mbyte.c 12.8 KB
Newer Older
1
/*
2 3
 * Copyright (C) 2000 Edmund Grimley Evans <edmundo@rano.org>
 *
4 5 6 7
 *     This program is free software; you can redistribute it and/or modify
 *     it under the terms of the GNU General Public License as published by
 *     the Free Software Foundation; either version 2 of the License, or
 *     (at your option) any later version.
8
 *
9 10 11 12
 *     This program is distributed in the hope that it will be useful,
 *     but WITHOUT ANY WARRANTY; without even the implied warranty of
 *     MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
 *     GNU General Public License for more details.
13
 *
14 15 16
 *     You should have received a copy of the GNU General Public License
 *     along with this program; if not, write to the Free Software
 *     Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA.
17 18 19 20 21 22
 */

/*
 * Japanese support by TAKIZAWA Takashi <taki@luna.email.ne.jp>.
 */

23 24 25 26 27
#if HAVE_CONFIG_H
# include "config.h"
#endif

#include "mutt.h"
28 29
#include "mbyte.h"
#include "charset.h"
30 31 32 33

#include <errno.h>

#include <ctype.h>
34 35 36 37 38 39 40 41 42 43 44 45

#ifndef EILSEQ
#define EILSEQ EINVAL
#endif

int Charset_is_utf8 = 0;
#ifndef HAVE_WC_FUNCS
static int charset_is_ja = 0;
static iconv_t charset_to_utf8 = (iconv_t)(-1);
static iconv_t charset_from_utf8 = (iconv_t)(-1);
#endif

46
void mutt_set_charset (char *charset)
47 48 49
{
  char buffer[STRING];

50
  mutt_canonical_charset (buffer, sizeof (buffer), charset);
51 52 53 54 55 56

  Charset_is_utf8 = 0;
#ifndef HAVE_WC_FUNCS
  charset_is_ja = 0;
  if (charset_to_utf8 != (iconv_t)(-1))
  {
57
    iconv_close (charset_to_utf8);
58 59 60 61
    charset_to_utf8 = (iconv_t)(-1);
  }
  if (charset_from_utf8 != (iconv_t)(-1))
  {
62
    iconv_close (charset_from_utf8);
63 64 65 66
    charset_from_utf8 = (iconv_t)(-1);
  }
#endif

67
  if (mutt_is_utf8 (buffer))
68 69
    Charset_is_utf8 = 1;
#ifndef HAVE_WC_FUNCS
70 71
  else if (!ascii_strcasecmp(buffer, "euc-jp") || !ascii_strcasecmp(buffer, "shift_jis")
  	|| !ascii_strcasecmp(buffer, "cp932") || !ascii_strcasecmp(buffer, "eucJP-ms"))
72 73 74 75 76 77 78 79 80
  {
    charset_is_ja = 1;

    /* Note flags=0 to skip charset-hooks: User masters the $charset
     * name, and we are sure of our "utf-8" constant. So there is no
     * possibility of wrong name that we would want to try to correct
     * with a charset-hook. Or rather: If $charset was wrong, we would
     * want to try to correct... $charset directly.
     */
81 82
    charset_to_utf8 = mutt_iconv_open ("utf-8", charset, 0);
    charset_from_utf8 = mutt_iconv_open (charset, "utf-8", 0);
83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103
  }
#endif

#if defined(HAVE_BIND_TEXTDOMAIN_CODESET) && defined(ENABLE_NLS)
  bind_textdomain_codeset(PACKAGE, buffer);
#endif
}

