18 import xml
.etree
.ElementTree
20 from ..compat
import functools
, re
# isort: split
21 from ..compat
import compat_etree_fromstring
, compat_expanduser
, compat_os_name
22 from ..downloader
import FileDownloader
23 from ..downloader
.f4m
import get_base_url
, remove_encrypted_media
58 parse_m3u8_attributes
,
83 """Information Extractor class.
85 Information extractors are the classes that, given a URL, extract
86 information about the video (or videos) the URL refers to. This
87 information includes the real video URL, the video title, author and
88 others. The information is stored in a dictionary which is then
89 passed to the YoutubeDL. The YoutubeDL processes this
90 information possibly downloading the video to the file system, among
91 other possible outcomes.
93 The type field determines the type of the result.
94 By far the most common value (and the default if _type is missing) is
95 "video", which indicates a single video.
97 For a video, the dictionaries must include the following fields:
100 title: Video title, unescaped. Set to an empty string if video has
101 no title as opposed to "None" which signifies that the
102 extractor failed to obtain a title
104 Additionally, it must contain either a formats entry or a url one:
106 formats: A list of dictionaries for each format available, ordered
107 from worst to best quality.
110 * url The mandatory URL representing the media:
111 for plain file media - HTTP URL of this file,
113 for HLS - URL of the M3U8 media playlist,
114 for HDS - URL of the F4M manifest,
116 - HTTP URL to plain file media (in case of
118 - URL of the MPD manifest or base URL
119 representing the media if MPD manifest
120 is parsed from a string (in case of
122 for MSS - URL of the ISM manifest.
124 The URL of the manifest file in case of
126 for HLS - URL of the M3U8 master playlist,
127 for HDS - URL of the F4M manifest,
128 for DASH - URL of the MPD manifest,
129 for MSS - URL of the ISM manifest.
130 * manifest_stream_number (For internal use only)
131 The index of the stream in the manifest file
132 * ext Will be calculated from URL if missing
133 * format A human-readable description of the format
134 ("mp4 container with h264/opus").
135 Calculated from the format_id, width, height.
136 and format_note fields if missing.
137 * format_id A short description of the format
138 ("mp4_h264_opus" or "19").
139 Technically optional, but strongly recommended.
140 * format_note Additional info about the format
141 ("3D" or "DASH video")
142 * width Width of the video, if known
143 * height Height of the video, if known
144 * resolution Textual description of width and height
145 * dynamic_range The dynamic range of the video. One of:
146 "SDR" (None), "HDR10", "HDR10+, "HDR12", "HLG, "DV"
147 * tbr Average bitrate of audio and video in KBit/s
148 * abr Average audio bitrate in KBit/s
149 * acodec Name of the audio codec in use
150 * asr Audio sampling rate in Hertz
151 * vbr Average video bitrate in KBit/s
153 * vcodec Name of the video codec in use
154 * container Name of the container format
155 * filesize The number of bytes, if known in advance
156 * filesize_approx An estimate for the number of bytes
157 * player_url SWF Player URL (used for rtmpdump).
158 * protocol The protocol that will be used for the actual
159 download, lower-case. One of "http", "https" or
160 one of the protocols defined in downloader.PROTOCOL_MAP
162 Base URL for fragments. Each fragment's path
163 value (if present) will be relative to
165 * fragments A list of fragments of a fragmented media.
166 Each fragment entry must contain either an url
167 or a path. If an url is present it should be
168 considered by a client. Otherwise both path and
169 fragment_base_url must be present. Here is
170 the list of all potential fields:
171 * "url" - fragment's URL
172 * "path" - fragment's path relative to
174 * "duration" (optional, int or float)
175 * "filesize" (optional, int)
176 * is_from_start Is a live format that can be downloaded
177 from the start. Boolean
178 * preference Order number of this format. If this field is
179 present and not None, the formats get sorted
180 by this field, regardless of all other values.
181 -1 for default (order by other properties),
182 -2 or smaller for less than default.
183 < -1000 to hide the format (if there is
184 another one which is strictly better)
185 * language Language code, e.g. "de" or "en-US".
186 * language_preference Is this in the language mentioned in
188 10 if it's what the URL is about,
189 -1 for default (don't know),
190 -10 otherwise, other values reserved for now.
191 * quality Order number of the video quality of this
192 format, irrespective of the file format.
193 -1 for default (order by other properties),
194 -2 or smaller for less than default.
195 * source_preference Order number for this video source
196 (quality takes higher priority)
197 -1 for default (order by other properties),
198 -2 or smaller for less than default.
199 * http_headers A dictionary of additional HTTP headers
200 to add to the request.
201 * stretched_ratio If given and not 1, indicates that the
202 video's pixels are not square.
203 width : height ratio as float.
204 * no_resume The server does not support resuming the
205 (HTTP or RTMP) download. Boolean.
206 * has_drm The format has DRM and cannot be downloaded. Boolean
207 * downloader_options A dictionary of downloader options
208 (For internal use only)
209 * http_chunk_size Chunk size for HTTP downloads
210 * ffmpeg_args Extra arguments for ffmpeg downloader
211 RTMP formats can also have the additional fields: page_url,
212 app, play_path, tc_url, flash_version, rtmp_live, rtmp_conn,
213 rtmp_protocol, rtmp_real_time
215 url: Final video URL.
216 ext: Video filename extension.
217 format: The video format, defaults to ext (used for --get-format)
218 player_url: SWF Player URL (used for rtmpdump).
220 The following fields are optional:
222 direct: True if a direct video file was given (must only be set by GenericIE)
223 alt_title: A secondary title of the video.
224 display_id An alternative identifier for the video, not necessarily
225 unique, but available before title. Typically, id is
226 something like "4234987", title "Dancing naked mole rats",
227 and display_id "dancing-naked-mole-rats"
228 thumbnails: A list of dictionaries, with the following entries:
229 * "id" (optional, string) - Thumbnail format ID
231 * "preference" (optional, int) - quality of the image
232 * "width" (optional, int)
233 * "height" (optional, int)
234 * "resolution" (optional, string "{width}x{height}",
236 * "filesize" (optional, int)
237 * "http_headers" (dict) - HTTP headers for the request
238 thumbnail: Full URL to a video thumbnail image.
239 description: Full video description.
240 uploader: Full name of the video uploader.
241 license: License name the video is licensed under.
242 creator: The creator of the video.
243 timestamp: UNIX timestamp of the moment the video was uploaded
244 upload_date: Video upload date in UTC (YYYYMMDD).
245 If not explicitly set, calculated from timestamp
246 release_timestamp: UNIX timestamp of the moment the video was released.
247 If it is not clear whether to use timestamp or this, use the former
248 release_date: The date (YYYYMMDD) when the video was released in UTC.
249 If not explicitly set, calculated from release_timestamp
250 modified_timestamp: UNIX timestamp of the moment the video was last modified.
251 modified_date: The date (YYYYMMDD) when the video was last modified in UTC.
252 If not explicitly set, calculated from modified_timestamp
253 uploader_id: Nickname or id of the video uploader.
254 uploader_url: Full URL to a personal webpage of the video uploader.
255 channel: Full name of the channel the video is uploaded on.
256 Note that channel fields may or may not repeat uploader
257 fields. This depends on a particular extractor.
258 channel_id: Id of the channel.
259 channel_url: Full URL to a channel webpage.
260 channel_follower_count: Number of followers of the channel.
261 location: Physical location where the video was filmed.
262 subtitles: The available subtitles as a dictionary in the format
263 {tag: subformats}. "tag" is usually a language code, and
264 "subformats" is a list sorted from lower to higher
265 preference, each element is a dictionary with the "ext"
267 * "data": The subtitles file contents
268 * "url": A URL pointing to the subtitles file
269 It can optionally also have:
270 * "name": Name or description of the subtitles
271 * "http_headers": A dictionary of additional HTTP headers
272 to add to the request.
273 "ext" will be calculated from URL if missing
274 automatic_captions: Like 'subtitles'; contains automatically generated
275 captions instead of normal subtitles
276 duration: Length of the video in seconds, as an integer or float.
277 view_count: How many users have watched the video on the platform.
278 like_count: Number of positive ratings of the video
279 dislike_count: Number of negative ratings of the video
280 repost_count: Number of reposts of the video
281 average_rating: Average rating give by users, the scale used depends on the webpage
282 comment_count: Number of comments on the video
283 comments: A list of comments, each with one or more of the following
284 properties (all but one of text or html optional):
285 * "author" - human-readable name of the comment author
286 * "author_id" - user ID of the comment author
287 * "author_thumbnail" - The thumbnail of the comment author
289 * "html" - Comment as HTML
290 * "text" - Plain text of the comment
291 * "timestamp" - UNIX timestamp of comment
292 * "parent" - ID of the comment this one is replying to.
293 Set to "root" to indicate that this is a
294 comment to the original video.
295 * "like_count" - Number of positive ratings of the comment
296 * "dislike_count" - Number of negative ratings of the comment
297 * "is_favorited" - Whether the comment is marked as
298 favorite by the video uploader
299 * "author_is_uploader" - Whether the comment is made by
301 age_limit: Age restriction for the video, as an integer (years)
302 webpage_url: The URL to the video webpage, if given to yt-dlp it
303 should allow to get the same result again. (It will be set
304 by YoutubeDL if it's missing)
305 categories: A list of categories that the video falls in, for example
307 tags: A list of tags assigned to the video, e.g. ["sweden", "pop music"]
308 cast: A list of the video cast
309 is_live: True, False, or None (=unknown). Whether this video is a
310 live stream that goes on instead of a fixed-length video.
311 was_live: True, False, or None (=unknown). Whether this video was
312 originally a live stream.
313 live_status: 'is_live', 'is_upcoming', 'was_live', 'not_live' or None (=unknown)
314 If absent, automatically set from is_live, was_live
315 start_time: Time in seconds where the reproduction should start, as
316 specified in the URL.
317 end_time: Time in seconds where the reproduction should end, as
318 specified in the URL.
319 chapters: A list of dictionaries, with the following entries:
320 * "start_time" - The start time of the chapter in seconds
321 * "end_time" - The end time of the chapter in seconds
322 * "title" (optional, string)
323 playable_in_embed: Whether this video is allowed to play in embedded
324 players on other sites. Can be True (=always allowed),
325 False (=never allowed), None (=unknown), or a string
326 specifying the criteria for embedability (Eg: 'whitelist')
327 availability: Under what condition the video is available. One of
328 'private', 'premium_only', 'subscriber_only', 'needs_auth',
329 'unlisted' or 'public'. Use 'InfoExtractor._availability'
331 __post_extractor: A function to be called just before the metadata is
332 written to either disk, logger or console. The function
333 must return a dict which will be added to the info_dict.
334 This is usefull for additional information that is
335 time-consuming to extract. Note that the fields thus
336 extracted will not be available to output template and
337 match_filter. So, only "comments" and "comment_count" are
338 currently allowed to be extracted via this method.
340 The following fields should only be used when the video belongs to some logical
343 chapter: Name or title of the chapter the video belongs to.
344 chapter_number: Number of the chapter the video belongs to, as an integer.
345 chapter_id: Id of the chapter the video belongs to, as a unicode string.
347 The following fields should only be used when the video is an episode of some
348 series, programme or podcast:
350 series: Title of the series or programme the video episode belongs to.
351 series_id: Id of the series or programme the video episode belongs to, as a unicode string.
352 season: Title of the season the video episode belongs to.
353 season_number: Number of the season the video episode belongs to, as an integer.
354 season_id: Id of the season the video episode belongs to, as a unicode string.
355 episode: Title of the video episode. Unlike mandatory video title field,
356 this field should denote the exact title of the video episode
357 without any kind of decoration.
358 episode_number: Number of the video episode within a season, as an integer.
359 episode_id: Id of the video episode, as a unicode string.
361 The following fields should only be used when the media is a track or a part of
364 track: Title of the track.
365 track_number: Number of the track within an album or a disc, as an integer.
366 track_id: Id of the track (useful in case of custom indexing, e.g. 6.iii),
368 artist: Artist(s) of the track.
369 genre: Genre(s) of the track.
370 album: Title of the album the track belongs to.
371 album_type: Type of the album (e.g. "Demo", "Full-length", "Split", "Compilation", etc).
372 album_artist: List of all artists appeared on the album (e.g.
373 "Ash Borer / Fell Voices" or "Various Artists", useful for splits
375 disc_number: Number of the disc or other physical medium the track belongs to,
377 release_year: Year (YYYY) when the album was released.
378 composer: Composer of the piece
380 The following fields should only be set for clips that should be cut from the original video:
382 section_start: Start time of the section in seconds
383 section_end: End time of the section in seconds
385 Unless mentioned otherwise, the fields should be Unicode strings.
387 Unless mentioned otherwise, None is equivalent to absence of information.
390 _type "playlist" indicates multiple videos.
391 There must be a key "entries", which is a list, an iterable, or a PagedList
392 object, each element of which is a valid dictionary by this specification.
394 Additionally, playlists can have "id", "title", and any other relevent
395 attributes with the same semantics as videos (see above).
397 It can also have the following optional fields:
399 playlist_count: The total number of videos in a playlist. If not given,
400 YoutubeDL tries to calculate it from "entries"
403 _type "multi_video" indicates that there are multiple videos that
404 form a single show, for examples multiple acts of an opera or TV episode.
405 It must have an entries key like a playlist and contain all the keys
406 required for a video at the same time.
409 _type "url" indicates that the video must be extracted from another
410 location, possibly by a different extractor. Its only required key is:
411 "url" - the next URL to extract.
412 The key "ie_key" can be set to the class name (minus the trailing "IE",
413 e.g. "Youtube") if the extractor class is known in advance.
414 Additionally, the dictionary may have any properties of the resolved entity
415 known in advance, for example "title" if the title of the referred video is
419 _type "url_transparent" entities have the same specification as "url", but
420 indicate that the given additional information is more precise than the one
421 associated with the resolved URL.
422 This is useful when a site employs a video service that hosts the video and
423 its technical metadata, but that video service does not embed a useful
424 title, description etc.
427 Subclasses of this should define a _VALID_URL regexp and, re-define the
428 _real_extract() and (optionally) _real_initialize() methods.
429 Probably, they should also be added to the list of extractors.
431 Subclasses may also override suitable() if necessary, but ensure the function
432 signature is preserved and that this function imports everything it needs
433 (except other extractors), so that lazy_extractors works correctly.
435 To support username + password (or netrc) login, the extractor must define a
436 _NETRC_MACHINE and re-define _perform_login(username, password) and
437 (optionally) _initialize_pre_login() methods. The _perform_login method will
438 be called between _initialize_pre_login and _real_initialize if credentials
439 are passed by the user. In cases where it is necessary to have the login
440 process as part of the extraction rather than initialization, _perform_login
441 can be left undefined.
443 _GEO_BYPASS attribute may be set to False in order to disable
444 geo restriction bypass mechanisms for a particular extractor.
445 Though it won't disable explicit geo restriction bypass based on
446 country code provided with geo_bypass_country.
448 _GEO_COUNTRIES attribute may contain a list of presumably geo unrestricted
449 countries for this extractor. One of these countries will be used by
450 geo restriction bypass mechanism right away in order to bypass
451 geo restriction, of course, if the mechanism is not disabled.
453 _GEO_IP_BLOCKS attribute may contain a list of presumably geo unrestricted
454 IP blocks in CIDR notation for this extractor. One of these IP blocks
455 will be used by geo restriction bypass mechanism similarly
458 The _WORKING attribute should be set to False for broken IEs
459 in order to warn the users and skip the tests.
464 _x_forwarded_for_ip
= None
466 _GEO_COUNTRIES
= None
467 _GEO_IP_BLOCKS
= None
469 _NETRC_MACHINE
= None
473 def _login_hint(self
, method
=NO_DEFAULT
, netrc
=None):
474 password_hint
= f
'--username and --password, or --netrc ({netrc or self._NETRC_MACHINE}) to provide account credentials'
477 'any': f
'Use --cookies, --cookies-from-browser, {password_hint}',
478 'password': f
'Use {password_hint}',
480 'Use --cookies-from-browser or --cookies for the authentication. '
481 'See https://github.com/ytdl-org/youtube-dl#how-do-i-pass-cookies-to-youtube-dl for how to manually pass cookies'),
482 }[method
if method
is not NO_DEFAULT
else 'any' if self
.supports_login() else 'cookies']
484 def __init__(self
, downloader
=None):
485 """Constructor. Receives an optional downloader (a YoutubeDL instance).
486 If a downloader is not passed during initialization,
487 it must be set using "set_downloader()" before "extract()" is called"""
489 self
._x
_forwarded
_for
_ip
= None
490 self
._printed
_messages
= set()
491 self
.set_downloader(downloader
)
494 def _match_valid_url(cls
, url
):
495 # This does not use has/getattr intentionally - we want to know whether
496 # we have cached the regexp for *this* class, whereas getattr would also
497 # match the superclass
498 if '_VALID_URL_RE' not in cls
.__dict
__:
499 if '_VALID_URL' not in cls
.__dict
__:
500 cls
._VALID
_URL
= cls
._make
_valid
_url
()
501 cls
._VALID
_URL
_RE
= re
.compile(cls
._VALID
_URL
)
502 return cls
._VALID
_URL
_RE
.match(url
)
505 def suitable(cls
, url
):
506 """Receives a URL and returns True if suitable for this IE."""
507 # This function must import everything it needs (except other extractors),
508 # so that lazy_extractors works correctly
509 return cls
._match
_valid
_url
(url
) is not None
512 def _match_id(cls
, url
):
513 return cls
._match
_valid
_url
(url
).group('id')
516 def get_temp_id(cls
, url
):
518 return cls
._match
_id
(url
)
519 except (IndexError, AttributeError):
524 """Getter method for _WORKING."""
528 def supports_login(cls
):
529 return bool(cls
._NETRC
_MACHINE
)
531 def initialize(self
):
532 """Initializes an instance (authentication, etc)."""
