release 2017.04.17

* Move extraction code to extractor
* Add extraction for LimelightEmbeddedPlayerFlash embeds
* Extract multiple video

.github/ISSUE_TEMPLATE.md

@@ -6,8 +6,8 @@
 
 ---
 
-### Make sure you are using the *latest* version: run `youtube-dl --version` and ensure your version is *2017.04.14*. If it's not read [this FAQ entry](https://github.com/rg3/youtube-dl/blob/master/README.md#how-do-i-update-youtube-dl) and update. Issues with outdated version will be rejected.
-- [ ] I've **verified** and **I assure** that I'm running youtube-dl **2017.04.14**
+### Make sure you are using the *latest* version: run `youtube-dl --version` and ensure your version is *2017.04.17*. If it's not read [this FAQ entry](https://github.com/rg3/youtube-dl/blob/master/README.md#how-do-i-update-youtube-dl) and update. Issues with outdated version will be rejected.
+- [ ] I've **verified** and **I assure** that I'm running youtube-dl **2017.04.17**
 
 ### Before submitting an *issue* make sure you have:
 - [ ] At least skimmed through [README](https://github.com/rg3/youtube-dl/blob/master/README.md) and **most notably** [FAQ](https://github.com/rg3/youtube-dl#faq) and [BUGS](https://github.com/rg3/youtube-dl#bugs) sections
@@ -35,7 +35,7 @@ $ youtube-dl -v <your command line>
 [debug] User config: []
 [debug] Command-line args: [u'-v', u'http://www.youtube.com/watch?v=BaW_jenozKcj']
 [debug] Encodings: locale cp1251, fs mbcs, out cp866, pref cp1251
-[debug] youtube-dl version 2017.04.14
+[debug] youtube-dl version 2017.04.17
 [debug] Python version 2.7.11 - Windows-2003Server-5.2.3790-SP2
 [debug] exe versions: ffmpeg N-75573-g1d0487f, ffprobe N-75573-g1d0487f, rtmpdump 2.4
 [debug] Proxy map: {}

AUTHORS

@@ -211,3 +211,4 @@ Juanjo Benages
 Xiao Di Guan
 Thomas Winant
 Daniel Twardowski
+Jeremie Jarosh

ChangeLog

@@ -1,3 +1,38 @@
+version 2017.04.17
+
+Extractors
+* [limelight] Improve extraction LimelightEmbeddedPlayerFlash media embeds and
+  add support for channel and channelList embeds
+* [generic] Extract multiple Limelight embeds (#12761)
++ [itv] Extract series metadata
+* [itv] Fix RTMP formats downloading (#12759)
+* [itv] Use native HLS downloader by default
++ [go90] Extract subtitles (#12752)
++ [go90] Extract series metadata (#12752)
+
+
+version 2017.04.16
+
+Core
+* [YoutubeDL] Apply expand_path after output template substitution
++ [YoutubeDL] Propagate overridden meta fields to extraction results of type
+  url (#11163)
+
+Extractors
++ [generic] Extract RSS entries as url_transparent (#11163)
++ [streamango] Add support for streamango.com (#12643)
++ [wsj:article] Add support for articles (#12558)
+* [brightcove] Relax video tag embeds extraction and validate ambiguous embeds'
+  URLs (#9163, #12005, #12178, #12480)
++ [udemy] Add support for react rendition (#12744)
+
+
+version 2017.04.15
+
+Extractors
+* [youku] Fix fileid extraction (#12741, #12743)
+
+
 version 2017.04.14
 
 Core

docs/supportedsites.md

@@ -745,6 +745,7 @@
  - **Steam**
  - **Stitcher**
  - **Streamable**
+ - **Streamango**
  - **streamcloud.eu**
  - **StreamCZ**
  - **StreetVoice**
@@ -966,6 +967,7 @@
  - **wrzuta.pl**
  - **wrzuta.pl:playlist**
  - **WSJ**: Wall Street Journal
+ - **WSJArticle**
  - **XBef**
  - **XboxClips**
  - **XFileShare**: XFileShare based sites: DaClips, FileHoot, GorillaVid, MovPod, PowerWatch, Rapidvideo.ws, TheVideoBee, Vidto, Streamin.To, XVIDSTAGE, Vid ABC, VidBom, vidlo

test/test_YoutubeDL.py

@@ -755,6 +755,7 @@ class TestYoutubeDL(unittest.TestCase):
                     '_type': 'url_transparent',
                     'url': 'foo2:',
                     'ie_key': 'Foo2',
+                    'title': 'foo1 title'
                 }
 
