12 import xml
.etree
.ElementTree
14 from ..compat
import functools
, re
# isort: split
15 from ..compat
import (
16 compat_cookiejar_Cookie
,
17 compat_cookies_SimpleCookie
,
18 compat_etree_fromstring
,
25 compat_urllib_parse_unquote
,
26 compat_urllib_parse_urlencode
,
27 compat_urllib_request
,
30 from ..downloader
import FileDownloader
31 from ..downloader
.f4m
import get_base_url
, remove_encrypted_media
65 parse_m3u8_attributes
,
90 """Information Extractor class.
92 Information extractors are the classes that, given a URL, extract
93 information about the video (or videos) the URL refers to. This
94 information includes the real video URL, the video title, author and
95 others. The information is stored in a dictionary which is then
96 passed to the YoutubeDL. The YoutubeDL processes this
97 information possibly downloading the video to the file system, among
98 other possible outcomes.
100 The type field determines the type of the result.
101 By far the most common value (and the default if _type is missing) is
102 "video", which indicates a single video.
104 For a video, the dictionaries must include the following fields:
106 id: Video identifier.
107 title: Video title, unescaped. Set to an empty string if video has
108 no title as opposed to "None" which signifies that the
109 extractor failed to obtain a title
111 Additionally, it must contain either a formats entry or a url one:
113 formats: A list of dictionaries for each format available, ordered
114 from worst to best quality.
117 * url The mandatory URL representing the media:
118 for plain file media - HTTP URL of this file,
120 for HLS - URL of the M3U8 media playlist,
121 for HDS - URL of the F4M manifest,
123 - HTTP URL to plain file media (in case of
125 - URL of the MPD manifest or base URL
126 representing the media if MPD manifest
127 is parsed from a string (in case of
129 for MSS - URL of the ISM manifest.
131 The URL of the manifest file in case of
133 for HLS - URL of the M3U8 master playlist,
134 for HDS - URL of the F4M manifest,
135 for DASH - URL of the MPD manifest,
136 for MSS - URL of the ISM manifest.
137 * manifest_stream_number (For internal use only)
138 The index of the stream in the manifest file
139 * ext Will be calculated from URL if missing
140 * format A human-readable description of the format
141 ("mp4 container with h264/opus").
142 Calculated from the format_id, width, height.
143 and format_note fields if missing.
144 * format_id A short description of the format
145 ("mp4_h264_opus" or "19").
146 Technically optional, but strongly recommended.
147 * format_note Additional info about the format
148 ("3D" or "DASH video")
149 * width Width of the video, if known
150 * height Height of the video, if known
151 * resolution Textual description of width and height
152 * dynamic_range The dynamic range of the video. One of:
153 "SDR" (None), "HDR10", "HDR10+, "HDR12", "HLG, "DV"
154 * tbr Average bitrate of audio and video in KBit/s
155 * abr Average audio bitrate in KBit/s
156 * acodec Name of the audio codec in use
157 * asr Audio sampling rate in Hertz
158 * vbr Average video bitrate in KBit/s
160 * vcodec Name of the video codec in use
161 * container Name of the container format
162 * filesize The number of bytes, if known in advance
163 * filesize_approx An estimate for the number of bytes
164 * player_url SWF Player URL (used for rtmpdump).
165 * protocol The protocol that will be used for the actual
166 download, lower-case. One of "http", "https" or
167 one of the protocols defined in downloader.PROTOCOL_MAP
169 Base URL for fragments. Each fragment's path
170 value (if present) will be relative to
172 * fragments A list of fragments of a fragmented media.
173 Each fragment entry must contain either an url
174 or a path. If an url is present it should be
175 considered by a client. Otherwise both path and
176 fragment_base_url must be present. Here is
177 the list of all potential fields:
178 * "url" - fragment's URL
179 * "path" - fragment's path relative to
181 * "duration" (optional, int or float)
182 * "filesize" (optional, int)
183 * is_from_start Is a live format that can be downloaded
184 from the start. Boolean
185 * preference Order number of this format. If this field is
186 present and not None, the formats get sorted
187 by this field, regardless of all other values.
188 -1 for default (order by other properties),
189 -2 or smaller for less than default.
190 < -1000 to hide the format (if there is
191 another one which is strictly better)
192 * language Language code, e.g. "de" or "en-US".
193 * language_preference Is this in the language mentioned in
195 10 if it's what the URL is about,
196 -1 for default (don't know),
197 -10 otherwise, other values reserved for now.
198 * quality Order number of the video quality of this
199 format, irrespective of the file format.
200 -1 for default (order by other properties),
201 -2 or smaller for less than default.
202 * source_preference Order number for this video source
203 (quality takes higher priority)
204 -1 for default (order by other properties),
205 -2 or smaller for less than default.
206 * http_headers A dictionary of additional HTTP headers
207 to add to the request.
208 * stretched_ratio If given and not 1, indicates that the
209 video's pixels are not square.
210 width : height ratio as float.
211 * no_resume The server does not support resuming the
212 (HTTP or RTMP) download. Boolean.
213 * has_drm The format has DRM and cannot be downloaded. Boolean
214 * downloader_options A dictionary of downloader options
215 (For internal use only)
216 * http_chunk_size Chunk size for HTTP downloads
217 * ffmpeg_args Extra arguments for ffmpeg downloader
218 RTMP formats can also have the additional fields: page_url,
219 app, play_path, tc_url, flash_version, rtmp_live, rtmp_conn,
220 rtmp_protocol, rtmp_real_time
222 url: Final video URL.
223 ext: Video filename extension.
224 format: The video format, defaults to ext (used for --get-format)
225 player_url: SWF Player URL (used for rtmpdump).
227 The following fields are optional:
229 direct: True if a direct video file was given (must only be set by GenericIE)
230 alt_title: A secondary title of the video.
231 display_id An alternative identifier for the video, not necessarily
232 unique, but available before title. Typically, id is
233 something like "4234987", title "Dancing naked mole rats",
234 and display_id "dancing-naked-mole-rats"
235 thumbnails: A list of dictionaries, with the following entries:
236 * "id" (optional, string) - Thumbnail format ID
238 * "preference" (optional, int) - quality of the image
239 * "width" (optional, int)
240 * "height" (optional, int)
241 * "resolution" (optional, string "{width}x{height}",
243 * "filesize" (optional, int)
244 * "http_headers" (dict) - HTTP headers for the request
245 thumbnail: Full URL to a video thumbnail image.
246 description: Full video description.
247 uploader: Full name of the video uploader.
248 license: License name the video is licensed under.
249 creator: The creator of the video.
250 timestamp: UNIX timestamp of the moment the video was uploaded
251 upload_date: Video upload date in UTC (YYYYMMDD).
252 If not explicitly set, calculated from timestamp
253 release_timestamp: UNIX timestamp of the moment the video was released.
254 If it is not clear whether to use timestamp or this, use the former
255 release_date: The date (YYYYMMDD) when the video was released in UTC.
256 If not explicitly set, calculated from release_timestamp
257 modified_timestamp: UNIX timestamp of the moment the video was last modified.
258 modified_date: The date (YYYYMMDD) when the video was last modified in UTC.
259 If not explicitly set, calculated from modified_timestamp
260 uploader_id: Nickname or id of the video uploader.
261 uploader_url: Full URL to a personal webpage of the video uploader.
262 channel: Full name of the channel the video is uploaded on.
263 Note that channel fields may or may not repeat uploader
264 fields. This depends on a particular extractor.
265 channel_id: Id of the channel.
266 channel_url: Full URL to a channel webpage.
267 channel_follower_count: Number of followers of the channel.
268 location: Physical location where the video was filmed.
269 subtitles: The available subtitles as a dictionary in the format
270 {tag: subformats}. "tag" is usually a language code, and
271 "subformats" is a list sorted from lower to higher
272 preference, each element is a dictionary with the "ext"
274 * "data": The subtitles file contents
275 * "url": A URL pointing to the subtitles file
276 It can optionally also have:
277 * "name": Name or description of the subtitles
278 * "http_headers": A dictionary of additional HTTP headers
279 to add to the request.
280 "ext" will be calculated from URL if missing
281 automatic_captions: Like 'subtitles'; contains automatically generated
282 captions instead of normal subtitles
283 duration: Length of the video in seconds, as an integer or float.
284 view_count: How many users have watched the video on the platform.
285 like_count: Number of positive ratings of the video
286 dislike_count: Number of negative ratings of the video
287 repost_count: Number of reposts of the video
288 average_rating: Average rating give by users, the scale used depends on the webpage
289 comment_count: Number of comments on the video
290 comments: A list of comments, each with one or more of the following
291 properties (all but one of text or html optional):
292 * "author" - human-readable name of the comment author
293 * "author_id" - user ID of the comment author
294 * "author_thumbnail" - The thumbnail of the comment author
296 * "html" - Comment as HTML
297 * "text" - Plain text of the comment
298 * "timestamp" - UNIX timestamp of comment
299 * "parent" - ID of the comment this one is replying to.
300 Set to "root" to indicate that this is a
301 comment to the original video.
302 * "like_count" - Number of positive ratings of the comment
303 * "dislike_count" - Number of negative ratings of the comment
304 * "is_favorited" - Whether the comment is marked as
305 favorite by the video uploader
306 * "author_is_uploader" - Whether the comment is made by
308 age_limit: Age restriction for the video, as an integer (years)
309 webpage_url: The URL to the video webpage, if given to yt-dlp it
310 should allow to get the same result again. (It will be set
311 by YoutubeDL if it's missing)
312 categories: A list of categories that the video falls in, for example
314 tags: A list of tags assigned to the video, e.g. ["sweden", "pop music"]
315 cast: A list of the video cast
316 is_live: True, False, or None (=unknown). Whether this video is a
317 live stream that goes on instead of a fixed-length video.
318 was_live: True, False, or None (=unknown). Whether this video was
319 originally a live stream.
320 live_status: 'is_live', 'is_upcoming', 'was_live', 'not_live' or None (=unknown)
321 If absent, automatically set from is_live, was_live
322 start_time: Time in seconds where the reproduction should start, as
323 specified in the URL.
324 end_time: Time in seconds where the reproduction should end, as
325 specified in the URL.
326 chapters: A list of dictionaries, with the following entries:
327 * "start_time" - The start time of the chapter in seconds
328 * "end_time" - The end time of the chapter in seconds
329 * "title" (optional, string)
330 playable_in_embed: Whether this video is allowed to play in embedded
331 players on other sites. Can be True (=always allowed),
332 False (=never allowed), None (=unknown), or a string
333 specifying the criteria for embedability (Eg: 'whitelist')
334 availability: Under what condition the video is available. One of
335 'private', 'premium_only', 'subscriber_only', 'needs_auth',
336 'unlisted' or 'public'. Use 'InfoExtractor._availability'
338 __post_extractor: A function to be called just before the metadata is
339 written to either disk, logger or console. The function
340 must return a dict which will be added to the info_dict.
341 This is usefull for additional information that is
342 time-consuming to extract. Note that the fields thus
343 extracted will not be available to output template and
344 match_filter. So, only "comments" and "comment_count" are
345 currently allowed to be extracted via this method.
347 The following fields should only be used when the video belongs to some logical
350 chapter: Name or title of the chapter the video belongs to.
351 chapter_number: Number of the chapter the video belongs to, as an integer.
352 chapter_id: Id of the chapter the video belongs to, as a unicode string.
354 The following fields should only be used when the video is an episode of some
355 series, programme or podcast:
357 series: Title of the series or programme the video episode belongs to.
358 series_id: Id of the series or programme the video episode belongs to, as a unicode string.
359 season: Title of the season the video episode belongs to.
360 season_number: Number of the season the video episode belongs to, as an integer.
361 season_id: Id of the season the video episode belongs to, as a unicode string.
362 episode: Title of the video episode. Unlike mandatory video title field,
363 this field should denote the exact title of the video episode
364 without any kind of decoration.
365 episode_number: Number of the video episode within a season, as an integer.
366 episode_id: Id of the video episode, as a unicode string.
368 The following fields should only be used when the media is a track or a part of
371 track: Title of the track.
372 track_number: Number of the track within an album or a disc, as an integer.
373 track_id: Id of the track (useful in case of custom indexing, e.g. 6.iii),
375 artist: Artist(s) of the track.
376 genre: Genre(s) of the track.
377 album: Title of the album the track belongs to.
378 album_type: Type of the album (e.g. "Demo", "Full-length", "Split", "Compilation", etc).
379 album_artist: List of all artists appeared on the album (e.g.
380 "Ash Borer / Fell Voices" or "Various Artists", useful for splits
382 disc_number: Number of the disc or other physical medium the track belongs to,
384 release_year: Year (YYYY) when the album was released.
385 composer: Composer of the piece
387 Unless mentioned otherwise, the fields should be Unicode strings.
389 Unless mentioned otherwise, None is equivalent to absence of information.
392 _type "playlist" indicates multiple videos.
393 There must be a key "entries", which is a list, an iterable, or a PagedList
394 object, each element of which is a valid dictionary by this specification.
396 Additionally, playlists can have "id", "title", and any other relevent
397 attributes with the same semantics as videos (see above).
399 It can also have the following optional fields:
401 playlist_count: The total number of videos in a playlist. If not given,
402 YoutubeDL tries to calculate it from "entries"
405 _type "multi_video" indicates that there are multiple videos that
406 form a single show, for examples multiple acts of an opera or TV episode.
407 It must have an entries key like a playlist and contain all the keys
408 required for a video at the same time.
411 _type "url" indicates that the video must be extracted from another
412 location, possibly by a different extractor. Its only required key is:
413 "url" - the next URL to extract.
414 The key "ie_key" can be set to the class name (minus the trailing "IE",
415 e.g. "Youtube") if the extractor class is known in advance.
416 Additionally, the dictionary may have any properties of the resolved entity
417 known in advance, for example "title" if the title of the referred video is
421 _type "url_transparent" entities have the same specification as "url", but
422 indicate that the given additional information is more precise than the one
423 associated with the resolved URL.
424 This is useful when a site employs a video service that hosts the video and
425 its technical metadata, but that video service does not embed a useful
426 title, description etc.
429 Subclasses of this should define a _VALID_URL regexp and, re-define the
430 _real_extract() and (optionally) _real_initialize() methods.
431 Probably, they should also be added to the list of extractors.
433 Subclasses may also override suitable() if necessary, but ensure the function
434 signature is preserved and that this function imports everything it needs
435 (except other extractors), so that lazy_extractors works correctly.
437 To support username + password (or netrc) login, the extractor must define a
438 _NETRC_MACHINE and re-define _perform_login(username, password) and
439 (optionally) _initialize_pre_login() methods. The _perform_login method will
440 be called between _initialize_pre_login and _real_initialize if credentials
441 are passed by the user. In cases where it is necessary to have the login
442 process as part of the extraction rather than initialization, _perform_login
443 can be left undefined.
445 _GEO_BYPASS attribute may be set to False in order to disable
446 geo restriction bypass mechanisms for a particular extractor.
447 Though it won't disable explicit geo restriction bypass based on
448 country code provided with geo_bypass_country.
450 _GEO_COUNTRIES attribute may contain a list of presumably geo unrestricted
451 countries for this extractor. One of these countries will be used by
452 geo restriction bypass mechanism right away in order to bypass
453 geo restriction, of course, if the mechanism is not disabled.
455 _GEO_IP_BLOCKS attribute may contain a list of presumably geo unrestricted
456 IP blocks in CIDR notation for this extractor. One of these IP blocks
457 will be used by geo restriction bypass mechanism similarly
460 The _WORKING attribute should be set to False for broken IEs
461 in order to warn the users and skip the tests.
466 _x_forwarded_for_ip
= None
468 _GEO_COUNTRIES
= None
469 _GEO_IP_BLOCKS
= None
471 _NETRC_MACHINE
= None
475 def _login_hint(self
, method
=NO_DEFAULT
, netrc
=None):
476 password_hint
= f
'--username and --password, or --netrc ({netrc or self._NETRC_MACHINE}) to provide account credentials'
479 'any': f
'Use --cookies, --cookies-from-browser, {password_hint}',
480 'password': f
'Use {password_hint}',
482 'Use --cookies-from-browser or --cookies for the authentication. '
483 'See https://github.com/ytdl-org/youtube-dl#how-do-i-pass-cookies-to-youtube-dl for how to manually pass cookies'),
484 }[method
if method
is not NO_DEFAULT
else 'any' if self
.supports_login() else 'cookies']
486 def __init__(self
, downloader
=None):
487 """Constructor. Receives an optional downloader (a YoutubeDL instance).
488 If a downloader is not passed during initialization,
489 it must be set using "set_downloader()" before "extract()" is called"""
491 self
._x
_forwarded
_for
_ip
= None
492 self
._printed
_messages
= set()
493 self
.set_downloader(downloader
)
496 def _match_valid_url(cls
, url
):
497 # This does not use has/getattr intentionally - we want to know whether
498 # we have cached the regexp for *this* class, whereas getattr would also
499 # match the superclass
500 if '_VALID_URL_RE' not in cls
.__dict
__:
501 if '_VALID_URL' not in cls
.__dict
__:
502 cls
._VALID
_URL
= cls
._make
_valid
_url
()
503 cls
._VALID
_URL
_RE
= re
.compile(cls
._VALID
_URL
)
504 return cls
._VALID
_URL
_RE
.match(url
)
507 def suitable(cls
, url
):
508 """Receives a URL and returns True if suitable for this IE."""
509 # This function must import everything it needs (except other extractors),
510 # so that lazy_extractors works correctly
511 return cls
._match
_valid
_url
(url
) is not None
514 def _match_id(cls
, url
):
515 return cls
._match
_valid
_url
(url
).group('id')
518 def get_temp_id(cls
, url
):
520 return cls
._match
_id
(url
)
521 except (IndexError, AttributeError):
526 """Getter method for _WORKING."""
530 def supports_login(cls
):
531 return bool(cls
._NETRC
_MACHINE
)
533 def initialize(self
):
534 """Initializes an instance (authentication, etc)."""
