Prepare for upload.

[youtubedl] / youtube_dl / extractor / vevo.py

diff --git a/youtube_dl/extractor/vevo.py b/youtube_dl/extractor/vevo.py
index 4c0de354f7d8ac467f0f027b201d785a38c77ffa..c4e37f694426c175b1f33d7795ca01baf6f7547b 100644 (file)
--- a/youtube_dl/extractor/vevo.py
+++ b/youtube_dl/extractor/vevo.py
@@ -4,22 +4,34 @@ import re
 
 from .common import InfoExtractor
 from ..compat import (
 
 from .common import InfoExtractor
 from ..compat import (

-    compat_etree_fromstring,
-    compat_urllib_request,
+    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 VevoIE(InfoExtractor):
-    """
+class VevoBaseIE(InfoExtractor):
+    def _extract_json(self, webpage, video_id, item):
+        return self._parse_json(
+            self._search_regex(
+                r'window\.__INITIAL_STORE__\s*=\s*({.+?});\s*</script>',
+                webpage, 'initial store'),
+            video_id)['default'][item]
+
+
+class VevoIE(VevoBaseIE):
+    '''

     Accepts urls from vevo.com or in the format 'vevo:{id}'
     (currently used by MTVIE and MySpaceIE)
     Accepts urls from vevo.com or in the format 'vevo:{id}'
     (currently used by MTVIE and MySpaceIE)

-    """
+    '''

     _VALID_URL = r'''(?x)
     _VALID_URL = r'''(?x)

-        (?:https?://www\.vevo\.com/watch/(?:[^/]+/(?:[^/]+/)?)?|
+        (?:https?://(?:www\.)?vevo\.com/watch/(?!playlist|genre)(?:[^/]+/(?:[^/]+/)?)?|

            https?://cache\.vevo\.com/m/html/embed\.html\?video=|
            https?://videoplayer\.vevo\.com/embed/embedded\?videoId=|
            vevo:)
            https?://cache\.vevo\.com/m/html/embed\.html\?video=|
            https?://videoplayer\.vevo\.com/embed/embedded\?videoId=|
            vevo:)


@@ -27,19 +39,19 @@ class VevoIE(InfoExtractor):
 
     _TESTS = [{
         'url': 'http://www.vevo.com/watch/hurts/somebody-to-die-for/GB1101300280',
 
     _TESTS = [{
         'url': 'http://www.vevo.com/watch/hurts/somebody-to-die-for/GB1101300280',

-        "md5": "95ee28ee45e70130e3ab02b0f579ae23",
+        'md5': '95ee28ee45e70130e3ab02b0f579ae23',

         'info_dict': {
             'id': 'GB1101300280',
             'ext': 'mp4',
         'info_dict': {
             'id': 'GB1101300280',
             'ext': 'mp4',

-            "upload_date": "20130624",
-            "uploader": "Hurts",
-            "title": "Somebody to Die For",
-            "duration": 230.12,
-            "width": 1920,
-            "height": 1080,
-            # timestamp and upload_date are often incorrect; seem to change randomly
-            'timestamp': int,
-        }
+            'title': 'Hurts - Somebody to Die For',
+            'timestamp': 1372057200,
+            'upload_date': '20130624',
+            'uploader': 'Hurts',
+            'track': 'Somebody to Die For',
+            'artist': 'Hurts',
+            'genre': 'Pop',
+        },
+        'expected_warnings': ['Unable to download SMIL file', 'Unable to download info'],

     }, {
         'note': 'v3 SMIL format',
         'url': 'http://www.vevo.com/watch/cassadee-pope/i-wish-i-could-break-your-heart/USUV71302923',
     }, {
         'note': 'v3 SMIL format',
         'url': 'http://www.vevo.com/watch/cassadee-pope/i-wish-i-could-break-your-heart/USUV71302923',


@@ -47,140 +59,207 @@ class VevoIE(InfoExtractor):
         'info_dict': {
             'id': 'USUV71302923',
             'ext': 'mp4',
         'info_dict': {
             'id': 'USUV71302923',
             'ext': 'mp4',

+            'title': 'Cassadee Pope - I Wish I Could Break Your Heart',
+            'timestamp': 1392796919,

             'upload_date': '20140219',
             'uploader': 'Cassadee Pope',
             'upload_date': '20140219',
             'uploader': 'Cassadee Pope',

