Merge 654173ac66 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/_extractors.py

@@ -1535,6 +1535,7 @@ from .qqmusic import (
     QQMusicAlbumIE,
     QQMusicToplistIE,
     QQMusicPlaylistIE,
+    QQMusicVideoIE,
 )
 from .r7 import (
     R7IE,

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/qqmusic.py

@@ -1,67 +1,32 @@
+import base64
+import json
 import random
-import re
 import time
 
 from .common import InfoExtractor
 from ..utils import (
     clean_html,
-    ExtractorError,
+    int_or_none,
+    join_nonempty,
+    js_to_json,
     strip_jsonp,
+    str_or_none,
+    traverse_obj,
     unescapeHTML,
+    url_or_none,
+    ExtractorError,
 )
 
 
-class QQMusicIE(InfoExtractor):
-    IE_NAME = 'qqmusic'
-    IE_DESC = 'QQ音乐'
-    _VALID_URL = r'https?://y\.qq\.com/n/yqq/song/(?P<id>[0-9A-Za-z]+)\.html'
-    _TESTS = [{
-        'url': 'https://y.qq.com/n/yqq/song/004295Et37taLD.html',
-        'md5': '5f1e6cea39e182857da7ffc5ef5e6bb8',
-        'info_dict': {
-            'id': '004295Et37taLD',
-            'ext': 'mp3',
-            'title': '可惜没如果',
-            'release_date': '20141227',
-            'creator': '林俊杰',
-            'description': 'md5:d85afb3051952ecc50a1ee8a286d1eac',
-            'thumbnail': r're:^https?://.*\.jpg$',
-        }
-    }, {
-        'note': 'There is no mp3-320 version of this song.',
-        'url': 'https://y.qq.com/n/yqq/song/004MsGEo3DdNxV.html',
-        'md5': 'fa3926f0c585cda0af8fa4f796482e3e',
-        'info_dict': {
-            'id': '004MsGEo3DdNxV',
-            'ext': 'mp3',
-            'title': '如果',
-            'release_date': '20050626',
-            'creator': '李季美',
-            'description': 'md5:46857d5ed62bc4ba84607a805dccf437',
-            'thumbnail': r're:^https?://.*\.jpg$',
-        }
-    }, {
-        'note': 'lyrics not in .lrc format',
-        'url': 'https://y.qq.com/n/yqq/song/001JyApY11tIp6.html',
-        'info_dict': {
-            'id': '001JyApY11tIp6',
-            'ext': 'mp3',
-            'title': 'Shadows Over Transylvania',
-            'release_date': '19970225',
-            'creator': 'Dark Funeral',
-            'description': 'md5:c9b20210587cbcd6836a1c597bab4525',
-            'thumbnail': r're:^https?://.*\.jpg$',
-        },
-        'params': {
-            'skip_download': True,
-        },
-    }]
+class QQMusicBaseIE(InfoExtractor):
+    def _get_g_tk(self):
+        n = 5381
+        for chr in self._get_cookies('https://y.qq.com').get('qqmusic_key', ''):
+            n += (n << 5) + ord(chr)
+        return n & 2147483647
 
-    _FORMATS = {
-        'mp3-320': {'prefix': 'M800', 'ext': 'mp3', 'preference': 40, 'abr': 320},
-        'mp3-128': {'prefix': 'M500', 'ext': 'mp3', 'preference': 30, 'abr': 128},
-        'm4a': {'prefix': 'C200', 'ext': 'm4a', 'preference': 10}
-    }
+    def _get_uin(self):
+        return int_or_none(self._get_cookies('https://y.qq.com').get('o_cookie')) or 0
 
     # Reference: m_r_GetRUin() in top_player.js
     # http://imgcache.gtimg.cn/music/portal_v3/y/top_player.js
@@ -70,166 +35,260 @@ class QQMusicIE(InfoExtractor):
         curMs = int(time.time() * 1000) % 1000
         return int(round(random.random() * 2147483647) * curMs % 1E10)
 
-    def _real_extract(self, url):
-        mid = self._match_id(url)
+    def download_init_data(self, url, mid, fatal=True):
+        webpage = self._download_webpage(url, mid, fatal=fatal)
+        return self._search_json(r'window\.__INITIAL_DATA__\s*=\s*', webpage,
+                                 'init data', mid, transform_source=js_to_json, fatal=fatal)
 
