import calendar
import copy
import datetime
+import enum
import hashlib
import itertools
import json
from ..utils import (
NO_DEFAULT,
ExtractorError,
+ LazyList,
UserNotLive,
bug_reports_message,
classproperty,
clean_html,
datetime_from_str,
dict_get,
+ filter_dict,
float_or_none,
format_field,
get_first,
build_innertube_clients()
+class BadgeType(enum.Enum):
+ AVAILABILITY_UNLISTED = enum.auto()
+ AVAILABILITY_PRIVATE = enum.auto()
+ AVAILABILITY_PUBLIC = enum.auto()
+ AVAILABILITY_PREMIUM = enum.auto()
+ AVAILABILITY_SUBSCRIPTION = enum.auto()
+ LIVE_NOW = enum.auto()
+
+
class YoutubeBaseInfoExtractor(InfoExtractor):
"""Provide base functions for Youtube extractors"""
_RESERVED_NAMES = (
r'channel|c|user|playlist|watch|w|v|embed|e|watch_popup|clip|'
r'shorts|movies|results|search|shared|hashtag|trending|explore|feed|feeds|'
- r'browse|oembed|get_video_info|iframe_api|s/player|'
- r'storefront|oops|index|account|reporthistory|t/terms|about|upload|signin|logout')
+ r'browse|oembed|get_video_info|iframe_api|s/player|source|'
+ r'storefront|oops|index|account|t/terms|about|upload|signin|logout')
_PLAYLIST_ID_RE = r'(?:(?:PL|LL|EC|UU|FL|RD|UL|TL|PU|OLAK5uy_)[0-9A-Za-z-_]{10,}|RDMM|WL|LL|LM)'
r'(?:www\.)?piped\.privacy\.com\.de',
)
+ # extracted from account/account_menu ep
+ # XXX: These are the supported YouTube UI and API languages,
+ # which is slightly different from languages supported for translation in YouTube studio
+ _SUPPORTED_LANG_CODES = [
+ 'af', 'az', 'id', 'ms', 'bs', 'ca', 'cs', 'da', 'de', 'et', 'en-IN', 'en-GB', 'en', 'es',
+ 'es-419', 'es-US', 'eu', 'fil', 'fr', 'fr-CA', 'gl', 'hr', 'zu', 'is', 'it', 'sw', 'lv',
+ 'lt', 'hu', 'nl', 'no', 'uz', 'pl', 'pt-PT', 'pt', 'ro', 'sq', 'sk', 'sl', 'sr-Latn', 'fi',
+ 'sv', 'vi', 'tr', 'be', 'bg', 'ky', 'kk', 'mk', 'mn', 'ru', 'sr', 'uk', 'el', 'hy', 'iw',
+ 'ur', 'ar', 'fa', 'ne', 'mr', 'hi', 'as', 'bn', 'pa', 'gu', 'or', 'ta', 'te', 'kn', 'ml',
+ 'si', 'th', 'lo', 'my', 'ka', 'am', 'km', 'zh-CN', 'zh-TW', 'zh-HK', 'ja', 'ko'
+ ]
+
+ _IGNORED_WARNINGS = {'Unavailable videos will be hidden during playback'}
+
+ @functools.cached_property
+ def _preferred_lang(self):
+ """
+ Returns a language code supported by YouTube for the user preferred language.
+ Returns None if no preferred language set.
+ """
+ preferred_lang = self._configuration_arg('lang', ie_key='Youtube', casesense=True, default=[''])[0]
+ if not preferred_lang:
+ return
+ if preferred_lang not in self._SUPPORTED_LANG_CODES:
+ raise ExtractorError(
+ f'Unsupported language code: {preferred_lang}. Supported language codes (case-sensitive): {join_nonempty(*self._SUPPORTED_LANG_CODES, delim=", ")}.',
+ expected=True)
+ elif preferred_lang != 'en':
+ self.report_warning(
+ f'Preferring "{preferred_lang}" translated fields. Note that some metadata extraction may fail or be incorrect.')
+ return preferred_lang
+
def _initialize_consent(self):
cookies = self._get_cookies('https://www.youtube.com/')
if cookies.get('__Secure-3PSID'):
pref = dict(urllib.parse.parse_qsl(pref_cookie.value))
except ValueError:
self.report_warning('Failed to parse user PREF cookie' + bug_reports_message())
- pref.update({'hl': 'en', 'tz': 'UTC'})
+ pref.update({'hl': self._preferred_lang or 'en', 'tz': 'UTC'})
self._set_cookie('.youtube.com', name='PREF', value=urllib.parse.urlencode(pref))
def _real_initialize(self):
(ytcfg, self._get_default_ytcfg(default_client)), 'INNERTUBE_CONTEXT', expected_type=dict)
# Enforce language and tz for extraction
client_context = traverse_obj(context, 'client', expected_type=dict, default={})
- client_context.update({'hl': 'en', 'timeZone': 'UTC', 'utcOffsetMinutes': 0})
+ client_context.update({'hl': self._preferred_lang or 'en', 'timeZone': 'UTC', 'utcOffsetMinutes': 0})
return context
_SAPISID = None
if auth is not None:
headers['Authorization'] = auth
headers['X-Origin'] = origin
- return {h: v for h, v in headers.items() if v is not None}
+ return filter_dict(headers)
def _download_ytcfg(self, client, video_id):
url = {
if next_continuation:
return next_continuation
- contents = []
- for key in ('contents', 'items'):
- contents.extend(try_get(renderer, lambda x: x[key], list) or [])
-
- for content in contents:
- if not isinstance(content, dict):
- continue
- continuation_ep = try_get(
- content, (lambda x: x['continuationItemRenderer']['continuationEndpoint'],
- lambda x: x['continuationItemRenderer']['button']['buttonRenderer']['command']),
- dict)
- continuation = cls._extract_continuation_ep_data(continuation_ep)
- if continuation:
- return continuation
+ return traverse_obj(renderer, (
+ ('contents', 'items', 'rows'), ..., 'continuationItemRenderer',
+ ('continuationEndpoint', ('button', 'buttonRenderer', 'command'))
+ ), get_all=False, expected_type=cls._extract_continuation_ep_data)
@classmethod
def _extract_alerts(cls, data):
yield alert_type, message
def _report_alerts(self, alerts, expected=True, fatal=True, only_once=False):
- errors = []
- warnings = []
+ errors, warnings = [], []
for alert_type, alert_message in alerts:
if alert_type.lower() == 'error' and fatal:
errors.append([alert_type, alert_message])
- else:
+ elif alert_message not in self._IGNORED_WARNINGS:
warnings.append([alert_type, alert_message])
for alert_type, alert_message in (warnings + errors[:-1]):
return self._report_alerts(self._extract_alerts(data), *args, **kwargs)
def _extract_badges(self, renderer: dict):
- badges = set()
- for badge in try_get(renderer, lambda x: x['badges'], list) or []:
- label = try_get(badge, lambda x: x['metadataBadgeRenderer']['label'], str)
- if label:
- badges.add(label.lower())
+ privacy_icon_map = {
+ 'PRIVACY_UNLISTED': BadgeType.AVAILABILITY_UNLISTED,
+ 'PRIVACY_PRIVATE': BadgeType.AVAILABILITY_PRIVATE,
+ 'PRIVACY_PUBLIC': BadgeType.AVAILABILITY_PUBLIC
+ }
+
+ badge_style_map = {
+ 'BADGE_STYLE_TYPE_MEMBERS_ONLY': BadgeType.AVAILABILITY_SUBSCRIPTION,
+ 'BADGE_STYLE_TYPE_PREMIUM': BadgeType.AVAILABILITY_PREMIUM,
+ 'BADGE_STYLE_TYPE_LIVE_NOW': BadgeType.LIVE_NOW
+ }
+
+ label_map = {
+ 'unlisted': BadgeType.AVAILABILITY_UNLISTED,
+ 'private': BadgeType.AVAILABILITY_PRIVATE,
+ 'members only': BadgeType.AVAILABILITY_SUBSCRIPTION,
+ 'live': BadgeType.LIVE_NOW,
+ 'premium': BadgeType.AVAILABILITY_PREMIUM
+ }
+
+ badges = []
+ for badge in traverse_obj(renderer, ('badges', ..., 'metadataBadgeRenderer'), default=[]):
+ badge_type = (
+ privacy_icon_map.get(traverse_obj(badge, ('icon', 'iconType'), expected_type=str))
+ or badge_style_map.get(traverse_obj(badge, 'style'))
+ )
+ if badge_type:
+ badges.append({'type': badge_type})
+ continue
+
+ # fallback, won't work in some languages
+ label = traverse_obj(badge, 'label', expected_type=str, default='')
+ for match, label_badge_type in label_map.items():
+ if match in label.lower():
+ badges.append({'type': badge_type})
+ continue
+
return badges
+ @staticmethod
+ def _has_badge(badges, badge_type):
+ return bool(traverse_obj(badges, lambda _, v: v['type'] == badge_type))
+
@staticmethod
def _get_text(data, *path_list, max_runs=None):
for path in path_list or [None]:
except ValueError:
return None
- def _extract_time_text(self, renderer, *path_list):
