2 from __future__
import unicode_literals
16 from ..compat
import (
17 compat_cookiejar_Cookie
,
18 compat_cookies_SimpleCookie
,
20 compat_etree_fromstring
,
26 compat_urllib_parse_unquote
,
27 compat_urllib_parse_urlencode
,
28 compat_urllib_request
,
30 compat_xml_parse_error
,
32 from ..downloader
import FileDownloader
33 from ..downloader
.f4m
import (
35 remove_encrypted_media
,
64 parse_m3u8_attributes
,
88 class InfoExtractor(object):
89 """Information Extractor class.
91 Information extractors are the classes that, given a URL, extract
92 information about the video (or videos) the URL refers to. This
93 information includes the real video URL, the video title, author and
94 others. The information is stored in a dictionary which is then
95 passed to the YoutubeDL. The YoutubeDL processes this
96 information possibly downloading the video to the file system, among
97 other possible outcomes.
99 The type field determines the type of the result.
100 By far the most common value (and the default if _type is missing) is
101 "video", which indicates a single video.
103 For a video, the dictionaries must include the following fields:
105 id: Video identifier.
106 title: Video title, unescaped.
108 Additionally, it must contain either a formats entry or a url one:
110 formats: A list of dictionaries for each format available, ordered
111 from worst to best quality.
114 * url The mandatory URL representing the media:
115 for plain file media - HTTP URL of this file,
117 for HLS - URL of the M3U8 media playlist,
118 for HDS - URL of the F4M manifest,
120 - HTTP URL to plain file media (in case of
122 - URL of the MPD manifest or base URL
123 representing the media if MPD manifest
124 is parsed from a string (in case of
126 for MSS - URL of the ISM manifest.
128 The URL of the manifest file in case of
130 for HLS - URL of the M3U8 master playlist,
131 for HDS - URL of the F4M manifest,
132 for DASH - URL of the MPD manifest,
133 for MSS - URL of the ISM manifest.
134 * ext Will be calculated from URL if missing
135 * format A human-readable description of the format
136 ("mp4 container with h264/opus").
137 Calculated from the format_id, width, height.
138 and format_note fields if missing.
139 * format_id A short description of the format
140 ("mp4_h264_opus" or "19").
141 Technically optional, but strongly recommended.
142 * format_note Additional info about the format
143 ("3D" or "DASH video")
144 * width Width of the video, if known
145 * height Height of the video, if known
146 * resolution Textual description of width and height
147 * tbr Average bitrate of audio and video in KBit/s
148 * abr Average audio bitrate in KBit/s
149 * acodec Name of the audio codec in use
150 * asr Audio sampling rate in Hertz
151 * vbr Average video bitrate in KBit/s
153 * vcodec Name of the video codec in use
154 * container Name of the container format
155 * filesize The number of bytes, if known in advance
156 * filesize_approx An estimate for the number of bytes
157 * player_url SWF Player URL (used for rtmpdump).
158 * protocol The protocol that will be used for the actual
159 download, lower-case.
160 "http", "https", "rtsp", "rtmp", "rtmp_ffmpeg", "rtmpe",
161 "m3u8", "m3u8_native" or "http_dash_segments".
163 Base URL for fragments. Each fragment's path
164 value (if present) will be relative to
166 * fragments A list of fragments of a fragmented media.
167 Each fragment entry must contain either an url
168 or a path. If an url is present it should be
169 considered by a client. Otherwise both path and
170 fragment_base_url must be present. Here is
171 the list of all potential fields:
172 * "url" - fragment's URL
173 * "path" - fragment's path relative to
175 * "duration" (optional, int or float)
176 * "filesize" (optional, int)
177 * preference Order number of this format. If this field is
178 present and not None, the formats get sorted
179 by this field, regardless of all other values.
180 -1 for default (order by other properties),
181 -2 or smaller for less than default.
182 < -1000 to hide the format (if there is
183 another one which is strictly better)
184 * language Language code, e.g. "de" or "en-US".
185 * language_preference Is this in the language mentioned in
187 10 if it's what the URL is about,
188 -1 for default (don't know),
189 -10 otherwise, other values reserved for now.
190 * quality Order number of the video quality of this
191 format, irrespective of the file format.
192 -1 for default (order by other properties),
193 -2 or smaller for less than default.
194 * source_preference Order number for this video source
195 (quality takes higher priority)
196 -1 for default (order by other properties),
197 -2 or smaller for less than default.
198 * http_headers A dictionary of additional HTTP headers
199 to add to the request.
200 * stretched_ratio If given and not 1, indicates that the
201 video's pixels are not square.
202 width : height ratio as float.
203 * no_resume The server does not support resuming the
204 (HTTP or RTMP) download. Boolean.
205 * downloader_options A dictionary of downloader options as
206 described in FileDownloader
207 RTMP formats can also have the additional fields: page_url,
208 app, play_path, tc_url, flash_version, rtmp_live, rtmp_conn,
209 rtmp_protocol, rtmp_real_time
211 url: Final video URL.
212 ext: Video filename extension.
213 format: The video format, defaults to ext (used for --get-format)
214 player_url: SWF Player URL (used for rtmpdump).
216 The following fields are optional:
218 alt_title: A secondary title of the video.
219 display_id An alternative identifier for the video, not necessarily
220 unique, but available before title. Typically, id is
221 something like "4234987", title "Dancing naked mole rats",
222 and display_id "dancing-naked-mole-rats"
223 thumbnails: A list of dictionaries, with the following entries:
224 * "id" (optional, string) - Thumbnail format ID
226 * "preference" (optional, int) - quality of the image
227 * "width" (optional, int)
228 * "height" (optional, int)
229 * "resolution" (optional, string "{width}x{height}",
231 * "filesize" (optional, int)
232 * "_test_url" (optional, bool) - If true, test the URL
233 thumbnail: Full URL to a video thumbnail image.
234 description: Full video description.
235 uploader: Full name of the video uploader.
236 license: License name the video is licensed under.
237 creator: The creator of the video.
238 release_timestamp: UNIX timestamp of the moment the video was released.
239 release_date: The date (YYYYMMDD) when the video was released.
240 timestamp: UNIX timestamp of the moment the video was uploaded
241 upload_date: Video upload date (YYYYMMDD).
242 If not explicitly set, calculated from timestamp.
243 uploader_id: Nickname or id of the video uploader.
244 uploader_url: Full URL to a personal webpage of the video uploader.
245 channel: Full name of the channel the video is uploaded on.
246 Note that channel fields may or may not repeat uploader
247 fields. This depends on a particular extractor.
248 channel_id: Id of the channel.
249 channel_url: Full URL to a channel webpage.
250 location: Physical location where the video was filmed.
251 subtitles: The available subtitles as a dictionary in the format
252 {tag: subformats}. "tag" is usually a language code, and
253 "subformats" is a list sorted from lower to higher
254 preference, each element is a dictionary with the "ext"
256 * "data": The subtitles file contents
257 * "url": A URL pointing to the subtitles file
258 It can optionally also have:
259 * "name": Name or description of the subtitles
260 "ext" will be calculated from URL if missing
261 automatic_captions: Like 'subtitles'; contains automatically generated
262 captions instead of normal subtitles
263 duration: Length of the video in seconds, as an integer or float.
264 view_count: How many users have watched the video on the platform.
265 like_count: Number of positive ratings of the video
266 dislike_count: Number of negative ratings of the video
267 repost_count: Number of reposts of the video
268 average_rating: Average rating give by users, the scale used depends on the webpage
269 comment_count: Number of comments on the video
270 comments: A list of comments, each with one or more of the following
271 properties (all but one of text or html optional):
272 * "author" - human-readable name of the comment author
273 * "author_id" - user ID of the comment author
274 * "author_thumbnail" - The thumbnail of the comment author
276 * "html" - Comment as HTML
277 * "text" - Plain text of the comment
278 * "timestamp" - UNIX timestamp of comment
279 * "parent" - ID of the comment this one is replying to.
280 Set to "root" to indicate that this is a
281 comment to the original video.
282 * "like_count" - Number of positive ratings of the comment
283 * "dislike_count" - Number of negative ratings of the comment
284 * "is_favorited" - Whether the comment is marked as
285 favorite by the video uploader
286 * "author_is_uploader" - Whether the comment is made by
288 age_limit: Age restriction for the video, as an integer (years)
289 webpage_url: The URL to the video webpage, if given to yt-dlp it
290 should allow to get the same result again. (It will be set
291 by YoutubeDL if it's missing)
292 categories: A list of categories that the video falls in, for example
294 tags: A list of tags assigned to the video, e.g. ["sweden", "pop music"]
295 cast: A list of the video cast
296 is_live: True, False, or None (=unknown). Whether this video is a
297 live stream that goes on instead of a fixed-length video.
298 was_live: True, False, or None (=unknown). Whether this video was
299 originally a live stream.
300 live_status: 'is_live', 'upcoming', 'was_live', 'not_live' or None (=unknown)
301 If absent, automatically set from is_live, was_live
302 start_time: Time in seconds where the reproduction should start, as
303 specified in the URL.
304 end_time: Time in seconds where the reproduction should end, as
305 specified in the URL.
306 chapters: A list of dictionaries, with the following entries:
307 * "start_time" - The start time of the chapter in seconds
308 * "end_time" - The end time of the chapter in seconds
309 * "title" (optional, string)
310 playable_in_embed: Whether this video is allowed to play in embedded
311 players on other sites. Can be True (=always allowed),
312 False (=never allowed), None (=unknown), or a string
313 specifying the criteria for embedability (Eg: 'whitelist')
314 availability: Under what condition the video is available. One of
315 'private', 'premium_only', 'subscriber_only', 'needs_auth',
316 'unlisted' or 'public'. Use 'InfoExtractor._availability'
318 __post_extractor: A function to be called just before the metadata is
319 written to either disk, logger or console. The function
320 must return a dict which will be added to the info_dict.
321 This is usefull for additional information that is
322 time-consuming to extract. Note that the fields thus
323 extracted will not be available to output template and
324 match_filter. So, only "comments" and "comment_count" are
325 currently allowed to be extracted via this method.
327 The following fields should only be used when the video belongs to some logical
330 chapter: Name or title of the chapter the video belongs to.
331 chapter_number: Number of the chapter the video belongs to, as an integer.
332 chapter_id: Id of the chapter the video belongs to, as a unicode string.
334 The following fields should only be used when the video is an episode of some
335 series, programme or podcast:
337 series: Title of the series or programme the video episode belongs to.
338 season: Title of the season the video episode belongs to.
339 season_number: Number of the season the video episode belongs to, as an integer.
340 season_id: Id of the season the video episode belongs to, as a unicode string.
341 episode: Title of the video episode. Unlike mandatory video title field,
342 this field should denote the exact title of the video episode
343 without any kind of decoration.
344 episode_number: Number of the video episode within a season, as an integer.
345 episode_id: Id of the video episode, as a unicode string.
347 The following fields should only be used when the media is a track or a part of
350 track: Title of the track.
351 track_number: Number of the track within an album or a disc, as an integer.
352 track_id: Id of the track (useful in case of custom indexing, e.g. 6.iii),
354 artist: Artist(s) of the track.
355 genre: Genre(s) of the track.
356 album: Title of the album the track belongs to.
357 album_type: Type of the album (e.g. "Demo", "Full-length", "Split", "Compilation", etc).
358 album_artist: List of all artists appeared on the album (e.g.
359 "Ash Borer / Fell Voices" or "Various Artists", useful for splits
361 disc_number: Number of the disc or other physical medium the track belongs to,
363 release_year: Year (YYYY) when the album was released.
365 Unless mentioned otherwise, the fields should be Unicode strings.
367 Unless mentioned otherwise, None is equivalent to absence of information.
370 _type "playlist" indicates multiple videos.
371 There must be a key "entries", which is a list, an iterable, or a PagedList
372 object, each element of which is a valid dictionary by this specification.
374 Additionally, playlists can have "id", "title", and any other relevent
375 attributes with the same semantics as videos (see above).
378 _type "multi_video" indicates that there are multiple videos that
379 form a single show, for examples multiple acts of an opera or TV episode.
380 It must have an entries key like a playlist and contain all the keys
381 required for a video at the same time.
384 _type "url" indicates that the video must be extracted from another
385 location, possibly by a different extractor. Its only required key is:
386 "url" - the next URL to extract.
387 The key "ie_key" can be set to the class name (minus the trailing "IE",
388 e.g. "Youtube") if the extractor class is known in advance.
389 Additionally, the dictionary may have any properties of the resolved entity
390 known in advance, for example "title" if the title of the referred video is
394 _type "url_transparent" entities have the same specification as "url", but
395 indicate that the given additional information is more precise than the one
396 associated with the resolved URL.
397 This is useful when a site employs a video service that hosts the video and
398 its technical metadata, but that video service does not embed a useful
399 title, description etc.
402 Subclasses of this one should re-define the _real_initialize() and
403 _real_extract() methods and define a _VALID_URL regexp.
404 Probably, they should also be added to the list of extractors.
406 _GEO_BYPASS attribute may be set to False in order to disable
407 geo restriction bypass mechanisms for a particular extractor.
408 Though it won't disable explicit geo restriction bypass based on
409 country code provided with geo_bypass_country.
411 _GEO_COUNTRIES attribute may contain a list of presumably geo unrestricted
412 countries for this extractor. One of these countries will be used by
413 geo restriction bypass mechanism right away in order to bypass
414 geo restriction, of course, if the mechanism is not disabled.
416 _GEO_IP_BLOCKS attribute may contain a list of presumably geo unrestricted
417 IP blocks in CIDR notation for this extractor. One of these IP blocks
418 will be used by geo restriction bypass mechanism similarly
421 Finally, the _WORKING attribute should be set to False for broken IEs
422 in order to warn the users and skip the tests.
427 _x_forwarded_for_ip
= None
429 _GEO_COUNTRIES
= None
430 _GEO_IP_BLOCKS
= None
434 'any': 'Use --cookies, --username and --password or --netrc to provide account credentials',
436 'Use --cookies for the authentication. '
437 'See https://github.com/ytdl-org/youtube-dl#how-do-i-pass-cookies-to-youtube-dl for how to pass cookies'),
438 'password': 'Use --username and --password or --netrc to provide account credentials',
441 def __init__(self
, downloader
=None):
442 """Constructor. Receives an optional downloader."""
444 self
._x
_forwarded
_for
_ip
= None
445 self
.set_downloader(downloader
)
448 def suitable(cls
, url
):
449 """Receives a URL and returns True if suitable for this IE."""
451 # This does not use has/getattr intentionally - we want to know whether
452 # we have cached the regexp for *this* class, whereas getattr would also
453 # match the superclass
454 if '_VALID_URL_RE' not in cls
.__dict
__:
455 cls
._VALID
_URL
_RE
= re
.compile(cls
._VALID
_URL
)
456 return cls
._VALID
_URL
_RE
.match(url
) is not None
459 def _match_id(cls
, url
):
460 if '_VALID_URL_RE' not in cls
.__dict
__:
461 cls
._VALID
_URL
_RE
= re
.compile(cls
._VALID
_URL
)
462 m
= cls
._VALID
_URL
_RE
.match(url
)
464 return compat_str(m
.group('id'))
468 """Getter method for _WORKING."""
471 def initialize(self
):
472 """Initializes an instance (authentication, etc)."""
473 self
._initialize
_geo
_bypass
({
474 'countries': self
._GEO
_COUNTRIES
,
475 'ip_blocks': self
._GEO
_IP
_BLOCKS
,
478 self
._real
_initialize
()
481 def _initialize_geo_bypass(self
, geo_bypass_context
):
483 Initialize geo restriction bypass mechanism.
485 This method is used to initialize geo bypass mechanism based on faking
486 X-Forwarded-For HTTP header. A random country from provided country list
487 is selected and a random IP belonging to this country is generated. This
488 IP will be passed as X-Forwarded-For HTTP header in all subsequent
491 This method will be used for initial geo bypass mechanism initialization
492 during the instance initialization with _GEO_COUNTRIES and
495 You may also manually call it from extractor's code if geo bypass
496 information is not available beforehand (e.g. obtained during
497 extraction) or due to some other reason. In this case you should pass
498 this information in geo bypass context passed as first argument. It may
499 contain following fields:
501 countries: List of geo unrestricted countries (similar
503 ip_blocks: List of geo unrestricted IP blocks in CIDR notation
504 (similar to _GEO_IP_BLOCKS)
507 if not self
._x
_forwarded
_for
_ip
:
509 # Geo bypass mechanism is explicitly disabled by user
510 if not self
.get_param('geo_bypass', True):
513 if not geo_bypass_context
:
514 geo_bypass_context
= {}
516 # Backward compatibility: previously _initialize_geo_bypass
517 # expected a list of countries, some 3rd party code may still use
519 if isinstance(geo_bypass_context
, (list, tuple)):
520 geo_bypass_context
= {
521 'countries': geo_bypass_context
,
524 # The whole point of geo bypass mechanism is to fake IP
525 # as X-Forwarded-For HTTP header based on some IP block or
528 # Path 1: bypassing based on IP block in CIDR notation
530 # Explicit IP block specified by user, use it right away
531 # regardless of whether extractor is geo bypassable or not
532 ip_block
= self
.get_param('geo_bypass_ip_block', None)
534 # Otherwise use random IP block from geo bypass context but only
535 # if extractor is known as geo bypassable
537 ip_blocks
= geo_bypass_context
.get('ip_blocks')
538 if self
._GEO
_BYPASS
and ip_blocks
:
539 ip_block
= random
.choice(ip_blocks
)
542 self
._x
_forwarded
_for
_ip
= GeoUtils
.random_ipv4(ip_block
)
543 self
._downloader
.write_debug(
544 '[debug] Using fake IP %s as X-Forwarded-For' % self
._x
_forwarded
_for
_ip
)
547 # Path 2: bypassing based on country code
549 # Explicit country code specified by user, use it right away
550 # regardless of whether extractor is geo bypassable or not
551 country
= self
.get_param('geo_bypass_country', None)
553 # Otherwise use random country code from geo bypass context but
554 # only if extractor is known as geo bypassable
556 countries
= geo_bypass_context
.get('countries')
557 if self
._GEO
_BYPASS
and countries
:
558 country
= random
.choice(countries
)
561 self
._x
_forwarded
_for
_ip
= GeoUtils
.random_ipv4(country
)
562 self
._downloader
.write_debug(
563 'Using fake IP %s (%s) as X-Forwarded-For' % (self
._x
_forwarded
_for
_ip
, country
.upper()))
565 def extract(self
, url
):
566 """Extracts URL information and returns it in list of dicts."""
