2 from __future__
import unicode_literals
17 from ..compat
import (
18 compat_cookiejar_Cookie
,
19 compat_cookies_SimpleCookie
,
21 compat_etree_fromstring
,
29 compat_urllib_parse_unquote
,
30 compat_urllib_parse_urlencode
,
31 compat_urllib_request
,
33 compat_xml_parse_error
,
35 from ..downloader
import FileDownloader
36 from ..downloader
.f4m
import (
38 remove_encrypted_media
,
70 parse_m3u8_attributes
,
96 class InfoExtractor(object):
97 """Information Extractor class.
99 Information extractors are the classes that, given a URL, extract
100 information about the video (or videos) the URL refers to. This
101 information includes the real video URL, the video title, author and
102 others. The information is stored in a dictionary which is then
103 passed to the YoutubeDL. The YoutubeDL processes this
104 information possibly downloading the video to the file system, among
105 other possible outcomes.
107 The type field determines the type of the result.
108 By far the most common value (and the default if _type is missing) is
109 "video", which indicates a single video.
111 For a video, the dictionaries must include the following fields:
113 id: Video identifier.
114 title: Video title, unescaped.
116 Additionally, it must contain either a formats entry or a url one:
118 formats: A list of dictionaries for each format available, ordered
119 from worst to best quality.
122 * url The mandatory URL representing the media:
123 for plain file media - HTTP URL of this file,
125 for HLS - URL of the M3U8 media playlist,
126 for HDS - URL of the F4M manifest,
128 - HTTP URL to plain file media (in case of
130 - URL of the MPD manifest or base URL
131 representing the media if MPD manifest
132 is parsed from a string (in case of
134 for MSS - URL of the ISM manifest.
136 The URL of the manifest file in case of
138 for HLS - URL of the M3U8 master playlist,
139 for HDS - URL of the F4M manifest,
140 for DASH - URL of the MPD manifest,
141 for MSS - URL of the ISM manifest.
142 * manifest_stream_number (For internal use only)
143 The index of the stream in the manifest file
144 * ext Will be calculated from URL if missing
145 * format A human-readable description of the format
146 ("mp4 container with h264/opus").
147 Calculated from the format_id, width, height.
148 and format_note fields if missing.
149 * format_id A short description of the format
150 ("mp4_h264_opus" or "19").
151 Technically optional, but strongly recommended.
152 * format_note Additional info about the format
153 ("3D" or "DASH video")
154 * width Width of the video, if known
155 * height Height of the video, if known
156 * resolution Textual description of width and height
157 * dynamic_range The dynamic range of the video. One of:
158 "SDR" (None), "HDR10", "HDR10+, "HDR12", "HLG, "DV"
159 * tbr Average bitrate of audio and video in KBit/s
160 * abr Average audio bitrate in KBit/s
161 * acodec Name of the audio codec in use
162 * asr Audio sampling rate in Hertz
163 * vbr Average video bitrate in KBit/s
165 * vcodec Name of the video codec in use
166 * container Name of the container format
167 * filesize The number of bytes, if known in advance
168 * filesize_approx An estimate for the number of bytes
169 * player_url SWF Player URL (used for rtmpdump).
170 * protocol The protocol that will be used for the actual
171 download, lower-case. One of "http", "https" or
172 one of the protocols defined in downloader.PROTOCOL_MAP
174 Base URL for fragments. Each fragment's path
175 value (if present) will be relative to
177 * fragments A list of fragments of a fragmented media.
178 Each fragment entry must contain either an url
179 or a path. If an url is present it should be
180 considered by a client. Otherwise both path and
181 fragment_base_url must be present. Here is
182 the list of all potential fields:
183 * "url" - fragment's URL
184 * "path" - fragment's path relative to
186 * "duration" (optional, int or float)
187 * "filesize" (optional, int)
188 * is_from_start Is a live format that can be downloaded
189 from the start. Boolean
190 * preference Order number of this format. If this field is
191 present and not None, the formats get sorted
192 by this field, regardless of all other values.
193 -1 for default (order by other properties),
194 -2 or smaller for less than default.
195 < -1000 to hide the format (if there is
196 another one which is strictly better)
197 * language Language code, e.g. "de" or "en-US".
198 * language_preference Is this in the language mentioned in
200 10 if it's what the URL is about,
201 -1 for default (don't know),
202 -10 otherwise, other values reserved for now.
203 * quality Order number of the video quality of this
204 format, irrespective of the file format.
205 -1 for default (order by other properties),
206 -2 or smaller for less than default.
207 * source_preference Order number for this video source
208 (quality takes higher priority)
209 -1 for default (order by other properties),
210 -2 or smaller for less than default.
211 * http_headers A dictionary of additional HTTP headers
212 to add to the request.
213 * stretched_ratio If given and not 1, indicates that the
214 video's pixels are not square.
215 width : height ratio as float.
216 * no_resume The server does not support resuming the
217 (HTTP or RTMP) download. Boolean.
218 * has_drm The format has DRM and cannot be downloaded. Boolean
219 * downloader_options A dictionary of downloader options as
220 described in FileDownloader (For internal use only)
221 RTMP formats can also have the additional fields: page_url,
222 app, play_path, tc_url, flash_version, rtmp_live, rtmp_conn,
223 rtmp_protocol, rtmp_real_time
225 url: Final video URL.
226 ext: Video filename extension.
227 format: The video format, defaults to ext (used for --get-format)
228 player_url: SWF Player URL (used for rtmpdump).
230 The following fields are optional:
232 direct: True if a direct video file was given (must only be set by GenericIE)
233 alt_title: A secondary title of the video.
234 display_id An alternative identifier for the video, not necessarily
235 unique, but available before title. Typically, id is
236 something like "4234987", title "Dancing naked mole rats",
237 and display_id "dancing-naked-mole-rats"
238 thumbnails: A list of dictionaries, with the following entries:
239 * "id" (optional, string) - Thumbnail format ID
241 * "preference" (optional, int) - quality of the image
242 * "width" (optional, int)
243 * "height" (optional, int)
244 * "resolution" (optional, string "{width}x{height}",
246 * "filesize" (optional, int)
247 * "http_headers" (dict) - HTTP headers for the request
248 thumbnail: Full URL to a video thumbnail image.
249 description: Full video description.
250 uploader: Full name of the video uploader.
251 license: License name the video is licensed under.
252 creator: The creator of the video.
253 timestamp: UNIX timestamp of the moment the video was uploaded
254 upload_date: Video upload date in UTC (YYYYMMDD).
255 If not explicitly set, calculated from timestamp
256 release_timestamp: UNIX timestamp of the moment the video was released.
257 If it is not clear whether to use timestamp or this, use the former
258 release_date: The date (YYYYMMDD) when the video was released in UTC.
259 If not explicitly set, calculated from release_timestamp
260 modified_timestamp: UNIX timestamp of the moment the video was last modified.
261 modified_date: The date (YYYYMMDD) when the video was last modified in UTC.
262 If not explicitly set, calculated from modified_timestamp
263 uploader_id: Nickname or id of the video uploader.
264 uploader_url: Full URL to a personal webpage of the video uploader.
265 channel: Full name of the channel the video is uploaded on.
266 Note that channel fields may or may not repeat uploader
267 fields. This depends on a particular extractor.
268 channel_id: Id of the channel.
269 channel_url: Full URL to a channel webpage.
270 channel_follower_count: Number of followers of the channel.
271 location: Physical location where the video was filmed.
272 subtitles: The available subtitles as a dictionary in the format
273 {tag: subformats}. "tag" is usually a language code, and
274 "subformats" is a list sorted from lower to higher
275 preference, each element is a dictionary with the "ext"
277 * "data": The subtitles file contents
278 * "url": A URL pointing to the subtitles file
279 It can optionally also have:
280 * "name": Name or description of the subtitles
281 * "http_headers": A dictionary of additional HTTP headers
282 to add to the request.
283 "ext" will be calculated from URL if missing
284 automatic_captions: Like 'subtitles'; contains automatically generated
285 captions instead of normal subtitles
286 duration: Length of the video in seconds, as an integer or float.
287 view_count: How many users have watched the video on the platform.
288 like_count: Number of positive ratings of the video
289 dislike_count: Number of negative ratings of the video
290 repost_count: Number of reposts of the video
291 average_rating: Average rating give by users, the scale used depends on the webpage
292 comment_count: Number of comments on the video
293 comments: A list of comments, each with one or more of the following
294 properties (all but one of text or html optional):
295 * "author" - human-readable name of the comment author
296 * "author_id" - user ID of the comment author
297 * "author_thumbnail" - The thumbnail of the comment author
299 * "html" - Comment as HTML
300 * "text" - Plain text of the comment
301 * "timestamp" - UNIX timestamp of comment
302 * "parent" - ID of the comment this one is replying to.
303 Set to "root" to indicate that this is a
304 comment to the original video.
305 * "like_count" - Number of positive ratings of the comment
306 * "dislike_count" - Number of negative ratings of the comment
307 * "is_favorited" - Whether the comment is marked as
308 favorite by the video uploader
309 * "author_is_uploader" - Whether the comment is made by
311 age_limit: Age restriction for the video, as an integer (years)
312 webpage_url: The URL to the video webpage, if given to yt-dlp it
313 should allow to get the same result again. (It will be set
314 by YoutubeDL if it's missing)
315 categories: A list of categories that the video falls in, for example
317 tags: A list of tags assigned to the video, e.g. ["sweden", "pop music"]
318 cast: A list of the video cast
319 is_live: True, False, or None (=unknown). Whether this video is a
320 live stream that goes on instead of a fixed-length video.
321 was_live: True, False, or None (=unknown). Whether this video was
322 originally a live stream.
323 live_status: 'is_live', 'is_upcoming', 'was_live', 'not_live' or None (=unknown)
324 If absent, automatically set from is_live, was_live
325 start_time: Time in seconds where the reproduction should start, as
326 specified in the URL.
327 end_time: Time in seconds where the reproduction should end, as
328 specified in the URL.
329 chapters: A list of dictionaries, with the following entries:
330 * "start_time" - The start time of the chapter in seconds
331 * "end_time" - The end time of the chapter in seconds
332 * "title" (optional, string)
333 playable_in_embed: Whether this video is allowed to play in embedded
334 players on other sites. Can be True (=always allowed),
335 False (=never allowed), None (=unknown), or a string
336 specifying the criteria for embedability (Eg: 'whitelist')
337 availability: Under what condition the video is available. One of
338 'private', 'premium_only', 'subscriber_only', 'needs_auth',
339 'unlisted' or 'public'. Use 'InfoExtractor._availability'
341 __post_extractor: A function to be called just before the metadata is
342 written to either disk, logger or console. The function
343 must return a dict which will be added to the info_dict.
344 This is usefull for additional information that is
345 time-consuming to extract. Note that the fields thus
346 extracted will not be available to output template and
347 match_filter. So, only "comments" and "comment_count" are
348 currently allowed to be extracted via this method.
350 The following fields should only be used when the video belongs to some logical
353 chapter: Name or title of the chapter the video belongs to.
354 chapter_number: Number of the chapter the video belongs to, as an integer.
355 chapter_id: Id of the chapter the video belongs to, as a unicode string.
357 The following fields should only be used when the video is an episode of some
358 series, programme or podcast:
360 series: Title of the series or programme the video episode belongs to.
361 series_id: Id of the series or programme the video episode belongs to, as a unicode string.
362 season: Title of the season the video episode belongs to.
363 season_number: Number of the season the video episode belongs to, as an integer.
364 season_id: Id of the season the video episode belongs to, as a unicode string.
365 episode: Title of the video episode. Unlike mandatory video title field,
366 this field should denote the exact title of the video episode
367 without any kind of decoration.
368 episode_number: Number of the video episode within a season, as an integer.
369 episode_id: Id of the video episode, as a unicode string.
371 The following fields should only be used when the media is a track or a part of
374 track: Title of the track.
375 track_number: Number of the track within an album or a disc, as an integer.
376 track_id: Id of the track (useful in case of custom indexing, e.g. 6.iii),
378 artist: Artist(s) of the track.
379 genre: Genre(s) of the track.
380 album: Title of the album the track belongs to.
381 album_type: Type of the album (e.g. "Demo", "Full-length", "Split", "Compilation", etc).
382 album_artist: List of all artists appeared on the album (e.g.
383 "Ash Borer / Fell Voices" or "Various Artists", useful for splits
385 disc_number: Number of the disc or other physical medium the track belongs to,
387 release_year: Year (YYYY) when the album was released.
388 composer: Composer of the piece
390 Unless mentioned otherwise, the fields should be Unicode strings.
392 Unless mentioned otherwise, None is equivalent to absence of information.
395 _type "playlist" indicates multiple videos.
396 There must be a key "entries", which is a list, an iterable, or a PagedList
397 object, each element of which is a valid dictionary by this specification.
399 Additionally, playlists can have "id", "title", and any other relevent
400 attributes with the same semantics as videos (see above).
402 It can also have the following optional fields:
404 playlist_count: The total number of videos in a playlist. If not given,
405 YoutubeDL tries to calculate it from "entries"
408 _type "multi_video" indicates that there are multiple videos that
409 form a single show, for examples multiple acts of an opera or TV episode.
410 It must have an entries key like a playlist and contain all the keys
411 required for a video at the same time.
414 _type "url" indicates that the video must be extracted from another
415 location, possibly by a different extractor. Its only required key is:
416 "url" - the next URL to extract.
417 The key "ie_key" can be set to the class name (minus the trailing "IE",
418 e.g. "Youtube") if the extractor class is known in advance.
419 Additionally, the dictionary may have any properties of the resolved entity
420 known in advance, for example "title" if the title of the referred video is
424 _type "url_transparent" entities have the same specification as "url", but
425 indicate that the given additional information is more precise than the one
426 associated with the resolved URL.
427 This is useful when a site employs a video service that hosts the video and
428 its technical metadata, but that video service does not embed a useful
429 title, description etc.
432 Subclasses of this should define a _VALID_URL regexp and, re-define the
433 _real_extract() and (optionally) _real_initialize() methods.
434 Probably, they should also be added to the list of extractors.
436 Subclasses may also override suitable() if necessary, but ensure the function
437 signature is preserved and that this function imports everything it needs
438 (except other extractors), so that lazy_extractors works correctly.
440 To support username + password (or netrc) login, the extractor must define a
441 _NETRC_MACHINE and re-define _perform_login(username, password) and
442 (optionally) _initialize_pre_login() methods. The _perform_login method will
443 be called between _initialize_pre_login and _real_initialize if credentials
444 are passed by the user. In cases where it is necessary to have the login
445 process as part of the extraction rather than initialization, _perform_login
446 can be left undefined.
448 _GEO_BYPASS attribute may be set to False in order to disable
449 geo restriction bypass mechanisms for a particular extractor.
450 Though it won't disable explicit geo restriction bypass based on
451 country code provided with geo_bypass_country.
453 _GEO_COUNTRIES attribute may contain a list of presumably geo unrestricted
454 countries for this extractor. One of these countries will be used by
455 geo restriction bypass mechanism right away in order to bypass
456 geo restriction, of course, if the mechanism is not disabled.
458 _GEO_IP_BLOCKS attribute may contain a list of presumably geo unrestricted
459 IP blocks in CIDR notation for this extractor. One of these IP blocks
460 will be used by geo restriction bypass mechanism similarly
463 The _WORKING attribute should be set to False for broken IEs
464 in order to warn the users and skip the tests.
469 _x_forwarded_for_ip
= None
471 _GEO_COUNTRIES
= None
472 _GEO_IP_BLOCKS
= None
474 _NETRC_MACHINE
= None
478 'any': 'Use --cookies, --cookies-from-browser, --username and --password, or --netrc to provide account credentials',
480 'Use --cookies-from-browser or --cookies for the authentication. '
481 'See https://github.com/ytdl-org/youtube-dl#how-do-i-pass-cookies-to-youtube-dl for how to manually pass cookies'),
482 'password': 'Use --username and --password, or --netrc to provide account credentials',
485 def __init__(self
, downloader
=None):
486 """Constructor. Receives an optional downloader (a YoutubeDL instance).
487 If a downloader is not passed during initialization,
488 it must be set using "set_downloader()" before "extract()" is called"""
490 self
._x
_forwarded
_for
_ip
= None
491 self
._printed
_messages
= set()
492 self
.set_downloader(downloader
)
495 def _match_valid_url(cls
, url
):
496 # This does not use has/getattr intentionally - we want to know whether
497 # we have cached the regexp for *this* class, whereas getattr would also
498 # match the superclass
499 if '_VALID_URL_RE' not in cls
.__dict
__:
500 if '_VALID_URL' not in cls
.__dict
__:
501 cls
._VALID
_URL
= cls
._make
_valid
_url
()
502 cls
._VALID
_URL
_RE
= re
.compile(cls
._VALID
_URL
)
503 return cls
._VALID
_URL
_RE
.match(url
)
506 def suitable(cls
, url
):
507 """Receives a URL and returns True if suitable for this IE."""
508 # This function must import everything it needs (except other extractors),
509 # so that lazy_extractors works correctly
510 return cls
._match
_valid
_url
(url
) is not None
513 def _match_id(cls
, url
):
514 return cls
._match
_valid
_url
(url
).group('id')
517 def get_temp_id(cls
, url
):
519 return cls
._match
_id
(url
)
520 except (IndexError, AttributeError):
525 """Getter method for _WORKING."""
529 def supports_login(cls
):
530 return bool(cls
._NETRC
_MACHINE
)
532 def initialize(self
):
533 """Initializes an instance (authentication, etc)."""
