lib/sjis.h

   1 /*
   2  * Copyright (C) 1999-2002, 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  * SHIFT_JIS
  22  */
  23
  24 /*
  25    Conversion between SJIS codes (s1,s2) and JISX0208 codes (c1,c2):
  26    Example. (s1,s2) = 0x8140, (c1,c2) = 0x2121.
  27    0x81 <= s1 <= 0x9F || 0xE0 <= s1 <= 0xEA,
  28    0x40 <= s2 <= 0x7E || 0x80 <= s2 <= 0xFC,
  29    0x21 <= c1 <= 0x74, 0x21 <= c2 <= 0x7E.
  30    Invariant:
  31      94*2*(s1 < 0xE0 ? s1-0x81 : s1-0xC1) + (s2 < 0x80 ? s2-0x40 : s2-0x41)
  32      = 94*(c1-0x21)+(c2-0x21)
  33    Conversion (s1,s2) -> (c1,c2):
  34      t1 := (s1 < 0xE0 ? s1-0x81 : s1-0xC1)
  35      t2 := (s2 < 0x80 ? s2-0x40 : s2-0x41)
  36      c1 := 2*t1 + (t2 < 0x5E ? 0 : 1) + 0x21
  37      c2 := (t2 < 0x5E ? t2 : t2-0x5E) + 0x21
  38    Conversion (c1,c2) -> (s1,s2):
  39      t1 := (c1 - 0x21) >> 1
  40      t2 := ((c1 - 0x21) & 1) * 0x5E + (c2 - 0x21)
  41      s1 := (t1 < 0x1F ? t1+0x81 : t1+0xC1)
  42      s2 := (t2 < 0x3F ? t2+0x40 : t2+0x41)
  43  */
  44
  45 static int
  46 sjis_mbtowc (conv_t conv, ucs4_t *pwc, const unsigned char *s, size_t n)
  47 {
  48   unsigned char c = *s;
  49   if (c < 0x80 || (c >= 0xa1 && c <= 0xdf))
  50     return jisx0201_mbtowc(conv,pwc,s,n);
  51   else {
  52     unsigned char s1, s2;
  53     s1 = c;
  54     if ((s1 >= 0x81 && s1 <= 0x9f) || (s1 >= 0xe0 && s1 <= 0xea)) {
  55       if (n < 2)
  56         return RET_TOOFEW(0);
  57       s2 = s[1];
  58       if ((s2 >= 0x40 && s2 <= 0x7e) || (s2 >= 0x80 && s2 <= 0xfc)) {
  59         unsigned char t1 = (s1 < 0xe0 ? s1-0x81 : s1-0xc1);
  60         unsigned char t2 = (s2 < 0x80 ? s2-0x40 : s2-0x41);
  61         unsigned char buf[2];
  62         buf[0] = 2*t1 + (t2 < 0x5e ? 0 : 1) + 0x21;
  63         buf[1] = (t2 < 0x5e ? t2 : t2-0x5e) + 0x21;
  64         return jisx0208_mbtowc(conv,pwc,buf,2);
  65       }
  66     } else if (s1 >= 0xf0 && s1 <= 0xf9) {
  67       /* User-defined range. See
  68        * Ken Lunde's "CJKV Information Processing", table 4-66, p. 206. */
  69       if (n < 2)
  70         return RET_TOOFEW(0);
  71       s2 = s[1];
  72       if ((s2 >= 0x40 && s2 <= 0x7e) || (s2 >= 0x80 && s2 <= 0xfc)) {
  73         *pwc = 0xe000 + 188*(s1 - 0xf0) + (s2 < 0x80 ? s2-0x40 : s2-0x41);
  74         return 2;
  75       }
  76     }
  77     return RET_ILSEQ;
  78   }
  79 }
  80
  81 static int
  82 sjis_wctomb (conv_t conv, unsigned char *r, ucs4_t wc, size_t n)
  83 {
  84   unsigned char buf[2];
  85   int ret;
  86
  87   /* Try JIS X 0201-1976. */
  88   ret = jisx0201_wctomb(conv,buf,wc,1);
  89   if (ret != RET_ILUNI) {
  90     unsigned char c;
  91     if (ret != 1) abort();
  92     c = buf[0];
  93     if (c < 0x80 || (c >= 0xa1 && c <= 0xdf)) {
  94       r[0] = c;
  95       return 1;
  96     }
  97   }
  98
  99   /* Try JIS X 0208-1990. */
 100   ret = jisx0208_wctomb(conv,buf,wc,2);
 101   if (ret != RET_ILUNI) {
 102     unsigned char c1, c2;
 103     if (ret != 2) abort();
 104     if (n < 2)
 105       return RET_TOOSMALL;
 106     c1 = buf[0];
 107     c2 = buf[1];
 108     if ((c1 >= 0x21 && c1 <= 0x74) && (c2 >= 0x21 && c2 <= 0x7e)) {
 109       unsigned char t1 = (c1 - 0x21) >> 1;
 110       unsigned char t2 = (((c1 - 0x21) & 1) ? 0x5e : 0) + (c2 - 0x21);
 111       r[0] = (t1 < 0x1f ? t1+0x81 : t1+0xc1);
 112       r[1] = (t2 < 0x3f ? t2+0x40 : t2+0x41);
 113       return 2;
 114     }
 115   }
 116
 117   /* User-defined range. See
 118    * Ken Lunde's "CJKV Information Processing", table 4-66, p. 206. */
 119   if (wc >= 0xe000 && wc < 0xe758) {
 120     unsigned char c1, c2;
 121     if (n < 2)
 122       return RET_TOOSMALL;
 123     c1 = (unsigned int) (wc - 0xe000) / 188;
 124     c2 = (unsigned int) (wc - 0xe000) % 188;
 125     r[0] = c1+0xf0;
 126     r[1] = (c2 < 0x3f ? c2+0x40 : c2+0x41);
 127     return 2;
 128   }
 129
 130   return RET_ILUNI;
 131 }