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 CharSet conv_guess_ja_encoding(const gchar *str)
 533 {
 534         const guchar *p = str;
 535         CharSet guessed = C_US_ASCII;
 536
 537         while (*p != '\0') {
 538                 if (*p == ESC && (*(p + 1) == '$' || *(p + 1) == '(')) {
 539                         if (guessed == C_US_ASCII)
 540                                 return C_ISO_2022_JP;
 541                         p += 2;
 542                 } else if (IS_ASCII(*p)) {
 543                         p++;
 544                 } else if (iseuckanji(*p) && iseuckanji(*(p + 1))) {
 545                         if (*p >= 0xfd && *p <= 0xfe)
 546                                 return C_EUC_JP;
 547                         else if (guessed == C_SHIFT_JIS) {
 548                                 if ((issjiskanji1(*p) &&
 549                                      issjiskanji2(*(p + 1))) ||
 550                                     issjishwkana(*p))
 551                                         guessed = C_SHIFT_JIS;
 552                                 else
 553                                         guessed = C_EUC_JP;
 554                         } else
 555                                 guessed = C_EUC_JP;
 556                         p += 2;
 557                 } else if (issjiskanji1(*p) && issjiskanji2(*(p + 1))) {
 558                         if (iseuchwkana1(*p) && iseuchwkana2(*(p + 1)))
 559                                 guessed = C_SHIFT_JIS;
 560                         else
 561                                 return C_SHIFT_JIS;
 562                         p += 2;
 563                 } else if (issjishwkana(*p)) {
 564                         guessed = C_SHIFT_JIS;
 565                         p++;
 566                 } else {
 567                         p++;
 568                 }
 569         }
 570
 571         return guessed;
 572 }
 573
 574 static void conv_jistodisp(gchar *outbuf, gint outlen, const gchar *inbuf)
 575 {
 576         conv_jistoutf8(outbuf, outlen, inbuf);
 577 }
 578
 579 static void conv_sjistodisp(gchar *outbuf, gint outlen, const gchar *inbuf)
 580 {
 581         conv_sjistoutf8(outbuf, outlen, inbuf);
 582 }
 583
 584 static void conv_euctodisp(gchar *outbuf, gint outlen, const gchar *inbuf)
 585 {
 586         conv_euctoutf8(outbuf, outlen, inbuf);
 587 }
 588
 589 void conv_utf8todisp(gchar *outbuf, gint outlen, const gchar *inbuf)
 590 {
 591         if (g_utf8_validate(inbuf, -1, NULL) == TRUE)
 592                 strncpy2(outbuf, inbuf, outlen);
 593         else
 594                 conv_ustodisp(outbuf, outlen, inbuf);
 595 }
 596
 597 static void conv_anytodisp(gchar *outbuf, gint outlen, const gchar *inbuf)
 598 {
 599         conv_anytoutf8(outbuf, outlen, inbuf);
 600         if (g_utf8_validate(outbuf, -1, NULL) != TRUE)
 601                 conv_unreadable_8bit(outbuf);
 602 }
 603
 604 static void conv_ustodisp(gchar *outbuf, gint outlen, const gchar *inbuf)
 605 {
 606         strncpy2(outbuf, inbuf, outlen);
 607         conv_unreadable_8bit(outbuf);
 608 }
 609
 610 void conv_localetodisp(gchar *outbuf, gint outlen, const gchar *inbuf)
 611 {
 612         gchar *tmpstr;
 613
 614         codeconv_set_strict(TRUE);
 615         tmpstr = conv_iconv_strdup(inbuf, conv_get_locale_charset_str(),
 616                                    CS_INTERNAL);
 617         codeconv_set_strict(FALSE);
 618         if (tmpstr && g_utf8_validate(tmpstr, -1, NULL)) {
 619                 strncpy2(outbuf, tmpstr, outlen);
 620                 g_free(tmpstr);
 621                 return;
 622         } else if (tmpstr && !g_utf8_validate(tmpstr, -1, NULL)) {
 623                 g_free(tmpstr);
 624                 codeconv_set_strict(TRUE);
 625                 tmpstr = conv_iconv_strdup(inbuf,
 626                                 conv_get_locale_charset_str_no_utf8(),
 627                                 CS_INTERNAL);
 628                 codeconv_set_strict(FALSE);
 629         }
 630         if (tmpstr && g_utf8_validate(tmpstr, -1, NULL)) {
 631                 strncpy2(outbuf, tmpstr, outlen);
 632                 g_free(tmpstr);
 633                 return;
 634         } else {
 635                 if (tmpstr)
 636                         g_free(tmpstr);
 637                 conv_utf8todisp(outbuf, outlen, inbuf);
 638         }
 639 }
 640
 641 static void conv_noconv(gchar *outbuf, gint outlen, const gchar *inbuf)
 642 {
 643         strncpy2(outbuf, inbuf, outlen);
 644 }
 645
 646 static const gchar *
 647 conv_get_fallback_for_private_encoding(const gchar *encoding)
 648 {
 649         if (encoding && (encoding[0] == 'X' || encoding[0] == 'x') &&
 650             encoding[1] == '-') {
 651                 if (!g_ascii_strcasecmp(encoding, CS_X_GBK))
 652                         return CS_GBK;
 653         }
 654
 655         return encoding;
 656 }
 657
 658 CodeConverter *conv_code_converter_new(const gchar *src_charset)
 659 {
 660         CodeConverter *conv;
 661
 662         src_charset = conv_get_fallback_for_private_encoding(src_charset);
 663
 664         conv = g_new0(CodeConverter, 1);
 665         conv->code_conv_func = conv_get_code_conv_func(src_charset, NULL);
 666         conv->charset_str = g_strdup(src_charset);
 667         conv->charset = conv_get_charset_from_str(src_charset);
 668
 669         return conv;
 670 }
 671
 672 void conv_code_converter_destroy(CodeConverter *conv)
 673 {
 674         g_free(conv->charset_str);
 675         g_free(conv);
 676 }
 677
 678 gint conv_convert(CodeConverter *conv, gchar *outbuf, gint outlen,
 679                   const gchar *inbuf)
 680 {
 681         if (conv->code_conv_func != conv_noconv)
 682                 conv->code_conv_func(outbuf, outlen, inbuf);
 683         else {
 684                 gchar *str;
 685
 686                 str = conv_iconv_strdup(inbuf, conv->charset_str, NULL);
 687                 if (!str)
 688                         return -1;
 689                 else {
 690                         strncpy2(outbuf, str, outlen);
 691                         g_free(str);
 692                 }
 693         }
 694
 695         return 0;
 696 }
 697
 698 gchar *conv_codeset_strdup(const gchar *inbuf,
