Return to johab.h CVS log

Up to [ELWIX - Embedded LightWeight unIX -] / embedaddon / libiconv / lib

libiconv

    1: /*
    2:  * Copyright (C) 1999-2001, 2007 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: /*
   22:  * JOHAB
   23:  */
   24: 
   25: /*
   26:    Conversion between JOHAB codes (s1,s2) and KSX1001 codes (c1,c2):
   27:    Example. (s1,s2) = 0xD931, (c1,c2) = 0x2121.
   28:             (s1,s2) = 0xDEF1, (c1,c2) = 0x2C71.
   29:             (s1,s2) = 0xE031, (c1,c2) = 0x4A21.
   30:             (s1,s2) = 0xF9FE, (c1,c2) = 0x7D7E.
   31:    0xD9 <= s1 <= 0xDE || 0xE0 <= s1 <= 0xF9,
   32:    0x31 <= s2 <= 0x7E || 0x91 <= s2 <= 0xFE,
   33:    0x21 <= c1 <= 0x2C || 0x4A <= c1 <= 0x7D,
   34:    0x21 <= c2 <= 0x7E.
   35:    Invariant:
   36:      94*(s1 < 0xE0 ? 2*s1-0x1B2 : 2*s1-0x197) + (s2 < 0x91 ? s2-0x31 : s2-0x43)
   37:      = 94*(c1-0x21)+(c2-0x21)
   38:    Conversion (s1,s2) -> (c1,c2):
   39:      t1 := (s1 < 0xE0 ? 2*s1-0x1B2 : 2*s1-0x197)
   40:      t2 := (s2 < 0x91 ? s2-0x31 : s2-0x43)
   41:      c1 := t1 + (t2 < 0x5E ? 0 : 1) + 0x21
   42:      c2 := (t2 < 0x5E ? t2 : t2-0x5E) + 0x21
   43:    Conversion (c1,c2) -> (s1,s2):
   44:      t := (c1 < 0x4A ? (c1-0x21+0x1B2) : (c1-0x21+0x197))
   45:      s1 := t >> 1
   46:      t2 := (t & 1) * 0x5E + (c2 - 0x21)
   47:      s2 := (t2 < 0x4E ? t2+0x31 : t2+0x43)
   48:  */
   49: 
   50: static int
   51: johab_mbtowc (conv_t conv, ucs4_t *pwc, const unsigned char *s, int n)
   52: {
   53:   unsigned char c = *s;
   54:   if (c < 0x80) {
   55:     if (c == 0x5c)
   56:       *pwc = (ucs4_t) 0x20a9;
   57:     else
   58:       *pwc = (ucs4_t) c;
   59:     return 1;
   60:   } else if (c < 0xd8) {
   61:     return johab_hangul_mbtowc(conv,pwc,s,n);
   62:   } else {
   63:     unsigned char s1, s2;
   64:     s1 = c;
   65:     if ((s1 >= 0xd9 && s1 <= 0xde) || (s1 >= 0xe0 && s1 <= 0xf9)) {
   66:       if (n < 2)
   67:         return RET_TOOFEW(0);
   68:       s2 = s[1];
   69:       if ((s2 >= 0x31 && s2 <= 0x7e) || (s2 >= 0x91 && s2 <= 0xfe)) {
   70:         /* In KSC 5601, now KS X 1001, the region s1 = 0xDA, 0xA1 <= s2 <= 0xD3
   71:            contains the 51 Jamo (Hangul letters). But in the Johab encoding,
   72:            they have been moved to the Hangul section, see
   73:            johab_hangul_page31. */
   74:         if (!(s1 == 0xda && (s2 >= 0xa1 && s2 <= 0xd3))) {
   75:           unsigned char t1 = (s1 < 0xe0 ? 2*(s1-0xd9) : 2*s1-0x197);
   76:           unsigned char t2 = (s2 < 0x91 ? s2-0x31 : s2-0x43);
   77:           unsigned char buf[2];
   78:           buf[0] = t1 + (t2 < 0x5e ? 0 : 1) + 0x21;
   79:           buf[1] = (t2 < 0x5e ? t2 : t2-0x5e) + 0x21;
   80:           return ksc5601_mbtowc(conv,pwc,buf,2);
   81:         }
   82:       }
   83:     }
   84:     return RET_ILSEQ;
   85:   }
   86: }
   87: 
   88: static int
   89: johab_wctomb (conv_t conv, unsigned char *r, ucs4_t wc, int n)
   90: {
   91:   unsigned char buf[2];
   92:   int ret;
   93: 
   94:   /* Try ASCII variation. */
   95:   if (wc < 0x0080 && wc != 0x005c) {
   96:     *r = wc;
   97:     return 1;
   98:   }
   99:   if (wc == 0x20a9) {
  100:     *r = 0x5c;
  101:     return 1;
  102:   }
  103: 
  104:   /* Try JOHAB Hangul table before KSC5601 table, because the KSC5601 table
  105:      contains some (2350 out of 11172) Hangul syllables (rows 0x30XX..0x48XX),
  106:      and we want the search to return the JOHAB Hangul table entry. */
  107: 
  108:   /* Try JOHAB Hangul. */
  109:   ret = johab_hangul_wctomb(conv,buf,wc,2);
  110:   if (ret != RET_ILUNI) {
  111:     if (ret != 2) abort();
  112:     if (n < 2)
  113:       return RET_TOOSMALL;
  114:     r[0] = buf[0];
  115:     r[1] = buf[1];
  116:     return 2;
  117:   }
  118: 
  119:   /* Try KSC5601, now KS X 1001. */
  120:   ret = ksc5601_wctomb(conv,buf,wc,2);
  121:   if (ret != RET_ILUNI) {
  122:     unsigned char c1, c2;
  123:     if (ret != 2) abort();
  124:     if (n < 2)
  125:       return RET_TOOSMALL;
  126:     c1 = buf[0];
  127:     c2 = buf[1];
  128:     if (((c1 >= 0x21 && c1 <= 0x2c) || (c1 >= 0x4a && c1 <= 0x7d))
  129:         && (c2 >= 0x21 && c2 <= 0x7e)) {
  130:       unsigned int t = (c1 < 0x4A ? (c1-0x21+0x1B2) : (c1-0x21+0x197));
  131:       unsigned char t2 = ((t & 1) ? 0x5e : 0) + (c2 - 0x21);
  132:       r[0] = t >> 1;
  133:       r[1] = (t2 < 0x4e ? t2+0x31 : t2+0x43);
  134:       return 2;
  135:     }
  136:   }
  137: 
  138:   return RET_ILUNI;
  139: }