X-Git-Url: http://www.privoxy.org/gitweb/?p=privoxy.git;a=blobdiff_plain;f=filters.c;h=e268d6ddcd2659c7aa121d5575d44b80e6381bc0;hp=e43ad99afd8cf74af5d0bc679ec57ba12ad625db;hb=cc7698e75316ef26c2e00492b500a3483a68f398;hpb=e30955dc472815457c4e10ed744fa7aa29ac4800 diff --git a/filters.c b/filters.c index e43ad99a..e268d6dd 100644 --- a/filters.c +++ b/filters.c @@ -1,4 +1,4 @@ -const char filters_rcs[] = "$Id: filters.c,v 1.15 2001/06/03 11:03:48 oes Exp $"; +const char filters_rcs[] = "$Id: filters.c,v 1.24 2001/07/25 17:22:51 oes Exp $"; /********************************************************************* * * File : $Source: /cvsroot/ijbswa/current/filters.c,v $ @@ -8,7 +8,7 @@ const char filters_rcs[] = "$Id: filters.c,v 1.15 2001/06/03 11:03:48 oes Exp $" * `acl_addr', `add_stats', `block_acl', `block_imageurl', * `block_url', `url_actions', `domaincmp', `dsplit', * `filter_popups', `forward_url', 'redirect_url', - * `ij_untrusted_url', `intercept_url', `re_process_buffer', + * `ij_untrusted_url', `intercept_url', `pcrs_filter_respose', * `show_proxy_args', 'ijb_send_banner', and `trust_url' * * Copyright : Written by and Copyright (C) 2001 the SourceForge @@ -38,59 +38,68 @@ const char filters_rcs[] = "$Id: filters.c,v 1.15 2001/06/03 11:03:48 oes Exp $" * * Revisions : * $Log: filters.c,v $ - * Revision 1.15 2001/06/03 11:03:48 oes - * Makefile/in - * - * introduced cgi.c - * - * actions.c: - * - * adapted to new enlist_unique arg format - * - * conf loadcfg.c - * - * introduced confdir option - * - * filters.c filtrers.h - * - * extracted-CGI relevant stuff - * - * jbsockets.c - * - * filled comment - * - * jcc.c - * - * support for new cgi mechansim - * - * list.c list.h - * - * functions for new list type: "map" - * extended enlist_unique - * - * miscutil.c .h - * introduced bindup() - * - * parsers.c parsers.h - * - * deleted const struct interceptors - * - * pcrs.c - * added FIXME - * - * project.h - * - * added struct map - * added struct http_response - * changes struct interceptors to struct cgi_dispatcher - * moved HTML stuff to cgi.h - * - * re_filterfile: - * - * changed - * - * showargs.c - * NO TIME LEFT + * Revision 1.24 2001/07/25 17:22:51 oes + * Added workaround for Netscape bug that prevents display of page when loading a component fails. + * + * Revision 1.23 2001/07/23 13:40:12 oes + * Fixed bug that caused document body to be dropped when pcrs joblist was empty. + * + * Revision 1.22 2001/07/18 12:29:34 oes + * - Made gif_deanimate_response respect + * csp->action->string[ACTION_STRING_DEANIMATE] + * - Logging cosmetics + * + * Revision 1.21 2001/07/13 13:59:53 oes + * - Introduced gif_deanimate_response which shares the + * generic content modification interface of pcrs_filter_response + * and acts as a wrapper to deanimate.c:gif_deanimate() + * - Renamed re_process_buffer to pcrs_filter_response + * - pcrs_filter_response now returns NULL on failiure + * - Removed all #ifdef PCRS + * + * Revision 1.20 2001/07/01 17:01:04 oes + * Added comments and missing return statement in is_untrusted_url() + * + * Revision 1.19 2001/06/29 21:45:41 oes + * Indentation, CRLF->LF, Tab-> Space + * + * Revision 1.18 2001/06/29 13:27:38 oes + * - Cleaned up, renamed and reorderd functions + * and improved comments + * + * - block_url: + * - Ported to CGI platform. Now delivers + * http_response or NULL + * - Unified HTML and GIF generation (moved image detection + * and GIF generation here from jcc.c:chat()) + * - Fixed HTTP status to: + * - 403 (Forbidden) for the "blocked" HTML message + * - 200 (OK) for GIF answers + * - 302 (Redirect) for