2005-09-27 [colin] 1.9.14cvs56
[claws.git] / src / codeconv.c
1 /*
2  * Sylpheed -- a GTK+ based, lightweight, and fast e-mail client
3  * Copyright (C) 1999-2005 Hiroyuki Yamamoto
4  *
5  * This program is free software; you can redistribute it and/or modify
6  * it under the terms of the GNU General Public License as published by
7  * the Free Software Foundation; either version 2 of the License, or
8  * (at your option) any later version.
9  *
10  * This program is distributed in the hope that it will be useful,
11  * but WITHOUT ANY WARRANTY; without even the implied warranty of
12  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
13  * GNU General Public License for more details.
14  *
15  * You should have received a copy of the GNU General Public License
16  * along with this program; if not, write to the Free Software
17  * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
18  */
19
20 #ifdef HAVE_CONFIG_H
21 #  include "config.h"
22 #endif
23
24 #include "defs.h"
25
26 #include <glib.h>
27 #include <glib/gi18n.h>
28 #include <string.h>
29 #include <ctype.h>
30 #include <stdlib.h>
31 #include <errno.h>
32
33 #if HAVE_LOCALE_H
34 #  include <locale.h>
35 #endif
36
37 #include <iconv.h>
38
39 #include "codeconv.h"
40 #include "unmime.h"
41 #include "base64.h"
42 #include "quoted-printable.h"
43 #include "utils.h"
44 #include "prefs_common.h"
45
46 typedef enum
47 {
48         JIS_ASCII,
49         JIS_KANJI,
50         JIS_HWKANA,
51         JIS_AUXKANJI
52 } JISState;
53
54 #define SUBST_CHAR      0x5f;
55 #define ESC             '\033'
56
57 #define iseuckanji(c) \
58         (((c) & 0xff) >= 0xa1 && ((c) & 0xff) <= 0xfe)
59 #define iseuchwkana1(c) \
60         (((c) & 0xff) == 0x8e)
61 #define iseuchwkana2(c) \
62         (((c) & 0xff) >= 0xa1 && ((c) & 0xff) <= 0xdf)
63 #define iseucaux(c) \
64         (((c) & 0xff) == 0x8f)
65 #define issjiskanji1(c) \
66         ((((c) & 0xff) >= 0x81 && ((c) & 0xff) <= 0x9f) || \
67          (((c) & 0xff) >= 0xe0 && ((c) & 0xff) <= 0xfc))
68 #define issjiskanji2(c) \
69         ((((c) & 0xff) >= 0x40 && ((c) & 0xff) <= 0x7e) || \
70          (((c) & 0xff) >= 0x80 && ((c) & 0xff) <= 0xfc))
71 #define issjishwkana(c) \
72         (((c) & 0xff) >= 0xa1 && ((c) & 0xff) <= 0xdf)
73
74 #define K_IN()                          \
75         if (state != JIS_KANJI) {       \
76                 *out++ = ESC;           \
77                 *out++ = '$';           \
78                 *out++ = 'B';           \
79                 state = JIS_KANJI;      \
80         }
81
82 #define K_OUT()                         \
83         if (state != JIS_ASCII) {       \
84                 *out++ = ESC;           \
85                 *out++ = '(';           \
86                 *out++ = 'B';           \
87                 state = JIS_ASCII;      \
88         }
89
90 #define HW_IN()                         \
91         if (state != JIS_HWKANA) {      \
92                 *out++ = ESC;           \
93                 *out++ = '(';           \
94                 *out++ = 'I';           \
95                 state = JIS_HWKANA;     \
96         }
97
98 #define AUX_IN()                        \
99         if (state != JIS_AUXKANJI) {    \
100                 *out++ = ESC;           \
101                 *out++ = '$';           \
102                 *out++ = '(';           \
103                 *out++ = 'D';           \
104                 state = JIS_AUXKANJI;   \
105         }
106
107 static void conv_jistoeuc(gchar *outbuf, gint outlen, const gchar *inbuf);
108 static void conv_euctojis(gchar *outbuf, gint outlen, const gchar *inbuf);
109 static void conv_sjistoeuc(gchar *outbuf, gint outlen, const gchar *inbuf);
110
111 static void conv_jistoutf8(gchar *outbuf, gint outlen, const gchar *inbuf);
112 static void conv_sjistoutf8(gchar *outbuf, gint outlen, const gchar *inbuf);
113 static void conv_euctoutf8(gchar *outbuf, gint outlen, const gchar *inbuf);
114 static void conv_anytoutf8(gchar *outbuf, gint outlen, const gchar *inbuf);
115
116 static void conv_utf8toeuc(gchar *outbuf, gint outlen, const gchar *inbuf);
117 static void conv_utf8tojis(gchar *outbuf, gint outlen, const gchar *inbuf);
118
119 static void conv_unreadable_8bit(gchar *str);
120
121 static void conv_jistodisp(gchar *outbuf, gint outlen, const gchar *inbuf);
122 static void conv_sjistodisp(gchar *outbuf, gint outlen, const gchar *inbuf);
123 static void conv_euctodisp(gchar *outbuf, gint outlen, const gchar *inbuf);
124
125 static void conv_anytodisp(gchar *outbuf, gint outlen, const gchar *inbuf);
126 static void conv_ustodisp(gchar *outbuf, gint outlen, const gchar *inbuf);
127 static void conv_noconv(gchar *outbuf, gint outlen, const gchar *inbuf);
128
129 static gboolean strict_mode = FALSE;
130
131 void codeconv_set_strict(gboolean mode)
132 {
133         strict_mode = mode;
134 }
135
136 static void conv_jistoeuc(gchar *outbuf, gint outlen, const gchar *inbuf)
137 {
138         const guchar *in = inbuf;
139         guchar *out = outbuf;
140         JISState state = JIS_ASCII;
141
142         while (*in != '\0') {
143                 if (*in == ESC) {
144                         in++;
145                         if (*in == '$') {
146                                 if (*(in + 1) == '@' || *(in + 1) == 'B') {
147                                         state = JIS_KANJI;
148                                         in += 2;
149                                 } else if (*(in + 1) == '(' &&
150                                            *(in + 2) == 'D') {
151                                         state = JIS_AUXKANJI;
152                                         in += 3;
153                                 } else {
154                                         /* unknown escape sequence */
155                                         state = JIS_ASCII;
156                                 }
157                         } else if (*in == '(') {
158                                 if (*(in + 1) == 'B' || *(in + 1) == 'J') {
159                                         state = JIS_ASCII;
160                                         in += 2;
161                                 } else if (*(in + 1) == 'I') {
162                                         state = JIS_HWKANA;
163                                         in += 2;
164                                 } else {
165                                         /* unknown escape sequence */
166                                         state = JIS_ASCII;
167                                 }
168                         } else {
169                                 /* unknown escape sequence */
170                                 state = JIS_ASCII;
171                         }
172                 } else if (*in == 0x0e) {
173                         state = JIS_HWKANA;
174                         in++;
175                 } else if (*in == 0x0f) {
176                         state = JIS_ASCII;
177                         in++;
178                 } else {
179                         switch (state) {
180                         case JIS_ASCII:
181                                 *out++ = *in++;
182                                 break;
183                         case JIS_KANJI:
184                                 *out++ = *in++ | 0x80;
185                                 if (*in == '\0') break;
186                                 *out++ = *in++ | 0x80;
187                                 break;
188                         case JIS_HWKANA:
189                                 *out++ = 0x8e;
190                                 *out++ = *in++ | 0x80;
191                                 break;
192                         case JIS_AUXKANJI:
193                                 *out++ = 0x8f;
194                                 *out++ = *in++ | 0x80;
195                                 if (*in == '\0') break;
196                                 *out++ = *in++ | 0x80;
197                                 break;
198                         }
199                 }
200         }
201
202         *out = '\0';
203 }
204
205 #define JIS_HWDAKUTEN           0x5e
206 #define JIS_HWHANDAKUTEN        0x5f
207
208 static gint conv_jis_hantozen(guchar *outbuf, guchar jis_code, guchar sound_sym)
209 {
210         static guint16 h2z_tbl[] = {
211                 /* 0x20 - 0x2f */
212                 0x0000, 0x2123, 0x2156, 0x2157, 0x2122, 0x2126, 0x2572, 0x2521,
213                 0x2523, 0x2525, 0x2527, 0x2529, 0x2563, 0x2565, 0x2567, 0x2543,
214                 /* 0x30 - 0x3f */
215                 0x213c, 0x2522, 0x2524, 0x2526, 0x2528, 0x252a, 0x252b, 0x252d,
216                 0x252f, 0x2531, 0x2533, 0x2535, 0x2537, 0x2539, 0x253b, 0x253d,
217                 /* 0x40 - 0x4f */
218                 0x253f, 0x2541, 0x2544, 0x2546, 0x2548, 0x254a, 0x254b, 0x254c,
