import http.client
import http.cookiejar
import http.cookies
+import inspect
import itertools
import json
import math
import netrc
import os
import random
+import re
import sys
import time
+import types
import urllib.parse
import urllib.request
import xml.etree.ElementTree
-from ..compat import functools, re # isort: split
+from ..compat import functools # isort: split
from ..compat import compat_etree_fromstring, compat_expanduser, compat_os_name
+from ..cookies import LenientSimpleCookie
from ..downloader import FileDownloader
from ..downloader.f4m import get_base_url, remove_encrypted_media
from ..utils import (
+ IDENTITY,
JSON_LD_RE,
NO_DEFAULT,
ExtractorError,
GeoUtils,
LenientJSONDecoder,
RegexNotFoundError,
+ RetryManager,
UnsupportedError,
age_restricted,
base_url,
parse_m3u8_attributes,
parse_resolution,
sanitize_filename,
+ sanitize_url,
sanitized_Request,
+ smuggle_url,
str_or_none,
str_to_int,
strip_or_none,
("3D" or "DASH video")
* width Width of the video, if known
* height Height of the video, if known
+ * aspect_ratio Aspect ratio of the video, if known
+ Automatically calculated from width and height
* resolution Textual description of width and height
+ Automatically calculated from width and height
* dynamic_range The dynamic range of the video. One of:
"SDR" (None), "HDR10", "HDR10+, "HDR12", "HLG, "DV"
* tbr Average bitrate of audio and video in KBit/s
* abr Average audio bitrate in KBit/s
* acodec Name of the audio codec in use
* asr Audio sampling rate in Hertz
+ * audio_channels Number of audio channels
* vbr Average video bitrate in KBit/s
* fps Frame rate
* vcodec Name of the video codec in use
captions instead of normal subtitles
duration: Length of the video in seconds, as an integer or float.
view_count: How many users have watched the video on the platform.
+ concurrent_view_count: How many users are currently watching the video on the platform.
like_count: Number of positive ratings of the video
dislike_count: Number of negative ratings of the video
repost_count: Number of reposts of the video
live stream that goes on instead of a fixed-length video.
was_live: True, False, or None (=unknown). Whether this video was
originally a live stream.
- live_status: 'is_live', 'is_upcoming', 'was_live', 'not_live' or None (=unknown)
+ live_status: None (=unknown), 'is_live', 'is_upcoming', 'was_live', 'not_live',
+ or 'post_live' (was live, but VOD is not yet processed)
If absent, automatically set from is_live, was_live
start_time: Time in seconds where the reproduction should start, as
specified in the URL.
playable_in_embed: Whether this video is allowed to play in embedded
players on other sites. Can be True (=always allowed),
False (=never allowed), None (=unknown), or a string
- specifying the criteria for embedability (Eg: 'whitelist')
+ specifying the criteria for embedability; e.g. 'whitelist'
availability: Under what condition the video is available. One of
'private', 'premium_only', 'subscriber_only', 'needs_auth',
'unlisted' or 'public'. Use 'InfoExtractor._availability'
to set it
+ _old_archive_ids: A list of old archive ids needed for backward compatibility
__post_extractor: A function to be called just before the metadata is
written to either disk, logger or console. The function
must return a dict which will be added to the info_dict.
section_start: Start time of the section in seconds
section_end: End time of the section in seconds
+ The following fields should only be set for storyboards:
+ rows: Number of rows in each storyboard fragment, as an integer
+ columns: Number of columns in each storyboard fragment, as an integer
+
Unless mentioned otherwise, the fields should be Unicode strings.
Unless mentioned otherwise, None is equivalent to absence of information.
title, description etc.
- Subclasses of this should define a _VALID_URL regexp and, re-define the
- _real_extract() and (optionally) _real_initialize() methods.
- Probably, they should also be added to the list of extractors.
+ Subclasses of this should also be added to the list of extractors and
+ should define a _VALID_URL regexp and, re-define the _real_extract() and
+ (optionally) _real_initialize() methods.
Subclasses may also override suitable() if necessary, but ensure the function
signature is preserved and that this function imports everything it needs
(except other extractors), so that lazy_extractors works correctly.
+ Subclasses can define a list of _EMBED_REGEX, which will be searched for in
+ the HTML of Generic webpages. It may also override _extract_embed_urls
+ or _extract_from_webpage as necessary. While these are normally classmethods,
+ _extract_from_webpage is allowed to be an instance method.
