12 import xml
.etree
.ElementTree
14 from ..compat
import functools
, re
# isort: split
15 from ..compat
import (
16 compat_cookiejar_Cookie
,
17 compat_cookies_SimpleCookie
,
18 compat_etree_fromstring
,
25 compat_urllib_parse_unquote
,
26 compat_urllib_parse_urlencode
,
27 compat_urllib_request
,
30 from ..downloader
import FileDownloader
31 from ..downloader
.f4m
import get_base_url
, remove_encrypted_media
66 parse_m3u8_attributes
,
91 """Information Extractor class.
93 Information extractors are the classes that, given a URL, extract
94 information about the video (or videos) the URL refers to. This
95 information includes the real video URL, the video title, author and
96 others. The information is stored in a dictionary which is then
97 passed to the YoutubeDL. The YoutubeDL processes this
98 information possibly downloading the video to the file system, among
99 other possible outcomes.
101 The type field determines the type of the result.
102 By far the most common value (and the default if _type is missing) is
103 "video", which indicates a single video.
105 For a video, the dictionaries must include the following fields:
107 id: Video identifier.
108 title: Video title, unescaped. Set to an empty string if video has
109 no title as opposed to "None" which signifies that the
110 extractor failed to obtain a title
112 Additionally, it must contain either a formats entry or a url one:
114 formats: A list of dictionaries for each format available, ordered
115 from worst to best quality.
118 * url The mandatory URL representing the media:
119 for plain file media - HTTP URL of this file,
121 for HLS - URL of the M3U8 media playlist,
122 for HDS - URL of the F4M manifest,
124 - HTTP URL to plain file media (in case of
126 - URL of the MPD manifest or base URL
127 representing the media if MPD manifest
128 is parsed from a string (in case of
130 for MSS - URL of the ISM manifest.
132 The URL of the manifest file in case of
134 for HLS - URL of the M3U8 master playlist,
135 for HDS - URL of the F4M manifest,
136 for DASH - URL of the MPD manifest,
137 for MSS - URL of the ISM manifest.
138 * manifest_stream_number (For internal use only)
139 The index of the stream in the manifest file
140 * ext Will be calculated from URL if missing
141 * format A human-readable description of the format
142 ("mp4 container with h264/opus").
143 Calculated from the format_id, width, height.
144 and format_note fields if missing.
145 * format_id A short description of the format
146 ("mp4_h264_opus" or "19").
147 Technically optional, but strongly recommended.
148 * format_note Additional info about the format
149 ("3D" or "DASH video")
150 * width Width of the video, if known
151 * height Height of the video, if known
152 * resolution Textual description of width and height
153 * dynamic_range The dynamic range of the video. One of:
154 "SDR" (None), "HDR10", "HDR10+, "HDR12", "HLG, "DV"
155 * tbr Average bitrate of audio and video in KBit/s
156 * abr Average audio bitrate in KBit/s
157 * acodec Name of the audio codec in use
158 * asr Audio sampling rate in Hertz
159 * vbr Average video bitrate in KBit/s
161 * vcodec Name of the video codec in use
162 * container Name of the container format
163 * filesize The number of bytes, if known in advance
164 * filesize_approx An estimate for the number of bytes
165 * player_url SWF Player URL (used for rtmpdump).
166 * protocol The protocol that will be used for the actual
167 download, lower-case. One of "http", "https" or
168 one of the protocols defined in downloader.PROTOCOL_MAP
170 Base URL for fragments. Each fragment's path
171 value (if present) will be relative to
173 * fragments A list of fragments of a fragmented media.
174 Each fragment entry must contain either an url
175 or a path. If an url is present it should be
176 considered by a client. Otherwise both path and
177 fragment_base_url must be present. Here is
178 the list of all potential fields:
179 * "url" - fragment's URL
180 * "path" - fragment's path relative to
182 * "duration" (optional, int or float)
183 * "filesize" (optional, int)
184 * is_from_start Is a live format that can be downloaded
185 from the start. Boolean
186 * preference Order number of this format. If this field is
187 present and not None, the formats get sorted
188 by this field, regardless of all other values.
189 -1 for default (order by other properties),
190 -2 or smaller for less than default.
191 < -1000 to hide the format (if there is
192 another one which is strictly better)
193 * language Language code, e.g. "de" or "en-US".
194 * language_preference Is this in the language mentioned in
196 10 if it's what the URL is about,
197 -1 for default (don't know),
198 -10 otherwise, other values reserved for now.
199 * quality Order number of the video quality of this
200 format, irrespective of the file format.
201 -1 for default (order by other properties),
202 -2 or smaller for less than default.
203 * source_preference Order number for this video source
204 (quality takes higher priority)
205 -1 for default (order by other properties),
206 -2 or smaller for less than default.
207 * http_headers A dictionary of additional HTTP headers
208 to add to the request.
209 * stretched_ratio If given and not 1, indicates that the
210 video's pixels are not square.
211 width : height ratio as float.
212 * no_resume The server does not support resuming the
213 (HTTP or RTMP) download. Boolean.
214 * has_drm The format has DRM and cannot be downloaded. Boolean
215 * downloader_options A dictionary of downloader options
216 (For internal use only)
217 * http_chunk_size Chunk size for HTTP downloads
218 * ffmpeg_args Extra arguments for ffmpeg downloader
219 RTMP formats can also have the additional fields: page_url,
220 app, play_path, tc_url, flash_version, rtmp_live, rtmp_conn,
221 rtmp_protocol, rtmp_real_time
223 url: Final video URL.
224 ext: Video filename extension.
225 format: The video format, defaults to ext (used for --get-format)
226 player_url: SWF Player URL (used for rtmpdump).
228 The following fields are optional:
230 direct: True if a direct video file was given (must only be set by GenericIE)
231 alt_title: A secondary title of the video.
232 display_id An alternative identifier for the video, not necessarily
233 unique, but available before title. Typically, id is
234 something like "4234987", title "Dancing naked mole rats",
235 and display_id "dancing-naked-mole-rats"
236 thumbnails: A list of dictionaries, with the following entries:
237 * "id" (optional, string) - Thumbnail format ID
239 * "preference" (optional, int) - quality of the image
240 * "width" (optional, int)
241 * "height" (optional, int)
242 * "resolution" (optional, string "{width}x{height}",
244 * "filesize" (optional, int)
245 * "http_headers" (dict) - HTTP headers for the request
246 thumbnail: Full URL to a video thumbnail image.
247 description: Full video description.
248 uploader: Full name of the video uploader.
249 license: License name the video is licensed under.
250 creator: The creator of the video.
251 timestamp: UNIX timestamp of the moment the video was uploaded
252 upload_date: Video upload date in UTC (YYYYMMDD).
253 If not explicitly set, calculated from timestamp
254 release_timestamp: UNIX timestamp of the moment the video was released.
255 If it is not clear whether to use timestamp or this, use the former
256 release_date: The date (YYYYMMDD) when the video was released in UTC.
257 If not explicitly set, calculated from release_timestamp
258 modified_timestamp: UNIX timestamp of the moment the video was last modified.
259 modified_date: The date (YYYYMMDD) when the video was last modified in UTC.
260 If not explicitly set, calculated from modified_timestamp
261 uploader_id: Nickname or id of the video uploader.
262 uploader_url: Full URL to a personal webpage of the video uploader.
263 channel: Full name of the channel the video is uploaded on.
264 Note that channel fields may or may not repeat uploader
265 fields. This depends on a particular extractor.
266 channel_id: Id of the channel.
267 channel_url: Full URL to a channel webpage.
268 channel_follower_count: Number of followers of the channel.
269 location: Physical location where the video was filmed.
270 subtitles: The available subtitles as a dictionary in the format
271 {tag: subformats}. "tag" is usually a language code, and
272 "subformats" is a list sorted from lower to higher
273 preference, each element is a dictionary with the "ext"
275 * "data": The subtitles file contents
276 * "url": A URL pointing to the subtitles file
277 It can optionally also have:
278 * "name": Name or description of the subtitles
279 * "http_headers": A dictionary of additional HTTP headers
280 to add to the request.
281 "ext" will be calculated from URL if missing
282 automatic_captions: Like 'subtitles'; contains automatically generated
283 captions instead of normal subtitles
284 duration: Length of the video in seconds, as an integer or float.
285 view_count: How many users have watched the video on the platform.
286 like_count: Number of positive ratings of the video
287 dislike_count: Number of negative ratings of the video
288 repost_count: Number of reposts of the video
289 average_rating: Average rating give by users, the scale used depends on the webpage
290 comment_count: Number of comments on the video
291 comments: A list of comments, each with one or more of the following
292 properties (all but one of text or html optional):
293 * "author" - human-readable name of the comment author
294 * "author_id" - user ID of the comment author
295 * "author_thumbnail" - The thumbnail of the comment author
297 * "html" - Comment as HTML
298 * "text" - Plain text of the comment
299 * "timestamp" - UNIX timestamp of comment
300 * "parent" - ID of the comment this one is replying to.
301 Set to "root" to indicate that this is a
302 comment to the original video.
303 * "like_count" - Number of positive ratings of the comment
304 * "dislike_count" - Number of negative ratings of the comment
305 * "is_favorited" - Whether the comment is marked as
306 favorite by the video uploader
307 * "author_is_uploader" - Whether the comment is made by
309 age_limit: Age restriction for the video, as an integer (years)
310 webpage_url: The URL to the video webpage, if given to yt-dlp it
311 should allow to get the same result again. (It will be set
312 by YoutubeDL if it's missing)
313 categories: A list of categories that the video falls in, for example
315 tags: A list of tags assigned to the video, e.g. ["sweden", "pop music"]
316 cast: A list of the video cast
317 is_live: True, False, or None (=unknown). Whether this video is a
318 live stream that goes on instead of a fixed-length video.
319 was_live: True, False, or None (=unknown). Whether this video was
320 originally a live stream.
321 live_status: 'is_live', 'is_upcoming', 'was_live', 'not_live' or None (=unknown)
322 If absent, automatically set from is_live, was_live
323 start_time: Time in seconds where the reproduction should start, as
324 specified in the URL.
325 end_time: Time in seconds where the reproduction should end, as
326 specified in the URL.
327 chapters: A list of dictionaries, with the following entries:
328 * "start_time" - The start time of the chapter in seconds
329 * "end_time" - The end time of the chapter in seconds
330 * "title" (optional, string)
331 playable_in_embed: Whether this video is allowed to play in embedded
332 players on other sites. Can be True (=always allowed),
333 False (=never allowed), None (=unknown), or a string
334 specifying the criteria for embedability (Eg: 'whitelist')
335 availability: Under what condition the video is available. One of
336 'private', 'premium_only', 'subscriber_only', 'needs_auth',
337 'unlisted' or 'public'. Use 'InfoExtractor._availability'
339 __post_extractor: A function to be called just before the metadata is
340 written to either disk, logger or console. The function
341 must return a dict which will be added to the info_dict.
342 This is usefull for additional information that is
343 time-consuming to extract. Note that the fields thus
344 extracted will not be available to output template and
345 match_filter. So, only "comments" and "comment_count" are
346 currently allowed to be extracted via this method.
348 The following fields should only be used when the video belongs to some logical
351 chapter: Name or title of the chapter the video belongs to.
352 chapter_number: Number of the chapter the video belongs to, as an integer.
353 chapter_id: Id of the chapter the video belongs to, as a unicode string.
355 The following fields should only be used when the video is an episode of some
356 series, programme or podcast:
358 series: Title of the series or programme the video episode belongs to.
359 series_id: Id of the series or programme the video episode belongs to, as a unicode string.
360 season: Title of the season the video episode belongs to.
361 season_number: Number of the season the video episode belongs to, as an integer.
362 season_id: Id of the season the video episode belongs to, as a unicode string.
363 episode: Title of the video episode. Unlike mandatory video title field,
364 this field should denote the exact title of the video episode
365 without any kind of decoration.
366 episode_number: Number of the video episode within a season, as an integer.
367 episode_id: Id of the video episode, as a unicode string.
369 The following fields should only be used when the media is a track or a part of
372 track: Title of the track.
373 track_number: Number of the track within an album or a disc, as an integer.
374 track_id: Id of the track (useful in case of custom indexing, e.g. 6.iii),
376 artist: Artist(s) of the track.
377 genre: Genre(s) of the track.
378 album: Title of the album the track belongs to.
379 album_type: Type of the album (e.g. "Demo", "Full-length", "Split", "Compilation", etc).
380 album_artist: List of all artists appeared on the album (e.g.
381 "Ash Borer / Fell Voices" or "Various Artists", useful for splits
383 disc_number: Number of the disc or other physical medium the track belongs to,
385 release_year: Year (YYYY) when the album was released.
386 composer: Composer of the piece
388 Unless mentioned otherwise, the fields should be Unicode strings.
390 Unless mentioned otherwise, None is equivalent to absence of information.
393 _type "playlist" indicates multiple videos.
394 There must be a key "entries", which is a list, an iterable, or a PagedList
395 object, each element of which is a valid dictionary by this specification.
397 Additionally, playlists can have "id", "title", and any other relevent
398 attributes with the same semantics as videos (see above).
400 It can also have the following optional fields:
402 playlist_count: The total number of videos in a playlist. If not given,
403 YoutubeDL tries to calculate it from "entries"
406 _type "multi_video" indicates that there are multiple videos that
407 form a single show, for examples multiple acts of an opera or TV episode.
408 It must have an entries key like a playlist and contain all the keys
409 required for a video at the same time.
412 _type "url" indicates that the video must be extracted from another
413 location, possibly by a different extractor. Its only required key is:
414 "url" - the next URL to extract.
415 The key "ie_key" can be set to the class name (minus the trailing "IE",
416 e.g. "Youtube") if the extractor class is known in advance.
417 Additionally, the dictionary may have any properties of the resolved entity
418 known in advance, for example "title" if the title of the referred video is
422 _type "url_transparent" entities have the same specification as "url", but
423 indicate that the given additional information is more precise than the one
424 associated with the resolved URL.
425 This is useful when a site employs a video service that hosts the video and
426 its technical metadata, but that video service does not embed a useful
427 title, description etc.
430 Subclasses of this should define a _VALID_URL regexp and, re-define the
431 _real_extract() and (optionally) _real_initialize() methods.
432 Probably, they should also be added to the list of extractors.
434 Subclasses may also override suitable() if necessary, but ensure the function
435 signature is preserved and that this function imports everything it needs
436 (except other extractors), so that lazy_extractors works correctly.
438 To support username + password (or netrc) login, the extractor must define a
439 _NETRC_MACHINE and re-define _perform_login(username, password) and
440 (optionally) _initialize_pre_login() methods. The _perform_login method will
441 be called between _initialize_pre_login and _real_initialize if credentials
442 are passed by the user. In cases where it is necessary to have the login
443 process as part of the extraction rather than initialization, _perform_login
444 can be left undefined.
446 _GEO_BYPASS attribute may be set to False in order to disable
447 geo restriction bypass mechanisms for a particular extractor.
448 Though it won't disable explicit geo restriction bypass based on
449 country code provided with geo_bypass_country.
451 _GEO_COUNTRIES attribute may contain a list of presumably geo unrestricted
452 countries for this extractor. One of these countries will be used by
453 geo restriction bypass mechanism right away in order to bypass
454 geo restriction, of course, if the mechanism is not disabled.
456 _GEO_IP_BLOCKS attribute may contain a list of presumably geo unrestricted
457 IP blocks in CIDR notation for this extractor. One of these IP blocks
458 will be used by geo restriction bypass mechanism similarly
461 The _WORKING attribute should be set to False for broken IEs
462 in order to warn the users and skip the tests.
467 _x_forwarded_for_ip
= None
469 _GEO_COUNTRIES
= None
470 _GEO_IP_BLOCKS
= None
472 _NETRC_MACHINE
= None
476 def _login_hint(self
, method
=NO_DEFAULT
, netrc
=None):
477 password_hint
= f
'--username and --password, or --netrc ({netrc or self._NETRC_MACHINE}) to provide account credentials'
480 'any': f
'Use --cookies, --cookies-from-browser, {password_hint}',
481 'password': f
'Use {password_hint}',
483 'Use --cookies-from-browser or --cookies for the authentication. '
484 'See https://github.com/ytdl-org/youtube-dl#how-do-i-pass-cookies-to-youtube-dl for how to manually pass cookies'),
485 }[method
if method
is not NO_DEFAULT
else 'any' if self
.supports_login() else 'cookies']
487 def __init__(self
, downloader
=None):
488 """Constructor. Receives an optional downloader (a YoutubeDL instance).
489 If a downloader is not passed during initialization,
490 it must be set using "set_downloader()" before "extract()" is called"""
492 self
._x
_forwarded
_for
_ip
= None
493 self
._printed
_messages
= set()
494 self
.set_downloader(downloader
)
497 def _match_valid_url(cls
, url
):
498 # This does not use has/getattr intentionally - we want to know whether
499 # we have cached the regexp for *this* class, whereas getattr would also
500 # match the superclass
501 if '_VALID_URL_RE' not in cls
.__dict
__:
502 if '_VALID_URL' not in cls
.__dict
__:
503 cls
._VALID
_URL
= cls
._make
_valid
_url
()
504 cls
._VALID
_URL
_RE
= re
.compile(cls
._VALID
_URL
)
505 return cls
._VALID
_URL
_RE
.match(url
)
508 def suitable(cls
, url
):
509 """Receives a URL and returns True if suitable for this IE."""
510 # This function must import everything it needs (except other extractors),
511 # so that lazy_extractors works correctly
512 return cls
._match
_valid
_url
(url
) is not None
515 def _match_id(cls
, url
):
516 return cls
._match
_valid
_url
(url
).group('id')
519 def get_temp_id(cls
, url
):
521 return cls
._match
_id
(url
)
522 except (IndexError, AttributeError):
527 """Getter method for _WORKING."""
531 def supports_login(cls
):
532 return bool(cls
._NETRC
_MACHINE
)
534 def initialize(self
):
535 """Initializes an instance (authentication, etc)."""
