Merge a177349759 into 96da952504

Authored by: seproDev, pukkandan

test/test_networking.py

@@ -785,6 +785,25 @@ class TestHTTPImpersonateRequestHandler(TestRequestHandlerBase):
                 assert res.status == 200
                 assert std_headers['user-agent'].lower() not in res.read().decode().lower()
 
+    def test_response_extensions(self, handler):
+        with handler() as rh:
+            for target in rh.supported_targets:
+                request = Request(
+                    f'http://127.0.0.1:{self.http_port}/gen_200', extensions={'impersonate': target})
+                res = validate_and_send(rh, request)
+                assert res.extensions['impersonate'] == rh._get_request_target(request)
+
+    def test_http_error_response_extensions(self, handler):
+        with handler() as rh:
+            for target in rh.supported_targets:
+                request = Request(
+                    f'http://127.0.0.1:{self.http_port}/gen_404', extensions={'impersonate': target})
+                try:
+                    validate_and_send(rh, request)
+                except HTTPError as e:
+                    res = e.response
+                assert res.extensions['impersonate'] == rh._get_request_target(request)
+
 
 class TestRequestHandlerMisc:
     """Misc generic tests for request handlers, not related to request or validation testing"""

yt_dlp/YoutubeDL.py

@@ -2136,6 +2136,11 @@ class YoutubeDL:
 
     def _check_formats(self, formats):
         for f in formats:
+            working = f.get('__working')
+            if working is not None:
+                if working:
+                    yield f
+                continue
             self.to_screen('[info] Testing format %s' % f['format_id'])
             path = self.get_output_path('temp')
             if not self._ensure_dir_exists(f'{path}/'):
@@ -2152,33 +2157,44 @@ class YoutubeDL:
                         os.remove(temp_file.name)
                     except OSError:
                         self.report_warning('Unable to delete temporary file "%s"' % temp_file.name)
+            f['__working'] = success
             if success:
                 yield f
             else:
                 self.to_screen('[info] Unable to download format %s. Skipping...' % f['format_id'])
 
+    def _select_formats(self, formats, selector):
+        return list(selector({
+            'formats': formats,
+            'has_merged_format': any('none' not in (f.get('acodec'), f.get('vcodec')) for f in formats),
+            'incomplete_formats': (all(f.get('vcodec') == 'none' for f in formats)  # No formats with video
+                                   or all(f.get('acodec') == 'none' for f in formats)),  # OR, No formats with audio
+        }))
+
     def _default_format_spec(self, info_dict, download=True):
+        download = download and not self.params.get('simulate')
+        prefer_best = download and (
+            self.params['outtmpl']['default'] == '-'
+            or info_dict.get('is_live') and not self.params.get('live_from_start'))
 
         def can_merge():
             merger = FFmpegMergerPP(self)
             return merger.available and merger.can_merge()
 
-        prefer_best = (
-            not self.params.get('simulate')
-            and download
-            and (
-                not can_merge()
-                or info_dict.get('is_live') and not self.params.get('live_from_start')
-                or self.params['outtmpl']['default'] == '-'))
-        compat = (
-            prefer_best
-            or self.params.get('allow_multiple_audio_streams', False)
-            or 'format-spec' in self.params['compat_opts'])
+        if not prefer_best and download and not can_merge():
+            prefer_best = True
+            formats = self._get_formats(info_dict)
+            evaluate_formats = lambda spec: self._select_formats(formats, self.build_format_selector(spec))
+            if evaluate_formats('b/bv+ba') != evaluate_formats('bv*+ba/b'):
+                self.report_warning('ffmpeg not found. The downloaded format may not be the best available. '
+                                    'Installing ffmpeg is strongly recommended: https://github.com/yt-dlp/yt-dlp#dependencies')
 