 699                            const gchar *src_code, const gchar *dest_code)
 700 {
 701         gchar *buf;
 702         size_t len;
 703         CodeConvFunc conv_func;
 704
 705         src_code = conv_get_fallback_for_private_encoding(src_code);
 706         conv_func = conv_get_code_conv_func(src_code, dest_code);
 707         if (conv_func != conv_noconv) {
 708                 len = (strlen(inbuf) + 1) * 3;
 709                 buf = g_malloc(len);
 710                 if (!buf) return NULL;
 711
 712                 conv_func(buf, len, inbuf);
 713                 return g_realloc(buf, strlen(buf) + 1);
 714         }
 715
 716         return conv_iconv_strdup(inbuf, src_code, dest_code);
 717 }
 718
 719 CodeConvFunc conv_get_code_conv_func(const gchar *src_charset_str,
 720                                      const gchar *dest_charset_str)
 721 {
 722         CodeConvFunc code_conv = conv_noconv;
 723         CharSet src_charset;
 724         CharSet dest_charset;
 725
 726         if (!src_charset_str)
 727                 src_charset = conv_get_locale_charset();
 728         else
 729                 src_charset = conv_get_charset_from_str(src_charset_str);
 730
 731         /* auto detection mode */
 732         if (!src_charset_str && !dest_charset_str) {
 733                 if (conv_is_ja_locale())
 734                         return conv_anytodisp;
 735                 else
 736                         return conv_noconv;
 737         }
 738
 739         dest_charset = conv_get_charset_from_str(dest_charset_str);
 740
 741         if (dest_charset == C_US_ASCII)
 742                 return conv_ustodisp;
 743
 744         switch (src_charset) {
 745         case C_US_ASCII:
 746         case C_ISO_8859_1:
 747         case C_ISO_8859_2:
 748         case C_ISO_8859_3:
 749         case C_ISO_8859_4:
 750         case C_ISO_8859_5:
 751         case C_ISO_8859_6:
 752         case C_ISO_8859_7:
 753         case C_ISO_8859_8:
 754         case C_ISO_8859_9:
 755         case C_ISO_8859_10:
 756         case C_ISO_8859_11:
 757         case C_ISO_8859_13:
 758         case C_ISO_8859_14:
 759         case C_ISO_8859_15:
 760                 break;
 761         case C_ISO_2022_JP:
 762         case C_ISO_2022_JP_2:
 763         case C_ISO_2022_JP_3:
 764                 if (dest_charset == C_AUTO)
 765                         code_conv = conv_jistodisp;
 766                 else if (dest_charset == C_EUC_JP)
 767                         code_conv = conv_jistoeuc;
 768                 else if (dest_charset == C_UTF_8)
 769                         code_conv = conv_jistoutf8;
 770                 break;
 771         case C_SHIFT_JIS:
 772                 if (dest_charset == C_AUTO)
 773                         code_conv = conv_sjistodisp;
 774                 else if (dest_charset == C_EUC_JP)
 775                         code_conv = conv_sjistoeuc;
 776                 else if (dest_charset == C_UTF_8)
 777                         code_conv = conv_sjistoutf8;
 778                 break;
 779         case C_EUC_JP:
 780                 if (dest_charset == C_AUTO)
 781                         code_conv = conv_euctodisp;
 782                 else if (dest_charset == C_ISO_2022_JP   ||
 783                          dest_charset == C_ISO_2022_JP_2 ||
 784                          dest_charset == C_ISO_2022_JP_3)
 785                         code_conv = conv_euctojis;
 786                 else if (dest_charset == C_UTF_8)
 787                         code_conv = conv_euctoutf8;
 788                 break;
 789         case C_UTF_8:
 790                 if (dest_charset == C_EUC_JP)
 791                         code_conv = conv_utf8toeuc;
 792                 else if (dest_charset == C_ISO_2022_JP   ||
 793                          dest_charset == C_ISO_2022_JP_2 ||
 794                          dest_charset == C_ISO_2022_JP_3)
 795                         code_conv = conv_utf8tojis;
 796                 break;
 797         default:
 798                 break;
 799         }
 800
 801         return code_conv;
 802 }
 803
 804 gchar *conv_iconv_strdup(const gchar *inbuf,
 805                          const gchar *src_code, const gchar *dest_code)
 806 {
 807         iconv_t cd;
 808         gchar *outbuf;
 809
 810         if (!src_code)
 811                 src_code = conv_get_outgoing_charset_str();
 812         if (!dest_code)
 813                 dest_code = CS_INTERNAL;
 814
 815         /* don't convert if src and dest codeset are identical */
 816         if (!strcasecmp(src_code, dest_code))
 817                 return g_strdup(inbuf);
 818
 819         /* don't convert if current codeset is US-ASCII */
 820         if (!strcasecmp(dest_code, CS_US_ASCII))
 821                 return g_strdup(inbuf);
 822
 823         cd = iconv_open(dest_code, src_code);
 824         if (cd == (iconv_t)-1)
 825                 return NULL;
 826
 827         outbuf = conv_iconv_strdup_with_cd(inbuf, cd);
 828
 829         iconv_close(cd);
 830
 831         return outbuf;
 832 }
 833
 834 gchar *conv_iconv_strdup_with_cd(const gchar *inbuf, iconv_t cd)
 835 {
 836         const gchar *inbuf_p;
 837         gchar *outbuf;
 838         gchar *outbuf_p;
 839         size_t in_size;
 840         size_t in_left;
 841         size_t out_size;
 842         size_t out_left;
 843         size_t n_conv;
 844         size_t len;
 845
 846         inbuf_p = inbuf;
 847         in_size = strlen(inbuf);
 848         in_left = in_size;
 849         out_size = (in_size + 1) * 2;
 850         outbuf = g_malloc(out_size);
 851         outbuf_p = outbuf;
 852         out_left = out_size;
 853
 854 #define EXPAND_BUF()                            \
 855 {                                               \
 856         len = outbuf_p - outbuf;                \
 857         out_size *= 2;                          \
 858         outbuf = g_realloc(outbuf, out_size);   \
 859         outbuf_p = outbuf + len;                \
 860         out_left = out_size - len;              \
 861 }
 862
 863         while ((n_conv = iconv(cd, (ICONV_CONST gchar **)&inbuf_p, &in_left,
 864                                &outbuf_p, &out_left)) == (size_t)-1) {
 865                 if (EILSEQ == errno) {
 866                         if (strict_mode) {
