12 import xml
.etree
.ElementTree
14 from ..compat
import (
15 compat_cookiejar_Cookie
,
16 compat_cookies_SimpleCookie
,
17 compat_etree_fromstring
,
24 compat_urllib_parse_unquote
,
25 compat_urllib_parse_urlencode
,
26 compat_urllib_request
,
30 from ..downloader
import FileDownloader
31 from ..downloader
.f4m
import get_base_url
, remove_encrypted_media
64 parse_m3u8_attributes
,
89 """Information Extractor class.
91 Information extractors are the classes that, given a URL, extract
92 information about the video (or videos) the URL refers to. This
93 information includes the real video URL, the video title, author and
94 others. The information is stored in a dictionary which is then
95 passed to the YoutubeDL. The YoutubeDL processes this
96 information possibly downloading the video to the file system, among
97 other possible outcomes.
99 The type field determines the type of the result.
100 By far the most common value (and the default if _type is missing) is
101 "video", which indicates a single video.
103 For a video, the dictionaries must include the following fields:
105 id: Video identifier.
106 title: Video title, unescaped.
108 Additionally, it must contain either a formats entry or a url one:
110 formats: A list of dictionaries for each format available, ordered
111 from worst to best quality.
114 * url The mandatory URL representing the media:
115 for plain file media - HTTP URL of this file,
117 for HLS - URL of the M3U8 media playlist,
118 for HDS - URL of the F4M manifest,
120 - HTTP URL to plain file media (in case of
122 - URL of the MPD manifest or base URL
123 representing the media if MPD manifest
124 is parsed from a string (in case of
126 for MSS - URL of the ISM manifest.
128 The URL of the manifest file in case of
130 for HLS - URL of the M3U8 master playlist,
131 for HDS - URL of the F4M manifest,
132 for DASH - URL of the MPD manifest,
133 for MSS - URL of the ISM manifest.
134 * manifest_stream_number (For internal use only)
135 The index of the stream in the manifest file
136 * ext Will be calculated from URL if missing
137 * format A human-readable description of the format
138 ("mp4 container with h264/opus").
139 Calculated from the format_id, width, height.
140 and format_note fields if missing.
141 * format_id A short description of the format
142 ("mp4_h264_opus" or "19").
143 Technically optional, but strongly recommended.
144 * format_note Additional info about the format
145 ("3D" or "DASH video")
146 * width Width of the video, if known
147 * height Height of the video, if known
148 * resolution Textual description of width and height
149 * dynamic_range The dynamic range of the video. One of:
150 "SDR" (None), "HDR10", "HDR10+, "HDR12", "HLG, "DV"
151 * tbr Average bitrate of audio and video in KBit/s
152 * abr Average audio bitrate in KBit/s
153 * acodec Name of the audio codec in use
154 * asr Audio sampling rate in Hertz
155 * vbr Average video bitrate in KBit/s
157 * vcodec Name of the video codec in use
158 * container Name of the container format
159 * filesize The number of bytes, if known in advance
160 * filesize_approx An estimate for the number of bytes
161 * player_url SWF Player URL (used for rtmpdump).
162 * protocol The protocol that will be used for the actual
163 download, lower-case. One of "http", "https" or
164 one of the protocols defined in downloader.PROTOCOL_MAP
166 Base URL for fragments. Each fragment's path
167 value (if present) will be relative to
169 * fragments A list of fragments of a fragmented media.
170 Each fragment entry must contain either an url
171 or a path. If an url is present it should be
172 considered by a client. Otherwise both path and
173 fragment_base_url must be present. Here is
174 the list of all potential fields:
175 * "url" - fragment's URL
176 * "path" - fragment's path relative to
178 * "duration" (optional, int or float)
179 * "filesize" (optional, int)
180 * is_from_start Is a live format that can be downloaded
181 from the start. Boolean
182 * preference Order number of this format. If this field is
183 present and not None, the formats get sorted
184 by this field, regardless of all other values.
185 -1 for default (order by other properties),
186 -2 or smaller for less than default.
187 < -1000 to hide the format (if there is
188 another one which is strictly better)
189 * language Language code, e.g. "de" or "en-US".
190 * language_preference Is this in the language mentioned in
192 10 if it's what the URL is about,
193 -1 for default (don't know),
194 -10 otherwise, other values reserved for now.
195 * quality Order number of the video quality of this
196 format, irrespective of the file format.
197 -1 for default (order by other properties),
198 -2 or smaller for less than default.
199 * source_preference Order number for this video source
200 (quality takes higher priority)
201 -1 for default (order by other properties),
202 -2 or smaller for less than default.
203 * http_headers A dictionary of additional HTTP headers
204 to add to the request.
205 * stretched_ratio If given and not 1, indicates that the
206 video's pixels are not square.
207 width : height ratio as float.
208 * no_resume The server does not support resuming the
209 (HTTP or RTMP) download. Boolean.
210 * has_drm The format has DRM and cannot be downloaded. Boolean
211 * downloader_options A dictionary of downloader options
212 (For internal use only)
213 * http_chunk_size Chunk size for HTTP downloads
214 * ffmpeg_args Extra arguments for ffmpeg downloader
215 RTMP formats can also have the additional fields: page_url,
216 app, play_path, tc_url, flash_version, rtmp_live, rtmp_conn,
217 rtmp_protocol, rtmp_real_time
219 url: Final video URL.
220 ext: Video filename extension.
221 format: The video format, defaults to ext (used for --get-format)
222 player_url: SWF Player URL (used for rtmpdump).
224 The following fields are optional:
226 direct: True if a direct video file was given (must only be set by GenericIE)
227 alt_title: A secondary title of the video.
228 display_id An alternative identifier for the video, not necessarily
229 unique, but available before title. Typically, id is
230 something like "4234987", title "Dancing naked mole rats",
231 and display_id "dancing-naked-mole-rats"
232 thumbnails: A list of dictionaries, with the following entries:
233 * "id" (optional, string) - Thumbnail format ID
235 * "preference" (optional, int) - quality of the image
236 * "width" (optional, int)
237 * "height" (optional, int)
238 * "resolution" (optional, string "{width}x{height}",
240 * "filesize" (optional, int)
241 * "http_headers" (dict) - HTTP headers for the request
242 thumbnail: Full URL to a video thumbnail image.
243 description: Full video description.
244 uploader: Full name of the video uploader.
245 license: License name the video is licensed under.
246 creator: The creator of the video.
247 timestamp: UNIX timestamp of the moment the video was uploaded
248 upload_date: Video upload date in UTC (YYYYMMDD).
249 If not explicitly set, calculated from timestamp
250 release_timestamp: UNIX timestamp of the moment the video was released.
251 If it is not clear whether to use timestamp or this, use the former
252 release_date: The date (YYYYMMDD) when the video was released in UTC.
253 If not explicitly set, calculated from release_timestamp
254 modified_timestamp: UNIX timestamp of the moment the video was last modified.
255 modified_date: The date (YYYYMMDD) when the video was last modified in UTC.
256 If not explicitly set, calculated from modified_timestamp
257 uploader_id: Nickname or id of the video uploader.
258 uploader_url: Full URL to a personal webpage of the video uploader.
259 channel: Full name of the channel the video is uploaded on.
260 Note that channel fields may or may not repeat uploader
261 fields. This depends on a particular extractor.
262 channel_id: Id of the channel.
263 channel_url: Full URL to a channel webpage.
264 channel_follower_count: Number of followers of the channel.
265 location: Physical location where the video was filmed.
266 subtitles: The available subtitles as a dictionary in the format
267 {tag: subformats}. "tag" is usually a language code, and
268 "subformats" is a list sorted from lower to higher
269 preference, each element is a dictionary with the "ext"
271 * "data": The subtitles file contents
272 * "url": A URL pointing to the subtitles file
273 It can optionally also have:
274 * "name": Name or description of the subtitles
275 * "http_headers": A dictionary of additional HTTP headers
276 to add to the request.
277 "ext" will be calculated from URL if missing
278 automatic_captions: Like 'subtitles'; contains automatically generated
279 captions instead of normal subtitles
280 duration: Length of the video in seconds, as an integer or float.
281 view_count: How many users have watched the video on the platform.
282 like_count: Number of positive ratings of the video
283 dislike_count: Number of negative ratings of the video
284 repost_count: Number of reposts of the video
285 average_rating: Average rating give by users, the scale used depends on the webpage
286 comment_count: Number of comments on the video
287 comments: A list of comments, each with one or more of the following
288 properties (all but one of text or html optional):
289 * "author" - human-readable name of the comment author
290 * "author_id" - user ID of the comment author
291 * "author_thumbnail" - The thumbnail of the comment author
293 * "html" - Comment as HTML
294 * "text" - Plain text of the comment
295 * "timestamp" - UNIX timestamp of comment
296 * "parent" - ID of the comment this one is replying to.
297 Set to "root" to indicate that this is a
298 comment to the original video.
299 * "like_count" - Number of positive ratings of the comment
300 * "dislike_count" - Number of negative ratings of the comment
301 * "is_favorited" - Whether the comment is marked as
302 favorite by the video uploader
303 * "author_is_uploader" - Whether the comment is made by
305 age_limit: Age restriction for the video, as an integer (years)
306 webpage_url: The URL to the video webpage, if given to yt-dlp it
307 should allow to get the same result again. (It will be set
308 by YoutubeDL if it's missing)
309 categories: A list of categories that the video falls in, for example
311 tags: A list of tags assigned to the video, e.g. ["sweden", "pop music"]
312 cast: A list of the video cast
313 is_live: True, False, or None (=unknown). Whether this video is a
314 live stream that goes on instead of a fixed-length video.
315 was_live: True, False, or None (=unknown). Whether this video was
316 originally a live stream.
317 live_status: 'is_live', 'is_upcoming', 'was_live', 'not_live' or None (=unknown)
318 If absent, automatically set from is_live, was_live
319 start_time: Time in seconds where the reproduction should start, as
320 specified in the URL.
321 end_time: Time in seconds where the reproduction should end, as
322 specified in the URL.
323 chapters: A list of dictionaries, with the following entries:
324 * "start_time" - The start time of the chapter in seconds
325 * "end_time" - The end time of the chapter in seconds
326 * "title" (optional, string)
327 playable_in_embed: Whether this video is allowed to play in embedded
328 players on other sites. Can be True (=always allowed),
329 False (=never allowed), None (=unknown), or a string
330 specifying the criteria for embedability (Eg: 'whitelist')
331 availability: Under what condition the video is available. One of
332 'private', 'premium_only', 'subscriber_only', 'needs_auth',
333 'unlisted' or 'public'. Use 'InfoExtractor._availability'
335 __post_extractor: A function to be called just before the metadata is
336 written to either disk, logger or console. The function
337 must return a dict which will be added to the info_dict.
338 This is usefull for additional information that is
339 time-consuming to extract. Note that the fields thus
340 extracted will not be available to output template and
341 match_filter. So, only "comments" and "comment_count" are
342 currently allowed to be extracted via this method.
344 The following fields should only be used when the video belongs to some logical
347 chapter: Name or title of the chapter the video belongs to.
348 chapter_number: Number of the chapter the video belongs to, as an integer.
349 chapter_id: Id of the chapter the video belongs to, as a unicode string.
351 The following fields should only be used when the video is an episode of some
352 series, programme or podcast:
354 series: Title of the series or programme the video episode belongs to.
355 series_id: Id of the series or programme the video episode belongs to, as a unicode string.
356 season: Title of the season the video episode belongs to.
357 season_number: Number of the season the video episode belongs to, as an integer.
358 season_id: Id of the season the video episode belongs to, as a unicode string.
359 episode: Title of the video episode. Unlike mandatory video title field,
360 this field should denote the exact title of the video episode
361 without any kind of decoration.
362 episode_number: Number of the video episode within a season, as an integer.
363 episode_id: Id of the video episode, as a unicode string.
365 The following fields should only be used when the media is a track or a part of
368 track: Title of the track.
369 track_number: Number of the track within an album or a disc, as an integer.
370 track_id: Id of the track (useful in case of custom indexing, e.g. 6.iii),
372 artist: Artist(s) of the track.
373 genre: Genre(s) of the track.
374 album: Title of the album the track belongs to.
375 album_type: Type of the album (e.g. "Demo", "Full-length", "Split", "Compilation", etc).
376 album_artist: List of all artists appeared on the album (e.g.
377 "Ash Borer / Fell Voices" or "Various Artists", useful for splits
379 disc_number: Number of the disc or other physical medium the track belongs to,
381 release_year: Year (YYYY) when the album was released.
382 composer: Composer of the piece
384 Unless mentioned otherwise, the fields should be Unicode strings.
386 Unless mentioned otherwise, None is equivalent to absence of information.
389 _type "playlist" indicates multiple videos.
390 There must be a key "entries", which is a list, an iterable, or a PagedList
391 object, each element of which is a valid dictionary by this specification.
393 Additionally, playlists can have "id", "title", and any other relevent
394 attributes with the same semantics as videos (see above).
396 It can also have the following optional fields:
398 playlist_count: The total number of videos in a playlist. If not given,
399 YoutubeDL tries to calculate it from "entries"
402 _type "multi_video" indicates that there are multiple videos that
403 form a single show, for examples multiple acts of an opera or TV episode.
404 It must have an entries key like a playlist and contain all the keys
405 required for a video at the same time.
408 _type "url" indicates that the video must be extracted from another
409 location, possibly by a different extractor. Its only required key is:
410 "url" - the next URL to extract.
411 The key "ie_key" can be set to the class name (minus the trailing "IE",
412 e.g. "Youtube") if the extractor class is known in advance.
413 Additionally, the dictionary may have any properties of the resolved entity
414 known in advance, for example "title" if the title of the referred video is
418 _type "url_transparent" entities have the same specification as "url", but
419 indicate that the given additional information is more precise than the one
420 associated with the resolved URL.
421 This is useful when a site employs a video service that hosts the video and
422 its technical metadata, but that video service does not embed a useful
423 title, description etc.
426 Subclasses of this should define a _VALID_URL regexp and, re-define the
427 _real_extract() and (optionally) _real_initialize() methods.
428 Probably, they should also be added to the list of extractors.
430 Subclasses may also override suitable() if necessary, but ensure the function
431 signature is preserved and that this function imports everything it needs
432 (except other extractors), so that lazy_extractors works correctly.
434 To support username + password (or netrc) login, the extractor must define a
435 _NETRC_MACHINE and re-define _perform_login(username, password) and
436 (optionally) _initialize_pre_login() methods. The _perform_login method will
437 be called between _initialize_pre_login and _real_initialize if credentials
438 are passed by the user. In cases where it is necessary to have the login
439 process as part of the extraction rather than initialization, _perform_login
440 can be left undefined.
442 _GEO_BYPASS attribute may be set to False in order to disable
443 geo restriction bypass mechanisms for a particular extractor.
444 Though it won't disable explicit geo restriction bypass based on
445 country code provided with geo_bypass_country.
447 _GEO_COUNTRIES attribute may contain a list of presumably geo unrestricted
448 countries for this extractor. One of these countries will be used by
449 geo restriction bypass mechanism right away in order to bypass
450 geo restriction, of course, if the mechanism is not disabled.
452 _GEO_IP_BLOCKS attribute may contain a list of presumably geo unrestricted
453 IP blocks in CIDR notation for this extractor. One of these IP blocks
454 will be used by geo restriction bypass mechanism similarly
457 The _WORKING attribute should be set to False for broken IEs
458 in order to warn the users and skip the tests.
463 _x_forwarded_for_ip
= None
465 _GEO_COUNTRIES
= None
466 _GEO_IP_BLOCKS
= None
468 _NETRC_MACHINE
= None
472 'any': 'Use --cookies, --cookies-from-browser, --username and --password, or --netrc to provide account credentials',
474 'Use --cookies-from-browser or --cookies for the authentication. '
475 'See https://github.com/ytdl-org/youtube-dl#how-do-i-pass-cookies-to-youtube-dl for how to manually pass cookies'),
476 'password': 'Use --username and --password, or --netrc to provide account credentials',
479 def __init__(self
, downloader
=None):
480 """Constructor. Receives an optional downloader (a YoutubeDL instance).
481 If a downloader is not passed during initialization,
482 it must be set using "set_downloader()" before "extract()" is called"""
484 self
._x
_forwarded
_for
_ip
= None
485 self
._printed
_messages
= set()
486 self
.set_downloader(downloader
)
489 def _match_valid_url(cls
, url
):
490 # This does not use has/getattr intentionally - we want to know whether
491 # we have cached the regexp for *this* class, whereas getattr would also
492 # match the superclass
493 if '_VALID_URL_RE' not in cls
.__dict
__:
494 if '_VALID_URL' not in cls
.__dict
__:
495 cls
._VALID
_URL
= cls
._make
_valid
_url
()
496 cls
._VALID
_URL
_RE
= re
.compile(cls
._VALID
_URL
)
497 return cls
._VALID
_URL
_RE
.match(url
)
500 def suitable(cls
, url
):
501 """Receives a URL and returns True if suitable for this IE."""
502 # This function must import everything it needs (except other extractors),
503 # so that lazy_extractors works correctly
504 return cls
._match
_valid
_url
(url
) is not None
507 def _match_id(cls
, url
):
508 return cls
._match
_valid
_url
(url
).group('id')
511 def get_temp_id(cls
, url
):
513 return cls
._match
_id
(url
)
514 except (IndexError, AttributeError):
519 """Getter method for _WORKING."""
523 def supports_login(cls
):
524 return bool(cls
._NETRC
_MACHINE
)
526 def initialize(self
):
527 """Initializes an instance (authentication, etc)."""
