Validate urls before adding to queue.

Prevents invalid url exceptions while parsing recursively
VIPnytt · Aug 10, 2019 · e793cae · e793cae
1 parent a10bd4a
commit e793cae
Showing 1 changed file with 9 additions and 3 deletions.
diff --git a/src/SitemapParser.php b/src/SitemapParser.php
@@ -146,7 +146,10 @@ public function parseRecursive($url)
     public function addToQueue(array $urlArray)
     {
         foreach ($urlArray as $url) {
-            $this->queue[] = $url;
+            $url = $this->urlEncode($url);
+            if ($this->urlValidate($url)) {
+                $this->queue[] = $url;
+            }
         }
     }
 
@@ -173,10 +176,13 @@ public function getQueue()
     public function parse($url, $urlContent = null)
     {
         $this->clean();
-        $this->currentURL = $url;
+        $this->currentURL = $this->urlEncode($url);
+        if (!$this->urlValidate($this->currentURL)) {
+            throw new Exceptions\SitemapParserException('Invalid URL');
+        }
         $this->history[] = $this->currentURL;
         $response = is_string($urlContent) ? $urlContent : $this->getContent();
-        if ($this->urlValidate($this->currentURL) && parse_url($this->currentURL, PHP_URL_PATH) === self::ROBOTSTXT_PATH) {
+        if (parse_url($this->currentURL, PHP_URL_PATH) === self::ROBOTSTXT_PATH) {
             $this->parseRobotstxt($response);
             return;
         }