19 import xml
.etree
.ElementTree
21 from ..compat
import functools
# isort: split
22 from ..compat
import compat_etree_fromstring
, compat_expanduser
, compat_os_name
23 from ..downloader
import FileDownloader
24 from ..downloader
.f4m
import get_base_url
, remove_encrypted_media
59 parse_m3u8_attributes
,
85 """Information Extractor class.
87 Information extractors are the classes that, given a URL, extract
88 information about the video (or videos) the URL refers to. This
89 information includes the real video URL, the video title, author and
90 others. The information is stored in a dictionary which is then
91 passed to the YoutubeDL. The YoutubeDL processes this
92 information possibly downloading the video to the file system, among
93 other possible outcomes.
95 The type field determines the type of the result.
96 By far the most common value (and the default if _type is missing) is
97 "video", which indicates a single video.
99 For a video, the dictionaries must include the following fields:
101 id: Video identifier.
102 title: Video title, unescaped. Set to an empty string if video has
103 no title as opposed to "None" which signifies that the
104 extractor failed to obtain a title
106 Additionally, it must contain either a formats entry or a url one:
108 formats: A list of dictionaries for each format available, ordered
109 from worst to best quality.
112 * url The mandatory URL representing the media:
113 for plain file media - HTTP URL of this file,
115 for HLS - URL of the M3U8 media playlist,
116 for HDS - URL of the F4M manifest,
118 - HTTP URL to plain file media (in case of
120 - URL of the MPD manifest or base URL
121 representing the media if MPD manifest
122 is parsed from a string (in case of
124 for MSS - URL of the ISM manifest.
126 The URL of the manifest file in case of
128 for HLS - URL of the M3U8 master playlist,
129 for HDS - URL of the F4M manifest,
130 for DASH - URL of the MPD manifest,
131 for MSS - URL of the ISM manifest.
132 * manifest_stream_number (For internal use only)
133 The index of the stream in the manifest file
134 * ext Will be calculated from URL if missing
135 * format A human-readable description of the format
136 ("mp4 container with h264/opus").
137 Calculated from the format_id, width, height.
138 and format_note fields if missing.
139 * format_id A short description of the format
140 ("mp4_h264_opus" or "19").
141 Technically optional, but strongly recommended.
142 * format_note Additional info about the format
143 ("3D" or "DASH video")
144 * width Width of the video, if known
145 * height Height of the video, if known
146 * resolution Textual description of width and height
147 * dynamic_range The dynamic range of the video. One of:
148 "SDR" (None), "HDR10", "HDR10+, "HDR12", "HLG, "DV"
149 * tbr Average bitrate of audio and video in KBit/s
150 * abr Average audio bitrate in KBit/s
151 * acodec Name of the audio codec in use
152 * asr Audio sampling rate in Hertz
153 * vbr Average video bitrate in KBit/s
155 * vcodec Name of the video codec in use
156 * container Name of the container format
157 * filesize The number of bytes, if known in advance
158 * filesize_approx An estimate for the number of bytes
159 * player_url SWF Player URL (used for rtmpdump).
160 * protocol The protocol that will be used for the actual
161 download, lower-case. One of "http", "https" or
162 one of the protocols defined in downloader.PROTOCOL_MAP
164 Base URL for fragments. Each fragment's path
165 value (if present) will be relative to
167 * fragments A list of fragments of a fragmented media.
168 Each fragment entry must contain either an url
169 or a path. If an url is present it should be
170 considered by a client. Otherwise both path and
171 fragment_base_url must be present. Here is
172 the list of all potential fields:
173 * "url" - fragment's URL
174 * "path" - fragment's path relative to
176 * "duration" (optional, int or float)
177 * "filesize" (optional, int)
178 * is_from_start Is a live format that can be downloaded
179 from the start. Boolean
180 * preference Order number of this format. If this field is
181 present and not None, the formats get sorted
182 by this field, regardless of all other values.
183 -1 for default (order by other properties),
184 -2 or smaller for less than default.
185 < -1000 to hide the format (if there is
186 another one which is strictly better)
187 * language Language code, e.g. "de" or "en-US".
188 * language_preference Is this in the language mentioned in
190 10 if it's what the URL is about,
191 -1 for default (don't know),
192 -10 otherwise, other values reserved for now.
193 * quality Order number of the video quality of this
194 format, irrespective of the file format.
195 -1 for default (order by other properties),
196 -2 or smaller for less than default.
197 * source_preference Order number for this video source
198 (quality takes higher priority)
199 -1 for default (order by other properties),
200 -2 or smaller for less than default.
201 * http_headers A dictionary of additional HTTP headers
202 to add to the request.
203 * stretched_ratio If given and not 1, indicates that the
204 video's pixels are not square.
205 width : height ratio as float.
206 * no_resume The server does not support resuming the
207 (HTTP or RTMP) download. Boolean.
208 * has_drm The format has DRM and cannot be downloaded. Boolean
209 * downloader_options A dictionary of downloader options
210 (For internal use only)
211 * http_chunk_size Chunk size for HTTP downloads
212 * ffmpeg_args Extra arguments for ffmpeg downloader
213 RTMP formats can also have the additional fields: page_url,
214 app, play_path, tc_url, flash_version, rtmp_live, rtmp_conn,
215 rtmp_protocol, rtmp_real_time
217 url: Final video URL.
218 ext: Video filename extension.
219 format: The video format, defaults to ext (used for --get-format)
220 player_url: SWF Player URL (used for rtmpdump).
222 The following fields are optional:
224 direct: True if a direct video file was given (must only be set by GenericIE)
225 alt_title: A secondary title of the video.
226 display_id An alternative identifier for the video, not necessarily
227 unique, but available before title. Typically, id is
228 something like "4234987", title "Dancing naked mole rats",
229 and display_id "dancing-naked-mole-rats"
230 thumbnails: A list of dictionaries, with the following entries:
231 * "id" (optional, string) - Thumbnail format ID
233 * "preference" (optional, int) - quality of the image
234 * "width" (optional, int)
235 * "height" (optional, int)
236 * "resolution" (optional, string "{width}x{height}",
238 * "filesize" (optional, int)
239 * "http_headers" (dict) - HTTP headers for the request
240 thumbnail: Full URL to a video thumbnail image.
241 description: Full video description.
242 uploader: Full name of the video uploader.
243 license: License name the video is licensed under.
244 creator: The creator of the video.
245 timestamp: UNIX timestamp of the moment the video was uploaded
246 upload_date: Video upload date in UTC (YYYYMMDD).
247 If not explicitly set, calculated from timestamp
248 release_timestamp: UNIX timestamp of the moment the video was released.
249 If it is not clear whether to use timestamp or this, use the former
250 release_date: The date (YYYYMMDD) when the video was released in UTC.
251 If not explicitly set, calculated from release_timestamp
252 modified_timestamp: UNIX timestamp of the moment the video was last modified.
253 modified_date: The date (YYYYMMDD) when the video was last modified in UTC.
254 If not explicitly set, calculated from modified_timestamp
255 uploader_id: Nickname or id of the video uploader.
256 uploader_url: Full URL to a personal webpage of the video uploader.
257 channel: Full name of the channel the video is uploaded on.
258 Note that channel fields may or may not repeat uploader
259 fields. This depends on a particular extractor.
260 channel_id: Id of the channel.
261 channel_url: Full URL to a channel webpage.
262 channel_follower_count: Number of followers of the channel.
263 location: Physical location where the video was filmed.
264 subtitles: The available subtitles as a dictionary in the format
265 {tag: subformats}. "tag" is usually a language code, and
266 "subformats" is a list sorted from lower to higher
267 preference, each element is a dictionary with the "ext"
269 * "data": The subtitles file contents
270 * "url": A URL pointing to the subtitles file
271 It can optionally also have:
272 * "name": Name or description of the subtitles
273 * "http_headers": A dictionary of additional HTTP headers
274 to add to the request.
275 "ext" will be calculated from URL if missing
276 automatic_captions: Like 'subtitles'; contains automatically generated
277 captions instead of normal subtitles
278 duration: Length of the video in seconds, as an integer or float.
279 view_count: How many users have watched the video on the platform.
280 like_count: Number of positive ratings of the video
281 dislike_count: Number of negative ratings of the video
282 repost_count: Number of reposts of the video
283 average_rating: Average rating give by users, the scale used depends on the webpage
284 comment_count: Number of comments on the video
285 comments: A list of comments, each with one or more of the following
286 properties (all but one of text or html optional):
287 * "author" - human-readable name of the comment author
288 * "author_id" - user ID of the comment author
289 * "author_thumbnail" - The thumbnail of the comment author
291 * "html" - Comment as HTML
292 * "text" - Plain text of the comment
293 * "timestamp" - UNIX timestamp of comment
294 * "parent" - ID of the comment this one is replying to.
295 Set to "root" to indicate that this is a
296 comment to the original video.
297 * "like_count" - Number of positive ratings of the comment
298 * "dislike_count" - Number of negative ratings of the comment
299 * "is_favorited" - Whether the comment is marked as
300 favorite by the video uploader
301 * "author_is_uploader" - Whether the comment is made by
303 age_limit: Age restriction for the video, as an integer (years)
304 webpage_url: The URL to the video webpage, if given to yt-dlp it
305 should allow to get the same result again. (It will be set
306 by YoutubeDL if it's missing)
307 categories: A list of categories that the video falls in, for example
309 tags: A list of tags assigned to the video, e.g. ["sweden", "pop music"]
310 cast: A list of the video cast
311 is_live: True, False, or None (=unknown). Whether this video is a
312 live stream that goes on instead of a fixed-length video.
313 was_live: True, False, or None (=unknown). Whether this video was
314 originally a live stream.
315 live_status: None (=unknown), 'is_live', 'is_upcoming', 'was_live', 'not_live'
316 or 'post_live' (was live, but VOD is not yet processed)
317 If absent, automatically set from is_live, was_live
318 start_time: Time in seconds where the reproduction should start, as
319 specified in the URL.
320 end_time: Time in seconds where the reproduction should end, as
321 specified in the URL.
322 chapters: A list of dictionaries, with the following entries:
323 * "start_time" - The start time of the chapter in seconds
324 * "end_time" - The end time of the chapter in seconds
325 * "title" (optional, string)
326 playable_in_embed: Whether this video is allowed to play in embedded
327 players on other sites. Can be True (=always allowed),
328 False (=never allowed), None (=unknown), or a string
329 specifying the criteria for embedability (Eg: 'whitelist')
330 availability: Under what condition the video is available. One of
331 'private', 'premium_only', 'subscriber_only', 'needs_auth',
332 'unlisted' or 'public'. Use 'InfoExtractor._availability'
334 __post_extractor: A function to be called just before the metadata is
335 written to either disk, logger or console. The function
336 must return a dict which will be added to the info_dict.
337 This is usefull for additional information that is
338 time-consuming to extract. Note that the fields thus
339 extracted will not be available to output template and
340 match_filter. So, only "comments" and "comment_count" are
341 currently allowed to be extracted via this method.
343 The following fields should only be used when the video belongs to some logical
346 chapter: Name or title of the chapter the video belongs to.
347 chapter_number: Number of the chapter the video belongs to, as an integer.
348 chapter_id: Id of the chapter the video belongs to, as a unicode string.
350 The following fields should only be used when the video is an episode of some
351 series, programme or podcast:
353 series: Title of the series or programme the video episode belongs to.
354 series_id: Id of the series or programme the video episode belongs to, as a unicode string.
355 season: Title of the season the video episode belongs to.
356 season_number: Number of the season the video episode belongs to, as an integer.
357 season_id: Id of the season the video episode belongs to, as a unicode string.
358 episode: Title of the video episode. Unlike mandatory video title field,
359 this field should denote the exact title of the video episode
360 without any kind of decoration.
361 episode_number: Number of the video episode within a season, as an integer.
362 episode_id: Id of the video episode, as a unicode string.
364 The following fields should only be used when the media is a track or a part of
367 track: Title of the track.
368 track_number: Number of the track within an album or a disc, as an integer.
369 track_id: Id of the track (useful in case of custom indexing, e.g. 6.iii),
371 artist: Artist(s) of the track.
372 genre: Genre(s) of the track.
373 album: Title of the album the track belongs to.
374 album_type: Type of the album (e.g. "Demo", "Full-length", "Split", "Compilation", etc).
375 album_artist: List of all artists appeared on the album (e.g.
376 "Ash Borer / Fell Voices" or "Various Artists", useful for splits
378 disc_number: Number of the disc or other physical medium the track belongs to,
380 release_year: Year (YYYY) when the album was released.
381 composer: Composer of the piece
383 The following fields should only be set for clips that should be cut from the original video:
385 section_start: Start time of the section in seconds
386 section_end: End time of the section in seconds
388 The following fields should only be set for storyboards:
389 rows: Number of rows in each storyboard fragment, as an integer
390 columns: Number of columns in each storyboard fragment, as an integer
392 Unless mentioned otherwise, the fields should be Unicode strings.
394 Unless mentioned otherwise, None is equivalent to absence of information.
397 _type "playlist" indicates multiple videos.
398 There must be a key "entries", which is a list, an iterable, or a PagedList
399 object, each element of which is a valid dictionary by this specification.
401 Additionally, playlists can have "id", "title", and any other relevant
402 attributes with the same semantics as videos (see above).
404 It can also have the following optional fields:
406 playlist_count: The total number of videos in a playlist. If not given,
407 YoutubeDL tries to calculate it from "entries"
410 _type "multi_video" indicates that there are multiple videos that
411 form a single show, for examples multiple acts of an opera or TV episode.
412 It must have an entries key like a playlist and contain all the keys
413 required for a video at the same time.
416 _type "url" indicates that the video must be extracted from another
417 location, possibly by a different extractor. Its only required key is:
418 "url" - the next URL to extract.
419 The key "ie_key" can be set to the class name (minus the trailing "IE",
420 e.g. "Youtube") if the extractor class is known in advance.
421 Additionally, the dictionary may have any properties of the resolved entity
422 known in advance, for example "title" if the title of the referred video is
426 _type "url_transparent" entities have the same specification as "url", but
427 indicate that the given additional information is more precise than the one
428 associated with the resolved URL.
429 This is useful when a site employs a video service that hosts the video and
430 its technical metadata, but that video service does not embed a useful
431 title, description etc.
434 Subclasses of this should define a _VALID_URL regexp and, re-define the
435 _real_extract() and (optionally) _real_initialize() methods.
436 Probably, they should also be added to the list of extractors.
438 Subclasses may also override suitable() if necessary, but ensure the function
439 signature is preserved and that this function imports everything it needs
440 (except other extractors), so that lazy_extractors works correctly.
442 To support username + password (or netrc) login, the extractor must define a
443 _NETRC_MACHINE and re-define _perform_login(username, password) and
444 (optionally) _initialize_pre_login() methods. The _perform_login method will
445 be called between _initialize_pre_login and _real_initialize if credentials
446 are passed by the user. In cases where it is necessary to have the login
447 process as part of the extraction rather than initialization, _perform_login
448 can be left undefined.
450 _GEO_BYPASS attribute may be set to False in order to disable
451 geo restriction bypass mechanisms for a particular extractor.
452 Though it won't disable explicit geo restriction bypass based on
453 country code provided with geo_bypass_country.
455 _GEO_COUNTRIES attribute may contain a list of presumably geo unrestricted
456 countries for this extractor. One of these countries will be used by
457 geo restriction bypass mechanism right away in order to bypass
458 geo restriction, of course, if the mechanism is not disabled.
460 _GEO_IP_BLOCKS attribute may contain a list of presumably geo unrestricted
461 IP blocks in CIDR notation for this extractor. One of these IP blocks
462 will be used by geo restriction bypass mechanism similarly
465 The _WORKING attribute should be set to False for broken IEs
466 in order to warn the users and skip the tests.
471 _x_forwarded_for_ip
= None
473 _GEO_COUNTRIES
= None
474 _GEO_IP_BLOCKS
= None
476 _NETRC_MACHINE
= None
480 def _login_hint(self
, method
=NO_DEFAULT
, netrc
=None):
481 password_hint
= f
'--username and --password, or --netrc ({netrc or self._NETRC_MACHINE}) to provide account credentials'
484 'any': f
'Use --cookies, --cookies-from-browser, {password_hint}',
485 'password': f
'Use {password_hint}',
487 'Use --cookies-from-browser or --cookies for the authentication. '
488 'See https://github.com/ytdl-org/youtube-dl#how-do-i-pass-cookies-to-youtube-dl for how to manually pass cookies'),
489 }[method
if method
is not NO_DEFAULT
else 'any' if self
.supports_login() else 'cookies']
491 def __init__(self
, downloader
=None):
492 """Constructor. Receives an optional downloader (a YoutubeDL instance).
493 If a downloader is not passed during initialization,
494 it must be set using "set_downloader()" before "extract()" is called"""
496 self
._x
_forwarded
_for
_ip
= None
497 self
._printed
_messages
= set()
498 self
.set_downloader(downloader
)
501 def _match_valid_url(cls
, url
):
502 # This does not use has/getattr intentionally - we want to know whether
503 # we have cached the regexp for *this* class, whereas getattr would also
504 # match the superclass
505 if '_VALID_URL_RE' not in cls
.__dict
__:
506 if '_VALID_URL' not in cls
.__dict
__:
507 cls
._VALID
_URL
= cls
._make
_valid
_url
()
508 cls
._VALID
_URL
_RE
= re
.compile(cls
._VALID
_URL
)
509 return cls
._VALID
_URL
_RE
.match(url
)
512 def suitable(cls
, url
):
513 """Receives a URL and returns True if suitable for this IE."""
514 # This function must import everything it needs (except other extractors),
515 # so that lazy_extractors works correctly
516 return cls
._match
_valid
_url
(url
) is not None
519 def _match_id(cls
, url
):
520 return cls
._match
_valid
_url
(url
).group('id')
523 def get_temp_id(cls
, url
):
525 return cls
._match
_id
(url
)
526 except (IndexError, AttributeError):
531 """Getter method for _WORKING."""
535 def supports_login(cls
):
536 return bool(cls
._NETRC
_MACHINE
)
538 def initialize(self
):
539 """Initializes an instance (authentication, etc)."""
