better episode number extraction

Authored by: bashonly

yt_dlp/extractor/nfb.py

@@ -5,6 +5,7 @@ from ..utils import (
     merge_dicts,
     parse_count,
     url_or_none,
+    urljoin,
 )
 from ..utils.traversal import traverse_obj
 
@@ -18,7 +19,7 @@ class NFBBaseIE(InfoExtractor):
             r'const\s+episodesData\s*=', webpage, 'episode data', video_id,
             contains_pattern=r'\[\s*{(?s:.+)}\s*\]', fatal=fatal) or []
 
-    def _extract_ep_info(self, data, video_id):
+    def _extract_ep_info(self, data, video_id, slug=None):
         info = traverse_obj(data, (lambda _, v: video_id in v['embed_url'], {
             'description': ('description', {str}),
             'thumbnail': ('thumbnail_url', {url_or_none}),
@@ -35,7 +36,7 @@ class NFBBaseIE(InfoExtractor):
             'id': video_id,
             'title': join_nonempty('series', 'episode', from_dict=info, delim=' - '),
             'episode_number': int_or_none(self._search_regex(
-                r'-e(?:pisode)?-?(\d+)(?:-|$)', video_id, 'episode number', default=None)),
+                r'[/-]e(?:pisode)?-?(\d+)(?:[/-]|$)', slug or video_id, 'episode number', default=None)),
         }
 
 
@@ -200,29 +201,33 @@ class NFBIE(NFBBaseIE):
             'thumbnail': r're:^https?://.*\.jpg$',
         },
         'params': {'skip_download': 'm3u8'},
+    }, {
+        'note': 'NFB film /embed/player/ page',
+        'url': 'https://www.nfb.ca/film/afterlife/embed/player/',
+        'info_dict': {
+            'id': 'afterlife',
+            'ext': 'mp4',
+            'title': 'Afterlife',
+            'description': 'md5:84951394f594f1fb1e62d9c43242fdf5',
+            'release_year': 1978,
+            'duration': 420.0,
+            'uploader': 'Ishu Patel',
+            'thumbnail': r're:^https?://.*\.jpg$',
+        },
+        'params': {'skip_download': 'm3u8'},
     }]
 
     def _real_extract(self, url):
         site, type_, slug = self._match_valid_url(url).group('site', 'type', 'id')
+        # Need to construct the URL since we match /embed/player/ URLs as well
         webpage, urlh = self._download_webpage_handle(f'https://www.{site}.ca/{type_}/{slug}/', slug)
-        json_ld = self._yield_json_ld(webpage, slug)
+        # 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')
 
-        # /film/ URLs have unique slugs used in the embed url
+        embed_url = urljoin(f'https://www.{site}.ca', self._html_search_regex(
-        video_id = slug if type_ == 'film' else traverse_obj(
+            r'<[^>]+\bid=["\']player-iframe["\'][^>]*\bsrc=["\']([^"\']+)', webpage, 'embed url'))
-            json_ld, (lambda _, v: 'VideoObject' in v['@type'], 'embedUrl', {self._match_id}),
+        video_id = self._match_id(embed_url)  # embed url has unique slug
-            get_all=False) or self._match_id(self._og_search_property('url', webpage, 'video id'))
+        player = self._download_webpage(embed_url, video_id, 'Downloading player page')
-
-        # type_ may have changed from film to serie(s) after redirect
-        type_ = self._match_valid_url(urlh.url).group('type')
-
-        player = self._download_webpage(
-            f'https://www.{site}.ca/film/{video_id}/embed/player/', video_id,
-            'Downloading player page', query={
-                'player_mode': '',
-                'embed_mode': '0',
-                'auto_focus': '1',
-                'context_type': type_ if type_ == 'film' else 'episode',
-            })
         if 'MESSAGE_GEOBLOCKED' in player:
             self.raise_geo_restricted(countries=self._GEO_COUNTRIES)
 
@@ -244,12 +249,12 @@ class NFBIE(NFBBaseIE):
             'release_year': int_or_none(self._html_search_regex(
                 r'<[^>]+\bitemprop=["\']datePublished["\'][^>]*>([^<]+)',
                 webpage, 'release_year', default=None)),
-        } if type_ == 'film' else self._extract_ep_info(self._extract_ep_data(webpage, video_id), video_id)
+        } if type_ == 'film' else self._extract_ep_info(self._extract_ep_data(webpage, video_id, slug), video_id)
 
         return merge_dicts({
             'formats': formats,
             'subtitles': subtitles,
-        }, info, self._json_ld(json_ld, video_id))
+        }, info, self._search_json_ld(webpage, video_id, default={}))
 
 
 class NFBSeriesIE(NFBBaseIE):