[ntvcojp] fix extractor

yt_dlp/extractor/ntvcojp.py

@@ -1,9 +1,13 @@
 from .common import InfoExtractor
+from ..networking.exceptions import HTTPError
 from ..utils import (
     ExtractorError,
-    smuggle_url,
+    clean_html,
-    traverse_obj,
+    int_or_none,
+    unified_timestamp,
+    urljoin,
 )
+from ..utils.traversal import find_element, traverse_obj
 
 
 class NTVCoJpCUIE(InfoExtractor):
@@ -11,45 +15,70 @@ class NTVCoJpCUIE(InfoExtractor):
     IE_DESC = 'Nippon Television Network'
     _VALID_URL = r'https?://cu\.ntv\.co\.jp/(?!program)(?P<id>[^/?&#]+)'
     _TEST = {
-        'url': 'https://cu.ntv.co.jp/televiva-chill-gohan_181031/',
+        'url': 'https://cu.ntv.co.jp/gaki_20250525/',
         'info_dict': {
-            'id': '5978891207001',
+            'title': '放送開始36年!方正ココリコが選ぶ神回&地獄回!',
+            'id': 'gaki_20250525',
             'ext': 'mp4',
-            'title': '桜エビと炒り卵がポイント！ 「中華風 エビチリおにぎり」──『美虎』五十嵐美幸',
+            'categories': ['ダウンタウンのガキの使いやあらへんで！'],
-            'upload_date': '20181213',
+            'description': '神回地獄回座談会!レギュラー放送1756回の中からココリコと方正が神回と地獄回をそれぞれ選んで発表!若手時代の遠藤がガキ使メンバーに振り回される!?田中が好きな懐かしの番組名物キャラに爆笑!?方正が思い出に残っている持ち込み回とは?笑ってはいけないシリーズから遠藤が大汗をかくほど追い詰められる企画が誕生していた!?3人のトラウマになっている過酷罰ゲームを振り返り!方正記念企画のはずがまさかの展開で涙!?',
-            'description': 'md5:1985b51a9abc285df0104d982a325f2a',
+            'timestamp': 1748145124,
-            'uploader_id': '3855502814001',
+            'release_timestamp': 1748145539,
-            'timestamp': 1544669941,
+            'duration': 1450,
-        },
+            'episode_number': 255,
-        'params': {
+            'episode': '放送開始36年!方正ココリコが選ぶ神回&地獄回!',
-            # m3u8 download
+            'upload_date': '20250525',
-            'skip_download': True,
+            'release_date': '20250525',
         },
     }
 
-    BRIGHTCOVE_URL_TEMPLATE = 'http://players.brightcove.net/%s/default_default/index.html?videoId=%s'
-
     def _real_extract(self, url):
-        display_id = self._match_id(url)
+        video_id = self._match_id(url)
-        webpage = self._download_webpage(url, display_id)
+        webpage = self._download_webpage(url, video_id)
-        player_config = self._search_nuxt_data(webpage, display_id)
+        meta = self._search_json(r'window\.app\s*=', webpage, 'episode info', video_id, fatal=False)
-        video_id = traverse_obj(player_config, ('movie', 'video_id'))
+        episode = traverse_obj(meta, ('falcorCache', 'catalog', 'episode', video_id, 'value'))
-        if not video_id:
+
-            raise ExtractorError('Failed to extract video ID for Brightcove')
+        nt_path = self._search_regex(r'<script[^>]+src=["\'](/assets/nt\.[^"\']+\.js)["\']', webpage, 'stream API config')
-        account_id = traverse_obj(player_config, ('player', 'account')) or '3855502814001'
+        nt_js = self._download_webpage(urljoin(url, nt_path), video_id, note='Downloading stream API config')
-        title = traverse_obj(player_config, ('movie', 'name'))
+        video_url = self._search_regex(r'videoPlaybackUrl:\s*[\'"]([^\'"]+)[\'"]', nt_js, 'stream API url')
-        if not title:
+        api_key = self._search_regex(r'api_key:\s*[\'"]([^\'"]+)[\'"]', nt_js, 'stream API key')
-            og_title = self._og_search_title(webpage, fatal=False) or traverse_obj(player_config, ('player', 'title'))
+
-            if og_title:
+        try:
-                title = og_title.split('(', 1)[0].strip()
+            source_meta = self._download_json(
-        description = (traverse_obj(player_config, ('movie', 'description'))
+                f'{video_url}ref:{video_id}',
-                       or self._html_search_meta(['description', 'og:description'], webpage))
+                video_id,
+                headers={'X-Streaks-Api-Key': api_key},
+                note='Downloading stream metadata',
+            )
+        except ExtractorError as e:
+            if isinstance(e.cause, HTTPError) and e.cause.status == 403:
+                self.raise_geo_restricted(countries=['JP'])
+            raise
+
+        formats, subtitles = [], {}
+        for src in traverse_obj(source_meta, ('sources', ..., 'src')):
+            fmts, subs = self._extract_m3u8_formats_and_subtitles(src, video_id, fatal=False)
+            formats.extend(fmts)
+            self._merge_subtitles(subs, target=subtitles)
+
         return {
-            '_type': 'url_transparent',
+            'title': traverse_obj(webpage, ({find_element(tag='h3')}, {clean_html})),
             'id': video_id,
-            'display_id': display_id,
+            **traverse_obj(
-            'title': title,
+                episode,
-            'description': description,
+                {
-            'url': smuggle_url(self.BRIGHTCOVE_URL_TEMPLATE % (account_id, video_id), {'geo_countries': ['JP']}),
+                    'categories': ('keywords', {list}),
-            'ie_key': 'BrightcoveNew',
+                    'id': ('content_id', {str}),
+                    'description': ('description', 0, 'value'),
+                    'timestamp': ('created_at', {unified_timestamp}),
+                    'release_timestamp': ('pub_date', {unified_timestamp}),
+                    'duration': ('tv_episode_info', 'duration', {int_or_none}),
+                    'episode_number': ('tv_episode_info', 'episode_number', {int_or_none}),
+                    'episode': ('title', lambda _, v: not v.get('is_phonetic'), 'value'),
+                    'series': ('custom_data', 'program_name'),
+                },
+                get_all=False,
+            ),
+            'formats': formats,
+            'subtitles': subtitles,
         }