Fix extraction from youtube.

[youtubedl] / youtube_dl / extractor / videofyme.py

diff --git a/youtube_dl/extractor/videofyme.py b/youtube_dl/extractor/videofyme.py
index 912802d9aa22082f2f39148db7920a6287c74ec6..cd3f50a63b70745b157dfd4d2f67549ead6d0de2 100644 (file)
--- a/youtube_dl/extractor/videofyme.py
+++ b/youtube_dl/extractor/videofyme.py
@@ -1,46 +1,52 @@
-import re
+from __future__ import unicode_literals
 
 from .common import InfoExtractor
 from ..utils import (
-    find_xpath_attr,
-    determine_ext,
+    int_or_none,
+    parse_iso8601,
 )
 
+
 class VideofyMeIE(InfoExtractor):
-    _VALID_URL = r'https?://(www.videofy.me/.+?|p.videofy.me/v)/(?P<id>\d+)(&|#|$)'
-    IE_NAME = u'videofy.me'
+    _VALID_URL = r'https?://(?:www\.videofy\.me/.+?|p\.videofy\.me/v)/(?P<id>\d+)(&|#|$)'
+    IE_NAME = 'videofy.me'
 
     _TEST = {
-        u'url': u'http://www.videofy.me/thisisvideofyme/1100701',
-        u'file':  u'1100701.mp4',
-        u'md5': u'c77d700bdc16ae2e9f3c26019bd96143',
-        u'info_dict': {
-            u'title': u'This is VideofyMe',
-            u'description': None,
-            u'uploader': u'VideofyMe',
-            u'uploader_id': u'thisisvideofyme',
+        'url': 'http://www.videofy.me/thisisvideofyme/1100701',
+        'md5': 'c77d700bdc16ae2e9f3c26019bd96143',
+        'info_dict': {
+            'id': '1100701',
+            'ext': 'mp4',
+            'title': 'This is VideofyMe',
+            'description': '',
+            'upload_date': '20130326',
+            'timestamp': 1364288959,
+            'uploader': 'VideofyMe',
+            'uploader_id': 'thisisvideofyme',
+            'view_count': int,
+            'likes': int,
+            'comment_count': int,
         },
-        
     }
 
     def _real_extract(self, url):
-        mobj = re.match(self._VALID_URL, url)
-        video_id = mobj.group('id')
-        config = self._download_xml('http://sunshine.videofy.me/?videoId=%s' % video_id,
-                                            video_id)
-        video = config.find('video')
-        sources = video.find('sources')
-        url_node = next(node for node in [find_xpath_attr(sources, 'source', 'id', 'HQ %s' % key) 
-            for key in ['on', 'av', 'off']] if node is not None)
-        video_url = url_node.find('url').text
-
-        return {'id': video_id,
-                'title': video.find('title').text,
-                'url': video_url,
-                'ext': determine_ext(video_url),
-                'thumbnail': video.find('thumb').text,
-                'description': video.find('description').text,
-                'uploader': config.find('blog/name').text,
-                'uploader_id': video.find('identifier').text,
-                'view_count': re.search(r'\d+', video.find('views').text).group(),
-                }
+        video_id = self._match_id(url)
+
+        config = self._download_json('http://vf-player-info-loader.herokuapp.com/%s.json' % video_id, video_id)['videoinfo']
+
+        video = config.get('video')
+        blog = config.get('blog', {})
+
+        return {
+            'id': video_id,
+            'title': video['title'],
+            'url': video['sources']['source']['url'],
+            'thumbnail': video.get('thumb'),
+            'description': video.get('description'),
+            'timestamp': parse_iso8601(video.get('date')),
+            'uploader': blog.get('name'),
+            'uploader_id': blog.get('identifier'),
+            'view_count': int_or_none(self._search_regex(r'([0-9]+)', video.get('views'), 'view count', fatal=False)),
+            'likes': int_or_none(video.get('likes')),
+            'comment_count': int_or_none(video.get('nrOfComments')),
+        }