Mercurial > audlegacy
comparison src/libguess/guess.c @ 2313:3149d4b1a9a9 trunk
[svn] - objective-make autodepend fixes
- move all sourcecode into src/ and adjust Makefiles accordingly
author | nenolod |
---|---|
date | Fri, 12 Jan 2007 11:43:40 -0800 |
parents | |
children | b474ecb5bde4 |
comparison
equal
deleted
inserted
replaced
2312:e1a5a66fb9cc | 2313:3149d4b1a9a9 |
---|---|
1 /* | |
2 * This code is derivative of guess.c of Gauche-0.8.3. | |
3 * The following is the original copyright notice. | |
4 */ | |
5 | |
6 /* | |
7 * guess.c - guessing character encoding | |
8 * | |
9 * Copyright (c) 2000-2003 Shiro Kawai, All rights reserved. | |
10 * | |
11 * Redistribution and use in source and binary forms, with or without | |
12 * modification, are permitted provided that the following conditions | |
13 * are met: | |
14 * | |
15 * 1. Redistributions of source code must retain the above copyright | |
16 * notice, this list of conditions and the following disclaimer. | |
17 * | |
18 * 2. Redistributions in binary form must reproduce the above copyright | |
19 * notice, this list of conditions and the following disclaimer in the | |
20 * documentation and/or other materials provided with the distribution. | |
21 * | |
22 * 3. Neither the name of the authors nor the names of its contributors | |
23 * may be used to endorse or promote products derived from this | |
24 * software without specific prior written permission. | |
25 * | |
26 * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS | |
27 * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT | |
28 * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR | |
29 * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT | |
30 * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, | |
31 * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED | |
32 * TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR | |
33 * PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF | |
34 * LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING | |
35 * NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS | |
36 * SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. | |
37 * | |
38 */ | |
39 | |
40 #include "libguess.h" | |
41 #define NULL ((void *)0) | |
42 | |
43 /* take precedence if scores are same. you can customize the order as: */ | |
44 /* ORDER_** &highest, &second, ... &lowest */ | |
45 #define ORDER_JP &utf8, &sjis, &eucj | |
46 #define ORDER_TW &utf8, &big5 | |
47 #define ORDER_CN &utf8, &gb2312, &gb18030 | |
48 #define ORDER_KR &utf8, &euck, &johab | |
49 | |
50 /* workaround for that glib's g_convert can't convert | |
51 properly from UCS-2BE/LE trailing after BOM. */ | |
52 #define WITH_G_CONVERT 1 | |
53 /* #undef WITH_G_CONVERT */ | |
54 | |
55 #ifdef WITH_G_CONVERT | |
56 const char UCS_2BE[] = "UTF-16"; | |
57 const char UCS_2LE[] = "UTF-16"; | |
58 #else | |
59 const char UCS_2BE[] = "UCS-2BE"; | |
60 const char UCS_2LE[] = "UCS-2LE"; | |
61 #endif | |
62 | |
63 /* data types */ | |
64 typedef struct guess_arc_rec | |
65 { | |
66 unsigned int next; /* next state */ | |
67 double score; /* score */ | |
68 } guess_arc; | |
69 | |
