FormatSorter,
GeoRestrictedError,
GeoUtils,
+ HEADRequest,
LenientJSONDecoder,
RegexNotFoundError,
RetryManager,
str_to_int,
strip_or_none,
traverse_obj,
+ truncate_string,
try_call,
try_get,
unescapeHTML,
update_url_query,
url_basename,
url_or_none,
+ urlhandle_detect_ext,
urljoin,
variadic,
xpath_element,
is parsed from a string (in case of
fragmented media)
for MSS - URL of the ISM manifest.
+ * request_data Data to send in POST request to the URL
* manifest_url
The URL of the manifest file in case of
fragmented media:
* no_resume The server does not support resuming the
(HTTP or RTMP) download. Boolean.
* has_drm The format has DRM and cannot be downloaded. Boolean
+ * extra_param_to_segment_url A query string to append to each
+ fragment's URL, or to update each existing query string
+ with. Only applied by the native HLS/DASH downloaders.
+ * hls_aes A dictionary of HLS AES-128 decryption information
+ used by the native HLS downloader to override the
+ values in the media playlist when an '#EXT-X-KEY' tag
+ is present in the playlist:
+ * uri The URI from which the key will be downloaded
+ * key The key (as hex) used to decrypt fragments.
+ If `key` is given, any key URI will be ignored
+ * iv The IV (as hex) used to decrypt fragments
* downloader_options A dictionary of downloader options
(For internal use only)
* http_chunk_size Chunk size for HTTP downloads
'unlisted' or 'public'. Use 'InfoExtractor._availability'
to set it
_old_archive_ids: A list of old archive ids needed for backward compatibility
+ _format_sort_fields: A list of fields to use for sorting formats
__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.
for _ in range(2):
try:
self.initialize()
- self.write_debug('Extracting URL: %s' % url)
+ self.to_screen('Extracting URL: %s' % (
+ url if self.get_param('verbose') else truncate_string(url, 100, 20)))
ie_result = self._real_extract(url)
if ie_result is None:
return None
except UnsupportedError:
raise
except ExtractorError as e:
- kwargs = {
- 'video_id': e.video_id or self.get_temp_id(url),
- 'ie': self.IE_NAME,
- 'tb': e.traceback or sys.exc_info()[2],
- 'expected': e.expected,
- 'cause': e.cause
- }
- if hasattr(e, 'countries'):
- kwargs['countries'] = e.countries
- raise type(e)(e.orig_msg, **kwargs)
+ e.video_id = e.video_id or self.get_temp_id(url),
+ e.ie = e.ie or self.IE_NAME,
+ e.traceback = e.traceback or sys.exc_info()[2]
+ raise
except http.client.IncompleteRead as e:
raise ExtractorError('A network error has occurred.', cause=e, expected=True, video_id=self.get_temp_id(url))
except (KeyError, StopIteration) as e:
Like _search_regex, but strips HTML tags and unescapes entities.
"""
res = self._search_regex(pattern, string, name, default, fatal, flags, group)
- if res:
- return clean_html(res).strip()
- else:
- return res
+ if isinstance(res, tuple):
+ return tuple(map(clean_html, res))
+ return clean_html(res)
def _get_netrc_login_info(self, netrc_machine=None):
username = None
# Helper functions for extracting OpenGraph info
@staticmethod
def _og_regexes(prop):
- content_re = r'content=(?:"([^"]+?)"|\'([^\']+?)\'|\s*([^\s"\'=<>`]+?))'