534 self
._printed
_messages
= set()
535 self
._initialize
_geo
_bypass
({
536 'countries': self
._GEO
_COUNTRIES
,
537 'ip_blocks': self
._GEO
_IP
_BLOCKS
,
540 self
._initialize
_pre
_login
()
541 if self
.supports_login():
542 username
, password
= self
._get
_login
_info
()
544 self
._perform
_login
(username
, password
)
545 elif self
.get_param('username') and False not in (self
.IE_DESC
, self
._NETRC
_MACHINE
):
546 self
.report_warning(f
'Login with password is not supported for this website. {self._LOGIN_HINTS["cookies"]}')
547 self
._real
_initialize
()
550 def _initialize_geo_bypass(self
, geo_bypass_context
):
552 Initialize geo restriction bypass mechanism.
554 This method is used to initialize geo bypass mechanism based on faking
555 X-Forwarded-For HTTP header. A random country from provided country list
556 is selected and a random IP belonging to this country is generated. This
557 IP will be passed as X-Forwarded-For HTTP header in all subsequent
560 This method will be used for initial geo bypass mechanism initialization
561 during the instance initialization with _GEO_COUNTRIES and
564 You may also manually call it from extractor's code if geo bypass
565 information is not available beforehand (e.g. obtained during
566 extraction) or due to some other reason. In this case you should pass
567 this information in geo bypass context passed as first argument. It may
568 contain following fields:
570 countries: List of geo unrestricted countries (similar
572 ip_blocks: List of geo unrestricted IP blocks in CIDR notation
573 (similar to _GEO_IP_BLOCKS)
576 if not self
._x
_forwarded
_for
_ip
:
578 # Geo bypass mechanism is explicitly disabled by user
579 if not self
.get_param('geo_bypass', True):
582 if not geo_bypass_context
:
583 geo_bypass_context
= {}
585 # Backward compatibility: previously _initialize_geo_bypass
586 # expected a list of countries, some 3rd party code may still use
588 if isinstance(geo_bypass_context
, (list, tuple)):
589 geo_bypass_context
= {
590 'countries': geo_bypass_context
,
593 # The whole point of geo bypass mechanism is to fake IP
594 # as X-Forwarded-For HTTP header based on some IP block or
597 # Path 1: bypassing based on IP block in CIDR notation
599 # Explicit IP block specified by user, use it right away
600 # regardless of whether extractor is geo bypassable or not
601 ip_block
= self
.get_param('geo_bypass_ip_block', None)
603 # Otherwise use random IP block from geo bypass context but only
604 # if extractor is known as geo bypassable
606 ip_blocks
= geo_bypass_context
.get('ip_blocks')
607 if self
._GEO
_BYPASS
and ip_blocks
:
608 ip_block
= random
.choice(ip_blocks
)
611 self
._x
_forwarded
_for
_ip
= GeoUtils
.random_ipv4(ip_block
)
612 self
._downloader
.write_debug(
613 '[debug] Using fake IP %s as X-Forwarded-For' % self
._x
_forwarded
_for
_ip
)
616 # Path 2: bypassing based on country code
618 # Explicit country code specified by user, use it right away
619 # regardless of whether extractor is geo bypassable or not
620 country
= self
.get_param('geo_bypass_country', None)
622 # Otherwise use random country code from geo bypass context but
623 # only if extractor is known as geo bypassable
625 countries
= geo_bypass_context
.get('countries')
626 if self
._GEO
_BYPASS
and countries
:
627 country
= random
.choice(countries
)
630 self
._x
_forwarded
_for
_ip
= GeoUtils
.random_ipv4(country
)
631 self
._downloader
.write_debug(
632 'Using fake IP %s (%s) as X-Forwarded-For' % (self
._x
_forwarded
_for
_ip
, country
.upper()))
634 def extract(self
, url
):
635 """Extracts URL information and returns it in list of dicts."""
640 self
.write_debug('Extracting URL: %s' % url
)
641 ie_result
= self
._real
_extract
(url
)
642 if ie_result
is None:
644 if self
._x
_forwarded
_for
_ip
:
645 ie_result
['__x_forwarded_for_ip'] = self
._x
_forwarded
_for
_ip
646 subtitles
= ie_result
.get('subtitles')
647 if (subtitles
and 'live_chat' in subtitles
648 and 'no-live-chat' in self
.get_param('compat_opts', [])):
649 del subtitles
['live_chat']
651 except GeoRestrictedError
as e
:
652 if self
.__maybe
_fake
_ip
_and
_retry
(e
.countries
):
655 except UnsupportedError
:
657 except ExtractorError
as e
:
659 'video_id': e
.video_id
or self
.get_temp_id(url
),
661 'tb': e
.traceback
or sys
.exc_info()[2],
662 'expected': e
.expected
,
665 if hasattr(e
, 'countries'):
666 kwargs
['countries'] = e
.countries
667 raise type(e
)(e
.orig_msg
, **kwargs
)
668 except compat_http_client
.IncompleteRead
as e
:
669 raise ExtractorError('A network error has occurred.', cause
=e
, expected
=True, video_id
=self
.get_temp_id(url
))
670 except (KeyError, StopIteration) as e
:
671 raise ExtractorError('An extractor error has occurred.', cause
=e
, video_id
=self
.get_temp_id(url
))
673 def __maybe_fake_ip_and_retry(self
, countries
):
674 if (not self
.get_param('geo_bypass_country', None)
676 and self
.get_param('geo_bypass', True)
677 and not self
._x
_forwarded
_for
_ip
679 country_code
= random
.choice(countries
)
680 self
._x
_forwarded
_for
_ip
= GeoUtils
.random_ipv4(country_code
)
681 if self
._x
_forwarded
_for
_ip
:
683 'Video is geo restricted. Retrying extraction with fake IP %s (%s) as X-Forwarded-For.'
684 % (self
._x
_forwarded
_for
_ip
, country_code
.upper()))
688 def set_downloader(self
, downloader
):
689 """Sets a YoutubeDL instance as the downloader for this IE."""
690 self
._downloader
= downloader
692 def _initialize_pre_login(self
):
693 """ Intialization before login. Redefine in subclasses."""
696 def _perform_login(self
, username
, password
):
697 """ Login with username and password. Redefine in subclasses."""
700 def _real_initialize(self
):
701 """Real initialization process. Redefine in subclasses."""
704 def _real_extract(self
, url
):
705 """Real extraction process. Redefine in subclasses."""
706 raise NotImplementedError('This method must be implemented by subclasses')
710 """A string for getting the InfoExtractor with get_info_extractor"""
711 return cls
.__name
__[:-2]
715 return compat_str(type(self
).__name
__[:-2])
718 def __can_accept_status_code(err
, expected_status
):
719 assert isinstance(err
, compat_urllib_error
.HTTPError
)
720 if expected_status
is None:
722 elif callable(expected_status
):
723 return expected_status(err
.code
) is True
725 return err
.code
in variadic(expected_status
)
727 def _request_webpage(self
, url_or_request
, video_id
, note
=None, errnote
=None, fatal
=True, data
=None, headers
={}, query={}
, expected_status
=None):
729 Return the response handle.
731 See _download_webpage docstring for arguments specification.
733 if not self
._downloader
._first
_webpage
_request
:
734 sleep_interval
= self
.get_param('sleep_interval_requests') or 0
735 if sleep_interval
> 0:
736 self
.to_screen('Sleeping %s seconds ...' % sleep_interval
)
737 time
.sleep(sleep_interval
)
739 self
._downloader
._first
_webpage
_request
= False
742 self
.report_download_webpage(video_id
)
743 elif note
is not False:
745 self
.to_screen('%s' % (note
,))
747 self
.to_screen('%s: %s' % (video_id
, note
))
749 # Some sites check X-Forwarded-For HTTP header in order to figure out
750 # the origin of the client behind proxy. This allows bypassing geo
751 # restriction by faking this header's value to IP that belongs to some
752 # geo unrestricted country. We will do so once we encounter any
753 # geo restriction error.
754 if self
._x
_forwarded
_for
_ip
:
755 if 'X-Forwarded-For' not in headers
:
756 headers
['X-Forwarded-For'] = self
._x
_forwarded
_for
_ip
758 if isinstance(url_or_request
, compat_urllib_request
.Request
):
759 url_or_request
= update_Request(
760 url_or_request
, data
=data
, headers
=headers
, query
=query
)
763 url_or_request
= update_url_query(url_or_request
, query
)
764 if data
is not None or headers
:
765 url_or_request
= sanitized_Request(url_or_request
, data
, headers
)
767 return self
._downloader
.urlopen(url_or_request
)
768 except network_exceptions
as err
:
769 if isinstance(err
, compat_urllib_error
.HTTPError
):
770 if self
.__can
_accept
_status
_code
(err
, expected_status
):
771 # Retain reference to error to prevent file object from
772 # being closed before it can be read. Works around the
773 # effects of <https://bugs.python.org/issue15002>
774 # introduced in Python 3.4.1.
781 errnote
= 'Unable to download webpage'
783 errmsg
= '%s: %s' % (errnote
, error_to_compat_str(err
))
785 raise ExtractorError(errmsg
, cause
=err
)
787 self
.report_warning(errmsg
)
790 def _download_webpage_handle(self
, url_or_request
, video_id
, note
=None, errnote
=None, fatal
=True, encoding
=None, data
=None, headers
={}, query={}
, expected_status
=None):
792 Return a tuple (page content as string, URL handle).
794 See _download_webpage docstring for arguments specification.
796 # Strip hashes from the URL (#1038)
797 if isinstance(url_or_request
, (compat_str
, str)):
798 url_or_request
= url_or_request
.partition('#')[0]
800 urlh
= self
._request
_webpage
(url_or_request
, video_id
, note
, errnote
, fatal
, data
=data
, headers
=headers
, query
=query
, expected_status
=expected_status
)
804 content
= self
._webpage
_read
_content
(urlh
, url_or_request
, video_id
, note
, errnote
, fatal
, encoding
=encoding
)
805 return (content
, urlh
)
808 def _guess_encoding_from_content(content_type
, webpage_bytes
):
809 m
= re
.match(r
'[a-zA-Z0-9_.-]+/[a-zA-Z0-9_.-]+\s*;\s*charset=(.+)', content_type
)
811 encoding
= m
.group(1)
813 m
= re
.search(br
'<meta[^>]+charset=[\'"]?([^\'")]+)[ /\'">]',
814 webpage_bytes[:1024])
816 encoding = m.group(1).decode('ascii')
817 elif webpage_bytes.startswith(b'\xff\xfe'):
824 def __check_blocked(self, content):
825 first_block = content[:512]
826 if ('<title>Access to this site is blocked</title>' in content
827 and 'Websense' in first_block):
828 msg = 'Access to this webpage has been blocked by Websense filtering software in your network.'
829 blocked_iframe = self._html_search_regex(
830 r'<iframe src="([^
"]+)"', content,
831 'Websense information URL
', default=None)
833 msg += ' Visit
%s for more details
' % blocked_iframe
834 raise ExtractorError(msg, expected=True)
835 if '<title
>The URL you requested has been blocked
</title
>' in first_block:
837 'Access to this webpage has been blocked by Indian censorship
. '
838 'Use a VPN
or proxy
server (with --proxy
) to route around it
.')
839 block_msg = self._html_search_regex(
840 r'</h1
><p
>(.*?
)</p
>',
841 content, 'block message
', default=None)
843 msg += ' (Message
: "%s")' % block_msg.replace('\n', ' ')
844 raise ExtractorError(msg, expected=True)
845 if ('<title
>TTK
:: Доступ к ресурсу ограничен
</title
>' in content
846 and 'blocklist
.rkn
.gov
.ru
' in content):
847 raise ExtractorError(
848 'Access to this webpage has been blocked by decision of the Russian government
. '
849 'Visit http
://blocklist
.rkn
.gov
.ru
/ for a block reason
.',
852 def _webpage_read_content(self, urlh, url_or_request, video_id, note=None, errnote=None, fatal=True, prefix=None, encoding=None):
853 content_type = urlh.headers.get('Content
-Type
', '')
854 webpage_bytes = urlh.read()
855 if prefix is not None:
856 webpage_bytes = prefix + webpage_bytes
858 encoding = self._guess_encoding_from_content(content_type, webpage_bytes)
859 if self.get_param('dump_intermediate_pages
', False):
860 self.to_screen('Dumping request to
' + urlh.geturl())
861 dump = base64.b64encode(webpage_bytes).decode('ascii
')
862 self._downloader.to_screen(dump)
863 if self.get_param('write_pages
', False):
864 basen = '%s_%s' % (video_id, urlh.geturl())
865 trim_length = self.get_param('trim_file_name
') or 240
866 if len(basen) > trim_length:
867 h = '___
' + hashlib.md5(basen.encode('utf
-8')).hexdigest()
868 basen = basen[:trim_length - len(h)] + h
869 raw_filename = basen + '.dump
'
870 filename = sanitize_filename(raw_filename, restricted=True)
871 self.to_screen('Saving request to
' + filename)
872 # Working around MAX_PATH limitation on Windows (see
873 # http://msdn.microsoft.com/en-us/library/windows/desktop/aa365247(v=vs.85).aspx)
874 if compat_os_name == 'nt
':
875 absfilepath = os.path.abspath(filename)
876 if len(absfilepath) > 259:
877 filename = '\\\\?
\\' + absfilepath
878 with open(filename, 'wb
') as outf:
879 outf.write(webpage_bytes)
882 content = webpage_bytes.decode(encoding, 'replace
')
884 content = webpage_bytes.decode('utf
-8', 'replace
')
886 self.__check_blocked(content)
890 def _download_webpage(
891 self, url_or_request, video_id, note=None, errnote=None,
892 fatal=True, tries=1, timeout=5, encoding=None, data=None,
893 headers={}, query={}, expected_status=None):
895 Return the data of the page as a string.
898 url_or_request -- plain text URL as a string or
899 a compat_urllib_request.Requestobject
900 video_id -- Video/playlist/item identifier (string)
903 note -- note printed before downloading (string)
904 errnote -- note printed in case of an error (string)
905 fatal -- flag denoting whether error should be considered fatal,
906 i.e. whether it should cause ExtractionError to be raised,
907 otherwise a warning will be reported and extraction continued
908 tries -- number of tries
909 timeout -- sleep interval between tries
910 encoding -- encoding for a page content decoding, guessed automatically
911 when not explicitly specified
912 data -- POST data (bytes)
913 headers -- HTTP headers (dict)
914 query -- URL query (dict)
915 expected_status -- allows to accept failed HTTP requests (non 2xx
916 status code) by explicitly specifying a set of accepted status
917 codes. Can be any of the following entities:
918 - an integer type specifying an exact failed status code to
920 - a list or a tuple of integer types specifying a list of
921 failed status codes to accept
922 - a callable accepting an actual failed status code and
923 returning True if it should be accepted
924 Note that this argument does not affect success status codes (2xx)
925 which are always accepted.
930 while success is False:
932 res = self._download_webpage_handle(
933 url_or_request, video_id, note, errnote, fatal,
934 encoding=encoding, data=data, headers=headers, query=query,
935 expected_status=expected_status)
937 except compat_http_client.IncompleteRead as e:
939 if try_count >= tries:
941 self._sleep(timeout, video_id)
948 def _download_xml_handle(
949 self, url_or_request, video_id, note='Downloading XML
',
950 errnote='Unable to download XML
', transform_source=None,
951 fatal=True, encoding=None, data=None, headers={}, query={},
952 expected_status=None):
954 Return a tuple (xml as an compat_etree_Element, URL handle).
956 See _download_webpage docstring for arguments specification.
958 res = self._download_webpage_handle(
959 url_or_request, video_id, note, errnote, fatal=fatal,
960 encoding=encoding, data=data, headers=headers, query=query,
961 expected_status=expected_status)
964 xml_string, urlh = res
965 return self._parse_xml(
966 xml_string, video_id, transform_source=transform_source,
970 self, url_or_request, video_id,
971 note='Downloading XML
', errnote='Unable to download XML
',
972 transform_source=None, fatal=True, encoding=None,
973 data=None, headers={}, query={}, expected_status=None):
975 Return the xml as an compat_etree_Element.
977 See _download_webpage docstring for arguments specification.
979 res = self._download_xml_handle(
980 url_or_request, video_id, note=note, errnote=errnote,
981 transform_source=transform_source, fatal=fatal, encoding=encoding,
982 data=data, headers=headers, query=query,
983 expected_status=expected_status)
984 return res if res is False else res[0]
986 def _parse_xml(self, xml_string, video_id, transform_source=None, fatal=True):
988 xml_string = transform_source(xml_string)
990 return compat_etree_fromstring(xml_string.encode('utf
-8'))
991 except compat_xml_parse_error as ve:
992 errmsg = '%s: Failed to parse XML
' % video_id
994 raise ExtractorError(errmsg, cause=ve)
996 self.report_warning(errmsg + str(ve))
998 def _download_json_handle(
999 self, url_or_request, video_id, note='Downloading JSON metadata
',
1000 errnote='Unable to download JSON metadata
', transform_source=None,
1001 fatal=True, encoding=None, data=None, headers={}, query={},
1002 expected_status=None):
1004 Return a tuple (JSON object, URL handle).
1006 See _download_webpage docstring for arguments specification.
1008 res = self._download_webpage_handle(
1009 url_or_request, video_id, note, errnote, fatal=fatal,
1010 encoding=encoding, data=data, headers=headers, query=query,
1011 expected_status=expected_status)
1014 json_string, urlh = res
1015 return self._parse_json(
1016 json_string, video_id, transform_source=transform_source,
1020 self, url_or_request, video_id, note='Downloading JSON metadata
',
1021 errnote='Unable to download JSON metadata
', transform_source=None,
1022 fatal=True, encoding=None, data=None, headers={}, query={},
1023 expected_status=None):
1025 Return the JSON object as a dict.
1027 See _download_webpage docstring for arguments specification.
1029 res = self._download_json_handle(
1030 url_or_request, video_id, note=note, errnote=errnote,
1031 transform_source=transform_source, fatal=fatal, encoding=encoding,
1032 data=data, headers=headers, query=query,
1033 expected_status=expected_status)
1034 return res if res is False else res[0]
1036 def _parse_json(self, json_string, video_id, transform_source=None, fatal=True):
1037 if transform_source:
1038 json_string = transform_source(json_string)
1040 return json.loads(json_string, strict=False)
1041 except ValueError as ve:
1042 errmsg = '%s: Failed to parse JSON
' % video_id
1044 raise ExtractorError(errmsg, cause=ve)
1046 self.report_warning(errmsg + str(ve))
1048 def _parse_socket_response_as_json(self, data, video_id, transform_source=None, fatal=True):
1049 return self._parse_json(
1050 data[data.find('{'):data.rfind('}
') + 1],
1051 video_id, transform_source, fatal)
1053 def _download_socket_json_handle(
1054 self, url_or_request, video_id, note='Polling socket
',
1055 errnote='Unable to poll socket
', transform_source=None,
1056 fatal=True, encoding=None, data=None, headers={}, query={},
1057 expected_status=None):
1059 Return a tuple (JSON object, URL handle).
1061 See _download_webpage docstring for arguments specification.
1063 res = self._download_webpage_handle(
1064 url_or_request, video_id, note, errnote, fatal=fatal,
1065 encoding=encoding, data=data, headers=headers, query=query,
1066 expected_status=expected_status)
1070 return self._parse_socket_response_as_json(
1071 webpage, video_id, transform_source=transform_source,
1074 def _download_socket_json(
1075 self, url_or_request, video_id, note='Polling socket
',
1076 errnote='Unable to poll socket
', transform_source=None,
1077 fatal=True, encoding=None, data=None, headers={}, query={},
1078 expected_status=None):
1080 Return the JSON object as a dict.
1082 See _download_webpage docstring for arguments specification.
1084 res = self._download_socket_json_handle(
1085 url_or_request, video_id, note=note, errnote=errnote,
1086 transform_source=transform_source, fatal=fatal, encoding=encoding,
1087 data=data, headers=headers, query=query,
1088 expected_status=expected_status)
1089 return res if res is False else res[0]
1091 def report_warning(self, msg, video_id=None, *args, only_once=False, **kwargs):
1092 idstr = format_field(video_id, template='%s: ')
1093 msg = f'[{self.IE_NAME}
] {idstr}{msg}
'
1095 if f'WARNING
: {msg}
' in self._printed_messages:
1097 self._printed_messages.add(f'WARNING
: {msg}
')
1098 self._downloader.report_warning(msg, *args, **kwargs)
1100 def to_screen(self, msg, *args, **kwargs):
1101 """Print msg to screen, prefixing it with '[ie_name
]'"""
1102 self._downloader.to_screen('[%s] %s' % (self.IE_NAME, msg), *args, **kwargs)
1104 def write_debug(self, msg, *args, **kwargs):
1105 self._downloader.write_debug('[%s] %s' % (self.IE_NAME, msg), *args, **kwargs)
1107 def get_param(self, name, default=None, *args, **kwargs):
1108 if self._downloader:
1109 return self._downloader.params.get(name, default, *args, **kwargs)
1112 def report_drm(self, video_id, partial=False):
1113 self.raise_no_formats('This video
is DRM protected
', expected=True, video_id=video_id)
1115 def report_extraction(self, id_or_name):
1116 """Report information extraction."""
