src/codeconv.c

   1 /*
   2  * Sylpheed -- a GTK+ based, lightweight, and fast e-mail client
   3  * Copyright (C) 1999-2009 Hiroyuki Yamamoto and the Claws Mail team
   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 3 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, see <http://www.gnu.org/licenses/>.
  17  *
  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 "codeconv.h"
  38 #include "unmime.h"
  39 #include "base64.h"
  40 #include "quoted-printable.h"
  41 #include "utils.h"
  42 #include "prefs_common.h"
  43
  44 /* For unknown reasons the inconv.m4 macro undefs that macro if no
  45    const is needed.  This would break the code below so we define it. */
  46 #ifndef ICONV_CONST
  47 #define ICONV_CONST
  48 #endif
  49
  50 typedef enum
  51 {
  52         JIS_ASCII,
  53         JIS_KANJI,
  54         JIS_HWKANA,
  55         JIS_AUXKANJI
  56 } JISState;
  57
  58 #define SUBST_CHAR      0x5f;
  59 #define ESC             '\033'
  60
  61 #define iseuckanji(c) \
  62         (((c) & 0xff) >= 0xa1 && ((c) & 0xff) <= 0xfe)
  63 #define iseuchwkana1(c) \
  64         (((c) & 0xff) == 0x8e)
  65 #define iseuchwkana2(c) \
  66         (((c) & 0xff) >= 0xa1 && ((c) & 0xff) <= 0xdf)
  67 #define iseucaux(c) \
  68         (((c) & 0xff) == 0x8f)
  69 #define issjiskanji1(c) \
  70         ((((c) & 0xff) >= 0x81 && ((c) & 0xff) <= 0x9f) || \
  71          (((c) & 0xff) >= 0xe0 && ((c) & 0xff) <= 0xfc))
  72 #define issjiskanji2(c) \
  73         ((((c) & 0xff) >= 0x40 && ((c) & 0xff) <= 0x7e) || \
  74          (((c) & 0xff) >= 0x80 && ((c) & 0xff) <= 0xfc))
  75 #define issjishwkana(c) \
  76         (((c) & 0xff) >= 0xa1 && ((c) & 0xff) <= 0xdf)
  77
  78 #define K_IN()                          \
  79         if (state != JIS_KANJI) {       \
  80                 *out++ = ESC;           \
  81                 *out++ = '$';           \
  82                 *out++ = 'B';           \
  83                 state = JIS_KANJI;      \
  84         }
  85
  86 #define K_OUT()                         \
  87         if (state != JIS_ASCII) {       \
  88                 *out++ = ESC;           \
  89                 *out++ = '(';           \
  90                 *out++ = 'B';           \
  91                 state = JIS_ASCII;      \
  92         }
  93
  94 #define HW_IN()                         \
  95         if (state != JIS_HWKANA) {      \
  96                 *out++ = ESC;           \
  97                 *out++ = '(';           \
  98                 *out++ = 'I';           \
  99                 state = JIS_HWKANA;     \
 100         }
 101
 102 #define AUX_IN()                        \
 103         if (state != JIS_AUXKANJI) {    \
 104                 *out++ = ESC;           \
 105                 *out++ = '$';           \
 106                 *out++ = '(';           \
 107                 *out++ = 'D';           \
 108                 state = JIS_AUXKANJI;   \
 109         }
 110
 111 static CodeConvFunc conv_get_code_conv_func     (const gchar    *src_charset_str,
 112                                          const gchar    *dest_charset_str);
 113
 114 static gchar *conv_iconv_strdup_with_cd (const gchar    *inbuf,
 115                                          iconv_t         cd);
 116
 117 static gchar *conv_iconv_strdup         (const gchar    *inbuf,
 118                                          const gchar    *src_code,
 119                                          const gchar    *dest_code);
 120
 121 static CharSet conv_get_locale_charset                  (void);
 122 static CharSet conv_get_outgoing_charset                (void);
 123 static CharSet conv_guess_ja_encoding(const gchar *str);
 124 static gboolean conv_is_ja_locale                       (void);
 125
 126 static gint conv_jistoeuc(gchar *outbuf, gint outlen, const gchar *inbuf);
 127 static gint conv_euctojis(gchar *outbuf, gint outlen, const gchar *inbuf);
 128 static gint conv_sjistoeuc(gchar *outbuf, gint outlen, const gchar *inbuf);
 129
 130 static gint conv_jistoutf8(gchar *outbuf, gint outlen, const gchar *inbuf);
 131 static gint conv_sjistoutf8(gchar *outbuf, gint outlen, const gchar *inbuf);
 132 static gint conv_euctoutf8(gchar *outbuf, gint outlen, const gchar *inbuf);
 133 static gint conv_anytoutf8(gchar *outbuf, gint outlen, const gchar *inbuf);
 134
 135 static gint conv_utf8toeuc(gchar *outbuf, gint outlen, const gchar *inbuf);
 136 static gint conv_utf8tojis(gchar *outbuf, gint outlen, const gchar *inbuf);
 137
 138 static void conv_unreadable_8bit(gchar *str);
 139
 140 static gint conv_jistodisp(gchar *outbuf, gint outlen, const gchar *inbuf);
 141 static gint conv_sjistodisp(gchar *outbuf, gint outlen, const gchar *inbuf);
 142 static gint conv_euctodisp(gchar *outbuf, gint outlen, const gchar *inbuf);
 143
 144 static gint conv_anytodisp(gchar *outbuf, gint outlen, const gchar *inbuf);
 145 static gint conv_ustodisp(gchar *outbuf, gint outlen, const gchar *inbuf);
 146 static gint conv_noconv(gchar *outbuf, gint outlen, const gchar *inbuf);
 147
 148 static gboolean strict_mode = FALSE;
 149
 150 void codeconv_set_strict(gboolean mode)
 151 {
 152         strict_mode = mode;
 153 }
 154
 155 static gint conv_jistoeuc(gchar *outbuf, gint outlen, const gchar *inbuf)
 156 {
 157         const guchar *in = inbuf;
 158         guchar *out = outbuf;
 159         JISState state = JIS_ASCII;
 160
 161         while (*in != '\0') {
 162                 if (*in == ESC) {
 163                         in++;
 164                         if (*in == '$') {
 165                                 if (*(in + 1) == '@' || *(in + 1) == 'B') {
 166                                         state = JIS_KANJI;
 167                                         in += 2;
 168                                 } else if (*(in + 1) == '(' &&
 169                                            *(in + 2) == 'D') {
 170                                         state = JIS_AUXKANJI;
 171                                         in += 3;
 172                                 } else {
 173                                         /* unknown escape sequence */
 174                                         state = JIS_ASCII;
 175                                 }
 176                         } else if (*in == '(') {
 177                                 if (*(in + 1) == 'B' || *(in + 1) == 'J') {
 178                                         state = JIS_ASCII;
 179                                         in += 2;
 180                                 } else if (*(in + 1) == 'I') {
 181                                         state = JIS_HWKANA;
 182                                         in += 2;
 183                                 } else {
 184                                         /* unknown escape sequence */
 185                                         state = JIS_ASCII;
 186                                 }
 187                         } else {
 188                                 /* unknown escape sequence */
 189                                 state = JIS_ASCII;
 190                         }
 191                 } else if (*in == 0x0e) {
 192                         state = JIS_HWKANA;
 193                         in++;
 194                 } else if (*in == 0x0f) {
 195                         state = JIS_ASCII;
 196                         in++;
 197                 } else {
 198                         switch (state) {
 199                         case JIS_ASCII:
 200                                 *out++ = *in++;
 201                                 break;
 202                         case JIS_KANJI:
 203                                 *out++ = *in++ | 0x80;
 204                                 if (*in == '\0') break;
 205                                 *out++ = *in++ | 0x80;
 206                                 break;
 207                         case JIS_HWKANA:
 208                                 *out++ = 0x8e;
 209                                 *out++ = *in++ | 0x80;
 210                                 break;
 211                         case JIS_AUXKANJI:
 212                                 *out++ = 0x8f;
 213                                 *out++ = *in++ | 0x80;
 214                                 if (*in == '\0') break;
 215                                 *out++ = *in++ | 0x80;
 216                                 break;
 217                         }
 218                 }
 219         }
 220
 221         *out = '\0';
 222         return 0;
 223 }
 224
 225 #define JIS_HWDAKUTEN           0x5e
 226 #define JIS_HWHANDAKUTEN        0x5f
 227
 228 static gint conv_jis_hantozen(guchar *outbuf, guchar jis_code, guchar sound_sym)
 229 {
 230         static guint16 h2z_tbl[] = {
 231                 /* 0x20 - 0x2f */
 232                 0x0000, 0x2123, 0x2156, 0x2157, 0x2122, 0x2126, 0x2572, 0x2521,
