12 import xml
.etree
.ElementTree
14 from ..compat
import (
15 compat_cookiejar_Cookie
,
16 compat_cookies_SimpleCookie
,
17 compat_etree_fromstring
,
24 compat_urllib_parse_unquote
,
25 compat_urllib_parse_urlencode
,
26 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
._downloader
.write_debug(
614 '[debug] Using fake IP %s as X-Forwarded-For' % self
._x
_forwarded
_for
_ip
)
617 # Path 2: bypassing based on country code
619 # Explicit country code specified by user, use it right away
620 # regardless of whether extractor is geo bypassable or not
621 country
= self
.get_param('geo_bypass_country', None)
623 # Otherwise use random country code from geo bypass context but
624 # only if extractor is known as geo bypassable
626 countries
= geo_bypass_context
.get('countries')
627 if self
._GEO
_BYPASS
and countries
:
628 country
= random
.choice(countries
)
631 self
._x
_forwarded
_for
_ip
= GeoUtils
.random_ipv4(country
)
632 self
._downloader
.write_debug(
633 f
'Using fake IP {self._x_forwarded_for_ip} ({country.upper()}) as X-Forwarded-For')
635 def extract(self
, url
):
636 """Extracts URL information and returns it in list of dicts."""
641 self
.write_debug('Extracting URL: %s' % url
)
642 ie_result
= self
._real
_extract
(url
)
643 if ie_result
is None:
645 if self
._x
_forwarded
_for
_ip
:
646 ie_result
['__x_forwarded_for_ip'] = self
._x
_forwarded
_for
_ip
647 subtitles
= ie_result
.get('subtitles')
648 if (subtitles
and 'live_chat' in subtitles
649 and 'no-live-chat' in self
.get_param('compat_opts', [])):
650 del subtitles
['live_chat']
652 except GeoRestrictedError
as e
:
653 if self
.__maybe
_fake
_ip
_and
_retry
(e
.countries
):
656 except UnsupportedError
:
658 except ExtractorError
as e
:
660 'video_id': e
.video_id
or self
.get_temp_id(url
),
662 'tb': e
.traceback
or sys
.exc_info()[2],
663 'expected': e
.expected
,
666 if hasattr(e
, 'countries'):
667 kwargs
['countries'] = e
.countries
668 raise type(e
)(e
.orig_msg
, **kwargs
)
669 except compat_http_client
.IncompleteRead
as e
:
670 raise ExtractorError('A network error has occurred.', cause
=e
, expected
=True, video_id
=self
.get_temp_id(url
))
671 except (KeyError, StopIteration) as e
:
672 raise ExtractorError('An extractor error has occurred.', cause
=e
, video_id
=self
.get_temp_id(url
))
674 def __maybe_fake_ip_and_retry(self
, countries
):
675 if (not self
.get_param('geo_bypass_country', None)
677 and self
.get_param('geo_bypass', True)
678 and not self
._x
_forwarded
_for
_ip
680 country_code
= random
.choice(countries
)
681 self
._x
_forwarded
_for
_ip
= GeoUtils
.random_ipv4(country_code
)
682 if self
._x
_forwarded
_for
_ip
:
684 'Video is geo restricted. Retrying extraction with fake IP %s (%s) as X-Forwarded-For.'
685 % (self
._x
_forwarded
_for
_ip
, country_code
.upper()))
689 def set_downloader(self
, downloader
):
690 """Sets a YoutubeDL instance as the downloader for this IE."""
691 self
._downloader
= downloader
693 def _initialize_pre_login(self
):
694 """ Intialization before login. Redefine in subclasses."""
697 def _perform_login(self
, username
, password
):
698 """ Login with username and password. Redefine in subclasses."""
701 def _real_initialize(self
):
702 """Real initialization process. Redefine in subclasses."""
705 def _real_extract(self
, url
):
706 """Real extraction process. Redefine in subclasses."""
707 raise NotImplementedError('This method must be implemented by subclasses')
711 """A string for getting the InfoExtractor with get_info_extractor"""
712 return cls
.__name
__[:-2]
716 return cls
.__name
__[:-2]
719 def __can_accept_status_code(err
, expected_status
):
720 assert isinstance(err
, compat_urllib_error
.HTTPError
)
721 if expected_status
is None:
723 elif callable(expected_status
):
724 return expected_status(err
.code
) is True
726 return err
.code
in variadic(expected_status
)
728 def _request_webpage(self
, url_or_request
, video_id
, note
=None, errnote
=None, fatal
=True, data
=None, headers
={}, query={}
, expected_status
=None):
730 Return the response handle.
732 See _download_webpage docstring for arguments specification.
734 if not self
._downloader
._first
_webpage
_request
:
735 sleep_interval
= self
.get_param('sleep_interval_requests') or 0
736 if sleep_interval
> 0:
737 self
.to_screen('Sleeping %s seconds ...' % sleep_interval
)
738 time
.sleep(sleep_interval
)
740 self
._downloader
._first
_webpage
_request
= False
743 self
.report_download_webpage(video_id
)
744 elif note
is not False:
746 self
.to_screen(str(note
))
748 self
.to_screen(f
'{video_id}: {note}')
750 # Some sites check X-Forwarded-For HTTP header in order to figure out
751 # the origin of the client behind proxy. This allows bypassing geo
752 # restriction by faking this header's value to IP that belongs to some
753 # geo unrestricted country. We will do so once we encounter any
754 # geo restriction error.
755 if self
._x
_forwarded
_for
_ip
:
756 if 'X-Forwarded-For' not in headers
:
757 headers
['X-Forwarded-For'] = self
._x
_forwarded
_for
_ip
759 if isinstance(url_or_request
, compat_urllib_request
.Request
):
760 url_or_request
= update_Request(
761 url_or_request
, data
=data
, headers
=headers
, query
=query
)
764 url_or_request
= update_url_query(url_or_request
, query
)
765 if data
is not None or headers
:
766 url_or_request
= sanitized_Request(url_or_request
, data
, headers
)
768 return self
._downloader
.urlopen(url_or_request
)
769 except network_exceptions
as err
:
770 if isinstance(err
, compat_urllib_error
.HTTPError
):
771 if self
.__can
_accept
_status
_code
(err
, expected_status
):
772 # Retain reference to error to prevent file object from
773 # being closed before it can be read. Works around the
774 # effects of <https://bugs.python.org/issue15002>
775 # introduced in Python 3.4.1.
782 errnote
= 'Unable to download webpage'
784 errmsg
= f
'{errnote}: {error_to_compat_str(err)}'
786 raise ExtractorError(errmsg
, cause
=err
)
788 self
.report_warning(errmsg
)
791 def _download_webpage_handle(self
, url_or_request
, video_id
, note
=None, errnote
=None, fatal
=True, encoding
=None, data
=None, headers
={}, query={}
, expected_status
=None):
793 Return a tuple (page content as string, URL handle).
795 See _download_webpage docstring for arguments specification.
797 # Strip hashes from the URL (#1038)
798 if isinstance(url_or_request
, (compat_str
, str)):
799 url_or_request
= url_or_request
.partition('#')[0]
801 urlh
= self
._request
_webpage
(url_or_request
, video_id
, note
, errnote
, fatal
, data
=data
, headers
=headers
, query
=query
, expected_status
=expected_status
)
805 content
= self
._webpage
_read
_content
(urlh
, url_or_request
, video_id
, note
, errnote
, fatal
, encoding
=encoding
)
806 return (content
, urlh
)
809 def _guess_encoding_from_content(content_type
, webpage_bytes
):
810 m
= re
.match(r
'[a-zA-Z0-9_.-]+/[a-zA-Z0-9_.-]+\s*;\s*charset=(.+)', content_type
)
812 encoding
= m
.group(1)
814 m
= re
.search(br
'<meta[^>]+charset=[\'"]?([^\'")]+)[ /\'">]',
815 webpage_bytes[:1024])
817 encoding = m.group(1).decode('ascii')
818 elif webpage_bytes.startswith(b'\xff\xfe'):
825 def __check_blocked(self, content):
826 first_block = content[:512]
827 if ('<title>Access to this site is blocked</title>' in content
828 and 'Websense' in first_block):
829 msg = 'Access to this webpage has been blocked by Websense filtering software in your network.'
830 blocked_iframe = self._html_search_regex(
831 r'<iframe src="([^
"]+)"', content,
832 'Websense information URL
', default=None)
834 msg += ' Visit
%s for more details
' % blocked_iframe
835 raise ExtractorError(msg, expected=True)
836 if '<title
>The URL you requested has been blocked
</title
>' in first_block:
838 'Access to this webpage has been blocked by Indian censorship
. '
839 'Use a VPN
or proxy
server (with --proxy
) to route around it
.')
840 block_msg = self._html_search_regex(
841 r'</h1
><p
>(.*?
)</p
>',
842 content, 'block message
', default=None)
844 msg += ' (Message
: "%s")' % block_msg.replace('\n', ' ')
845 raise ExtractorError(msg, expected=True)
846 if ('<title
>TTK
:: Доступ к ресурсу ограничен
</title
>' in content
847 and 'blocklist
.rkn
.gov
.ru
' in content):
848 raise ExtractorError(
849 'Access to this webpage has been blocked by decision of the Russian government
. '
850 'Visit http
://blocklist
.rkn
.gov
.ru
/ for a block reason
.',
853 def _webpage_read_content(self, urlh, url_or_request, video_id, note=None, errnote=None, fatal=True, prefix=None, encoding=None):
854 content_type = urlh.headers.get('Content
-Type
', '')
855 webpage_bytes = urlh.read()
856 if prefix is not None:
857 webpage_bytes = prefix + webpage_bytes
859 encoding = self._guess_encoding_from_content(content_type, webpage_bytes)
860 if self.get_param('dump_intermediate_pages
', False):
861 self.to_screen('Dumping request to
' + urlh.geturl())
862 dump = base64.b64encode(webpage_bytes).decode('ascii
')
863 self._downloader.to_screen(dump)
864 if self.get_param('write_pages
', False):
865 basen = f'{video_id}_{urlh.geturl()}
'
866 trim_length = self.get_param('trim_file_name
') or 240
867 if len(basen) > trim_length:
868 h = '___
' + hashlib.md5(basen.encode('utf
-8')).hexdigest()
869 basen = basen[:trim_length - len(h)] + h
870 raw_filename = basen + '.dump
'
871 filename = sanitize_filename(raw_filename, restricted=True)
872 self.to_screen('Saving request to
' + filename)
873 # Working around MAX_PATH limitation on Windows (see
874 # http://msdn.microsoft.com/en-us/library/windows/desktop/aa365247(v=vs.85).aspx)
875 if compat_os_name == 'nt
':
876 absfilepath = os.path.abspath(filename)
877 if len(absfilepath) > 259:
878 filename = '\\\\?
\\' + absfilepath
879 with open(filename, 'wb
') as outf:
880 outf.write(webpage_bytes)
883 content = webpage_bytes.decode(encoding, 'replace
')
885 content = webpage_bytes.decode('utf
-8', 'replace
')
887 self.__check_blocked(content)
891 def _download_webpage(
892 self, url_or_request, video_id, note=None, errnote=None,
893 fatal=True, tries=1, timeout=5, encoding=None, data=None,
894 headers={}, query={}, expected_status=None):
896 Return the data of the page as a string.
899 url_or_request -- plain text URL as a string or
900 a compat_urllib_request.Requestobject
901 video_id -- Video/playlist/item identifier (string)
904 note -- note printed before downloading (string)
905 errnote -- note printed in case of an error (string)
906 fatal -- flag denoting whether error should be considered fatal,
907 i.e. whether it should cause ExtractionError to be raised,
908 otherwise a warning will be reported and extraction continued
909 tries -- number of tries
910 timeout -- sleep interval between tries
911 encoding -- encoding for a page content decoding, guessed automatically
912 when not explicitly specified
913 data -- POST data (bytes)
914 headers -- HTTP headers (dict)
915 query -- URL query (dict)
916 expected_status -- allows to accept failed HTTP requests (non 2xx
917 status code) by explicitly specifying a set of accepted status
918 codes. Can be any of the following entities:
919 - an integer type specifying an exact failed status code to
921 - a list or a tuple of integer types specifying a list of
922 failed status codes to accept
923 - a callable accepting an actual failed status code and
924 returning True if it should be accepted
925 Note that this argument does not affect success status codes (2xx)
926 which are always accepted.
931 while success is False:
933 res = self._download_webpage_handle(
934 url_or_request, video_id, note, errnote, fatal,
935 encoding=encoding, data=data, headers=headers, query=query,
936 expected_status=expected_status)
938 except compat_http_client.IncompleteRead as e:
940 if try_count >= tries:
942 self._sleep(timeout, video_id)
949 def _download_xml_handle(
950 self, url_or_request, video_id, note='Downloading XML
',
951 errnote='Unable to download XML
', transform_source=None,
952 fatal=True, encoding=None, data=None, headers={}, query={},
953 expected_status=None):
955 Return a tuple (xml as an xml.etree.ElementTree.Element, URL handle).
957 See _download_webpage docstring for arguments specification.
959 res = self._download_webpage_handle(
960 url_or_request, video_id, note, errnote, fatal=fatal,
961 encoding=encoding, data=data, headers=headers, query=query,
962 expected_status=expected_status)
965 xml_string, urlh = res
966 return self._parse_xml(
967 xml_string, video_id, transform_source=transform_source,
971 self, url_or_request, video_id,
972 note='Downloading XML
', errnote='Unable to download XML
',
973 transform_source=None, fatal=True, encoding=None,
974 data=None, headers={}, query={}, expected_status=None):
976 Return the xml as an xml.etree.ElementTree.Element.
978 See _download_webpage docstring for arguments specification.
980 res = self._download_xml_handle(
981 url_or_request, video_id, note=note, errnote=errnote,
982 transform_source=transform_source, fatal=fatal, encoding=encoding,
983 data=data, headers=headers, query=query,
984 expected_status=expected_status)
985 return res if res is False else res[0]
987 def _parse_xml(self, xml_string, video_id, transform_source=None, fatal=True):
989 xml_string = transform_source(xml_string)
991 return compat_etree_fromstring(xml_string.encode('utf
-8'))
992 except xml.etree.ElementTree.ParseError as ve:
993 errmsg = '%s: Failed to parse XML
' % video_id
995 raise ExtractorError(errmsg, cause=ve)
997 self.report_warning(errmsg + str(ve))
999 def _download_json_handle(
1000 self, url_or_request, video_id, note='Downloading JSON metadata
',
1001 errnote='Unable to download JSON metadata
', transform_source=None,
1002 fatal=True, encoding=None, data=None, headers={}, query={},
1003 expected_status=None):
1005 Return a tuple (JSON object, URL handle).
1007 See _download_webpage docstring for arguments specification.
1009 res = self._download_webpage_handle(
1010 url_or_request, video_id, note, errnote, fatal=fatal,
1011 encoding=encoding, data=data, headers=headers, query=query,
1012 expected_status=expected_status)
1015 json_string, urlh = res
1016 return self._parse_json(
1017 json_string, video_id, transform_source=transform_source,
1021 self, url_or_request, video_id, note='Downloading JSON metadata
',
1022 errnote='Unable to download JSON metadata
', transform_source=None,
1023 fatal=True, encoding=None, data=None, headers={}, query={},
1024 expected_status=None):
1026 Return the JSON object as a dict.
1028 See _download_webpage docstring for arguments specification.
1030 res = self._download_json_handle(
1031 url_or_request, video_id, note=note, errnote=errnote,
1032 transform_source=transform_source, fatal=fatal, encoding=encoding,
1033 data=data, headers=headers, query=query,
1034 expected_status=expected_status)
1035 return res if res is False else res[0]
1037 def _parse_json(self, json_string, video_id, transform_source=None, fatal=True):
1038 if transform_source:
1039 json_string = transform_source(json_string)
1041 return json.loads(json_string, strict=False)
1042 except ValueError as ve:
1043 errmsg = '%s: Failed to parse JSON
' % video_id
1045 raise ExtractorError(errmsg, cause=ve)
1047 self.report_warning(errmsg + str(ve))
1049 def _parse_socket_response_as_json(self, data, video_id, transform_source=None, fatal=True):
1050 return self._parse_json(
1051 data[data.find('{'):data.rfind('}
') + 1],
1052 video_id, transform_source, fatal)
1054 def _download_socket_json_handle(
1055 self, url_or_request, video_id, note='Polling socket
',
1056 errnote='Unable to poll socket
', transform_source=None,
1057 fatal=True, encoding=None, data=None, headers={}, query={},
1058 expected_status=None):
1060 Return a tuple (JSON object, URL handle).
1062 See _download_webpage docstring for arguments specification.
1064 res = self._download_webpage_handle(
1065 url_or_request, video_id, note, errnote, fatal=fatal,
1066 encoding=encoding, data=data, headers=headers, query=query,
1067 expected_status=expected_status)
1071 return self._parse_socket_response_as_json(
1072 webpage, video_id, transform_source=transform_source,
1075 def _download_socket_json(
1076 self, url_or_request, video_id, note='Polling socket
',
1077 errnote='Unable to poll socket
', transform_source=None,
1078 fatal=True, encoding=None, data=None, headers={}, query={},
1079 expected_status=None):
1081 Return the JSON object as a dict.
1083 See _download_webpage docstring for arguments specification.
1085 res = self._download_socket_json_handle(
1086 url_or_request, video_id, note=note, errnote=errnote,
1087 transform_source=transform_source, fatal=fatal, encoding=encoding,
1088 data=data, headers=headers, query=query,
1089 expected_status=expected_status)
1090 return res if res is False else res[0]
1092 def report_warning(self, msg, video_id=None, *args, only_once=False, **kwargs):
1093 idstr = format_field(video_id, template='%s: ')
1094 msg = f'[{self.IE_NAME}
] {idstr}{msg}
'
1096 if f'WARNING
: {msg}
' in self._printed_messages:
1098 self._printed_messages.add(f'WARNING
: {msg}
')
1099 self._downloader.report_warning(msg, *args, **kwargs)
1101 def to_screen(self, msg, *args, **kwargs):
1102 """Print msg to screen, prefixing it with '[ie_name
]'"""
1103 self._downloader.to_screen(f'[{self.IE_NAME}
] {msg}
', *args, **kwargs)
1105 def write_debug(self, msg, *args, **kwargs):
1106 self._downloader.write_debug(f'[{self.IE_NAME}
] {msg}
', *args, **kwargs)
1108 def get_param(self, name, default=None, *args, **kwargs):
1109 if self._downloader:
1110 return self._downloader.params.get(name, default, *args, **kwargs)
1113 def report_drm(self, video_id, partial=False):
1114 self.raise_no_formats('This video
is DRM protected
', expected=True, video_id=video_id)
1116 def report_extraction(self, id_or_name):
1117 """Report information extraction."""