1117 self.to_screen('%s: Extracting information
' % id_or_name)
1119 def report_download_webpage(self, video_id):
1120 """Report webpage download."""
1121 self.to_screen('%s: Downloading webpage
' % video_id)
1123 def report_age_confirmation(self):
1124 """Report attempt to confirm age."""
1125 self.to_screen('Confirming age
')
1127 def report_login(self):
1128 """Report attempt to log in."""
1129 self.to_screen('Logging
in')
1131 def raise_login_required(
1132 self, msg='This video
is only available
for registered users
',
1133 metadata_available=False, method=NO_DEFAULT):
1134 if metadata_available and (
1135 self.get_param('ignore_no_formats_error
') or self.get_param('wait_for_video
')):
1136 self.report_warning(msg)
1138 if method is NO_DEFAULT:
1139 method = 'any
' if self.supports_login() else 'cookies
'
1140 if method is not None:
1141 assert method in self._LOGIN_HINTS, 'Invalid login method
'
1142 msg = '%s. %s' % (msg, self._LOGIN_HINTS[method])
1143 raise ExtractorError(msg, expected=True)
1145 def raise_geo_restricted(
1146 self, msg='This video
is not available
from your location due to geo restriction
',
1147 countries=None, metadata_available=False):
1148 if metadata_available and (
1149 self.get_param('ignore_no_formats_error
') or self.get_param('wait_for_video
')):
1150 self.report_warning(msg)
1152 raise GeoRestrictedError(msg, countries=countries)
1154 def raise_no_formats(self, msg, expected=False, video_id=None):
1156 self.get_param('ignore_no_formats_error
') or self.get_param('wait_for_video
')):
1157 self.report_warning(msg, video_id)
1158 elif isinstance(msg, ExtractorError):
1161 raise ExtractorError(msg, expected=expected, video_id=video_id)
1163 # Methods for following #608
1165 def url_result(url, ie=None, video_id=None, video_title=None, *, url_transparent=False, **kwargs):
1166 """Returns a URL that points to a page that should be processed"""
1168 kwargs['ie_key
'] = ie if isinstance(ie, str) else ie.ie_key()
1169 if video_id is not None:
1170 kwargs['id'] = video_id
1171 if video_title is not None:
1172 kwargs['title
'] = video_title
1175 '_type
': 'url_transparent
' if url_transparent else 'url
',
1179 def playlist_from_matches(self, matches, playlist_id=None, playlist_title=None, getter=None, ie=None, video_kwargs=None, **kwargs):
1180 urls = (self.url_result(self._proto_relative_url(m), ie, **(video_kwargs or {}))
1181 for m in orderedSet(map(getter, matches) if getter else matches))
1182 return self.playlist_result(urls, playlist_id, playlist_title, **kwargs)
1185 def playlist_result(entries, playlist_id=None, playlist_title=None, playlist_description=None, *, multi_video=False, **kwargs):
1186 """Returns a playlist"""
1188 kwargs['id'] = playlist_id
1190 kwargs['title
'] = playlist_title
1191 if playlist_description is not None:
1192 kwargs['description
'] = playlist_description
1195 '_type
': 'multi_video
' if multi_video else 'playlist
',
1199 def _search_regex(self, pattern, string, name, default=NO_DEFAULT, fatal=True, flags=0, group=None):
1201 Perform a regex search on the given string, using a single or a list of
1202 patterns returning the first matching group.
1203 In case of failure return a default value or raise a WARNING or a
1204 RegexNotFoundError, depending on fatal, specifying the field name.
1208 elif isinstance(pattern, (str, compat_Pattern)):
1209 mobj = re.search(pattern, string, flags)
1212 mobj = re.search(p, string, flags)
1216 _name = self._downloader._format_err(name, self._downloader.Styles.EMPHASIS)
1220 # return the first matching group
1221 return next(g for g in mobj.groups() if g is not None)
1222 elif isinstance(group, (list, tuple)):
1223 return tuple(mobj.group(g) for g in group)
1225 return mobj.group(group)
1226 elif default is not NO_DEFAULT:
1229 raise RegexNotFoundError('Unable to extract
%s' % _name)
1231 self.report_warning('unable to extract
%s' % _name + bug_reports_message())
1234 def _html_search_regex(self, pattern, string, name, default=NO_DEFAULT, fatal=True, flags=0, group=None):
1236 Like _search_regex, but strips HTML tags and unescapes entities.
1238 res = self._search_regex(pattern, string, name, default, fatal, flags, group)
1240 return clean_html(res).strip()
1244 def _get_netrc_login_info(self, netrc_machine=None):
1247 netrc_machine = netrc_machine or self._NETRC_MACHINE
1249 if self.get_param('usenetrc
', False):
1251 netrc_file = compat_expanduser(self.get_param('netrc_location
') or '~
')
1252 if os.path.isdir(netrc_file):
1253 netrc_file = os.path.join(netrc_file, '.netrc
')
1254 info = netrc.netrc(file=netrc_file).authenticators(netrc_machine)
1255 if info is not None:
1259 raise netrc.NetrcParseError(
1260 'No authenticators
for %s' % netrc_machine)
1261 except (IOError, netrc.NetrcParseError) as err:
1262 self.report_warning(
1263 'parsing
.netrc
: %s' % error_to_compat_str(err))
1265 return username, password
1267 def _get_login_info(self, username_option='username
', password_option='password
', netrc_machine=None):
1269 Get the login info as (username, password)
1270 First look for the manually specified credentials using username_option
1271 and password_option as keys in params dictionary. If no such credentials
1272 available look in the netrc file using the netrc_machine or _NETRC_MACHINE
1274 If there's no info available
, return (None, None)
1277 # Attempt to use provided username and password or .netrc data
1278 username = self.get_param(username_option)
1279 if username is not None:
1280 password = self.get_param(password_option)
1282 username, password = self._get_netrc_login_info(netrc_machine)
1284 return username, password
1286 def _get_tfa_info(self, note='two-factor verification code'):
1288 Get the two
-factor authentication info
1289 TODO
- asking the user will be required
for sms
/phone verify
1290 currently just uses the command line option
1291 If there
's no info available, return None
1294 tfa = self.get_param('twofactor
')
1298 return compat_getpass('Type
%s and press
[Return
]: ' % note)
1300 # Helper functions for extracting OpenGraph info
1302 def _og_regexes(prop):
1303 content_re = r'content
=(?
:"([^"]+?
)"|\'([^\']+?)\'|\s*([^\s"\'=<>`
]+?
))'
1304 property_re = (r'(?
:name|
property)=(?
:\'og
%(sep)s%(prop)s\'|
"og%(sep)s%(prop)s"|\s
*og
%(sep)s%(prop)s\b)'
1305 % {'prop': re.escape(prop), 'sep': '(?::|[:-])'})
1306 template = r'<meta
[^
>]+?
%s[^
>]+?
%s'
1308 template % (property_re, content_re),
1309 template % (content_re, property_re),
1313 def _meta_regex(prop):
1314 return r'''(?isx)<meta
1315 (?=[^>]+(?:itemprop|name|property|id|http-equiv)=(["\']?)%s\1)
1316 [^>]+?content=(["\'])(?P<content>.*?)\2''' % re.escape(prop)
1318 def _og_search_property(self, prop, html, name=None, **kargs):
1319 prop = variadic(prop)
1321 name = 'OpenGraph
%s' % prop[0]
1324 og_regexes.extend(self._og_regexes(p))
1325 escaped = self._search_regex(og_regexes, html, name, flags=re.DOTALL, **kargs)
1328 return unescapeHTML(escaped)
1330 def _og_search_thumbnail(self, html, **kargs):
1331 return self._og_search_property('image
', html, 'thumbnail URL
', fatal=False, **kargs)
1333 def _og_search_description(self, html, **kargs):
1334 return self._og_search_property('description
', html, fatal=False, **kargs)
1336 def _og_search_title(self, html, *, fatal=False, **kargs):
1337 return self._og_search_property('title
', html, fatal=fatal, **kargs)
1339 def _og_search_video_url(self, html, name='video url
', secure=True, **kargs):
1340 regexes = self._og_regexes('video
') + self._og_regexes('video
:url
')
1342 regexes = self._og_regexes('video
:secure_url
') + regexes
1343 return self._html_search_regex(regexes, html, name, **kargs)
1345 def _og_search_url(self, html, **kargs):
1346 return self._og_search_property('url
', html, **kargs)
1348 def _html_extract_title(self, html, name='title
', *, fatal=False, **kwargs):
1349 return self._html_search_regex(r'(?s
)<title
>([^
<]+)</title
>', html, name, fatal=fatal, **kwargs)
1351 def _html_search_meta(self, name, html, display_name=None, fatal=False, **kwargs):
1352 name = variadic(name)
1353 if display_name is None:
1354 display_name = name[0]
1355 return self._html_search_regex(
1356 [self._meta_regex(n) for n in name],
1357 html, display_name, fatal=fatal, group='content
', **kwargs)
1359 def _dc_search_uploader(self, html):
1360 return self._html_search_meta('dc
.creator
', html, 'uploader
')
1362 def _rta_search(self, html):
1363 # See http://www.rtalabel.org/index.php?content=howtofaq#single
1364 if re.search(r'(?ix
)<meta\s
+name
="rating"\s
+'
1365 r' content
="RTA-5042-1996-1400-1577-RTA"',
1370 def _media_rating_search(self, html):
1371 # See http://www.tjg-designs.com/WP/metadata-code-examples-adding-metadata-to-your-web-pages/
1372 rating = self._html_search_meta('rating
', html)
1384 return RATING_TABLE.get(rating.lower())
1386 def _family_friendly_search(self, html):
1387 # See http://schema.org/VideoObject
1388 family_friendly = self._html_search_meta(
1389 'isFamilyFriendly
', html, default=None)
1391 if not family_friendly:
1400 return RATING_TABLE.get(family_friendly.lower())
1402 def _twitter_search_player(self, html):
1403 return self._html_search_meta('twitter
:player
', html,
1404 'twitter card player
')
1406 def _search_json_ld(self, html, video_id, expected_type=None, **kwargs):
1407 json_ld_list = list(re.finditer(JSON_LD_RE, html))
1408 default = kwargs.get('default
', NO_DEFAULT)
1409 # JSON-LD may be malformed and thus `fatal` should be respected.
1410 # At the same time `default` may be passed that assumes `fatal=False`
1411 # for _search_regex. Let's simulate the same behavior here
as well
.
1412 fatal
= kwargs
.get('fatal', True) if default
is NO_DEFAULT
else False
1414 for mobj
in json_ld_list
:
1415 json_ld_item
= self
._parse
_json
(
1416 mobj
.group('json_ld'), video_id
, fatal
=fatal
)
1417 if not json_ld_item
:
1419 if isinstance(json_ld_item
, dict):
1420 json_ld
.append(json_ld_item
)
1421 elif isinstance(json_ld_item
, (list, tuple)):
1422 json_ld
.extend(json_ld_item
)
1424 json_ld
= self
._json
_ld
(json_ld
, video_id
, fatal
=fatal
, expected_type
=expected_type
)
1427 if default
is not NO_DEFAULT
:
1430 raise RegexNotFoundError('Unable to extract JSON-LD')
1432 self
.report_warning('unable to extract JSON-LD %s' % bug_reports_message())
1435 def _json_ld(self
, json_ld
, video_id
, fatal
=True, expected_type
=None):
1436 if isinstance(json_ld
, compat_str
):
1437 json_ld
= self
._parse
_json
(json_ld
, video_id
, fatal
=fatal
)
1441 if not isinstance(json_ld
, (list, tuple, dict)):
1443 if isinstance(json_ld
, dict):
1446 INTERACTION_TYPE_MAP
= {
1447 'CommentAction': 'comment',
1448 'AgreeAction': 'like',
1449 'DisagreeAction': 'dislike',
1450 'LikeAction': 'like',
1451 'DislikeAction': 'dislike',
1452 'ListenAction': 'view',
1453 'WatchAction': 'view',
1454 'ViewAction': 'view',
1457 def extract_interaction_type(e
):
1458 interaction_type
= e
.get('interactionType')
1459 if isinstance(interaction_type
, dict):
1460 interaction_type
= interaction_type
.get('@type')
1461 return str_or_none(interaction_type
)
1463 def extract_interaction_statistic(e
):
1464 interaction_statistic
= e
.get('interactionStatistic')
1465 if isinstance(interaction_statistic
, dict):
1466 interaction_statistic
= [interaction_statistic
]
1467 if not isinstance(interaction_statistic
, list):
1469 for is_e
in interaction_statistic
:
1470 if not isinstance(is_e
, dict):
1472 if is_e
.get('@type') != 'InteractionCounter':
1474 interaction_type
= extract_interaction_type(is_e
)
1475 if not interaction_type
:
1477 # For interaction count some sites provide string instead of
1478 # an integer (as per spec) with non digit characters (e.g. ",")
1479 # so extracting count with more relaxed str_to_int
1480 interaction_count
= str_to_int(is_e
.get('userInteractionCount'))
1481 if interaction_count
is None:
1483 count_kind
= INTERACTION_TYPE_MAP
.get(interaction_type
.split('/')[-1])
1486 count_key
= '%s_count' % count_kind
1487 if info
.get(count_key
) is not None:
1489 info
[count_key
] = interaction_count
1491 def extract_chapter_information(e
):
1493 'title': part
.get('name'),
1494 'start_time': part
.get('startOffset'),
1495 'end_time': part
.get('endOffset'),
1496 } for part
in variadic(e
.get('hasPart') or []) if part
.get('@type') == 'Clip']
1497 for idx
, (last_c
, current_c
, next_c
) in enumerate(zip(
1498 [{'end_time': 0}
] + chapters
, chapters
, chapters
[1:])):
1499 current_c
['end_time'] = current_c
['end_time'] or next_c
['start_time']
1500 current_c
['start_time'] = current_c
['start_time'] or last_c
['end_time']
1501 if None in current_c
.values():
1502 self
.report_warning(f
'Chapter {idx} contains broken data. Not extracting chapters')
1505 chapters
[-1]['end_time'] = chapters
[-1]['end_time'] or info
['duration']
1506 info
['chapters'] = chapters
1508 def extract_video_object(e
):
1509 assert e
['@type'] == 'VideoObject'
1510 author
= e
.get('author')
1512 'url': url_or_none(e
.get('contentUrl')),
1513 'title': unescapeHTML(e
.get('name')),
1514 'description': unescapeHTML(e
.get('description')),
1515 'thumbnails': [{'url': url_or_none(url)}
1516 for url
in variadic(traverse_obj(e
, 'thumbnailUrl', 'thumbnailURL'))],
1517 'duration': parse_duration(e
.get('duration')),
1518 'timestamp': unified_timestamp(e
.get('uploadDate')),
1519 # author can be an instance of 'Organization' or 'Person' types.
1520 # both types can have 'name' property(inherited from 'Thing' type). [1]
1521 # however some websites are using 'Text' type instead.
1522 # 1. https://schema.org/VideoObject
1523 'uploader': author
.get('name') if isinstance(author
, dict) else author
if isinstance(author
, compat_str
) else None,
1524 'filesize': float_or_none(e
.get('contentSize')),
1525 'tbr': int_or_none(e
.get('bitrate')),
1526 'width': int_or_none(e
.get('width')),
1527 'height': int_or_none(e
.get('height')),
1528 'view_count': int_or_none(e
.get('interactionCount')),
1530 extract_interaction_statistic(e
)
1531 extract_chapter_information(e
)
1533 def traverse_json_ld(json_ld
, at_top_level
=True):
1535 if at_top_level
and '@context' not in e
:
1537 if at_top_level
and set(e
.keys()) == {'@context', '@graph'}
:
1538 traverse_json_ld(variadic(e
['@graph'], allowed_types
=(dict,)), at_top_level
=False)
1540 item_type
= e
.get('@type')
1541 if expected_type
is not None and expected_type
!= item_type
:
1543 rating
= traverse_obj(e
, ('aggregateRating', 'ratingValue'), expected_type
=float_or_none
)
1544 if rating
is not None:
1545 info
['average_rating'] = rating
1546 if item_type
in ('TVEpisode', 'Episode'):
1547 episode_name
= unescapeHTML(e
.get('name'))
1549 'episode': episode_name
,
1550 'episode_number': int_or_none(e
.get('episodeNumber')),
1551 'description': unescapeHTML(e
.get('description')),
1553 if not info
.get('title') and episode_name
:
1554 info
['title'] = episode_name
1555 part_of_season
= e
.get('partOfSeason')
1556 if isinstance(part_of_season
, dict) and part_of_season
.get('@type') in ('TVSeason', 'Season', 'CreativeWorkSeason'):
1558 'season': unescapeHTML(part_of_season
.get('name')),
1559 'season_number': int_or_none(part_of_season
.get('seasonNumber')),
1561 part_of_series
= e
.get('partOfSeries') or e
.get('partOfTVSeries')
1562 if isinstance(part_of_series
, dict) and part_of_series
.get('@type') in ('TVSeries', 'Series', 'CreativeWorkSeries'):
1563 info
['series'] = unescapeHTML(part_of_series
.get('name'))
1564 elif item_type
== 'Movie':
1566 'title': unescapeHTML(e
.get('name')),
1567 'description': unescapeHTML(e
.get('description')),
1568 'duration': parse_duration(e
.get('duration')),
1569 'timestamp': unified_timestamp(e
.get('dateCreated')),
1571 elif item_type
in ('Article', 'NewsArticle'):
1573 'timestamp': parse_iso8601(e
.get('datePublished')),
1574 'title': unescapeHTML(e
.get('headline')),
1575 'description': unescapeHTML(e
.get('articleBody') or e
.get('description')),
1577 if traverse_obj(e
, ('video', 0, '@type')) == 'VideoObject':
1578 extract_video_object(e
['video'][0])
1579 elif item_type
== 'VideoObject':
1580 extract_video_object(e
)
1581 if expected_type
is None:
1585 video
= e
.get('video')
1586 if isinstance(video
, dict) and video
.get('@type') == 'VideoObject':
1587 extract_video_object(video
)
1588 if expected_type
is None:
1592 traverse_json_ld(json_ld
)
1594 return filter_dict(info
)
1596 def _search_nextjs_data(self
, webpage
, video_id
, *, transform_source
=None, fatal
=True, **kw
):
1597 return self
._parse
_json
(
1599 r
'(?s)<script[^>]+id=[\'"]__NEXT_DATA__[\'"][^
>]*>([^
<]+)</script
>',
1600 webpage, 'next
.js data
', fatal=fatal, **kw),
1601 video_id, transform_source=transform_source, fatal=fatal)
1603 def _search_nuxt_data(self, webpage, video_id, context_name='__NUXT__
'):
1604 ''' Parses Nuxt.js metadata. This works as long as the function __NUXT__ invokes is a pure function. '''
1605 # not all website do this, but it can be changed
1606 # https://stackoverflow.com/questions/67463109/how-to-change-or-hide-nuxt-and-nuxt-keyword-in-page-source
1607 rectx = re.escape(context_name)
1608 js, arg_keys, arg_vals = self._search_regex(
1609 (r'<script
>window\
.%s=\
(function\
((?P
<arg_keys
>.*?
