[ie/tver:olympic] Add extractor (#15885)

Authored by: doe1080

yt_dlp/extractor/_extractors.py

@@ -2180,7 +2180,10 @@ from .tvc import (
     TVCIE,
     TVCArticleIE,
 )
-from .tver import TVerIE
+from .tver import (
+    TVerIE,
+    TVerOlympicIE,
+)
 from .tvigle import TvigleIE
 from .tviplayer import TVIPlayerIE
 from .tvn24 import TVN24IE

yt_dlp/extractor/streaks.py

@@ -22,7 +22,7 @@ class StreaksBaseIE(InfoExtractor):
     _GEO_BYPASS = False
     _GEO_COUNTRIES = ['JP']
 
-    def _extract_from_streaks_api(self, project_id, media_id, headers=None, query=None, ssai=False):
+    def _extract_from_streaks_api(self, project_id, media_id, headers=None, query=None, ssai=False, live_from_start=False):
         try:
             response = self._download_json(
                 self._API_URL_TEMPLATE.format('playback', project_id, media_id, ''),
@@ -83,6 +83,10 @@ class StreaksBaseIE(InfoExtractor):
 
             fmts, subs = self._extract_m3u8_formats_and_subtitles(
                 src_url, media_id, 'mp4', m3u8_id='hls', fatal=False, live=is_live, query=query)
+            for fmt in fmts:
+                if live_from_start:
+                    fmt.setdefault('downloader_options', {}).update({'ffmpeg_args': ['-live_start_index', '0']})
+                    fmt['is_from_start'] = True
             formats.extend(fmts)
             self._merge_subtitles(subs, target=subtitles)
 

yt_dlp/extractor/tver.py

@@ -4,6 +4,7 @@ from .streaks import StreaksBaseIE
 from ..utils import (
     ExtractorError,
     GeoRestrictedError,
+    clean_html,
     int_or_none,
     join_nonempty,
     make_archive_id,
@@ -11,7 +12,9 @@ from ..utils import (
     str_or_none,
     strip_or_none,
     time_seconds,
+    unified_timestamp,
     update_url_query,
+    url_or_none,
 )
 from ..utils.traversal import require, traverse_obj
 