-        detail_info_page = self._download_webpage(
-            'http://s.plcloud.music.qq.com/fcgi-bin/fcg_yqq_song_detail_info.fcg?songmid=%s&play=0' % mid,
-            mid, note='Download song detail info',
-            errnote='Unable to get song detail info', encoding='gbk')
-
-        song_name = self._html_search_regex(
-            r"songname:\s*'([^']+)'", detail_info_page, 'song name')
-
-        publish_time = self._html_search_regex(
-            r'发行时间：(\d{4}-\d{2}-\d{2})', detail_info_page,
-            'publish time', default=None)
-        if publish_time:
-            publish_time = publish_time.replace('-', '')
-
-        singer = self._html_search_regex(
-            r"singer:\s*'([^']+)", detail_info_page, 'singer', default=None)
-
-        lrc_content = self._html_search_regex(
-            r'<div class="content" id="lrc_content"[^<>]*>([^<>]+)</div>',
-            detail_info_page, 'LRC lyrics', default=None)
-        if lrc_content:
-            lrc_content = lrc_content.replace('\\n', '\n')
-
-        thumbnail_url = None
-        albummid = self._search_regex(
-            [r'albummid:\'([0-9a-zA-Z]+)\'', r'"albummid":"([0-9a-zA-Z]+)"'],
-            detail_info_page, 'album mid', default=None)
-        if albummid:
-            thumbnail_url = 'http://i.gtimg.cn/music/photo/mid_album_500/%s/%s/%s.jpg' \
-                            % (albummid[-2:-1], albummid[-1], albummid)
-
-        guid = self.m_r_get_ruin()
-
-        vkey = self._download_json(
-            'http://base.music.qq.com/fcgi-bin/fcg_musicexpress.fcg?json=3&guid=%s' % guid,
-            mid, note='Retrieve vkey', errnote='Unable to get vkey',
-            transform_source=strip_jsonp)['key']
-
-        formats = []
-        for format_id, details in self._FORMATS.items():
-            formats.append({
-                'url': 'http://cc.stream.qqmusic.qq.com/%s%s.%s?vkey=%s&guid=%s&fromtag=0'
-                       % (details['prefix'], mid, details['ext'], vkey, guid),
-                'format': format_id,
-                'format_id': format_id,
-                'quality': details['preference'],
-                'abr': details.get('abr'),
-            })
-        self._check_formats(formats, mid)
-
-        actual_lrc_lyrics = ''.join(
-            line + '\n' for line in re.findall(
-                r'(?m)^(\[[0-9]{2}:[0-9]{2}(?:\.[0-9]{2,})?\][^\n]*|\[[^\]]*\])', lrc_content))
-
-        info_dict = {
-            'id': mid,
-            'formats': formats,
-            'title': song_name,
-            'release_date': publish_time,
-            'creator': singer,
-            'description': lrc_content,
-            'thumbnail': thumbnail_url
-        }
-        if actual_lrc_lyrics:
-            info_dict['subtitles'] = {
-                'origin': [{
-                    'ext': 'lrc',
-                    'data': actual_lrc_lyrics,
-                }]
-            }
-        return info_dict
-
-
-class QQPlaylistBaseIE(InfoExtractor):
-    @staticmethod
-    def qq_static_url(category, mid):
-        return 'http://y.qq.com/y/static/%s/%s/%s/%s.html' % (category, mid[-2], mid[-1], mid)
-
-    def get_singer_all_songs(self, singmid, num):
-        return self._download_webpage(
-            r'https://c.y.qq.com/v8/fcg-bin/fcg_v8_singer_track_cp.fcg', singmid,
-            query={
+    def make_fcu_req(self, req_dict, mid, **kwargs):
+        payload = json.dumps({
+            'comm': {
+                'cv': 0,
+                'ct': 24,
                 'format': 'json',
-                'inCharset': 'utf8',
-                'outCharset': 'utf-8',
-                'platform': 'yqq',
-                'needNewCode': 0,
-                'singermid': singmid,
-                'order': 'listen',
-                'begin': 0,
-                'num': num,
-                'songstatus': 1,
-            })
+                'uin': self._get_uin(),
+            },
+            **req_dict
+        }, separators=(',', ':')).encode('utf-8')
 