1118 self.to_screen('%s: Extracting information
' % id_or_name)
1120 def report_download_webpage(self, video_id):
1121 """Report webpage download."""
1122 self.to_screen('%s: Downloading webpage
' % video_id)
1124 def report_age_confirmation(self):
1125 """Report attempt to confirm age."""
1126 self.to_screen('Confirming age
')
1128 def report_login(self):
1129 """Report attempt to log in."""
1130 self.to_screen('Logging
in')
1132 def raise_login_required(
1133 self, msg='This video
is only available
for registered users
',
1134 metadata_available=False, method=NO_DEFAULT):
1135 if metadata_available and (
1136 self.get_param('ignore_no_formats_error
') or self.get_param('wait_for_video
')):
1137 self.report_warning(msg)
1139 msg += format_field(self._login_hint(method), template='. %s')
1140 raise ExtractorError(msg, expected=True)
1142 def raise_geo_restricted(
1143 self, msg='This video
is not available
from your location due to geo restriction
',
1144 countries=None, metadata_available=False):
1145 if metadata_available and (
1146 self.get_param('ignore_no_formats_error
') or self.get_param('wait_for_video
')):
1147 self.report_warning(msg)
1149 raise GeoRestrictedError(msg, countries=countries)
1151 def raise_no_formats(self, msg, expected=False, video_id=None):
1153 self.get_param('ignore_no_formats_error
') or self.get_param('wait_for_video
')):
1154 self.report_warning(msg, video_id)
1155 elif isinstance(msg, ExtractorError):
1158 raise ExtractorError(msg, expected=expected, video_id=video_id)
1160 # Methods for following #608
1162 def url_result(url, ie=None, video_id=None, video_title=None, *, url_transparent=False, **kwargs):
1163 """Returns a URL that points to a page that should be processed"""
1165 kwargs['ie_key
'] = ie if isinstance(ie, str) else ie.ie_key()
1166 if video_id is not None:
1167 kwargs['id'] = video_id
1168 if video_title is not None:
1169 kwargs['title
'] = video_title
1172 '_type
': 'url_transparent
' if url_transparent else 'url
',
1176 def playlist_from_matches(self, matches, playlist_id=None, playlist_title=None, getter=None, ie=None, video_kwargs=None, **kwargs):
1177 urls = (self.url_result(self._proto_relative_url(m), ie, **(video_kwargs or {}))
1178 for m in orderedSet(map(getter, matches) if getter else matches))
1179 return self.playlist_result(urls, playlist_id, playlist_title, **kwargs)
1182 def playlist_result(entries, playlist_id=None, playlist_title=None, playlist_description=None, *, multi_video=False, **kwargs):
1183 """Returns a playlist"""
1185 kwargs['id'] = playlist_id
1187 kwargs['title
'] = playlist_title
1188 if playlist_description is not None:
1189 kwargs['description
'] = playlist_description
1192 '_type
': 'multi_video
' if multi_video else 'playlist
',
1196 def _search_regex(self, pattern, string, name, default=NO_DEFAULT, fatal=True, flags=0, group=None):
1198 Perform a regex search on the given string, using a single or a list of
1199 patterns returning the first matching group.
1200 In case of failure return a default value or raise a WARNING or a
1201 RegexNotFoundError, depending on fatal, specifying the field name.
1205 elif isinstance(pattern, (str, re.Pattern)):
1206 mobj = re.search(pattern, string, flags)
1209 mobj = re.search(p, string, flags)
1213 _name = self._downloader._format_err(name, self._downloader.Styles.EMPHASIS)
1217 # return the first matching group
1218 return next(g for g in mobj.groups() if g is not None)
1219 elif isinstance(group, (list, tuple)):
1220 return tuple(mobj.group(g) for g in group)
1222 return mobj.group(group)
1223 elif default is not NO_DEFAULT:
1226 raise RegexNotFoundError('Unable to extract
%s' % _name)
1228 self.report_warning('unable to extract
%s' % _name + bug_reports_message())
1231 def _html_search_regex(self, pattern, string, name, default=NO_DEFAULT, fatal=True, flags=0, group=None):
1233 Like _search_regex, but strips HTML tags and unescapes entities.
1235 res = self._search_regex(pattern, string, name, default, fatal, flags, group)
1237 return clean_html(res).strip()
1241 def _get_netrc_login_info(self, netrc_machine=None):
1244 netrc_machine = netrc_machine or self._NETRC_MACHINE
1246 if self.get_param('usenetrc
', False):
1248 netrc_file = compat_expanduser(self.get_param('netrc_location
') or '~
')
1249 if os.path.isdir(netrc_file):
1250 netrc_file = os.path.join(netrc_file, '.netrc
')
1251 info = netrc.netrc(file=netrc_file).authenticators(netrc_machine)
1252 if info is not None:
1256 raise netrc.NetrcParseError(
1257 'No authenticators
for %s' % netrc_machine)
1258 except (OSError, netrc.NetrcParseError) as err:
1259 self.report_warning(
1260 'parsing
.netrc
: %s' % error_to_compat_str(err))
1262 return username, password
1264 def _get_login_info(self, username_option='username
', password_option='password
', netrc_machine=None):
1266 Get the login info as (username, password)
1267 First look for the manually specified credentials using username_option
1268 and password_option as keys in params dictionary. If no such credentials
1269 available look in the netrc file using the netrc_machine or _NETRC_MACHINE
1271 If there's no info available
, return (None, None)
1274 # Attempt to use provided username and password or .netrc data
1275 username = self.get_param(username_option)
1276 if username is not None:
1277 password = self.get_param(password_option)
1279 username, password = self._get_netrc_login_info(netrc_machine)
1281 return username, password
1283 def _get_tfa_info(self, note='two-factor verification code'):
1285 Get the two
-factor authentication info
1286 TODO
- asking the user will be required
for sms
/phone verify
1287 currently just uses the command line option
1288 If there
's no info available, return None
1291 tfa = self.get_param('twofactor
')
1295 return compat_getpass('Type
%s and press
[Return
]: ' % note)
1297 # Helper functions for extracting OpenGraph info
1299 def _og_regexes(prop):
1300 content_re = r'content
=(?
:"([^"]+?
)"|\'([^\']+?)\'|\s*([^\s"\'=<>`
]+?
))'
1301 property_re = (r'(?
:name|
property)=(?
:\'og
%(sep)s%(prop)s\'|
"og%(sep)s%(prop)s"|\s
*og
%(sep)s%(prop)s\b)'
1302 % {'prop': re.escape(prop), 'sep': '(?::|[:-])'})
1303 template = r'<meta
[^
>]+?
%s[^
>]+?
%s'
1305 template % (property_re, content_re),
1306 template % (content_re, property_re),
1310 def _meta_regex(prop):
1311 return r'''(?isx)<meta
1312 (?=[^>]+(?:itemprop|name|property|id|http-equiv)=(["\']?)%s\1)
1313 [^>]+?content=(["\'])(?P<content>.*?)\2''' % re.escape(prop)
1315 def _og_search_property(self, prop, html, name=None, **kargs):
1316 prop = variadic(prop)
1318 name = 'OpenGraph
%s' % prop[0]
1321 og_regexes.extend(self._og_regexes(p))
1322 escaped = self._search_regex(og_regexes, html, name, flags=re.DOTALL, **kargs)
1325 return unescapeHTML(escaped)
1327 def _og_search_thumbnail(self, html, **kargs):
1328 return self._og_search_property('image
', html, 'thumbnail URL
', fatal=False, **kargs)
1330 def _og_search_description(self, html, **kargs):
1331 return self._og_search_property('description
', html, fatal=False, **kargs)
1333 def _og_search_title(self, html, *, fatal=False, **kargs):
1334 return self._og_search_property('title
', html, fatal=fatal, **kargs)
1336 def _og_search_video_url(self, html, name='video url
', secure=True, **kargs):
1337 regexes = self._og_regexes('video
') + self._og_regexes('video
:url
')
1339 regexes = self._og_regexes('video
:secure_url
') + regexes
1340 return self._html_search_regex(regexes, html, name, **kargs)
1342 def _og_search_url(self, html, **kargs):
1343 return self._og_search_property('url
', html, **kargs)
1345 def _html_extract_title(self, html, name='title
', *, fatal=False, **kwargs):
1346 return self._html_search_regex(r'(?s
)<title
>([^
<]+)</title
>', html, name, fatal=fatal, **kwargs)
1348 def _html_search_meta(self, name, html, display_name=None, fatal=False, **kwargs):
1349 name = variadic(name)
1350 if display_name is None:
1351 display_name = name[0]
1352 return self._html_search_regex(
1353 [self._meta_regex(n) for n in name],
1354 html, display_name, fatal=fatal, group='content
', **kwargs)
1356 def _dc_search_uploader(self, html):
1357 return self._html_search_meta('dc
.creator
', html, 'uploader
')
1359 def _rta_search(self, html):
1360 # See http://www.rtalabel.org/index.php?content=howtofaq#single
1361 if re.search(r'(?ix
)<meta\s
+name
="rating"\s
+'
1362 r' content
="RTA-5042-1996-1400-1577-RTA"',
1367 def _media_rating_search(self, html):
1368 # See http://www.tjg-designs.com/WP/metadata-code-examples-adding-metadata-to-your-web-pages/
1369 rating = self._html_search_meta('rating
', html)
1381 return RATING_TABLE.get(rating.lower())
1383 def _family_friendly_search(self, html):
1384 # See http://schema.org/VideoObject
1385 family_friendly = self._html_search_meta(
1386 'isFamilyFriendly
', html, default=None)
1388 if not family_friendly:
1397 return RATING_TABLE.get(family_friendly.lower())
1399 def _twitter_search_player(self, html):
1400 return self._html_search_meta('twitter
:player
', html,
1401 'twitter card player
')
1403 def _search_json_ld(self, html, video_id, expected_type=None, **kwargs):
1404 json_ld_list = list(re.finditer(JSON_LD_RE, html))
1405 default = kwargs.get('default
', NO_DEFAULT)
1406 # JSON-LD may be malformed and thus `fatal` should be respected.
1407 # At the same time `default` may be passed that assumes `fatal=False`
1408 # for _search_regex. Let's simulate the same behavior here
as well
.
1409 fatal
= kwargs
.get('fatal', True) if default
is NO_DEFAULT
else False
1411 for mobj
in json_ld_list
:
1412 json_ld_item
= self
._parse
_json
(
1413 mobj
.group('json_ld'), video_id
, fatal
=fatal
)
1414 if not json_ld_item
:
1416 if isinstance(json_ld_item
, dict):
1417 json_ld
.append(json_ld_item
)
1418 elif isinstance(json_ld_item
, (list, tuple)):
1419 json_ld
.extend(json_ld_item
)
1421 json_ld
= self
._json
_ld
(json_ld
, video_id
, fatal
=fatal
, expected_type
=expected_type
)
1424 if default
is not NO_DEFAULT
:
1427 raise RegexNotFoundError('Unable to extract JSON-LD')
1429 self
.report_warning('unable to extract JSON-LD %s' % bug_reports_message())
1432 def _json_ld(self
, json_ld
, video_id
, fatal
=True, expected_type
=None):
1433 if isinstance(json_ld
, compat_str
):
1434 json_ld
= self
._parse
_json
(json_ld
, video_id
, fatal
=fatal
)
1438 if not isinstance(json_ld
, (list, tuple, dict)):
1440 if isinstance(json_ld
, dict):
1443 INTERACTION_TYPE_MAP
= {
1444 'CommentAction': 'comment',
1445 'AgreeAction': 'like',
1446 'DisagreeAction': 'dislike',
1447 'LikeAction': 'like',
1448 'DislikeAction': 'dislike',
1449 'ListenAction': 'view',
1450 'WatchAction': 'view',
1451 'ViewAction': 'view',
1454 def extract_interaction_type(e
):
1455 interaction_type
= e
.get('interactionType')
1456 if isinstance(interaction_type
, dict):
1457 interaction_type
= interaction_type
.get('@type')
1458 return str_or_none(interaction_type
)
1460 def extract_interaction_statistic(e
):
1461 interaction_statistic
= e
.get('interactionStatistic')
1462 if isinstance(interaction_statistic
, dict):
1463 interaction_statistic
= [interaction_statistic
]
1464 if not isinstance(interaction_statistic
, list):
1466 for is_e
in interaction_statistic
:
1467 if not isinstance(is_e
, dict):
1469 if is_e
.get('@type') != 'InteractionCounter':
1471 interaction_type
= extract_interaction_type(is_e
)
1472 if not interaction_type
:
1474 # For interaction count some sites provide string instead of
1475 # an integer (as per spec) with non digit characters (e.g. ",")
1476 # so extracting count with more relaxed str_to_int
1477 interaction_count
= str_to_int(is_e
.get('userInteractionCount'))
1478 if interaction_count
is None:
1480 count_kind
= INTERACTION_TYPE_MAP
.get(interaction_type
.split('/')[-1])
1483 count_key
= '%s_count' % count_kind
1484 if info
.get(count_key
) is not None:
1486 info
[count_key
] = interaction_count
1488 def extract_chapter_information(e
):
1490 'title': part
.get('name'),
1491 'start_time': part
.get('startOffset'),
1492 'end_time': part
.get('endOffset'),
1493 } for part
in variadic(e
.get('hasPart') or []) if part
.get('@type') == 'Clip']
1494 for idx
, (last_c
, current_c
, next_c
) in enumerate(zip(
1495 [{'end_time': 0}
] + chapters
, chapters
, chapters
[1:])):
1496 current_c
['end_time'] = current_c
['end_time'] or next_c
['start_time']
1497 current_c
['start_time'] = current_c
['start_time'] or last_c
['end_time']
1498 if None in current_c
.values():
1499 self
.report_warning(f
'Chapter {idx} contains broken data. Not extracting chapters')
1502 chapters
[-1]['end_time'] = chapters
[-1]['end_time'] or info
['duration']
1503 info
['chapters'] = chapters
1505 def extract_video_object(e
):
1506 assert e
['@type'] == 'VideoObject'
1507 author
= e
.get('author')
1509 'url': url_or_none(e
.get('contentUrl')),
1510 'title': unescapeHTML(e
.get('name')),
1511 'description': unescapeHTML(e
.get('description')),
1512 'thumbnails': [{'url': url_or_none(url)}
1513 for url
in variadic(traverse_obj(e
, 'thumbnailUrl', 'thumbnailURL'))],
1514 'duration': parse_duration(e
.get('duration')),
1515 'timestamp': unified_timestamp(e
.get('uploadDate')),
1516 # author can be an instance of 'Organization' or 'Person' types.
