X-Git-Url: http://www.privoxy.org/gitweb/?p=privoxy.git;a=blobdiff_plain;f=parsers.c;h=eef1fd613bd427503a26893df9e21663fe9afdcf;hp=4f51e0bfe08dfde7f77019fc1b955a7552828116;hb=eaf36cae45ade9aabc5651da1735766b692fcf15;hpb=1952a87fe55ceeff77c376d61e281df7aee11db7 diff --git a/parsers.c b/parsers.c index 4f51e0bf..eef1fd61 100644 --- a/parsers.c +++ b/parsers.c @@ -1,4 +1,4 @@ -const char parsers_rcs[] = "$Id: parsers.c,v 1.3 2001/05/20 01:21:20 jongfoster Exp $"; +const char parsers_rcs[] = "$Id: parsers.c,v 1.35 2001/10/09 22:39:21 jongfoster Exp $"; /********************************************************************* * * File : $Source: /cvsroot/ijbswa/current/parsers.c,v $ @@ -11,7 +11,7 @@ const char parsers_rcs[] = "$Id: parsers.c,v 1.3 2001/05/20 01:21:20 jongfoster * `client_x_forwarded_adder', `client_xtra_adder', * `content_type', `crumble', `destroy_list', `enlist', * `flush_socket', `free_http_request', `get_header', - * `list_to_text', `match', `parse_http_request', `sed', + * `list_to_text', `parse_http_request', `sed', * and `server_set_cookie'. * * Copyright : Written by and Copyright (C) 2001 the SourceForge @@ -41,6 +41,230 @@ const char parsers_rcs[] = "$Id: parsers.c,v 1.3 2001/05/20 01:21:20 jongfoster * * Revisions : * $Log: parsers.c,v $ + * Revision 1.35 2001/10/09 22:39:21 jongfoster + * assert.h is also required under Win32, so moving out of #ifndef _WIN32 + * block. + * + * Revision 1.34 2001/10/07 18:50:55 oes + * Added server_content_encoding, renamed server_transfer_encoding + * + * Revision 1.33 2001/10/07 18:04:49 oes + * Changed server_http11 to server_http and its pattern to "HTTP". + * Additional functionality: it now saves the HTTP status into + * csp->http->status and sets CT_TABOO for Status 206 (partial range) + * + * Revision 1.32 2001/10/07 15:43:28 oes + * Removed FEATURE_DENY_GZIP and replaced it with client_accept_encoding, + * client_te and client_accept_encoding_adder, triggered by the new + * +no-compression action. For HTTP/1.1 the Accept-Encoding header is + * changed to allow only identity and chunked, and the TE header is + * crunched. For HTTP/1.0, Accept-Encoding is crunched. + * + * parse_http_request no longer does anything than parsing. The rewriting + * of http->cmd and version mangling are gone. It now also recognizes + * the put and delete methods and saves the url in http->url. Removed + * unused variable. + * + * renamed content_type and content_length to have the server_ prefix + * + * server_content_type now only works if csp->content_type != CT_TABOO + * + * added server_transfer_encoding, which + * - Sets CT_TABOO to prohibit filtering if encoding compresses + * - Raises the CSP_FLAG_CHUNKED flag if Encoding is "chunked" + * - Change from "chunked" to "identity" if body was chunked + * but has been de-chunked for filtering. + * + * added server_content_md5 which crunches any Content-MD5 headers + * if the body was modified. + * + * made server_http11 conditional on +downgrade action + * + * Replaced 6 boolean members of csp with one bitmap (csp->flags) + * + * Revision 1.31 2001/10/05 14:25:02 oes + * Crumble Keep-Alive from Server + * + * Revision 1.30 2001/09/29 12:56:03 joergs + * IJB now changes HTTP/1.1 to HTTP/1.0 in requests and answers. + * + * Revision 1.29 2001/09/24 21:09:24 jongfoster + * Fixing 2 memory leaks that Guy spotted, where the paramater to + * enlist() was not being free()d. + * + * Revision 1.28 2001/09/22 16:32:28 jongfoster + * Removing unused #includes. + * + * Revision 1.27 2001/09/20 15:45:25 steudten + * + * add casting from size_t to int for printf() + * remove local variable shadow s2 + * + * Revision 1.26 2001/09/16 17:05:14 jongfoster + * Removing unused #include showarg.h + * + * Revision 1.25 2001/09/16 13:21:27 jongfoster + * Changes to use new list functions. + * + * Revision 1.24 2001/09/13 23:05:50 jongfoster + * Changing the string paramater to the header parsers a "const". + * + * Revision 1.23 2001/09/12 18:08:19 steudten + * + * In parse_http_request() header rewriting miss the host value, so + * from http://www.mydomain.com the result was just " / " not + * http://www.mydomain.com/ in case we forward. + * + * Revision 1.22 2001/09/10 10:58:53 oes + * Silenced compiler warnings + * + * Revision 1.21 2001/07/31 14:46:00 oes + * - Persistant connections now suppressed + * - sed() no longer appends empty header to csp->headers + * + * Revision 1.20 2001/07/30 22:08:36 jongfoster + * Tidying up #defines: + * - All feature #defines are now of the form FEATURE_xxx + * - Permanently turned off WIN_GUI_EDIT + * - Permanently turned on WEBDAV and SPLIT_PROXY_ARGS + * + * Revision 1.19 2001/07/25 17:21:54 oes + * client_uagent now saves copy of User-Agent: header value + * + * Revision 1.18 2001/07/13 14:02:46 oes + * - Included fix to repair broken HTTP requests that + * don't contain a path, not even '/'. + * - Removed all #ifdef PCRS + * - content_type now always inspected and classified as + * text, gif or other. + * - formatting / comments + * + * Revision 1.17 2001/06/29 21:45:41 oes + * Indentation, CRLF->LF, Tab-> Space + * + * Revision 1.16 2001/06/29 13:32:42 oes + * - Fixed a comment + * - Adapted free_http_request + * - Removed logentry from cancelled commit + * + * Revision 1.15 2001/06/03 19:12:38 oes + * deleted const struct interceptors + * + * Revision 1.14 2001/06/01 18:49:17 jongfoster + * Replaced "list_share" with "list" - the tiny memory gain was not + * worth the extra complexity. + * + * Revision 1.13 2001/05/31 21:30:33 jongfoster + * Removed list code - it's now in list.