From 3bf38e47aea11a9866d9676a66298458a22c0b6e Mon Sep 17 00:00:00 2001 From: Kevin O'Connor Date: Sat, 3 Oct 2020 17:23:08 -0400 Subject: [PATCH] [AENetworks] Update extractor for the current site(s) This gets the AENetworks extractor back into a working state. The highlights are: * Support for the `play` subdomain that the content is now hosted on * Extract ThePlatform URLs from the GraphQL endpoint they now use * Properly extract video information from the updates made to the non-play subdomain sites. * Properly extract the series name, season number, and episode numbers * Added more explicit error messages for when extraction fails Fixes #23363 --- youtube_dl/extractor/aenetworks.py | 162 +++++++++++++++++++++++------ 1 file changed, 131 insertions(+), 31 deletions(-) diff --git a/youtube_dl/extractor/aenetworks.py b/youtube_dl/extractor/aenetworks.py index 611b948f510..3a70d47aabb 100644 --- a/youtube_dl/extractor/aenetworks.py +++ b/youtube_dl/extractor/aenetworks.py @@ -1,6 +1,7 @@ # coding: utf-8 from __future__ import unicode_literals +import json import re from .theplatform import ThePlatformIE @@ -20,6 +21,13 @@ class AENetworksBaseIE(ThePlatformIE): _THEPLATFORM_KEY = 'crazyjava' _THEPLATFORM_SECRET = 's3cr3t' + def _parse_theplatform_metadata(self, info): + metadata = super()._parse_theplatform_metadata(info) + metadata['season_number'] = int(info.get('AETN$season')) + metadata['episode_number'] = int(info.get('AETN$episode')) + metadata['series'] = info.get('AETN$seriesNameGlobal') + return metadata + def _extract_aen_smil(self, smil_url, video_id, auth=None): query = {'mbr': 'true'} if auth: @@ -63,7 +71,7 @@ class AENetworksIE(AENetworksBaseIE): IE_DESC = 'A+E Networks: A&E, Lifetime, History.com, FYI Network and History Vault' _VALID_URL = r'''(?x) https?:// - (?:www\.)? + (?:(?Pwww|play)\.)? (?P (?:history(?:vault)?|aetv|mylifetime|lifetimemovieclub)\.com| fyi\.tv @@ -75,11 +83,26 @@ class AENetworksIE(AENetworksBaseIE): collections/[^/]+/(?P[^/]+) ) ''' + _GRAPHQL_QUERY = """ +fragment video on Video { + publicUrl +} + +query getUserVideo($videoId: ID!) { + video(id: $videoId) { + ...video + } +} +""" + _TESTS = [{ 'url': 'http://www.history.com/shows/mountain-men/season-1/episode-1', 'info_dict': { 'id': '22253814', 'ext': 'mp4', + 'series': 'Mountain Men', + 'season_number': 1, + 'episode_number': 1, 'title': 'Winter is Coming', 'description': 'md5:641f424b7a19d8e24f26dea22cf59d74', 'timestamp': 1338306241, @@ -91,19 +114,60 @@ class AENetworksIE(AENetworksBaseIE): 'skip_download': True, }, 'add_ie': ['ThePlatform'], + }, { + 'url': 'http://play.history.com/shows/mountain-men/season-1/episode-1', + 'info_dict': { + 'id': '22253814', + 'ext': 'mp4', + 'series': 'Mountain Men', + 'season_number': 1, + 'episode_number': 1, + 'title': 'Winter Is Coming', + 'description': 'md5:a40e370925074260b1c8a633c632c63a', + 'timestamp': 1338306241, + 'upload_date': '20120529', + 'uploader': 'AENE-NEW', + }, + 'params': { + # m3u8 download + 'skip_download': True, + }, + 'add_ie': ['ThePlatform'], }, { 'url': 'http://www.history.com/shows/ancient-aliens/season-1', 'info_dict': { 'id': '71889446852', + 'title': 'Ancient Aliens' }, 'playlist_mincount': 5, }, { - 'url': 'http://www.mylifetime.com/shows/atlanta-plastic', + 'url': 'http://www.mylifetime.com/shows/marrying-millions', 'info_dict': { - 'id': 'SERIES4317', - 'title': 'Atlanta Plastic', + 'id': 'SERIES6093', + 'title': 'Marrying Millions', }, - 'playlist_mincount': 2, + 'playlist_mincount': 1, + }, { + 'url': 'http://www.mylifetime.com/shows/marrying-millions/season-1', + 'info_dict': { + 'id': '269343782619', + 'title': 'Marrying Millions', + }, + 'playlist_mincount': 10, + }, { + 'url': 'https://play.mylifetime.com/shows/marrying-millions/season-1', + 'info_dict': { + 'id': 'SERIES6093', + 'title': 'Marrying Millions', + }, + 'playlist_mincount': 10, + }, { + 'url': 'https://play.mylifetime.com/shows/marrying-millions', + 'info_dict': { + 'id': 'SERIES6093', + 'title': 'Marrying Millions', + }, + 'playlist_mincount': 11, }, { 'url': 'http://www.aetv.com/shows/duck-dynasty/season-9/episode-1', 'only_matching': True @@ -137,56 +201,92 @@ class AENetworksIE(AENetworksBaseIE): 'fyi.tv': 'FYI', } - def _real_extract(self, url): - domain, show_path, movie_display_id, special_display_id, collection_display_id = re.match(self._VALID_URL, url).groups() - display_id = show_path or movie_display_id or special_display_id or collection_display_id - webpage = self._download_webpage(url, display_id, headers=self.geo_verification_headers()) - if show_path: - url_parts = show_path.split('/') + def _extract_playlist(self, url, webpage, display_id, subdomain, url_parts): + # The "play" is pretty distinct from the normal sites, however, it contains all the data we need in a JSON blob. + if subdomain == 'play': + series_id = self._search_regex(r'showid/(SERIES[0-9]+)', webpage, 'series id') + season_num = int_or_none(self._search_regex(r'/season-([0-9]+)/?', url, 'season number', fatal=False, default=None)) + show_data = self._parse_json( + self._search_regex(r'(?s)]+id="__NEXT_DATA__"[^>]*>(.