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
,
65 parse_m3u8_attributes
,
89 class InfoExtractor(object):
90 """Information Extractor class.
92 Information extractors are the classes that, given a URL, extract
93 information about the video (or videos) the URL refers to. This
94 information includes the real video URL, the video title, author and
95 others. The information is stored in a dictionary which is then
96 passed to the YoutubeDL. The YoutubeDL processes this
97 information possibly downloading the video to the file system, among
98 other possible outcomes.
100 The type field determines the type of the result.
101 By far the most common value (and the default if _type is missing) is
102 "video", which indicates a single video.
104 For a video, the dictionaries must include the following fields:
106 id: Video identifier.
107 title: Video title, unescaped.
109 Additionally, it must contain either a formats entry or a url one:
111 formats: A list of dictionaries for each format available, ordered
112 from worst to best quality.
115 * url The mandatory URL representing the media:
116 for plain file media - HTTP URL of this file,
118 for HLS - URL of the M3U8 media playlist,
119 for HDS - URL of the F4M manifest,
121 - HTTP URL to plain file media (in case of
123 - URL of the MPD manifest or base URL
124 representing the media if MPD manifest
125 is parsed from a string (in case of
127 for MSS - URL of the ISM manifest.
129 The URL of the manifest file in case of
131 for HLS - URL of the M3U8 master playlist,
132 for HDS - URL of the F4M manifest,
133 for DASH - URL of the MPD manifest,
134 for MSS - URL of the ISM manifest.
135 * ext Will be calculated from URL if missing
136 * format A human-readable description of the format
137 ("mp4 container with h264/opus").
138 Calculated from the format_id, width, height.
139 and format_note fields if missing.
140 * format_id A short description of the format
141 ("mp4_h264_opus" or "19").
142 Technically optional, but strongly recommended.
143 * format_note Additional info about the format
144 ("3D" or "DASH video")
145 * width Width of the video, if known
146 * height Height of the video, if known
147 * resolution Textual description of width and height
148 * tbr Average bitrate of audio and video in KBit/s
149 * abr Average audio bitrate in KBit/s
150 * acodec Name of the audio codec in use
151 * asr Audio sampling rate in Hertz
152 * vbr Average video bitrate in KBit/s
154 * vcodec Name of the video codec in use
155 * container Name of the container format
156 * filesize The number of bytes, if known in advance
157 * filesize_approx An estimate for the number of bytes
158 * player_url SWF Player URL (used for rtmpdump).
159 * protocol The protocol that will be used for the actual
160 download, lower-case.
161 "http", "https", "rtsp", "rtmp", "rtmp_ffmpeg", "rtmpe",
162 "m3u8", "m3u8_native" or "http_dash_segments".
164 Base URL for fragments. Each fragment's path
165 value (if present) will be relative to
167 * fragments A list of fragments of a fragmented media.
168 Each fragment entry must contain either an url
169 or a path. If an url is present it should be
170 considered by a client. Otherwise both path and
171 fragment_base_url must be present. Here is
172 the list of all potential fields:
173 * "url" - fragment's URL
174 * "path" - fragment's path relative to
176 * "duration" (optional, int or float)
177 * "filesize" (optional, int)
178 * preference Order number of this format. If this field is
179 present and not None, the formats get sorted
180 by this field, regardless of all other values.
181 -1 for default (order by other properties),
182 -2 or smaller for less than default.
183 < -1000 to hide the format (if there is
184 another one which is strictly better)
185 * language Language code, e.g. "de" or "en-US".
186 * language_preference Is this in the language mentioned in
188 10 if it's what the URL is about,
189 -1 for default (don't know),
190 -10 otherwise, other values reserved for now.
191 * quality Order number of the video quality of this
192 format, irrespective of the file format.
193 -1 for default (order by other properties),
194 -2 or smaller for less than default.
195 * source_preference Order number for this video source
196 (quality takes higher priority)
197 -1 for default (order by other properties),
198 -2 or smaller for less than default.
199 * http_headers A dictionary of additional HTTP headers
200 to add to the request.
201 * stretched_ratio If given and not 1, indicates that the
202 video's pixels are not square.
203 width : height ratio as float.
204 * no_resume The server does not support resuming the
205 (HTTP or RTMP) download. Boolean.
206 * downloader_options A dictionary of downloader options as
207 described in FileDownloader
208 RTMP formats can also have the additional fields: page_url,
209 app, play_path, tc_url, flash_version, rtmp_live, rtmp_conn,
210 rtmp_protocol, rtmp_real_time
212 url: Final video URL.
213 ext: Video filename extension.
214 format: The video format, defaults to ext (used for --get-format)
215 player_url: SWF Player URL (used for rtmpdump).
217 The following fields are optional:
219 alt_title: A secondary title of the video.
220 display_id An alternative identifier for the video, not necessarily
221 unique, but available before title. Typically, id is
222 something like "4234987", title "Dancing naked mole rats",
223 and display_id "dancing-naked-mole-rats"
224 thumbnails: A list of dictionaries, with the following entries:
225 * "id" (optional, string) - Thumbnail format ID
227 * "preference" (optional, int) - quality of the image
228 * "width" (optional, int)
229 * "height" (optional, int)
230 * "resolution" (optional, string "{width}x{height}",
232 * "filesize" (optional, int)
233 * "_test_url" (optional, bool) - If true, test the URL
234 thumbnail: Full URL to a video thumbnail image.
235 description: Full video description.
236 uploader: Full name of the video uploader.
237 license: License name the video is licensed under.
238 creator: The creator of the video.
239 release_timestamp: UNIX timestamp of the moment the video was released.
240 release_date: The date (YYYYMMDD) when the video was released.
241 timestamp: UNIX timestamp of the moment the video was uploaded
242 upload_date: Video upload date (YYYYMMDD).
243 If not explicitly set, calculated from timestamp.
244 uploader_id: Nickname or id of the video uploader.
245 uploader_url: Full URL to a personal webpage of the video uploader.
246 channel: Full name of the channel the video is uploaded on.
247 Note that channel fields may or may not repeat uploader
248 fields. This depends on a particular extractor.
249 channel_id: Id of the channel.
250 channel_url: Full URL to a channel webpage.
251 location: Physical location where the video was filmed.
252 subtitles: The available subtitles as a dictionary in the format
253 {tag: subformats}. "tag" is usually a language code, and
254 "subformats" is a list sorted from lower to higher
255 preference, each element is a dictionary with the "ext"
257 * "data": The subtitles file contents
258 * "url": A URL pointing to the subtitles file
259 It can optionally also have:
260 * "name": Name or description of the subtitles
261 "ext" will be calculated from URL if missing
262 automatic_captions: Like 'subtitles'; contains automatically generated
263 captions instead of normal subtitles
264 duration: Length of the video in seconds, as an integer or float.
265 view_count: How many users have watched the video on the platform.
266 like_count: Number of positive ratings of the video
267 dislike_count: Number of negative ratings of the video
268 repost_count: Number of reposts of the video
269 average_rating: Average rating give by users, the scale used depends on the webpage
270 comment_count: Number of comments on the video
271 comments: A list of comments, each with one or more of the following
272 properties (all but one of text or html optional):
273 * "author" - human-readable name of the comment author
274 * "author_id" - user ID of the comment author
275 * "author_thumbnail" - The thumbnail of the comment author
277 * "html" - Comment as HTML
278 * "text" - Plain text of the comment
279 * "timestamp" - UNIX timestamp of comment
280 * "parent" - ID of the comment this one is replying to.
281 Set to "root" to indicate that this is a
282 comment to the original video.
283 * "like_count" - Number of positive ratings of the comment
284 * "dislike_count" - Number of negative ratings of the comment
285 * "is_favorited" - Whether the comment is marked as
286 favorite by the video uploader
287 * "author_is_uploader" - Whether the comment is made by
289 age_limit: Age restriction for the video, as an integer (years)
290 webpage_url: The URL to the video webpage, if given to yt-dlp it
291 should allow to get the same result again. (It will be set
292 by YoutubeDL if it's missing)
293 categories: A list of categories that the video falls in, for example
295 tags: A list of tags assigned to the video, e.g. ["sweden", "pop music"]
296 cast: A list of the video cast
297 is_live: True, False, or None (=unknown). Whether this video is a
298 live stream that goes on instead of a fixed-length video.
299 was_live: True, False, or None (=unknown). Whether this video was
300 originally a live stream.
301 live_status: 'is_live', 'is_upcoming', 'was_live', 'not_live' or None (=unknown)
302 If absent, automatically set from is_live, was_live
303 start_time: Time in seconds where the reproduction should start, as
304 specified in the URL.
305 end_time: Time in seconds where the reproduction should end, as
306 specified in the URL.
307 chapters: A list of dictionaries, with the following entries:
308 * "start_time" - The start time of the chapter in seconds
309 * "end_time" - The end time of the chapter in seconds
310 * "title" (optional, string)
311 playable_in_embed: Whether this video is allowed to play in embedded
312 players on other sites. Can be True (=always allowed),
313 False (=never allowed), None (=unknown), or a string
314 specifying the criteria for embedability (Eg: 'whitelist')
315 availability: Under what condition the video is available. One of
316 'private', 'premium_only', 'subscriber_only', 'needs_auth',
317 'unlisted' or 'public'. Use 'InfoExtractor._availability'
319 __post_extractor: A function to be called just before the metadata is
320 written to either disk, logger or console. The function
321 must return a dict which will be added to the info_dict.
322 This is usefull for additional information that is
323 time-consuming to extract. Note that the fields thus
324 extracted will not be available to output template and
325 match_filter. So, only "comments" and "comment_count" are
326 currently allowed to be extracted via this method.
328 The following fields should only be used when the video belongs to some logical
331 chapter: Name or title of the chapter the video belongs to.
332 chapter_number: Number of the chapter the video belongs to, as an integer.
333 chapter_id: Id of the chapter the video belongs to, as a unicode string.
335 The following fields should only be used when the video is an episode of some
336 series, programme or podcast:
338 series: Title of the series or programme the video episode belongs to.
339 season: Title of the season the video episode belongs to.
340 season_number: Number of the season the video episode belongs to, as an integer.
341 season_id: Id of the season the video episode belongs to, as a unicode string.
342 episode: Title of the video episode. Unlike mandatory video title field,
343 this field should denote the exact title of the video episode
344 without any kind of decoration.
345 episode_number: Number of the video episode within a season, as an integer.
346 episode_id: Id of the video episode, as a unicode string.
348 The following fields should only be used when the media is a track or a part of
351 track: Title of the track.
352 track_number: Number of the track within an album or a disc, as an integer.
353 track_id: Id of the track (useful in case of custom indexing, e.g. 6.iii),
355 artist: Artist(s) of the track.
356 genre: Genre(s) of the track.
357 album: Title of the album the track belongs to.
358 album_type: Type of the album (e.g. "Demo", "Full-length", "Split", "Compilation", etc).
359 album_artist: List of all artists appeared on the album (e.g.
360 "Ash Borer / Fell Voices" or "Various Artists", useful for splits
362 disc_number: Number of the disc or other physical medium the track belongs to,
364 release_year: Year (YYYY) when the album was released.
366 Unless mentioned otherwise, the fields should be Unicode strings.
368 Unless mentioned otherwise, None is equivalent to absence of information.
371 _type "playlist" indicates multiple videos.
372 There must be a key "entries", which is a list, an iterable, or a PagedList
373 object, each element of which is a valid dictionary by this specification.
375 Additionally, playlists can have "id", "title", and any other relevent
376 attributes with the same semantics as videos (see above).
379 _type "multi_video" indicates that there are multiple videos that
380 form a single show, for examples multiple acts of an opera or TV episode.
381 It must have an entries key like a playlist and contain all the keys
382 required for a video at the same time.
385 _type "url" indicates that the video must be extracted from another
386 location, possibly by a different extractor. Its only required key is:
387 "url" - the next URL to extract.
388 The key "ie_key" can be set to the class name (minus the trailing "IE",
389 e.g. "Youtube") if the extractor class is known in advance.
390 Additionally, the dictionary may have any properties of the resolved entity
391 known in advance, for example "title" if the title of the referred video is
395 _type "url_transparent" entities have the same specification as "url", but
396 indicate that the given additional information is more precise than the one
397 associated with the resolved URL.
398 This is useful when a site employs a video service that hosts the video and
399 its technical metadata, but that video service does not embed a useful
400 title, description etc.
403 Subclasses of this one should re-define the _real_initialize() and
404 _real_extract() methods and define a _VALID_URL regexp.
405 Probably, they should also be added to the list of extractors.
407 _GEO_BYPASS attribute may be set to False in order to disable
408 geo restriction bypass mechanisms for a particular extractor.
409 Though it won't disable explicit geo restriction bypass based on
410 country code provided with geo_bypass_country.
412 _GEO_COUNTRIES attribute may contain a list of presumably geo unrestricted
413 countries for this extractor. One of these countries will be used by
414 geo restriction bypass mechanism right away in order to bypass
415 geo restriction, of course, if the mechanism is not disabled.
417 _GEO_IP_BLOCKS attribute may contain a list of presumably geo unrestricted
418 IP blocks in CIDR notation for this extractor. One of these IP blocks
419 will be used by geo restriction bypass mechanism similarly
422 Finally, the _WORKING attribute should be set to False for broken IEs
423 in order to warn the users and skip the tests.
428 _x_forwarded_for_ip
= None
430 _GEO_COUNTRIES
= None
431 _GEO_IP_BLOCKS
= None
435 'any': 'Use --cookies, --username and --password or --netrc to provide account credentials',
437 'Use --cookies for the authentication. '
438 'See https://github.com/ytdl-org/youtube-dl#how-do-i-pass-cookies-to-youtube-dl for how to pass cookies'),
439 'password': 'Use --username and --password or --netrc to provide account credentials',
442 def __init__(self
, downloader
=None):
443 """Constructor. Receives an optional downloader."""
445 self
._x
_forwarded
_for
_ip
= None
446 self
._printed
_messages
= set()
447 self
.set_downloader(downloader
)
450 def _match_valid_url(cls
, url
):
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
)
459 def suitable(cls
, url
):
460 """Receives a URL and returns True if suitable for this IE."""
461 # This function must import everything it needs (except other extractors),
462 # so that lazy_extractors works correctly
463 return cls
._match
_valid
_url
(url
) is not None
466 def _match_id(cls
, url
):
467 return cls
._match
_valid
_url
(url
).group('id')
471 """Getter method for _WORKING."""
474 def initialize(self
):
475 """Initializes an instance (authentication, etc)."""
476 self
._printed
_messages
= set()
477 self
._initialize
_geo
_bypass
({
478 'countries': self
._GEO
_COUNTRIES
,
479 'ip_blocks': self
._GEO
_IP
_BLOCKS
,
482 self
._real
_initialize
()
485 def _initialize_geo_bypass(self
, geo_bypass_context
):
487 Initialize geo restriction bypass mechanism.
489 This method is used to initialize geo bypass mechanism based on faking
490 X-Forwarded-For HTTP header. A random country from provided country list
491 is selected and a random IP belonging to this country is generated. This
492 IP will be passed as X-Forwarded-For HTTP header in all subsequent
495 This method will be used for initial geo bypass mechanism initialization
496 during the instance initialization with _GEO_COUNTRIES and
499 You may also manually call it from extractor's code if geo bypass
500 information is not available beforehand (e.g. obtained during
501 extraction) or due to some other reason. In this case you should pass
502 this information in geo bypass context passed as first argument. It may
503 contain following fields:
505 countries: List of geo unrestricted countries (similar
507 ip_blocks: List of geo unrestricted IP blocks in CIDR notation
508 (similar to _GEO_IP_BLOCKS)
511 if not self
._x
_forwarded
_for
_ip
:
513 # Geo bypass mechanism is explicitly disabled by user
514 if not self
.get_param('geo_bypass', True):
517 if not geo_bypass_context
:
518 geo_bypass_context
= {}
520 # Backward compatibility: previously _initialize_geo_bypass
521 # expected a list of countries, some 3rd party code may still use
523 if isinstance(geo_bypass_context
, (list, tuple)):
524 geo_bypass_context
= {
525 'countries': geo_bypass_context
,
528 # The whole point of geo bypass mechanism is to fake IP
529 # as X-Forwarded-For HTTP header based on some IP block or
532 # Path 1: bypassing based on IP block in CIDR notation
534 # Explicit IP block specified by user, use it right away
535 # regardless of whether extractor is geo bypassable or not
536 ip_block
= self
.get_param('geo_bypass_ip_block', None)
538 # Otherwise use random IP block from geo bypass context but only
539 # if extractor is known as geo bypassable
541 ip_blocks
= geo_bypass_context
.get('ip_blocks')
542 if self
._GEO
_BYPASS
and ip_blocks
:
543 ip_block
= random
.choice(ip_blocks
)
546 self
._x
_forwarded
_for
_ip
= GeoUtils
.random_ipv4(ip_block
)
547 self
._downloader
.write_debug(
548 '[debug] Using fake IP %s as X-Forwarded-For' % self
._x
_forwarded
_for
_ip
)
551 # Path 2: bypassing based on country code
553 # Explicit country code specified by user, use it right away
554 # regardless of whether extractor is geo bypassable or not
555 country
= self
.get_param('geo_bypass_country', None)
557 # Otherwise use random country code from geo bypass context but
558 # only if extractor is known as geo bypassable
560 countries
= geo_bypass_context
.get('countries')
561 if self
._GEO
_BYPASS
and countries
:
562 country
= random
.choice(countries
)
565 self
._x
_forwarded
_for
_ip
= GeoUtils
.random_ipv4(country
)
566 self
._downloader
.write_debug(
567 'Using fake IP %s (%s) as X-Forwarded-For' % (self
._x
_forwarded
_for
_ip
, country
.upper()))
569 def extract(self
, url
):
570 """Extracts URL information and returns it in list of dicts."""
