lib/iso2022_jp1.h

   1 /*
   2  * Copyright (C) 1999-2000 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-1
  23  */
  24
  25 /* Specification: RFC 2237 */
  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 #define STATE_JISX0212       3
  36
  37 static int
  38 iso2022_jp1_mbtowc (conv_t conv, ucs4_t *pwc, const unsigned char *s, int n)
  39 {
  40   state_t state = conv->istate;
  41   int count = 0;
  42   unsigned char c;
  43   for (;;) {
  44     c = *s;
  45     if (c == ESC) {
  46       if (n < count+3)
  47         goto none;
  48       if (s[1] == '(') {
  49         if (s[2] == 'B') {
  50           state = STATE_ASCII;
  51           s += 3; count += 3;
  52           if (n < count+1)
  53             goto none;
  54           continue;
  55         }
  56         if (s[2] == 'J') {
  57           state = STATE_JISX0201ROMAN;
  58           s += 3; count += 3;
  59           if (n < count+1)
  60             goto none;
  61           continue;
  62         }
  63         return RET_ILSEQ;
  64       }
  65       if (s[1] == '$') {
  66         if (s[2] == '@' || s[2] == 'B') {
  67           /* We don't distinguish JIS X 0208-1978 and JIS X 0208-1983. */
  68           state = STATE_JISX0208;
  69           s += 3; count += 3;
  70           if (n < count+1)
  71             goto none;
  72           continue;
  73         }
  74         if (s[2] == '(') {
  75           if (n < count+4)
  76             goto none;
  77           if (s[3] == 'D') {
  78             state = STATE_JISX0212;
  79             s += 4; count += 4;
  80             if (n < count+1)
  81               goto none;
  82             continue;
  83           }
  84         }
  85         return RET_ILSEQ;
  86       }
  87       return RET_ILSEQ;
  88     }
  89     break;
  90   }
  91   switch (state) {
  92     case STATE_ASCII:
  93       if (c < 0x80) {
  94         int ret = ascii_mbtowc(conv,pwc,s,1);
  95         if (ret == RET_ILSEQ)
  96           return RET_ILSEQ;
  97         if (ret != 1) abort();
  98         conv->istate = state;
  99         return count+1;
 100       } else
 101         return RET_ILSEQ;
 102     case STATE_JISX0201ROMAN:
 103       if (c < 0x80) {
 104         int ret = jisx0201_mbtowc(conv,pwc,s,1);
 105         if (ret == RET_ILSEQ)
 106           return RET_ILSEQ;
 107         if (ret != 1) abort();
 108         conv->istate = state;
 109         return count+1;
 110       } else
 111         return RET_ILSEQ;
 112     case STATE_JISX0208:
 113       if (n < count+2)
 114         goto none;
 115       if (s[0] < 0x80 && s[1] < 0x80) {
 116         int ret = jisx0208_mbtowc(conv,pwc,s,2);
 117         if (ret == RET_ILSEQ)
 118           return RET_ILSEQ;
 119         if (ret != 2) abort();
 120         conv->istate = state;
 121         return count+2;
 122       } else
 123         return RET_ILSEQ;
 124     case STATE_JISX0212:
 125       if (n < count+2)
 126         goto none;
 127       if (s[0] < 0x80 && s[1] < 0x80) {
 128         int ret = jisx0212_mbtowc(conv,pwc,s,2);
 129         if (ret == RET_ILSEQ)
 130           return RET_ILSEQ;
 131         if (ret != 2) abort();
 132         conv->istate = state;
 133         return count+2;
 134       } else
 135         return RET_ILSEQ;
 136     default: abort();
 137   }