+?)]+data-href="(/shows/%s/season-\d+)"' % url_parts[0], webpage): + for season_url_path in re.findall(r'(?s)]+href="(/shows/%s/season-\d+)"' % url_parts[0], webpage): entries.append(self.url_result( compat_urlparse.urljoin(url, season_url_path), 'AENetworks')) if entries: return self.playlist_result( - entries, self._html_search_meta('aetn:SeriesId', webpage), - self._html_search_meta('aetn:SeriesTitle', webpage)) - else: - # single season - url_parts_len = 2 + entries, self._html_search_meta('aetn:SeriesId', webpage), series_title) + raise ExtractorError('Failed to extract seasons for show: %s' % url_parts[0]) if url_parts_len == 2: entries = [] - for episode_item in re.findall(r'(?s)<[^>]+class="[^"]*(?:episode|program)-item[^"]*"[^>]*>', webpage): + for episode_item in re.findall(r'(?s)<[^>]+data-episodetype[^>]*>', webpage): episode_attributes = extract_attributes(episode_item) episode_url = compat_urlparse.urljoin( url, episode_attributes['data-canonical']) - entries.append(self.url_result( - episode_url, 'AENetworks', - episode_attributes.get('data-videoid') or episode_attributes.get('data-video-id'))) + video_id = episode_attributes.get('data-videoid') or episode_attributes.get('data-video-id') + episode_title = episode_attributes.get('aria-label') + entries.append(self.url_result(episode_url, 'AENetworks', video_id, episode_title)) return self.playlist_result( - entries, self._html_search_meta('aetn:SeasonId', webpage)) - - video_id = self._html_search_meta('aetn:VideoID', webpage) - media_url = self._search_regex( - [r"media_url\s*=\s*'(?P[^']+)'", - r'data-media-url=(?P(?:https?:)?//[^\s>]+)', - r'data-media-url=(["\'])(?P(?:(?!\1).)+?)\1'], - webpage, 'video url', group='url') + entries, self._html_search_meta('aetn:SeasonId', webpage), series_title) + raise ExtractorError('Failed to extract playlist', video_id=display_id) + + def _real_extract(self, url): + subdomain, domain, show_path, movie_display_id, special_display_id, collection_display_id = re.match(self._VALID_URL, url).groups() + display_id = show_path or movie_display_id or special_display_id or collection_display_id + webpage = self._download_webpage(url, display_id, headers=self.geo_verification_headers()) + + if show_path: + url_parts = show_path.split('/') + # If there's only the show name and/or season number then we'll need to extract a playlist. + if len(url_parts) < 3: + return self._extract_playlist(url, webpage, display_id, subdomain, url_parts) + + requestor_id = self._DOMAIN_TO_REQUESTOR_ID[domain] + video_id = self._html_search_meta(['videoId', 'aetn:VideoID'], webpage) + # Make a GraphQL query to get the episode URL as they no longer directly embed it in the response webpage. + video_data = self._download_json( + 'https://yoga.appsvcs.aetnd.com/graphql?brand=%s&mode=live&platform=web' % (requestor_id.lower()), video_id, + data=json.dumps( + { + 'operationName': 'getUserVideo', + 'variables': {'videoId': video_id}, + 'query': self._GRAPHQL_QUERY, + }).encode('utf-8'), + headers={'Content-Type': 'application/json'}) + media_url = video_data.get('data', {}).get('video', {}).get('publicUrl') + if not media_url: + raise ExtractorError('Failed to extract media URL', video_id=video_id) theplatform_metadata = self._download_theplatform_metadata(self._search_regex( r'https?://link\.theplatform\.com/s/([^?]+)', media_url, 'theplatform_path'), video_id) info = self._parse_theplatform_metadata(theplatform_metadata) auth = None if theplatform_metadata.get('AETN$isBehindWall'): - requestor_id = self._DOMAIN_TO_REQUESTOR_ID[domain] resource = self._get_mvpd_resource( requestor_id, theplatform_metadata['title'], theplatform_metadata.get('AETN$PPL_pplProgramId') or theplatform_metadata.get('AETN$PPL_pplProgramId_OLD'), theplatform_metadata['ratings'][0]['rating']) auth = self._extract_mvpd_auth( url, video_id, requestor_id, resource) - info.update(self._search_json_ld(webpage, video_id, fatal=False)) + # JSON-LD data isn't present on the play subdomain webpages. + if subdomain != 'play': + info.update(self._search_json_ld(webpage, video_id, fatal=False)) info.update(self._extract_aen_smil(media_url, video_id, auth)) return info