1517 # both types can have 'name' property(inherited from 'Thing' type). [1]
1518 # however some websites are using 'Text' type instead.
1519 # 1. https://schema.org/VideoObject
1520 'uploader': author
.get('name') if isinstance(author
, dict) else author
if isinstance(author
, compat_str
) else None,
1521 'filesize': float_or_none(e
.get('contentSize')),
1522 'tbr': int_or_none(e
.get('bitrate')),
1523 'width': int_or_none(e
.get('width')),
1524 'height': int_or_none(e
.get('height')),
1525 'view_count': int_or_none(e
.get('interactionCount')),
1527 extract_interaction_statistic(e
)
1528 extract_chapter_information(e
)
1530 def traverse_json_ld(json_ld
, at_top_level
=True):
1532 if at_top_level
and '@context' not in e
:
1534 if at_top_level
and set(e
.keys()) == {'@context', '@graph'}
:
1535 traverse_json_ld(variadic(e
['@graph'], allowed_types
=(dict,)), at_top_level
=False)
1537 item_type
= e
.get('@type')
1538 if expected_type
is not None and expected_type
!= item_type
:
1540 rating
= traverse_obj(e
, ('aggregateRating', 'ratingValue'), expected_type
=float_or_none
)
1541 if rating
is not None:
1542 info
['average_rating'] = rating
1543 if item_type
in ('TVEpisode', 'Episode'):
1544 episode_name
= unescapeHTML(e
.get('name'))
1546 'episode': episode_name
,
1547 'episode_number': int_or_none(e
.get('episodeNumber')),
1548 'description': unescapeHTML(e
.get('description')),
1550 if not info
.get('title') and episode_name
:
1551 info
['title'] = episode_name
1552 part_of_season
= e
.get('partOfSeason')
1553 if isinstance(part_of_season
, dict) and part_of_season
.get('@type') in ('TVSeason', 'Season', 'CreativeWorkSeason'):
1555 'season': unescapeHTML(part_of_season
.get('name')),
1556 'season_number': int_or_none(part_of_season
.get('seasonNumber')),
1558 part_of_series
= e
.get('partOfSeries') or e
.get('partOfTVSeries')
1559 if isinstance(part_of_series
, dict) and part_of_series
.get('@type') in ('TVSeries', 'Series', 'CreativeWorkSeries'):
1560 info
['series'] = unescapeHTML(part_of_series
.get('name'))
1561 elif item_type
== 'Movie':
1563 'title': unescapeHTML(e
.get('name')),
1564 'description': unescapeHTML(e
.get('description')),
1565 'duration': parse_duration(e
.get('duration')),
1566 'timestamp': unified_timestamp(e
.get('dateCreated')),
1568 elif item_type
in ('Article', 'NewsArticle'):
1570 'timestamp': parse_iso8601(e
.get('datePublished')),
1571 'title': unescapeHTML(e
.get('headline')),
1572 'description': unescapeHTML(e
.get('articleBody') or e
.get('description')),
1574 if traverse_obj(e
, ('video', 0, '@type')) == 'VideoObject':
1575 extract_video_object(e
['video'][0])
1576 elif item_type
== 'VideoObject':
1577 extract_video_object(e
)
1578 if expected_type
is None:
1582 video
= e
.get('video')
1583 if isinstance(video
, dict) and video
.get('@type') == 'VideoObject':
1584 extract_video_object(video
)
1585 if expected_type
is None:
1589 traverse_json_ld(json_ld
)
1591 return filter_dict(info
)
1593 def _search_nextjs_data(self
, webpage
, video_id
, *, transform_source
=None, fatal
=True, **kw
):
1594 return self
._parse
_json
(
1596 r
'(?s)<script[^>]+id=[\'"]__NEXT_DATA__[\'"][^
>]*>([^
<]+)</script
>',
1597 webpage, 'next
.js data
', fatal=fatal, **kw),
1598 video_id, transform_source=transform_source, fatal=fatal)
1600 def _search_nuxt_data(self, webpage, video_id, context_name='__NUXT__
'):
1601 ''' Parses Nuxt.js metadata. This works as long as the function __NUXT__ invokes is a pure function. '''
1602 # not all website do this, but it can be changed
1603 # https://stackoverflow.com/questions/67463109/how-to-change-or-hide-nuxt-and-nuxt-keyword-in-page-source
1604 rectx = re.escape(context_name)
1605 js, arg_keys, arg_vals = self._search_regex(
1606 (r'<script
>window\
.%s=\
(function\
((?P
<arg_keys
>.*?
)\
)\{return\s(?P<js>\{.*?\}
)\
}\
((?P
<arg_vals
>.+?
)\
)\
);?
</script
>' % rectx,
1607 r'%s\
(.*?\
(function\
((?P
<arg_keys
>.*?
)\
)\{return\s(?P<js>\{.*?\}
)\
}\
((?P
<arg_vals
>.*?
)\
)' % rectx),
1608 webpage, context_name, group=['js
', 'arg_keys
', 'arg_vals
'])
1610 args = dict(zip(arg_keys.split(','), arg_vals.split(',')))
1612 for key, val in args.items():
1613 if val in ('undefined
', 'void
0'):
1616 return self._parse_json(js_to_json(js, args), video_id)['data
'][0]
1619 def _hidden_inputs(html):
1620 html = re.sub(r'<!--(?
:(?
!<!--).)*-->', '', html)
1622 for input in re.findall(r'(?i
)(<input[^
>]+>)', html):
1623 attrs = extract_attributes(input)
1626 if attrs.get('type') not in ('hidden
', 'submit
'):
1628 name = attrs.get('name
') or attrs.get('id')
1629 value = attrs.get('value
')
1630 if name and value is not None:
1631 hidden_inputs[name] = value
1632 return hidden_inputs
1634 def _form_hidden_inputs(self, form_id, html):
1635 form = self._search_regex(
1636 r'(?
is)<form
[^
>]+?
id=(["\'])%s\1[^>]*>(?P<form>.+?)</form>' % form_id,
1637 html, '%s form' % form_id, group='form')
1638 return self._hidden_inputs(form)
1641 regex = r' *((?P<reverse>\+)?(?P<field>[a-zA-Z0-9_]+)((?P<separator>[~:])(?P<limit>.*?))?)? *$'
1643 default = ('hidden', 'aud_or_vid', 'hasvid', 'ie_pref', 'lang', 'quality',
1644 'res', 'fps', 'hdr:12', 'codec:vp9.2', 'size', 'br', 'asr',
1645 'proto', 'ext', 'hasaud', 'source', 'id') # These must not be aliases
1646 ytdl_default = ('hasaud', 'lang', 'quality', 'tbr', 'filesize', 'vbr',
1647 'height', 'width', 'proto', 'vext', 'abr', 'aext',
1648 'fps', 'fs_approx', 'source', 'id')
1651 'vcodec': {'type': 'ordered', 'regex': True,
1652 'order': ['av0?1', 'vp0?9.2', 'vp0?9', '[hx]265|he?vc?', '[hx]264|avc', 'vp0?8', 'mp4v|h263', 'theora', '', None, 'none']},
1653 'acodec': {'type': 'ordered', 'regex': True,
1654 'order': ['[af]lac', 'wav|aiff', 'opus', 'vorbis|ogg', 'aac', 'mp?4a?', 'mp3', 'e-?a?c-?3', 'ac-?3', 'dts', '', None, 'none']},
1655 'hdr': {'type': 'ordered', 'regex': True, 'field': 'dynamic_range',
1656 'order': ['dv', '(hdr)?12', r'(hdr)?10\+', '(hdr)?10', 'hlg', '', 'sdr', None]},
1657 'proto': {'type': 'ordered', 'regex': True, 'field': 'protocol',
1658 'order': ['(ht|f)tps', '(ht|f)tp$', 'm3u8.*', '.*dash', 'websocket_frag', 'rtmpe?', '', 'mms|rtsp', 'ws|websocket', 'f4']},
1659 'vext': {'type': 'ordered', 'field': 'video_ext',
1660 'order': ('mp4', 'webm', 'flv', '', 'none'),
1661 'order_free': ('webm', 'mp4', 'flv', '', 'none')},
1662 'aext': {'type': 'ordered', 'field': 'audio_ext',
1663 'order': ('m4a', 'aac', 'mp3', 'ogg', 'opus', 'webm', '', 'none'),
1664 'order_free': ('opus', 'ogg', 'webm', 'm4a', 'mp3', 'aac', '', 'none')},
1665 'hidden': {'visible': False, 'forced': True, 'type': 'extractor', 'max': -1000},
1666 'aud_or_vid': {'visible': False, 'forced': True, 'type': 'multiple',
1667 'field': ('vcodec', 'acodec'),
1668 'function': lambda it: int(any(v != 'none' for v in it))},
1669 'ie_pref': {'priority': True, 'type': 'extractor'},
1670 'hasvid': {'priority': True, 'field': 'vcodec', 'type': 'boolean', 'not_in_list': ('none',)},
1671 'hasaud': {'field': 'acodec', 'type': 'boolean', 'not_in_list': ('none',)},
1672 'lang': {'convert': 'float', 'field': 'language_preference', 'default': -1},
1673 'quality': {'convert': 'float', 'default': -1},
1674 'filesize': {'convert': 'bytes'},
1675 'fs_approx': {'convert': 'bytes', 'field': 'filesize_approx'},
1676 'id': {'convert': 'string', 'field': 'format_id'},
1677 'height': {'convert': 'float_none'},
1678 'width': {'convert': 'float_none'},
1679 'fps': {'convert': 'float_none'},
1680 'tbr': {'convert': 'float_none'},
1681 'vbr': {'convert': 'float_none'},
1682 'abr': {'convert': 'float_none'},
1683 'asr': {'convert': 'float_none'},
1684 'source': {'convert': 'float', 'field': 'source_preference', 'default': -1},
1686 'codec': {'type': 'combined', 'field': ('vcodec', 'acodec')},
1687 'br': {'type': 'combined', 'field': ('tbr', 'vbr', 'abr'), 'same_limit': True},
1688 'size': {'type': 'combined', 'same_limit': True, 'field': ('filesize', 'fs_approx')},
1689 'ext': {'type': 'combined', 'field': ('vext', 'aext')},
1690 'res': {'type': 'multiple', 'field': ('height', 'width'),
1691 'function': lambda it: (lambda l: min(l) if l else 0)(tuple(filter(None, it)))},
1693 # For compatibility with youtube-dl
1694 'format_id': {'type': 'alias', 'field': 'id'},
1695 'preference': {'type': 'alias', 'field': 'ie_pref'},
1696 'language_preference': {'type': 'alias', 'field': 'lang'},
1697 'source_preference': {'type': 'alias', 'field': 'source'},
1698 'protocol': {'type': 'alias', 'field': 'proto'},
1699 'filesize_approx': {'type': 'alias', 'field': 'fs_approx'},
1702 'dimension': {'type': 'alias', 'field': 'res', 'deprecated': True},
1703 'resolution': {'type': 'alias', 'field': 'res', 'deprecated': True},
1704 'extension': {'type': 'alias', 'field': 'ext', 'deprecated': True},
1705 'bitrate': {'type': 'alias', 'field': 'br', 'deprecated': True},
1706 'total_bitrate': {'type': 'alias', 'field': 'tbr', 'deprecated': True},
1707 'video_bitrate': {'type': 'alias', 'field': 'vbr', 'deprecated': True},
1708 'audio_bitrate': {'type': 'alias', 'field': 'abr', 'deprecated': True},
1709 'framerate': {'type': 'alias', 'field': 'fps', 'deprecated': True},
1710 'filesize_estimate': {'type': 'alias', 'field': 'size', 'deprecated': True},
1711 'samplerate': {'type': 'alias', 'field': 'asr', 'deprecated': True},
1712 'video_ext': {'type': 'alias', 'field': 'vext', 'deprecated': True},
1713 'audio_ext': {'type': 'alias', 'field': 'aext', 'deprecated': True},
1714 'video_codec': {'type': 'alias', 'field': 'vcodec', 'deprecated': True},
1715 'audio_codec': {'type': 'alias', 'field': 'acodec', 'deprecated': True},
1716 'video': {'type': 'alias', 'field': 'hasvid', 'deprecated': True},
1717 'has_video': {'type': 'alias', 'field': 'hasvid', 'deprecated': True},
1718 'audio': {'type': 'alias', 'field': 'hasaud', 'deprecated': True},
1719 'has_audio': {'type': 'alias', 'field': 'hasaud', 'deprecated': True},
1720 'extractor': {'type': 'alias', 'field': 'ie_pref', 'deprecated': True},
1721 'extractor_preference': {'type': 'alias', 'field': 'ie_pref', 'deprecated': True},
1724 def __init__(self, ie, field_preference):
1726 self.ydl = ie._downloader
1727 self.evaluate_params(self.ydl.params, field_preference)
1728 if ie.get_param('verbose'):
1729 self.print_verbose_info(self.ydl.write_debug)
1731 def _get_field_setting(self, field, key):
1732 if field not in self.settings:
1733 if key in ('forced', 'priority'):
1735 self.ydl.deprecation_warning(
1736 f'Using arbitrary fields ({field}) for format sorting is deprecated '
1737 'and may be removed in a future version')
1738 self.settings[field] = {}
1739 propObj = self.settings[field]
1740 if key not in propObj:
1741 type = propObj.get('type')
1743 default = 'preference' if type == 'extractor' else (field,) if type in ('combined', 'multiple') else field
1744 elif key == 'convert':
1745 default = 'order' if type == 'ordered' else 'float_string' if field else 'ignore'
1747 default = {'type': 'field', 'visible': True, 'order': [], 'not_in_list': (None,)}.get(key, None)
1748 propObj[key] = default
1751 def _resolve_field_value(self, field, value, convertNone=False):
1756 value = value.lower()
1757 conversion = self._get_field_setting(field, 'convert')
1758 if conversion == 'ignore':
1760 if conversion == 'string':
1762 elif conversion == 'float_none':
1763 return float_or_none(value)
1764 elif conversion == 'bytes':
1765 return FileDownloader.parse_bytes(value)
1766 elif conversion == 'order':
1767 order_list = (self._use_free_order and self._get_field_setting(field, 'order_free')) or self._get_field_setting(field, 'order')
1768 use_regex = self._get_field_setting(field, 'regex')
1769 list_length = len(order_list)
1770 empty_pos = order_list.index('') if '' in order_list else list_length + 1
1771 if use_regex and value is not None:
1772 for i, regex in enumerate(order_list):
1773 if regex and re.match(regex, value):
1774 return list_length - i
1775 return list_length - empty_pos # not in list
1776 else: # not regex or value = None
1777 return list_length - (order_list.index(value) if value in order_list else empty_pos)
1779 if value.isnumeric():
1782 self.settings[field]['convert'] = 'string'
1785 def evaluate_params(self, params, sort_extractor):
1786 self._use_free_order = params.get('prefer_free_formats', False)
1787 self._sort_user = params.get('format_sort', [])
1788 self._sort_extractor = sort_extractor
1790 def add_item(field, reverse, closest, limit_text):
1791 field = field.lower()
1792 if field in self._order:
1794 self._order.append(field)
1795 limit = self._resolve_field_value(field, limit_text)
1798 'closest': False if limit is None else closest,
1799 'limit_text': limit_text,
1801 if field in self.settings:
1802 self.settings[field].update(data)
1804 self.settings[field] = data
1807 tuple(field for field in self.default if self._get_field_setting(field, 'forced'))
1808 + (tuple() if params.get('format_sort_force', False)
1809 else tuple(field for field in self.default if self._get_field_setting(field, 'priority')))
1810 + tuple(self._sort_user) + tuple(sort_extractor) + self.default)
1812 for item in sort_list:
1813 match = re.match(self.regex, item)
1815 raise ExtractorError('Invalid format sort string "%s" given by extractor' % item)
1816 field = match.group('field')
1819 if self._get_field_setting(field, 'type') == 'alias':
1820 alias, field = field, self._get_field_setting(field, 'field')
1821 if self._get_field_setting(alias, 'deprecated'):
1822 self.ydl.deprecation_warning(
1823 f'Format sorting alias {alias} is deprecated '
1824 f'and may be removed in a future version. Please use {field} instead')
1825 reverse = match.group('reverse') is not None
1826 closest = match.group('separator') == '~'
1827 limit_text = match.group('limit')
1829 has_limit = limit_text is not None
1830 has_multiple_fields = self._get_field_setting(field, 'type') == 'combined'
1831 has_multiple_limits = has_limit and has_multiple_fields and not self._get_field_setting(field, 'same_limit')
1833 fields = self._get_field_setting(field, 'field') if has_multiple_fields else (field,)
1834 limits = limit_text.split(':') if has_multiple_limits else (limit_text,) if has_limit else tuple()
1835 limit_count = len(limits)
1836 for (i, f) in enumerate(fields):
1837 add_item(f, reverse, closest,
1838 limits[i] if i < limit_count
1839 else limits[0] if has_limit and not has_multiple_limits
1842 def print_verbose_info(self, write_debug):
1844 write_debug('Sort order given by user: %s' % ', '.join(self._sort_user))
1845 if self._sort_extractor:
1846 write_debug('Sort order given by extractor: %s' % ', '.join(self._sort_extractor))
1847 write_debug('Formats sorted by: %s' % ', '.join(['%s%s%s' % (
1848 '+' if self._get_field_setting(field, 'reverse') else '', field,
1849 '%s%s(%s)' % ('~' if self._get_field_setting(field, 'closest') else ':',
1850 self._get_field_setting(field, 'limit_text'),
1851 self._get_field_setting(field, 'limit'))
1852 if self._get_field_setting(field, 'limit_text') is not None else '')
1853 for field in self._order if self._get_field_setting(field, 'visible')]))
1855 def _calculate_field_preference_from_value(self, format, field, type, value):
1856 reverse = self._get_field_setting(field, 'reverse')
1857 closest = self._get_field_setting(field, 'closest')
1858 limit = self._get_field_setting(field, 'limit')
1860 if type == 'extractor':
1861 maximum = self._get_field_setting(field, 'max')
1862 if value is None or (maximum is not None and value >= maximum):
1864 elif type == 'boolean':
1865 in_list = self._get_field_setting(field, 'in_list')
1866 not_in_list = self._get_field_setting(field, 'not_in_list')
1867 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
1868 elif type == 'ordered':
1869 value = self._resolve_field_value(field, value, True)
1871 # try to convert to number
1872 val_num = float_or_none(value, default=self._get_field_setting(field, 'default'))
1873 is_num = self._get_field_setting(field, 'convert') != 'string' and val_num is not None
1877 return ((-10, 0) if value is None
1878 else (1, value, 0) if not is_num # if a field has mixed strings and numbers, strings are sorted higher
1879 else (0, -abs(value - limit), value - limit if reverse else limit - value) if closest
1880 else (0, value, 0) if not reverse and (limit is None or value <= limit)
1881 else (0, -value, 0) if limit is None or (reverse and value == limit) or value > limit
1882 else (-1, value, 0))
1884 def _calculate_field_preference(self, format, field):
1885 type = self._get_field_setting(field, 'type') # extractor, boolean, ordered, field, multiple
1886 get_value = lambda f: format.get(self._get_field_setting(f, 'field'))
1887 if type == 'multiple':
1888 type = 'field' # Only 'field' is allowed in multiple for now
1889 actual_fields = self._get_field_setting(field, 'field')
1891 value = self._get_field_setting(field, 'function')(get_value(f) for f in actual_fields)
1893 value = get_value(field)
1894 return self._calculate_field_preference_from_value(format, field, type, value)
1896 def calculate_preference(self, format):
1897 # Determine missing protocol
1898 if not format.get('protocol'):
1899 format['protocol'] = determine_protocol(format)
1901 # Determine missing ext
1902 if not format.get('ext') and 'url' in format:
1903 format['ext'] = determine_ext(format['url'])
1904 if format.get('vcodec') == 'none':
1905 format['audio_ext'] = format['ext'] if format.get('acodec') != 'none' else 'none'
1906 format['video_ext'] = 'none'
1908 format['video_ext'] = format['ext']
1909 format['audio_ext'] = 'none'
1910 # if format.get('preference') is None and format.get('ext') in ('f4f', 'f4m'): # Not supported?