+
+ _extract_from_webpage may raise self.StopExtraction() to stop further
+ processing of the webpage and obtain exclusive rights to it. This is useful
+ when the extractor cannot reliably be matched using just the URL,
+ e.g. invidious/peertube instances
+
+ Embed-only extractors can be defined by setting _VALID_URL = False.
+
To support username + password (or netrc) login, the extractor must define a
_NETRC_MACHINE and re-define _perform_login(username, password) and
(optionally) _initialize_pre_login() methods. The _perform_login method will
will be used by geo restriction bypass mechanism similarly
to _GEO_COUNTRIES.
+ The _ENABLED attribute should be set to False for IEs that
+ are disabled by default and must be explicitly enabled.
+
The _WORKING attribute should be set to False for broken IEs
in order to warn the users and skip the tests.
"""
_GEO_COUNTRIES = None
_GEO_IP_BLOCKS = None
_WORKING = True
+ _ENABLED = True
_NETRC_MACHINE = None
IE_DESC = None
SEARCH_KEY = None
+ _VALID_URL = None
+ _EMBED_REGEX = []
def _login_hint(self, method=NO_DEFAULT, netrc=None):
password_hint = f'--username and --password, or --netrc ({netrc or self._NETRC_MACHINE}) to provide account credentials'
'password': f'Use {password_hint}',
'cookies': (
'Use --cookies-from-browser or --cookies for the authentication. '
- 'See https://github.com/ytdl-org/youtube-dl#how-do-i-pass-cookies-to-youtube-dl for how to manually pass cookies'),
+ 'See https://github.com/yt-dlp/yt-dlp/wiki/FAQ#how-do-i-pass-cookies-to-yt-dlp for how to manually pass cookies'),
}[method if method is not NO_DEFAULT else 'any' if self.supports_login() else 'cookies']
def __init__(self, downloader=None):
@classmethod
def _match_valid_url(cls, url):
+ if cls._VALID_URL is False:
+ return None
# This does not use has/getattr intentionally - we want to know whether
# we have cached the regexp for *this* class, whereas getattr would also
# match the superclass
if '_VALID_URL_RE' not in cls.__dict__:
- if '_VALID_URL' not in cls.__dict__:
- cls._VALID_URL = cls._make_valid_url()
cls._VALID_URL_RE = re.compile(cls._VALID_URL)
return cls._VALID_URL_RE.match(url)
return None
if self._x_forwarded_for_ip:
ie_result['__x_forwarded_for_ip'] = self._x_forwarded_for_ip
- subtitles = ie_result.get('subtitles')
- if (subtitles and 'live_chat' in subtitles
- and 'no-live-chat' in self.get_param('compat_opts', [])):
- del subtitles['live_chat']
+ subtitles = ie_result.get('subtitles') or {}
+ if 'no-live-chat' in self.get_param('compat_opts'):
+ for lang in ('live_chat', 'comments', 'danmaku'):
+ subtitles.pop(lang, None)
return ie_result
except GeoRestrictedError as e:
if self.__maybe_fake_ip_and_retry(e.countries):
return content
- def _parse_xml(self, xml_string, video_id, transform_source=None, fatal=True):
+ def __print_error(self, errnote, fatal, video_id, err):
+ if fatal:
+ raise ExtractorError(f'{video_id}: {errnote}', cause=err)
+ elif errnote:
+ self.report_warning(f'{video_id}: {errnote}: {err}')
+
+ def _parse_xml(self, xml_string, video_id, transform_source=None, fatal=True, errnote=None):
if transform_source:
xml_string = transform_source(xml_string)
try:
return compat_etree_fromstring(xml_string.encode('utf-8'))
except xml.etree.ElementTree.ParseError as ve:
- errmsg = '%s: Failed to parse XML ' % video_id
- if fatal:
- raise ExtractorError(errmsg, cause=ve)
- else:
- self.report_warning(errmsg + str(ve))
+ self.__print_error('Failed to parse XML' if errnote is None else errnote, fatal, video_id, ve)
- def _parse_json(self, json_string, video_id, transform_source=None, fatal=True, **parser_kwargs):
+ def _parse_json(self, json_string, video_id, transform_source=None, fatal=True, errnote=None, **parser_kwargs):
try:
return json.loads(
json_string, cls=LenientJSONDecoder, strict=False, transform_source=transform_source, **parser_kwargs)
except ValueError as ve:
- errmsg = f'{video_id}: Failed to parse JSON'
- if fatal:
- raise ExtractorError(errmsg, cause=ve)
- else:
- self.report_warning(f'{errmsg}: {ve}')