70 typedef struct guess_dfa_rec | |
71 { | |
72 signed char (*states)[256]; | |
73 guess_arc *arcs; | |
74 int state; | |
75 double score; | |
76 } guess_dfa; | |
77 | |
78 /* macros */ | |
79 #define DFA_INIT(st, ar) \ | |
80 { st, ar, 0, 1.0 } | |
81 | |
82 #define DFA_NEXT(dfa, ch) \ | |
83 do { \ | |
84 int arc__; \ | |
85 if (dfa.state >= 0) { \ | |
86 arc__ = dfa.states[dfa.state][ch]; \ | |
87 if (arc__ < 0) { \ | |
88 dfa.state = -1; \ | |
89 } else { \ | |
90 dfa.state = dfa.arcs[arc__].next; \ | |
91 dfa.score *= dfa.arcs[arc__].score; \ | |
92 } \ | |
93 } \ | |
94 } while (0) | |
95 | |
96 #define DFA_ALIVE(dfa) (dfa.state >= 0) | |
97 | |
98 /* include DFA table generated by guess.scm */ | |
99 #include "guess_tab.c" | |
100 | |
101 const char *guess_jp(const char *buf, int buflen) | |
102 { | |
103 int i; | |
104 guess_dfa eucj = DFA_INIT(guess_eucj_st, guess_eucj_ar); | |
105 guess_dfa sjis = DFA_INIT(guess_sjis_st, guess_sjis_ar); | |
106 guess_dfa utf8 = DFA_INIT(guess_utf8_st, guess_utf8_ar); | |
107 guess_dfa *top = NULL; | |
108 | |
109 guess_dfa *order[] = { ORDER_JP, NULL }; | |
110 | |
111 for (i = 0; i < buflen; i++) { | |
112 int c = (unsigned char) buf[i]; | |
113 | |
114 /* special treatment of iso-2022 escape sequence */ | |
115 if (c == 0x1b) { | |
116 if (i < buflen - 1) { | |
117 c = (unsigned char) buf[++i]; | |
118 if (c == '$' || c == '(') | |
119 return "ISO-2022-JP"; | |
120 } | |
121 } | |
122 | |
123 /* special treatment of BOM */ | |
124 if (i == 0 && c == 0xff) { | |
125 if (i < buflen - 1) { | |
126 c = (unsigned char) buf[i + 1]; | |
127 if (c == 0xfe) | |
128 return UCS_2LE; | |
129 } | |
130 } | |
131 if (i == 0 && c == 0xfe) { | |
132 if (i < buflen - 1) { | |
133 c = (unsigned char) buf[i + 1]; | |
134 if (c == 0xff) | |
135 return UCS_2BE; | |
136 } | |
137 } | |
138 | |
139 if (DFA_ALIVE(eucj)) { | |
140 if (!DFA_ALIVE(sjis) && !DFA_ALIVE(utf8)) | |
141 return "EUC-JP"; | |
142 DFA_NEXT(eucj, c); | |
143 } | |
144 if (DFA_ALIVE(sjis)) { | |
145 if (!DFA_ALIVE(eucj) && !DFA_ALIVE(utf8)) | |
146 return "SJIS"; | |
147 DFA_NEXT(sjis, c); | |
148 } | |
149 if (DFA_ALIVE(utf8)) { | |
150 if (!DFA_ALIVE(sjis) && !DFA_ALIVE(eucj)) | |
151 return "UTF-8"; | |
152 DFA_NEXT(utf8, c); | |
153 } | |
154 | |
155 if (!DFA_ALIVE(eucj) && !DFA_ALIVE(sjis) && !DFA_ALIVE(utf8)) { | |
156 /* we ran out the possibilities */ | |
157 return NULL; | |
158 } | |
159 } | |
160 | |
161 /* Now, we have ambigous code. Pick the highest score. If more than | |
162 one candidate tie, pick the default encoding. */ | |
163 for (i = 0; order[i] != NULL; i++) { | |
164 if (order[i]->state >= 0) { //DFA_ALIVE() | |
165 if (top == NULL || order[i]->score > top->score) | |
166 top = order[i]; | |
167 } | |
168 } | |
169 | |
170 if (top == &eucj) | |
171 return "EUC-JP"; | |
172 if (top == &utf8) | |
173 return "UTF-8"; | |
174 if (top == &sjis) | |
175 return "SJIS"; | |
176 return NULL; | |
177 } | |
178 | |
179 const char *guess_tw(const char *buf, int buflen) | |
180 { | |
181 int i; | |
182 guess_dfa big5 = DFA_INIT(guess_big5_st, guess_big5_ar); | |
183 guess_dfa utf8 = DFA_INIT(guess_utf8_st, guess_utf8_ar); | |
