tweeper.php: fix validation when generating enclosure elements

[tweeper.git] / tweeper.php

diff --git a/tweeper.php b/tweeper.php
index 17fff8b..d084398 100644 (file)
--- a/tweeper.php
+++ b/tweeper.php
@@ -96,7 +96,11 @@ class Tweeper {
       "video/ogg",
     );
 
-    $url_info = Tweeper::get_info($url);
+    // The RSS specification says that the enclosure element url must be http.
+    // See http://sourceforge.net/p/feedvalidator/bugs/72/
+    $http_url = preg_replace("/^https/", "http", $url);
+
+    $url_info = Tweeper::get_info($http_url);
 
     $supported = in_array($url_info['content_type'], $supported_content_types);
     if (!$supported) {
@@ -164,6 +168,21 @@ class Tweeper {
     return $xsltProcessor;
   }
 
+  private function html_to_xml($html) {
+    $xmlDoc = new DOMDocument();
+
+    // Handle warnings and errors when loading invalid HTML.
+    $xml_errors_value = libxml_use_internal_errors(true);
+    $xmlDoc->loadHTML($html);
+    foreach (libxml_get_errors() as $xml_error) {
+      $this->log_xml_error($xml_error);
+    }
+    libxml_clear_errors();
+    libxml_use_internal_errors($xml_errors_value);
+
+    return $xmlDoc;
+  }
+
   public function tweep($src_url) {
     $url = parse_url($src_url);
     if (FALSE === $url || empty($url["host"])) {
@@ -176,18 +195,15 @@ class Tweeper {
       return NULL;
     }
 
-    $html = Tweeper::get_contents($src_url);
-
-    $xmlDoc = new DOMDocument();
+    $html = $this->get_contents($src_url);
+    if (FALSE === $html) {
+      return NULL;
+    }
 
-    // Handle warnings and errors when loading invalid HTML.
-    $xml_errors_value = libxml_use_internal_errors(true);
-    $xmlDoc->loadHTML($html);
-    foreach (libxml_get_errors() as $xml_error) {
-      $this->log_xml_error($xml_error);
+    $xmlDoc = $this->html_to_xml($html);
+    if (NULL === $xmlDoc) {
+      return NULL;
     }
-    libxml_clear_errors();
-    libxml_use_internal_errors($xml_errors_value);
 
     $output = $xsltProcessor->transformToXML($xmlDoc);