535 self
._printed
_messages
= set()
536 self
._initialize
_geo
_bypass
({
537 'countries': self
._GEO
_COUNTRIES
,
538 'ip_blocks': self
._GEO
_IP
_BLOCKS
,
541 self
._initialize
_pre
_login
()
542 if self
.supports_login():
543 username
, password
= self
._get
_login
_info
()
545 self
._perform
_login
(username
, password
)
546 elif self
.get_param('username') and False not in (self
.IE_DESC
, self
._NETRC
_MACHINE
):
547 self
.report_warning(f
'Login with password is not supported for this website. {self._login_hint("cookies")}')
548 self
._real
_initialize
()
551 def _initialize_geo_bypass(self
, geo_bypass_context
):
553 Initialize geo restriction bypass mechanism.
555 This method is used to initialize geo bypass mechanism based on faking
556 X-Forwarded-For HTTP header. A random country from provided country list
557 is selected and a random IP belonging to this country is generated. This
558 IP will be passed as X-Forwarded-For HTTP header in all subsequent
561 This method will be used for initial geo bypass mechanism initialization
562 during the instance initialization with _GEO_COUNTRIES and
565 You may also manually call it from extractor's code if geo bypass
566 information is not available beforehand (e.g. obtained during
567 extraction) or due to some other reason. In this case you should pass
568 this information in geo bypass context passed as first argument. It may
569 contain following fields:
571 countries: List of geo unrestricted countries (similar
573 ip_blocks: List of geo unrestricted IP blocks in CIDR notation
574 (similar to _GEO_IP_BLOCKS)
577 if not self
._x
_forwarded
_for
_ip
:
579 # Geo bypass mechanism is explicitly disabled by user
580 if not self
.get_param('geo_bypass', True):
583 if not geo_bypass_context
:
584 geo_bypass_context
= {}
586 # Backward compatibility: previously _initialize_geo_bypass
587 # expected a list of countries, some 3rd party code may still use
589 if isinstance(geo_bypass_context
, (list, tuple)):
590 geo_bypass_context
= {
591 'countries': geo_bypass_context
,
594 # The whole point of geo bypass mechanism is to fake IP
595 # as X-Forwarded-For HTTP header based on some IP block or
598 # Path 1: bypassing based on IP block in CIDR notation
600 # Explicit IP block specified by user, use it right away
601 # regardless of whether extractor is geo bypassable or not
602 ip_block
= self
.get_param('geo_bypass_ip_block', None)
604 # Otherwise use random IP block from geo bypass context but only
605 # if extractor is known as geo bypassable
607 ip_blocks
= geo_bypass_context
.get('ip_blocks')
608 if self
._GEO
_BYPASS
and ip_blocks
:
609 ip_block
= random
.choice(ip_blocks
)
612 self
._x
_forwarded
_for
_ip
= GeoUtils
.random_ipv4(ip_block
)
613 self
.write_debug(f
'Using fake IP {self._x_forwarded_for_ip} as X-Forwarded-For')
616 # Path 2: bypassing based on country code
618 # Explicit country code specified by user, use it right away
619 # regardless of whether extractor is geo bypassable or not
620 country
= self
.get_param('geo_bypass_country', None)
622 # Otherwise use random country code from geo bypass context but
623 # only if extractor is known as geo bypassable
625 countries
= geo_bypass_context
.get('countries')
626 if self
._GEO
_BYPASS
and countries
:
627 country
= random
.choice(countries
)
630 self
._x
_forwarded
_for
_ip
= GeoUtils
.random_ipv4(country
)
631 self
._downloader
.write_debug(
632 f
'Using fake IP {self._x_forwarded_for_ip} ({country.upper()}) as X-Forwarded-For')
634 def extract(self
, url
):
635 """Extracts URL information and returns it in list of dicts."""
640 self
.write_debug('Extracting URL: %s' % url
)
641 ie_result
= self
._real
_extract
(url
)
642 if ie_result
is None:
644 if self
._x
_forwarded
_for
_ip
:
645 ie_result
['__x_forwarded_for_ip'] = self
._x
_forwarded
_for
_ip
646 subtitles
= ie_result
.get('subtitles')
647 if (subtitles
and 'live_chat' in subtitles
648 and 'no-live-chat' in self
.get_param('compat_opts', [])):
649 del subtitles
['live_chat']
651 except GeoRestrictedError
as e
:
652 if self
.__maybe
_fake
_ip
_and
_retry
(e
.countries
):
655 except UnsupportedError
:
657 except ExtractorError
as e
:
659 'video_id': e
.video_id
or self
.get_temp_id(url
),
661 'tb': e
.traceback
or sys
.exc_info()[2],
662 'expected': e
.expected
,
665 if hasattr(e
, 'countries'):
666 kwargs
['countries'] = e
.countries
667 raise type(e
)(e
.orig_msg
, **kwargs
)
668 except compat_http_client
.IncompleteRead
as e
:
669 raise ExtractorError('A network error has occurred.', cause
=e
, expected
=True, video_id
=self
.get_temp_id(url
))
670 except (KeyError, StopIteration) as e
:
671 raise ExtractorError('An extractor error has occurred.', cause
=e
, video_id
=self
.get_temp_id(url
))
673 def __maybe_fake_ip_and_retry(self
, countries
):
674 if (not self
.get_param('geo_bypass_country', None)
676 and self
.get_param('geo_bypass', True)
677 and not self
._x
_forwarded
_for
_ip
679 country_code
= random
.choice(countries
)
680 self
._x
_forwarded
_for
_ip
= GeoUtils
.random_ipv4(country_code
)
681 if self
._x
_forwarded
_for
_ip
:
683 'Video is geo restricted. Retrying extraction with fake IP %s (%s) as X-Forwarded-For.'
684 % (self
._x
_forwarded
_for
_ip
, country_code
.upper()))
688 def set_downloader(self
, downloader
):
689 """Sets a YoutubeDL instance as the downloader for this IE."""
690 self
._downloader
= downloader
692 def _initialize_pre_login(self
):
693 """ Intialization before login. Redefine in subclasses."""
696 def _perform_login(self
, username
, password
):
697 """ Login with username and password. Redefine in subclasses."""
700 def _real_initialize(self
):
701 """Real initialization process. Redefine in subclasses."""
704 def _real_extract(self
, url
):
705 """Real extraction process. Redefine in subclasses."""
706 raise NotImplementedError('This method must be implemented by subclasses')
710 """A string for getting the InfoExtractor with get_info_extractor"""
711 return cls
.__name
__[:-2]
715 return cls
.__name
__[:-2]
718 def __can_accept_status_code(err
, expected_status
):
719 assert isinstance(err
, compat_urllib_error
.HTTPError
)
720 if expected_status
is None:
722 elif callable(expected_status
):
723 return expected_status(err
.code
) is True
725 return err
.code
in variadic(expected_status
)
727 def _create_request(self
, url_or_request
, data
=None, headers
={}, query={}
):
728 if isinstance(url_or_request
, compat_urllib_request
.Request
):
729 return update_Request(url_or_request
, data
=data
, headers
=headers
, query
=query
)
731 url_or_request
= update_url_query(url_or_request
, query
)
732 return sanitized_Request(url_or_request
, data
, headers
)
734 def _request_webpage(self
, url_or_request
, video_id
, note
=None, errnote
=None, fatal
=True, data
=None, headers
={}, query={}
, expected_status
=None):
736 Return the response handle.
738 See _download_webpage docstring for arguments specification.
740 if not self
._downloader
._first
_webpage
_request
:
741 sleep_interval
= self
.get_param('sleep_interval_requests') or 0
742 if sleep_interval
> 0:
743 self
.to_screen('Sleeping %s seconds ...' % sleep_interval
)
744 time
.sleep(sleep_interval
)
746 self
._downloader
._first
_webpage
_request
= False
749 self
.report_download_webpage(video_id
)
750 elif note
is not False:
752 self
.to_screen(str(note
))
754 self
.to_screen(f
'{video_id}: {note}')
756 # Some sites check X-Forwarded-For HTTP header in order to figure out
757 # the origin of the client behind proxy. This allows bypassing geo
758 # restriction by faking this header's value to IP that belongs to some
759 # geo unrestricted country. We will do so once we encounter any
760 # geo restriction error.
761 if self
._x
_forwarded
_for
_ip
:
762 if 'X-Forwarded-For' not in headers
:
763 headers
['X-Forwarded-For'] = self
._x
_forwarded
_for
_ip
766 return self
._downloader
.urlopen(self
._create
_request
(url_or_request
, data
, headers
, query
))
767 except network_exceptions
as err
:
768 if isinstance(err
, compat_urllib_error
.HTTPError
):
769 if self
.__can
_accept
_status
_code
(err
, expected_status
):
770 # Retain reference to error to prevent file object from
771 # being closed before it can be read. Works around the
772 # effects of <https://bugs.python.org/issue15002>
773 # introduced in Python 3.4.1.
780 errnote
= 'Unable to download webpage'
782 errmsg
= f
'{errnote}: {error_to_compat_str(err)}'
784 raise ExtractorError(errmsg
, cause
=err
)
786 self
.report_warning(errmsg
)
789 def _download_webpage_handle(self
, url_or_request
, video_id
, note
=None, errnote
=None, fatal
=True,
790 encoding
=None, data
=None, headers
={}, query={}
, expected_status
=None):
792 Return a tuple (page content as string, URL handle).
795 url_or_request -- plain text URL as a string or
796 a compat_urllib_request.Requestobject
797 video_id -- Video/playlist/item identifier (string)
800 note -- note printed before downloading (string)
801 errnote -- note printed in case of an error (string)
802 fatal -- flag denoting whether error should be considered fatal,
803 i.e. whether it should cause ExtractionError to be raised,
804 otherwise a warning will be reported and extraction continued
805 encoding -- encoding for a page content decoding, guessed automatically
806 when not explicitly specified
807 data -- POST data (bytes)
808 headers -- HTTP headers (dict)
809 query -- URL query (dict)
810 expected_status -- allows to accept failed HTTP requests (non 2xx
811 status code) by explicitly specifying a set of accepted status
812 codes. Can be any of the following entities:
813 - an integer type specifying an exact failed status code to
815 - a list or a tuple of integer types specifying a list of
816 failed status codes to accept
817 - a callable accepting an actual failed status code and
818 returning True if it should be accepted
819 Note that this argument does not affect success status codes (2xx)
820 which are always accepted.
823 # Strip hashes from the URL (#1038)
824 if isinstance(url_or_request
, (compat_str
, str)):
825 url_or_request
= url_or_request
.partition('#')[0]
827 urlh
= self
._request
_webpage
(url_or_request
, video_id
, note
, errnote
, fatal
, data
=data
, headers
=headers
, query
=query
, expected_status
=expected_status
)
831 content
= self
._webpage
_read
_content
(urlh
, url_or_request
, video_id
, note
, errnote
, fatal
, encoding
=encoding
)
832 return (content
, urlh
)
835 def _guess_encoding_from_content(content_type
, webpage_bytes
):
836 m
= re
.match(r
'[a-zA-Z0-9_.-]+/[a-zA-Z0-9_.-]+\s*;\s*charset=(.+)', content_type
)
838 encoding
= m
.group(1)
840 m
= re
.search(br
'<meta[^>]+charset=[\'"]?([^\'")]+)[ /\'">]',
841 webpage_bytes[:1024])
843 encoding = m.group(1).decode('ascii')
844 elif webpage_bytes.startswith(b'\xff\xfe'):
851 def __check_blocked(self, content):
852 first_block = content[:512]
853 if ('<title>Access to this site is blocked</title>' in content
854 and 'Websense' in first_block):
855 msg = 'Access to this webpage has been blocked by Websense filtering software in your network.'
856 blocked_iframe = self._html_search_regex(
857 r'<iframe src="([^
"]+)"', content,
858 'Websense information URL
', default=None)
860 msg += ' Visit
%s for more details
' % blocked_iframe
861 raise ExtractorError(msg, expected=True)
862 if '<title
>The URL you requested has been blocked
</title
>' in first_block:
864 'Access to this webpage has been blocked by Indian censorship
. '
865 'Use a VPN
or proxy
server (with --proxy
) to route around it
.')
866 block_msg = self._html_search_regex(
867 r'</h1
><p
>(.*?
)</p
>',
868 content, 'block message
', default=None)
870 msg += ' (Message
: "%s")' % block_msg.replace('\n', ' ')
871 raise ExtractorError(msg, expected=True)
872 if ('<title
>TTK
:: Доступ к ресурсу ограничен
</title
>' in content
873 and 'blocklist
.rkn
.gov
.ru
' in content):
874 raise ExtractorError(
875 'Access to this webpage has been blocked by decision of the Russian government
. '
876 'Visit http
://blocklist
.rkn
.gov
.ru
/ for a block reason
.',
879 def _request_dump_filename(self, url, video_id):
880 basen = f'{video_id}_{url}
'
881 trim_length = self.get_param('trim_file_name
') or 240
882 if len(basen) > trim_length:
883 h = '___
' + hashlib.md5(basen.encode('utf
-8')).hexdigest()
884 basen = basen[:trim_length - len(h)] + h
885 filename = sanitize_filename(f'{basen}
.dump
', restricted=True)
886 # Working around MAX_PATH limitation on Windows (see
887 # http://msdn.microsoft.com/en-us/library/windows/desktop/aa365247(v=vs.85).aspx)
888 if compat_os_name == 'nt
':
889 absfilepath = os.path.abspath(filename)
890 if len(absfilepath) > 259:
891 filename = fR'\\?\{absfilepath}
'
894 def __decode_webpage(self, webpage_bytes, encoding, headers):
896 encoding = self._guess_encoding_from_content(headers.get('Content
-Type
', ''), webpage_bytes)
898 return webpage_bytes.decode(encoding, 'replace
')
900 return webpage_bytes.decode('utf
-8', 'replace
')
902 def _webpage_read_content(self, urlh, url_or_request, video_id, note=None, errnote=None, fatal=True, prefix=None, encoding=None):
903 webpage_bytes = urlh.read()
904 if prefix is not None:
905 webpage_bytes = prefix + webpage_bytes
906 if self.get_param('dump_intermediate_pages
', False):
907 self.to_screen('Dumping request to
' + urlh.geturl())
908 dump = base64.b64encode(webpage_bytes).decode('ascii
')
909 self._downloader.to_screen(dump)
910 if self.get_param('write_pages
'):
911 filename = self._request_dump_filename(video_id, urlh.geturl())
912 self.to_screen(f'Saving request to {filename}
')
913 with open(filename, 'wb
') as outf:
914 outf.write(webpage_bytes)
916 content = self.__decode_webpage(webpage_bytes, encoding, urlh.headers)
917 self.__check_blocked(content)
921 def _parse_xml(self, xml_string, video_id, transform_source=None, fatal=True):
923 xml_string = transform_source(xml_string)
925 return compat_etree_fromstring(xml_string.encode('utf
-8'))
926 except xml.etree.ElementTree.ParseError as ve:
927 errmsg = '%s: Failed to parse XML
' % video_id
929 raise ExtractorError(errmsg, cause=ve)
931 self.report_warning(errmsg + str(ve))
933 def _parse_json(self, json_string, video_id, transform_source=None, fatal=True, lenient=False):
935 json_string = transform_source(json_string)
938 return json.loads(json_string, strict=False)
939 except json.JSONDecodeError as e:
943 return json.loads(json_string[:e.pos], strict=False)
946 except ValueError as ve:
947 errmsg = f'{video_id}
: Failed to parse JSON
'
949 raise ExtractorError(errmsg, cause=ve)
951 self.report_warning(errmsg + str(ve))
953 def _parse_socket_response_as_json(self, data, video_id, transform_source=None, fatal=True):
954 return self._parse_json(
955 data[data.find('{'):data.rfind('}
') + 1],
956 video_id, transform_source, fatal)
958 def __create_download_methods(name, parser, note, errnote, return_value):
960 def parse(ie, content, *args, **kwargs):
963 # parser is fetched by name so subclasses can override it
964 return getattr(ie, parser)(content, *args, **kwargs)
966 def download_handle(self, url_or_request, video_id, note=note, errnote=errnote, transform_source=None,
967 fatal=True, encoding=None, data=None, headers={}, query={}, expected_status=None):
968 res = self._download_webpage_handle(
969 url_or_request, video_id, note=note, errnote=errnote, fatal=fatal, encoding=encoding,
970 data=data, headers=headers, query=query, expected_status=expected_status)
974 return parse(self, content, video_id, transform_source=transform_source, fatal=fatal), urlh
976 def download_content(self, url_or_request, video_id, note=note, errnote=errnote, transform_source=None,
977 fatal=True, encoding=None, data=None, headers={}, query={}, expected_status=None):
978 if self.get_param('load_pages
'):
979 url_or_request = self._create_request(url_or_request, data, headers, query)
980 filename = self._request_dump_filename(url_or_request.full_url, video_id)
981 self.to_screen(f'Loading request
from {filename}
')
983 with open(filename, 'rb
') as dumpf:
984 webpage_bytes = dumpf.read()
986 self.report_warning(f'Unable to load request
from disk
: {e}
')
988 content = self.__decode_webpage(webpage_bytes, encoding, url_or_request.headers)
989 return parse(self, content, video_id, transform_source, fatal)
993 'transform_source
': transform_source,
995 'encoding
': encoding,
999 'expected_status
': expected_status,
1002 kwargs.pop('transform_source
')
1003 # The method is fetched by name so subclasses can override _download_..._handle
1004 res = getattr(self, download_handle.__name__)(url_or_request, video_id, **kwargs)
1005 return res if res is False else res[0]
1007 def impersonate(func, name, return_value):
1008 func.__name__, func.__qualname__ = name, f'InfoExtractor
.{name}
'
1010 @param transform_source Apply this transformation before parsing
1011 @returns {return_value}
1013 See _download_webpage_handle docstring for other arguments specification
1016 impersonate(download_handle, f'_download_{name}_handle
', f'({return_value}
, URL handle
)')
1017 impersonate(download_content, f'_download_{name}
', f'{return_value}
')
1018 return download_handle, download_content
1020 _download_xml_handle, _download_xml = __create_download_methods(
1021 'xml
', '_parse_xml
', 'Downloading XML
', 'Unable to download XML
', 'xml
as an xml
.etree
.ElementTree
.Element
')
1022 _download_json_handle, _download_json = __create_download_methods(
1023 'json
', '_parse_json
', 'Downloading JSON metadata
', 'Unable to download JSON metadata
', 'JSON
object as a
dict')
1024 _download_socket_json_handle, _download_socket_json = __create_download_methods(
1025 'socket_json
', '_parse_socket_response_as_json
', 'Polling socket
', 'Unable to poll socket
', 'JSON
object as a
dict')
1026 __download_webpage = __create_download_methods('webpage
', None, None, None, 'data of the page
as a string
')[1]
1028 def _download_webpage(
1029 self, url_or_request, video_id, note=None, errnote=None,
1030 fatal=True, tries=1, timeout=NO_DEFAULT, *args, **kwargs):
1032 Return the data of the page as a string.
1035 tries -- number of tries
1036 timeout -- sleep interval between tries
1038 See _download_webpage_handle docstring for other arguments specification.
1041 R''' # NB: These are unused; should they be deprecated?
1043 self._downloader.deprecation_warning('tries argument
is deprecated
in InfoExtractor
._download
_webpage
')
1044 if timeout is NO_DEFAULT:
1047 self._downloader.deprecation_warning('timeout argument
is deprecated
in InfoExtractor
._download
_webpage
')
1053 return self.__download_webpage(url_or_request, video_id, note, errnote, None, fatal, *args, **kwargs)
1054 except compat_http_client.IncompleteRead as e:
1056 if try_count >= tries:
1058 self._sleep(timeout, video_id)
1060 def report_warning(self, msg, video_id=None, *args, only_once=False, **kwargs):
1061 idstr = format_field(video_id, template='%s: ')
1062 msg = f'[{self.IE_NAME}
] {idstr}{msg}
'
1064 if f'WARNING
: {msg}
' in self._printed_messages:
1066 self._printed_messages.add(f'WARNING
: {msg}
')
1067 self._downloader.report_warning(msg, *args, **kwargs)
1069 def to_screen(self, msg, *args, **kwargs):
1070 """Print msg to screen, prefixing it with '[ie_name
]'"""
1071 self._downloader.to_screen(f'[{self.IE_NAME}
] {msg}
', *args, **kwargs)
1073 def write_debug(self, msg, *args, **kwargs):
1074 self._downloader.write_debug(f'[{self.IE_NAME}
] {msg}
', *args, **kwargs)
1076 def get_param(self, name, default=None, *args, **kwargs):
1077 if self._downloader:
1078 return self._downloader.params.get(name, default, *args, **kwargs)
1081 def report_drm(self, video_id, partial=False):
1082 self.raise_no_formats('This video
is DRM protected
', expected=True, video_id=video_id)
1084 def report_extraction(self, id_or_name):
1085 """Report information extraction."""
