X-Git-Url: http://www.privoxy.org/gitweb/?p=privoxy.git;a=blobdiff_plain;f=urlmatch.c;h=36699fc248471749d14cbd6badea70e78917b8c5;hp=418a949ac1e324454e39246eab73689e2150ab06;hb=8ff919206b9740c46e7fe93e01d62f594d14262e;hpb=452b882d3844f7d6ba93155010aea2c406161f8e diff --git a/urlmatch.c b/urlmatch.c index 418a949a..36699fc2 100644 --- a/urlmatch.c +++ b/urlmatch.c @@ -1,4 +1,3 @@ -const char urlmatch_rcs[] = "$Id: urlmatch.c,v 1.86 2015/12/27 12:47:17 fabiankeil Exp $"; /********************************************************************* * * File : $Source: /cvsroot/ijbswa/current/urlmatch.c,v $ @@ -7,7 +6,7 @@ const char urlmatch_rcs[] = "$Id: urlmatch.c,v 1.86 2015/12/27 12:47:17 fabianke * patterns. * * Copyright : Written by and Copyright (C) 2001-2014 - * the Privoxy team. http://www.privoxy.org/ + * the Privoxy team. https://www.privoxy.org/ * * Based on the Internet Junkbuster originally written * by and Copyright (C) 1997 Anonymous Coders and @@ -56,8 +55,6 @@ const char urlmatch_rcs[] = "$Id: urlmatch.c,v 1.86 2015/12/27 12:47:17 fabianke #include "miscutil.h" #include "errlog.h" -const char urlmatch_h_rcs[] = URLMATCH_H_VERSION; - enum regex_anchoring { NO_ANCHORING, @@ -65,7 +62,10 @@ enum regex_anchoring RIGHT_ANCHORED, RIGHT_ANCHORED_HOST }; -static jb_err compile_host_pattern(struct pattern_spec *url, const char *host_pattern); +static jb_err compile_vanilla_host_pattern(struct pattern_spec *url, const char *host_pattern); +#ifdef FEATURE_EXTENDED_HOST_PATTERNS +static jb_err compile_pcre_host_pattern(struct pattern_spec *url, const char *host_pattern); +#endif /********************************************************************* * @@ -90,17 +90,14 @@ void free_http_request(struct http_request *http) freez(http->url); freez(http->hostport); freez(http->path); - freez(http->ver); + freez(http->version); freez(http->host_ip_addr_str); -#ifndef FEATURE_EXTENDED_HOST_PATTERNS freez(http->dbuffer); freez(http->dvec); http->dcount = 0; -#endif } -#ifndef FEATURE_EXTENDED_HOST_PATTERNS /********************************************************************* * * Function : init_domain_components @@ -155,7 +152,6 @@ jb_err init_domain_components(struct http_request *http) return JB_ERR_OK; } -#endif /* ndef FEATURE_EXTENDED_HOST_PATTERNS */ /********************************************************************* @@ -294,12 +290,17 @@ jb_err parse_http_url(const char *url, struct http_request *http, int require_pr /* * Got a path. * - * NOTE: The following line ignores the path for HTTPS URLS. - * This means that you get consistent behaviour if you type a - * https URL in and it's parsed by the function. (When the - * URL is actually retrieved, SSL hides the path part). + * If FEATURE_HTTPS_INSPECTION isn't available, ignore the + * path for https URLs so that we get consistent behaviour + * if a https URL is parsed. When the URL is actually + * retrieved, https hides the path part. */ - http->path = strdup_or_die(http->ssl ? "/" : url_path); + http->path = strdup_or_die( +#ifndef FEATURE_HTTPS_INSPECTION + http->ssl ? "/" : +#endif + url_path + ); *url_path = '\0'; http->hostport = strdup_or_die(url_noproto); } @@ -406,12 +407,8 @@ jb_err parse_http_url(const char *url, struct http_request *http, int require_pr freez(buf); } -#ifdef FEATURE_EXTENDED_HOST_PATTERNS - return JB_ERR_OK; -#else /* Split domain name so we can compare it against wildcards */ return init_domain_components(http); -#endif /* def FEATURE_EXTENDED_HOST_PATTERNS */ } @@ -494,7 +491,7 @@ static int unknown_method(const char *method) * JB_ERR_PARSE if the HTTP version is unsupported * *********************************************************************/ -jb_err static normalize_http_version(char *http_version) +static jb_err normalize_http_version(char *http_version) { unsigned int