528 self
._printed
_messages
= set()
529 self
._initialize
_geo
_bypass
({
530 'countries': self
._GEO
_COUNTRIES
,
531 'ip_blocks': self
._GEO
_IP
_BLOCKS
,
534 self
._initialize
_pre
_login
()
535 if self
.supports_login():
536 username
, password
= self
._get
_login
_info
()
538 self
._perform
_login
(username
, password
)
539 elif self
.get_param('username') and False not in (self
.IE_DESC
, self
._NETRC
_MACHINE
):
540 self
.report_warning(f
'Login with password is not supported for this website. {self._LOGIN_HINTS["cookies"]}')
541 self
._real
_initialize
()
544 def _initialize_geo_bypass(self
, geo_bypass_context
):
546 Initialize geo restriction bypass mechanism.
548 This method is used to initialize geo bypass mechanism based on faking
549 X-Forwarded-For HTTP header. A random country from provided country list
550 is selected and a random IP belonging to this country is generated. This
551 IP will be passed as X-Forwarded-For HTTP header in all subsequent
554 This method will be used for initial geo bypass mechanism initialization
555 during the instance initialization with _GEO_COUNTRIES and
558 You may also manually call it from extractor's code if geo bypass
559 information is not available beforehand (e.g. obtained during
560 extraction) or due to some other reason. In this case you should pass
561 this information in geo bypass context passed as first argument. It may
562 contain following fields:
564 countries: List of geo unrestricted countries (similar
566 ip_blocks: List of geo unrestricted IP blocks in CIDR notation
567 (similar to _GEO_IP_BLOCKS)
570 if not self
._x
_forwarded
_for
_ip
:
572 # Geo bypass mechanism is explicitly disabled by user
573 if not self
.get_param('geo_bypass', True):
576 if not geo_bypass_context
:
577 geo_bypass_context
= {}
579 # Backward compatibility: previously _initialize_geo_bypass
580 # expected a list of countries, some 3rd party code may still use
582 if isinstance(geo_bypass_context
, (list, tuple)):
583 geo_bypass_context
= {
584 'countries': geo_bypass_context
,
587 # The whole point of geo bypass mechanism is to fake IP
588 # as X-Forwarded-For HTTP header based on some IP block or
591 # Path 1: bypassing based on IP block in CIDR notation
593 # Explicit IP block specified by user, use it right away
594 # regardless of whether extractor is geo bypassable or not
595 ip_block
= self
.get_param('geo_bypass_ip_block', None)
597 # Otherwise use random IP block from geo bypass context but only
598 # if extractor is known as geo bypassable
600 ip_blocks
= geo_bypass_context
.get('ip_blocks')
601 if self
._GEO
_BYPASS
and ip_blocks
:
602 ip_block
= random
.choice(ip_blocks
)
605 self
._x
_forwarded
_for
_ip
= GeoUtils
.random_ipv4(ip_block
)
606 self
._downloader
.write_debug(
607 '[debug] Using fake IP %s as X-Forwarded-For' % self
._x
_forwarded
_for
_ip
)
610 # Path 2: bypassing based on country code
612 # Explicit country code specified by user, use it right away
613 # regardless of whether extractor is geo bypassable or not
614 country
= self
.get_param('geo_bypass_country', None)
616 # Otherwise use random country code from geo bypass context but
617 # only if extractor is known as geo bypassable
619 countries
= geo_bypass_context
.get('countries')
620 if self
._GEO
_BYPASS
and countries
:
621 country
= random
.choice(countries
)
624 self
._x
_forwarded
_for
_ip
= GeoUtils
.random_ipv4(country
)
625 self
._downloader
.write_debug(
626 f
'Using fake IP {self._x_forwarded_for_ip} ({country.upper()}) as X-Forwarded-For')
628 def extract(self
, url
):
629 """Extracts URL information and returns it in list of dicts."""
634 self
.write_debug('Extracting URL: %s' % url
)
635 ie_result
= self
._real
_extract
(url
)
636 if ie_result
is None:
638 if self
._x
_forwarded
_for
_ip
:
639 ie_result
['__x_forwarded_for_ip'] = self
._x
_forwarded
_for
_ip
640 subtitles
= ie_result
.get('subtitles')
641 if (subtitles
and 'live_chat' in subtitles
642 and 'no-live-chat' in self
.get_param('compat_opts', [])):
643 del subtitles
['live_chat']
645 except GeoRestrictedError
as e
:
646 if self
.__maybe
_fake
_ip
_and
_retry
(e
.countries
):
649 except UnsupportedError
:
651 except ExtractorError
as e
:
653 'video_id': e
.video_id
or self
.get_temp_id(url
),
655 'tb': e
.traceback
or sys
.exc_info()[2],
656 'expected': e
.expected
,
659 if hasattr(e
, 'countries'):
660 kwargs
['countries'] = e
.countries
661 raise type(e
)(e
.orig_msg
, **kwargs
)
662 except compat_http_client
.IncompleteRead
as e
:
663 raise ExtractorError('A network error has occurred.', cause
=e
, expected
=True, video_id
=self
.get_temp_id(url
))
664 except (KeyError, StopIteration) as e
:
665 raise ExtractorError('An extractor error has occurred.', cause
=e
, video_id
=self
.get_temp_id(url
))
667 def __maybe_fake_ip_and_retry(self
, countries
):
668 if (not self
.get_param('geo_bypass_country', None)
670 and self
.get_param('geo_bypass', True)
671 and not self
._x
_forwarded
_for
_ip
673 country_code
= random
.choice(countries
)
674 self
._x
_forwarded
_for
_ip
= GeoUtils
.random_ipv4(country_code
)
675 if self
._x
_forwarded
_for
_ip
:
677 'Video is geo restricted. Retrying extraction with fake IP %s (%s) as X-Forwarded-For.'
678 % (self
._x
_forwarded
_for
_ip
, country_code
.upper()))
682 def set_downloader(self
, downloader
):
683 """Sets a YoutubeDL instance as the downloader for this IE."""
684 self
._downloader
= downloader
686 def _initialize_pre_login(self
):
687 """ Intialization before login. Redefine in subclasses."""
690 def _perform_login(self
, username
, password
):
691 """ Login with username and password. Redefine in subclasses."""
694 def _real_initialize(self
):
695 """Real initialization process. Redefine in subclasses."""
698 def _real_extract(self
, url
):
699 """Real extraction process. Redefine in subclasses."""
700 raise NotImplementedError('This method must be implemented by subclasses')
704 """A string for getting the InfoExtractor with get_info_extractor"""
705 return cls
.__name
__[:-2]
709 return compat_str(type(self
).__name
__[:-2])
712 def __can_accept_status_code(err
, expected_status
):
713 assert isinstance(err
, compat_urllib_error
.HTTPError
)
714 if expected_status
is None:
716 elif callable(expected_status
):
717 return expected_status(err
.code
) is True
719 return err
.code
in variadic(expected_status
)
721 def _request_webpage(self
, url_or_request
, video_id
, note
=None, errnote
=None, fatal
=True, data
=None, headers
={}, query={}
, expected_status
=None):
723 Return the response handle.
725 See _download_webpage docstring for arguments specification.
727 if not self
._downloader
._first
_webpage
_request
:
728 sleep_interval
= self
.get_param('sleep_interval_requests') or 0
729 if sleep_interval
> 0:
730 self
.to_screen('Sleeping %s seconds ...' % sleep_interval
)
731 time
.sleep(sleep_interval
)
733 self
._downloader
._first
_webpage
_request
= False
736 self
.report_download_webpage(video_id
)
737 elif note
is not False:
739 self
.to_screen(str(note
))
741 self
.to_screen(f
'{video_id}: {note}')
743 # Some sites check X-Forwarded-For HTTP header in order to figure out
744 # the origin of the client behind proxy. This allows bypassing geo
745 # restriction by faking this header's value to IP that belongs to some
746 # geo unrestricted country. We will do so once we encounter any
747 # geo restriction error.
748 if self
._x
_forwarded
_for
_ip
:
749 if 'X-Forwarded-For' not in headers
:
750 headers
['X-Forwarded-For'] = self
._x
_forwarded
_for
_ip
752 if isinstance(url_or_request
, compat_urllib_request
.Request
):
753 url_or_request
= update_Request(
754 url_or_request
, data
=data
, headers
=headers
, query
=query
)
757 url_or_request
= update_url_query(url_or_request
, query
)
758 if data
is not None or headers
:
759 url_or_request
= sanitized_Request(url_or_request
, data
, headers
)
761 return self
._downloader
.urlopen(url_or_request
)
762 except network_exceptions
as err
:
763 if isinstance(err
, compat_urllib_error
.HTTPError
):
764 if self
.__can
_accept
_status
_code
(err
, expected_status
):
765 # Retain reference to error to prevent file object from
766 # being closed before it can be read. Works around the
767 # effects of <https://bugs.python.org/issue15002>
768 # introduced in Python 3.4.1.
775 errnote
= 'Unable to download webpage'
777 errmsg
= f
'{errnote}: {error_to_compat_str(err)}'
779 raise ExtractorError(errmsg
, cause
=err
)
781 self
.report_warning(errmsg
)
784 def _download_webpage_handle(self
, url_or_request
, video_id
, note
=None, errnote
=None, fatal
=True, encoding
=None, data
=None, headers
={}, query={}
, expected_status
=None):
786 Return a tuple (page content as string, URL handle).
788 See _download_webpage docstring for arguments specification.
790 # Strip hashes from the URL (#1038)
791 if isinstance(url_or_request
, (compat_str
, str)):
792 url_or_request
= url_or_request
.partition('#')[0]
794 urlh
= self
._request
_webpage
(url_or_request
, video_id
, note
, errnote
, fatal
, data
=data
, headers
=headers
, query
=query
, expected_status
=expected_status
)
798 content
= self
._webpage
_read
_content
(urlh
, url_or_request
, video_id
, note
, errnote
, fatal
, encoding
=encoding
)
799 return (content
, urlh
)
802 def _guess_encoding_from_content(content_type
, webpage_bytes
):
803 m
= re
.match(r
'[a-zA-Z0-9_.-]+/[a-zA-Z0-9_.-]+\s*;\s*charset=(.+)', content_type
)
805 encoding
= m
.group(1)
807 m
= re
.search(br
'<meta[^>]+charset=[\'"]?([^\'")]+)[ /\'">]',
808 webpage_bytes[:1024])
810 encoding = m.group(1).decode('ascii')
811 elif webpage_bytes.startswith(b'\xff\xfe'):
818 def __check_blocked(self, content):
819 first_block = content[:512]
820 if ('<title>Access to this site is blocked</title>' in content
821 and 'Websense' in first_block):
822 msg = 'Access to this webpage has been blocked by Websense filtering software in your network.'
823 blocked_iframe = self._html_search_regex(
824 r'<iframe src="([^
"]+)"', content,
825 'Websense information URL
', default=None)
827 msg += ' Visit
%s for more details
' % blocked_iframe
828 raise ExtractorError(msg, expected=True)
829 if '<title
>The URL you requested has been blocked
</title
>' in first_block:
831 'Access to this webpage has been blocked by Indian censorship
. '
832 'Use a VPN
or proxy
server (with --proxy
) to route around it
.')
833 block_msg = self._html_search_regex(
834 r'</h1
><p
>(.*?
)</p
>',
835 content, 'block message
', default=None)
837 msg += ' (Message
: "%s")' % block_msg.replace('\n', ' ')
838 raise ExtractorError(msg, expected=True)
839 if ('<title
>TTK
:: Доступ к ресурсу ограничен
</title
>' in content
840 and 'blocklist
.rkn
.gov
.ru
' in content):
841 raise ExtractorError(
842 'Access to this webpage has been blocked by decision of the Russian government
. '
843 'Visit http
://blocklist
.rkn
.gov
.ru
/ for a block reason
.',
846 def _webpage_read_content(self, urlh, url_or_request, video_id, note=None, errnote=None, fatal=True, prefix=None, encoding=None):
847 content_type = urlh.headers.get('Content
-Type
', '')
848 webpage_bytes = urlh.read()
849 if prefix is not None:
850 webpage_bytes = prefix + webpage_bytes
852 encoding = self._guess_encoding_from_content(content_type, webpage_bytes)
853 if self.get_param('dump_intermediate_pages
', False):
854 self.to_screen('Dumping request to
' + urlh.geturl())
855 dump = base64.b64encode(webpage_bytes).decode('ascii
')
856 self._downloader.to_screen(dump)
857 if self.get_param('write_pages
', False):
858 basen = f'{video_id}_{urlh.geturl()}
'
859 trim_length = self.get_param('trim_file_name
') or 240
860 if len(basen) > trim_length:
861 h = '___
' + hashlib.md5(basen.encode('utf
-8')).hexdigest()
862 basen = basen[:trim_length - len(h)] + h
863 raw_filename = basen + '.dump
'
864 filename = sanitize_filename(raw_filename, restricted=True)
865 self.to_screen('Saving request to
' + filename)
866 # Working around MAX_PATH limitation on Windows (see
867 # http://msdn.microsoft.com/en-us/library/windows/desktop/aa365247(v=vs.85).aspx)
868 if compat_os_name == 'nt
':
869 absfilepath = os.path.abspath(filename)
870 if len(absfilepath) > 259:
871 filename = '\\\\?
\\' + absfilepath
872 with open(filename, 'wb
') as outf:
873 outf.write(webpage_bytes)
876 content = webpage_bytes.decode(encoding, 'replace
')
878 content = webpage_bytes.decode('utf
-8', 'replace
')
880 self.__check_blocked(content)
884 def _download_webpage(
885 self, url_or_request, video_id, note=None, errnote=None,
886 fatal=True, tries=1, timeout=5, encoding=None, data=None,
887 headers={}, query={}, expected_status=None):
889 Return the data of the page as a string.
892 url_or_request -- plain text URL as a string or
893 a compat_urllib_request.Requestobject
894 video_id -- Video/playlist/item identifier (string)
897 note -- note printed before downloading (string)
898 errnote -- note printed in case of an error (string)
899 fatal -- flag denoting whether error should be considered fatal,
900 i.e. whether it should cause ExtractionError to be raised,
901 otherwise a warning will be reported and extraction continued
902 tries -- number of tries
903 timeout -- sleep interval between tries
904 encoding -- encoding for a page content decoding, guessed automatically
905 when not explicitly specified
906 data -- POST data (bytes)
907 headers -- HTTP headers (dict)
908 query -- URL query (dict)
909 expected_status -- allows to accept failed HTTP requests (non 2xx
910 status code) by explicitly specifying a set of accepted status
911 codes. Can be any of the following entities:
912 - an integer type specifying an exact failed status code to
914 - a list or a tuple of integer types specifying a list of
915 failed status codes to accept
916 - a callable accepting an actual failed status code and
917 returning True if it should be accepted
918 Note that this argument does not affect success status codes (2xx)
919 which are always accepted.
924 while success is False:
926 res = self._download_webpage_handle(
927 url_or_request, video_id, note, errnote, fatal,
928 encoding=encoding, data=data, headers=headers, query=query,
929 expected_status=expected_status)
931 except compat_http_client.IncompleteRead as e:
933 if try_count >= tries:
935 self._sleep(timeout, video_id)
942 def _download_xml_handle(
943 self, url_or_request, video_id, note='Downloading XML
',
944 errnote='Unable to download XML
', transform_source=None,
945 fatal=True, encoding=None, data=None, headers={}, query={},
946 expected_status=None):
948 Return a tuple (xml as an xml.etree.ElementTree.Element, URL handle).
950 See _download_webpage docstring for arguments specification.
952 res = self._download_webpage_handle(
953 url_or_request, video_id, note, errnote, fatal=fatal,
954 encoding=encoding, data=data, headers=headers, query=query,
955 expected_status=expected_status)
958 xml_string, urlh = res
959 return self._parse_xml(
960 xml_string, video_id, transform_source=transform_source,
964 self, url_or_request, video_id,
965 note='Downloading XML
', errnote='Unable to download XML
',
966 transform_source=None, fatal=True, encoding=None,
967 data=None, headers={}, query={}, expected_status=None):
969 Return the xml as an xml.etree.ElementTree.Element.
971 See _download_webpage docstring for arguments specification.
973 res = self._download_xml_handle(
974 url_or_request, video_id, note=note, errnote=errnote,
975 transform_source=transform_source, fatal=fatal, encoding=encoding,
976 data=data, headers=headers, query=query,
977 expected_status=expected_status)
978 return res if res is False else res[0]
980 def _parse_xml(self, xml_string, video_id, transform_source=None, fatal=True):
982 xml_string = transform_source(xml_string)
984 return compat_etree_fromstring(xml_string.encode('utf
-8'))
985 except xml.etree.ElementTree.ParseError as ve:
986 errmsg = '%s: Failed to parse XML
' % video_id
988 raise ExtractorError(errmsg, cause=ve)
990 self.report_warning(errmsg + str(ve))
992 def _download_json_handle(
993 self, url_or_request, video_id, note='Downloading JSON metadata
',
994 errnote='Unable to download JSON metadata
', transform_source=None,
995 fatal=True, encoding=None, data=None, headers={}, query={},
996 expected_status=None):
998 Return a tuple (JSON object, URL handle).
1000 See _download_webpage docstring for arguments specification.
1002 res = self._download_webpage_handle(
1003 url_or_request, video_id, note, errnote, fatal=fatal,
1004 encoding=encoding, data=data, headers=headers, query=query,
1005 expected_status=expected_status)
1008 json_string, urlh = res
1009 return self._parse_json(
1010 json_string, video_id, transform_source=transform_source,
1014 self, url_or_request, video_id, note='Downloading JSON metadata
',
1015 errnote='Unable to download JSON metadata
', transform_source=None,
1016 fatal=True, encoding=None, data=None, headers={}, query={},
1017 expected_status=None):
1019 Return the JSON object as a dict.
1021 See _download_webpage docstring for arguments specification.
1023 res = self._download_json_handle(
1024 url_or_request, video_id, note=note, errnote=errnote,
1025 transform_source=transform_source, fatal=fatal, encoding=encoding,
1026 data=data, headers=headers, query=query,
1027 expected_status=expected_status)
1028 return res if res is False else res[0]
1030 def _parse_json(self, json_string, video_id, transform_source=None, fatal=True):
1031 if transform_source:
1032 json_string = transform_source(json_string)
1034 return json.loads(json_string, strict=False)
1035 except ValueError as ve:
1036 errmsg = '%s: Failed to parse JSON
' % video_id
1038 raise ExtractorError(errmsg, cause=ve)
1040 self.report_warning(errmsg + str(ve))
1042 def _parse_socket_response_as_json(self, data, video_id, transform_source=None, fatal=True):
1043 return self._parse_json(
1044 data[data.find('{'):data.rfind('}
') + 1],
1045 video_id, transform_source, fatal)
1047 def _download_socket_json_handle(
1048 self, url_or_request, video_id, note='Polling socket
',
1049 errnote='Unable to poll socket
', transform_source=None,
1050 fatal=True, encoding=None, data=None, headers={}, query={},
1051 expected_status=None):
1053 Return a tuple (JSON object, URL handle).
1055 See _download_webpage docstring for arguments specification.
1057 res = self._download_webpage_handle(
1058 url_or_request, video_id, note, errnote, fatal=fatal,
1059 encoding=encoding, data=data, headers=headers, query=query,
1060 expected_status=expected_status)
1064 return self._parse_socket_response_as_json(
1065 webpage, video_id, transform_source=transform_source,
1068 def _download_socket_json(
1069 self, url_or_request, video_id, note='Polling socket
',
1070 errnote='Unable to poll socket
', transform_source=None,
1071 fatal=True, encoding=None, data=None, headers={}, query={},
1072 expected_status=None):
1074 Return the JSON object as a dict.
1076 See _download_webpage docstring for arguments specification.
1078 res = self._download_socket_json_handle(
1079 url_or_request, video_id, note=note, errnote=errnote,
1080 transform_source=transform_source, fatal=fatal, encoding=encoding,
1081 data=data, headers=headers, query=query,
1082 expected_status=expected_status)
1083 return res if res is False else res[0]
1085 def report_warning(self, msg, video_id=None, *args, only_once=False, **kwargs):
1086 idstr = format_field(video_id, template='%s: ')
1087 msg = f'[{self.IE_NAME}
] {idstr}{msg}
'
1089 if f'WARNING
: {msg}
' in self._printed_messages:
1091 self._printed_messages.add(f'WARNING
: {msg}
')
1092 self._downloader.report_warning(msg, *args, **kwargs)
1094 def to_screen(self, msg, *args, **kwargs):
1095 """Print msg to screen, prefixing it with '[ie_name
]'"""
1096 self._downloader.to_screen(f'[{self.IE_NAME}
] {msg}
', *args, **kwargs)
1098 def write_debug(self, msg, *args, **kwargs):
1099 self._downloader.write_debug(f'[{self.IE_NAME}
] {msg}
', *args, **kwargs)
1101 def get_param(self, name, default=None, *args, **kwargs):
1102 if self._downloader:
1103 return self._downloader.params.get(name, default, *args, **kwargs)
1106 def report_drm(self, video_id, partial=False):
1107 self.raise_no_formats('This video
is DRM protected
', expected=True, video_id=video_id)
1109 def report_extraction(self, id_or_name):
1110 """Report information extraction."""
