[ie/nfl] Fix extractors (#11409)

Authored by: bashonly

yt_dlp/extractor/anvato.py

@@ -33,24 +33,6 @@ class AnvatoIE(InfoExtractor):
     _AUTH_KEY = b'\x31\xc2\x42\x84\x9e\x73\xa0\xce'  # from anvplayer.min.js
 
     _TESTS = [{
-        # from https://www.nfl.com/videos/baker-mayfield-s-game-changing-plays-from-3-td-game-week-14
-        'url': 'anvato:GXvEgwyJeWem8KCYXfeoHWknwP48Mboj:899441',
-        'md5': '921919dab3cd0b849ff3d624831ae3e2',
-        'info_dict': {
-            'id': '899441',
-            'ext': 'mp4',
-            'title': 'Baker Mayfield\'s game-changing plays from 3-TD game Week 14',
-            'description': 'md5:85e05a3cc163f8c344340f220521136d',
-            'upload_date': '20201215',
-            'timestamp': 1608009755,
-            'thumbnail': r're:^https?://.*\.jpg',
-            'uploader': 'NFL',
-            'tags': ['Baltimore Ravens at Cleveland Browns (2020-REG-14)', 'Baker Mayfield', 'Game Highlights',
-                     'Player Highlights', 'Cleveland Browns', 'league'],
-            'duration': 157,
-            'categories': ['Entertainment', 'Game', 'Highlights'],
-        },
-    }, {
         # from https://ktla.com/news/99-year-old-woman-learns-to-fly-in-torrance-checks-off-bucket-list-dream/
         'url': 'anvato:X8POa4zpGZMmeiq0wqiO8IP5rMqQM9VN:8032455',
         'md5': '837718bcfb3a7778d022f857f7a9b19e',
@@ -241,31 +223,6 @@ class AnvatoIE(InfoExtractor):
         'telemundo': 'anvato_mcp_telemundo_web_prod_c5278d51ad46fda4b6ca3d0ea44a7846a054f582',
     }
 
-    def _generate_nfl_token(self, anvack, mcp_id):
-        reroute = self._download_json(
-            'https://api.nfl.com/v1/reroute', mcp_id, data=b'grant_type=client_credentials',
-            headers={'X-Domain-Id': 100}, note='Fetching token info')
-        token_type = reroute.get('token_type') or 'Bearer'
-        auth_token = f'{token_type} {reroute["access_token"]}'
-        response = self._download_json(
-            'https://api.nfl.com/v3/shield/', mcp_id, data=json.dumps({
-                'query': '''{
-  viewer {
-    mediaToken(anvack: "%s", id: %s) {
-      token
-    }
-  }
-}''' % (anvack, mcp_id),  # noqa: UP031
-            }).encode(), headers={
-                'Authorization': auth_token,
-                'Content-Type': 'application/json',
-            }, note='Fetching NFL API token')
-        return traverse_obj(response, ('data', 'viewer', 'mediaToken', 'token'))
-
-    _TOKEN_GENERATORS = {
-        'GXvEgwyJeWem8KCYXfeoHWknwP48Mboj': _generate_nfl_token,
-    }
-
     def _server_time(self, access_key, video_id):
         return int_or_none(traverse_obj(self._download_json(
             f'{self._API_BASE_URL}/server_time', video_id, query={'anvack': access_key},
@@ -290,8 +247,6 @@ def _get_video_json(self, access_key, video_id, extracted_token):
         }
         if extracted_token is not None:
             api['anvstk2'] = extracted_token
-        elif self._TOKEN_GENERATORS.get(access_key) is not None:
-            api['anvstk2'] = self._TOKEN_GENERATORS[access_key](self, access_key, video_id)
         elif self._ANVACK_TABLE.get(access_key) is not None:
             api['anvstk'] = md5_text(f'{access_key}|{anvrid}|{server_time}|{self._ANVACK_TABLE[access_key]}')
         else:

