release/src/router/libiconv/lib/iconv_open1.h

   1 /*
   2  * Copyright (C) 1999-2008, 2011 Free Software Foundation, Inc.
   3  * This file is part of the GNU LIBICONV Library.
   4  *
   5  * The GNU LIBICONV Library is free software; you can redistribute it
   6  * and/or modify it under the terms of the GNU Library General Public
   7  * License as published by the Free Software Foundation; either version 2
   8  * of the License, or (at your option) any later version.
   9  *
  10  * The GNU LIBICONV Library is distributed in the hope that it will be
  11  * useful, but WITHOUT ANY WARRANTY; without even the implied warranty of
  12  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
  13  * Library General Public License for more details.
  14  *
  15  * You should have received a copy of the GNU Library General Public
  16  * License along with the GNU LIBICONV Library; see the file COPYING.LIB.
  17  * If not, write to the Free Software Foundation, Inc., 51 Franklin Street,
  18  * Fifth Floor, Boston, MA 02110-1301, USA.
  19  */
  20
  21 /* Part 1 of iconv_open.
  22    Input: const char* tocode, const char* fromcode.
  23    Output:
  24      unsigned int from_index;
  25      int from_wchar;
  26      unsigned int to_index;
  27      int to_wchar;
  28      int transliterate;
  29      int discard_ilseq;
  30    Jumps to 'invalid' in case of errror.
  31  */
  32 {
  33   char buf[MAX_WORD_LENGTH+10+1];
  34   const char* cp;
  35   char* bp;
  36   const struct alias * ap;
  37   unsigned int count;
  38
  39   transliterate = 0;
  40   discard_ilseq = 0;
  41
  42   /* Before calling aliases_lookup, convert the input string to upper case,
  43    * and check whether it's entirely ASCII (we call gperf with option "-7"
  44    * to achieve a smaller table) and non-empty. If it's not entirely ASCII,
  45    * or if it's too long, it is not a valid encoding name.
  46    */
  47   for (to_wchar = 0;;) {
  48     /* Search tocode in the table. */
  49     for (cp = tocode, bp = buf, count = MAX_WORD_LENGTH+10+1; ; cp++, bp++) {
  50       unsigned char c = * (unsigned char *) cp;
  51       if (c >= 0x80)
  52         goto invalid;
  53       if (c >= 'a' && c <= 'z')
  54         c -= 'a'-'A';
  55       *bp = c;
  56       if (c == '\0')
  57         break;
  58       if (--count == 0)
  59         goto invalid;
  60     }
  61     for (;;) {
  62       if (bp-buf >= 10 && memcmp(bp-10,"//TRANSLIT",10)==0) {
  63         bp -= 10;
  64         *bp = '\0';
  65         transliterate = 1;
  66         continue;
  67       }
  68       if (bp-buf >= 8 && memcmp(bp-8,"//IGNORE",8)==0) {
  69         bp -= 8;
  70         *bp = '\0';
  71         discard_ilseq = 1;
  72         continue;
  73       }
  74       break;
  75     }
  76     if (buf[0] == '\0') {
  77       tocode = locale_charset();
  78       /* Avoid an endless loop that could occur when using an older version
  79          of localcharset.c. */
  80       if (tocode[0] == '\0')
  81         goto invalid;
  82       continue;
  83     }
  84     ap = aliases_lookup(buf,bp-buf);
  85     if (ap == NULL) {
  86       ap = aliases2_lookup(buf);
  87       if (ap == NULL)
  88         goto invalid;
  89     }
  90     if (ap->encoding_index == ei_local_char) {
  91       tocode = locale_charset();
  92       /* Avoid an endless loop that could occur when using an older version
  93          of localcharset.c. */
  94       if (tocode[0] == '\0')
  95         goto invalid;
  96       continue;
  97     }
  98     if (ap->encoding_index == ei_local_wchar_t) {
  99       /* On systems which define __STDC_ISO_10646__, wchar_t is Unicode.