540 self
._printed
_messages
= set()
541 self
._initialize
_geo
_bypass
({
542 'countries': self
._GEO
_COUNTRIES
,
543 'ip_blocks': self
._GEO
_IP
_BLOCKS
,
546 self
._initialize
_pre
_login
()
547 if self
.supports_login():
548 username
, password
= self
._get
_login
_info
()
550 self
._perform
_login
(username
, password
)
551 elif self
.get_param('username') and False not in (self
.IE_DESC
, self
._NETRC
_MACHINE
):
552 self
.report_warning(f
'Login with password is not supported for this website. {self._login_hint("cookies")}')
553 self
._real
_initialize
()
556 def _initialize_geo_bypass(self
, geo_bypass_context
):
558 Initialize geo restriction bypass mechanism.
560 This method is used to initialize geo bypass mechanism based on faking
561 X-Forwarded-For HTTP header. A random country from provided country list
562 is selected and a random IP belonging to this country is generated. This
563 IP will be passed as X-Forwarded-For HTTP header in all subsequent
566 This method will be used for initial geo bypass mechanism initialization
567 during the instance initialization with _GEO_COUNTRIES and
570 You may also manually call it from extractor's code if geo bypass
571 information is not available beforehand (e.g. obtained during
572 extraction) or due to some other reason. In this case you should pass
573 this information in geo bypass context passed as first argument. It may
574 contain following fields:
576 countries: List of geo unrestricted countries (similar
578 ip_blocks: List of geo unrestricted IP blocks in CIDR notation
579 (similar to _GEO_IP_BLOCKS)
582 if not self
._x
_forwarded
_for
_ip
:
584 # Geo bypass mechanism is explicitly disabled by user
585 if not self
.get_param('geo_bypass', True):
588 if not geo_bypass_context
:
589 geo_bypass_context
= {}
591 # Backward compatibility: previously _initialize_geo_bypass
592 # expected a list of countries, some 3rd party code may still use
594 if isinstance(geo_bypass_context
, (list, tuple)):
595 geo_bypass_context
= {
596 'countries': geo_bypass_context
,
599 # The whole point of geo bypass mechanism is to fake IP
600 # as X-Forwarded-For HTTP header based on some IP block or
603 # Path 1: bypassing based on IP block in CIDR notation
605 # Explicit IP block specified by user, use it right away
606 # regardless of whether extractor is geo bypassable or not
607 ip_block
= self
.get_param('geo_bypass_ip_block', None)
609 # Otherwise use random IP block from geo bypass context but only
610 # if extractor is known as geo bypassable
612 ip_blocks
= geo_bypass_context
.get('ip_blocks')
613 if self
._GEO
_BYPASS
and ip_blocks
:
614 ip_block
= random
.choice(ip_blocks
)
617 self
._x
_forwarded
_for
_ip
= GeoUtils
.random_ipv4(ip_block
)
618 self
.write_debug(f
'Using fake IP {self._x_forwarded_for_ip} as X-Forwarded-For')
621 # Path 2: bypassing based on country code
623 # Explicit country code specified by user, use it right away
624 # regardless of whether extractor is geo bypassable or not
625 country
= self
.get_param('geo_bypass_country', None)
627 # Otherwise use random country code from geo bypass context but
628 # only if extractor is known as geo bypassable
630 countries
= geo_bypass_context
.get('countries')
631 if self
._GEO
_BYPASS
and countries
:
632 country
= random
.choice(countries
)
635 self
._x
_forwarded
_for
_ip
= GeoUtils
.random_ipv4(country
)
636 self
._downloader
.write_debug(
637 f
'Using fake IP {self._x_forwarded_for_ip} ({country.upper()}) as X-Forwarded-For')
639 def extract(self
, url
):
640 """Extracts URL information and returns it in list of dicts."""
645 self
.write_debug('Extracting URL: %s' % url
)
646 ie_result
= self
._real
_extract
(url
)
647 if ie_result
is None:
649 if self
._x
_forwarded
_for
_ip
:
650 ie_result
['__x_forwarded_for_ip'] = self
._x
_forwarded
_for
_ip
651 subtitles
= ie_result
.get('subtitles') or {}
652 if 'no-live-chat' in self
.get_param('compat_opts'):
653 for lang
in ('live_chat', 'comments', 'danmaku'):
654 subtitles
.pop(lang
, None)
656 except GeoRestrictedError
as e
:
657 if self
.__maybe
_fake
_ip
_and
_retry
(e
.countries
):
660 except UnsupportedError
:
662 except ExtractorError
as e
:
664 'video_id': e
.video_id
or self
.get_temp_id(url
),
666 'tb': e
.traceback
or sys
.exc_info()[2],
667 'expected': e
.expected
,
670 if hasattr(e
, 'countries'):
671 kwargs
['countries'] = e
.countries
672 raise type(e
)(e
.orig_msg
, **kwargs
)
673 except http
.client
.IncompleteRead
as e
:
674 raise ExtractorError('A network error has occurred.', cause
=e
, expected
=True, video_id
=self
.get_temp_id(url
))
675 except (KeyError, StopIteration) as e
:
676 raise ExtractorError('An extractor error has occurred.', cause
=e
, video_id
=self
.get_temp_id(url
))
678 def __maybe_fake_ip_and_retry(self
, countries
):
679 if (not self
.get_param('geo_bypass_country', None)
681 and self
.get_param('geo_bypass', True)
682 and not self
._x
_forwarded
_for
_ip
684 country_code
= random
.choice(countries
)
685 self
._x
_forwarded
_for
_ip
= GeoUtils
.random_ipv4(country_code
)
686 if self
._x
_forwarded
_for
_ip
:
688 'Video is geo restricted. Retrying extraction with fake IP %s (%s) as X-Forwarded-For.'
689 % (self
._x
_forwarded
_for
_ip
, country_code
.upper()))
693 def set_downloader(self
, downloader
):
694 """Sets a YoutubeDL instance as the downloader for this IE."""
695 self
._downloader
= downloader
699 return self
._downloader
.cache
703 return self
._downloader
.cookiejar
705 def _initialize_pre_login(self
):
706 """ Initialization before login. Redefine in subclasses."""
709 def _perform_login(self
, username
, password
):
710 """ Login with username and password. Redefine in subclasses."""
713 def _real_initialize(self
):
714 """Real initialization process. Redefine in subclasses."""
717 def _real_extract(self
, url
):
718 """Real extraction process. Redefine in subclasses."""
719 raise NotImplementedError('This method must be implemented by subclasses')
723 """A string for getting the InfoExtractor with get_info_extractor"""
724 return cls
.__name
__[:-2]
728 return cls
.__name
__[:-2]
731 def __can_accept_status_code(err
, expected_status
):
732 assert isinstance(err
, urllib
.error
.HTTPError
)
733 if expected_status
is None:
735 elif callable(expected_status
):
736 return expected_status(err
.code
) is True
738 return err
.code
in variadic(expected_status
)
740 def _create_request(self
, url_or_request
, data
=None, headers
=None, query
=None):
741 if isinstance(url_or_request
, urllib
.request
.Request
):
742 return update_Request(url_or_request
, data
=data
, headers
=headers
, query
=query
)
744 url_or_request
= update_url_query(url_or_request
, query
)
745 return sanitized_Request(url_or_request
, data
, headers
or {})
747 def _request_webpage(self
, url_or_request
, video_id
, note
=None, errnote
=None, fatal
=True, data
=None, headers
=None, query
=None, expected_status
=None):
749 Return the response handle.
751 See _download_webpage docstring for arguments specification.
753 if not self
._downloader
._first
_webpage
_request
:
754 sleep_interval
= self
.get_param('sleep_interval_requests') or 0
755 if sleep_interval
> 0:
756 self
.to_screen('Sleeping %s seconds ...' % sleep_interval
)
757 time
.sleep(sleep_interval
)
759 self
._downloader
._first
_webpage
_request
= False
762 self
.report_download_webpage(video_id
)
763 elif note
is not False:
765 self
.to_screen(str(note
))
767 self
.to_screen(f
'{video_id}: {note}')
769 # Some sites check X-Forwarded-For HTTP header in order to figure out
770 # the origin of the client behind proxy. This allows bypassing geo
771 # restriction by faking this header's value to IP that belongs to some
772 # geo unrestricted country. We will do so once we encounter any
773 # geo restriction error.
774 if self
._x
_forwarded
_for
_ip
:
775 headers
= (headers
or {}).copy()
776 headers
.setdefault('X-Forwarded-For', self
._x
_forwarded
_for
_ip
)
779 return self
._downloader
.urlopen(self
._create
_request
(url_or_request
, data
, headers
, query
))
780 except network_exceptions
as err
:
781 if isinstance(err
, urllib
.error
.HTTPError
):
782 if self
.__can
_accept
_status
_code
(err
, expected_status
):
783 # Retain reference to error to prevent file object from
784 # being closed before it can be read. Works around the
785 # effects of <https://bugs.python.org/issue15002>
786 # introduced in Python 3.4.1.
793 errnote
= 'Unable to download webpage'
795 errmsg
= f
'{errnote}: {error_to_compat_str(err)}'
797 raise ExtractorError(errmsg
, cause
=err
)
799 self
.report_warning(errmsg
)
802 def _download_webpage_handle(self
, url_or_request
, video_id
, note
=None, errnote
=None, fatal
=True,
803 encoding
=None, data
=None, headers
={}, query={}
, expected_status
=None):
805 Return a tuple (page content as string, URL handle).
808 url_or_request -- plain text URL as a string or
809 a urllib.request.Request object
810 video_id -- Video/playlist/item identifier (string)
813 note -- note printed before downloading (string)
814 errnote -- note printed in case of an error (string)
815 fatal -- flag denoting whether error should be considered fatal,
816 i.e. whether it should cause ExtractionError to be raised,
817 otherwise a warning will be reported and extraction continued
818 encoding -- encoding for a page content decoding, guessed automatically
819 when not explicitly specified
820 data -- POST data (bytes)
821 headers -- HTTP headers (dict)
822 query -- URL query (dict)
823 expected_status -- allows to accept failed HTTP requests (non 2xx
824 status code) by explicitly specifying a set of accepted status
825 codes. Can be any of the following entities:
826 - an integer type specifying an exact failed status code to
828 - a list or a tuple of integer types specifying a list of
829 failed status codes to accept
830 - a callable accepting an actual failed status code and
831 returning True if it should be accepted
832 Note that this argument does not affect success status codes (2xx)
833 which are always accepted.
836 # Strip hashes from the URL (#1038)
837 if isinstance(url_or_request
, str):
838 url_or_request
= url_or_request
.partition('#')[0]
840 urlh
= self
._request
_webpage
(url_or_request
, video_id
, note
, errnote
, fatal
, data
=data
, headers
=headers
, query
=query
, expected_status
=expected_status
)
844 content
= self
._webpage
_read
_content
(urlh
, url_or_request
, video_id
, note
, errnote
, fatal
, encoding
=encoding
)
845 return (content
, urlh
)
848 def _guess_encoding_from_content(content_type
, webpage_bytes
):
849 m
= re
.match(r
'[a-zA-Z0-9_.-]+/[a-zA-Z0-9_.-]+\s*;\s*charset=(.+)', content_type
)
851 encoding
= m
.group(1)
853 m
= re
.search(br
'<meta[^>]+charset=[\'"]?([^\'")]+)[ /\'">]',
854 webpage_bytes[:1024])
856 encoding = m.group(1).decode('ascii')
857 elif webpage_bytes.startswith(b'\xff\xfe'):
864 def __check_blocked(self, content):
865 first_block = content[:512]
866 if ('<title>Access to this site is blocked</title>' in content
867 and 'Websense' in first_block):
868 msg = 'Access to this webpage has been blocked by Websense filtering software in your network.'
869 blocked_iframe = self._html_search_regex(
870 r'<iframe src="([^
"]+)"', content,
871 'Websense information URL
', default=None)
873 msg += ' Visit
%s for more details
' % blocked_iframe
874 raise ExtractorError(msg, expected=True)
875 if '<title
>The URL you requested has been blocked
</title
>' in first_block:
877 'Access to this webpage has been blocked by Indian censorship
. '
878 'Use a VPN
or proxy
server (with --proxy
) to route around it
.')
879 block_msg = self._html_search_regex(
880 r'</h1
><p
>(.*?
)</p
>',
881 content, 'block message
', default=None)
883 msg += ' (Message
: "%s")' % block_msg.replace('\n', ' ')
884 raise ExtractorError(msg, expected=True)
885 if ('<title
>TTK
:: Доступ к ресурсу ограничен
</title
>' in content
886 and 'blocklist
.rkn
.gov
.ru
' in content):
887 raise ExtractorError(
888 'Access to this webpage has been blocked by decision of the Russian government
. '
889 'Visit http
://blocklist
.rkn
.gov
.ru
/ for a block reason
.',
892 def _request_dump_filename(self, url, video_id):
893 basen = f'{video_id}_{url}
'
894 trim_length = self.get_param('trim_file_name
') or 240
895 if len(basen) > trim_length:
896 h = '___
' + hashlib.md5(basen.encode('utf
-8')).hexdigest()
897 basen = basen[:trim_length - len(h)] + h
898 filename = sanitize_filename(f'{basen}
.dump
', restricted=True)
899 # Working around MAX_PATH limitation on Windows (see
900 # http://msdn.microsoft.com/en-us/library/windows/desktop/aa365247(v=vs.85).aspx)
901 if compat_os_name == 'nt
':
902 absfilepath = os.path.abspath(filename)
903 if len(absfilepath) > 259:
904 filename = fR'\\?\{absfilepath}
'
907 def __decode_webpage(self, webpage_bytes, encoding, headers):
909 encoding = self._guess_encoding_from_content(headers.get('Content
-Type
', ''), webpage_bytes)
911 return webpage_bytes.decode(encoding, 'replace
')
913 return webpage_bytes.decode('utf
-8', 'replace
')
915 def _webpage_read_content(self, urlh, url_or_request, video_id, note=None, errnote=None, fatal=True, prefix=None, encoding=None):
916 webpage_bytes = urlh.read()
917 if prefix is not None:
918 webpage_bytes = prefix + webpage_bytes
919 if self.get_param('dump_intermediate_pages
', False):
920 self.to_screen('Dumping request to
' + urlh.geturl())
921 dump = base64.b64encode(webpage_bytes).decode('ascii
')
922 self._downloader.to_screen(dump)
923 if self.get_param('write_pages
'):
924 filename = self._request_dump_filename(urlh.geturl(), video_id)
925 self.to_screen(f'Saving request to {filename}
')
926 with open(filename, 'wb
') as outf:
927 outf.write(webpage_bytes)
929 content = self.__decode_webpage(webpage_bytes, encoding, urlh.headers)
930 self.__check_blocked(content)
934 def __print_error(self, errnote, fatal, video_id, err):
936 raise ExtractorError(f'{video_id}
: {errnote}
', cause=err)
938 self.report_warning(f'{video_id}
: {errnote}
: {err}
')
940 def _parse_xml(self, xml_string, video_id, transform_source=None, fatal=True, errnote=None):
942 xml_string = transform_source(xml_string)
944 return compat_etree_fromstring(xml_string.encode('utf
-8'))
945 except xml.etree.ElementTree.ParseError as ve:
946 self.__print_error('Failed to parse XML
' if errnote is None else errnote, fatal, video_id, ve)
948 def _parse_json(self, json_string, video_id, transform_source=None, fatal=True, errnote=None, **parser_kwargs):
951 json_string, cls=LenientJSONDecoder, strict=False, transform_source=transform_source, **parser_kwargs)
952 except ValueError as ve:
953 self.__print_error('Failed to parse JSON
' if errnote is None else errnote, fatal, video_id, ve)
955 def _parse_socket_response_as_json(self, data, *args, **kwargs):
956 return self._parse_json(data[data.find('{'):data.rfind('}
') + 1], *args, **kwargs)
958 def __create_download_methods(name, parser, note, errnote, return_value):
960 def parse(ie, content, *args, errnote=errnote, **kwargs):
964 kwargs['errnote
'] = errnote
965 # parser is fetched by name so subclasses can override it
966 return getattr(ie, parser)(content, *args, **kwargs)
968 def download_handle(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 res = self._download_webpage_handle(
971 url_or_request, video_id, note=note, errnote=errnote, fatal=fatal, encoding=encoding,
972 data=data, headers=headers, query=query, expected_status=expected_status)
976 return parse(self, content, video_id, transform_source=transform_source, fatal=fatal, errnote=errnote), urlh
978 def download_content(self, url_or_request, video_id, note=note, errnote=errnote, transform_source=None,
979 fatal=True, encoding=None, data=None, headers={}, query={}, expected_status=None):
980 if self.get_param('load_pages
'):
981 url_or_request = self._create_request(url_or_request, data, headers, query)
982 filename = self._request_dump_filename(url_or_request.full_url, video_id)
983 self.to_screen(f'Loading request
from {filename}
')
985 with open(filename, 'rb
') as dumpf:
986 webpage_bytes = dumpf.read()
988 self.report_warning(f'Unable to load request
from disk
: {e}
')
990 content = self.__decode_webpage(webpage_bytes, encoding, url_or_request.headers)
991 return parse(self, content, video_id, transform_source=transform_source, fatal=fatal, errnote=errnote)
995 'transform_source
': transform_source,
997 'encoding
': encoding,
1001 'expected_status
': expected_status,
1004 kwargs.pop('transform_source
')
1005 # The method is fetched by name so subclasses can override _download_..._handle
1006 res = getattr(self, download_handle.__name__)(url_or_request, video_id, **kwargs)
1007 return res if res is False else res[0]
1009 def impersonate(func, name, return_value):
1010 func.__name__, func.__qualname__ = name, f'InfoExtractor
.{name}
'
1012 @param transform_source Apply this transformation before parsing
1013 @returns {return_value}
1015 See _download_webpage_handle docstring for other arguments specification
1018 impersonate(download_handle, f'_download_{name}_handle
', f'({return_value}
, URL handle
)')
1019 impersonate(download_content, f'_download_{name}
', f'{return_value}
')
1020 return download_handle, download_content
1022 _download_xml_handle, _download_xml = __create_download_methods(
1023 'xml
', '_parse_xml
', 'Downloading XML
', 'Unable to download XML
', 'xml
as an xml
.etree
.ElementTree
.Element
')
1024 _download_json_handle, _download_json = __create_download_methods(
1025 'json
', '_parse_json
', 'Downloading JSON metadata
', 'Unable to download JSON metadata
', 'JSON
object as a
dict')
1026 _download_socket_json_handle, _download_socket_json = __create_download_methods(
1027 'socket_json
', '_parse_socket_response_as_json
', 'Polling socket
', 'Unable to poll socket
', 'JSON
object as a
dict')
1028 __download_webpage = __create_download_methods('webpage
', None, None, None, 'data of the page
as a string
')[1]
1030 def _download_webpage(
1031 self, url_or_request, video_id, note=None, errnote=None,
1032 fatal=True, tries=1, timeout=NO_DEFAULT, *args, **kwargs):
1034 Return the data of the page as a string.
1037 tries -- number of tries
1038 timeout -- sleep interval between tries
1040 See _download_webpage_handle docstring for other arguments specification.
1043 R''' # NB: These are unused; should they be deprecated?
1045 self._downloader.deprecation_warning('tries argument
is deprecated
in InfoExtractor
._download
_webpage
')
1046 if timeout is NO_DEFAULT:
1049 self._downloader.deprecation_warning('timeout argument
is deprecated
in InfoExtractor
._download
_webpage
')
1055 return self.__download_webpage(url_or_request, video_id, note, errnote, None, fatal, *args, **kwargs)
1056 except http.client.IncompleteRead as e:
1058 if try_count >= tries:
1060 self._sleep(timeout, video_id)
1062 def report_warning(self, msg, video_id=None, *args, only_once=False, **kwargs):
1063 idstr = format_field(video_id, None, '%s: ')
1064 msg = f'[{self.IE_NAME}
] {idstr}{msg}
'
1066 if f'WARNING
: {msg}
' in self._printed_messages:
1068 self._printed_messages.add(f'WARNING
: {msg}
')
1069 self._downloader.report_warning(msg, *args, **kwargs)
1071 def to_screen(self, msg, *args, **kwargs):
1072 """Print msg to screen, prefixing it with '[ie_name
]'"""
1073 self._downloader.to_screen(f'[{self.IE_NAME}
] {msg}
', *args, **kwargs)
1075 def write_debug(self, msg, *args, **kwargs):
1076 self._downloader.write_debug(f'[{self.IE_NAME}
] {msg}
', *args, **kwargs)
1078 def get_param(self, name, default=None, *args, **kwargs):
1079 if self._downloader:
1080 return self._downloader.params.get(name, default, *args, **kwargs)
1083 def report_drm(self, video_id, partial=False):
1084 self.raise_no_formats('This video
is DRM protected
', expected=True, video_id=video_id)
1086 def report_extraction(self, id_or_name):
1087 """Report information extraction."""