redirect to GIF + * + * - trust_url: + * - Ported to CGI platform. Now delivers + * http_response or NULL + * - Separated detection of untrusted URL into + * (bool)is_untrusted_url + * - Added enforcement of untrusted requests + * + * - Moved redirect_url() from cgi.c to here + * and ported it to the CGI platform + * + * - Removed logentry from cancelled commit + * + * Revision 1.17 2001/06/09 10:55:28 jongfoster + * Changing BUFSIZ ==> BUFFER_SIZE + * + * Revision 1.16 2001/06/07 23:10:26 jongfoster + * Allowing unanchored domain patterns to back off and retry + * if they partially match. Optimized right-anchored patterns. + * Moving ACL and forward files into config file. + * Replacing struct gateway with struct forward_spec + * + * Revision 1.15 2001/06/03 19:12:00 oes + * extracted-CGI relevant stuff * * Revision 1.14 2001/06/01 10:30:55 oes * Added optional left-anchoring to domaincmp @@ -258,6 +267,8 @@ const char filters_rcs[] = "$Id: filters.c,v 1.15 2001/06/03 11:03:48 oes Exp $" #include "miscutil.h" #include "actions.h" #include "cgi.h" +#include "list.h" +#include "deanimate.h" #ifdef _WIN32 #include "win32.h" @@ -292,36 +303,33 @@ const char filters_h_rcs[] = FILTERS_H_VERSION; * Returns : 0 = FALSE (don't block) and 1 = TRUE (do block) * *********************************************************************/ -int block_acl(struct access_control_addr *dst, - struct client_state *csp) +int block_acl(struct access_control_addr *dst, struct client_state *csp) { - struct file_list *fl; - struct access_control_list *a, *acl; + struct access_control_list *acl = csp->config->acl; /* if not using an access control list, then permit the connection */ - if (((fl = csp->alist) == NULL) || - ((acl = (struct access_control_list *) fl->f) == NULL)) + if (acl == NULL) { return(0); } /* search the list */ - for (a = acl->next ; a ; a = a->next) + while (acl != NULL) { - if ((csp->ip_addr_long & a->src->mask) == a->src->addr) + if ((csp->ip_addr_long & acl->src->mask) == acl->src->addr) { if (dst == NULL) { /* Just want to check if they have any access */ - if (a->action == ACL_PERMIT) + if (acl->action == ACL_PERMIT) { return(0); } } - else if ( ((dst->addr & a->dst->mask) == a->dst->addr) - && ((dst->port == a->dst->port) || (a->dst->port == 0))) + else if ( ((dst->addr & acl->dst->mask) == acl->dst->addr) + && ((dst->port == acl->dst->port) || (acl->dst->port == 0))) { - if (a->action == ACL_PERMIT) + if (acl->action == ACL_PERMIT) { return(0); } @@ -331,6 +339,7 @@ int block_acl(struct access_control_addr *dst, } } } + acl = acl->next; } return(1); @@ -421,63 +430,295 @@ int acl_addr(char *aspec, struct access_control_addr *aca) * Description : Called from `chat'. Check to see if we need to block this. * * Parameters : - * 1 : http = http_request request to "check" for blocked - * 2 : csp = Current client state (buffers, headers, etc...) + * 1 : csp = Current client state (buffers, headers, etc...) * - * Returns : NULL => unblocked, else string to HTML block description. + * Returns : NULL => unblocked, else HTTP block response * *********************************************************************/ -char *block_url(struct http_request *http, struct client_state *csp) +struct http_response *block_url(struct client_state *csp) { char *p; - int n; - int factor = 2; + struct http_response *rsp; + struct map *exports = NULL; + /* + * If it's not blocked, don't block it ;-) + */ if ((csp->action->flags & ACTION_BLOCK) == 0) { return(NULL); } + + /* + * Else, prepare a response + */ + if (NULL == ( rsp = (struct http_response *)zalloc(sizeof(*rsp)))) + { + return