[TVer] Fix extractor (#3268)

Authored by: Lesmiscore

yt_dlp/extractor/tver.py

@@ -1,77 +1,94 @@
 # coding: utf-8
 from __future__ import unicode_literals
 
-
 from .common import InfoExtractor
-from ..compat import compat_str
 from ..utils import (
     ExtractorError,
-    int_or_none,
-    remove_start,
     smuggle_url,
+    str_or_none,
     traverse_obj,
 )
 
 
 class TVerIE(InfoExtractor):
-    _VALID_URL = r'https?://(?:www\.)?tver\.jp/(?P<path>corner|episode|feature|lp|tokyo2020/video)/(?P<id>[fc]?\d+)'
+    _VALID_URL = r'https?://(?:www\.)?tver\.jp/(?:(?P<type>lp|corner|series|episodes?|feature|tokyo2020/video)/)+(?P<id>[a-zA-Z0-9]+)'
-    # videos are only available for 7 days
     _TESTS = [{
-        'url': 'https://tver.jp/corner/f0062178',
+        'skip': 'videos are only available for 7 days',
+        'url': 'https://tver.jp/episodes/ephss8yveb',
+        'info_dict': {
+            'title': '#44　料理と値段と店主にびっくり　オモてなしすぎウマい店　2時間SP',
+            'description': 'md5:66985373a66fed8ad3cd595a3cfebb13',
+        },
+        'add_ie': ['BrightcoveNew'],
+    }, {
+        'skip': 'videos are only available for 7 days',
+        'url': 'https://tver.jp/lp/episodes/ep6f16g26p',
+        'info_dict': {
+            # sorry but this is "correct"
+            'title': '4月11日(月)23時06分 ~ 放送予定',
+            'description': 'md5:4029cc5f4b1e8090dfc5b7bd2bc5cd0b',
+        },
+        'add_ie': ['BrightcoveNew'],
+    }, {
+        'url': 'https://tver.jp/corner/f0103888',
         'only_matching': True,
     }, {
-        'url': 'https://tver.jp/feature/f0062413',
+        'url': 'https://tver.jp/lp/f0033031',
-        'only_matching': True,
-    }, {
-        'url': 'https://tver.jp/episode/79622438',
-        'only_matching': True,
-    }, {
-        # subtitle = ' '
-        'url': 'https://tver.jp/corner/f0068870',
-        'only_matching': True,
-    }, {
-        'url': 'https://tver.jp/lp/f0009694',
-        'only_matching': True,
-    }, {
-        'url': 'https://tver.jp/lp/c0000239',
-        'only_matching': True,
-    }, {
-        'url': 'https://tver.jp/tokyo2020/video/6264525510001',
         'only_matching': True,
     }]
-    _TOKEN = None
     BRIGHTCOVE_URL_TEMPLATE = 'http://players.brightcove.net/%s/default_default/index.html?videoId=%s'
+    _PLATFORM_UID = None
+    _PLATFORM_TOKEN = None
 
     def _real_initialize(self):
-        self._TOKEN = self._download_json(
+        create_response = self._download_json(
-            'https://tver.jp/api/access_token.php', None)['token']
+            'https://platform-api.tver.jp/v2/api/platform_users/browser/create', None,
+            note='Creating session', data=b'device_type=pc', headers={
+                'Origin': 'https://s.tver.jp',
+                'Referer': 'https://s.tver.jp/',
+                'Content-Type': 'application/x-www-form-urlencoded',
+            })
+        self._PLATFORM_UID = traverse_obj(create_response, ('result', 'platform_uid'))
+        self._PLATFORM_TOKEN = traverse_obj(create_response, ('result', 'platform_token'))
 
     def _real_extract(self, url):
-        path, video_id = self._match_valid_url(url).groups()
+        video_id, video_type = self._match_valid_url(url).group('id', 'type')
-        if path == 'lp':
+        if video_type not in {'series', 'episodes'}:
-            webpage = self._download_webpage(url, video_id)
+            webpage = self._download_webpage(url, video_id, note='Resolving to new URL')
-            redirect_path = self._search_regex(r'to_href="([^"]+)', webpage, 'redirect path')
+            video_id = self._match_id(self._search_regex(
-            path, video_id = self._match_valid_url(f'https://tver.jp{redirect_path}').groups()
+                (r'canonical"\s*href="(https?://tver\.jp/[^"]+)"', r'&link=(https?://tver\.jp/[^?&]+)[?&]'),
-        api_response = self._download_json(f'https://api.tver.jp/v4/{path}/{video_id}', video_id, query={'token': self._TOKEN})
+                webpage, 'url regex'))
-        p_id = traverse_obj(api_response, ('main', 'publisher_id'))
+        video_info = self._download_json(
-        if not p_id:
+            f'https://statics.tver.jp/content/episode/{video_id}.json', video_id,
-            error_msg, expected = traverse_obj(api_response, ('episode', 0, 'textbar', 0, ('text', 'longer')), get_all=False), True
+            query={'v': '5'}, headers={
-            if not error_msg:
+                'Origin': 'https://tver.jp',
-                error_msg, expected = 'Failed to extract publisher ID', False
+                'Referer': 'https://tver.jp/',
-            raise ExtractorError(error_msg, expected=expected)
+            })
-        service = remove_start(traverse_obj(api_response, ('main', 'service')), 'ts_')
+        p_id = video_info['video']['accountID']
+        r_id = traverse_obj(video_info, ('video', ('videoRefID', 'videoID')), get_all=False)
+        if not r_id:
+            raise ExtractorError('Failed to extract reference ID for Brightcove')
+        if not r_id.isdigit():
+            r_id = f'ref:{r_id}'
 
-        r_id = traverse_obj(api_response, ('main', 'reference_id'))
+        additional_info = self._download_json(
-        if service not in ('tx', 'russia2018', 'sebare2018live', 'gorin'):
+            f'https://platform-api.tver.jp/service/api/v1/callEpisode/{video_id}?require_data=mylist,later[epefy106ur],good[epefy106ur],resume[epefy106ur]',
-            r_id = 'ref:' + r_id
+            video_id, fatal=False,
-        bc_url = smuggle_url(
+            query={
-            self.BRIGHTCOVE_URL_TEMPLATE % (p_id, r_id),
+                'platform_uid': self._PLATFORM_UID,
-            {'geo_countries': ['JP']})
+                'platform_token': self._PLATFORM_TOKEN,
+            }, headers={
+                'x-tver-platform-type': 'web'
+            })
 
         return {
             '_type': 'url_transparent',
-            'description': traverse_obj(api_response, ('main', 'note', 0, 'text'), expected_type=compat_str),
+            'title': str_or_none(video_info.get('title')),
-            'episode_number': int_or_none(traverse_obj(api_response, ('main', 'ext', 'episode_number'), expected_type=compat_str)),
+            'description': str_or_none(video_info.get('description')),
-            'url': bc_url,
+            'url': smuggle_url(
+                self.BRIGHTCOVE_URL_TEMPLATE % (p_id, r_id), {'geo_countries': ['JP']}),
+            'series': traverse_obj(
+                additional_info, ('result', ('episode', 'series'), 'content', ('seriesTitle', 'title')),
+                get_all=False),
             'ie_key': 'BrightcoveNew',
         }