571 self
.write_debug('Extracting URL: %s' % url
)
572 ie_result
= self
._real
_extract
(url
)
573 if ie_result
is None:
575 if self
._x
_forwarded
_for
_ip
:
576 ie_result
['__x_forwarded_for_ip'] = self
._x
_forwarded
_for
_ip
577 subtitles
= ie_result
.get('subtitles')
578 if (subtitles
and 'live_chat' in subtitles
579 and 'no-live-chat' in self
.get_param('compat_opts', [])):
580 del subtitles
['live_chat']
582 except GeoRestrictedError
as e
:
583 if self
.__maybe
_fake
_ip
_and
_retry
(e
.countries
):
586 except ExtractorError
:
588 except compat_http_client
.IncompleteRead
as e
:
589 raise ExtractorError('A network error has occurred.', cause
=e
, expected
=True)
590 except (KeyError, StopIteration) as e
:
591 raise ExtractorError('An extractor error has occurred.', cause
=e
)
593 def __maybe_fake_ip_and_retry(self
, countries
):
594 if (not self
.get_param('geo_bypass_country', None)
596 and self
.get_param('geo_bypass', True)
597 and not self
._x
_forwarded
_for
_ip
599 country_code
= random
.choice(countries
)
600 self
._x
_forwarded
_for
_ip
= GeoUtils
.random_ipv4(country_code
)
601 if self
._x
_forwarded
_for
_ip
:
603 'Video is geo restricted. Retrying extraction with fake IP %s (%s) as X-Forwarded-For.'
604 % (self
._x
_forwarded
_for
_ip
, country_code
.upper()))
608 def set_downloader(self
, downloader
):
609 """Sets the downloader for this IE."""
610 self
._downloader
= downloader
612 def _real_initialize(self
):
613 """Real initialization process. Redefine in subclasses."""
616 def _real_extract(self
, url
):
617 """Real extraction process. Redefine in subclasses."""
622 """A string for getting the InfoExtractor with get_info_extractor"""
623 return compat_str(cls
.__name
__[:-2])
627 return compat_str(type(self
).__name
__[:-2])
630 def __can_accept_status_code(err
, expected_status
):
631 assert isinstance(err
, compat_urllib_error
.HTTPError
)
632 if expected_status
is None:
634 elif callable(expected_status
):
635 return expected_status(err
.code
) is True
637 return err
.code
in variadic(expected_status
)
639 def _request_webpage(self
, url_or_request
, video_id
, note
=None, errnote
=None, fatal
=True, data
=None, headers
={}, query={}
, expected_status
=None):
641 Return the response handle.
643 See _download_webpage docstring for arguments specification.
645 if not self
._downloader
._first
_webpage
_request
:
646 sleep_interval
= float_or_none(self
.get_param('sleep_interval_requests')) or 0
647 if sleep_interval
> 0:
648 self
.to_screen('Sleeping %s seconds ...' % sleep_interval
)
649 time
.sleep(sleep_interval
)
651 self
._downloader
._first
_webpage
_request
= False
654 self
.report_download_webpage(video_id
)
655 elif note
is not False:
657 self
.to_screen('%s' % (note
,))
659 self
.to_screen('%s: %s' % (video_id
, note
))
661 # Some sites check X-Forwarded-For HTTP header in order to figure out
662 # the origin of the client behind proxy. This allows bypassing geo
663 # restriction by faking this header's value to IP that belongs to some
664 # geo unrestricted country. We will do so once we encounter any
665 # geo restriction error.
666 if self
._x
_forwarded
_for
_ip
:
667 if 'X-Forwarded-For' not in headers
:
668 headers
['X-Forwarded-For'] = self
._x
_forwarded
_for
_ip
670 if isinstance(url_or_request
, compat_urllib_request
.Request
):
671 url_or_request
= update_Request(
672 url_or_request
, data
=data
, headers
=headers
, query
=query
)
675 url_or_request
= update_url_query(url_or_request
, query
)
676 if data
is not None or headers
:
677 url_or_request
= sanitized_Request(url_or_request
, data
, headers
)
679 return self
._downloader
.urlopen(url_or_request
)
680 except network_exceptions
as err
:
681 if isinstance(err
, compat_urllib_error
.HTTPError
):
682 if self
.__can
_accept
_status
_code
(err
, expected_status
):
683 # Retain reference to error to prevent file object from
684 # being closed before it can be read. Works around the
685 # effects of <https://bugs.python.org/issue15002>
686 # introduced in Python 3.4.1.
693 errnote
= 'Unable to download webpage'
695 errmsg
= '%s: %s' % (errnote
, error_to_compat_str(err
))
697 raise ExtractorError(errmsg
, sys
.exc_info()[2], cause
=err
)
699 self
.report_warning(errmsg
)
702 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):
704 Return a tuple (page content as string, URL handle).
706 See _download_webpage docstring for arguments specification.
708 # Strip hashes from the URL (#1038)
709 if isinstance(url_or_request
, (compat_str
, str)):
710 url_or_request
= url_or_request
.partition('#')[0]
712 urlh
= self
._request
_webpage
(url_or_request
, video_id
, note
, errnote
, fatal
, data
=data
, headers
=headers
, query
=query
, expected_status
=expected_status
)
716 content
= self
._webpage
_read
_content
(urlh
, url_or_request
, video_id
, note
, errnote
, fatal
, encoding
=encoding
)
717 return (content
, urlh
)
720 def _guess_encoding_from_content(content_type
, webpage_bytes
):
721 m
= re
.match(r
'[a-zA-Z0-9_.-]+/[a-zA-Z0-9_.-]+\s*;\s*charset=(.+)', content_type
)
723 encoding
= m
.group(1)
725 m
= re
.search(br
'<meta[^>]+charset=[\'"]?([^\'")]+)[ /\'">]',
726 webpage_bytes[:1024])
728 encoding = m.group(1).decode('ascii')
729 elif webpage_bytes.startswith(b'\xff\xfe'):
736 def __check_blocked(self, content):
737 first_block = content[:512]
738 if ('<title>Access to this site is blocked</title>' in content
739 and 'Websense' in first_block):
740 msg = 'Access to this webpage has been blocked by Websense filtering software in your network.'
741 blocked_iframe = self._html_search_regex(
742 r'<iframe src="([^
"]+)"', content,
743 'Websense information URL
', default=None)
745 msg += ' Visit
%s for more details
' % blocked_iframe
746 raise ExtractorError(msg, expected=True)
747 if '<title
>The URL you requested has been blocked
</title
>' in first_block:
749 'Access to this webpage has been blocked by Indian censorship
. '
750 'Use a VPN
or proxy
server (with --proxy
) to route around it
.')
751 block_msg = self._html_search_regex(
752 r'</h1
><p
>(.*?
)</p
>',
753 content, 'block message
', default=None)
755 msg += ' (Message
: "%s")' % block_msg.replace('\n', ' ')
756 raise ExtractorError(msg, expected=True)
757 if ('<title
>TTK
:: Доступ к ресурсу ограничен
</title
>' in content
758 and 'blocklist
.rkn
.gov
.ru
' in content):
759 raise ExtractorError(
760 'Access to this webpage has been blocked by decision of the Russian government
. '
761 'Visit http
://blocklist
.rkn
.gov
.ru
/ for a block reason
.',
764 def _webpage_read_content(self, urlh, url_or_request, video_id, note=None, errnote=None, fatal=True, prefix=None, encoding=None):
765 content_type = urlh.headers.get('Content
-Type
', '')
766 webpage_bytes = urlh.read()
767 if prefix is not None:
768 webpage_bytes = prefix + webpage_bytes
770 encoding = self._guess_encoding_from_content(content_type, webpage_bytes)
771 if self.get_param('dump_intermediate_pages
', False):
772 self.to_screen('Dumping request to
' + urlh.geturl())
773 dump = base64.b64encode(webpage_bytes).decode('ascii
')
774 self._downloader.to_screen(dump)
775 if self.get_param('write_pages
', False):
776 basen = '%s_%s' % (video_id, urlh.geturl())
778 h = '___
' + hashlib.md5(basen.encode('utf
-8')).hexdigest()
779 basen = basen[:240 - len(h)] + h
780 raw_filename = basen + '.dump
'
781 filename = sanitize_filename(raw_filename, restricted=True)
782 self.to_screen('Saving request to
' + filename)
783 # Working around MAX_PATH limitation on Windows (see
784 # http://msdn.microsoft.com/en-us/library/windows/desktop/aa365247(v=vs.85).aspx)
785 if compat_os_name == 'nt
':
786 absfilepath = os.path.abspath(filename)
787 if len(absfilepath) > 259:
788 filename = '\\\\?
\\' + absfilepath
789 with open(filename, 'wb
') as outf:
790 outf.write(webpage_bytes)
793 content = webpage_bytes.decode(encoding, 'replace
')
795 content = webpage_bytes.decode('utf
-8', 'replace
')
797 self.__check_blocked(content)
801 def _download_webpage(
802 self, url_or_request, video_id, note=None, errnote=None,
803 fatal=True, tries=1, timeout=5, encoding=None, data=None,
804 headers={}, query={}, expected_status=None):
806 Return the data of the page as a string.
809 url_or_request -- plain text URL as a string or
810 a compat_urllib_request.Requestobject
811 video_id -- Video/playlist/item identifier (string)
814 note -- note printed before downloading (string)
815 errnote -- note printed in case of an error (string)
816 fatal -- flag denoting whether error should be considered fatal,
817 i.e. whether it should cause ExtractionError to be raised,
818 otherwise a warning will be reported and extraction continued
819 tries -- number of tries
820 timeout -- sleep interval between tries
821 encoding -- encoding for a page content decoding, guessed automatically
822 when not explicitly specified
823 data -- POST data (bytes)
824 headers -- HTTP headers (dict)
825 query -- URL query (dict)
826 expected_status -- allows to accept failed HTTP requests (non 2xx
827 status code) by explicitly specifying a set of accepted status
828 codes. Can be any of the following entities:
829 - an integer type specifying an exact failed status code to
831 - a list or a tuple of integer types specifying a list of
832 failed status codes to accept
833 - a callable accepting an actual failed status code and
834 returning True if it should be accepted
835 Note that this argument does not affect success status codes (2xx)
836 which are always accepted.
841 while success is False:
843 res = self._download_webpage_handle(
844 url_or_request, video_id, note, errnote, fatal,
845 encoding=encoding, data=data, headers=headers, query=query,
846 expected_status=expected_status)
848 except compat_http_client.IncompleteRead as e:
850 if try_count >= tries:
852 self._sleep(timeout, video_id)
859 def _download_xml_handle(
860 self, url_or_request, video_id, note='Downloading XML
',
861 errnote='Unable to download XML
', transform_source=None,
862 fatal=True, encoding=None, data=None, headers={}, query={},
863 expected_status=None):
865 Return a tuple (xml as an compat_etree_Element, URL handle).
867 See _download_webpage docstring for arguments specification.
869 res = self._download_webpage_handle(
870 url_or_request, video_id, note, errnote, fatal=fatal,
871 encoding=encoding, data=data, headers=headers, query=query,
872 expected_status=expected_status)
875 xml_string, urlh = res
876 return self._parse_xml(
877 xml_string, video_id, transform_source=transform_source,
881 self, url_or_request, video_id,
882 note='Downloading XML
', errnote='Unable to download XML
',
883 transform_source=None, fatal=True, encoding=None,
884 data=None, headers={}, query={}, expected_status=None):
886 Return the xml as an compat_etree_Element.
888 See _download_webpage docstring for arguments specification.
890 res = self._download_xml_handle(
891 url_or_request, video_id, note=note, errnote=errnote,
892 transform_source=transform_source, fatal=fatal, encoding=encoding,
893 data=data, headers=headers, query=query,
894 expected_status=expected_status)
895 return res if res is False else res[0]
897 def _parse_xml(self, xml_string, video_id, transform_source=None, fatal=True):
899 xml_string = transform_source(xml_string)
901 return compat_etree_fromstring(xml_string.encode('utf
-8'))
902 except compat_xml_parse_error as ve:
903 errmsg = '%s: Failed to parse XML
' % video_id
905 raise ExtractorError(errmsg, cause=ve)
907 self.report_warning(errmsg + str(ve))
909 def _download_json_handle(
910 self, url_or_request, video_id, note='Downloading JSON metadata
',
911 errnote='Unable to download JSON metadata
', transform_source=None,
912 fatal=True, encoding=None, data=None, headers={}, query={},
913 expected_status=None):
915 Return a tuple (JSON object, URL handle).
917 See _download_webpage docstring for arguments specification.
919 res = self._download_webpage_handle(
920 url_or_request, video_id, note, errnote, fatal=fatal,
921 encoding=encoding, data=data, headers=headers, query=query,
922 expected_status=expected_status)
925 json_string, urlh = res
926 return self._parse_json(
927 json_string, video_id, transform_source=transform_source,
931 self, url_or_request, video_id, note='Downloading JSON metadata
',
932 errnote='Unable to download JSON metadata
', transform_source=None,
933 fatal=True, encoding=None, data=None, headers={}, query={},
934 expected_status=None):
936 Return the JSON object as a dict.
938 See _download_webpage docstring for arguments specification.
940 res = self._download_json_handle(
941 url_or_request, video_id, note=note, errnote=errnote,
942 transform_source=transform_source, fatal=fatal, encoding=encoding,
943 data=data, headers=headers, query=query,
944 expected_status=expected_status)
945 return res if res is False else res[0]
947 def _parse_json(self, json_string, video_id, transform_source=None, fatal=True):
949 json_string = transform_source(json_string)
951 return json.loads(json_string)
952 except ValueError as ve:
953 errmsg = '%s: Failed to parse JSON
' % video_id
955 raise ExtractorError(errmsg, cause=ve)
957 self.report_warning(errmsg + str(ve))
959 def _parse_socket_response_as_json(self, data, video_id, transform_source=None, fatal=True):
960 return self._parse_json(
961 data[data.find('{'):data.rfind('}
') + 1],
962 video_id, transform_source, fatal)
964 def _download_socket_json_handle(
965 self, url_or_request, video_id, note='Polling socket
',
966 errnote='Unable to poll socket
', transform_source=None,
967 fatal=True, encoding=None, data=None, headers={}, query={},
968 expected_status=None):
970 Return a tuple (JSON object, URL handle).
972 See _download_webpage docstring for arguments specification.
974 res = self._download_webpage_handle(
975 url_or_request, video_id, note, errnote, fatal=fatal,
976 encoding=encoding, data=data, headers=headers, query=query,
977 expected_status=expected_status)
981 return self._parse_socket_response_as_json(
982 webpage, video_id, transform_source=transform_source,
985 def _download_socket_json(
986 self, url_or_request, video_id, note='Polling socket
',
987 errnote='Unable to poll socket
', transform_source=None,
988 fatal=True, encoding=None, data=None, headers={}, query={},
989 expected_status=None):
991 Return the JSON object as a dict.
993 See _download_webpage docstring for arguments specification.
995 res = self._download_socket_json_handle(
996 url_or_request, video_id, note=note, errnote=errnote,
997 transform_source=transform_source, fatal=fatal, encoding=encoding,
998 data=data, headers=headers, query=query,
999 expected_status=expected_status)
1000 return res if res is False else res[0]
1002 def report_warning(self, msg, video_id=None, *args, **kwargs):
1003 idstr = '' if video_id is None else '%s: ' % video_id
1004 self._downloader.report_warning(
1005 '[%s] %s%s' % (self.IE_NAME, idstr, msg), *args, **kwargs)
1007 def to_screen(self, msg, *args, **kwargs):
1008 """Print msg to screen, prefixing it with '[ie_name
]'"""
1009 self._downloader.to_screen('[%s] %s' % (self.IE_NAME, msg), *args, **kwargs)
1011 def write_debug(self, msg, *args, **kwargs):
1012 self._downloader.write_debug('[%s] %s' % (self.IE_NAME, msg), *args, **kwargs)
1014 def get_param(self, name, default=None, *args, **kwargs):
1015 if self._downloader:
1016 return self._downloader.params.get(name, default, *args, **kwargs)
1019 def report_extraction(self, id_or_name):
1020 """Report information extraction."""