575 self
.write_debug('Extracting URL: %s' % url
)
576 ie_result
= self
._real
_extract
(url
)
577 if ie_result
is None:
579 if self
._x
_forwarded
_for
_ip
:
580 ie_result
['__x_forwarded_for_ip'] = self
._x
_forwarded
_for
_ip
581 subtitles
= ie_result
.get('subtitles')
582 if (subtitles
and 'live_chat' in subtitles
583 and 'no-live-chat' in self
.get_param('compat_opts', [])):
584 del subtitles
['live_chat']
586 except GeoRestrictedError
as e
:
587 if self
.__maybe
_fake
_ip
_and
_retry
(e
.countries
):
590 except ExtractorError
:
592 except compat_http_client
.IncompleteRead
as e
:
593 raise ExtractorError('A network error has occurred.', cause
=e
, expected
=True)
594 except (KeyError, StopIteration) as e
:
595 raise ExtractorError('An extractor error has occurred.', cause
=e
)
597 def __maybe_fake_ip_and_retry(self
, countries
):
598 if (not self
.get_param('geo_bypass_country', None)
600 and self
.get_param('geo_bypass', True)
601 and not self
._x
_forwarded
_for
_ip
603 country_code
= random
.choice(countries
)
604 self
._x
_forwarded
_for
_ip
= GeoUtils
.random_ipv4(country_code
)
605 if self
._x
_forwarded
_for
_ip
:
607 'Video is geo restricted. Retrying extraction with fake IP %s (%s) as X-Forwarded-For.'
608 % (self
._x
_forwarded
_for
_ip
, country_code
.upper()))
612 def set_downloader(self
, downloader
):
613 """Sets the downloader for this IE."""
614 self
._downloader
= downloader
616 def _real_initialize(self
):
617 """Real initialization process. Redefine in subclasses."""
620 def _real_extract(self
, url
):
621 """Real extraction process. Redefine in subclasses."""
626 """A string for getting the InfoExtractor with get_info_extractor"""
627 return cls
.__name
__[:-2]
631 return compat_str(type(self
).__name
__[:-2])
634 def __can_accept_status_code(err
, expected_status
):
635 assert isinstance(err
, compat_urllib_error
.HTTPError
)
636 if expected_status
is None:
638 elif callable(expected_status
):
639 return expected_status(err
.code
) is True
641 return err
.code
in variadic(expected_status
)
643 def _request_webpage(self
, url_or_request
, video_id
, note
=None, errnote
=None, fatal
=True, data
=None, headers
={}, query={}
, expected_status
=None):
645 Return the response handle.
647 See _download_webpage docstring for arguments specification.
649 if not self
._downloader
._first
_webpage
_request
:
650 sleep_interval
= float_or_none(self
.get_param('sleep_interval_requests')) or 0
651 if sleep_interval
> 0:
652 self
.to_screen('Sleeping %s seconds ...' % sleep_interval
)
653 time
.sleep(sleep_interval
)
655 self
._downloader
._first
_webpage
_request
= False
658 self
.report_download_webpage(video_id
)
659 elif note
is not False:
661 self
.to_screen('%s' % (note
,))
663 self
.to_screen('%s: %s' % (video_id
, note
))
665 # Some sites check X-Forwarded-For HTTP header in order to figure out
666 # the origin of the client behind proxy. This allows bypassing geo
667 # restriction by faking this header's value to IP that belongs to some
668 # geo unrestricted country. We will do so once we encounter any
669 # geo restriction error.
670 if self
._x
_forwarded
_for
_ip
:
671 if 'X-Forwarded-For' not in headers
:
672 headers
['X-Forwarded-For'] = self
._x
_forwarded
_for
_ip
674 if isinstance(url_or_request
, compat_urllib_request
.Request
):
675 url_or_request
= update_Request(
676 url_or_request
, data
=data
, headers
=headers
, query
=query
)
679 url_or_request
= update_url_query(url_or_request
, query
)
680 if data
is not None or headers
:
681 url_or_request
= sanitized_Request(url_or_request
, data
, headers
)
683 return self
._downloader
.urlopen(url_or_request
)
684 except network_exceptions
as err
:
685 if isinstance(err
, compat_urllib_error
.HTTPError
):
686 if self
.__can
_accept
_status
_code
(err
, expected_status
):
687 # Retain reference to error to prevent file object from
688 # being closed before it can be read. Works around the
689 # effects of <https://bugs.python.org/issue15002>
690 # introduced in Python 3.4.1.
697 errnote
= 'Unable to download webpage'
699 errmsg
= '%s: %s' % (errnote
, error_to_compat_str(err
))
701 raise ExtractorError(errmsg
, sys
.exc_info()[2], cause
=err
)
703 self
.report_warning(errmsg
)
706 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):
708 Return a tuple (page content as string, URL handle).
710 See _download_webpage docstring for arguments specification.
712 # Strip hashes from the URL (#1038)
713 if isinstance(url_or_request
, (compat_str
, str)):
714 url_or_request
= url_or_request
.partition('#')[0]
716 urlh
= self
._request
_webpage
(url_or_request
, video_id
, note
, errnote
, fatal
, data
=data
, headers
=headers
, query
=query
, expected_status
=expected_status
)
720 content
= self
._webpage
_read
_content
(urlh
, url_or_request
, video_id
, note
, errnote
, fatal
, encoding
=encoding
)
721 return (content
, urlh
)
724 def _guess_encoding_from_content(content_type
, webpage_bytes
):
725 m
= re
.match(r
'[a-zA-Z0-9_.-]+/[a-zA-Z0-9_.-]+\s*;\s*charset=(.+)', content_type
)
727 encoding
= m
.group(1)
729 m
= re
.search(br
'<meta[^>]+charset=[\'"]?([^\'")]+)[ /\'">]',
730 webpage_bytes[:1024])
732 encoding = m.group(1).decode('ascii')
733 elif webpage_bytes.startswith(b'\xff\xfe'):
740 def __check_blocked(self, content):
741 first_block = content[:512]
742 if ('<title>Access to this site is blocked</title>' in content
743 and 'Websense' in first_block):
744 msg = 'Access to this webpage has been blocked by Websense filtering software in your network.'
745 blocked_iframe = self._html_search_regex(
746 r'<iframe src="([^
"]+)"', content,
747 'Websense information URL
', default=None)
749 msg += ' Visit
%s for more details
' % blocked_iframe
750 raise ExtractorError(msg, expected=True)
751 if '<title
>The URL you requested has been blocked
</title
>' in first_block:
753 'Access to this webpage has been blocked by Indian censorship
. '
754 'Use a VPN
or proxy
server (with --proxy
) to route around it
.')
755 block_msg = self._html_search_regex(
756 r'</h1
><p
>(.*?
)</p
>',
757 content, 'block message
', default=None)
759 msg += ' (Message
: "%s")' % block_msg.replace('\n', ' ')
760 raise ExtractorError(msg, expected=True)
761 if ('<title
>TTK
:: Доступ к ресурсу ограничен
</title
>' in content
762 and 'blocklist
.rkn
.gov
.ru
' in content):
763 raise ExtractorError(
764 'Access to this webpage has been blocked by decision of the Russian government
. '
765 'Visit http
://blocklist
.rkn
.gov
.ru
/ for a block reason
.',
768 def _webpage_read_content(self, urlh, url_or_request, video_id, note=None, errnote=None, fatal=True, prefix=None, encoding=None):
769 content_type = urlh.headers.get('Content
-Type
', '')
770 webpage_bytes = urlh.read()
771 if prefix is not None:
772 webpage_bytes = prefix + webpage_bytes
774 encoding = self._guess_encoding_from_content(content_type, webpage_bytes)
775 if self.get_param('dump_intermediate_pages
', False):
776 self.to_screen('Dumping request to
' + urlh.geturl())
777 dump = base64.b64encode(webpage_bytes).decode('ascii
')
778 self._downloader.to_screen(dump)
779 if self.get_param('write_pages
', False):
780 basen = '%s_%s' % (video_id, urlh.geturl())
782 h = '___
' + hashlib.md5(basen.encode('utf
-8')).hexdigest()
783 basen = basen[:240 - len(h)] + h
784 raw_filename = basen + '.dump
'
785 filename = sanitize_filename(raw_filename, restricted=True)
786 self.to_screen('Saving request to
' + filename)
787 # Working around MAX_PATH limitation on Windows (see
788 # http://msdn.microsoft.com/en-us/library/windows/desktop/aa365247(v=vs.85).aspx)
789 if compat_os_name == 'nt
':
790 absfilepath = os.path.abspath(filename)
791 if len(absfilepath) > 259:
792 filename = '\\\\?
\\' + absfilepath
793 with open(filename, 'wb
') as outf:
794 outf.write(webpage_bytes)
797 content = webpage_bytes.decode(encoding, 'replace
')
799 content = webpage_bytes.decode('utf
-8', 'replace
')
801 self.__check_blocked(content)
805 def _download_webpage(
806 self, url_or_request, video_id, note=None, errnote=None,
807 fatal=True, tries=1, timeout=5, encoding=None, data=None,
808 headers={}, query={}, expected_status=None):
810 Return the data of the page as a string.
813 url_or_request -- plain text URL as a string or
814 a compat_urllib_request.Requestobject
815 video_id -- Video/playlist/item identifier (string)
818 note -- note printed before downloading (string)
819 errnote -- note printed in case of an error (string)
820 fatal -- flag denoting whether error should be considered fatal,
821 i.e. whether it should cause ExtractionError to be raised,
822 otherwise a warning will be reported and extraction continued
823 tries -- number of tries
824 timeout -- sleep interval between tries
825 encoding -- encoding for a page content decoding, guessed automatically
826 when not explicitly specified
827 data -- POST data (bytes)
828 headers -- HTTP headers (dict)
829 query -- URL query (dict)
830 expected_status -- allows to accept failed HTTP requests (non 2xx
831 status code) by explicitly specifying a set of accepted status
832 codes. Can be any of the following entities:
833 - an integer type specifying an exact failed status code to
835 - a list or a tuple of integer types specifying a list of
836 failed status codes to accept
837 - a callable accepting an actual failed status code and
838 returning True if it should be accepted
839 Note that this argument does not affect success status codes (2xx)
840 which are always accepted.
845 while success is False:
847 res = self._download_webpage_handle(
848 url_or_request, video_id, note, errnote, fatal,
849 encoding=encoding, data=data, headers=headers, query=query,
850 expected_status=expected_status)
852 except compat_http_client.IncompleteRead as e:
854 if try_count >= tries:
856 self._sleep(timeout, video_id)
863 def _download_xml_handle(
864 self, url_or_request, video_id, note='Downloading XML
',
865 errnote='Unable to download XML
', transform_source=None,
866 fatal=True, encoding=None, data=None, headers={}, query={},
867 expected_status=None):
869 Return a tuple (xml as an compat_etree_Element, URL handle).
871 See _download_webpage docstring for arguments specification.
873 res = self._download_webpage_handle(
874 url_or_request, video_id, note, errnote, fatal=fatal,
875 encoding=encoding, data=data, headers=headers, query=query,
876 expected_status=expected_status)
879 xml_string, urlh = res
880 return self._parse_xml(
881 xml_string, video_id, transform_source=transform_source,
885 self, url_or_request, video_id,
886 note='Downloading XML
', errnote='Unable to download XML
',
887 transform_source=None, fatal=True, encoding=None,
888 data=None, headers={}, query={}, expected_status=None):
890 Return the xml as an compat_etree_Element.
892 See _download_webpage docstring for arguments specification.
894 res = self._download_xml_handle(
895 url_or_request, video_id, note=note, errnote=errnote,
896 transform_source=transform_source, fatal=fatal, encoding=encoding,
897 data=data, headers=headers, query=query,
898 expected_status=expected_status)
899 return res if res is False else res[0]
901 def _parse_xml(self, xml_string, video_id, transform_source=None, fatal=True):
903 xml_string = transform_source(xml_string)
905 return compat_etree_fromstring(xml_string.encode('utf
-8'))
906 except compat_xml_parse_error as ve:
907 errmsg = '%s: Failed to parse XML
' % video_id
909 raise ExtractorError(errmsg, cause=ve)
911 self.report_warning(errmsg + str(ve))
913 def _download_json_handle(
914 self, url_or_request, video_id, note='Downloading JSON metadata
',
915 errnote='Unable to download JSON metadata
', transform_source=None,
916 fatal=True, encoding=None, data=None, headers={}, query={},
917 expected_status=None):
919 Return a tuple (JSON object, URL handle).
921 See _download_webpage docstring for arguments specification.
923 res = self._download_webpage_handle(
924 url_or_request, video_id, note, errnote, fatal=fatal,
925 encoding=encoding, data=data, headers=headers, query=query,
926 expected_status=expected_status)
929 json_string, urlh = res
930 return self._parse_json(
931 json_string, video_id, transform_source=transform_source,
935 self, url_or_request, video_id, note='Downloading JSON metadata
',
936 errnote='Unable to download JSON metadata
', transform_source=None,
937 fatal=True, encoding=None, data=None, headers={}, query={},
938 expected_status=None):
940 Return the JSON object as a dict.
942 See _download_webpage docstring for arguments specification.
944 res = self._download_json_handle(
945 url_or_request, video_id, note=note, errnote=errnote,
946 transform_source=transform_source, fatal=fatal, encoding=encoding,
947 data=data, headers=headers, query=query,
948 expected_status=expected_status)
949 return res if res is False else res[0]
951 def _parse_json(self, json_string, video_id, transform_source=None, fatal=True):
953 json_string = transform_source(json_string)
955 return json.loads(json_string)
956 except ValueError as ve:
957 errmsg = '%s: Failed to parse JSON
' % video_id
959 raise ExtractorError(errmsg, cause=ve)
961 self.report_warning(errmsg + str(ve))
963 def _parse_socket_response_as_json(self, data, video_id, transform_source=None, fatal=True):
964 return self._parse_json(
965 data[data.find('{'):data.rfind('}
') + 1],
966 video_id, transform_source, fatal)
968 def _download_socket_json_handle(
969 self, url_or_request, video_id, note='Polling socket
',
970 errnote='Unable to poll socket
', transform_source=None,
971 fatal=True, encoding=None, data=None, headers={}, query={},
972 expected_status=None):
974 Return a tuple (JSON object, URL handle).
976 See _download_webpage docstring for arguments specification.
978 res = self._download_webpage_handle(
979 url_or_request, video_id, note, errnote, fatal=fatal,
980 encoding=encoding, data=data, headers=headers, query=query,
981 expected_status=expected_status)
985 return self._parse_socket_response_as_json(
986 webpage, video_id, transform_source=transform_source,
989 def _download_socket_json(
990 self, url_or_request, video_id, note='Polling socket
',
991 errnote='Unable to poll socket
', transform_source=None,
992 fatal=True, encoding=None, data=None, headers={}, query={},
993 expected_status=None):
995 Return the JSON object as a dict.
997 See _download_webpage docstring for arguments specification.
999 res = self._download_socket_json_handle(
1000 url_or_request, video_id, note=note, errnote=errnote,
1001 transform_source=transform_source, fatal=fatal, encoding=encoding,
1002 data=data, headers=headers, query=query,
1003 expected_status=expected_status)
1004 return res if res is False else res[0]
1006 def report_warning(self, msg, video_id=None, *args, only_once=False, **kwargs):
1007 idstr = format_field(video_id, template='%s: ')
1008 msg = f'[{self.IE_NAME}
] {idstr}{msg}
'
1010 if f'WARNING
: {msg}
' in self._printed_messages:
1012 self._printed_messages.add(f'WARNING
: {msg}
')
1013 self._downloader.report_warning(msg, *args, **kwargs)
1015 def to_screen(self, msg, *args, **kwargs):
1016 """Print msg to screen, prefixing it with '[ie_name
]'"""
1017 self._downloader.to_screen('[%s] %s' % (self.IE_NAME, msg), *args, **kwargs)
1019 def write_debug(self, msg, *args, **kwargs):
1020 self._downloader.write_debug('[%s] %s' % (self.IE_NAME, msg), *args, **kwargs)
1022 def get_param(self, name, default=None, *args, **kwargs):
1023 if self._downloader:
1024 return self._downloader.params.get(name, default, *args, **kwargs)
1027 def report_extraction(self, id_or_name):
1028 """Report information extraction."""
