Merge 0724bfea8b into 5904853ae5

Closes #7442
Authored by: bashonly

yt_dlp/extractor/cbc.py

@@ -151,7 +151,7 @@ def _real_extract(self, url):
 
 class CBCPlayerIE(InfoExtractor):
     IE_NAME = 'cbc.ca:player'
-    _VALID_URL = r'(?:cbcplayer:|https?://(?:www\.)?cbc\.ca/(?:player/play/|i/caffeine/syndicate/\?mediaId=))(?P<id>(?:\d\.)?\d+)'
+    _VALID_URL = r'(?:cbcplayer:|https?://(?:www\.)?cbc\.ca/(?:player/play/(?:video/)?|i/caffeine/syndicate/\?mediaId=))(?P<id>(?:\d\.)?\d+)'
     _TESTS = [{
         'url': 'http://www.cbc.ca/player/play/2683190193',
         'md5': '64d25f841ddf4ddb28a235338af32e2c',
@@ -277,6 +277,28 @@ class CBCPlayerIE(InfoExtractor):
             'location': 'Canada',
             'media_type': 'Full Program',
         },
+    }, {
+        'url': 'https://www.cbc.ca/player/play/video/1.7194274',
+        'md5': '188b96cf6bdcb2540e178a6caa957128',
+        'info_dict': {
+            'id': '2334524995812',
+            'ext': 'mp4',
+            'title': '#TheMoment a rare white spirit moose was spotted in Alberta',
+            'description': 'md5:18ae269a2d0265c5b0bbe4b2e1ac61a3',
+            'timestamp': 1714788791,
+            'duration': 77.678,
+            'subtitles': {'eng': [{'ext': 'vtt', 'protocol': 'm3u8_native'}]},
+            'thumbnail': 'https://thumbnails.cbc.ca/maven_legacy/thumbnails/201/543/THE_MOMENT.jpg',
+            'uploader': 'CBCC-NEW',
+            'chapters': 'count:0',
+            'upload_date': '20240504',
+            'categories': 'count:3',
+            'series': 'The National',
+            'tags': 'count:15',
+            'creators': ['encoder'],
+            'location': 'Canada',
+            'media_type': 'Excerpt',
+        },
     }, {
         'url': 'cbcplayer:1.7159484',
         'only_matching': True,

yt_dlp/extractor/crunchyroll.py

@@ -53,15 +53,19 @@ def _set_auth_info(self, response):
         CrunchyrollBaseIE._AUTH_EXPIRY = time_seconds(seconds=traverse_obj(response, ('expires_in', {float_or_none}), default=300) - 10)
 
     def _request_token(self, headers, data, note='Requesting token', errnote='Failed to request token'):
-        try:  # TODO: Add impersonation support here
+        try:
             return self._download_json(
                 f'{self._BASE_URL}/auth/v1/token', None, note=note, errnote=errnote,
-                headers=headers, data=urlencode_postdata(data))
+                headers=headers, data=urlencode_postdata(data), impersonate=True)
         except ExtractorError as error:
             if not isinstance(error.cause, HTTPError) or error.cause.status != 403:
                 raise
+            if target := error.cause.response.extensions.get('impersonate'):
+                raise ExtractorError(f'Got HTTP Error 403 when using impersonate target "{target}"')
             raise ExtractorError(
-                'Request blocked by Cloudflare; navigate to Crunchyroll in your browser, '
+                'Request blocked by Cloudflare. '
+                'Install the required impersonation dependency if possible, '
+                'or else navigate to Crunchyroll in your browser, '
                 'then pass the fresh cookies (with --cookies-from-browser or --cookies) '
                 'and your browser\'s User-Agent (with --user-agent)', expected=True)
 

yt_dlp/extractor/nfb.py

@@ -5,7 +5,6 @@
     merge_dicts,
     parse_count,
     url_or_none,
-    urljoin,
 )
 from ..utils.traversal import traverse_obj
 
@@ -16,8 +15,8 @@ class NFBBaseIE(InfoExtractor):
 
     def _extract_ep_data(self, webpage, video_id, fatal=False):
         return self._search_json(
-            r'const\s+episodesData\s*=', webpage, 'episode data', video_id,
-            contains_pattern=r'\[\s*{(?s:.+)}\s*\]', fatal=fatal) or []
+            r'episodesData\s*:', webpage, 'episode data', video_id,
+            contains_pattern=r'\s*{(?s:.+)}\s*', fatal=fatal) or []
 
     def _extract_ep_info(self, data, video_id, slug=None):
         info = traverse_obj(data, (lambda _, v: video_id in v['embed_url'], {
@@ -224,15 +223,15 @@ def _real_extract(self, url):
         # type_ can change from film to serie(s) after redirect; new slug may have episode number
         type_, slug = self._match_valid_url(urlh.url).group('type', 'id')
 
