12 import xml
.etree
.ElementTree
14 from ..compat
import functools
, re
# isort: split
15 from ..compat
import (
16 compat_cookiejar_Cookie
,
17 compat_cookies_SimpleCookie
,
18 compat_etree_fromstring
,
25 compat_urllib_parse_unquote
,
26 compat_urllib_parse_urlencode
,
27 compat_urllib_request
,
30 from ..downloader
import FileDownloader
31 from ..downloader
.f4m
import get_base_url
, remove_encrypted_media
66 parse_m3u8_attributes
,
91 """Information Extractor class.
93 Information extractors are the classes that, given a URL, extract
94 information about the video (or videos) the URL refers to. This
95 information includes the real video URL, the video title, author and
96 others. The information is stored in a dictionary which is then
97 passed to the YoutubeDL. The YoutubeDL processes this
98 information possibly downloading the video to the file system, among
99 other possible outcomes.
101 The type field determines the type of the result.
102 By far the most common value (and the default if _type is missing) is
103 "video", which indicates a single video.
105 For a video, the dictionaries must include the following fields:
107 id: Video identifier.
108 title: Video title, unescaped. Set to an empty string if video has
109 no title as opposed to "None" which signifies that the
110 extractor failed to obtain a title
112 Additionally, it must contain either a formats entry or a url one:
114 formats: A list of dictionaries for each format available, ordered
115 from worst to best quality.
118 * url The mandatory URL representing the media:
119 for plain file media - HTTP URL of this file,
121 for HLS - URL of the M3U8 media playlist,
122 for HDS - URL of the F4M manifest,
124 - HTTP URL to plain file media (in case of
126 - URL of the MPD manifest or base URL
127 representing the media if MPD manifest
128 is parsed from a string (in case of
130 for MSS - URL of the ISM manifest.
132 The URL of the manifest file in case of
134 for HLS - URL of the M3U8 master playlist,
135 for HDS - URL of the F4M manifest,
136 for DASH - URL of the MPD manifest,
137 for MSS - URL of the ISM manifest.
138 * manifest_stream_number (For internal use only)
139 The index of the stream in the manifest file
140 * ext Will be calculated from URL if missing
141 * format A human-readable description of the format
142 ("mp4 container with h264/opus").
143 Calculated from the format_id, width, height.
144 and format_note fields if missing.
145 * format_id A short description of the format
146 ("mp4_h264_opus" or "19").
147 Technically optional, but strongly recommended.
148 * format_note Additional info about the format
149 ("3D" or "DASH video")
150 * width Width of the video, if known
151 * height Height of the video, if known
152 * resolution Textual description of width and height
153 * dynamic_range The dynamic range of the video. One of:
154 "SDR" (None), "HDR10", "HDR10+, "HDR12", "HLG, "DV"
155 * tbr Average bitrate of audio and video in KBit/s
156 * abr Average audio bitrate in KBit/s
157 * acodec Name of the audio codec in use
158 * asr Audio sampling rate in Hertz
159 * vbr Average video bitrate in KBit/s
161 * vcodec Name of the video codec in use
162 * container Name of the container format
163 * filesize The number of bytes, if known in advance
164 * filesize_approx An estimate for the number of bytes
165 * player_url SWF Player URL (used for rtmpdump).
166 * protocol The protocol that will be used for the actual
167 download, lower-case. One of "http", "https" or
168 one of the protocols defined in downloader.PROTOCOL_MAP
170 Base URL for fragments. Each fragment's path
171 value (if present) will be relative to
173 * fragments A list of fragments of a fragmented media.
174 Each fragment entry must contain either an url
175 or a path. If an url is present it should be
176 considered by a client. Otherwise both path and
177 fragment_base_url must be present. Here is
178 the list of all potential fields:
179 * "url" - fragment's URL
180 * "path" - fragment's path relative to
182 * "duration" (optional, int or float)
183 * "filesize" (optional, int)
184 * is_from_start Is a live format that can be downloaded
185 from the start. Boolean
186 * preference Order number of this format. If this field is
187 present and not None, the formats get sorted
188 by this field, regardless of all other values.
189 -1 for default (order by other properties),
190 -2 or smaller for less than default.
191 < -1000 to hide the format (if there is
192 another one which is strictly better)
193 * language Language code, e.g. "de" or "en-US".
194 * language_preference Is this in the language mentioned in
196 10 if it's what the URL is about,
197 -1 for default (don't know),
198 -10 otherwise, other values reserved for now.
199 * quality Order number of the video quality of this
200 format, irrespective of the file format.
201 -1 for default (order by other properties),
202 -2 or smaller for less than default.
203 * source_preference Order number for this video source
204 (quality takes higher priority)
205 -1 for default (order by other properties),
206 -2 or smaller for less than default.
207 * http_headers A dictionary of additional HTTP headers
208 to add to the request.
209 * stretched_ratio If given and not 1, indicates that the
210 video's pixels are not square.
211 width : height ratio as float.
212 * no_resume The server does not support resuming the
213 (HTTP or RTMP) download. Boolean.
214 * has_drm The format has DRM and cannot be downloaded. Boolean
215 * downloader_options A dictionary of downloader options
216 (For internal use only)
217 * http_chunk_size Chunk size for HTTP downloads
218 * ffmpeg_args Extra arguments for ffmpeg downloader
219 RTMP formats can also have the additional fields: page_url,
220 app, play_path, tc_url, flash_version, rtmp_live, rtmp_conn,
221 rtmp_protocol, rtmp_real_time
223 url: Final video URL.
224 ext: Video filename extension.
225 format: The video format, defaults to ext (used for --get-format)
226 player_url: SWF Player URL (used for rtmpdump).
228 The following fields are optional:
230 direct: True if a direct video file was given (must only be set by GenericIE)
231 alt_title: A secondary title of the video.
232 display_id An alternative identifier for the video, not necessarily
233 unique, but available before title. Typically, id is
234 something like "4234987", title "Dancing naked mole rats",
235 and display_id "dancing-naked-mole-rats"
236 thumbnails: A list of dictionaries, with the following entries:
237 * "id" (optional, string) - Thumbnail format ID
239 * "preference" (optional, int) - quality of the image
240 * "width" (optional, int)
241 * "height" (optional, int)
242 * "resolution" (optional, string "{width}x{height}",
244 * "filesize" (optional, int)
245 * "http_headers" (dict) - HTTP headers for the request
246 thumbnail: Full URL to a video thumbnail image.
247 description: Full video description.
248 uploader: Full name of the video uploader.
249 license: License name the video is licensed under.
250 creator: The creator of the video.
251 timestamp: UNIX timestamp of the moment the video was uploaded
252 upload_date: Video upload date in UTC (YYYYMMDD).
253 If not explicitly set, calculated from timestamp
254 release_timestamp: UNIX timestamp of the moment the video was released.
255 If it is not clear whether to use timestamp or this, use the former
256 release_date: The date (YYYYMMDD) when the video was released in UTC.
257 If not explicitly set, calculated from release_timestamp
258 modified_timestamp: UNIX timestamp of the moment the video was last modified.
259 modified_date: The date (YYYYMMDD) when the video was last modified in UTC.
260 If not explicitly set, calculated from modified_timestamp
261 uploader_id: Nickname or id of the video uploader.
262 uploader_url: Full URL to a personal webpage of the video uploader.
263 channel: Full name of the channel the video is uploaded on.
264 Note that channel fields may or may not repeat uploader
265 fields. This depends on a particular extractor.
266 channel_id: Id of the channel.
267 channel_url: Full URL to a channel webpage.
268 channel_follower_count: Number of followers of the channel.
269 location: Physical location where the video was filmed.
270 subtitles: The available subtitles as a dictionary in the format
271 {tag: subformats}. "tag" is usually a language code, and
272 "subformats" is a list sorted from lower to higher
273 preference, each element is a dictionary with the "ext"
275 * "data": The subtitles file contents
276 * "url": A URL pointing to the subtitles file
277 It can optionally also have:
278 * "name": Name or description of the subtitles
279 * "http_headers": A dictionary of additional HTTP headers
280 to add to the request.
281 "ext" will be calculated from URL if missing
282 automatic_captions: Like 'subtitles'; contains automatically generated
283 captions instead of normal subtitles
284 duration: Length of the video in seconds, as an integer or float.
285 view_count: How many users have watched the video on the platform.
286 like_count: Number of positive ratings of the video
287 dislike_count: Number of negative ratings of the video
288 repost_count: Number of reposts of the video
289 average_rating: Average rating give by users, the scale used depends on the webpage
290 comment_count: Number of comments on the video
291 comments: A list of comments, each with one or more of the following
292 properties (all but one of text or html optional):
293 * "author" - human-readable name of the comment author
294 * "author_id" - user ID of the comment author
295 * "author_thumbnail" - The thumbnail of the comment author
297 * "html" - Comment as HTML
298 * "text" - Plain text of the comment
299 * "timestamp" - UNIX timestamp of comment
300 * "parent" - ID of the comment this one is replying to.
301 Set to "root" to indicate that this is a
302 comment to the original video.
303 * "like_count" - Number of positive ratings of the comment
304 * "dislike_count" - Number of negative ratings of the comment
305 * "is_favorited" - Whether the comment is marked as
306 favorite by the video uploader
307 * "author_is_uploader" - Whether the comment is made by
309 age_limit: Age restriction for the video, as an integer (years)
310 webpage_url: The URL to the video webpage, if given to yt-dlp it
311 should allow to get the same result again. (It will be set
312 by YoutubeDL if it's missing)
313 categories: A list of categories that the video falls in, for example
315 tags: A list of tags assigned to the video, e.g. ["sweden", "pop music"]
316 cast: A list of the video cast
317 is_live: True, False, or None (=unknown). Whether this video is a
318 live stream that goes on instead of a fixed-length video.
319 was_live: True, False, or None (=unknown). Whether this video was
320 originally a live stream.
321 live_status: 'is_live', 'is_upcoming', 'was_live', 'not_live' or None (=unknown)
322 If absent, automatically set from is_live, was_live
323 start_time: Time in seconds where the reproduction should start, as
324 specified in the URL.
325 end_time: Time in seconds where the reproduction should end, as
326 specified in the URL.
327 chapters: A list of dictionaries, with the following entries:
328 * "start_time" - The start time of the chapter in seconds
329 * "end_time" - The end time of the chapter in seconds
330 * "title" (optional, string)
331 playable_in_embed: Whether this video is allowed to play in embedded
332 players on other sites. Can be True (=always allowed),
333 False (=never allowed), None (=unknown), or a string
334 specifying the criteria for embedability (Eg: 'whitelist')
335 availability: Under what condition the video is available. One of
336 'private', 'premium_only', 'subscriber_only', 'needs_auth',
337 'unlisted' or 'public'. Use 'InfoExtractor._availability'
339 __post_extractor: A function to be called just before the metadata is
340 written to either disk, logger or console. The function
341 must return a dict which will be added to the info_dict.
342 This is usefull for additional information that is
343 time-consuming to extract. Note that the fields thus
344 extracted will not be available to output template and
345 match_filter. So, only "comments" and "comment_count" are
346 currently allowed to be extracted via this method.
348 The following fields should only be used when the video belongs to some logical
351 chapter: Name or title of the chapter the video belongs to.
352 chapter_number: Number of the chapter the video belongs to, as an integer.
353 chapter_id: Id of the chapter the video belongs to, as a unicode string.
355 The following fields should only be used when the video is an episode of some
356 series, programme or podcast:
358 series: Title of the series or programme the video episode belongs to.
359 series_id: Id of the series or programme the video episode belongs to, as a unicode string.
360 season: Title of the season the video episode belongs to.
361 season_number: Number of the season the video episode belongs to, as an integer.
362 season_id: Id of the season the video episode belongs to, as a unicode string.
363 episode: Title of the video episode. Unlike mandatory video title field,
364 this field should denote the exact title of the video episode
365 without any kind of decoration.
366 episode_number: Number of the video episode within a season, as an integer.
367 episode_id: Id of the video episode, as a unicode string.
369 The following fields should only be used when the media is a track or a part of
372 track: Title of the track.
373 track_number: Number of the track within an album or a disc, as an integer.
374 track_id: Id of the track (useful in case of custom indexing, e.g. 6.iii),
376 artist: Artist(s) of the track.
377 genre: Genre(s) of the track.
378 album: Title of the album the track belongs to.
379 album_type: Type of the album (e.g. "Demo", "Full-length", "Split", "Compilation", etc).
380 album_artist: List of all artists appeared on the album (e.g.
381 "Ash Borer / Fell Voices" or "Various Artists", useful for splits
383 disc_number: Number of the disc or other physical medium the track belongs to,
385 release_year: Year (YYYY) when the album was released.
386 composer: Composer of the piece
388 The following fields should only be set for clips that should be cut from the original video:
390 section_start: Start time of the section in seconds
391 section_end: End time of the section in seconds
393 Unless mentioned otherwise, the fields should be Unicode strings.
395 Unless mentioned otherwise, None is equivalent to absence of information.
398 _type "playlist" indicates multiple videos.
399 There must be a key "entries", which is a list, an iterable, or a PagedList
400 object, each element of which is a valid dictionary by this specification.
402 Additionally, playlists can have "id", "title", and any other relevent
403 attributes with the same semantics as videos (see above).
405 It can also have the following optional fields:
407 playlist_count: The total number of videos in a playlist. If not given,
408 YoutubeDL tries to calculate it from "entries"
411 _type "multi_video" indicates that there are multiple videos that
412 form a single show, for examples multiple acts of an opera or TV episode.
413 It must have an entries key like a playlist and contain all the keys
414 required for a video at the same time.
417 _type "url" indicates that the video must be extracted from another
418 location, possibly by a different extractor. Its only required key is:
419 "url" - the next URL to extract.
420 The key "ie_key" can be set to the class name (minus the trailing "IE",
421 e.g. "Youtube") if the extractor class is known in advance.
422 Additionally, the dictionary may have any properties of the resolved entity
423 known in advance, for example "title" if the title of the referred video is
427 _type "url_transparent" entities have the same specification as "url", but
428 indicate that the given additional information is more precise than the one
429 associated with the resolved URL.
430 This is useful when a site employs a video service that hosts the video and
431 its technical metadata, but that video service does not embed a useful
432 title, description etc.
435 Subclasses of this should define a _VALID_URL regexp and, re-define the
436 _real_extract() and (optionally) _real_initialize() methods.
437 Probably, they should also be added to the list of extractors.
439 Subclasses may also override suitable() if necessary, but ensure the function
440 signature is preserved and that this function imports everything it needs
441 (except other extractors), so that lazy_extractors works correctly.
443 To support username + password (or netrc) login, the extractor must define a
444 _NETRC_MACHINE and re-define _perform_login(username, password) and
445 (optionally) _initialize_pre_login() methods. The _perform_login method will
446 be called between _initialize_pre_login and _real_initialize if credentials
447 are passed by the user. In cases where it is necessary to have the login
448 process as part of the extraction rather than initialization, _perform_login
449 can be left undefined.
451 _GEO_BYPASS attribute may be set to False in order to disable
452 geo restriction bypass mechanisms for a particular extractor.
453 Though it won't disable explicit geo restriction bypass based on
454 country code provided with geo_bypass_country.
456 _GEO_COUNTRIES attribute may contain a list of presumably geo unrestricted
457 countries for this extractor. One of these countries will be used by
458 geo restriction bypass mechanism right away in order to bypass
459 geo restriction, of course, if the mechanism is not disabled.
461 _GEO_IP_BLOCKS attribute may contain a list of presumably geo unrestricted
462 IP blocks in CIDR notation for this extractor. One of these IP blocks
463 will be used by geo restriction bypass mechanism similarly
466 The _WORKING attribute should be set to False for broken IEs
467 in order to warn the users and skip the tests.
