diff options
Diffstat (limited to 'youtube_dl/extractor/brightcove.py')
| -rw-r--r-- | youtube_dl/extractor/brightcove.py | 557 |
1 files changed, 315 insertions, 242 deletions
diff --git a/youtube_dl/extractor/brightcove.py b/youtube_dl/extractor/brightcove.py index ef560b592..6022076ac 100644 --- a/youtube_dl/extractor/brightcove.py +++ b/youtube_dl/extractor/brightcove.py @@ -1,38 +1,45 @@ -# encoding: utf-8 +# coding: utf-8 from __future__ import unicode_literals +import base64 import re -import json +import struct +from .adobepass import AdobePassIE from .common import InfoExtractor from ..compat import ( compat_etree_fromstring, + compat_HTTPError, compat_parse_qs, - compat_str, compat_urllib_parse_urlparse, compat_urlparse, compat_xml_parse_error, - compat_HTTPError, ) from ..utils import ( - determine_ext, + clean_html, + extract_attributes, ExtractorError, find_xpath_attr, fix_xml_ampersands, float_or_none, - js_to_json, int_or_none, + js_to_json, + mimetype2ext, parse_iso8601, + smuggle_url, + str_or_none, + try_get, unescapeHTML, unsmuggle_url, + UnsupportedError, update_url_query, + url_or_none, ) class BrightcoveLegacyIE(InfoExtractor): IE_NAME = 'brightcove:legacy' _VALID_URL = r'(?:https?://.*brightcove\.com/(services|viewer).*?\?|brightcove:)(?P<query>.*)' - _FEDERATED_URL = 'http://c.brightcove.com/services/viewer/htmlFederated' _TESTS = [ { @@ -49,7 +56,8 @@ class BrightcoveLegacyIE(InfoExtractor): 'timestamp': 1368213670, 'upload_date': '20130510', 'uploader_id': '1589608506001', - } + }, + 'skip': 'The player has been deactivated by the content owner', }, { # From http://medianetwork.oracle.com/video/player/1785452137001 @@ -64,6 +72,7 @@ class BrightcoveLegacyIE(InfoExtractor): 'upload_date': '20120814', 'uploader_id': '1460825906', }, + 'skip': 'video not playable', }, { # From http://mashable.com/2013/10/26/thermoelectric-bracelet-lets-you-control-your-body-temperature/ @@ -73,7 +82,7 @@ class BrightcoveLegacyIE(InfoExtractor): 'ext': 'mp4', 'title': 'This Bracelet Acts as a Personal Thermostat', 'description': 'md5:547b78c64f4112766ccf4e151c20b6a0', - 'uploader': 'Mashable', + # 'uploader': 'Mashable', 'timestamp': 1382041798, 'upload_date': '20131017', 'uploader_id': '1130468786001', @@ -90,6 +99,7 @@ class BrightcoveLegacyIE(InfoExtractor): 'description': 'md5:363109c02998fee92ec02211bd8000df', 'uploader': 'National Ballet of Canada', }, + 'skip': 'Video gone', }, { # test flv videos served by akamaihd.net @@ -106,9 +116,10 @@ class BrightcoveLegacyIE(InfoExtractor): 'upload_date': '20140827', 'uploader_id': '710858724001', }, + 'skip': 'Video gone', }, { - # playlist test + # playlist with 'videoList' # from http://support.brightcove.com/en/video-cloud/docs/playlist-support-single-video-players 'url': 'http://c.brightcove.com/services/viewer/htmlFederated?playerID=3550052898001&playerKey=AQ%7E%7E%2CAAABmA9XpXk%7E%2C-Kp7jNgisre1fG5OdqpAFUTcs0lP_ZoL', 'info_dict': { @@ -116,26 +127,37 @@ class BrightcoveLegacyIE(InfoExtractor): 'id': '3550319591001', }, 'playlist_mincount': 7, + 'skip': 'Unsupported URL', + }, + { + # playlist with 'playlistTab' (https://github.com/ytdl-org/youtube-dl/issues/9965) + 'url': 'http://c.brightcove.com/services/json/experience/runtime/?command=get_programming_for_experience&playerKey=AQ%7E%7E,AAABXlLMdok%7E,NJ4EoMlZ4rZdx9eU1rkMVd8EaYPBBUlg', + 'info_dict': { + 'id': '1522758701001', + 'title': 'Lesson 08', + }, + 'playlist_mincount': 10, + 'skip': 'Unsupported URL', }, + { + # playerID inferred from bcpid + # from http://www.un.org/chinese/News/story.asp?NewsID=27724 + 'url': 