-    def get_entries_from_page(self, singmid):
-        entries = []
-
-        default_num = 1
-        json_text = self.get_singer_all_songs(singmid, default_num)
-        json_obj_all_songs = self._parse_json(json_text, singmid)
-
-        if json_obj_all_songs['code'] == 0:
-            total = json_obj_all_songs['data']['total']
-            json_text = self.get_singer_all_songs(singmid, total)
-            json_obj_all_songs = self._parse_json(json_text, singmid)
-
-        for item in json_obj_all_songs['data']['list']:
-            if item['musicData'].get('songmid') is not None:
-                songmid = item['musicData']['songmid']
-                entries.append(self.url_result(
-                    r'https://y.qq.com/n/yqq/song/%s.html' % songmid, 'QQMusic', songmid))
-
-        return entries
+        return self._download_json('https://u.y.qq.com/cgi-bin/musicu.fcg',
+                                   mid, data=payload, **kwargs)
 
 
-class QQMusicSingerIE(QQPlaylistBaseIE):
-    IE_NAME = 'qqmusic:singer'
-    IE_DESC = 'QQ音乐 - 歌手'
-    _VALID_URL = r'https?://y\.qq\.com/n/yqq/singer/(?P<id>[0-9A-Za-z]+)\.html'
-    _TEST = {
-        'url': 'https://y.qq.com/n/yqq/singer/001BLpXF2DyJe2.html',
+class QQMusicIE(QQMusicBaseIE):
+    IE_NAME = 'qqmusic'
+    IE_DESC = 'QQ音乐'
+    _VALID_URL = r'https?://y\.qq\.com/n/ryqq/songDetail/(?P<id>[0-9A-Za-z]+)'
+    _TESTS = [{
+        'url': 'https://y.qq.com/n/ryqq/songDetail/004Ti8rT003TaZ',
+        'md5': 'd7adc5c438d12e2cb648cca81593fd47',
         'info_dict': {
-            'id': '001BLpXF2DyJe2',
-            'title': '林俊杰',
-            'description': 'md5:870ec08f7d8547c29c93010899103751',
+            'id': '004Ti8rT003TaZ',
+            'ext': 'mp3',
+            'title': '永夜のパレード (永夜的游行)',
+            'album': '幻想遊園郷 -Fantastic Park-',
+            'release_date': '20111230',
+            'duration': 281,
+            'creators': ['ケーキ姫', 'JUMA'],
+            'genres': ['Pop'],
+            'description': 'md5:b5261f3d595657ae561e9e6aee7eb7d9',
+            'size': 4501244,
+            'thumbnail': r're:^https?://.*\.jpg(?:$|[#?])',
+            'subtitles': 'count:1',
         },
-        'playlist_mincount': 12,
+        'params': {'listsubtitles': True},
+    }, {
+        'url': 'https://y.qq.com/n/ryqq/songDetail/004295Et37taLD',
+        'md5': '5f1e6cea39e182857da7ffc5ef5e6bb8',
+        'info_dict': {
+            'id': '004295Et37taLD',
+            'ext': 'mp3',
+            'title': '可惜没如果',
+            'album': '新地球 - 人 (Special Edition)',
+            'release_date': '20150129',
+            'duration': 298,
+            'creators': ['林俊杰'],
+            'genres': ['Pop'],
+            'description': 'md5:f568421ff618d2066e74b65a04149c4e',
+            'thumbnail': r're:^https?://.*\.jpg(?:$|[#?])',
+        },
+        'skip': 'premium member only',
+    }, {
+        'note': 'There is no mp3-320 version of this song.',
+        'url': 'https://y.qq.com/n/ryqq/songDetail/004MsGEo3DdNxV',
+        'md5': '028aaef1ae13d8a9f4861a92614887f9',
+        'info_dict': {
+            'id': '004MsGEo3DdNxV',
+            'ext': 'mp3',
+            'title': '如果',
+            'album': '新传媒电视连续剧金曲系列II',
+            'release_date': '20050626',
+            'duration': 220,
+            'creators': ['李季美'],
+            'description': 'md5:fc711212aa623b28534954dc4bd67385',
+            'size': 3535730,
+            'thumbnail': r're:^https?://.*\.jpg(?:$|[#?])',
+        },
+    }, {
+        'note': 'lyrics not in .lrc format',
+        'url': 'https://y.qq.com/n/ryqq/songDetail/001JyApY11tIp6',
+        'info_dict': {
+            'id': '001JyApY11tIp6',
+            'ext': 'mp3',
+            'title': 'Shadows Over Transylvania',
+            'release_date': '19970225',
+            'creator': 'Dark Funeral',
+            'description': 'md5:c9b20210587cbcd6836a1c597bab4525',
+            'thumbnail': r're:^https?://.*\.jpg(?:$|[#?])',
+        },
+        'params': {'skip_download': True},
+        'skip': 'no longer available',
+    }]
+
+    _FORMATS = {
+        'F000': {'name': 'flac', 'prefix': 'F000', 'ext': 'flac', 'preference': 60},
+        'A000': {'name': 'ape', 'prefix': 'A000', 'ext': 'ape', 'preference': 50},
+        'M800': {'name': '320mp3', 'prefix': 'M800', 'ext': 'mp3', 'preference': 40, 'abr': 320},
+        'M500': {'name': '128mp3', 'prefix': 'M500', 'ext': 'mp3', 'preference': 30, 'abr': 128},
+        'C400': {'name': '96aac', 'prefix': 'C400', 'ext': 'm4a', 'preference': 20, 'abr': 96},
+        'C200': {'name': '48aac', 'prefix': 'C200', 'ext': 'm4a', 'preference': 20, 'abr': 48},
     }
 