-        return (
-            'best/bestvideo+bestaudio' if prefer_best
-            else 'bestvideo*+bestaudio/best' if not compat
-            else 'bestvideo+bestaudio/best')
+        compat = (self.params.get('allow_multiple_audio_streams')
+                  or 'format-spec' in self.params['compat_opts'])
+
+        return ('best/bestvideo+bestaudio' if prefer_best
+                else 'bestvideo+bestaudio/best' if compat
+                else 'bestvideo*+bestaudio/best')
 
     def build_format_selector(self, format_spec):
         def syntax_error(note, start):
@@ -2928,12 +2944,7 @@ class YoutubeDL:
                 self.write_debug(f'Default format spec: {req_format}')
                 format_selector = self.build_format_selector(req_format)
 
-            formats_to_download = list(format_selector({
-                'formats': formats,
-                'has_merged_format': any('none' not in (f.get('acodec'), f.get('vcodec')) for f in formats),
-                'incomplete_formats': (all(f.get('vcodec') == 'none' for f in formats)  # No formats with video
-                                       or all(f.get('acodec') == 'none' for f in formats)),  # OR, No formats with audio
-            }))
+            formats_to_download = self._select_formats(formats, format_selector)
             if interactive_format_selection and not formats_to_download:
                 self.report_error('Requested format is not available', tb=False, is_error=False)
                 continue

yt_dlp/extractor/patreon.py

@@ -219,7 +219,29 @@ class PatreonIE(PatreonBaseIE):
             'thumbnail': r're:^https?://.+',
         },
         'params': {'skip_download': 'm3u8'},
+    }, {
+        # multiple attachments/embeds
+        'url': 'https://www.patreon.com/posts/holy-wars-solos-100601977',
+        'playlist_count': 3,
+        'info_dict': {
+            'id': '100601977',
+            'title': '"Holy Wars" (Megadeth) Solos Transcription & Lesson/Analysis',
+            'description': 'md5:d099ab976edfce6de2a65c2b169a88d3',
+            'uploader': 'Bradley Hall',
+            'uploader_id': '24401883',
+            'uploader_url': 'https://www.patreon.com/bradleyhallguitar',
+            'channel_id': '3193932',
+            'channel_url': 'https://www.patreon.com/bradleyhallguitar',
+            'channel_follower_count': int,
+            'timestamp': 1710777855,
+            'upload_date': '20240318',
+            'like_count': int,
+            'comment_count': int,
+            'thumbnail': r're:^https?://.+',
+        },
+        'skip': 'Patron-only content',
     }]
+    _RETURN_TYPE = 'video'
 
     def _real_extract(self, url):
         video_id = self._match_id(url)
@@ -234,58 +256,54 @@ class PatreonIE(PatreonBaseIE):
                 'include': 'audio,user,user_defined_tags,campaign,attachments_media',
             })
         attributes = post['data']['attributes']
-        title = attributes['title'].strip()
-        image = attributes.get('image') or {}
-        info = {
-            'id': video_id,
-            'title': title,
-            'description': clean_html(attributes.get('content')),
-            'thumbnail': image.get('large_url') or image.get('url'),
-            'timestamp': parse_iso8601(attributes.get('published_at')),
-            'like_count': int_or_none(attributes.get('like_count')),
-            'comment_count': int_or_none(attributes.get('comment_count')),
-        }
-        can_view_post = traverse_obj(attributes, 'current_user_can_view')
-        if can_view_post and info['comment_count']:
-            info['__post_extractor'] = self.extract_comments(video_id)
+        info = traverse_obj(attributes, {
+            'title': ('title', {str.strip}),
+            'description': ('content', {clean_html}),
+            'thumbnail': ('image', ('large_url', 'url'), {url_or_none}, any),
+            'timestamp': ('published_at', {parse_iso8601}),
+            'like_count': ('like_count', {int_or_none}),
+            'comment_count': ('comment_count', {int_or_none}),
+        })
 
-        for i in post.get('included', []):
-            i_type = i.get('type')
-            if i_type == 'media':
-                media_attributes = i.get('attributes') or {}
-                download_url = media_attributes.get('download_url')
+        entries = []
+        idx = 0
+        for include in traverse_obj(post, ('included', lambda _, v: v['type'])):
+            include_type = include['type']
+            if include_type == 'media':
+                media_attributes = traverse_obj(include, ('attributes', {dict})) or {}
+                download_url = url_or_none(media_attributes.get('download_url'))
                 ext = mimetype2ext(media_attributes.get('mimetype'))
 