 100          This is also the case on native Woe32 systems and Cygwin >= 1.7, where
 101          we know that it is UTF-16.  */
 102 #if ((defined _WIN32 || defined __WIN32__) && !defined __CYGWIN__) || (defined __CYGWIN__ && CYGWIN_VERSION_DLL_MAJOR >= 1007)
 103       if (sizeof(wchar_t) == 4) {
 104         to_index = ei_ucs4internal;
 105         break;
 106       }
 107       if (sizeof(wchar_t) == 2) {
 108 # if WORDS_LITTLEENDIAN
 109         to_index = ei_utf16le;
 110 # else
 111         to_index = ei_utf16be;
 112 # endif
 113         break;
 114       }
 115 #elif __STDC_ISO_10646__
 116       if (sizeof(wchar_t) == 4) {
 117         to_index = ei_ucs4internal;
 118         break;
 119       }
 120       if (sizeof(wchar_t) == 2) {
 121         to_index = ei_ucs2internal;
 122         break;
 123       }
 124       if (sizeof(wchar_t) == 1) {
 125         to_index = ei_iso8859_1;
 126         break;
 127       }
 128 #endif
 129 #if HAVE_MBRTOWC
 130       to_wchar = 1;
 131       tocode = locale_charset();
 132       continue;
 133 #endif
 134       goto invalid;
 135     }
 136     to_index = ap->encoding_index;
 137     break;
 138   }
 139   for (from_wchar = 0;;) {
 140     /* Search fromcode in the table. */
 141     for (cp = fromcode, bp = buf, count = MAX_WORD_LENGTH+10+1; ; cp++, bp++) {
 142       unsigned char c = * (unsigned char *) cp;
 143       if (c >= 0x80)
 144         goto invalid;
 145       if (c >= 'a' && c <= 'z')
 146         c -= 'a'-'A';
 147       *bp = c;
 148       if (c == '\0')
 149         break;
 150       if (--count == 0)
 151         goto invalid;
 152     }
 153     for (;;) {
 154       if (bp-buf >= 10 && memcmp(bp-10,"//TRANSLIT",10)==0) {
 155         bp -= 10;
 156         *bp = '\0';
 157         continue;
 158       }
 159       if (bp-buf >= 8 && memcmp(bp-8,"//IGNORE",8)==0) {
 160         bp -= 8;
 161         *bp = '\0';
 162         continue;
 163       }
 164       break;
 165     }
 166     if (buf[0] == '\0') {
 167       fromcode = locale_charset();
 168       /* Avoid an endless loop that could occur when using an older version
 169          of localcharset.c. */
 170       if (fromcode[0] == '\0')
 171         goto invalid;
 172       continue;
 173     }
 174     ap = aliases_lookup(buf,bp-buf);
 175     if (ap == NULL) {
 176       ap = aliases2_lookup(buf);
 177       if (ap == NULL)
 178         goto invalid;
 179     }
 180     if (ap->encoding_index == ei_local_char) {
 181       fromcode = locale_charset();
 182       /* Avoid an endless loop that could occur when using an older version
 183          of localcharset.c. */
 184       if (fromcode[0] == '\0')
 185         goto invalid;
 186       continue;
 187     }
 188     if (ap->encoding_index == ei_local_wchar_t) {
 189       /* On systems which define __STDC_ISO_10646__, wchar_t is Unicode.
 190          This is also the case on native Woe32 systems and Cygwin >= 1.7, where
 191          we know that it is UTF-16.  */
 192 #if ((defined _WIN32 || defined __WIN32__) && !defined __CYGWIN__) || (defined __CYGWIN__ && CYGWIN_VERSION_DLL_MAJOR >= 1007)
 193       if (sizeof(wchar_t) == 4) {
 194         from_index = ei_ucs4internal;
 195         break;
 196       }
 197       if (sizeof(wchar_t) == 2) {
 198 # if WORDS_LITTLEENDIAN
 199         from_index = ei_utf16le;
 200 # else
 201         from_index = ei_utf16be;
 202 # endif
 203         break;
 204       }
 205 #elif __STDC_ISO_10646__
 206       if (sizeof(wchar_t) == 4) {
 207         from_index = ei_ucs4internal;
 208         break;
 209       }
 210       if (sizeof(wchar_t) == 2) {
 211         from_index = ei_ucs2internal;
 212         break;
 213       }
 214       if (sizeof(wchar_t) == 1) {
 215         from_index = ei_iso8859_1;
 216         break;
 217       }
 218 #endif
 219 #if HAVE_WCRTOMB
 220       from_wchar = 1;
 221       fromcode = locale_charset();
 222       continue;
 223 #endif
 224       goto invalid;
 225     }
 226     from_index = ap->encoding_index;
 227     break;
 228   }
 229 }