1088 self.to_screen('%s: Extracting information
' % id_or_name)
1090 def report_download_webpage(self, video_id):
1091 """Report webpage download."""
1092 self.to_screen('%s: Downloading webpage
' % video_id)
1094 def report_age_confirmation(self):
1095 """Report attempt to confirm age."""
1096 self.to_screen('Confirming age
')
1098 def report_login(self):
1099 """Report attempt to log in."""
1100 self.to_screen('Logging
in')
1102 def raise_login_required(
1103 self, msg='This video
is only available
for registered users
',
1104 metadata_available=False, method=NO_DEFAULT):
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 msg += format_field(self._login_hint(method), None, '. %s')
1110 raise ExtractorError(msg, expected=True)
1112 def raise_geo_restricted(
1113 self, msg='This video
is not available
from your location due to geo restriction
',
1114 countries=None, metadata_available=False):
1115 if metadata_available and (
1116 self.get_param('ignore_no_formats_error
') or self.get_param('wait_for_video
')):
1117 self.report_warning(msg)
1119 raise GeoRestrictedError(msg, countries=countries)
1121 def raise_no_formats(self, msg, expected=False, video_id=None):
1123 self.get_param('ignore_no_formats_error
') or self.get_param('wait_for_video
')):
1124 self.report_warning(msg, video_id)
1125 elif isinstance(msg, ExtractorError):
1128 raise ExtractorError(msg, expected=expected, video_id=video_id)
1130 # Methods for following #608
1132 def url_result(url, ie=None, video_id=None, video_title=None, *, url_transparent=False, **kwargs):
1133 """Returns a URL that points to a page that should be processed"""
1135 kwargs['ie_key
'] = ie if isinstance(ie, str) else ie.ie_key()
1136 if video_id is not None:
1137 kwargs['id'] = video_id
1138 if video_title is not None:
1139 kwargs['title
'] = video_title
1142 '_type
': 'url_transparent
' if url_transparent else 'url
',
1146 def playlist_from_matches(self, matches, playlist_id=None, playlist_title=None, getter=None, ie=None, video_kwargs=None, **kwargs):
1147 urls = (self.url_result(self._proto_relative_url(m), ie, **(video_kwargs or {}))
1148 for m in orderedSet(map(getter, matches) if getter else matches))
1149 return self.playlist_result(urls, playlist_id, playlist_title, **kwargs)
1152 def playlist_result(entries, playlist_id=None, playlist_title=None, playlist_description=None, *, multi_video=False, **kwargs):
1153 """Returns a playlist"""
1155 kwargs['id'] = playlist_id
1157 kwargs['title
'] = playlist_title
1158 if playlist_description is not None:
1159 kwargs['description
'] = playlist_description
1162 '_type
': 'multi_video
' if multi_video else 'playlist
',
1166 def _search_regex(self, pattern, string, name, default=NO_DEFAULT, fatal=True, flags=0, group=None):
1168 Perform a regex search on the given string, using a single or a list of
1169 patterns returning the first matching group.
1170 In case of failure return a default value or raise a WARNING or a
1171 RegexNotFoundError, depending on fatal, specifying the field name.
1175 elif isinstance(pattern, (str, re.Pattern)):
1176 mobj = re.search(pattern, string, flags)
1179 mobj = re.search(p, string, flags)
1183 _name = self._downloader._format_err(name, self._downloader.Styles.EMPHASIS)
1187 # return the first matching group
1188 return next(g for g in mobj.groups() if g is not None)
1189 elif isinstance(group, (list, tuple)):
1190 return tuple(mobj.group(g) for g in group)
1192 return mobj.group(group)
1193 elif default is not NO_DEFAULT:
1196 raise RegexNotFoundError('Unable to extract
%s' % _name)
1198 self.report_warning('unable to extract
%s' % _name + bug_reports_message())
1201 def _search_json(self, start_pattern, string, name, video_id, *, end_pattern='',
1202 contains_pattern='(?s
:.+)', fatal=True, default=NO_DEFAULT, **kwargs):
1203 """Searches string for the JSON object specified by start_pattern"""
1204 # NB: end_pattern is only used to reduce the size of the initial match
1205 if default is NO_DEFAULT:
1206 default, has_default = {}, False
1208 fatal, has_default = False, True
1210 json_string = self._search_regex(
1211 rf'{start_pattern}\s
*(?P
<json
>{{\s*{contains_pattern}\s
*}})\s
*{end_pattern}
',
1212 string, name, group='json
', fatal=fatal, default=None if has_default else NO_DEFAULT)
1216 _name = self._downloader._format_err(name, self._downloader.Styles.EMPHASIS)
1218 return self._parse_json(json_string, video_id, ignore_extra=True, **kwargs)
1219 except ExtractorError as e:
1221 raise ExtractorError(
1222 f'Unable to extract {_name}
- Failed to parse JSON
', cause=e.cause, video_id=video_id)
1223 elif not has_default:
1224 self.report_warning(
1225 f'Unable to extract {_name}
- Failed to parse JSON
: {e}
', video_id=video_id)
1228 def _html_search_regex(self, pattern, string, name, default=NO_DEFAULT, fatal=True, flags=0, group=None):
1230 Like _search_regex, but strips HTML tags and unescapes entities.
1232 res = self._search_regex(pattern, string, name, default, fatal, flags, group)
1234 return clean_html(res).strip()
1238 def _get_netrc_login_info(self, netrc_machine=None):
1241 netrc_machine = netrc_machine or self._NETRC_MACHINE
1243 if self.get_param('usenetrc
', False):
1245 netrc_file = compat_expanduser(self.get_param('netrc_location
') or '~
')
1246 if os.path.isdir(netrc_file):
1247 netrc_file = os.path.join(netrc_file, '.netrc
')
1248 info = netrc.netrc(file=netrc_file).authenticators(netrc_machine)
1249 if info is not None:
1253 raise netrc.NetrcParseError(
1254 'No authenticators
for %s' % netrc_machine)
1255 except (OSError, netrc.NetrcParseError) as err:
1256 self.report_warning(
1257 'parsing
.netrc
: %s' % error_to_compat_str(err))
1259 return username, password
1261 def _get_login_info(self, username_option='username
', password_option='password
', netrc_machine=None):
1263 Get the login info as (username, password)
1264 First look for the manually specified credentials using username_option
1265 and password_option as keys in params dictionary. If no such credentials
1266 available look in the netrc file using the netrc_machine or _NETRC_MACHINE
1268 If there's no info available
, return (None, None)
1271 # Attempt to use provided username and password or .netrc data
1272 username = self.get_param(username_option)
1273 if username is not None:
1274 password = self.get_param(password_option)
1276 username, password = self._get_netrc_login_info(netrc_machine)
1278 return username, password
1280 def _get_tfa_info(self, note='two-factor verification code'):
1282 Get the two
-factor authentication info
1283 TODO
- asking the user will be required
for sms
/phone verify
1284 currently just uses the command line option
1285 If there
's no info available, return None
1288 tfa = self.get_param('twofactor
')
1292 return getpass.getpass('Type
%s and press
[Return
]: ' % note)
1294 # Helper functions for extracting OpenGraph info
1296 def _og_regexes(prop):
1297 content_re = r'content
=(?
:"([^"]+?
)"|\'([^\']+?)\'|\s*([^\s"\'=<>`
]+?
))'
1298 property_re = (r'(?
:name|
property)=(?
:\'og
%(sep)s%(prop)s\'|
"og%(sep)s%(prop)s"|\s
*og
%(sep)s%(prop)s\b)'
1299 % {'prop': re.escape(prop), 'sep': '(?::|[:-])'})
1300 template = r'<meta
[^
>]+?
%s[^
>]+?
%s'
1302 template % (property_re, content_re),
1303 template % (content_re, property_re),
1307 def _meta_regex(prop):
1308 return r'''(?isx)<meta
1309 (?=[^>]+(?:itemprop|name|property|id|http-equiv)=(["\']?)%s\1)
1310 [^>]+?content=(["\'])(?P<content>.*?)\2''' % re.escape(prop)
1312 def _og_search_property(self, prop, html, name=None, **kargs):
1313 prop = variadic(prop)
1315 name = 'OpenGraph
%s' % prop[0]
1318 og_regexes.extend(self._og_regexes(p))
1319 escaped = self._search_regex(og_regexes, html, name, flags=re.DOTALL, **kargs)
1322 return unescapeHTML(escaped)
1324 def _og_search_thumbnail(self, html, **kargs):
1325 return self._og_search_property('image
', html, 'thumbnail URL
', fatal=False, **kargs)
1327 def _og_search_description(self, html, **kargs):
1328 return self._og_search_property('description
', html, fatal=False, **kargs)
1330 def _og_search_title(self, html, *, fatal=False, **kargs):
1331 return self._og_search_property('title
', html, fatal=fatal, **kargs)
1333 def _og_search_video_url(self, html, name='video url
', secure=True, **kargs):
1334 regexes = self._og_regexes('video
') + self._og_regexes('video
:url
')
1336 regexes = self._og_regexes('video
:secure_url
') + regexes
1337 return self._html_search_regex(regexes, html, name, **kargs)
1339 def _og_search_url(self, html, **kargs):
1340 return self._og_search_property('url
', html, **kargs)
1342 def _html_extract_title(self, html, name='title
', *, fatal=False, **kwargs):
1343 return self._html_search_regex(r'(?s
)<title
\b[^
>]*>([^
<]+)</title
>', html, name, fatal=fatal, **kwargs)
1345 def _html_search_meta(self, name, html, display_name=None, fatal=False, **kwargs):
1346 name = variadic(name)
1347 if display_name is None:
1348 display_name = name[0]
1349 return self._html_search_regex(
1350 [self._meta_regex(n) for n in name],
1351 html, display_name, fatal=fatal, group='content
', **kwargs)
1353 def _dc_search_uploader(self, html):
1354 return self._html_search_meta('dc
.creator
', html, 'uploader
')
1356 def _rta_search(self, html):
1357 # See http://www.rtalabel.org/index.php?content=howtofaq#single
1358 if re.search(r'(?ix
)<meta\s
+name
="rating"\s
+'
1359 r' content
="RTA-5042-1996-1400-1577-RTA"',
1364 def _media_rating_search(self, html):
1365 # See http://www.tjg-designs.com/WP/metadata-code-examples-adding-metadata-to-your-web-pages/
1366 rating = self._html_search_meta('rating
', html)
1378 return RATING_TABLE.get(rating.lower())
1380 def _family_friendly_search(self, html):
1381 # See http://schema.org/VideoObject
1382 family_friendly = self._html_search_meta(
1383 'isFamilyFriendly
', html, default=None)
1385 if not family_friendly:
1394 return RATING_TABLE.get(family_friendly.lower())
1396 def _twitter_search_player(self, html):
1397 return self._html_search_meta('twitter
:player
', html,
1398 'twitter card player
')
1400 def _yield_json_ld(self, html, video_id, *, fatal=True, default=NO_DEFAULT):
1401 """Yield all json ld objects in the html"""
1402 if default is not NO_DEFAULT:
1404 for mobj in re.finditer(JSON_LD_RE, html):
1405 json_ld_item = self._parse_json(mobj.group('json_ld
'), video_id, fatal=fatal)
1406 for json_ld in variadic(json_ld_item):
1407 if isinstance(json_ld, dict):
1410 def _search_json_ld(self, html, video_id, expected_type=None, *, fatal=True, default=NO_DEFAULT):
1411 """Search for a video in any json ld in the html"""
1412 if default is not NO_DEFAULT:
1414 info = self._json_ld(
1415 list(self._yield_json_ld(html, video_id, fatal=fatal, default=default)),
1416 video_id, fatal=fatal, expected_type=expected_type)
1419 if default is not NO_DEFAULT:
1422 raise RegexNotFoundError('Unable to extract JSON
-LD
')
1424 self.report_warning('unable to extract JSON
-LD
%s' % bug_reports_message())
1427 def _json_ld(self, json_ld, video_id, fatal=True, expected_type=None):
1428 if isinstance(json_ld, str):
1429 json_ld = self._parse_json(json_ld, video_id, fatal=fatal)
1433 if not isinstance(json_ld, (list, tuple, dict)):
1435 if isinstance(json_ld, dict):
1438 INTERACTION_TYPE_MAP = {
1439 'CommentAction
': 'comment
',
1440 'AgreeAction
': 'like
',
1441 'DisagreeAction
': 'dislike
',
1442 'LikeAction
': 'like
',
1443 'DislikeAction
': 'dislike
',
1444 'ListenAction
': 'view
',
1445 'WatchAction
': 'view
',
1446 'ViewAction
': 'view
',
1449 def is_type(e, *expected_types):
1450 type = variadic(traverse_obj(e, '@type'))
1451 return any(x in type for x in expected_types)
1453 def extract_interaction_type(e):
1454 interaction_type = e.get('interactionType
')
1455 if isinstance(interaction_type, dict):
1456 interaction_type = interaction_type.get('@type')
1457 return str_or_none(interaction_type)
1459 def extract_interaction_statistic(e):
1460 interaction_statistic = e.get('interactionStatistic
')
1461 if isinstance(interaction_statistic, dict):
1462 interaction_statistic = [interaction_statistic]
1463 if not isinstance(interaction_statistic, list):
1465 for is_e in interaction_statistic:
1466 if not is_type(is_e, 'InteractionCounter
'):
1468 interaction_type = extract_interaction_type(is_e)
1469 if not interaction_type:
1471 # For interaction count some sites provide string instead of
1472 # an integer (as per spec) with non digit characters (e.g. ",")
1473 # so extracting count with more relaxed str_to_int
1474 interaction_count = str_to_int(is_e.get('userInteractionCount
'))
1475 if interaction_count is None:
1477 count_kind = INTERACTION_TYPE_MAP.get(interaction_type.split('/')[-1])
1480 count_key = '%s_count
' % count_kind
1481 if info.get(count_key) is not None:
1483 info[count_key] = interaction_count
1485 def extract_chapter_information(e):
1487 'title
': part.get('name
'),
1488 'start_time
': part.get('startOffset
'),
1489 'end_time
': part.get('endOffset
'),
1490 } for part in variadic(e.get('hasPart
') or []) if part.get('@type') == 'Clip
']
1491 for idx, (last_c, current_c, next_c) in enumerate(zip(
1492 [{'end_time': 0}] + chapters, chapters, chapters[1:])):
1493 current_c['end_time
'] = current_c['end_time
'] or next_c['start_time
']
1494 current_c['start_time
'] = current_c['start_time
'] or last_c['end_time
']
1495 if None in current_c.values():
1496 self.report_warning(f'Chapter {idx} contains broken data
. Not extracting chapters
')
1499 chapters[-1]['end_time
'] = chapters[-1]['end_time
'] or info['duration
']
1500 info['chapters
'] = chapters
1502 def extract_video_object(e):
1503 assert is_type(e, 'VideoObject
')
1504 author = e.get('author
')
1506 'url
': url_or_none(e.get('contentUrl
')),
1507 'title
': unescapeHTML(e.get('name
')),
1508 'description
': unescapeHTML(e.get('description
')),
1509 'thumbnails
': [{'url': unescapeHTML(url)}
1510 for url in variadic(traverse_obj(e, 'thumbnailUrl
', 'thumbnailURL
'))
1511 if url_or_none(url)],
1512 'duration
': parse_duration(e.get('duration
')),
1513 'timestamp
': unified_timestamp(e.get('uploadDate
')),
1514 # author can be an instance of 'Organization
' or 'Person
' types.
1515 # both types can have 'name
' property(inherited from 'Thing
' type). [1]
1516 # however some websites are using 'Text
' type instead.
1517 # 1. https://schema.org/VideoObject
1518 'uploader
': author.get('name
') if isinstance(author, dict) else author if isinstance(author, str) else None,
1519 'filesize
': int_or_none(float_or_none(e.get('contentSize
'))),
1520 'tbr
': int_or_none(e.get('bitrate
')),
1521 'width
': int_or_none(e.get('width
')),
1522 'height
': int_or_none(e.get('height
')),
1523 'view_count
': int_or_none(e.get('interactionCount
')),
1525 extract_interaction_statistic(e)
1526 extract_chapter_information(e)
1528 def traverse_json_ld(json_ld, at_top_level=True):
1530 if at_top_level and '@context' not in e:
1532 if at_top_level and set(e.keys()) == {'@context', '@graph'}:
1533 traverse_json_ld(variadic(e['@graph'], allowed_types=(dict,)), at_top_level=False)
1535 if expected_type is not None and not is_type(e, expected_type):
1537 rating = traverse_obj(e, ('aggregateRating
', 'ratingValue
'), expected_type=float_or_none)
1538 if rating is not None:
1539 info['average_rating
'] = rating
1540 if is_type(e, 'TVEpisode
', 'Episode
'):
1541 episode_name = unescapeHTML(e.get('name
'))
1543 'episode
': episode_name,
1544 'episode_number
': int_or_none(e.get('episodeNumber
')),
1545 'description
': unescapeHTML(e.get('description
')),
1547 if not info.get('title
') and episode_name:
1548 info['title
'] = episode_name
1549 part_of_season = e.get('partOfSeason
')
1550 if is_type(part_of_season, 'TVSeason
', 'Season
', 'CreativeWorkSeason
'):
1552 'season
': unescapeHTML(part_of_season.get('name
')),
1553 'season_number
': int_or_none(part_of_season.get('seasonNumber
')),
1555 part_of_series = e.get('partOfSeries
') or e.get('partOfTVSeries
')
1556 if is_type(part_of_series, 'TVSeries
', 'Series
', 'CreativeWorkSeries
'):
1557 info['series
'] = unescapeHTML(part_of_series.get('name
'))
1558 elif is_type(e, 'Movie
'):
1560 'title
': unescapeHTML(e.get('name
')),
1561 'description
': unescapeHTML(e.get('description
')),
1562 'duration
': parse_duration(e.get('duration
')),
1563 'timestamp
': unified_timestamp(e.get('dateCreated
')),
1565 elif is_type(e, 'Article
', 'NewsArticle
'):
1567 'timestamp
': parse_iso8601(e.get('datePublished
')),
1568 'title
': unescapeHTML(e.get('headline
')),
1569 'description
': unescapeHTML(e.get('articleBody
') or e.get('description
')),
1571 if is_type(traverse_obj(e, ('video
', 0)), 'VideoObject
'):
1572 extract_video_object(e['video
'][0])
1573 elif is_type(traverse_obj(e, ('subjectOf
', 0)), 'VideoObject
'):
1574 extract_video_object(e['subjectOf
'][0])
1575 elif is_type(e, 'VideoObject
'):
1576 extract_video_object(e)
1577 if expected_type is None:
1581 video = e.get('video
')
1582 if is_type(video, 'VideoObject
'):
1583 extract_video_object(video)
1584 if expected_type is None:
1588 traverse_json_ld(json_ld)
1590 return filter_dict(info)
1592 def _search_nextjs_data(self, webpage, video_id, *, transform_source=None, fatal=True, **kw):
1593 return self._parse_json(
1595 r'(?s
)<script
[^
>]+id=[\'"]__NEXT_DATA__[\'"][^
>]*>([^
<]+)</script
>',
1596 webpage, 'next
.js data
', fatal=fatal, **kw),
1597 video_id, transform_source=transform_source, fatal=fatal)
1599 def _search_nuxt_data(self, webpage, video_id, context_name='__NUXT__
', *, fatal=True, traverse=('data
', 0)):
1600 """Parses Nuxt.js metadata. This works as long as the function __NUXT__ invokes is a pure function"""
1601 rectx = re.escape(context_name)
1602 FUNCTION_RE = r'\
(function\
((?P
<arg_keys
>.*?
