2 from __future__
import unicode_literals
18 from ..compat
import (
19 compat_cookiejar_Cookie
,
22 compat_etree_fromstring
,
29 compat_urllib_parse_unquote
,
30 compat_urllib_parse_urlencode
,
31 compat_urllib_request
,
33 compat_xml_parse_error
,
35 from ..downloader
import FileDownloader
36 from ..downloader
.f4m
import (
38 remove_encrypted_media
,
66 parse_m3u8_attributes
,
88 class InfoExtractor(object):
89 """Information Extractor class.
91 Information extractors are the classes that, given a URL, extract
92 information about the video (or videos) the URL refers to. This
93 information includes the real video URL, the video title, author and
94 others. The information is stored in a dictionary which is then
95 passed to the YoutubeDL. The YoutubeDL processes this
96 information possibly downloading the video to the file system, among
97 other possible outcomes.
99 The type field determines the type of the result.
100 By far the most common value (and the default if _type is missing) is
101 "video", which indicates a single video.
103 For a video, the dictionaries must include the following fields:
105 id: Video identifier.
106 title: Video title, unescaped.
108 Additionally, it must contain either a formats entry or a url one:
110 formats: A list of dictionaries for each format available, ordered
111 from worst to best quality.
114 * url The mandatory URL representing the media:
115 for plain file media - HTTP URL of this file,
117 for HLS - URL of the M3U8 media playlist,
118 for HDS - URL of the F4M manifest,
120 - HTTP URL to plain file media (in case of
122 - URL of the MPD manifest or base URL
123 representing the media if MPD manifest
124 is parsed from a string (in case of
126 for MSS - URL of the ISM manifest.
128 The URL of the manifest file in case of
130 for HLS - URL of the M3U8 master playlist,
131 for HDS - URL of the F4M manifest,
132 for DASH - URL of the MPD manifest,
133 for MSS - URL of the ISM manifest.
134 * ext Will be calculated from URL if missing
135 * format A human-readable description of the format
136 ("mp4 container with h264/opus").
137 Calculated from the format_id, width, height.
138 and format_note fields if missing.
139 * format_id A short description of the format
140 ("mp4_h264_opus" or "19").
141 Technically optional, but strongly recommended.
142 * format_note Additional info about the format
143 ("3D" or "DASH video")
144 * width Width of the video, if known
145 * height Height of the video, if known
146 * resolution Textual description of width and height
147 * tbr Average bitrate of audio and video in KBit/s
148 * abr Average audio bitrate in KBit/s
149 * acodec Name of the audio codec in use
150 * asr Audio sampling rate in Hertz
151 * vbr Average video bitrate in KBit/s
153 * vcodec Name of the video codec in use
154 * container Name of the container format
155 * filesize The number of bytes, if known in advance
156 * filesize_approx An estimate for the number of bytes
157 * player_url SWF Player URL (used for rtmpdump).
158 * protocol The protocol that will be used for the actual
159 download, lower-case.
160 "http", "https", "rtsp", "rtmp", "rtmpe",
161 "m3u8", "m3u8_native" or "http_dash_segments".
163 Base URL for fragments. Each fragment's path
164 value (if present) will be relative to
166 * fragments A list of fragments of a fragmented media.
167 Each fragment entry must contain either an url
168 or a path. If an url is present it should be
169 considered by a client. Otherwise both path and
170 fragment_base_url must be present. Here is
171 the list of all potential fields:
172 * "url" - fragment's URL
173 * "path" - fragment's path relative to
175 * "duration" (optional, int or float)
176 * "filesize" (optional, int)
177 * preference Order number of this format. If this field is
178 present and not None, the formats get sorted
179 by this field, regardless of all other values.
180 -1 for default (order by other properties),
181 -2 or smaller for less than default.
182 < -1000 to hide the format (if there is
183 another one which is strictly better)
184 * language Language code, e.g. "de" or "en-US".
185 * language_preference Is this in the language mentioned in
187 10 if it's what the URL is about,
188 -1 for default (don't know),
189 -10 otherwise, other values reserved for now.
190 * quality Order number of the video quality of this
191 format, irrespective of the file format.
192 -1 for default (order by other properties),
193 -2 or smaller for less than default.
194 * source_preference Order number for this video source
195 (quality takes higher priority)
196 -1 for default (order by other properties),
197 -2 or smaller for less than default.
198 * http_headers A dictionary of additional HTTP headers
199 to add to the request.
200 * stretched_ratio If given and not 1, indicates that the
201 video's pixels are not square.
202 width : height ratio as float.
203 * no_resume The server does not support resuming the
204 (HTTP or RTMP) download. Boolean.
205 * downloader_options A dictionary of downloader options as
206 described in FileDownloader
208 url: Final video URL.
209 ext: Video filename extension.
210 format: The video format, defaults to ext (used for --get-format)
211 player_url: SWF Player URL (used for rtmpdump).
213 The following fields are optional:
215 alt_title: A secondary title of the video.
216 display_id An alternative identifier for the video, not necessarily
217 unique, but available before title. Typically, id is
218 something like "4234987", title "Dancing naked mole rats",
219 and display_id "dancing-naked-mole-rats"
220 thumbnails: A list of dictionaries, with the following entries:
221 * "id" (optional, string) - Thumbnail format ID
223 * "preference" (optional, int) - quality of the image
224 * "width" (optional, int)
225 * "height" (optional, int)
226 * "resolution" (optional, string "{width}x{height}",
228 * "filesize" (optional, int)
229 thumbnail: Full URL to a video thumbnail image.
230 description: Full video description.
231 uploader: Full name of the video uploader.
232 license: License name the video is licensed under.
233 creator: The creator of the video.
234 release_timestamp: UNIX timestamp of the moment the video was released.
235 release_date: The date (YYYYMMDD) when the video was released.
236 timestamp: UNIX timestamp of the moment the video was uploaded
237 upload_date: Video upload date (YYYYMMDD).
238 If not explicitly set, calculated from timestamp.
239 uploader_id: Nickname or id of the video uploader.
240 uploader_url: Full URL to a personal webpage of the video uploader.
241 channel: Full name of the channel the video is uploaded on.
242 Note that channel fields may or may not repeat uploader
243 fields. This depends on a particular extractor.
244 channel_id: Id of the channel.
245 channel_url: Full URL to a channel webpage.
246 location: Physical location where the video was filmed.