536 self
._printed
_messages
= set()
537 self
._initialize
_geo
_bypass
({
538 'countries': self
._GEO
_COUNTRIES
,
539 'ip_blocks': self
._GEO
_IP
_BLOCKS
,
542 self
._initialize
_pre
_login
()
543 if self
.supports_login():
544 username
, password
= self
._get
_login
_info
()
546 self
._perform
_login
(username
, password
)
547 elif self
.get_param('username') and False not in (self
.IE_DESC
, self
._NETRC
_MACHINE
):
548 self
.report_warning(f
'Login with password is not supported for this website. {self._login_hint("cookies")}')
549 self
._real
_initialize
()
552 def _initialize_geo_bypass(self
, geo_bypass_context
):
554 Initialize geo restriction bypass mechanism.
556 This method is used to initialize geo bypass mechanism based on faking
557 X-Forwarded-For HTTP header. A random country from provided country list
558 is selected and a random IP belonging to this country is generated. This
559 IP will be passed as X-Forwarded-For HTTP header in all subsequent
562 This method will be used for initial geo bypass mechanism initialization
563 during the instance initialization with _GEO_COUNTRIES and
566 You may also manually call it from extractor's code if geo bypass
567 information is not available beforehand (e.g. obtained during
568 extraction) or due to some other reason. In this case you should pass
569 this information in geo bypass context passed as first argument. It may
570 contain following fields:
572 countries: List of geo unrestricted countries (similar
574 ip_blocks: List of geo unrestricted IP blocks in CIDR notation
575 (similar to _GEO_IP_BLOCKS)
578 if not self
._x
_forwarded
_for
_ip
:
580 # Geo bypass mechanism is explicitly disabled by user
581 if not self
.get_param('geo_bypass', True):
584 if not geo_bypass_context
:
585 geo_bypass_context
= {}
587 # Backward compatibility: previously _initialize_geo_bypass
588 # expected a list of countries, some 3rd party code may still use
590 if isinstance(geo_bypass_context
, (list, tuple)):
591 geo_bypass_context
= {
592 'countries': geo_bypass_context
,
595 # The whole point of geo bypass mechanism is to fake IP
596 # as X-Forwarded-For HTTP header based on some IP block or
599 # Path 1: bypassing based on IP block in CIDR notation
601 # Explicit IP block specified by user, use it right away
602 # regardless of whether extractor is geo bypassable or not
603 ip_block
= self
.get_param('geo_bypass_ip_block', None)
605 # Otherwise use random IP block from geo bypass context but only
606 # if extractor is known as geo bypassable
608 ip_blocks
= geo_bypass_context
.get('ip_blocks')
609 if self
._GEO
_BYPASS
and ip_blocks
:
610 ip_block
= random
.choice(ip_blocks
)
613 self
._x
_forwarded
_for
_ip
= GeoUtils
.random_ipv4(ip_block
)
614 self
.write_debug(f
'Using fake IP {self._x_forwarded_for_ip} as X-Forwarded-For')
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 _create_request(self
, url_or_request
, data
=None, headers
={}, query={}
):
729 if isinstance(url_or_request
, compat_urllib_request
.Request
):
730 return update_Request(url_or_request
, data
=data
, headers
=headers
, query
=query
)
732 url_or_request
= update_url_query(url_or_request
, query
)
733 return sanitized_Request(url_or_request
, data
, headers
)
735 def _request_webpage(self
, url_or_request
, video_id
, note
=None, errnote
=None, fatal
=True, data
=None, headers
={}, query={}
, expected_status
=None):
737 Return the response handle.
739 See _download_webpage docstring for arguments specification.
741 if not self
._downloader
._first
_webpage
_request
:
742 sleep_interval
= self
.get_param('sleep_interval_requests') or 0
743 if sleep_interval
> 0:
744 self
.to_screen('Sleeping %s seconds ...' % sleep_interval
)
745 time
.sleep(sleep_interval
)
747 self
._downloader
._first
_webpage
_request
= False
750 self
.report_download_webpage(video_id
)
751 elif note
is not False:
753 self
.to_screen(str(note
))
755 self
.to_screen(f
'{video_id}: {note}')
757 # Some sites check X-Forwarded-For HTTP header in order to figure out
758 # the origin of the client behind proxy. This allows bypassing geo
759 # restriction by faking this header's value to IP that belongs to some
760 # geo unrestricted country. We will do so once we encounter any
761 # geo restriction error.
762 if self
._x
_forwarded
_for
_ip
:
763 if 'X-Forwarded-For' not in headers
:
764 headers
['X-Forwarded-For'] = self
._x
_forwarded
_for
_ip
767 return self
._downloader
.urlopen(self
._create
_request
(url_or_request
, data
, headers
, query
))
768 except network_exceptions
as err
:
769 if isinstance(err
, compat_urllib_error
.HTTPError
):
770 if self
.__can
_accept
_status
_code
(err
, expected_status
):
771 # Retain reference to error to prevent file object from
772 # being closed before it can be read. Works around the
773 # effects of <https://bugs.python.org/issue15002>
774 # introduced in Python 3.4.1.
781 errnote
= 'Unable to download webpage'
783 errmsg
= f
'{errnote}: {error_to_compat_str(err)}'
785 raise ExtractorError(errmsg
, cause
=err
)
787 self
.report_warning(errmsg
)
790 def _download_webpage_handle(self
, url_or_request
, video_id
, note
=None, errnote
=None, fatal
=True,
791 encoding
=None, data
=None, headers
={}, query={}
, expected_status
=None):
793 Return a tuple (page content as string, URL handle).
796 url_or_request -- plain text URL as a string or
797 a compat_urllib_request.Requestobject
798 video_id -- Video/playlist/item identifier (string)
801 note -- note printed before downloading (string)
802 errnote -- note printed in case of an error (string)
803 fatal -- flag denoting whether error should be considered fatal,
804 i.e. whether it should cause ExtractionError to be raised,
805 otherwise a warning will be reported and extraction continued
806 encoding -- encoding for a page content decoding, guessed automatically
807 when not explicitly specified
808 data -- POST data (bytes)
809 headers -- HTTP headers (dict)
810 query -- URL query (dict)
811 expected_status -- allows to accept failed HTTP requests (non 2xx
812 status code) by explicitly specifying a set of accepted status
813 codes. Can be any of the following entities:
814 - an integer type specifying an exact failed status code to
816 - a list or a tuple of integer types specifying a list of
817 failed status codes to accept
818 - a callable accepting an actual failed status code and
819 returning True if it should be accepted
820 Note that this argument does not affect success status codes (2xx)
821 which are always accepted.
824 # Strip hashes from the URL (#1038)
825 if isinstance(url_or_request
, (compat_str
, str)):
826 url_or_request
= url_or_request
.partition('#')[0]
828 urlh
= self
._request
_webpage
(url_or_request
, video_id
, note
, errnote
, fatal
, data
=data
, headers
=headers
, query
=query
, expected_status
=expected_status
)
832 content
= self
._webpage
_read
_content
(urlh
, url_or_request
, video_id
, note
, errnote
, fatal
, encoding
=encoding
)
833 return (content
, urlh
)
836 def _guess_encoding_from_content(content_type
, webpage_bytes
):
837 m
= re
.match(r
'[a-zA-Z0-9_.-]+/[a-zA-Z0-9_.-]+\s*;\s*charset=(.+)', content_type
)
839 encoding
= m
.group(1)
841 m
= re
.search(br
'<meta[^>]+charset=[\'"]?([^\'")]+)[ /\'">]',
842 webpage_bytes[:1024])
844 encoding = m.group(1).decode('ascii')
845 elif webpage_bytes.startswith(b'\xff\xfe'):
852 def __check_blocked(self, content):
853 first_block = content[:512]
854 if ('<title>Access to this site is blocked</title>' in content
855 and 'Websense' in first_block):
856 msg = 'Access to this webpage has been blocked by Websense filtering software in your network.'
857 blocked_iframe = self._html_search_regex(
858 r'<iframe src="([^
"]+)"', content,
859 'Websense information URL
', default=None)
861 msg += ' Visit
%s for more details
' % blocked_iframe
862 raise ExtractorError(msg, expected=True)
863 if '<title
>The URL you requested has been blocked
</title
>' in first_block:
865 'Access to this webpage has been blocked by Indian censorship
. '
866 'Use a VPN
or proxy
server (with --proxy
) to route around it
.')
867 block_msg = self._html_search_regex(
868 r'</h1
><p
>(.*?
)</p
>',
869 content, 'block message
', default=None)
871 msg += ' (Message
: "%s")' % block_msg.replace('\n', ' ')
872 raise ExtractorError(msg, expected=True)
873 if ('<title
>TTK
:: Доступ к ресурсу ограничен
</title
>' in content
874 and 'blocklist
.rkn
.gov
.ru
' in content):
875 raise ExtractorError(
876 'Access to this webpage has been blocked by decision of the Russian government
. '
877 'Visit http
://blocklist
.rkn
.gov
.ru
/ for a block reason
.',
880 def _request_dump_filename(self, url, video_id):
881 basen = f'{video_id}_{url}
'
882 trim_length = self.get_param('trim_file_name
') or 240
883 if len(basen) > trim_length:
884 h = '___
' + hashlib.md5(basen.encode('utf
-8')).hexdigest()
885 basen = basen[:trim_length - len(h)] + h
886 filename = sanitize_filename(f'{basen}
.dump
', restricted=True)
887 # Working around MAX_PATH limitation on Windows (see
888 # http://msdn.microsoft.com/en-us/library/windows/desktop/aa365247(v=vs.85).aspx)
889 if compat_os_name == 'nt
':
890 absfilepath = os.path.abspath(filename)
891 if len(absfilepath) > 259:
892 filename = fR'\\?\{absfilepath}
'
895 def __decode_webpage(self, webpage_bytes, encoding, headers):
897 encoding = self._guess_encoding_from_content(headers.get('Content
-Type
', ''), webpage_bytes)
899 return webpage_bytes.decode(encoding, 'replace
')
901 return webpage_bytes.decode('utf
-8', 'replace
')
903 def _webpage_read_content(self, urlh, url_or_request, video_id, note=None, errnote=None, fatal=True, prefix=None, encoding=None):
904 webpage_bytes = urlh.read()
905 if prefix is not None:
906 webpage_bytes = prefix + webpage_bytes
907 if self.get_param('dump_intermediate_pages
', False):
908 self.to_screen('Dumping request to
' + urlh.geturl())
909 dump = base64.b64encode(webpage_bytes).decode('ascii
')
910 self._downloader.to_screen(dump)
911 if self.get_param('write_pages
'):
912 filename = self._request_dump_filename(urlh.geturl(), video_id)
913 self.to_screen(f'Saving request to {filename}
')
914 with open(filename, 'wb
') as outf:
915 outf.write(webpage_bytes)
917 content = self.__decode_webpage(webpage_bytes, encoding, urlh.headers)
918 self.__check_blocked(content)
922 def _parse_xml(self, xml_string, video_id, transform_source=None, fatal=True):
924 xml_string = transform_source(xml_string)
926 return compat_etree_fromstring(xml_string.encode('utf
-8'))
927 except xml.etree.ElementTree.ParseError as ve:
928 errmsg = '%s: Failed to parse XML
' % video_id
930 raise ExtractorError(errmsg, cause=ve)
932 self.report_warning(errmsg + str(ve))
934 def _parse_json(self, json_string, video_id, transform_source=None, fatal=True, **parser_kwargs):
937 json_string, cls=LenientJSONDecoder, strict=False, transform_source=transform_source, **parser_kwargs)
938 except ValueError as ve:
939 errmsg = f'{video_id}
: Failed to parse JSON
'
941 raise ExtractorError(errmsg, cause=ve)
943 self.report_warning(f'{errmsg}
: {ve}
')
945 def _parse_socket_response_as_json(self, data, video_id, transform_source=None, fatal=True):
946 return self._parse_json(
947 data[data.find('{'):data.rfind('}
') + 1],
948 video_id, transform_source, fatal)
950 def __create_download_methods(name, parser, note, errnote, return_value):
952 def parse(ie, content, *args, **kwargs):
955 # parser is fetched by name so subclasses can override it
956 return getattr(ie, parser)(content, *args, **kwargs)
958 def download_handle(self, url_or_request, video_id, note=note, errnote=errnote, transform_source=None,
959 fatal=True, encoding=None, data=None, headers={}, query={}, expected_status=None):
960 res = self._download_webpage_handle(
961 url_or_request, video_id, note=note, errnote=errnote, fatal=fatal, encoding=encoding,
962 data=data, headers=headers, query=query, expected_status=expected_status)
966 return parse(self, content, video_id, transform_source=transform_source, fatal=fatal), urlh
968 def download_content(self, url_or_request, video_id, note=note, errnote=errnote, transform_source=None,
969 fatal=True, encoding=None, data=None, headers={}, query={}, expected_status=None):
970 if self.get_param('load_pages
'):
971 url_or_request = self._create_request(url_or_request, data, headers, query)
972 filename = self._request_dump_filename(url_or_request.full_url, video_id)
973 self.to_screen(f'Loading request
from {filename}
')
975 with open(filename, 'rb
') as dumpf:
976 webpage_bytes = dumpf.read()
978 self.report_warning(f'Unable to load request
from disk
: {e}
')
980 content = self.__decode_webpage(webpage_bytes, encoding, url_or_request.headers)
981 return parse(self, content, video_id, transform_source, fatal)
985 'transform_source
': transform_source,
987 'encoding
': encoding,
991 'expected_status
': expected_status,
994 kwargs.pop('transform_source
')
995 # The method is fetched by name so subclasses can override _download_..._handle
996 res = getattr(self, download_handle.__name__)(url_or_request, video_id, **kwargs)
997 return res if res is False else res[0]
999 def impersonate(func, name, return_value):
1000 func.__name__, func.__qualname__ = name, f'InfoExtractor
.{name}
'
1002 @param transform_source Apply this transformation before parsing
1003 @returns {return_value}
1005 See _download_webpage_handle docstring for other arguments specification
1008 impersonate(download_handle, f'_download_{name}_handle
', f'({return_value}
, URL handle
)')
1009 impersonate(download_content, f'_download_{name}
', f'{return_value}
')
1010 return download_handle, download_content
1012 _download_xml_handle, _download_xml = __create_download_methods(
1013 'xml
', '_parse_xml
', 'Downloading XML
', 'Unable to download XML
', 'xml
as an xml
.etree
.ElementTree
.Element
')
1014 _download_json_handle, _download_json = __create_download_methods(
1015 'json
', '_parse_json
', 'Downloading JSON metadata
', 'Unable to download JSON metadata
', 'JSON
object as a
dict')
1016 _download_socket_json_handle, _download_socket_json = __create_download_methods(
1017 'socket_json
', '_parse_socket_response_as_json
', 'Polling socket
', 'Unable to poll socket
', 'JSON
object as a
dict')
1018 __download_webpage = __create_download_methods('webpage
', None, None, None, 'data of the page
as a string
')[1]
1020 def _download_webpage(
1021 self, url_or_request, video_id, note=None, errnote=None,
1022 fatal=True, tries=1, timeout=NO_DEFAULT, *args, **kwargs):
1024 Return the data of the page as a string.
1027 tries -- number of tries
1028 timeout -- sleep interval between tries
1030 See _download_webpage_handle docstring for other arguments specification.
1033 R''' # NB: These are unused; should they be deprecated?
1035 self._downloader.deprecation_warning('tries argument
is deprecated
in InfoExtractor
._download
_webpage
')
1036 if timeout is NO_DEFAULT:
1039 self._downloader.deprecation_warning('timeout argument
is deprecated
in InfoExtractor
._download
_webpage
')
1045 return self.__download_webpage(url_or_request, video_id, note, errnote, None, fatal, *args, **kwargs)
1046 except compat_http_client.IncompleteRead as e:
1048 if try_count >= tries:
1050 self._sleep(timeout, video_id)
1052 def report_warning(self, msg, video_id=None, *args, only_once=False, **kwargs):
1053 idstr = format_field(video_id, None, '%s: ')
1054 msg = f'[{self.IE_NAME}
] {idstr}{msg}
'
1056 if f'WARNING
: {msg}
' in self._printed_messages:
1058 self._printed_messages.add(f'WARNING
: {msg}
')
1059 self._downloader.report_warning(msg, *args, **kwargs)
1061 def to_screen(self, msg, *args, **kwargs):
1062 """Print msg to screen, prefixing it with '[ie_name
]'"""
1063 self._downloader.to_screen(f'[{self.IE_NAME}
] {msg}
', *args, **kwargs)
1065 def write_debug(self, msg, *args, **kwargs):
1066 self._downloader.write_debug(f'[{self.IE_NAME}
] {msg}
', *args, **kwargs)
1068 def get_param(self, name, default=None, *args, **kwargs):
1069 if self._downloader:
1070 return self._downloader.params.get(name, default, *args, **kwargs)
1073 def report_drm(self, video_id, partial=False):
1074 self.raise_no_formats('This video
is DRM protected
', expected=True, video_id=video_id)
1076 def report_extraction(self, id_or_name):
1077 """Report information extraction."""
1078 self.to_screen('%s: Extracting information
' % id_or_name)
1080 def report_download_webpage(self, video_id):
1081 """Report webpage download."""
1082 self.to_screen('%s: Downloading webpage
' % video_id)
1084 def report_age_confirmation(self):
1085 """Report attempt to confirm age."""
1086 self.to_screen('Confirming age
')
1088 def report_login(self):
1089 """Report attempt to log in."""
1090 self.to_screen('Logging
in')
1092 def raise_login_required(
1093 self, msg='This video
is only available
for registered users
',
1094 metadata_available=False, method=NO_DEFAULT):
1095 if metadata_available and (
1096 self.get_param('ignore_no_formats_error
') or self.get_param('wait_for_video
')):
1097 self.report_warning(msg)
1099 msg += format_field(self._login_hint(method), None, '. %s')
1100 raise ExtractorError(msg, expected=True)
1102 def raise_geo_restricted(
1103 self, msg='This video
is not available
from your location due to geo restriction
',
1104 countries=None, metadata_available=False):
1105 if metadata_available and (
1106 self.get_param('ignore_no_formats_error
') or self.get_param('wait_for_video
')):
1107 self.report_warning(msg)
1109 raise GeoRestrictedError(msg, countries=countries)
1111 def raise_no_formats(self, msg, expected=False, video_id=None):
1113 self.get_param('ignore_no_formats_error
') or self.get_param('wait_for_video
')):
1114 self.report_warning(msg, video_id)
1115 elif isinstance(msg, ExtractorError):
1118 raise ExtractorError(msg, expected=expected, video_id=video_id)
1120 # Methods for following #608
1122 def url_result(url, ie=None, video_id=None, video_title=None, *, url_transparent=False, **kwargs):
1123 """Returns a URL that points to a page that should be processed"""
1125 kwargs['ie_key
'] = ie if isinstance(ie, str) else ie.ie_key()
1126 if video_id is not None:
1127 kwargs['id'] = video_id
1128 if video_title is not None:
1129 kwargs['title
'] = video_title
1132 '_type
': 'url_transparent
' if url_transparent else 'url
',
1136 def playlist_from_matches(self, matches, playlist_id=None, playlist_title=None, getter=None, ie=None, video_kwargs=None, **kwargs):
1137 urls = (self.url_result(self._proto_relative_url(m), ie, **(video_kwargs or {}))
1138 for m in orderedSet(map(getter, matches) if getter else matches))
1139 return self.playlist_result(urls, playlist_id, playlist_title, **kwargs)
1142 def playlist_result(entries, playlist_id=None, playlist_title=None, playlist_description=None, *, multi_video=False, **kwargs):
1143 """Returns a playlist"""
1145 kwargs['id'] = playlist_id
1147 kwargs['title
'] = playlist_title
1148 if playlist_description is not None:
1149 kwargs['description
'] = playlist_description
1152 '_type
': 'multi_video
' if multi_video else 'playlist
',
1156 def _search_regex(self, pattern, string, name, default=NO_DEFAULT, fatal=True, flags=0, group=None):
1158 Perform a regex search on the given string, using a single or a list of
1159 patterns returning the first matching group.