1029 self.to_screen('%s: Extracting information
' % id_or_name)
1031 def report_download_webpage(self, video_id):
1032 """Report webpage download."""
1033 self.to_screen('%s: Downloading webpage
' % video_id)
1035 def report_age_confirmation(self):
1036 """Report attempt to confirm age."""
1037 self.to_screen('Confirming age
')
1039 def report_login(self):
1040 """Report attempt to log in."""
1041 self.to_screen('Logging
in')
1043 def raise_login_required(
1044 self, msg='This video
is only available
for registered users
',
1045 metadata_available=False, method='any
'):
1046 if metadata_available and self.get_param('ignore_no_formats_error
'):
1047 self.report_warning(msg)
1048 if method is not None:
1049 msg = '%s. %s' % (msg, self._LOGIN_HINTS[method])
1050 raise ExtractorError(msg, expected=True)
1052 def raise_geo_restricted(
1053 self, msg='This video
is not available
from your location due to geo restriction
',
1054 countries=None, metadata_available=False):
1055 if metadata_available and self.get_param('ignore_no_formats_error
'):
1056 self.report_warning(msg)
1058 raise GeoRestrictedError(msg, countries=countries)
1060 def raise_no_formats(self, msg, expected=False, video_id=None):
1061 if expected and self.get_param('ignore_no_formats_error
'):
1062 self.report_warning(msg, video_id)
1063 elif isinstance(msg, ExtractorError):
1066 raise ExtractorError(msg, expected=expected, video_id=video_id)
1068 # Methods for following #608
1070 def url_result(url, ie=None, video_id=None, video_title=None):
1071 """Returns a URL that points to a page that should be processed"""
1072 # TODO: ie should be the class used for getting the info
1073 video_info = {'_type
': 'url
',
1076 if video_id is not None:
1077 video_info['id'] = video_id
1078 if video_title is not None:
1079 video_info['title
'] = video_title
1082 def playlist_from_matches(self, matches, playlist_id=None, playlist_title=None, getter=None, ie=None):
1084 self.url_result(self._proto_relative_url(getter(m) if getter else m), ie)
1086 return self.playlist_result(
1087 urls, playlist_id=playlist_id, playlist_title=playlist_title)
1090 def playlist_result(entries, playlist_id=None, playlist_title=None, playlist_description=None, **kwargs):
1091 """Returns a playlist"""
1092 video_info = {'_type
': 'playlist
',
1094 video_info.update(kwargs)
1096 video_info['id'] = playlist_id
1098 video_info['title
'] = playlist_title
1099 if playlist_description is not None:
1100 video_info['description
'] = playlist_description
1103 def _search_regex(self, pattern, string, name, default=NO_DEFAULT, fatal=True, flags=0, group=None):
1105 Perform a regex search on the given string, using a single or a list of
1106 patterns returning the first matching group.
1107 In case of failure return a default value or raise a WARNING or a
1108 RegexNotFoundError, depending on fatal, specifying the field name.
1110 if isinstance(pattern, (str, compat_str, compiled_regex_type)):
1111 mobj = re.search(pattern, string, flags)
1114 mobj = re.search(p, string, flags)
1118 if not self.get_param('no_color
') and compat_os_name != 'nt
' and sys.stderr.isatty():
1119 _name = '\033[0;34m
%s\033[0m
' % name
1125 # return the first matching group
1126 return next(g for g in mobj.groups() if g is not None)
1127 elif isinstance(group, (list, tuple)):
1128 return tuple(mobj.group(g) for g in group)
1130 return mobj.group(group)
1131 elif default is not NO_DEFAULT:
1134 raise RegexNotFoundError('Unable to extract
%s' % _name)
1136 self.report_warning('unable to extract
%s' % _name + bug_reports_message())
1139 def _html_search_regex(self, pattern, string, name, default=NO_DEFAULT, fatal=True, flags=0, group=None):
1141 Like _search_regex, but strips HTML tags and unescapes entities.
1143 res = self._search_regex(pattern, string, name, default, fatal, flags, group)
1145 return clean_html(res).strip()
1149 def _get_netrc_login_info(self, netrc_machine=None):
1152 netrc_machine = netrc_machine or self._NETRC_MACHINE
1154 if self.get_param('usenetrc
', False):
1156 info = netrc.netrc().authenticators(netrc_machine)
1157 if info is not None:
1161 raise netrc.NetrcParseError(
1162 'No authenticators
for %s' % netrc_machine)
1163 except (IOError, netrc.NetrcParseError) as err:
1164 self.report_warning(
1165 'parsing
.netrc
: %s' % error_to_compat_str(err))
1167 return username, password
1169 def _get_login_info(self, username_option='username
', password_option='password
', netrc_machine=None):
1171 Get the login info as (username, password)
1172 First look for the manually specified credentials using username_option
1173 and password_option as keys in params dictionary. If no such credentials
1174 available look in the netrc file using the netrc_machine or _NETRC_MACHINE
1176 If there's no info available
, return (None, None)
1179 # Attempt to use provided username and password or .netrc data
1180 username = self.get_param(username_option)
1181 if username is not None:
1182 password = self.get_param(password_option)
1184 username, password = self._get_netrc_login_info(netrc_machine)
1186 return username, password
1188 def _get_tfa_info(self, note='two-factor verification code'):
1190 Get the two
-factor authentication info
1191 TODO
- asking the user will be required
for sms
/phone verify
1192 currently just uses the command line option
1193 If there
's no info available, return None
1196 tfa = self.get_param('twofactor
')
1200 return compat_getpass('Type
%s and press
[Return
]: ' % note)
1202 # Helper functions for extracting OpenGraph info
1204 def _og_regexes(prop):
1205 content_re = r'content
=(?
:"([^"]+?
)"|\'([^\']+?)\'|\s*([^\s"\'=<>`
]+?
))'
1206 property_re = (r'(?
:name|
property)=(?
:\'og
[:-]%(prop)s\'|
"og[:-]%(prop)s"|\s
*og
[:-]%(prop)s\b)'
1207 % {'prop': re.escape(prop)})
1208 template = r'<meta
[^
>]+?
%s[^
>]+?
%s'
1210 template % (property_re, content_re),
1211 template % (content_re, property_re),
1215 def _meta_regex(prop):
1216 return r'''(?isx)<meta
1217 (?=[^>]+(?:itemprop|name|property|id|http-equiv)=(["\']?)%s\1)
1218 [^>]+?content=(["\'])(?P<content>.*?)\2''' % re.escape(prop)
1220 def _og_search_property(self, prop, html, name=None, **kargs):
1221 prop = variadic(prop)
1223 name = 'OpenGraph
%s' % prop[0]
1226 og_regexes.extend(self._og_regexes(p))
1227 escaped = self._search_regex(og_regexes, html, name, flags=re.DOTALL, **kargs)
1230 return unescapeHTML(escaped)
1232 def _og_search_thumbnail(self, html, **kargs):
1233 return self._og_search_property('image
', html, 'thumbnail URL
', fatal=False, **kargs)
1235 def _og_search_description(self, html, **kargs):
1236 return self._og_search_property('description
', html, fatal=False, **kargs)
1238 def _og_search_title(self, html, **kargs):
1239 return self._og_search_property('title
', html, **kargs)
1241 def _og_search_video_url(self, html, name='video url
', secure=True, **kargs):
1242 regexes = self._og_regexes('video
') + self._og_regexes('video
:url
')
1244 regexes = self._og_regexes('video
:secure_url
') + regexes
1245 return self._html_search_regex(regexes, html, name, **kargs)
1247 def _og_search_url(self, html, **kargs):
1248 return self._og_search_property('url
', html, **kargs)
1250 def _html_search_meta(self, name, html, display_name=None, fatal=False, **kwargs):
1251 name = variadic(name)
1252 if display_name is None:
1253 display_name = name[0]
1254 return self._html_search_regex(
1255 [self._meta_regex(n) for n in name],
1256 html, display_name, fatal=fatal, group='content
', **kwargs)
1258 def _dc_search_uploader(self, html):
1259 return self._html_search_meta('dc
.creator
', html, 'uploader
')
1261 def _rta_search(self, html):
1262 # See http://www.rtalabel.org/index.php?content=howtofaq#single
1263 if re.search(r'(?ix
)<meta\s
+name
="rating"\s
+'
1264 r' content
="RTA-5042-1996-1400-1577-RTA"',
1269 def _media_rating_search(self, html):
1270 # See http://www.tjg-designs.com/WP/metadata-code-examples-adding-metadata-to-your-web-pages/
1271 rating = self._html_search_meta('rating
', html)
1283 return RATING_TABLE.get(rating.lower())
1285 def _family_friendly_search(self, html):
1286 # See http://schema.org/VideoObject
1287 family_friendly = self._html_search_meta(
1288 'isFamilyFriendly
', html, default=None)
1290 if not family_friendly:
1299 return RATING_TABLE.get(family_friendly.lower())
1301 def _twitter_search_player(self, html):
1302 return self._html_search_meta('twitter
:player
', html,
1303 'twitter card player
')
1305 def _search_json_ld(self, html, video_id, expected_type=None, **kwargs):
1306 json_ld_list = list(re.finditer(JSON_LD_RE, html))
1307 default = kwargs.get('default
', NO_DEFAULT)
1308 # JSON-LD may be malformed and thus `fatal` should be respected.
1309 # At the same time `default` may be passed that assumes `fatal=False`
1310 # for _search_regex. Let's simulate the same behavior here
as well
.
1311 fatal
= kwargs
.get('fatal', True) if default
is NO_DEFAULT
else False
1313 for mobj
in json_ld_list
:
1314 json_ld_item
= self
._parse
_json
(
1315 mobj
.group('json_ld'), video_id
, fatal
=fatal
)
1316 if not json_ld_item
:
1318 if isinstance(json_ld_item
, dict):
1319 json_ld
.append(json_ld_item
)
1320 elif isinstance(json_ld_item
, (list, tuple)):
1321 json_ld
.extend(json_ld_item
)
1323 json_ld
= self
._json
_ld
(json_ld
, video_id
, fatal
=fatal
, expected_type
=expected_type
)
1326 if default
is not NO_DEFAULT
:
1329 raise RegexNotFoundError('Unable to extract JSON-LD')
1331 self
.report_warning('unable to extract JSON-LD %s' % bug_reports_message())
1334 def _json_ld(self
, json_ld
, video_id
, fatal
=True, expected_type
=None):
1335 if isinstance(json_ld
, compat_str
):
1336 json_ld
= self
._parse
_json
(json_ld
, video_id
, fatal
=fatal
)
1340 if not isinstance(json_ld
, (list, tuple, dict)):
1342 if isinstance(json_ld
, dict):
1345 INTERACTION_TYPE_MAP
= {
1346 'CommentAction': 'comment',
1347 'AgreeAction': 'like',
1348 'DisagreeAction': 'dislike',
1349 'LikeAction': 'like',
1350 'DislikeAction': 'dislike',
1351 'ListenAction': 'view',
1352 'WatchAction': 'view',
1353 'ViewAction': 'view',
1356 def extract_interaction_type(e
):
1357 interaction_type
= e
.get('interactionType')
1358 if isinstance(interaction_type
, dict):
1359 interaction_type
= interaction_type
.get('@type')
1360 return str_or_none(interaction_type
)
1362 def extract_interaction_statistic(e
):
1363 interaction_statistic
= e
.get('interactionStatistic')
1364 if isinstance(interaction_statistic
, dict):
1365 interaction_statistic
= [interaction_statistic
]
1366 if not isinstance(interaction_statistic
, list):
1368 for is_e
in interaction_statistic
:
1369 if not isinstance(is_e
, dict):
1371 if is_e
.get('@type') != 'InteractionCounter':
1373 interaction_type
= extract_interaction_type(is_e
)
1374 if not interaction_type
:
1376 # For interaction count some sites provide string instead of
1377 # an integer (as per spec) with non digit characters (e.g. ",")
1378 # so extracting count with more relaxed str_to_int
1379 interaction_count
= str_to_int(is_e
.get('userInteractionCount'))
1380 if interaction_count
is None:
1382 count_kind
= INTERACTION_TYPE_MAP
.get(interaction_type
.split('/')[-1])
1385 count_key
= '%s_count' % count_kind
1386 if info
.get(count_key
) is not None:
1388 info
[count_key
] = interaction_count
1390 def extract_video_object(e
):
1391 assert e
['@type'] == 'VideoObject'
1392 author
= e
.get('author')
1394 'url': url_or_none(e
.get('contentUrl')),
1395 'title': unescapeHTML(e
.get('name')),
1396 'description': unescapeHTML(e
.get('description')),
1397 'thumbnail': url_or_none(e
.get('thumbnailUrl') or e
.get('thumbnailURL')),
1398 'duration': parse_duration(e
.get('duration')),
1399 'timestamp': unified_timestamp(e
.get('uploadDate')),
1400 # author can be an instance of 'Organization' or 'Person' types.
1401 # both types can have 'name' property(inherited from 'Thing' type). [1]
1402 # however some websites are using 'Text' type instead.
1403 # 1. https://schema.org/VideoObject
1404 'uploader': author
.get('name') if isinstance(author
, dict) else author
if isinstance(author
, compat_str
) else None,
1405 'filesize': float_or_none(e
.get('contentSize')),
1406 'tbr': int_or_none(e
.get('bitrate')),
1407 'width': int_or_none(e
.get('width')),
1408 'height': int_or_none(e
.get('height')),
1409 'view_count': int_or_none(e
.get('interactionCount')),
1411 extract_interaction_statistic(e
)
1415 item_type
= e
.get('@type')
1416 if expected_type
is not None and expected_type
!= item_type
:
1418 if item_type
in ('TVEpisode', 'Episode'):
1419 episode_name
= unescapeHTML(e
.get('name'))
1421 'episode': episode_name
,
1422 'episode_number': int_or_none(e
.get('episodeNumber')),
1423 'description': unescapeHTML(e
.get('description')),
1425 if not info
.get('title') and episode_name
:
1426 info
['title'] = episode_name
1427 part_of_season
= e
.get('partOfSeason')
1428 if isinstance(part_of_season
, dict) and part_of_season
.get('@type') in ('TVSeason', 'Season', 'CreativeWorkSeason'):
1430 'season': unescapeHTML(part_of_season
.get('name')),
1431 'season_number': int_or_none(part_of_season
.get('seasonNumber')),
1433 part_of_series
= e
.get('partOfSeries') or e
.get('partOfTVSeries')
1434 if isinstance(part_of_series
, dict) and part_of_series
.get('@type') in ('TVSeries', 'Series', 'CreativeWorkSeries'):
1435 info
['series'] = unescapeHTML(part_of_series
.get('name'))
1436 elif item_type
== 'Movie':
1438 'title': unescapeHTML(e
.get('name')),
1439 'description': unescapeHTML(e
.get('description')),
1440 'duration': parse_duration(e
.get('duration')),
1441 'timestamp': unified_timestamp(e
.get('dateCreated')),
1443 elif item_type
in ('Article', 'NewsArticle'):
1445 'timestamp': parse_iso8601(e
.get('datePublished')),
1446 'title': unescapeHTML(e
.get('headline')),
1447 'description': unescapeHTML(e
.get('articleBody')),
1449 elif item_type
== 'VideoObject':
1450 extract_video_object(e
)
1451 if expected_type
is None:
1455 video
= e
.get('video')
1456 if isinstance(video
, dict) and video
.get('@type') == 'VideoObject':
1457 extract_video_object(video
)
1458 if expected_type
is None:
1462 return dict((k
, v
) for k
, v
in info
.items() if v
is not None)
1465 def _hidden_inputs(html
):
1466 html
= re
.sub(r
'<!--(?:(?!<!--).)*-->', '', html
)
1468 for input in re
.findall(r
'(?i)(<input[^>]+>)', html
):
1469 attrs
= extract_attributes(input)
1472 if attrs
.get('type') not in ('hidden', 'submit'):
1474 name
= attrs
.get('name') or attrs
.get('id')
1475 value
= attrs
.get('value')
1476 if name
and value
is not None:
1477 hidden_inputs
[name
] = value
1478 return hidden_inputs
1480 def _form_hidden_inputs(self
, form_id
, html
):
1481 form
= self
._search
_regex
(
1482 r
'(?is)<form[^>]+?id=(["\'])%s\
1[^
>]*>(?P
<form
>.+?
)</form
>' % form_id,
1483 html, '%s form
' % form_id, group='form
')
1484 return self._hidden_inputs(form)
1487 regex = r' *((?P
<reverse
>\
+)?
(?P
<field
>[a
-zA
-Z0
-9_]+)((?P
<separator
>[~
:])(?P
<limit
>.*?
