lib/iso2022_jp.h

   1 /*
   2  * Copyright (C) 1999-2001, 2008, 2016 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 Lesser General Public
   7  * License as published by the Free Software Foundation; either version 2.1
   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  * Lesser General Public License for more details.
  14  *
  15  * You should have received a copy of the GNU Lesser General Public
  16  * License along with the GNU LIBICONV Library; see the file COPYING.LIB.
  17  * If not, see <https://www.gnu.org/licenses/>.
  18  */
  19
  20 /*
  21  * ISO-2022-JP
  22  */
  23
  24 /* Specification: RFC 1468 */
  25
  26 #define ESC 0x1b
  27
  28 /*
  29  * The state can be one of the following values.
  30  */
  31 #define STATE_ASCII          0
  32 #define STATE_JISX0201ROMAN  1
  33 #define STATE_JISX0208       2
  34
  35 static int
  36 iso2022_jp_mbtowc (conv_t conv, ucs4_t *pwc, const unsigned char *s, size_t n)
  37 {
  38   state_t state = conv->istate;
  39   int count = 0;
  40   unsigned char c;
  41   for (;;) {
  42     c = *s;
  43     if (c == ESC) {
  44       if (n < count+3)
  45         goto none;
  46       if (s[1] == '(') {
  47         if (s[2] == 'B') {
  48           state = STATE_ASCII;
  49           s += 3; count += 3;
  50           if (n < count+1)
  51             goto none;
  52           continue;
  53         }
  54         if (s[2] == 'J') {
  55           state = STATE_JISX0201ROMAN;
  56           s += 3; count += 3;
  57           if (n < count+1)
  58             goto none;
  59           continue;
  60         }
  61         goto ilseq;
  62       }
  63       if (s[1] == '$') {
  64         if (s[2] == '@' || s[2] == 'B') {
  65           /* We don't distinguish JIS X 0208-1978 and JIS X 0208-1983. */
  66           state = STATE_JISX0208;
  67           s += 3; count += 3;
  68           if (n < count+1)
  69             goto none;
  70           continue;
  71         }
  72         goto ilseq;
  73       }
  74       goto ilseq;
  75     }
  76     break;
  77   }
  78   switch (state) {
  79     case STATE_ASCII:
  80       if (c < 0x80) {
  81         int ret = ascii_mbtowc(conv,pwc,s,1);
  82         if (ret == RET_ILSEQ)
  83           goto ilseq;
  84         if (ret != 1) abort();
  85         conv->istate = state;
  86         return count+1;
  87       } else
  88         goto ilseq;
  89     case STATE_JISX0201ROMAN:
  90       if (c < 0x80) {
  91         int ret = jisx0201_mbtowc(conv,pwc,s,1);
  92         if (ret == RET_ILSEQ)
  93           goto ilseq;
  94         if (ret != 1) abort();
  95         conv->istate = state;
  96         return count+1;
  97       } else
  98         goto ilseq;
  99     case STATE_JISX0208:
 100       if (n < count+2)
 101         goto none;
 102       if (s[0] < 0x80 && s[1] < 0x80) {
 103         int ret = jisx0208_mbtowc(conv,pwc,s,2);
 104         if (ret == RET_ILSEQ)
 105           goto ilseq;
 106         if (ret != 2) abort();
 107         conv->istate = state;
 108         return count+2;
 109       } else
 110         goto ilseq;
 111     default: abort();
 112   }
 113
 114 none:
 115   conv->istate = state;
 116   return RET_TOOFEW(count);
 117
 118 ilseq:
 119   conv->istate = state;
 120   return RET_SHIFT_ILSEQ(count);
 121 }
 122
 123 static int
 124 iso2022_jp_wctomb (conv_t conv, unsigned char *r, ucs4_t wc, size_t n)
 125 {
 126   state_t state = conv->ostate;
 127   unsigned char buf[2];
 128   int ret;
 129
 130   /* Try ASCII. */
 131   ret = ascii_wctomb(conv,buf,wc,1);
 132   if (ret != RET_ILUNI) {
 133     if (ret != 1) abort();
 134     if (buf[0] < 0x80) {
 135       int count = (state == STATE_ASCII ? 1 : 4);
 136       if (n < count)
 137         return RET_TOOSMALL;
 138       if (state != STATE_ASCII) {
 139         r[0] = ESC;
 140         r[1] = '(';
 141         r[2] = 'B';
 142         r += 3;
 143         state = STATE_ASCII;
 144       }
 145       r[0] = buf[0];
 146       conv->ostate = state;
 147       return count;
 148     }
 149   }
 150
 151   /* Try JIS X 0201-1976 Roman. */
 152   ret = jisx0201_wctomb(conv,buf,wc,1);
 153   if (ret != RET_ILUNI) {
 154     if (ret != 1) abort();
 155     if (buf[0] < 0x80) {
 156       int count = (state == STATE_JISX0201ROMAN ? 1 : 4);
 157       if (n < count)
 158         return RET_TOOSMALL;
 159       if (state != STATE_JISX0201ROMAN) {
 160         r[0] = ESC;
 161         r[1] = '(';
 162         r[2] = 'J';
 163         r += 3;
 164         state = STATE_JISX0201ROMAN;
 165       }
 166       r[0] = buf[0];
 167       conv->ostate = state;
 168       return count;
 169     }
 170   }
 171
 172   /* Try JIS X 0208-1990 in place of JIS X 0208-1978 and JIS X 0208-1983. */
 173   ret = jisx0208_wctomb(conv,buf,wc,2);
 174   if (ret != RET_ILUNI) {
 175     if (ret != 2) abort();
 176     if (buf[0] < 0x80 && buf[1] < 0x80) {
 177       int count = (state == STATE_JISX0208 ? 2 : 5);
 178       if (n < count)
 179         return RET_TOOSMALL;
 180       if (state != STATE_JISX0208) {
 181         r[0] = ESC;
 182         r[1] = '$';
 183         r[2] = 'B';
 184         r += 3;
 185         state = STATE_JISX0208;
 186       }
 187       r[0] = buf[0];
 188       r[1] = buf[1];
 189       conv->ostate = state;
 190       return count;
 191     }
 192   }
 193
 194   return RET_ILUNI;
 195 }
 196
 197 static int
 198 iso2022_jp_reset (conv_t conv, unsigned char *r, size_t n)
 199 {
 200   state_t state = conv->ostate;
 201   if (state != STATE_ASCII) {
 202     if (n < 3)
 203       return RET_TOOSMALL;
 204     r[0] = ESC;
 205     r[1] = '(';
 206     r[2] = 'B';
 207     /* conv->ostate = 0; will be done by the caller */
 208     return 3;
 209   } else
 210     return 0;
 211 }
 212
 213 #undef STATE_JISX0208
 214 #undef STATE_JISX0201ROMAN
 215 #undef STATE_ASCII