         class Foo2IE(InfoExtractor):
@@ -771,7 +772,7 @@ class TestYoutubeDL(unittest.TestCase):
             _VALID_URL = r'foo3:'
 
             def _real_extract(self, url):
-                return _make_result([{'url': TEST_URL}])
+                return _make_result([{'url': TEST_URL}], title='foo3 title')
 
         ydl.add_info_extractor(Foo1IE(ydl))
         ydl.add_info_extractor(Foo2IE(ydl))
@@ -779,6 +780,7 @@ class TestYoutubeDL(unittest.TestCase):
         ydl.extract_info('foo1:')
         downloaded = ydl.downloaded_info_dicts[0]
         self.assertEqual(downloaded['url'], TEST_URL)
+        self.assertEqual(downloaded['title'], 'foo1 title')
 
 
 if __name__ == '__main__':

youtube_dl/YoutubeDL.py

@@ -672,8 +672,7 @@ class YoutubeDL(object):
                         FORMAT_RE.format(numeric_field),
                         r'%({0})s'.format(numeric_field), outtmpl)
 
-            tmpl = expand_path(outtmpl)
-            filename = tmpl % template_dict
+            filename = expand_path(outtmpl % template_dict)
             # Temporary fix for #4787
             # 'Treat' all problem characters by passing filename through preferredencoding
             # to workaround encoding issues with subprocess on python2 @ Windows
@@ -851,7 +850,14 @@ class YoutubeDL(object):
             new_result = info.copy()
             new_result.update(force_properties)
 
-            assert new_result.get('_type') != 'url_transparent'
+            # Extracted info may not be a video result (i.e.
+            # info.get('_type', 'video') != video) but rather an url or
+            # url_transparent. In such cases outer metadata (from ie_result)
+            # should be propagated to inner one (info). For this to happen
+            # _type of info should be overridden with url_transparent. This
+            # fixes issue from https://github.com/rg3/youtube-dl/pull/11163.
+            if new_result.get('_type') == 'url':
+                new_result['_type'] = 'url_transparent'
 
             return self.process_ie_result(
                 new_result, download=download, extra_info=extra_info)

youtube_dl/extractor/brightcove.py

@@ -17,6 +17,7 @@ from ..compat import (
 from ..utils import (
     determine_ext,
     ExtractorError,
+    extract_attributes,
     find_xpath_attr,
     fix_xml_ampersands,
     float_or_none,
@@ -109,6 +110,7 @@ class BrightcoveLegacyIE(InfoExtractor):
                 'upload_date': '20140827',
                 'uploader_id': '710858724001',
             },
+            'skip': 'Video gone',
         },
         {
             # playlist with 'videoList'
@@ -487,12 +489,13 @@ class BrightcoveNewIE(InfoExtractor):
         return urls[0] if urls else None
 
     @staticmethod
-    def _extract_urls(webpage):
+    def _extract_urls(ie, webpage):
         # Reference:
         # 1. http://docs.brightcove.com/en/video-cloud/brightcove-player/guides/publish-video.html#setvideoiniframe
-        # 2. http://docs.brightcove.com/en/video-cloud/brightcove-player/guides/publish-video.html#setvideousingjavascript
-        # 3. http://docs.brightcove.com/en/video-cloud/brightcove-player/guides/embed-in-page.html
-        # 4. https://support.brightcove.com/en/video-cloud/docs/dynamically-assigning-videos-player
+        # 2. http://docs.brightcove.com/en/video-cloud/brightcove-player/guides/publish-video.html#tag
+        # 3. http://docs.brightcove.com/en/video-cloud/brightcove-player/guides/publish-video.html#setvideousingjavascript
+        # 4. http://docs.brightcove.com/en/video-cloud/brightcove-player/guides/in-page-embed-player-implementation.html
+        # 5. https://support.brightcove.com/en/video-cloud/docs/dynamically-assigning-videos-player
 
         entries = []
 
@@ -501,22 +504,48 @@ class BrightcoveNewIE(InfoExtractor):
                 r'<iframe[^>]+src=(["\'])((?:https?:)?//players\.brightcove\.net/\d+/[^/]+/index\.html.+?)\1', webpage):
             entries.append(url if url.startswith('http') else 'http:' + url)
 