1021 self.to_screen('%s: Extracting information
' % id_or_name)
1023 def report_download_webpage(self, video_id):
1024 """Report webpage download."""
1025 self.to_screen('%s: Downloading webpage
' % video_id)
1027 def report_age_confirmation(self):
1028 """Report attempt to confirm age."""
1029 self.to_screen('Confirming age
')
1031 def report_login(self):
1032 """Report attempt to log in."""
1033 self.to_screen('Logging
in')
1035 def raise_login_required(
1036 self, msg='This video
is only available
for registered users
',
1037 metadata_available=False, method='any
'):
1038 if metadata_available and self.get_param('ignore_no_formats_error
'):
1039 self.report_warning(msg)
1040 if method is not None:
1041 msg = '%s. %s' % (msg, self._LOGIN_HINTS[method])
1042 raise ExtractorError(msg, expected=True)
1044 def raise_geo_restricted(
1045 self, msg='This video
is not available
from your location due to geo restriction
',
1046 countries=None, metadata_available=False):
1047 if metadata_available and self.get_param('ignore_no_formats_error
'):
1048 self.report_warning(msg)
1050 raise GeoRestrictedError(msg, countries=countries)
1052 def raise_no_formats(self, msg, expected=False, video_id=None):
1053 if expected and self.get_param('ignore_no_formats_error
'):
1054 self.report_warning(msg, video_id)
1056 raise ExtractorError(msg, expected=expected, video_id=video_id)
1058 # Methods for following #608
1060 def url_result(url, ie=None, video_id=None, video_title=None):
1061 """Returns a URL that points to a page that should be processed"""
1062 # TODO: ie should be the class used for getting the info
1063 video_info = {'_type
': 'url
',
1066 if video_id is not None:
1067 video_info['id'] = video_id
1068 if video_title is not None:
1069 video_info['title
'] = video_title
1072 def playlist_from_matches(self, matches, playlist_id=None, playlist_title=None, getter=None, ie=None):
1074 self.url_result(self._proto_relative_url(getter(m) if getter else m), ie)
1076 return self.playlist_result(
1077 urls, playlist_id=playlist_id, playlist_title=playlist_title)
1080 def playlist_result(entries, playlist_id=None, playlist_title=None, playlist_description=None, **kwargs):
1081 """Returns a playlist"""
1082 video_info = {'_type
': 'playlist
',
1084 video_info.update(kwargs)
1086 video_info['id'] = playlist_id
1088 video_info['title
'] = playlist_title
1089 if playlist_description is not None:
1090 video_info['description
'] = playlist_description
1093 def _search_regex(self, pattern, string, name, default=NO_DEFAULT, fatal=True, flags=0, group=None):
1095 Perform a regex search on the given string, using a single or a list of
1096 patterns returning the first matching group.
1097 In case of failure return a default value or raise a WARNING or a
1098 RegexNotFoundError, depending on fatal, specifying the field name.
1100 if isinstance(pattern, (str, compat_str, compiled_regex_type)):
1101 mobj = re.search(pattern, string, flags)
1104 mobj = re.search(p, string, flags)
1108 if not self.get_param('no_color
') and compat_os_name != 'nt
' and sys.stderr.isatty():
1109 _name = '\033[0;34m
%s\033[0m
' % name
1115 # return the first matching group
1116 return next(g for g in mobj.groups() if g is not None)
1117 elif isinstance(group, (list, tuple)):
1118 return tuple(mobj.group(g) for g in group)
1120 return mobj.group(group)
1121 elif default is not NO_DEFAULT:
1124 raise RegexNotFoundError('Unable to extract
%s' % _name)
1126 self.report_warning('unable to extract
%s' % _name + bug_reports_message())
1129 def _html_search_regex(self, pattern, string, name, default=NO_DEFAULT, fatal=True, flags=0, group=None):
1131 Like _search_regex, but strips HTML tags and unescapes entities.
1133 res = self._search_regex(pattern, string, name, default, fatal, flags, group)
1135 return clean_html(res).strip()
1139 def _get_netrc_login_info(self, netrc_machine=None):
1142 netrc_machine = netrc_machine or self._NETRC_MACHINE
1144 if self.get_param('usenetrc
', False):
1146 info = netrc.netrc().authenticators(netrc_machine)
1147 if info is not None:
1151 raise netrc.NetrcParseError(
1152 'No authenticators
for %s' % netrc_machine)
1153 except (IOError, netrc.NetrcParseError) as err:
1154 self.report_warning(
1155 'parsing
.netrc
: %s' % error_to_compat_str(err))
1157 return username, password
1159 def _get_login_info(self, username_option='username
', password_option='password
', netrc_machine=None):
1161 Get the login info as (username, password)
1162 First look for the manually specified credentials using username_option
1163 and password_option as keys in params dictionary. If no such credentials
1164 available look in the netrc file using the netrc_machine or _NETRC_MACHINE
1166 If there's no info available
, return (None, None)
1169 # Attempt to use provided username and password or .netrc data
1170 username = self.get_param(username_option)
1171 if username is not None:
1172 password = self.get_param(password_option)
1174 username, password = self._get_netrc_login_info(netrc_machine)
1176 return username, password
1178 def _get_tfa_info(self, note='two-factor verification code'):
1180 Get the two
-factor authentication info
1181 TODO
- asking the user will be required
for sms
/phone verify
1182 currently just uses the command line option
1183 If there
's no info available, return None
1186 tfa = self.get_param('twofactor
')
1190 return compat_getpass('Type
%s and press
[Return
]: ' % note)
1192 # Helper functions for extracting OpenGraph info
1194 def _og_regexes(prop):
1195 content_re = r'content
=(?
:"([^"]+?
)"|\'([^\']+?)\'|\s*([^\s"\'=<>`
]+?
))'
1196 property_re = (r'(?
:name|
property)=(?
:\'og
[:-]%(prop)s\'|
"og[:-]%(prop)s"|\s
*og
[:-]%(prop)s\b)'
1197 % {'prop': re.escape(prop)})
1198 template = r'<meta
[^
>]+?
%s[^
>]+?
%s'
1200 template % (property_re, content_re),
1201 template % (content_re, property_re),
1205 def _meta_regex(prop):
1206 return r'''(?isx)<meta
1207 (?=[^>]+(?:itemprop|name|property|id|http-equiv)=(["\']?)%s\1)
1208 [^>]+?content=(["\'])(?P<content>.*?)\2''' % re.escape(prop)
1210 def _og_search_property(self, prop, html, name=None, **kargs):
1211 prop = variadic(prop)
1213 name = 'OpenGraph
%s' % prop[0]
1216 og_regexes.extend(self._og_regexes(p))
1217 escaped = self._search_regex(og_regexes, html, name, flags=re.DOTALL, **kargs)
1220 return unescapeHTML(escaped)
1222 def _og_search_thumbnail(self, html, **kargs):
1223 return self._og_search_property('image
', html, 'thumbnail URL
', fatal=False, **kargs)
1225 def _og_search_description(self, html, **kargs):
1226 return self._og_search_property('description
', html, fatal=False, **kargs)
1228 def _og_search_title(self, html, **kargs):
1229 return self._og_search_property('title
', html, **kargs)
1231 def _og_search_video_url(self, html, name='video url
', secure=True, **kargs):
1232 regexes = self._og_regexes('video
') + self._og_regexes('video
:url
')
1234 regexes = self._og_regexes('video
:secure_url
') + regexes
1235 return self._html_search_regex(regexes, html, name, **kargs)
1237 def _og_search_url(self, html, **kargs):
1238 return self._og_search_property('url
', html, **kargs)
1240 def _html_search_meta(self, name, html, display_name=None, fatal=False, **kwargs):
1241 name = variadic(name)
1242 if display_name is None:
1243 display_name = name[0]
1244 return self._html_search_regex(
1245 [self._meta_regex(n) for n in name],
1246 html, display_name, fatal=fatal, group='content
', **kwargs)
1248 def _dc_search_uploader(self, html):
1249 return self._html_search_meta('dc
.creator
', html, 'uploader
')
1251 def _rta_search(self, html):
1252 # See http://www.rtalabel.org/index.php?content=howtofaq#single
1253 if re.search(r'(?ix
)<meta\s
+name
="rating"\s
+'
1254 r' content
="RTA-5042-1996-1400-1577-RTA"',
1259 def _media_rating_search(self, html):
1260 # See http://www.tjg-designs.com/WP/metadata-code-examples-adding-metadata-to-your-web-pages/
1261 rating = self._html_search_meta('rating
', html)
1273 return RATING_TABLE.get(rating.lower())
1275 def _family_friendly_search(self, html):
1276 # See http://schema.org/VideoObject
1277 family_friendly = self._html_search_meta(
1278 'isFamilyFriendly
', html, default=None)
1280 if not family_friendly:
1289 return RATING_TABLE.get(family_friendly.lower())
1291 def _twitter_search_player(self, html):
1292 return self._html_search_meta('twitter
:player
', html,
1293 'twitter card player
')
1295 def _search_json_ld(self, html, video_id, expected_type=None, **kwargs):
1296 json_ld_list = list(re.finditer(JSON_LD_RE, html))
1297 default = kwargs.get('default
', NO_DEFAULT)
1298 # JSON-LD may be malformed and thus `fatal` should be respected.
1299 # At the same time `default` may be passed that assumes `fatal=False`
1300 # for _search_regex. Let's simulate the same behavior here
as well
.
1301 fatal
= kwargs
.get('fatal', True) if default
is NO_DEFAULT
else False
1303 for mobj
in json_ld_list
:
1304 json_ld_item
= self
._parse
_json
(
1305 mobj
.group('json_ld'), video_id
, fatal
=fatal
)
1306 if not json_ld_item
:
1308 if isinstance(json_ld_item
, dict):
1309 json_ld
.append(json_ld_item
)
1310 elif isinstance(json_ld_item
, (list, tuple)):
1311 json_ld
.extend(json_ld_item
)
1313 json_ld
= self
._json
_ld
(json_ld
, video_id
, fatal
=fatal
, expected_type
=expected_type
)
1316 if default
is not NO_DEFAULT
:
1319 raise RegexNotFoundError('Unable to extract JSON-LD')
1321 self
.report_warning('unable to extract JSON-LD %s' % bug_reports_message())
1324 def _json_ld(self
, json_ld
, video_id
, fatal
=True, expected_type
=None):
1325 if isinstance(json_ld
, compat_str
):
1326 json_ld
= self
._parse
_json
(json_ld
, video_id
, fatal
=fatal
)
1330 if not isinstance(json_ld
, (list, tuple, dict)):
1332 if isinstance(json_ld
, dict):
1335 INTERACTION_TYPE_MAP
= {
1336 'CommentAction': 'comment',
1337 'AgreeAction': 'like',
1338 'DisagreeAction': 'dislike',
1339 'LikeAction': 'like',
1340 'DislikeAction': 'dislike',
1341 'ListenAction': 'view',
1342 'WatchAction': 'view',
1343 'ViewAction': 'view',
1346 def extract_interaction_type(e
):
1347 interaction_type
= e
.get('interactionType')
1348 if isinstance(interaction_type
, dict):
1349 interaction_type
= interaction_type
.get('@type')
1350 return str_or_none(interaction_type
)
1352 def extract_interaction_statistic(e
):
1353 interaction_statistic
= e
.get('interactionStatistic')
1354 if isinstance(interaction_statistic
, dict):
1355 interaction_statistic
= [interaction_statistic
]
1356 if not isinstance(interaction_statistic
, list):
1358 for is_e
in interaction_statistic
:
1359 if not isinstance(is_e
, dict):
1361 if is_e
.get('@type') != 'InteractionCounter':
1363 interaction_type
= extract_interaction_type(is_e
)
1364 if not interaction_type
:
1366 # For interaction count some sites provide string instead of
1367 # an integer (as per spec) with non digit characters (e.g. ",")
1368 # so extracting count with more relaxed str_to_int
1369 interaction_count
= str_to_int(is_e
.get('userInteractionCount'))
1370 if interaction_count
is None:
1372 count_kind
= INTERACTION_TYPE_MAP
.get(interaction_type
.split('/')[-1])
1375 count_key
= '%s_count' % count_kind
1376 if info
.get(count_key
) is not None:
1378 info
[count_key
] = interaction_count
1380 def extract_video_object(e
):
1381 assert e
['@type'] == 'VideoObject'
1382 author
= e
.get('author')
1384 'url': url_or_none(e
.get('contentUrl')),
1385 'title': unescapeHTML(e
.get('name')),
1386 'description': unescapeHTML(e
.get('description')),
1387 'thumbnail': url_or_none(e
.get('thumbnailUrl') or e
.get('thumbnailURL')),
1388 'duration': parse_duration(e
.get('duration')),
1389 'timestamp': unified_timestamp(e
.get('uploadDate')),
1390 # author can be an instance of 'Organization' or 'Person' types.
1391 # both types can have 'name' property(inherited from 'Thing' type). [1]
1392 # however some websites are using 'Text' type instead.
1393 # 1. https://schema.org/VideoObject
1394 'uploader': author
.get('name') if isinstance(author
, dict) else author
if isinstance(author
, compat_str
) else None,
1395 'filesize': float_or_none(e
.get('contentSize')),
1396 'tbr': int_or_none(e
.get('bitrate')),
1397 'width': int_or_none(e
.get('width')),
1398 'height': int_or_none(e
.get('height')),
1399 'view_count': int_or_none(e
.get('interactionCount')),
1401 extract_interaction_statistic(e
)
1405 item_type
= e
.get('@type')
1406 if expected_type
is not None and expected_type
!= item_type
:
1408 if item_type
in ('TVEpisode', 'Episode'):
1409 episode_name
= unescapeHTML(e
.get('name'))
1411 'episode': episode_name
,
1412 'episode_number': int_or_none(e
.get('episodeNumber')),
1413 'description': unescapeHTML(e
.get('description')),
1415 if not info
.get('title') and episode_name
:
1416 info
['title'] = episode_name
1417 part_of_season
= e
.get('partOfSeason')
1418 if isinstance(part_of_season
, dict) and part_of_season
.get('@type') in ('TVSeason', 'Season', 'CreativeWorkSeason'):
1420 'season': unescapeHTML(part_of_season
.get('name')),
1421 'season_number': int_or_none(part_of_season
.get('seasonNumber')),
1423 part_of_series
= e
.get('partOfSeries') or e
.get('partOfTVSeries')
1424 if isinstance(part_of_series
, dict) and part_of_series
.get('@type') in ('TVSeries', 'Series', 'CreativeWorkSeries'):
1425 info
['series'] = unescapeHTML(part_of_series
.get('name'))
1426 elif item_type
== 'Movie':
1428 'title': unescapeHTML(e
.get('name')),
1429 'description': unescapeHTML(e
.get('description')),
1430 'duration': parse_duration(e
.get('duration')),
1431 'timestamp': unified_timestamp(e
.get('dateCreated')),
1433 elif item_type
in ('Article', 'NewsArticle'):
1435 'timestamp': parse_iso8601(e
.get('datePublished')),
1436 'title': unescapeHTML(e
.get('headline')),
1437 'description': unescapeHTML(e
.get('articleBody')),
1439 elif item_type
== 'VideoObject':
1440 extract_video_object(e
)
1441 if expected_type
is None:
1445 video
= e
.get('video')
1446 if isinstance(video
, dict) and video
.get('@type') == 'VideoObject':
1447 extract_video_object(video
)
1448 if expected_type
is None:
1452 return dict((k
, v
) for k
, v
in info
.items() if v
is not None)
1455 def _hidden_inputs(html
):
1456 html
= re
.sub(r
'<!--(?:(?!<!--).)*-->', '', html
)
1458 for input in re
.findall(r
'(?i)(<input[^>]+>)', html
):
1459 attrs
= extract_attributes(input)
1462 if attrs
.get('type') not in ('hidden', 'submit'):
1464 name
= attrs
.get('name') or attrs
.get('id')
1465 value
= attrs
.get('value')
1466 if name
and value
is not None:
1467 hidden_inputs
[name
] = value
1468 return hidden_inputs
1470 def _form_hidden_inputs(self
, form_id
, html
):
1471 form
= self
._search
_regex
(
1472 r
'(?is)<form[^>]+?id=(["\'])%s\
1[^
>]*>(?P
<form
>.+?