 867                                 g_free(outbuf);
 868                                 return NULL;
 869                         }
 870                         //g_print("iconv(): at %d: %s\n", in_size - in_left, g_strerror(errno));
 871                         inbuf_p++;
 872                         in_left--;
 873                         if (out_left == 0) {
 874                                 EXPAND_BUF();
 875                         }
 876                         *outbuf_p++ = SUBST_CHAR;
 877                         out_left--;
 878                 } else if (EINVAL == errno) {
 879                         break;
 880                 } else if (E2BIG == errno) {
 881                         EXPAND_BUF();
 882                 } else {
 883                         g_warning("conv_iconv_strdup(): %s\n",
 884                                   g_strerror(errno));
 885                         break;
 886                 }
 887         }
 888
 889         while ((n_conv = iconv(cd, NULL, NULL, &outbuf_p, &out_left)) ==
 890                (size_t)-1) {
 891                 if (E2BIG == errno) {
 892                         EXPAND_BUF();
 893                 } else {
 894                         g_warning("conv_iconv_strdup(): %s\n",
 895                                   g_strerror(errno));
 896                         break;
 897                 }
 898         }
 899
 900 #undef EXPAND_BUF
 901
 902         len = outbuf_p - outbuf;
 903         outbuf = g_realloc(outbuf, len + 1);
 904         outbuf[len] = '\0';
 905
 906         return outbuf;
 907 }
 908
 909 static const struct {
 910         CharSet charset;
 911         gchar *const name;
 912 } charsets[] = {
 913         {C_US_ASCII,            CS_US_ASCII},
 914         {C_US_ASCII,            CS_ANSI_X3_4_1968},
 915         {C_UTF_8,               CS_UTF_8},
 916         {C_UTF_7,               CS_UTF_7},
 917         {C_ISO_8859_1,          CS_ISO_8859_1},
 918         {C_ISO_8859_2,          CS_ISO_8859_2},
 919         {C_ISO_8859_3,          CS_ISO_8859_3},
 920         {C_ISO_8859_4,          CS_ISO_8859_4},
 921         {C_ISO_8859_5,          CS_ISO_8859_5},
 922         {C_ISO_8859_6,          CS_ISO_8859_6},
 923         {C_ISO_8859_7,          CS_ISO_8859_7},
 924         {C_ISO_8859_8,          CS_ISO_8859_8},
 925         {C_ISO_8859_9,          CS_ISO_8859_9},
 926         {C_ISO_8859_10,         CS_ISO_8859_10},
 927         {C_ISO_8859_11,         CS_ISO_8859_11},
 928         {C_ISO_8859_13,         CS_ISO_8859_13},
 929         {C_ISO_8859_14,         CS_ISO_8859_14},
 930         {C_ISO_8859_15,         CS_ISO_8859_15},
 931         {C_BALTIC,              CS_BALTIC},
 932         {C_CP1250,              CS_CP1250},
 933         {C_CP1251,              CS_CP1251},
 934         {C_CP1252,              CS_CP1252},
 935         {C_CP1253,              CS_CP1253},
 936         {C_CP1254,              CS_CP1254},
 937         {C_CP1255,              CS_CP1255},
 938         {C_CP1256,              CS_CP1256},
 939         {C_CP1257,              CS_CP1257},
 940         {C_CP1258,              CS_CP1258},
 941         {C_WINDOWS_1250,        CS_WINDOWS_1250},
 942         {C_WINDOWS_1251,        CS_WINDOWS_1251},
 943         {C_WINDOWS_1252,        CS_WINDOWS_1252},
 944         {C_WINDOWS_1253,        CS_WINDOWS_1253},
 945         {C_WINDOWS_1254,        CS_WINDOWS_1254},
 946         {C_WINDOWS_1255,        CS_WINDOWS_1255},
 947         {C_WINDOWS_1256,        CS_WINDOWS_1256},
 948         {C_WINDOWS_1257,        CS_WINDOWS_1257},
 949         {C_WINDOWS_1258,        CS_WINDOWS_1258},
 950         {C_KOI8_R,              CS_KOI8_R},
 951         {C_KOI8_T,              CS_KOI8_T},
 952         {C_KOI8_U,              CS_KOI8_U},
 953         {C_ISO_2022_JP,         CS_ISO_2022_JP},
 954         {C_ISO_2022_JP_2,       CS_ISO_2022_JP_2},
 955         {C_ISO_2022_JP_3,       CS_ISO_2022_JP_3},
 956         {C_EUC_JP,              CS_EUC_JP},
 957         {C_EUC_JP,              CS_EUCJP},
 958         {C_EUC_JP_MS,           CS_EUC_JP_MS},
 959         {C_SHIFT_JIS,           CS_SHIFT_JIS},
 960         {C_SHIFT_JIS,           CS_SHIFT__JIS},
 961         {C_SHIFT_JIS,           CS_SJIS},
 962         {C_ISO_2022_KR,         CS_ISO_2022_KR},
 963         {C_EUC_KR,              CS_EUC_KR},
 964         {C_ISO_2022_CN,         CS_ISO_2022_CN},
 965         {C_EUC_CN,              CS_EUC_CN},
 966         {C_GB2312,              CS_GB2312},
 967         {C_GBK,                 CS_GBK},
 968         {C_EUC_TW,              CS_EUC_TW},
 969         {C_BIG5,                CS_BIG5},
 970         {C_BIG5_HKSCS,          CS_BIG5_HKSCS},
 971         {C_TIS_620,             CS_TIS_620},
 972         {C_WINDOWS_874,         CS_WINDOWS_874},
 973         {C_GEORGIAN_PS,         CS_GEORGIAN_PS},
 974         {C_TCVN5712_1,          CS_TCVN5712_1},
 975 };
 976
 977 static const struct {
 978         gchar *const locale;
 979         CharSet charset;
 980         CharSet out_charset;
 981 } locale_table[] = {
 982         {"ja_JP.eucJP"  , C_EUC_JP      , C_ISO_2022_JP},
 983         {"ja_JP.EUC-JP" , C_EUC_JP      , C_ISO_2022_JP},
 984         {"ja_JP.EUC"    , C_EUC_JP      , C_ISO_2022_JP},
 985         {"ja_JP.ujis"   , C_EUC_JP      , C_ISO_2022_JP},
 986         {"ja_JP.SJIS"   , C_SHIFT_JIS   , C_ISO_2022_JP},
 987         {"ja_JP.JIS"    , C_ISO_2022_JP , C_ISO_2022_JP},
 988 #ifdef G_OS_WIN32
 989         {"ja_JP"        , C_SHIFT_JIS   , C_ISO_2022_JP},
 990 #else
 991         {"ja_JP"        , C_EUC_JP      , C_ISO_2022_JP},
 992 #endif
 993         {"ko_KR.EUC-KR" , C_EUC_KR      , C_EUC_KR},
 994         {"ko_KR"        , C_EUC_KR      , C_EUC_KR},
 995         {"zh_CN.GB2312" , C_GB2312      , C_GB2312},
 996         {"zh_CN.GBK"    , C_GBK         , C_GBK},
 997         {"zh_CN"        , C_GB2312      , C_GB2312},
 998         {"zh_HK"        , C_BIG5_HKSCS  , C_BIG5_HKSCS},
 999         {"zh_TW.eucTW"  , C_EUC_TW      , C_BIG5},
1000         {"zh_TW.EUC-TW" , C_EUC_TW      , C_BIG5},
1001         {"zh_TW.Big5"   , C_BIG5        , C_BIG5},
1002         {"zh_TW"        , C_BIG5        , C_BIG5},
1003
1004         {"ru_RU.KOI8-R" , C_KOI8_R      , C_KOI8_R},