-        # Look for embed_in_page embeds [2]
-        for video_id, account_id, player_id, embed in re.findall(
-                # According to examples from [3] it's unclear whether video id
-                # may be optional and what to do when it is
-                # According to [4] data-video-id may be prefixed with ref:
-                r'''(?sx)
-                    <video[^>]+
-                        data-video-id=["\'](\d+|ref:[^"\']+)["\'][^>]*>.*?
-                    </video>.*?
-                    <script[^>]+
-                        src=["\'](?:https?:)?//players\.brightcove\.net/
-                        (\d+)/([^/]+)_([^/]+)/index(?:\.min)?\.js
+        # Look for <video> tags [2] and embed_in_page embeds [3]
+        # [2] looks like:
+        for video, script_tag, account_id, player_id, embed in re.findall(
+                r'''(?isx)
+                    (<video\s+[^>]+>)
+                    (?:.*?
+                        (<script[^>]+
+                            src=["\'](?:https?:)?//players\.brightcove\.net/
+                            (\d+)/([^/]+)_([^/]+)/index(?:\.min)?\.js
+                        )
+                    )?
                 ''', webpage):
-            entries.append(
-                'http://players.brightcove.net/%s/%s_%s/index.html?videoId=%s'
-                % (account_id, player_id, embed, video_id))
+            attrs = extract_attributes(video)
+
+            # According to examples from [4] it's unclear whether video id
+            # may be optional and what to do when it is
+            video_id = attrs.get('data-video-id')
+            if not video_id:
+                continue
+
+            account_id = account_id or attrs.get('data-account')
+            if not account_id:
+                continue
+
+            player_id = player_id or attrs.get('data-player') or 'default'
+            embed = embed or attrs.get('data-embed') or 'default'
+
+            bc_url = 'http://players.brightcove.net/%s/%s_%s/index.html?videoId=%s' % (
+                account_id, player_id, embed, video_id)
+
+            # Some brightcove videos may be embedded with video tag only and
+            # without script tag or any mentioning of brightcove at all. Such
+            # embeds are considered ambiguous since they are matched based only
+            # on data-video-id and data-account attributes and in the wild may
+            # not be brightcove embeds at all. Let's check reconstructed
+            # brightcove URLs in case of such embeds and only process valid
+            # ones. By this we ensure there is indeed a brightcove embed.
+            if not script_tag and not ie._is_valid_url(
+                    bc_url, video_id, 'possible brightcove video'):
+                continue
+
+            entries.append(bc_url)
 
         return entries
 

youtube_dl/extractor/canalplus.py

@@ -7,8 +7,8 @@ from .common import InfoExtractor
 from ..compat import compat_urllib_parse_urlparse
 from ..utils import (
     dict_get,
-    ExtractorError,
-    HEADRequest,
+    # ExtractorError,
+    # HEADRequest,
     int_or_none,
     qualities,
     remove_end,

youtube_dl/extractor/curiositystream.py

@@ -48,7 +48,6 @@ class CuriosityStreamBaseIE(InfoExtractor):
 
     def _extract_media_info(self, media):
         video_id = compat_str(media['id'])
-        limelight_media_id = media['limelight_media_id']
         title = media['title']
 
         formats = []

youtube_dl/extractor/extractors.py

@@ -939,6 +939,7 @@ from .srmediathek import SRMediathekIE
 from .stanfordoc import StanfordOpenClassroomIE
 from .steam import SteamIE
 from .streamable import StreamableIE
+from .streamango import StreamangoIE
 from .streamcloud import StreamcloudIE
 from .streamcz import StreamCZIE
 from .streetvoice import StreetVoiceIE
@@ -1233,7 +1234,10 @@ from .wrzuta import (
     WrzutaIE,
     WrzutaPlaylistIE,
 )
-from .wsj import WSJIE
+from .wsj import (
+    WSJIE,
+    WSJArticleIE,
+)
 from .xbef import XBefIE
 from .xboxclips import XboxClipsIE
 from .xfileshare import XFileShareIE

youtube_dl/extractor/generic.py

@@ -85,6 +85,7 @@ from .ustream import UstreamIE
 from .openload import OpenloadIE
 from .videopress import VideoPressIE
 from .rutube import RutubeIE
+from .limelight import LimelightBaseIE
 
