xref: /freebsd/contrib/libarchive/tar/subst.c (revision fcb560670601b2a4d87bb31d7531c8dcc37ee71b)
1 /*-
2  * Copyright (c) 2008 Joerg Sonnenberger
3  * All rights reserved.
4  *
5  * Redistribution and use in source and binary forms, with or without
6  * modification, are permitted provided that the following conditions
7  * are met:
8  * 1. Redistributions of source code must retain the above copyright
9  *    notice, this list of conditions and the following disclaimer.
10  * 2. Redistributions in binary form must reproduce the above copyright
11  *    notice, this list of conditions and the following disclaimer in the
12  *    documentation and/or other materials provided with the distribution.
13  *
14  * THIS SOFTWARE IS PROVIDED BY THE AUTHOR(S) ``AS IS'' AND ANY EXPRESS OR
15  * IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES
16  * OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED.
17  * IN NO EVENT SHALL THE AUTHOR(S) BE LIABLE FOR ANY DIRECT, INDIRECT,
18  * INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT
19  * NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
20  * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
21  * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
22  * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF
23  * THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
24  */
25 
26 #include "bsdtar_platform.h"
27 __FBSDID("$FreeBSD$");
28 
29 #if defined(HAVE_REGEX_H) || defined(HAVE_PCREPOSIX_H)
30 #include "bsdtar.h"
31 
32 #include <errno.h>
33 #ifdef HAVE_PCREPOSIX_H
34 #include <pcreposix.h>
35 #else
36 #include <regex.h>
37 #endif
38 #include <stdlib.h>
39 #include <string.h>
40 
41 #ifndef REG_BASIC
42 #define	REG_BASIC 0
43 #endif
44 
45 #include "err.h"
46 
47 struct subst_rule {
48 	struct subst_rule *next;
49 	regex_t re;
50 	char *result;
51 	unsigned int global:1, print:1, regular:1, symlink:1, hardlink:1;
52 };
53 
54 struct substitution {
55 	struct subst_rule *first_rule, *last_rule;
56 };
57 
58 static void
59 init_substitution(struct bsdtar *bsdtar)
60 {
61 	struct substitution *subst;
62 
63 	bsdtar->substitution = subst = malloc(sizeof(*subst));
64 	if (subst == NULL)
65 		lafe_errc(1, errno, "Out of memory");
66 	subst->first_rule = subst->last_rule = NULL;
67 }
68 
69 void
70 add_substitution(struct bsdtar *bsdtar, const char *rule_text)
71 {
72 	struct subst_rule *rule;
73 	struct substitution *subst;
74 	const char *end_pattern, *start_subst;
75 	char *pattern;
76 	int r;
77 
78 	if ((subst = bsdtar->substitution) == NULL) {
79 		init_substitution(bsdtar);
80 		subst = bsdtar->substitution;
81 	}
82 
83 	rule = malloc(sizeof(*rule));
84 	if (rule == NULL)
85 		lafe_errc(1, errno, "Out of memory");
86 	rule->next = NULL;
87 
88 	if (subst->last_rule == NULL)
89 		subst->first_rule = rule;
90 	else
91 		subst->last_rule->next = rule;
92 	subst->last_rule = rule;
93 
94 	if (*rule_text == '\0')
95 		lafe_errc(1, 0, "Empty replacement string");
96 	end_pattern = strchr(rule_text + 1, *rule_text);
97 	if (end_pattern == NULL)
98 		lafe_errc(1, 0, "Invalid replacement string");
99 
100 	pattern = malloc(end_pattern - rule_text);
101 	if (pattern == NULL)
102 		lafe_errc(1, errno, "Out of memory");
103 	memcpy(pattern, rule_text + 1, end_pattern - rule_text - 1);
104 	pattern[end_pattern - rule_text - 1] = '\0';
105 
106 	if ((r = regcomp(&rule->re, pattern, REG_BASIC)) != 0) {
107 		char buf[80];
108 		regerror(r, &rule->re, buf, sizeof(buf));
109 		lafe_errc(1, 0, "Invalid regular expression: %s", buf);
110 	}
111 	free(pattern);
112 
113 	start_subst = end_pattern + 1;
114 	end_pattern = strchr(start_subst, *rule_text);
115 	if (end_pattern == NULL)
116 		lafe_errc(1, 0, "Invalid replacement string");
117 
118 	rule->result = malloc(end_pattern - start_subst + 1);
119 	if (rule->result == NULL)
120 		lafe_errc(1, errno, "Out of memory");
121 	memcpy(rule->result, start_subst, end_pattern - start_subst);
122 	rule->result[end_pattern - start_subst] = '\0';
123 
124 	/* Defaults */
125 	rule->global = 0; /* Don't do multiple replacements. */
126 	rule->print = 0; /* Don't print. */
127 	rule->regular = 1; /* Rewrite regular filenames. */
128 	rule->symlink = 1; /* Rewrite symlink targets. */
129 	rule->hardlink = 1; /* Rewrite hardlink targets. */
130 
131 	while (*++end_pattern) {
132 		switch (*end_pattern) {
133 		case 'g':
134 		case 'G':
135 			rule->global = 1;
