Update upstream source from tag 'upstream/2019.09.28'

[youtubedl] / youtube_dl / extractor / vevo.py

diff --git a/youtube_dl/extractor/vevo.py b/youtube_dl/extractor/vevo.py
index d82261e5eec5f3c575bc48f23b23c64aa0355f83..4ea9f1b4bc02d5fc19777124b1f7d5b6cc44abba 100644 (file)
--- a/youtube_dl/extractor/vevo.py
+++ b/youtube_dl/extractor/vevo.py
@@ -1,28 +1,28 @@
 from __future__ import unicode_literals
 
 import re
 from __future__ import unicode_literals
 
 import re

+import json

 
 from .common import InfoExtractor
 from ..compat import (
 
 from .common import InfoExtractor
 from ..compat import (

-    compat_etree_fromstring,

     compat_str,
     compat_urlparse,
     compat_str,
     compat_urlparse,

+    compat_HTTPError,

 )
 from ..utils import (
     ExtractorError,
     int_or_none,
 )
 from ..utils import (
     ExtractorError,
     int_or_none,

-    sanitized_Request,

     parse_iso8601,
 )
 
 
 class VevoBaseIE(InfoExtractor):
     parse_iso8601,
 )
 
 
 class VevoBaseIE(InfoExtractor):

-    def _extract_json(self, webpage, video_id, item):
+    def _extract_json(self, webpage, video_id):

         return self._parse_json(
             self._search_regex(
                 r'window\.__INITIAL_STORE__\s*=\s*({.+?});\s*</script>',
                 webpage, 'initial store'),
         return self._parse_json(
             self._search_regex(
                 r'window\.__INITIAL_STORE__\s*=\s*({.+?});\s*</script>',
                 webpage, 'initial store'),

-            video_id)['default'][item]
+            video_id)

 
 
 class VevoIE(VevoBaseIE):
 
 
 class VevoIE(VevoBaseIE):


@@ -34,6 +34,7 @@ class VevoIE(VevoBaseIE):
         (?:https?://(?:www\.)?vevo\.com/watch/(?!playlist|genre)(?:[^/]+/(?:[^/]+/)?)?|
            https?://cache\.vevo\.com/m/html/embed\.html\?video=|
            https?://videoplayer\.vevo\.com/embed/embedded\?videoId=|
         (?:https?://(?:www\.)?vevo\.com/watch/(?!playlist|genre)(?:[^/]+/(?:[^/]+/)?)?|
            https?://cache\.vevo\.com/m/html/embed\.html\?video=|
            https?://videoplayer\.vevo\.com/embed/embedded\?videoId=|

+           https?://embed\.vevo\.com/.*?[?&]isrc=|

            vevo:)
         (?P<id>[^&?#]+)'''
 
            vevo:)
         (?P<id>[^&?#]+)'''
 


@@ -139,22 +140,15 @@ class VevoIE(VevoBaseIE):
         # no genres available
         'url': 'http://www.vevo.com/watch/INS171400764',
         'only_matching': True,
         # no genres available
         'url': 'http://www.vevo.com/watch/INS171400764',
         'only_matching': True,

+    }, {
+        # Another case available only via the webpage; using streams/streamsV3 formats
+        # Geo-restricted to Netherlands/Germany
+        'url': 'http://www.vevo.com/watch/boostee/pop-corn-clip-officiel/FR1A91600909',
+        'only_matching': True,
+    }, {
+        'url': 'https://embed.vevo.com/?isrc=USH5V1923499&partnerId=4d61b777-8023-4191-9ede-497ed6c24647&partnerAdCode=',
+        'only_matching': True,

     }]
     }]

-    _SMIL_BASE_URL = 'http://smil.lvl3.vevo.com'
-    _SOURCE_TYPES = {
-        0: 'youtube',
-        1: 'brightcove',
-        2: 'http',
-        3: 'hls_ios',
-        4: 'hls',
-        5: 'smil',  # http
-        7: 'f4m_cc',
-        8: 'f4m_ak',
-        9: 'f4m_l3',
-        10: 'ism',
-        13: 'smil',  # rtmp
-        18: 'dash',
-    }

