X-Git-Url: https://git.rapsys.eu/youtubedl/blobdiff_plain/03342304420e5daeb428ffdcc7bbd2bbfecfa61a..1bce1fcde7a8d835418c1a8b8d5ae441e055306a:/youtube_dl/extractor/brightcove.py diff --git a/youtube_dl/extractor/brightcove.py b/youtube_dl/extractor/brightcove.py index c947337..5c6e99d 100644 --- a/youtube_dl/extractor/brightcove.py +++ b/youtube_dl/extractor/brightcove.py @@ -1,4 +1,4 @@ -# encoding: utf-8 +# coding: utf-8 from __future__ import unicode_literals import re @@ -9,10 +9,10 @@ from ..compat import ( compat_etree_fromstring, compat_parse_qs, compat_str, - compat_urllib_parse, compat_urllib_parse_urlparse, compat_urlparse, compat_xml_parse_error, + compat_HTTPError, ) from ..utils import ( determine_ext, @@ -23,16 +23,18 @@ from ..utils import ( js_to_json, int_or_none, parse_iso8601, - sanitized_Request, unescapeHTML, unsmuggle_url, + update_url_query, + clean_html, + mimetype2ext, ) class BrightcoveLegacyIE(InfoExtractor): IE_NAME = 'brightcove:legacy' _VALID_URL = r'(?:https?://.*brightcove\.com/(services|viewer).*?\?|brightcove:)(?P.*)' - _FEDERATED_URL_TEMPLATE = 'http://c.brightcove.com/services/viewer/htmlFederated?%s' + _FEDERATED_URL = 'http://c.brightcove.com/services/viewer/htmlFederated' _TESTS = [ { @@ -46,6 +48,9 @@ class BrightcoveLegacyIE(InfoExtractor): 'title': 'Xavier Sala i Martín: “Un banc que no presta és un banc zombi que no serveix per a res”', 'uploader': '8TV', 'description': 'md5:a950cc4285c43e44d763d036710cd9cd', + 'timestamp': 1368213670, + 'upload_date': '20130510', + 'uploader_id': '1589608506001', } }, { @@ -57,6 +62,9 @@ class BrightcoveLegacyIE(InfoExtractor): 'title': 'JVMLS 2012: Arrays 2.0 - Opportunities and Challenges', 'description': 'John Rose speaks at the JVM Language Summit, August 1, 2012.', 'uploader': 'Oracle', + 'timestamp': 1344975024, + 'upload_date': '20120814', + 'uploader_id': '1460825906', }, }, { @@ -68,6 +76,9 @@ class BrightcoveLegacyIE(InfoExtractor): 'title': 'This Bracelet Acts as a Personal Thermostat', 'description': 'md5:547b78c64f4112766ccf4e151c20b6a0', 'uploader': 'Mashable', + 'timestamp': 1382041798, + 'upload_date': '20131017', + 'uploader_id': '1130468786001', }, }, { @@ -81,22 +92,26 @@ class BrightcoveLegacyIE(InfoExtractor): 'description': 'md5:363109c02998fee92ec02211bd8000df', 'uploader': 'National Ballet of Canada', }, + 'skip': 'Video gone', }, { # test flv videos served by akamaihd.net # From http://www.redbull.com/en/bike/stories/1331655643987/replay-uci-dh-world-cup-2014-from-fort-william - 'url': 'http://c.brightcove.com/services/viewer/htmlFederated?%40videoPlayer=ref%3ABC2996102916001&linkBaseURL=http%3A%2F%2Fwww.redbull.com%2Fen%2Fbike%2Fvideos%2F1331655630249%2Freplay-uci-fort-william-2014-dh&playerKey=AQ%7E%7E%2CAAAApYJ7UqE%7E%2Cxqr_zXk0I-zzNndy8NlHogrCb5QdyZRf&playerID=1398061561001#__youtubedl_smuggle=%7B%22Referer%22%3A+%22http%3A%2F%2Fwww.redbull.com%2Fen%2Fbike%2Fstories%2F1331655643987%2Freplay-uci-dh-world-cup-2014-from-fort-william%22%7D', + 'url': 'http://c.brightcove.com/services/viewer/htmlFederated?%40videoPlayer=ref%3Aevent-stream-356&linkBaseURL=http%3A%2F%2Fwww.redbull.com%2Fen%2Fbike%2Fvideos%2F1331655630249%2Freplay-uci-fort-william-2014-dh&playerKey=AQ%7E%7E%2CAAAApYJ7UqE%7E%2Cxqr_zXk0I-zzNndy8NlHogrCb5QdyZRf&playerID=1398061561001#__youtubedl_smuggle=%7B%22Referer%22%3A+%22http%3A%2F%2Fwww.redbull.com%2Fen%2Fbike%2Fstories%2F1331655643987%2Freplay-uci-dh-world-cup-2014-from-fort-william%22%7D', # The md5 checksum changes on each download 'info_dict': { - 'id': '2996102916001', + 'id': '3750436379001', 'ext': 'flv', 'title': 'UCI MTB World Cup 2014: Fort