533 self
._printed
_messages
= set()
534 self
._initialize
_geo
_bypass
({
535 'countries': self
._GEO
_COUNTRIES
,
536 'ip_blocks': self
._GEO
_IP
_BLOCKS
,
539 self
._initialize
_pre
_login
()
540 if self
.supports_login():
541 username
, password
= self
._get
_login
_info
()
543 self
._perform
_login
(username
, password
)
544 elif self
.get_param('username') and False not in (self
.IE_DESC
, self
._NETRC
_MACHINE
):
545 self
.report_warning(f
'Login with password is not supported for this website. {self._login_hint("cookies")}')
546 self
._real
_initialize
()
549 def _initialize_geo_bypass(self
, geo_bypass_context
):
551 Initialize geo restriction bypass mechanism.
553 This method is used to initialize geo bypass mechanism based on faking
554 X-Forwarded-For HTTP header. A random country from provided country list
555 is selected and a random IP belonging to this country is generated. This
556 IP will be passed as X-Forwarded-For HTTP header in all subsequent
559 This method will be used for initial geo bypass mechanism initialization
560 during the instance initialization with _GEO_COUNTRIES and
563 You may also manually call it from extractor's code if geo bypass
564 information is not available beforehand (e.g. obtained during
565 extraction) or due to some other reason. In this case you should pass
566 this information in geo bypass context passed as first argument. It may
567 contain following fields:
569 countries: List of geo unrestricted countries (similar
571 ip_blocks: List of geo unrestricted IP blocks in CIDR notation
572 (similar to _GEO_IP_BLOCKS)
575 if not self
._x
_forwarded
_for
_ip
:
577 # Geo bypass mechanism is explicitly disabled by user
578 if not self
.get_param('geo_bypass', True):
581 if not geo_bypass_context
:
582 geo_bypass_context
= {}
584 # Backward compatibility: previously _initialize_geo_bypass
585 # expected a list of countries, some 3rd party code may still use
587 if isinstance(geo_bypass_context
, (list, tuple)):
588 geo_bypass_context
= {
589 'countries': geo_bypass_context
,
592 # The whole point of geo bypass mechanism is to fake IP
593 # as X-Forwarded-For HTTP header based on some IP block or
596 # Path 1: bypassing based on IP block in CIDR notation
598 # Explicit IP block specified by user, use it right away
599 # regardless of whether extractor is geo bypassable or not
600 ip_block
= self
.get_param('geo_bypass_ip_block', None)
602 # Otherwise use random IP block from geo bypass context but only
603 # if extractor is known as geo bypassable
605 ip_blocks
= geo_bypass_context
.get('ip_blocks')
606 if self
._GEO
_BYPASS
and ip_blocks
:
607 ip_block
= random
.choice(ip_blocks
)
610 self
._x
_forwarded
_for
_ip
= GeoUtils
.random_ipv4(ip_block
)
611 self
.write_debug(f
'Using fake IP {self._x_forwarded_for_ip} as X-Forwarded-For')
614 # Path 2: bypassing based on country code
616 # Explicit country code specified by user, use it right away
617 # regardless of whether extractor is geo bypassable or not
618 country
= self
.get_param('geo_bypass_country', None)
620 # Otherwise use random country code from geo bypass context but
621 # only if extractor is known as geo bypassable
623 countries
= geo_bypass_context
.get('countries')
624 if self
._GEO
_BYPASS
and countries
:
625 country
= random
.choice(countries
)
628 self
._x
_forwarded
_for
_ip
= GeoUtils
.random_ipv4(country
)
629 self
._downloader
.write_debug(
630 f
'Using fake IP {self._x_forwarded_for_ip} ({country.upper()}) as X-Forwarded-For')
632 def extract(self
, url
):
633 """Extracts URL information and returns it in list of dicts."""
638 self
.write_debug('Extracting URL: %s' % url
)
639 ie_result
= self
._real
_extract
(url
)
640 if ie_result
is None:
642 if self
._x
_forwarded
_for
_ip
:
643 ie_result
['__x_forwarded_for_ip'] = self
._x
_forwarded
_for
_ip
644 subtitles
= ie_result
.get('subtitles')
645 if (subtitles
and 'live_chat' in subtitles
646 and 'no-live-chat' in self
.get_param('compat_opts', [])):
647 del subtitles
['live_chat']
649 except GeoRestrictedError
as e
:
650 if self
.__maybe
_fake
_ip
_and
_retry
(e
.countries
):
653 except UnsupportedError
:
655 except ExtractorError
as e
:
657 'video_id': e
.video_id
or self
.get_temp_id(url
),
659 'tb': e
.traceback
or sys
.exc_info()[2],
660 'expected': e
.expected
,
663 if hasattr(e
, 'countries'):
664 kwargs
['countries'] = e
.countries
665 raise type(e
)(e
.orig_msg
, **kwargs
)
666 except http
.client
.IncompleteRead
as e
:
667 raise ExtractorError('A network error has occurred.', cause
=e
, expected
=True, video_id
=self
.get_temp_id(url
))
668 except (KeyError, StopIteration) as e
:
669 raise ExtractorError('An extractor error has occurred.', cause
=e
, video_id
=self
.get_temp_id(url
))
671 def __maybe_fake_ip_and_retry(self
, countries
):
672 if (not self
.get_param('geo_bypass_country', None)
674 and self
.get_param('geo_bypass', True)
675 and not self
._x
_forwarded
_for
_ip
677 country_code
= random
.choice(countries
)
678 self
._x
_forwarded
_for
_ip
= GeoUtils
.random_ipv4(country_code
)
679 if self
._x
_forwarded
_for
_ip
:
681 'Video is geo restricted. Retrying extraction with fake IP %s (%s) as X-Forwarded-For.'
682 % (self
._x
_forwarded
_for
_ip
, country_code
.upper()))
686 def set_downloader(self
, downloader
):
687 """Sets a YoutubeDL instance as the downloader for this IE."""
688 self
._downloader
= downloader
692 return self
._downloader
.cache
696 return self
._downloader
.cookiejar
698 def _initialize_pre_login(self
):
699 """ Intialization before login. Redefine in subclasses."""
702 def _perform_login(self
, username
, password
):
703 """ Login with username and password. Redefine in subclasses."""
706 def _real_initialize(self
):
707 """Real initialization process. Redefine in subclasses."""
710 def _real_extract(self
, url
):
711 """Real extraction process. Redefine in subclasses."""
712 raise NotImplementedError('This method must be implemented by subclasses')
716 """A string for getting the InfoExtractor with get_info_extractor"""
717 return cls
.__name
__[:-2]
721 return cls
.__name
__[:-2]
724 def __can_accept_status_code(err
, expected_status
):
725 assert isinstance(err
, urllib
.error
.HTTPError
)
726 if expected_status
is None:
728 elif callable(expected_status
):
729 return expected_status(err
.code
) is True
731 return err
.code
in variadic(expected_status
)
733 def _create_request(self
, url_or_request
, data
=None, headers
={}, query={}
):
734 if isinstance(url_or_request
, urllib
.request
.Request
):
735 return update_Request(url_or_request
, data
=data
, headers
=headers
, query
=query
)
737 url_or_request
= update_url_query(url_or_request
, query
)
738 return sanitized_Request(url_or_request
, data
, headers
)
740 def _request_webpage(self
, url_or_request
, video_id
, note
=None, errnote
=None, fatal
=True, data
=None, headers
={}, query={}
, expected_status
=None):
742 Return the response handle.
744 See _download_webpage docstring for arguments specification.
746 if not self
._downloader
._first
_webpage
_request
:
747 sleep_interval
= self
.get_param('sleep_interval_requests') or 0
748 if sleep_interval
> 0:
749 self
.to_screen('Sleeping %s seconds ...' % sleep_interval
)
750 time
.sleep(sleep_interval
)
752 self
._downloader
._first
_webpage
_request
= False
755 self
.report_download_webpage(video_id
)
756 elif note
is not False:
758 self
.to_screen(str(note
))
760 self
.to_screen(f
'{video_id}: {note}')
762 # Some sites check X-Forwarded-For HTTP header in order to figure out
763 # the origin of the client behind proxy. This allows bypassing geo
764 # restriction by faking this header's value to IP that belongs to some
765 # geo unrestricted country. We will do so once we encounter any
766 # geo restriction error.
767 if self
._x
_forwarded
_for
_ip
:
768 if 'X-Forwarded-For' not in headers
:
769 headers
['X-Forwarded-For'] = self
._x
_forwarded
_for
_ip
772 return self
._downloader
.urlopen(self
._create
_request
(url_or_request
, data
, headers
, query
))
773 except network_exceptions
as err
:
774 if isinstance(err
, urllib
.error
.HTTPError
):
775 if self
.__can
_accept
_status
_code
(err
, expected_status
):
776 # Retain reference to error to prevent file object from
777 # being closed before it can be read. Works around the
778 # effects of <https://bugs.python.org/issue15002>
779 # introduced in Python 3.4.1.
786 errnote
= 'Unable to download webpage'
788 errmsg
= f
'{errnote}: {error_to_compat_str(err)}'
790 raise ExtractorError(errmsg
, cause
=err
)
792 self
.report_warning(errmsg
)
795 def _download_webpage_handle(self
, url_or_request
, video_id
, note
=None, errnote
=None, fatal
=True,
796 encoding
=None, data
=None, headers
={}, query={}
, expected_status
=None):
798 Return a tuple (page content as string, URL handle).
801 url_or_request -- plain text URL as a string or
802 a urllib.request.Request object
803 video_id -- Video/playlist/item identifier (string)
806 note -- note printed before downloading (string)
807 errnote -- note printed in case of an error (string)
808 fatal -- flag denoting whether error should be considered fatal,
809 i.e. whether it should cause ExtractionError to be raised,
810 otherwise a warning will be reported and extraction continued
811 encoding -- encoding for a page content decoding, guessed automatically
812 when not explicitly specified
813 data -- POST data (bytes)
814 headers -- HTTP headers (dict)
815 query -- URL query (dict)
816 expected_status -- allows to accept failed HTTP requests (non 2xx
817 status code) by explicitly specifying a set of accepted status
818 codes. Can be any of the following entities:
819 - an integer type specifying an exact failed status code to
821 - a list or a tuple of integer types specifying a list of
822 failed status codes to accept
823 - a callable accepting an actual failed status code and
824 returning True if it should be accepted
825 Note that this argument does not affect success status codes (2xx)
826 which are always accepted.
829 # Strip hashes from the URL (#1038)
830 if isinstance(url_or_request
, str):
831 url_or_request
= url_or_request
.partition('#')[0]
833 urlh
= self
._request
_webpage
(url_or_request
, video_id
, note
, errnote
, fatal
, data
=data
, headers
=headers
, query
=query
, expected_status
=expected_status
)
837 content
= self
._webpage
_read
_content
(urlh
, url_or_request
, video_id
, note
, errnote
, fatal
, encoding
=encoding
)
838 return (content
, urlh
)
841 def _guess_encoding_from_content(content_type
, webpage_bytes
):
842 m
= re
.match(r
'[a-zA-Z0-9_.-]+/[a-zA-Z0-9_.-]+\s*;\s*charset=(.+)', content_type
)
844 encoding
= m
.group(1)
846 m
= re
.search(br
'<meta[^>]+charset=[\'"]?([^\'")]+)[ /\'">]',
847 webpage_bytes[:1024])
849 encoding = m.group(1).decode('ascii')
850 elif webpage_bytes.startswith(b'\xff\xfe'):
857 def __check_blocked(self, content):
858 first_block = content[:512]
859 if ('<title>Access to this site is blocked</title>' in content
860 and 'Websense' in first_block):
861 msg = 'Access to this webpage has been blocked by Websense filtering software in your network.'
862 blocked_iframe = self._html_search_regex(
863 r'<iframe src="([^
"]+)"', content,
864 'Websense information URL
', default=None)
866 msg += ' Visit
%s for more details
' % blocked_iframe
867 raise ExtractorError(msg, expected=True)
868 if '<title
>The URL you requested has been blocked
</title
>' in first_block:
870 'Access to this webpage has been blocked by Indian censorship
. '
871 'Use a VPN
or proxy
server (with --proxy
) to route around it
.')
872 block_msg = self._html_search_regex(
873 r'</h1
><p
>(.*?
)</p
>',
874 content, 'block message
', default=None)
876 msg += ' (Message
: "%s")' % block_msg.replace('\n', ' ')
877 raise ExtractorError(msg, expected=True)
878 if ('<title
>TTK
:: Доступ к ресурсу ограничен
</title
>' in content
879 and 'blocklist
.rkn
.gov
.ru
' in content):
880 raise ExtractorError(
881 'Access to this webpage has been blocked by decision of the Russian government
. '
882 'Visit http
://blocklist
.rkn
.gov
.ru
/ for a block reason
.',
885 def _request_dump_filename(self, url, video_id):
886 basen = f'{video_id}_{url}
'
887 trim_length = self.get_param('trim_file_name
') or 240
888 if len(basen) > trim_length:
889 h = '___
' + hashlib.md5(basen.encode('utf
-8')).hexdigest()
890 basen = basen[:trim_length - len(h)] + h
891 filename = sanitize_filename(f'{basen}
.dump
', restricted=True)
892 # Working around MAX_PATH limitation on Windows (see
893 # http://msdn.microsoft.com/en-us/library/windows/desktop/aa365247(v=vs.85).aspx)
894 if compat_os_name == 'nt
':
895 absfilepath = os.path.abspath(filename)
896 if len(absfilepath) > 259:
897 filename = fR'\\?\{absfilepath}
'
900 def __decode_webpage(self, webpage_bytes, encoding, headers):
902 encoding = self._guess_encoding_from_content(headers.get('Content
-Type
', ''), webpage_bytes)
904 return webpage_bytes.decode(encoding, 'replace
')
906 return webpage_bytes.decode('utf
-8', 'replace
')
908 def _webpage_read_content(self, urlh, url_or_request, video_id, note=None, errnote=None, fatal=True, prefix=None, encoding=None):
909 webpage_bytes = urlh.read()
910 if prefix is not None:
911 webpage_bytes = prefix + webpage_bytes
912 if self.get_param('dump_intermediate_pages
', False):
913 self.to_screen('Dumping request to
' + urlh.geturl())
914 dump = base64.b64encode(webpage_bytes).decode('ascii
')
915 self._downloader.to_screen(dump)
916 if self.get_param('write_pages
'):
917 filename = self._request_dump_filename(urlh.geturl(), video_id)
918 self.to_screen(f'Saving request to {filename}
')
919 with open(filename, 'wb
') as outf:
920 outf.write(webpage_bytes)
922 content = self.__decode_webpage(webpage_bytes, encoding, urlh.headers)
923 self.__check_blocked(content)
927 def _parse_xml(self, xml_string, video_id, transform_source=None, fatal=True):
929 xml_string = transform_source(xml_string)
931 return compat_etree_fromstring(xml_string.encode('utf
-8'))
932 except xml.etree.ElementTree.ParseError as ve:
933 errmsg = '%s: Failed to parse XML
' % video_id
935 raise ExtractorError(errmsg, cause=ve)
937 self.report_warning(errmsg + str(ve))
939 def _parse_json(self, json_string, video_id, transform_source=None, fatal=True, **parser_kwargs):
942 json_string, cls=LenientJSONDecoder, strict=False, transform_source=transform_source, **parser_kwargs)
943 except ValueError as ve:
944 errmsg = f'{video_id}
: Failed to parse JSON
'
946 raise ExtractorError(errmsg, cause=ve)
948 self.report_warning(f'{errmsg}
: {ve}
')
950 def _parse_socket_response_as_json(self, data, video_id, transform_source=None, fatal=True):
951 return self._parse_json(
952 data[data.find('{'):data.rfind('}
') + 1],
953 video_id, transform_source, fatal)
955 def __create_download_methods(name, parser, note, errnote, return_value):
957 def parse(ie, content, *args, **kwargs):
960 # parser is fetched by name so subclasses can override it
961 return getattr(ie, parser)(content, *args, **kwargs)
963 def download_handle(self, url_or_request, video_id, note=note, errnote=errnote, transform_source=None,
964 fatal=True, encoding=None, data=None, headers={}, query={}, expected_status=None):
965 res = self._download_webpage_handle(
966 url_or_request, video_id, note=note, errnote=errnote, fatal=fatal, encoding=encoding,
967 data=data, headers=headers, query=query, expected_status=expected_status)
971 return parse(self, content, video_id, transform_source=transform_source, fatal=fatal), urlh
973 def download_content(self, url_or_request, video_id, note=note, errnote=errnote, transform_source=None,
974 fatal=True, encoding=None, data=None, headers={}, query={}, expected_status=None):
975 if self.get_param('load_pages
'):
976 url_or_request = self._create_request(url_or_request, data, headers, query)
977 filename = self._request_dump_filename(url_or_request.full_url, video_id)
978 self.to_screen(f'Loading request
from {filename}
')
980 with open(filename, 'rb
') as dumpf:
981 webpage_bytes = dumpf.read()
983 self.report_warning(f'Unable to load request
from disk
: {e}
')
985 content = self.__decode_webpage(webpage_bytes, encoding, url_or_request.headers)
986 return parse(self, content, video_id, transform_source, fatal)
990 'transform_source
': transform_source,
992 'encoding
': encoding,
996 'expected_status
': expected_status,
999 kwargs.pop('transform_source
')
1000 # The method is fetched by name so subclasses can override _download_..._handle
1001 res = getattr(self, download_handle.__name__)(url_or_request, video_id, **kwargs)
1002 return res if res is False else res[0]
1004 def impersonate(func, name, return_value):
1005 func.__name__, func.__qualname__ = name, f'InfoExtractor
.{name}
'
1007 @param transform_source Apply this transformation before parsing
1008 @returns {return_value}
1010 See _download_webpage_handle docstring for other arguments specification
1013 impersonate(download_handle, f'_download_{name}_handle
', f'({return_value}
, URL handle
)')
1014 impersonate(download_content, f'_download_{name}
', f'{return_value}
')
1015 return download_handle, download_content
1017 _download_xml_handle, _download_xml = __create_download_methods(
1018 'xml
', '_parse_xml
', 'Downloading XML
', 'Unable to download XML
', 'xml
as an xml
.etree
.ElementTree
.Element
')
1019 _download_json_handle, _download_json = __create_download_methods(
1020 'json
', '_parse_json
', 'Downloading JSON metadata
', 'Unable to download JSON metadata
', 'JSON
object as a
dict')
1021 _download_socket_json_handle, _download_socket_json = __create_download_methods(
1022 'socket_json
', '_parse_socket_response_as_json
', 'Polling socket
', 'Unable to poll socket
', 'JSON
object as a
dict')
1023 __download_webpage = __create_download_methods('webpage
', None, None, None, 'data of the page
as a string
')[1]
1025 def _download_webpage(
1026 self, url_or_request, video_id, note=None, errnote=None,
1027 fatal=True, tries=1, timeout=NO_DEFAULT, *args, **kwargs):
1029 Return the data of the page as a string.
1032 tries -- number of tries
1033 timeout -- sleep interval between tries
1035 See _download_webpage_handle docstring for other arguments specification.
1038 R''' # NB: These are unused; should they be deprecated?
1040 self._downloader.deprecation_warning('tries argument
is deprecated
in InfoExtractor
._download
_webpage
')
1041 if timeout is NO_DEFAULT:
1044 self._downloader.deprecation_warning('timeout argument
is deprecated
in InfoExtractor
._download
_webpage
')
1050 return self.__download_webpage(url_or_request, video_id, note, errnote, None, fatal, *args, **kwargs)
1051 except http.client.IncompleteRead as e:
1053 if try_count >= tries:
1055 self._sleep(timeout, video_id)
1057 def report_warning(self, msg, video_id=None, *args, only_once=False, **kwargs):
1058 idstr = format_field(video_id, None, '%s: ')
1059 msg = f'[{self.IE_NAME}
] {idstr}{msg}
'
1061 if f'WARNING
: {msg}
' in self._printed_messages:
1063 self._printed_messages.add(f'WARNING
: {msg}
')
1064 self._downloader.report_warning(msg, *args, **kwargs)
1066 def to_screen(self, msg, *args, **kwargs):
1067 """Print msg to screen, prefixing it with '[ie_name
]'"""
1068 self._downloader.to_screen(f'[{self.IE_NAME}
] {msg}
', *args, **kwargs)
1070 def write_debug(self, msg, *args, **kwargs):
1071 self._downloader.write_debug(f'[{self.IE_NAME}
] {msg}
', *args, **kwargs)
1073 def get_param(self, name, default=None, *args, **kwargs):
1074 if self._downloader:
1075 return self._downloader.params.get(name, default, *args, **kwargs)
1078 def report_drm(self, video_id, partial=False):
1079 self.raise_no_formats('This video
is DRM protected
', expected=True, video_id=video_id)
1081 def report_extraction(self, id_or_name):
1082 """Report information extraction."""