 233                 0x2523, 0x2525, 0x2527, 0x2529, 0x2563, 0x2565, 0x2567, 0x2543,
 234                 /* 0x30 - 0x3f */
 235                 0x213c, 0x2522, 0x2524, 0x2526, 0x2528, 0x252a, 0x252b, 0x252d,
 236                 0x252f, 0x2531, 0x2533, 0x2535, 0x2537, 0x2539, 0x253b, 0x253d,
 237                 /* 0x40 - 0x4f */
 238                 0x253f, 0x2541, 0x2544, 0x2546, 0x2548, 0x254a, 0x254b, 0x254c,
 239                 0x254d, 0x254e, 0x254f, 0x2552, 0x2555, 0x2558, 0x255b, 0x255e,
 240                 /* 0x50 - 0x5f */
 241                 0x255f, 0x2560, 0x2561, 0x2562, 0x2564, 0x2566, 0x2568, 0x2569,
 242                 0x256a, 0x256b, 0x256c, 0x256d, 0x256f, 0x2573, 0x212b, 0x212c
 243         };
 244
 245         static guint16 dakuten_tbl[] = {
 246                 /* 0x30 - 0x3f */
 247                 0x0000, 0x0000, 0x0000, 0x0000, 0x0000, 0x0000, 0x252c, 0x252e,
 248                 0x2530, 0x2532, 0x2534, 0x2536, 0x2538, 0x253a, 0x253c, 0x253e,
 249                 /* 0x40 - 0x4f */
 250                 0x2540, 0x2542, 0x2545, 0x2547, 0x2549, 0x0000, 0x0000, 0x0000,
 251                 0x0000, 0x0000, 0x2550, 0x2553, 0x2556, 0x2559, 0x255c, 0x0000
 252         };
 253
 254         static guint16 handakuten_tbl[] = {
 255                 /* 0x4a - 0x4e */
 256                 0x2551, 0x2554, 0x2557, 0x255a, 0x255d
 257         };
 258
 259         guint16 out_code;
 260
 261         jis_code &= 0x7f;
 262         sound_sym &= 0x7f;
 263
 264         if (jis_code < 0x21 || jis_code > 0x5f)
 265                 return 0;
 266
 267         if (sound_sym == JIS_HWDAKUTEN &&
 268             jis_code >= 0x36 && jis_code <= 0x4e) {
 269                 out_code = dakuten_tbl[jis_code - 0x30];
 270                 if (out_code != 0) {
 271                         *outbuf = out_code >> 8;
 272                         *(outbuf + 1) = out_code & 0xff;
 273                         return 2;
 274                 }
 275         }
 276
 277         if (sound_sym == JIS_HWHANDAKUTEN &&
 278             jis_code >= 0x4a && jis_code <= 0x4e) {
 279                 out_code = handakuten_tbl[jis_code - 0x4a];
 280                 *outbuf = out_code >> 8;
 281                 *(outbuf + 1) = out_code & 0xff;
 282                 return 2;
 283         }
 284
 285         out_code = h2z_tbl[jis_code - 0x20];
 286         *outbuf = out_code >> 8;
 287         *(outbuf + 1) = out_code & 0xff;
 288         return 1;
 289 }
 290
 291 static gint conv_euctojis(gchar *outbuf, gint outlen, const gchar *inbuf)
 292 {
 293         const guchar *in = inbuf;
 294         guchar *out = outbuf;
 295         JISState state = JIS_ASCII;
 296
 297         while (*in != '\0') {
 298                 if (IS_ASCII(*in)) {
 299                         K_OUT();
 300                         *out++ = *in++;
 301                 } else if (iseuckanji(*in)) {
 302                         if (iseuckanji(*(in + 1))) {
 303                                 K_IN();
 304                                 *out++ = *in++ & 0x7f;
 305                                 *out++ = *in++ & 0x7f;
 306                         } else {
 307                                 K_OUT();
 308                                 *out++ = SUBST_CHAR;
 309                                 in++;
 310                                 if (*in != '\0' && !IS_ASCII(*in)) {
 311                                         *out++ = SUBST_CHAR;
 312                                         in++;
 313                                 }
 314                         }
 315                 } else if (iseuchwkana1(*in)) {
 316                         if (iseuchwkana2(*(in + 1))) {
 317                                 if (prefs_common.allow_jisx0201_kana) {
 318                                         HW_IN();
 319                                         in++;
 320                                         *out++ = *in++ & 0x7f;
 321                                 } else {
 322                                         guchar jis_ch[2];
 323                                         gint len;
 324
 325                                         if (iseuchwkana1(*(in + 2)) &&
 326                                             iseuchwkana2(*(in + 3)))
 327                                                 len = conv_jis_hantozen
 328                                                         (jis_ch,
 329                                                          *(in + 1), *(in + 3));
 330                                         else
 331                                                 len = conv_jis_hantozen
 332                                                         (jis_ch,
 333                                                          *(in + 1), '\0');
 334                                         if (len == 0)
 335                                                 in += 2;
 336                                         else {
 337                                                 K_IN();
 338                                                 in += len * 2;
 339                                                 *out++ = jis_ch[0];
 340                                                 *out++ = jis_ch[1];
 341                                         }
 342                                 }
 343                         } else {
 344                                 K_OUT();
 345                                 in++;
 346                                 if (*in != '\0' && !IS_ASCII(*in)) {
 347                                         *out++ = SUBST_CHAR;
 348                                         in++;
 349                                 }
 350                         }
 351                 } else if (iseucaux(*in)) {
 352                         in++;
 353                         if (iseuckanji(*in) && iseuckanji(*(in + 1))) {
 354                                 AUX_IN();
 355                                 *out++ = *in++ & 0x7f;
 356                                 *out++ = *in++ & 0x7f;
 357                         } else {
 358                                 K_OUT();
 359                                 if (*in != '\0' && !IS_ASCII(*in)) {
 360                                         *out++ = SUBST_CHAR;
 361                                         in++;
 362                                         if (*in != '\0' && !IS_ASCII(*in)) {
 363                                                 *out++ = SUBST_CHAR;
 364                                                 in++;
 365                                         }
 366                                 }
 367                         }
 368                 } else {
 369                         K_OUT();
 370                         *out++ = SUBST_CHAR;
 371                         in++;
 372                 }
 373         }
 374
 375         K_OUT();
 376         *out = '\0';
 377         return 0;
 378 }
 379
 380 static gint conv_sjistoeuc(gchar *outbuf, gint outlen, const gchar *inbuf)
 381 {
 382         const guchar *in = inbuf;
 383         guchar *out = outbuf;
 384
 385         while (*in != '\0') {
 386                 if (IS_ASCII(*in)) {
 387                         *out++ = *in++;
 388                 } else if (issjiskanji1(*in)) {
 389                         if (issjiskanji2(*(in + 1))) {
 390                                 guchar out1 = *in;
 391                                 guchar out2 = *(in + 1);
 392                                 guchar row;
 393
 394                                 row = out1 < 0xa0 ? 0x70 : 0xb0;
 395                                 if (out2 < 0x9f) {
 396                                         out1 = (out1 - row) * 2 - 1;
 397                                         out2 -= out2 > 0x7f ? 0x20 : 0x1f;
 398                                 } else {
 399                                         out1 = (out1 - row) * 2;
 400                                         out2 -= 0x7e;
 401                                 }
 402
 403                                 *out++ = out1 | 0x80;
 404                                 *out++ = out2 | 0x80;
 405                                 in += 2;
 406                         } else {
 407                                 *out++ = SUBST_CHAR;
 408                                 in++;
 409                                 if (*in != '\0' && !IS_ASCII(*in)) {
 410                                         *out++ = SUBST_CHAR;
 411                                         in++;
 412                                 }
 413                         }
 414                 } else if (issjishwkana(*in)) {
 415                         *out++ = 0x8e;
 416                         *out++ = *in++;
 417                 } else {
 418                         *out++ = SUBST_CHAR;
 419                         in++;
 420                 }
 421         }
 422
 423         *out = '\0';
 424         return 0;