1086 self.to_screen('%s: Extracting information
' % id_or_name)
1088 def report_download_webpage(self, video_id):
1089 """Report webpage download."""
1090 self.to_screen('%s: Downloading webpage
' % video_id)
1092 def report_age_confirmation(self):
1093 """Report attempt to confirm age."""
1094 self.to_screen('Confirming age
')
1096 def report_login(self):
1097 """Report attempt to log in."""
1098 self.to_screen('Logging
in')
1100 def raise_login_required(
1101 self, msg='This video
is only available
for registered users
',
1102 metadata_available=False, method=NO_DEFAULT):
1103 if metadata_available and (
1104 self.get_param('ignore_no_formats_error
') or self.get_param('wait_for_video
')):
1105 self.report_warning(msg)
1107 msg += format_field(self._login_hint(method), template='. %s')
1108 raise ExtractorError(msg, expected=True)
1110 def raise_geo_restricted(
1111 self, msg='This video
is not available
from your location due to geo restriction
',
1112 countries=None, metadata_available=False):
1113 if metadata_available and (
1114 self.get_param('ignore_no_formats_error
') or self.get_param('wait_for_video
')):
1115 self.report_warning(msg)
1117 raise GeoRestrictedError(msg, countries=countries)
1119 def raise_no_formats(self, msg, expected=False, video_id=None):
1121 self.get_param('ignore_no_formats_error
') or self.get_param('wait_for_video
')):
1122 self.report_warning(msg, video_id)
1123 elif isinstance(msg, ExtractorError):
1126 raise ExtractorError(msg, expected=expected, video_id=video_id)
1128 # Methods for following #608
1130 def url_result(url, ie=None, video_id=None, video_title=None, *, url_transparent=False, **kwargs):
1131 """Returns a URL that points to a page that should be processed"""
1133 kwargs['ie_key
'] = ie if isinstance(ie, str) else ie.ie_key()
1134 if video_id is not None:
1135 kwargs['id'] = video_id
1136 if video_title is not None:
1137 kwargs['title
'] = video_title
1140 '_type
': 'url_transparent
' if url_transparent else 'url
',
1144 def playlist_from_matches(self, matches, playlist_id=None, playlist_title=None, getter=None, ie=None, video_kwargs=None, **kwargs):
1145 urls = (self.url_result(self._proto_relative_url(m), ie, **(video_kwargs or {}))
1146 for m in orderedSet(map(getter, matches) if getter else matches))
1147 return self.playlist_result(urls, playlist_id, playlist_title, **kwargs)
1150 def playlist_result(entries, playlist_id=None, playlist_title=None, playlist_description=None, *, multi_video=False, **kwargs):
1151 """Returns a playlist"""
1153 kwargs['id'] = playlist_id
1155 kwargs['title
'] = playlist_title
1156 if playlist_description is not None:
1157 kwargs['description
'] = playlist_description
1160 '_type
': 'multi_video
' if multi_video else 'playlist
',
1164 def _search_regex(self, pattern, string, name, default=NO_DEFAULT, fatal=True, flags=0, group=None):
1166 Perform a regex search on the given string, using a single or a list of
1167 patterns returning the first matching group.
1168 In case of failure return a default value or raise a WARNING or a
1169 RegexNotFoundError, depending on fatal, specifying the field name.
1173 elif isinstance(pattern, (str, re.Pattern)):
1174 mobj = re.search(pattern, string, flags)
1177 mobj = re.search(p, string, flags)
1181 _name = self._downloader._format_err(name, self._downloader.Styles.EMPHASIS)
1185 # return the first matching group
1186 return next(g for g in mobj.groups() if g is not None)
1187 elif isinstance(group, (list, tuple)):
1188 return tuple(mobj.group(g) for g in group)
1190 return mobj.group(group)
1191 elif default is not NO_DEFAULT:
1194 raise RegexNotFoundError('Unable to extract
%s' % _name)
1196 self.report_warning('unable to extract
%s' % _name + bug_reports_message())
1199 def _html_search_regex(self, pattern, string, name, default=NO_DEFAULT, fatal=True, flags=0, group=None):
1201 Like _search_regex, but strips HTML tags and unescapes entities.
1203 res = self._search_regex(pattern, string, name, default, fatal, flags, group)
1205 return clean_html(res).strip()
1209 def _get_netrc_login_info(self, netrc_machine=None):
1212 netrc_machine = netrc_machine or self._NETRC_MACHINE
1214 if self.get_param('usenetrc
', False):
1216 netrc_file = compat_expanduser(self.get_param('netrc_location
') or '~
')
1217 if os.path.isdir(netrc_file):
1218 netrc_file = os.path.join(netrc_file, '.netrc
')
1219 info = netrc.netrc(file=netrc_file).authenticators(netrc_machine)
1220 if info is not None:
1224 raise netrc.NetrcParseError(
1225 'No authenticators
for %s' % netrc_machine)
1226 except (OSError, netrc.NetrcParseError) as err:
1227 self.report_warning(
1228 'parsing
.netrc
: %s' % error_to_compat_str(err))
1230 return username, password
1232 def _get_login_info(self, username_option='username
', password_option='password
', netrc_machine=None):
1234 Get the login info as (username, password)
1235 First look for the manually specified credentials using username_option
1236 and password_option as keys in params dictionary. If no such credentials
1237 available look in the netrc file using the netrc_machine or _NETRC_MACHINE
1239 If there's no info available
, return (None, None)
1242 # Attempt to use provided username and password or .netrc data
1243 username = self.get_param(username_option)
1244 if username is not None:
1245 password = self.get_param(password_option)
1247 username, password = self._get_netrc_login_info(netrc_machine)
1249 return username, password
1251 def _get_tfa_info(self, note='two-factor verification code'):
1253 Get the two
-factor authentication info
1254 TODO
- asking the user will be required
for sms
/phone verify
1255 currently just uses the command line option
1256 If there
's no info available, return None
1259 tfa = self.get_param('twofactor
')
1263 return compat_getpass('Type
%s and press
[Return
]: ' % note)
1265 # Helper functions for extracting OpenGraph info
1267 def _og_regexes(prop):
1268 content_re = r'content
=(?
:"([^"]+?
)"|\'([^\']+?)\'|\s*([^\s"\'=<>`
]+?
))'
1269 property_re = (r'(?
:name|
property)=(?
:\'og
%(sep)s%(prop)s\'|
"og%(sep)s%(prop)s"|\s
*og
%(sep)s%(prop)s\b)'
1270 % {'prop': re.escape(prop), 'sep': '(?::|[:-])'})
1271 template = r'<meta
[^
>]+?
%s[^
>]+?
%s'
1273 template % (property_re, content_re),
1274 template % (content_re, property_re),
1278 def _meta_regex(prop):
1279 return r'''(?isx)<meta
1280 (?=[^>]+(?:itemprop|name|property|id|http-equiv)=(["\']?)%s\1)
1281 [^>]+?content=(["\'])(?P<content>.*?)\2''' % re.escape(prop)
1283 def _og_search_property(self, prop, html, name=None, **kargs):
1284 prop = variadic(prop)
1286 name = 'OpenGraph
%s' % prop[0]
1289 og_regexes.extend(self._og_regexes(p))
1290 escaped = self._search_regex(og_regexes, html, name, flags=re.DOTALL, **kargs)
1293 return unescapeHTML(escaped)
1295 def _og_search_thumbnail(self, html, **kargs):
1296 return self._og_search_property('image
', html, 'thumbnail URL
', fatal=False, **kargs)
1298 def _og_search_description(self, html, **kargs):
1299 return self._og_search_property('description
', html, fatal=False, **kargs)
1301 def _og_search_title(self, html, *, fatal=False, **kargs):
1302 return self._og_search_property('title
', html, fatal=fatal, **kargs)
1304 def _og_search_video_url(self, html, name='video url
', secure=True, **kargs):
1305 regexes = self._og_regexes('video
') + self._og_regexes('video
:url
')
1307 regexes = self._og_regexes('video
:secure_url
') + regexes
1308 return self._html_search_regex(regexes, html, name, **kargs)
1310 def _og_search_url(self, html, **kargs):
1311 return self._og_search_property('url
', html, **kargs)
1313 def _html_extract_title(self, html, name='title
', *, fatal=False, **kwargs):
1314 return self._html_search_regex(r'(?s
)<title
\b[^
>]*>([^
<]+)</title
>', html, name, fatal=fatal, **kwargs)
1316 def _html_search_meta(self, name, html, display_name=None, fatal=False, **kwargs):
1317 name = variadic(name)
1318 if display_name is None:
1319 display_name = name[0]
1320 return self._html_search_regex(
1321 [self._meta_regex(n) for n in name],
1322 html, display_name, fatal=fatal, group='content
', **kwargs)
1324 def _dc_search_uploader(self, html):
1325 return self._html_search_meta('dc
.creator
', html, 'uploader
')
1327 def _rta_search(self, html):
1328 # See http://www.rtalabel.org/index.php?content=howtofaq#single
1329 if re.search(r'(?ix
)<meta\s
+name
="rating"\s
+'
1330 r' content
="RTA-5042-1996-1400-1577-RTA"',
1335 def _media_rating_search(self, html):
1336 # See http://www.tjg-designs.com/WP/metadata-code-examples-adding-metadata-to-your-web-pages/
1337 rating = self._html_search_meta('rating
', html)
1349 return RATING_TABLE.get(rating.lower())
1351 def _family_friendly_search(self, html):
1352 # See http://schema.org/VideoObject
1353 family_friendly = self._html_search_meta(
1354 'isFamilyFriendly
', html, default=None)
1356 if not family_friendly:
1365 return RATING_TABLE.get(family_friendly.lower())
1367 def _twitter_search_player(self, html):
1368 return self._html_search_meta('twitter
:player
', html,
1369 'twitter card player
')
1371 def _search_json_ld(self, html, video_id, expected_type=None, **kwargs):
1372 json_ld_list = list(re.finditer(JSON_LD_RE, html))
1373 default = kwargs.get('default
', NO_DEFAULT)
1374 # JSON-LD may be malformed and thus `fatal` should be respected.
1375 # At the same time `default` may be passed that assumes `fatal=False`
1376 # for _search_regex. Let's simulate the same behavior here
as well
.
1377 fatal
= kwargs
.get('fatal', True) if default
is NO_DEFAULT
else False
1379 for mobj
in json_ld_list
:
1380 json_ld_item
= self
._parse
_json
(
1381 mobj
.group('json_ld'), video_id
, fatal
=fatal
)
1382 if not json_ld_item
:
1384 if isinstance(json_ld_item
, dict):
1385 json_ld
.append(json_ld_item
)
1386 elif isinstance(json_ld_item
, (list, tuple)):
1387 json_ld
.extend(json_ld_item
)
1389 json_ld
= self
._json
_ld
(json_ld
, video_id
, fatal
=fatal
, expected_type
=expected_type
)
1392 if default
is not NO_DEFAULT
:
1395 raise RegexNotFoundError('Unable to extract JSON-LD')
1397 self
.report_warning('unable to extract JSON-LD %s' % bug_reports_message())
1400 def _json_ld(self
, json_ld
, video_id
, fatal
=True, expected_type
=None):
1401 if isinstance(json_ld
, compat_str
):
1402 json_ld
= self
._parse
_json
(json_ld
, video_id
, fatal
=fatal
)
1406 if not isinstance(json_ld
, (list, tuple, dict)):
1408 if isinstance(json_ld
, dict):
1411 INTERACTION_TYPE_MAP
= {
1412 'CommentAction': 'comment',
1413 'AgreeAction': 'like',
1414 'DisagreeAction': 'dislike',
1415 'LikeAction': 'like',
1416 'DislikeAction': 'dislike',
1417 'ListenAction': 'view',
1418 'WatchAction': 'view',
1419 'ViewAction': 'view',
1422 def extract_interaction_type(e
):
1423 interaction_type
= e
.get('interactionType')
1424 if isinstance(interaction_type
, dict):
1425 interaction_type
= interaction_type
.get('@type')
1426 return str_or_none(interaction_type
)
1428 def extract_interaction_statistic(e
):
1429 interaction_statistic
= e
.get('interactionStatistic')
1430 if isinstance(interaction_statistic
, dict):
1431 interaction_statistic
= [interaction_statistic
]
1432 if not isinstance(interaction_statistic
, list):
1434 for is_e
in interaction_statistic
:
1435 if not isinstance(is_e
, dict):
1437 if is_e
.get('@type') != 'InteractionCounter':
1439 interaction_type
= extract_interaction_type(is_e
)
1440 if not interaction_type
:
1442 # For interaction count some sites provide string instead of
1443 # an integer (as per spec) with non digit characters (e.g. ",")
1444 # so extracting count with more relaxed str_to_int
1445 interaction_count
= str_to_int(is_e
.get('userInteractionCount'))
1446 if interaction_count
is None:
1448 count_kind
= INTERACTION_TYPE_MAP
.get(interaction_type
.split('/')[-1])
1451 count_key
= '%s_count' % count_kind
1452 if info
.get(count_key
) is not None:
1454 info
[count_key
] = interaction_count
1456 def extract_chapter_information(e
):
1458 'title': part
.get('name'),
1459 'start_time': part
.get('startOffset'),
1460 'end_time': part
.get('endOffset'),
1461 } for part
in variadic(e
.get('hasPart') or []) if part
.get('@type') == 'Clip']
1462 for idx
, (last_c
, current_c
, next_c
) in enumerate(zip(
1463 [{'end_time': 0}
] + chapters
, chapters
, chapters
[1:])):
1464 current_c
['end_time'] = current_c
['end_time'] or next_c
['start_time']
1465 current_c
['start_time'] = current_c
['start_time'] or last_c
['end_time']
1466 if None in current_c
.values():
1467 self
.report_warning(f
'Chapter {idx} contains broken data. Not extracting chapters')
1470 chapters
[-1]['end_time'] = chapters
[-1]['end_time'] or info
['duration']
1471 info
['chapters'] = chapters
1473 def extract_video_object(e
):
1474 assert e
['@type'] == 'VideoObject'
1475 author
= e
.get('author')
1477 'url': traverse_obj(e
, 'contentUrl', 'embedUrl', expected_type
=url_or_none
),
1478 'title': unescapeHTML(e
.get('name')),
1479 'description': unescapeHTML(e
.get('description')),
1480 'thumbnails': [{'url': url}
1481 for url
in variadic(traverse_obj(e
, 'thumbnailUrl', 'thumbnailURL'))
1482 if url_or_none(url
)],
1483 'duration': parse_duration(e
.get('duration')),
1484 'timestamp': unified_timestamp(e
.get('uploadDate')),
1485 # author can be an instance of 'Organization' or 'Person' types.
1486 # both types can have 'name' property(inherited from 'Thing' type). [1]
1487 # however some websites are using 'Text' type instead.
1488 # 1. https://schema.org/VideoObject
1489 'uploader': author
.get('name') if isinstance(author
, dict) else author
if isinstance(author
, compat_str
) else None,
1490 'filesize': float_or_none(e
.get('contentSize')),
1491 'tbr': int_or_none(e
.get('bitrate')),
1492 'width': int_or_none(e
.get('width')),
1493 'height': int_or_none(e
.get('height')),
1494 'view_count': int_or_none(e
.get('interactionCount')),
1496 extract_interaction_statistic(e
)
1497 extract_chapter_information(e
)
1499 def traverse_json_ld(json_ld
, at_top_level
=True):
1501 if at_top_level
and '@context' not in e
:
1503 if at_top_level
and set(e
.keys()) == {'@context', '@graph'}
:
1504 traverse_json_ld(variadic(e
['@graph'], allowed_types
=(dict,)), at_top_level
=False)
1506 item_type
= e
.get('@type')
1507 if expected_type
is not None and expected_type
!= item_type
:
1509 rating
= traverse_obj(e
, ('aggregateRating', 'ratingValue'), expected_type
=float_or_none
)
1510 if rating
is not None:
1511 info
['average_rating'] = rating
1512 if item_type
in ('TVEpisode', 'Episode'):
1513 episode_name
= unescapeHTML(e
.get('name'))
1515 'episode': episode_name
,
1516 'episode_number': int_or_none(e
.get('episodeNumber')),
1517 'description': unescapeHTML(e
.get('description')),
1519 if not info
.get('title') and episode_name
:
1520 info
['title'] = episode_name
1521 part_of_season
= e
.get('partOfSeason')
1522 if isinstance(part_of_season
, dict) and part_of_season
.get('@type') in ('TVSeason', 'Season', 'CreativeWorkSeason'):
1524 'season': unescapeHTML(part_of_season
.get('name')),
1525 'season_number': int_or_none(part_of_season
.get('seasonNumber')),
1527 part_of_series
= e
.get('partOfSeries') or e
.get('partOfTVSeries')
1528 if isinstance(part_of_series
, dict) and part_of_series
.get('@type') in ('TVSeries', 'Series', 'CreativeWorkSeries'):
1529 info
['series'] = unescapeHTML(part_of_series
.get('name'))
1530 elif item_type
== 'Movie':
1532 'title': unescapeHTML(e
.get('name')),
1533 'description': unescapeHTML(e
.get('description')),
1534 'duration': parse_duration(e
.get('duration')),
1535 'timestamp': unified_timestamp(e
.get('dateCreated')),
1537 elif item_type
in ('Article', 'NewsArticle'):
1539 'timestamp': parse_iso8601(e
.get('datePublished')),
1540 'title': unescapeHTML(e
.get('headline')),
1541 'description': unescapeHTML(e
.get('articleBody') or e
.get('description')),
1543 if traverse_obj(e
, ('video', 0, '@type')) == 'VideoObject':
1544 extract_video_object(e
['video'][0])
1545 elif traverse_obj(e
, ('subjectOf', 0, '@type')) == 'VideoObject':
1546 extract_video_object(e
['subjectOf'][0])
1547 elif item_type
== 'VideoObject':
1548 extract_video_object(e
)
1549 if expected_type
is None:
1553 video
= e
.get('video')
1554 if isinstance(video
, dict) and video
.get('@type') == 'VideoObject':
1555 extract_video_object(video
)
1556 if expected_type
is None:
1560 traverse_json_ld(json_ld
)
1562 return filter_dict(info
)
1564 def _search_nextjs_data(self
, webpage
, video_id
, *, transform_source
=None, fatal
=True, **kw
):
1565 return self
._parse
_json
(
1567 r
'(?s)<script[^>]+id=[\'"]__NEXT_DATA__[\'"][^
>]*>([^
<]+)</script
>',
1568 webpage, 'next
.js data
', fatal=fatal, **kw),
1569 video_id, transform_source=transform_source, fatal=fatal)
1571 def _search_nuxt_data(self, webpage, video_id, context_name='__NUXT__
'):
1572 ''' Parses Nuxt.js metadata. This works as long as the function __NUXT__ invokes is a pure function. '''
1573 # not all website do this, but it can be changed
1574 # https://stackoverflow.com/questions/67463109/how-to-change-or-hide-nuxt-and-nuxt-keyword-in-page-source
1575 rectx = re.escape(context_name)
1576 js, arg_keys, arg_vals = self._search_regex(
1577 (r'<script
>window\
.%s=\
(function\
((?P
<arg_keys
>.*?