1160 In case of failure return a default value or raise a WARNING or a
1161 RegexNotFoundError, depending on fatal, specifying the field name.
1165 elif isinstance(pattern, (str, re.Pattern)):
1166 mobj = re.search(pattern, string, flags)
1169 mobj = re.search(p, string, flags)
1173 _name = self._downloader._format_err(name, self._downloader.Styles.EMPHASIS)
1177 # return the first matching group
1178 return next(g for g in mobj.groups() if g is not None)
1179 elif isinstance(group, (list, tuple)):
1180 return tuple(mobj.group(g) for g in group)
1182 return mobj.group(group)
1183 elif default is not NO_DEFAULT:
1186 raise RegexNotFoundError('Unable to extract
%s' % _name)
1188 self.report_warning('unable to extract
%s' % _name + bug_reports_message())
1191 def _search_json(self, start_pattern, string, name, video_id, *, end_pattern='', contains_pattern='(?s
:.+)', fatal=True, **kwargs):
1192 """Searches string for the JSON object specified by start_pattern"""
1193 # NB: end_pattern is only used to reduce the size of the initial match
1194 return self._parse_json(
1195 self._search_regex(rf'{start_pattern}\s
*(?P
<json
>{{{contains_pattern}
}})\s
*{end_pattern}
',
1196 string, name, group='json
', fatal=fatal) or '{}',
1197 video_id, fatal=fatal, ignore_extra=True, **kwargs) or {}
1199 def _html_search_regex(self, pattern, string, name, default=NO_DEFAULT, fatal=True, flags=0, group=None):
1201 Like _search_regex, but strips HTML tags and unescapes entities.
1203 res = self._search_regex(pattern, string, name, default, fatal, flags, group)
1205 return clean_html(res).strip()
1209 def _get_netrc_login_info(self, netrc_machine=None):
1212 netrc_machine = netrc_machine or self._NETRC_MACHINE
1214 if self.get_param('usenetrc
', False):
1216 netrc_file = compat_expanduser(self.get_param('netrc_location
') or '~
')
1217 if os.path.isdir(netrc_file):
1218 netrc_file = os.path.join(netrc_file, '.netrc
')
1219 info = netrc.netrc(file=netrc_file).authenticators(netrc_machine)
1220 if info is not None:
1224 raise netrc.NetrcParseError(
1225 'No authenticators
for %s' % netrc_machine)
1226 except (OSError, netrc.NetrcParseError) as err:
1227 self.report_warning(
1228 'parsing
.netrc
: %s' % error_to_compat_str(err))
1230 return username, password
1232 def _get_login_info(self, username_option='username
', password_option='password
', netrc_machine=None):
1234 Get the login info as (username, password)
1235 First look for the manually specified credentials using username_option
1236 and password_option as keys in params dictionary. If no such credentials
1237 available look in the netrc file using the netrc_machine or _NETRC_MACHINE
1239 If there's no info available
, return (None, None)
1242 # Attempt to use provided username and password or .netrc data
1243 username = self.get_param(username_option)
1244 if username is not None:
1245 password = self.get_param(password_option)
1247 username, password = self._get_netrc_login_info(netrc_machine)
1249 return username, password
1251 def _get_tfa_info(self, note='two-factor verification code'):
1253 Get the two
-factor authentication info
1254 TODO
- asking the user will be required
for sms
/phone verify
1255 currently just uses the command line option
1256 If there
's no info available, return None
1259 tfa = self.get_param('twofactor
')
1263 return compat_getpass('Type
%s and press
[Return
]: ' % note)
1265 # Helper functions for extracting OpenGraph info
1267 def _og_regexes(prop):
1268 content_re = r'content
=(?
:"([^"]+?
)"|\'([^\']+?)\'|\s*([^\s"\'=<>`
]+?
))'
1269 property_re = (r'(?
:name|
property)=(?
:\'og
%(sep)s%(prop)s\'|
"og%(sep)s%(prop)s"|\s
*og
%(sep)s%(prop)s\b)'
1270 % {'prop': re.escape(prop), 'sep': '(?::|[:-])'})
1271 template = r'<meta
[^
>]+?
%s[^
>]+?
%s'
1273 template % (property_re, content_re),
1274 template % (content_re, property_re),
1278 def _meta_regex(prop):
1279 return r'''(?isx)<meta
1280 (?=[^>]+(?:itemprop|name|property|id|http-equiv)=(["\']?)%s\1)
1281 [^>]+?content=(["\'])(?P<content>.*?)\2''' % re.escape(prop)
1283 def _og_search_property(self, prop, html, name=None, **kargs):
1284 prop = variadic(prop)
1286 name = 'OpenGraph
%s' % prop[0]
1289 og_regexes.extend(self._og_regexes(p))
1290 escaped = self._search_regex(og_regexes, html, name, flags=re.DOTALL, **kargs)
1293 return unescapeHTML(escaped)
1295 def _og_search_thumbnail(self, html, **kargs):
1296 return self._og_search_property('image
', html, 'thumbnail URL
', fatal=False, **kargs)
1298 def _og_search_description(self, html, **kargs):
1299 return self._og_search_property('description
', html, fatal=False, **kargs)
1301 def _og_search_title(self, html, *, fatal=False, **kargs):
1302 return self._og_search_property('title
', html, fatal=fatal, **kargs)
1304 def _og_search_video_url(self, html, name='video url
', secure=True, **kargs):
1305 regexes = self._og_regexes('video
') + self._og_regexes('video
:url
')
1307 regexes = self._og_regexes('video
:secure_url
') + regexes
1308 return self._html_search_regex(regexes, html, name, **kargs)
1310 def _og_search_url(self, html, **kargs):
1311 return self._og_search_property('url
', html, **kargs)
1313 def _html_extract_title(self, html, name='title
', *, fatal=False, **kwargs):
1314 return self._html_search_regex(r'(?s
)<title
\b[^
>]*>([^
<]+)</title
>', html, name, fatal=fatal, **kwargs)
1316 def _html_search_meta(self, name, html, display_name=None, fatal=False, **kwargs):
1317 name = variadic(name)
1318 if display_name is None:
1319 display_name = name[0]
1320 return self._html_search_regex(
1321 [self._meta_regex(n) for n in name],
1322 html, display_name, fatal=fatal, group='content
', **kwargs)
1324 def _dc_search_uploader(self, html):
1325 return self._html_search_meta('dc
.creator
', html, 'uploader
')
1327 def _rta_search(self, html):
1328 # See http://www.rtalabel.org/index.php?content=howtofaq#single
1329 if re.search(r'(?ix
)<meta\s
+name
="rating"\s
+'
1330 r' content
="RTA-5042-1996-1400-1577-RTA"',
1335 def _media_rating_search(self, html):
1336 # See http://www.tjg-designs.com/WP/metadata-code-examples-adding-metadata-to-your-web-pages/
1337 rating = self._html_search_meta('rating
', html)
1349 return RATING_TABLE.get(rating.lower())
1351 def _family_friendly_search(self, html):
1352 # See http://schema.org/VideoObject
1353 family_friendly = self._html_search_meta(
1354 'isFamilyFriendly
', html, default=None)
1356 if not family_friendly:
1365 return RATING_TABLE.get(family_friendly.lower())
1367 def _twitter_search_player(self, html):
1368 return self._html_search_meta('twitter
:player
', html,
1369 'twitter card player
')
1371 def _search_json_ld(self, html, video_id, expected_type=None, **kwargs):
1372 json_ld_list = list(re.finditer(JSON_LD_RE, html))
1373 default = kwargs.get('default
', NO_DEFAULT)
1374 # JSON-LD may be malformed and thus `fatal` should be respected.
1375 # At the same time `default` may be passed that assumes `fatal=False`
1376 # for _search_regex. Let's simulate the same behavior here
as well
.
1377 fatal
= kwargs
.get('fatal', True) if default
is NO_DEFAULT
else False
1379 for mobj
in json_ld_list
:
1380 json_ld_item
= self
._parse
_json
(
1381 mobj
.group('json_ld'), video_id
, fatal
=fatal
)
1382 if not json_ld_item
:
1384 if isinstance(json_ld_item
, dict):
1385 json_ld
.append(json_ld_item
)
1386 elif isinstance(json_ld_item
, (list, tuple)):
1387 json_ld
.extend(json_ld_item
)
1389 json_ld
= self
._json
_ld
(json_ld
, video_id
, fatal
=fatal
, expected_type
=expected_type
)
1392 if default
is not NO_DEFAULT
:
1395 raise RegexNotFoundError('Unable to extract JSON-LD')
1397 self
.report_warning('unable to extract JSON-LD %s' % bug_reports_message())
1400 def _json_ld(self
, json_ld
, video_id
, fatal
=True, expected_type
=None):
1401 if isinstance(json_ld
, compat_str
):
1402 json_ld
= self
._parse
_json
(json_ld
, video_id
, fatal
=fatal
)
1406 if not isinstance(json_ld
, (list, tuple, dict)):
1408 if isinstance(json_ld
, dict):
1411 INTERACTION_TYPE_MAP
= {
1412 'CommentAction': 'comment',
1413 'AgreeAction': 'like',
1414 'DisagreeAction': 'dislike',
1415 'LikeAction': 'like',
1416 'DislikeAction': 'dislike',
1417 'ListenAction': 'view',
1418 'WatchAction': 'view',
1419 'ViewAction': 'view',
1422 def is_type(e
, *expected_types
):
1423 type = variadic(traverse_obj(e
, '@type'))
1424 return any(x
in type for x
in expected_types
)
1426 def extract_interaction_type(e
):
1427 interaction_type
= e
.get('interactionType')
1428 if isinstance(interaction_type
, dict):
1429 interaction_type
= interaction_type
.get('@type')
1430 return str_or_none(interaction_type
)
1432 def extract_interaction_statistic(e
):
1433 interaction_statistic
= e
.get('interactionStatistic')
1434 if isinstance(interaction_statistic
, dict):
1435 interaction_statistic
= [interaction_statistic
]
1436 if not isinstance(interaction_statistic
, list):
1438 for is_e
in interaction_statistic
:
1439 if not is_type(is_e
, 'InteractionCounter'):
1441 interaction_type
= extract_interaction_type(is_e
)
1442 if not interaction_type
:
1444 # For interaction count some sites provide string instead of
1445 # an integer (as per spec) with non digit characters (e.g. ",")
1446 # so extracting count with more relaxed str_to_int
1447 interaction_count
= str_to_int(is_e
.get('userInteractionCount'))
1448 if interaction_count
is None:
1450 count_kind
= INTERACTION_TYPE_MAP
.get(interaction_type
.split('/')[-1])
1453 count_key
= '%s_count' % count_kind
1454 if info
.get(count_key
) is not None:
1456 info
[count_key
] = interaction_count
1458 def extract_chapter_information(e
):
1460 'title': part
.get('name'),
1461 'start_time': part
.get('startOffset'),
1462 'end_time': part
.get('endOffset'),
1463 } for part
in variadic(e
.get('hasPart') or []) if part
.get('@type') == 'Clip']
1464 for idx
, (last_c
, current_c
, next_c
) in enumerate(zip(
1465 [{'end_time': 0}
] + chapters
, chapters
, chapters
[1:])):
1466 current_c
['end_time'] = current_c
['end_time'] or next_c
['start_time']
1467 current_c
['start_time'] = current_c
['start_time'] or last_c
['end_time']
1468 if None in current_c
.values():
1469 self
.report_warning(f
'Chapter {idx} contains broken data. Not extracting chapters')
1472 chapters
[-1]['end_time'] = chapters
[-1]['end_time'] or info
['duration']
1473 info
['chapters'] = chapters
1475 def extract_video_object(e
):
1476 assert is_type(e
, 'VideoObject')
1477 author
= e
.get('author')
1479 'url': traverse_obj(e
, 'contentUrl', 'embedUrl', expected_type
=url_or_none
),
1480 'title': unescapeHTML(e
.get('name')),
1481 'description': unescapeHTML(e
.get('description')),
1482 'thumbnails': [{'url': url}
1483 for url
in variadic(traverse_obj(e
, 'thumbnailUrl', 'thumbnailURL'))
1484 if url_or_none(url
)],
1485 'duration': parse_duration(e
.get('duration')),
1486 'timestamp': unified_timestamp(e
.get('uploadDate')),
1487 # author can be an instance of 'Organization' or 'Person' types.
1488 # both types can have 'name' property(inherited from 'Thing' type). [1]
1489 # however some websites are using 'Text' type instead.
1490 # 1. https://schema.org/VideoObject
1491 'uploader': author
.get('name') if isinstance(author
, dict) else author
if isinstance(author
, compat_str
) else None,
1492 'filesize': int_or_none(float_or_none(e
.get('contentSize'))),
1493 'tbr': int_or_none(e
.get('bitrate')),
1494 'width': int_or_none(e
.get('width')),
1495 'height': int_or_none(e
.get('height')),
1496 'view_count': int_or_none(e
.get('interactionCount')),
1498 extract_interaction_statistic(e
)
1499 extract_chapter_information(e
)
1501 def traverse_json_ld(json_ld
, at_top_level
=True):
1503 if at_top_level
and '@context' not in e
:
1505 if at_top_level
and set(e
.keys()) == {'@context', '@graph'}
:
1506 traverse_json_ld(variadic(e
['@graph'], allowed_types
=(dict,)), at_top_level
=False)
1508 if expected_type
is not None and not is_type(e
, expected_type
):
1510 rating
= traverse_obj(e
, ('aggregateRating', 'ratingValue'), expected_type
=float_or_none
)
1511 if rating
is not None:
1512 info
['average_rating'] = rating
1513 if is_type(e
, 'TVEpisode', 'Episode'):
1514 episode_name
= unescapeHTML(e
.get('name'))
1516 'episode': episode_name
,
1517 'episode_number': int_or_none(e
.get('episodeNumber')),
1518 'description': unescapeHTML(e
.get('description')),
1520 if not info
.get('title') and episode_name
:
1521 info
['title'] = episode_name
1522 part_of_season
= e
.get('partOfSeason')
1523 if is_type(part_of_season
, 'TVSeason', 'Season', 'CreativeWorkSeason'):
1525 'season': unescapeHTML(part_of_season
.get('name')),
1526 'season_number': int_or_none(part_of_season
.get('seasonNumber')),
1528 part_of_series
= e
.get('partOfSeries') or e
.get('partOfTVSeries')
1529 if is_type(part_of_series
, 'TVSeries', 'Series', 'CreativeWorkSeries'):
1530 info
['series'] = unescapeHTML(part_of_series
.get('name'))
1531 elif is_type(e
, 'Movie'):
1533 'title': unescapeHTML(e
.get('name')),
1534 'description': unescapeHTML(e
.get('description')),
1535 'duration': parse_duration(e
.get('duration')),
1536 'timestamp': unified_timestamp(e
.get('dateCreated')),
1538 elif is_type(e
, 'Article', 'NewsArticle'):
1540 'timestamp': parse_iso8601(e
.get('datePublished')),
1541 'title': unescapeHTML(e
.get('headline')),
1542 'description': unescapeHTML(e
.get('articleBody') or e
.get('description')),
1544 if is_type(traverse_obj(e
, ('video', 0)), 'VideoObject'):
1545 extract_video_object(e
['video'][0])
1546 elif is_type(traverse_obj(e
, ('subjectOf', 0)), 'VideoObject'):
1547 extract_video_object(e
['subjectOf'][0])
1548 elif is_type(e
, 'VideoObject'):
1549 extract_video_object(e
)
1550 if expected_type
is None:
1554 video
= e
.get('video')
1555 if is_type(video
, 'VideoObject'):
1556 extract_video_object(video
)
1557 if expected_type
is None:
1561 traverse_json_ld(json_ld
)
1563 return filter_dict(info
)
1565 def _search_nextjs_data(self
, webpage
, video_id
, *, transform_source
=None, fatal
=True, **kw
):
1566 return self
._parse
_json
(
1568 r
'(?s)<script[^>]+id=[\'"]__NEXT_DATA__[\'"][^
>]*>([^
<]+)</script
>',
1569 webpage, 'next
.js data
', fatal=fatal, **kw),
1570 video_id, transform_source=transform_source, fatal=fatal)
1572 def _search_nuxt_data(self, webpage, video_id, context_name='__NUXT__
', return_full_data=False):
1573 ''' Parses Nuxt.js metadata. This works as long as the function __NUXT__ invokes is a pure function. '''
1574 # not all website do this, but it can be changed
1575 # https://stackoverflow.com/questions/67463109/how-to-change-or-hide-nuxt-and-nuxt-keyword-in-page-source
1576 rectx = re.escape(context_name)
1577 js, arg_keys, arg_vals = self._search_regex(
1578 (r'<script
>window\
.%s=\
(function\
((?P
<arg_keys
>.*?
)\
)\{return\s(?P<js>\{.*?\}
)\
}\
((?P
<arg_vals
>.+?
)\
)\
);?
</script
>' % rectx,
1579 r'%s\
(.*?\
(function\
((?P
<arg_keys
>.*?
)\
)\{return\s(?P<js>\{.*?\}
)\
}\
((?P
<arg_vals
>.*?
)\
)' % rectx),
1580 webpage, context_name, group=['js
', 'arg_keys
', 'arg_vals
'])
1582 args = dict(zip(arg_keys.split(','), arg_vals.split(',')))
1584 for key, val in args.items():
1585 if val in ('undefined
', 'void
0'):
1588 ret = self._parse_json(js_to_json(js, args), video_id)
1589 if return_full_data:
1591 return ret['data
'][0]
1594 def _hidden_inputs(html):
1595 html = re.sub(r'<!--(?
:(?
