lib/cp950ext.h

   1 /*
   2  * Copyright (C) 1999-2001 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  * CP950 extensions
  23  */
  24
  25 static const unsigned short cp950ext_2uni_pagef9[157-116] = {
  26   /* 0xf9 */
  27                                   0x7881, 0x92b9, 0x88cf, 0x58bb,
  28   0x6052, 0x7ca7, 0x5afa, 0x2554, 0x2566, 0x2557, 0x2560, 0x256c,
  29   0x2563, 0x255a, 0x2569, 0x255d, 0x2552, 0x2564, 0x2555, 0x255e,
  30   0x256a, 0x2561, 0x2558, 0x2567, 0x255b, 0x2553, 0x2565, 0x2556,
  31   0x255f, 0x256b, 0x2562, 0x2559, 0x2568, 0x255c, 0x2551, 0x2550,
  32   0x256d, 0x256e, 0x2570, 0x256f, 0x2593,
  33 };
  34
  35 static int
  36 cp950ext_mbtowc (conv_t conv, ucs4_t *pwc, const unsigned char *s, int n)
  37 {
  38   unsigned char c1 = s[0];
  39   if ((c1 == 0xf9)) {
  40     if (n >= 2) {
  41       unsigned char c2 = s[1];
  42       if ((c2 >= 0x40 && c2 < 0x7f) || (c2 >= 0xa1 && c2 < 0xff)) {
  43         unsigned int i = 157 * (c1 - 0xa1) + (c2 - (c2 >= 0xa1 ? 0x62 : 0x40));
  44         unsigned short wc = 0xfffd;
  45         {
  46           if (i >= 13932 && i < 13973)
  47             wc = cp950ext_2uni_pagef9[i-13932];
  48         }
  49         if (wc != 0xfffd) {
  50           *pwc = (ucs4_t) wc;
  51           return 2;
  52         }
  53       }
  54       return RET_ILSEQ;
  55     }
  56     return RET_TOOFEW(0);
  57   }
  58   return RET_ILSEQ;
  59 }
  60
  61 static const unsigned short cp950ext_2charset[41] = {
  62   0xf9f9, 0xf9f8, 0xf9e6, 0xf9ef, 0xf9dd, 0xf9e8, 0xf9f1, 0xf9df,
  63   0xf9ec, 0xf9f5, 0xf9e3, 0xf9ee, 0xf9f7, 0xf9e5, 0xf9e9, 0xf9f2,
  64   0xf9e0, 0xf9eb, 0xf9f4, 0xf9e2, 0xf9e7, 0xf9f0, 0xf9de, 0xf9ed,
  65   0xf9f6, 0xf9e4, 0xf9ea, 0xf9f3, 0xf9e1, 0xf9fa, 0xf9fb, 0xf9fd,
  66   0xf9fc, 0xf9fe, 0xf9d9, 0xf9dc, 0xf9da, 0xf9d6, 0xf9db, 0xf9d8,
  67   0xf9d7,
  68 };
  69
  70 static const Summary16 cp950ext_uni2indx_page25[10] = {
  71   /* 0x2500 */
  72   {    0, 0x0000 }, {    0, 0x0000 }, {    0, 0x0000 }, {    0, 0x0000 },
  73   {    0, 0x0000 }, {    0, 0xffff }, {   16, 0xffff }, {   32, 0x0001 },
  74   {   33, 0x0000 }, {   33, 0x0008 },
  75 };
  76 static const Summary16 cp950ext_uni2indx_page58[12] = {
  77   /* 0x5800 */
  78   {   34, 0x0000 }, {   34, 0x0000 }, {   34, 0x0000 }, {   34, 0x0000 },
  79   {   34, 0x0000 }, {   34, 0x0000 }, {   34, 0x0000 }, {   34, 0x0000 },
  80   {   34, 0x0000 }, {   34, 0x0000 }, {   34, 0x0000 }, {   34, 0x0800 },
  81 };
  82 static const Summary16 cp950ext_uni2indx_page5a[16] = {
  83   /* 0x5a00 */
  84   {   35, 0x0000 }, {   35, 0x0000 }, {   35, 0x0000 }, {   35, 0x0000 },
  85   {   35, 0x0000 }, {   35, 0x0000 }, {   35, 0x0000 }, {   35, 0x0000 },
  86   {   35, 0x0000 }, {   35, 0x0000 }, {   35, 0x0000 }, {   35, 0x0000 },
  87   {   35, 0x0000 }, {   35, 0x0000 }, {   35, 0x0000 }, {   35, 0x0400 },
  88 };
  89 static const Summary16 cp950ext_uni2indx_page60[6] = {
  90   /* 0x6000 */
  91   {   36, 0x0000 }, {   36, 0x0000 }, {   36, 0x0000 }, {   36, 0x0000 },
  92   {   36, 0x0000 }, {   36, 0x0004 },
  93 };
