Fixes #410 - Retrieve title fails in multiple cases

  * `get_http_url()` renamed to `get_http_response()`.
  * Use the same HTTP context to retrieve response headers and content.
  * Follow HTTP 301 and 302 redirections to retrieve the title (default max 3 redirections).
  * Add `LinkUtils` to extract titles and charset.
  * Try to retrieve charset from HTTP headers first (new), then HTML content.
  * Use mb_string to re-encode title if necessary.

1 files changed, 42 insertions, 7 deletions

diff --git a/application/HttpUtils.php b/application/HttpUtils.php
index 499220c5..e2c1cb47 100644..100755
--- a/application/HttpUtils.php
+++ b/application/HttpUtils.php
@@ -13,7 +13,7 @@
  *  [1] = URL content (downloaded data)
  *
  * Example:
- *  list($headers, $data) = get_http_url('http://sebauvage.net/');
+ *  list($headers, $data) = get_http_response('http://sebauvage.net/');
  *  if (strpos($headers[0], '200 OK') !== false) {
  *      echo 'Data type: '.htmlspecialchars($headers['Content-Type']);
  *  } else {
@@ -24,31 +24,66 @@
  * @see http://php.net/manual/en/function.stream-context-create.php
  * @see http://php.net/manual/en/function.get-headers.php
  */
-function get_http_url($url, $timeout = 30, $maxBytes = 4194304)
+function get_http_response($url, $timeout = 30, $maxBytes = 4194304)
 {
+    $urlObj = new Url($url);
+    if (! filter_var($url, FILTER_VALIDATE_URL) || ! $urlObj->isHttp()) {
+        return array(array(0 => 'Invalid HTTP Url'), false);
+    }
+
     $options = array(
         'http' => array(
             'method' => 'GET',
             'timeout' => $timeout,
             'user_agent' => 'Mozilla/5.0 (X11; Ubuntu; Linux x86_64; rv:23.0)'
-                         .' Gecko/20100101 Firefox/23.0'
+                         .' Gecko/20100101 Firefox/23.0',
+            'request_fulluri' => true,
         )
     );
 
     $context = stream_context_create($options);
+    stream_context_set_default($options);
+
+    list($headers, $finalUrl) = get_redirected_headers($urlObj->cleanup());
+    if (! $headers || strpos($headers[0], '200 OK') === false) {
+        return array($headers, false);
+    }
 
     try {
         // TODO: catch Exception in calling code (thumbnailer)
-        $content = file_get_contents($url, false, $context, -1, $maxBytes);
+        $content = file_get_contents($finalUrl, false, $context, -1, $maxBytes);
     } catch (Exception $exc) {
         return array(array(0 => 'HTTP Error'), $exc->getMessage());
     }
 
-    if (!$content) {
-        return array(array(0 => 'HTTP Error'), '');
+    return array($headers, $content);
+}
+
+/**
+ * Retrieve HTTP headers, following n redirections (temporary and permanent).
+ *
+ * @param string $url initial URL to reach.
+ * @param int $redirectionLimit max redirection follow..
+ *
+ * @return array
+ */
+function get_redirected_headers($url, $redirectionLimit = 3)
+{
+    $headers = get_headers($url, 1);
+
+    // Headers found, redirection found, and limit not reached.
+    if ($redirectionLimit-- > 0
+        && !empty($headers)
+        && (strpos($headers[0], '301') !== false || strpos($headers[0], '302') !== false)
+        && !empty($headers['Location'])) {
+
+        $redirection = is_array($headers['Location']) ? end($headers['Location']) : $headers['Location'];
+        if ($redirection != $url) {
+            return get_redirected_headers($redirection, $redirectionLimit);
+        }
     }
 
-    return array(get_headers($url, 1), $content);
+    return array($headers, $url);
 }
 
 /**