lib/cp950ext.h

   1 /*
   2  * Copyright (C) 1999-2001, 2012, 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  * CP950 extensions
  22  */
  23
  24 static const unsigned short cp950ext_2uni_pagef9[157-116] = {
  25   /* 0xf9 */
  26                                   0x7881, 0x92b9, 0x88cf, 0x58bb,
  27   0x6052, 0x7ca7, 0x5afa, 0x2554, 0x2566, 0x2557, 0x2560, 0x256c,
  28   0x2563, 0x255a, 0x2569, 0x255d, 0x2552, 0x2564, 0x2555, 0x255e,
  29   0x256a, 0x2561, 0x2558, 0x2567, 0x255b, 0x2553, 0x2565, 0x2556,
  30   0x255f, 0x256b, 0x2562, 0x2559, 0x2568, 0x255c, 0x2551, 0x2550,
  31   0x256d, 0x256e, 0x2570, 0x256f, 0x2593,
  32 };
  33
  34 static int
  35 cp950ext_mbtowc (conv_t conv, ucs4_t *pwc, const unsigned char *s, size_t n)
  36 {
  37   unsigned char c1 = s[0];
  38   if ((c1 == 0xf9)) {
  39     if (n >= 2) {
  40       unsigned char c2 = s[1];
  41       if ((c2 >= 0x40 && c2 < 0x7f) || (c2 >= 0xa1 && c2 < 0xff)) {
  42         unsigned int i = 157 * (c1 - 0xa1) + (c2 - (c2 >= 0xa1 ? 0x62 : 0x40));
  43         unsigned short wc = 0xfffd;
  44         {
  45           if (i >= 13932 && i < 13973)
  46             wc = cp950ext_2uni_pagef9[i-13932];
  47         }
  48         if (wc != 0xfffd) {
  49           *pwc = (ucs4_t) wc;
  50           return 2;
  51         }
  52       }
  53       return RET_ILSEQ;
  54     }
  55     return RET_TOOFEW(0);
  56   }
  57   return RET_ILSEQ;
  58 }
  59
  60 static const unsigned short cp950ext_2charset[41] = {
  61   0xf9f9, 0xf9f8, 0xf9e6, 0xf9ef, 0xf9dd, 0xf9e8, 0xf9f1, 0xf9df,
  62   0xf9ec, 0xf9f5, 0xf9e3, 0xf9ee, 0xf9f7, 0xf9e5, 0xf9e9, 0xf9f2,
  63   0xf9e0, 0xf9eb, 0xf9f4, 0xf9e2, 0xf9e7, 0xf9f0, 0xf9de, 0xf9ed,
  64   0xf9f6, 0xf9e4, 0xf9ea, 0xf9f3, 0xf9e1, 0xf9fa, 0xf9fb, 0xf9fd,
  65   0xf9fc, 0xf9fe, 0xf9d9, 0xf9dc, 0xf9da, 0xf9d6, 0xf9db, 0xf9d8,
  66   0xf9d7,
  67 };
  68
  69 static const Summary16 cp950ext_uni2indx_page25[10] = {
  70   /* 0x2500 */
  71   {    0, 0x0000 }, {    0, 0x0000 }, {    0, 0x0000 }, {    0, 0x0000 },
  72   {    0, 0x0000 }, {    0, 0xffff }, {   16, 0xffff }, {   32, 0x0001 },
  73   {   33, 0x0000 }, {   33, 0x0008 },
  74 };
  75 static const Summary16 cp950ext_uni2indx_page58[12] = {
  76   /* 0x5800 */
  77   {   34, 0x0000 }, {   34, 0x0000 }, {   34, 0x0000 }, {   34, 0x0000 },
  78   {   34, 0x0000 }, {   34, 0x0000 }, {   34, 0x0000 }, {   34, 0x0000 },
  79   {   34, 0x0000 }, {   34, 0x0000 }, {   34, 0x0000 }, {   34, 0x0800 },
  80 };
  81 static const Summary16 cp950ext_uni2indx_page5a[16] = {
  82   /* 0x5a00 */
  83   {   35, 0x0000 }, {   35, 0x0000 }, {   35, 0x0000 }, {   35, 0x0000 },
  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, 0x0400 },
  87 };
  88 static const Summary16 cp950ext_uni2indx_page60[6] = {
  89   /* 0x6000 */
  90   {   36, 0x0000 }, {   36, 0x0000 }, {   36, 0x0000 }, {   36, 0x0000 },
  91   {   36, 0x0000 }, {   36, 0x0004 },
  92 };
  93 static const Summary16 cp950ext_uni2indx_page78[9] = {
  94   /* 0x7800 */