1911 # format['preference'] = -1000
1913 # Determine missing bitrates
1914 if format.get('tbr') is None:
1915 if format.get('vbr') is not None and format.get('abr') is not None:
1916 format['tbr'] = format.get('vbr', 0) + format.get('abr', 0)
1918 if format.get('vcodec') != 'none' and format.get('vbr') is None:
1919 format['vbr'] = format.get('tbr') - format.get('abr', 0)
1920 if format.get('acodec') != 'none' and format.get('abr') is None:
1921 format['abr'] = format.get('tbr') - format.get('vbr', 0)
1923 return tuple(self._calculate_field_preference(format, field) for field in self._order)
1925 def _sort_formats(self, formats, field_preference=[]):
1928 formats.sort(key=self.FormatSort(self, field_preference).calculate_preference)
1930 def _check_formats(self, formats, video_id):
1932 formats[:] = filter(
1933 lambda f: self._is_valid_url(
1935 item='%s video format' % f.get('format_id') if f.get('format_id') else 'video'),
1939 def _remove_duplicate_formats(formats):
1943 if f['url'] not in format_urls:
1944 format_urls.add(f['url'])
1945 unique_formats.append(f)
1946 formats[:] = unique_formats
1948 def _is_valid_url(self, url, video_id, item='video', headers={}):
1949 url = self._proto_relative_url(url, scheme='http:')
1950 # For now assume non HTTP(S) URLs always valid
1951 if not (url.startswith('http://') or url.startswith('https://')):
1954 self._request_webpage(url, video_id, 'Checking %s URL' % item, headers=headers)
1956 except ExtractorError as e:
1958 '%s: %s URL is invalid, skipping: %s'
1959 % (video_id, item, error_to_compat_str(e.cause)))
1962 def http_scheme(self):
1963 """ Either "http
:" or "https
:", depending on the user's preferences """
1966 if self.get_param('prefer_insecure', False)
1969 def _proto_relative_url(self, url, scheme=None):
1972 if url.startswith('//'):
1974 scheme = self.http_scheme()
1979 def _sleep(self, timeout, video_id, msg_template=None):
1980 if msg_template is None:
1981 msg_template = '%(video_id)s: Waiting for %(timeout)s seconds'
1982 msg = msg_template % {'video_id': video_id, 'timeout': timeout}
1986 def _extract_f4m_formats(self, manifest_url, video_id, preference=None, quality=None, f4m_id=None,
1987 transform_source=lambda s: fix_xml_ampersands(s).strip(),
1988 fatal=True, m3u8_id=None, data=None, headers={}, query={}):
1989 res = self._download_xml_handle(
1990 manifest_url, video_id, 'Downloading f4m manifest',
1991 'Unable to download f4m manifest',
1992 # Some manifests may be malformed, e.g. prosiebensat1 generated manifests
1993 # (see https://github.com/ytdl-org/youtube-dl/issues/6215#issuecomment-121704244)
1994 transform_source=transform_source,
1995 fatal=fatal, data=data, headers=headers, query=query)
1999 manifest, urlh = res
2000 manifest_url = urlh.geturl()
2002 return self._parse_f4m_formats(
2003 manifest, manifest_url, video_id, preference=preference, quality=quality, f4m_id=f4m_id,
2004 transform_source=transform_source, fatal=fatal, m3u8_id=m3u8_id)
2006 def _parse_f4m_formats(self, manifest, manifest_url, video_id, preference=None, quality=None, f4m_id=None,
2007 transform_source=lambda s: fix_xml_ampersands(s).strip(),
2008 fatal=True, m3u8_id=None):
2009 if not isinstance(manifest, xml.etree.ElementTree.Element) and not fatal:
2012 # currently yt-dlp cannot decode the playerVerificationChallenge as Akamai uses Adobe Alchemy
2013 akamai_pv = manifest.find('{http://ns.adobe.com/f4m/1.0}pv-2.0')
2014 if akamai_pv is not None and ';' in akamai_pv.text:
2015 playerVerificationChallenge = akamai_pv.text.split(';')[0]
2016 if playerVerificationChallenge.strip() != '':
2020 manifest_version = '1.0'
2021 media_nodes = manifest.findall('{http://ns.adobe.com/f4m/1.0}media')
2023 manifest_version = '2.0'
2024 media_nodes = manifest.findall('{http://ns.adobe.com/f4m/2.0}media')
2025 # Remove unsupported DRM protected media from final formats
2026 # rendition (see https://github.com/ytdl-org/youtube-dl/issues/8573).
2027 media_nodes = remove_encrypted_media(media_nodes)
2031 manifest_base_url = get_base_url(manifest)
2033 bootstrap_info = xpath_element(
2034 manifest, ['{http://ns.adobe.com/f4m/1.0}bootstrapInfo', '{http://ns.adobe.com/f4m/2.0}bootstrapInfo'],
2035 'bootstrap info', default=None)
2038 mime_type = xpath_text(
2039 manifest, ['{http://ns.adobe.com/f4m/1.0}mimeType', '{http://ns.adobe.com/f4m/2.0}mimeType'],
2040 'base URL', default=None)
2041 if mime_type and mime_type.startswith('audio/'):
2044 for i, media_el in enumerate(media_nodes):
2045 tbr = int_or_none(media_el.attrib.get('bitrate'))
2046 width = int_or_none(media_el.attrib.get('width'))
2047 height = int_or_none(media_el.attrib.get('height'))
2048 format_id = join_nonempty(f4m_id, tbr or i)
2049 # If <bootstrapInfo> is present, the specified f4m is a
2050 # stream-level manifest, and only set-level manifests may refer to
2051 # external resources. See section 11.4 and section 4 of F4M spec
2052 if bootstrap_info is None:
2054 # @href is introduced in 2.0, see section 11.6 of F4M spec
2055 if manifest_version == '2.0':
2056 media_url = media_el.attrib.get('href')
2057 if media_url is None:
2058 media_url = media_el.attrib.get('url')
2062 media_url if media_url.startswith('http://') or media_url.startswith('https://')
2063 else ((manifest_base_url or '/'.join(manifest_url.split('/')[:-1])) + '/' + media_url))
2064 # If media_url is itself a f4m manifest do the recursive extraction
2065 # since bitrates in parent manifest (this one) and media_url manifest
2066 # may differ leading to inability to resolve the format by requested
2067 # bitrate in f4m downloader
2068 ext = determine_ext(manifest_url)
2070 f4m_formats = self._extract_f4m_formats(
2071 manifest_url, video_id, preference=preference, quality=quality, f4m_id=f4m_id,
2072 transform_source=transform_source, fatal=fatal)
2073 # Sometimes stream-level manifest contains single media entry that
2074 # does not contain any quality metadata (e.g. http://matchtv.ru/#live-player).
2075 # At the same time parent's media entry in set-level manifest may
2076 # contain it. We will copy it from parent in such cases.
2077 if len(f4m_formats) == 1:
2080 'tbr': f.get('tbr') or tbr,
2081 'width': f.get('width') or width,
2082 'height': f.get('height') or height,
2083 'format_id': f.get('format_id') if not tbr else format_id,
2086 formats.extend(f4m_formats)
2089 formats.extend(self._extract_m3u8_formats(
2090 manifest_url, video_id, 'mp4', preference=preference,
2091 quality=quality, m3u8_id=m3u8_id, fatal=fatal))
2094 'format_id': format_id,
2095 'url': manifest_url,
2096 'manifest_url': manifest_url,
2097 'ext': 'flv' if bootstrap_info is not None else None,
2103 'preference': preference,
2108 def _m3u8_meta_format(self, m3u8_url, ext=None, preference=None, quality=None, m3u8_id=None):
2110 'format_id': join_nonempty(m3u8_id, 'meta'),
2114 'preference': preference - 100 if preference else -100,
2116 'resolution': 'multiple',
2117 'format_note': 'Quality selection URL',
2120 def _report_ignoring_subs(self, name):
2121 self.report_warning(bug_reports_message(
2122 f'Ignoring subtitle tracks found in the {name} manifest; '
2123 'if any subtitle tracks are missing,'
2126 def _extract_m3u8_formats(self, *args, **kwargs):
2127 fmts, subs = self._extract_m3u8_formats_and_subtitles(*args, **kwargs)
2129 self._report_ignoring_subs('HLS')
2132 def _extract_m3u8_formats_and_subtitles(
2133 self, m3u8_url, video_id, ext=None, entry_protocol='m3u8_native',
2134 preference=None, quality=None, m3u8_id=None, note=None,
2135 errnote=None, fatal=True, live=False, data=None, headers={},
2138 res = self._download_webpage_handle(
2140 note='Downloading m3u8 information' if note is None else note,
2141 errnote='Failed to download m3u8 information' if errnote is None else errnote,
2142 fatal=fatal, data=data, headers=headers, query=query)
2147 m3u8_doc, urlh = res
2148 m3u8_url = urlh.geturl()
2150 return self._parse_m3u8_formats_and_subtitles(
2151 m3u8_doc, m3u8_url, ext=ext, entry_protocol=entry_protocol,
2152 preference=preference, quality=quality, m3u8_id=m3u8_id,
2153 note=note, errnote=errnote, fatal=fatal, live=live, data=data,
2154 headers=headers, query=query, video_id=video_id)
2156 def _parse_m3u8_formats_and_subtitles(
2157 self, m3u8_doc, m3u8_url=None, ext=None, entry_protocol='m3u8_native',
2158 preference=None, quality=None, m3u8_id=None, live=False, note=None,
2159 errnote=None, fatal=True, data=None, headers={}, query={},
2161 formats, subtitles = [], {}
2163 has_drm = re.search('|'.join([
2164 r'#EXT-X-FAXS-CM:', # Adobe Flash Access
2165 r'#EXT-X-(?:SESSION-)?KEY:.*?URI="skd
://', # Apple FairPlay
2168 def format_url(url):
2169 return url if re.match(r'^https?
://', url) else compat_urlparse.urljoin(m3u8_url, url)
2171 if self.get_param('hls_split_discontinuity
', False):
2172 def _extract_m3u8_playlist_indices(manifest_url=None, m3u8_doc=None):
2174 if not manifest_url:
2176 m3u8_doc = self._download_webpage(
2177 manifest_url, video_id, fatal=fatal, data=data, headers=headers,
2178 note=False, errnote='Failed to download m3u8 playlist information
')
2179 if m3u8_doc is False:
2181 return range(1 + sum(line.startswith('#EXT-X-DISCONTINUITY') for line in m3u8_doc.splitlines()))
2184 def _extract_m3u8_playlist_indices(*args
, **kwargs
):
2188 # 1. https://tools.ietf.org/html/draft-pantos-http-live-streaming-21
2189 # 2. https://github.com/ytdl-org/youtube-dl/issues/12211
2190 # 3. https://github.com/ytdl-org/youtube-dl/issues/18923
2192 # We should try extracting formats only from master playlists [1, 4.3.4],
2193 # i.e. playlists that describe available qualities. On the other hand
2194 # media playlists [1, 4.3.3] should be returned as is since they contain
2195 # just the media without qualities renditions.
2196 # Fortunately, master playlist can be easily distinguished from media
2197 # playlist based on particular tags availability. As of [1, 4.3.3, 4.3.4]
2198 # master playlist tags MUST NOT appear in a media playlist and vice versa.
2199 # As of [1, 4.3.3.1] #EXT-X-TARGETDURATION tag is REQUIRED for every
2200 # media playlist and MUST NOT appear in master playlist thus we can
2201 # clearly detect media playlist with this criterion.
2203 if '#EXT-X-TARGETDURATION' in m3u8_doc
: # media playlist, return as is
2205 'format_id': join_nonempty(m3u8_id
, idx
),
2206 'format_index': idx
,
2207 'url': m3u8_url
or encode_data_uri(m3u8_doc
.encode('utf-8'), 'application/x-mpegurl'),
2209 'protocol': entry_protocol
,
2210 'preference': preference
,
2213 } for idx
in _extract_m3u8_playlist_indices(m3u8_doc
=m3u8_doc
)]
2215 return formats
, subtitles
2218 last_stream_inf
= {}
2220 def extract_media(x_media_line
):
2221 media
= parse_m3u8_attributes(x_media_line
)
2222 # As per [1, 4.3.4.1] TYPE, GROUP-ID and NAME are REQUIRED
2223 media_type
, group_id
, name
= media
.get('TYPE'), media
.get('GROUP-ID'), media
.get('NAME')
2224 if not (media_type
and group_id
and name
):
2226 groups
.setdefault(group_id
, []).append(media
)
2227 # <https://tools.ietf.org/html/rfc8216#section-4.3.4.1>
2228 if media_type
== 'SUBTITLES':
2229 # According to RFC 8216 §4.3.4.2.1, URI is REQUIRED in the
2230 # EXT-X-MEDIA tag if the media type is SUBTITLES.