472 _x_forwarded_for_ip
= None
474 _GEO_COUNTRIES
= None
475 _GEO_IP_BLOCKS
= None
477 _NETRC_MACHINE
= None
481 def _login_hint(self
, method
=NO_DEFAULT
, netrc
=None):
482 password_hint
= f
'--username and --password, or --netrc ({netrc or self._NETRC_MACHINE}) to provide account credentials'
485 'any': f
'Use --cookies, --cookies-from-browser, {password_hint}',
486 'password': f
'Use {password_hint}',
488 'Use --cookies-from-browser or --cookies for the authentication. '
489 'See https://github.com/ytdl-org/youtube-dl#how-do-i-pass-cookies-to-youtube-dl for how to manually pass cookies'),
490 }[method
if method
is not NO_DEFAULT
else 'any' if self
.supports_login() else 'cookies']
492 def __init__(self
, downloader
=None):
493 """Constructor. Receives an optional downloader (a YoutubeDL instance).
494 If a downloader is not passed during initialization,
495 it must be set using "set_downloader()" before "extract()" is called"""
497 self
._x
_forwarded
_for
_ip
= None
498 self
._printed
_messages
= set()
499 self
.set_downloader(downloader
)
502 def _match_valid_url(cls
, url
):
503 # This does not use has/getattr intentionally - we want to know whether
504 # we have cached the regexp for *this* class, whereas getattr would also
505 # match the superclass
506 if '_VALID_URL_RE' not in cls
.__dict
__:
507 if '_VALID_URL' not in cls
.__dict
__:
508 cls
._VALID
_URL
= cls
._make
_valid
_url
()
509 cls
._VALID
_URL
_RE
= re
.compile(cls
._VALID
_URL
)
510 return cls
._VALID
_URL
_RE
.match(url
)
513 def suitable(cls
, url
):
514 """Receives a URL and returns True if suitable for this IE."""
515 # This function must import everything it needs (except other extractors),
516 # so that lazy_extractors works correctly
517 return cls
._match
_valid
_url
(url
) is not None
520 def _match_id(cls
, url
):
521 return cls
._match
_valid
_url
(url
).group('id')
524 def get_temp_id(cls
, url
):
526 return cls
._match
_id
(url
)
527 except (IndexError, AttributeError):
532 """Getter method for _WORKING."""
536 def supports_login(cls
):
537 return bool(cls
._NETRC
_MACHINE
)
539 def initialize(self
):
540 """Initializes an instance (authentication, etc)."""
541 self
._printed
_messages
= set()
542 self
._initialize
_geo
_bypass
({
543 'countries': self
._GEO
_COUNTRIES
,
544 'ip_blocks': self
._GEO
_IP
_BLOCKS
,
547 self
._initialize
_pre
_login
()
548 if self
.supports_login():
549 username
, password
= self
._get
_login
_info
()
551 self
._perform
_login
(username
, password
)
552 elif self
.get_param('username') and False not in (self
.IE_DESC
, self
._NETRC
_MACHINE
):
553 self
.report_warning(f
'Login with password is not supported for this website. {self._login_hint("cookies")}')
554 self
._real
_initialize
()
557 def _initialize_geo_bypass(self
, geo_bypass_context
):
559 Initialize geo restriction bypass mechanism.
561 This method is used to initialize geo bypass mechanism based on faking
562 X-Forwarded-For HTTP header. A random country from provided country list
563 is selected and a random IP belonging to this country is generated. This
564 IP will be passed as X-Forwarded-For HTTP header in all subsequent
567 This method will be used for initial geo bypass mechanism initialization
568 during the instance initialization with _GEO_COUNTRIES and
571 You may also manually call it from extractor's code if geo bypass
572 information is not available beforehand (e.g. obtained during
573 extraction) or due to some other reason. In this case you should pass
574 this information in geo bypass context passed as first argument. It may
575 contain following fields:
577 countries: List of geo unrestricted countries (similar
579 ip_blocks: List of geo unrestricted IP blocks in CIDR notation
580 (similar to _GEO_IP_BLOCKS)
583 if not self
._x
_forwarded
_for
_ip
:
585 # Geo bypass mechanism is explicitly disabled by user
586 if not self
.get_param('geo_bypass', True):
589 if not geo_bypass_context
:
590 geo_bypass_context
= {}
592 # Backward compatibility: previously _initialize_geo_bypass
593 # expected a list of countries, some 3rd party code may still use
595 if isinstance(geo_bypass_context
, (list, tuple)):
596 geo_bypass_context
= {
597 'countries': geo_bypass_context
,
600 # The whole point of geo bypass mechanism is to fake IP
601 # as X-Forwarded-For HTTP header based on some IP block or
604 # Path 1: bypassing based on IP block in CIDR notation
606 # Explicit IP block specified by user, use it right away
607 # regardless of whether extractor is geo bypassable or not
608 ip_block
= self
.get_param('geo_bypass_ip_block', None)
610 # Otherwise use random IP block from geo bypass context but only
611 # if extractor is known as geo bypassable
613 ip_blocks
= geo_bypass_context
.get('ip_blocks')
614 if self
._GEO
_BYPASS
and ip_blocks
:
615 ip_block
= random
.choice(ip_blocks
)
618 self
._x
_forwarded
_for
_ip
= GeoUtils
.random_ipv4(ip_block
)
619 self
.write_debug(f
'Using fake IP {self._x_forwarded_for_ip} as X-Forwarded-For')
622 # Path 2: bypassing based on country code
624 # Explicit country code specified by user, use it right away
625 # regardless of whether extractor is geo bypassable or not
626 country
= self
.get_param('geo_bypass_country', None)
628 # Otherwise use random country code from geo bypass context but
629 # only if extractor is known as geo bypassable
631 countries
= geo_bypass_context
.get('countries')
632 if self
._GEO
_BYPASS
and countries
:
633 country
= random
.choice(countries
)
636 self
._x
_forwarded
_for
_ip
= GeoUtils
.random_ipv4(country
)
637 self
._downloader
.write_debug(
638 f
'Using fake IP {self._x_forwarded_for_ip} ({country.upper()}) as X-Forwarded-For')
640 def extract(self
, url
):
641 """Extracts URL information and returns it in list of dicts."""
646 self
.write_debug('Extracting URL: %s' % url
)
647 ie_result
= self
._real
_extract
(url
)
648 if ie_result
is None:
650 if self
._x
_forwarded
_for
_ip
:
651 ie_result
['__x_forwarded_for_ip'] = self
._x
_forwarded
_for
_ip
652 subtitles
= ie_result
.get('subtitles')
653 if (subtitles
and 'live_chat' in subtitles
654 and 'no-live-chat' in self
.get_param('compat_opts', [])):
655 del subtitles
['live_chat']
657 except GeoRestrictedError
as e
:
658 if self
.__maybe
_fake
_ip
_and
_retry
(e
.countries
):
661 except UnsupportedError
:
663 except ExtractorError
as e
:
665 'video_id': e
.video_id
or self
.get_temp_id(url
),
667 'tb': e
.traceback
or sys
.exc_info()[2],
668 'expected': e
.expected
,
671 if hasattr(e
, 'countries'):
672 kwargs
['countries'] = e
.countries
673 raise type(e
)(e
.orig_msg
, **kwargs
)
674 except compat_http_client
.IncompleteRead
as e
:
675 raise ExtractorError('A network error has occurred.', cause
=e
, expected
=True, video_id
=self
.get_temp_id(url
))
676 except (KeyError, StopIteration) as e
:
677 raise ExtractorError('An extractor error has occurred.', cause
=e
, video_id
=self
.get_temp_id(url
))
679 def __maybe_fake_ip_and_retry(self
, countries
):
680 if (not self
.get_param('geo_bypass_country', None)
682 and self
.get_param('geo_bypass', True)
683 and not self
._x
_forwarded
_for
_ip
685 country_code
= random
.choice(countries
)
686 self
._x
_forwarded
_for
_ip
= GeoUtils
.random_ipv4(country_code
)
687 if self
._x
_forwarded
_for
_ip
:
689 'Video is geo restricted. Retrying extraction with fake IP %s (%s) as X-Forwarded-For.'
690 % (self
._x
_forwarded
_for
_ip
, country_code
.upper()))
694 def set_downloader(self
, downloader
):
695 """Sets a YoutubeDL instance as the downloader for this IE."""
696 self
._downloader
= downloader
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
, compat_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
, compat_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
, compat_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 compat_urllib_request.Requestobject
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
, (compat_str
, 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 compat_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 compat_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 _search_json_ld(self, html, video_id, expected_type=None, **kwargs):
1396 json_ld_list = list(re.finditer(JSON_LD_RE, html))
1397 default = kwargs.get('default
', NO_DEFAULT)
1398 # JSON-LD may be malformed and thus `fatal` should be respected.
1399 # At the same time `default` may be passed that assumes `fatal=False`
1400 # for _search_regex. Let's simulate the same behavior here
as well
.
1401 fatal
= kwargs
.get('fatal', True) if default
is NO_DEFAULT
else False
1403 for mobj
in json_ld_list
:
1404 json_ld_item
= self
._parse
_json
(
1405 mobj
.group('json_ld'), video_id
, fatal
=fatal
)
1406 if not json_ld_item
:
1408 if isinstance(json_ld_item
, dict):
1409 json_ld
.append(json_ld_item
)
1410 elif isinstance(json_ld_item
, (list, tuple)):
1411 json_ld
.extend(json_ld_item
)
1413 json_ld
= self
._json
_ld
(json_ld
, video_id
, fatal
=fatal
, expected_type
=expected_type
)
1416 if default
is not NO_DEFAULT
:
1419 raise RegexNotFoundError('Unable to extract JSON-LD')
1421 self
.report_warning('unable to extract JSON-LD %s' % bug_reports_message())
1424 def _json_ld(self
, json_ld
, video_id
, fatal
=True, expected_type
=None):
1425 if isinstance(json_ld
, compat_str
):
1426 json_ld
= self
._parse
_json
(json_ld
, video_id
, fatal
=fatal
)
1430 if not isinstance(json_ld
, (list, tuple, dict)):
1432 if isinstance(json_ld
, dict):
1435 INTERACTION_TYPE_MAP
= {
1436 'CommentAction': 'comment',
1437 'AgreeAction': 'like',
1438 'DisagreeAction': 'dislike',
1439 'LikeAction': 'like',
1440 'DislikeAction': 'dislike',
1441 'ListenAction': 'view',
1442 'WatchAction': 'view',
1443 'ViewAction': 'view',
1446 def is_type(e
, *expected_types
):
1447 type = variadic(traverse_obj(e
, '@type'))
1448 return any(x
in type for x
in expected_types
)
1450 def extract_interaction_type(e
):
1451 interaction_type
= e
.get('interactionType')
1452 if isinstance(interaction_type
, dict):
1453 interaction_type
= interaction_type
.get('@type')
1454 return str_or_none(interaction_type
)
1456 def extract_interaction_statistic(e
):
1457 interaction_statistic
= e
.get('interactionStatistic')
1458 if isinstance(interaction_statistic
, dict):
1459 interaction_statistic
= [interaction_statistic
]
1460 if not isinstance(interaction_statistic
, list):
1462 for is_e
in interaction_statistic
:
1463 if not is_type(is_e
, 'InteractionCounter'):
1465 interaction_type
= extract_interaction_type(is_e
)
1466 if not interaction_type
:
1468 # For interaction count some sites provide string instead of
1469 # an integer (as per spec) with non digit characters (e.g. ",")
1470 # so extracting count with more relaxed str_to_int
1471 interaction_count
= str_to_int(is_e
.get('userInteractionCount'))
1472 if interaction_count
is None:
1474 count_kind
= INTERACTION_TYPE_MAP
.get(interaction_type
.split('/')[-1])
1477 count_key
= '%s_count' % count_kind
1478 if info
.get(count_key
) is not None:
1480 info
[count_key
] = interaction_count
1482 def extract_chapter_information(e
):
1484 'title': part
.get('name'),
1485 'start_time': part
.get('startOffset'),
1486 'end_time': part
.get('endOffset'),
1487 } for part
in variadic(e
.get('hasPart') or []) if part
.get('@type') == 'Clip']
1488 for idx
, (last_c
, current_c
, next_c
) in enumerate(zip(
1489 [{'end_time': 0}
] + chapters
, chapters
, chapters
[1:])):
1490 current_c
['end_time'] = current_c
['end_time'] or next_c
['start_time']
1491 current_c
['start_time'] = current_c
['start_time'] or last_c
['end_time']
1492 if None in current_c
.values():
1493 self
.report_warning(f
'Chapter {idx} contains broken data. Not extracting chapters')
1496 chapters
[-1]['end_time'] = chapters
[-1]['end_time'] or info
['duration']
1497 info
['chapters'] = chapters
1499 def extract_video_object(e
):
1500 assert is_type(e
, 'VideoObject')
1501 author
= e
.get('author')
1503 'url': traverse_obj(e
, 'contentUrl', 'embedUrl', expected_type
=url_or_none
),
1504 'title': unescapeHTML(e
.get('name')),
1505 'description': unescapeHTML(e
.get('description')),
1506 'thumbnails': [{'url': url}
1507 for url
in variadic(traverse_obj(e
, 'thumbnailUrl', 'thumbnailURL'))
1508 if url_or_none(url
)],
1509 'duration': parse_duration(e
.get('duration')),
1510 'timestamp': unified_timestamp(e
.get('uploadDate')),
1511 # author can be an instance of 'Organization' or 'Person' types.
1512 # both types can have 'name' property(inherited from 'Thing' type). [1]
1513 # however some websites are using 'Text' type instead.
1514 # 1. https://schema.org/VideoObject
1515 'uploader': author
.get('name') if isinstance(author
, dict) else author
if isinstance(author
, compat_str
) else None,
1516 'filesize': int_or_none(float_or_none(e
.get('contentSize'))),
1517 'tbr': int_or_none(e
.get('bitrate')),
1518 'width': int_or_none(e
.get('width')),
1519 'height': int_or_none(e
.get('height')),
1520 'view_count': int_or_none(e
.get('interactionCount')),
1522 extract_interaction_statistic(e
)
1523 extract_chapter_information(e
)
1525 def traverse_json_ld(json_ld
, at_top_level
=True):
1527 if at_top_level
and '@context' not in e
:
1529 if at_top_level
and set(e
.keys()) == {'@context', '@graph'}
:
1530 traverse_json_ld(variadic(e
['@graph'], allowed_types
=(dict,)), at_top_level
=False)
1532 if expected_type
is not None and not is_type(e
, expected_type
):
1534 rating
= traverse_obj(e
, ('aggregateRating', 'ratingValue'), expected_type
=float_or_none
)
1535 if rating
is not None:
1536 info
['average_rating'] = rating
1537 if is_type(e
, 'TVEpisode', 'Episode'):
1538 episode_name
= unescapeHTML(e
.get('name'))
1540 'episode': episode_name
,
1541 'episode_number': int_or_none(e
.get('episodeNumber')),
1542 'description': unescapeHTML(e
.get('description')),
1544 if not info
.get('title') and episode_name
:
1545 info
['title'] = episode_name
1546 part_of_season
= e
.get('partOfSeason')
1547 if is_type(part_of_season
, 'TVSeason', 'Season', 'CreativeWorkSeason'):
1549 'season': unescapeHTML(part_of_season
.get('name')),
1550 'season_number': int_or_none(part_of_season
.get('seasonNumber')),
1552 part_of_series
= e
.get('partOfSeries') or e
.get('partOfTVSeries')
1553 if is_type(part_of_series
, 'TVSeries', 'Series', 'CreativeWorkSeries'):
1554 info
['series'] = unescapeHTML(part_of_series
.get('name'))
1555 elif is_type(e
, 'Movie'):
1557 'title': unescapeHTML(e
.get('name')),
1558 'description': unescapeHTML(e
.get('description')),
1559 'duration': parse_duration(e
.get('duration')),
1560 'timestamp': unified_timestamp(e
.get('dateCreated')),
1562 elif is_type(e
, 'Article', 'NewsArticle'):
1564 'timestamp': parse_iso8601(e
.get('datePublished')),
1565 'title': unescapeHTML(e
.get('headline')),
1566 'description': unescapeHTML(e
.get('articleBody') or e
.get('description')),
1568 if is_type(traverse_obj(e
, ('video', 0)), 'VideoObject'):
1569 extract_video_object(e
['video'][0])
1570 elif is_type(traverse_obj(e
, ('subjectOf', 0)), 'VideoObject'):
1571 extract_video_object(e
['subjectOf'][0])
1572 elif is_type(e
, 'VideoObject'):
1573 extract_video_object(e
)
1574 if expected_type
is None:
1578 video
= e
.get('video')
1579 if is_type(video
, 'VideoObject'):
1580 extract_video_object(video
)
1581 if expected_type
is None:
1585 traverse_json_ld(json_ld
)
1587 return filter_dict(info
)
1589 def _search_nextjs_data(self
, webpage
, video_id
, *, transform_source
=None, fatal
=True, **kw
):
1590 return self
._parse
_json
(
1592 r
'(?s)<script[^>]+id=[\'"]__NEXT_DATA__[\'"][^
>]*>([^
<]+)</script
>',
1593 webpage, 'next
.js data
', fatal=fatal, **kw),
1594 video_id, transform_source=transform_source, fatal=fatal)
1596 def _search_nuxt_data(self, webpage, video_id, context_name='__NUXT__
', *, fatal=True, traverse=('data
', 0)):
1597 """Parses Nuxt.js metadata. This works as long as the function __NUXT__ invokes is a pure function"""
1598 rectx = re.escape(context_name)
1599 FUNCTION_RE = r'\
(function\
((?P
<arg_keys
>.*?
