[cleanup] Misc fixes

Closes #4027

yt_dlp/extractor/common.py

@@ -1487,7 +1487,7 @@ class InfoExtractor:
                 # however some websites are using 'Text' type instead.
                 # 1. https://schema.org/VideoObject
                 'uploader': author.get('name') if isinstance(author, dict) else author if isinstance(author, compat_str) else None,
-                'filesize': float_or_none(e.get('contentSize')),
+                'filesize': int_or_none(float_or_none(e.get('contentSize'))),
                 'tbr': int_or_none(e.get('bitrate')),
                 'width': int_or_none(e.get('width')),
                 'height': int_or_none(e.get('height')),

yt_dlp/extractor/extractors.py

@@ -534,7 +534,6 @@ from .foxnews import (
 )
 from .foxsports import FoxSportsIE
 from .fptplay import FptplayIE
-from .franceculture import FranceCultureIE
 from .franceinter import FranceInterIE
 from .francetv import (
     FranceTVIE,
@@ -1348,7 +1347,7 @@ from .radiocanada import (
 from .radiode import RadioDeIE
 from .radiojavan import RadioJavanIE
 from .radiobremen import RadioBremenIE
-from .radiofrance import RadioFranceIE
+from .radiofrance import FranceCultureIE, RadioFranceIE
 from .radiozet import RadioZetPodcastIE
 from .radiokapital import (
     RadioKapitalIE,

yt_dlp/extractor/franceculture.py

@@ -1,46 +0,0 @@
-from .common import InfoExtractor
-from ..utils import int_or_none, parse_duration, unified_strdate
-
-
-class FranceCultureIE(InfoExtractor):
-    _VALID_URL = r'https?://(?:www\.)?radiofrance\.fr/franceculture/podcasts/(?:[^?#]+/)?(?P<display_id>[^?#]+)-(?P<id>\d+)($|[?#])'
-    _TESTS = [
-        {
-            'url': 'https://www.radiofrance.fr/franceculture/podcasts/science-en-questions/la-physique-d-einstein-aiderait-elle-a-comprendre-le-cerveau-8440487',
-            'info_dict': {
-                'id': '8440487',
-                'display_id': 'la-physique-d-einstein-aiderait-elle-a-comprendre-le-cerveau',
-                'ext': 'mp3',
-                'title': 'La physique d’Einstein aiderait-elle à comprendre le cerveau ?',
-                'description': 'Existerait-il un pont conceptuel entre la physique de l’espace-temps et les neurosciences ?',
-                'thumbnail': 'https://cdn.radiofrance.fr/s3/cruiser-production/2022/05/d184e7a3-4827-4494-bf94-04ed7b120db4/1200x630_gettyimages-200171095-001.jpg',
-                'upload_date': '20220514',
-                'duration': 2750,
-            },
-        },
-    ]
-
-    def _real_extract(self, url):
-        video_id, display_id = self._match_valid_url(url).group('id', 'display_id')
-        webpage = self._download_webpage(url, display_id)
-
-        # _search_json_ld doesn't correctly handle this. See https://github.com/yt-dlp/yt-dlp/pull/3874#discussion_r891903846
-        video_data = self._search_json('', webpage, 'audio data', display_id, contains_pattern=r'\s*"@type"\s*:\s*"AudioObject"\s*.+')
-
-        return {
-            'id': video_id,
-            'display_id': display_id,
-            'url': video_data['contentUrl'],
-            'ext': video_data.get('encodingFormat'),
-            'vcodec': 'none' if video_data.get('encodingFormat') == 'mp3' else None,
-            'duration': parse_duration(video_data.get('duration')),
-            'title': self._html_search_regex(r'(?s)<h1[^>]*itemprop="[^"]*name[^"]*"[^>]*>(.+?)</h1>',
-                                             webpage, 'title', default=self._og_search_title(webpage)),
-            'description': self._html_search_regex(
-                r'(?s)<meta name="description"\s*content="([^"]+)', webpage, 'description', default=None),
-            'thumbnail': self._og_search_thumbnail(webpage),
-            'uploader': self._html_search_regex(
-                r'(?s)<span class="author">(.*?)</span>', webpage, 'uploader', default=None),
-            'upload_date': unified_strdate(self._search_regex(
-                r'"datePublished"\s*:\s*"([^"]+)', webpage, 'timestamp', fatal=False))
-        }