)\
)\{return\s(?P<js>\{.*?\}
)\
}\
((?P
<arg_vals
>.+?
)\
)\
);?
</script
>' % rectx,
1610 r'%s\
(.*?\
(function\
((?P
<arg_keys
>.*?
)\
)\{return\s(?P<js>\{.*?\}
)\
}\
((?P
<arg_vals
>.*?
)\
)' % rectx),
1611 webpage, context_name, group=['js
', 'arg_keys
', 'arg_vals
'])
1613 args = dict(zip(arg_keys.split(','), arg_vals.split(',')))
1615 for key, val in args.items():
1616 if val in ('undefined
', 'void
0'):
1619 return self._parse_json(js_to_json(js, args), video_id)['data
'][0]
1622 def _hidden_inputs(html):
1623 html = re.sub(r'<!--(?
:(?
!<!--).)*-->', '', html)
1625 for input in re.findall(r'(?i
)(<input[^
>]+>)', html):
1626 attrs = extract_attributes(input)
1629 if attrs.get('type') not in ('hidden
', 'submit
'):
1631 name = attrs.get('name
') or attrs.get('id')
1632 value = attrs.get('value
')
1633 if name and value is not None:
1634 hidden_inputs[name] = value
1635 return hidden_inputs
1637 def _form_hidden_inputs(self, form_id, html):
1638 form = self._search_regex(
1639 r'(?
is)<form
[^
>]+?
id=(["\'])%s\1[^>]*>(?P<form>.+?)</form>' % form_id,
1640 html, '%s form' % form_id, group='form')
1641 return self._hidden_inputs(form)
1644 regex = r' *((?P<reverse>\+)?(?P<field>[a-zA-Z0-9_]+)((?P<separator>[~:])(?P<limit>.*?))?)? *$'
1646 default = ('hidden', 'aud_or_vid', 'hasvid', 'ie_pref', 'lang', 'quality',
1647 'res', 'fps', 'hdr:12', 'codec:vp9.2', 'size', 'br', 'asr',
1648 'proto', 'ext', 'hasaud', 'source', 'id') # These must not be aliases
1649 ytdl_default = ('hasaud', 'lang', 'quality', 'tbr', 'filesize', 'vbr',
1650 'height', 'width', 'proto', 'vext', 'abr', 'aext',
1651 'fps', 'fs_approx', 'source', 'id')
1654 'vcodec': {'type': 'ordered', 'regex': True,
1655 'order': ['av0?1', 'vp0?9.2', 'vp0?9', '[hx]265|he?vc?', '[hx]264|avc', 'vp0?8', 'mp4v|h263', 'theora', '', None, 'none']},
1656 'acodec': {'type': 'ordered', 'regex': True,
1657 'order': ['[af]lac', 'wav|aiff', 'opus', 'vorbis|ogg', 'aac', 'mp?4a?', 'mp3', 'e-?a?c-?3', 'ac-?3', 'dts', '', None, 'none']},
1658 'hdr': {'type': 'ordered', 'regex': True, 'field': 'dynamic_range',
1659 'order': ['dv', '(hdr)?12', r'(hdr)?10\+', '(hdr)?10', 'hlg', '', 'sdr', None]},
1660 'proto': {'type': 'ordered', 'regex': True, 'field': 'protocol',
1661 'order': ['(ht|f)tps', '(ht|f)tp$', 'm3u8.*', '.*dash', 'websocket_frag', 'rtmpe?', '', 'mms|rtsp', 'ws|websocket', 'f4']},
1662 'vext': {'type': 'ordered', 'field': 'video_ext',
1663 'order': ('mp4', 'webm', 'flv', '', 'none'),
1664 'order_free': ('webm', 'mp4', 'flv', '', 'none')},
1665 'aext': {'type': 'ordered', 'field': 'audio_ext',
1666 'order': ('m4a', 'aac', 'mp3', 'ogg', 'opus', 'webm', '', 'none'),
1667 'order_free': ('opus', 'ogg', 'webm', 'm4a', 'mp3', 'aac', '', 'none')},
1668 'hidden': {'visible': False, 'forced': True, 'type': 'extractor', 'max': -1000},
1669 'aud_or_vid': {'visible': False, 'forced': True, 'type': 'multiple',
1670 'field': ('vcodec', 'acodec'),
1671 'function': lambda it: int(any(v != 'none' for v in it))},
1672 'ie_pref': {'priority': True, 'type': 'extractor'},
1673 'hasvid': {'priority': True, 'field': 'vcodec', 'type': 'boolean', 'not_in_list': ('none',)},
1674 'hasaud': {'field': 'acodec', 'type': 'boolean', 'not_in_list': ('none',)},
1675 'lang': {'convert': 'float', 'field': 'language_preference', 'default': -1},
1676 'quality': {'convert': 'float', 'default': -1},
1677 'filesize': {'convert': 'bytes'},
1678 'fs_approx': {'convert': 'bytes', 'field': 'filesize_approx'},
1679 'id': {'convert': 'string', 'field': 'format_id'},
1680 'height': {'convert': 'float_none'},
1681 'width': {'convert': 'float_none'},
1682 'fps': {'convert': 'float_none'},
1683 'tbr': {'convert': 'float_none'},
1684 'vbr': {'convert': 'float_none'},
1685 'abr': {'convert': 'float_none'},
1686 'asr': {'convert': 'float_none'},
1687 'source': {'convert': 'float', 'field': 'source_preference', 'default': -1},
1689 'codec': {'type': 'combined', 'field': ('vcodec', 'acodec')},
1690 'br': {'type': 'combined', 'field': ('tbr', 'vbr', 'abr'), 'same_limit': True},
1691 'size': {'type': 'combined', 'same_limit': True, 'field': ('filesize', 'fs_approx')},
1692 'ext': {'type': 'combined', 'field': ('vext', 'aext')},
1693 'res': {'type': 'multiple', 'field': ('height', 'width'),
1694 'function': lambda it: (lambda l: min(l) if l else 0)(tuple(filter(None, it)))},
1696 # For compatibility with youtube-dl
1697 'format_id': {'type': 'alias', 'field': 'id'},
1698 'preference': {'type': 'alias', 'field': 'ie_pref'},
1699 'language_preference': {'type': 'alias', 'field': 'lang'},
1700 'source_preference': {'type': 'alias', 'field': 'source'},
1701 'protocol': {'type': 'alias', 'field': 'proto'},
1702 'filesize_approx': {'type': 'alias', 'field': 'fs_approx'},
1705 'dimension': {'type': 'alias', 'field': 'res', 'deprecated': True},
1706 'resolution': {'type': 'alias', 'field': 'res', 'deprecated': True},
1707 'extension': {'type': 'alias', 'field': 'ext', 'deprecated': True},
1708 'bitrate': {'type': 'alias', 'field': 'br', 'deprecated': True},
1709 'total_bitrate': {'type': 'alias', 'field': 'tbr', 'deprecated': True},
1710 'video_bitrate': {'type': 'alias', 'field': 'vbr', 'deprecated': True},
1711 'audio_bitrate': {'type': 'alias', 'field': 'abr', 'deprecated': True},
1712 'framerate': {'type': 'alias', 'field': 'fps', 'deprecated': True},
1713 'filesize_estimate': {'type': 'alias', 'field': 'size', 'deprecated': True},
1714 'samplerate': {'type': 'alias', 'field': 'asr', 'deprecated': True},
1715 'video_ext': {'type': 'alias', 'field': 'vext', 'deprecated': True},
1716 'audio_ext': {'type': 'alias', 'field': 'aext', 'deprecated': True},
1717 'video_codec': {'type': 'alias', 'field': 'vcodec', 'deprecated': True},
1718 'audio_codec': {'type': 'alias', 'field': 'acodec', 'deprecated': True},
1719 'video': {'type': 'alias', 'field': 'hasvid', 'deprecated': True},
1720 'has_video': {'type': 'alias', 'field': 'hasvid', 'deprecated': True},
1721 'audio': {'type': 'alias', 'field': 'hasaud', 'deprecated': True},
1722 'has_audio': {'type': 'alias', 'field': 'hasaud', 'deprecated': True},
1723 'extractor': {'type': 'alias', 'field': 'ie_pref', 'deprecated': True},
1724 'extractor_preference': {'type': 'alias', 'field': 'ie_pref', 'deprecated': True},
1727 def __init__(self, ie, field_preference):
1729 self.ydl = ie._downloader
1730 self.evaluate_params(self.ydl.params, field_preference)
1731 if ie.get_param('verbose'):
1732 self.print_verbose_info(self.ydl.write_debug)
1734 def _get_field_setting(self, field, key):
1735 if field not in self.settings:
1736 if key in ('forced', 'priority'):
1738 self.ydl.deprecation_warning(
1739 f'Using arbitrary fields ({field}) for format sorting is deprecated '
1740 'and may be removed in a future version')
1741 self.settings[field] = {}
1742 propObj = self.settings[field]
1743 if key not in propObj:
1744 type = propObj.get('type')
1746 default = 'preference' if type == 'extractor' else (field,) if type in ('combined', 'multiple') else field
1747 elif key == 'convert':
1748 default = 'order' if type == 'ordered' else 'float_string' if field else 'ignore'
1750 default = {'type': 'field', 'visible': True, 'order': [], 'not_in_list': (None,)}.get(key, None)
1751 propObj[key] = default
1754 def _resolve_field_value(self, field, value, convertNone=False):
1759 value = value.lower()
1760 conversion = self._get_field_setting(field, 'convert')
1761 if conversion == 'ignore':
1763 if conversion == 'string':
1765 elif conversion == 'float_none':
1766 return float_or_none(value)
1767 elif conversion == 'bytes':
1768 return FileDownloader.parse_bytes(value)
1769 elif conversion == 'order':
1770 order_list = (self._use_free_order and self._get_field_setting(field, 'order_free')) or self._get_field_setting(field, 'order')
1771 use_regex = self._get_field_setting(field, 'regex')
1772 list_length = len(order_list)
1773 empty_pos = order_list.index('') if '' in order_list else list_length + 1
1774 if use_regex and value is not None:
1775 for i, regex in enumerate(order_list):
1776 if regex and re.match(regex, value):
1777 return list_length - i
1778 return list_length - empty_pos # not in list
1779 else: # not regex or value = None
1780 return list_length - (order_list.index(value) if value in order_list else empty_pos)
1782 if value.isnumeric():
1785 self.settings[field]['convert'] = 'string'
1788 def evaluate_params(self, params, sort_extractor):
1789 self._use_free_order = params.get('prefer_free_formats', False)
1790 self._sort_user = params.get('format_sort', [])
1791 self._sort_extractor = sort_extractor
1793 def add_item(field, reverse, closest, limit_text):
1794 field = field.lower()
1795 if field in self._order:
1797 self._order.append(field)
1798 limit = self._resolve_field_value(field, limit_text)
1801 'closest': False if limit is None else closest,
1802 'limit_text': limit_text,
1804 if field in self.settings:
1805 self.settings[field].update(data)
1807 self.settings[field] = data
1810 tuple(field for field in self.default if self._get_field_setting(field, 'forced'))
1811 + (tuple() if params.get('format_sort_force', False)
1812 else tuple(field for field in self.default if self._get_field_setting(field, 'priority')))
1813 + tuple(self._sort_user) + tuple(sort_extractor) + self.default)
1815 for item in sort_list:
1816 match = re.match(self.regex, item)
1818 raise ExtractorError('Invalid format sort string "%s" given by extractor' % item)
1819 field = match.group('field')
1822 if self._get_field_setting(field, 'type') == 'alias':
1823 alias, field = field, self._get_field_setting(field, 'field')
1824 if self._get_field_setting(alias, 'deprecated'):
1825 self.ydl.deprecation_warning(
1826 f'Format sorting alias {alias} is deprecated '
1827 f'and may be removed in a future version. Please use {field} instead')
1828 reverse = match.group('reverse') is not None
1829 closest = match.group('separator') == '~'
1830 limit_text = match.group('limit')
1832 has_limit = limit_text is not None
1833 has_multiple_fields = self._get_field_setting(field, 'type') == 'combined'
1834 has_multiple_limits = has_limit and has_multiple_fields and not self._get_field_setting(field, 'same_limit')
1836 fields = self._get_field_setting(field, 'field') if has_multiple_fields else (field,)
1837 limits = limit_text.split(':') if has_multiple_limits else (limit_text,) if has_limit else tuple()
1838 limit_count = len(limits)
1839 for (i, f) in enumerate(fields):
1840 add_item(f, reverse, closest,
1841 limits[i] if i < limit_count
1842 else limits[0] if has_limit and not has_multiple_limits
1845 def print_verbose_info(self, write_debug):
1847 write_debug('Sort order given by user: %s' % ', '.join(self._sort_user))
1848 if self._sort_extractor:
1849 write_debug('Sort order given by extractor: %s' % ', '.join(self._sort_extractor))
1850 write_debug('Formats sorted by: %s' % ', '.join(['%s%s%s' % (
1851 '+' if self._get_field_setting(field, 'reverse') else '', field,
1852 '%s%s(%s)' % ('~' if self._get_field_setting(field, 'closest') else ':',
1853 self._get_field_setting(field, 'limit_text'),
1854 self._get_field_setting(field, 'limit'))
1855 if self._get_field_setting(field, 'limit_text') is not None else '')
1856 for field in self._order if self._get_field_setting(field, 'visible')]))
1858 def _calculate_field_preference_from_value(self, format, field, type, value):
1859 reverse = self._get_field_setting(field, 'reverse')
1860 closest = self._get_field_setting(field, 'closest')
1861 limit = self._get_field_setting(field, 'limit')
1863 if type == 'extractor':
1864 maximum = self._get_field_setting(field, 'max')
1865 if value is None or (maximum is not None and value >= maximum):
1867 elif type == 'boolean':
1868 in_list = self._get_field_setting(field, 'in_list')
1869 not_in_list = self._get_field_setting(field, 'not_in_list')
1870 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
1871 elif type == 'ordered':
1872 value = self._resolve_field_value(field, value, True)
1874 # try to convert to number
1875 val_num = float_or_none(value, default=self._get_field_setting(field, 'default'))
1876 is_num = self._get_field_setting(field, 'convert') != 'string' and val_num is not None
1880 return ((-10, 0) if value is None
1881 else (1, value, 0) if not is_num # if a field has mixed strings and numbers, strings are sorted higher
1882 else (0, -abs(value - limit), value - limit if reverse else limit - value) if closest
1883 else (0, value, 0) if not reverse and (limit is None or value <= limit)
1884 else (0, -value, 0) if limit is None or (reverse and value == limit) or value > limit
1885 else (-1, value, 0))
1887 def _calculate_field_preference(self, format, field):
1888 type = self._get_field_setting(field, 'type') # extractor, boolean, ordered, field, multiple
1889 get_value = lambda f: format.get(self._get_field_setting(f, 'field'))
1890 if type == 'multiple':
1891 type = 'field' # Only 'field' is allowed in multiple for now
1892 actual_fields = self._get_field_setting(field, 'field')
1894 value = self._get_field_setting(field, 'function')(get_value(f) for f in actual_fields)
1896 value = get_value(field)
1897 return self._calculate_field_preference_from_value(format, field, type, value)
1899 def calculate_preference(self, format):
1900 # Determine missing protocol
1901 if not format.get('protocol'):
1902 format['protocol'] = determine_protocol(format)
1904 # Determine missing ext
1905 if not format.get('ext') and 'url' in format:
1906 format['ext'] = determine_ext(format['url'])
1907 if format.get('vcodec') == 'none':
1908 format['audio_ext'] = format['ext'] if format.get('acodec') != 'none' else 'none'
1909 format['video_ext'] = 'none'
1911 format['video_ext'] = format['ext']
1912 format['audio_ext'] = 'none'
1913 # if format.get('preference') is None and format.get('ext') in ('f4f', 'f4m'): # Not supported?