- """@returns (timestamp, time_text)"""
- text = self._get_text(renderer, *path_list) or ''
+ def _parse_time_text(self, text):
+ if not text:
+ return
dt = self.extract_relative_time(text)
timestamp = None
if isinstance(dt, datetime.datetime):
(r'([a-z]+\s*\d{1,2},?\s*20\d{2})', r'(?:.+|^)(?:live|premieres|ed|ing)(?:\s*(?:on|for))?\s*(.+\d)'),
text.lower(), 'time text', default=None)))
- if text and timestamp is None:
- self.report_warning(f"Cannot parse localized time text '{text}'" + bug_reports_message(), only_once=True)
- return timestamp, text
+ if text and timestamp is None and self._preferred_lang in (None, 'en'):
+ self.report_warning(
+ f'Cannot parse localized time text "{text}"', only_once=True)
+ return timestamp
def _extract_response(self, item_id, query, note='Downloading API JSON', headers=None,
ytcfg=None, check_get_keys=None, ep='browse', fatal=True, api_hostname=None,
video_id = renderer.get('videoId')
title = self._get_text(renderer, 'title')
description = self._get_text(renderer, 'descriptionSnippet')
- duration = parse_duration(self._get_text(
- renderer, 'lengthText', ('thumbnailOverlays', ..., 'thumbnailOverlayTimeStatusRenderer', 'text')))
+
+ duration = int_or_none(renderer.get('lengthSeconds'))
+ if duration is None:
+ duration = parse_duration(self._get_text(
+ renderer, 'lengthText', ('thumbnailOverlays', ..., 'thumbnailOverlayTimeStatusRenderer', 'text')))
if duration is None:
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)?$',
traverse_obj(renderer, ('title', 'accessibility', 'accessibilityData', 'label'), default='', expected_type=str),
video_id, default=None, group='duration'))
- view_count = self._get_count(renderer, 'viewCountText')
-
+ # 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(
renderer, ('shortBylineText', 'runs', ..., 'navigationEndpoint', 'browseEndpoint', 'browseId'),
expected_type=str, get_all=False)
- timestamp, time_text = self._extract_time_text(renderer, 'publishedTimeText')
+ time_text = self._get_text(renderer, 'publishedTimeText', 'videoInfo') or ''
scheduled_timestamp = str_to_int(traverse_obj(renderer, ('upcomingEventData', 'startTime'), get_all=False))
overlay_style = traverse_obj(
renderer, ('thumbnailOverlays', ..., 'thumbnailOverlayTimeStatusRenderer', 'style'),
if overlay_style == 'SHORTS' or '/shorts/' in navigation_url:
url = f'https://www.youtube.com/shorts/{video_id}'
+ live_status = (
+ 'is_upcoming' if scheduled_timestamp is not None
+ 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 None)
+
return {
'_type': 'url',
'ie_key': YoutubeIE.ie_key(),
'title': title,
'description': description,
'duration': duration,
- 'view_count': view_count,
'uploader': uploader,
'channel_id': channel_id,
'thumbnails': thumbnails,
- 'upload_date': (strftime_or_none(timestamp, '%Y%m%d')
- if self._configuration_arg('approximate_date', ie_key='youtubetab')
- else None),
- 'live_status': ('is_upcoming' if scheduled_timestamp is not None
- else 'was_live' if 'streamed' in time_text.lower()
- else 'is_live' if overlay_style == 'LIVE' or 'live now' in badges
- else None),
+ 'timestamp': (self._parse_time_text(time_text)
+ if self._configuration_arg('approximate_date', ie_key=YoutubeTabIE)
+ else None),
'release_timestamp': scheduled_timestamp,
- 'availability': self._availability(needs_premium='premium' in badges, needs_subscription='members only' in badges)
+ 'availability':
+ 'public' if self._has_badge(badges, BadgeType.AVAILABILITY_PUBLIC)
+ else self._availability(
+ is_private=self._has_badge(badges, BadgeType.AVAILABILITY_PRIVATE) or None,
+ needs_premium=self._has_badge(badges, BadgeType.AVAILABILITY_PREMIUM) or None,
+ needs_subscription=self._has_badge(badges, BadgeType.AVAILABILITY_SUBSCRIPTION) 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,
+ 'live_status': live_status
}
_EMBED_REGEX = [
r'''(?x)
(?:
- <iframe[^>]+?src=|
+ <(?:[0-9A-Za-z-]+?)?iframe[^>]+?src=|
data-video-url=|
<embed[^>]+?src=|
embedSWF\(?:\s*|
'live_status': 'not_live',
'playable_in_embed': True,
'comment_count': int,
- 'channel_follower_count': int
+ 'channel_follower_count': int,
+ 'chapters': list,
},
'params': {
'skip_download': True,
'live_status': 'not_live',
'channel_url': 'https://www.youtube.com/channel/UCH1dpzjCEiGAt8CXkryhkZg',
'comment_count': int,
- 'channel_follower_count': int
+ 'channel_follower_count': int,
+ 'chapters': list,
},
'params': {
'skip_download': True,
'duration': 522,
'channel': 'kudvenkat',
'comment_count': int,
- 'channel_follower_count': int
+ 'channel_follower_count': int,
+ 'chapters': list,
},
'params': {
'skip_download': True,
'like_count': int,
'live_status': 'not_live',
'playable_in_embed': True,
- 'channel_follower_count': int
+ 'channel_follower_count': int,
+ 'chapters': list,
},
'params': {
'format': '17', # 3gp format available on android
'duration': 248,
'categories': ['Education'],
'age_limit': 0,
- 'channel_follower_count': int
+ 'channel_follower_count': int,
+ 'chapters': list,
}, 'params': {'format': 'mhtml', 'skip_download': True}
}, {
# Ensure video upload_date is in UTC timezone (video was uploaded 1641170939)
'view_count': int,
'playable_in_embed': True,
'description': 'md5:2ef1d002cad520f65825346e2084e49d',
+ 'concurrent_view_count': int,
},
'params': {'skip_download': True}
}, {
'tags': [],
'uploader_url': 'http://www.youtube.com/user/nao20010128nao',
}
+ }, {
+ # Prefer primary title+description language metadata by default
+ # Do not prefer translated description if primary is empty
+ 'url': 'https://www.youtube.com/watch?v=el3E4MbxRqQ',
+ 'info_dict': {
+ 'id': 'el3E4MbxRqQ',
+ 'ext': 'mp4',
+ 'title': 'dlp test video 2 - primary sv no desc',
+ 'description': '',
+ 'channel': 'cole-dlp-test-acc',
+ 'tags': [],
+ 'view_count': int,
+ 'channel_url': 'https://www.youtube.com/channel/UCiu-3thuViMebBjw_5nWYrA',
+ 'like_count': int,
+ 'playable_in_embed': True,
+ 'availability': 'unlisted',
+ 'thumbnail': 'https://i.ytimg.com/vi_webp/el3E4MbxRqQ/maxresdefault.webp',
+ 'age_limit': 0,
+ 'duration': 5,
+ 'uploader_id': 'UCiu-3thuViMebBjw_5nWYrA',
+ 'uploader_url': 'http://www.youtube.com/channel/UCiu-3thuViMebBjw_5nWYrA',
+ 'live_status': 'not_live',
+ 'upload_date': '20220908',
+ 'categories': ['People & Blogs'],
+ 'uploader': 'cole-dlp-test-acc',
+ 'channel_id': 'UCiu-3thuViMebBjw_5nWYrA',
+ },
+ 'params': {'skip_download': True}
+ }, {
+ # Extractor argument: prefer translated title+description
+ 'url': 'https://www.youtube.com/watch?v=gHKT4uU8Zng',
+ 'info_dict': {
+ 'id': 'gHKT4uU8Zng',
+ 'ext': 'mp4',
+ 'channel': 'cole-dlp-test-acc',
+ 'tags': [],
+ 'duration': 5,
+ 'live_status': 'not_live',
+ 'channel_id': 'UCiu-3thuViMebBjw_5nWYrA',
+ 'upload_date': '20220728',
+ 'uploader_id': 'UCiu-3thuViMebBjw_5nWYrA',
+ 'view_count': int,
+ 'categories': ['People & Blogs'],
+ 'thumbnail': 'https://i.ytimg.com/vi_webp/gHKT4uU8Zng/maxresdefault.webp',
+ 'title': 'dlp test video title translated (fr)',
+ 'availability': 'public',
+ 'uploader': 'cole-dlp-test-acc',
+ 'age_limit': 0,
+ 'description': 'dlp test video description translated (fr)',
+ 'playable_in_embed': True,