NULL; + } + + /* + * If it's an image-url, send back an image or redirect + * as specified by the relevant +image action + */ +#ifdef IMAGE_BLOCKING + if (((csp->action->flags & ACTION_IMAGE_BLOCKER) != 0) + && is_imageurl(csp)) + { + /* determine HOW images should be blocked */ + p = csp->action->string[ACTION_STRING_IMAGE_BLOCKER]; + + /* and handle accordingly: */ + if ((p == NULL) || (0 == strcmpic(p, "logo"))) + { + rsp->body = bindup(JBGIF, sizeof(JBGIF)); + rsp->content_length = sizeof(JBGIF); + enlist_unique_header(rsp->headers, "Content-Type", "image/gif"); + } + + else if (0 == strcmpic(p, "blank")) + { + rsp->body = bindup(BLANKGIF, sizeof(BLANKGIF)); + rsp->content_length = sizeof(BLANKGIF); + enlist_unique_header(rsp->headers, "Content-Type", "image/gif"); + } + + else + { + rsp->status = strdup("302 Local Redirect from Junkbuster"); + enlist_unique_header(rsp->headers, "Location", p); + } + } else +#endif /* def IMAGE_BLOCKING */ + + /* + * Else, generate an HTML "blocked" message: + */ { + + exports = default_exports(csp, NULL); #ifdef FORCE_LOAD - factor++; -#endif /* def FORCE_LOAD */ + exports = map(exports, "force-prefix", 1, FORCE_PREFIX, 1); +#else + exports = map_block_killer(exports, "force-support"); +#endif /* ndef FORCE_LOAD */ + + exports = map(exports, "hostport", 1, csp->http->hostport, 1); + exports = map(exports, "hostport-html", 1, html_encode(csp->http->hostport), 0); + exports = map(exports, "path", 1, csp->http->path, 1); + exports = map(exports, "path-html", 1, html_encode(csp->http->path), 0); + + rsp->body = fill_template(csp, "blocked", exports); + free_map(exports); + + /* + * Workaround for stupid Netscape bug which prevents + * pages from being displayed if loading a referenced + * JavaScript or style sheet fails. So make it appear + * as if it succeeded. + */ + if (csp->http->user_agent + && !strncmpic(csp->http->user_agent, "mozilla", 7) + && !strstr(csp->http->user_agent, "compatible") + && !strstr(csp->http->user_agent, "Opera")) + { + rsp->status = strdup("200 Request for blocked URL"); + } + else + { + rsp->status = strdup("404 Request for blocked URL"); + } + + } + + return(finish_http_response(rsp)); + +} + + +#ifdef TRUST_FILES +/********************************************************************* + * + * Function : trust_url FIXME: I should be called distrust_url + * + * Description : Calls is_untrusted_url to determine if the URL is trusted + * and if not, returns a HTTP 304 response with a reject message. + * + * Parameters : + * 1 : csp = Current client state (buffers, headers, etc...) + * + * Returns : NULL => trusted, else http_response. + * + *********************************************************************/ +struct http_response *trust_url(struct client_state *csp) +{ + struct http_response *rsp; + struct map *exports = NULL; + char buf[BUFFER_SIZE], *p = NULL; + struct url_spec **tl, *t; + + /* + * Don't bother to work on trusted URLs + */ + if (!is_untrusted_url(csp)) + { + return NULL; + } + + /* + * Else, prepare a response: + */ + if (NULL == ( rsp = (struct http_response *)zalloc(sizeof(*rsp)))) + { + return NULL; + } + exports = default_exports(csp, NULL); - n = strlen(CBLOCK); - n += factor * strlen(http->hostport); - n += factor * strlen(http->path); + /* + * Export the host, port, and referrer information + */ + exports = map(exports, "hostport", 1, csp->http->hostport, 1); + exports = map(exports, "path", 1, csp->http->path, 1); + exports = map(exports, "hostport-html", 1, html_encode(csp->http->hostport), 0); + exports = map(exports, "path-html", 1, html_encode(csp->http->path), 0); - p = (char *)malloc(n); + if (csp->referrer && strlen(csp->referrer) > 9) + { + exports = map(exports, "referrer", 1, csp->referrer + 9, 1); + exports = map(exports, "referrer-html", 1, html_encode(csp->referrer + 9), 0); + } + else + { + exports = map(exports, "referrer", 1, "unknown", 1); + exports = map(exports, "referrer-html", 1, "unknown", 1); + } + /* + * Export the trust list + */ + for (tl = csp->config->trust_list; (t = *tl) ; tl++) + { + sprintf(buf, "