)\
){return\s+(?P<js>{.*?}
)\s
*;?\s
*}\
((?P
<arg_vals
>.*?
)\
)'
1603 js, arg_keys, arg_vals = self._search_regex(
1604 (rf'<script
>\s
*window\
.{rectx}
={FUNCTION_RE}\s
*\
)\s
*;?\s
*</script
>', rf'{rectx}\
(.*?{FUNCTION_RE}
'),
1605 webpage, context_name, group=('js
', 'arg_keys
', 'arg_vals
'), fatal=fatal)
1607 args = dict(zip(arg_keys.split(','), arg_vals.split(',')))
1609 for key, val in args.items():
1610 if val in ('undefined
', 'void
0'):
1613 ret = self._parse_json(js, video_id, transform_source=functools.partial(js_to_json, vars=args), fatal=fatal)
1614 return traverse_obj(ret, traverse) or {}
1617 def _hidden_inputs(html):
1618 html = re.sub(r'<!--(?
:(?
!<!--).)*-->', '', html)
1620 for input in re.findall(r'(?i
)(<input[^
>]+>)', html):
1621 attrs = extract_attributes(input)
1624 if attrs.get('type') not in ('hidden
', 'submit
'):
1626 name = attrs.get('name
') or attrs.get('id')
1627 value = attrs.get('value
')
1628 if name and value is not None:
1629 hidden_inputs[name] = value
1630 return hidden_inputs
1632 def _form_hidden_inputs(self, form_id, html):
1633 form = self._search_regex(
1634 r'(?
is)<form
[^
>]+?
id=(["\'])%s\1[^>]*>(?P<form>.+?)</form>' % form_id,
1635 html, '%s form' % form_id, group='form')
1636 return self._hidden_inputs(form)
1639 regex = r' *((?P<reverse>\+)?(?P<field>[a-zA-Z0-9_]+)((?P<separator>[~:])(?P<limit>.*?))?)? *$'
1641 default = ('hidden', 'aud_or_vid', 'hasvid', 'ie_pref', 'lang', 'quality',
1642 'res', 'fps', 'hdr:12', 'codec:vp9.2', 'size', 'br', 'asr',
1643 'proto', 'ext', 'hasaud', 'source', 'id') # These must not be aliases
1644 ytdl_default = ('hasaud', 'lang', 'quality', 'tbr', 'filesize', 'vbr',
1645 'height', 'width', 'proto', 'vext', 'abr', 'aext',
1646 'fps', 'fs_approx', 'source', 'id')
1649 'vcodec': {'type': 'ordered', 'regex': True,
1650 'order': ['av0?1', 'vp0?9.2', 'vp0?9', '[hx]265|he?vc?', '[hx]264|avc', 'vp0?8', 'mp4v|h263', 'theora', '', None, 'none']},
1651 'acodec': {'type': 'ordered', 'regex': True,
1652 'order': ['[af]lac', 'wav|aiff', 'opus', 'vorbis|ogg', 'aac', 'mp?4a?', 'mp3', 'e-?a?c-?3', 'ac-?3', 'dts', '', None, 'none']},
1653 'hdr': {'type': 'ordered', 'regex': True, 'field': 'dynamic_range',
1654 'order': ['dv', '(hdr)?12', r'(hdr)?10\+', '(hdr)?10', 'hlg', '', 'sdr', None]},
1655 'proto': {'type': 'ordered', 'regex': True, 'field': 'protocol',
1656 'order': ['(ht|f)tps', '(ht|f)tp$', 'm3u8.*', '.*dash', 'websocket_frag', 'rtmpe?', '', 'mms|rtsp', 'ws|websocket', 'f4']},
1657 'vext': {'type': 'ordered', 'field': 'video_ext',
1658 'order': ('mp4', 'webm', 'flv', '', 'none'),
1659 'order_free': ('webm', 'mp4', 'flv', '', 'none')},
1660 'aext': {'type': 'ordered', 'field': 'audio_ext',
1661 'order': ('m4a', 'aac', 'mp3', 'ogg', 'opus', 'webm', '', 'none'),
1662 'order_free': ('opus', 'ogg', 'webm', 'm4a', 'mp3', 'aac', '', 'none')},
1663 'hidden': {'visible': False, 'forced': True, 'type': 'extractor', 'max': -1000},
1664 'aud_or_vid': {'visible': False, 'forced': True, 'type': 'multiple',
1665 'field': ('vcodec', 'acodec'),
1666 'function': lambda it: int(any(v != 'none' for v in it))},
1667 'ie_pref': {'priority': True, 'type': 'extractor'},
1668 'hasvid': {'priority': True, 'field': 'vcodec', 'type': 'boolean', 'not_in_list': ('none',)},
1669 'hasaud': {'field': 'acodec', 'type': 'boolean', 'not_in_list': ('none',)},
1670 'lang': {'convert': 'float', 'field': 'language_preference', 'default': -1},
1671 'quality': {'convert': 'float', 'default': -1},
1672 'filesize': {'convert': 'bytes'},
1673 'fs_approx': {'convert': 'bytes', 'field': 'filesize_approx'},
1674 'id': {'convert': 'string', 'field': 'format_id'},
1675 'height': {'convert': 'float_none'},
1676 'width': {'convert': 'float_none'},
1677 'fps': {'convert': 'float_none'},
1678 'tbr': {'convert': 'float_none'},
1679 'vbr': {'convert': 'float_none'},
1680 'abr': {'convert': 'float_none'},
1681 'asr': {'convert': 'float_none'},
1682 'source': {'convert': 'float', 'field': 'source_preference', 'default': -1},
1684 'codec': {'type': 'combined', 'field': ('vcodec', 'acodec')},
1685 'br': {'type': 'combined', 'field': ('tbr', 'vbr', 'abr'), 'same_limit': True},
1686 'size': {'type': 'combined', 'same_limit': True, 'field': ('filesize', 'fs_approx')},
1687 'ext': {'type': 'combined', 'field': ('vext', 'aext')},
1688 'res': {'type': 'multiple', 'field': ('height', 'width'),
1689 'function': lambda it: (lambda l: min(l) if l else 0)(tuple(filter(None, it)))},
1691 # For compatibility with youtube-dl
1692 'format_id': {'type': 'alias', 'field': 'id'},
1693 'preference': {'type': 'alias', 'field': 'ie_pref'},
1694 'language_preference': {'type': 'alias', 'field': 'lang'},
1695 'source_preference': {'type': 'alias', 'field': 'source'},
1696 'protocol': {'type': 'alias', 'field': 'proto'},
1697 'filesize_approx': {'type': 'alias', 'field': 'fs_approx'},
1700 'dimension': {'type': 'alias', 'field': 'res', 'deprecated': True},
1701 'resolution': {'type': 'alias', 'field': 'res', 'deprecated': True},
1702 'extension': {'type': 'alias', 'field': 'ext', 'deprecated': True},
1703 'bitrate': {'type': 'alias', 'field': 'br', 'deprecated': True},
1704 'total_bitrate': {'type': 'alias', 'field': 'tbr', 'deprecated': True},
1705 'video_bitrate': {'type': 'alias', 'field': 'vbr', 'deprecated': True},
1706 'audio_bitrate': {'type': 'alias', 'field': 'abr', 'deprecated': True},
1707 'framerate': {'type': 'alias', 'field': 'fps', 'deprecated': True},
1708 'filesize_estimate': {'type': 'alias', 'field': 'size', 'deprecated': True},
1709 'samplerate': {'type': 'alias', 'field': 'asr', 'deprecated': True},
1710 'video_ext': {'type': 'alias', 'field': 'vext', 'deprecated': True},
1711 'audio_ext': {'type': 'alias', 'field': 'aext', 'deprecated': True},
1712 'video_codec': {'type': 'alias', 'field': 'vcodec', 'deprecated': True},
1713 'audio_codec': {'type': 'alias', 'field': 'acodec', 'deprecated': True},
1714 'video': {'type': 'alias', 'field': 'hasvid', 'deprecated': True},
1715 'has_video': {'type': 'alias', 'field': 'hasvid', 'deprecated': True},
1716 'audio': {'type': 'alias', 'field': 'hasaud', 'deprecated': True},
1717 'has_audio': {'type': 'alias', 'field': 'hasaud', 'deprecated': True},
1718 'extractor': {'type': 'alias', 'field': 'ie_pref', 'deprecated': True},
1719 'extractor_preference': {'type': 'alias', 'field': 'ie_pref', 'deprecated': True},
1722 def __init__(self, ie, field_preference):
1724 self.ydl = ie._downloader
1725 self.evaluate_params(self.ydl.params, field_preference)
1726 if ie.get_param('verbose'):
1727 self.print_verbose_info(self.ydl.write_debug)
1729 def _get_field_setting(self, field, key):
1730 if field not in self.settings:
1731 if key in ('forced', 'priority'):
1733 self.ydl.deprecation_warning(
1734 f'Using arbitrary fields ({field}) for format sorting is deprecated '
1735 'and may be removed in a future version')
1736 self.settings[field] = {}
1737 propObj = self.settings[field]
1738 if key not in propObj:
1739 type = propObj.get('type')
1741 default = 'preference' if type == 'extractor' else (field,) if type in ('combined', 'multiple') else field
1742 elif key == 'convert':
1743 default = 'order' if type == 'ordered' else 'float_string' if field else 'ignore'
1745 default = {'type': 'field', 'visible': True, 'order': [], 'not_in_list': (None,)}.get(key, None)
1746 propObj[key] = default
1749 def _resolve_field_value(self, field, value, convertNone=False):
1754 value = value.lower()
1755 conversion = self._get_field_setting(field, 'convert')
1756 if conversion == 'ignore':
1758 if conversion == 'string':
1760 elif conversion == 'float_none':
1761 return float_or_none(value)
1762 elif conversion == 'bytes':
1763 return FileDownloader.parse_bytes(value)
1764 elif conversion == 'order':
1765 order_list = (self._use_free_order and self._get_field_setting(field, 'order_free')) or self._get_field_setting(field, 'order')
1766 use_regex = self._get_field_setting(field, 'regex')
1767 list_length = len(order_list)
1768 empty_pos = order_list.index('') if '' in order_list else list_length + 1
1769 if use_regex and value is not None:
1770 for i, regex in enumerate(order_list):
1771 if regex and re.match(regex, value):
1772 return list_length - i
1773 return list_length - empty_pos # not in list
1774 else: # not regex or value = None
1775 return list_length - (order_list.index(value) if value in order_list else empty_pos)
1777 if value.isnumeric():
1780 self.settings[field]['convert'] = 'string'
1783 def evaluate_params(self, params, sort_extractor):
1784 self._use_free_order = params.get('prefer_free_formats', False)
1785 self._sort_user = params.get('format_sort', [])
1786 self._sort_extractor = sort_extractor
1788 def add_item(field, reverse, closest, limit_text):
1789 field = field.lower()
1790 if field in self._order:
1792 self._order.append(field)
1793 limit = self._resolve_field_value(field, limit_text)
1796 'closest': False if limit is None else closest,
1797 'limit_text': limit_text,
1799 if field in self.settings:
1800 self.settings[field].update(data)
1802 self.settings[field] = data
1805 tuple(field for field in self.default if self._get_field_setting(field, 'forced'))
1806 + (tuple() if params.get('format_sort_force', False)
1807 else tuple(field for field in self.default if self._get_field_setting(field, 'priority')))
1808 + tuple(self._sort_user) + tuple(sort_extractor) + self.default)
1810 for item in sort_list:
1811 match = re.match(self.regex, item)
1813 raise ExtractorError('Invalid format sort string "%s" given by extractor' % item)
1814 field = match.group('field')
1817 if self._get_field_setting(field, 'type') == 'alias':
1818 alias, field = field, self._get_field_setting(field, 'field')
1819 if self._get_field_setting(alias, 'deprecated'):
1820 self.ydl.deprecation_warning(
1821 f'Format sorting alias {alias} is deprecated '
1822 f'and may be removed in a future version. Please use {field} instead')
1823 reverse = match.group('reverse') is not None
1824 closest = match.group('separator') == '~'
1825 limit_text = match.group('limit')
1827 has_limit = limit_text is not None
1828 has_multiple_fields = self._get_field_setting(field, 'type') == 'combined'
1829 has_multiple_limits = has_limit and has_multiple_fields and not self._get_field_setting(field, 'same_limit')
1831 fields = self._get_field_setting(field, 'field') if has_multiple_fields else (field,)
1832 limits = limit_text.split(':') if has_multiple_limits else (limit_text,) if has_limit else tuple()
1833 limit_count = len(limits)
1834 for (i, f) in enumerate(fields):
1835 add_item(f, reverse, closest,
1836 limits[i] if i < limit_count
1837 else limits[0] if has_limit and not has_multiple_limits
1840 def print_verbose_info(self, write_debug):
1842 write_debug('Sort order given by user: %s' % ', '.join(self._sort_user))
1843 if self._sort_extractor:
1844 write_debug('Sort order given by extractor: %s' % ', '.join(self._sort_extractor))
1845 write_debug('Formats sorted by: %s' % ', '.join(['%s%s%s' % (
1846 '+' if self._get_field_setting(field, 'reverse') else '', field,
1847 '%s%s(%s)' % ('~' if self._get_field_setting(field, 'closest') else ':',
1848 self._get_field_setting(field, 'limit_text'),
1849 self._get_field_setting(field, 'limit'))
1850 if self._get_field_setting(field, 'limit_text') is not None else '')
1851 for field in self._order if self._get_field_setting(field, 'visible')]))
1853 def _calculate_field_preference_from_value(self, format, field, type, value):
1854 reverse = self._get_field_setting(field, 'reverse')
1855 closest = self._get_field_setting(field, 'closest')
1856 limit = self._get_field_setting(field, 'limit')
1858 if type == 'extractor':
1859 maximum = self._get_field_setting(field, 'max')
1860 if value is None or (maximum is not None and value >= maximum):
1862 elif type == 'boolean':
1863 in_list = self._get_field_setting(field, 'in_list')
1864 not_in_list = self._get_field_setting(field, 'not_in_list')
1865 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
1866 elif type == 'ordered':
1867 value = self._resolve_field_value(field, value, True)
1869 # try to convert to number
1870 val_num = float_or_none(value, default=self._get_field_setting(field, 'default'))
1871 is_num = self._get_field_setting(field, 'convert') != 'string' and val_num is not None
1875 return ((-10, 0) if value is None
1876 else (1, value, 0) if not is_num # if a field has mixed strings and numbers, strings are sorted higher
1877 else (0, -abs(value - limit), value - limit if reverse else limit - value) if closest
1878 else (0, value, 0) if not reverse and (limit is None or value <= limit)
1879 else (0, -value, 0) if limit is None or (reverse and value == limit) or value > limit
1880 else (-1, value, 0))
1882 def _calculate_field_preference(self, format, field):
1883 type = self._get_field_setting(field, 'type') # extractor, boolean, ordered, field, multiple
1884 get_value = lambda f: format.get(self._get_field_setting(f, 'field'))
1885 if type == 'multiple':
1886 type = 'field' # Only 'field' is allowed in multiple for now
1887 actual_fields = self._get_field_setting(field, 'field')
1889 value = self._get_field_setting(field, 'function')(get_value(f) for f in actual_fields)
1891 value = get_value(field)
1892 return self._calculate_field_preference_from_value(format, field, type, value)
1894 def calculate_preference(self, format):
1895 # Determine missing protocol
1896 if not format.get('protocol'):
1897 format['protocol'] = determine_protocol(format)
1899 # Determine missing ext
1900 if not format.get('ext') and 'url' in format:
1901 format['ext'] = determine_ext(format['url'])
1902 if format.get('vcodec') == 'none':
1903 format['audio_ext'] = format['ext'] if format.get('acodec') != 'none' else 'none'
1904 format['video_ext'] = 'none'
1906 format['video_ext'] = format['ext']
1907 format['audio_ext'] = 'none'
1908 # if format.get('preference') is None and format.get('ext') in ('f4f', 'f4m'): # Not supported?