247 subtitles: The available subtitles as a dictionary in the format
248 {tag: subformats}. "tag" is usually a language code, and
249 "subformats" is a list sorted from lower to higher
250 preference, each element is a dictionary with the "ext"
252 * "data": The subtitles file contents
253 * "url": A URL pointing to the subtitles file
254 "ext" will be calculated from URL if missing
255 automatic_captions: Like 'subtitles'; contains automatically generated
256 captions instead of normal subtitles
257 duration: Length of the video in seconds, as an integer or float.
258 view_count: How many users have watched the video on the platform.
259 like_count: Number of positive ratings of the video
260 dislike_count: Number of negative ratings of the video
261 repost_count: Number of reposts of the video
262 average_rating: Average rating give by users, the scale used depends on the webpage
263 comment_count: Number of comments on the video
264 comments: A list of comments, each with one or more of the following
265 properties (all but one of text or html optional):
266 * "author" - human-readable name of the comment author
267 * "author_id" - user ID of the comment author
268 * "author_thumbnail" - The thumbnail of the comment author
270 * "html" - Comment as HTML
271 * "text" - Plain text of the comment
272 * "timestamp" - UNIX timestamp of comment
273 * "parent" - ID of the comment this one is replying to.
274 Set to "root" to indicate that this is a
275 comment to the original video.
276 * "like_count" - Number of positive ratings of the comment
277 * "dislike_count" - Number of negative ratings of the comment
278 * "is_favorited" - Whether the comment is marked as
279 favorite by the video uploader
280 * "author_is_uploader" - Whether the comment is made by
282 age_limit: Age restriction for the video, as an integer (years)
283 webpage_url: The URL to the video webpage, if given to yt-dlp it
284 should allow to get the same result again. (It will be set
285 by YoutubeDL if it's missing)
286 categories: A list of categories that the video falls in, for example
288 tags: A list of tags assigned to the video, e.g. ["sweden", "pop music"]
289 is_live: True, False, or None (=unknown). Whether this video is a
290 live stream that goes on instead of a fixed-length video.
291 was_live: True, False, or None (=unknown). Whether this video was
292 originally a live stream.
293 start_time: Time in seconds where the reproduction should start, as
294 specified in the URL.
295 end_time: Time in seconds where the reproduction should end, as
296 specified in the URL.
297 chapters: A list of dictionaries, with the following entries:
298 * "start_time" - The start time of the chapter in seconds
299 * "end_time" - The end time of the chapter in seconds
300 * "title" (optional, string)
301 playable_in_embed: Whether this video is allowed to play in embedded
302 players on other sites. Can be True (=always allowed),
303 False (=never allowed), None (=unknown), or a string
304 specifying the criteria for embedability (Eg: 'whitelist')
305 availability: Under what condition the video is available. One of
306 'private', 'premium_only', 'subscriber_only', 'needs_auth',
307 'unlisted' or 'public'. Use 'InfoExtractor._availability'
309 __post_extractor: A function to be called just before the metadata is
310 written to either disk, logger or console. The function
311 must return a dict which will be added to the info_dict.
312 This is usefull for additional information that is
313 time-consuming to extract. Note that the fields thus
314 extracted will not be available to output template and
315 match_filter. So, only "comments" and "comment_count" are
316 currently allowed to be extracted via this method.
318 The following fields should only be used when the video belongs to some logical
321 chapter: Name or title of the chapter the video belongs to.
322 chapter_number: Number of the chapter the video belongs to, as an integer.
323 chapter_id: Id of the chapter the video belongs to, as a unicode string.
325 The following fields should only be used when the video is an episode of some
326 series, programme or podcast:
328 series: Title of the series or programme the video episode belongs to.
329 season: Title of the season the video episode belongs to.
330 season_number: Number of the season the video episode belongs to, as an integer.
331 season_id: Id of the season the video episode belongs to, as a unicode string.
332 episode: Title of the video episode. Unlike mandatory video title field,
333 this field should denote the exact title of the video episode
334 without any kind of decoration.
335 episode_number: Number of the video episode within a season, as an integer.
336 episode_id: Id of the video episode, as a unicode string.
338 The following fields should only be used when the media is a track or a part of
341 track: Title of the track.
342 track_number: Number of the track within an album or a disc, as an integer.
343 track_id: Id of the track (useful in case of custom indexing, e.g. 6.iii),
345 artist: Artist(s) of the track.
346 genre: Genre(s) of the track.
347 album: Title of the album the track belongs to.
348 album_type: Type of the album (e.g. "Demo", "Full-length", "Split", "Compilation", etc).
349 album_artist: List of all artists appeared on the album (e.g.
350 "Ash Borer / Fell Voices" or "Various Artists", useful for splits
352 disc_number: Number of the disc or other physical medium the track belongs to,
354 release_year: Year (YYYY) when the album was released.
356 Unless mentioned otherwise, the fields should be Unicode strings.
358 Unless mentioned otherwise, None is equivalent to absence of information.
361 _type "playlist" indicates multiple videos.
362 There must be a key "entries", which is a list, an iterable, or a PagedList
363 object, each element of which is a valid dictionary by this specification.
365 Additionally, playlists can have "id", "title", and any other relevent
366 attributes with the same semantics as videos (see above).
369 _type "multi_video" indicates that there are multiple videos that
370 form a single show, for examples multiple acts of an opera or TV episode.
371 It must have an entries key like a playlist and contain all the keys
372 required for a video at the same time.
375 _type "url" indicates that the video must be extracted from another
376 location, possibly by a different extractor. Its only required key is:
377 "url" - the next URL to extract.
378 The key "ie_key" can be set to the class name (minus the trailing "IE",
379 e.g. "Youtube") if the extractor class is known in advance.
380 Additionally, the dictionary may have any properties of the resolved entity
381 known in advance, for example "title" if the title of the referred video is
385 _type "url_transparent" entities have the same specification as "url", but
386 indicate that the given additional information is more precise than the one
387 associated with the resolved URL.
388 This is useful when a site employs a video service that hosts the video and
389 its technical metadata, but that video service does not embed a useful
390 title, description etc.
393 Subclasses of this one should re-define the _real_initialize() and
394 _real_extract() methods and define a _VALID_URL regexp.
395 Probably, they should also be added to the list of extractors.
397 _GEO_BYPASS attribute may be set to False in order to disable
398 geo restriction bypass mechanisms for a particular extractor.
399 Though it won't disable explicit geo restriction bypass based on
400 country code provided with geo_bypass_country.