219                 0x254d, 0x254e, 0x254f, 0x2552, 0x2555, 0x2558, 0x255b, 0x255e,
220                 /* 0x50 - 0x5f */
221                 0x255f, 0x2560, 0x2561, 0x2562, 0x2564, 0x2566, 0x2568, 0x2569,
222                 0x256a, 0x256b, 0x256c, 0x256d, 0x256f, 0x2573, 0x212b, 0x212c
223         };
224
225         static guint16 dakuten_tbl[] = {
226                 /* 0x30 - 0x3f */
227                 0x0000, 0x0000, 0x0000, 0x0000, 0x0000, 0x0000, 0x252c, 0x252e,
228                 0x2530, 0x2532, 0x2534, 0x2536, 0x2538, 0x253a, 0x253c, 0x253e,
229                 /* 0x40 - 0x4f */
230                 0x2540, 0x2542, 0x2545, 0x2547, 0x2549, 0x0000, 0x0000, 0x0000,
231                 0x0000, 0x0000, 0x2550, 0x2553, 0x2556, 0x2559, 0x255c, 0x0000
232         };
233
234         static guint16 handakuten_tbl[] = {
235                 /* 0x4a - 0x4e */
236                 0x2551, 0x2554, 0x2557, 0x255a, 0x255d
237         };
238
239         guint16 out_code;
240
241         jis_code &= 0x7f;
242         sound_sym &= 0x7f;
243
244         if (jis_code < 0x21 || jis_code > 0x5f)
245                 return 0;
246
247         if (sound_sym == JIS_HWDAKUTEN &&
248             jis_code >= 0x36 && jis_code <= 0x4e) {
249                 out_code = dakuten_tbl[jis_code - 0x30];
250                 if (out_code != 0) {
251                         *outbuf = out_code >> 8;
252                         *(outbuf + 1) = out_code & 0xff;
253                         return 2;
254                 }
255         }
256
257         if (sound_sym == JIS_HWHANDAKUTEN &&
258             jis_code >= 0x4a && jis_code <= 0x4e) {
259                 out_code = handakuten_tbl[jis_code - 0x4a];
260                 *outbuf = out_code >> 8;
261                 *(outbuf + 1) = out_code & 0xff;
262                 return 2;
263         }
264
265         out_code = h2z_tbl[jis_code - 0x20];
266         *outbuf = out_code >> 8;
267         *(outbuf + 1) = out_code & 0xff;
268         return 1;
269 }
270
271 static void conv_euctojis(gchar *outbuf, gint outlen, const gchar *inbuf)
272 {
273         const guchar *in = inbuf;
274         guchar *out = outbuf;
275         JISState state = JIS_ASCII;
276
277         while (*in != '\0') {
278                 if (IS_ASCII(*in)) {
279                         K_OUT();
280                         *out++ = *in++;
281                 } else if (iseuckanji(*in)) {
282                         if (iseuckanji(*(in + 1))) {
283                                 K_IN();
284                                 *out++ = *in++ & 0x7f;
285                                 *out++ = *in++ & 0x7f;
286                         } else {
287                                 K_OUT();
288                                 *out++ = SUBST_CHAR;
289                                 in++;
290                                 if (*in != '\0' && !IS_ASCII(*in)) {
291                                         *out++ = SUBST_CHAR;
292                                         in++;
293                                 }
294                         }
295                 } else if (iseuchwkana1(*in)) {
296                         if (iseuchwkana2(*(in + 1))) {
297                                 if (prefs_common.allow_jisx0201_kana) {
298                                         HW_IN();
299                                         in++;
300                                         *out++ = *in++ & 0x7f;
301                                 } else {
302                                         guchar jis_ch[2];
303                                         gint len;
304
305                                         if (iseuchwkana1(*(in + 2)) &&
306                                             iseuchwkana2(*(in + 3)))
307                                                 len = conv_jis_hantozen
308                                                         (jis_ch,
309                                                          *(in + 1), *(in + 3));
310                                         else
311                                                 len = conv_jis_hantozen
312                                                         (jis_ch,
313                                                          *(in + 1), '\0');
314                                         if (len == 0)
315                                                 in += 2;
316                                         else {
317                                                 K_IN();
318                                                 in += len * 2;
319                                                 *out++ = jis_ch[0];
320                                                 *out++ = jis_ch[1];
321                                         }
322                                 }
323                         } else {
324                                 K_OUT();
325                                 in++;
326                                 if (*in != '\0' && !IS_ASCII(*in)) {
327                                         *out++ = SUBST_CHAR;
328                                         in++;
329                                 }
330                         }
331                 } else if (iseucaux(*in)) {
332                         in++;
333                         if (iseuckanji(*in) && iseuckanji(*(in + 1))) {
334                                 AUX_IN();
335                                 *out++ = *in++ & 0x7f;
336                                 *out++ = *in++ & 0x7f;
337                         } else {
338                                 K_OUT();
339                                 if (*in != '\0' && !IS_ASCII(*in)) {
340                                         *out++ = SUBST_CHAR;
341                                         in++;
342                                         if (*in != '\0' && !IS_ASCII(*in)) {
343                                                 *out++ = SUBST_CHAR;
344                                                 in++;
345                                         }
346                                 }
347                         }
348                 } else {
349                         K_OUT();
350                         *out++ = SUBST_CHAR;
351                         in++;
352                 }
353         }
354
355         K_OUT();
356         *out = '\0';
357 }
358
359 static void conv_sjistoeuc(gchar *outbuf, gint outlen, const gchar *inbuf)
360 {
361         const guchar *in = inbuf;
362         guchar *out = outbuf;
363
364         while (*in != '\0') {
365                 if (IS_ASCII(*in)) {
366                         *out++ = *in++;
367                 } else if (issjiskanji1(*in)) {
368                         if (issjiskanji2(*(in + 1))) {
369                                 guchar out1 = *in;
370                                 guchar out2 = *(in + 1);
371                                 guchar row;
372
373                                 row = out1 < 0xa0 ? 0x70 : 0xb0;
374                                 if (out2 < 0x9f) {
375                                         out1 = (out1 - row) * 2 - 1;
376                                         out2 -= out2 > 0x7f ? 0x20 : 0x1f;
377                                 } else {
378                                         out1 = (out1 - row) * 2;
379                                         out2 -= 0x7e;
380                                 }
381
382                                 *out++ = out1 | 0x80;
383                                 *out++ = out2 | 0x80;
384                                 in += 2;
385                         } else {
386                                 *out++ = SUBST_CHAR;
387                                 in++;
388                                 if (*in != '\0' && !IS_ASCII(*in)) {
389                                         *out++ = SUBST_CHAR;
390                                         in++;
391                                 }
392                         }
393                 } else if (issjishwkana(*in)) {
394                         *out++ = 0x8e;
395                         *out++ = *in++;
396                 } else {
397                         *out++ = SUBST_CHAR;
398                         in++;
399                 }
400         }
401
402         *out = '\0';
403 }
404
405 static void conv_jistoutf8(gchar *outbuf, gint outlen, const gchar *inbuf)
406 {
407         gchar *eucstr;
408
409         Xalloca(eucstr, outlen, return);
410
411         conv_jistoeuc(eucstr, outlen, inbuf);
412         conv_euctoutf8(outbuf, outlen, eucstr);
413 }
414
415 static void conv_sjistoutf8(gchar *outbuf, gint outlen, const gchar *inbuf)
416 {
417         gchar *tmpstr;
418
419         tmpstr = conv_iconv_strdup(inbuf, CS_SHIFT_JIS, CS_UTF_8);
420         if (tmpstr) {
421                 strncpy2(outbuf, tmpstr, outlen);
422                 g_free(tmpstr);
423         } else
424                 strncpy2(outbuf, inbuf, outlen);
