Willy Tarreau | baaee00 | 2006-06-26 02:48:02 +0200 | [diff] [blame] | 1 | /* |
Willy Tarreau | f4f0412 | 2010-01-28 18:10:50 +0100 | [diff] [blame] | 2 | * include/common/regex.h |
| 3 | * This file defines everything related to regular expressions. |
| 4 | * |
| 5 | * Copyright (C) 2000-2010 Willy Tarreau - w@1wt.eu |
| 6 | * |
| 7 | * This library is free software; you can redistribute it and/or |
| 8 | * modify it under the terms of the GNU Lesser General Public |
| 9 | * License as published by the Free Software Foundation, version 2.1 |
| 10 | * exclusively. |
| 11 | * |
| 12 | * This library is distributed in the hope that it will be useful, |
| 13 | * but WITHOUT ANY WARRANTY; without even the implied warranty of |
| 14 | * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU |
| 15 | * Lesser General Public License for more details. |
| 16 | * |
| 17 | * You should have received a copy of the GNU Lesser General Public |
| 18 | * License along with this library; if not, write to the Free Software |
| 19 | * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA |
| 20 | */ |
Willy Tarreau | baaee00 | 2006-06-26 02:48:02 +0200 | [diff] [blame] | 21 | |
Willy Tarreau | 2dd0d47 | 2006-06-29 17:53:05 +0200 | [diff] [blame] | 22 | #ifndef _COMMON_REGEX_H |
| 23 | #define _COMMON_REGEX_H |
Willy Tarreau | baaee00 | 2006-06-26 02:48:02 +0200 | [diff] [blame] | 24 | |
Thierry FOURNIER | e28f1ec | 2013-10-09 15:23:01 +0200 | [diff] [blame] | 25 | #include <stdlib.h> |
| 26 | |
Willy Tarreau | e3ba5f0 | 2006-06-29 18:54:54 +0200 | [diff] [blame] | 27 | #include <common/config.h> |
Willy Tarreau | baaee00 | 2006-06-26 02:48:02 +0200 | [diff] [blame] | 28 | |
| 29 | #ifdef USE_PCRE |
| 30 | #include <pcre.h> |
| 31 | #include <pcreposix.h> |
Thierry FOURNIER | 799c042 | 2013-12-06 20:36:20 +0100 | [diff] [blame] | 32 | #else /* no PCRE */ |
| 33 | #include <regex.h> |
| 34 | #endif |
Hiroaki Nakamura | 7035132 | 2013-01-13 15:00:42 +0900 | [diff] [blame] | 35 | |
Thierry FOURNIER | 799c042 | 2013-12-06 20:36:20 +0100 | [diff] [blame] | 36 | struct my_regex { |
| 37 | #ifdef USE_PCRE |
Thierry FOURNIER | 2620276 | 2014-06-18 11:50:51 +0200 | [diff] [blame] | 38 | pcre *reg; |
| 39 | pcre_extra *extra; |
Hiroaki Nakamura | 7035132 | 2013-01-13 15:00:42 +0900 | [diff] [blame] | 40 | #ifdef USE_PCRE_JIT |
Thierry FOURNIER | ed5a4ae | 2013-10-14 14:07:36 +0200 | [diff] [blame] | 41 | #ifndef PCRE_CONFIG_JIT |
| 42 | #error "The PCRE lib doesn't support JIT. Change your lib, or remove the option USE_PCRE_JIT." |
| 43 | #endif |
Hiroaki Nakamura | 7035132 | 2013-01-13 15:00:42 +0900 | [diff] [blame] | 44 | #endif |
Hiroaki Nakamura | 7035132 | 2013-01-13 15:00:42 +0900 | [diff] [blame] | 45 | #else /* no PCRE */ |
Thierry FOURNIER | 799c042 | 2013-12-06 20:36:20 +0100 | [diff] [blame] | 46 | regex_t regex; |
Willy Tarreau | baaee00 | 2006-06-26 02:48:02 +0200 | [diff] [blame] | 47 | #endif |
Thierry FOURNIER | 799c042 | 2013-12-06 20:36:20 +0100 | [diff] [blame] | 48 | }; |
Willy Tarreau | baaee00 | 2006-06-26 02:48:02 +0200 | [diff] [blame] | 49 | |
| 50 | /* what to do when a header matches a regex */ |
| 51 | #define ACT_ALLOW 0 /* allow the request */ |
| 52 | #define ACT_REPLACE 1 /* replace the matching header */ |
| 53 | #define ACT_REMOVE 2 /* remove the matching header */ |