+ 'channel_url': 'https://www.youtube.com/channel/UCiu-3thuViMebBjw_5nWYrA',
+ 'uploader_url': 'http://www.youtube.com/channel/UCiu-3thuViMebBjw_5nWYrA',
+ },
+ 'params': {'skip_download': True, 'extractor_args': {'youtube': {'lang': ['fr']}}},
+ 'expected_warnings': [r'Preferring "fr" translated fields'],
}, {
'note': '6 channel audio',
'url': 'https://www.youtube.com/watch?v=zgdo7-RRjgo',
self._code_cache = {}
self._player_cache = {}
- def _prepare_live_from_start_formats(self, formats, video_id, live_start_time, url, webpage_url, smuggled_data):
+ def _prepare_live_from_start_formats(self, formats, video_id, live_start_time, url, webpage_url, smuggled_data, is_live):
lock = threading.Lock()
-
- is_live = True
start_time = time.time()
formats = [f for f in formats if f.get('is_from_start')]
microformats = traverse_obj(
prs, (..., 'microformat', 'playerMicroformatRenderer'),
expected_type=dict, default=[])
- _, is_live, _, formats, _ = self._list_formats(video_id, microformats, video_details, prs, player_url)
+ _, live_status, _, formats, _ = self._list_formats(video_id, microformats, video_details, prs, player_url)
+ is_live = live_status == 'is_live'
start_time = time.time()
def mpd_feed(format_id, delay):
return f['manifest_url'], f['manifest_stream_number'], is_live
for f in formats:
- f['is_live'] = True
- f['protocol'] = 'http_dash_segments_generator'
- f['fragments'] = functools.partial(
- self._live_dash_fragments, f['format_id'], live_start_time, mpd_feed)
+ f['is_live'] = is_live
+ gen = functools.partial(self._live_dash_fragments, video_id, f['format_id'],
+ live_start_time, mpd_feed, not is_live and f.copy())
+ if is_live:
+ f['fragments'] = gen
+ f['protocol'] = 'http_dash_segments_generator'
+ else:
+ f['fragments'] = LazyList(gen({}))
+ del f['is_from_start']
- def _live_dash_fragments(self, format_id, live_start_time, mpd_feed, ctx):
+ def _live_dash_fragments(self, video_id, format_id, live_start_time, mpd_feed, manifestless_orig_fmt, ctx):
FETCH_SPAN, MAX_DURATION = 5, 432000
mpd_url, stream_number, is_live = None, None, True
return False, last_seq
elif old_mpd_url == mpd_url:
return True, last_seq
- try:
- fmts, _ = self._extract_mpd_formats_and_subtitles(
- mpd_url, None, note=False, errnote=False, fatal=False)
- except ExtractorError:
- fmts = None
- if not fmts:
- no_fragment_score += 2
- return False, last_seq
- fmt_info = next(x for x in fmts if x['manifest_stream_number'] == stream_number)
+ if manifestless_orig_fmt:
+ fmt_info = manifestless_orig_fmt
+ else:
+ try:
+ fmts, _ = self._extract_mpd_formats_and_subtitles(
+ mpd_url, None, note=False, errnote=False, fatal=False)
+ except ExtractorError:
+ fmts = None
+ if not fmts:
+ no_fragment_score += 2
+ return False, last_seq
+ fmt_info = next(x for x in fmts if x['manifest_stream_number'] == stream_number)
fragments = fmt_info['fragments']
fragment_base_url = fmt_info['fragment_base_url']
assert fragment_base_url
_last_seq = int(re.search(r'(?:/|^)sq/(\d+)', fragments[-1]['path']).group(1))
return True, _last_seq
+ self.write_debug(f'[{video_id}] Generating fragments for format {format_id}')
while is_live:
fetch_time = time.time()
if no_fragment_score > 30:
except ExtractorError:
continue
+ if manifestless_orig_fmt:
+ # Stop at the first iteration if running for post-live manifestless;
+ # fragment count no longer increase since it starts
+ break
+
time.sleep(max(0, FETCH_SPAN + fetch_time - time.time()))
def _extract_player_url(self, *ytcfgs, webpage=None):
self.report_warning(
f'Native nsig extraction failed: Trying with PhantomJS\n'
f' n = {s} ; player = {player_url}', video_id)
- self.write_debug(e)
+ self.write_debug(e, only_once=True)
args, func_body = func_code
ret = jsi.execute(
# these seem to mark watchtime "history" in the real world
# they're required, so send in a single value
qs.update({
- 'st': video_length,
+ 'st': 0,
'et': video_length,
})
for contents in content_list)), [])
def _extract_chapters_from_description(self, description, duration):
+ duration_re = r'(?:\d+:)?\d{1,2}:\d{2}'
+ sep_re = r'(?m)^\s*(%s)\b\W*\s(%s)\s*$'
return self._extract_chapters(
- re.findall(r'(?m)^((?:\d+:)?\d{1,2}:\d{2})\b\W*\s(.+?)\s*$', description or ''),
+ re.findall(sep_re % (duration_re, r'.+?'), description or ''),
chapter_time=lambda x: parse_duration(x[0]), chapter_title=lambda x: x[1],
+ duration=duration, strict=False) or self._extract_chapters(
+ re.findall(sep_re % (r'.+?', duration_re), description or ''),
+ chapter_time=lambda x: parse_duration(x[1]), chapter_title=lambda x: x[0],
duration=duration, strict=False)
def _extract_chapters(self, chapter_list, chapter_time, chapter_title, duration, strict=True):
self.report_warning(f'Incomplete chapter {idx}')
elif chapters[-1]['start_time'] <= chapter['start_time'] <= duration:
chapters.append(chapter)
- else:
- self.report_warning(f'Invalid start time for chapter "{chapter["title"]}"')
+ elif chapter not in chapters:
+ self.report_warning(
+ f'Invalid start time ({chapter["start_time"]} < {chapters[-1]["start_time"]}) for chapter "{chapter["title"]}"')
return chapters[1:]
def _extract_comment(self, comment_renderer, parent=None):
text = self._get_text(comment_renderer, 'contentText')
- # note: timestamp is an estimate calculated from the current time and time_text
- timestamp, time_text = self._extract_time_text(comment_renderer, 'publishedTimeText')
+ # Timestamp is an estimate calculated from the current time and time_text
+ time_text = self._get_text(comment_renderer, 'publishedTimeText') or ''
+ timestamp = self._parse_time_text(time_text)
+
author = self._get_text(comment_renderer, 'authorText')
author_id = try_get(comment_renderer,
lambda x: x['authorEndpoint']['browseEndpoint']['browseId'], str)
self.report_warning(last_error)
return prs, player_url
- def _extract_formats_and_subtitles(self, streaming_data, video_id, player_url, is_live, duration):
+ def _needs_live_processing(self, live_status, duration):
+ if (live_status == 'is_live' and self.get_param('live_from_start')
+ or live_status == 'post_live' and (duration or 0) > 4 * 3600):
+ return live_status
+
+ def _extract_formats_and_subtitles(self, streaming_data, video_id, player_url, live_status, duration):
itags, stream_ids = {}, []
itag_qualities, res_qualities = {}, {0: None}
q = qualities([
dct['container'] = dct['ext'] + '_dash'
yield dct
- live_from_start = is_live and self.get_param('live_from_start')
- skip_manifests = self._configuration_arg('skip')
- if not self.get_param('youtube_include_hls_manifest', True):
- skip_manifests.append('hls')
+ needs_live_processing = self._needs_live_processing(live_status, duration)
+ skip_bad_formats = not self._configuration_arg('include_incomplete_formats')
+
+ skip_manifests = set(self._configuration_arg('skip'))
+ if (not self.get_param('youtube_include_hls_manifest', True)
+ or needs_live_processing == 'is_live' # These will be filtered out by YoutubeDL anyway
+ or needs_live_processing and skip_bad_formats):
+ skip_manifests.add('hls')
+
if not self.get_param('youtube_include_dash_manifest', True):
- skip_manifests.append('dash')
- get_dash = 'dash' not in skip_manifests and (