2231 # However, lack of URI has been spotted in the wild.
2232 # e.g. NebulaIE; see https://github.com/yt-dlp/yt-dlp/issues/339
2233 if not media
.get('URI'):
2235 url
= format_url(media
['URI'])
2238 'ext': determine_ext(url
),
2240 if sub_info
['ext'] == 'm3u8':
2241 # Per RFC 8216 §3.1, the only possible subtitle format m3u8
2242 # files may contain is WebVTT:
2243 # <https://tools.ietf.org/html/rfc8216#section-3.1>
2244 sub_info
['ext'] = 'vtt'
2245 sub_info
['protocol'] = 'm3u8_native'
2246 lang
= media
.get('LANGUAGE') or 'und'
2247 subtitles
.setdefault(lang
, []).append(sub_info
)
2248 if media_type
not in ('VIDEO', 'AUDIO'):
2250 media_url
= media
.get('URI')
2252 manifest_url
= format_url(media_url
)
2254 'format_id': join_nonempty(m3u8_id
, group_id
, name
, idx
),
2255 'format_note': name
,
2256 'format_index': idx
,
2257 'url': manifest_url
,
2258 'manifest_url': m3u8_url
,
2259 'language': media
.get('LANGUAGE'),
2261 'protocol': entry_protocol
,
2262 'preference': preference
,
2264 'vcodec': 'none' if media_type
== 'AUDIO' else None,
2265 } for idx
in _extract_m3u8_playlist_indices(manifest_url
))
2267 def build_stream_name():
2268 # Despite specification does not mention NAME attribute for
2269 # EXT-X-STREAM-INF tag it still sometimes may be present (see [1]
2270 # or vidio test in TestInfoExtractor.test_parse_m3u8_formats)
2271 # 1. http://www.vidio.com/watch/165683-dj_ambred-booyah-live-2015
2272 stream_name
= last_stream_inf
.get('NAME')
2275 # If there is no NAME in EXT-X-STREAM-INF it will be obtained
2276 # from corresponding rendition group
2277 stream_group_id
= last_stream_inf
.get('VIDEO')
2278 if not stream_group_id
:
2280 stream_group
= groups
.get(stream_group_id
)
2281 if not stream_group
:
2282 return stream_group_id
2283 rendition
= stream_group
[0]
2284 return rendition
.get('NAME') or stream_group_id
2286 # parse EXT-X-MEDIA tags before EXT-X-STREAM-INF in order to have the
2287 # chance to detect video only formats when EXT-X-STREAM-INF tags
2288 # precede EXT-X-MEDIA tags in HLS manifest such as [3].
2289 for line
in m3u8_doc
.splitlines():
2290 if line
.startswith('#EXT-X-MEDIA:'):
2293 for line
in m3u8_doc
.splitlines():
2294 if line
.startswith('#EXT-X-STREAM-INF:'):
2295 last_stream_inf
= parse_m3u8_attributes(line
)
2296 elif line
.startswith('#') or not line
.strip():
2299 tbr
= float_or_none(
2300 last_stream_inf
.get('AVERAGE-BANDWIDTH')
2301 or last_stream_inf
.get('BANDWIDTH'), scale
=1000)
2302 manifest_url
= format_url(line
.strip())
2304 for idx
in _extract_m3u8_playlist_indices(manifest_url
):
2305 format_id
= [m3u8_id
, None, idx
]
2306 # Bandwidth of live streams may differ over time thus making
2307 # format_id unpredictable. So it's better to keep provided
2310 stream_name
= build_stream_name()
2311 format_id
[1] = stream_name
or '%d' % (tbr
or len(formats
))
2313 'format_id': join_nonempty(*format_id
),
2314 'format_index': idx
,
2315 'url': manifest_url
,
2316 'manifest_url': m3u8_url
,
2319 'fps': float_or_none(last_stream_inf
.get('FRAME-RATE')),
2320 'protocol': entry_protocol
,
2321 'preference': preference
,
2324 resolution
= last_stream_inf
.get('RESOLUTION')
2326 mobj
= re
.search(r
'(?P<width>\d+)[xX](?P<height>\d+)', resolution
)
2328 f
['width'] = int(mobj
.group('width'))
2329 f
['height'] = int(mobj
.group('height'))
2330 # Unified Streaming Platform
2332 r
'audio.*?(?:%3D|=)(\d+)(?:-video.*?(?:%3D|=)(\d+))?', f
['url'])
2334 abr
, vbr
= mobj
.groups()
2335 abr
, vbr
= float_or_none(abr
, 1000), float_or_none(vbr
, 1000)
2340 codecs
= parse_codecs(last_stream_inf
.get('CODECS'))
2342 audio_group_id
= last_stream_inf
.get('AUDIO')
2343 # As per [1, 4.3.4.1.1] any EXT-X-STREAM-INF tag which
2344 # references a rendition group MUST have a CODECS attribute.
2345 # However, this is not always respected, for example, [2]
2346 # contains EXT-X-STREAM-INF tag which references AUDIO
2347 # rendition group but does not have CODECS and despite
2348 # referencing an audio group it represents a complete
2349 # (with audio and video) format. So, for such cases we will
2350 # ignore references to rendition groups and treat them
2351 # as complete formats.
2352 if audio_group_id
and codecs
and f
.get('vcodec') != 'none':
2353 audio_group
= groups
.get(audio_group_id
)
2354 if audio_group
and audio_group
[0].get('URI'):
2355 # TODO: update acodec for audio only formats with
2357 f
['acodec'] = 'none'
2358 if not f
.get('ext'):
2359 f
['ext'] = 'm4a' if f
.get('vcodec') == 'none' else 'mp4'
2363 progressive_uri
= last_stream_inf
.get('PROGRESSIVE-URI')
2366 del http_f
['manifest_url']
2368 'format_id': f
['format_id'].replace('hls-', 'http-'),
2370 'url': progressive_uri
,
2372 formats
.append(http_f
)
2374 last_stream_inf
= {}
2375 return formats
, subtitles
2377 def _extract_m3u8_vod_duration(
2378 self
, m3u8_vod_url
, video_id
, note
=None, errnote
=None, data
=None, headers
={}, query={}
):
2380 m3u8_vod
= self
._download
_webpage
(
2381 m3u8_vod_url
, video_id
,
2382 note
='Downloading m3u8 VOD manifest' if note
is None else note
,
2383 errnote
='Failed to download VOD manifest' if errnote
is None else errnote
,
2384 fatal
=False, data
=data
, headers
=headers
, query
=query
)
2386 return self
._parse
_m
3u8_vod
_duration
(m3u8_vod
or '', video_id
)
2388 def _parse_m3u8_vod_duration(self
, m3u8_vod
, video_id
):
2389 if '#EXT-X-PLAYLIST-TYPE:VOD' not in m3u8_vod
:
2393 float(line
[len('#EXTINF:'):].split(',')[0])
2394 for line
in m3u8_vod
.splitlines() if line
.startswith('#EXTINF:'))) or None
2397 def _xpath_ns(path
, namespace
=None):
2401 for c
in path
.split('/'):
2402 if not c
or c
== '.':
2405 out
.append('{%s}%s' % (namespace
, c
))
2406 return '/'.join(out
)
2408 def _extract_smil_formats_and_subtitles(self
, smil_url
, video_id
, fatal
=True, f4m_params
=None, transform_source
=None):
2409 res
= self
._download
_smil
(smil_url
, video_id
, fatal
=fatal
, transform_source
=transform_source
)
2415 smil_url
= urlh
.geturl()
2417 namespace
= self
._parse
_smil
_namespace
(smil
)
2419 fmts
= self
._parse
_smil
_formats
(
2420 smil
, smil_url
, video_id
, namespace
=namespace
, f4m_params
=f4m_params
)
2421 subs
= self
._parse
_smil
_subtitles
(
2422 smil
, namespace
=namespace
)
2426 def _extract_smil_formats(self
, *args
, **kwargs
):
2427 fmts
, subs
= self
._extract
_smil
_formats
_and
_subtitles
(*args
, **kwargs
)
2429 self
._report
_ignoring
_subs
('SMIL')
2432 def _extract_smil_info(self
, smil_url
, video_id
, fatal
=True, f4m_params
=None):
2433 res
= self
._download
_smil
(smil_url
, video_id
, fatal
=fatal
)
2438 smil_url
= urlh
.geturl()
2440 return self
._parse
_smil
(smil
, smil_url
, video_id
, f4m_params
=f4m_params
)
2442 def _download_smil(self
, smil_url
, video_id
, fatal
=True, transform_source
=None):
2443 return self
._download
_xml
_handle
(
2444 smil_url
, video_id
, 'Downloading SMIL file',
2445 'Unable to download SMIL file', fatal
=fatal
, transform_source
=transform_source
)
2447 def _parse_smil(self
, smil
, smil_url
, video_id
, f4m_params
=None):
2448 namespace
= self
._parse
_smil
_namespace
(smil
)
2450 formats
= self
._parse
_smil
_formats
(
2451 smil
, smil_url
, video_id
, namespace
=namespace
, f4m_params
=f4m_params
)
2452 subtitles
= self
._parse
_smil
_subtitles
(smil
, namespace
=namespace
)
2454 video_id
= os
.path
.splitext(url_basename(smil_url
))[0]
2458 for meta
in smil
.findall(self
._xpath
_ns
('./head/meta', namespace
)):
2459 name
= meta
.attrib
.get('name')
2460 content
= meta
.attrib
.get('content')
2461 if not name
or not content
:
2463 if not title
and name
== 'title':
2465 elif not description
and name
in ('description', 'abstract'):
2466 description
= content
2467 elif not upload_date
and name
== 'date':
2468 upload_date
= unified_strdate(content
)
2471 'id': image
.get('type'),
2472 'url': image
.get('src'),
2473 'width': int_or_none(image
.get('width')),
2474 'height': int_or_none(image
.get('height')),
2475 } for image
in smil
.findall(self
._xpath
_ns
('.//image', namespace
)) if image
.get('src')]
2479 'title': title
or video_id
,
2480 'description': description
,
2481 'upload_date': upload_date
,
2482 'thumbnails': thumbnails
,
2484 'subtitles': subtitles
,
2487 def _parse_smil_namespace(self
, smil
):
2488 return self
._search
_regex
(
2489 r
'(?i)^{([^}]+)?}smil$', smil
.tag
, 'namespace', default
=None)
2491 def _parse_smil_formats(self
, smil
, smil_url
, video_id
, namespace
=None, f4m_params
=None, transform_rtmp_url
=None):
2493 for meta
in smil
.findall(self
._xpath
_ns
('./head/meta', namespace
)):
2494 b
= meta
.get('base') or meta
.get('httpBase')
2506 media
= smil
.findall(self
._xpath
_ns
('.//video', namespace
)) + smil
.findall(self
._xpath
_ns
('.//audio', namespace
))
2507 for medium
in media
:
2508 src
= medium
.get('src')
2509 if not src
or src
in srcs
:
2513 bitrate
= float_or_none(medium
.get('system-bitrate') or medium
.get('systemBitrate'), 1000)
2514 filesize
= int_or_none(medium
.get('size') or medium
.get('fileSize'))
2515 width
= int_or_none(medium
.get('width'))
2516 height
= int_or_none(medium
.get('height'))
2517 proto
= medium
.get('proto')
2518 ext
= medium
.get('ext')
2519 src_ext
= determine_ext(src
)
2520 streamer
= medium
.get('streamer') or base
2522 if proto
== 'rtmp' or streamer
.startswith('rtmp'):
2528 'format_id': 'rtmp-%d' % (rtmp_count
if bitrate
is None else bitrate
),
2530 'filesize': filesize
,
2534 if transform_rtmp_url
:
2535 streamer
, src
= transform_rtmp_url(streamer
, src
)
2536 formats
[-1].update({
2542 src_url
= src
if src
.startswith('http') else compat_urlparse
.urljoin(base
, src
)
2543 src_url
= src_url
.strip()
2545 if proto
== 'm3u8' or src_ext
== 'm3u8':
2546 m3u8_formats
= self
._extract
_m
3u8_formats
(
2547 src_url
, video_id
, ext
or 'mp4', m3u8_id
='hls', fatal
=False)
2548 if len(m3u8_formats
) == 1:
2550 m3u8_formats
[0].update({
2551 'format_id': 'hls-%d' % (m3u8_count
if bitrate
is None else bitrate
),
2556 formats
.extend(m3u8_formats
)
2557 elif src_ext
== 'f4m':
2562 'plugin': 'flowplayer-3.2.0.1',
2564 f4m_url
+= '&' if '?' in f4m_url
else '?'
2565 f4m_url
+= compat_urllib_parse_urlencode(f4m_params
)
2566 formats
.extend(self
._extract
_f
4m
_formats
(f4m_url
, video_id
, f4m_id
='hds', fatal
=False))
2567 elif src_ext
== 'mpd':
2568 formats
.extend(self
._extract
_mpd
_formats
(
2569 src_url
, video_id
, mpd_id
='dash', fatal
=False))
2570 elif re
.search(r
'\.ism/[Mm]anifest', src_url
):
2571 formats
.extend(self
._extract
_ism
_formats
(
2572 src_url
, video_id
, ism_id
='mss', fatal
=False))
2573 elif src_url
.startswith('http') and self
._is
_valid
_url
(src
, video_id
):
2577 'ext': ext
or src_ext
or 'flv',
2578 'format_id': 'http-%d' % (bitrate
or http_count
),
2580 'filesize': filesize
,
2585 for medium
in smil
.findall(self
._xpath
_ns
('.//imagestream', namespace
)):
2586 src
= medium
.get('src')
2587 if not src
or src
in srcs
:
2593 'format_id': 'imagestream-%d' % (imgs_count
),
2595 'ext': mimetype2ext(medium
.get('type')),
2598 'width': int_or_none(medium
.get('width')),
2599 'height': int_or_none(medium
.get('height')),
2600 'format_note': 'SMIL storyboards',
2605 def _parse_smil_subtitles(self
, smil
, namespace
=None, subtitles_lang
='en'):
2608 for num
, textstream
in enumerate(smil
.findall(self
._xpath
_ns
('.//textstream', namespace
))):
2609 src
= textstream
.get('src')
2610 if not src
or src
in urls
:
2613 ext
= textstream
.get('ext') or mimetype2ext(textstream
.get('type')) or determine_ext(src
)
2614 lang
= textstream
.get('systemLanguage') or textstream
.get('systemLanguageName') or textstream
.get('lang') or subtitles_lang
2615 subtitles
.setdefault(lang
, []).append({
2621 def _extract_xspf_playlist(self
, xspf_url
, playlist_id
, fatal
=True):
2622 res
= self
._download
_xml
_handle
(
2623 xspf_url
, playlist_id
, 'Downloading xpsf playlist',
2624 'Unable to download xspf manifest', fatal
=fatal
)
2629 xspf_url
= urlh
.geturl()
2631 return self
._parse
_xspf
(
2632 xspf
, playlist_id
, xspf_url
=xspf_url
,
2633 xspf_base_url
=base_url(xspf_url
))
2635 def _parse_xspf(self
, xspf_doc
, playlist_id
, xspf_url
=None, xspf_base_url
=None):
2637 'xspf': 'http://xspf.org/ns/0/',
2638 's1': 'http://static.streamone.nl/player/ns/0',
2642 for track
in xspf_doc
.findall(xpath_with_ns('./xspf:trackList/xspf:track', NS_MAP
)):
2644 track
, xpath_with_ns('./xspf:title', NS_MAP
), 'title', default
=playlist_id
)
2645 description
= xpath_text(
2646 track
, xpath_with_ns('./xspf:annotation', NS_MAP
), 'description')
2647 thumbnail
= xpath_text(
2648 track
, xpath_with_ns('./xspf:image', NS_MAP
), 'thumbnail')
2649 duration
= float_or_none(
2650 xpath_text(track
, xpath_with_ns('./xspf:duration', NS_MAP
), 'duration'), 1000)
2653 for location
in track
.findall(xpath_with_ns('./xspf:location', NS_MAP
)):
2654 format_url
= urljoin(xspf_base_url
, location
.text
)
2659 'manifest_url': xspf_url
,
2660 'format_id': location
.get(xpath_with_ns('s1:label', NS_MAP
)),
2661 'width': int_or_none(location
.get(xpath_with_ns('s1:width', NS_MAP
))),
2662 'height': int_or_none(location
.get(xpath_with_ns('s1:height', NS_MAP
))),
2664 self
._sort
_formats
(formats
)
2669 'description': description
,
2670 'thumbnail': thumbnail
,
2671 'duration': duration
,
2676 def _extract_mpd_formats(self
, *args
, **kwargs
):
2677 fmts
, subs
= self
._extract
_mpd
_formats
_and
_subtitles
(*args
, **kwargs
)
2679 self
._report
_ignoring
_subs
('DASH')
2682 def _extract_mpd_formats_and_subtitles(
2683 self
, mpd_url
, video_id
, mpd_id
=None, note
=None, errnote
=None,
2684 fatal
=True, data
=None, headers
={}, query={}
):
2685 res
= self
._download
_xml
_handle
(
2687 note
='Downloading MPD manifest' if note
is None else note
,
2688 errnote
='Failed to download MPD manifest' if errnote
is None else errnote
,
2689 fatal
=fatal
, data
=data
, headers
=headers
, query
=query
)
2696 # We could have been redirected to a new url when we retrieved our mpd file.
