|
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;
|
|
985
|
602 int seen_euc_char;
|
|
771
|
603 unsigned int seen_iso2022_esc:1;
|
|
|
604 unsigned int seen_bad_first_byte:1;
|
|
|
605 unsigned int seen_bad_second_byte:1;
|
|
|
606
|
|
|
607 /* temporary */
|
|
|
608 unsigned int in_second_byte:1;
|
|
|
609 };
|
|
|
610
|
|
|
611 static void
|
|
|
612 big5_detect (struct detection_state *st, const UExtbyte *src,
|
|
|
613 Bytecount n)
|
|
|
614 {
|
|
|
615 struct big5_detector *data = DETECTION_STATE_DATA (st, big5);
|
|
|
616
|
|
|
617 while (n--)
|
|
|
618 {
|
|
|
619 UExtbyte c = *src++;
|
|
|
620 if (!data->in_second_byte)
|
|
|
621 {
|
|
|
622 if (c >= 0xA1 && c <= 0xFE)
|
|
|
623 data->in_second_byte = 1;
|
|
|
624 else if (c == ISO_CODE_ESC || c == ISO_CODE_SI || c == ISO_CODE_SO)
|
|
|
625 data->seen_iso2022_esc = 1;
|
|
|
626 else if (c >= 0x80)
|
|
|
627 data->seen_bad_first_byte = 1;
|
|
|
628 }
|
|
|
629 else
|
|
|
630 {
|
|
|
631 data->in_second_byte = 0;
|
|
985
|
632 if (c >= 0xA1 && c <= 0xFE)
|
|
|
633 data->seen_euc_char++;
|
|
|
634 else if (c >= 0x40 && c <= 0x7E)
|
|
771
|
635 data->seen_big5_char++;
|
|
|
636 else
|
|
|
637 data->seen_bad_second_byte = 1;
|
|
|
638 }
|
|
|
639 }
|
|
|
640
|
|
|
641 if (data->seen_bad_second_byte)
|
|
|
642 DET_RESULT (st, big5) = DET_NEARLY_IMPOSSIBLE;
|
|
|
643 else if (data->seen_bad_first_byte)
|
|
|
644 DET_RESULT (st, big5) = DET_QUITE_IMPROBABLE;
|
|
|
645 else if (data->seen_iso2022_esc)
|
|
|
646 DET_RESULT (st, big5) = DET_SOMEWHAT_UNLIKELY;
|
|
|
647 else if (data->seen_big5_char >= 4)
|
|
|
648 DET_RESULT (st, big5) = DET_SOMEWHAT_LIKELY;
|
|
985
|
649 else if (data->seen_euc_char)
|
|
|
650 DET_RESULT (st, big5) = DET_SLIGHTLY_LIKELY;
|
|
771
|
651 else
|
|
|
652 DET_RESULT (st, big5) = DET_AS_LIKELY_AS_UNLIKELY;
|
|
|
653 }
|
|
|
654
|
|
|
655
|
|
|
656 /************************************************************************/
|
|
|
657 /* ISO2022 methods */
|
|
|
658 /************************************************************************/
|
|
|
659
|
|
|
660 /* Any ISO-2022-compliant coding system. Includes JIS, EUC, CTEXT
|
|
|
661 (Compound Text, the encoding of selections in X Windows). See below for
|
|
|
662 a complete description of ISO-2022. */
|
|
|
663
|
|
|
664 /* Flags indicating what we've seen so far when parsing an
|
|
|
665 ISO2022 escape sequence. */
|
|
|
666 enum iso_esc_flag
|
|
|
667 {
|
|
|
668 /* Partial sequences */
|
|
|
669 ISO_ESC_NOTHING, /* Nothing has been seen. */
|
|
|
670 ISO_ESC, /* We've seen ESC. */
|
|
|
671 ISO_ESC_2_4, /* We've seen ESC $. This indicates
|
|
|
672 that we're designating a multi-byte, rather
|
|
|
673 than a single-byte, character set. */
|
|
|
674 ISO_ESC_2_8, /* We've seen ESC 0x28, i.e. ESC (.
|
|
|
675 This means designate a 94-character
|
|
|
676 character set into G0. */
|
|
|
677 ISO_ESC_2_9, /* We've seen ESC 0x29 -- designate a
|
|
|
678 94-character character set into G1. */
|
|
|
679 ISO_ESC_2_10, /* We've seen ESC 0x2A. */
|
|
|
680 ISO_ESC_2_11, /* We've seen ESC 0x2B. */
|
|
|
681 ISO_ESC_2_12, /* We've seen ESC 0x2C -- designate a
|
|
|
682 96-character character set into G0.
|
|
|
683 (This is not ISO2022-standard.
|
|
|
684 The following 96-character
|
|
|
685 control sequences are standard,
|
|
|
686 though.) */
|
|
|
687 ISO_ESC_2_13, /* We've seen ESC 0x2D -- designate a
|
|
|
688 96-character character set into G1.
|
|
|
689 */
|
|
|
690 ISO_ESC_2_14, /* We've seen ESC 0x2E. */
|
|
|
691 ISO_ESC_2_15, /* We've seen ESC 0x2F. */
|
|
|
692 ISO_ESC_2_4_8, /* We've seen ESC $ 0x28 -- designate
|
|
|
693 a 94^N character set into G0. */
|
|
|
694 ISO_ESC_2_4_9, /* We've seen ESC $ 0x29. */
|
|
|
695 ISO_ESC_2_4_10, /* We've seen ESC $ 0x2A. */
|
|
|
696 ISO_ESC_2_4_11, /* We've seen ESC $ 0x2B. */
|
|
|
697 ISO_ESC_2_4_12, /* We've seen ESC $ 0x2C. */
|
|
|
698 ISO_ESC_2_4_13, /* We've seen ESC $ 0x2D. */
|
|
|
699 ISO_ESC_2_4_14, /* We've seen ESC $ 0x2E. */
|
|
|
700 ISO_ESC_2_4_15, /* We've seen ESC $ 0x2F. */
|
|
|
701 ISO_ESC_5_11, /* We've seen ESC [ or 0x9B. This
|
|
|
702 starts a directionality-control
|
|
|
703 sequence. The next character
|
|
|
704 must be 0, 1, 2, or ]. */
|
|
|
705 ISO_ESC_5_11_0, /* We've seen 0x9B 0. The next character must be ]. */
|
|
|
706 ISO_ESC_5_11_1, /* We've seen 0x9B 1. The next character must be ]. */
|
|
|
707 ISO_ESC_5_11_2, /* We've seen 0x9B 2. The next character must be ]. */
|
|
|
708
|
|
|
709 /* Full sequences. */
|
|
|
710 ISO_ESC_START_COMPOSITE, /* Private usage for START COMPOSING */
|
|
|
711 ISO_ESC_END_COMPOSITE, /* Private usage for END COMPOSING */
|
|
|
712 ISO_ESC_SINGLE_SHIFT, /* We've seen a complete single-shift sequence. */
|
|
|
713 ISO_ESC_LOCKING_SHIFT,/* We've seen a complete locking-shift sequence. */
|
|
|
714 ISO_ESC_DESIGNATE, /* We've seen a complete designation sequence. */
|
|
|
715 ISO_ESC_DIRECTIONALITY,/* We've seen a complete ISO6429 directionality
|
|
|
716 sequence. */
|
|
|
717 ISO_ESC_LITERAL /* We've seen a literal character ala
|
|
|
718 escape-quoting. */
|
|
|
719 };
|
|
|
720
|
|
|
721 enum iso_error
|
|
|
722 {
|
|
|
723 ISO_ERROR_BAD_FINAL,
|
|
|
724 ISO_ERROR_UNKWOWN_ESC_SEQUENCE,
|
|
|
725 ISO_ERROR_INVALID_CODE_POINT_CHARACTER,
|
|
|
726 };
|
|
|
727
|
|
|
728
|
|
|
729 /* Flags indicating current state while converting code. */
|
|
|
730
|
|
|
731 /************ Used during encoding and decoding: ************/
|
|
|
732 /* If set, the current directionality is right-to-left. Otherwise, it's
|
|
|
733 left-to-right. */
|
|
|
734 #define ISO_STATE_R2L (1 << 0)
|
|
|
735
|
|
|
736 /************ Used during encoding: ************/
|
|
|
737 /* If set, we just saw a CR. */
|
|
|
738 #define ISO_STATE_CR (1 << 1)
|
|
|
739
|
|
|
740 /************ Used during decoding: ************/
|
|
|
741 /* If set, we're currently parsing an escape sequence and the upper 16 bits
|
|
|
742 should be looked at to indicate what partial escape sequence we've seen
|
|
|
743 so far. Otherwise, we're running through actual text. */
|
|
|
744 #define ISO_STATE_ESCAPE (1 << 2)
|
|
|
745 /* If set, G2 is invoked into GL, but only for the next character. */
|
|
|
746 #define ISO_STATE_SS2 (1 << 3)
|
|
|
747 /* If set, G3 is invoked into GL, but only for the next character. If both
|
|
|
748 ISO_STATE_SS2 and ISO_STATE_SS3 are set, ISO_STATE_SS2 overrides; but
|
|
|
749 this probably indicates an error in the text encoding. */
|
|
|
750 #define ISO_STATE_SS3 (1 << 4)
|
|
|
751 /* If set, we're currently processing a composite character (i.e. a
|
|
|
752 character constructed by overstriking two or more characters). */
|
|
|
753 #define ISO_STATE_COMPOSITE (1 << 5)
|
|
|
754
|
|
|
755 /* ISO_STATE_LOCK is the mask of flags that remain on until explicitly
|
|
|
756 turned off when in the ISO2022 encoder/decoder. Other flags are turned
|
|
|
757 off at the end of processing each character or escape sequence. */
|
|
|
758 # define ISO_STATE_LOCK \
|
|
|
759 (ISO_STATE_COMPOSITE | ISO_STATE_R2L)
|
|
|
760
|
|
|
761 typedef struct charset_conversion_spec
|
|
|
762 {
|
|
|
763 Lisp_Object from_charset;
|
|
|
764 Lisp_Object to_charset;
|
|
|
765 } charset_conversion_spec;
|
|
|
766
|
|
|
767 typedef struct
|
|
|
768 {
|
|
|
769 Dynarr_declare (charset_conversion_spec);
|
|
|
770 } charset_conversion_spec_dynarr;
|
|
|
771
|
|
|
772 struct iso2022_coding_system
|
|
|
773 {
|
|
|
774 /* What are the charsets to be initially designated to G0, G1,
|
|
|
775 G2, G3? If t, no charset is initially designated. If nil,
|
|
|
776 no charset is initially designated and no charset is allowed
|
|
|
777 to be designated. */
|
|
|
778 Lisp_Object initial_charset[4];
|
|
|
779
|
|
|
780 /* If true, a designation escape sequence needs to be sent on output
|
|
|
781 for the charset in G[0-3] before that charset is used. */
|
|
|
782 unsigned char force_charset_on_output[4];
|
|
|
783
|
|
|
784 charset_conversion_spec_dynarr *input_conv;
|
|
|
785 charset_conversion_spec_dynarr *output_conv;
|
|
|
786
|
|
|
787 unsigned int shoort :1; /* C makes you speak Dutch */
|
|
|
788 unsigned int no_ascii_eol :1;
|
|
|
789 unsigned int no_ascii_cntl :1;
|
|
|
790 unsigned int seven :1;
|
|
|
791 unsigned int lock_shift :1;
|
|
|
792 unsigned int no_iso6429 :1;
|
|
|
793 unsigned int escape_quoted :1;
|
|
|
794 };
|
|
|
795
|
|
|
796 #define CODING_SYSTEM_ISO2022_INITIAL_CHARSET(codesys, g) \
|
|
|
797 (CODING_SYSTEM_TYPE_DATA (codesys, iso2022)->initial_charset[g])
|
|
|
798 #define CODING_SYSTEM_ISO2022_FORCE_CHARSET_ON_OUTPUT(codesys, g) \
|
|
|
799 (CODING_SYSTEM_TYPE_DATA (codesys, iso2022)->force_charset_on_output[g])
|
|
|
800 #define CODING_SYSTEM_ISO2022_SHORT(codesys) \
|
|
|
801 (CODING_SYSTEM_TYPE_DATA (codesys, iso2022)->shoort)
|
|
|
802 #define CODING_SYSTEM_ISO2022_NO_ASCII_EOL(codesys) \
|
|
|
803 (CODING_SYSTEM_TYPE_DATA (codesys, iso2022)->no_ascii_eol)
|
|
|
804 #define CODING_SYSTEM_ISO2022_NO_ASCII_CNTL(codesys) \
|
|
|
805 (CODING_SYSTEM_TYPE_DATA (codesys, iso2022)->no_ascii_cntl)
|
|
|
806 #define CODING_SYSTEM_ISO2022_SEVEN(codesys) \
|
|
|
807 (CODING_SYSTEM_TYPE_DATA (codesys, iso2022)->seven)
|
|
|
808 #define CODING_SYSTEM_ISO2022_LOCK_SHIFT(codesys) \
|
|
|
809 (CODING_SYSTEM_TYPE_DATA (codesys, iso2022)->lock_shift)
|
|
|
810 #define CODING_SYSTEM_ISO2022_NO_ISO6429(codesys) \
|
|
|
811 (CODING_SYSTEM_TYPE_DATA (codesys, iso2022)->no_iso6429)
|
|
|
812 #define CODING_SYSTEM_ISO2022_ESCAPE_QUOTED(codesys) \
|
|
|
813 (CODING_SYSTEM_TYPE_DATA (codesys, iso2022)->escape_quoted)
|
|
|
814 #define CODING_SYSTEM_ISO2022_INPUT_CONV(codesys) \
|
|
|
815 (CODING_SYSTEM_TYPE_DATA (codesys, iso2022)->input_conv)
|
|
|
816 #define CODING_SYSTEM_ISO2022_OUTPUT_CONV(codesys) \
|
|
|
817 (CODING_SYSTEM_TYPE_DATA (codesys, iso2022)->output_conv)
|
|
|
818
|
|
|
819 #define XCODING_SYSTEM_ISO2022_INITIAL_CHARSET(codesys, g) \
|
|
|
820 CODING_SYSTEM_ISO2022_INITIAL_CHARSET (XCODING_SYSTEM (codesys), g)
|
|
|
821 #define XCODING_SYSTEM_ISO2022_FORCE_CHARSET_ON_OUTPUT(codesys, g) \
|
|
|
822 CODING_SYSTEM_ISO2022_FORCE_CHARSET_ON_OUTPUT (XCODING_SYSTEM (codesys), g)
|
|
|
823 #define XCODING_SYSTEM_ISO2022_SHORT(codesys) \
|
|
|
824 CODING_SYSTEM_ISO2022_SHORT (XCODING_SYSTEM (codesys))
|
|
|
825 #define XCODING_SYSTEM_ISO2022_NO_ASCII_EOL(codesys) \
|
|
|
826 CODING_SYSTEM_ISO2022_NO_ASCII_EOL (XCODING_SYSTEM (codesys))
|
|
|
827 #define XCODING_SYSTEM_ISO2022_NO_ASCII_CNTL(codesys) \
|
|
|
828 CODING_SYSTEM_ISO2022_NO_ASCII_CNTL (XCODING_SYSTEM (codesys))
|
|
|
829 #define XCODING_SYSTEM_ISO2022_SEVEN(codesys) \
|
|
|
830 CODING_SYSTEM_ISO2022_SEVEN (XCODING_SYSTEM (codesys))
|
|
|
831 #define XCODING_SYSTEM_ISO2022_LOCK_SHIFT(codesys) \
|
|
|
832 CODING_SYSTEM_ISO2022_LOCK_SHIFT (XCODING_SYSTEM (codesys))
|
|
|
833 #define XCODING_SYSTEM_ISO2022_NO_ISO6429(codesys) \
|
|
|
834 CODING_SYSTEM_ISO2022_NO_ISO6429 (XCODING_SYSTEM (codesys))
|
|
|
835 #define XCODING_SYSTEM_ISO2022_ESCAPE_QUOTED(codesys) \
|
|
|
836 CODING_SYSTEM_ISO2022_ESCAPE_QUOTED (XCODING_SYSTEM (codesys))
|
|
|
837 #define XCODING_SYSTEM_ISO2022_INPUT_CONV(codesys) \
|
|
|
838 CODING_SYSTEM_ISO2022_INPUT_CONV (XCODING_SYSTEM (codesys))
|
|
|
839 #define XCODING_SYSTEM_ISO2022_OUTPUT_CONV(codesys) \
|
|
|
840 CODING_SYSTEM_ISO2022_OUTPUT_CONV (XCODING_SYSTEM (codesys))
|
|
|
841
|
|
|
842 /* Additional information used by the ISO2022 decoder and detector. */
|
|
|
843 struct iso2022_coding_stream
|
|
|
844 {
|
|
|
845 /* CHARSET holds the character sets currently assigned to the G0
|
|
|
846 through G3 variables. It is initialized from the array
|
|
|
847 INITIAL_CHARSET in CODESYS. */
|
|
|
848 Lisp_Object charset[4];
|
|
|
849
|
|
|
850 /* Which registers are currently invoked into the left (GL) and
|
|
|
851 right (GR) halves of the 8-bit encoding space? */
|
|
|
852 int register_left, register_right;
|
|
|
853
|
|
|
854 /* FLAGS holds flags indicating the current state of the encoding. Some of
|
|
|
855 these flags are actually part of the state-dependent data and should be
|
|
|
856 moved there. */
|
|
|
857 unsigned int flags;
|
|
|
858
|
|
|
859 /**************** for decoding ****************/
|
|
|
860
|
|
|
861 /* ISO_ESC holds a value indicating part of an escape sequence
|
|
|
862 that has already been seen. */
|
|
|
863 enum iso_esc_flag esc;
|
|
|
864
|
|
|
865 /* This records the bytes we've seen so far in an escape sequence,
|
|
|
866 in case the sequence is invalid (we spit out the bytes unchanged). */
|
|
|
867 unsigned char esc_bytes[8];
|
|
|
868
|
|
|
869 /* Index for next byte to store in ISO escape sequence. */
|
|
|
870 int esc_bytes_index;
|
|
|
871
|
|
|
872 #ifdef ENABLE_COMPOSITE_CHARS
|
|
|
873 /* Stuff seen so far when composing a string. */
|
|
|
874 unsigned_char_dynarr *composite_chars;
|
|
|
875 #endif
|
|
|
876
|
|
|
877 /* If we saw an invalid designation sequence for a particular
|
|
|
878 register, we flag it here and switch to ASCII. The next time we
|
|
|
879 see a valid designation for this register, we turn off the flag
|
|
|
880 and do the designation normally, but pretend the sequence was
|
|
|
881 invalid. The effect of all this is that (most of the time) the
|
|
|
882 escape sequences for both the switch to the unknown charset, and
|
|
|
883 the switch back to the known charset, get inserted literally into
|
|
|
884 the buffer and saved out as such. The hope is that we can
|
|
|
885 preserve the escape sequences so that the resulting written out
|
|
|
886 file makes sense. If we don't do any of this, the designation
|
|
|
887 to the invalid charset will be preserved but that switch back
|
|
|
888 to the known charset will probably get eaten because it was
|
|
|
889 the same charset that was already present in the register. */
|
|
|
890 unsigned char invalid_designated[4];
|
|
|
891
|
|
|
892 /* We try to do similar things as above for direction-switching
|
|
|
893 sequences. If we encountered a direction switch while an
|
|
|
894 invalid designation was present, or an invalid designation
|
|
|
895 just after a direction switch (i.e. no valid designation
|
|
|
896 encountered yet), we insert the direction-switch escape
|
|
|
897 sequence literally into the output stream, and later on
|
|
|
898 insert the corresponding direction-restoring escape sequence
|
|
|
899 literally also. */
|
|
|
900 unsigned int switched_dir_and_no_valid_charset_yet :1;
|
|
|
901 unsigned int invalid_switch_dir :1;
|
|
|
902
|
|
|
903 /* Tells the decoder to output the escape sequence literally
|
|
|
904 even though it was valid. Used in the games we play to
|
|
|
905 avoid lossage when we encounter invalid designations. */
|
|
|
906 unsigned int output_literally :1;
|
|
|
907 /* We encountered a direction switch followed by an invalid
|
|
|
908 designation. We didn't output the direction switch
|
|
|
909 literally because we didn't know about the invalid designation;
|
|
|
910 but we have to do so now. */
|
|
|
911 unsigned int output_direction_sequence :1;
|
|
|
912
|
|
|
913 /**************** for encoding ****************/
|
|
|
914
|
|
|
915 /* Whether we need to explicitly designate the charset in the
|
|
|
916 G? register before using it. It is initialized from the
|
|
|
917 array FORCE_CHARSET_ON_OUTPUT in CODESYS. */
|
|
|
918 unsigned char force_charset_on_output[4];
|
|
|
919
|
|
|
920 /* Other state variables that need to be preserved across
|
|
|
921 invocations. */
|
|
|
922 Lisp_Object current_charset;
|
|
|
923 int current_half;
|
|
|
924 int current_char_boundary;
|
|
|
925 };
|
|
|
926
|
|
1204
|
927 static const struct memory_description ccs_description_1[] =
|
|
771
|
928 {
|
|
|
929 { XD_LISP_OBJECT, offsetof (charset_conversion_spec, from_charset) },
|
|
|
930 { XD_LISP_OBJECT, offsetof (charset_conversion_spec, to_charset) },
|
|
|
931 { XD_END }
|
|
|
932 };
|
|
|
933
|
|
1204
|
934 static const struct sized_memory_description ccs_description =
|
|
771
|
935 {
|
|
|
936 sizeof (charset_conversion_spec),
|
|
|
937 ccs_description_1
|
|
|
938 };
|
|
|
939
|
|
1204
|
940 static const struct memory_description ccsd_description_1[] =
|
|
771
|
941 {
|
|
|
942 XD_DYNARR_DESC (charset_conversion_spec_dynarr, &ccs_description),
|
|
|
943 { XD_END }
|
|
|
944 };
|
|
|
945
|
|
1204
|
946 static const struct sized_memory_description ccsd_description =
|
|
771
|
947 {
|
|
|
948 sizeof (charset_conversion_spec_dynarr),
|
|
|
949 ccsd_description_1
|
|
|
950 };
|
|
|
951
|
|
1204
|
952 static const struct memory_description iso2022_coding_system_description[] = {
|
|
|
953 { XD_LISP_OBJECT_ARRAY, offsetof (struct iso2022_coding_system,
|
|
|
954 initial_charset), 4 },
|
|
|
955 { XD_STRUCT_PTR, offsetof (struct iso2022_coding_system, input_conv),
|
|
771
|
956 1, &ccsd_description },
|
|
1204
|
957 { XD_STRUCT_PTR, offsetof (struct iso2022_coding_system, output_conv),
|
|
771
|
958 1, &ccsd_description },
|
|
|
959 { XD_END }
|
|
|
960 };
|
|
|
961
|
|
1204
|
962 DEFINE_CODING_SYSTEM_TYPE_WITH_DATA (iso2022);
|
|
|
963
|
|
771
|
964 /* The following note taken directly from FSF 21.0.103. */
|
|
|
965
|
|
|
966 /* The following note describes the coding system ISO2022 briefly.