)\
){return\s+(?P<js>{.*?}
)\s
*;?\s
*}\
((?P
<arg_vals
>.*?
)\
)'
1600 js, arg_keys, arg_vals = self._search_regex(
1601 (rf'<script
>\s
*window\
.{rectx}
={FUNCTION_RE}\s
*\
)\s
*;?\s
*</script
>', rf'{rectx}\
(.*?{FUNCTION_RE}
'),
1602 webpage, context_name, group=('js
', 'arg_keys
', 'arg_vals
'), fatal=fatal)
1604 args = dict(zip(arg_keys.split(','), arg_vals.split(',')))
1606 for key, val in args.items():
1607 if val in ('undefined
', 'void
0'):
1610 ret = self._parse_json(js, video_id, transform_source=functools.partial(js_to_json, vars=args), fatal=fatal)
1611 return traverse_obj(ret, traverse) or {}
1614 def _hidden_inputs(html):
1615 html = re.sub(r'<!--(?
:(?
!<!--).)*-->', '', html)
1617 for input in re.findall(r'(?i
)(<input[^
>]+>)', html):
1618 attrs = extract_attributes(input)
1621 if attrs.get('type') not in ('hidden
', 'submit
'):
1623 name = attrs.get('name
') or attrs.get('id')
1624 value = attrs.get('value
')
1625 if name and value is not None:
1626 hidden_inputs[name] = value
1627 return hidden_inputs
1629 def _form_hidden_inputs(self, form_id, html):
1630 form = self._search_regex(
1631 r'(?
is)<form
[^
>]+?
id=(["\'])%s\1[^>]*>(?P<form>.+?)</form>' % form_id,
1632 html, '%s form' % form_id, group='form')
1633 return self._hidden_inputs(form)
1636 regex = r' *((?P<reverse>\+)?(?P<field>[a-zA-Z0-9_]+)((?P<separator>[~:])(?P<limit>.*?))?)? *$'
1638 default = ('hidden', 'aud_or_vid', 'hasvid', 'ie_pref', 'lang', 'quality',
1639 'res', 'fps', 'hdr:12', 'codec:vp9.2', 'size', 'br', 'asr',
1640 'proto', 'ext', 'hasaud', 'source', 'id') # These must not be aliases
1641 ytdl_default = ('hasaud', 'lang', 'quality', 'tbr', 'filesize', 'vbr',
1642 'height', 'width', 'proto', 'vext', 'abr', 'aext',
1643 'fps', 'fs_approx', 'source', 'id')
1646 'vcodec': {'type': 'ordered', 'regex': True,
1647 'order': ['av0?1', 'vp0?9.2', 'vp0?9', '[hx]265|he?vc?', '[hx]264|avc', 'vp0?8', 'mp4v|h263', 'theora', '', None, 'none']},
1648 'acodec': {'type': 'ordered', 'regex': True,
1649 'order': ['[af]lac', 'wav|aiff', 'opus', 'vorbis|ogg', 'aac', 'mp?4a?', 'mp3', 'e-?a?c-?3', 'ac-?3', 'dts', '', None, 'none']},
1650 'hdr': {'type': 'ordered', 'regex': True, 'field': 'dynamic_range',
1651 'order': ['dv', '(hdr)?12', r'(hdr)?10\+', '(hdr)?10', 'hlg', '', 'sdr', None]},
1652 'proto': {'type': 'ordered', 'regex': True, 'field': 'protocol',
1653 'order': ['(ht|f)tps', '(ht|f)tp$', 'm3u8.*', '.*dash', 'websocket_frag', 'rtmpe?', '', 'mms|rtsp', 'ws|websocket', 'f4']},
1654 'vext': {'type': 'ordered', 'field': 'video_ext',
1655 'order': ('mp4', 'webm', 'flv', '', 'none'),
1656 'order_free': ('webm', 'mp4', 'flv', '', 'none')},
1657 'aext': {'type': 'ordered', 'field': 'audio_ext',
1658 'order': ('m4a', 'aac', 'mp3', 'ogg', 'opus', 'webm', '', 'none'),
1659 'order_free': ('opus', 'ogg', 'webm', 'm4a', 'mp3', 'aac', '', 'none')},
1660 'hidden': {'visible': False, 'forced': True, 'type': 'extractor', 'max': -1000},
1661 'aud_or_vid': {'visible': False, 'forced': True, 'type': 'multiple',
1662 'field': ('vcodec', 'acodec'),
1663 'function': lambda it: int(any(v != 'none' for v in it))},
1664 'ie_pref': {'priority': True, 'type': 'extractor'},
1665 'hasvid': {'priority': True, 'field': 'vcodec', 'type': 'boolean', 'not_in_list': ('none',)},
1666 'hasaud': {'field': 'acodec', 'type': 'boolean', 'not_in_list': ('none',)},
1667 'lang': {'convert': 'float', 'field': 'language_preference', 'default': -1},
1668 'quality': {'convert': 'float', 'default': -1},
1669 'filesize': {'convert': 'bytes'},
1670 'fs_approx': {'convert': 'bytes', 'field': 'filesize_approx'},
1671 'id': {'convert': 'string', 'field': 'format_id'},
1672 'height': {'convert': 'float_none'},
1673 'width': {'convert': 'float_none'},
1674 'fps': {'convert': 'float_none'},
1675 'tbr': {'convert': 'float_none'},
1676 'vbr': {'convert': 'float_none'},
1677 'abr': {'convert': 'float_none'},
1678 'asr': {'convert': 'float_none'},
1679 'source': {'convert': 'float', 'field': 'source_preference', 'default': -1},
1681 'codec': {'type': 'combined', 'field': ('vcodec', 'acodec')},
1682 'br': {'type': 'combined', 'field': ('tbr', 'vbr', 'abr'), 'same_limit': True},
1683 'size': {'type': 'combined', 'same_limit': True, 'field': ('filesize', 'fs_approx')},
1684 'ext': {'type': 'combined', 'field': ('vext', 'aext')},
1685 'res': {'type': 'multiple', 'field': ('height', 'width'),
1686 'function': lambda it: (lambda l: min(l) if l else 0)(tuple(filter(None, it)))},
1688 # For compatibility with youtube-dl
1689 'format_id': {'type': 'alias', 'field': 'id'},
1690 'preference': {'type': 'alias', 'field': 'ie_pref'},
1691 'language_preference': {'type': 'alias', 'field': 'lang'},
1692 'source_preference': {'type': 'alias', 'field': 'source'},
1693 'protocol': {'type': 'alias', 'field': 'proto'},
1694 'filesize_approx': {'type': 'alias', 'field': 'fs_approx'},
1697 'dimension': {'type': 'alias', 'field': 'res', 'deprecated': True},
1698 'resolution': {'type': 'alias', 'field': 'res', 'deprecated': True},
1699 'extension': {'type': 'alias', 'field': 'ext', 'deprecated': True},
1700 'bitrate': {'type': 'alias', 'field': 'br', 'deprecated': True},
1701 'total_bitrate': {'type': 'alias', 'field': 'tbr', 'deprecated': True},
1702 'video_bitrate': {'type': 'alias', 'field': 'vbr', 'deprecated': True},
1703 'audio_bitrate': {'type': 'alias', 'field': 'abr', 'deprecated': True},
1704 'framerate': {'type': 'alias', 'field': 'fps', 'deprecated': True},
1705 'filesize_estimate': {'type': 'alias', 'field': 'size', 'deprecated': True},
1706 'samplerate': {'type': 'alias', 'field': 'asr', 'deprecated': True},
1707 'video_ext': {'type': 'alias', 'field': 'vext', 'deprecated': True},
1708 'audio_ext': {'type': 'alias', 'field': 'aext', 'deprecated': True},
1709 'video_codec': {'type': 'alias', 'field': 'vcodec', 'deprecated': True},
1710 'audio_codec': {'type': 'alias', 'field': 'acodec', 'deprecated': True},
1711 'video': {'type': 'alias', 'field': 'hasvid', 'deprecated': True},
1712 'has_video': {'type': 'alias', 'field': 'hasvid', 'deprecated': True},
1713 'audio': {'type': 'alias', 'field': 'hasaud', 'deprecated': True},
1714 'has_audio': {'type': 'alias', 'field': 'hasaud', 'deprecated': True},
1715 'extractor': {'type': 'alias', 'field': 'ie_pref', 'deprecated': True},
1716 'extractor_preference': {'type': 'alias', 'field': 'ie_pref', 'deprecated': True},
1719 def __init__(self, ie, field_preference):
1721 self.ydl = ie._downloader
1722 self.evaluate_params(self.ydl.params, field_preference)
1723 if ie.get_param('verbose'):
1724 self.print_verbose_info(self.ydl.write_debug)
1726 def _get_field_setting(self, field, key):
1727 if field not in self.settings:
1728 if key in ('forced', 'priority'):
1730 self.ydl.deprecation_warning(
1731 f'Using arbitrary fields ({field}) for format sorting is deprecated '
1732 'and may be removed in a future version')
1733 self.settings[field] = {}
1734 propObj = self.settings[field]
1735 if key not in propObj:
1736 type = propObj.get('type')
1738 default = 'preference' if type == 'extractor' else (field,) if type in ('combined', 'multiple') else field
1739 elif key == 'convert':
1740 default = 'order' if type == 'ordered' else 'float_string' if field else 'ignore'
1742 default = {'type': 'field', 'visible': True, 'order': [], 'not_in_list': (None,)}.get(key, None)
1743 propObj[key] = default
1746 def _resolve_field_value(self, field, value, convertNone=False):
1751 value = value.lower()
1752 conversion = self._get_field_setting(field, 'convert')
1753 if conversion == 'ignore':
1755 if conversion == 'string':
1757 elif conversion == 'float_none':
1758 return float_or_none(value)
1759 elif conversion == 'bytes':
1760 return FileDownloader.parse_bytes(value)
1761 elif conversion == 'order':
1762 order_list = (self._use_free_order and self._get_field_setting(field, 'order_free')) or self._get_field_setting(field, 'order')
1763 use_regex = self._get_field_setting(field, 'regex')
1764 list_length = len(order_list)
1765 empty_pos = order_list.index('') if '' in order_list else list_length + 1
1766 if use_regex and value is not None:
1767 for i, regex in enumerate(order_list):
1768 if regex and re.match(regex, value):
1769 return list_length - i
1770 return list_length - empty_pos # not in list
1771 else: # not regex or value = None
1772 return list_length - (order_list.index(value) if value in order_list else empty_pos)
1774 if value.isnumeric():
1777 self.settings[field]['convert'] = 'string'
1780 def evaluate_params(self, params, sort_extractor):
1781 self._use_free_order = params.get('prefer_free_formats', False)
1782 self._sort_user = params.get('format_sort', [])
1783 self._sort_extractor = sort_extractor
1785 def add_item(field, reverse, closest, limit_text):
1786 field = field.lower()
1787 if field in self._order:
1789 self._order.append(field)
1790 limit = self._resolve_field_value(field, limit_text)
1793 'closest': False if limit is None else closest,
1794 'limit_text': limit_text,
1796 if field in self.settings:
1797 self.settings[field].update(data)
1799 self.settings[field] = data
1802 tuple(field for field in self.default if self._get_field_setting(field, 'forced'))
1803 + (tuple() if params.get('format_sort_force', False)
1804 else tuple(field for field in self.default if self._get_field_setting(field, 'priority')))
1805 + tuple(self._sort_user) + tuple(sort_extractor) + self.default)
1807 for item in sort_list:
1808 match = re.match(self.regex, item)
1810 raise ExtractorError('Invalid format sort string "%s" given by extractor' % item)
1811 field = match.group('field')
1814 if self._get_field_setting(field, 'type') == 'alias':
1815 alias, field = field, self._get_field_setting(field, 'field')
1816 if self._get_field_setting(alias, 'deprecated'):
1817 self.ydl.deprecation_warning(
1818 f'Format sorting alias {alias} is deprecated '
1819 f'and may be removed in a future version. Please use {field} instead')
1820 reverse = match.group('reverse') is not None
1821 closest = match.group('separator') == '~'
1822 limit_text = match.group('limit')
1824 has_limit = limit_text is not None
1825 has_multiple_fields = self._get_field_setting(field, 'type') == 'combined'
1826 has_multiple_limits = has_limit and has_multiple_fields and not self._get_field_setting(field, 'same_limit')
1828 fields = self._get_field_setting(field, 'field') if has_multiple_fields else (field,)
1829 limits = limit_text.split(':') if has_multiple_limits else (limit_text,) if has_limit else tuple()
1830 limit_count = len(limits)
1831 for (i, f) in enumerate(fields):
1832 add_item(f, reverse, closest,
1833 limits[i] if i < limit_count
1834 else limits[0] if has_limit and not has_multiple_limits
1837 def print_verbose_info(self, write_debug):
1839 write_debug('Sort order given by user: %s' % ', '.join(self._sort_user))
1840 if self._sort_extractor:
1841 write_debug('Sort order given by extractor: %s' % ', '.join(self._sort_extractor))
1842 write_debug('Formats sorted by: %s' % ', '.join(['%s%s%s' % (
1843 '+' if self._get_field_setting(field, 'reverse') else '', field,
1844 '%s%s(%s)' % ('~' if self._get_field_setting(field, 'closest') else ':',
1845 self._get_field_setting(field, 'limit_text'),
1846 self._get_field_setting(field, 'limit'))
1847 if self._get_field_setting(field, 'limit_text') is not None else '')
1848 for field in self._order if self._get_field_setting(field, 'visible')]))
1850 def _calculate_field_preference_from_value(self, format, field, type, value):
1851 reverse = self._get_field_setting(field, 'reverse')
1852 closest = self._get_field_setting(field, 'closest')
1853 limit = self._get_field_setting(field, 'limit')
1855 if type == 'extractor':
1856 maximum = self._get_field_setting(field, 'max')
1857 if value is None or (maximum is not None and value >= maximum):
1859 elif type == 'boolean':
1860 in_list = self._get_field_setting(field, 'in_list')
1861 not_in_list = self._get_field_setting(field, 'not_in_list')
1862 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
1863 elif type == 'ordered':
1864 value = self._resolve_field_value(field, value, True)
1866 # try to convert to number
1867 val_num = float_or_none(value, default=self._get_field_setting(field, 'default'))
1868 is_num = self._get_field_setting(field, 'convert') != 'string' and val_num is not None
1872 return ((-10, 0) if value is None
1873 else (1, value, 0) if not is_num # if a field has mixed strings and numbers, strings are sorted higher
1874 else (0, -abs(value - limit), value - limit if reverse else limit - value) if closest
1875 else (0, value, 0) if not reverse and (limit is None or value <= limit)
1876 else (0, -value, 0) if limit is None or (reverse and value == limit) or value > limit
1877 else (-1, value, 0))
1879 def _calculate_field_preference(self, format, field):
1880 type = self._get_field_setting(field, 'type') # extractor, boolean, ordered, field, multiple
1881 get_value = lambda f: format.get(self._get_field_setting(f, 'field'))
1882 if type == 'multiple':
1883 type = 'field' # Only 'field' is allowed in multiple for now
1884 actual_fields = self._get_field_setting(field, 'field')
1886 value = self._get_field_setting(field, 'function')(get_value(f) for f in actual_fields)
1888 value = get_value(field)
1889 return self._calculate_field_preference_from_value(format, field, type, value)
1891 def calculate_preference(self, format):
1892 # Determine missing protocol
1893 if not format.get('protocol'):
1894 format['protocol'] = determine_protocol(format)
1896 # Determine missing ext
1897 if not format.get('ext') and 'url' in format:
1898 format['ext'] = determine_ext(format['url'])
1899 if format.get('vcodec') == 'none':
1900 format['audio_ext'] = format['ext'] if format.get('acodec') != 'none' else 'none'
1901 format['video_ext'] = 'none'
1903 format['video_ext'] = format['ext']
1904 format['audio_ext'] = 'none'
1905 # if format.get('preference') is None and format.get('ext') in ('f4f', 'f4m'): # Not supported?
