[youtube] Fix extraction.

[youtube-dl] / youtube_dl / extractor / jeuxvideo.py

diff --git a/youtube_dl/extractor/jeuxvideo.py b/youtube_dl/extractor/jeuxvideo.py
index 4327bc13dab9c913eca8c4c263b6405e0dcc9076..e9f4ed7384225456edf87baf0c24ce5c6ecc4b20 100644 (file)
--- a/youtube_dl/extractor/jeuxvideo.py
+++ b/youtube_dl/extractor/jeuxvideo.py
@@ -1,47 +1,56 @@
 # coding: utf-8
 
-import json
+from __future__ import unicode_literals
+
 import re
-import xml.etree.ElementTree
 
 from .common import InfoExtractor
 
+
 class JeuxVideoIE(InfoExtractor):
-    _VALID_URL = r'http://.*?\.jeuxvideo\.com/.*/(.*?)-\d+\.htm'
-
-    _TEST = {
-        u'url': u'http://www.jeuxvideo.com/reportages-videos-jeux/0004/00046170/tearaway-playstation-vita-gc-2013-tearaway-nous-presente-ses-papiers-d-identite-00115182.htm',
-        u'file': u'5182.mp4',
-        u'md5': u'e0fdb0cd3ce98713ef9c1e1e025779d0',
-        u'info_dict': {
-            u'title': u'GC 2013 : Tearaway nous présente ses papiers d\'identité',
-            u'description': u'Lorsque les développeurs de LittleBigPlanet proposent un nouveau titre, on ne peut que s\'attendre à un résultat original et fort attrayant.\n',
+    _VALID_URL = r'https?://.*?\.jeuxvideo\.com/.*/(.*?)\.htm'
+
+    _TESTS = [{
+        'url': 'http://www.jeuxvideo.com/reportages-videos-jeux/0004/00046170/tearaway-playstation-vita-gc-2013-tearaway-nous-presente-ses-papiers-d-identite-00115182.htm',
+        'md5': '046e491afb32a8aaac1f44dd4ddd54ee',
+        'info_dict': {
+            'id': '114765',
+            'ext': 'mp4',
+            'title': 'Tearaway : GC 2013 : Tearaway nous présente ses papiers d\'identité',
+            'description': 'Lorsque les développeurs de LittleBigPlanet proposent un nouveau titre, on ne peut que s\'attendre à un résultat original et fort attrayant.',
         },
-    }
+    }, {
+        'url': 'http://www.jeuxvideo.com/videos/chroniques/434220/l-histoire-du-jeu-video-la-saturn.htm',
+        'only_matching': True,
+    }]
 
     def _real_extract(self, url):
         mobj = re.match(self._VALID_URL, url)
-        title = re.match(self._VALID_URL, url).group(1)
+        title = mobj.group(1)
         webpage = self._download_webpage(url, title)
-        m_download = re.search(r'<param name="flashvars" value="config=(.*?)" />', webpage)
-
-        xml_link = m_download.group(1)
-        
-        id = re.search(r'http://www.jeuxvideo.com/config/\w+/0011/(.*?)/\d+_player\.xml', xml_link).group(1)
-
-        xml_config = self._download_webpage(xml_link, title,
-                                                  'Downloading XML config')
-        config = xml.etree.ElementTree.fromstring(xml_config.encode('utf-8'))
-        info = re.search(r'<format\.json>(.*?)</format\.json>',
-                         xml_config, re.MULTILINE|re.DOTALL).group(1)
-        info = json.loads(info)['versions'][0]
-        
-        video_url = 'http://video720.jeuxvideo.com/' + info['file']
-
-        return {'id': id,
-                'title' : config.find('titre_video').text,
-                'ext' : 'mp4',
-                'url' : video_url,
-                'description': self._og_search_description(webpage),
-                'thumbnail': config.find('image').text,
-                }
+        title = self._html_search_meta('name', webpage) or self._og_search_title(webpage)
+        config_url = self._html_search_regex(
+            r'data-src(?:set-video)?="(/contenu/medias/video\.php.*?)"',
+            webpage, 'config URL')
+        config_url = 'http://www.jeuxvideo.com' + config_url
+
+        video_id = self._search_regex(
+            r'id=(\d+)',
+            config_url, 'video ID')
+
+        config = self._download_json(
+            config_url, title, 'Downloading JSON config')
+
+        formats = [{
+            'url': source['file'],
+            'format_id': source['label'],
+            'resolution': source['label'],
+        } for source in reversed(config['sources'])]
+
+        return {
+            'id': video_id,
+            'title': title,
+            'formats': formats,
+            'description': self._og_search_description(webpage),
+            'thumbnail': config.get('image'),
+        }