/src/sudo/lib/util/regex.c
Line | Count | Source (jump to first uncovered line) |
1 | | /* |
2 | | * SPDX-License-Identifier: ISC |
3 | | * |
4 | | * Copyright (c) 2023 Todd C. Miller <Todd.Miller@sudo.ws> |
5 | | * |
6 | | * Permission to use, copy, modify, and distribute this software for any |
7 | | * purpose with or without fee is hereby granted, provided that the above |
8 | | * copyright notice and this permission notice appear in all copies. |
9 | | * |
10 | | * THE SOFTWARE IS PROVIDED "AS IS" AND THE AUTHOR DISCLAIMS ALL WARRANTIES |
11 | | * WITH REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED WARRANTIES OF |
12 | | * MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR |
13 | | * ANY SPECIAL, DIRECT, INDIRECT, OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES |
14 | | * WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER IN AN |
15 | | * ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF |
16 | | * OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE. |
17 | | */ |
18 | | |
19 | | /* |
20 | | * This is an open source non-commercial project. Dear PVS-Studio, please check it. |
21 | | * PVS-Studio Static Code Analyzer for C, C++ and C#: http://www.viva64.com |
22 | | */ |
23 | | |
24 | | #include <config.h> |
25 | | |
26 | | #include <stddef.h> |
27 | | #include <stdlib.h> |
28 | | #include <string.h> |
29 | | #include <ctype.h> |
30 | | #include <limits.h> |
31 | | #include <regex.h> |
32 | | |
33 | | #include "sudo_compat.h" |
34 | | #include "sudo_debug.h" |
35 | | #include "sudo_util.h" |
36 | | #include "sudo_gettext.h" |
37 | | |
38 | | static char errbuf[1024]; |
39 | | |
40 | | /* |
41 | | * Parse a number between 0 and INT_MAX, handling escaped digits. |
42 | | * Returns the number on success or -1 on error. |
43 | | * Sets endp to the first non-matching character. |
44 | | */ |
45 | | static int |
46 | | parse_num(const char *str, char **endp) |
47 | 0 | { |
48 | 0 | debug_decl(check_pattern, SUDO_DEBUG_UTIL); |
49 | 0 | const unsigned int lastval = INT_MAX / 10; |
50 | 0 | const unsigned int remainder = INT_MAX % 10; |
51 | 0 | unsigned int result = 0; |
52 | 0 | unsigned char ch; |
53 | |
|
54 | 0 | while ((ch = *str++) != '\0') { |
55 | 0 | if (ch == '\\' && isdigit((unsigned int)str[0])) |
56 | 0 | ch = *str++; |
57 | 0 | else if (!isdigit(ch)) |
58 | 0 | break; |
59 | 0 | ch -= '0'; |
60 | 0 | if (result > lastval || (result == lastval && ch > remainder)) { |
61 | 0 | result = -1; |
62 | 0 | break; |
63 | 0 | } |
64 | 0 | result *= 10; |
65 | 0 | result += ch; |
66 | 0 | } |
67 | 0 | *endp = (char *)(str - 1); |
68 | |
|
69 | 0 | debug_return_int(result); |
70 | 0 | } |
71 | | |
72 | | /* |
73 | | * Check pattern for invalid repetition sequences. |
74 | | * This is implementation-specific behavior, not all regcomp(3) forbid them. |
75 | | * Glibc allows it but uses excessive memory for repeated '+' ops. |
76 | | */ |
77 | | static int |
78 | | check_pattern(const char *pattern) |
79 | 0 | { |
80 | 0 | debug_decl(check_pattern, SUDO_DEBUG_UTIL); |
81 | 0 | const char *cp = pattern; |
82 | 0 | int b1, b2 = 0; |
83 | 0 | char ch, *ep, prev = '\0'; |
84 | |
|
85 | 0 | while ((ch = *cp++) != '\0') { |
86 | 0 | switch (ch) { |
87 | 0 | case '\\': |
88 | 0 | if (*cp != '\0') { |
89 | | /* Skip escaped character. */ |