1906 # format['preference'] = -1000
1908 # Determine missing bitrates
1909 if format.get('tbr') is None:
1910 if format.get('vbr') is not None and format.get('abr') is not None:
1911 format['tbr'] = format.get('vbr', 0) + format.get('abr', 0)
1913 if format.get('vcodec') != 'none' and format.get('vbr') is None:
1914 format['vbr'] = format.get('tbr') - format.get('abr', 0)
1915 if format.get('acodec') != 'none' and format.get('abr') is None:
1916 format['abr'] = format.get('tbr') - format.get('vbr', 0)
1918 return tuple(self._calculate_field_preference(format, field) for field in self._order)
1920 def _sort_formats(self, formats, field_preference=[]):
1923 formats.sort(key=self.FormatSort(self, field_preference).calculate_preference)
1925 def _check_formats(self, formats, video_id):
1927 formats[:] = filter(
1928 lambda f: self._is_valid_url(
1930 item='%s video format' % f.get('format_id') if f.get('format_id') else 'video'),
1934 def _remove_duplicate_formats(formats):
1938 if f['url'] not in format_urls:
1939 format_urls.add(f['url'])
1940 unique_formats.append(f)
1941 formats[:] = unique_formats
1943 def _is_valid_url(self, url, video_id, item='video', headers={}):
1944 url = self._proto_relative_url(url, scheme='http:')
1945 # For now assume non HTTP(S) URLs always valid
1946 if not (url.startswith('http://') or url.startswith('https://')):
1949 self._request_webpage(url, video_id, 'Checking %s URL' % item, headers=headers)
1951 except ExtractorError as e:
1953 '%s: %s URL is invalid, skipping: %s'
1954 % (video_id, item, error_to_compat_str(e.cause)))
1957 def http_scheme(self):
1958 """ Either "http
:" or "https
:", depending on the user's preferences """
1961 if self.get_param('prefer_insecure', False)
1964 def _proto_relative_url(self, url, scheme=None):
1967 if url.startswith('//'):
1969 scheme = self.http_scheme()
1974 def _sleep(self, timeout, video_id, msg_template=None):
1975 if msg_template is None:
1976 msg_template = '%(video_id)s: Waiting for %(timeout)s seconds'
1977 msg = msg_template % {'video_id': video_id, 'timeout': timeout}
1981 def _extract_f4m_formats(self, manifest_url, video_id, preference=None, quality=None, f4m_id=None,
1982 transform_source=lambda s: fix_xml_ampersands(s).strip(),
1983 fatal=True, m3u8_id=None, data=None, headers={}, query={}):
1984 res = self._download_xml_handle(
1985 manifest_url, video_id, 'Downloading f4m manifest',
1986 'Unable to download f4m manifest',
1987 # Some manifests may be malformed, e.g. prosiebensat1 generated manifests
1988 # (see https://github.com/ytdl-org/youtube-dl/issues/6215#issuecomment-121704244)
1989 transform_source=transform_source,
1990 fatal=fatal, data=data, headers=headers, query=query)
1994 manifest, urlh = res
1995 manifest_url = urlh.geturl()
1997 return self._parse_f4m_formats(
1998 manifest, manifest_url, video_id, preference=preference, quality=quality, f4m_id=f4m_id,
1999 transform_source=transform_source, fatal=fatal, m3u8_id=m3u8_id)
2001 def _parse_f4m_formats(self, manifest, manifest_url, video_id, preference=None, quality=None, f4m_id=None,
2002 transform_source=lambda s: fix_xml_ampersands(s).strip(),
2003 fatal=True, m3u8_id=None):
2004 if not isinstance(manifest, xml.etree.ElementTree.Element) and not fatal:
2007 # currently yt-dlp cannot decode the playerVerificationChallenge as Akamai uses Adobe Alchemy
2008 akamai_pv = manifest.find('{http://ns.adobe.com/f4m/1.0}pv-2.0')
2009 if akamai_pv is not None and ';' in akamai_pv.text:
2010 playerVerificationChallenge = akamai_pv.text.split(';')[0]
2011 if playerVerificationChallenge.strip() != '':
2015 manifest_version = '1.0'
2016 media_nodes = manifest.findall('{http://ns.adobe.com/f4m/1.0}media')
2018 manifest_version = '2.0'
2019 media_nodes = manifest.findall('{http://ns.adobe.com/f4m/2.0}media')
2020 # Remove unsupported DRM protected media from final formats
2021 # rendition (see https://github.com/ytdl-org/youtube-dl/issues/8573).
2022 media_nodes = remove_encrypted_media(media_nodes)
2026 manifest_base_url = get_base_url(manifest)
2028 bootstrap_info = xpath_element(
2029 manifest, ['{http://ns.adobe.com/f4m/1.0}bootstrapInfo', '{http://ns.adobe.com/f4m/2.0}bootstrapInfo'],
2030 'bootstrap info', default=None)
2033 mime_type = xpath_text(
2034 manifest, ['{http://ns.adobe.com/f4m/1.0}mimeType', '{http://ns.adobe.com/f4m/2.0}mimeType'],
2035 'base URL', default=None)
2036 if mime_type and mime_type.startswith('audio/'):
2039 for i, media_el in enumerate(media_nodes):
2040 tbr = int_or_none(media_el.attrib.get('bitrate'))
2041 width = int_or_none(media_el.attrib.get('width'))
2042 height = int_or_none(media_el.attrib.get('height'))
2043 format_id = join_nonempty(f4m_id, tbr or i)
2044 # If <bootstrapInfo> is present, the specified f4m is a
2045 # stream-level manifest, and only set-level manifests may refer to
2046 # external resources. See section 11.4 and section 4 of F4M spec
2047 if bootstrap_info is None:
2049 # @href is introduced in 2.0, see section 11.6 of F4M spec
2050 if manifest_version == '2.0':
2051 media_url = media_el.attrib.get('href')
2052 if media_url is None:
2053 media_url = media_el.attrib.get('url')
2057 media_url if media_url.startswith('http://') or media_url.startswith('https://')
2058 else ((manifest_base_url or '/'.join(manifest_url.split('/')[:-1])) + '/' + media_url))
2059 # If media_url is itself a f4m manifest do the recursive extraction
2060 # since bitrates in parent manifest (this one) and media_url manifest
2061 # may differ leading to inability to resolve the format by requested
2062 # bitrate in f4m downloader
2063 ext = determine_ext(manifest_url)
2065 f4m_formats = self._extract_f4m_formats(
2066 manifest_url, video_id, preference=preference, quality=quality, f4m_id=f4m_id,
2067 transform_source=transform_source, fatal=fatal)
2068 # Sometimes stream-level manifest contains single media entry that
2069 # does not contain any quality metadata (e.g. http://matchtv.ru/#live-player).
2070 # At the same time parent's media entry in set-level manifest may
2071 # contain it. We will copy it from parent in such cases.
2072 if len(f4m_formats) == 1:
2075 'tbr': f.get('tbr') or tbr,
2076 'width': f.get('width') or width,
2077 'height': f.get('height') or height,
2078 'format_id': f.get('format_id') if not tbr else format_id,
2081 formats.extend(f4m_formats)
2084 formats.extend(self._extract_m3u8_formats(
2085 manifest_url, video_id, 'mp4', preference=preference,
2086 quality=quality, m3u8_id=m3u8_id, fatal=fatal))
2089 'format_id': format_id,
2090 'url': manifest_url,
2091 'manifest_url': manifest_url,
2092 'ext': 'flv' if bootstrap_info is not None else None,
2098 'preference': preference,
2103 def _m3u8_meta_format(self, m3u8_url, ext=None, preference=None, quality=None, m3u8_id=None):
2105 'format_id': join_nonempty(m3u8_id, 'meta'),
2109 'preference': preference - 100 if preference else -100,
2111 'resolution': 'multiple',
2112 'format_note': 'Quality selection URL',
2115 def _report_ignoring_subs(self, name):
2116 self.report_warning(bug_reports_message(
2117 f'Ignoring subtitle tracks found in the {name} manifest; '
2118 'if any subtitle tracks are missing,'
2121 def _extract_m3u8_formats(self, *args, **kwargs):
2122 fmts, subs = self._extract_m3u8_formats_and_subtitles(*args, **kwargs)
2124 self._report_ignoring_subs('HLS')
2127 def _extract_m3u8_formats_and_subtitles(
2128 self, m3u8_url, video_id, ext=None, entry_protocol='m3u8_native',
2129 preference=None, quality=None, m3u8_id=None, note=None,
2130 errnote=None, fatal=True, live=False, data=None, headers={},
2133 res = self._download_webpage_handle(
2135 note='Downloading m3u8 information' if note is None else note,
2136 errnote='Failed to download m3u8 information' if errnote is None else errnote,
2137 fatal=fatal, data=data, headers=headers, query=query)
2142 m3u8_doc, urlh = res
2143 m3u8_url = urlh.geturl()
2145 return self._parse_m3u8_formats_and_subtitles(
2146 m3u8_doc, m3u8_url, ext=ext, entry_protocol=entry_protocol,
2147 preference=preference, quality=quality, m3u8_id=m3u8_id,
2148 note=note, errnote=errnote, fatal=fatal, live=live, data=data,
2149 headers=headers, query=query, video_id=video_id)
2151 def _parse_m3u8_formats_and_subtitles(
2152 self, m3u8_doc, m3u8_url=None, ext=None, entry_protocol='m3u8_native',
2153 preference=None, quality=None, m3u8_id=None, live=False, note=None,
2154 errnote=None, fatal=True, data=None, headers={}, query={},
2156 formats, subtitles = [], {}
2158 has_drm = re.search('|'.join([
2159 r'#EXT-X-FAXS-CM:', # Adobe Flash Access
2160 r'#EXT-X-(?:SESSION-)?KEY:.*?URI="skd
://', # Apple FairPlay
2163 def format_url(url):
2164 return url if re.match(r'^https?
://', url) else compat_urlparse.urljoin(m3u8_url, url)
2166 if self.get_param('hls_split_discontinuity
', False):
2167 def _extract_m3u8_playlist_indices(manifest_url=None, m3u8_doc=None):
2169 if not manifest_url:
2171 m3u8_doc = self._download_webpage(
2172 manifest_url, video_id, fatal=fatal, data=data, headers=headers,
2173 note=False, errnote='Failed to download m3u8 playlist information
')
2174 if m3u8_doc is False:
2176 return range(1 + sum(line.startswith('#EXT-X-DISCONTINUITY') for line in m3u8_doc.splitlines()))
2179 def _extract_m3u8_playlist_indices(*args
, **kwargs
):
2183 # 1. https://tools.ietf.org/html/draft-pantos-http-live-streaming-21
2184 # 2. https://github.com/ytdl-org/youtube-dl/issues/12211
2185 # 3. https://github.com/ytdl-org/youtube-dl/issues/18923
2187 # We should try extracting formats only from master playlists [1, 4.3.4],
2188 # i.e. playlists that describe available qualities. On the other hand
2189 # media playlists [1, 4.3.3] should be returned as is since they contain
2190 # just the media without qualities renditions.
2191 # Fortunately, master playlist can be easily distinguished from media
2192 # playlist based on particular tags availability. As of [1, 4.3.3, 4.3.4]
2193 # master playlist tags MUST NOT appear in a media playlist and vice versa.
2194 # As of [1, 4.3.3.1] #EXT-X-TARGETDURATION tag is REQUIRED for every
2195 # media playlist and MUST NOT appear in master playlist thus we can
2196 # clearly detect media playlist with this criterion.
2198 if '#EXT-X-TARGETDURATION' in m3u8_doc
: # media playlist, return as is
2200 'format_id': join_nonempty(m3u8_id
, idx
),
2201 'format_index': idx
,
2202 'url': m3u8_url
or encode_data_uri(m3u8_doc
.encode('utf-8'), 'application/x-mpegurl'),
2204 'protocol': entry_protocol
,
2205 'preference': preference
,
2208 } for idx
in _extract_m3u8_playlist_indices(m3u8_doc
=m3u8_doc
)]
2210 return formats
, subtitles
2213 last_stream_inf
= {}
2215 def extract_media(x_media_line
):
2216 media
= parse_m3u8_attributes(x_media_line
)
2217 # As per [1, 4.3.4.1] TYPE, GROUP-ID and NAME are REQUIRED
2218 media_type
, group_id
, name
= media
.get('TYPE'), media
.get('GROUP-ID'), media
.get('NAME')
2219 if not (media_type
and group_id
and name
):
2221 groups
.setdefault(group_id
, []).append(media
)
2222 # <https://tools.ietf.org/html/rfc8216#section-4.3.4.1>
2223 if media_type
== 'SUBTITLES':
2224 # According to RFC 8216 §4.3.4.2.1, URI is REQUIRED in the
2225 # EXT-X-MEDIA tag if the media type is SUBTITLES.
2226 # However, lack of URI has been spotted in the wild.
2227 # e.g. NebulaIE; see https://github.com/yt-dlp/yt-dlp/issues/339
2228 if not media
.get('URI'):
2230 url
= format_url(media
['URI'])
2233 'ext': determine_ext(url
),
2235 if sub_info
['ext'] == 'm3u8':
2236 # Per RFC 8216 §3.1, the only possible subtitle format m3u8
2237 # files may contain is WebVTT:
2238 # <https://tools.ietf.org/html/rfc8216#section-3.1>
2239 sub_info
['ext'] = 'vtt'
2240 sub_info
['protocol'] = 'm3u8_native'
2241 lang
= media
.get('LANGUAGE') or 'und'
2242 subtitles
.setdefault(lang
, []).append(sub_info
)
2243 if media_type
not in ('VIDEO', 'AUDIO'):
2245 media_url
= media
.get('URI')
2247 manifest_url
= format_url(media_url
)
2249 'format_id': join_nonempty(m3u8_id
, group_id
, name
, idx
),
2250 'format_note': name
,
2251 'format_index': idx
,
2252 'url': manifest_url
,
2253 'manifest_url': m3u8_url
,
2254 'language': media
.get('LANGUAGE'),
2256 'protocol': entry_protocol
,
2257 'preference': preference
,
2259 'vcodec': 'none' if media_type
== 'AUDIO' else None,
2260 } for idx
in _extract_m3u8_playlist_indices(manifest_url
))
2262 def build_stream_name():
2263 # Despite specification does not mention NAME attribute for
2264 # EXT-X-STREAM-INF tag it still sometimes may be present (see [1]
2265 # or vidio test in TestInfoExtractor.test_parse_m3u8_formats)
2266 # 1. http://www.vidio.com/watch/165683-dj_ambred-booyah-live-2015
2267 stream_name
= last_stream_inf
.get('NAME')
2270 # If there is no NAME in EXT-X-STREAM-INF it will be obtained
2271 # from corresponding rendition group
2272 stream_group_id
= last_stream_inf
.get('VIDEO')
2273 if not stream_group_id
:
2275 stream_group
= groups
.get(stream_group_id
)
2276 if not stream_group
:
2277 return stream_group_id
2278 rendition
= stream_group
[0]
2279 return rendition
.get('NAME') or stream_group_id
2281 # parse EXT-X-MEDIA tags before EXT-X-STREAM-INF in order to have the
2282 # chance to detect video only formats when EXT-X-STREAM-INF tags
2283 # precede EXT-X-MEDIA tags in HLS manifest such as [3].