1111 self.to_screen('%s: Extracting information
' % id_or_name)
1113 def report_download_webpage(self, video_id):
1114 """Report webpage download."""
1115 self.to_screen('%s: Downloading webpage
' % video_id)
1117 def report_age_confirmation(self):
1118 """Report attempt to confirm age."""
1119 self.to_screen('Confirming age
')
1121 def report_login(self):
1122 """Report attempt to log in."""
1123 self.to_screen('Logging
in')
1125 def raise_login_required(
1126 self, msg='This video
is only available
for registered users
',
1127 metadata_available=False, method=NO_DEFAULT):
1128 if metadata_available and (
1129 self.get_param('ignore_no_formats_error
') or self.get_param('wait_for_video
')):
1130 self.report_warning(msg)
1132 if method is NO_DEFAULT:
1133 method = 'any
' if self.supports_login() else 'cookies
'
1134 if method is not None:
1135 assert method in self._LOGIN_HINTS, 'Invalid login method
'
1136 msg = f'{msg}
. {self._LOGIN_HINTS[method]}
'
1137 raise ExtractorError(msg, expected=True)
1139 def raise_geo_restricted(
1140 self, msg='This video
is not available
from your location due to geo restriction
',
1141 countries=None, metadata_available=False):
1142 if metadata_available and (
1143 self.get_param('ignore_no_formats_error
') or self.get_param('wait_for_video
')):
1144 self.report_warning(msg)
1146 raise GeoRestrictedError(msg, countries=countries)
1148 def raise_no_formats(self, msg, expected=False, video_id=None):
1150 self.get_param('ignore_no_formats_error
') or self.get_param('wait_for_video
')):
1151 self.report_warning(msg, video_id)
1152 elif isinstance(msg, ExtractorError):
1155 raise ExtractorError(msg, expected=expected, video_id=video_id)
1157 # Methods for following #608
1159 def url_result(url, ie=None, video_id=None, video_title=None, *, url_transparent=False, **kwargs):
1160 """Returns a URL that points to a page that should be processed"""
1162 kwargs['ie_key
'] = ie if isinstance(ie, str) else ie.ie_key()
1163 if video_id is not None:
1164 kwargs['id'] = video_id
1165 if video_title is not None:
1166 kwargs['title
'] = video_title
1169 '_type
': 'url_transparent
' if url_transparent else 'url
',
1173 def playlist_from_matches(self, matches, playlist_id=None, playlist_title=None, getter=None, ie=None, video_kwargs=None, **kwargs):
1174 urls = (self.url_result(self._proto_relative_url(m), ie, **(video_kwargs or {}))
1175 for m in orderedSet(map(getter, matches) if getter else matches))
1176 return self.playlist_result(urls, playlist_id, playlist_title, **kwargs)
1179 def playlist_result(entries, playlist_id=None, playlist_title=None, playlist_description=None, *, multi_video=False, **kwargs):
1180 """Returns a playlist"""
1182 kwargs['id'] = playlist_id
1184 kwargs['title
'] = playlist_title
1185 if playlist_description is not None:
1186 kwargs['description
'] = playlist_description
1189 '_type
': 'multi_video
' if multi_video else 'playlist
',
1193 def _search_regex(self, pattern, string, name, default=NO_DEFAULT, fatal=True, flags=0, group=None):
1195 Perform a regex search on the given string, using a single or a list of
1196 patterns returning the first matching group.
1197 In case of failure return a default value or raise a WARNING or a
1198 RegexNotFoundError, depending on fatal, specifying the field name.
1202 elif isinstance(pattern, (str, re.Pattern)):
1203 mobj = re.search(pattern, string, flags)
1206 mobj = re.search(p, string, flags)
1210 _name = self._downloader._format_err(name, self._downloader.Styles.EMPHASIS)
1214 # return the first matching group
1215 return next(g for g in mobj.groups() if g is not None)
1216 elif isinstance(group, (list, tuple)):
1217 return tuple(mobj.group(g) for g in group)
1219 return mobj.group(group)
1220 elif default is not NO_DEFAULT:
1223 raise RegexNotFoundError('Unable to extract
%s' % _name)
1225 self.report_warning('unable to extract
%s' % _name + bug_reports_message())
1228 def _html_search_regex(self, pattern, string, name, default=NO_DEFAULT, fatal=True, flags=0, group=None):
1230 Like _search_regex, but strips HTML tags and unescapes entities.
1232 res = self._search_regex(pattern, string, name, default, fatal, flags, group)
1234 return clean_html(res).strip()
1238 def _get_netrc_login_info(self, netrc_machine=None):
1241 netrc_machine = netrc_machine or self._NETRC_MACHINE
1243 if self.get_param('usenetrc
', False):
1245 netrc_file = compat_expanduser(self.get_param('netrc_location
') or '~
')
1246 if os.path.isdir(netrc_file):
1247 netrc_file = os.path.join(netrc_file, '.netrc
')
1248 info = netrc.netrc(file=netrc_file).authenticators(netrc_machine)
1249 if info is not None:
1253 raise netrc.NetrcParseError(
1254 'No authenticators
for %s' % netrc_machine)
1255 except (OSError, netrc.NetrcParseError) as err:
1256 self.report_warning(
1257 'parsing
.netrc
: %s' % error_to_compat_str(err))
1259 return username, password
1261 def _get_login_info(self, username_option='username
', password_option='password
', netrc_machine=None):
1263 Get the login info as (username, password)
1264 First look for the manually specified credentials using username_option
1265 and password_option as keys in params dictionary. If no such credentials
1266 available look in the netrc file using the netrc_machine or _NETRC_MACHINE
1268 If there's no info available
, return (None, None)
1271 # Attempt to use provided username and password or .netrc data
1272 username = self.get_param(username_option)
1273 if username is not None:
1274 password = self.get_param(password_option)
1276 username, password = self._get_netrc_login_info(netrc_machine)
1278 return username, password
1280 def _get_tfa_info(self, note='two-factor verification code'):
1282 Get the two
-factor authentication info
1283 TODO
- asking the user will be required
for sms
/phone verify
1284 currently just uses the command line option
1285 If there
's no info available, return None
1288 tfa = self.get_param('twofactor
')
1292 return compat_getpass('Type
%s and press
[Return
]: ' % note)
1294 # Helper functions for extracting OpenGraph info
1296 def _og_regexes(prop):
1297 content_re = r'content
=(?
:"([^"]+?
)"|\'([^\']+?)\'|\s*([^\s"\'=<>`
]+?
))'
1298 property_re = (r'(?
:name|
property)=(?
:\'og
%(sep)s%(prop)s\'|
"og%(sep)s%(prop)s"|\s
*og
%(sep)s%(prop)s\b)'
1299 % {'prop': re.escape(prop), 'sep': '(?::|[:-])'})
1300 template = r'<meta
[^
>]+?
%s[^
>]+?
%s'
1302 template % (property_re, content_re),
1303 template % (content_re, property_re),
1307 def _meta_regex(prop):
1308 return r'''(?isx)<meta
1309 (?=[^>]+(?:itemprop|name|property|id|http-equiv)=(["\']?)%s\1)
1310 [^>]+?content=(["\'])(?P<content>.*?)\2''' % re.escape(prop)
1312 def _og_search_property(self, prop, html, name=None, **kargs):
1313 prop = variadic(prop)
1315 name = 'OpenGraph
%s' % prop[0]
1318 og_regexes.extend(self._og_regexes(p))
1319 escaped = self._search_regex(og_regexes, html, name, flags=re.DOTALL, **kargs)
1322 return unescapeHTML(escaped)
1324 def _og_search_thumbnail(self, html, **kargs):
1325 return self._og_search_property('image
', html, 'thumbnail URL
', fatal=False, **kargs)
1327 def _og_search_description(self, html, **kargs):
1328 return self._og_search_property('description
', html, fatal=False, **kargs)
1330 def _og_search_title(self, html, *, fatal=False, **kargs):
1331 return self._og_search_property('title
', html, fatal=fatal, **kargs)
1333 def _og_search_video_url(self, html, name='video url
', secure=True, **kargs):
1334 regexes = self._og_regexes('video
') + self._og_regexes('video
:url
')
1336 regexes = self._og_regexes('video
:secure_url
') + regexes
1337 return self._html_search_regex(regexes, html, name, **kargs)
1339 def _og_search_url(self, html, **kargs):
1340 return self._og_search_property('url
', html, **kargs)
1342 def _html_extract_title(self, html, name='title
', *, fatal=False, **kwargs):
1343 return self._html_search_regex(r'(?s
)<title
>([^
<]+)</title
>', html, name, fatal=fatal, **kwargs)
1345 def _html_search_meta(self, name, html, display_name=None, fatal=False, **kwargs):
1346 name = variadic(name)
1347 if display_name is None:
1348 display_name = name[0]
1349 return self._html_search_regex(
1350 [self._meta_regex(n) for n in name],
1351 html, display_name, fatal=fatal, group='content
', **kwargs)
1353 def _dc_search_uploader(self, html):
1354 return self._html_search_meta('dc
.creator
', html, 'uploader
')
1356 def _rta_search(self, html):
1357 # See http://www.rtalabel.org/index.php?content=howtofaq#single
1358 if re.search(r'(?ix
)<meta\s
+name
="rating"\s
+'
1359 r' content
="RTA-5042-1996-1400-1577-RTA"',
1364 def _media_rating_search(self, html):
1365 # See http://www.tjg-designs.com/WP/metadata-code-examples-adding-metadata-to-your-web-pages/
1366 rating = self._html_search_meta('rating
', html)
1378 return RATING_TABLE.get(rating.lower())
1380 def _family_friendly_search(self, html):
1381 # See http://schema.org/VideoObject
1382 family_friendly = self._html_search_meta(
1383 'isFamilyFriendly
', html, default=None)
1385 if not family_friendly:
1394 return RATING_TABLE.get(family_friendly.lower())
1396 def _twitter_search_player(self, html):
1397 return self._html_search_meta('twitter
:player
', html,
1398 'twitter card player
')
1400 def _search_json_ld(self, html, video_id, expected_type=None, **kwargs):
1401 json_ld_list = list(re.finditer(JSON_LD_RE, html))
1402 default = kwargs.get('default
', NO_DEFAULT)
1403 # JSON-LD may be malformed and thus `fatal` should be respected.
1404 # At the same time `default` may be passed that assumes `fatal=False`
1405 # for _search_regex. Let's simulate the same behavior here
as well
.
1406 fatal
= kwargs
.get('fatal', True) if default
is NO_DEFAULT
else False
1408 for mobj
in json_ld_list
:
1409 json_ld_item
= self
._parse
_json
(
1410 mobj
.group('json_ld'), video_id
, fatal
=fatal
)
1411 if not json_ld_item
:
1413 if isinstance(json_ld_item
, dict):
1414 json_ld
.append(json_ld_item
)
1415 elif isinstance(json_ld_item
, (list, tuple)):
1416 json_ld
.extend(json_ld_item
)
1418 json_ld
= self
._json
_ld
(json_ld
, video_id
, fatal
=fatal
, expected_type
=expected_type
)
1421 if default
is not NO_DEFAULT
:
1424 raise RegexNotFoundError('Unable to extract JSON-LD')
1426 self
.report_warning('unable to extract JSON-LD %s' % bug_reports_message())
1429 def _json_ld(self
, json_ld
, video_id
, fatal
=True, expected_type
=None):
1430 if isinstance(json_ld
, compat_str
):
1431 json_ld
= self
._parse
_json
(json_ld
, video_id
, fatal
=fatal
)
1435 if not isinstance(json_ld
, (list, tuple, dict)):
1437 if isinstance(json_ld
, dict):
1440 INTERACTION_TYPE_MAP
= {
1441 'CommentAction': 'comment',
1442 'AgreeAction': 'like',
1443 'DisagreeAction': 'dislike',
1444 'LikeAction': 'like',
1445 'DislikeAction': 'dislike',
1446 'ListenAction': 'view',
1447 'WatchAction': 'view',
1448 'ViewAction': 'view',
1451 def extract_interaction_type(e
):
1452 interaction_type
= e
.get('interactionType')
1453 if isinstance(interaction_type
, dict):
1454 interaction_type
= interaction_type
.get('@type')
1455 return str_or_none(interaction_type
)
1457 def extract_interaction_statistic(e
):
1458 interaction_statistic
= e
.get('interactionStatistic')
1459 if isinstance(interaction_statistic
, dict):
1460 interaction_statistic
= [interaction_statistic
]
1461 if not isinstance(interaction_statistic
, list):
1463 for is_e
in interaction_statistic
:
1464 if not isinstance(is_e
, dict):
1466 if is_e
.get('@type') != 'InteractionCounter':
1468 interaction_type
= extract_interaction_type(is_e
)
1469 if not interaction_type
:
1471 # For interaction count some sites provide string instead of
1472 # an integer (as per spec) with non digit characters (e.g. ",")
1473 # so extracting count with more relaxed str_to_int
1474 interaction_count
= str_to_int(is_e
.get('userInteractionCount'))
1475 if interaction_count
is None:
1477 count_kind
= INTERACTION_TYPE_MAP
.get(interaction_type
.split('/')[-1])
1480 count_key
= '%s_count' % count_kind
1481 if info
.get(count_key
) is not None:
1483 info
[count_key
] = interaction_count
1485 def extract_chapter_information(e
):
1487 'title': part
.get('name'),
1488 'start_time': part
.get('startOffset'),
1489 'end_time': part
.get('endOffset'),
1490 } for part
in variadic(e
.get('hasPart') or []) if part
.get('@type') == 'Clip']
1491 for idx
, (last_c
, current_c
, next_c
) in enumerate(zip(
1492 [{'end_time': 0}
] + chapters
, chapters
, chapters
[1:])):
1493 current_c
['end_time'] = current_c
['end_time'] or next_c
['start_time']
1494 current_c
['start_time'] = current_c
['start_time'] or last_c
['end_time']
1495 if None in current_c
.values():
1496 self
.report_warning(f
'Chapter {idx} contains broken data. Not extracting chapters')
1499 chapters
[-1]['end_time'] = chapters
[-1]['end_time'] or info
['duration']
1500 info
['chapters'] = chapters
1502 def extract_video_object(e
):
1503 assert e
['@type'] == 'VideoObject'
1504 author
= e
.get('author')
1506 'url': url_or_none(e
.get('contentUrl')),
1507 'title': unescapeHTML(e
.get('name')),
1508 'description': unescapeHTML(e
.get('description')),
1509 'thumbnails': [{'url': url_or_none(url)}
1510 for url
in variadic(traverse_obj(e
, 'thumbnailUrl', 'thumbnailURL'))],
1511 'duration': parse_duration(e
.get('duration')),
1512 'timestamp': unified_timestamp(e
.get('uploadDate')),
1513 # author can be an instance of 'Organization' or 'Person' types.
1514 # both types can have 'name' property(inherited from 'Thing' type). [1]
1515 # however some websites are using 'Text' type instead.
1516 # 1. https://schema.org/VideoObject
1517 'uploader': author
.get('name') if isinstance(author
, dict) else author
if isinstance(author
, compat_str
) else None,
1518 'filesize': float_or_none(e
.get('contentSize')),
1519 'tbr': int_or_none(e
.get('bitrate')),
1520 'width': int_or_none(e
.get('width')),
1521 'height': int_or_none(e
.get('height')),
1522 'view_count': int_or_none(e
.get('interactionCount')),
1524 extract_interaction_statistic(e
)
1525 extract_chapter_information(e
)
1527 def traverse_json_ld(json_ld
, at_top_level
=True):
1529 if at_top_level
and '@context' not in e
:
1531 if at_top_level
and set(e
.keys()) == {'@context', '@graph'}
:
1532 traverse_json_ld(variadic(e
['@graph'], allowed_types
=(dict,)), at_top_level
=False)
1534 item_type
= e
.get('@type')
1535 if expected_type
is not None and expected_type
!= item_type
:
1537 rating
= traverse_obj(e
, ('aggregateRating', 'ratingValue'), expected_type
=float_or_none
)
1538 if rating
is not None:
1539 info
['average_rating'] = rating
1540 if item_type
in ('TVEpisode', 'Episode'):
1541 episode_name
= unescapeHTML(e
.get('name'))
1543 'episode': episode_name
,
1544 'episode_number': int_or_none(e
.get('episodeNumber')),
1545 'description': unescapeHTML(e
.get('description')),
1547 if not info
.get('title') and episode_name
:
1548 info
['title'] = episode_name
1549 part_of_season
= e
.get('partOfSeason')
1550 if isinstance(part_of_season
, dict) and part_of_season
.get('@type') in ('TVSeason', 'Season', 'CreativeWorkSeason'):
1552 'season': unescapeHTML(part_of_season
.get('name')),
1553 'season_number': int_or_none(part_of_season
.get('seasonNumber')),
1555 part_of_series
= e
.get('partOfSeries') or e
.get('partOfTVSeries')
1556 if isinstance(part_of_series
, dict) and part_of_series
.get('@type') in ('TVSeries', 'Series', 'CreativeWorkSeries'):
1557 info
['series'] = unescapeHTML(part_of_series
.get('name'))
1558 elif item_type
== 'Movie':
1560 'title': unescapeHTML(e
.get('name')),
1561 'description': unescapeHTML(e
.get('description')),
1562 'duration': parse_duration(e
.get('duration')),
1563 'timestamp': unified_timestamp(e
.get('dateCreated')),
1565 elif item_type
in ('Article', 'NewsArticle'):
1567 'timestamp': parse_iso8601(e
.get('datePublished')),
1568 'title': unescapeHTML(e
.get('headline')),
1569 'description': unescapeHTML(e
.get('articleBody') or e
.get('description')),
1571 if traverse_obj(e
, ('video', 0, '@type')) == 'VideoObject':
1572 extract_video_object(e
['video'][0])
1573 elif item_type
== 'VideoObject':
1574 extract_video_object(e
)
1575 if expected_type
is None:
1579 video
= e
.get('video')
1580 if isinstance(video
, dict) and video
.get('@type') == 'VideoObject':
1581 extract_video_object(video
)
1582 if expected_type
is None:
1586 traverse_json_ld(json_ld
)
1588 return filter_dict(info
)
1590 def _search_nextjs_data(self
, webpage
, video_id
, *, transform_source
=None, fatal
=True, **kw
):
1591 return self
._parse
_json
(
1593 r
'(?s)<script[^>]+id=[\'"]__NEXT_DATA__[\'"][^
>]*>([^
<]+)</script
>',
1594 webpage, 'next
.js data
', fatal=fatal, **kw),
1595 video_id, transform_source=transform_source, fatal=fatal)
1597 def _search_nuxt_data(self, webpage, video_id, context_name='__NUXT__
'):
1598 ''' Parses Nuxt.js metadata. This works as long as the function __NUXT__ invokes is a pure function. '''
1599 # not all website do this, but it can be changed
1600 # https://stackoverflow.com/questions/67463109/how-to-change-or-hide-nuxt-and-nuxt-keyword-in-page-source
1601 rectx = re.escape(context_name)
1602 js, arg_keys, arg_vals = self._search_regex(
1603 (r'<script
>window\
.%s=\
(function\
((?P
<arg_keys
>.*?