425 }
426
427 static void conv_euctoutf8(gchar *outbuf, gint outlen, const gchar *inbuf)
428 {
429         static iconv_t cd = (iconv_t)-1;
430         static gboolean iconv_ok = TRUE;
431         gchar *tmpstr;
432
433         if (cd == (iconv_t)-1) {
434                 if (!iconv_ok) {
435                         strncpy2(outbuf, inbuf, outlen);
436                         return;
437                 }
438                 cd = iconv_open(CS_UTF_8, CS_EUC_JP_MS);
439                 if (cd == (iconv_t)-1) {
440                         cd = iconv_open(CS_UTF_8, CS_EUC_JP);
441                         if (cd == (iconv_t)-1) {
442                                 g_warning("conv_euctoutf8(): %s\n",
443                                           g_strerror(errno));
444                                 iconv_ok = FALSE;
445                                 strncpy2(outbuf, inbuf, outlen);
446                                 return;
447                         }
448                 }
449         }
450
451         tmpstr = conv_iconv_strdup_with_cd(inbuf, cd);
452         if (tmpstr) {
453                 strncpy2(outbuf, tmpstr, outlen);
454                 g_free(tmpstr);
455         } else
456                 strncpy2(outbuf, inbuf, outlen);
457 }
458
459 static void conv_anytoutf8(gchar *outbuf, gint outlen, const gchar *inbuf)
460 {
461         switch (conv_guess_ja_encoding(inbuf)) {
462         case C_ISO_2022_JP:
463                 conv_jistoutf8(outbuf, outlen, inbuf);
464                 break;
465         case C_SHIFT_JIS:
466                 conv_sjistoutf8(outbuf, outlen, inbuf);
467                 break;
468         case C_EUC_JP:
469                 conv_euctoutf8(outbuf, outlen, inbuf);
470                 break;
471         default:
472                 strncpy2(outbuf, inbuf, outlen);
473                 break;
474         }
475 }
476
477 static void conv_utf8toeuc(gchar *outbuf, gint outlen, const gchar *inbuf)
478 {
479         static iconv_t cd = (iconv_t)-1;
480         static gboolean iconv_ok = TRUE;
481         gchar *tmpstr;
482
483         if (cd == (iconv_t)-1) {
484                 if (!iconv_ok) {
485                         strncpy2(outbuf, inbuf, outlen);
486                         return;
487                 }
488                 cd = iconv_open(CS_EUC_JP_MS, CS_UTF_8);
489                 if (cd == (iconv_t)-1) {
490                         cd = iconv_open(CS_EUC_JP, CS_UTF_8);
491                         if (cd == (iconv_t)-1) {
492                                 g_warning("conv_utf8toeuc(): %s\n",
493                                           g_strerror(errno));
494                                 iconv_ok = FALSE;
495                                 strncpy2(outbuf, inbuf, outlen);
496                                 return;
497                         }
498                 }
499         }
500
501         tmpstr = conv_iconv_strdup_with_cd(inbuf, cd);
502         if (tmpstr) {
503                 strncpy2(outbuf, tmpstr, outlen);
504                 g_free(tmpstr);
505         } else
506                 strncpy2(outbuf, inbuf, outlen);
507 }
508
509 static void conv_utf8tojis(gchar *outbuf, gint outlen, const gchar *inbuf)
510 {
511         gchar *eucstr;
512
513         Xalloca(eucstr, outlen, return);
514
515         conv_utf8toeuc(eucstr, outlen, inbuf);
516         conv_euctojis(outbuf, outlen, eucstr);
517 }
518
519 static void conv_unreadable_8bit(gchar *str)
520 {
521         register guchar *p = str;
522
523         while (*p != '\0') {
524                 /* convert CR+LF -> LF */
525                 if (*p == '\r' && *(p + 1) == '\n')
526                         memmove(p, p + 1, strlen(p));
527                 else if (!IS_ASCII(*p)) *p = SUBST_CHAR;
528                 p++;
529         }
530 }
531
532 #define NCV     '\0'
533
534 void conv_mb_alnum(gchar *str)
535 {
536         static guchar char_tbl[] = {
537                 /* 0xa0 - 0xaf */
538                 NCV, ' ', NCV, NCV, ',', '.', NCV, ':',
539                 ';', '?', '!', NCV, NCV, NCV, NCV, NCV,
540                 /* 0xb0 - 0xbf */
541                 NCV, NCV, NCV, NCV, NCV, NCV, NCV, NCV,
542                 NCV, NCV, NCV, NCV, NCV, NCV, NCV, NCV,
543                 /* 0xc0 - 0xcf */
544                 NCV, NCV, NCV, NCV, NCV, NCV, NCV, NCV,
545                 NCV, NCV, '(', ')', NCV, NCV, '[', ']',
546                 /* 0xd0 - 0xdf */
547                 '{', '}', NCV, NCV, NCV, NCV, NCV, NCV,
548                 NCV, NCV, NCV, NCV, '+', '-', NCV, NCV,
549                 /* 0xe0 - 0xef */
550                 NCV, '=', NCV, '<', '>', NCV, NCV, NCV,
551                 NCV, NCV, NCV, NCV, NCV, NCV, NCV, NCV
552         };
553
554         register guchar *p = str;
555         register gint len;
556
557         len = strlen(str);
558
559         while (len > 1) {
560                 if (*p == 0xa3) {
561                         register guchar ch = *(p + 1);
562
563                         if (ch >= 0xb0 && ch <= 0xfa) {
564                                 /* [a-zA-Z] */
565                                 *p = ch & 0x7f;
566                                 p++;
567                                 len--;
568                                 memmove(p, p + 1, len);
569                                 len--;
570                         } else  {
571                                 p += 2;
572                                 len -= 2;
573                         }
574                 } else if (*p == 0xa1) {
575                         register guchar ch = *(p + 1);
576
577                         if (ch >= 0xa0 && ch <= 0xef &&
578                             NCV != char_tbl[ch - 0xa0]) {
579                                 *p = char_tbl[ch - 0xa0];
580                                 p++;
581                                 len--;
582                                 memmove(p, p + 1, len);
583                                 len--;
584                         } else {
585                                 p += 2;
586                                 len -= 2;
587                         }
588                 } else if (iseuckanji(*p)) {
589                         p += 2;
590                         len -= 2;
591                 } else {
592                         p++;
593                         len--;
594                 }
595         }
596 }
597
598 CharSet conv_guess_ja_encoding(const gchar *str)
599 {
600         const guchar *p = str;
601         CharSet guessed = C_US_ASCII;
602
603         while (*p != '\0') {
604                 if (*p == ESC && (*(p + 1) == '$' || *(p + 1) == '(')) {
605                         if (guessed == C_US_ASCII)
606                                 return C_ISO_2022_JP;
607                         p += 2;
608                 } else if (IS_ASCII(*p)) {
609                         p++;
610                 } else if (iseuckanji(*p) && iseuckanji(*(p + 1))) {
611                         if (*p >= 0xfd && *p <= 0xfe)
612                                 return C_EUC_JP;
613                         else if (guessed == C_SHIFT_JIS) {
614                                 if ((issjiskanji1(*p) &&
615                                      issjiskanji2(*(p + 1))) ||
616                                     issjishwkana(*p))
617                                         guessed = C_SHIFT_JIS;
618                                 else
619                                         guessed = C_EUC_JP;
620                         } else
621                                 guessed = C_EUC_JP;
622                         p += 2;
623                 } else if (issjiskanji1(*p) && issjiskanji2(*(p + 1))) {
624                         if (iseuchwkana1(*p) && iseuchwkana2(*(p + 1)))
625                                 guessed = C_SHIFT_JIS;
626                         else
627                                 return C_SHIFT_JIS;
628                         p += 2;
629                 } else if (issjishwkana(*p)) {
630                         guessed = C_SHIFT_JIS;
631                         p++;
632                 } else {
633                         p++;
634                 }
635         }
636
637         return guessed;
638 }
639
640 static void conv_jistodisp(gchar *outbuf, gint outlen, const gchar *inbuf)
641 {
642         conv_jistoutf8(outbuf, outlen, inbuf);
643 }
644
645 static void conv_sjistodisp(gchar *outbuf, gint outlen, const gchar *inbuf)
646 {
647         conv_sjistoutf8(outbuf, outlen, inbuf);
648 }
649
650 static void conv_euctodisp(gchar *outbuf, gint outlen, const gchar *inbuf)
651 {
652         conv_euctoutf8(outbuf, outlen, inbuf);
653 }
654
655 void conv_utf8todisp(gchar *outbuf, gint outlen, const gchar *inbuf)
656 {
657         if (g_utf8_validate(inbuf, -1, NULL) == TRUE)
658                 strncpy2(outbuf, inbuf, outlen);
659         else