     _VERSIONS = {
         0: 'youtube',  # only in AuthenticateVideo videoVersions
         1: 'level3',
     _VERSIONS = {
         0: 'youtube',  # only in AuthenticateVideo videoVersions
         1: 'level3',


@@ -163,199 +157,119 @@ class VevoIE(VevoBaseIE):
         4: 'amazon',
     }
 
         4: 'amazon',
     }
 

-    def _parse_smil_formats(self, smil, smil_url, video_id, namespace=None, f4m_params=None, transform_rtmp_url=None):
-        formats = []
-        els = smil.findall('.//{http://www.w3.org/2001/SMIL20/Language}video')
-        for el in els:
-            src = el.attrib['src']
-            m = re.match(r'''(?xi)
-                (?P<ext>[a-z0-9]+):
-                (?P<path>
-                    [/a-z0-9]+     # The directory and main part of the URL
-                    _(?P<tbr>[0-9]+)k
-                    _(?P<width>[0-9]+)x(?P<height>[0-9]+)
-                    _(?P<vcodec>[a-z0-9]+)
-                    _(?P<vbr>[0-9]+)
-                    _(?P<acodec>[a-z0-9]+)
-                    _(?P<abr>[0-9]+)
-                    \.[a-z0-9]+  # File extension
-                )''', src)
-            if not m:
-                continue
-
-            format_url = self._SMIL_BASE_URL + m.group('path')
-            formats.append({
-                'url': format_url,
-                'format_id': 'smil_' + m.group('tbr'),
-                'vcodec': m.group('vcodec'),
-                'acodec': m.group('acodec'),
-                'tbr': int(m.group('tbr')),
-                'vbr': int(m.group('vbr')),
-                'abr': int(m.group('abr')),
-                'ext': m.group('ext'),
-                'width': int(m.group('width')),
-                'height': int(m.group('height')),
-            })
-        return formats
-

     def _initialize_api(self, video_id):
     def _initialize_api(self, video_id):

-        req = sanitized_Request(
-            'http://www.vevo.com/auth', data=b'')

         webpage = self._download_webpage(
         webpage = self._download_webpage(

-            req, None,
+            'https://accounts.vevo.com/token', None,

             note='Retrieving oauth token',
             note='Retrieving oauth token',

-            errnote='Unable to retrieve oauth token')
+            errnote='Unable to retrieve oauth token',
+            data=json.dumps({
+                'client_id': 'SPupX1tvqFEopQ1YS6SS',
+                'grant_type': 'urn:vevo:params:oauth:grant-type:anonymous',
+            }).encode('utf-8'),
+            headers={
+                'Content-Type': 'application/json',
+            })

 
 

-        if 'THIS PAGE IS CURRENTLY UNAVAILABLE IN YOUR REGION' in webpage:
+        if re.search(r'(?i)THIS PAGE IS CURRENTLY UNAVAILABLE IN YOUR REGION', webpage):

             self.raise_geo_restricted(
                 '%s said: This page is currently unavailable in your region' % self.IE_NAME)
 
         auth_info = self._parse_json(webpage, video_id)
             self.raise_geo_restricted(
                 '%s said: This page is currently unavailable in your region' % self.IE_NAME)
 
         auth_info = self._parse_json(webpage, video_id)

-        self._api_url_template = self.http_scheme() + '//apiv2.vevo.com/%s?token=' + auth_info['access_token']
+        self._api_url_template = self.http_scheme() + '//apiv2.vevo.com/%s?token=' + auth_info['legacy_token']

 
     def _call_api(self, path, *args, **kwargs):
 
     def _call_api(self, path, *args, **kwargs):