@@ -257,3 +260,113 @@ class TVerIE(StreaksBaseIE):
             'id': video_id,
             '_old_archive_ids': [make_archive_id('BrightcoveNew', brightcove_id)] if brightcove_id else None,
         }
+
+
+class TVerOlympicIE(StreaksBaseIE):
+    IE_NAME = 'tver:olympic'
+
+    _API_BASE = 'https://olympic-data.tver.jp/api'
+    _VALID_URL = r'https?://(?:www\.)?tver\.jp/olympic/milanocortina2026/(?P<type>live|video)/play/(?P<id>\w+)'
+    _TESTS = [{
+        'url': 'https://tver.jp/olympic/milanocortina2026/video/play/3b1d4462150b42558d9cc8aabb5238d0/',
+        'info_dict': {
+            'id': '3b1d4462150b42558d9cc8aabb5238d0',
+            'ext': 'mp4',
+            'title': '【開会式】ぎゅっと凝縮ハイライト',
+            'display_id': 'ref:3b1d4462150b42558d9cc8aabb5238d0',
+            'duration': 712.045,
+            'live_status': 'not_live',
+            'modified_date': r're:\d{8}',
+            'modified_timestamp': int,
+            'tags': 'count:1',
+            'thumbnail': r're:https://.+\.(?:jpg|png)',
+            'timestamp': 1770420187,
+            'upload_date': '20260206',
+            'uploader_id': 'tver-olympic',
+        },
+    }, {
+        'url': 'https://tver.jp/olympic/milanocortina2026/live/play/glts313itwvj/',
+        'info_dict': {
+            'id': 'glts313itwvj',
+            'ext': 'mp4',
+            'title': '開会式ハイライト',
+            'channel_id': 'ntv',
+            'display_id': 'ref:sp_260207_spc_01_dvr',
+            'duration': 7680,
+            'live_status': 'was_live',
+            'modified_date': r're:\d{8}',
+            'modified_timestamp': int,
+            'thumbnail': r're:https://.+\.(?:jpg|png)',
+            'timestamp': 1770420300,
+            'upload_date': '20260206',
+            'uploader_id': 'tver-olympic-live',
+        },
+    }]
+
+    def _real_extract(self, url):
+        video_type, video_id = self._match_valid_url(url).group('type', 'id')
+        live_from_start = self.get_param('live_from_start')
+
+        if video_type == 'live':
+            project_id = 'tver-olympic-live'
+            api_key = 'a35ebb1ca7d443758dc7fcc5d99b1f72'
+            olympic_data = traverse_obj(self._download_json(
+                f'{self._API_BASE}/live/{video_id}', video_id), ('contents', 'live', {dict}))
+            media_id = traverse_obj(olympic_data, ('video_id', {str}))
+
+            now = time_seconds()
+            start_timestamp_str = traverse_obj(olympic_data, ('onair_start_date', {str}))
+            start_timestamp = unified_timestamp(start_timestamp_str, tz_offset=9)
+            if not start_timestamp:
+                raise ExtractorError('Unable to extract on-air start time')
+            end_timestamp = traverse_obj(olympic_data, (
+                'onair_end_date', {unified_timestamp(tz_offset=9)}, {require('on-air end time')}))
+
+            if now < start_timestamp:
+                self.raise_no_formats(
+                    f'This program is scheduled to start at {start_timestamp_str} JST', expected=True)
+
+                return {
+                    'id': video_id,
+                    'live_status': 'is_upcoming',
+                    'release_timestamp': start_timestamp,
+                }
+            elif start_timestamp <= now < end_timestamp:
+                live_status = 'is_live'
+                if live_from_start:
+                    media_id += '_dvr'
+            elif end_timestamp <= now:
+                dvr_end_timestamp = traverse_obj(olympic_data, (
+                    'dvr_end_date', {unified_timestamp(tz_offset=9)}))
+                if dvr_end_timestamp and now < dvr_end_timestamp:
+                    live_status = 'was_live'
+                    media_id += '_dvr'
+                else:
+                    raise ExtractorError(
+                        'This program is no longer available', expected=True)
+        else:
+            project_id = 'tver-olympic'
+            api_key = '4b55a4db3cce4ad38df6dd8543e3e46a'
+            media_id = video_id
+            live_status = 'not_live'
+            olympic_data = traverse_obj(self._download_json(
+                f'{self._API_BASE}/video/{video_id}', video_id), ('contents', 'video', {dict}))
+
+        return {
+            **self._extract_from_streaks_api(project_id, f'ref:{media_id}', {
+                'Origin': 'https://tver.jp',
+                'Referer': 'https://tver.jp/',
+                'X-Streaks-Api-Key': api_key,
+            }, live_from_start=live_from_start),
+            **traverse_obj(olympic_data, {
+                'title': ('title', {clean_html}, filter),
+                'alt_title': ('sub_title', {clean_html}, filter),
+                'channel': ('channel', {clean_html}, filter),
+                'channel_id': ('channel_id', {clean_html}, filter),
+                'description': (('description', 'description_l', 'description_s'), {clean_html}, filter, any),
+                'timestamp': ('onair_start_date', {unified_timestamp(tz_offset=9)}),
+                'thumbnail': (('picture_l_url', 'picture_m_url', 'picture_s_url'), {url_or_none}, any),
+            }),
+            'id': video_id,
+            'live_status': live_status,
+        }

yt_dlp/options.py

@@ -511,7 +511,7 @@ def create_parser():
     general.add_option(
         '--live-from-start',
         action='store_true', dest='live_from_start',
-        help='Download livestreams from the start. Currently experimental and only supported for YouTube and Twitch')
+        help='Download livestreams from the start. Currently experimental and only supported for YouTube, Twitch, and TVer')
     general.add_option(
         '--no-live-from-start',
         action='store_false', dest='live_from_start',