!<!--).)*-->', '', html)
1597 for input in re.findall(r'(?i
)(<input[^
>]+>)', html):
1598 attrs = extract_attributes(input)
1601 if attrs.get('type') not in ('hidden
', 'submit
'):
1603 name = attrs.get('name
') or attrs.get('id')
1604 value = attrs.get('value
')
1605 if name and value is not None:
1606 hidden_inputs[name] = value
1607 return hidden_inputs
1609 def _form_hidden_inputs(self, form_id, html):
1610 form = self._search_regex(
1611 r'(?
is)<form
[^
>]+?
id=(["\'])%s\1[^>]*>(?P<form>.+?)</form>' % form_id,
1612 html, '%s form' % form_id, group='form')
1613 return self._hidden_inputs(form)
1616 regex = r' *((?P<reverse>\+)?(?P<field>[a-zA-Z0-9_]+)((?P<separator>[~:])(?P<limit>.*?))?)? *$'
1618 default = ('hidden', 'aud_or_vid', 'hasvid', 'ie_pref', 'lang', 'quality',
1619 'res', 'fps', 'hdr:12', 'codec:vp9.2', 'size', 'br', 'asr',
1620 'proto', 'ext', 'hasaud', 'source', 'id') # These must not be aliases
1621 ytdl_default = ('hasaud', 'lang', 'quality', 'tbr', 'filesize', 'vbr',
1622 'height', 'width', 'proto', 'vext', 'abr', 'aext',
1623 'fps', 'fs_approx', 'source', 'id')
1626 'vcodec': {'type': 'ordered', 'regex': True,
1627 'order': ['av0?1', 'vp0?9.2', 'vp0?9', '[hx]265|he?vc?', '[hx]264|avc', 'vp0?8', 'mp4v|h263', 'theora', '', None, 'none']},
1628 'acodec': {'type': 'ordered', 'regex': True,
1629 'order': ['[af]lac', 'wav|aiff', 'opus', 'vorbis|ogg', 'aac', 'mp?4a?', 'mp3', 'e-?a?c-?3', 'ac-?3', 'dts', '', None, 'none']},
1630 'hdr': {'type': 'ordered', 'regex': True, 'field': 'dynamic_range',
1631 'order': ['dv', '(hdr)?12', r'(hdr)?10\+', '(hdr)?10', 'hlg', '', 'sdr', None]},
1632 'proto': {'type': 'ordered', 'regex': True, 'field': 'protocol',
1633 'order': ['(ht|f)tps', '(ht|f)tp$', 'm3u8.*', '.*dash', 'websocket_frag', 'rtmpe?', '', 'mms|rtsp', 'ws|websocket', 'f4']},
1634 'vext': {'type': 'ordered', 'field': 'video_ext',
1635 'order': ('mp4', 'webm', 'flv', '', 'none'),
1636 'order_free': ('webm', 'mp4', 'flv', '', 'none')},
1637 'aext': {'type': 'ordered', 'field': 'audio_ext',
1638 'order': ('m4a', 'aac', 'mp3', 'ogg', 'opus', 'webm', '', 'none'),
1639 'order_free': ('opus', 'ogg', 'webm', 'm4a', 'mp3', 'aac', '', 'none')},
1640 'hidden': {'visible': False, 'forced': True, 'type': 'extractor', 'max': -1000},
1641 'aud_or_vid': {'visible': False, 'forced': True, 'type': 'multiple',
1642 'field': ('vcodec', 'acodec'),
1643 'function': lambda it: int(any(v != 'none' for v in it))},
1644 'ie_pref': {'priority': True, 'type': 'extractor'},
1645 'hasvid': {'priority': True, 'field': 'vcodec', 'type': 'boolean', 'not_in_list': ('none',)},
1646 'hasaud': {'field': 'acodec', 'type': 'boolean', 'not_in_list': ('none',)},
1647 'lang': {'convert': 'float', 'field': 'language_preference', 'default': -1},
1648 'quality': {'convert': 'float', 'default': -1},
1649 'filesize': {'convert': 'bytes'},
1650 'fs_approx': {'convert': 'bytes', 'field': 'filesize_approx'},
1651 'id': {'convert': 'string', 'field': 'format_id'},
1652 'height': {'convert': 'float_none'},
1653 'width': {'convert': 'float_none'},
1654 'fps': {'convert': 'float_none'},
1655 'tbr': {'convert': 'float_none'},
1656 'vbr': {'convert': 'float_none'},
1657 'abr': {'convert': 'float_none'},
1658 'asr': {'convert': 'float_none'},
1659 'source': {'convert': 'float', 'field': 'source_preference', 'default': -1},
1661 'codec': {'type': 'combined', 'field': ('vcodec', 'acodec')},
1662 'br': {'type': 'combined', 'field': ('tbr', 'vbr', 'abr'), 'same_limit': True},
1663 'size': {'type': 'combined', 'same_limit': True, 'field': ('filesize', 'fs_approx')},
1664 'ext': {'type': 'combined', 'field': ('vext', 'aext')},
1665 'res': {'type': 'multiple', 'field': ('height', 'width'),
1666 'function': lambda it: (lambda l: min(l) if l else 0)(tuple(filter(None, it)))},
1668 # For compatibility with youtube-dl
1669 'format_id': {'type': 'alias', 'field': 'id'},
1670 'preference': {'type': 'alias', 'field': 'ie_pref'},
1671 'language_preference': {'type': 'alias', 'field': 'lang'},
1672 'source_preference': {'type': 'alias', 'field': 'source'},
1673 'protocol': {'type': 'alias', 'field': 'proto'},
1674 'filesize_approx': {'type': 'alias', 'field': 'fs_approx'},
1677 'dimension': {'type': 'alias', 'field': 'res', 'deprecated': True},
1678 'resolution': {'type': 'alias', 'field': 'res', 'deprecated': True},
1679 'extension': {'type': 'alias', 'field': 'ext', 'deprecated': True},
1680 'bitrate': {'type': 'alias', 'field': 'br', 'deprecated': True},
1681 'total_bitrate': {'type': 'alias', 'field': 'tbr', 'deprecated': True},
1682 'video_bitrate': {'type': 'alias', 'field': 'vbr', 'deprecated': True},
1683 'audio_bitrate': {'type': 'alias', 'field': 'abr', 'deprecated': True},
1684 'framerate': {'type': 'alias', 'field': 'fps', 'deprecated': True},
1685 'filesize_estimate': {'type': 'alias', 'field': 'size', 'deprecated': True},
1686 'samplerate': {'type': 'alias', 'field': 'asr', 'deprecated': True},
1687 'video_ext': {'type': 'alias', 'field': 'vext', 'deprecated': True},
1688 'audio_ext': {'type': 'alias', 'field': 'aext', 'deprecated': True},
1689 'video_codec': {'type': 'alias', 'field': 'vcodec', 'deprecated': True},
1690 'audio_codec': {'type': 'alias', 'field': 'acodec', 'deprecated': True},
1691 'video': {'type': 'alias', 'field': 'hasvid', 'deprecated': True},
1692 'has_video': {'type': 'alias', 'field': 'hasvid', 'deprecated': True},
1693 'audio': {'type': 'alias', 'field': 'hasaud', 'deprecated': True},
1694 'has_audio': {'type': 'alias', 'field': 'hasaud', 'deprecated': True},
1695 'extractor': {'type': 'alias', 'field': 'ie_pref', 'deprecated': True},
1696 'extractor_preference': {'type': 'alias', 'field': 'ie_pref', 'deprecated': True},
1699 def __init__(self, ie, field_preference):
1701 self.ydl = ie._downloader
1702 self.evaluate_params(self.ydl.params, field_preference)
1703 if ie.get_param('verbose'):
1704 self.print_verbose_info(self.ydl.write_debug)
1706 def _get_field_setting(self, field, key):
1707 if field not in self.settings:
1708 if key in ('forced', 'priority'):
1710 self.ydl.deprecation_warning(
1711 f'Using arbitrary fields ({field}) for format sorting is deprecated '
1712 'and may be removed in a future version')
1713 self.settings[field] = {}
1714 propObj = self.settings[field]
1715 if key not in propObj:
1716 type = propObj.get('type')
1718 default = 'preference' if type == 'extractor' else (field,) if type in ('combined', 'multiple') else field
1719 elif key == 'convert':
1720 default = 'order' if type == 'ordered' else 'float_string' if field else 'ignore'
1722 default = {'type': 'field', 'visible': True, 'order': [], 'not_in_list': (None,)}.get(key, None)
1723 propObj[key] = default
1726 def _resolve_field_value(self, field, value, convertNone=False):
1731 value = value.lower()
1732 conversion = self._get_field_setting(field, 'convert')
1733 if conversion == 'ignore':
1735 if conversion == 'string':
1737 elif conversion == 'float_none':
1738 return float_or_none(value)
1739 elif conversion == 'bytes':
1740 return FileDownloader.parse_bytes(value)
1741 elif conversion == 'order':
1742 order_list = (self._use_free_order and self._get_field_setting(field, 'order_free')) or self._get_field_setting(field, 'order')
1743 use_regex = self._get_field_setting(field, 'regex')
1744 list_length = len(order_list)
1745 empty_pos = order_list.index('') if '' in order_list else list_length + 1
1746 if use_regex and value is not None:
1747 for i, regex in enumerate(order_list):
1748 if regex and re.match(regex, value):
1749 return list_length - i
1750 return list_length - empty_pos # not in list
1751 else: # not regex or value = None
1752 return list_length - (order_list.index(value) if value in order_list else empty_pos)
1754 if value.isnumeric():
1757 self.settings[field]['convert'] = 'string'
1760 def evaluate_params(self, params, sort_extractor):
1761 self._use_free_order = params.get('prefer_free_formats', False)
1762 self._sort_user = params.get('format_sort', [])
1763 self._sort_extractor = sort_extractor
1765 def add_item(field, reverse, closest, limit_text):
1766 field = field.lower()
1767 if field in self._order:
1769 self._order.append(field)
1770 limit = self._resolve_field_value(field, limit_text)
1773 'closest': False if limit is None else closest,
1774 'limit_text': limit_text,
1776 if field in self.settings:
1777 self.settings[field].update(data)
1779 self.settings[field] = data
1782 tuple(field for field in self.default if self._get_field_setting(field, 'forced'))
1783 + (tuple() if params.get('format_sort_force', False)
1784 else tuple(field for field in self.default if self._get_field_setting(field, 'priority')))
1785 + tuple(self._sort_user) + tuple(sort_extractor) + self.default)
1787 for item in sort_list:
1788 match = re.match(self.regex, item)
1790 raise ExtractorError('Invalid format sort string "%s" given by extractor' % item)
1791 field = match.group('field')
1794 if self._get_field_setting(field, 'type') == 'alias':
1795 alias, field = field, self._get_field_setting(field, 'field')
1796 if self._get_field_setting(alias, 'deprecated'):
1797 self.ydl.deprecation_warning(
1798 f'Format sorting alias {alias} is deprecated '
1799 f'and may be removed in a future version. Please use {field} instead')
1800 reverse = match.group('reverse') is not None
1801 closest = match.group('separator') == '~'
1802 limit_text = match.group('limit')
1804 has_limit = limit_text is not None
1805 has_multiple_fields = self._get_field_setting(field, 'type') == 'combined'
1806 has_multiple_limits = has_limit and has_multiple_fields and not self._get_field_setting(field, 'same_limit')
1808 fields = self._get_field_setting(field, 'field') if has_multiple_fields else (field,)
1809 limits = limit_text.split(':') if has_multiple_limits else (limit_text,) if has_limit else tuple()
1810 limit_count = len(limits)
1811 for (i, f) in enumerate(fields):
1812 add_item(f, reverse, closest,
1813 limits[i] if i < limit_count
1814 else limits[0] if has_limit and not has_multiple_limits
1817 def print_verbose_info(self, write_debug):
1819 write_debug('Sort order given by user: %s' % ', '.join(self._sort_user))
1820 if self._sort_extractor:
1821 write_debug('Sort order given by extractor: %s' % ', '.join(self._sort_extractor))
1822 write_debug('Formats sorted by: %s' % ', '.join(['%s%s%s' % (
1823 '+' if self._get_field_setting(field, 'reverse') else '', field,
1824 '%s%s(%s)' % ('~' if self._get_field_setting(field, 'closest') else ':',
1825 self._get_field_setting(field, 'limit_text'),
1826 self._get_field_setting(field, 'limit'))
1827 if self._get_field_setting(field, 'limit_text') is not None else '')
1828 for field in self._order if self._get_field_setting(field, 'visible')]))
1830 def _calculate_field_preference_from_value(self, format, field, type, value):
1831 reverse = self._get_field_setting(field, 'reverse')
1832 closest = self._get_field_setting(field, 'closest')
1833 limit = self._get_field_setting(field, 'limit')
1835 if type == 'extractor':
1836 maximum = self._get_field_setting(field, 'max')
1837 if value is None or (maximum is not None and value >= maximum):
1839 elif type == 'boolean':
1840 in_list = self._get_field_setting(field, 'in_list')
1841 not_in_list = self._get_field_setting(field, 'not_in_list')
1842 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
1843 elif type == 'ordered':
1844 value = self._resolve_field_value(field, value, True)
1846 # try to convert to number
1847 val_num = float_or_none(value, default=self._get_field_setting(field, 'default'))
1848 is_num = self._get_field_setting(field, 'convert') != 'string' and val_num is not None
1852 return ((-10, 0) if value is None
1853 else (1, value, 0) if not is_num # if a field has mixed strings and numbers, strings are sorted higher
1854 else (0, -abs(value - limit), value - limit if reverse else limit - value) if closest
1855 else (0, value, 0) if not reverse and (limit is None or value <= limit)
1856 else (0, -value, 0) if limit is None or (reverse and value == limit) or value > limit
1857 else (-1, value, 0))
1859 def _calculate_field_preference(self, format, field):
1860 type = self._get_field_setting(field, 'type') # extractor, boolean, ordered, field, multiple
1861 get_value = lambda f: format.get(self._get_field_setting(f, 'field'))
1862 if type == 'multiple':
1863 type = 'field' # Only 'field' is allowed in multiple for now
1864 actual_fields = self._get_field_setting(field, 'field')
1866 value = self._get_field_setting(field, 'function')(get_value(f) for f in actual_fields)
1868 value = get_value(field)
1869 return self._calculate_field_preference_from_value(format, field, type, value)
1871 def calculate_preference(self, format):
1872 # Determine missing protocol
1873 if not format.get('protocol'):
1874 format['protocol'] = determine_protocol(format)
1876 # Determine missing ext
1877 if not format.get('ext') and 'url' in format:
1878 format['ext'] = determine_ext(format['url'])
1879 if format.get('vcodec') == 'none':
1880 format['audio_ext'] = format['ext'] if format.get('acodec') != 'none' else 'none'
1881 format['video_ext'] = 'none'
1883 format['video_ext'] = format['ext']
1884 format['audio_ext'] = 'none'
1885 # if format.get('preference') is None and format.get('ext') in ('f4f', 'f4m'): # Not supported?
1886 # format['preference'] = -1000
1888 # Determine missing bitrates
1889 if format.get('tbr') is None:
1890 if format.get('vbr') is not None and format.get('abr') is not None:
1891 format['tbr'] = format.get('vbr', 0) + format.get('abr', 0)
1893 if format.get('vcodec') != 'none' and format.get('vbr') is None:
1894 format['vbr'] = format.get('tbr') - format.get('abr', 0)
1895 if format.get('acodec') != 'none' and format.get('abr') is None:
1896 format['abr'] = format.get('tbr') - format.get('vbr', 0)
1898 return tuple(self._calculate_field_preference(format, field) for field in self._order)
1900 def _sort_formats(self, formats, field_preference=[]):
1903 formats.sort(key=self.FormatSort(self, field_preference).calculate_preference)
1905 def _check_formats(self, formats, video_id):
1907 formats[:] = filter(
1908 lambda f: self._is_valid_url(
1910 item='%s video format' % f.get('format_id') if f.get('format_id') else 'video'),
1914 def _remove_duplicate_formats(formats):
1918 if f['url'] not in format_urls:
1919 format_urls.add(f['url'])
1920 unique_formats.append(f)
1921 formats[:] = unique_formats
1923 def _is_valid_url(self, url, video_id, item='video', headers={}):
1924 url = self._proto_relative_url(url, scheme='http:')
1925 # For now assume non HTTP(S) URLs always valid
1926 if not (url.startswith('http://') or url.startswith('https://')):
1929 self._request_webpage(url, video_id, 'Checking %s URL' % item, headers=headers)
1931 except ExtractorError as e:
1933 '%s: %s URL is invalid, skipping: %s'
1934 % (video_id, item, error_to_compat_str(e.cause)))
1937 def http_scheme(self):
1938 """ Either "http
:" or "https
:", depending on the user's preferences """
1941 if self.get_param('prefer_insecure', False)
1944 def _proto_relative_url(self, url, scheme=None):
1947 if url.startswith('//'):
1949 scheme = self.http_scheme()
1954 def _sleep(self, timeout, video_id, msg_template=None):
1955 if msg_template is None:
1956 msg_template = '%(video_id)s: Waiting for %(timeout)s seconds'
1957 msg = msg_template % {'video_id': video_id, 'timeout': timeout}
1961 def _extract_f4m_formats(self, manifest_url, video_id, preference=None, quality=None, f4m_id=None,
1962 transform_source=lambda s: fix_xml_ampersands(s).strip(),
1963 fatal=True, m3u8_id=None, data=None, headers={}, query={}):
1964 res = self._download_xml_handle(
1965 manifest_url, video_id, 'Downloading f4m manifest',
1966 'Unable to download f4m manifest',
1967 # Some manifests may be malformed, e.g. prosiebensat1 generated manifests
1968 # (see https://github.com/ytdl-org/youtube-dl/issues/6215#issuecomment-121704244)
1969 transform_source=transform_source,
1970 fatal=fatal, data=data, headers=headers, query=query)
1974 manifest, urlh = res
1975 manifest_url = urlh.geturl()
1977 return self._parse_f4m_formats(
1978 manifest, manifest_url, video_id, preference=preference, quality=quality, f4m_id=f4m_id,
1979 transform_source=transform_source, fatal=fatal, m3u8_id=m3u8_id)
1981 def _parse_f4m_formats(self, manifest, manifest_url, video_id, preference=None, quality=None, f4m_id=None,
1982 transform_source=lambda s: fix_xml_ampersands(s).strip(),
1983 fatal=True, m3u8_id=None):
1984 if not isinstance(manifest, xml.etree.ElementTree.Element) and not fatal:
1987 # currently yt-dlp cannot decode the playerVerificationChallenge as Akamai uses Adobe Alchemy
1988 akamai_pv = manifest.find('{http://ns.adobe.com/f4m/1.0}pv-2.0')
1989 if akamai_pv is not None and ';' in akamai_pv.text:
1990 playerVerificationChallenge = akamai_pv.text.split(';')[0]
1991 if playerVerificationChallenge.strip() != '':
1995 manifest_version = '1.0'
1996 media_nodes = manifest.findall('{http://ns.adobe.com/f4m/1.0}media')
1998 manifest_version = '2.0'
1999 media_nodes = manifest.findall('{http://ns.adobe.com/f4m/2.0}media')
2000 # Remove unsupported DRM protected media from final formats
2001 # rendition (see https://github.com/ytdl-org/youtube-dl/issues/8573).
