Merge pull request #7185 from remitamine/ooyala

[ooyala] extract more formats and metadata

youtube_dl/extractor/byutv.py

@@ -14,9 +14,10 @@ class BYUtvIE(InfoExtractor):
         'info_dict': {
             'id': 'studio-c-season-5-episode-5',
             'ext': 'mp4',
-            'description': 'md5:5438d33774b6bdc662f9485a340401cc',
+            'description': 'md5:e07269172baff037f8e8bf9956bc9747',
             'title': 'Season 5 Episode 5',
-            'thumbnail': 're:^https?://.*\.jpg$'
+            'thumbnail': 're:^https?://.*\.jpg$',
+            'duration': 1486486,
         },
         'params': {
             'skip_download': True,

youtube_dl/extractor/generic.py

@@ -339,6 +339,7 @@ class GenericIE(InfoExtractor):
                 'id': 'BwY2RxaTrTkslxOfcan0UCf0YqyvWysJ',
                 'ext': 'mp4',
                 'title': '2cc213299525360.mov',  # that's what we get
+                'duration': 238231,
             },
             'add_ie': ['Ooyala'],
         },
@@ -350,6 +351,7 @@ class GenericIE(InfoExtractor):
                 'ext': 'mp4',
                 'title': '"Steve Jobs: Man in the Machine" trailer',
                 'description': 'The first trailer for the Alex Gibney documentary "Steve Jobs: Man in the Machine."',
+                'duration': 135427,
             },
             'params': {
                 'skip_download': True,
@@ -960,8 +962,9 @@ class GenericIE(InfoExtractor):
             'info_dict': {
                 'id': '50YnY4czr4ms1vJ7yz3xzq0excz_pUMs',
                 'ext': 'mp4',
-                'description': 'VIDEO: Index/Match versus VLOOKUP.',
+                'description': 'VIDEO: INDEX/MATCH versus VLOOKUP.',
                 'title': 'This is what separates the Excel masters from the wannabes',
+                'duration': 191933,
             },
             'params': {
                 # m3u8 downloads
@@ -1501,7 +1504,7 @@ class GenericIE(InfoExtractor):
                 re.search(r'SBN\.VideoLinkset\.ooyala\([\'"](?P<ec>.{32})[\'"]\)', webpage) or
                 re.search(r'data-ooyala-video-id\s*=\s*[\'"](?P<ec>.{32})[\'"]', webpage))
         if mobj is not None:
-            return OoyalaIE._build_url_result(mobj.group('ec'))
+            return OoyalaIE._build_url_result(smuggle_url(mobj.group('ec'), {'domain': url}))
 
         # Look for multiple Ooyala embeds on SBN network websites
         mobj = re.search(r'SBN\.VideoLinkset\.entryGroup\((\[.*?\])', webpage)
@@ -1509,7 +1512,7 @@ class GenericIE(InfoExtractor):
             embeds = self._parse_json(mobj.group(1), video_id, fatal=False)
             if embeds:
                 return _playlist_from_matches(
-                    embeds, getter=lambda v: OoyalaIE._url_for_embed_code(v['provider_video_id']), ie='Ooyala')
+                    embeds, getter=lambda v: OoyalaIE._url_for_embed_code(smuggle_url(v['provider_video_id'], {'domain': url})), ie='Ooyala')
 
         # Look for Aparat videos
         mobj = re.search(r'<iframe .*?src="(http://www\.aparat\.com/video/[^"]+)"', webpage)

youtube_dl/extractor/groupon.py

@@ -18,6 +18,8 @@ class GrouponIE(InfoExtractor):
                 'id': 'tubGNycTo_9Uxg82uESj4i61EYX8nyuf',
                 'ext': 'mp4',
                 'title': 'Bikram Yoga Huntington Beach | Orange County',
+                'description': 'md5:d41d8cd98f00b204e9800998ecf8427e',
+                'duration': 44961,
             },
         }],
         'params': {

youtube_dl/extractor/howcast.py

@@ -16,6 +16,7 @@ class HowcastIE(InfoExtractor):
             'description': 'md5:dbe792e5f6f1489027027bf2eba188a3',
             'timestamp': 1276081287,
             'upload_date': '20100609',
+            'duration': 56823,
         },
         'params': {
             # m3u8 download

youtube_dl/extractor/ooyala.py

@@ -1,108 +1,69 @@
 from __future__ import unicode_literals
 import re
-import json
 import base64
 
 from .common import InfoExtractor
 from ..utils import (
-    unescapeHTML,
-    ExtractorError,
-    determine_ext,
     int_or_none,
+    float_or_none,
+    ExtractorError,
+    unsmuggle_url,
 )
+from ..compat import compat_urllib_parse
 
 
 class OoyalaBaseIE(InfoExtractor):
 