                 # if size_bytes is None, this media file is likely unavailable
                 # See: https://github.com/yt-dlp/yt-dlp/issues/4608
                 size_bytes = int_or_none(media_attributes.get('size_bytes'))
                 if download_url and ext in KNOWN_EXTENSIONS and size_bytes is not None:
-                    # XXX: what happens if there are multiple attachments?
-                    return {
-                        **info,
+                    idx += 1
+                    entries.append({
+                        'id': f'{video_id}-{idx}',
                         'ext': ext,
                         'filesize': size_bytes,
                         'url': download_url,
-                    }
-            elif i_type == 'user':
-                user_attributes = i.get('attributes')
-                if user_attributes:
-                    info.update({
-                        'uploader': user_attributes.get('full_name'),
-                        'uploader_id': str_or_none(i.get('id')),
-                        'uploader_url': user_attributes.get('url'),
                     })
 
-            elif i_type == 'post_tag':
-                info.setdefault('tags', []).append(traverse_obj(i, ('attributes', 'value')))
+            elif include_type == 'user':
+                info.update(traverse_obj(include, {
+                    'uploader': ('attributes', 'full_name', {str}),
+                    'uploader_id': ('id', {str_or_none}),
+                    'uploader_url': ('attributes', 'url', {url_or_none}),
+                }))
 
-            elif i_type == 'campaign':
-                info.update({
-                    'channel': traverse_obj(i, ('attributes', 'title')),
-                    'channel_id': str_or_none(i.get('id')),
-                    'channel_url': traverse_obj(i, ('attributes', 'url')),
-                    'channel_follower_count': int_or_none(traverse_obj(i, ('attributes', 'patron_count'))),
-                })
+            elif include_type == 'post_tag':
+                if post_tag := traverse_obj(include, ('attributes', 'value', {str})):
+                    info.setdefault('tags', []).append(post_tag)
+
+            elif include_type == 'campaign':
+                info.update(traverse_obj(include, {
+                    'channel': ('attributes', 'title', {str}),
+                    'channel_id': ('id', {str_or_none}),
+                    'channel_url': ('attributes', 'url', {url_or_none}),
+                    'channel_follower_count': ('attributes', 'patron_count', {int_or_none}),
+                }))
 
         # handle Vimeo embeds
         if traverse_obj(attributes, ('embed', 'provider')) == 'Vimeo':
@@ -296,36 +314,50 @@ class PatreonIE(PatreonBaseIE):
                     v_url, video_id, 'Checking Vimeo embed URL',
                     headers={'Referer': 'https://patreon.com/'},
                     fatal=False, errnote=False):
-                return self.url_result(
+                entries.append(self.url_result(
                     VimeoIE._smuggle_referrer(v_url, 'https://patreon.com/'),
-                    VimeoIE, url_transparent=True, **info)
+                    VimeoIE, url_transparent=True))
 
         embed_url = traverse_obj(attributes, ('embed', 'url', {url_or_none}))
         if embed_url and self._request_webpage(embed_url, video_id, 'Checking embed URL', fatal=False, errnote=False):
-            return self.url_result(embed_url, **info)
+            entries.append(self.url_result(embed_url))
 
-        post_file = traverse_obj(attributes, 'post_file')
+        post_file = traverse_obj(attributes, ('post_file', {dict}))
         if post_file:
             name = post_file.get('name')
             ext = determine_ext(name)
             if ext in KNOWN_EXTENSIONS:
-                return {
-                    **info,
+                entries.append({
+                    'id': video_id,
                     'ext': ext,
                     'url': post_file['url'],
-                }
+                })
             elif name == 'video' or determine_ext(post_file.get('url')) == 'm3u8':
                 formats, subtitles = self._extract_m3u8_formats_and_subtitles(post_file['url'], video_id)
-                return {
-                    **info,
+                entries.append({
+                    'id': video_id,
                     'formats': formats,
                     'subtitles': subtitles,
-                }
+                })
 