'https://link.brightcove.com/services/player/bcpid1722935254001/?bctid=5360463607001&autoStart=false&secureConnections=true&width=650&height=350', + 'only_matching': True, # Tested in GenericIE + } ] - FLV_VCODECS = { - 1: 'SORENSON', - 2: 'ON2', - 3: 'H264', - 4: 'VP8', - } @classmethod - def _build_brighcove_url(cls, object_str): + def _build_brightcove_url(cls, object_str): """ Build a Brightcove url from a xml string containing <object class="BrightcoveExperience">{params}</object> """ - # Fix up some stupid HTML, see https://github.com/rg3/youtube-dl/issues/1553 + # Fix up some stupid HTML, see https://github.com/ytdl-org/youtube-dl/issues/1553 object_str = re.sub(r'(<param(?:\s+[a-zA-Z0-9_]+="[^"]*")*)>', lambda m: m.group(1) + '/>', object_str) - # Fix up some stupid XML, see https://github.com/rg3/youtube-dl/issues/1608 + # Fix up some stupid XML, see https://github.com/ytdl-org/youtube-dl/issues/1608 object_str = object_str.replace('<--', '<!--') # remove namespace to simplify extraction object_str = re.sub(r'(<object[^>]*)(xmlns=".*?")', r'\1', object_str) @@ -167,7 +189,7 @@ class BrightcoveLegacyIE(InfoExtractor): params = {} - playerID = find_param('playerID') + playerID = find_param('playerID') or find_param('playerId') if playerID is None: raise ExtractorError('Cannot find player ID') params['playerID'] = playerID @@ -179,6 +201,16 @@ class BrightcoveLegacyIE(InfoExtractor): # These fields hold the id of the video videoPlayer = find_param('@videoPlayer') or find_param('videoId') or find_param('videoID') or find_param('@videoList') if videoPlayer is not None: + if isinstance(videoPlayer, list): + videoPlayer = videoPlayer[0] + videoPlayer = videoPlayer.strip() + # UUID is also possible for videoPlayer (e.g. + # http://www.popcornflix.com/hoodies-vs-hooligans/7f2d2b87-bbf2-4623-acfb-ea942b4f01dd + # or http://www8.hp.com/cn/zh/home.html) + if not (re.match( + r'^(?:\d+|[\da-fA-F]{8}-?[\da-fA-F]{4}-?[\da-fA-F]{4}-?[\da-fA-F]{4}-?[\da-fA-F]{12})$', + videoPlayer) or videoPlayer.startswith('ref:')): + return None params['@videoPlayer'] = videoPlayer linkBase = find_param('linkBaseURL') if linkBase is not None: @@ -186,13 +218,13 @@ class BrightcoveLegacyIE(InfoExtractor): return cls._make_brightcove_url(params) @classmethod - def _build_brighcove_url_from_js(cls, object_js): + def _build_brightcove_url_from_js(cls, object_js): # The layout of JS is as follows: # customBC.createVideo = function (width, height, playerID, playerKey, videoPlayer, VideoRandomID) { # // build Brightcove <object /> XML # } m = re.search( - r'''(?x)customBC.\createVideo\( + r'''(?x)customBC\.createVideo\( .*? # skipping width and height ["\'](?P<playerID>\d+)["\']\s*,\s* # playerID ["\'](?P<playerKey>AQ[^"\']{48})[^"\']*["\']\s*,\s* # playerKey begins with AQ and is 50 characters @@ -205,7 +237,8 @@ class BrightcoveLegacyIE(InfoExtractor): @classmethod def _make_brightcove_url(cls, params): - return update_url_query(cls._FEDERATED_URL, params) + return update_url_query( + 'http://c.brightcove.com/services/viewer/htmlFederated', params) @classmethod def _extract_brightcove_url(cls, webpage): @@ -220,13 +253,16 @@ class BrightcoveLegacyIE(InfoExtractor): """Return a list of all Brightcove URLs from the webpage """ url_m = re.search( - r'<meta\s+property=[\'"]og:video[\'"]\s+content=[\'"](https?://(?:secure|c)\.brightcove.com/[^\'"]+)[\'"]', - webpage) + r'''(?x) + <meta\s+ + (?:property|itemprop)=([\'"])(?:og:video|embedURL)\1[^>]+ + content=([\'"])(?P<url>https?://(?:secure|c)\.brightcove.com/(?:(?!\2).)