X-Git-Url: http://www.privoxy.org/gitweb/?p=privoxy.git;a=blobdiff_plain;f=urlmatch.c;h=0ddd57a788d7ae835f325553503f85941aafd3f6;hp=96d8877cd484d15f11d20dd008b681d978ecc9d6;hb=68e4d4cba2e37322878d79832aa5943fad679325;hpb=073ebc2e7fc82aad52174bf2a9ec354a3b382850 diff --git a/urlmatch.c b/urlmatch.c index 96d8877c..0ddd57a7 100644 --- a/urlmatch.c +++ b/urlmatch.c @@ -1,4 +1,4 @@ -const char urlmatch_rcs[] = "$Id: urlmatch.c,v 1.18 2007/07/30 16:42:21 fabiankeil Exp $"; +const char urlmatch_rcs[] = "$Id: urlmatch.c,v 1.33 2008/04/12 14:03:13 fabiankeil Exp $"; /********************************************************************* * * File : $Source: /cvsroot/ijbswa/current/urlmatch.c,v $ @@ -6,7 +6,7 @@ const char urlmatch_rcs[] = "$Id: urlmatch.c,v 1.18 2007/07/30 16:42:21 fabianke * Purpose : Declares functions to match URLs against URL * patterns. * - * Copyright : Written by and Copyright (C) 2001-2003, 2006-2007 the SourceForge + * Copyright : Written by and Copyright (C) 2001-2003, 2006-2008 the SourceForge * Privoxy team. http://www.privoxy.org/ * * Based on the Internet Junkbuster originally written @@ -33,6 +33,65 @@ const char urlmatch_rcs[] = "$Id: urlmatch.c,v 1.18 2007/07/30 16:42:21 fabianke * * Revisions : * $Log: urlmatch.c,v $ + * Revision 1.33 2008/04/12 14:03:13 fabiankeil + * Remove an obvious comment and improve another one. + * + * Revision 1.32 2008/04/12 12:38:06 fabiankeil + * Factor out duplicated code to compile host, path and tag patterns. + * + * Revision 1.31 2008/04/10 14:41:04 fabiankeil + * Ditch url_spec's path member now that it's no longer used. + * + * Revision 1.30 2008/04/10 04:24:24 fabiankeil + * Stop duplicating the plain text representation of the path regex + * (and keeping the copy around). Once the regex is compiled it's no + * longer useful. + * + * Revision 1.29 2008/04/10 04:17:56 fabiankeil + * In url_match(), check the right member for NULL when determining + * whether there's a path regex to execute. Looking for a plain-text + * representation works as well, but it looks "interesting" and that + * member will be removed soonish anyway. + * + * Revision 1.28 2008/04/08 16:07:39 fabiankeil + * Make it harder to mistake url_match()'s + * second parameter for an url_spec. + * + * Revision 1.27 2008/04/08 15:44:33 fabiankeil + * Save a bit of memory (and a few cpu cycles) by not bothering to + * compile slash-only path regexes that don't affect the result. + * + * Revision 1.26 2008/04/07 16:57:18 fabiankeil + * - Use free_url_spec() more consistently. + * - Let it reset url->dcount just in case. + * + * Revision 1.25 2008/04/06 15:18:38 fabiankeil + * Oh well, rename the --enable-pcre-host-patterns option to + * --enable-extended-host-patterns as it's not really PCRE syntax. + * + * Revision 1.24 2008/04/06 14:54:26 fabiankeil + * Use PCRE syntax in host patterns when configured + * with --enable-pcre-host-patterns. + * + * Revision 1.23 2008/04/05 12:19:20 fabiankeil + * Factor compile_host_pattern() out of create_url_spec(). + * + * Revision 1.22 2008/03/30 15:02:32 fabiankeil + * SZitify unknown_method(). + * + * Revision 1.21 2007/12/24 16:34:23 fabiankeil + * Band-aid (and micro-optimization) that makes it less likely to run out of + * stack space with overly-complex path patterns. Probably masks the problem + * reported by Lee in #1856679. Hohoho. + * + * Revision 1.20 2007/09/02 15:31:20 fabiankeil + * Move match_portlist() from filter.c to urlmatch.c. + * It's used for url matching, not for filtering. + * + * Revision 1.19 2007/09/02 13:42:11 fabiankeil + * - Allow port lists in url patterns. + * - Ditch unused url_spec member pathlen. + * * Revision 1.18 2007/07/30 16:42:21 fabiankeil * Move the method check into unknown_method() * and loop through the known methods instead @@ -166,13 +225,11 @@ const char urlmatch_rcs[] = "$Id: urlmatch.c,v 1.18 2007/07/30 16:42:21 fabianke #include "ssplit.h" #include "miscutil.h" #include "errlog.h" -/* - * XXX: only for match_portlist() which I will relocate soonish. - */ -#include "filters.h" const char urlmatch_h_rcs[] = URLMATCH_H_VERSION; +enum regex_anchoring {NO_ANCHORING, LEFT_ANCHORED, RIGHT_ANCHORED}; +static jb_err compile_host_pattern(struct url_spec *url, const char *host_pattern); /********************************************************************* * @@ -204,6 +261,7 @@ void free_http_request(struct http_request *http) http->dcount = 0; } + /********************************************************************* * * Function : init_domain_components @@ -268,6 +326,7 @@ jb_err init_domain_components(struct http_request *http) return JB_ERR_OK; } + /********************************************************************* * * Function : parse_http_url @@ -509,11 +568,10 @@ static int unknown_method(const char *method) */ "VERSION-CONTROL", "REPORT", "CHECKOUT", "CHECKIN", "UNCHECKOUT", "MKWORKSPACE", "UPDATE", "LABEL", "MERGE", "BASELINE-CONTROL", "MKACTIVITY", - NULL }; int i; - for (i = 0; NULL != known_http_methods[i]; i++) + for (i = 0; i < SZ(known_http_methods); i++) { if (0 == strcmpic(method, known_http_methods[i])) { @@ -619,6 +677,264 @@ jb_err parse_http_request(const char *req, } +/********************************************************************* + * + * Function : compile_pattern + * + * Description : Compiles a host, domain or TAG pattern. + * + * Parameters : + * 1 : pattern = The pattern to compile. + * 2 : anchoring = How the regex should be anchored. + * Can be either one of NO_ANCHORING, + * LEFT_ANCHORED or RIGHT_ANCHORED. + * 3 : url = In case of failures, the spec member is + * logged and the structure freed. + * 4 : regex = Where the compiled regex should be stored. + * + * Returns : JB_ERR_OK - Success + * JB_ERR_MEMORY - Out of memory + * JB_ERR_PARSE - Cannot parse regex + * + *********************************************************************/ +static jb_err compile_pattern(const char *pattern, enum regex_anchoring anchoring, + struct url_spec *url, regex_t **regex) +{ + int errcode; + char rebuf[BUFFER_SIZE]; + const char *fmt; + + assert(pattern); + assert(strlen(pattern) < sizeof(rebuf) - 2); + + if (pattern[0] == '\0') + { + *regex = NULL; + return JB_ERR_OK; + } + + switch (anchoring) + { + case NO_ANCHORING: + fmt = "%s"; + break; + case RIGHT_ANCHORED: + fmt = "%s$"; + break; + case LEFT_ANCHORED: + fmt = "^%s"; + break; + default: + log_error(LOG_LEVEL_FATAL, + "Invalid anchoring in compile_pattern %d", anchoring); + } + + *regex = zalloc(sizeof(**regex)); + if (NULL == *regex) + { + free_url_spec(url); + return JB_ERR_MEMORY; + } + + snprintf(rebuf, sizeof(rebuf), fmt, pattern); + + errcode = regcomp(*regex, rebuf, (REG_EXTENDED|REG_NOSUB|REG_ICASE)); + + if (errcode) + { + size_t errlen = regerror(errcode, *regex, rebuf, sizeof(rebuf)); + if (errlen > (sizeof(rebuf) - (size_t)1)) + { + errlen = sizeof(rebuf) - (size_t)1; + } + rebuf[errlen] = '\0'; + log_error(LOG_LEVEL_ERROR, "error compiling %s from %s: %s", + pattern, url->spec, rebuf); + free_url_spec(url); + + return JB_ERR_PARSE; + } + + return JB_ERR_OK; + +} + + +/********************************************************************* + * + * Function : compile_url_pattern + * + * Description : Compiles the three parts of an URL pattern. + * + * Parameters : + * 1 : url = Target url_spec to be filled in. + * 2 : buf = The url pattern to compile. Will be messed up. + * + * Returns : JB_ERR_OK - Success + * JB_ERR_MEMORY - Out of memory + * JB_ERR_PARSE - Cannot parse regex + * + *********************************************************************/ +static jb_err