/* urls can't contain null pointer, caller must ensure this */
static enum phish_status phishingCheck(const struct cl_engine* engine,struct url_check* urls)
{
struct url_check host_url;
enum phish_status rc=CL_PHISH_NODECISION;
int phishy=0;
const struct phishcheck* pchk = (const struct phishcheck*) engine->phishcheck;
if(!urls->realLink.data)
return CL_PHISH_CLEAN;
cli_dbgmsg("Phishcheck:Checking url %s->%s\n", urls->realLink.data,
urls->displayLink.data);
if(!strcmp(urls->realLink.data,urls->displayLink.data))
return CL_PHISH_CLEAN;/* displayed and real URL are identical -> clean */
if((rc = cleanupURLs(urls))) {
if(isPhishing(rc))/* not allowed to decide this is phishing */
return CL_PHISH_CLEAN;
return rc;/* URLs identical after cleanup */
}
if(whitelist_check(engine,urls,0))
return CL_PHISH_WHITELISTED;/* if url is whitelist don't perform further checks */
if((!isURL(pchk, urls->displayLink.data) || !isRealURL(pchk, urls->realLink.data) )&&
( (phishy&PHISHY_NUMERIC_IP && !isNumericURL(pchk, urls->displayLink.data)) ||
!(phishy&PHISHY_NUMERIC_IP))) {
cli_dbgmsg("Displayed 'url' is not url:%s\n",urls->displayLink.data);
return CL_PHISH_TEXTURL;
}
if(urls->flags&DOMAINLIST_REQUIRED && domainlist_match(engine,urls->realLink.data,urls->displayLink.data,NULL,0,&urls->flags))
phishy |= DOMAIN_LISTED;
else {
/* although entire url is not listed, the host might be,
* so defer phishing decisions till we know if host is listed*/
}
url_check_init(&host_url);
if((rc = url_get_host(pchk, urls,&host_url,DOMAIN_DISPLAY,&phishy))) {
free_if_needed(&host_url);
if(isPhishing(rc))
return CL_PHISH_CLEAN;
return rc;
}
if(urls->flags&DOMAINLIST_REQUIRED) {
if(!(phishy&DOMAIN_LISTED)) {
if(domainlist_match(engine,host_url.displayLink.data,host_url.realLink.data,&urls->pre_fixup,1,&urls->flags))
phishy |= DOMAIN_LISTED;
else {
}
}
}
/* link type filtering must occur after last domainlist_match */
if(urls->link_type & LINKTYPE_IMAGE && !(urls->flags&CHECK_IMG_URL))
return CL_PHISH_HOST_NOT_LISTED;/* its listed, but this link type is filtered */
if(urls->flags & DOMAINLIST_REQUIRED && !(phishy & DOMAIN_LISTED) ) {
urls->flags &= urls->always_check_flags;
if(!urls->flags) {
free_if_needed(&host_url);
return CL_PHISH_HOST_NOT_LISTED;
}
}
if(urls->flags&CHECK_CLOAKING) {
/*Checks if URL is cloaked.
Should we check if it contains another http://, https://?
No because we might get false positives from redirect services.*/
if(strchr(urls->realLink.data,0x1)) {
free_if_needed(&host_url);
return CL_PHISH_CLOAKED_NULL;
}
if(isEncoded(urls->displayLink.data)) {
free_if_needed(&host_url);
return CL_PHISH_HEX_URL;
}
}
if(urls->displayLink.data[0]=='\0') {
free_if_needed(&host_url);
return CL_PHISH_CLEAN;
}
if(urls->flags&CHECK_SSL && isSSL(urls->displayLink.data) && !isSSL(urls->realLink.data)) {
free_if_needed(&host_url);
return CL_PHISH_SSL_SPOOF;
}
if(!urls->flags&CHECK_CLOAKING && urls->flags & DOMAINLIST_REQUIRED && !(phishy&DOMAIN_LISTED) ) {
free_if_needed(&host_url);
return CL_PHISH_HOST_NOT_LISTED;
}
if((rc = url_get_host(pchk, urls,&host_url,DOMAIN_REAL,&phishy)))
{
free_if_needed(&host_url);
return rc;
}
if(urls->flags&DOMAINLIST_REQUIRED && !(phishy&DOMAIN_LISTED)) {
free_if_needed(&host_url);
return CL_PHISH_HOST_NOT_LISTED;
}
if(whitelist_check(engine,&host_url,1)) {
free_if_needed(&host_url);
return CL_PHISH_HOST_WHITELISTED;
}
if(urls->flags&HOST_SUFFICIENT) {
if(!strcmp(urls->realLink.data,urls->displayLink.data)) {
free_if_needed(&host_url);
return CL_PHISH_HOST_OK;
}
if(urls->flags&DOMAIN_SUFFICIENT) {
struct url_check domain_url;
url_check_init(&domain_url);
url_get_domain(pchk, &host_url,&domain_url);
if(!strcmp(domain_url.realLink.data,domain_url.displayLink.data)) {
free_if_needed(&host_url);
free_if_needed(&domain_url);
return CL_PHISH_DOMAIN_OK;
}
free_if_needed(&domain_url);
}
free_if_needed(&host_url);
}/*HOST_SUFFICIENT*/
