subscribe: verify XML before adding to the database; fetch: try to work around entity problems if initial parsing fails

2 files changed, 47 insertions, 1 deletions

diff --git a/include/functions.php b/include/functions.php
index 4cc8f134d..8ac5753c9 100644
--- a/include/functions.php
+++ b/include/functions.php
@@ -1558,6 +1558,7 @@
 	 *                     Here you should call extractfeedurls in rpc-backend
 	 *                     to get all possible feeds.
 	 *                 5 - Couldn't download the URL content.
+	 *                 6 - Content is an invalid XML.
 	 */
 	function subscribe_to_feed($url, $cat_id = 0,
 			$auth_login = '', $auth_pass = '') {
@@ -1588,6 +1589,18 @@
 			$url = key($feedUrls);
 		}
 
+		libxml_use_internal_errors(true);
+		$doc = new DOMDocument();
+		$doc->loadXML(html_entity_decode($contents));
+		$error = libxml_get_last_error();
+		libxml_clear_errors();
+
+		if ($error) {
+			$error_message = format_libxml_error($error);
+
+			return array("code" => 6, "message" => $error_message);
+		}
+
 		if ($cat_id == "0" || !$cat_id) {
 			$cat_qpart = "NULL";
 		} else {
@@ -4203,4 +4216,10 @@
 		return LABEL_BASE_INDEX - 1 + abs($feed);
 	}
 
+	function format_libxml_error($error) {
+		return T_sprintf("LibXML error %s at line %d (column %d): %s",
+				$error->code, $error->line, $error->column,
+				$error->message);
+	}
+
 ?>
diff --git a/include/rssfuncs.php b/include/rssfuncs.php
index 31d35bf8e..47d622169 100644
--- a/include/rssfuncs.php
+++ b/include/rssfuncs.php
@@ -316,6 +316,25 @@
 					_debug("update_rss_feed: fetch done.");
 				}
 
+				$error = verify_feed_xml($feed_data);
+
+				if ($error) {
+					if ($debug_enabled) {
+						_debug("update_rss_feed: error verifying XML, code: " . $error->code);
+					}
+
+					if ($error->code == 26) {
+						if ($debug_enabled) {
+							_debug("update_rss_feed: got error 26, trying to decode entities...");
+						}
+
+						$feed_data = html_entity_decode($feed_data, ENT_COMPAT, 'UTF-8');
+
+						$error = verify_feed_xml($feed_data);
+
+						if ($error) $feed_data = '';
+					}
+				}
 			}
 
 			if (!$feed_data) {
@@ -559,7 +578,7 @@
 					_debug("update_rss_feed: date $entry_timestamp [$entry_timestamp_fmt]");
 				}
 
-				$entry_title = html_entity_decode($item->get_title());
+				$entry_title = html_entity_decode($item->get_title(), ENT_COMPAT, 'UTF-8');
 
 				$entry_link = rewrite_relative_url($site_url, $item->get_link());
 
@@ -1421,5 +1440,13 @@
 			mb_strtolower(strip_tags($title), 'utf-8'));
 	}
 
+	function verify_feed_xml($feed_data) {
+		libxml_use_internal_errors(true);
+		$doc = new DOMDocument();
+		$doc->loadXML($feed_data);
+		$error = libxml_get_last_error();
+		libxml_clear_errors();
+		return $error;
+	}
 
 ?>