402 _GEO_COUNTRIES attribute may contain a list of presumably geo unrestricted
403 countries for this extractor. One of these countries will be used by
404 geo restriction bypass mechanism right away in order to bypass
405 geo restriction, of course, if the mechanism is not disabled.
407 _GEO_IP_BLOCKS attribute may contain a list of presumably geo unrestricted
408 IP blocks in CIDR notation for this extractor. One of these IP blocks
409 will be used by geo restriction bypass mechanism similarly
412 Finally, the _WORKING attribute should be set to False for broken IEs
413 in order to warn the users and skip the tests.
418 _x_forwarded_for_ip
= None
420 _GEO_COUNTRIES
= None
421 _GEO_IP_BLOCKS
= None
424 def __init__(self
, downloader
=None):
425 """Constructor. Receives an optional downloader."""
427 self
._x
_forwarded
_for
_ip
= None
428 self
.set_downloader(downloader
)
431 def suitable(cls
, url
):
432 """Receives a URL and returns True if suitable for this IE."""
434 # This does not use has/getattr intentionally - we want to know whether
435 # we have cached the regexp for *this* class, whereas getattr would also
436 # match the superclass
437 if '_VALID_URL_RE' not in cls
.__dict
__:
438 cls
._VALID
_URL
_RE
= re
.compile(cls
._VALID
_URL
)
439 return cls
._VALID
_URL
_RE
.match(url
) is not None
442 def _match_id(cls
, url
):
443 if '_VALID_URL_RE' not in cls
.__dict
__:
444 cls
._VALID
_URL
_RE
= re
.compile(cls
._VALID
_URL
)
445 m
= cls
._VALID
_URL
_RE
.match(url
)
447 return compat_str(m
.group('id'))
451 """Getter method for _WORKING."""
454 def initialize(self
):
455 """Initializes an instance (authentication, etc)."""
456 self
._initialize
_geo
_bypass
({
457 'countries': self
._GEO
_COUNTRIES
,
458 'ip_blocks': self
._GEO
_IP
_BLOCKS
,
461 self
._real
_initialize
()
464 def _initialize_geo_bypass(self
, geo_bypass_context
):
466 Initialize geo restriction bypass mechanism.
468 This method is used to initialize geo bypass mechanism based on faking
469 X-Forwarded-For HTTP header. A random country from provided country list
470 is selected and a random IP belonging to this country is generated. This
471 IP will be passed as X-Forwarded-For HTTP header in all subsequent
474 This method will be used for initial geo bypass mechanism initialization
475 during the instance initialization with _GEO_COUNTRIES and
478 You may also manually call it from extractor's code if geo bypass
479 information is not available beforehand (e.g. obtained during
480 extraction) or due to some other reason. In this case you should pass
481 this information in geo bypass context passed as first argument. It may
482 contain following fields:
484 countries: List of geo unrestricted countries (similar
486 ip_blocks: List of geo unrestricted IP blocks in CIDR notation
487 (similar to _GEO_IP_BLOCKS)
490 if not self
._x
_forwarded
_for
_ip
:
492 # Geo bypass mechanism is explicitly disabled by user
493 if not self
._downloader
.params
.get('geo_bypass', True):
496 if not geo_bypass_context
:
497 geo_bypass_context
= {}
499 # Backward compatibility: previously _initialize_geo_bypass
500 # expected a list of countries, some 3rd party code may still use
502 if isinstance(geo_bypass_context
, (list, tuple)):
503 geo_bypass_context
= {
504 'countries': geo_bypass_context
,
507 # The whole point of geo bypass mechanism is to fake IP
508 # as X-Forwarded-For HTTP header based on some IP block or
511 # Path 1: bypassing based on IP block in CIDR notation
513 # Explicit IP block specified by user, use it right away
514 # regardless of whether extractor is geo bypassable or not
515 ip_block
= self
._downloader
.params
.get('geo_bypass_ip_block', None)
517 # Otherwise use random IP block from geo bypass context but only
518 # if extractor is known as geo bypassable
520 ip_blocks
= geo_bypass_context
.get('ip_blocks')
521 if self
._GEO
_BYPASS
and ip_blocks
:
522 ip_block
= random
.choice(ip_blocks
)
525 self
._x
_forwarded
_for
_ip
= GeoUtils
.random_ipv4(ip_block
)
526 if self
._downloader
.params
.get('verbose', False):
527 self
._downloader
.to_screen(
528 '[debug] Using fake IP %s as X-Forwarded-For.'
529 % self
._x
_forwarded
_for
_ip
)
532 # Path 2: bypassing based on country code
534 # Explicit country code specified by user, use it right away
535 # regardless of whether extractor is geo bypassable or not
536 country
= self
._downloader
.params
.get('geo_bypass_country', None)
538 # Otherwise use random country code from geo bypass context but
539 # only if extractor is known as geo bypassable
541 countries
= geo_bypass_context
.get('countries')
542 if self
._GEO
_BYPASS
and countries
:
543 country
= random
.choice(countries
)
546 self
._x
_forwarded
_for
_ip
= GeoUtils
.random_ipv4(country
)
547 if self
._downloader
.params
.get('verbose', False):
548 self
._downloader
.to_screen(
549 '[debug] Using fake IP %s (%s) as X-Forwarded-For.'
550 % (self
._x
_forwarded
_for
_ip
, country
.upper()))
552 def extract(self
, url
):
553 """Extracts URL information and returns it in list of dicts."""
558 ie_result
= self
._real
_extract
(url
)
559 if self
._x
_forwarded
_for
_ip
:
560 ie_result
['__x_forwarded_for_ip'] = self
._x
_forwarded
_for
_ip
562 except GeoRestrictedError
as e
:
563 if self
.__maybe
_fake
_ip
_and
_retry
(e
.countries
):
566 except ExtractorError
:
568 except compat_http_client
.IncompleteRead
as e
:
569 raise ExtractorError('A network error has occurred.', cause
=e
, expected
=True)
570 except (KeyError, StopIteration) as e
:
571 raise ExtractorError('An extractor error has occurred.', cause
=e
)
573 def __maybe_fake_ip_and_retry(self
, countries
):
574 if (not self
._downloader
.params
.get('geo_bypass_country', None)
576 and self
._downloader
.params
.get('geo_bypass', True)
577 and not self
._x
_forwarded
_for
_ip
579 country_code
= random
.choice(countries
)
580 self
._x
_forwarded
_for
_ip
= GeoUtils
.random_ipv4(country_code
)
581 if self
._x
_forwarded
_for
_ip
:
583 'Video is geo restricted. Retrying extraction with fake IP %s (%s) as X-Forwarded-For.'