2284 for line
in m3u8_doc
.splitlines():
2285 if line
.startswith('#EXT-X-MEDIA:'):
2288 for line
in m3u8_doc
.splitlines():
2289 if line
.startswith('#EXT-X-STREAM-INF:'):
2290 last_stream_inf
= parse_m3u8_attributes(line
)
2291 elif line
.startswith('#') or not line
.strip():
2294 tbr
= float_or_none(
2295 last_stream_inf
.get('AVERAGE-BANDWIDTH')
2296 or last_stream_inf
.get('BANDWIDTH'), scale
=1000)
2297 manifest_url
= format_url(line
.strip())
2299 for idx
in _extract_m3u8_playlist_indices(manifest_url
):
2300 format_id
= [m3u8_id
, None, idx
]
2301 # Bandwidth of live streams may differ over time thus making
2302 # format_id unpredictable. So it's better to keep provided
2305 stream_name
= build_stream_name()
2306 format_id
[1] = stream_name
or '%d' % (tbr
or len(formats
))
2308 'format_id': join_nonempty(*format_id
),
2309 'format_index': idx
,
2310 'url': manifest_url
,
2311 'manifest_url': m3u8_url
,
2314 'fps': float_or_none(last_stream_inf
.get('FRAME-RATE')),
2315 'protocol': entry_protocol
,
2316 'preference': preference
,
2319 resolution
= last_stream_inf
.get('RESOLUTION')
2321 mobj
= re
.search(r
'(?P<width>\d+)[xX](?P<height>\d+)', resolution
)
2323 f
['width'] = int(mobj
.group('width'))
2324 f
['height'] = int(mobj
.group('height'))
2325 # Unified Streaming Platform
2327 r
'audio.*?(?:%3D|=)(\d+)(?:-video.*?(?:%3D|=)(\d+))?', f
['url'])
2329 abr
, vbr
= mobj
.groups()
2330 abr
, vbr
= float_or_none(abr
, 1000), float_or_none(vbr
, 1000)
2335 codecs
= parse_codecs(last_stream_inf
.get('CODECS'))
2337 audio_group_id
= last_stream_inf
.get('AUDIO')
2338 # As per [1, 4.3.4.1.1] any EXT-X-STREAM-INF tag which
2339 # references a rendition group MUST have a CODECS attribute.
2340 # However, this is not always respected, for example, [2]
2341 # contains EXT-X-STREAM-INF tag which references AUDIO
2342 # rendition group but does not have CODECS and despite
2343 # referencing an audio group it represents a complete
2344 # (with audio and video) format. So, for such cases we will
2345 # ignore references to rendition groups and treat them
2346 # as complete formats.
2347 if audio_group_id
and codecs
and f
.get('vcodec') != 'none':
2348 audio_group
= groups
.get(audio_group_id
)
2349 if audio_group
and audio_group
[0].get('URI'):
2350 # TODO: update acodec for audio only formats with
2352 f
['acodec'] = 'none'
2353 if not f
.get('ext'):
2354 f
['ext'] = 'm4a' if f
.get('vcodec') == 'none' else 'mp4'
2358 progressive_uri
= last_stream_inf
.get('PROGRESSIVE-URI')
2361 del http_f
['manifest_url']
2363 'format_id': f
['format_id'].replace('hls-', 'http-'),
2365 'url': progressive_uri
,
2367 formats
.append(http_f
)
2369 last_stream_inf
= {}
2370 return formats
, subtitles
2372 def _extract_m3u8_vod_duration(
2373 self
, m3u8_vod_url
, video_id
, note
=None, errnote
=None, data
=None, headers
={}, query={}
):
2375 m3u8_vod
= self
._download
_webpage
(
2376 m3u8_vod_url
, video_id
,
2377 note
='Downloading m3u8 VOD manifest' if note
is None else note
,
2378 errnote
='Failed to download VOD manifest' if errnote
is None else errnote
,
2379 fatal
=False, data
=data
, headers
=headers
, query
=query
)
2381 return self
._parse
_m
3u8_vod
_duration
(m3u8_vod
or '', video_id
)
2383 def _parse_m3u8_vod_duration(self
, m3u8_vod
, video_id
):
2384 if '#EXT-X-PLAYLIST-TYPE:VOD' not in m3u8_vod
:
2388 float(line
[len('#EXTINF:'):].split(',')[0])
2389 for line
in m3u8_vod
.splitlines() if line
.startswith('#EXTINF:'))) or None
2392 def _xpath_ns(path
, namespace
=None):
2396 for c
in path
.split('/'):
2397 if not c
or c
== '.':
2400 out
.append('{%s}%s' % (namespace
, c
))
2401 return '/'.join(out
)
2403 def _extract_smil_formats_and_subtitles(self
, smil_url
, video_id
, fatal
=True, f4m_params
=None, transform_source
=None):
2404 res
= self
._download
_smil
(smil_url
, video_id
, fatal
=fatal
, transform_source
=transform_source
)
2410 smil_url
= urlh
.geturl()
2412 namespace
= self
._parse
_smil
_namespace
(smil
)
2414 fmts
= self
._parse
_smil
_formats
(
2415 smil
, smil_url
, video_id
, namespace
=namespace
, f4m_params
=f4m_params
)
2416 subs
= self
._parse
_smil
_subtitles
(
2417 smil
, namespace
=namespace
)
2421 def _extract_smil_formats(self
, *args
, **kwargs
):
2422 fmts
, subs
= self
._extract
_smil
_formats
_and
_subtitles
(*args
, **kwargs
)
2424 self
._report
_ignoring
_subs
('SMIL')
2427 def _extract_smil_info(self
, smil_url
, video_id
, fatal
=True, f4m_params
=None):
2428 res
= self
._download
_smil
(smil_url
, video_id
, fatal
=fatal
)
2433 smil_url
= urlh
.geturl()
2435 return self
._parse
_smil
(smil
, smil_url
, video_id
, f4m_params
=f4m_params
)
2437 def _download_smil(self
, smil_url
, video_id
, fatal
=True, transform_source
=None):
2438 return self
._download
_xml
_handle
(
2439 smil_url
, video_id
, 'Downloading SMIL file',
2440 'Unable to download SMIL file', fatal
=fatal
, transform_source
=transform_source
)
2442 def _parse_smil(self
, smil
, smil_url
, video_id
, f4m_params
=None):
2443 namespace
= self
._parse
_smil
_namespace
(smil
)
2445 formats
= self
._parse
_smil
_formats
(
2446 smil
, smil_url
, video_id
, namespace
=namespace
, f4m_params
=f4m_params
)
2447 subtitles
= self
._parse
_smil
_subtitles
(smil
, namespace
=namespace
)
2449 video_id
= os
.path
.splitext(url_basename(smil_url
))[0]
2453 for meta
in smil
.findall(self
._xpath
_ns
('./head/meta', namespace
)):
2454 name
= meta
.attrib
.get('name')
2455 content
= meta
.attrib
.get('content')
2456 if not name
or not content
:
2458 if not title
and name
== 'title':
2460 elif not description
and name
in ('description', 'abstract'):
2461 description
= content
2462 elif not upload_date
and name
== 'date':
2463 upload_date
= unified_strdate(content
)
2466 'id': image
.get('type'),
2467 'url': image
.get('src'),
2468 'width': int_or_none(image
.get('width')),
2469 'height': int_or_none(image
.get('height')),
2470 } for image
in smil
.findall(self
._xpath
_ns
('.//image', namespace
)) if image
.get('src')]
2474 'title': title
or video_id
,
2475 'description': description
,
2476 'upload_date': upload_date
,
2477 'thumbnails': thumbnails
,
2479 'subtitles': subtitles
,
2482 def _parse_smil_namespace(self
, smil
):
2483 return self
._search
_regex
(
2484 r
'(?i)^{([^}]+)?}smil$', smil
.tag
, 'namespace', default
=None)
2486 def _parse_smil_formats(self
, smil
, smil_url
, video_id
, namespace
=None, f4m_params
=None, transform_rtmp_url
=None):
2488 for meta
in smil
.findall(self
._xpath
_ns
('./head/meta', namespace
)):
2489 b
= meta
.get('base') or meta
.get('httpBase')
2501 media
= smil
.findall(self
._xpath
_ns
('.//video', namespace
)) + smil
.findall(self
._xpath
_ns
('.//audio', namespace
))
2502 for medium
in media
:
2503 src
= medium
.get('src')
2504 if not src
or src
in srcs
:
2508 bitrate
= float_or_none(medium
.get('system-bitrate') or medium
.get('systemBitrate'), 1000)
2509 filesize
= int_or_none(medium
.get('size') or medium
.get('fileSize'))
2510 width
= int_or_none(medium
.get('width'))
2511 height
= int_or_none(medium
.get('height'))
2512 proto
= medium
.get('proto')
2513 ext
= medium
.get('ext')
2514 src_ext
= determine_ext(src
)
2515 streamer
= medium
.get('streamer') or base
2517 if proto
== 'rtmp' or streamer
.startswith('rtmp'):
2523 'format_id': 'rtmp-%d' % (rtmp_count
if bitrate
is None else bitrate
),
2525 'filesize': filesize
,
2529 if transform_rtmp_url
:
2530 streamer
, src
= transform_rtmp_url(streamer
, src
)
2531 formats
[-1].update({
2537 src_url
= src
if src
.startswith('http') else compat_urlparse
.urljoin(base
, src
)
2538 src_url
= src_url
.strip()
2540 if proto
== 'm3u8' or src_ext
== 'm3u8':
2541 m3u8_formats
= self
._extract
_m
3u8_formats
(
2542 src_url
, video_id
, ext
or 'mp4', m3u8_id
='hls', fatal
=False)
2543 if len(m3u8_formats
) == 1:
2545 m3u8_formats
[0].update({
2546 'format_id': 'hls-%d' % (m3u8_count
if bitrate
is None else bitrate
),
2551 formats
.extend(m3u8_formats
)
2552 elif src_ext
== 'f4m':
2557 'plugin': 'flowplayer-3.2.0.1',
2559 f4m_url
+= '&' if '?' in f4m_url
else '?'
2560 f4m_url
+= compat_urllib_parse_urlencode(f4m_params
)
2561 formats
.extend(self
._extract
_f
4m
_formats
(f4m_url
, video_id
, f4m_id
='hds', fatal
=False))
2562 elif src_ext
== 'mpd':
2563 formats
.extend(self
._extract
_mpd
_formats
(
2564 src_url
, video_id
, mpd_id
='dash', fatal
=False))
2565 elif re
.search(r
'\.ism/[Mm]anifest', src_url
):
2566 formats
.extend(self
._extract
_ism
_formats
(
2567 src_url
, video_id
, ism_id
='mss', fatal
=False))
2568 elif src_url
.startswith('http') and self
._is
_valid
_url
(src
, video_id
):
2572 'ext': ext
or src_ext
or 'flv',
2573 'format_id': 'http-%d' % (bitrate
or http_count
),
2575 'filesize': filesize
,
2580 for medium
in smil
.findall(self
._xpath
_ns
('.//imagestream', namespace
)):
2581 src
= medium
.get('src')
2582 if not src
or src
in srcs
:
2588 'format_id': 'imagestream-%d' % (imgs_count
),
2590 'ext': mimetype2ext(medium
.get('type')),
2593 'width': int_or_none(medium
.get('width')),
2594 'height': int_or_none(medium
.get('height')),
2595 'format_note': 'SMIL storyboards',
2600 def _parse_smil_subtitles(self
, smil
, namespace
=None, subtitles_lang
='en'):
2603 for num
, textstream
in enumerate(smil
.findall(self
._xpath
_ns
('.//textstream', namespace
))):
2604 src
= textstream
.get('src')
2605 if not src
or src
in urls
:
2608 ext
= textstream
.get('ext') or mimetype2ext(textstream
.get('type')) or determine_ext(src
)
2609 lang
= textstream
.get('systemLanguage') or textstream
.get('systemLanguageName') or textstream
.get('lang') or subtitles_lang
2610 subtitles
.setdefault(lang
, []).append({
2616 def _extract_xspf_playlist(self
, xspf_url
, playlist_id
, fatal
=True):
2617 res
= self
._download
_xml
_handle
(
2618 xspf_url
, playlist_id
, 'Downloading xpsf playlist',
2619 'Unable to download xspf manifest', fatal
=fatal
)
2624 xspf_url
= urlh
.geturl()
2626 return self
._parse
_xspf
(
2627 xspf
, playlist_id
, xspf_url
=xspf_url
,
2628 xspf_base_url
=base_url(xspf_url
))
2630 def _parse_xspf(self
, xspf_doc
, playlist_id
, xspf_url
=None, xspf_base_url
=None):
2632 'xspf': 'http://xspf.org/ns/0/',
2633 's1': 'http://static.streamone.nl/player/ns/0',
2637 for track
in xspf_doc
.findall(xpath_with_ns('./xspf:trackList/xspf:track', NS_MAP
)):
2639 track
, xpath_with_ns('./xspf:title', NS_MAP
), 'title', default
=playlist_id
)
2640 description
= xpath_text(
2641 track
, xpath_with_ns('./xspf:annotation', NS_MAP
), 'description')
2642 thumbnail
= xpath_text(
2643 track
, xpath_with_ns('./xspf:image', NS_MAP
), 'thumbnail')
2644 duration
= float_or_none(
2645 xpath_text(track
, xpath_with_ns('./xspf:duration', NS_MAP
), 'duration'), 1000)
2648 for location
in track
.findall(xpath_with_ns('./xspf:location', NS_MAP
)):
2649 format_url
= urljoin(xspf_base_url
, location
.text
)
2654 'manifest_url': xspf_url
,
2655 'format_id': location
.get(xpath_with_ns('s1:label', NS_MAP
)),
2656 'width': int_or_none(location
.get(xpath_with_ns('s1:width', NS_MAP
))),
2657 'height': int_or_none(location
.get(xpath_with_ns('s1:height', NS_MAP
))),
2659 self
._sort
_formats
(formats
)
2664 'description': description
,
2665 'thumbnail': thumbnail
,
2666 'duration': duration
,
2671 def _extract_mpd_formats(self
, *args
, **kwargs
):
2672 fmts
, subs
= self
._extract
_mpd
_formats
_and
_subtitles
(*args
, **kwargs
)
2674 self
._report
_ignoring
_subs
('DASH')
2677 def _extract_mpd_formats_and_subtitles(
2678 self
, mpd_url
, video_id
, mpd_id
=None, note
=None, errnote
=None,
2679 fatal
=True, data
=None, headers
={}, query={}
):
2680 res
= self
._download
_xml
_handle
(
2682 note
='Downloading MPD manifest' if note
is None else note
,
2683 errnote
='Failed to download MPD manifest' if errnote
is None else errnote
,
2684 fatal
=fatal
, data
=data
, headers
=headers
, query
=query
)
2691 # We could have been redirected to a new url when we retrieved our mpd file.
