Mercurial > libguess
comparison cjk_impl.c @ 0:d9b6ff839eab
initial import
author | Yoshiki Yazawa <yaz@cc.rim.or.jp> |
---|---|
date | Fri, 30 Nov 2007 19:34:51 +0900 |
parents | |
children | 754a4550c64e |
comparison
equal
deleted
inserted
replaced
-1:000000000000 | 0:d9b6ff839eab |
---|---|
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 | |
42 /* take precedence if scores are same. you can customize the order as: */ | |
43 /* ORDER_** &highest, &second, ... &lowest */ | |
44 #define ORDER_JP &utf8, &sjis, &eucj | |
45 #define ORDER_TW &utf8, &big5 | |
46 #define ORDER_CN &utf8, &gb2312, &gb18030 | |
47 #define ORDER_KR &utf8, &euck, &johab | |
48 | |
49 /* workaround for that glib's g_convert can't convert | |
50 properly from UCS-2BE/LE trailing after BOM. */ | |
51 #define WITH_G_CONVERT 1 | |
52 /* #undef WITH_G_CONVERT */ | |
53 | |
54 #ifdef WITH_G_CONVERT | |
55 const char UCS_2BE[] = "UTF-16"; | |
56 const char UCS_2LE[] = "UTF-16"; | |
57 #else | |
58 const char UCS_2BE[] = "UCS-2BE"; | |
59 const char UCS_2LE[] = "UCS-2LE"; | |
60 #endif | |
61 | |
62 /* data types */ | |
63 typedef struct guess_arc_rec | |
64 { | |
65 unsigned int next; /* next state */ | |
66 double score; /* score */ | |
67 } guess_arc; | |
68 | |
69 typedef struct guess_dfa_rec | |
70 { | |
71 signed char (*states)[256]; | |
72 guess_arc *arcs; | |
73 int state; | |
74 double score; | |
75 } guess_dfa; | |
76 | |
77 /* macros */ | |
78 #define DFA_INIT(st, ar) \ | |
79 { st, ar, 0, 1.0 } | |
80 | |
81 #define DFA_NEXT(dfa, ch) \ | |
82 do { \ | |
83 int arc__; \ | |
84 if (dfa.state >= 0) { \ | |
85 arc__ = dfa.states[dfa.state][ch]; \ | |
86 if (arc__ < 0) { \ | |
87 dfa.state = -1; \ | |
88 } else { \ | |
89 dfa.state = dfa.arcs[arc__].next; \ | |
90 dfa.score *= dfa.arcs[arc__].score; \ | |
91 } \ | |
92 } \ | |
93 } while (0) | |
94 | |
95 #define DFA_ALIVE(dfa) (dfa.state >= 0) | |
96 | |
97 /* include DFA table generated by guess.scm */ | |
98 #include "guess_tab.c" | |
99 | |
100 | |
101 int dfa_validate_utf8(const char *buf, int buflen) | |
102 { | |
103 int i; | |
104 guess_dfa utf8 = DFA_INIT(guess_utf8_st, guess_utf8_ar); | |
105 | |
106 for (i = 0; i < buflen; i++) { | |
107 int c = (unsigned char) buf[i]; | |
108 | |
109 if (DFA_ALIVE(utf8)) | |
110 DFA_NEXT(utf8, c); | |
111 else | |
112 break; | |
113 } | |
114 | |
115 DFA_NEXT(utf8, '\0'); //Bug #53 | |
116 | |
117 if(DFA_ALIVE(utf8)) | |
118 return 1; | |
119 else | |
120 return 0; | |
121 } | |
122 | |
123 const char *guess_jp(const char *buf, int buflen) | |
124 { | |
125 int i; | |
126 guess_dfa eucj = DFA_INIT(guess_eucj_st, guess_eucj_ar); | |
127 guess_dfa sjis = DFA_INIT(guess_sjis_st, guess_sjis_ar); | |
128 guess_dfa utf8 = DFA_INIT(guess_utf8_st, guess_utf8_ar); | |
129 guess_dfa *top = NULL; | |
