libunibreak 8.0
Loading...
Searching...
No Matches
unibreakdef.h
Go to the documentation of this file.
1/* vim: set expandtab tabstop=4 softtabstop=4 shiftwidth=4: */
2
3/*
4 * Break processing in a Unicode sequence. Designed to be used in a
5 * generic text renderer.
6 *
7 * Copyright (C) 2015-2026 Wu Yongwei <wuyongwei at gmail dot com>
8 *
9 * This software is provided 'as-is', without any express or implied
10 * warranty. In no event will the author be held liable for any damages
11 * arising from the use of this software.
12 *
13 * Permission is granted to anyone to use this software for any purpose,
14 * including commercial applications, and to alter it and redistribute
15 * it freely, subject to the following restrictions:
16 *
17 * 1. The origin of this software must not be misrepresented; you must
18 * not claim that you wrote the original software. If you use this
19 * software in a product, an acknowledgement in the product
20 * documentation would be appreciated but is not required.
21 * 2. Altered source versions must be plainly marked as such, and must
22 * not be misrepresented as being the original software.
23 * 3. This notice may not be removed or altered from any source
24 * distribution.
25 */
26
34
35#ifndef UNIBREAKDEF_H
36#define UNIBREAKDEF_H
37
38#include <stdbool.h>
39#include <stddef.h>
40#include "unibreakbase.h"
41
42#define ARRAY_LEN(x) (sizeof(x) / sizeof(x[0]))
43
44#ifdef __cplusplus
45extern "C" {
46#endif
47
52#define EOS 0xFFFFFFFF
53
58typedef utf32_t (*get_next_char_t)(const void *, size_t, size_t *);
59
60/* Function Prototype */
61utf32_t ub_get_next_char_utf8 (const void *sv, size_t len, size_t *ip);
62utf32_t ub_get_next_char_utf16(const void *sv, size_t len, size_t *ip);
63utf32_t ub_get_next_char_utf32(const void *sv, size_t len, size_t *ip);
64
65inline const void *ub_bsearch(utf32_t ch, const void *ptr, size_t count,
66 size_t size)
67{
68 int min = 0;
69 int max = (int)(count - 1);;
70 int mid;
71
72 do
73 {
74 mid = (min + max) / 2;
75 const unsigned char *mid_ptr =
76 (const unsigned char *)ptr + mid * size;
77 utf32_t mid_start = *(const utf32_t *)mid_ptr;
78 utf32_t mid_end = *((const utf32_t *)mid_ptr + 1);
79
80 if (ch < mid_start)
81 {
82 max = mid - 1;
83 }
84 else if (ch > mid_end)
85 {
86 min = mid + 1;
87 }
88 else
89 {
90 return mid_ptr;
91 }
92 } while (min <= max);
93
94 return NULL;
95}
96
97#ifdef __cplusplus
98}
99#endif
100
101#endif /* UNIBREAKDEF_H */
Header file for common definitions in the libunibreak library.
unsigned int utf32_t
Type for UTF-32 data points.
Definition unibreakbase.h:49
utf32_t ub_get_next_char_utf8(const void *sv, size_t len, size_t *ip)
Gets the next Unicode character in a UTF-8 sequence.
Definition unibreakdef.c:50
utf32_t ub_get_next_char_utf32(const void *sv, size_t len, size_t *ip)
Gets the next Unicode character in a UTF-32 sequence.
Definition unibreakdef.c:161
const void * ub_bsearch(utf32_t ch, const void *ptr, size_t count, size_t size)
Definition unibreakdef.h:65
utf32_t ub_get_next_char_utf16(const void *sv, size_t len, size_t *ip)
Gets the next Unicode character in a UTF-16 sequence.
Definition unibreakdef.c:119
utf32_t(*) get_next_char_t(const void *, size_t, size_t *)
Abstract function interface for ub_get_next_char_utf8, ub_get_next_char_utf16, and ub_get_next_char_u...
Definition unibreakdef.h:58