| 54 | #define ACT_DENY 3 /* deny the request */ |
| 55 | #define ACT_PASS 4 /* pass this header without allowing or denying the request */ |
Willy Tarreau | b8750a8 | 2006-09-03 09:56:00 +0200 | [diff] [blame] | 56 | #define ACT_TARPIT 5 /* tarpit the connection matching this request */ |
Willy Tarreau | a496b60 | 2006-12-17 23:15:24 +0100 | [diff] [blame] | 57 | #define ACT_SETBE 6 /* switch the backend */ |
Willy Tarreau | baaee00 | 2006-06-26 02:48:02 +0200 | [diff] [blame] | 58 | |
| 59 | struct hdr_exp { |
| 60 | struct hdr_exp *next; |
Thierry FOURNIER | 09af0d6 | 2014-06-18 11:35:54 +0200 | [diff] [blame] | 61 | struct my_regex *preg; /* expression to look for */ |
Willy Tarreau | baaee00 | 2006-06-26 02:48:02 +0200 | [diff] [blame] | 62 | int action; /* ACT_ALLOW, ACT_REPLACE, ACT_REMOVE, ACT_DENY */ |
Willy Tarreau | b17916e | 2006-10-15 15:17:57 +0200 | [diff] [blame] | 63 | const char *replace; /* expression to set instead */ |
Willy Tarreau | f4f0412 | 2010-01-28 18:10:50 +0100 | [diff] [blame] | 64 | void *cond; /* a possible condition or NULL */ |
Willy Tarreau | baaee00 | 2006-06-26 02:48:02 +0200 | [diff] [blame] | 65 | }; |
| 66 | |
| 67 | extern regmatch_t pmatch[MAX_MATCH]; |
| 68 | |
Thierry FOURNIER | ed5a4ae | 2013-10-14 14:07:36 +0200 | [diff] [blame] | 69 | /* "str" is the string that contain the regex to compile. |
| 70 | * "regex" is preallocated memory. After the execution of this function, this |
| 71 | * struct contain the compiled regex. |
| 72 | * "cs" is the case sensitive flag. If cs is true, case sensitive is enabled. |
| 73 | * "cap" is capture flag. If cap if true the regex can capture into |
| 74 | * parenthesis strings. |
| 75 | * "err" is the standar error message pointer. |
| 76 | * |
| 77 | * The function return 1 is succes case, else return 0 and err is filled. |
| 78 | */ |
Thierry FOURNIER | 799c042 | 2013-12-06 20:36:20 +0100 | [diff] [blame] | 79 | int regex_comp(const char *str, struct my_regex *regex, int cs, int cap, char **err); |
Willy Tarreau | c874653 | 2014-05-28 23:05:07 +0200 | [diff] [blame] | 80 | int exp_replace(char *dst, unsigned int dst_size, char *src, const char *str, const regmatch_t *matches); |
Willy Tarreau | b17916e | 2006-10-15 15:17:57 +0200 | [diff] [blame] | 81 | const char *check_replace_string(const char *str); |
Thierry FOURNIER | 09af0d6 | 2014-06-18 11:35:54 +0200 | [diff] [blame] | 82 | const char *chain_regex(struct hdr_exp **head, struct my_regex *preg, |
Willy Tarreau | f4f0412 | 2010-01-28 18:10:50 +0100 | [diff] [blame] | 83 | int action, const char *replace, void *cond); |
Willy Tarreau | baaee00 | 2006-06-26 02:48:02 +0200 | [diff] [blame] | 84 | |
Thierry FOURNIER | b8f980c | 2014-06-11 13:59:05 +0200 | [diff] [blame] | 85 | /* If the function doesn't match, it returns false, else it returns true. |
| 86 | */ |
| 87 | static inline int regex_exec(const struct my_regex *preg, char *subject) { |
Thierry FOURNIER | 2620276 | 2014-06-18 11:50:51 +0200 | [diff] [blame] | 88 | #if defined(USE_PCRE) || defined(USE_PCRE_JIT) |
Thierry FOURNIER | b8f980c | 2014-06-11 13:59:05 +0200 | [diff] [blame] | 89 | if (pcre_exec(preg->reg, preg->extra, subject, strlen(subject), 0, 0, NULL, 0) < 0) |
| 90 | return 0; |
| 91 | return 1; |
| 92 | #else |
| 93 | int match; |
| 94 | match = regexec(&preg->regex, subject, 0, NULL, 0); |
| 95 | if (match == REG_NOMATCH) |