90 | 0 | cp++; |
91 | 0 | prev = '\0'; |
92 | 0 | continue; |
93 | 0 | } |
94 | 0 | break; |
95 | 0 | case '?': |
96 | 0 | case '*': |
97 | 0 | case '+': |
98 | 0 | if (prev == '?' || prev == '*' || prev == '+' || prev == '{' ) { |
99 | | /* Invalid repetition operator. */ |
100 | 0 | debug_return_int(REG_BADRPT); |
101 | 0 | } |
102 | 0 | break; |
103 | 0 | case '{': |
104 | | /* |
105 | | * Try to match bound: {[0-9\\]*\?,[0-9\\]*} |
106 | | * GNU libc supports escaped digits and commas. |
107 | | */ |
108 | 0 | b1 = parse_num(cp, &ep); |
109 | 0 | switch (ep[0]) { |
110 | 0 | case '\\': |
111 | 0 | if (ep[1] != ',') |
112 | 0 | break; |
113 | 0 | ep++; |
114 | 0 | FALLTHROUGH; |
115 | 0 | case ',': |
116 | 0 | cp = ep + 1; |
117 | 0 | b2 = parse_num(cp, &ep); |
118 | 0 | break; |
119 | 0 | } |
120 | 0 | cp = ep; |
121 | 0 | if (*cp == '}') { |
122 | 0 | if (b1 < 0 || b1 > 255 || b2 < 0 || b2 > 255) { |
123 | | /* Invalid bound value. */ |
124 | 0 | debug_return_int(REG_BADBR); |
125 | 0 | } |
126 | 0 | if (prev == '?' || prev == '*' || prev == '+' || prev == '{' ) { |
127 | | /* Invalid repetition operator. */ |
128 | 0 | debug_return_int(REG_BADRPT); |
129 | 0 | } |
130 | | /* Skip past '}', prev will be set to '{' below */ |
131 | 0 | cp++; |
132 | 0 | break; |
133 | 0 | } |
134 | 0 | prev = '\0'; |
135 | 0 | continue; |
136 | 0 | } |
137 | 0 | prev = ch; |
138 | 0 | } |
139 | | |
140 | 0 | debug_return_int(0); |
141 | 0 | } |
142 | | |
143 | | /* |
144 | | * Wrapper around regcomp() that handles a regex starting with (?i). |
145 | | * Avoid using regex_t in the function args so we don't need to |
146 | | * include regex.h everywhere. |
147 | | */ |
148 | | bool |
149 | | sudo_regex_compile_v1(void *v, const char *pattern, const char **errstr) |
150 | 0 | { |
151 | 0 | int errcode, cflags = REG_EXTENDED|REG_NOSUB; |
152 | 0 | regex_t *preg; |
153 | 0 | char *copy = NULL; |
154 | 0 | const char *cp; |
155 | 0 | regex_t rebuf; |
156 | 0 | debug_decl(regex_compile, SUDO_DEBUG_UTIL); |
157 | | |
158 | | /* Some callers just want to check the validity of the pattern. */ |
159 | 0 | preg = v ? v : &rebuf; |
160 | | |
161 | | /* Limit the length of regular expressions to avoid fuzzer issues. */ |
162 | 0 | if (strlen(pattern) > 1024) { |
163 | 0 | *errstr = N_("regular expression too large"); |
164 | 0 | debug_return_bool(false); |
165 | 0 | } |
166 | | |
167 | | /* Check for (?i) to enable case-insensitive matching. */ |
168 | 0 | cp = pattern[0] == '^' ? pattern + 1 : pattern; |
169 | 0 | if (strncmp(cp, "(?i)", 4) == 0) { |
170 | 0 | cflags |= REG_ICASE; |
171 | 0 | copy = strdup(pattern + 4); |
172 | 0 | if (copy == NULL) { |
173 | 0 | *errstr = N_("unable to allocate memory"); |
174 | 0 | debug_return_bool(false); |
175 | 0 | } |
176 | 0 | if (pattern[0] == '^') |
177 | 0 | copy[0] = '^'; |
178 | 0 | pattern = copy; |
179 | 0 | } |
180 | | |
181 | 0 | errcode = check_pattern(pattern); |
182 | 0 | if (errcode == 0) |
183 | 0 | errcode = regcomp(preg, pattern, cflags); |
184 | 0 | if (errcode == 0) { |
185 | 0 | if (preg == &rebuf) |
186 | 0 | regfree(&rebuf); |
187 | 0 | } else { |
188 | 0 | regerror(errcode, preg, errbuf, sizeof(errbuf)); |
189 | 0 | *errstr = errbuf; |
190 | 0 | } |
191 | 0 | free(copy); |
192 | |
|
193 | 0 | debug_return_bool(errcode == 0); |
194 | 0 | } |