1083 self.to_screen('%s: Extracting information
' % id_or_name)
1085 def report_download_webpage(self, video_id):
1086 """Report webpage download."""
1087 self.to_screen('%s: Downloading webpage
' % video_id)
1089 def report_age_confirmation(self):
1090 """Report attempt to confirm age."""
1091 self.to_screen('Confirming age
')
1093 def report_login(self):
1094 """Report attempt to log in."""
1095 self.to_screen('Logging
in')
1097 def raise_login_required(
1098 self, msg='This video
is only available
for registered users
',
1099 metadata_available=False, method=NO_DEFAULT):
1100 if metadata_available and (
1101 self.get_param('ignore_no_formats_error
') or self.get_param('wait_for_video
')):
1102 self.report_warning(msg)
1104 msg += format_field(self._login_hint(method), None, '. %s')
1105 raise ExtractorError(msg, expected=True)
1107 def raise_geo_restricted(
1108 self, msg='This video
is not available
from your location due to geo restriction
',
1109 countries=None, metadata_available=False):
1110 if metadata_available and (
1111 self.get_param('ignore_no_formats_error
') or self.get_param('wait_for_video
')):
1112 self.report_warning(msg)
1114 raise GeoRestrictedError(msg, countries=countries)
1116 def raise_no_formats(self, msg, expected=False, video_id=None):
1118 self.get_param('ignore_no_formats_error
') or self.get_param('wait_for_video
')):
1119 self.report_warning(msg, video_id)
1120 elif isinstance(msg, ExtractorError):
1123 raise ExtractorError(msg, expected=expected, video_id=video_id)
1125 # Methods for following #608
1127 def url_result(url, ie=None, video_id=None, video_title=None, *, url_transparent=False, **kwargs):
1128 """Returns a URL that points to a page that should be processed"""
1130 kwargs['ie_key
'] = ie if isinstance(ie, str) else ie.ie_key()
1131 if video_id is not None:
1132 kwargs['id'] = video_id
1133 if video_title is not None:
1134 kwargs['title
'] = video_title
1137 '_type
': 'url_transparent
' if url_transparent else 'url
',
1141 def playlist_from_matches(self, matches, playlist_id=None, playlist_title=None, getter=None, ie=None, video_kwargs=None, **kwargs):
1142 urls = (self.url_result(self._proto_relative_url(m), ie, **(video_kwargs or {}))
1143 for m in orderedSet(map(getter, matches) if getter else matches))
1144 return self.playlist_result(urls, playlist_id, playlist_title, **kwargs)
1147 def playlist_result(entries, playlist_id=None, playlist_title=None, playlist_description=None, *, multi_video=False, **kwargs):
1148 """Returns a playlist"""
1150 kwargs['id'] = playlist_id
1152 kwargs['title
'] = playlist_title
1153 if playlist_description is not None:
1154 kwargs['description
'] = playlist_description
1157 '_type
': 'multi_video
' if multi_video else 'playlist
',
1161 def _search_regex(self, pattern, string, name, default=NO_DEFAULT, fatal=True, flags=0, group=None):
1163 Perform a regex search on the given string, using a single or a list of
1164 patterns returning the first matching group.
1165 In case of failure return a default value or raise a WARNING or a
1166 RegexNotFoundError, depending on fatal, specifying the field name.
1170 elif isinstance(pattern, (str, re.Pattern)):
1171 mobj = re.search(pattern, string, flags)
1174 mobj = re.search(p, string, flags)
1178 _name = self._downloader._format_err(name, self._downloader.Styles.EMPHASIS)
1182 # return the first matching group
1183 return next(g for g in mobj.groups() if g is not None)
1184 elif isinstance(group, (list, tuple)):
1185 return tuple(mobj.group(g) for g in group)
1187 return mobj.group(group)
1188 elif default is not NO_DEFAULT:
1191 raise RegexNotFoundError('Unable to extract
%s' % _name)
1193 self.report_warning('unable to extract
%s' % _name + bug_reports_message())
1196 def _search_json(self, start_pattern, string, name, video_id, *, end_pattern='',
1197 contains_pattern='(?s
:.+)', fatal=True, default=NO_DEFAULT, **kwargs):
1198 """Searches string for the JSON object specified by start_pattern"""
1199 # NB: end_pattern is only used to reduce the size of the initial match
1200 if default is NO_DEFAULT:
1201 default, has_default = {}, False
1203 fatal, has_default = False, True
1205 json_string = self._search_regex(
1206 rf'{start_pattern}\s
*(?P
<json
>{{\s*{contains_pattern}\s
*}})\s
*{end_pattern}
',
1207 string, name, group='json
', fatal=fatal, default=None if has_default else NO_DEFAULT)
1211 _name = self._downloader._format_err(name, self._downloader.Styles.EMPHASIS)
1213 return self._parse_json(json_string, video_id, ignore_extra=True, **kwargs)
1214 except ExtractorError as e:
1216 raise ExtractorError(
1217 f'Unable to extract {_name}
- Failed to parse JSON
', cause=e.cause, video_id=video_id)
1218 elif not has_default:
1219 self.report_warning(
1220 f'Unable to extract {_name}
- Failed to parse JSON
: {e}
', video_id=video_id)
1223 def _html_search_regex(self, pattern, string, name, default=NO_DEFAULT, fatal=True, flags=0, group=None):
1225 Like _search_regex, but strips HTML tags and unescapes entities.
1227 res = self._search_regex(pattern, string, name, default, fatal, flags, group)
1229 return clean_html(res).strip()
1233 def _get_netrc_login_info(self, netrc_machine=None):
1236 netrc_machine = netrc_machine or self._NETRC_MACHINE
1238 if self.get_param('usenetrc
', False):
1240 netrc_file = compat_expanduser(self.get_param('netrc_location
') or '~
')
1241 if os.path.isdir(netrc_file):
1242 netrc_file = os.path.join(netrc_file, '.netrc
')
1243 info = netrc.netrc(file=netrc_file).authenticators(netrc_machine)
1244 if info is not None:
1248 raise netrc.NetrcParseError(
1249 'No authenticators
for %s' % netrc_machine)
1250 except (OSError, netrc.NetrcParseError) as err:
1251 self.report_warning(
1252 'parsing
.netrc
: %s' % error_to_compat_str(err))
1254 return username, password
1256 def _get_login_info(self, username_option='username
', password_option='password
', netrc_machine=None):
1258 Get the login info as (username, password)
1259 First look for the manually specified credentials using username_option
1260 and password_option as keys in params dictionary. If no such credentials
1261 available look in the netrc file using the netrc_machine or _NETRC_MACHINE
1263 If there's no info available
, return (None, None)
1266 # Attempt to use provided username and password or .netrc data
1267 username = self.get_param(username_option)
1268 if username is not None:
1269 password = self.get_param(password_option)
1271 username, password = self._get_netrc_login_info(netrc_machine)
1273 return username, password
1275 def _get_tfa_info(self, note='two-factor verification code'):
1277 Get the two
-factor authentication info
1278 TODO
- asking the user will be required
for sms
/phone verify
1279 currently just uses the command line option
1280 If there
's no info available, return None
1283 tfa = self.get_param('twofactor
')
1287 return getpass.getpass('Type
%s and press
[Return
]: ' % note)
1289 # Helper functions for extracting OpenGraph info
1291 def _og_regexes(prop):
1292 content_re = r'content
=(?
:"([^"]+?
)"|\'([^\']+?)\'|\s*([^\s"\'=<>`
]+?
))'
1293 property_re = (r'(?
:name|
property)=(?
:\'og
%(sep)s%(prop)s\'|
"og%(sep)s%(prop)s"|\s
*og
%(sep)s%(prop)s\b)'
1294 % {'prop': re.escape(prop), 'sep': '(?::|[:-])'})
1295 template = r'<meta
[^
>]+?
%s[^
>]+?
%s'
1297 template % (property_re, content_re),
1298 template % (content_re, property_re),
1302 def _meta_regex(prop):
1303 return r'''(?isx)<meta
1304 (?=[^>]+(?:itemprop|name|property|id|http-equiv)=(["\']?)%s\1)
1305 [^>]+?content=(["\'])(?P<content>.*?)\2''' % re.escape(prop)
1307 def _og_search_property(self, prop, html, name=None, **kargs):
1308 prop = variadic(prop)
1310 name = 'OpenGraph
%s' % prop[0]
1313 og_regexes.extend(self._og_regexes(p))
1314 escaped = self._search_regex(og_regexes, html, name, flags=re.DOTALL, **kargs)
1317 return unescapeHTML(escaped)
1319 def _og_search_thumbnail(self, html, **kargs):
1320 return self._og_search_property('image
', html, 'thumbnail URL
', fatal=False, **kargs)
1322 def _og_search_description(self, html, **kargs):
1323 return self._og_search_property('description
', html, fatal=False, **kargs)
1325 def _og_search_title(self, html, *, fatal=False, **kargs):
1326 return self._og_search_property('title
', html, fatal=fatal, **kargs)
1328 def _og_search_video_url(self, html, name='video url
', secure=True, **kargs):
1329 regexes = self._og_regexes('video
') + self._og_regexes('video
:url
')
1331 regexes = self._og_regexes('video
:secure_url
') + regexes
1332 return self._html_search_regex(regexes, html, name, **kargs)
1334 def _og_search_url(self, html, **kargs):
1335 return self._og_search_property('url
', html, **kargs)
1337 def _html_extract_title(self, html, name='title
', *, fatal=False, **kwargs):
1338 return self._html_search_regex(r'(?s
)<title
\b[^
>]*>([^
<]+)</title
>', html, name, fatal=fatal, **kwargs)
1340 def _html_search_meta(self, name, html, display_name=None, fatal=False, **kwargs):
1341 name = variadic(name)
1342 if display_name is None:
1343 display_name = name[0]
1344 return self._html_search_regex(
1345 [self._meta_regex(n) for n in name],
1346 html, display_name, fatal=fatal, group='content
', **kwargs)
1348 def _dc_search_uploader(self, html):
1349 return self._html_search_meta('dc
.creator
', html, 'uploader
')
1351 def _rta_search(self, html):
1352 # See http://www.rtalabel.org/index.php?content=howtofaq#single
1353 if re.search(r'(?ix
)<meta\s
+name
="rating"\s
+'
1354 r' content
="RTA-5042-1996-1400-1577-RTA"',
1359 def _media_rating_search(self, html):
1360 # See http://www.tjg-designs.com/WP/metadata-code-examples-adding-metadata-to-your-web-pages/
1361 rating = self._html_search_meta('rating
', html)
1373 return RATING_TABLE.get(rating.lower())
1375 def _family_friendly_search(self, html):
1376 # See http://schema.org/VideoObject
1377 family_friendly = self._html_search_meta(
1378 'isFamilyFriendly
', html, default=None)
1380 if not family_friendly:
1389 return RATING_TABLE.get(family_friendly.lower())
1391 def _twitter_search_player(self, html):
1392 return self._html_search_meta('twitter
:player
', html,
1393 'twitter card player
')
1395 def _yield_json_ld(self, html, video_id, *, fatal=True, default=NO_DEFAULT):
1396 """Yield all json ld objects in the html"""
1397 if default is not NO_DEFAULT:
1399 for mobj in re.finditer(JSON_LD_RE, html):
1400 json_ld_item = self._parse_json(mobj.group('json_ld
'), video_id, fatal=fatal)
1401 for json_ld in variadic(json_ld_item):
1402 if isinstance(json_ld, dict):
1405 def _search_json_ld(self, html, video_id, expected_type=None, *, fatal=True, default=NO_DEFAULT):
1406 """Search for a video in any json ld in the html"""
1407 if default is not NO_DEFAULT:
1409 info = self._json_ld(
1410 list(self._yield_json_ld(html, video_id, fatal=fatal, default=default)),
1411 video_id, fatal=fatal, expected_type=expected_type)
1414 if default is not NO_DEFAULT:
1417 raise RegexNotFoundError('Unable to extract JSON
-LD
')
1419 self.report_warning('unable to extract JSON
-LD
%s' % bug_reports_message())
1422 def _json_ld(self, json_ld, video_id, fatal=True, expected_type=None):
1423 if isinstance(json_ld, str):
1424 json_ld = self._parse_json(json_ld, video_id, fatal=fatal)
1428 if not isinstance(json_ld, (list, tuple, dict)):
1430 if isinstance(json_ld, dict):
1433 INTERACTION_TYPE_MAP = {
1434 'CommentAction
': 'comment
',
1435 'AgreeAction
': 'like
',
1436 'DisagreeAction
': 'dislike
',
1437 'LikeAction
': 'like
',
1438 'DislikeAction
': 'dislike
',
1439 'ListenAction
': 'view
',
1440 'WatchAction
': 'view
',
1441 'ViewAction
': 'view
',
1444 def is_type(e, *expected_types):
1445 type = variadic(traverse_obj(e, '@type'))
1446 return any(x in type for x in expected_types)
1448 def extract_interaction_type(e):
1449 interaction_type = e.get('interactionType
')
1450 if isinstance(interaction_type, dict):
1451 interaction_type = interaction_type.get('@type')
1452 return str_or_none(interaction_type)
1454 def extract_interaction_statistic(e):
1455 interaction_statistic = e.get('interactionStatistic
')
1456 if isinstance(interaction_statistic, dict):
1457 interaction_statistic = [interaction_statistic]
1458 if not isinstance(interaction_statistic, list):
1460 for is_e in interaction_statistic:
1461 if not is_type(is_e, 'InteractionCounter
'):
1463 interaction_type = extract_interaction_type(is_e)
1464 if not interaction_type:
1466 # For interaction count some sites provide string instead of
1467 # an integer (as per spec) with non digit characters (e.g. ",")
1468 # so extracting count with more relaxed str_to_int
1469 interaction_count = str_to_int(is_e.get('userInteractionCount
'))
1470 if interaction_count is None:
1472 count_kind = INTERACTION_TYPE_MAP.get(interaction_type.split('/')[-1])
1475 count_key = '%s_count
' % count_kind
1476 if info.get(count_key) is not None:
1478 info[count_key] = interaction_count
1480 def extract_chapter_information(e):
1482 'title
': part.get('name
'),
1483 'start_time
': part.get('startOffset
'),
1484 'end_time
': part.get('endOffset
'),
1485 } for part in variadic(e.get('hasPart
') or []) if part.get('@type') == 'Clip
']
1486 for idx, (last_c, current_c, next_c) in enumerate(zip(
1487 [{'end_time': 0}] + chapters, chapters, chapters[1:])):
1488 current_c['end_time
'] = current_c['end_time
'] or next_c['start_time
']
1489 current_c['start_time
'] = current_c['start_time
'] or last_c['end_time
']
1490 if None in current_c.values():
1491 self.report_warning(f'Chapter {idx} contains broken data
. Not extracting chapters
')
1494 chapters[-1]['end_time
'] = chapters[-1]['end_time
'] or info['duration
']
1495 info['chapters
'] = chapters
1497 def extract_video_object(e):
1498 assert is_type(e, 'VideoObject
')
1499 author = e.get('author
')
1501 'url
': url_or_none(e.get('contentUrl
')),
1502 'title
': unescapeHTML(e.get('name
')),
1503 'description
': unescapeHTML(e.get('description
')),
1504 'thumbnails
': [{'url': url}
1505 for url in variadic(traverse_obj(e, 'thumbnailUrl
', 'thumbnailURL
'))
1506 if url_or_none(url)],
1507 'duration
': parse_duration(e.get('duration
')),
1508 'timestamp
': unified_timestamp(e.get('uploadDate
')),
1509 # author can be an instance of 'Organization
' or 'Person
' types.
1510 # both types can have 'name
' property(inherited from 'Thing
' type). [1]
1511 # however some websites are using 'Text
' type instead.
1512 # 1. https://schema.org/VideoObject
1513 'uploader
': author.get('name
') if isinstance(author, dict) else author if isinstance(author, str) else None,
1514 'filesize
': int_or_none(float_or_none(e.get('contentSize
'))),
1515 'tbr
': int_or_none(e.get('bitrate
')),
1516 'width
': int_or_none(e.get('width
')),
1517 'height
': int_or_none(e.get('height
')),
1518 'view_count
': int_or_none(e.get('interactionCount
')),
1520 extract_interaction_statistic(e)
1521 extract_chapter_information(e)
1523 def traverse_json_ld(json_ld, at_top_level=True):
1525 if at_top_level and '@context' not in e:
1527 if at_top_level and set(e.keys()) == {'@context', '@graph'}:
1528 traverse_json_ld(variadic(e['@graph'], allowed_types=(dict,)), at_top_level=False)
1530 if expected_type is not None and not is_type(e, expected_type):
1532 rating = traverse_obj(e, ('aggregateRating
', 'ratingValue
'), expected_type=float_or_none)
1533 if rating is not None:
1534 info['average_rating
'] = rating
1535 if is_type(e, 'TVEpisode
', 'Episode
'):
1536 episode_name = unescapeHTML(e.get('name
'))
1538 'episode
': episode_name,
1539 'episode_number
': int_or_none(e.get('episodeNumber
')),
1540 'description
': unescapeHTML(e.get('description
')),
1542 if not info.get('title
') and episode_name:
1543 info['title
'] = episode_name
1544 part_of_season = e.get('partOfSeason
')
1545 if is_type(part_of_season, 'TVSeason
', 'Season
', 'CreativeWorkSeason
'):
1547 'season
': unescapeHTML(part_of_season.get('name
')),
1548 'season_number
': int_or_none(part_of_season.get('seasonNumber
')),
1550 part_of_series = e.get('partOfSeries
') or e.get('partOfTVSeries
')
1551 if is_type(part_of_series, 'TVSeries
', 'Series
', 'CreativeWorkSeries
'):
1552 info['series
'] = unescapeHTML(part_of_series.get('name
'))
1553 elif is_type(e, 'Movie
'):
1555 'title
': unescapeHTML(e.get('name
')),
1556 'description
': unescapeHTML(e.get('description
')),
1557 'duration
': parse_duration(e.get('duration
')),
1558 'timestamp
': unified_timestamp(e.get('dateCreated
')),
1560 elif is_type(e, 'Article
', 'NewsArticle
'):
1562 'timestamp
': parse_iso8601(e.get('datePublished
')),
1563 'title
': unescapeHTML(e.get('headline
')),
1564 'description
': unescapeHTML(e.get('articleBody
') or e.get('description
')),
1566 if is_type(traverse_obj(e, ('video
', 0)), 'VideoObject
'):
1567 extract_video_object(e['video
'][0])
1568 elif is_type(traverse_obj(e, ('subjectOf
', 0)), 'VideoObject
'):
1569 extract_video_object(e['subjectOf
'][0])
1570 elif is_type(e, 'VideoObject
'):
1571 extract_video_object(e)
1572 if expected_type is None:
1576 video = e.get('video
')
1577 if is_type(video, 'VideoObject
'):
1578 extract_video_object(video)
1579 if expected_type is None:
1583 traverse_json_ld(json_ld)
1585 return filter_dict(info)
1587 def _search_nextjs_data(self, webpage, video_id, *, transform_source=None, fatal=True, **kw):
1588 return self._parse_json(
1590 r'(?s
)<script
[^
>]+id=[\'"]__NEXT_DATA__[\'"][^
>]*>([^
<]+)</script
>',
1591 webpage, 'next
.js data
', fatal=fatal, **kw),
1592 video_id, transform_source=transform_source, fatal=fatal)
1594 def _search_nuxt_data(self, webpage, video_id, context_name='__NUXT__
', *, fatal=True, traverse=('data
', 0)):
1595 """Parses Nuxt.js metadata. This works as long as the function __NUXT__ invokes is a pure function"""
1596 rectx = re.escape(context_name)
1597 FUNCTION_RE = r'\
(function\
((?P
<arg_keys
>.*?