)\
)\{return\s(?P<js>\{.*?\}
)\
}\
((?P
<arg_vals
>.+?
)\
)\
);?
</script
>' % rectx,
1604 r'%s\
(.*?\
(function\
((?P
<arg_keys
>.*?
)\
)\{return\s(?P<js>\{.*?\}
)\
}\
((?P
<arg_vals
>.*?
)\
)' % rectx),
1605 webpage, context_name, group=['js
', 'arg_keys
', 'arg_vals
'])
1607 args = dict(zip(arg_keys.split(','), arg_vals.split(',')))
1609 for key, val in args.items():
1610 if val in ('undefined
', 'void
0'):
1613 return self._parse_json(js_to_json(js, args), video_id)['data
'][0]
1616 def _hidden_inputs(html):
1617 html = re.sub(r'<!--(?
:(?
!<!--).)*-->', '', html)
1619 for input in re.findall(r'(?i
)(<input[^
>]+>)', html):
1620 attrs = extract_attributes(input)
1623 if attrs.get('type') not in ('hidden
', 'submit
'):
1625 name = attrs.get('name
') or attrs.get('id')
1626 value = attrs.get('value
')
1627 if name and value is not None:
1628 hidden_inputs[name] = value
1629 return hidden_inputs
1631 def _form_hidden_inputs(self, form_id, html):
1632 form = self._search_regex(
1633 r'(?
is)<form
[^
>]+?
id=(["\'])%s\1[^>]*>(?P<form>.+?)</form>' % form_id,
1634 html, '%s form' % form_id, group='form')
1635 return self._hidden_inputs(form)
1638 regex = r' *((?P<reverse>\+)?(?P<field>[a-zA-Z0-9_]+)((?P<separator>[~:])(?P<limit>.*?))?)? *$'
1640 default = ('hidden', 'aud_or_vid', 'hasvid', 'ie_pref', 'lang', 'quality',
1641 'res', 'fps', 'hdr:12', 'codec:vp9.2', 'size', 'br', 'asr',
1642 'proto', 'ext', 'hasaud', 'source', 'id') # These must not be aliases
1643 ytdl_default = ('hasaud', 'lang', 'quality', 'tbr', 'filesize', 'vbr',
1644 'height', 'width', 'proto', 'vext', 'abr', 'aext',
1645 'fps', 'fs_approx', 'source', 'id')
1648 'vcodec': {'type': 'ordered', 'regex': True,
1649 'order': ['av0?1', 'vp0?9.2', 'vp0?9', '[hx]265|he?vc?', '[hx]264|avc', 'vp0?8', 'mp4v|h263', 'theora', '', None, 'none']},
1650 'acodec': {'type': 'ordered', 'regex': True,
1651 'order': ['[af]lac', 'wav|aiff', 'opus', 'vorbis|ogg', 'aac', 'mp?4a?', 'mp3', 'e-?a?c-?3', 'ac-?3', 'dts', '', None, 'none']},
1652 'hdr': {'type': 'ordered', 'regex': True, 'field': 'dynamic_range',
1653 'order': ['dv', '(hdr)?12', r'(hdr)?10\+', '(hdr)?10', 'hlg', '', 'sdr', None]},
1654 'proto': {'type': 'ordered', 'regex': True, 'field': 'protocol',
1655 'order': ['(ht|f)tps', '(ht|f)tp$', 'm3u8.*', '.*dash', 'websocket_frag', 'rtmpe?', '', 'mms|rtsp', 'ws|websocket', 'f4']},
1656 'vext': {'type': 'ordered', 'field': 'video_ext',
1657 'order': ('mp4', 'webm', 'flv', '', 'none'),
1658 'order_free': ('webm', 'mp4', 'flv', '', 'none')},
1659 'aext': {'type': 'ordered', 'field': 'audio_ext',
1660 'order': ('m4a', 'aac', 'mp3', 'ogg', 'opus', 'webm', '', 'none'),
1661 'order_free': ('opus', 'ogg', 'webm', 'm4a', 'mp3', 'aac', '', 'none')},
1662 'hidden': {'visible': False, 'forced': True, 'type': 'extractor', 'max': -1000},
1663 'aud_or_vid': {'visible': False, 'forced': True, 'type': 'multiple',
1664 'field': ('vcodec', 'acodec'),
1665 'function': lambda it: int(any(v != 'none' for v in it))},
1666 'ie_pref': {'priority': True, 'type': 'extractor'},
1667 'hasvid': {'priority': True, 'field': 'vcodec', 'type': 'boolean', 'not_in_list': ('none',)},
1668 'hasaud': {'field': 'acodec', 'type': 'boolean', 'not_in_list': ('none',)},
1669 'lang': {'convert': 'float', 'field': 'language_preference', 'default': -1},
1670 'quality': {'convert': 'float', 'default': -1},
1671 'filesize': {'convert': 'bytes'},
1672 'fs_approx': {'convert': 'bytes', 'field': 'filesize_approx'},
1673 'id': {'convert': 'string', 'field': 'format_id'},
1674 'height': {'convert': 'float_none'},
1675 'width': {'convert': 'float_none'},
1676 'fps': {'convert': 'float_none'},
1677 'tbr': {'convert': 'float_none'},
1678 'vbr': {'convert': 'float_none'},
1679 'abr': {'convert': 'float_none'},
1680 'asr': {'convert': 'float_none'},
1681 'source': {'convert': 'float', 'field': 'source_preference', 'default': -1},
1683 'codec': {'type': 'combined', 'field': ('vcodec', 'acodec')},
1684 'br': {'type': 'combined', 'field': ('tbr', 'vbr', 'abr'), 'same_limit': True},
1685 'size': {'type': 'combined', 'same_limit': True, 'field': ('filesize', 'fs_approx')},
1686 'ext': {'type': 'combined', 'field': ('vext', 'aext')},
1687 'res': {'type': 'multiple', 'field': ('height', 'width'),
1688 'function': lambda it: (lambda l: min(l) if l else 0)(tuple(filter(None, it)))},
1690 # For compatibility with youtube-dl
1691 'format_id': {'type': 'alias', 'field': 'id'},
1692 'preference': {'type': 'alias', 'field': 'ie_pref'},
1693 'language_preference': {'type': 'alias', 'field': 'lang'},
1694 'source_preference': {'type': 'alias', 'field': 'source'},
1695 'protocol': {'type': 'alias', 'field': 'proto'},
1696 'filesize_approx': {'type': 'alias', 'field': 'fs_approx'},
1699 'dimension': {'type': 'alias', 'field': 'res', 'deprecated': True},
1700 'resolution': {'type': 'alias', 'field': 'res', 'deprecated': True},
1701 'extension': {'type': 'alias', 'field': 'ext', 'deprecated': True},
1702 'bitrate': {'type': 'alias', 'field': 'br', 'deprecated': True},
1703 'total_bitrate': {'type': 'alias', 'field': 'tbr', 'deprecated': True},
1704 'video_bitrate': {'type': 'alias', 'field': 'vbr', 'deprecated': True},
1705 'audio_bitrate': {'type': 'alias', 'field': 'abr', 'deprecated': True},
1706 'framerate': {'type': 'alias', 'field': 'fps', 'deprecated': True},
1707 'filesize_estimate': {'type': 'alias', 'field': 'size', 'deprecated': True},
1708 'samplerate': {'type': 'alias', 'field': 'asr', 'deprecated': True},
1709 'video_ext': {'type': 'alias', 'field': 'vext', 'deprecated': True},
1710 'audio_ext': {'type': 'alias', 'field': 'aext', 'deprecated': True},
1711 'video_codec': {'type': 'alias', 'field': 'vcodec', 'deprecated': True},
1712 'audio_codec': {'type': 'alias', 'field': 'acodec', 'deprecated': True},
1713 'video': {'type': 'alias', 'field': 'hasvid', 'deprecated': True},
1714 'has_video': {'type': 'alias', 'field': 'hasvid', 'deprecated': True},
1715 'audio': {'type': 'alias', 'field': 'hasaud', 'deprecated': True},
1716 'has_audio': {'type': 'alias', 'field': 'hasaud', 'deprecated': True},
1717 'extractor': {'type': 'alias', 'field': 'ie_pref', 'deprecated': True},
1718 'extractor_preference': {'type': 'alias', 'field': 'ie_pref', 'deprecated': True},
1721 def __init__(self, ie, field_preference):
1723 self.ydl = ie._downloader
1724 self.evaluate_params(self.ydl.params, field_preference)
1725 if ie.get_param('verbose'):
1726 self.print_verbose_info(self.ydl.write_debug)
1728 def _get_field_setting(self, field, key):
1729 if field not in self.settings:
1730 if key in ('forced', 'priority'):
1732 self.ydl.deprecation_warning(
1733 f'Using arbitrary fields ({field}) for format sorting is deprecated '
1734 'and may be removed in a future version')
1735 self.settings[field] = {}
1736 propObj = self.settings[field]
1737 if key not in propObj:
1738 type = propObj.get('type')
1740 default = 'preference' if type == 'extractor' else (field,) if type in ('combined', 'multiple') else field
1741 elif key == 'convert':
1742 default = 'order' if type == 'ordered' else 'float_string' if field else 'ignore'
1744 default = {'type': 'field', 'visible': True, 'order': [], 'not_in_list': (None,)}.get(key, None)
1745 propObj[key] = default
1748 def _resolve_field_value(self, field, value, convertNone=False):
1753 value = value.lower()
1754 conversion = self._get_field_setting(field, 'convert')
1755 if conversion == 'ignore':
1757 if conversion == 'string':
1759 elif conversion == 'float_none':
1760 return float_or_none(value)
1761 elif conversion == 'bytes':
1762 return FileDownloader.parse_bytes(value)
1763 elif conversion == 'order':
1764 order_list = (self._use_free_order and self._get_field_setting(field, 'order_free')) or self._get_field_setting(field, 'order')
1765 use_regex = self._get_field_setting(field, 'regex')
1766 list_length = len(order_list)
1767 empty_pos = order_list.index('') if '' in order_list else list_length + 1
1768 if use_regex and value is not None:
1769 for i, regex in enumerate(order_list):
1770 if regex and re.match(regex, value):
1771 return list_length - i
1772 return list_length - empty_pos # not in list
1773 else: # not regex or value = None
1774 return list_length - (order_list.index(value) if value in order_list else empty_pos)
1776 if value.isnumeric():
1779 self.settings[field]['convert'] = 'string'
1782 def evaluate_params(self, params, sort_extractor):
1783 self._use_free_order = params.get('prefer_free_formats', False)
1784 self._sort_user = params.get('format_sort', [])
1785 self._sort_extractor = sort_extractor
1787 def add_item(field, reverse, closest, limit_text):
1788 field = field.lower()
1789 if field in self._order:
1791 self._order.append(field)
1792 limit = self._resolve_field_value(field, limit_text)
1795 'closest': False if limit is None else closest,
1796 'limit_text': limit_text,
1798 if field in self.settings:
1799 self.settings[field].update(data)
1801 self.settings[field] = data
1804 tuple(field for field in self.default if self._get_field_setting(field, 'forced'))
1805 + (tuple() if params.get('format_sort_force', False)
1806 else tuple(field for field in self.default if self._get_field_setting(field, 'priority')))
1807 + tuple(self._sort_user) + tuple(sort_extractor) + self.default)
1809 for item in sort_list:
1810 match = re.match(self.regex, item)
1812 raise ExtractorError('Invalid format sort string "%s" given by extractor' % item)
1813 field = match.group('field')
1816 if self._get_field_setting(field, 'type') == 'alias':
1817 alias, field = field, self._get_field_setting(field, 'field')
1818 if self._get_field_setting(alias, 'deprecated'):
1819 self.ydl.deprecation_warning(
1820 f'Format sorting alias {alias} is deprecated '
1821 f'and may be removed in a future version. Please use {field} instead')
1822 reverse = match.group('reverse') is not None
1823 closest = match.group('separator') == '~'
1824 limit_text = match.group('limit')
1826 has_limit = limit_text is not None
1827 has_multiple_fields = self._get_field_setting(field, 'type') == 'combined'
1828 has_multiple_limits = has_limit and has_multiple_fields and not self._get_field_setting(field, 'same_limit')
1830 fields = self._get_field_setting(field, 'field') if has_multiple_fields else (field,)
1831 limits = limit_text.split(':') if has_multiple_limits else (limit_text,) if has_limit else tuple()
1832 limit_count = len(limits)
1833 for (i, f) in enumerate(fields):
1834 add_item(f, reverse, closest,
1835 limits[i] if i < limit_count
1836 else limits[0] if has_limit and not has_multiple_limits
1839 def print_verbose_info(self, write_debug):
1841 write_debug('Sort order given by user: %s' % ', '.join(self._sort_user))
1842 if self._sort_extractor:
1843 write_debug('Sort order given by extractor: %s' % ', '.join(self._sort_extractor))
1844 write_debug('Formats sorted by: %s' % ', '.join(['%s%s%s' % (
1845 '+' if self._get_field_setting(field, 'reverse') else '', field,
1846 '%s%s(%s)' % ('~' if self._get_field_setting(field, 'closest') else ':',
1847 self._get_field_setting(field, 'limit_text'),
1848 self._get_field_setting(field, 'limit'))
1849 if self._get_field_setting(field, 'limit_text') is not None else '')
1850 for field in self._order if self._get_field_setting(field, 'visible')]))
1852 def _calculate_field_preference_from_value(self, format, field, type, value):
1853 reverse = self._get_field_setting(field, 'reverse')
1854 closest = self._get_field_setting(field, 'closest')
1855 limit = self._get_field_setting(field, 'limit')
1857 if type == 'extractor':
1858 maximum = self._get_field_setting(field, 'max')
1859 if value is None or (maximum is not None and value >= maximum):
1861 elif type == 'boolean':
1862 in_list = self._get_field_setting(field, 'in_list')
1863 not_in_list = self._get_field_setting(field, 'not_in_list')
1864 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
1865 elif type == 'ordered':
1866 value = self._resolve_field_value(field, value, True)
1868 # try to convert to number
1869 val_num = float_or_none(value, default=self._get_field_setting(field, 'default'))
1870 is_num = self._get_field_setting(field, 'convert') != 'string' and val_num is not None
1874 return ((-10, 0) if value is None
1875 else (1, value, 0) if not is_num # if a field has mixed strings and numbers, strings are sorted higher
1876 else (0, -abs(value - limit), value - limit if reverse else limit - value) if closest
1877 else (0, value, 0) if not reverse and (limit is None or value <= limit)
1878 else (0, -value, 0) if limit is None or (reverse and value == limit) or value > limit
1879 else (-1, value, 0))
1881 def _calculate_field_preference(self, format, field):
1882 type = self._get_field_setting(field, 'type') # extractor, boolean, ordered, field, multiple
1883 get_value = lambda f: format.get(self._get_field_setting(f, 'field'))
1884 if type == 'multiple':
1885 type = 'field' # Only 'field' is allowed in multiple for now
1886 actual_fields = self._get_field_setting(field, 'field')
1888 value = self._get_field_setting(field, 'function')(get_value(f) for f in actual_fields)
1890 value = get_value(field)
1891 return self._calculate_field_preference_from_value(format, field, type, value)
1893 def calculate_preference(self, format):
1894 # Determine missing protocol
1895 if not format.get('protocol'):
1896 format['protocol'] = determine_protocol(format)
1898 # Determine missing ext
1899 if not format.get('ext') and 'url' in format:
1900 format['ext'] = determine_ext(format['url'])
1901 if format.get('vcodec') == 'none':
1902 format['audio_ext'] = format['ext'] if format.get('acodec') != 'none' else 'none'
1903 format['video_ext'] = 'none'
1905 format['video_ext'] = format['ext']
1906 format['audio_ext'] = 'none'
1907 # if format.get('preference') is None and format.get('ext') in ('f4f', 'f4m'): # Not supported?