- not is_live or live_from_start or self._configuration_arg('include_live_dash'))
- get_hls = not live_from_start and 'hls' not in skip_manifests
+ skip_manifests.add('dash')
+ if self._configuration_arg('include_live_dash'):
+ self._downloader.deprecated_feature('[youtube] include_live_dash extractor argument is deprecated. '
+ 'Use include_incomplete_formats extractor argument instead')
+ elif skip_bad_formats and live_status == 'is_live' and needs_live_processing != 'is_live':
+ skip_manifests.add('dash')
def process_manifest_format(f, proto, itag):
if itag in itags:
subtitles = {}
for sd in streaming_data:
- hls_manifest_url = get_hls and sd.get('hlsManifestUrl')
+ hls_manifest_url = 'hls' not in skip_manifests and sd.get('hlsManifestUrl')
if hls_manifest_url:
- fmts, subs = self._extract_m3u8_formats_and_subtitles(hls_manifest_url, video_id, 'mp4', fatal=False, live=is_live)
+ fmts, subs = self._extract_m3u8_formats_and_subtitles(
+ hls_manifest_url, video_id, 'mp4', fatal=False, live=live_status == 'is_live')
subtitles = self._merge_subtitles(subs, subtitles)
for f in fmts:
if process_manifest_format(f, 'hls', self._search_regex(
r'/itag/(\d+)', f['url'], 'itag', default=None)):
yield f
- dash_manifest_url = get_dash and sd.get('dashManifestUrl')
+ dash_manifest_url = 'dash' not in skip_manifests and sd.get('dashManifestUrl')
if dash_manifest_url:
formats, subs = self._extract_mpd_formats_and_subtitles(dash_manifest_url, video_id, fatal=False)
subtitles = self._merge_subtitles(subs, subtitles) # Prioritize HLS subs over DASH
if process_manifest_format(f, 'dash', f['format_id']):
f['filesize'] = int_or_none(self._search_regex(
r'/clen/(\d+)', f.get('fragment_base_url') or f['url'], 'file size', default=None))
- if live_from_start:
+ if needs_live_processing:
f['is_from_start'] = True
yield f
is_live = get_first(video_details, 'isLive')
if is_live is None:
is_live = get_first(live_broadcast_details, 'isLiveNow')
-
+ live_content = get_first(video_details, 'isLiveContent')
+ is_upcoming = get_first(video_details, 'isUpcoming')
+ post_live = get_first(video_details, 'isPostLiveDvr')
+ live_status = ('post_live' if post_live
+ else 'is_live' if is_live
+ else 'is_upcoming' if is_upcoming
+ else 'was_live' if live_content
+ else 'not_live' if False in (is_live, live_content)
+ else None)
streaming_data = traverse_obj(player_responses, (..., 'streamingData'), default=[])
- *formats, subtitles = self._extract_formats_and_subtitles(streaming_data, video_id, player_url, is_live, duration)
+ *formats, subtitles = self._extract_formats_and_subtitles(streaming_data, video_id, player_url, live_status, duration)
- return live_broadcast_details, is_live, streaming_data, formats, subtitles
+ return live_broadcast_details, live_status, streaming_data, formats, subtitles
def _real_extract(self, url):
url, smuggled_data = unsmuggle_url(url, {})
microformats = traverse_obj(
player_responses, (..., 'microformat', 'playerMicroformatRenderer'),
expected_type=dict, default=[])
- video_title = (
- get_first(video_details, 'title')
- or self._get_text(microformats, (..., 'title'))
- or search_meta(['og:title', 'twitter:title', 'title']))
- video_description = get_first(video_details, 'shortDescription')
+
+ translated_title = self._get_text(microformats, (..., 'title'))
+ video_title = (self._preferred_lang and translated_title
+ or get_first(video_details, 'title') # primary
+ or translated_title
+ or search_meta(['og:title', 'twitter:title', 'title']))
+ translated_description = self._get_text(microformats, (..., 'description'))
+ original_description = get_first(video_details, 'shortDescription')
+ video_description = (
+ self._preferred_lang and translated_description
+ # If original description is blank, it will be an empty string.
+ # Do not prefer translated description in this case.
+ or original_description if original_description is not None else translated_description)
multifeed_metadata_list = get_first(
player_responses,
return self.playlist_result(
entries, video_id, video_title, video_description)
- duration = int_or_none(
- get_first(video_details, 'lengthSeconds')
- or get_first(microformats, 'lengthSeconds')
- or parse_duration(search_meta('duration'))) or None
+ duration = (int_or_none(get_first(video_details, 'lengthSeconds'))
+ or int_or_none(get_first(microformats, 'lengthSeconds'))
+ or parse_duration(search_meta('duration')) or None)
- live_broadcast_details, is_live, streaming_data, formats, automatic_captions = \
- self._list_formats(video_id, microformats, video_details, player_responses, player_url)
+ live_broadcast_details, live_status, streaming_data, formats, automatic_captions = \
+ self._list_formats(video_id, microformats, video_details, player_responses, player_url, duration)
+ if live_status == 'post_live':
+ self.write_debug(f'{video_id}: Video is in Post-Live Manifestless mode')
if not formats:
if not self.get_param('allow_unplayable_formats') and traverse_obj(streaming_data, (..., 'licenseInfos')):
thumbnails.extend({
'url': 'https://i.ytimg.com/vi{webp}/{video_id}/{name}{live}.{ext}'.format(
video_id=video_id, name=name, ext=ext,
- webp='_webp' if ext == 'webp' else '', live='_live' if is_live else ''),
+ webp='_webp' if ext == 'webp' else '', live='_live' if live_status == 'is_live' else ''),
} for name in thumbnail_names for ext in ('webp', 'jpg'))
for thumb in thumbnails:
i = next((i for i, t in enumerate(thumbnail_names) if f'/{video_id}/{t}' in thumb['url']), n_thumbnail_names)
or search_meta('channelId'))
owner_profile_url = get_first(microformats, 'ownerProfileUrl')
- live_content = get_first(video_details, 'isLiveContent')
- is_upcoming = get_first(video_details, 'isUpcoming')
- if is_live is None:
- if is_upcoming or live_content is False:
- is_live = False
- if is_upcoming is None and (live_content or is_live):
- is_upcoming = False
live_start_time = parse_iso8601(get_first(live_broadcast_details, 'startTimestamp'))
live_end_time = parse_iso8601(get_first(live_broadcast_details, 'endTimestamp'))
if not duration and live_end_time and live_start_time:
duration = live_end_time - live_start_time
- if is_live and self.get_param('live_from_start'):
- self._prepare_live_from_start_formats(formats, video_id, live_start_time, url, webpage_url, smuggled_data)
+ needs_live_processing = self._needs_live_processing(live_status, duration)
+
+ def is_bad_format(fmt):
+ if needs_live_processing and not fmt.get('is_from_start'):
+ return True
+ elif (live_status == 'is_live' and needs_live_processing != 'is_live'
+ and fmt.get('protocol') == 'http_dash_segments'):
+ return True
+
+ for fmt in filter(is_bad_format, formats):
+ fmt['preference'] = (fmt.get('preference') or -1) - 10
+ fmt['format_note'] = join_nonempty(fmt.get('format_note'), '(Last 4 hours)', delim=' ')
+
+ if needs_live_processing:
+ self._prepare_live_from_start_formats(
+ formats, video_id, live_start_time, url, webpage_url, smuggled_data, live_status == 'is_live')
formats.extend(self._extract_storyboard(player_responses, duration))
'categories': [category] if category else None,
'tags': keywords,
'playable_in_embed': get_first(playability_statuses, 'playableInEmbed'),
