Fix urls interaction.

4 files changed, 58 insertions, 69 deletions

diff --git a/src/libmime/message.c b/src/libmime/message.c
index 702b148cb..2f96447bf 100644
--- a/src/libmime/message.c
+++ b/src/libmime/message.c
@@ -44,7 +44,7 @@ strip_html_tags (struct rspamd_task *task,
 	GByteArray * src,
 	gint *stateptr)
 {
-	uint8_t *p, *rp, *tbegin = NULL, *end, c, lc, *estart;
+	uint8_t *p, *rp, *tbegin = NULL, *end, c, lc, *estart = NULL;
 	gint br, i = 0, depth = 0, in_q = 0;
 	gint state = 0;
 	guint dlen;
@@ -1639,25 +1639,22 @@ process_message (struct rspamd_task *task)
 				if (url_str != NULL) {
 					subject_url = rspamd_mempool_alloc0 (task->task_pool,
 							sizeof (struct rspamd_url));
-					if (subject_url != NULL) {
-						/* Try to parse url */
-						rc = rspamd_url_parse (subject_url, url_str, task->task_pool);
-						if ((rc == URI_ERRNO_OK || rc == URI_ERRNO_NO_SLASHES ||
-							rc == URI_ERRNO_NO_HOST_SLASH) &&
-							subject_url->hostlen > 0) {
-							if (subject_url->protocol != PROTOCOL_MAILTO) {
-								if (!g_tree_lookup (task->urls, subject_url)) {
-									g_tree_insert (task->urls,
+					rc = rspamd_url_parse (subject_url, url_str,
+							strlen (url_str), task->task_pool);
+
+					if ((rc == URI_ERRNO_OK) && subject_url->hostlen > 0) {
+						if (subject_url->protocol != PROTOCOL_MAILTO) {
+							if (!g_tree_lookup (task->urls, subject_url)) {
+								g_tree_insert (task->urls,
 										subject_url,
 										subject_url);
-								}
 							}
 						}
-						else if (rc != URI_ERRNO_OK) {
-							msg_info ("extract of url '%s' failed: %s",
+					}
+					else if (rc != URI_ERRNO_OK) {
+						msg_info ("extract of url '%s' failed: %s",
 								url_str,
 								rspamd_url_strerror (rc));
-						}
 					}
 				}
 			}
diff --git a/src/libserver/html.c b/src/libserver/html.c
index 7df9270c3..fe90b7dc4 100644
--- a/src/libserver/html.c
+++ b/src/libserver/html.c
@@ -575,10 +575,7 @@ void
 decode_entitles (gchar *s, guint * len)
 {
 	guint l, rep_len;
-	gchar *t = s;                           /* t - tortoise */
-	gchar *h = s;                           /* h - hare     */
-	gchar *e = s;
-	gchar *end_ptr;
+	gchar *t = s, *h = s, *e = s, *end_ptr;
 	gint state = 0, val, base;
 	entity *found, key;
 
@@ -735,45 +732,38 @@ check_phishing (struct rspamd_task *task,
 	if (rspamd_url_find (task->task_pool, url_text, len, NULL, NULL, &url_str,
 		TRUE) && url_str != NULL) {
 		new = rspamd_mempool_alloc0 (task->task_pool, sizeof (struct rspamd_url));
-		if (new != NULL) {
-			g_strstrip (url_str);
-			rc = rspamd_url_parse (new, url_str, task->task_pool);
+		g_strstrip (url_str);
+		rc = rspamd_url_parse (new, url_str, strlen (url_str), task->task_pool);
 
-			if (rc == URI_ERRNO_OK || rc == URI_ERRNO_NO_SLASHES || rc ==
-				URI_ERRNO_NO_HOST_SLASH) {
-				if (g_ascii_strncasecmp (href_url->host, new->host,
+		if (rc == URI_ERRNO_OK) {
+			if (g_ascii_strncasecmp (href_url->host, new->host,
 					MAX (href_url->hostlen, new->hostlen)) != 0) {
-					/* Special check for urls beginning with 'www' */
-					if (new->hostlen > 4 && href_url->hostlen > 4) {
-						p = new->host;
-						c = NULL;
-						if ((p[0] == 'w' || p[0] == 'W') &&
+				/* Special check for urls beginning with 'www' */
+				if (new->hostlen > 4 && href_url->hostlen > 4) {
+					p = new->host;
+					c = NULL;
+					if ((p[0] == 'w' || p[0] == 'W') &&
 							(p[1] == 'w' || p[1] == 'W') &&
 							(p[2] == 'w' || p[2] == 'W') &&
 							(p[3] == '.')) {
-							p += 4;
-							c = href_url->host;
-							len = MAX (href_url->hostlen, new->hostlen - 4);
-						}
-						else {
-							p = href_url->host;
-							if ((p[0] == 'w' || p[0] == 'W') &&
+						p += 4;
+						c = href_url->host;
+						len = MAX (href_url->hostlen, new->hostlen - 4);
+					}
+					else {
+						p = href_url->host;
+						if ((p[0] == 'w' || p[0] == 'W') &&
 								(p[1] == 'w' || p[1] == 'W') &&
 								(p[2] == 'w' || p[2] == 'W') &&
 								(p[3] == '.')) {
-								p += 4;
-								c = new->host;
-								len = MAX (href_url->hostlen - 4, new->hostlen);
-							}
-						}
-						/* Compare parts and check for phished hostname */
-						if (c != NULL) {
-							if (g_ascii_strncasecmp (p, c, len) != 0) {
-								href_url->is_phished = TRUE;
-								href_url->phished_url = new;
-							}
+							p += 4;
+							c = new->host;
+							len = MAX (href_url->hostlen - 4, new->hostlen);
 						}
-						else {
+					}
+					/* Compare parts and check for phished hostname */
+					if (c != NULL) {
+						if (g_ascii_strncasecmp (p, c, len) != 0) {
 							href_url->is_phished = TRUE;
 							href_url->phished_url = new;
 						}
@@ -783,12 +773,16 @@ check_phishing (struct rspamd_task *task,
 						href_url->phished_url = new;
 					}
 				}
+				else {
+					href_url->is_phished = TRUE;
+					href_url->phished_url = new;
+				}
 			}
-			else {
-				msg_info ("extract of url '%s' failed: %s",
+		}
+		else {
+			msg_info ("extract of url '%s' failed: %s",
 					url_str,
 					rspamd_url_strerror (rc));
-			}
 		}
 	}
 