1908 # format['preference'] = -1000
1910 # Determine missing bitrates
1911 if format.get('tbr') is None:
1912 if format.get('vbr') is not None and format.get('abr') is not None:
1913 format['tbr'] = format.get('vbr', 0) + format.get('abr', 0)
1915 if format.get('vcodec') != 'none' and format.get('vbr') is None:
1916 format['vbr'] = format.get('tbr') - format.get('abr', 0)
1917 if format.get('acodec') != 'none' and format.get('abr') is None:
1918 format['abr'] = format.get('tbr') - format.get('vbr', 0)
1920 return tuple(self._calculate_field_preference(format, field) for field in self._order)
1922 def _sort_formats(self, formats, field_preference=[]):
1925 formats.sort(key=self.FormatSort(self, field_preference).calculate_preference)
1927 def _check_formats(self, formats, video_id):
1929 formats[:] = filter(
1930 lambda f: self._is_valid_url(
1932 item='%s video format' % f.get('format_id') if f.get('format_id') else 'video'),
1936 def _remove_duplicate_formats(formats):
1940 if f['url'] not in format_urls:
1941 format_urls.add(f['url'])
1942 unique_formats.append(f)
1943 formats[:] = unique_formats
1945 def _is_valid_url(self, url, video_id, item='video', headers={}):
1946 url = self._proto_relative_url(url, scheme='http:')
1947 # For now assume non HTTP(S) URLs always valid
1948 if not (url.startswith('http://') or url.startswith('https://')):
1951 self._request_webpage(url, video_id, 'Checking %s URL' % item, headers=headers)
1953 except ExtractorError as e:
1955 '%s: %s URL is invalid, skipping: %s'
1956 % (video_id, item, error_to_compat_str(e.cause)))
1959 def http_scheme(self):
1960 """ Either "http
:" or "https
:", depending on the user's preferences """
1963 if self.get_param('prefer_insecure', False)
1966 def _proto_relative_url(self, url, scheme=None):
1969 if url.startswith('//'):
1971 scheme = self.http_scheme()
1976 def _sleep(self, timeout, video_id, msg_template=None):
1977 if msg_template is None:
1978 msg_template = '%(video_id)s: Waiting for %(timeout)s seconds'
1979 msg = msg_template % {'video_id': video_id, 'timeout': timeout}
1983 def _extract_f4m_formats(self, manifest_url, video_id, preference=None, quality=None, f4m_id=None,
1984 transform_source=lambda s: fix_xml_ampersands(s).strip(),
1985 fatal=True, m3u8_id=None, data=None, headers={}, query={}):
1986 res = self._download_xml_handle(
1987 manifest_url, video_id, 'Downloading f4m manifest',
1988 'Unable to download f4m manifest',
1989 # Some manifests may be malformed, e.g. prosiebensat1 generated manifests
1990 # (see https://github.com/ytdl-org/youtube-dl/issues/6215#issuecomment-121704244)
1991 transform_source=transform_source,
1992 fatal=fatal, data=data, headers=headers, query=query)
1996 manifest, urlh = res
1997 manifest_url = urlh.geturl()
1999 return self._parse_f4m_formats(
2000 manifest, manifest_url, video_id, preference=preference, quality=quality, f4m_id=f4m_id,
2001 transform_source=transform_source, fatal=fatal, m3u8_id=m3u8_id)
2003 def _parse_f4m_formats(self, manifest, manifest_url, video_id, preference=None, quality=None, f4m_id=None,
2004 transform_source=lambda s: fix_xml_ampersands(s).strip(),
2005 fatal=True, m3u8_id=None):
2006 if not isinstance(manifest, xml.etree.ElementTree.Element) and not fatal:
2009 # currently yt-dlp cannot decode the playerVerificationChallenge as Akamai uses Adobe Alchemy
2010 akamai_pv = manifest.find('{http://ns.adobe.com/f4m/1.0}pv-2.0')
2011 if akamai_pv is not None and ';' in akamai_pv.text:
2012 playerVerificationChallenge = akamai_pv.text.split(';')[0]
2013 if playerVerificationChallenge.strip() != '':
2017 manifest_version = '1.0'
2018 media_nodes = manifest.findall('{http://ns.adobe.com/f4m/1.0}media')
2020 manifest_version = '2.0'
2021 media_nodes = manifest.findall('{http://ns.adobe.com/f4m/2.0}media')
2022 # Remove unsupported DRM protected media from final formats
2023 # rendition (see https://github.com/ytdl-org/youtube-dl/issues/8573).
2024 media_nodes = remove_encrypted_media(media_nodes)
2028 manifest_base_url = get_base_url(manifest)
2030 bootstrap_info = xpath_element(
2031 manifest, ['{http://ns.adobe.com/f4m/1.0}bootstrapInfo', '{http://ns.adobe.com/f4m/2.0}bootstrapInfo'],
2032 'bootstrap info', default=None)
2035 mime_type = xpath_text(
2036 manifest, ['{http://ns.adobe.com/f4m/1.0}mimeType', '{http://ns.adobe.com/f4m/2.0}mimeType'],
2037 'base URL', default=None)
2038 if mime_type and mime_type.startswith('audio/'):
2041 for i, media_el in enumerate(media_nodes):
2042 tbr = int_or_none(media_el.attrib.get('bitrate'))
2043 width = int_or_none(media_el.attrib.get('width'))
2044 height = int_or_none(media_el.attrib.get('height'))
2045 format_id = join_nonempty(f4m_id, tbr or i)
2046 # If <bootstrapInfo> is present, the specified f4m is a
2047 # stream-level manifest, and only set-level manifests may refer to
2048 # external resources. See section 11.4 and section 4 of F4M spec
2049 if bootstrap_info is None:
2051 # @href is introduced in 2.0, see section 11.6 of F4M spec
2052 if manifest_version == '2.0':
2053 media_url = media_el.attrib.get('href')
2054 if media_url is None:
2055 media_url = media_el.attrib.get('url')
2059 media_url if media_url.startswith('http://') or media_url.startswith('https://')
2060 else ((manifest_base_url or '/'.join(manifest_url.split('/')[:-1])) + '/' + media_url))
2061 # If media_url is itself a f4m manifest do the recursive extraction
2062 # since bitrates in parent manifest (this one) and media_url manifest
2063 # may differ leading to inability to resolve the format by requested
2064 # bitrate in f4m downloader
2065 ext = determine_ext(manifest_url)
2067 f4m_formats = self._extract_f4m_formats(
2068 manifest_url, video_id, preference=preference, quality=quality, f4m_id=f4m_id,
2069 transform_source=transform_source, fatal=fatal)
2070 # Sometimes stream-level manifest contains single media entry that
2071 # does not contain any quality metadata (e.g. http://matchtv.ru/#live-player).
2072 # At the same time parent's media entry in set-level manifest may
2073 # contain it. We will copy it from parent in such cases.
2074 if len(f4m_formats) == 1:
2077 'tbr': f.get('tbr') or tbr,
2078 'width': f.get('width') or width,
2079 'height': f.get('height') or height,
2080 'format_id': f.get('format_id') if not tbr else format_id,
2083 formats.extend(f4m_formats)
2086 formats.extend(self._extract_m3u8_formats(
2087 manifest_url, video_id, 'mp4', preference=preference,
2088 quality=quality, m3u8_id=m3u8_id, fatal=fatal))
2091 'format_id': format_id,
2092 'url': manifest_url,
2093 'manifest_url': manifest_url,
2094 'ext': 'flv' if bootstrap_info is not None else None,
2100 'preference': preference,
2105 def _m3u8_meta_format(self, m3u8_url, ext=None, preference=None, quality=None, m3u8_id=None):
2107 'format_id': join_nonempty(m3u8_id, 'meta'),
2111 'preference': preference - 100 if preference else -100,
2113 'resolution': 'multiple',
2114 'format_note': 'Quality selection URL',
2117 def _report_ignoring_subs(self, name):
2118 self.report_warning(bug_reports_message(
2119 f'Ignoring subtitle tracks found in the {name} manifest; '
2120 'if any subtitle tracks are missing,'
2123 def _extract_m3u8_formats(self, *args, **kwargs):
2124 fmts, subs = self._extract_m3u8_formats_and_subtitles(*args, **kwargs)
2126 self._report_ignoring_subs('HLS')
2129 def _extract_m3u8_formats_and_subtitles(
2130 self, m3u8_url, video_id, ext=None, entry_protocol='m3u8_native',
2131 preference=None, quality=None, m3u8_id=None, note=None,
2132 errnote=None, fatal=True, live=False, data=None, headers={},
2135 res = self._download_webpage_handle(
2137 note='Downloading m3u8 information' if note is None else note,
2138 errnote='Failed to download m3u8 information' if errnote is None else errnote,
2139 fatal=fatal, data=data, headers=headers, query=query)
2144 m3u8_doc, urlh = res
2145 m3u8_url = urlh.geturl()
2147 return self._parse_m3u8_formats_and_subtitles(
2148 m3u8_doc, m3u8_url, ext=ext, entry_protocol=entry_protocol,
2149 preference=preference, quality=quality, m3u8_id=m3u8_id,
2150 note=note, errnote=errnote, fatal=fatal, live=live, data=data,
2151 headers=headers, query=query, video_id=video_id)
2153 def _parse_m3u8_formats_and_subtitles(
2154 self, m3u8_doc, m3u8_url=None, ext=None, entry_protocol='m3u8_native',
2155 preference=None, quality=None, m3u8_id=None, live=False, note=None,
2156 errnote=None, fatal=True, data=None, headers={}, query={},
2158 formats, subtitles = [], {}
2160 has_drm = re.search('|'.join([
2161 r'#EXT-X-FAXS-CM:', # Adobe Flash Access
2162 r'#EXT-X-(?:SESSION-)?KEY:.*?URI="skd
://', # Apple FairPlay
2165 def format_url(url):
2166 return url if re.match(r'^https?
://', url) else compat_urlparse.urljoin(m3u8_url, url)
2168 if self.get_param('hls_split_discontinuity
', False):
2169 def _extract_m3u8_playlist_indices(manifest_url=None, m3u8_doc=None):
2171 if not manifest_url:
2173 m3u8_doc = self._download_webpage(
2174 manifest_url, video_id, fatal=fatal, data=data, headers=headers,
2175 note=False, errnote='Failed to download m3u8 playlist information
')
2176 if m3u8_doc is False:
2178 return range(1 + sum(line.startswith('#EXT-X-DISCONTINUITY') for line in m3u8_doc.splitlines()))
2181 def _extract_m3u8_playlist_indices(*args
, **kwargs
):
2185 # 1. https://tools.ietf.org/html/draft-pantos-http-live-streaming-21
2186 # 2. https://github.com/ytdl-org/youtube-dl/issues/12211
2187 # 3. https://github.com/ytdl-org/youtube-dl/issues/18923
2189 # We should try extracting formats only from master playlists [1, 4.3.4],
2190 # i.e. playlists that describe available qualities. On the other hand
2191 # media playlists [1, 4.3.3] should be returned as is since they contain
2192 # just the media without qualities renditions.
2193 # Fortunately, master playlist can be easily distinguished from media
2194 # playlist based on particular tags availability. As of [1, 4.3.3, 4.3.4]
2195 # master playlist tags MUST NOT appear in a media playlist and vice versa.
2196 # As of [1, 4.3.3.1] #EXT-X-TARGETDURATION tag is REQUIRED for every
2197 # media playlist and MUST NOT appear in master playlist thus we can
2198 # clearly detect media playlist with this criterion.
2200 if '#EXT-X-TARGETDURATION' in m3u8_doc
: # media playlist, return as is
2202 'format_id': join_nonempty(m3u8_id
, idx
),
2203 'format_index': idx
,
2204 'url': m3u8_url
or encode_data_uri(m3u8_doc
.encode('utf-8'), 'application/x-mpegurl'),
2206 'protocol': entry_protocol
,
2207 'preference': preference
,
2210 } for idx
in _extract_m3u8_playlist_indices(m3u8_doc
=m3u8_doc
)]
2212 return formats
, subtitles
2215 last_stream_inf
= {}
2217 def extract_media(x_media_line
):
2218 media
= parse_m3u8_attributes(x_media_line
)
2219 # As per [1, 4.3.4.1] TYPE, GROUP-ID and NAME are REQUIRED
2220 media_type
, group_id
, name
= media
.get('TYPE'), media
.get('GROUP-ID'), media
.get('NAME')
2221 if not (media_type
and group_id
and name
):
2223 groups
.setdefault(group_id
, []).append(media
)
2224 # <https://tools.ietf.org/html/rfc8216#section-4.3.4.1>
2225 if media_type
== 'SUBTITLES':
2226 # According to RFC 8216 §4.3.4.2.1, URI is REQUIRED in the
2227 # EXT-X-MEDIA tag if the media type is SUBTITLES.
2228 # However, lack of URI has been spotted in the wild.
2229 # e.g. NebulaIE; see https://github.com/yt-dlp/yt-dlp/issues/339
2230 if not media
.get('URI'):
2232 url
= format_url(media
['URI'])
2235 'ext': determine_ext(url
),
2237 if sub_info
['ext'] == 'm3u8':
2238 # Per RFC 8216 §3.1, the only possible subtitle format m3u8
2239 # files may contain is WebVTT:
2240 # <https://tools.ietf.org/html/rfc8216#section-3.1>
2241 sub_info
['ext'] = 'vtt'
2242 sub_info
['protocol'] = 'm3u8_native'
2243 lang
= media
.get('LANGUAGE') or 'und'
2244 subtitles
.setdefault(lang
, []).append(sub_info
)
2245 if media_type
not in ('VIDEO', 'AUDIO'):
2247 media_url
= media
.get('URI')
2249 manifest_url
= format_url(media_url
)
2251 'format_id': join_nonempty(m3u8_id
, group_id
, name
, idx
),
2252 'format_note': name
,
2253 'format_index': idx
,
2254 'url': manifest_url
,
2255 'manifest_url': m3u8_url
,
2256 'language': media
.get('LANGUAGE'),
2258 'protocol': entry_protocol
,
2259 'preference': preference
,
2261 'vcodec': 'none' if media_type
== 'AUDIO' else None,
2262 } for idx
in _extract_m3u8_playlist_indices(manifest_url
))
2264 def build_stream_name():
2265 # Despite specification does not mention NAME attribute for
2266 # EXT-X-STREAM-INF tag it still sometimes may be present (see [1]
2267 # or vidio test in TestInfoExtractor.test_parse_m3u8_formats)
2268 # 1. http://www.vidio.com/watch/165683-dj_ambred-booyah-live-2015
2269 stream_name
= last_stream_inf
.get('NAME')
2272 # If there is no NAME in EXT-X-STREAM-INF it will be obtained
2273 # from corresponding rendition group
2274 stream_group_id
= last_stream_inf
.get('VIDEO')
2275 if not stream_group_id
:
2277 stream_group
= groups
.get(stream_group_id
)
2278 if not stream_group
:
2279 return stream_group_id
2280 rendition
= stream_group
[0]
2281 return rendition
.get('NAME') or stream_group_id
2283 # parse EXT-X-MEDIA tags before EXT-X-STREAM-INF in order to have the
2284 # chance to detect video only formats when EXT-X-STREAM-INF tags
2285 # precede EXT-X-MEDIA tags in HLS manifest such as [3].
2286 for line
in m3u8_doc
.splitlines():
2287 if line
.startswith('#EXT-X-MEDIA:'):
2290 for line
in m3u8_doc
.splitlines():
2291 if line
.startswith('#EXT-X-STREAM-INF:'):
2292 last_stream_inf
= parse_m3u8_attributes(line
)
2293 elif line
.startswith('#') or not line
.strip():
2296 tbr
= float_or_none(
2297 last_stream_inf
.get('AVERAGE-BANDWIDTH')
2298 or last_stream_inf
.get('BANDWIDTH'), scale
=1000)
2299 manifest_url
= format_url(line
.strip())
2301 for idx
in _extract_m3u8_playlist_indices(manifest_url
):
2302 format_id
= [m3u8_id
, None, idx
]
2303 # Bandwidth of live streams may differ over time thus making
2304 # format_id unpredictable. So it's better to keep provided
2307 stream_name
= build_stream_name()
2308 format_id
[1] = stream_name
or '%d' % (tbr
or len(formats
))
2310 'format_id': join_nonempty(*format_id
),
2311 'format_index': idx
,
2312 'url': manifest_url
,
2313 'manifest_url': m3u8_url
,
2316 'fps': float_or_none(last_stream_inf
.get('FRAME-RATE')),
2317 'protocol': entry_protocol
,
2318 'preference': preference
,
2321 resolution
= last_stream_inf
.get('RESOLUTION')
2323 mobj
= re
.search(r
'(?P<width>\d+)[xX](?P<height>\d+)', resolution
)
2325 f
['width'] = int(mobj
.group('width'))
2326 f
['height'] = int(mobj
.group('height'))
2327 # Unified Streaming Platform
2329 r
'audio.*?(?:%3D|=)(\d+)(?:-video.*?(?:%3D|=)(\d+))?', f
['url'])
2331 abr
, vbr
= mobj
.groups()
2332 abr
, vbr
= float_or_none(abr
, 1000), float_or_none(vbr
, 1000)
2337 codecs
= parse_codecs(last_stream_inf
.get('CODECS'))
2339 audio_group_id
= last_stream_inf
.get('AUDIO')
2340 # As per [1, 4.3.4.1.1] any EXT-X-STREAM-INF tag which
2341 # references a rendition group MUST have a CODECS attribute.
2342 # However, this is not always respected, for example, [2]
2343 # contains EXT-X-STREAM-INF tag which references AUDIO
2344 # rendition group but does not have CODECS and despite
2345 # referencing an audio group it represents a complete
2346 # (with audio and video) format. So, for such cases we will
2347 # ignore references to rendition groups and treat them
2348 # as complete formats.