2692 mpd_url
= urlh
.geturl()
2693 mpd_base_url
= base_url(mpd_url
)
2695 return self
._parse
_mpd
_formats
_and
_subtitles
(
2696 mpd_doc
, mpd_id
, mpd_base_url
, mpd_url
)
2698 def _parse_mpd_formats(self
, *args
, **kwargs
):
2699 fmts
, subs
= self
._parse
_mpd
_formats
_and
_subtitles
(*args
, **kwargs
)
2701 self
._report
_ignoring
_subs
('DASH')
2704 def _parse_mpd_formats_and_subtitles(
2705 self
, mpd_doc
, mpd_id
=None, mpd_base_url
='', mpd_url
=None):
2707 Parse formats from MPD manifest.
2709 1. MPEG-DASH Standard, ISO/IEC 23009-1:2014(E),
2710 http://standards.iso.org/ittf/PubliclyAvailableStandards/c065274_ISO_IEC_23009-1_2014.zip
2711 2. https://en.wikipedia.org/wiki/Dynamic_Adaptive_Streaming_over_HTTP
2713 if not self
.get_param('dynamic_mpd', True):
2714 if mpd_doc
.get('type') == 'dynamic':
2717 namespace
= self
._search
_regex
(r
'(?i)^{([^}]+)?}MPD$', mpd_doc
.tag
, 'namespace', default
=None)
2720 return self
._xpath
_ns
(path
, namespace
)
2722 def is_drm_protected(element
):
2723 return element
.find(_add_ns('ContentProtection')) is not None
2725 def extract_multisegment_info(element
, ms_parent_info
):
2726 ms_info
= ms_parent_info
.copy()
2728 # As per [1, 5.3.9.2.2] SegmentList and SegmentTemplate share some
2729 # common attributes and elements. We will only extract relevant
2731 def extract_common(source
):
2732 segment_timeline
= source
.find(_add_ns('SegmentTimeline'))
2733 if segment_timeline
is not None:
2734 s_e
= segment_timeline
.findall(_add_ns('S'))
2736 ms_info
['total_number'] = 0
2739 r
= int(s
.get('r', 0))
2740 ms_info
['total_number'] += 1 + r
2741 ms_info
['s'].append({
2742 't': int(s
.get('t', 0)),
2743 # @d is mandatory (see [1, 5.3.9.6.2, Table 17, page 60])
2744 'd': int(s
.attrib
['d']),
2747 start_number
= source
.get('startNumber')
2749 ms_info
['start_number'] = int(start_number
)
2750 timescale
= source
.get('timescale')
2752 ms_info
['timescale'] = int(timescale
)
2753 segment_duration
= source
.get('duration')
2754 if segment_duration
:
2755 ms_info
['segment_duration'] = float(segment_duration
)
2757 def extract_Initialization(source
):
2758 initialization
= source
.find(_add_ns('Initialization'))
2759 if initialization
is not None:
2760 ms_info
['initialization_url'] = initialization
.attrib
['sourceURL']
2762 segment_list
= element
.find(_add_ns('SegmentList'))
2763 if segment_list
is not None:
2764 extract_common(segment_list
)
2765 extract_Initialization(segment_list
)
2766 segment_urls_e
= segment_list
.findall(_add_ns('SegmentURL'))
2768 ms_info
['segment_urls'] = [segment
.attrib
['media'] for segment
in segment_urls_e
]
2770 segment_template
= element
.find(_add_ns('SegmentTemplate'))
2771 if segment_template
is not None:
2772 extract_common(segment_template
)
2773 media
= segment_template
.get('media')
2775 ms_info
['media'] = media
2776 initialization
= segment_template
.get('initialization')
2778 ms_info
['initialization'] = initialization
2780 extract_Initialization(segment_template
)
2783 mpd_duration
= parse_duration(mpd_doc
.get('mediaPresentationDuration'))
2784 formats
, subtitles
= [], {}
2785 stream_numbers
= collections
.defaultdict(int)
2786 for period
in mpd_doc
.findall(_add_ns('Period')):
2787 period_duration
= parse_duration(period
.get('duration')) or mpd_duration
2788 period_ms_info
= extract_multisegment_info(period
, {
2792 for adaptation_set
in period
.findall(_add_ns('AdaptationSet')):
2793 adaption_set_ms_info
= extract_multisegment_info(adaptation_set
, period_ms_info
)
2794 for representation
in adaptation_set
.findall(_add_ns('Representation')):
2795 representation_attrib
= adaptation_set
.attrib
.copy()
2796 representation_attrib
.update(representation
.attrib
)
2797 # According to [1, 5.3.7.2, Table 9, page 41], @mimeType is mandatory
2798 mime_type
= representation_attrib
['mimeType']
2799 content_type
= representation_attrib
.get('contentType', mime_type
.split('/')[0])
2801 codec_str
= representation_attrib
.get('codecs', '')
2802 # Some kind of binary subtitle found in some youtube livestreams
2803 if mime_type
== 'application/x-rawcc':
2804 codecs
= {'scodec': codec_str}
2806 codecs
= parse_codecs(codec_str
)
2807 if content_type
not in ('video', 'audio', 'text'):
2808 if mime_type
== 'image/jpeg':
2809 content_type
= mime_type
2810 elif codecs
.get('vcodec', 'none') != 'none':
2811 content_type
= 'video'
2812 elif codecs
.get('acodec', 'none') != 'none':
2813 content_type
= 'audio'
2814 elif codecs
.get('scodec', 'none') != 'none':
2815 content_type
= 'text'
2816 elif mimetype2ext(mime_type
) in ('tt', 'dfxp', 'ttml', 'xml', 'json'):
2817 content_type
= 'text'
2819 self
.report_warning('Unknown MIME type %s in DASH manifest' % mime_type
)
2823 for element
in (representation
, adaptation_set
, period
, mpd_doc
):
2824 base_url_e
= element
.find(_add_ns('BaseURL'))
2825 if base_url_e
is not None:
2826 base_url
= base_url_e
.text
+ base_url
2827 if re
.match(r
'^https?://', base_url
):
2829 if mpd_base_url
and base_url
.startswith('/'):
2830 base_url
= compat_urlparse
.urljoin(mpd_base_url
, base_url
)
2831 elif mpd_base_url
and not re
.match(r
'^https?://', base_url
):
2832 if not mpd_base_url
.endswith('/'):
2834 base_url
= mpd_base_url
+ base_url
2835 representation_id
= representation_attrib
.get('id')
2836 lang
= representation_attrib
.get('lang')
2837 url_el
= representation
.find(_add_ns('BaseURL'))
2838 filesize
= int_or_none(url_el
.attrib
.get('{http://youtube.com/yt/2012/10/10}contentLength') if url_el
is not None else None)
2839 bandwidth
= int_or_none(representation_attrib
.get('bandwidth'))
2840 if representation_id
is not None:
2841 format_id
= representation_id
2843 format_id
= content_type
2845 format_id
= mpd_id
+ '-' + format_id
2846 if content_type
in ('video', 'audio'):
2848 'format_id': format_id
,
2849 'manifest_url': mpd_url
,
2850 'ext': mimetype2ext(mime_type
),
2851 'width': int_or_none(representation_attrib
.get('width')),
2852 'height': int_or_none(representation_attrib
.get('height')),
2853 'tbr': float_or_none(bandwidth
, 1000),
2854 'asr': int_or_none(representation_attrib
.get('audioSamplingRate')),
2855 'fps': int_or_none(representation_attrib
.get('frameRate')),
2856 'language': lang
if lang
not in ('mul', 'und', 'zxx', 'mis') else None,
2857 'format_note': 'DASH %s' % content_type
,
2858 'filesize': filesize
,
2859 'container': mimetype2ext(mime_type
) + '_dash',
2862 elif content_type
== 'text':
2864 'ext': mimetype2ext(mime_type
),
2865 'manifest_url': mpd_url
,
2866 'filesize': filesize
,
2868 elif content_type
== 'image/jpeg':
2869 # See test case in VikiIE
2870 # https://www.viki.com/videos/1175236v-choosing-spouse-by-lottery-episode-1
2872 'format_id': format_id
,
2874 'manifest_url': mpd_url
,
2875 'format_note': 'DASH storyboards (jpeg)',
2879 if is_drm_protected(adaptation_set
) or is_drm_protected(representation
):
2881 representation_ms_info
= extract_multisegment_info(representation
, adaption_set_ms_info
)
2883 def prepare_template(template_name
, identifiers
):
2884 tmpl
= representation_ms_info
[template_name
]
2885 # First of, % characters outside $...$ templates
2886 # must be escaped by doubling for proper processing
2887 # by % operator string formatting used further (see
2888 # https://github.com/ytdl-org/youtube-dl/issues/16867).
2894 in_template
= not in_template
2895 elif c
== '%' and not in_template
:
2897 # Next, $...$ templates are translated to their
2898 # %(...) counterparts to be used with % operator
2899 if representation_id
is not None:
2900 t
= t
.replace('$RepresentationID$', representation_id
)
2901 t
= re
.sub(r
'\$(%s)\$' % '|'.join(identifiers
), r
'%(\1)d', t
)
2902 t
= re
.sub(r
'\$(%s)%%([^$]+)\$' % '|'.join(identifiers
), r
'%(\1)\2', t
)
2903 t
.replace('$$', '$')
2906 # @initialization is a regular template like @media one
2907 # so it should be handled just the same way (see
2908 # https://github.com/ytdl-org/youtube-dl/issues/11605)
2909 if 'initialization' in representation_ms_info
:
2910 initialization_template
= prepare_template(
2912 # As per [1, 5.3.9.4.2, Table 15, page 54] $Number$ and
2913 # $Time$ shall not be included for @initialization thus
2914 # only $Bandwidth$ remains
2916 representation_ms_info
['initialization_url'] = initialization_template
% {
2917 'Bandwidth': bandwidth
,
2920 def location_key(location
):
2921 return 'url' if re
.match(r
'^https?://', location
) else 'path'
2923 if 'segment_urls' not in representation_ms_info
and 'media' in representation_ms_info
:
2925 media_template
= prepare_template('media', ('Number', 'Bandwidth', 'Time'))
2926 media_location_key
= location_key(media_template
)
2928 # As per [1, 5.3.9.4.4, Table 16, page 55] $Number$ and $Time$
2929 # can't be used at the same time
2930 if '%(Number' in media_template
and 's' not in representation_ms_info
:
2931 segment_duration
= None
2932 if 'total_number' not in representation_ms_info
and 'segment_duration' in representation_ms_info
:
2933 segment_duration
= float_or_none(representation_ms_info
['segment_duration'], representation_ms_info
['timescale'])
2934 representation_ms_info
['total_number'] = int(math
.ceil(
2935 float_or_none(period_duration
, segment_duration
, default
=0)))
2936 representation_ms_info
['fragments'] = [{
2937 media_location_key
: media_template
% {
2938 'Number': segment_number
,
2939 'Bandwidth': bandwidth
,
2941 'duration': segment_duration
,
2942 } for segment_number
in range(
2943 representation_ms_info
['start_number'],
2944 representation_ms_info
['total_number'] + representation_ms_info
['start_number'])]
2946 # $Number*$ or $Time$ in media template with S list available
2947 # Example $Number*$: http://www.svtplay.se/klipp/9023742/stopptid-om-bjorn-borg
2948 # Example $Time$: https://play.arkena.com/embed/avp/v2/player/media/b41dda37-d8e7-4d3f-b1b5-9a9db578bdfe/1/129411
2949 representation_ms_info
['fragments'] = []
2952 segment_number
= representation_ms_info
['start_number']
2954 def add_segment_url():
2955 segment_url
= media_template
% {
2956 'Time': segment_time
,
2957 'Bandwidth': bandwidth
,
2958 'Number': segment_number
,
2960 representation_ms_info
['fragments'].append({
2961 media_location_key
: segment_url
,
2962 'duration': float_or_none(segment_d
, representation_ms_info
['timescale']),
2965 for num
, s
in enumerate(representation_ms_info
['s']):
2966 segment_time
= s
.get('t') or segment_time
2970 for r
in range(s
.get('r', 0)):
2971 segment_time
+= segment_d
2974 segment_time
+= segment_d
2975 elif 'segment_urls' in representation_ms_info
and 's' in representation_ms_info
:
2977 # Example: https://www.youtube.com/watch?v=iXZV5uAYMJI
2978 # or any YouTube dashsegments video
2981 timescale
= representation_ms_info
['timescale']
2982 for s
in representation_ms_info
['s']:
2983 duration
= float_or_none(s
['d'], timescale
)
2984 for r
in range(s
.get('r', 0) + 1):
2985 segment_uri
= representation_ms_info
['segment_urls'][segment_index
]
2987 location_key(segment_uri
): segment_uri
,
2988 'duration': duration
,
2991 representation_ms_info
['fragments'] = fragments
2992 elif 'segment_urls' in representation_ms_info
:
2993 # Segment URLs with no SegmentTimeline
2994 # Example: https://www.seznam.cz/zpravy/clanek/cesko-zasahne-vitr-o-sile-vichrice-muze-byt-i-zivotu-nebezpecny-39091
2995 # https://github.com/ytdl-org/youtube-dl/pull/14844
2997 segment_duration
= float_or_none(
2998 representation_ms_info
['segment_duration'],
2999 representation_ms_info
['timescale']) if 'segment_duration' in representation_ms_info
else None
3000 for segment_url
in representation_ms_info
['segment_urls']:
3002 location_key(segment_url
): segment_url
,
3004 if segment_duration
:
3005 fragment
['duration'] = segment_duration
3006 fragments
.append(fragment
)
3007 representation_ms_info
['fragments'] = fragments
3008 # If there is a fragments key available then we correctly recognized fragmented media.
3009 # Otherwise we will assume unfragmented media with direct access. Technically, such
3010 # assumption is not necessarily correct since we may simply have no support for
3011 # some forms of fragmented media renditions yet, but for now we'll use this fallback.
3012 if 'fragments' in representation_ms_info
:
3014 # NB: mpd_url may be empty when MPD manifest is parsed from a string
3015 'url': mpd_url
or base_url
,
3016 'fragment_base_url': base_url
,
3018 'protocol': 'http_dash_segments' if mime_type
!= 'image/jpeg' else 'mhtml',
3020 if 'initialization_url' in representation_ms_info
:
3021 initialization_url
= representation_ms_info
['initialization_url']
3022 if not f
.get('url'):
3023 f
['url'] = initialization_url
3024 f
['fragments'].append({location_key(initialization_url): initialization_url}
)
3025 f
['fragments'].extend(representation_ms_info
['fragments'])
3026 if not period_duration
:
3027 period_duration
= try_get(
3028 representation_ms_info
,
3029 lambda r
: sum(frag
['duration'] for frag
in r
['fragments']), float)
3031 # Assuming direct URL to unfragmented media.