yt_dlp/extractor/nfl.py

@@ -11,9 +11,12 @@
     clean_html,
     determine_ext,
     get_element_by_class,
-    traverse_obj,
+    int_or_none,
+    make_archive_id,
+    url_or_none,
     urlencode_postdata,
 )
+from ..utils.traversal import traverse_obj
 
 
 class NFLBaseIE(InfoExtractor):
@@ -75,22 +78,15 @@ class NFLBaseIE(InfoExtractor):
             'osVersion': '10.0',
         }, separators=(',', ':')).encode()).decode(),
         'networkType': 'other',
-        'nflClaimGroupsToAdd': [],
+        'peacockUUID': 'undefined',
-        'nflClaimGroupsToRemove': [],
     }
     _ACCOUNT_INFO = {}
-    _API_KEY = None
+    _API_KEY = '3_Qa8TkWpIB8ESCBT8tY2TukbVKgO5F6BJVc7N1oComdwFzI7H2L9NOWdm11i_BY9f'
 
     _TOKEN = None
     _TOKEN_EXPIRY = 0
 
-    def _get_account_info(self, url, slug):
+    def _get_account_info(self):
-        if not self._API_KEY:
-            webpage = self._download_webpage(url, slug, fatal=False) or ''
-            self._API_KEY = self._search_regex(
-                r'window\.gigyaApiKey\s*=\s*["\'](\w+)["\'];', webpage, 'API key',
-                fatal=False) or '3_Qa8TkWpIB8ESCBT8tY2TukbVKgO5F6BJVc7N1oComdwFzI7H2L9NOWdm11i_BY9f'
-
         cookies = self._get_cookies('https://auth-id.nfl.com/')
         login_token = traverse_obj(cookies, (
             (f'glt_{self._API_KEY}', lambda k, _: k.startswith('glt_')), {lambda x: x.value}), get_all=False)
@@ -103,7 +99,7 @@ def _get_account_info(self, url, slug):
                 'or else try using --cookies-from-browser instead', expected=True)
 
         account = self._download_json(
-            'https://auth-id.nfl.com/accounts.getAccountInfo', slug,
+            'https://auth-id.nfl.com/accounts.getAccountInfo', None,
             note='Downloading account info', data=urlencode_postdata({
                 'include': 'profile,data',
                 'lang': 'en',
@@ -111,7 +107,7 @@ def _get_account_info(self, url, slug):
                 'sdk': 'js_latest',
                 'login_token': login_token,
                 'authMode': 'cookie',
-                'pageURL': url,
+                'pageURL': 'https://www.nfl.com/',
                 'sdkBuild': traverse_obj(cookies, (
                     'gig_canary_ver', {lambda x: x.value.partition('-')[0]}), default='15170'),
                 'format': 'json',
@@ -126,53 +122,76 @@ def _get_account_info(self, url, slug):
         if len(self._ACCOUNT_INFO) != 3:
             raise ExtractorError('Failed to retrieve account info with provided cookies', expected=True)
 
-    def _get_auth_token(self, url, slug):
+    def _get_auth_token(self):
         if self._TOKEN and self._TOKEN_EXPIRY > int(time.time() + 30):
             return
 
-        if not self._ACCOUNT_INFO:
-            self._get_account_info(url, slug)
-
         token = self._download_json(
             'https://api.nfl.com/identity/v3/token%s' % (
                 '/refresh' if self._ACCOUNT_INFO.get('refreshToken') else ''),
-            slug, headers={'Content-Type': 'application/json'}, note='Downloading access token',
+            None, headers={'Content-Type': 'application/json'}, note='Downloading access token',
             data=json.dumps({**self._CLIENT_DATA, **self._ACCOUNT_INFO}, separators=(',', ':')).encode())
 
         self._TOKEN = token['accessToken']
         self._TOKEN_EXPIRY = token['expiresIn']
         self._ACCOUNT_INFO['refreshToken'] = token['refreshToken']
 