| 96 | return 0; |
| 97 | return 1; |
| 98 | #endif |
| 99 | } |
| 100 | |
Thierry FOURNIER | ef37a66 | 2013-10-15 13:41:44 +0200 | [diff] [blame] | 101 | /* Note that <subject> MUST be at least <length+1> characters long and must |
| 102 | * be writable because the function will temporarily force a zero past the |
| 103 | * last character. |
Thierry FOURNIER | b8f980c | 2014-06-11 13:59:05 +0200 | [diff] [blame] | 104 | * |
| 105 | * If the function doesn't match, it returns false, else it returns true. |
Thierry FOURNIER | ef37a66 | 2013-10-15 13:41:44 +0200 | [diff] [blame] | 106 | */ |
Thierry FOURNIER | b8f980c | 2014-06-11 13:59:05 +0200 | [diff] [blame] | 107 | static inline int regex_exec2(const struct my_regex *preg, char *subject, int length) { |
Thierry FOURNIER | 2620276 | 2014-06-18 11:50:51 +0200 | [diff] [blame] | 108 | #if defined(USE_PCRE) || defined(USE_PCRE_JIT) |
Thierry FOURNIER | b8f980c | 2014-06-11 13:59:05 +0200 | [diff] [blame] | 109 | if (pcre_exec(preg->reg, preg->extra, subject, length, 0, 0, NULL, 0) < 0) |
| 110 | return 0; |
| 111 | return 1; |
Hiroaki Nakamura | 7035132 | 2013-01-13 15:00:42 +0900 | [diff] [blame] | 112 | #else |
Thierry FOURNIER | ef37a66 | 2013-10-15 13:41:44 +0200 | [diff] [blame] | 113 | int match; |
| 114 | char old_char = subject[length]; |
| 115 | subject[length] = 0; |
Thierry FOURNIER | 799c042 | 2013-12-06 20:36:20 +0100 | [diff] [blame] | 116 | match = regexec(&preg->regex, subject, 0, NULL, 0); |
Thierry FOURNIER | ef37a66 | 2013-10-15 13:41:44 +0200 | [diff] [blame] | 117 | subject[length] = old_char; |
Thierry FOURNIER | b8f980c | 2014-06-11 13:59:05 +0200 | [diff] [blame] | 118 | if (match == REG_NOMATCH) |
| 119 | return 0; |
| 120 | return 1; |
Hiroaki Nakamura | 7035132 | 2013-01-13 15:00:42 +0900 | [diff] [blame] | 121 | #endif |
| 122 | } |
| 123 | |
Thierry FOURNIER | b8f980c | 2014-06-11 13:59:05 +0200 | [diff] [blame] | 124 | int regex_exec_match(const struct my_regex *preg, const char *subject, |
| 125 | size_t nmatch, regmatch_t pmatch[]); |
| 126 | int regex_exec_match2(const struct my_regex *preg, char *subject, int length, |
| 127 | size_t nmatch, regmatch_t pmatch[]); |
| 128 | |
Thierry FOURNIER | 799c042 | 2013-12-06 20:36:20 +0100 | [diff] [blame] | 129 | static inline void regex_free(struct my_regex *preg) { |
Thierry FOURNIER | 2620276 | 2014-06-18 11:50:51 +0200 | [diff] [blame] | 130 | #if defined(USE_PCRE) || defined(USE_PCRE_JIT) |
| 131 | pcre_free(preg->reg); |
Hiroaki Nakamura | 7035132 | 2013-01-13 15:00:42 +0900 | [diff] [blame] | 132 | #ifdef USE_PCRE_JIT |
| 133 | pcre_free_study(preg->extra); |
Thierry FOURNIER | 2620276 | 2014-06-18 11:50:51 +0200 | [diff] [blame] | 134 | #endif |
Hiroaki Nakamura | 7035132 | 2013-01-13 15:00:42 +0900 | [diff] [blame] | 135 | #else |
Thierry FOURNIER | 799c042 | 2013-12-06 20:36:20 +0100 | [diff] [blame] | 136 | regfree(&preg->regex); |
Hiroaki Nakamura | 7035132 | 2013-01-13 15:00:42 +0900 | [diff] [blame] | 137 | #endif |
| 138 | } |
| 139 | |
Willy Tarreau | 2dd0d47 | 2006-06-29 17:53:05 +0200 | [diff] [blame] | 140 | #endif /* _COMMON_REGEX_H */ |
Willy Tarreau | baaee00 | 2006-06-26 02:48:02 +0200 | [diff] [blame] | 141 | |
| 142 | /* |
| 143 | * Local variables: |
| 144 | * c-indent-level: 8 |
| 145 | * c-basic-offset: 8 |
| 146 | * End: |
| 147 | */ |