[ch] + * Renamed "permission" to "action", and changed many features + * to use the actions file rather than the global config. + * + * Revision 1.12 2001/05/31 17:33:13 oes + * + * CRLF -> LF + * + * Revision 1.11 2001/05/29 20:11:19 joergs + * '/ * inside comment' warning removed. + * + * Revision 1.10 2001/05/29 09:50:24 jongfoster + * Unified blocklist/imagelist/permissionslist. + * File format is still under discussion, but the internal changes + * are (mostly) done. + * + * Also modified interceptor behaviour: + * - We now intercept all URLs beginning with one of the following + * prefixes (and *only* these prefixes): + * * http://i.j.b/ + * * http://ijbswa.sf.net/config/ + * * http://ijbswa.sourceforge.net/config/ + * - New interceptors "home page" - go to http://i.j.b/ to see it. + * - Internal changes so that intercepted and fast redirect pages + * are not replaced with an image. + * - Interceptors now have the option to send a binary page direct + * to the client. (i.e. ijb-send-banner uses this) + * - Implemented show-url-info interceptor. (Which is why I needed + * the above interceptors changes - a typical URL is + * "http://i.j.b/show-url-info?url=www.somesite.com/banner.gif". + * The previous mechanism would not have intercepted that, and + * if it had been intercepted then it then it would have replaced + * it with an image.) + * + * Revision 1.9 2001/05/28 17:26:33 jongfoster + * Fixing segfault if last header was crunched. + * Fixing Windows build (snprintf() is _snprintf() under Win32, but we + * can use the cross-platform sprintf() instead.) + * + * Revision 1.8 2001/05/27 22:17:04 oes + * + * - re_process_buffer no longer writes the modified buffer + * to the client, which was very ugly. It now returns the + * buffer, which it is then written by chat. + * + * - content_length now adjusts the Content-Length: header + * for modified documents rather than crunch()ing it. + * (Length info in csp->content_length, which is 0 for + * unmodified documents) + * + * - For this to work, sed() is called twice when filtering. + * + * Revision 1.7 2001/05/27 13:19:06 oes + * Patched Joergs solution for the content-length in. + * + * Revision 1.6 2001/05/26 13:39:32 jongfoster + * Only crunches Content-Length header if applying RE filtering. + * Without this fix, Microsoft Windows Update wouldn't work. + * + * Revision 1.5 2001/05/26 00:28:36 jongfoster + * Automatic reloading of config file. + * Removed obsolete SIGHUP support (Unix) and Reload menu option (Win32). + * Most of the global variables have been moved to a new + * struct configuration_spec, accessed through csp->config->globalname + * Most of the globals remaining are used by the Win32 GUI. + * + * Revision 1.4 2001/05/22 18:46:04 oes + * + * - Enabled filtering banners by size rather than URL + * by adding patterns that replace all standard banner + * sizes with the "Junkbuster" gif to the re_filterfile + * + * - Enabled filtering WebBugs by providing a pattern + * which kills all 1x1 images + * + * - Added support for PCRE_UNGREEDY behaviour to pcrs, + * which is selected by the (nonstandard and therefore + * capital) letter 'U' in the option string. + * It causes the quantifiers to be ungreedy by default. + * Appending a ? turns back to greedy (!). + * + * - Added a new interceptor ijb-send-banner, which + * sends back the "Junkbuster" gif. Without imagelist or + * MSIE detection support, or if tinygif = 1, or the + * URL isn't recognized as an imageurl, a lame HTML + * explanation is sent instead. + * + * - Added new feature, which permits blocking remote + * script redirects and firing back a local redirect + * to the browser. + * The feature is conditionally compiled, i.e. it + * can be disabled with --disable-fast-redirects, + * plus it must be activated by a "fast-redirects" + * line in the config file, has its own log level + * and of course wants to be displayed by show-proxy-args + * Note: Boy, all the #ifdefs in 1001 locations and + * all the fumbling with configure.in and acconfig.h + * were *way* more work than the feature itself :-( + * + * - Because a generic redirect template was needed for + * this, tinygif = 3 now uses the same. + * + * - Moved GIFs, and other static HTTP response templates + * to project.h + * + * - Some minor fixes + * + * - Removed some >400 CRs again (Jon, you really worked + * a lot! ;-) + * * Revision 1.3 2001/05/20 01:21:20 jongfoster * Version 2.9.4 checkin. * - Merged popupfile and cookiefile, and added control over PCRS @@ -66,10 +290,14 @@ const char parsers_rcs[] = "$Id: parsers.c,v 1.3 2001/05/20 01:21:20 jongfoster #include "config.h" +#ifndef _WIN32 #include #include #include #include +#endif + +#include #include #ifndef _WIN32 @@ -77,12 +305,9 @@ const char parsers_rcs[] = "$Id: parsers.c,v 1.3 2001/05/20 01:21:20 jongfoster #endif #include "project.h" +#include "list.h" #include "parsers.h" #include "encode.h" -#include "filters.h" -#include "loaders.h" -#include "showargs.h" -#include "jcc.h" #include "ssplit.h" #include "errlog.h" #include "jbsockets.h" @@ -111,90 +336,51 @@ const struct parsers client_patterns[] = { { "from:", 5, client_from }, { "cookie:", 7, client_send_cookie }, { "x-forwarded-for:", 16, client_x_forwarded }, - { "proxy-connection:", 17, crumble }, -#ifdef DENY_GZIP - { "Accept-Encoding: gzip", 21, crumble }, -#endif /* def DENY_GZIP */ -#if