yt_dlp/extractor/freetv.py

@@ -2,11 +2,7 @@ import itertools
 import re
 
 from .common import InfoExtractor
-from ..utils import (
-    int_or_none,
-    traverse_obj,
-    urlencode_postdata,
-)
+from ..utils import int_or_none, traverse_obj, urlencode_postdata
 
 
 class FreeTvBaseIE(InfoExtractor):

yt_dlp/extractor/radiofrance.py

@@ -1,6 +1,7 @@
 import re
 
 from .common import InfoExtractor
+from ..utils import parse_duration, unified_strdate
 
 
 class RadioFranceIE(InfoExtractor):
@@ -54,3 +55,47 @@ class RadioFranceIE(InfoExtractor):
             'description': description,
             'uploader': uploader,
         }
+
+
+class FranceCultureIE(InfoExtractor):
+    _VALID_URL = r'https?://(?:www\.)?radiofrance\.fr/franceculture/podcasts/(?:[^?#]+/)?(?P<display_id>[^?#]+)-(?P<id>\d+)($|[?#])'
+    _TESTS = [
+        {
+            'url': 'https://www.radiofrance.fr/franceculture/podcasts/science-en-questions/la-physique-d-einstein-aiderait-elle-a-comprendre-le-cerveau-8440487',
+            'info_dict': {
+                'id': '8440487',
+                'display_id': 'la-physique-d-einstein-aiderait-elle-a-comprendre-le-cerveau',
+                'ext': 'mp3',
+                'title': 'La physique d’Einstein aiderait-elle à comprendre le cerveau ?',
+                'description': 'Existerait-il un pont conceptuel entre la physique de l’espace-temps et les neurosciences ?',
+                'thumbnail': 'https://cdn.radiofrance.fr/s3/cruiser-production/2022/05/d184e7a3-4827-4494-bf94-04ed7b120db4/1200x630_gettyimages-200171095-001.jpg',
+                'upload_date': '20220514',
+                'duration': 2750,
+            },
+        },
+    ]
+
+    def _real_extract(self, url):
+        video_id, display_id = self._match_valid_url(url).group('id', 'display_id')
+        webpage = self._download_webpage(url, display_id)
+
+        # _search_json_ld doesn't correctly handle this. See https://github.com/yt-dlp/yt-dlp/pull/3874#discussion_r891903846
+        video_data = self._search_json('', webpage, 'audio data', display_id, contains_pattern=r'\s*"@type"\s*:\s*"AudioObject"\s*.+')
+
+        return {
+            'id': video_id,
+            'display_id': display_id,
+            'url': video_data['contentUrl'],
+            'ext': video_data.get('encodingFormat'),
+            'vcodec': 'none' if video_data.get('encodingFormat') == 'mp3' else None,
+            'duration': parse_duration(video_data.get('duration')),
+            'title': self._html_search_regex(r'(?s)<h1[^>]*itemprop="[^"]*name[^"]*"[^>]*>(.+?)</h1>',
+                                             webpage, 'title', default=self._og_search_title(webpage)),
+            'description': self._html_search_regex(
+                r'(?s)<meta name="description"\s*content="([^"]+)', webpage, 'description', default=None),
+            'thumbnail': self._og_search_thumbnail(webpage),
+            'uploader': self._html_search_regex(
+                r'(?s)<span class="author">(.*?)</span>', webpage, 'uploader', default=None),
+            'upload_date': unified_strdate(self._search_regex(
+                r'"datePublished"\s*:\s*"([^"]+)', webpage, 'timestamp', fatal=False))
+        }

yt_dlp/extractor/youtube.py

@@ -3674,8 +3674,7 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
 
         initial_data = None
         if webpage:
-            initial_data = self._search_json(
-                self._YT_INITIAL_DATA_RE, webpage, 'yt initial data', video_id, fatal=False)
+            initial_data = self.extract_yt_initial_data(video_id, webpage, fatal=False)
         if not initial_data:
             query = {'videoId': video_id}
             query.update(self._get_checkok_params())