  • %s
  • \n", t->spec); + p = strsav(p, buf); + } + exports = map(exports, "trusted-referrers", 1, p, 0); + p = NULL; + + /* + * Export the trust info, if available + */ + if (csp->config->trust_info->next) + { + struct list *l; + + for (l = csp->config->trust_info->next; l ; l = l->next) + { + sprintf(buf, "
  • %s
    \n",l->str, l->str); + p = strsav(p, buf); + } + exports = map(exports, "trust-info", 1, p, 0); + } + else + { + exports = map_block_killer(exports, "have-trust-info"); + } + + /* + * Export the force prefix or the force conditional block killer + */ #ifdef FORCE_LOAD - sprintf(p, CBLOCK, http->hostport, http->path, http->hostport, http->path, - http->hostport, http->path); + exports = map(exports, "force-prefix", 1, FORCE_PREFIX, 1); #else - sprintf(p, CBLOCK, http->hostport, http->path, http->hostport, http->path); -#endif /* def FORCE_LOAD */ + exports = map_block_killer(exports, "force-support"); +#endif /* ndef FORCE_LOAD */ - return(p); + /* + * Build the response + */ + rsp->body = fill_template(csp, "untrusted", exports); + free_map(exports); + + return(finish_http_response(rsp)); + +} +#endif /* def TRUST_FILES */ + + +#ifdef FAST_REDIRECTS +/********************************************************************* + * + * Function : redirect_url + * + * Description : Checks for redirection URLs and returns a HTTP redirect + * to the destination URL, if necessary + * + * Parameters : + * 1 : csp = Current client state (buffers, headers, etc...) + * + * Returns : NULL if URL was clean, HTTP redirect otherwise. + * + *********************************************************************/ +struct http_response *redirect_url(struct client_state *csp) +{ + char *p, *q; + struct http_response *rsp; + + p = q = csp->http->path; + log_error(LOG_LEVEL_REDIRECTS, "checking path for redirects: %s", p); + + /* + * find the last URL encoded in the request + */ + while (p = strstr(p, "http://")) + { + q = p++; } + + /* + * if there was any, generate and return a HTTP redirect + */ + if (q != csp->http->path) + { + log_error(LOG_LEVEL_REDIRECTS, "redirecting to: %s", q); + + if (NULL == ( rsp = zalloc(sizeof(*rsp)))) + { + return NULL; + } + + rsp->status = strdup("302 Local Redirect from Junkbuster"); + enlist_unique_header(rsp->headers, "Location", q); + + return(finish_http_response(rsp)); + } + else + { + return(NULL); + } + } +#endif /* def FAST_REDIRECTS */ #ifdef IMAGE_BLOCKING /********************************************************************* * - * Function : block_imageurl - * - * Description : Given a URL which is blocked, decide whether to - * send the "blocked" image or HTML. + * Function : is_imageurl * + * Description : Given a URL, decide whether it is an image or not, + * using either the info from a previous +image action + * or, #ifdef DETECT_MSIE_IMAGES, the info from the + * browser's accept header. + * * Parameters : - * 1 : http = URL to check. - * 2 : csp = Current client state (buffers, headers, etc...) + * 1 : csp = Current client state (buffers, headers, etc...) * - * Returns : True (nonzero) if URL is in image list, false (0) + * Returns : True (nonzero) if URL is an image, false (0) * otherwise * *********************************************************************/ -int block_imageurl(struct http_request *http, struct client_state *csp) +int is_imageurl(struct client_state *csp) { #ifdef DETECT_MSIE_IMAGES if ((csp->accept_types @@ -495,6 +736,7 @@ int block_imageurl(struct http_request *http, struct client_state *csp) #endif return ((csp->action->flags & ACTION_IMAGE) != 0); + } #endif /* def IMAGE_BLOCKING */ @@ -502,84 +744,68 @@ int block_imageurl(struct http_request *http, struct client_state *csp) #ifdef TRUST_FILES /********************************************************************* * - * Function : trust_url + * Function : is_untrusted_url + * + * Description : Should we "distrust" this URL (and block it)? * - * Description : Should we "trust" this URL? See "trustfile" line in config. + * Yes if it matches a line in the trustfile, or if the + * referrer matches a line starting with "+" in the + * trustfile. + * No otherwise. * * Parameters : - * 1 : http = http_request request for requested URL - * 2 : csp = Current client state (buffers, headers, etc...) + * 1 : csp = Current client state (buffers, headers, etc...) * - * Returns : NULL => trusted, else string to HTML "untrusted" description. + * Returns : 0 => trusted, 1 => untrusted * *********************************************************************/ -char *trust_url(struct http_request *http, struct client_state *csp) +int is_untrusted_url(struct client_state *csp) { struct file_list *fl; struct block_spec *b; struct url_spec url[1], **tl, *t; - char *p, *h; - char *hostport, *path, *refer; struct http_request rhttp[1]; - int n; + char *p, *h; + /* + * If we don't have a trustlist, we trust everybody + */ if (((fl = csp->tlist) == NULL) || ((b = fl->f) == NULL)) { - return(NULL); + return(0); } - *url = dsplit(http->host); + + /* + * Do we trust the request URL itself? + */ + *url = dsplit(csp->http->host); /* if splitting the domain fails, punt */ - if (url->dbuf == NULL) return(NULL); + if (url->dbuf == NULL) return(0); memset(rhttp, '\0', sizeof(*rhttp)); for (b = b->next; b ; b = b->next) { - if ((b->url->port == 0) || (b->url->port == http->port)) + if ((b->url->port == 0) || (b->url->port == csp->http->port)) { if ((b->url->domain[0] == '\0') || (domaincmp(b->url, url) == 0)) { if ((b->url->path == NULL) || #ifdef REGEX - (regexec(b->url->preg, http->path, 0, NULL, 0) == 0) + (regexec(b->url->preg, csp->http->path, 0, NULL, 0) == 0) #else - (strncmp(b->url->path, http->path, b->url->pathlen) == 0) + (strncmp(b->url->path, csp->http->path, b->url->pathlen) == 0) #endif ) { freez(url->dbuf); freez(url->dvec); - if (b->reject == 0) return(NULL); - - hostport = url_encode(http->hostport); - path = url_encode(http->path); - - if (csp->referrer) - { - refer = url_encode(csp->referrer); - } - else - { - refer = url_encode("undefined"); - } - - n = strlen(CTRUST); - n += strlen(hostport); - n += strlen(path); - n += strlen(refer); - - p = (char *)malloc(n); + if (b->reject == 0) return(0); - sprintf(p, CTRUST, hostport, path, refer); - - freez(hostport); - freez(path); - freez(refer); - - return(p); + return(1); } } } @@ -591,7 +817,7 @@ char *trust_url(struct http_request *http, struct client_state *csp) if ((csp->referrer == NULL)|| (strlen(csp->referrer) <= 9)) { /* no referrer was supplied */ - goto trust_url_not_trusted; + return(1); } /* forge a URL from the referrer so we can use @@ -604,19 +830,21 @@ char *trust_url(struct http_request *http, struct client_state *csp) p = strsav(p, " HTTP/1.0"); parse_http_request(p, rhttp, csp); + freez(p); if (rhttp->cmd == NULL) { - freez(p); - goto trust_url_not_trusted; + return(1); } - freez(p); + /* + * If not, do we maybe trust its referrer? + */ *url = dsplit(rhttp->host); /* if splitting the domain fails, punt */ - if (url->dbuf == NULL) goto trust_url_not_trusted; + if (url->dbuf == NULL) return(1); for (tl = csp->config->trust_list; (t = *tl) ; tl++) { @@ -647,9 +875,9 @@ char *trust_url(struct http_request *http, struct client_state *csp) h = NULL; h = strsav(h, "~"); - h = strsav(h, http->hostport); + h = strsav(h, csp->http->hostport); - p = http->path; + p = csp->http->path; if ((*p++ == '/') && (*p++ == '~')) { @@ -659,78 +887,48 @@ char *trust_url(struct http_request *http, struct client_state *csp) if ((p = strchr(p, '/'))) { *p = '\0'; - h = strsav(h, http->path); + h = strsav(h, csp->http->path); /* FIXME: p?! */ h = strsav(h, "/"); } } - free_http_request(rhttp); - fprintf(fp, "%s\n", h); freez(h); fclose(fp); } - return(NULL); + return(0); } } } } - -trust_url_not_trusted: - free_http_request(rhttp); - - hostport = url_encode(http->hostport); - path = url_encode(http->path); - - if (csp->referrer) - { - refer = url_encode(csp->referrer); - } - else - { - refer = url_encode("undefined"); - } - - n = strlen(CTRUST); - n += strlen(hostport); - n += strlen(path); - n += strlen(refer); - - p = (char *)malloc(n); - sprintf(p, CTRUST, hostport, path, refer); - - freez(hostport); - freez(path); - freez(refer); - - return(p); - + return(1); } #endif /* def TRUST_FILES */ -#ifdef PCRS /********************************************************************* * - * Function : re_process_buffer + * Function : pcrs_filter_response * - * Description : Apply all jobs from the joblist (aka. Perl regexp's) to - * the text buffer that's been accumulated in csp->iob->buf - * and set csp->content_length to the modified size. + * Description : Apply all the pcrs jobs from the joblist (re_filterfile) + * to the text buffer that's been accumulated in + * csp->iob->buf and set csp->content_length to the modified + * size. * * Parameters : * 1 : csp = Current client state (buffers, headers, etc...) * * Returns : a pointer to the (newly allocated) modified buffer. + * or NULL in case something went wrong * - * *********************************************************************/ -char *re_process_buffer(struct client_state *csp) +char *pcrs_filter_response(struct client_state *csp) { int hits=0; int size = csp->iob->eod - csp->iob->cur; - char *old=csp->iob->cur, *new = NULL; - pcrs_job *job, *joblist; + + char *old = csp->iob->cur, *new = NULL; + pcrs_job *job; struct file_list *fl; struct re_filterfile_spec *b; @@ -738,25 +936,28 @@ char *re_process_buffer(struct client_state *csp) /* Sanity first ;-) */ if (size <= 0) { - return(strdup("")); + return(NULL); } if ( ( NULL == (fl = csp->rlist) ) || ( NULL == (b = fl->f) ) ) { log_error(LOG_LEVEL_ERROR, "Unable to get current state of regexp filtering."); - return(strdup("")); + return(NULL); } - joblist = b->joblist; - + if ( NULL == b->joblist ) + { + log_error(LOG_LEVEL_RE_FILTER, "Empty joblist. Nothing to do."); + return(NULL); + } log_error(LOG_LEVEL_RE_FILTER, "re_filtering %s%s (size %d) ...", csp->http->hostport, csp->http->path, size); /* Apply all jobs from the joblist */ - for (job = joblist; NULL != job; job = job->next) + for (job = b->joblist; NULL != job; job = job->next) { - hits += pcrs_exec_substitution(job, old, size, &new, &size); + hits += pcrs_execute(job, old, size, &new, &size); if (old != csp->iob->cur) free(old); old=new; } @@ -770,7 +971,57 @@ char *re_process_buffer(struct client_state *csp) return(new); } -#endif /* def PCRS */ + + +/********************************************************************* + * + * Function : gif_deanimate_response + * + * Description : Deanimate the GIF image that has been accumulated in + * csp->iob->buf and set csp->content_length to the modified + * size. + * + * Parameters : + * 1 : csp = Current client state (buffers, headers, etc...) + * + * Returns : a pointer to the (newly allocated) modified buffer. + * or NULL in case something went wrong. + * + *********************************************************************/ +char *gif_deanimate_response(struct client_state *csp) +{ + struct binbuffer *in, *out; + char *p; + int size = csp->iob->eod - csp->iob->cur; + + if ( (NULL == (in = (struct binbuffer *)zalloc(sizeof *in ))) + || (NULL == (out = (struct binbuffer *)zalloc(sizeof *out))) ) + { + log_error(LOG_LEVEL_DEANIMATE, "failed! (no mem)"); + return NULL; + } + + in->buffer = csp->iob->cur; + in->size = size; + + if (gif_deanimate(in, out, strncmp("last", csp->action->string[ACTION_STRING_DEANIMATE], 4))) + { + log_error(LOG_LEVEL_DEANIMATE, "failed! (gif parsing)"); + free(in); + buf_free(out); + return(NULL); + } + else + { + log_error(LOG_LEVEL_DEANIMATE, "Success! GIF shrunk from %d bytes to %d.", size, out->offset); + csp->content_length = out->offset; + p = out->buffer; + free(in); + free(out); + return(p); + } + +} /********************************************************************* @@ -800,6 +1051,7 @@ void url_actions(struct http_request *http, } apply_url_actions(csp->action, http, b); + } @@ -872,51 +1124,55 @@ void apply_url_actions(struct current_action_spec *action, * 1 : http = http_request request for current URL * 2 : csp = Current client state (buffers, headers, etc...) * - * Returns : Return gw_default for no forward match, - * else a gateway pointer to a specific forwarding proxy. + * Returns : Pointer to forwarding information. * *********************************************************************/ -const struct gateway *forward_url(struct http_request *http, struct client_state *csp) +const struct forward_spec * forward_url(struct http_request *http, + struct client_state *csp) { - struct file_list *fl; - struct forward_spec *b; + static const struct forward_spec fwd_default[1] = { 0 }; /* All zeroes */ + struct forward_spec *fwd = csp->config->forward; struct url_spec url[1]; - if (((fl = csp->flist) == NULL) || ((b = fl->f) == NULL)) + if (fwd == NULL) { - return(gw_default); + return(fwd_default); } *url = dsplit(http->host); /* if splitting the domain fails, punt */ - if (url->dbuf == NULL) return(gw_default); + if (url->dbuf == NULL) + { + return(fwd_default); + } - for (b = b->next; b ; b = b->next) + while (fwd != NULL) { - if ((b->url->port == 0) || (b->url->port == http->port)) + if ((fwd->url->port == 0) || (fwd->url->port == http->port)) { - if ((b->url->domain[0] == '\0') || (domaincmp(b->url, url) == 0)) + if ((fwd->url->domain[0] == '\0') || (domaincmp(fwd->url, url) == 0)) { - if ((b->url->path == NULL) || + if ((fwd->url->path == NULL) || #ifdef REGEX - (regexec(b->url->preg, http->path, 0, NULL, 0) == 0) + (regexec(fwd->url->preg, http->path, 0, NULL, 0) == 0) #else - (strncmp(b->url->path, http->path, b->url->pathlen) == 0) + (strncmp(fwd->url->path, http->path, fwd->url->pathlen) == 0) #endif ) { freez(url->dbuf); freez(url->dvec); - return(b->gw); + return(fwd); } } } + fwd = fwd->next; } freez(url->dbuf); freez(url->dvec); - return(gw_default); + return(fwd_default); } @@ -938,7 +1194,7 @@ const struct gateway *forward_url(struct http_request *http, struct client_state struct url_spec dsplit(char *domain) { struct url_spec ret[1]; - char *v[BUFSIZ]; + char *v[BUFFER_SIZE]; int size; char *p; @@ -946,12 +1202,13 @@ struct url_spec dsplit(char *domain) if (domain[strlen(domain) - 1] == '.') { - ret->unanchored |= ANCHOR_RIGHT; - } - if (domain[0] == '.') + ret->unanchored |= ANCHOR_RIGHT; + } + + if (domain[0] == '.') { - ret->unanchored |= ANCHOR_LEFT; - } + ret->unanchored |= ANCHOR_LEFT; + } ret->dbuf = strdup(domain); @@ -975,12 +1232,47 @@ struct url_spec dsplit(char *domain) memcpy(ret->dvec, v, size); } - return(*ret); } +/********************************************************************* + * + * Function : simple_domaincmp + * + * Description : Domain-wise Compare fqdn's. The comparison is + * both left- and right-anchored. The individual + * domain names are compared with simplematch(). + * This is only used by domaincmp. + * + * Parameters : + * 1 : pv = array of patterns to compare + * 2 : fv = array of domain components to compare + * 3 : len = length of the arrays (both arrays are the + * same length - if they weren't, it couldn't + * possibly be a match). + * + * Returns : 0 => domains are equivalent, else no match. + * + *********************************************************************/ +static int simple_domaincmp(char **pv, char **fv, int len) +{ + int n; + + for (n = 0; n < len; n++) + { + if (simplematch(pv[n], fv[n])) + { + return 1; + } + } + + return 0; + +} + + /********************************************************************* * * Function : domaincmp @@ -989,7 +1281,7 @@ struct url_spec dsplit(char *domain) * pattern->unachored, the comparison is un-, left-, * right-anchored, or both. * The individual domain names are compared with - * trivialmatch(). + * simplematch(). * * Parameters : * 1 : pattern = a domain that may contain a '*' as a wildcard. @@ -1001,38 +1293,69 @@ struct url_spec dsplit(char *domain) int domaincmp(struct url_spec *pattern, struct url_spec *fqdn) { char **pv, **fv; /* vectors */ - int pn, fn; /* counters */ - char *p, *f; /* chars */ + int plen, flen; + int unanchored = pattern->unanchored & (ANCHOR_RIGHT | ANCHOR_LEFT); - pv = pattern->dvec; - fv = fqdn->dvec; - fn = pn = 0; + plen = pattern->dcnt; + flen = fqdn->dcnt; - while (fn < fqdn->dcnt && pn < pattern->dcnt) + if (flen < plen) { - p = pv[pn]; - f = fv[fn]; + /* fqdn is too short to match this pattern */ + return 1; + } - if (simplematch(p, f)) + pv = pattern->dvec; + fv = fqdn->dvec; + + if (unanchored == ANCHOR_LEFT) + { + /* + * Right anchored. + * + * Convert this into a fully anchored pattern with + * the fqdn and pattern the same length + */ + fv += (flen - plen); /* flen - plen >= 0 due to check above */ + return simple_domaincmp(pv, fv, plen); + } + else if (unanchored == 0) + { + /* Fully anchored, check length */ + if (flen != plen) { - if(pn || !(pattern->unanchored & ANCHOR_LEFT)) - { - return 1; - } + return 1; } - else + return simple_domaincmp(pv, fv, plen); + } + else if (unanchored == ANCHOR_RIGHT) + { + /* Left anchored, ignore all extra in fqdn */ + return simple_domaincmp(pv, fv, plen); + } + else + { + /* Unanchored */ + int n; + int maxn = flen - plen; + for (n = 0; n <= maxn; n++) { - pn++; + if (!simple_domaincmp(pv, fv, plen)) + { + return 0; + } + /* + * Doesn't match from start of fqdn + * Try skipping first part of fqdn + */ + fv++; } - fn++; + return 1; } - return ((pn < pattern->dcnt) || ((fn < fqdn->dcnt) && !(pattern->unanchored & ANCHOR_RIGHT))); - } - /* Local Variables: tab-width: 3