)\
)\{return\s(?P<js>\{.*?\}
)\
}\
((?P
<arg_vals
>.+?
)\
)\
);?
</script
>' % rectx,
1578 r'%s\
(.*?\
(function\
((?P
<arg_keys
>.*?
)\
)\{return\s(?P<js>\{.*?\}
)\
}\
((?P
<arg_vals
>.*?
)\
)' % rectx),
1579 webpage, context_name, group=['js
', 'arg_keys
', 'arg_vals
'])
1581 args = dict(zip(arg_keys.split(','), arg_vals.split(',')))
1583 for key, val in args.items():
1584 if val in ('undefined
', 'void
0'):
1587 return self._parse_json(js_to_json(js, args), video_id)['data
'][0]
1590 def _hidden_inputs(html):
1591 html = re.sub(r'<!--(?
:(?
!<!--).)*-->', '', html)
1593 for input in re.findall(r'(?i
)(<input[^
>]+>)', html):
1594 attrs = extract_attributes(input)
1597 if attrs.get('type') not in ('hidden
', 'submit
'):
1599 name = attrs.get('name
') or attrs.get('id')
1600 value = attrs.get('value
')
1601 if name and value is not None:
1602 hidden_inputs[name] = value
1603 return hidden_inputs
1605 def _form_hidden_inputs(self, form_id, html):
1606 form = self._search_regex(
1607 r'(?
is)<form
[^
>]+?
id=(["\'])%s\1[^>]*>(?P<form>.+?)</form>' % form_id,
1608 html, '%s form' % form_id, group='form')
1609 return self._hidden_inputs(form)
1612 regex = r' *((?P<reverse>\+)?(?P<field>[a-zA-Z0-9_]+)((?P<separator>[~:])(?P<limit>.*?))?)? *$'
1614 default = ('hidden', 'aud_or_vid', 'hasvid', 'ie_pref', 'lang', 'quality',
1615 'res', 'fps', 'hdr:12', 'codec:vp9.2', 'size', 'br', 'asr',
1616 'proto', 'ext', 'hasaud', 'source', 'id') # These must not be aliases
1617 ytdl_default = ('hasaud', 'lang', 'quality', 'tbr', 'filesize', 'vbr',
1618 'height', 'width', 'proto', 'vext', 'abr', 'aext',
1619 'fps', 'fs_approx', 'source', 'id')
1622 'vcodec': {'type': 'ordered', 'regex': True,
1623 'order': ['av0?1', 'vp0?9.2', 'vp0?9', '[hx]265|he?vc?', '[hx]264|avc', 'vp0?8', 'mp4v|h263', 'theora', '', None, 'none']},
1624 'acodec': {'type': 'ordered', 'regex': True,
1625 'order': ['[af]lac', 'wav|aiff', 'opus', 'vorbis|ogg', 'aac', 'mp?4a?', 'mp3', 'e-?a?c-?3', 'ac-?3', 'dts', '', None, 'none']},
1626 'hdr': {'type': 'ordered', 'regex': True, 'field': 'dynamic_range',
1627 'order': ['dv', '(hdr)?12', r'(hdr)?10\+', '(hdr)?10', 'hlg', '', 'sdr', None]},
1628 'proto': {'type': 'ordered', 'regex': True, 'field': 'protocol',
1629 'order': ['(ht|f)tps', '(ht|f)tp$', 'm3u8.*', '.*dash', 'websocket_frag', 'rtmpe?', '', 'mms|rtsp', 'ws|websocket', 'f4']},
1630 'vext': {'type': 'ordered', 'field': 'video_ext',
1631 'order': ('mp4', 'webm', 'flv', '', 'none'),
1632 'order_free': ('webm', 'mp4', 'flv', '', 'none')},
1633 'aext': {'type': 'ordered', 'field': 'audio_ext',
1634 'order': ('m4a', 'aac', 'mp3', 'ogg', 'opus', 'webm', '', 'none'),
1635 'order_free': ('opus', 'ogg', 'webm', 'm4a', 'mp3', 'aac', '', 'none')},
1636 'hidden': {'visible': False, 'forced': True, 'type': 'extractor', 'max': -1000},
1637 'aud_or_vid': {'visible': False, 'forced': True, 'type': 'multiple',
1638 'field': ('vcodec', 'acodec'),
1639 'function': lambda it: int(any(v != 'none' for v in it))},
1640 'ie_pref': {'priority': True, 'type': 'extractor'},
1641 'hasvid': {'priority': True, 'field': 'vcodec', 'type': 'boolean', 'not_in_list': ('none',)},
1642 'hasaud': {'field': 'acodec', 'type': 'boolean', 'not_in_list': ('none',)},
1643 'lang': {'convert': 'float', 'field': 'language_preference', 'default': -1},
1644 'quality': {'convert': 'float', 'default': -1},
1645 'filesize': {'convert': 'bytes'},
1646 'fs_approx': {'convert': 'bytes', 'field': 'filesize_approx'},
1647 'id': {'convert': 'string', 'field': 'format_id'},
1648 'height': {'convert': 'float_none'},
1649 'width': {'convert': 'float_none'},
1650 'fps': {'convert': 'float_none'},
1651 'tbr': {'convert': 'float_none'},
1652 'vbr': {'convert': 'float_none'},
1653 'abr': {'convert': 'float_none'},
1654 'asr': {'convert': 'float_none'},
1655 'source': {'convert': 'float', 'field': 'source_preference', 'default': -1},
1657 'codec': {'type': 'combined', 'field': ('vcodec', 'acodec')},
1658 'br': {'type': 'combined', 'field': ('tbr', 'vbr', 'abr'), 'same_limit': True},
1659 'size': {'type': 'combined', 'same_limit': True, 'field': ('filesize', 'fs_approx')},
1660 'ext': {'type': 'combined', 'field': ('vext', 'aext')},
1661 'res': {'type': 'multiple', 'field': ('height', 'width'),
1662 'function': lambda it: (lambda l: min(l) if l else 0)(tuple(filter(None, it)))},
1664 # For compatibility with youtube-dl
1665 'format_id': {'type': 'alias', 'field': 'id'},
1666 'preference': {'type': 'alias', 'field': 'ie_pref'},
1667 'language_preference': {'type': 'alias', 'field': 'lang'},
1668 'source_preference': {'type': 'alias', 'field': 'source'},
1669 'protocol': {'type': 'alias', 'field': 'proto'},
1670 'filesize_approx': {'type': 'alias', 'field': 'fs_approx'},
1673 'dimension': {'type': 'alias', 'field': 'res', 'deprecated': True},
1674 'resolution': {'type': 'alias', 'field': 'res', 'deprecated': True},
1675 'extension': {'type': 'alias', 'field': 'ext', 'deprecated': True},
1676 'bitrate': {'type': 'alias', 'field': 'br', 'deprecated': True},
1677 'total_bitrate': {'type': 'alias', 'field': 'tbr', 'deprecated': True},
1678 'video_bitrate': {'type': 'alias', 'field': 'vbr', 'deprecated': True},
1679 'audio_bitrate': {'type': 'alias', 'field': 'abr', 'deprecated': True},
1680 'framerate': {'type': 'alias', 'field': 'fps', 'deprecated': True},
1681 'filesize_estimate': {'type': 'alias', 'field': 'size', 'deprecated': True},
1682 'samplerate': {'type': 'alias', 'field': 'asr', 'deprecated': True},
1683 'video_ext': {'type': 'alias', 'field': 'vext', 'deprecated': True},
1684 'audio_ext': {'type': 'alias', 'field': 'aext', 'deprecated': True},
1685 'video_codec': {'type': 'alias', 'field': 'vcodec', 'deprecated': True},
1686 'audio_codec': {'type': 'alias', 'field': 'acodec', 'deprecated': True},
1687 'video': {'type': 'alias', 'field': 'hasvid', 'deprecated': True},
1688 'has_video': {'type': 'alias', 'field': 'hasvid', 'deprecated': True},
1689 'audio': {'type': 'alias', 'field': 'hasaud', 'deprecated': True},
1690 'has_audio': {'type': 'alias', 'field': 'hasaud', 'deprecated': True},
1691 'extractor': {'type': 'alias', 'field': 'ie_pref', 'deprecated': True},
1692 'extractor_preference': {'type': 'alias', 'field': 'ie_pref', 'deprecated': True},
1695 def __init__(self, ie, field_preference):
1697 self.ydl = ie._downloader
1698 self.evaluate_params(self.ydl.params, field_preference)
1699 if ie.get_param('verbose'):
1700 self.print_verbose_info(self.ydl.write_debug)
1702 def _get_field_setting(self, field, key):
1703 if field not in self.settings:
1704 if key in ('forced', 'priority'):
1706 self.ydl.deprecation_warning(
1707 f'Using arbitrary fields ({field}) for format sorting is deprecated '
1708 'and may be removed in a future version')
1709 self.settings[field] = {}
1710 propObj = self.settings[field]
1711 if key not in propObj:
1712 type = propObj.get('type')
1714 default = 'preference' if type == 'extractor' else (field,) if type in ('combined', 'multiple') else field
1715 elif key == 'convert':
1716 default = 'order' if type == 'ordered' else 'float_string' if field else 'ignore'
1718 default = {'type': 'field', 'visible': True, 'order': [], 'not_in_list': (None,)}.get(key, None)
1719 propObj[key] = default
1722 def _resolve_field_value(self, field, value, convertNone=False):
1727 value = value.lower()
1728 conversion = self._get_field_setting(field, 'convert')
1729 if conversion == 'ignore':
1731 if conversion == 'string':
1733 elif conversion == 'float_none':
1734 return float_or_none(value)
1735 elif conversion == 'bytes':
1736 return FileDownloader.parse_bytes(value)
1737 elif conversion == 'order':
1738 order_list = (self._use_free_order and self._get_field_setting(field, 'order_free')) or self._get_field_setting(field, 'order')
1739 use_regex = self._get_field_setting(field, 'regex')
1740 list_length = len(order_list)
1741 empty_pos = order_list.index('') if '' in order_list else list_length + 1
1742 if use_regex and value is not None:
1743 for i, regex in enumerate(order_list):
1744 if regex and re.match(regex, value):
1745 return list_length - i
1746 return list_length - empty_pos # not in list
1747 else: # not regex or value = None
1748 return list_length - (order_list.index(value) if value in order_list else empty_pos)
1750 if value.isnumeric():
1753 self.settings[field]['convert'] = 'string'
1756 def evaluate_params(self, params, sort_extractor):
1757 self._use_free_order = params.get('prefer_free_formats', False)
1758 self._sort_user = params.get('format_sort', [])
1759 self._sort_extractor = sort_extractor
1761 def add_item(field, reverse, closest, limit_text):
1762 field = field.lower()
1763 if field in self._order:
1765 self._order.append(field)
1766 limit = self._resolve_field_value(field, limit_text)
1769 'closest': False if limit is None else closest,
1770 'limit_text': limit_text,
1772 if field in self.settings:
1773 self.settings[field].update(data)
1775 self.settings[field] = data
1778 tuple(field for field in self.default if self._get_field_setting(field, 'forced'))
1779 + (tuple() if params.get('format_sort_force', False)
1780 else tuple(field for field in self.default if self._get_field_setting(field, 'priority')))
1781 + tuple(self._sort_user) + tuple(sort_extractor) + self.default)
1783 for item in sort_list:
1784 match = re.match(self.regex, item)
1786 raise ExtractorError('Invalid format sort string "%s" given by extractor' % item)
1787 field = match.group('field')
1790 if self._get_field_setting(field, 'type') == 'alias':
1791 alias, field = field, self._get_field_setting(field, 'field')
1792 if self._get_field_setting(alias, 'deprecated'):
1793 self.ydl.deprecation_warning(
1794 f'Format sorting alias {alias} is deprecated '
1795 f'and may be removed in a future version. Please use {field} instead')
1796 reverse = match.group('reverse') is not None
1797 closest = match.group('separator') == '~'
1798 limit_text = match.group('limit')
1800 has_limit = limit_text is not None
1801 has_multiple_fields = self._get_field_setting(field, 'type') == 'combined'
1802 has_multiple_limits = has_limit and has_multiple_fields and not self._get_field_setting(field, 'same_limit')
1804 fields = self._get_field_setting(field, 'field') if has_multiple_fields else (field,)
1805 limits = limit_text.split(':') if has_multiple_limits else (limit_text,) if has_limit else tuple()
1806 limit_count = len(limits)
1807 for (i, f) in enumerate(fields):
1808 add_item(f, reverse, closest,
1809 limits[i] if i < limit_count
1810 else limits[0] if has_limit and not has_multiple_limits
1813 def print_verbose_info(self, write_debug):
1815 write_debug('Sort order given by user: %s' % ', '.join(self._sort_user))
1816 if self._sort_extractor:
1817 write_debug('Sort order given by extractor: %s' % ', '.join(self._sort_extractor))
1818 write_debug('Formats sorted by: %s' % ', '.join(['%s%s%s' % (
1819 '+' if self._get_field_setting(field, 'reverse') else '', field,
1820 '%s%s(%s)' % ('~' if self._get_field_setting(field, 'closest') else ':',
1821 self._get_field_setting(field, 'limit_text'),
1822 self._get_field_setting(field, 'limit'))
1823 if self._get_field_setting(field, 'limit_text') is not None else '')
1824 for field in self._order if self._get_field_setting(field, 'visible')]))
1826 def _calculate_field_preference_from_value(self, format, field, type, value):
1827 reverse = self._get_field_setting(field, 'reverse')
1828 closest = self._get_field_setting(field, 'closest')
1829 limit = self._get_field_setting(field, 'limit')
1831 if type == 'extractor':
1832 maximum = self._get_field_setting(field, 'max')
1833 if value is None or (maximum is not None and value >= maximum):
1835 elif type == 'boolean':
1836 in_list = self._get_field_setting(field, 'in_list')
1837 not_in_list = self._get_field_setting(field, 'not_in_list')
1838 value = 0 if ((in_list is None or value in in_list) and (not_in_list is None or value not in not_in_list)) else -1
1839 elif type == 'ordered':
1840 value = self._resolve_field_value(field, value, True)
1842 # try to convert to number
1843 val_num = float_or_none(value, default=self._get_field_setting(field, 'default'))
1844 is_num = self._get_field_setting(field, 'convert') != 'string' and val_num is not None
1848 return ((-10, 0) if value is None
1849 else (1, value, 0) if not is_num # if a field has mixed strings and numbers, strings are sorted higher
1850 else (0, -abs(value - limit), value - limit if reverse else limit - value) if closest
1851 else (0, value, 0) if not reverse and (limit is None or value <= limit)
1852 else (0, -value, 0) if limit is None or (reverse and value == limit) or value > limit
1853 else (-1, value, 0))
1855 def _calculate_field_preference(self, format, field):
1856 type = self._get_field_setting(field, 'type') # extractor, boolean, ordered, field, multiple
1857 get_value = lambda f: format.get(self._get_field_setting(f, 'field'))
1858 if type == 'multiple':
1859 type = 'field' # Only 'field' is allowed in multiple for now
1860 actual_fields = self._get_field_setting(field, 'field')
1862 value = self._get_field_setting(field, 'function')(get_value(f) for f in actual_fields)
1864 value = get_value(field)
1865 return self._calculate_field_preference_from_value(format, field, type, value)
1867 def calculate_preference(self, format):
1868 # Determine missing protocol
1869 if not format.get('protocol'):
1870 format['protocol'] = determine_protocol(format)
1872 # Determine missing ext
1873 if not format.get('ext') and 'url' in format:
1874 format['ext'] = determine_ext(format['url'])
1875 if format.get('vcodec') == 'none':
1876 format['audio_ext'] = format['ext'] if format.get('acodec') != 'none' else 'none'
1877 format['video_ext'] = 'none'
1879 format['video_ext'] = format['ext']
1880 format['audio_ext'] = 'none'
1881 # if format.get('preference') is None and format.get('ext') in ('f4f', 'f4m'): # Not supported?
1882 # format['preference'] = -1000
1884 # Determine missing bitrates
1885 if format.get('tbr') is None:
1886 if format.get('vbr') is not None and format.get('abr') is not None:
1887 format['tbr'] = format.get('vbr', 0) + format.get('abr', 0)
1889 if format.get('vcodec') != 'none' and format.get('vbr') is None:
1890 format['vbr'] = format.get('tbr') - format.get('abr', 0)
1891 if format.get('acodec') != 'none' and format.get('abr') is None:
1892 format['abr'] = format.get('tbr') - format.get('vbr', 0)
1894 return tuple(self._calculate_field_preference(format, field) for field in self._order)
1896 def _sort_formats(self, formats, field_preference=[]):
1899 formats.sort(key=self.FormatSort(self, field_preference).calculate_preference)
1901 def _check_formats(self, formats, video_id):
1903 formats[:] = filter(
1904 lambda f: self._is_valid_url(
1906 item='%s video format' % f.get('format_id') if f.get('format_id') else 'video'),
1910 def _remove_duplicate_formats(formats):
1914 if f['url'] not in format_urls:
1915 format_urls.add(f['url'])
1916 unique_formats.append(f)
1917 formats[:] = unique_formats
1919 def _is_valid_url(self, url, video_id, item='video', headers={}):
1920 url = self._proto_relative_url(url, scheme='http:')
1921 # For now assume non HTTP(S) URLs always valid
1922 if not (url.startswith('http://') or url.startswith('https://')):
1925 self._request_webpage(url, video_id, 'Checking %s URL' % item, headers=headers)
1927 except ExtractorError as e:
1929 '%s: %s URL is invalid, skipping: %s'
1930 % (video_id, item, error_to_compat_str(e.cause)))
1933 def http_scheme(self):
1934 """ Either "http
:" or "https
:", depending on the user's preferences """
1937 if self.get_param('prefer_insecure', False)
1940 def _proto_relative_url(self, url, scheme=None):
1943 if url.startswith('//'):
1945 scheme = self.http_scheme()
1950 def _sleep(self, timeout, video_id, msg_template=None):
1951 if msg_template is None:
1952 msg_template = '%(video_id)s: Waiting for %(timeout)s seconds'
1953 msg = msg_template % {'video_id': video_id, 'timeout': timeout}
1957 def _extract_f4m_formats(self, manifest_url, video_id, preference=None, quality=None, f4m_id=None,
1958 transform_source=lambda s: fix_xml_ampersands(s).strip(),
1959 fatal=True, m3u8_id=None, data=None, headers={}, query={}):
1960 res = self._download_xml_handle(
1961 manifest_url, video_id, 'Downloading f4m manifest',
1962 'Unable to download f4m manifest',
1963 # Some manifests may be malformed, e.g. prosiebensat1 generated manifests
1964 # (see https://github.com/ytdl-org/youtube-dl/issues/6215#issuecomment-121704244)
1965 transform_source=transform_source,
1966 fatal=fatal, data=data, headers=headers, query=query)
1970 manifest, urlh = res
1971 manifest_url = urlh.geturl()
1973 return self._parse_f4m_formats(
1974 manifest, manifest_url, video_id, preference=preference, quality=quality, f4m_id=f4m_id,
1975 transform_source=transform_source, fatal=fatal, m3u8_id=m3u8_id)
1977 def _parse_f4m_formats(self, manifest, manifest_url, video_id, preference=None, quality=None, f4m_id=None,
1978 transform_source=lambda s: fix_xml_ampersands(s).strip(),
1979 fatal=True, m3u8_id=None):
1980 if not isinstance(manifest, xml.etree.ElementTree.Element) and not fatal:
1983 # currently yt-dlp cannot decode the playerVerificationChallenge as Akamai uses Adobe Alchemy
1984 akamai_pv = manifest.find('{http://ns.adobe.com/f4m/1.0}pv-2.0')
1985 if akamai_pv is not None and ';' in akamai_pv.text:
1986 playerVerificationChallenge = akamai_pv.text.split(';')[0]
1987 if playerVerificationChallenge.strip() != '':
1991 manifest_version = '1.0'
1992 media_nodes = manifest.findall('{http://ns.adobe.com/f4m/1.0}media')
1994 manifest_version = '2.0'
1995 media_nodes = manifest.findall('{http://ns.adobe.com/f4m/2.0}media')
1996 # Remove unsupported DRM protected media from final formats
1997 # rendition (see https://github.com/ytdl-org/youtube-dl/issues/8573).