@@ -871,8 +865,7 @@ parse_tag_url (struct rspamd_task *task,
 
 		url_text = rspamd_mempool_alloc (task->task_pool, len + 1);
 		rspamd_strlcpy (url_text, c, len + 1);
-		rspamd_url_unescape (url_text);
-		decode_entitles (url_text, NULL);
+		decode_entitles (url_text, &len);
 
 		if (g_ascii_strncasecmp (url_text, "http",
 			sizeof ("http") - 1) != 0 &&
@@ -882,14 +875,14 @@ parse_tag_url (struct rspamd_task *task,
 			sizeof ("ftp://") - 1) != 0 &&
 			g_ascii_strncasecmp (url_text, "mailto:",
 			sizeof ("mailto:") - 1) != 0) {
+
 			return;
 		}
 
 		url = rspamd_mempool_alloc (task->task_pool, sizeof (struct rspamd_url));
-		rc = rspamd_url_parse (url, url_text, task->task_pool);
+		rc = rspamd_url_parse (url, url_text, len, task->task_pool);
 
-		if (rc != URI_ERRNO_EMPTY && rc != URI_ERRNO_NO_HOST && url->hostlen !=
-			0) {
+		if (rc != URI_ERRNO_EMPTY && url->hostlen != 0) {
 			/*
 			 * Check for phishing
 			 */
diff --git a/src/libserver/url.h b/src/libserver/url.h
index db3a3472c..f2aed42c6 100644
--- a/src/libserver/url.h
+++ b/src/libserver/url.h
@@ -108,9 +108,4 @@ gboolean rspamd_url_find (rspamd_mempool_t *pool,
  */
 const gchar * rspamd_url_strerror (enum uri_errno err);
 
-/*
- * URL unescape characters in the specified string
- */
-void rspamd_url_unescape (gchar *s);
-
 #endif
diff --git a/src/plugins/surbl.c b/src/plugins/surbl.c
index df9227c08..119a4f8cb 100644
--- a/src/plugins/surbl.c
+++ b/src/plugins/surbl.c
@@ -867,10 +867,10 @@ redirector_callback (gint fd, short what, void *arg)
 {
 	struct redirector_param *param = (struct redirector_param *)arg;
 	gchar url_buf[512];
-	gint r;
+	gint r, urllen;
 	struct timeval *timeout;
 	struct rspamd_task *task;
-	gchar *p, *c;
+	gchar *p, *c, *urlstr;
 	gboolean found = FALSE;
 
 	task = param->task;
@@ -947,17 +947,21 @@ redirector_callback (gint fd, short what, void *arg)
 						break;
 					}
 				}
+
 				if (found) {
 					debug_task ("<%s> got reply from redirector: '%s' -> '%s'",
 						param->task->message_id,
 						struri (param->url),
 						c);
-					r =
-						rspamd_url_parse (param->url,
-							rspamd_mempool_strdup (param->task->task_pool,
-							c), param->task->task_pool);
-					if (r == URI_ERRNO_OK || r == URI_ERRNO_NO_SLASHES || r ==
-						URI_ERRNO_NO_HOST_SLASH) {
+
+					urllen = strlen (c);
+					urlstr = rspamd_mempool_alloc (param->task->task_pool,
+							urllen + 1);
+					rspamd_strlcpy (urlstr, c, urllen + 1);
+					r = rspamd_url_parse (param->url, urlstr, urllen,
+							param->task->task_pool);
+
+					if (r == URI_ERRNO_OK) {
 						make_surbl_requests (param->url,
 							param->task,
 							param->suffix,