iconv.c

libiconv是一个很不错的字符集转换库。程序接口也很简单

    }
    to_index = ap->encoding_index;
    break;
  }
  for (from_wchar = 0;;) {
    /* Search fromcode in the table. */
    for (cp = fromcode, bp = buf, count = MAX_WORD_LENGTH+10+1; ; cp++, bp++) {
      unsigned char c = * (unsigned char *) cp;
      if (c >= 0x80)
        goto invalid;
      if (c >= 'a' && c <= 'z')
        c -= 'a'-'A';
      *bp = c;
      if (c == '\0')
        break;
      if (--count == 0)
        goto invalid;
    }
    if (bp-buf >= 10 && memcmp(bp-10,"//TRANSLIT",10)==0) {
      bp -= 10;
      *bp = '\0';
    }
    if (bp-buf >= 8 && memcmp(bp-8,"//IGNORE",8)==0) {
      bp -= 8;
      *bp = '\0';
    }
    if (buf[0] == '\0') {
      fromcode = locale_charset();
      /* Avoid an endless loop that could occur when using an older version
         of localcharset.c. */
      if (fromcode[0] == '\0')
        goto invalid;
      continue;
    }
    ap = aliases_lookup(buf,bp-buf);
    if (ap == NULL) {
      ap = aliases2_lookup(buf);
      if (ap == NULL)
        goto invalid;
    }
    if (ap->encoding_index == ei_local_char) {
      fromcode = locale_charset();
      /* Avoid an endless loop that could occur when using an older version
         of localcharset.c. */
      if (fromcode[0] == '\0')
        goto invalid;
      continue;
    }
    if (ap->encoding_index == ei_local_wchar_t) {
#if __STDC_ISO_10646__
      if (sizeof(wchar_t) == 4) {
        from_index = ei_ucs4internal;
        break;
      }
      if (sizeof(wchar_t) == 2) {
        from_index = ei_ucs2internal;
        break;
      }
      if (sizeof(wchar_t) == 1) {
        from_index = ei_iso8859_1;
        break;
      }
#endif
#if HAVE_WCRTOMB
      from_wchar = 1;
      fromcode = locale_charset();
      continue;
#endif
      goto invalid;
    }
    from_index = ap->encoding_index;
    break;
  }
  cd = (struct conv_struct *) malloc(from_wchar != to_wchar
                                     ? sizeof(struct wchar_conv_struct)
                                     : sizeof(struct conv_struct));
  if (cd == NULL) {
    errno = ENOMEM;
    return (iconv_t)(-1);
  }
  cd->iindex = from_index;
  cd->ifuncs = all_encodings[from_index].ifuncs;
  cd->oindex = to_index;
  cd->ofuncs = all_encodings[to_index].ofuncs;
  cd->oflags = all_encodings[to_index].oflags;
  /* Initialize the loop functions. */
#if HAVE_MBRTOWC
  if (to_wchar) {
#if HAVE_WCRTOMB
    if (from_wchar) {
      cd->lfuncs.loop_convert = wchar_id_loop_convert;
      cd->lfuncs.loop_reset = wchar_id_loop_reset;
    } else
#endif
    {
      cd->lfuncs.loop_convert = wchar_to_loop_convert;
      cd->lfuncs.loop_reset = wchar_to_loop_reset;
    }
  } else
#endif
  {
#if HAVE_WCRTOMB
    if (from_wchar) {
      cd->lfuncs.loop_convert = wchar_from_loop_convert;
      cd->lfuncs.loop_reset = wchar_from_loop_reset;
    } else
#endif
    {
      cd->lfuncs.loop_convert = unicode_loop_convert;
      cd->lfuncs.loop_reset = unicode_loop_reset;
    }
  }
  /* Initialize the states. */
  memset(&cd->istate,'\0',sizeof(state_t));
  memset(&cd->ostate,'\0',sizeof(state_t));
  /* Initialize the operation flags. */
  cd->transliterate = transliterate;
  cd->discard_ilseq = discard_ilseq;
  /* Initialize additional fields. */
  if (from_wchar != to_wchar) {
    struct wchar_conv_struct * wcd = (struct wchar_conv_struct *) cd;
    memset(&wcd->state,'\0',sizeof(mbstate_t));
  }
  /* Done. */
  return (iconv_t)cd;
invalid:
  errno = EINVAL;
  return (iconv_t)(-1);
}

size_t iconv (iconv_t icd,
              ICONV_CONST char* * inbuf, size_t *inbytesleft,
              char* * outbuf, size_t *outbytesleft)
{
  conv_t cd = (conv_t) icd;
  if (inbuf == NULL || *inbuf == NULL)
    return cd->lfuncs.loop_reset(icd,outbuf,outbytesleft);
  else
    return cd->lfuncs.loop_convert(icd,
                                   (const char* *)inbuf,inbytesleft,
                                   outbuf,outbytesleft);
}

int iconv_close (iconv_t icd)
{
  conv_t cd = (conv_t) icd;
  free(cd);
  return 0;
}