2349 if audio_group_id
and codecs
and f
.get('vcodec') != 'none':
2350 audio_group
= groups
.get(audio_group_id
)
2351 if audio_group
and audio_group
[0].get('URI'):
2352 # TODO: update acodec for audio only formats with
2354 f
['acodec'] = 'none'
2355 if not f
.get('ext'):
2356 f
['ext'] = 'm4a' if f
.get('vcodec') == 'none' else 'mp4'
2360 progressive_uri
= last_stream_inf
.get('PROGRESSIVE-URI')
2363 del http_f
['manifest_url']
2365 'format_id': f
['format_id'].replace('hls-', 'http-'),
2367 'url': progressive_uri
,
2369 formats
.append(http_f
)
2371 last_stream_inf
= {}
2372 return formats
, subtitles
2374 def _extract_m3u8_vod_duration(
2375 self
, m3u8_vod_url
, video_id
, note
=None, errnote
=None, data
=None, headers
={}, query={}
):
2377 m3u8_vod
= self
._download
_webpage
(
2378 m3u8_vod_url
, video_id
,
2379 note
='Downloading m3u8 VOD manifest' if note
is None else note
,
2380 errnote
='Failed to download VOD manifest' if errnote
is None else errnote
,
2381 fatal
=False, data
=data
, headers
=headers
, query
=query
)
2383 return self
._parse
_m
3u8_vod
_duration
(m3u8_vod
or '', video_id
)
2385 def _parse_m3u8_vod_duration(self
, m3u8_vod
, video_id
):
2386 if '#EXT-X-PLAYLIST-TYPE:VOD' not in m3u8_vod
:
2390 float(line
[len('#EXTINF:'):].split(',')[0])
2391 for line
in m3u8_vod
.splitlines() if line
.startswith('#EXTINF:'))) or None
2394 def _xpath_ns(path
, namespace
=None):
2398 for c
in path
.split('/'):
2399 if not c
or c
== '.':
2402 out
.append('{%s}%s' % (namespace
, c
))
2403 return '/'.join(out
)
2405 def _extract_smil_formats_and_subtitles(self
, smil_url
, video_id
, fatal
=True, f4m_params
=None, transform_source
=None):
2406 res
= self
._download
_smil
(smil_url
, video_id
, fatal
=fatal
, transform_source
=transform_source
)
2412 smil_url
= urlh
.geturl()
2414 namespace
= self
._parse
_smil
_namespace
(smil
)
2416 fmts
= self
._parse
_smil
_formats
(
2417 smil
, smil_url
, video_id
, namespace
=namespace
, f4m_params
=f4m_params
)
2418 subs
= self
._parse
_smil
_subtitles
(
2419 smil
, namespace
=namespace
)
2423 def _extract_smil_formats(self
, *args
, **kwargs
):
2424 fmts
, subs
= self
._extract
_smil
_formats
_and
_subtitles
(*args
, **kwargs
)
2426 self
._report
_ignoring
_subs
('SMIL')
2429 def _extract_smil_info(self
, smil_url
, video_id
, fatal
=True, f4m_params
=None):
2430 res
= self
._download
_smil
(smil_url
, video_id
, fatal
=fatal
)
2435 smil_url
= urlh
.geturl()
2437 return self
._parse
_smil
(smil
, smil_url
, video_id
, f4m_params
=f4m_params
)
2439 def _download_smil(self
, smil_url
, video_id
, fatal
=True, transform_source
=None):
2440 return self
._download
_xml
_handle
(
2441 smil_url
, video_id
, 'Downloading SMIL file',
2442 'Unable to download SMIL file', fatal
=fatal
, transform_source
=transform_source
)
2444 def _parse_smil(self
, smil
, smil_url
, video_id
, f4m_params
=None):
2445 namespace
= self
._parse
_smil
_namespace
(smil
)
2447 formats
= self
._parse
_smil
_formats
(
2448 smil
, smil_url
, video_id
, namespace
=namespace
, f4m_params
=f4m_params
)
2449 subtitles
= self
._parse
_smil
_subtitles
(smil
, namespace
=namespace
)
2451 video_id
= os
.path
.splitext(url_basename(smil_url
))[0]
2455 for meta
in smil
.findall(self
._xpath
_ns
('./head/meta', namespace
)):
2456 name
= meta
.attrib
.get('name')
2457 content
= meta
.attrib
.get('content')
2458 if not name
or not content
:
2460 if not title
and name
== 'title':
2462 elif not description
and name
in ('description', 'abstract'):
2463 description
= content
2464 elif not upload_date
and name
== 'date':
2465 upload_date
= unified_strdate(content
)
2468 'id': image
.get('type'),
2469 'url': image
.get('src'),
2470 'width': int_or_none(image
.get('width')),
2471 'height': int_or_none(image
.get('height')),
2472 } for image
in smil
.findall(self
._xpath
_ns
('.//image', namespace
)) if image
.get('src')]
2476 'title': title
or video_id
,
2477 'description': description
,
2478 'upload_date': upload_date
,
2479 'thumbnails': thumbnails
,
2481 'subtitles': subtitles
,
2484 def _parse_smil_namespace(self
, smil
):
2485 return self
._search
_regex
(
2486 r
'(?i)^{([^}]+)?}smil$', smil
.tag
, 'namespace', default
=None)
2488 def _parse_smil_formats(self
, smil
, smil_url
, video_id
, namespace
=None, f4m_params
=None, transform_rtmp_url
=None):
2490 for meta
in smil
.findall(self
._xpath
_ns
('./head/meta', namespace
)):
2491 b
= meta
.get('base') or meta
.get('httpBase')
2503 media
= smil
.findall(self
._xpath
_ns
('.//video', namespace
)) + smil
.findall(self
._xpath
_ns
('.//audio', namespace
))
2504 for medium
in media
:
2505 src
= medium
.get('src')
2506 if not src
or src
in srcs
:
2510 bitrate
= float_or_none(medium
.get('system-bitrate') or medium
.get('systemBitrate'), 1000)
2511 filesize
= int_or_none(medium
.get('size') or medium
.get('fileSize'))
2512 width
= int_or_none(medium
.get('width'))
2513 height
= int_or_none(medium
.get('height'))
2514 proto
= medium
.get('proto')
2515 ext
= medium
.get('ext')
2516 src_ext
= determine_ext(src
)
2517 streamer
= medium
.get('streamer') or base
2519 if proto
== 'rtmp' or streamer
.startswith('rtmp'):
2525 'format_id': 'rtmp-%d' % (rtmp_count
if bitrate
is None else bitrate
),
2527 'filesize': filesize
,
2531 if transform_rtmp_url
:
2532 streamer
, src
= transform_rtmp_url(streamer
, src
)
2533 formats
[-1].update({
2539 src_url
= src
if src
.startswith('http') else compat_urlparse
.urljoin(base
, src
)
2540 src_url
= src_url
.strip()
2542 if proto
== 'm3u8' or src_ext
== 'm3u8':
2543 m3u8_formats
= self
._extract
_m
3u8_formats
(
2544 src_url
, video_id
, ext
or 'mp4', m3u8_id
='hls', fatal
=False)
2545 if len(m3u8_formats
) == 1:
2547 m3u8_formats
[0].update({
2548 'format_id': 'hls-%d' % (m3u8_count
if bitrate
is None else bitrate
),
2553 formats
.extend(m3u8_formats
)
2554 elif src_ext
== 'f4m':
2559 'plugin': 'flowplayer-3.2.0.1',
2561 f4m_url
+= '&' if '?' in f4m_url
else '?'
2562 f4m_url
+= compat_urllib_parse_urlencode(f4m_params
)
2563 formats
.extend(self
._extract
_f
4m
_formats
(f4m_url
, video_id
, f4m_id
='hds', fatal
=False))
2564 elif src_ext
== 'mpd':
2565 formats
.extend(self
._extract
_mpd
_formats
(
2566 src_url
, video_id
, mpd_id
='dash', fatal
=False))
2567 elif re
.search(r
'\.ism/[Mm]anifest', src_url
):
2568 formats
.extend(self
._extract
_ism
_formats
(
2569 src_url
, video_id
, ism_id
='mss', fatal
=False))
2570 elif src_url
.startswith('http') and self
._is
_valid
_url
(src
, video_id
):
2574 'ext': ext
or src_ext
or 'flv',
2575 'format_id': 'http-%d' % (bitrate
or http_count
),
2577 'filesize': filesize
,
2582 for medium
in smil
.findall(self
._xpath
_ns
('.//imagestream', namespace
)):
2583 src
= medium
.get('src')
2584 if not src
or src
in srcs
:
2590 'format_id': 'imagestream-%d' % (imgs_count
),
2592 'ext': mimetype2ext(medium
.get('type')),
2595 'width': int_or_none(medium
.get('width')),
2596 'height': int_or_none(medium
.get('height')),
2597 'format_note': 'SMIL storyboards',
2602 def _parse_smil_subtitles(self
, smil
, namespace
=None, subtitles_lang
='en'):
2605 for num
, textstream
in enumerate(smil
.findall(self
._xpath
_ns
('.//textstream', namespace
))):
2606 src
= textstream
.get('src')
2607 if not src
or src
in urls
:
2610 ext
= textstream
.get('ext') or mimetype2ext(textstream
.get('type')) or determine_ext(src
)
2611 lang
= textstream
.get('systemLanguage') or textstream
.get('systemLanguageName') or textstream
.get('lang') or subtitles_lang
2612 subtitles
.setdefault(lang
, []).append({
2618 def _extract_xspf_playlist(self
, xspf_url
, playlist_id
, fatal
=True):
2619 res
= self
._download
_xml
_handle
(
2620 xspf_url
, playlist_id
, 'Downloading xpsf playlist',
2621 'Unable to download xspf manifest', fatal
=fatal
)
2626 xspf_url
= urlh
.geturl()
2628 return self
._parse
_xspf
(
2629 xspf
, playlist_id
, xspf_url
=xspf_url
,
2630 xspf_base_url
=base_url(xspf_url
))
2632 def _parse_xspf(self
, xspf_doc
, playlist_id
, xspf_url
=None, xspf_base_url
=None):
2634 'xspf': 'http://xspf.org/ns/0/',
2635 's1': 'http://static.streamone.nl/player/ns/0',
2639 for track
in xspf_doc
.findall(xpath_with_ns('./xspf:trackList/xspf:track', NS_MAP
)):
2641 track
, xpath_with_ns('./xspf:title', NS_MAP
), 'title', default
=playlist_id
)
2642 description
= xpath_text(
2643 track
, xpath_with_ns('./xspf:annotation', NS_MAP
), 'description')
2644 thumbnail
= xpath_text(
2645 track
, xpath_with_ns('./xspf:image', NS_MAP
), 'thumbnail')
2646 duration
= float_or_none(
2647 xpath_text(track
, xpath_with_ns('./xspf:duration', NS_MAP
), 'duration'), 1000)
2650 for location
in track
.findall(xpath_with_ns('./xspf:location', NS_MAP
)):
2651 format_url
= urljoin(xspf_base_url
, location
.text
)
2656 'manifest_url': xspf_url
,
2657 'format_id': location
.get(xpath_with_ns('s1:label', NS_MAP
)),
2658 'width': int_or_none(location
.get(xpath_with_ns('s1:width', NS_MAP
))),
2659 'height': int_or_none(location
.get(xpath_with_ns('s1:height', NS_MAP
))),
2661 self
._sort
_formats
(formats
)
2666 'description': description
,
2667 'thumbnail': thumbnail
,
2668 'duration': duration
,
2673 def _extract_mpd_formats(self
, *args
, **kwargs
):
2674 fmts
, subs
= self
._extract
_mpd
_formats
_and
_subtitles
(*args
, **kwargs
)
2676 self
._report
_ignoring
_subs
('DASH')
2679 def _extract_mpd_formats_and_subtitles(
2680 self
, mpd_url
, video_id
, mpd_id
=None, note
=None, errnote
=None,
2681 fatal
=True, data
=None, headers
={}, query={}
):
2682 res
= self
._download
_xml
_handle
(
2684 note
='Downloading MPD manifest' if note
is None else note
,
2685 errnote
='Failed to download MPD manifest' if errnote
is None else errnote
,
2686 fatal
=fatal
, data
=data
, headers
=headers
, query
=query
)
2693 # We could have been redirected to a new url when we retrieved our mpd file.
2694 mpd_url
= urlh
.geturl()
2695 mpd_base_url
= base_url(mpd_url
)
2697 return self
._parse
_mpd
_formats
_and
_subtitles
(
2698 mpd_doc
, mpd_id
, mpd_base_url
, mpd_url
)
2700 def _parse_mpd_formats(self
, *args
, **kwargs
):
2701 fmts
, subs
= self
._parse
_mpd
_formats
_and
_subtitles
(*args
, **kwargs
)
2703 self
._report
_ignoring
_subs
('DASH')
2706 def _parse_mpd_formats_and_subtitles(
2707 self
, mpd_doc
, mpd_id
=None, mpd_base_url
='', mpd_url
=None):
2709 Parse formats from MPD manifest.
2711 1. MPEG-DASH Standard, ISO/IEC 23009-1:2014(E),
2712 http://standards.iso.org/ittf/PubliclyAvailableStandards/c065274_ISO_IEC_23009-1_2014.zip
2713 2. https://en.wikipedia.org/wiki/Dynamic_Adaptive_Streaming_over_HTTP
2715 if not self
.get_param('dynamic_mpd', True):
2716 if mpd_doc
.get('type') == 'dynamic':
2719 namespace
= self
._search
_regex
(r
'(?i)^{([^}]+)?}MPD$', mpd_doc
.tag
, 'namespace', default
=None)
2722 return self
._xpath
_ns
(path
, namespace
)
2724 def is_drm_protected(element
):
2725 return element
.find(_add_ns('ContentProtection')) is not None
2727 def extract_multisegment_info(element
, ms_parent_info
):
2728 ms_info
= ms_parent_info
.copy()
2730 # As per [1, 5.3.9.2.2] SegmentList and SegmentTemplate share some
2731 # common attributes and elements. We will only extract relevant
2733 def extract_common(source
):
2734 segment_timeline
= source
.find(_add_ns('SegmentTimeline'))
2735 if segment_timeline
is not None:
2736 s_e
= segment_timeline
.findall(_add_ns('S'))
2738 ms_info
['total_number'] = 0
2741 r
= int(s
.get('r', 0))
2742 ms_info
['total_number'] += 1 + r
2743 ms_info
['s'].append({
2744 't': int(s
.get('t', 0)),
2745 # @d is mandatory (see [1, 5.3.9.6.2, Table 17, page 60])
2746 'd': int(s
.attrib
['d']),
2749 start_number
= source
.get('startNumber')
2751 ms_info
['start_number'] = int(start_number
)
2752 timescale
= source
.get('timescale')
2754 ms_info
['timescale'] = int(timescale
)
2755 segment_duration
= source
.get('duration')
2756 if segment_duration
:
2757 ms_info
['segment_duration'] = float(segment_duration
)
2759 def extract_Initialization(source
):
2760 initialization
= source
.find(_add_ns('Initialization'))
2761 if initialization
is not None:
2762 ms_info
['initialization_url'] = initialization
.attrib
['sourceURL']
2764 segment_list
= element
.find(_add_ns('SegmentList'))
2765 if segment_list
is not None:
2766 extract_common(segment_list
)
2767 extract_Initialization(segment_list
)
2768 segment_urls_e
= segment_list
.findall(_add_ns('SegmentURL'))
2770 ms_info
['segment_urls'] = [segment
.attrib
['media'] for segment
in segment_urls_e
]
2772 segment_template
= element
.find(_add_ns('SegmentTemplate'))
2773 if segment_template
is not None:
2774 extract_common(segment_template
)
2775 media
= segment_template
.get('media')
2777 ms_info
['media'] = media
2778 initialization
= segment_template
.get('initialization')
2780 ms_info
['initialization'] = initialization
2782 extract_Initialization(segment_template
)
2785 mpd_duration
= parse_duration(mpd_doc
.get('mediaPresentationDuration'))
2786 formats
, subtitles
= [], {}
2787 stream_numbers
= collections
.defaultdict(int)
2788 for period
in mpd_doc
.findall(_add_ns('Period')):
2789 period_duration
= parse_duration(period
.get('duration')) or mpd_duration
2790 period_ms_info
= extract_multisegment_info(period
, {
2794 for adaptation_set
in period
.findall(_add_ns('AdaptationSet')):
2795 adaption_set_ms_info
= extract_multisegment_info(adaptation_set
, period_ms_info
)
2796 for representation
in adaptation_set
.findall(_add_ns('Representation')):
2797 representation_attrib
= adaptation_set
.attrib
.copy()
2798 representation_attrib
.update(representation
.attrib
)
2799 # According to [1, 5.3.7.2, Table 9, page 41], @mimeType is mandatory
2800 mime_type
= representation_attrib
['mimeType']
2801 content_type
= representation_attrib
.get('contentType', mime_type
.split('/')[0])
2803 codecs
= parse_codecs(representation_attrib
.get('codecs', ''))
2804 if content_type
not in ('video', 'audio', 'text'):
2805 if mime_type
== 'image/jpeg':
2806 content_type
= mime_type
2807 elif codecs
['vcodec'] != 'none':
2808 content_type
= 'video'
2809 elif codecs
['acodec'] != 'none':
2810 content_type
= 'audio'
2811 elif codecs
.get('tcodec', 'none') != 'none':
2812 content_type
= 'text'
2813 elif mimetype2ext(mime_type
) in ('tt', 'dfxp', 'ttml', 'xml', 'json'):
2814 content_type
= 'text'
2816 self
.report_warning('Unknown MIME type %s in DASH manifest' % mime_type
)
2820 for element
in (representation
, adaptation_set
, period
, mpd_doc
):
2821 base_url_e
= element
.find(_add_ns('BaseURL'))
2822 if base_url_e
is not None:
2823 base_url
= base_url_e
.text
+ base_url
2824 if re
.match(r
'^https?://', base_url
):
2826 if mpd_base_url
and base_url
.startswith('/'):
2827 base_url
= compat_urlparse
.urljoin(mpd_base_url
, base_url
)
2828 elif mpd_base_url
and not re
.match(r
'^https?://', base_url
):
2829 if not mpd_base_url
.endswith('/'):
2831 base_url
= mpd_base_url
+ base_url
2832 representation_id
= representation_attrib
.get('id')
2833 lang
= representation_attrib
.get('lang')
2834 url_el
= representation
.find(_add_ns('BaseURL'))
2835 filesize
= int_or_none(url_el
.attrib
.get('{http://youtube.com/yt/2012/10/10}contentLength') if url_el
is not None else None)
2836 bandwidth
= int_or_none(representation_attrib
.get('bandwidth'))
2837 if representation_id
is not None:
2838 format_id
= representation_id
2840 format_id
= content_type
2842 format_id
= mpd_id
+ '-' + format_id
2843 if content_type
in ('video', 'audio'):
2845 'format_id': format_id
,
2846 'manifest_url': mpd_url
,
2847 'ext': mimetype2ext(mime_type
),
2848 'width': int_or_none(representation_attrib
.get('width')),
2849 'height': int_or_none(representation_attrib
.get('height')),
2850 'tbr': float_or_none(bandwidth
, 1000),
2851 'asr': int_or_none(representation_attrib
.get('audioSamplingRate')),
2852 'fps': int_or_none(representation_attrib
.get('frameRate')),
2853 'language': lang
if lang
not in ('mul', 'und', 'zxx', 'mis') else None,
2854 'format_note': 'DASH %s' % content_type
,
2855 'filesize': filesize
,
2856 'container': mimetype2ext(mime_type
) + '_dash',
2859 elif content_type
== 'text':
2861 'ext': mimetype2ext(mime_type
),
2862 'manifest_url': mpd_url
,
2863 'filesize': filesize
,
2865 elif content_type
== 'image/jpeg':
2866 # See test case in VikiIE
2867 # https://www.viki.com/videos/1175236v-choosing-spouse-by-lottery-episode-1
2869 'format_id': format_id
,
2871 'manifest_url': mpd_url
,
2872 'format_note': 'DASH storyboards (jpeg)',
2876 if is_drm_protected(adaptation_set
) or is_drm_protected(representation
):
2878 representation_ms_info
= extract_multisegment_info(representation
, adaption_set_ms_info
)
2880 def prepare_template(template_name
, identifiers
):
2881 tmpl
= representation_ms_info
[template_name
]
2882 # First of, % characters outside $...$ templates
2883 # must be escaped by doubling for proper processing
2884 # by % operator string formatting used further (see
2885 # https://github.com/ytdl-org/youtube-dl/issues/16867).