|
|
|
967 Since the intention of this note is to help understand the
|
|
|
968 functions in this file, some parts are NOT ACCURATE or are OVERLY
|
|
|
969 SIMPLIFIED. For thorough understanding, please refer to the
|
|
|
970 original document of ISO2022. This is equivalent to the standard
|
|
|
971 ECMA-35, obtainable from <URL:http://www.ecma.ch/> (*).
|
|
|
972
|
|
|
973 ISO2022 provides many mechanisms to encode several character sets
|
|
|
974 in 7-bit and 8-bit environments. For 7-bit environments, all text
|
|
|
975 is encoded using bytes less than 128. This may make the encoded
|
|
|
976 text a little bit longer, but the text passes more easily through
|
|
|
977 several types of gateway, some of which strip off the MSB (Most
|
|
|
978 Significant Bit).
|
|
|
979
|
|
|
980 There are two kinds of character sets: control character sets and
|
|
|
981 graphic character sets. The former contain control characters such
|
|
|
982 as `newline' and `escape' to provide control functions (control
|
|
|
983 functions are also provided by escape sequences). The latter
|
|
|
984 contain graphic characters such as 'A' and '-'. Emacs recognizes
|
|
|
985 two control character sets and many graphic character sets.
|
|
|
986
|
|
|
987 Graphic character sets are classified into one of the following
|
|
|
988 four classes, according to the number of bytes (DIMENSION) and
|
|
|
989 number of characters in one dimension (CHARS) of the set:
|
|
|
990 - DIMENSION1_CHARS94
|
|
|
991 - DIMENSION1_CHARS96
|
|
|
992 - DIMENSION2_CHARS94
|
|
|
993 - DIMENSION2_CHARS96
|
|
|
994
|
|
|
995 In addition, each character set is assigned an identification tag,
|
|
|
996 unique for each set, called the "final character" (denoted as <F>
|
|
|
997 hereafter). The <F> of each character set is decided by ECMA(*)
|
|
|
998 when it is registered in ISO. The code range of <F> is 0x30..0x7F
|
|
|
999 (0x30..0x3F are for private use only).
|
|
|
1000
|
|
|
1001 Note (*): ECMA = European Computer Manufacturers Association
|
|
|
1002
|
|
|
1003 Here are examples of graphic character sets [NAME(<F>)]:
|
|
|
1004 o DIMENSION1_CHARS94 -- ASCII('B'), right-half-of-JISX0201('I'), ...
|
|
|
1005 o DIMENSION1_CHARS96 -- right-half-of-ISO8859-1('A'), ...
|
|
|
1006 o DIMENSION2_CHARS94 -- GB2312('A'), JISX0208('B'), ...
|
|
|
1007 o DIMENSION2_CHARS96 -- none for the moment
|
|
|
1008
|
|
|
1009 A code area (1 byte=8 bits) is divided into 4 areas, C0, GL, C1, and GR.
|
|
|
1010 C0 [0x00..0x1F] -- control character plane 0
|
|
|
1011 GL [0x20..0x7F] -- graphic character plane 0
|
|
|
1012 C1 [0x80..0x9F] -- control character plane 1
|
|
|
1013 GR [0xA0..0xFF] -- graphic character plane 1
|
|
|
1014
|
|
|
1015 A control character set is directly designated and invoked to C0 or
|
|
|
1016 C1 by an escape sequence. The most common case is that:
|
|
|
1017 - ISO646's control character set is designated/invoked to C0, and
|
|
|
1018 - ISO6429's control character set is designated/invoked to C1,
|
|
|
1019 and usually these designations/invocations are omitted in encoded
|
|
|
1020 text. In a 7-bit environment, only C0 can be used, and a control
|
|
|
1021 character for C1 is encoded by an appropriate escape sequence to
|
|
|
1022 fit into the environment. All control characters for C1 are
|
|
|
1023 defined to have corresponding escape sequences.
|
|
|
1024
|
|
|
1025 A graphic character set is at first designated to one of four
|
|
|
1026 graphic registers (G0 through G3), then these graphic registers are
|
|
|
1027 invoked to GL or GR. These designations and invocations can be
|
|
|
1028 done independently. The most common case is that G0 is invoked to
|
|
|
1029 GL, G1 is invoked to GR, and ASCII is designated to G0. Usually
|
|
|
1030 these invocations and designations are omitted in encoded text.
|
|
|
1031 In a 7-bit environment, only GL can be used.
|
|
|
1032
|
|
|
1033 When a graphic character set of CHARS94 is invoked to GL, codes
|
|
|
1034 0x20 and 0x7F of the GL area work as control characters SPACE and
|
|
|
1035 DEL respectively, and codes 0xA0 and 0xFF of the GR area should not
|
|
|
1036 be used.
|
|
|
1037
|
|
|
1038 There are two ways of invocation: locking-shift and single-shift.
|
|
|
1039 With locking-shift, the invocation lasts until the next different
|
|
|
1040 invocation, whereas with single-shift, the invocation affects the
|
|
|
1041 following character only and doesn't affect the locking-shift
|
|
|
1042 state. Invocations are done by the following control characters or
|
|
|
1043 escape sequences:
|
|
|
1044
|
|
|
1045 ----------------------------------------------------------------------
|
|
|
1046 abbrev function cntrl escape seq description
|
|
|
1047 ----------------------------------------------------------------------
|
|
|
1048 SI/LS0 (shift-in) 0x0F none invoke G0 into GL
|
|
|
1049 SO/LS1 (shift-out) 0x0E none invoke G1 into GL
|
|
|
1050 LS2 (locking-shift-2) none ESC 'n' invoke G2 into GL
|
|
|
1051 LS3 (locking-shift-3) none ESC 'o' invoke G3 into GL
|
|
|
1052 LS1R (locking-shift-1 right) none ESC '~' invoke G1 into GR (*)
|
|
|
1053 LS2R (locking-shift-2 right) none ESC '}' invoke G2 into GR (*)
|
|
|
1054 LS3R (locking-shift 3 right) none ESC '|' invoke G3 into GR (*)
|
|
|
1055 SS2 (single-shift-2) 0x8E ESC 'N' invoke G2 for one char
|
|
|
1056 SS3 (single-shift-3) 0x8F ESC 'O' invoke G3 for one char
|
|
|
1057 ----------------------------------------------------------------------
|
|
|
1058 (*) These are not used by any known coding system.
|
|
|
1059
|
|
|
1060 Control characters for these functions are defined by macros
|
|
|
1061 ISO_CODE_XXX in `coding.h'.
|
|
|
1062
|
|
|
1063 Designations are done by the following escape sequences:
|
|
|
1064 ----------------------------------------------------------------------
|
|
|
1065 escape sequence description
|
|
|
1066 ----------------------------------------------------------------------
|
|
|
1067 ESC '(' <F> designate DIMENSION1_CHARS94<F> to G0
|
|
|
1068 ESC ')' <F> designate DIMENSION1_CHARS94<F> to G1
|
|
|
1069 ESC '*' <F> designate DIMENSION1_CHARS94<F> to G2
|
|
|
1070 ESC '+' <F> designate DIMENSION1_CHARS94<F> to G3
|
|
|
1071 ESC ',' <F> designate DIMENSION1_CHARS96<F> to G0 (*)
|
|
|
1072 ESC '-' <F> designate DIMENSION1_CHARS96<F> to G1
|
|
|
1073 ESC '.' <F> designate DIMENSION1_CHARS96<F> to G2
|
|
|
1074 ESC '/' <F> designate DIMENSION1_CHARS96<F> to G3
|
|
|
1075 ESC '$' '(' <F> designate DIMENSION2_CHARS94<F> to G0 (**)
|
|
|
1076 ESC '$' ')' <F> designate DIMENSION2_CHARS94<F> to G1
|
|
|
1077 ESC '$' '*' <F> designate DIMENSION2_CHARS94<F> to G2
|
|
|
1078 ESC '$' '+' <F> designate DIMENSION2_CHARS94<F> to G3
|
|
|
1079 ESC '$' ',' <F> designate DIMENSION2_CHARS96<F> to G0 (*)
|
|
|
1080 ESC '$' '-' <F> designate DIMENSION2_CHARS96<F> to G1
|
|
|
1081 ESC '$' '.' <F> designate DIMENSION2_CHARS96<F> to G2
|
|
|
1082 ESC '$' '/' <F> designate DIMENSION2_CHARS96<F> to G3
|
|
|
1083 ----------------------------------------------------------------------
|
|
|
1084
|
|
|
1085 In this list, "DIMENSION1_CHARS94<F>" means a graphic character set
|
|
|
1086 of dimension 1, chars 94, and final character <F>, etc...
|
|
|
1087
|
|
|
1088 Note (*): Although these designations are not allowed in ISO2022,
|
|
|
1089 Emacs accepts them on decoding, and produces them on encoding
|
|
|
1090 CHARS96 character sets in a coding system which is characterized as
|
|
|
1091 7-bit environment, non-locking-shift, and non-single-shift.
|
|
|
1092
|
|
|
1093 Note (**): If <F> is '@', 'A', or 'B', the intermediate character
|
|
|
1094 '(' can be omitted. We refer to this as "short-form" hereafter.
|
|
|
1095
|
|
|
1096 Now you may notice that there are a lot of ways of encoding the
|
|
|
1097 same multilingual text in ISO2022. Actually, there exist many
|
|
|
1098 coding systems such as Compound Text (used in X11's inter client
|
|
|
1099 communication, ISO-2022-JP (used in Japanese Internet), ISO-2022-KR
|
|
|
1100 (used in Korean Internet), EUC (Extended UNIX Code, used in Asian
|
|
|
1101 localized platforms), and all of these are variants of ISO2022.
|
|
|
1102
|
|
|
1103 In addition to the above, Emacs handles two more kinds of escape
|
|
|
1104 sequences: ISO6429's direction specification and Emacs' private
|
|
|
1105 sequence for specifying character composition.
|
|
|
1106
|
|
|
1107 ISO6429's direction specification takes the following form:
|
|
|
1108 o CSI ']' -- end of the current direction
|
|
|
1109 o CSI '0' ']' -- end of the current direction
|
|
|
1110 o CSI '1' ']' -- start of left-to-right text
|
|
|
1111 o CSI '2' ']' -- start of right-to-left text
|
|
|
1112 The control character CSI (0x9B: control sequence introducer) is
|
|
|
1113 abbreviated to the escape sequence ESC '[' in a 7-bit environment.
|
|
|
1114
|
|
|
1115 Character composition specification takes the following form:
|
|
|
1116 o ESC '0' -- start relative composition
|
|
|
1117 o ESC '1' -- end composition
|
|
|
1118 o ESC '2' -- start rule-base composition (*)
|
|
|
1119 o ESC '3' -- start relative composition with alternate chars (**)
|
|
|
1120 o ESC '4' -- start rule-base composition with alternate chars (**)
|
|
|
1121 Since these are not standard escape sequences of any ISO standard,
|
|
|
1122 the use of them with these meanings is restricted to Emacs only.
|
|
|
1123
|
|
|
1124 (*) This form is used only in Emacs 20.5 and older versions,
|
|
|
1125 but the newer versions can safely decode it.
|
|
|
1126 (**) This form is used only in Emacs 21.1 and newer versions,
|
|
|
1127 and the older versions can't decode it.
|
|
|
1128
|
|
|
1129 Here's a list of example usages of these composition escape
|
|
|
1130 sequences (categorized by `enum composition_method').