 
 class GenericIE(InfoExtractor):
@@ -449,6 +450,59 @@ class GenericIE(InfoExtractor):
                 },
             }],
         },
+        {
+            # Brightcove with UUID in videoPlayer
+            'url': 'http://www8.hp.com/cn/zh/home.html',
+            'info_dict': {
+                'id': '5255815316001',
+                'ext': 'mp4',
+                'title': 'Sprocket Video - China',
+                'description': 'Sprocket Video - China',
+                'uploader': 'HP-Video Gallery',
+                'timestamp': 1482263210,
+                'upload_date': '20161220',
+                'uploader_id': '1107601872001',
+            },
+            'params': {
+                'skip_download': True,  # m3u8 download
+            },
+            'skip': 'video rotates...weekly?',
+        },
+        {
+            # Brightcove:new type [2].
+            'url': 'http://www.delawaresportszone.com/video-st-thomas-more-earns-first-trip-to-basketball-semis',
+            'md5': '2b35148fcf48da41c9fb4591650784f3',
+            'info_dict': {
+                'id': '5348741021001',
+                'ext': 'mp4',
+                'upload_date': '20170306',
+                'uploader_id': '4191638492001',
+                'timestamp': 1488769918,
+                'title': 'VIDEO:  St. Thomas More earns first trip to basketball semis',
+
+            },
+        },
+        {
+            # Alternative brightcove <video> attributes
+            'url': 'http://www.programme-tv.net/videos/extraits/81095-guillaume-canet-evoque-les-rumeurs-d-infidelite-de-marion-cotillard-avec-brad-pitt-dans-vivement-dimanche/',
+            'info_dict': {
+                'id': '81095-guillaume-canet-evoque-les-rumeurs-d-infidelite-de-marion-cotillard-avec-brad-pitt-dans-vivement-dimanche',
+                'title': "Guillaume Canet évoque les rumeurs d'infidélité de Marion Cotillard avec Brad Pitt dans Vivement Dimanche, Extraits : toutes les vidéos avec Télé-Loisirs",
+            },
+            'playlist': [{
+                'md5': '732d22ba3d33f2f3fc253c39f8f36523',
+                'info_dict': {
+                    'id': '5311302538001',
+                    'ext': 'mp4',
+                    'title': "Guillaume Canet évoque les rumeurs d'infidélité de Marion Cotillard avec Brad Pitt dans Vivement Dimanche",
+                    'description': "Guillaume Canet évoque les rumeurs d'infidélité de Marion Cotillard avec Brad Pitt dans Vivement Dimanche (France 2, 5 février 2017)",
+                    'timestamp': 1486321708,
+                    'upload_date': '20170205',
+                    'uploader_id': '800000640001',
+                },
+                'only_matching': True,
+            }],
+        },
         {
             # Brightcove with UUID in videoPlayer
             'url': 'http://www8.hp.com/cn/zh/home.html',
@@ -1598,6 +1652,15 @@ class GenericIE(InfoExtractor):
             },
             'add_ie': [SenateISVPIE.ie_key()],
         },
+        {
+            # Limelight embeds (1 channel embed + 4 media embeds)
+            'url': 'http://www.sedona.com/FacilitatorTraining2017',
+            'info_dict': {
+                'id': 'FacilitatorTraining2017',
+                'title': 'Facilitator Training 2017',
+            },
+            'playlist_mincount': 5,
+        },
         # {
         #     # TODO: find another test
         #     # http://schema.org/VideoObject
@@ -1640,7 +1703,7 @@ class GenericIE(InfoExtractor):
                 continue
 
             entries.append({
-                '_type': 'url',
+                '_type': 'url_transparent',
                 'url': next_url,
                 'title': it.find('title').text,
             })
@@ -1900,7 +1963,6 @@ class GenericIE(InfoExtractor):
         # Look for Brightcove Legacy Studio embeds
         bc_urls = BrightcoveLegacyIE._extract_brightcove_urls(webpage)
         if bc_urls:
-            self.to_screen('Brightcove video detected.')
             entries = [{
                 '_type': 'url',
                 'url': smuggle_url(bc_url, {'Referer': url}),
@@ -1915,7 +1977,7 @@ class GenericIE(InfoExtractor):
             }
 
         # Look for Brightcove New Studio embeds
-        bc_urls = BrightcoveNewIE._extract_urls(webpage)
+        bc_urls = BrightcoveNewIE._extract_urls(self, webpage)
         if bc_urls:
             return self.playlist_from_matches(bc_urls, video_id, video_title, ie='BrightcoveNew')
 
@@ -2431,6 +2493,11 @@ class GenericIE(InfoExtractor):
             return self.url_result(piksel_url, PikselIE.ie_key())
 