 425 }
 426
 427 static gint conv_jistoutf8(gchar *outbuf, gint outlen, const gchar *inbuf)
 428 {
 429         gchar *eucstr;
 430
 431         Xalloca(eucstr, outlen, return -1);
 432
 433         if (conv_jistoeuc(eucstr, outlen, inbuf) <0)
 434                 return -1;
 435         if (conv_euctoutf8(outbuf, outlen, eucstr) < 0)
 436                 return -1;
 437         return 0;
 438 }
 439
 440 static gint conv_sjistoutf8(gchar *outbuf, gint outlen, const gchar *inbuf)
 441 {
 442         gchar *tmpstr;
 443
 444         tmpstr = conv_iconv_strdup(inbuf, CS_SHIFT_JIS, CS_UTF_8);
 445         if (tmpstr) {
 446                 strncpy2(outbuf, tmpstr, outlen);
 447                 g_free(tmpstr);
 448                 return 0;
 449         } else {
 450                 strncpy2(outbuf, inbuf, outlen);
 451                 return -1;
 452         }
 453 }
 454
 455 static gint conv_euctoutf8(gchar *outbuf, gint outlen, const gchar *inbuf)
 456 {
 457         static iconv_t cd = (iconv_t)-1;
 458         static gboolean iconv_ok = TRUE;
 459         gchar *tmpstr;
 460
 461         if (cd == (iconv_t)-1) {
 462                 if (!iconv_ok) {
 463                         strncpy2(outbuf, inbuf, outlen);
 464                         return -1;
 465                 }
 466                 cd = iconv_open(CS_UTF_8, CS_EUC_JP_MS);
 467                 if (cd == (iconv_t)-1) {
 468                         cd = iconv_open(CS_UTF_8, CS_EUC_JP);
 469                         if (cd == (iconv_t)-1) {
 470                                 g_warning("conv_euctoutf8(): %s\n",
 471                                           g_strerror(errno));
 472                                 iconv_ok = FALSE;
 473                                 strncpy2(outbuf, inbuf, outlen);
 474                                 return -1;
 475                         }
 476                 }
 477         }
 478
 479         tmpstr = conv_iconv_strdup_with_cd(inbuf, cd);
 480         if (tmpstr) {
 481                 strncpy2(outbuf, tmpstr, outlen);
 482                 g_free(tmpstr);
 483                 return 0;
 484         } else {
 485                 strncpy2(outbuf, inbuf, outlen);
 486                 return -1;
 487         }
 488 }
 489
 490 static gint conv_anytoutf8(gchar *outbuf, gint outlen, const gchar *inbuf)
 491 {
 492         gint r = -1;
 493         switch (conv_guess_ja_encoding(inbuf)) {
 494         case C_ISO_2022_JP:
 495                 r = conv_jistoutf8(outbuf, outlen, inbuf);
 496                 break;
 497         case C_SHIFT_JIS:
 498                 r = conv_sjistoutf8(outbuf, outlen, inbuf);
 499                 break;
 500         case C_EUC_JP:
 501                 r = conv_euctoutf8(outbuf, outlen, inbuf);
 502                 break;
 503         default:
 504                 r = 0;
 505                 strncpy2(outbuf, inbuf, outlen);
 506                 break;
 507         }
 508
 509         return r;
 510 }
 511
 512 static gint conv_utf8toeuc(gchar *outbuf, gint outlen, const gchar *inbuf)
 513 {
 514         static iconv_t cd = (iconv_t)-1;
 515         static gboolean iconv_ok = TRUE;
 516         gchar *tmpstr;
 517
 518         if (cd == (iconv_t)-1) {
 519                 if (!iconv_ok) {
 520                         strncpy2(outbuf, inbuf, outlen);
 521                         return -1;
 522                 }
 523                 cd = iconv_open(CS_EUC_JP_MS, CS_UTF_8);
 524                 if (cd == (iconv_t)-1) {
 525                         cd = iconv_open(CS_EUC_JP, CS_UTF_8);
 526                         if (cd == (iconv_t)-1) {
 527                                 g_warning("conv_utf8toeuc(): %s\n",
 528                                           g_strerror(errno));
 529                                 iconv_ok = FALSE;
 530                                 strncpy2(outbuf, inbuf, outlen);
 531                                 return -1;
 532                         }
 533                 }
 534         }
 535
 536         tmpstr = conv_iconv_strdup_with_cd(inbuf, cd);
 537         if (tmpstr) {
 538                 strncpy2(outbuf, tmpstr, outlen);
 539                 g_free(tmpstr);
 540                 return 0;
 541         } else {
 542                 strncpy2(outbuf, inbuf, outlen);
 543                 return -1;
 544         }
 545 }
 546
 547 static gint conv_utf8tojis(gchar *outbuf, gint outlen, const gchar *inbuf)
 548 {
 549         gchar *eucstr;
 550
 551         Xalloca(eucstr, outlen, return -1);
 552
 553         if (conv_utf8toeuc(eucstr, outlen, inbuf) < 0)
 554                 return -1;
 555         if (conv_euctojis(outbuf, outlen, eucstr) < 0)
 556                 return -1;
 557
 558         return 0;
 559 }
 560
 561 static void conv_unreadable_8bit(gchar *str)
 562 {
 563         register guchar *p = str;
 564
 565         while (*p != '\0') {
 566                 /* convert CR+LF -> LF */
 567                 if (*p == '\r' && *(p + 1) == '\n')
 568                         memmove(p, p + 1, strlen(p));
 569                 else if (!IS_ASCII(*p)) *p = SUBST_CHAR;
 570                 p++;
 571         }
 572 }
 573
 574 static CharSet conv_guess_ja_encoding(const gchar *str)
 575 {
 576         const guchar *p = str;
 577         CharSet guessed = C_US_ASCII;
 578
 579         while (*p != '\0') {
 580                 if (*p == ESC && (*(p + 1) == '$' || *(p + 1) == '(')) {
 581                         if (guessed == C_US_ASCII)
 582                                 return C_ISO_2022_JP;
 583                         p += 2;
 584                 } else if (IS_ASCII(*p)) {
 585                         p++;
 586                 } else if (iseuckanji(*p) && iseuckanji(*(p + 1))) {
 587                         if (*p >= 0xfd && *p <= 0xfe)
 588                                 return C_EUC_JP;
 589                         else if (guessed == C_SHIFT_JIS) {
 590                                 if ((issjiskanji1(*p) &&
 591                                      issjiskanji2(*(p + 1))) ||
 592                                     issjishwkana(*p))
 593                                         guessed = C_SHIFT_JIS;
 594                                 else
 595                                         guessed = C_EUC_JP;
 596                         } else
 597                                 guessed = C_EUC_JP;
 598                         p += 2;
 599                 } else if (issjiskanji1(*p) && issjiskanji2(*(p + 1))) {
 600                         if (iseuchwkana1(*p) && iseuchwkana2(*(p + 1)))
 601                                 guessed = C_SHIFT_JIS;
 602                         else
 603                                 return C_SHIFT_JIS;
 604                         p += 2;
 605                 } else if (issjishwkana(*p)) {
 606                         guessed = C_SHIFT_JIS;
 607                         p++;
 608                 } else {
 609                         p++;
 610                 }
 611         }
 612
 613         return guessed;
 614 }
 615
 616 static gint conv_jistodisp(gchar *outbuf, gint outlen, const gchar *inbuf)
 617 {
 618         return conv_jistoutf8(outbuf, outlen, inbuf);
 619 }
 620
 621 static gint conv_sjistodisp(gchar *outbuf, gint outlen, const gchar *inbuf)
 622 {
 623         return conv_sjistoutf8(outbuf, outlen, inbuf);
 624 }
 625
 626 static gint conv_euctodisp(gchar *outbuf, gint outlen, const gchar *inbuf)
 627 {
 628         return conv_euctoutf8(outbuf, outlen, inbuf);
 629 }
 630
 631 void conv_utf8todisp(gchar *outbuf, gint outlen, const gchar *inbuf)
 632 {
 633         if (g_utf8_validate(inbuf, -1, NULL) == TRUE)
 634                 strncpy2(outbuf, inbuf, outlen);
 635         else
 636                 conv_ustodisp(outbuf, outlen, inbuf);
 637 }
 638
 639 static gint conv_anytodisp(gchar *outbuf, gint outlen, const gchar *inbuf)
 640 {
 641         gint r = 0;
 642         if (conv_anytoutf8(outbuf, outlen, inbuf) < 0)
 643                 r = -1;
 644         if (g_utf8_validate(outbuf, -1, NULL) != TRUE)
 645                 conv_unreadable_8bit(outbuf);
 646         return r;
 647 }
 648
 649 static gint conv_ustodisp(gchar *outbuf, gint outlen, const gchar *inbuf)
 650 {
 651         strncpy2(outbuf, inbuf, outlen);
 652         conv_unreadable_8bit(outbuf);
 653
 654         return 0;
 655 }
 656
 657 void conv_localetodisp(gchar *outbuf, gint outlen, const gchar *inbuf)
 658 {
 659         gchar *tmpstr;
 660
 661         codeconv_set_strict(TRUE);
 662         tmpstr = conv_iconv_strdup(inbuf, conv_get_locale_charset_str(),
 663                                    CS_INTERNAL);
 664         codeconv_set_strict(FALSE);
 665         if (tmpstr && g_utf8_validate(tmpstr, -1, NULL)) {
 666                 strncpy2(outbuf, tmpstr, outlen);
 667                 g_free(tmpstr);
 668                 return;
