X-Git-Url: http://www.privoxy.org/gitweb/?p=privoxy.git;a=blobdiff_plain;f=urlmatch.c;h=1c4963a2d89755a922899bde5642e2b8d0028a05;hp=2f39258879c7e31a885a1ccf5f4eb617853a09a1;hb=11050ad03e957405c4ee436b28ead7b12e98bf25;hpb=460cbef0b136baa4ca9d590884503859dac16661 diff --git a/urlmatch.c b/urlmatch.c index 2f392588..1c4963a2 100644 --- a/urlmatch.c +++ b/urlmatch.c @@ -1,4 +1,4 @@ -const char urlmatch_rcs[] = "$Id: urlmatch.c,v 1.14 2007/01/06 14:23:56 fabiankeil Exp $"; +const char urlmatch_rcs[] = "$Id: urlmatch.c,v 1.27 2008/04/08 15:44:33 fabiankeil Exp $"; /********************************************************************* * * File : $Source: /cvsroot/ijbswa/current/urlmatch.c,v $ @@ -6,7 +6,7 @@ const char urlmatch_rcs[] = "$Id: urlmatch.c,v 1.14 2007/01/06 14:23:56 fabianke * Purpose : Declares functions to match URLs against URL * patterns. * - * Copyright : Written by and Copyright (C) 2001-2003, 2006-2007 the SourceForge + * Copyright : Written by and Copyright (C) 2001-2003, 2006-2008 the SourceForge * Privoxy team. http://www.privoxy.org/ * * Based on the Internet Junkbuster originally written @@ -33,6 +33,58 @@ const char urlmatch_rcs[] = "$Id: urlmatch.c,v 1.14 2007/01/06 14:23:56 fabianke * * Revisions : * $Log: urlmatch.c,v $ + * 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 + * of using a screen-long OR chain. + * + * Revision 1.17 2007/04/15 16:39:21 fabiankeil + * Introduce tags as alternative way to specify which + * actions apply to a request. At the moment tags can be + * created based on client and server headers. + * + * Revision 1.16 2007/02/13 13:59:24 fabiankeil + * Remove redundant log message. + * + * Revision 1.15 2007/01/28 16:11:23 fabiankeil + * Accept WebDAV methods for subversion + * in parse_http_request(). Closes FR 1581425. + * * Revision 1.14 2007/01/06 14:23:56 fabiankeil * Fix gcc43 warnings. Mark *csp as immutable * for parse_http_url() and url_match(). @@ -183,6 +235,7 @@ void free_http_request(struct http_request *http) http->dcount = 0; } + /********************************************************************* * * Function : init_domain_components @@ -247,6 +300,7 @@ jb_err init_domain_components(struct http_request *http) return JB_ERR_OK; } + /********************************************************************* * * Function : parse_http_url @@ -450,6 +504,60 @@ jb_err parse_http_url(const char * url, } +/********************************************************************* + * + * Function : unknown_method + * + * Description : Checks whether a method is unknown. + * + * Parameters : + * 1 : method = points to a http method + * + * Returns : TRUE if it's unknown, FALSE otherwise. + * + *********************************************************************/ +static int unknown_method(const char *method) +{ + static const char *known_http_methods[] = { + /* Basic HTTP request type */ + "GET", "HEAD", "POST", "PUT", "DELETE", "OPTIONS", "TRACE", "CONNECT", + /* webDAV extensions (RFC2518) */ + "PROPFIND", "PROPPATCH", "MOVE", "COPY", "MKCOL", "LOCK", "UNLOCK", + /* + * Microsoft webDAV extension for Exchange 2000. See: + * http://lists.w3.org/Archives/Public/w3c-dist-auth/2002JanMar/0001.html + * http://msdn.microsoft.com/library/en-us/wss/wss/_webdav_methods.asp + */ + "BCOPY", "BMOVE", "BDELETE", "BPROPFIND", "BPROPPATCH", + /* + * Another Microsoft webDAV extension for Exchange 2000. See: + * http://systems.cs.colorado.edu/grunwald/MobileComputing/Papers/draft-cohen-gena-p-base-00.txt + * http://lists.w3.org/Archives/Public/w3c-dist-auth/2002JanMar/0001.html + * http://msdn.microsoft.com/library/en-us/wss/wss/_webdav_methods.asp + */ + "SUBSCRIBE", "UNSUBSCRIBE", "NOTIFY", "POLL", + /* + * Yet another WebDAV extension, this time for + * Web Distributed Authoring and Versioning (RFC3253) + */ + "VERSION-CONTROL", "REPORT", "CHECKOUT", "CHECKIN", "UNCHECKOUT", + "MKWORKSPACE", "UPDATE", "LABEL", "MERGE", "BASELINE-CONTROL", "MKACTIVITY", + }; + int i; + + for (i = 0; i < SZ(known_http_methods); i++) + { + if (0 == strcmpic(method, known_http_methods[i])) + { + return FALSE; + } + } + + return TRUE; + +} + + /********************************************************************* * * Function : parse_http_request @@ -473,7 +581,7 @@ jb_err parse_http_request(const char *req, const struct client_state *csp) { char *buf; - char *v[10]; + char *v[10]; /* XXX: Why 10? We should only need three. */ int n; jb_err err; int is_connect = 0; @@ -489,81 +597,31 @@ jb_err parse_http_request(const char *req, n = ssplit(buf, " \r\n", v, SZ(v), 1, 1); if (n != 3) { - log_error(LOG_LEVEL_ERROR, "Trouble ssplitting: %s", buf); free(buf); return JB_ERR_PARSE; } - /* this could be a CONNECT request */ - if (strcmpic(v[0], "connect") == 0) - { - /* Secure */ - is_connect = 1; - } - /* or it could be any other basic HTTP request type */ - else if ((0 == strcmpic(v[0], "get")) - || (0 == strcmpic(v[0], "head")) - || (0 == strcmpic(v[0], "post")) - || (0 == strcmpic(v[0], "put")) - || (0 == strcmpic(v[0], "delete")) - || (0 == strcmpic(v[0], "options")) - || (0 == strcmpic(v[0], "trace")) - - /* or a webDAV extension (RFC2518) */ - || (0 == strcmpic(v[0], "propfind")) - || (0 == strcmpic(v[0], "proppatch")) - || (0 == strcmpic(v[0], "move")) - || (0 == strcmpic(v[0], "copy")) - || (0 == strcmpic(v[0], "mkcol")) - || (0 == strcmpic(v[0], "lock")) - || (0 == strcmpic(v[0], "unlock")) - - /* Or a Microsoft webDAV extension for Exchange 2000. See: */ - /* http://lists.w3.org/Archives/Public/w3c-dist-auth/2002JanMar/0001.html */ - /* http://msdn.microsoft.com/library/en-us/wss/wss/_webdav_methods.asp */ - || (0 == strcmpic(v[0], "bcopy")) - || (0 == strcmpic(v[0], "bmove")) - || (0 == strcmpic(v[0], "bdelete")) - || (0 == strcmpic(v[0], "bpropfind")) - || (0 == strcmpic(v[0], "bproppatch")) - - /* Or another Microsoft webDAV extension for Exchange 2000. See: */ - /* http://systems.cs.colorado.edu/grunwald/MobileComputing/Papers/draft-cohen-gena-p-base-00.txt */ - /* http://lists.w3.org/Archives/Public/w3c-dist-auth/2002JanMar/0001.html */ - /* http://msdn.microsoft.com/library/en-us/wss/wss/_webdav_methods.asp */ - || (0 == strcmpic(v[0], "subscribe")) - || (0 == strcmpic(v[0], "unsubscribe")) - || (0 == strcmpic(v[0], "notify")) - || (0 == strcmpic(v[0], "poll")) - - /* - * Or yet another WebDAV extension, this time for - * Web Distributed Authoring and Versioning (RFC3253) - */ - || (0 == strcmpic(v[0], "version-control")) - || (0 == strcmpic(v[0], "report")) - || (0 == strcmpic(v[0], "checkout")) - || (0 == strcmpic(v[0], "checkin")) - || (0 == strcmpic(v[0], "uncheckout")) - || (0 == strcmpic(v[0], "mkworkspace")) - || (0 == strcmpic(v[0], "update")) - || (0 == strcmpic(v[0], "label")) - || (0 == strcmpic(v[0], "merge")) - || (0 == strcmpic(v[0], "baseline-control")) - || (0 == strcmpic(v[0], "mkactivity")) - ) - { - /* Normal */ - is_connect = 0; - } - else + /* + * Fail in case of unknown methods + * which we might not handle correctly. + * + * XXX: There should be a config option + * to forward requests with unknown methods + * anyway. Most of them don't need special + * steps. + */ + if (unknown_method(v[0])) { - /* Unknown HTTP method */ log_error(LOG_LEVEL_ERROR, "Unknown HTTP method detected: %s", v[0]); free(buf); return JB_ERR_PARSE; } + if (strcmpic(v[0], "CONNECT") == 0) + { + is_connect = 1; + } + err = parse_http_url(v[1], http, csp); if (err) { @@ -593,6 +651,148 @@ jb_err parse_http_request(const char *req, } +#ifdef FEATURE_EXTENDED_HOST_PATTERNS +/********************************************************************* + * + * Function : compile_host_pattern + * + * Description : Parses and compiles a PCRE 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) +{ + int errcode; + char rebuf[BUFFER_SIZE]; + + assert(host_pattern); + assert(strlen(host_pattern) < sizeof(rebuf) - 2); + + url->host_regex = zalloc(sizeof(*url->host_regex)); + if (NULL == url->host_regex) + { + free_url_spec(url); + return