584 % (self
._x
_forwarded
_for
_ip
, country_code
.upper()))
588 def set_downloader(self
, downloader
):
589 """Sets the downloader for this IE."""
590 self
._downloader
= downloader
592 def _real_initialize(self
):
593 """Real initialization process. Redefine in subclasses."""
596 def _real_extract(self
, url
):
597 """Real extraction process. Redefine in subclasses."""
602 """A string for getting the InfoExtractor with get_info_extractor"""
603 return compat_str(cls
.__name
__[:-2])
607 return compat_str(type(self
).__name
__[:-2])
610 def __can_accept_status_code(err
, expected_status
):
611 assert isinstance(err
, compat_urllib_error
.HTTPError
)
612 if expected_status
is None:
614 if isinstance(expected_status
, compat_integer_types
):
615 return err
.code
== expected_status
616 elif isinstance(expected_status
, (list, tuple)):
617 return err
.code
in expected_status
618 elif callable(expected_status
):
619 return expected_status(err
.code
) is True
623 def _request_webpage(self
, url_or_request
, video_id
, note
=None, errnote
=None, fatal
=True, data
=None, headers
={}, query={}
, expected_status
=None):
625 Return the response handle.
627 See _download_webpage docstring for arguments specification.
629 if not self
._downloader
._first
_webpage
_request
:
630 sleep_interval
= float_or_none(self
._downloader
.params
.get('sleep_interval_requests')) or 0
631 if sleep_interval
> 0:
632 self
.to_screen('Sleeping %s seconds ...' % sleep_interval
)
633 time
.sleep(sleep_interval
)
635 self
._downloader
._first
_webpage
_request
= False
638 self
.report_download_webpage(video_id
)
639 elif note
is not False:
641 self
.to_screen('%s' % (note
,))
643 self
.to_screen('%s: %s' % (video_id
, note
))
645 # Some sites check X-Forwarded-For HTTP header in order to figure out
646 # the origin of the client behind proxy. This allows bypassing geo
647 # restriction by faking this header's value to IP that belongs to some
648 # geo unrestricted country. We will do so once we encounter any
649 # geo restriction error.
650 if self
._x
_forwarded
_for
_ip
:
651 if 'X-Forwarded-For' not in headers
:
652 headers
['X-Forwarded-For'] = self
._x
_forwarded
_for
_ip
654 if isinstance(url_or_request
, compat_urllib_request
.Request
):
655 url_or_request
= update_Request(
656 url_or_request
, data
=data
, headers
=headers
, query
=query
)
659 url_or_request
= update_url_query(url_or_request
, query
)
660 if data
is not None or headers
:
661 url_or_request
= sanitized_Request(url_or_request
, data
, headers
)
662 exceptions
= [compat_urllib_error
.URLError
, compat_http_client
.HTTPException
, socket
.error
]
663 if hasattr(ssl
, 'CertificateError'):
664 exceptions
.append(ssl
.CertificateError
)
666 return self
._downloader
.urlopen(url_or_request
)
667 except tuple(exceptions
) as err
:
668 if isinstance(err
, compat_urllib_error
.HTTPError
):
669 if self
.__can
_accept
_status
_code
(err
, expected_status
):
670 # Retain reference to error to prevent file object from
671 # being closed before it can be read. Works around the
672 # effects of <https://bugs.python.org/issue15002>
673 # introduced in Python 3.4.1.
680 errnote
= 'Unable to download webpage'
682 errmsg
= '%s: %s' % (errnote
, error_to_compat_str(err
))
684 raise ExtractorError(errmsg
, sys
.exc_info()[2], cause
=err
)
686 self
._downloader
.report_warning(errmsg
)
689 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):
691 Return a tuple (page content as string, URL handle).
693 See _download_webpage docstring for arguments specification.
695 # Strip hashes from the URL (#1038)
696 if isinstance(url_or_request
, (compat_str
, str)):
697 url_or_request
= url_or_request
.partition('#')[0]
699 urlh
= self
._request
_webpage
(url_or_request
, video_id
, note
, errnote
, fatal
, data
=data
, headers
=headers
, query
=query
, expected_status
=expected_status
)
703 content
= self
._webpage
_read
_content
(urlh
, url_or_request
, video_id
, note
, errnote
, fatal
, encoding
=encoding
)
704 return (content
, urlh
)
707 def _guess_encoding_from_content(content_type
, webpage_bytes
):
708 m
= re
.match(r
'[a-zA-Z0-9_.-]+/[a-zA-Z0-9_.-]+\s*;\s*charset=(.+)', content_type
)
710 encoding
= m
.group(1)
712 m
= re
.search(br
'<meta[^>]+charset=[\'"]?([^\'")]+)[ /\'">]',
713 webpage_bytes[:1024])
715 encoding = m.group(1).decode('ascii')
716 elif webpage_bytes.startswith(b'\xff\xfe'):
723 def __check_blocked(self, content):
724 first_block = content[:512]
725 if ('<title>Access to this site is blocked</title>' in content
726 and 'Websense' in first_block):
727 msg = 'Access to this webpage has been blocked by Websense filtering software in your network.'
728 blocked_iframe = self._html_search_regex(
729 r'<iframe src="([^
"]+)"', content,
730 'Websense information URL
', default=None)
732 msg += ' Visit
%s for more details
' % blocked_iframe
733 raise ExtractorError(msg, expected=True)
734 if '<title
>The URL you requested has been blocked
</title
>' in first_block:
736 'Access to this webpage has been blocked by Indian censorship
. '
737 'Use a VPN
or proxy
server (with --proxy
) to route around it
.')
738 block_msg = self._html_search_regex(
739 r'</h1
><p
>(.*?
