[youtube] Simplify the decryption process for the manifest urls and add a test (close...

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

diff --git a/youtube_dl/extractor/youtube.py b/youtube_dl/extractor/youtube.py
index 8e768ea4fbad4cf7bd58703e43e853e478d8c450..e1ef90e384c437f635f0727014505bfde7b7b34e 100644 (file)
--- a/youtube_dl/extractor/youtube.py
+++ b/youtube_dl/extractor/youtube.py
@@ -297,6 +297,23 @@ class YoutubeIE(YoutubeBaseInfoExtractor, SubtitlesInfoExtractor):
                 u"format": "141",
             },
         },
+        # DASH manifest with encrypted signature
+        {
+            u'url': u'https://www.youtube.com/watch?v=IB3lcPjvWLA',
+            u'info_dict': {
+                u'id': u'IB3lcPjvWLA',
+                u'ext': u'm4a',
+                u'title': u'Afrojack - The Spark ft. Spree Wilson',
+                u'description': u'md5:3199ed45ee8836572865580804d7ac0f',
+                u'uploader': u'AfrojackVEVO',
+                u'uploader_id': u'AfrojackVEVO',
+                u'upload_date': u'20131011',
+            },
+            u"params": {
+                u'youtube_include_dash_manifest': True,
+                u'format': '141',
+            },
+        },
     ]
 
 
@@ -1272,8 +1289,8 @@ class YoutubeIE(YoutubeBaseInfoExtractor, SubtitlesInfoExtractor):
             mobj = re.search(r';ytplayer.config = ({.*?});', video_webpage)
             if not mobj:
                 raise ValueError('Could not find vevo ID')
-            info = json.loads(mobj.group(1))
-            args = info['args']
+            ytplayer_config = json.loads(mobj.group(1))
+            args = ytplayer_config['args']
             # Easy way to know if the 's' value is in url_encoded_fmt_stream_map
             # this signatures are encrypted
             if 'url_encoded_fmt_stream_map' not in args:
@@ -1366,12 +1383,24 @@ class YoutubeIE(YoutubeBaseInfoExtractor, SubtitlesInfoExtractor):
             raise ExtractorError(u'no conn, hlsvp or url_encoded_fmt_stream_map information found in video info')
 
         # Look for the DASH manifest
-        dash_manifest_url_lst = video_info.get('dashmpd')
-        if (dash_manifest_url_lst and dash_manifest_url_lst[0] and
-                self._downloader.params.get('youtube_include_dash_manifest', False)):
+        if (self._downloader.params.get('youtube_include_dash_manifest', False)):
             try:
+                # The DASH manifest used needs to be the one from the original video_webpage.
+                # The one found in get_video_info seems to be using different signatures.
+                # However, in the case of an age restriction there won't be any embedded dashmpd in the video_webpage.
+                # Luckily, it seems, this case uses some kind of default signature (len == 86), so the
+                # combination of get_video_info and the _static_decrypt_signature() decryption fallback will work here.
+                if age_gate:
+                    dash_manifest_url = video_info.get('dashmpd')[0]
+                else:
+                    dash_manifest_url = ytplayer_config['args']['dashmpd']
+                def decrypt_sig(mobj):
+                    s = mobj.group(1)
+                    dec_s = self._decrypt_signature(s, video_id, player_url, age_gate)
+                    return '/signature/%s' % dec_s
+                dash_manifest_url = re.sub(r'/s/([\w\.]+)', decrypt_sig, dash_manifest_url)
                 dash_doc = self._download_xml(
-                    dash_manifest_url_lst[0], video_id,
+                    dash_manifest_url, video_id,
                     note=u'Downloading DASH manifest',
                     errnote=u'Could not download DASH manifest')
                 for r in dash_doc.findall(u'.//{urn:mpeg:DASH:schema:MPD:2011}Representation'):