1914 # format['preference'] = -1000
1916 # Determine missing bitrates
1917 if format.get('tbr') is None:
1918 if format.get('vbr') is not None and format.get('abr') is not None:
1919 format['tbr'] = format.get('vbr', 0) + format.get('abr', 0)
1921 if format.get('vcodec') != 'none' and format.get('vbr') is None:
1922 format['vbr'] = format.get('tbr') - format.get('abr', 0)
1923 if format.get('acodec') != 'none' and format.get('abr') is None:
1924 format['abr'] = format.get('tbr') - format.get('vbr', 0)
1926 return tuple(self._calculate_field_preference(format, field) for field in self._order)
1928 def _sort_formats(self, formats, field_preference=[]):
1931 format_sort = self.FormatSort(self, field_preference)
1932 formats.sort(key=lambda f: format_sort.calculate_preference(f))
1934 def _check_formats(self, formats, video_id):
1936 formats[:] = filter(
1937 lambda f: self._is_valid_url(
1939 item='%s video format' % f.get('format_id') if f.get('format_id') else 'video'),
1943 def _remove_duplicate_formats(formats):
1947 if f['url'] not in format_urls:
1948 format_urls.add(f['url'])
1949 unique_formats.append(f)
1950 formats[:] = unique_formats
1952 def _is_valid_url(self, url, video_id, item='video', headers={}):
1953 url = self._proto_relative_url(url, scheme='http:')
1954 # For now assume non HTTP(S) URLs always valid
1955 if not (url.startswith('http://') or url.startswith('https://')):
1958 self._request_webpage(url, video_id, 'Checking %s URL' % item, headers=headers)
1960 except ExtractorError as e:
1962 '%s: %s URL is invalid, skipping: %s'
1963 % (video_id, item, error_to_compat_str(e.cause)))
1966 def http_scheme(self):
1967 """ Either "http
:" or "https
:", depending on the user's preferences """
1970 if self.get_param('prefer_insecure', False)
1973 def _proto_relative_url(self, url, scheme=None):
1976 if url.startswith('//'):
1978 scheme = self.http_scheme()
1983 def _sleep(self, timeout, video_id, msg_template=None):
1984 if msg_template is None:
1985 msg_template = '%(video_id)s: Waiting for %(timeout)s seconds'
1986 msg = msg_template % {'video_id': video_id, 'timeout': timeout}
1990 def _extract_f4m_formats(self, manifest_url, video_id, preference=None, quality=None, f4m_id=None,
1991 transform_source=lambda s: fix_xml_ampersands(s).strip(),
1992 fatal=True, m3u8_id=None, data=None, headers={}, query={}):
1993 manifest = self._download_xml(
1994 manifest_url, video_id, 'Downloading f4m manifest',
1995 'Unable to download f4m manifest',
1996 # Some manifests may be malformed, e.g. prosiebensat1 generated manifests
1997 # (see https://github.com/ytdl-org/youtube-dl/issues/6215#issuecomment-121704244)
1998 transform_source=transform_source,
1999 fatal=fatal, data=data, headers=headers, query=query)
2001 if manifest is False:
2004 return self._parse_f4m_formats(
2005 manifest, manifest_url, video_id, preference=preference, quality=quality, f4m_id=f4m_id,
2006 transform_source=transform_source, fatal=fatal, m3u8_id=m3u8_id)
2008 def _parse_f4m_formats(self, manifest, manifest_url, video_id, preference=None, quality=None, f4m_id=None,
2009 transform_source=lambda s: fix_xml_ampersands(s).strip(),
2010 fatal=True, m3u8_id=None):
2011 if not isinstance(manifest, compat_etree_Element) and not fatal:
2014 # currently yt-dlp cannot decode the playerVerificationChallenge as Akamai uses Adobe Alchemy
2015 akamai_pv = manifest.find('{http://ns.adobe.com/f4m/1.0}pv-2.0')
2016 if akamai_pv is not None and ';' in akamai_pv.text:
2017 playerVerificationChallenge = akamai_pv.text.split(';')[0]
2018 if playerVerificationChallenge.strip() != '':
2022 manifest_version = '1.0'
2023 media_nodes = manifest.findall('{http://ns.adobe.com/f4m/1.0}media')
2025 manifest_version = '2.0'
2026 media_nodes = manifest.findall('{http://ns.adobe.com/f4m/2.0}media')
2027 # Remove unsupported DRM protected media from final formats
2028 # rendition (see https://github.com/ytdl-org/youtube-dl/issues/8573).
2029 media_nodes = remove_encrypted_media(media_nodes)
2033 manifest_base_url = get_base_url(manifest)
2035 bootstrap_info = xpath_element(
2036 manifest, ['{http://ns.adobe.com/f4m/1.0}bootstrapInfo', '{http://ns.adobe.com/f4m/2.0}bootstrapInfo'],
2037 'bootstrap info', default=None)
2040 mime_type = xpath_text(
2041 manifest, ['{http://ns.adobe.com/f4m/1.0}mimeType', '{http://ns.adobe.com/f4m/2.0}mimeType'],
2042 'base URL', default=None)
2043 if mime_type and mime_type.startswith('audio/'):
2046 for i, media_el in enumerate(media_nodes):
2047 tbr = int_or_none(media_el.attrib.get('bitrate'))
2048 width = int_or_none(media_el.attrib.get('width'))
2049 height = int_or_none(media_el.attrib.get('height'))
2050 format_id = join_nonempty(f4m_id, tbr or i)
2051 # If <bootstrapInfo> is present, the specified f4m is a
2052 # stream-level manifest, and only set-level manifests may refer to
2053 # external resources. See section 11.4 and section 4 of F4M spec
2054 if bootstrap_info is None:
2056 # @href is introduced in 2.0, see section 11.6 of F4M spec
2057 if manifest_version == '2.0':
2058 media_url = media_el.attrib.get('href')
2059 if media_url is None:
2060 media_url = media_el.attrib.get('url')
2064 media_url if media_url.startswith('http://') or media_url.startswith('https://')
2065 else ((manifest_base_url or '/'.join(manifest_url.split('/')[:-1])) + '/' + media_url))
2066 # If media_url is itself a f4m manifest do the recursive extraction
2067 # since bitrates in parent manifest (this one) and media_url manifest
2068 # may differ leading to inability to resolve the format by requested
2069 # bitrate in f4m downloader
2070 ext = determine_ext(manifest_url)
2072 f4m_formats = self._extract_f4m_formats(
2073 manifest_url, video_id, preference=preference, quality=quality, f4m_id=f4m_id,
2074 transform_source=transform_source, fatal=fatal)
2075 # Sometimes stream-level manifest contains single media entry that
2076 # does not contain any quality metadata (e.g. http://matchtv.ru/#live-player).
2077 # At the same time parent's media entry in set-level manifest may
2078 # contain it. We will copy it from parent in such cases.
2079 if len(f4m_formats) == 1:
2082 'tbr': f.get('tbr') or tbr,
2083 'width': f.get('width') or width,
2084 'height': f.get('height') or height,
2085 'format_id': f.get('format_id') if not tbr else format_id,
2088 formats.extend(f4m_formats)
2091 formats.extend(self._extract_m3u8_formats(
2092 manifest_url, video_id, 'mp4', preference=preference,
2093 quality=quality, m3u8_id=m3u8_id, fatal=fatal))
2096 'format_id': format_id,
2097 'url': manifest_url,
2098 'manifest_url': manifest_url,
2099 'ext': 'flv' if bootstrap_info is not None else None,
2105 'preference': preference,
2110 def _m3u8_meta_format(self, m3u8_url, ext=None, preference=None, quality=None, m3u8_id=None):
2112 'format_id': join_nonempty(m3u8_id, 'meta'),
2116 'preference': preference - 100 if preference else -100,
2118 'resolution': 'multiple',
2119 'format_note': 'Quality selection URL',
2122 def _report_ignoring_subs(self, name):
2123 self.report_warning(bug_reports_message(
2124 f'Ignoring subtitle tracks found in the {name} manifest; '
2125 'if any subtitle tracks are missing,'
2128 def _extract_m3u8_formats(self, *args, **kwargs):
2129 fmts, subs = self._extract_m3u8_formats_and_subtitles(*args, **kwargs)
2131 self._report_ignoring_subs('HLS')
2134 def _extract_m3u8_formats_and_subtitles(
2135 self, m3u8_url, video_id, ext=None, entry_protocol='m3u8_native',
2136 preference=None, quality=None, m3u8_id=None, note=None,
2137 errnote=None, fatal=True, live=False, data=None, headers={},
2140 res = self._download_webpage_handle(
2142 note='Downloading m3u8 information' if note is None else note,
2143 errnote='Failed to download m3u8 information' if errnote is None else errnote,
2144 fatal=fatal, data=data, headers=headers, query=query)
2149 m3u8_doc, urlh = res
2150 m3u8_url = urlh.geturl()
2152 return self._parse_m3u8_formats_and_subtitles(
2153 m3u8_doc, m3u8_url, ext=ext, entry_protocol=entry_protocol,
2154 preference=preference, quality=quality, m3u8_id=m3u8_id,
2155 note=note, errnote=errnote, fatal=fatal, live=live, data=data,
2156 headers=headers, query=query, video_id=video_id)
2158 def _parse_m3u8_formats_and_subtitles(
2159 self, m3u8_doc, m3u8_url=None, ext=None, entry_protocol='m3u8_native',
2160 preference=None, quality=None, m3u8_id=None, live=False, note=None,
2161 errnote=None, fatal=True, data=None, headers={}, query={},
2163 formats, subtitles = [], {}
2165 has_drm = re.search('|'.join([
2166 r'#EXT-X-FAXS-CM:', # Adobe Flash Access
2167 r'#EXT-X-(?:SESSION-)?KEY:.*?URI="skd
://', # Apple FairPlay
2170 def format_url(url):
2171 return url if re.match(r'^https?
://', url) else compat_urlparse.urljoin(m3u8_url, url)
2173 if self.get_param('hls_split_discontinuity
', False):
2174 def _extract_m3u8_playlist_indices(manifest_url=None, m3u8_doc=None):
2176 if not manifest_url:
2178 m3u8_doc = self._download_webpage(
2179 manifest_url, video_id, fatal=fatal, data=data, headers=headers,
2180 note=False, errnote='Failed to download m3u8 playlist information
')
2181 if m3u8_doc is False:
2183 return range(1 + sum(line.startswith('#EXT-X-DISCONTINUITY') for line in m3u8_doc.splitlines()))
2186 def _extract_m3u8_playlist_indices(*args
, **kwargs
):
2190 # 1. https://tools.ietf.org/html/draft-pantos-http-live-streaming-21
2191 # 2. https://github.com/ytdl-org/youtube-dl/issues/12211
2192 # 3. https://github.com/ytdl-org/youtube-dl/issues/18923
2194 # We should try extracting formats only from master playlists [1, 4.3.4],
2195 # i.e. playlists that describe available qualities. On the other hand
2196 # media playlists [1, 4.3.3] should be returned as is since they contain
2197 # just the media without qualities renditions.
2198 # Fortunately, master playlist can be easily distinguished from media
2199 # playlist based on particular tags availability. As of [1, 4.3.3, 4.3.4]
2200 # master playlist tags MUST NOT appear in a media playlist and vice versa.
2201 # As of [1, 4.3.3.1] #EXT-X-TARGETDURATION tag is REQUIRED for every
2202 # media playlist and MUST NOT appear in master playlist thus we can
2203 # clearly detect media playlist with this criterion.
2205 if '#EXT-X-TARGETDURATION' in m3u8_doc
: # media playlist, return as is
2207 'format_id': join_nonempty(m3u8_id
, idx
),
2208 'format_index': idx
,
2209 'url': m3u8_url
or encode_data_uri(m3u8_doc
.encode('utf-8'), 'application/x-mpegurl'),
2211 'protocol': entry_protocol
,
2212 'preference': preference
,
2215 } for idx
in _extract_m3u8_playlist_indices(m3u8_doc
=m3u8_doc
)]
2217 return formats
, subtitles
2220 last_stream_inf
= {}
2222 def extract_media(x_media_line
):
2223 media
= parse_m3u8_attributes(x_media_line
)
2224 # As per [1, 4.3.4.1] TYPE, GROUP-ID and NAME are REQUIRED
2225 media_type
, group_id
, name
= media
.get('TYPE'), media
.get('GROUP-ID'), media
.get('NAME')
2226 if not (media_type
and group_id
and name
):
2228 groups
.setdefault(group_id
, []).append(media
)
2229 # <https://tools.ietf.org/html/rfc8216#section-4.3.4.1>
2230 if media_type
== 'SUBTITLES':
2231 # According to RFC 8216 §4.3.4.2.1, URI is REQUIRED in the
2232 # EXT-X-MEDIA tag if the media type is SUBTITLES.
2233 # However, lack of URI has been spotted in the wild.
2234 # e.g. NebulaIE; see https://github.com/yt-dlp/yt-dlp/issues/339
2235 if not media
.get('URI'):
2237 url
= format_url(media
['URI'])
2240 'ext': determine_ext(url
),
2242 if sub_info
['ext'] == 'm3u8':
2243 # Per RFC 8216 §3.1, the only possible subtitle format m3u8
2244 # files may contain is WebVTT:
2245 # <https://tools.ietf.org/html/rfc8216#section-3.1>
2246 sub_info
['ext'] = 'vtt'
2247 sub_info
['protocol'] = 'm3u8_native'
2248 lang
= media
.get('LANGUAGE') or 'und'
2249 subtitles
.setdefault(lang
, []).append(sub_info
)
2250 if media_type
not in ('VIDEO', 'AUDIO'):
2252 media_url
= media
.get('URI')
2254 manifest_url
= format_url(media_url
)
2256 'format_id': join_nonempty(m3u8_id
, group_id
, name
, idx
),
2257 'format_note': name
,
2258 'format_index': idx
,
2259 'url': manifest_url
,
2260 'manifest_url': m3u8_url
,
2261 'language': media
.get('LANGUAGE'),
2263 'protocol': entry_protocol
,
2264 'preference': preference
,
2266 'vcodec': 'none' if media_type
== 'AUDIO' else None,
2267 } for idx
in _extract_m3u8_playlist_indices(manifest_url
))
2269 def build_stream_name():
2270 # Despite specification does not mention NAME attribute for
2271 # EXT-X-STREAM-INF tag it still sometimes may be present (see [1]
2272 # or vidio test in TestInfoExtractor.test_parse_m3u8_formats)
2273 # 1. http://www.vidio.com/watch/165683-dj_ambred-booyah-live-2015
2274 stream_name
= last_stream_inf
.get('NAME')
2277 # If there is no NAME in EXT-X-STREAM-INF it will be obtained
2278 # from corresponding rendition group
2279 stream_group_id
= last_stream_inf
.get('VIDEO')
2280 if not stream_group_id
:
2282 stream_group
= groups
.get(stream_group_id
)
2283 if not stream_group
:
2284 return stream_group_id
2285 rendition
= stream_group
[0]
2286 return rendition
.get('NAME') or stream_group_id
2288 # parse EXT-X-MEDIA tags before EXT-X-STREAM-INF in order to have the
2289 # chance to detect video only formats when EXT-X-STREAM-INF tags
2290 # precede EXT-X-MEDIA tags in HLS manifest such as [3].
2291 for line
in m3u8_doc
.splitlines():
2292 if line
.startswith('#EXT-X-MEDIA:'):
2295 for line
in m3u8_doc
.splitlines():
2296 if line
.startswith('#EXT-X-STREAM-INF:'):
2297 last_stream_inf
= parse_m3u8_attributes(line
)
2298 elif line
.startswith('#') or not line
.strip():
2301 tbr
= float_or_none(
2302 last_stream_inf
.get('AVERAGE-BANDWIDTH')
2303 or last_stream_inf
.get('BANDWIDTH'), scale
=1000)
2304 manifest_url
= format_url(line
.strip())
2306 for idx
in _extract_m3u8_playlist_indices(manifest_url
):
2307 format_id
= [m3u8_id
, None, idx
]
2308 # Bandwidth of live streams may differ over time thus making
2309 # format_id unpredictable. So it's better to keep provided
2312 stream_name
= build_stream_name()
2313 format_id
[1] = stream_name
or '%d' % (tbr
or len(formats
))
2315 'format_id': join_nonempty(*format_id
),
2316 'format_index': idx
,
2317 'url': manifest_url
,
2318 'manifest_url': m3u8_url
,
2321 'fps': float_or_none(last_stream_inf
.get('FRAME-RATE')),
2322 'protocol': entry_protocol
,
2323 'preference': preference
,
2326 resolution
= last_stream_inf
.get('RESOLUTION')
2328 mobj
= re
.search(r
'(?P<width>\d+)[xX](?P<height>\d+)', resolution
)
2330 f
['width'] = int(mobj
.group('width'))
2331 f
['height'] = int(mobj
.group('height'))
2332 # Unified Streaming Platform
2334 r
'audio.*?(?:%3D|=)(\d+)(?:-video.*?(?:%3D|=)(\d+))?', f
['url'])
2336 abr
, vbr
= mobj
.groups()
2337 abr
, vbr
= float_or_none(abr
, 1000), float_or_none(vbr
, 1000)
2342 codecs
= parse_codecs(last_stream_inf
.get('CODECS'))
2344 audio_group_id
= last_stream_inf
.get('AUDIO')
2345 # As per [1, 4.3.4.1.1] any EXT-X-STREAM-INF tag which
2346 # references a rendition group MUST have a CODECS attribute.
2347 # However, this is not always respected, for example, [2]
2348 # contains EXT-X-STREAM-INF tag which references AUDIO
2349 # rendition group but does not have CODECS and despite
2350 # referencing an audio group it represents a complete
2351 # (with audio and video) format. So, for such cases we will
2352 # ignore references to rendition groups and treat them
2353 # as complete formats.
