[cnn] Improve thumbnail extraction

rzr · Jun 7, 2014 · be6d722 · be6d722
1 parent d551980
commit be6d722
Show file tree

Hide file tree

Showing 2 changed files with 8 additions and 4 deletions.
diff --git a/youtube_dl/YoutubeDL.py b/youtube_dl/YoutubeDL.py
@@ -719,6 +719,8 @@ def process_video_result(self, info_dict, download=True):
 
         thumbnails = info_dict.get('thumbnails')
         if thumbnails:
+            thumbnails.sort(key=lambda t: (
+                t.get('width'), t.get('height'), t.get('url')))
             for t in thumbnails:
                 if 'width' in t and 'height' in t:
                     t['resolution'] = '%dx%d' % (t['width'], t['height'])

diff --git a/youtube_dl/extractor/cnn.py b/youtube_dl/extractor/cnn.py
@@ -79,8 +79,11 @@ def _real_extract(self, url):
 
         self._sort_formats(formats)
 
-        thumbnails = sorted([((int(t.attrib['height']),int(t.attrib['width'])), t.text) for t in info.findall('images/image')])
-        thumbs_dict = [{'resolution': res, 'url': t_url} for (res, t_url) in thumbnails]
+        thumbnails = [{
+            'height': int(t.attrib['height']),
+            'width': int(t.attrib['width']),
+            'url': t.text,
+        } for t in info.findall('images/image')]
 
         metas_el = info.find('metas')
         upload_date = (
@@ -93,8 +96,7 @@ def _real_extract(self, url):
             'id': info.attrib['id'],
             'title': info.find('headline').text,
             'formats': formats,
-            'thumbnail': thumbnails[-1][1],
-            'thumbnails': thumbs_dict,
+            'thumbnails': thumbnails,
             'description': info.find('description').text,
             'duration': duration,
             'upload_date': upload_date,