1998 media_nodes = remove_encrypted_media(media_nodes)
2002 manifest_base_url = get_base_url(manifest)
2004 bootstrap_info = xpath_element(
2005 manifest, ['{http://ns.adobe.com/f4m/1.0}bootstrapInfo', '{http://ns.adobe.com/f4m/2.0}bootstrapInfo'],
2006 'bootstrap info', default=None)
2009 mime_type = xpath_text(
2010 manifest, ['{http://ns.adobe.com/f4m/1.0}mimeType', '{http://ns.adobe.com/f4m/2.0}mimeType'],
2011 'base URL', default=None)
2012 if mime_type and mime_type.startswith('audio/'):
2015 for i, media_el in enumerate(media_nodes):
2016 tbr = int_or_none(media_el.attrib.get('bitrate'))
2017 width = int_or_none(media_el.attrib.get('width'))
2018 height = int_or_none(media_el.attrib.get('height'))
2019 format_id = join_nonempty(f4m_id, tbr or i)
2020 # If <bootstrapInfo> is present, the specified f4m is a
2021 # stream-level manifest, and only set-level manifests may refer to
2022 # external resources. See section 11.4 and section 4 of F4M spec
2023 if bootstrap_info is None:
2025 # @href is introduced in 2.0, see section 11.6 of F4M spec
2026 if manifest_version == '2.0':
2027 media_url = media_el.attrib.get('href')
2028 if media_url is None:
2029 media_url = media_el.attrib.get('url')
2033 media_url if media_url.startswith('http://') or media_url.startswith('https://')
2034 else ((manifest_base_url or '/'.join(manifest_url.split('/')[:-1])) + '/' + media_url))
2035 # If media_url is itself a f4m manifest do the recursive extraction
2036 # since bitrates in parent manifest (this one) and media_url manifest
2037 # may differ leading to inability to resolve the format by requested
2038 # bitrate in f4m downloader
2039 ext = determine_ext(manifest_url)
2041 f4m_formats = self._extract_f4m_formats(
2042 manifest_url, video_id, preference=preference, quality=quality, f4m_id=f4m_id,
2043 transform_source=transform_source, fatal=fatal)
2044 # Sometimes stream-level manifest contains single media entry that
2045 # does not contain any quality metadata (e.g. http://matchtv.ru/#live-player).
2046 # At the same time parent's media entry in set-level manifest may
2047 # contain it. We will copy it from parent in such cases.
2048 if len(f4m_formats) == 1:
2051 'tbr': f.get('tbr') or tbr,
2052 'width': f.get('width') or width,
2053 'height': f.get('height') or height,
2054 'format_id': f.get('format_id') if not tbr else format_id,
2057 formats.extend(f4m_formats)
2060 formats.extend(self._extract_m3u8_formats(
2061 manifest_url, video_id, 'mp4', preference=preference,
2062 quality=quality, m3u8_id=m3u8_id, fatal=fatal))
2065 'format_id': format_id,
2066 'url': manifest_url,
2067 'manifest_url': manifest_url,
2068 'ext': 'flv' if bootstrap_info is not None else None,
2074 'preference': preference,
2079 def _m3u8_meta_format(self, m3u8_url, ext=None, preference=None, quality=None, m3u8_id=None):
2081 'format_id': join_nonempty(m3u8_id, 'meta'),
2085 'preference': preference - 100 if preference else -100,
2087 'resolution': 'multiple',
2088 'format_note': 'Quality selection URL',
2091 def _report_ignoring_subs(self, name):
2092 self.report_warning(bug_reports_message(
2093 f'Ignoring subtitle tracks found in the {name} manifest; '
2094 'if any subtitle tracks are missing,'
2097 def _extract_m3u8_formats(self, *args, **kwargs):
2098 fmts, subs = self._extract_m3u8_formats_and_subtitles(*args, **kwargs)
2100 self._report_ignoring_subs('HLS')
2103 def _extract_m3u8_formats_and_subtitles(
2104 self, m3u8_url, video_id, ext=None, entry_protocol='m3u8_native',
2105 preference=None, quality=None, m3u8_id=None, note=None,
2106 errnote=None, fatal=True, live=False, data=None, headers={},
2109 res = self._download_webpage_handle(
2111 note='Downloading m3u8 information' if note is None else note,
2112 errnote='Failed to download m3u8 information' if errnote is None else errnote,
2113 fatal=fatal, data=data, headers=headers, query=query)
2118 m3u8_doc, urlh = res
2119 m3u8_url = urlh.geturl()
2121 return self._parse_m3u8_formats_and_subtitles(
2122 m3u8_doc, m3u8_url, ext=ext, entry_protocol=entry_protocol,
2123 preference=preference, quality=quality, m3u8_id=m3u8_id,
2124 note=note, errnote=errnote, fatal=fatal, live=live, data=data,
2125 headers=headers, query=query, video_id=video_id)
2127 def _parse_m3u8_formats_and_subtitles(
2128 self, m3u8_doc, m3u8_url=None, ext=None, entry_protocol='m3u8_native',
2129 preference=None, quality=None, m3u8_id=None, live=False, note=None,
2130 errnote=None, fatal=True, data=None, headers={}, query={},
2132 formats, subtitles = [], {}
2134 has_drm = re.search('|'.join([
2135 r'#EXT-X-FAXS-CM:', # Adobe Flash Access
2136 r'#EXT-X-(?:SESSION-)?KEY:.*?URI="skd
://', # Apple FairPlay
2139 def format_url(url):
2140 return url if re.match(r'^https?
://', url) else compat_urlparse.urljoin(m3u8_url, url)
2142 if self.get_param('hls_split_discontinuity
', False):
2143 def _extract_m3u8_playlist_indices(manifest_url=None, m3u8_doc=None):
2145 if not manifest_url:
2147 m3u8_doc = self._download_webpage(
2148 manifest_url, video_id, fatal=fatal, data=data, headers=headers,
2149 note=False, errnote='Failed to download m3u8 playlist information
')
2150 if m3u8_doc is False:
2152 return range(1 + sum(line.startswith('#EXT-X-DISCONTINUITY') for line in m3u8_doc.splitlines()))
2155 def _extract_m3u8_playlist_indices(*args
, **kwargs
):
2159 # 1. https://tools.ietf.org/html/draft-pantos-http-live-streaming-21
2160 # 2. https://github.com/ytdl-org/youtube-dl/issues/12211
2161 # 3. https://github.com/ytdl-org/youtube-dl/issues/18923
2163 # We should try extracting formats only from master playlists [1, 4.3.4],
2164 # i.e. playlists that describe available qualities. On the other hand
2165 # media playlists [1, 4.3.3] should be returned as is since they contain
2166 # just the media without qualities renditions.
2167 # Fortunately, master playlist can be easily distinguished from media
2168 # playlist based on particular tags availability. As of [1, 4.3.3, 4.3.4]
2169 # master playlist tags MUST NOT appear in a media playlist and vice versa.
2170 # As of [1, 4.3.3.1] #EXT-X-TARGETDURATION tag is REQUIRED for every
2171 # media playlist and MUST NOT appear in master playlist thus we can
2172 # clearly detect media playlist with this criterion.
2174 if '#EXT-X-TARGETDURATION' in m3u8_doc
: # media playlist, return as is
2176 'format_id': join_nonempty(m3u8_id
, idx
),
2177 'format_index': idx
,
2178 'url': m3u8_url
or encode_data_uri(m3u8_doc
.encode('utf-8'), 'application/x-mpegurl'),
2180 'protocol': entry_protocol
,
2181 'preference': preference
,
2184 } for idx
in _extract_m3u8_playlist_indices(m3u8_doc
=m3u8_doc
)]
2186 return formats
, subtitles
2189 last_stream_inf
= {}
2191 def extract_media(x_media_line
):
2192 media
= parse_m3u8_attributes(x_media_line
)
2193 # As per [1, 4.3.4.1] TYPE, GROUP-ID and NAME are REQUIRED
2194 media_type
, group_id
, name
= media
.get('TYPE'), media
.get('GROUP-ID'), media
.get('NAME')
2195 if not (media_type
and group_id
and name
):
2197 groups
.setdefault(group_id
, []).append(media
)
2198 # <https://tools.ietf.org/html/rfc8216#section-4.3.4.1>
2199 if media_type
== 'SUBTITLES':
2200 # According to RFC 8216 §4.3.4.2.1, URI is REQUIRED in the
2201 # EXT-X-MEDIA tag if the media type is SUBTITLES.
2202 # However, lack of URI has been spotted in the wild.
2203 # e.g. NebulaIE; see https://github.com/yt-dlp/yt-dlp/issues/339
2204 if not media
.get('URI'):
2206 url
= format_url(media
['URI'])
2209 'ext': determine_ext(url
),
2211 if sub_info
['ext'] == 'm3u8':
2212 # Per RFC 8216 §3.1, the only possible subtitle format m3u8
2213 # files may contain is WebVTT:
2214 # <https://tools.ietf.org/html/rfc8216#section-3.1>
2215 sub_info
['ext'] = 'vtt'
2216 sub_info
['protocol'] = 'm3u8_native'
2217 lang
= media
.get('LANGUAGE') or 'und'
2218 subtitles
.setdefault(lang
, []).append(sub_info
)
2219 if media_type
not in ('VIDEO', 'AUDIO'):
2221 media_url
= media
.get('URI')
2223 manifest_url
= format_url(media_url
)
2225 'format_id': join_nonempty(m3u8_id
, group_id
, name
, idx
),
2226 'format_note': name
,
2227 'format_index': idx
,
2228 'url': manifest_url
,
2229 'manifest_url': m3u8_url
,
2230 'language': media
.get('LANGUAGE'),
2232 'protocol': entry_protocol
,
2233 'preference': preference
,
2235 'vcodec': 'none' if media_type
== 'AUDIO' else None,
2236 } for idx
in _extract_m3u8_playlist_indices(manifest_url
))
2238 def build_stream_name():
2239 # Despite specification does not mention NAME attribute for
2240 # EXT-X-STREAM-INF tag it still sometimes may be present (see [1]
2241 # or vidio test in TestInfoExtractor.test_parse_m3u8_formats)
2242 # 1. http://www.vidio.com/watch/165683-dj_ambred-booyah-live-2015
2243 stream_name
= last_stream_inf
.get('NAME')
2246 # If there is no NAME in EXT-X-STREAM-INF it will be obtained
2247 # from corresponding rendition group
2248 stream_group_id
= last_stream_inf
.get('VIDEO')
2249 if not stream_group_id
:
2251 stream_group
= groups
.get(stream_group_id
)
2252 if not stream_group
:
2253 return stream_group_id
2254 rendition
= stream_group
[0]
2255 return rendition
.get('NAME') or stream_group_id
2257 # parse EXT-X-MEDIA tags before EXT-X-STREAM-INF in order to have the
2258 # chance to detect video only formats when EXT-X-STREAM-INF tags
2259 # precede EXT-X-MEDIA tags in HLS manifest such as [3].
2260 for line
in m3u8_doc
.splitlines():
2261 if line
.startswith('#EXT-X-MEDIA:'):
2264 for line
in m3u8_doc
.splitlines():
2265 if line
.startswith('#EXT-X-STREAM-INF:'):
2266 last_stream_inf
= parse_m3u8_attributes(line
)
2267 elif line
.startswith('#') or not line
.strip():
2270 tbr
= float_or_none(
2271 last_stream_inf
.get('AVERAGE-BANDWIDTH')
2272 or last_stream_inf
.get('BANDWIDTH'), scale
=1000)
2273 manifest_url
= format_url(line
.strip())
2275 for idx
in _extract_m3u8_playlist_indices(manifest_url
):
2276 format_id
= [m3u8_id
, None, idx
]
2277 # Bandwidth of live streams may differ over time thus making
2278 # format_id unpredictable. So it's better to keep provided
2281 stream_name
= build_stream_name()
2282 format_id
[1] = stream_name
or '%d' % (tbr
or len(formats
))
2284 'format_id': join_nonempty(*format_id
),
2285 'format_index': idx
,
2286 'url': manifest_url
,
2287 'manifest_url': m3u8_url
,
2290 'fps': float_or_none(last_stream_inf
.get('FRAME-RATE')),
2291 'protocol': entry_protocol
,
2292 'preference': preference
,
2295 resolution
= last_stream_inf
.get('RESOLUTION')
2297 mobj
= re
.search(r
'(?P<width>\d+)[xX](?P<height>\d+)', resolution
)
2299 f
['width'] = int(mobj
.group('width'))
2300 f
['height'] = int(mobj
.group('height'))
2301 # Unified Streaming Platform
2303 r
'audio.*?(?:%3D|=)(\d+)(?:-video.*?(?:%3D|=)(\d+))?', f
['url'])
2305 abr
, vbr
= mobj
.groups()
2306 abr
, vbr
= float_or_none(abr
, 1000), float_or_none(vbr
, 1000)
2311 codecs
= parse_codecs(last_stream_inf
.get('CODECS'))
2313 audio_group_id
= last_stream_inf
.get('AUDIO')
2314 # As per [1, 4.3.4.1.1] any EXT-X-STREAM-INF tag which
2315 # references a rendition group MUST have a CODECS attribute.
2316 # However, this is not always respected, for example, [2]
2317 # contains EXT-X-STREAM-INF tag which references AUDIO
2318 # rendition group but does not have CODECS and despite
2319 # referencing an audio group it represents a complete
2320 # (with audio and video) format. So, for such cases we will
2321 # ignore references to rendition groups and treat them
2322 # as complete formats.
2323 if audio_group_id
and codecs
and f
.get('vcodec') != 'none':
2324 audio_group
= groups
.get(audio_group_id
)
2325 if audio_group
and audio_group
[0].get('URI'):
2326 # TODO: update acodec for audio only formats with
2328 f
['acodec'] = 'none'
2329 if not f
.get('ext'):
2330 f
['ext'] = 'm4a' if f
.get('vcodec') == 'none' else 'mp4'
2334 progressive_uri
= last_stream_inf
.get('PROGRESSIVE-URI')
2337 del http_f
['manifest_url']
2339 'format_id': f
['format_id'].replace('hls-', 'http-'),
2341 'url': progressive_uri
,
2343 formats
.append(http_f
)
2345 last_stream_inf
= {}
2346 return formats
, subtitles
2348 def _extract_m3u8_vod_duration(
2349 self
, m3u8_vod_url
, video_id
, note
=None, errnote
=None, data
=None, headers
={}, query={}
):
2351 m3u8_vod
= self
._download
_webpage
(
2352 m3u8_vod_url
, video_id
,
2353 note
='Downloading m3u8 VOD manifest' if note
is None else note
,
2354 errnote
='Failed to download VOD manifest' if errnote
is None else errnote
,
2355 fatal
=False, data
=data
, headers
=headers
, query
=query
)
2357 return self
._parse
_m
3u8_vod
_duration
(m3u8_vod
or '', video_id
)
2359 def _parse_m3u8_vod_duration(self
, m3u8_vod
, video_id
):
2360 if '#EXT-X-PLAYLIST-TYPE:VOD' not in m3u8_vod
:
2364 float(line
[len('#EXTINF:'):].split(',')[0])
2365 for line
in m3u8_vod
.splitlines() if line
.startswith('#EXTINF:'))) or None
2368 def _xpath_ns(path
, namespace
=None):
2372 for c
in path
.split('/'):
2373 if not c
or c
== '.':
2376 out
.append('{%s}%s' % (namespace
, c
))
2377 return '/'.join(out
)
2379 def _extract_smil_formats_and_subtitles(self
, smil_url
, video_id
, fatal
=True, f4m_params
=None, transform_source
=None):
2380 res
= self
._download
_smil
(smil_url
, video_id
, fatal
=fatal
, transform_source
=transform_source
)
2386 smil_url
= urlh
.geturl()
2388 namespace
= self
._parse
_smil
_namespace
(smil
)
2390 fmts
= self
._parse
_smil
_formats
(
2391 smil
, smil_url
, video_id
, namespace
=namespace
, f4m_params
=f4m_params
)
2392 subs
= self
._parse
_smil
_subtitles
(
2393 smil
, namespace
=namespace
)
2397 def _extract_smil_formats(self
, *args
, **kwargs
):
2398 fmts
, subs
= self
._extract
_smil
_formats
_and
_subtitles
(*args
, **kwargs
)
2400 self
._report
_ignoring
_subs
('SMIL')
2403 def _extract_smil_info(self
, smil_url
, video_id
, fatal
=True, f4m_params
=None):
2404 res
= self
._download
_smil
(smil_url
, video_id
, fatal
=fatal
)
2409 smil_url
= urlh
.geturl()
2411 return self
._parse
_smil
(smil
, smil_url
, video_id
, f4m_params
=f4m_params
)
2413 def _download_smil(self
, smil_url
, video_id
, fatal
=True, transform_source
=None):
2414 return self
._download
_xml
_handle
(
2415 smil_url
, video_id
, 'Downloading SMIL file',
2416 'Unable to download SMIL file', fatal
=fatal
, transform_source
=transform_source
)
2418 def _parse_smil(self
, smil
, smil_url
, video_id
, f4m_params
=None):
2419 namespace
= self
._parse
_smil
_namespace
(smil
)
2421 formats
= self
._parse
_smil
_formats
(
2422 smil
, smil_url
, video_id
, namespace
=namespace
, f4m_params
=f4m_params
)
2423 subtitles
= self
._parse
_smil
_subtitles
(smil
, namespace
=namespace
)
2425 video_id
= os
.path
.splitext(url_basename(smil_url
))[0]
2429 for meta
in smil
.findall(self
._xpath
_ns
('./head/meta', namespace
)):
2430 name
= meta
.attrib
.get('name')
2431 content
= meta
.attrib
.get('content')
2432 if not name
or not content
:
2434 if not title
and name
== 'title':
2436 elif not description
and name
in ('description', 'abstract'):
2437 description
= content
2438 elif not upload_date
and name
== 'date':
2439 upload_date
= unified_strdate(content
)
2442 'id': image
.get('type'),
2443 'url': image
.get('src'),
2444 'width': int_or_none(image
.get('width')),
2445 'height': int_or_none(image
.get('height')),
2446 } for image
in smil
.findall(self
._xpath
_ns
('.//image', namespace
)) if image
.get('src')]
2450 'title': title
or video_id
,
2451 'description': description
,
2452 'upload_date': upload_date
,
2453 'thumbnails': thumbnails
,
2455 'subtitles': subtitles
,
2458 def _parse_smil_namespace(self
, smil
):
2459 return self
._search
_regex
(
2460 r
'(?i)^{([^}]+)?}smil$', smil
.tag
, 'namespace', default
=None)
2462 def _parse_smil_formats(self
, smil
, smil_url
, video_id
, namespace
=None, f4m_params
=None, transform_rtmp_url
=None):
2464 for meta
in smil
.findall(self
._xpath
_ns
('./head/meta', namespace
)):
2465 b
= meta
.get('base') or meta
.get('httpBase')
2477 media
= smil
.findall(self
._xpath
_ns
('.//video', namespace
)) + smil
.findall(self
._xpath
_ns
('.//audio', namespace
))
2478 for medium
in media
:
2479 src
= medium
.get('src')
2480 if not src
or src
in srcs
:
2484 bitrate
= float_or_none(medium
.get('system-bitrate') or medium
.get('systemBitrate'), 1000)
2485 filesize
= int_or_none(medium
.get('size') or medium
.get('fileSize'))
2486 width
= int_or_none(medium
.get('width'))
2487 height
= int_or_none(medium
.get('height'))
2488 proto
= medium
.get('proto')
2489 ext
= medium
.get('ext')
2490 src_ext
= determine_ext(src
)
2491 streamer
= medium
.get('streamer') or base
2493 if proto
== 'rtmp' or streamer
.startswith('rtmp'):
2499 'format_id': 'rtmp-%d' % (rtmp_count
if bitrate
is None else bitrate
),
2501 'filesize': filesize
,
2505 if transform_rtmp_url
:
2506 streamer
, src
= transform_rtmp_url(streamer
, src
)
2507 formats
[-1].update({
2513 src_url
= src
if src
.startswith('http') else compat_urlparse
.urljoin(base
, src
)
2514 src_url
= src_url
.strip()
2516 if proto
== 'm3u8' or src_ext
== 'm3u8':
2517 m3u8_formats
= self
._extract
_m
3u8_formats
(
2518 src_url
, video_id
, ext
or 'mp4', m3u8_id
='hls', fatal
=False)
2519 if len(m3u8_formats
) == 1:
2521 m3u8_formats
[0].update({
2522 'format_id': 'hls-%d' % (m3u8_count
if bitrate
is None else bitrate
),
2527 formats
.extend(m3u8_formats
)
2528 elif src_ext
== 'f4m':
2533 'plugin': 'flowplayer-3.2.0.1',
2535 f4m_url
+= '&' if '?' in f4m_url
else '?'