+ self.__print_error('Failed to parse JSON' if errnote is None else errnote, fatal, video_id, ve)
- def _parse_socket_response_as_json(self, data, video_id, transform_source=None, fatal=True):
- return self._parse_json(
- data[data.find('{'):data.rfind('}') + 1],
- video_id, transform_source, fatal)
+ def _parse_socket_response_as_json(self, data, *args, **kwargs):
+ return self._parse_json(data[data.find('{'):data.rfind('}') + 1], *args, **kwargs)
def __create_download_methods(name, parser, note, errnote, return_value):
- def parse(ie, content, *args, **kwargs):
+ def parse(ie, content, *args, errnote=errnote, **kwargs):
if parser is None:
return content
+ if errnote is False:
+ kwargs['errnote'] = errnote
# parser is fetched by name so subclasses can override it
return getattr(ie, parser)(content, *args, **kwargs)
if res is False:
return res
content, urlh = res
- return parse(self, content, video_id, transform_source=transform_source, fatal=fatal), urlh
+ return parse(self, content, video_id, transform_source=transform_source, fatal=fatal, errnote=errnote), urlh
def download_content(self, url_or_request, video_id, note=note, errnote=errnote, transform_source=None,
fatal=True, encoding=None, data=None, headers={}, query={}, expected_status=None):
self.report_warning(f'Unable to load request from disk: {e}')
else:
content = self.__decode_webpage(webpage_bytes, encoding, url_or_request.headers)
- return parse(self, content, video_id, transform_source, fatal)
+ return parse(self, content, video_id, transform_source=transform_source, fatal=fatal, errnote=errnote)
kwargs = {
'note': note,
'errnote': errnote,
return self._downloader.params.get(name, default, *args, **kwargs)
return default
- def report_drm(self, video_id, partial=False):
+ def report_drm(self, video_id, partial=NO_DEFAULT):
+ if partial is not NO_DEFAULT:
+ self._downloader.deprecation_warning('InfoExtractor.report_drm no longer accepts the argument partial')
self.raise_no_formats('This video is DRM protected', expected=True, video_id=video_id)
def report_extraction(self, id_or_name):
'url': url,
}
- def playlist_from_matches(self, matches, playlist_id=None, playlist_title=None, getter=None, ie=None, video_kwargs=None, **kwargs):
- urls = (self.url_result(self._proto_relative_url(m), ie, **(video_kwargs or {}))
- for m in orderedSet(map(getter, matches) if getter else matches))
- return self.playlist_result(urls, playlist_id, playlist_title, **kwargs)
+ @classmethod
+ def playlist_from_matches(cls, matches, playlist_id=None, playlist_title=None,
+ getter=IDENTITY, ie=None, video_kwargs=None, **kwargs):
+ return cls.playlist_result(
+ (cls.url_result(m, ie, **(video_kwargs or {})) for m in orderedSet(map(getter, matches), lazy=True)),
+ playlist_id, playlist_title, **kwargs)
@staticmethod
def playlist_result(entries, playlist_id=None, playlist_title=None, playlist_description=None, *, multi_video=False, **kwargs):
return None
def _search_json(self, start_pattern, string, name, video_id, *, end_pattern='',
- contains_pattern='(?s:.+)', fatal=True, default=NO_DEFAULT, **kwargs):
+ contains_pattern=r'{(?s:.+)}', fatal=True, default=NO_DEFAULT, **kwargs):
"""Searches string for the JSON object specified by start_pattern"""
# NB: end_pattern is only used to reduce the size of the initial match
if default is NO_DEFAULT:
fatal, has_default = False, True
json_string = self._search_regex(
- rf'{start_pattern}\s*(?P<json>{{\s*{contains_pattern}\s*}})\s*{end_pattern}',
+ rf'(?:{start_pattern})\s*(?P<json>{contains_pattern})\s*(?:{end_pattern})',
string, name, group='json', fatal=fatal, default=None if has_default else NO_DEFAULT)
if not json_string:
return default
def _dc_search_uploader(self, html):
return self._html_search_meta('dc.creator', html, 'uploader')
- def _rta_search(self, html):
+ @staticmethod
+ def _rta_search(html):
# See http://www.rtalabel.org/index.php?content=howtofaq#single
if re.search(r'(?ix)<meta\s+name="rating"\s+'
r' content="RTA-5042-1996-1400-1577-RTA"',
html):
return 18
+
+ # And then there are the jokers who advertise that they use RTA, but actually don't.