)</form
>' % form_id,
1473 html, '%s form
' % form_id, group='form
')
1474 return self._hidden_inputs(form)
1477 regex = r' *((?P
<reverse
>\
+)?
(?P
<field
>[a
-zA
-Z0
-9_]+)((?P
<separator
>[~
:])(?P
<limit
>.*?
))?
)?
*$
'
1479 default = ('hidden
', 'aud_or_vid
', 'hasvid
', 'ie_pref
', 'lang
', 'quality
',
1480 'res
', 'fps
', 'codec
:vp9
.2
', 'size
', 'br
', 'asr
',
1481 'proto
', 'ext
', 'hasaud
', 'source
', 'format_id
') # These must not be aliases
1482 ytdl_default = ('hasaud
', 'quality
', 'tbr
', 'filesize
', 'vbr
',
1483 'height
', 'width
', 'proto
', 'vext
', 'abr
', 'aext
',
1484 'fps
', 'fs_approx
', 'source
', 'format_id
')
1487 'vcodec
': {'type': 'ordered
', 'regex
': True,
1488 'order
': ['av0?
1', 'vp0?
9.2', 'vp0?
9', '[hx
]265|he?vc?
', '[hx
]264|avc
', 'vp0?
8', 'mp4v|h263
', 'theora
', '', None, 'none
']},
1489 'acodec
': {'type': 'ordered
', 'regex
': True,
1490 'order
': ['opus
', 'vorbis
', 'aac
', 'mp?
4a?
', 'mp3
', 'e?a?c
-?
3', 'dts
', '', None, 'none
']},
1491 'proto
': {'type': 'ordered
', 'regex
': True, 'field
': 'protocol
',
1492 'order
': ['(ht|f
)tps
', '(ht|f
)tp$
', 'm3u8
.+', '.*dash
', 'ws|websocket
', '', 'mms|rtsp
', 'none
', 'f4
']},
1493 'vext
': {'type': 'ordered
', 'field
': 'video_ext
',
1494 'order
': ('mp4
', 'webm
', 'flv
', '', 'none
'),
1495 'order_free
': ('webm
', 'mp4
', 'flv
', '', 'none
')},
1496 'aext
': {'type': 'ordered
', 'field
': 'audio_ext
',
1497 'order
': ('m4a
', 'aac
', 'mp3
', 'ogg
', 'opus
', 'webm
', '', 'none
'),
1498 'order_free
': ('opus
', 'ogg
', 'webm
', 'm4a
', 'mp3
', 'aac
', '', 'none
')},
1499 'hidden
': {'visible': False, 'forced': True, 'type': 'extractor', 'max': -1000},
1500 'aud_or_vid
': {'visible
': False, 'forced
': True, 'type': 'multiple
',
1501 'field
': ('vcodec
', 'acodec
'),
1502 'function
': lambda it: int(any(v != 'none
' for v in it))},
1503 'ie_pref
': {'priority': True, 'type': 'extractor'},
1504 'hasvid
': {'priority': True, 'field': 'vcodec', 'type': 'boolean', 'not_in_list': ('none',)},
1505 'hasaud
': {'field': 'acodec', 'type': 'boolean', 'not_in_list': ('none',)},
1506 'lang
': {'priority': True, 'convert': 'ignore', 'field': 'language_preference'},
1507 'quality
': {'convert': 'float_none', 'default': -1},
1508 'filesize
': {'convert': 'bytes'},
1509 'fs_approx
': {'convert': 'bytes', 'field': 'filesize_approx'},
1510 'id': {'convert': 'string', 'field': 'format_id'},
1511 'height
': {'convert': 'float_none'},
1512 'width
': {'convert': 'float_none'},
1513 'fps
': {'convert': 'float_none'},
1514 'tbr
': {'convert': 'float_none'},
1515 'vbr
': {'convert': 'float_none'},
1516 'abr
': {'convert': 'float_none'},
1517 'asr
': {'convert': 'float_none'},
1518 'source
': {'convert': 'ignore', 'field': 'source_preference'},
1520 'codec
': {'type': 'combined', 'field': ('vcodec', 'acodec')},
1521 'br
': {'type': 'combined', 'field': ('tbr', 'vbr', 'abr'), 'same_limit': True},
1522 'size
': {'type': 'combined', 'same_limit': True, 'field': ('filesize', 'fs_approx')},
1523 'ext
': {'type': 'combined', 'field': ('vext', 'aext')},
1524 'res
': {'type': 'multiple
', 'field
': ('height
', 'width
'),
1525 'function
': lambda it: (lambda l: min(l) if l else 0)(tuple(filter(None, it)))},
1527 # Most of these exist only for compatibility reasons
1528 'dimension
': {'type': 'alias', 'field': 'res'},
1529 'resolution
': {'type': 'alias', 'field': 'res'},
1530 'extension
': {'type': 'alias', 'field': 'ext'},
1531 'bitrate
': {'type': 'alias', 'field': 'br'},
1532 'total_bitrate
': {'type': 'alias', 'field': 'tbr'},
1533 'video_bitrate
': {'type': 'alias', 'field': 'vbr'},
1534 'audio_bitrate
': {'type': 'alias', 'field': 'abr'},
1535 'framerate
': {'type': 'alias', 'field': 'fps'},
1536 'language_preference
': {'type': 'alias', 'field': 'lang'}, # not named as 'language
' because such a field exists
1537 'protocol
': {'type': 'alias', 'field': 'proto'},
1538 'source_preference
': {'type': 'alias', 'field': 'source'},
1539 'filesize_approx
': {'type': 'alias', 'field': 'fs_approx'},
1540 'filesize_estimate
': {'type': 'alias', 'field': 'size'},
1541 'samplerate
': {'type': 'alias', 'field': 'asr'},
1542 'video_ext
': {'type': 'alias', 'field': 'vext'},
1543 'audio_ext
': {'type': 'alias', 'field': 'aext'},
1544 'video_codec
': {'type': 'alias', 'field': 'vcodec'},
1545 'audio_codec
': {'type': 'alias', 'field': 'acodec'},
1546 'video
': {'type': 'alias', 'field': 'hasvid'},
1547 'has_video
': {'type': 'alias', 'field': 'hasvid'},
1548 'audio
': {'type': 'alias', 'field': 'hasaud'},
1549 'has_audio
': {'type': 'alias', 'field': 'hasaud'},
1550 'extractor
': {'type': 'alias', 'field': 'ie_pref'},
1551 'preference
': {'type': 'alias', 'field': 'ie_pref'},
1552 'extractor_preference
': {'type': 'alias', 'field': 'ie_pref'},
1553 'format_id
': {'type': 'alias', 'field': 'id'},
1558 def _get_field_setting(self, field, key):
1559 if field not in self.settings:
1560 self.settings[field] = {}
1561 propObj = self.settings[field]
1562 if key not in propObj:
1563 type = propObj.get('type')
1565 default = 'preference
' if type == 'extractor
' else (field,) if type in ('combined
', 'multiple
') else field
1566 elif key == 'convert
':
1567 default = 'order
' if type == 'ordered
' else 'float_string
' if field else 'ignore
'
1569 default = {'type': 'field', 'visible': True, 'order': [], 'not_in_list': (None,)}.get(key, None)
1570 propObj[key] = default
1573 def _resolve_field_value(self, field, value, convertNone=False):
1578 value = value.lower()
1579 conversion = self._get_field_setting(field, 'convert
')
1580 if conversion == 'ignore
':
1582 if conversion == 'string
':
1584 elif conversion == 'float_none
':
1585 return float_or_none(value)
1586 elif conversion == 'bytes':
1587 return FileDownloader.parse_bytes(value)
1588 elif conversion == 'order
':
1589 order_list = (self._use_free_order and self._get_field_setting(field, 'order_free
')) or self._get_field_setting(field, 'order
')
1590 use_regex = self._get_field_setting(field, 'regex
')
1591 list_length = len(order_list)
1592 empty_pos = order_list.index('') if '' in order_list else list_length + 1
1593 if use_regex and value is not None:
1594 for i, regex in enumerate(order_list):
1595 if regex and re.match(regex, value):
1596 return list_length - i
1597 return list_length - empty_pos # not in list
1598 else: # not regex or value = None
1599 return list_length - (order_list.index(value) if value in order_list else empty_pos)
1601 if value.isnumeric():
1604 self.settings[field]['convert
'] = 'string
'
1607 def evaluate_params(self, params, sort_extractor):
1608 self._use_free_order = params.get('prefer_free_formats
', False)
1609 self._sort_user = params.get('format_sort
', [])
1610 self._sort_extractor = sort_extractor
1612 def add_item(field, reverse, closest, limit_text):
1613 field = field.lower()
1614 if field in self._order:
1616 self._order.append(field)
1617 limit = self._resolve_field_value(field, limit_text)
1620 'closest
': False if limit is None else closest,
1621 'limit_text
': limit_text,
1623 if field in self.settings:
1624 self.settings[field].update(data)
1626 self.settings[field] = data
1629 tuple(field for field in self.default if self._get_field_setting(field, 'forced
'))
1630 + (tuple() if params.get('format_sort_force
', False)
1631 else tuple(field for field in self.default if self._get_field_setting(field, 'priority
')))
1632 + tuple(self._sort_user) + tuple(sort_extractor) + self.default)
1634 for item in sort_list:
1635 match = re.match(self.regex, item)
1637 raise ExtractorError('Invalid format sort string
"%s" given by extractor
' % item)
1638 field = match.group('field
')
1641 if self._get_field_setting(field, 'type') == 'alias
':
1642 field = self._get_field_setting(field, 'field
')
1643 reverse = match.group('reverse
') is not None
1644 closest = match.group('separator
') == '~
'
1645 limit_text = match.group('limit
')
1647 has_limit = limit_text is not None
1648 has_multiple_fields = self._get_field_setting(field, 'type') == 'combined
'
1649 has_multiple_limits = has_limit and has_multiple_fields and not self._get_field_setting(field, 'same_limit
')
1651 fields = self._get_field_setting(field, 'field
') if has_multiple_fields else (field,)
1652 limits = limit_text.split(":") if has_multiple_limits else (limit_text,) if has_limit else tuple()
1653 limit_count = len(limits)
1654 for (i, f) in enumerate(fields):
1655 add_item(f, reverse, closest,
1656 limits[i] if i < limit_count
1657 else limits[0] if has_limit and not has_multiple_limits
1660 def print_verbose_info(self, write_debug):
1662 write_debug('Sort order given by user
: %s' % ', '.join(self._sort_user))
1663 if self._sort_extractor:
1664 write_debug('Sort order given by extractor
: %s' % ', '.join(self._sort_extractor))
1665 write_debug('Formats
sorted by
: %s' % ', '.join(['%s%s%s' % (
1666 '+' if self._get_field_setting(field, 'reverse
') else '', field,
1667 '%s%s(%s)' % ('~
' if self._get_field_setting(field, 'closest
') else ':',
1668 self._get_field_setting(field, 'limit_text
'),
1669 self._get_field_setting(field, 'limit
'))
1670 if self._get_field_setting(field, 'limit_text
') is not None else '')
1671 for field in self._order if self._get_field_setting(field, 'visible
')]))
1673 def _calculate_field_preference_from_value(self, format, field, type, value):
1674 reverse = self._get_field_setting(field, 'reverse
')
1675 closest = self._get_field_setting(field, 'closest
')
1676 limit = self._get_field_setting(field, 'limit
')
1678 if type == 'extractor
':
1679 maximum = self._get_field_setting(field, 'max')
1680 if value is None or (maximum is not None and value >= maximum):
1682 elif type == 'boolean
':
1683 in_list = self._get_field_setting(field, 'in_list
')
1684 not_in_list = self._get_field_setting(field, 'not_in_list
')
1685 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
1686 elif type == 'ordered
':
1687 value = self._resolve_field_value(field, value, True)
1689 # try to convert to number
1690 val_num = float_or_none(value, default=self._get_field_setting(field, 'default
'))
1691 is_num = self._get_field_setting(field, 'convert
') != 'string
' and val_num is not None
1695 return ((-10, 0) if value is None
1696 else (1, value, 0) if not is_num # if a field has mixed strings and numbers, strings are sorted higher
1697 else (0, -abs(value - limit), value - limit if reverse else limit - value) if closest
1698 else (0, value, 0) if not reverse and (limit is None or value <= limit)
1699 else (0, -value, 0) if limit is None or (reverse and value == limit) or value > limit
1700 else (-1, value, 0))
1702 def _calculate_field_preference(self, format, field):
1703 type = self._get_field_setting(field, 'type') # extractor, boolean, ordered, field, multiple
1704 get_value = lambda f: format.get(self._get_field_setting(f, 'field
'))
1705 if type == 'multiple
':
1706 type = 'field
' # Only 'field
' is allowed in multiple for now
1707 actual_fields = self._get_field_setting(field, 'field
')
1709 value = self._get_field_setting(field, 'function
')(get_value(f) for f in actual_fields)
1711 value = get_value(field)
1712 return self._calculate_field_preference_from_value(format, field, type, value)
1714 def calculate_preference(self, format):
1715 # Determine missing protocol
1716 if not format.get('protocol
'):
1717 format['protocol
'] = determine_protocol(format)
1719 # Determine missing ext
1720 if not format.get('ext
') and 'url
' in format:
1721 format['ext
'] = determine_ext(format['url
'])
1722 if format.get('vcodec
') == 'none
':
1723 format['audio_ext
'] = format['ext
'] if format.get('acodec
') != 'none
' else 'none
'
1724 format['video_ext
'] = 'none
'
1726 format['video_ext
'] = format['ext
']
1727 format['audio_ext
'] = 'none
'
1728 # if format.get('preference
') is None and format.get('ext
') in ('f4f
', 'f4m
'): # Not supported?
1729 # format['preference
'] = -1000
1731 # Determine missing bitrates
1732 if format.get('tbr
') is None:
1733 if format.get('vbr
') is not None and format.get('abr
') is not None:
1734 format['tbr
'] = format.get('vbr
', 0) + format.get('abr
', 0)
1736 if format.get('vcodec
') != "none" and format.get('vbr
') is None:
1737 format['vbr
'] = format.get('tbr
') - format.get('abr
', 0)
1738 if format.get('acodec
') != "none" and format.get('abr
') is None:
1739 format['abr
'] = format.get('tbr
') - format.get('vbr
', 0)
1741 return tuple(self._calculate_field_preference(format, field) for field in self._order)
1743 def _sort_formats(self, formats, field_preference=[]):
1745 if self.get_param('ignore_no_formats_error
'):
1747 raise ExtractorError('No video formats found
')
1748 format_sort = self.FormatSort() # params and to_screen are taken from the downloader
1749 format_sort.evaluate_params(self._downloader.params, field_preference)
1750 if self.get_param('verbose
', False):
1751 format_sort.print_verbose_info(self._downloader.write_debug)
1752 formats.sort(key=lambda f: format_sort.calculate_preference(f))
1754 def _check_formats(self, formats, video_id):
1756 formats[:] = filter(
1757 lambda f: self._is_valid_url(
1759 item='%s video format
' % f.get('format_id
') if f.get('format_id
') else 'video
'),
1763 def _remove_duplicate_formats(formats):
1767 if f['url
'] not in format_urls:
1768 format_urls.add(f['url
'])
1769 unique_formats.append(f)
1770 formats[:] = unique_formats
1772 def _is_valid_url(self, url, video_id, item='video
', headers={}):
1773 url = self._proto_relative_url(url, scheme='http
:')
1774 # For now assume non HTTP(S) URLs always valid
1775 if not (url.startswith('http
://') or url.startswith('https
://')):
1778 self._request_webpage(url, video_id, 'Checking
%s URL
' % item, headers=headers)
1780 except ExtractorError as e:
1782 '%s: %s URL
is invalid
, skipping
: %s'
1783 % (video_id, item, error_to_compat_str(e.cause)))
1786 def http_scheme(self):
1787 """ Either "http:" or "https:", depending on the user's preferences
"""
1790 if self.get_param('prefer_insecure', False)
1793 def _proto_relative_url(self, url, scheme=None):
1796 if url.startswith('//'):
1798 scheme = self.http_scheme()
1803 def _sleep(self, timeout, video_id, msg_template=None):
1804 if msg_template is None:
1805 msg_template = '%(video_id)s: Waiting for %(timeout)s seconds'
1806 msg = msg_template % {'video_id': video_id, 'timeout': timeout}
1810 def _extract_f4m_formats(self, manifest_url, video_id, preference=None, quality=None, f4m_id=None,
1811 transform_source=lambda s: fix_xml_ampersands(s).strip(),
1812 fatal=True, m3u8_id=None, data=None, headers={}, query={}):
1813 manifest = self._download_xml(
1814 manifest_url, video_id, 'Downloading f4m manifest',
1815 'Unable to download f4m manifest',
1816 # Some manifests may be malformed, e.g. prosiebensat1 generated manifests
1817 # (see https://github.com/ytdl-org/youtube-dl/issues/6215#issuecomment-121704244)
1818 transform_source=transform_source,
1819 fatal=fatal, data=data, headers=headers, query=query)
1821 if manifest is False:
1824 return self._parse_f4m_formats(
1825 manifest, manifest_url, video_id, preference=preference, quality=quality, f4m_id=f4m_id,
1826 transform_source=transform_source, fatal=fatal, m3u8_id=m3u8_id)
1828 def _parse_f4m_formats(self, manifest, manifest_url, video_id, preference=None, quality=None, f4m_id=None,
1829 transform_source=lambda s: fix_xml_ampersands(s).strip(),
1830 fatal=True, m3u8_id=None):
1831 if not isinstance(manifest, compat_etree_Element) and not fatal:
1834 # currently yt-dlp cannot decode the playerVerificationChallenge as Akamai uses Adobe Alchemy
1835 akamai_pv = manifest.find('{http://ns.adobe.com/f4m/1.0}pv-2.0')
1836 if akamai_pv is not None and ';' in akamai_pv.text:
1837 playerVerificationChallenge = akamai_pv.text.split(';')[0]
1838 if playerVerificationChallenge.strip() != '':
1842 manifest_version = '1.0'
1843 media_nodes = manifest.findall('{http://ns.adobe.com/f4m/1.0}media')
1845 manifest_version = '2.0'
1846 media_nodes = manifest.findall('{http://ns.adobe.com/f4m/2.0}media')
1847 # Remove unsupported DRM protected media from final formats
1848 # rendition (see https://github.com/ytdl-org/youtube-dl/issues/8573).