defined(DETECT_MSIE_IMAGES) + { "Accept-Encoding:", 16, client_accept_encoding }, + { "TE:", 3, client_te }, +#if defined(FEATURE_IMAGE_DETECT_MSIE) { "Accept:", 7, client_accept }, -#endif /* defined(DETECT_MSIE_IMAGES) */ -#ifdef FORCE_LOAD - { "Host:", 5, client_host }, -#endif /* def FORCE_LOAD */ +#endif /* defined(FEATURE_IMAGE_DETECT_MSIE) */ + { "Host:", 5, crumble }, /* { "if-modified-since:", 18, crumble }, */ + { "Keep-Alive:", 11, crumble }, + { "connection:", 11, crumble }, + { "proxy-connection:", 17, crumble }, { NULL, 0, NULL } }; -const struct interceptors intercept_patterns[] = { - { "show-proxy-args", 14, show_proxy_args }, - { "ijb-send-banner", 14, ijb_send_banner }, -#ifdef TRUST_FILES - { "ij-untrusted-url", 15, ij_untrusted_url }, -#endif /* def TRUST_FILES */ - { NULL, 0, NULL } -}; const struct parsers server_patterns[] = { + { "HTTP", 4, server_http }, { "set-cookie:", 11, server_set_cookie }, { "connection:", 11, crumble }, -#if defined(PCRS) || defined(KILLPOPUPS) - { "Content-Type:", 13, content_type }, - { "Content-Length:", 15, crumble }, -#endif /* defined(PCRS) || defined(KILLPOPUPS) */ + { "Content-Type:", 13, server_content_type }, + { "Content-Length:", 15, server_content_length }, + { "Content-MD5:", 12, server_content_md5 }, + { "Content-Encoding:", 17, server_content_encoding }, + { "Transfer-Encoding:", 18, server_transfer_coding }, + { "Keep-Alive:", 11, crumble }, { NULL, 0, NULL } }; void (* const add_client_headers[])(struct client_state *) = { + client_host_adder, client_cookie_adder, client_x_forwarded_adder, client_xtra_adder, + client_accept_encoding_adder, + connection_close_adder, NULL }; void (* const add_server_headers[])(struct client_state *) = { + connection_close_adder, NULL }; -/********************************************************************* - * - * Function : match - * - * Description : Do a `strncmpic' on every pattern in pats. - * - * Parameters : - * 1 : buf = a string to match to a list of patterns - * 2 : pats = list of strings to compare against buf. - * - * Returns : Return the matching "struct parsers *", - * or NULL if no pattern matches. - * - *********************************************************************/ -static const struct parsers *match(char *buf, const struct parsers *pats) -{ - const struct parsers *v; - - if (buf == NULL) - { - /* hit me */ - log_error(LOG_LEVEL_ERROR, "NULL parameter to match()"); - return(NULL); - } - - for (v = pats; v->str ; v++) - { - if (strncmpic(buf, v->str, v->len) == 0) - { - return(v); - } - } - return(NULL); - -} - - /********************************************************************* * * Function : flush_socket @@ -351,125 +537,6 @@ char *get_header(struct client_state *csp) } -/********************************************************************* - * - * Function : enlist - * - * Description : Append a string into a specified string list. - * - * Parameters : - * 1 : h = pointer to list 'dummy' header - * 2 : s = string to add to the list - * - * Returns : N/A - * - *********************************************************************/ -void enlist(struct list *h, const char *s) -{ - struct list *n = (struct list *)malloc(sizeof(*n)); - struct list *l; - - if (n) - { - n->str = strdup(s); - n->next = NULL; - - if ((l = h->last)) - { - l->next = n; - } - else - { - h->next = n; - } - - h->last = n; - } - -} - - -/********************************************************************* - * - * Function : destroy_list - * - * Description : Destroy a string list (opposite of enlist) - * - * Parameters : - * 1 : h = pointer to list 'dummy' header - * - * Returns : N/A - * - *********************************************************************/ -void destroy_list(struct list *h) -{ - struct list *p, *n; - - for (p = h->next; p ; p = n) - { - n = p->next; - freez(p->str); - freez(p); - } - - memset(h, '\0', sizeof(*h)); - -} - - -/********************************************************************* - * - * Function : list_to_text - * - * Description : "Flaten" a string list into 1 long \r\n delimited string. - * - * Parameters : - * 1 : h = pointer to list 'dummy' header - * - * Returns : NULL on malloc error, else new long string. - * - *********************************************************************/ -static char *list_to_text(struct list *h) -{ - struct list *p; - char *ret = NULL; - char *s; - int size; - - size = 0; - - for (p = h->next; p ; p = p->next) - { - if (p->str) - { - size += strlen(p->str) + 2; - } - } - - if ((ret = (char *)malloc(size + 1)) == NULL) - { - return(NULL); - } - - ret[size] = '\0'; - - s = ret; - - for (p = h->next; p ; p = p->next) - { - if (p->str) - { - strcpy(s, p->str); - s += strlen(s); - *s++ = '\r'; *s++ = '\n'; - } - } - - return(ret); - -} - - /********************************************************************* * * Function : sed @@ -493,22 +560,27 @@ static char *list_to_text(struct list *h) *********************************************************************/ char *sed(const struct parsers pats[], void (* const more_headers[])(struct client_state *), struct client_state *csp) { - struct list *p; + struct list_entry *p; const struct parsers *v; char *hdr; void (* const *f)(); - for (p = csp->headers->next; p ; p = p->next) + for (v = pats; v->str ; v++) { - log_error(LOG_LEVEL_HEADER, "scan: %s", p->str); - - if ((v = match(p->str, pats))) + for (p = csp->headers->first; p ; p = p->next) { - hdr = v->parser(v, p->str, csp); - freez(p->str); - p->str = hdr; - } + /* Header crunch()ed in previous run? -> ignore */ + if (p->str == NULL) continue; + + if (v == pats) log_error(LOG_LEVEL_HEADER, "scan: %s", p->str); + if (strncmpic(p->str, v->str, v->len) == 0) + { + hdr = v->parser(v, p->str, csp); + freez((char *)p->str); /* FIXME: Yuck! patching a list...