+ AGE_LIMIT_MARKERS = [
+ r'Proudly Labeled <a href="http://www\.rtalabel\.org/" title="Restricted to Adults">RTA</a>',
+ ]
+ if any(re.search(marker, html) for marker in AGE_LIMIT_MARKERS):
+ return 18
return 0
def _media_rating_search(self, html):
if not json_ld:
return {}
info = {}
- if not isinstance(json_ld, (list, tuple, dict)):
- return info
- if isinstance(json_ld, dict):
- json_ld = [json_ld]
INTERACTION_TYPE_MAP = {
'CommentAction': 'comment',
info['chapters'] = chapters
def extract_video_object(e):
- assert is_type(e, 'VideoObject')
author = e.get('author')
info.update({
'url': url_or_none(e.get('contentUrl')),
+ 'ext': mimetype2ext(e.get('encodingFormat')),
'title': unescapeHTML(e.get('name')),
'description': unescapeHTML(e.get('description')),
- 'thumbnails': [{'url': url}
+ 'thumbnails': [{'url': unescapeHTML(url)}
for url in variadic(traverse_obj(e, 'thumbnailUrl', 'thumbnailURL'))
if url_or_none(url)],
'duration': parse_duration(e.get('duration')),
# however some websites are using 'Text' type instead.
# 1. https://schema.org/VideoObject
'uploader': author.get('name') if isinstance(author, dict) else author if isinstance(author, str) else None,
+ 'artist': traverse_obj(e, ('byArtist', 'name'), expected_type=str),
'filesize': int_or_none(float_or_none(e.get('contentSize'))),
'tbr': int_or_none(e.get('bitrate')),
'width': int_or_none(e.get('width')),
'height': int_or_none(e.get('height')),
'view_count': int_or_none(e.get('interactionCount')),
+ 'tags': try_call(lambda: e.get('keywords').split(',')),
})
+ if is_type(e, 'AudioObject'):
+ info.update({
+ 'vcodec': 'none',
+ 'abr': int_or_none(e.get('bitrate')),
+ })
extract_interaction_statistic(e)
extract_chapter_information(e)
def traverse_json_ld(json_ld, at_top_level=True):
- for e in json_ld:
+ for e in variadic(json_ld):
+ if not isinstance(e, dict):
+ continue
if at_top_level and '@context' not in e:
continue
if at_top_level and set(e.keys()) == {'@context', '@graph'}:
- traverse_json_ld(variadic(e['@graph'], allowed_types=(dict,)), at_top_level=False)
- break
+ traverse_json_ld(e['@graph'], at_top_level=False)
+ continue
if expected_type is not None and not is_type(e, expected_type):
continue
rating = traverse_obj(e, ('aggregateRating', 'ratingValue'), expected_type=float_or_none)
extract_video_object(e['video'][0])
elif is_type(traverse_obj(e, ('subjectOf', 0)), 'VideoObject'):
extract_video_object(e['subjectOf'][0])
- elif is_type(e, 'VideoObject'):
+ elif is_type(e, 'VideoObject', 'AudioObject'):
extract_video_object(e)
if expected_type is None:
continue
continue
else:
break
- traverse_json_ld(json_ld)
+ traverse_json_ld(json_ld)
return filter_dict(info)
def _search_nextjs_data(self, webpage, video_id, *, transform_source=None, fatal=True, **kw):
FUNCTION_RE = r'\(function\((?P<arg_keys>.*?)\){return\s+(?P<js>{.*?})\s*;?\s*}\((?P<arg_vals>.*?)\)'
js, arg_keys, arg_vals = self._search_regex(
(rf'<script>\s*window\.{rectx}={FUNCTION_RE}\s*\)\s*;?\s*</script>', rf'{rectx}\(.*?{FUNCTION_RE}'),
- webpage, context_name, group=('js', 'arg_keys', 'arg_vals'), fatal=fatal)
+ webpage, context_name, group=('js', 'arg_keys', 'arg_vals'),
+ default=NO_DEFAULT if fatal else (None, None, None))
+ if js is None:
+ return {}
args = dict(zip(arg_keys.split(','), arg_vals.split(',')))
regex = r' *((?P<reverse>\+)?(?P<field>[a-zA-Z0-9_]+)((?P<separator>[~:])(?P<limit>.*?))?)? *$'