         # Look for Limelight embeds
+        limelight_urls = LimelightBaseIE._extract_urls(webpage, url)
+        if limelight_urls:
+            return self.playlist_result(
+                limelight_urls, video_id, video_title, video_description)
+
         mobj = re.search(r'LimelightPlayer\.doLoad(Media|Channel|ChannelList)\(["\'](?P<id>[a-z0-9]{32})', webpage)
         if mobj:
             lm = {

youtube_dl/extractor/go90.py

@@ -5,6 +5,7 @@ import re
 
 from .common import InfoExtractor
 from ..utils import (
+    determine_ext,
     int_or_none,
     parse_iso8601,
 )
@@ -18,7 +19,7 @@ class Go90IE(InfoExtractor):
         'info_dict': {
             'id': '84BUqjLpf9D',
             'ext': 'mp4',
-            'title': 'Inside The Utah Coalition Against Pornography Convention',
+            'title': 'Daily VICE - Inside The Utah Coalition Against Pornography Convention',
             'description': 'VICE\'s Karley Sciortino meets with activists who discuss the state\'s strong anti-porn stance. Then, VICE Sports explains NFL contracts.',
             'timestamp': 1491868800,
             'upload_date': '20170411',
@@ -32,11 +33,28 @@ class Go90IE(InfoExtractor):
             video_id, headers={
                 'Content-Type': 'application/json; charset=utf-8',
             }, data=b'{"client":"web","device_type":"pc"}')
-        title = video_data['title']
         main_video_asset = video_data['main_video_asset']
 
+        episode_number = int_or_none(video_data.get('episode_number'))
+        series = None
+        season = None
+        season_id = None
+        season_number = None
+        for metadata in video_data.get('__children', {}).get('Item', {}).values():
+            if metadata.get('type') == 'show':
+                series = metadata.get('title')
+            elif metadata.get('type') == 'season':
+                season = metadata.get('title')
+                season_id = metadata.get('id')
+                season_number = int_or_none(metadata.get('season_number'))
+
+        title = episode = video_data.get('title') or series
+        if series and series != title:
+            title = '%s - %s' % (series, title)
+
         thumbnails = []
         formats = []
+        subtitles = {}
         for asset in video_data.get('assets'):
             if asset.get('id') == main_video_asset:
                 for source in asset.get('sources', []):
@@ -70,6 +88,15 @@ class Go90IE(InfoExtractor):
                             'height': int_or_none(source.get('height')),
                             'tbr': int_or_none(source.get('bitrate')),
                         })
+
+                for caption in asset.get('caption_metadata', []):
+                    caption_url = caption.get('source_url')
+                    if not caption_url:
+                        continue
+                    subtitles.setdefault(caption.get('language', 'en'), []).append({
+                        'url': caption_url,
+                        'ext': determine_ext(caption_url, 'vtt'),
+                    })
             elif asset.get('type') == 'image':
                 asset_location = asset.get('location')
                 if not asset_location:
@@ -89,4 +116,11 @@ class Go90IE(InfoExtractor):
             'description': video_data.get('short_description'),
             'like_count': int_or_none(video_data.get('like_count')),
             'timestamp': parse_iso8601(video_data.get('released_at')),
+            'series': series,
+            'episode': episode,
+            'season': season,
+            'season_id': season_id,
+            'season_number': season_number,
+            'episode_number': episode_number,
+            'subtitles': subtitles,
         }

youtube_dl/extractor/itv.py

@@ -116,13 +116,25 @@ class ITVIE(InfoExtractor):
             if not play_path:
                 continue
             tbr = int_or_none(media_file.get('bitrate'), 1000)
-            formats.append({
+            f = {
                 'format_id': 'rtmp' + ('-%d' % tbr if tbr else ''),
-                'url': rtmp_url,
                 'play_path': play_path,
+                # Providing this swfVfy allows to avoid truncated downloads
+                'player_url': 'http://www.itv.com/mercury/Mercury_VideoPlayer.swf',
+                'page_url': url,
                 'tbr': tbr,
                 'ext': 'flv',
-            })
+            }
+            app = self._search_regex(
+                'rtmpe?://[^/]+/(.+)$', rtmp_url, 'app', default=None)
+            if app:
+                f.update({
+                    'url': rtmp_url.split('?', 1)[0],
+                    'app': app,
+                })
+            else:
+                f['url'] = rtmp_url
+            formats.append(f)
 
         ios_playlist_url = params.get('data-video-playlist')
         hmac = params.get('data-video-hmac')
@@ -172,7 +184,9 @@ class ITVIE(InfoExtractor):
                         href = ios_base_url + href
                     ext = determine_ext(href)
                     if ext == 'm3u8':
-                        formats.extend(self._extract_m3u8_formats(href, video_id, 'mp4', m3u8_id='hls', fatal=False))
+                        formats.extend(self._extract_m3u8_formats(
+                            href, video_id, 'mp4', entry_protocol='m3u8_native',
+                            m3u8_id='hls', fatal=False))
                     else:
                         formats.append({
                             'url': href,
@@ -189,7 +203,8 @@ class ITVIE(InfoExtractor):
                 'ext': 'ttml' if ext == 'xml' else ext,
             })
 