1849 media_nodes = remove_encrypted_media(media_nodes)
1853 manifest_base_url = get_base_url(manifest)
1855 bootstrap_info = xpath_element(
1856 manifest, ['{http://ns.adobe.com/f4m/1.0}bootstrapInfo', '{http://ns.adobe.com/f4m/2.0}bootstrapInfo'],
1857 'bootstrap info', default=None)
1860 mime_type = xpath_text(
1861 manifest, ['{http://ns.adobe.com/f4m/1.0}mimeType', '{http://ns.adobe.com/f4m/2.0}mimeType'],
1862 'base URL', default=None)
1863 if mime_type and mime_type.startswith('audio/'):
1866 for i, media_el in enumerate(media_nodes):
1867 tbr = int_or_none(media_el.attrib.get('bitrate'))
1868 width = int_or_none(media_el.attrib.get('width'))
1869 height = int_or_none(media_el.attrib.get('height'))
1870 format_id = '-'.join(filter(None, [f4m_id, compat_str(i if tbr is None else tbr)]))
1871 # If <bootstrapInfo> is present, the specified f4m is a
1872 # stream-level manifest, and only set-level manifests may refer to
1873 # external resources. See section 11.4 and section 4 of F4M spec
1874 if bootstrap_info is None:
1876 # @href is introduced in 2.0, see section 11.6 of F4M spec
1877 if manifest_version == '2.0':
1878 media_url = media_el.attrib.get('href')
1879 if media_url is None:
1880 media_url = media_el.attrib.get('url')
1884 media_url if media_url.startswith('http://') or media_url.startswith('https://')
1885 else ((manifest_base_url or '/'.join(manifest_url.split('/')[:-1])) + '/' + media_url))
1886 # If media_url is itself a f4m manifest do the recursive extraction
1887 # since bitrates in parent manifest (this one) and media_url manifest
1888 # may differ leading to inability to resolve the format by requested
1889 # bitrate in f4m downloader
1890 ext = determine_ext(manifest_url)
1892 f4m_formats = self._extract_f4m_formats(
1893 manifest_url, video_id, preference=preference, quality=quality, f4m_id=f4m_id,
1894 transform_source=transform_source, fatal=fatal)
1895 # Sometimes stream-level manifest contains single media entry that
1896 # does not contain any quality metadata (e.g. http://matchtv.ru/#live-player).
1897 # At the same time parent's media entry in set-level manifest may
1898 # contain it. We will copy it from parent in such cases.
1899 if len(f4m_formats) == 1:
1902 'tbr': f.get('tbr') or tbr,
1903 'width': f.get('width') or width,
1904 'height': f.get('height') or height,
1905 'format_id': f.get('format_id') if not tbr else format_id,
1908 formats.extend(f4m_formats)
1911 formats.extend(self._extract_m3u8_formats(
1912 manifest_url, video_id, 'mp4', preference=preference,
1913 quality=quality, m3u8_id=m3u8_id, fatal=fatal))
1916 'format_id': format_id,
1917 'url': manifest_url,
1918 'manifest_url': manifest_url,
1919 'ext': 'flv' if bootstrap_info is not None else None,
1925 'preference': preference,
1930 def _m3u8_meta_format(self, m3u8_url, ext=None, preference=None, quality=None, m3u8_id=None):
1932 'format_id': '-'.join(filter(None, [m3u8_id, 'meta'])),
1936 'preference': preference - 100 if preference else -100,
1938 'resolution': 'multiple',
1939 'format_note': 'Quality selection URL',
1942 def _extract_m3u8_formats(self, *args, **kwargs):
1943 fmts, subs = self._extract_m3u8_formats_and_subtitles(*args, **kwargs)
1945 self.report_warning(bug_reports_message(
1946 "Ignoring subtitle tracks found in the HLS manifest; "
1947 "if any subtitle tracks are missing,"
1951 def _extract_m3u8_formats_and_subtitles(
1952 self, m3u8_url, video_id, ext=None, entry_protocol='m3u8_native',
1953 preference=None, quality=None, m3u8_id=None, note=None,
1954 errnote=None, fatal=True, live=False, data=None, headers={},
1957 res = self._download_webpage_handle(
1959 note='Downloading m3u8 information' if note is None else note,
1960 errnote='Failed to download m3u8 information' if errnote is None else errnote,
1961 fatal=fatal, data=data, headers=headers, query=query)
1966 m3u8_doc, urlh = res
1967 m3u8_url = urlh.geturl()
1969 return self._parse_m3u8_formats_and_subtitles(
1970 m3u8_doc, m3u8_url, ext=ext, entry_protocol=entry_protocol,
1971 preference=preference, quality=quality, m3u8_id=m3u8_id,
1972 note=note, errnote=errnote, fatal=fatal, live=live, data=data,
1973 headers=headers, query=query, video_id=video_id)
1975 def _parse_m3u8_formats_and_subtitles(
1976 self, m3u8_doc, m3u8_url, ext=None, entry_protocol='m3u8_native',
1977 preference=None, quality=None, m3u8_id=None, live=False, note=None,
1978 errnote=None, fatal=True, data=None, headers={}, query={},
1980 formats, subtitles = [], {}
1982 if '#EXT-X-FAXS-CM:' in m3u8_doc: # Adobe Flash Access
1983 return formats, subtitles
1985 if (not self.get_param('allow_unplayable_formats')
1986 and re.search(r'#EXT-X-SESSION-KEY:.*?URI="skd://', m3u8_doc)): # Apple FairPlay
1987 return formats, subtitles
1989 def format_url(url):
1990 return url if re.match(r'^https?://', url) else compat_urlparse.urljoin(m3u8_url, url)
1992 if self.get_param('hls_split_discontinuity', False):
1993 def _extract_m3u8_playlist_indices(manifest_url=None, m3u8_doc=None):
1995 if not manifest_url:
1997 m3u8_doc = self._download_webpage(
1998 manifest_url, video_id, fatal=fatal, data=data, headers=headers,
1999 note=False, errnote='Failed to download m3u8 playlist information')
2000 if m3u8_doc is False:
2002 return range(1 + sum(line.startswith('#EXT-X-DISCONTINUITY') for line in m3u8_doc.splitlines()))
2005 def _extract_m3u8_playlist_indices(*args, **kwargs):
2009 # 1. https://tools.ietf.org/html/draft-pantos-http-live-streaming-21
2010 # 2. https://github.com/ytdl-org/youtube-dl/issues/12211
2011 # 3. https://github.com/ytdl-org/youtube-dl/issues/18923
2013 # We should try extracting formats only from master playlists [1, 4.3.4],
2014 # i.e. playlists that describe available qualities. On the other hand
2015 # media playlists [1, 4.3.3] should be returned as is since they contain
2016 # just the media without qualities renditions.
2017 # Fortunately, master playlist can be easily distinguished from media
2018 # playlist based on particular tags availability. As of [1, 4.3.3, 4.3.4]
2019 # master playlist tags MUST NOT appear in a media playlist and vice versa.
2020 # As of [1, 4.3.3.1] #EXT-X-TARGETDURATION tag is REQUIRED for every
2021 # media playlist and MUST NOT appear in master playlist thus we can
2022 # clearly detect media playlist with this criterion.
2024 if '#EXT-X-TARGETDURATION' in m3u8_doc: # media playlist, return as is
2026 'format_id': '-'.join(map(str, filter(None, [m3u8_id, idx]))),
2027 'format_index': idx,
2030 'protocol': entry_protocol,
2031 'preference': preference,
2033 } for idx in _extract_m3u8_playlist_indices(m3u8_doc=m3u8_doc)]
2035 return formats, subtitles
2038 last_stream_inf = {}
2040 def extract_media(x_media_line):
2041 media = parse_m3u8_attributes(x_media_line)
2042 # As per [1, 4.3.4.1] TYPE, GROUP-ID and NAME are REQUIRED
2043 media_type, group_id, name = media.get('TYPE'), media.get('GROUP-ID'), media.get('NAME')
2044 if not (media_type and group_id and name):
2046 groups.setdefault(group_id, []).append(media)
2047 # <https://tools.ietf.org/html/rfc8216#section-4.3.4.1>
2048 if media_type == 'SUBTITLES':
2049 # According to RFC 8216 §4.3.4.2.1, URI is REQUIRED in the
2050 # EXT-X-MEDIA tag if the media type is SUBTITLES.
2051 # However, lack of URI has been spotted in the wild.
2052 # e.g. NebulaIE; see https://github.com/yt-dlp/yt-dlp/issues/339
2053 if not media.get('URI'):
2055 url = format_url(media['URI'])
2058 'ext': determine_ext(url),
2060 if sub_info['ext'] == 'm3u8':
2061 # Per RFC 8216 §3.1, the only possible subtitle format m3u8
2062 # files may contain is WebVTT:
2063 # <https://tools.ietf.org/html/rfc8216#section-3.1>
2064 sub_info['ext'] = 'vtt'
2065 sub_info['protocol'] = 'm3u8_native'
2066 lang = media.get('LANGUAGE') or 'und'
2067 subtitles.setdefault(lang, []).append(sub_info)
2068 if media_type not in ('VIDEO', 'AUDIO'):
2070 media_url = media.get('URI')
2072 manifest_url = format_url(media_url)
2074 'format_id': '-'.join(map(str, filter(None, (m3u8_id, group_id, name, idx)))),
2075 'format_note': name,
2076 'format_index': idx,
2077 'url': manifest_url,
2078 'manifest_url': m3u8_url,
2079 'language': media.get('LANGUAGE'),
2081 'protocol': entry_protocol,
2082 'preference': preference,
2084 'vcodec': 'none' if media_type == 'AUDIO' else None,
2085 } for idx in _extract_m3u8_playlist_indices(manifest_url))
2087 def build_stream_name():
2088 # Despite specification does not mention NAME attribute for
2089 # EXT-X-STREAM-INF tag it still sometimes may be present (see [1]
2090 # or vidio test in TestInfoExtractor.test_parse_m3u8_formats)
2091 # 1. http://www.vidio.com/watch/165683-dj_ambred-booyah-live-2015
2092 stream_name = last_stream_inf.get('NAME')
2095 # If there is no NAME in EXT-X-STREAM-INF it will be obtained
2096 # from corresponding rendition group
2097 stream_group_id = last_stream_inf.get('VIDEO')
2098 if not stream_group_id:
2100 stream_group = groups.get(stream_group_id)
2101 if not stream_group:
2102 return stream_group_id
2103 rendition = stream_group[0]
2104 return rendition.get('NAME') or stream_group_id
2106 # parse EXT-X-MEDIA tags before EXT-X-STREAM-INF in order to have the
2107 # chance to detect video only formats when EXT-X-STREAM-INF tags
2108 # precede EXT-X-MEDIA tags in HLS manifest such as [3].
2109 for line in m3u8_doc.splitlines():
2110 if line.startswith('#EXT-X-MEDIA:'):
2113 for line in m3u8_doc.splitlines():
2114 if line.startswith('#EXT-X-STREAM-INF:'):
2115 last_stream_inf = parse_m3u8_attributes(line)
2116 elif line.startswith('#') or not line.strip():
2119 tbr = float_or_none(
2120 last_stream_inf.get('AVERAGE-BANDWIDTH')
2121 or last_stream_inf.get('BANDWIDTH'), scale=1000)
2122 manifest_url = format_url(line.strip())
2124 for idx in _extract_m3u8_playlist_indices(manifest_url):
2125 format_id = [m3u8_id, None, idx]
2126 # Bandwidth of live streams may differ over time thus making
2127 # format_id unpredictable. So it's better to keep provided
2130 stream_name = build_stream_name()
2131 format_id[1] = stream_name if stream_name else '%d' % (tbr if tbr else len(formats))
2133 'format_id': '-'.join(map(str, filter(None, format_id))),
2134 'format_index': idx,
2135 'url': manifest_url,
2136 'manifest_url': m3u8_url,
2139 'fps': float_or_none(last_stream_inf.get('FRAME-RATE')),
2140 'protocol': entry_protocol,
2141 'preference': preference,
2144 resolution = last_stream_inf.get('RESOLUTION')
2146 mobj = re.search(r'(?P<width>\d+)[xX](?P<height>\d+)', resolution)
2148 f['width'] = int(mobj.group('width'))
2149 f['height'] = int(mobj.group('height'))
2150 # Unified Streaming Platform
2152 r'audio.*?(?:%3D|=)(\d+)(?:-video.*?(?:%3D|=)(\d+))?', f['url'])
2154 abr, vbr = mobj.groups()
2155 abr, vbr = float_or_none(abr, 1000), float_or_none(vbr, 1000)
2160 codecs = parse_codecs(last_stream_inf.get('CODECS'))
2162 audio_group_id = last_stream_inf.get('AUDIO')
2163 # As per [1, 4.3.4.1.1] any EXT-X-STREAM-INF tag which
2164 # references a rendition group MUST have a CODECS attribute.
2165 # However, this is not always respected, for example, [2]
2166 # contains EXT-X-STREAM-INF tag which references AUDIO
2167 # rendition group but does not have CODECS and despite
2168 # referencing an audio group it represents a complete
2169 # (with audio and video) format. So, for such cases we will
2170 # ignore references to rendition groups and treat them
2171 # as complete formats.