-        return self._download_json(self._api_url_template % path, *args, **kwargs)
+        try:
+            data = self._download_json(self._api_url_template % path, *args, **kwargs)
+        except ExtractorError as e:
+            if isinstance(e.cause, compat_HTTPError):
+                errors = self._parse_json(e.cause.read().decode(), None)['errors']
+                error_message = ', '.join([error['message'] for error in errors])
+                raise ExtractorError('%s said: %s' % (self.IE_NAME, error_message), expected=True)
+            raise
+        return data

 
     def _real_extract(self, url):
         video_id = self._match_id(url)
 
 
     def _real_extract(self, url):
         video_id = self._match_id(url)
 

-        json_url = 'http://api.vevo.com/VideoService/AuthenticateVideo?isrc=%s' % video_id
-        response = self._download_json(
-            json_url, video_id, 'Downloading video info',
-            'Unable to download info', fatal=False) or {}
-        video_info = response.get('video') or {}
+        self._initialize_api(video_id)
+
+        video_info = self._call_api(
+            'video/%s' % video_id, video_id, 'Downloading api video info',
+            'Failed to download video info')
+
+        video_versions = self._call_api(
+            'video/%s/streams' % video_id, video_id,
+            'Downloading video versions info',
+            'Failed to download video versions info',
+            fatal=False)
+
+        # Some videos are only available via webpage (e.g.
+        # https://github.com/ytdl-org/youtube-dl/issues/9366)
+        if not video_versions:
+            webpage = self._download_webpage(url, video_id)
+            json_data = self._extract_json(webpage, video_id)
+            if 'streams' in json_data.get('default', {}):
+                video_versions = json_data['default']['streams'][video_id][0]
+            else:
+                video_versions = [
+                    value
+                    for key, value in json_data['apollo']['data'].items()
+                    if key.startswith('%s.streams' % video_id)]
+
+        uploader = None

         artist = None
         featured_artist = None
         artist = None
         featured_artist = None

-        uploader = None
-        view_count = None
+        artists = video_info.get('artists')
+        for curr_artist in artists:
+            if curr_artist.get('role') == 'Featured':
+                featured_artist = curr_artist['name']
+            else:
+                artist = uploader = curr_artist['name']
+

         formats = []
         formats = []

+        for video_version in video_versions:
+            version = self._VERSIONS.get(video_version.get('version'), 'generic')
+            version_url = video_version.get('url')
+            if not version_url:
+                continue

 
 

-        if not video_info:
-            try:
-                self._initialize_api(video_id)
-            except ExtractorError:
-                ytid = response.get('errorInfo', {}).get('ytid')
-                if ytid:
-                    self.report_warning(
-                        'Video is geoblocked, trying with the YouTube video %s' % ytid)
-                    return self.url_result(ytid, 'Youtube', ytid)
-
-                raise
-
-            video_info = self._call_api(
-                'video/%s' % video_id, video_id, 'Downloading api video info',
-                'Failed to download video info')
-
-            video_versions = self._call_api(
-                'video/%s/streams' % video_id, video_id,
-                'Downloading video versions info',
-                'Failed to download video versions info',
-                fatal=False)
-
-            # Some videos are only available via webpage (e.g.
-            # https://github.com/rg3/youtube-dl/issues/9366)
-            if not video_versions:
-                webpage = self._download_webpage(url, video_id)
-                video_versions = self._extract_json(webpage, video_id, 'streams')[video_id][0]
-
-            timestamp = parse_iso8601(video_info.get('releaseDate'))
-            artists = video_info.get('artists')
-            for curr_artist in artists:
-                if curr_artist.get('role') == 'Featured':
-                    featured_artist = curr_artist['name']
-                else:
-                    artist = uploader = curr_artist['name']
-            view_count = int_or_none(video_info.get('views', {}).get('total'))
-
-            for video_version in video_versions:
-                version = self._VERSIONS.get(video_version['version'])
-                version_url = video_version.get('url')
-                if not version_url:
+            if '.ism' in version_url:
+                continue
+            elif '.mpd' in version_url:
+                formats.extend(self._extract_mpd_formats(
+                    version_url, video_id, mpd_id='dash-%s' % version,
+                    note='Downloading %s MPD information' % version,
+                    errnote='Failed to download %s MPD information' % version,
+                    fatal=False))
+            elif '.m3u8' in version_url:
+                formats.extend(self._extract_m3u8_formats(
+                    version_url, video_id, 'mp4', 'm3u8_native',
+                    m3u8_id='hls-%s' % version,
+                    note='Downloading %s m3u8 information' % version,
+                    errnote='Failed to download %s m3u8 information' % version,
+                    fatal=False))
+            else:
+                m = re.search(r'''(?xi)
+                    _(?P<width>[0-9]+)x(?P<height>[0-9]+)
+                    _(?P<vcodec>[a-z0-9]+)
+                    _(?P<vbr>[0-9]+)
+                    _(?P<acodec>[a-z0-9]+)
+                    _(?P<abr>[0-9]+)
+                    \.(?P<ext>[a-z0-9]+)''', version_url)
+                if not m:

                     continue
 
                     continue
 

-                if '.ism' in version_url:
-                    continue
-                elif '.mpd' in version_url:
-                    formats.extend(self._extract_mpd_formats(
-                        version_url, video_id, mpd_id='dash-%s' % version,
-                        note='Downloading %s MPD information' % version,
-                        errnote='Failed to download %s MPD information' % version,
-                        fatal=False))
-                elif '.m3u8' in version_url:
-                    formats.extend(self._extract_m3u8_formats(
-                        version_url, video_id, 'mp4', 'm3u8_native',
-                        m3u8_id='hls-%s' % version,
-                        note='Downloading %s m3u8 information' % version,
-                        errnote='Failed to download %s m3u8 information' % version,
-                        fatal=False))
-                else:
-                    m = re.search(r'''(?xi)
-                        _(?P<width>[0-9]+)x(?P<height>[0-9]+)
-                        _(?P<vcodec>[a-z0-9]+)
-                        _(?P<vbr>[0-9]+)
-                        _(?P<acodec>[a-z0-9]+)
-                        _(?P<abr>[0-9]+)
-                        \.(?P<ext>[a-z0-9]+)''', version_url)
-                    if not m:
-                        continue
-
-                    formats.append({
-                        'url': version_url,
-                        'format_id': 'http-%s-%s' % (version, video_version['quality']),
-                        'vcodec': m.group('vcodec'),
-                        'acodec': m.group('acodec'),
-                        'vbr': int(m.group('vbr')),
-                        'abr': int(m.group('abr')),
-                        'ext': m.group('ext'),
-                        'width': int(m.group('width')),
-                        'height': int(m.group('height')),
-                    })
-        else:
-            timestamp = int_or_none(self._search_regex(
-                r'/Date\((\d+)\)/',
-                video_info['releaseDate'], 'release date', fatal=False),
-                scale=1000)
-            artists = video_info.get('mainArtists')
-            if artists:
-                artist = uploader = artists[0]['artistName']
-
-            featured_artists = video_info.get('featuredArtists')
-            if featured_artists:
-                featured_artist = featured_artists[0]['artistName']
-
-            smil_parsed = False
-            for video_version in video_info['videoVersions']:
-                version = self._VERSIONS.get(video_version['version'])
-                if version == 'youtube':
-                    continue
-                else:
-                    source_type = self._SOURCE_TYPES.get(video_version['sourceType'])
-                    renditions = compat_etree_fromstring(video_version['data'])
-                    if source_type == 'http':
-                        for rend in renditions.findall('rendition'):
-                            attr = rend.attrib
-                            formats.append({
-                                'url': attr['url'],
-                                'format_id': 'http-%s-%s' % (version, attr['name']),
-                                'height': int_or_none(attr.get('frameheight')),
-                                'width': int_or_none(attr.get('frameWidth')),
-                                'tbr': int_or_none(attr.get('totalBitrate')),
-                                'vbr': int_or_none(attr.get('videoBitrate')),
-                                'abr': int_or_none(attr.get('audioBitrate')),
-                                'vcodec': attr.get('videoCodec'),
-                                'acodec': attr.get('audioCodec'),
-                            })
-                    elif source_type == 'hls':
-                        formats.extend(self._extract_m3u8_formats(
-                            renditions.find('rendition').attrib['url'], video_id,
-                            'mp4', 'm3u8_native', m3u8_id='hls-%s' % version,
-                            note='Downloading %s m3u8 information' % version,
-                            errnote='Failed to download %s m3u8 information' % version,
-                            fatal=False))
-                    elif source_type == 'smil' and version == 'level3' and not smil_parsed:
-                        formats.extend(self._extract_smil_formats(
-                            renditions.find('rendition').attrib['url'], video_id, False))
-                        smil_parsed = True
+                formats.append({
+                    'url': version_url,
+                    'format_id': 'http-%s-%s' % (version, video_version['quality']),
+                    'vcodec': m.group('vcodec'),
+                    'acodec': m.group('acodec'),
+                    'vbr': int(m.group('vbr')),
+                    'abr': int(m.group('abr')),
+                    'ext': m.group('ext'),
+                    'width': int(m.group('width')),
+                    'height': int(m.group('height')),
+                })