-        embed_url = urljoin(f'https://www.{site}.ca', self._html_search_regex(
-            r'<[^>]+\bid=["\']player-iframe["\'][^>]*\bsrc=["\']([^"\']+)', webpage, 'embed url'))
+        embed_url = self._html_search_regex(
+            r'<script.*<iframe\s+src=[\\"\']+(https?:[^\\"\']+)', webpage, 'embed url')
         video_id = self._match_id(embed_url)  # embed url has unique slug
         player = self._download_webpage(embed_url, video_id, 'Downloading player page')
         if 'MESSAGE_GEOBLOCKED' in player:
             self.raise_geo_restricted(countries=self._GEO_COUNTRIES)
 
         formats, subtitles = self._extract_m3u8_formats_and_subtitles(
-            self._html_search_regex(r'source:\s*\'([^\']+)', player, 'm3u8 url'),
+            self._html_search_regex(r'["\']?source["\']?:\s*["\']([^"\']+\.m3u8)', player, 'm3u8 url'),
             video_id, 'mp4', m3u8_id='hls')
 
         if dv_source := self._html_search_regex(r'dvSource:\s*\'([^\']+)', player, 'dv', default=None):
@@ -246,7 +245,7 @@ def _real_extract(self, url):
         info = {
             'id': video_id,
             'title': self._html_search_regex(
-                r'<[^>]+\bid=["\']titleHeader["\'][^>]*>\s*<h1[^>]*>\s*([^<]+?)\s*</h1>',
+                r'["\']nfb_version_title["\']\s*:\s*["\']([^"\']+)',
                 webpage, 'title', default=None),
             'description': self._html_search_regex(
                 r'<[^>]+\bid=["\']tabSynopsis["\'][^>]*>\s*<p[^>]*>\s*([^<]+)',
@@ -254,9 +253,9 @@ def _real_extract(self, url):
             'thumbnail': self._html_search_regex(
                 r'poster:\s*\'([^\']+)', player, 'thumbnail', default=None),
             'uploader': self._html_search_regex(
-                r'<[^>]+\bitemprop=["\']name["\'][^>]*>([^<]+)', webpage, 'uploader', default=None),
+                r'<[^>]+\bitemprop=["\']director["\'][^>]*>([^<]+)', webpage, 'uploader', default=None),
             'release_year': int_or_none(self._html_search_regex(
-                r'<[^>]+\bitemprop=["\']datePublished["\'][^>]*>([^<]+)',
+                r'["\']nfb_version_year["\']\s*:\s*["\']([^"\']+)',
                 webpage, 'release_year', default=None)),
         } if type_ == 'film' else self._extract_ep_info(self._extract_ep_data(webpage, video_id, slug), video_id)
 

yt_dlp/extractor/youporn.py

@@ -72,15 +72,15 @@ class YouPornIE(InfoExtractor):
             'id': '16290308',
             'age_limit': 18,
             'categories': [],
-            'description': 'md5:00ea70f642f431c379763c17c2f396bc',
+            'description': str,  # TODO: detect/remove SEO spam description in ytdl backport
             'display_id': 'tinderspecial-trailer1',
             'duration': 298.0,
             'ext': 'mp4',
             'upload_date': '20201123',
             'uploader': 'Ersties',
             'tags': [],
-            'thumbnail': 'https://fi1.ypncdn.com/202011/23/16290308/original/8/tinderspecial-trailer1-8(m=eaAaaEPbaaaa).jpg',
-            'timestamp': 1606089600,
+            'thumbnail': r're:https://.+\.jpg',
+            'timestamp': 1606147564,
             'title': 'Tinder In Real Life',
             'view_count': int,
         }
@@ -88,11 +88,17 @@ class YouPornIE(InfoExtractor):
 
     def _real_extract(self, url):
         video_id, display_id = self._match_valid_url(url).group('id', 'display_id')
-        definitions = self._download_json(
-            f'https://www.youporn.com/api/video/media_definitions/{video_id}/', display_id or video_id)
+        self._set_cookie('.youporn.com', 'age_verified', '1')
+        webpage = self._download_webpage(f'https://www.youporn.com/watch/{video_id}', video_id)
+        definitions = self._search_json(r'\bplayervars\s*:', webpage, 'player vars', video_id)['mediaDefinitions']
 
-        def get_format_data(data, f):
-            return traverse_obj(data, lambda _, v: v['format'] == f and url_or_none(v['videoUrl']))
+        def get_format_data(data, stream_type):
+            info_url = traverse_obj(data, (lambda _, v: v['format'] == stream_type, 'videoUrl', {url_or_none}, any))
+            if not info_url:
+                return []
+            return traverse_obj(
+                self._download_json(info_url, video_id, f'Downloading {stream_type} info JSON', fatal=False),
+                lambda _, v: v['format'] == stream_type and url_or_none(v['videoUrl']))
 
         formats = []
         # Try to extract only the actual master m3u8 first, avoiding the duplicate single resolution "master" m3u8s
@@ -123,10 +129,6 @@ def get_format_data(data, f):
             f['height'] = height
             formats.append(f)
 
-        webpage = self._download_webpage(
-            'http://www.youporn.com/watch/%s' % video_id, display_id,
-            headers={'Cookie': 'age_verified=1'})
-
         title = self._html_search_regex(
             r'(?s)<div[^>]+class=["\']watchVideoTitle[^>]+>(.+?)</div>',
             webpage, 'title', default=None) or self._og_search_title(