2002 media_nodes = remove_encrypted_media(media_nodes)
2006 manifest_base_url = get_base_url(manifest)
2008 bootstrap_info = xpath_element(
2009 manifest, ['{http://ns.adobe.com/f4m/1.0}bootstrapInfo', '{http://ns.adobe.com/f4m/2.0}bootstrapInfo'],
2010 'bootstrap info', default=None)
2013 mime_type = xpath_text(
2014 manifest, ['{http://ns.adobe.com/f4m/1.0}mimeType', '{http://ns.adobe.com/f4m/2.0}mimeType'],
2015 'base URL', default=None)
2016 if mime_type and mime_type.startswith('audio/'):
2019 for i, media_el in enumerate(media_nodes):
2020 tbr = int_or_none(media_el.attrib.get('bitrate'))
2021 width = int_or_none(media_el.attrib.get('width'))
2022 height = int_or_none(media_el.attrib.get('height'))
2023 format_id = join_nonempty(f4m_id, tbr or i)
2024 # If <bootstrapInfo> is present, the specified f4m is a
2025 # stream-level manifest, and only set-level manifests may refer to
2026 # external resources. See section 11.4 and section 4 of F4M spec
2027 if bootstrap_info is None:
2029 # @href is introduced in 2.0, see section 11.6 of F4M spec
2030 if manifest_version == '2.0':
2031 media_url = media_el.attrib.get('href')
2032 if media_url is None:
2033 media_url = media_el.attrib.get('url')
2037 media_url if media_url.startswith('http://') or media_url.startswith('https://')
2038 else ((manifest_base_url or '/'.join(manifest_url.split('/')[:-1])) + '/' + media_url))
2039 # If media_url is itself a f4m manifest do the recursive extraction
2040 # since bitrates in parent manifest (this one) and media_url manifest
2041 # may differ leading to inability to resolve the format by requested
2042 # bitrate in f4m downloader
2043 ext = determine_ext(manifest_url)
2045 f4m_formats = self._extract_f4m_formats(
2046 manifest_url, video_id, preference=preference, quality=quality, f4m_id=f4m_id,
2047 transform_source=transform_source, fatal=fatal)
2048 # Sometimes stream-level manifest contains single media entry that
2049 # does not contain any quality metadata (e.g. http://matchtv.ru/#live-player).
2050 # At the same time parent's media entry in set-level manifest may
2051 # contain it. We will copy it from parent in such cases.
2052 if len(f4m_formats) == 1:
2055 'tbr': f.get('tbr') or tbr,
2056 'width': f.get('width') or width,
2057 'height': f.get('height') or height,
2058 'format_id': f.get('format_id') if not tbr else format_id,
2061 formats.extend(f4m_formats)
2064 formats.extend(self._extract_m3u8_formats(
2065 manifest_url, video_id, 'mp4', preference=preference,
2066 quality=quality, m3u8_id=m3u8_id, fatal=fatal))
2069 'format_id': format_id,
2070 'url': manifest_url,
2071 'manifest_url': manifest_url,
2072 'ext': 'flv' if bootstrap_info is not None else None,
2078 'preference': preference,
2083 def _m3u8_meta_format(self, m3u8_url, ext=None, preference=None, quality=None, m3u8_id=None):
2085 'format_id': join_nonempty(m3u8_id, 'meta'),
2089 'preference': preference - 100 if preference else -100,
2091 'resolution': 'multiple',
2092 'format_note': 'Quality selection URL',
2095 def _report_ignoring_subs(self, name):
2096 self.report_warning(bug_reports_message(
2097 f'Ignoring subtitle tracks found in the {name} manifest; '
2098 'if any subtitle tracks are missing,'
2101 def _extract_m3u8_formats(self, *args, **kwargs):
2102 fmts, subs = self._extract_m3u8_formats_and_subtitles(*args, **kwargs)
2104 self._report_ignoring_subs('HLS')
2107 def _extract_m3u8_formats_and_subtitles(
2108 self, m3u8_url, video_id, ext=None, entry_protocol='m3u8_native',
2109 preference=None, quality=None, m3u8_id=None, note=None,
2110 errnote=None, fatal=True, live=False, data=None, headers={},
2113 res = self._download_webpage_handle(
2115 note='Downloading m3u8 information' if note is None else note,
2116 errnote='Failed to download m3u8 information' if errnote is None else errnote,
2117 fatal=fatal, data=data, headers=headers, query=query)
2122 m3u8_doc, urlh = res
2123 m3u8_url = urlh.geturl()
2125 return self._parse_m3u8_formats_and_subtitles(
2126 m3u8_doc, m3u8_url, ext=ext, entry_protocol=entry_protocol,
2127 preference=preference, quality=quality, m3u8_id=m3u8_id,
2128 note=note, errnote=errnote, fatal=fatal, live=live, data=data,
2129 headers=headers, query=query, video_id=video_id)
2131 def _parse_m3u8_formats_and_subtitles(
2132 self, m3u8_doc, m3u8_url=None, ext=None, entry_protocol='m3u8_native',
2133 preference=None, quality=None, m3u8_id=None, live=False, note=None,
2134 errnote=None, fatal=True, data=None, headers={}, query={},
2136 formats, subtitles = [], {}
2138 has_drm = re.search('|'.join([
2139 r'#EXT-X-FAXS-CM:', # Adobe Flash Access
2140 r'#EXT-X-(?:SESSION-)?KEY:.*?URI="skd
://', # Apple FairPlay
2143 def format_url(url):
2144 return url if re.match(r'^https?
://', url) else compat_urlparse.urljoin(m3u8_url, url)
2146 if self.get_param('hls_split_discontinuity
', False):
2147 def _extract_m3u8_playlist_indices(manifest_url=None, m3u8_doc=None):
2149 if not manifest_url:
2151 m3u8_doc = self._download_webpage(
2152 manifest_url, video_id, fatal=fatal, data=data, headers=headers,
2153 note=False, errnote='Failed to download m3u8 playlist information
')
2154 if m3u8_doc is False:
2156 return range(1 + sum(line.startswith('#EXT-X-DISCONTINUITY') for line in m3u8_doc.splitlines()))
2159 def _extract_m3u8_playlist_indices(*args
, **kwargs
):
2163 # 1. https://tools.ietf.org/html/draft-pantos-http-live-streaming-21
2164 # 2. https://github.com/ytdl-org/youtube-dl/issues/12211
2165 # 3. https://github.com/ytdl-org/youtube-dl/issues/18923
2167 # We should try extracting formats only from master playlists [1, 4.3.4],
2168 # i.e. playlists that describe available qualities. On the other hand
2169 # media playlists [1, 4.3.3] should be returned as is since they contain
2170 # just the media without qualities renditions.
2171 # Fortunately, master playlist can be easily distinguished from media
2172 # playlist based on particular tags availability. As of [1, 4.3.3, 4.3.4]
2173 # master playlist tags MUST NOT appear in a media playlist and vice versa.
2174 # As of [1, 4.3.3.1] #EXT-X-TARGETDURATION tag is REQUIRED for every
2175 # media playlist and MUST NOT appear in master playlist thus we can
2176 # clearly detect media playlist with this criterion.
2178 if '#EXT-X-TARGETDURATION' in m3u8_doc
: # media playlist, return as is
2180 'format_id': join_nonempty(m3u8_id
, idx
),
2181 'format_index': idx
,
2182 'url': m3u8_url
or encode_data_uri(m3u8_doc
.encode('utf-8'), 'application/x-mpegurl'),
2184 'protocol': entry_protocol
,
2185 'preference': preference
,
2188 } for idx
in _extract_m3u8_playlist_indices(m3u8_doc
=m3u8_doc
)]
2190 return formats
, subtitles
2193 last_stream_inf
= {}
2195 def extract_media(x_media_line
):
2196 media
= parse_m3u8_attributes(x_media_line
)
2197 # As per [1, 4.3.4.1] TYPE, GROUP-ID and NAME are REQUIRED
2198 media_type
, group_id
, name
= media
.get('TYPE'), media
.get('GROUP-ID'), media
.get('NAME')
2199 if not (media_type
and group_id
and name
):
2201 groups
.setdefault(group_id
, []).append(media
)
2202 # <https://tools.ietf.org/html/rfc8216#section-4.3.4.1>
2203 if media_type
== 'SUBTITLES':
2204 # According to RFC 8216 §4.3.4.2.1, URI is REQUIRED in the
2205 # EXT-X-MEDIA tag if the media type is SUBTITLES.
2206 # However, lack of URI has been spotted in the wild.
2207 # e.g. NebulaIE; see https://github.com/yt-dlp/yt-dlp/issues/339
2208 if not media
.get('URI'):
2210 url
= format_url(media
['URI'])
2213 'ext': determine_ext(url
),
2215 if sub_info
['ext'] == 'm3u8':
2216 # Per RFC 8216 §3.1, the only possible subtitle format m3u8
2217 # files may contain is WebVTT:
2218 # <https://tools.ietf.org/html/rfc8216#section-3.1>
2219 sub_info
['ext'] = 'vtt'
2220 sub_info
['protocol'] = 'm3u8_native'
2221 lang
= media
.get('LANGUAGE') or 'und'
2222 subtitles
.setdefault(lang
, []).append(sub_info
)
2223 if media_type
not in ('VIDEO', 'AUDIO'):
2225 media_url
= media
.get('URI')
2227 manifest_url
= format_url(media_url
)
2229 'format_id': join_nonempty(m3u8_id
, group_id
, name
, idx
),
2230 'format_note': name
,
2231 'format_index': idx
,
2232 'url': manifest_url
,
2233 'manifest_url': m3u8_url
,
2234 'language': media
.get('LANGUAGE'),
2236 'protocol': entry_protocol
,
2237 'preference': preference
,
2239 'vcodec': 'none' if media_type
== 'AUDIO' else None,
2240 } for idx
in _extract_m3u8_playlist_indices(manifest_url
))
2242 def build_stream_name():
2243 # Despite specification does not mention NAME attribute for
2244 # EXT-X-STREAM-INF tag it still sometimes may be present (see [1]
2245 # or vidio test in TestInfoExtractor.test_parse_m3u8_formats)
2246 # 1. http://www.vidio.com/watch/165683-dj_ambred-booyah-live-2015
2247 stream_name
= last_stream_inf
.get('NAME')
2250 # If there is no NAME in EXT-X-STREAM-INF it will be obtained
2251 # from corresponding rendition group
2252 stream_group_id
= last_stream_inf
.get('VIDEO')
2253 if not stream_group_id
:
2255 stream_group
= groups
.get(stream_group_id
)
2256 if not stream_group
:
2257 return stream_group_id
2258 rendition
= stream_group
[0]
2259 return rendition
.get('NAME') or stream_group_id
2261 # parse EXT-X-MEDIA tags before EXT-X-STREAM-INF in order to have the
2262 # chance to detect video only formats when EXT-X-STREAM-INF tags
2263 # precede EXT-X-MEDIA tags in HLS manifest such as [3].
2264 for line
in m3u8_doc
.splitlines():
2265 if line
.startswith('#EXT-X-MEDIA:'):
2268 for line
in m3u8_doc
.splitlines():
2269 if line
.startswith('#EXT-X-STREAM-INF:'):
2270 last_stream_inf
= parse_m3u8_attributes(line
)
2271 elif line
.startswith('#') or not line
.strip():
2274 tbr
= float_or_none(
2275 last_stream_inf
.get('AVERAGE-BANDWIDTH')
2276 or last_stream_inf
.get('BANDWIDTH'), scale
=1000)
2277 manifest_url
= format_url(line
.strip())
2279 for idx
in _extract_m3u8_playlist_indices(manifest_url
):
2280 format_id
= [m3u8_id
, None, idx
]
2281 # Bandwidth of live streams may differ over time thus making
2282 # format_id unpredictable. So it's better to keep provided
2285 stream_name
= build_stream_name()
2286 format_id
[1] = stream_name
or '%d' % (tbr
or len(formats
))
2288 'format_id': join_nonempty(*format_id
),
2289 'format_index': idx
,
2290 'url': manifest_url
,
2291 'manifest_url': m3u8_url
,
2294 'fps': float_or_none(last_stream_inf
.get('FRAME-RATE')),
2295 'protocol': entry_protocol
,
2296 'preference': preference
,
2299 resolution
= last_stream_inf
.get('RESOLUTION')
2301 mobj
= re
.search(r
'(?P<width>\d+)[xX](?P<height>\d+)', resolution
)
2303 f
['width'] = int(mobj
.group('width'))
2304 f
['height'] = int(mobj
.group('height'))
2305 # Unified Streaming Platform
2307 r
'audio.*?(?:%3D|=)(\d+)(?:-video.*?(?:%3D|=)(\d+))?', f
['url'])
2309 abr
, vbr
= mobj
.groups()
2310 abr
, vbr
= float_or_none(abr
, 1000), float_or_none(vbr
, 1000)
2315 codecs
= parse_codecs(last_stream_inf
.get('CODECS'))
2317 audio_group_id
= last_stream_inf
.get('AUDIO')
2318 # As per [1, 4.3.4.1.1] any EXT-X-STREAM-INF tag which
2319 # references a rendition group MUST have a CODECS attribute.
2320 # However, this is not always respected, for example, [2]
2321 # contains EXT-X-STREAM-INF tag which references AUDIO
2322 # rendition group but does not have CODECS and despite
2323 # referencing an audio group it represents a complete
2324 # (with audio and video) format. So, for such cases we will
2325 # ignore references to rendition groups and treat them
2326 # as complete formats.
2327 if audio_group_id
and codecs
and f
.get('vcodec') != 'none':
2328 audio_group
= groups
.get(audio_group_id
)
2329 if audio_group
and audio_group
[0].get('URI'):
2330 # TODO: update acodec for audio only formats with
2332 f
['acodec'] = 'none'
2333 if not f
.get('ext'):
2334 f
['ext'] = 'm4a' if f
.get('vcodec') == 'none' else 'mp4'
2338 progressive_uri
= last_stream_inf
.get('PROGRESSIVE-URI')
2341 del http_f
['manifest_url']
2343 'format_id': f
['format_id'].replace('hls-', 'http-'),
2345 'url': progressive_uri
,
2347 formats
.append(http_f
)
2349 last_stream_inf
= {}
2350 return formats
, subtitles
2352 def _extract_m3u8_vod_duration(
2353 self
, m3u8_vod_url
, video_id
, note
=None, errnote
=None, data
=None, headers
={}, query={}
):
2355 m3u8_vod
= self
._download
_webpage
(
2356 m3u8_vod_url
, video_id
,
2357 note
='Downloading m3u8 VOD manifest' if note
is None else note
,
2358 errnote
='Failed to download VOD manifest' if errnote
is None else errnote
,
2359 fatal
=False, data
=data
, headers
=headers
, query
=query
)
2361 return self
._parse
_m
3u8_vod
_duration
(m3u8_vod
or '', video_id
)
2363 def _parse_m3u8_vod_duration(self
, m3u8_vod
, video_id
):
2364 if '#EXT-X-PLAYLIST-TYPE:VOD' not in m3u8_vod
:
2368 float(line
[len('#EXTINF:'):].split(',')[0])
2369 for line
in m3u8_vod
.splitlines() if line
.startswith('#EXTINF:'))) or None
2372 def _xpath_ns(path
, namespace
=None):
2376 for c
in path
.split('/'):
2377 if not c
or c
== '.':
2380 out
.append('{%s}%s' % (namespace
, c
))
2381 return '/'.join(out
)
2383 def _extract_smil_formats_and_subtitles(self
, smil_url
, video_id
, fatal
=True, f4m_params
=None, transform_source
=None):
2384 res
= self
._download
_smil
(smil_url
, video_id
, fatal
=fatal
, transform_source
=transform_source
)
2390 smil_url
= urlh
.geturl()
2392 namespace
= self
._parse
_smil
_namespace
(smil
)
2394 fmts
= self
._parse
_smil
_formats
(
2395 smil
, smil_url
, video_id
, namespace
=namespace
, f4m_params
=f4m_params
)
2396 subs
= self
._parse
_smil
_subtitles
(
2397 smil
, namespace
=namespace
)
2401 def _extract_smil_formats(self
, *args
, **kwargs
):
2402 fmts
, subs
= self
._extract
_smil
_formats
_and
_subtitles
(*args
, **kwargs
)
2404 self
._report
_ignoring
_subs
('SMIL')
2407 def _extract_smil_info(self
, smil_url
, video_id
, fatal
=True, f4m_params
=None):
2408 res
= self
._download
_smil
(smil_url
, video_id
, fatal
=fatal
)
2413 smil_url
= urlh
.geturl()
2415 return self
._parse
_smil
(smil
, smil_url
, video_id
, f4m_params
=f4m_params
)
2417 def _download_smil(self
, smil_url
, video_id
, fatal
=True, transform_source
=None):
2418 return self
._download
_xml
_handle
(
2419 smil_url
, video_id
, 'Downloading SMIL file',
2420 'Unable to download SMIL file', fatal
=fatal
, transform_source
=transform_source
)
2422 def _parse_smil(self
, smil
, smil_url
, video_id
, f4m_params
=None):
2423 namespace
= self
._parse
_smil
_namespace
(smil
)
2425 formats
= self
._parse
_smil
_formats
(
2426 smil
, smil_url
, video_id
, namespace
=namespace
, f4m_params
=f4m_params
)
2427 subtitles
= self
._parse
_smil
_subtitles
(smil
, namespace
=namespace
)
2429 video_id
= os
.path
.splitext(url_basename(smil_url
))[0]
2433 for meta
in smil
.findall(self
._xpath
_ns
('./head/meta', namespace
)):
2434 name
= meta
.attrib
.get('name')
2435 content
= meta
.attrib
.get('content')
2436 if not name
or not content
:
2438 if not title
and name
== 'title':
2440 elif not description
and name
in ('description', 'abstract'):
2441 description
= content
2442 elif not upload_date
and name
== 'date':
2443 upload_date
= unified_strdate(content
)
2446 'id': image
.get('type'),
2447 'url': image
.get('src'),
2448 'width': int_or_none(image
.get('width')),
2449 'height': int_or_none(image
.get('height')),
2450 } for image
in smil
.findall(self
._xpath
_ns
('.//image', namespace
)) if image
.get('src')]
2454 'title': title
or video_id
,
2455 'description': description
,
2456 'upload_date': upload_date
,
2457 'thumbnails': thumbnails
,
2459 'subtitles': subtitles
,
2462 def _parse_smil_namespace(self
, smil
):
2463 return self
._search
_regex
(
2464 r
'(?i)^{([^}]+)?}smil$', smil
.tag
, 'namespace', default
=None)
2466 def _parse_smil_formats(self
, smil
, smil_url
, video_id
, namespace
=None, f4m_params
=None, transform_rtmp_url
=None):
2468 for meta
in smil
.findall(self
._xpath
_ns
('./head/meta', namespace
)):
2469 b
= meta
.get('base') or meta
.get('httpBase')
2481 media
= smil
.findall(self
._xpath
_ns
('.//video', namespace
)) + smil
.findall(self
._xpath
_ns
('.//audio', namespace
))
2482 for medium
in media
:
2483 src
= medium
.get('src')
2484 if not src
or src
in srcs
:
2488 bitrate
= float_or_none(medium
.get('system-bitrate') or medium
.get('systemBitrate'), 1000)
2489 filesize
= int_or_none(medium
.get('size') or medium
.get('fileSize'))
2490 width
= int_or_none(medium
.get('width'))
2491 height
= int_or_none(medium
.get('height'))
2492 proto
= medium
.get('proto')
2493 ext
= medium
.get('ext')
2494 src_ext
= determine_ext(src
)
2495 streamer
= medium
.get('streamer') or base
2497 if proto
== 'rtmp' or streamer
.startswith('rtmp'):
2503 'format_id': 'rtmp-%d' % (rtmp_count
if bitrate
is None else bitrate
),
2505 'filesize': filesize
,
2509 if transform_rtmp_url
:
2510 streamer
, src
= transform_rtmp_url(streamer
, src
)
2511 formats
[-1].update({
2517 src_url
= src
if src
.startswith('http') else compat_urlparse
.urljoin(base
, src
)
2518 src_url
= src_url
.strip()
2520 if proto
== 'm3u8' or src_ext
== 'm3u8':
2521 m3u8_formats
= self
._extract
_m
3u8_formats
(
2522 src_url
, video_id
, ext
or 'mp4', m3u8_id
='hls', fatal
=False)
2523 if len(m3u8_formats
) == 1:
2525 m3u8_formats
[0].update({
2526 'format_id': 'hls-%d' % (m3u8_count
if bitrate
is None else bitrate
),
2531 formats
.extend(m3u8_formats
)
2532 elif src_ext
== 'f4m':
2537 'plugin': 'flowplayer-3.2.0.1',
2539 f4m_url
+= '&' if '?' in f4m_url
else '?'
