
|
|
3038 {
|
|
|
3039 XCODING_SYSTEM_CCL_DECODE (codesys) = Qnil;
|
|
|
3040 XCODING_SYSTEM_CCL_ENCODE (codesys) = Qnil;
|
|
|
3041 }
|
|
|
3042
|
|
|
3043 static int
|
|
|
3044 ccl_putprop (Lisp_Object codesys, Lisp_Object key, Lisp_Object value)
|
|
|
3045 {
|
|
|
3046 Lisp_Object sym;
|
|
|
3047 struct ccl_program test_ccl;
|
|
|
3048 Char_ASCII *suffix;
|
|
|
3049
|
|
|
3050 /* Check key first. */
|
|
|
3051 if (EQ (key, Qdecode))
|
|
|
3052 suffix = "-ccl-decode";
|
|
|
3053 else if (EQ (key, Qencode))
|
|
|
3054 suffix = "-ccl-encode";
|
|
|
3055 else
|
|
|
3056 return 0;
|
|
|
3057
|
|
|
3058 /* If value is vector, register it as a ccl program
|
|
|
3059 associated with a newly created symbol for
|
|
|
3060 backward compatibility.
|
|
|
3061
|
|
|
3062 #### Bogosity alert! Do we really have to do this crap???? --ben */
|
|
|
3063 if (VECTORP (value))
|
|
|
3064 {
|
|
|
3065 sym = Fintern (concat2 (Fsymbol_name (XCODING_SYSTEM_NAME (codesys)),
|
|
|
3066 build_string (suffix)),
|
|
|
3067 Qnil);
|
|
|
3068 Fregister_ccl_program (sym, value);
|
|
|
3069 }
|
|
|
3070 else
|
|
|
3071 {
|
|
|
3072 CHECK_SYMBOL (value);
|
|
|
3073 sym = value;
|
|
|
3074 }
|
|
|
3075 /* check if the given ccl programs are valid. */
|
|
|
3076 if (setup_ccl_program (&test_ccl, sym) < 0)
|
|
|
3077 invalid_argument ("Invalid CCL program", value);
|
|
|
3078
|
|
|
3079 if (EQ (key, Qdecode))
|
|
|
3080 XCODING_SYSTEM_CCL_DECODE (codesys) = sym;
|
|
|
3081 else if (EQ (key, Qencode))
|
|
|
3082 XCODING_SYSTEM_CCL_ENCODE (codesys) = sym;
|
|
|
3083
|
|
|
3084 return 1;
|
|
|
3085 }
|
|
|
3086
|
|
|
3087 static Lisp_Object
|
|
|
3088 ccl_getprop (Lisp_Object coding_system, Lisp_Object prop)
|
|
|
3089 {
|
|
|
3090 if (EQ (prop, Qdecode))
|
|
|
3091 return XCODING_SYSTEM_CCL_DECODE (coding_system);
|
|
|
3092 else if (EQ (prop, Qencode))
|
|
|
3093 return XCODING_SYSTEM_CCL_ENCODE (coding_system);
|
|
|
3094 else
|
|
|
3095 return Qunbound;
|
|
|
3096 }
|
|
|
3097
|
|
|
3098
|
|
|
3099 /************************************************************************/
|
|
|
3100 /* Initialization */
|
|
|
3101 /************************************************************************/
|
|
|
3102
|
|
|
3103 void
|
|
|
3104 syms_of_mule_coding (void)
|
|
|
3105 {
|
|
|
3106 DEFSUBR (Fdecode_shift_jis_char);
|
|
|
3107 DEFSUBR (Fencode_shift_jis_char);
|
|
|
3108 DEFSUBR (Fdecode_big5_char);
|
|
|
3109 DEFSUBR (Fencode_big5_char);
|
|
|
3110
|
|
|
3111 DEFSYMBOL (Qbig5);
|
|
|
3112 DEFSYMBOL (Qshift_jis);
|
|
|
3113 DEFSYMBOL (Qccl);
|
|
|
3114 DEFSYMBOL (Qiso2022);
|
|
|
3115
|
|
|
3116 DEFSYMBOL (Qcharset_g0);
|
|
|
3117 DEFSYMBOL (Qcharset_g1);
|
|
|
3118 DEFSYMBOL (Qcharset_g2);
|
|
|
3119 DEFSYMBOL (Qcharset_g3);
|
|
|
3120 DEFSYMBOL (Qforce_g0_on_output);
|
|
|
3121 DEFSYMBOL (Qforce_g1_on_output);
|
|
|
3122 DEFSYMBOL (Qforce_g2_on_output);
|
|
|
3123 DEFSYMBOL (Qforce_g3_on_output);
|
|
|
3124 DEFSYMBOL (Qno_iso6429);
|
|
|
3125 DEFSYMBOL (Qinput_charset_conversion);
|
|
|
3126 DEFSYMBOL (Qoutput_charset_conversion);
|
|
|
3127
|
|
|
3128 DEFSYMBOL (Qshort);
|
|
|
3129 DEFSYMBOL (Qno_ascii_eol);
|
|
|
3130 DEFSYMBOL (Qno_ascii_cntl);