#ifndef HAVE_WC_FUNCS

/*
 * For systems that don't have them, we provide here our own
 * implementations of wcrtomb(), mbrtowc(), iswprint() and wcwidth().
 * Instead of using the locale, as these functions normally would,
 * we use Mutt's Charset variable. We support 3 types of charset:
 * (1) For 8-bit charsets, wchar_t uses the same encoding as char.
 * (2) For UTF-8, wchar_t uses UCS.
 * (3) For stateless Japanese encodings, we use UCS and convert
 *     via UTF-8 using iconv.
 * Unfortunately, we can't handle non-stateless encodings.
 */
104 105

static size_t wcrtomb_iconv (char *s, wchar_t wc, iconv_t cd)
106
{
107 108 109 110
  char buf[MB_LEN_MAX+1];
  ICONV_CONST char *ib;
  char *ob;
  size_t ibl, obl, r;
111 112 113

  if (s)
  {
114
    ibl = mutt_wctoutf8 (buf, wc, sizeof (buf));
115 116 117 118 119
    if (ibl == (size_t)(-1))
      return (size_t)(-1);
    ib = buf;
    ob = s;
    obl = MB_LEN_MAX;
120
    r = iconv (cd, &ib, &ibl, &ob, &obl);
121 122 123 124 125 126
  }
  else
  {
    ib = "";
    ibl = 1;
    ob = buf;
127 128
    obl = sizeof (buf);
    r = iconv (cd, &ib, &ibl, &ob, &obl);
129 130 131 132
  }
  return ob - s;
}

133
size_t wcrtomb (char *s, wchar_t wc, mbstate_t *ps)
134 135 136 137
{
  /* We only handle stateless encodings, so we can ignore ps. */

  if (Charset_is_utf8)
138
    return mutt_wctoutf8 (s, wc, MB_LEN_MAX);
139
  else if (charset_from_utf8 != (iconv_t)(-1))
140
    return wcrtomb_iconv (s, wc, charset_from_utf8);
141 142 143 144 145 146 147 148 149 150 151 152 153 154
  else
  {
    if (!s)
      return 1;
    if (wc < 0x100)
    {
      *s = wc;
      return 1;
    }
    errno = EILSEQ;
    return (size_t)(-1);
  }
}

155 156
size_t mbrtowc_iconv (wchar_t *pwc, const char *s, size_t n,
		      mbstate_t *ps, iconv_t cd)
157 158
{
  static mbstate_t mbstate;
159 160
  ICONV_CONST char *ib, *ibmax;
  char *ob, *t;
161 162 163 164 165 166
  size_t ibl, obl, k, r;
  char bufi[8], bufo[6];

  if (!n)
    return (size_t)(-2);

167 168 169
  t = memchr (ps, 0, sizeof (*ps));
  k = t ? (t - (char *)ps) : sizeof (*ps);
  if (k > sizeof (bufi))
170 171 172 173
    k = 0;
  if (k)
  {
    /* use the buffer for input */
174
    memcpy (bufi, ps, k);
175
    ib = bufi;
176 177
    ibmax = bufi + (k + n < sizeof (bufi) ? k + n : sizeof (bufi));
    memcpy (bufi + k, s, ibmax - bufi - k);
178 179 180 181
  }
  else
  {
    /* use the real input */
182 183
    ib = (ICONV_CONST char*) s;
    ibmax = (ICONV_CONST char*) s + n;
184 185 186
  }

  ob = bufo;
187
  obl = sizeof (bufo);
188 189 190 191
  ibl = 1;

  for (;;)
  {
192
    r = iconv (cd, &ib, &ibl, &ob, &obl);
193 194 195
    if (ob > bufo && (!k || ib > bufi + k))
    {
      /* we have a character */
196 197
      memset (ps, 0, sizeof (*ps));
      utf8rtowc (pwc, bufo, ob - bufo, &mbstate);
198 199 200 201 202
      return (pwc && *pwc) ? (ib - (k ? bufi + k : s)) : 0;
    }
    else if (!r || (r == (size_t)(-1) && errno == EINVAL))
    {
      if (ib + ibl < ibmax)
203 204
	/* try using more input */
	++ibl;
205 206
      else if (k && ib > bufi + k && bufi + k + n > ibmax)
      {
207 208 209 210 211
	/* switch to using real input */
	ib = (ICONV_CONST char*) s + (ib - bufi - k);
	ibmax = (ICONV_CONST char*) s + n;
	k = 0;
	++ibl;
212 213 214
      }
      else
      {
215 216 217 218 219
	/* save the state and give up */
	memset (ps, 0, sizeof (*ps));
	if (ibl <= sizeof (mbstate_t)) /* need extra condition here! */
	  memcpy (ps, ib, ibl);
	return (size_t)(-2);
220 221 222 223 224 225 226 227 228 229 230
      }
    }
    else
    {
      /* bad input */
      errno = EILSEQ;
      return (size_t)(-1);
    }
  }
}