 669         } else if (tmpstr && !g_utf8_validate(tmpstr, -1, NULL)) {
 670                 g_free(tmpstr);
 671                 codeconv_set_strict(TRUE);
 672                 tmpstr = conv_iconv_strdup(inbuf,
 673                                 conv_get_locale_charset_str_no_utf8(),
 674                                 CS_INTERNAL);
 675                 codeconv_set_strict(FALSE);
 676         }
 677         if (tmpstr && g_utf8_validate(tmpstr, -1, NULL)) {
 678                 strncpy2(outbuf, tmpstr, outlen);
 679                 g_free(tmpstr);
 680                 return;
 681         } else {
 682                 g_free(tmpstr);
 683                 conv_utf8todisp(outbuf, outlen, inbuf);
 684         }
 685 }
 686
 687 static gint conv_noconv(gchar *outbuf, gint outlen, const gchar *inbuf)
 688 {
 689         strncpy2(outbuf, inbuf, outlen);
 690         return 0;
 691 }
 692
 693 static const gchar *
 694 conv_get_fallback_for_private_encoding(const gchar *encoding)
 695 {
 696         if (encoding && (encoding[0] == 'X' || encoding[0] == 'x') &&
 697             encoding[1] == '-') {
 698                 if (!g_ascii_strcasecmp(encoding, CS_X_GBK))
 699                         return CS_GBK;
 700         }
 701
 702         return encoding;
 703 }
 704
 705 CodeConverter *conv_code_converter_new(const gchar *src_charset)
 706 {
 707         CodeConverter *conv;
 708
 709         src_charset = conv_get_fallback_for_private_encoding(src_charset);
 710
 711         conv = g_new0(CodeConverter, 1);
 712         conv->code_conv_func = conv_get_code_conv_func(src_charset, NULL);
 713         conv->charset_str = g_strdup(src_charset);
 714         conv->charset = conv_get_charset_from_str(src_charset);
 715
 716         return conv;
 717 }
 718
 719 void conv_code_converter_destroy(CodeConverter *conv)
 720 {
 721         g_free(conv->charset_str);
 722         g_free(conv);
 723 }
 724
 725 gint conv_convert(CodeConverter *conv, gchar *outbuf, gint outlen,
 726                   const gchar *inbuf)
 727 {
 728         if (conv->code_conv_func != conv_noconv)
 729                 return conv->code_conv_func(outbuf, outlen, inbuf);
 730         else {
 731                 gchar *str;
 732
 733                 str = conv_iconv_strdup(inbuf, conv->charset_str, NULL);
 734                 if (!str)
 735                         return -1;
 736                 else {
 737                         strncpy2(outbuf, str, outlen);
 738                         g_free(str);
 739                 }
 740         }
 741
 742         return 0;
 743 }
 744
 745 gchar *conv_codeset_strdup(const gchar *inbuf,
 746                            const gchar *src_code, const gchar *dest_code)
 747 {
 748         gchar *buf;
 749         size_t len;
 750         CodeConvFunc conv_func;
 751
 752         if (!strcmp2(src_code, dest_code))
 753                 return g_strdup(inbuf);
 754
 755         src_code = conv_get_fallback_for_private_encoding(src_code);
 756         conv_func = conv_get_code_conv_func(src_code, dest_code);
 757         if (conv_func == conv_ustodisp && strict_mode && !is_ascii_str(inbuf))
 758                 return NULL;
 759
 760         if (conv_func != conv_noconv) {
 761                 len = (strlen(inbuf) + 1) * 3;
 762                 buf = g_malloc(len);
 763                 if (!buf) return NULL;
 764
 765                 if (conv_func(buf, len, inbuf) == 0 || !strict_mode)
 766                         return g_realloc(buf, strlen(buf) + 1);
 767                 else {
 768                         g_free(buf);
 769                         return NULL;
 770                 }
 771         }
 772
 773         return conv_iconv_strdup(inbuf, src_code, dest_code);
 774 }
 775
 776 static CodeConvFunc conv_get_code_conv_func(const gchar *src_charset_str,
 777                                      const gchar *dest_charset_str)
 778 {
 779         CodeConvFunc code_conv = conv_noconv;
 780         CharSet src_charset;
 781         CharSet dest_charset;
 782
 783         if (!src_charset_str)
 784                 src_charset = conv_get_locale_charset();
 785         else
 786                 src_charset = conv_get_charset_from_str(src_charset_str);
 787
 788         /* auto detection mode */
 789         if (!src_charset_str && !dest_charset_str) {
 790                 if (conv_is_ja_locale())
 791                         return conv_anytodisp;
 792                 else
 793                         return conv_noconv;
 794         }
 795
 796         dest_charset = conv_get_charset_from_str(dest_charset_str);
 797
 798         if (dest_charset == C_US_ASCII)
 799                 return conv_ustodisp;
 800
 801         switch (src_charset) {
 802         case C_US_ASCII:
 803         case C_ISO_8859_1:
 804         case C_ISO_8859_2:
 805         case C_ISO_8859_3:
 806         case C_ISO_8859_4:
 807         case C_ISO_8859_5:
 808         case C_ISO_8859_6:
 809         case C_ISO_8859_7:
 810         case C_ISO_8859_8:
 811         case C_ISO_8859_9:
 812         case C_ISO_8859_10:
 813         case C_ISO_8859_11:
 814         case C_ISO_8859_13:
 815         case C_ISO_8859_14:
 816         case C_ISO_8859_15:
 817                 break;
 818         case C_ISO_2022_JP:
 819         case C_ISO_2022_JP_2:
 820         case C_ISO_2022_JP_3:
 821                 if (dest_charset == C_AUTO)
 822                         code_conv = conv_jistodisp;
 823                 else if (dest_charset == C_EUC_JP)
 824                         code_conv = conv_jistoeuc;
 825                 else if (dest_charset == C_UTF_8)
 826                         code_conv = conv_jistoutf8;
 827                 break;
 828         case C_SHIFT_JIS:
 829                 if (dest_charset == C_AUTO)
 830                         code_conv = conv_sjistodisp;
 831                 else if (dest_charset == C_EUC_JP)
 832                         code_conv = conv_sjistoeuc;
 833                 else if (dest_charset == C_UTF_8)
 834                         code_conv = conv_sjistoutf8;
 835                 break;
 836         case C_EUC_JP:
 837                 if (dest_charset == C_AUTO)
 838                         code_conv = conv_euctodisp;
 839                 else if (dest_charset == C_ISO_2022_JP   ||
 840                          dest_charset == C_ISO_2022_JP_2 ||
 841                          dest_charset == C_ISO_2022_JP_3)
 842                         code_conv = conv_euctojis;
 843                 else if (dest_charset == C_UTF_8)
 844                         code_conv = conv_euctoutf8;
 845                 break;
 846         case C_UTF_8:
 847                 if (dest_charset == C_EUC_JP)
 848                         code_conv = conv_utf8toeuc;
 849                 else if (dest_charset == C_ISO_2022_JP   ||
 850                          dest_charset == C_ISO_2022_JP_2 ||
 851                          dest_charset == C_ISO_2022_JP_3)
 852                         code_conv = conv_utf8tojis;
 853                 break;
 854         default:
 855                 break;
 856         }
 857
 858         return code_conv;
 859 }
 860
 861 static gchar *conv_iconv_strdup(const gchar *inbuf,
 862                          const gchar *src_code, const gchar *dest_code)
 863 {
 864         iconv_t cd;
 865         gchar *outbuf;
 866
 867         if (!src_code && !dest_code &&
 868             g_utf8_validate(inbuf, -1, NULL))
 869                 return g_strdup(inbuf);
 870
 871         if (!src_code)
 872                 src_code = conv_get_outgoing_charset_str();
 873         if (!dest_code)
 874                 dest_code = CS_INTERNAL;
 875
 876         /* don't convert if src and dest codeset are identical */
 877         if (!strcasecmp(src_code, dest_code))
 878                 return g_strdup(inbuf);
 879
 880         /* don't convert if dest codeset is US-ASCII */
 881         if (!strcasecmp(src_code, CS_US_ASCII))
 882                 return g_strdup(inbuf);
 883
 884         /* don't convert if dest codeset is US-ASCII */
 885         if (!strcasecmp(dest_code, CS_US_ASCII))
 886                 return g_strdup(inbuf);
 887
 888         cd = iconv_open(dest_code, src_code);
 889         if (cd == (iconv_t)-1)
 890                 return NULL;
 891
 892         outbuf = conv_iconv_strdup_with_cd(inbuf, cd);
 893
 894         iconv_close(cd);
 895
 896         return outbuf;
 897 }
 898
 899 gchar *conv_iconv_strdup_with_cd(const gchar *inbuf, iconv_t cd)
 900 {
 901         const gchar *inbuf_p;
 902         gchar *outbuf;
 903         gchar *outbuf_p;
 904         size_t in_size;
 905         size_t in_left;
 906         size_t out_size;
 907         size_t out_left;
 908         size_t n_conv;
 909         size_t len;
 910
 911         inbuf_p = inbuf;
 912         in_size = strlen(inbuf);
 913         in_left = in_size;