- 'is_live': is_live,
- 'was_live': (False if is_live or is_upcoming or live_content is False
- else None if is_live is None or is_upcoming is None
- else live_content),
- 'live_status': 'is_upcoming' if is_upcoming else None, # rest will be set by YoutubeDL
+ 'live_status': live_status,
'release_timestamp': live_start_time,
}
- if get_first(video_details, 'isPostLiveDvr'):
- self.write_debug('Video is in Post-Live Manifestless mode')
- info['live_status'] = 'post_live'
- if (duration or 0) > 4 * 3600:
- self.report_warning(
- 'The livestream has not finished processing. Only 4 hours of the video can be currently downloaded. '
- 'This is a known issue and patches are welcome')
-
subtitles = {}
pctr = traverse_obj(player_responses, (..., 'captions', 'playerCaptionsTracklistRenderer'), expected_type=dict)
if pctr:
'url': f'https://www.youtube.com/watch?v={video_id}&bpctr=9999999999&has_verified=1',
'video_id': video_id,
'ext': 'json',
- 'protocol': 'youtube_live_chat' if is_live or is_upcoming else 'youtube_live_chat_replay',
+ 'protocol': ('youtube_live_chat' if live_status in ('is_live', 'is_upcoming')
+ else 'youtube_live_chat_replay'),
}]
if initial_data:
vpir,
lambda x: x['videoActions']['menuRenderer']['topLevelButtons'],
list) or []):
- tbr = tlb.get('toggleButtonRenderer') or {}
- for getter, regex in [(
- lambda x: x['defaultText']['accessibility']['accessibilityData'],
- r'(?P<count>[\d,]+)\s*(?P<type>(?:dis)?like)'), ([
- lambda x: x['accessibility'],
- lambda x: x['accessibilityData']['accessibilityData'],
- ], r'(?P<type>(?:dis)?like) this video along with (?P<count>[\d,]+) other people')]:
- label = (try_get(tbr, getter, dict) or {}).get('label')
- if label:
- mobj = re.match(regex, label)
- if mobj:
- info[mobj.group('type') + '_count'] = str_to_int(mobj.group('count'))
- break
+ tbrs = variadic(
+ traverse_obj(
+ tlb, 'toggleButtonRenderer',
+ ('segmentedLikeDislikeButtonRenderer', ..., 'toggleButtonRenderer'),
+ default=[]))
+ for tbr in tbrs:
+ for getter, regex in [(
+ lambda x: x['defaultText']['accessibility']['accessibilityData'],
+ r'(?P<count>[\d,]+)\s*(?P<type>(?:dis)?like)'), ([
+ lambda x: x['accessibility'],
+ lambda x: x['accessibilityData']['accessibilityData'],
+ ], r'(?P<type>(?:dis)?like) this video along with (?P<count>[\d,]+) other people')]:
+ label = (try_get(tbr, getter, dict) or {}).get('label')
+ if label:
+ mobj = re.match(regex, label)
+ if mobj:
+ info[mobj.group('type') + '_count'] = str_to_int(mobj.group('count'))
+ break
sbr_tooltip = try_get(
vpir, lambda x: x['sentimentBar']['sentimentBarRenderer']['tooltip'])
if sbr_tooltip:
'like_count': str_to_int(like_count),
'dislike_count': str_to_int(dislike_count),
})
+ vcr = traverse_obj(vpir, ('viewCount', 'videoViewCountRenderer'))
+ if vcr:
+ vc = self._get_count(vcr, 'viewCount')
+ # Upcoming premieres with waiting count are treated as live here
+ if vcr.get('isLive'):
+ info['concurrent_view_count'] = vc
+ elif info.get('view_count') is None:
+ info['view_count'] = vc
+
vsir = get_first(contents, 'videoSecondaryInfoRenderer')
if vsir:
vor = traverse_obj(vsir, ('owner', 'videoOwnerRenderer'))
unified_strdate(get_first(microformats, 'uploadDate'))
or unified_strdate(search_meta('uploadDate')))
if not upload_date or (
- not info.get('is_live')
- and not info.get('was_live')
- and info.get('live_status') != 'is_upcoming'
+ live_status in ('not_live', None)
and 'no-youtube-prefer-utc-upload-date' not in self.get_param('compat_opts', [])
):
- upload_date = strftime_or_none(self._extract_time_text(vpir, 'dateText')[0], '%Y%m%d') or upload_date
+ upload_date = strftime_or_none(
+ self._parse_time_text(self._get_text(vpir, 'dateText')), '%Y%m%d') or upload_date
info['upload_date'] = upload_date
for to, frm in fallbacks.items():
if v:
info[d_k] = v
- is_private = get_first(video_details, 'isPrivate', expected_type=bool)
- is_unlisted = get_first(microformats, 'isUnlisted', expected_type=bool)
- is_membersonly = None
- is_premium = None
- if initial_data and is_private is not None:
- is_membersonly = False
- is_premium = False
- contents = try_get(initial_data, lambda x: x['contents']['twoColumnWatchNextResults']['results']['results']['contents'], list) or []
- badge_labels = set()
- for content in contents:
- if not isinstance(content, dict):
- continue
- badge_labels.update(self._extract_badges(content.get('videoPrimaryInfoRenderer')))
- for badge_label in badge_labels:
- if badge_label.lower() == 'members only':
- is_membersonly = True
- elif badge_label.lower() == 'premium':
- is_premium = True
- elif badge_label.lower() == 'unlisted':
- is_unlisted = True
-
- info['availability'] = self._availability(
- is_private=is_private,
- needs_premium=is_premium,
- needs_subscription=is_membersonly,
- needs_auth=info['age_limit'] >= 18,
- is_unlisted=None if is_private is None else is_unlisted)
+ badges = self._extract_badges(traverse_obj(contents, (..., 'videoPrimaryInfoRenderer'), get_all=False))
+
+ is_private = (self._has_badge(badges, BadgeType.AVAILABILITY_PRIVATE)
+ or get_first(video_details, 'isPrivate', expected_type=bool))
+
+ info['availability'] = (
+ 'public' if self._has_badge(badges, BadgeType.AVAILABILITY_PUBLIC)
+ else self._availability(
+ is_private=is_private,
+ needs_premium=(
+ self._has_badge(badges, BadgeType.AVAILABILITY_PREMIUM)
+ or False if initial_data and is_private is not None else None),
+ needs_subscription=(
+ self._has_badge(badges, BadgeType.AVAILABILITY_SUBSCRIPTION)
+ or False if initial_data and is_private is not None else None),
+ needs_auth=info['age_limit'] >= 18,
+ is_unlisted=None if is_private is None else (
+ self._has_badge(badges, BadgeType.AVAILABILITY_UNLISTED)
+ or get_first(microformats, 'isUnlisted', expected_type=bool))))
info['__post_extractor'] = self.extract_comments(master_ytcfg, video_id, contents, webpage)
yield self._extract_video(renderer)
def _rich_entries(self, rich_grid_renderer):
- renderer = try_get(
- rich_grid_renderer, lambda x: x['content']['videoRenderer'], dict) or {}
+ renderer = traverse_obj(
+ rich_grid_renderer, ('content', ('videoRenderer', 'reelItemRenderer')), get_all=False) or {}
video_id = renderer.get('videoId')
if not video_id:
return
yield entry
'''
+ def _report_history_entries(self, renderer):
+ for url in traverse_obj(renderer, (
+ 'rows', ..., 'reportHistoryTableRowRenderer', 'cells', ...,
+ 'reportHistoryTableCellRenderer', 'cell', 'reportHistoryTableTextCellRenderer', 'text', 'runs', ...,
+ 'navigationEndpoint', 'commandMetadata', 'webCommandMetadata', 'url')):
+ yield self.url_result(urljoin('https://www.youtube.com', url), YoutubeIE)
+
def _extract_entries(self, parent_renderer, continuation_list):
# continuation_list is modified in-place with continuation_list = [continuation_token]
continuation_list[:] = [None]
content, 'itemSectionRenderer', 'musicShelfRenderer', 'musicShelfContinuation',
expected_type=dict)
if not is_renderer:
- renderer = content.get('richItemRenderer')
- if renderer:
- for entry in self._rich_entries(renderer):
+ if content.get('richItemRenderer'):
+ for entry in self._rich_entries(content['richItemRenderer']):
