X-Git-Url: https://git.rapsys.eu/youtubedl/blobdiff_plain/f46044c66663049e286c20ee015db99d47d9dd8a..dd50658396011c8e98b7201f035408b7959d56e1:/youtube_dl/extractor/brightcove.py?ds=sidebyside diff --git a/youtube_dl/extractor/brightcove.py b/youtube_dl/extractor/brightcove.py index 0733bec..465ae39 100644 --- a/youtube_dl/extractor/brightcove.py +++ b/youtube_dl/extractor/brightcove.py @@ -1,32 +1,44 @@ -# encoding: utf-8 +# coding: utf-8 from __future__ import unicode_literals -import re +import base64 import json -import xml.etree.ElementTree +import re +import struct from .common import InfoExtractor +from .adobepass import AdobePassIE from ..compat import ( + compat_etree_fromstring, compat_parse_qs, compat_str, - compat_urllib_parse, compat_urllib_parse_urlparse, - compat_urllib_request, compat_urlparse, + compat_xml_parse_error, + compat_HTTPError, ) from ..utils import ( determine_ext, ExtractorError, + extract_attributes, find_xpath_attr, fix_xml_ampersands, + float_or_none, + js_to_json, + int_or_none, + parse_iso8601, unescapeHTML, unsmuggle_url, + update_url_query, + clean_html, + mimetype2ext, ) -class BrightcoveIE(InfoExtractor): +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 = [ { @@ -40,6 +52,9 @@ class BrightcoveIE(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', } }, { @@ -51,6 +66,9 @@ class BrightcoveIE(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', }, }, { @@ -62,6 +80,9 @@ class BrightcoveIE(InfoExtractor): 'title': 'This Bracelet Acts as a Personal Thermostat', 'description': 'md5:547b78c64f4112766ccf4e151c20b6a0', 'uploader': 'Mashable', + 'timestamp': 1382041798, + 'upload_date': '20131017', + 'uploader_id': '1130468786001', }, }, { @@ -75,22 +96,27 @@ class BrightcoveIE(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', }, + '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': { @@ -99,7 +125,28 @@ class BrightcoveIE(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, + }, + { + # 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): @@ -117,7 +164,10 @@ class BrightcoveIE(InfoExtractor): object_str = re.sub(r'(]*)(xmlns=".*?")', r'\1', object_str) object_str = fix_xml_ampersands(object_str) - object_doc = xml.etree.ElementTree.fromstring(object_str.encode('utf-8')) + try: + object_doc = compat_etree_fromstring(object_str.encode('utf-8')) + except compat_xml_parse_error: + return fv_el = find_xpath_attr(object_doc, './param', 'name', 'flashVars') if fv_el is not None: @@ -127,17 +177,20 @@ class BrightcoveIE(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 @@ -146,15 +199,46 @@ class BrightcoveIE(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: + 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: params['linkBaseURL'] = linkBase - data = compat_urllib_parse.urlencode(params) - return cls._FEDERATED_URL_TEMPLATE % data + return cls._make_brightcove_url(params) + + @classmethod + def _build_brighcove_url_from_js(cls, object_js): + # The layout of JS is as follows: + # customBC.createVideo = function (width, height, playerID, playerKey, videoPlayer, VideoRandomID) { + # // build Brightcove XML + # } + m = re.search( + 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 + # in length, however it's appended to itself + # in places, so truncate + ["\'](?P\d+)["\'] # @videoPlayer + ''', object_js) + if m: + return cls._make_brightcove_url(m.groupdict()) + + @classmethod + def _make_brightcove_url(cls, params): + return update_url_query(cls._FEDERATED_URL, params) @classmethod def _extract_brightcove_url(cls, webpage): @@ -169,13 +253,16 @@ class BrightcoveIE(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( @@ -183,15 +270,24 @@ class BrightcoveIE(InfoExtractor): (?: [^>]+?class=[\'"][^>]*?BrightcoveExperience.*?[\'"] | [^>]*?>\s*''', + ).+?