231
size_t mbrtowc (wchar_t *pwc, const char *s, size_t n, mbstate_t *ps)
232 233 234 235 236 237 238
{
  static mbstate_t mbstate;

  if (!ps)
    ps = &mbstate;

  if (Charset_is_utf8)
239
    return utf8rtowc (pwc, s, n, ps);
240
  else if (charset_to_utf8 != (iconv_t)(-1))
241
    return mbrtowc_iconv (pwc, s, n, ps, charset_to_utf8);
242 243 244 245 246 247 248 249
  else
  {
    if (!s)
    {
      memset(ps, 0, sizeof(*ps));
      return 0;
    }
    if (!n)
250
      return (size_t)-2;
251
    if (pwc)
252
      *pwc = (wchar_t)(unsigned char)*s;
253 254 255 256
    return (*s != 0);
  }
}

257
int iswprint (wint_t wc)
258 259 260 261
{
  if (Charset_is_utf8 || charset_is_ja)
    return ((0x20 <= wc && wc < 0x7f) || 0xa0 <= wc);
  else
262
    return (0 <= wc && wc < 256) ? IsPrint (wc) : 0;
263 264
}

265
int iswspace (wint_t wc)
266 267 268 269
{
  if (Charset_is_utf8 || charset_is_ja)
    return (9 <= wc && wc <= 13) || wc == 32;
  else
270
    return (0 <= wc && wc < 256) ? isspace (wc) : 0;
271 272
}

273
static wint_t towupper_ucs (wint_t x)
274 275 276 277 278 279 280 281 282 283 284 285 286 287
{
  /* Only works for x < 0x130 */
  if ((0x60 < x && x < 0x7b) || (0xe0 <= x && x < 0xff && x != 0xf7))
    return x - 32;
  else if (0x100 <= x && x < 0x130)
    return x & ~1;
  else if (x == 0xb5)
    return 0x39c;
  else if (x == 0xff)
    return 0x178;
  else
    return x;
}

288
static int iswupper_ucs (wint_t x)
289 290 291
{
  /* Only works for x < 0x130 */
  if ((0x60 < x && x < 0x7b) || (0xe0 <= x && x < 0xff && x != 0xf7))
292 293
    return 0;
  else if ((0x40 < x && x < 0x5b) || (0xbf < x && x < 0xde))
294 295 296 297 298 299
    return 1;
  else if (0x100 <= x && x < 0x130)
    return 1;
  else if (x == 0xb5)
    return 1;
  else if (x == 0xff)
300
    return 0;
301 302 303 304
  else
    return 0;
}

305
static wint_t towlower_ucs (wint_t x)
306 307 308 309 310 311 312 313 314 315
{
  /* Only works for x < 0x130 */
  if ((0x40 < x && x < 0x5b) || (0xc0 <= x && x < 0xdf && x != 0xd7))
    return x + 32;
  else if (0x100 <= x && x < 0x130)
    return x | 1;
  else
    return x;
}

316
static int iswalnum_ucs (wint_t wc)
317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332
{
  /* Only works for x < 0x220 */
  if (wc >= 0x100)
    return 1;
  else if (wc < 0x30)
    return 0;
  else if (wc < 0x3a)
    return 1;
  else if (wc < 0xa0)
    return (0x40 < (wc & ~0x20) && (wc & ~0x20) < 0x5b);
  else if (wc < 0xc0)
    return (wc == 0xaa || wc == 0xb5 || wc == 0xba);
  else
    return !(wc == 0xd7 || wc == 0xf7);
}