-        return {
+        info = self._search_json_ld(webpage, video_id, default={})
+        info.update({
             'id': video_id,
             'title': title,
             'formats': formats,
@@ -198,4 +213,5 @@ class ITVIE(InfoExtractor):
             'episode_number': int_or_none(xpath_text(playlist, 'EpisodeNumber')),
             'series': xpath_text(playlist, 'ProgrammeTitle'),
             'duartion': parse_duration(xpath_text(playlist, 'Duration')),
-        }
+        })
+        return info

youtube_dl/extractor/limelight.py

@@ -9,6 +9,7 @@ from ..utils import (
     determine_ext,
     float_or_none,
     int_or_none,
+    smuggle_url,
     unsmuggle_url,
     ExtractorError,
 )
@@ -18,6 +19,42 @@ class LimelightBaseIE(InfoExtractor):
     _PLAYLIST_SERVICE_URL = 'http://production-ps.lvp.llnw.net/r/PlaylistService/%s/%s/%s'
     _API_URL = 'http://api.video.limelight.com/rest/organizations/%s/%s/%s/%s.json'
 
+    @classmethod
+    def _extract_urls(cls, webpage, source_url):
+        lm = {
+            'Media': 'media',
+            'Channel': 'channel',
+            'ChannelList': 'channel_list',
+        }
+        entries = []
+        for kind, video_id in re.findall(
+                r'LimelightPlayer\.doLoad(Media|Channel|ChannelList)\(["\'](?P<id>[a-z0-9]{32})',
+                webpage):
+            entries.append(cls.url_result(
+                smuggle_url(
+                    'limelight:%s:%s' % (lm[kind], video_id),
+                    {'source_url': source_url}),
+                'Limelight%s' % kind, video_id))
+        for mobj in re.finditer(
+                # As per [1] class attribute should be exactly equal to
+                # LimelightEmbeddedPlayerFlash but numerous examples seen
+                # that don't exactly match it (e.g. [2]).
+                # 1. http://support.3playmedia.com/hc/en-us/articles/227732408-Limelight-Embedding-the-Captions-Plugin-with-the-Limelight-Player-on-Your-Webpage
+                # 2. http://www.sedona.com/FacilitatorTraining2017
+                r'''(?sx)
+                    <object[^>]+class=(["\'])(?:(?!\1).)*\bLimelightEmbeddedPlayerFlash\b(?:(?!\1).)*\1[^>]*>.*?
+                        <param[^>]+
+                            name=(["\'])flashVars\2[^>]+
+                            value=(["\'])(?:(?!\3).)*(?P<kind>media|channel(?:List)?)Id=(?P<id>[a-z0-9]{32})
+                ''', webpage):
+            kind, video_id = mobj.group('kind'), mobj.group('id')
+            entries.append(cls.url_result(
+                smuggle_url(
+                    'limelight:%s:%s' % (kind, video_id),
+                    {'source_url': source_url}),
+                'Limelight%s' % kind.capitalize(), video_id))
+        return entries
+
     def _call_playlist_service(self, item_id, method, fatal=True, referer=None):
         headers = {}
         if referer:

youtube_dl/extractor/streamango.py

@@ -0,0 +1,64 @@
+# coding: utf-8
+from __future__ import unicode_literals
+
+import re
+
+from .common import InfoExtractor
+from ..utils import (
+    determine_ext,
+    int_or_none,
+    js_to_json,
+)
+
+
+class StreamangoIE(InfoExtractor):
+    _VALID_URL = r'https?://(?:www\.)?streamango\.com/(?:f|embed)/(?P<id>[^/?#&]+)'
+    _TESTS = [{
+        'url': 'https://streamango.com/f/clapasobsptpkdfe/20170315_150006_mp4',
+        'md5': 'e992787515a182f55e38fc97588d802a',
+        'info_dict': {
+            'id': 'clapasobsptpkdfe',
+            'ext': 'mp4',
+            'title': '20170315_150006.mp4',
+        }
+    }, {
+        'url': 'https://streamango.com/embed/clapasobsptpkdfe/20170315_150006_mp4',
+        'only_matching': True,
+    }]
+
+    def _real_extract(self, url):
+        video_id = self._match_id(url)
+
+        webpage = self._download_webpage(url, video_id)
+
+        title = self._og_search_title(webpage)
+
+        formats = []
+        for format_ in re.findall(r'({[^}]*\bsrc\s*:\s*[^}]*})', webpage):
+            video = self._parse_json(
+                format_, video_id, transform_source=js_to_json, fatal=False)
+            if not video:
+                continue
+            src = video.get('src')
+            if not src:
+                continue
+            ext = determine_ext(src, default_ext=None)
+            if video.get('type') == 'application/dash+xml' or ext == 'mpd':
+                formats.extend(self._extract_mpd_formats(
+                    src, video_id, mpd_id='dash', fatal=False))
+            else:
+                formats.append({
+                    'url': src,
+                    'ext': ext or 'mp4',
+                    'width': int_or_none(video.get('width')),
+                    'height': int_or_none(video.get('height')),
+                    'tbr': int_or_none(video.get('bitrate')),
+                })
+        self._sort_formats(formats)
+
+        return {
+            'id': video_id,
+            'url': url,
+            'title': title,
+            'formats': formats,
+        }