))?
)?
*$
'
1489 default = ('hidden
', 'aud_or_vid
', 'hasvid
', 'ie_pref
', 'lang
', 'quality
',
1490 'res
', 'fps
', 'codec
:vp9
.2
', 'size
', 'br
', 'asr
',
1491 'proto
', 'ext
', 'hasaud
', 'source
', 'format_id
') # These must not be aliases
1492 ytdl_default = ('hasaud
', 'quality
', 'tbr
', 'filesize
', 'vbr
',
1493 'height
', 'width
', 'proto
', 'vext
', 'abr
', 'aext
',
1494 'fps
', 'fs_approx
', 'source
', 'format_id
')
1497 'vcodec
': {'type': 'ordered
', 'regex
': True,
1498 'order
': ['av0?
1', 'vp0?
9.2', 'vp0?
9', '[hx
]265|he?vc?
', '[hx
]264|avc
', 'vp0?
8', 'mp4v|h263
', 'theora
', '', None, 'none
']},
1499 'acodec
': {'type': 'ordered
', 'regex
': True,
1500 'order
': ['opus
', 'vorbis
', 'aac
', 'mp?
4a?
', 'mp3
', 'e?a?c
-?
3', 'dts
', '', None, 'none
']},
1501 'proto
': {'type': 'ordered
', 'regex
': True, 'field
': 'protocol
',
1502 'order
': ['(ht|f
)tps
', '(ht|f
)tp$
', 'm3u8
.+', '.*dash
', 'ws|websocket
', '', 'mms|rtsp
', 'none
', 'f4
']},
1503 'vext
': {'type': 'ordered
', 'field
': 'video_ext
',
1504 'order
': ('mp4
', 'webm
', 'flv
', '', 'none
'),
1505 'order_free
': ('webm
', 'mp4
', 'flv
', '', 'none
')},
1506 'aext
': {'type': 'ordered
', 'field
': 'audio_ext
',
1507 'order
': ('m4a
', 'aac
', 'mp3
', 'ogg
', 'opus
', 'webm
', '', 'none
'),
1508 'order_free
': ('opus
', 'ogg
', 'webm
', 'm4a
', 'mp3
', 'aac
', '', 'none
')},
1509 'hidden
': {'visible': False, 'forced': True, 'type': 'extractor', 'max': -1000},
1510 'aud_or_vid
': {'visible
': False, 'forced
': True, 'type': 'multiple
',
1511 'field
': ('vcodec
', 'acodec
'),
1512 'function
': lambda it: int(any(v != 'none
' for v in it))},
1513 'ie_pref
': {'priority': True, 'type': 'extractor'},
1514 'hasvid
': {'priority': True, 'field': 'vcodec', 'type': 'boolean', 'not_in_list': ('none',)},
1515 'hasaud
': {'field': 'acodec', 'type': 'boolean', 'not_in_list': ('none',)},
1516 'lang
': {'priority': True, 'convert': 'ignore', 'field': 'language_preference'},
1517 'quality
': {'convert': 'float_none', 'default': -1},
1518 'filesize
': {'convert': 'bytes'},
1519 'fs_approx
': {'convert': 'bytes', 'field': 'filesize_approx'},
1520 'id': {'convert': 'string', 'field': 'format_id'},
1521 'height
': {'convert': 'float_none'},
1522 'width
': {'convert': 'float_none'},
1523 'fps
': {'convert': 'float_none'},
1524 'tbr
': {'convert': 'float_none'},
1525 'vbr
': {'convert': 'float_none'},
1526 'abr
': {'convert': 'float_none'},
1527 'asr
': {'convert': 'float_none'},
1528 'source
': {'convert': 'ignore', 'field': 'source_preference'},
1530 'codec
': {'type': 'combined', 'field': ('vcodec', 'acodec')},
1531 'br
': {'type': 'combined', 'field': ('tbr', 'vbr', 'abr'), 'same_limit': True},
1532 'size
': {'type': 'combined', 'same_limit': True, 'field': ('filesize', 'fs_approx')},
1533 'ext
': {'type': 'combined', 'field': ('vext', 'aext')},
1534 'res
': {'type': 'multiple
', 'field
': ('height
', 'width
'),
1535 'function
': lambda it: (lambda l: min(l) if l else 0)(tuple(filter(None, it)))},
1537 # Most of these exist only for compatibility reasons
1538 'dimension
': {'type': 'alias', 'field': 'res'},
1539 'resolution
': {'type': 'alias', 'field': 'res'},
1540 'extension
': {'type': 'alias', 'field': 'ext'},
1541 'bitrate
': {'type': 'alias', 'field': 'br'},
1542 'total_bitrate
': {'type': 'alias', 'field': 'tbr'},
1543 'video_bitrate
': {'type': 'alias', 'field': 'vbr'},
1544 'audio_bitrate
': {'type': 'alias', 'field': 'abr'},
1545 'framerate
': {'type': 'alias', 'field': 'fps'},
1546 'language_preference
': {'type': 'alias', 'field': 'lang'}, # not named as 'language
' because such a field exists
1547 'protocol
': {'type': 'alias', 'field': 'proto'},
1548 'source_preference
': {'type': 'alias', 'field': 'source'},
1549 'filesize_approx
': {'type': 'alias', 'field': 'fs_approx'},
1550 'filesize_estimate
': {'type': 'alias', 'field': 'size'},
1551 'samplerate
': {'type': 'alias', 'field': 'asr'},
1552 'video_ext
': {'type': 'alias', 'field': 'vext'},
1553 'audio_ext
': {'type': 'alias', 'field': 'aext'},
1554 'video_codec
': {'type': 'alias', 'field': 'vcodec'},
1555 'audio_codec
': {'type': 'alias', 'field': 'acodec'},
1556 'video
': {'type': 'alias', 'field': 'hasvid'},
1557 'has_video
': {'type': 'alias', 'field': 'hasvid'},
1558 'audio
': {'type': 'alias', 'field': 'hasaud'},
1559 'has_audio
': {'type': 'alias', 'field': 'hasaud'},
1560 'extractor
': {'type': 'alias', 'field': 'ie_pref'},
1561 'preference
': {'type': 'alias', 'field': 'ie_pref'},
1562 'extractor_preference
': {'type': 'alias', 'field': 'ie_pref'},
1563 'format_id
': {'type': 'alias', 'field': 'id'},
1568 def _get_field_setting(self, field, key):
1569 if field not in self.settings:
1570 self.settings[field] = {}
1571 propObj = self.settings[field]
1572 if key not in propObj:
1573 type = propObj.get('type')
1575 default = 'preference
' if type == 'extractor
' else (field,) if type in ('combined
', 'multiple
') else field
1576 elif key == 'convert
':
1577 default = 'order
' if type == 'ordered
' else 'float_string
' if field else 'ignore
'
1579 default = {'type': 'field', 'visible': True, 'order': [], 'not_in_list': (None,)}.get(key, None)
1580 propObj[key] = default
1583 def _resolve_field_value(self, field, value, convertNone=False):
1588 value = value.lower()
1589 conversion = self._get_field_setting(field, 'convert
')
1590 if conversion == 'ignore
':
1592 if conversion == 'string
':
1594 elif conversion == 'float_none
':
1595 return float_or_none(value)
1596 elif conversion == 'bytes':
1597 return FileDownloader.parse_bytes(value)
1598 elif conversion == 'order
':
1599 order_list = (self._use_free_order and self._get_field_setting(field, 'order_free
')) or self._get_field_setting(field, 'order
')
1600 use_regex = self._get_field_setting(field, 'regex
')
1601 list_length = len(order_list)
1602 empty_pos = order_list.index('') if '' in order_list else list_length + 1
1603 if use_regex and value is not None:
1604 for i, regex in enumerate(order_list):
1605 if regex and re.match(regex, value):
1606 return list_length - i
1607 return list_length - empty_pos # not in list
1608 else: # not regex or value = None
1609 return list_length - (order_list.index(value) if value in order_list else empty_pos)
1611 if value.isnumeric():
1614 self.settings[field]['convert
'] = 'string
'
1617 def evaluate_params(self, params, sort_extractor):
1618 self._use_free_order = params.get('prefer_free_formats
', False)
1619 self._sort_user = params.get('format_sort
', [])
1620 self._sort_extractor = sort_extractor
1622 def add_item(field, reverse, closest, limit_text):
1623 field = field.lower()
1624 if field in self._order:
1626 self._order.append(field)
1627 limit = self._resolve_field_value(field, limit_text)
1630 'closest
': False if limit is None else closest,
1631 'limit_text
': limit_text,
1633 if field in self.settings:
1634 self.settings[field].update(data)
1636 self.settings[field] = data
1639 tuple(field for field in self.default if self._get_field_setting(field, 'forced
'))
1640 + (tuple() if params.get('format_sort_force
', False)
1641 else tuple(field for field in self.default if self._get_field_setting(field, 'priority
')))
1642 + tuple(self._sort_user) + tuple(sort_extractor) + self.default)
1644 for item in sort_list:
1645 match = re.match(self.regex, item)
1647 raise ExtractorError('Invalid format sort string
"%s" given by extractor
' % item)
1648 field = match.group('field
')
1651 if self._get_field_setting(field, 'type') == 'alias
':
1652 field = self._get_field_setting(field, 'field
')
1653 reverse = match.group('reverse
') is not None
1654 closest = match.group('separator
') == '~
'
1655 limit_text = match.group('limit
')
1657 has_limit = limit_text is not None
1658 has_multiple_fields = self._get_field_setting(field, 'type') == 'combined
'
1659 has_multiple_limits = has_limit and has_multiple_fields and not self._get_field_setting(field, 'same_limit
')
1661 fields = self._get_field_setting(field, 'field
') if has_multiple_fields else (field,)
1662 limits = limit_text.split(":") if has_multiple_limits else (limit_text,) if has_limit else tuple()
1663 limit_count = len(limits)
1664 for (i, f) in enumerate(fields):
1665 add_item(f, reverse, closest,
1666 limits[i] if i < limit_count
1667 else limits[0] if has_limit and not has_multiple_limits
1670 def print_verbose_info(self, write_debug):
1672 write_debug('Sort order given by user
: %s' % ', '.join(self._sort_user))
1673 if self._sort_extractor:
1674 write_debug('Sort order given by extractor
: %s' % ', '.join(self._sort_extractor))
1675 write_debug('Formats
sorted by
: %s' % ', '.join(['%s%s%s' % (
1676 '+' if self._get_field_setting(field, 'reverse
') else '', field,
1677 '%s%s(%s)' % ('~
' if self._get_field_setting(field, 'closest
') else ':',
1678 self._get_field_setting(field, 'limit_text
'),
1679 self._get_field_setting(field, 'limit
'))
1680 if self._get_field_setting(field, 'limit_text
') is not None else '')
1681 for field in self._order if self._get_field_setting(field, 'visible
')]))
1683 def _calculate_field_preference_from_value(self, format, field, type, value):
1684 reverse = self._get_field_setting(field, 'reverse
')
1685 closest = self._get_field_setting(field, 'closest
')
1686 limit = self._get_field_setting(field, 'limit
')
1688 if type == 'extractor
':
1689 maximum = self._get_field_setting(field, 'max')
1690 if value is None or (maximum is not None and value >= maximum):
1692 elif type == 'boolean
':
1693 in_list = self._get_field_setting(field, 'in_list
')
1694 not_in_list = self._get_field_setting(field, 'not_in_list
')
1695 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
1696 elif type == 'ordered
':
1697 value = self._resolve_field_value(field, value, True)
1699 # try to convert to number
1700 val_num = float_or_none(value, default=self._get_field_setting(field, 'default
'))
1701 is_num = self._get_field_setting(field, 'convert
') != 'string
' and val_num is not None
1705 return ((-10, 0) if value is None
1706 else (1, value, 0) if not is_num # if a field has mixed strings and numbers, strings are sorted higher
1707 else (0, -abs(value - limit), value - limit if reverse else limit - value) if closest
1708 else (0, value, 0) if not reverse and (limit is None or value <= limit)
1709 else (0, -value, 0) if limit is None or (reverse and value == limit) or value > limit
1710 else (-1, value, 0))
1712 def _calculate_field_preference(self, format, field):
1713 type = self._get_field_setting(field, 'type') # extractor, boolean, ordered, field, multiple
1714 get_value = lambda f: format.get(self._get_field_setting(f, 'field
'))
1715 if type == 'multiple
':
1716 type = 'field
' # Only 'field
' is allowed in multiple for now
1717 actual_fields = self._get_field_setting(field, 'field
')
1719 value = self._get_field_setting(field, 'function
')(get_value(f) for f in actual_fields)
1721 value = get_value(field)
1722 return self._calculate_field_preference_from_value(format, field, type, value)
1724 def calculate_preference(self, format):
1725 # Determine missing protocol
1726 if not format.get('protocol
'):
1727 format['protocol
'] = determine_protocol(format)
1729 # Determine missing ext
1730 if not format.get('ext
') and 'url
' in format:
1731 format['ext
'] = determine_ext(format['url
'])
1732 if format.get('vcodec
') == 'none
':
1733 format['audio_ext
'] = format['ext
'] if format.get('acodec
') != 'none
' else 'none
'
1734 format['video_ext
'] = 'none
'
1736 format['video_ext
'] = format['ext
']
1737 format['audio_ext
'] = 'none
'
1738 # if format.get('preference
') is None and format.get('ext
') in ('f4f
', 'f4m
'): # Not supported?
1739 # format['preference
'] = -1000
1741 # Determine missing bitrates
1742 if format.get('tbr
') is None:
1743 if format.get('vbr
') is not None and format.get('abr
') is not None:
1744 format['tbr
'] = format.get('vbr
', 0) + format.get('abr
', 0)
1746 if format.get('vcodec
') != "none" and format.get('vbr
') is None:
1747 format['vbr
'] = format.get('tbr
') - format.get('abr
', 0)
1748 if format.get('acodec
') != "none" and format.get('abr
') is None:
1749 format['abr
'] = format.get('tbr
') - format.get('vbr
', 0)
1751 return tuple(self._calculate_field_preference(format, field) for field in self._order)
1753 def _sort_formats(self, formats, field_preference=[]):
1755 if self.get_param('ignore_no_formats_error
'):
1757 raise ExtractorError('No video formats found
')
1758 format_sort = self.FormatSort() # params and to_screen are taken from the downloader
1759 format_sort.evaluate_params(self._downloader.params, field_preference)
1760 if self.get_param('verbose
', False):
1761 format_sort.print_verbose_info(self._downloader.write_debug)
1762 formats.sort(key=lambda f: format_sort.calculate_preference(f))
1764 def _check_formats(self, formats, video_id):
1766 formats[:] = filter(
1767 lambda f: self._is_valid_url(
1769 item='%s video format
' % f.get('format_id
') if f.get('format_id
') else 'video
'),
1773 def _remove_duplicate_formats(formats):
1777 if f['url
'] not in format_urls:
1778 format_urls.add(f['url
'])
1779 unique_formats.append(f)
1780 formats[:] = unique_formats
1782 def _is_valid_url(self, url, video_id, item='video
', headers={}):
1783 url = self._proto_relative_url(url, scheme='http
:')
1784 # For now assume non HTTP(S) URLs always valid
1785 if not (url.startswith('http
://') or url.startswith('https
://')):
1788 self._request_webpage(url, video_id, 'Checking
%s URL
' % item, headers=headers)
1790 except ExtractorError as e:
1792 '%s: %s URL
is invalid
, skipping
: %s'
1793 % (video_id, item, error_to_compat_str(e.cause)))
1796 def http_scheme(self):
1797 """ Either "http:" or "https:", depending on the user's preferences
"""
1800 if self.get_param('prefer_insecure', False)
1803 def _proto_relative_url(self, url, scheme=None):
1806 if url.startswith('//'):
1808 scheme = self.http_scheme()
1813 def _sleep(self, timeout, video_id, msg_template=None):
1814 if msg_template is None:
1815 msg_template = '%(video_id)s: Waiting for %(timeout)s seconds'
1816 msg = msg_template % {'video_id': video_id, 'timeout': timeout}
1820 def _extract_f4m_formats(self, manifest_url, video_id, preference=None, quality=None, f4m_id=None,
1821 transform_source=lambda s: fix_xml_ampersands(s).strip(),
1822 fatal=True, m3u8_id=None, data=None, headers={}, query={}):
1823 manifest = self._download_xml(
1824 manifest_url, video_id, 'Downloading f4m manifest',
1825 'Unable to download f4m manifest',
1826 # Some manifests may be malformed, e.g. prosiebensat1 generated manifests
1827 # (see https://github.com/ytdl-org/youtube-dl/issues/6215#issuecomment-121704244)
1828 transform_source=transform_source,
1829 fatal=fatal, data=data, headers=headers, query=query)
1831 if manifest is False:
1834 return self._parse_f4m_formats(
1835 manifest, manifest_url, video_id, preference=preference, quality=quality, f4m_id=f4m_id,
1836 transform_source=transform_source, fatal=fatal, m3u8_id=m3u8_id)
1838 def _parse_f4m_formats(self, manifest, manifest_url, video_id, preference=None, quality=None, f4m_id=None,
1839 transform_source=lambda s: fix_xml_ampersands(s).strip(),
1840 fatal=True, m3u8_id=None):
1841 if not isinstance(manifest, compat_etree_Element) and not fatal:
1844 # currently yt-dlp cannot decode the playerVerificationChallenge as Akamai uses Adobe Alchemy
1845 akamai_pv = manifest.find('{http://ns.adobe.com/f4m/1.0}pv-2.0')
1846 if akamai_pv is not None and ';' in akamai_pv.text:
1847 playerVerificationChallenge = akamai_pv.text.split(';')[0]
1848 if playerVerificationChallenge.strip() != '':
1852 manifest_version = '1.0'
1853 media_nodes = manifest.findall('{http://ns.adobe.com/f4m/1.0}media')
1855 manifest_version = '2.0'
1856 media_nodes = manifest.findall('{http://ns.adobe.com/f4m/2.0}media')
1857 # Remove unsupported DRM protected media from final formats
1858 # rendition (see https://github.com/ytdl-org/youtube-dl/issues/8573).