default = ('hidden', 'aud_or_vid', 'hasvid', 'ie_pref', 'lang', 'quality',
- 'res', 'fps', 'hdr:12', 'codec:vp9.2', 'size', 'br', 'asr',
- 'proto', 'ext', 'hasaud', 'source', 'id') # These must not be aliases
+ 'res', 'fps', 'hdr:12', 'vcodec:vp9.2', 'channels', 'acodec',
+ 'size', 'br', 'asr', 'proto', 'ext', 'hasaud', 'source', 'id') # These must not be aliases
ytdl_default = ('hasaud', 'lang', 'quality', 'tbr', 'filesize', 'vbr',
'height', 'width', 'proto', 'vext', 'abr', 'aext',
'fps', 'fs_approx', 'source', 'id')
'order_free': ('webm', 'mp4', 'flv', '', 'none')},
'aext': {'type': 'ordered', 'field': 'audio_ext',
'order': ('m4a', 'aac', 'mp3', 'ogg', 'opus', 'webm', '', 'none'),
- 'order_free': ('opus', 'ogg', 'webm', 'm4a', 'mp3', 'aac', '', 'none')},
+ 'order_free': ('ogg', 'opus', 'webm', 'mp3', 'm4a', 'aac', '', 'none')},
'hidden': {'visible': False, 'forced': True, 'type': 'extractor', 'max': -1000},
'aud_or_vid': {'visible': False, 'forced': True, 'type': 'multiple',
'field': ('vcodec', 'acodec'),
'height': {'convert': 'float_none'},
'width': {'convert': 'float_none'},
'fps': {'convert': 'float_none'},
+ 'channels': {'convert': 'float_none', 'field': 'audio_channels'},
'tbr': {'convert': 'float_none'},
'vbr': {'convert': 'float_none'},
'abr': {'convert': 'float_none'},
'res': {'type': 'multiple', 'field': ('height', 'width'),
'function': lambda it: (lambda l: min(l) if l else 0)(tuple(filter(None, it)))},
- # For compatibility with youtube-dl
+ # Actual field names
'format_id': {'type': 'alias', 'field': 'id'},
'preference': {'type': 'alias', 'field': 'ie_pref'},
'language_preference': {'type': 'alias', 'field': 'lang'},
'source_preference': {'type': 'alias', 'field': 'source'},
'protocol': {'type': 'alias', 'field': 'proto'},
'filesize_approx': {'type': 'alias', 'field': 'fs_approx'},
+ 'audio_channels': {'type': 'alias', 'field': 'channels'},
# Deprecated
'dimension': {'type': 'alias', 'field': 'res', 'deprecated': True},
if field not in self.settings:
if key in ('forced', 'priority'):
return False
- self.ydl.deprecation_warning(
- f'Using arbitrary fields ({field}) for format sorting is deprecated '
- 'and may be removed in a future version')
+ self.ydl.deprecated_feature(f'Using arbitrary fields ({field}) for format sorting is '
+ 'deprecated and may be removed in a future version')
self.settings[field] = {}
propObj = self.settings[field]
if key not in propObj:
if self._get_field_setting(field, 'type') == 'alias':
alias, field = field, self._get_field_setting(field, 'field')
if self._get_field_setting(alias, 'deprecated'):
- self.ydl.deprecation_warning(
- f'Format sorting alias {alias} is deprecated '
- f'and may be removed in a future version. Please use {field} instead')
+ self.ydl.deprecated_feature(f'Format sorting alias {alias} is deprecated and may '
+ f'be removed in a future version. Please use {field} instead')
reverse = match.group('reverse') is not None
closest = match.group('separator') == '~'
limit_text = match.group('limit')
else 'https:')
def _proto_relative_url(self, url, scheme=None):
- if url is None:
- return url
- if url.startswith('//'):
- if scheme is None:
- scheme = self.http_scheme()
- return scheme + url
- else:
- return url
+ scheme = scheme or self.http_scheme()
+ assert scheme.endswith(':')
+ return sanitize_url(url, scheme=scheme[:-1])
def _sleep(self, timeout, video_id, msg_template=None):
if msg_template is None:
audio_group_id = last_stream_inf.get('AUDIO')