William, UK - Downhill Finals', - 'uploader': 'Red Bull TV', + 'uploader': 'RBTV Old (do not use)', 'description': 'UCI MTB World Cup 2014: Fort William, UK - Downhill Finals', + 'timestamp': 1409122195, + 'upload_date': '20140827', + 'uploader_id': '710858724001', }, }, { - # 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': { @@ -105,7 +120,22 @@ class BrightcoveLegacyIE(InfoExtractor): }, 'playlist_mincount': 7, }, + { + # playlist with 'playlistTab' (https://github.com/rg3/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, + }, ] + FLV_VCODECS = { + 1: 'SORENSON', + 2: 'ON2', + 3: 'H264', + 4: 'VP8', + } @classmethod def _build_brighcove_url(cls, object_str): @@ -136,17 +166,20 @@ class BrightcoveLegacyIE(InfoExtractor): else: flashvars = {} + data_url = object_doc.attrib.get('data', '') + data_url_params = compat_parse_qs(compat_urllib_parse_urlparse(data_url).query) + def find_param(name): if name in flashvars: return flashvars[name] node = find_xpath_attr(object_doc, './param', 'name', name) if node is not None: return node.attrib['value'] - return None + return data_url_params.get(name) 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 @@ -155,8 +188,8 @@ class BrightcoveLegacyIE(InfoExtractor): # Not all pages define this value if playerKey is not None: params['playerKey'] = playerKey - # The three fields hold the id of the video - videoPlayer = find_param('@videoPlayer') or find_param('videoId') or find_param('videoID') + # 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: params['@videoPlayer'] = videoPlayer linkBase = find_param('linkBaseURL') @@ -171,7 +204,7 @@ class BrightcoveLegacyIE(InfoExtractor): # // build Brightcove XML # } m = re.search( - r'''(?x)customBC.\createVideo\( + r'''(?x)customBC\.createVideo\( .*? # skipping width and height ["\'](?P\d+)["\']\s*,\s* # playerID ["\'](?PAQ[^"\']{48})[^"\']*["\']\s*,\s* # playerKey begins with AQ and is 50 characters @@ -184,8 +217,7 @@ class BrightcoveLegacyIE(InfoExtractor): @classmethod def _make_brightcove_url(cls, params): - data = compat_urllib_parse.urlencode(params) - return cls._FEDERATED_URL_TEMPLATE % data + return update_url_query(cls._FEDERATED_URL, params) @classmethod def _extract_brightcove_url(cls, webpage): @@ -200,13 +232,16 @@ class BrightcoveLegacyIE(InfoExtractor): """Return a list of all Brightcove URLs from the webpage """ url_m = re.search( - r']+ + content=([\'"])(?Phttps?://(?: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( @@ -227,7 +262,7 @@ class BrightcoveLegacyIE(InfoExtractor): 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) @@ -239,7 +274,7 @@ class BrightcoveLegacyIE(InfoExtractor): # We set the original url as the default 'Referer' header referer = smuggled_data.get('Referer', url) return self._get_video_info( - videoPlayer[0], query_str, query, referer=referer) + videoPlayer[0], query, referer=referer) elif 'playerKey' in query: player_key = query['playerKey'] return self._get_playlist_info(player_key[0]) @@ -248,15 +283,14 @@ class BrightcoveLegacyIE(InfoExtractor): 'Cannot find playerKey= variable. Did you forget quotes in a shell invocation?', expected=True) - def _get_video_info(self, video_id, query_str, query, referer=None): - request_url = self._FEDERATED_URL_TEMPLATE % query_str - req = sanitized_Request(request_url) + 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: - req.add_header('Referer', referer) - webpage = self._download_webpage(req, video_id) + headers['Referer'] = referer + webpage = self._download_webpage(self._FEDERATED_URL, video_id, headers=headers, query=query) error_msg = self._html_search_regex( r"

We're sorry.