-            'title': 'I Wish I Could Break Your Heart',
-            'duration': 226.101,
-            'age_limit': 0,
-            'timestamp': int,
-        }
+            'track': 'I Wish I Could Break Your Heart',
+            'artist': 'Cassadee Pope',
+            'genre': 'Country',
+        },
+        'expected_warnings': ['Unable to download SMIL file', 'Unable to download info'],

     }, {
         'note': 'Age-limited video',
         'url': 'https://www.vevo.com/watch/justin-timberlake/tunnel-vision-explicit/USRV81300282',
         'info_dict': {
             'id': 'USRV81300282',
             'ext': 'mp4',
     }, {
         'note': 'Age-limited video',
         'url': 'https://www.vevo.com/watch/justin-timberlake/tunnel-vision-explicit/USRV81300282',
         'info_dict': {
             'id': 'USRV81300282',
             'ext': 'mp4',

+            'title': 'Justin Timberlake - Tunnel Vision (Explicit)',

             'age_limit': 18,
             'age_limit': 18,

-            'title': 'Tunnel Vision (Explicit)',
+            'timestamp': 1372888800,
+            'upload_date': '20130703',

             'uploader': 'Justin Timberlake',
             'uploader': 'Justin Timberlake',

-            'upload_date': 're:2013070[34]',
-            'timestamp': int,
+            'track': 'Tunnel Vision (Explicit)',
+            'artist': 'Justin Timberlake',
+            'genre': 'Pop',

         },
         },

-        'params': {
-            'skip_download': 'true',
-        }
+        'expected_warnings': ['Unable to download SMIL file', 'Unable to download info'],
+    }, {
+        'note': 'No video_info',
+        'url': 'http://www.vevo.com/watch/k-camp-1/Till-I-Die/USUV71503000',
+        'md5': '8b83cc492d72fc9cf74a02acee7dc1b0',
+        'info_dict': {
+            'id': 'USUV71503000',
+            'ext': 'mp4',
+            'title': 'K Camp ft. T.I. - Till I Die',
+            'age_limit': 18,
+            'timestamp': 1449468000,
+            'upload_date': '20151207',
+            'uploader': 'K Camp',
+            'track': 'Till I Die',
+            'artist': 'K Camp',
+            'genre': 'Hip-Hop',
+        },
+        'expected_warnings': ['Unable to download SMIL file', 'Unable to download info'],
+    }, {
+        'note': 'Featured test',
+        'url': 'https://www.vevo.com/watch/lemaitre/Wait/USUV71402190',
+        'md5': 'd28675e5e8805035d949dc5cf161071d',
+        'info_dict': {
+            'id': 'USUV71402190',
+            'ext': 'mp4',
+            'title': 'Lemaitre ft. LoLo - Wait',
+            'age_limit': 0,
+            'timestamp': 1413432000,
+            'upload_date': '20141016',
+            'uploader': 'Lemaitre',
+            'track': 'Wait',
+            'artist': 'Lemaitre',
+            'genre': 'Electronic',
+        },
+        'expected_warnings': ['Unable to download SMIL file', 'Unable to download info'],
+    }, {
+        'note': 'Only available via webpage',
+        'url': 'http://www.vevo.com/watch/GBUV71600656',
+        'md5': '67e79210613865b66a47c33baa5e37fe',
+        'info_dict': {
+            'id': 'GBUV71600656',
+            'ext': 'mp4',
+            'title': 'ABC - Viva Love',
+            'age_limit': 0,
+            'timestamp': 1461830400,
+            'upload_date': '20160428',
+            'uploader': 'ABC',
+            'track': 'Viva Love',
+            'artist': 'ABC',
+            'genre': 'Pop',
+        },
+        'expected_warnings': ['Failed to download video versions info'],
+    }, {
+        # no genres available
+        'url': 'http://www.vevo.com/watch/INS171400764',
+        'only_matching': True,

     }]
     }]

-    _SMIL_BASE_URL = 'http://smil.lvl3.vevo.com/'
+    _VERSIONS = {
+        0: 'youtube',  # only in AuthenticateVideo videoVersions
+        1: 'level3',
+        2: 'akamai',
+        3: 'level3',
+        4: 'amazon',
+    }

 
 

-    def _real_initialize(self):
-        req = compat_urllib_request.Request(
+    def _initialize_api(self, video_id):
+        req = sanitized_Request(

             'http://www.vevo.com/auth', data=b'')
         webpage = self._download_webpage(
             req, None,
             note='Retrieving oauth token',
             'http://www.vevo.com/auth', data=b'')
         webpage = self._download_webpage(
             req, None,
             note='Retrieving oauth token',