+ content_re = r'content=(?:"([^"]+?)"|\'([^\']+?)\'|\s*([^\s"\'=<>`]+?)(?=\s|/?>))'
property_re = (r'(?:name|property)=(?:\'og%(sep)s%(prop)s\'|"og%(sep)s%(prop)s"|\s*og%(sep)s%(prop)s\b)'
% {'prop': re.escape(prop), 'sep': '(?::|[:-])'})
template = r'<meta[^>]+?%s[^>]+?%s'
# 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>',
+ r'>[^<]*you acknowledge you are at least (\d+) years old',
+ r'>\s*(?:18\s+U(?:\.S\.C\.|SC)\s+)?(?:ยง+\s*)?2257\b',
]
- if any(re.search(marker, html) for marker in AGE_LIMIT_MARKERS):
- return 18
- return 0
+
+ age_limit = 0
+ for marker in AGE_LIMIT_MARKERS:
+ mobj = re.search(marker, html)
+ if mobj:
+ age_limit = max(age_limit, int(traverse_obj(mobj, 1, default=18)))
+ return age_limit
def _media_rating_search(self, html):
# See http://www.tjg-designs.com/WP/metadata-code-examples-adding-metadata-to-your-web-pages/
if js is None:
return {}
- args = dict(zip(arg_keys.split(','), arg_vals.split(',')))
-
- for key, val in args.items():
- if val in ('undefined', 'void 0'):
- args[key] = 'null'
+ args = dict(zip(arg_keys.split(','), map(json.dumps, self._parse_json(
+ f'[{arg_vals}]', video_id, transform_source=js_to_json, fatal=fatal) or ())))
ret = self._parse_json(js, video_id, transform_source=functools.partial(js_to_json, vars=args), fatal=fatal)
return traverse_obj(ret, traverse) or {}
return FormatSort
def _sort_formats(self, formats, field_preference=[]):
- if not formats:
+ if not field_preference:
+ self._downloader.deprecation_warning(
+ 'yt_dlp.InfoExtractor._sort_formats is deprecated and is no longer required')
return
- formats.sort(key=FormatSorter(self._downloader, field_preference).calculate_preference)
+ self._downloader.deprecation_warning(
+ 'yt_dlp.InfoExtractor._sort_formats is deprecated and no longer works as expected. '
+ 'Return _format_sort_fields in the info_dict instead')
+ if formats:
+ formats[0]['__sort_fields'] = field_preference
def _check_formats(self, formats, video_id):
if formats:
def _extract_f4m_formats(self, manifest_url, video_id, preference=None, quality=None, f4m_id=None,
transform_source=lambda s: fix_xml_ampersands(s).strip(),
fatal=True, m3u8_id=None, data=None, headers={}, query={}):
+ if self.get_param('ignore_no_formats_error'):
+ fatal = False
+
res = self._download_xml_handle(
manifest_url, video_id, 'Downloading f4m manifest',
'Unable to download f4m manifest',
errnote=None, fatal=True, live=False, data=None, headers={},
query={}):
+ if self.get_param('ignore_no_formats_error'):
+ fatal = False
+
+ if not m3u8_url:
+ if errnote is not False:
+ errnote = errnote or 'Failed to obtain m3u8 URL'
+ if fatal:
+ raise ExtractorError(errnote, video_id=video_id)
+ self.report_warning(f'{errnote}{bug_reports_message()}')
+ return [], {}
+
res = self._download_webpage_handle(
m3u8_url, video_id,
note='Downloading m3u8 information' if note is None else note,
'protocol': entry_protocol,
'preference': preference,
'quality': quality,
+ 'has_drm': has_drm,
'vcodec': 'none' if media_type == 'AUDIO' else None,
} for idx in _extract_m3u8_playlist_indices(manifest_url))
'protocol': entry_protocol,
'preference': preference,
'quality': quality,
+ 'has_drm': has_drm,
}
resolution = last_stream_inf.get('RESOLUTION')
if resolution:
return self._parse_m3u8_vod_duration(m3u8_vod or '', video_id)
def _parse_m3u8_vod_duration(self, m3u8_vod, video_id):
- if '#EXT-X-PLAYLIST-TYPE:VOD' not in m3u8_vod:
+ if '#EXT-X-ENDLIST' not in m3u8_vod:
return None
return int(sum(
float(line[len('#EXTINF:'):].split(',')[0])
for line in m3u8_vod.splitlines() if line.startswith('#EXTINF:'))) or None
+ def _extract_mpd_vod_duration(
+ self, mpd_url, video_id, note=None, errnote=None, data=None, headers={}, query={}):
+
+ mpd_doc = self._download_xml(
+ mpd_url, video_id,
+ note='Downloading MPD VOD manifest' if note is None else note,
+ errnote='Failed to download VOD manifest' if errnote is None else errnote,
+ fatal=False, data=data, headers=headers, query=query) or {}
+ return int_or_none(parse_duration(mpd_doc.get('mediaPresentationDuration')))
+
@staticmethod
def _xpath_ns(path, namespace=None):
if not namespace:
return '/'.join(out)
def _extract_smil_formats_and_subtitles(self, smil_url, video_id, fatal=True, f4m_params=None, transform_source=None):
+ if self.get_param('ignore_no_formats_error'):
+ fatal = False
+
res = self._download_smil(smil_url, video_id, fatal=fatal, transform_source=transform_source)
if res is False:
assert not fatal
height = int_or_none(medium.get('height'))
proto = medium.get('proto')
ext = medium.get('ext')
- src_ext = determine_ext(src)
+ src_ext = determine_ext(src, default_ext=None) or ext or urlhandle_detect_ext(
+ self._request_webpage(HEADRequest(src), video_id, note='Requesting extension info', fatal=False))
