xref: /freebsd/contrib/libarchive/tar/subst.c (revision 5ca8e32633c4ffbbcd6762e5888b6a4ba0708c6c)
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 	rule->result = NULL;
88 
89 	if (subst->last_rule == NULL)
90 		subst->first_rule = rule;
91 	else
92 		subst->last_rule->next = rule;
93 	subst->last_rule = rule;
94 
95 	if (*rule_text == '\0')
96 		lafe_errc(1, 0, "Empty replacement string");
97 	end_pattern = strchr(rule_text + 1, *rule_text);
98 	if (end_pattern == NULL)
99 		lafe_errc(1, 0, "Invalid replacement string");
100 
101 	pattern = malloc(end_pattern - rule_text);
102 	if (pattern == NULL)
103 		lafe_errc(1, errno, "Out of memory");
104 	memcpy(pattern, rule_text + 1, end_pattern - rule_text - 1);
105 	pattern[end_pattern - rule_text - 1] = '\0';
106 
107 	if ((r = regcomp(&rule->re, pattern, REG_BASIC)) != 0) {
108 		char buf[80];
109 		regerror(r, &rule->re, buf, sizeof(buf));
110 		lafe_errc(1, 0, "Invalid regular expression: %s", buf);
111 	}
112 	free(pattern);
113 
114 	start_subst = end_pattern + 1;
115 	end_pattern = strchr(start_subst, *rule_text);
116 	if (end_pattern == NULL)
117 		lafe_errc(1, 0, "Invalid replacement string");
118 
119 	rule->result = malloc(end_pattern - start_subst + 1);
120 	if (rule->result == NULL)
121 		lafe_errc(1, errno, "Out of memory");
122 	memcpy(rule->result, start_subst, end_pattern - start_subst);
123 	rule->result[end_pattern - start_subst] = '\0';
124 
125 	/* Defaults */
126 	rule->global = 0; /* Don't do multiple replacements. */
127 	rule->print = 0; /* Don't print. */
128 	rule->regular = 1; /* Rewrite regular filenames. */
129 	rule->symlink = 1; /* Rewrite symlink targets. */
130 	rule->hardlink = 1; /* Rewrite hardlink targets. */
131 
132 	while (*++end_pattern) {
133 		switch (*end_pattern) {
134 		case 'g':
135 		case 'G':
136 			rule->global = 1;
137 			break;
138 		case 'h':
139 			rule->hardlink = 1;
140 			break;
141 		case 'H':
142 			rule->hardlink = 0;
143 			break;
144 		case 'p':
145 		case 'P':
146 			rule->print = 1;
147 			break;
148 		case 'r':
149 			rule->regular = 1;
150 			break;
151 		case 'R':
152 			rule->regular = 0;
153 			break;
154 		case 's':
155 			rule->symlink = 1;
156 			break;
157 		case 'S':
158 			rule->symlink = 0;
159 			break;
160 		default:
161 			lafe_errc(1, 0, "Invalid replacement flag %c", *end_pattern);
162 		}
163 	}
164 }
165 
166 static void
167 realloc_strncat(char **str, const char *append, size_t len)
168 {
169 	char *new_str;
170 	size_t old_len;
171 
172 	if (*str == NULL)
173 		old_len = 0;
174 	else
175 		old_len = strlen(*str);
176 
177 	new_str = malloc(old_len + len + 1);
178 	if (new_str == NULL)
179 		lafe_errc(1, errno, "Out of memory");
180 	if (*str != NULL)
181 		memcpy(new_str, *str, old_len);
182 	memcpy(new_str + old_len, append, len);
183 	new_str[old_len + len] = '\0';
184 	free(*str);
185 	*str = new_str;
186 }
187 
188 static void
189 realloc_strcat(char **str, const char *append)
190 {
191 	char *new_str;
192 	size_t old_len;
193 
194 	if (*str == NULL)
195 		old_len = 0;
196 	else
197 		old_len = strlen(*str);
198 
199 	new_str = malloc(old_len + strlen(append) + 1);
200 	if (new_str == NULL)