1005         {"ru_RU.KOI8R"  , C_KOI8_R      , C_KOI8_R},
1006         {"ru_RU.CP1251" , C_WINDOWS_1251, C_KOI8_R},
1007         {"ru_RU"        , C_ISO_8859_5  , C_KOI8_R},
1008         {"tg_TJ"        , C_KOI8_T      , C_KOI8_T},
1009         {"ru_UA"        , C_KOI8_U      , C_KOI8_U},
1010         {"uk_UA.CP1251" , C_WINDOWS_1251, C_KOI8_U},
1011         {"uk_UA"        , C_KOI8_U      , C_KOI8_U},
1012
1013         {"be_BY"        , C_WINDOWS_1251, C_WINDOWS_1251},
1014         {"bg_BG"        , C_WINDOWS_1251, C_WINDOWS_1251},
1015
1016         {"yi_US"        , C_WINDOWS_1255, C_WINDOWS_1255},
1017
1018         {"af_ZA"        , C_ISO_8859_1  , C_ISO_8859_1},
1019         {"br_FR"        , C_ISO_8859_1  , C_ISO_8859_1},
1020         {"ca_ES"        , C_ISO_8859_1  , C_ISO_8859_1},
1021         {"da_DK"        , C_ISO_8859_1  , C_ISO_8859_1},
1022         {"de_AT"        , C_ISO_8859_1  , C_ISO_8859_1},
1023         {"de_BE"        , C_ISO_8859_1  , C_ISO_8859_1},
1024         {"de_CH"        , C_ISO_8859_1  , C_ISO_8859_1},
1025         {"de_DE"        , C_ISO_8859_1  , C_ISO_8859_1},
1026         {"de_LU"        , C_ISO_8859_1  , C_ISO_8859_1},
1027         {"en_AU"        , C_ISO_8859_1  , C_ISO_8859_1},
1028         {"en_BW"        , C_ISO_8859_1  , C_ISO_8859_1},
1029         {"en_CA"        , C_ISO_8859_1  , C_ISO_8859_1},
1030         {"en_DK"        , C_ISO_8859_1  , C_ISO_8859_1},
1031         {"en_GB"        , C_ISO_8859_1  , C_ISO_8859_1},
1032         {"en_HK"        , C_ISO_8859_1  , C_ISO_8859_1},
1033         {"en_IE"        , C_ISO_8859_1  , C_ISO_8859_1},
1034         {"en_NZ"        , C_ISO_8859_1  , C_ISO_8859_1},
1035         {"en_PH"        , C_ISO_8859_1  , C_ISO_8859_1},
1036         {"en_SG"        , C_ISO_8859_1  , C_ISO_8859_1},
1037         {"en_US"        , C_ISO_8859_1  , C_ISO_8859_1},
1038         {"en_ZA"        , C_ISO_8859_1  , C_ISO_8859_1},
1039         {"en_ZW"        , C_ISO_8859_1  , C_ISO_8859_1},
1040         {"es_AR"        , C_ISO_8859_1  , C_ISO_8859_1},
1041         {"es_BO"        , C_ISO_8859_1  , C_ISO_8859_1},
1042         {"es_CL"        , C_ISO_8859_1  , C_ISO_8859_1},
1043         {"es_CO"        , C_ISO_8859_1  , C_ISO_8859_1},
1044         {"es_CR"        , C_ISO_8859_1  , C_ISO_8859_1},
1045         {"es_DO"        , C_ISO_8859_1  , C_ISO_8859_1},
1046         {"es_EC"        , C_ISO_8859_1  , C_ISO_8859_1},
1047         {"es_ES"        , C_ISO_8859_1  , C_ISO_8859_1},
1048         {"es_GT"        , C_ISO_8859_1  , C_ISO_8859_1},
1049         {"es_HN"        , C_ISO_8859_1  , C_ISO_8859_1},
1050         {"es_MX"        , C_ISO_8859_1  , C_ISO_8859_1},
1051         {"es_NI"        , C_ISO_8859_1  , C_ISO_8859_1},
1052         {"es_PA"        , C_ISO_8859_1  , C_ISO_8859_1},
1053         {"es_PE"        , C_ISO_8859_1  , C_ISO_8859_1},
1054         {"es_PR"        , C_ISO_8859_1  , C_ISO_8859_1},
1055         {"es_PY"        , C_ISO_8859_1  , C_ISO_8859_1},
1056         {"es_SV"        , C_ISO_8859_1  , C_ISO_8859_1},
1057         {"es_US"        , C_ISO_8859_1  , C_ISO_8859_1},
1058         {"es_UY"        , C_ISO_8859_1  , C_ISO_8859_1},
1059         {"es_VE"        , C_ISO_8859_1  , C_ISO_8859_1},
1060         {"et_EE"        , C_ISO_8859_1  , C_ISO_8859_1},
1061         {"eu_ES"        , C_ISO_8859_1  , C_ISO_8859_1},
1062         {"fi_FI"        , C_ISO_8859_1  , C_ISO_8859_1},
1063         {"fo_FO"        , C_ISO_8859_1  , C_ISO_8859_1},
1064         {"fr_BE"        , C_ISO_8859_1  , C_ISO_8859_1},
1065         {"fr_CA"        , C_ISO_8859_1  , C_ISO_8859_1},
1066         {"fr_CH"        , C_ISO_8859_1  , C_ISO_8859_1},
1067         {"fr_FR"        , C_ISO_8859_1  , C_ISO_8859_1},
1068         {"fr_LU"        , C_ISO_8859_1  , C_ISO_8859_1},
1069         {"ga_IE"        , C_ISO_8859_1  , C_ISO_8859_1},
1070         {"gl_ES"        , C_ISO_8859_1  , C_ISO_8859_1},
1071         {"gv_GB"        , C_ISO_8859_1  , C_ISO_8859_1},
1072         {"id_ID"        , C_ISO_8859_1  , C_ISO_8859_1},
1073         {"is_IS"        , C_ISO_8859_1  , C_ISO_8859_1},
1074         {"it_CH"        , C_ISO_8859_1  , C_ISO_8859_1},
1075         {"it_IT"        , C_ISO_8859_1  , C_ISO_8859_1},
1076         {"kl_GL"        , C_ISO_8859_1  , C_ISO_8859_1},
1077         {"kw_GB"        , C_ISO_8859_1  , C_ISO_8859_1},
1078         {"ms_MY"        , C_ISO_8859_1  , C_ISO_8859_1},
1079         {"nl_BE"        , C_ISO_8859_1  , C_ISO_8859_1},
1080         {"nl_NL"        , C_ISO_8859_1  , C_ISO_8859_1},
1081         {"nn_NO"        , C_ISO_8859_1  , C_ISO_8859_1},
1082         {"no_NO"        , C_ISO_8859_1  , C_ISO_8859_1},
1083         {"oc_FR"        , C_ISO_8859_1  , C_ISO_8859_1},
1084         {"pt_BR"        , C_ISO_8859_1  , C_ISO_8859_1},
1085         {"pt_PT"        , C_ISO_8859_1  , C_ISO_8859_1},
1086         {"sq_AL"        , C_ISO_8859_1  , C_ISO_8859_1},
1087         {"sv_FI"        , C_ISO_8859_1  , C_ISO_8859_1},
1088         {"sv_SE"        , C_ISO_8859_1  , C_ISO_8859_1},
1089         {"tl_PH"        , C_ISO_8859_1  , C_ISO_8859_1},
1090         {"uz_UZ"        , C_ISO_8859_1  , C_ISO_8859_1},
1091         {"wa_BE"        , C_ISO_8859_1  , C_ISO_8859_1},
1092
1093         {"bs_BA"        , C_ISO_8859_2  , C_ISO_8859_2},
1094         {"cs_CZ"        , C_ISO_8859_2  , C_ISO_8859_2},
1095         {"hr_HR"        , C_ISO_8859_2  , C_ISO_8859_2},
1096         {"hu_HU"        , C_ISO_8859_2  , C_ISO_8859_2},
1097         {"pl_PL"        , C_ISO_8859_2  , C_ISO_8859_2},
1098         {"ro_RO"        , C_ISO_8859_2  , C_ISO_8859_2},
1099         {"sk_SK"        , C_ISO_8859_2  , C_ISO_8859_2},
1100         {"sl_SI"        , C_ISO_8859_2  , C_ISO_8859_2},
1101
1102         {"sr_YU@cyrillic"       , C_ISO_8859_5  , C_ISO_8859_5},
1103         {"sr_YU"                , C_ISO_8859_2  , C_ISO_8859_2},
1104
1105         {"mt_MT"                , C_ISO_8859_3  , C_ISO_8859_3},
1106
1107         {"lt_LT.iso88594"       , C_ISO_8859_4  , C_ISO_8859_4},
1108         {"lt_LT.ISO8859-4"      , C_ISO_8859_4  , C_ISO_8859_4},
1109         {"lt_LT.ISO_8859-4"     , C_ISO_8859_4  , C_ISO_8859_4},
1110         {"lt_LT"                , C_ISO_8859_13 , C_ISO_8859_13},
1111
1112         {"mk_MK"        , C_ISO_8859_5  , C_ISO_8859_5},
1113
1114         {"ar_AE"        , C_ISO_8859_6  , C_ISO_8859_6},