2536 f4m_url
+= compat_urllib_parse_urlencode(f4m_params
)
2537 formats
.extend(self
._extract
_f
4m
_formats
(f4m_url
, video_id
, f4m_id
='hds', fatal
=False))
2538 elif src_ext
== 'mpd':
2539 formats
.extend(self
._extract
_mpd
_formats
(
2540 src_url
, video_id
, mpd_id
='dash', fatal
=False))
2541 elif re
.search(r
'\.ism/[Mm]anifest', src_url
):
2542 formats
.extend(self
._extract
_ism
_formats
(
2543 src_url
, video_id
, ism_id
='mss', fatal
=False))
2544 elif src_url
.startswith('http') and self
._is
_valid
_url
(src
, video_id
):
2548 'ext': ext
or src_ext
or 'flv',
2549 'format_id': 'http-%d' % (bitrate
or http_count
),
2551 'filesize': filesize
,
2556 for medium
in smil
.findall(self
._xpath
_ns
('.//imagestream', namespace
)):
2557 src
= medium
.get('src')
2558 if not src
or src
in srcs
:
2564 'format_id': 'imagestream-%d' % (imgs_count
),
2566 'ext': mimetype2ext(medium
.get('type')),
2569 'width': int_or_none(medium
.get('width')),
2570 'height': int_or_none(medium
.get('height')),
2571 'format_note': 'SMIL storyboards',
2576 def _parse_smil_subtitles(self
, smil
, namespace
=None, subtitles_lang
='en'):
2579 for num
, textstream
in enumerate(smil
.findall(self
._xpath
_ns
('.//textstream', namespace
))):
2580 src
= textstream
.get('src')
2581 if not src
or src
in urls
:
2584 ext
= textstream
.get('ext') or mimetype2ext(textstream
.get('type')) or determine_ext(src
)
2585 lang
= textstream
.get('systemLanguage') or textstream
.get('systemLanguageName') or textstream
.get('lang') or subtitles_lang
2586 subtitles
.setdefault(lang
, []).append({
2592 def _extract_xspf_playlist(self
, xspf_url
, playlist_id
, fatal
=True):
2593 res
= self
._download
_xml
_handle
(
2594 xspf_url
, playlist_id
, 'Downloading xpsf playlist',
2595 'Unable to download xspf manifest', fatal
=fatal
)
2600 xspf_url
= urlh
.geturl()
2602 return self
._parse
_xspf
(
2603 xspf
, playlist_id
, xspf_url
=xspf_url
,
2604 xspf_base_url
=base_url(xspf_url
))
2606 def _parse_xspf(self
, xspf_doc
, playlist_id
, xspf_url
=None, xspf_base_url
=None):
2608 'xspf': 'http://xspf.org/ns/0/',
2609 's1': 'http://static.streamone.nl/player/ns/0',
2613 for track
in xspf_doc
.findall(xpath_with_ns('./xspf:trackList/xspf:track', NS_MAP
)):
2615 track
, xpath_with_ns('./xspf:title', NS_MAP
), 'title', default
=playlist_id
)
2616 description
= xpath_text(
2617 track
, xpath_with_ns('./xspf:annotation', NS_MAP
), 'description')
2618 thumbnail
= xpath_text(
2619 track
, xpath_with_ns('./xspf:image', NS_MAP
), 'thumbnail')
2620 duration
= float_or_none(
2621 xpath_text(track
, xpath_with_ns('./xspf:duration', NS_MAP
), 'duration'), 1000)
2624 for location
in track
.findall(xpath_with_ns('./xspf:location', NS_MAP
)):
2625 format_url
= urljoin(xspf_base_url
, location
.text
)
2630 'manifest_url': xspf_url
,
2631 'format_id': location
.get(xpath_with_ns('s1:label', NS_MAP
)),
2632 'width': int_or_none(location
.get(xpath_with_ns('s1:width', NS_MAP
))),
2633 'height': int_or_none(location
.get(xpath_with_ns('s1:height', NS_MAP
))),
2635 self
._sort
_formats
(formats
)
2640 'description': description
,
2641 'thumbnail': thumbnail
,
2642 'duration': duration
,
2647 def _extract_mpd_formats(self
, *args
, **kwargs
):
2648 fmts
, subs
= self
._extract
_mpd
_formats
_and
_subtitles
(*args
, **kwargs
)
2650 self
._report
_ignoring
_subs
('DASH')
2653 def _extract_mpd_formats_and_subtitles(
2654 self
, mpd_url
, video_id
, mpd_id
=None, note
=None, errnote
=None,
2655 fatal
=True, data
=None, headers
={}, query={}
):
2656 res
= self
._download
_xml
_handle
(
2658 note
='Downloading MPD manifest' if note
is None else note
,
2659 errnote
='Failed to download MPD manifest' if errnote
is None else errnote
,
2660 fatal
=fatal
, data
=data
, headers
=headers
, query
=query
)
2667 # We could have been redirected to a new url when we retrieved our mpd file.
2668 mpd_url
= urlh
.geturl()
2669 mpd_base_url
= base_url(mpd_url
)
2671 return self
._parse
_mpd
_formats
_and
_subtitles
(
2672 mpd_doc
, mpd_id
, mpd_base_url
, mpd_url
)
2674 def _parse_mpd_formats(self
, *args
, **kwargs
):
2675 fmts
, subs
= self
._parse
_mpd
_formats
_and
_subtitles
(*args
, **kwargs
)
2677 self
._report
_ignoring
_subs
('DASH')
2680 def _parse_mpd_formats_and_subtitles(
2681 self
, mpd_doc
, mpd_id
=None, mpd_base_url
='', mpd_url
=None):
2683 Parse formats from MPD manifest.
2685 1. MPEG-DASH Standard, ISO/IEC 23009-1:2014(E),
2686 http://standards.iso.org/ittf/PubliclyAvailableStandards/c065274_ISO_IEC_23009-1_2014.zip
2687 2. https://en.wikipedia.org/wiki/Dynamic_Adaptive_Streaming_over_HTTP
2689 if not self
.get_param('dynamic_mpd', True):
2690 if mpd_doc
.get('type') == 'dynamic':
2693 namespace
= self
._search
_regex
(r
'(?i)^{([^}]+)?}MPD$', mpd_doc
.tag
, 'namespace', default
=None)
2696 return self
._xpath
_ns
(path
, namespace
)
2698 def is_drm_protected(element
):
2699 return element
.find(_add_ns('ContentProtection')) is not None
2701 def extract_multisegment_info(element
, ms_parent_info
):
2702 ms_info
= ms_parent_info
.copy()
2704 # As per [1, 5.3.9.2.2] SegmentList and SegmentTemplate share some
2705 # common attributes and elements. We will only extract relevant
2707 def extract_common(source
):
2708 segment_timeline
= source
.find(_add_ns('SegmentTimeline'))
2709 if segment_timeline
is not None:
2710 s_e
= segment_timeline
.findall(_add_ns('S'))
2712 ms_info
['total_number'] = 0
2715 r
= int(s
.get('r', 0))
2716 ms_info
['total_number'] += 1 + r
2717 ms_info
['s'].append({
2718 't': int(s
.get('t', 0)),
2719 # @d is mandatory (see [1, 5.3.9.6.2, Table 17, page 60])
2720 'd': int(s
.attrib
['d']),
2723 start_number
= source
.get('startNumber')
2725 ms_info
['start_number'] = int(start_number
)
2726 timescale
= source
.get('timescale')
2728 ms_info
['timescale'] = int(timescale
)
2729 segment_duration
= source
.get('duration')
2730 if segment_duration
:
2731 ms_info
['segment_duration'] = float(segment_duration
)
2733 def extract_Initialization(source
):
2734 initialization
= source
.find(_add_ns('Initialization'))
2735 if initialization
is not None:
2736 ms_info
['initialization_url'] = initialization
.attrib
['sourceURL']
2738 segment_list
= element
.find(_add_ns('SegmentList'))
2739 if segment_list
is not None:
2740 extract_common(segment_list
)
2741 extract_Initialization(segment_list
)
2742 segment_urls_e
= segment_list
.findall(_add_ns('SegmentURL'))
2744 ms_info
['segment_urls'] = [segment
.attrib
['media'] for segment
in segment_urls_e
]
2746 segment_template
= element
.find(_add_ns('SegmentTemplate'))
2747 if segment_template
is not None:
2748 extract_common(segment_template
)
2749 media
= segment_template
.get('media')
2751 ms_info
['media'] = media
2752 initialization
= segment_template
.get('initialization')
2754 ms_info
['initialization'] = initialization
2756 extract_Initialization(segment_template
)
2759 mpd_duration
= parse_duration(mpd_doc
.get('mediaPresentationDuration'))
2760 formats
, subtitles
= [], {}
2761 stream_numbers
= collections
.defaultdict(int)
2762 for period
in mpd_doc
.findall(_add_ns('Period')):
2763 period_duration
= parse_duration(period
.get('duration')) or mpd_duration
2764 period_ms_info
= extract_multisegment_info(period
, {
2768 for adaptation_set
in period
.findall(_add_ns('AdaptationSet')):
2769 adaption_set_ms_info
= extract_multisegment_info(adaptation_set
, period_ms_info
)
2770 for representation
in adaptation_set
.findall(_add_ns('Representation')):
2771 representation_attrib
= adaptation_set
.attrib
.copy()
2772 representation_attrib
.update(representation
.attrib
)
2773 # According to [1, 5.3.7.2, Table 9, page 41], @mimeType is mandatory
2774 mime_type
= representation_attrib
['mimeType']
2775 content_type
= representation_attrib
.get('contentType', mime_type
.split('/')[0])
2777 codec_str
= representation_attrib
.get('codecs', '')
2778 # Some kind of binary subtitle found in some youtube livestreams
2779 if mime_type
== 'application/x-rawcc':
2780 codecs
= {'scodec': codec_str}
2782 codecs
= parse_codecs(codec_str
)
2783 if content_type
not in ('video', 'audio', 'text'):
2784 if mime_type
== 'image/jpeg':
2785 content_type
= mime_type
2786 elif codecs
.get('vcodec', 'none') != 'none':
2787 content_type
= 'video'
2788 elif codecs
.get('acodec', 'none') != 'none':
2789 content_type
= 'audio'
2790 elif codecs
.get('scodec', 'none') != 'none':
2791 content_type
= 'text'
2792 elif mimetype2ext(mime_type
) in ('tt', 'dfxp', 'ttml', 'xml', 'json'):
2793 content_type
= 'text'
2795 self
.report_warning('Unknown MIME type %s in DASH manifest' % mime_type
)
2799 for element
in (representation
, adaptation_set
, period
, mpd_doc
):
2800 base_url_e
= element
.find(_add_ns('BaseURL'))
2801 if base_url_e
is not None:
2802 base_url
= base_url_e
.text
+ base_url
2803 if re
.match(r
'^https?://', base_url
):
2805 if mpd_base_url
and base_url
.startswith('/'):
2806 base_url
= compat_urlparse
.urljoin(mpd_base_url
, base_url
)
2807 elif mpd_base_url
and not re
.match(r
'^https?://', base_url
):
2808 if not mpd_base_url
.endswith('/'):
2810 base_url
= mpd_base_url
+ base_url
2811 representation_id
= representation_attrib
.get('id')
2812 lang
= representation_attrib
.get('lang')
2813 url_el
= representation
.find(_add_ns('BaseURL'))
2814 filesize
= int_or_none(url_el
.attrib
.get('{http://youtube.com/yt/2012/10/10}contentLength') if url_el
is not None else None)
2815 bandwidth
= int_or_none(representation_attrib
.get('bandwidth'))
2816 if representation_id
is not None:
2817 format_id
= representation_id
2819 format_id
= content_type
2821 format_id
= mpd_id
+ '-' + format_id
2822 if content_type
in ('video', 'audio'):
2824 'format_id': format_id
,
2825 'manifest_url': mpd_url
,
2826 'ext': mimetype2ext(mime_type
),
2827 'width': int_or_none(representation_attrib
.get('width')),
2828 'height': int_or_none(representation_attrib
.get('height')),
2829 'tbr': float_or_none(bandwidth
, 1000),
2830 'asr': int_or_none(representation_attrib
.get('audioSamplingRate')),
2831 'fps': int_or_none(representation_attrib
.get('frameRate')),
2832 'language': lang
if lang
not in ('mul', 'und', 'zxx', 'mis') else None,
2833 'format_note': 'DASH %s' % content_type
,
2834 'filesize': filesize
,
2835 'container': mimetype2ext(mime_type
) + '_dash',
2838 elif content_type
== 'text':
2840 'ext': mimetype2ext(mime_type
),
2841 'manifest_url': mpd_url
,
2842 'filesize': filesize
,
2844 elif content_type
== 'image/jpeg':
2845 # See test case in VikiIE
2846 # https://www.viki.com/videos/1175236v-choosing-spouse-by-lottery-episode-1
2848 'format_id': format_id
,
2850 'manifest_url': mpd_url
,
2851 'format_note': 'DASH storyboards (jpeg)',
2855 if is_drm_protected(adaptation_set
) or is_drm_protected(representation
):
2857 representation_ms_info
= extract_multisegment_info(representation
, adaption_set_ms_info
)
2859 def prepare_template(template_name
, identifiers
):
2860 tmpl
= representation_ms_info
[template_name
]
2861 # First of, % characters outside $...$ templates
2862 # must be escaped by doubling for proper processing
2863 # by % operator string formatting used further (see
2864 # https://github.com/ytdl-org/youtube-dl/issues/16867).