youtube_dl/extractor/udemy.py

@@ -212,12 +212,15 @@ class UdemyIE(InfoExtractor):
         thumbnail = asset.get('thumbnail_url') or asset.get('thumbnailUrl')
         duration = float_or_none(asset.get('data', {}).get('duration'))
 
+        subtitles = {}
+        automatic_captions = {}
+
         formats = []
 
-        def extract_output_format(src):
+        def extract_output_format(src, f_id):
             return {
                 'url': src['url'],
-                'format_id': '%sp' % (src.get('height') or format_id),
+                'format_id': '%sp' % (src.get('height') or f_id),
                 'width': int_or_none(src.get('width')),
                 'height': int_or_none(src.get('height')),
                 'vbr': int_or_none(src.get('video_bitrate_in_kbps')),
@@ -237,30 +240,33 @@ class UdemyIE(InfoExtractor):
         def add_output_format_meta(f, key):
             output = outputs.get(key)
             if isinstance(output, dict):
-                output_format = extract_output_format(output)
+                output_format = extract_output_format(output, key)
                 output_format.update(f)
                 return output_format
             return f
 
+        def extract_formats(source_list):
+            if not isinstance(source_list, list):
+                return
+            for source in source_list:
+                video_url = source.get('file') or source.get('src')
+                if not video_url or not isinstance(video_url, compat_str):
+                    continue
+                format_id = source.get('label')
+                f = {
+                    'url': video_url,
+                    'format_id': '%sp' % format_id,
+                    'height': int_or_none(format_id),
+                }
+                if format_id:
+                    # Some videos contain additional metadata (e.g.
+                    # https://www.udemy.com/ios9-swift/learn/#/lecture/3383208)
+                    f = add_output_format_meta(f, format_id)
+                formats.append(f)
+
         download_urls = asset.get('download_urls')
         if isinstance(download_urls, dict):
-            video = download_urls.get('Video')
-            if isinstance(video, list):
-                for format_ in video:
-                    video_url = format_.get('file')
-                    if not video_url:
-                        continue
-                    format_id = format_.get('label')
-                    f = {
-                        'url': format_['file'],
-                        'format_id': '%sp' % format_id,
-                        'height': int_or_none(format_id),
-                    }
-                    if format_id:
-                        # Some videos contain additional metadata (e.g.
-                        # https://www.udemy.com/ios9-swift/learn/#/lecture/3383208)
-                        f = add_output_format_meta(f, format_id)
-                    formats.append(f)
+            extract_formats(download_urls.get('Video'))
 
         view_html = lecture.get('view_html')
         if view_html:
@@ -294,6 +300,35 @@ class UdemyIE(InfoExtractor):
                         'height': height,
                     }, res))
 
+            # react rendition since 2017.04.15 (see
+            # https://github.com/rg3/youtube-dl/issues/12744)
+            data = self._parse_json(
+                self._search_regex(
+                    r'videojs-setup-data=(["\'])(?P<data>{.+?})\1', view_html,
+                    'setup data', default='{}', group='data'), video_id,
+                transform_source=unescapeHTML, fatal=False)
+            if data and isinstance(data, dict):
+                extract_formats(data.get('sources'))
+                if not duration:
+                    duration = int_or_none(data.get('duration'))
+                tracks = data.get('tracks')
+                if isinstance(tracks, list):
+                    for track in tracks:
+                        if not isinstance(track, dict):
+                            continue
+                        if track.get('kind') != 'captions':
+                            continue
+                        src = track.get('src')
+                        if not src or not isinstance(src, compat_str):
+                            continue
+                        lang = track.get('language') or track.get(
+                            'srclang') or track.get('label')
+                        sub_dict = automatic_captions if track.get(
+                            'autogenerated') is True else subtitles
+                        sub_dict.setdefault(lang, []).append({
+                            'url': src,
+                        })
+
         self._sort_formats(formats, field_preference=('height', 'width', 'tbr', 'format_id'))
 
         return {
@@ -302,7 +337,9 @@ class UdemyIE(InfoExtractor):
             'description': description,
             'thumbnail': thumbnail,
             'duration': duration,
-            'formats': formats
+            'formats': formats,
+            'subtitles': subtitles,
+            'automatic_captions': automatic_captions,
         }
 