|
|
|
1131
|
|
|
1132 COMPOSITION_RELATIVE:
|
|
|
1133 ESC 0 CHAR [ CHAR ] ESC 1
|
|
|
1134 COMPOSITION_WITH_RULE:
|
|
|
1135 ESC 2 CHAR [ RULE CHAR ] ESC 1
|
|
|
1136 COMPOSITION_WITH_ALTCHARS:
|
|
|
1137 ESC 3 ALTCHAR [ ALTCHAR ] ESC 0 CHAR [ CHAR ] ESC 1
|
|
|
1138 COMPOSITION_WITH_RULE_ALTCHARS:
|
|
|
1139 ESC 4 ALTCHAR [ RULE ALTCHAR ] ESC 0 CHAR [ CHAR ] ESC 1 */
|
|
|
1140
|
|
|
1141 static void
|
|
|
1142 reset_iso2022_decode (Lisp_Object coding_system,
|
|
|
1143 struct iso2022_coding_stream *data)
|
|
|
1144 {
|
|
|
1145 int i;
|
|
|
1146 #ifdef ENABLE_COMPOSITE_CHARS
|
|
|
1147 unsigned_char_dynarr *old_composite_chars = data->composite_chars;
|
|
|
1148 #endif
|
|
|
1149
|
|
|
1150 xzero (*data);
|
|
|
1151
|
|
|
1152 for (i = 0; i < 4; i++)
|
|
|
1153 {
|
|
|
1154 if (!NILP (coding_system))
|
|
|
1155 data->charset[i] =
|
|
|
1156 XCODING_SYSTEM_ISO2022_INITIAL_CHARSET (coding_system, i);
|
|
|
1157 else
|
|
|
1158 data->charset[i] = Qt;
|
|
|
1159 }
|
|
|
1160 data->esc = ISO_ESC_NOTHING;
|
|
|
1161 data->register_right = 1;
|
|
|
1162 #ifdef ENABLE_COMPOSITE_CHARS
|
|
|
1163 if (old_composite_chars)
|
|
|
1164 {
|
|
|
1165 data->composite_chars = old_composite_chars;
|
|
|
1166 Dynarr_reset (data->composite_chars);
|
|
|
1167 }
|
|
|
1168 #endif
|
|
|
1169 }
|
|
|
1170
|
|
|
1171 static void
|
|
|
1172 reset_iso2022_encode (Lisp_Object coding_system,
|
|
|
1173 struct iso2022_coding_stream *data)
|
|
|
1174 {
|
|
|
1175 int i;
|
|
|
1176
|
|
|
1177 xzero (*data);
|
|
|
1178
|
|
|
1179 for (i = 0; i < 4; i++)
|
|
|
1180 {
|
|
|
1181 data->charset[i] =
|
|
|
1182 XCODING_SYSTEM_ISO2022_INITIAL_CHARSET (coding_system, i);
|
|
|
1183 data->force_charset_on_output[i] =
|
|
|
1184 XCODING_SYSTEM_ISO2022_FORCE_CHARSET_ON_OUTPUT (coding_system, i);
|
|
|
1185 }
|
|
|
1186 data->register_right = 1;
|
|
|
1187 data->current_charset = Qnil;
|
|
|
1188 data->current_char_boundary = 1;
|
|
|
1189 }
|
|
|
1190
|
|
|
1191 static void
|
|
|
1192 iso2022_init_coding_stream (struct coding_stream *str)
|
|
|
1193 {
|
|
|
1194 if (str->direction == CODING_DECODE)
|
|
|
1195 reset_iso2022_decode (str->codesys,
|
|
|
1196 CODING_STREAM_TYPE_DATA (str, iso2022));
|
|
|
1197 else
|
|
|
1198 reset_iso2022_encode (str->codesys,
|
|
|
1199 CODING_STREAM_TYPE_DATA (str, iso2022));
|
|
|
1200 }
|
|
|
1201
|
|
|
1202 static void
|
|
|
1203 iso2022_rewind_coding_stream (struct coding_stream *str)
|
|
|
1204 {
|
|
|
1205 iso2022_init_coding_stream (str);
|
|
|
1206 }
|
|
|
1207
|
|
|
1208 static int
|
|
|
1209 fit_to_be_escape_quoted (unsigned char c)
|
|
|
1210 {
|
|
|
1211 switch (c)
|
|
|
1212 {
|
|
|
1213 case ISO_CODE_ESC:
|
|
|
1214 case ISO_CODE_CSI:
|
|
|
1215 case ISO_CODE_SS2:
|
|
|
1216 case ISO_CODE_SS3:
|
|
|
1217 case ISO_CODE_SO:
|
|
|
1218 case ISO_CODE_SI:
|
|
|
1219 return 1;
|
|
|
1220
|
|
|
1221 default:
|
|
|
1222 return 0;
|
|
|
1223 }
|
|
|
1224 }
|
|
|
1225
|
|
|
1226 static Lisp_Object
|
|
867
|
1227 charset_by_attributes_or_create_one (int type, Ibyte final, int dir)
|
|
771
|
1228 {
|
|
826
|
1229 Lisp_Object charset = charset_by_attributes (type, final, dir);
|
|
771
|
1230
|
|
|
1231 if (NILP (charset))
|
|
|
1232 {
|
|
|
1233 int chars, dim;
|
|
|
1234
|
|
|
1235 switch (type)
|
|
|
1236 {
|
|
|
1237 case CHARSET_TYPE_94:
|
|
|
1238 chars = 94; dim = 1;
|
|
|
1239 break;
|
|
|
1240 case CHARSET_TYPE_96:
|
|
|
1241 chars = 96; dim = 1;
|
|
|
1242 break;
|
|
|
1243 case CHARSET_TYPE_94X94:
|
|
|
1244 chars = 94; dim = 2;
|
|
|
1245 break;
|
|
|
1246 case CHARSET_TYPE_96X96:
|
|
|
1247 chars = 96; dim = 2;
|
|
|
1248 break;
|
|
|
1249 default:
|
|
|
1250 abort (); chars = 0; dim = 0;
|
|
|
1251 }
|
|
|
1252
|
|
|
1253 charset = Fmake_charset (Qunbound, Qnil,
|
|
|
1254 nconc2 (list6 (Qfinal, make_char (final),
|
|
|
1255 Qchars, make_int (chars),
|
|
|
1256 Qdimension, make_int (dim)),
|
|
|
1257 list2 (Qdirection,
|
|
|
1258 dir == CHARSET_LEFT_TO_RIGHT ?
|
|
|
1259 Ql2r : Qr2l)));
|
|
|
1260 }
|
|
|
1261
|
|
|
1262 return charset;
|
|
|
1263 }
|
|
|
1264
|
|
|
1265 /* Parse one byte of an ISO2022 escape sequence.
|
|
|
1266 If the result is an invalid escape sequence, return 0 and
|
|
|
1267 do not change anything in STR. Otherwise, if the result is
|
|
|
1268 an incomplete escape sequence, update ISO2022.ESC and
|
|
|
1269 ISO2022.ESC_BYTES and return -1. Otherwise, update
|
|
|
1270 all the state variables (but not ISO2022.ESC_BYTES) and
|
|
|
1271 return 1.
|
|
|
1272
|
|
|
1273 If CHECK_INVALID_CHARSETS is non-zero, check for designation
|
|
|
1274 or invocation of an invalid character set and treat that as
|
|
|
1275 an unrecognized escape sequence.
|
|
|
1276
|
|
|
1277 ********************************************************************
|
|
|
1278
|
|
|
1279 #### Strategies for error annotation and coding orthogonalization
|
|
|
1280
|
|
|
1281 We really want to separate out a number of things. Conceptually,
|
|
|
1282 there is a nested syntax.
|
|
|
1283
|
|
|
1284 At the top level is the ISO 2022 extension syntax, including charset
|
|
|
1285 designation and invocation, and certain auxiliary controls such as the
|
|
|
1286 ISO 6429 direction specification. These are octet-oriented, with the
|
|
|
1287 single exception (AFAIK) of the "exit Unicode" sequence which uses the
|
|
|
1288 UTF's natural width (1 byte for UTF-7 and UTF-8, 2 bytes for UCS-2 and
|
|
|
1289 UTF-16, and 4 bytes for UCS-4 and UTF-32). This will be treated as a
|
|
|
1290 (deprecated) special case in Unicode processing.
|
|
|
1291
|
|
|
1292 The middle layer is ISO 2022 character interpretation. This will depend
|
|
|
1293 on the current state of the ISO 2022 registers, and assembles octets
|
|
|
1294 into the character's internal representation.
|
|
|
1295
|
|
|
1296 The lowest level is translating system control conventions. At present
|
|
|
1297 this is restricted to newline translation, but one could imagine doing
|
|
|
1298 tab conversion or line wrapping here. "Escape from Unicode" processing
|
|
|
1299 would be done at this level.
|
|
|
1300
|
|
|
1301 At each level the parser will verify the syntax. In the case of a
|
|
|
1302 syntax error or warning (such as a redundant escape sequence that affects
|
|
|
1303 no characters), the parser will take some action, typically inserting the
|
|
|
1304 erroneous octets directly into the output and creating an annotation
|
|
|
1305 which can be used by higher level I/O to mark the affected region.
|
|
|
1306
|
|
|
1307 This should make it possible to do something sensible about separating
|
|
|
1308 newline convention processing from character construction, and about
|
|
|
1309 preventing ISO 2022 escape sequences from being recognized
|
|
|
1310 inappropriately.
|
|
|
1311
|
|
|
1312 The basic strategy will be to have octet classification tables, and
|
|
|
1313 switch processing according to the table entry.
|
|
|
1314
|
|
|
1315 It's possible that, by doing the processing with tables of functions or
|
|
|
1316 the like, the parser can be used for both detection and translation. */
|
|
|
1317
|
|
|
1318 static int
|
|
|
1319 parse_iso2022_esc (Lisp_Object codesys, struct iso2022_coding_stream *iso,
|
|
|
1320 unsigned char c, unsigned int *flags,
|
|
|
1321 int check_invalid_charsets)
|
|
|
1322 {
|
|
|
1323 /* (1) If we're at the end of a designation sequence, CS is the
|
|
|
1324 charset being designated and REG is the register to designate
|
|
|
1325 it to.
|
|
|
1326
|
|
|
1327 (2) If we're at the end of a locking-shift sequence, REG is
|
|
|
1328 the register to invoke and HALF (0 == left, 1 == right) is
|
|
|
1329 the half to invoke it into.
|
|
|
1330
|
|
|
1331 (3) If we're at the end of a single-shift sequence, REG is
|
|
|
1332 the register to invoke. */
|
|
|
1333 Lisp_Object cs = Qnil;
|
|
|
1334 int reg, half;
|
|
|
1335
|
|
|
1336 /* NOTE: This code does goto's all over the fucking place.
|
|
|
1337 The reason for this is that we're basically implementing
|
|
|
1338 a state machine here, and hierarchical languages like C
|
|
|
1339 don't really provide a clean way of doing this. */
|
|
|
1340
|
|
|
1341 if (! (*flags & ISO_STATE_ESCAPE))
|
|
|
1342 /* At beginning of escape sequence; we need to reset our
|
|
|
1343 escape-state variables. */
|
|
|
1344 iso->esc = ISO_ESC_NOTHING;
|
|
|
1345
|
|
|
1346 iso->output_literally = 0;
|
|
|
1347 iso->output_direction_sequence = 0;
|
|
|
1348
|
|
|
1349 switch (iso->esc)
|
|
|
1350 {
|
|
|
1351 case ISO_ESC_NOTHING:
|
|
|
1352 iso->esc_bytes_index = 0;
|
|
|
1353 switch (c)
|
|
|
1354 {
|
|
|
1355 case ISO_CODE_ESC: /* Start escape sequence */
|
|
|
1356 *flags |= ISO_STATE_ESCAPE;
|
|
|
1357 iso->esc = ISO_ESC;
|
|
|
1358 goto not_done;
|
|
|
1359
|
|
|
1360 case ISO_CODE_CSI: /* ISO6429 (specifying directionality) */
|
|
|
1361 *flags |= ISO_STATE_ESCAPE;
|
|
|
1362 iso->esc = ISO_ESC_5_11;
|
|
|
1363 goto not_done;
|
|
|
1364
|
|
|
1365 case ISO_CODE_SO: /* locking shift 1 */
|
|
|
1366 reg = 1; half = 0;
|
|
|
1367 goto locking_shift;
|
|
|
1368 case ISO_CODE_SI: /* locking shift 0 */
|
|
|
1369 reg = 0; half = 0;
|
|
|
1370 goto locking_shift;
|
|
|
1371
|
|
|
1372 case ISO_CODE_SS2: /* single shift */
|
|
|
1373 reg = 2;
|
|
|
1374 goto single_shift;
|
|
|
1375 case ISO_CODE_SS3: /* single shift */
|
|
|
1376 reg = 3;
|
|
|
1377 goto single_shift;
|
|
|
1378
|
|
|
1379 default: /* Other control characters */
|
|
|
1380 error:
|
|
|
1381 *flags &= ISO_STATE_LOCK;
|
|
|
1382 return 0;
|
|
|
1383 }
|
|
|
1384
|
|
|
1385 case ISO_ESC:
|
|
|
1386 switch (c)
|
|
|
1387 {
|
|
|
1388 /**** single shift ****/
|
|
|
1389
|
|
|
1390 case 'N': /* single shift 2 */
|
|
|
1391 reg = 2;
|
|
|
1392 goto single_shift;
|
|
|
1393 case 'O': /* single shift 3 */
|
|
|
1394 reg = 3;
|
|
|
1395 goto single_shift;
|
|
|
1396
|
|
|
1397 /**** locking shift ****/
|
|
|
1398
|
|
|
1399 case '~': /* locking shift 1 right */
|
|
|
1400 reg = 1; half = 1;
|
|
|
1401 goto locking_shift;
|
|
|
1402 case 'n': /* locking shift 2 */
|
|
|
1403 reg = 2; half = 0;
|
|
|
1404 goto locking_shift;
|
|
|
1405 case '}': /* locking shift 2 right */
|
|
|
1406 reg = 2; half = 1;
|
|
|
1407 goto locking_shift;
|
|
|
1408 case 'o': /* locking shift 3 */
|
|
|
1409 reg = 3; half = 0;
|
|
|
1410 goto locking_shift;
|
|
|
1411 case '|': /* locking shift 3 right */
|
|
|
1412 reg = 3; half = 1;
|
|
|
1413 goto locking_shift;
|
|
|
1414
|
|
|
1415 /**** composite ****/
|
|
|
1416
|
|
|
1417 #ifdef ENABLE_COMPOSITE_CHARS
|
|
|
1418 case '0':
|
|
|
1419 iso->esc = ISO_ESC_START_COMPOSITE;
|
|
|
1420 *flags = (*flags & ISO_STATE_LOCK) |
|
|
|
1421 ISO_STATE_COMPOSITE;
|
|
|
1422 return 1;
|
|
|
1423
|
|
|
1424 case '1':
|
|
|
1425 iso->esc = ISO_ESC_END_COMPOSITE;
|
|
|
1426 *flags = (*flags & ISO_STATE_LOCK) &
|
|
|
1427 ~ISO_STATE_COMPOSITE;
|
|
|
1428 return 1;
|
|
|
1429 #else
|
|
|
1430 case '0': case '1': case '2': case '3': case '4':
|
|
|
1431 /* We simply return a flag indicating that some composite
|
|
|
1432 escape was seen. The caller will use the particular
|
|
|
1433 character to encode the appropriate "composite hack"
|
|
|
1434 character out of Vcharset_composite, so that we will
|
|
|
1435 preserve these values on output. */
|
|
|
1436 iso->esc = ISO_ESC_START_COMPOSITE;
|
|
|
1437 *flags &= ISO_STATE_LOCK;
|
|
|
1438 return 1;
|
|
|
1439 #endif /* ENABLE_COMPOSITE_CHARS */
|
|
|
1440
|
|
|
1441 /**** directionality ****/
|
|
|
1442
|
|
|
1443 case '[':
|
|
|
1444 iso->esc = ISO_ESC_5_11;
|
|
|
1445 goto not_done;
|
|
|
1446
|
|
|
1447 /**** designation ****/
|
|
|
1448
|
|
|
1449 case '$': /* multibyte charset prefix */
|
|
|
1450 iso->esc = ISO_ESC_2_4;
|
|
|
1451 goto not_done;
|
|
|
1452
|
|
|
1453 default:
|
|
|
1454 if (0x28 <= c && c <= 0x2F)
|
|
|
1455 {
|
|
|
1456 iso->esc = (enum iso_esc_flag) (c - 0x28 + ISO_ESC_2_8);
|
|
|
1457 goto not_done;
|
|
|
1458 }
|
|
|
1459
|
|
|
1460 /* This function is called with CODESYS equal to nil when
|
|
|
1461 doing coding-system detection. */
|
|
|
1462 if (!NILP (codesys)
|
|
|
1463 && XCODING_SYSTEM_ISO2022_ESCAPE_QUOTED (codesys)
|
|
|
1464 && fit_to_be_escape_quoted (c))
|
|
|
1465 {
|
|
|
1466 iso->esc = ISO_ESC_LITERAL;
|
|
|
1467 *flags &= ISO_STATE_LOCK;
|
|
|
1468 return 1;
|
|
|
1469 }
|
|
|
1470
|
|
|
1471 /* bzzzt! */
|
|
|
1472 goto error;
|
|
|
1473 }
|
|
|
1474
|
|
|
1475
|
|
|
1476
|
|
|
1477 /**** directionality ****/
|
|
|
1478
|
|
|
1479 case ISO_ESC_5_11: /* ISO6429 direction control */
|
|
|
1480 if (c == ']')
|
|
|
1481 {
|
|
|
1482 *flags &= (ISO_STATE_LOCK & ~ISO_STATE_R2L);
|
|
|
1483 goto directionality;
|
|
|
1484 }
|
|
|
1485 if (c == '0') iso->esc = ISO_ESC_5_11_0;
|
|
|
1486 else if (c == '1') iso->esc = ISO_ESC_5_11_1;
|
|
|
1487 else if (c == '2') iso->esc = ISO_ESC_5_11_2;
|
|
|
1488 else goto error;
|
|
|
1489 goto not_done;
|
|
|
1490
|
|
|
1491 case ISO_ESC_5_11_0:
|
|
|
1492 if (c == ']')
|
|
|
1493 {
|
|
|
1494 *flags &= (ISO_STATE_LOCK & ~ISO_STATE_R2L);
|
|
|
1495 goto directionality;
|
|
|
1496 }
|
|
|
1497 goto error;
|
|
|
1498
|
|
|
1499 case ISO_ESC_5_11_1:
|
|
|
1500 if (c == ']')
|
|
|
1501 {
|
|
|
1502 *flags = (ISO_STATE_LOCK & ~ISO_STATE_R2L);
|
|
|
1503 goto directionality;
|
|
|
1504 }
|
|
|
1505 goto error;
|
|
|
1506
|
|
|
1507 case ISO_ESC_5_11_2:
|
|
|
1508 if (c == ']')
|
|
|
1509 {
|
|
|
1510 *flags = (*flags & ISO_STATE_LOCK) | ISO_STATE_R2L;
|
|
|
1511 goto directionality;
|
|
|
1512 }
|
|
|
1513 goto error;
|
|
|
1514
|
|
|
1515 directionality:
|
|
|
1516 iso->esc = ISO_ESC_DIRECTIONALITY;
|
|
|
1517 /* Various junk here to attempt to preserve the direction sequences
|
|
|
1518 literally in the text if they would otherwise be swallowed due
|
|
|
1519 to invalid designations that don't show up as actual charset
|
|
|
1520 changes in the text. */
|
|
|
1521 if (iso->invalid_switch_dir)
|
|
|
1522 {
|
|
|
1523 /* We already inserted a direction switch literally into the
|
|
|
1524 text. We assume (#### this may not be right) that the
|
|
|
1525 next direction switch is the one going the other way,
|
|
|
1526 and we need to output that literally as well. */
|
|
|
1527 iso->output_literally = 1;
|
|
|
1528 iso->invalid_switch_dir = 0;
|
|
|
1529 }
|
|
|
1530 else
|
|
|
1531 {
|
|
|
1532 int jj;
|
|
|
1533
|
|
|
1534 /* If we are in the thrall of an invalid designation,
|
|
|
1535 then stick the directionality sequence literally into the
|
|
|
1536 output stream so it ends up in the original text again. */
|
|
|
1537 for (jj = 0; jj < 4; jj++)
|
|
|
1538 if (iso->invalid_designated[jj])
|
|
|
1539 break;
|
|
|
1540 if (jj < 4)
|
|
|
1541 {
|
|
|
1542 iso->output_literally = 1;
|
|
|
1543 iso->invalid_switch_dir = 1;
|
|
|
1544 }
|
|
|
1545 else
|
|
|
1546 /* Indicate that we haven't yet seen a valid designation,
|
|
|
1547 so that if a switch-dir is directly followed by an
|
|
|
1548 invalid designation, both get inserted literally. */
|
|
|
1549 iso->switched_dir_and_no_valid_charset_yet = 1;
|
|
|
1550 }
|
|
|
1551 return 1;
|
|
|
1552
|
|
|
1553
|
|
|
1554 /**** designation ****/
|
|
|
1555
|
|
|
1556 case ISO_ESC_2_4:
|
|
|
1557 if (0x28 <= c && c <= 0x2F)
|
|
|
1558 {
|
|
|
1559 iso->esc = (enum iso_esc_flag) (c - 0x28 + ISO_ESC_2_4_8);
|
|
|
1560 goto not_done;
|
|
|
1561 }
|
|
|
1562 if (0x40 <= c && c <= 0x42)
|
|
|
1563 {
|
|
|
1564 cs = charset_by_attributes_or_create_one (CHARSET_TYPE_94X94, c,
|
|
|
1565 *flags & ISO_STATE_R2L ?