)\
){return\s+(?P<js>{.*?}
)\s
*;?\s
*}\
((?P
<arg_vals
>.*?
)\
)'
1598 js, arg_keys, arg_vals = self._search_regex(
1599 (rf'<script
>\s
*window\
.{rectx}
={FUNCTION_RE}\s
*\
)\s
*;?\s
*</script
>', rf'{rectx}\
(.*?{FUNCTION_RE}
'),
1600 webpage, context_name, group=('js
', 'arg_keys
', 'arg_vals
'), fatal=fatal)
1602 args = dict(zip(arg_keys.split(','), arg_vals.split(',')))
1604 for key, val in args.items():
1605 if val in ('undefined
', 'void
0'):
1608 ret = self._parse_json(js, video_id, transform_source=functools.partial(js_to_json, vars=args), fatal=fatal)
1609 return traverse_obj(ret, traverse) or {}
1612 def _hidden_inputs(html):
1613 html = re.sub(r'<!--(?
:(?
!<!--).)*-->', '', html)
1615 for input in re.findall(r'(?i
)(<input[^
>]+>)', html):
1616 attrs = extract_attributes(input)
1619 if attrs.get('type') not in ('hidden
', 'submit
'):
1621 name = attrs.get('name
') or attrs.get('id')
1622 value = attrs.get('value
')
1623 if name and value is not None:
1624 hidden_inputs[name] = value
1625 return hidden_inputs
1627 def _form_hidden_inputs(self, form_id, html):
1628 form = self._search_regex(
1629 r'(?
is)<form
[^
>]+?
id=(["\'])%s\1[^>]*>(?P<form>.+?)</form>' % form_id,
1630 html, '%s form' % form_id, group='form')
1631 return self._hidden_inputs(form)
1634 regex = r' *((?P<reverse>\+)?(?P<field>[a-zA-Z0-9_]+)((?P<separator>[~:])(?P<limit>.*?))?)? *$'
1636 default = ('hidden', 'aud_or_vid', 'hasvid', 'ie_pref', 'lang', 'quality',
1637 'res', 'fps', 'hdr:12', 'codec:vp9.2', 'size', 'br', 'asr',
1638 'proto', 'ext', 'hasaud', 'source', 'id') # These must not be aliases
1639 ytdl_default = ('hasaud', 'lang', 'quality', 'tbr', 'filesize', 'vbr',
1640 'height', 'width', 'proto', 'vext', 'abr', 'aext',
1641 'fps', 'fs_approx', 'source', 'id')
1644 'vcodec': {'type': 'ordered', 'regex': True,
1645 'order': ['av0?1', 'vp0?9.2', 'vp0?9', '[hx]265|he?vc?', '[hx]264|avc', 'vp0?8', 'mp4v|h263', 'theora', '', None, 'none']},
1646 'acodec': {'type': 'ordered', 'regex': True,
1647 'order': ['[af]lac', 'wav|aiff', 'opus', 'vorbis|ogg', 'aac', 'mp?4a?', 'mp3', 'e-?a?c-?3', 'ac-?3', 'dts', '', None, 'none']},
1648 'hdr': {'type': 'ordered', 'regex': True, 'field': 'dynamic_range',
1649 'order': ['dv', '(hdr)?12', r'(hdr)?10\+', '(hdr)?10', 'hlg', '', 'sdr', None]},
1650 'proto': {'type': 'ordered', 'regex': True, 'field': 'protocol',
1651 'order': ['(ht|f)tps', '(ht|f)tp$', 'm3u8.*', '.*dash', 'websocket_frag', 'rtmpe?', '', 'mms|rtsp', 'ws|websocket', 'f4']},
1652 'vext': {'type': 'ordered', 'field': 'video_ext',
1653 'order': ('mp4', 'webm', 'flv', '', 'none'),
1654 'order_free': ('webm', 'mp4', 'flv', '', 'none')},
1655 'aext': {'type': 'ordered', 'field': 'audio_ext',
1656 'order': ('m4a', 'aac', 'mp3', 'ogg', 'opus', 'webm', '', 'none'),
1657 'order_free': ('opus', 'ogg', 'webm', 'm4a', 'mp3', 'aac', '', 'none')},
1658 'hidden': {'visible': False, 'forced': True, 'type': 'extractor', 'max': -1000},
1659 'aud_or_vid': {'visible': False, 'forced': True, 'type': 'multiple',
1660 'field': ('vcodec', 'acodec'),
1661 'function': lambda it: int(any(v != 'none' for v in it))},
1662 'ie_pref': {'priority': True, 'type': 'extractor'},
1663 'hasvid': {'priority': True, 'field': 'vcodec', 'type': 'boolean', 'not_in_list': ('none',)},
1664 'hasaud': {'field': 'acodec', 'type': 'boolean', 'not_in_list': ('none',)},
1665 'lang': {'convert': 'float', 'field': 'language_preference', 'default': -1},
1666 'quality': {'convert': 'float', 'default': -1},
1667 'filesize': {'convert': 'bytes'},
1668 'fs_approx': {'convert': 'bytes', 'field': 'filesize_approx'},
1669 'id': {'convert': 'string', 'field': 'format_id'},
1670 'height': {'convert': 'float_none'},
1671 'width': {'convert': 'float_none'},
1672 'fps': {'convert': 'float_none'},
1673 'tbr': {'convert': 'float_none'},
1674 'vbr': {'convert': 'float_none'},
1675 'abr': {'convert': 'float_none'},
1676 'asr': {'convert': 'float_none'},
1677 'source': {'convert': 'float', 'field': 'source_preference', 'default': -1},
1679 'codec': {'type': 'combined', 'field': ('vcodec', 'acodec')},
1680 'br': {'type': 'combined', 'field': ('tbr', 'vbr', 'abr'), 'same_limit': True},
1681 'size': {'type': 'combined', 'same_limit': True, 'field': ('filesize', 'fs_approx')},
1682 'ext': {'type': 'combined', 'field': ('vext', 'aext')},
1683 'res': {'type': 'multiple', 'field': ('height', 'width'),
1684 'function': lambda it: (lambda l: min(l) if l else 0)(tuple(filter(None, it)))},
1686 # For compatibility with youtube-dl
1687 'format_id': {'type': 'alias', 'field': 'id'},
1688 'preference': {'type': 'alias', 'field': 'ie_pref'},
1689 'language_preference': {'type': 'alias', 'field': 'lang'},
1690 'source_preference': {'type': 'alias', 'field': 'source'},
1691 'protocol': {'type': 'alias', 'field': 'proto'},
1692 'filesize_approx': {'type': 'alias', 'field': 'fs_approx'},
1695 'dimension': {'type': 'alias', 'field': 'res', 'deprecated': True},
1696 'resolution': {'type': 'alias', 'field': 'res', 'deprecated': True},
1697 'extension': {'type': 'alias', 'field': 'ext', 'deprecated': True},
1698 'bitrate': {'type': 'alias', 'field': 'br', 'deprecated': True},
1699 'total_bitrate': {'type': 'alias', 'field': 'tbr', 'deprecated': True},
1700 'video_bitrate': {'type': 'alias', 'field': 'vbr', 'deprecated': True},
1701 'audio_bitrate': {'type': 'alias', 'field': 'abr', 'deprecated': True},
1702 'framerate': {'type': 'alias', 'field': 'fps', 'deprecated': True},
1703 'filesize_estimate': {'type': 'alias', 'field': 'size', 'deprecated': True},
1704 'samplerate': {'type': 'alias', 'field': 'asr', 'deprecated': True},
1705 'video_ext': {'type': 'alias', 'field': 'vext', 'deprecated': True},
1706 'audio_ext': {'type': 'alias', 'field': 'aext', 'deprecated': True},
1707 'video_codec': {'type': 'alias', 'field': 'vcodec', 'deprecated': True},
1708 'audio_codec': {'type': 'alias', 'field': 'acodec', 'deprecated': True},
1709 'video': {'type': 'alias', 'field': 'hasvid', 'deprecated': True},
1710 'has_video': {'type': 'alias', 'field': 'hasvid', 'deprecated': True},
1711 'audio': {'type': 'alias', 'field': 'hasaud', 'deprecated': True},
1712 'has_audio': {'type': 'alias', 'field': 'hasaud', 'deprecated': True},
1713 'extractor': {'type': 'alias', 'field': 'ie_pref', 'deprecated': True},
1714 'extractor_preference': {'type': 'alias', 'field': 'ie_pref', 'deprecated': True},
1717 def __init__(self, ie, field_preference):
1719 self.ydl = ie._downloader
1720 self.evaluate_params(self.ydl.params, field_preference)
1721 if ie.get_param('verbose'):
1722 self.print_verbose_info(self.ydl.write_debug)
1724 def _get_field_setting(self, field, key):
1725 if field not in self.settings:
1726 if key in ('forced', 'priority'):
1728 self.ydl.deprecation_warning(
1729 f'Using arbitrary fields ({field}) for format sorting is deprecated '
1730 'and may be removed in a future version')
1731 self.settings[field] = {}
1732 propObj = self.settings[field]
1733 if key not in propObj:
1734 type = propObj.get('type')
1736 default = 'preference' if type == 'extractor' else (field,) if type in ('combined', 'multiple') else field
1737 elif key == 'convert':
1738 default = 'order' if type == 'ordered' else 'float_string' if field else 'ignore'
1740 default = {'type': 'field', 'visible': True, 'order': [], 'not_in_list': (None,)}.get(key, None)
1741 propObj[key] = default
1744 def _resolve_field_value(self, field, value, convertNone=False):
1749 value = value.lower()
1750 conversion = self._get_field_setting(field, 'convert')
1751 if conversion == 'ignore':
1753 if conversion == 'string':
1755 elif conversion == 'float_none':
1756 return float_or_none(value)
1757 elif conversion == 'bytes':
1758 return FileDownloader.parse_bytes(value)
1759 elif conversion == 'order':
1760 order_list = (self._use_free_order and self._get_field_setting(field, 'order_free')) or self._get_field_setting(field, 'order')
1761 use_regex = self._get_field_setting(field, 'regex')
1762 list_length = len(order_list)
1763 empty_pos = order_list.index('') if '' in order_list else list_length + 1
1764 if use_regex and value is not None:
1765 for i, regex in enumerate(order_list):
1766 if regex and re.match(regex, value):
1767 return list_length - i
1768 return list_length - empty_pos # not in list
1769 else: # not regex or value = None
1770 return list_length - (order_list.index(value) if value in order_list else empty_pos)
1772 if value.isnumeric():
1775 self.settings[field]['convert'] = 'string'
1778 def evaluate_params(self, params, sort_extractor):
1779 self._use_free_order = params.get('prefer_free_formats', False)
1780 self._sort_user = params.get('format_sort', [])
1781 self._sort_extractor = sort_extractor
1783 def add_item(field, reverse, closest, limit_text):
1784 field = field.lower()
1785 if field in self._order:
1787 self._order.append(field)
1788 limit = self._resolve_field_value(field, limit_text)
1791 'closest': False if limit is None else closest,
1792 'limit_text': limit_text,
1794 if field in self.settings:
1795 self.settings[field].update(data)
1797 self.settings[field] = data
1800 tuple(field for field in self.default if self._get_field_setting(field, 'forced'))
1801 + (tuple() if params.get('format_sort_force', False)
1802 else tuple(field for field in self.default if self._get_field_setting(field, 'priority')))
1803 + tuple(self._sort_user) + tuple(sort_extractor) + self.default)
1805 for item in sort_list:
1806 match = re.match(self.regex, item)
1808 raise ExtractorError('Invalid format sort string "%s" given by extractor' % item)
1809 field = match.group('field')
1812 if self._get_field_setting(field, 'type') == 'alias':
1813 alias, field = field, self._get_field_setting(field, 'field')
1814 if self._get_field_setting(alias, 'deprecated'):
1815 self.ydl.deprecation_warning(
1816 f'Format sorting alias {alias} is deprecated '
1817 f'and may be removed in a future version. Please use {field} instead')
1818 reverse = match.group('reverse') is not None
1819 closest = match.group('separator') == '~'
1820 limit_text = match.group('limit')
1822 has_limit = limit_text is not None
1823 has_multiple_fields = self._get_field_setting(field, 'type') == 'combined'
1824 has_multiple_limits = has_limit and has_multiple_fields and not self._get_field_setting(field, 'same_limit')
1826 fields = self._get_field_setting(field, 'field') if has_multiple_fields else (field,)
1827 limits = limit_text.split(':') if has_multiple_limits else (limit_text,) if has_limit else tuple()
1828 limit_count = len(limits)
1829 for (i, f) in enumerate(fields):
1830 add_item(f, reverse, closest,
1831 limits[i] if i < limit_count
1832 else limits[0] if has_limit and not has_multiple_limits
1835 def print_verbose_info(self, write_debug):
1837 write_debug('Sort order given by user: %s' % ', '.join(self._sort_user))
1838 if self._sort_extractor:
1839 write_debug('Sort order given by extractor: %s' % ', '.join(self._sort_extractor))
1840 write_debug('Formats sorted by: %s' % ', '.join(['%s%s%s' % (
1841 '+' if self._get_field_setting(field, 'reverse') else '', field,
1842 '%s%s(%s)' % ('~' if self._get_field_setting(field, 'closest') else ':',
1843 self._get_field_setting(field, 'limit_text'),
1844 self._get_field_setting(field, 'limit'))
1845 if self._get_field_setting(field, 'limit_text') is not None else '')
1846 for field in self._order if self._get_field_setting(field, 'visible')]))
1848 def _calculate_field_preference_from_value(self, format, field, type, value):
1849 reverse = self._get_field_setting(field, 'reverse')
1850 closest = self._get_field_setting(field, 'closest')
1851 limit = self._get_field_setting(field, 'limit')
1853 if type == 'extractor':
1854 maximum = self._get_field_setting(field, 'max')
1855 if value is None or (maximum is not None and value >= maximum):
1857 elif type == 'boolean':
1858 in_list = self._get_field_setting(field, 'in_list')
1859 not_in_list = self._get_field_setting(field, 'not_in_list')
1860 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
1861 elif type == 'ordered':
1862 value = self._resolve_field_value(field, value, True)
1864 # try to convert to number
1865 val_num = float_or_none(value, default=self._get_field_setting(field, 'default'))
1866 is_num = self._get_field_setting(field, 'convert') != 'string' and val_num is not None
1870 return ((-10, 0) if value is None
1871 else (1, value, 0) if not is_num # if a field has mixed strings and numbers, strings are sorted higher
1872 else (0, -abs(value - limit), value - limit if reverse else limit - value) if closest
1873 else (0, value, 0) if not reverse and (limit is None or value <= limit)
1874 else (0, -value, 0) if limit is None or (reverse and value == limit) or value > limit
1875 else (-1, value, 0))
1877 def _calculate_field_preference(self, format, field):
1878 type = self._get_field_setting(field, 'type') # extractor, boolean, ordered, field, multiple
1879 get_value = lambda f: format.get(self._get_field_setting(f, 'field'))
1880 if type == 'multiple':
1881 type = 'field' # Only 'field' is allowed in multiple for now
1882 actual_fields = self._get_field_setting(field, 'field')
1884 value = self._get_field_setting(field, 'function')(get_value(f) for f in actual_fields)
1886 value = get_value(field)
1887 return self._calculate_field_preference_from_value(format, field, type, value)
1889 def calculate_preference(self, format):
1890 # Determine missing protocol
1891 if not format.get('protocol'):
1892 format['protocol'] = determine_protocol(format)
1894 # Determine missing ext
1895 if not format.get('ext') and 'url' in format:
1896 format['ext'] = determine_ext(format['url'])
1897 if format.get('vcodec') == 'none':
1898 format['audio_ext'] = format['ext'] if format.get('acodec') != 'none' else 'none'
1899 format['video_ext'] = 'none'
1901 format['video_ext'] = format['ext']
1902 format['audio_ext'] = 'none'
1903 # if format.get('preference') is None and format.get('ext') in ('f4f', 'f4m'): # Not supported?
