lib/iso2022_jp.h

   1 /*
   2  * Copyright (C) 1999-2001 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., 59 Temple Place -
  18  * Suite 330, Boston, MA 02111-1307, USA.
  19  */
  20
  21 /*
  22  * ISO-2022-JP
  23  */
  24
  25 /* Specification: RFC 1468 */
  26
  27 #define ESC 0x1b
  28
  29 /*
  30  * The state can be one of the following values.
  31  */
  32 #define STATE_ASCII          0
  33 #define STATE_JISX0201ROMAN  1
  34 #define STATE_JISX0208       2
  35
  36 static int
  37 iso2022_jp_mbtowc (conv_t conv, ucs4_t *pwc, const unsigned char *s, int n)
  38 {
  39   state_t state = conv->istate;
  40   int count = 0;
  41   unsigned char c;
  42   for (;;) {
  43     c = *s;
  44     if (c == ESC) {
  45       if (n < count+3)
  46         goto none;
  47       if (s[1] == '(') {
  48         if (s[2] == 'B') {
  49           state = STATE_ASCII;
  50           s += 3; count += 3;
  51           if (n < count+1)
  52             goto none;
  53           continue;
  54         }
  55         if (s[2] == 'J') {
  56           state = STATE_JISX0201ROMAN;
  57           s += 3; count += 3;
  58           if (n < count+1)
  59             goto none;
  60           continue;
  61         }
  62         return RET_ILSEQ;
  63       }
  64       if (s[1] == '$') {
  65         if (s[2] == '@' || s[2] == 'B') {
  66           /* We don't distinguish JIS X 0208-1978 and JIS X 0208-1983. */
  67           state = STATE_JISX0208;
  68           s += 3; count += 3;
  69           if (n < count+1)
  70             goto none;
  71           continue;
  72         }
  73         return RET_ILSEQ;
  74       }
  75       return RET_ILSEQ;
  76     }
  77     break;
  78   }
  79   switch (state) {
  80     case STATE_ASCII:
  81       if (c < 0x80) {
  82         int ret = ascii_mbtowc(conv,pwc,s,1);
  83         if (ret == RET_ILSEQ)
  84           return RET_ILSEQ;
  85         if (ret != 1) abort();
  86         conv->istate = state;
  87         return count+1;
  88       } else
  89         return RET_ILSEQ;
  90     case STATE_JISX0201ROMAN:
  91       if (c < 0x80) {
  92         int ret = jisx0201_mbtowc(conv,pwc,s,1);
  93         if (ret == RET_ILSEQ)
  94           return RET_ILSEQ;
  95         if (ret != 1) abort();
  96         conv->istate = state;
  97         return count+1;
  98       } else
  99         return RET_ILSEQ;
 100     case STATE_JISX0208:
 101       if (n < count+2)
 102         goto none;
 103       if (s[0] < 0x80 && s[1] < 0x80) {
 104         int ret = jisx0208_mbtowc(conv,pwc,s,2);
 105         if (ret == RET_ILSEQ)
 106           return RET_ILSEQ;
 107         if (ret != 2) abort();
 108         conv->istate = state;
 109         return count+2;
 110       } else
 111         return RET_ILSEQ;
 112     default: abort();
 113   }
 114
 115 none:
 116   conv->istate = state;
 117   return RET_TOOFEW(count);
 118 }
 119
 120 static int
 121 iso2022_jp_wctomb (conv_t conv, unsigned char *r, ucs4_t wc, int n)
 122 {
 123   state_t state = conv->ostate;
 124   unsigned char buf[2];
 125   int ret;
 126
 127   /* Try ASCII. */
 128   ret = ascii_wctomb(conv,buf,wc,1);
 129   if (ret != RET_ILUNI) {
 130     if (ret != 1) abort();
 131     if (buf[0] < 0x80) {
 132       int count = (state == STATE_ASCII ? 1 : 4);
 133       if (n < count)
 134         return RET_TOOSMALL;
 135       if (state != STATE_ASCII) {
 136         r[0] = ESC;
 137         r[1] = '(';
 138         r[2] = 'B';
 139         r += 3;
 140         state = STATE_ASCII;
 141       }
 142       r[0] = buf[0];
 143       conv->ostate = state;
 144       return count;
 145     }
 146   }
 147
 148   /* Try JIS X 0201-1976 Roman. */
 149   ret = jisx0201_wctomb(conv,buf,wc,1);
 150   if (ret != RET_ILUNI) {
 151     if (ret != 1) abort();
 152     if (buf[0] < 0x80) {
 153       int count = (state == STATE_JISX0201ROMAN ? 1 : 4);
 154       if (n < count)
 155         return RET_TOOSMALL;
 156       if (state != STATE_JISX0201ROMAN) {
 157         r[0] = ESC;
 158         r[1] = '(';
 159         r[2] = 'J';
 160         r += 3;
 161         state = STATE_JISX0201ROMAN;
 162       }
 163       r[0] = buf[0];
 164       conv->ostate = state;
 165       return count;
 166     }
 167   }
 168
 169   /* Try JIS X 0208-1990 in place of JIS X 0208-1978 and JIS X 0208-1983. */
 170   ret = jisx0208_wctomb(conv,buf,wc,2);
 171   if (ret != RET_ILUNI) {
 172     if (ret != 2) abort();
 173     if (buf[0] < 0x80 && buf[1] < 0x80) {
 174       int count = (state == STATE_JISX0208 ? 2 : 5);
 175       if (n < count)
 176         return RET_TOOSMALL;
 177       if (state != STATE_JISX0208) {
 178         r[0] = ESC;
 179         r[1] = '$';
 180         r[2] = 'B';
 181         r += 3;
 182         state = STATE_JISX0208;
 183       }
 184       r[0] = buf[0];
 185       r[1] = buf[1];
 186       conv->ostate = state;
 187       return count;
 188     }
 189   }
 190
 191   return RET_ILUNI;
 192 }
 193
 194 static int
 195 iso2022_jp_reset (conv_t conv, unsigned char *r, int n)
 196 {
 197   state_t state = conv->ostate;
 198   if (state != STATE_ASCII) {
 199     if (n < 3)
 200       return RET_TOOSMALL;
 201     r[0] = ESC;
 202     r[1] = '(';
 203     r[2] = 'B';
 204     /* conv->ostate = 0; will be done by the caller */
 205     return 3;
 206   } else
 207     return 0;
 208 }
 209
 210 #undef STATE_JISX0208
 211 #undef STATE_JISX0201ROMAN
 212 #undef STATE_ASCII