     def _real_extract(self, url):
         mid = self._match_id(url)
 
-        entries = self.get_entries_from_page(mid)
-        singer_page = self._download_webpage(url, mid, 'Download singer page')
-        singer_name = self._html_search_regex(
-            r"singername\s*:\s*'(.*?)'", singer_page, 'singer name', default=None)
-        singer_desc = None
+        init_data = self.download_init_data(url, mid, fatal=False)
+        info_data = self.make_fcu_req({'info': {
+            'module': 'music.pf_song_detail_svr',
+            'method': 'get_song_detail_yqq',
+            'param': {
+                'song_mid': mid,
+                'song_type': 0,
+            }
+        }}, mid, note='Downloading song info')['info']['data']['track_info']
 
-        if mid:
-            singer_desc_page = self._download_xml(
-                'http://s.plcloud.music.qq.com/fcgi-bin/fcg_get_singer_desc.fcg', mid,
-                'Donwload singer description XML',
-                query={'utf8': 1, 'outCharset': 'utf-8', 'format': 'xml', 'singermid': mid},
-                headers={'Referer': 'https://y.qq.com/n/yqq/singer/'})
+        media_mid = info_data['file']['media_mid']
 
-            singer_desc = singer_desc_page.find('./data/info/desc').text
+        data = self.make_fcu_req({
+            'req_1': {
+                'module': 'vkey.GetVkeyServer',
+                'method': 'CgiGetVkey',
+                'param': {
+                    'guid': str(self.m_r_get_ruin()),
+                    'songmid': [mid] * len(self._FORMATS),
+                    'songtype': [0] * len(self._FORMATS),
+                    'uin': str(self._get_uin()),
+                    'loginflag': 1,
+                    'platform': '20',
+                    'filename': [f'{f["prefix"]}{media_mid}.{f["ext"]}' for f in self._FORMATS.values()],
+                }
+            },
+            'req_2': {
+                'module': 'music.musichallSong.PlayLyricInfo',
+                'method': 'GetPlayLyricInfo',
+                'param': {'songMID': mid},
+            },
+        }, mid, note='Downloading formats and lyric')
 