|
|
|
1566 CHARSET_RIGHT_TO_LEFT :
|
|
|
1567 CHARSET_LEFT_TO_RIGHT);
|
|
|
1568 reg = 0;
|
|
|
1569 goto designated;
|
|
|
1570 }
|
|
|
1571 goto error;
|
|
|
1572
|
|
|
1573 default:
|
|
|
1574 {
|
|
|
1575 int type = -1;
|
|
|
1576
|
|
|
1577 if (iso->esc >= ISO_ESC_2_8 &&
|
|
|
1578 iso->esc <= ISO_ESC_2_15)
|
|
|
1579 {
|
|
|
1580 type = ((iso->esc >= ISO_ESC_2_12) ?
|
|
|
1581 CHARSET_TYPE_96 : CHARSET_TYPE_94);
|
|
|
1582 reg = (iso->esc - ISO_ESC_2_8) & 3;
|
|
|
1583 }
|
|
|
1584 else if (iso->esc >= ISO_ESC_2_4_8 &&
|
|
|
1585 iso->esc <= ISO_ESC_2_4_15)
|
|
|
1586 {
|
|
|
1587 type = ((iso->esc >= ISO_ESC_2_4_12) ?
|
|
|
1588 CHARSET_TYPE_96X96 : CHARSET_TYPE_94X94);
|
|
|
1589 reg = (iso->esc - ISO_ESC_2_4_8) & 3;
|
|
|
1590 }
|
|
|
1591 else
|
|
|
1592 {
|
|
|
1593 /* Can this ever be reached? -slb */
|
|
|
1594 abort ();
|
|
|
1595 goto error;
|
|
|
1596 }
|
|
|
1597
|
|
|
1598 if (c < '0' || c > '~' ||
|
|
|
1599 (c > 0x5F && (type == CHARSET_TYPE_94X94 ||
|
|
|
1600 type == CHARSET_TYPE_96X96)))
|
|
|
1601 goto error; /* bad final byte */
|
|
|
1602
|
|
|
1603 cs = charset_by_attributes_or_create_one (type, c,
|
|
|
1604 *flags & ISO_STATE_R2L ?
|
|
|
1605 CHARSET_RIGHT_TO_LEFT :
|
|
|
1606 CHARSET_LEFT_TO_RIGHT);
|
|
|
1607 goto designated;
|
|
|
1608 }
|
|
|
1609 }
|
|
|
1610
|
|
|
1611 not_done:
|
|
|
1612 iso->esc_bytes[iso->esc_bytes_index++] = (unsigned char) c;
|
|
|
1613 return -1;
|
|
|
1614
|
|
|
1615 single_shift:
|
|
|
1616 if (check_invalid_charsets && !CHARSETP (iso->charset[reg]))
|
|
|
1617 /* can't invoke something that ain't there. */
|
|
|
1618 goto error;
|
|
|
1619 iso->esc = ISO_ESC_SINGLE_SHIFT;
|
|
|
1620 *flags &= ISO_STATE_LOCK;
|
|
|
1621 if (reg == 2)
|
|
|
1622 *flags |= ISO_STATE_SS2;
|
|
|
1623 else
|
|
|
1624 *flags |= ISO_STATE_SS3;
|
|
|
1625 return 1;
|
|
|
1626
|
|
|
1627 locking_shift:
|
|
|
1628 if (check_invalid_charsets &&
|
|
|
1629 !CHARSETP (iso->charset[reg]))
|
|
|
1630 /* can't invoke something that ain't there. */
|
|
|
1631 goto error;
|
|
|
1632 if (half)
|
|
|
1633 iso->register_right = reg;
|
|
|
1634 else
|
|
|
1635 iso->register_left = reg;
|
|
|
1636 *flags &= ISO_STATE_LOCK;
|
|
|
1637 iso->esc = ISO_ESC_LOCKING_SHIFT;
|
|
|
1638 return 1;
|
|
|
1639
|
|
|
1640 designated:
|
|
|
1641 if (NILP (cs) && check_invalid_charsets)
|
|
|
1642 {
|
|
|
1643 abort ();
|
|
|
1644 /* #### This should never happen now that we automatically create
|
|
|
1645 temporary charsets as necessary. We should probably remove
|
|
|
1646 this code. --ben */
|
|
|
1647 iso->invalid_designated[reg] = 1;
|
|
|
1648 iso->charset[reg] = Vcharset_ascii;
|
|
|
1649 iso->esc = ISO_ESC_DESIGNATE;
|
|
|
1650 *flags &= ISO_STATE_LOCK;
|
|
|
1651 iso->output_literally = 1;
|
|
|
1652 if (iso->switched_dir_and_no_valid_charset_yet)
|
|
|
1653 {
|
|
|
1654 /* We encountered a switch-direction followed by an
|
|
|
1655 invalid designation. Ensure that the switch-direction
|
|
|
1656 gets outputted; otherwise it will probably get eaten
|
|
|
1657 when the text is written out again. */
|
|
|
1658 iso->switched_dir_and_no_valid_charset_yet = 0;
|
|
|
1659 iso->output_direction_sequence = 1;
|
|
|
1660 /* And make sure that the switch-dir going the other
|
|
|
1661 way gets outputted, as well. */
|
|
|
1662 iso->invalid_switch_dir = 1;
|
|
|
1663 }
|
|
|
1664 return 1;
|
|
|
1665 }
|
|
|
1666 /* This function is called with CODESYS equal to nil when
|
|
|
1667 doing coding-system detection. */
|
|
|
1668 if (!NILP (codesys))
|
|
|
1669 {
|
|
|
1670 charset_conversion_spec_dynarr *dyn =
|
|
|
1671 XCODING_SYSTEM_ISO2022_INPUT_CONV (codesys);
|
|
|
1672
|
|
|
1673 if (dyn)
|
|
|
1674 {
|
|
|
1675 int i;
|
|
|
1676
|
|
|
1677 for (i = 0; i < Dynarr_length (dyn); i++)
|
|
|
1678 {
|
|
|
1679 struct charset_conversion_spec *spec = Dynarr_atp (dyn, i);
|
|
|
1680 if (EQ (cs, spec->from_charset))
|
|
|
1681 cs = spec->to_charset;
|
|
|
1682 }
|
|
|
1683 }
|
|
|
1684 }
|
|
|
1685
|
|
|
1686 iso->charset[reg] = cs;
|
|
|
1687 iso->esc = ISO_ESC_DESIGNATE;
|
|
|
1688 *flags &= ISO_STATE_LOCK;
|
|
|
1689 if (iso->invalid_designated[reg])
|
|
|
1690 {
|
|
|
1691 iso->invalid_designated[reg] = 0;
|
|
|
1692 iso->output_literally = 1;
|
|
|
1693 }
|
|
|
1694 if (iso->switched_dir_and_no_valid_charset_yet)
|
|
|
1695 iso->switched_dir_and_no_valid_charset_yet = 0;
|
|
|
1696 return 1;
|
|
|
1697 }
|
|
|
1698
|
|
|
1699 /* If FLAGS is a null pointer or specifies right-to-left motion,
|
|
|
1700 output a switch-dir-to-left-to-right sequence to DST.
|
|
|
1701 Also update FLAGS if it is not a null pointer.
|
|
|
1702 If INTERNAL_P is set, we are outputting in internal format and
|
|
|
1703 need to handle the CSI differently. */
|
|
|
1704
|
|
|
1705 static void
|
|
|
1706 restore_left_to_right_direction (Lisp_Object codesys,
|
|
|
1707 unsigned_char_dynarr *dst,
|
|
|
1708 unsigned int *flags,
|
|
|
1709 int internal_p)
|
|
|
1710 {
|
|
|
1711 if (!flags || (*flags & ISO_STATE_R2L))
|
|
|
1712 {
|
|
|
1713 if (XCODING_SYSTEM_ISO2022_SEVEN (codesys))
|
|
|
1714 {
|
|
|
1715 Dynarr_add (dst, ISO_CODE_ESC);
|
|
|
1716 Dynarr_add (dst, '[');
|
|
|
1717 }
|
|
|
1718 else if (internal_p)
|
|
|
1719 DECODE_ADD_BINARY_CHAR (ISO_CODE_CSI, dst);
|
|
|
1720 else
|
|
|
1721 Dynarr_add (dst, ISO_CODE_CSI);
|
|
|
1722 Dynarr_add (dst, '0');
|
|
|
1723 Dynarr_add (dst, ']');
|
|
|
1724 if (flags)
|
|
|
1725 *flags &= ~ISO_STATE_R2L;
|
|
|
1726 }
|
|
|
1727 }
|
|
|
1728
|
|
|
1729 /* If FLAGS is a null pointer or specifies a direction different from
|
|
|
1730 DIRECTION (which should be either CHARSET_RIGHT_TO_LEFT or
|
|
|
1731 CHARSET_LEFT_TO_RIGHT), output the appropriate switch-dir escape
|
|
|
1732 sequence to DST. Also update FLAGS if it is not a null pointer.
|
|
|
1733 If INTERNAL_P is set, we are outputting in internal format and
|
|
|
1734 need to handle the CSI differently. */
|
|
|
1735
|
|
|
1736 static void
|
|
|
1737 ensure_correct_direction (int direction, Lisp_Object codesys,
|
|
|
1738 unsigned_char_dynarr *dst, unsigned int *flags,
|
|
|
1739 int internal_p)
|
|
|
1740 {
|
|
|
1741 if ((!flags || (*flags & ISO_STATE_R2L)) &&
|
|
|
1742 direction == CHARSET_LEFT_TO_RIGHT)
|
|
|
1743 restore_left_to_right_direction (codesys, dst, flags, internal_p);
|
|
|
1744 else if (!XCODING_SYSTEM_ISO2022_NO_ISO6429 (codesys)
|
|
|
1745 && (!flags || !(*flags & ISO_STATE_R2L)) &&
|
|
|
1746 direction == CHARSET_RIGHT_TO_LEFT)
|
|
|
1747 {
|
|
|
1748 if (XCODING_SYSTEM_ISO2022_SEVEN (codesys))
|
|
|
1749 {
|
|
|
1750 Dynarr_add (dst, ISO_CODE_ESC);
|
|
|
1751 Dynarr_add (dst, '[');
|
|
|
1752 }
|
|
|
1753 else if (internal_p)
|
|
|
1754 DECODE_ADD_BINARY_CHAR (ISO_CODE_CSI, dst);
|
|
|
1755 else
|
|
|
1756 Dynarr_add (dst, ISO_CODE_CSI);
|
|
|
1757 Dynarr_add (dst, '2');
|
|
|
1758 Dynarr_add (dst, ']');
|
|
|
1759 if (flags)
|
|
|
1760 *flags |= ISO_STATE_R2L;
|
|
|
1761 }
|
|
|
1762 }
|
|
|
1763
|
|
|
1764 /* Convert ISO2022-format data to internal format. */
|
|
|
1765
|
|
|
1766 static Bytecount
|
|
|
1767 iso2022_decode (struct coding_stream *str, const UExtbyte *src,
|
|
|
1768 unsigned_char_dynarr *dst, Bytecount n)
|
|
|
1769 {
|
|
|
1770 unsigned int ch = str->ch;
|
|
|
1771 #ifdef ENABLE_COMPOSITE_CHARS
|
|
|
1772 unsigned_char_dynarr *real_dst = dst;
|
|
|
1773 #endif
|
|
|
1774 struct iso2022_coding_stream *data =
|
|
|
1775 CODING_STREAM_TYPE_DATA (str, iso2022);
|
|
|
1776 unsigned int flags = data->flags;
|
|
|
1777 Bytecount orign = n;
|
|
|
1778
|
|
|
1779 #ifdef ENABLE_COMPOSITE_CHARS
|
|
|
1780 if (flags & ISO_STATE_COMPOSITE)
|
|
|
1781 dst = data->composite_chars;
|
|
|
1782 #endif /* ENABLE_COMPOSITE_CHARS */
|
|
|
1783
|
|
|
1784 while (n--)
|
|
|
1785 {
|
|
|
1786 UExtbyte c = *src++;
|
|
|
1787 if (flags & ISO_STATE_ESCAPE)
|
|
|
1788 { /* Within ESC sequence */
|
|
|
1789 int retval = parse_iso2022_esc (str->codesys, data,
|
|
|
1790 c, &flags, 1);
|
|
|
1791
|
|
|
1792 if (retval)
|
|
|
1793 {
|
|
|
1794 switch (data->esc)
|
|
|
1795 {
|
|
|
1796 #ifdef ENABLE_COMPOSITE_CHARS
|
|
|
1797 case ISO_ESC_START_COMPOSITE:
|
|
|
1798 if (data->composite_chars)
|
|
|
1799 Dynarr_reset (data->composite_chars);
|
|
|
1800 else
|
|
|
1801 data->composite_chars = Dynarr_new (unsigned_char);
|
|
|
1802 dst = data->composite_chars;
|
|
|
1803 break;
|
|
|
1804 case ISO_ESC_END_COMPOSITE:
|
|
|
1805 {
|
|
867
|
1806 Ibyte comstr[MAX_ICHAR_LEN];
|
|
771
|
1807 Bytecount len;
|
|
867
|
1808 Ichar emch = lookup_composite_char (Dynarr_atp (dst, 0),
|
|
771
|
1809 Dynarr_length (dst));
|
|
|
1810 dst = real_dst;
|
|
867
|
1811 len = set_itext_ichar (comstr, emch);
|
|
771
|
1812 Dynarr_add_many (dst, comstr, len);
|
|
|
1813 break;
|
|
|
1814 }
|
|
|
1815 #else
|
|
|
1816 case ISO_ESC_START_COMPOSITE:
|
|
|
1817 {
|
|
867
|
1818 Ibyte comstr[MAX_ICHAR_LEN];
|
|
771
|
1819 Bytecount len;
|
|
867
|
1820 Ichar emch = make_ichar (Vcharset_composite, c - '0' + ' ',
|
|
771
|
1821 0);
|
|
867
|
1822 len = set_itext_ichar (comstr, emch);
|
|
771
|
1823 Dynarr_add_many (dst, comstr, len);
|
|
|
1824 break;
|
|
|
1825 }
|
|
|
1826 #endif /* ENABLE_COMPOSITE_CHARS */
|
|
|
1827
|
|
|
1828 case ISO_ESC_LITERAL:
|
|
|
1829 DECODE_ADD_BINARY_CHAR (c, dst);
|
|
|
1830 break;
|
|
|
1831
|
|
|
1832 default:
|
|
|
1833 /* Everything else handled already */
|
|
|
1834 break;
|
|
|
1835 }
|
|
|
1836 }
|
|
|
1837
|
|
|
1838 /* Attempted error recovery. */
|
|
|
1839 if (data->output_direction_sequence)
|
|
|
1840 ensure_correct_direction (flags & ISO_STATE_R2L ?
|
|
|
1841 CHARSET_RIGHT_TO_LEFT :
|
|
|
1842 CHARSET_LEFT_TO_RIGHT,
|
|
|
1843 str->codesys, dst, 0, 1);
|
|
|
1844 /* More error recovery. */
|
|
|
1845 if (!retval || data->output_literally)
|
|
|
1846 {
|
|
|
1847 /* Output the (possibly invalid) sequence */
|
|
|
1848 int i;
|
|
|
1849 for (i = 0; i < data->esc_bytes_index; i++)
|
|
|
1850 DECODE_ADD_BINARY_CHAR (data->esc_bytes[i], dst);
|
|
|
1851 flags &= ISO_STATE_LOCK;
|
|
|
1852 if (!retval)
|
|
|
1853 n++, src--;/* Repeat the loop with the same character. */
|
|
|
1854 else
|
|
|
1855 {
|
|
|
1856 /* No sense in reprocessing the final byte of the
|
|
|
1857 escape sequence; it could mess things up anyway.
|
|
|
1858 Just add it now. */
|
|
|
1859 DECODE_ADD_BINARY_CHAR (c, dst);
|
|
|
1860 }
|
|
|
1861 }
|
|
|
1862 ch = 0;
|
|
|
1863 }
|
|
826
|
1864 else if (byte_c0_p (c) || byte_c1_p (c))
|
|
771
|
1865 { /* Control characters */
|
|
|
1866
|
|
|
1867 /***** Error-handling *****/
|
|
|
1868
|
|
|
1869 /* If we were in the middle of a character, dump out the
|
|
|
1870 partial character. */
|
|
|
1871 DECODE_OUTPUT_PARTIAL_CHAR (ch, dst);
|
|
|
1872
|
|
|
1873 /* If we just saw a single-shift character, dump it out.