)</p
>',
740 content, 'block message
', default=None)
742 msg += ' (Message
: "%s")' % block_msg.replace('\n', ' ')
743 raise ExtractorError(msg, expected=True)
744 if ('<title
>TTK
:: Доступ к ресурсу ограничен
</title
>' in content
745 and 'blocklist
.rkn
.gov
.ru
' in content):
746 raise ExtractorError(
747 'Access to this webpage has been blocked by decision of the Russian government
. '
748 'Visit http
://blocklist
.rkn
.gov
.ru
/ for a block reason
.',
751 def _webpage_read_content(self, urlh, url_or_request, video_id, note=None, errnote=None, fatal=True, prefix=None, encoding=None):
752 content_type = urlh.headers.get('Content
-Type
', '')
753 webpage_bytes = urlh.read()
754 if prefix is not None:
755 webpage_bytes = prefix + webpage_bytes
757 encoding = self._guess_encoding_from_content(content_type, webpage_bytes)
758 if self._downloader.params.get('dump_intermediate_pages
', False):
759 self.to_screen('Dumping request to
' + urlh.geturl())
760 dump = base64.b64encode(webpage_bytes).decode('ascii
')
761 self._downloader.to_screen(dump)
762 if self._downloader.params.get('write_pages
', False):
763 basen = '%s_%s' % (video_id, urlh.geturl())
765 h = '___
' + hashlib.md5(basen.encode('utf
-8')).hexdigest()
766 basen = basen[:240 - len(h)] + h
767 raw_filename = basen + '.dump
'
768 filename = sanitize_filename(raw_filename, restricted=True)
769 self.to_screen('Saving request to
' + filename)
770 # Working around MAX_PATH limitation on Windows (see
771 # http://msdn.microsoft.com/en-us/library/windows/desktop/aa365247(v=vs.85).aspx)
772 if compat_os_name == 'nt
':
773 absfilepath = os.path.abspath(filename)
774 if len(absfilepath) > 259:
775 filename = '\\\\?
\\' + absfilepath
776 with open(filename, 'wb
') as outf:
777 outf.write(webpage_bytes)
780 content = webpage_bytes.decode(encoding, 'replace
')
782 content = webpage_bytes.decode('utf
-8', 'replace
')
784 self.__check_blocked(content)
788 def _download_webpage(
789 self, url_or_request, video_id, note=None, errnote=None,
790 fatal=True, tries=1, timeout=5, encoding=None, data=None,
791 headers={}, query={}, expected_status=None):
793 Return the data of the page as a string.
796 url_or_request -- plain text URL as a string or
797 a compat_urllib_request.Requestobject
798 video_id -- Video/playlist/item identifier (string)
801 note -- note printed before downloading (string)
802 errnote -- note printed in case of an error (string)
803 fatal -- flag denoting whether error should be considered fatal,
804 i.e. whether it should cause ExtractionError to be raised,
805 otherwise a warning will be reported and extraction continued
806 tries -- number of tries
807 timeout -- sleep interval between tries
808 encoding -- encoding for a page content decoding, guessed automatically
809 when not explicitly specified
810 data -- POST data (bytes)
811 headers -- HTTP headers (dict)
812 query -- URL query (dict)
813 expected_status -- allows to accept failed HTTP requests (non 2xx
814 status code) by explicitly specifying a set of accepted status
815 codes. Can be any of the following entities:
816 - an integer type specifying an exact failed status code to
818 - a list or a tuple of integer types specifying a list of
819 failed status codes to accept
820 - a callable accepting an actual failed status code and
821 returning True if it should be accepted
822 Note that this argument does not affect success status codes (2xx)
823 which are always accepted.
828 while success is False:
830 res = self._download_webpage_handle(
831 url_or_request, video_id, note, errnote, fatal,
832 encoding=encoding, data=data, headers=headers, query=query,
833 expected_status=expected_status)
835 except compat_http_client.IncompleteRead as e:
837 if try_count >= tries:
839 self._sleep(timeout, video_id)
846 def _download_xml_handle(
847 self, url_or_request, video_id, note='Downloading XML
',
848 errnote='Unable to download XML
', transform_source=None,
849 fatal=True, encoding=None, data=None, headers={}, query={},
850 expected_status=None):
852 Return a tuple (xml as an compat_etree_Element, URL handle).
854 See _download_webpage docstring for arguments specification.
856 res = self._download_webpage_handle(
857 url_or_request, video_id, note, errnote, fatal=fatal,
858 encoding=encoding, data=data, headers=headers, query=query,
859 expected_status=expected_status)
862 xml_string, urlh = res
863 return self._parse_xml(
864 xml_string, video_id, transform_source=transform_source,
868 self, url_or_request, video_id,
869 note='Downloading XML
', errnote='Unable to download XML
',
870 transform_source=None, fatal=True, encoding=None,
871 data=None, headers={}, query={}, expected_status=None):
873 Return the xml as an compat_etree_Element.
875 See _download_webpage docstring for arguments specification.
877 res = self._download_xml_handle(
878 url_or_request, video_id, note=note, errnote=errnote,
879 transform_source=transform_source, fatal=fatal, encoding=encoding,
880 data=data, headers=headers, query=query,
881 expected_status=expected_status)
882 return res if res is False else res[0]
884 def _parse_xml(self, xml_string, video_id, transform_source=None, fatal=True):
886 xml_string = transform_source(xml_string)
888 return compat_etree_fromstring(xml_string.encode('utf
-8'))
889 except compat_xml_parse_error as ve:
890 errmsg = '%s: Failed to parse XML
' % video_id
892 raise ExtractorError(errmsg, cause=ve)
894 self.report_warning(errmsg + str(ve))
896 def _download_json_handle(
897 self, url_or_request, video_id, note='Downloading JSON metadata
',
898 errnote='Unable to download JSON metadata
', transform_source=None,
899 fatal=True, encoding=None, data=None, headers={}, query={},
900 expected_status=None):
902 Return a tuple (JSON object, URL handle).
904 See _download_webpage docstring for arguments specification.
906 res = self._download_webpage_handle(
907 url_or_request, video_id, note, errnote, fatal=fatal,
908 encoding=encoding, data=data, headers=headers, query=query,
909 expected_status=expected_status)
912 json_string, urlh = res
913 return self._parse_json(
914 json_string, video_id, transform_source=transform_source,
918 self, url_or_request, video_id, note='Downloading JSON metadata
',
919 errnote='Unable to download JSON metadata
', transform_source=None,
920 fatal=True, encoding=None, data=None, headers={}, query={},
921 expected_status=None):
923 Return the JSON object as a dict.
925 See _download_webpage docstring for arguments specification.
927 res = self._download_json_handle(
928 url_or_request, video_id, note=note, errnote=errnote,
929 transform_source=transform_source, fatal=fatal, encoding=encoding,
930 data=data, headers=headers, query=query,
931 expected_status=expected_status)
932 return res if res is False else res[0]
934 def _parse_json(self, json_string, video_id, transform_source=None, fatal=True):
936 json_string = transform_source(json_string)
938 return json.loads(json_string)
939 except ValueError as ve:
940 errmsg = '%s: Failed to parse JSON
' % video_id
942 raise ExtractorError(errmsg, cause=ve)
944 self.report_warning(errmsg + str(ve))
946 def report_warning(self, msg, video_id=None):
947 idstr = '' if video_id is None else '%s: ' % video_id
948 self._downloader.report_warning(
949 '[%s] %s%s' % (self.IE_NAME, idstr, msg))
951 def to_screen(self, msg):
952 """Print msg to screen, prefixing it with '[ie_name
]'"""
953 self._downloader.to_screen('[%s] %s' % (self.IE_NAME, msg))
955 def report_extraction(self, id_or_name):
956 """Report information extraction."""