yield entry
continuation_list[0] = self._extract_continuation(parent_renderer)
+ elif content.get('reportHistorySectionRenderer'): # https://www.youtube.com/reporthistory
+ table = traverse_obj(content, ('reportHistorySectionRenderer', 'table', 'tableRenderer'))
+ yield from self._report_history_entries(table)
+ continuation_list[0] = self._extract_continuation(table)
continue
+
isr_contents = try_get(is_renderer, lambda x: x['contents'], list) or []
for isr_content in isr_contents:
if not isinstance(isr_content, dict):
# See: https://github.com/ytdl-org/youtube-dl/issues/28702
visitor_data = self._extract_visitor_data(response) or visitor_data
- known_continuation_renderers = {
- 'playlistVideoListContinuation': self._playlist_entries,
- 'gridContinuation': self._grid_entries,
- 'itemSectionContinuation': self._post_thread_continuation_entries,
- 'sectionListContinuation': extract_entries, # for feeds
- }
- continuation_contents = try_get(
- response, lambda x: x['continuationContents'], dict) or {}
- continuation_renderer = None
- for key, value in continuation_contents.items():
- if key not in known_continuation_renderers:
- continue
- continuation_renderer = value
- continuation_list = [None]
- yield from known_continuation_renderers[key](continuation_renderer)
- continuation = continuation_list[0] or self._extract_continuation(continuation_renderer)
- break
- if continuation_renderer:
- continue
-
known_renderers = {
'videoRenderer': (self._grid_entries, 'items'), # for membership tab
'gridPlaylistRenderer': (self._grid_entries, 'items'),
'playlistVideoRenderer': (self._playlist_entries, 'contents'),
'itemSectionRenderer': (extract_entries, 'contents'), # for feeds
'richItemRenderer': (extract_entries, 'contents'), # for hashtag
- 'backstagePostThreadRenderer': (self._post_thread_continuation_entries, 'contents')
+ 'backstagePostThreadRenderer': (self._post_thread_continuation_entries, 'contents'),
+ 'reportHistoryTableRowRenderer': (self._report_history_entries, 'rows'),
+ 'playlistVideoListContinuation': (self._playlist_entries, None),
+ 'gridContinuation': (self._grid_entries, None),
+ 'itemSectionContinuation': (self._post_thread_continuation_entries, None),
+ 'sectionListContinuation': (extract_entries, None), # for feeds
}
- on_response_received = dict_get(response, ('onResponseReceivedActions', 'onResponseReceivedEndpoints'))
- continuation_items = try_get(
- on_response_received, lambda x: x[0]['appendContinuationItemsAction']['continuationItems'], list)
- continuation_item = try_get(continuation_items, lambda x: x[0], dict) or {}
+
+ continuation_items = traverse_obj(response, (
+ ('onResponseReceivedActions', 'onResponseReceivedEndpoints'), ...,
+ 'appendContinuationItemsAction', 'continuationItems'
+ ), 'continuationContents', get_all=False)
+ continuation_item = traverse_obj(continuation_items, 0, None, expected_type=dict, default={})
+
video_items_renderer = None
- for key, value in continuation_item.items():
+ for key in continuation_item.keys():
if key not in known_renderers:
continue
- video_items_renderer = {known_renderers[key][1]: continuation_items}
+ func, parent_key = known_renderers[key]
+ video_items_renderer = {parent_key: continuation_items} if parent_key else continuation_items
continuation_list = [None]
- yield from known_renderers[key][0](video_items_renderer)
+ yield from func(video_items_renderer)
continuation = continuation_list[0] or self._extract_continuation(video_items_renderer)
+
+ if not video_items_renderer:
break
- if video_items_renderer:
- continue
- break
@staticmethod
def _extract_selected_tab(tabs, fatal=True):
if fatal:
raise ExtractorError('Unable to find selected tab')
- def _extract_uploader(self, data):
- uploader = {}
- renderer = self._extract_sidebar_info_renderer(data, 'playlistSidebarSecondaryInfoRenderer') or {}
- owner = try_get(
- renderer, lambda x: x['videoOwner']['videoOwnerRenderer']['title']['runs'][0], dict)
- if owner:
- owner_text = owner.get('text')
- uploader['uploader'] = self._search_regex(
- r'^by (.+) and \d+ others?$', owner_text, 'uploader', default=owner_text)
- uploader['uploader_id'] = try_get(
- owner, lambda x: x['navigationEndpoint']['browseEndpoint']['browseId'], str)
- uploader['uploader_url'] = urljoin(
- 'https://www.youtube.com/',
- try_get(owner, lambda x: x['navigationEndpoint']['browseEndpoint']['canonicalBaseUrl'], str))
- return {k: v for k, v in uploader.items() if v is not None}
-
def _extract_from_tabs(self, item_id, ytcfg, data, tabs):
playlist_id = title = description = channel_url = channel_name = channel_id = None
tags = []
selected_tab = self._extract_selected_tab(tabs)
+ # Deprecated - remove when layout discontinued
primary_sidebar_renderer = self._extract_sidebar_info_renderer(data, 'playlistSidebarPrimaryInfoRenderer')
- renderer = try_get(
+ playlist_header_renderer = traverse_obj(data, ('header', 'playlistHeaderRenderer'), expected_type=dict)
+ metadata_renderer = try_get(
data, lambda x: x['metadata']['channelMetadataRenderer'], dict)
- if renderer:
- channel_name = renderer.get('title')
- channel_url = renderer.get('channelUrl')
- channel_id = renderer.get('externalId')
+ if metadata_renderer:
+ channel_name = metadata_renderer.get('title')
+ channel_url = metadata_renderer.get('channelUrl')
+ channel_id = metadata_renderer.get('externalId')
else:
- renderer = try_get(
+ metadata_renderer = try_get(
data, lambda x: x['metadata']['playlistMetadataRenderer'], dict)
- if renderer:
- title = renderer.get('title')
- description = renderer.get('description', '')
+ if metadata_renderer:
+ title = metadata_renderer.get('title')
+ description = metadata_renderer.get('description', '')
playlist_id = channel_id
- tags = renderer.get('keywords', '').split()
+ tags = metadata_renderer.get('keywords', '').split()
# We can get the uncropped banner/avatar by replacing the crop params with '=s0'
# See: https://github.com/yt-dlp/yt-dlp/issues/2237#issuecomment-1013694714
def _get_uncropped(url):
return url_or_none((url or '').split('=')[0] + '=s0')
- avatar_thumbnails = self._extract_thumbnails(renderer, 'avatar')
+ avatar_thumbnails = self._extract_thumbnails(metadata_renderer, 'avatar')
if avatar_thumbnails:
uncropped_avatar = _get_uncropped(avatar_thumbnails[0]['url'])
if uncropped_avatar:
'preference': -5
})
+ # Deprecated - remove when old layout is discontinued
primary_thumbnails = self._extract_thumbnails(
primary_sidebar_renderer, ('thumbnailRenderer', ('playlistVideoThumbnailRenderer', 'playlistCustomThumbnailRenderer'), 'thumbnail'))
+ playlist_thumbnails = self._extract_thumbnails(
+ playlist_header_renderer, ('playlistHeaderBanner', 'heroPlaylistThumbnailRenderer', 'thumbnail'))
+
if playlist_id is None:
playlist_id = item_id
- playlist_stats = traverse_obj(primary_sidebar_renderer, 'stats')
- last_updated_unix, _ = self._extract_time_text(playlist_stats, 2)
+ # Deprecated - remove primary_sidebar_renderer when old layout discontinued
+ # Playlist stats is a text runs array containing [video count, view count, last updated].
+ # last updated or (view count and last updated) may be missing.