# As per [1, 4.3.4.1.1] any EXT-X-STREAM-INF tag which
# references a rendition group MUST have a CODECS attribute.
- # However, this is not always respected, for example, [2]
+ # However, this is not always respected. E.g. [2]
# contains EXT-X-STREAM-INF tag which references AUDIO
# rendition group but does not have CODECS and despite
# referencing an audio group it represents a complete
def prepare_template(template_name, identifiers):
tmpl = representation_ms_info[template_name]
+ if representation_id is not None:
+ tmpl = tmpl.replace('$RepresentationID$', representation_id)
# First of, % characters outside $...$ templates
# must be escaped by doubling for proper processing
# by % operator string formatting used further (see
t += c
# Next, $...$ templates are translated to their
# %(...) counterparts to be used with % operator
- if representation_id is not None:
- t = t.replace('$RepresentationID$', representation_id)
t = re.sub(r'\$(%s)\$' % '|'.join(identifiers), r'%(\1)d', t)
t = re.sub(r'\$(%s)%%([^$]+)\$' % '|'.join(identifiers), r'%(\1)\2', t)
t.replace('$$', '$')
segment_number += 1
segment_time += segment_d
elif 'segment_urls' in representation_ms_info and 's' in representation_ms_info:
- # No media template
- # Example: https://www.youtube.com/watch?v=iXZV5uAYMJI
+ # No media template,
+ # e.g. https://www.youtube.com/watch?v=iXZV5uAYMJI
# or any YouTube dashsegments video
fragments = []
segment_index = 0
representation_ms_info['fragments'] = fragments
elif 'segment_urls' in representation_ms_info:
# Segment URLs with no SegmentTimeline
- # Example: https://www.seznam.cz/zpravy/clanek/cesko-zasahne-vitr-o-sile-vichrice-muze-byt-i-zivotu-nebezpecny-39091
+ # E.g. https://www.seznam.cz/zpravy/clanek/cesko-zasahne-vitr-o-sile-vichrice-muze-byt-i-zivotu-nebezpecny-39091
# https://github.com/ytdl-org/youtube-dl/pull/14844
fragments = []
segment_duration = float_or_none(
stream_name = stream.get('Name')
stream_language = stream.get('Language', 'und')
for track in stream.findall('QualityLevel'):
- fourcc = track.get('FourCC') or ('AACL' if track.get('AudioTag') == '255' else None)
+ KNOWN_TAGS = {'255': 'AACL', '65534': 'EC-3'}
+ fourcc = track.get('FourCC') or KNOWN_TAGS.get(track.get('AudioTag'))
# TODO: add support for WVC1 and WMAP
- if fourcc not in ('H264', 'AVC1', 'AACL', 'TTML'):
+ if fourcc not in ('H264', 'AVC1', 'AACL', 'TTML', 'EC-3'):
self.report_warning('%s is not a supported codec' % fourcc)
continue
tbr = int(track.attrib['Bitrate']) // 1000
media_tags.extend(re.findall(
# We only allow video|audio followed by a whitespace or '>'.
# Allowing more characters may end up in significant slow down (see
- # https://github.com/ytdl-org/youtube-dl/issues/11979, example URL:
- # http://www.porntrex.com/maps/videositemap.xml).
+ # https://github.com/ytdl-org/youtube-dl/issues/11979,
+ # e.g. http://www.porntrex.com/maps/videositemap.xml).
r'(?s)(<(?P<tag>%s)(?:\s+[^>]*)?>)(.*?)</(?P=tag)>' % _MEDIA_TAG_NAME_RE, webpage))
for media_tag, _, media_type, media_content in media_tags:
media_info = {
'subtitles': {},
}
media_attributes = extract_attributes(media_tag)
- src = strip_or_none(media_attributes.get('src'))
+ src = strip_or_none(dict_get(media_attributes, ('src', 'data-video-src', 'data-src', 'data-source')))
if src:
f = parse_content_type(media_attributes.get('type'))
_, formats = _media_formats(src, media_type, f)
s_attr = extract_attributes(source_tag)
# data-video-src and data-src are non standard but seen
# several times in the wild
- src = strip_or_none(dict_get(s_attr, ('src', 'data-video-src', 'data-src')))
+ src = strip_or_none(dict_get(s_attr, ('src', 'data-video-src', 'data-src', 'data-source')))
if not src:
continue
f = parse_content_type(s_attr.get('type'))
'url': source_url,
'width': int_or_none(source.get('width')),
'height': height,
- 'tbr': int_or_none(source.get('bitrate')),
+ 'tbr': int_or_none(source.get('bitrate'), scale=1000),
+ 'filesize': int_or_none(source.get('filesize')),
'ext': ext,
}
if source_url.startswith('rtmp'):