JB_ERR_MEMORY; + } + + snprintf(rebuf, sizeof(rebuf), "%s$", host_pattern); + + errcode = regcomp(url->host_regex, rebuf, + (REG_EXTENDED|REG_NOSUB|REG_ICASE)); + + if (errcode) + { + size_t errlen = regerror(errcode, url->host_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: %s", url->spec, rebuf); + free_url_spec(url); + + return JB_ERR_PARSE; + } + + return JB_ERR_OK; + +} + +#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 @@ -710,6 +910,7 @@ static int domain_match(const struct url_spec *pattern, const struct http_reques } } +#endif /* def FEATURE_EXTENDED_HOST_PATTERNS */ /********************************************************************* @@ -738,6 +939,9 @@ 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); @@ -755,30 +959,66 @@ jb_err create_url_spec(struct url_spec * url, const char * buf) return JB_ERR_MEMORY; } - if ((p = strchr(buf, '/')) != NULL) + /* Is it tag pattern? */ + if (0 == strncmpic("TAG:", url->spec, 4)) { - if (NULL == (url->path = strdup(p))) + if (NULL == (url->tag_regex = zalloc(sizeof(*url->tag_regex)))) { freez(url->spec); return JB_ERR_MEMORY; } - url->pathlen = strlen(url->path); + + /* 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); + free_url_spec(url); + + return JB_ERR_PARSE; + } + return JB_ERR_OK; + } + + /* Only reached for URL patterns */ + p = strchr(buf, '/'); + if (NULL != p) + { + if (*(p+1) != '\0') + { + url->path = strdup(p); + if (NULL == url->path) + { + free_url_spec(url); + return JB_ERR_MEMORY; + } + } + else + { + /* + * The path pattern is a single slash and can + * be ignored as it won't affect the result. + */ + assert(NULL == url->path); + url->path = NULL; + } *p = '\0'; } else { - url->path = NULL; - url->pathlen = 0; + url->path = NULL; } if (url->path) { - int errcode; - char rebuf[BUFFER_SIZE]; - if (NULL == (url->preg = zalloc(sizeof(*url->preg)))) { - freez(url->spec); - freez(url->path); + free_url_spec(url); return JB_ERR_MEMORY; } @@ -788,116 +1028,39 @@ jb_err create_url_spec(struct url_spec * url, const char * buf) (REG_EXTENDED|REG_NOSUB|REG_ICASE)); if (errcode) { - size_t errlen = regerror(errcode, - url->preg, rebuf, sizeof(rebuf)); + 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); + free_url_spec(url); return JB_ERR_PARSE; } } - if ((p = strchr(buf, ':')) == NULL) + + p = strchr(buf, ':'); + if (NULL != p) { - url->port = 0; + *p++ = '\0'; + url->port_list = strdup(p); + if (NULL == url->port_list) + { + return JB_ERR_MEMORY; + } } else { - *p++ = '\0'; - url->port = atoi(p); + 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. - */ + return compile_host_pattern(url, buf); } return JB_ERR_OK; @@ -923,14 +1086,29 @@ 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); + url->dcount = 0; +#endif /* ndef FEATURE_EXTENDED_HOST_PATTERNS */ freez(url->path); + freez(url->port_list); if (url->preg) { regfree(url->preg); freez(url->preg); } + if (url->tag_regex) + { + regfree(url->tag_regex); + freez(url->tag_regex); + } } @@ -944,17 +1122,109 @@ void free_url_spec(struct url_spec *url) * 1 : pattern = a URL pattern * 2 : url = URL to match * - * Returns : 0 iff the URL matches the pattern, else nonzero. + * Returns : Nonzero if the URL matches the pattern, else 0. * *********************************************************************/ int url_match(const struct url_spec *pattern, - const struct http_request *url) + const struct http_request *http) { - return ((pattern->port == 0) || (pattern->port == url->port)) - && ((pattern->dbuffer == NULL) || (domain_match(pattern, url) == 0)) - && ((pattern->path == NULL) || - (regexec(pattern->preg, url->path, 0, NULL, 0) == 0) - ); + /* 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->path) || (0 == regexec(pattern->preg, http->path, 0, NULL, 0))) + + if (pattern->tag_regex != NULL) + { + /* It's a tag pattern and shouldn't be matched against URLs */ + return 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); + + /* + * 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; + }