2540 f4m_url
+= compat_urllib_parse_urlencode(f4m_params
)
2541 formats
.extend(self
._extract
_f
4m
_formats
(f4m_url
, video_id
, f4m_id
='hds', fatal
=False))
2542 elif src_ext
== 'mpd':
2543 formats
.extend(self
._extract
_mpd
_formats
(
2544 src_url
, video_id
, mpd_id
='dash', fatal
=False))
2545 elif re
.search(r
'\.ism/[Mm]anifest', src_url
):
2546 formats
.extend(self
._extract
_ism
_formats
(
2547 src_url
, video_id
, ism_id
='mss', fatal
=False))
2548 elif src_url
.startswith('http') and self
._is
_valid
_url
(src
, video_id
):
2552 'ext': ext
or src_ext
or 'flv',
2553 'format_id': 'http-%d' % (bitrate
or http_count
),
2555 'filesize': filesize
,
2560 for medium
in smil
.findall(self
._xpath
_ns
('.//imagestream', namespace
)):
2561 src
= medium
.get('src')
2562 if not src
or src
in srcs
:
2568 'format_id': 'imagestream-%d' % (imgs_count
),
2570 'ext': mimetype2ext(medium
.get('type')),
2573 'width': int_or_none(medium
.get('width')),
2574 'height': int_or_none(medium
.get('height')),
2575 'format_note': 'SMIL storyboards',
2580 def _parse_smil_subtitles(self
, smil
, namespace
=None, subtitles_lang
='en'):
2583 for num
, textstream
in enumerate(smil
.findall(self
._xpath
_ns
('.//textstream', namespace
))):
2584 src
= textstream
.get('src')
2585 if not src
or src
in urls
:
2588 ext
= textstream
.get('ext') or mimetype2ext(textstream
.get('type')) or determine_ext(src
)
2589 lang
= textstream
.get('systemLanguage') or textstream
.get('systemLanguageName') or textstream
.get('lang') or subtitles_lang
2590 subtitles
.setdefault(lang
, []).append({
2596 def _extract_xspf_playlist(self
, xspf_url
, playlist_id
, fatal
=True):
2597 res
= self
._download
_xml
_handle
(
2598 xspf_url
, playlist_id
, 'Downloading xpsf playlist',
2599 'Unable to download xspf manifest', fatal
=fatal
)
2604 xspf_url
= urlh
.geturl()
2606 return self
._parse
_xspf
(
2607 xspf
, playlist_id
, xspf_url
=xspf_url
,
2608 xspf_base_url
=base_url(xspf_url
))
2610 def _parse_xspf(self
, xspf_doc
, playlist_id
, xspf_url
=None, xspf_base_url
=None):
2612 'xspf': 'http://xspf.org/ns/0/',
2613 's1': 'http://static.streamone.nl/player/ns/0',
2617 for track
in xspf_doc
.findall(xpath_with_ns('./xspf:trackList/xspf:track', NS_MAP
)):
2619 track
, xpath_with_ns('./xspf:title', NS_MAP
), 'title', default
=playlist_id
)
2620 description
= xpath_text(
2621 track
, xpath_with_ns('./xspf:annotation', NS_MAP
), 'description')
2622 thumbnail
= xpath_text(
2623 track
, xpath_with_ns('./xspf:image', NS_MAP
), 'thumbnail')
2624 duration
= float_or_none(
2625 xpath_text(track
, xpath_with_ns('./xspf:duration', NS_MAP
), 'duration'), 1000)
2628 for location
in track
.findall(xpath_with_ns('./xspf:location', NS_MAP
)):
2629 format_url
= urljoin(xspf_base_url
, location
.text
)
2634 'manifest_url': xspf_url
,
2635 'format_id': location
.get(xpath_with_ns('s1:label', NS_MAP
)),
2636 'width': int_or_none(location
.get(xpath_with_ns('s1:width', NS_MAP
))),
2637 'height': int_or_none(location
.get(xpath_with_ns('s1:height', NS_MAP
))),
2639 self
._sort
_formats
(formats
)
2644 'description': description
,
2645 'thumbnail': thumbnail
,
2646 'duration': duration
,
2651 def _extract_mpd_formats(self
, *args
, **kwargs
):
2652 fmts
, subs
= self
._extract
_mpd
_formats
_and
_subtitles
(*args
, **kwargs
)
2654 self
._report
_ignoring
_subs
('DASH')
2657 def _extract_mpd_formats_and_subtitles(
2658 self
, mpd_url
, video_id
, mpd_id
=None, note
=None, errnote
=None,
2659 fatal
=True, data
=None, headers
={}, query={}
):
2660 res
= self
._download
_xml
_handle
(
2662 note
='Downloading MPD manifest' if note
is None else note
,
2663 errnote
='Failed to download MPD manifest' if errnote
is None else errnote
,
2664 fatal
=fatal
, data
=data
, headers
=headers
, query
=query
)
2671 # We could have been redirected to a new url when we retrieved our mpd file.
2672 mpd_url
= urlh
.geturl()
2673 mpd_base_url
= base_url(mpd_url
)
2675 return self
._parse
_mpd
_formats
_and
_subtitles
(
2676 mpd_doc
, mpd_id
, mpd_base_url
, mpd_url
)
2678 def _parse_mpd_formats(self
, *args
, **kwargs
):
2679 fmts
, subs
= self
._parse
_mpd
_formats
_and
_subtitles
(*args
, **kwargs
)
2681 self
._report
_ignoring
_subs
('DASH')
2684 def _parse_mpd_formats_and_subtitles(
2685 self
, mpd_doc
, mpd_id
=None, mpd_base_url
='', mpd_url
=None):
2687 Parse formats from MPD manifest.
2689 1. MPEG-DASH Standard, ISO/IEC 23009-1:2014(E),
2690 http://standards.iso.org/ittf/PubliclyAvailableStandards/c065274_ISO_IEC_23009-1_2014.zip
2691 2. https://en.wikipedia.org/wiki/Dynamic_Adaptive_Streaming_over_HTTP
2693 if not self
.get_param('dynamic_mpd', True):
2694 if mpd_doc
.get('type') == 'dynamic':
2697 namespace
= self
._search
_regex
(r
'(?i)^{([^}]+)?}MPD$', mpd_doc
.tag
, 'namespace', default
=None)
2700 return self
._xpath
_ns
(path
, namespace
)
2702 def is_drm_protected(element
):
2703 return element
.find(_add_ns('ContentProtection')) is not None
2705 def extract_multisegment_info(element
, ms_parent_info
):
2706 ms_info
= ms_parent_info
.copy()
2708 # As per [1, 5.3.9.2.2] SegmentList and SegmentTemplate share some
2709 # common attributes and elements. We will only extract relevant
2711 def extract_common(source
):
2712 segment_timeline
= source
.find(_add_ns('SegmentTimeline'))
2713 if segment_timeline
is not None:
2714 s_e
= segment_timeline
.findall(_add_ns('S'))
2716 ms_info
['total_number'] = 0
2719 r
= int(s
.get('r', 0))
2720 ms_info
['total_number'] += 1 + r
2721 ms_info
['s'].append({
2722 't': int(s
.get('t', 0)),
2723 # @d is mandatory (see [1, 5.3.9.6.2, Table 17, page 60])
2724 'd': int(s
.attrib
['d']),
2727 start_number
= source
.get('startNumber')
2729 ms_info
['start_number'] = int(start_number
)
2730 timescale
= source
.get('timescale')
2732 ms_info
['timescale'] = int(timescale
)
2733 segment_duration
= source
.get('duration')
2734 if segment_duration
:
2735 ms_info
['segment_duration'] = float(segment_duration
)
2737 def extract_Initialization(source
):
2738 initialization
= source
.find(_add_ns('Initialization'))
2739 if initialization
is not None:
2740 ms_info
['initialization_url'] = initialization
.attrib
['sourceURL']
2742 segment_list
= element
.find(_add_ns('SegmentList'))
2743 if segment_list
is not None:
2744 extract_common(segment_list
)
2745 extract_Initialization(segment_list
)
2746 segment_urls_e
= segment_list
.findall(_add_ns('SegmentURL'))
2748 ms_info
['segment_urls'] = [segment
.attrib
['media'] for segment
in segment_urls_e
]
2750 segment_template
= element
.find(_add_ns('SegmentTemplate'))
2751 if segment_template
is not None:
2752 extract_common(segment_template
)
2753 media
= segment_template
.get('media')
2755 ms_info
['media'] = media
2756 initialization
= segment_template
.get('initialization')
2758 ms_info
['initialization'] = initialization
2760 extract_Initialization(segment_template
)
2763 mpd_duration
= parse_duration(mpd_doc
.get('mediaPresentationDuration'))
2764 formats
, subtitles
= [], {}
2765 stream_numbers
= collections
.defaultdict(int)
2766 for period
in mpd_doc
.findall(_add_ns('Period')):
2767 period_duration
= parse_duration(period
.get('duration')) or mpd_duration
2768 period_ms_info
= extract_multisegment_info(period
, {
2772 for adaptation_set
in period
.findall(_add_ns('AdaptationSet')):
2773 adaption_set_ms_info
= extract_multisegment_info(adaptation_set
, period_ms_info
)
2774 for representation
in adaptation_set
.findall(_add_ns('Representation')):
2775 representation_attrib
= adaptation_set
.attrib
.copy()
2776 representation_attrib
.update(representation
.attrib
)
2777 # According to [1, 5.3.7.2, Table 9, page 41], @mimeType is mandatory
2778 mime_type
= representation_attrib
['mimeType']
2779 content_type
= representation_attrib
.get('contentType', mime_type
.split('/')[0])
2781 codec_str
= representation_attrib
.get('codecs', '')
2782 # Some kind of binary subtitle found in some youtube livestreams
2783 if mime_type
== 'application/x-rawcc':
2784 codecs
= {'scodec': codec_str}
2786 codecs
= parse_codecs(codec_str
)
2787 if content_type
not in ('video', 'audio', 'text'):
2788 if mime_type
== 'image/jpeg':
2789 content_type
= mime_type
2790 elif codecs
.get('vcodec', 'none') != 'none':
2791 content_type
= 'video'
2792 elif codecs
.get('acodec', 'none') != 'none':
2793 content_type
= 'audio'
2794 elif codecs
.get('scodec', 'none') != 'none':
2795 content_type
= 'text'
2796 elif mimetype2ext(mime_type
) in ('tt', 'dfxp', 'ttml', 'xml', 'json'):
2797 content_type
= 'text'
2799 self
.report_warning('Unknown MIME type %s in DASH manifest' % mime_type
)
2803 for element
in (representation
, adaptation_set
, period
, mpd_doc
):
2804 base_url_e
= element
.find(_add_ns('BaseURL'))
2805 if base_url_e
is not None:
2806 base_url
= base_url_e
.text
+ base_url
2807 if re
.match(r
'^https?://', base_url
):
2809 if mpd_base_url
and base_url
.startswith('/'):
2810 base_url
= compat_urlparse
.urljoin(mpd_base_url
, base_url
)
2811 elif mpd_base_url
and not re
.match(r
'^https?://', base_url
):
2812 if not mpd_base_url
.endswith('/'):
2814 base_url
= mpd_base_url
+ base_url
2815 representation_id
= representation_attrib
.get('id')
2816 lang
= representation_attrib
.get('lang')
2817 url_el
= representation
.find(_add_ns('BaseURL'))
2818 filesize
= int_or_none(url_el
.attrib
.get('{http://youtube.com/yt/2012/10/10}contentLength') if url_el
is not None else None)
2819 bandwidth
= int_or_none(representation_attrib
.get('bandwidth'))
2820 if representation_id
is not None:
2821 format_id
= representation_id
2823 format_id
= content_type
2825 format_id
= mpd_id
+ '-' + format_id
2826 if content_type
in ('video', 'audio'):
2828 'format_id': format_id
,
2829 'manifest_url': mpd_url
,
2830 'ext': mimetype2ext(mime_type
),
2831 'width': int_or_none(representation_attrib
.get('width')),
2832 'height': int_or_none(representation_attrib
.get('height')),
2833 'tbr': float_or_none(bandwidth
, 1000),
2834 'asr': int_or_none(representation_attrib
.get('audioSamplingRate')),
2835 'fps': int_or_none(representation_attrib
.get('frameRate')),
2836 'language': lang
if lang
not in ('mul', 'und', 'zxx', 'mis') else None,
2837 'format_note': 'DASH %s' % content_type
,
2838 'filesize': filesize
,
2839 'container': mimetype2ext(mime_type
) + '_dash',
2842 elif content_type
== 'text':
2844 'ext': mimetype2ext(mime_type
),
2845 'manifest_url': mpd_url
,
2846 'filesize': filesize
,
2848 elif content_type
== 'image/jpeg':
2849 # See test case in VikiIE
2850 # https://www.viki.com/videos/1175236v-choosing-spouse-by-lottery-episode-1
2852 'format_id': format_id
,
2854 'manifest_url': mpd_url
,
2855 'format_note': 'DASH storyboards (jpeg)',
2859 if is_drm_protected(adaptation_set
) or is_drm_protected(representation
):
2861 representation_ms_info
= extract_multisegment_info(representation
, adaption_set_ms_info
)
2863 def prepare_template(template_name
, identifiers
):
2864 tmpl
= representation_ms_info
[template_name
]
2865 # First of, % characters outside $...$ templates
2866 # must be escaped by doubling for proper processing
2867 # by % operator string formatting used further (see
2868 # https://github.com/ytdl-org/youtube-dl/issues/16867).
2874 in_template
= not in_template
2875 elif c
== '%' and not in_template
:
2877 # Next, $...$ templates are translated to their
2878 # %(...) counterparts to be used with % operator
2879 if representation_id
is not None:
2880 t
= t
.replace('$RepresentationID$', representation_id
)
2881 t
= re
.sub(r
'\$(%s)\$' % '|'.join(identifiers
), r
'%(\1)d', t
)
2882 t
= re
.sub(r
'\$(%s)%%([^$]+)\$' % '|'.join(identifiers
), r
'%(\1)\2', t
)
2883 t
.replace('$$', '$')
2886 # @initialization is a regular template like @media one
2887 # so it should be handled just the same way (see
2888 # https://github.com/ytdl-org/youtube-dl/issues/11605)
2889 if 'initialization' in representation_ms_info
:
2890 initialization_template
= prepare_template(
2892 # As per [1, 5.3.9.4.2, Table 15, page 54] $Number$ and
2893 # $Time$ shall not be included for @initialization thus
2894 # only $Bandwidth$ remains
2896 representation_ms_info
['initialization_url'] = initialization_template
% {
2897 'Bandwidth': bandwidth
,
2900 def location_key(location
):
2901 return 'url' if re
.match(r
'^https?://', location
) else 'path'
2903 if 'segment_urls' not in representation_ms_info
and 'media' in representation_ms_info
:
2905 media_template
= prepare_template('media', ('Number', 'Bandwidth', 'Time'))
2906 media_location_key
= location_key(media_template
)
2908 # As per [1, 5.3.9.4.4, Table 16, page 55] $Number$ and $Time$
2909 # can't be used at the same time
2910 if '%(Number' in media_template
and 's' not in representation_ms_info
:
2911 segment_duration
= None
2912 if 'total_number' not in representation_ms_info
and 'segment_duration' in representation_ms_info
:
2913 segment_duration
= float_or_none(representation_ms_info
['segment_duration'], representation_ms_info
['timescale'])
2914 representation_ms_info
['total_number'] = int(math
.ceil(
2915 float_or_none(period_duration
, segment_duration
, default
=0)))
2916 representation_ms_info
['fragments'] = [{
2917 media_location_key
: media_template
% {
2918 'Number': segment_number
,
2919 'Bandwidth': bandwidth
,
2921 'duration': segment_duration
,
2922 } for segment_number
in range(
2923 representation_ms_info
['start_number'],
2924 representation_ms_info
['total_number'] + representation_ms_info
['start_number'])]
2926 # $Number*$ or $Time$ in media template with S list available
2927 # Example $Number*$: http://www.svtplay.se/klipp/9023742/stopptid-om-bjorn-borg
2928 # Example $Time$: https://play.arkena.com/embed/avp/v2/player/media/b41dda37-d8e7-4d3f-b1b5-9a9db578bdfe/1/129411
2929 representation_ms_info
['fragments'] = []
2932 segment_number
= representation_ms_info
['start_number']
2934 def add_segment_url():
2935 segment_url
= media_template
% {
2936 'Time': segment_time
,
2937 'Bandwidth': bandwidth
,
2938 'Number': segment_number
,
2940 representation_ms_info
['fragments'].append({
2941 media_location_key
: segment_url
,
2942 'duration': float_or_none(segment_d
, representation_ms_info
['timescale']),
2945 for num
, s
in enumerate(representation_ms_info
['s']):
2946 segment_time
= s
.get('t') or segment_time
2950 for r
in range(s
.get('r', 0)):
2951 segment_time
+= segment_d
2954 segment_time
+= segment_d
2955 elif 'segment_urls' in representation_ms_info
and 's' in representation_ms_info
:
2957 # Example: https://www.youtube.com/watch?v=iXZV5uAYMJI
2958 # or any YouTube dashsegments video
2961 timescale
= representation_ms_info
['timescale']
2962 for s
in representation_ms_info
['s']:
2963 duration
= float_or_none(s
['d'], timescale
)
2964 for r
in range(s
.get('r', 0) + 1):
2965 segment_uri
= representation_ms_info
['segment_urls'][segment_index
]
2967 location_key(segment_uri
): segment_uri
,
2968 'duration': duration
,
2971 representation_ms_info
['fragments'] = fragments
2972 elif 'segment_urls' in representation_ms_info
:
2973 # Segment URLs with no SegmentTimeline
2974 # Example: https://www.seznam.cz/zpravy/clanek/cesko-zasahne-vitr-o-sile-vichrice-muze-byt-i-zivotu-nebezpecny-39091
2975 # https://github.com/ytdl-org/youtube-dl/pull/14844
2977 segment_duration
= float_or_none(
2978 representation_ms_info
['segment_duration'],
2979 representation_ms_info
['timescale']) if 'segment_duration' in representation_ms_info
else None
2980 for segment_url
in representation_ms_info
['segment_urls']:
2982 location_key(segment_url
): segment_url
,
2984 if segment_duration
:
2985 fragment
['duration'] = segment_duration
2986 fragments
.append(fragment
)
2987 representation_ms_info
['fragments'] = fragments
2988 # If there is a fragments key available then we correctly recognized fragmented media.