|
|
|
1874 This may dump out the wrong sort of single-shift character,
|
|
|
1875 but least it will give an indication that something went
|
|
|
1876 wrong. */
|
|
|
1877 if (flags & ISO_STATE_SS2)
|
|
|
1878 {
|
|
|
1879 DECODE_ADD_BINARY_CHAR (ISO_CODE_SS2, dst);
|
|
|
1880 flags &= ~ISO_STATE_SS2;
|
|
|
1881 }
|
|
|
1882 if (flags & ISO_STATE_SS3)
|
|
|
1883 {
|
|
|
1884 DECODE_ADD_BINARY_CHAR (ISO_CODE_SS3, dst);
|
|
|
1885 flags &= ~ISO_STATE_SS3;
|
|
|
1886 }
|
|
|
1887
|
|
|
1888 /***** Now handle the control characters. *****/
|
|
|
1889
|
|
|
1890 flags &= ISO_STATE_LOCK;
|
|
|
1891
|
|
|
1892 if (!parse_iso2022_esc (str->codesys, data, c, &flags, 1))
|
|
|
1893 DECODE_ADD_BINARY_CHAR (c, dst);
|
|
|
1894 }
|
|
|
1895 else
|
|
|
1896 { /* Graphic characters */
|
|
|
1897 Lisp_Object charset;
|
|
|
1898 int lb;
|
|
|
1899 int reg;
|
|
|
1900
|
|
|
1901 /* Now determine the charset. */
|
|
|
1902 reg = ((flags & ISO_STATE_SS2) ? 2
|
|
|
1903 : (flags & ISO_STATE_SS3) ? 3
|
|
826
|
1904 : !byte_ascii_p (c) ? data->register_right
|
|
771
|
1905 : data->register_left);
|
|
|
1906 charset = data->charset[reg];
|
|
|
1907
|
|
|
1908 /* Error checking: */
|
|
|
1909 if (! CHARSETP (charset)
|
|
|
1910 || data->invalid_designated[reg]
|
|
|
1911 || (((c & 0x7F) == ' ' || (c & 0x7F) == ISO_CODE_DEL)
|
|
|
1912 && XCHARSET_CHARS (charset) == 94))
|
|
|
1913 /* Mrmph. We are trying to invoke a register that has no
|
|
|
1914 or an invalid charset in it, or trying to add a character
|
|
|
1915 outside the range of the charset. Insert that char literally
|
|
|
1916 to preserve it for the output. */
|
|
|
1917 {
|
|
|
1918 DECODE_OUTPUT_PARTIAL_CHAR (ch, dst);
|
|
|
1919 DECODE_ADD_BINARY_CHAR (c, dst);
|
|
|
1920 }
|
|
|
1921
|
|
|
1922 else
|
|
|
1923 {
|
|
|
1924 /* Things are probably hunky-dorey. */
|
|
|
1925
|
|
|
1926 /* Fetch reverse charset, maybe. */
|
|
|
1927 if (((flags & ISO_STATE_R2L) &&
|
|
|
1928 XCHARSET_DIRECTION (charset) == CHARSET_LEFT_TO_RIGHT)
|
|
|
1929 ||
|
|
|
1930 (!(flags & ISO_STATE_R2L) &&
|
|
|
1931 XCHARSET_DIRECTION (charset) == CHARSET_RIGHT_TO_LEFT))
|
|
|
1932 {
|
|
|
1933 Lisp_Object new_charset =
|
|
|
1934 XCHARSET_REVERSE_DIRECTION_CHARSET (charset);
|
|
|
1935 if (!NILP (new_charset))
|
|
|
1936 charset = new_charset;
|
|
|
1937 }
|
|
|
1938
|
|
|
1939 lb = XCHARSET_LEADING_BYTE (charset);
|
|
|
1940 switch (XCHARSET_REP_BYTES (charset))
|
|
|
1941 {
|
|
|
1942 case 1: /* ASCII */
|
|
|
1943 DECODE_OUTPUT_PARTIAL_CHAR (ch, dst);
|
|
|
1944 Dynarr_add (dst, c & 0x7F);
|
|
|
1945 break;
|
|
|
1946
|
|
|
1947 case 2: /* one-byte official */
|
|
|
1948 DECODE_OUTPUT_PARTIAL_CHAR (ch, dst);
|
|
|
1949 Dynarr_add (dst, lb);
|
|
|
1950 Dynarr_add (dst, c | 0x80);
|
|
|
1951 break;
|
|
|
1952
|
|
|
1953 case 3: /* one-byte private or two-byte official */
|
|
|
1954 if (XCHARSET_PRIVATE_P (charset))
|
|
|
1955 {
|
|
|
1956 DECODE_OUTPUT_PARTIAL_CHAR (ch, dst);
|
|
|
1957 Dynarr_add (dst, PRE_LEADING_BYTE_PRIVATE_1);
|
|
|
1958 Dynarr_add (dst, lb);
|
|
|
1959 Dynarr_add (dst, c | 0x80);
|
|
|
1960 }
|
|
|
1961 else
|
|
|
1962 {
|
|
|
1963 if (ch)
|
|
|
1964 {
|
|
|
1965 Dynarr_add (dst, lb);
|
|
|
1966 Dynarr_add (dst, ch | 0x80);
|
|
|
1967 Dynarr_add (dst, c | 0x80);
|
|
|
1968 ch = 0;
|
|
|
1969 }
|
|
|
1970 else
|
|
|
1971 ch = c;
|
|
|
1972 }
|
|
|
1973 break;
|
|
|
1974
|
|
|
1975 default: /* two-byte private */
|
|
|
1976 if (ch)
|
|
|
1977 {
|
|
|
1978 Dynarr_add (dst, PRE_LEADING_BYTE_PRIVATE_2);
|
|
|
1979 Dynarr_add (dst, lb);
|
|
|
1980 Dynarr_add (dst, ch | 0x80);
|
|
|
1981 Dynarr_add (dst, c | 0x80);
|
|
|
1982 ch = 0;
|
|
|
1983 }
|
|
|
1984 else
|
|
|
1985 ch = c;
|
|
|
1986 }
|
|
|
1987 }
|
|
|
1988
|
|
|
1989 if (!ch)
|
|
|
1990 flags &= ISO_STATE_LOCK;
|
|
|
1991 }
|
|
|
1992
|
|
|
1993 }
|
|
|
1994
|
|
|
1995 if (str->eof)
|
|
|
1996 DECODE_OUTPUT_PARTIAL_CHAR (ch, dst);
|
|
|
1997
|
|
|
1998 data->flags = flags;
|
|
|
1999 str->ch = ch;
|
|
|
2000 return orign;
|
|
|
2001 }
|
|
|
2002
|
|
|
2003
|
|
|
2004 /***** ISO2022 encoder *****/
|
|
|
2005
|
|
|
2006 /* Designate CHARSET into register REG. */
|
|
|
2007
|
|
|
2008 static void
|
|
|
2009 iso2022_designate (Lisp_Object charset, int reg,
|
|
|
2010 struct coding_stream *str, unsigned_char_dynarr *dst)
|
|
|
2011 {
|
|
|
2012 static const char inter94[] = "()*+";
|
|
|
2013 static const char inter96[] = ",-./";
|
|
|
2014 int type;
|
|
|
2015 unsigned char final;
|
|
|
2016 struct iso2022_coding_stream *data =
|
|
|
2017 CODING_STREAM_TYPE_DATA (str, iso2022);
|
|
|
2018 Lisp_Object old_charset = data->charset[reg];
|
|
|
2019
|
|
|
2020 data->charset[reg] = charset;
|
|
|
2021 if (!CHARSETP (charset))
|
|
|
2022 /* charset might be an initial nil or t. */
|
|
|
2023 return;
|
|
|
2024 type = XCHARSET_TYPE (charset);
|
|
|
2025 final = XCHARSET_FINAL (charset);
|
|
|
2026 if (!data->force_charset_on_output[reg] &&
|
|
|
2027 CHARSETP (old_charset) &&
|
|
|
2028 XCHARSET_TYPE (old_charset) == type &&
|
|
|
2029 XCHARSET_FINAL (old_charset) == final)
|
|
|
2030 return;
|
|
|
2031
|
|
|
2032 data->force_charset_on_output[reg] = 0;
|
|
|
2033
|
|
|
2034 {
|
|
|
2035 charset_conversion_spec_dynarr *dyn =
|
|
|
2036 XCODING_SYSTEM_ISO2022_OUTPUT_CONV (str->codesys);
|
|
|
2037
|
|
|
2038 if (dyn)
|
|
|
2039 {
|
|
|
2040 int i;
|
|
|
2041
|
|
|
2042 for (i = 0; i < Dynarr_length (dyn); i++)
|
|
|
2043 {
|
|
|
2044 struct charset_conversion_spec *spec = Dynarr_atp (dyn, i);
|
|
|
2045 if (EQ (charset, spec->from_charset))
|
|
|
2046 charset = spec->to_charset;
|
|
|
2047 }
|
|
|
2048 }
|
|
|
2049 }
|
|
|
2050
|
|
|
2051 Dynarr_add (dst, ISO_CODE_ESC);
|
|
|
2052 switch (type)
|
|
|
2053 {
|
|
|
2054 case CHARSET_TYPE_94:
|
|
|
2055 Dynarr_add (dst, inter94[reg]);
|
|
|
2056 break;
|
|
|
2057 case CHARSET_TYPE_96:
|
|
|
2058 Dynarr_add (dst, inter96[reg]);
|
|
|
2059 break;
|
|
|
2060 case CHARSET_TYPE_94X94:
|
|
|
2061 Dynarr_add (dst, '$');
|
|
|
2062 if (reg != 0
|
|
|
2063 || !(XCODING_SYSTEM_ISO2022_SHORT (str->codesys))
|
|
|
2064 || final < '@'
|
|
|
2065 || final > 'B')
|
|
|
2066 Dynarr_add (dst, inter94[reg]);
|
|
|
2067 break;
|
|
|
2068 case CHARSET_TYPE_96X96:
|
|
|
2069 Dynarr_add (dst, '$');
|
|
|
2070 Dynarr_add (dst, inter96[reg]);
|
|
|
2071 break;
|
|
|
2072 }
|
|
|
2073 Dynarr_add (dst, final);
|
|
|
2074 }
|
|
|
2075
|
|
|
2076 static void
|
|
|
2077 ensure_normal_shift (struct coding_stream *str, unsigned_char_dynarr *dst)
|
|
|
2078 {
|
|
|
2079 struct iso2022_coding_stream *data =
|
|
|
2080 CODING_STREAM_TYPE_DATA (str, iso2022);
|
|
|
2081
|
|
|
2082 if (data->register_left != 0)
|
|
|
2083 {
|
|
|
2084 Dynarr_add (dst, ISO_CODE_SI);
|
|
|
2085 data->register_left = 0;
|
|
|
2086 }
|
|
|
2087 }
|
|
|
2088
|
|
|
2089 static void
|
|
|
2090 ensure_shift_out (struct coding_stream *str, unsigned_char_dynarr *dst)
|
|
|
2091 {
|
|
|
2092 struct iso2022_coding_stream *data =
|
|
|
2093 CODING_STREAM_TYPE_DATA (str, iso2022);
|
|
|
2094
|
|
|
2095 if (data->register_left != 1)
|
|
|
2096 {
|
|
|
2097 Dynarr_add (dst, ISO_CODE_SO);
|
|
|
2098 data->register_left = 1;
|
|
|
2099 }
|
|
|
2100 }
|
|
|
2101
|
|
|
2102 /* Convert internally-formatted data to ISO2022 format. */
|
|
|
2103
|
|
|
2104 static Bytecount
|
|
867
|
2105 iso2022_encode (struct coding_stream *str, const Ibyte *src,
|
|
771
|
2106 unsigned_char_dynarr *dst, Bytecount n)
|
|
|
2107 {
|
|
|
2108 unsigned char charmask;
|
|
867
|
2109 Ibyte c;
|
|
771
|
2110 unsigned char char_boundary;
|
|
|
2111 unsigned int ch = str->ch;
|
|
|
2112 Lisp_Object codesys = str->codesys;
|
|
|
2113 int i;
|
|
|
2114 Lisp_Object charset;
|
|
|
2115 int half;
|
|
|
2116 struct iso2022_coding_stream *data =
|
|
|
2117 CODING_STREAM_TYPE_DATA (str, iso2022);
|
|
|
2118 unsigned int flags = data->flags;
|
|
|
2119 Bytecount orign = n;
|
|
|
2120
|
|
|
2121 #ifdef ENABLE_COMPOSITE_CHARS
|
|
|
2122 /* flags for handling composite chars. We do a little switcheroo
|
|
|
2123 on the source while we're outputting the composite char. */
|
|
|
2124 Bytecount saved_n = 0;
|
|
867
|
2125 const Ibyte *saved_src = NULL;
|
|
771
|
2126 int in_composite = 0;
|
|
|
2127 #endif /* ENABLE_COMPOSITE_CHARS */
|
|
|
2128
|
|
|
2129 char_boundary = data->current_char_boundary;
|
|
|
2130 charset = data->current_charset;
|
|
|
2131 half = data->current_half;
|
|
|
2132
|
|
|
2133 #ifdef ENABLE_COMPOSITE_CHARS
|
|
|
2134 back_to_square_n:
|
|
|
2135 #endif
|
|
|
2136 while (n--)
|
|
|
2137 {
|
|
|
2138 c = *src++;
|
|
|
2139
|
|
826
|
2140 if (byte_ascii_p (c))
|
|
771
|
2141 { /* Processing ASCII character */
|
|
|
2142 ch = 0;
|
|
|
2143
|
|
|
2144 restore_left_to_right_direction (codesys, dst, &flags, 0);
|
|
|
2145
|
|
|
2146 /* Make sure G0 contains ASCII */
|
|
|
2147 if ((c > ' ' && c < ISO_CODE_DEL) ||
|
|
|
2148 !XCODING_SYSTEM_ISO2022_NO_ASCII_CNTL (codesys))
|
|
|
2149 {
|
|
|
2150 ensure_normal_shift (str, dst);
|
|
|
2151 iso2022_designate (Vcharset_ascii, 0, str, dst);
|
|
|
2152 }
|
|
|
2153
|
|
|
2154 /* If necessary, restore everything to the default state
|
|
|
2155 at end-of-line */
|
|
|
2156 if (!(XCODING_SYSTEM_ISO2022_NO_ASCII_EOL (codesys)))
|
|
|
2157 {
|
|
|
2158 /* NOTE: CRLF encoding happens *BEFORE* other encoding.