major_version; unsigned int minor_version; @@ -590,7 +587,7 @@ jb_err parse_http_request(const char *req, struct http_request *http) */ http->cmd = strdup_or_die(req); http->gpc = strdup_or_die(v[0]); - http->ver = strdup_or_die(v[2]); + http->version = strdup_or_die(v[2]); http->ocmd = strdup_or_die(http->cmd); freez(buf); @@ -624,11 +621,11 @@ static jb_err compile_pattern(const char *pattern, enum regex_anchoring anchorin struct pattern_spec *url, regex_t **regex) { int errcode; - char rebuf[BUFFER_SIZE]; const char *fmt = NULL; + char *rebuf; + size_t rebuf_size; assert(pattern); - assert(strlen(pattern) < sizeof(rebuf) - 2); if (pattern[0] == '\0') { @@ -654,27 +651,30 @@ static jb_err compile_pattern(const char *pattern, enum regex_anchoring anchorin log_error(LOG_LEVEL_FATAL, "Invalid anchoring in compile_pattern %d", anchoring); } - + rebuf_size = strlen(pattern) + strlen(fmt); + rebuf = malloc_or_die(rebuf_size); *regex = zalloc_or_die(sizeof(**regex)); - snprintf(rebuf, sizeof(rebuf), fmt, pattern); + snprintf(rebuf, rebuf_size, 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)) + size_t errlen = regerror(errcode, *regex, rebuf, rebuf_size); + if (errlen > (rebuf_size - (size_t)1)) { - errlen = sizeof(rebuf) - (size_t)1; + errlen = rebuf_size - (size_t)1; } rebuf[errlen] = '\0'; log_error(LOG_LEVEL_ERROR, "error compiling %s from %s: %s", pattern, url->spec, rebuf); free_pattern_spec(url); + freez(rebuf); return JB_ERR_PARSE; } + freez(rebuf); return JB_ERR_OK; @@ -700,6 +700,20 @@ static jb_err compile_url_pattern(struct pattern_spec *url, char *buf) { char *p; +#ifdef FEATURE_EXTENDED_HOST_PATTERNS + const size_t prefix_length = 18; + if (strncmpic(buf, "PCRE-HOST-PATTERN:", prefix_length) == 0) + { + url->pattern.url_spec.host_regex_type = EXTENDED_HOST_PATTERN; + /* Overwrite the "PCRE-HOST-PATTERN:" prefix */ + memmove(buf, buf+prefix_length, strlen(buf+prefix_length)+1); + } + else + { + url->pattern.url_spec.host_regex_type = VANILLA_HOST_PATTERN; + } +#endif + p = strchr(buf, '/'); if (NULL != p) { @@ -761,7 +775,16 @@ static jb_err compile_url_pattern(struct pattern_spec *url, char *buf) if (buf[0] != '\0') { - return compile_host_pattern(url, buf); +#ifdef FEATURE_EXTENDED_HOST_PATTERNS + if (url->pattern.url_spec.host_regex_type == EXTENDED_HOST_PATTERN) + { + return compile_pcre_host_pattern(url, buf); + } + else +#endif + { + return compile_vanilla_host_pattern(url, buf); + } } return JB_ERR_OK; @@ -772,9 +795,9 @@ static jb_err compile_url_pattern(struct pattern_spec *url, char *buf) #ifdef FEATURE_EXTENDED_HOST_PATTERNS /********************************************************************* * - * Function : compile_host_pattern + * Function : compile_pcre_host_pattern * - * Description : Parses and compiles a host pattern. + * Description : Parses and compiles a pcre host pattern. * * Parameters : * 1 : url = Target pattern_spec to be filled in. @@ -785,16 +808,16 @@ static jb_err compile_url_pattern(struct pattern_spec *url, char *buf) * JB_ERR_PARSE - Cannot parse regex * *********************************************************************/ -static jb_err compile_host_pattern(struct pattern_spec *url, const char *host_pattern) +static jb_err compile_pcre_host_pattern(struct pattern_spec *url, const char *host_pattern) { return compile_pattern(host_pattern, RIGHT_ANCHORED_HOST, url, &url->pattern.url_spec.host_regex); } +#endif /* def FEATURE_EXTENDED_HOST_PATTERNS */ -#else /********************************************************************* * - * Function : compile_host_pattern + * Function : compile_vanilla_host_pattern * * Description : Parses and "compiles" an old-school host pattern. * @@ -806,7 +829,7 @@ static jb_err compile_host_pattern(struct pattern_spec *url, const char *host_pa * JB_ERR_PARSE - Cannot parse regex * *********************************************************************/ -static jb_err compile_host_pattern(struct pattern_spec *url, const char *host_pattern) +static jb_err compile_vanilla_host_pattern(struct pattern_spec *url, const char *host_pattern) { char *v[150]; size_t size; @@ -1122,7 +1145,6 @@ static int domain_match(const struct pattern_spec *p, const struct http_request } } -#endif /* def FEATURE_EXTENDED_HOST_PATTERNS */ /********************************************************************* @@ -1161,6 +1183,9 @@ jb_err create_pattern_spec(struct pattern_spec *pattern, char *buf) const unsigned flag; } tag_pattern[] = { { "TAG:", 4, PATTERN_SPEC_TAG_PATTERN}, + #ifdef FEATURE_CLIENT_TAGS + { "CLIENT-TAG:", 11, PATTERN_SPEC_CLIENT_TAG_PATTERN}, + #endif { "NO-REQUEST-TAG:", 15, PATTERN_SPEC_NO_REQUEST_TAG_PATTERN}, { "NO-RESPONSE-TAG:", 16, PATTERN_SPEC_NO_RESPONSE_TAG_PATTERN} }; @@ -1221,11 +1246,10 @@ void free_pattern_spec(struct pattern_spec *pattern) regfree(pattern->pattern.url_spec.host_regex); freez(pattern->pattern.url_spec.host_regex); } -#else +#endif /* def FEATURE_EXTENDED_HOST_PATTERNS */ freez(pattern->pattern.url_spec.dbuffer); freez(pattern->pattern.url_spec.dvec); pattern->pattern.url_spec.dcount = 0; -#endif /* ndef FEATURE_EXTENDED_HOST_PATTERNS */ freez(pattern->pattern.url_spec.port_list); if (pattern->pattern.url_spec.preg) { @@ -1277,11 +1301,13 @@ static int host_matches(const struct http_request *http, { assert(http->host != NULL); #ifdef FEATURE_EXTENDED_HOST_PATTERNS - return ((NULL == pattern->pattern.url_spec.host_regex) - || (0 == regexec(pattern->pattern.url_spec.host_regex, http->host, 0, NULL, 0))); -#else - return ((NULL == pattern->pattern.url_spec.dbuffer) || (0 == domain_match(pattern, http))); + if (pattern->pattern.url_spec.host_regex_type == EXTENDED_HOST_PATTERN) { + return ((NULL == pattern->pattern.url_spec.host_regex) + || (0 == regexec(pattern->pattern.url_spec.host_regex, + http->host, 0, NULL, 0))); + } #endif + return ((NULL == pattern->pattern.url_spec.dbuffer) || (0 == domain_match(pattern, http))); } @@ -1417,31 +1443,53 @@ int match_portlist(const char *portlist, int port) * * Function : parse_forwarder_address * - * Description : Parse out the host and port from a forwarder address. + * Description : Parse out the username, password, host and port from + * a forwarder address. * * Parameters : * 1 : address = The forwarder address to parse. * 2 : hostname = Used to return the hostname. NULL on error. * 3 : port = Used to return the port. Untouched if no port * is specified. + * 4 : username = Used to return the username if any. + * 5 : password = Used to return the password if any. * * Returns : JB_ERR_OK on success * JB_ERR_MEMORY on out of memory * JB_ERR_PARSE on malformed address. * *********************************************************************/ -jb_err parse_forwarder_address(char *address, char **hostname, int *port) +jb_err parse_forwarder_address(char *address, char **hostname, int *port, + char **username, char **password) { - char *p = address; + char *p; + char *tmp; + + tmp = *hostname = strdup_or_die(address); + + /* Parse username and password */ + if (username && password && (NULL != (p = strchr(*hostname, '@')))) + { + *p++ = '\0'; + *username = strdup_or_die(*hostname); + *hostname = strdup_or_die(p); - if ((*address == '[') && (NULL == strchr(address, ']'))) + if (NULL != (p = strchr(*username, ':'))) + { + *p++ = '\0'; + *password = strdup_or_die(p); + } + freez(tmp); + } + + /* Parse hostname and port */ + p = *hostname; + if ((*p == '[') && (NULL == strchr(p, ']'))) { /* XXX: Should do some more validity checks here. */ return JB_ERR_PARSE; } - *hostname = strdup_or_die(address); - if ((**hostname == '[') && (NULL != (p = strchr(*hostname, ']')))) { *p++ = '\0';