1859 media_nodes = remove_encrypted_media(media_nodes)
1863 manifest_base_url = get_base_url(manifest)
1865 bootstrap_info = xpath_element(
1866 manifest, ['{http://ns.adobe.com/f4m/1.0}bootstrapInfo', '{http://ns.adobe.com/f4m/2.0}bootstrapInfo'],
1867 'bootstrap info', default=None)
1870 mime_type = xpath_text(
1871 manifest, ['{http://ns.adobe.com/f4m/1.0}mimeType', '{http://ns.adobe.com/f4m/2.0}mimeType'],
1872 'base URL', default=None)
1873 if mime_type and mime_type.startswith('audio/'):
1876 for i, media_el in enumerate(media_nodes):
1877 tbr = int_or_none(media_el.attrib.get('bitrate'))
1878 width = int_or_none(media_el.attrib.get('width'))
1879 height = int_or_none(media_el.attrib.get('height'))
1880 format_id = '-'.join(filter(None, [f4m_id, compat_str(i if tbr is None else tbr)]))
1881 # If <bootstrapInfo> is present, the specified f4m is a
1882 # stream-level manifest, and only set-level manifests may refer to
1883 # external resources. See section 11.4 and section 4 of F4M spec
1884 if bootstrap_info is None:
1886 # @href is introduced in 2.0, see section 11.6 of F4M spec
1887 if manifest_version == '2.0':
1888 media_url = media_el.attrib.get('href')
1889 if media_url is None:
1890 media_url = media_el.attrib.get('url')
1894 media_url if media_url.startswith('http://') or media_url.startswith('https://')
1895 else ((manifest_base_url or '/'.join(manifest_url.split('/')[:-1])) + '/' + media_url))
1896 # If media_url is itself a f4m manifest do the recursive extraction
1897 # since bitrates in parent manifest (this one) and media_url manifest
1898 # may differ leading to inability to resolve the format by requested
1899 # bitrate in f4m downloader
1900 ext = determine_ext(manifest_url)
1902 f4m_formats = self._extract_f4m_formats(
1903 manifest_url, video_id, preference=preference, quality=quality, f4m_id=f4m_id,
1904 transform_source=transform_source, fatal=fatal)
1905 # Sometimes stream-level manifest contains single media entry that
1906 # does not contain any quality metadata (e.g. http://matchtv.ru/#live-player).
1907 # At the same time parent's media entry in set-level manifest may
1908 # contain it. We will copy it from parent in such cases.
1909 if len(f4m_formats) == 1:
1912 'tbr': f.get('tbr') or tbr,
1913 'width': f.get('width') or width,
1914 'height': f.get('height') or height,
1915 'format_id': f.get('format_id') if not tbr else format_id,
1918 formats.extend(f4m_formats)
1921 formats.extend(self._extract_m3u8_formats(
1922 manifest_url, video_id, 'mp4', preference=preference,
1923 quality=quality, m3u8_id=m3u8_id, fatal=fatal))
1926 'format_id': format_id,
1927 'url': manifest_url,
1928 'manifest_url': manifest_url,
1929 'ext': 'flv' if bootstrap_info is not None else None,
1935 'preference': preference,
1940 def _m3u8_meta_format(self, m3u8_url, ext=None, preference=None, quality=None, m3u8_id=None):
1942 'format_id': '-'.join(filter(None, [m3u8_id, 'meta'])),
1946 'preference': preference - 100 if preference else -100,
1948 'resolution': 'multiple',
1949 'format_note': 'Quality selection URL',
1952 def _extract_m3u8_formats(self, *args, **kwargs):
1953 fmts, subs = self._extract_m3u8_formats_and_subtitles(*args, **kwargs)
1955 self.report_warning(bug_reports_message(
1956 "Ignoring subtitle tracks found in the HLS manifest; "
1957 "if any subtitle tracks are missing,"
1961 def _extract_m3u8_formats_and_subtitles(
1962 self, m3u8_url, video_id, ext=None, entry_protocol='m3u8_native',
1963 preference=None, quality=None, m3u8_id=None, note=None,
1964 errnote=None, fatal=True, live=False, data=None, headers={},
1967 res = self._download_webpage_handle(
1969 note='Downloading m3u8 information' if note is None else note,
1970 errnote='Failed to download m3u8 information' if errnote is None else errnote,
1971 fatal=fatal, data=data, headers=headers, query=query)
1976 m3u8_doc, urlh = res
1977 m3u8_url = urlh.geturl()
1979 return self._parse_m3u8_formats_and_subtitles(
1980 m3u8_doc, m3u8_url, ext=ext, entry_protocol=entry_protocol,
1981 preference=preference, quality=quality, m3u8_id=m3u8_id,
1982 note=note, errnote=errnote, fatal=fatal, live=live, data=data,
1983 headers=headers, query=query, video_id=video_id)
1985 def _parse_m3u8_formats_and_subtitles(
1986 self, m3u8_doc, m3u8_url, ext=None, entry_protocol='m3u8_native',
1987 preference=None, quality=None, m3u8_id=None, live=False, note=None,
1988 errnote=None, fatal=True, data=None, headers={}, query={},
1990 formats, subtitles = [], {}
1992 if '#EXT-X-FAXS-CM:' in m3u8_doc: # Adobe Flash Access
1993 return formats, subtitles
1995 if (not self.get_param('allow_unplayable_formats')
1996 and re.search(r'#EXT-X-SESSION-KEY:.*?URI="skd://', m3u8_doc)): # Apple FairPlay
1997 return formats, subtitles
1999 def format_url(url):
2000 return url if re.match(r'^https?://', url) else compat_urlparse.urljoin(m3u8_url, url)
2002 if self.get_param('hls_split_discontinuity', False):
2003 def _extract_m3u8_playlist_indices(manifest_url=None, m3u8_doc=None):
2005 if not manifest_url:
2007 m3u8_doc = self._download_webpage(
2008 manifest_url, video_id, fatal=fatal, data=data, headers=headers,
2009 note=False, errnote='Failed to download m3u8 playlist information')
2010 if m3u8_doc is False:
2012 return range(1 + sum(line.startswith('#EXT-X-DISCONTINUITY') for line in m3u8_doc.splitlines()))
2015 def _extract_m3u8_playlist_indices(*args, **kwargs):
2019 # 1. https://tools.ietf.org/html/draft-pantos-http-live-streaming-21
2020 # 2. https://github.com/ytdl-org/youtube-dl/issues/12211
2021 # 3. https://github.com/ytdl-org/youtube-dl/issues/18923
2023 # We should try extracting formats only from master playlists [1, 4.3.4],
2024 # i.e. playlists that describe available qualities. On the other hand
2025 # media playlists [1, 4.3.3] should be returned as is since they contain
2026 # just the media without qualities renditions.
2027 # Fortunately, master playlist can be easily distinguished from media
2028 # playlist based on particular tags availability. As of [1, 4.3.3, 4.3.4]
2029 # master playlist tags MUST NOT appear in a media playlist and vice versa.
2030 # As of [1, 4.3.3.1] #EXT-X-TARGETDURATION tag is REQUIRED for every
2031 # media playlist and MUST NOT appear in master playlist thus we can
2032 # clearly detect media playlist with this criterion.
2034 if '#EXT-X-TARGETDURATION' in m3u8_doc: # media playlist, return as is
2036 'format_id': '-'.join(map(str, filter(None, [m3u8_id, idx]))),
2037 'format_index': idx,
2040 'protocol': entry_protocol,
2041 'preference': preference,
2043 } for idx in _extract_m3u8_playlist_indices(m3u8_doc=m3u8_doc)]
2045 return formats, subtitles
2048 last_stream_inf = {}
2050 def extract_media(x_media_line):
2051 media = parse_m3u8_attributes(x_media_line)
2052 # As per [1, 4.3.4.1] TYPE, GROUP-ID and NAME are REQUIRED
2053 media_type, group_id, name = media.get('TYPE'), media.get('GROUP-ID'), media.get('NAME')
2054 if not (media_type and group_id and name):
2056 groups.setdefault(group_id, []).append(media)
2057 # <https://tools.ietf.org/html/rfc8216#section-4.3.4.1>
2058 if media_type == 'SUBTITLES':
2059 # According to RFC 8216 §4.3.4.2.1, URI is REQUIRED in the
2060 # EXT-X-MEDIA tag if the media type is SUBTITLES.
2061 # However, lack of URI has been spotted in the wild.
2062 # e.g. NebulaIE; see https://github.com/yt-dlp/yt-dlp/issues/339
2063 if not media.get('URI'):
2065 url = format_url(media['URI'])
2068 'ext': determine_ext(url),
2070 if sub_info['ext'] == 'm3u8':
2071 # Per RFC 8216 §3.1, the only possible subtitle format m3u8
2072 # files may contain is WebVTT:
2073 # <https://tools.ietf.org/html/rfc8216#section-3.1>
2074 sub_info['ext'] = 'vtt'
2075 sub_info['protocol'] = 'm3u8_native'
2076 lang = media.get('LANGUAGE') or 'und'
2077 subtitles.setdefault(lang, []).append(sub_info)
2078 if media_type not in ('VIDEO', 'AUDIO'):
2080 media_url = media.get('URI')
2082 manifest_url = format_url(media_url)
2084 'format_id': '-'.join(map(str, filter(None, (m3u8_id, group_id, name, idx)))),
2085 'format_note': name,
2086 'format_index': idx,
2087 'url': manifest_url,
2088 'manifest_url': m3u8_url,
2089 'language': media.get('LANGUAGE'),
2091 'protocol': entry_protocol,
2092 'preference': preference,
2094 'vcodec': 'none' if media_type == 'AUDIO' else None,
2095 } for idx in _extract_m3u8_playlist_indices(manifest_url))
2097 def build_stream_name():
2098 # Despite specification does not mention NAME attribute for
2099 # EXT-X-STREAM-INF tag it still sometimes may be present (see [1]
2100 # or vidio test in TestInfoExtractor.test_parse_m3u8_formats)
2101 # 1. http://www.vidio.com/watch/165683-dj_ambred-booyah-live-2015
2102 stream_name = last_stream_inf.get('NAME')
2105 # If there is no NAME in EXT-X-STREAM-INF it will be obtained
2106 # from corresponding rendition group
2107 stream_group_id = last_stream_inf.get('VIDEO')
2108 if not stream_group_id:
2110 stream_group = groups.get(stream_group_id)
2111 if not stream_group:
2112 return stream_group_id
2113 rendition = stream_group[0]
2114 return rendition.get('NAME') or stream_group_id
2116 # parse EXT-X-MEDIA tags before EXT-X-STREAM-INF in order to have the
2117 # chance to detect video only formats when EXT-X-STREAM-INF tags
2118 # precede EXT-X-MEDIA tags in HLS manifest such as [3].
2119 for line in m3u8_doc.splitlines():
2120 if line.startswith('#EXT-X-MEDIA:'):
2123 for line in m3u8_doc.splitlines():
2124 if line.startswith('#EXT-X-STREAM-INF:'):
2125 last_stream_inf = parse_m3u8_attributes(line)
2126 elif line.startswith('#') or not line.strip():
2129 tbr = float_or_none(
2130 last_stream_inf.get('AVERAGE-BANDWIDTH')
2131 or last_stream_inf.get('BANDWIDTH'), scale=1000)
2132 manifest_url = format_url(line.strip())
2134 for idx in _extract_m3u8_playlist_indices(manifest_url):
2135 format_id = [m3u8_id, None, idx]
2136 # Bandwidth of live streams may differ over time thus making
2137 # format_id unpredictable. So it's better to keep provided
2140 stream_name = build_stream_name()
2141 format_id[1] = stream_name if stream_name else '%d' % (tbr if tbr else len(formats))
2143 'format_id': '-'.join(map(str, filter(None, format_id))),
2144 'format_index': idx,
2145 'url': manifest_url,
2146 'manifest_url': m3u8_url,
2149 'fps': float_or_none(last_stream_inf.get('FRAME-RATE')),
2150 'protocol': entry_protocol,
2151 'preference': preference,
2154 resolution = last_stream_inf.get('RESOLUTION')
2156 mobj = re.search(r'(?P<width>\d+)[xX](?P<height>\d+)', resolution)
2158 f['width'] = int(mobj.group('width'))
2159 f['height'] = int(mobj.group('height'))
2160 # Unified Streaming Platform
2162 r'audio.*?(?:%3D|=)(\d+)(?:-video.*?(?:%3D|=)(\d+))?', f['url'])
2164 abr, vbr = mobj.groups()
2165 abr, vbr = float_or_none(abr, 1000), float_or_none(vbr, 1000)
2170 codecs = parse_codecs(last_stream_inf.get('CODECS'))
2172 audio_group_id = last_stream_inf.get('AUDIO')
2173 # As per [1, 4.3.4.1.1] any EXT-X-STREAM-INF tag which
2174 # references a rendition group MUST have a CODECS attribute.
2175 # However, this is not always respected, for example, [2]
2176 # contains EXT-X-STREAM-INF tag which references AUDIO
2177 # rendition group but does not have CODECS and despite
2178 # referencing an audio group it represents a complete
2179 # (with audio and video) format. So, for such cases we will
2180 # ignore references to rendition groups and treat them
2181 # as complete formats.