130 | |
131 guess_dfa *order[] = { ORDER_JP, NULL }; | |
132 | |
133 for (i = 0; i < buflen; i++) { | |
134 int c = (unsigned char) buf[i]; | |
135 | |
136 /* special treatment of iso-2022 escape sequence */ | |
137 if (c == 0x1b) { | |
138 if (i < buflen - 1) { | |
139 c = (unsigned char) buf[++i]; | |
140 if (c == '$' || c == '(') | |
141 return "ISO-2022-JP"; | |
142 } | |
143 } | |
144 | |
145 /* special treatment of BOM */ | |
146 if (i == 0 && c == 0xff) { | |
147 if (i < buflen - 1) { | |
148 c = (unsigned char) buf[i + 1]; | |
149 if (c == 0xfe) | |
150 return UCS_2LE; | |
151 } | |
152 } | |
153 if (i == 0 && c == 0xfe) { | |
154 if (i < buflen - 1) { | |
155 c = (unsigned char) buf[i + 1]; | |
156 if (c == 0xff) | |
157 return UCS_2BE; | |
158 } | |
159 } | |
160 | |
161 if (DFA_ALIVE(eucj)) { | |
162 if (!DFA_ALIVE(sjis) && !DFA_ALIVE(utf8)) | |
163 return "EUC-JP"; | |
164 DFA_NEXT(eucj, c); | |
165 } | |
166 if (DFA_ALIVE(sjis)) { | |
167 if (!DFA_ALIVE(eucj) && !DFA_ALIVE(utf8)) | |
168 return "SJIS"; | |
169 DFA_NEXT(sjis, c); | |
170 } | |
171 if (DFA_ALIVE(utf8)) { | |
172 if (!DFA_ALIVE(sjis) && !DFA_ALIVE(eucj)) | |
173 return "UTF-8"; | |
174 DFA_NEXT(utf8, c); | |
175 } | |
176 | |
177 if (!DFA_ALIVE(eucj) && !DFA_ALIVE(sjis) && !DFA_ALIVE(utf8)) { | |
178 /* we ran out the possibilities */ | |
179 return NULL; | |
180 } | |
181 } | |
182 | |
183 /* Now, we have ambigous code. Pick the highest score. If more than | |
184 one candidate tie, pick the default encoding. */ | |
185 for (i = 0; order[i] != NULL; i++) { | |
186 if (order[i]->state >= 0) { //DFA_ALIVE() | |
187 if (top == NULL || order[i]->score > top->score) | |
188 top = order[i]; | |
189 } | |
190 } | |
191 | |
192 if (top == &eucj) | |
193 return "EUC-JP"; | |
194 if (top == &utf8) | |
195 return "UTF-8"; | |
196 if (top == &sjis) | |
197 return "SJIS"; | |
198 return NULL; | |
199 } | |
200 | |
201 const char *guess_tw(const char *buf, int buflen) | |
202 { | |
203 int i; | |
204 guess_dfa big5 = DFA_INIT(guess_big5_st, guess_big5_ar); | |
205 guess_dfa utf8 = DFA_INIT(guess_utf8_st, guess_utf8_ar); | |
206 guess_dfa *top = NULL; | |
207 | |
208 guess_dfa *order[] = { ORDER_TW, NULL }; | |
209 | |
210 for (i = 0; i < buflen; i++) { | |
211 int c = (unsigned char) buf[i]; | |
212 | |
213 /* special treatment of iso-2022 escape sequence */ | |
214 if (c == 0x1b) { | |
215 if (i < buflen - 1) { | |
216 c = (unsigned char) buf[++i]; | |
217 if (c == '$' || c == '(') | |
218 return "ISO-2022-TW"; | |
219 } | |
220 } | |
221 | |
222 /* special treatment of BOM */ | |
223 if (i == 0 && c == 0xff) { | |
224 if (i < buflen - 1) { | |
225 c = (unsigned char) buf[i + 1]; | |
226 if (c == 0xfe) | |
227 return UCS_2LE; | |
228 } | |
229 } | |
230 if (i == 0 && c == 0xfe) { | |
231 if (i < buflen - 1) { | |