1115         {"ar_BH"        , C_ISO_8859_6  , C_ISO_8859_6},
1116         {"ar_DZ"        , C_ISO_8859_6  , C_ISO_8859_6},
1117         {"ar_EG"        , C_ISO_8859_6  , C_ISO_8859_6},
1118         {"ar_IQ"        , C_ISO_8859_6  , C_ISO_8859_6},
1119         {"ar_JO"        , C_ISO_8859_6  , C_ISO_8859_6},
1120         {"ar_KW"        , C_ISO_8859_6  , C_ISO_8859_6},
1121         {"ar_LB"        , C_ISO_8859_6  , C_ISO_8859_6},
1122         {"ar_LY"        , C_ISO_8859_6  , C_ISO_8859_6},
1123         {"ar_MA"        , C_ISO_8859_6  , C_ISO_8859_6},
1124         {"ar_OM"        , C_ISO_8859_6  , C_ISO_8859_6},
1125         {"ar_QA"        , C_ISO_8859_6  , C_ISO_8859_6},
1126         {"ar_SA"        , C_ISO_8859_6  , C_ISO_8859_6},
1127         {"ar_SD"        , C_ISO_8859_6  , C_ISO_8859_6},
1128         {"ar_SY"        , C_ISO_8859_6  , C_ISO_8859_6},
1129         {"ar_TN"        , C_ISO_8859_6  , C_ISO_8859_6},
1130         {"ar_YE"        , C_ISO_8859_6  , C_ISO_8859_6},
1131
1132         {"el_GR"        , C_ISO_8859_7  , C_ISO_8859_7},
1133         {"he_IL"        , C_ISO_8859_8  , C_ISO_8859_8},
1134         {"iw_IL"        , C_ISO_8859_8  , C_ISO_8859_8},
1135         {"tr_TR"        , C_ISO_8859_9  , C_ISO_8859_9},
1136
1137         {"lv_LV"        , C_ISO_8859_13 , C_ISO_8859_13},
1138         {"mi_NZ"        , C_ISO_8859_13 , C_ISO_8859_13},
1139
1140         {"cy_GB"        , C_ISO_8859_14 , C_ISO_8859_14},
1141
1142         {"ar_IN"        , C_UTF_8       , C_UTF_8},
1143         {"en_IN"        , C_UTF_8       , C_UTF_8},
1144         {"se_NO"        , C_UTF_8       , C_UTF_8},
1145         {"ta_IN"        , C_UTF_8       , C_UTF_8},
1146         {"te_IN"        , C_UTF_8       , C_UTF_8},
1147         {"ur_PK"        , C_UTF_8       , C_UTF_8},
1148
1149         {"th_TH"        , C_TIS_620     , C_TIS_620},
1150         /* {"th_TH"     , C_WINDOWS_874}, */
1151         /* {"th_TH"     , C_ISO_8859_11}, */
1152
1153         {"ka_GE"        , C_GEORGIAN_PS , C_GEORGIAN_PS},
1154         {"vi_VN.TCVN"   , C_TCVN5712_1  , C_TCVN5712_1},
1155
1156         {"C"                    , C_US_ASCII    , C_US_ASCII},
1157         {"POSIX"                , C_US_ASCII    , C_US_ASCII},
1158         {"ANSI_X3.4-1968"       , C_US_ASCII    , C_US_ASCII},
1159 };
1160
1161 static GHashTable *conv_get_charset_to_str_table(void)
1162 {
1163         static GHashTable *table;
1164         gint i;
1165
1166         if (table)
1167                 return table;
1168
1169         table = g_hash_table_new(NULL, g_direct_equal);
1170
1171         for (i = 0; i < sizeof(charsets) / sizeof(charsets[0]); i++) {
1172                 if (g_hash_table_lookup(table, GUINT_TO_POINTER(charsets[i].charset))
1173                     == NULL) {
1174                         g_hash_table_insert
1175                                 (table, GUINT_TO_POINTER(charsets[i].charset),
1176                                  charsets[i].name);
1177                 }
1178         }
1179
1180         return table;
1181 }
1182
1183 static GHashTable *conv_get_charset_from_str_table(void)
1184 {
1185         static GHashTable *table;
1186         gint i;
1187
1188         if (table)
1189                 return table;
1190
1191         table = g_hash_table_new(str_case_hash, str_case_equal);
1192
1193         for (i = 0; i < sizeof(charsets) / sizeof(charsets[0]); i++) {
1194                 g_hash_table_insert(table, charsets[i].name,
1195                                     GUINT_TO_POINTER(charsets[i].charset));
1196         }
1197
1198         return table;
1199 }
1200
1201 const gchar *conv_get_charset_str(CharSet charset)
1202 {
1203         GHashTable *table;
1204
1205         table = conv_get_charset_to_str_table();
1206         return g_hash_table_lookup(table, GUINT_TO_POINTER(charset));
1207 }
1208
1209 CharSet conv_get_charset_from_str(const gchar *charset)
1210 {
1211         GHashTable *table;
1212
1213         if (!charset) return C_AUTO;
1214
1215         table = conv_get_charset_from_str_table();
1216         return GPOINTER_TO_UINT(g_hash_table_lookup(table, charset));
1217 }
1218
1219 CharSet conv_get_locale_charset(void)
1220 {
1221         static CharSet cur_charset = -1;
1222         const gchar *cur_locale;
1223         const gchar *p;
1224         gint i;
1225
1226         if (cur_charset != -1)
1227                 return cur_charset;
1228
1229         cur_locale = conv_get_current_locale();
1230         if (!cur_locale) {
1231                 cur_charset = C_US_ASCII;
1232                 return cur_charset;
1233         }
1234
1235         if (strcasestr(cur_locale, "UTF-8")) {
1236                 cur_charset = C_UTF_8;
1237                 return cur_charset;
1238         }
1239
1240         if ((p = strcasestr(cur_locale, "@euro")) && p[5] == '\0') {
1241                 cur_charset = C_ISO_8859_15;
1242                 return cur_charset;
1243         }
1244
1245         for (i = 0; i < sizeof(locale_table) / sizeof(locale_table[0]); i++) {
1246                 const gchar *p;
1247
1248                 /* "ja_JP.EUC" matches with "ja_JP.eucJP", "ja_JP.EUC" and
1249                    "ja_JP". "ja_JP" matches with "ja_JP.xxxx" and "ja" */
1250                 if (!g_ascii_strncasecmp(cur_locale, locale_table[i].locale,
1251                                  strlen(locale_table[i].locale))) {
1252                         cur_charset = locale_table[i].charset;
1253                         return cur_charset;
1254                 } else if ((p = strchr(locale_table[i].locale, '_')) &&
1255                          !strchr(p + 1, '.')) {
1256                         if (strlen(cur_locale) == 2 &&
1257                             !g_ascii_strncasecmp(cur_locale, locale_table[i].locale, 2)) {
1258                                 cur_charset = locale_table[i].charset;
1259                                 return cur_charset;
1260                         }
1261                 }
1262         }
1263
1264         cur_charset = C_AUTO;
1265         return cur_charset;
1266 }
1267
1268 static CharSet conv_get_locale_charset_no_utf8(void)
1269 {
1270         static CharSet cur_charset = -1;
1271         const gchar *cur_locale;
1272         const gchar *p;
1273         gchar *tmp;