660                 conv_ustodisp(outbuf, outlen, inbuf);
661 }
662
663 static void conv_anytodisp(gchar *outbuf, gint outlen, const gchar *inbuf)
664 {
665         conv_anytoutf8(outbuf, outlen, inbuf);
666         if (g_utf8_validate(outbuf, -1, NULL) != TRUE)
667                 conv_unreadable_8bit(outbuf);
668 }
669
670 static void conv_ustodisp(gchar *outbuf, gint outlen, const gchar *inbuf)
671 {
672         strncpy2(outbuf, inbuf, outlen);
673         conv_unreadable_8bit(outbuf);
674 }
675
676 void conv_localetodisp(gchar *outbuf, gint outlen, const gchar *inbuf)
677 {
678         gchar *tmpstr;
679
680         tmpstr = conv_iconv_strdup(inbuf, conv_get_locale_charset_str(),
681                                    CS_INTERNAL);
682         if (tmpstr && g_utf8_validate(tmpstr, -1, NULL)) {
683                 strncpy2(outbuf, tmpstr, outlen);
684                 g_free(tmpstr);
685         } else if (tmpstr && !g_utf8_validate(tmpstr, -1, NULL)) {
686                 g_free(tmpstr);
687                 conv_utf8todisp(outbuf, outlen, inbuf);
688         } else
689                 conv_utf8todisp(outbuf, outlen, inbuf);
690 }
691
692 static void conv_noconv(gchar *outbuf, gint outlen, const gchar *inbuf)
693 {
694         strncpy2(outbuf, inbuf, outlen);
695 }
696
697 static const gchar *
698 conv_get_fallback_for_private_encoding(const gchar *encoding)
699 {
700         if (encoding && (encoding[0] == 'X' || encoding[0] == 'x') &&
701             encoding[1] == '-') {
702                 if (!g_ascii_strcasecmp(encoding, CS_X_GBK))
703                         return CS_GBK;
704         }
705
706         return encoding;
707 }
708
709 CodeConverter *conv_code_converter_new(const gchar *src_charset)
710 {
711         CodeConverter *conv;
712
713         src_charset = conv_get_fallback_for_private_encoding(src_charset);
714
715         conv = g_new0(CodeConverter, 1);
716         conv->code_conv_func = conv_get_code_conv_func(src_charset, NULL);
717         conv->charset_str = g_strdup(src_charset);
718         conv->charset = conv_get_charset_from_str(src_charset);
719
720         return conv;
721 }
722
723 void conv_code_converter_destroy(CodeConverter *conv)
724 {
725         g_free(conv->charset_str);
726         g_free(conv);
727 }
728
729 gint conv_convert(CodeConverter *conv, gchar *outbuf, gint outlen,
730                   const gchar *inbuf)
731 {
732         if (conv->code_conv_func != conv_noconv)
733                 conv->code_conv_func(outbuf, outlen, inbuf);
734         else {
735                 gchar *str;
736
737                 str = conv_iconv_strdup(inbuf, conv->charset_str, NULL);
738                 if (!str)
739                         return -1;
740                 else {
741                         strncpy2(outbuf, str, outlen);
742                         g_free(str);
743                 }
744         }
745
746         return 0;
747 }
748
749 gchar *conv_codeset_strdup(const gchar *inbuf,
750                            const gchar *src_code, const gchar *dest_code)
751 {
752         gchar *buf;
753         size_t len;
754         CodeConvFunc conv_func;
755
756         src_code = conv_get_fallback_for_private_encoding(src_code);
757         conv_func = conv_get_code_conv_func(src_code, dest_code);
758         if (conv_func != conv_noconv) {
759                 len = (strlen(inbuf) + 1) * 3;
760                 buf = g_malloc(len);
761                 if (!buf) return NULL;
762
763                 conv_func(buf, len, inbuf);
764                 return g_realloc(buf, strlen(buf) + 1);
765         }
766
767         return conv_iconv_strdup(inbuf, src_code, dest_code);
768 }
769
770 CodeConvFunc conv_get_code_conv_func(const gchar *src_charset_str,
771                                      const gchar *dest_charset_str)
772 {
773         CodeConvFunc code_conv = conv_noconv;
774         CharSet src_charset;
775         CharSet dest_charset;
776
777         if (!src_charset_str)
778                 src_charset = conv_get_locale_charset();
779         else
780                 src_charset = conv_get_charset_from_str(src_charset_str);
781
782         /* auto detection mode */
783         if (!src_charset_str && !dest_charset_str) {
784                 if (conv_is_ja_locale())
785                         return conv_anytodisp;
786                 else
787                         return conv_noconv;
788         }
789
790         dest_charset = conv_get_charset_from_str(dest_charset_str);
791
792         if (dest_charset == C_US_ASCII)
793                 return conv_ustodisp;
794
795         switch (src_charset) {
796         case C_US_ASCII:
797         case C_ISO_8859_1:
798         case C_ISO_8859_2:
799         case C_ISO_8859_3:
800         case C_ISO_8859_4:
801         case C_ISO_8859_5:
802         case C_ISO_8859_6:
803         case C_ISO_8859_7:
804         case C_ISO_8859_8:
805         case C_ISO_8859_9:
806         case C_ISO_8859_10:
807         case C_ISO_8859_11:
808         case C_ISO_8859_13:
809         case C_ISO_8859_14:
810         case C_ISO_8859_15:
811                 break;
812         case C_ISO_2022_JP:
813         case C_ISO_2022_JP_2:
814         case C_ISO_2022_JP_3:
815                 if (dest_charset == C_AUTO)
816                         code_conv = conv_jistodisp;
817                 else if (dest_charset == C_EUC_JP)
818                         code_conv = conv_jistoeuc;
819                 else if (dest_charset == C_UTF_8)
820                         code_conv = conv_jistoutf8;
821                 break;
822         case C_SHIFT_JIS:
823                 if (dest_charset == C_AUTO)
824                         code_conv = conv_sjistodisp;
825                 else if (dest_charset == C_EUC_JP)
826                         code_conv = conv_sjistoeuc;
827                 else if (dest_charset == C_UTF_8)
828                         code_conv = conv_sjistoutf8;
829                 break;
830         case C_EUC_JP:
831                 if (dest_charset == C_AUTO)
832                         code_conv = conv_euctodisp;
833                 else if (dest_charset == C_ISO_2022_JP   ||
834                          dest_charset == C_ISO_2022_JP_2 ||
835                          dest_charset == C_ISO_2022_JP_3)
836                         code_conv = conv_euctojis;
837                 else if (dest_charset == C_UTF_8)
838                         code_conv = conv_euctoutf8;
839                 break;
840         case C_UTF_8:
841                 if (dest_charset == C_EUC_JP)
842                         code_conv = conv_utf8toeuc;
843                 else if (dest_charset == C_ISO_2022_JP   ||
844                          dest_charset == C_ISO_2022_JP_2 ||
845                          dest_charset == C_ISO_2022_JP_3)
846                         code_conv = conv_utf8tojis;
847                 break;
848         default:
849                 break;
850         }
851
852         return code_conv;
853 }
854
855 gchar *conv_iconv_strdup(const gchar *inbuf,
856                          const gchar *src_code, const gchar *dest_code)
857 {
858         iconv_t cd;
859         gchar *outbuf;
860
861         if (!src_code)
862                 src_code = conv_get_outgoing_charset_str();
863         if (!dest_code)
864                 dest_code = CS_INTERNAL;
865
866         /* don't convert if src and dest codeset are identical */
867         if (!strcasecmp(src_code, dest_code))
868                 return g_strdup(inbuf);
869
870         /* don't convert if current codeset is US-ASCII */
871         if (!strcasecmp(dest_code, CS_US_ASCII))
872                 return g_strdup(inbuf);
873
874         cd = iconv_open(dest_code, src_code);
875         if (cd == (iconv_t)-1)
876                 return NULL;
877
878         outbuf = conv_iconv_strdup_with_cd(inbuf, cd);
879
880         iconv_close(cd);
881
882         return outbuf;
883 }
884
885 gchar *conv_iconv_strdup_with_cd(const gchar *inbuf, iconv_t cd)
886 {
887         const gchar *inbuf_p;
888         gchar *outbuf;
889         gchar *outbuf_p;
890         size_t in_size;
891         size_t in_left;
892         size_t out_size;
893         size_t out_left;
894         size_t n_conv;
895         size_t len;
896
897         inbuf_p = inbuf;
898         in_size = strlen(inbuf);
899         in_left = in_size;
900         out_size = (in_size + 1) * 2;
901         outbuf = g_malloc(out_size);
902         outbuf_p = outbuf;
903         out_left = out_size;
904
905 #define EXPAND_BUF()                            \
906 {                                               \
907         len = outbuf_p - outbuf;                \
908         out_size *= 2;                          \
909         outbuf = g_realloc(outbuf, out_size);   \
910         outbuf_p = outbuf + len;                \
911         out_left = out_size - len;              \
912 }
913