+)\2 + ''', webpage) if url_m: - url = unescapeHTML(url_m.group(1)) + url = unescapeHTML(url_m.group('url')) # Some sites don't add it, we can't download with this url, for example: # http://www.ktvu.com/videos/news/raw-video-caltrain-releases-video-of-man-almost/vCTZdY/ - if 'playerKey' in url or 'videoId' in url: + if 'playerKey' in url or 'videoId' in url or 'idVideo' in url: return [url] matches = re.findall( @@ -237,17 +273,21 @@ class BrightcoveLegacyIE(InfoExtractor): ).+?>\s*</object>''', webpage) if matches: - return list(filter(None, [cls._build_brighcove_url(m) for m in matches])) + return list(filter(None, [cls._build_brightcove_url(m) for m in matches])) - return list(filter(None, [ - cls._build_brighcove_url_from_js(custom_bc) - for custom_bc in re.findall(r'(customBC\.createVideo\(.+?\);)', webpage)])) + matches = re.findall(r'(customBC\.createVideo\(.+?\);)', webpage) + if matches: + return list(filter(None, [ + cls._build_brightcove_url_from_js(custom_bc) + for custom_bc in matches])) + return [src for _, src in re.findall( + r'<iframe[^>]+src=([\'"])((?:https?:)?//link\.brightcove\.com/services/player/(?!\1).+)\1', webpage)] def _real_extract(self, url): url, smuggled_data = unsmuggle_url(url, {}) # Change the 'videoId' and others field to '@videoPlayer' - url = re.sub(r'(?<=[?&])(videoI(d|D)|bctid)', '%40videoPlayer', url) + url = re.sub(r'(?<=[?&])(videoI(d|D)|idVideo|bctid)', '%40videoPlayer', url) # Change bckey (used by bcove.me urls) to playerKey url = re.sub(r'(?<=[?&])bckey', 'playerKey', url) mobj = re.match(self._VALID_URL, url) @@ -257,153 +297,50 @@ class BrightcoveLegacyIE(InfoExtractor): videoPlayer = query.get('@videoPlayer') if videoPlayer: # We set the original url as the default 'Referer' header - referer = smuggled_data.get('Referer', url) - return self._get_video_info( - videoPlayer[0], query, referer=referer) - elif 'playerKey' in query: - player_key = query['playerKey'] - return self._get_playlist_info(player_key[0]) - else: - raise ExtractorError( - 'Cannot find playerKey= variable. Did you forget quotes in a shell invocation?', - expected=True) - - def _get_video_info(self, video_id, query, referer=None): - headers = {} - linkBase = query.get('linkBaseURL') - if linkBase is not None: - referer = linkBase[0] - if referer is not None: - headers['Referer'] = referer - webpage = self._download_webpage(self._FEDERATED_URL, video_id, headers=headers, query=query) - - error_msg = self._html_search_regex( - r"<h1>We're sorry.</h1>([\s\n]*<p>.*?</p>)+", webpage, - 'error message', default=None) - if error_msg is not None: - raise ExtractorError( - 'brightcove said: %s' % error_msg, expected=True) - - self.report_extraction(video_id) - info = self._search_regex(r'var experienceJSON = ({.*});', webpage, 'json') - info = json.loads(info)['data'] - video_info = info['programmedContent']['videoPlayer']['mediaDTO'] - video_info['_youtubedl_adServerURL'] = info.get('adServerURL') - - return self._extract_video_info(video_info) - - def _get_playlist_info(self, player_key): - info_url = 'http://c.brightcove.com/services/json/experience/runtime/?command=get_programming_for_experience&playerKey=%s' % player_key - playlist_info = self._download_webpage( - info_url, player_key, 'Downloading playlist information') - - json_data = json.loads(playlist_info) - if 'videoList' not in json_data: - raise ExtractorError('Empty playlist') - playlist_info = json_data['videoList'] - videos = [self._extract_video_info(video_info) for video_info in playlist_info['mediaCollectionDTO']['videoDTOs']] - - return self.playlist_result(videos, playlist_id='%s' % playlist_info['id'], - playlist_title=playlist_info['mediaCollectionDTO']['displayName']) - - def _extract_video_info(self, video_info): - video_id = compat_str(video_info['id']) - publisher_id = video_info.get('publisherId') - info = { - 'id': video_id, - 'title': video_info['displayName'].strip(), - 'description': video_info.get('shortDescription'), - 'thumbnail': video_info.get('videoStillURL') or video_info.get('thumbnailURL'), - 'uploader': video_info.get('publisherName'), - 'uploader_id': compat_str(publisher_id) if publisher_id else None, - 'duration': float_or_none(video_info.get('length'), 1000), - 'timestamp': int_or_none(video_info.get('creationDate'), 1000), - } - - renditions = video_info.get('renditions', []) + video_info.get('IOSRenditions', []) - if renditions: - formats = [] - for rend in renditions: - url = rend['defaultURL'] - if not url: - continue - ext = None - if rend['remote']: - url_comp = compat_urllib_parse_urlparse(url) - if url_comp.path.endswith('.m3u8'): - formats.extend( - self._extract_m3u8_formats( - url, video_id, 'mp4', 'm3u8_native', m3u8_id='hls', fatal=False)) - continue - elif 'akamaihd.net' in url_comp.netloc: - # This type of renditions are served through - # akamaihd.net, but they don't use f4m manifests - url = url.replace('control/', '') + '?&v=3.3.0&fp=13&r=FEEFJ&g=RTSJIMBMPFPB' - ext = 'flv' - if ext is None: - ext = determine_ext(url) - tbr = int_or_none(rend.get('encodingRate'), 1000) - a_format = { - 'format_id': 'http%s' % ('-%s' % tbr if tbr else ''), - 'url': url, - 'ext': ext, - 'filesize': int_or_none(rend.get('size')) or None, - 'tbr': tbr, - } - if rend.get('audioOnly'): - a_format.update({ - 'vcodec': 'none', - }) + referer = query.get('linkBaseURL', [None])[0] or smuggled_data.get('Referer', url) + video_id = videoPlayer[0] + if 'playerID' not in query: + mobj = re.search(r'/bcpid(\d+)', url) + if mobj is not None: + query['playerID'] = [mobj.group(1)] + publisher_id = query.get('publisherId') + if publisher_id and publisher_id[0].isdigit(): + publisher_id = publisher_id[0] + if not publisher_id: + player_key = query.get('playerKey') + if player_key and ',' in player_key[0]: + player_key = player_key[0] else: - a_format.update({ - 'height': int_or_none(rend.get('frameHeight')), - 'width': int_or_none(rend.get('frameWidth')), - 'vcodec': rend.get('videoCodec'), - }) - - # m3u8 manifests with remote == false are media playlists - # Not calling _extract_m3u8_formats here to save network traffic - if ext == 'm3u8': - a_format.update({ - 'format_id': 'hls%s' % ('-%s' % tbr if tbr else ''), - 'ext': 'mp4', - 'protocol': 'm3u8_native', - }) - - formats.append(a_format) - self._sort_formats(formats) - info['formats'] = formats - elif video_info.get('FLVFullLengthURL') is not None: - info.update({ - 'url': video_info['FLVFullLengthURL'], - 'vcodec': self.FLV_VCODECS.get(video_info.get('FLVFullCodec')), - 'filesize': int_or_none(video_info.get('FLVFullSize')), - }) - - if self._downloader.params.get('include_ads', False): - adServerURL = video_info.get('_youtubedl_adServerURL') - if adServerURL: - ad_info = { - '_type': 'url', - 'url': adServerURL, - } - if 'url' in info: - return { - '_type': 'playlist', - 'title': info['title'], - 'entries': [ad_info, info], - } - else: - return ad_info - - if 'url' not in info and not info.get('formats'): - raise ExtractorError('Unable to extract video url for %s' % video_id) - return info - - -class BrightcoveNewIE(InfoExtractor): + player_id = query.get('playerID') + if player_id and player_id[0].isdigit(): + headers = {} + if referer: + headers['Referer'] = referer + player_page = self._download_webpage( + 'http://link.brightcove.com/services/player/bcpid' + player_id[0], + video_id, headers=headers, fatal=False) + if player_page: + player_key = self._search_regex( + r'<param\s+name="playerKey"\s+value="([\w~,-]+)"', + player_page, 'player key', fatal=False) + if player_key: + enc_pub_id = player_key.split(',')[1].replace('~', '=') + publisher_id = struct.unpack('>Q', base64.urlsafe_b64decode(enc_pub_id))[0] + if publisher_id: + brightcove_new_url = 'http://players.brightcove.net/%s/default_default/index.html?videoId=%s' % (publisher_id, video_id) + if referer: + brightcove_new_url = smuggle_url(brightcove_new_url, {'referrer': referer}) + return self.url_result(brightcove_new_url, BrightcoveNewIE.ie_key(), video_id) + # TODO: figure out if it's possible to extract playlistId from playerKey + # elif 'playerKey' in query: + # player_key = query['playerKey'] + # return self._get_playlist_info(player_key[0]) + raise UnsupportedError(url) + + +class BrightcoveNewIE(AdobePassIE): IE_NAME = 'brightcove:new' - _VALID_URL = r'https?://players\.brightcove\.net/(?P<account_id>\d+)/(?P<player_id>[^/]+)_(?P<embed>[^/]+)/index\.html\?.*videoId=(?P<video_id>\d+|ref:[^&]+)' + _VALID_URL = r'https?://players\.brightcove\.net/(?P<account_id>\d+)/(?P<player_id>[^/]+)_(?P<embed>[^/]+)/index\.html\?.*(?P<content_type>video|playlist)Id=(?P<video_id>\d+|ref:[^&]+)' _TESTS = [{ 'url': 'http://players.brightcove.net/929656772001/e41d32dc-ec74-459e-a845-6c69f7b724ea_default/index.html?videoId=4463358922001', 'md5': 'c8100925723840d4b0d243f7025703be', @@ -416,7 +353,7 @@ class BrightcoveNewIE(InfoExtractor): 'timestamp': 1441391203, 'upload_date': '20150904', 'uploader_id': '929656772001', - 'formats': 'mincount:22', + 'formats': 'mincount:20', }, }, { # with rtmp streams @@ -430,13 +367,28 @@ class BrightcoveNewIE(InfoExtractor): 'timestamp': 1433556729, 'upload_date': '20150606', 'uploader_id': '4036320279001', - 'formats': 'mincount:41', + 'formats': 'mincount:39', + }, + 'params': { + # m3u8 download + 'skip_download': True, + } + }, { + # playlist stream + 'url': 'https://players.brightcove.net/1752604059001/S13cJdUBz_default/index.html?playlistId=5718313430001', + 'info_dict': { + 'id': '5718313430001', + 'title': 'No Audio Playlist', }, + 'playlist_count': 7, 'params': { # m3u8 download 'skip_download': True, } }, { + 'url': 'http://players.brightcove.net/5690807595001/HyZNerRl7_default/index.html?playlistId=5743160747001', + 'only_matching': True, + }, { # ref: prefixed video id 'url': 'http://players.brightcove.net/3910869709001/21519b5c-4b3b-4363-accb-bdc8f358f823_default/index.html?videoId=ref:7069442', 'only_matching': True, @@ -451,17 +403,18 @@ class BrightcoveNewIE(InfoExtractor): }] @staticmethod - def _extract_url(webpage): - urls = BrightcoveNewIE._extract_urls(webpage) + def _extract_url(ie, webpage): + urls = BrightcoveNewIE._extract_urls(ie, webpage) 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 = [] @@ -470,72 +423,73 @@ 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(?:-js)?\s+[^>]*\bdata-video-id\s*=\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) - return entries + # 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 - def _real_extract(self, url): - account_id, player_id, embed, video_id = re.match(self._VALID_URL, url).groups() + account_id = account_id or attrs.get('data-account') + if not account_id: + continue - webpage = self._download_webpage( - 'http://players.brightcove.net/%s/%s_%s/index.min.js' - % (account_id, player_id, embed), video_id) + player_id = player_id or attrs.get('data-player') or 'default' + embed = embed or attrs.get('data-embed') or 'default' - policy_key = None + bc_url = 'http://players.brightcove.net/%s/%s_%s/index.html?videoId=%s' % ( + account_id, player_id, embed, video_id) - catalog = self._search_regex( - r'catalog\(({.