-        return self.playlist_result(entries, mid, singer_name, singer_desc)
+        if data['req_1']['code'] != 0:
+            raise ExtractorError(f'Failed to download formats, error {data["req_1"]["code"]}')
+        formats = traverse_obj(data, ('req_1', 'data', 'midurlinfo', lambda _, v: v['songmid'] == mid and v['purl'], {
+            'url': ('purl', {str}, {lambda x: f'https://dl.stream.qqmusic.qq.com/{x}'}),
+            'format': ('filename', {lambda x: self._FORMATS[x[:4]]['name']}),
+            'format_id': ('filename', {lambda x: self._FORMATS[x[:4]]['name']}),
+            'size': ('filename', {lambda x: self._FORMATS[x[:4]]['name']},
+                     {lambda x: traverse_obj(info_data, ('file', f'size_{x}'), get_all=False)}),
+            'quality': ('filename', {lambda x: self._FORMATS[x[:4]]['preference']}),
+            'abr': ('filename', {lambda x: self._FORMATS[x[:4]]['abr']}),
+        }))
+
+        if traverse_obj(data, ('req_2', 'code')):
+            self.report_warning(f'Failed to download lyric, error {data["req_2"]["code"]}')
+        lrc_content = traverse_obj(data, ('req_2', 'data', 'lyric', {lambda x: base64.b64decode(x).decode('utf-8')}))
+
+        info_dict = {
+            'id': mid,
+            'formats': formats,
+            **traverse_obj(info_data, {
+                'title': ('title', {str}),
+                'album': ('album', 'title', {str}, {lambda x: x or None}),
+                'release_date': ('time_public', {lambda x: x.replace('-', '') or None}),
+                'creators': ('singer', ..., 'name', {str}),
+                'alt_title': ('subtitle', {str}, {lambda x: x or None}),
+                'duration': ('interval', {int_or_none}),
+            }),
+            **traverse_obj(init_data, ('detail', {
+                'thumbnail': ('picurl', {url_or_none}),
+                'description': ('info', 'intro', 'content', ..., 'value', {str}),
+                'genres': ('info', 'genre', 'content', ..., 'value', {str}, {lambda x: [x]}),
+            }), get_all=False),
+        }
+        if lrc_content:
+            info_dict['subtitles'] = {'origin': [{'ext': 'lrc', 'data': lrc_content}]}
+            info_dict['description'] = join_nonempty(info_dict.get('description'), lrc_content, delim='\n')
+        return info_dict
+
+
+class QQMusicSingerIE(QQMusicBaseIE):
+    IE_NAME = 'qqmusic:singer'
+    IE_DESC = 'QQ音乐 - 歌手'
+    _VALID_URL = r'https?://y\.qq\.com/n/ryqq/singer/(?P<id>[0-9A-Za-z]+)'
+    _TESTS = [{
+        'url': 'https://y.qq.com/n/ryqq/singer/001BLpXF2DyJe2',
+        'info_dict': {
+            'id': '001BLpXF2DyJe2',
+            'title': '林俊杰',
+            'description': 'md5:10624ce73b06fa400bc846f59b0305fa',
+            'thumbnail': r're:^https?://.*\.jpg(?:$|[#?])',
+        },
+        'playlist_mincount': 100,
+    }, {
+        'url': 'https://y.qq.com/n/ryqq/singer/000Q00f213YzNV',
+        'info_dict': {
+            'id': '000Q00f213YzNV',
+            'title': '桃几OvO',
+            'description': '小破站小唱见~希望大家喜欢听我唱歌~！',
+            'thumbnail': r're:^https?://.*\.jpg(?:$|[#?])',
+        },
+        'playlist_count': 12,
+        'playlist': [{
+            'info_dict': {
+                'id': '0016cvsy02mmCl',
+                'ext': 'mp3',
+                'title': '群青',
+                'album': '桃几2021年翻唱集',
+                'release_date': '20210913',
+                'duration': 248,
+                'creators': ['桃几OvO'],
+                'genres': ['Pop'],
+                'description': 'md5:4296005a04edcb5cdbe0889d5055a7ae',
+                'size': 3970822,
+                'thumbnail': r're:^https?://.*\.jpg(?:$|[#?])',
+            },
+        }],
+    }]
+
+    def _entries(self, mid, init_data):
+        page_size = 50
+        max_num = traverse_obj(init_data, ('singerDetail', 'songTotalNum'))
+        for page in range(0, max_num // page_size + 1):
+            data = self.make_fcu_req({'req_1': {
+                'module': 'music.web_singer_info_svr',
+                'method': 'get_singer_detail_info',
+                'param': {
+                    'sort': 5,
+                    'singermid': mid,
+                    'sin': page * page_size,
+                    'num': page_size,
+                }}}, mid, note=f'Downloading page {page}')
+            yield from traverse_obj(data, ('req_1', 'data', 'songlist', ..., {lambda x: self.url_result(
+                f'https://y.qq.com/n/ryqq/songDetail/{x["mid"]}', QQMusicIE, x['mid'], x.get('title'))}))
+
+    def _real_extract(self, url):
+        mid = self._match_id(url)
+
+        init_data = self.download_init_data(url, mid)
+        return self.playlist_result(
+            self._entries(mid, init_data), mid, **traverse_obj(init_data, ('singerDetail', {
+                'title': ('basic_info', 'name', {str}),
+                'description': ('ex_info', 'desc', {str}),
+                'thumbnail': ('pic', 'pic', {url_or_none}),
+            })))
+
+
+class QQPlaylistBaseIE(InfoExtractor):
+    def _extract_entries(self, info_json, path):
+        return traverse_obj(info_json, (*path, {lambda song: self.url_result(
+            f'https://y.qq.com/n/ryqq/songDetail/{song["songmid"]}', QQMusicIE, song['songmid'], song['songname'])}))
 