|
|
|
2159 Thus, even though we're working with internal-format
|
|
|
2160 data, there may be CR's or CRLF sequences representing
|
|
|
2161 newlines. */
|
|
|
2162 if (c == '\r' || (c == '\n' && !(flags & ISO_STATE_CR)))
|
|
|
2163 {
|
|
|
2164 restore_left_to_right_direction (codesys, dst, &flags, 0);
|
|
|
2165
|
|
|
2166 ensure_normal_shift (str, dst);
|
|
|
2167
|
|
|
2168 for (i = 0; i < 4; i++)
|
|
|
2169 {
|
|
|
2170 Lisp_Object initial_charset =
|
|
|
2171 XCODING_SYSTEM_ISO2022_INITIAL_CHARSET (codesys, i);
|
|
|
2172 iso2022_designate (initial_charset, i, str, dst);
|
|
|
2173 }
|
|
|
2174 }
|
|
|
2175 if (c == '\r')
|
|
|
2176 flags |= ISO_STATE_CR;
|
|
|
2177 else
|
|
|
2178 flags &= ~ISO_STATE_CR;
|
|
|
2179 }
|
|
|
2180
|
|
|
2181 if (XCODING_SYSTEM_ISO2022_ESCAPE_QUOTED (codesys)
|
|
|
2182 && fit_to_be_escape_quoted (c))
|
|
|
2183 Dynarr_add (dst, ISO_CODE_ESC);
|
|
|
2184 Dynarr_add (dst, c);
|
|
|
2185 char_boundary = 1;
|
|
|
2186 }
|
|
|
2187
|
|
867
|
2188 else if (ibyte_leading_byte_p (c) || ibyte_leading_byte_p (ch))
|
|
771
|
2189 { /* Processing Leading Byte */
|
|
|
2190 ch = 0;
|
|
826
|
2191 charset = charset_by_leading_byte (c);
|
|
|
2192 if (leading_byte_prefix_p (c))
|
|
771
|
2193 ch = c;
|
|
|
2194 else if (!EQ (charset, Vcharset_control_1)
|
|
|
2195 && !EQ (charset, Vcharset_composite))
|
|
|
2196 {
|
|
|
2197 int reg;
|
|
|
2198
|
|
|
2199 ensure_correct_direction (XCHARSET_DIRECTION (charset),
|
|
|
2200 codesys, dst, &flags, 0);
|
|
|
2201
|
|
|
2202 /* Now determine which register to use. */
|
|
|
2203 reg = -1;
|
|
|
2204 for (i = 0; i < 4; i++)
|
|
|
2205 {
|
|
|
2206 if (EQ (charset, data->charset[i]) ||
|
|
|
2207 EQ (charset,
|
|
|
2208 XCODING_SYSTEM_ISO2022_INITIAL_CHARSET (codesys, i)))
|
|
|
2209 {
|
|
|
2210 reg = i;
|
|
|
2211 break;
|
|
|
2212 }
|
|
|
2213 }
|
|
|
2214
|
|
|
2215 if (reg == -1)
|
|
|
2216 {
|
|
|
2217 if (XCHARSET_GRAPHIC (charset) != 0)
|
|
|
2218 {
|
|
|
2219 if (!NILP (data->charset[1]) &&
|
|
|
2220 (!XCODING_SYSTEM_ISO2022_SEVEN (codesys) ||
|
|
|
2221 XCODING_SYSTEM_ISO2022_LOCK_SHIFT (codesys)))
|
|
|
2222 reg = 1;
|
|
|
2223 else if (!NILP (data->charset[2]))
|
|
|
2224 reg = 2;
|
|
|
2225 else if (!NILP (data->charset[3]))
|
|
|
2226 reg = 3;
|
|
|
2227 else
|
|
|
2228 reg = 0;
|
|
|
2229 }
|
|
|
2230 else
|
|
|
2231 reg = 0;
|
|
|
2232 }
|
|
|
2233
|
|
|
2234 iso2022_designate (charset, reg, str, dst);
|
|
|
2235
|
|
|
2236 /* Now invoke that register. */
|
|
|
2237 switch (reg)
|
|
|
2238 {
|
|
|
2239 case 0:
|
|
|
2240 ensure_normal_shift (str, dst);
|
|
|
2241 half = 0;
|
|
|
2242 break;
|
|
|
2243
|
|
|
2244 case 1:
|
|
|
2245 if (XCODING_SYSTEM_ISO2022_SEVEN (codesys))
|
|
|
2246 {
|
|
|
2247 ensure_shift_out (str, dst);
|
|
|
2248 half = 0;
|
|
|
2249 }
|
|
|
2250 else
|
|
|
2251 half = 1;
|
|
|
2252 break;
|
|
|
2253
|
|
|
2254 case 2:
|
|
|
2255 if (XCODING_SYSTEM_ISO2022_SEVEN (str->codesys))
|
|
|
2256 {
|
|
|
2257 Dynarr_add (dst, ISO_CODE_ESC);
|
|
|
2258 Dynarr_add (dst, 'N');
|
|
|
2259 half = 0;
|
|
|
2260 }
|
|
|
2261 else
|
|
|
2262 {
|
|
|
2263 Dynarr_add (dst, ISO_CODE_SS2);
|
|
|
2264 half = 1;
|
|
|
2265 }
|
|
|
2266 break;
|
|
|
2267
|
|
|
2268 case 3:
|
|
|
2269 if (XCODING_SYSTEM_ISO2022_SEVEN (str->codesys))
|
|
|
2270 {
|
|
|
2271 Dynarr_add (dst, ISO_CODE_ESC);
|
|
|
2272 Dynarr_add (dst, 'O');
|
|
|
2273 half = 0;
|
|
|
2274 }
|
|
|
2275 else
|
|
|
2276 {
|
|
|
2277 Dynarr_add (dst, ISO_CODE_SS3);
|
|
|
2278 half = 1;
|
|
|
2279 }
|
|
|
2280 break;
|
|
|
2281
|
|
|
2282 default:
|
|
|
2283 abort ();
|
|
|
2284 }
|
|
|
2285 }
|
|
|
2286 char_boundary = 0;
|
|
|
2287 }
|
|
|
2288 else
|
|
|
2289 { /* Processing Non-ASCII character */
|
|
|
2290 charmask = (half == 0 ? 0x7F : 0xFF);
|
|
|
2291 char_boundary = 1;
|
|
|
2292 if (EQ (charset, Vcharset_control_1))
|
|
|
2293 {
|
|
|
2294 if (XCODING_SYSTEM_ISO2022_ESCAPE_QUOTED (codesys)
|
|
|
2295 && fit_to_be_escape_quoted (c))
|
|
|
2296 Dynarr_add (dst, ISO_CODE_ESC);
|
|
|
2297 /* you asked for it ... */
|
|
|
2298 Dynarr_add (dst, c - 0x20);
|
|
|
2299 }
|
|
|
2300 #ifndef ENABLE_COMPOSITE_CHARS
|
|
|
2301 else if (EQ (charset, Vcharset_composite))
|
|
|
2302 {
|
|
|
2303 if (c >= 160 || c <= 164) /* Someone might have stuck in
|
|
|
2304 something else */
|
|
|
2305 {
|
|
|
2306 Dynarr_add (dst, ISO_CODE_ESC);
|
|
|
2307 Dynarr_add (dst, c - 160 + '0');
|
|
|
2308 }
|
|
|
2309 }
|
|
|
2310 #endif
|
|
|
2311 else
|
|
|
2312 {
|
|
|
2313 switch (XCHARSET_REP_BYTES (charset))
|
|
|
2314 {
|
|
|
2315 case 2:
|
|
|
2316 Dynarr_add (dst, c & charmask);
|
|
|
2317 break;
|
|
|
2318 case 3:
|
|
|
2319 if (XCHARSET_PRIVATE_P (charset))
|
|
|
2320 {
|
|
|
2321 Dynarr_add (dst, c & charmask);
|
|
|
2322 ch = 0;
|
|
|
2323 }
|
|
|
2324 else if (ch)
|
|
|
2325 {
|
|
|
2326 #ifdef ENABLE_COMPOSITE_CHARS
|
|
|
2327 if (EQ (charset, Vcharset_composite))
|
|
|
2328 {
|
|
|
2329 if (in_composite)
|
|
|
2330 {
|
|
|
2331 /* #### Bother! We don't know how to
|
|
|
2332 handle this yet. */
|
|
|
2333 Dynarr_add (dst, '~');
|
|
|
2334 }
|
|
|
2335 else
|
|
|
2336 {
|
|
867
|
2337 Ichar emch = make_ichar (Vcharset_composite,
|
|
771
|
2338 ch & 0x7F, c & 0x7F);
|
|
|
2339 Lisp_Object lstr = composite_char_string (emch);
|
|
|
2340 saved_n = n;
|
|
|
2341 saved_src = src;
|
|
|
2342 in_composite = 1;
|
|
|
2343 src = XSTRING_DATA (lstr);
|
|
|
2344 n = XSTRING_LENGTH (lstr);
|
|
|
2345 Dynarr_add (dst, ISO_CODE_ESC);
|
|
|
2346 Dynarr_add (dst, '0'); /* start composing */
|
|
|
2347 }
|
|
|
2348 }
|
|
|
2349 else
|
|
|
2350 #endif /* ENABLE_COMPOSITE_CHARS */
|
|
|
2351 {
|
|
|
2352 Dynarr_add (dst, ch & charmask);
|
|
|
2353 Dynarr_add (dst, c & charmask);
|
|
|
2354 }
|
|
|
2355 ch = 0;
|
|
|
2356 }
|
|
|
2357 else
|
|
|
2358 {
|
|
|
2359 ch = c;
|
|
|
2360 char_boundary = 0;
|
|
|
2361 }
|
|
|
2362 break;
|
|
|
2363 case 4:
|
|
|
2364 if (ch)
|
|
|
2365 {
|
|
|
2366 Dynarr_add (dst, ch & charmask);
|
|
|
2367 Dynarr_add (dst, c & charmask);
|
|
|
2368 ch = 0;
|
|
|
2369 }
|
|
|
2370 else
|
|
|
2371 {
|
|
|
2372 ch = c;
|
|
|
2373 char_boundary = 0;
|
|
|
2374 }
|
|
|
2375 break;
|
|
|
2376 default:
|
|
|
2377 abort ();
|
|
|
2378 }
|
|
|
2379 }
|
|
|
2380 }
|
|
|
2381 }
|
|
|
2382
|
|
|
2383 #ifdef ENABLE_COMPOSITE_CHARS
|
|
|
2384 if (in_composite)
|
|
|
2385 {
|
|
|
2386 n = saved_n;
|
|
|
2387 src = saved_src;
|
|
|
2388 in_composite = 0;
|
|
|
2389 Dynarr_add (dst, ISO_CODE_ESC);
|
|
|
2390 Dynarr_add (dst, '1'); /* end composing */
|
|
|
2391 goto back_to_square_n; /* Wheeeeeeeee ..... */
|
|
|
2392 }
|
|
|
2393 #endif /* ENABLE_COMPOSITE_CHARS */
|
|
|
2394
|
|
|
2395 if (char_boundary && str->eof)
|
|
|
2396 {
|
|
|
2397 restore_left_to_right_direction (codesys, dst, &flags, 0);
|
|
|
2398 ensure_normal_shift (str, dst);
|
|
|
2399 for (i = 0; i < 4; i++)
|
|
|
2400 {
|
|
|
2401 Lisp_Object initial_charset =
|
|
|
2402 XCODING_SYSTEM_ISO2022_INITIAL_CHARSET (codesys, i);
|
|
|
2403 iso2022_designate (initial_charset, i, str, dst);
|
|
|
2404 }
|
|
|
2405 }
|
|
|
2406
|
|
|
2407 data->flags = flags;
|
|
|
2408 str->ch = ch;
|
|
|
2409 data->current_char_boundary = char_boundary;
|
|
|
2410 data->current_charset = charset;
|
|
|
2411 data->current_half = half;
|
|
|
2412
|
|
|
2413 /* Verbum caro factum est! */
|
|
|
2414 return orign;
|
|
|
2415 }
|
|
|
2416
|
|
|
2417 static Bytecount
|
|
|
2418 iso2022_convert (struct coding_stream *str,
|
|
|
2419 const UExtbyte *src,
|
|
|
2420 unsigned_char_dynarr *dst, Bytecount n)
|
|
|
2421 {
|
|
|
2422 if (str->direction == CODING_DECODE)
|
|
|
2423 return iso2022_decode (str, src, dst, n);
|
|
|
2424 else
|
|
|
2425 return iso2022_encode (str, src, dst, n);
|
|
|
2426 }
|
|
|
2427
|
|
|
2428 static void
|
|
|
2429 iso2022_mark (Lisp_Object codesys)
|
|
|
2430 {
|
|
|
2431 int i;
|
|
|
2432
|
|
|
2433 for (i = 0; i < 4; i++)
|
|
|
2434 mark_object (XCODING_SYSTEM_ISO2022_INITIAL_CHARSET (codesys, i));
|
|
|
2435 if (XCODING_SYSTEM_ISO2022_INPUT_CONV (codesys))
|
|
|
2436 {
|
|
|
2437 for (i = 0;
|
|
|
2438 i < Dynarr_length (XCODING_SYSTEM_ISO2022_INPUT_CONV (codesys));
|
|
|
2439 i++)
|
|
|
2440 {
|
|
|
2441 struct charset_conversion_spec *ccs =
|
|
|
2442 Dynarr_atp (XCODING_SYSTEM_ISO2022_INPUT_CONV (codesys), i);
|
|
|
2443 mark_object (ccs->from_charset);
|
|
|
2444 mark_object (ccs->to_charset);
|
|
|
2445 }
|
|
|
2446 }
|
|
|
2447 if (XCODING_SYSTEM_ISO2022_OUTPUT_CONV (codesys))
|
|
|
2448 {
|
|
|
2449 for (i = 0;
|
|
|
2450 i < Dynarr_length (XCODING_SYSTEM_ISO2022_OUTPUT_CONV (codesys));
|
|
|
2451 i++)
|
|
|
2452 {
|
|
|
2453 struct charset_conversion_spec *ccs =
|
|
|
2454 Dynarr_atp (XCODING_SYSTEM_ISO2022_OUTPUT_CONV (codesys), i);
|
|
|
2455 mark_object (ccs->from_charset);
|
|
|
2456 mark_object (ccs->to_charset);
|
|
|
2457 }
|
|
|
2458 }
|
|
|
2459 }
|
|
|
2460
|
|
|
2461 static void
|
|
|
2462 iso2022_finalize (Lisp_Object cs)
|
|
|
2463 {
|
|
|
2464 if (XCODING_SYSTEM_ISO2022_INPUT_CONV (cs))
|
|
|
2465 {
|
|
|
2466 Dynarr_free (XCODING_SYSTEM_ISO2022_INPUT_CONV (cs));
|
|
|
2467 XCODING_SYSTEM_ISO2022_INPUT_CONV (cs) = 0;
|
|
|
2468 }
|
|
|
2469 if (XCODING_SYSTEM_ISO2022_OUTPUT_CONV (cs))
|
|
|
2470 {
|
|
|
2471 Dynarr_free (XCODING_SYSTEM_ISO2022_OUTPUT_CONV (cs));
|
|
|
2472 XCODING_SYSTEM_ISO2022_OUTPUT_CONV (cs) = 0;
|
|
|
2473 }
|
|
|
2474 }
|
|
|
2475
|
|
|
2476 /* Given a list of charset conversion specs as specified in a Lisp
|
|
|
2477 program, parse it into STORE_HERE. */
|
|
|
2478
|
|
|
2479 static void
|
|
|
2480 parse_charset_conversion_specs (charset_conversion_spec_dynarr *store_here,
|
|
|
2481 Lisp_Object spec_list)
|
|
|
2482 {
|
|
|
2483 Lisp_Object rest;
|
|
|
2484
|
|
|
2485 EXTERNAL_LIST_LOOP (rest, spec_list)
|
|
|
2486 {
|
|
|
2487 Lisp_Object car = XCAR (rest);
|
|
|
2488 Lisp_Object from, to;
|
|
|
2489 struct charset_conversion_spec spec;
|
|
|
2490
|
|
|
2491 if (!CONSP (car) || !CONSP (XCDR (car)) || !NILP (XCDR (XCDR (car))))
|
|
|
2492 invalid_argument ("Invalid charset conversion spec", car);
|
|
|
2493 from = Fget_charset (XCAR (car));
|
|
|
2494 to = Fget_charset (XCAR (XCDR (car)));
|
|
|
2495 if (XCHARSET_TYPE (from) != XCHARSET_TYPE (to))
|
|
|
2496 invalid_operation_2
|
|
|
2497 ("Attempted conversion between different charset types",
|
|
|
2498 from, to);
|
|
|
2499 spec.from_charset = from;
|
|
|
2500 spec.to_charset = to;
|
|
|
2501
|
|
|
2502 Dynarr_add (store_here, spec);
|
|
|
2503 }
|
|
|
2504 }
|
|
|
2505
|
|
|
2506 /* Given a dynarr LOAD_HERE of internally-stored charset conversion
|
|
|
2507 specs, return the equivalent as the Lisp programmer would see it.
|
|
|
2508
|
|
|
2509 If LOAD_HERE is 0, return Qnil. */
|
|
|
2510
|
|
|
2511 static Lisp_Object
|
|
|
2512 unparse_charset_conversion_specs (charset_conversion_spec_dynarr *load_here,
|
|
|
2513 int names)
|
|
|
2514 {
|
|
|
2515 int i;
|
|
|
2516 Lisp_Object result;
|
|
|
2517
|
|
|
2518 if (!load_here)
|
|
|
2519 return Qnil;
|
|
|
2520 for (i = 0, result = Qnil; i < Dynarr_length (load_here); i++)
|
|
|
2521 {
|
|
|
2522 struct charset_conversion_spec *ccs = Dynarr_atp (load_here, i);
|
|
|
2523 if (names)
|
|
|
2524 result = Fcons (list2 (XCHARSET_NAME (ccs->from_charset),
|
|
|
2525 XCHARSET_NAME (ccs->to_charset)), result);
|
|
|
2526 else
|
|
|
2527 result = Fcons (list2 (ccs->from_charset, ccs->to_charset), result);
|
|
|
2528 }
|
|
|
2529
|
|
|
2530 return Fnreverse (result);
|
|
|
2531 }
|
|
|
2532
|
|
|
2533 static int
|
|
|
2534 iso2022_putprop (Lisp_Object codesys,
|
|
|
2535 Lisp_Object key,
|
|
|
2536 Lisp_Object value)
|
|
|
2537 {
|
|
|
2538 #define FROB_INITIAL_CHARSET(charset_num) \
|
|
|
2539 XCODING_SYSTEM_ISO2022_INITIAL_CHARSET (codesys, charset_num) = \
|
|
|
2540 ((EQ (value, Qt) || EQ (value, Qnil)) ? value : Fget_charset (value))
|
|
|
2541
|
|
|
2542 if (EQ (key, Qcharset_g0)) FROB_INITIAL_CHARSET (0);
|
|
|
2543 else if (EQ (key, Qcharset_g1)) FROB_INITIAL_CHARSET (1);
|
|
|
2544 else if (EQ (key, Qcharset_g2)) FROB_INITIAL_CHARSET (2);
|
|
|
2545 else if (EQ (key, Qcharset_g3)) FROB_INITIAL_CHARSET (3);
|
|
|
2546
|
|
|
2547 #define FROB_FORCE_CHARSET(charset_num) \
|
|
|
2548 XCODING_SYSTEM_ISO2022_FORCE_CHARSET_ON_OUTPUT (codesys, charset_num) = \
|
|
|
2549 !NILP (value)
|
|
|
2550
|
|
|
2551 else if (EQ (key, Qforce_g0_on_output)) FROB_FORCE_CHARSET (0);
|
|
|
2552 else if (EQ (key, Qforce_g1_on_output)) FROB_FORCE_CHARSET (1);
|
|
|
2553 else if (EQ (key, Qforce_g2_on_output)) FROB_FORCE_CHARSET (2);
|
|
|
2554 else if (EQ (key, Qforce_g3_on_output)) FROB_FORCE_CHARSET (3);
|
|
|
2555
|
|
|
2556 #define FROB_BOOLEAN_PROPERTY(prop) \
|
|
|
2557 XCODING_SYSTEM_ISO2022_##prop (codesys) = !NILP (value)
|
|
|
2558
|
|
|
2559 else if (EQ (key, Qshort)) FROB_BOOLEAN_PROPERTY (SHORT);
|
|
|
2560 else if (EQ (key, Qno_ascii_eol)) FROB_BOOLEAN_PROPERTY (NO_ASCII_EOL);
|
|
|
2561 else if (EQ (key, Qno_ascii_cntl)) FROB_BOOLEAN_PROPERTY (NO_ASCII_CNTL);