*/ + p->str = hdr; + } + } } /* place any additional headers on the csp->headers list */ @@ -517,9 +589,6 @@ char *sed(const struct parsers pats[], void (* const more_headers[])(struct clie (*f)(csp); } - /* add the blank line at the end of the header */ - enlist(csp->headers, ""); - hdr = list_to_text(csp->headers); return(hdr); @@ -541,12 +610,17 @@ char *sed(const struct parsers pats[], void (* const more_headers[])(struct clie *********************************************************************/ void free_http_request(struct http_request *http) { + assert(http); + freez(http->cmd); freez(http->gpc); freez(http->host); + freez(http->url); freez(http->hostport); freez(http->path); freez(http->ver); + freez(http->host_ip_addr_str); + freez(http->user_agent); } @@ -572,11 +646,9 @@ void parse_http_request(char *req, struct http_request *http, struct client_stat int n; memset(http, '\0', sizeof(*http)); - - http->cmd = strdup(req); + http->cmd = strdup(req); buf = strdup(req); - n = ssplit(buf, " \r\n", v, SZ(v), 1, 1); if (n == 3) @@ -590,41 +662,29 @@ void parse_http_request(char *req, struct http_request *http, struct client_stat http->ver = strdup(v[2]); } -#ifdef WEBDAV - -/* This next line is a little ugly, but it simplifies the if statement below. */ -/* Basically if using webDAV, we want the OR condition to use these too. */ - -/* - * by haroon - * These are the headers as defined in RFC2518 to add webDAV support - */ - -#define OR_WEBDAV || \ - (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")) - -#else /* No webDAV support is enabled. Provide an empty OR_WEBDAV macro. */ - -#define OR_WEBDAV - -#endif - - /* or it could be a GET or a POST (possibly webDAV too) */ - if ((strcmpic(v[0], "get") == 0) || - (strcmpic(v[0], "head") == 0) OR_WEBDAV || - (strcmpic(v[0], "post") == 0)) + /* or it could be any other basic HTTP request type */ + 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")) + + /* 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")) + ) { - http->ssl = 0; - http->gpc = strdup(v[0]); - url = v[1]; - http->ver = strdup(v[2]); + http->ssl = 0; + http->gpc = strdup(v[0]); + http->url = strdup(v[1]); + http->ver = strdup(v[2]); + url = v[1]; if (strncmpic(url, "http://", 7) == 0) { url += 7; @@ -638,11 +698,22 @@ void parse_http_request(char *req, struct http_request *http, struct client_stat url = NULL; } - if (url && (p = strchr(url, '/'))) + if (url) { - http->path = strdup(p); - *p = '\0'; - http->hostport = strdup(url); + if ((p = strchr(url, '/'))) + { + http->path = strdup(p); + *p = '\0'; + http->hostport = strdup(url); + } + /* + * Repair broken HTTP requests that don't contain a path + */ + else + { + http->path = strdup("/"); + http->hostport = strdup(url); + } } } } @@ -692,7 +763,7 @@ void parse_http_request(char *req, struct http_request *http, struct client_stat if (http->path == NULL) { - http->path = strdup(""); + http->path = strdup("/"); } } @@ -716,7 +787,7 @@ void parse_http_request(char *req, struct http_request *http, struct client_stat * Returns : Always NULL. * *********************************************************************/ -char *crumble(const struct parsers *v, char *s, struct client_state *csp) +char *crumble(const struct parsers *v, const char *s, struct client_state *csp) { log_error(LOG_LEVEL_HEADER, "crunch!"); return(NULL); @@ -724,13 +795,93 @@ char *crumble(const struct parsers *v, char *s, struct client_state *csp) } -#if defined(PCRS) || defined(KILLPOPUPS) +/********************************************************************* + * + * Function : server_content_type + * + * Description : Set the content-type for filterable types (text/.*, + * javascript and image/gif) unless filtering has been + * forbidden (CT_TABOO) while parsing earlier headers. + * + * Parameters : + * 1 : v = ignored + * 2 : s = header string we are "considering" + * 3 : csp = Current client state (buffers, headers, etc...) + * + * Returns : A duplicate string pointer to this header (ie. pass thru) + * + *********************************************************************/ +char *server_content_type(const struct parsers *v, const char *s, struct client_state *csp) +{ + if (csp->content_type != CT_TABOO) + { + if (strstr(s, " text/") || strstr(s, "application/x-javascript")) + csp->content_type = CT_TEXT; + else if (strstr(s, " image/gif")) + csp->content_type = CT_GIF; + else + csp->content_type = 0; + } + + return(strdup(s)); + +} + + +/********************************************************************* + * + * Function : server_transfer_coding + * + * Description : - Prohibit filtering (CT_TABOO) if transfer coding compresses + * - Raise the CSP_FLAG_CHUNKED flag if coding is "chunked" + * - Change from "chunked" to "identity" if body was chunked + * but has been de-chunked for filtering. + * + * Parameters : + * 1 : v = ignored + * 2 : s = header string we are "considering" + * 3 : csp = Current client state (buffers, headers, etc...) + * + * Returns : A duplicate string pointer to this header (ie. pass thru) + * + *********************************************************************/ +char *server_transfer_coding(const struct parsers *v, const char *s, struct