-    def _extract_result(self, info, more_info):
+    def _extract(self, content_tree_url, video_id, domain='example.org'):
-        embedCode = info['embedCode']
+        content_tree = self._download_json(content_tree_url, video_id)['content_tree']
-        video_url = info.get('ipad_url') or info['url']
+        metadata = content_tree[list(content_tree)[0]]
-
+        embed_code = metadata['embed_code']
-        if determine_ext(video_url) == 'm3u8':
+        pcode = metadata.get('asset_pcode') or embed_code
-            formats = self._extract_m3u8_formats(video_url, embedCode, ext='mp4')
+        video_info = {
-        else:
+            'id': embed_code,
-            formats = [{
+            'title': metadata['title'],
-                'url': video_url,
+            'description': metadata.get('description'),
-                'ext': 'mp4',
+            'thumbnail': metadata.get('thumbnail_image') or metadata.get('promo_image'),
-            }]
+            'duration': int_or_none(metadata.get('duration')),
-
-        return {
-            'id': embedCode,
-            'title': unescapeHTML(info['title']),
-            'formats': formats,
-            'description': unescapeHTML(more_info['description']),
-            'thumbnail': more_info['promo'],
         }
 
-    def _extract(self, player_url, video_id):
+        formats = []
-        player = self._download_webpage(player_url, video_id)
+        for supported_format in ('mp4', 'm3u8', 'hds', 'rtmp'):
-        mobile_url = self._search_regex(r'mobile_player_url="(.+?)&device="',
-                                        player, 'mobile player url')
-        # Looks like some videos are only available for particular devices
-        # (e.g. http://player.ooyala.com/player.js?embedCode=x1b3lqZDq9y_7kMyC2Op5qo-p077tXD0
-        # is only available for ipad)
-        # Working around with fetching URLs for all the devices found starting with 'unknown'
-        # until we succeed or eventually fail for each device.
-        devices = re.findall(r'device\s*=\s*"([^"]+)";', player)
-        devices.remove('unknown')
-        devices.insert(0, 'unknown')
-        for device in devices:
-            mobile_player = self._download_webpage(
-                '%s&device=%s' % (mobile_url, device), video_id,
-                'Downloading mobile player JS for %s device' % device)
-            videos_info = self._search_regex(
-                r'var streams=window.oo_testEnv\?\[\]:eval\("\((\[{.*?}\])\)"\);',
-                mobile_player, 'info', fatal=False, default=None)
-            if videos_info:
-                break
-
-        if not videos_info:
-            formats = []
             auth_data = self._download_json(
-                'http://player.ooyala.com/sas/player_api/v1/authorization/embed_code/%s/%s?domain=www.example.org&supportedFormats=mp4,webm' % (video_id, video_id),
+                'http://player.ooyala.com/sas/player_api/v1/authorization/embed_code/%s/%s?' % (pcode, embed_code) + compat_urllib_parse.urlencode({'domain': domain, 'supportedFormats': supported_format}),
-                video_id)
+                video_id, 'Downloading %s JSON' % supported_format)
 
-            cur_auth_data = auth_data['authorization_data'][video_id]
+            cur_auth_data = auth_data['authorization_data'][embed_code]
 
-            for stream in cur_auth_data['streams']:
+            if cur_auth_data['authorized']:
-                formats.append({
+                for stream in cur_auth_data['streams']:
-                    'url': base64.b64decode(stream['url']['data'].encode('ascii')).decode('utf-8'),
+                    url = base64.b64decode(stream['url']['data'].encode('ascii')).decode('utf-8')
-                    'ext': stream.get('delivery_type'),
+                    delivery_type = stream['delivery_type']
-                    'format': stream.get('video_codec'),
+                    if delivery_type == 'remote_asset':
-                    'format_id': stream.get('profile'),
+                        video_info['url'] = url
-                    'width': int_or_none(stream.get('width')),
+                        return video_info
-                    'height': int_or_none(stream.get('height')),
+                    if delivery_type == 'hls':
-                    'abr': int_or_none(stream.get('audio_bitrate')),
+                        formats.extend(self._extract_m3u8_formats(url, embed_code, 'mp4', 'm3u8_native', m3u8_id='hls', fatal=False))
-                    'vbr': int_or_none(stream.get('video_bitrate')),
+                    elif delivery_type == 'hds':
-                })
+                        formats.extend(self._extract_f4m_formats(url, embed_code, -1, 'hds', fatal=False))
-            if formats:
+                    else:
-                return {
+                        formats.append({
-                    'id': video_id,
+                            'url': url,
-                    'formats': formats,
+                            'ext': stream.get('delivery_type'),
-                    'title': 'Ooyala video',
+                            'vcodec': stream.get('video_codec'),
-                }
+                            'format_id': '%s-%s-%sp' % (stream.get('profile'), delivery_type, stream.get('height')),
+                            'width': int_or_none(stream.get('width')),
+                            'height': int_or_none(stream.get('height')),
+                            'abr': int_or_none(stream.get('audio_bitrate')),
+                            'vbr': int_or_none(stream.get('video_bitrate')),
+                            'fps': float_or_none(stream.get('framerate')),
+                        })
+            else:
+                raise ExtractorError('%s said: %s' % (self.IE_NAME, cur_auth_data['message']), expected=True)
+        self._sort_formats(formats)
 