1904 # format['preference'] = -1000
1906 # Determine missing bitrates
1907 if format.get('tbr') is None:
1908 if format.get('vbr') is not None and format.get('abr') is not None:
1909 format['tbr'] = format.get('vbr', 0) + format.get('abr', 0)
1911 if format.get('vcodec') != 'none' and format.get('vbr') is None:
1912 format['vbr'] = format.get('tbr') - format.get('abr', 0)
1913 if format.get('acodec') != 'none' and format.get('abr') is None:
1914 format['abr'] = format.get('tbr') - format.get('vbr', 0)
1916 return tuple(self._calculate_field_preference(format, field) for field in self._order)
1918 def _sort_formats(self, formats, field_preference=[]):
1921 formats.sort(key=self.FormatSort(self, field_preference).calculate_preference)
1923 def _check_formats(self, formats, video_id):
1925 formats[:] = filter(
1926 lambda f: self._is_valid_url(
1928 item='%s video format' % f.get('format_id') if f.get('format_id') else 'video'),
1932 def _remove_duplicate_formats(formats):
1936 if f['url'] not in format_urls:
1937 format_urls.add(f['url'])
1938 unique_formats.append(f)
1939 formats[:] = unique_formats
1941 def _is_valid_url(self, url, video_id, item='video', headers={}):
1942 url = self._proto_relative_url(url, scheme='http:')
1943 # For now assume non HTTP(S) URLs always valid
1944 if not (url.startswith('http://') or url.startswith('https://')):
1947 self._request_webpage(url, video_id, 'Checking %s URL' % item, headers=headers)
1949 except ExtractorError as e:
1951 '%s: %s URL is invalid, skipping: %s'
1952 % (video_id, item, error_to_compat_str(e.cause)))
1955 def http_scheme(self):
1956 """ Either "http
:" or "https
:", depending on the user's preferences """
1959 if self.get_param('prefer_insecure', False)
1962 def _proto_relative_url(self, url, scheme=None):
1965 if url.startswith('//'):
1967 scheme = self.http_scheme()
1972 def _sleep(self, timeout, video_id, msg_template=None):
1973 if msg_template is None:
1974 msg_template = '%(video_id)s: Waiting for %(timeout)s seconds'
1975 msg = msg_template % {'video_id': video_id, 'timeout': timeout}
1979 def _extract_f4m_formats(self, manifest_url, video_id, preference=None, quality=None, f4m_id=None,
1980 transform_source=lambda s: fix_xml_ampersands(s).strip(),
1981 fatal=True, m3u8_id=None, data=None, headers={}, query={}):
1982 res = self._download_xml_handle(
1983 manifest_url, video_id, 'Downloading f4m manifest',
1984 'Unable to download f4m manifest',
1985 # Some manifests may be malformed, e.g. prosiebensat1 generated manifests
1986 # (see https://github.com/ytdl-org/youtube-dl/issues/6215#issuecomment-121704244)
1987 transform_source=transform_source,
1988 fatal=fatal, data=data, headers=headers, query=query)
1992 manifest, urlh = res
1993 manifest_url = urlh.geturl()
1995 return self._parse_f4m_formats(
1996 manifest, manifest_url, video_id, preference=preference, quality=quality, f4m_id=f4m_id,
1997 transform_source=transform_source, fatal=fatal, m3u8_id=m3u8_id)
1999 def _parse_f4m_formats(self, manifest, manifest_url, video_id, preference=None, quality=None, f4m_id=None,
2000 transform_source=lambda s: fix_xml_ampersands(s).strip(),
2001 fatal=True, m3u8_id=None):
2002 if not isinstance(manifest, xml.etree.ElementTree.Element) and not fatal:
2005 # currently yt-dlp cannot decode the playerVerificationChallenge as Akamai uses Adobe Alchemy
2006 akamai_pv = manifest.find('{http://ns.adobe.com/f4m/1.0}pv-2.0')
2007 if akamai_pv is not None and ';' in akamai_pv.text:
2008 playerVerificationChallenge = akamai_pv.text.split(';')[0]
2009 if playerVerificationChallenge.strip() != '':
2013 manifest_version = '1.0'
2014 media_nodes = manifest.findall('{http://ns.adobe.com/f4m/1.0}media')
2016 manifest_version = '2.0'
2017 media_nodes = manifest.findall('{http://ns.adobe.com/f4m/2.0}media')
2018 # Remove unsupported DRM protected media from final formats
2019 # rendition (see https://github.com/ytdl-org/youtube-dl/issues/8573).
2020 media_nodes = remove_encrypted_media(media_nodes)
2024 manifest_base_url = get_base_url(manifest)
2026 bootstrap_info = xpath_element(
2027 manifest, ['{http://ns.adobe.com/f4m/1.0}bootstrapInfo', '{http://ns.adobe.com/f4m/2.0}bootstrapInfo'],
2028 'bootstrap info', default=None)
2031 mime_type = xpath_text(
2032 manifest, ['{http://ns.adobe.com/f4m/1.0}mimeType', '{http://ns.adobe.com/f4m/2.0}mimeType'],
2033 'base URL', default=None)
2034 if mime_type and mime_type.startswith('audio/'):
2037 for i, media_el in enumerate(media_nodes):
2038 tbr = int_or_none(media_el.attrib.get('bitrate'))
2039 width = int_or_none(media_el.attrib.get('width'))
2040 height = int_or_none(media_el.attrib.get('height'))
2041 format_id = join_nonempty(f4m_id, tbr or i)
2042 # If <bootstrapInfo> is present, the specified f4m is a
2043 # stream-level manifest, and only set-level manifests may refer to
2044 # external resources. See section 11.4 and section 4 of F4M spec
2045 if bootstrap_info is None:
2047 # @href is introduced in 2.0, see section 11.6 of F4M spec
2048 if manifest_version == '2.0':
2049 media_url = media_el.attrib.get('href')
2050 if media_url is None:
2051 media_url = media_el.attrib.get('url')
2055 media_url if media_url.startswith('http://') or media_url.startswith('https://')
2056 else ((manifest_base_url or '/'.join(manifest_url.split('/')[:-1])) + '/' + media_url))
2057 # If media_url is itself a f4m manifest do the recursive extraction
2058 # since bitrates in parent manifest (this one) and media_url manifest
2059 # may differ leading to inability to resolve the format by requested
2060 # bitrate in f4m downloader
2061 ext = determine_ext(manifest_url)
2063 f4m_formats = self._extract_f4m_formats(
2064 manifest_url, video_id, preference=preference, quality=quality, f4m_id=f4m_id,
2065 transform_source=transform_source, fatal=fatal)
2066 # Sometimes stream-level manifest contains single media entry that
2067 # does not contain any quality metadata (e.g. http://matchtv.ru/#live-player).
2068 # At the same time parent's media entry in set-level manifest may
2069 # contain it. We will copy it from parent in such cases.
2070 if len(f4m_formats) == 1:
2073 'tbr': f.get('tbr') or tbr,
2074 'width': f.get('width') or width,
2075 'height': f.get('height') or height,
2076 'format_id': f.get('format_id') if not tbr else format_id,
2079 formats.extend(f4m_formats)
2082 formats.extend(self._extract_m3u8_formats(
2083 manifest_url, video_id, 'mp4', preference=preference,
2084 quality=quality, m3u8_id=m3u8_id, fatal=fatal))
2087 'format_id': format_id,
2088 'url': manifest_url,
2089 'manifest_url': manifest_url,
2090 'ext': 'flv' if bootstrap_info is not None else None,
2096 'preference': preference,
2101 def _m3u8_meta_format(self, m3u8_url, ext=None, preference=None, quality=None, m3u8_id=None):
2103 'format_id': join_nonempty(m3u8_id, 'meta'),
2107 'preference': preference - 100 if preference else -100,
2109 'resolution': 'multiple',
2110 'format_note': 'Quality selection URL',
2113 def _report_ignoring_subs(self, name):
2114 self.report_warning(bug_reports_message(
2115 f'Ignoring subtitle tracks found in the {name} manifest; '
2116 'if any subtitle tracks are missing,'
2119 def _extract_m3u8_formats(self, *args, **kwargs):
2120 fmts, subs = self._extract_m3u8_formats_and_subtitles(*args, **kwargs)
2122 self._report_ignoring_subs('HLS')
2125 def _extract_m3u8_formats_and_subtitles(
2126 self, m3u8_url, video_id, ext=None, entry_protocol='m3u8_native',
2127 preference=None, quality=None, m3u8_id=None, note=None,
2128 errnote=None, fatal=True, live=False, data=None, headers={},
2131 res = self._download_webpage_handle(
2133 note='Downloading m3u8 information' if note is None else note,
2134 errnote='Failed to download m3u8 information' if errnote is None else errnote,
2135 fatal=fatal, data=data, headers=headers, query=query)
2140 m3u8_doc, urlh = res
2141 m3u8_url = urlh.geturl()
2143 return self._parse_m3u8_formats_and_subtitles(
2144 m3u8_doc, m3u8_url, ext=ext, entry_protocol=entry_protocol,
2145 preference=preference, quality=quality, m3u8_id=m3u8_id,
2146 note=note, errnote=errnote, fatal=fatal, live=live, data=data,
2147 headers=headers, query=query, video_id=video_id)
2149 def _parse_m3u8_formats_and_subtitles(
2150 self, m3u8_doc, m3u8_url=None, ext=None, entry_protocol='m3u8_native',
2151 preference=None, quality=None, m3u8_id=None, live=False, note=None,
2152 errnote=None, fatal=True, data=None, headers={}, query={},
2154 formats, subtitles = [], {}
2156 has_drm = re.search('|'.join([
2157 r'#EXT-X-FAXS-CM:', # Adobe Flash Access
2158 r'#EXT-X-(?:SESSION-)?KEY:.*?URI="skd
://', # Apple FairPlay
2161 def format_url(url):
2162 return url if re.match(r'^https?
://', url) else urllib.parse.urljoin(m3u8_url, url)
2164 if self.get_param('hls_split_discontinuity
', False):
2165 def _extract_m3u8_playlist_indices(manifest_url=None, m3u8_doc=None):
2167 if not manifest_url:
2169 m3u8_doc = self._download_webpage(
2170 manifest_url, video_id, fatal=fatal, data=data, headers=headers,
2171 note=False, errnote='Failed to download m3u8 playlist information
')
2172 if m3u8_doc is False:
2174 return range(1 + sum(line.startswith('#EXT-X-DISCONTINUITY') for line in m3u8_doc.splitlines()))
2177 def _extract_m3u8_playlist_indices(*args
, **kwargs
):
2181 # 1. https://tools.ietf.org/html/draft-pantos-http-live-streaming-21
2182 # 2. https://github.com/ytdl-org/youtube-dl/issues/12211
2183 # 3. https://github.com/ytdl-org/youtube-dl/issues/18923
2185 # We should try extracting formats only from master playlists [1, 4.3.4],
2186 # i.e. playlists that describe available qualities. On the other hand
2187 # media playlists [1, 4.3.3] should be returned as is since they contain
2188 # just the media without qualities renditions.
2189 # Fortunately, master playlist can be easily distinguished from media
2190 # playlist based on particular tags availability. As of [1, 4.3.3, 4.3.4]
2191 # master playlist tags MUST NOT appear in a media playlist and vice versa.
2192 # As of [1, 4.3.3.1] #EXT-X-TARGETDURATION tag is REQUIRED for every
2193 # media playlist and MUST NOT appear in master playlist thus we can
2194 # clearly detect media playlist with this criterion.
2196 if '#EXT-X-TARGETDURATION' in m3u8_doc
: # media playlist, return as is
2198 'format_id': join_nonempty(m3u8_id
, idx
),
2199 'format_index': idx
,
2200 'url': m3u8_url
or encode_data_uri(m3u8_doc
.encode('utf-8'), 'application/x-mpegurl'),
2202 'protocol': entry_protocol
,
2203 'preference': preference
,
2206 } for idx
in _extract_m3u8_playlist_indices(m3u8_doc
=m3u8_doc
)]
2208 return formats
, subtitles
2211 last_stream_inf
= {}
2213 def extract_media(x_media_line
):
2214 media
= parse_m3u8_attributes(x_media_line
)
2215 # As per [1, 4.3.4.1] TYPE, GROUP-ID and NAME are REQUIRED
2216 media_type
, group_id
, name
= media
.get('TYPE'), media
.get('GROUP-ID'), media
.get('NAME')
2217 if not (media_type
and group_id
and name
):
2219 groups
.setdefault(group_id
, []).append(media
)
2220 # <https://tools.ietf.org/html/rfc8216#section-4.3.4.1>
2221 if media_type
== 'SUBTITLES':
2222 # According to RFC 8216 §4.3.4.2.1, URI is REQUIRED in the
2223 # EXT-X-MEDIA tag if the media type is SUBTITLES.
2224 # However, lack of URI has been spotted in the wild.
2225 # e.g. NebulaIE; see https://github.com/yt-dlp/yt-dlp/issues/339
2226 if not media
.get('URI'):
2228 url
= format_url(media
['URI'])
2231 'ext': determine_ext(url
),
2233 if sub_info
['ext'] == 'm3u8':
2234 # Per RFC 8216 §3.1, the only possible subtitle format m3u8
2235 # files may contain is WebVTT:
2236 # <https://tools.ietf.org/html/rfc8216#section-3.1>
2237 sub_info
['ext'] = 'vtt'
2238 sub_info
['protocol'] = 'm3u8_native'
2239 lang
= media
.get('LANGUAGE') or 'und'
2240 subtitles
.setdefault(lang
, []).append(sub_info
)
2241 if media_type
not in ('VIDEO', 'AUDIO'):
2243 media_url
= media
.get('URI')
2245 manifest_url
= format_url(media_url
)
2247 'format_id': join_nonempty(m3u8_id
, group_id
, name
, idx
),
2248 'format_note': name
,
2249 'format_index': idx
,
2250 'url': manifest_url
,
2251 'manifest_url': m3u8_url
,
2252 'language': media
.get('LANGUAGE'),
2254 'protocol': entry_protocol
,
2255 'preference': preference
,
2257 'vcodec': 'none' if media_type
== 'AUDIO' else None,
2258 } for idx
in _extract_m3u8_playlist_indices(manifest_url
))
2260 def build_stream_name():
2261 # Despite specification does not mention NAME attribute for
2262 # EXT-X-STREAM-INF tag it still sometimes may be present (see [1]
2263 # or vidio test in TestInfoExtractor.test_parse_m3u8_formats)
2264 # 1. http://www.vidio.com/watch/165683-dj_ambred-booyah-live-2015
2265 stream_name
= last_stream_inf
.get('NAME')
2268 # If there is no NAME in EXT-X-STREAM-INF it will be obtained
2269 # from corresponding rendition group
2270 stream_group_id
= last_stream_inf
.get('VIDEO')
2271 if not stream_group_id
:
2273 stream_group
= groups
.get(stream_group_id
)
2274 if not stream_group
:
2275 return stream_group_id
2276 rendition
= stream_group
[0]
2277 return rendition
.get('NAME') or stream_group_id
2279 # parse EXT-X-MEDIA tags before EXT-X-STREAM-INF in order to have the
2280 # chance to detect video only formats when EXT-X-STREAM-INF tags
2281 # precede EXT-X-MEDIA tags in HLS manifest such as [3].
2282 for line
in m3u8_doc
.splitlines():
2283 if line
.startswith('#EXT-X-MEDIA:'):
2286 for line
in m3u8_doc
.splitlines():
2287 if line
.startswith('#EXT-X-STREAM-INF:'):
2288 last_stream_inf
= parse_m3u8_attributes(line
)
2289 elif line
.startswith('#') or not line
.strip():
2292 tbr
= float_or_none(
2293 last_stream_inf
.get('AVERAGE-BANDWIDTH')
2294 or last_stream_inf
.get('BANDWIDTH'), scale
=1000)
2295 manifest_url
= format_url(line
.strip())
2297 for idx
in _extract_m3u8_playlist_indices(manifest_url
):
2298 format_id
= [m3u8_id
, None, idx
]
2299 # Bandwidth of live streams may differ over time thus making
2300 # format_id unpredictable. So it's better to keep provided
2303 stream_name
= build_stream_name()
2304 format_id
[1] = stream_name
or '%d' % (tbr
or len(formats
))
2306 'format_id': join_nonempty(*format_id
),
2307 'format_index': idx
,
2308 'url': manifest_url
,
2309 'manifest_url': m3u8_url
,
2312 'fps': float_or_none(last_stream_inf
.get('FRAME-RATE')),
2313 'protocol': entry_protocol
,
2314 'preference': preference
,
2317 resolution
= last_stream_inf
.get('RESOLUTION')
2319 mobj
= re
.search(r
'(?P<width>\d+)[xX](?P<height>\d+)', resolution
)
2321 f
['width'] = int(mobj
.group('width'))
2322 f
['height'] = int(mobj
.group('height'))
2323 # Unified Streaming Platform
2325 r
'audio.*?(?:%3D|=)(\d+)(?:-video.*?(?:%3D|=)(\d+))?', f
['url'])
2327 abr
, vbr
= mobj
.groups()
2328 abr
, vbr
= float_or_none(abr
, 1000), float_or_none(vbr
, 1000)
2333 codecs
= parse_codecs(last_stream_inf
.get('CODECS'))
2335 audio_group_id
= last_stream_inf
.get('AUDIO')
2336 # As per [1, 4.3.4.1.1] any EXT-X-STREAM-INF tag which
2337 # references a rendition group MUST have a CODECS attribute.
2338 # However, this is not always respected, for example, [2]
2339 # contains EXT-X-STREAM-INF tag which references AUDIO
2340 # rendition group but does not have CODECS and despite
2341 # referencing an audio group it represents a complete
2342 # (with audio and video) format. So, for such cases we will
2343 # ignore references to rendition groups and treat them
2344 # as complete formats.