-        if can_view_post is False:
+        can_view_post = traverse_obj(attributes, 'current_user_can_view')
+        comments = None
+        if can_view_post and info.get('comment_count'):
+            comments = self.extract_comments(video_id)
+
+        if not entries and can_view_post is False:
             self.raise_no_formats('You do not have access to this post', video_id=video_id, expected=True)
-        else:
+        elif not entries:
             self.raise_no_formats('No supported media found in this post', video_id=video_id, expected=True)
+        elif len(entries) == 1:
+            info.update(entries[0])
+        else:
+            for entry in entries:
+                entry.update(info)
+            return self.playlist_result(entries, video_id, **info, __post_extractor=comments)
+
+        info['id'] = video_id
+        info['__post_extractor'] = comments
         return info
 
     def _get_comments(self, post_id):

yt_dlp/extractor/yandexmusic.py

@@ -1,12 +1,14 @@
 import hashlib
 import itertools
+from datetime import datetime, timezone
 
 from .common import InfoExtractor
 from ..compat import compat_str
 from ..utils import (
     ExtractorError,
-    int_or_none,
     float_or_none,
+    int_or_none,
+    traverse_obj,
     try_get,
 )
 
@@ -57,6 +59,41 @@ class YandexMusicBaseIE(InfoExtractor):
             },
             query=query)
 
+    def _extract_timestamp(self, album):
+        dt = None
+        if release_date := album.get('releaseDate'):
+            dt = datetime.fromisoformat(release_date)
+        elif year := album.get('year'):
+            dt = datetime(year, 1, 1, tzinfo=timezone.utc)
+        if dt:
+            return int(dt.timestamp())
+
+    def _extract_thumbnail(self, data):
+        if cover_uri := data.get('ogImage'):
+            thumbnail = cover_uri.replace('%%', 'orig')
+            if not thumbnail.startswith('http'):
+                thumbnail = 'https://' + thumbnail
+            return thumbnail
+
+    def _extract_artists(self, data):
+        def artist_name(artist):
+            decomposed = artist.get('decomposed')
+            if not isinstance(decomposed, list):
+                return artist['name']
+            parts = [artist['name']]
+            for element in decomposed:
+                if isinstance(element, dict) and element.get('name'):
+                    parts.append(element['name'])
+                elif isinstance(element, compat_str):
+                    parts.append(element)
+            return ''.join(parts)
+
+        artist_list = data.get('artists')
+        if artist_list and isinstance(artist_list, list):
+            artists_names = [artist_name(a) for a in artist_list if a.get('name')]
+            if artists_names:
+                return ', '.join(artists_names)
+
 
 class YandexMusicTrackIE(YandexMusicBaseIE):
     IE_NAME = 'yandexmusic:track'
@@ -76,29 +113,34 @@ class YandexMusicTrackIE(YandexMusicBaseIE):
             'album': 'md5:cd04fb13c4efeafdfa0a6a6aca36d01a',
             'album_artist': 'md5:5f54c35462c07952df33d97cfb5fc200',
             'artist': 'md5:e6fd86621825f14dc0b25db3acd68160',
-            'release_year': 2009,
+            'timestamp': 1241208000,
+            'disc_number': 1,
+            'track_number': 11,
+            'thumbnail': 're:^https://.*$',
+            'upload_date': str
         },
         # 'skip': 'Travis CI servers blocked by YandexMusic',
     }, {
         # multiple disks
-        'url': 'http://music.yandex.ru/album/3840501/track/705105',
-        'md5': '82a54e9e787301dd45aba093cf6e58c0',
+        'url': 'https://music.yandex.ru/album/19235277/track/94721644',
+        'md5': 'f239d4923c5fdbc2cc8f6c80b4fec721',
         'info_dict': {
-            'id': '705105',
+            'id': '94721644',
             'ext': 'mp3',
-            'title': 'md5:f86d4a9188279860a83000277024c1a6',
+            'title': 'Nirvana - Sliver',
             'filesize': int,
-            'duration': 239.27,
-            'track': 'md5:40f887f0666ba1aa10b835aca44807d1',
-            'album': 'md5:624f5224b14f5c88a8e812fd7fbf1873',
-            'album_artist': 'md5:dd35f2af4e8927100cbe6f5e62e1fb12',
-            'artist': 'md5:dd35f2af4e8927100cbe6f5e62e1fb12',
-            'release_year': 2016,
-            'genre': 'pop',
+            'duration': 133.49,
+            'track': 'Sliver',
+            'album': 'Nevermind',
+            'album_artist': 'Nirvana',
+            'artist': 'Nirvana',
+            'timestamp': 685659600,
             'disc_number': 2,
             'track_number': 9,
-        },
-        # 'skip': 'Travis CI servers blocked by YandexMusic',
+            'thumbnail': 're:^https://.*$',
+            'upload_date': str,
+            'genre': 'rock'
+        }
     }, {
         'url': 'http://music.yandex.com/album/540508/track/4878838',
         'only_matching': True,
@@ -124,12 +166,8 @@ class YandexMusicTrackIE(YandexMusicBaseIE):
         key = hashlib.md5(('XGRlBW9FXlekgbPrRHuSiA' + fd_data['path'][1:] + fd_data['s']).encode('utf-8')).hexdigest()
         f_url = 'http://%s/get-mp3/%s/%s?track-id=%s ' % (fd_data['host'], key, fd_data['ts'] + fd_data['path'], track['id'])
 