1909 # format['preference'] = -1000
1911 # Determine missing bitrates
1912 if format.get('tbr') is None:
1913 if format.get('vbr') is not None and format.get('abr') is not None:
1914 format['tbr'] = format.get('vbr', 0) + format.get('abr', 0)
1916 if format.get('vcodec') != 'none' and format.get('vbr') is None:
1917 format['vbr'] = format.get('tbr') - format.get('abr', 0)
1918 if format.get('acodec') != 'none' and format.get('abr') is None:
1919 format['abr'] = format.get('tbr') - format.get('vbr', 0)
1921 return tuple(self._calculate_field_preference(format, field) for field in self._order)
1923 def _sort_formats(self, formats, field_preference=[]):
1926 formats.sort(key=self.FormatSort(self, field_preference).calculate_preference)
1928 def _check_formats(self, formats, video_id):
1930 formats[:] = filter(
1931 lambda f: self._is_valid_url(
1933 item='%s video format' % f.get('format_id') if f.get('format_id') else 'video'),
1937 def _remove_duplicate_formats(formats):
1941 if f['url'] not in format_urls:
1942 format_urls.add(f['url'])
1943 unique_formats.append(f)
1944 formats[:] = unique_formats
1946 def _is_valid_url(self, url, video_id, item='video', headers={}):
1947 url = self._proto_relative_url(url, scheme='http:')
1948 # For now assume non HTTP(S) URLs always valid
1949 if not (url.startswith('http://') or url.startswith('https://')):
1952 self._request_webpage(url, video_id, 'Checking %s URL' % item, headers=headers)
1954 except ExtractorError as e:
1956 '%s: %s URL is invalid, skipping: %s'
1957 % (video_id, item, error_to_compat_str(e.cause)))
1960 def http_scheme(self):
1961 """ Either "http
:" or "https
:", depending on the user's preferences """
1964 if self.get_param('prefer_insecure', False)
1967 def _proto_relative_url(self, url, scheme=None):
1970 if url.startswith('//'):
1972 scheme = self.http_scheme()
1977 def _sleep(self, timeout, video_id, msg_template=None):
1978 if msg_template is None:
1979 msg_template = '%(video_id)s: Waiting for %(timeout)s seconds'
1980 msg = msg_template % {'video_id': video_id, 'timeout': timeout}
1984 def _extract_f4m_formats(self, manifest_url, video_id, preference=None, quality=None, f4m_id=None,
1985 transform_source=lambda s: fix_xml_ampersands(s).strip(),
1986 fatal=True, m3u8_id=None, data=None, headers={}, query={}):
1987 res = self._download_xml_handle(
1988 manifest_url, video_id, 'Downloading f4m manifest',
1989 'Unable to download f4m manifest',
1990 # Some manifests may be malformed, e.g. prosiebensat1 generated manifests
1991 # (see https://github.com/ytdl-org/youtube-dl/issues/6215#issuecomment-121704244)
1992 transform_source=transform_source,
1993 fatal=fatal, data=data, headers=headers, query=query)
1997 manifest, urlh = res
1998 manifest_url = urlh.geturl()
2000 return self._parse_f4m_formats(
2001 manifest, manifest_url, video_id, preference=preference, quality=quality, f4m_id=f4m_id,
2002 transform_source=transform_source, fatal=fatal, m3u8_id=m3u8_id)
2004 def _parse_f4m_formats(self, manifest, manifest_url, video_id, preference=None, quality=None, f4m_id=None,
2005 transform_source=lambda s: fix_xml_ampersands(s).strip(),
2006 fatal=True, m3u8_id=None):
2007 if not isinstance(manifest, xml.etree.ElementTree.Element) and not fatal:
2010 # currently yt-dlp cannot decode the playerVerificationChallenge as Akamai uses Adobe Alchemy
2011 akamai_pv = manifest.find('{http://ns.adobe.com/f4m/1.0}pv-2.0')
2012 if akamai_pv is not None and ';' in akamai_pv.text:
2013 playerVerificationChallenge = akamai_pv.text.split(';')[0]
2014 if playerVerificationChallenge.strip() != '':
2018 manifest_version = '1.0'
2019 media_nodes = manifest.findall('{http://ns.adobe.com/f4m/1.0}media')
2021 manifest_version = '2.0'
2022 media_nodes = manifest.findall('{http://ns.adobe.com/f4m/2.0}media')
2023 # Remove unsupported DRM protected media from final formats
2024 # rendition (see https://github.com/ytdl-org/youtube-dl/issues/8573).
2025 media_nodes = remove_encrypted_media(media_nodes)
2029 manifest_base_url = get_base_url(manifest)
2031 bootstrap_info = xpath_element(
2032 manifest, ['{http://ns.adobe.com/f4m/1.0}bootstrapInfo', '{http://ns.adobe.com/f4m/2.0}bootstrapInfo'],
2033 'bootstrap info', default=None)
2036 mime_type = xpath_text(
2037 manifest, ['{http://ns.adobe.com/f4m/1.0}mimeType', '{http://ns.adobe.com/f4m/2.0}mimeType'],
2038 'base URL', default=None)
2039 if mime_type and mime_type.startswith('audio/'):
2042 for i, media_el in enumerate(media_nodes):
2043 tbr = int_or_none(media_el.attrib.get('bitrate'))
2044 width = int_or_none(media_el.attrib.get('width'))
2045 height = int_or_none(media_el.attrib.get('height'))
2046 format_id = join_nonempty(f4m_id, tbr or i)
2047 # If <bootstrapInfo> is present, the specified f4m is a
2048 # stream-level manifest, and only set-level manifests may refer to
2049 # external resources. See section 11.4 and section 4 of F4M spec
2050 if bootstrap_info is None:
2052 # @href is introduced in 2.0, see section 11.6 of F4M spec
2053 if manifest_version == '2.0':
2054 media_url = media_el.attrib.get('href')
2055 if media_url is None:
2056 media_url = media_el.attrib.get('url')
2060 media_url if media_url.startswith('http://') or media_url.startswith('https://')
2061 else ((manifest_base_url or '/'.join(manifest_url.split('/')[:-1])) + '/' + media_url))
2062 # If media_url is itself a f4m manifest do the recursive extraction
2063 # since bitrates in parent manifest (this one) and media_url manifest
2064 # may differ leading to inability to resolve the format by requested
2065 # bitrate in f4m downloader
2066 ext = determine_ext(manifest_url)
2068 f4m_formats = self._extract_f4m_formats(
2069 manifest_url, video_id, preference=preference, quality=quality, f4m_id=f4m_id,
2070 transform_source=transform_source, fatal=fatal)
2071 # Sometimes stream-level manifest contains single media entry that
2072 # does not contain any quality metadata (e.g. http://matchtv.ru/#live-player).
2073 # At the same time parent's media entry in set-level manifest may
2074 # contain it. We will copy it from parent in such cases.
2075 if len(f4m_formats) == 1:
2078 'tbr': f.get('tbr') or tbr,
2079 'width': f.get('width') or width,
2080 'height': f.get('height') or height,
2081 'format_id': f.get('format_id') if not tbr else format_id,
2084 formats.extend(f4m_formats)
2087 formats.extend(self._extract_m3u8_formats(
2088 manifest_url, video_id, 'mp4', preference=preference,
2089 quality=quality, m3u8_id=m3u8_id, fatal=fatal))
2092 'format_id': format_id,
2093 'url': manifest_url,
2094 'manifest_url': manifest_url,
2095 'ext': 'flv' if bootstrap_info is not None else None,
2101 'preference': preference,
2106 def _m3u8_meta_format(self, m3u8_url, ext=None, preference=None, quality=None, m3u8_id=None):
2108 'format_id': join_nonempty(m3u8_id, 'meta'),
2112 'preference': preference - 100 if preference else -100,
2114 'resolution': 'multiple',
2115 'format_note': 'Quality selection URL',
2118 def _report_ignoring_subs(self, name):
2119 self.report_warning(bug_reports_message(
2120 f'Ignoring subtitle tracks found in the {name} manifest; '
2121 'if any subtitle tracks are missing,'
2124 def _extract_m3u8_formats(self, *args, **kwargs):
2125 fmts, subs = self._extract_m3u8_formats_and_subtitles(*args, **kwargs)
2127 self._report_ignoring_subs('HLS')
2130 def _extract_m3u8_formats_and_subtitles(
2131 self, m3u8_url, video_id, ext=None, entry_protocol='m3u8_native',
2132 preference=None, quality=None, m3u8_id=None, note=None,
2133 errnote=None, fatal=True, live=False, data=None, headers={},
2136 res = self._download_webpage_handle(
2138 note='Downloading m3u8 information' if note is None else note,
2139 errnote='Failed to download m3u8 information' if errnote is None else errnote,
2140 fatal=fatal, data=data, headers=headers, query=query)
2145 m3u8_doc, urlh = res
2146 m3u8_url = urlh.geturl()
2148 return self._parse_m3u8_formats_and_subtitles(
2149 m3u8_doc, m3u8_url, ext=ext, entry_protocol=entry_protocol,
2150 preference=preference, quality=quality, m3u8_id=m3u8_id,
2151 note=note, errnote=errnote, fatal=fatal, live=live, data=data,
2152 headers=headers, query=query, video_id=video_id)
2154 def _parse_m3u8_formats_and_subtitles(
2155 self, m3u8_doc, m3u8_url=None, ext=None, entry_protocol='m3u8_native',
2156 preference=None, quality=None, m3u8_id=None, live=False, note=None,
2157 errnote=None, fatal=True, data=None, headers={}, query={},
2159 formats, subtitles = [], {}
2161 has_drm = re.search('|'.join([
2162 r'#EXT-X-FAXS-CM:', # Adobe Flash Access
2163 r'#EXT-X-(?:SESSION-)?KEY:.*?URI="skd
://', # Apple FairPlay
2166 def format_url(url):
2167 return url if re.match(r'^https?
://', url) else urllib.parse.urljoin(m3u8_url, url)
2169 if self.get_param('hls_split_discontinuity
', False):
2170 def _extract_m3u8_playlist_indices(manifest_url=None, m3u8_doc=None):
2172 if not manifest_url:
2174 m3u8_doc = self._download_webpage(
2175 manifest_url, video_id, fatal=fatal, data=data, headers=headers,
2176 note=False, errnote='Failed to download m3u8 playlist information
')
2177 if m3u8_doc is False:
2179 return range(1 + sum(line.startswith('#EXT-X-DISCONTINUITY') for line in m3u8_doc.splitlines()))
2182 def _extract_m3u8_playlist_indices(*args
, **kwargs
):
2186 # 1. https://tools.ietf.org/html/draft-pantos-http-live-streaming-21
2187 # 2. https://github.com/ytdl-org/youtube-dl/issues/12211
2188 # 3. https://github.com/ytdl-org/youtube-dl/issues/18923
2190 # We should try extracting formats only from master playlists [1, 4.3.4],
2191 # i.e. playlists that describe available qualities. On the other hand
2192 # media playlists [1, 4.3.3] should be returned as is since they contain
2193 # just the media without qualities renditions.
2194 # Fortunately, master playlist can be easily distinguished from media
2195 # playlist based on particular tags availability. As of [1, 4.3.3, 4.3.4]
2196 # master playlist tags MUST NOT appear in a media playlist and vice versa.
2197 # As of [1, 4.3.3.1] #EXT-X-TARGETDURATION tag is REQUIRED for every
2198 # media playlist and MUST NOT appear in master playlist thus we can
2199 # clearly detect media playlist with this criterion.
2201 if '#EXT-X-TARGETDURATION' in m3u8_doc
: # media playlist, return as is
2203 'format_id': join_nonempty(m3u8_id
, idx
),
2204 'format_index': idx
,
2205 'url': m3u8_url
or encode_data_uri(m3u8_doc
.encode('utf-8'), 'application/x-mpegurl'),
2207 'protocol': entry_protocol
,
2208 'preference': preference
,
2211 } for idx
in _extract_m3u8_playlist_indices(m3u8_doc
=m3u8_doc
)]
2213 return formats
, subtitles
2216 last_stream_inf
= {}
2218 def extract_media(x_media_line
):
2219 media
= parse_m3u8_attributes(x_media_line
)
2220 # As per [1, 4.3.4.1] TYPE, GROUP-ID and NAME are REQUIRED
2221 media_type
, group_id
, name
= media
.get('TYPE'), media
.get('GROUP-ID'), media
.get('NAME')
2222 if not (media_type
and group_id
and name
):
2224 groups
.setdefault(group_id
, []).append(media
)
2225 # <https://tools.ietf.org/html/rfc8216#section-4.3.4.1>
2226 if media_type
== 'SUBTITLES':
2227 # According to RFC 8216 §4.3.4.2.1, URI is REQUIRED in the
2228 # EXT-X-MEDIA tag if the media type is SUBTITLES.
2229 # However, lack of URI has been spotted in the wild.
2230 # e.g. NebulaIE; see https://github.com/yt-dlp/yt-dlp/issues/339
2231 if not media
.get('URI'):
2233 url
= format_url(media
['URI'])
2236 'ext': determine_ext(url
),
2238 if sub_info
['ext'] == 'm3u8':
2239 # Per RFC 8216 §3.1, the only possible subtitle format m3u8
2240 # files may contain is WebVTT:
2241 # <https://tools.ietf.org/html/rfc8216#section-3.1>
2242 sub_info
['ext'] = 'vtt'
2243 sub_info
['protocol'] = 'm3u8_native'
2244 lang
= media
.get('LANGUAGE') or 'und'
2245 subtitles
.setdefault(lang
, []).append(sub_info
)
2246 if media_type
not in ('VIDEO', 'AUDIO'):
2248 media_url
= media
.get('URI')
2250 manifest_url
= format_url(media_url
)
2252 'format_id': join_nonempty(m3u8_id
, group_id
, name
, idx
),
2253 'format_note': name
,
2254 'format_index': idx
,
2255 'url': manifest_url
,
2256 'manifest_url': m3u8_url
,
2257 'language': media
.get('LANGUAGE'),
2259 'protocol': entry_protocol
,
2260 'preference': preference
,
2262 'vcodec': 'none' if media_type
== 'AUDIO' else None,
2263 } for idx
in _extract_m3u8_playlist_indices(manifest_url
))
2265 def build_stream_name():
2266 # Despite specification does not mention NAME attribute for
2267 # EXT-X-STREAM-INF tag it still sometimes may be present (see [1]
2268 # or vidio test in TestInfoExtractor.test_parse_m3u8_formats)
2269 # 1. http://www.vidio.com/watch/165683-dj_ambred-booyah-live-2015
2270 stream_name
= last_stream_inf
.get('NAME')
2273 # If there is no NAME in EXT-X-STREAM-INF it will be obtained
2274 # from corresponding rendition group
2275 stream_group_id
= last_stream_inf
.get('VIDEO')
2276 if not stream_group_id
:
2278 stream_group
= groups
.get(stream_group_id
)
2279 if not stream_group
:
2280 return stream_group_id
2281 rendition
= stream_group
[0]
2282 return rendition
.get('NAME') or stream_group_id
2284 # parse EXT-X-MEDIA tags before EXT-X-STREAM-INF in order to have the
2285 # chance to detect video only formats when EXT-X-STREAM-INF tags
2286 # precede EXT-X-MEDIA tags in HLS manifest such as [3].
2287 for line
in m3u8_doc
.splitlines():
2288 if line
.startswith('#EXT-X-MEDIA:'):
2291 for line
in m3u8_doc
.splitlines():
2292 if line
.startswith('#EXT-X-STREAM-INF:'):
2293 last_stream_inf
= parse_m3u8_attributes(line
)
2294 elif line
.startswith('#') or not line
.strip():
2297 tbr
= float_or_none(
2298 last_stream_inf
.get('AVERAGE-BANDWIDTH')
2299 or last_stream_inf
.get('BANDWIDTH'), scale
=1000)
2300 manifest_url
= format_url(line
.strip())
2302 for idx
in _extract_m3u8_playlist_indices(manifest_url
):
2303 format_id
= [m3u8_id
, None, idx
]
2304 # Bandwidth of live streams may differ over time thus making
2305 # format_id unpredictable. So it's better to keep provided
2308 stream_name
= build_stream_name()
2309 format_id
[1] = stream_name
or '%d' % (tbr
or len(formats
))
2311 'format_id': join_nonempty(*format_id
),
2312 'format_index': idx
,
2313 'url': manifest_url
,
2314 'manifest_url': m3u8_url
,
2317 'fps': float_or_none(last_stream_inf
.get('FRAME-RATE')),
2318 'protocol': entry_protocol
,
2319 'preference': preference
,
2322 resolution
= last_stream_inf
.get('RESOLUTION')
2324 mobj
= re
.search(r
'(?P<width>\d+)[xX](?P<height>\d+)', resolution
)
2326 f
['width'] = int(mobj
.group('width'))
2327 f
['height'] = int(mobj
.group('height'))
2328 # Unified Streaming Platform
2330 r
'audio.*?(?:%3D|=)(\d+)(?:-video.*?(?:%3D|=)(\d+))?', f
['url'])
2332 abr
, vbr
= mobj
.groups()
2333 abr
, vbr
= float_or_none(abr
, 1000), float_or_none(vbr
, 1000)
2338 codecs
= parse_codecs(last_stream_inf
.get('CODECS'))
2340 audio_group_id
= last_stream_inf
.get('AUDIO')
2341 # As per [1, 4.3.4.1.1] any EXT-X-STREAM-INF tag which
2342 # references a rendition group MUST have a CODECS attribute.
2343 # However, this is not always respected, for example, [2]
2344 # contains EXT-X-STREAM-INF tag which references AUDIO
2345 # rendition group but does not have CODECS and despite
2346 # referencing an audio group it represents a complete
2347 # (with audio and video) format. So, for such cases we will
2348 # ignore references to rendition groups and treat them
2349 # as complete formats.