([\s\n]*

.*?

)+", webpage, @@ -279,24 +313,35 @@ class BrightcoveLegacyIE(InfoExtractor): info_url, player_key, 'Downloading playlist information') json_data = json.loads(playlist_info) - if 'videoList' not in json_data: + if 'videoList' in json_data: + playlist_info = json_data['videoList'] + playlist_dto = playlist_info['mediaCollectionDTO'] + elif 'playlistTabs' in json_data: + playlist_info = json_data['playlistTabs'] + playlist_dto = playlist_info['lineupListDTO']['playlistDTOs'][0] + else: raise ExtractorError('Empty playlist') - playlist_info = json_data['videoList'] - videos = [self._extract_video_info(video_info) for video_info in playlist_info['mediaCollectionDTO']['videoDTOs']] + + videos = [self._extract_video_info(video_info) for video_info in playlist_dto['videoDTOs']] return self.playlist_result(videos, playlist_id='%s' % playlist_info['id'], - playlist_title=playlist_info['mediaCollectionDTO']['displayName']) + playlist_title=playlist_dto['displayName']) def _extract_video_info(self, video_info): + video_id = compat_str(video_info['id']) + publisher_id = video_info.get('publisherId') info = { - 'id': compat_str(video_info['id']), + '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') + renditions = video_info.get('renditions', []) + video_info.get('IOSRenditions', []) if renditions: formats = [] for rend in renditions: @@ -308,7 +353,8 @@ class BrightcoveLegacyIE(InfoExtractor): url_comp = compat_urllib_parse_urlparse(url) if url_comp.path.endswith('.m3u8'): formats.extend( - self._extract_m3u8_formats(url, info['id'], 'mp4')) + 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 @@ -317,19 +363,42 @@ class BrightcoveLegacyIE(InfoExtractor): ext = 'flv' if ext is None: ext = determine_ext(url) - size = rend.get('size') - formats.append({ + tbr = int_or_none(rend.get('encodingRate'), 1000) + a_format = { + 'format_id': 'http%s' % ('-%s' % tbr if tbr else ''), 'url': url, 'ext': ext, - 'height': rend.get('frameHeight'), - 'width': rend.get('frameWidth'), - 'filesize': size if size != 0 else None, - }) + 'filesize': int_or_none(rend.get('size')) or None, + 'tbr': tbr, + } + if rend.get('audioOnly'): + a_format.update({ + 'vcodec': 'none', + }) + 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): @@ -349,13 +418,13 @@ class BrightcoveLegacyIE(InfoExtractor): return ad_info if 'url' not in info and not info.get('formats'): - raise ExtractorError('Unable to extract video url for %s' % info['id']) + raise ExtractorError('Unable to extract video url for %s' % video_id) return info class BrightcoveNewIE(InfoExtractor): IE_NAME = 'brightcove:new' - _VALID_URL = r'https?://players\.brightcove\.net/(?P\d+)/(?P[^/]+)_(?P[^/]+)/index\.html\?.*videoId=(?P(?:ref:)?\d+)' + _VALID_URL = r'https?://players\.brightcove\.net/(?P\d+)/(?P[^/]+)_(?P[^/]+)/index\.html\?.*videoId=(?P\d+|ref:[^&]+)' _TESTS = [{ 'url': 'http://players.brightcove.net/929656772001/e41d32dc-ec74-459e-a845-6c69f7b724ea_default/index.html?videoId=4463358922001', 'md5': 'c8100925723840d4b0d243f7025703be', @@ -385,12 +454,21 @@ class BrightcoveNewIE(InfoExtractor): 'formats': 'mincount:41', }, 'params': { + # m3u8 download 'skip_download': 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, + }, { + # non numeric ref: prefixed video id + 'url': 'http://players.brightcove.net/710858724001/default_default/index.html?videoId=ref:event-stream-356', + 'only_matching': True, + }, { + # unavailable video without message but with error_code + 'url': 'http://players.brightcove.net/1305187701/c832abfb-641b-44eb-9da0-2fe76786505f_default/index.html?videoId=4377407326001', + 'only_matching': True, }] @staticmethod @@ -410,8 +488,8 @@ class BrightcoveNewIE(InfoExtractor): # Look for iframe embeds [1] for _, url in re.findall( - r']+src=(["\'])((?:https?:)//players\.brightcove\.net/\d+/[^/]+/index\.html.+?)