914         while ((n_conv = iconv(cd, (ICONV_CONST gchar **)&inbuf_p, &in_left,
915                                &outbuf_p, &out_left)) == (size_t)-1) {
916                 if (EILSEQ == errno) {
917                         if (strict_mode) {
918                                 g_free(outbuf);
919                                 return NULL;
920                         }
921                         //g_print("iconv(): at %d: %s\n", in_size - in_left, g_strerror(errno));
922                         inbuf_p++;
923                         in_left--;
924                         if (out_left == 0) {
925                                 EXPAND_BUF();
926                         }
927                         *outbuf_p++ = SUBST_CHAR;
928                         out_left--;
929                 } else if (EINVAL == errno) {
930                         break;
931                 } else if (E2BIG == errno) {
932                         EXPAND_BUF();
933                 } else {
934                         g_warning("conv_iconv_strdup(): %s\n",
935                                   g_strerror(errno));
936                         break;
937                 }
938         }
939
940         while ((n_conv = iconv(cd, NULL, NULL, &outbuf_p, &out_left)) ==
941                (size_t)-1) {
942                 if (E2BIG == errno) {
943                         EXPAND_BUF();
944                 } else {
945                         g_warning("conv_iconv_strdup(): %s\n",
946                                   g_strerror(errno));
947                         break;
948                 }
949         }
950
951 #undef EXPAND_BUF
952
953         len = outbuf_p - outbuf;
954         outbuf = g_realloc(outbuf, len + 1);
955         outbuf[len] = '\0';
956
957         return outbuf;
958 }
959
960 static const struct {
961         CharSet charset;
962         gchar *const name;
963 } charsets[] = {
964         {C_US_ASCII,            CS_US_ASCII},
965         {C_US_ASCII,            CS_ANSI_X3_4_1968},
966         {C_UTF_8,               CS_UTF_8},
967         {C_UTF_7,               CS_UTF_7},
968         {C_ISO_8859_1,          CS_ISO_8859_1},
969         {C_ISO_8859_2,          CS_ISO_8859_2},
970         {C_ISO_8859_3,          CS_ISO_8859_3},
971         {C_ISO_8859_4,          CS_ISO_8859_4},
972         {C_ISO_8859_5,          CS_ISO_8859_5},
973         {C_ISO_8859_6,          CS_ISO_8859_6},
974         {C_ISO_8859_7,          CS_ISO_8859_7},
975         {C_ISO_8859_8,          CS_ISO_8859_8},
976         {C_ISO_8859_9,          CS_ISO_8859_9},
977         {C_ISO_8859_10,         CS_ISO_8859_10},
978         {C_ISO_8859_11,         CS_ISO_8859_11},
979         {C_ISO_8859_13,         CS_ISO_8859_13},
980         {C_ISO_8859_14,         CS_ISO_8859_14},
981         {C_ISO_8859_15,         CS_ISO_8859_15},
982         {C_BALTIC,              CS_BALTIC},
983         {C_CP1250,              CS_CP1250},
984         {C_CP1251,              CS_CP1251},
985         {C_CP1252,              CS_CP1252},
986         {C_CP1253,              CS_CP1253},
987         {C_CP1254,              CS_CP1254},
988         {C_CP1255,              CS_CP1255},
989         {C_CP1256,              CS_CP1256},
990         {C_CP1257,              CS_CP1257},
991         {C_CP1258,              CS_CP1258},
992         {C_WINDOWS_1250,        CS_WINDOWS_1250},
993         {C_WINDOWS_1251,        CS_WINDOWS_1251},
994         {C_WINDOWS_1252,        CS_WINDOWS_1252},
995         {C_WINDOWS_1253,        CS_WINDOWS_1253},
996         {C_WINDOWS_1254,        CS_WINDOWS_1254},
997         {C_WINDOWS_1255,        CS_WINDOWS_1255},
998         {C_WINDOWS_1256,        CS_WINDOWS_1256},
999         {C_WINDOWS_1257,        CS_WINDOWS_1257},
1000         {C_WINDOWS_1258,        CS_WINDOWS_1258},
1001         {C_KOI8_R,              CS_KOI8_R},
1002         {C_KOI8_T,              CS_KOI8_T},
1003         {C_KOI8_U,              CS_KOI8_U},
1004         {C_ISO_2022_JP,         CS_ISO_2022_JP},
1005         {C_ISO_2022_JP_2,       CS_ISO_2022_JP_2},
1006         {C_ISO_2022_JP_3,       CS_ISO_2022_JP_3},
1007         {C_EUC_JP,              CS_EUC_JP},
1008         {C_EUC_JP,              CS_EUCJP},
1009         {C_EUC_JP_MS,           CS_EUC_JP_MS},
1010         {C_SHIFT_JIS,           CS_SHIFT_JIS},
1011         {C_SHIFT_JIS,           CS_SHIFT__JIS},
1012         {C_SHIFT_JIS,           CS_SJIS},
1013         {C_ISO_2022_KR,         CS_ISO_2022_KR},
1014         {C_EUC_KR,              CS_EUC_KR},
1015         {C_ISO_2022_CN,         CS_ISO_2022_CN},
1016         {C_EUC_CN,              CS_EUC_CN},
1017         {C_GB2312,              CS_GB2312},
1018         {C_GBK,                 CS_GBK},
1019         {C_EUC_TW,              CS_EUC_TW},
1020         {C_BIG5,                CS_BIG5},
1021         {C_BIG5_HKSCS,          CS_BIG5_HKSCS},
1022         {C_TIS_620,             CS_TIS_620},
1023         {C_WINDOWS_874,         CS_WINDOWS_874},
1024         {C_GEORGIAN_PS,         CS_GEORGIAN_PS},
1025         {C_TCVN5712_1,          CS_TCVN5712_1},
1026 };
1027
1028 static const struct {
1029         gchar *const locale;
1030         CharSet charset;
1031         CharSet out_charset;
1032 } locale_table[] = {
1033         {"ja_JP.eucJP"  , C_EUC_JP      , C_ISO_2022_JP},
1034         {"ja_JP.EUC-JP" , C_EUC_JP      , C_ISO_2022_JP},
1035         {"ja_JP.EUC"    , C_EUC_JP      , C_ISO_2022_JP},
1036         {"ja_JP.ujis"   , C_EUC_JP      , C_ISO_2022_JP},
1037         {"ja_JP.SJIS"   , C_SHIFT_JIS   , C_ISO_2022_JP},
1038         {"ja_JP.JIS"    , C_ISO_2022_JP , C_ISO_2022_JP},
1039 #ifdef G_OS_WIN32
1040         {"ja_JP"        , C_SHIFT_JIS   , C_ISO_2022_JP},
1041 #else
1042         {"ja_JP"        , C_EUC_JP      , C_ISO_2022_JP},
1043 #endif
1044         {"ko_KR.EUC-KR" , C_EUC_KR      , C_EUC_KR},
1045         {"ko_KR"        , C_EUC_KR      , C_EUC_KR},
1046         {"zh_CN.GB2312" , C_GB2312      , C_GB2312},
1047         {"zh_CN.GBK"    , C_GBK         , C_GBK},
1048         {"zh_CN"        , C_GB2312      , C_GB2312},
1049         {"zh_HK"        , C_BIG5_HKSCS  , C_BIG5_HKSCS},
1050         {"zh_TW.eucTW"  , C_EUC_TW      , C_BIG5},
1051         {"zh_TW.EUC-TW" , C_EUC_TW      , C_BIG5},
1052         {"zh_TW.Big5"   , C_BIG5        , C_BIG5},
1053         {"zh_TW"        , C_BIG5        , C_BIG5},
1054
1055         {"ru_RU.KOI8-R" , C_KOI8_R      , C_KOI8_R},
1056         {"ru_RU.KOI8R"  , C_KOI8_R      , C_KOI8_R},
1057         {"ru_RU.CP1251" , C_WINDOWS_1251, C_KOI8_R},
1058         {"ru_RU"        , C_ISO_8859_5  , C_KOI8_R},
1059         {"tg_TJ"        , C_KOI8_T      , C_KOI8_T},
1060         {"ru_UA"        , C_KOI8_U      , C_KOI8_U},
1061         {"uk_UA.CP1251" , C_WINDOWS_1251, C_KOI8_U},
1062         {"uk_UA"        , C_KOI8_U      , C_KOI8_U},
1063
1064         {"be_BY"        , C_WINDOWS_1251, C_WINDOWS_1251},
1065         {"bg_BG"        , C_WINDOWS_1251, C_WINDOWS_1251},
1066
1067         {"yi_US"        , C_WINDOWS_1255, C_WINDOWS_1255},
1068
1069         {"af_ZA"        , C_ISO_8859_1  , C_ISO_8859_1},
1070         {"br_FR"        , C_ISO_8859_1  , C_ISO_8859_1},
1071         {"ca_ES"        , C_ISO_8859_1  , C_ISO_8859_1},
1072         {"da_DK"        , C_ISO_8859_1  , C_ISO_8859_1},
1073         {"de_AT"        , C_ISO_8859_1  , C_ISO_8859_1},
1074         {"de_BE"        , C_ISO_8859_1  , C_ISO_8859_1},
1075         {"de_CH"        , C_ISO_8859_1  , C_ISO_8859_1},
1076         {"de_DE"        , C_ISO_8859_1  , C_ISO_8859_1},
1077         {"de_LU"        , C_ISO_8859_1  , C_ISO_8859_1},
1078         {"en_AU"        , C_ISO_8859_1  , C_ISO_8859_1},
1079         {"en_BW"        , C_ISO_8859_1  , C_ISO_8859_1},
1080         {"en_CA"        , C_ISO_8859_1  , C_ISO_8859_1},
1081         {"en_DK"        , C_ISO_8859_1  , C_ISO_8859_1},
1082         {"en_GB"        , C_ISO_8859_1  , C_ISO_8859_1},
1083         {"en_HK"        , C_ISO_8859_1  , C_ISO_8859_1},
1084         {"en_IE"        , C_ISO_8859_1  , C_ISO_8859_1},
1085         {"en_NZ"        , C_ISO_8859_1  , C_ISO_8859_1},
1086         {"en_PH"        , C_ISO_8859_1  , C_ISO_8859_1},
1087         {"en_SG"        , C_ISO_8859_1  , C_ISO_8859_1},
1088         {"en_US"        , C_ISO_8859_1  , C_ISO_8859_1},
1089         {"en_ZA"        , C_ISO_8859_1  , C_ISO_8859_1},
1090         {"en_ZW"        , C_ISO_8859_1  , C_ISO_8859_1},
1091         {"es_AR"        , C_ISO_8859_1  , C_ISO_8859_1},
1092         {"es_BO"        , C_ISO_8859_1  , C_ISO_8859_1},
1093         {"es_CL"        , C_ISO_8859_1  , C_ISO_8859_1},
1094         {"es_CO"        , C_ISO_8859_1  , C_ISO_8859_1},
1095         {"es_CR"        , C_ISO_8859_1  , C_ISO_8859_1},
1096         {"es_DO"        , C_ISO_8859_1  , C_ISO_8859_1},
1097         {"es_EC"        , C_ISO_8859_1  , C_ISO_8859_1},