client_state *csp) +{ + /* + * Turn off pcrs and gif filtering if body compressed + */ + if (strstr(s, "gzip") || strstr(s, "compress") || strstr(s, "deflate")) + { + csp->content_type = CT_TABOO; + } + + /* + * Raise flag if body chunked + */ + if (strstr(s, "chunked")) + { + csp->flags |= CSP_FLAG_CHUNKED; + + /* + * If the body was modified, it has been + * de-chunked first, so adjust the header: + */ + if (csp->flags & CSP_FLAG_MODIFIED) + { + return(strdup("Transfer-Encoding: identity")); + } + } + + return(strdup(s)); + +} + /********************************************************************* * - * Function : content_type + * Function : server_content_encoding * - * Description : Is this a text/* or javascript MIME Type? + * Description : Prohibit filtering (CT_TABOO) if content encoding compresses * * Parameters : * 1 : v = ignored @@ -740,20 +891,152 @@ char *crumble(const struct parsers *v, char *s, struct client_state *csp) * Returns : A duplicate string pointer to this header (ie. pass thru) * *********************************************************************/ -char *content_type(const struct parsers *v, char *s, struct client_state *csp) +char *server_content_encoding(const struct parsers *v, const char *s, struct client_state *csp) { - if (strstr (s, " text/") || strstr (s, "application/x-javascript")) - csp->is_text = 1; - else - csp->is_text = 0; + /* + * Turn off pcrs and gif filtering if body compressed + */ + if (strstr(s, "gzip") || strstr(s, "compress") || strstr(s, "deflate")) + { + csp->content_type = CT_TABOO; + } return(strdup(s)); } -#endif /* defined(PCRS) || defined(KILLPOPUPS) */ + +/********************************************************************* + * + * Function : server_content_length + * + * Description : Adjust Content-Length header if we modified + * the body. + * + * Parameters : + * 1 : v = ignored + * 2 : s = header string we are "considering" + * 3 : csp = Current client state (buffers, headers, etc...) + * + * Returns : A duplicate string pointer to this header (ie. pass thru) + * + *********************************************************************/ +char *server_content_length(const struct parsers *v, const char *s, struct client_state *csp) +{ + if (csp->content_length != 0) /* Content length has been modified */ + { + char * s2 = (char *) zalloc(100); + sprintf(s2, "Content-Length: %d", (int) csp->content_length); + + log_error(LOG_LEVEL_HEADER, "Adjust Content-Length to %d", (int) csp->content_length); + return(s2); + } + else + { + return(strdup(s)); + } + +} + + +/********************************************************************* + * + * Function : server_content_md5 + * + * Description : Crumble any Content-MD5 headers if the document was + * modified. FIXME: Should we re-compute instead? + * + * Parameters : + * 1 : v = ignored + * 2 : s = header string we are "considering" + * 3 : csp = Current client state (buffers, headers, etc...) + * + * Returns : A duplicate string pointer to this header (ie. pass thru) + * + *********************************************************************/ +char *server_content_md5(const struct parsers *v, const char *s, struct client_state *csp) +{ + if (csp->flags & CSP_FLAG_MODIFIED) + { + log_error(LOG_LEVEL_HEADER, "Crunching Content-MD5"); + return(NULL); + } + else + { + return(strdup(s)); + } + +} + + +/********************************************************************* + * + * Function : client_accept_encoding + * + * Description : Rewrite the client's Accept-Encoding header so that + * if doesn't allow compression, if the action applies. + * Note: For HTTP/1.0 the absence of the header is enough. + * + * Parameters : + * 1 : v = ignored + * 2 : s = header string we are "considering" + * 3 : csp = Current client state (buffers, headers, etc...) + * + * Returns : A copy of the client's original or the modified header. + * + *********************************************************************/ +char *client_accept_encoding(const struct parsers *v, const char *s, struct client_state *csp) +{ + if ((csp->action->flags & ACTION_NO_COMPRESSION) == 0) + { + return(strdup(s)); + } + else + { + log_error(LOG_LEVEL_HEADER, "Supressed offer to compress content"); + + if (!strcmpic(csp->http->ver, "HTTP/1.1")) + { + return(strdup("Accept-Encoding: identity;q=1.0, *;q=0")); + } + else + { + return(NULL); + } + } + +} +/********************************************************************* + * + * Function : client_te + * + * Description : Rewrite the client's TE header so that + * if doesn't allow compression, if the action applies. + * + * Parameters : + * 1 : v = ignored + * 2 : s = header string we are "considering" + * 3 : csp = Current client state (buffers, headers, etc...) + * + * Returns : A copy of the client's original or the modified header. + * + *********************************************************************/ +char *client_te(const struct parsers *v, const char *s, struct client_state *csp) +{ + if ((csp->action->flags & ACTION_NO_COMPRESSION) == 0) + { + return(strdup(s)); + } + else + { + log_error(LOG_LEVEL_HEADER, "Supressed offer to compress transfer"); + return(NULL); + } + +} + /********************************************************************* * * Function : client_referrer @@ -770,70 +1053,90 @@ char *content_type(const struct parsers *v, char *s, struct client_state *csp) * or modified header * *********************************************************************/ -char *client_referrer(const struct parsers *v, char *s, struct client_state *csp) +char *client_referrer(const struct parsers *v, const char *s, struct client_state *csp) { -#ifdef FORCE_LOAD + const char * newval; + char * s2; +#ifdef FEATURE_FORCE_LOAD /* Since the referrer can include the prefix even * even if the request itself is non-forced, we must * clean it unconditionally */ strclean(s, FORCE_PREFIX); -#endif /* def FORCE_LOAD */ +#endif /* def FEATURE_FORCE_LOAD */ +#ifdef FEATURE_TRUST csp->referrer = strdup(s); +#endif /* def FEATURE_TRUST */ - if (referrer == NULL) + /* + * Are we sending referer? + */ + if ((csp->action->flags & ACTION_HIDE_REFERER) == 0) { - log_error(LOG_LEVEL_HEADER, "crunch!"); - return(NULL); + return(strdup(s)); } - if (*referrer == '.') + newval = csp->action->string[ACTION_STRING_REFERER]; + + /* + * Are we blocking referer? + */ + if ((newval == NULL) || (0 == strcmpic(newval, "block")) ) { - return(strdup(s)); + log_error(LOG_LEVEL_HEADER, "crunch!"); + return(NULL); } - if (*referrer == '@') + /* + * Are we forging referer? + */ + if (0 == strcmpic(newval, "forge")) { - if (csp->permissions & PERMIT_COOKIE_READ) - { - return(strdup(s)); - } - else - { - log_error(LOG_LEVEL_HEADER, "crunch!"); - return(NULL); - } + /* + * Forge a referer as http://[hostname:port of REQUEST]/ + * to fool stupid checks for in-site links + */ + log_error(LOG_LEVEL_HEADER, "crunch+forge!"); + s2 = strsav(NULL, "Referer: "); + s2 = strsav(s2, "http://"); + s2 = strsav(s2, csp->http->hostport); + s2 = strsav(s2, "/"); + return(s2); } /* - * New option § or L: Forge a referer as http://[hostname:port of REQUEST]/ - * to fool stupid checks for in-site links + * Have we got a fixed referer? */ - - if (*referrer == '§' || *referrer == 'L') + if (0 == strncmpic(newval, "http://", 7)) { - if (csp->permissions & PERMIT_COOKIE_READ) - { - return(strdup(s)); - } - else - { - log_error(LOG_LEVEL_HEADER, "crunch+forge!"); - s = strsav(NULL, "Referer: "); - s = strsav(s, "http://"); - s = strsav(s, csp->http->hostport); - s = strsav(s, "/"); - return(s); - } - } + /* + * We have a specific (fixed) referer we want to send. + */ + char * s3; - log_error(LOG_LEVEL_HEADER, "modified"); + log_error(LOG_LEVEL_HEADER, "modified"); + + s3 = strsav( NULL, "Referer: " ); + s3 = strsav( s3, newval ); + return(s3); + } - s = strsav( NULL, "Referer: " ); - s = strsav( s, referrer ); - return(s); + /* Should never get here! */ + log_error(LOG_LEVEL_ERROR, "Bad parameter: +referer{%s}", newval); + /* + * Forge is probably the best default. + * + * Forge a referer as http://[hostname:port of REQUEST]/ + * to fool stupid checks for in-site links + */ + log_error(LOG_LEVEL_HEADER, "crunch+forge!"); + s2 = strsav(NULL, "Referer: "); + s2 = strsav(s2, "http://"); + s2 = strsav(s2, csp->http->hostport); + s2 = strsav(s2, "/"); + return(s2); } @@ -841,8 +1144,9 @@ char *client_referrer(const struct parsers *v, char *s, struct client_state *csp * * Function : client_uagent * - * Description : Handle the "user-agent" config setting properly. - * Called from `sed'. + * Description : Handle the "user-agent" config setting properly + * and remember its original value to enable browser + * bug workarounds. Called from `sed'. * * Parameters : * 1 : v = ignored @@ -853,9 +1157,18 @@ char *client_referrer(const struct parsers *v, char *s, struct client_state *csp * a malloc'ed string pointer to this header (ie. pass thru). * *********************************************************************/ -char *client_uagent(const struct parsers *v, char *s, struct client_state *csp) +char *client_uagent(const struct parsers *v, const char *s, struct client_state *csp) { -#ifdef DETECT_MSIE_IMAGES + const char * newval; + char * s2; + + /* Save the client's User-Agent: value */ + if (strlen(s) >= 12) + { + csp->http->user_agent = strdup(s + 12); + } + +#ifdef FEATURE_IMAGE_DETECT_MSIE if (strstr (s, "MSIE ")) { /* This is Microsoft Internet Explorer. @@ -863,37 +1176,24 @@ char *client_uagent(const struct parsers *v, char *s, struct client_state *csp) */ csp->accept_types |= ACCEPT_TYPE_IS_MSIE; } -#endif /* def DETECT_MSIE_IMAGES */ +#endif /* def FEATURE_IMAGE_DETECT_MSIE */ - if (uagent == NULL) - { - log_error(LOG_LEVEL_HEADER, "default"); - return(strdup(DEFAULT_USER_AGENT)); - } - - if (*uagent == '.') + if ((csp->action->flags & ACTION_HIDE_USER_AGENT) == 0) { return(strdup(s)); } - if (*uagent == '@') + newval = csp->action->string[ACTION_STRING_USER_AGENT]; + if (newval == NULL) { - if (csp->permissions & PERMIT_COOKIE_READ) - { - return(strdup(s)); - } - else - { - log_error(LOG_LEVEL_HEADER, "default"); - return(strdup(DEFAULT_USER_AGENT)); - } + return(strdup(s)); } log_error(LOG_LEVEL_HEADER, "modified"); - s = strsav( NULL, "User-Agent: " ); - s = strsav( s, uagent ); - return(s); + s2 = strsav( NULL, "User-Agent: " ); + s2 = strsav( s2, newval ); + return(s2); } @@ -912,35 +1212,17 @@ char *client_uagent(const struct parsers *v, char *s, struct client_state *csp) * Returns : NULL if crunched, or a malloc'ed string to original header * *********************************************************************/ -char *client_ua(const struct parsers *v, char *s, struct client_state *csp) +char *client_ua(const struct parsers *v, const char *s, struct client_state *csp) { - if (uagent == NULL) - { - log_error(LOG_LEVEL_HEADER, "crunch!"); - return(NULL); - } - - if (*uagent == '.') + if ((csp->action->flags & ACTION_HIDE_USER_AGENT) == 