333
static int iswalpha_ucs (wint_t wc)
334 335 336 337 338 339 340 341 342 343 344 345 346 347
{
  /* Only works for x < 0x220 */
  if (wc >= 0x100)
    return 1;
  else if (wc < 0x3a)
    return 0;
  else if (wc < 0xa0)
    return (0x40 < (wc & ~0x20) && (wc & ~0x20) < 0x5b);
  else if (wc < 0xc0)
    return (wc == 0xaa || wc == 0xb5 || wc == 0xba);
  else
    return !(wc == 0xd7 || wc == 0xf7);
}

348
wint_t towupper (wint_t wc)
349 350
{
  if (Charset_is_utf8 || charset_is_ja)
351
    return towupper_ucs (wc);
352
  else
353
    return (0 <= wc && wc < 256) ? toupper (wc) : wc;
354 355
}

356
wint_t towlower (wint_t wc)
357 358
{
  if (Charset_is_utf8 || charset_is_ja)
359
    return towlower_ucs (wc);
360
  else
361
    return (0 <= wc && wc < 256) ? tolower (wc) : wc;
362 363
}

364
int iswalnum (wint_t wc)
365 366
{
  if (Charset_is_utf8 || charset_is_ja)
367
    return iswalnum_ucs (wc);
368
  else
369
    return (0 <= wc && wc < 256) ? isalnum (wc) : 0;
370 371
}

372
int iswalpha (wint_t wc)
373 374
{
  if (Charset_is_utf8 || charset_is_ja)
375
    return iswalpha_ucs (wc);
376
  else
377
    return (0 <= wc && wc < 256) ? isalpha (wc) : 0;
378 379
}

380
int iswupper (wint_t wc)
381 382
{
  if (Charset_is_utf8 || charset_is_ja)
383
    return iswupper_ucs (wc);
384
  else
385
    return (0 <= wc && wc < 256) ? isupper (wc) : 0;
386 387
}

388 389 390 391 392
/*
 * l10n for Japanese:
 *   Symbols, Greek and Cyrillic in JIS X 0208, Japanese Kanji
 *   Character Set, have a column width of 2.
 */
393
int wcwidth_ja (wchar_t ucs)
394 395 396 397 398 399 400 401 402 403 404 405 406 407 408
{
  if (ucs >= 0x3021)
    return -1; /* continue with the normal check */
  /* a rough range for quick check */
  if ((ucs >= 0x00a1 && ucs <= 0x00fe) || /* Latin-1 Supplement */
      (ucs >= 0x0391 && ucs <= 0x0451) || /* Greek and Cyrillic */
      (ucs >= 0x2010 && ucs <= 0x266f) || /* Symbols */
      (ucs >= 0x3000 && ucs <= 0x3020))   /* CJK Symbols and Punctuation */
    return 2;
  else
    return -1;
}

int wcwidth_ucs(wchar_t ucs);

409
int wcwidth (wchar_t wc)
410 411 412 413 414 415 416
{
  if (!Charset_is_utf8)
  {
    if (!charset_is_ja)
    {
      /* 8-bit case */
      if (!wc)
417 418 419
	return 0;
      else if ((0 <= wc && wc < 256) && IsPrint (wc))
	return 1;
420
      else
421
	return -1;
422 423 424 425
    }
    else
    {
      /* Japanese */
426
      int k = wcwidth_ja (wc);
427
      if (k != -1)
428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514
	return k;
    }
  }
  return wcwidth_ucs (wc);
}

size_t utf8rtowc (wchar_t *pwc, const char *s, size_t n, mbstate_t *_ps)
{
  static wchar_t mbstate;
  wchar_t *ps = (wchar_t *)_ps;
  size_t k = 1;
  unsigned char c;
  wchar_t wc;
  int count;

  if (!ps)
    ps = &mbstate;

  if (!s)
  {
    *ps = 0;
    return 0;
  }
  if (!n)
    return (size_t)-2;