2172 if audio_group_id and codecs and f.get('vcodec') != 'none':
2173 audio_group = groups.get(audio_group_id)
2174 if audio_group and audio_group[0].get('URI'):
2175 # TODO: update acodec for audio only formats with
2177 f['acodec'] = 'none'
2178 if not f.get('ext'):
2179 f['ext'] = 'm4a' if f.get('vcodec') == 'none' else 'mp4'
2183 progressive_uri = last_stream_inf.get('PROGRESSIVE-URI')
2186 del http_f['manifest_url']
2188 'format_id': f['format_id'].replace('hls-', 'http-'),
2190 'url': progressive_uri,
2192 formats.append(http_f)
2194 last_stream_inf = {}
2195 return formats, subtitles
2198 def _xpath_ns(path, namespace=None):
2202 for c in path.split('/'):
2203 if not c or c == '.':
2206 out.append('{%s}%s' % (namespace, c))
2207 return '/'.join(out)
2209 def _extract_smil_formats_and_subtitles(self, smil_url, video_id, fatal=True, f4m_params=None, transform_source=None):
2210 smil = self._download_smil(smil_url, video_id, fatal=fatal, transform_source=transform_source)
2216 namespace = self._parse_smil_namespace(smil)
2218 fmts = self._parse_smil_formats(
2219 smil, smil_url, video_id, namespace=namespace, f4m_params=f4m_params)
2220 subs = self._parse_smil_subtitles(
2221 smil, namespace=namespace)
2225 def _extract_smil_formats(self, *args, **kwargs):
2226 fmts, subs = self._extract_smil_formats_and_subtitles(*args, **kwargs)
2228 self.report_warning(bug_reports_message(
2229 "Ignoring subtitle tracks found in the SMIL manifest; "
2230 "if any subtitle tracks are missing,"
2234 def _extract_smil_info(self, smil_url, video_id, fatal=True, f4m_params=None):
2235 smil = self._download_smil(smil_url, video_id, fatal=fatal)
2238 return self._parse_smil(smil, smil_url, video_id, f4m_params=f4m_params)
2240 def _download_smil(self, smil_url, video_id, fatal=True, transform_source=None):
2241 return self._download_xml(
2242 smil_url, video_id, 'Downloading SMIL file',
2243 'Unable to download SMIL file', fatal=fatal, transform_source=transform_source)
2245 def _parse_smil(self, smil, smil_url, video_id, f4m_params=None):
2246 namespace = self._parse_smil_namespace(smil)
2248 formats = self._parse_smil_formats(
2249 smil, smil_url, video_id, namespace=namespace, f4m_params=f4m_params)
2250 subtitles = self._parse_smil_subtitles(smil, namespace=namespace)
2252 video_id = os.path.splitext(url_basename(smil_url))[0]
2256 for meta in smil.findall(self._xpath_ns('./head/meta', namespace)):
2257 name = meta.attrib.get('name')
2258 content = meta.attrib.get('content')
2259 if not name or not content:
2261 if not title and name == 'title':
2263 elif not description and name in ('description', 'abstract'):
2264 description = content
2265 elif not upload_date and name == 'date':
2266 upload_date = unified_strdate(content)
2269 'id': image.get('type'),
2270 'url': image.get('src'),
2271 'width': int_or_none(image.get('width')),
2272 'height': int_or_none(image.get('height')),
2273 } for image in smil.findall(self._xpath_ns('.//image', namespace)) if image.get('src')]
2277 'title': title or video_id,
2278 'description': description,
2279 'upload_date': upload_date,
2280 'thumbnails': thumbnails,
2282 'subtitles': subtitles,
2285 def _parse_smil_namespace(self, smil):
2286 return self._search_regex(
2287 r'(?i)^{([^}]+)?}smil$', smil.tag, 'namespace', default=None)
2289 def _parse_smil_formats(self, smil, smil_url, video_id, namespace=None, f4m_params=None, transform_rtmp_url=None):
2291 for meta in smil.findall(self._xpath_ns('./head/meta', namespace)):
2292 b = meta.get('base') or meta.get('httpBase')
2303 media = smil.findall(self._xpath_ns('.//video', namespace)) + smil.findall(self._xpath_ns('.//audio', namespace))
2304 for medium in media:
2305 src = medium.get('src')
2306 if not src or src in srcs:
2310 bitrate = float_or_none(medium.get('system-bitrate') or medium.get('systemBitrate'), 1000)
2311 filesize = int_or_none(medium.get('size') or medium.get('fileSize'))
2312 width = int_or_none(medium.get('width'))
2313 height = int_or_none(medium.get('height'))
2314 proto = medium.get('proto')
2315 ext = medium.get('ext')
2316 src_ext = determine_ext(src)
2317 streamer = medium.get('streamer') or base
2319 if proto == 'rtmp' or streamer.startswith('rtmp'):
2325 'format_id': 'rtmp-%d' % (rtmp_count if bitrate is None else bitrate),
2327 'filesize': filesize,
2331 if transform_rtmp_url:
2332 streamer, src = transform_rtmp_url(streamer, src)
2333 formats[-1].update({
2339 src_url = src if src.startswith('http') else compat_urlparse.urljoin(base, src)
2340 src_url = src_url.strip()
2342 if proto == 'm3u8' or src_ext == 'm3u8':
2343 m3u8_formats = self._extract_m3u8_formats(
2344 src_url, video_id, ext or 'mp4', m3u8_id='hls', fatal=False)
2345 if len(m3u8_formats) == 1:
2347 m3u8_formats[0].update({
2348 'format_id': 'hls-%d' % (m3u8_count if bitrate is None else bitrate),
2353 formats.extend(m3u8_formats)
2354 elif src_ext == 'f4m':
2359 'plugin': 'flowplayer-3.2.0.1',
2361 f4m_url += '&' if '?' in f4m_url else '?'
2362 f4m_url += compat_urllib_parse_urlencode(f4m_params)
2363 formats.extend(self._extract_f4m_formats(f4m_url, video_id, f4m_id='hds', fatal=False))
2364 elif src_ext == 'mpd':
2365 formats.extend(self._extract_mpd_formats(
2366 src_url, video_id, mpd_id='dash', fatal=False))
2367 elif re.search(r'\.ism/[Mm]anifest', src_url):
2368 formats.extend(self._extract_ism_formats(
2369 src_url, video_id, ism_id='mss', fatal=False))
2370 elif src_url.startswith('http') and self._is_valid_url(src, video_id):
2374 'ext': ext or src_ext or 'flv',
2375 'format_id': 'http-%d' % (bitrate or http_count),
2377 'filesize': filesize,
2384 def _parse_smil_subtitles(self, smil, namespace=None, subtitles_lang='en'):
2387 for num, textstream in enumerate(smil.findall(self._xpath_ns('.//textstream', namespace))):
2388 src = textstream.get('src')
2389 if not src or src in urls:
2392 ext = textstream.get('ext') or mimetype2ext(textstream.get('type')) or determine_ext(src)
2393 lang = textstream.get('systemLanguage') or textstream.get('systemLanguageName') or textstream.get('lang') or subtitles_lang
2394 subtitles.setdefault(lang, []).append({
2400 def _extract_xspf_playlist(self, xspf_url, playlist_id, fatal=True):
2401 xspf = self._download_xml(
2402 xspf_url, playlist_id, 'Downloading xpsf playlist',
2403 'Unable to download xspf manifest', fatal=fatal)
2406 return self._parse_xspf(
2407 xspf, playlist_id, xspf_url=xspf_url,
2408 xspf_base_url=base_url(xspf_url))
2410 def _parse_xspf(self, xspf_doc, playlist_id, xspf_url=None, xspf_base_url=None):
2412 'xspf': 'http://xspf.org/ns/0/',
2413 's1': 'http://static.streamone.nl/player/ns/0',
2417 for track in xspf_doc.findall(xpath_with_ns('./xspf:trackList/xspf:track', NS_MAP)):
2419 track, xpath_with_ns('./xspf:title', NS_MAP), 'title', default=playlist_id)
2420 description = xpath_text(
2421 track, xpath_with_ns('./xspf:annotation', NS_MAP), 'description')
2422 thumbnail = xpath_text(
2423 track, xpath_with_ns('./xspf:image', NS_MAP), 'thumbnail')
2424 duration = float_or_none(
2425 xpath_text(track, xpath_with_ns('./xspf:duration', NS_MAP), 'duration'), 1000)
2428 for location in track.findall(xpath_with_ns('./xspf:location', NS_MAP)):
2429 format_url = urljoin(xspf_base_url, location.text)
2434 'manifest_url': xspf_url,
2435 'format_id': location.get(xpath_with_ns('s1:label', NS_MAP)),
2436 'width': int_or_none(location.get(xpath_with_ns('s1:width', NS_MAP))),
2437 'height': int_or_none(location.get(xpath_with_ns('s1:height', NS_MAP))),
2439 self._sort_formats(formats)
2444 'description': description,
2445 'thumbnail': thumbnail,
2446 'duration': duration,
2451 def _extract_mpd_formats(self, *args, **kwargs):
2452 fmts, subs = self._extract_mpd_formats_and_subtitles(*args, **kwargs)
2454 self.report_warning(bug_reports_message(
2455 "Ignoring subtitle tracks found in the DASH manifest; "
2456 "if any subtitle tracks are missing,"
2460 def _extract_mpd_formats_and_subtitles(
2461 self, mpd_url, video_id, mpd_id=None, note=None, errnote=None,
2462 fatal=True, data=None, headers={}, query={}):
2463 res = self._download_xml_handle(
2465 note='Downloading MPD manifest' if note is None else note,
2466 errnote='Failed to download MPD manifest' if errnote is None else errnote,
2467 fatal=fatal, data=data, headers=headers, query=query)
2473 mpd_base_url = base_url(urlh.geturl())
2475 return self._parse_mpd_formats_and_subtitles(
2476 mpd_doc, mpd_id, mpd_base_url, mpd_url)
2478 def _parse_mpd_formats(self, *args, **kwargs):
2479 fmts, subs = self._parse_mpd_formats_and_subtitles(*args, **kwargs)
2481 self.report_warning(bug_reports_message(
2482 "Ignoring subtitle tracks found in the DASH manifest; "
2483 "if any subtitle tracks are missing,"
2487 def _parse_mpd_formats_and_subtitles(
2488 self, mpd_doc, mpd_id=None, mpd_base_url='', mpd_url=None):
2490 Parse formats
from MPD manifest
.
2492 1. MPEG
-DASH Standard
, ISO
/IEC
23009-1:2014(E
),
2493 http
://standards
.iso
.org
/ittf
/PubliclyAvailableStandards
/c065274_ISO_IEC_23009
-1_2014.zip
2494 2. https
://en
.wikipedia
.org
/wiki
/Dynamic_Adaptive_Streaming_over_HTTP
2496 if not self.get_param('dynamic_mpd', True):
2497 if mpd_doc.get('type') == 'dynamic':
2500 namespace = self._search_regex(r'(?i)^{([^}]+)?}MPD$', mpd_doc.tag, 'namespace', default=None)
2503 return self._xpath_ns(path, namespace)
2505 def is_drm_protected(element):
2506 return element.find(_add_ns('ContentProtection')) is not None
2508 def extract_multisegment_info(element, ms_parent_info):
2509 ms_info = ms_parent_info.copy()
2511 # As per [1, 5.3.9.2.2] SegmentList and SegmentTemplate share some
2512 # common attributes and elements. We will only extract relevant
2514 def extract_common(source):
2515 segment_timeline = source.find(_add_ns('SegmentTimeline'))
2516 if segment_timeline is not None:
2517 s_e = segment_timeline.findall(_add_ns('S'))
2519 ms_info['total_number'] = 0
2522 r = int(s.get('r', 0))
2523 ms_info['total_number'] += 1 + r
2524 ms_info['s'].append({
2525 't': int(s.get('t', 0)),
2526 # @d is mandatory (see [1, 5.3.9.6.2, Table 17, page 60])
2527 'd': int(s.attrib['d']),
2530 start_number = source.get('startNumber')
2532 ms_info['start_number'] = int(start_number)
2533 timescale = source.get('timescale')
2535 ms_info['timescale'] = int(timescale)
2536 segment_duration = source.get('duration')
2537 if segment_duration:
2538 ms_info['segment_duration'] = float(segment_duration)
2540 def extract_Initialization(source):
2541 initialization = source.find(_add_ns('Initialization'))
2542 if initialization is not None:
2543 ms_info['initialization_url'] = initialization.attrib['sourceURL']
2545 segment_list = element.find(_add_ns('SegmentList'))
2546 if segment_list is not None:
2547 extract_common(segment_list)
2548 extract_Initialization(segment_list)
2549 segment_urls_e = segment_list.findall(_add_ns('SegmentURL'))
2551 ms_info['segment_urls'] = [segment.attrib['media'] for segment in segment_urls_e]
2553 segment_template = element.find(_add_ns('SegmentTemplate'))
2554 if segment_template is not None:
2555 extract_common(segment_template)
2556 media = segment_template.get('media')
2558 ms_info['media'] = media
2559 initialization = segment_template.get('initialization')
2561 ms_info['initialization'] = initialization
2563 extract_Initialization(segment_template)
2566 skip_unplayable = not self.get_param('allow_unplayable_formats')
2568 mpd_duration = parse_duration(mpd_doc.get('mediaPresentationDuration'))
2571 for period in mpd_doc.findall(_add_ns('Period')):
2572 period_duration = parse_duration(period.get('duration')) or mpd_duration
2573 period_ms_info = extract_multisegment_info(period, {
2577 for adaptation_set in period.findall(_add_ns('AdaptationSet')):
2578 if skip_unplayable and is_drm_protected(adaptation_set):
2580 adaption_set_ms_info = extract_multisegment_info(adaptation_set, period_ms_info)
2581 for representation in adaptation_set.findall(_add_ns('Representation')):
2582 if skip_unplayable and is_drm_protected(representation):
2584 representation_attrib = adaptation_set.attrib.copy()
2585 representation_attrib.update(representation.attrib)
2586 # According to [1, 5.3.7.2, Table 9, page 41], @mimeType is mandatory
2587 mime_type = representation_attrib['mimeType']
2588 content_type = representation_attrib.get('contentType', mime_type.split('/')[0])
2590 if content_type in ('video', 'audio', 'text') or mime_type == 'image/jpeg':
2592 for element in (representation, adaptation_set, period, mpd_doc):
2593 base_url_e = element.find(_add_ns('BaseURL'))
2594 if base_url_e is not None:
2595 base_url = base_url_e.text + base_url
2596 if re.match(r'^https?://', base_url):
2598 if mpd_base_url and not re.match(r'^https?://', base_url):
2599 if not mpd_base_url.endswith('/') and not base_url.startswith('/'):
2601 base_url = mpd_base_url + base_url
2602 representation_id = representation_attrib.get('id')
2603 lang = representation_attrib.get('lang')
2604 url_el = representation.find(_add_ns('BaseURL'))
2605 filesize = int_or_none(url_el.attrib.get('{http://youtube.com/yt/2012/10/10}contentLength') if url_el is not None else None)
2606 bandwidth = int_or_none(representation_attrib.get('bandwidth'))
2607 if representation_id is not None:
2608 format_id = representation_id
2610 format_id = content_type
2612 format_id = mpd_id + '-' + format_id
2613 if content_type in ('video', 'audio'):
2615 'format_id': format_id,
2616 'manifest_url': mpd_url,
2617 'ext': mimetype2ext(mime_type),
2618 'width': int_or_none(representation_attrib.get('width')),
2619 'height': int_or_none(representation_attrib.get('height')),
2620 'tbr': float_or_none(bandwidth, 1000),
2621 'asr': int_or_none(representation_attrib.get('audioSamplingRate')),
2622 'fps': int_or_none(representation_attrib.get('frameRate')),
2623 'language': lang if lang not in ('mul', 'und', 'zxx', 'mis') else None,
2624 'format_note': 'DASH %s' % content_type,
2625 'filesize': filesize,
2626 'container': mimetype2ext(mime_type) + '_dash',
2628 f.update(parse_codecs(representation_attrib.get('codecs')))
2629 elif content_type == 'text':
2631 'ext': mimetype2ext(mime_type),
2632 'manifest_url': mpd_url,
2633 'filesize': filesize,
2635 elif mime_type == 'image/jpeg':
2636 # See test case in VikiIE
2637 # https://www.viki.com/videos/1175236v-choosing-spouse-by-lottery-episode-1
2639 'format_id': format_id,
2641 'manifest_url': mpd_url,
2642 'format_note': 'DASH storyboards (jpeg)',
2646 representation_ms_info = extract_multisegment_info(representation, adaption_set_ms_info)
2648 def prepare_template(template_name, identifiers):
2649 tmpl = representation_ms_info[template_name]
2650 # First of, % characters outside $...$ templates
2651 # must be escaped by doubling for proper processing
2652 # by % operator string formatting used further (see
2653 # https://github.com/ytdl-org/youtube-dl/issues/16867).