3033 if content_type
in ('video', 'audio', 'image/jpeg'):
3034 f
['manifest_stream_number'] = stream_numbers
[f
['url']]
3035 stream_numbers
[f
['url']] += 1
3037 elif content_type
== 'text':
3038 subtitles
.setdefault(lang
or 'und', []).append(f
)
3040 return formats
, subtitles
3042 def _extract_ism_formats(self
, *args
, **kwargs
):
3043 fmts
, subs
= self
._extract
_ism
_formats
_and
_subtitles
(*args
, **kwargs
)
3045 self
._report
_ignoring
_subs
('ISM')
3048 def _extract_ism_formats_and_subtitles(self
, ism_url
, video_id
, ism_id
=None, note
=None, errnote
=None, fatal
=True, data
=None, headers
={}, query={}
):
3049 res
= self
._download
_xml
_handle
(
3051 note
='Downloading ISM manifest' if note
is None else note
,
3052 errnote
='Failed to download ISM manifest' if errnote
is None else errnote
,
3053 fatal
=fatal
, data
=data
, headers
=headers
, query
=query
)
3060 return self
._parse
_ism
_formats
_and
_subtitles
(ism_doc
, urlh
.geturl(), ism_id
)
3062 def _parse_ism_formats_and_subtitles(self
, ism_doc
, ism_url
, ism_id
=None):
3064 Parse formats from ISM manifest.
3066 1. [MS-SSTR]: Smooth Streaming Protocol,
3067 https://msdn.microsoft.com/en-us/library/ff469518.aspx
3069 if ism_doc
.get('IsLive') == 'TRUE':
3072 duration
= int(ism_doc
.attrib
['Duration'])
3073 timescale
= int_or_none(ism_doc
.get('TimeScale')) or 10000000
3077 for stream
in ism_doc
.findall('StreamIndex'):
3078 stream_type
= stream
.get('Type')
3079 if stream_type
not in ('video', 'audio', 'text'):
3081 url_pattern
= stream
.attrib
['Url']
3082 stream_timescale
= int_or_none(stream
.get('TimeScale')) or timescale
3083 stream_name
= stream
.get('Name')
3084 stream_language
= stream
.get('Language', 'und')
3085 for track
in stream
.findall('QualityLevel'):
3086 fourcc
= track
.get('FourCC') or ('AACL' if track
.get('AudioTag') == '255' else None)
3087 # TODO: add support for WVC1 and WMAP
3088 if fourcc
not in ('H264', 'AVC1', 'AACL', 'TTML'):
3089 self
.report_warning('%s is not a supported codec' % fourcc
)
3091 tbr
= int(track
.attrib
['Bitrate']) // 1000
3092 # [1] does not mention Width and Height attributes. However,
3093 # they're often present while MaxWidth and MaxHeight are
3094 # missing, so should be used as fallbacks
3095 width
= int_or_none(track
.get('MaxWidth') or track
.get('Width'))
3096 height
= int_or_none(track
.get('MaxHeight') or track
.get('Height'))
3097 sampling_rate
= int_or_none(track
.get('SamplingRate'))
3099 track_url_pattern
= re
.sub(r
'{[Bb]itrate}', track
.attrib
['Bitrate'], url_pattern
)
3100 track_url_pattern
= compat_urlparse
.urljoin(ism_url
, track_url_pattern
)
3106 stream_fragments
= stream
.findall('c')
3107 for stream_fragment_index
, stream_fragment
in enumerate(stream_fragments
):
3108 fragment_ctx
['time'] = int_or_none(stream_fragment
.get('t')) or fragment_ctx
['time']
3109 fragment_repeat
= int_or_none(stream_fragment
.get('r')) or 1
3110 fragment_ctx
['duration'] = int_or_none(stream_fragment
.get('d'))
3111 if not fragment_ctx
['duration']:
3113 next_fragment_time
= int(stream_fragment
[stream_fragment_index
+ 1].attrib
['t'])
3115 next_fragment_time
= duration
3116 fragment_ctx
['duration'] = (next_fragment_time
- fragment_ctx
['time']) / fragment_repeat
3117 for _
in range(fragment_repeat
):
3119 'url': re
.sub(r
'{start[ _]time}', compat_str(fragment_ctx
['time']), track_url_pattern
),
3120 'duration': fragment_ctx
['duration'] / stream_timescale
,
3122 fragment_ctx
['time'] += fragment_ctx
['duration']
3124 if stream_type
== 'text':
3125 subtitles
.setdefault(stream_language
, []).append({
3129 'manifest_url': ism_url
,
3130 'fragments': fragments
,
3131 '_download_params': {
3132 'stream_type': stream_type
,
3133 'duration': duration
,
3134 'timescale': stream_timescale
,
3136 'language': stream_language
,
3137 'codec_private_data': track
.get('CodecPrivateData'),
3140 elif stream_type
in ('video', 'audio'):
3142 'format_id': join_nonempty(ism_id
, stream_name
, tbr
),
3144 'manifest_url': ism_url
,
3145 'ext': 'ismv' if stream_type
== 'video' else 'isma',
3149 'asr': sampling_rate
,
3150 'vcodec': 'none' if stream_type
== 'audio' else fourcc
,
3151 'acodec': 'none' if stream_type
== 'video' else fourcc
,
3153 'fragments': fragments
,
3154 'has_drm': ism_doc
.find('Protection') is not None,
3155 '_download_params': {
3156 'stream_type': stream_type
,
3157 'duration': duration
,
3158 'timescale': stream_timescale
,
3159 'width': width
or 0,
3160 'height': height
or 0,
3162 'language': stream_language
,
3163 'codec_private_data': track
.get('CodecPrivateData'),
3164 'sampling_rate': sampling_rate
,
3165 'channels': int_or_none(track
.get('Channels', 2)),
3166 'bits_per_sample': int_or_none(track
.get('BitsPerSample', 16)),
3167 'nal_unit_length_field': int_or_none(track
.get('NALUnitLengthField', 4)),
3170 return formats
, subtitles
3172 def _parse_html5_media_entries(self
, base_url
, webpage
, video_id
, m3u8_id
=None, m3u8_entry_protocol
='m3u8_native', mpd_id
=None, preference
=None, quality
=None):
3173 def absolute_url(item_url
):
3174 return urljoin(base_url
, item_url
)
3176 def parse_content_type(content_type
):
3177 if not content_type
:
3179 ctr
= re
.search(r
'(?P<mimetype>[^/]+/[^;]+)(?:;\s*codecs="?(?P<codecs>[^"]+))?', content_type
)
3181 mimetype
, codecs
= ctr
.groups()
3182 f
= parse_codecs(codecs
)
3183 f
['ext'] = mimetype2ext(mimetype
)
3187 def _media_formats(src
, cur_media_type
, type_info
=None):
3188 type_info
= type_info
or {}
3189 full_url
= absolute_url(src
)
3190 ext
= type_info
.get('ext') or determine_ext(full_url
)
3192 is_plain_url
= False
3193 formats
= self
._extract
_m
3u8_formats
(
3194 full_url
, video_id
, ext
='mp4',
3195 entry_protocol
=m3u8_entry_protocol
, m3u8_id
=m3u8_id
,
3196 preference
=preference
, quality
=quality
, fatal
=False)
3198 is_plain_url
= False
3199 formats
= self
._extract
_mpd
_formats
(
3200 full_url
, video_id
, mpd_id
=mpd_id
, fatal
=False)
3205 'vcodec': 'none' if cur_media_type
== 'audio' else None,
3208 return is_plain_url
, formats
3211 # amp-video and amp-audio are very similar to their HTML5 counterparts
3212 # so we wll include them right here (see
3213 # https://www.ampproject.org/docs/reference/components/amp-video)
3214 # For dl8-* tags see https://delight-vr.com/documentation/dl8-video/
3215 _MEDIA_TAG_NAME_RE
= r
'(?:(?:amp|dl8(?:-live)?)-)?(video|audio)'
3216 media_tags
= [(media_tag
, media_tag_name
, media_type
, '')
3217 for media_tag
, media_tag_name
, media_type
3218 in re
.findall(r
'(?s)(<(%s)[^>]*/>)' % _MEDIA_TAG_NAME_RE
, webpage
)]
3219 media_tags
.extend(re
.findall(
3220 # We only allow video|audio followed by a whitespace or '>'.
3221 # Allowing more characters may end up in significant slow down (see
3222 # https://github.com/ytdl-org/youtube-dl/issues/11979, example URL:
3223 # http://www.porntrex.com/maps/videositemap.xml).
3224 r
'(?s)(<(?P<tag>%s)(?:\s+[^>]*)?>)(.*?)</(?P=tag)>' % _MEDIA_TAG_NAME_RE
, webpage
))
3225 for media_tag
, _
, media_type
, media_content
in media_tags
:
3230 media_attributes
= extract_attributes(media_tag
)
3231 src
= strip_or_none(media_attributes
.get('src'))
3233 f
= parse_content_type(media_attributes
.get('type'))
3234 _
, formats
= _media_formats(src
, media_type
, f
)
3235 media_info
['formats'].extend(formats
)
3236 media_info
['thumbnail'] = absolute_url(media_attributes
.get('poster'))
3238 for source_tag
in re
.findall(r
'<source[^>]+>', media_content
):
3239 s_attr
= extract_attributes(source_tag
)
3240 # data-video-src and data-src are non standard but seen
3241 # several times in the wild
3242 src
= strip_or_none(dict_get(s_attr
, ('src', 'data-video-src', 'data-src')))
3245 f
= parse_content_type(s_attr
.get('type'))
3246 is_plain_url
, formats
= _media_formats(src
, media_type
, f
)
3248 # width, height, res, label and title attributes are
3249 # all not standard but seen several times in the wild
3252 for lbl
in ('label', 'title')
3253 if str_or_none(s_attr
.get(lbl
))
3255 width
= int_or_none(s_attr
.get('width'))
3256 height
= (int_or_none(s_attr
.get('height'))
3257 or int_or_none(s_attr
.get('res')))
3258 if not width
or not height
:
3260 resolution
= parse_resolution(lbl
)
3263 width
= width
or resolution
.get('width')
3264 height
= height
or resolution
.get('height')
3266 tbr
= parse_bitrate(lbl
)
3275 'format_id': s_attr
.get('label') or s_attr
.get('title'),
3277 f
.update(formats
[0])
3278 media_info
['formats'].append(f
)
3280 media_info
['formats'].extend(formats
)
3281 for track_tag
in re
.findall(r
'<track[^>]+>', media_content
):
3282 track_attributes
= extract_attributes(track_tag
)
3283 kind
= track_attributes
.get('kind')
3284 if not kind
or kind
in ('subtitles', 'captions'):
3285 src
= strip_or_none(track_attributes
.get('src'))
3288 lang
= track_attributes
.get('srclang') or track_attributes
.get('lang') or track_attributes
.get('label')
3289 media_info
['subtitles'].setdefault(lang
, []).append({
3290 'url': absolute_url(src
),
3292 for f
in media_info
['formats']:
3293 f
.setdefault('http_headers', {})['Referer'] = base_url
3294 if media_info
['formats'] or media_info
['subtitles']:
3295 entries
.append(media_info
)
3298 def _extract_akamai_formats(self
, *args
, **kwargs
):
3299 fmts
, subs
= self
._extract
_akamai
_formats
_and
_subtitles
(*args
, **kwargs
)
3301 self
._report
_ignoring
_subs
('akamai')
3304 def _extract_akamai_formats_and_subtitles(self
, manifest_url
, video_id
, hosts
={}):
3305 signed
= 'hdnea=' in manifest_url
3307 # https://learn.akamai.com/en-us/webhelp/media-services-on-demand/stream-packaging-user-guide/GUID-BE6C0F73-1E06-483B-B0EA-57984B91B7F9.html
3308 manifest_url
= re
.sub(
3309 r
'(?:b=[\d,-]+|(?:__a__|attributes)=off|__b__=\d+)&?',
3310 '', manifest_url
).strip('?')
3315 hdcore_sign
= 'hdcore=3.7.0'
3316 f4m_url
= re
.sub(r
'(https?://[^/]+)/i/', r
'\1/z/', manifest_url
).replace('/master.m3u8', '/manifest.f4m')
3317 hds_host
= hosts
.get('hds')
3319 f4m_url
= re
.sub(r
'(https?://)[^/]+', r
'\1' + hds_host
, f4m_url
)
3320 if 'hdcore=' not in f4m_url
:
3321 f4m_url
+= ('&' if '?' in f4m_url
else '?') + hdcore_sign
3322 f4m_formats
= self
._extract
_f
4m
_formats
(
3323 f4m_url
, video_id
, f4m_id
='hds', fatal
=False)
3324 for entry
in f4m_formats
:
3325 entry
.update({'extra_param_to_segment_url': hdcore_sign}
)
3326 formats
.extend(f4m_formats
)
3328 m3u8_url
= re
.sub(r
'(https?://[^/]+)/z/', r
'\1/i/', manifest_url
).replace('/manifest.f4m', '/master.m3u8')
3329 hls_host
= hosts
.get('hls')
3331 m3u8_url
= re
.sub(r
'(https?://)[^/]+', r
'\1' + hls_host
, m3u8_url
)
3332 m3u8_formats
, m3u8_subtitles
= self
._extract
_m
3u8_formats
_and
_subtitles
(
3333 m3u8_url
, video_id
, 'mp4', 'm3u8_native',
3334 m3u8_id
='hls', fatal
=False)
3335 formats
.extend(m3u8_formats
)
3336 subtitles
= self
._merge
_subtitles
(subtitles
, m3u8_subtitles
)
3338 http_host
= hosts
.get('http')
3339 if http_host
and m3u8_formats
and not signed
:
3340 REPL_REGEX
= r
'https?://[^/]+/i/([^,]+),([^/]+),([^/]+)\.csmil/.+'
3341 qualities
= re
.match(REPL_REGEX
, m3u8_url
).group(2).split(',')
3342 qualities_length
= len(qualities
)
3343 if len(m3u8_formats
) in (qualities_length
, qualities_length
+ 1):
3345 for f
in m3u8_formats
:
3346 if f
['vcodec'] != 'none':
3347 for protocol
in ('http', 'https'):
3349 del http_f
['manifest_url']
3351 REPL_REGEX
, protocol
+ fr
'://{http_host}/\g<1>{qualities[i]}\3', f
['url'])
3353 'format_id': http_f
['format_id'].replace('hls-', protocol
+ '-'),
3355 'protocol': protocol
,
3357 formats
.append(http_f
)
3360 return formats
, subtitles
3362 def _extract_wowza_formats(self
, url
, video_id
, m3u8_entry_protocol
='m3u8_native', skip_protocols
=[]):
3363 query
= compat_urlparse
.urlparse(url
).query
3364 url
= re
.sub(r
'/(?:manifest|playlist|jwplayer)\.(?:m3u8|f4m|mpd|smil)', '', url
)
3366 r
'(?:(?:http|rtmp|rtsp)(?P<s>s)?:)?(?P<url>//[^?]+)', url
)
3367 url_base
= mobj
.group('url')
3368 http_base_url
= '%s%s:%s' % ('http', mobj
.group('s') or '', url_base
)
3371 def manifest_url(manifest
):
3372 m_url
= f
'{http_base_url}/{manifest}'
3374 m_url
+= '?%s' % query
3377 if 'm3u8' not in skip_protocols
:
3378 formats
.extend(self
._extract
_m
3u8_formats
(
3379 manifest_url('playlist.m3u8'), video_id
, 'mp4',
3380 m3u8_entry_protocol
, m3u8_id
='hls', fatal
=False))
3381 if 'f4m' not in skip_protocols
:
3382 formats
.extend(self
._extract
_f
4m
_formats
(
3383 manifest_url('manifest.f4m'),
3384 video_id
, f4m_id
='hds', fatal
=False))
3385 if 'dash' not in skip_protocols
:
3386 formats
.extend(self
._extract
_mpd
_formats
(
3387 manifest_url('manifest.mpd'),
3388 video_id
, mpd_id
='dash', fatal
=False))
3389 if re
.search(r
'(?:/smil:|\.smil)', url_base
):
3390 if 'smil' not in skip_protocols
:
3391 rtmp_formats
= self
._extract
_smil
_formats
(
3392 manifest_url('jwplayer.smil'),
3393 video_id
, fatal
=False)
3394 for rtmp_format
in rtmp_formats
:
3395 rtsp_format
= rtmp_format
.copy()
3396 rtsp_format
['url'] = '%s/%s' % (rtmp_format
['url'], rtmp_format
['play_path'])
3397 del rtsp_format
['play_path']
3398 del rtsp_format
['ext']
3399 rtsp_format
.update({
3400 'url': rtsp_format
['url'].replace('rtmp://', 'rtsp://'),
3401 'format_id': rtmp_format
['format_id'].replace('rtmp', 'rtsp'),
3404 formats
.extend([rtmp_format
, rtsp_format
])
3406 for protocol
in ('rtmp', 'rtsp'):
3407 if protocol
not in skip_protocols
:
3409 'url': f
'{protocol}:{url_base}',
3410 'format_id': protocol
,
3411 'protocol': protocol
,
3415 def _find_jwplayer_data(self
, webpage
, video_id
=None, transform_source
=js_to_json
):
3417 r
'(?s)jwplayer\((?P<quote>[\'"])[^\'" ]+(?P
=quote
)\
)(?