+    def _extract_video(self, mcp_id, is_live=False):
+        self._get_auth_token()
+        data = self._download_json(
+            f'https://api.nfl.com/play/v1/asset/{mcp_id}', mcp_id, headers={
+                'Authorization': f'Bearer {self._TOKEN}',
+                'Accept': 'application/json',
+                'Content-Type': 'application/json',
+            }, data=json.dumps({'init': True, 'live': is_live}, separators=(',', ':')).encode())
+        formats, subtitles = self._extract_m3u8_formats_and_subtitles(
+            data['accessUrl'], mcp_id, 'mp4', m3u8_id='hls')
+
+        return {
+            'id': mcp_id,
+            'formats': formats,
+            'subtitles': subtitles,
+            'is_live': is_live,
+            '_old_archive_ids': [make_archive_id(AnvatoIE, mcp_id)],
+            **traverse_obj(data, ('metadata', {
+                'title': ('event', ('def_title', 'friendlyName'), {str}, any),
+                'description': ('event', 'def_description', {str}),
+                'duration': ('event', 'duration', {int_or_none}),
+                'thumbnails': ('thumbnails', ..., 'url', {'url': {url_or_none}}),
+            })),
+        }
+
     def _parse_video_config(self, video_config, display_id):
         video_config = self._parse_json(video_config, display_id)
+        is_live = traverse_obj(video_config, ('live', {bool})) or False
         item = video_config['playlist'][0]
-        mcp_id = item.get('mcpID')
+        if mcp_id := item.get('mcpID'):
-        if mcp_id:
+            return self._extract_video(mcp_id, is_live=is_live)
-            info = self.url_result(f'{self._ANVATO_PREFIX}{mcp_id}', AnvatoIE, mcp_id)
+
-        else:
+        info = {'id': item.get('id') or item['entityId']}
-            media_id = item.get('id') or item['entityId']
+
-            title = item.get('title')
         item_url = item['url']
-            info = {'id': media_id}
         ext = determine_ext(item_url)
         if ext == 'm3u8':
-                info['formats'] = self._extract_m3u8_formats(item_url, media_id, 'mp4')
+            info['formats'] = self._extract_m3u8_formats(item_url, info['id'], 'mp4')
         else:
             info['url'] = item_url
             if item.get('audio') is True:
                 info['vcodec'] = 'none'
-            is_live = video_config.get('live') is True
+
         thumbnails = None
-            image_url = item.get(item.get('imageSrc')) or item.get(item.get('posterImage'))
+        if image_url := traverse_obj(item, 'imageSrc', 'posterImage', expected_type=url_or_none):
-            if image_url:
             thumbnails = [{
                 'url': image_url,
                 'ext': determine_ext(image_url, 'jpg'),
             }]
+
         info.update({
-                'title': title,
+            **traverse_obj(item, {
+                'title': ('title', {str}),
+                'description': ('description', {clean_html}),
+            }),
             'is_live': is_live,
-                'description': clean_html(item.get('description')),
             'thumbnails': thumbnails,
         })
         return info
@@ -188,24 +207,20 @@ class NFLIE(NFLBaseIE):
             'ext': 'mp4',
             'title': "Baker Mayfield's game-changing plays from 3-TD game Week 14",
             'description': 'md5:85e05a3cc163f8c344340f220521136d',
-            'upload_date': '20201215',
+            'thumbnail': r're:https?://.+\.jpg',
-            'timestamp': 1608009755,
-            'thumbnail': r're:^https?://.*\.jpg$',
-            'uploader': 'NFL',
-            'tags': 'count:6',
             'duration': 157,
-            'categories': 'count:3',
+            '_old_archive_ids': ['anvato 899441'],
         },
     }, {
         'url': 'https://www.chiefs.com/listen/patrick-mahomes-travis-kelce-react-to-win-over-dolphins-the-breakdown',
-        'md5': '6886b32c24b463038c760ceb55a34566',
+        'md5': '92a517f05bd3eb50fe50244bc621aec8',
         'info_dict': {
-            'id': 'd87e8790-3e14-11eb-8ceb-ff05c2867f99',
+            'id': '8b7c3625-a461-4751-8db4-85f536f2bbd0',
             'ext': 'mp3',
             'title': 'Patrick Mahomes, Travis Kelce React to Win Over Dolphins | The Breakdown',
             'description': 'md5:12ada8ee70e6762658c30e223e095075',
+            'thumbnail': 'https://static.clubs.nfl.com/image/private/t_editorial_landscape_12_desktop/v1571153441/chiefs/rfljejccnyhhkpkfq855',
         },
-        'skip': 'HTTP Error 404: Not Found',
     }, {
         'url': 'https://www.buffalobills.com/video/buffalo-bills-military-recognition-week-14',
         'only_matching': True,
@@ -236,13 +251,16 @@ class NFLArticleIE(NFLBaseIE):
     def _real_extract(self, url):
         display_id = self._match_id(url)
         webpage = self._download_webpage(url, display_id)
-        entries = []
+
+        def entries():
             for video_config in re.findall(self._VIDEO_CONFIG_REGEX, webpage):
-            entries.append(self._parse_video_config(video_config, display_id))
+                yield self._parse_video_config(video_config, display_id)
+
         title = clean_html(get_element_by_class(
             'nfl-c-article__title', webpage)) or self._html_search_meta(
             ['og:title', 'twitter:title'], webpage)
-        return self.playlist_result(entries, display_id, title)
+
+        return self.playlist_result(entries(), display_id, title)
 
