Merge 7fa603bcc3 into f2a4983df7

yt_dlp/extractor/litv.py

@@ -1,4 +1,3 @@
-import functools
 import json
 import uuid
 
@@ -6,12 +5,10 @@ from .common import InfoExtractor
 from ..utils import (
     ExtractorError,
     int_or_none,
-    join_nonempty,
     smuggle_url,
     traverse_obj,
     try_call,
     unsmuggle_url,
-    urljoin,
 )
 
 
@@ -73,52 +70,71 @@ class LiTVIE(InfoExtractor):
 
         webpage = self._download_webpage(url, video_id)
 
-        vod_data = self._search_nextjs_data(webpage, video_id, default={})
+        vod_data = self._parse_json(self._search_regex(
+            r'<script\s+id="__NEXT_DATA__"[^>]*>(.+)</script>', webpage, 'VOD data', default='{}'),
+            video_id)
 
-        program_info = traverse_obj(vod_data, ('props', 'pageProps', 'programInformation', {dict})) or {}
+        program_info = traverse_obj(
+            vod_data,
+            ('props', 'pageProps', 'programInformation'),
+            default={}, expected_type=dict)
+        series_id = program_info.get('series_id')
         playlist_data = traverse_obj(vod_data, ('props', 'pageProps', 'seriesTree'))
-        if playlist_data is not None and self._yes_playlist(program_info.get('series_id'), video_id, smuggled_data):
+        if playlist_data is not None and self._yes_playlist(series_id, video_id, smuggled_data):
-            return self._extract_playlist(playlist_data, program_info.get('content_type'))
+            content_type = program_info.get('content_type')
+            return self._extract_playlist(playlist_data, content_type)
 
-        asset_id = traverse_obj(program_info, ('assets', 0, 'asset_id', {str}))
+        asset_id = traverse_obj(program_info, ('assets', 0, 'asset_id'))
-        if asset_id:  # This is a live stream
+        if asset_id is None:  # live stream case
+            asset_id = program_info.get('content_id')
+            media_type = program_info.get('content_type')
+        else:  # vod case
             media_type = 'vod'
-        else:  # This is a VOD
-            asset_id = program_info['content_id']
-            media_type = program_info['content_type']
         puid = try_call(lambda: self._get_cookies('https://www.litv.tv/')['PUID'].value)
-        if puid:
+        if puid is None:
-            endpoint = 'get-urls'
-        else:
             puid = str(uuid.uuid4())
             endpoint = 'get-urls-no-auth'
+        else:
+            endpoint = 'get-urls'
+        payload = {'AssetId': asset_id, 'MediaType': media_type, 'puid': puid}
         video_data = self._download_json(
             f'https://www.litv.tv/api/{endpoint}', video_id,
-            data=json.dumps({'AssetId': asset_id, 'MediaType': media_type, 'puid': puid}).encode(),
+            data=json.dumps(payload).encode(),
             headers={'Content-Type': 'application/json'})
 
-        if error := traverse_obj(video_data, ('error', {dict})):
+        if video_data.get('error'):
-            error_msg = traverse_obj(error, ('message', {str}))
+            error_msg = traverse_obj(video_data, ('error', 'message'))
-            if error_msg and 'OutsideRegionError' in error_msg:
+            if error_msg == 'OutsideRegionError: ':
                 self.raise_geo_restricted('This video is available in Taiwan only')
-            elif error_msg:
+            if error_msg:
                 raise ExtractorError(f'{self.IE_NAME} said: {error_msg}', expected=True)
             raise ExtractorError(f'Unexpected error from {self.IE_NAME}')
 
+        try:
+            asset_url = video_data['result']['AssetURLs'][0]
+        except KeyError:
+            raise ExtractorError(f'Unexpected result from {self.IE_NAME}')
         formats = self._extract_m3u8_formats(
-            video_data['result']['AssetURLs'][0], video_id, ext='mp4', m3u8_id='hls')
+            asset_url, video_id, ext='mp4',
+            entry_protocol='m3u8_native', m3u8_id='hls')
         for a_format in formats:
             # LiTV HLS segments doesn't like compressions
             a_format.setdefault('http_headers', {})['Accept-Encoding'] = 'identity'
 
+        title = program_info['title'] + program_info.get('secondary_mark', '')
+        description = program_info.get('description')
+        thumbnail = program_info.get('picture')
+        if thumbnail is not None:
+            thumbnail = 'https://p-cdnstatic.svc.litv.tv/' + thumbnail
+        categories = [item['name'] for item in program_info.get('genres', [])]
+        episode = int_or_none(program_info.get('episode'))
+
         return {
             'id': video_id,
             'formats': formats,
-            'title': join_nonempty('title', 'secondary_mark', delim='', from_dict=program_info),
+            'title': title,
-            **traverse_obj(program_info, {
+            'description': description,
-                'description': ('description', {str}),
+            'thumbnail': thumbnail,
-                'thumbnail': ('picture', {functools.partial(urljoin, 'https://p-cdnstatic.svc.litv.tv/')}),
+            'categories': categories,
-                'categories': ('genres', ..., 'name', {str}),
+            'episode_number': episode,
-                'episode_number': ('episode', {int_or_none}),
-            }),
         }