1098         {"es_ES"        , C_ISO_8859_1  , C_ISO_8859_1},
1099         {"es_GT"        , C_ISO_8859_1  , C_ISO_8859_1},
1100         {"es_HN"        , C_ISO_8859_1  , C_ISO_8859_1},
1101         {"es_MX"        , C_ISO_8859_1  , C_ISO_8859_1},
1102         {"es_NI"        , C_ISO_8859_1  , C_ISO_8859_1},
1103         {"es_PA"        , C_ISO_8859_1  , C_ISO_8859_1},
1104         {"es_PE"        , C_ISO_8859_1  , C_ISO_8859_1},
1105         {"es_PR"        , C_ISO_8859_1  , C_ISO_8859_1},
1106         {"es_PY"        , C_ISO_8859_1  , C_ISO_8859_1},
1107         {"es_SV"        , C_ISO_8859_1  , C_ISO_8859_1},
1108         {"es_US"        , C_ISO_8859_1  , C_ISO_8859_1},
1109         {"es_UY"        , C_ISO_8859_1  , C_ISO_8859_1},
1110         {"es_VE"        , C_ISO_8859_1  , C_ISO_8859_1},
1111         {"et_EE"        , C_ISO_8859_1  , C_ISO_8859_1},
1112         {"eu_ES"        , C_ISO_8859_1  , C_ISO_8859_1},
1113         {"fi_FI"        , C_ISO_8859_1  , C_ISO_8859_1},
1114         {"fo_FO"        , C_ISO_8859_1  , C_ISO_8859_1},
1115         {"fr_BE"        , C_ISO_8859_1  , C_ISO_8859_1},
1116         {"fr_CA"        , C_ISO_8859_1  , C_ISO_8859_1},
1117         {"fr_CH"        , C_ISO_8859_1  , C_ISO_8859_1},
1118         {"fr_FR"        , C_ISO_8859_1  , C_ISO_8859_1},
1119         {"fr_LU"        , C_ISO_8859_1  , C_ISO_8859_1},
1120         {"ga_IE"        , C_ISO_8859_1  , C_ISO_8859_1},
1121         {"gl_ES"        , C_ISO_8859_1  , C_ISO_8859_1},
1122         {"gv_GB"        , C_ISO_8859_1  , C_ISO_8859_1},
1123         {"id_ID"        , C_ISO_8859_1  , C_ISO_8859_1},
1124         {"is_IS"        , C_ISO_8859_1  , C_ISO_8859_1},
1125         {"it_CH"        , C_ISO_8859_1  , C_ISO_8859_1},
1126         {"it_IT"        , C_ISO_8859_1  , C_ISO_8859_1},
1127         {"kl_GL"        , C_ISO_8859_1  , C_ISO_8859_1},
1128         {"kw_GB"        , C_ISO_8859_1  , C_ISO_8859_1},
1129         {"ms_MY"        , C_ISO_8859_1  , C_ISO_8859_1},
1130         {"nl_BE"        , C_ISO_8859_1  , C_ISO_8859_1},
1131         {"nl_NL"        , C_ISO_8859_1  , C_ISO_8859_1},
1132         {"nn_NO"        , C_ISO_8859_1  , C_ISO_8859_1},
1133         {"no_NO"        , C_ISO_8859_1  , C_ISO_8859_1},
1134         {"oc_FR"        , C_ISO_8859_1  , C_ISO_8859_1},
1135         {"pt_BR"        , C_ISO_8859_1  , C_ISO_8859_1},
1136         {"pt_PT"        , C_ISO_8859_1  , C_ISO_8859_1},
1137         {"sq_AL"        , C_ISO_8859_1  , C_ISO_8859_1},
1138         {"sv_FI"        , C_ISO_8859_1  , C_ISO_8859_1},
1139         {"sv_SE"        , C_ISO_8859_1  , C_ISO_8859_1},
1140         {"tl_PH"        , C_ISO_8859_1  , C_ISO_8859_1},
1141         {"uz_UZ"        , C_ISO_8859_1  , C_ISO_8859_1},
1142         {"wa_BE"        , C_ISO_8859_1  , C_ISO_8859_1},
1143
1144         {"bs_BA"        , C_ISO_8859_2  , C_ISO_8859_2},
1145         {"cs_CZ"        , C_ISO_8859_2  , C_ISO_8859_2},
1146         {"hr_HR"        , C_ISO_8859_2  , C_ISO_8859_2},
1147         {"hu_HU"        , C_ISO_8859_2  , C_ISO_8859_2},
1148         {"pl_PL"        , C_ISO_8859_2  , C_ISO_8859_2},
1149         {"ro_RO"        , C_ISO_8859_2  , C_ISO_8859_2},
1150         {"sk_SK"        , C_ISO_8859_2  , C_ISO_8859_2},
1151         {"sl_SI"        , C_ISO_8859_2  , C_ISO_8859_2},
1152
1153         {"sr_YU@cyrillic"       , C_ISO_8859_5  , C_ISO_8859_5},
1154         {"sr_YU"                , C_ISO_8859_2  , C_ISO_8859_2},
1155
1156         {"mt_MT"                , C_ISO_8859_3  , C_ISO_8859_3},
1157
1158         {"lt_LT.iso88594"       , C_ISO_8859_4  , C_ISO_8859_4},
1159         {"lt_LT.ISO8859-4"      , C_ISO_8859_4  , C_ISO_8859_4},
1160         {"lt_LT.ISO_8859-4"     , C_ISO_8859_4  , C_ISO_8859_4},
1161         {"lt_LT"                , C_ISO_8859_13 , C_ISO_8859_13},
1162
1163         {"mk_MK"        , C_ISO_8859_5  , C_ISO_8859_5},
1164
1165         {"ar_AE"        , C_ISO_8859_6  , C_ISO_8859_6},
1166         {"ar_BH"        , C_ISO_8859_6  , C_ISO_8859_6},
1167         {"ar_DZ"        , C_ISO_8859_6  , C_ISO_8859_6},
1168         {"ar_EG"        , C_ISO_8859_6  , C_ISO_8859_6},
1169         {"ar_IQ"        , C_ISO_8859_6  , C_ISO_8859_6},
1170         {"ar_JO"        , C_ISO_8859_6  , C_ISO_8859_6},
1171         {"ar_KW"        , C_ISO_8859_6  , C_ISO_8859_6},
1172         {"ar_LB"        , C_ISO_8859_6  , C_ISO_8859_6},
1173         {"ar_LY"        , C_ISO_8859_6  , C_ISO_8859_6},
1174         {"ar_MA"        , C_ISO_8859_6  , C_ISO_8859_6},
1175         {"ar_OM"        , C_ISO_8859_6  , C_ISO_8859_6},
1176         {"ar_QA"        , C_ISO_8859_6  , C_ISO_8859_6},
1177         {"ar_SA"        , C_ISO_8859_6  , C_ISO_8859_6},
1178         {"ar_SD"        , C_ISO_8859_6  , C_ISO_8859_6},
1179         {"ar_SY"        , C_ISO_8859_6  , C_ISO_8859_6},
1180         {"ar_TN"        , C_ISO_8859_6  , C_ISO_8859_6},
1181         {"ar_YE"        , C_ISO_8859_6  , C_ISO_8859_6},
1182
1183         {"el_GR"        , C_ISO_8859_7  , C_ISO_8859_7},
1184         {"he_IL"        , C_ISO_8859_8  , C_ISO_8859_8},
1185         {"iw_IL"        , C_ISO_8859_8  , C_ISO_8859_8},
1186         {"tr_TR"        , C_ISO_8859_9  , C_ISO_8859_9},
1187
1188         {"lv_LV"        , C_ISO_8859_13 , C_ISO_8859_13},
1189         {"mi_NZ"        , C_ISO_8859_13 , C_ISO_8859_13},
1190
1191         {"cy_GB"        , C_ISO_8859_14 , C_ISO_8859_14},
1192
1193         {"ar_IN"        , C_UTF_8       , C_UTF_8},
1194         {"en_IN"        , C_UTF_8       , C_UTF_8},
1195         {"se_NO"        , C_UTF_8       , C_UTF_8},
1196         {"ta_IN"        , C_UTF_8       , C_UTF_8},
1197         {"te_IN"        , C_UTF_8       , C_UTF_8},
1198         {"ur_PK"        , C_UTF_8       , C_UTF_8},
1199
1200         {"th_TH"        , C_TIS_620     , C_TIS_620},
1201         /* {"th_TH"     , C_WINDOWS_874}, */
1202         /* {"th_TH"     , C_ISO_8859_11}, */
1203
1204         {"ka_GE"        , C_GEORGIAN_PS , C_GEORGIAN_PS},
1205         {"vi_VN.TCVN"   , C_TCVN5712_1  , C_TCVN5712_1},
1206
1207         {"C"                    , C_US_ASCII    , C_US_ASCII},
1208         {"POSIX"                , C_US_ASCII    , C_US_ASCII},
1209         {"ANSI_X3.4-1968"       , C_US_ASCII    , C_US_ASCII},
1210 };
1211
1212 static GHashTable *conv_get_charset_to_str_table(void)
1213 {
1214         static GHashTable *table;
1215         gint i;
1216
1217         if (table)
1218                 return table;
1219
1220         table = g_hash_table_new(NULL, g_direct_equal);
1221
1222         for (i = 0; i < sizeof(charsets) / sizeof(charsets[0]); i++) {
1223                 if (g_hash_table_lookup(table, GUINT_TO_POINTER(charsets[i].charset))
1224                     == NULL) {
1225                         g_hash_table_insert
1226                                 (table, GUINT_TO_POINTER(charsets[i].charset),
1227                                  charsets[i].name);
1228                 }
1229         }
1230
1231         return table;
1232 }
1233
1234 static GHashTable *conv_get_charset_from_str_table(void)
1235 {
1236         static GHashTable *table;
1237         gint i;
1238
1239         if (table)
1240                 return table;
1241
1242         table = g_hash_table_new(str_case_hash, str_case_equal);
1243
1244         for (i = 0; i < sizeof(charsets) / sizeof(charsets[0]); i++) {
1245                 g_hash_table_insert(table, charsets[i].name,
1246                                     GUINT_TO_POINTER(charsets[i].charset));
1247         }
1248
1249         return table;
1250 }
1251
1252 const gchar *conv_get_charset_str(CharSet charset)
1253 {
1254         GHashTable *table;
1255
1256         table = conv_get_charset_to_str_table();
1257         return g_hash_table_lookup(table, GUINT_TO_POINTER(charset));
1258 }
1259
1260 CharSet conv_get_charset_from_str(const gchar *charset)
1261 {
1262         GHashTable *table;
1263
1264         if (!charset) return C_AUTO;
1265
1266         table = conv_get_charset_from_str_table();
1267         return GPOINTER_TO_UINT(g_hash_table_lookup(table, charset));
1268 }
1269
1270 CharSet conv_get_locale_charset(void)
1271 {
1272         static CharSet cur_charset = -1;
1273         const gchar *cur_locale;
1274         const gchar *p;
1275         gint i;
1276
1277         if (cur_charset != -1)
1278                 return cur_charset;
1279
1280         cur_locale = conv_get_current_locale();
1281         if (!cur_locale) {
1282                 cur_charset = C_US_ASCII;
1283                 return cur_charset;
1284         }
1285
1286         if (strcasestr(cur_locale, "UTF-8")) {
1287                 cur_charset = C_UTF_8;
1288                 return cur_charset;
1289         }
1290
1291         if ((p = strcasestr(cur_locale, "@euro")) && p[5] == '\0') {
1292                 cur_charset = C_ISO_8859_15;
1293                 return cur_charset;
1294         }
1295
