[extractor/Douyin] fix Douyin

yt_dlp/extractor/tiktok.py

@@ -6,7 +6,7 @@
 import time
 
 from .common import InfoExtractor
-from ..compat import compat_urllib_parse_unquote, compat_urllib_parse_urlparse
+from ..compat import compat_urllib_parse_urlparse
 from ..networking import HEADRequest
 from ..utils import (
     ExtractorError,
@@ -15,7 +15,6 @@
     UserNotLive,
     determine_ext,
     format_field,
-    get_first,
     int_or_none,
     join_nonempty,
     merge_dicts,
@@ -242,7 +241,7 @@ def extract_addr(addr, add_meta={}):
                 'format_id': 'play_addr',
                 'format_note': 'Direct video',
                 'vcodec': 'h265' if traverse_obj(
-                    video_info, 'is_bytevc1', 'is_h265') else 'h264',  # TODO: Check for "direct iOS" videos, like https://www.tiktok.com/@cookierun_dev/video/7039716639834656002
+                    video_info, 'is_bytevc1', 'is_h265') else 'h264', # TODO: Check for "direct iOS" videos, like https://www.tiktok.com/@cookierun_dev/video/7039716639834656002
                 'width': video_info.get('width'),
                 'height': video_info.get('height'),
             }))
@@ -767,7 +766,7 @@ def _video_entries_api(self, webpage, user_id, username):
             'max_cursor': 0,
             'min_cursor': 0,
             'retry_type': 'no_retry',
-            'device_id': ''.join(random.choices(string.digits, k=19)),  # Some endpoints don't like randomized device_id, so it isn't directly set in _call_api.
+            'device_id': ''.join(random.choices(string.digits, k=19)), # Some endpoints don't like randomized device_id, so it isn't directly set in _call_api.
         }
 
         for page in itertools.count(1):
@@ -1035,24 +1034,17 @@ class DouyinIE(TikTokBaseIE):
     def _real_extract(self, url):
         video_id = self._match_id(url)
 
-        try:
-            return self._extract_aweme_app(video_id)
-        except ExtractorError as e:
-            e.expected = True
-            self.to_screen(f'{e}; trying with webpage')
-
-        webpage = self._download_webpage(url, video_id)
-        render_data = self._search_json(
-            r'<script [^>]*\bid=[\'"]RENDER_DATA[\'"][^>]*>', webpage, 'render data', video_id,
-            contains_pattern=r'%7B(?s:.+)%7D', fatal=False, transform_source=compat_urllib_parse_unquote)
-        if not render_data:
+        detail = traverse_obj(self._download_json(
+            'https://www.douyin.com/aweme/v1/web/aweme/detail/', video_id,
+            'Downloading web detail JSON', 'Failed to download web detail JSON',
+            query={'aweme_id': video_id}, fatal=False), ('aweme_detail', {dict}))
+        if not detail:
             # TODO: Run verification challenge code to generate signature cookies
-            cookies = self._get_cookies(self._WEBPAGE_HOST)
-            expected = not cookies.get('s_v_web_id') or not cookies.get('ttwid')
             raise ExtractorError(
-                'Fresh cookies (not necessarily logged in) are needed', expected=expected)
+                'Fresh cookies (not necessarily logged in) are needed',
+                expected=not self._get_cookies(self._WEBPAGE_HOST).get('s_v_web_id'))
 
-        return self._parse_aweme_video_web(get_first(render_data, ('aweme', 'detail')), url, video_id)
+        return self._parse_aweme_video_app(detail)
 
 
 class TikTokVMIE(InfoExtractor):