  if (!*ps)
  {
    c = (unsigned char)*s;
    if (c < 0x80)
    {
      if (pwc)
	*pwc = c;
      return (c != 0);
    }
    else if (c < 0xc2)
    {
      errno = EILSEQ;
      return (size_t)-1;
    }
    else if (c < 0xe0)
      wc = ((c & 0x1f) << 6) + (count = 0);
    else if (c < 0xf0)
      wc = ((c & 0x0f) << 12) + (count = 1);
    else if (c < 0xf8)
      wc = ((c & 0x07) << 18) + (count = 2);
    else if (c < 0xfc)
      wc = ((c & 0x03) << 24) + (count = 3);
    else if (c < 0xfe)
      wc = ((c & 0x01) << 30) + (count = 4);
    else
    {
      errno = EILSEQ;
      return (size_t)-1;
    }
    ++s, --n, ++k;
  }
  else
  {
    wc = *ps & 0x7fffffff;
    count = wc & 7; /* if count > 4 it will be caught below */
  }

  for (; n; ++s, --n, ++k)
  {
    c = (unsigned char)*s;
    if (0x80 <= c && c < 0xc0)
    {
      wc |= (c & 0x3f) << (6 * count);
      if (!count)
      {
	if (pwc)
	  *pwc = wc;
	*ps = 0;
	return wc ? k : 0;
      }
      --count, --wc;
      if (!(wc >> (11+count*5)))
      {
	errno = count < 4 ? EILSEQ : EINVAL;
	return (size_t)-1;
      }
    }
    else
    {
      errno = EILSEQ;
      return (size_t)-1;
515 516
    }
  }
517 518
  *ps = wc;
  return (size_t)-2;
519 520 521 522
}

#endif /* !HAVE_WC_FUNCS */

523
wchar_t replacement_char (void)
524 525 526 527
{
  return Charset_is_utf8 ? 0xfffd : '?';
}

528
int is_display_corrupting_utf8 (wchar_t wc)
529
{
530 531 532 533 534 535 536 537 538
  if (wc == (wchar_t)0x200f ||   /* bidi markers: #3827 */
      wc == (wchar_t)0x200e ||
      wc == (wchar_t)0x00ad ||   /* soft hyphen: #3848 */
      wc == (wchar_t)0xfeff ||   /* zero width no-break space */
      (wc >= (wchar_t)0x2066 &&  /* misc directional markers */
       wc <= (wchar_t)0x2069) ||
      (wc >= (wchar_t)0x202a &&  /* misc directional markers: #3854 */
       wc <= (wchar_t)0x202e))
    return 1;
539
  else
540
    return 0;
541 542
}

543
int mutt_filter_unprintable (char **s)
544
{
545
  BUFFER *b = NULL;
546 547 548 549 550 551
  wchar_t wc;
  size_t k, k2;
  char scratch[MB_LEN_MAX + 1];
  char *p = *s;
  mbstate_t mbstate1, mbstate2;

552
  if (!(b = mutt_buffer_new ()))
553
    return -1;
554 555 556
  memset (&mbstate1, 0, sizeof (mbstate1));
  memset (&mbstate2, 0, sizeof (mbstate2));
  for (; (k = mbrtowc (&wc, p, MB_LEN_MAX, &mbstate1)); p += k)
557 558 559 560
  {
    if (k == (size_t)(-1) || k == (size_t)(-2))
    {
      k = 1;
561
      memset (&mbstate1, 0, sizeof (mbstate1));
562 563
      wc = replacement_char();
    }
564
    if (!IsWPrint (wc))
565
      wc = '?';
566 567
    else if (Charset_is_utf8 &&
             is_display_corrupting_utf8 (wc))
568
      continue;
569
    k2 = wcrtomb (scratch, wc, &mbstate2);
570
    scratch[k2] = '\0';
571
    mutt_buffer_addstr (b, scratch);
572
  }
573 574 575
  FREE (s);  /* __FREE_CHECKED__ */
  *s = b->data ? b->data : safe_calloc (1, 1);
  FREE (&b);
576 577
  return 0;
}
578