184 guess_dfa *top = NULL; | |
185 | |
186 guess_dfa *order[] = { ORDER_TW, NULL }; | |
187 | |
188 for (i = 0; i < buflen; i++) { | |
189 int c = (unsigned char) buf[i]; | |
190 | |
191 /* special treatment of iso-2022 escape sequence */ | |
192 if (c == 0x1b) { | |
193 if (i < buflen - 1) { | |
194 c = (unsigned char) buf[++i]; | |
195 if (c == '$' || c == '(') | |
196 return "ISO-2022-TW"; | |
197 } | |
198 } | |
199 | |
200 /* special treatment of BOM */ | |
201 if (i == 0 && c == 0xff) { | |
202 if (i < buflen - 1) { | |
203 c = (unsigned char) buf[i + 1]; | |
204 if (c == 0xfe) | |
205 return UCS_2LE; | |
206 } | |
207 } | |
208 if (i == 0 && c == 0xfe) { | |
209 if (i < buflen - 1) { | |
210 c = (unsigned char) buf[i + 1]; | |
211 if (c == 0xff) | |
212 return UCS_2BE; | |
213 } | |
214 } | |
215 | |
216 if (DFA_ALIVE(big5)) { | |
217 if (!DFA_ALIVE(utf8)) | |
218 return "BIG5"; | |
219 DFA_NEXT(big5, c); | |
220 } | |
221 if (DFA_ALIVE(utf8)) { | |
222 if (!DFA_ALIVE(big5)) | |
223 return "UTF-8"; | |
224 DFA_NEXT(utf8, c); | |
225 } | |
226 | |
227 if (!DFA_ALIVE(big5) && !DFA_ALIVE(utf8)) { | |
228 /* we ran out the possibilities */ | |
229 return NULL; | |
230 } | |
231 } | |
232 | |
233 /* Now, we have ambigous code. Pick the highest score. If more than | |
234 one candidate tie, pick the default encoding. */ | |
235 for (i = 0; order[i] != NULL; i++) { | |
236 if (order[i]->state >= 0) { //DFA_ALIVE() | |
237 if (top == NULL || order[i]->score > top->score) | |
238 top = order[i]; | |
239 } | |
240 } | |
241 | |
242 if (top == &big5) | |
243 return "BIG5"; | |
244 if (top == &utf8) | |
245 return "UTF-8"; | |
246 return NULL; | |
247 } | |
248 | |
249 const char *guess_cn(const char *buf, int buflen) | |
250 { | |
251 int i; | |
252 guess_dfa gb2312 = DFA_INIT(guess_gb2312_st, guess_gb2312_ar); | |
253 guess_dfa utf8 = DFA_INIT(guess_utf8_st, guess_utf8_ar); | |
254 guess_dfa gb18030 = DFA_INIT(guess_gb18030_st, guess_gb18030_ar); | |
255 guess_dfa *top = NULL; | |
256 | |
257 guess_dfa *order[] = { ORDER_CN, NULL }; | |
258 | |
259 for (i = 0; i < buflen; i++) { | |
260 int c = (unsigned char) buf[i]; | |
261 int c2; | |
262 | |
263 /* special treatment of iso-2022 escape sequence */ | |
264 if (c == 0x1b) { | |
265 if (i < buflen - 1) { | |
266 c = (unsigned char) buf[i + 1]; | |
267 c2 = (unsigned char) buf[i + 2]; | |
268 if (c == '$' && (c2 == ')' || c2 == '+')) | |
269 return "ISO-2022-CN"; | |
270 } | |
271 } | |
272 | |
273 /* special treatment of BOM */ | |
274 if (i == 0 && c == 0xff) { | |
275 if (i < buflen - 1) { | |
276 c = (unsigned char) buf[i + 1]; | |
277 if (c == 0xfe) | |
278 return UCS_2LE; | |
279 } | |
280 } | |
281 if (i == 0 && c == 0xfe) { | |
282 if (i < buflen - 1) { | |
283 c = (unsigned char) buf[i + 1]; | |
284 if (c == 0xff) | |
285 return UCS_2BE; | |
286 } | |
287 } | |
288 | |
289 if (DFA_ALIVE(gb2312)) { | |
290 if (!DFA_ALIVE(utf8) && !DFA_ALIVE(gb18030)) | |
291 return "GB2312"; | |
292 DFA_NEXT(gb2312, c); | |
293 } | |
294 if (DFA_ALIVE(utf8)) { | |
295 if (!DFA_ALIVE(gb2312) && !DFA_ALIVE(gb18030)) | |
296 return "UTF-8"; | |
297 DFA_NEXT(utf8, c); | |