2182 if audio_group_id and codecs and f.get('vcodec') != 'none':
2183 audio_group = groups.get(audio_group_id)
2184 if audio_group and audio_group[0].get('URI'):
2185 # TODO: update acodec for audio only formats with
2187 f['acodec'] = 'none'
2188 if not f.get('ext'):
2189 f['ext'] = 'm4a' if f.get('vcodec') == 'none' else 'mp4'
2193 progressive_uri = last_stream_inf.get('PROGRESSIVE-URI')
2196 del http_f['manifest_url']
2198 'format_id': f['format_id'].replace('hls-', 'http-'),
2200 'url': progressive_uri,
2202 formats.append(http_f)
2204 last_stream_inf = {}
2205 return formats, subtitles
2208 def _xpath_ns(path, namespace=None):
2212 for c in path.split('/'):
2213 if not c or c == '.':
2216 out.append('{%s}%s' % (namespace, c))
2217 return '/'.join(out)
2219 def _extract_smil_formats_and_subtitles(self, smil_url, video_id, fatal=True, f4m_params=None, transform_source=None):
2220 smil = self._download_smil(smil_url, video_id, fatal=fatal, transform_source=transform_source)
2226 namespace = self._parse_smil_namespace(smil)
2228 fmts = self._parse_smil_formats(
2229 smil, smil_url, video_id, namespace=namespace, f4m_params=f4m_params)
2230 subs = self._parse_smil_subtitles(
2231 smil, namespace=namespace)
2235 def _extract_smil_formats(self, *args, **kwargs):
2236 fmts, subs = self._extract_smil_formats_and_subtitles(*args, **kwargs)
2238 self.report_warning(bug_reports_message(
2239 "Ignoring subtitle tracks found in the SMIL manifest; "
2240 "if any subtitle tracks are missing,"
2244 def _extract_smil_info(self, smil_url, video_id, fatal=True, f4m_params=None):
2245 smil = self._download_smil(smil_url, video_id, fatal=fatal)
2248 return self._parse_smil(smil, smil_url, video_id, f4m_params=f4m_params)
2250 def _download_smil(self, smil_url, video_id, fatal=True, transform_source=None):
2251 return self._download_xml(
2252 smil_url, video_id, 'Downloading SMIL file',
2253 'Unable to download SMIL file', fatal=fatal, transform_source=transform_source)
2255 def _parse_smil(self, smil, smil_url, video_id, f4m_params=None):
2256 namespace = self._parse_smil_namespace(smil)
2258 formats = self._parse_smil_formats(
2259 smil, smil_url, video_id, namespace=namespace, f4m_params=f4m_params)
2260 subtitles = self._parse_smil_subtitles(smil, namespace=namespace)
2262 video_id = os.path.splitext(url_basename(smil_url))[0]
2266 for meta in smil.findall(self._xpath_ns('./head/meta', namespace)):
2267 name = meta.attrib.get('name')
2268 content = meta.attrib.get('content')
2269 if not name or not content:
2271 if not title and name == 'title':
2273 elif not description and name in ('description', 'abstract'):
2274 description = content
2275 elif not upload_date and name == 'date':
2276 upload_date = unified_strdate(content)
2279 'id': image.get('type'),
2280 'url': image.get('src'),
2281 'width': int_or_none(image.get('width')),
2282 'height': int_or_none(image.get('height')),
2283 } for image in smil.findall(self._xpath_ns('.//image', namespace)) if image.get('src')]
2287 'title': title or video_id,
2288 'description': description,
2289 'upload_date': upload_date,
2290 'thumbnails': thumbnails,
2292 'subtitles': subtitles,
2295 def _parse_smil_namespace(self, smil):
2296 return self._search_regex(
2297 r'(?i)^{([^}]+)?}smil$', smil.tag, 'namespace', default=None)
2299 def _parse_smil_formats(self, smil, smil_url, video_id, namespace=None, f4m_params=None, transform_rtmp_url=None):
2301 for meta in smil.findall(self._xpath_ns('./head/meta', namespace)):
2302 b = meta.get('base') or meta.get('httpBase')
2313 media = smil.findall(self._xpath_ns('.//video', namespace)) + smil.findall(self._xpath_ns('.//audio', namespace))
2314 for medium in media:
2315 src = medium.get('src')
2316 if not src or src in srcs:
2320 bitrate = float_or_none(medium.get('system-bitrate') or medium.get('systemBitrate'), 1000)
2321 filesize = int_or_none(medium.get('size') or medium.get('fileSize'))
2322 width = int_or_none(medium.get('width'))
2323 height = int_or_none(medium.get('height'))
2324 proto = medium.get('proto')
2325 ext = medium.get('ext')
2326 src_ext = determine_ext(src)
2327 streamer = medium.get('streamer') or base
2329 if proto == 'rtmp' or streamer.startswith('rtmp'):
2335 'format_id': 'rtmp-%d' % (rtmp_count if bitrate is None else bitrate),
2337 'filesize': filesize,
2341 if transform_rtmp_url:
2342 streamer, src = transform_rtmp_url(streamer, src)
2343 formats[-1].update({
2349 src_url = src if src.startswith('http') else compat_urlparse.urljoin(base, src)
2350 src_url = src_url.strip()
2352 if proto == 'm3u8' or src_ext == 'm3u8':
2353 m3u8_formats = self._extract_m3u8_formats(
2354 src_url, video_id, ext or 'mp4', m3u8_id='hls', fatal=False)
2355 if len(m3u8_formats) == 1:
2357 m3u8_formats[0].update({
2358 'format_id': 'hls-%d' % (m3u8_count if bitrate is None else bitrate),
2363 formats.extend(m3u8_formats)
2364 elif src_ext == 'f4m':
2369 'plugin': 'flowplayer-3.2.0.1',
2371 f4m_url += '&' if '?' in f4m_url else '?'
2372 f4m_url += compat_urllib_parse_urlencode(f4m_params)
2373 formats.extend(self._extract_f4m_formats(f4m_url, video_id, f4m_id='hds', fatal=False))
2374 elif src_ext == 'mpd':
2375 formats.extend(self._extract_mpd_formats(
2376 src_url, video_id, mpd_id='dash', fatal=False))
2377 elif re.search(r'\.ism/[Mm]anifest', src_url):
2378 formats.extend(self._extract_ism_formats(
2379 src_url, video_id, ism_id='mss', fatal=False))
2380 elif src_url.startswith('http') and self._is_valid_url(src, video_id):
2384 'ext': ext or src_ext or 'flv',
2385 'format_id': 'http-%d' % (bitrate or http_count),
2387 'filesize': filesize,
2394 def _parse_smil_subtitles(self, smil, namespace=None, subtitles_lang='en'):
2397 for num, textstream in enumerate(smil.findall(self._xpath_ns('.//textstream', namespace))):
2398 src = textstream.get('src')
2399 if not src or src in urls:
2402 ext = textstream.get('ext') or mimetype2ext(textstream.get('type')) or determine_ext(src)
2403 lang = textstream.get('systemLanguage') or textstream.get('systemLanguageName') or textstream.get('lang') or subtitles_lang
2404 subtitles.setdefault(lang, []).append({
2410 def _extract_xspf_playlist(self, xspf_url, playlist_id, fatal=True):
2411 xspf = self._download_xml(
2412 xspf_url, playlist_id, 'Downloading xpsf playlist',
2413 'Unable to download xspf manifest', fatal=fatal)
2416 return self._parse_xspf(
2417 xspf, playlist_id, xspf_url=xspf_url,
2418 xspf_base_url=base_url(xspf_url))
2420 def _parse_xspf(self, xspf_doc, playlist_id, xspf_url=None, xspf_base_url=None):
2422 'xspf': 'http://xspf.org/ns/0/',
2423 's1': 'http://static.streamone.nl/player/ns/0',
2427 for track in xspf_doc.findall(xpath_with_ns('./xspf:trackList/xspf:track', NS_MAP)):
2429 track, xpath_with_ns('./xspf:title', NS_MAP), 'title', default=playlist_id)
2430 description = xpath_text(
2431 track, xpath_with_ns('./xspf:annotation', NS_MAP), 'description')
2432 thumbnail = xpath_text(
2433 track, xpath_with_ns('./xspf:image', NS_MAP), 'thumbnail')
2434 duration = float_or_none(
2435 xpath_text(track, xpath_with_ns('./xspf:duration', NS_MAP), 'duration'), 1000)
2438 for location in track.findall(xpath_with_ns('./xspf:location', NS_MAP)):
2439 format_url = urljoin(xspf_base_url, location.text)
2444 'manifest_url': xspf_url,
2445 'format_id': location.get(xpath_with_ns('s1:label', NS_MAP)),
2446 'width': int_or_none(location.get(xpath_with_ns('s1:width', NS_MAP))),
2447 'height': int_or_none(location.get(xpath_with_ns('s1:height', NS_MAP))),
2449 self._sort_formats(formats)
2454 'description': description,
2455 'thumbnail': thumbnail,
2456 'duration': duration,
2461 def _extract_mpd_formats(self, *args, **kwargs):
2462 fmts, subs = self._extract_mpd_formats_and_subtitles(*args, **kwargs)
2464 self.report_warning(bug_reports_message(
2465 "Ignoring subtitle tracks found in the DASH manifest; "
2466 "if any subtitle tracks are missing,"
2470 def _extract_mpd_formats_and_subtitles(
2471 self, mpd_url, video_id, mpd_id=None, note=None, errnote=None,
2472 fatal=True, data=None, headers={}, query={}):
2473 res = self._download_xml_handle(
2475 note='Downloading MPD manifest' if note is None else note,
2476 errnote='Failed to download MPD manifest' if errnote is None else errnote,
2477 fatal=fatal, data=data, headers=headers, query=query)
2483 mpd_base_url = base_url(urlh.geturl())
2485 return self._parse_mpd_formats_and_subtitles(
2486 mpd_doc, mpd_id, mpd_base_url, mpd_url)
2488 def _parse_mpd_formats(self, *args, **kwargs):
2489 fmts, subs = self._parse_mpd_formats_and_subtitles(*args, **kwargs)
2491 self.report_warning(bug_reports_message(
2492 "Ignoring subtitle tracks found in the DASH manifest; "
2493 "if any subtitle tracks are missing,"
2497 def _parse_mpd_formats_and_subtitles(
2498 self, mpd_doc, mpd_id=None, mpd_base_url='', mpd_url=None):
2500 Parse formats
from MPD manifest
.
2502 1. MPEG
-DASH Standard
, ISO
/IEC
23009-1:2014(E
),
2503 http
://standards
.iso
.org
/ittf
/PubliclyAvailableStandards
/c065274_ISO_IEC_23009
-1_2014.zip
2504 2. https
://en
.wikipedia
.org
/wiki
/Dynamic_Adaptive_Streaming_over_HTTP
2506 if not self.get_param('dynamic_mpd', True):
2507 if mpd_doc.get('type') == 'dynamic':
2510 namespace = self._search_regex(r'(?i)^{([^}]+)?}MPD$', mpd_doc.tag, 'namespace', default=None)
2513 return self._xpath_ns(path, namespace)
2515 def is_drm_protected(element):
2516 return element.find(_add_ns('ContentProtection')) is not None
2518 def extract_multisegment_info(element, ms_parent_info):
2519 ms_info = ms_parent_info.copy()
2521 # As per [1, 5.3.9.2.2] SegmentList and SegmentTemplate share some
2522 # common attributes and elements. We will only extract relevant
2524 def extract_common(source):
2525 segment_timeline = source.find(_add_ns('SegmentTimeline'))
2526 if segment_timeline is not None:
2527 s_e = segment_timeline.findall(_add_ns('S'))
2529 ms_info['total_number'] = 0
2532 r = int(s.get('r', 0))
2533 ms_info['total_number'] += 1 + r
2534 ms_info['s'].append({
2535 't': int(s.get('t', 0)),
2536 # @d is mandatory (see [1, 5.3.9.6.2, Table 17, page 60])
2537 'd': int(s.attrib['d']),
2540 start_number = source.get('startNumber')
2542 ms_info['start_number'] = int(start_number)
2543 timescale = source.get('timescale')
2545 ms_info['timescale'] = int(timescale)
2546 segment_duration = source.get('duration')
2547 if segment_duration:
2548 ms_info['segment_duration'] = float(segment_duration)
2550 def extract_Initialization(source):
2551 initialization = source.find(_add_ns('Initialization'))
2552 if initialization is not None:
2553 ms_info['initialization_url'] = initialization.attrib['sourceURL']
2555 segment_list = element.find(_add_ns('SegmentList'))
2556 if segment_list is not None:
2557 extract_common(segment_list)
2558 extract_Initialization(segment_list)
2559 segment_urls_e = segment_list.findall(_add_ns('SegmentURL'))
2561 ms_info['segment_urls'] = [segment.attrib['media'] for segment in segment_urls_e]
2563 segment_template = element.find(_add_ns('SegmentTemplate'))
2564 if segment_template is not None:
2565 extract_common(segment_template)
2566 media = segment_template.get('media')
2568 ms_info['media'] = media
2569 initialization = segment_template.get('initialization')
2571 ms_info['initialization'] = initialization
2573 extract_Initialization(segment_template)
2576 skip_unplayable = not self.get_param('allow_unplayable_formats')
2578 mpd_duration = parse_duration(mpd_doc.get('mediaPresentationDuration'))
2581 for period in mpd_doc.findall(_add_ns('Period')):
2582 period_duration = parse_duration(period.get('duration')) or mpd_duration
2583 period_ms_info = extract_multisegment_info(period, {
2587 for adaptation_set in period.findall(_add_ns('AdaptationSet')):
2588 if skip_unplayable and is_drm_protected(adaptation_set):
2590 adaption_set_ms_info = extract_multisegment_info(adaptation_set, period_ms_info)
2591 for representation in adaptation_set.findall(_add_ns('Representation')):
2592 if skip_unplayable and is_drm_protected(representation):
2594 representation_attrib = adaptation_set.attrib.copy()
2595 representation_attrib.update(representation.attrib)
2596 # According to [1, 5.3.7.2, Table 9, page 41], @mimeType is mandatory
2597 mime_type = representation_attrib['mimeType']
2598 content_type = representation_attrib.get('contentType', mime_type.split('/')[0])
2600 codecs = representation_attrib.get('codecs', '')
2601 if content_type not in ('video', 'audio', 'text'):
2602 if mime_type == 'image/jpeg':
2603 content_type = mime_type
2604 elif codecs.split('.')[0] == 'stpp':
2605 content_type = 'text'
2607 self.report_warning('Unknown MIME type %s in DASH manifest' % mime_type)
2611 for element in (representation, adaptation_set, period, mpd_doc):
2612 base_url_e = element.find(_add_ns('BaseURL'))
2613 if base_url_e is not None:
2614 base_url = base_url_e.text + base_url
2615 if re.match(r'^https?://', base_url):
2617 if mpd_base_url and not re.match(r'^https?://', base_url):
2618 if not mpd_base_url.endswith('/') and not base_url.startswith('/'):
2620 base_url = mpd_base_url + base_url
2621 representation_id = representation_attrib.get('id')
2622 lang = representation_attrib.get('lang')
2623 url_el = representation.find(_add_ns('BaseURL'))
2624 filesize = int_or_none(url_el.attrib.get('{http://youtube.com/yt/2012/10/10}contentLength') if url_el is not None else None)
2625 bandwidth = int_or_none(representation_attrib.get('bandwidth'))
2626 if representation_id is not None:
2627 format_id = representation_id
2629 format_id = content_type
2631 format_id = mpd_id + '-' + format_id
2632 if content_type in ('video', 'audio'):
2634 'format_id': format_id,
2635 'manifest_url': mpd_url,
2636 'ext': mimetype2ext(mime_type),
2637 'width': int_or_none(representation_attrib.get('width')),
2638 'height': int_or_none(representation_attrib.get('height')),
2639 'tbr': float_or_none(bandwidth, 1000),
2640 'asr': int_or_none(representation_attrib.get('audioSamplingRate')),
2641 'fps': int_or_none(representation_attrib.get('frameRate')),
2642 'language': lang if lang not in ('mul', 'und', 'zxx', 'mis') else None,
2643 'format_note': 'DASH %s' % content_type,
2644 'filesize': filesize,
2645 'container': mimetype2ext(mime_type) + '_dash',
2647 f.update(parse_codecs(codecs))
2648 elif content_type == 'text':
2650 'ext': mimetype2ext(mime_type),
2651 'manifest_url': mpd_url,
2652 'filesize': filesize,
2654 elif content_type == 'image/jpeg':
2655 # See test case in VikiIE
2656 # https://www.viki.com/videos/1175236v-choosing-spouse-by-lottery-episode-1
2658 'format_id': format_id,
2660 'manifest_url': mpd_url,
2661 'format_note': 'DASH storyboards (jpeg)',
2665 representation_ms_info = extract_multisegment_info(representation, adaption_set_ms_info)
2667 def prepare_template(template_name, identifiers):
2668 tmpl = representation_ms_info[template_name]
2669 # First of, % characters outside $...$ templates
2670 # must be escaped by doubling for proper processing
2671 # by % operator string formatting used further (see
2672 # https://github.com/ytdl-org/youtube-dl/issues/16867).