2989 # Otherwise we will assume unfragmented media with direct access. Technically, such
2990 # assumption is not necessarily correct since we may simply have no support for
2991 # some forms of fragmented media renditions yet, but for now we'll use this fallback.
2992 if 'fragments' in representation_ms_info
:
2994 # NB: mpd_url may be empty when MPD manifest is parsed from a string
2995 'url': mpd_url
or base_url
,
2996 'fragment_base_url': base_url
,
2998 'protocol': 'http_dash_segments' if mime_type
!= 'image/jpeg' else 'mhtml',
3000 if 'initialization_url' in representation_ms_info
:
3001 initialization_url
= representation_ms_info
['initialization_url']
3002 if not f
.get('url'):
3003 f
['url'] = initialization_url
3004 f
['fragments'].append({location_key(initialization_url): initialization_url}
)
3005 f
['fragments'].extend(representation_ms_info
['fragments'])
3006 if not period_duration
:
3007 period_duration
= try_get(
3008 representation_ms_info
,
3009 lambda r
: sum(frag
['duration'] for frag
in r
['fragments']), float)
3011 # Assuming direct URL to unfragmented media.
3013 if content_type
in ('video', 'audio', 'image/jpeg'):
3014 f
['manifest_stream_number'] = stream_numbers
[f
['url']]
3015 stream_numbers
[f
['url']] += 1
3017 elif content_type
== 'text':
3018 subtitles
.setdefault(lang
or 'und', []).append(f
)
3020 return formats
, subtitles
3022 def _extract_ism_formats(self
, *args
, **kwargs
):
3023 fmts
, subs
= self
._extract
_ism
_formats
_and
_subtitles
(*args
, **kwargs
)
3025 self
._report
_ignoring
_subs
('ISM')
3028 def _extract_ism_formats_and_subtitles(self
, ism_url
, video_id
, ism_id
=None, note
=None, errnote
=None, fatal
=True, data
=None, headers
={}, query={}
):
3029 res
= self
._download
_xml
_handle
(
3031 note
='Downloading ISM manifest' if note
is None else note
,
3032 errnote
='Failed to download ISM manifest' if errnote
is None else errnote
,
3033 fatal
=fatal
, data
=data
, headers
=headers
, query
=query
)
3040 return self
._parse
_ism
_formats
_and
_subtitles
(ism_doc
, urlh
.geturl(), ism_id
)
3042 def _parse_ism_formats_and_subtitles(self
, ism_doc
, ism_url
, ism_id
=None):
3044 Parse formats from ISM manifest.
3046 1. [MS-SSTR]: Smooth Streaming Protocol,
3047 https://msdn.microsoft.com/en-us/library/ff469518.aspx
3049 if ism_doc
.get('IsLive') == 'TRUE':
3052 duration
= int(ism_doc
.attrib
['Duration'])
3053 timescale
= int_or_none(ism_doc
.get('TimeScale')) or 10000000
3057 for stream
in ism_doc
.findall('StreamIndex'):
3058 stream_type
= stream
.get('Type')
3059 if stream_type
not in ('video', 'audio', 'text'):
3061 url_pattern
= stream
.attrib
['Url']
3062 stream_timescale
= int_or_none(stream
.get('TimeScale')) or timescale
3063 stream_name
= stream
.get('Name')
3064 stream_language
= stream
.get('Language', 'und')
3065 for track
in stream
.findall('QualityLevel'):
3066 fourcc
= track
.get('FourCC') or ('AACL' if track
.get('AudioTag') == '255' else None)
3067 # TODO: add support for WVC1 and WMAP
3068 if fourcc
not in ('H264', 'AVC1', 'AACL', 'TTML'):
3069 self
.report_warning('%s is not a supported codec' % fourcc
)
3071 tbr
= int(track
.attrib
['Bitrate']) // 1000
3072 # [1] does not mention Width and Height attributes. However,
3073 # they're often present while MaxWidth and MaxHeight are
3074 # missing, so should be used as fallbacks
3075 width
= int_or_none(track
.get('MaxWidth') or track
.get('Width'))
3076 height
= int_or_none(track
.get('MaxHeight') or track
.get('Height'))
3077 sampling_rate
= int_or_none(track
.get('SamplingRate'))
3079 track_url_pattern
= re
.sub(r
'{[Bb]itrate}', track
.attrib
['Bitrate'], url_pattern
)
3080 track_url_pattern
= compat_urlparse
.urljoin(ism_url
, track_url_pattern
)
3086 stream_fragments
= stream
.findall('c')
3087 for stream_fragment_index
, stream_fragment
in enumerate(stream_fragments
):
3088 fragment_ctx
['time'] = int_or_none(stream_fragment
.get('t')) or fragment_ctx
['time']
3089 fragment_repeat
= int_or_none(stream_fragment
.get('r')) or 1
3090 fragment_ctx
['duration'] = int_or_none(stream_fragment
.get('d'))
3091 if not fragment_ctx
['duration']:
3093 next_fragment_time
= int(stream_fragment
[stream_fragment_index
+ 1].attrib
['t'])
3095 next_fragment_time
= duration
3096 fragment_ctx
['duration'] = (next_fragment_time
- fragment_ctx
['time']) / fragment_repeat
3097 for _
in range(fragment_repeat
):
3099 'url': re
.sub(r
'{start[ _]time}', compat_str(fragment_ctx
['time']), track_url_pattern
),
3100 'duration': fragment_ctx
['duration'] / stream_timescale
,
3102 fragment_ctx
['time'] += fragment_ctx
['duration']
3104 if stream_type
== 'text':
3105 subtitles
.setdefault(stream_language
, []).append({
3109 'manifest_url': ism_url
,
3110 'fragments': fragments
,
3111 '_download_params': {
3112 'stream_type': stream_type
,
3113 'duration': duration
,
3114 'timescale': stream_timescale
,
3116 'language': stream_language
,
3117 'codec_private_data': track
.get('CodecPrivateData'),
3120 elif stream_type
in ('video', 'audio'):
3122 'format_id': join_nonempty(ism_id
, stream_name
, tbr
),
3124 'manifest_url': ism_url
,
3125 'ext': 'ismv' if stream_type
== 'video' else 'isma',
3129 'asr': sampling_rate
,
3130 'vcodec': 'none' if stream_type
== 'audio' else fourcc
,
3131 'acodec': 'none' if stream_type
== 'video' else fourcc
,
3133 'fragments': fragments
,
3134 'has_drm': ism_doc
.find('Protection') is not None,
3135 '_download_params': {
3136 'stream_type': stream_type
,
3137 'duration': duration
,
3138 'timescale': stream_timescale
,
3139 'width': width
or 0,
3140 'height': height
or 0,
3142 'language': stream_language
,
3143 'codec_private_data': track
.get('CodecPrivateData'),
3144 'sampling_rate': sampling_rate
,
3145 'channels': int_or_none(track
.get('Channels', 2)),
3146 'bits_per_sample': int_or_none(track
.get('BitsPerSample', 16)),
3147 'nal_unit_length_field': int_or_none(track
.get('NALUnitLengthField', 4)),
3150 return formats
, subtitles
3152 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):
3153 def absolute_url(item_url
):
3154 return urljoin(base_url
, item_url
)
3156 def parse_content_type(content_type
):
3157 if not content_type
:
3159 ctr
= re
.search(r
'(?P<mimetype>[^/]+/[^;]+)(?:;\s*codecs="?(?P<codecs>[^"]+))?', content_type
)
3161 mimetype
, codecs
= ctr
.groups()
3162 f
= parse_codecs(codecs
)
3163 f
['ext'] = mimetype2ext(mimetype
)
3167 def _media_formats(src
, cur_media_type
, type_info
=None):
3168 type_info
= type_info
or {}
3169 full_url
= absolute_url(src
)
3170 ext
= type_info
.get('ext') or determine_ext(full_url
)
3172 is_plain_url
= False
3173 formats
= self
._extract
_m
3u8_formats
(
3174 full_url
, video_id
, ext
='mp4',
3175 entry_protocol
=m3u8_entry_protocol
, m3u8_id
=m3u8_id
,
3176 preference
=preference
, quality
=quality
, fatal
=False)
3178 is_plain_url
= False
3179 formats
= self
._extract
_mpd
_formats
(
3180 full_url
, video_id
, mpd_id
=mpd_id
, fatal
=False)
3185 'vcodec': 'none' if cur_media_type
== 'audio' else None,
3188 return is_plain_url
, formats
3191 # amp-video and amp-audio are very similar to their HTML5 counterparts
3192 # so we wll include them right here (see
3193 # https://www.ampproject.org/docs/reference/components/amp-video)
3194 # For dl8-* tags see https://delight-vr.com/documentation/dl8-video/
3195 _MEDIA_TAG_NAME_RE
= r
'(?:(?:amp|dl8(?:-live)?)-)?(video|audio)'
3196 media_tags
= [(media_tag
, media_tag_name
, media_type
, '')
3197 for media_tag
, media_tag_name
, media_type
3198 in re
.findall(r
'(?s)(<(%s)[^>]*/>)' % _MEDIA_TAG_NAME_RE
, webpage
)]
3199 media_tags
.extend(re
.findall(
3200 # We only allow video|audio followed by a whitespace or '>'.
3201 # Allowing more characters may end up in significant slow down (see
3202 # https://github.com/ytdl-org/youtube-dl/issues/11979, example URL:
3203 # http://www.porntrex.com/maps/videositemap.xml).
3204 r
'(?s)(<(?P<tag>%s)(?:\s+[^>]*)?>)(.*?)</(?P=tag)>' % _MEDIA_TAG_NAME_RE
, webpage
))
3205 for media_tag
, _
, media_type
, media_content
in media_tags
:
3210 media_attributes
= extract_attributes(media_tag
)
3211 src
= strip_or_none(media_attributes
.get('src'))
3213 f
= parse_content_type(media_attributes
.get('type'))
3214 _
, formats
= _media_formats(src
, media_type
, f
)
3215 media_info
['formats'].extend(formats
)
3216 media_info
['thumbnail'] = absolute_url(media_attributes
.get('poster'))
3218 for source_tag
in re
.findall(r
'<source[^>]+>', media_content
):
3219 s_attr
= extract_attributes(source_tag
)
3220 # data-video-src and data-src are non standard but seen
3221 # several times in the wild
3222 src
= strip_or_none(dict_get(s_attr
, ('src', 'data-video-src', 'data-src')))
3225 f
= parse_content_type(s_attr
.get('type'))
3226 is_plain_url
, formats
= _media_formats(src
, media_type
, f
)
3228 # width, height, res, label and title attributes are
3229 # all not standard but seen several times in the wild
3232 for lbl
in ('label', 'title')
3233 if str_or_none(s_attr
.get(lbl
))
3235 width
= int_or_none(s_attr
.get('width'))
3236 height
= (int_or_none(s_attr
.get('height'))
3237 or int_or_none(s_attr
.get('res')))
3238 if not width
or not height
:
3240 resolution
= parse_resolution(lbl
)
3243 width
= width
or resolution
.get('width')
3244 height
= height
or resolution
.get('height')
3246 tbr
= parse_bitrate(lbl
)
3255 'format_id': s_attr
.get('label') or s_attr
.get('title'),
3257 f
.update(formats
[0])
3258 media_info
['formats'].append(f
)
3260 media_info
['formats'].extend(formats
)
3261 for track_tag
in re
.findall(r
'<track[^>]+>', media_content
):
3262 track_attributes
= extract_attributes(track_tag
)
3263 kind
= track_attributes
.get('kind')
3264 if not kind
or kind
in ('subtitles', 'captions'):
3265 src
= strip_or_none(track_attributes
.get('src'))
3268 lang
= track_attributes
.get('srclang') or track_attributes
.get('lang') or track_attributes
.get('label')
3269 media_info
['subtitles'].setdefault(lang
, []).append({
3270 'url': absolute_url(src
),
3272 for f
in media_info
['formats']:
3273 f
.setdefault('http_headers', {})['Referer'] = base_url
3274 if media_info
['formats'] or media_info
['subtitles']:
3275 entries
.append(media_info
)
3278 def _extract_akamai_formats(self
, *args
, **kwargs
):
3279 fmts
, subs
= self
._extract
_akamai
_formats
_and
_subtitles
(*args
, **kwargs
)
3281 self
._report
_ignoring
_subs
('akamai')
3284 def _extract_akamai_formats_and_subtitles(self
, manifest_url
, video_id
, hosts
={}):
3285 signed
= 'hdnea=' in manifest_url
3287 # https://learn.akamai.com/en-us/webhelp/media-services-on-demand/stream-packaging-user-guide/GUID-BE6C0F73-1E06-483B-B0EA-57984B91B7F9.html
3288 manifest_url
= re
.sub(
3289 r
'(?:b=[\d,-]+|(?:__a__|attributes)=off|__b__=\d+)&?',
3290 '', manifest_url
).strip('?')
3295 hdcore_sign
= 'hdcore=3.7.0'
3296 f4m_url
= re
.sub(r
'(https?://[^/]+)/i/', r
'\1/z/', manifest_url
).replace('/master.m3u8', '/manifest.f4m')
3297 hds_host
= hosts
.get('hds')
3299 f4m_url
= re
.sub(r
'(https?://)[^/]+', r
'\1' + hds_host
, f4m_url
)
3300 if 'hdcore=' not in f4m_url
:
3301 f4m_url
+= ('&' if '?' in f4m_url
else '?') + hdcore_sign
3302 f4m_formats
= self
._extract
_f
4m
_formats
(
3303 f4m_url
, video_id
, f4m_id
='hds', fatal
=False)
3304 for entry
in f4m_formats
:
3305 entry
.update({'extra_param_to_segment_url': hdcore_sign}
)
3306 formats
.extend(f4m_formats
)
3308 m3u8_url
= re
.sub(r
'(https?://[^/]+)/z/', r
'\1/i/', manifest_url
).replace('/manifest.f4m', '/master.m3u8')
3309 hls_host
= hosts
.get('hls')
3311 m3u8_url
= re
.sub(r
'(https?://)[^/]+', r
'\1' + hls_host
, m3u8_url
)
3312 m3u8_formats
, m3u8_subtitles
= self
._extract
_m
3u8_formats
_and
_subtitles
(
3313 m3u8_url
, video_id
, 'mp4', 'm3u8_native',
3314 m3u8_id
='hls', fatal
=False)
3315 formats
.extend(m3u8_formats
)
3316 subtitles
= self
._merge
_subtitles
(subtitles
, m3u8_subtitles
)
3318 http_host
= hosts
.get('http')
3319 if http_host
and m3u8_formats
and not signed
:
3320 REPL_REGEX
= r
'https?://[^/]+/i/([^,]+),([^/]+),([^/]+)\.csmil/.+'
3321 qualities
= re
.match(REPL_REGEX
, m3u8_url
).group(2).split(',')
3322 qualities_length
= len(qualities
)
3323 if len(m3u8_formats
) in (qualities_length
, qualities_length
+ 1):
3325 for f
in m3u8_formats
:
3326 if f
['vcodec'] != 'none':
3327 for protocol
in ('http', 'https'):
3329 del http_f
['manifest_url']
3331 REPL_REGEX
, protocol
+ fr
'://{http_host}/\g<1>{qualities[i]}\3', f
['url'])
3333 'format_id': http_f
['format_id'].replace('hls-', protocol
+ '-'),
3335 'protocol': protocol
,
3337 formats
.append(http_f
)
3340 return formats
, subtitles
3342 def _extract_wowza_formats(self
, url
, video_id
, m3u8_entry_protocol
='m3u8_native', skip_protocols
=[]):
3343 query
= compat_urlparse
.urlparse(url
).query
3344 url
= re
.sub(r
'/(?:manifest|playlist|jwplayer)\.(?:m3u8|f4m|mpd|smil)', '', url
)
3346 r
'(?:(?:http|rtmp|rtsp)(?P<s>s)?:)?(?P<url>//[^?]+)', url
)
3347 url_base
= mobj
.group('url')
3348 http_base_url
= '%s%s:%s' % ('http', mobj
.group('s') or '', url_base
)
3351 def manifest_url(manifest
):
3352 m_url
= f
'{http_base_url}/{manifest}'
3354 m_url
+= '?%s' % query
3357 if 'm3u8' not in skip_protocols
:
3358 formats
.extend(self
._extract
_m
3u8_formats
(
3359 manifest_url('playlist.m3u8'), video_id
, 'mp4',
3360 m3u8_entry_protocol
, m3u8_id
='hls', fatal
=False))
3361 if 'f4m' not in skip_protocols
:
3362 formats
.extend(self
._extract
_f
4m
_formats
(
3363 manifest_url('manifest.f4m'),
3364 video_id
, f4m_id
='hds', fatal
=False))
3365 if 'dash' not in skip_protocols
:
3366 formats
.extend(self
._extract
_mpd
_formats
(
3367 manifest_url('manifest.mpd'),
3368 video_id
, mpd_id
='dash', fatal
=False))
3369 if re
.search(r
'(?:/smil:|\.smil)', url_base
):
3370 if 'smil' not in skip_protocols
:
3371 rtmp_formats
= self
._extract
_smil
_formats
(
3372 manifest_url('jwplayer.smil'),
3373 video_id
, fatal
=False)
3374 for rtmp_format
in rtmp_formats
:
3375 rtsp_format
= rtmp_format
.copy()
3376 rtsp_format
['url'] = '%s/%s' % (rtmp_format
['url'], rtmp_format
['play_path'])
3377 del rtsp_format
['play_path']
3378 del rtsp_format
['ext']
3379 rtsp_format
.update({
3380 'url': rtsp_format
['url'].replace('rtmp://', 'rtsp://'),
3381 'format_id': rtmp_format
['format_id'].replace('rtmp', 'rtsp'),
3384 formats
.extend([rtmp_format
, rtsp_format
])
3386 for protocol
in ('rtmp', 'rtsp'):
3387 if protocol
not in skip_protocols
:
3389 'url': f
'{protocol}:{url_base}',
3390 'format_id': protocol
,
3391 'protocol': protocol
,
3395 def _find_jwplayer_data(self
, webpage
, video_id
=None, transform_source
=js_to_json
):
3397 r
'(?s)jwplayer\((?P<quote>[\'"])[^\'" ]+(?P
=quote
)\
)(?