2345 if audio_group_id
and codecs
and f
.get('vcodec') != 'none':
2346 audio_group
= groups
.get(audio_group_id
)
2347 if audio_group
and audio_group
[0].get('URI'):
2348 # TODO: update acodec for audio only formats with
2350 f
['acodec'] = 'none'
2351 if not f
.get('ext'):
2352 f
['ext'] = 'm4a' if f
.get('vcodec') == 'none' else 'mp4'
2356 progressive_uri
= last_stream_inf
.get('PROGRESSIVE-URI')
2359 del http_f
['manifest_url']
2361 'format_id': f
['format_id'].replace('hls-', 'http-'),
2363 'url': progressive_uri
,
2365 formats
.append(http_f
)
2367 last_stream_inf
= {}
2368 return formats
, subtitles
2370 def _extract_m3u8_vod_duration(
2371 self
, m3u8_vod_url
, video_id
, note
=None, errnote
=None, data
=None, headers
={}, query={}
):
2373 m3u8_vod
= self
._download
_webpage
(
2374 m3u8_vod_url
, video_id
,
2375 note
='Downloading m3u8 VOD manifest' if note
is None else note
,
2376 errnote
='Failed to download VOD manifest' if errnote
is None else errnote
,
2377 fatal
=False, data
=data
, headers
=headers
, query
=query
)
2379 return self
._parse
_m
3u8_vod
_duration
(m3u8_vod
or '', video_id
)
2381 def _parse_m3u8_vod_duration(self
, m3u8_vod
, video_id
):
2382 if '#EXT-X-PLAYLIST-TYPE:VOD' not in m3u8_vod
:
2386 float(line
[len('#EXTINF:'):].split(',')[0])
2387 for line
in m3u8_vod
.splitlines() if line
.startswith('#EXTINF:'))) or None
2390 def _xpath_ns(path
, namespace
=None):
2394 for c
in path
.split('/'):
2395 if not c
or c
== '.':
2398 out
.append('{%s}%s' % (namespace
, c
))
2399 return '/'.join(out
)
2401 def _extract_smil_formats_and_subtitles(self
, smil_url
, video_id
, fatal
=True, f4m_params
=None, transform_source
=None):
2402 res
= self
._download
_smil
(smil_url
, video_id
, fatal
=fatal
, transform_source
=transform_source
)
2408 smil_url
= urlh
.geturl()
2410 namespace
= self
._parse
_smil
_namespace
(smil
)
2412 fmts
= self
._parse
_smil
_formats
(
2413 smil
, smil_url
, video_id
, namespace
=namespace
, f4m_params
=f4m_params
)
2414 subs
= self
._parse
_smil
_subtitles
(
2415 smil
, namespace
=namespace
)
2419 def _extract_smil_formats(self
, *args
, **kwargs
):
2420 fmts
, subs
= self
._extract
_smil
_formats
_and
_subtitles
(*args
, **kwargs
)
2422 self
._report
_ignoring
_subs
('SMIL')
2425 def _extract_smil_info(self
, smil_url
, video_id
, fatal
=True, f4m_params
=None):
2426 res
= self
._download
_smil
(smil_url
, video_id
, fatal
=fatal
)
2431 smil_url
= urlh
.geturl()
2433 return self
._parse
_smil
(smil
, smil_url
, video_id
, f4m_params
=f4m_params
)
2435 def _download_smil(self
, smil_url
, video_id
, fatal
=True, transform_source
=None):
2436 return self
._download
_xml
_handle
(
2437 smil_url
, video_id
, 'Downloading SMIL file',
2438 'Unable to download SMIL file', fatal
=fatal
, transform_source
=transform_source
)
2440 def _parse_smil(self
, smil
, smil_url
, video_id
, f4m_params
=None):
2441 namespace
= self
._parse
_smil
_namespace
(smil
)
2443 formats
= self
._parse
_smil
_formats
(
2444 smil
, smil_url
, video_id
, namespace
=namespace
, f4m_params
=f4m_params
)
2445 subtitles
= self
._parse
_smil
_subtitles
(smil
, namespace
=namespace
)
2447 video_id
= os
.path
.splitext(url_basename(smil_url
))[0]
2451 for meta
in smil
.findall(self
._xpath
_ns
('./head/meta', namespace
)):
2452 name
= meta
.attrib
.get('name')
2453 content
= meta
.attrib
.get('content')
2454 if not name
or not content
:
2456 if not title
and name
== 'title':
2458 elif not description
and name
in ('description', 'abstract'):
2459 description
= content
2460 elif not upload_date
and name
== 'date':
2461 upload_date
= unified_strdate(content
)
2464 'id': image
.get('type'),
2465 'url': image
.get('src'),
2466 'width': int_or_none(image
.get('width')),
2467 'height': int_or_none(image
.get('height')),
2468 } for image
in smil
.findall(self
._xpath
_ns
('.//image', namespace
)) if image
.get('src')]
2472 'title': title
or video_id
,
2473 'description': description
,
2474 'upload_date': upload_date
,
2475 'thumbnails': thumbnails
,
2477 'subtitles': subtitles
,
2480 def _parse_smil_namespace(self
, smil
):
2481 return self
._search
_regex
(
2482 r
'(?i)^{([^}]+)?}smil$', smil
.tag
, 'namespace', default
=None)
2484 def _parse_smil_formats(self
, smil
, smil_url
, video_id
, namespace
=None, f4m_params
=None, transform_rtmp_url
=None):
2486 for meta
in smil
.findall(self
._xpath
_ns
('./head/meta', namespace
)):
2487 b
= meta
.get('base') or meta
.get('httpBase')
2499 media
= smil
.findall(self
._xpath
_ns
('.//video', namespace
)) + smil
.findall(self
._xpath
_ns
('.//audio', namespace
))
2500 for medium
in media
:
2501 src
= medium
.get('src')
2502 if not src
or src
in srcs
:
2506 bitrate
= float_or_none(medium
.get('system-bitrate') or medium
.get('systemBitrate'), 1000)
2507 filesize
= int_or_none(medium
.get('size') or medium
.get('fileSize'))
2508 width
= int_or_none(medium
.get('width'))
2509 height
= int_or_none(medium
.get('height'))
2510 proto
= medium
.get('proto')
2511 ext
= medium
.get('ext')
2512 src_ext
= determine_ext(src
)
2513 streamer
= medium
.get('streamer') or base
2515 if proto
== 'rtmp' or streamer
.startswith('rtmp'):
2521 'format_id': 'rtmp-%d' % (rtmp_count
if bitrate
is None else bitrate
),
2523 'filesize': filesize
,
2527 if transform_rtmp_url
:
2528 streamer
, src
= transform_rtmp_url(streamer
, src
)
2529 formats
[-1].update({
2535 src_url
= src
if src
.startswith('http') else urllib
.parse
.urljoin(base
, src
)
2536 src_url
= src_url
.strip()
2538 if proto
== 'm3u8' or src_ext
== 'm3u8':
2539 m3u8_formats
= self
._extract
_m
3u8_formats
(
2540 src_url
, video_id
, ext
or 'mp4', m3u8_id
='hls', fatal
=False)
2541 if len(m3u8_formats
) == 1:
2543 m3u8_formats
[0].update({
2544 'format_id': 'hls-%d' % (m3u8_count
if bitrate
is None else bitrate
),
2549 formats
.extend(m3u8_formats
)
2550 elif src_ext
== 'f4m':
2555 'plugin': 'flowplayer-3.2.0.1',
2557 f4m_url
+= '&' if '?' in f4m_url
else '?'
2558 f4m_url
+= urllib
.parse
.urlencode(f4m_params
)
2559 formats
.extend(self
._extract
_f
4m
_formats
(f4m_url
, video_id
, f4m_id
='hds', fatal
=False))
2560 elif src_ext
== 'mpd':
2561 formats
.extend(self
._extract
_mpd
_formats
(
2562 src_url
, video_id
, mpd_id
='dash', fatal
=False))
2563 elif re
.search(r
'\.ism/[Mm]anifest', src_url
):
2564 formats
.extend(self
._extract
_ism
_formats
(
2565 src_url
, video_id
, ism_id
='mss', fatal
=False))
2566 elif src_url
.startswith('http') and self
._is
_valid
_url
(src
, video_id
):
2570 'ext': ext
or src_ext
or 'flv',
2571 'format_id': 'http-%d' % (bitrate
or http_count
),
2573 'filesize': filesize
,
2578 for medium
in smil
.findall(self
._xpath
_ns
('.//imagestream', namespace
)):
2579 src
= medium
.get('src')
2580 if not src
or src
in srcs
:
2586 'format_id': 'imagestream-%d' % (imgs_count
),
2588 'ext': mimetype2ext(medium
.get('type')),
2591 'width': int_or_none(medium
.get('width')),
2592 'height': int_or_none(medium
.get('height')),
2593 'format_note': 'SMIL storyboards',
2598 def _parse_smil_subtitles(self
, smil
, namespace
=None, subtitles_lang
='en'):
2601 for num
, textstream
in enumerate(smil
.findall(self
._xpath
_ns
('.//textstream', namespace
))):
2602 src
= textstream
.get('src')
2603 if not src
or src
in urls
:
2606 ext
= textstream
.get('ext') or mimetype2ext(textstream
.get('type')) or determine_ext(src
)
2607 lang
= textstream
.get('systemLanguage') or textstream
.get('systemLanguageName') or textstream
.get('lang') or subtitles_lang
2608 subtitles
.setdefault(lang
, []).append({
2614 def _extract_xspf_playlist(self
, xspf_url
, playlist_id
, fatal
=True):
2615 res
= self
._download
_xml
_handle
(
2616 xspf_url
, playlist_id
, 'Downloading xpsf playlist',
2617 'Unable to download xspf manifest', fatal
=fatal
)
2622 xspf_url
= urlh
.geturl()
2624 return self
._parse
_xspf
(
2625 xspf
, playlist_id
, xspf_url
=xspf_url
,
2626 xspf_base_url
=base_url(xspf_url
))
2628 def _parse_xspf(self
, xspf_doc
, playlist_id
, xspf_url
=None, xspf_base_url
=None):
2630 'xspf': 'http://xspf.org/ns/0/',
2631 's1': 'http://static.streamone.nl/player/ns/0',
2635 for track
in xspf_doc
.findall(xpath_with_ns('./xspf:trackList/xspf:track', NS_MAP
)):
2637 track
, xpath_with_ns('./xspf:title', NS_MAP
), 'title', default
=playlist_id
)
2638 description
= xpath_text(
2639 track
, xpath_with_ns('./xspf:annotation', NS_MAP
), 'description')
2640 thumbnail
= xpath_text(
2641 track
, xpath_with_ns('./xspf:image', NS_MAP
), 'thumbnail')
2642 duration
= float_or_none(
2643 xpath_text(track
, xpath_with_ns('./xspf:duration', NS_MAP
), 'duration'), 1000)
2646 for location
in track
.findall(xpath_with_ns('./xspf:location', NS_MAP
)):
2647 format_url
= urljoin(xspf_base_url
, location
.text
)
2652 'manifest_url': xspf_url
,
2653 'format_id': location
.get(xpath_with_ns('s1:label', NS_MAP
)),
2654 'width': int_or_none(location
.get(xpath_with_ns('s1:width', NS_MAP
))),
2655 'height': int_or_none(location
.get(xpath_with_ns('s1:height', NS_MAP
))),
2657 self
._sort
_formats
(formats
)
2662 'description': description
,
2663 'thumbnail': thumbnail
,
2664 'duration': duration
,
2669 def _extract_mpd_formats(self
, *args
, **kwargs
):
2670 fmts
, subs
= self
._extract
_mpd
_formats
_and
_subtitles
(*args
, **kwargs
)
2672 self
._report
_ignoring
_subs
('DASH')
2675 def _extract_mpd_formats_and_subtitles(
2676 self
, mpd_url
, video_id
, mpd_id
=None, note
=None, errnote
=None,
2677 fatal
=True, data
=None, headers
={}, query={}
):
2678 res
= self
._download
_xml
_handle
(
2680 note
='Downloading MPD manifest' if note
is None else note
,
2681 errnote
='Failed to download MPD manifest' if errnote
is None else errnote
,
2682 fatal
=fatal
, data
=data
, headers
=headers
, query
=query
)
2689 # We could have been redirected to a new url when we retrieved our mpd file.
2690 mpd_url
= urlh
.geturl()
2691 mpd_base_url
= base_url(mpd_url
)
2693 return self
._parse
_mpd
_formats
_and
_subtitles
(
2694 mpd_doc
, mpd_id
, mpd_base_url
, mpd_url
)
2696 def _parse_mpd_formats(self
, *args
, **kwargs
):
2697 fmts
, subs
= self
._parse
_mpd
_formats
_and
_subtitles
(*args
, **kwargs
)
2699 self
._report
_ignoring
_subs
('DASH')
2702 def _parse_mpd_formats_and_subtitles(
2703 self
, mpd_doc
, mpd_id
=None, mpd_base_url
='', mpd_url
=None):
2705 Parse formats from MPD manifest.
2707 1. MPEG-DASH Standard, ISO/IEC 23009-1:2014(E),
2708 http://standards.iso.org/ittf/PubliclyAvailableStandards/c065274_ISO_IEC_23009-1_2014.zip
2709 2. https://en.wikipedia.org/wiki/Dynamic_Adaptive_Streaming_over_HTTP
2711 if not self
.get_param('dynamic_mpd', True):
2712 if mpd_doc
.get('type') == 'dynamic':
2715 namespace
= self
._search
_regex
(r
'(?i)^{([^}]+)?}MPD$', mpd_doc
.tag
, 'namespace', default
=None)
2718 return self
._xpath
_ns
(path
, namespace
)
2720 def is_drm_protected(element
):
2721 return element
.find(_add_ns('ContentProtection')) is not None
2723 def extract_multisegment_info(element
, ms_parent_info
):
2724 ms_info
= ms_parent_info
.copy()
2726 # As per [1, 5.3.9.2.2] SegmentList and SegmentTemplate share some
2727 # common attributes and elements. We will only extract relevant
2729 def extract_common(source
):
2730 segment_timeline
= source
.find(_add_ns('SegmentTimeline'))
2731 if segment_timeline
is not None:
2732 s_e
= segment_timeline
.findall(_add_ns('S'))
2734 ms_info
['total_number'] = 0
2737 r
= int(s
.get('r', 0))
2738 ms_info
['total_number'] += 1 + r
2739 ms_info
['s'].append({
2740 't': int(s
.get('t', 0)),
2741 # @d is mandatory (see [1, 5.3.9.6.2, Table 17, page 60])
2742 'd': int(s
.attrib
['d']),
2745 start_number
= source
.get('startNumber')
2747 ms_info
['start_number'] = int(start_number
)
2748 timescale
= source
.get('timescale')
2750 ms_info
['timescale'] = int(timescale
)
2751 segment_duration
= source
.get('duration')
2752 if segment_duration
:
2753 ms_info
['segment_duration'] = float(segment_duration
)
2755 def extract_Initialization(source
):
2756 initialization
= source
.find(_add_ns('Initialization'))
2757 if initialization
is not None:
2758 ms_info
['initialization_url'] = initialization
.attrib
['sourceURL']
2760 segment_list
= element
.find(_add_ns('SegmentList'))
2761 if segment_list
is not None:
2762 extract_common(segment_list
)
2763 extract_Initialization(segment_list
)
2764 segment_urls_e
= segment_list
.findall(_add_ns('SegmentURL'))
2766 ms_info
['segment_urls'] = [segment
.attrib
['media'] for segment
in segment_urls_e
]
2768 segment_template
= element
.find(_add_ns('SegmentTemplate'))
2769 if segment_template
is not None:
2770 extract_common(segment_template
)
2771 media
= segment_template
.get('media')
2773 ms_info
['media'] = media
2774 initialization
= segment_template
.get('initialization')
2776 ms_info
['initialization'] = initialization
2778 extract_Initialization(segment_template
)
2781 mpd_duration
= parse_duration(mpd_doc
.get('mediaPresentationDuration'))
2782 formats
, subtitles
= [], {}
2783 stream_numbers
= collections
.defaultdict(int)
2784 for period
in mpd_doc
.findall(_add_ns('Period')):
2785 period_duration
= parse_duration(period
.get('duration')) or mpd_duration
2786 period_ms_info
= extract_multisegment_info(period
, {
2790 for adaptation_set
in period
.findall(_add_ns('AdaptationSet')):
2791 adaption_set_ms_info
= extract_multisegment_info(adaptation_set
, period_ms_info
)
2792 for representation
in adaptation_set
.findall(_add_ns('Representation')):
2793 representation_attrib
= adaptation_set
.attrib
.copy()
2794 representation_attrib
.update(representation
.attrib
)
2795 # According to [1, 5.3.7.2, Table 9, page 41], @mimeType is mandatory
2796 mime_type
= representation_attrib
['mimeType']
2797 content_type
= representation_attrib
.get('contentType', mime_type
.split('/')[0])
2799 codec_str
= representation_attrib
.get('codecs', '')
2800 # Some kind of binary subtitle found in some youtube livestreams
2801 if mime_type
== 'application/x-rawcc':
2802 codecs
= {'scodec': codec_str}
2804 codecs
= parse_codecs(codec_str
)
2805 if content_type
not in ('video', 'audio', 'text'):
2806 if mime_type
== 'image/jpeg':
2807 content_type
= mime_type
2808 elif codecs
.get('vcodec', 'none') != 'none':
2809 content_type
= 'video'
2810 elif codecs
.get('acodec', 'none') != 'none':
2811 content_type
= 'audio'
2812 elif codecs
.get('scodec', 'none') != 'none':
2813 content_type
= 'text'
2814 elif mimetype2ext(mime_type
) in ('tt', 'dfxp', 'ttml', 'xml', 'json'):
2815 content_type
= 'text'
2817 self
.report_warning('Unknown MIME type %s in DASH manifest' % mime_type
)
2821 for element
in (representation
, adaptation_set
, period
, mpd_doc
):
2822 base_url_e
= element
.find(_add_ns('BaseURL'))
2823 if base_url_e
is not None:
2824 base_url
= base_url_e
.text
+ base_url
2825 if re
.match(r
'^https?://', base_url
):
2827 if mpd_base_url
and base_url
.startswith('/'):
2828 base_url
= urllib
.parse
.urljoin(mpd_base_url
, base_url
)
2829 elif mpd_base_url
and not re
.match(r
'^https?://', base_url
):
2830 if not mpd_base_url
.endswith('/'):
2832 base_url
= mpd_base_url
+ base_url
2833 representation_id
= representation_attrib
.get('id')
2834 lang
= representation_attrib
.get('lang')
2835 url_el
= representation
.find(_add_ns('BaseURL'))
2836 filesize
= int_or_none(url_el
.attrib
.get('{http://youtube.com/yt/2012/10/10}contentLength') if url_el
is not None else None)
2837 bandwidth
= int_or_none(representation_attrib
.get('bandwidth'))
2838 if representation_id
is not None:
2839 format_id
= representation_id
2841 format_id
= content_type
2843 format_id
= mpd_id
+ '-' + format_id
2844 if content_type
in ('video', 'audio'):
2846 'format_id': format_id
,
2847 'manifest_url': mpd_url
,
2848 'ext': mimetype2ext(mime_type
),
2849 'width': int_or_none(representation_attrib
.get('width')),
2850 'height': int_or_none(representation_attrib
.get('height')),
2851 'tbr': float_or_none(bandwidth
, 1000),
2852 'asr': int_or_none(representation_attrib
.get('audioSamplingRate')),
2853 'fps': int_or_none(representation_attrib
.get('frameRate')),
2854 'language': lang
if lang
not in ('mul', 'und', 'zxx', 'mis') else None,
2855 'format_note': 'DASH %s' % content_type
,
2856 'filesize': filesize
,
2857 'container': mimetype2ext(mime_type
) + '_dash',
2860 elif content_type
== 'text':
2862 'ext': mimetype2ext(mime_type
),
2863 'manifest_url': mpd_url
,
2864 'filesize': filesize
,
2866 elif content_type
== 'image/jpeg':
2867 # See test case in VikiIE
2868 # https://www.viki.com/videos/1175236v-choosing-spouse-by-lottery-episode-1
2870 'format_id': format_id
,
2872 'manifest_url': mpd_url
,
2873 'format_note': 'DASH storyboards (jpeg)',
2877 if is_drm_protected(adaptation_set
) or is_drm_protected(representation
):
2879 representation_ms_info
= extract_multisegment_info(representation
, adaption_set_ms_info
)
2881 def prepare_template(template_name
, identifiers
):
2882 tmpl
= representation_ms_info
[template_name
]
2883 # First of, % characters outside $...$ templates
2884 # must be escaped by doubling for proper processing
2885 # by % operator string formatting used further (see
2886 # https://github.com/ytdl-org/youtube-dl/issues/16867).