2350 if audio_group_id
and codecs
and f
.get('vcodec') != 'none':
2351 audio_group
= groups
.get(audio_group_id
)
2352 if audio_group
and audio_group
[0].get('URI'):
2353 # TODO: update acodec for audio only formats with
2355 f
['acodec'] = 'none'
2356 if not f
.get('ext'):
2357 f
['ext'] = 'm4a' if f
.get('vcodec') == 'none' else 'mp4'
2361 progressive_uri
= last_stream_inf
.get('PROGRESSIVE-URI')
2364 del http_f
['manifest_url']
2366 'format_id': f
['format_id'].replace('hls-', 'http-'),
2368 'url': progressive_uri
,
2370 formats
.append(http_f
)
2372 last_stream_inf
= {}
2373 return formats
, subtitles
2375 def _extract_m3u8_vod_duration(
2376 self
, m3u8_vod_url
, video_id
, note
=None, errnote
=None, data
=None, headers
={}, query={}
):
2378 m3u8_vod
= self
._download
_webpage
(
2379 m3u8_vod_url
, video_id
,
2380 note
='Downloading m3u8 VOD manifest' if note
is None else note
,
2381 errnote
='Failed to download VOD manifest' if errnote
is None else errnote
,
2382 fatal
=False, data
=data
, headers
=headers
, query
=query
)
2384 return self
._parse
_m
3u8_vod
_duration
(m3u8_vod
or '', video_id
)
2386 def _parse_m3u8_vod_duration(self
, m3u8_vod
, video_id
):
2387 if '#EXT-X-PLAYLIST-TYPE:VOD' not in m3u8_vod
:
2391 float(line
[len('#EXTINF:'):].split(',')[0])
2392 for line
in m3u8_vod
.splitlines() if line
.startswith('#EXTINF:'))) or None
2395 def _xpath_ns(path
, namespace
=None):
2399 for c
in path
.split('/'):
2400 if not c
or c
== '.':
2403 out
.append('{%s}%s' % (namespace
, c
))
2404 return '/'.join(out
)
2406 def _extract_smil_formats_and_subtitles(self
, smil_url
, video_id
, fatal
=True, f4m_params
=None, transform_source
=None):
2407 res
= self
._download
_smil
(smil_url
, video_id
, fatal
=fatal
, transform_source
=transform_source
)
2413 smil_url
= urlh
.geturl()
2415 namespace
= self
._parse
_smil
_namespace
(smil
)
2417 fmts
= self
._parse
_smil
_formats
(
2418 smil
, smil_url
, video_id
, namespace
=namespace
, f4m_params
=f4m_params
)
2419 subs
= self
._parse
_smil
_subtitles
(
2420 smil
, namespace
=namespace
)
2424 def _extract_smil_formats(self
, *args
, **kwargs
):
2425 fmts
, subs
= self
._extract
_smil
_formats
_and
_subtitles
(*args
, **kwargs
)
2427 self
._report
_ignoring
_subs
('SMIL')
2430 def _extract_smil_info(self
, smil_url
, video_id
, fatal
=True, f4m_params
=None):
2431 res
= self
._download
_smil
(smil_url
, video_id
, fatal
=fatal
)
2436 smil_url
= urlh
.geturl()
2438 return self
._parse
_smil
(smil
, smil_url
, video_id
, f4m_params
=f4m_params
)
2440 def _download_smil(self
, smil_url
, video_id
, fatal
=True, transform_source
=None):
2441 return self
._download
_xml
_handle
(
2442 smil_url
, video_id
, 'Downloading SMIL file',
2443 'Unable to download SMIL file', fatal
=fatal
, transform_source
=transform_source
)
2445 def _parse_smil(self
, smil
, smil_url
, video_id
, f4m_params
=None):
2446 namespace
= self
._parse
_smil
_namespace
(smil
)
2448 formats
= self
._parse
_smil
_formats
(
2449 smil
, smil_url
, video_id
, namespace
=namespace
, f4m_params
=f4m_params
)
2450 subtitles
= self
._parse
_smil
_subtitles
(smil
, namespace
=namespace
)
2452 video_id
= os
.path
.splitext(url_basename(smil_url
))[0]
2456 for meta
in smil
.findall(self
._xpath
_ns
('./head/meta', namespace
)):
2457 name
= meta
.attrib
.get('name')
2458 content
= meta
.attrib
.get('content')
2459 if not name
or not content
:
2461 if not title
and name
== 'title':
2463 elif not description
and name
in ('description', 'abstract'):
2464 description
= content
2465 elif not upload_date
and name
== 'date':
2466 upload_date
= unified_strdate(content
)
2469 'id': image
.get('type'),
2470 'url': image
.get('src'),
2471 'width': int_or_none(image
.get('width')),
2472 'height': int_or_none(image
.get('height')),
2473 } for image
in smil
.findall(self
._xpath
_ns
('.//image', namespace
)) if image
.get('src')]
2477 'title': title
or video_id
,
2478 'description': description
,
2479 'upload_date': upload_date
,
2480 'thumbnails': thumbnails
,
2482 'subtitles': subtitles
,
2485 def _parse_smil_namespace(self
, smil
):
2486 return self
._search
_regex
(
2487 r
'(?i)^{([^}]+)?}smil$', smil
.tag
, 'namespace', default
=None)
2489 def _parse_smil_formats(self
, smil
, smil_url
, video_id
, namespace
=None, f4m_params
=None, transform_rtmp_url
=None):
2491 for meta
in smil
.findall(self
._xpath
_ns
('./head/meta', namespace
)):
2492 b
= meta
.get('base') or meta
.get('httpBase')
2504 media
= smil
.findall(self
._xpath
_ns
('.//video', namespace
)) + smil
.findall(self
._xpath
_ns
('.//audio', namespace
))
2505 for medium
in media
:
2506 src
= medium
.get('src')
2507 if not src
or src
in srcs
:
2511 bitrate
= float_or_none(medium
.get('system-bitrate') or medium
.get('systemBitrate'), 1000)
2512 filesize
= int_or_none(medium
.get('size') or medium
.get('fileSize'))
2513 width
= int_or_none(medium
.get('width'))
2514 height
= int_or_none(medium
.get('height'))
2515 proto
= medium
.get('proto')
2516 ext
= medium
.get('ext')
2517 src_ext
= determine_ext(src
)
2518 streamer
= medium
.get('streamer') or base
2520 if proto
== 'rtmp' or streamer
.startswith('rtmp'):
2526 'format_id': 'rtmp-%d' % (rtmp_count
if bitrate
is None else bitrate
),
2528 'filesize': filesize
,
2532 if transform_rtmp_url
:
2533 streamer
, src
= transform_rtmp_url(streamer
, src
)
2534 formats
[-1].update({
2540 src_url
= src
if src
.startswith('http') else urllib
.parse
.urljoin(base
, src
)
2541 src_url
= src_url
.strip()
2543 if proto
== 'm3u8' or src_ext
== 'm3u8':
2544 m3u8_formats
= self
._extract
_m
3u8_formats
(
2545 src_url
, video_id
, ext
or 'mp4', m3u8_id
='hls', fatal
=False)
2546 if len(m3u8_formats
) == 1:
2548 m3u8_formats
[0].update({
2549 'format_id': 'hls-%d' % (m3u8_count
if bitrate
is None else bitrate
),
2554 formats
.extend(m3u8_formats
)
2555 elif src_ext
== 'f4m':
2560 'plugin': 'flowplayer-3.2.0.1',
2562 f4m_url
+= '&' if '?' in f4m_url
else '?'
2563 f4m_url
+= urllib
.parse
.urlencode(f4m_params
)
2564 formats
.extend(self
._extract
_f
4m
_formats
(f4m_url
, video_id
, f4m_id
='hds', fatal
=False))
2565 elif src_ext
== 'mpd':
2566 formats
.extend(self
._extract
_mpd
_formats
(
2567 src_url
, video_id
, mpd_id
='dash', fatal
=False))
2568 elif re
.search(r
'\.ism/[Mm]anifest', src_url
):
2569 formats
.extend(self
._extract
_ism
_formats
(
2570 src_url
, video_id
, ism_id
='mss', fatal
=False))
2571 elif src_url
.startswith('http') and self
._is
_valid
_url
(src
, video_id
):
2575 'ext': ext
or src_ext
or 'flv',
2576 'format_id': 'http-%d' % (bitrate
or http_count
),
2578 'filesize': filesize
,
2583 for medium
in smil
.findall(self
._xpath
_ns
('.//imagestream', namespace
)):
2584 src
= medium
.get('src')
2585 if not src
or src
in srcs
:
2591 'format_id': 'imagestream-%d' % (imgs_count
),
2593 'ext': mimetype2ext(medium
.get('type')),
2596 'width': int_or_none(medium
.get('width')),
2597 'height': int_or_none(medium
.get('height')),
2598 'format_note': 'SMIL storyboards',
2603 def _parse_smil_subtitles(self
, smil
, namespace
=None, subtitles_lang
='en'):
2606 for num
, textstream
in enumerate(smil
.findall(self
._xpath
_ns
('.//textstream', namespace
))):
2607 src
= textstream
.get('src')
2608 if not src
or src
in urls
:
2611 ext
= textstream
.get('ext') or mimetype2ext(textstream
.get('type')) or determine_ext(src
)
2612 lang
= textstream
.get('systemLanguage') or textstream
.get('systemLanguageName') or textstream
.get('lang') or subtitles_lang
2613 subtitles
.setdefault(lang
, []).append({
2619 def _extract_xspf_playlist(self
, xspf_url
, playlist_id
, fatal
=True):
2620 res
= self
._download
_xml
_handle
(
2621 xspf_url
, playlist_id
, 'Downloading xpsf playlist',
2622 'Unable to download xspf manifest', fatal
=fatal
)
2627 xspf_url
= urlh
.geturl()
2629 return self
._parse
_xspf
(
2630 xspf
, playlist_id
, xspf_url
=xspf_url
,
2631 xspf_base_url
=base_url(xspf_url
))
2633 def _parse_xspf(self
, xspf_doc
, playlist_id
, xspf_url
=None, xspf_base_url
=None):
2635 'xspf': 'http://xspf.org/ns/0/',
2636 's1': 'http://static.streamone.nl/player/ns/0',
2640 for track
in xspf_doc
.findall(xpath_with_ns('./xspf:trackList/xspf:track', NS_MAP
)):
2642 track
, xpath_with_ns('./xspf:title', NS_MAP
), 'title', default
=playlist_id
)
2643 description
= xpath_text(
2644 track
, xpath_with_ns('./xspf:annotation', NS_MAP
), 'description')
2645 thumbnail
= xpath_text(
2646 track
, xpath_with_ns('./xspf:image', NS_MAP
), 'thumbnail')
2647 duration
= float_or_none(
2648 xpath_text(track
, xpath_with_ns('./xspf:duration', NS_MAP
), 'duration'), 1000)
2651 for location
in track
.findall(xpath_with_ns('./xspf:location', NS_MAP
)):
2652 format_url
= urljoin(xspf_base_url
, location
.text
)
2657 'manifest_url': xspf_url
,
2658 'format_id': location
.get(xpath_with_ns('s1:label', NS_MAP
)),
2659 'width': int_or_none(location
.get(xpath_with_ns('s1:width', NS_MAP
))),
2660 'height': int_or_none(location
.get(xpath_with_ns('s1:height', NS_MAP
))),
2662 self
._sort
_formats
(formats
)
2667 'description': description
,
2668 'thumbnail': thumbnail
,
2669 'duration': duration
,
2674 def _extract_mpd_formats(self
, *args
, **kwargs
):
2675 fmts
, subs
= self
._extract
_mpd
_formats
_and
_subtitles
(*args
, **kwargs
)
2677 self
._report
_ignoring
_subs
('DASH')
2680 def _extract_mpd_formats_and_subtitles(
2681 self
, mpd_url
, video_id
, mpd_id
=None, note
=None, errnote
=None,
2682 fatal
=True, data
=None, headers
={}, query={}
):
2683 res
= self
._download
_xml
_handle
(
2685 note
='Downloading MPD manifest' if note
is None else note
,
2686 errnote
='Failed to download MPD manifest' if errnote
is None else errnote
,
2687 fatal
=fatal
, data
=data
, headers
=headers
, query
=query
)
2694 # We could have been redirected to a new url when we retrieved our mpd file.
2695 mpd_url
= urlh
.geturl()
2696 mpd_base_url
= base_url(mpd_url
)
2698 return self
._parse
_mpd
_formats
_and
_subtitles
(
2699 mpd_doc
, mpd_id
, mpd_base_url
, mpd_url
)
2701 def _parse_mpd_formats(self
, *args
, **kwargs
):
2702 fmts
, subs
= self
._parse
_mpd
_formats
_and
_subtitles
(*args
, **kwargs
)
2704 self
._report
_ignoring
_subs
('DASH')
2707 def _parse_mpd_formats_and_subtitles(
2708 self
, mpd_doc
, mpd_id
=None, mpd_base_url
='', mpd_url
=None):
2710 Parse formats from MPD manifest.
2712 1. MPEG-DASH Standard, ISO/IEC 23009-1:2014(E),
2713 http://standards.iso.org/ittf/PubliclyAvailableStandards/c065274_ISO_IEC_23009-1_2014.zip
2714 2. https://en.wikipedia.org/wiki/Dynamic_Adaptive_Streaming_over_HTTP
2716 if not self
.get_param('dynamic_mpd', True):
2717 if mpd_doc
.get('type') == 'dynamic':
2720 namespace
= self
._search
_regex
(r
'(?i)^{([^}]+)?}MPD$', mpd_doc
.tag
, 'namespace', default
=None)
2723 return self
._xpath
_ns
(path
, namespace
)
2725 def is_drm_protected(element
):
2726 return element
.find(_add_ns('ContentProtection')) is not None
2728 def extract_multisegment_info(element
, ms_parent_info
):
2729 ms_info
= ms_parent_info
.copy()
2731 # As per [1, 5.3.9.2.2] SegmentList and SegmentTemplate share some
2732 # common attributes and elements. We will only extract relevant
2734 def extract_common(source
):
2735 segment_timeline
= source
.find(_add_ns('SegmentTimeline'))
2736 if segment_timeline
is not None:
2737 s_e
= segment_timeline
.findall(_add_ns('S'))
2739 ms_info
['total_number'] = 0
2742 r
= int(s
.get('r', 0))
2743 ms_info
['total_number'] += 1 + r
2744 ms_info
['s'].append({
2745 't': int(s
.get('t', 0)),
2746 # @d is mandatory (see [1, 5.3.9.6.2, Table 17, page 60])
2747 'd': int(s
.attrib
['d']),
2750 start_number
= source
.get('startNumber')
2752 ms_info
['start_number'] = int(start_number
)
2753 timescale
= source
.get('timescale')
2755 ms_info
['timescale'] = int(timescale
)
2756 segment_duration
= source
.get('duration')
2757 if segment_duration
:
2758 ms_info
['segment_duration'] = float(segment_duration
)
2760 def extract_Initialization(source
):
2761 initialization
= source
.find(_add_ns('Initialization'))
2762 if initialization
is not None:
2763 ms_info
['initialization_url'] = initialization
.attrib
['sourceURL']
2765 segment_list
= element
.find(_add_ns('SegmentList'))
2766 if segment_list
is not None:
2767 extract_common(segment_list
)
2768 extract_Initialization(segment_list
)
2769 segment_urls_e
= segment_list
.findall(_add_ns('SegmentURL'))
2771 ms_info
['segment_urls'] = [segment
.attrib
['media'] for segment
in segment_urls_e
]
2773 segment_template
= element
.find(_add_ns('SegmentTemplate'))
2774 if segment_template
is not None:
2775 extract_common(segment_template
)
2776 media
= segment_template
.get('media')
2778 ms_info
['media'] = media
2779 initialization
= segment_template
.get('initialization')
2781 ms_info
['initialization'] = initialization
2783 extract_Initialization(segment_template
)
2786 mpd_duration
= parse_duration(mpd_doc
.get('mediaPresentationDuration'))
2787 formats
, subtitles
= [], {}
2788 stream_numbers
= collections
.defaultdict(int)
2789 for period
in mpd_doc
.findall(_add_ns('Period')):
2790 period_duration
= parse_duration(period
.get('duration')) or mpd_duration
2791 period_ms_info
= extract_multisegment_info(period
, {
2795 for adaptation_set
in period
.findall(_add_ns('AdaptationSet')):
2796 adaption_set_ms_info
= extract_multisegment_info(adaptation_set
, period_ms_info
)
2797 for representation
in adaptation_set
.findall(_add_ns('Representation')):
2798 representation_attrib
= adaptation_set
.attrib
.copy()
2799 representation_attrib
.update(representation
.attrib
)
2800 # According to [1, 5.3.7.2, Table 9, page 41], @mimeType is mandatory
2801 mime_type
= representation_attrib
['mimeType']
2802 content_type
= representation_attrib
.get('contentType', mime_type
.split('/')[0])
2804 codec_str
= representation_attrib
.get('codecs', '')
2805 # Some kind of binary subtitle found in some youtube livestreams
2806 if mime_type
== 'application/x-rawcc':
2807 codecs
= {'scodec': codec_str}
2809 codecs
= parse_codecs(codec_str
)
2810 if content_type
not in ('video', 'audio', 'text'):
2811 if mime_type
== 'image/jpeg':
2812 content_type
= mime_type
2813 elif codecs
.get('vcodec', 'none') != 'none':
2814 content_type
= 'video'
2815 elif codecs
.get('acodec', 'none') != 'none':
2816 content_type
= 'audio'
2817 elif codecs
.get('scodec', 'none') != 'none':
2818 content_type
= 'text'
2819 elif mimetype2ext(mime_type
) in ('tt', 'dfxp', 'ttml', 'xml', 'json'):
2820 content_type
= 'text'
2822 self
.report_warning('Unknown MIME type %s in DASH manifest' % mime_type
)
2826 for element
in (representation
, adaptation_set
, period
, mpd_doc
):
2827 base_url_e
= element
.find(_add_ns('BaseURL'))
2828 if try_call(lambda: base_url_e
.text
) is not None:
2829 base_url
= base_url_e
.text
+ base_url
2830 if re
.match(r
'^https?://', base_url
):
2832 if mpd_base_url
and base_url
.startswith('/'):
2833 base_url
= urllib
.parse
.urljoin(mpd_base_url
, base_url
)
2834 elif mpd_base_url
and not re
.match(r
'^https?://', base_url
):
2835 if not mpd_base_url
.endswith('/'):
2837 base_url
= mpd_base_url
+ base_url
2838 representation_id
= representation_attrib
.get('id')
2839 lang
= representation_attrib
.get('lang')
2840 url_el
= representation
.find(_add_ns('BaseURL'))
2841 filesize
= int_or_none(url_el
.attrib
.get('{http://youtube.com/yt/2012/10/10}contentLength') if url_el
is not None else None)
2842 bandwidth
= int_or_none(representation_attrib
.get('bandwidth'))
2843 if representation_id
is not None:
2844 format_id
= representation_id
2846 format_id
= content_type
2848 format_id
= mpd_id
+ '-' + format_id
2849 if content_type
in ('video', 'audio'):
2851 'format_id': format_id
,
2852 'manifest_url': mpd_url
,
2853 'ext': mimetype2ext(mime_type
),
2854 'width': int_or_none(representation_attrib
.get('width')),
2855 'height': int_or_none(representation_attrib
.get('height')),
2856 'tbr': float_or_none(bandwidth
, 1000),
2857 'asr': int_or_none(representation_attrib
.get('audioSamplingRate')),
2858 'fps': int_or_none(representation_attrib
.get('frameRate')),
2859 'language': lang
if lang
not in ('mul', 'und', 'zxx', 'mis') else None,
2860 'format_note': 'DASH %s' % content_type
,
2861 'filesize': filesize
,
2862 'container': mimetype2ext(mime_type
) + '_dash',
2865 elif content_type
== 'text':
2867 'ext': mimetype2ext(mime_type
),
2868 'manifest_url': mpd_url
,
2869 'filesize': filesize
,
2871 elif content_type
== 'image/jpeg':
2872 # See test case in VikiIE
2873 # https://www.viki.com/videos/1175236v-choosing-spouse-by-lottery-episode-1
2875 'format_id': format_id
,
2877 'manifest_url': mpd_url
,
2878 'format_note': 'DASH storyboards (jpeg)',
2882 if is_drm_protected(adaptation_set
) or is_drm_protected(representation
):
2884 representation_ms_info
= extract_multisegment_info(representation
, adaption_set_ms_info
)
2886 def prepare_template(template_name
, identifiers
):
2887 tmpl
= representation_ms_info
[template_name
]
2888 # First of, % characters outside $...$ templates
2889 # must be escaped by doubling for proper processing
2890 # by % operator string formatting used further (see
2891 # https://github.com/ytdl-org/youtube-dl/issues/16867).
