Merge pull request #3965 from nicofrand/previewPic

Preview picture: use the 1st pic retrieved if no og:image set

src/Wallabag/CoreBundle/Helper/ContentProxy.php

@@ -12,8 +12,8 @@ use Wallabag\CoreBundle\Entity\Entry;
 use Wallabag\CoreBundle\Tools\Utils;
 
 /**
- * This kind of proxy class take care of getting the content from an url
- * and update the entry with what it found.
+ * This kind of proxy class takes care of getting the content from an url
+ * and updates the entry with what it found.
  */
 class ContentProxy
 {
@@ -289,13 +289,25 @@ class ContentProxy
             $this->updateLanguage($entry, $content['language']);
         }
 
+        $previewPictureUrl = '';
         if (!empty($content['open_graph']['og_image'])) {
-            $this->updatePreviewPicture($entry, $content['open_graph']['og_image']);
+            $previewPictureUrl = $content['open_graph']['og_image'];
         }
 
         // if content is an image, define it as a preview too
         if (!empty($content['content_type']) && \in_array($this->mimeGuesser->guess($content['content_type']), ['jpeg', 'jpg', 'gif', 'png'], true)) {
-            $this->updatePreviewPicture($entry, $content['url']);
+            $previewPictureUrl = $content['url'];
+        } elseif (empty($previewPictureUrl)) {
+            $this->logger->debug('Extracting images from content to provide a default preview picture');
+            $imagesUrls = DownloadImages::extractImagesUrlsFromHtml($content['html']);
+            $this->logger->debug(\count($imagesUrls) . ' pictures found');
+            if (!empty($imagesUrls)) {
+                $previewPictureUrl = $imagesUrls[0];
+            }
+        }
+
+        if (!empty($previewPictureUrl)) {
+            $this->updatePreviewPicture($entry, $previewPictureUrl);
         }
 
         if (!empty($content['content_type'])) {

src/Wallabag/CoreBundle/Helper/DownloadImages.php

@@ -30,6 +30,25 @@ class DownloadImages
         $this->setFolder();
     }
 
+    /**
+     * Process the html and extract images URLs from it.
+     *
+     * @param string $html
+     *
+     * @return string[]
+     */
+    public static function extractImagesUrlsFromHtml($html)
+    {
+        $crawler = new Crawler($html);
+        $imagesCrawler = $crawler
+            ->filterXpath('//img');
+        $imagesUrls = $imagesCrawler
+            ->extract(['src']);
+        $imagesSrcsetUrls = self::getSrcsetUrls($imagesCrawler);
+
+        return array_unique(array_merge($imagesUrls, $imagesSrcsetUrls));
+    }
+
     /**
      * Process the html and extract image from it, save them to local and return the updated html.
      *
@@ -41,13 +60,7 @@ class DownloadImages
      */
     public function processHtml($entryId, $html, $url)
     {
-        $crawler = new Crawler($html);
-        $imagesCrawler = $crawler
-            ->filterXpath('//img');
-        $imagesUrls = $imagesCrawler
-            ->extract(['src']);
-        $imagesSrcsetUrls = $this->getSrcsetUrls($imagesCrawler);
-        $imagesUrls = array_unique(array_merge($imagesUrls, $imagesSrcsetUrls));
+        $imagesUrls = self::extractImagesUrlsFromHtml($html);
 
         $relativePath = $this->getRelativePath($entryId);
 
@@ -199,7 +212,7 @@ class DownloadImages
      *
      * @return array An array of urls
      */
-    private function getSrcsetUrls(Crawler $imagesCrawler)
+    private static function getSrcsetUrls(Crawler $imagesCrawler)
     {
         $urls = [];
         $iterator = $imagesCrawler