0) { return(strdup(s)); } - - if (*uagent == '@') + else { - if (csp->permissions & PERMIT_COOKIE_READ) - { - return(strdup(s)); - } - else - { - log_error(LOG_LEVEL_HEADER, "crunch!"); - return(NULL); - } + log_error(LOG_LEVEL_HEADER, "crunch!"); + return(NULL); } - - log_error(LOG_LEVEL_HEADER, "crunch!"); - return(NULL); - } @@ -960,25 +1242,32 @@ char *client_ua(const struct parsers *v, char *s, struct client_state *csp) * modified/original header. * *********************************************************************/ -char *client_from(const struct parsers *v, char *s, struct client_state *csp) +char *client_from(const struct parsers *v, const char *s, struct client_state *csp) { - /* if not set, zap it */ - if (from == NULL) + const char * newval; + char * s2; + + if ((csp->action->flags & ACTION_HIDE_FROM) == 0) { - log_error(LOG_LEVEL_HEADER, "crunch!"); - return(NULL); + return(strdup(s)); } - if (*from == '.') + newval = csp->action->string[ACTION_STRING_FROM]; + + /* + * Are we blocking referer? + */ + if ((newval == NULL) || (0 == strcmpic(newval, "block")) ) { - return(strdup(s)); + log_error(LOG_LEVEL_HEADER, "crunch!"); + return(NULL); } log_error(LOG_LEVEL_HEADER, " modified"); - s = strsav( NULL, "From: " ); - s = strsav( s, from ); - return(s); + s2 = strsav( NULL, "From: " ); + s2 = strsav( s2, newval ); + return(s2); } @@ -999,9 +1288,9 @@ char *client_from(const struct parsers *v, char *s, struct client_state *csp) * Returns : Always NULL. * *********************************************************************/ -char *client_send_cookie(const struct parsers *v, char *s, struct client_state *csp) +char *client_send_cookie(const struct parsers *v, const char *s, struct client_state *csp) { - if (csp->permissions & PERMIT_COOKIE_READ) + if ((csp->action->flags & ACTION_NO_COOKIE_READ) == 0) { enlist(csp->cookie_list, s + v->len + 1); } @@ -1034,10 +1323,11 @@ char *client_send_cookie(const struct parsers *v, char *s, struct client_state * * Returns : Always NULL. * *********************************************************************/ -char *client_x_forwarded(const struct parsers *v, char *s, struct client_state *csp) +char *client_x_forwarded(const struct parsers *v, const char *s, struct client_state *csp) { - if (add_forwarded) + if ((csp->action->flags & ACTION_HIDE_FORWARDED) == 0) { + /* Save it so we can re-add it later */ csp->x_forwarded = strdup(s); } @@ -1050,7 +1340,7 @@ char *client_x_forwarded(const struct parsers *v, char *s, struct client_state * } -#if defined(DETECT_MSIE_IMAGES) +#if defined(FEATURE_IMAGE_DETECT_MSIE) /********************************************************************* * * Function : client_accept @@ -1068,9 +1358,9 @@ char *client_x_forwarded(const struct parsers *v, char *s, struct client_state * * Returns : Duplicate of argument s. * *********************************************************************/ -char *client_accept(const struct parsers *v, char *s, struct client_state *csp) +char *client_accept(const struct parsers *v, const char *s, struct client_state *csp) { -#ifdef DETECT_MSIE_IMAGES +#ifdef FEATURE_IMAGE_DETECT_MSIE if (strstr (s, "image/gif")) { /* Client will accept HTML. If this seems counterintuitive, @@ -1082,18 +1372,44 @@ char *client_accept(const struct parsers *v, char *s, struct client_state *csp) { csp->accept_types |= ACCEPT_TYPE_MSIE_IMAGE; } -#endif /* def DETECT_MSIE_IMAGES */ +#endif /* def FEATURE_IMAGE_DETECT_MSIE */ return(strdup(s)); } -#endif /* defined(DETECT_MSIE_IMAGES) */ +#endif /* defined(FEATURE_IMAGE_DETECT_MSIE) */ /* the following functions add headers directly to the header list */ +/********************************************************************* + * + * Function : client_host_adder + * + * Description : (re)adds the host header. Called from `sed'. + * + * Parameters : + * 1 : csp = Current client state (buffers, headers, etc...) + * + * Returns : N/A + * + *********************************************************************/ +void client_host_adder(struct client_state *csp) +{ + char *p = NULL; + + p = strsav(p, "Host: "); + p = strsav(p, csp->http->hostport); + + log_error(LOG_LEVEL_HEADER, "addh: %s", p); + enlist(csp->headers, p); + + freez(p); +} + + /********************************************************************* * * Function : client_cookie_adder @@ -1108,27 +1424,27 @@ char *client_accept(const struct parsers *v, char *s, struct client_state *csp) *********************************************************************/ void client_cookie_adder(struct client_state *csp) { - struct list *l; + struct list_entry *lst; char *tmp = NULL; char *e; - for (l = csp->cookie_list->next; l ; l = l->next) + for (lst = csp->cookie_list->first; lst ; lst = lst->next) { if (tmp) { tmp = strsav(tmp, "; "); } - tmp = strsav(tmp, l->str); + tmp = strsav(tmp, lst->str); } - for (l = wafer_list->next; l ; l = l->next) + for (lst = csp->action->multi[ACTION_MULTI_WAFER]->first; lst ; lst = lst->next) { if (tmp) { tmp = strsav(tmp, "; "); } - if ((e = cookie_encode(l->str))) + if ((e = cookie_encode(lst->str))) { tmp = strsav(tmp, e); freez(e); @@ -1150,6 +1466,32 @@ void client_cookie_adder(struct client_state *csp) } +/********************************************************************* + * + * Function : client_accept_encoding_adder + * + * Description : Add an Accept-Encoding header to the client's request + * that disables compression if the action applies, and + * the header is not already