 914         out_size = (in_size + 1) * 2;
 915         outbuf = g_malloc(out_size);
 916         outbuf_p = outbuf;
 917         out_left = out_size;
 918
 919 #define EXPAND_BUF()                            \
 920 {                                               \
 921         len = outbuf_p - outbuf;                \
 922         out_size *= 2;                          \
 923         outbuf = g_realloc(outbuf, out_size);   \
 924         outbuf_p = outbuf + len;                \
 925         out_left = out_size - len;              \
 926 }
 927
 928         while ((n_conv = iconv(cd, (ICONV_CONST gchar **)&inbuf_p, &in_left,
 929                                &outbuf_p, &out_left)) == (size_t)-1) {
 930                 if (EILSEQ == errno) {
 931                         if (strict_mode) {
 932                                 g_free(outbuf);
 933                                 return NULL;
 934                         }
 935                         //g_print("iconv(): at %d: %s\n", in_size - in_left, g_strerror(errno));
 936                         inbuf_p++;
 937                         in_left--;
 938                         if (out_left == 0) {
 939                                 EXPAND_BUF();
 940                         }
 941                         *outbuf_p++ = SUBST_CHAR;
 942                         out_left--;
 943                 } else if (EINVAL == errno) {
 944                         break;
 945                 } else if (E2BIG == errno) {
 946                         EXPAND_BUF();
 947                 } else {
 948                         g_warning("conv_iconv_strdup(): %s\n",
 949                                   g_strerror(errno));
 950                         break;
 951                 }
 952         }
 953
 954         while ((n_conv = iconv(cd, NULL, NULL, &outbuf_p, &out_left)) ==
 955                (size_t)-1) {
 956                 if (E2BIG == errno) {
 957                         EXPAND_BUF();
 958                 } else {
 959                         g_warning("conv_iconv_strdup(): %s\n",
 960                                   g_strerror(errno));
 961                         break;
 962                 }
 963         }
 964
 965 #undef EXPAND_BUF
 966
 967         len = outbuf_p - outbuf;
 968         outbuf = g_realloc(outbuf, len + 1);
 969         outbuf[len] = '\0';
 970
 971         return outbuf;
 972 }
 973
 974 static const struct {
 975         CharSet charset;
 976         gchar *const name;
 977 } charsets[] = {
 978         {C_US_ASCII,            CS_US_ASCII},
 979         {C_US_ASCII,            CS_ANSI_X3_4_1968},
 980         {C_UTF_8,               CS_UTF_8},
 981         {C_UTF_7,               CS_UTF_7},
 982         {C_ISO_8859_1,          CS_ISO_8859_1},
 983         {C_ISO_8859_2,          CS_ISO_8859_2},
 984         {C_ISO_8859_3,          CS_ISO_8859_3},
 985         {C_ISO_8859_4,          CS_ISO_8859_4},
 986         {C_ISO_8859_5,          CS_ISO_8859_5},
 987         {C_ISO_8859_6,          CS_ISO_8859_6},
 988         {C_ISO_8859_7,          CS_ISO_8859_7},
 989         {C_ISO_8859_8,          CS_ISO_8859_8},
 990         {C_ISO_8859_9,          CS_ISO_8859_9},
 991         {C_ISO_8859_10,         CS_ISO_8859_10},
 992         {C_ISO_8859_11,         CS_ISO_8859_11},
 993         {C_ISO_8859_13,         CS_ISO_8859_13},
 994         {C_ISO_8859_14,         CS_ISO_8859_14},
 995         {C_ISO_8859_15,         CS_ISO_8859_15},
 996         {C_BALTIC,              CS_BALTIC},
 997         {C_CP1250,              CS_CP1250},
 998         {C_CP1251,              CS_CP1251},
 999         {C_CP1252,              CS_CP1252},
1000         {C_CP1253,              CS_CP1253},
1001         {C_CP1254,              CS_CP1254},
1002         {C_CP1255,              CS_CP1255},
1003         {C_CP1256,              CS_CP1256},
1004         {C_CP1257,              CS_CP1257},
1005         {C_CP1258,              CS_CP1258},
1006         {C_WINDOWS_1250,        CS_WINDOWS_1250},
1007         {C_WINDOWS_1251,        CS_WINDOWS_1251},
1008         {C_WINDOWS_1252,        CS_WINDOWS_1252},
1009         {C_WINDOWS_1253,        CS_WINDOWS_1253},
1010         {C_WINDOWS_1254,        CS_WINDOWS_1254},
1011         {C_WINDOWS_1255,        CS_WINDOWS_1255},
1012         {C_WINDOWS_1256,        CS_WINDOWS_1256},
1013         {C_WINDOWS_1257,        CS_WINDOWS_1257},
1014         {C_WINDOWS_1258,        CS_WINDOWS_1258},
1015         {C_KOI8_R,              CS_KOI8_R},
1016         {C_KOI8_T,              CS_KOI8_T},
1017         {C_KOI8_U,              CS_KOI8_U},
1018         {C_ISO_2022_JP,         CS_ISO_2022_JP},
1019         {C_ISO_2022_JP_2,       CS_ISO_2022_JP_2},
1020         {C_ISO_2022_JP_3,       CS_ISO_2022_JP_3},
1021         {C_EUC_JP,              CS_EUC_JP},
1022         {C_EUC_JP,              CS_EUCJP},
1023         {C_EUC_JP_MS,           CS_EUC_JP_MS},
1024         {C_SHIFT_JIS,           CS_SHIFT_JIS},
1025         {C_SHIFT_JIS,           CS_SHIFT__JIS},
1026         {C_SHIFT_JIS,           CS_SJIS},
1027         {C_ISO_2022_KR,         CS_ISO_2022_KR},
1028         {C_EUC_KR,              CS_EUC_KR},
1029         {C_ISO_2022_CN,         CS_ISO_2022_CN},
1030         {C_EUC_CN,              CS_EUC_CN},
1031         {C_GB2312,              CS_GB2312},
1032         {C_GBK,                 CS_GBK},
1033         {C_EUC_TW,              CS_EUC_TW},
1034         {C_BIG5,                CS_BIG5},
1035         {C_BIG5_HKSCS,          CS_BIG5_HKSCS},
1036         {C_TIS_620,             CS_TIS_620},
1037         {C_WINDOWS_874,         CS_WINDOWS_874},
1038         {C_GEORGIAN_PS,         CS_GEORGIAN_PS},
1039         {C_TCVN5712_1,          CS_TCVN5712_1},
1040 };
1041
1042 static const struct {
1043         gchar *const locale;
1044         CharSet charset;
1045         CharSet out_charset;
1046 } locale_table[] = {
1047         {"ja_JP.eucJP"  , C_EUC_JP      , C_ISO_2022_JP},
1048         {"ja_JP.EUC-JP" , C_EUC_JP      , C_ISO_2022_JP},
1049         {"ja_JP.EUC"    , C_EUC_JP      , C_ISO_2022_JP},
1050         {"ja_JP.ujis"   , C_EUC_JP      , C_ISO_2022_JP},
1051         {"ja_JP.SJIS"   , C_SHIFT_JIS   , C_ISO_2022_JP},
1052         {"ja_JP.JIS"    , C_ISO_2022_JP , C_ISO_2022_JP},
1053 #ifdef G_OS_WIN32
1054         {"ja_JP"        , C_SHIFT_JIS   , C_ISO_2022_JP},
1055 #else
1056         {"ja_JP"        , C_EUC_JP      , C_ISO_2022_JP},
1057 #endif
1058         {"ko_KR.EUC-KR" , C_EUC_KR      , C_EUC_KR},
1059         {"ko_KR"        , C_EUC_KR      , C_EUC_KR},
1060         {"zh_CN.GB2312" , C_GB2312      , C_GB2312},
1061         {"zh_CN.GBK"    , C_GBK         , C_GBK},
1062         {"zh_CN"        , C_GB2312      , C_GB2312},
1063         {"zh_HK"        , C_BIG5_HKSCS  , C_BIG5_HKSCS},
1064         {"zh_TW.eucTW"  , C_EUC_TW      , C_BIG5},
1065         {"zh_TW.EUC-TW" , C_EUC_TW      , C_BIG5},
1066         {"zh_TW.Big5"   , C_BIG5        , C_BIG5},
1067         {"zh_TW"        , C_BIG5        , C_BIG5},
1068
1069         {"ru_RU.KOI8-R" , C_KOI8_R      , C_KOI8_R},
1070         {"ru_RU.KOI8R"  , C_KOI8_R      , C_KOI8_R},
1071         {"ru_RU.CP1251" , C_WINDOWS_1251, C_KOI8_R},
1072         {"ru_RU"        , C_ISO_8859_5  , C_KOI8_R},
1073         {"tg_TJ"        , C_KOI8_T      , C_KOI8_T},
1074         {"ru_UA"        , C_KOI8_U      , C_KOI8_U},
1075         {"uk_UA.CP1251" , C_WINDOWS_1251, C_KOI8_U},
1076         {"uk_UA"        , C_KOI8_U      , C_KOI8_U},
1077
1078         {"be_BY"        , C_WINDOWS_1251, C_WINDOWS_1251},
1079         {"bg_BG"        , C_WINDOWS_1251, C_WINDOWS_1251},
1080
1081         {"yi_US"        , C_WINDOWS_1255, C_WINDOWS_1255},
1082
1083         {"af_ZA"        , C_ISO_8859_1  , C_ISO_8859_1},
1084         {"br_FR"        , C_ISO_8859_1  , C_ISO_8859_1},
1085         {"ca_ES"        , C_ISO_8859_1  , C_ISO_8859_1},
1086         {"da_DK"        , C_ISO_8859_1  , C_ISO_8859_1},
1087         {"de_AT"        , C_ISO_8859_1  , C_ISO_8859_1},
1088         {"de_BE"        , C_ISO_8859_1  , C_ISO_8859_1},
1089         {"de_CH"        , C_ISO_8859_1  , C_ISO_8859_1},
1090         {"de_DE"        , C_ISO_8859_1  , C_ISO_8859_1},
1091         {"de_LU"        , C_ISO_8859_1  , C_ISO_8859_1},
1092         {"en_AU"        , C_ISO_8859_1  , C_ISO_8859_1},
1093         {"en_BW"        , C_ISO_8859_1  , C_ISO_8859_1},
1094         {"en_CA"        , C_ISO_8859_1  , C_ISO_8859_1},
1095         {"en_DK"        , C_ISO_8859_1  , C_ISO_8859_1},
1096         {"en_GB"        , C_ISO_8859_1  , C_ISO_8859_1},
1097         {"en_HK"        , C_ISO_8859_1  , C_ISO_8859_1},
1098         {"en_IE"        , C_ISO_8859_1  , C_ISO_8859_1},
1099         {"en_NZ"        , C_ISO_8859_1  , C_ISO_8859_1},
1100         {"en_PH"        , C_ISO_8859_1  , C_ISO_8859_1},
1101         {"en_SG"        , C_ISO_8859_1  , C_ISO_8859_1},