2678 in_template = not in_template
2679 elif c == '%' and not in_template:
2681 # Next, $...$ templates are translated to their
2682 # %(...) counterparts to be used with % operator
2683 if representation_id is not None:
2684 t = t.replace('$RepresentationID$', representation_id)
2685 t = re.sub(r'\$(%s)\$' % '|'.join(identifiers), r'%(\1)d', t)
2686 t = re.sub(r'\$(%s)%%([^$]+)\$' % '|'.join(identifiers), r'%(\1)\2', t)
2687 t.replace('$$', '$')
2690 # @initialization is a regular template like @media one
2691 # so it should be handled just the same way (see
2692 # https://github.com/ytdl-org/youtube-dl/issues/11605)
2693 if 'initialization' in representation_ms_info:
2694 initialization_template = prepare_template(
2696 # As per [1, 5.3.9.4.2, Table 15, page 54] $Number$ and
2697 # $Time$ shall not be included for @initialization thus
2698 # only $Bandwidth$ remains
2700 representation_ms_info['initialization_url'] = initialization_template % {
2701 'Bandwidth': bandwidth,
2704 def location_key(location):
2705 return 'url' if re.match(r'^https?://', location) else 'path'
2707 if 'segment_urls' not in representation_ms_info and 'media' in representation_ms_info:
2709 media_template = prepare_template('media', ('Number', 'Bandwidth', 'Time'))
2710 media_location_key = location_key(media_template)
2712 # As per [1, 5.3.9.4.4, Table 16, page 55] $Number$ and $Time$
2713 # can't be used at the same time
2714 if '%(Number' in media_template and 's' not in representation_ms_info:
2715 segment_duration = None
2716 if 'total_number' not in representation_ms_info and 'segment_duration' in representation_ms_info:
2717 segment_duration = float_or_none(representation_ms_info['segment_duration'], representation_ms_info['timescale'])
2718 representation_ms_info['total_number'] = int(math.ceil(float(period_duration) / segment_duration))
2719 representation_ms_info['fragments'] = [{
2720 media_location_key: media_template % {
2721 'Number': segment_number,
2722 'Bandwidth': bandwidth,
2724 'duration': segment_duration,
2725 } for segment_number in range(
2726 representation_ms_info['start_number'],
2727 representation_ms_info['total_number'] + representation_ms_info['start_number'])]
2729 # $Number*$ or $Time$ in media template with S list available
2730 # Example $Number*$: http://www.svtplay.se/klipp/9023742/stopptid-om-bjorn-borg
2731 # Example $Time$: https://play.arkena.com/embed/avp/v2/player/media/b41dda37-d8e7-4d3f-b1b5-9a9db578bdfe/1/129411
2732 representation_ms_info['fragments'] = []
2735 segment_number = representation_ms_info['start_number']
2737 def add_segment_url():
2738 segment_url = media_template % {
2739 'Time': segment_time,
2740 'Bandwidth': bandwidth,
2741 'Number': segment_number,
2743 representation_ms_info['fragments'].append({
2744 media_location_key: segment_url,
2745 'duration': float_or_none(segment_d, representation_ms_info['timescale']),
2748 for num, s in enumerate(representation_ms_info['s']):
2749 segment_time = s.get('t') or segment_time
2753 for r in range(s.get('r', 0)):
2754 segment_time += segment_d
2757 segment_time += segment_d
2758 elif 'segment_urls' in representation_ms_info and 's' in representation_ms_info:
2760 # Example: https://www.youtube.com/watch?v=iXZV5uAYMJI
2761 # or any YouTube dashsegments video
2764 timescale = representation_ms_info['timescale']
2765 for s in representation_ms_info['s']:
2766 duration = float_or_none(s['d'], timescale)
2767 for r in range(s.get('r', 0) + 1):
2768 segment_uri = representation_ms_info['segment_urls'][segment_index]
2770 location_key(segment_uri): segment_uri,
2771 'duration': duration,
2774 representation_ms_info['fragments'] = fragments
2775 elif 'segment_urls' in representation_ms_info:
2776 # Segment URLs with no SegmentTimeline
2777 # Example: https://www.seznam.cz/zpravy/clanek/cesko-zasahne-vitr-o-sile-vichrice-muze-byt-i-zivotu-nebezpecny-39091
2778 # https://github.com/ytdl-org/youtube-dl/pull/14844
2780 segment_duration = float_or_none(
2781 representation_ms_info['segment_duration'],
2782 representation_ms_info['timescale']) if 'segment_duration' in representation_ms_info else None
2783 for segment_url in representation_ms_info['segment_urls']:
2785 location_key(segment_url): segment_url,
2787 if segment_duration:
2788 fragment['duration'] = segment_duration
2789 fragments.append(fragment)
2790 representation_ms_info['fragments'] = fragments
2791 # If there is a fragments key available then we correctly recognized fragmented media.
2792 # Otherwise we will assume unfragmented media with direct access. Technically, such
2793 # assumption is not necessarily correct since we may simply have no support for
2794 # some forms of fragmented media renditions yet, but for now we'll use this fallback.
2795 if 'fragments' in representation_ms_info:
2797 # NB: mpd_url may be empty when MPD manifest is parsed from a string
2798 'url': mpd_url or base_url,
2799 'fragment_base_url': base_url,
2801 'protocol': 'http_dash_segments' if mime_type != 'image/jpeg' else 'mhtml',
2803 if 'initialization_url' in representation_ms_info:
2804 initialization_url = representation_ms_info['initialization_url']
2805 if not f.get('url'):
2806 f['url'] = initialization_url
2807 f['fragments'].append({location_key(initialization_url): initialization_url})
2808 f['fragments'].extend(representation_ms_info['fragments'])
2810 # Assuming direct URL to unfragmented media.
2812 if content_type in ('video', 'audio') or mime_type == 'image/jpeg':
2814 elif content_type == 'text':
2815 subtitles.setdefault(lang or 'und', []).append(f)
2817 return formats, subtitles
2819 def _extract_ism_formats(self, *args, **kwargs):
2820 fmts, subs = self._extract_ism_formats_and_subtitles(*args, **kwargs)
2822 self.report_warning(bug_reports_message(
2823 "Ignoring subtitle tracks found in the ISM manifest; "
2824 "if any subtitle tracks are missing,"
2828 def _extract_ism_formats_and_subtitles(self, ism_url, video_id, ism_id=None, note=None, errnote=None, fatal=True, data=None, headers={}, query={}):
2829 res = self._download_xml_handle(
2831 note='Downloading ISM manifest' if note is None else note,
2832 errnote='Failed to download ISM manifest' if errnote is None else errnote,
2833 fatal=fatal, data=data, headers=headers, query=query)
2840 return self._parse_ism_formats_and_subtitles(ism_doc, urlh.geturl(), ism_id)
2842 def _parse_ism_formats_and_subtitles(self, ism_doc, ism_url, ism_id=None):
2844 Parse formats
from ISM manifest
.
2846 1. [MS
-SSTR
]: Smooth Streaming Protocol
,
2847 https
://msdn
.microsoft
.com
/en
-us
/library
/ff469518
.aspx
2849 if ism_doc.get('IsLive') == 'TRUE':
2851 if (not self.get_param('allow_unplayable_formats')
2852 and ism_doc.find('Protection') is not None):
2855 duration = int(ism_doc.attrib['Duration'])
2856 timescale = int_or_none(ism_doc.get('TimeScale')) or 10000000
2860 for stream in ism_doc.findall('StreamIndex'):
2861 stream_type = stream.get('Type')
2862 if stream_type not in ('video', 'audio', 'text'):
2864 url_pattern = stream.attrib['Url']
2865 stream_timescale = int_or_none(stream.get('TimeScale')) or timescale
2866 stream_name = stream.get('Name')
2867 stream_language = stream.get('Language', 'und')
2868 for track in stream.findall('QualityLevel'):
2869 fourcc = track.get('FourCC') or ('AACL' if track.get('AudioTag') == '255' else None)
2870 # TODO: add support for WVC1 and WMAP
2871 if fourcc not in ('H264', 'AVC1', 'AACL', 'TTML'):
2872 self.report_warning('%s is not a supported codec' % fourcc)
2874 tbr = int(track.attrib['Bitrate']) // 1000
2875 # [1] does not mention Width and Height attributes. However,
2876 # they're often present while MaxWidth and MaxHeight are
2877 # missing, so should be used as fallbacks
2878 width = int_or_none(track.get('MaxWidth') or track.get('Width'))
2879 height = int_or_none(track.get('MaxHeight') or track.get('Height'))
2880 sampling_rate = int_or_none(track.get('SamplingRate'))
2882 track_url_pattern = re.sub(r'{[Bb]itrate}', track.attrib['Bitrate'], url_pattern)
2883 track_url_pattern = compat_urlparse.urljoin(ism_url, track_url_pattern)
2889 stream_fragments = stream.findall('c')
2890 for stream_fragment_index, stream_fragment in enumerate(stream_fragments):
2891 fragment_ctx['time'] = int_or_none(stream_fragment.get('t')) or fragment_ctx['time']
2892 fragment_repeat = int_or_none(stream_fragment.get('r')) or 1
2893 fragment_ctx['duration'] = int_or_none(stream_fragment.get('d'))
2894 if not fragment_ctx['duration']:
2896 next_fragment_time = int(stream_fragment[stream_fragment_index + 1].attrib['t'])
2898 next_fragment_time = duration
2899 fragment_ctx['duration'] = (next_fragment_time - fragment_ctx['time']) / fragment_repeat
2900 for _ in range(fragment_repeat):
2902 'url': re.sub(r'{start[ _]time}', compat_str(fragment_ctx['time']), track_url_pattern),
2903 'duration': fragment_ctx['duration'] / stream_timescale,
2905 fragment_ctx['time'] += fragment_ctx['duration']
2909 format_id.append(ism_id)
2911 format_id.append(stream_name)
2912 format_id.append(compat_str(tbr))
2914 if stream_type == 'text':
2915 subtitles.setdefault(stream_language, []).append({
2919 'manifest_url': ism_url,
2920 'fragments': fragments,
2921 '_download_params': {
2922 'stream_type': stream_type,
2923 'duration': duration,
2924 'timescale': stream_timescale,
2926 'language': stream_language,
2927 'codec_private_data': track.get('CodecPrivateData'),
2930 elif stream_type in ('video', 'audio'):
2932 'format_id': '-'.join(format_id),
2934 'manifest_url': ism_url,
2935 'ext': 'ismv' if stream_type == 'video' else 'isma',
2939 'asr': sampling_rate,
2940 'vcodec': 'none' if stream_type == 'audio' else fourcc,
2941 'acodec': 'none' if stream_type == 'video' else fourcc,
2943 'fragments': fragments,
2944 '_download_params': {
2945 'stream_type': stream_type,
2946 'duration': duration,
2947 'timescale': stream_timescale,
2948 'width': width or 0,
2949 'height': height or 0,
2951 'language': stream_language,
2952 'codec_private_data': track.get('CodecPrivateData'),
2953 'sampling_rate': sampling_rate,
2954 'channels': int_or_none(track.get('Channels', 2)),
2955 'bits_per_sample': int_or_none(track.get('BitsPerSample', 16)),
2956 'nal_unit_length_field': int_or_none(track.get('NALUnitLengthField', 4)),
2959 return formats, subtitles
2961 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):
2962 def absolute_url(item_url):
2963 return urljoin(base_url, item_url)
2965 def parse_content_type(content_type):
2966 if not content_type:
2968 ctr = re.search(r'(?P<mimetype>[^/]+/[^;]+)(?:;\s*codecs="?(?P<codecs>[^"]+))?', content_type)
2970 mimetype, codecs = ctr.groups()
2971 f = parse_codecs(codecs)
2972 f['ext'] = mimetype2ext(mimetype)
2976 def _media_formats(src, cur_media_type, type_info={}):
2977 full_url = absolute_url(src)
2978 ext = type_info.get('ext') or determine_ext(full_url)
2980 is_plain_url = False
2981 formats = self._extract_m3u8_formats(
2982 full_url, video_id, ext='mp4',
2983 entry_protocol=m3u8_entry_protocol, m3u8_id=m3u8_id,
2984 preference=preference, quality=quality, fatal=False)
2986 is_plain_url = False
2987 formats = self._extract_mpd_formats(
2988 full_url, video_id, mpd_id=mpd_id, fatal=False)
2993 'vcodec': 'none' if cur_media_type == 'audio' else None,
2995 return is_plain_url, formats
2998 # amp-video and amp-audio are very similar to their HTML5 counterparts
2999 # so we wll include them right here (see
3000 # https://www.ampproject.org/docs/reference/components/amp-video)
3001 # For dl8-* tags see https://delight-vr.com/documentation/dl8-video/
3002 _MEDIA_TAG_NAME_RE = r'(?:(?:amp|dl8(?:-live)?)-)?(video|audio)'
3003 media_tags = [(media_tag, media_tag_name, media_type, '')
3004 for media_tag, media_tag_name, media_type
3005 in re.findall(r'(?s)(<(%s)[^>]*/>)' % _MEDIA_TAG_NAME_RE, webpage)]
3006 media_tags.extend(re.findall(
3007 # We only allow video|audio followed by a whitespace or '>'.
3008 # Allowing more characters may end up in significant slow down (see
3009 # https://github.com/ytdl-org/youtube-dl/issues/11979, example URL:
3010 # http://www.porntrex.com/maps/videositemap.xml).
3011 r'(?s)(<(?P<tag>%s)(?:\s+[^>]*)?>)(.*?)</(?P=tag)>' % _MEDIA_TAG_NAME_RE, webpage))
3012 for media_tag, _, media_type, media_content in media_tags:
3017 media_attributes = extract_attributes(media_tag)
3018 src = strip_or_none(media_attributes.get('src'))
3020 _, formats = _media_formats(src, media_type)
3021 media_info['formats'].extend(formats)
3022 media_info['thumbnail'] = absolute_url(media_attributes.get('poster'))
3024 for source_tag in re.findall(r'<source[^>]+>', media_content):
3025 s_attr = extract_attributes(source_tag)
3026 # data-video-src and data-src are non standard but seen
3027 # several times in the wild
3028 src = strip_or_none(dict_get(s_attr, ('src', 'data-video-src', 'data-src')))
3031 f = parse_content_type(s_attr.get('type'))
3032 is_plain_url, formats = _media_formats(src, media_type, f)
3034 # width, height, res, label and title attributes are
3035 # all not standard but seen several times in the wild
3038 for lbl in ('label', 'title')
3039 if str_or_none(s_attr.get(lbl))
3041 width = int_or_none(s_attr.get('width'))
3042 height = (int_or_none(s_attr.get('height'))
3043 or int_or_none(s_attr.get('res')))
3044 if not width or not height:
3046 resolution = parse_resolution(lbl)
3049 width = width or resolution.get('width')
3050 height = height or resolution.get('height')
3052 tbr = parse_bitrate(lbl)
3061 'format_id': s_attr.get('label') or s_attr.get('title'),
3063 f.update(formats[0])
3064 media_info['formats'].append(f)
3066 media_info['formats'].extend(formats)
3067 for track_tag in re.findall(r'<track[^>]+>', media_content):
3068 track_attributes = extract_attributes(track_tag)
3069 kind = track_attributes.get('kind')
3070 if not kind or kind in ('subtitles', 'captions'):
3071 src = strip_or_none(track_attributes.get('src'))
3074 lang = track_attributes.get('srclang') or track_attributes.get('lang') or track_attributes.get('label')
3075 media_info['subtitles'].setdefault(lang, []).append({
3076 'url': absolute_url(src),
3078 for f in media_info['formats']:
3079 f.setdefault('http_headers', {})['Referer'] = base_url
3080 if media_info['formats'] or media_info['subtitles']:
3081 entries.append(media_info)
3084 def _extract_akamai_formats(self, *args, **kwargs):
3085 fmts, subs = self._extract_akamai_formats_and_subtitles(*args, **kwargs)
3087 self.report_warning(bug_reports_message(
3088 "Ignoring subtitle tracks found in the manifests; "
3089 "if any subtitle tracks are missing,"
3093 def _extract_akamai_formats_and_subtitles(self, manifest_url, video_id, hosts={}):
3094 signed = 'hdnea=' in manifest_url
3096 # https://learn.akamai.com/en-us/webhelp/media-services-on-demand/stream-packaging-user-guide/GUID-BE6C0F73-1E06-483B-B0EA-57984B91B7F9.html
3097 manifest_url = re.sub(
3098 r'(?:b=[\d,-]+|(?:__a__|attributes)=off|__b__=\d+)&?',
3099 '', manifest_url).strip('?')
