mdw@git.distorted.org.uk Git - disorder/blob - lib/charset.c

   1 /*
   2  * This file is part of DisOrder.
   3  * Copyright (C) 2004, 2005, 2007, 2008 Richard Kettlewell
   4  *
   5  * This program is free software: you can redistribute it and/or modify
   6  * it under the terms of the GNU General Public License as published by
   7  * 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
  13  * GNU General Public License for more details.
  14  *
  15  * You should have received a copy of the GNU General Public License
  16  * along with this program.  If not, see <http://www.gnu.org/licenses/>.
  17  */
  18 /** @file lib/charset.c @brief Character set conversion */
  19
  20 #include "common.h"
  21
  22 #include <iconv.h>
  23 #include <errno.h>
  24 #include <langinfo.h>
  25
  26 #include "mem.h"
  27 #include "log.h"
  28 #include "charset.h"
  29 #include "configuration.h"
  30 #include "vector.h"
  31 #include "unicode.h"
  32
  33 /** @brief Low-level converstion routine
  34  * @param from Source encoding
  35  * @param to Destination encoding
  36  * @param ptr First byte to convert
  37  * @param n Number of bytes to convert
  38  * @return Converted text, 0-terminated; or NULL on error.
  39  */
  40 static void *convert(const char *from, const char *to,
  41                      const void *ptr, size_t n) {
  42   iconv_t i;
  43   size_t len;
  44   char *buf = 0, *s, *d;
  45   size_t bufsize = 0, sl, dl;
  46
  47   if((i = iconv_open(to, from)) == (iconv_t)-1)
  48     fatal(errno, "error calling iconv_open");
  49   do {
  50     bufsize = bufsize ? 2 * bufsize : 32;
  51     buf = xrealloc_noptr(buf, bufsize);
  52     iconv(i, 0, 0, 0, 0);
  53     s = (char *)ptr;
  54     sl = n;
  55     d = buf;
  56     dl = bufsize;
  57     /* (void *) to work around FreeBSD's nonstandard iconv prototype */
  58     len = iconv(i, (void *)&s, &sl, &d, &dl);
  59   } while(len == (size_t)-1 && errno == E2BIG);
  60   iconv_close(i);
  61   if(len == (size_t)-1) {
  62     error(errno, "error converting from %s to %s", from, to);
  63     return 0;
  64   }
  65   return buf;
  66 }
  67
  68 /** @brief Convert from the local multibyte encoding to UTF-8 */
  69 char *mb2utf8(const char *mb) {
  70   return convert(nl_langinfo(CODESET), "UTF-8", mb, strlen(mb) + 1);
  71 }
  72
  73 /** @brief Convert from UTF-8 to the local multibyte encoding */
  74 char *utf82mb(const char *utf8) {
  75   return convert("UTF-8", nl_langinfo(CODESET), utf8, strlen(utf8) + 1);
  76 }
  77
  78 /** @brief Convert from encoding @p from to UTF-8 */
  79 char *any2utf8(const char *from, const char *any) {
  80   return convert(from, "UTF-8", any, strlen(any) + 1);
  81 }
  82
  83 /** @brief Convert from encoding @p from to the local multibyte encoding */
  84 char *any2mb(const char *from, const char *any) {
  85   if(from) return convert(from, nl_langinfo(CODESET), any, strlen(any) + 1);
  86   else return xstrdup(any);
  87 }
  88
  89 /** @brief Convert from encoding @p from to encoding @p to */
  90 char *any2any(const char *from,
  91               const char *to,
  92               const char *any) {
  93   if(from || to) return convert(from, to, any, strlen(any) + 1);
  94   else return xstrdup(any);
  95 }
  96
  97 /** @brief Truncate a string for display purposes
  98  * @param s Pointer to UTF-8 string
  99  * @param max Maximum number of columns
 100  * @return @p or truncated string (never NULL)
 101  *
 102  * Returns a string that is no longer than @p max graphemes long and is either
 103  * (canonically) equal to @p s or is a truncated form of it with an ellipsis
 104  * appended.
 105  *
 106  * We don't take display width into account (tricky for HTML!) and we don't
 107  * attempt to implement the Bidi algorithm.  If you have track names for which
 108  * either of these matter in practice then get in touch.
 109  */
 110 const char *truncate_for_display(const char *s, long max) {
 111   uint32_t *s32;
 112   size_t l32, cut;
 113   utf32_iterator it;
 114   long graphemes;
 115
 116   /* Convert to UTF-32 for processing */
 117   if(!(s32 = utf8_to_utf32(s, strlen(s), &l32)))
 118     return 0;
 119   it = utf32_iterator_new(s32, l32);
 120   cut = l32;
 121   graphemes = 0;                        /* # of graphemes left of it */
 122   while(graphemes <= max && utf32_iterator_where(it) < l32) {
 123     if(graphemes == max - 1)
 124       cut = utf32_iterator_where(it);
 125     utf32_iterator_advance(it, 1);
 126     if(utf32_iterator_grapheme_boundary(it))
 127       ++graphemes;
 128   }
 129   if(graphemes > max) {                 /* we need to cut */
 130     s32[cut] = 0x2026;                  /* HORIZONTAL ELLIPSIS */
 131     l32 = cut + 1;
 132     s = utf32_to_utf8(s32, l32, 0);
 133   }
 134   xfree(s32);
 135   return s;
 136 }
 137
 138 /*
 139 Local Variables:
 140 c-basic-offset:2
 141 comment-column:40
 142 fill-column:79
 143 indent-tabs-mode:nil
 144 End:
 145 */