2354 if audio_group_id
and codecs
and f
.get('vcodec') != 'none':
2355 audio_group
= groups
.get(audio_group_id
)
2356 if audio_group
and audio_group
[0].get('URI'):
2357 # TODO: update acodec for audio only formats with
2359 f
['acodec'] = 'none'
2360 if not f
.get('ext'):
2361 f
['ext'] = 'm4a' if f
.get('vcodec') == 'none' else 'mp4'
2365 progressive_uri
= last_stream_inf
.get('PROGRESSIVE-URI')
2368 del http_f
['manifest_url']
2370 'format_id': f
['format_id'].replace('hls-', 'http-'),
2372 'url': progressive_uri
,
2374 formats
.append(http_f
)
2376 last_stream_inf
= {}
2377 return formats
, subtitles
2379 def _extract_m3u8_vod_duration(
2380 self
, m3u8_vod_url
, video_id
, note
=None, errnote
=None, data
=None, headers
={}, query={}
):
2382 m3u8_vod
= self
._download
_webpage
(
2383 m3u8_vod_url
, video_id
,
2384 note
='Downloading m3u8 VOD manifest' if note
is None else note
,
2385 errnote
='Failed to download VOD manifest' if errnote
is None else errnote
,
2386 fatal
=False, data
=data
, headers
=headers
, query
=query
)
2388 return self
._parse
_m
3u8_vod
_duration
(m3u8_vod
or '', video_id
)
2390 def _parse_m3u8_vod_duration(self
, m3u8_vod
, video_id
):
2391 if '#EXT-X-PLAYLIST-TYPE:VOD' not in m3u8_vod
:
2395 float(line
[len('#EXTINF:'):].split(',')[0])
2396 for line
in m3u8_vod
.splitlines() if line
.startswith('#EXTINF:'))) or None
2399 def _xpath_ns(path
, namespace
=None):
2403 for c
in path
.split('/'):
2404 if not c
or c
== '.':
2407 out
.append('{%s}%s' % (namespace
, c
))
2408 return '/'.join(out
)
2410 def _extract_smil_formats_and_subtitles(self
, smil_url
, video_id
, fatal
=True, f4m_params
=None, transform_source
=None):
2411 smil
= self
._download
_smil
(smil_url
, video_id
, fatal
=fatal
, transform_source
=transform_source
)
2417 namespace
= self
._parse
_smil
_namespace
(smil
)
2419 fmts
= self
._parse
_smil
_formats
(
2420 smil
, smil_url
, video_id
, namespace
=namespace
, f4m_params
=f4m_params
)
2421 subs
= self
._parse
_smil
_subtitles
(
2422 smil
, namespace
=namespace
)
2426 def _extract_smil_formats(self
, *args
, **kwargs
):
2427 fmts
, subs
= self
._extract
_smil
_formats
_and
_subtitles
(*args
, **kwargs
)
2429 self
._report
_ignoring
_subs
('SMIL')
2432 def _extract_smil_info(self
, smil_url
, video_id
, fatal
=True, f4m_params
=None):
2433 smil
= self
._download
_smil
(smil_url
, video_id
, fatal
=fatal
)
2436 return self
._parse
_smil
(smil
, smil_url
, video_id
, f4m_params
=f4m_params
)
2438 def _download_smil(self
, smil_url
, video_id
, fatal
=True, transform_source
=None):
2439 return self
._download
_xml
(
2440 smil_url
, video_id
, 'Downloading SMIL file',
2441 'Unable to download SMIL file', fatal
=fatal
, transform_source
=transform_source
)
2443 def _parse_smil(self
, smil
, smil_url
, video_id
, f4m_params
=None):
2444 namespace
= self
._parse
_smil
_namespace
(smil
)
2446 formats
= self
._parse
_smil
_formats
(
2447 smil
, smil_url
, video_id
, namespace
=namespace
, f4m_params
=f4m_params
)
2448 subtitles
= self
._parse
_smil
_subtitles
(smil
, namespace
=namespace
)
2450 video_id
= os
.path
.splitext(url_basename(smil_url
))[0]
2454 for meta
in smil
.findall(self
._xpath
_ns
('./head/meta', namespace
)):
2455 name
= meta
.attrib
.get('name')
2456 content
= meta
.attrib
.get('content')
2457 if not name
or not content
:
2459 if not title
and name
== 'title':
2461 elif not description
and name
in ('description', 'abstract'):
2462 description
= content
2463 elif not upload_date
and name
== 'date':
2464 upload_date
= unified_strdate(content
)
2467 'id': image
.get('type'),
2468 'url': image
.get('src'),
2469 'width': int_or_none(image
.get('width')),
2470 'height': int_or_none(image
.get('height')),
2471 } for image
in smil
.findall(self
._xpath
_ns
('.//image', namespace
)) if image
.get('src')]
2475 'title': title
or video_id
,
2476 'description': description
,
2477 'upload_date': upload_date
,
2478 'thumbnails': thumbnails
,
2480 'subtitles': subtitles
,
2483 def _parse_smil_namespace(self
, smil
):
2484 return self
._search
_regex
(
2485 r
'(?i)^{([^}]+)?}smil$', smil
.tag
, 'namespace', default
=None)
2487 def _parse_smil_formats(self
, smil
, smil_url
, video_id
, namespace
=None, f4m_params
=None, transform_rtmp_url
=None):
2489 for meta
in smil
.findall(self
._xpath
_ns
('./head/meta', namespace
)):
2490 b
= meta
.get('base') or meta
.get('httpBase')
2502 media
= smil
.findall(self
._xpath
_ns
('.//video', namespace
)) + smil
.findall(self
._xpath
_ns
('.//audio', namespace
))
2503 for medium
in media
:
2504 src
= medium
.get('src')
2505 if not src
or src
in srcs
:
2509 bitrate
= float_or_none(medium
.get('system-bitrate') or medium
.get('systemBitrate'), 1000)
2510 filesize
= int_or_none(medium
.get('size') or medium
.get('fileSize'))
2511 width
= int_or_none(medium
.get('width'))
2512 height
= int_or_none(medium
.get('height'))
2513 proto
= medium
.get('proto')
2514 ext
= medium
.get('ext')
2515 src_ext
= determine_ext(src
)
2516 streamer
= medium
.get('streamer') or base
2518 if proto
== 'rtmp' or streamer
.startswith('rtmp'):
2524 'format_id': 'rtmp-%d' % (rtmp_count
if bitrate
is None else bitrate
),
2526 'filesize': filesize
,
2530 if transform_rtmp_url
:
2531 streamer
, src
= transform_rtmp_url(streamer
, src
)
2532 formats
[-1].update({
2538 src_url
= src
if src
.startswith('http') else compat_urlparse
.urljoin(base
, src
)
2539 src_url
= src_url
.strip()
2541 if proto
== 'm3u8' or src_ext
== 'm3u8':
2542 m3u8_formats
= self
._extract
_m
3u8_formats
(
2543 src_url
, video_id
, ext
or 'mp4', m3u8_id
='hls', fatal
=False)
2544 if len(m3u8_formats
) == 1:
2546 m3u8_formats
[0].update({
2547 'format_id': 'hls-%d' % (m3u8_count
if bitrate
is None else bitrate
),
2552 formats
.extend(m3u8_formats
)
2553 elif src_ext
== 'f4m':
2558 'plugin': 'flowplayer-3.2.0.1',
2560 f4m_url
+= '&' if '?' in f4m_url
else '?'
2561 f4m_url
+= compat_urllib_parse_urlencode(f4m_params
)
2562 formats
.extend(self
._extract
_f
4m
_formats
(f4m_url
, video_id
, f4m_id
='hds', fatal
=False))
2563 elif src_ext
== 'mpd':
2564 formats
.extend(self
._extract
_mpd
_formats
(
2565 src_url
, video_id
, mpd_id
='dash', fatal
=False))
2566 elif re
.search(r
'\.ism/[Mm]anifest', src_url
):
2567 formats
.extend(self
._extract
_ism
_formats
(
2568 src_url
, video_id
, ism_id
='mss', fatal
=False))
2569 elif src_url
.startswith('http') and self
._is
_valid
_url
(src
, video_id
):
2573 'ext': ext
or src_ext
or 'flv',
2574 'format_id': 'http-%d' % (bitrate
or http_count
),
2576 'filesize': filesize
,
2581 for medium
in smil
.findall(self
._xpath
_ns
('.//imagestream', namespace
)):
2582 src
= medium
.get('src')
2583 if not src
or src
in srcs
:
2589 'format_id': 'imagestream-%d' % (imgs_count
),
2591 'ext': mimetype2ext(medium
.get('type')),
2594 'width': int_or_none(medium
.get('width')),
2595 'height': int_or_none(medium
.get('height')),
2596 'format_note': 'SMIL storyboards',
2601 def _parse_smil_subtitles(self
, smil
, namespace
=None, subtitles_lang
='en'):
2604 for num
, textstream
in enumerate(smil
.findall(self
._xpath
_ns
('.//textstream', namespace
))):
2605 src
= textstream
.get('src')
2606 if not src
or src
in urls
:
2609 ext
= textstream
.get('ext') or mimetype2ext(textstream
.get('type')) or determine_ext(src
)
2610 lang
= textstream
.get('systemLanguage') or textstream
.get('systemLanguageName') or textstream
.get('lang') or subtitles_lang
2611 subtitles
.setdefault(lang
, []).append({
2617 def _extract_xspf_playlist(self
, xspf_url
, playlist_id
, fatal
=True):
2618 xspf
= self
._download
_xml
(
2619 xspf_url
, playlist_id
, 'Downloading xpsf playlist',
2620 'Unable to download xspf manifest', fatal
=fatal
)
2623 return self
._parse
_xspf
(
2624 xspf
, playlist_id
, xspf_url
=xspf_url
,
2625 xspf_base_url
=base_url(xspf_url
))
2627 def _parse_xspf(self
, xspf_doc
, playlist_id
, xspf_url
=None, xspf_base_url
=None):
2629 'xspf': 'http://xspf.org/ns/0/',
2630 's1': 'http://static.streamone.nl/player/ns/0',
2634 for track
in xspf_doc
.findall(xpath_with_ns('./xspf:trackList/xspf:track', NS_MAP
)):
2636 track
, xpath_with_ns('./xspf:title', NS_MAP
), 'title', default
=playlist_id
)
2637 description
= xpath_text(
2638 track
, xpath_with_ns('./xspf:annotation', NS_MAP
), 'description')
2639 thumbnail
= xpath_text(
2640 track
, xpath_with_ns('./xspf:image', NS_MAP
), 'thumbnail')
2641 duration
= float_or_none(
2642 xpath_text(track
, xpath_with_ns('./xspf:duration', NS_MAP
), 'duration'), 1000)
2645 for location
in track
.findall(xpath_with_ns('./xspf:location', NS_MAP
)):
2646 format_url
= urljoin(xspf_base_url
, location
.text
)
2651 'manifest_url': xspf_url
,
2652 'format_id': location
.get(xpath_with_ns('s1:label', NS_MAP
)),
2653 'width': int_or_none(location
.get(xpath_with_ns('s1:width', NS_MAP
))),
2654 'height': int_or_none(location
.get(xpath_with_ns('s1:height', NS_MAP
))),
2656 self
._sort
_formats
(formats
)
2661 'description': description
,
2662 'thumbnail': thumbnail
,
2663 'duration': duration
,
2668 def _extract_mpd_formats(self
, *args
, **kwargs
):
2669 fmts
, subs
= self
._extract
_mpd
_formats
_and
_subtitles
(*args
, **kwargs
)
2671 self
._report
_ignoring
_subs
('DASH')
2674 def _extract_mpd_formats_and_subtitles(
2675 self
, mpd_url
, video_id
, mpd_id
=None, note
=None, errnote
=None,
2676 fatal
=True, data
=None, headers
={}, query={}
):
2677 res
= self
._download
_xml
_handle
(
2679 note
='Downloading MPD manifest' if note
is None else note
,
2680 errnote
='Failed to download MPD manifest' if errnote
is None else errnote
,
2681 fatal
=fatal
, data
=data
, headers
=headers
, query
=query
)
2687 mpd_base_url
= base_url(urlh
.geturl())
2689 return self
._parse
_mpd
_formats
_and
_subtitles
(
2690 mpd_doc
, mpd_id
, mpd_base_url
, mpd_url
)
2692 def _parse_mpd_formats(self
, *args
, **kwargs
):
2693 fmts
, subs
= self
._parse
_mpd
_formats
_and
_subtitles
(*args
, **kwargs
)
2695 self
._report
_ignoring
_subs
('DASH')
2698 def _parse_mpd_formats_and_subtitles(
2699 self
, mpd_doc
, mpd_id
=None, mpd_base_url
='', mpd_url
=None):
2701 Parse formats from MPD manifest.
2703 1. MPEG-DASH Standard, ISO/IEC 23009-1:2014(E),
2704 http://standards.iso.org/ittf/PubliclyAvailableStandards/c065274_ISO_IEC_23009-1_2014.zip
2705 2. https://en.wikipedia.org/wiki/Dynamic_Adaptive_Streaming_over_HTTP
2707 if not self
.get_param('dynamic_mpd', True):
2708 if mpd_doc
.get('type') == 'dynamic':
2711 namespace
= self
._search
_regex
(r
'(?i)^{([^}]+)?}MPD$', mpd_doc
.tag
, 'namespace', default
=None)
2714 return self
._xpath
_ns
(path
, namespace
)
2716 def is_drm_protected(element
):
2717 return element
.find(_add_ns('ContentProtection')) is not None
2719 def extract_multisegment_info(element
, ms_parent_info
):
2720 ms_info
= ms_parent_info
.copy()
2722 # As per [1, 5.3.9.2.2] SegmentList and SegmentTemplate share some
2723 # common attributes and elements. We will only extract relevant
2725 def extract_common(source
):
2726 segment_timeline
= source
.find(_add_ns('SegmentTimeline'))
2727 if segment_timeline
is not None:
2728 s_e
= segment_timeline
.findall(_add_ns('S'))
2730 ms_info
['total_number'] = 0
2733 r
= int(s
.get('r', 0))
2734 ms_info
['total_number'] += 1 + r
2735 ms_info
['s'].append({
2736 't': int(s
.get('t', 0)),
2737 # @d is mandatory (see [1, 5.3.9.6.2, Table 17, page 60])
2738 'd': int(s
.attrib
['d']),
2741 start_number
= source
.get('startNumber')
2743 ms_info
['start_number'] = int(start_number
)
2744 timescale
= source
.get('timescale')
2746 ms_info
['timescale'] = int(timescale
)
2747 segment_duration
= source
.get('duration')
2748 if segment_duration
:
2749 ms_info
['segment_duration'] = float(segment_duration
)
2751 def extract_Initialization(source
):
2752 initialization
= source
.find(_add_ns('Initialization'))
2753 if initialization
is not None:
2754 ms_info
['initialization_url'] = initialization
.attrib
['sourceURL']
2756 segment_list
= element
.find(_add_ns('SegmentList'))
2757 if segment_list
is not None:
2758 extract_common(segment_list
)
2759 extract_Initialization(segment_list
)
2760 segment_urls_e
= segment_list
.findall(_add_ns('SegmentURL'))
2762 ms_info
['segment_urls'] = [segment
.attrib
['media'] for segment
in segment_urls_e
]
2764 segment_template
= element
.find(_add_ns('SegmentTemplate'))
2765 if segment_template
is not None:
2766 extract_common(segment_template
)
2767 media
= segment_template
.get('media')
2769 ms_info
['media'] = media
2770 initialization
= segment_template
.get('initialization')
2772 ms_info
['initialization'] = initialization
2774 extract_Initialization(segment_template
)
2777 mpd_duration
= parse_duration(mpd_doc
.get('mediaPresentationDuration'))
2778 formats
, subtitles
= [], {}
2779 stream_numbers
= collections
.defaultdict(int)
2780 for period
in mpd_doc
.findall(_add_ns('Period')):
2781 period_duration
= parse_duration(period
.get('duration')) or mpd_duration
2782 period_ms_info
= extract_multisegment_info(period
, {
2786 for adaptation_set
in period
.findall(_add_ns('AdaptationSet')):
2787 adaption_set_ms_info
= extract_multisegment_info(adaptation_set
, period_ms_info
)
2788 for representation
in adaptation_set
.findall(_add_ns('Representation')):
2789 representation_attrib
= adaptation_set
.attrib
.copy()
2790 representation_attrib
.update(representation
.attrib
)
2791 # According to [1, 5.3.7.2, Table 9, page 41], @mimeType is mandatory
2792 mime_type
= representation_attrib
['mimeType']
2793 content_type
= representation_attrib
.get('contentType', mime_type
.split('/')[0])
2795 codecs
= parse_codecs(representation_attrib
.get('codecs', ''))
2796 if content_type
not in ('video', 'audio', 'text'):
2797 if mime_type
== 'image/jpeg':
2798 content_type
= mime_type
2799 elif codecs
['vcodec'] != 'none':
2800 content_type
= 'video'
2801 elif codecs
['acodec'] != 'none':
2802 content_type
= 'audio'
2803 elif codecs
.get('tcodec', 'none') != 'none':
2804 content_type
= 'text'
2805 elif mimetype2ext(mime_type
) in ('tt', 'dfxp', 'ttml', 'xml', 'json'):
2806 content_type
= 'text'
2808 self
.report_warning('Unknown MIME type %s in DASH manifest' % mime_type
)
2812 for element
in (representation
, adaptation_set
, period
, mpd_doc
):
2813 base_url_e
= element
.find(_add_ns('BaseURL'))
2814 if base_url_e
is not None:
2815 base_url
= base_url_e
.text
+ base_url
2816 if re
.match(r
'^https?://', base_url
):
2818 if mpd_base_url
and base_url
.startswith('/'):
2819 base_url
= compat_urlparse
.urljoin(mpd_base_url
, base_url
)
2820 elif mpd_base_url
and not re
.match(r
'^https?://', base_url
):
2821 if not mpd_base_url
.endswith('/'):
2823 base_url
= mpd_base_url
+ base_url
2824 representation_id
= representation_attrib
.get('id')
2825 lang
= representation_attrib
.get('lang')
2826 url_el
= representation
.find(_add_ns('BaseURL'))
2827 filesize
= int_or_none(url_el
.attrib
.get('{http://youtube.com/yt/2012/10/10}contentLength') if url_el
is not None else None)
2828 bandwidth
= int_or_none(representation_attrib
.get('bandwidth'))
2829 if representation_id
is not None:
2830 format_id
= representation_id
2832 format_id
= content_type
2834 format_id
= mpd_id
+ '-' + format_id
2835 if content_type
in ('video', 'audio'):
2837 'format_id': format_id
,
2838 'manifest_url': mpd_url
,
2839 'ext': mimetype2ext(mime_type
),
2840 'width': int_or_none(representation_attrib
.get('width')),
2841 'height': int_or_none(representation_attrib
.get('height')),
2842 'tbr': float_or_none(bandwidth
, 1000),
2843 'asr': int_or_none(representation_attrib
.get('audioSamplingRate')),
2844 'fps': int_or_none(representation_attrib
.get('frameRate')),
2845 'language': lang
if lang
not in ('mul', 'und', 'zxx', 'mis') else None,
2846 'format_note': 'DASH %s' % content_type
,
2847 'filesize': filesize
,
2848 'container': mimetype2ext(mime_type
) + '_dash',
2851 elif content_type
== 'text':
2853 'ext': mimetype2ext(mime_type
),
2854 'manifest_url': mpd_url
,
2855 'filesize': filesize
,
2857 elif content_type
== 'image/jpeg':
2858 # See test case in VikiIE
2859 # https://www.viki.com/videos/1175236v-choosing-spouse-by-lottery-episode-1
2861 'format_id': format_id
,
2863 'manifest_url': mpd_url
,
2864 'format_note': 'DASH storyboards (jpeg)',
2868 if is_drm_protected(adaptation_set
) or is_drm_protected(representation
):
2870 representation_ms_info
= extract_multisegment_info(representation
, adaption_set_ms_info
)
2872 def prepare_template(template_name
, identifiers
):
2873 tmpl
= representation_ms_info
[template_name
]
2874 # First of, % characters outside $...$ templates
2875 # must be escaped by doubling for proper processing
2876 # by % operator string formatting used further (see
2877 # https://github.com/ytdl-org/youtube-dl/issues/16867).