2659 in_template = not in_template
2660 elif c == '%' and not in_template:
2662 # Next, $...$ templates are translated to their
2663 # %(...) counterparts to be used with % operator
2664 if representation_id is not None:
2665 t = t.replace('$RepresentationID$', representation_id)
2666 t = re.sub(r'\$(%s)\$' % '|'.join(identifiers), r'%(\1)d', t)
2667 t = re.sub(r'\$(%s)%%([^$]+)\$' % '|'.join(identifiers), r'%(\1)\2', t)
2668 t.replace('$$', '$')
2671 # @initialization is a regular template like @media one
2672 # so it should be handled just the same way (see
2673 # https://github.com/ytdl-org/youtube-dl/issues/11605)
2674 if 'initialization' in representation_ms_info:
2675 initialization_template = prepare_template(
2677 # As per [1, 5.3.9.4.2, Table 15, page 54] $Number$ and
2678 # $Time$ shall not be included for @initialization thus
2679 # only $Bandwidth$ remains
2681 representation_ms_info['initialization_url'] = initialization_template % {
2682 'Bandwidth': bandwidth,
2685 def location_key(location):
2686 return 'url' if re.match(r'^https?://', location) else 'path'
2688 if 'segment_urls' not in representation_ms_info and 'media' in representation_ms_info:
2690 media_template = prepare_template('media', ('Number', 'Bandwidth', 'Time'))
2691 media_location_key = location_key(media_template)
2693 # As per [1, 5.3.9.4.4, Table 16, page 55] $Number$ and $Time$
2694 # can't be used at the same time
2695 if '%(Number' in media_template and 's' not in representation_ms_info:
2696 segment_duration = None
2697 if 'total_number' not in representation_ms_info and 'segment_duration' in representation_ms_info:
2698 segment_duration = float_or_none(representation_ms_info['segment_duration'], representation_ms_info['timescale'])
2699 representation_ms_info['total_number'] = int(math.ceil(float(period_duration) / segment_duration))
2700 representation_ms_info['fragments'] = [{
2701 media_location_key: media_template % {
2702 'Number': segment_number,
2703 'Bandwidth': bandwidth,
2705 'duration': segment_duration,
2706 } for segment_number in range(
2707 representation_ms_info['start_number'],
2708 representation_ms_info['total_number'] + representation_ms_info['start_number'])]
2710 # $Number*$ or $Time$ in media template with S list available
2711 # Example $Number*$: http://www.svtplay.se/klipp/9023742/stopptid-om-bjorn-borg
2712 # Example $Time$: https://play.arkena.com/embed/avp/v2/player/media/b41dda37-d8e7-4d3f-b1b5-9a9db578bdfe/1/129411
2713 representation_ms_info['fragments'] = []
2716 segment_number = representation_ms_info['start_number']
2718 def add_segment_url():
2719 segment_url = media_template % {
2720 'Time': segment_time,
2721 'Bandwidth': bandwidth,
2722 'Number': segment_number,
2724 representation_ms_info['fragments'].append({
2725 media_location_key: segment_url,
2726 'duration': float_or_none(segment_d, representation_ms_info['timescale']),
2729 for num, s in enumerate(representation_ms_info['s']):
2730 segment_time = s.get('t') or segment_time
2734 for r in range(s.get('r', 0)):
2735 segment_time += segment_d
2738 segment_time += segment_d
2739 elif 'segment_urls' in representation_ms_info and 's' in representation_ms_info:
2741 # Example: https://www.youtube.com/watch?v=iXZV5uAYMJI
2742 # or any YouTube dashsegments video
2745 timescale = representation_ms_info['timescale']
2746 for s in representation_ms_info['s']:
2747 duration = float_or_none(s['d'], timescale)
2748 for r in range(s.get('r', 0) + 1):
2749 segment_uri = representation_ms_info['segment_urls'][segment_index]
2751 location_key(segment_uri): segment_uri,
2752 'duration': duration,
2755 representation_ms_info['fragments'] = fragments
2756 elif 'segment_urls' in representation_ms_info:
2757 # Segment URLs with no SegmentTimeline
2758 # Example: https://www.seznam.cz/zpravy/clanek/cesko-zasahne-vitr-o-sile-vichrice-muze-byt-i-zivotu-nebezpecny-39091
2759 # https://github.com/ytdl-org/youtube-dl/pull/14844
2761 segment_duration = float_or_none(
2762 representation_ms_info['segment_duration'],
2763 representation_ms_info['timescale']) if 'segment_duration' in representation_ms_info else None
2764 for segment_url in representation_ms_info['segment_urls']:
2766 location_key(segment_url): segment_url,
2768 if segment_duration:
2769 fragment['duration'] = segment_duration
2770 fragments.append(fragment)
2771 representation_ms_info['fragments'] = fragments
2772 # If there is a fragments key available then we correctly recognized fragmented media.
2773 # Otherwise we will assume unfragmented media with direct access. Technically, such
2774 # assumption is not necessarily correct since we may simply have no support for
2775 # some forms of fragmented media renditions yet, but for now we'll use this fallback.
2776 if 'fragments' in representation_ms_info:
2778 # NB: mpd_url may be empty when MPD manifest is parsed from a string
2779 'url': mpd_url or base_url,
2780 'fragment_base_url': base_url,
2782 'protocol': 'http_dash_segments' if mime_type != 'image/jpeg' else 'mhtml',
2784 if 'initialization_url' in representation_ms_info:
2785 initialization_url = representation_ms_info['initialization_url']
2786 if not f.get('url'):
2787 f['url'] = initialization_url
2788 f['fragments'].append({location_key(initialization_url): initialization_url})
2789 f['fragments'].extend(representation_ms_info['fragments'])
2791 # Assuming direct URL to unfragmented media.
2793 if content_type in ('video', 'audio') or mime_type == 'image/jpeg':
2795 elif content_type == 'text':
2796 subtitles.setdefault(lang or 'und', []).append(f)
2798 self.report_warning('Unknown MIME type %s in DASH manifest' % mime_type)
2799 return formats, subtitles
2801 def _extract_ism_formats(self, *args, **kwargs):
2802 fmts, subs = self._extract_ism_formats_and_subtitles(*args, **kwargs)
2804 self.report_warning(bug_reports_message(
2805 "Ignoring subtitle tracks found in the ISM manifest; "
2806 "if any subtitle tracks are missing,"
2810 def _extract_ism_formats_and_subtitles(self, ism_url, video_id, ism_id=None, note=None, errnote=None, fatal=True, data=None, headers={}, query={}):
2811 res = self._download_xml_handle(
2813 note='Downloading ISM manifest' if note is None else note,
2814 errnote='Failed to download ISM manifest' if errnote is None else errnote,
2815 fatal=fatal, data=data, headers=headers, query=query)
2822 return self._parse_ism_formats_and_subtitles(ism_doc, urlh.geturl(), ism_id)
2824 def _parse_ism_formats_and_subtitles(self, ism_doc, ism_url, ism_id=None):
2826 Parse formats
from ISM manifest
.
2828 1. [MS
-SSTR
]: Smooth Streaming Protocol
,
2829 https
://msdn
.microsoft
.com
/en
-us
/library
/ff469518
.aspx
2831 if ism_doc.get('IsLive') == 'TRUE':
2833 if (not self.get_param('allow_unplayable_formats')
2834 and ism_doc.find('Protection') is not None):
2837 duration = int(ism_doc.attrib['Duration'])
2838 timescale = int_or_none(ism_doc.get('TimeScale')) or 10000000
2842 for stream in ism_doc.findall('StreamIndex'):
2843 stream_type = stream.get('Type')
2844 if stream_type not in ('video', 'audio', 'text'):
2846 url_pattern = stream.attrib['Url']
2847 stream_timescale = int_or_none(stream.get('TimeScale')) or timescale
2848 stream_name = stream.get('Name')
2849 stream_language = stream.get('Language', 'und')
2850 for track in stream.findall('QualityLevel'):
2851 fourcc = track.get('FourCC') or ('AACL' if track.get('AudioTag') == '255' else None)
2852 # TODO: add support for WVC1 and WMAP
2853 if fourcc not in ('H264', 'AVC1', 'AACL', 'TTML'):
2854 self.report_warning('%s is not a supported codec' % fourcc)
2856 tbr = int(track.attrib['Bitrate']) // 1000
2857 # [1] does not mention Width and Height attributes. However,
2858 # they're often present while MaxWidth and MaxHeight are
2859 # missing, so should be used as fallbacks
2860 width = int_or_none(track.get('MaxWidth') or track.get('Width'))
2861 height = int_or_none(track.get('MaxHeight') or track.get('Height'))
2862 sampling_rate = int_or_none(track.get('SamplingRate'))
2864 track_url_pattern = re.sub(r'{[Bb]itrate}', track.attrib['Bitrate'], url_pattern)
2865 track_url_pattern = compat_urlparse.urljoin(ism_url, track_url_pattern)
2871 stream_fragments = stream.findall('c')
2872 for stream_fragment_index, stream_fragment in enumerate(stream_fragments):
2873 fragment_ctx['time'] = int_or_none(stream_fragment.get('t')) or fragment_ctx['time']
2874 fragment_repeat = int_or_none(stream_fragment.get('r')) or 1
2875 fragment_ctx['duration'] = int_or_none(stream_fragment.get('d'))
2876 if not fragment_ctx['duration']:
2878 next_fragment_time = int(stream_fragment[stream_fragment_index + 1].attrib['t'])
2880 next_fragment_time = duration
2881 fragment_ctx['duration'] = (next_fragment_time - fragment_ctx['time']) / fragment_repeat
2882 for _ in range(fragment_repeat):
2884 'url': re.sub(r'{start[ _]time}', compat_str(fragment_ctx['time']), track_url_pattern),
2885 'duration': fragment_ctx['duration'] / stream_timescale,
2887 fragment_ctx['time'] += fragment_ctx['duration']
2891 format_id.append(ism_id)
2893 format_id.append(stream_name)
2894 format_id.append(compat_str(tbr))
2896 if stream_type == 'text':
2897 subtitles.setdefault(stream_language, []).append({
2901 'manifest_url': ism_url,
2902 'fragments': fragments,
2903 '_download_params': {
2904 'stream_type': stream_type,
2905 'duration': duration,
2906 'timescale': stream_timescale,
2908 'language': stream_language,
2909 'codec_private_data': track.get('CodecPrivateData'),
2912 elif stream_type in ('video', 'audio'):
2914 'format_id': '-'.join(format_id),
2916 'manifest_url': ism_url,
2917 'ext': 'ismv' if stream_type == 'video' else 'isma',
2921 'asr': sampling_rate,
2922 'vcodec': 'none' if stream_type == 'audio' else fourcc,
2923 'acodec': 'none' if stream_type == 'video' else fourcc,
2925 'fragments': fragments,
2926 '_download_params': {
2927 'stream_type': stream_type,
2928 'duration': duration,
2929 'timescale': stream_timescale,
2930 'width': width or 0,
2931 'height': height or 0,
2933 'language': stream_language,
2934 'codec_private_data': track.get('CodecPrivateData'),
2935 'sampling_rate': sampling_rate,
2936 'channels': int_or_none(track.get('Channels', 2)),
2937 'bits_per_sample': int_or_none(track.get('BitsPerSample', 16)),
2938 'nal_unit_length_field': int_or_none(track.get('NALUnitLengthField', 4)),
2941 return formats, subtitles
2943 def _parse_html5_media_entries(self, base_url, webpage, video_id, m3u8_id=None, m3u8_entry_protocol='m3u8', mpd_id=None, preference=None, quality=None):
2944 def absolute_url(item_url):
2945 return urljoin(base_url, item_url)
2947 def parse_content_type(content_type):
2948 if not content_type:
2950 ctr = re.search(r'(?P<mimetype>[^/]+/[^;]+)(?:;\s*codecs="?(?P<codecs>[^"]+))?', content_type)
2952 mimetype, codecs = ctr.groups()
2953 f = parse_codecs(codecs)
2954 f['ext'] = mimetype2ext(mimetype)
2958 def _media_formats(src, cur_media_type, type_info={}):
2959 full_url = absolute_url(src)
2960 ext = type_info.get('ext') or determine_ext(full_url)
2962 is_plain_url = False
2963 formats = self._extract_m3u8_formats(
2964 full_url, video_id, ext='mp4',
2965 entry_protocol=m3u8_entry_protocol, m3u8_id=m3u8_id,
2966 preference=preference, quality=quality, fatal=False)
2968 is_plain_url = False
2969 formats = self._extract_mpd_formats(
2970 full_url, video_id, mpd_id=mpd_id, fatal=False)
2975 'vcodec': 'none' if cur_media_type == 'audio' else None,
2977 return is_plain_url, formats
2980 # amp-video and amp-audio are very similar to their HTML5 counterparts
2981 # so we wll include them right here (see
2982 # https://www.ampproject.org/docs/reference/components/amp-video)
2983 # For dl8-* tags see https://delight-vr.com/documentation/dl8-video/
2984 _MEDIA_TAG_NAME_RE = r'(?:(?:amp|dl8(?:-live)?)-)?(video|audio)'
2985 media_tags = [(media_tag, media_tag_name, media_type, '')
2986 for media_tag, media_tag_name, media_type
2987 in re.findall(r'(?s)(<(%s)[^>]*/>)' % _MEDIA_TAG_NAME_RE, webpage)]
2988 media_tags.extend(re.findall(
2989 # We only allow video|audio followed by a whitespace or '>'.
2990 # Allowing more characters may end up in significant slow down (see
2991 # https://github.com/ytdl-org/youtube-dl/issues/11979, example URL:
2992 # http://www.porntrex.com/maps/videositemap.xml).
2993 r'(?s)(<(?P<tag>%s)(?:\s+[^>]*)?>)(.*?)</(?P=tag)>' % _MEDIA_TAG_NAME_RE, webpage))
2994 for media_tag, _, media_type, media_content in media_tags:
2999 media_attributes = extract_attributes(media_tag)
3000 src = strip_or_none(media_attributes.get('src'))
3002 _, formats = _media_formats(src, media_type)
3003 media_info['formats'].extend(formats)
3004 media_info['thumbnail'] = absolute_url(media_attributes.get('poster'))
3006 for source_tag in re.findall(r'<source[^>]+>', media_content):
3007 s_attr = extract_attributes(source_tag)
3008 # data-video-src and data-src are non standard but seen
3009 # several times in the wild
3010 src = strip_or_none(dict_get(s_attr, ('src', 'data-video-src', 'data-src')))
3013 f = parse_content_type(s_attr.get('type'))
3014 is_plain_url, formats = _media_formats(src, media_type, f)
3016 # width, height, res, label and title attributes are
3017 # all not standard but seen several times in the wild
3020 for lbl in ('label', 'title')
3021 if str_or_none(s_attr.get(lbl))
3023 width = int_or_none(s_attr.get('width'))
3024 height = (int_or_none(s_attr.get('height'))
3025 or int_or_none(s_attr.get('res')))
3026 if not width or not height:
3028 resolution = parse_resolution(lbl)
3031 width = width or resolution.get('width')
3032 height = height or resolution.get('height')
3034 tbr = parse_bitrate(lbl)
3043 'format_id': s_attr.get('label') or s_attr.get('title'),
3045 f.update(formats[0])
3046 media_info['formats'].append(f)
3048 media_info['formats'].extend(formats)
3049 for track_tag in re.findall(r'<track[^>]+>', media_content):
3050 track_attributes = extract_attributes(track_tag)
3051 kind = track_attributes.get('kind')
3052 if not kind or kind in ('subtitles', 'captions'):
3053 src = strip_or_none(track_attributes.get('src'))
3056 lang = track_attributes.get('srclang') or track_attributes.get('lang') or track_attributes.get('label')
3057 media_info['subtitles'].setdefault(lang, []).append({
3058 'url': absolute_url(src),
3060 for f in media_info['formats']:
3061 f.setdefault('http_headers', {})['Referer'] = base_url
3062 if media_info['formats'] or media_info['subtitles']:
3063 entries.append(media_info)
3066 def _extract_akamai_formats(self, *args, **kwargs):
3067 fmts, subs = self._extract_akamai_formats_and_subtitles(*args, **kwargs)
3069 self.report_warning(bug_reports_message(
3070 "Ignoring subtitle tracks found in the manifests; "
3071 "if any subtitle tracks are missing,"
3075 def _extract_akamai_formats_and_subtitles(self, manifest_url, video_id, hosts={}):
3076 signed = 'hdnea=' in manifest_url
3078 # https://learn.akamai.com/en-us/webhelp/media-services-on-demand/stream-packaging-user-guide/GUID-BE6C0F73-1E06-483B-B0EA-57984B91B7F9.html
3079 manifest_url = re.sub(
3080 r'(?:b=[\d,-]+|(?:__a__|attributes)=off|__b__=\d+)&?',
3081 '', manifest_url).strip('?')
