lib/utf8.h

   1 /*
   2  * Copyright (C) 1999-2000 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  * UTF-8
  23  */
  24
  25 /* Specification: RFC 2279 */
  26
  27 static int
  28 utf8_mbtowc (conv_t conv, ucs4_t *pwc, const unsigned char *s, int n)
  29 {
  30   unsigned char c = s[0];
  31
  32   if (c < 0x80) {
  33     *pwc = c;
  34     return 1;
  35   } else if (c < 0xc2) {
  36     return RET_ILSEQ;
  37   } else if (c < 0xe0) {
  38     if (n < 2)
  39       return RET_TOOFEW(0);
  40     if (!((s[1] ^ 0x80) < 0x40))
  41       return RET_ILSEQ;
  42     *pwc = ((ucs4_t) (c & 0x1f) << 6)
  43            | (ucs4_t) (s[1] ^ 0x80);
  44     return 2;
  45   } else if (c < 0xf0) {
  46     if (n < 3)
  47       return RET_TOOFEW(0);
  48     if (!((s[1] ^ 0x80) < 0x40 && (s[2] ^ 0x80) < 0x40
  49           && (c >= 0xe1 || s[1] >= 0xa0)))
  50       return RET_ILSEQ;
  51     *pwc = ((ucs4_t) (c & 0x0f) << 12)
  52            | ((ucs4_t) (s[1] ^ 0x80) << 6)
  53            | (ucs4_t) (s[2] ^ 0x80);
  54     return 3;
  55   } else if (c < 0xf8 && sizeof(ucs4_t)*8 >= 32) {
  56     if (n < 4)
  57       return RET_TOOFEW(0);
  58     if (!((s[1] ^ 0x80) < 0x40 && (s[2] ^ 0x80) < 0x40
  59           && (s[3] ^ 0x80) < 0x40
  60           && (c >= 0xf1 || s[1] >= 0x90)))
  61       return RET_ILSEQ;
  62     *pwc = ((ucs4_t) (c & 0x07) << 18)
  63            | ((ucs4_t) (s[1] ^ 0x80) << 12)
  64            | ((ucs4_t) (s[2] ^ 0x80) << 6)
  65            | (ucs4_t) (s[3] ^ 0x80);
  66     return 4;
  67   } else if (c < 0xfc && sizeof(ucs4_t)*8 >= 32) {
  68     if (n < 5)
  69       return RET_TOOFEW(0);
  70     if (!((s[1] ^ 0x80) < 0x40 && (s[2] ^ 0x80) < 0x40
  71           && (s[3] ^ 0x80) < 0x40 && (s[4] ^ 0x80) < 0x40
  72           && (c >= 0xf9 || s[1] >= 0x88)))
  73       return RET_ILSEQ;
  74     *pwc = ((ucs4_t) (c & 0x03) << 24)
  75            | ((ucs4_t) (s[1] ^ 0x80) << 18)
  76            | ((ucs4_t) (s[2] ^ 0x80) << 12)
  77            | ((ucs4_t) (s[3] ^ 0x80) << 6)
  78            | (ucs4_t) (s[4] ^ 0x80);
  79     return 5;
  80   } else if (c < 0xfe && sizeof(ucs4_t)*8 >= 32) {
  81     if (n < 6)
  82       return RET_TOOFEW(0);
  83     if (!((s[1] ^ 0x80) < 0x40 && (s[2] ^ 0x80) < 0x40
  84           && (s[3] ^ 0x80) < 0x40 && (s[4] ^ 0x80) < 0x40
  85           && (s[5] ^ 0x80) < 0x40
  86           && (c >= 0xfd || s[1] >= 0x84)))
  87       return RET_ILSEQ;
  88     *pwc = ((ucs4_t) (c & 0x01) << 30)
  89            | ((ucs4_t) (s[1] ^ 0x80) << 24)
  90            | ((ucs4_t) (s[2] ^ 0x80) << 18)
  91            | ((ucs4_t) (s[3] ^ 0x80) << 12)
  92            | ((ucs4_t) (s[4] ^ 0x80) << 6)
  93            | (ucs4_t) (s[5] ^ 0x80);
  94     return 6;
  95   } else
  96     return RET_ILSEQ;
  97 }
  98
  99 static int
 100 utf8_wctomb (conv_t conv, unsigned char *r, ucs4_t wc, int n) /* n == 0 is acceptable */
 101 {
 102   int count;
 103   if (wc < 0x80)
 104     count = 1;
 105   else if (wc < 0x800)
 106     count = 2;
 107   else if (wc < 0x10000)
 108     count = 3;
 109   else if (wc < 0x200000)
 110     count = 4;
 111   else if (wc < 0x4000000)
 112     count = 5;
 113   else if (wc <= 0x7fffffff)
 114     count = 6;
 115   else
 116     return RET_ILSEQ;
 117   if (n < count)
 118     return RET_TOOSMALL;
 119   switch (count) { /* note: code falls through cases! */
 120     case 6: r[5] = 0x80 | (wc & 0x3f); wc = wc >> 6; wc |= 0x4000000;
 121     case 5: r[4] = 0x80 | (wc & 0x3f); wc = wc >> 6; wc |= 0x200000;
 122     case 4: r[3] = 0x80 | (wc & 0x3f); wc = wc >> 6; wc |= 0x10000;
 123     case 3: r[2] = 0x80 | (wc & 0x3f); wc = wc >> 6; wc |= 0x800;
 124     case 2: r[1] = 0x80 | (wc & 0x3f); wc = wc >> 6; wc |= 0xc0;
 125     case 1: r[0] = wc;
 126   }
 127   return count;
 128 }