-            if not cur_auth_data['authorized']:
+        video_info['formats'] = formats
-                raise ExtractorError(cur_auth_data['message'], expected=True)
+        return video_info
-
-        if not videos_info:
-            raise ExtractorError('Unable to extract info')
-        videos_info = videos_info.replace('\\"', '"')
-        videos_more_info = self._search_regex(
-            r'eval\("\(({.*?\\"promo\\".*?})\)"', mobile_player, 'more info').replace('\\"', '"')
-        videos_info = json.loads(videos_info)
-        videos_more_info = json.loads(videos_more_info)
-
-        if videos_more_info.get('lineup'):
-            videos = [self._extract_result(info, more_info) for (info, more_info) in zip(videos_info, videos_more_info['lineup'])]
-            return {
-                '_type': 'playlist',
-                'id': video_id,
-                'title': unescapeHTML(videos_more_info['title']),
-                'entries': videos,
-            }
-        else:
-            return self._extract_result(videos_info[0], videos_more_info)
 
 
 class OoyalaIE(OoyalaBaseIE):
@@ -117,6 +78,7 @@ class OoyalaIE(OoyalaBaseIE):
                 'ext': 'mp4',
                 'title': 'Explaining Data Recovery from Hard Drives and SSDs',
                 'description': 'How badly damaged does a drive have to be to defeat Russell and his crew? Apparently, smashed to bits.',
+                'duration': 853386,
             },
         }, {
             # Only available for ipad
@@ -125,7 +87,7 @@ class OoyalaIE(OoyalaBaseIE):
                 'id': 'x1b3lqZDq9y_7kMyC2Op5qo-p077tXD0',
                 'ext': 'mp4',
                 'title': 'Simulation Overview - Levels of Simulation',
-                'description': '',
+                'duration': 194948,
             },
         },
         {
@@ -136,7 +98,8 @@ class OoyalaIE(OoyalaBaseIE):
             'info_dict': {
                 'id': 'FiOG81ZTrvckcchQxmalf4aQj590qTEx',
                 'ext': 'mp4',
-                'title': 'Ooyala video',
+                'title': 'Divide Tool Path.mp4',
+                'duration': 204405,
             }
         }
     ]
@@ -151,9 +114,11 @@ class OoyalaIE(OoyalaBaseIE):
                               ie=cls.ie_key())
 
     def _real_extract(self, url):
+        url, smuggled_data = unsmuggle_url(url, {})
         embed_code = self._match_id(url)
-        player_url = 'http://player.ooyala.com/player.js?embedCode=%s' % embed_code
+        domain = smuggled_data.get('domain')
-        return self._extract(player_url, embed_code)
+        content_tree_url = 'http://player.ooyala.com/player_api/v1/content_tree/embed_code/%s/%s' % (embed_code, embed_code)
+        return self._extract(content_tree_url, embed_code, domain)
 
 
 class OoyalaExternalIE(OoyalaBaseIE):
@@ -170,7 +135,7 @@ class OoyalaExternalIE(OoyalaBaseIE):
                         .*?&pcode=
                     )
                     (?P<pcode>.+?)
-                    (&|$)
+                    (?:&|$)
                     '''
 
     _TEST = {
@@ -179,7 +144,7 @@ class OoyalaExternalIE(OoyalaBaseIE):
             'id': 'FkYWtmazr6Ed8xmvILvKLWjd4QvYZpzG',
             'ext': 'mp4',
             'title': 'dm_140128_30for30Shorts___JudgingJewellv2',
-            'description': '',
+            'duration': 1302000,
         },
         'params': {
             # m3u8 download
@@ -188,9 +153,6 @@ class OoyalaExternalIE(OoyalaBaseIE):
     }
 
     def _real_extract(self, url):
-        mobj = re.match(self._VALID_URL, url)
+        partner_id, video_id, pcode = re.match(self._VALID_URL, url).groups()
-        partner_id = mobj.group('partner_id')
+        content_tree_url = 'http://player.ooyala.com/player_api/v1/content_tree/external_id/%s/%s:%s' % (pcode, partner_id, video_id)
-        video_id = mobj.group('id')
+        return self._extract(content_tree_url, video_id)
-        pcode = mobj.group('pcode')
-        player_url = 'http://player.ooyala.com/player.js?externalId=%s:%s&pcode=%s' % (partner_id, video_id, pcode)
-        return self._extract(player_url, video_id)

youtube_dl/extractor/teachingchannel.py

@@ -16,6 +16,7 @@ class TeachingChannelIE(InfoExtractor):
             'ext': 'mp4',
             'title': 'A History of Teaming',
             'description': 'md5:2a9033db8da81f2edffa4c99888140b3',
+            'duration': 422255,
         },
         'params': {
             # m3u8 download

youtube_dl/extractor/vice.py

@@ -15,6 +15,7 @@ class ViceIE(InfoExtractor):
                 'id': '43cW1mYzpia9IlestBjVpd23Yu3afAfp',
                 'ext': 'mp4',
                 'title': 'VICE_COWBOYCAPITALISTS_PART01_v1_VICE_WM_1080p.mov',
+                'duration': 725983,
             },
             'params': {
                 # Requires ffmpeg (m3u8 manifest)