/*we failed to find a reason why the 2 URLs are different, this is definitely phishing*/
if(urls->flags&DOMAINLIST_REQUIRED && !(phishy&DOMAIN_LISTED))
return CL_PHISH_HOST_NOT_LISTED;
return phishy_map(phishy,CL_PHISH_NOMATCH);
}
/* urls can't contain null pointer, caller must ensure this */
static enum phish_status phishingCheck(const struct cl_engine* engine,struct url_check* urls)
{
struct url_check host_url;
int rc = CL_PHISH_NODECISION;
int phishy=0;
const struct phishcheck* pchk = (const struct phishcheck*) engine->phishcheck;
if(!urls->realLink.data)
return CL_PHISH_CLEAN;
cli_dbgmsg("Phishcheck:Checking url %s->%s\n", urls->realLink.data,
urls->displayLink.data);
if(!isURL(urls->realLink.data, 0)) {
cli_dbgmsg("Real 'url' is not url:%s\n",urls->realLink.data);
return CL_PHISH_CLEAN;
}
if(( rc = url_hash_match(engine->domainlist_matcher, urls->realLink.data, strlen(urls->realLink.data)) )) {
if (rc == CL_PHISH_CLEAN) {
cli_dbgmsg("not analyzing, not a real url: %s\n", urls->realLink.data);
return CL_PHISH_CLEAN;
} else {
cli_dbgmsg("Hash matched for: %s\n", urls->realLink.data);
return rc;
}
}
if(!strcmp(urls->realLink.data,urls->displayLink.data))
return CL_PHISH_CLEAN;/* displayed and real URL are identical -> clean */
if (urls->displayLink.data[0] == '\0') {
return CL_PHISH_CLEAN;
}
if((rc = cleanupURLs(urls))) {
/* it can only return an error, or say its clean;
* it is not allowed to decide it is phishing */
return rc < 0 ? rc : CL_PHISH_CLEAN;
}
cli_dbgmsg("Phishcheck:URL after cleanup: %s->%s\n", urls->realLink.data,
urls->displayLink.data);
if((!isURL(urls->displayLink.data, 1) ) &&
( (phishy&PHISHY_NUMERIC_IP && !isNumericURL(pchk, urls->displayLink.data)) ||
!(phishy&PHISHY_NUMERIC_IP))) {
cli_dbgmsg("Displayed 'url' is not url:%s\n",urls->displayLink.data);
return CL_PHISH_CLEAN;
}
if(whitelist_check(engine, urls, 0))
return CL_PHISH_CLEAN;/* if url is whitelisted don't perform further checks */
url_check_init(&host_url);
if((rc = url_get_host(urls, &host_url, DOMAIN_DISPLAY, &phishy))) {
free_if_needed(&host_url);
return rc < 0 ? rc : CL_PHISH_CLEAN;
}
if (domainlist_match(engine, host_url.displayLink.data,host_url.realLink.data,&urls->pre_fixup,1)) {
phishy |= DOMAIN_LISTED;
} else {
urls->flags &= urls->always_check_flags;
/* don't return, we may need to check for ssl/cloaking */
}
/* link type filtering must occur after last domainlist_match */
if(urls->link_type & LINKTYPE_IMAGE && !(urls->flags&CHECK_IMG_URL)) {
free_if_needed(&host_url);
return CL_PHISH_CLEAN;/* its listed, but this link type is filtered */
}
if(urls->flags&CHECK_CLOAKING) {
/*Checks if URL is cloaked.
Should we check if it contains another http://, https://?
No because we might get false positives from redirect services.*/
if(strchr(urls->realLink.data,0x1)) {
free_if_needed(&host_url);
return CL_PHISH_CLOAKED_NULL;
}
}
if(urls->flags&CHECK_SSL && isSSL(urls->displayLink.data) && !isSSL(urls->realLink.data)) {
free_if_needed(&host_url);
return CL_PHISH_SSL_SPOOF;
}
if (!(phishy & DOMAIN_LISTED)) {
free_if_needed(&host_url);
return CL_PHISH_CLEAN;
}
if((rc = url_get_host(urls,&host_url,DOMAIN_REAL,&phishy)))
{
free_if_needed(&host_url);
return rc < 0 ? rc : CL_PHISH_CLEAN;
}
if(whitelist_check(engine,&host_url,1)) {
free_if_needed(&host_url);
return CL_PHISH_CLEAN;
}
if(!strcmp(urls->realLink.data,urls->displayLink.data)) {
free_if_needed(&host_url);
return CL_PHISH_CLEAN;
}
{
struct url_check domain_url;
url_check_init(&domain_url);
url_get_domain(&host_url,&domain_url);
if(!strcmp(domain_url.realLink.data,domain_url.displayLink.data)) {
free_if_needed(&host_url);
free_if_needed(&domain_url);
return CL_PHISH_CLEAN;
}
free_if_needed(&domain_url);
}
free_if_needed(&host_url);
/*we failed to find a reason why the 2 URLs are different, this is definitely phishing*/
return phishy_map(phishy,CL_PHISH_NOMATCH);
}