chiark / gitweb /
distinguish the different decompositions via a flag bit
[disorder] / lib / charset.c
CommitLineData
460b9539 1/*
2 * This file is part of DisOrder.
3 * Copyright (C) 2004, 2005 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 */
14ad73b9 20/** @file lib/charset.c @brief Character set conversion */
460b9539 21
22#include <config.h>
23#include "types.h"
24
25#include <iconv.h>
26#include <string.h>
27#include <errno.h>
28#include <langinfo.h>
29
30#include "mem.h"
31#include "log.h"
32#include "charset.h"
33#include "configuration.h"
34#include "utf8.h"
35#include "vector.h"
61507e3c 36#include "unidata.h"
460b9539 37
14ad73b9
RK
38/** @brief Low-level converstion routine
39 * @param from Source encoding
40 * @param to Destination encoding
41 * @param ptr First byte to convert
42 * @param n Number of bytes to convert
43 * @return Converted text, 0-terminated; or NULL on error.
44 */
460b9539 45static void *convert(const char *from, const char *to,
46 const void *ptr, size_t n) {
47 iconv_t i;
48 size_t len;
49 char *buf = 0, *s, *d;
50 size_t bufsize = 0, sl, dl;
51
52 if((i = iconv_open(to, from)) == (iconv_t)-1)
53 fatal(errno, "error calling iconv_open");
54 do {
55 bufsize = bufsize ? 2 * bufsize : 32;
56 buf = xrealloc_noptr(buf, bufsize);
57 iconv(i, 0, 0, 0, 0);
58 s = (char *)ptr;
59 sl = n;
60 d = buf;
61 dl = bufsize;
62 /* (void *) to work around FreeBSD's nonstandard iconv prototype */
63 len = iconv(i, (void *)&s, &sl, &d, &dl);
64 } while(len == (size_t)-1 && errno == E2BIG);
65 iconv_close(i);
66 if(len == (size_t)-1) {
67 error(errno, "error converting from %s to %s", from, to);
68 return 0;
69 }
70 return buf;
71}
72
14ad73b9 73/** @brief Convert from the local multibyte encoding to UTF-8 */
460b9539 74char *mb2utf8(const char *mb) {
75 return convert(nl_langinfo(CODESET), "UTF-8", mb, strlen(mb) + 1);
76}
77
14ad73b9 78/** @brief Convert from UTF-8 to the local multibyte encoding */
460b9539 79char *utf82mb(const char *utf8) {
80 return convert("UTF-8", nl_langinfo(CODESET), utf8, strlen(utf8) + 1);
81}
82
14ad73b9 83/** @brief Convert from encoding @p from to UTF-8 */
460b9539 84char *any2utf8(const char *from, const char *any) {
85 return convert(from, "UTF-8", any, strlen(any) + 1);
86}
87
14ad73b9 88/** @brief Convert from encoding @p from to the local multibyte encoding */
460b9539 89char *any2mb(const char *from, const char *any) {
90 if(from) return convert(from, nl_langinfo(CODESET), any, strlen(any) + 1);
91 else return xstrdup(any);
92}
93
14ad73b9 94/** @brief Convert from encoding @p from to encoding @p to */
460b9539 95char *any2any(const char *from,
96 const char *to,
97 const char *any) {
98 if(from || to) return convert(from, to, any, strlen(any) + 1);
99 else return xstrdup(any);
100}
101
61507e3c
RK
102/** @brief Return nonzero if @p c is a combining character */
103static int combining(int c) {
104 if(c < UNICODE_NCHARS) {
e5a5a138 105 const struct unidata *const ud = &unidata[c / UNICODE_MODULUS][c % UNICODE_MODULUS];
61507e3c 106
14523635 107 return ud->general_category == unicode_General_Category_Mn || ud->ccc != 0;
61507e3c
RK
108 }
109 /* Assume unknown characters are noncombining */
110 return 0;
111}
112
113/** @brief Truncate a string for display purposes
114 * @param s Pointer to UTF-8 string
115 * @param max Maximum number of columns
116 * @return @p or truncated string (never NULL)
117 *
118 * We don't correctly support bidi or double-width characters yet, nor
119 * locate default grapheme cluster boundaries for saner truncation.
120 */
121const char *truncate_for_display(const char *s, long max) {
122 const char *t = s, *r, *cut = 0;
123 char *truncated;
124 uint32_t c;
125 long n = 0;
126
127 /* We need to discover two things: firstly whether the string is
128 * longer than @p max glyphs and secondly if it is not, where to cut
129 * the string.
130 *
131 * Combining characters follow their base character (unicode
132 * standard 5.0 s2.11), so after each base character we must
133 */
134 while(*t) {
135 PARSE_UTF8(t, c, return s);
136 if(combining(c))
137 /* This must be an initial combining character. We just skip it. */
138 continue;
139 /* So c must be a base character. It may be followed by any
140 * number of combining characters. We advance past them. */
141 do {
142 r = t;
143 PARSE_UTF8(t, c, return s);
144 } while(combining(c));
145 /* Last character wasn't a combining character so back up */
146 t = r;
147 ++n;
148 /* So now there are N glyphs before position T. We might
149 * therefore have reached the cut position. */
150 if(n == max - 3)
151 cut = t;
152 }
153 /* If the string is short enough we return it unmodified */
154 if(n < max)
155 return s;
156 truncated = xmalloc_noptr(cut - s + 4);
157 memcpy(truncated, s, cut - s);
158 strcpy(truncated + (cut - s), "...");
159 return truncated;
160}
161
460b9539 162/*
163Local Variables:
164c-basic-offset:2
165comment-column:40
166End:
167*/