2891 in_template
= not in_template
2892 elif c
== '%' and not in_template
:
2894 # Next, $...$ templates are translated to their
2895 # %(...) counterparts to be used with % operator
2896 if representation_id
is not None:
2897 t
= t
.replace('$RepresentationID$', representation_id
)
2898 t
= re
.sub(r
'\$(%s)\$' % '|'.join(identifiers
), r
'%(\1)d', t
)
2899 t
= re
.sub(r
'\$(%s)%%([^$]+)\$' % '|'.join(identifiers
), r
'%(\1)\2', t
)
2900 t
.replace('$$', '$')
2903 # @initialization is a regular template like @media one
2904 # so it should be handled just the same way (see
2905 # https://github.com/ytdl-org/youtube-dl/issues/11605)
2906 if 'initialization' in representation_ms_info
:
2907 initialization_template
= prepare_template(
2909 # As per [1, 5.3.9.4.2, Table 15, page 54] $Number$ and
2910 # $Time$ shall not be included for @initialization thus
2911 # only $Bandwidth$ remains
2913 representation_ms_info
['initialization_url'] = initialization_template
% {
2914 'Bandwidth': bandwidth
,
2917 def location_key(location
):
2918 return 'url' if re
.match(r
'^https?://', location
) else 'path'
2920 if 'segment_urls' not in representation_ms_info
and 'media' in representation_ms_info
:
2922 media_template
= prepare_template('media', ('Number', 'Bandwidth', 'Time'))
2923 media_location_key
= location_key(media_template
)
2925 # As per [1, 5.3.9.4.4, Table 16, page 55] $Number$ and $Time$
2926 # can't be used at the same time
2927 if '%(Number' in media_template
and 's' not in representation_ms_info
:
2928 segment_duration
= None
2929 if 'total_number' not in representation_ms_info
and 'segment_duration' in representation_ms_info
:
2930 segment_duration
= float_or_none(representation_ms_info
['segment_duration'], representation_ms_info
['timescale'])
2931 representation_ms_info
['total_number'] = int(math
.ceil(
2932 float_or_none(period_duration
, segment_duration
, default
=0)))
2933 representation_ms_info
['fragments'] = [{
2934 media_location_key
: media_template
% {
2935 'Number': segment_number
,
2936 'Bandwidth': bandwidth
,
2938 'duration': segment_duration
,
2939 } for segment_number
in range(
2940 representation_ms_info
['start_number'],
2941 representation_ms_info
['total_number'] + representation_ms_info
['start_number'])]
2943 # $Number*$ or $Time$ in media template with S list available
2944 # Example $Number*$: http://www.svtplay.se/klipp/9023742/stopptid-om-bjorn-borg
2945 # Example $Time$: https://play.arkena.com/embed/avp/v2/player/media/b41dda37-d8e7-4d3f-b1b5-9a9db578bdfe/1/129411
2946 representation_ms_info
['fragments'] = []
2949 segment_number
= representation_ms_info
['start_number']
2951 def add_segment_url():
2952 segment_url
= media_template
% {
2953 'Time': segment_time
,
2954 'Bandwidth': bandwidth
,
2955 'Number': segment_number
,
2957 representation_ms_info
['fragments'].append({
2958 media_location_key
: segment_url
,
2959 'duration': float_or_none(segment_d
, representation_ms_info
['timescale']),
2962 for num
, s
in enumerate(representation_ms_info
['s']):
2963 segment_time
= s
.get('t') or segment_time
2967 for r
in range(s
.get('r', 0)):
2968 segment_time
+= segment_d
2971 segment_time
+= segment_d
2972 elif 'segment_urls' in representation_ms_info
and 's' in representation_ms_info
:
2974 # Example: https://www.youtube.com/watch?v=iXZV5uAYMJI
2975 # or any YouTube dashsegments video
2978 timescale
= representation_ms_info
['timescale']
2979 for s
in representation_ms_info
['s']:
2980 duration
= float_or_none(s
['d'], timescale
)
2981 for r
in range(s
.get('r', 0) + 1):
2982 segment_uri
= representation_ms_info
['segment_urls'][segment_index
]
2984 location_key(segment_uri
): segment_uri
,
2985 'duration': duration
,
2988 representation_ms_info
['fragments'] = fragments
2989 elif 'segment_urls' in representation_ms_info
:
2990 # Segment URLs with no SegmentTimeline
2991 # Example: https://www.seznam.cz/zpravy/clanek/cesko-zasahne-vitr-o-sile-vichrice-muze-byt-i-zivotu-nebezpecny-39091
2992 # https://github.com/ytdl-org/youtube-dl/pull/14844
2994 segment_duration
= float_or_none(
2995 representation_ms_info
['segment_duration'],
2996 representation_ms_info
['timescale']) if 'segment_duration' in representation_ms_info
else None
2997 for segment_url
in representation_ms_info
['segment_urls']:
2999 location_key(segment_url
): segment_url
,
3001 if segment_duration
:
3002 fragment
['duration'] = segment_duration
3003 fragments
.append(fragment
)
3004 representation_ms_info
['fragments'] = fragments
3005 # If there is a fragments key available then we correctly recognized fragmented media.
3006 # Otherwise we will assume unfragmented media with direct access. Technically, such
3007 # assumption is not necessarily correct since we may simply have no support for
3008 # some forms of fragmented media renditions yet, but for now we'll use this fallback.
3009 if 'fragments' in representation_ms_info
:
3011 # NB: mpd_url may be empty when MPD manifest is parsed from a string
3012 'url': mpd_url
or base_url
,
3013 'fragment_base_url': base_url
,
3015 'protocol': 'http_dash_segments' if mime_type
!= 'image/jpeg' else 'mhtml',
3017 if 'initialization_url' in representation_ms_info
:
3018 initialization_url
= representation_ms_info
['initialization_url']
3019 if not f
.get('url'):
3020 f
['url'] = initialization_url
3021 f
['fragments'].append({location_key(initialization_url): initialization_url}
)
3022 f
['fragments'].extend(representation_ms_info
['fragments'])
3023 if not period_duration
:
3024 period_duration
= try_get(
3025 representation_ms_info
,
3026 lambda r
: sum(frag
['duration'] for frag
in r
['fragments']), float)
3028 # Assuming direct URL to unfragmented media.
3030 if content_type
in ('video', 'audio', 'image/jpeg'):
3031 f
['manifest_stream_number'] = stream_numbers
[f
['url']]
3032 stream_numbers
[f
['url']] += 1
3034 elif content_type
== 'text':
3035 subtitles
.setdefault(lang
or 'und', []).append(f
)
3037 return formats
, subtitles
3039 def _extract_ism_formats(self
, *args
, **kwargs
):
3040 fmts
, subs
= self
._extract
_ism
_formats
_and
_subtitles
(*args
, **kwargs
)
3042 self
._report
_ignoring
_subs
('ISM')
3045 def _extract_ism_formats_and_subtitles(self
, ism_url
, video_id
, ism_id
=None, note
=None, errnote
=None, fatal
=True, data
=None, headers
={}, query={}
):
3046 res
= self
._download
_xml
_handle
(
3048 note
='Downloading ISM manifest' if note
is None else note
,
3049 errnote
='Failed to download ISM manifest' if errnote
is None else errnote
,
3050 fatal
=fatal
, data
=data
, headers
=headers
, query
=query
)
3057 return self
._parse
_ism
_formats
_and
_subtitles
(ism_doc
, urlh
.geturl(), ism_id
)
3059 def _parse_ism_formats_and_subtitles(self
, ism_doc
, ism_url
, ism_id
=None):
3061 Parse formats from ISM manifest.
3063 1. [MS-SSTR]: Smooth Streaming Protocol,
3064 https://msdn.microsoft.com/en-us/library/ff469518.aspx
3066 if ism_doc
.get('IsLive') == 'TRUE':
3069 duration
= int(ism_doc
.attrib
['Duration'])
3070 timescale
= int_or_none(ism_doc
.get('TimeScale')) or 10000000
3074 for stream
in ism_doc
.findall('StreamIndex'):
3075 stream_type
= stream
.get('Type')
3076 if stream_type
not in ('video', 'audio', 'text'):
3078 url_pattern
= stream
.attrib
['Url']
3079 stream_timescale
= int_or_none(stream
.get('TimeScale')) or timescale
3080 stream_name
= stream
.get('Name')
3081 stream_language
= stream
.get('Language', 'und')
3082 for track
in stream
.findall('QualityLevel'):
3083 fourcc
= track
.get('FourCC') or ('AACL' if track
.get('AudioTag') == '255' else None)
3084 # TODO: add support for WVC1 and WMAP
3085 if fourcc
not in ('H264', 'AVC1', 'AACL', 'TTML'):
3086 self
.report_warning('%s is not a supported codec' % fourcc
)
3088 tbr
= int(track
.attrib
['Bitrate']) // 1000
3089 # [1] does not mention Width and Height attributes. However,
3090 # they're often present while MaxWidth and MaxHeight are
3091 # missing, so should be used as fallbacks
3092 width
= int_or_none(track
.get('MaxWidth') or track
.get('Width'))
3093 height
= int_or_none(track
.get('MaxHeight') or track
.get('Height'))
3094 sampling_rate
= int_or_none(track
.get('SamplingRate'))
3096 track_url_pattern
= re
.sub(r
'{[Bb]itrate}', track
.attrib
['Bitrate'], url_pattern
)
3097 track_url_pattern
= compat_urlparse
.urljoin(ism_url
, track_url_pattern
)
3103 stream_fragments
= stream
.findall('c')
3104 for stream_fragment_index
, stream_fragment
in enumerate(stream_fragments
):
3105 fragment_ctx
['time'] = int_or_none(stream_fragment
.get('t')) or fragment_ctx
['time']
3106 fragment_repeat
= int_or_none(stream_fragment
.get('r')) or 1
3107 fragment_ctx
['duration'] = int_or_none(stream_fragment
.get('d'))
3108 if not fragment_ctx
['duration']:
3110 next_fragment_time
= int(stream_fragment
[stream_fragment_index
+ 1].attrib
['t'])
3112 next_fragment_time
= duration
3113 fragment_ctx
['duration'] = (next_fragment_time
- fragment_ctx
['time']) / fragment_repeat
3114 for _
in range(fragment_repeat
):
3116 'url': re
.sub(r
'{start[ _]time}', compat_str(fragment_ctx
['time']), track_url_pattern
),
3117 'duration': fragment_ctx
['duration'] / stream_timescale
,
3119 fragment_ctx
['time'] += fragment_ctx
['duration']
3121 if stream_type
== 'text':
3122 subtitles
.setdefault(stream_language
, []).append({
3126 'manifest_url': ism_url
,
3127 'fragments': fragments
,
3128 '_download_params': {
3129 'stream_type': stream_type
,
3130 'duration': duration
,
3131 'timescale': stream_timescale
,
3133 'language': stream_language
,
3134 'codec_private_data': track
.get('CodecPrivateData'),
3137 elif stream_type
in ('video', 'audio'):
3139 'format_id': join_nonempty(ism_id
, stream_name
, tbr
),
3141 'manifest_url': ism_url
,
3142 'ext': 'ismv' if stream_type
== 'video' else 'isma',
3146 'asr': sampling_rate
,
3147 'vcodec': 'none' if stream_type
== 'audio' else fourcc
,
3148 'acodec': 'none' if stream_type
== 'video' else fourcc
,
3150 'fragments': fragments
,
3151 'has_drm': ism_doc
.find('Protection') is not None,
3152 '_download_params': {
3153 'stream_type': stream_type
,
3154 'duration': duration
,
3155 'timescale': stream_timescale
,
3156 'width': width
or 0,
3157 'height': height
or 0,
3159 'language': stream_language
,
3160 'codec_private_data': track
.get('CodecPrivateData'),
3161 'sampling_rate': sampling_rate
,
3162 'channels': int_or_none(track
.get('Channels', 2)),
3163 'bits_per_sample': int_or_none(track
.get('BitsPerSample', 16)),
3164 'nal_unit_length_field': int_or_none(track
.get('NALUnitLengthField', 4)),
3167 return formats
, subtitles
3169 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):
3170 def absolute_url(item_url
):
3171 return urljoin(base_url
, item_url
)
3173 def parse_content_type(content_type
):
3174 if not content_type
:
3176 ctr
= re
.search(r
'(?P<mimetype>[^/]+/[^;]+)(?:;\s*codecs="?(?P<codecs>[^"]+))?', content_type
)
3178 mimetype
, codecs
= ctr
.groups()
3179 f
= parse_codecs(codecs
)
3180 f
['ext'] = mimetype2ext(mimetype
)
3184 def _media_formats(src
, cur_media_type
, type_info
={}):
3185 full_url
= absolute_url(src
)
3186 ext
= type_info
.get('ext') or determine_ext(full_url
)
3188 is_plain_url
= False
3189 formats
= self
._extract
_m
3u8_formats
(
3190 full_url
, video_id
, ext
='mp4',
3191 entry_protocol
=m3u8_entry_protocol
, m3u8_id
=m3u8_id
,
3192 preference
=preference
, quality
=quality
, fatal
=False)
3194 is_plain_url
= False
3195 formats
= self
._extract
_mpd
_formats
(
3196 full_url
, video_id
, mpd_id
=mpd_id
, fatal
=False)
3201 'vcodec': 'none' if cur_media_type
== 'audio' else None,
3203 return is_plain_url
, formats
3206 # amp-video and amp-audio are very similar to their HTML5 counterparts
3207 # so we wll include them right here (see
3208 # https://www.ampproject.org/docs/reference/components/amp-video)
3209 # For dl8-* tags see https://delight-vr.com/documentation/dl8-video/
3210 _MEDIA_TAG_NAME_RE
= r
'(?:(?:amp|dl8(?:-live)?)-)?(video|audio)'
3211 media_tags
= [(media_tag
, media_tag_name
, media_type
, '')
3212 for media_tag
, media_tag_name
, media_type
3213 in re
.findall(r
'(?s)(<(%s)[^>]*/>)' % _MEDIA_TAG_NAME_RE
, webpage
)]
3214 media_tags
.extend(re
.findall(
3215 # We only allow video|audio followed by a whitespace or '>'.
3216 # Allowing more characters may end up in significant slow down (see
3217 # https://github.com/ytdl-org/youtube-dl/issues/11979, example URL:
3218 # http://www.porntrex.com/maps/videositemap.xml).
3219 r
'(?s)(<(?P<tag>%s)(?:\s+[^>]*)?>)(.*?)</(?P=tag)>' % _MEDIA_TAG_NAME_RE
, webpage
))
3220 for media_tag
, _
, media_type
, media_content
in media_tags
:
3225 media_attributes
= extract_attributes(media_tag
)
3226 src
= strip_or_none(media_attributes
.get('src'))
3228 _
, formats
= _media_formats(src
, media_type
)
3229 media_info
['formats'].extend(formats
)
3230 media_info
['thumbnail'] = absolute_url(media_attributes
.get('poster'))
3232 for source_tag
in re
.findall(r
'<source[^>]+>', media_content
):
3233 s_attr
= extract_attributes(source_tag
)
3234 # data-video-src and data-src are non standard but seen
3235 # several times in the wild
3236 src
= strip_or_none(dict_get(s_attr
, ('src', 'data-video-src', 'data-src')))
3239 f
= parse_content_type(s_attr
.get('type'))
3240 is_plain_url
, formats
= _media_formats(src
, media_type
, f
)
3242 # width, height, res, label and title attributes are
3243 # all not standard but seen several times in the wild
3246 for lbl
in ('label', 'title')
3247 if str_or_none(s_attr
.get(lbl
))
3249 width
= int_or_none(s_attr
.get('width'))
3250 height
= (int_or_none(s_attr
.get('height'))
3251 or int_or_none(s_attr
.get('res')))
3252 if not width
or not height
:
3254 resolution
= parse_resolution(lbl
)
3257 width
= width
or resolution
.get('width')
3258 height
= height
or resolution
.get('height')
3260 tbr
= parse_bitrate(lbl
)
3269 'format_id': s_attr
.get('label') or s_attr
.get('title'),
3271 f
.update(formats
[0])
3272 media_info
['formats'].append(f
)
3274 media_info
['formats'].extend(formats
)
3275 for track_tag
in re
.findall(r
'<track[^>]+>', media_content
):
3276 track_attributes
= extract_attributes(track_tag
)
3277 kind
= track_attributes
.get('kind')
3278 if not kind
or kind
in ('subtitles', 'captions'):
3279 src
= strip_or_none(track_attributes
.get('src'))
3282 lang
= track_attributes
.get('srclang') or track_attributes
.get('lang') or track_attributes
.get('label')
3283 media_info
['subtitles'].setdefault(lang
, []).append({
3284 'url': absolute_url(src
),
3286 for f
in media_info
['formats']:
3287 f
.setdefault('http_headers', {})['Referer'] = base_url
3288 if media_info
['formats'] or media_info
['subtitles']:
3289 entries
.append(media_info
)
3292 def _extract_akamai_formats(self
, *args
, **kwargs
):
3293 fmts
, subs
= self
._extract
_akamai
_formats
_and
_subtitles
(*args
, **kwargs
)
3295 self
._report
_ignoring
_subs
('akamai')
3298 def _extract_akamai_formats_and_subtitles(self
, manifest_url
, video_id
, hosts
={}):
3299 signed
= 'hdnea=' in manifest_url
3301 # https://learn.akamai.com/en-us/webhelp/media-services-on-demand/stream-packaging-user-guide/GUID-BE6C0F73-1E06-483B-B0EA-57984B91B7F9.html
3302 manifest_url
= re
.sub(
3303 r
'(?:b=[\d,-]+|(?:__a__|attributes)=off|__b__=\d+)&?',
3304 '', manifest_url
).strip('?')
