chiark / gitweb /
build fixes
[disorder] / lib / unicode.h
CommitLineData
e5a5a138
RK
1/*
2 * This file is part of DisOrde
3 * Copyright (C) 2007 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/unicode.h
21 * @brief Unicode support functions
22 */
23
24#ifndef UNICODE_H
25#define UNICODE_H
26
092f426f
RK
27/** @brief Smart pointer into a string
28 *
29 * Iterators can be efficiently moved either forwards or back to the start of
30 * the string. They cannot (currently) efficiently be moved backwards. Their
31 * advantage is that they remember internal state to speed up boundary
32 * detection.
33 *
34 * Iterators can point to any code point of the string, or to a hypothetical
35 * post-final code point of value 0, but not outside the string.
36 */
37typedef struct utf32_iterator_data *utf32_iterator;
38
c85b7022
RK
39/** @brief Property tailor function
40 * @param c Code point
41 * @return Tailored property or -1 to use standard value
42 *
43 * See also utf32_iterator_tailor_word_break().
44 */
45typedef int unicode_property_tailor(uint32_t c);
46
e5a5a138
RK
47char *utf32_to_utf8(const uint32_t *s, size_t ns, size_t *nd);
48uint32_t *utf8_to_utf32(const char *s, size_t ns, size_t *nd);
18cda350 49int utf8_valid(const char *s, size_t ns);
e5a5a138
RK
50
51size_t utf32_len(const uint32_t *s);
52int utf32_cmp(const uint32_t *a, const uint32_t *b);
53
54uint32_t *utf32_decompose_canon(const uint32_t *s, size_t ns, size_t *ndp);
55char *utf8_decompose_canon(const char *s, size_t ns, size_t *ndp);
56
57uint32_t *utf32_decompose_compat(const uint32_t *s, size_t ns, size_t *ndp);
58char *utf8_decompose_compat(const char *s, size_t ns, size_t *ndp);
59
16506c9d
RK
60uint32_t *utf32_compose_canon(const uint32_t *s, size_t ns, size_t *ndp);
61char *utf8_compose_canon(const char *s, size_t ns, size_t *ndp);
62
63uint32_t *utf32_compose_compat(const uint32_t *s, size_t ns, size_t *ndp);
64char *utf8_compose_compat(const char *s, size_t ns, size_t *ndp);
65
e5a5a138
RK
66uint32_t *utf32_casefold_canon(const uint32_t *s, size_t ns, size_t *ndp);
67char *utf8_casefold_canon(const char *s, size_t ns, size_t *ndp);
68
69uint32_t *utf32_casefold_compat(const uint32_t *s, size_t ns, size_t *ndp);
70char *utf8_casefold_compat(const char *s, size_t ns, size_t *ndp);
71
1625e11a 72int utf32_is_grapheme_boundary(const uint32_t *s, size_t ns, size_t n);
0b7052da 73int utf32_is_word_boundary(const uint32_t *s, size_t ns, size_t n);
e5a5a138 74
092f426f
RK
75utf32_iterator utf32_iterator_new(const uint32_t *s, size_t ns);
76void utf32_iterator_destroy(utf32_iterator it);
77
78size_t utf32_iterator_where(utf32_iterator it);
79int utf32_iterator_set(utf32_iterator it, size_t n);
80int utf32_iterator_advance(utf32_iterator it, size_t n);
81uint32_t utf32_iterator_code(utf32_iterator it);
82int utf32_iterator_grapheme_boundary(utf32_iterator it);
83int utf32_iterator_word_boundary(utf32_iterator it);
c85b7022
RK
84void utf32_iterator_tailor_word_break(utf32_iterator it,
85 unicode_property_tailor *pt);
092f426f 86
c85b7022
RK
87uint32_t **utf32_word_split(const uint32_t *s, size_t ns, size_t *nwp,
88 unicode_property_tailor *wbreak);
89char **utf8_word_split(const char *s, size_t ns, size_t *nwp,
90 unicode_property_tailor *wbreak);
8818b7fc 91
caecd4f4
RK
92/** @brief Convert 0-terminated UTF-32 to UTF-8
93 * @param s 0-terminated UTF-32 string
94 * @return 0-terminated UTF-8 string or 0 on error
95 *
96 * See utf32_to_utf8() for possible causes of errors.
97 */
98static inline char *utf32nt_to_utf8(const uint32_t *s) {
99 return utf32_to_utf8(s, utf32_len(s), 0);
100}
101
102/** @brief Convert 0-terminated UTF-8 to UTF-32
103 * @param s 0-terminated UTF-8 string
104 * @return 0-terminated UTF-32 string or 0 on error
105 *
106 * See utf8_to_utf32() for possible causes of errors.
107 */
108static inline uint32_t *utf8nt_to_utf32(const char *s) {
109 return utf8_to_utf32(s, strlen(s), 0);
110}
111
e5a5a138
RK
112#endif /* UNICODE_H */
113
114/*
115Local Variables:
116c-basic-offset:2
117comment-column:40
118fill-column:79
119indent-tabs-mode:nil
120End:
121*/