def _get_cookies(self, url):
""" Return a http.cookies.SimpleCookie with the cookies for the url """
- return http.cookies.SimpleCookie(self._downloader._calc_cookies(url))
+ return LenientSimpleCookie(self._downloader._calc_cookies(url))
def _apply_first_set_cookie_header(self, url_handle, cookie):
"""
@classmethod
def get_testcases(cls, include_onlymatching=False):
- t = getattr(cls, '_TEST', None)
+ # Do not look in super classes
+ t = vars(cls).get('_TEST')
if t:
assert not hasattr(cls, '_TESTS'), f'{cls.ie_key()}IE has _TEST and _TESTS'
tests = [t]
else:
- tests = getattr(cls, '_TESTS', [])
+ tests = vars(cls).get('_TESTS', [])
for t in tests:
if not include_onlymatching and t.get('only_matching', False):
continue
t['name'] = cls.ie_key()
yield t
- @classproperty
+ @classmethod
+ def get_webpage_testcases(cls):
+ tests = vars(cls).get('_WEBPAGE_TESTS', [])
+ for t in tests:
+ t['name'] = cls.ie_key()
+ return tests
+
+ @classproperty(cache=True)
def age_limit(cls):
"""Get age limit from the testcases"""
return max(traverse_obj(
- tuple(cls.get_testcases(include_onlymatching=False)),
+ (*cls.get_testcases(include_onlymatching=False), *cls.get_webpage_testcases()),
(..., (('playlist', 0), None), 'info_dict', 'age_limit')) or [0])
+ @classproperty(cache=True)
+ def _RETURN_TYPE(cls):
+ """What the extractor returns: "video", "playlist", "any", or None (Unknown)"""
+ tests = tuple(cls.get_testcases(include_onlymatching=False))
+ if not tests:
+ return None
+ elif not any(k.startswith('playlist') for test in tests for k in test):
+ return 'video'
+ elif all(any(k.startswith('playlist') for k in test) for test in tests):
+ return 'playlist'
+ return 'any'
+
+ @classmethod
+ def is_single_video(cls, url):
+ """Returns whether the URL is of a single video, None if unknown"""
+ assert cls.suitable(url), 'The URL must be suitable for the extractor'
+ return {'video': True, 'playlist': False}.get(cls._RETURN_TYPE)
+
@classmethod
def is_suitable(cls, age_limit):
"""Test whether the extractor is generally suitable for the given age limit"""
desc += f'; "{cls.SEARCH_KEY}:" prefix'
if search_examples:
_COUNTS = ('', '5', '10', 'all')
- desc += f' (Example: "{cls.SEARCH_KEY}{random.choice(_COUNTS)}:{random.choice(search_examples)}")'
+ desc += f' (e.g. "{cls.SEARCH_KEY}{random.choice(_COUNTS)}:{random.choice(search_examples)}")'
if not cls.working():
desc += ' (**Currently broken**)' if markdown else ' (Currently broken)'
- name = f' - **{cls.IE_NAME}**' if markdown else cls.IE_NAME
+ # Escape emojis. Ref: https://github.com/github/markup/issues/1153
+ name = (' - **%s**' % re.sub(r':(\w+:)', ':\u200B\\g<1>', cls.IE_NAME)) if markdown else cls.IE_NAME
return f'{name}:{desc}' if desc else name
def extract_subtitles(self, *args, **kwargs):
def _get_subtitles(self, *args, **kwargs):
raise NotImplementedError('This method must be implemented by subclasses')
+ class CommentsDisabled(Exception):
+ """Raise in _get_comments if comments are disabled for the video"""
+
def extract_comments(self, *args, **kwargs):
if not self.get_param('getcomments'):
return None
interrupted = False
except KeyboardInterrupt:
self.to_screen('Interrupted by user')
+ except self.CommentsDisabled:
+ return {'comments': None, 'comment_count': None}
except Exception as e:
if self.get_param('ignoreerrors') is not True:
raise
headers['Ytdl-request-proxy'] = geo_verification_proxy
return headers
- def _generic_id(self, url):
+ @staticmethod
+ def _generic_id(url):
return urllib.parse.unquote(os.path.splitext(url.rstrip('/').split('/')[-1])[0])
- def _generic_title(self, url):
- return urllib.parse.unquote(os.path.splitext(url_basename(url))[0])
+ def _generic_title(self, url='', webpage='', *, default=None):
+ return (self._og_search_title(webpage, default=None)
+ or self._html_extract_title(webpage, default=None)
+ or urllib.parse.unquote(os.path.splitext(url_basename(url))[0])
+ or default)
@staticmethod
def _availability(is_private=None, needs_premium=None, needs_subscription=None, needs_auth=None, is_unlisted=None):