-        thumbnail = None
-        cover_uri = track.get('albums', [{}])[0].get('coverUri')
-        if cover_uri:
-            thumbnail = cover_uri.replace('%%', 'orig')
-            if not thumbnail.startswith('http'):
-                thumbnail = 'http://' + thumbnail
+        album = traverse_obj(track, ('albums', 0), expected_type=dict)
+        thumbnail = self._extract_thumbnail(track)
 
         track_info = {
             'id': track_id,
@@ -143,43 +181,24 @@ class YandexMusicTrackIE(YandexMusicBaseIE):
             'abr': int_or_none(download_data.get('bitrate')),
         }
 
-        def extract_artist_name(artist):
-            decomposed = artist.get('decomposed')
-            if not isinstance(decomposed, list):
-                return artist['name']
-            parts = [artist['name']]
-            for element in decomposed:
-                if isinstance(element, dict) and element.get('name'):
-                    parts.append(element['name'])
-                elif isinstance(element, compat_str):
-                    parts.append(element)
-            return ''.join(parts)
-
-        def extract_artist(artist_list):
-            if artist_list and isinstance(artist_list, list):
-                artists_names = [extract_artist_name(a) for a in artist_list if a.get('name')]
-                if artists_names:
-                    return ', '.join(artists_names)
-
         albums = track.get('albums')
         if albums and isinstance(albums, list):
             album = albums[0]
             if isinstance(album, dict):
-                year = album.get('year')
                 disc_number = int_or_none(try_get(
                     album, lambda x: x['trackPosition']['volume']))
                 track_number = int_or_none(try_get(
                     album, lambda x: x['trackPosition']['index']))
                 track_info.update({
                     'album': album.get('title'),
-                    'album_artist': extract_artist(album.get('artists')),
-                    'release_year': int_or_none(year),
+                    'album_artist': self._extract_artists(album),
+                    'timestamp': self._extract_timestamp(album),
                     'genre': album.get('genre'),
                     'disc_number': disc_number,
                     'track_number': track_number,
                 })
 
-        track_artist = extract_artist(track.get('artists'))
+        track_artist = self._extract_artists(track)
         if track_artist:
             track_info.update({
                 'artist': track_artist,
@@ -259,26 +278,23 @@ class YandexMusicAlbumIE(YandexMusicPlaylistBaseIE):
         'url': 'http://music.yandex.ru/album/540508',
         'info_dict': {
             'id': '540508',
-            'title': 'md5:7ed1c3567f28d14be9f61179116f5571',
+            'title': 'Gypsy Soul',
+            'artist': 'Carlo Ambrosio',
+            'thumbnail': 're:^https://.*$',
+            'timestamp': 1241208000,
+            'upload_date': str
         },
         'playlist_count': 50,
         # 'skip': 'Travis CI servers blocked by YandexMusic',
-    }, {
-        'url': 'https://music.yandex.ru/album/3840501',
-        'info_dict': {
-            'id': '3840501',
-            'title': 'md5:36733472cdaa7dcb1fd9473f7da8e50f',
-        },
-        'playlist_count': 33,
-        # 'skip': 'Travis CI servers blocked by YandexMusic',
-    }, {
-        # empty artists
+    }, {  # empty artists
         'url': 'https://music.yandex.ru/album/9091882',
         'info_dict': {
             'id': '9091882',
             'title': 'ТЕД на русском',
+            'thumbnail': 're:^https://.*$',
+            'artist': None,
         },
-        'playlist_count': 187,
+        'playlist_count': 166,
     }]
 