957 self.to_screen('%s: Extracting information
' % id_or_name)
959 def report_download_webpage(self, video_id):
960 """Report webpage download."""
961 self.to_screen('%s: Downloading webpage
' % video_id)
963 def report_age_confirmation(self):
964 """Report attempt to confirm age."""
965 self.to_screen('Confirming age
')
967 def report_login(self):
968 """Report attempt to log in."""
969 self.to_screen('Logging
in')
972 def raise_login_required(msg='This video
is only available
for registered users
'):
973 raise ExtractorError(
974 '%s. Use
--username
and --password
or --netrc to provide account credentials
.' % msg,
978 def raise_geo_restricted(msg='This video
is not available
from your location due to geo restriction
', countries=None):
979 raise GeoRestrictedError(msg, countries=countries)
981 # Methods for following #608
983 def url_result(url, ie=None, video_id=None, video_title=None):
984 """Returns a URL that points to a page that should be processed"""
985 # TODO: ie should be the class used for getting the info
986 video_info = {'_type
': 'url
',
989 if video_id is not None:
990 video_info['id'] = video_id
991 if video_title is not None:
992 video_info['title
'] = video_title
995 def playlist_from_matches(self, matches, playlist_id=None, playlist_title=None, getter=None, ie=None):
997 self.url_result(self._proto_relative_url(getter(m) if getter else m), ie)
999 return self.playlist_result(
1000 urls, playlist_id=playlist_id, playlist_title=playlist_title)
1003 def playlist_result(entries, playlist_id=None, playlist_title=None, playlist_description=None, **kwargs):
1004 """Returns a playlist"""
1005 video_info = {'_type
': 'playlist
',
1007 video_info.update(kwargs)
1009 video_info['id'] = playlist_id
1011 video_info['title
'] = playlist_title
1012 if playlist_description is not None:
1013 video_info['description
'] = playlist_description
1016 def _search_regex(self, pattern, string, name, default=NO_DEFAULT, fatal=True, flags=0, group=None):
1018 Perform a regex search on the given string, using a single or a list of
1019 patterns returning the first matching group.
1020 In case of failure return a default value or raise a WARNING or a
1021 RegexNotFoundError, depending on fatal, specifying the field name.
1023 if isinstance(pattern, (str, compat_str, compiled_regex_type)):
1024 mobj = re.search(pattern, string, flags)
1027 mobj = re.search(p, string, flags)
1031 if not self._downloader.params.get('no_color
') and compat_os_name != 'nt
' and sys.stderr.isatty():
1032 _name = '\033[0;34m
%s\033[0m
' % name
1038 # return the first matching group
1039 return next(g for g in mobj.groups() if g is not None)
1041 return mobj.group(group)
1042 elif default is not NO_DEFAULT:
1045 raise RegexNotFoundError('Unable to extract
%s' % _name)
1047 self._downloader.report_warning('unable to extract
%s' % _name + bug_reports_message())
1050 def _html_search_regex(self, pattern, string, name, default=NO_DEFAULT, fatal=True, flags=0, group=None):
1052 Like _search_regex, but strips HTML tags and unescapes entities.
1054 res = self._search_regex(pattern, string, name, default, fatal, flags, group)
1056 return clean_html(res).strip()
1060 def _get_netrc_login_info(self, netrc_machine=None):
1063 netrc_machine = netrc_machine or self._NETRC_MACHINE
1065 if self._downloader.params.get('usenetrc
', False):
1067 info = netrc.netrc().authenticators(netrc_machine)
1068 if info is not None:
1072 raise netrc.NetrcParseError(
1073 'No authenticators
for %s' % netrc_machine)
1074 except (IOError, netrc.NetrcParseError) as err:
1075 self._downloader.report_warning(
1076 'parsing
.netrc
: %s' % error_to_compat_str(err))
1078 return username, password
1080 def _get_login_info(self, username_option='username
', password_option='password
', netrc_machine=None):
1082 Get the login info as (username, password)
1083 First look for the manually specified credentials using username_option
1084 and password_option as keys in params dictionary. If no such credentials
1085 available look in the netrc file using the netrc_machine or _NETRC_MACHINE
1087 If there's no info available
, return (None, None)
1089 if self._downloader is None:
1092 downloader_params = self._downloader.params
1094 # Attempt to use provided username and password or .netrc data
1095 if downloader_params.get(username_option) is not None:
1096 username = downloader_params[username_option]
1097 password = downloader_params[password_option]
1099 username, password = self._get_netrc_login_info(netrc_machine)
1101 return username, password
1103 def _get_tfa_info(self, note='two-factor verification code'):
1105 Get the two
-factor authentication info
1106 TODO
- asking the user will be required
for sms
/phone verify
1107 currently just uses the command line option
1108 If there
's no info available, return None
1110 if self._downloader is None:
1112 downloader_params = self._downloader.params
1114 if downloader_params.get('twofactor
') is not None:
1115 return downloader_params['twofactor
']
1117 return compat_getpass('Type
%s and press
[Return
]: ' % note)
1119 # Helper functions for extracting OpenGraph info
1121 def _og_regexes(prop):
1122 content_re = r'content
=(?
:"([^"]+?
)"|\'([^\']+?)\'|\s*([^\s"\'=<>`
]+?
))'
1123 property_re = (r'(?
:name|
property)=(?
:\'og
[:-]%(prop)s\'|
"og[:-]%(prop)s"|\s
*og
[:-]%(prop)s\b)'
1124 % {'prop': re.escape(prop)})
1125 template = r'<meta
[^
>]+?
%s[^
>]+?
