[extractor/youtube:tab] Fix video metadata from tabs (#5489)

Closes #5488
Authored by: coletdjnz
This commit is contained in:
Matthew 2022-11-10 15:44:12 +13:00 committed by GitHub
parent 495322b95b
commit 4dc23a8051
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23

View File

@ -912,7 +912,12 @@ def is_music_url(url):
def _extract_video(self, renderer): def _extract_video(self, renderer):
video_id = renderer.get('videoId') video_id = renderer.get('videoId')
title = self._get_text(renderer, 'title')
reel_header_renderer = traverse_obj(renderer, (
'navigationEndpoint', 'reelWatchEndpoint', 'overlay', 'reelPlayerOverlayRenderer',
'reelPlayerHeaderSupportedRenderers', 'reelPlayerHeaderRenderer'))
title = self._get_text(renderer, 'title', 'headline') or self._get_text(reel_header_renderer, 'reelTitleText')
description = self._get_text(renderer, 'descriptionSnippet') description = self._get_text(renderer, 'descriptionSnippet')
duration = int_or_none(renderer.get('lengthSeconds')) duration = int_or_none(renderer.get('lengthSeconds'))
@ -920,24 +925,23 @@ def _extract_video(self, renderer):
duration = parse_duration(self._get_text( duration = parse_duration(self._get_text(
renderer, 'lengthText', ('thumbnailOverlays', ..., 'thumbnailOverlayTimeStatusRenderer', 'text'))) renderer, 'lengthText', ('thumbnailOverlays', ..., 'thumbnailOverlayTimeStatusRenderer', 'text')))
if duration is None: if duration is None:
# XXX: should write a parser to be more general to support more cases (e.g. shorts in shorts tab)
duration = parse_duration(self._search_regex( duration = parse_duration(self._search_regex(
r'(?i)(ago)(?!.*\1)\s+(?P<duration>[a-z0-9 ,]+?)(?:\s+[\d,]+\s+views)?(?:\s+-\s+play\s+short)?$', r'(?i)(ago)(?!.*\1)\s+(?P<duration>[a-z0-9 ,]+?)(?:\s+[\d,]+\s+views)?(?:\s+-\s+play\s+short)?$',
traverse_obj(renderer, ('title', 'accessibility', 'accessibilityData', 'label'), default='', expected_type=str), traverse_obj(renderer, ('title', 'accessibility', 'accessibilityData', 'label'), default='', expected_type=str),
video_id, default=None, group='duration')) video_id, default=None, group='duration'))
# videoInfo is a string like '50K views • 10 years ago'.
view_count = self._get_count(renderer, 'viewCountText', 'shortViewCountText', 'videoInfo')
uploader = self._get_text(renderer, 'ownerText', 'shortBylineText')
channel_id = traverse_obj( channel_id = traverse_obj(
renderer, ('shortBylineText', 'runs', ..., 'navigationEndpoint', 'browseEndpoint', 'browseId'), renderer, ('shortBylineText', 'runs', ..., 'navigationEndpoint', 'browseEndpoint', 'browseId'),
expected_type=str, get_all=False) expected_type=str, get_all=False)
time_text = self._get_text(renderer, 'publishedTimeText', 'videoInfo') or '' if not channel_id:
scheduled_timestamp = str_to_int(traverse_obj(renderer, ('upcomingEventData', 'startTime'), get_all=False)) channel_id = traverse_obj(reel_header_renderer, ('channelNavigationEndpoint', 'browseEndpoint', 'browseId'))
overlay_style = traverse_obj( overlay_style = traverse_obj(
renderer, ('thumbnailOverlays', ..., 'thumbnailOverlayTimeStatusRenderer', 'style'), renderer, ('thumbnailOverlays', ..., 'thumbnailOverlayTimeStatusRenderer', 'style'),
get_all=False, expected_type=str) get_all=False, expected_type=str)