3086 hdcore_sign = 'hdcore=3.7.0'
3087 f4m_url = re.sub(r'(https?://[^/]+)/i/', r'\1/z/', manifest_url).replace('/master.m3u8', '/manifest.f4m')
3088 hds_host = hosts.get('hds')
3090 f4m_url = re.sub(r'(https?://)[^/]+', r'\1' + hds_host, f4m_url)
3091 if 'hdcore=' not in f4m_url:
3092 f4m_url += ('&' if '?' in f4m_url else '?') + hdcore_sign
3093 f4m_formats = self._extract_f4m_formats(
3094 f4m_url, video_id, f4m_id='hds', fatal=False)
3095 for entry in f4m_formats:
3096 entry.update({'extra_param_to_segment_url': hdcore_sign})
3097 formats.extend(f4m_formats)
3099 m3u8_url = re.sub(r'(https?://[^/]+)/z/', r'\1/i/', manifest_url).replace('/manifest.f4m', '/master.m3u8')
3100 hls_host = hosts.get('hls')
3102 m3u8_url = re.sub(r'(https?://)[^/]+', r'\1' + hls_host, m3u8_url)
3103 m3u8_formats, m3u8_subtitles = self._extract_m3u8_formats_and_subtitles(
3104 m3u8_url, video_id, 'mp4', 'm3u8_native',
3105 m3u8_id='hls', fatal=False)
3106 formats.extend(m3u8_formats)
3107 subtitles = self._merge_subtitles(subtitles, m3u8_subtitles)
3109 http_host = hosts.get('http')
3110 if http_host and m3u8_formats and not signed:
3111 REPL_REGEX = r'https?://[^/]+/i/([^,]+),([^/]+),([^/]+)\.csmil/.+'
3112 qualities = re.match(REPL_REGEX, m3u8_url).group(2).split(',')
3113 qualities_length = len(qualities)
3114 if len(m3u8_formats) in (qualities_length, qualities_length + 1):
3116 for f in m3u8_formats:
3117 if f['vcodec'] != 'none':
3118 for protocol in ('http', 'https'):
3120 del http_f['manifest_url']
3122 REPL_REGEX, protocol + r'://%s/\g<1>%s\3' % (http_host, qualities[i]), f['url'])
3124 'format_id': http_f['format_id'].replace('hls-', protocol + '-'),
3126 'protocol': protocol,
3128 formats.append(http_f)
3131 return formats, subtitles
3133 def _extract_wowza_formats(self, url, video_id, m3u8_entry_protocol='m3u8_native', skip_protocols=[]):
3134 query = compat_urlparse.urlparse(url).query
3135 url = re.sub(r'/(?:manifest|playlist|jwplayer)\.(?:m3u8|f4m|mpd|smil)', '', url)
3137 r'(?:(?:http|rtmp|rtsp)(?P<s>s)?:)?(?P<url>//[^?]+)', url)
3138 url_base = mobj.group('url')
3139 http_base_url = '%s%s:%s' % ('http', mobj.group('s') or '', url_base)
3142 def manifest_url(manifest):
3143 m_url = '%s/%s' % (http_base_url, manifest)
3145 m_url += '?%s' % query
3148 if 'm3u8' not in skip_protocols:
3149 formats.extend(self._extract_m3u8_formats(
3150 manifest_url('playlist.m3u8'), video_id, 'mp4',
3151 m3u8_entry_protocol, m3u8_id='hls', fatal=False))
3152 if 'f4m' not in skip_protocols:
3153 formats.extend(self._extract_f4m_formats(
3154 manifest_url('manifest.f4m'),
3155 video_id, f4m_id='hds', fatal=False))
3156 if 'dash' not in skip_protocols:
3157 formats.extend(self._extract_mpd_formats(
3158 manifest_url('manifest.mpd'),
3159 video_id, mpd_id='dash', fatal=False))
3160 if re.search(r'(?:/smil:|\.smil)', url_base):
3161 if 'smil' not in skip_protocols:
3162 rtmp_formats = self._extract_smil_formats(
3163 manifest_url('jwplayer.smil'),
3164 video_id, fatal=False)
3165 for rtmp_format in rtmp_formats:
3166 rtsp_format = rtmp_format.copy()
3167 rtsp_format['url'] = '%s/%s' % (rtmp_format['url'], rtmp_format['play_path'])
3168 del rtsp_format['play_path']
3169 del rtsp_format['ext']
3170 rtsp_format.update({
3171 'url': rtsp_format['url'].replace('rtmp://', 'rtsp://'),
3172 'format_id': rtmp_format['format_id'].replace('rtmp', 'rtsp'),
3175 formats.extend([rtmp_format, rtsp_format])
3177 for protocol in ('rtmp', 'rtsp'):
3178 if protocol not in skip_protocols:
3180 'url': '%s:%s' % (protocol, url_base),
3181 'format_id': protocol,
3182 'protocol': protocol,
3186 def _find_jwplayer_data(self, webpage, video_id=None, transform_source=js_to_json):
3188 r'(?s)jwplayer\((?P<quote>[\'"])[^\'" ]+(?P=quote)\)(?!</script>).*?\.setup\s*\((?P<options>[^)]+)\)',
3192 jwplayer_data = self._parse_json(mobj.group('options'),
3194 transform_source=transform_source)
3195 except ExtractorError:
3198 if isinstance(jwplayer_data, dict):
3199 return jwplayer_data
3201 def _extract_jwplayer_data(self, webpage, video_id, *args, **kwargs):
3202 jwplayer_data = self._find_jwplayer_data(
3203 webpage, video_id, transform_source=js_to_json)
3204 return self._parse_jwplayer_data(
3205 jwplayer_data, video_id, *args, **kwargs)
3207 def _parse_jwplayer_data(self, jwplayer_data, video_id=None, require_title=True,
3208 m3u8_id=None, mpd_id=None, rtmp_params=None, base_url=None):
3209 # JWPlayer backward compatibility: flattened playlists
3210 # https://github.com/jwplayer/jwplayer/blob/v7.4.3/src/js/api/config.js#L81-L96
3211 if 'playlist' not in jwplayer_data:
3212 jwplayer_data = {'playlist': [jwplayer_data]}
3216 # JWPlayer backward compatibility: single playlist item
3217 # https://github.com/jwplayer/jwplayer/blob/v7.7.0/src/js/playlist/playlist.js#L10
3218 if not isinstance(jwplayer_data['playlist'], list):
3219 jwplayer_data['playlist'] = [jwplayer_data['playlist']]
3221 for video_data in jwplayer_data['playlist']:
3222 # JWPlayer backward compatibility: flattened sources
3223 # https://github.com/jwplayer/jwplayer/blob/v7.4.3/src/js/playlist/item.js#L29-L35
3224 if 'sources' not in video_data:
3225 video_data['sources'] = [video_data]
3227 this_video_id = video_id or video_data['mediaid']
3229 formats = self._parse_jwplayer_formats(
3230 video_data['sources'], video_id=this_video_id, m3u8_id=m3u8_id,
3231 mpd_id=mpd_id, rtmp_params=rtmp_params, base_url=base_url)
3234 tracks = video_data.get('tracks')
3235 if tracks and isinstance(tracks, list):
3236 for track in tracks:
3237 if not isinstance(track, dict):
3239 track_kind = track.get('kind')
3240 if not track_kind or not isinstance(track_kind, compat_str):
3242 if track_kind.lower() not in ('captions', 'subtitles'):
3244 track_url = urljoin(base_url, track.get('file'))
3247 subtitles.setdefault(track.get('label') or 'en', []).append({
3248 'url': self._proto_relative_url(track_url)
3252 'id': this_video_id,
3253 'title': unescapeHTML(video_data['title'] if require_title else video_data.get('title')),
3254 'description': clean_html(video_data.get('description')),
3255 'thumbnail': urljoin(base_url, self._proto_relative_url(video_data.get('image'))),
3256 'timestamp': int_or_none(video_data.get('pubdate')),
3257 'duration': float_or_none(jwplayer_data.get('duration') or video_data.get('duration')),
3258 'subtitles': subtitles,
3260 # https://github.com/jwplayer/jwplayer/blob/master/src/js/utils/validator.js#L32
3261 if len(formats) == 1 and re.search(r'^(?:http|//).*(?:youtube\.com|youtu\.be)/.+', formats[0]['url']):
3263 '_type': 'url_transparent',
3264 'url': formats[0]['url'],
3267 self._sort_formats(formats)
3268 entry['formats'] = formats
3269 entries.append(entry)
3270 if len(entries) == 1:
3273 return self.playlist_result(entries)
3275 def _parse_jwplayer_formats(self, jwplayer_sources_data, video_id=None,
3276 m3u8_id=None, mpd_id=None, rtmp_params=None, base_url=None):
3279 for source in jwplayer_sources_data:
3280 if not isinstance(source, dict):
3282 source_url = urljoin(
3283 base_url, self._proto_relative_url(source.get('file')))
3284 if not source_url or source_url in urls:
3286 urls.append(source_url)
3287 source_type = source.get('type') or ''
3288 ext = mimetype2ext(source_type) or determine_ext(source_url)
3289 if source_type == 'hls' or ext == 'm3u8':
3290 formats.extend(self._extract_m3u8_formats(
3291 source_url, video_id, 'mp4', entry_protocol='m3u8_native',
3292 m3u8_id=m3u8_id, fatal=False))
3293 elif source_type == 'dash' or ext == 'mpd':
3294 formats.extend(self._extract_mpd_formats(
3295 source_url, video_id, mpd_id=mpd_id, fatal=False))
3297 formats.extend(self._extract_smil_formats(
3298 source_url, video_id, fatal=False))
3299 # https://github.com/jwplayer/jwplayer/blob/master/src/js/providers/default.js#L67
3300 elif source_type.startswith('audio') or ext in (
3301 'oga', 'aac', 'mp3', 'mpeg', 'vorbis'):
3308 height = int_or_none(source.get('height'))
3310 # Often no height is provided but there is a label in
3311 # format like "1080p", "720p SD", or 1080.
3312 height = int_or_none(self._search_regex(
3313 r'^(\d{3,4})[pP]?(?:\b|$)', compat_str(source.get('label') or ''),
3314 'height', default=None))
3317 'width': int_or_none(source.get('width')),
3319 'tbr': int_or_none(source.get('bitrate')),
3322 if source_url.startswith('rtmp'):
3323 a_format['ext'] = 'flv'
3324 # See com/longtailvideo/jwplayer/media/RTMPMediaProvider.as
3325 # of jwplayer.flash.swf
3326 rtmp_url_parts = re.split(
3327 r'((?:mp4|mp3|flv):)', source_url, 1)
3328 if len(rtmp_url_parts) == 3:
3329 rtmp_url, prefix, play_path = rtmp_url_parts
3332 'play_path': prefix + play_path,
3335 a_format.update(rtmp_params)
3336 formats.append(a_format)
3339 def _live_title(self, name):
3340 """ Generate the title
for a live video
"""
3341 now = datetime.datetime.now()
3342 now_str = now.strftime('%Y-%m-%d %H:%M')
3343 return name + ' ' + now_str
3345 def _int(self, v, name, fatal=False, **kwargs):
3346 res = int_or_none(v, **kwargs)
3347 if 'get_attr' in kwargs:
3348 print(getattr(v, kwargs['get_attr']))
3350 msg = 'Failed to extract %s: Could not parse value %r' % (name, v)
3352 raise ExtractorError(msg)
3354 self.report_warning(msg)
3357 def _float(self, v, name, fatal=False, **kwargs):
3358 res = float_or_none(v, **kwargs)
3360 msg = 'Failed to extract %s: Could not parse value %r' % (name, v)
3362 raise ExtractorError(msg)
3364 self.report_warning(msg)
3367 def _set_cookie(self, domain, name, value, expire_time=None, port=None,
3368 path='/', secure=False, discard=False, rest={}, **kwargs):
3369 cookie = compat_cookiejar_Cookie(
3370 0, name, value, port, port is not None, domain, True,
3371 domain.startswith('.'), path, True, secure, expire_time,
3372 discard, None, None, rest)
3373 self._downloader.cookiejar.set_cookie(cookie)
3375 def _get_cookies(self, url):
3376 """ Return a compat_cookies_SimpleCookie
with the cookies
for the url
"""
3377 req = sanitized_Request(url)
3378 self._downloader.cookiejar.add_cookie_header(req)
3379 return compat_cookies_SimpleCookie(req.get_header('Cookie'))
3381 def _apply_first_set_cookie_header(self, url_handle, cookie):
3383 Apply first Set
-Cookie header instead of the last
. Experimental
.
3385 Some
sites (e
.g
. [1-3]) may serve two cookies under the same name
3386 in Set
-Cookie header
and expect the
first (old
) one to be
set rather
3387 than
second (new
). However
, as of RFC6265 the newer one cookie
3388 should be
set into cookie store what actually happens
.
3389 We will workaround this issue by resetting the cookie to
3390 the first one manually
.
3391 1. https
://new
.vk
.com
/
3392 2. https
://github
.com
/ytdl
-org
/youtube
-dl
/issues
/9841#issuecomment-227871201
3393 3. https
://learning
.oreilly
.com
/
3395 for header, cookies in url_handle.headers.items():
3396 if header.lower() != 'set-cookie':
3398 if sys.version_info[0] >= 3:
3399 cookies = cookies.encode('iso-8859-1')
3400 cookies = cookies.decode('utf-8')
3401 cookie_value = re.search(
3402 r'%s=(.+?);.*?\b[Dd]omain=(.+?)(?:[,;]|$)' % cookie, cookies)
3404 value, domain = cookie_value.groups()
3405 self._set_cookie(domain, cookie, value)
3408 def get_testcases(self, include_onlymatching=False):
3409 t = getattr(self, '_TEST', None)
3411 assert not hasattr(self, '_TESTS'), \
3412 '%s has _TEST and _TESTS' % type(self).__name__
3415 tests = getattr(self, '_TESTS', [])
3417 if not include_onlymatching and t.get('only_matching', False):
3419 t['name'] = type(self).__name__[:-len('IE')]
3422 def is_suitable(self, age_limit):
3423 """ Test whether the extractor
is generally suitable
for the given
3424 age
limit (i
.e
. pornographic sites are
not, all others usually are
) """
3426 any_restricted = False
3427 for tc in self.get_testcases(include_onlymatching=False):
3428 if tc.get('playlist', []):
3429 tc = tc['playlist'][0]
3430 is_restricted = age_restricted(
3431 tc.get('info_dict', {}).get('age_limit'), age_limit)
3432 if not is_restricted:
3434 any_restricted = any_restricted or is_restricted
3435 return not any_restricted
3437 def extract_subtitles(self, *args, **kwargs):
3438 if (self.get_param('writesubtitles', False)
3439 or self.get_param('listsubtitles')):
3440 return self._get_subtitles(*args, **kwargs)
3443 def _get_subtitles(self, *args, **kwargs):
3444 raise NotImplementedError('This method must be implemented by subclasses')
3447 def _merge_subtitle_items(subtitle_list1, subtitle_list2):
3448 """ Merge subtitle items
for one language
. Items
with duplicated URLs
3449 will be dropped
. """
3450 list1_urls = set([item['url'] for item in subtitle_list1])
3451 ret = list(subtitle_list1)
3452 ret.extend([item for item in subtitle_list2 if item['url'] not in list1_urls])
3456 def _merge_subtitles(cls, *dicts, target=None):
3457 """ Merge subtitle dictionaries
, language by language
. """
3461 for lang, subs in d.items():
3462 target[lang] = cls._merge_subtitle_items(target.get(lang, []), subs)
3465 def extract_automatic_captions(self, *args, **kwargs):
3466 if (self.get_param('writeautomaticsub', False)
3467 or self.get_param('listsubtitles')):
3468 return self._get_automatic_captions(*args, **kwargs)
3471 def _get_automatic_captions(self, *args, **kwargs):
3472 raise NotImplementedError('This method must be implemented by subclasses')
3474 def mark_watched(self, *args, **kwargs):
3475 if (self.get_param('mark_watched', False)
3476 and (self._get_login_info()[0] is not None
3477 or self.get_param('cookiefile') is not None)):
3478 self._mark_watched(*args, **kwargs)
3480 def _mark_watched(self, *args, **kwargs):
3481 raise NotImplementedError('This method must be implemented by subclasses')
3483 def geo_verification_headers(self):
3485 geo_verification_proxy = self.get_param('geo_verification_proxy')
3486 if geo_verification_proxy:
3487 headers['Ytdl-request-proxy'] = geo_verification_proxy
3490 def _generic_id(self, url):
3491 return compat_urllib_parse_unquote(os.path.splitext(url.rstrip('/').split('/')[-1])[0])
3493 def _generic_title(self, url):
3494 return compat_urllib_parse_unquote(os.path.splitext(url_basename(url))[0])
3497 def _availability(is_private=None, needs_premium=None, needs_subscription=None, needs_auth=None, is_unlisted=None):
3498 all_known = all(map(
3499 lambda x: x is not None,
3500 (is_private, needs_premium, needs_subscription, needs_auth, is_unlisted)))
3502 'private' if is_private
3503 else 'premium_only' if needs_premium
3504 else 'subscriber_only' if needs_subscription
3505 else 'needs_auth' if needs_auth
3506 else 'unlisted' if is_unlisted
3507 else 'public' if all_known
3510 def _configuration_arg(self, key, default=NO_DEFAULT, casesense=False):
3512 @returns A list of values for the extractor argument given by "key"
3513 or "default" if no such key is present
3514 @param default The default value to return when the key is not present (default: [])
3515 @param casesense When false, the values are converted to lower case
3518 self._downloader.params, ('extractor_args', self.ie_key().lower(), key))
3520 return [] if default is NO_DEFAULT else default
3521 return list(val) if casesense else [x.lower() for x in val]
3524 class SearchInfoExtractor(InfoExtractor):
3526 Base
class for paged search queries extractors
.
3527 They accept URLs
in the format
_SEARCH_KEY(|all|
[0-9]):{query}
3528 Instances should define _SEARCH_KEY
and _MAX_RESULTS
.
3532 def _make_valid_url(cls):
3533 return r'%s(?P<prefix>|[1-9][0-9]*|all):(?P<query>[\s\S]+)' % cls._SEARCH_KEY
3536 def suitable(cls, url):
3537 return re.match(cls._make_valid_url(), url) is not None
3539 def _real_extract(self, query):
3540 mobj = re.match(self._make_valid_url(), query)
3542 raise ExtractorError('Invalid search query "%s"' % query)
3544 prefix = mobj.group('prefix')
3545 query = mobj.group('query')
3547 return self._get_n_results(query, 1)
3548 elif prefix == 'all':
3549 return self._get_n_results(query, self._MAX_RESULTS)
3553 raise ExtractorError('invalid download number %s for query "%s"' % (n, query))
3554 elif n > self._MAX_RESULTS:
3555 self.report_warning('%s returns max %i results (you requested %i)' % (self._SEARCH_KEY, self._MAX_RESULTS, n))
3556 n = self._MAX_RESULTS
3557 return self._get_n_results(query, n)
3559 def _get_n_results(self, query, n):
3560 """Get a specified number of results
for a query
"""
3561 raise NotImplementedError('This method must be implemented by subclasses')
3564 def SEARCH_KEY(self):
3565 return self._SEARCH_KEY