%s'
1127 template % (property_re, content_re),
1128 template % (content_re, property_re),
1132 def _meta_regex(prop):
1133 return r'''(?isx)<meta
1134 (?=[^>]+(?:itemprop|name|property|id|http-equiv)=(["\']?)%s\1)
1135 [^>]+?content=(["\'])(?P<content>.*?)\2''' % re.escape(prop)
1137 def _og_search_property(self, prop, html, name=None, **kargs):
1138 if not isinstance(prop, (list, tuple)):
1141 name = 'OpenGraph
%s' % prop[0]
1144 og_regexes.extend(self._og_regexes(p))
1145 escaped = self._search_regex(og_regexes, html, name, flags=re.DOTALL, **kargs)
1148 return unescapeHTML(escaped)
1150 def _og_search_thumbnail(self, html, **kargs):
1151 return self._og_search_property('image
', html, 'thumbnail URL
', fatal=False, **kargs)
1153 def _og_search_description(self, html, **kargs):
1154 return self._og_search_property('description
', html, fatal=False, **kargs)
1156 def _og_search_title(self, html, **kargs):
1157 return self._og_search_property('title
', html, **kargs)
1159 def _og_search_video_url(self, html, name='video url
', secure=True, **kargs):
1160 regexes = self._og_regexes('video
') + self._og_regexes('video
:url
')
1162 regexes = self._og_regexes('video
:secure_url
') + regexes
1163 return self._html_search_regex(regexes, html, name, **kargs)
1165 def _og_search_url(self, html, **kargs):
1166 return self._og_search_property('url
', html, **kargs)
1168 def _html_search_meta(self, name, html, display_name=None, fatal=False, **kwargs):
1169 if not isinstance(name, (list, tuple)):
1171 if display_name is None:
1172 display_name = name[0]
1173 return self._html_search_regex(
1174 [self._meta_regex(n) for n in name],
1175 html, display_name, fatal=fatal, group='content
', **kwargs)
1177 def _dc_search_uploader(self, html):
1178 return self._html_search_meta('dc
.creator
', html, 'uploader
')
1180 def _rta_search(self, html):
1181 # See http://www.rtalabel.org/index.php?content=howtofaq#single
1182 if re.search(r'(?ix
)<meta\s
+name
="rating"\s
+'
1183 r' content
="RTA-5042-1996-1400-1577-RTA"',
1188 def _media_rating_search(self, html):
1189 # See http://www.tjg-designs.com/WP/metadata-code-examples-adding-metadata-to-your-web-pages/
1190 rating = self._html_search_meta('rating
', html)
1202 return RATING_TABLE.get(rating.lower())
1204 def _family_friendly_search(self, html):
1205 # See http://schema.org/VideoObject
1206 family_friendly = self._html_search_meta(
1207 'isFamilyFriendly
', html, default=None)
1209 if not family_friendly:
1218 return RATING_TABLE.get(family_friendly.lower())
1220 def _twitter_search_player(self, html):
1221 return self._html_search_meta('twitter
:player
', html,
1222 'twitter card player
')
1224 def _search_json_ld(self, html, video_id, expected_type=None, **kwargs):
1225 json_ld_list = list(re.finditer(JSON_LD_RE, html))
1226 default = kwargs.get('default
', NO_DEFAULT)
1227 # JSON-LD may be malformed and thus `fatal` should be respected.
1228 # At the same time `default` may be passed that assumes `fatal=False`
1229 # for _search_regex. Let's simulate the same behavior here
as well
.
1230 fatal
= kwargs
.get('fatal', True) if default
== NO_DEFAULT
else False
1232 for mobj
in json_ld_list
:
1233 json_ld_item
= self
._parse
_json
(
1234 mobj
.group('json_ld'), video_id
, fatal
=fatal
)
1235 if not json_ld_item
:
1237 if isinstance(json_ld_item
, dict):
1238 json_ld
.append(json_ld_item
)
1239 elif isinstance(json_ld_item
, (list, tuple)):
1240 json_ld
.extend(json_ld_item
)
1242 json_ld
= self
._json
_ld
(json_ld
, video_id
, fatal
=fatal
, expected_type
=expected_type
)
1245 if default
is not NO_DEFAULT
:
1248 raise RegexNotFoundError('Unable to extract JSON-LD')
1250 self
._downloader
.report_warning('unable to extract JSON-LD %s' % bug_reports_message())
1253 def _json_ld(self
, json_ld
, video_id
, fatal
=True, expected_type
=None):
1254 if isinstance(json_ld
, compat_str
):
1255 json_ld
= self
._parse
_json
(json_ld
, video_id
, fatal
=fatal
)
1259 if not isinstance(json_ld
, (list, tuple, dict)):
1261 if isinstance(json_ld
, dict):
1264 INTERACTION_TYPE_MAP
= {
1265 'CommentAction': 'comment',
1266 'AgreeAction': 'like',
1267 'DisagreeAction': 'dislike',
1268 'LikeAction': 'like',
1269 'DislikeAction': 'dislike',
1270 'ListenAction': 'view',
1271 'WatchAction': 'view',
1272 'ViewAction': 'view',
1275 def extract_interaction_type(e
):
1276 interaction_type
= e
.get('interactionType')
1277 if isinstance(interaction_type
, dict):
1278 interaction_type
= interaction_type
.get('@type')
1279 return str_or_none(interaction_type
)
1281 def extract_interaction_statistic(e
):
1282 interaction_statistic
= e
.get('interactionStatistic')
1283 if isinstance(interaction_statistic
, dict):
1284 interaction_statistic
= [interaction_statistic
]
1285 if not isinstance(interaction_statistic
, list):
1287 for is_e
in interaction_statistic
:
1288 if not isinstance(is_e
, dict):
1290 if is_e
.get('@type') != 'InteractionCounter':
1292 interaction_type
= extract_interaction_type(is_e
)
1293 if not interaction_type
:
1295 # For interaction count some sites provide string instead of
1296 # an integer (as per spec) with non digit characters (e.g. ",")
1297 # so extracting count with more relaxed str_to_int
1298 interaction_count
= str_to_int(is_e
.get('userInteractionCount'))