2892 in_template
= not in_template
2893 elif c
== '%' and not in_template
:
2895 # Next, $...$ templates are translated to their
2896 # %(...) counterparts to be used with % operator
2897 if representation_id
is not None:
2898 t
= t
.replace('$RepresentationID$', representation_id
)
2899 t
= re
.sub(r
'\$(%s)\$' % '|'.join(identifiers
), r
'%(\1)d', t
)
2900 t
= re
.sub(r
'\$(%s)%%([^$]+)\$' % '|'.join(identifiers
), r
'%(\1)\2', t
)
2901 t
.replace('$$', '$')
2904 # @initialization is a regular template like @media one
2905 # so it should be handled just the same way (see
2906 # https://github.com/ytdl-org/youtube-dl/issues/11605)
2907 if 'initialization' in representation_ms_info
:
2908 initialization_template
= prepare_template(
2910 # As per [1, 5.3.9.4.2, Table 15, page 54] $Number$ and
2911 # $Time$ shall not be included for @initialization thus
2912 # only $Bandwidth$ remains
2914 representation_ms_info
['initialization_url'] = initialization_template
% {
2915 'Bandwidth': bandwidth
,
2918 def location_key(location
):
2919 return 'url' if re
.match(r
'^https?://', location
) else 'path'
2921 if 'segment_urls' not in representation_ms_info
and 'media' in representation_ms_info
:
2923 media_template
= prepare_template('media', ('Number', 'Bandwidth', 'Time'))
2924 media_location_key
= location_key(media_template
)
2926 # As per [1, 5.3.9.4.4, Table 16, page 55] $Number$ and $Time$
2927 # can't be used at the same time
2928 if '%(Number' in media_template
and 's' not in representation_ms_info
:
2929 segment_duration
= None
2930 if 'total_number' not in representation_ms_info
and 'segment_duration' in representation_ms_info
:
2931 segment_duration
= float_or_none(representation_ms_info
['segment_duration'], representation_ms_info
['timescale'])
2932 representation_ms_info
['total_number'] = int(math
.ceil(
2933 float_or_none(period_duration
, segment_duration
, default
=0)))
2934 representation_ms_info
['fragments'] = [{
2935 media_location_key
: media_template
% {
2936 'Number': segment_number
,
2937 'Bandwidth': bandwidth
,
2939 'duration': segment_duration
,
2940 } for segment_number
in range(
2941 representation_ms_info
['start_number'],
2942 representation_ms_info
['total_number'] + representation_ms_info
['start_number'])]
2944 # $Number*$ or $Time$ in media template with S list available
2945 # Example $Number*$: http://www.svtplay.se/klipp/9023742/stopptid-om-bjorn-borg
2946 # Example $Time$: https://play.arkena.com/embed/avp/v2/player/media/b41dda37-d8e7-4d3f-b1b5-9a9db578bdfe/1/129411
2947 representation_ms_info
['fragments'] = []
2950 segment_number
= representation_ms_info
['start_number']
2952 def add_segment_url():
2953 segment_url
= media_template
% {
2954 'Time': segment_time
,
2955 'Bandwidth': bandwidth
,
2956 'Number': segment_number
,
2958 representation_ms_info
['fragments'].append({
2959 media_location_key
: segment_url
,
2960 'duration': float_or_none(segment_d
, representation_ms_info
['timescale']),
2963 for num
, s
in enumerate(representation_ms_info
['s']):
2964 segment_time
= s
.get('t') or segment_time
2968 for r
in range(s
.get('r', 0)):
2969 segment_time
+= segment_d
2972 segment_time
+= segment_d
2973 elif 'segment_urls' in representation_ms_info
and 's' in representation_ms_info
:
2975 # Example: https://www.youtube.com/watch?v=iXZV5uAYMJI
2976 # or any YouTube dashsegments video
2979 timescale
= representation_ms_info
['timescale']
2980 for s
in representation_ms_info
['s']:
2981 duration
= float_or_none(s
['d'], timescale
)
2982 for r
in range(s
.get('r', 0) + 1):
2983 segment_uri
= representation_ms_info
['segment_urls'][segment_index
]
2985 location_key(segment_uri
): segment_uri
,
2986 'duration': duration
,
2989 representation_ms_info
['fragments'] = fragments
2990 elif 'segment_urls' in representation_ms_info
:
2991 # Segment URLs with no SegmentTimeline
2992 # Example: https://www.seznam.cz/zpravy/clanek/cesko-zasahne-vitr-o-sile-vichrice-muze-byt-i-zivotu-nebezpecny-39091
2993 # https://github.com/ytdl-org/youtube-dl/pull/14844
2995 segment_duration
= float_or_none(
2996 representation_ms_info
['segment_duration'],
2997 representation_ms_info
['timescale']) if 'segment_duration' in representation_ms_info
else None
2998 for segment_url
in representation_ms_info
['segment_urls']:
3000 location_key(segment_url
): segment_url
,
3002 if segment_duration
:
3003 fragment
['duration'] = segment_duration
3004 fragments
.append(fragment
)
3005 representation_ms_info
['fragments'] = fragments
3006 # If there is a fragments key available then we correctly recognized fragmented media.
3007 # Otherwise we will assume unfragmented media with direct access. Technically, such
3008 # assumption is not necessarily correct since we may simply have no support for
3009 # some forms of fragmented media renditions yet, but for now we'll use this fallback.
3010 if 'fragments' in representation_ms_info
:
3012 # NB: mpd_url may be empty when MPD manifest is parsed from a string
3013 'url': mpd_url
or base_url
,
3014 'fragment_base_url': base_url
,
3016 'protocol': 'http_dash_segments' if mime_type
!= 'image/jpeg' else 'mhtml',
3018 if 'initialization_url' in representation_ms_info
:
3019 initialization_url
= representation_ms_info
['initialization_url']
3020 if not f
.get('url'):
3021 f
['url'] = initialization_url
3022 f
['fragments'].append({location_key(initialization_url): initialization_url}
)
3023 f
['fragments'].extend(representation_ms_info
['fragments'])
3024 if not period_duration
:
3025 period_duration
= try_get(
3026 representation_ms_info
,
3027 lambda r
: sum(frag
['duration'] for frag
in r
['fragments']), float)
3029 # Assuming direct URL to unfragmented media.
3031 if content_type
in ('video', 'audio', 'image/jpeg'):
3032 f
['manifest_stream_number'] = stream_numbers
[f
['url']]
3033 stream_numbers
[f
['url']] += 1
3035 elif content_type
== 'text':
3036 subtitles
.setdefault(lang
or 'und', []).append(f
)
3038 return formats
, subtitles
3040 def _extract_ism_formats(self
, *args
, **kwargs
):
3041 fmts
, subs
= self
._extract
_ism
_formats
_and
_subtitles
(*args
, **kwargs
)
3043 self
._report
_ignoring
_subs
('ISM')
3046 def _extract_ism_formats_and_subtitles(self
, ism_url
, video_id
, ism_id
=None, note
=None, errnote
=None, fatal
=True, data
=None, headers
={}, query={}
):
3047 res
= self
._download
_xml
_handle
(
3049 note
='Downloading ISM manifest' if note
is None else note
,
3050 errnote
='Failed to download ISM manifest' if errnote
is None else errnote
,
3051 fatal
=fatal
, data
=data
, headers
=headers
, query
=query
)
3058 return self
._parse
_ism
_formats
_and
_subtitles
(ism_doc
, urlh
.geturl(), ism_id
)
3060 def _parse_ism_formats_and_subtitles(self
, ism_doc
, ism_url
, ism_id
=None):
3062 Parse formats from ISM manifest.
3064 1. [MS-SSTR]: Smooth Streaming Protocol,
3065 https://msdn.microsoft.com/en-us/library/ff469518.aspx
3067 if ism_doc
.get('IsLive') == 'TRUE':
3070 duration
= int(ism_doc
.attrib
['Duration'])
3071 timescale
= int_or_none(ism_doc
.get('TimeScale')) or 10000000
3075 for stream
in ism_doc
.findall('StreamIndex'):
3076 stream_type
= stream
.get('Type')
3077 if stream_type
not in ('video', 'audio', 'text'):
3079 url_pattern
= stream
.attrib
['Url']
3080 stream_timescale
= int_or_none(stream
.get('TimeScale')) or timescale
3081 stream_name
= stream
.get('Name')
3082 stream_language
= stream
.get('Language', 'und')
3083 for track
in stream
.findall('QualityLevel'):
3084 fourcc
= track
.get('FourCC') or ('AACL' if track
.get('AudioTag') == '255' else None)
3085 # TODO: add support for WVC1 and WMAP
3086 if fourcc
not in ('H264', 'AVC1', 'AACL', 'TTML'):
3087 self
.report_warning('%s is not a supported codec' % fourcc
)
3089 tbr
= int(track
.attrib
['Bitrate']) // 1000
3090 # [1] does not mention Width and Height attributes. However,
3091 # they're often present while MaxWidth and MaxHeight are
3092 # missing, so should be used as fallbacks
3093 width
= int_or_none(track
.get('MaxWidth') or track
.get('Width'))
3094 height
= int_or_none(track
.get('MaxHeight') or track
.get('Height'))
3095 sampling_rate
= int_or_none(track
.get('SamplingRate'))
3097 track_url_pattern
= re
.sub(r
'{[Bb]itrate}', track
.attrib
['Bitrate'], url_pattern
)
3098 track_url_pattern
= urllib
.parse
.urljoin(ism_url
, track_url_pattern
)
3104 stream_fragments
= stream
.findall('c')
3105 for stream_fragment_index
, stream_fragment
in enumerate(stream_fragments
):
3106 fragment_ctx
['time'] = int_or_none(stream_fragment
.get('t')) or fragment_ctx
['time']
3107 fragment_repeat
= int_or_none(stream_fragment
.get('r')) or 1
3108 fragment_ctx
['duration'] = int_or_none(stream_fragment
.get('d'))
3109 if not fragment_ctx
['duration']:
3111 next_fragment_time
= int(stream_fragment
[stream_fragment_index
+ 1].attrib
['t'])
3113 next_fragment_time
= duration
3114 fragment_ctx
['duration'] = (next_fragment_time
- fragment_ctx
['time']) / fragment_repeat
3115 for _
in range(fragment_repeat
):
3117 'url': re
.sub(r
'{start[ _]time}', str(fragment_ctx
['time']), track_url_pattern
),
3118 'duration': fragment_ctx
['duration'] / stream_timescale
,
3120 fragment_ctx
['time'] += fragment_ctx
['duration']
3122 if stream_type
== 'text':
3123 subtitles
.setdefault(stream_language
, []).append({
3127 'manifest_url': ism_url
,
3128 'fragments': fragments
,
3129 '_download_params': {
3130 'stream_type': stream_type
,
3131 'duration': duration
,
3132 'timescale': stream_timescale
,
3134 'language': stream_language
,
3135 'codec_private_data': track
.get('CodecPrivateData'),
3138 elif stream_type
in ('video', 'audio'):
3140 'format_id': join_nonempty(ism_id
, stream_name
, tbr
),
3142 'manifest_url': ism_url
,
3143 'ext': 'ismv' if stream_type
== 'video' else 'isma',
3147 'asr': sampling_rate
,
3148 'vcodec': 'none' if stream_type
== 'audio' else fourcc
,
3149 'acodec': 'none' if stream_type
== 'video' else fourcc
,
3151 'fragments': fragments
,
3152 'has_drm': ism_doc
.find('Protection') is not None,
3153 '_download_params': {
3154 'stream_type': stream_type
,
3155 'duration': duration
,
3156 'timescale': stream_timescale
,
3157 'width': width
or 0,
3158 'height': height
or 0,
3160 'language': stream_language
,
3161 'codec_private_data': track
.get('CodecPrivateData'),
3162 'sampling_rate': sampling_rate
,
3163 'channels': int_or_none(track
.get('Channels', 2)),
3164 'bits_per_sample': int_or_none(track
.get('BitsPerSample', 16)),
3165 'nal_unit_length_field': int_or_none(track
.get('NALUnitLengthField', 4)),
3168 return formats
, subtitles
3170 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):
3171 def absolute_url(item_url
):
3172 return urljoin(base_url
, item_url
)
3174 def parse_content_type(content_type
):
3175 if not content_type
:
3177 ctr
= re
.search(r
'(?P<mimetype>[^/]+/[^;]+)(?:;\s*codecs="?(?P<codecs>[^"]+))?', content_type
)
3179 mimetype
, codecs
= ctr
.groups()
3180 f
= parse_codecs(codecs
)
3181 f
['ext'] = mimetype2ext(mimetype
)
3185 def _media_formats(src
, cur_media_type
, type_info
=None):
3186 type_info
= type_info
or {}
3187 full_url
= absolute_url(src
)
3188 ext
= type_info
.get('ext') or determine_ext(full_url
)
3190 is_plain_url
= False
3191 formats
= self
._extract
_m
3u8_formats
(
3192 full_url
, video_id
, ext
='mp4',
3193 entry_protocol
=m3u8_entry_protocol
, m3u8_id
=m3u8_id
,
3194 preference
=preference
, quality
=quality
, fatal
=False)
3196 is_plain_url
= False
3197 formats
= self
._extract
_mpd
_formats
(
3198 full_url
, video_id
, mpd_id
=mpd_id
, fatal
=False)
3203 'vcodec': 'none' if cur_media_type
== 'audio' else None,
3206 return is_plain_url
, formats
3209 # amp-video and amp-audio are very similar to their HTML5 counterparts
3210 # so we wll include them right here (see
3211 # https://www.ampproject.org/docs/reference/components/amp-video)
3212 # For dl8-* tags see https://delight-vr.com/documentation/dl8-video/
3213 _MEDIA_TAG_NAME_RE
= r
'(?:(?:amp|dl8(?:-live)?)-)?(video|audio)'
3214 media_tags
= [(media_tag
, media_tag_name
, media_type
, '')
3215 for media_tag
, media_tag_name
, media_type
3216 in re
.findall(r
'(?s)(<(%s)[^>]*/>)' % _MEDIA_TAG_NAME_RE
, webpage
)]
3217 media_tags
.extend(re
.findall(
3218 # We only allow video|audio followed by a whitespace or '>'.
3219 # Allowing more characters may end up in significant slow down (see
3220 # https://github.com/ytdl-org/youtube-dl/issues/11979, example URL:
3221 # http://www.porntrex.com/maps/videositemap.xml).
3222 r
'(?s)(<(?P<tag>%s)(?:\s+[^>]*)?>)(.*?)</(?P=tag)>' % _MEDIA_TAG_NAME_RE
, webpage
))
3223 for media_tag
, _
, media_type
, media_content
in media_tags
:
3228 media_attributes
= extract_attributes(media_tag
)
3229 src
= strip_or_none(media_attributes
.get('src'))
3231 f
= parse_content_type(media_attributes
.get('type'))
3232 _
, formats
= _media_formats(src
, media_type
, f
)
3233 media_info
['formats'].extend(formats
)
3234 media_info
['thumbnail'] = absolute_url(media_attributes
.get('poster'))
3236 for source_tag
in re
.findall(r
'<source[^>]+>', media_content
):
3237 s_attr
= extract_attributes(source_tag
)
3238 # data-video-src and data-src are non standard but seen
3239 # several times in the wild
3240 src
= strip_or_none(dict_get(s_attr
, ('src', 'data-video-src', 'data-src')))
3243 f
= parse_content_type(s_attr
.get('type'))
3244 is_plain_url
, formats
= _media_formats(src
, media_type
, f
)
3246 # width, height, res, label and title attributes are
3247 # all not standard but seen several times in the wild
3250 for lbl
in ('label', 'title')
3251 if str_or_none(s_attr
.get(lbl
))
3253 width
= int_or_none(s_attr
.get('width'))
3254 height
= (int_or_none(s_attr
.get('height'))
3255 or int_or_none(s_attr
.get('res')))
3256 if not width
or not height
:
3258 resolution
= parse_resolution(lbl
)
3261 width
= width
or resolution
.get('width')
3262 height
= height
or resolution
.get('height')
3264 tbr
= parse_bitrate(lbl
)
3273 'format_id': s_attr
.get('label') or s_attr
.get('title'),
3275 f
.update(formats
[0])
3276 media_info
['formats'].append(f
)
3278 media_info
['formats'].extend(formats
)
3279 for track_tag
in re
.findall(r
'<track[^>]+>', media_content
):
3280 track_attributes
= extract_attributes(track_tag
)
3281 kind
= track_attributes
.get('kind')
3282 if not kind
or kind
in ('subtitles', 'captions'):
3283 src
= strip_or_none(track_attributes
.get('src'))
3286 lang
= track_attributes
.get('srclang') or track_attributes
.get('lang') or track_attributes
.get('label')
3287 media_info
['subtitles'].setdefault(lang
, []).append({
3288 'url': absolute_url(src
),
3290 for f
in media_info
['formats']:
3291 f
.setdefault('http_headers', {})['Referer'] = base_url
3292 if media_info
['formats'] or media_info
['subtitles']:
3293 entries
.append(media_info
)
3296 def _extract_akamai_formats(self
, *args
, **kwargs
):
3297 fmts
, subs
= self
._extract
_akamai
_formats
_and
_subtitles
(*args
, **kwargs
)
3299 self
._report
_ignoring
_subs
('akamai')
3302 def _extract_akamai_formats_and_subtitles(self
, manifest_url
, video_id
, hosts
={}):
3303 signed
= 'hdnea=' in manifest_url
3305 # https://learn.akamai.com/en-us/webhelp/media-services-on-demand/stream-packaging-user-guide/GUID-BE6C0F73-1E06-483B-B0EA-57984B91B7F9.html
3306 manifest_url
= re
.sub(
3307 r
'(?:b=[\d,-]+|(?:__a__|attributes)=off|__b__=\d+)&?',
3308 '', manifest_url
).strip('?')
3313 hdcore_sign
= 'hdcore=3.7.0'
3314 f4m_url
= re
.sub(r
'(https?://[^/]+)/i/', r
'\1/z/', manifest_url
).replace('/master.m3u8', '/manifest.f4m')
3315 hds_host
= hosts
.get('hds')
3317 f4m_url
= re
.sub(r
'(https?://)[^/]+', r
'\1' + hds_host
, f4m_url
)
3318 if 'hdcore=' not in f4m_url
:
3319 f4m_url
+= ('&' if '?' in f4m_url
else '?') + hdcore_sign
3320 f4m_formats
= self
._extract
_f
4m
_formats
(
3321 f4m_url
, video_id
, f4m_id
='hds', fatal
=False)
3322 for entry
in f4m_formats
:
3323 entry
.update({'extra_param_to_segment_url': hdcore_sign}
)
3324 formats
.extend(f4m_formats
)
3326 m3u8_url
= re
.sub(r
'(https?://[^/]+)/z/', r
'\1/i/', manifest_url
).replace('/manifest.f4m', '/master.m3u8')
3327 hls_host
= hosts
.get('hls')
3329 m3u8_url
= re
.sub(r
'(https?://)[^/]+', r
'\1' + hls_host
, m3u8_url
)
3330 m3u8_formats
, m3u8_subtitles
= self
._extract
_m
3u8_formats
_and
_subtitles
(
3331 m3u8_url
, video_id
, 'mp4', 'm3u8_native',
3332 m3u8_id
='hls', fatal
=False)
3333 formats
.extend(m3u8_formats
)
3334 subtitles
= self
._merge
_subtitles
(subtitles
, m3u8_subtitles
)
3336 http_host
= hosts
.get('http')
3337 if http_host
and m3u8_formats
and not signed
:
3338 REPL_REGEX
= r
'https?://[^/]+/i/([^,]+),([^/]+),([^/]+)\.csmil/.+'
3339 qualities
= re
.match(REPL_REGEX
, m3u8_url
).group(2).split(',')
3340 qualities_length
= len(qualities
)
3341 if len(m3u8_formats
) in (qualities_length
, qualities_length
+ 1):
3343 for f
in m3u8_formats
:
3344 if f
['vcodec'] != 'none':
3345 for protocol
in ('http', 'https'):
3347 del http_f
['manifest_url']
3349 REPL_REGEX
, protocol
+ fr
'://{http_host}/\g<1>{qualities[i]}\3', f
['url'])
3351 'format_id': http_f
['format_id'].replace('hls-', protocol
+ '-'),
3353 'protocol': protocol
,
3355 formats
.append(http_f
)
3358 return formats
, subtitles
3360 def _extract_wowza_formats(self
, url
, video_id
, m3u8_entry_protocol
='m3u8_native', skip_protocols
=[]):
3361 query
= urllib
.parse
.urlparse(url
).query
3362 url
= re
.sub(r
'/(?:manifest|playlist|jwplayer)\.(?:m3u8|f4m|mpd|smil)', '', url
)
3364 r
'(?:(?:http|rtmp|rtsp)(?P<s>s)?:)?(?P<url>//[^?]+)', url
)
3365 url_base
= mobj
.group('url')
3366 http_base_url
= '%s%s:%s' % ('http', mobj
.group('s') or '', url_base
)
3369 def manifest_url(manifest
):
3370 m_url
= f
'{http_base_url}/{manifest}'
3372 m_url
+= '?%s' % query
3375 if 'm3u8' not in skip_protocols
:
3376 formats
.extend(self
._extract
_m
3u8_formats
(
3377 manifest_url('playlist.m3u8'), video_id
, 'mp4',
3378 m3u8_entry_protocol
, m3u8_id
='hls', fatal
=False))
3379 if 'f4m' not in skip_protocols
:
3380 formats
.extend(self
._extract
_f
4m
_formats
(
3381 manifest_url('manifest.f4m'),
3382 video_id
, f4m_id
='hds', fatal
=False))
3383 if 'dash' not in skip_protocols
:
3384 formats
.extend(self
._extract
_mpd
_formats
(
3385 manifest_url('manifest.mpd'),
3386 video_id
, mpd_id
='dash', fatal
=False))
3387 if re
.search(r
'(?:/smil:|\.smil)', url_base
):
3388 if 'smil' not in skip_protocols
:
3389 rtmp_formats
= self
._extract
_smil
_formats
(
3390 manifest_url('jwplayer.smil'),
3391 video_id
, fatal
=False)
3392 for rtmp_format
in rtmp_formats
:
3393 rtsp_format
= rtmp_format
.copy()
3394 rtsp_format
['url'] = '%s/%s' % (rtmp_format
['url'], rtmp_format
['play_path'])
3395 del rtsp_format
['play_path']
3396 del rtsp_format
['ext']
3397 rtsp_format
.update({
3398 'url': rtsp_format
['url'].replace('rtmp://', 'rtsp://'),
3399 'format_id': rtmp_format
['format_id'].replace('rtmp', 'rtsp'),
3402 formats
.extend([rtmp_format
, rtsp_format
])
3404 for protocol
in ('rtmp', 'rtsp'):
3405 if protocol
not in skip_protocols
:
3407 'url': f
'{protocol}:{url_base}',
3408 'format_id': protocol
,
3409 'protocol': protocol
,
3413 def _find_jwplayer_data(self
, webpage
, video_id
=None, transform_source
=js_to_json
):
3415 r
'(?s)jwplayer\((?P<quote>[\'"])[^\'" ]+(?P
=quote
)\
)(?