1274         gint i;
1275
1276         if (prefs_common.broken_are_utf8)
1277                 return conv_get_locale_charset();
1278
1279         if (cur_charset != -1)
1280                 return cur_charset;
1281
1282         cur_locale = conv_get_current_locale();
1283         if (!cur_locale) {
1284                 cur_charset = C_US_ASCII;
1285                 return cur_charset;
1286         }
1287
1288         if (strcasestr(cur_locale, "UTF-8")) {
1289                 tmp = g_strdup(cur_locale);
1290                 *(strcasestr(tmp, ".UTF-8")) = '\0';
1291                 cur_locale = tmp;
1292         }
1293
1294         if ((p = strcasestr(cur_locale, "@euro")) && p[5] == '\0') {
1295                 cur_charset = C_ISO_8859_15;
1296                 return cur_charset;
1297         }
1298
1299         for (i = 0; i < sizeof(locale_table) / sizeof(locale_table[0]); i++) {
1300                 const gchar *p;
1301
1302                 /* "ja_JP.EUC" matches with "ja_JP.eucJP", "ja_JP.EUC" and
1303                    "ja_JP". "ja_JP" matches with "ja_JP.xxxx" and "ja" */
1304                 if (!g_ascii_strncasecmp(cur_locale, locale_table[i].locale,
1305                                  strlen(locale_table[i].locale))) {
1306                         cur_charset = locale_table[i].charset;
1307                         return cur_charset;
1308                 } else if ((p = strchr(locale_table[i].locale, '_')) &&
1309                          !strchr(p + 1, '.')) {
1310                         if (strlen(cur_locale) == 2 &&
1311                             !g_ascii_strncasecmp(cur_locale, locale_table[i].locale, 2)) {
1312                                 cur_charset = locale_table[i].charset;
1313                                 return cur_charset;
1314                         }
1315                 }
1316         }
1317
1318         cur_charset = C_AUTO;
1319         return cur_charset;
1320 }
1321
1322 const gchar *conv_get_locale_charset_str(void)
1323 {
1324         static const gchar *codeset = NULL;
1325
1326         if (!codeset)
1327                 codeset = conv_get_charset_str(conv_get_locale_charset());
1328
1329         return codeset ? codeset : CS_INTERNAL;
1330 }
1331
1332 const gchar *conv_get_locale_charset_str_no_utf8(void)
1333 {
1334         static const gchar *codeset = NULL;
1335
1336         if (!codeset)
1337                 codeset = conv_get_charset_str(conv_get_locale_charset_no_utf8());
1338
1339         return codeset ? codeset : CS_INTERNAL;
1340 }
1341
1342 CharSet conv_get_internal_charset(void)
1343 {
1344         return C_INTERNAL;
1345 }
1346
1347 const gchar *conv_get_internal_charset_str(void)
1348 {
1349         return CS_INTERNAL;
1350 }
1351
1352 CharSet conv_get_outgoing_charset(void)
1353 {
1354         static CharSet out_charset = -1;
1355         const gchar *cur_locale;
1356         const gchar *p;
1357         gint i;
1358
1359         if (out_charset != -1)
1360                 return out_charset;
1361
1362         cur_locale = conv_get_current_locale();
1363         if (!cur_locale) {
1364                 out_charset = C_AUTO;
1365                 return out_charset;
1366         }
1367
1368         if (strcasestr(cur_locale, "UTF-8")) {
1369                 out_charset = C_UTF_8;
1370                 return out_charset;
1371         }
1372
1373         if ((p = strcasestr(cur_locale, "@euro")) && p[5] == '\0') {
1374                 out_charset = C_ISO_8859_15;
1375                 return out_charset;
1376         }
1377
1378         for (i = 0; i < sizeof(locale_table) / sizeof(locale_table[0]); i++) {
1379                 const gchar *p;
1380
1381                 if (!g_ascii_strncasecmp(cur_locale, locale_table[i].locale,
1382                                  strlen(locale_table[i].locale))) {
1383                         out_charset = locale_table[i].out_charset;
1384                         break;
1385                 } else if ((p = strchr(locale_table[i].locale, '_')) &&
1386                          !strchr(p + 1, '.')) {
1387                         if (strlen(cur_locale) == 2 &&
1388                             !g_ascii_strncasecmp(cur_locale, locale_table[i].locale, 2)) {
1389                                 out_charset = locale_table[i].out_charset;
1390                                 break;
1391                         }
1392                 }
1393         }
1394
1395         return out_charset;
1396 }
1397
1398 const gchar *conv_get_outgoing_charset_str(void)
1399 {
1400         CharSet out_charset;
1401         const gchar *str;
1402
1403         out_charset = conv_get_outgoing_charset();
1404         str = conv_get_charset_str(out_charset);
1405
1406         return str ? str : CS_UTF_8;
1407 }
1408
1409 gboolean conv_is_multibyte_encoding(CharSet encoding)
1410 {
1411         switch (encoding) {
1412         case C_EUC_JP:
1413         case C_EUC_JP_MS:
1414         case C_EUC_KR:
1415         case C_EUC_TW:
1416         case C_EUC_CN:
1417         case C_ISO_2022_JP:
1418         case C_ISO_2022_JP_2:
1419         case C_ISO_2022_JP_3:
1420         case C_ISO_2022_KR:
1421         case C_ISO_2022_CN:
1422         case C_SHIFT_JIS:
1423         case C_GB2312:
1424         case C_GBK:
1425         case C_BIG5:
1426         case C_UTF_8:
1427         case C_UTF_7:
1428                 return TRUE;
1429         default:
1430                 return FALSE;
1431         }
1432 }
1433
1434 const gchar *conv_get_current_locale(void)
1435 {
1436         const gchar *cur_locale;
1437
1438 #ifdef G_OS_WIN32
1439         cur_locale = g_win32_getlocale();
1440 #else
1441         cur_locale = g_getenv("LC_ALL");
1442         if (!cur_locale) cur_locale = g_getenv("LC_CTYPE");
1443         if (!cur_locale) cur_locale = g_getenv("LANG");
1444         if (!cur_locale) cur_locale = setlocale(LC_CTYPE, NULL);
1445 #endif /* G_OS_WIN32 */
1446
1447         debug_print("current locale: %s\n",
1448                     cur_locale ? cur_locale : "(none)");
1449
1450         return cur_locale;
1451 }
1452
1453 gboolean conv_is_ja_locale(void)
1454 {
1455         static gint is_ja_locale = -1;
1456         const gchar *cur_locale;
1457
1458         if (is_ja_locale != -1)
1459                 return is_ja_locale != 0;