232 c = (unsigned char) buf[i + 1]; | |
233 if (c == 0xff) | |
234 return UCS_2BE; | |
235 } | |
236 } | |
237 | |
238 if (DFA_ALIVE(big5)) { | |
239 if (!DFA_ALIVE(utf8)) | |
240 return "BIG5"; | |
241 DFA_NEXT(big5, c); | |
242 } | |
243 if (DFA_ALIVE(utf8)) { | |
244 if (!DFA_ALIVE(big5)) | |
245 return "UTF-8"; | |
246 DFA_NEXT(utf8, c); | |
247 } | |
248 | |
249 if (!DFA_ALIVE(big5) && !DFA_ALIVE(utf8)) { | |
250 /* we ran out the possibilities */ | |
251 return NULL; | |
252 } | |
253 } | |
254 | |
255 /* Now, we have ambigous code. Pick the highest score. If more than | |
256 one candidate tie, pick the default encoding. */ | |
257 for (i = 0; order[i] != NULL; i++) { | |
258 if (order[i]->state >= 0) { //DFA_ALIVE() | |
259 if (top == NULL || order[i]->score > top->score) | |
260 top = order[i]; | |
261 } | |
262 } | |
263 | |
264 if (top == &big5) | |
265 return "BIG5"; | |
266 if (top == &utf8) | |
267 return "UTF-8"; | |
268 return NULL; | |
269 } | |
270 | |
271 const char *guess_cn(const char *buf, int buflen) | |
272 { | |
273 int i; | |
274 guess_dfa gb2312 = DFA_INIT(guess_gb2312_st, guess_gb2312_ar); | |
275 guess_dfa utf8 = DFA_INIT(guess_utf8_st, guess_utf8_ar); | |
276 guess_dfa gb18030 = DFA_INIT(guess_gb18030_st, guess_gb18030_ar); | |
277 guess_dfa *top = NULL; | |
278 | |
279 guess_dfa *order[] = { ORDER_CN, NULL }; | |
280 | |
281 for (i = 0; i < buflen; i++) { | |
282 int c = (unsigned char) buf[i]; | |
283 int c2; | |
284 | |
285 /* special treatment of iso-2022 escape sequence */ | |
286 if (c == 0x1b) { | |
287 if (i < buflen - 1) { | |
288 c = (unsigned char) buf[i + 1]; | |
289 c2 = (unsigned char) buf[i + 2]; | |
290 if (c == '$' && (c2 == ')' || c2 == '+')) | |
291 return "ISO-2022-CN"; | |
292 } | |
293 } | |
294 | |
295 /* special treatment of BOM */ | |
296 if (i == 0 && c == 0xff) { | |
297 if (i < buflen - 1) { | |
298 c = (unsigned char) buf[i + 1]; | |
299 if (c == 0xfe) | |
300 return UCS_2LE; | |
301 } | |
302 } | |
303 if (i == 0 && c == 0xfe) { | |
304 if (i < buflen - 1) { | |
305 c = (unsigned char) buf[i + 1]; | |
306 if (c == 0xff) | |
307 return UCS_2BE; | |
308 } | |
309 } | |
310 | |
311 if (DFA_ALIVE(gb2312)) { | |
312 if (!DFA_ALIVE(utf8) && !DFA_ALIVE(gb18030)) | |
313 return "GB2312"; | |
314 DFA_NEXT(gb2312, c); | |
315 } | |
316 if (DFA_ALIVE(utf8)) { | |
317 if (!DFA_ALIVE(gb2312) && !DFA_ALIVE(gb18030)) | |
318 return "UTF-8"; | |
319 DFA_NEXT(utf8, c); | |
320 } | |
321 if (DFA_ALIVE(gb18030)) { | |
322 if (!DFA_ALIVE(utf8) && !DFA_ALIVE(gb2312)) | |
323 return "GB18030"; | |
324 DFA_NEXT(gb18030, c); | |
325 } | |
326 | |
327 if (!DFA_ALIVE(gb2312) && !DFA_ALIVE(utf8) && !DFA_ALIVE(gb18030)) { | |
328 /* we ran out the possibilities */ | |
329 return NULL; | |
330 } | |
331 } | |
332 | |