 
 class NFLPlusReplayIE(NFLBaseIE):
@@ -307,6 +325,9 @@ class NFLPlusReplayIE(NFLBaseIE):
         'all_22': 'All-22',
     }
 
+    def _real_initialize(self):
+        self._get_account_info()
+
     def _real_extract(self, url):
         slug, video_id = self._match_valid_url(url).group('slug', 'id')
         requested_types = self._configuration_arg('type', ['all'])
@@ -315,7 +336,7 @@ def _real_extract(self, url):
         requested_types = traverse_obj(self._REPLAY_TYPES, (None, requested_types))
 
         if not video_id:
-            self._get_auth_token(url, slug)
+            self._get_auth_token()
             headers = {'Authorization': f'Bearer {self._TOKEN}'}
             game_id = self._download_json(
                 f'https://api.nfl.com/football/v2/games/externalId/slug/{slug}', slug,
@@ -328,14 +349,13 @@ def _real_extract(self, url):
                     'items', lambda _, v: v['subType'] == requested_types[0], 'mcpPlaybackId'), get_all=False)
 
         if video_id:
-            return self.url_result(f'{self._ANVATO_PREFIX}{video_id}', AnvatoIE, video_id)
+            return self._extract_video(video_id)
 
         def entries():
             for replay in traverse_obj(
                 replays, ('items', lambda _, v: v['mcpPlaybackId'] and v['subType'] in requested_types),
             ):
-                video_id = replay['mcpPlaybackId']
+                yield self._extract_video(replay['mcpPlaybackId'])
-                yield self.url_result(f'{self._ANVATO_PREFIX}{video_id}', AnvatoIE, video_id)
 
         return self.playlist_result(entries(), slug)
 
@@ -362,12 +382,15 @@ class NFLPlusEpisodeIE(NFLBaseIE):
         'params': {'skip_download': 'm3u8'},
     }]
 
+    def _real_initialize(self):
+        self._get_account_info()
+
     def _real_extract(self, url):
         slug = self._match_id(url)
-        self._get_auth_token(url, slug)
+        self._get_auth_token()
         video_id = self._download_json(
             f'https://api.nfl.com/content/v1/videos/episodes/{slug}', slug, headers={
                 'Authorization': f'Bearer {self._TOKEN}',
             })['mcpPlaybackId']
 
-        return self.url_result(f'{self._ANVATO_PREFIX}{video_id}', AnvatoIE, video_id)
+        return self._extract_video(video_id)