1460
1461         is_ja_locale = 0;
1462         cur_locale = conv_get_current_locale();
1463         if (cur_locale) {
1464                 if (g_ascii_strncasecmp(cur_locale, "ja", 2) == 0)
1465                         is_ja_locale = 1;
1466         }
1467
1468         return is_ja_locale != 0;
1469 }
1470
1471 gchar *conv_unmime_header(const gchar *str, const gchar *default_encoding)
1472 {
1473         gchar buf[BUFFSIZE];
1474
1475         if (is_ascii_str(str))
1476                 return unmime_header(str);
1477
1478         if (default_encoding) {
1479                 gchar *utf8_buf;
1480
1481                 utf8_buf = conv_codeset_strdup
1482                         (str, default_encoding, CS_INTERNAL);
1483                 if (utf8_buf) {
1484                         gchar *decoded_str;
1485
1486                         decoded_str = unmime_header(utf8_buf);
1487                         g_free(utf8_buf);
1488                         return decoded_str;
1489                 }
1490         }
1491
1492         if (conv_is_ja_locale())
1493                 conv_anytodisp(buf, sizeof(buf), str);
1494         else
1495                 conv_localetodisp(buf, sizeof(buf), str);
1496
1497         return unmime_header(buf);
1498 }
1499
1500 #define MAX_LINELEN             76
1501 #define MAX_HARD_LINELEN        996
1502 #define MIMESEP_BEGIN           "=?"
1503 #define MIMESEP_END             "?="
1504
1505 #define LBREAK_IF_REQUIRED(cond, is_plain_text)                         \
1506 {                                                                       \
1507         if (len - (destp - (guchar *)dest) < MAX_LINELEN + 2) {         \
1508                 *destp = '\0';                                          \
1509                 return;                                                 \
1510         }                                                               \
1511                                                                         \
1512         if ((cond) && *srcp) {                                          \
1513                 if (destp > (guchar *)dest && left < MAX_LINELEN - 1) { \
1514                         if (isspace(*(destp - 1)))                      \
1515                                 destp--;                                \
1516                         else if (is_plain_text && isspace(*srcp))       \
1517                                 srcp++;                                 \
1518                         if (*srcp) {                                    \
1519                                 *destp++ = '\n';                        \
1520                                 *destp++ = ' ';                         \
1521                                 left = MAX_LINELEN - 1;                 \
1522                         }                                               \
1523                 }                                                       \
1524         }                                                               \
1525 }
1526
1527 void conv_encode_header_full(gchar *dest, gint len, const gchar *src,
1528                         gint header_len, gboolean addr_field,
1529                         const gchar *out_encoding_)
1530 {
1531         const gchar *cur_encoding;
1532         const gchar *out_encoding;
1533         gint mimestr_len;
1534         gchar *mimesep_enc;
1535         gint left;
1536         const guchar *srcp = src;
1537         guchar *destp = dest;
1538         gboolean use_base64;
1539
1540         g_return_if_fail(g_utf8_validate(src, -1, NULL) == TRUE);
1541
1542         if (MB_CUR_MAX > 1) {
1543                 use_base64 = TRUE;
1544                 mimesep_enc = "?B?";
1545         } else {
1546                 use_base64 = FALSE;
1547                 mimesep_enc = "?Q?";
1548         }
1549
1550         cur_encoding = CS_INTERNAL;
1551
1552         if (out_encoding_)
1553                 out_encoding = out_encoding_;
1554         else
1555                 out_encoding = conv_get_outgoing_charset_str();
1556
1557         if (!strcmp(out_encoding, CS_US_ASCII))
1558                 out_encoding = CS_ISO_8859_1;
1559
1560         mimestr_len = strlen(MIMESEP_BEGIN) + strlen(out_encoding) +
1561                 strlen(mimesep_enc) + strlen(MIMESEP_END);
1562
1563         left = MAX_LINELEN - header_len;
1564
1565         while (*srcp) {
1566                 LBREAK_IF_REQUIRED(left <= 0, TRUE);
1567
1568                 while (isspace(*srcp)) {
1569                         *destp++ = *srcp++;
1570                         left--;
1571                         LBREAK_IF_REQUIRED(left <= 0, TRUE);
1572                 }
1573
1574                 /* output as it is if the next word is ASCII string */
1575                 if (!is_next_nonascii(srcp)) {
1576                         gint word_len;
1577
1578                         word_len = get_next_word_len(srcp);
1579                         LBREAK_IF_REQUIRED(left < word_len, TRUE);
1580                         while (word_len > 0) {
1581                                 LBREAK_IF_REQUIRED(left + (MAX_HARD_LINELEN - MAX_LINELEN) <= 0, TRUE)
1582                                 *destp++ = *srcp++;
1583                                 left--;
1584                                 word_len--;
1585                         }
1586
1587                         continue;
1588                 }
1589
1590                 /* don't include parentheses in encoded strings */
1591                 if (addr_field && (*srcp == '(' || *srcp == ')')) {
1592                         LBREAK_IF_REQUIRED(left < 2, FALSE);
1593                         *destp++ = *srcp++;
1594                         left--;
1595                 }
1596
1597                 while (1) {
1598                         gint mb_len = 0;
1599                         gint cur_len = 0;
1600                         gchar *part_str;
1601                         gchar *out_str;
1602                         gchar *enc_str;
1603                         const guchar *p = srcp;
1604                         gint out_str_len;
1605                         gint out_enc_str_len;
1606                         gint mime_block_len;
1607                         gboolean cont = FALSE;
1608
1609                         while (*p != '\0') {