compile_url_pattern(struct url_spec *url, char *buf) +{ + char *p; + + p = strchr(buf, '/'); + if (NULL != p) + { + /* + * Only compile the regex if it consists of more than + * a single slash, otherwise it wouldn't affect the result. + */ + if (p[1] != '\0') + { + /* + * XXX: does it make sense to compile the slash at the beginning? + */ + jb_err err = compile_pattern(p, LEFT_ANCHORED, url, &url->preg); + + if (JB_ERR_OK != err) + { + return err; + } + } + *p = '\0'; + } + + p = strchr(buf, ':'); + if (NULL != p) + { + *p++ = '\0'; + url->port_list = strdup(p); + if (NULL == url->port_list) + { + return JB_ERR_MEMORY; + } + } + else + { + url->port_list = NULL; + } + + if (buf[0] != '\0') + { + return compile_host_pattern(url, buf); + } + + return JB_ERR_OK; + +} + + +#ifdef FEATURE_EXTENDED_HOST_PATTERNS +/********************************************************************* + * + * Function : compile_host_pattern + * + * Description : Parses and compiles a host pattern.. + * + * Parameters : + * 1 : url = Target url_spec to be filled in. + * 2 : host_pattern = Host pattern to compile. + * + * Returns : JB_ERR_OK - Success + * JB_ERR_MEMORY - Out of memory + * JB_ERR_PARSE - Cannot parse regex + * + *********************************************************************/ +static jb_err compile_host_pattern(struct url_spec *url, const char *host_pattern) +{ + return compile_pattern(host_pattern, RIGHT_ANCHORED, url, &url->host_regex); +} + +#else + +/********************************************************************* + * + * Function : compile_host_pattern + * + * Description : Parses and "compiles" an old-school host pattern. + * + * Parameters : + * 1 : url = Target url_spec to be filled in. + * 2 : host_pattern = Host pattern to parse. + * + * Returns : JB_ERR_OK - Success + * JB_ERR_MEMORY - Out of memory + * JB_ERR_PARSE - Cannot parse regex + * + *********************************************************************/ +static jb_err compile_host_pattern(struct url_spec *url, const char *host_pattern) +{ + char *v[150]; + size_t size; + char *p; + + /* + * Parse domain part + */ + if (host_pattern[strlen(host_pattern) - 1] == '.') + { + url->unanchored |= ANCHOR_RIGHT; + } + if (host_pattern[0] == '.') + { + url->unanchored |= ANCHOR_LEFT; + } + + /* + * Split domain into components + */ + url->dbuffer = strdup(host_pattern); + if (NULL == url->dbuffer) + { + free_url_spec(url); + return JB_ERR_MEMORY; + } + + /* + * Map to lower case + */ + for (p = url->dbuffer; *p ; p++) + { + *p = (char)tolower((int)(unsigned char)*p); + } + + /* + * Split the domain name into components + */ + url->dcount = ssplit(url->dbuffer, ".", v, SZ(v), 1, 1); + + if (url->dcount < 0) + { + free_url_spec(url); + return JB_ERR_MEMORY; + } + else if (url->dcount != 0) + { + /* + * Save a copy of the pointers in dvec + */ + size = (size_t)url->dcount * sizeof(*url->dvec); + + url->dvec = (char **)malloc(size); + if (NULL == url->dvec) + { + free_url_spec(url); + return JB_ERR_MEMORY; + } + + memcpy(url->dvec, v, size); + } + /* + * else dcount == 0 in which case we needn't do anything, + * since dvec will never be accessed and the pattern will + * match all domains. + */ + return JB_ERR_OK; +} + + /********************************************************************* * * Function : simple_domaincmp @@ -736,6 +1052,7 @@ static int domain_match(const struct url_spec *pattern, const struct http_reques } } +#endif /* def FEATURE_EXTENDED_HOST_PATTERNS */ /********************************************************************* @@ -753,7 +1070,7 @@ static int domain_match(const struct url_spec *pattern, const struct http_reques * function. If this function succeeds, the * buffer is copied to url->spec. If this * function fails, the contents of the buffer - * are lost