333 /* Now, we have ambigous code. Pick the highest score. If more than | |
334 one candidate tie, pick the default encoding. */ | |
335 for (i = 0; order[i] != NULL; i++) { | |
336 if (order[i]->state >= 0) { //DFA_ALIVE() | |
337 if (top == NULL || order[i]->score > top->score) | |
338 top = order[i]; | |
339 } | |
340 } | |
341 | |
342 if (top == &gb2312) | |
343 return "GB2312"; | |
344 if (top == &utf8) | |
345 return "UTF-8"; | |
346 if (top == &gb18030) | |
347 return "GB18030"; | |
348 return NULL; | |
349 } | |
350 | |
351 const char *guess_kr(const char *buf, int buflen) | |
352 { | |
353 int i; | |
354 guess_dfa euck = DFA_INIT(guess_euck_st, guess_euck_ar); | |
355 guess_dfa utf8 = DFA_INIT(guess_utf8_st, guess_utf8_ar); | |
356 guess_dfa johab = DFA_INIT(guess_johab_st, guess_johab_ar); | |
357 guess_dfa *top = NULL; | |
358 | |
359 guess_dfa *order[] = { ORDER_KR, NULL }; | |
360 | |
361 for (i = 0; i < buflen; i++) { | |
362 int c = (unsigned char) buf[i]; | |
363 int c2; | |
364 | |
365 /* special treatment of iso-2022 escape sequence */ | |
366 if (c == 0x1b) { | |
367 if (i < buflen - 1) { | |
368 c = (unsigned char) buf[i + 1]; | |
369 c2 = (unsigned char) buf[i + 2]; | |
370 if (c == '$' && c2 == ')') | |
371 return "ISO-2022-KR"; | |
372 } | |
373 } | |
374 | |
375 /* special treatment of BOM */ | |
376 if (i == 0 && c == 0xff) { | |
377 if (i < buflen - 1) { | |
378 c = (unsigned char) buf[i + 1]; | |
379 if (c == 0xfe) | |
380 return UCS_2LE; | |
381 } | |
382 } | |
383 if (i == 0 && c == 0xfe) { | |
384 if (i < buflen - 1) { | |
385 c = (unsigned char) buf[i + 1]; | |
386 if (c == 0xff) | |
387 return UCS_2BE; | |
388 } | |
389 } | |
390 | |
391 if (DFA_ALIVE(euck)) { | |
392 if (!DFA_ALIVE(johab) && !DFA_ALIVE(utf8)) | |
393 return "EUC-KR"; | |
394 DFA_NEXT(euck, c); | |
395 } | |
396 if (DFA_ALIVE(johab)) { | |
397 if (!DFA_ALIVE(euck) && !DFA_ALIVE(utf8)) | |
398 return "JOHAB"; | |
399 DFA_NEXT(johab, c); | |
400 } | |
401 if (DFA_ALIVE(utf8)) { | |
402 if (!DFA_ALIVE(euck) && !DFA_ALIVE(johab)) | |
403 return "UTF-8"; | |
404 DFA_NEXT(utf8, c); | |
405 } | |
406 | |
407 if (!DFA_ALIVE(euck) && !DFA_ALIVE(johab) && !DFA_ALIVE(utf8)) { | |
408 /* we ran out the possibilities */ | |
409 return NULL; | |
410 } | |
411 } | |
412 | |
413 /* Now, we have ambigous code. Pick the highest score. If more than | |
414 one candidate tie, pick the default encoding. */ | |
415 for (i = 0; order[i] != NULL; i++) { | |
416 if (order[i]->state >= 0) { //DFA_ALIVE() | |
417 if (top == NULL || order[i]->score > top->score) | |
418 top = order[i]; | |
419 } | |
420 } | |
421 | |
422 if (top == &euck) | |
423 return "EUC-KR"; | |
424 if (top == &utf8) | |
425 return "UTF-8"; | |
426 if (top == &johab) | |
427 return "JOHAB"; | |
428 return NULL; | |
429 } |