[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

@@ -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$seriesNameGlobal')
+        return metadata
+
     def _extract_aen_smil(self, smil_url, video_id, auth=None):
         query = {'mbr': 'true'}
         if auth:
@@ -63,7 +71,7 @@ class AENetworksIE(AENetworksBaseIE):
     IE_DESC = 'A+E Networks: A&E, Lifetime, History.com, FYI Network and History Vault'
     _VALID_URL = r'''(?x)
                     https?://
-                        (?:www\.)?
+                        (?:(?P<subdomain>www|play)\.)?
                         (?P<domain>
                             (?:history(?:vault)?|aetv|mylifetime|lifetimemovieclub)\.com|
                             fyi\.tv
@@ -75,11 +83,26 @@ class AENetworksIE(AENetworksBaseIE):
                             collections/[^/]+/(?P<collection_display_id>[^/]+)
                         )
                     '''
+    _GRAPHQL_QUERY = """
+fragment video on Video {
+  publicUrl
+}
+
+query getUserVideo($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',
+            'id': 'SERIES6093',
-            'title': 'Atlanta Plastic',
+            '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
@@ -137,55 +201,91 @@ class AENetworksIE(AENetworksBaseIE):
         'fyi.tv': 'FYI',
     }
 
-    def _real_extract(self, url):
+    def _extract_playlist(self, url, webpage, display_id, subdomain, url_parts):
-        domain, show_path, movie_display_id, special_display_id, collection_display_id = re.match(self._VALID_URL, url).groups()
+        # The "play" is pretty distinct from the normal sites, however, it contains all the data we need in a JSON blob.
-        display_id = show_path or movie_display_id or special_display_id or collection_display_id
+        if subdomain == 'play':
-        webpage = self._download_webpage(url, display_id, headers=self.geo_verification_headers())
+            series_id = self._search_regex(r'showid/(SERIES[0-9]+)', webpage, 'series id')
-        if show_path:
+            season_num = int_or_none(self._search_regex(r'/season-([0-9]+)/?', url, 'season number', fatal=False, default=None))
-            url_parts = show_path.split('/')
+            show_data = self._parse_json(
+                self._search_regex(r'(?s)<script[^>]+id="__NEXT_DATA__"[^>]*>(.+?)</script', webpage, 'show data'),
+                series_id, fatal=True)
+            if show_data:
+                apolloState = show_data.get('props', {}).get('apolloState', {})
+                entries = []
+                for key, episode in apolloState.items():
+                    if not key.startswith('Episode:') or series_id != episode.get('seriesId'):
+                        continue
+                    # If a season number was specified in the URL, filter out any episodes that don't match.
+                    if season_num and season_num != episode.get('tvSeasonNumber'):
+                        continue
+                    episode_url = compat_urlparse.urljoin(url, episode.get('canonical'))
+                    entries.append(self.url_result(episode_url, 'AENetworks', episode.get('id'), episode.get('title')))
+                series_name = apolloState.get('Series:%s' % series_id, {}).get('title')
+                return self.playlist_result(entries, series_id, series_name)
+        else:
+            series_title = self._html_search_meta('aetn:SeriesTitle', webpage)
             url_parts_len = len(url_parts)
             if url_parts_len == 1:
                 entries = []
-                for season_url_path in re.findall(r'(?s)<li[^>]+data-href="(/shows/%s/season-\d+)"' % url_parts[0], webpage):
+                for season_url_path in re.findall(r'(?s)<a[^>]+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),
+                        entries, self._html_search_meta('aetn:SeriesId', webpage), series_title)
-                        self._html_search_meta('aetn:SeriesTitle', webpage))
+                raise ExtractorError('Failed to extract seasons for show: %s' % url_parts[0])
-                else:
-                    # single season
-                    url_parts_len = 2
             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(
+                    video_id = episode_attributes.get('data-videoid') or episode_attributes.get('data-video-id')
-                        episode_url, 'AENetworks',
+                    episode_title = episode_attributes.get('aria-label')
-                        episode_attributes.get('data-videoid') or episode_attributes.get('data-video-id')))
+                    entries.append(self.url_result(episode_url, 'AENetworks', video_id, episode_title))
                 return self.playlist_result(
-                    entries, self._html_search_meta('aetn:SeasonId', webpage))
+                    entries, self._html_search_meta('aetn:SeasonId', webpage), series_title)
+        raise ExtractorError('Failed to extract playlist', video_id=display_id)
 
-        video_id = self._html_search_meta('aetn:VideoID', webpage)
+    def _real_extract(self, url):
-        media_url = self._search_regex(
+        subdomain, domain, show_path, movie_display_id, special_display_id, collection_display_id = re.match(self._VALID_URL, url).groups()
-            [r"media_url\s*=\s*'(?P<url>[^']+)'",
+        display_id = show_path or movie_display_id or special_display_id or collection_display_id
-             r'data-media-url=(?P<url>(?:https?:)?//[^\s>]+)',
+        webpage = self._download_webpage(url, display_id, headers=self.geo_verification_headers())
-             r'data-media-url=(["\'])(?P<url>(?:(?!\1).)+?)\1'],
+
-            webpage, 'video url', group='url')
+        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)
+        # 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