3309 hdcore_sign
= 'hdcore=3.7.0'
3310 f4m_url
= re
.sub(r
'(https?://[^/]+)/i/', r
'\1/z/', manifest_url
).replace('/master.m3u8', '/manifest.f4m')
3311 hds_host
= hosts
.get('hds')
3313 f4m_url
= re
.sub(r
'(https?://)[^/]+', r
'\1' + hds_host
, f4m_url
)
3314 if 'hdcore=' not in f4m_url
:
3315 f4m_url
+= ('&' if '?' in f4m_url
else '?') + hdcore_sign
3316 f4m_formats
= self
._extract
_f
4m
_formats
(
3317 f4m_url
, video_id
, f4m_id
='hds', fatal
=False)
3318 for entry
in f4m_formats
:
3319 entry
.update({'extra_param_to_segment_url': hdcore_sign}
)
3320 formats
.extend(f4m_formats
)
3322 m3u8_url
= re
.sub(r
'(https?://[^/]+)/z/', r
'\1/i/', manifest_url
).replace('/manifest.f4m', '/master.m3u8')
3323 hls_host
= hosts
.get('hls')
3325 m3u8_url
= re
.sub(r
'(https?://)[^/]+', r
'\1' + hls_host
, m3u8_url
)
3326 m3u8_formats
, m3u8_subtitles
= self
._extract
_m
3u8_formats
_and
_subtitles
(
3327 m3u8_url
, video_id
, 'mp4', 'm3u8_native',
3328 m3u8_id
='hls', fatal
=False)
3329 formats
.extend(m3u8_formats
)
3330 subtitles
= self
._merge
_subtitles
(subtitles
, m3u8_subtitles
)
3332 http_host
= hosts
.get('http')
3333 if http_host
and m3u8_formats
and not signed
:
3334 REPL_REGEX
= r
'https?://[^/]+/i/([^,]+),([^/]+),([^/]+)\.csmil/.+'
3335 qualities
= re
.match(REPL_REGEX
, m3u8_url
).group(2).split(',')
3336 qualities_length
= len(qualities
)
3337 if len(m3u8_formats
) in (qualities_length
, qualities_length
+ 1):
3339 for f
in m3u8_formats
:
3340 if f
['vcodec'] != 'none':
3341 for protocol
in ('http', 'https'):
3343 del http_f
['manifest_url']
3345 REPL_REGEX
, protocol
+ fr
'://{http_host}/\g<1>{qualities[i]}\3', f
['url'])
3347 'format_id': http_f
['format_id'].replace('hls-', protocol
+ '-'),
3349 'protocol': protocol
,
3351 formats
.append(http_f
)
3354 return formats
, subtitles
3356 def _extract_wowza_formats(self
, url
, video_id
, m3u8_entry_protocol
='m3u8_native', skip_protocols
=[]):
3357 query
= compat_urlparse
.urlparse(url
).query
3358 url
= re
.sub(r
'/(?:manifest|playlist|jwplayer)\.(?:m3u8|f4m|mpd|smil)', '', url
)
3360 r
'(?:(?:http|rtmp|rtsp)(?P<s>s)?:)?(?P<url>//[^?]+)', url
)
3361 url_base
= mobj
.group('url')
3362 http_base_url
= '%s%s:%s' % ('http', mobj
.group('s') or '', url_base
)
3365 def manifest_url(manifest
):
3366 m_url
= f
'{http_base_url}/{manifest}'
3368 m_url
+= '?%s' % query
3371 if 'm3u8' not in skip_protocols
:
3372 formats
.extend(self
._extract
_m
3u8_formats
(
3373 manifest_url('playlist.m3u8'), video_id
, 'mp4',
3374 m3u8_entry_protocol
, m3u8_id
='hls', fatal
=False))
3375 if 'f4m' not in skip_protocols
:
3376 formats
.extend(self
._extract
_f
4m
_formats
(
3377 manifest_url('manifest.f4m'),
3378 video_id
, f4m_id
='hds', fatal
=False))
3379 if 'dash' not in skip_protocols
:
3380 formats
.extend(self
._extract
_mpd
_formats
(
3381 manifest_url('manifest.mpd'),
3382 video_id
, mpd_id
='dash', fatal
=False))
3383 if re
.search(r
'(?:/smil:|\.smil)', url_base
):
3384 if 'smil' not in skip_protocols
:
3385 rtmp_formats
= self
._extract
_smil
_formats
(
3386 manifest_url('jwplayer.smil'),
3387 video_id
, fatal
=False)
3388 for rtmp_format
in rtmp_formats
:
3389 rtsp_format
= rtmp_format
.copy()
3390 rtsp_format
['url'] = '%s/%s' % (rtmp_format
['url'], rtmp_format
['play_path'])
3391 del rtsp_format
['play_path']
3392 del rtsp_format
['ext']
3393 rtsp_format
.update({
3394 'url': rtsp_format
['url'].replace('rtmp://', 'rtsp://'),
3395 'format_id': rtmp_format
['format_id'].replace('rtmp', 'rtsp'),
3398 formats
.extend([rtmp_format
, rtsp_format
])
3400 for protocol
in ('rtmp', 'rtsp'):
3401 if protocol
not in skip_protocols
:
3403 'url': f
'{protocol}:{url_base}',
3404 'format_id': protocol
,
3405 'protocol': protocol
,
3409 def _find_jwplayer_data(self
, webpage
, video_id
=None, transform_source
=js_to_json
):
3411 r
'(?s)jwplayer\((?P<quote>[\'"])[^\'" ]+(?P
=quote
)\
)(?
!</script
>).*?\
.setup\s
*\
((?P
<options
>[^
)]+)\
)',
3415 jwplayer_data = self._parse_json(mobj.group('options
'),
3417 transform_source=transform_source)
3418 except ExtractorError:
3421 if isinstance(jwplayer_data, dict):
3422 return jwplayer_data
3424 def _extract_jwplayer_data(self, webpage, video_id, *args, **kwargs):
3425 jwplayer_data = self._find_jwplayer_data(
3426 webpage, video_id, transform_source=js_to_json)
3427 return self._parse_jwplayer_data(
3428 jwplayer_data, video_id, *args, **kwargs)
3430 def _parse_jwplayer_data(self, jwplayer_data, video_id=None, require_title=True,
3431 m3u8_id=None, mpd_id=None, rtmp_params=None, base_url=None):
3432 # JWPlayer backward compatibility: flattened playlists
3433 # https://github.com/jwplayer/jwplayer/blob/v7.4.3/src/js/api/config.js#L81-L96
3434 if 'playlist
' not in jwplayer_data:
3435 jwplayer_data = {'playlist': [jwplayer_data]}
3439 # JWPlayer backward compatibility: single playlist item
3440 # https://github.com/jwplayer/jwplayer/blob/v7.7.0/src/js/playlist/playlist.js#L10
3441 if not isinstance(jwplayer_data['playlist
'], list):
3442 jwplayer_data['playlist
'] = [jwplayer_data['playlist
']]
3444 for video_data in jwplayer_data['playlist
']:
3445 # JWPlayer backward compatibility: flattened sources
3446 # https://github.com/jwplayer/jwplayer/blob/v7.4.3/src/js/playlist/item.js#L29-L35
3447 if 'sources
' not in video_data:
3448 video_data['sources
'] = [video_data]
3450 this_video_id = video_id or video_data['mediaid
']
3452 formats = self._parse_jwplayer_formats(
3453 video_data['sources
'], video_id=this_video_id, m3u8_id=m3u8_id,
3454 mpd_id=mpd_id, rtmp_params=rtmp_params, base_url=base_url)
3457 tracks = video_data.get('tracks
')
3458 if tracks and isinstance(tracks, list):
3459 for track in tracks:
3460 if not isinstance(track, dict):
3462 track_kind = track.get('kind
')
3463 if not track_kind or not isinstance(track_kind, compat_str):
3465 if track_kind.lower() not in ('captions
', 'subtitles
'):
3467 track_url = urljoin(base_url, track.get('file'))
3470 subtitles.setdefault(track.get('label
') or 'en
', []).append({
3471 'url
': self._proto_relative_url(track_url)
3475 'id': this_video_id,
3476 'title
': unescapeHTML(video_data['title
'] if require_title else video_data.get('title
')),
3477 'description
': clean_html(video_data.get('description
')),
3478 'thumbnail
': urljoin(base_url, self._proto_relative_url(video_data.get('image
'))),
3479 'timestamp
': int_or_none(video_data.get('pubdate
')),
3480 'duration
': float_or_none(jwplayer_data.get('duration
') or video_data.get('duration
')),
3481 'subtitles
': subtitles,
3483 # https://github.com/jwplayer/jwplayer/blob/master/src/js/utils/validator.js#L32
3484 if len(formats) == 1 and re.search(r'^
(?
:http|
//).*(?
:youtube\
.com|youtu\
.be
)/.+', formats[0]['url
']):
3486 '_type
': 'url_transparent
',
3487 'url
': formats[0]['url
'],
3490 self._sort_formats(formats)
3491 entry['formats
'] = formats
3492 entries.append(entry)
3493 if len(entries) == 1:
3496 return self.playlist_result(entries)
3498 def _parse_jwplayer_formats(self, jwplayer_sources_data, video_id=None,
3499 m3u8_id=None, mpd_id=None, rtmp_params=None, base_url=None):
3502 for source in jwplayer_sources_data:
3503 if not isinstance(source, dict):
3505 source_url = urljoin(
3506 base_url, self._proto_relative_url(source.get('file')))
3507 if not source_url or source_url in urls:
3509 urls.append(source_url)
3510 source_type = source.get('type') or ''
3511 ext = mimetype2ext(source_type) or determine_ext(source_url)
3512 if source_type == 'hls
' or ext == 'm3u8
':
3513 formats.extend(self._extract_m3u8_formats(
3514 source_url, video_id, 'mp4
', entry_protocol='m3u8_native
',
3515 m3u8_id=m3u8_id, fatal=False))
3516 elif source_type == 'dash
' or ext == 'mpd
':
3517 formats.extend(self._extract_mpd_formats(
3518 source_url, video_id, mpd_id=mpd_id, fatal=False))
3520 formats.extend(self._extract_smil_formats(
3521 source_url, video_id, fatal=False))
3522 # https://github.com/jwplayer/jwplayer/blob/master/src/js/providers/default.js#L67
3523 elif source_type.startswith('audio
') or ext in (
3524 'oga
', 'aac
', 'mp3
', 'mpeg
', 'vorbis
'):
3531 height = int_or_none(source.get('height
'))
3533 # Often no height is provided but there is a label in
3534 # format like "1080p", "720p SD", or 1080.
3535 height = int_or_none(self._search_regex(
3536 r'^
(\d{3,4}
)[pP
]?
(?
:\b|$
)', compat_str(source.get('label
') or ''),
3537 'height
', default=None))
3540 'width
': int_or_none(source.get('width
')),
3542 'tbr
': int_or_none(source.get('bitrate
')),
3545 if source_url.startswith('rtmp
'):
3546 a_format['ext
'] = 'flv
'
3547 # See com/longtailvideo/jwplayer/media/RTMPMediaProvider.as
3548 # of jwplayer.flash.swf
3549 rtmp_url_parts = re.split(
3550 r'((?
:mp4|mp3|flv
):)', source_url, 1)
3551 if len(rtmp_url_parts) == 3:
3552 rtmp_url, prefix, play_path = rtmp_url_parts
3555 'play_path
': prefix + play_path,
3558 a_format.update(rtmp_params)
3559 formats.append(a_format)
3562 def _live_title(self, name):
3563 self._downloader.deprecation_warning('yt_dlp
.InfoExtractor
._live
_title
is deprecated
and does
not work
as expected
')
3566 def _int(self, v, name, fatal=False, **kwargs):
3567 res = int_or_none(v, **kwargs)
3569 msg = f'Failed to extract {name}
: Could
not parse value {v!r}
'
3571 raise ExtractorError(msg)
3573 self.report_warning(msg)
3576 def _float(self, v, name, fatal=False, **kwargs):
3577 res = float_or_none(v, **kwargs)
3579 msg = f'Failed to extract {name}
: Could
not parse value {v!r}
'
3581 raise ExtractorError(msg)
3583 self.report_warning(msg)
3586 def _set_cookie(self, domain, name, value, expire_time=None, port=None,
3587 path='/', secure=False, discard=False, rest={}, **kwargs):
3588 cookie = compat_cookiejar_Cookie(
3589 0, name, value, port, port is not None, domain, True,
3590 domain.startswith('.'), path, True, secure, expire_time,
3591 discard, None, None, rest)
3592 self._downloader.cookiejar.set_cookie(cookie)
3594 def _get_cookies(self, url):
3595 """ Return a compat_cookies_SimpleCookie with the cookies for the url """
3596 req = sanitized_Request(url)
3597 self._downloader.cookiejar.add_cookie_header(req)
3598 return compat_cookies_SimpleCookie(req.get_header('Cookie
'))
3600 def _apply_first_set_cookie_header(self, url_handle, cookie):
3602 Apply first Set-Cookie header instead of the last. Experimental.
3604 Some sites (e.g. [1-3]) may serve two cookies under the same name
3605 in Set-Cookie header and expect the first (old) one to be set rather
3606 than second (new). However, as of RFC6265 the newer one cookie
3607 should be set into cookie store what actually happens.
3608 We will workaround this issue by resetting the cookie to
3609 the first one manually.
3610 1. https://new.vk.com/
3611 2. https://github.com/ytdl-org/youtube-dl/issues/9841#issuecomment-227871201
3612 3. https://learning.oreilly.com/
3614 for header, cookies in url_handle.headers.items():
3615 if header.lower() != 'set-cookie
':
3617 cookies = cookies.encode('iso
-8859-1').decode('utf
-8')
3618 cookie_value = re.search(
3619 r'%s=(.+?
);.*?
\b[Dd
]omain
=(.+?
)(?
:[,;]|$
)' % cookie, cookies)
3621 value, domain = cookie_value.groups()
3622 self._set_cookie(domain, cookie, value)
3625 def get_testcases(self, include_onlymatching=False):
3626 t = getattr(self, '_TEST
', None)
3628 assert not hasattr(self, '_TESTS
'), \
3629 '%s has _TEST
and _TESTS
' % type(self).__name__
3632 tests = getattr(self, '_TESTS
', [])
3634 if not include_onlymatching and t.get('only_matching
', False):
3636 t['name
'] = type(self).__name__[:-len('IE
')]
3639 def is_suitable(self, age_limit):
3640 """ Test whether the extractor is generally suitable for the given
3641 age limit (i.e. pornographic sites are not, all others usually are) """
3643 any_restricted = False
3644 for tc in self.get_testcases(include_onlymatching=False):
3645 if tc.get('playlist
', []):
3646 tc = tc['playlist
'][0]
3647 is_restricted = age_restricted(
3648 tc.get('info_dict
', {}).get('age_limit
'), age_limit)
3649 if not is_restricted:
3651 any_restricted = any_restricted or is_restricted
3652 return not any_restricted
3654 def extract_subtitles(self, *args, **kwargs):
3655 if (self.get_param('writesubtitles
', False)
3656 or self.get_param('listsubtitles
')):
3657 return self._get_subtitles(*args, **kwargs)
3660 def _get_subtitles(self, *args, **kwargs):
3661 raise NotImplementedError('This method must be implemented by subclasses
')
3663 def extract_comments(self, *args, **kwargs):
3664 if not self.get_param('getcomments
'):
3666 generator = self._get_comments(*args, **kwargs)
3673 comments.append(next(generator))
3674 except StopIteration:
3676 except KeyboardInterrupt:
3677 self.to_screen('Interrupted by user
')
3678 except Exception as e:
3679 if self.get_param('ignoreerrors
') is not True:
3681 self._downloader.report_error(e)
3682 comment_count = len(comments)
3683 self.to_screen(f'Extracted {comment_count} comments
')
3685 'comments
': comments,
3686 'comment_count
': None if interrupted else comment_count
3690 def _get_comments(self, *args, **kwargs):
3691 raise NotImplementedError('This method must be implemented by subclasses
')
3694 def _merge_subtitle_items(subtitle_list1, subtitle_list2):
3695 """ Merge subtitle items for one language. Items with duplicated URLs/data
3696 will be dropped. """
3697 list1_data = {(item.get('url'), item.get('data')) for item in subtitle_list1}
3698 ret = list(subtitle_list1)
3699 ret.extend(item for item in subtitle_list2 if (item.get('url
'), item.get('data
')) not in list1_data)
3703 def _merge_subtitles(cls, *dicts, target=None):
3704 """ Merge subtitle dictionaries, language by language. """
3708 for lang, subs in d.items():
3709 target[lang] = cls._merge_subtitle_items(target.get(lang, []), subs)
3712 def extract_automatic_captions(self, *args, **kwargs):
3713 if (self.get_param('writeautomaticsub
', False)
3714 or self.get_param('listsubtitles
')):
3715 return self._get_automatic_captions(*args, **kwargs)
3718 def _get_automatic_captions(self, *args, **kwargs):
3719 raise NotImplementedError('This method must be implemented by subclasses
')
3721 def mark_watched(self, *args, **kwargs):
3722 if not self.get_param('mark_watched
', False):
3724 if (self.supports_login() and self._get_login_info()[0] is not None
3725 or self.get_param('cookiefile
') or self.get_param('cookiesfrombrowser
')):
3726 self._mark_watched(*args, **kwargs)
3728 def _mark_watched(self, *args, **kwargs):
3729 raise NotImplementedError('This method must be implemented by subclasses
')
3731 def geo_verification_headers(self):
3733 geo_verification_proxy = self.get_param('geo_verification_proxy
')
3734 if geo_verification_proxy:
3735 headers['Ytdl
-request
-proxy
'] = geo_verification_proxy
3738 def _generic_id(self, url):
3739 return compat_urllib_parse_unquote(os.path.splitext(url.rstrip('/').split('/')[-1])[0])
3741 def _generic_title(self, url):
3742 return compat_urllib_parse_unquote(os.path.splitext(url_basename(url))[0])
3745 def _availability(is_private=None, needs_premium=None, needs_subscription=None, needs_auth=None, is_unlisted=None):
3746 all_known = all(map(
3747 lambda x: x is not None,
3748 (is_private, needs_premium, needs_subscription, needs_auth, is_unlisted)))
3750 'private
' if is_private
3751 else 'premium_only
' if needs_premium
3752 else 'subscriber_only
' if needs_subscription
3753 else 'needs_auth
' if needs_auth
3754 else 'unlisted
' if is_unlisted
3755 else 'public
' if all_known
3758 def _configuration_arg(self, key, default=NO_DEFAULT, *, ie_key=None, casesense=False):
3760 @returns A list of values for the extractor argument given by "key"
3761 or "default" if no such key is present
3762 @param default The default value to return when the key is not present (default: [])
3763 @param casesense When false, the values are converted to lower case
3766 self._downloader.params, ('extractor_args
', (ie_key or self.ie_key()).lower(), key))
3768 return [] if default is NO_DEFAULT else default
3769 return list(val) if casesense else [x.lower() for x in val]
3771 def _yes_playlist(self, playlist_id, video_id, smuggled_data=None, *, playlist_label='playlist
', video_label='video
'):
3772 if not playlist_id or not video_id:
3775 no_playlist = (smuggled_data or {}).get('force_noplaylist
')
3776 if no_playlist is not None:
3777 return not no_playlist
3779 video_id = '' if video_id is True else f' {video_id}
'
3780 playlist_id = '' if playlist_id is True else f' {playlist_id}
'
3781 if self.get_param('noplaylist
'):
3782 self.to_screen(f'Downloading just the {video_label}{video_id} because of
--no
-playlist
')
3784 self.to_screen(f'Downloading {playlist_label}{playlist_id}
- add
--no
-playlist to download just the {video_label}{video_id}
')
3788 class SearchInfoExtractor(InfoExtractor):
3790 Base class for paged search queries extractors.
3791 They accept URLs in the format _SEARCH_KEY(|all|[0-9]):{query}
3792 Instances should define _SEARCH_KEY and optionally _MAX_RESULTS
3795 _MAX_RESULTS = float('inf
')
3798 def _make_valid_url(cls):
3799 return r'%s(?P
<prefix
>|
[1-9][0-9]*|all
):(?P
<query
>[\s\S
]+)' % cls._SEARCH_KEY
3801 def _real_extract(self, query):
3802 prefix, query = self._match_valid_url(query).group('prefix
', 'query
')
3804 return self._get_n_results(query, 1)
3805 elif prefix == 'all
':
3806 return self._get_n_results(query, self._MAX_RESULTS)
3810 raise ExtractorError(f'invalid download number {n}
for query
"{query}"')
3811 elif n > self._MAX_RESULTS:
3812 self.report_warning('%s returns
max %i results (you requested
%i)' % (self._SEARCH_KEY, self._MAX_RESULTS, n))
3813 n = self._MAX_RESULTS
3814 return self._get_n_results(query, n)
3816 def _get_n_results(self, query, n):
3817 """Get a specified number of results for a query.
3818 Either this function or _search_results must be overridden by subclasses """
3819 return self.playlist_result(
3820 itertools.islice(self._search_results(query), 0, None if n == float('inf
') else n),
3823 def _search_results(self, query):
3824 """Returns an iterator of search results"""
3825 raise NotImplementedError('This method must be implemented by subclasses
')
3828 def SEARCH_KEY(self):
3829 return self._SEARCH_KEY