\1', webpage): - entries.append(url) + r']+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( @@ -420,11 +498,11 @@ class BrightcoveNewIE(InfoExtractor): # According to [4] data-video-id may be prefixed with ref: r'''(?sx) ]+ - data-video-id=["\']((?:ref:)?\d+)["\'][^>]*>.*? + data-video-id=["\'](\d+|ref:[^"\']+)["\'][^>]*>.*? .*? ]+ src=["\'](?:https?:)?//players\.brightcove\.net/ - (\d+)/([\da-f-]+)_([^/]+)/index\.min\.js + (\d+)/([^/]+)_([^/]+)/index(?:\.min)?\.js ''', webpage): entries.append( 'http://players.brightcove.net/%s/%s_%s/index.html?videoId=%s' @@ -454,24 +532,36 @@ class BrightcoveNewIE(InfoExtractor): r'policyKey\s*:\s*(["\'])(?P.+?)\1', webpage, 'policy key', group='pk') - req = sanitized_Request( - 'https://edge.api.brightcove.com/playback/v1/accounts/%s/videos/%s' - % (account_id, video_id), - headers={'Accept': 'application/json;pk=%s' % policy_key}) - json_data = self._download_json(req, video_id) + 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 - title = json_data['name'] + title = json_data['name'].strip() formats = [] for source in json_data.get('sources', []): - source_type = source.get('type') + container = source.get('container') + ext = mimetype2ext(source.get('type')) src = source.get('src') - if source_type == 'application/x-mpegURL': + if ext == 'ism' or container == 'WVM': + continue + elif ext == 'm3u8' or container == 'M2TS': if not src: continue formats.extend(self._extract_m3u8_formats( - src, video_id, 'mp4', entry_protocol='m3u8_native', - m3u8_id='hls', fatal=False)) + src, video_id, 'mp4', 'm3u8_native', m3u8_id='hls', fatal=False)) + elif ext == 'mpd': + if not src: + continue + formats.extend(self._extract_mpd_formats(src, video_id, 'dash', fatal=False)) else: streaming_src = source.get('streaming_src') stream_name, app_name = source.get('stream_name'), source.get('app_name') @@ -479,15 +569,23 @@ class BrightcoveNewIE(InfoExtractor): continue tbr = float_or_none(source.get('avg_bitrate'), 1000) height = int_or_none(source.get('height')) + width = int_or_none(source.get('width')) f = { 'tbr': tbr, - 'width': int_or_none(source.get('width')), - 'height': height, 'filesize': int_or_none(source.get('size')), - 'container': source.get('container'), - 'vcodec': source.get('codec'), - 'ext': source.get('container').lower(), + 'container': container, + 'ext': ext or container.lower(), } + if width == 0 and height == 0: + f.update({ + 'vcodec': 'none', + }) + else: + f.update({ + 'width': width, + 'height': height, + 'vcodec': source.get('codec'), + }) def build_format_id(kind): format_id = kind @@ -501,7 +599,7 @@ class BrightcoveNewIE(InfoExtractor): f.update({ 'url': src or streaming_src, 'format_id': build_format_id('http' if src else 'http-streaming'), - 'preference': 2 if src else 1, + 'source_preference': 0 if src else -1, }) else: f.update({ @@ -510,22 +608,37 @@ class BrightcoveNewIE(InfoExtractor): 'format_id': build_format_id('rtmp'), }) formats.append(f) + + errors = json_data.get('errors') + if not formats and errors: + error = errors[0] + raise ExtractorError( + error.get('message') or error.get('error_subcode') or error['error_code'], expected=True) + self._sort_formats(formats) - description = json_data.get('description') - thumbnail = json_data.get('thumbnail') - timestamp = parse_iso8601(json_data.get('published_at')) + 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'], + }) + + is_live = False duration = float_or_none(json_data.get('duration'), 1000) - tags = json_data.get('tags', []) + if duration and duration < 0: + is_live = True return { 'id': video_id, - 'title': title, - 'description': description, - 'thumbnail': thumbnail, + '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': duration, - 'timestamp': timestamp, + 'timestamp': parse_iso8601(json_data.get('published_at')), 'uploader_id': account_id, 'formats': formats, - 'tags': tags, + 'subtitles': subtitles, + 'tags': json_data.get('tags', []), + 'is_live': is_live, }