2697 mpd_url
= urlh
.geturl()
2698 mpd_base_url
= base_url(mpd_url
)
2700 return self
._parse
_mpd
_formats
_and
_subtitles
(
2701 mpd_doc
, mpd_id
, mpd_base_url
, mpd_url
)
2703 def _parse_mpd_formats(self
, *args
, **kwargs
):
2704 fmts
, subs
= self
._parse
_mpd
_formats
_and
_subtitles
(*args
, **kwargs
)
2706 self
._report
_ignoring
_subs
('DASH')
2709 def _parse_mpd_formats_and_subtitles(
2710 self
, mpd_doc
, mpd_id
=None, mpd_base_url
='', mpd_url
=None):
2712 Parse formats from MPD manifest.
2714 1. MPEG-DASH Standard, ISO/IEC 23009-1:2014(E),
2715 http://standards.iso.org/ittf/PubliclyAvailableStandards/c065274_ISO_IEC_23009-1_2014.zip
2716 2. https://en.wikipedia.org/wiki/Dynamic_Adaptive_Streaming_over_HTTP
2718 if not self
.get_param('dynamic_mpd', True):
2719 if mpd_doc
.get('type') == 'dynamic':
2722 namespace
= self
._search
_regex
(r
'(?i)^{([^}]+)?}MPD$', mpd_doc
.tag
, 'namespace', default
=None)
2725 return self
._xpath
_ns
(path
, namespace
)
2727 def is_drm_protected(element
):
2728 return element
.find(_add_ns('ContentProtection')) is not None
2730 def extract_multisegment_info(element
, ms_parent_info
):
2731 ms_info
= ms_parent_info
.copy()
2733 # As per [1, 5.3.9.2.2] SegmentList and SegmentTemplate share some
2734 # common attributes and elements. We will only extract relevant
2736 def extract_common(source
):
2737 segment_timeline
= source
.find(_add_ns('SegmentTimeline'))
2738 if segment_timeline
is not None:
2739 s_e
= segment_timeline
.findall(_add_ns('S'))
2741 ms_info
['total_number'] = 0
2744 r
= int(s
.get('r', 0))
2745 ms_info
['total_number'] += 1 + r
2746 ms_info
['s'].append({
2747 't': int(s
.get('t', 0)),
2748 # @d is mandatory (see [1, 5.3.9.6.2, Table 17, page 60])
2749 'd': int(s
.attrib
['d']),
2752 start_number
= source
.get('startNumber')
2754 ms_info
['start_number'] = int(start_number
)
2755 timescale
= source
.get('timescale')
2757 ms_info
['timescale'] = int(timescale
)
2758 segment_duration
= source
.get('duration')
2759 if segment_duration
:
2760 ms_info
['segment_duration'] = float(segment_duration
)
2762 def extract_Initialization(source
):
2763 initialization
= source
.find(_add_ns('Initialization'))
2764 if initialization
is not None:
2765 ms_info
['initialization_url'] = initialization
.attrib
['sourceURL']
2767 segment_list
= element
.find(_add_ns('SegmentList'))
2768 if segment_list
is not None:
2769 extract_common(segment_list
)
2770 extract_Initialization(segment_list
)
2771 segment_urls_e
= segment_list
.findall(_add_ns('SegmentURL'))
2773 ms_info
['segment_urls'] = [segment
.attrib
['media'] for segment
in segment_urls_e
]
2775 segment_template
= element
.find(_add_ns('SegmentTemplate'))
2776 if segment_template
is not None:
2777 extract_common(segment_template
)
2778 media
= segment_template
.get('media')
2780 ms_info
['media'] = media
2781 initialization
= segment_template
.get('initialization')
2783 ms_info
['initialization'] = initialization
2785 extract_Initialization(segment_template
)
2788 mpd_duration
= parse_duration(mpd_doc
.get('mediaPresentationDuration'))
2789 formats
, subtitles
= [], {}
2790 stream_numbers
= collections
.defaultdict(int)
2791 for period
in mpd_doc
.findall(_add_ns('Period')):
2792 period_duration
= parse_duration(period
.get('duration')) or mpd_duration
2793 period_ms_info
= extract_multisegment_info(period
, {
2797 for adaptation_set
in period
.findall(_add_ns('AdaptationSet')):
2798 adaption_set_ms_info
= extract_multisegment_info(adaptation_set
, period_ms_info
)
2799 for representation
in adaptation_set
.findall(_add_ns('Representation')):
2800 representation_attrib
= adaptation_set
.attrib
.copy()
2801 representation_attrib
.update(representation
.attrib
)
2802 # According to [1, 5.3.7.2, Table 9, page 41], @mimeType is mandatory
2803 mime_type
= representation_attrib
['mimeType']
2804 content_type
= representation_attrib
.get('contentType', mime_type
.split('/')[0])
2806 codecs
= parse_codecs(representation_attrib
.get('codecs', ''))
2807 if content_type
not in ('video', 'audio', 'text'):
2808 if mime_type
== 'image/jpeg':
2809 content_type
= mime_type
2810 elif codecs
['vcodec'] != 'none':
2811 content_type
= 'video'
2812 elif codecs
['acodec'] != 'none':
2813 content_type
= 'audio'
2814 elif codecs
.get('scodec', 'none') != 'none':
2815 content_type
= 'text'
2816 elif mimetype2ext(mime_type
) in ('tt', 'dfxp', 'ttml', 'xml', 'json'):
2817 content_type
= 'text'
2819 self
.report_warning('Unknown MIME type %s in DASH manifest' % mime_type
)
2823 for element
in (representation
, adaptation_set
, period
, mpd_doc
):
2824 base_url_e
= element
.find(_add_ns('BaseURL'))
2825 if base_url_e
is not None:
2826 base_url
= base_url_e
.text
+ base_url
2827 if re
.match(r
'^https?://', base_url
):
2829 if mpd_base_url
and base_url
.startswith('/'):
2830 base_url
= compat_urlparse
.urljoin(mpd_base_url
, base_url
)
2831 elif mpd_base_url
and not re
.match(r
'^https?://', base_url
):
2832 if not mpd_base_url
.endswith('/'):
2834 base_url
= mpd_base_url
+ base_url
2835 representation_id
= representation_attrib
.get('id')
2836 lang
= representation_attrib
.get('lang')
2837 url_el
= representation
.find(_add_ns('BaseURL'))
2838 filesize
= int_or_none(url_el
.attrib
.get('{http://youtube.com/yt/2012/10/10}contentLength') if url_el
is not None else None)
2839 bandwidth
= int_or_none(representation_attrib
.get('bandwidth'))
2840 if representation_id
is not None:
2841 format_id
= representation_id
2843 format_id
= content_type
2845 format_id
= mpd_id
+ '-' + format_id
2846 if content_type
in ('video', 'audio'):
2848 'format_id': format_id
,
2849 'manifest_url': mpd_url
,
2850 'ext': mimetype2ext(mime_type
),
2851 'width': int_or_none(representation_attrib
.get('width')),
2852 'height': int_or_none(representation_attrib
.get('height')),
2853 'tbr': float_or_none(bandwidth
, 1000),
2854 'asr': int_or_none(representation_attrib
.get('audioSamplingRate')),
2855 'fps': int_or_none(representation_attrib
.get('frameRate')),
2856 'language': lang
if lang
not in ('mul', 'und', 'zxx', 'mis') else None,
2857 'format_note': 'DASH %s' % content_type
,
2858 'filesize': filesize
,
2859 'container': mimetype2ext(mime_type
) + '_dash',
2862 elif content_type
== 'text':
2864 'ext': mimetype2ext(mime_type
),
2865 'manifest_url': mpd_url
,
2866 'filesize': filesize
,
2868 elif content_type
== 'image/jpeg':
2869 # See test case in VikiIE
2870 # https://www.viki.com/videos/1175236v-choosing-spouse-by-lottery-episode-1
2872 'format_id': format_id
,
2874 'manifest_url': mpd_url
,
2875 'format_note': 'DASH storyboards (jpeg)',
2879 if is_drm_protected(adaptation_set
) or is_drm_protected(representation
):
2881 representation_ms_info
= extract_multisegment_info(representation
, adaption_set_ms_info
)
2883 def prepare_template(template_name
, identifiers
):
2884 tmpl
= representation_ms_info
[template_name
]
2885 # First of, % characters outside $...$ templates
2886 # must be escaped by doubling for proper processing
2887 # by % operator string formatting used further (see
2888 # https://github.com/ytdl-org/youtube-dl/issues/16867).
2894 in_template
= not in_template
2895 elif c
== '%' and not in_template
:
2897 # Next, $...$ templates are translated to their
2898 # %(...) counterparts to be used with % operator
2899 if representation_id
is not None:
2900 t
= t
.replace('$RepresentationID$', representation_id
)
2901 t
= re
.sub(r
'\$(%s)\$' % '|'.join(identifiers
), r
'%(\1)d', t
)
2902 t
= re
.sub(r
'\$(%s)%%([^$]+)\$' % '|'.join(identifiers
), r
'%(\1)\2', t
)
2903 t
.replace('$$', '$')
2906 # @initialization is a regular template like @media one
2907 # so it should be handled just the same way (see
2908 # https://github.com/ytdl-org/youtube-dl/issues/11605)
2909 if 'initialization' in representation_ms_info
:
2910 initialization_template
= prepare_template(
2912 # As per [1, 5.3.9.4.2, Table 15, page 54] $Number$ and
2913 # $Time$ shall not be included for @initialization thus
2914 # only $Bandwidth$ remains
2916 representation_ms_info
['initialization_url'] = initialization_template
% {
2917 'Bandwidth': bandwidth
,
2920 def location_key(location
):
2921 return 'url' if re
.match(r
'^https?://', location
) else 'path'
2923 if 'segment_urls' not in representation_ms_info
and 'media' in representation_ms_info
:
2925 media_template
= prepare_template('media', ('Number', 'Bandwidth', 'Time'))
2926 media_location_key
= location_key(media_template
)
2928 # As per [1, 5.3.9.4.4, Table 16, page 55] $Number$ and $Time$
2929 # can't be used at the same time
2930 if '%(Number' in media_template
and 's' not in representation_ms_info
:
2931 segment_duration
= None
2932 if 'total_number' not in representation_ms_info
and 'segment_duration' in representation_ms_info
:
2933 segment_duration
= float_or_none(representation_ms_info
['segment_duration'], representation_ms_info
['timescale'])
2934 representation_ms_info
['total_number'] = int(math
.ceil(
2935 float_or_none(period_duration
, segment_duration
, default
=0)))
2936 representation_ms_info
['fragments'] = [{
2937 media_location_key
: media_template
% {
2938 'Number': segment_number
,
2939 'Bandwidth': bandwidth
,
2941 'duration': segment_duration
,
2942 } for segment_number
in range(
2943 representation_ms_info
['start_number'],
2944 representation_ms_info
['total_number'] + representation_ms_info
['start_number'])]
2946 # $Number*$ or $Time$ in media template with S list available
2947 # Example $Number*$: http://www.svtplay.se/klipp/9023742/stopptid-om-bjorn-borg
2948 # Example $Time$: https://play.arkena.com/embed/avp/v2/player/media/b41dda37-d8e7-4d3f-b1b5-9a9db578bdfe/1/129411
2949 representation_ms_info
['fragments'] = []
2952 segment_number
= representation_ms_info
['start_number']
2954 def add_segment_url():
2955 segment_url
= media_template
% {
2956 'Time': segment_time
,
2957 'Bandwidth': bandwidth
,
2958 'Number': segment_number
,
2960 representation_ms_info
['fragments'].append({
2961 media_location_key
: segment_url
,
2962 'duration': float_or_none(segment_d
, representation_ms_info
['timescale']),
2965 for num
, s
in enumerate(representation_ms_info
['s']):
2966 segment_time
= s
.get('t') or segment_time
2970 for r
in range(s
.get('r', 0)):
2971 segment_time
+= segment_d
2974 segment_time
+= segment_d
2975 elif 'segment_urls' in representation_ms_info
and 's' in representation_ms_info
:
2977 # Example: https://www.youtube.com/watch?v=iXZV5uAYMJI
2978 # or any YouTube dashsegments video
2981 timescale
= representation_ms_info
['timescale']
2982 for s
in representation_ms_info
['s']:
2983 duration
= float_or_none(s
['d'], timescale
)
2984 for r
in range(s
.get('r', 0) + 1):
2985 segment_uri
= representation_ms_info
['segment_urls'][segment_index
]
2987 location_key(segment_uri
): segment_uri
,
2988 'duration': duration
,
2991 representation_ms_info
['fragments'] = fragments
2992 elif 'segment_urls' in representation_ms_info
:
2993 # Segment URLs with no SegmentTimeline
2994 # Example: https://www.seznam.cz/zpravy/clanek/cesko-zasahne-vitr-o-sile-vichrice-muze-byt-i-zivotu-nebezpecny-39091
2995 # https://github.com/ytdl-org/youtube-dl/pull/14844
2997 segment_duration
= float_or_none(
2998 representation_ms_info
['segment_duration'],
2999 representation_ms_info
['timescale']) if 'segment_duration' in representation_ms_info
else None
3000 for segment_url
in representation_ms_info
['segment_urls']:
3002 location_key(segment_url
): segment_url
,
3004 if segment_duration
:
3005 fragment
['duration'] = segment_duration
3006 fragments
.append(fragment
)
3007 representation_ms_info
['fragments'] = fragments
3008 # If there is a fragments key available then we correctly recognized fragmented media.
3009 # Otherwise we will assume unfragmented media with direct access. Technically, such
3010 # assumption is not necessarily correct since we may simply have no support for
3011 # some forms of fragmented media renditions yet, but for now we'll use this fallback.
3012 if 'fragments' in representation_ms_info
:
3014 # NB: mpd_url may be empty when MPD manifest is parsed from a string
3015 'url': mpd_url
or base_url
,
3016 'fragment_base_url': base_url
,
3018 'protocol': 'http_dash_segments' if mime_type
!= 'image/jpeg' else 'mhtml',
3020 if 'initialization_url' in representation_ms_info
:
3021 initialization_url
= representation_ms_info
['initialization_url']
3022 if not f
.get('url'):
3023 f
['url'] = initialization_url
3024 f
['fragments'].append({location_key(initialization_url): initialization_url}
)
3025 f
['fragments'].extend(representation_ms_info
['fragments'])
3026 if not period_duration
:
3027 period_duration
= try_get(
3028 representation_ms_info
,
3029 lambda r
: sum(frag
['duration'] for frag
in r
['fragments']), float)
3031 # Assuming direct URL to unfragmented media.