!</script
>).*?\
.setup\s
*\
((?P
<options
>[^
)]+)\
)',
3421 jwplayer_data = self._parse_json(mobj.group('options
'),
3423 transform_source=transform_source)
3424 except ExtractorError:
3427 if isinstance(jwplayer_data, dict):
3428 return jwplayer_data
3430 def _extract_jwplayer_data(self, webpage, video_id, *args, **kwargs):
3431 jwplayer_data = self._find_jwplayer_data(
3432 webpage, video_id, transform_source=js_to_json)
3433 return self._parse_jwplayer_data(
3434 jwplayer_data, video_id, *args, **kwargs)
3436 def _parse_jwplayer_data(self, jwplayer_data, video_id=None, require_title=True,
3437 m3u8_id=None, mpd_id=None, rtmp_params=None, base_url=None):
3438 # JWPlayer backward compatibility: flattened playlists
3439 # https://github.com/jwplayer/jwplayer/blob/v7.4.3/src/js/api/config.js#L81-L96
3440 if 'playlist
' not in jwplayer_data:
3441 jwplayer_data = {'playlist': [jwplayer_data]}
3445 # JWPlayer backward compatibility: single playlist item
3446 # https://github.com/jwplayer/jwplayer/blob/v7.7.0/src/js/playlist/playlist.js#L10
3447 if not isinstance(jwplayer_data['playlist
'], list):
3448 jwplayer_data['playlist
'] = [jwplayer_data['playlist
']]
3450 for video_data in jwplayer_data['playlist
']:
3451 # JWPlayer backward compatibility: flattened sources
3452 # https://github.com/jwplayer/jwplayer/blob/v7.4.3/src/js/playlist/item.js#L29-L35
3453 if 'sources
' not in video_data:
3454 video_data['sources
'] = [video_data]
3456 this_video_id = video_id or video_data['mediaid
']
3458 formats = self._parse_jwplayer_formats(
3459 video_data['sources
'], video_id=this_video_id, m3u8_id=m3u8_id,
3460 mpd_id=mpd_id, rtmp_params=rtmp_params, base_url=base_url)
3463 tracks = video_data.get('tracks
')
3464 if tracks and isinstance(tracks, list):
3465 for track in tracks:
3466 if not isinstance(track, dict):
3468 track_kind = track.get('kind
')
3469 if not track_kind or not isinstance(track_kind, compat_str):
3471 if track_kind.lower() not in ('captions
', 'subtitles
'):
3473 track_url = urljoin(base_url, track.get('file'))
3476 subtitles.setdefault(track.get('label
') or 'en
', []).append({
3477 'url
': self._proto_relative_url(track_url)
3481 'id': this_video_id,
3482 'title
': unescapeHTML(video_data['title
'] if require_title else video_data.get('title
')),
3483 'description
': clean_html(video_data.get('description
')),
3484 'thumbnail
': urljoin(base_url, self._proto_relative_url(video_data.get('image
'))),
3485 'timestamp
': int_or_none(video_data.get('pubdate
')),
3486 'duration
': float_or_none(jwplayer_data.get('duration
') or video_data.get('duration
')),
3487 'subtitles
': subtitles,
3489 # https://github.com/jwplayer/jwplayer/blob/master/src/js/utils/validator.js#L32
3490 if len(formats) == 1 and re.search(r'^
(?
:http|
//).*(?
:youtube\
.com|youtu\
.be
)/.+', formats[0]['url
']):
3492 '_type
': 'url_transparent
',
3493 'url
': formats[0]['url
'],
3496 self._sort_formats(formats)
3497 entry['formats
'] = formats
3498 entries.append(entry)
3499 if len(entries) == 1:
3502 return self.playlist_result(entries)
3504 def _parse_jwplayer_formats(self, jwplayer_sources_data, video_id=None,
3505 m3u8_id=None, mpd_id=None, rtmp_params=None, base_url=None):
3508 for source in jwplayer_sources_data:
3509 if not isinstance(source, dict):
3511 source_url = urljoin(
3512 base_url, self._proto_relative_url(source.get('file')))
3513 if not source_url or source_url in urls:
3515 urls.append(source_url)
3516 source_type = source.get('type') or ''
3517 ext = mimetype2ext(source_type) or determine_ext(source_url)
3518 if source_type == 'hls
' or ext == 'm3u8
':
3519 formats.extend(self._extract_m3u8_formats(
3520 source_url, video_id, 'mp4
', entry_protocol='m3u8_native
',
3521 m3u8_id=m3u8_id, fatal=False))
3522 elif source_type == 'dash
' or ext == 'mpd
':
3523 formats.extend(self._extract_mpd_formats(
3524 source_url, video_id, mpd_id=mpd_id, fatal=False))
3526 formats.extend(self._extract_smil_formats(
3527 source_url, video_id, fatal=False))
3528 # https://github.com/jwplayer/jwplayer/blob/master/src/js/providers/default.js#L67
3529 elif source_type.startswith('audio
') or ext in (
3530 'oga
', 'aac
', 'mp3
', 'mpeg
', 'vorbis
'):
3537 height = int_or_none(source.get('height
'))
3539 # Often no height is provided but there is a label in
3540 # format like "1080p", "720p SD", or 1080.
3541 height = int_or_none(self._search_regex(
3542 r'^
(\d{3,4}
)[pP
]?
(?
:\b|$
)', compat_str(source.get('label
') or ''),
3543 'height
', default=None))
3546 'width
': int_or_none(source.get('width
')),
3548 'tbr
': int_or_none(source.get('bitrate
')),
3551 if source_url.startswith('rtmp
'):
3552 a_format['ext
'] = 'flv
'
3553 # See com/longtailvideo/jwplayer/media/RTMPMediaProvider.as
3554 # of jwplayer.flash.swf
3555 rtmp_url_parts = re.split(
3556 r'((?
:mp4|mp3|flv
):)', source_url, 1)
3557 if len(rtmp_url_parts) == 3:
3558 rtmp_url, prefix, play_path = rtmp_url_parts
3561 'play_path
': prefix + play_path,
3564 a_format.update(rtmp_params)
3565 formats.append(a_format)
3568 def _live_title(self, name):
3569 self._downloader.deprecation_warning('yt_dlp
.InfoExtractor
._live
_title
is deprecated
and does
not work
as expected
')
3572 def _int(self, v, name, fatal=False, **kwargs):
3573 res = int_or_none(v, **kwargs)
3575 msg = f'Failed to extract {name}
: Could
not parse value {v!r}
'
3577 raise ExtractorError(msg)
3579 self.report_warning(msg)
3582 def _float(self, v, name, fatal=False, **kwargs):
3583 res = float_or_none(v, **kwargs)
3585 msg = f'Failed to extract {name}
: Could
not parse value {v!r}
'
3587 raise ExtractorError(msg)
3589 self.report_warning(msg)
3592 def _set_cookie(self, domain, name, value, expire_time=None, port=None,
3593 path='/', secure=False, discard=False, rest={}, **kwargs):
3594 cookie = compat_cookiejar_Cookie(
3595 0, name, value, port, port is not None, domain, True,
3596 domain.startswith('.'), path, True, secure, expire_time,
3597 discard, None, None, rest)
3598 self._downloader.cookiejar.set_cookie(cookie)
3600 def _get_cookies(self, url):
3601 """ Return a compat_cookies_SimpleCookie with the cookies for the url """
3602 return compat_cookies_SimpleCookie(self._downloader._calc_cookies(url))
3604 def _apply_first_set_cookie_header(self, url_handle, cookie):
3606 Apply first Set-Cookie header instead of the last. Experimental.
3608 Some sites (e.g. [1-3]) may serve two cookies under the same name
3609 in Set-Cookie header and expect the first (old) one to be set rather
3610 than second (new). However, as of RFC6265 the newer one cookie
3611 should be set into cookie store what actually happens.
3612 We will workaround this issue by resetting the cookie to
3613 the first one manually.
3614 1. https://new.vk.com/
3615 2. https://github.com/ytdl-org/youtube-dl/issues/9841#issuecomment-227871201
3616 3. https://learning.oreilly.com/
3618 for header, cookies in url_handle.headers.items():
3619 if header.lower() != 'set-cookie
':
3621 cookies = cookies.encode('iso
-8859-1').decode('utf
-8')
3622 cookie_value = re.search(
3623 r'%s=(.+?
);.*?
\b[Dd
]omain
=(.+?
)(?
:[,;]|$
)' % cookie, cookies)
3625 value, domain = cookie_value.groups()
3626 self._set_cookie(domain, cookie, value)
3630 def get_testcases(cls, include_onlymatching=False):
3631 t = getattr(cls, '_TEST
', None)
3633 assert not hasattr(cls, '_TESTS
'), f'{cls.ie_key()}IE has _TEST
and _TESTS
'
3636 tests = getattr(cls, '_TESTS
', [])
3638 if not include_onlymatching and t.get('only_matching
', False):
3640 t['name
'] = cls.ie_key()
3645 """Get age limit from the testcases"""
3646 return max(traverse_obj(
3647 tuple(cls.get_testcases(include_onlymatching=False)),
3648 (..., (('playlist
', 0), None), 'info_dict
', 'age_limit
')) or [0])
3651 def is_suitable(cls, age_limit):
3652 """Test whether the extractor is generally suitable for the given age limit"""
3653 return not age_restricted(cls.age_limit, age_limit)
3656 def description(cls, *, markdown=True, search_examples=None):
3657 """Description of the extractor"""
3659 if cls._NETRC_MACHINE:
3661 desc += f' [<abbr title
="netrc machine"><em
>{cls._NETRC_MACHINE}
</em
></abbr
>]'
3663 desc += f' [{cls._NETRC_MACHINE}
]'
3664 if cls.IE_DESC is False:
3667 desc += f' {cls.IE_DESC}
'
3669 desc += f'; "{cls.SEARCH_KEY}:" prefix
'
3671 _COUNTS = ('', '5', '10', 'all
')
3672 desc += f' (Example
: "{cls.SEARCH_KEY}{random.choice(_COUNTS)}:{random.choice(search_examples)}")'
3673 if not cls.working():
3674 desc += ' (**Currently broken
**)' if markdown else ' (Currently broken
)'
3676 name = f' - **{cls.IE_NAME}
**' if markdown else cls.IE_NAME
3677 return f'{name}
:{desc}
' if desc else name
3679 def extract_subtitles(self, *args, **kwargs):
3680 if (self.get_param('writesubtitles
', False)
3681 or self.get_param('listsubtitles
')):
3682 return self._get_subtitles(*args, **kwargs)
3685 def _get_subtitles(self, *args, **kwargs):
3686 raise NotImplementedError('This method must be implemented by subclasses
')
3688 def extract_comments(self, *args, **kwargs):
3689 if not self.get_param('getcomments
'):
3691 generator = self._get_comments(*args, **kwargs)
3698 comments.append(next(generator))
3699 except StopIteration:
3701 except KeyboardInterrupt:
3702 self.to_screen('Interrupted by user
')
3703 except Exception as e:
3704 if self.get_param('ignoreerrors
') is not True:
3706 self._downloader.report_error(e)
3707 comment_count = len(comments)
3708 self.to_screen(f'Extracted {comment_count} comments
')
3710 'comments
': comments,
3711 'comment_count
': None if interrupted else comment_count
3715 def _get_comments(self, *args, **kwargs):
3716 raise NotImplementedError('This method must be implemented by subclasses
')
3719 def _merge_subtitle_items(subtitle_list1, subtitle_list2):
3720 """ Merge subtitle items for one language. Items with duplicated URLs/data
3721 will be dropped. """
3722 list1_data = {(item.get('url'), item.get('data')) for item in subtitle_list1}
3723 ret = list(subtitle_list1)
3724 ret.extend(item for item in subtitle_list2 if (item.get('url
'), item.get('data
')) not in list1_data)
3728 def _merge_subtitles(cls, *dicts, target=None):
3729 """ Merge subtitle dictionaries, language by language. """
3733 for lang, subs in d.items():
3734 target[lang] = cls._merge_subtitle_items(target.get(lang, []), subs)
3737 def extract_automatic_captions(self, *args, **kwargs):
3738 if (self.get_param('writeautomaticsub
', False)
3739 or self.get_param('listsubtitles
')):
3740 return self._get_automatic_captions(*args, **kwargs)
3743 def _get_automatic_captions(self, *args, **kwargs):
3744 raise NotImplementedError('This method must be implemented by subclasses
')
3746 @functools.cached_property
3747 def _cookies_passed(self):
3748 """Whether cookies have been passed to YoutubeDL"""
3749 return self.get_param('cookiefile
') is not None or self.get_param('cookiesfrombrowser
') is not None
3751 def mark_watched(self, *args, **kwargs):
3752 if not self.get_param('mark_watched
', False):
3754 if self.supports_login() and self._get_login_info()[0] is not None or self._cookies_passed:
3755 self._mark_watched(*args, **kwargs)
3757 def _mark_watched(self, *args, **kwargs):
3758 raise NotImplementedError('This method must be implemented by subclasses
')
3760 def geo_verification_headers(self):
3762 geo_verification_proxy = self.get_param('geo_verification_proxy
')
3763 if geo_verification_proxy:
3764 headers['Ytdl
-request
-proxy
'] = geo_verification_proxy
3767 def _generic_id(self, url):
3768 return compat_urllib_parse_unquote(os.path.splitext(url.rstrip('/').split('/')[-1])[0])
3770 def _generic_title(self, url):
3771 return compat_urllib_parse_unquote(os.path.splitext(url_basename(url))[0])
3774 def _availability(is_private=None, needs_premium=None, needs_subscription=None, needs_auth=None, is_unlisted=None):
3775 all_known = all(map(
3776 lambda x: x is not None,
3777 (is_private, needs_premium, needs_subscription, needs_auth, is_unlisted)))
3779 'private
' if is_private
3780 else 'premium_only
' if needs_premium
3781 else 'subscriber_only
' if needs_subscription
3782 else 'needs_auth
' if needs_auth
3783 else 'unlisted
' if is_unlisted
3784 else 'public
' if all_known
3787 def _configuration_arg(self, key, default=NO_DEFAULT, *, ie_key=None, casesense=False):
3789 @returns A list of values for the extractor argument given by "key"
3790 or "default" if no such key is present
3791 @param default The default value to return when the key is not present (default: [])
3792 @param casesense When false, the values are converted to lower case
3795 self._downloader.params, ('extractor_args
', (ie_key or self.ie_key()).lower(), key))
3797 return [] if default is NO_DEFAULT else default
3798 return list(val) if casesense else [x.lower() for x in val]
3800 def _yes_playlist(self, playlist_id, video_id, smuggled_data=None, *, playlist_label='playlist
', video_label='video
'):
3801 if not playlist_id or not video_id:
3804 no_playlist = (smuggled_data or {}).get('force_noplaylist
')
3805 if no_playlist is not None:
3806 return not no_playlist
3808 video_id = '' if video_id is True else f' {video_id}
'
3809 playlist_id = '' if playlist_id is True else f' {playlist_id}
'
3810 if self.get_param('noplaylist
'):
3811 self.to_screen(f'Downloading just the {video_label}{video_id} because of
--no
-playlist
')
3813 self.to_screen(f'Downloading {playlist_label}{playlist_id}
- add
--no
-playlist to download just the {video_label}{video_id}
')
3817 class SearchInfoExtractor(InfoExtractor):
3819 Base class for paged search queries extractors.
3820 They accept URLs in the format _SEARCH_KEY(|all|[0-9]):{query}
3821 Instances should define _SEARCH_KEY and optionally _MAX_RESULTS
3824 _MAX_RESULTS = float('inf
')
3827 def _make_valid_url(cls):
3828 return r'%s(?P
<prefix
>|
[1-9][0-9]*|all
):(?P
<query
>[\s\S
]+)' % cls._SEARCH_KEY
3830 def _real_extract(self, query):
3831 prefix, query = self._match_valid_url(query).group('prefix
', 'query
')
3833 return self._get_n_results(query, 1)
3834 elif prefix == 'all
':
3835 return self._get_n_results(query, self._MAX_RESULTS)
3839 raise ExtractorError(f'invalid download number {n}
for query
"{query}"')
3840 elif n > self._MAX_RESULTS:
3841 self.report_warning('%s returns
max %i results (you requested
%i)' % (self._SEARCH_KEY, self._MAX_RESULTS, n))
3842 n = self._MAX_RESULTS
3843 return self._get_n_results(query, n)
3845 def _get_n_results(self, query, n):
3846 """Get a specified number of results for a query.
3847 Either this function or _search_results must be overridden by subclasses """
3848 return self.playlist_result(
3849 itertools.islice(self._search_results(query), 0, None if n == float('inf
') else n),
3852 def _search_results(self, query):
3853 """Returns an iterator of search results"""
3854 raise NotImplementedError('This method must be implemented by subclasses
')
3857 def SEARCH_KEY(cls):
3858 return cls._SEARCH_KEY