1296         for (i = 0; i < sizeof(locale_table) / sizeof(locale_table[0]); i++) {
1297                 const gchar *p;
1298
1299                 /* "ja_JP.EUC" matches with "ja_JP.eucJP", "ja_JP.EUC" and
1300                    "ja_JP". "ja_JP" matches with "ja_JP.xxxx" and "ja" */
1301                 if (!g_ascii_strncasecmp(cur_locale, locale_table[i].locale,
1302                                  strlen(locale_table[i].locale))) {
1303                         cur_charset = locale_table[i].charset;
1304                         return cur_charset;
1305                 } else if ((p = strchr(locale_table[i].locale, '_')) &&
1306                          !strchr(p + 1, '.')) {
1307                         if (strlen(cur_locale) == 2 &&
1308                             !g_ascii_strncasecmp(cur_locale, locale_table[i].locale, 2)) {
1309                                 cur_charset = locale_table[i].charset;
1310                                 return cur_charset;
1311                         }
1312                 }
1313         }
1314
1315         cur_charset = C_AUTO;
1316         return cur_charset;
1317 }
1318
1319 static CharSet conv_get_locale_charset_no_utf8(void)
1320 {
1321         static CharSet cur_charset = -1;
1322         const gchar *cur_locale;
1323         const gchar *p;
1324         gchar *tmp;
1325         gint i;
1326
1327         if (cur_charset != -1)
1328                 return cur_charset;
1329
1330         cur_locale = conv_get_current_locale();
1331         if (!cur_locale) {
1332                 cur_charset = C_US_ASCII;
1333                 return cur_charset;
1334         }
1335
1336         if (strcasestr(cur_locale, "UTF-8")) {
1337                 tmp = g_strdup(cur_locale);
1338                 *(strcasestr(tmp, ".UTF-8")) = '\0';
1339                 cur_locale = tmp;
1340         }
1341
1342         if ((p = strcasestr(cur_locale, "@euro")) && p[5] == '\0') {
1343                 cur_charset = C_ISO_8859_15;
1344                 return cur_charset;
1345         }
1346
1347         for (i = 0; i < sizeof(locale_table) / sizeof(locale_table[0]); i++) {
1348                 const gchar *p;
1349
1350                 /* "ja_JP.EUC" matches with "ja_JP.eucJP", "ja_JP.EUC" and
1351                    "ja_JP". "ja_JP" matches with "ja_JP.xxxx" and "ja" */
1352                 if (!g_ascii_strncasecmp(cur_locale, locale_table[i].locale,
1353                                  strlen(locale_table[i].locale))) {
1354                         cur_charset = locale_table[i].charset;
1355                         return cur_charset;
1356                 } else if ((p = strchr(locale_table[i].locale, '_')) &&
1357                          !strchr(p + 1, '.')) {
1358                         if (strlen(cur_locale) == 2 &&
1359                             !g_ascii_strncasecmp(cur_locale, locale_table[i].locale, 2)) {
1360                                 cur_charset = locale_table[i].charset;
1361                                 return cur_charset;
1362                         }
1363                 }
1364         }
1365
1366         cur_charset = C_AUTO;
1367         return cur_charset;
1368 }
1369
1370 const gchar *conv_get_locale_charset_str(void)
1371 {
1372         static const gchar *codeset = NULL;
1373
1374         if (!codeset)
1375                 codeset = conv_get_charset_str(conv_get_locale_charset());
1376
1377         return codeset ? codeset : CS_INTERNAL;
1378 }
1379
1380 const gchar *conv_get_locale_charset_str_no_utf8(void)
1381 {
1382         static const gchar *codeset = NULL;
1383
1384         if (!codeset)
1385                 codeset = conv_get_charset_str(conv_get_locale_charset_no_utf8());
1386
1387         return codeset ? codeset : CS_INTERNAL;
1388 }
1389
1390 CharSet conv_get_internal_charset(void)
1391 {
1392         return C_INTERNAL;
1393 }
1394
1395 const gchar *conv_get_internal_charset_str(void)
1396 {
1397         return CS_INTERNAL;
1398 }
1399
1400 CharSet conv_get_outgoing_charset(void)
1401 {
1402         static CharSet out_charset = -1;
1403         const gchar *cur_locale;
1404         const gchar *p;
1405         gint i;
1406
1407         if (out_charset != -1)
1408                 return out_charset;
1409
1410         cur_locale = conv_get_current_locale();
1411         if (!cur_locale) {
1412                 out_charset = C_AUTO;
1413                 return out_charset;
1414         }
1415
1416         if (strcasestr(cur_locale, "UTF-8")) {
1417                 out_charset = C_UTF_8;
1418                 return out_charset;
1419         }
1420
1421         if ((p = strcasestr(cur_locale, "@euro")) && p[5] == '\0') {
1422                 out_charset = C_ISO_8859_15;
1423                 return out_charset;
1424         }
1425
1426         for (i = 0; i < sizeof(locale_table) / sizeof(locale_table[0]); i++) {
1427                 const gchar *p;
1428
1429                 if (!g_ascii_strncasecmp(cur_locale, locale_table[i].locale,
1430                                  strlen(locale_table[i].locale))) {
1431                         out_charset = locale_table[i].out_charset;
1432                         break;
1433                 } else if ((p = strchr(locale_table[i].locale, '_')) &&
1434                          !strchr(p + 1, '.')) {
1435                         if (strlen(cur_locale) == 2 &&
1436                             !g_ascii_strncasecmp(cur_locale, locale_table[i].locale, 2)) {
1437                                 out_charset = locale_table[i].out_charset;
1438                                 break;
1439                         }
1440                 }
1441         }
1442
1443         return out_charset;
1444 }
1445
1446 const gchar *conv_get_outgoing_charset_str(void)
1447 {
1448         CharSet out_charset;
1449         const gchar *str;
1450
1451         out_charset = conv_get_outgoing_charset();
1452         str = conv_get_charset_str(out_charset);
1453
1454         return str ? str : CS_UTF_8;
1455 }
1456
1457 gboolean conv_is_multibyte_encoding(CharSet encoding)
1458 {
1459         switch (encoding) {
1460         case C_EUC_JP:
1461         case C_EUC_JP_MS:
1462         case C_EUC_KR:
1463         case C_EUC_TW:
1464         case C_EUC_CN:
1465         case C_ISO_2022_JP:
1466         case C_ISO_2022_JP_2:
1467         case C_ISO_2022_JP_3:
1468         case C_ISO_2022_KR:
1469         case C_ISO_2022_CN:
1470         case C_SHIFT_JIS:
1471         case C_GB2312:
1472         case C_GBK:
1473         case C_BIG5:
1474         case C_UTF_8:
1475         case C_UTF_7:
1476                 return TRUE;
1477         default:
1478                 return FALSE;
1479         }
1480 }
1481
1482 const gchar *conv_get_current_locale(void)
1483 {
1484         const gchar *cur_locale;
1485
1486 #ifdef G_OS_WIN32
1487         cur_locale = g_win32_getlocale();
1488 #else
1489         cur_locale = g_getenv("LC_ALL");
1490         if (!cur_locale) cur_locale = g_getenv("LC_CTYPE");
1491         if (!cur_locale) cur_locale = g_getenv("LANG");
1492         if (!cur_locale) cur_locale = setlocale(LC_CTYPE, NULL);
1493 #endif /* G_OS_WIN32 */
1494
1495         debug_print("current locale: %s\n",
1496                     cur_locale ? cur_locale : "(none)");
1497
1498         return cur_locale;
1499 }
1500
1501 gboolean conv_is_ja_locale(void)
1502 {
1503         static gint is_ja_locale = -1;
1504         const gchar *cur_locale;
1505
1506         if (is_ja_locale != -1)
1507                 return is_ja_locale != 0;
1508
1509         is_ja_locale = 0;
1510         cur_locale = conv_get_current_locale();
1511         if (cur_locale) {
1512                 if (g_ascii_strncasecmp(cur_locale, "ja", 2) == 0)
1513                         is_ja_locale = 1;
1514         }
1515
1516         return is_ja_locale != 0;
1517 }
1518
1519 gchar *conv_unmime_header(const gchar *str, const gchar *default_encoding)
1520 {
1521         gchar buf[BUFFSIZE];
1522
1523         if (is_ascii_str(str))
1524                 return unmime_header(str);
1525
1526         if (default_encoding) {
1527                 gchar *utf8_buf;
1528
1529                 utf8_buf = conv_codeset_strdup
1530                         (str, default_encoding, CS_INTERNAL);
1531                 if (utf8_buf) {
1532                         gchar *decoded_str;
1533
1534                         decoded_str = unmime_header(utf8_buf);
1535                         g_free(utf8_buf);
1536                         return decoded_str;
1537                 }
1538         }
1539
1540         if (conv_is_ja_locale())
1541                 conv_anytodisp(buf, sizeof(buf), str);
1542         else
1543                 conv_localetodisp(buf, sizeof(buf), str);
1544
1545         return unmime_header(buf);
1546 }
1547
1548 #define MAX_LINELEN             76
1549 #define MAX_HARD_LINELEN        996
1550 #define MIMESEP_BEGIN           "=?"