+ playlist_stats = get_first(
+ (primary_sidebar_renderer, playlist_header_renderer), (('stats', 'briefStats', 'numVideosText'),))
+ last_updated_unix = self._parse_time_text(
+ self._get_text(playlist_stats, 2) # deprecated, remove when old layout discontinued
+ or self._get_text(playlist_header_renderer, ('byline', 1, 'playlistBylineRenderer', 'text')))
+
+ view_count = self._get_count(playlist_stats, 1)
+ if view_count is None:
+ view_count = self._get_count(playlist_header_renderer, 'viewCountText')
+
+ playlist_count = self._get_count(playlist_stats, 0)
+ if playlist_count is None:
+ playlist_count = self._get_count(playlist_header_renderer, ('byline', 0, 'playlistBylineRenderer', 'text'))
+
if title is None:
title = self._get_text(data, ('header', 'hashtagHeaderRenderer', 'hashtag')) or playlist_id
title += format_field(selected_tab, 'title', ' - %s')
'uploader': channel_name,
'uploader_id': channel_id,
'uploader_url': channel_url,
- 'thumbnails': primary_thumbnails + avatar_thumbnails + channel_banners,
+ 'thumbnails': (primary_thumbnails or playlist_thumbnails) + avatar_thumbnails + channel_banners,
'tags': tags,
- 'view_count': self._get_count(playlist_stats, 1),
+ 'view_count': view_count,
'availability': self._extract_availability(data),
'modified_date': strftime_or_none(last_updated_unix, '%Y%m%d'),
- 'playlist_count': self._get_count(playlist_stats, 0),
+ 'playlist_count': playlist_count,
'channel_follower_count': self._get_count(data, ('header', ..., 'subscriberCountText')),
}
if not channel_id:
- metadata.update(self._extract_uploader(data))
+ owner = traverse_obj(playlist_header_renderer, 'ownerText')
+ if not owner:
+ # Deprecated
+ owner = traverse_obj(
+ self._extract_sidebar_info_renderer(data, 'playlistSidebarSecondaryInfoRenderer'),
+ ('videoOwner', 'videoOwnerRenderer', 'title'))
+ owner_text = self._get_text(owner)
+ browse_ep = traverse_obj(owner, ('runs', 0, 'navigationEndpoint', 'browseEndpoint')) or {}
+ metadata.update(filter_dict({
+ 'uploader': self._search_regex(r'^by (.+) and \d+ others?$', owner_text, 'uploader', default=owner_text),
+ 'uploader_id': browse_ep.get('browseId'),
+ 'uploader_url': urljoin('https://www.youtube.com', browse_ep.get('canonicalBaseUrl'))
+ }))
+
metadata.update({
'channel': metadata['uploader'],
'channel_id': metadata['uploader_id'],
Note: Unless YouTube tells us explicitly, we do not assume it is public
@param data: response
"""
- is_private = is_unlisted = None
- renderer = self._extract_sidebar_info_renderer(data, 'playlistSidebarPrimaryInfoRenderer') or {}
- badge_labels = self._extract_badges(renderer)
+ sidebar_renderer = self._extract_sidebar_info_renderer(data, 'playlistSidebarPrimaryInfoRenderer') or {}
+ playlist_header_renderer = traverse_obj(data, ('header', 'playlistHeaderRenderer')) or {}
+ player_header_privacy = playlist_header_renderer.get('privacy')
+
+ badges = self._extract_badges(sidebar_renderer)
# Personal playlists, when authenticated, have a dropdown visibility selector instead of a badge
- privacy_dropdown_entries = try_get(
- renderer, lambda x: x['privacyForm']['dropdownFormFieldRenderer']['dropdown']['dropdownRenderer']['entries'], list) or []
- for renderer_dict in privacy_dropdown_entries:
- is_selected = try_get(
- renderer_dict, lambda x: x['privacyDropdownItemRenderer']['isSelected'], bool) or False
- if not is_selected:
- continue
- label = self._get_text(renderer_dict, ('privacyDropdownItemRenderer', 'label'))
- if label:
- badge_labels.add(label.lower())
- break
+ privacy_setting_icon = get_first(
+ (playlist_header_renderer, sidebar_renderer),
+ ('privacyForm', 'dropdownFormFieldRenderer', 'dropdown', 'dropdownRenderer', 'entries',
+ lambda _, v: v['privacyDropdownItemRenderer']['isSelected'], 'privacyDropdownItemRenderer', 'icon', 'iconType'),
+ expected_type=str)
- for badge_label in badge_labels:
- if badge_label == 'unlisted':
- is_unlisted = True
- elif badge_label == 'private':
- is_private = True
- elif badge_label == 'public':
- is_unlisted = is_private = False
- return self._availability(is_private, False, False, False, is_unlisted)
+ microformats_is_unlisted = traverse_obj(
+ data, ('microformat', 'microformatDataRenderer', 'unlisted'), expected_type=bool)
+
+ return (
+ 'public' if (
+ self._has_badge(badges, BadgeType.AVAILABILITY_PUBLIC)
+ or player_header_privacy == 'PUBLIC'
+ or privacy_setting_icon == 'PRIVACY_PUBLIC')
+ else self._availability(
+ is_private=(
+ self._has_badge(badges, BadgeType.AVAILABILITY_PRIVATE)
+ or player_header_privacy == 'PRIVATE' if player_header_privacy is not None
+ else privacy_setting_icon == 'PRIVACY_PRIVATE' if privacy_setting_icon is not None else None),
+ is_unlisted=(
+ self._has_badge(badges, BadgeType.AVAILABILITY_UNLISTED)
+ or player_header_privacy == 'UNLISTED' if player_header_privacy is not None
+ else privacy_setting_icon == 'PRIVACY_UNLISTED' if privacy_setting_icon is not None
+ else microformats_is_unlisted if microformats_is_unlisted is not None else None),
+ needs_subscription=self._has_badge(badges, BadgeType.AVAILABILITY_SUBSCRIPTION) or None,
+ needs_premium=self._has_badge(badges, BadgeType.AVAILABILITY_PREMIUM) or None,
+ needs_auth=False))
@staticmethod
def _extract_sidebar_info_renderer(data, info_renderer, expected_type=dict):
def _reload_with_unavailable_videos(self, item_id, data, ytcfg):
"""
- Get playlist with unavailable videos if the 'show unavailable videos' button exists.
+ Reload playlists with unavailable videos (e.g. private videos, region blocked, etc.)
"""
- browse_id = params = None
- renderer = self._extract_sidebar_info_renderer(data, 'playlistSidebarPrimaryInfoRenderer')
- if not renderer:
+ is_playlist = bool(traverse_obj(
+ data, ('metadata', 'playlistMetadataRenderer'), ('header', 'playlistHeaderRenderer')))
+ if not is_playlist:
return
- menu_renderer = try_get(
- renderer, lambda x: x['menu']['menuRenderer']['items'], list) or []
- for menu_item in menu_renderer:
- if not isinstance(menu_item, dict):
- continue
- nav_item_renderer = menu_item.get('menuNavigationItemRenderer')
- text = try_get(
- nav_item_renderer, lambda x: x['text']['simpleText'], str)
- if not text or text.lower() != 'show unavailable videos':
- continue
- browse_endpoint = try_get(
- nav_item_renderer, lambda x: x['navigationEndpoint']['browseEndpoint'], dict) or {}
- browse_id = browse_endpoint.get('browseId')
- params = browse_endpoint.get('params')
- break
-
headers = self.generate_api_headers(
ytcfg=ytcfg, account_syncid=self._extract_account_syncid(ytcfg, data),
visitor_data=self._extract_visitor_data(data, ytcfg))
query = {
- 'params': params or 'wgYCCAA=',
- 'browseId': browse_id or 'VL%s' % item_id
+ 'params': 'wgYCCAA=',
+ 'browseId': f'VL{item_id}'
}
return self._extract_response(
item_id=item_id, headers=headers, query=query,
check_get_keys='contents', fatal=False, ytcfg=ytcfg,
- note='Downloading API JSON with unavailable videos')
+ note='Redownloading playlist API JSON with unavailable videos')
@functools.cached_property
def skip_webpage(self):
'channel_id': 'UCmlqkdCBesrv2Lak1mF_MxA',
'uploader_url': 'https://www.youtube.com/channel/UCmlqkdCBesrv2Lak1mF_MxA',
'channel_url': 'https://www.youtube.com/channel/UCmlqkdCBesrv2Lak1mF_MxA',
+ 'availability': 'public',
},
'playlist_count': 1,
}, {
'channel_id': 'UCmlqkdCBesrv2Lak1mF_MxA',
'channel_url': 'https://www.youtube.com/channel/UCmlqkdCBesrv2Lak1mF_MxA',
