From 9165d6bab9ae3cc98984ebeb249ef320b9f4db83 Mon Sep 17 00:00:00 2001 From: remitamine Date: Tue, 26 Jan 2016 13:46:58 +0100 Subject: [PATCH] [vevo] extract metadata and formats from api if videoinfo is empty these was fixed by @yan12125 in ff51983e1563db08734e43a07d5671a517f99ef6 i only added some code to extract video metadata and more formats from api --- youtube_dl/extractor/vevo.py | 189 ++++++++++++++++++++++++++--------- 1 file changed, 140 insertions(+), 49 deletions(-) diff --git a/youtube_dl/extractor/vevo.py b/youtube_dl/extractor/vevo.py index ff854e74b..9c955c895 100644 --- a/youtube_dl/extractor/vevo.py +++ b/youtube_dl/extractor/vevo.py @@ -7,6 +7,8 @@ from ..utils import ( ExtractorError, int_or_none, + sanitized_Request, + parse_iso8601, ) @@ -64,6 +66,7 @@ class VevoIE(InfoExtractor): 'uploader_id': 'justintimberlakeVEVO', 'uploader': 'justintimberlakeVEVO', 'upload_date': '20130703', + 'duration': 419, }, 'params': { 'skip_download': 'true', @@ -72,13 +75,18 @@ class VevoIE(InfoExtractor): }, { 'note': 'No video_info', 'url': 'http://www.vevo.com/watch/k-camp-1/Till-I-Die/USUV71503000', - 'md5': '8b83cc492d72fc9cf74a02acee7dc1b0', + 'md5': 'a8b84d1d1957cd01046441b701b270fb', 'info_dict': { - 'id': 'USUV71503000', + 'id': 'Lad2jHtJCqY', 'ext': 'mp4', - 'title': 'Till I Die - K Camp ft. T.I.', + 'title': 'K Camp - Till I Die ft. T.I.', + 'description': 'md5:0694920ededdee4a14cfc39695cc8ec3', + 'uploader_id': 'KCampVEVO', + 'uploader': 'KCampVEVO', + 'upload_date': '20151207', 'duration': 193, }, + 'add_ie': ['Youtube'], }] _SMIL_BASE_URL = 'http://smil.lvl3.vevo.com' _SOURCE_TYPES = { @@ -96,7 +104,7 @@ class VevoIE(InfoExtractor): 18: 'dash', } _VERSIONS = { - 0: 'youtube', + 0: 'youtube', # only in AuthenticateVideo videoVersions 1: 'level3', 2: 'akamai', 3: 'level3', @@ -138,14 +146,34 @@ def _parse_smil_formats(self, smil, smil_url, video_id, namespace=None, f4m_para }) return formats + def _initialize_api(self, video_url, video_id): + req = sanitized_Request( + 'http://www.vevo.com/auth', data=b'') + webpage = self._download_webpage( + req, None, + note='Retrieving oauth token', + errnote='Unable to retrieve oauth token') + + if 'THIS PAGE IS CURRENTLY UNAVAILABLE IN YOUR REGION' in webpage: + raise ExtractorError('%s said: This page is currently unavailable in your region.' % self.IE_NAME, expected=True) + + 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, video_id, note, errnote, fatal=True): + return self._download_json(self._api_url_template % path, video_id, note, errnote) + def _real_extract(self, url): video_id = self._match_id(url) - webpage = None - json_url = 'http://videoplayer.vevo.com/VideoService/AuthenticateVideo?isrc=%s' % video_id - response = self._download_json(json_url, video_id) - video_info = response['video'] or {} + response = self._download_json(json_url, video_id, 'Downloading video info', 'Unable to download info') + video_info = response.get('video') or {} + video_versions = video_info.get('videoVersions') + uploader = None + timestamp = None + view_count = None + formats = [] if not video_info: ytid = response.get('errorInfo', {}).get('ytid') @@ -161,44 +189,112 @@ def _real_extract(self, url): if url.startswith('vevo:'): raise ExtractorError( 'Please specify full Vevo URL for downloading', expected=True) - webpage = self._download_webpage(url, video_id) - title = video_info.get('title') or self._og_search_title(webpage) + self._initialize_api(url, video_id) + video_info = self._call_api( + 'video/%s' % video_id, video_id, 'Downloading api video info', + 'Failed to download video info') - smil_parsed = False - formats = [] - for video_version in video_info['videoVersions']: - version = self._VERSIONS.get(video_version['version']) - if version == 'youtube': + ytid = video_info.get('youTubeId') + if ytid: return self.url_result( - video_version['id'], 'Youtube', video_version['id']) - 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': '%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': + ytid, 'Youtube', ytid) + + video_versions = self._call_api( + 'video/%s/streams' % video_id, video_id, + 'Downloading video versions info', + 'Failed to download video versions info') + + timestamp = parse_iso8601(video_info.get('releaseDate')) + artists = video_info.get('artists') + if artists: + uploader = artists[0]['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: + continue + + if '.mpd' in version_url or '.ism' in version_url: + continue + elif '.m3u8' in version_url: formats.extend(self._extract_m3u8_formats( - renditions.find('rendition').attrib['url'], video_id, - 'mp4', 'm3u8_native', m3u8_id='hls-%s' % version, fatal=False)) - elif source_type == 'smil' and not smil_parsed: - formats.extend(self._extract_smil_formats( - renditions.find('rendition').attrib['url'], video_id, False)) - smil_parsed = True + 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[0-9]+)x(?P[0-9]+) + _(?P[a-z0-9]+) + _(?P[0-9]+) + _(?P[a-z0-9]+) + _(?P[0-9]+) + \.(?P[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: + uploader = artists[0]['artistName'] + + smil_parsed = False + for video_version in video_info['videoVersions']: + version = self._VERSIONS.get(video_version['version']) + if version == 'youtube': + return self.url_result( + video_version['id'], 'Youtube', video_version['id']) + 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 not smil_parsed: + formats.extend(self._extract_smil_formats( + renditions.find('rendition').attrib['url'], video_id, False)) + smil_parsed = True self._sort_formats(formats) + title = video_info['title'] + is_explicit = video_info.get('isExplicit') if is_explicit is True: age_limit = 18 @@ -207,21 +303,16 @@ def _real_extract(self, url): else: age_limit = None - timestamp = int_or_none(self._search_regex( - r'/Date\((\d+)\)/', - video_info['launchDate'], 'launch date', fatal=False), - scale=1000) if video_info else None - - duration = video_info.get('duration') or int_or_none( - self._html_search_meta('video:duration', webpage)) + duration = video_info.get('duration') return { 'id': video_id, 'title': title, 'formats': formats, - 'thumbnail': video_info.get('imageUrl'), + 'thumbnail': video_info.get('imageUrl') or video_info.get('thumbnailUrl'), 'timestamp': timestamp, - 'uploader': video_info['mainArtists'][0]['artistName'] if video_info else None, + 'uploader': uploader, 'duration': duration, + 'view_count': view_count, 'age_limit': age_limit, }