 138
 139 none:
 140   conv->istate = state;
 141   return RET_TOOFEW(count);
 142 }
 143
 144 static int
 145 iso2022_jp1_wctomb (conv_t conv, unsigned char *r, ucs4_t wc, int n)
 146 {
 147   state_t state = conv->ostate;
 148   unsigned char buf[2];
 149   int ret;
 150
 151   /* Try ASCII. */
 152   ret = ascii_wctomb(conv,buf,wc,1);
 153   if (ret != RET_ILSEQ) {
 154     if (ret != 1) abort();
 155     if (buf[0] < 0x80) {
 156       int count = (state == STATE_ASCII ? 1 : 4);
 157       if (n < count)
 158         return RET_TOOSMALL;
 159       if (state != STATE_ASCII) {
 160         r[0] = ESC;
 161         r[1] = '(';
 162         r[2] = 'B';
 163         r += 3;
 164         state = STATE_ASCII;
 165       }
 166       r[0] = buf[0];
 167       conv->ostate = state;
 168       return count;
 169     }
 170   }
 171
 172   /* Try JIS X 0201-1976 Roman. */
 173   ret = jisx0201_wctomb(conv,buf,wc,1);
 174   if (ret != RET_ILSEQ) {
 175     if (ret != 1) abort();
 176     if (buf[0] < 0x80) {
 177       int count = (state == STATE_JISX0201ROMAN ? 1 : 4);
 178       if (n < count)
 179         return RET_TOOSMALL;
 180       if (state != STATE_JISX0201ROMAN) {
 181         r[0] = ESC;
 182         r[1] = '(';
 183         r[2] = 'J';
 184         r += 3;
 185         state = STATE_JISX0201ROMAN;
 186       }
 187       r[0] = buf[0];
 188       conv->ostate = state;
 189       return count;
 190     }
 191   }
 192
 193   /* Try JIS X 0208-1990 in place of JIS X 0208-1978 and JIS X 0208-1983. */
 194   ret = jisx0208_wctomb(conv,buf,wc,2);
 195   if (ret != RET_ILSEQ) {
 196     if (ret != 2) abort();
 197     if (buf[0] < 0x80 && buf[1] < 0x80) {
 198       int count = (state == STATE_JISX0208 ? 2 : 5);
 199       if (n < count)
 200         return RET_TOOSMALL;
 201       if (state != STATE_JISX0208) {
 202         r[0] = ESC;
 203         r[1] = '$';
 204         r[2] = 'B';
 205         r += 3;
 206         state = STATE_JISX0208;
 207       }
 208       r[0] = buf[0];
 209       r[1] = buf[1];
 210       conv->ostate = state;
 211       return count;
 212     }
 213   }
 214
 215   /* Try JIS X 0212-1990. */
 216   ret = jisx0212_wctomb(conv,buf,wc,2);
 217   if (ret != RET_ILSEQ) {
 218     if (ret != 2) abort();
 219     if (buf[0] < 0x80 && buf[1] < 0x80) {
 220       int count = (state == STATE_JISX0212 ? 2 : 6);
 221       if (n < count)
 222         return RET_TOOSMALL;
 223       if (state != STATE_JISX0212) {
 224         r[0] = ESC;
 225         r[1] = '$';
 226         r[2] = '(';
 227         r[3] = 'D';
 228         r += 4;
 229         state = STATE_JISX0212;
 230       }
 231       r[0] = buf[0];
 232       r[1] = buf[1];
 233       conv->ostate = state;
 234       return count;
 235     }
 236   }
 237
 238   return RET_ILSEQ;
 239 }
 240
 241 static int
 242 iso2022_jp1_reset (conv_t conv, unsigned char *r, int n)
 243 {
 244   state_t state = conv->ostate;
 245   if (state != STATE_ASCII) {
 246     if (n < 3)
 247       return RET_TOOSMALL;
 248     r[0] = ESC;
 249     r[1] = '(';
 250     r[2] = 'B';
 251     /* conv->ostate = 0; will be done by the caller */
 252     return 3;
 253   } else
 254     return 0;
 255 }
 256
 257 #undef STATE_JISX0212
 258 #undef STATE_JISX0208
 259 #undef STATE_JISX0201ROMAN
 260 #undef STATE_ASCII