lib/unicodeio.c

   1 /* Unicode character output to streams with locale dependent encoding.
   2
   3    Copyright (C) 2000, 2001 Free Software Foundation, Inc.
   4
   5    This program is free software; you can redistribute it and/or modify it
   6    under the terms of the GNU Library General Public License as published
   7    by the Free Software Foundation; either version 2, or (at your option)
   8    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    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 this program; if not, write to the Free Software
  17    Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307,
  18    USA.  */
  19
  20 /* Written by Bruno Haible <haible@clisp.cons.org>.  */
  21
  22 #ifdef HAVE_CONFIG_H
  23 # include <config.h>
  24 #endif
  25
  26 #if HAVE_STDDEF_H
  27 # include <stddef.h>
  28 #endif
  29
  30 #include <stdio.h>
  31 #if HAVE_STRING_H
  32 # include <string.h>
  33 #else
  34 # include <strings.h>
  35 #endif
  36
  37 #include <errno.h>
  38 #ifndef errno
  39 extern int errno;
  40 #endif
  41
  42 #if HAVE_ICONV
  43 # include <iconv.h>
  44 #endif
  45
  46 #include <error.h>
  47
  48 #if ENABLE_NLS
  49 # include <libintl.h>
  50 # define _(Text) gettext (Text)
  51 #else
  52 # define _(Text) Text
  53 #endif
  54
  55 #include "unicodeio.h"
  56
  57 /* When we pass a Unicode character to iconv(), we must pass it in a
  58    suitable encoding. The standardized Unicode encodings are
  59    UTF-8, UCS-2, UCS-4, UTF-16, UTF-16BE, UTF-16LE, UTF-7.
  60    UCS-2 supports only characters up to \U0000FFFF.
  61    UTF-16 and variants support only characters up to \U0010FFFF.
  62    UTF-7 is way too complex and not supported by glibc-2.1.
  63    UCS-4 specification leaves doubts about endianness and byte order
  64    mark. glibc currently interprets it as big endian without byte order
  65    mark, but this is not backed by an RFC.
  66    So we use UTF-8. It supports characters up to \U7FFFFFFF and is
  67    unambiguously defined.  */
  68
  69 /* Stores the UTF-8 representation of the Unicode character wc in r[0..5].
  70    Returns the number of bytes stored, or -1 if wc is out of range.  */
  71 static int
  72 utf8_wctomb (unsigned char *r, unsigned int wc)
  73 {
  74   int count;
  75
  76   if (wc < 0x80)
  77     count = 1;
  78   else if (wc < 0x800)
  79     count = 2;
  80   else if (wc < 0x10000)
  81     count = 3;
  82   else if (wc < 0x200000)
  83     count = 4;
  84   else if (wc < 0x4000000)
  85     count = 5;
  86   else if (wc <= 0x7fffffff)
  87     count = 6;
  88   else
  89     return -1;
  90
  91   switch (count)
  92     {
  93       /* Note: code falls through cases! */
  94       case 6: r[5] = 0x80 | (wc & 0x3f); wc = wc >> 6; wc |= 0x4000000;
  95       case 5: r[4] = 0x80 | (wc & 0x3f); wc = wc >> 6; wc |= 0x200000;
  96       case 4: r[3] = 0x80 | (wc & 0x3f); wc = wc >> 6; wc |= 0x10000;
  97       case 3: r[2] = 0x80 | (wc & 0x3f); wc = wc >> 6; wc |= 0x800;
  98       case 2: r[1] = 0x80 | (wc & 0x3f); wc = wc >> 6; wc |= 0xc0;
  99       case 1: r[0] = wc;
 100     }
 101
 102   return count;
 103 }
 104
 105 /* Luckily, the encoding's name is platform independent.  */
 106 #define UTF8_NAME "UTF-8"
 107
 108 /* Outputs the Unicode character CODE to the output stream STREAM.
 109    Assumes that the locale doesn't change between two calls.  */
 110 void
 111 print_unicode_char (FILE *stream, unsigned int code)
 112 {
 113   static int initialized;
 114   static int is_utf8;
 115 #if HAVE_ICONV
 116   static iconv_t utf8_to_local;
 117 #endif
 118
 119   char inbuf[6];
 120   int count;
 121
 122   if (!initialized)
 123     {
 124       extern const char *locale_charset PARAMS ((void));
 125       const char *charset = locale_charset ();
 126
 127       is_utf8 = !strcmp (charset, UTF8_NAME);
 128 #if HAVE_ICONV
 129       if (!is_utf8)
 130         {
 131           utf8_to_local = iconv_open (charset, UTF8_NAME);
 132           if (utf8_to_local == (iconv_t)(-1))
 133             {
 134               /* For an unknown encoding, assume ASCII.  */
 135               utf8_to_local = iconv_open ("ASCII", UTF8_NAME);
 136               if (utf8_to_local == (iconv_t)(-1))
 137                 error (1, 0,
 138                        _("cannot output U+%04X: iconv function not usable"),
 139                        code);
 140             }
 141         }
 142 #endif
 143       initialized = 1;
 144     }
 145
 146   /* Convert the character to UTF-8.  */
 147   count = utf8_wctomb ((unsigned char *) inbuf, code);
 148   if (count < 0)
 149     error (1, 0, _("U+%04X: character out of range"), code);
 150
 151   if (is_utf8)
 152     {
 153       fwrite (inbuf, 1, count, stream);
 154     }
 155   else
 156     {
 157 #if HAVE_ICONV
 158       char outbuf[25];
 159       const char *inptr;
 160       size_t inbytesleft;
 161       char *outptr;
 162       size_t outbytesleft;
 163       size_t res;
 164
 165       inptr = inbuf;
 166       inbytesleft = count;
 167       outptr = outbuf;
 168       outbytesleft = sizeof (outbuf);
 169
 170       /* Convert the character from UTF-8 to the locale's charset.  */
 171       res = iconv (utf8_to_local,
 172                    (ICONV_CONST char **)&inptr, &inbytesleft,
 173                    &outptr, &outbytesleft);
 174       if (inbytesleft > 0 || res == (size_t)(-1)
 175           /* Irix iconv() inserts a NUL byte if it cannot convert. */
 176 # if !defined _LIBICONV_VERSION && (defined sgi || defined __sgi)
 177           || (res > 0 && code != 0 && outptr - outbuf == 1 && *outbuf == '\0')
 178 # endif
 179          )
 180         error (1, res == (size_t)(-1) ? errno : 0,
 181                _("cannot convert U+%04X to local character set"), code);
 182
 183       /* Avoid glibc-2.1 bug and Solaris 2.7 bug.  */
 184 # if defined _LIBICONV_VERSION \
 185     || !((__GLIBC__ - 0 == 2 && __GLIBC_MINOR__ - 0 <= 1) || defined __sun)
 186
 187       /* Get back to the initial shift state.  */
 188       res = iconv (utf8_to_local, NULL, NULL, &outptr, &outbytesleft);
 189       if (res == (size_t)(-1))
 190         error (1, errno, _("cannot convert U+%04X to local character set"),
 191                code);
 192 # endif
 193
 194       fwrite (outbuf, 1, outptr - outbuf, stream);
 195 #else
 196       error (1, 0, _("cannot output U+%04X: iconv function not available"),
 197              code);
 198 #endif
 199     }
 200 }