2870 in_template
= not in_template
2871 elif c
== '%' and not in_template
:
2873 # Next, $...$ templates are translated to their
2874 # %(...) counterparts to be used with % operator
2875 if representation_id
is not None:
2876 t
= t
.replace('$RepresentationID$', representation_id
)
2877 t
= re
.sub(r
'\$(%s)\$' % '|'.join(identifiers
), r
'%(\1)d', t
)
2878 t
= re
.sub(r
'\$(%s)%%([^$]+)\$' % '|'.join(identifiers
), r
'%(\1)\2', t
)
2879 t
.replace('$$', '$')
2882 # @initialization is a regular template like @media one
2883 # so it should be handled just the same way (see
2884 # https://github.com/ytdl-org/youtube-dl/issues/11605)
2885 if 'initialization' in representation_ms_info
:
2886 initialization_template
= prepare_template(
2888 # As per [1, 5.3.9.4.2, Table 15, page 54] $Number$ and
2889 # $Time$ shall not be included for @initialization thus
2890 # only $Bandwidth$ remains
2892 representation_ms_info
['initialization_url'] = initialization_template
% {
2893 'Bandwidth': bandwidth
,
2896 def location_key(location
):
2897 return 'url' if re
.match(r
'^https?://', location
) else 'path'
2899 if 'segment_urls' not in representation_ms_info
and 'media' in representation_ms_info
:
2901 media_template
= prepare_template('media', ('Number', 'Bandwidth', 'Time'))
2902 media_location_key
= location_key(media_template
)
2904 # As per [1, 5.3.9.4.4, Table 16, page 55] $Number$ and $Time$
2905 # can't be used at the same time
2906 if '%(Number' in media_template
and 's' not in representation_ms_info
:
2907 segment_duration
= None
2908 if 'total_number' not in representation_ms_info
and 'segment_duration' in representation_ms_info
:
2909 segment_duration
= float_or_none(representation_ms_info
['segment_duration'], representation_ms_info
['timescale'])
2910 representation_ms_info
['total_number'] = int(math
.ceil(
2911 float_or_none(period_duration
, segment_duration
, default
=0)))
2912 representation_ms_info
['fragments'] = [{
2913 media_location_key
: media_template
% {
2914 'Number': segment_number
,
2915 'Bandwidth': bandwidth
,
2917 'duration': segment_duration
,
2918 } for segment_number
in range(
2919 representation_ms_info
['start_number'],
2920 representation_ms_info
['total_number'] + representation_ms_info
['start_number'])]
2922 # $Number*$ or $Time$ in media template with S list available
2923 # Example $Number*$: http://www.svtplay.se/klipp/9023742/stopptid-om-bjorn-borg
2924 # Example $Time$: https://play.arkena.com/embed/avp/v2/player/media/b41dda37-d8e7-4d3f-b1b5-9a9db578bdfe/1/129411
2925 representation_ms_info
['fragments'] = []
2928 segment_number
= representation_ms_info
['start_number']
2930 def add_segment_url():
2931 segment_url
= media_template
% {
2932 'Time': segment_time
,
2933 'Bandwidth': bandwidth
,
2934 'Number': segment_number
,
2936 representation_ms_info
['fragments'].append({
2937 media_location_key
: segment_url
,
2938 'duration': float_or_none(segment_d
, representation_ms_info
['timescale']),
2941 for num
, s
in enumerate(representation_ms_info
['s']):
2942 segment_time
= s
.get('t') or segment_time
2946 for r
in range(s
.get('r', 0)):
2947 segment_time
+= segment_d
2950 segment_time
+= segment_d
2951 elif 'segment_urls' in representation_ms_info
and 's' in representation_ms_info
:
2953 # Example: https://www.youtube.com/watch?v=iXZV5uAYMJI
2954 # or any YouTube dashsegments video
2957 timescale
= representation_ms_info
['timescale']
2958 for s
in representation_ms_info
['s']:
2959 duration
= float_or_none(s
['d'], timescale
)
2960 for r
in range(s
.get('r', 0) + 1):
2961 segment_uri
= representation_ms_info
['segment_urls'][segment_index
]
2963 location_key(segment_uri
): segment_uri
,
2964 'duration': duration
,
2967 representation_ms_info
['fragments'] = fragments
2968 elif 'segment_urls' in representation_ms_info
:
2969 # Segment URLs with no SegmentTimeline
2970 # Example: https://www.seznam.cz/zpravy/clanek/cesko-zasahne-vitr-o-sile-vichrice-muze-byt-i-zivotu-nebezpecny-39091
2971 # https://github.com/ytdl-org/youtube-dl/pull/14844
2973 segment_duration
= float_or_none(
2974 representation_ms_info
['segment_duration'],
2975 representation_ms_info
['timescale']) if 'segment_duration' in representation_ms_info
else None
2976 for segment_url
in representation_ms_info
['segment_urls']:
2978 location_key(segment_url
): segment_url
,
2980 if segment_duration
:
2981 fragment
['duration'] = segment_duration
2982 fragments
.append(fragment
)
2983 representation_ms_info
['fragments'] = fragments
2984 # If there is a fragments key available then we correctly recognized fragmented media.
2985 # Otherwise we will assume unfragmented media with direct access. Technically, such
2986 # assumption is not necessarily correct since we may simply have no support for
2987 # some forms of fragmented media renditions yet, but for now we'll use this fallback.
2988 if 'fragments' in representation_ms_info
:
2990 # NB: mpd_url may be empty when MPD manifest is parsed from a string
2991 'url': mpd_url
or base_url
,
2992 'fragment_base_url': base_url
,
2994 'protocol': 'http_dash_segments' if mime_type
!= 'image/jpeg' else 'mhtml',
2996 if 'initialization_url' in representation_ms_info
:
2997 initialization_url
= representation_ms_info
['initialization_url']
2998 if not f
.get('url'):
2999 f
['url'] = initialization_url
3000 f
['fragments'].append({location_key(initialization_url): initialization_url}
)
3001 f
['fragments'].extend(representation_ms_info
['fragments'])
3002 if not period_duration
:
3003 period_duration
= try_get(
3004 representation_ms_info
,
3005 lambda r
: sum(frag
['duration'] for frag
in r
['fragments']), float)
3007 # Assuming direct URL to unfragmented media.
3009 if content_type
in ('video', 'audio', 'image/jpeg'):
3010 f
['manifest_stream_number'] = stream_numbers
[f
['url']]
3011 stream_numbers
[f
['url']] += 1
3013 elif content_type
== 'text':
3014 subtitles
.setdefault(lang
or 'und', []).append(f
)
3016 return formats
, subtitles
3018 def _extract_ism_formats(self
, *args
, **kwargs
):
3019 fmts
, subs
= self
._extract
_ism
_formats
_and
_subtitles
(*args
, **kwargs
)
3021 self
._report
_ignoring
_subs
('ISM')
3024 def _extract_ism_formats_and_subtitles(self
, ism_url
, video_id
, ism_id
=None, note
=None, errnote
=None, fatal
=True, data
=None, headers
={}, query={}
):
3025 res
= self
._download
_xml
_handle
(
3027 note
='Downloading ISM manifest' if note
is None else note
,
3028 errnote
='Failed to download ISM manifest' if errnote
is None else errnote
,
3029 fatal
=fatal
, data
=data
, headers
=headers
, query
=query
)
3036 return self
._parse
_ism
_formats
_and
_subtitles
(ism_doc
, urlh
.geturl(), ism_id
)
3038 def _parse_ism_formats_and_subtitles(self
, ism_doc
, ism_url
, ism_id
=None):
3040 Parse formats from ISM manifest.
3042 1. [MS-SSTR]: Smooth Streaming Protocol,
3043 https://msdn.microsoft.com/en-us/library/ff469518.aspx
3045 if ism_doc
.get('IsLive') == 'TRUE':
3048 duration
= int(ism_doc
.attrib
['Duration'])
3049 timescale
= int_or_none(ism_doc
.get('TimeScale')) or 10000000
3053 for stream
in ism_doc
.findall('StreamIndex'):
3054 stream_type
= stream
.get('Type')
3055 if stream_type
not in ('video', 'audio', 'text'):
3057 url_pattern
= stream
.attrib
['Url']
3058 stream_timescale
= int_or_none(stream
.get('TimeScale')) or timescale
3059 stream_name
= stream
.get('Name')
3060 stream_language
= stream
.get('Language', 'und')
3061 for track
in stream
.findall('QualityLevel'):
3062 fourcc
= track
.get('FourCC') or ('AACL' if track
.get('AudioTag') == '255' else None)
3063 # TODO: add support for WVC1 and WMAP
3064 if fourcc
not in ('H264', 'AVC1', 'AACL', 'TTML'):
3065 self
.report_warning('%s is not a supported codec' % fourcc
)
3067 tbr
= int(track
.attrib
['Bitrate']) // 1000
3068 # [1] does not mention Width and Height attributes. However,
3069 # they're often present while MaxWidth and MaxHeight are
3070 # missing, so should be used as fallbacks
3071 width
= int_or_none(track
.get('MaxWidth') or track
.get('Width'))
3072 height
= int_or_none(track
.get('MaxHeight') or track
.get('Height'))
3073 sampling_rate
= int_or_none(track
.get('SamplingRate'))
3075 track_url_pattern
= re
.sub(r
'{[Bb]itrate}', track
.attrib
['Bitrate'], url_pattern
)
3076 track_url_pattern
= compat_urlparse
.urljoin(ism_url
, track_url_pattern
)
3082 stream_fragments
= stream
.findall('c')
3083 for stream_fragment_index
, stream_fragment
in enumerate(stream_fragments
):
3084 fragment_ctx
['time'] = int_or_none(stream_fragment
.get('t')) or fragment_ctx
['time']
3085 fragment_repeat
= int_or_none(stream_fragment
.get('r')) or 1
3086 fragment_ctx
['duration'] = int_or_none(stream_fragment
.get('d'))
3087 if not fragment_ctx
['duration']:
3089 next_fragment_time
= int(stream_fragment
[stream_fragment_index
+ 1].attrib
['t'])
3091 next_fragment_time
= duration
3092 fragment_ctx
['duration'] = (next_fragment_time
- fragment_ctx
['time']) / fragment_repeat
3093 for _
in range(fragment_repeat
):
3095 'url': re
.sub(r
'{start[ _]time}', compat_str(fragment_ctx
['time']), track_url_pattern
),
3096 'duration': fragment_ctx
['duration'] / stream_timescale
,
3098 fragment_ctx
['time'] += fragment_ctx
['duration']
3100 if stream_type
== 'text':
3101 subtitles
.setdefault(stream_language
, []).append({
3105 'manifest_url': ism_url
,
3106 'fragments': fragments
,
3107 '_download_params': {
3108 'stream_type': stream_type
,
3109 'duration': duration
,
3110 'timescale': stream_timescale
,
3112 'language': stream_language
,
3113 'codec_private_data': track
.get('CodecPrivateData'),
3116 elif stream_type
in ('video', 'audio'):
3118 'format_id': join_nonempty(ism_id
, stream_name
, tbr
),
3120 'manifest_url': ism_url
,
3121 'ext': 'ismv' if stream_type
== 'video' else 'isma',
3125 'asr': sampling_rate
,
3126 'vcodec': 'none' if stream_type
== 'audio' else fourcc
,
3127 'acodec': 'none' if stream_type
== 'video' else fourcc
,
3129 'fragments': fragments
,
3130 'has_drm': ism_doc
.find('Protection') is not None,
3131 '_download_params': {
3132 'stream_type': stream_type
,
3133 'duration': duration
,
3134 'timescale': stream_timescale
,
3135 'width': width
or 0,
3136 'height': height
or 0,
3138 'language': stream_language
,
3139 'codec_private_data': track
.get('CodecPrivateData'),
3140 'sampling_rate': sampling_rate
,
3141 'channels': int_or_none(track
.get('Channels', 2)),
3142 'bits_per_sample': int_or_none(track
.get('BitsPerSample', 16)),
3143 'nal_unit_length_field': int_or_none(track
.get('NALUnitLengthField', 4)),
3146 return formats
, subtitles
3148 def _parse_html5_media_entries(self
, base_url
, webpage
, video_id
, m3u8_id
=None, m3u8_entry_protocol
='m3u8_native', mpd_id
=None, preference
=None, quality
=None):
3149 def absolute_url(item_url
):
3150 return urljoin(base_url
, item_url
)
3152 def parse_content_type(content_type
):
3153 if not content_type
:
3155 ctr
= re
.search(r
'(?P<mimetype>[^/]+/[^;]+)(?:;\s*codecs="?(?P<codecs>[^"]+))?', content_type
)
3157 mimetype
, codecs
= ctr
.groups()
3158 f
= parse_codecs(codecs
)
3159 f
['ext'] = mimetype2ext(mimetype
)
3163 def _media_formats(src
, cur_media_type
, type_info
=None):
3164 type_info
= type_info
or {}
3165 full_url
= absolute_url(src
)
3166 ext
= type_info
.get('ext') or determine_ext(full_url
)
3168 is_plain_url
= False
3169 formats
= self
._extract
_m
3u8_formats
(
3170 full_url
, video_id
, ext
='mp4',
3171 entry_protocol
=m3u8_entry_protocol
, m3u8_id
=m3u8_id
,
3172 preference
=preference
, quality
=quality
, fatal
=False)
3174 is_plain_url
= False
3175 formats
= self
._extract
_mpd
_formats
(
3176 full_url
, video_id
, mpd_id
=mpd_id
, fatal
=False)
3181 'vcodec': 'none' if cur_media_type
== 'audio' else None,
3184 return is_plain_url
, formats
3187 # amp-video and amp-audio are very similar to their HTML5 counterparts
3188 # so we wll include them right here (see
3189 # https://www.ampproject.org/docs/reference/components/amp-video)
3190 # For dl8-* tags see https://delight-vr.com/documentation/dl8-video/
3191 _MEDIA_TAG_NAME_RE
= r
'(?:(?:amp|dl8(?:-live)?)-)?(video|audio)'
3192 media_tags
= [(media_tag
, media_tag_name
, media_type
, '')
3193 for media_tag
, media_tag_name
, media_type
3194 in re
.findall(r
'(?s)(<(%s)[^>]*/>)' % _MEDIA_TAG_NAME_RE
, webpage
)]
3195 media_tags
.extend(re
.findall(
3196 # We only allow video|audio followed by a whitespace or '>'.
3197 # Allowing more characters may end up in significant slow down (see
3198 # https://github.com/ytdl-org/youtube-dl/issues/11979, example URL:
3199 # http://www.porntrex.com/maps/videositemap.xml).
3200 r
'(?s)(<(?P<tag>%s)(?:\s+[^>]*)?>)(.*?)</(?P=tag)>' % _MEDIA_TAG_NAME_RE
, webpage
))
3201 for media_tag
, _
, media_type
, media_content
in media_tags
:
3206 media_attributes
= extract_attributes(media_tag
)
3207 src
= strip_or_none(media_attributes
.get('src'))
3209 f
= parse_content_type(media_attributes
.get('type'))
3210 _
, formats
= _media_formats(src
, media_type
, f
)
3211 media_info
['formats'].extend(formats
)
3212 media_info
['thumbnail'] = absolute_url(media_attributes
.get('poster'))
3214 for source_tag
in re
.findall(r
'<source[^>]+>', media_content
):
3215 s_attr
= extract_attributes(source_tag
)
3216 # data-video-src and data-src are non standard but seen
3217 # several times in the wild
3218 src
= strip_or_none(dict_get(s_attr
, ('src', 'data-video-src', 'data-src')))
3221 f
= parse_content_type(s_attr
.get('type'))
3222 is_plain_url
, formats
= _media_formats(src
, media_type
, f
)
3224 # width, height, res, label and title attributes are
3225 # all not standard but seen several times in the wild
3228 for lbl
in ('label', 'title')
3229 if str_or_none(s_attr
.get(lbl
))
3231 width
= int_or_none(s_attr
.get('width'))
3232 height
= (int_or_none(s_attr
.get('height'))
3233 or int_or_none(s_attr
.get('res')))
3234 if not width
or not height
:
3236 resolution
= parse_resolution(lbl
)
3239 width
= width
or resolution
.get('width')
3240 height
= height
or resolution
.get('height')
3242 tbr
= parse_bitrate(lbl
)
3251 'format_id': s_attr
.get('label') or s_attr
.get('title'),
3253 f
.update(formats
[0])
3254 media_info
['formats'].append(f
)
3256 media_info
['formats'].extend(formats
)
3257 for track_tag
in re
.findall(r
'<track[^>]+>', media_content
):
3258 track_attributes
= extract_attributes(track_tag
)
3259 kind
= track_attributes
.get('kind')
3260 if not kind
or kind
in ('subtitles', 'captions'):
3261 src
= strip_or_none(track_attributes
.get('src'))
3264 lang
= track_attributes
.get('srclang') or track_attributes
.get('lang') or track_attributes
.get('label')
3265 media_info
['subtitles'].setdefault(lang
, []).append({
3266 'url': absolute_url(src
),
3268 for f
in media_info
['formats']:
3269 f
.setdefault('http_headers', {})['Referer'] = base_url
3270 if media_info
['formats'] or media_info
['subtitles']:
3271 entries
.append(media_info
)
3274 def _extract_akamai_formats(self
, *args
, **kwargs
):
3275 fmts
, subs
= self
._extract
_akamai
_formats
_and
_subtitles
(*args
, **kwargs
)
3277 self
._report
_ignoring
_subs
('akamai')
3280 def _extract_akamai_formats_and_subtitles(self
, manifest_url
, video_id
, hosts
={}):
3281 signed
= 'hdnea=' in manifest_url
3283 # https://learn.akamai.com/en-us/webhelp/media-services-on-demand/stream-packaging-user-guide/GUID-BE6C0F73-1E06-483B-B0EA-57984B91B7F9.html
3284 manifest_url
= re
.sub(
3285 r
'(?:b=[\d,-]+|(?:__a__|attributes)=off|__b__=\d+)&?',
3286 '', manifest_url
).strip('?')
3291 hdcore_sign
= 'hdcore=3.7.0'
3292 f4m_url
= re
.sub(r
'(https?://[^/]+)/i/', r
'\1/z/', manifest_url
).replace('/master.m3u8', '/manifest.f4m')
3293 hds_host
= hosts
.get('hds')
3295 f4m_url
= re
.sub(r
'(https?://)[^/]+', r
'\1' + hds_host
, f4m_url
)
3296 if 'hdcore=' not in f4m_url
:
3297 f4m_url
+= ('&' if '?' in f4m_url
else '?') + hdcore_sign
3298 f4m_formats
= self
._extract
_f
4m
_formats
(
3299 f4m_url
, video_id
, f4m_id
='hds', fatal
=False)
3300 for entry
in f4m_formats
:
3301 entry
.update({'extra_param_to_segment_url': hdcore_sign}
)
3302 formats
.extend(f4m_formats
)
3304 m3u8_url
= re
.sub(r
'(https?://[^/]+)/z/', r
'\1/i/', manifest_url
).replace('/manifest.f4m', '/master.m3u8')
3305 hls_host
= hosts
.get('hls')
3307 m3u8_url
= re
.sub(r
'(https?://)[^/]+', r
'\1' + hls_host
, m3u8_url
)
3308 m3u8_formats
, m3u8_subtitles
= self
._extract
_m
3u8_formats
_and
_subtitles
(
3309 m3u8_url
, video_id
, 'mp4', 'm3u8_native',
3310 m3u8_id
='hls', fatal
=False)
3311 formats
.extend(m3u8_formats
)
3312 subtitles
= self
._merge
_subtitles
(subtitles
, m3u8_subtitles
)
3314 http_host
= hosts
.get('http')
3315 if http_host
and m3u8_formats
and not signed
:
3316 REPL_REGEX
= r
'https?://[^/]+/i/([^,]+),([^/]+),([^/]+)\.csmil/.+'
3317 qualities
= re
.match(REPL_REGEX
, m3u8_url
).group(2).split(',')
3318 qualities_length
= len(qualities
)
3319 if len(m3u8_formats
) in (qualities_length
, qualities_length
+ 1):
3321 for f
in m3u8_formats
:
3322 if f
['vcodec'] != 'none':
3323 for protocol
in ('http', 'https'):
3325 del http_f
['manifest_url']
3327 REPL_REGEX
, protocol
+ fr
'://{http_host}/\g<1>{qualities[i]}\3', f
['url'])
3329 'format_id': http_f
['format_id'].replace('hls-', protocol
+ '-'),
3331 'protocol': protocol
,
3333 formats
.append(http_f
)
3336 return formats
, subtitles
3338 def _extract_wowza_formats(self
, url
, video_id
, m3u8_entry_protocol
='m3u8_native', skip_protocols
=[]):
3339 query
= compat_urlparse
.urlparse(url
).query
3340 url
= re
.sub(r
'/(?:manifest|playlist|jwplayer)\.(?:m3u8|f4m|mpd|smil)', '', url
)
3342 r
'(?:(?:http|rtmp|rtsp)(?P<s>s)?:)?(?P<url>//[^?]+)', url
)
3343 url_base
= mobj
.group('url')
3344 http_base_url
= '%s%s:%s' % ('http', mobj
.group('s') or '', url_base
)
3347 def manifest_url(manifest
):
3348 m_url
= f
'{http_base_url}/{manifest}'
3350 m_url
+= '?%s' % query
3353 if 'm3u8' not in skip_protocols
:
3354 formats
.extend(self
._extract
_m
3u8_formats
(
3355 manifest_url('playlist.m3u8'), video_id
, 'mp4',
3356 m3u8_entry_protocol
, m3u8_id
='hls', fatal
=False))
3357 if 'f4m' not in skip_protocols
:
3358 formats
.extend(self
._extract
_f
4m
_formats
(
3359 manifest_url('manifest.f4m'),
3360 video_id
, f4m_id
='hds', fatal
=False))
3361 if 'dash' not in skip_protocols
:
3362 formats
.extend(self
._extract
_mpd
_formats
(
3363 manifest_url('manifest.mpd'),
3364 video_id
, mpd_id
='dash', fatal
=False))
3365 if re
.search(r
'(?:/smil:|\.smil)', url_base
):
3366 if 'smil' not in skip_protocols
:
3367 rtmp_formats
= self
._extract
_smil
_formats
(
3368 manifest_url('jwplayer.smil'),
3369 video_id
, fatal
=False)
3370 for rtmp_format
in rtmp_formats
:
3371 rtsp_format
= rtmp_format
.copy()
3372 rtsp_format
['url'] = '%s/%s' % (rtmp_format
['url'], rtmp_format
['play_path'])
3373 del rtsp_format
['play_path']
3374 del rtsp_format
['ext']
3375 rtsp_format
.update({
3376 'url': rtsp_format
['url'].replace('rtmp://', 'rtsp://'),
3377 'format_id': rtmp_format
['format_id'].replace('rtmp', 'rtsp'),
3380 formats
.extend([rtmp_format
, rtsp_format
])
3382 for protocol
in ('rtmp', 'rtsp'):
3383 if protocol
not in skip_protocols
:
3385 'url': f
'{protocol}:{url_base}',
3386 'format_id': protocol
,
3387 'protocol': protocol
,
3391 def _find_jwplayer_data(self
, webpage
, video_id
=None, transform_source
=js_to_json
):
3393 r
'(?s)jwplayer\((?P<quote>[\'"])[^\'" ]+(?P
=quote
)\
)(?