!</script
>).*?\
.setup\s
*\
((?P
<options
>[^
)]+)\
)',
3419 jwplayer_data = self._parse_json(mobj.group('options
'),
3421 transform_source=transform_source)
3422 except ExtractorError:
3425 if isinstance(jwplayer_data, dict):
3426 return jwplayer_data
3428 def _extract_jwplayer_data(self, webpage, video_id, *args, **kwargs):
3429 jwplayer_data = self._find_jwplayer_data(
3430 webpage, video_id, transform_source=js_to_json)
3431 return self._parse_jwplayer_data(
3432 jwplayer_data, video_id, *args, **kwargs)
3434 def _parse_jwplayer_data(self, jwplayer_data, video_id=None, require_title=True,
3435 m3u8_id=None, mpd_id=None, rtmp_params=None, base_url=None):
3436 # JWPlayer backward compatibility: flattened playlists
3437 # https://github.com/jwplayer/jwplayer/blob/v7.4.3/src/js/api/config.js#L81-L96
3438 if 'playlist
' not in jwplayer_data:
3439 jwplayer_data = {'playlist': [jwplayer_data]}
3443 # JWPlayer backward compatibility: single playlist item
3444 # https://github.com/jwplayer/jwplayer/blob/v7.7.0/src/js/playlist/playlist.js#L10
3445 if not isinstance(jwplayer_data['playlist
'], list):
3446 jwplayer_data['playlist
'] = [jwplayer_data['playlist
']]
3448 for video_data in jwplayer_data['playlist
']:
3449 # JWPlayer backward compatibility: flattened sources
3450 # https://github.com/jwplayer/jwplayer/blob/v7.4.3/src/js/playlist/item.js#L29-L35
3451 if 'sources
' not in video_data:
3452 video_data['sources
'] = [video_data]
3454 this_video_id = video_id or video_data['mediaid
']
3456 formats = self._parse_jwplayer_formats(
3457 video_data['sources
'], video_id=this_video_id, m3u8_id=m3u8_id,
3458 mpd_id=mpd_id, rtmp_params=rtmp_params, base_url=base_url)
3461 tracks = video_data.get('tracks
')
3462 if tracks and isinstance(tracks, list):
3463 for track in tracks:
3464 if not isinstance(track, dict):
3466 track_kind = track.get('kind
')
3467 if not track_kind or not isinstance(track_kind, str):
3469 if track_kind.lower() not in ('captions
', 'subtitles
'):
3471 track_url = urljoin(base_url, track.get('file'))
3474 subtitles.setdefault(track.get('label
') or 'en
', []).append({
3475 'url
': self._proto_relative_url(track_url)
3479 'id': this_video_id,
3480 'title
': unescapeHTML(video_data['title
'] if require_title else video_data.get('title
')),
3481 'description
': clean_html(video_data.get('description
')),
3482 'thumbnail
': urljoin(base_url, self._proto_relative_url(video_data.get('image
'))),
3483 'timestamp
': int_or_none(video_data.get('pubdate
')),
3484 'duration
': float_or_none(jwplayer_data.get('duration
') or video_data.get('duration
')),
3485 'subtitles
': subtitles,
3487 # https://github.com/jwplayer/jwplayer/blob/master/src/js/utils/validator.js#L32
3488 if len(formats) == 1 and re.search(r'^
(?
:http|
//).*(?
:youtube\
.com|youtu\
.be
)/.+', formats[0]['url
']):
3490 '_type
': 'url_transparent
',
3491 'url
': formats[0]['url
'],
3494 self._sort_formats(formats)
3495 entry['formats
'] = formats
3496 entries.append(entry)
3497 if len(entries) == 1:
3500 return self.playlist_result(entries)
3502 def _parse_jwplayer_formats(self, jwplayer_sources_data, video_id=None,
3503 m3u8_id=None, mpd_id=None, rtmp_params=None, base_url=None):
3506 for source in jwplayer_sources_data:
3507 if not isinstance(source, dict):
3509 source_url = urljoin(
3510 base_url, self._proto_relative_url(source.get('file')))
3511 if not source_url or source_url in urls:
3513 urls.append(source_url)
3514 source_type = source.get('type') or ''
3515 ext = mimetype2ext(source_type) or determine_ext(source_url)
3516 if source_type == 'hls
' or ext == 'm3u8
':
3517 formats.extend(self._extract_m3u8_formats(
3518 source_url, video_id, 'mp4
', entry_protocol='m3u8_native
',
3519 m3u8_id=m3u8_id, fatal=False))
3520 elif source_type == 'dash
' or ext == 'mpd
':
3521 formats.extend(self._extract_mpd_formats(
3522 source_url, video_id, mpd_id=mpd_id, fatal=False))
3524 formats.extend(self._extract_smil_formats(
3525 source_url, video_id, fatal=False))
3526 # https://github.com/jwplayer/jwplayer/blob/master/src/js/providers/default.js#L67
3527 elif source_type.startswith('audio
') or ext in (
3528 'oga
', 'aac
', 'mp3
', 'mpeg
', 'vorbis
'):
3535 height = int_or_none(source.get('height
'))
3537 # Often no height is provided but there is a label in
3538 # format like "1080p", "720p SD", or 1080.
3539 height = int_or_none(self._search_regex(
3540 r'^
(\d{3,4}
)[pP
]?
(?
:\b|$
)', str(source.get('label
') or ''),
3541 'height
', default=None))
3544 'width
': int_or_none(source.get('width
')),
3546 'tbr
': int_or_none(source.get('bitrate
')),
3549 if source_url.startswith('rtmp
'):
3550 a_format['ext
'] = 'flv
'
3551 # See com/longtailvideo/jwplayer/media/RTMPMediaProvider.as
3552 # of jwplayer.flash.swf
3553 rtmp_url_parts = re.split(
3554 r'((?
:mp4|mp3|flv
):)', source_url, 1)
3555 if len(rtmp_url_parts) == 3:
3556 rtmp_url, prefix, play_path = rtmp_url_parts
3559 'play_path
': prefix + play_path,
3562 a_format.update(rtmp_params)
3563 formats.append(a_format)
3566 def _live_title(self, name):
3567 self._downloader.deprecation_warning('yt_dlp
.InfoExtractor
._live
_title
is deprecated
and does
not work
as expected
')
3570 def _int(self, v, name, fatal=False, **kwargs):
3571 res = int_or_none(v, **kwargs)
3573 msg = f'Failed to extract {name}
: Could
not parse value {v!r}
'
3575 raise ExtractorError(msg)
3577 self.report_warning(msg)
3580 def _float(self, v, name, fatal=False, **kwargs):
3581 res = float_or_none(v, **kwargs)
3583 msg = f'Failed to extract {name}
: Could
not parse value {v!r}
'
3585 raise ExtractorError(msg)
3587 self.report_warning(msg)
3590 def _set_cookie(self, domain, name, value, expire_time=None, port=None,
3591 path='/', secure=False, discard=False, rest={}, **kwargs):
3592 cookie = http.cookiejar.Cookie(
3593 0, name, value, port, port is not None, domain, True,
3594 domain.startswith('.'), path, True, secure, expire_time,
3595 discard, None, None, rest)
3596 self.cookiejar.set_cookie(cookie)
3598 def _get_cookies(self, url):
3599 """ Return a http.cookies.SimpleCookie with the cookies for the url """
3600 return http.cookies.SimpleCookie(self._downloader._calc_cookies(url))
3602 def _apply_first_set_cookie_header(self, url_handle, cookie):
3604 Apply first Set-Cookie header instead of the last. Experimental.
3606 Some sites (e.g. [1-3]) may serve two cookies under the same name
3607 in Set-Cookie header and expect the first (old) one to be set rather
3608 than second (new). However, as of RFC6265 the newer one cookie
3609 should be set into cookie store what actually happens.
3610 We will workaround this issue by resetting the cookie to
3611 the first one manually.
3612 1. https://new.vk.com/
3613 2. https://github.com/ytdl-org/youtube-dl/issues/9841#issuecomment-227871201
3614 3. https://learning.oreilly.com/
3616 for header, cookies in url_handle.headers.items():
3617 if header.lower() != 'set-cookie
':
3619 cookies = cookies.encode('iso
-8859-1').decode('utf
-8')
3620 cookie_value = re.search(
3621 r'%s=(.+?
);.*?
\b[Dd
]omain
=(.+?
)(?
:[,;]|$
)' % cookie, cookies)
3623 value, domain = cookie_value.groups()
3624 self._set_cookie(domain, cookie, value)
3628 def get_testcases(cls, include_onlymatching=False):
3629 t = getattr(cls, '_TEST
', None)
3631 assert not hasattr(cls, '_TESTS
'), f'{cls.ie_key()}IE has _TEST
and _TESTS
'
3634 tests = getattr(cls, '_TESTS
', [])
3636 if not include_onlymatching and t.get('only_matching
', False):
3638 t['name
'] = cls.ie_key()
3643 """Get age limit from the testcases"""
3644 return max(traverse_obj(
3645 tuple(cls.get_testcases(include_onlymatching=False)),
3646 (..., (('playlist
', 0), None), 'info_dict
', 'age_limit
')) or [0])
3649 def is_suitable(cls, age_limit):
3650 """Test whether the extractor is generally suitable for the given age limit"""
3651 return not age_restricted(cls.age_limit, age_limit)
3654 def description(cls, *, markdown=True, search_examples=None):
3655 """Description of the extractor"""
3657 if cls._NETRC_MACHINE:
3659 desc += f' [<abbr title
="netrc machine"><em
>{cls._NETRC_MACHINE}
</em
></abbr
>]'
3661 desc += f' [{cls._NETRC_MACHINE}
]'
3662 if cls.IE_DESC is False:
3665 desc += f' {cls.IE_DESC}
'
3667 desc += f'; "{cls.SEARCH_KEY}:" prefix
'
3669 _COUNTS = ('', '5', '10', 'all
')
3670 desc += f' (Example
: "{cls.SEARCH_KEY}{random.choice(_COUNTS)}:{random.choice(search_examples)}")'
3671 if not cls.working():
3672 desc += ' (**Currently broken
**)' if markdown else ' (Currently broken
)'
3674 name = f' - **{cls.IE_NAME}
**' if markdown else cls.IE_NAME
3675 return f'{name}
:{desc}
' if desc else name
3677 def extract_subtitles(self, *args, **kwargs):
3678 if (self.get_param('writesubtitles
', False)
3679 or self.get_param('listsubtitles
')):
3680 return self._get_subtitles(*args, **kwargs)
3683 def _get_subtitles(self, *args, **kwargs):
3684 raise NotImplementedError('This method must be implemented by subclasses
')
3686 def extract_comments(self, *args, **kwargs):
3687 if not self.get_param('getcomments
'):
3689 generator = self._get_comments(*args, **kwargs)
3696 comments.append(next(generator))
3697 except StopIteration:
3699 except KeyboardInterrupt:
3700 self.to_screen('Interrupted by user
')
3701 except Exception as e:
3702 if self.get_param('ignoreerrors
') is not True:
3704 self._downloader.report_error(e)
3705 comment_count = len(comments)
3706 self.to_screen(f'Extracted {comment_count} comments
')
3708 'comments
': comments,
3709 'comment_count
': None if interrupted else comment_count
3713 def _get_comments(self, *args, **kwargs):
3714 raise NotImplementedError('This method must be implemented by subclasses
')
3717 def _merge_subtitle_items(subtitle_list1, subtitle_list2):
3718 """ Merge subtitle items for one language. Items with duplicated URLs/data
3719 will be dropped. """
3720 list1_data = {(item.get('url'), item.get('data')) for item in subtitle_list1}
3721 ret = list(subtitle_list1)
3722 ret.extend(item for item in subtitle_list2 if (item.get('url
'), item.get('data
')) not in list1_data)
3726 def _merge_subtitles(cls, *dicts, target=None):
3727 """ Merge subtitle dictionaries, language by language. """
3731 for lang, subs in d.items():
3732 target[lang] = cls._merge_subtitle_items(target.get(lang, []), subs)
3735 def extract_automatic_captions(self, *args, **kwargs):
3736 if (self.get_param('writeautomaticsub
', False)
3737 or self.get_param('listsubtitles
')):
3738 return self._get_automatic_captions(*args, **kwargs)
3741 def _get_automatic_captions(self, *args, **kwargs):
3742 raise NotImplementedError('This method must be implemented by subclasses
')
3744 @functools.cached_property
3745 def _cookies_passed(self):
3746 """Whether cookies have been passed to YoutubeDL"""
3747 return self.get_param('cookiefile
') is not None or self.get_param('cookiesfrombrowser
') is not None
3749 def mark_watched(self, *args, **kwargs):
3750 if not self.get_param('mark_watched
', False):
3752 if self.supports_login() and self._get_login_info()[0] is not None or self._cookies_passed:
3753 self._mark_watched(*args, **kwargs)
3755 def _mark_watched(self, *args, **kwargs):
3756 raise NotImplementedError('This method must be implemented by subclasses
')
3758 def geo_verification_headers(self):
3760 geo_verification_proxy = self.get_param('geo_verification_proxy
')
3761 if geo_verification_proxy:
3762 headers['Ytdl
-request
-proxy
'] = geo_verification_proxy
3765 def _generic_id(self, url):
3766 return urllib.parse.unquote(os.path.splitext(url.rstrip('/').split('/')[-1])[0])
3768 def _generic_title(self, url):
3769 return urllib.parse.unquote(os.path.splitext(url_basename(url))[0])
3772 def _availability(is_private=None, needs_premium=None, needs_subscription=None, needs_auth=None, is_unlisted=None):
3773 all_known = all(map(
3774 lambda x: x is not None,
3775 (is_private, needs_premium, needs_subscription, needs_auth, is_unlisted)))
3777 'private
' if is_private
3778 else 'premium_only
' if needs_premium
3779 else 'subscriber_only
' if needs_subscription
3780 else 'needs_auth
' if needs_auth
3781 else 'unlisted
' if is_unlisted
3782 else 'public
' if all_known
3785 def _configuration_arg(self, key, default=NO_DEFAULT, *, ie_key=None, casesense=False):
3787 @returns A list of values for the extractor argument given by "key"
3788 or "default" if no such key is present
3789 @param default The default value to return when the key is not present (default: [])
3790 @param casesense When false, the values are converted to lower case
3793 self._downloader.params, ('extractor_args
', (ie_key or self.ie_key()).lower(), key))
3795 return [] if default is NO_DEFAULT else default
3796 return list(val) if casesense else [x.lower() for x in val]
3798 def _yes_playlist(self, playlist_id, video_id, smuggled_data=None, *, playlist_label='playlist
', video_label='video
'):
3799 if not playlist_id or not video_id:
3802 no_playlist = (smuggled_data or {}).get('force_noplaylist
')
3803 if no_playlist is not None:
3804 return not no_playlist
3806 video_id = '' if video_id is True else f' {video_id}
'
3807 playlist_id = '' if playlist_id is True else f' {playlist_id}
'
3808 if self.get_param('noplaylist
'):
3809 self.to_screen(f'Downloading just the {video_label}{video_id} because of
--no
-playlist
')
3811 self.to_screen(f'Downloading {playlist_label}{playlist_id}
- add
--no
-playlist to download just the {video_label}{video_id}
')
3815 class SearchInfoExtractor(InfoExtractor):
3817 Base class for paged search queries extractors.
3818 They accept URLs in the format _SEARCH_KEY(|all|[0-9]):{query}
3819 Instances should define _SEARCH_KEY and optionally _MAX_RESULTS
3822 _MAX_RESULTS = float('inf
')
3825 def _make_valid_url(cls):
3826 return r'%s(?P
<prefix
>|
[1-9][0-9]*|all
):(?P
<query
>[\s\S
]+)' % cls._SEARCH_KEY
3828 def _real_extract(self, query):
3829 prefix, query = self._match_valid_url(query).group('prefix
', 'query
')
3831 return self._get_n_results(query, 1)
3832 elif prefix == 'all
':
3833 return self._get_n_results(query, self._MAX_RESULTS)
3837 raise ExtractorError(f'invalid download number {n}
for query
"{query}"')
3838 elif n > self._MAX_RESULTS:
3839 self.report_warning('%s returns
max %i results (you requested
%i)' % (self._SEARCH_KEY, self._MAX_RESULTS, n))
3840 n = self._MAX_RESULTS
3841 return self._get_n_results(query, n)
3843 def _get_n_results(self, query, n):
3844 """Get a specified number of results for a query.
3845 Either this function or _search_results must be overridden by subclasses """
3846 return self.playlist_result(
3847 itertools.islice(self._search_results(query), 0, None if n == float('inf
') else n),
3850 def _search_results(self, query):
3851 """Returns an iterator of search results"""
3852 raise NotImplementedError('This method must be implemented by subclasses
')
3855 def SEARCH_KEY(cls):
3856 return cls._SEARCH_KEY