!</script
>).*?\
.setup\s
*\
((?P
<options
>[^
)]+)\
)',
3401 jwplayer_data = self._parse_json(mobj.group('options
'),
3403 transform_source=transform_source)
3404 except ExtractorError:
3407 if isinstance(jwplayer_data, dict):
3408 return jwplayer_data
3410 def _extract_jwplayer_data(self, webpage, video_id, *args, **kwargs):
3411 jwplayer_data = self._find_jwplayer_data(
3412 webpage, video_id, transform_source=js_to_json)
3413 return self._parse_jwplayer_data(
3414 jwplayer_data, video_id, *args, **kwargs)
3416 def _parse_jwplayer_data(self, jwplayer_data, video_id=None, require_title=True,
3417 m3u8_id=None, mpd_id=None, rtmp_params=None, base_url=None):
3418 # JWPlayer backward compatibility: flattened playlists
3419 # https://github.com/jwplayer/jwplayer/blob/v7.4.3/src/js/api/config.js#L81-L96
3420 if 'playlist
' not in jwplayer_data:
3421 jwplayer_data = {'playlist': [jwplayer_data]}
3425 # JWPlayer backward compatibility: single playlist item
3426 # https://github.com/jwplayer/jwplayer/blob/v7.7.0/src/js/playlist/playlist.js#L10
3427 if not isinstance(jwplayer_data['playlist
'], list):
3428 jwplayer_data['playlist
'] = [jwplayer_data['playlist
']]
3430 for video_data in jwplayer_data['playlist
']:
3431 # JWPlayer backward compatibility: flattened sources
3432 # https://github.com/jwplayer/jwplayer/blob/v7.4.3/src/js/playlist/item.js#L29-L35
3433 if 'sources
' not in video_data:
3434 video_data['sources
'] = [video_data]
3436 this_video_id = video_id or video_data['mediaid
']
3438 formats = self._parse_jwplayer_formats(
3439 video_data['sources
'], video_id=this_video_id, m3u8_id=m3u8_id,
3440 mpd_id=mpd_id, rtmp_params=rtmp_params, base_url=base_url)
3443 tracks = video_data.get('tracks
')
3444 if tracks and isinstance(tracks, list):
3445 for track in tracks:
3446 if not isinstance(track, dict):
3448 track_kind = track.get('kind
')
3449 if not track_kind or not isinstance(track_kind, compat_str):
3451 if track_kind.lower() not in ('captions
', 'subtitles
'):
3453 track_url = urljoin(base_url, track.get('file'))
3456 subtitles.setdefault(track.get('label
') or 'en
', []).append({
3457 'url
': self._proto_relative_url(track_url)
3461 'id': this_video_id,
3462 'title
': unescapeHTML(video_data['title
'] if require_title else video_data.get('title
')),
3463 'description
': clean_html(video_data.get('description
')),
3464 'thumbnail
': urljoin(base_url, self._proto_relative_url(video_data.get('image
'))),
3465 'timestamp
': int_or_none(video_data.get('pubdate
')),
3466 'duration
': float_or_none(jwplayer_data.get('duration
') or video_data.get('duration
')),
3467 'subtitles
': subtitles,
3469 # https://github.com/jwplayer/jwplayer/blob/master/src/js/utils/validator.js#L32
3470 if len(formats) == 1 and re.search(r'^
(?
:http|
//).*(?
:youtube\
.com|youtu\
.be
)/.+', formats[0]['url
']):
3472 '_type
': 'url_transparent
',
3473 'url
': formats[0]['url
'],
3476 self._sort_formats(formats)
3477 entry['formats
'] = formats
3478 entries.append(entry)
3479 if len(entries) == 1:
3482 return self.playlist_result(entries)
3484 def _parse_jwplayer_formats(self, jwplayer_sources_data, video_id=None,
3485 m3u8_id=None, mpd_id=None, rtmp_params=None, base_url=None):
3488 for source in jwplayer_sources_data:
3489 if not isinstance(source, dict):
3491 source_url = urljoin(
3492 base_url, self._proto_relative_url(source.get('file')))
3493 if not source_url or source_url in urls:
3495 urls.append(source_url)
3496 source_type = source.get('type') or ''
3497 ext = mimetype2ext(source_type) or determine_ext(source_url)
3498 if source_type == 'hls
' or ext == 'm3u8
':
3499 formats.extend(self._extract_m3u8_formats(
3500 source_url, video_id, 'mp4
', entry_protocol='m3u8_native
',
3501 m3u8_id=m3u8_id, fatal=False))
3502 elif source_type == 'dash
' or ext == 'mpd
':
3503 formats.extend(self._extract_mpd_formats(
3504 source_url, video_id, mpd_id=mpd_id, fatal=False))
3506 formats.extend(self._extract_smil_formats(
3507 source_url, video_id, fatal=False))
3508 # https://github.com/jwplayer/jwplayer/blob/master/src/js/providers/default.js#L67
3509 elif source_type.startswith('audio
') or ext in (
3510 'oga
', 'aac
', 'mp3
', 'mpeg
', 'vorbis
'):
3517 height = int_or_none(source.get('height
'))
3519 # Often no height is provided but there is a label in
3520 # format like "1080p", "720p SD", or 1080.
3521 height = int_or_none(self._search_regex(
3522 r'^
(\d{3,4}
)[pP
]?
(?
:\b|$
)', compat_str(source.get('label
') or ''),
3523 'height
', default=None))
3526 'width
': int_or_none(source.get('width
')),
3528 'tbr
': int_or_none(source.get('bitrate
')),
3531 if source_url.startswith('rtmp
'):
3532 a_format['ext
'] = 'flv
'
3533 # See com/longtailvideo/jwplayer/media/RTMPMediaProvider.as
3534 # of jwplayer.flash.swf
3535 rtmp_url_parts = re.split(
3536 r'((?
:mp4|mp3|flv
):)', source_url, 1)
3537 if len(rtmp_url_parts) == 3:
3538 rtmp_url, prefix, play_path = rtmp_url_parts
3541 'play_path
': prefix + play_path,
3544 a_format.update(rtmp_params)
3545 formats.append(a_format)
3548 def _live_title(self, name):
3549 self._downloader.deprecation_warning('yt_dlp
.InfoExtractor
._live
_title
is deprecated
and does
not work
as expected
')
3552 def _int(self, v, name, fatal=False, **kwargs):
3553 res = int_or_none(v, **kwargs)
3555 msg = f'Failed to extract {name}
: Could
not parse value {v!r}
'
3557 raise ExtractorError(msg)
3559 self.report_warning(msg)
3562 def _float(self, v, name, fatal=False, **kwargs):
3563 res = float_or_none(v, **kwargs)
3565 msg = f'Failed to extract {name}
: Could
not parse value {v!r}
'
3567 raise ExtractorError(msg)
3569 self.report_warning(msg)
3572 def _set_cookie(self, domain, name, value, expire_time=None, port=None,
3573 path='/', secure=False, discard=False, rest={}, **kwargs):
3574 cookie = compat_cookiejar_Cookie(
3575 0, name, value, port, port is not None, domain, True,
3576 domain.startswith('.'), path, True, secure, expire_time,
3577 discard, None, None, rest)
3578 self._downloader.cookiejar.set_cookie(cookie)
3580 def _get_cookies(self, url):
3581 """ Return a compat_cookies_SimpleCookie with the cookies for the url """
3582 return compat_cookies_SimpleCookie(self._downloader._calc_cookies(url))
3584 def _apply_first_set_cookie_header(self, url_handle, cookie):
3586 Apply first Set-Cookie header instead of the last. Experimental.
3588 Some sites (e.g. [1-3]) may serve two cookies under the same name
3589 in Set-Cookie header and expect the first (old) one to be set rather
3590 than second (new). However, as of RFC6265 the newer one cookie
3591 should be set into cookie store what actually happens.
3592 We will workaround this issue by resetting the cookie to
3593 the first one manually.
3594 1. https://new.vk.com/
3595 2. https://github.com/ytdl-org/youtube-dl/issues/9841#issuecomment-227871201
3596 3. https://learning.oreilly.com/
3598 for header, cookies in url_handle.headers.items():
3599 if header.lower() != 'set-cookie
':
3601 cookies = cookies.encode('iso
-8859-1').decode('utf
-8')
3602 cookie_value = re.search(
3603 r'%s=(.+?
);.*?
\b[Dd
]omain
=(.+?
)(?
:[,;]|$
)' % cookie, cookies)
3605 value, domain = cookie_value.groups()
3606 self._set_cookie(domain, cookie, value)
3610 def get_testcases(cls, include_onlymatching=False):
3611 t = getattr(cls, '_TEST
', None)
3613 assert not hasattr(cls, '_TESTS
'), f'{cls.ie_key()}IE has _TEST
and _TESTS
'
3616 tests = getattr(cls, '_TESTS
', [])
3618 if not include_onlymatching and t.get('only_matching
', False):
3620 t['name
'] = cls.ie_key()
3625 """Get age limit from the testcases"""
3626 return max(traverse_obj(
3627 tuple(cls.get_testcases(include_onlymatching=False)),
3628 (..., (('playlist
', 0), None), 'info_dict
', 'age_limit
')) or [0])
3631 def is_suitable(cls, age_limit):
3632 """Test whether the extractor is generally suitable for the given age limit"""
3633 return not age_restricted(cls.age_limit, age_limit)
3636 def description(cls, *, markdown=True, search_examples=None):
3637 """Description of the extractor"""
3639 if cls._NETRC_MACHINE:
3641 desc += f' [<abbr title
="netrc machine"><em
>{cls._NETRC_MACHINE}
</em
></abbr
>]'
3643 desc += f' [{cls._NETRC_MACHINE}
]'
3644 if cls.IE_DESC is False:
3647 desc += f' {cls.IE_DESC}
'
3649 desc += f'; "{cls.SEARCH_KEY}:" prefix
'
3651 _COUNTS = ('', '5', '10', 'all
')
3652 desc += f' (Example
: "{cls.SEARCH_KEY}{random.choice(_COUNTS)}:{random.choice(search_examples)}")'
3653 if not cls.working():
3654 desc += ' (**Currently broken
**)' if markdown else ' (Currently broken
)'
3656 name = f' - **{cls.IE_NAME}
**' if markdown else cls.IE_NAME
3657 return f'{name}
:{desc}
' if desc else name
3659 def extract_subtitles(self, *args, **kwargs):
3660 if (self.get_param('writesubtitles
', False)
3661 or self.get_param('listsubtitles
')):
3662 return self._get_subtitles(*args, **kwargs)
3665 def _get_subtitles(self, *args, **kwargs):
3666 raise NotImplementedError('This method must be implemented by subclasses
')
3668 def extract_comments(self, *args, **kwargs):
3669 if not self.get_param('getcomments
'):
3671 generator = self._get_comments(*args, **kwargs)
3678 comments.append(next(generator))
3679 except StopIteration:
3681 except KeyboardInterrupt:
3682 self.to_screen('Interrupted by user
')
3683 except Exception as e:
3684 if self.get_param('ignoreerrors
') is not True:
3686 self._downloader.report_error(e)
3687 comment_count = len(comments)
3688 self.to_screen(f'Extracted {comment_count} comments
')
3690 'comments
': comments,
3691 'comment_count
': None if interrupted else comment_count
3695 def _get_comments(self, *args, **kwargs):
3696 raise NotImplementedError('This method must be implemented by subclasses
')
3699 def _merge_subtitle_items(subtitle_list1, subtitle_list2):
3700 """ Merge subtitle items for one language. Items with duplicated URLs/data
3701 will be dropped. """
3702 list1_data = {(item.get('url'), item.get('data')) for item in subtitle_list1}
3703 ret = list(subtitle_list1)
3704 ret.extend(item for item in subtitle_list2 if (item.get('url
'), item.get('data
')) not in list1_data)
3708 def _merge_subtitles(cls, *dicts, target=None):
3709 """ Merge subtitle dictionaries, language by language. """
3713 for lang, subs in d.items():
3714 target[lang] = cls._merge_subtitle_items(target.get(lang, []), subs)
3717 def extract_automatic_captions(self, *args, **kwargs):
3718 if (self.get_param('writeautomaticsub
', False)
3719 or self.get_param('listsubtitles
')):
3720 return self._get_automatic_captions(*args, **kwargs)
3723 def _get_automatic_captions(self, *args, **kwargs):
3724 raise NotImplementedError('This method must be implemented by subclasses
')
3726 @functools.cached_property
3727 def _cookies_passed(self):
3728 """Whether cookies have been passed to YoutubeDL"""
3729 return self.get_param('cookiefile
') is not None or self.get_param('cookiesfrombrowser
') is not None
3731 def mark_watched(self, *args, **kwargs):
3732 if not self.get_param('mark_watched
', False):
3734 if self.supports_login() and self._get_login_info()[0] is not None or self._cookies_passed:
3735 self._mark_watched(*args, **kwargs)
3737 def _mark_watched(self, *args, **kwargs):
3738 raise NotImplementedError('This method must be implemented by subclasses
')
3740 def geo_verification_headers(self):
3742 geo_verification_proxy = self.get_param('geo_verification_proxy
')
3743 if geo_verification_proxy:
3744 headers['Ytdl
-request
-proxy
'] = geo_verification_proxy
3747 def _generic_id(self, url):
3748 return compat_urllib_parse_unquote(os.path.splitext(url.rstrip('/').split('/')[-1])[0])
3750 def _generic_title(self, url):
3751 return compat_urllib_parse_unquote(os.path.splitext(url_basename(url))[0])
3754 def _availability(is_private=None, needs_premium=None, needs_subscription=None, needs_auth=None, is_unlisted=None):
3755 all_known = all(map(
3756 lambda x: x is not None,
3757 (is_private, needs_premium, needs_subscription, needs_auth, is_unlisted)))
3759 'private
' if is_private
3760 else 'premium_only
' if needs_premium
3761 else 'subscriber_only
' if needs_subscription
3762 else 'needs_auth
' if needs_auth
3763 else 'unlisted
' if is_unlisted
3764 else 'public
' if all_known
3767 def _configuration_arg(self, key, default=NO_DEFAULT, *, ie_key=None, casesense=False):
3769 @returns A list of values for the extractor argument given by "key"
3770 or "default" if no such key is present
3771 @param default The default value to return when the key is not present (default: [])
3772 @param casesense When false, the values are converted to lower case
3775 self._downloader.params, ('extractor_args
', (ie_key or self.ie_key()).lower(), key))
3777 return [] if default is NO_DEFAULT else default
3778 return list(val) if casesense else [x.lower() for x in val]
3780 def _yes_playlist(self, playlist_id, video_id, smuggled_data=None, *, playlist_label='playlist
', video_label='video
'):
3781 if not playlist_id or not video_id:
3784 no_playlist = (smuggled_data or {}).get('force_noplaylist
')
3785 if no_playlist is not None:
3786 return not no_playlist
3788 video_id = '' if video_id is True else f' {video_id}
'
3789 playlist_id = '' if playlist_id is True else f' {playlist_id}
'
3790 if self.get_param('noplaylist
'):
3791 self.to_screen(f'Downloading just the {video_label}{video_id} because of
--no
-playlist
')
3793 self.to_screen(f'Downloading {playlist_label}{playlist_id}
- add
--no
-playlist to download just the {video_label}{video_id}
')
3797 class SearchInfoExtractor(InfoExtractor):
3799 Base class for paged search queries extractors.
3800 They accept URLs in the format _SEARCH_KEY(|all|[0-9]):{query}
3801 Instances should define _SEARCH_KEY and optionally _MAX_RESULTS
3804 _MAX_RESULTS = float('inf
')
3807 def _make_valid_url(cls):
3808 return r'%s(?P
<prefix
>|
[1-9][0-9]*|all
):(?P
<query
>[\s\S
]+)' % cls._SEARCH_KEY
3810 def _real_extract(self, query):
3811 prefix, query = self._match_valid_url(query).group('prefix
', 'query
')
3813 return self._get_n_results(query, 1)
3814 elif prefix == 'all
':
3815 return self._get_n_results(query, self._MAX_RESULTS)
3819 raise ExtractorError(f'invalid download number {n}
for query
"{query}"')
3820 elif n > self._MAX_RESULTS:
3821 self.report_warning('%s returns
max %i results (you requested
%i)' % (self._SEARCH_KEY, self._MAX_RESULTS, n))
3822 n = self._MAX_RESULTS
3823 return self._get_n_results(query, n)
3825 def _get_n_results(self, query, n):
3826 """Get a specified number of results for a query.
3827 Either this function or _search_results must be overridden by subclasses """
3828 return self.playlist_result(
3829 itertools.islice(self._search_results(query), 0, None if n == float('inf
') else n),
3832 def _search_results(self, query):
3833 """Returns an iterator of search results"""
3834 raise NotImplementedError('This method must be implemented by subclasses
')
3837 def SEARCH_KEY(cls):
3838 return cls._SEARCH_KEY