     @classmethod
@@ -296,15 +312,11 @@ class YandexMusicAlbumIE(YandexMusicPlaylistBaseIE):
 
         entries = self._build_playlist([track for volume in album['volumes'] for track in volume])
 
-        title = album['title']
-        artist = try_get(album, lambda x: x['artists'][0]['name'], compat_str)
-        if artist:
-            title = '%s - %s' % (artist, title)
-        year = album.get('year')
-        if year:
-            title += ' (%s)' % year
-
-        return self.playlist_result(entries, compat_str(album['id']), title)
+        return self.playlist_result(entries, compat_str(album['id']),
+                                    album.get('title'),
+                                    thumbnail=self._extract_thumbnail(album),
+                                    artist=self._extract_artists(album),
+                                    timestamp=self._extract_timestamp(album))
 
 
 class YandexMusicPlaylistIE(YandexMusicPlaylistBaseIE):
@@ -318,6 +330,7 @@ class YandexMusicPlaylistIE(YandexMusicPlaylistBaseIE):
             'id': '1245',
             'title': 'md5:841559b3fe2b998eca88d0d2e22a3097',
             'description': 'md5:3b9f27b0efbe53f2ee1e844d07155cc9',
+            'thumbnail': 're:^https://.*$',
         },
         'playlist_count': 5,
         # 'skip': 'Travis CI servers blocked by YandexMusic',
@@ -331,8 +344,10 @@ class YandexMusicPlaylistIE(YandexMusicPlaylistBaseIE):
         'info_dict': {
             'id': '1364',
             'title': 'md5:b3b400f997d3f878a13ae0699653f7db',
+            'thumbnail': 're:^https://.*$',
+            'description': ''
         },
-        'playlist_mincount': 437,
+        'playlist_mincount': 300,
         # 'skip': 'Travis CI servers blocked by YandexMusic',
     }]
 
@@ -357,7 +372,8 @@ class YandexMusicPlaylistIE(YandexMusicPlaylistBaseIE):
         return self.playlist_result(
             self._build_playlist(tracks),
             compat_str(playlist_id),
-            playlist.get('title'), playlist.get('description'))
+            playlist.get('title'), playlist.get('description'),
+            thumbnail=self._extract_thumbnail(playlist))
 
 
 class YandexMusicArtistBaseIE(YandexMusicPlaylistBaseIE):
@@ -396,9 +412,20 @@ class YandexMusicArtistTracksIE(YandexMusicArtistBaseIE):
         'info_dict': {
             'id': '617526',
             'title': 'md5:131aef29d45fd5a965ca613e708c040b',
+            'artist': 'Ringtone Hits',
+            'thumbnail': 're:^https://.*$',
         },
         'playlist_count': 507,
         # 'skip': 'Travis CI servers blocked by YandexMusic',
+    }, {
+        'url': 'https://music.yandex.ru/artist/9262/tracks',
+        'info_dict': {
+            'id': '9262',
+            'title': 'Nirvana - Треки',
+            'artist': 'Nirvana',
+            'thumbnail': 're:^https://.*$',
+        },
+        'playlist_count': 101
     }]
 