+?})\);', webpage, 'catalog', default=None) - if catalog: - catalog = self._parse_json( - js_to_json(catalog), video_id, fatal=False) - if catalog: - policy_key = catalog.get('policyKey') + # 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 - if not policy_key: - policy_key = self._search_regex( - r'policyKey\s*:\s*(["\'])(?P<pk>.+?)\1', - webpage, 'policy key', group='pk') + entries.append(bc_url) - api_url = 'https://edge.api.brightcove.com/playback/v1/accounts/%s/videos/%s' % (account_id, video_id) - try: - json_data = self._download_json(api_url, video_id, headers={ - 'Accept': 'application/json;pk=%s' % policy_key - }) - except ExtractorError as e: - if isinstance(e.cause, compat_HTTPError) and e.cause.code == 403: - json_data = self._parse_json(e.cause.read().decode(), video_id)[0] - raise ExtractorError( - json_data.get('message') or json_data['error_code'], expected=True) - raise + return entries + def _parse_brightcove_metadata(self, json_data, video_id, headers={}): title = json_data['name'].strip() + num_drm_sources = 0 formats = [] - for source in json_data.get('sources', []): + sources = json_data.get('sources') or [] + for source in sources: container = source.get('container') - source_type = source.get('type') + ext = mimetype2ext(source.get('type')) src = source.get('src') - if source_type == 'application/x-mpegURL' or container == 'M2TS': + # https://support.brightcove.com/playback-api-video-fields-reference#key_systems_object + if container == 'WVM' or source.get('key_systems'): + num_drm_sources += 1 + continue + elif ext == 'ism': + continue + elif ext == 'm3u8' or container == 'M2TS': if not src: continue formats.extend(self._extract_m3u8_formats( src, video_id, 'mp4', 'm3u8_native', m3u8_id='hls', fatal=False)) - elif source_type == 'application/dash+xml': + elif ext == 'mpd': if not src: continue formats.extend(self._extract_mpd_formats(src, video_id, 'dash', fatal=False)) @@ -551,7 +505,7 @@ class BrightcoveNewIE(InfoExtractor): 'tbr': tbr, 'filesize': int_or_none(source.get('size')), 'container': container, - 'ext': container.lower(), + 'ext': ext or container.lower(), } if width == 0 and height == 0: f.update({ @@ -585,24 +539,143 @@ class BrightcoveNewIE(InfoExtractor): 'format_id': build_format_id('rtmp'), }) formats.append(f) + + if not formats: + errors = json_data.get('errors') + if errors: + error = errors[0] + raise ExtractorError( + error.get('message') or error.get('error_subcode') or error['error_code'], expected=True) + if sources and num_drm_sources == len(sources): + raise ExtractorError('This video is DRM protected.', expected=True) + self._sort_formats(formats) + for f in formats: + f.setdefault('http_headers', {}).update(headers) + subtitles = {} for text_track in json_data.get('text_tracks', []): - if text_track.get('src'): - subtitles.setdefault(text_track.get('srclang'), []).append({ - 'url': text_track['src'], - }) + if text_track.get('kind') != 'captions': + continue + text_track_url = url_or_none(text_track.get('src')) + if not text_track_url: + continue + lang = (str_or_none(text_track.get('srclang')) + or str_or_none(text_track.get('label')) or 'en').lower() + subtitles.setdefault(lang, []).append({ + 'url': text_track_url, + }) + + is_live = False + duration = float_or_none(json_data.get('duration'), 1000) + if duration is not None and duration <= 0: + is_live = True return { 'id': video_id, - 'title': title, - 'description': json_data.get('description'), + 'title': self._live_title(title) if is_live else title, + 'description': clean_html(json_data.get('description')), 