1610                                 if (isspace(*p) && !is_next_nonascii(p + 1))
1611                                         break;
1612                                 /* don't include parentheses in encoded
1613                                    strings */
1614                                 if (addr_field && (*p == '(' || *p == ')'))
1615                                         break;
1616
1617                                 mb_len = g_utf8_skip[*p];
1618
1619                                 Xstrndup_a(part_str, srcp, cur_len + mb_len, );
1620                                 out_str = conv_codeset_strdup
1621                                         (part_str, cur_encoding, out_encoding);
1622                                 if (!out_str) {
1623                                         if (strict_mode) {
1624                                                 *dest = '\0';
1625                                                 return;
1626                                         } else {
1627                                                 g_warning("conv_encode_header(): code conversion failed\n");
1628                                                 conv_unreadable_8bit(part_str);
1629                                                 out_str = g_strdup(part_str);
1630                                         }
1631                                 }
1632                                 out_str_len = strlen(out_str);
1633
1634                                 if (use_base64)
1635                                         out_enc_str_len = B64LEN(out_str_len);
1636                                 else
1637                                         out_enc_str_len =
1638                                                 qp_get_q_encoding_len(out_str);
1639
1640                                 g_free(out_str);
1641
1642                                 if (mimestr_len + out_enc_str_len <= left) {
1643                                         cur_len += mb_len;
1644                                         p += mb_len;
1645                                 } else if (cur_len == 0) {
1646                                         LBREAK_IF_REQUIRED(1, FALSE);
1647                                         continue;
1648                                 } else {
1649                                         cont = TRUE;
1650                                         break;
1651                                 }
1652                         }
1653
1654                         if (cur_len > 0) {
1655                                 Xstrndup_a(part_str, srcp, cur_len, );
1656                                 out_str = conv_codeset_strdup
1657                                         (part_str, cur_encoding, out_encoding);
1658                                 if (!out_str) {
1659                                         g_warning("conv_encode_header(): code conversion failed\n");
1660                                         conv_unreadable_8bit(part_str);
1661                                         out_str = g_strdup(part_str);
1662                                 }
1663                                 out_str_len = strlen(out_str);
1664
1665                                 if (use_base64)
1666                                         out_enc_str_len = B64LEN(out_str_len);
1667                                 else
1668                                         out_enc_str_len =
1669                                                 qp_get_q_encoding_len(out_str);
1670
1671                                 Xalloca(enc_str, out_enc_str_len + 1, );
1672                                 if (use_base64)
1673                                         base64_encode(enc_str, out_str, out_str_len);
1674                                 else
1675                                         qp_q_encode(enc_str, out_str);
1676
1677                                 g_free(out_str);
1678
1679                                 /* output MIME-encoded string block */
1680                                 mime_block_len = mimestr_len + strlen(enc_str);
1681                                 g_snprintf(destp, mime_block_len + 1,
1682                                            MIMESEP_BEGIN "%s%s%s" MIMESEP_END,
1683                                            out_encoding, mimesep_enc, enc_str);
1684                                 destp += mime_block_len;
1685                                 srcp += cur_len;
1686
1687                                 left -= mime_block_len;
1688                         }
1689
1690                         LBREAK_IF_REQUIRED(cont, FALSE);
1691
1692                         if (cur_len == 0)
1693                                 break;
1694                 }
1695         }
1696
1697         *destp = '\0';
1698 }
1699
1700 void conv_encode_header(gchar *dest, gint len, const gchar *src,
1701                         gint header_len, gboolean addr_field)
1702 {
1703         conv_encode_header_full(dest,len,src,header_len,addr_field,NULL);
1704 }
1705
1706 #undef LBREAK_IF_REQUIRED
1707 gchar *conv_filename_from_utf8(const gchar *utf8_file)
1708 {
1709         gchar *fs_file;
1710         GError *error = NULL;
1711
1712         fs_file = g_filename_from_utf8(utf8_file, -1, NULL, NULL, &error);
1713         if (error) {
1714                 g_warning("failed to convert encoding of file name: %s\n",
1715                           error->message);
1716                 g_error_free(error);
1717         }
1718         if (!fs_file)
1719                 fs_file = g_strdup(utf8_file);
1720
1721         return fs_file;
1722 }
1723
1724 gchar *conv_filename_to_utf8(const gchar *fs_file)
1725 {
1726         gchar *utf8_file = NULL;
1727         GError *error = NULL;
1728
1729         utf8_file = g_filename_to_utf8(fs_file, -1, NULL, NULL, &error);
1730         if (error) {
1731                 g_warning("failed to convert encoding of file name: %s\n",
1732                           error->message);
1733                 g_error_free(error);
1734         }
1735
1736         if (!utf8_file || !g_utf8_validate(utf8_file, -1, NULL)) {
1737                 g_free(utf8_file);
1738                 utf8_file = g_strdup(fs_file);
1739                 conv_unreadable_8bit(utf8_file);
1740         }
1741
1742         return utf8_file;
1743 }