1102         {"en_US"        , C_ISO_8859_1  , C_ISO_8859_1},
1103         {"en_ZA"        , C_ISO_8859_1  , C_ISO_8859_1},
1104         {"en_ZW"        , C_ISO_8859_1  , C_ISO_8859_1},
1105         {"es_AR"        , C_ISO_8859_1  , C_ISO_8859_1},
1106         {"es_BO"        , C_ISO_8859_1  , C_ISO_8859_1},
1107         {"es_CL"        , C_ISO_8859_1  , C_ISO_8859_1},
1108         {"es_CO"        , C_ISO_8859_1  , C_ISO_8859_1},
1109         {"es_CR"        , C_ISO_8859_1  , C_ISO_8859_1},
1110         {"es_DO"        , C_ISO_8859_1  , C_ISO_8859_1},
1111         {"es_EC"        , C_ISO_8859_1  , C_ISO_8859_1},
1112         {"es_ES"        , C_ISO_8859_1  , C_ISO_8859_1},
1113         {"es_GT"        , C_ISO_8859_1  , C_ISO_8859_1},
1114         {"es_HN"        , C_ISO_8859_1  , C_ISO_8859_1},
1115         {"es_MX"        , C_ISO_8859_1  , C_ISO_8859_1},
1116         {"es_NI"        , C_ISO_8859_1  , C_ISO_8859_1},
1117         {"es_PA"        , C_ISO_8859_1  , C_ISO_8859_1},
1118         {"es_PE"        , C_ISO_8859_1  , C_ISO_8859_1},
1119         {"es_PR"        , C_ISO_8859_1  , C_ISO_8859_1},
1120         {"es_PY"        , C_ISO_8859_1  , C_ISO_8859_1},
1121         {"es_SV"        , C_ISO_8859_1  , C_ISO_8859_1},
1122         {"es_US"        , C_ISO_8859_1  , C_ISO_8859_1},
1123         {"es_UY"        , C_ISO_8859_1  , C_ISO_8859_1},
1124         {"es_VE"        , C_ISO_8859_1  , C_ISO_8859_1},
1125         {"et_EE"        , C_ISO_8859_1  , C_ISO_8859_1},
1126         {"eu_ES"        , C_ISO_8859_1  , C_ISO_8859_1},
1127         {"fi_FI"        , C_ISO_8859_1  , C_ISO_8859_1},
1128         {"fo_FO"        , C_ISO_8859_1  , C_ISO_8859_1},
1129         {"fr_BE"        , C_ISO_8859_1  , C_ISO_8859_1},
1130         {"fr_CA"        , C_ISO_8859_1  , C_ISO_8859_1},
1131         {"fr_CH"        , C_ISO_8859_1  , C_ISO_8859_1},
1132         {"fr_FR"        , C_ISO_8859_1  , C_ISO_8859_1},
1133         {"fr_LU"        , C_ISO_8859_1  , C_ISO_8859_1},
1134         {"ga_IE"        , C_ISO_8859_1  , C_ISO_8859_1},
1135         {"gl_ES"        , C_ISO_8859_1  , C_ISO_8859_1},
1136         {"gv_GB"        , C_ISO_8859_1  , C_ISO_8859_1},
1137         {"id_ID"        , C_ISO_8859_1  , C_ISO_8859_1},
1138         {"is_IS"        , C_ISO_8859_1  , C_ISO_8859_1},
1139         {"it_CH"        , C_ISO_8859_1  , C_ISO_8859_1},
1140         {"it_IT"        , C_ISO_8859_1  , C_ISO_8859_1},
1141         {"kl_GL"        , C_ISO_8859_1  , C_ISO_8859_1},
1142         {"kw_GB"        , C_ISO_8859_1  , C_ISO_8859_1},
1143         {"ms_MY"        , C_ISO_8859_1  , C_ISO_8859_1},
1144         {"nl_BE"        , C_ISO_8859_1  , C_ISO_8859_1},
1145         {"nl_NL"        , C_ISO_8859_1  , C_ISO_8859_1},
1146         {"nb_NO"        , C_ISO_8859_1  , C_ISO_8859_1},
1147         {"nn_NO"        , C_ISO_8859_1  , C_ISO_8859_1},
1148         {"no_NO"        , C_ISO_8859_1  , C_ISO_8859_1},
1149         {"oc_FR"        , C_ISO_8859_1  , C_ISO_8859_1},
1150         {"pt_BR"        , C_ISO_8859_1  , C_ISO_8859_1},
1151         {"pt_PT"        , C_ISO_8859_1  , C_ISO_8859_1},
1152         {"sq_AL"        , C_ISO_8859_1  , C_ISO_8859_1},
1153         {"sv_FI"        , C_ISO_8859_1  , C_ISO_8859_1},
1154         {"sv_SE"        , C_ISO_8859_1  , C_ISO_8859_1},
1155         {"tl_PH"        , C_ISO_8859_1  , C_ISO_8859_1},
1156         {"uz_UZ"        , C_ISO_8859_1  , C_ISO_8859_1},
1157         {"wa_BE"        , C_ISO_8859_1  , C_ISO_8859_1},
1158
1159         {"bs_BA"        , C_ISO_8859_2  , C_ISO_8859_2},
1160         {"cs_CZ"        , C_ISO_8859_2  , C_ISO_8859_2},
1161         {"hr_HR"        , C_ISO_8859_2  , C_ISO_8859_2},
1162         {"hu_HU"        , C_ISO_8859_2  , C_ISO_8859_2},
1163         {"pl_PL"        , C_ISO_8859_2  , C_ISO_8859_2},
1164         {"ro_RO"        , C_ISO_8859_2  , C_ISO_8859_2},
1165         {"sk_SK"        , C_ISO_8859_2  , C_ISO_8859_2},
1166         {"sl_SI"        , C_ISO_8859_2  , C_ISO_8859_2},
1167
1168         {"sr_YU@cyrillic"       , C_ISO_8859_5  , C_ISO_8859_5},
1169         {"sr_YU"                , C_ISO_8859_2  , C_ISO_8859_2},
1170
1171         {"mt_MT"                , C_ISO_8859_3  , C_ISO_8859_3},
1172
1173         {"lt_LT.iso88594"       , C_ISO_8859_4  , C_ISO_8859_4},
1174         {"lt_LT.ISO8859-4"      , C_ISO_8859_4  , C_ISO_8859_4},
1175         {"lt_LT.ISO_8859-4"     , C_ISO_8859_4  , C_ISO_8859_4},
1176         {"lt_LT"                , C_ISO_8859_13 , C_ISO_8859_13},
1177
1178         {"mk_MK"        , C_ISO_8859_5  , C_ISO_8859_5},
1179
1180         {"ar_AE"        , C_ISO_8859_6  , C_ISO_8859_6},
1181         {"ar_BH"        , C_ISO_8859_6  , C_ISO_8859_6},
1182         {"ar_DZ"        , C_ISO_8859_6  , C_ISO_8859_6},
1183         {"ar_EG"        , C_ISO_8859_6  , C_ISO_8859_6},
1184         {"ar_IQ"        , C_ISO_8859_6  , C_ISO_8859_6},
1185         {"ar_JO"        , C_ISO_8859_6  , C_ISO_8859_6},
1186         {"ar_KW"        , C_ISO_8859_6  , C_ISO_8859_6},
1187         {"ar_LB"        , C_ISO_8859_6  , C_ISO_8859_6},
1188         {"ar_LY"        , C_ISO_8859_6  , C_ISO_8859_6},
1189         {"ar_MA"        , C_ISO_8859_6  , C_ISO_8859_6},
1190         {"ar_OM"        , C_ISO_8859_6  , C_ISO_8859_6},
1191         {"ar_QA"        , C_ISO_8859_6  , C_ISO_8859_6},
1192         {"ar_SA"        , C_ISO_8859_6  , C_ISO_8859_6},
1193         {"ar_SD"        , C_ISO_8859_6  , C_ISO_8859_6},
1194         {"ar_SY"        , C_ISO_8859_6  , C_ISO_8859_6},
1195         {"ar_TN"        , C_ISO_8859_6  , C_ISO_8859_6},
1196         {"ar_YE"        , C_ISO_8859_6  , C_ISO_8859_6},
1197
1198         {"el_GR"        , C_ISO_8859_7  , C_ISO_8859_7},
1199         {"he_IL"        , C_ISO_8859_8  , C_ISO_8859_8},
1200         {"iw_IL"        , C_ISO_8859_8  , C_ISO_8859_8},
1201         {"tr_TR"        , C_ISO_8859_9  , C_ISO_8859_9},
1202
1203         {"lv_LV"        , C_ISO_8859_13 , C_ISO_8859_13},
1204         {"mi_NZ"        , C_ISO_8859_13 , C_ISO_8859_13},
1205
1206         {"cy_GB"        , C_ISO_8859_14 , C_ISO_8859_14},
1207
1208         {"ar_IN"        , C_UTF_8       , C_UTF_8},
1209         {"en_IN"        , C_UTF_8       , C_UTF_8},
1210         {"se_NO"        , C_UTF_8       , C_UTF_8},
1211         {"ta_IN"        , C_UTF_8       , C_UTF_8},
1212         {"te_IN"        , C_UTF_8       , C_UTF_8},
1213         {"ur_PK"        , C_UTF_8       , C_UTF_8},
1214
1215         {"th_TH"        , C_TIS_620     , C_TIS_620},
1216         /* {"th_TH"     , C_WINDOWS_874}, */
1217         /* {"th_TH"     , C_ISO_8859_11}, */
1218
1219         {"ka_GE"        , C_GEORGIAN_PS , C_GEORGIAN_PS},
1220         {"vi_VN.TCVN"   , C_TCVN5712_1  , C_TCVN5712_1},
1221
1222         {"C"                    , C_US_ASCII    , C_US_ASCII},
1223         {"POSIX"                , C_US_ASCII    , C_US_ASCII},
1224         {"ANSI_X3.4-1968"       , C_US_ASCII    , C_US_ASCII},
1225 };
1226
1227 static GHashTable *conv_get_charset_to_str_table(void)
1228 {
1229         static GHashTable *table;
1230         gint i;
1231
1232         if (table)
1233                 return table;
1234
1235         table = g_hash_table_new(NULL, g_direct_equal);
1236
1237         for (i = 0; i < sizeof(charsets) / sizeof(charsets[0]); i++) {
1238                 if (g_hash_table_lookup(table, GUINT_TO_POINTER(charsets[i].charset))
1239                     == NULL) {
1240                         g_hash_table_insert
1241                                 (table, GUINT_TO_POINTER(charsets[i].charset),
1242                                  charsets[i].name);
1243                 }
1244         }
1245
1246         return table;
1247 }
1248
1249 static GHashTable *conv_get_charset_from_str_table(void)
1250 {
1251         static GHashTable *table;
1252         gint i;
1253
1254         if (table)
1255                 return table;
1256
1257         table = g_hash_table_new(str_case_hash, str_case_equal);
1258
1259         for (i = 0; i < sizeof(charsets) / sizeof(charsets[0]); i++) {
1260                 g_hash_table_insert(table, charsets[i].name,
1261                                     GUINT_TO_POINTER(charsets[i].charset));
1262         }
1263
1264         return table;
1265 }
1266
1267 const gchar *conv_get_charset_str(CharSet charset)
1268 {
1269         GHashTable *table;
1270
1271         table = conv_get_charset_to_str_table();
1272         return g_hash_table_lookup(table, GUINT_TO_POINTER(charset));
1273 }
1274
1275 CharSet conv_get_charset_from_str(const gchar *charset)
1276 {
1277         GHashTable *table;
1278
1279         if (!charset) return C_AUTO;
1280
1281         table = conv_get_charset_from_str_table();
1282         return GPOINTER_TO_UINT(g_hash_table_lookup(table, charset));
1283 }
1284
1285 static CharSet conv_get_locale_charset(void)
1286 {
1287         static CharSet cur_charset = -1;