201 		lafe_errc(1, errno, "Out of memory");
202 	if (*str != NULL)
203 		memcpy(new_str, *str, old_len);
204 	strcpy(new_str + old_len, append);
205 	free(*str);
206 	*str = new_str;
207 }
208 
209 int
210 apply_substitution(struct bsdtar *bsdtar, const char *name, char **result,
211     int symlink_target, int hardlink_target)
212 {
213 	const char *path = name;
214 	regmatch_t matches[10];
215 	size_t i, j;
216 	struct subst_rule *rule;
217 	struct substitution *subst;
218 	int c, got_match, print_match;
219 
220 	*result = NULL;
221 
222 	if ((subst = bsdtar->substitution) == NULL)
223 		return 0;
224 
225 	got_match = 0;
226 	print_match = 0;
227 
228 	for (rule = subst->first_rule; rule != NULL; rule = rule->next) {
229 		if (symlink_target) {
230 			if (!rule->symlink)
231 				continue;
232 		} else if (hardlink_target) {
233 			if (!rule->hardlink)
234 				continue;
235 		} else { /* Regular filename. */
236 			if (!rule->regular)
237 				continue;
238 		}
239 
240 		while (1) {
241 			if (regexec(&rule->re, name, 10, matches, 0))
242 				break;
243 
244 			got_match = 1;
245 			print_match |= rule->print;
246 			realloc_strncat(result, name, matches[0].rm_so);
247 
248 			for (i = 0, j = 0; rule->result[i] != '\0'; ++i) {
249 				if (rule->result[i] == '~') {
250 					realloc_strncat(result, rule->result + j, i - j);
251 					realloc_strncat(result,
252 					    name + matches[0].rm_so,
253 					    matches[0].rm_eo - matches[0].rm_so);
254 					j = i + 1;
255 					continue;
256 				}
257 				if (rule->result[i] != '\\')
258 					continue;
259 
260 				++i;
261 				c = rule->result[i];
262 				switch (c) {
263 				case '~':
264 				case '\\':
265 					realloc_strncat(result, rule->result + j, i - j - 1);
266 					j = i;
267 					break;
268 				case '1':
269 				case '2':
270 				case '3':
271 				case '4':
272 				case '5':
273 				case '6':
274 				case '7':
275 				case '8':
276 				case '9':
277 					realloc_strncat(result, rule->result + j, i - j - 1);
278 					if ((size_t)(c - '0') > (size_t)(rule->re.re_nsub)) {
279 						free(*result);
280 						*result = NULL;
281 						return -1;
282 					}
283 					realloc_strncat(result, name + matches[c - '0'].rm_so, matches[c - '0'].rm_eo - matches[c - '0'].rm_so);
284 					j = i + 1;
285 					break;
286 				default:
287 					/* Just continue; */
288 					break;
289 				}
290 
291 			}
292 
293 			realloc_strcat(result, rule->result + j);
294 
295 			name += matches[0].rm_eo;
296 
297 			if (!rule->global)
298 				break;
299 		}
300 	}
301 
302 	if (got_match)
303 		realloc_strcat(result, name);
304 
305 	if (print_match)
306 		fprintf(stderr, "%s >> %s\n", path, *result);
307 
308 	return got_match;
309 }
310 
311 void
312 cleanup_substitution(struct bsdtar *bsdtar)
313 {
314 	struct subst_rule *rule;
315 	struct substitution *subst;
316 
317 	if ((subst = bsdtar->substitution) == NULL)
318 		return;
319 
320 	while ((rule = subst->first_rule) != NULL) {
321 		subst->first_rule = rule->next;
322 		free(rule->result);
323 		regfree(&rule->re);
324 		free(rule);
325 	}
326 	free(subst);
327 }
328 #endif /* defined(HAVE_REGEX_H) || defined(HAVE_PCREPOSIX_H) */
329