2883 in_template
= not in_template
2884 elif c
== '%' and not in_template
:
2886 # Next, $...$ templates are translated to their
2887 # %(...) counterparts to be used with % operator
2888 if representation_id
is not None:
2889 t
= t
.replace('$RepresentationID$', representation_id
)
2890 t
= re
.sub(r
'\$(%s)\$' % '|'.join(identifiers
), r
'%(\1)d', t
)
2891 t
= re
.sub(r
'\$(%s)%%([^$]+)\$' % '|'.join(identifiers
), r
'%(\1)\2', t
)
2892 t
.replace('$$', '$')
2895 # @initialization is a regular template like @media one
2896 # so it should be handled just the same way (see
2897 # https://github.com/ytdl-org/youtube-dl/issues/11605)
2898 if 'initialization' in representation_ms_info
:
2899 initialization_template
= prepare_template(
2901 # As per [1, 5.3.9.4.2, Table 15, page 54] $Number$ and
2902 # $Time$ shall not be included for @initialization thus
2903 # only $Bandwidth$ remains
2905 representation_ms_info
['initialization_url'] = initialization_template
% {
2906 'Bandwidth': bandwidth
,
2909 def location_key(location
):
2910 return 'url' if re
.match(r
'^https?://', location
) else 'path'
2912 if 'segment_urls' not in representation_ms_info
and 'media' in representation_ms_info
:
2914 media_template
= prepare_template('media', ('Number', 'Bandwidth', 'Time'))
2915 media_location_key
= location_key(media_template
)
2917 # As per [1, 5.3.9.4.4, Table 16, page 55] $Number$ and $Time$
2918 # can't be used at the same time
2919 if '%(Number' in media_template
and 's' not in representation_ms_info
:
2920 segment_duration
= None
2921 if 'total_number' not in representation_ms_info
and 'segment_duration' in representation_ms_info
:
2922 segment_duration
= float_or_none(representation_ms_info
['segment_duration'], representation_ms_info
['timescale'])
2923 representation_ms_info
['total_number'] = int(math
.ceil(
2924 float_or_none(period_duration
, segment_duration
, default
=0)))
2925 representation_ms_info
['fragments'] = [{
2926 media_location_key
: media_template
% {
2927 'Number': segment_number
,
2928 'Bandwidth': bandwidth
,
2930 'duration': segment_duration
,
2931 } for segment_number
in range(
2932 representation_ms_info
['start_number'],
2933 representation_ms_info
['total_number'] + representation_ms_info
['start_number'])]
2935 # $Number*$ or $Time$ in media template with S list available
2936 # Example $Number*$: http://www.svtplay.se/klipp/9023742/stopptid-om-bjorn-borg
2937 # Example $Time$: https://play.arkena.com/embed/avp/v2/player/media/b41dda37-d8e7-4d3f-b1b5-9a9db578bdfe/1/129411
2938 representation_ms_info
['fragments'] = []
2941 segment_number
= representation_ms_info
['start_number']
2943 def add_segment_url():
2944 segment_url
= media_template
% {
2945 'Time': segment_time
,
2946 'Bandwidth': bandwidth
,
2947 'Number': segment_number
,
2949 representation_ms_info
['fragments'].append({
2950 media_location_key
: segment_url
,
2951 'duration': float_or_none(segment_d
, representation_ms_info
['timescale']),
2954 for num
, s
in enumerate(representation_ms_info
['s']):
2955 segment_time
= s
.get('t') or segment_time
2959 for r
in range(s
.get('r', 0)):
2960 segment_time
+= segment_d
2963 segment_time
+= segment_d
2964 elif 'segment_urls' in representation_ms_info
and 's' in representation_ms_info
:
2966 # Example: https://www.youtube.com/watch?v=iXZV5uAYMJI
2967 # or any YouTube dashsegments video
2970 timescale
= representation_ms_info
['timescale']
2971 for s
in representation_ms_info
['s']:
2972 duration
= float_or_none(s
['d'], timescale
)
2973 for r
in range(s
.get('r', 0) + 1):
2974 segment_uri
= representation_ms_info
['segment_urls'][segment_index
]
2976 location_key(segment_uri
): segment_uri
,
2977 'duration': duration
,
2980 representation_ms_info
['fragments'] = fragments
2981 elif 'segment_urls' in representation_ms_info
:
2982 # Segment URLs with no SegmentTimeline
2983 # Example: https://www.seznam.cz/zpravy/clanek/cesko-zasahne-vitr-o-sile-vichrice-muze-byt-i-zivotu-nebezpecny-39091
2984 # https://github.com/ytdl-org/youtube-dl/pull/14844
2986 segment_duration
= float_or_none(
2987 representation_ms_info
['segment_duration'],
2988 representation_ms_info
['timescale']) if 'segment_duration' in representation_ms_info
else None
2989 for segment_url
in representation_ms_info
['segment_urls']:
2991 location_key(segment_url
): segment_url
,
2993 if segment_duration
:
2994 fragment
['duration'] = segment_duration
2995 fragments
.append(fragment
)
2996 representation_ms_info
['fragments'] = fragments
2997 # If there is a fragments key available then we correctly recognized fragmented media.
2998 # Otherwise we will assume unfragmented media with direct access. Technically, such
2999 # assumption is not necessarily correct since we may simply have no support for
3000 # some forms of fragmented media renditions yet, but for now we'll use this fallback.
3001 if 'fragments' in representation_ms_info
:
3003 # NB: mpd_url may be empty when MPD manifest is parsed from a string
3004 'url': mpd_url
or base_url
,
3005 'fragment_base_url': base_url
,
3007 'protocol': 'http_dash_segments' if mime_type
!= 'image/jpeg' else 'mhtml',
3009 if 'initialization_url' in representation_ms_info
:
3010 initialization_url
= representation_ms_info
['initialization_url']
3011 if not f
.get('url'):
3012 f
['url'] = initialization_url
3013 f
['fragments'].append({location_key(initialization_url): initialization_url}
)
3014 f
['fragments'].extend(representation_ms_info
['fragments'])
3015 if not period_duration
:
3016 period_duration
= try_get(
3017 representation_ms_info
,
3018 lambda r
: sum(frag
['duration'] for frag
in r
['fragments']), float)
3020 # Assuming direct URL to unfragmented media.
3022 if content_type
in ('video', 'audio', 'image/jpeg'):
3023 f
['manifest_stream_number'] = stream_numbers
[f
['url']]
3024 stream_numbers
[f
['url']] += 1
3026 elif content_type
== 'text':
3027 subtitles
.setdefault(lang
or 'und', []).append(f
)
3029 return formats
, subtitles
3031 def _extract_ism_formats(self
, *args
, **kwargs
):
3032 fmts
, subs
= self
._extract
_ism
_formats
_and
_subtitles
(*args
, **kwargs
)
3034 self
._report
_ignoring
_subs
('ISM')
3037 def _extract_ism_formats_and_subtitles(self
, ism_url
, video_id
, ism_id
=None, note
=None, errnote
=None, fatal
=True, data
=None, headers
={}, query={}
):
3038 res
= self
._download
_xml
_handle
(
3040 note
='Downloading ISM manifest' if note
is None else note
,
3041 errnote
='Failed to download ISM manifest' if errnote
is None else errnote
,
3042 fatal
=fatal
, data
=data
, headers
=headers
, query
=query
)
3049 return self
._parse
_ism
_formats
_and
_subtitles
(ism_doc
, urlh
.geturl(), ism_id
)
3051 def _parse_ism_formats_and_subtitles(self
, ism_doc
, ism_url
, ism_id
=None):
3053 Parse formats from ISM manifest.
3055 1. [MS-SSTR]: Smooth Streaming Protocol,
3056 https://msdn.microsoft.com/en-us/library/ff469518.aspx
3058 if ism_doc
.get('IsLive') == 'TRUE':
3061 duration
= int(ism_doc
.attrib
['Duration'])
3062 timescale
= int_or_none(ism_doc
.get('TimeScale')) or 10000000
3066 for stream
in ism_doc
.findall('StreamIndex'):
3067 stream_type
= stream
.get('Type')
3068 if stream_type
not in ('video', 'audio', 'text'):
3070 url_pattern
= stream
.attrib
['Url']
3071 stream_timescale
= int_or_none(stream
.get('TimeScale')) or timescale
3072 stream_name
= stream
.get('Name')
3073 stream_language
= stream
.get('Language', 'und')
3074 for track
in stream
.findall('QualityLevel'):
3075 fourcc
= track
.get('FourCC') or ('AACL' if track
.get('AudioTag') == '255' else None)
3076 # TODO: add support for WVC1 and WMAP
3077 if fourcc
not in ('H264', 'AVC1', 'AACL', 'TTML'):
3078 self
.report_warning('%s is not a supported codec' % fourcc
)
3080 tbr
= int(track
.attrib
['Bitrate']) // 1000
3081 # [1] does not mention Width and Height attributes. However,
3082 # they're often present while MaxWidth and MaxHeight are
3083 # missing, so should be used as fallbacks
3084 width
= int_or_none(track
.get('MaxWidth') or track
.get('Width'))
3085 height
= int_or_none(track
.get('MaxHeight') or track
.get('Height'))
3086 sampling_rate
= int_or_none(track
.get('SamplingRate'))
3088 track_url_pattern
= re
.sub(r
'{[Bb]itrate}', track
.attrib
['Bitrate'], url_pattern
)
3089 track_url_pattern
= compat_urlparse
.urljoin(ism_url
, track_url_pattern
)
3095 stream_fragments
= stream
.findall('c')
3096 for stream_fragment_index
, stream_fragment
in enumerate(stream_fragments
):
3097 fragment_ctx
['time'] = int_or_none(stream_fragment
.get('t')) or fragment_ctx
['time']
3098 fragment_repeat
= int_or_none(stream_fragment
.get('r')) or 1
3099 fragment_ctx
['duration'] = int_or_none(stream_fragment
.get('d'))
3100 if not fragment_ctx
['duration']:
3102 next_fragment_time
= int(stream_fragment
[stream_fragment_index
+ 1].attrib
['t'])
3104 next_fragment_time
= duration
3105 fragment_ctx
['duration'] = (next_fragment_time
- fragment_ctx
['time']) / fragment_repeat
3106 for _
in range(fragment_repeat
):
3108 'url': re
.sub(r
'{start[ _]time}', compat_str(fragment_ctx
['time']), track_url_pattern
),
3109 'duration': fragment_ctx
['duration'] / stream_timescale
,
3111 fragment_ctx
['time'] += fragment_ctx
['duration']
3113 if stream_type
== 'text':
3114 subtitles
.setdefault(stream_language
, []).append({
3118 'manifest_url': ism_url
,
3119 'fragments': fragments
,
3120 '_download_params': {
3121 'stream_type': stream_type
,
3122 'duration': duration
,
3123 'timescale': stream_timescale
,
3125 'language': stream_language
,
3126 'codec_private_data': track
.get('CodecPrivateData'),
3129 elif stream_type
in ('video', 'audio'):
3131 'format_id': join_nonempty(ism_id
, stream_name
, tbr
),
3133 'manifest_url': ism_url
,
3134 'ext': 'ismv' if stream_type
== 'video' else 'isma',
3138 'asr': sampling_rate
,
3139 'vcodec': 'none' if stream_type
== 'audio' else fourcc
,
3140 'acodec': 'none' if stream_type
== 'video' else fourcc
,
3142 'fragments': fragments
,
3143 'has_drm': ism_doc
.find('Protection') is not None,
3144 '_download_params': {
3145 'stream_type': stream_type
,
3146 'duration': duration
,
3147 'timescale': stream_timescale
,
3148 'width': width
or 0,
3149 'height': height
or 0,
3151 'language': stream_language
,
3152 'codec_private_data': track
.get('CodecPrivateData'),
3153 'sampling_rate': sampling_rate
,
3154 'channels': int_or_none(track
.get('Channels', 2)),
3155 'bits_per_sample': int_or_none(track
.get('BitsPerSample', 16)),
3156 'nal_unit_length_field': int_or_none(track
.get('NALUnitLengthField', 4)),
3159 return formats
, subtitles
3161 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):
3162 def absolute_url(item_url
):
3163 return urljoin(base_url
, item_url
)
3165 def parse_content_type(content_type
):
3166 if not content_type
:
3168 ctr
= re
.search(r
'(?P<mimetype>[^/]+/[^;]+)(?:;\s*codecs="?(?P<codecs>[^"]+))?', content_type
)
3170 mimetype
, codecs
= ctr
.groups()
3171 f
= parse_codecs(codecs
)
3172 f
['ext'] = mimetype2ext(mimetype
)
3176 def _media_formats(src
, cur_media_type
, type_info
={}):
3177 full_url
= absolute_url(src
)
3178 ext
= type_info
.get('ext') or determine_ext(full_url
)
3180 is_plain_url
= False
3181 formats
= self
._extract
_m
3u8_formats
(
3182 full_url
, video_id
, ext
='mp4',
3183 entry_protocol
=m3u8_entry_protocol
, m3u8_id
=m3u8_id
,
3184 preference
=preference
, quality
=quality
, fatal
=False)
3186 is_plain_url
= False
3187 formats
= self
._extract
_mpd
_formats
(
3188 full_url
, video_id
, mpd_id
=mpd_id
, fatal
=False)
3193 'vcodec': 'none' if cur_media_type
== 'audio' else None,
3195 return is_plain_url
, formats
3198 # amp-video and amp-audio are very similar to their HTML5 counterparts
3199 # so we wll include them right here (see
3200 # https://www.ampproject.org/docs/reference/components/amp-video)
3201 # For dl8-* tags see https://delight-vr.com/documentation/dl8-video/
3202 _MEDIA_TAG_NAME_RE
= r
'(?:(?:amp|dl8(?:-live)?)-)?(video|audio)'
3203 media_tags
= [(media_tag
, media_tag_name
, media_type
, '')
3204 for media_tag
, media_tag_name
, media_type
3205 in re
.findall(r
'(?s)(<(%s)[^>]*/>)' % _MEDIA_TAG_NAME_RE
, webpage
)]
3206 media_tags
.extend(re
.findall(
3207 # We only allow video|audio followed by a whitespace or '>'.
3208 # Allowing more characters may end up in significant slow down (see
3209 # https://github.com/ytdl-org/youtube-dl/issues/11979, example URL:
3210 # http://www.porntrex.com/maps/videositemap.xml).
3211 r
'(?s)(<(?P<tag>%s)(?:\s+[^>]*)?>)(.*?)</(?P=tag)>' % _MEDIA_TAG_NAME_RE
, webpage
))
3212 for media_tag
, _
, media_type
, media_content
in media_tags
:
3217 media_attributes
= extract_attributes(media_tag
)
3218 src
= strip_or_none(media_attributes
.get('src'))
3220 _
, formats
= _media_formats(src
, media_type
)
3221 media_info
['formats'].extend(formats
)
3222 media_info
['thumbnail'] = absolute_url(media_attributes
.get('poster'))
3224 for source_tag
in re
.findall(r
'<source[^>]+>', media_content
):
3225 s_attr
= extract_attributes(source_tag
)
3226 # data-video-src and data-src are non standard but seen
3227 # several times in the wild
3228 src
= strip_or_none(dict_get(s_attr
, ('src', 'data-video-src', 'data-src')))
3231 f
= parse_content_type(s_attr
.get('type'))
3232 is_plain_url
, formats
= _media_formats(src
, media_type
, f
)
3234 # width, height, res, label and title attributes are
3235 # all not standard but seen several times in the wild
3238 for lbl
in ('label', 'title')
3239 if str_or_none(s_attr
.get(lbl
))
3241 width
= int_or_none(s_attr
.get('width'))
3242 height
= (int_or_none(s_attr
.get('height'))
3243 or int_or_none(s_attr
.get('res')))
3244 if not width
or not height
:
3246 resolution
= parse_resolution(lbl
)
3249 width
= width
or resolution
.get('width')
3250 height
= height
or resolution
.get('height')
3252 tbr
= parse_bitrate(lbl
)
3261 'format_id': s_attr
.get('label') or s_attr
.get('title'),
3263 f
.update(formats
[0])
3264 media_info
['formats'].append(f
)
3266 media_info
['formats'].extend(formats
)
3267 for track_tag
in re
.findall(r
'<track[^>]+>', media_content
):
3268 track_attributes
= extract_attributes(track_tag
)
3269 kind
= track_attributes
.get('kind')
3270 if not kind
or kind
in ('subtitles', 'captions'):
3271 src
= strip_or_none(track_attributes
.get('src'))
3274 lang
= track_attributes
.get('srclang') or track_attributes
.get('lang') or track_attributes
.get('label')
3275 media_info
['subtitles'].setdefault(lang
, []).append({
3276 'url': absolute_url(src
),
3278 for f
in media_info
['formats']:
3279 f
.setdefault('http_headers', {})['Referer'] = base_url
3280 if media_info
['formats'] or media_info
['subtitles']:
3281 entries
.append(media_info
)
3284 def _extract_akamai_formats(self
, *args
, **kwargs
):
3285 fmts
, subs
= self
._extract
_akamai
_formats
_and
_subtitles
(*args
, **kwargs
)
3287 self
._report
_ignoring
_subs
('akamai')
3290 def _extract_akamai_formats_and_subtitles(self
, manifest_url
, video_id
, hosts
={}):
3291 signed
= 'hdnea=' in manifest_url
3293 # https://learn.akamai.com/en-us/webhelp/media-services-on-demand/stream-packaging-user-guide/GUID-BE6C0F73-1E06-483B-B0EA-57984B91B7F9.html
3294 manifest_url
= re
.sub(
3295 r
'(?:b=[\d,-]+|(?:__a__|attributes)=off|__b__=\d+)&?',
3296 '', manifest_url
).strip('?')