|
|
|
2562 else if (EQ (key, Qseven)) FROB_BOOLEAN_PROPERTY (SEVEN);
|
|
|
2563 else if (EQ (key, Qlock_shift)) FROB_BOOLEAN_PROPERTY (LOCK_SHIFT);
|
|
|
2564 else if (EQ (key, Qno_iso6429)) FROB_BOOLEAN_PROPERTY (NO_ISO6429);
|
|
|
2565 else if (EQ (key, Qescape_quoted)) FROB_BOOLEAN_PROPERTY (ESCAPE_QUOTED);
|
|
|
2566
|
|
|
2567 else if (EQ (key, Qinput_charset_conversion))
|
|
|
2568 {
|
|
|
2569 XCODING_SYSTEM_ISO2022_INPUT_CONV (codesys) =
|
|
|
2570 Dynarr_new (charset_conversion_spec);
|
|
|
2571 parse_charset_conversion_specs
|
|
|
2572 (XCODING_SYSTEM_ISO2022_INPUT_CONV (codesys), value);
|
|
|
2573 }
|
|
|
2574 else if (EQ (key, Qoutput_charset_conversion))
|
|
|
2575 {
|
|
|
2576 XCODING_SYSTEM_ISO2022_OUTPUT_CONV (codesys) =
|
|
|
2577 Dynarr_new (charset_conversion_spec);
|
|
|
2578 parse_charset_conversion_specs
|
|
|
2579 (XCODING_SYSTEM_ISO2022_OUTPUT_CONV (codesys), value);
|
|
|
2580 }
|
|
|
2581 else
|
|
|
2582 return 0;
|
|
|
2583
|
|
|
2584 return 1;
|
|
|
2585 }
|
|
|
2586
|
|
|
2587 static void
|
|
|
2588 iso2022_finalize_coding_stream (struct coding_stream *str)
|
|
|
2589 {
|
|
|
2590 #ifdef ENABLE_COMPOSITE_CHARS
|
|
|
2591 struct iso2022_coding_stream *data =
|
|
|
2592 CODING_STREAM_TYPE_DATA (str, iso2022);
|
|
|
2593
|
|
|
2594 if (data->composite_chars)
|
|
|
2595 Dynarr_free (data->composite_chars);
|
|
|
2596 #endif
|
|
|
2597 }
|
|
|
2598
|
|
|
2599 static void
|
|
|
2600 iso2022_init (Lisp_Object codesys)
|
|
|
2601 {
|
|
|
2602 int i;
|
|
|
2603 for (i = 0; i < 4; i++)
|
|
|
2604 XCODING_SYSTEM_ISO2022_INITIAL_CHARSET (codesys, i) = Qnil;
|
|
|
2605 }
|
|
|
2606
|
|
|
2607 static Lisp_Object
|
|
|
2608 coding_system_charset (Lisp_Object coding_system, int gnum)
|
|
|
2609 {
|
|
|
2610 Lisp_Object cs
|
|
|
2611 = XCODING_SYSTEM_ISO2022_INITIAL_CHARSET (coding_system, gnum);
|
|
|
2612
|
|
|
2613 return CHARSETP (cs) ? XCHARSET_NAME (cs) : Qnil;
|
|
|
2614 }
|
|
|
2615
|
|
|
2616 static Lisp_Object
|
|
|
2617 iso2022_getprop (Lisp_Object coding_system, Lisp_Object prop)
|
|
|
2618 {
|
|
|
2619 if (EQ (prop, Qcharset_g0))
|
|
|
2620 return coding_system_charset (coding_system, 0);
|
|
|
2621 else if (EQ (prop, Qcharset_g1))
|
|
|
2622 return coding_system_charset (coding_system, 1);
|
|
|
2623 else if (EQ (prop, Qcharset_g2))
|
|
|
2624 return coding_system_charset (coding_system, 2);
|
|
|
2625 else if (EQ (prop, Qcharset_g3))
|
|
|
2626 return coding_system_charset (coding_system, 3);
|
|
|
2627
|
|
|
2628 #define FORCE_CHARSET(charset_num) \
|
|
|
2629 (XCODING_SYSTEM_ISO2022_FORCE_CHARSET_ON_OUTPUT \
|
|
|
2630 (coding_system, charset_num) ? Qt : Qnil)
|
|
|
2631
|
|
|
2632 else if (EQ (prop, Qforce_g0_on_output))
|
|
|
2633 return FORCE_CHARSET (0);
|
|
|
2634 else if (EQ (prop, Qforce_g1_on_output))
|
|
|
2635 return FORCE_CHARSET (1);
|
|
|
2636 else if (EQ (prop, Qforce_g2_on_output))
|
|
|
2637 return FORCE_CHARSET (2);
|
|
|
2638 else if (EQ (prop, Qforce_g3_on_output))
|
|
|
2639 return FORCE_CHARSET (3);
|
|
|
2640
|
|
|
2641 #define LISP_BOOLEAN(prop) \
|
|
|
2642 (XCODING_SYSTEM_ISO2022_##prop (coding_system) ? Qt : Qnil)
|
|
|
2643
|
|
|
2644 else if (EQ (prop, Qshort)) return LISP_BOOLEAN (SHORT);
|
|
|
2645 else if (EQ (prop, Qno_ascii_eol)) return LISP_BOOLEAN (NO_ASCII_EOL);
|
|
|
2646 else if (EQ (prop, Qno_ascii_cntl)) return LISP_BOOLEAN (NO_ASCII_CNTL);
|
|
|
2647 else if (EQ (prop, Qseven)) return LISP_BOOLEAN (SEVEN);
|
|
|
2648 else if (EQ (prop, Qlock_shift)) return LISP_BOOLEAN (LOCK_SHIFT);
|
|
|
2649 else if (EQ (prop, Qno_iso6429)) return LISP_BOOLEAN (NO_ISO6429);
|
|
|
2650 else if (EQ (prop, Qescape_quoted)) return LISP_BOOLEAN (ESCAPE_QUOTED);
|
|
|
2651
|
|
|
2652 else if (EQ (prop, Qinput_charset_conversion))
|
|
|
2653 return
|
|
|
2654 unparse_charset_conversion_specs
|
|
|
2655 (XCODING_SYSTEM_ISO2022_INPUT_CONV (coding_system), 0);
|
|
|
2656 else if (EQ (prop, Qoutput_charset_conversion))
|
|
|
2657 return
|
|
|
2658 unparse_charset_conversion_specs
|
|
|
2659 (XCODING_SYSTEM_ISO2022_OUTPUT_CONV (coding_system), 0);
|
|
|
2660 else
|
|
|
2661 return Qunbound;
|
|
|
2662 }
|
|
|
2663
|
|
|
2664 static void
|
|
|
2665 iso2022_print (Lisp_Object cs, Lisp_Object printcharfun, int escapeflag)
|
|
|
2666 {
|
|
|
2667 int i;
|
|
|
2668
|
|
826
|
2669 write_c_string (printcharfun, "(");
|
|
771
|
2670 for (i = 0; i < 4; i++)
|
|
|
2671 {
|
|
|
2672 Lisp_Object charset = coding_system_charset (cs, i);
|
|
|
2673 if (i > 0)
|
|
826
|
2674 write_c_string (printcharfun, ", ");
|
|
771
|
2675 write_fmt_string (printcharfun, "g%d=", i);
|
|
800
|
2676 print_internal (CHARSETP (charset) ? XCHARSET_NAME (charset) : charset, printcharfun, 0);
|
|
771
|
2677 if (XCODING_SYSTEM_ISO2022_FORCE_CHARSET_ON_OUTPUT (cs, i))
|
|
826
|
2678 write_c_string (printcharfun, "(force)");
|
|
771
|
2679 }
|
|
|
2680
|
|
800
|
2681 #define FROB(prop) \
|
|
|
2682 if (!NILP (iso2022_getprop (cs, prop))) \
|
|
|
2683 { \
|
|
|
2684 write_fmt_string (printcharfun, ", %s", prop); \
|
|
771
|
2685 }
|
|
|
2686
|
|
|
2687 FROB (Qshort);
|
|
|
2688 FROB (Qno_ascii_eol);
|
|
|
2689 FROB (Qno_ascii_cntl);
|
|
|
2690 FROB (Qseven);
|
|
|
2691 FROB (Qlock_shift);
|
|
|
2692 FROB (Qno_iso6429);
|
|
|
2693 FROB (Qescape_quoted);
|
|
|
2694
|
|
|
2695 {
|
|
|
2696 Lisp_Object val =
|
|
|
2697 unparse_charset_conversion_specs
|
|
|
2698 (XCODING_SYSTEM_ISO2022_INPUT_CONV (cs), 1);
|
|
|
2699 if (!NILP (val))
|
|
|
2700 {
|
|
800
|
2701 write_fmt_string_lisp (printcharfun, ", input-charset-conversion=%s", 1, val);
|
|
771
|
2702 }
|
|
|
2703 val =
|
|
|
2704 unparse_charset_conversion_specs
|
|
|
2705 (XCODING_SYSTEM_ISO2022_OUTPUT_CONV (cs), 1);
|
|
|
2706 if (!NILP (val))
|
|
|
2707 {
|
|
800
|
2708 write_fmt_string_lisp (printcharfun, ", output-charset-conversion=%s", 1, val);
|
|
771
|
2709 }
|
|
826
|
2710 write_c_string (printcharfun, ")");
|
|
771
|
2711 }
|
|
|
2712 }
|
|
|
2713
|
|
|
2714
|
|
|
2715 /************************************************************************/
|
|
|
2716 /* ISO2022 detector */
|
|
|
2717 /************************************************************************/
|
|
|
2718
|
|
|
2719 DEFINE_DETECTOR (iso2022);
|
|
|
2720 /* ISO2022 system using only seven-bit bytes, no locking shift */
|
|
|
2721 DEFINE_DETECTOR_CATEGORY (iso2022, iso_7);
|
|
|
2722 /* ISO2022 system using eight-bit bytes, no locking shift, no single shift,
|
|
|
2723 using designation to switch charsets */
|
|
|
2724 DEFINE_DETECTOR_CATEGORY (iso2022, iso_8_designate);
|
|
|
2725 /* ISO2022 system using eight-bit bytes, no locking shift, no designation
|
|
|
2726 sequences, one-dimension characters in the upper half. */
|
|
|
2727 DEFINE_DETECTOR_CATEGORY (iso2022, iso_8_1);
|
|
|
2728 /* ISO2022 system using eight-bit bytes, no locking shift, no designation
|
|
|
2729 sequences, two-dimension characters in the upper half. */
|
|
|
2730 DEFINE_DETECTOR_CATEGORY (iso2022, iso_8_2);
|
|
|
2731 /* ISO2022 system using locking shift */
|
|
|
2732 DEFINE_DETECTOR_CATEGORY (iso2022, iso_lock_shift);
|
|
|
2733
|
|
|
2734 struct iso2022_detector
|
|
|
2735 {
|
|
|
2736 int initted;
|
|
|
2737 struct iso2022_coding_stream *iso;
|
|
|
2738 unsigned int flags;
|
|
|
2739
|
|
|
2740 /* for keeping temporary track of high-byte groups */
|
|
|
2741 int high_byte_count;
|
|
|
2742 unsigned int saw_single_shift_just_now:1;
|
|
|
2743
|
|
|
2744 /* running state; we set the likelihoods at the end */
|
|
|
2745 unsigned int seen_high_byte:1;
|
|
|
2746 unsigned int seen_single_shift:1;
|
|
|
2747 unsigned int seen_locking_shift:1;
|
|
|
2748 unsigned int seen_designate:1;
|
|
|
2749 unsigned int bad_single_byte_sequences;
|
|
|
2750 unsigned int bad_multibyte_escape_sequences;
|
|
|
2751 unsigned int good_multibyte_escape_sequences;
|
|
|
2752 int even_high_byte_groups;
|
|
985
|
2753 int longest_even_high_byte;
|
|
771
|
2754 int odd_high_byte_groups;
|
|
|
2755 };
|
|
|
2756
|
|
|
2757 static void
|
|
|
2758 iso2022_detect (struct detection_state *st, const UExtbyte *src,
|
|
|
2759 Bytecount n)
|
|
|
2760 {
|
|
|
2761 Bytecount orign = n;
|
|
|
2762 struct iso2022_detector *data = DETECTION_STATE_DATA (st, iso2022);
|
|
|
2763
|
|
|
2764 /* #### There are serious deficiencies in the recognition mechanism
|
|
|
2765 here. This needs to be much smarter if it's going to cut it.
|
|
|
2766 The sequence "\xff\x0f" is currently detected as LOCK_SHIFT while
|
|
|
2767 it should be detected as Latin-1.
|
|
|
2768 All the ISO2022 stuff in this file should be synced up with the
|
|
|
2769 code from FSF Emacs-21.0, in which Mule should be more or less stable.
|
|
|
2770 Perhaps we should wait till R2L works in FSF Emacs? */
|
|
|
2771
|
|
|
2772 /* We keep track of running state on our own, and set the categories at the
|
|
|
2773 end; that way we can reflect the correct state each time we finish, but
|
|
|
2774 not get confused by those results the next time around. */
|
|
|
2775
|
|
|
2776 if (!data->initted)
|
|
|
2777 {
|
|
|
2778 xzero (*data);
|
|
|
2779 data->iso = xnew_and_zero (struct iso2022_coding_stream);
|
|
|
2780 reset_iso2022_decode (Qnil, data->iso);
|
|
|
2781 data->initted = 1;
|
|
|
2782 }
|
|
|
2783
|
|
|
2784 while (n--)
|
|
|
2785 {
|
|
|
2786 UExtbyte c = *src++;
|
|
|
2787 if (c >= 0x80)
|
|
|
2788 data->seen_high_byte = 1;
|
|
|
2789 if (c >= 0xA0)
|
|
|
2790 data->high_byte_count++;
|
|
|
2791 else
|
|
|
2792 {
|
|
|
2793 if (data->high_byte_count &&
|
|
|
2794 !data->saw_single_shift_just_now)
|
|
|
2795 {
|
|
|
2796 if (data->high_byte_count & 1)
|
|
|
2797 data->odd_high_byte_groups++;
|
|
|
2798 else
|
|
985
|
2799 {
|
|
|
2800 data->even_high_byte_groups++;
|
|
|
2801 if (data->longest_even_high_byte < data->high_byte_count)
|
|
|
2802 data->longest_even_high_byte = data->high_byte_count;
|
|
|
2803 }
|
|
771
|
2804 }
|
|
|
2805 data->high_byte_count = 0;
|
|
|
2806 data->saw_single_shift_just_now = 0;
|
|
|
2807 }
|
|
|
2808 if (!(data->flags & ISO_STATE_ESCAPE)
|
|
826
|
2809 && (byte_c0_p (c) || byte_c1_p (c)))
|
|
771
|
2810 { /* control chars */
|
|
|
2811 switch (c)
|
|
|
2812 {
|
|
|
2813 /* Allow and ignore control characters that you might
|
|
|
2814 reasonably see in a text file */
|
|
|
2815 case '\r':
|
|
|
2816 case '\n':
|
|
|
2817 case '\t':
|
|
|
2818 case 7: /* bell */
|
|
|
2819 case 8: /* backspace */
|
|
|
2820 case 11: /* vertical tab */
|
|
|
2821 case 12: /* form feed */
|
|
|
2822 case 26: /* MS-DOS C-z junk */
|
|
|
2823 case 31: /* '^_' -- for info */
|
|
|
2824 goto label_continue_loop;
|
|
|
2825
|
|
|
2826 default:
|
|
|
2827 break;
|
|
|
2828 }
|
|
|
2829 }
|
|
|
2830
|
|
826
|
2831 if ((data->flags & ISO_STATE_ESCAPE) || byte_c0_p (c)
|
|
|
2832 || byte_c1_p (c))
|
|
771
|
2833 {
|
|
|
2834 switch (parse_iso2022_esc (Qnil, data->iso, c,
|
|
|
2835 &data->flags, 0))
|
|
|
2836 {
|
|
|
2837 case 1: /* done */
|
|
|
2838 if (data->iso->esc_bytes_index > 0)
|
|
|
2839 data->good_multibyte_escape_sequences++;
|
|
|
2840 switch (data->iso->esc)
|
|
|
2841 {
|
|
|
2842 case ISO_ESC_DESIGNATE:
|
|
|
2843 data->seen_designate = 1;
|
|
|
2844 break;
|
|
|
2845 case ISO_ESC_LOCKING_SHIFT:
|
|
|
2846 data->seen_locking_shift = 1;
|
|
|
2847 break;
|
|
|
2848 case ISO_ESC_SINGLE_SHIFT:
|
|
|
2849 data->saw_single_shift_just_now = 1;
|
|
|
2850 data->seen_single_shift = 1;
|
|
|
2851 break;
|
|
|
2852 default:
|
|
|
2853 break;
|
|
|
2854 }
|
|
|
2855 break;
|
|
|
2856
|
|
|
2857 case -1: /* not done */
|
|
|
2858 break;
|
|
|
2859
|
|
|
2860 case 0: /* error */
|
|
|
2861 if (data->iso->esc == ISO_ESC_NOTHING)
|
|
|
2862 data->bad_single_byte_sequences++;
|
|
|
2863 else
|
|
|
2864 data->bad_multibyte_escape_sequences++;
|
|
|
2865 }
|
|
|
2866 }
|
|
|
2867 label_continue_loop:;
|
|
|
2868 }
|
|
|
2869
|
|
985
|
2870 if (data->high_byte_count &&
|
|
|
2871 !data->saw_single_shift_just_now)
|
|
|
2872 {
|
|
|
2873 if (data->high_byte_count & 1)
|
|
|
2874 data->odd_high_byte_groups++;
|
|
|
2875 else
|
|
|
2876 {
|
|
|
2877 data->even_high_byte_groups++;
|
|
|
2878 if (data->longest_even_high_byte < data->high_byte_count)
|
|
|
2879 data->longest_even_high_byte = data->high_byte_count;
|
|
|
2880 }
|
|
|
2881 }
|
|
|
2882
|
|
771
|
2883 if (data->bad_multibyte_escape_sequences > 2 ||
|
|
|
2884 (data->bad_multibyte_escape_sequences > 0 &&
|
|
|
2885 data->good_multibyte_escape_sequences /
|
|
|
2886 data->bad_multibyte_escape_sequences < 10))
|
|
|
2887 /* Just making it up ... */
|
|
|
2888 SET_DET_RESULTS (st, iso2022, DET_NEARLY_IMPOSSIBLE);
|
|
|
2889 else if (data->bad_single_byte_sequences > 5 ||
|
|
|
2890 (data->bad_single_byte_sequences > 0 &&
|
|
|
2891 (data->good_multibyte_escape_sequences +
|
|
|
2892 data->even_high_byte_groups +
|
|
|
2893 data->odd_high_byte_groups) /
|
|
|
2894 data->bad_single_byte_sequences < 10))
|
|
|
2895 SET_DET_RESULTS (st, iso2022, DET_SOMEWHAT_UNLIKELY);
|
|
|