 
 class QQMusicAlbumIE(QQPlaylistBaseIE):
     IE_NAME = 'qqmusic:album'
     IE_DESC = 'QQ音乐 - 专辑'
-    _VALID_URL = r'https?://y\.qq\.com/n/yqq/album/(?P<id>[0-9A-Za-z]+)\.html'
+    _VALID_URL = r'https?://y\.qq\.com/n/ryqq/albumDetail/(?P<id>[0-9A-Za-z]+)'
 
     _TESTS = [{
-        'url': 'https://y.qq.com/n/yqq/album/000gXCTb2AhRR1.html',
+        'url': 'https://y.qq.com/n/ryqq/albumDetail/000gXCTb2AhRR1',
         'info_dict': {
             'id': '000gXCTb2AhRR1',
             'title': '我们都是这样长大的',
@@ -237,10 +296,10 @@ class QQMusicAlbumIE(QQPlaylistBaseIE):
         },
         'playlist_count': 4,
     }, {
-        'url': 'https://y.qq.com/n/yqq/album/002Y5a3b3AlCu3.html',
+        'url': 'https://y.qq.com/n/ryqq/albumDetail/002Y5a3b3AlCu3',
         'info_dict': {
             'id': '002Y5a3b3AlCu3',
-            'title': '그리고...',
+            'title': '그리고…',
             'description': 'md5:a48823755615508a95080e81b51ba729',
         },
         'playlist_count': 8,
@@ -249,49 +308,45 @@ class QQMusicAlbumIE(QQPlaylistBaseIE):
     def _real_extract(self, url):
         mid = self._match_id(url)
 
-        album = self._download_json(
-            'http://i.y.qq.com/v8/fcg-bin/fcg_v8_album_info_cp.fcg?albummid=%s&format=json' % mid,
-            mid, 'Download album page')['data']
+        album_json = self._download_json(
+            'http://i.y.qq.com/v8/fcg-bin/fcg_v8_album_info_cp.fcg',
+            mid, 'Download album page',
+            query={'albummid': mid, 'format': 'json'})['data']
 
-        entries = [
-            self.url_result(
-                'https://y.qq.com/n/yqq/song/' + song['songmid'] + '.html', 'QQMusic', song['songmid']
-            ) for song in album['list']
-        ]
-        album_name = album.get('name')
-        album_detail = album.get('desc')
-        if album_detail is not None:
-            album_detail = album_detail.strip()
+        entries = self._extract_entries(album_json, ('list', ...))
 