badges = self._extract_badges(renderer) badges = self._extract_badges(renderer)
thumbnails = self._extract_thumbnails(renderer, 'thumbnail')
navigation_url = urljoin('https://www.youtube.com/', traverse_obj( navigation_url = urljoin('https://www.youtube.com/', traverse_obj(
renderer, ('navigationEndpoint', 'commandMetadata', 'webCommandMetadata', 'url'), renderer, ('navigationEndpoint', 'commandMetadata', 'webCommandMetadata', 'url'),
expected_type=str)) or '' expected_type=str)) or ''
@ -945,12 +949,22 @@ def _extract_video(self, renderer):
if overlay_style == 'SHORTS' or '/shorts/' in navigation_url: if overlay_style == 'SHORTS' or '/shorts/' in navigation_url:
url = f'https://www.youtube.com/shorts/{video_id}' url = f'https://www.youtube.com/shorts/{video_id}'
time_text = (self._get_text(renderer, 'publishedTimeText', 'videoInfo')
or self._get_text(reel_header_renderer, 'timestampText') or '')
scheduled_timestamp = str_to_int(traverse_obj(renderer, ('upcomingEventData', 'startTime'), get_all=False))
live_status = ( live_status = (
'is_upcoming' if scheduled_timestamp is not None 'is_upcoming' if scheduled_timestamp is not None
else 'was_live' if 'streamed' in time_text.lower() else 'was_live' if 'streamed' in time_text.lower()
else 'is_live' if overlay_style == 'LIVE' or self._has_badge(badges, BadgeType.LIVE_NOW) else 'is_live' if overlay_style == 'LIVE' or self._has_badge(badges, BadgeType.LIVE_NOW)
else None) else None)
# videoInfo is a string like '50K views • 10 years ago'.
view_count_text = self._get_text(renderer, 'viewCountText', 'shortViewCountText', 'videoInfo') or ''
view_count = (0 if 'no views' in view_count_text.lower()
else self._get_count({'simpleText': view_count_text}))
view_count_field = 'concurrent_view_count' if live_status in ('is_live', 'is_upcoming') else 'view_count'
return { return {
'_type': 'url', '_type': 'url',
'ie_key': YoutubeIE.ie_key(), 'ie_key': YoutubeIE.ie_key(),
@ -959,9 +973,11 @@ def _extract_video(self, renderer):
'title': title, 'title': title,
'description': description, 'description': description,
'duration': duration, 'duration': duration,
'uploader': uploader,
'channel_id': channel_id, 'channel_id': channel_id,
'thumbnails': thumbnails, 'channel': (self._get_text(renderer, 'ownerText', 'shortBylineText')
or self._get_text(reel_header_renderer, 'channelTitleText')),
'channel_url': f'https://www.youtube.com/channel/{channel_id}' if channel_id else None,
'thumbnails': self._extract_thumbnails(renderer, 'thumbnail'),
'timestamp': (self._parse_time_text(time_text) 'timestamp': (self._parse_time_text(time_text)
if self._configuration_arg('approximate_date', ie_key=YoutubeTabIE) if self._configuration_arg('approximate_date', ie_key=YoutubeTabIE)
else None), else None),
@ -973,7 +989,7 @@ def _extract_video(self, renderer):
needs_premium=self._has_badge(badges, BadgeType.AVAILABILITY_PREMIUM) or None, needs_premium=self._has_badge(badges, BadgeType.AVAILABILITY_PREMIUM) or None,
needs_subscription=self._has_badge(badges, BadgeType.AVAILABILITY_SUBSCRIPTION) or None, needs_subscription=self._has_badge(badges, BadgeType.AVAILABILITY_SUBSCRIPTION) or None,