2897 in_template
= not in_template
2898 elif c
== '%' and not in_template
:
2900 # Next, $...$ templates are translated to their
2901 # %(...) counterparts to be used with % operator
2902 if representation_id
is not None:
2903 t
= t
.replace('$RepresentationID$', representation_id
)
2904 t
= re
.sub(r
'\$(%s)\$' % '|'.join(identifiers
), r
'%(\1)d', t
)
2905 t
= re
.sub(r
'\$(%s)%%([^$]+)\$' % '|'.join(identifiers
), r
'%(\1)\2', t
)
2906 t
.replace('$$', '$')
2909 # @initialization is a regular template like @media one
2910 # so it should be handled just the same way (see
2911 # https://github.com/ytdl-org/youtube-dl/issues/11605)
2912 if 'initialization' in representation_ms_info
:
2913 initialization_template
= prepare_template(
2915 # As per [1, 5.3.9.4.2, Table 15, page 54] $Number$ and
2916 # $Time$ shall not be included for @initialization thus
2917 # only $Bandwidth$ remains
2919 representation_ms_info
['initialization_url'] = initialization_template
% {
2920 'Bandwidth': bandwidth
,
2923 def location_key(location
):
2924 return 'url' if re
.match(r
'^https?://', location
) else 'path'
2926 if 'segment_urls' not in representation_ms_info
and 'media' in representation_ms_info
:
2928 media_template
= prepare_template('media', ('Number', 'Bandwidth', 'Time'))
2929 media_location_key
= location_key(media_template
)
2931 # As per [1, 5.3.9.4.4, Table 16, page 55] $Number$ and $Time$
2932 # can't be used at the same time
2933 if '%(Number' in media_template
and 's' not in representation_ms_info
:
2934 segment_duration
= None
2935 if 'total_number' not in representation_ms_info
and 'segment_duration' in representation_ms_info
:
2936 segment_duration
= float_or_none(representation_ms_info
['segment_duration'], representation_ms_info
['timescale'])
2937 representation_ms_info
['total_number'] = int(math
.ceil(
2938 float_or_none(period_duration
, segment_duration
, default
=0)))
2939 representation_ms_info
['fragments'] = [{
2940 media_location_key
: media_template
% {
2941 'Number': segment_number
,
2942 'Bandwidth': bandwidth
,
2944 'duration': segment_duration
,
2945 } for segment_number
in range(
2946 representation_ms_info
['start_number'],
2947 representation_ms_info
['total_number'] + representation_ms_info
['start_number'])]
2949 # $Number*$ or $Time$ in media template with S list available
2950 # Example $Number*$: http://www.svtplay.se/klipp/9023742/stopptid-om-bjorn-borg
2951 # Example $Time$: https://play.arkena.com/embed/avp/v2/player/media/b41dda37-d8e7-4d3f-b1b5-9a9db578bdfe/1/129411
2952 representation_ms_info
['fragments'] = []
2955 segment_number
= representation_ms_info
['start_number']
2957 def add_segment_url():
2958 segment_url
= media_template
% {
2959 'Time': segment_time
,
2960 'Bandwidth': bandwidth
,
2961 'Number': segment_number
,
2963 representation_ms_info
['fragments'].append({
2964 media_location_key
: segment_url
,
2965 'duration': float_or_none(segment_d
, representation_ms_info
['timescale']),
2968 for num
, s
in enumerate(representation_ms_info
['s']):
2969 segment_time
= s
.get('t') or segment_time
2973 for r
in range(s
.get('r', 0)):
2974 segment_time
+= segment_d
2977 segment_time
+= segment_d
2978 elif 'segment_urls' in representation_ms_info
and 's' in representation_ms_info
:
2980 # Example: https://www.youtube.com/watch?v=iXZV5uAYMJI
2981 # or any YouTube dashsegments video
2984 timescale
= representation_ms_info
['timescale']
2985 for s
in representation_ms_info
['s']:
2986 duration
= float_or_none(s
['d'], timescale
)
2987 for r
in range(s
.get('r', 0) + 1):
2988 segment_uri
= representation_ms_info
['segment_urls'][segment_index
]
2990 location_key(segment_uri
): segment_uri
,
2991 'duration': duration
,
2994 representation_ms_info
['fragments'] = fragments
2995 elif 'segment_urls' in representation_ms_info
:
2996 # Segment URLs with no SegmentTimeline
2997 # Example: https://www.seznam.cz/zpravy/clanek/cesko-zasahne-vitr-o-sile-vichrice-muze-byt-i-zivotu-nebezpecny-39091
2998 # https://github.com/ytdl-org/youtube-dl/pull/14844
3000 segment_duration
= float_or_none(
3001 representation_ms_info
['segment_duration'],
3002 representation_ms_info
['timescale']) if 'segment_duration' in representation_ms_info
else None
3003 for segment_url
in representation_ms_info
['segment_urls']:
3005 location_key(segment_url
): segment_url
,
3007 if segment_duration
:
3008 fragment
['duration'] = segment_duration
3009 fragments
.append(fragment
)
3010 representation_ms_info
['fragments'] = fragments
3011 # If there is a fragments key available then we correctly recognized fragmented media.
3012 # Otherwise we will assume unfragmented media with direct access. Technically, such
3013 # assumption is not necessarily correct since we may simply have no support for
3014 # some forms of fragmented media renditions yet, but for now we'll use this fallback.
3015 if 'fragments' in representation_ms_info
:
3017 # NB: mpd_url may be empty when MPD manifest is parsed from a string
3018 'url': mpd_url
or base_url
,
3019 'fragment_base_url': base_url
,
3021 'protocol': 'http_dash_segments' if mime_type
!= 'image/jpeg' else 'mhtml',
3023 if 'initialization_url' in representation_ms_info
:
3024 initialization_url
= representation_ms_info
['initialization_url']
3025 if not f
.get('url'):
3026 f
['url'] = initialization_url
3027 f
['fragments'].append({location_key(initialization_url): initialization_url}
)
3028 f
['fragments'].extend(representation_ms_info
['fragments'])
3029 if not period_duration
:
3030 period_duration
= try_get(
3031 representation_ms_info
,
3032 lambda r
: sum(frag
['duration'] for frag
in r
['fragments']), float)
3034 # Assuming direct URL to unfragmented media.
3036 if content_type
in ('video', 'audio', 'image/jpeg'):
3037 f
['manifest_stream_number'] = stream_numbers
[f
['url']]
3038 stream_numbers
[f
['url']] += 1
3040 elif content_type
== 'text':
3041 subtitles
.setdefault(lang
or 'und', []).append(f
)
3043 return formats
, subtitles
3045 def _extract_ism_formats(self
, *args
, **kwargs
):
3046 fmts
, subs
= self
._extract
_ism
_formats
_and
_subtitles
(*args
, **kwargs
)
3048 self
._report
_ignoring
_subs
('ISM')
3051 def _extract_ism_formats_and_subtitles(self
, ism_url
, video_id
, ism_id
=None, note
=None, errnote
=None, fatal
=True, data
=None, headers
={}, query={}
):
3052 res
= self
._download
_xml
_handle
(
3054 note
='Downloading ISM manifest' if note
is None else note
,
3055 errnote
='Failed to download ISM manifest' if errnote
is None else errnote
,
3056 fatal
=fatal
, data
=data
, headers
=headers
, query
=query
)
3063 return self
._parse
_ism
_formats
_and
_subtitles
(ism_doc
, urlh
.geturl(), ism_id
)
3065 def _parse_ism_formats_and_subtitles(self
, ism_doc
, ism_url
, ism_id
=None):
3067 Parse formats from ISM manifest.
3069 1. [MS-SSTR]: Smooth Streaming Protocol,
3070 https://msdn.microsoft.com/en-us/library/ff469518.aspx
3072 if ism_doc
.get('IsLive') == 'TRUE':
3075 duration
= int(ism_doc
.attrib
['Duration'])
3076 timescale
= int_or_none(ism_doc
.get('TimeScale')) or 10000000
3080 for stream
in ism_doc
.findall('StreamIndex'):
3081 stream_type
= stream
.get('Type')
3082 if stream_type
not in ('video', 'audio', 'text'):
3084 url_pattern
= stream
.attrib
['Url']
3085 stream_timescale
= int_or_none(stream
.get('TimeScale')) or timescale
3086 stream_name
= stream
.get('Name')
3087 stream_language
= stream
.get('Language', 'und')
3088 for track
in stream
.findall('QualityLevel'):
3089 fourcc
= track
.get('FourCC') or ('AACL' if track
.get('AudioTag') == '255' else None)
3090 # TODO: add support for WVC1 and WMAP
3091 if fourcc
not in ('H264', 'AVC1', 'AACL', 'TTML'):
3092 self
.report_warning('%s is not a supported codec' % fourcc
)
3094 tbr
= int(track
.attrib
['Bitrate']) // 1000
3095 # [1] does not mention Width and Height attributes. However,
3096 # they're often present while MaxWidth and MaxHeight are
3097 # missing, so should be used as fallbacks
3098 width
= int_or_none(track
.get('MaxWidth') or track
.get('Width'))
3099 height
= int_or_none(track
.get('MaxHeight') or track
.get('Height'))
3100 sampling_rate
= int_or_none(track
.get('SamplingRate'))
3102 track_url_pattern
= re
.sub(r
'{[Bb]itrate}', track
.attrib
['Bitrate'], url_pattern
)
3103 track_url_pattern
= urllib
.parse
.urljoin(ism_url
, track_url_pattern
)
3109 stream_fragments
= stream
.findall('c')
3110 for stream_fragment_index
, stream_fragment
in enumerate(stream_fragments
):
3111 fragment_ctx
['time'] = int_or_none(stream_fragment
.get('t')) or fragment_ctx
['time']
3112 fragment_repeat
= int_or_none(stream_fragment
.get('r')) or 1
3113 fragment_ctx
['duration'] = int_or_none(stream_fragment
.get('d'))
3114 if not fragment_ctx
['duration']:
3116 next_fragment_time
= int(stream_fragment
[stream_fragment_index
+ 1].attrib
['t'])
3118 next_fragment_time
= duration
3119 fragment_ctx
['duration'] = (next_fragment_time
- fragment_ctx
['time']) / fragment_repeat
3120 for _
in range(fragment_repeat
):
3122 'url': re
.sub(r
'{start[ _]time}', str(fragment_ctx
['time']), track_url_pattern
),
3123 'duration': fragment_ctx
['duration'] / stream_timescale
,
3125 fragment_ctx
['time'] += fragment_ctx
['duration']
3127 if stream_type
== 'text':
3128 subtitles
.setdefault(stream_language
, []).append({
3132 'manifest_url': ism_url
,
3133 'fragments': fragments
,
3134 '_download_params': {
3135 'stream_type': stream_type
,
3136 'duration': duration
,
3137 'timescale': stream_timescale
,
3139 'language': stream_language
,
3140 'codec_private_data': track
.get('CodecPrivateData'),
3143 elif stream_type
in ('video', 'audio'):
3145 'format_id': join_nonempty(ism_id
, stream_name
, tbr
),
3147 'manifest_url': ism_url
,
3148 'ext': 'ismv' if stream_type
== 'video' else 'isma',
3152 'asr': sampling_rate
,
3153 'vcodec': 'none' if stream_type
== 'audio' else fourcc
,
3154 'acodec': 'none' if stream_type
== 'video' else fourcc
,
3156 'fragments': fragments
,
3157 'has_drm': ism_doc
.find('Protection') is not None,
3158 '_download_params': {
3159 'stream_type': stream_type
,
3160 'duration': duration
,
3161 'timescale': stream_timescale
,
3162 'width': width
or 0,
3163 'height': height
or 0,
3165 'language': stream_language
,
3166 'codec_private_data': track
.get('CodecPrivateData'),
3167 'sampling_rate': sampling_rate
,
3168 'channels': int_or_none(track
.get('Channels', 2)),
3169 'bits_per_sample': int_or_none(track
.get('BitsPerSample', 16)),
3170 'nal_unit_length_field': int_or_none(track
.get('NALUnitLengthField', 4)),
3173 return formats
, subtitles
3175 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):
3176 def absolute_url(item_url
):
3177 return urljoin(base_url
, item_url
)
3179 def parse_content_type(content_type
):
3180 if not content_type
:
3182 ctr
= re
.search(r
'(?P<mimetype>[^/]+/[^;]+)(?:;\s*codecs="?(?P<codecs>[^"]+))?', content_type
)
3184 mimetype
, codecs
= ctr
.groups()
3185 f
= parse_codecs(codecs
)
3186 f
['ext'] = mimetype2ext(mimetype
)
3190 def _media_formats(src
, cur_media_type
, type_info
=None):
3191 type_info
= type_info
or {}
3192 full_url
= absolute_url(src
)
3193 ext
= type_info
.get('ext') or determine_ext(full_url
)
3195 is_plain_url
= False
3196 formats
= self
._extract
_m
3u8_formats
(
3197 full_url
, video_id
, ext
='mp4',
3198 entry_protocol
=m3u8_entry_protocol
, m3u8_id
=m3u8_id
,
3199 preference
=preference
, quality
=quality
, fatal
=False)
3201 is_plain_url
= False
3202 formats
= self
._extract
_mpd
_formats
(
3203 full_url
, video_id
, mpd_id
=mpd_id
, fatal
=False)
3208 'vcodec': 'none' if cur_media_type
== 'audio' else None,
3211 return is_plain_url
, formats
3214 # amp-video and amp-audio are very similar to their HTML5 counterparts
3215 # so we will include them right here (see
3216 # https://www.ampproject.org/docs/reference/components/amp-video)
3217 # For dl8-* tags see https://delight-vr.com/documentation/dl8-video/
3218 _MEDIA_TAG_NAME_RE
= r
'(?:(?:amp|dl8(?:-live)?)-)?(video|audio)'
3219 media_tags
= [(media_tag
, media_tag_name
, media_type
, '')
3220 for media_tag
, media_tag_name
, media_type
3221 in re
.findall(r
'(?s)(<(%s)[^>]*/>)' % _MEDIA_TAG_NAME_RE
, webpage
)]
3222 media_tags
.extend(re
.findall(
3223 # We only allow video|audio followed by a whitespace or '>'.
3224 # Allowing more characters may end up in significant slow down (see
3225 # https://github.com/ytdl-org/youtube-dl/issues/11979, example URL:
3226 # http://www.porntrex.com/maps/videositemap.xml).
3227 r
'(?s)(<(?P<tag>%s)(?:\s+[^>]*)?>)(.*?)</(?P=tag)>' % _MEDIA_TAG_NAME_RE
, webpage
))
3228 for media_tag
, _
, media_type
, media_content
in media_tags
:
3233 media_attributes
= extract_attributes(media_tag
)
3234 src
= strip_or_none(media_attributes
.get('src'))
3236 f
= parse_content_type(media_attributes
.get('type'))
3237 _
, formats
= _media_formats(src
, media_type
, f
)
3238 media_info
['formats'].extend(formats
)
3239 media_info
['thumbnail'] = absolute_url(media_attributes
.get('poster'))
3241 for source_tag
in re
.findall(r
'<source[^>]+>', media_content
):
3242 s_attr
= extract_attributes(source_tag
)
3243 # data-video-src and data-src are non standard but seen
3244 # several times in the wild
3245 src
= strip_or_none(dict_get(s_attr
, ('src', 'data-video-src', 'data-src')))
3248 f
= parse_content_type(s_attr
.get('type'))
3249 is_plain_url
, formats
= _media_formats(src
, media_type
, f
)
3251 # width, height, res, label and title attributes are
3252 # all not standard but seen several times in the wild
3255 for lbl
in ('label', 'title')
3256 if str_or_none(s_attr
.get(lbl
))
3258 width
= int_or_none(s_attr
.get('width'))
3259 height
= (int_or_none(s_attr
.get('height'))
3260 or int_or_none(s_attr
.get('res')))
3261 if not width
or not height
:
3263 resolution
= parse_resolution(lbl
)
3266 width
= width
or resolution
.get('width')
3267 height
= height
or resolution
.get('height')
3269 tbr
= parse_bitrate(lbl
)
3278 'format_id': s_attr
.get('label') or s_attr
.get('title'),
3280 f
.update(formats
[0])
3281 media_info
['formats'].append(f
)
3283 media_info
['formats'].extend(formats
)
3284 for track_tag
in re
.findall(r
'<track[^>]+>', media_content
):
3285 track_attributes
= extract_attributes(track_tag
)
3286 kind
= track_attributes
.get('kind')
3287 if not kind
or kind
in ('subtitles', 'captions'):
3288 src
= strip_or_none(track_attributes
.get('src'))
3291 lang
= track_attributes
.get('srclang') or track_attributes
.get('lang') or track_attributes
.get('label')
3292 media_info
['subtitles'].setdefault(lang
, []).append({
3293 'url': absolute_url(src
),
3295 for f
in media_info
['formats']:
3296 f
.setdefault('http_headers', {})['Referer'] = base_url
3297 if media_info
['formats'] or media_info
['subtitles']:
3298 entries
.append(media_info
)
3301 def _extract_akamai_formats(self
, *args
, **kwargs
):
3302 fmts
, subs
= self
._extract
_akamai
_formats
_and
_subtitles
(*args
, **kwargs
)
3304 self
._report
_ignoring
_subs
('akamai')
3307 def _extract_akamai_formats_and_subtitles(self
, manifest_url
, video_id
, hosts
={}):
3308 signed
= 'hdnea=' in manifest_url
3310 # https://learn.akamai.com/en-us/webhelp/media-services-on-demand/stream-packaging-user-guide/GUID-BE6C0F73-1E06-483B-B0EA-57984B91B7F9.html
3311 manifest_url
= re
.sub(
3312 r
'(?:b=[\d,-]+|(?:__a__|attributes)=off|__b__=\d+)&?',
3313 '', manifest_url
).strip('?')
3318 hdcore_sign
= 'hdcore=3.7.0'
3319 f4m_url
= re
.sub(r
'(https?://[^/]+)/i/', r
'\1/z/', manifest_url
).replace('/master.m3u8', '/manifest.f4m')
3320 hds_host
= hosts
.get('hds')
3322 f4m_url
= re
.sub(r
'(https?://)[^/]+', r
'\1' + hds_host
, f4m_url
)
3323 if 'hdcore=' not in f4m_url
:
3324 f4m_url
+= ('&' if '?' in f4m_url
else '?') + hdcore_sign
3325 f4m_formats
= self
._extract
_f
4m
_formats
(
3326 f4m_url
, video_id
, f4m_id
='hds', fatal
=False)
3327 for entry
in f4m_formats
:
3328 entry
.update({'extra_param_to_segment_url': hdcore_sign}
)
3329 formats
.extend(f4m_formats
)
3331 m3u8_url
= re
.sub(r
'(https?://[^/]+)/z/', r
'\1/i/', manifest_url
).replace('/manifest.f4m', '/master.m3u8')
3332 hls_host
= hosts
.get('hls')
3334 m3u8_url
= re
.sub(r
'(https?://)[^/]+', r
'\1' + hls_host
, m3u8_url
)
3335 m3u8_formats
, m3u8_subtitles
= self
._extract
_m
3u8_formats
_and
_subtitles
(
3336 m3u8_url
, video_id
, 'mp4', 'm3u8_native',
3337 m3u8_id
='hls', fatal
=False)
3338 formats
.extend(m3u8_formats
)
3339 subtitles
= self
._merge
_subtitles
(subtitles
, m3u8_subtitles
)
3341 http_host
= hosts
.get('http')
3342 if http_host
and m3u8_formats
and not signed
:
3343 REPL_REGEX
= r
'https?://[^/]+/i/([^,]+),([^/]+),([^/]+)\.csmil/.+'
3344 qualities
= re
.match(REPL_REGEX
, m3u8_url
).group(2).split(',')
3345 qualities_length
= len(qualities
)
3346 if len(m3u8_formats
) in (qualities_length
, qualities_length
+ 1):
3348 for f
in m3u8_formats
:
3349 if f
['vcodec'] != 'none':
3350 for protocol
in ('http', 'https'):
3352 del http_f
['manifest_url']
3354 REPL_REGEX
, protocol
+ fr
'://{http_host}/\g<1>{qualities[i]}\3', f
['url'])
3356 'format_id': http_f
['format_id'].replace('hls-', protocol
+ '-'),
3358 'protocol': protocol
,
3360 formats
.append(http_f
)
3363 return formats
, subtitles
3365 def _extract_wowza_formats(self
, url
, video_id
, m3u8_entry_protocol
='m3u8_native', skip_protocols
=[]):
3366 query
= urllib
.parse
.urlparse(url
).query
3367 url
= re
.sub(r
'/(?:manifest|playlist|jwplayer)\.(?:m3u8|f4m|mpd|smil)', '', url
)
3369 r
'(?:(?:http|rtmp|rtsp)(?P<s>s)?:)?(?P<url>//[^?]+)', url
)
3370 url_base
= mobj
.group('url')
3371 http_base_url
= '%s%s:%s' % ('http', mobj
.group('s') or '', url_base
)
3374 def manifest_url(manifest
):
3375 m_url
= f
'{http_base_url}/{manifest}'
3377 m_url
+= '?%s' % query
3380 if 'm3u8' not in skip_protocols
:
3381 formats
.extend(self
._extract
_m
3u8_formats
(
3382 manifest_url('playlist.m3u8'), video_id
, 'mp4',
3383 m3u8_entry_protocol
, m3u8_id
='hls', fatal
=False))
3384 if 'f4m' not in skip_protocols
:
3385 formats
.extend(self
._extract
_f
4m
_formats
(
3386 manifest_url('manifest.f4m'),
3387 video_id
, f4m_id
='hds', fatal
=False))
3388 if 'dash' not in skip_protocols
:
3389 formats
.extend(self
._extract
_mpd
_formats
(
3390 manifest_url('manifest.mpd'),
3391 video_id
, mpd_id
='dash', fatal
=False))
3392 if re
.search(r
'(?:/smil:|\.smil)', url_base
):
3393 if 'smil' not in skip_protocols
:
3394 rtmp_formats
= self
._extract
_smil
_formats
(
3395 manifest_url('jwplayer.smil'),
3396 video_id
, fatal
=False)
3397 for rtmp_format
in rtmp_formats
:
3398 rtsp_format
= rtmp_format
.copy()
3399 rtsp_format
['url'] = '%s/%s' % (rtmp_format
['url'], rtmp_format
['play_path'])
3400 del rtsp_format
['play_path']
3401 del rtsp_format
['ext']
3402 rtsp_format
.update({
3403 'url': rtsp_format
['url'].replace('rtmp://', 'rtsp://'),
3404 'format_id': rtmp_format
['format_id'].replace('rtmp', 'rtsp'),
3407 formats
.extend([rtmp_format
, rtsp_format
])
3409 for protocol
in ('rtmp', 'rtsp'):
3410 if protocol
not in skip_protocols
:
3412 'url': f
'{protocol}:{url_base}',
3413 'format_id': protocol
,
3414 'protocol': protocol
,
3418 def _find_jwplayer_data(self
, webpage
, video_id
=None, transform_source
=js_to_json
):
3420 r
'(?s)jwplayer\((?P<quote>[\'"])[^\'" ]+(?P
=quote
)\
)(?