136 			break;
137 		case 'h':
138 			rule->hardlink = 1;
139 			break;
140 		case 'H':
141 			rule->hardlink = 0;
142 			break;
143 		case 'p':
144 		case 'P':
145 			rule->print = 1;
146 			break;
147 		case 'r':
148 			rule->regular = 1;
149 			break;
150 		case 'R':
151 			rule->regular = 0;
152 			break;
153 		case 's':
154 			rule->symlink = 1;
155 			break;
156 		case 'S':
157 			rule->symlink = 0;
158 			break;
159 		default:
160 			lafe_errc(1, 0, "Invalid replacement flag %c", *end_pattern);
161 		}
162 	}
163 }
164 
165 static void
166 realloc_strncat(char **str, const char *append, size_t len)
167 {
168 	char *new_str;
169 	size_t old_len;
170 
171 	if (*str == NULL)
172 		old_len = 0;
173 	else
174 		old_len = strlen(*str);
175 
176 	new_str = malloc(old_len + len + 1);
177 	if (new_str == NULL)
178 		lafe_errc(1, errno, "Out of memory");
179 	if (*str != NULL)
180 		memcpy(new_str, *str, old_len);
181 	memcpy(new_str + old_len, append, len);
182 	new_str[old_len + len] = '\0';
183 	free(*str);
184 	*str = new_str;
185 }
186 
187 static void
188 realloc_strcat(char **str, const char *append)
189 {
190 	char *new_str;
191 	size_t old_len;
192 
193 	if (*str == NULL)
194 		old_len = 0;
195 	else
196 		old_len = strlen(*str);
197 
198 	new_str = malloc(old_len + strlen(append) + 1);
199 	if (new_str == NULL)
200 		lafe_errc(1, errno, "Out of memory");
201 	if (*str != NULL)
202 		memcpy(new_str, *str, old_len);
203 	strcpy(new_str + old_len, append);
204 	free(*str);
205 	*str = new_str;
206 }
207 
208 int
209 apply_substitution(struct bsdtar *bsdtar, const char *name, char **result,
210     int symlink_target, int hardlink_target)
211 {
212 	const char *path = name;
213 	regmatch_t matches[10];
214 	size_t i, j;
215 	struct subst_rule *rule;
216 	struct substitution *subst;
217 	int c, got_match, print_match;
218 
219 	*result = NULL;
220 
221 	if ((subst = bsdtar->substitution) == NULL)
222 		return 0;
223 
224 	got_match = 0;
225 	print_match = 0;
226 
227 	for (rule = subst->first_rule; rule != NULL; rule = rule->next) {
228 		if (symlink_target) {
229 			if (!rule->symlink)
230 				continue;
231 		} else if (hardlink_target) {
232 			if (!rule->hardlink)
233 				continue;
234 		} else { /* Regular filename. */
235 			if (!rule->regular)
236 				continue;
237 		}
238 
239 		if (regexec(&rule->re, name, 10, matches, 0))
240 			continue;
241 
242 		got_match = 1;
243 		print_match |= rule->print;
244 		realloc_strncat(result, name, matches[0].rm_so);
245 
246 		for (i = 0, j = 0; rule->result[i] != '\0'; ++i) {
247 			if (rule->result[i] == '~') {
248 				realloc_strncat(result, rule->result + j, i - j);
249 				realloc_strncat(result,
250 				    name + matches[0].rm_so,
251 				    matches[0].rm_eo - matches[0].rm_so);
252 				j = i + 1;
253 				continue;
254 			}
255 			if (rule->result[i] != '\\')
256 				continue;
257 
258 			++i;
259 			c = rule->result[i];
260 			switch (c) {
261 			case '~':
262 			case '\\':
263 				realloc_strncat(result, rule->result + j, i - j - 1);
264 				j = i;
265 				break;
266 			case '1':
267 			case '2':
268 			case '3':
269 			case '4':
270 			case '5':
271 			case '6':
272 			case '7':
273 			case '8':
274 			case '9':
275 				realloc_strncat(result, rule->result + j, i - j - 1);
276 				if ((size_t)(c - '0') > (size_t)(rule->re.re_nsub)) {
277 					free(*result);
278 					*result = NULL;
279 					return -1;
280 				}
281 				realloc_strncat(result, name + matches[c - '0'].rm_so, matches[c - '0'].rm_eo - matches[c - '0'].rm_so);
282 				j = i + 1;
283 				break;
284 			default:
285 				/* Just continue; */
286 				break;
287 			}
288 
289 		}
290 
291 		realloc_strcat(result, rule->result + j);
292 
293 		name += matches[0].rm_eo;
294 
295 		if (!rule->global)
296 			break;
297 	}
298 
299 	if (got_match)
300 		realloc_strcat(result, name);
301 
302 	if (print_match)
303 		fprintf(stderr, "%s >> %s\n", path, *result);
304 
305 	return got_match;
306 }
307 
308 void
309 cleanup_substitution(struct bsdtar *bsdtar)
310 {
311 	struct subst_rule *rule;
312 	struct substitution *subst;
313 
314 	if ((subst = bsdtar->substitution) == NULL)
315 		return;
316 
317 	while ((rule = subst->first_rule) != NULL) {
318 		subst->first_rule = rule->next;
319 		free(rule->result);
320 		free(rule);
321 	}
322 	free(subst);
323 }
324 #endif /* defined(HAVE_REGEX_H) || defined(HAVE_PCREPOSIX_H) */
325