 

youtube_dl/extractor/wsj.py

@@ -10,12 +10,14 @@ from ..utils import (
 
 
 class WSJIE(InfoExtractor):
-    _VALID_URL = r'''(?x)https?://
-        (?:
-            video-api\.wsj\.com/api-video/player/iframe\.html\?guid=|
-            (?:www\.)?wsj\.com/video/[^/]+/
-        )
-        (?P<id>[a-zA-Z0-9-]+)'''
+    _VALID_URL = r'''(?x)
+                        (?:
+                            https?://video-api\.wsj\.com/api-video/player/iframe\.html\?.*?\bguid=|
+                            https?://(?:www\.)?wsj\.com/video/[^/]+/|
+                            wsj:
+                        )
+                        (?P<id>[a-fA-F0-9-]{36})
+                    '''
     IE_DESC = 'Wall Street Journal'
     _TESTS = [{
         'url': 'http://video-api.wsj.com/api-video/player/iframe.html?guid=1BD01A4C-BFE8-40A5-A42F-8A8AF9898B1A',
@@ -38,12 +40,17 @@ class WSJIE(InfoExtractor):
     def _real_extract(self, url):
         video_id = self._match_id(url)
 
-        api_url = (
-            'http://video-api.wsj.com/api-video/find_all_videos.asp?'
-            'type=guid&count=1&query=%s&fields=type,hls,videoMP4List,'
-            'thumbnailList,author,description,name,duration,videoURL,'
-            'titletag,formattedCreationDate,keywords,editor' % video_id)
-        info = self._download_json(api_url, video_id)['items'][0]
+        info = self._download_json(
+            'http://video-api.wsj.com/api-video/find_all_videos.asp', video_id,
+            query={
+                'type': 'guid',
+                'count': 1,
+                'query': video_id,
+                'fields': ','.join((
+                    'type', 'hls', 'videoMP4List', 'thumbnailList', 'author',
+                    'description', 'name', 'duration', 'videoURL', 'titletag',
+                    'formattedCreationDate', 'keywords', 'editor')),
+            })['items'][0]
         title = info.get('name', info.get('titletag'))
 
         formats = []
@@ -87,3 +94,24 @@ class WSJIE(InfoExtractor):
             'title': title,
             'categories': info.get('keywords'),
         }
+
+
+class WSJArticleIE(InfoExtractor):
+    _VALID_URL = r'(?i)https?://(?:www\.)?wsj\.com/articles/(?P<id>[^/?#&]+)'
+    _TEST = {
+        'url': 'https://www.wsj.com/articles/dont-like-china-no-pandas-for-you-1490366939?',
+        'info_dict': {
+            'id': '4B13FA62-1D8C-45DB-8EA1-4105CB20B362',
+            'ext': 'mp4',
+            'upload_date': '20170221',
+            'uploader_id': 'ralcaraz',
+            'title': 'Bao Bao the Panda Leaves for China',
+        }
+    }
+
+    def _real_extract(self, url):
+        article_id = self._match_id(url)
+        webpage = self._download_webpage(url, article_id)
+        video_id = self._search_regex(
+            r'data-src=["\']([a-fA-F0-9-]{36})', webpage, 'video id')
+        return self.url_result('wsj:%s' % video_id, WSJIE.ie_key(), video_id)

youtube_dl/extractor/youku.py

@@ -10,12 +10,14 @@ import time
 
 from .common import InfoExtractor
 from ..compat import (
-    compat_urllib_parse_urlencode,
     compat_ord,
+    compat_str,
+    compat_urllib_parse_urlencode,
 )
 from ..utils import (
     ExtractorError,
     get_element_by_attribute,
+    try_get,
 )
 
 
@@ -105,7 +107,9 @@ class YoukuIE(InfoExtractor):
             if stream.get('channel_type') == 'tail':
                 continue
             format = stream.get('stream_type')
-            fileid = stream['stream_fileid']
+            fileid = try_get(
+                stream, lambda x: x['segs'][0]['fileid'],
+                compat_str) or stream['stream_fileid']
             fileid_dict[format] = fileid
 
         def get_fileid(format, n):

youtube_dl/version.py

@@ -1,3 +1,3 @@
 from __future__ import unicode_literals
 
-__version__ = '2017.04.14'
+__version__ = '2017.04.17'