is_unlisted=self._has_badge(badges, BadgeType.AVAILABILITY_UNLISTED) or None), is_unlisted=self._has_badge(badges, BadgeType.AVAILABILITY_UNLISTED) or None),
'concurrent_view_count' if live_status in ('is_live', 'is_upcoming') else 'view_count': view_count, view_count_field: view_count,
'live_status': live_status 'live_status': live_status
} }
@ -5484,7 +5500,7 @@ class YoutubeTabIE(YoutubeTabBaseInfoExtractor):
'title': '#cctv9', 'title': '#cctv9',
'tags': [], 'tags': [],
}, },
'playlist_mincount': 350, 'playlist_mincount': 300, # not consistent but should be over 300
}, { }, {
'url': 'https://www.youtube.com/watch?list=PLW4dVinRY435CBE_JD3t-0SRXKfnZHS1P&feature=youtu.be&v=M9cJMXmQ_ZU', 'url': 'https://www.youtube.com/watch?list=PLW4dVinRY435CBE_JD3t-0SRXKfnZHS1P&feature=youtu.be&v=M9cJMXmQ_ZU',
'only_matching': True, 'only_matching': True,
@ -5671,7 +5687,7 @@ class YoutubeTabIE(YoutubeTabBaseInfoExtractor):
'tags': [], 'tags': [],
'uploader_id': 'UCiu-3thuViMebBjw_5nWYrA', 'uploader_id': 'UCiu-3thuViMebBjw_5nWYrA',
'channel_url': 'https://www.youtube.com/channel/UCiu-3thuViMebBjw_5nWYrA', 'channel_url': 'https://www.youtube.com/channel/UCiu-3thuViMebBjw_5nWYrA',
'description': '', 'description': 'test description',
'title': 'cole-dlp-test-acc - 再生リスト', 'title': 'cole-dlp-test-acc - 再生リスト',
'uploader_url': 'https://www.youtube.com/channel/UCiu-3thuViMebBjw_5nWYrA', 'uploader_url': 'https://www.youtube.com/channel/UCiu-3thuViMebBjw_5nWYrA',
'uploader': 'cole-dlp-test-acc', 'uploader': 'cole-dlp-test-acc',
@ -5828,6 +5844,62 @@ class YoutubeTabIE(YoutubeTabBaseInfoExtractor):
'tags': [], 'tags': [],
}, },
'playlist_mincount': 30, 'playlist_mincount': 30,
}, {
# Shorts url result in shorts tab
'url': 'https://www.youtube.com/channel/UCiu-3thuViMebBjw_5nWYrA/shorts',
'info_dict': {
'id': 'UCiu-3thuViMebBjw_5nWYrA',
'title': 'cole-dlp-test-acc - Shorts',
'uploader_id': 'UCiu-3thuViMebBjw_5nWYrA',
'channel': 'cole-dlp-test-acc',
'channel_follower_count': int,
'description': 'test description',
'channel_id': 'UCiu-3thuViMebBjw_5nWYrA',
'channel_url': 'https://www.youtube.com/channel/UCiu-3thuViMebBjw_5nWYrA',
'tags': [],
'uploader': 'cole-dlp-test-acc',
'uploader_url': 'https://www.youtube.com/channel/UCiu-3thuViMebBjw_5nWYrA',
},
'playlist': [{
'info_dict': {
'_type': 'url',
'ie_key': 'Youtube',
'url': 'https://www.youtube.com/shorts/sSM9J5YH_60',
'id': 'sSM9J5YH_60',
'channel_id': 'UCiu-3thuViMebBjw_5nWYrA',
'title': 'SHORT short',
'channel': 'cole-dlp-test-acc',
'channel_url': 'https://www.youtube.com/channel/UCiu-3thuViMebBjw_5nWYrA',
'view_count': int,
'thumbnails': list,
}
}],
'params': {'extract_flat': True},
}, {
# Live video status should be extracted
'url': 'https://www.youtube.com/channel/UCQvWX73GQygcwXOTSf_VDVg/live',
'info_dict': {
'id': 'UCQvWX73GQygcwXOTSf_VDVg',
'title': 'UCQvWX73GQygcwXOTSf_VDVg - Live', # TODO, should be Minecraft - Live or Minecraft - Topic - Live
'tags': []
},
'playlist': [{
'info_dict': {
'_type': 'url',
'ie_key': 'Youtube',
'url': 'startswith:https://www.youtube.com/watch?v=',
'id': str,
'title': str,
'live_status': 'is_live',
'channel_id': str,
'channel_url': str,
'concurrent_view_count': int,
'channel': str,
}
}],
'params': {'extract_flat': True},
'playlist_mincount': 1
}] }]
@classmethod @classmethod