From 24d41e2c6e759baf17a10aa2e48d4b1907d5c7a0 Mon Sep 17 00:00:00 2001
From: Friendika <info@friendika.com>
Date: Mon, 4 Jul 2011 23:02:04 -0700
Subject: [PATCH] purify html before trying to parse wild urls. This way at
 least it should parse.

---
 mod/parse_url.php | 17 +++++++++++++----
 1 file changed, 13 insertions(+), 4 deletions(-)

diff --git a/mod/parse_url.php b/mod/parse_url.php
index 15a6aced0e..ec28d74111 100644
--- a/mod/parse_url.php
+++ b/mod/parse_url.php
@@ -1,6 +1,7 @@
 <?php
 
 require_once('library/HTML5/Parser.php');
+require_once('library/HTMLPurifier.auto.php');
 
 function parse_url_content(&$a) {
 
@@ -31,16 +32,25 @@ function parse_url_content(&$a) {
 		killme();
 	}
 
+	logger('parse_url: data: ' . $s, LOGGER_DATA);
 
 	if(! $s) {
 		echo sprintf($template,$url,$url,'');
 		killme();
 	}
 
+	$config = HTMLPurifier_Config::createDefault();
+	$config->set('Cache.DefinitionImpl', null);
+
+	$purifier = new HTMLPurifier($config);
+	$s = $purifier->purify($s);
+
 	$dom = @HTML5_Parser::parse($s);
 
-	if(! $dom)
-		return $ret;
+	if(! $dom) {
+		echo sprintf($template,$url,$url,'');
+		killme();
+	}
 
 	$items = $dom->getElementsByTagName('title');
 
@@ -51,7 +61,6 @@ function parse_url_content(&$a) {
 		}
 	}
 
-
 	$divs = $dom->getElementsByTagName('div');
 	if($divs) {
 		foreach($divs as $div) {
@@ -94,6 +103,6 @@ function parse_url_content(&$a) {
 		$text = '<br />' . $text;
 	}
 
-	echo sprintf($template,$url,$title,$text);
+	echo sprintf($template,$url,($title) ? $title : $url,$text);
 	killme();
 }