2896 else if (data->seen_locking_shift)
|
|
|
2897 {
|
|
|
2898 SET_DET_RESULTS (st, iso2022, DET_QUITE_IMPROBABLE);
|
|
|
2899 DET_RESULT (st, iso_lock_shift) = DET_QUITE_PROBABLE;
|
|
|
2900 }
|
|
|
2901 else if (!data->seen_high_byte)
|
|
|
2902 {
|
|
|
2903 SET_DET_RESULTS (st, iso2022, DET_SOMEWHAT_UNLIKELY);
|
|
|
2904 if (data->good_multibyte_escape_sequences)
|
|
|
2905 DET_RESULT (st, iso_7) = DET_QUITE_PROBABLE;
|
|
|
2906 else if (data->seen_single_shift)
|
|
|
2907 DET_RESULT (st, iso_7) = DET_SOMEWHAT_LIKELY;
|
|
|
2908 else
|
|
|
2909 {
|
|
|
2910 /* If we've just seen pure 7-bit data, no escape sequences,
|
|
|
2911 then we can't give much likelihood; but if we've seen enough
|
|
|
2912 of this data, we can assume some unlikelihood of any 8-bit
|
|
|
2913 encoding */
|
|
|
2914 if (orign + st->bytes_seen >= 1000)
|
|
|
2915 DET_RESULT (st, iso_7) = DET_AS_LIKELY_AS_UNLIKELY;
|
|
|
2916 else
|
|
|
2917 SET_DET_RESULTS (st, iso2022, DET_AS_LIKELY_AS_UNLIKELY);
|
|
|
2918 }
|
|
|
2919 }
|
|
|
2920 else if (data->seen_designate)
|
|
|
2921 {
|
|
|
2922 SET_DET_RESULTS (st, iso2022, DET_QUITE_IMPROBABLE);
|
|
|
2923 if (data->seen_single_shift)
|
|
|
2924 /* #### Does this really make sense? */
|
|
|
2925 DET_RESULT (st, iso_8_designate) = DET_SOMEWHAT_UNLIKELY;
|
|
|
2926 else
|
|
|
2927 DET_RESULT (st, iso_8_designate) = DET_QUITE_PROBABLE;
|
|
|
2928 }
|
|
|
2929 else if (data->odd_high_byte_groups > 0 &&
|
|
|
2930 data->even_high_byte_groups == 0)
|
|
|
2931 {
|
|
|
2932 SET_DET_RESULTS (st, iso2022, DET_SOMEWHAT_UNLIKELY);
|
|
|
2933 if (data->seen_single_shift)
|
|
|
2934 DET_RESULT (st, iso_8_1) = DET_QUITE_PROBABLE;
|
|
|
2935 else
|
|
|
2936 DET_RESULT (st, iso_8_1) = DET_SOMEWHAT_LIKELY;
|
|
|
2937 }
|
|
|
2938 else if (data->odd_high_byte_groups == 0 &&
|
|
|
2939 data->even_high_byte_groups > 0)
|
|
|
2940 {
|
|
985
|
2941 #if 0
|
|
771
|
2942 SET_DET_RESULTS (st, iso2022, DET_SOMEWHAT_UNLIKELY);
|
|
|
2943 if (data->even_high_byte_groups > 10)
|
|
|
2944 {
|
|
|
2945 if (data->seen_single_shift)
|
|
|
2946 DET_RESULT (st, iso_8_2) = DET_QUITE_PROBABLE;
|
|
|
2947 else
|
|
|
2948 DET_RESULT (st, iso_8_2) = DET_SOMEWHAT_LIKELY;
|
|
|
2949 if (data->even_high_byte_groups < 50)
|
|
|
2950 DET_RESULT (st, iso_8_1) = DET_SOMEWHAT_UNLIKELY;
|
|
|
2951 /* else it stays at quite improbable */
|
|
|
2952 }
|
|
985
|
2953 #else
|
|
|
2954 SET_DET_RESULTS (st, iso2022, DET_SOMEWHAT_UNLIKELY);
|
|
|
2955 if (data->seen_single_shift)
|
|
|
2956 DET_RESULT (st, iso_8_2) = DET_QUITE_PROBABLE;
|
|
|
2957 else if (data->even_high_byte_groups > 10)
|
|
|
2958 DET_RESULT (st, iso_8_2) = DET_SOMEWHAT_LIKELY;
|
|
|
2959 else if (data->longest_even_high_byte > 6)
|
|
|
2960 DET_RESULT (st, iso_8_2) = DET_SLIGHTLY_LIKELY;
|
|
|
2961 #endif
|
|
771
|
2962 }
|
|
|
2963 else if (data->odd_high_byte_groups > 0 &&
|
|
|
2964 data->even_high_byte_groups > 0)
|
|
|
2965 SET_DET_RESULTS (st, iso2022, DET_SOMEWHAT_UNLIKELY);
|
|
|
2966 else
|
|
|
2967 SET_DET_RESULTS (st, iso2022, DET_AS_LIKELY_AS_UNLIKELY);
|
|
|
2968 }
|
|
|
2969
|
|
|
2970 static void
|
|
|
2971 iso2022_finalize_detection_state (struct detection_state *st)
|
|
|
2972 {
|
|
|
2973 struct iso2022_detector *data = DETECTION_STATE_DATA (st, iso2022);
|
|
|
2974 if (data->iso)
|
|
|
2975 xfree (data->iso);
|
|
|
2976 }
|
|
|
2977
|
|
|
2978
|
|
|
2979 /************************************************************************/
|
|
|
2980 /* CCL methods */
|
|
|
2981 /************************************************************************/
|
|
|
2982
|
|
|
2983 /* Converter written in CCL. */
|
|
|
2984
|
|
|
2985 struct ccl_coding_system
|
|
|
2986 {
|
|
|
2987 /* For a CCL coding system, these specify the CCL programs used for
|
|
|
2988 decoding (input) and encoding (output). */
|
|
|
2989 Lisp_Object decode;
|
|
|
2990 Lisp_Object encode;
|
|
|
2991 };
|
|
|
2992
|
|
|
2993 #define CODING_SYSTEM_CCL_DECODE(codesys) \
|
|
|
2994 (CODING_SYSTEM_TYPE_DATA (codesys, ccl)->decode)
|
|
|
2995 #define CODING_SYSTEM_CCL_ENCODE(codesys) \
|
|
|
2996 (CODING_SYSTEM_TYPE_DATA (codesys, ccl)->encode)
|
|
|
2997 #define XCODING_SYSTEM_CCL_DECODE(codesys) \
|
|
|
2998 CODING_SYSTEM_CCL_DECODE (XCODING_SYSTEM (codesys))
|
|
|
2999 #define XCODING_SYSTEM_CCL_ENCODE(codesys) \
|
|
|
3000 CODING_SYSTEM_CCL_ENCODE (XCODING_SYSTEM (codesys))
|
|
|
3001
|
|
|
3002 struct ccl_coding_stream
|
|
|
3003 {
|
|
|
3004 /* state of the running CCL program */
|
|
|
3005 struct ccl_program ccl;
|
|
|
3006 };
|
|
|
3007
|
|
1204
|
3008 static const struct memory_description ccl_coding_system_description[] = {
|
|
|
3009 { XD_LISP_OBJECT, offsetof (struct ccl_coding_system, decode) },
|
|
|
3010 { XD_LISP_OBJECT, offsetof (struct ccl_coding_system, encode) },
|
|
771
|
3011 { XD_END }
|
|
|
3012 };
|
|
|
3013
|
|
1204
|
3014 DEFINE_CODING_SYSTEM_TYPE_WITH_DATA (ccl);
|
|
|
3015
|
|
771
|
3016 static void
|
|
|
3017 ccl_mark (Lisp_Object codesys)
|
|
|
3018 {
|
|
|
3019 mark_object (XCODING_SYSTEM_CCL_DECODE (codesys));
|
|
|
3020 mark_object (XCODING_SYSTEM_CCL_ENCODE (codesys));
|
|
|
3021 }
|
|
|
3022
|
|
|
3023 static Bytecount
|
|
|
3024 ccl_convert (struct coding_stream *str, const UExtbyte *src,
|
|
|
3025 unsigned_char_dynarr *dst, Bytecount n)
|
|
|
3026 {
|
|
|
3027 struct ccl_coding_stream *data =
|
|
|
3028 CODING_STREAM_TYPE_DATA (str, ccl);
|
|
|
3029 Bytecount orign = n;
|
|
|
3030
|
|
|
3031 data->ccl.last_block = str->eof;
|
|
|
3032 /* When applying a CCL program to a stream, SRC must not be NULL -- this
|
|
|
3033 is a special signal to the driver that read and write operations are
|
|
|
3034 not allowed. The code does not actually look at what SRC points to if
|
|
|
3035 N == 0.
|
|
|
3036 */
|
|
|
3037 ccl_driver (&data->ccl, src ? src : (const unsigned char *) "",
|
|
|
3038 dst, n, 0,
|
|
|
3039 str->direction == CODING_DECODE ? CCL_MODE_DECODING :
|
|
|
3040 CCL_MODE_ENCODING);
|
|
|
3041 return orign;
|
|
|
3042 }
|
|
|
3043
|
|
|
3044 static void
|
|
|
3045 ccl_init_coding_stream (struct coding_stream *str)
|
|
|
3046 {
|
|
|
3047 struct ccl_coding_stream *data =
|
|
|
3048 CODING_STREAM_TYPE_DATA (str, ccl);
|
|
|
3049
|
|
|
3050 setup_ccl_program (&data->ccl,
|
|
|
3051 str->direction == CODING_DECODE ?
|
|
|
3052 XCODING_SYSTEM_CCL_DECODE (str->codesys) :
|
|
|
3053 XCODING_SYSTEM_CCL_ENCODE (str->codesys));
|
|
|
3054 }
|
|
|
3055
|
|
|
3056 static void
|
|
|
3057 ccl_rewind_coding_stream (struct coding_stream *str)
|
|
|
3058 {
|
|
|
3059 ccl_init_coding_stream (str);
|
|
|
3060 }
|
|
|
3061
|
|
|
3062 static void
|
|
|
3063 ccl_init (Lisp_Object codesys)
|
|
|
3064 {
|
|
|
3065 XCODING_SYSTEM_CCL_DECODE (codesys) = Qnil;
|
|
|
3066 XCODING_SYSTEM_CCL_ENCODE (codesys) = Qnil;
|
|
|
3067 }
|
|
|
3068
|
|
|
3069 static int
|
|
|
3070 ccl_putprop (Lisp_Object codesys, Lisp_Object key, Lisp_Object value)
|
|
|
3071 {
|
|
|
3072 Lisp_Object sym;
|
|
|
3073 struct ccl_program test_ccl;
|
|
|
3074 Char_ASCII *suffix;
|
|
|
3075
|
|
|
3076 /* Check key first. */
|
|
|
3077 if (EQ (key, Qdecode))
|
|
|
3078 suffix = "-ccl-decode";
|
|
|
3079 else if (EQ (key, Qencode))
|
|
|
3080 suffix = "-ccl-encode";
|
|
|
3081 else
|
|
|
3082 return 0;
|
|
|
3083
|
|
|
3084 /* If value is vector, register it as a ccl program
|
|
|
3085 associated with a newly created symbol for
|
|
|
3086 backward compatibility.
|
|
|
3087
|
|
|
3088 #### Bogosity alert! Do we really have to do this crap???? --ben */
|
|
|
3089 if (VECTORP (value))
|
|
|
3090 {
|
|
|
3091 sym = Fintern (concat2 (Fsymbol_name (XCODING_SYSTEM_NAME (codesys)),
|
|
|
3092 build_string (suffix)),
|
|
|
3093 Qnil);
|
|
|
3094 Fregister_ccl_program (sym, value);
|
|
|
3095 }
|
|
|
3096 else
|
|
|
3097 {
|
|
|
3098 CHECK_SYMBOL (value);
|
|
|
3099 sym = value;
|
|
|
3100 }
|
|
|
3101 /* check if the given ccl programs are valid. */
|
|
|
3102 if (setup_ccl_program (&test_ccl, sym) < 0)
|
|
|
3103 invalid_argument ("Invalid CCL program", value);
|
|
|
3104
|
|
|
3105 if (EQ (key, Qdecode))
|
|
|
3106 XCODING_SYSTEM_CCL_DECODE (codesys) = sym;
|
|
|
3107 else if (EQ (key, Qencode))
|
|
|
3108 XCODING_SYSTEM_CCL_ENCODE (codesys) = sym;
|
|
|
3109
|
|
|
3110 return 1;
|
|
|
3111 }
|
|
|
3112
|
|
|
3113 static Lisp_Object
|
|
|
3114 ccl_getprop (Lisp_Object coding_system, Lisp_Object prop)
|
|
|
3115 {
|
|
|
3116 if (EQ (prop, Qdecode))
|
|
|
3117 return XCODING_SYSTEM_CCL_DECODE (coding_system);
|
|
|
3118 else if (EQ (prop, Qencode))
|
|
|
3119 return XCODING_SYSTEM_CCL_ENCODE (coding_system);
|
|
|
3120 else
|
|
|
3121 return Qunbound;
|
|
|
3122 }
|
|
|
3123
|
|
|
3124
|
|
|
3125 /************************************************************************/
|
|
|
3126 /* Initialization */
|
|
|
3127 /************************************************************************/
|
|
|
3128
|
|
|
3129 void
|
|
|
3130 syms_of_mule_coding (void)
|
|
|
3131 {
|
|
|
3132 DEFSUBR (Fdecode_shift_jis_char);
|
|
|
3133 DEFSUBR (Fencode_shift_jis_char);
|
|
|
3134 DEFSUBR (Fdecode_big5_char);
|
|
|
3135 DEFSUBR (Fencode_big5_char);
|
|
|
3136
|
|
|
3137 DEFSYMBOL (Qbig5);
|
|
|
3138 DEFSYMBOL (Qshift_jis);
|
|
|
3139 DEFSYMBOL (Qccl);
|
|
|
3140 DEFSYMBOL (Qiso2022);
|
|
|
3141
|
|
|
3142 DEFSYMBOL (Qcharset_g0);
|
|
|
3143 DEFSYMBOL (Qcharset_g1);
|
|
|
3144 DEFSYMBOL (Qcharset_g2);
|
|
|
3145 DEFSYMBOL (Qcharset_g3);
|
|
|
3146 DEFSYMBOL (Qforce_g0_on_output);
|
|
|
3147 DEFSYMBOL (Qforce_g1_on_output);
|
|
|
3148 DEFSYMBOL (Qforce_g2_on_output);
|
|
|
3149 DEFSYMBOL (Qforce_g3_on_output);
|
|
|
3150 DEFSYMBOL (Qno_iso6429);
|
|
|
3151 DEFSYMBOL (Qinput_charset_conversion);
|
|
|
3152 DEFSYMBOL (Qoutput_charset_conversion);
|
|
|
3153
|
|
|
3154 DEFSYMBOL (Qshort);
|
|
|
3155 DEFSYMBOL (Qno_ascii_eol);
|
|
|
3156 DEFSYMBOL (Qno_ascii_cntl);
|
|
|
3157 DEFSYMBOL (Qseven);
|
|
|
3158 DEFSYMBOL (Qlock_shift);
|
|
|
3159
|
|
|
3160 DEFSYMBOL (Qiso_7);
|
|
|
3161 DEFSYMBOL (Qiso_8_designate);
|
|
|
3162 DEFSYMBOL (Qiso_8_1);
|
|
|
3163 DEFSYMBOL (Qiso_8_2);
|
|
|
3164 DEFSYMBOL (Qiso_lock_shift);
|
|
|
3165 }
|
|
|
3166
|
|
|
3167 void
|
|
|
3168 coding_system_type_create_mule_coding (void)
|
|
|
3169 {
|
|
|
3170 INITIALIZE_CODING_SYSTEM_TYPE_WITH_DATA (iso2022, "iso2022-coding-system-p");
|
|
|
3171 CODING_SYSTEM_HAS_METHOD (iso2022, mark);
|
|
|
3172 CODING_SYSTEM_HAS_METHOD (iso2022, convert);
|
|
|
3173 CODING_SYSTEM_HAS_METHOD (iso2022, finalize_coding_stream);
|
|
|
3174 CODING_SYSTEM_HAS_METHOD (iso2022, init_coding_stream);
|
|
|
3175 CODING_SYSTEM_HAS_METHOD (iso2022, rewind_coding_stream);
|
|
|
3176 CODING_SYSTEM_HAS_METHOD (iso2022, init);
|
|
|
3177 CODING_SYSTEM_HAS_METHOD (iso2022, print);
|
|
|
3178 CODING_SYSTEM_HAS_METHOD (iso2022, finalize);
|
|
|
3179 CODING_SYSTEM_HAS_METHOD (iso2022, putprop);
|
|
|
3180 CODING_SYSTEM_HAS_METHOD (iso2022, getprop);
|
|
|
3181
|
|
|
3182 INITIALIZE_DETECTOR (iso2022);
|
|
|
3183 DETECTOR_HAS_METHOD (iso2022, detect);
|
|
|
3184 DETECTOR_HAS_METHOD (iso2022, finalize_detection_state);
|
|
|
3185 INITIALIZE_DETECTOR_CATEGORY (iso2022, iso_7);
|
|
|
3186 INITIALIZE_DETECTOR_CATEGORY (iso2022, iso_8_designate);
|
|
|
3187 INITIALIZE_DETECTOR_CATEGORY (iso2022, iso_8_1);
|
|
|
3188 INITIALIZE_DETECTOR_CATEGORY (iso2022, iso_8_2);
|
|
|
3189 INITIALIZE_DETECTOR_CATEGORY (iso2022, iso_lock_shift);
|
|
|
3190
|
|
|
3191 INITIALIZE_CODING_SYSTEM_TYPE_WITH_DATA (ccl, "ccl-coding-system-p");
|
|
|
3192 CODING_SYSTEM_HAS_METHOD (ccl, mark);
|
|
|
3193 CODING_SYSTEM_HAS_METHOD (ccl, convert);
|
|
|
3194 CODING_SYSTEM_HAS_METHOD (ccl, init);
|
|
|
3195 CODING_SYSTEM_HAS_METHOD (ccl, init_coding_stream);
|
|
|
3196 CODING_SYSTEM_HAS_METHOD (ccl, rewind_coding_stream);
|
|
|
3197 CODING_SYSTEM_HAS_METHOD (ccl, putprop);
|
|
|
3198 CODING_SYSTEM_HAS_METHOD (ccl, getprop);
|
|
|
3199
|
|
|
3200 INITIALIZE_CODING_SYSTEM_TYPE (shift_jis, "shift-jis-coding-system-p");
|
|
|
3201 CODING_SYSTEM_HAS_METHOD (shift_jis, convert);
|
|
|
3202
|
|
|
3203 INITIALIZE_DETECTOR (shift_jis);
|
|
|
3204 DETECTOR_HAS_METHOD (shift_jis, detect);
|
|
|
3205 INITIALIZE_DETECTOR_CATEGORY (shift_jis, shift_jis);
|
|
|
3206
|
|
|
3207 INITIALIZE_CODING_SYSTEM_TYPE (big5, "big5-coding-system-p");
|
|
|
3208 CODING_SYSTEM_HAS_METHOD (big5, convert);
|
|
|
3209
|
|
|
3210 INITIALIZE_DETECTOR (big5);
|
|
|
3211 DETECTOR_HAS_METHOD (big5, detect);
|
|
|
3212 INITIALIZE_DETECTOR_CATEGORY (big5, big5);
|
|
|
3213 }
|
|
|
3214
|
|
|
3215 void
|
|
|
3216 reinit_coding_system_type_create_mule_coding (void)
|
|
|
3217 {
|
|
|
3218 REINITIALIZE_CODING_SYSTEM_TYPE (iso2022);
|
|
|
3219 REINITIALIZE_CODING_SYSTEM_TYPE (ccl);
|
|
|
3220 REINITIALIZE_CODING_SYSTEM_TYPE (shift_jis);
|
|
|
3221 REINITIALIZE_CODING_SYSTEM_TYPE (big5);
|
|
|
3222 }
|
|
|
3223
|
|
|
3224 void
|
|
|
3225 reinit_vars_of_mule_coding (void)
|
|
|
3226 {
|
|
|
3227 }
|
|
|
3228
|
|
|
3229 void
|
|
|
3230 vars_of_mule_coding (void)
|
|
|
3231 {
|
|
|
3232 }
|