     _ARTIST_SORT = ''
@@ -410,10 +437,15 @@ class YandexMusicArtistTracksIE(YandexMusicArtistBaseIE):
         artist_id = mobj.group('id')
         data = self._call_artist(tld, url, artist_id)
         tracks = self._extract_tracks(data, artist_id, url, tld)
-        artist = try_get(data, lambda x: x['artist']['name'], compat_str)
-        title = '%s - %s' % (artist or artist_id, 'Треки')
+        thumbnail = None
+        artist_name = None
+        if artist := data.get('artist'):
+            artist_name = artist.get('name')
+            thumbnail = self._extract_thumbnail(artist)
+        title = '%s - %s' % (artist_name or artist_id, 'Треки')
         return self.playlist_result(
-            self._build_playlist(tracks), artist_id, title)
+            self._build_playlist(tracks), artist_id, title,
+            thumbnail=thumbnail, artist=artist_name)
 
 
 class YandexMusicArtistAlbumsIE(YandexMusicArtistBaseIE):
@@ -426,9 +458,20 @@ class YandexMusicArtistAlbumsIE(YandexMusicArtistBaseIE):
         'info_dict': {
             'id': '617526',
             'title': 'md5:55dc58d5c85699b7fb41ee926700236c',
+            'artist': 'Ringtone Hits',
+            'thumbnail': 're:^https://.*$',
         },
         'playlist_count': 8,
         # 'skip': 'Travis CI servers blocked by YandexMusic',
+    }, {  # artist with thumbnail
+        'url': 'https://music.yandex.ru/artist/9262/albums',
+        'info_dict': {
+            'id': '9262',
+            'title': 'Nirvana - Альбомы',
+            'artist': 'Nirvana',
+            'thumbnail': 're:^https://.*$',
+        },
+        'playlist_count': 15
     }]
 
     _ARTIST_SORT = 'year'
@@ -449,6 +492,11 @@ class YandexMusicArtistAlbumsIE(YandexMusicArtistBaseIE):
             entries.append(self.url_result(
                 'http://music.yandex.ru/album/%s' % album_id,
                 ie=YandexMusicAlbumIE.ie_key(), video_id=album_id))
-        artist = try_get(data, lambda x: x['artist']['name'], compat_str)
-        title = '%s - %s' % (artist or artist_id, 'Альбомы')
-        return self.playlist_result(entries, artist_id, title)
+        thumbnail = None
+        artist_name = None
+        if artist := data.get('artist'):
+            artist_name = artist.get('name')
+            thumbnail = self._extract_thumbnail(artist)
+        title = '%s - %s' % (artist_name or artist_id, 'Альбомы')
+        return self.playlist_result(entries, artist_id, title,
+                                    thumbnail=thumbnail, artist=artist_name)

yt_dlp/networking/_curlcffi.py

@@ -132,6 +132,16 @@ class CurlCFFIRH(ImpersonateRequestHandler, InstanceStoreMixin):
         extensions.pop('cookiejar', None)
         extensions.pop('timeout', None)
 
+    def send(self, request: Request) -> Response:
+        target = self._get_request_target(request)
+        try:
+            response = super().send(request)
+        except HTTPError as e:
+            e.response.extensions['impersonate'] = target
+            raise
+        response.extensions['impersonate'] = target
+        return response
+
     def _send(self, request: Request):
         max_redirects_exceeded = False
         session: curl_cffi.requests.Session = self._get_instance(

yt_dlp/networking/common.py

@@ -497,6 +497,7 @@ class Response(io.IOBase):
     @param headers: response headers.
     @param status: Response HTTP status code. Default is 200 OK.
     @param reason: HTTP status reason. Will use built-in reasons based on status code if not provided.
+    @param extensions: Dictionary of handler-specific response extensions.
     """
 
     def __init__(
@@ -505,7 +506,9 @@ class Response(io.IOBase):
             url: str,
             headers: Mapping[str, str],
             status: int = 200,
-            reason: str = None):
+            reason: str = None,
+            extensions: dict = None
+    ):
 
         self.fp = fp
         self.headers = Message()
@@ -517,6 +520,7 @@ class Response(io.IOBase):
             self.reason = reason or HTTPStatus(status).phrase
         except ValueError:
             self.reason = None
+        self.extensions = extensions or {}
 
     def readable(self):
         return self.fp.readable()