!</script
>).*?\
.setup\s
*\
((?P
<options
>[^
)]+)\
)',
3424 jwplayer_data = self._parse_json(mobj.group('options
'),
3426 transform_source=transform_source)
3427 except ExtractorError:
3430 if isinstance(jwplayer_data, dict):
3431 return jwplayer_data
3433 def _extract_jwplayer_data(self, webpage, video_id, *args, **kwargs):
3434 jwplayer_data = self._find_jwplayer_data(
3435 webpage, video_id, transform_source=js_to_json)
3436 return self._parse_jwplayer_data(
3437 jwplayer_data, video_id, *args, **kwargs)
3439 def _parse_jwplayer_data(self, jwplayer_data, video_id=None, require_title=True,
3440 m3u8_id=None, mpd_id=None, rtmp_params=None, base_url=None):
3441 # JWPlayer backward compatibility: flattened playlists
3442 # https://github.com/jwplayer/jwplayer/blob/v7.4.3/src/js/api/config.js#L81-L96
3443 if 'playlist
' not in jwplayer_data:
3444 jwplayer_data = {'playlist': [jwplayer_data]}
3448 # JWPlayer backward compatibility: single playlist item
3449 # https://github.com/jwplayer/jwplayer/blob/v7.7.0/src/js/playlist/playlist.js#L10
3450 if not isinstance(jwplayer_data['playlist
'], list):
3451 jwplayer_data['playlist
'] = [jwplayer_data['playlist
']]
3453 for video_data in jwplayer_data['playlist
']:
3454 # JWPlayer backward compatibility: flattened sources
3455 # https://github.com/jwplayer/jwplayer/blob/v7.4.3/src/js/playlist/item.js#L29-L35
3456 if 'sources
' not in video_data:
3457 video_data['sources
'] = [video_data]
3459 this_video_id = video_id or video_data['mediaid
']
3461 formats = self._parse_jwplayer_formats(
3462 video_data['sources
'], video_id=this_video_id, m3u8_id=m3u8_id,
3463 mpd_id=mpd_id, rtmp_params=rtmp_params, base_url=base_url)
3466 tracks = video_data.get('tracks
')
3467 if tracks and isinstance(tracks, list):
3468 for track in tracks:
3469 if not isinstance(track, dict):
3471 track_kind = track.get('kind
')
3472 if not track_kind or not isinstance(track_kind, str):
3474 if track_kind.lower() not in ('captions
', 'subtitles
'):
3476 track_url = urljoin(base_url, track.get('file'))
3479 subtitles.setdefault(track.get('label
') or 'en
', []).append({
3480 'url
': self._proto_relative_url(track_url)
3484 'id': this_video_id,
3485 'title
': unescapeHTML(video_data['title
'] if require_title else video_data.get('title
')),
3486 'description
': clean_html(video_data.get('description
')),
3487 'thumbnail
': urljoin(base_url, self._proto_relative_url(video_data.get('image
'))),
3488 'timestamp
': int_or_none(video_data.get('pubdate
')),
3489 'duration
': float_or_none(jwplayer_data.get('duration
') or video_data.get('duration
')),
3490 'subtitles
': subtitles,
3492 # https://github.com/jwplayer/jwplayer/blob/master/src/js/utils/validator.js#L32
3493 if len(formats) == 1 and re.search(r'^
(?
:http|
//).*(?
:youtube\
.com|youtu\
.be
)/.+', formats[0]['url
']):
3495 '_type
': 'url_transparent
',
3496 'url
': formats[0]['url
'],
3499 self._sort_formats(formats)
3500 entry['formats
'] = formats
3501 entries.append(entry)
3502 if len(entries) == 1:
3505 return self.playlist_result(entries)
3507 def _parse_jwplayer_formats(self, jwplayer_sources_data, video_id=None,
3508 m3u8_id=None, mpd_id=None, rtmp_params=None, base_url=None):
3511 for source in jwplayer_sources_data:
3512 if not isinstance(source, dict):
3514 source_url = urljoin(
3515 base_url, self._proto_relative_url(source.get('file')))
3516 if not source_url or source_url in urls:
3518 urls.append(source_url)
3519 source_type = source.get('type') or ''
3520 ext = mimetype2ext(source_type) or determine_ext(source_url)
3521 if source_type == 'hls
' or ext == 'm3u8
':
3522 formats.extend(self._extract_m3u8_formats(
3523 source_url, video_id, 'mp4
', entry_protocol='m3u8_native
',
3524 m3u8_id=m3u8_id, fatal=False))
3525 elif source_type == 'dash
' or ext == 'mpd
':
3526 formats.extend(self._extract_mpd_formats(
3527 source_url, video_id, mpd_id=mpd_id, fatal=False))
3529 formats.extend(self._extract_smil_formats(
3530 source_url, video_id, fatal=False))
3531 # https://github.com/jwplayer/jwplayer/blob/master/src/js/providers/default.js#L67
3532 elif source_type.startswith('audio
') or ext in (
3533 'oga
', 'aac
', 'mp3
', 'mpeg
', 'vorbis
'):
3540 height = int_or_none(source.get('height
'))
3542 # Often no height is provided but there is a label in
3543 # format like "1080p", "720p SD", or 1080.
3544 height = int_or_none(self._search_regex(
3545 r'^
(\d{3,4}
)[pP
]?
(?
:\b|$
)', str(source.get('label
') or ''),
3546 'height
', default=None))
3549 'width
': int_or_none(source.get('width
')),
3551 'tbr
': int_or_none(source.get('bitrate
')),
3554 if source_url.startswith('rtmp
'):
3555 a_format['ext
'] = 'flv
'
3556 # See com/longtailvideo/jwplayer/media/RTMPMediaProvider.as
3557 # of jwplayer.flash.swf
3558 rtmp_url_parts = re.split(
3559 r'((?
:mp4|mp3|flv
):)', source_url, 1)
3560 if len(rtmp_url_parts) == 3:
3561 rtmp_url, prefix, play_path = rtmp_url_parts
3564 'play_path
': prefix + play_path,
3567 a_format.update(rtmp_params)
3568 formats.append(a_format)
3571 def _live_title(self, name):
3572 self._downloader.deprecation_warning('yt_dlp
.InfoExtractor
._live
_title
is deprecated
and does
not work
as expected
')
3575 def _int(self, v, name, fatal=False, **kwargs):
3576 res = int_or_none(v, **kwargs)
3578 msg = f'Failed to extract {name}
: Could
not parse value {v!r}
'
3580 raise ExtractorError(msg)
3582 self.report_warning(msg)
3585 def _float(self, v, name, fatal=False, **kwargs):
3586 res = float_or_none(v, **kwargs)
3588 msg = f'Failed to extract {name}
: Could
not parse value {v!r}
'
3590 raise ExtractorError(msg)
3592 self.report_warning(msg)
3595 def _set_cookie(self, domain, name, value, expire_time=None, port=None,
3596 path='/', secure=False, discard=False, rest={}, **kwargs):
3597 cookie = http.cookiejar.Cookie(
3598 0, name, value, port, port is not None, domain, True,
3599 domain.startswith('.'), path, True, secure, expire_time,
3600 discard, None, None, rest)
3601 self.cookiejar.set_cookie(cookie)
3603 def _get_cookies(self, url):
3604 """ Return a http.cookies.SimpleCookie with the cookies for the url """
3605 return http.cookies.SimpleCookie(self._downloader._calc_cookies(url))
3607 def _apply_first_set_cookie_header(self, url_handle, cookie):
3609 Apply first Set-Cookie header instead of the last. Experimental.
3611 Some sites (e.g. [1-3]) may serve two cookies under the same name
3612 in Set-Cookie header and expect the first (old) one to be set rather
3613 than second (new). However, as of RFC6265 the newer one cookie
3614 should be set into cookie store what actually happens.
3615 We will workaround this issue by resetting the cookie to
3616 the first one manually.
3617 1. https://new.vk.com/
3618 2. https://github.com/ytdl-org/youtube-dl/issues/9841#issuecomment-227871201
3619 3. https://learning.oreilly.com/
3621 for header, cookies in url_handle.headers.items():
3622 if header.lower() != 'set-cookie
':
3624 cookies = cookies.encode('iso
-8859-1').decode('utf
-8')
3625 cookie_value = re.search(
3626 r'%s=(.+?
);.*?
\b[Dd
]omain
=(.+?
)(?
:[,;]|$
)' % cookie, cookies)
3628 value, domain = cookie_value.groups()
3629 self._set_cookie(domain, cookie, value)
3633 def get_testcases(cls, include_onlymatching=False):
3634 t = getattr(cls, '_TEST
', None)
3636 assert not hasattr(cls, '_TESTS
'), f'{cls.ie_key()}IE has _TEST
and _TESTS
'
3639 tests = getattr(cls, '_TESTS
', [])
3641 if not include_onlymatching and t.get('only_matching
', False):
3643 t['name
'] = cls.ie_key()
3648 """Get age limit from the testcases"""
3649 return max(traverse_obj(
3650 tuple(cls.get_testcases(include_onlymatching=False)),
3651 (..., (('playlist
', 0), None), 'info_dict
', 'age_limit
')) or [0])
3654 def is_suitable(cls, age_limit):
3655 """Test whether the extractor is generally suitable for the given age limit"""
3656 return not age_restricted(cls.age_limit, age_limit)
3659 def description(cls, *, markdown=True, search_examples=None):
3660 """Description of the extractor"""
3662 if cls._NETRC_MACHINE:
3664 desc += f' [<abbr title
="netrc machine"><em
>{cls._NETRC_MACHINE}
</em
></abbr
>]'
3666 desc += f' [{cls._NETRC_MACHINE}
]'
3667 if cls.IE_DESC is False:
3670 desc += f' {cls.IE_DESC}
'
3672 desc += f'; "{cls.SEARCH_KEY}:" prefix
'
3674 _COUNTS = ('', '5', '10', 'all
')
3675 desc += f' (Example
: "{cls.SEARCH_KEY}{random.choice(_COUNTS)}:{random.choice(search_examples)}")'
3676 if not cls.working():
3677 desc += ' (**Currently broken
**)' if markdown else ' (Currently broken
)'
3679 name = f' - **{cls.IE_NAME}
**' if markdown else cls.IE_NAME
3680 return f'{name}
:{desc}
' if desc else name
3682 def extract_subtitles(self, *args, **kwargs):
3683 if (self.get_param('writesubtitles
', False)
3684 or self.get_param('listsubtitles
')):
3685 return self._get_subtitles(*args, **kwargs)
3688 def _get_subtitles(self, *args, **kwargs):
3689 raise NotImplementedError('This method must be implemented by subclasses
')
3691 def extract_comments(self, *args, **kwargs):
3692 if not self.get_param('getcomments
'):
3694 generator = self._get_comments(*args, **kwargs)
3701 comments.append(next(generator))
3702 except StopIteration:
3704 except KeyboardInterrupt:
3705 self.to_screen('Interrupted by user
')
3706 except Exception as e:
3707 if self.get_param('ignoreerrors
') is not True:
3709 self._downloader.report_error(e)
3710 comment_count = len(comments)
3711 self.to_screen(f'Extracted {comment_count} comments
')
3713 'comments
': comments,
3714 'comment_count
': None if interrupted else comment_count
3718 def _get_comments(self, *args, **kwargs):
3719 raise NotImplementedError('This method must be implemented by subclasses
')
3722 def _merge_subtitle_items(subtitle_list1, subtitle_list2):
3723 """ Merge subtitle items for one language. Items with duplicated URLs/data
3724 will be dropped. """
3725 list1_data = {(item.get('url'), item.get('data')) for item in subtitle_list1}
3726 ret = list(subtitle_list1)
3727 ret.extend(item for item in subtitle_list2 if (item.get('url
'), item.get('data
')) not in list1_data)
3731 def _merge_subtitles(cls, *dicts, target=None):
3732 """ Merge subtitle dictionaries, language by language. """
3736 for lang, subs in d.items():
3737 target[lang] = cls._merge_subtitle_items(target.get(lang, []), subs)
3740 def extract_automatic_captions(self, *args, **kwargs):
3741 if (self.get_param('writeautomaticsub
', False)
3742 or self.get_param('listsubtitles
')):
3743 return self._get_automatic_captions(*args, **kwargs)
3746 def _get_automatic_captions(self, *args, **kwargs):
3747 raise NotImplementedError('This method must be implemented by subclasses
')
3749 @functools.cached_property
3750 def _cookies_passed(self):
3751 """Whether cookies have been passed to YoutubeDL"""
3752 return self.get_param('cookiefile
') is not None or self.get_param('cookiesfrombrowser
') is not None
3754 def mark_watched(self, *args, **kwargs):
3755 if not self.get_param('mark_watched
', False):
3757 if self.supports_login() and self._get_login_info()[0] is not None or self._cookies_passed:
3758 self._mark_watched(*args, **kwargs)
3760 def _mark_watched(self, *args, **kwargs):
3761 raise NotImplementedError('This method must be implemented by subclasses
')
3763 def geo_verification_headers(self):
3765 geo_verification_proxy = self.get_param('geo_verification_proxy
')
3766 if geo_verification_proxy:
3767 headers['Ytdl
-request
-proxy
'] = geo_verification_proxy
3770 def _generic_id(self, url):
3771 return urllib.parse.unquote(os.path.splitext(url.rstrip('/').split('/')[-1])[0])
3773 def _generic_title(self, url):
3774 return urllib.parse.unquote(os.path.splitext(url_basename(url))[0])
3777 def _availability(is_private=None, needs_premium=None, needs_subscription=None, needs_auth=None, is_unlisted=None):
3778 all_known = all(map(
3779 lambda x: x is not None,
3780 (is_private, needs_premium, needs_subscription, needs_auth, is_unlisted)))
3782 'private
' if is_private
3783 else 'premium_only
' if needs_premium
3784 else 'subscriber_only
' if needs_subscription
3785 else 'needs_auth
' if needs_auth
3786 else 'unlisted
' if is_unlisted
3787 else 'public
' if all_known
3790 def _configuration_arg(self, key, default=NO_DEFAULT, *, ie_key=None, casesense=False):
3792 @returns A list of values for the extractor argument given by "key"
3793 or "default" if no such key is present
3794 @param default The default value to return when the key is not present (default: [])
3795 @param casesense When false, the values are converted to lower case
3798 self._downloader.params, ('extractor_args
', (ie_key or self.ie_key()).lower(), key))
3800 return [] if default is NO_DEFAULT else default
3801 return list(val) if casesense else [x.lower() for x in val]
3803 def _yes_playlist(self, playlist_id, video_id, smuggled_data=None, *, playlist_label='playlist
', video_label='video
'):
3804 if not playlist_id or not video_id:
3807 no_playlist = (smuggled_data or {}).get('force_noplaylist
')
3808 if no_playlist is not None:
3809 return not no_playlist
3811 video_id = '' if video_id is True else f' {video_id}
'
3812 playlist_id = '' if playlist_id is True else f' {playlist_id}
'
3813 if self.get_param('noplaylist
'):
3814 self.to_screen(f'Downloading just the {video_label}{video_id} because of
--no
-playlist
')
3816 self.to_screen(f'Downloading {playlist_label}{playlist_id}
- add
--no
-playlist to download just the {video_label}{video_id}
')
3820 class SearchInfoExtractor(InfoExtractor):
3822 Base class for paged search queries extractors.
3823 They accept URLs in the format _SEARCH_KEY(|all|[0-9]):{query}
3824 Instances should define _SEARCH_KEY and optionally _MAX_RESULTS
3827 _MAX_RESULTS = float('inf
')
3830 def _make_valid_url(cls):
3831 return r'%s(?P
<prefix
>|
[1-9][0-9]*|all
):(?P
<query
>[\s\S
]+)' % cls._SEARCH_KEY
3833 def _real_extract(self, query):
3834 prefix, query = self._match_valid_url(query).group('prefix
', 'query
')
3836 return self._get_n_results(query, 1)
3837 elif prefix == 'all
':
3838 return self._get_n_results(query, self._MAX_RESULTS)
3842 raise ExtractorError(f'invalid download number {n}
for query
"{query}"')
3843 elif n > self._MAX_RESULTS:
3844 self.report_warning('%s returns
max %i results (you requested
%i)' % (self._SEARCH_KEY, self._MAX_RESULTS, n))
3845 n = self._MAX_RESULTS
3846 return self._get_n_results(query, n)
3848 def _get_n_results(self, query, n):
3849 """Get a specified number of results for a query.
3850 Either this function or _search_results must be overridden by subclasses """
3851 return self.playlist_result(
3852 itertools.islice(self._search_results(query), 0, None if n == float('inf
') else n),
3855 def _search_results(self, query):
3856 """Returns an iterator of search results"""
3857 raise NotImplementedError('This method must be implemented by subclasses
')
3860 def SEARCH_KEY(cls):
3861 return cls._SEARCH_KEY