there. Called from `sed'. + * Note: For HTTP/1.0, the absence of the header is enough. + * + * Parameters : + * 1 : csp = Current client state (buffers, headers, etc...) + * + * Returns : N/A + * + *********************************************************************/ +void client_accept_encoding_adder(struct client_state *csp) +{ + if ( ((csp->action->flags & ACTION_NO_COMPRESSION) != 0) + && (!strcmpic(csp->http->ver, "HTTP/1.1")) ) + { + enlist_unique(csp->headers, "Accept-Encoding: identity;q=1.0, *;q=0", 16); + } + +} + + /********************************************************************* * * Function : client_xtra_adder @@ -1164,12 +1506,13 @@ void client_cookie_adder(struct client_state *csp) *********************************************************************/ void client_xtra_adder(struct client_state *csp) { - struct list *l; + struct list_entry *lst; - for (l = xtra_list->next; l ; l = l->next) + for (lst = csp->action->multi[ACTION_MULTI_ADD_HEADER]->first; + lst ; lst = lst->next) { - log_error(LOG_LEVEL_HEADER, "addh: %s", l->str); - enlist(csp->headers, l->str); + log_error(LOG_LEVEL_HEADER, "addh: %s", lst->str); + enlist(csp->headers, lst->str); } } @@ -1191,7 +1534,10 @@ void client_x_forwarded_adder(struct client_state *csp) { char *p = NULL; - if (add_forwarded == 0) return; + if ((csp->action->flags & ACTION_HIDE_FORWARDED) != 0) + { + return; + } if (csp->x_forwarded) { @@ -1208,73 +1554,105 @@ void client_x_forwarded_adder(struct client_state *csp) log_error(LOG_LEVEL_HEADER, "addh: %s", p); enlist(csp->headers, p); + freez(p); } /********************************************************************* * - * Function : server_set_cookie + * Function : connection_close_adder * - * Description : Handle the server "cookie" header properly. - * Log cookie to the jar file. Then "crunch" it, - * or accept it. Called from `sed'. + * Description : Adds a "Connection: close" header to csp->headers + * as a temporary fix for the needed but missing HTTP/1.1 + * support. Called from `sed'. + * FIXME: This whole function shouldn't be neccessary! + * + * Parameters : + * 1 : csp = Current client state (buffers, headers, etc...) + * + * Returns : N/A + * + *********************************************************************/ +void connection_close_adder(struct client_state *csp) +{ + enlist(csp->headers, "Connection: close"); +} + + +/********************************************************************* + * + * Function : server_http + * + * Description : - Save the HTTP Status into csp->http->status + * - Set CT_TABOO to prevent filtering if the answer + * is a partial range (HTTP status 206) + * - Rewrite HTTP/1.1 answers to HTTP/1.0 if +downgrade + * action applies. * * Parameters : * 1 : v = parser pattern that matched this header * 2 : s = header that matched this pattern * 3 : csp = Current client state (buffers, headers, etc...) * - * Returns : `crumble' or a newly malloc'ed string. + * Returns : Copy of changed or original answer. * *********************************************************************/ -char *server_set_cookie(const struct parsers *v, char *s, struct client_state *csp) +char *server_http(const struct parsers *v, const char *s, struct client_state *csp) { -#ifdef JAR_FILES - if (jar) + char *ret = strdup(s); + + sscanf(ret, "HTTP/%*d.%*d %d", &(csp->http->status)); + if (csp->http->status == 206) { - fprintf(jar, "%s\t%s\n", csp->http->host, (s + v->len + 1)); + csp->content_type = CT_TABOO; } -#endif /* def JAR_FILES */ - if (!(csp->permissions & PERMIT_COOKIE_SET)) + if ((csp->action->flags & ACTION_DOWNGRADE) != 0) { - return(crumble(v, s, csp)); + ret[7] = '0'; + log_error(LOG_LEVEL_HEADER, "Downgraded answer to HTTP/1.0"); } - - return(strdup(s)); + return(ret); } -#ifdef FORCE_LOAD /********************************************************************* * - * Function : client_host + * Function : server_set_cookie * - * Description : Clean the FORCE_PREFIX out of the 'host' http - * header, if we use force + * Description : Handle the server "cookie" header properly. + * Log cookie to the jar file. Then "crunch" it, + * or accept it. Called from `sed'. * * Parameters : - * 1 : v = ignored - * 2 : s = header (from sed) to clean + * 1 : v = parser pattern that matched this header + * 2 : s = header that matched this pattern * 3 : csp = Current client state (buffers, headers, etc...) * - * Returns : A malloc'ed pointer to the cleaned host header + * Returns : `crumble' or a newly malloc'ed string. * *********************************************************************/ -char *client_host(const struct parsers *v, char *s, struct client_state *csp) +char *server_set_cookie(const struct parsers *v, const char *s, struct client_state *csp) { - char *cleanhost = strdup(s); - - if(csp->force) - strclean(cleanhost, FORCE_PREFIX); - - return(cleanhost); +#ifdef FEATURE_COOKIE_JAR + if (csp->config->jar) + { + fprintf(csp->config->jar, "%s\t%s\n", csp->http->host, (s + v->len + 1)); + } +#endif /* def FEATURE_COOKIE_JAR */ + + if ((csp->action->flags & ACTION_NO_COOKIE_SET) != 0) + { + return(crumble(v, s, csp)); + } + + return(strdup(s)); + } -#endif /* def FORCE_LOAD */ - - -#ifdef FORCE_LOAD + + +#ifdef FEATURE_FORCE_LOAD /********************************************************************* * * Function : strclean @@ -1308,7 +1686,7 @@ int strclean(const char *string, const char *substring) return(hits); } -#endif /* def FORCE_LOAD */ +#endif /* def FEATURE_FORCE_LOAD */ /*