1551 #define MIMESEP_END             "?="
1552
1553 #define LBREAK_IF_REQUIRED(cond, is_plain_text)                         \
1554 {                                                                       \
1555         if (len - (destp - (guchar *)dest) < MAX_LINELEN + 2) {         \
1556                 *destp = '\0';                                          \
1557                 return;                                                 \
1558         }                                                               \
1559                                                                         \
1560         if ((cond) && *srcp) {                                          \
1561                 if (destp > (guchar *)dest && left < MAX_LINELEN - 1) { \
1562                         if (isspace(*(destp - 1)))                      \
1563                                 destp--;                                \
1564                         else if (is_plain_text && isspace(*srcp))       \
1565                                 srcp++;                                 \
1566                         if (*srcp) {                                    \
1567                                 *destp++ = '\n';                        \
1568                                 *destp++ = ' ';                         \
1569                                 left = MAX_LINELEN - 1;                 \
1570                         }                                               \
1571                 }                                                       \
1572         }                                                               \
1573 }
1574
1575 void conv_encode_header_full(gchar *dest, gint len, const gchar *src,
1576                         gint header_len, gboolean addr_field,
1577                         const gchar *out_encoding_)
1578 {
1579         const gchar *cur_encoding;
1580         const gchar *out_encoding;
1581         gint mimestr_len;
1582         gchar *mimesep_enc;
1583         gint left;
1584         const guchar *srcp = src;
1585         guchar *destp = dest;
1586         gboolean use_base64;
1587
1588         g_return_if_fail(g_utf8_validate(src, -1, NULL) == TRUE);
1589
1590         if (MB_CUR_MAX > 1) {
1591                 use_base64 = TRUE;
1592                 mimesep_enc = "?B?";
1593         } else {
1594                 use_base64 = FALSE;
1595                 mimesep_enc = "?Q?";
1596         }
1597
1598         cur_encoding = CS_INTERNAL;
1599
1600         if (out_encoding_)
1601                 out_encoding = out_encoding_;
1602         else
1603                 out_encoding = conv_get_outgoing_charset_str();
1604
1605         if (!strcmp(out_encoding, CS_US_ASCII))
1606                 out_encoding = CS_ISO_8859_1;
1607
1608         mimestr_len = strlen(MIMESEP_BEGIN) + strlen(out_encoding) +
1609                 strlen(mimesep_enc) + strlen(MIMESEP_END);
1610
1611         left = MAX_LINELEN - header_len;
1612
1613         while (*srcp) {
1614                 LBREAK_IF_REQUIRED(left <= 0, TRUE);
1615
1616                 while (isspace(*srcp)) {
1617                         *destp++ = *srcp++;
1618                         left--;
1619                         LBREAK_IF_REQUIRED(left <= 0, TRUE);
1620                 }
1621
1622                 /* output as it is if the next word is ASCII string */
1623                 if (!is_next_nonascii(srcp)) {
1624                         gint word_len;
1625
1626                         word_len = get_next_word_len(srcp);
1627                         LBREAK_IF_REQUIRED(left < word_len, TRUE);
1628                         while (word_len > 0) {
1629                                 LBREAK_IF_REQUIRED(left + (MAX_HARD_LINELEN - MAX_LINELEN) <= 0, TRUE)
1630                                 *destp++ = *srcp++;
1631                                 left--;
1632                                 word_len--;
1633                         }
1634
1635                         continue;
1636                 }
1637
1638                 /* don't include parentheses in encoded strings */
1639                 if (addr_field && (*srcp == '(' || *srcp == ')')) {
1640                         LBREAK_IF_REQUIRED(left < 2, FALSE);
1641                         *destp++ = *srcp++;
1642                         left--;
1643                 }
1644
1645                 while (1) {
1646                         gint mb_len = 0;
1647                         gint cur_len = 0;
1648                         gchar *part_str;
1649                         gchar *out_str;
1650                         gchar *enc_str;
1651                         const guchar *p = srcp;
1652                         gint out_str_len;
1653                         gint out_enc_str_len;
1654                         gint mime_block_len;
1655                         gboolean cont = FALSE;
1656
1657                         while (*p != '\0') {
1658                                 if (isspace(*p) && !is_next_nonascii(p + 1))
1659                                         break;
1660                                 /* don't include parentheses in encoded
1661                                    strings */
1662                                 if (addr_field && (*p == '(' || *p == ')'))
1663                                         break;
1664
1665                                 mb_len = g_utf8_skip[*p];
1666
1667                                 Xstrndup_a(part_str, srcp, cur_len + mb_len, );
1668                                 out_str = conv_codeset_strdup
1669                                         (part_str, cur_encoding, out_encoding);
1670                                 if (!out_str) {
1671                                         if (strict_mode) {
1672                                                 *dest = '\0';
1673                                                 return;
1674                                         } else {
1675                                                 g_warning("conv_encode_header(): code conversion failed\n");
1676                                                 conv_unreadable_8bit(part_str);
1677                                                 out_str = g_strdup(part_str);
1678                                         }
1679                                 }
1680                                 out_str_len = strlen(out_str);
1681
1682                                 if (use_base64)
1683                                         out_enc_str_len = B64LEN(out_str_len);
1684                                 else
1685                                         out_enc_str_len =
1686                                                 qp_get_q_encoding_len(out_str);
1687
1688                                 g_free(out_str);
1689
1690                                 if (mimestr_len + out_enc_str_len <= left) {
1691                                         cur_len += mb_len;
1692                                         p += mb_len;
1693                                 } else if (cur_len == 0) {
1694                                         LBREAK_IF_REQUIRED(1, FALSE);
1695                                         continue;
1696                                 } else {
1697                                         cont = TRUE;
1698                                         break;
1699                                 }
1700                         }
1701
1702                         if (cur_len > 0) {
1703                                 Xstrndup_a(part_str, srcp, cur_len, );
1704                                 out_str = conv_codeset_strdup
1705                                         (part_str, cur_encoding, out_encoding);
1706                                 if (!out_str) {
1707                                         g_warning("conv_encode_header(): code conversion failed\n");
1708                                         conv_unreadable_8bit(part_str);
1709                                         out_str = g_strdup(part_str);
1710                                 }
1711                                 out_str_len = strlen(out_str);
1712
1713                                 if (use_base64)
1714                                         out_enc_str_len = B64LEN(out_str_len);
1715                                 else
1716                                         out_enc_str_len =
1717                                                 qp_get_q_encoding_len(out_str);
1718
1719                                 Xalloca(enc_str, out_enc_str_len + 1, );
1720                                 if (use_base64)
1721                                         base64_encode(enc_str, out_str, out_str_len);
1722                                 else
1723                                         qp_q_encode(enc_str, out_str);
1724
1725                                 g_free(out_str);
1726
1727                                 /* output MIME-encoded string block */
1728                                 mime_block_len = mimestr_len + strlen(enc_str);
1729                                 g_snprintf(destp, mime_block_len + 1,
1730                                            MIMESEP_BEGIN "%s%s%s" MIMESEP_END,
1731                                            out_encoding, mimesep_enc, enc_str);
1732                                 destp += mime_block_len;
1733                                 srcp += cur_len;
1734
1735                                 left -= mime_block_len;
1736                         }
1737
1738                         LBREAK_IF_REQUIRED(cont, FALSE);
1739
1740                         if (cur_len == 0)
1741                                 break;
1742                 }
1743         }
1744
1745         *destp = '\0';
1746 }
1747
1748 void conv_encode_header(gchar *dest, gint len, const gchar *src,
1749                         gint header_len, gboolean addr_field)
1750 {
1751         conv_encode_header_full(dest,len,src,header_len,addr_field,NULL);
1752 }
1753
1754 #undef LBREAK_IF_REQUIRED
1755 gchar *conv_filename_from_utf8(const gchar *utf8_file)
1756 {
1757         gchar *fs_file;
1758         GError *error = NULL;
1759
1760         fs_file = g_filename_from_utf8(utf8_file, -1, NULL, NULL, &error);
1761         if (error) {
1762                 g_warning("failed to convert encoding of file name: %s\n",
1763                           error->message);
1764                 g_error_free(error);
1765         }
1766         if (!fs_file)
1767                 fs_file = g_strdup(utf8_file);
1768
1769         return fs_file;
1770 }
1771
1772 gchar *conv_filename_to_utf8(const gchar *fs_file)
1773 {
1774         gchar *utf8_file = NULL;
1775         GError *error = NULL;
1776
1777         utf8_file = g_filename_to_utf8(fs_file, -1, NULL, NULL, &error);
1778         if (error) {
1779                 g_warning("failed to convert encoding of file name: %s\n",
1780                           error->message);
1781                 g_error_free(error);
1782         }
1783
1784         if (!utf8_file || !g_utf8_validate(utf8_file, -1, NULL)) {
1785                 g_free(utf8_file);
1786                 utf8_file = g_strdup(fs_file);
1787                 conv_unreadable_8bit(utf8_file);
1788         }
1789
1790         return utf8_file;
1791 }