'uploader_url': 'https://www.youtube.com/channel/UCmlqkdCBesrv2Lak1mF_MxA',
+ 'availability': 'public',
},
'playlist_count': 0,
}, {
'channel_id': 'UCEPzS1rYsrkqzSLNp76nrcg',
'channel_url': 'https://www.youtube.com/c/ChRiStIaAn008',
'channel': 'Christiaan008',
+ 'availability': 'public',
},
'playlist_count': 96,
}, {
'view_count': int,
'description': '',
'channel_id': 'UCBABnxM4Ar9ten8Mdjj1j0Q',
+ 'availability': 'public',
},
'playlist_mincount': 1123,
'expected_warnings': [r'[Uu]navailable videos (are|will be) hidden'],
'channel': 'Interstellar Movie',
'description': '',
'modified_date': r're:\d{8}',
+ 'availability': 'public',
},
'playlist_mincount': 21,
}, {
'channel_url': 'https://www.youtube.com/channel/UCTYLiWFZy8xtPwxFwX9rV7Q',
'channel_id': 'UCTYLiWFZy8xtPwxFwX9rV7Q',
'modified_date': r're:\d{8}',
+ 'availability': 'public',
},
'playlist_mincount': 200,
'expected_warnings': [r'[Uu]navailable videos (are|will be) hidden'],
'uploader_url': 'https://www.youtube.com/c/blanktv',
'modified_date': r're:\d{8}',
'description': '',
+ 'availability': 'public',
},
'playlist_mincount': 1000,
'expected_warnings': [r'[Uu]navailable videos (are|will be) hidden'],
'channel_id': 'UC9-y-6csu5WGm29I7JiwpnA',
'channel_url': 'https://www.youtube.com/user/Computerphile',
'channel': 'Computerphile',
+ 'availability': 'public',
+ 'modified_date': '20190712',
},
'playlist_mincount': 11,
}, {
'channel_id': 'UC_aEa8K-EOJ3D6gOs7HcyNg',
'tags': [],
'channel': 'NoCopyrightSounds',
+ 'availability': 'public',
},
'playlist_mincount': 166,
'expected_warnings': [r'[Uu]navailable videos (are|will be) hidden'],
'modified_date': r're:\d{8}',
'uploader_url': 'https://www.youtube.com/channel/UC9ALqqC4aIeG5iDs7i90Bfw',
'description': '',
+ 'availability': 'public',
},
'expected_warnings': [
'The URL does not have a videos tab',
'channel': 'Royalty Free Music - Topic',
'view_count': int,
'uploader_url': 'https://www.youtube.com/channel/UC9ALqqC4aIeG5iDs7i90Bfw',
+ 'availability': 'public',
},
'expected_warnings': [
'does not have a videos tab',
'uploader_url': 'https://www.youtube.com/channel/UCKcqXmCcyqnhgpA5P0oHH_Q',
},
'playlist_mincount': 2
+ }, {
+ 'note': 'translated tab name',
+ 'url': 'https://www.youtube.com/channel/UCiu-3thuViMebBjw_5nWYrA/playlists',
+ 'info_dict': {
+ 'id': 'UCiu-3thuViMebBjw_5nWYrA',
+ 'tags': [],
+ 'uploader_id': 'UCiu-3thuViMebBjw_5nWYrA',
+ 'channel_url': 'https://www.youtube.com/channel/UCiu-3thuViMebBjw_5nWYrA',
+ 'description': '',
+ 'title': 'cole-dlp-test-acc - 再生リスト',
+ 'uploader_url': 'https://www.youtube.com/channel/UCiu-3thuViMebBjw_5nWYrA',
+ 'uploader': 'cole-dlp-test-acc',
+ 'channel_id': 'UCiu-3thuViMebBjw_5nWYrA',
+ 'channel': 'cole-dlp-test-acc',
+ 'channel_follower_count': int,
+ },
+ 'playlist_mincount': 1,
+ 'params': {'extractor_args': {'youtube': {'lang': ['ja']}}},
+ 'expected_warnings': ['Preferring "ja"'],
+ }, {
+ # XXX: this should really check flat playlist entries, but the test suite doesn't support that
+ 'note': 'preferred lang set with playlist with translated video titles',
+ 'url': 'https://www.youtube.com/playlist?list=PLt5yu3-wZAlQAaPZ5Z-rJoTdbT-45Q7c0',
+ 'info_dict': {
+ 'id': 'PLt5yu3-wZAlQAaPZ5Z-rJoTdbT-45Q7c0',
+ 'tags': [],
+ 'view_count': int,
+ 'channel_url': 'https://www.youtube.com/channel/UCiu-3thuViMebBjw_5nWYrA',
+ 'uploader': 'cole-dlp-test-acc',
+ 'uploader_id': 'UCiu-3thuViMebBjw_5nWYrA',
+ 'channel': 'cole-dlp-test-acc',
+ 'channel_id': 'UCiu-3thuViMebBjw_5nWYrA',
+ 'description': 'test',
+ 'uploader_url': 'https://www.youtube.com/channel/UCiu-3thuViMebBjw_5nWYrA',
+ 'title': 'dlp test playlist',
+ 'availability': 'public',
+ },
+ 'playlist_mincount': 1,
+ 'params': {'extractor_args': {'youtube': {'lang': ['ja']}}},
+ 'expected_warnings': ['Preferring "ja"'],
+ }, {
+ # shorts audio pivot for 2GtVksBMYFM.
+ 'url': 'https://www.youtube.com/feed/sfv_audio_pivot?bp=8gUrCikSJwoLMkd0VmtzQk1ZRk0SCzJHdFZrc0JNWUZNGgsyR3RWa3NCTVlGTQ==',
+ 'info_dict': {
+ 'id': 'sfv_audio_pivot',
+ 'title': 'sfv_audio_pivot',
+ 'tags': [],
+ },
+ 'playlist_mincount': 50,
+
}]
@classmethod
tabs = traverse_obj(data, ('contents', 'twoColumnBrowseResultsRenderer', 'tabs'), expected_type=list)
if tabs:
selected_tab = self._extract_selected_tab(tabs)
- selected_tab_name = selected_tab.get('title', '').lower()
+ selected_tab_url = urljoin(
+ url, traverse_obj(selected_tab, ('endpoint', 'commandMetadata', 'webCommandMetadata', 'url')))
+ translated_tab_name = selected_tab.get('title', '').lower()
+
+ # Prefer tab name from tab url as it is always in en,
+ # but only when preferred lang is set as it may not extract reliably in all cases.
+ selected_tab_name = (self._preferred_lang in (None, 'en') and translated_tab_name
+ or selected_tab_url and get_mobj(selected_tab_url)['tab'][1:] # primary
+ or translated_tab_name)
+
if selected_tab_name == 'home':
selected_tab_name = 'featured'
requested_tab_name = mobj['tab'][1:]
+
if 'no-youtube-channel-redirect' not in compat_opts:
if requested_tab_name == 'live': # Live tab should have redirected to the video
raise UserNotLive(video_id=mobj['id'])
'channel': 'milan',
'channel_id': 'UCEI1-PVPcYXjB73Hfelbmaw',
'uploader_url': 'https://www.youtube.com/channel/UCEI1-PVPcYXjB73Hfelbmaw',
+ 'availability': 'public',
},
'expected_warnings': [r'[Uu]navailable videos (are|will be) hidden'],
}, {
'uploader_url': 'https://www.youtube.com/c/愛低音的國王',
'channel_id': 'UC21nz3_MesPLqtDqwdvnoxA',
'modified_date': r're:\d{8}',
+ 'availability': 'public',
},
'expected_warnings': [r'[Uu]navailable videos (are|will be) hidden'],
}, {
title = self._search_regex(
rf'{re.escape(channel or "")}[^:]+: (.+)', notification_title,
'video title', default=None)
- upload_date = (strftime_or_none(self._extract_time_text(notification, 'sentTimeText')[0], '%Y%m%d')
- if self._configuration_arg('approximate_date', ie_key=YoutubeTabIE.ie_key())
- else None)
+ timestamp = (self._parse_time_text(self._get_text(notification, 'sentTimeText'))
+ if self._configuration_arg('approximate_date', ie_key=YoutubeTabIE)
+ else None)
return {
'_type': 'url',
'url': url,
'channel_id': channel_id,
'channel': channel,
'thumbnails': self._extract_thumbnails(notification, 'videoThumbnail'),
- 'upload_date': upload_date,
+ 'timestamp': timestamp,
}
def _notification_menu_entries(self, ytcfg):
ie=YoutubeTabIE, video_id=playlist_id)
+class YoutubeShortsAudioPivotIE(InfoExtractor):
+ IE_DESC = 'YouTube Shorts audio pivot (Shorts using audio of a given video)'
+ IE_NAME = 'youtube:shorts:pivot:audio'
+ _VALID_URL = r'https?://(?:www\.)?youtube\.com/source/(?P<id>[\w-]{11})/shorts'
+ _TESTS = [{
+ 'url': 'https://www.youtube.com/source/Lyj-MZSAA9o/shorts',
+ 'only_matching': True,
+ }]
+
+ @staticmethod
+ def _generate_audio_pivot_params(video_id):
+ """
+ Generates sfv_audio_pivot browse params for this video id
+ """
+ pb_params = b'\xf2\x05+\n)\x12\'\n\x0b%b\x12\x0b%b\x1a\x0b%b' % ((video_id.encode(),) * 3)
+ return urllib.parse.quote(base64.b64encode(pb_params).decode())
+
+ def _real_extract(self, url):
+ video_id = self._match_id(url)
+ return self.url_result(
+ f'https://www.youtube.com/feed/sfv_audio_pivot?bp={self._generate_audio_pivot_params(video_id)}',
+ ie=YoutubeTabIE)
+
+
class YoutubeTruncatedURLIE(InfoExtractor):
IE_NAME = 'youtube:truncated_url'
IE_DESC = False # Do not list
projects / yt-dlp.git / blobdiff