  94 static const Summary16 cp950ext_uni2indx_page78[9] = {
  95   /* 0x7800 */
  96   {   37, 0x0000 }, {   37, 0x0000 }, {   37, 0x0000 }, {   37, 0x0000 },
  97   {   37, 0x0000 }, {   37, 0x0000 }, {   37, 0x0000 }, {   37, 0x0000 },
  98   {   37, 0x0002 },
  99 };
 100 static const Summary16 cp950ext_uni2indx_page7c[11] = {
 101   /* 0x7c00 */
 102   {   38, 0x0000 }, {   38, 0x0000 }, {   38, 0x0000 }, {   38, 0x0000 },
 103   {   38, 0x0000 }, {   38, 0x0000 }, {   38, 0x0000 }, {   38, 0x0000 },
 104   {   38, 0x0000 }, {   38, 0x0000 }, {   38, 0x0080 },
 105 };
 106 static const Summary16 cp950ext_uni2indx_page88[13] = {
 107   /* 0x8800 */
 108   {   39, 0x0000 }, {   39, 0x0000 }, {   39, 0x0000 }, {   39, 0x0000 },
 109   {   39, 0x0000 }, {   39, 0x0000 }, {   39, 0x0000 }, {   39, 0x0000 },
 110   {   39, 0x0000 }, {   39, 0x0000 }, {   39, 0x0000 }, {   39, 0x0000 },
 111   {   39, 0x8000 },
 112 };
 113 static const Summary16 cp950ext_uni2indx_page92[12] = {
 114   /* 0x9200 */
 115   {   40, 0x0000 }, {   40, 0x0000 }, {   40, 0x0000 }, {   40, 0x0000 },
 116   {   40, 0x0000 }, {   40, 0x0000 }, {   40, 0x0000 }, {   40, 0x0000 },
 117   {   40, 0x0000 }, {   40, 0x0000 }, {   40, 0x0000 }, {   40, 0x0200 },
 118 };
 119
 120 static int
 121 cp950ext_wctomb (conv_t conv, unsigned char *r, ucs4_t wc, int n)
 122 {
 123   if (n >= 2) {
 124     const Summary16 *summary = NULL;
 125     if (wc >= 0x2500 && wc < 0x25a0)
 126       summary = &cp950ext_uni2indx_page25[(wc>>4)-0x250];
 127     else if (wc >= 0x5800 && wc < 0x58c0)
 128       summary = &cp950ext_uni2indx_page58[(wc>>4)-0x580];
 129     else if (wc >= 0x5a00 && wc < 0x5b00)
 130       summary = &cp950ext_uni2indx_page5a[(wc>>4)-0x5a0];
 131     else if (wc >= 0x6000 && wc < 0x6060)
 132       summary = &cp950ext_uni2indx_page60[(wc>>4)-0x600];
 133     else if (wc >= 0x7800 && wc < 0x7890)
 134       summary = &cp950ext_uni2indx_page78[(wc>>4)-0x780];
 135     else if (wc >= 0x7c00 && wc < 0x7cb0)
 136       summary = &cp950ext_uni2indx_page7c[(wc>>4)-0x7c0];
 137     else if (wc >= 0x8800 && wc < 0x88d0)
 138       summary = &cp950ext_uni2indx_page88[(wc>>4)-0x880];
 139     else if (wc >= 0x9200 && wc < 0x92c0)
 140       summary = &cp950ext_uni2indx_page92[(wc>>4)-0x920];
 141     if (summary) {
 142       unsigned short used = summary->used;
 143       unsigned int i = wc & 0x0f;
 144       if (used & ((unsigned short) 1 << i)) {
 145         unsigned short c;
 146         /* Keep in `used' only the bits 0..i-1. */
 147         used &= ((unsigned short) 1 << i) - 1;
 148         /* Add `summary->indx' and the number of bits set in `used'. */
 149         used = (used & 0x5555) + ((used & 0xaaaa) >> 1);
 150         used = (used & 0x3333) + ((used & 0xcccc) >> 2);
 151         used = (used & 0x0f0f) + ((used & 0xf0f0) >> 4);
 152         used = (used & 0x00ff) + (used >> 8);
 153         c = cp950ext_2charset[summary->indx + used];
 154         r[0] = (c >> 8); r[1] = (c & 0xff);
 155         return 2;
 156       }
 157     }
 158     return RET_ILUNI;
 159   }
 160   return RET_TOOSMALL;
 161 }