298 } | |
299 if (DFA_ALIVE(gb18030)) { | |
300 if (!DFA_ALIVE(utf8) && !DFA_ALIVE(gb2312)) | |
301 return "GB18030"; | |
302 DFA_NEXT(gb18030, c); | |
303 } | |
304 | |
305 if (!DFA_ALIVE(gb2312) && !DFA_ALIVE(utf8) && !DFA_ALIVE(gb18030)) { | |
306 /* we ran out the possibilities */ | |
307 return NULL; | |
308 } | |
309 } | |
310 | |
311 /* Now, we have ambigous code. Pick the highest score. If more than | |
312 one candidate tie, pick the default encoding. */ | |
313 for (i = 0; order[i] != NULL; i++) { | |
314 if (order[i]->state >= 0) { //DFA_ALIVE() | |
315 if (top == NULL || order[i]->score > top->score) | |
316 top = order[i]; | |
317 } | |
318 } | |
319 | |
320 if (top == &gb2312) | |
321 return "GB2312"; | |
322 if (top == &utf8) | |
323 return "UTF-8"; | |
324 if (top == &gb18030) | |
325 return "GB18030"; | |
326 return NULL; | |
327 } | |
328 | |
329 const char *guess_kr(const char *buf, int buflen) | |
330 { | |
331 int i; | |
332 guess_dfa euck = DFA_INIT(guess_euck_st, guess_euck_ar); | |
333 guess_dfa utf8 = DFA_INIT(guess_utf8_st, guess_utf8_ar); | |
334 guess_dfa johab = DFA_INIT(guess_johab_st, guess_johab_ar); | |
335 guess_dfa *top = NULL; | |
336 | |
337 guess_dfa *order[] = { ORDER_KR, NULL }; | |
338 | |
339 for (i = 0; i < buflen; i++) { | |
340 int c = (unsigned char) buf[i]; | |
341 int c2; | |
342 | |
343 /* special treatment of iso-2022 escape sequence */ | |
344 if (c == 0x1b) { | |
345 if (i < buflen - 1) { | |
346 c = (unsigned char) buf[i + 1]; | |
347 c2 = (unsigned char) buf[i + 2]; | |
348 if (c == '$' && c2 == ')') | |
349 return "ISO-2022-KR"; | |
350 } | |
351 } | |
352 | |
353 /* special treatment of BOM */ | |
354 if (i == 0 && c == 0xff) { | |
355 if (i < buflen - 1) { | |
356 c = (unsigned char) buf[i + 1]; | |
357 if (c == 0xfe) | |
358 return UCS_2LE; | |
359 } | |
360 } | |
361 if (i == 0 && c == 0xfe) { | |
362 if (i < buflen - 1) { | |
363 c = (unsigned char) buf[i + 1]; | |
364 if (c == 0xff) | |
365 return UCS_2BE; | |
366 } | |
367 } | |
368 | |
369 if (DFA_ALIVE(euck)) { | |
370 if (!DFA_ALIVE(johab) && !DFA_ALIVE(utf8)) | |
371 return "EUC-KR"; | |
372 DFA_NEXT(euck, c); | |
373 } | |
374 if (DFA_ALIVE(johab)) { | |
375 if (!DFA_ALIVE(euck) && !DFA_ALIVE(utf8)) | |
376 return "JOHAB"; | |
377 DFA_NEXT(johab, c); | |
378 } | |
379 if (DFA_ALIVE(utf8)) { | |
380 if (!DFA_ALIVE(euck) && !DFA_ALIVE(johab)) | |
381 return "UTF-8"; | |
382 DFA_NEXT(utf8, c); | |
383 } | |
384 | |
385 if (!DFA_ALIVE(euck) && !DFA_ALIVE(johab) && !DFA_ALIVE(utf8)) { | |
386 /* we ran out the possibilities */ | |
387 return NULL; | |
388 } | |
389 } | |
390 | |
391 /* Now, we have ambigous code. Pick the highest score. If more than | |
392 one candidate tie, pick the default encoding. */ | |
393 for (i = 0; order[i] != NULL; i++) { | |
394 if (order[i]->state >= 0) { //DFA_ALIVE() | |
395 if (top == NULL || order[i]->score > top->score) | |
396 top = order[i]; | |
397 } | |
398 } | |
399 | |
400 if (top == &euck) | |
401 return "EUC-KR"; | |
402 if (top == &utf8) | |
403 return "UTF-8"; | |
404 if (top == &johab) | |
405 return "JOHAB"; | |
406 return NULL; | |
407 } |