3104 hdcore_sign = 'hdcore=3.7.0'
3105 f4m_url = re.sub(r'(https?://[^/]+)/i/', r'\1/z/', manifest_url).replace('/master.m3u8', '/manifest.f4m')
3106 hds_host = hosts.get('hds')
3108 f4m_url = re.sub(r'(https?://)[^/]+', r'\1' + hds_host, f4m_url)
3109 if 'hdcore=' not in f4m_url:
3110 f4m_url += ('&' if '?' in f4m_url else '?') + hdcore_sign
3111 f4m_formats = self._extract_f4m_formats(
3112 f4m_url, video_id, f4m_id='hds', fatal=False)
3113 for entry in f4m_formats:
3114 entry.update({'extra_param_to_segment_url': hdcore_sign})
3115 formats.extend(f4m_formats)
3117 m3u8_url = re.sub(r'(https?://[^/]+)/z/', r'\1/i/', manifest_url).replace('/manifest.f4m', '/master.m3u8')
3118 hls_host = hosts.get('hls')
3120 m3u8_url = re.sub(r'(https?://)[^/]+', r'\1' + hls_host, m3u8_url)
3121 m3u8_formats, m3u8_subtitles = self._extract_m3u8_formats_and_subtitles(
3122 m3u8_url, video_id, 'mp4', 'm3u8_native',
3123 m3u8_id='hls', fatal=False)
3124 formats.extend(m3u8_formats)
3125 subtitles = self._merge_subtitles(subtitles, m3u8_subtitles)
3127 http_host = hosts.get('http')
3128 if http_host and m3u8_formats and not signed:
3129 REPL_REGEX = r'https?://[^/]+/i/([^,]+),([^/]+),([^/]+)\.csmil/.+'
3130 qualities = re.match(REPL_REGEX, m3u8_url).group(2).split(',')
3131 qualities_length = len(qualities)
3132 if len(m3u8_formats) in (qualities_length, qualities_length + 1):
3134 for f in m3u8_formats:
3135 if f['vcodec'] != 'none':
3136 for protocol in ('http', 'https'):
3138 del http_f['manifest_url']
3140 REPL_REGEX, protocol + r'://%s/\g<1>%s\3' % (http_host, qualities[i]), f['url'])
3142 'format_id': http_f['format_id'].replace('hls-', protocol + '-'),
3144 'protocol': protocol,
3146 formats.append(http_f)
3149 return formats, subtitles
3151 def _extract_wowza_formats(self, url, video_id, m3u8_entry_protocol='m3u8_native', skip_protocols=[]):
3152 query = compat_urlparse.urlparse(url).query
3153 url = re.sub(r'/(?:manifest|playlist|jwplayer)\.(?:m3u8|f4m|mpd|smil)', '', url)
3155 r'(?:(?:http|rtmp|rtsp)(?P<s>s)?:)?(?P<url>//[^?]+)', url)
3156 url_base = mobj.group('url')
3157 http_base_url = '%s%s:%s' % ('http', mobj.group('s') or '', url_base)
3160 def manifest_url(manifest):
3161 m_url = '%s/%s' % (http_base_url, manifest)
3163 m_url += '?%s' % query
3166 if 'm3u8' not in skip_protocols:
3167 formats.extend(self._extract_m3u8_formats(
3168 manifest_url('playlist.m3u8'), video_id, 'mp4',
3169 m3u8_entry_protocol, m3u8_id='hls', fatal=False))
3170 if 'f4m' not in skip_protocols:
3171 formats.extend(self._extract_f4m_formats(
3172 manifest_url('manifest.f4m'),
3173 video_id, f4m_id='hds', fatal=False))
3174 if 'dash' not in skip_protocols:
3175 formats.extend(self._extract_mpd_formats(
3176 manifest_url('manifest.mpd'),
3177 video_id, mpd_id='dash', fatal=False))
3178 if re.search(r'(?:/smil:|\.smil)', url_base):
3179 if 'smil' not in skip_protocols:
3180 rtmp_formats = self._extract_smil_formats(
3181 manifest_url('jwplayer.smil'),
3182 video_id, fatal=False)
3183 for rtmp_format in rtmp_formats:
3184 rtsp_format = rtmp_format.copy()
3185 rtsp_format['url'] = '%s/%s' % (rtmp_format['url'], rtmp_format['play_path'])
3186 del rtsp_format['play_path']
3187 del rtsp_format['ext']
3188 rtsp_format.update({
3189 'url': rtsp_format['url'].replace('rtmp://', 'rtsp://'),
3190 'format_id': rtmp_format['format_id'].replace('rtmp', 'rtsp'),
3193 formats.extend([rtmp_format, rtsp_format])
3195 for protocol in ('rtmp', 'rtsp'):
3196 if protocol not in skip_protocols:
3198 'url': '%s:%s' % (protocol, url_base),
3199 'format_id': protocol,
3200 'protocol': protocol,
3204 def _find_jwplayer_data(self, webpage, video_id=None, transform_source=js_to_json):
3206 r'(?s)jwplayer\((?P<quote>[\'"])[^\'" ]+(?P=quote)\)(?!</script>).*?\.setup\s*\((?P<options>[^)]+)\)',
3210 jwplayer_data = self._parse_json(mobj.group('options'),
3212 transform_source=transform_source)
3213 except ExtractorError:
3216 if isinstance(jwplayer_data, dict):
3217 return jwplayer_data
3219 def _extract_jwplayer_data(self, webpage, video_id, *args, **kwargs):
3220 jwplayer_data = self._find_jwplayer_data(
3221 webpage, video_id, transform_source=js_to_json)
3222 return self._parse_jwplayer_data(
3223 jwplayer_data, video_id, *args, **kwargs)
3225 def _parse_jwplayer_data(self, jwplayer_data, video_id=None, require_title=True,
3226 m3u8_id=None, mpd_id=None, rtmp_params=None, base_url=None):
3227 # JWPlayer backward compatibility: flattened playlists
3228 # https://github.com/jwplayer/jwplayer/blob/v7.4.3/src/js/api/config.js#L81-L96
3229 if 'playlist' not in jwplayer_data:
3230 jwplayer_data = {'playlist': [jwplayer_data]}
3234 # JWPlayer backward compatibility: single playlist item
3235 # https://github.com/jwplayer/jwplayer/blob/v7.7.0/src/js/playlist/playlist.js#L10
3236 if not isinstance(jwplayer_data['playlist'], list):
3237 jwplayer_data['playlist'] = [jwplayer_data['playlist']]
3239 for video_data in jwplayer_data['playlist']:
3240 # JWPlayer backward compatibility: flattened sources
3241 # https://github.com/jwplayer/jwplayer/blob/v7.4.3/src/js/playlist/item.js#L29-L35
3242 if 'sources' not in video_data:
3243 video_data['sources'] = [video_data]
3245 this_video_id = video_id or video_data['mediaid']
3247 formats = self._parse_jwplayer_formats(
3248 video_data['sources'], video_id=this_video_id, m3u8_id=m3u8_id,
3249 mpd_id=mpd_id, rtmp_params=rtmp_params, base_url=base_url)
3252 tracks = video_data.get('tracks')
3253 if tracks and isinstance(tracks, list):
3254 for track in tracks:
3255 if not isinstance(track, dict):
3257 track_kind = track.get('kind')
3258 if not track_kind or not isinstance(track_kind, compat_str):
3260 if track_kind.lower() not in ('captions', 'subtitles'):
3262 track_url = urljoin(base_url, track.get('file'))
3265 subtitles.setdefault(track.get('label') or 'en', []).append({
3266 'url': self._proto_relative_url(track_url)
3270 'id': this_video_id,
3271 'title': unescapeHTML(video_data['title'] if require_title else video_data.get('title')),
3272 'description': clean_html(video_data.get('description')),
3273 'thumbnail': urljoin(base_url, self._proto_relative_url(video_data.get('image'))),
3274 'timestamp': int_or_none(video_data.get('pubdate')),
3275 'duration': float_or_none(jwplayer_data.get('duration') or video_data.get('duration')),
3276 'subtitles': subtitles,
3278 # https://github.com/jwplayer/jwplayer/blob/master/src/js/utils/validator.js#L32
3279 if len(formats) == 1 and re.search(r'^(?:http|//).*(?:youtube\.com|youtu\.be)/.+', formats[0]['url']):
3281 '_type': 'url_transparent',
3282 'url': formats[0]['url'],
3285 self._sort_formats(formats)
3286 entry['formats'] = formats
3287 entries.append(entry)
3288 if len(entries) == 1:
3291 return self.playlist_result(entries)
3293 def _parse_jwplayer_formats(self, jwplayer_sources_data, video_id=None,
3294 m3u8_id=None, mpd_id=None, rtmp_params=None, base_url=None):
3297 for source in jwplayer_sources_data:
3298 if not isinstance(source, dict):
3300 source_url = urljoin(
3301 base_url, self._proto_relative_url(source.get('file')))
3302 if not source_url or source_url in urls:
3304 urls.append(source_url)
3305 source_type = source.get('type') or ''
3306 ext = mimetype2ext(source_type) or determine_ext(source_url)
3307 if source_type == 'hls' or ext == 'm3u8':
3308 formats.extend(self._extract_m3u8_formats(
3309 source_url, video_id, 'mp4', entry_protocol='m3u8_native',
3310 m3u8_id=m3u8_id, fatal=False))
3311 elif source_type == 'dash' or ext == 'mpd':
3312 formats.extend(self._extract_mpd_formats(
3313 source_url, video_id, mpd_id=mpd_id, fatal=False))
3315 formats.extend(self._extract_smil_formats(
3316 source_url, video_id, fatal=False))
3317 # https://github.com/jwplayer/jwplayer/blob/master/src/js/providers/default.js#L67
3318 elif source_type.startswith('audio') or ext in (
3319 'oga', 'aac', 'mp3', 'mpeg', 'vorbis'):
3326 height = int_or_none(source.get('height'))
3328 # Often no height is provided but there is a label in
3329 # format like "1080p", "720p SD", or 1080.
3330 height = int_or_none(self._search_regex(
3331 r'^(\d{3,4})[pP]?(?:\b|$)', compat_str(source.get('label') or ''),
3332 'height', default=None))
3335 'width': int_or_none(source.get('width')),
3337 'tbr': int_or_none(source.get('bitrate')),
3340 if source_url.startswith('rtmp'):
3341 a_format['ext'] = 'flv'
3342 # See com/longtailvideo/jwplayer/media/RTMPMediaProvider.as
3343 # of jwplayer.flash.swf
3344 rtmp_url_parts = re.split(
3345 r'((?:mp4|mp3|flv):)', source_url, 1)
3346 if len(rtmp_url_parts) == 3:
3347 rtmp_url, prefix, play_path = rtmp_url_parts
3350 'play_path': prefix + play_path,
3353 a_format.update(rtmp_params)
3354 formats.append(a_format)
3357 def _live_title(self, name):
3358 """ Generate the title
for a live video
"""
3359 now = datetime.datetime.now()
3360 now_str = now.strftime('%Y-%m-%d %H:%M')
3361 return name + ' ' + now_str
3363 def _int(self, v, name, fatal=False, **kwargs):
3364 res = int_or_none(v, **kwargs)
3365 if 'get_attr' in kwargs:
3366 print(getattr(v, kwargs['get_attr']))
3368 msg = 'Failed to extract %s: Could not parse value %r' % (name, v)
3370 raise ExtractorError(msg)
3372 self.report_warning(msg)
3375 def _float(self, v, name, fatal=False, **kwargs):
3376 res = float_or_none(v, **kwargs)
3378 msg = 'Failed to extract %s: Could not parse value %r' % (name, v)
3380 raise ExtractorError(msg)
3382 self.report_warning(msg)
3385 def _set_cookie(self, domain, name, value, expire_time=None, port=None,
3386 path='/', secure=False, discard=False, rest={}, **kwargs):
3387 cookie = compat_cookiejar_Cookie(
3388 0, name, value, port, port is not None, domain, True,
3389 domain.startswith('.'), path, True, secure, expire_time,
3390 discard, None, None, rest)
3391 self._downloader.cookiejar.set_cookie(cookie)
3393 def _get_cookies(self, url):
3394 """ Return a compat_cookies_SimpleCookie
with the cookies
for the url
"""
3395 req = sanitized_Request(url)
3396 self._downloader.cookiejar.add_cookie_header(req)
3397 return compat_cookies_SimpleCookie(req.get_header('Cookie'))
3399 def _apply_first_set_cookie_header(self, url_handle, cookie):
3401 Apply first Set
-Cookie header instead of the last
. Experimental
.
3403 Some
sites (e
.g
. [1-3]) may serve two cookies under the same name
3404 in Set
-Cookie header
and expect the
first (old
) one to be
set rather
3405 than
second (new
). However
, as of RFC6265 the newer one cookie
3406 should be
set into cookie store what actually happens
.
3407 We will workaround this issue by resetting the cookie to
3408 the first one manually
.
3409 1. https
://new
.vk
.com
/
3410 2. https
://github
.com
/ytdl
-org
/youtube
-dl
/issues
/9841#issuecomment-227871201
3411 3. https
://learning
.oreilly
.com
/
3413 for header, cookies in url_handle.headers.items():
3414 if header.lower() != 'set-cookie':
3416 if sys.version_info[0] >= 3:
3417 cookies = cookies.encode('iso-8859-1')
3418 cookies = cookies.decode('utf-8')
3419 cookie_value = re.search(
3420 r'%s=(.+?);.*?\b[Dd]omain=(.+?)(?:[,;]|$)' % cookie, cookies)
3422 value, domain = cookie_value.groups()
3423 self._set_cookie(domain, cookie, value)
3426 def get_testcases(self, include_onlymatching=False):
3427 t = getattr(self, '_TEST', None)
3429 assert not hasattr(self, '_TESTS'), \
3430 '%s has _TEST and _TESTS' % type(self).__name__
3433 tests = getattr(self, '_TESTS', [])
3435 if not include_onlymatching and t.get('only_matching', False):
3437 t['name'] = type(self).__name__[:-len('IE')]
3440 def is_suitable(self, age_limit):
3441 """ Test whether the extractor
is generally suitable
for the given
3442 age
limit (i
.e
. pornographic sites are
not, all others usually are
) """
3444 any_restricted = False
3445 for tc in self.get_testcases(include_onlymatching=False):
3446 if tc.get('playlist', []):
3447 tc = tc['playlist'][0]
3448 is_restricted = age_restricted(
3449 tc.get('info_dict', {}).get('age_limit'), age_limit)
3450 if not is_restricted:
3452 any_restricted = any_restricted or is_restricted
3453 return not any_restricted
3455 def extract_subtitles(self, *args, **kwargs):
3456 if (self.get_param('writesubtitles', False)
3457 or self.get_param('listsubtitles')):
3458 return self._get_subtitles(*args, **kwargs)
3461 def _get_subtitles(self, *args, **kwargs):
3462 raise NotImplementedError('This method must be implemented by subclasses')
3465 def _merge_subtitle_items(subtitle_list1, subtitle_list2):
3466 """ Merge subtitle items
for one language
. Items
with duplicated URLs
3467 will be dropped
. """
3468 list1_urls = set([item['url'] for item in subtitle_list1])
3469 ret = list(subtitle_list1)
3470 ret.extend([item for item in subtitle_list2 if item['url'] not in list1_urls])
3474 def _merge_subtitles(cls, *dicts, target=None):
3475 """ Merge subtitle dictionaries
, language by language
. """
3479 for lang, subs in d.items():
3480 target[lang] = cls._merge_subtitle_items(target.get(lang, []), subs)
3483 def extract_automatic_captions(self, *args, **kwargs):
3484 if (self.get_param('writeautomaticsub', False)
3485 or self.get_param('listsubtitles')):
3486 return self._get_automatic_captions(*args, **kwargs)
3489 def _get_automatic_captions(self, *args, **kwargs):
3490 raise NotImplementedError('This method must be implemented by subclasses')
3492 def mark_watched(self, *args, **kwargs):
3493 if (self.get_param('mark_watched', False)
3494 and (self._get_login_info()[0] is not None
3495 or self.get_param('cookiefile') is not None)):
3496 self._mark_watched(*args, **kwargs)
3498 def _mark_watched(self, *args, **kwargs):
3499 raise NotImplementedError('This method must be implemented by subclasses')
3501 def geo_verification_headers(self):
3503 geo_verification_proxy = self.get_param('geo_verification_proxy')
3504 if geo_verification_proxy:
3505 headers['Ytdl-request-proxy'] = geo_verification_proxy
3508 def _generic_id(self, url):
3509 return compat_urllib_parse_unquote(os.path.splitext(url.rstrip('/').split('/')[-1])[0])
3511 def _generic_title(self, url):
3512 return compat_urllib_parse_unquote(os.path.splitext(url_basename(url))[0])
3515 def _availability(is_private=None, needs_premium=None, needs_subscription=None, needs_auth=None, is_unlisted=None):
3516 all_known = all(map(
3517 lambda x: x is not None,
3518 (is_private, needs_premium, needs_subscription, needs_auth, is_unlisted)))
3520 'private' if is_private
3521 else 'premium_only' if needs_premium
3522 else 'subscriber_only' if needs_subscription
3523 else 'needs_auth' if needs_auth
3524 else 'unlisted' if is_unlisted
3525 else 'public' if all_known
3528 def _configuration_arg(self, key, default=NO_DEFAULT, casesense=False):
3530 @returns A list of values for the extractor argument given by "key"
3531 or "default" if no such key is present
3532 @param default The default value to return when the key is not present (default: [])
3533 @param casesense When false, the values are converted to lower case
3536 self._downloader.params, ('extractor_args', self.ie_key().lower(), key))
3538 return [] if default is NO_DEFAULT else default
3539 return list(val) if casesense else [x.lower() for x in val]
3542 class SearchInfoExtractor(InfoExtractor):
3544 Base
class for paged search queries extractors
.
3545 They accept URLs
in the format
_SEARCH_KEY(|all|
[0-9]):{query}
3546 Instances should define _SEARCH_KEY
and _MAX_RESULTS
.
3550 def _make_valid_url(cls):
3551 return r'%s(?P<prefix>|[1-9][0-9]*|all):(?P<query>[\s\S]+)' % cls._SEARCH_KEY
3554 def suitable(cls, url):
3555 return re.match(cls._make_valid_url(), url) is not None
3557 def _real_extract(self, query):
3558 mobj = re.match(self._make_valid_url(), query)
3560 raise ExtractorError('Invalid search query "%s"' % query)
3562 prefix = mobj.group('prefix')
3563 query = mobj.group('query')
3565 return self._get_n_results(query, 1)
3566 elif prefix == 'all':
3567 return self._get_n_results(query, self._MAX_RESULTS)
3571 raise ExtractorError('invalid download number %s for query "%s"' % (n, query))
3572 elif n > self._MAX_RESULTS:
3573 self.report_warning('%s returns max %i results (you requested %i)' % (self._SEARCH_KEY, self._MAX_RESULTS, n))
3574 n = self._MAX_RESULTS
3575 return self._get_n_results(query, n)
3577 def _get_n_results(self, query, n):
3578 """Get a specified number of results
for a query
"""
3579 raise NotImplementedError('This method must be implemented by subclasses')
3582 def SEARCH_KEY(self):
3583 return self._SEARCH_KEY