streamer = medium.get('streamer') or base
if proto == 'rtmp' or streamer.startswith('rtmp'):
'width': int_or_none(location.get(xpath_with_ns('s1:width', NS_MAP))),
'height': int_or_none(location.get(xpath_with_ns('s1:height', NS_MAP))),
})
- self._sort_formats(formats)
entries.append({
'id': playlist_id,
def _extract_mpd_formats_and_subtitles(
self, mpd_url, video_id, mpd_id=None, note=None, errnote=None,
fatal=True, data=None, headers={}, query={}):
+
+ if self.get_param('ignore_no_formats_error'):
+ fatal = False
+
res = self._download_xml_handle(
mpd_url, video_id,
note='Downloading MPD manifest' if note is None else note,
return fmts
def _extract_ism_formats_and_subtitles(self, ism_url, video_id, ism_id=None, note=None, errnote=None, fatal=True, data=None, headers={}, query={}):
+ if self.get_param('ignore_no_formats_error'):
+ fatal = False
+
res = self._download_xml_handle(
ism_url, video_id,
note='Downloading ISM manifest' if note is None else note,
def _find_jwplayer_data(self, webpage, video_id=None, transform_source=js_to_json):
mobj = re.search(
- r'(?s)jwplayer\((?P<quote>[\'"])[^\'" ]+(?P=quote)\)(?!</script>).*?\.setup\s*\((?P<options>[^)]+)\)',
+ r'''(?s)jwplayer\s*\(\s*(?P<q>'|")(?!(?P=q)).+(?P=q)\s*\)(?!</script>).*?\.\s*setup\s*\(\s*(?P<options>(?:\([^)]*\)|[^)])+)\s*\)''',
webpage)
if mobj:
try:
def _parse_jwplayer_data(self, jwplayer_data, video_id=None, require_title=True,
m3u8_id=None, mpd_id=None, rtmp_params=None, base_url=None):
- # JWPlayer backward compatibility: flattened playlists
- # https://github.com/jwplayer/jwplayer/blob/v7.4.3/src/js/api/config.js#L81-L96
- if 'playlist' not in jwplayer_data:
- jwplayer_data = {'playlist': [jwplayer_data]}
-
entries = []
+ if not isinstance(jwplayer_data, dict):
+ return entries
- # JWPlayer backward compatibility: single playlist item
+ playlist_items = jwplayer_data.get('playlist')
+ # JWPlayer backward compatibility: single playlist item/flattened playlists
# https://github.com/jwplayer/jwplayer/blob/v7.7.0/src/js/playlist/playlist.js#L10
- if not isinstance(jwplayer_data['playlist'], list):
- jwplayer_data['playlist'] = [jwplayer_data['playlist']]
+ # https://github.com/jwplayer/jwplayer/blob/v7.4.3/src/js/api/config.js#L81-L96
+ if not isinstance(playlist_items, list):
+ playlist_items = (playlist_items or jwplayer_data, )
- for video_data in jwplayer_data['playlist']:
+ for video_data in playlist_items:
+ if not isinstance(video_data, dict):
+ continue
# JWPlayer backward compatibility: flattened sources
# https://github.com/jwplayer/jwplayer/blob/v7.4.3/src/js/playlist/item.js#L29-L35
if 'sources' not in video_data:
'timestamp': int_or_none(video_data.get('pubdate')),
'duration': float_or_none(jwplayer_data.get('duration') or video_data.get('duration')),
'subtitles': subtitles,
+ 'alt_title': clean_html(video_data.get('subtitle')), # attributes used e.g. by Tele5 ...
+ 'genre': clean_html(video_data.get('genre')),
+ 'channel': clean_html(dict_get(video_data, ('category', 'channel'))),
+ 'season_number': int_or_none(video_data.get('season')),
+ 'episode_number': int_or_none(video_data.get('episode')),
+ 'release_year': int_or_none(video_data.get('releasedate')),
+ 'age_limit': int_or_none(video_data.get('age_restriction')),
}
# https://github.com/jwplayer/jwplayer/blob/master/src/js/utils/validator.js#L32
if len(formats) == 1 and re.search(r'^(?:http|//).*(?:youtube\.com|youtu\.be)/.+', formats[0]['url']):
'url': formats[0]['url'],
})
else:
- self._sort_formats(formats)
entry['formats'] = formats
entries.append(entry)
if len(entries) == 1:
def _parse_jwplayer_formats(self, jwplayer_sources_data, video_id=None,
m3u8_id=None, mpd_id=None, rtmp_params=None, base_url=None):
- urls = []
+ urls = set()
formats = []
for source in jwplayer_sources_data:
if not isinstance(source, dict):
base_url, self._proto_relative_url(source.get('file')))
if not source_url or source_url in urls:
continue
- urls.append(source_url)
+ urls.add(source_url)
source_type = source.get('type') or ''
ext = mimetype2ext(source_type) or determine_ext(source_url)
- if source_type == 'hls' or ext == 'm3u8':
+ if source_type == 'hls' or ext == 'm3u8' or 'format=m3u8-aapl' in source_url:
formats.extend(self._extract_m3u8_formats(
source_url, video_id, 'mp4', entry_protocol='m3u8_native',
m3u8_id=m3u8_id, fatal=False))
- elif source_type == 'dash' or ext == 'mpd':
+ elif source_type == 'dash' or ext == 'mpd' or 'format=mpd-time-csf' in source_url:
formats.extend(self._extract_mpd_formats(
source_url, video_id, mpd_id=mpd_id, fatal=False))
elif ext == 'smil':
'ext': ext,
})
else:
+ format_id = str_or_none(source.get('label'))
height = int_or_none(source.get('height'))
- if height is None:
+ if height is None and format_id:
# Often no height is provided but there is a label in
# format like "1080p", "720p SD", or 1080.
- height = int_or_none(self._search_regex(
- r'^(\d{3,4})[pP]?(?:\b|$)', str(source.get('label') or ''),
- 'height', default=None))
+ height = parse_resolution(format_id).get('height')
a_format = {
'url': source_url,
'width': int_or_none(source.get('width')),
'tbr': int_or_none(source.get('bitrate'), scale=1000),
'filesize': int_or_none(source.get('filesize')),
'ext': ext,
+ 'format_id': format_id
}
if source_url.startswith('rtmp'):
a_format['ext'] = 'flv'
continue
t['name'] = cls.ie_key()
yield t
+ if getattr(cls, '__wrapped__', None):
+ yield from cls.__wrapped__.get_testcases(include_onlymatching)
@classmethod
def get_webpage_testcases(cls):
tests = vars(cls).get('_WEBPAGE_TESTS', [])
for t in tests:
t['name'] = cls.ie_key()
- return tests
+ yield t
+ if getattr(cls, '__wrapped__', None):
+ yield from cls.__wrapped__.get_webpage_testcases()
@classproperty(cache=True)
def age_limit(cls):
@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)
+ if cls.suitable(url):
+ return {'video': True, 'playlist': False}.get(cls._RETURN_TYPE)
@classmethod
def is_suitable(cls, age_limit):
desc = ''
if cls._NETRC_MACHINE:
if markdown:
- desc += f' [<abbr title="netrc machine"><em>{cls._NETRC_MACHINE}</em></abbr>]'
+ desc += f' [*{cls._NETRC_MACHINE}*](## "netrc machine")'
else:
desc += f' [{cls._NETRC_MACHINE}]'
if cls.IE_DESC is False:
elif cls.IE_DESC:
desc += f' {cls.IE_DESC}'
if cls.SEARCH_KEY:
- desc += f'; "{cls.SEARCH_KEY}:" prefix'
+ desc += f'{";" if cls.IE_DESC else ""} "{cls.SEARCH_KEY}:" prefix'
if search_examples:
_COUNTS = ('', '5', '10', 'all')
desc += f' (e.g. "{cls.SEARCH_KEY}{random.choice(_COUNTS)}:{random.choice(search_examples)}")'
or urllib.parse.unquote(os.path.splitext(url_basename(url))[0])
or default)
+ def _extract_chapters_helper(self, chapter_list, start_function, title_function, duration, strict=True):
+ if not duration:
+ return
+ chapter_list = [{
+ 'start_time': start_function(chapter),
+ 'title': title_function(chapter),
+ } for chapter in chapter_list or []]
+ if not strict:
+ chapter_list.sort(key=lambda c: c['start_time'] or 0)
+
+ chapters = [{'start_time': 0}]
+ for idx, chapter in enumerate(chapter_list):
+ if chapter['start_time'] is None:
+ self.report_warning(f'Incomplete chapter {idx}')
+ elif chapters[-1]['start_time'] <= chapter['start_time'] <= duration:
+ chapters.append(chapter)
+ 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_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_helper(
+ re.findall(sep_re % (duration_re, r'.+?'), description or ''),
+ start_function=lambda x: parse_duration(x[0]), title_function=lambda x: x[1],
+ duration=duration, strict=False) or self._extract_chapters_helper(
+ re.findall(sep_re % (r'.+?', duration_re), description or ''),
+ start_function=lambda x: parse_duration(x[1]), title_function=lambda x: x[0],
+ duration=duration, strict=False)
+
@staticmethod
def _availability(is_private=None, needs_premium=None, needs_subscription=None, needs_auth=None, is_unlisted=None):
all_known = all(map(
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
+ cls.PLUGIN_NAME, cls.ie_key = plugin_name, super_class.ie_key
+ cls.IE_NAME = f'{super_class.IE_NAME}+{plugin_name}'
while getattr(super_class, '__wrapped__', None):
super_class = super_class.__wrapped__
setattr(sys.modules[super_class.__module__], super_class.__name__, cls)
+ _PLUGIN_OVERRIDES[super_class].append(cls)
return super().__init_subclass__(**kwargs)
def _real_extract(self, url):
raise UnsupportedError(url)
+
+
+_PLUGIN_OVERRIDES = collections.defaultdict(list)