lib/unistr/u8-mbtouc-unsafe.c

   1 /* Look at first character in UTF-8 string.
   2    Copyright (C) 1999-2002, 2006-2007, 2009-2010 Free Software Foundation, Inc.
   3    Written by Bruno Haible <bruno@clisp.org>, 2001.
   4
   5    This program is free software: you can redistribute it and/or modify it
   6    under the terms of the GNU Lesser General Public License as published
   7    by the Free Software Foundation; either version 3 of the License, or
   8    (at your option) any later version.
   9
  10    This program is distributed in the hope that it will be useful,
  11    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 License
  16    along with this program.  If not, see <http://www.gnu.org/licenses/>.  */
  17
  18 #include <config.h>
  19
  20 #if defined IN_LIBUNISTRING
  21 /* Tell unistr.h to declare u8_mbtouc_unsafe as 'extern', not
  22    'static inline'.  */
  23 # include "unistring-notinline.h"
  24 #endif
  25
  26 /* Specification.  */
  27 #include "unistr.h"
  28
  29 #if !HAVE_INLINE
  30
  31 int
  32 u8_mbtouc_unsafe (ucs4_t *puc, const uint8_t *s, size_t n)
  33 {
  34   uint8_t c = *s;
  35
  36   if (c < 0x80)
  37     {
  38       *puc = c;
  39       return 1;
  40     }
  41   else if (c >= 0xc2)
  42     {
  43       if (c < 0xe0)
  44         {
  45           if (n >= 2)
  46             {
  47 #if CONFIG_UNICODE_SAFETY
  48               if ((s[1] ^ 0x80) < 0x40)
  49 #endif
  50                 {
  51                   *puc = ((unsigned int) (c & 0x1f) << 6)
  52                          | (unsigned int) (s[1] ^ 0x80);
  53                   return 2;
  54                 }
  55               /* invalid multibyte character */
  56             }
  57           else
  58             {
  59               /* incomplete multibyte character */
  60               *puc = 0xfffd;
  61               return n;
  62             }
  63         }
  64       else if (c < 0xf0)
  65         {
  66           if (n >= 3)
  67             {
  68 #if CONFIG_UNICODE_SAFETY
  69               if ((s[1] ^ 0x80) < 0x40 && (s[2] ^ 0x80) < 0x40
  70                   && (c >= 0xe1 || s[1] >= 0xa0)
  71                   && (c != 0xed || s[1] < 0xa0))
  72 #endif
  73                 {
  74                   *puc = ((unsigned int) (c & 0x0f) << 12)
  75                          | ((unsigned int) (s[1] ^ 0x80) << 6)
  76                          | (unsigned int) (s[2] ^ 0x80);
  77                   return 3;
  78                 }
  79               /* invalid multibyte character */
  80             }
  81           else
  82             {
  83               /* incomplete multibyte character */
  84               *puc = 0xfffd;
  85               return n;
  86             }
  87         }
  88       else if (c < 0xf8)
  89         {
  90           if (n >= 4)
  91             {
  92 #if CONFIG_UNICODE_SAFETY
  93               if ((s[1] ^ 0x80) < 0x40 && (s[2] ^ 0x80) < 0x40
  94                   && (s[3] ^ 0x80) < 0x40
  95                   && (c >= 0xf1 || s[1] >= 0x90)
  96 #if 1
  97                   && (c < 0xf4 || (c == 0xf4 && s[1] < 0x90))
  98 #endif
  99                  )
 100 #endif
 101                 {
 102                   *puc = ((unsigned int) (c & 0x07) << 18)
 103                          | ((unsigned int) (s[1] ^ 0x80) << 12)
 104                          | ((unsigned int) (s[2] ^ 0x80) << 6)
 105                          | (unsigned int) (s[3] ^ 0x80);
 106                   return 4;
 107                 }
 108               /* invalid multibyte character */
 109             }
 110           else
 111             {
 112               /* incomplete multibyte character */
 113               *puc = 0xfffd;
 114               return n;
 115             }
 116         }
 117 #if 0
 118       else if (c < 0xfc)
 119         {
 120           if (n >= 5)
 121             {
 122 #if CONFIG_UNICODE_SAFETY
 123               if ((s[1] ^ 0x80) < 0x40 && (s[2] ^ 0x80) < 0x40
 124                   && (s[3] ^ 0x80) < 0x40 && (s[4] ^ 0x80) < 0x40
 125                   && (c >= 0xf9 || s[1] >= 0x88))
 126 #endif
 127                 {
 128                   *puc = ((unsigned int) (c & 0x03) << 24)
 129                          | ((unsigned int) (s[1] ^ 0x80) << 18)
 130                          | ((unsigned int) (s[2] ^ 0x80) << 12)
 131                          | ((unsigned int) (s[3] ^ 0x80) << 6)
 132                          | (unsigned int) (s[4] ^ 0x80);
 133                   return 5;
 134                 }
 135               /* invalid multibyte character */
 136             }
 137           else
 138             {
 139               /* incomplete multibyte character */
 140               *puc = 0xfffd;
 141               return n;
 142             }
 143         }
 144       else if (c < 0xfe)
 145         {
 146           if (n >= 6)
 147             {
 148 #if CONFIG_UNICODE_SAFETY
 149               if ((s[1] ^ 0x80) < 0x40 && (s[2] ^ 0x80) < 0x40
 150                   && (s[3] ^ 0x80) < 0x40 && (s[4] ^ 0x80) < 0x40
 151                   && (s[5] ^ 0x80) < 0x40
 152                   && (c >= 0xfd || s[1] >= 0x84))
 153 #endif
 154                 {
 155                   *puc = ((unsigned int) (c & 0x01) << 30)
 156                          | ((unsigned int) (s[1] ^ 0x80) << 24)
 157                          | ((unsigned int) (s[2] ^ 0x80) << 18)
 158                          | ((unsigned int) (s[3] ^ 0x80) << 12)
 159                          | ((unsigned int) (s[4] ^ 0x80) << 6)
 160                          | (unsigned int) (s[5] ^ 0x80);
 161                   return 6;
 162                 }
 163               /* invalid multibyte character */
 164             }
 165           else
 166             {
 167               /* incomplete multibyte character */
 168               *puc = 0xfffd;
 169               return n;
 170             }
 171         }
 172 #endif
 173     }
 174   /* invalid multibyte character */
 175   *puc = 0xfffd;
 176   return 1;
 177 }
 178
 179 #endif