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 2 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, but
  11  * WITHOUT ANY WARRANTY; without even the implied warranty of
  12  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
  13  * 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, write to the Free Software
  17  * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307
  18  * USA
  19  */
  20 /** @file lib/charset.c @brief Character set conversion */
  21
  22 #include "common.h"
  23
  24 #include <iconv.h>
  25 #include <errno.h>
  26 #include <langinfo.h>
  27
  28 #include "mem.h"
  29 #include "log.h"
  30 #include "charset.h"
  31 #include "configuration.h"
  32 #include "vector.h"
  33 #include "unicode.h"
  34
  35 /** @brief Low-level converstion routine
  36  * @param from Source encoding
  37  * @param to Destination encoding
  38  * @param ptr First byte to convert
  39  * @param n Number of bytes to convert
  40  * @return Converted text, 0-terminated; or NULL on error.
  41  */
  42 static void *convert(const char *from, const char *to,
  43                      const void *ptr, size_t n) {
  44   iconv_t i;
  45   size_t len;
  46   char *buf = 0, *s, *d;
  47   size_t bufsize = 0, sl, dl;
  48
  49   if((i = iconv_open(to, from)) == (iconv_t)-1)
  50     fatal(errno, "error calling iconv_open");
  51   do {
  52     bufsize = bufsize ? 2 * bufsize : 32;
  53     buf = xrealloc_noptr(buf, bufsize);
  54     iconv(i, 0, 0, 0, 0);
  55     s = (char *)ptr;
  56     sl = n;
  57     d = buf;
  58     dl = bufsize;
  59     /* (void *) to work around FreeBSD's nonstandard iconv prototype */
  60     len = iconv(i, (void *)&s, &sl, &d, &dl);
  61   } while(len == (size_t)-1 && errno == E2BIG);
  62   iconv_close(i);
  63   if(len == (size_t)-1) {
  64     error(errno, "error converting from %s to %s", from, to);
  65     return 0;
  66   }
  67   return buf;
  68 }
  69
  70 /** @brief Convert from the local multibyte encoding to UTF-8 */
  71 char *mb2utf8(const char *mb) {
  72   return convert(nl_langinfo(CODESET), "UTF-8", mb, strlen(mb) + 1);
  73 }
  74
  75 /** @brief Convert from UTF-8 to the local multibyte encoding */
  76 char *utf82mb(const char *utf8) {
  77   return convert("UTF-8", nl_langinfo(CODESET), utf8, strlen(utf8) + 1);
  78 }
  79
  80 /** @brief Convert from encoding @p from to UTF-8 */
  81 char *any2utf8(const char *from, const char *any) {
  82   return convert(from, "UTF-8", any, strlen(any) + 1);
  83 }
  84
  85 /** @brief Convert from encoding @p from to the local multibyte encoding */
  86 char *any2mb(const char *from, const char *any) {
  87   if(from) return convert(from, nl_langinfo(CODESET), any, strlen(any) + 1);
  88   else return xstrdup(any);
  89 }
  90
  91 /** @brief Convert from encoding @p from to encoding @p to */
  92 char *any2any(const char *from,
  93               const char *to,
  94               const char *any) {
  95   if(from || to) return convert(from, to, any, strlen(any) + 1);
  96   else return xstrdup(any);
  97 }
  98
  99 /** @brief Truncate a string for display purposes
 100  * @param s Pointer to UTF-8 string
 101  * @param max Maximum number of columns
 102  * @return @p or truncated string (never NULL)
 103  *
 104  * Returns a string that is no longer than @p max graphemes long and is either
 105  * (canonically) equal to @p s or is a truncated form of it with an ellipsis
 106  * appended.
 107  *
 108  * We don't take display width into account (tricky for HTML!) and we don't
 109  * attempt to implement the Bidi algorithm.  If you have track names for which
 110  * either of these matter in practice then get in touch.
 111  */
 112 const char *truncate_for_display(const char *s, long max) {
 113   uint32_t *s32;
 114   size_t l32, cut;
 115   utf32_iterator it;
 116   long graphemes;
 117
 118   /* Convert to UTF-32 for processing */
 119   if(!(s32 = utf8_to_utf32(s, strlen(s), &l32)))
 120     return 0;
 121   it = utf32_iterator_new(s32, l32);
 122   cut = l32;
 123   graphemes = 0;                        /* # of graphemes left of it */
 124   while(graphemes <= max && utf32_iterator_where(it) < l32) {
 125     if(graphemes == max - 1)
 126       cut = utf32_iterator_where(it);
 127     utf32_iterator_advance(it, 1);
 128     if(utf32_iterator_grapheme_boundary(it))
 129       ++graphemes;
 130   }
 131   if(graphemes > max) {                 /* we need to cut */
 132     s32[cut] = 0x2026;                  /* HORIZONTAL ELLIPSIS */
 133     l32 = cut + 1;
 134     s = utf32_to_utf8(s32, l32, 0);
 135   }
 136   xfree(s32);
 137   return s;
 138 }
 139
 140 /*
 141 Local Variables:
 142 c-basic-offset:2
 143 comment-column:40
 144 fill-column:79
 145 indent-tabs-mode:nil
 146 End:
 147 */