@param default The default value to return when the key is not present (default: [])
@param casesense When false, the values are converted to lower case
'''
- val = traverse_obj(
- self._downloader.params, ('extractor_args', (ie_key or self.ie_key()).lower(), key))
+ ie_key = ie_key if isinstance(ie_key, str) else (ie_key or self).ie_key()
+ val = traverse_obj(self._downloader.params, ('extractor_args', ie_key.lower(), key))
if val is None:
return [] if default is NO_DEFAULT else default
return list(val) if casesense else [x.lower() for x in val]
self.to_screen(f'Downloading {playlist_label}{playlist_id} - add --no-playlist to download just the {video_label}{video_id}')
return True
+ def _error_or_warning(self, err, _count=None, _retries=0, *, fatal=True):
+ RetryManager.report_retry(
+ err, _count or int(fatal), _retries,
+ info=self.to_screen, warn=self.report_warning, error=None if fatal else self.report_warning,
+ sleep_func=self.get_param('retry_sleep_functions', {}).get('extractor'))
+
+ def RetryManager(self, **kwargs):
+ return RetryManager(self.get_param('extractor_retries', 3), self._error_or_warning, **kwargs)
+
+ def _extract_generic_embeds(self, url, *args, info_dict={}, note='Extracting generic embeds', **kwargs):
+ display_id = traverse_obj(info_dict, 'display_id', 'id')
+ self.to_screen(f'{format_field(display_id, None, "%s: ")}{note}')
+ return self._downloader.get_info_extractor('Generic')._extract_embeds(
+ smuggle_url(url, {'block_ies': [self.ie_key()]}), *args, **kwargs)
+
+ @classmethod
+ def extract_from_webpage(cls, ydl, url, webpage):
+ ie = (cls if isinstance(cls._extract_from_webpage, types.MethodType)
+ else ydl.get_info_extractor(cls.ie_key()))
+ for info in ie._extract_from_webpage(url, webpage) or []:
+ # url = None since we do not want to set (webpage/original)_url
+ ydl.add_default_extra_info(info, ie, None)
+ yield info
+
+ @classmethod
+ def _extract_from_webpage(cls, url, webpage):
+ for embed_url in orderedSet(
+ cls._extract_embed_urls(url, webpage) or [], lazy=True):
+ yield cls.url_result(embed_url, None if cls._VALID_URL is False else cls)
+
+ @classmethod
+ def _extract_embed_urls(cls, url, webpage):
+ """@returns all the embed urls on the webpage"""
+ if '_EMBED_URL_RE' not in cls.__dict__:
+ assert isinstance(cls._EMBED_REGEX, (list, tuple))
+ for idx, regex in enumerate(cls._EMBED_REGEX):
+ assert regex.count('(?P<url>') == 1, \
+ f'{cls.__name__}._EMBED_REGEX[{idx}] must have exactly 1 url group\n\t{regex}'
+ cls._EMBED_URL_RE = tuple(map(re.compile, cls._EMBED_REGEX))
+
+ for regex in cls._EMBED_URL_RE:
+ for mobj in regex.finditer(webpage):
+ embed_url = urllib.parse.urljoin(url, unescapeHTML(mobj.group('url')))
+ if cls._VALID_URL is False or cls.suitable(embed_url):
+ yield embed_url
+
+ class StopExtraction(Exception):
+ pass
+
+ @classmethod
+ def _extract_url(cls, webpage): # TODO: Remove
+ """Only for compatibility with some older extractors"""
+ return next(iter(cls._extract_embed_urls(None, webpage) or []), None)
+
+ @classmethod
+ def __init_subclass__(cls, *, plugin_name=None, **kwargs):
+ if plugin_name:
+ mro = inspect.getmro(cls)
+ super_class = cls.__wrapped__ = mro[mro.index(cls) + 1]
+ cls.IE_NAME, cls.ie_key = f'{super_class.IE_NAME}+{plugin_name}', super_class.ie_key
+ while getattr(super_class, '__wrapped__', None):
+ super_class = super_class.__wrapped__
+ setattr(sys.modules[super_class.__module__], super_class.__name__, cls)
+
+ return super().__init_subclass__(**kwargs)
+
class SearchInfoExtractor(InfoExtractor):
"""
"""
_MAX_RESULTS = float('inf')
+ _RETURN_TYPE = 'playlist'
- @classmethod
- def _make_valid_url(cls):
+ @classproperty
+ def _VALID_URL(cls):
return r'%s(?P<prefix>|[1-9][0-9]*|all):(?P<query>[\s\S]+)' % cls._SEARCH_KEY
def _real_extract(self, query):
@classproperty
def SEARCH_KEY(cls):
return cls._SEARCH_KEY
+
+
+class UnsupportedURLIE(InfoExtractor):
+ _VALID_URL = '.*'
+ _ENABLED = False
+ IE_DESC = False
+
+ def _real_extract(self, url):
+ raise UnsupportedError(url)