3301 hdcore_sign
= 'hdcore=3.7.0'
3302 f4m_url
= re
.sub(r
'(https?://[^/]+)/i/', r
'\1/z/', manifest_url
).replace('/master.m3u8', '/manifest.f4m')
3303 hds_host
= hosts
.get('hds')
3305 f4m_url
= re
.sub(r
'(https?://)[^/]+', r
'\1' + hds_host
, f4m_url
)
3306 if 'hdcore=' not in f4m_url
:
3307 f4m_url
+= ('&' if '?' in f4m_url
else '?') + hdcore_sign
3308 f4m_formats
= self
._extract
_f
4m
_formats
(
3309 f4m_url
, video_id
, f4m_id
='hds', fatal
=False)
3310 for entry
in f4m_formats
:
3311 entry
.update({'extra_param_to_segment_url': hdcore_sign}
)
3312 formats
.extend(f4m_formats
)
3314 m3u8_url
= re
.sub(r
'(https?://[^/]+)/z/', r
'\1/i/', manifest_url
).replace('/manifest.f4m', '/master.m3u8')
3315 hls_host
= hosts
.get('hls')
3317 m3u8_url
= re
.sub(r
'(https?://)[^/]+', r
'\1' + hls_host
, m3u8_url
)
3318 m3u8_formats
, m3u8_subtitles
= self
._extract
_m
3u8_formats
_and
_subtitles
(
3319 m3u8_url
, video_id
, 'mp4', 'm3u8_native',
3320 m3u8_id
='hls', fatal
=False)
3321 formats
.extend(m3u8_formats
)
3322 subtitles
= self
._merge
_subtitles
(subtitles
, m3u8_subtitles
)
3324 http_host
= hosts
.get('http')
3325 if http_host
and m3u8_formats
and not signed
:
3326 REPL_REGEX
= r
'https?://[^/]+/i/([^,]+),([^/]+),([^/]+)\.csmil/.+'
3327 qualities
= re
.match(REPL_REGEX
, m3u8_url
).group(2).split(',')
3328 qualities_length
= len(qualities
)
3329 if len(m3u8_formats
) in (qualities_length
, qualities_length
+ 1):
3331 for f
in m3u8_formats
:
3332 if f
['vcodec'] != 'none':
3333 for protocol
in ('http', 'https'):
3335 del http_f
['manifest_url']
3337 REPL_REGEX
, protocol
+ r
'://%s/\g<1>%s\3' % (http_host
, qualities
[i
]), f
['url'])
3339 'format_id': http_f
['format_id'].replace('hls-', protocol
+ '-'),
3341 'protocol': protocol
,
3343 formats
.append(http_f
)
3346 return formats
, subtitles
3348 def _extract_wowza_formats(self
, url
, video_id
, m3u8_entry_protocol
='m3u8_native', skip_protocols
=[]):
3349 query
= compat_urlparse
.urlparse(url
).query
3350 url
= re
.sub(r
'/(?:manifest|playlist|jwplayer)\.(?:m3u8|f4m|mpd|smil)', '', url
)
3352 r
'(?:(?:http|rtmp|rtsp)(?P<s>s)?:)?(?P<url>//[^?]+)', url
)
3353 url_base
= mobj
.group('url')
3354 http_base_url
= '%s%s:%s' % ('http', mobj
.group('s') or '', url_base
)
3357 def manifest_url(manifest
):
3358 m_url
= '%s/%s' % (http_base_url
, manifest
)
3360 m_url
+= '?%s' % query
3363 if 'm3u8' not in skip_protocols
:
3364 formats
.extend(self
._extract
_m
3u8_formats
(
3365 manifest_url('playlist.m3u8'), video_id
, 'mp4',
3366 m3u8_entry_protocol
, m3u8_id
='hls', fatal
=False))
3367 if 'f4m' not in skip_protocols
:
3368 formats
.extend(self
._extract
_f
4m
_formats
(
3369 manifest_url('manifest.f4m'),
3370 video_id
, f4m_id
='hds', fatal
=False))
3371 if 'dash' not in skip_protocols
:
3372 formats
.extend(self
._extract
_mpd
_formats
(
3373 manifest_url('manifest.mpd'),
3374 video_id
, mpd_id
='dash', fatal
=False))
3375 if re
.search(r
'(?:/smil:|\.smil)', url_base
):
3376 if 'smil' not in skip_protocols
:
3377 rtmp_formats
= self
._extract
_smil
_formats
(
3378 manifest_url('jwplayer.smil'),
3379 video_id
, fatal
=False)
3380 for rtmp_format
in rtmp_formats
:
3381 rtsp_format
= rtmp_format
.copy()
3382 rtsp_format
['url'] = '%s/%s' % (rtmp_format
['url'], rtmp_format
['play_path'])
3383 del rtsp_format
['play_path']
3384 del rtsp_format
['ext']
3385 rtsp_format
.update({
3386 'url': rtsp_format
['url'].replace('rtmp://', 'rtsp://'),
3387 'format_id': rtmp_format
['format_id'].replace('rtmp', 'rtsp'),
3390 formats
.extend([rtmp_format
, rtsp_format
])
3392 for protocol
in ('rtmp', 'rtsp'):
3393 if protocol
not in skip_protocols
:
3395 'url': '%s:%s' % (protocol
, url_base
),
3396 'format_id': protocol
,
3397 'protocol': protocol
,
3401 def _find_jwplayer_data(self
, webpage
, video_id
=None, transform_source
=js_to_json
):
3403 r
'(?s)jwplayer\((?P<quote>[\'"])[^\'" ]+(?P
=quote
)\
)(?
!</script
>).*?\
.setup\s
*\
((?P
<options
>[^
)]+)\
)',
3407 jwplayer_data = self._parse_json(mobj.group('options
'),
3409 transform_source=transform_source)
3410 except ExtractorError:
3413 if isinstance(jwplayer_data, dict):
3414 return jwplayer_data
3416 def _extract_jwplayer_data(self, webpage, video_id, *args, **kwargs):
3417 jwplayer_data = self._find_jwplayer_data(
3418 webpage, video_id, transform_source=js_to_json)
3419 return self._parse_jwplayer_data(
3420 jwplayer_data, video_id, *args, **kwargs)
3422 def _parse_jwplayer_data(self, jwplayer_data, video_id=None, require_title=True,
3423 m3u8_id=None, mpd_id=None, rtmp_params=None, base_url=None):
3424 # JWPlayer backward compatibility: flattened playlists
3425 # https://github.com/jwplayer/jwplayer/blob/v7.4.3/src/js/api/config.js#L81-L96
3426 if 'playlist
' not in jwplayer_data:
3427 jwplayer_data = {'playlist': [jwplayer_data]}
3431 # JWPlayer backward compatibility: single playlist item
3432 # https://github.com/jwplayer/jwplayer/blob/v7.7.0/src/js/playlist/playlist.js#L10
3433 if not isinstance(jwplayer_data['playlist
'], list):
3434 jwplayer_data['playlist
'] = [jwplayer_data['playlist
']]
3436 for video_data in jwplayer_data['playlist
']:
3437 # JWPlayer backward compatibility: flattened sources
3438 # https://github.com/jwplayer/jwplayer/blob/v7.4.3/src/js/playlist/item.js#L29-L35
3439 if 'sources
' not in video_data:
3440 video_data['sources
'] = [video_data]
3442 this_video_id = video_id or video_data['mediaid
']
3444 formats = self._parse_jwplayer_formats(
3445 video_data['sources
'], video_id=this_video_id, m3u8_id=m3u8_id,
3446 mpd_id=mpd_id, rtmp_params=rtmp_params, base_url=base_url)
3449 tracks = video_data.get('tracks
')
3450 if tracks and isinstance(tracks, list):
3451 for track in tracks:
3452 if not isinstance(track, dict):
3454 track_kind = track.get('kind
')
3455 if not track_kind or not isinstance(track_kind, compat_str):
3457 if track_kind.lower() not in ('captions
', 'subtitles
'):
3459 track_url = urljoin(base_url, track.get('file'))
3462 subtitles.setdefault(track.get('label
') or 'en
', []).append({
3463 'url
': self._proto_relative_url(track_url)
3467 'id': this_video_id,
3468 'title
': unescapeHTML(video_data['title
'] if require_title else video_data.get('title
')),
3469 'description
': clean_html(video_data.get('description
')),
3470 'thumbnail
': urljoin(base_url, self._proto_relative_url(video_data.get('image
'))),
3471 'timestamp
': int_or_none(video_data.get('pubdate
')),
3472 'duration
': float_or_none(jwplayer_data.get('duration
') or video_data.get('duration
')),
3473 'subtitles
': subtitles,
3475 # https://github.com/jwplayer/jwplayer/blob/master/src/js/utils/validator.js#L32
3476 if len(formats) == 1 and re.search(r'^
(?
:http|
//).*(?
:youtube\
.com|youtu\
.be
)/.+', formats[0]['url
']):
3478 '_type
': 'url_transparent
',
3479 'url
': formats[0]['url
'],
3482 self._sort_formats(formats)
3483 entry['formats
'] = formats
3484 entries.append(entry)
3485 if len(entries) == 1:
3488 return self.playlist_result(entries)
3490 def _parse_jwplayer_formats(self, jwplayer_sources_data, video_id=None,
3491 m3u8_id=None, mpd_id=None, rtmp_params=None, base_url=None):
3494 for source in jwplayer_sources_data:
3495 if not isinstance(source, dict):
3497 source_url = urljoin(
3498 base_url, self._proto_relative_url(source.get('file')))
3499 if not source_url or source_url in urls:
3501 urls.append(source_url)
3502 source_type = source.get('type') or ''
3503 ext = mimetype2ext(source_type) or determine_ext(source_url)
3504 if source_type == 'hls
' or ext == 'm3u8
':
3505 formats.extend(self._extract_m3u8_formats(
3506 source_url, video_id, 'mp4
', entry_protocol='m3u8_native
',
3507 m3u8_id=m3u8_id, fatal=False))
3508 elif source_type == 'dash
' or ext == 'mpd
':
3509 formats.extend(self._extract_mpd_formats(
3510 source_url, video_id, mpd_id=mpd_id, fatal=False))
3512 formats.extend(self._extract_smil_formats(
3513 source_url, video_id, fatal=False))
3514 # https://github.com/jwplayer/jwplayer/blob/master/src/js/providers/default.js#L67
3515 elif source_type.startswith('audio
') or ext in (
3516 'oga
', 'aac
', 'mp3
', 'mpeg
', 'vorbis
'):
3523 height = int_or_none(source.get('height
'))
3525 # Often no height is provided but there is a label in
3526 # format like "1080p", "720p SD", or 1080.
3527 height = int_or_none(self._search_regex(
3528 r'^
(\d{3,4}
)[pP
]?
(?
:\b|$
)', compat_str(source.get('label
') or ''),
3529 'height
', default=None))
3532 'width
': int_or_none(source.get('width
')),
3534 'tbr
': int_or_none(source.get('bitrate
')),
3537 if source_url.startswith('rtmp
'):
3538 a_format['ext
'] = 'flv
'
3539 # See com/longtailvideo/jwplayer/media/RTMPMediaProvider.as
3540 # of jwplayer.flash.swf
3541 rtmp_url_parts = re.split(
3542 r'((?
:mp4|mp3|flv
):)', source_url, 1)
3543 if len(rtmp_url_parts) == 3:
3544 rtmp_url, prefix, play_path = rtmp_url_parts
3547 'play_path
': prefix + play_path,
3550 a_format.update(rtmp_params)
3551 formats.append(a_format)
3554 def _live_title(self, name):
3555 self._downloader.deprecation_warning('yt_dlp
.InfoExtractor
._live
_title
is deprecated
and does
not work
as expected
')
3558 def _int(self, v, name, fatal=False, **kwargs):
3559 res = int_or_none(v, **kwargs)
3561 msg = 'Failed to extract
%s: Could
not parse value
%r' % (name, v)
3563 raise ExtractorError(msg)
3565 self.report_warning(msg)
3568 def _float(self, v, name, fatal=False, **kwargs):
3569 res = float_or_none(v, **kwargs)
3571 msg = 'Failed to extract
%s: Could
not parse value
%r' % (name, v)
3573 raise ExtractorError(msg)
3575 self.report_warning(msg)
3578 def _set_cookie(self, domain, name, value, expire_time=None, port=None,
3579 path='/', secure=False, discard=False, rest={}, **kwargs):
3580 cookie = compat_cookiejar_Cookie(
3581 0, name, value, port, port is not None, domain, True,
3582 domain.startswith('.'), path, True, secure, expire_time,
3583 discard, None, None, rest)
3584 self._downloader.cookiejar.set_cookie(cookie)
3586 def _get_cookies(self, url):
3587 """ Return a compat_cookies_SimpleCookie with the cookies for the url """
3588 req = sanitized_Request(url)
3589 self._downloader.cookiejar.add_cookie_header(req)
3590 return compat_cookies_SimpleCookie(req.get_header('Cookie
'))
3592 def _apply_first_set_cookie_header(self, url_handle, cookie):
3594 Apply first Set-Cookie header instead of the last. Experimental.
3596 Some sites (e.g. [1-3]) may serve two cookies under the same name
3597 in Set-Cookie header and expect the first (old) one to be set rather
3598 than second (new). However, as of RFC6265 the newer one cookie
3599 should be set into cookie store what actually happens.
3600 We will workaround this issue by resetting the cookie to
3601 the first one manually.
3602 1. https://new.vk.com/
3603 2. https://github.com/ytdl-org/youtube-dl/issues/9841#issuecomment-227871201
3604 3. https://learning.oreilly.com/
3606 for header, cookies in url_handle.headers.items():
3607 if header.lower() != 'set-cookie
':
3609 if sys.version_info[0] >= 3:
3610 cookies = cookies.encode('iso
-8859-1')
3611 cookies = cookies.decode('utf
-8')
3612 cookie_value = re.search(
3613 r'%s=(.+?
);.*?
\b[Dd
]omain
=(.+?
)(?
:[,;]|$
)' % cookie, cookies)
3615 value, domain = cookie_value.groups()
3616 self._set_cookie(domain, cookie, value)
3619 def get_testcases(self, include_onlymatching=False):
3620 t = getattr(self, '_TEST
', None)
3622 assert not hasattr(self, '_TESTS
'), \
3623 '%s has _TEST
and _TESTS
' % type(self).__name__
3626 tests = getattr(self, '_TESTS
', [])
3628 if not include_onlymatching and t.get('only_matching
', False):
3630 t['name
'] = type(self).__name__[:-len('IE
')]
3633 def is_suitable(self, age_limit):
3634 """ Test whether the extractor is generally suitable for the given
3635 age limit (i.e. pornographic sites are not, all others usually are) """
3637 any_restricted = False
3638 for tc in self.get_testcases(include_onlymatching=False):
3639 if tc.get('playlist
', []):
3640 tc = tc['playlist
'][0]
3641 is_restricted = age_restricted(
3642 tc.get('info_dict
', {}).get('age_limit
'), age_limit)
3643 if not is_restricted:
3645 any_restricted = any_restricted or is_restricted
3646 return not any_restricted
3648 def extract_subtitles(self, *args, **kwargs):
3649 if (self.get_param('writesubtitles
', False)
3650 or self.get_param('listsubtitles
')):
3651 return self._get_subtitles(*args, **kwargs)
3654 def _get_subtitles(self, *args, **kwargs):
3655 raise NotImplementedError('This method must be implemented by subclasses
')
3657 def extract_comments(self, *args, **kwargs):
3658 if not self.get_param('getcomments
'):
3660 generator = self._get_comments(*args, **kwargs)
3667 comments.append(next(generator))
3668 except StopIteration:
3670 except KeyboardInterrupt:
3671 self.to_screen('Interrupted by user
')
3672 except Exception as e:
3673 if self.get_param('ignoreerrors
') is not True:
3675 self._downloader.report_error(e)
3676 comment_count = len(comments)
3677 self.to_screen(f'Extracted {comment_count} comments
')
3679 'comments
': comments,
3680 'comment_count
': None if interrupted else comment_count
3684 def _get_comments(self, *args, **kwargs):
3685 raise NotImplementedError('This method must be implemented by subclasses
')
3688 def _merge_subtitle_items(subtitle_list1, subtitle_list2):
3689 """ Merge subtitle items for one language. Items with duplicated URLs/data
3690 will be dropped. """
3691 list1_data = set((item.get('url
'), item.get('data
')) for item in subtitle_list1)
3692 ret = list(subtitle_list1)
3693 ret.extend(item for item in subtitle_list2 if (item.get('url
'), item.get('data
')) not in list1_data)
3697 def _merge_subtitles(cls, *dicts, target=None):
3698 """ Merge subtitle dictionaries, language by language. """
3702 for lang, subs in d.items():
3703 target[lang] = cls._merge_subtitle_items(target.get(lang, []), subs)
3706 def extract_automatic_captions(self, *args, **kwargs):
3707 if (self.get_param('writeautomaticsub
', False)
3708 or self.get_param('listsubtitles
')):
3709 return self._get_automatic_captions(*args, **kwargs)
3712 def _get_automatic_captions(self, *args, **kwargs):
3713 raise NotImplementedError('This method must be implemented by subclasses
')
3715 def mark_watched(self, *args, **kwargs):
3716 if not self.get_param('mark_watched
', False):
3718 if (self.supports_login() and self._get_login_info()[0] is not None
3719 or self.get_param('cookiefile
') or self.get_param('cookiesfrombrowser
')):
3720 self._mark_watched(*args, **kwargs)
3722 def _mark_watched(self, *args, **kwargs):
3723 raise NotImplementedError('This method must be implemented by subclasses
')
3725 def geo_verification_headers(self):
3727 geo_verification_proxy = self.get_param('geo_verification_proxy
')
3728 if geo_verification_proxy:
3729 headers['Ytdl
-request
-proxy
'] = geo_verification_proxy
3732 def _generic_id(self, url):
3733 return compat_urllib_parse_unquote(os.path.splitext(url.rstrip('/').split('/')[-1])[0])
3735 def _generic_title(self, url):
3736 return compat_urllib_parse_unquote(os.path.splitext(url_basename(url))[0])
3739 def _availability(is_private=None, needs_premium=None, needs_subscription=None, needs_auth=None, is_unlisted=None):
3740 all_known = all(map(
3741 lambda x: x is not None,
3742 (is_private, needs_premium, needs_subscription, needs_auth, is_unlisted)))
3744 'private
' if is_private
3745 else 'premium_only
' if needs_premium
3746 else 'subscriber_only
' if needs_subscription
3747 else 'needs_auth
' if needs_auth
3748 else 'unlisted
' if is_unlisted
3749 else 'public
' if all_known
3752 def _configuration_arg(self, key, default=NO_DEFAULT, *, ie_key=None, casesense=False):
3754 @returns A list of values for the extractor argument given by "key"
3755 or "default" if no such key is present
3756 @param default The default value to return when the key is not present (default: [])
3757 @param casesense When false, the values are converted to lower case
3760 self._downloader.params, ('extractor_args
', (ie_key or self.ie_key()).lower(), key))
3762 return [] if default is NO_DEFAULT else default
3763 return list(val) if casesense else [x.lower() for x in val]
3765 def _yes_playlist(self, playlist_id, video_id, smuggled_data=None, *, playlist_label='playlist
', video_label='video
'):
3766 if not playlist_id or not video_id:
3769 no_playlist = (smuggled_data or {}).get('force_noplaylist
')
3770 if no_playlist is not None:
3771 return not no_playlist
3773 video_id = '' if video_id is True else f' {video_id}
'
3774 playlist_id = '' if playlist_id is True else f' {playlist_id}
'
3775 if self.get_param('noplaylist
'):
3776 self.to_screen(f'Downloading just the {video_label}{video_id} because of
--no
-playlist
')
3778 self.to_screen(f'Downloading {playlist_label}{playlist_id}
- add
--no
-playlist to download just the {video_label}{video_id}
')
3782 class SearchInfoExtractor(InfoExtractor):
3784 Base class for paged search queries extractors.
3785 They accept URLs in the format _SEARCH_KEY(|all|[0-9]):{query}
3786 Instances should define _SEARCH_KEY and optionally _MAX_RESULTS
3789 _MAX_RESULTS = float('inf
')
3792 def _make_valid_url(cls):
3793 return r'%s(?P
<prefix
>|
[1-9][0-9]*|all
):(?P
<query
>[\s\S
]+)' % cls._SEARCH_KEY
3795 def _real_extract(self, query):
3796 prefix, query = self._match_valid_url(query).group('prefix
', 'query
')
3798 return self._get_n_results(query, 1)
3799 elif prefix == 'all
':
3800 return self._get_n_results(query, self._MAX_RESULTS)
3804 raise ExtractorError('invalid download number
%s for query
"%s"' % (n, query))
3805 elif n > self._MAX_RESULTS:
3806 self.report_warning('%s returns
max %i results (you requested
%i)' % (self._SEARCH_KEY, self._MAX_RESULTS, n))
3807 n = self._MAX_RESULTS
3808 return self._get_n_results(query, n)
3810 def _get_n_results(self, query, n):
3811 """Get a specified number of results for a query.
3812 Either this function or _search_results must be overridden by subclasses """
3813 return self.playlist_result(
3814 itertools.islice(self._search_results(query), 0, None if n == float('inf
') else n),
3817 def _search_results(self, query):
3818 """Returns an iterator of search results"""
3819 raise NotImplementedError('This method must be implemented by subclasses
')
3822 def SEARCH_KEY(self):
3823 return self._SEARCH_KEY