>\s*''', webpage) - return [cls._build_brighcove_url(m) for m in matches] + if matches: + return list(filter(None, [cls._build_brighcove_url(m) for m in matches])) + + matches = re.findall(r'(customBC\.createVideo\(.+?\);)', webpage) + if matches: + return list(filter(None, [ + cls._build_brighcove_url_from_js(custom_bc) + for custom_bc in matches])) + return [src for _, src in re.findall( + r']+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) @@ -202,8 +298,12 @@ class BrightcoveIE(InfoExtractor): if videoPlayer: # We set the original url as the default 'Referer' header referer = smuggled_data.get('Referer', url) + if 'playerID' not in query: + mobj = re.search(r'/bcpid(\d+)', url) + if mobj is not None: + query['playerID'] = [mobj.group(1)] 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]) @@ -212,20 +312,45 @@ class BrightcoveIE(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 = compat_urllib_request.Request(request_url) + def _brightcove_new_url_result(self, publisher_id, video_id): + brightcove_new_url = 'http://players.brightcove.net/%s/default_default/index.html?videoId=%s' % (publisher_id, video_id) + return self.url_result(brightcove_new_url, BrightcoveNewIE.ie_key(), video_id) + + 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, 'error message', default=None) if error_msg is not None: + 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: + player_id = query.get('playerID') + if player_id and player_id[0].isdigit(): + 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'Q', base64.urlsafe_b64decode(enc_pub_id))[0] + if publisher_id: + return self._brightcove_new_url_result(publisher_id, video_id) raise ExtractorError( 'brightcove said: %s' % error_msg, expected=True) @@ -243,24 +368,35 @@ class BrightcoveIE(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: @@ -272,7 +408,8 @@ class BrightcoveIE(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 @@ -281,19 +418,42 @@ class BrightcoveIE(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): @@ -312,6 +472,304 @@ class BrightcoveIE(InfoExtractor): else: return ad_info - if 'url' not in info and not info.get('formats'): - raise ExtractorError('Unable to extract video url for %s' % info['id']) + if not info.get('url') and not info.get('formats'): + uploader_id = info.get('uploader_id') + if uploader_id: + info.update(self._brightcove_new_url_result(uploader_id, video_id)) + else: + raise ExtractorError('Unable to extract video url for %s' % video_id) return info + + +class BrightcoveNewIE(AdobePassIE): + IE_NAME = 'brightcove:new' + _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', + 'info_dict': { + 'id': '4463358922001', + 'ext': 'mp4', + 'title': 'Meet the man behind Popcorn Time', + 'description': 'md5:eac376a4fe366edc70279bfb681aea16', + 'duration': 165.768, + 'timestamp': 1441391203, + 'upload_date': '20150904', + 'uploader_id': '929656772001', + 'formats': 'mincount:20', + }, + }, { + # with rtmp streams + 'url': 'http://players.brightcove.net/4036320279001/5d112ed9-283f-485f-a7f9-33f42e8bc042_default/index.html?videoId=4279049078001', + 'info_dict': { + 'id': '4279049078001', + 'ext': 'mp4', + 'title': 'Titansgrave: Chapter 0', + 'description': 'Titansgrave: Chapter 0', + 'duration': 1242.058, + 'timestamp': 1433556729, + 'upload_date': '20150606', + 'uploader_id': '4036320279001', + 'formats': 'mincount:39', + }, + '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 + def _extract_url(ie, webpage): + urls = BrightcoveNewIE._extract_urls(ie, webpage) + return urls[0] if urls else None + + @staticmethod + 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#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 = [] + + # Look for iframe embeds [1] + for _, url in re.findall( + r']+src=(["\'])((?:https?:)?//players\.brightcove\.net/\d+/[^/]+/index\.html.+?)\1', webpage): + entries.append(url if url.startswith('http') else 'http:' + url) + + # Look for