-        return self.playlist_result(entries, mid, album_name, album_detail)
+        return self.playlist_result(entries, mid, **traverse_obj(album_json, {
+            'title': ('name', {str}),
+            'description': ('desc', {str}, {str.strip}),
+        }))
 
 
 class QQMusicToplistIE(QQPlaylistBaseIE):
     IE_NAME = 'qqmusic:toplist'
     IE_DESC = 'QQ音乐 - 排行榜'
-    _VALID_URL = r'https?://y\.qq\.com/n/yqq/toplist/(?P<id>[0-9]+)\.html'
+    _VALID_URL = r'https?://y\.qq\.com/n/ryqq/toplist/(?P<id>[0-9]+)'
 
     _TESTS = [{
-        'url': 'https://y.qq.com/n/yqq/toplist/123.html',
+        'url': 'https://y.qq.com/n/ryqq/toplist/123',
         'info_dict': {
             'id': '123',
-            'title': '美国iTunes榜',
-            'description': 'md5:89db2335fdbb10678dee2d43fe9aba08',
+            'title': r're:美国热门音乐榜 \d{4}-\d{2}-\d{2}',
+            'description': '美国热门音乐榜，每周一更新。',
         },
-        'playlist_count': 100,
+        'playlist_count': 95,
     }, {
-        'url': 'https://y.qq.com/n/yqq/toplist/3.html',
+        'url': 'https://y.qq.com/n/ryqq/toplist/3',
         'info_dict': {
             'id': '3',
-            'title': '巅峰榜·欧美',
-            'description': 'md5:5a600d42c01696b26b71f8c4d43407da',
+            'title': r're:巅峰榜·欧美 \d{4}-\d{2}-\d{2}',
+            'description': 'md5:4def03b60d3644be4c9a36f21fd33857',
         },
         'playlist_count': 100,
     }, {
-        'url': 'https://y.qq.com/n/yqq/toplist/106.html',
+        'url': 'https://y.qq.com/n/ryqq/toplist/106',
         'info_dict': {
             'id': '106',
-            'title': '韩国Mnet榜',
+            'title': r're:韩国Mnet榜 \d{4}-\d{2}-\d{2}',
             'description': 'md5:cb84b325215e1d21708c615cac82a6e7',
         },
         'playlist_count': 50,
@@ -305,33 +360,21 @@ class QQMusicToplistIE(QQPlaylistBaseIE):
             note='Download toplist page',
             query={'type': 'toplist', 'topid': list_id, 'format': 'json'})
 
-        entries = [self.url_result(
-            'https://y.qq.com/n/yqq/song/' + song['data']['songmid'] + '.html', 'QQMusic',
-            song['data']['songmid'])
-            for song in toplist_json['songlist']]
+        entries = self._extract_entries(toplist_json, ('songlist', ..., 'data'))
+
+        list_name = join_nonempty(*traverse_obj(toplist_json, ((('topinfo', 'ListName'), 'update_time'),)), delim=' ')
+        list_description = traverse_obj(toplist_json, ('topinfo', 'info'))
 
-        topinfo = toplist_json.get('topinfo', {})
-        list_name = topinfo.get('ListName')
-        list_description = topinfo.get('info')
         return self.playlist_result(entries, list_id, list_name, list_description)
 
 
 class QQMusicPlaylistIE(QQPlaylistBaseIE):
     IE_NAME = 'qqmusic:playlist'
     IE_DESC = 'QQ音乐 - 歌单'
-    _VALID_URL = r'https?://y\.qq\.com/n/yqq/playlist/(?P<id>[0-9]+)\.html'
+    _VALID_URL = r'https?://y\.qq\.com/n/ryqq/playlist/(?P<id>[0-9]+)'
 
     _TESTS = [{
-        'url': 'http://y.qq.com/n/yqq/playlist/3462654915.html',
-        'info_dict': {
-            'id': '3462654915',
-            'title': '韩国5月新歌精选下旬',
-            'description': 'md5:d2c9d758a96b9888cf4fe82f603121d4',
-        },
-        'playlist_count': 40,
-        'skip': 'playlist gone',
-    }, {
-        'url': 'https://y.qq.com/n/yqq/playlist/1374105607.html',
+        'url': 'https://y.qq.com/n/ryqq/playlist/1374105607',
         'info_dict': {
             'id': '1374105607',
             'title': '易入人心的华语民谣',
@@ -347,19 +390,84 @@ class QQMusicPlaylistIE(QQPlaylistBaseIE):
             'http://i.y.qq.com/qzone-music/fcg-bin/fcg_ucc_getcdinfo_byids_cp.fcg',
             list_id, 'Download list page',
             query={'type': 1, 'json': 1, 'utf8': 1, 'onlysong': 0, 'disstid': list_id},
-            transform_source=strip_jsonp)
+            transform_source=strip_jsonp, headers={'Referer': url})
         if not len(list_json.get('cdlist', [])):
-            if list_json.get('code'):
-                raise ExtractorError(
-                    'QQ Music said: error %d in fetching playlist info' % list_json['code'],
-                    expected=True)
-            raise ExtractorError('Unable to get playlist info')
+            error_msg = ''
+            if list_json.get('code') or list_json.get('subcode'):
+                error_msg = f': Error {list_json.get("code")}-{list_json["subcode"]}: {list_json.get("msg", "")}'
+            raise ExtractorError(f'Unable to get playlist info{error_msg}')
 
