Return to gbk.h CVS log

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

libiconv

    1: /*
    2:  * Copyright (C) 1999-2001, 2005, 2008 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:  * GBK
   23:  */
   24: 
   25: /*
   26:  * GBK, as described in Ken Lunde's book, is an extension of GB 2312-1980
   27:  * (shifted by adding 0x8080 to the range 0xA1A1..0xFEFE, as used in EUC-CN).
   28:  * It adds the following ranges:
   29:  *
   30:  * (part of GBK/1)  0xA2A1-0xA2AA  Small Roman numerals
   31:  * GBK/3   0x{81-A0}{40-7E,80-FE}  6080 new characters, all in Unicode
   32:  * GBK/4   0x{AA-FE}{40-7E,80-A0}  8160 new characters, 8080 in Unicode
   33:  * GBK/5   0x{A8-A9}{40-7E,80-A0}  166 new characters, 153 in Unicode
   34:  *
   35:  * Furthermore, all four tables I have looked at
   36:  *   - the CP936 table by Microsoft, found on ftp.unicode.org in 1999,
   37:  *   - the GBK table by Sun, investigated on a Solaris 2.7 machine,
   38:  *   - the GBK tables by CWEX, found in the Big5+ package,
   39:  *   - the GB18030 standard (second printing),
   40:  * agree in the following extensions. (Ken Lunde must have overlooked these
   41:  * differences between GB2312 and GBK. Also, the CWEX tables have additional
   42:  * differences.)
   43:  *
   44:  * 1. Some characters in the GB2312 range are defined differently:
   45:  *
   46:  *     code    GB2312                         GBK
   47:  *    0xA1A4   0x30FB # KATAKANA MIDDLE DOT   0x00B7 # MIDDLE DOT
   48:  *    0xA1AA   0x2015 # HORIZONTAL BAR        0x2014 # EM DASH
   49:  *
   50:  * 2. 19 characters added in the range 0xA6E0-0xA6F5.
   51:  *
   52:  * 3. 4 characters added in the range 0xA8BB-0xA8C0.
   53:  *
   54:  * CP936 as of 1999 was identical to GBK. However, since 1999, Microsoft has
   55:  * added new mappings to CP936...
   56:  */
   57: 
   58: #include "gbkext1.h"
   59: #include "gbkext2.h"
   60: #include "gbkext_inv.h"
   61: #include "cp936ext.h"
   62: 
   63: static int
   64: gbk_mbtowc (conv_t conv, ucs4_t *pwc, const unsigned char *s, int n)
   65: {
   66:   unsigned char c = *s;
   67: 
   68:   if (c >= 0x81 && c < 0xff) {
   69:     if (n < 2)
   70:       return RET_TOOFEW(0);
   71:     if (c >= 0xa1 && c <= 0xf7) {
   72:       unsigned char c2 = s[1];
   73:       if (c == 0xa1) {
   74:         if (c2 == 0xa4) {
   75:           *pwc = 0x00b7;
   76:           return 2;
   77:         }
   78:         if (c2 == 0xaa) {
   79:           *pwc = 0x2014;
   80:           return 2;
   81:         }
   82:       }
   83:       if (c2 >= 0xa1 && c2 < 0xff) {
   84:         unsigned char buf[2];
   85:         int ret;
   86:         buf[0] = c-0x80; buf[1] = c2-0x80;
   87:         ret = gb2312_mbtowc(conv,pwc,buf,2);
   88:         if (ret != RET_ILSEQ)
   89:           return ret;
   90:         buf[0] = c; buf[1] = c2;
   91:         ret = cp936ext_mbtowc(conv,pwc,buf,2);
   92:         if (ret != RET_ILSEQ)
   93:           return ret;
   94:       }
   95:     }
   96:     if (c >= 0x81 && c <= 0xa0)
   97:       return gbkext1_mbtowc(conv,pwc,s,2);
   98:     if (c >= 0xa8 && c <= 0xfe)
   99:       return gbkext2_mbtowc(conv,pwc,s,2);
  100:     if (c == 0xa2) {
  101:       unsigned char c2 = s[1];
  102:       if (c2 >= 0xa1 && c2 <= 0xaa) {
  103:         *pwc = 0x2170+(c2-0xa1);
  104:         return 2;
  105:       }
  106:     }
  107:   }
  108:   return RET_ILSEQ;
  109: }
  110: 
  111: static int
  112: gbk_wctomb (conv_t conv, unsigned char *r, ucs4_t wc, int n)
  113: {
  114:   unsigned char buf[2];
  115:   int ret;
  116: 
  117:   if (wc != 0x30fb && wc != 0x2015) {
  118:     ret = gb2312_wctomb(conv,buf,wc,2);
  119:     if (ret != RET_ILUNI) {
  120:       if (ret != 2) abort();
  121:       if (n < 2)
  122:         return RET_TOOSMALL;
  123:       r[0] = buf[0]+0x80;
  124:       r[1] = buf[1]+0x80;
  125:       return 2;
  126:     }
  127:   }
  128:   ret = gbkext_inv_wctomb(conv,buf,wc,2);
  129:   if (ret != RET_ILUNI) {
  130:     if (ret != 2) abort();
  131:     if (n < 2)
  132:       return RET_TOOSMALL;
  133:     r[0] = buf[0];
  134:     r[1] = buf[1];
  135:     return 2;
  136:   }
  137:   if (wc >= 0x2170 && wc <= 0x2179) {
  138:     if (n < 2)
  139:       return RET_TOOSMALL;
  140:     r[0] = 0xa2;
  141:     r[1] = 0xa1 + (wc-0x2170);
  142:     return 2;
  143:   }
  144:   ret = cp936ext_wctomb(conv,buf,wc,2);
  145:   if (ret != RET_ILUNI) {
  146:     if (ret != 2) abort();
  147:     if (n < 2)
  148:       return RET_TOOSMALL;
  149:     r[0] = buf[0];
  150:     r[1] = buf[1];
  151:     return 2;
  152:   }
  153:   if (wc == 0x00b7) {
  154:     if (n < 2)
  155:       return RET_TOOSMALL;
  156:     r[0] = 0xa1;
  157:     r[1] = 0xa4;
  158:     return 2;
  159:   }
  160:   if (wc == 0x2014) {
  161:     if (n < 2)
  162:       return RET_TOOSMALL;
  163:     r[0] = 0xa1;
  164:     r[1] = 0xaa;
  165:     return 2;
  166:   }
  167: 
  168:   return RET_ILUNI;
  169: }