|
|
|
3131 DEFSYMBOL (Qseven);
|
|
|
3132 DEFSYMBOL (Qlock_shift);
|
|
|
3133
|
|
|
3134 DEFSYMBOL (Qiso_7);
|
|
|
3135 DEFSYMBOL (Qiso_8_designate);
|
|
|
3136 DEFSYMBOL (Qiso_8_1);
|
|
|
3137 DEFSYMBOL (Qiso_8_2);
|
|
|
3138 DEFSYMBOL (Qiso_lock_shift);
|
|
|
3139 }
|
|
|
3140
|
|
|
3141 void
|
|
|
3142 coding_system_type_create_mule_coding (void)
|
|
|
3143 {
|
|
|
3144 INITIALIZE_CODING_SYSTEM_TYPE_WITH_DATA (iso2022, "iso2022-coding-system-p");
|
|
|
3145 CODING_SYSTEM_HAS_METHOD (iso2022, mark);
|
|
|
3146 CODING_SYSTEM_HAS_METHOD (iso2022, convert);
|
|
|
3147 CODING_SYSTEM_HAS_METHOD (iso2022, finalize_coding_stream);
|
|
|
3148 CODING_SYSTEM_HAS_METHOD (iso2022, init_coding_stream);
|
|
|
3149 CODING_SYSTEM_HAS_METHOD (iso2022, rewind_coding_stream);
|
|
|
3150 CODING_SYSTEM_HAS_METHOD (iso2022, init);
|
|
|
3151 CODING_SYSTEM_HAS_METHOD (iso2022, print);
|
|
|
3152 CODING_SYSTEM_HAS_METHOD (iso2022, finalize);
|
|
|
3153 CODING_SYSTEM_HAS_METHOD (iso2022, putprop);
|
|
|
3154 CODING_SYSTEM_HAS_METHOD (iso2022, getprop);
|
|
|
3155
|
|
|
3156 INITIALIZE_DETECTOR (iso2022);
|
|
|
3157 DETECTOR_HAS_METHOD (iso2022, detect);
|
|
|
3158 DETECTOR_HAS_METHOD (iso2022, finalize_detection_state);
|
|
|
3159 INITIALIZE_DETECTOR_CATEGORY (iso2022, iso_7);
|
|
|
3160 INITIALIZE_DETECTOR_CATEGORY (iso2022, iso_8_designate);
|
|
|
3161 INITIALIZE_DETECTOR_CATEGORY (iso2022, iso_8_1);
|
|
|
3162 INITIALIZE_DETECTOR_CATEGORY (iso2022, iso_8_2);
|
|
|
3163 INITIALIZE_DETECTOR_CATEGORY (iso2022, iso_lock_shift);
|
|
|
3164
|
|
|
3165 INITIALIZE_CODING_SYSTEM_TYPE_WITH_DATA (ccl, "ccl-coding-system-p");
|
|
|
3166 CODING_SYSTEM_HAS_METHOD (ccl, mark);
|
|
|
3167 CODING_SYSTEM_HAS_METHOD (ccl, convert);
|
|
|
3168 CODING_SYSTEM_HAS_METHOD (ccl, init);
|
|
|
3169 CODING_SYSTEM_HAS_METHOD (ccl, init_coding_stream);
|
|
|
3170 CODING_SYSTEM_HAS_METHOD (ccl, rewind_coding_stream);
|
|
|
3171 CODING_SYSTEM_HAS_METHOD (ccl, putprop);
|
|
|
3172 CODING_SYSTEM_HAS_METHOD (ccl, getprop);
|
|
|
3173
|
|
|
3174 INITIALIZE_CODING_SYSTEM_TYPE (shift_jis, "shift-jis-coding-system-p");
|
|
|
3175 CODING_SYSTEM_HAS_METHOD (shift_jis, convert);
|
|
|
3176
|
|
|
3177 INITIALIZE_DETECTOR (shift_jis);
|
|
|
3178 DETECTOR_HAS_METHOD (shift_jis, detect);
|
|
|
3179 INITIALIZE_DETECTOR_CATEGORY (shift_jis, shift_jis);
|
|
|
3180
|
|
|
3181 INITIALIZE_CODING_SYSTEM_TYPE (big5, "big5-coding-system-p");
|
|
|
3182 CODING_SYSTEM_HAS_METHOD (big5, convert);
|
|
|
3183
|
|
|
3184 INITIALIZE_DETECTOR (big5);
|
|
|
3185 DETECTOR_HAS_METHOD (big5, detect);
|
|
|
3186 INITIALIZE_DETECTOR_CATEGORY (big5, big5);
|
|
|
3187 }
|
|
|
3188
|
|
|
3189 void
|
|
|
3190 reinit_coding_system_type_create_mule_coding (void)
|
|
|
3191 {
|
|
|
3192 REINITIALIZE_CODING_SYSTEM_TYPE (iso2022);
|
|
|
3193 REINITIALIZE_CODING_SYSTEM_TYPE (ccl);
|
|
|
3194 REINITIALIZE_CODING_SYSTEM_TYPE (shift_jis);
|
|
|
3195 REINITIALIZE_CODING_SYSTEM_TYPE (big5);
|
|
|
3196 }
|
|
|
3197
|
|
|
3198 void
|
|
|
3199 reinit_vars_of_mule_coding (void)
|
|
|
3200 {
|
|
|
3201 }
|
|
|
3202
|
|
|
3203 void
|
|
|
3204 vars_of_mule_coding (void)
|
|
|
3205 {
|
|
|
3206 }
|
