1*4882a593Smuzhiyun /* 2*4882a593Smuzhiyun * Copyright (c) 2004-2005 Sergey Lyubka <valenok@gmail.com> 3*4882a593Smuzhiyun * All rights reserved 4*4882a593Smuzhiyun * 5*4882a593Smuzhiyun * "THE BEER-WARE LICENSE" (Revision 42): 6*4882a593Smuzhiyun * Sergey Lyubka wrote this file. As long as you retain this notice you 7*4882a593Smuzhiyun * can do whatever you want with this stuff. If we meet some day, and you think 8*4882a593Smuzhiyun * this stuff is worth it, you can buy me a beer in return. 9*4882a593Smuzhiyun */ 10*4882a593Smuzhiyun 11*4882a593Smuzhiyun /* 12*4882a593Smuzhiyun * Downloaded Sat Nov 5 17:42:08 CET 2011 at 13*4882a593Smuzhiyun * http://slre.sourceforge.net/1.0/slre.h 14*4882a593Smuzhiyun */ 15*4882a593Smuzhiyun 16*4882a593Smuzhiyun /* 17*4882a593Smuzhiyun * This is a regular expression library that implements a subset of Perl RE. 18*4882a593Smuzhiyun * Please refer to http://slre.sourceforge.net for detailed description. 19*4882a593Smuzhiyun * 20*4882a593Smuzhiyun * Usage example (parsing HTTP request): 21*4882a593Smuzhiyun * 22*4882a593Smuzhiyun * struct slre slre; 23*4882a593Smuzhiyun * struct cap captures[4 + 1]; // Number of braket pairs + 1 24*4882a593Smuzhiyun * ... 25*4882a593Smuzhiyun * 26*4882a593Smuzhiyun * slre_compile(&slre,"^(GET|POST) (\S+) HTTP/(\S+?)\r\n"); 27*4882a593Smuzhiyun * 28*4882a593Smuzhiyun * if (slre_match(&slre, buf, len, captures)) { 29*4882a593Smuzhiyun * printf("Request line length: %d\n", captures[0].len); 30*4882a593Smuzhiyun * printf("Method: %.*s\n", captures[1].len, captures[1].ptr); 31*4882a593Smuzhiyun * printf("URI: %.*s\n", captures[2].len, captures[2].ptr); 32*4882a593Smuzhiyun * } 33*4882a593Smuzhiyun * 34*4882a593Smuzhiyun * Supported syntax: 35*4882a593Smuzhiyun * ^ Match beginning of a buffer 36*4882a593Smuzhiyun * $ Match end of a buffer 37*4882a593Smuzhiyun * () Grouping and substring capturing 38*4882a593Smuzhiyun * [...] Match any character from set 39*4882a593Smuzhiyun * [^...] Match any character but ones from set 40*4882a593Smuzhiyun * \s Match whitespace 41*4882a593Smuzhiyun * \S Match non-whitespace 42*4882a593Smuzhiyun * \d Match decimal digit 43*4882a593Smuzhiyun * \r Match carriage return 44*4882a593Smuzhiyun * \n Match newline 45*4882a593Smuzhiyun * + Match one or more times (greedy) 46*4882a593Smuzhiyun * +? Match one or more times (non-greedy) 47*4882a593Smuzhiyun * * Match zero or more times (greedy) 48*4882a593Smuzhiyun * *? Match zero or more times (non-greedy) 49*4882a593Smuzhiyun * ? Match zero or once 50*4882a593Smuzhiyun * \xDD Match byte with hex value 0xDD 51*4882a593Smuzhiyun * \meta Match one of the meta character: ^$().[*+?\ 52*4882a593Smuzhiyun */ 53*4882a593Smuzhiyun 54*4882a593Smuzhiyun #ifndef SLRE_HEADER_DEFINED 55*4882a593Smuzhiyun #define SLRE_HEADER_DEFINED 56*4882a593Smuzhiyun 57*4882a593Smuzhiyun /* 58*4882a593Smuzhiyun * Compiled regular expression 59*4882a593Smuzhiyun */ 60*4882a593Smuzhiyun struct slre { 61*4882a593Smuzhiyun unsigned char code[256]; 62*4882a593Smuzhiyun unsigned char data[256]; 63*4882a593Smuzhiyun int code_size; 64*4882a593Smuzhiyun int data_size; 65*4882a593Smuzhiyun int num_caps; /* Number of bracket pairs */ 66*4882a593Smuzhiyun int anchored; /* Must match from string start */ 67*4882a593Smuzhiyun const char *err_str; /* Error string */ 68*4882a593Smuzhiyun }; 69*4882a593Smuzhiyun 70*4882a593Smuzhiyun /* 71*4882a593Smuzhiyun * Captured substring 72*4882a593Smuzhiyun */ 73*4882a593Smuzhiyun struct cap { 74*4882a593Smuzhiyun const char *ptr; /* Pointer to the substring */ 75*4882a593Smuzhiyun int len; /* Substring length */ 76*4882a593Smuzhiyun }; 77*4882a593Smuzhiyun 78*4882a593Smuzhiyun /* 79*4882a593Smuzhiyun * Compile regular expression. If success, 1 is returned. 80*4882a593Smuzhiyun * If error, 0 is returned and slre.err_str points to the error message. 81*4882a593Smuzhiyun */ 82*4882a593Smuzhiyun int slre_compile(struct slre *, const char *re); 83*4882a593Smuzhiyun 84*4882a593Smuzhiyun /* 85*4882a593Smuzhiyun * Return 1 if match, 0 if no match. 86*4882a593Smuzhiyun * If `captured_substrings' array is not NULL, then it is filled with the 87*4882a593Smuzhiyun * values of captured substrings. captured_substrings[0] element is always 88*4882a593Smuzhiyun * a full matched substring. The round bracket captures start from 89*4882a593Smuzhiyun * captured_substrings[1]. 90*4882a593Smuzhiyun * It is assumed that the size of captured_substrings array is enough to 91*4882a593Smuzhiyun * hold all captures. The caller function must make sure it is! So, the 92*4882a593Smuzhiyun * array_size = number_of_round_bracket_pairs + 1 93*4882a593Smuzhiyun */ 94*4882a593Smuzhiyun int slre_match(const struct slre *, const char *buf, int buf_len, 95*4882a593Smuzhiyun struct cap *captured_substrings); 96*4882a593Smuzhiyun 97*4882a593Smuzhiyun #ifdef SLRE_TEST 98*4882a593Smuzhiyun void slre_dump(const struct slre *r, FILE *fp); 99*4882a593Smuzhiyun #endif /* SLRE_TEST */ 100*4882a593Smuzhiyun #endif /* SLRE_HEADER_DEFINED */ 101