         self._sort_formats(formats)
 
         track = video_info['title']
         self._sort_formats(formats)
 
         track = video_info['title']


@@ -365,8 +279,8 @@ class VevoIE(VevoBaseIE):
 
         genres = video_info.get('genres')
         genre = (
 
         genres = video_info.get('genres')
         genre = (

-            genres[0] if genres and isinstance(genres, list) and
-            isinstance(genres[0], compat_str) else None)
+            genres[0] if genres and isinstance(genres, list)
+            and isinstance(genres[0], compat_str) else None)

 
         is_explicit = video_info.get('isExplicit')
         if is_explicit is True:
 
         is_explicit = video_info.get('isExplicit')
         if is_explicit is True:


@@ -376,17 +290,15 @@ class VevoIE(VevoBaseIE):
         else:
             age_limit = None
 
         else:
             age_limit = None
 

-        duration = video_info.get('duration')
-

         return {
             'id': video_id,
             'title': title,
             'formats': formats,
             'thumbnail': video_info.get('imageUrl') or video_info.get('thumbnailUrl'),
         return {
             'id': video_id,
             'title': title,
             'formats': formats,
             'thumbnail': video_info.get('imageUrl') or video_info.get('thumbnailUrl'),

-            'timestamp': timestamp,
+            'timestamp': parse_iso8601(video_info.get('releaseDate')),

             'uploader': uploader,
             'uploader': uploader,

-            'duration': duration,
-            'view_count': view_count,
+            'duration': int_or_none(video_info.get('duration')),
+            'view_count': int_or_none(video_info.get('views', {}).get('total')),

             'age_limit': age_limit,
             'track': track,
             'artist': uploader,
             'age_limit': age_limit,
             'track': track,
             'artist': uploader,


@@ -448,7 +360,7 @@ class VevoPlaylistIE(VevoBaseIE):
             if video_id:
                 return self.url_result('vevo:%s' % video_id, VevoIE.ie_key())
 
             if video_id:
                 return self.url_result('vevo:%s' % video_id, VevoIE.ie_key())
 

-        playlists = self._extract_json(webpage, playlist_id, '%ss' % playlist_kind)
+        playlists = self._extract_json(webpage, playlist_id)['default']['%ss' % playlist_kind]

 
         playlist = (list(playlists.values())[0]
                     if playlist_kind == 'playlist' else playlists[playlist_id])
 
         playlist = (list(playlists.values())[0]
                     if playlist_kind == 'playlist' else playlists[playlist_id])