-            errnote='Unable to retrieve oauth token',
+            errnote='Unable to retrieve oauth token')
+
+        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._api_url_template = self.http_scheme() + '//apiv2.vevo.com/%s?token=' + auth_info['access_token']
+
+    def _call_api(self, 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)
+
+        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)
             fatal=False)

-        if webpage is False:
-            self._oauth_token = None
-        else:
-            self._oauth_token = self._search_regex(
-                r'access_token":\s*"([^"]+)"',
-                webpage, 'access token', fatal=False)
-
-    def _formats_from_json(self, video_info):
-        last_version = {'version': -1}
-        for version in video_info['videoVersions']:
-            # These are the HTTP downloads, other types are for different manifests
-            if version['sourceType'] == 2:
-                if version['version'] > last_version['version']:
-                    last_version = version
-        if last_version['version'] == -1:
-            raise ExtractorError('Unable to extract last version of the video')
-
-        renditions = compat_etree_fromstring(last_version['data'])
-        formats = []
-        # Already sorted from worst to best quality
-        for rend in renditions.findall('rendition'):
-            attr = rend.attrib
-            format_note = '%(videoCodec)s@%(videoBitrate)4sk, %(audioCodec)s@%(audioBitrate)3sk' % attr
-            formats.append({
-                'url': attr['url'],
-                'format_id': attr['name'],
-                'format_note': format_note,
-                'height': int(attr['frameheight']),
-                'width': int(attr['frameWidth']),
-            })
-        return formats
-
-    def _formats_from_smil(self, smil_xml):
+
+        # 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]
+
+        uploader = None
+        artist = None
+        featured_artist = 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 = []

-        smil_doc = compat_etree_fromstring(smil_xml.encode('utf-8'))
-        els = smil_doc.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<cbr>[0-9]+)k
+        for video_version in video_versions:
+            version = self._VERSIONS.get(video_version['version'])
+            version_url = video_version.get('url')
+            if not version_url:
+                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<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('cbr'),
-                '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')),
-            })
-        return formats
-
-    def _download_api_formats(self, video_id):
-        if not self._oauth_token:
-            self._downloader.report_warning(
-                'No oauth token available, skipping API HLS download')
-            return []
-
-        api_url = 'https://apiv2.vevo.com/video/%s/streams/hls?token=%s' % (
-            video_id, self._oauth_token)
-        api_data = self._download_json(
-            api_url, video_id,
-            note='Downloading HLS formats',
-            errnote='Failed to download HLS format list', fatal=False)
-        if api_data is None:
-            return []
-
-        m3u8_url = api_data[0]['url']
-        return self._extract_m3u8_formats(
-            m3u8_url, video_id, entry_protocol='m3u8_native', ext='mp4',
-            preference=0)
+                    \.(?P<ext>[a-z0-9]+)''', version_url)
+                if not m:
+                    continue

 
 

-    def _real_extract(self, url):
-        mobj = re.match(self._VALID_URL, url)
-        video_id = mobj.group('id')
-
-        json_url = 'http://videoplayer.vevo.com/VideoService/AuthenticateVideo?isrc=%s' % video_id
-        response = self._download_json(json_url, video_id)
-        video_info = response['video']
+                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)

 
 

-        if not video_info:
-            if 'statusMessage' in response:
-                raise ExtractorError('%s said: %s' % (self.IE_NAME, response['statusMessage']), expected=True)
-            raise ExtractorError('Unable to extract videos')
+        track = video_info['title']
+        if featured_artist:
+            artist = '%s ft. %s' % (artist, featured_artist)
+        title = '%s - %s' % (artist, track) if artist else track

 
 

-        formats = self._formats_from_json(video_info)
+        genres = video_info.get('genres')
+        genre = (
+            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:


@@ -190,40 +269,85 @@ class VevoIE(InfoExtractor):
         else:
             age_limit = None
 
         else:
             age_limit = None
 

-        # Download via HLS API
-        formats.extend(self._download_api_formats(video_id))
-
-        # Download SMIL
-        smil_blocks = sorted((
-            f for f in video_info['videoVersions']
-            if f['sourceType'] == 13),
-            key=lambda f: f['version'])
-        smil_url = '%s/Video/V2/VFILE/%s/%sr.smil' % (
-            self._SMIL_BASE_URL, video_id, video_id.lower())
-        if smil_blocks:
-            smil_url_m = self._search_regex(
-                r'url="([^"]+)"', smil_blocks[-1]['data'], 'SMIL URL',
-                default=None)
-            if smil_url_m is not None:
-                smil_url = smil_url_m
-        if smil_url:
-            smil_xml = self._download_webpage(
-                smil_url, video_id, 'Downloading SMIL info', fatal=False)
-            if smil_xml:
-                formats.extend(self._formats_from_smil(smil_xml))
-
-        self._sort_formats(formats)
-        timestamp_ms = int_or_none(self._search_regex(
-            r'/Date\((\d+)\)/',
-            video_info['launchDate'], 'launch date', fatal=False))
-

         return {
             'id': video_id,
         return {
             'id': video_id,

-            'title': video_info['title'],
+            'title': title,

             'formats': formats,
             'formats': formats,

-            'thumbnail': video_info['imageUrl'],
-            'timestamp': timestamp_ms // 1000,
-            'uploader': video_info['mainArtists'][0]['artistName'],
-            'duration': video_info['duration'],
+            'thumbnail': video_info.get('imageUrl') or video_info.get('thumbnailUrl'),
+            'timestamp': parse_iso8601(video_info.get('releaseDate')),
+            'uploader': uploader,
+            'duration': int_or_none(video_info.get('duration')),
+            'view_count': int_or_none(video_info.get('views', {}).get('total')),

             'age_limit': age_limit,
             'age_limit': age_limit,

+            'track': track,
+            'artist': uploader,
+            'genre': genre,

         }
         }

+
+
+class VevoPlaylistIE(VevoBaseIE):
+    _VALID_URL = r'https?://(?:www\.)?vevo\.com/watch/(?P<kind>playlist|genre)/(?P<id>[^/?#&]+)'
+
+    _TESTS = [{
+        'url': 'http://www.vevo.com/watch/playlist/dadbf4e7-b99f-4184-9670-6f0e547b6a29',
+        'info_dict': {
+            'id': 'dadbf4e7-b99f-4184-9670-6f0e547b6a29',
+            'title': 'Best-Of: Birdman',
+        },
+        'playlist_count': 10,
+    }, {
+        'url': 'http://www.vevo.com/watch/genre/rock',
+        'info_dict': {
+            'id': 'rock',
+            'title': 'Rock',
+        },
+        'playlist_count': 20,
+    }, {
+        'url': 'http://www.vevo.com/watch/playlist/dadbf4e7-b99f-4184-9670-6f0e547b6a29?index=0',
+        'md5': '32dcdfddddf9ec6917fc88ca26d36282',
+        'info_dict': {
+            'id': 'USCMV1100073',
+            'ext': 'mp4',
+            'title': 'Birdman - Y.U. MAD',
+            'timestamp': 1323417600,
+            'upload_date': '20111209',
+            'uploader': 'Birdman',
+            'track': 'Y.U. MAD',
+            'artist': 'Birdman',
+            'genre': 'Rap/Hip-Hop',
+        },
+        'expected_warnings': ['Unable to download SMIL file'],
+    }, {
+        'url': 'http://www.vevo.com/watch/genre/rock?index=0',
+        'only_matching': True,
+    }]
+
+    def _real_extract(self, url):
+        mobj = re.match(self._VALID_URL, url)
+        playlist_id = mobj.group('id')
+        playlist_kind = mobj.group('kind')
+
+        webpage = self._download_webpage(url, playlist_id)
+
+        qs = compat_urlparse.parse_qs(compat_urlparse.urlparse(url).query)
+        index = qs.get('index', [None])[0]
+
+        if index:
+            video_id = self._search_regex(
+                r'<meta[^>]+content=(["\'])vevo://video/(?P<id>.+?)\1[^>]*>',
+                webpage, 'video id', default=None, group='id')
+            if video_id:
+                return self.url_result('vevo:%s' % video_id, VevoIE.ie_key())
+
+        playlists = self._extract_json(webpage, playlist_id, '%ss' % playlist_kind)
+
+        playlist = (list(playlists.values())[0]
+                    if playlist_kind == 'playlist' else playlists[playlist_id])
+
+        entries = [
+            self.url_result('vevo:%s' % src, VevoIE.ie_key())
+            for src in playlist['isrcs']]
+
+        return self.playlist_result(
+            entries, playlist.get('playlistId') or playlist_id,
+            playlist.get('name'), playlist.get('description'))