  95   {   37, 0x0000 }, {   37, 0x0000 }, {   37, 0x0000 }, {   37, 0x0000 },
  96   {   37, 0x0000 }, {   37, 0x0000 }, {   37, 0x0000 }, {   37, 0x0000 },
  97   {   37, 0x0002 },
  98 };
  99 static const Summary16 cp950ext_uni2indx_page7c[11] = {
 100   /* 0x7c00 */
 101   {   38, 0x0000 }, {   38, 0x0000 }, {   38, 0x0000 }, {   38, 0x0000 },
 102   {   38, 0x0000 }, {   38, 0x0000 }, {   38, 0x0000 }, {   38, 0x0000 },
 103   {   38, 0x0000 }, {   38, 0x0000 }, {   38, 0x0080 },
 104 };
 105 static const Summary16 cp950ext_uni2indx_page88[13] = {
 106   /* 0x8800 */
 107   {   39, 0x0000 }, {   39, 0x0000 }, {   39, 0x0000 }, {   39, 0x0000 },
 108   {   39, 0x0000 }, {   39, 0x0000 }, {   39, 0x0000 }, {   39, 0x0000 },
 109   {   39, 0x0000 }, {   39, 0x0000 }, {   39, 0x0000 }, {   39, 0x0000 },
 110   {   39, 0x8000 },
 111 };
 112 static const Summary16 cp950ext_uni2indx_page92[12] = {
 113   /* 0x9200 */
 114   {   40, 0x0000 }, {   40, 0x0000 }, {   40, 0x0000 }, {   40, 0x0000 },
 115   {   40, 0x0000 }, {   40, 0x0000 }, {   40, 0x0000 }, {   40, 0x0000 },
 116   {   40, 0x0000 }, {   40, 0x0000 }, {   40, 0x0000 }, {   40, 0x0200 },
 117 };
 118
 119 static int
 120 cp950ext_wctomb (conv_t conv, unsigned char *r, ucs4_t wc, size_t n)
 121 {
 122   if (n >= 2) {
 123     const Summary16 *summary = NULL;
 124     if (wc >= 0x2500 && wc < 0x25a0)
 125       summary = &cp950ext_uni2indx_page25[(wc>>4)-0x250];
 126     else if (wc >= 0x5800 && wc < 0x58c0)
 127       summary = &cp950ext_uni2indx_page58[(wc>>4)-0x580];
 128     else if (wc >= 0x5a00 && wc < 0x5b00)
 129       summary = &cp950ext_uni2indx_page5a[(wc>>4)-0x5a0];
 130     else if (wc >= 0x6000 && wc < 0x6060)
 131       summary = &cp950ext_uni2indx_page60[(wc>>4)-0x600];
 132     else if (wc >= 0x7800 && wc < 0x7890)
 133       summary = &cp950ext_uni2indx_page78[(wc>>4)-0x780];
 134     else if (wc >= 0x7c00 && wc < 0x7cb0)
 135       summary = &cp950ext_uni2indx_page7c[(wc>>4)-0x7c0];
 136     else if (wc >= 0x8800 && wc < 0x88d0)
 137       summary = &cp950ext_uni2indx_page88[(wc>>4)-0x880];
 138     else if (wc >= 0x9200 && wc < 0x92c0)
 139       summary = &cp950ext_uni2indx_page92[(wc>>4)-0x920];
 140     if (summary) {
 141       unsigned short used = summary->used;
 142       unsigned int i = wc & 0x0f;
 143       if (used & ((unsigned short) 1 << i)) {
 144         unsigned short c;
 145         /* Keep in 'used' only the bits 0..i-1. */
 146         used &= ((unsigned short) 1 << i) - 1;
 147         /* Add 'summary->indx' and the number of bits set in 'used'. */
 148         used = (used & 0x5555) + ((used & 0xaaaa) >> 1);
 149         used = (used & 0x3333) + ((used & 0xcccc) >> 2);
 150         used = (used & 0x0f0f) + ((used & 0xf0f0) >> 4);
 151         used = (used & 0x00ff) + (used >> 8);
 152         c = cp950ext_2charset[summary->indx + used];
 153         r[0] = (c >> 8); r[1] = (c & 0xff);
 154         return 2;
 155       }
 156     }
 157     return RET_ILUNI;
 158   }
 159   return RET_TOOSMALL;
 160 }