!</script
>).*?\
.setup\s
*\
((?P
<options
>[^
)]+)\
)',
3397 jwplayer_data = self._parse_json(mobj.group('options
'),
3399 transform_source=transform_source)
3400 except ExtractorError:
3403 if isinstance(jwplayer_data, dict):
3404 return jwplayer_data
3406 def _extract_jwplayer_data(self, webpage, video_id, *args, **kwargs):
3407 jwplayer_data = self._find_jwplayer_data(
3408 webpage, video_id, transform_source=js_to_json)
3409 return self._parse_jwplayer_data(
3410 jwplayer_data, video_id, *args, **kwargs)
3412 def _parse_jwplayer_data(self, jwplayer_data, video_id=None, require_title=True,
3413 m3u8_id=None, mpd_id=None, rtmp_params=None, base_url=None):
3414 # JWPlayer backward compatibility: flattened playlists
3415 # https://github.com/jwplayer/jwplayer/blob/v7.4.3/src/js/api/config.js#L81-L96
3416 if 'playlist
' not in jwplayer_data:
3417 jwplayer_data = {'playlist': [jwplayer_data]}
3421 # JWPlayer backward compatibility: single playlist item
3422 # https://github.com/jwplayer/jwplayer/blob/v7.7.0/src/js/playlist/playlist.js#L10
3423 if not isinstance(jwplayer_data['playlist
'], list):
3424 jwplayer_data['playlist
'] = [jwplayer_data['playlist
']]
3426 for video_data in jwplayer_data['playlist
']:
3427 # JWPlayer backward compatibility: flattened sources
3428 # https://github.com/jwplayer/jwplayer/blob/v7.4.3/src/js/playlist/item.js#L29-L35
3429 if 'sources
' not in video_data:
3430 video_data['sources
'] = [video_data]
3432 this_video_id = video_id or video_data['mediaid
']
3434 formats = self._parse_jwplayer_formats(
3435 video_data['sources
'], video_id=this_video_id, m3u8_id=m3u8_id,
3436 mpd_id=mpd_id, rtmp_params=rtmp_params, base_url=base_url)
3439 tracks = video_data.get('tracks
')
3440 if tracks and isinstance(tracks, list):
3441 for track in tracks:
3442 if not isinstance(track, dict):
3444 track_kind = track.get('kind
')
3445 if not track_kind or not isinstance(track_kind, compat_str):
3447 if track_kind.lower() not in ('captions
', 'subtitles
'):
3449 track_url = urljoin(base_url, track.get('file'))
3452 subtitles.setdefault(track.get('label
') or 'en
', []).append({
3453 'url
': self._proto_relative_url(track_url)
3457 'id': this_video_id,
3458 'title
': unescapeHTML(video_data['title
'] if require_title else video_data.get('title
')),
3459 'description
': clean_html(video_data.get('description
')),
3460 'thumbnail
': urljoin(base_url, self._proto_relative_url(video_data.get('image
'))),
3461 'timestamp
': int_or_none(video_data.get('pubdate
')),
3462 'duration
': float_or_none(jwplayer_data.get('duration
') or video_data.get('duration
')),
3463 'subtitles
': subtitles,
3465 # https://github.com/jwplayer/jwplayer/blob/master/src/js/utils/validator.js#L32
3466 if len(formats) == 1 and re.search(r'^
(?
:http|
//).*(?
:youtube\
.com|youtu\
.be
)/.+', formats[0]['url
']):
3468 '_type
': 'url_transparent
',
3469 'url
': formats[0]['url
'],
3472 self._sort_formats(formats)
3473 entry['formats
'] = formats
3474 entries.append(entry)
3475 if len(entries) == 1:
3478 return self.playlist_result(entries)
3480 def _parse_jwplayer_formats(self, jwplayer_sources_data, video_id=None,
3481 m3u8_id=None, mpd_id=None, rtmp_params=None, base_url=None):
3484 for source in jwplayer_sources_data:
3485 if not isinstance(source, dict):
3487 source_url = urljoin(
3488 base_url, self._proto_relative_url(source.get('file')))
3489 if not source_url or source_url in urls:
3491 urls.append(source_url)
3492 source_type = source.get('type') or ''
3493 ext = mimetype2ext(source_type) or determine_ext(source_url)
3494 if source_type == 'hls
' or ext == 'm3u8
':
3495 formats.extend(self._extract_m3u8_formats(
3496 source_url, video_id, 'mp4
', entry_protocol='m3u8_native
',
3497 m3u8_id=m3u8_id, fatal=False))
3498 elif source_type == 'dash
' or ext == 'mpd
':
3499 formats.extend(self._extract_mpd_formats(
3500 source_url, video_id, mpd_id=mpd_id, fatal=False))
3502 formats.extend(self._extract_smil_formats(
3503 source_url, video_id, fatal=False))
3504 # https://github.com/jwplayer/jwplayer/blob/master/src/js/providers/default.js#L67
3505 elif source_type.startswith('audio
') or ext in (
3506 'oga
', 'aac
', 'mp3
', 'mpeg
', 'vorbis
'):
3513 height = int_or_none(source.get('height
'))
3515 # Often no height is provided but there is a label in
3516 # format like "1080p", "720p SD", or 1080.
3517 height = int_or_none(self._search_regex(
3518 r'^
(\d{3,4}
)[pP
]?
(?
:\b|$
)', compat_str(source.get('label
') or ''),
3519 'height
', default=None))
3522 'width
': int_or_none(source.get('width
')),
3524 'tbr
': int_or_none(source.get('bitrate
')),
3527 if source_url.startswith('rtmp
'):
3528 a_format['ext
'] = 'flv
'
3529 # See com/longtailvideo/jwplayer/media/RTMPMediaProvider.as
3530 # of jwplayer.flash.swf
3531 rtmp_url_parts = re.split(
3532 r'((?
:mp4|mp3|flv
):)', source_url, 1)
3533 if len(rtmp_url_parts) == 3:
3534 rtmp_url, prefix, play_path = rtmp_url_parts
3537 'play_path
': prefix + play_path,
3540 a_format.update(rtmp_params)
3541 formats.append(a_format)
3544 def _live_title(self, name):
3545 self._downloader.deprecation_warning('yt_dlp
.InfoExtractor
._live
_title
is deprecated
and does
not work
as expected
')
3548 def _int(self, v, name, fatal=False, **kwargs):
3549 res = int_or_none(v, **kwargs)
3551 msg = f'Failed to extract {name}
: Could
not parse value {v!r}
'
3553 raise ExtractorError(msg)
3555 self.report_warning(msg)
3558 def _float(self, v, name, fatal=False, **kwargs):
3559 res = float_or_none(v, **kwargs)
3561 msg = f'Failed to extract {name}
: Could
not parse value {v!r}
'
3563 raise ExtractorError(msg)
3565 self.report_warning(msg)
3568 def _set_cookie(self, domain, name, value, expire_time=None, port=None,
3569 path='/', secure=False, discard=False, rest={}, **kwargs):
3570 cookie = compat_cookiejar_Cookie(
3571 0, name, value, port, port is not None, domain, True,
3572 domain.startswith('.'), path, True, secure, expire_time,
3573 discard, None, None, rest)
3574 self._downloader.cookiejar.set_cookie(cookie)
3576 def _get_cookies(self, url):
3577 """ Return a compat_cookies_SimpleCookie with the cookies for the url """
3578 return compat_cookies_SimpleCookie(self._downloader._calc_cookies(url))
3580 def _apply_first_set_cookie_header(self, url_handle, cookie):
3582 Apply first Set-Cookie header instead of the last. Experimental.
3584 Some sites (e.g. [1-3]) may serve two cookies under the same name
3585 in Set-Cookie header and expect the first (old) one to be set rather
3586 than second (new). However, as of RFC6265 the newer one cookie
3587 should be set into cookie store what actually happens.
3588 We will workaround this issue by resetting the cookie to
3589 the first one manually.
3590 1. https://new.vk.com/
3591 2. https://github.com/ytdl-org/youtube-dl/issues/9841#issuecomment-227871201
3592 3. https://learning.oreilly.com/
3594 for header, cookies in url_handle.headers.items():
3595 if header.lower() != 'set-cookie
':
3597 cookies = cookies.encode('iso
-8859-1').decode('utf
-8')
3598 cookie_value = re.search(
3599 r'%s=(.+?
);.*?
\b[Dd
]omain
=(.+?
)(?
:[,;]|$
)' % cookie, cookies)
3601 value, domain = cookie_value.groups()
3602 self._set_cookie(domain, cookie, value)
3606 def get_testcases(cls, include_onlymatching=False):
3607 t = getattr(cls, '_TEST
', None)
3609 assert not hasattr(cls, '_TESTS
'), f'{cls.ie_key()}IE has _TEST
and _TESTS
'
3612 tests = getattr(cls, '_TESTS
', [])
3614 if not include_onlymatching and t.get('only_matching
', False):
3616 t['name
'] = cls.ie_key()
3621 """Get age limit from the testcases"""
3622 return max(traverse_obj(
3623 tuple(cls.get_testcases(include_onlymatching=False)),
3624 (..., (('playlist
', 0), None), 'info_dict
', 'age_limit
')) or [0])
3627 def is_suitable(cls, age_limit):
3628 """Test whether the extractor is generally suitable for the given age limit"""
3629 return not age_restricted(cls.age_limit, age_limit)
3632 def description(cls, *, markdown=True, search_examples=None):
3633 """Description of the extractor"""
3635 if cls._NETRC_MACHINE:
3637 desc += f' [<abbr title
="netrc machine"><em
>{cls._NETRC_MACHINE}
</em
></abbr
>]'
3639 desc += f' [{cls._NETRC_MACHINE}
]'
3640 if cls.IE_DESC is False:
3643 desc += f' {cls.IE_DESC}
'
3645 desc += f'; "{cls.SEARCH_KEY}:" prefix
'
3647 _COUNTS = ('', '5', '10', 'all
')
3648 desc += f' (Example
: "{cls.SEARCH_KEY}{random.choice(_COUNTS)}:{random.choice(search_examples)}")'
3649 if not cls.working():
3650 desc += ' (**Currently broken
**)' if markdown else ' (Currently broken
)'
3652 name = f' - **{cls.IE_NAME}
**' if markdown else cls.IE_NAME
3653 return f'{name}
:{desc}
' if desc else name
3655 def extract_subtitles(self, *args, **kwargs):
3656 if (self.get_param('writesubtitles
', False)
3657 or self.get_param('listsubtitles
')):
3658 return self._get_subtitles(*args, **kwargs)
3661 def _get_subtitles(self, *args, **kwargs):
3662 raise NotImplementedError('This method must be implemented by subclasses
')
3664 def extract_comments(self, *args, **kwargs):
3665 if not self.get_param('getcomments
'):
3667 generator = self._get_comments(*args, **kwargs)
3674 comments.append(next(generator))
3675 except StopIteration:
3677 except KeyboardInterrupt:
3678 self.to_screen('Interrupted by user
')
3679 except Exception as e:
3680 if self.get_param('ignoreerrors
') is not True:
3682 self._downloader.report_error(e)
3683 comment_count = len(comments)
3684 self.to_screen(f'Extracted {comment_count} comments
')
3686 'comments
': comments,
3687 'comment_count
': None if interrupted else comment_count
3691 def _get_comments(self, *args, **kwargs):
3692 raise NotImplementedError('This method must be implemented by subclasses
')
3695 def _merge_subtitle_items(subtitle_list1, subtitle_list2):
3696 """ Merge subtitle items for one language. Items with duplicated URLs/data
3697 will be dropped. """
3698 list1_data = {(item.get('url'), item.get('data')) for item in subtitle_list1}
3699 ret = list(subtitle_list1)
3700 ret.extend(item for item in subtitle_list2 if (item.get('url
'), item.get('data
')) not in list1_data)
3704 def _merge_subtitles(cls, *dicts, target=None):
3705 """ Merge subtitle dictionaries, language by language. """
3709 for lang, subs in d.items():
3710 target[lang] = cls._merge_subtitle_items(target.get(lang, []), subs)
3713 def extract_automatic_captions(self, *args, **kwargs):
3714 if (self.get_param('writeautomaticsub
', False)
3715 or self.get_param('listsubtitles
')):
3716 return self._get_automatic_captions(*args, **kwargs)
3719 def _get_automatic_captions(self, *args, **kwargs):
3720 raise NotImplementedError('This method must be implemented by subclasses
')
3722 @functools.cached_property
3723 def _cookies_passed(self):
3724 """Whether cookies have been passed to YoutubeDL"""
3725 return self.get_param('cookiefile
') is not None or self.get_param('cookiesfrombrowser
') is not None
3727 def mark_watched(self, *args, **kwargs):
3728 if not self.get_param('mark_watched
', False):
3730 if self.supports_login() and self._get_login_info()[0] is not None or self._cookies_passed:
3731 self._mark_watched(*args, **kwargs)
3733 def _mark_watched(self, *args, **kwargs):
3734 raise NotImplementedError('This method must be implemented by subclasses
')
3736 def geo_verification_headers(self):
3738 geo_verification_proxy = self.get_param('geo_verification_proxy
')
3739 if geo_verification_proxy:
3740 headers['Ytdl
-request
-proxy
'] = geo_verification_proxy
3743 def _generic_id(self, url):
3744 return compat_urllib_parse_unquote(os.path.splitext(url.rstrip('/').split('/')[-1])[0])
3746 def _generic_title(self, url):
3747 return compat_urllib_parse_unquote(os.path.splitext(url_basename(url))[0])
3750 def _availability(is_private=None, needs_premium=None, needs_subscription=None, needs_auth=None, is_unlisted=None):
3751 all_known = all(map(
3752 lambda x: x is not None,
3753 (is_private, needs_premium, needs_subscription, needs_auth, is_unlisted)))
3755 'private
' if is_private
3756 else 'premium_only
' if needs_premium
3757 else 'subscriber_only
' if needs_subscription
3758 else 'needs_auth
' if needs_auth
3759 else 'unlisted
' if is_unlisted
3760 else 'public
' if all_known
3763 def _configuration_arg(self, key, default=NO_DEFAULT, *, ie_key=None, casesense=False):
3765 @returns A list of values for the extractor argument given by "key"
3766 or "default" if no such key is present
3767 @param default The default value to return when the key is not present (default: [])
3768 @param casesense When false, the values are converted to lower case
3771 self._downloader.params, ('extractor_args
', (ie_key or self.ie_key()).lower(), key))
3773 return [] if default is NO_DEFAULT else default
3774 return list(val) if casesense else [x.lower() for x in val]
3776 def _yes_playlist(self, playlist_id, video_id, smuggled_data=None, *, playlist_label='playlist
', video_label='video
'):
3777 if not playlist_id or not video_id:
3780 no_playlist = (smuggled_data or {}).get('force_noplaylist
')
3781 if no_playlist is not None:
3782 return not no_playlist
3784 video_id = '' if video_id is True else f' {video_id}
'
3785 playlist_id = '' if playlist_id is True else f' {playlist_id}
'
3786 if self.get_param('noplaylist
'):
3787 self.to_screen(f'Downloading just the {video_label}{video_id} because of
--no
-playlist
')
3789 self.to_screen(f'Downloading {playlist_label}{playlist_id}
- add
--no
-playlist to download just the {video_label}{video_id}
')
3793 class SearchInfoExtractor(InfoExtractor):
3795 Base class for paged search queries extractors.
3796 They accept URLs in the format _SEARCH_KEY(|all|[0-9]):{query}
3797 Instances should define _SEARCH_KEY and optionally _MAX_RESULTS
3800 _MAX_RESULTS = float('inf
')
3803 def _make_valid_url(cls):
3804 return r'%s(?P
<prefix
>|
[1-9][0-9]*|all
):(?P
<query
>[\s\S
]+)' % cls._SEARCH_KEY
3806 def _real_extract(self, query):
3807 prefix, query = self._match_valid_url(query).group('prefix
', 'query
')
3809 return self._get_n_results(query, 1)
3810 elif prefix == 'all
':
3811 return self._get_n_results(query, self._MAX_RESULTS)
3815 raise ExtractorError(f'invalid download number {n}
for query
"{query}"')
3816 elif n > self._MAX_RESULTS:
3817 self.report_warning('%s returns
max %i results (you requested
%i)' % (self._SEARCH_KEY, self._MAX_RESULTS, n))
3818 n = self._MAX_RESULTS
3819 return self._get_n_results(query, n)
3821 def _get_n_results(self, query, n):
3822 """Get a specified number of results for a query.
3823 Either this function or _search_results must be overridden by subclasses """
3824 return self.playlist_result(
3825 itertools.islice(self._search_results(query), 0, None if n == float('inf
') else n),
3828 def _search_results(self, query):
3829 """Returns an iterator of search results"""
3830 raise NotImplementedError('This method must be implemented by subclasses
')
3833 def SEARCH_KEY(cls):
3834 return cls._SEARCH_KEY