3033 if content_type
in ('video', 'audio', 'image/jpeg'):
3034 f
['manifest_stream_number'] = stream_numbers
[f
['url']]
3035 stream_numbers
[f
['url']] += 1
3037 elif content_type
== 'text':
3038 subtitles
.setdefault(lang
or 'und', []).append(f
)
3040 return formats
, subtitles
3042 def _extract_ism_formats(self
, *args
, **kwargs
):
3043 fmts
, subs
= self
._extract
_ism
_formats
_and
_subtitles
(*args
, **kwargs
)
3045 self
._report
_ignoring
_subs
('ISM')
3048 def _extract_ism_formats_and_subtitles(self
, ism_url
, video_id
, ism_id
=None, note
=None, errnote
=None, fatal
=True, data
=None, headers
={}, query={}
):
3049 res
= self
._download
_xml
_handle
(
3051 note
='Downloading ISM manifest' if note
is None else note
,
3052 errnote
='Failed to download ISM manifest' if errnote
is None else errnote
,
3053 fatal
=fatal
, data
=data
, headers
=headers
, query
=query
)
3060 return self
._parse
_ism
_formats
_and
_subtitles
(ism_doc
, urlh
.geturl(), ism_id
)
3062 def _parse_ism_formats_and_subtitles(self
, ism_doc
, ism_url
, ism_id
=None):
3064 Parse formats from ISM manifest.
3066 1. [MS-SSTR]: Smooth Streaming Protocol,
3067 https://msdn.microsoft.com/en-us/library/ff469518.aspx
3069 if ism_doc
.get('IsLive') == 'TRUE':
3072 duration
= int(ism_doc
.attrib
['Duration'])
3073 timescale
= int_or_none(ism_doc
.get('TimeScale')) or 10000000
3077 for stream
in ism_doc
.findall('StreamIndex'):
3078 stream_type
= stream
.get('Type')
3079 if stream_type
not in ('video', 'audio', 'text'):
3081 url_pattern
= stream
.attrib
['Url']
3082 stream_timescale
= int_or_none(stream
.get('TimeScale')) or timescale
3083 stream_name
= stream
.get('Name')
3084 stream_language
= stream
.get('Language', 'und')
3085 for track
in stream
.findall('QualityLevel'):
3086 fourcc
= track
.get('FourCC') or ('AACL' if track
.get('AudioTag') == '255' else None)
3087 # TODO: add support for WVC1 and WMAP
3088 if fourcc
not in ('H264', 'AVC1', 'AACL', 'TTML'):
3089 self
.report_warning('%s is not a supported codec' % fourcc
)
3091 tbr
= int(track
.attrib
['Bitrate']) // 1000
3092 # [1] does not mention Width and Height attributes. However,
3093 # they're often present while MaxWidth and MaxHeight are
3094 # missing, so should be used as fallbacks
3095 width
= int_or_none(track
.get('MaxWidth') or track
.get('Width'))
3096 height
= int_or_none(track
.get('MaxHeight') or track
.get('Height'))
3097 sampling_rate
= int_or_none(track
.get('SamplingRate'))
3099 track_url_pattern
= re
.sub(r
'{[Bb]itrate}', track
.attrib
['Bitrate'], url_pattern
)
3100 track_url_pattern
= compat_urlparse
.urljoin(ism_url
, track_url_pattern
)
3106 stream_fragments
= stream
.findall('c')
3107 for stream_fragment_index
, stream_fragment
in enumerate(stream_fragments
):
3108 fragment_ctx
['time'] = int_or_none(stream_fragment
.get('t')) or fragment_ctx
['time']
3109 fragment_repeat
= int_or_none(stream_fragment
.get('r')) or 1
3110 fragment_ctx
['duration'] = int_or_none(stream_fragment
.get('d'))
3111 if not fragment_ctx
['duration']:
3113 next_fragment_time
= int(stream_fragment
[stream_fragment_index
+ 1].attrib
['t'])
3115 next_fragment_time
= duration
3116 fragment_ctx
['duration'] = (next_fragment_time
- fragment_ctx
['time']) / fragment_repeat
3117 for _
in range(fragment_repeat
):
3119 'url': re
.sub(r
'{start[ _]time}', compat_str(fragment_ctx
['time']), track_url_pattern
),
3120 'duration': fragment_ctx
['duration'] / stream_timescale
,
3122 fragment_ctx
['time'] += fragment_ctx
['duration']
3124 if stream_type
== 'text':
3125 subtitles
.setdefault(stream_language
, []).append({
3129 'manifest_url': ism_url
,
3130 'fragments': fragments
,
3131 '_download_params': {
3132 'stream_type': stream_type
,
3133 'duration': duration
,
3134 'timescale': stream_timescale
,
3136 'language': stream_language
,
3137 'codec_private_data': track
.get('CodecPrivateData'),
3140 elif stream_type
in ('video', 'audio'):
3142 'format_id': join_nonempty(ism_id
, stream_name
, tbr
),
3144 'manifest_url': ism_url
,
3145 'ext': 'ismv' if stream_type
== 'video' else 'isma',
3149 'asr': sampling_rate
,
3150 'vcodec': 'none' if stream_type
== 'audio' else fourcc
,
3151 'acodec': 'none' if stream_type
== 'video' else fourcc
,
3153 'fragments': fragments
,
3154 'has_drm': ism_doc
.find('Protection') is not None,
3155 '_download_params': {
3156 'stream_type': stream_type
,
3157 'duration': duration
,
3158 'timescale': stream_timescale
,
3159 'width': width
or 0,
3160 'height': height
or 0,
3162 'language': stream_language
,
3163 'codec_private_data': track
.get('CodecPrivateData'),
3164 'sampling_rate': sampling_rate
,
3165 'channels': int_or_none(track
.get('Channels', 2)),
3166 'bits_per_sample': int_or_none(track
.get('BitsPerSample', 16)),
3167 'nal_unit_length_field': int_or_none(track
.get('NALUnitLengthField', 4)),
3170 return formats
, subtitles
3172 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):
3173 def absolute_url(item_url
):
3174 return urljoin(base_url
, item_url
)
3176 def parse_content_type(content_type
):
3177 if not content_type
:
3179 ctr
= re
.search(r
'(?P<mimetype>[^/]+/[^;]+)(?:;\s*codecs="?(?P<codecs>[^"]+))?', content_type
)
3181 mimetype
, codecs
= ctr
.groups()
3182 f
= parse_codecs(codecs
)
3183 f
['ext'] = mimetype2ext(mimetype
)
3187 def _media_formats(src
, cur_media_type
, type_info
={}):
3188 full_url
= absolute_url(src
)
3189 ext
= type_info
.get('ext') or determine_ext(full_url
)
3191 is_plain_url
= False
3192 formats
= self
._extract
_m
3u8_formats
(
3193 full_url
, video_id
, ext
='mp4',
3194 entry_protocol
=m3u8_entry_protocol
, m3u8_id
=m3u8_id
,
3195 preference
=preference
, quality
=quality
, fatal
=False)
3197 is_plain_url
= False
3198 formats
= self
._extract
_mpd
_formats
(
3199 full_url
, video_id
, mpd_id
=mpd_id
, fatal
=False)
3204 'vcodec': 'none' if cur_media_type
== 'audio' else None,
3206 return is_plain_url
, formats
3209 # amp-video and amp-audio are very similar to their HTML5 counterparts
3210 # so we wll include them right here (see
3211 # https://www.ampproject.org/docs/reference/components/amp-video)
3212 # For dl8-* tags see https://delight-vr.com/documentation/dl8-video/
3213 _MEDIA_TAG_NAME_RE
= r
'(?:(?:amp|dl8(?:-live)?)-)?(video|audio)'
3214 media_tags
= [(media_tag
, media_tag_name
, media_type
, '')
3215 for media_tag
, media_tag_name
, media_type
3216 in re
.findall(r
'(?s)(<(%s)[^>]*/>)' % _MEDIA_TAG_NAME_RE
, webpage
)]
3217 media_tags
.extend(re
.findall(
3218 # We only allow video|audio followed by a whitespace or '>'.
3219 # Allowing more characters may end up in significant slow down (see
3220 # https://github.com/ytdl-org/youtube-dl/issues/11979, example URL:
3221 # http://www.porntrex.com/maps/videositemap.xml).
3222 r
'(?s)(<(?P<tag>%s)(?:\s+[^>]*)?>)(.*?)</(?P=tag)>' % _MEDIA_TAG_NAME_RE
, webpage
))
3223 for media_tag
, _
, media_type
, media_content
in media_tags
:
3228 media_attributes
= extract_attributes(media_tag
)
3229 src
= strip_or_none(media_attributes
.get('src'))
3231 _
, formats
= _media_formats(src
, media_type
)
3232 media_info
['formats'].extend(formats
)
3233 media_info
['thumbnail'] = absolute_url(media_attributes
.get('poster'))
3235 for source_tag
in re
.findall(r
'<source[^>]+>', media_content
):
3236 s_attr
= extract_attributes(source_tag
)
3237 # data-video-src and data-src are non standard but seen
3238 # several times in the wild
3239 src
= strip_or_none(dict_get(s_attr
, ('src', 'data-video-src', 'data-src')))
3242 f
= parse_content_type(s_attr
.get('type'))
3243 is_plain_url
, formats
= _media_formats(src
, media_type
, f
)
3245 # width, height, res, label and title attributes are
3246 # all not standard but seen several times in the wild
3249 for lbl
in ('label', 'title')
3250 if str_or_none(s_attr
.get(lbl
))
3252 width
= int_or_none(s_attr
.get('width'))
3253 height
= (int_or_none(s_attr
.get('height'))
3254 or int_or_none(s_attr
.get('res')))
3255 if not width
or not height
:
3257 resolution
= parse_resolution(lbl
)
3260 width
= width
or resolution
.get('width')
3261 height
= height
or resolution
.get('height')
3263 tbr
= parse_bitrate(lbl
)
3272 'format_id': s_attr
.get('label') or s_attr
.get('title'),
3274 f
.update(formats
[0])
3275 media_info
['formats'].append(f
)
3277 media_info
['formats'].extend(formats
)
3278 for track_tag
in re
.findall(r
'<track[^>]+>', media_content
):
3279 track_attributes
= extract_attributes(track_tag
)
3280 kind
= track_attributes
.get('kind')
3281 if not kind
or kind
in ('subtitles', 'captions'):
3282 src
= strip_or_none(track_attributes
.get('src'))
3285 lang
= track_attributes
.get('srclang') or track_attributes
.get('lang') or track_attributes
.get('label')
3286 media_info
['subtitles'].setdefault(lang
, []).append({
3287 'url': absolute_url(src
),
3289 for f
in media_info
['formats']:
3290 f
.setdefault('http_headers', {})['Referer'] = base_url
3291 if media_info
['formats'] or media_info
['subtitles']:
3292 entries
.append(media_info
)
3295 def _extract_akamai_formats(self
, *args
, **kwargs
):
3296 fmts
, subs
= self
._extract
_akamai
_formats
_and
_subtitles
(*args
, **kwargs
)
3298 self
._report
_ignoring
_subs
('akamai')
3301 def _extract_akamai_formats_and_subtitles(self
, manifest_url
, video_id
, hosts
={}):
3302 signed
= 'hdnea=' in manifest_url
3304 # https://learn.akamai.com/en-us/webhelp/media-services-on-demand/stream-packaging-user-guide/GUID-BE6C0F73-1E06-483B-B0EA-57984B91B7F9.html
3305 manifest_url
= re
.sub(
3306 r
'(?:b=[\d,-]+|(?:__a__|attributes)=off|__b__=\d+)&?',
3307 '', manifest_url
).strip('?')
3312 hdcore_sign
= 'hdcore=3.7.0'
3313 f4m_url
= re
.sub(r
'(https?://[^/]+)/i/', r
'\1/z/', manifest_url
).replace('/master.m3u8', '/manifest.f4m')
3314 hds_host
= hosts
.get('hds')
3316 f4m_url
= re
.sub(r
'(https?://)[^/]+', r
'\1' + hds_host
, f4m_url
)
3317 if 'hdcore=' not in f4m_url
:
3318 f4m_url
+= ('&' if '?' in f4m_url
else '?') + hdcore_sign
3319 f4m_formats
= self
._extract
_f
4m
_formats
(
3320 f4m_url
, video_id
, f4m_id
='hds', fatal
=False)
3321 for entry
in f4m_formats
:
3322 entry
.update({'extra_param_to_segment_url': hdcore_sign}
)
3323 formats
.extend(f4m_formats
)
3325 m3u8_url
= re
.sub(r
'(https?://[^/]+)/z/', r
'\1/i/', manifest_url
).replace('/manifest.f4m', '/master.m3u8')
3326 hls_host
= hosts
.get('hls')
3328 m3u8_url
= re
.sub(r
'(https?://)[^/]+', r
'\1' + hls_host
, m3u8_url
)
3329 m3u8_formats
, m3u8_subtitles
= self
._extract
_m
3u8_formats
_and
_subtitles
(
3330 m3u8_url
, video_id
, 'mp4', 'm3u8_native',
3331 m3u8_id
='hls', fatal
=False)
3332 formats
.extend(m3u8_formats
)
3333 subtitles
= self
._merge
_subtitles
(subtitles
, m3u8_subtitles
)
3335 http_host
= hosts
.get('http')
3336 if http_host
and m3u8_formats
and not signed
:
3337 REPL_REGEX
= r
'https?://[^/]+/i/([^,]+),([^/]+),([^/]+)\.csmil/.+'
3338 qualities
= re
.match(REPL_REGEX
, m3u8_url
).group(2).split(',')
3339 qualities_length
= len(qualities
)
3340 if len(m3u8_formats
) in (qualities_length
, qualities_length
+ 1):
3342 for f
in m3u8_formats
:
3343 if f
['vcodec'] != 'none':
3344 for protocol
in ('http', 'https'):
3346 del http_f
['manifest_url']
3348 REPL_REGEX
, protocol
+ fr
'://{http_host}/\g<1>{qualities[i]}\3', f
['url'])
3350 'format_id': http_f
['format_id'].replace('hls-', protocol
+ '-'),
3352 'protocol': protocol
,
3354 formats
.append(http_f
)
3357 return formats
, subtitles
3359 def _extract_wowza_formats(self
, url
, video_id
, m3u8_entry_protocol
='m3u8_native', skip_protocols
=[]):
3360 query
= compat_urlparse
.urlparse(url
).query
3361 url
= re
.sub(r
'/(?:manifest|playlist|jwplayer)\.(?:m3u8|f4m|mpd|smil)', '', url
)
3363 r
'(?:(?:http|rtmp|rtsp)(?P<s>s)?:)?(?P<url>//[^?]+)', url
)
3364 url_base
= mobj
.group('url')
3365 http_base_url
= '%s%s:%s' % ('http', mobj
.group('s') or '', url_base
)
3368 def manifest_url(manifest
):
3369 m_url
= f
'{http_base_url}/{manifest}'
3371 m_url
+= '?%s' % query
3374 if 'm3u8' not in skip_protocols
:
3375 formats
.extend(self
._extract
_m
3u8_formats
(
3376 manifest_url('playlist.m3u8'), video_id
, 'mp4',
3377 m3u8_entry_protocol
, m3u8_id
='hls', fatal
=False))
3378 if 'f4m' not in skip_protocols
:
3379 formats
.extend(self
._extract
_f
4m
_formats
(
3380 manifest_url('manifest.f4m'),
3381 video_id
, f4m_id
='hds', fatal
=False))
3382 if 'dash' not in skip_protocols
:
3383 formats
.extend(self
._extract
_mpd
_formats
(
3384 manifest_url('manifest.mpd'),
3385 video_id
, mpd_id
='dash', fatal
=False))
3386 if re
.search(r
'(?:/smil:|\.smil)', url_base
):
3387 if 'smil' not in skip_protocols
:
3388 rtmp_formats
= self
._extract
_smil
_formats
(
3389 manifest_url('jwplayer.smil'),
3390 video_id
, fatal
=False)
3391 for rtmp_format
in rtmp_formats
:
3392 rtsp_format
= rtmp_format
.copy()
3393 rtsp_format
['url'] = '%s/%s' % (rtmp_format
['url'], rtmp_format
['play_path'])
3394 del rtsp_format
['play_path']
3395 del rtsp_format
['ext']
3396 rtsp_format
.update({
3397 'url': rtsp_format
['url'].replace('rtmp://', 'rtsp://'),
3398 'format_id': rtmp_format
['format_id'].replace('rtmp', 'rtsp'),
3401 formats
.extend([rtmp_format
, rtsp_format
])
3403 for protocol
in ('rtmp', 'rtsp'):
3404 if protocol
not in skip_protocols
:
3406 'url': f
'{protocol}:{url_base}',
3407 'format_id': protocol
,
3408 'protocol': protocol
,
3412 def _find_jwplayer_data(self
, webpage
, video_id
=None, transform_source
=js_to_json
):
3414 r
'(?s)jwplayer\((?P<quote>[\'"])[^\'" ]+(?P
=quote
)\
)(?