-        cdlist = list_json['cdlist'][0]
-        entries = [self.url_result(
-            'https://y.qq.com/n/yqq/song/' + song['songmid'] + '.html', 'QQMusic', song['songmid'])
-            for song in cdlist['songlist']]
+        entries = self._extract_entries(list_json, ('cdlist', 0, 'songlist', ...))
 
-        list_name = cdlist.get('dissname')
-        list_description = clean_html(unescapeHTML(cdlist.get('desc')))
-        return self.playlist_result(entries, list_id, list_name, list_description)
+        return self.playlist_result(entries, list_id, **traverse_obj(list_json, ('cdlist', 0, {
+            'title': ('dissname', {str}),
+            'description': ('desc', {lambda x: clean_html(unescapeHTML(x))}),
+        })))
+
+
+class QQMusicVideoIE(QQMusicBaseIE):
+    IE_NAME = 'qqmusic:mv'
+    IE_DESC = 'QQ音乐 - MV'
+    _VALID_URL = r'https?://y\.qq\.com/n/ryqq/mv/(?P<id>[0-9A-Za-z]+)'
+
+    _TESTS = [{
+        'url': 'https://y.qq.com/n/ryqq/mv/002Vsarh3SVU8K',
+        'info_dict': {
+            'id': '002Vsarh3SVU8K',
+            'ext': 'mp4',
+            'title': 'The Chant (Extended Mix / Audio)',
+            'description': '',
+            'thumbnail': r're:^https?://.*\.jpg(?:$|[#?])',
+            'release_timestamp': 1688918400,
+            'release_date': '20230709',
+            'duration': 313,
+            'creators': ['Duke Dumont'],
+            'view_count': int,
+        },
+    }]
+
+    def _parse_url_formats(self, url_data):
+        formats = traverse_obj(url_data, ('mp4', lambda _, v: v.get('freeflow_url'), {
+            'url': ('freeflow_url', 0, {url_or_none}),
+            'filesize': ('fileSize', {int_or_none}),
+            'format_id': ('newFileType', {str_or_none}),
+        }))
+        return formats
+
+    def _real_extract(self, url):
+        video_id = self._match_id(url)
+
+        video_info = self.make_fcu_req({
+            'mvInfo': {
+                'module': 'music.video.VideoData',
+                'method': 'get_video_info_batch',
+                'param': {
+                    'vidlist': [video_id],
+                    'required': [
+                        'vid', 'type', 'sid', 'cover_pic', 'duration', 'singers',
+                        'video_pay', 'hint', 'code', 'msg', 'name', 'desc',
+                        'playcnt', 'pubdate', 'play_forbid_reason']
+                }
+            },
+            'mvUrl': {
+                'module': 'music.stream.MvUrlProxy',
+                'method': 'GetMvUrls',
+                'param': {'vids': [video_id]}
+            }
+        }, video_id)
+        if traverse_obj(video_info, ('mvInfo', 'data', video_id, 'play_forbid_reason')) == 3:
+            self.raise_geo_restricted()
+
+        return {
+            'id': video_id,
+            'formats': self._parse_url_formats(traverse_obj(video_info, ('mvUrl', 'data', video_id))),
+            **traverse_obj(video_info, ('mvInfo', 'data', video_id, {
+                'title': ('name', {str}),
+                'description': ('desc', {str}),
+                'thumbnail': ('cover_pic', {url_or_none}),
+                'release_timestamp': ('pubdate', {int_or_none}),
+                'duration': ('duration', {int_or_none}),
+                'creators': ('singers', ..., 'name', {str}),
+                'view_count': ('playcnt', {int_or_none}),
+            })),
+        }

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()