xref: /illumos-gate/usr/src/lib/libc/port/locale/wcscoll.c (revision efcfb316ebac90ddc614edc285349bc5c55e52bd)
1 /*
2  * Copyright 2013 Garrett D'Amore <garrett@damore.org>
3  * Copyright 2017 Nexenta Systems, Inc.
4  * Copyright (c) 2002 Tim J. Robbins
5  * All rights reserved.
6  *
7  * Redistribution and use in source and binary forms, with or without
8  * modification, are permitted provided that the following conditions
9  * are met:
10  * 1. Redistributions of source code must retain the above copyright
11  *    notice, this list of conditions and the following disclaimer.
12  * 2. Redistributions in binary form must reproduce the above copyright
13  *    notice, this list of conditions and the following disclaimer in the
14  *    documentation and/or other materials provided with the distribution.
15  *
16  * THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND
17  * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
18  * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
19  * ARE DISCLAIMED.  IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE
20  * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
21  * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
22  * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
23  * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
24  * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
25  * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
26  * SUCH DAMAGE.
27  */
28 
29 #include "lint.h"
30 #include <errno.h>
31 #include <stdlib.h>
32 #include <string.h>
33 #include <wchar.h>
34 #include <assert.h>
35 #include "collate.h"
36 #include "localeimpl.h"
37 
38 int
wcscoll_l(const wchar_t * ws1,const wchar_t * ws2,locale_t loc)39 wcscoll_l(const wchar_t *ws1, const wchar_t *ws2, locale_t loc)
40 {
41 	int len1, len2, pri1, pri2;
42 	wchar_t *tr1 = NULL, *tr2 = NULL;
43 	int direc, pass;
44 	const struct lc_collate *lcc = loc->collate;
45 	int ret = wcscmp(ws1, ws2);
46 
47 	if (lcc->lc_is_posix || ret == 0)
48 		return (ret);
49 
50 	if (*ws1 == 0 && *ws2 != 0)
51 		return (-1);
52 	if (*ws1 != 0 && *ws2 == 0)
53 		return (1);
54 
55 	/*
56 	 * Once upon a time we had code to try to optimize this, but
57 	 * it turns out that you really can't make many assumptions
58 	 * safely.  You absolutely have to run this pass by pass,
59 	 * because some passes will be ignored for a given character,
60 	 * while others will not.  Simpler locales will benefit from
61 	 * having fewer passes, and most comparisions should resolve
62 	 * during the primary pass anyway.
63 	 *
64 	 * Note that we do one final extra pass at the end to pick
65 	 * up UNDEFINED elements.  There is special handling for them.
66 	 */
67 	for (pass = 0; pass <= lcc->lc_directive_count; pass++) {
68 		const int32_t *st1 = NULL;
69 		const int32_t *st2 = NULL;
70 		const wchar_t	*w1 = ws1;
71 		const wchar_t	*w2 = ws2;
72 
73 		/* special pass for UNDEFINED */
74 		if (pass == lcc->lc_directive_count) {
75 			direc = DIRECTIVE_FORWARD;
76 		} else {
77 			direc = lcc->lc_directive[pass];
78 		}
79 
80 		if (direc & DIRECTIVE_BACKWARD) {
81 			wchar_t *bp, *fp, c;
82 			free(tr1);
83 			if ((tr1 = wcsdup(w1)) == NULL)
84 				goto end;
85 			bp = tr1;
86 			fp = tr1 + wcslen(tr1) - 1;
87 			while (bp < fp) {
88 				c = *bp;
89 				*bp++ = *fp;
90 				*fp-- = c;
91 			}
92 			free(tr2);
93 			if ((tr2 = wcsdup(w2)) == NULL)
94 				goto end;
95 			bp = tr2;
96 			fp = tr2 + wcslen(tr2) - 1;
97 			while (bp < fp) {
98 				c = *bp;
99 				*bp++ = *fp;
100 				*fp-- = c;
101 			}
102 			w1 = tr1;
103 			w2 = tr2;
104 		}
105 
106 		if (direc & DIRECTIVE_POSITION) {
107 			int check1, check2;
108 			while (*w1 && *w2) {
109 				pri1 = pri2 = 0;
110 				check1 = check2 = 1;
111 				while ((pri1 == pri2) && (check1 || check2)) {
112 					if (check1) {
113 						_collate_lookup(lcc, w1, &len1,
114 						    &pri1, pass, &st1);
115 						if (pri1 < 0) {
116 							errno = EINVAL;
117 							goto end;
118 						}
119 						if (!pri1) {
120 							/*CSTYLED*/
121 							pri1 = COLLATE_MAX_PRIORITY;
122 							st1 = NULL;
123 						}
124 						check1 = (st1 != NULL);
125 					}
126 					if (check2) {
127 						_collate_lookup(lcc, w2, &len2,
128 						    &pri2, pass, &st2);
129 						if (pri2 < 0) {
130 							errno = EINVAL;
131 							goto end;
132 						}
133 						if (!pri2) {
134 							/*CSTYLED*/
135 							pri2 = COLLATE_MAX_PRIORITY;
136 							st2 = NULL;
137 						}
138 						check2 = (st2 != NULL);
139 					}
140 				}
141 				if (pri1 != pri2) {
142 					ret = pri1 - pri2;
143 					goto end;
144 				}
145 				w1 += len1;
146 				w2 += len2;
147 			}
148 			if (!*w1) {
149 				if (*w2) {
150 					ret = -(int)*w2;
151 					goto end;
152 				}
153 			} else {
154 				ret = *w1;
155 				goto end;
156 			}
157 		} else {
158 			int vpri1 = 0, vpri2 = 0;
159 			while (*w1 || *w2 || st1 || st2) {
160 				pri1 = 1;
161 				while (*w1 || st1) {
162 					_collate_lookup(lcc, w1, &len1, &pri1,
163 					    pass, &st1);
164 					w1 += len1;
165 					if (pri1 > 0) {
166 						vpri1++;
167 						break;
168 					}
169 					if (pri1 < 0) {
170 						errno = EINVAL;
171 						goto end;
172 					}
173 					st1 = NULL;
174 				}
175 				pri2 = 1;
176 				while (*w2 || st2) {
177 					_collate_lookup(lcc, w2, &len2, &pri2,
178 					    pass, &st2);
179 					w2 += len2;
180 					if (pri2 > 0) {
181 						vpri2++;
182 						break;
183 					}
184 					if (pri2 < 0) {
185 						errno = EINVAL;
186 						goto end;
187 					}
188 					st2 = NULL;
189 				}
190 				if ((!pri1 || !pri2) && (vpri1 == vpri2))
191 					break;
192 				if (pri1 != pri2) {
193 					ret = pri1 - pri2;
194 					goto end;
195 				}
196 			}
197 			if (vpri1 && !vpri2) {
198 				ret = 1;
199 				goto end;
200 			}
201 			if (!vpri1 && vpri2) {
202 				ret = -1;
203 				goto end;
204 			}
205 		}
206 	}
207 	ret = 0;
208 
209 end:
210 	free(tr1);
211 	free(tr2);
212 
213 	return (ret);
214 }
215 
216 int
wcscoll(const wchar_t * ws1,const wchar_t * ws2)217 wcscoll(const wchar_t *ws1, const wchar_t *ws2)
218 {
219 	return (wcscoll_l(ws1, ws2, uselocale(NULL)));
220 }
221