!</script
>).*?\
.setup\s
*\
((?P
<options
>[^
)]+)\
)',
3418 jwplayer_data = self._parse_json(mobj.group('options
'),
3420 transform_source=transform_source)
3421 except ExtractorError:
3424 if isinstance(jwplayer_data, dict):
3425 return jwplayer_data
3427 def _extract_jwplayer_data(self, webpage, video_id, *args, **kwargs):
3428 jwplayer_data = self._find_jwplayer_data(
3429 webpage, video_id, transform_source=js_to_json)
3430 return self._parse_jwplayer_data(
3431 jwplayer_data, video_id, *args, **kwargs)
3433 def _parse_jwplayer_data(self, jwplayer_data, video_id=None, require_title=True,
3434 m3u8_id=None, mpd_id=None, rtmp_params=None, base_url=None):
3435 # JWPlayer backward compatibility: flattened playlists
3436 # https://github.com/jwplayer/jwplayer/blob/v7.4.3/src/js/api/config.js#L81-L96
3437 if 'playlist
' not in jwplayer_data:
3438 jwplayer_data = {'playlist': [jwplayer_data]}
3442 # JWPlayer backward compatibility: single playlist item
3443 # https://github.com/jwplayer/jwplayer/blob/v7.7.0/src/js/playlist/playlist.js#L10
3444 if not isinstance(jwplayer_data['playlist
'], list):
3445 jwplayer_data['playlist
'] = [jwplayer_data['playlist
']]
3447 for video_data in jwplayer_data['playlist
']:
3448 # JWPlayer backward compatibility: flattened sources
3449 # https://github.com/jwplayer/jwplayer/blob/v7.4.3/src/js/playlist/item.js#L29-L35
3450 if 'sources
' not in video_data:
3451 video_data['sources
'] = [video_data]
3453 this_video_id = video_id or video_data['mediaid
']
3455 formats = self._parse_jwplayer_formats(
3456 video_data['sources
'], video_id=this_video_id, m3u8_id=m3u8_id,
3457 mpd_id=mpd_id, rtmp_params=rtmp_params, base_url=base_url)
3460 tracks = video_data.get('tracks
')
3461 if tracks and isinstance(tracks, list):
3462 for track in tracks:
3463 if not isinstance(track, dict):
3465 track_kind = track.get('kind
')
3466 if not track_kind or not isinstance(track_kind, compat_str):
3468 if track_kind.lower() not in ('captions
', 'subtitles
'):
3470 track_url = urljoin(base_url, track.get('file'))
3473 subtitles.setdefault(track.get('label
') or 'en
', []).append({
3474 'url
': self._proto_relative_url(track_url)
3478 'id': this_video_id,
3479 'title
': unescapeHTML(video_data['title
'] if require_title else video_data.get('title
')),
3480 'description
': clean_html(video_data.get('description
')),
3481 'thumbnail
': urljoin(base_url, self._proto_relative_url(video_data.get('image
'))),
3482 'timestamp
': int_or_none(video_data.get('pubdate
')),
3483 'duration
': float_or_none(jwplayer_data.get('duration
') or video_data.get('duration
')),
3484 'subtitles
': subtitles,
3486 # https://github.com/jwplayer/jwplayer/blob/master/src/js/utils/validator.js#L32
3487 if len(formats) == 1 and re.search(r'^
(?
:http|
//).*(?
:youtube\
.com|youtu\
.be
)/.+', formats[0]['url
']):
3489 '_type
': 'url_transparent
',
3490 'url
': formats[0]['url
'],
3493 self._sort_formats(formats)
3494 entry['formats
'] = formats
3495 entries.append(entry)
3496 if len(entries) == 1:
3499 return self.playlist_result(entries)
3501 def _parse_jwplayer_formats(self, jwplayer_sources_data, video_id=None,
3502 m3u8_id=None, mpd_id=None, rtmp_params=None, base_url=None):
3505 for source in jwplayer_sources_data:
3506 if not isinstance(source, dict):
3508 source_url = urljoin(
3509 base_url, self._proto_relative_url(source.get('file')))
3510 if not source_url or source_url in urls:
3512 urls.append(source_url)
3513 source_type = source.get('type') or ''
3514 ext = mimetype2ext(source_type) or determine_ext(source_url)
3515 if source_type == 'hls
' or ext == 'm3u8
':
3516 formats.extend(self._extract_m3u8_formats(
3517 source_url, video_id, 'mp4
', entry_protocol='m3u8_native
',
3518 m3u8_id=m3u8_id, fatal=False))
3519 elif source_type == 'dash
' or ext == 'mpd
':
3520 formats.extend(self._extract_mpd_formats(
3521 source_url, video_id, mpd_id=mpd_id, fatal=False))
3523 formats.extend(self._extract_smil_formats(
3524 source_url, video_id, fatal=False))
3525 # https://github.com/jwplayer/jwplayer/blob/master/src/js/providers/default.js#L67
3526 elif source_type.startswith('audio
') or ext in (
3527 'oga
', 'aac
', 'mp3
', 'mpeg
', 'vorbis
'):
3534 height = int_or_none(source.get('height
'))
3536 # Often no height is provided but there is a label in
3537 # format like "1080p", "720p SD", or 1080.
3538 height = int_or_none(self._search_regex(
3539 r'^
(\d{3,4}
)[pP
]?
(?
:\b|$
)', compat_str(source.get('label
') or ''),
3540 'height
', default=None))
3543 'width
': int_or_none(source.get('width
')),
3545 'tbr
': int_or_none(source.get('bitrate
')),
3548 if source_url.startswith('rtmp
'):
3549 a_format['ext
'] = 'flv
'
3550 # See com/longtailvideo/jwplayer/media/RTMPMediaProvider.as
3551 # of jwplayer.flash.swf
3552 rtmp_url_parts = re.split(
3553 r'((?
:mp4|mp3|flv
):)', source_url, 1)
3554 if len(rtmp_url_parts) == 3:
3555 rtmp_url, prefix, play_path = rtmp_url_parts
3558 'play_path
': prefix + play_path,
3561 a_format.update(rtmp_params)
3562 formats.append(a_format)
3565 def _live_title(self, name):
3566 self._downloader.deprecation_warning('yt_dlp
.InfoExtractor
._live
_title
is deprecated
and does
not work
as expected
')
3569 def _int(self, v, name, fatal=False, **kwargs):
3570 res = int_or_none(v, **kwargs)
3572 msg = f'Failed to extract {name}
: Could
not parse value {v!r}
'
3574 raise ExtractorError(msg)
3576 self.report_warning(msg)
3579 def _float(self, v, name, fatal=False, **kwargs):
3580 res = float_or_none(v, **kwargs)
3582 msg = f'Failed to extract {name}
: Could
not parse value {v!r}
'
3584 raise ExtractorError(msg)
3586 self.report_warning(msg)
3589 def _set_cookie(self, domain, name, value, expire_time=None, port=None,
3590 path='/', secure=False, discard=False, rest={}, **kwargs):
3591 cookie = compat_cookiejar_Cookie(
3592 0, name, value, port, port is not None, domain, True,
3593 domain.startswith('.'), path, True, secure, expire_time,
3594 discard, None, None, rest)
3595 self._downloader.cookiejar.set_cookie(cookie)
3597 def _get_cookies(self, url):
3598 """ Return a compat_cookies_SimpleCookie with the cookies for the url """
3599 req = sanitized_Request(url)
3600 self._downloader.cookiejar.add_cookie_header(req)
3601 return compat_cookies_SimpleCookie(req.get_header('Cookie
'))
3603 def _apply_first_set_cookie_header(self, url_handle, cookie):
3605 Apply first Set-Cookie header instead of the last. Experimental.
3607 Some sites (e.g. [1-3]) may serve two cookies under the same name
3608 in Set-Cookie header and expect the first (old) one to be set rather
3609 than second (new). However, as of RFC6265 the newer one cookie
3610 should be set into cookie store what actually happens.
3611 We will workaround this issue by resetting the cookie to
3612 the first one manually.
3613 1. https://new.vk.com/
3614 2. https://github.com/ytdl-org/youtube-dl/issues/9841#issuecomment-227871201
3615 3. https://learning.oreilly.com/
3617 for header, cookies in url_handle.headers.items():
3618 if header.lower() != 'set-cookie
':
3620 cookies = cookies.encode('iso
-8859-1').decode('utf
-8')
3621 cookie_value = re.search(
3622 r'%s=(.+?
);.*?
\b[Dd
]omain
=(.+?
)(?
:[,;]|$
)' % cookie, cookies)
3624 value, domain = cookie_value.groups()
3625 self._set_cookie(domain, cookie, value)
3629 def get_testcases(cls, include_onlymatching=False):
3630 t = getattr(cls, '_TEST
', None)
3632 assert not hasattr(cls, '_TESTS
'), f'{cls.ie_key()}IE has _TEST
and _TESTS
'
3635 tests = getattr(cls, '_TESTS
', [])
3637 if not include_onlymatching and t.get('only_matching
', False):
3639 t['name
'] = cls.ie_key()
3644 """Get age limit from the testcases"""
3645 return max(traverse_obj(
3646 tuple(cls.get_testcases(include_onlymatching=False)),
3647 (..., (('playlist
', 0), None), 'info_dict
', 'age_limit
')) or [0])
3650 def is_suitable(cls, age_limit):
3651 """Test whether the extractor is generally suitable for the given age limit"""
3652 return not age_restricted(cls.age_limit, age_limit)
3655 def description(cls, *, markdown=True, search_examples=None):
3656 """Description of the extractor"""
3658 if cls._NETRC_MACHINE:
3660 desc += f' [<abbr title
="netrc machine"><em
>{cls._NETRC_MACHINE}
</em
></abbr
>]'
3662 desc += f' [{cls._NETRC_MACHINE}
]'
3663 if cls.IE_DESC is False:
3666 desc += f' {cls.IE_DESC}
'
3668 desc += f'; "{cls.SEARCH_KEY}:" prefix
'
3670 _COUNTS = ('', '5', '10', 'all
')
3671 desc += f' (Example
: "{cls.SEARCH_KEY}{random.choice(_COUNTS)}:{random.choice(search_examples)}")'
3672 if not cls.working():
3673 desc += ' (**Currently broken
**)' if markdown else ' (Currently broken
)'
3675 name = f' - **{cls.IE_NAME}
**' if markdown else cls.IE_NAME
3676 return f'{name}
:{desc}
' if desc else name
3678 def extract_subtitles(self, *args, **kwargs):
3679 if (self.get_param('writesubtitles
', False)
3680 or self.get_param('listsubtitles
')):
3681 return self._get_subtitles(*args, **kwargs)
3684 def _get_subtitles(self, *args, **kwargs):
3685 raise NotImplementedError('This method must be implemented by subclasses
')
3687 def extract_comments(self, *args, **kwargs):
3688 if not self.get_param('getcomments
'):
3690 generator = self._get_comments(*args, **kwargs)
3697 comments.append(next(generator))
3698 except StopIteration:
3700 except KeyboardInterrupt:
3701 self.to_screen('Interrupted by user
')
3702 except Exception as e:
3703 if self.get_param('ignoreerrors
') is not True:
3705 self._downloader.report_error(e)
3706 comment_count = len(comments)
3707 self.to_screen(f'Extracted {comment_count} comments
')
3709 'comments
': comments,
3710 'comment_count
': None if interrupted else comment_count
3714 def _get_comments(self, *args, **kwargs):
3715 raise NotImplementedError('This method must be implemented by subclasses
')
3718 def _merge_subtitle_items(subtitle_list1, subtitle_list2):
3719 """ Merge subtitle items for one language. Items with duplicated URLs/data
3720 will be dropped. """
3721 list1_data = {(item.get('url'), item.get('data')) for item in subtitle_list1}
3722 ret = list(subtitle_list1)
3723 ret.extend(item for item in subtitle_list2 if (item.get('url
'), item.get('data
')) not in list1_data)
3727 def _merge_subtitles(cls, *dicts, target=None):
3728 """ Merge subtitle dictionaries, language by language. """
3732 for lang, subs in d.items():
3733 target[lang] = cls._merge_subtitle_items(target.get(lang, []), subs)
3736 def extract_automatic_captions(self, *args, **kwargs):
3737 if (self.get_param('writeautomaticsub
', False)
3738 or self.get_param('listsubtitles
')):
3739 return self._get_automatic_captions(*args, **kwargs)
3742 def _get_automatic_captions(self, *args, **kwargs):
3743 raise NotImplementedError('This method must be implemented by subclasses
')
3746 def _cookies_passed(self):
3747 """Whether cookies have been passed to YoutubeDL"""
3748 return self.get_param('cookiefile
') is not None or self.get_param('cookiesfrombrowser
') is not None
3750 def mark_watched(self, *args, **kwargs):
3751 if not self.get_param('mark_watched
', False):
3753 if self.supports_login() and self._get_login_info()[0] is not None or self._cookies_passed:
3754 self._mark_watched(*args, **kwargs)
3756 def _mark_watched(self, *args, **kwargs):
3757 raise NotImplementedError('This method must be implemented by subclasses
')
3759 def geo_verification_headers(self):
3761 geo_verification_proxy = self.get_param('geo_verification_proxy
')
3762 if geo_verification_proxy:
3763 headers['Ytdl
-request
-proxy
'] = geo_verification_proxy
3766 def _generic_id(self, url):
3767 return compat_urllib_parse_unquote(os.path.splitext(url.rstrip('/').split('/')[-1])[0])
3769 def _generic_title(self, url):
3770 return compat_urllib_parse_unquote(os.path.splitext(url_basename(url))[0])
3773 def _availability(is_private=None, needs_premium=None, needs_subscription=None, needs_auth=None, is_unlisted=None):
3774 all_known = all(map(
3775 lambda x: x is not None,
3776 (is_private, needs_premium, needs_subscription, needs_auth, is_unlisted)))
3778 'private
' if is_private
3779 else 'premium_only
' if needs_premium
3780 else 'subscriber_only
' if needs_subscription
3781 else 'needs_auth
' if needs_auth
3782 else 'unlisted
' if is_unlisted
3783 else 'public
' if all_known
3786 def _configuration_arg(self, key, default=NO_DEFAULT, *, ie_key=None, casesense=False):
3788 @returns A list of values for the extractor argument given by "key"
3789 or "default" if no such key is present
3790 @param default The default value to return when the key is not present (default: [])
3791 @param casesense When false, the values are converted to lower case
3794 self._downloader.params, ('extractor_args
', (ie_key or self.ie_key()).lower(), key))
3796 return [] if default is NO_DEFAULT else default
3797 return list(val) if casesense else [x.lower() for x in val]
3799 def _yes_playlist(self, playlist_id, video_id, smuggled_data=None, *, playlist_label='playlist
', video_label='video
'):
3800 if not playlist_id or not video_id:
3803 no_playlist = (smuggled_data or {}).get('force_noplaylist
')
3804 if no_playlist is not None:
3805 return not no_playlist
3807 video_id = '' if video_id is True else f' {video_id}
'
3808 playlist_id = '' if playlist_id is True else f' {playlist_id}
'
3809 if self.get_param('noplaylist
'):
3810 self.to_screen(f'Downloading just the {video_label}{video_id} because of
--no
-playlist
')
3812 self.to_screen(f'Downloading {playlist_label}{playlist_id}
- add
--no
-playlist to download just the {video_label}{video_id}
')
3816 class SearchInfoExtractor(InfoExtractor):
3818 Base class for paged search queries extractors.
3819 They accept URLs in the format _SEARCH_KEY(|all|[0-9]):{query}
3820 Instances should define _SEARCH_KEY and optionally _MAX_RESULTS
3823 _MAX_RESULTS = float('inf
')
3826 def _make_valid_url(cls):
3827 return r'%s(?P
<prefix
>|
[1-9][0-9]*|all
):(?P
<query
>[\s\S
]+)' % cls._SEARCH_KEY
3829 def _real_extract(self, query):
3830 prefix, query = self._match_valid_url(query).group('prefix
', 'query
')
3832 return self._get_n_results(query, 1)
3833 elif prefix == 'all
':
3834 return self._get_n_results(query, self._MAX_RESULTS)
3838 raise ExtractorError(f'invalid download number {n}
for query
"{query}"')
3839 elif n > self._MAX_RESULTS:
3840 self.report_warning('%s returns
max %i results (you requested
%i)' % (self._SEARCH_KEY, self._MAX_RESULTS, n))
3841 n = self._MAX_RESULTS
3842 return self._get_n_results(query, n)
3844 def _get_n_results(self, query, n):
3845 """Get a specified number of results for a query.
3846 Either this function or _search_results must be overridden by subclasses """
3847 return self.playlist_result(
3848 itertools.islice(self._search_results(query), 0, None if n == float('inf
') else n),
3851 def _search_results(self, query):
3852 """Returns an iterator of search results"""
3853 raise NotImplementedError('This method must be implemented by subclasses
')
3856 def SEARCH_KEY(cls):
3857 return cls._SEARCH_KEY