forever. + * are lost forever. XXX: Why is this const? * * Returns : JB_ERR_OK - Success * JB_ERR_MEMORY - Out of memory @@ -763,23 +1080,14 @@ static int domain_match(const struct url_spec *pattern, const struct http_reques *********************************************************************/ jb_err create_url_spec(struct url_spec * url, const char * buf) { - char *p; - int errcode; - size_t errlen; - char rebuf[BUFFER_SIZE]; - assert(url); assert(buf); - /* - * Zero memory - */ memset(url, '\0', sizeof(*url)); - /* - * Save a copy of the orignal specification - */ - if ((url->spec = strdup(buf)) == NULL) + /* Remember the original specification for the CGI pages. */ + url->spec = strdup(buf); + if (NULL == url->spec) { return JB_ERR_MEMORY; } @@ -787,184 +1095,13 @@ jb_err create_url_spec(struct url_spec * url, const char * buf) /* Is it tag pattern? */ if (0 == strncmpic("TAG:", url->spec, 4)) { - if (NULL == (url->tag_regex = zalloc(sizeof(*url->tag_regex)))) - { - freez(url->spec); - return JB_ERR_MEMORY; - } - - /* buf + 4 to skip "TAG:" */ - errcode = regcomp(url->tag_regex, buf + 4, (REG_EXTENDED|REG_NOSUB|REG_ICASE)); - if (errcode) - { - errlen = regerror(errcode, url->preg, rebuf, sizeof(rebuf)); - if (errlen > (sizeof(rebuf) - 1)) - { - errlen = sizeof(rebuf) - 1; - } - rebuf[errlen] = '\0'; - - log_error(LOG_LEVEL_ERROR, "error compiling %s: %s", url->spec, rebuf); - - freez(url->spec); - regfree(url->tag_regex); - freez(url->tag_regex); - - return JB_ERR_PARSE; - } - return JB_ERR_OK; - } - - /* Only reached for URL patterns */ - p = strchr(buf, '/'); - if (NULL != p) - { - url->path = strdup(p); - if (NULL == url->path) - { - freez(url->spec); - return JB_ERR_MEMORY; - } - *p = '\0'; - } - else - { - url->path = NULL; - } - if (url->path) - { - if (NULL == (url->preg = zalloc(sizeof(*url->preg)))) - { - freez(url->spec); - freez(url->path); - return JB_ERR_MEMORY; - } - - snprintf(rebuf, sizeof(rebuf), "^(%s)", url->path); - - errcode = regcomp(url->preg, rebuf, - (REG_EXTENDED|REG_NOSUB|REG_ICASE)); - if (errcode) - { - errlen = regerror(errcode, url->preg, rebuf, sizeof(rebuf)); - - if (errlen > (sizeof(rebuf) - (size_t)1)) - { - errlen = sizeof(rebuf) - (size_t)1; - } - rebuf[errlen] = '\0'; - - log_error(LOG_LEVEL_ERROR, "error compiling %s: %s", - url->spec, rebuf); - - freez(url->spec); - freez(url->path); - regfree(url->preg); - freez(url->preg); - - return JB_ERR_PARSE; - } - } - - p = strchr(buf, ':'); - if (NULL != p) - { - *p++ = '\0'; - url->port_list = strdup(p); - if (NULL == url->port_list) - { - return JB_ERR_MEMORY; - } - } - else - { - url->port_list = NULL; - } - - if (buf[0] != '\0') - { - char *v[150]; - size_t size; - - /* - * Parse domain part - */ - if (buf[strlen(buf) - 1] == '.') - { - url->unanchored |= ANCHOR_RIGHT; - } - if (buf[0] == '.') - { - url->unanchored |= ANCHOR_LEFT; - } - - /* - * Split domain into components - */ - url->dbuffer = strdup(buf); - if (NULL == url->dbuffer) - { - freez(url->spec); - freez(url->path); - regfree(url->preg); - freez(url->preg); - return JB_ERR_MEMORY; - } - - /* - * Map to lower case - */ - for (p = url->dbuffer; *p ; p++) - { - *p = (char)tolower((int)(unsigned char)*p); - } - - /* - * Split the domain name into components - */ - url->dcount = ssplit(url->dbuffer, ".", v, SZ(v), 1, 1); - - if (url->dcount < 0) - { - freez(url->spec); - freez(url->path); - regfree(url->preg); - freez(url->preg); - freez(url->dbuffer); - url->dcount = 0; - return JB_ERR_MEMORY; - } - else if (url->dcount != 0) - { - - /* - * Save a copy of the pointers in dvec - */ - size = (size_t)url->dcount * sizeof(*url->dvec); - - url->dvec = (char **)malloc(size); - if (NULL == url->dvec) - { - freez(url->spec); - freez(url->path); - regfree(url->preg); - freez(url->preg); - freez(url->dbuffer); - url->dcount = 0; - return JB_ERR_MEMORY; - } - - memcpy(url->dvec, v, size); - } - /* - * else dcount == 0 in which case we needn't do anything, - * since dvec will never be accessed and the pattern will - * match all domains. - */ + /* The pattern starts with the first character after "TAG:" */ + const char *tag_pattern = buf + 4; + return compile_pattern(tag_pattern, NO_ANCHORING, url, &url->tag_regex); } - return JB_ERR_OK; - + /* If it isn't a tag pattern it must be a URL pattern. */ + return compile_url_pattern(url, (char *)buf); } @@ -986,9 +1123,17 @@ void free_url_spec(struct url_spec *url) if (url == NULL) return; freez(url->spec); +#ifdef FEATURE_EXTENDED_HOST_PATTERNS + if (url->host_regex) + { + regfree(url->host_regex); + freez(url->host_regex); + } +#else freez(url->dbuffer); freez(url->dvec); - freez(url->path); + url->dcount = 0; +#endif /* ndef FEATURE_EXTENDED_HOST_PATTERNS */ freez(url->port_list); if (url->preg) { @@ -1017,11 +1162,16 @@ void free_url_spec(struct url_spec *url) * *********************************************************************/ int url_match(const struct url_spec *pattern, - const struct http_request *url) + const struct http_request *http) { - int port_matches; - int domain_matches; - int path_matches; + /* XXX: these should probably be functions. */ +#define PORT_MATCHES ((NULL == pattern->port_list) || match_portlist(pattern->port_list, http->port)) +#ifdef FEATURE_EXTENDED_HOST_PATTERNS +#define DOMAIN_MATCHES ((NULL == pattern->host_regex) || (0 == regexec(pattern->host_regex, http->host, 0, NULL, 0))) +#else +#define DOMAIN_MATCHES ((NULL == pattern->dbuffer) || (0 == domain_match(pattern, http))) +#endif +#define PATH_MATCHES ((NULL == pattern->preg) || (0 == regexec(pattern->preg, http->path, 0, NULL, 0))) if (pattern->tag_regex != NULL) { @@ -1029,11 +1179,87 @@ int url_match(const struct url_spec *pattern, return 0; } - port_matches = (NULL == pattern->port_list) || match_portlist(pattern->port_list, url->port); - domain_matches = (NULL == pattern->dbuffer) || (0 == domain_match(pattern, url)); - path_matches = (NULL == pattern->path) || (0 == regexec(pattern->preg, url->path, 0, NULL, 0)); + return (PORT_MATCHES && DOMAIN_MATCHES && PATH_MATCHES); + +} + + +/********************************************************************* + * + * Function : match_portlist + * + * Description : Check if a given number is covered by a comma + * separated list of numbers and ranges (a,b-c,d,..) + * + * Parameters : + * 1 : portlist = String with list + * 2 : port = port to check + * + * Returns : 0 => no match + * 1 => match + * + *********************************************************************/ +int match_portlist(const char *portlist, int port) +{ + char *min, *max, *next, *portlist_copy; + + min = next = portlist_copy = strdup(portlist); - return (port_matches && domain_matches && path_matches); + /* + * Zero-terminate first item and remember offset for next + */ + if (NULL != (next = strchr(portlist_copy, (int) ','))) + { + *next++ = '\0'; + } + + /* + * Loop through all items, checking for match + */ + while(min) + { + if (NULL == (max = strchr(min, (int) '-'))) + { + /* + * No dash, check for equality + */ + if (port == atoi(min)) + { + free(portlist_copy); + return(1); + } + } + else + { + /* + * This is a range, so check if between min and max, + * or, if max was omitted, between min and 65K + */ + *max++ = '\0'; + if(port >= atoi(min) && port <= (atoi(max) ? atoi(max) : 65535)) + { + free(portlist_copy); + return(1); + } + + } + + /* + * Jump to next item + */ + min = next; + + /* + * Zero-terminate next item and remember offset for n+1 + */ + if ((NULL != next) && (NULL != (next = strchr(next, (int) ',')))) + { + *next++ = '\0'; + } + } + + free(portlist_copy); + return 0; }