1288         const gchar *cur_locale;
1289         const gchar *p;
1290         gint i;
1291
1292         if (cur_charset != -1)
1293                 return cur_charset;
1294
1295         cur_locale = conv_get_current_locale();
1296         if (!cur_locale) {
1297                 cur_charset = C_US_ASCII;
1298                 return cur_charset;
1299         }
1300
1301         if (strcasestr(cur_locale, ".UTF-8") ||
1302             strcasestr(cur_locale, ".utf8")) {
1303                 cur_charset = C_UTF_8;
1304                 return cur_charset;
1305         }
1306
1307         if ((p = strcasestr(cur_locale, "@euro")) && p[5] == '\0') {
1308                 cur_charset = C_ISO_8859_15;
1309                 return cur_charset;
1310         }
1311
1312         for (i = 0; i < sizeof(locale_table) / sizeof(locale_table[0]); i++) {
1313                 const gchar *p;
1314
1315                 /* "ja_JP.EUC" matches with "ja_JP.eucJP", "ja_JP.EUC" and
1316                    "ja_JP". "ja_JP" matches with "ja_JP.xxxx" and "ja" */
1317                 if (!g_ascii_strncasecmp(cur_locale, locale_table[i].locale,
1318                                  strlen(locale_table[i].locale))) {
1319                         cur_charset = locale_table[i].charset;
1320                         return cur_charset;
1321                 } else if ((p = strchr(locale_table[i].locale, '_')) &&
1322                          !strchr(p + 1, '.')) {
1323                         if (strlen(cur_locale) == 2 &&
1324                             !g_ascii_strncasecmp(cur_locale, locale_table[i].locale, 2)) {
1325                                 cur_charset = locale_table[i].charset;
1326                                 return cur_charset;
1327                         }
1328                 }
1329         }
1330
1331         cur_charset = C_AUTO;
1332         return cur_charset;
1333 }
1334
1335 static CharSet conv_get_locale_charset_no_utf8(void)
1336 {
1337         static CharSet cur_charset = -1;
1338         const gchar *cur_locale;
1339         const gchar *p;
1340         gchar *tmp;
1341         gint i;
1342
1343         if (prefs_common.broken_are_utf8)
1344                 return conv_get_locale_charset();
1345
1346         if (cur_charset != -1)
1347                 return cur_charset;
1348
1349         cur_locale = conv_get_current_locale();
1350         if (!cur_locale) {
1351                 cur_charset = C_US_ASCII;
1352                 return cur_charset;
1353         }
1354
1355         if (strcasestr(cur_locale, "UTF-8")) {
1356                 tmp = g_strdup(cur_locale);
1357                 *(strcasestr(tmp, ".UTF-8")) = '\0';
1358                 cur_locale = tmp;
1359         }
1360
1361         if ((p = strcasestr(cur_locale, "@euro")) && p[5] == '\0') {
1362                 cur_charset = C_ISO_8859_15;
1363                 return cur_charset;
1364         }
1365
1366         for (i = 0; i < sizeof(locale_table) / sizeof(locale_table[0]); i++) {
1367                 const gchar *p;
1368
1369                 /* "ja_JP.EUC" matches with "ja_JP.eucJP", "ja_JP.EUC" and
1370                    "ja_JP". "ja_JP" matches with "ja_JP.xxxx" and "ja" */
1371                 if (!g_ascii_strncasecmp(cur_locale, locale_table[i].locale,
1372                                  strlen(locale_table[i].locale))) {
1373                         cur_charset = locale_table[i].charset;
1374                         return cur_charset;
1375                 } else if ((p = strchr(locale_table[i].locale, '_')) &&
1376                          !strchr(p + 1, '.')) {
1377                         if (strlen(cur_locale) == 2 &&
1378                             !g_ascii_strncasecmp(cur_locale, locale_table[i].locale, 2)) {
1379                                 cur_charset = locale_table[i].charset;
1380                                 return cur_charset;
1381                         }
1382                 }
1383         }
1384
1385         cur_charset = C_AUTO;
1386         return cur_charset;
1387 }
1388
1389 const gchar *conv_get_locale_charset_str(void)
1390 {
1391         static const gchar *codeset = NULL;
1392
1393         if (!codeset)
1394                 codeset = conv_get_charset_str(conv_get_locale_charset());
1395
1396         return codeset ? codeset : CS_INTERNAL;
1397 }
1398
1399 const gchar *conv_get_locale_charset_str_no_utf8(void)
1400 {
1401         static const gchar *codeset = NULL;
1402
1403         if (!codeset)
1404                 codeset = conv_get_charset_str(conv_get_locale_charset_no_utf8());
1405
1406         return codeset ? codeset : CS_INTERNAL;
1407 }
1408
1409 static CharSet conv_get_outgoing_charset(void)
1410 {
1411         static CharSet out_charset = -1;
1412         const gchar *cur_locale;
1413         const gchar *p;
1414         gint i;
1415
1416         if (out_charset != -1)
1417                 return out_charset;
1418
1419         cur_locale = conv_get_current_locale();
1420         if (!cur_locale) {
1421                 out_charset = C_AUTO;
1422                 return out_charset;
1423         }
1424
1425         if (strcasestr(cur_locale, "UTF-8")) {
1426                 out_charset = C_UTF_8;
1427                 return out_charset;
1428         }
1429
1430         if ((p = strcasestr(cur_locale, "@euro")) && p[5] == '\0') {
1431                 out_charset = C_ISO_8859_15;
1432                 return out_charset;
1433         }
1434
1435         for (i = 0; i < sizeof(locale_table) / sizeof(locale_table[0]); i++) {
1436                 const gchar *p;
1437
1438                 if (!g_ascii_strncasecmp(cur_locale, locale_table[i].locale,
1439                                  strlen(locale_table[i].locale))) {
1440                         out_charset = locale_table[i].out_charset;
1441                         break;
1442                 } else if ((p = strchr(locale_table[i].locale, '_')) &&
1443                          !strchr(p + 1, '.')) {
1444                         if (strlen(cur_locale) == 2 &&
1445                             !g_ascii_strncasecmp(cur_locale, locale_table[i].locale, 2)) {
1446                                 out_charset = locale_table[i].out_charset;
1447                                 break;
1448                         }
1449                 }
1450         }
1451
1452         return out_charset;
1453 }
1454
1455 const gchar *conv_get_outgoing_charset_str(void)
1456 {
1457         CharSet out_charset;
1458         const gchar *str;
1459
1460         out_charset = conv_get_outgoing_charset();
1461         str = conv_get_charset_str(out_charset);
1462
1463         return str ? str : CS_UTF_8;
1464 }
1465
1466 const gchar *conv_get_current_locale(void)
1467 {
1468         const gchar *cur_locale;
1469
1470 #ifdef G_OS_WIN32
1471         cur_locale = g_win32_getlocale();
1472 #else
1473         cur_locale = g_getenv("LC_ALL");
1474         if (!cur_locale) cur_locale = g_getenv("LC_CTYPE");
1475         if (!cur_locale) cur_locale = g_getenv("LANG");
1476         if (!cur_locale) cur_locale = setlocale(LC_CTYPE, NULL);
1477 #endif /* G_OS_WIN32 */
1478
1479         debug_print("current locale: %s\n",
1480                     cur_locale ? cur_locale : "(none)");
1481
1482         return cur_locale;
1483 }
1484
1485 static gboolean conv_is_ja_locale(void)
1486 {
1487         static gint is_ja_locale = -1;
1488         const gchar *cur_locale;
1489
1490         if (is_ja_locale != -1)
1491                 return is_ja_locale != 0;
1492
1493         is_ja_locale = 0;
1494         cur_locale = conv_get_current_locale();
1495         if (cur_locale) {
1496                 if (g_ascii_strncasecmp(cur_locale, "ja", 2) == 0)
1497                         is_ja_locale = 1;
1498         }
1499
1500         return is_ja_locale != 0;
1501 }
1502
1503 gchar *conv_unmime_header(const gchar *str, const gchar *default_encoding)
1504 {
1505         gchar buf[BUFFSIZE];
1506
1507         if (is_ascii_str(str))
1508                 return unmime_header(str);
1509
1510         if (default_encoding) {
1511                 gchar *utf8_buf;
1512
1513                 utf8_buf = conv_codeset_strdup
1514                         (str, default_encoding, CS_INTERNAL);
1515                 if (utf8_buf) {
1516                         gchar *decoded_str;
1517
1518                         decoded_str = unmime_header(utf8_buf);
1519                         g_free(utf8_buf);
1520                         return decoded_str;
1521                 }
1522         }
1523
1524         if (conv_is_ja_locale())
1525                 conv_anytodisp(buf, sizeof(buf), str);
1526         else
1527                 conv_localetodisp(buf, sizeof(buf), str);
1528
1529         return unmime_header(buf);
1530 }
1531
1532 #define MAX_LINELEN             76
1533 #define MAX_HARD_LINELEN        996
1534 #define MIMESEP_BEGIN           "=?"
