xref: /illumos-gate/usr/src/lib/libc/port/locale/strcoll.c (revision eb9a1df2aeb866bf1de4494433b6d7e5fa07b3ae)
1 /*
2  * Copyright 2013 Garrett D'Amore <garrett@damore.org>
3  * Copyright 2010 Nexenta Systems, Inc.  All rights reserved.
4  * Copyright (c) 1995 Alex Tatmanjants <alex@elvisti.kiev.ua>
5  *		at Electronni Visti IA, Kiev, Ukraine.
6  *			All rights reserved.
7  *
8  * Redistribution and use in source and binary forms, with or without
9  * modification, are permitted provided that the following conditions
10  * are met:
11  * 1. Redistributions of source code must retain the above copyright
12  *    notice, this list of conditions and the following disclaimer.
13  * 2. Redistributions in binary form must reproduce the above copyright
14  *    notice, this list of conditions and the following disclaimer in the
15  *    documentation and/or other materials provided with the distribution.
16  *
17  * THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``AS IS'' AND
18  * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
19  * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
20  * ARE DISCLAIMED.  IN NO EVENT SHALL THE AUTHOR BE LIABLE
21  * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
22  * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
23  * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
24  * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
25  * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
26  * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
27  * SUCH DAMAGE.
28  */
29 
30 #include "lint.h"
31 #include "file64.h"
32 #include <alloca.h>
33 #include <stdlib.h>
34 #include <string.h>
35 #include <errno.h>
36 #include <wchar.h>
37 #include <xlocale.h>
38 #include "localeimpl.h"
39 #include "collate.h"
40 
41 #define	ALLOCA_LIMIT	16
42 
43 /*
44  * In order to properly handle multibyte locales, its easiet to just
45  * convert to wide characters and then use wcscoll.  However if an
46  * error occurs, we gracefully fall back to simple strcmp.  Caller
47  * should check errno.
48  */
49 int
50 strcoll_l(const char *s1, const char *s2, locale_t loc)
51 {
52 	int ret;
53 	wchar_t *t1 = NULL, *t2 = NULL;
54 	wchar_t *w1 = NULL, *w2 = NULL;
55 	size_t sz1, sz2;
56 	const struct lc_collate *lcc = loc->collate;
57 
58 	if (lcc->lc_is_posix)
59 		return (strcmp(s1, s2));
60 
61 	sz1 = strlen(s1) + 1;
62 	sz2 = strlen(s2) + 1;
63 
64 	/*
65 	 * Simple assumption: conversion to wide format is strictly
66 	 * reducing, i.e. a single byte (or multibyte character)
67 	 * cannot result in multiple wide characters.
68 	 *
69 	 * We gain a bit of performance by giving preference to alloca
70 	 * for small string allocations.
71 	 */
72 	if (sz1 > ALLOCA_LIMIT) {
73 		if ((t1 = malloc(sz1 * sizeof (wchar_t))) == NULL)
74 			goto error;
75 		w1 = t1;
76 	} else {
77 		if ((w1 = alloca(sz1 * sizeof (wchar_t))) == NULL)
78 			goto error;
79 	}
80 	if (sz2 > ALLOCA_LIMIT) {
81 		if ((t2 = malloc(sz2 * sizeof (wchar_t))) == NULL)
82 			goto error;
83 		w2 = t2;
84 	} else {
85 		if ((w2 = alloca(sz2 * sizeof (wchar_t))) == NULL)
86 			goto error;
87 	}
88 
89 	if ((mbstowcs_l(w1, s1, sz1, loc)) == (size_t)-1)
90 		goto error;
91 
92 	if ((mbstowcs_l(w2, s2, sz2, loc)) == (size_t)-1)
93 		goto error;
94 
95 	ret = wcscoll_l(w1, w2, loc);
96 	if (t1)
97 		free(t1);
98 	if (t2)
99 		free(t2);
100 
101 	return (ret);
102 
103 error:
104 	if (t1)
105 		free(t1);
106 	if (t2)
107 		free(t2);
108 	return (strcmp(s1, s2));
109 }
110 
111 int
112 strcoll(const char *s1, const char *s2)
113 {
114 	return (strcoll_l(s1, s2, uselocale(NULL)));
115 }
116