#ifndef LIBICONV_PLUG

int iconvctl (iconv_t icd, int request, void* argument)
{
  conv_t cd = (conv_t) icd;
  switch (request) {
    case ICONV_TRIVIALP:
      *(int *)argument =
        ((cd->lfuncs.loop_convert == unicode_loop_convert
          && cd->iindex == cd->oindex)
         || cd->lfuncs.loop_convert == wchar_id_loop_convert
         ? 1 : 0);
      return 0;
    case ICONV_GET_TRANSLITERATE:
      *(int *)argument = cd->transliterate;
      return 0;
    case ICONV_SET_TRANSLITERATE:
      cd->transliterate = (*(const int *)argument ? 1 : 0);
      return 0;
    case ICONV_GET_DISCARD_ILSEQ:
      *(int *)argument = cd->discard_ilseq;
      return 0;
    case ICONV_SET_DISCARD_ILSEQ:
      cd->discard_ilseq = (*(const int *)argument ? 1 : 0);
      return 0;
    default:
      errno = EINVAL;
      return -1;
  }
}

/* An alias after its name has been converted from 'int' to 'const char*'. */
struct nalias { const char* name; unsigned int encoding_index; };

static int compare_by_index (const void * arg1, const void * arg2)
{
  const struct nalias * alias1 = (const struct nalias *) arg1;
  const struct nalias * alias2 = (const struct nalias *) arg2;
  return (int)alias1->encoding_index - (int)alias2->encoding_index;
}

static int compare_by_name (const void * arg1, const void * arg2)
{
  const char * name1 = *(const char **)arg1;
  const char * name2 = *(const char **)arg2;
  /* Compare alphabetically, but put "CS" names at the end. */
  int sign = strcmp(name1,name2);
  if (sign != 0) {
    sign = ((name1[0]=='C' && name1[1]=='S') - (name2[0]=='C' && name2[1]=='S'))
           * 4 + (sign >= 0 ? 1 : -1);
  }
  return sign;
}

void iconvlist (int (*do_one) (unsigned int namescount,
                               const char * const * names,
                               void* data),
                void* data)
{
#define aliascount1  sizeof(aliases)/sizeof(aliases[0])
#ifndef aliases2_lookup
#define aliascount2  sizeof(sysdep_aliases)/sizeof(sysdep_aliases[0])
#else
#define aliascount2  0
#endif
#define aliascount  (aliascount1+aliascount2)
  struct nalias aliasbuf[aliascount];
  const char * namesbuf[aliascount];
  size_t num_aliases;
  {
    /* Put all existing aliases into a buffer. */
    size_t i;
    size_t j;
    j = 0;
    for (i = 0; i < aliascount1; i++) {
      const struct alias * p = &aliases[i];
      if (p->name >= 0
          && p->encoding_index != ei_local_char
          && p->encoding_index != ei_local_wchar_t) {
        aliasbuf[j].name = stringpool + p->name;
        aliasbuf[j].encoding_index = p->encoding_index;
        j++;
      }
    }
#ifndef aliases2_lookup
    for (i = 0; i < aliascount2; i++) {
      aliasbuf[j].name = stringpool2 + sysdep_aliases[i].name;
      aliasbuf[j].encoding_index = sysdep_aliases[i].encoding_index;
      j++;
    }
#endif
    num_aliases = j;
  }
  /* Sort by encoding_index. */
  if (num_aliases > 1)
    qsort(aliasbuf, num_aliases, sizeof(struct nalias), compare_by_index);
  {
    /* Process all aliases with the same encoding_index together. */
    size_t j;
    j = 0;
    while (j < num_aliases) {
      unsigned int ei = aliasbuf[j].encoding_index;
      size_t i = 0;
      do
        namesbuf[i++] = aliasbuf[j++].name;
      while (j < num_aliases && aliasbuf[j].encoding_index == ei);
      if (i > 1)
        qsort(namesbuf, i, sizeof(const char *), compare_by_name);
      /* Call the callback. */
      if (do_one(i,namesbuf,data))
        break;
    }
  }
#undef aliascount
#undef aliascount2
#undef aliascount1
}

int _libiconv_version = _LIBICONV_VERSION;

#if defined __FreeBSD__ && !defined __gnu_freebsd__
/* GNU libiconv is the native FreeBSD iconv implementation since 2002.
   It wants to define the symbols 'iconv_open', 'iconv', 'iconv_close'.  */
#define strong_alias(name, aliasname) _strong_alias(name, aliasname)
#define _strong_alias(name, aliasname) \
  extern __typeof (name) aliasname __attribute__ ((alias (#name)));
#undef iconv_open
#undef iconv
#undef iconv_close
strong_alias (libiconv_open, iconv_open)
strong_alias (libiconv, iconv)
strong_alias (libiconv_close, iconv_close)
#endif

#endif