xref: /freebsd/lib/libc/locale/collate.h (revision 2e3f49888ec8851bafb22011533217487764fdb0)
1 /*-
2  * SPDX-License-Identifier: BSD-2-Clause
3  *
4  * Copyright 2010 Nexenta Systems, Inc.  All rights reserved.
5  * Copyright (c) 1995 Alex Tatmanjants <alex@elvisti.kiev.ua>
6  *		at Electronni Visti IA, Kiev, Ukraine.
7  *			All rights reserved.
8  *
9  * Copyright (c) 2011 The FreeBSD Foundation
10  *
11  * Portions of this software were developed by David Chisnall
12  * under sponsorship from the FreeBSD Foundation.
13  *
14  * Redistribution and use in source and binary forms, with or without
15  * modification, are permitted provided that the following conditions
16  * are met:
17  * 1. Redistributions of source code must retain the above copyright
18  *    notice, this list of conditions and the following disclaimer.
19  * 2. Redistributions in binary form must reproduce the above copyright
20  *    notice, this list of conditions and the following disclaimer in the
21  *    documentation and/or other materials provided with the distribution.
22  *
23  * THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``AS IS'' AND
24  * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
25  * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
26  * ARE DISCLAIMED.  IN NO EVENT SHALL THE AUTHOR BE LIABLE
27  * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
28  * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
29  * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
30  * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
31  * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
32  * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
33  * SUCH DAMAGE.
34  */
35 
36 #ifndef _COLLATE_H_
37 #define	_COLLATE_H_
38 
39 #include <sys/types.h>
40 #include <limits.h>
41 #include "xlocale_private.h"
42 
43 /*
44  * Work around buildworld bootstrapping from older systems whose limits.h
45  * sets COLL_WEIGHTS_MAX to 0.
46  */
47 #if COLL_WEIGHTS_MAX == 0
48 #undef COLL_WEIGHTS_MAX
49 #define COLL_WEIGHTS_MAX 10
50 #endif
51 
52 #define	COLLATE_STR_LEN		24		/* should be 64-bit multiple */
53 
54 #define	COLLATE_FMT_VERSION_LEN	12
55 #define	COLLATE_FMT_VERSION	"BSD 1.0\n"
56 
57 #define	COLLATE_MAX_PRIORITY	(0x7fffffff)	/* max signed value */
58 #define	COLLATE_SUBST_PRIORITY	(0x40000000)	/* bit indicates subst table */
59 
60 #define	DIRECTIVE_UNDEF		0x00
61 #define	DIRECTIVE_FORWARD	0x01
62 #define	DIRECTIVE_BACKWARD	0x02
63 #define	DIRECTIVE_POSITION	0x04
64 #define	DIRECTIVE_UNDEFINED	0x08	/* special last weight for UNDEFINED */
65 
66 #define	DIRECTIVE_DIRECTION_MASK (DIRECTIVE_FORWARD | DIRECTIVE_BACKWARD)
67 
68 /*
69  * The collate file format is as follows:
70  *
71  * char	fmt_version[COLLATE_FMT_VERSION_LEN];	// must be COLLATE_FMT_VERSION
72  * char	def_version[XLOCALE_DEF_VERSION_LEN];	// NUL-terminated, may be empty
73  * collate_info_t	info;			// see below, includes padding
74  * collate_char_pri_t	char_data[256];		// 8 bit char values
75  * collate_subst_t	subst[*];		// 0 or more substitutions
76  * collate_chain_pri_t	chains[*];		// 0 or more chains
77  * collate_large_pri_t	large[*];		// extended char priorities
78  *
79  * Note that all structures must be 32-bit aligned, as each structure
80  * contains 32-bit member fields.  The entire file is mmap'd, so its
81  * critical that alignment be observed.  It is not generally safe to
82  * use any 64-bit values in the structures.
83  */
84 
85 typedef struct collate_info {
86 	uint8_t directive_count;
87 	uint8_t directive[COLL_WEIGHTS_MAX];
88 	int32_t pri_count[COLL_WEIGHTS_MAX];
89 	int32_t flags;
90 	int32_t chain_count;
91 	int32_t large_count;
92 	int32_t subst_count[COLL_WEIGHTS_MAX];
93 	int32_t undef_pri[COLL_WEIGHTS_MAX];
94 } collate_info_t;
95 
96 typedef struct collate_char {
97 	int32_t pri[COLL_WEIGHTS_MAX];
98 } collate_char_t;
99 
100 typedef struct collate_chain {
101 	wchar_t str[COLLATE_STR_LEN];
102 	int32_t pri[COLL_WEIGHTS_MAX];
103 } collate_chain_t;
104 
105 typedef struct collate_large {
106 	int32_t val;
107 	collate_char_t pri;
108 } collate_large_t;
109 
110 typedef struct collate_subst {
111 	int32_t key;
112 	int32_t pri[COLLATE_STR_LEN];
113 } collate_subst_t;
114 
115 struct xlocale_collate {
116 	struct xlocale_component header;
117 	int __collate_load_error;
118 	char * map;
119 	size_t maplen;
120 
121 	collate_info_t	*info;
122 	collate_char_t	*char_pri_table;
123 	collate_large_t	*large_pri_table;
124 	collate_chain_t	*chain_pri_table;
125 	collate_subst_t	*subst_table[COLL_WEIGHTS_MAX];
126 };
127 
128 __BEGIN_DECLS
129 int	__collate_load_tables(const char *);
130 int	__collate_equiv_value(locale_t, const wchar_t *, size_t);
131 void	_collate_lookup(struct xlocale_collate *,const wchar_t *, int *, int *,
132 	int, const int **);
133 int	__collate_range_cmp(char, char);
134 int	__wcollate_range_cmp(wchar_t, wchar_t);
135 size_t	_collate_wxfrm(struct xlocale_collate *, const wchar_t *, wchar_t *,
136 	size_t);
137 size_t	_collate_sxfrm(struct xlocale_collate *, const wchar_t *, char *,
138 	size_t);
139 __END_DECLS
140 
141 #endif /* !_COLLATE_H_ */
142