'thumbnail': json_data.get('thumbnail') or json_data.get('poster'), - 'duration': float_or_none(json_data.get('duration'), 1000), + 'duration': duration, 'timestamp': parse_iso8601(json_data.get('published_at')), - 'uploader_id': account_id, + 'uploader_id': json_data.get('account_id'), 'formats': formats, 'subtitles': subtitles, 'tags': json_data.get('tags', []), + 'is_live': is_live, } + + def _real_extract(self, url): + url, smuggled_data = unsmuggle_url(url, {}) + self._initialize_geo_bypass({ + 'countries': smuggled_data.get('geo_countries'), + 'ip_blocks': smuggled_data.get('geo_ip_blocks'), + }) + + account_id, player_id, embed, content_type, video_id = re.match(self._VALID_URL, url).groups() + + policy_key_id = '%s_%s' % (account_id, player_id) + policy_key = self._downloader.cache.load('brightcove', policy_key_id) + policy_key_extracted = False + store_pk = lambda x: self._downloader.cache.store('brightcove', policy_key_id, x) + + def extract_policy_key(): + base_url = 'http://players.brightcove.net/%s/%s_%s/' % (account_id, player_id, embed) + config = self._download_json( + base_url + 'config.json', video_id, fatal=False) or {} + policy_key = try_get( + config, lambda x: x['video_cloud']['policy_key']) + if not policy_key: + webpage = self._download_webpage( + base_url + 'index.min.js', video_id) + + catalog = self._search_regex( + r'catalog\(({.+?})\);', webpage, 'catalog', default=None) + if catalog: + catalog = self._parse_json( + js_to_json(catalog), video_id, fatal=False) + if catalog: + policy_key = catalog.get('policyKey') + + if not policy_key: + policy_key = self._search_regex( + r'policyKey\s*:\s*(["\'])(?P<pk>.+?)\1', + webpage, 'policy key', group='pk') + + store_pk(policy_key) + return policy_key + + api_url = 'https://edge.api.brightcove.com/playback/v1/accounts/%s/%ss/%s' % (account_id, content_type, video_id) + headers = {} + referrer = smuggled_data.get('referrer') + if referrer: + headers.update({ + 'Referer': referrer, + 'Origin': re.search(r'https?://[^/]+', referrer).group(0), + }) + + for _ in range(2): + if not policy_key: + policy_key = extract_policy_key() + policy_key_extracted = True + headers['Accept'] = 'application/json;pk=%s' % policy_key + try: + json_data = self._download_json(api_url, video_id, headers=headers) + break + except ExtractorError as e: + if isinstance(e.cause, compat_HTTPError) and e.cause.code in (401, 403): + json_data = self._parse_json(e.cause.read().decode(), video_id)[0] + message = json_data.get('message') or json_data['error_code'] + if json_data.get('error_subcode') == 'CLIENT_GEO': + self.raise_geo_restricted(msg=message) + elif json_data.get('error_code') == 'INVALID_POLICY_KEY' and not policy_key_extracted: + policy_key = None + store_pk(None) + continue + raise ExtractorError(message, expected=True) + raise + + errors = json_data.get('errors') + if errors and errors[0].get('error_subcode') == 'TVE_AUTH': + custom_fields = json_data['custom_fields'] + tve_token = self._extract_mvpd_auth( + smuggled_data['source_url'], video_id, + custom_fields['bcadobepassrequestorid'], + custom_fields['bcadobepassresourceid']) + json_data = self._download_json( + api_url, video_id, headers={ + 'Accept': 'application/json;pk=%s' % policy_key + }, query={ + 'tveToken': tve_token, + }) + + if content_type == 'playlist': + return self.playlist_result( + [self._parse_brightcove_metadata(vid, vid.get('id'), headers) + for vid in json_data.get('videos', []) if vid.get('id')], + json_data.get('id'), json_data.get('name'), + json_data.get('description')) + + return self._parse_brightcove_metadata( + json_data, video_id, headers=headers) |