1535 #define MIMESEP_END             "?="
1536
1537 #define LBREAK_IF_REQUIRED(cond, is_plain_text)                         \
1538 {                                                                       \
1539         if (len - (destp - (guchar *)dest) < MAX_LINELEN + 2) {         \
1540                 *destp = '\0';                                          \
1541                 return;                                                 \
1542         }                                                               \
1543                                                                         \
1544         if ((cond) && *srcp) {                                          \
1545                 if (destp > (guchar *)dest && left < MAX_LINELEN - 1) { \
1546                         if (isspace(*(destp - 1)))                      \
1547                                 destp--;                                \
1548                         else if (is_plain_text && isspace(*srcp))       \
1549                                 srcp++;                                 \
1550                         if (*srcp) {                                    \
1551                                 *destp++ = '\n';                        \
1552                                 *destp++ = ' ';                         \
1553                                 left = MAX_LINELEN - 1;                 \
1554                         }                                               \
1555                 } else if (destp == (guchar *)dest && left < 7) {       \
1556                         if (isspace(*(destp - 1)))                      \
1557                                 destp--;                                \
1558                         else if (is_plain_text && isspace(*srcp))       \
1559                                 srcp++;                                 \
1560                         if (*srcp) {                                    \
1561                                 *destp++ = '\n';                        \
1562                                 *destp++ = ' ';                         \
1563                                 left = MAX_LINELEN - 1;                 \
1564                         }                                               \
1565                 }                                                       \
1566         }                                                               \
1567 }
1568
1569 void conv_encode_header_full(gchar *dest, gint len, const gchar *src,
1570                         gint header_len, gboolean addr_field,
1571                         const gchar *out_encoding_)
1572 {
1573         const gchar *cur_encoding;
1574         const gchar *out_encoding;
1575         gint mimestr_len;
1576         gchar *mimesep_enc;
1577         gint left;
1578         const guchar *srcp = src;
1579         guchar *destp = dest;
1580         gboolean use_base64;
1581
1582         cm_return_if_fail(g_utf8_validate(src, -1, NULL) == TRUE);
1583         cm_return_if_fail(destp != NULL);
1584
1585         if (MB_CUR_MAX > 1) {
1586                 use_base64 = TRUE;
1587                 mimesep_enc = "?B?";
1588         } else {
1589                 use_base64 = FALSE;
1590                 mimesep_enc = "?Q?";
1591         }
1592
1593         cur_encoding = CS_INTERNAL;
1594
1595         if (out_encoding_)
1596                 out_encoding = out_encoding_;
1597         else
1598                 out_encoding = conv_get_outgoing_charset_str();
1599
1600         if (!strcmp(out_encoding, CS_US_ASCII))
1601                 out_encoding = CS_ISO_8859_1;
1602
1603         mimestr_len = strlen(MIMESEP_BEGIN) + strlen(out_encoding) +
1604                 strlen(mimesep_enc) + strlen(MIMESEP_END);
1605
1606         left = MAX_LINELEN - header_len;
1607
1608         while (*srcp) {
1609                 LBREAK_IF_REQUIRED(left <= 0, TRUE);
1610
1611                 while (isspace(*srcp)) {
1612                         *destp++ = *srcp++;
1613                         left--;
1614                         LBREAK_IF_REQUIRED(left <= 0, TRUE);
1615                 }
1616
1617                 /* output as it is if the next word is ASCII string */
1618                 if (!is_next_nonascii(srcp)) {
1619                         gint word_len;
1620
1621                         word_len = get_next_word_len(srcp);
1622                         LBREAK_IF_REQUIRED(left < word_len, TRUE);
1623                         while (word_len > 0) {
1624                                 LBREAK_IF_REQUIRED(left + (MAX_HARD_LINELEN - MAX_LINELEN) <= 0, TRUE)
1625                                 *destp++ = *srcp++;
1626                                 left--;
1627                                 word_len--;
1628                         }
1629
1630                         continue;
1631                 }
1632
1633                 /* don't include parentheses and quotes in encoded strings */
1634                 if (addr_field && (*srcp == '(' || *srcp == ')' || *srcp == '"')) {
1635                         LBREAK_IF_REQUIRED(left < 2, FALSE);
1636                         *destp++ = *srcp++;
1637                         left--;
1638                 }
1639
1640                 while (1) {
1641                         gint mb_len = 0;
1642                         gint cur_len = 0;
1643                         gchar *part_str;
1644                         gchar *out_str;
1645                         gchar *enc_str;
1646                         const guchar *p = srcp;
1647                         gint out_str_len;
1648                         gint out_enc_str_len;
1649                         gint mime_block_len;
1650                         gboolean cont = FALSE;
1651
1652                         while (*p != '\0') {
1653                                 if (isspace(*p) && !is_next_nonascii(p + 1))
1654                                         break;
1655                                 /* don't include parentheses in encoded
1656                                    strings */
1657                                 if (addr_field && (*p == '(' || *p == ')' || *p == '"'))
1658                                         break;
1659
1660                                 mb_len = g_utf8_skip[*p];
1661
1662                                 Xstrndup_a(part_str, srcp, cur_len + mb_len, );
1663                                 out_str = conv_codeset_strdup
1664                                         (part_str, cur_encoding, out_encoding);
1665                                 if (!out_str) {
1666                                         if (strict_mode) {
1667                                                 *dest = '\0';
1668                                                 return;
1669                                         } else {
1670                                                 g_warning("conv_encode_header(): code conversion failed\n");
1671                                                 conv_unreadable_8bit(part_str);
1672                                                 out_str = g_strdup(part_str);
1673                                         }
1674                                 }
1675                                 out_str_len = strlen(out_str);
1676
1677                                 if (use_base64)
1678                                         out_enc_str_len = B64LEN(out_str_len);
1679                                 else
1680                                         out_enc_str_len =
1681                                                 qp_get_q_encoding_len(out_str);
1682
1683                                 g_free(out_str);
1684
1685                                 if (mimestr_len + out_enc_str_len <= left) {
1686                                         cur_len += mb_len;
1687                                         p += mb_len;
1688                                 } else if (cur_len == 0) {
1689                                         LBREAK_IF_REQUIRED(1, FALSE);
1690                                         continue;
1691                                 } else {
1692                                         cont = TRUE;
1693                                         break;
1694                                 }
1695                         }
1696
1697                         if (cur_len > 0) {
1698                                 Xstrndup_a(part_str, srcp, cur_len, );
1699                                 out_str = conv_codeset_strdup
1700                                         (part_str, cur_encoding, out_encoding);
1701                                 if (!out_str) {
1702                                         g_warning("conv_encode_header(): code conversion failed\n");
1703                                         conv_unreadable_8bit(part_str);
1704                                         out_str = g_strdup(part_str);
1705                                 }
1706                                 out_str_len = strlen(out_str);
1707
1708                                 if (use_base64)
1709                                         out_enc_str_len = B64LEN(out_str_len);
1710                                 else
1711                                         out_enc_str_len =
1712                                                 qp_get_q_encoding_len(out_str);
1713
1714                                 Xalloca(enc_str, out_enc_str_len + 1, );
1715                                 if (use_base64)
1716                                         base64_encode(enc_str, out_str, out_str_len);
1717                                 else
1718                                         qp_q_encode(enc_str, out_str);
1719
1720                                 g_free(out_str);
1721
1722                                 /* output MIME-encoded string block */
1723                                 mime_block_len = mimestr_len + strlen(enc_str);
1724                                 g_snprintf(destp, mime_block_len + 1,
1725                                            MIMESEP_BEGIN "%s%s%s" MIMESEP_END,
1726                                            out_encoding, mimesep_enc, enc_str);
1727                                 destp += mime_block_len;
1728                                 srcp += cur_len;
1729
1730                                 left -= mime_block_len;
1731                         }
1732
1733                         LBREAK_IF_REQUIRED(cont, FALSE);
1734
1735                         if (cur_len == 0)
1736                                 break;
1737                 }
1738         }
1739
1740         *destp = '\0';
1741 }
1742
1743 void conv_encode_header(gchar *dest, gint len, const gchar *src,
1744                         gint header_len, gboolean addr_field)
1745 {
1746         conv_encode_header_full(dest,len,src,header_len,addr_field,NULL);
1747 }
1748
1749 #undef LBREAK_IF_REQUIRED
1750 gchar *conv_filename_from_utf8(const gchar *utf8_file)
1751 {
1752         gchar *fs_file;
1753         GError *error = NULL;
1754
1755         fs_file = g_filename_from_utf8(utf8_file, -1, NULL, NULL, &error);
1756         if (error) {
1757                 debug_print("failed to convert encoding of file name: %s\n",
1758                           error->message);
1759                 g_error_free(error);
1760         }
1761         if (!fs_file)
1762                 fs_file = g_strdup(utf8_file);
1763
1764         return fs_file;
1765 }
1766
1767 gchar *conv_filename_to_utf8(const gchar *fs_file)
1768 {
1769         gchar *utf8_file = NULL;
1770         GError *error = NULL;
1771
1772         utf8_file = g_filename_to_utf8(fs_file, -1, NULL, NULL, &error);
1773         if (error) {
1774                 g_warning("failed to convert encoding of file name: %s\n",
1775                           error->message);
1776                 g_error_free(error);
1777         }
1778
1779         if (!utf8_file || !g_utf8_validate(utf8_file, -1, NULL)) {
1780                 g_free(utf8_file);
1781                 utf8_file = g_strdup(fs_file);
1782                 conv_unreadable_8bit(utf8_file);
1783         }
1784
1785         return utf8_file;
1786 }