1299 if interaction_count
is None:
1301 count_kind
= INTERACTION_TYPE_MAP
.get(interaction_type
.split('/')[-1])
1304 count_key
= '%s_count' % count_kind
1305 if info
.get(count_key
) is not None:
1307 info
[count_key
] = interaction_count
1309 def extract_video_object(e
):
1310 assert e
['@type'] == 'VideoObject'
1312 'url': url_or_none(e
.get('contentUrl')),
1313 'title': unescapeHTML(e
.get('name')),
1314 'description': unescapeHTML(e
.get('description')),
1315 'thumbnail': url_or_none(e
.get('thumbnailUrl') or e
.get('thumbnailURL')),
1316 'duration': parse_duration(e
.get('duration')),
1317 'timestamp': unified_timestamp(e
.get('uploadDate')),
1318 'uploader': str_or_none(e
.get('author')),
1319 'filesize': float_or_none(e
.get('contentSize')),
1320 'tbr': int_or_none(e
.get('bitrate')),
1321 'width': int_or_none(e
.get('width')),
1322 'height': int_or_none(e
.get('height')),
1323 'view_count': int_or_none(e
.get('interactionCount')),
1325 extract_interaction_statistic(e
)
1329 item_type
= e
.get('@type')
1330 if expected_type
is not None and expected_type
!= item_type
:
1332 if item_type
in ('TVEpisode', 'Episode'):
1333 episode_name
= unescapeHTML(e
.get('name'))
1335 'episode': episode_name
,
1336 'episode_number': int_or_none(e
.get('episodeNumber')),
1337 'description': unescapeHTML(e
.get('description')),
1339 if not info
.get('title') and episode_name
:
1340 info
['title'] = episode_name
1341 part_of_season
= e
.get('partOfSeason')
1342 if isinstance(part_of_season
, dict) and part_of_season
.get('@type') in ('TVSeason', 'Season', 'CreativeWorkSeason'):
1344 'season': unescapeHTML(part_of_season
.get('name')),
1345 'season_number': int_or_none(part_of_season
.get('seasonNumber')),
1347 part_of_series
= e
.get('partOfSeries') or e
.get('partOfTVSeries')
1348 if isinstance(part_of_series
, dict) and part_of_series
.get('@type') in ('TVSeries', 'Series', 'CreativeWorkSeries'):
1349 info
['series'] = unescapeHTML(part_of_series
.get('name'))
1350 elif item_type
== 'Movie':
1352 'title': unescapeHTML(e
.get('name')),
1353 'description': unescapeHTML(e
.get('description')),
1354 'duration': parse_duration(e
.get('duration')),
1355 'timestamp': unified_timestamp(e
.get('dateCreated')),
1357 elif item_type
in ('Article', 'NewsArticle'):
1359 'timestamp': parse_iso8601(e
.get('datePublished')),
1360 'title': unescapeHTML(e
.get('headline')),
1361 'description': unescapeHTML(e
.get('articleBody')),
1363 elif item_type
== 'VideoObject':
1364 extract_video_object(e
)
1365 if expected_type
is None:
1369 video
= e
.get('video')
1370 if isinstance(video
, dict) and video
.get('@type') == 'VideoObject':
1371 extract_video_object(video
)
1372 if expected_type
is None:
1376 return dict((k
, v
) for k
, v
in info
.items() if v
is not None)
1379 def _hidden_inputs(html
):
1380 html
= re
.sub(r
'<!--(?:(?!<!--).)*-->', '', html
)
1382 for input in re
.findall(r
'(?i)(<input[^>]+>)', html
):
1383 attrs
= extract_attributes(input)
1386 if attrs
.get('type') not in ('hidden', 'submit'):
1388 name
= attrs
.get('name') or attrs
.get('id')
1389 value
= attrs
.get('value')
1390 if name
and value
is not None:
1391 hidden_inputs
[name
] = value
1392 return hidden_inputs
1394 def _form_hidden_inputs(self
, form_id
, html
):
1395 form
= self
._search
_regex
(
1396 r
'(?is)<form[^>]+?id=(["\'])%s\
1[^
>]*>(?P
<form
>.+?
)</form
>' % form_id,
1397 html, '%s form
' % form_id, group='form
')
1398 return self._hidden_inputs(form)
1401 regex = r' *((?P
<reverse
>\
+)?
(?P
<field
>[a
-zA
-Z0
-9_]+)((?P
<separator
>[~
:])(?P
<limit
>.*?
))?
)?
*$
'
1403 default = ('hidden
', 'hasvid
', 'ie_pref
', 'lang
', 'quality
',
1404 'res
', 'fps
', 'codec
:vp9
.2
', 'size
', 'br
', 'asr
',
1405 'proto
', 'ext
', 'has_audio
', 'source
', 'format_id
') # These must not be aliases
1408 'vcodec
': {'type': 'ordered
', 'regex
': True,
1409 'order
': ['av0?
1', 'vp0?
9.2', 'vp0?
9', '[hx
]265|he?vc?
', '[hx
]264|avc
', 'vp0?
8', 'mp4v|h263
', 'theora
', '', None, 'none
']},
1410 'acodec
': {'type': 'ordered
', 'regex
': True,
1411 'order
': ['opus
', 'vorbis
', 'aac
', 'mp?
4a?
', 'mp3
', 'e?a?c
-?
3', 'dts
', '', None, 'none
']},
1412 'proto
': {'type': 'ordered
', 'regex
': True, 'field
': 'protocol
',
1413 'order
': ['(ht|f
)tps
', '(ht|f
)tp$
', 'm3u8
.+', 'm3u8
', '.*dash
', '', 'mms|rtsp
', 'none
', 'f4
']},
1414 'vext
': {'type': 'ordered
', 'field
': 'video_ext
',
1415 'order
': ('mp4
', 'webm
', 'flv
', '', 'none
'),
1416 'order_free
': ('webm
', 'mp4
', 'flv
', '', 'none
')},
1417 'aext
': {'type': 'ordered
', 'field
': 'audio_ext
',
1418 'order
': ('m4a
', 'aac
', 'mp3
', 'ogg
', 'opus
', 'webm
', '', 'none
'),
1419 'order_free
': ('opus
', 'ogg
', 'webm
', 'm4a
', 'mp3
', 'aac
', '', 'none
')},
1420 'hidden
': {'visible': False, 'forced': True, 'type': 'extractor', 'max': -1000},
1421 'ie_pref
': {'priority': True, 'type': 'extractor'},
1422 'hasvid
': {'priority': True, 'field': 'vcodec', 'type': 'boolean', 'not_in_list': ('none',)},
1423 'hasaud
': {'field': 'acodec', 'type': 'boolean', 'not_in_list': ('none',)},
1424 'lang
': {'priority': True, 'convert': 'ignore', 'field': 'language_preference'},
1425 'quality
': {'convert': 'float_none'},
1426 'filesize
': {'convert': 'bytes'},
1427 'fs_approx
': {'convert': 'bytes', 'field': 'filesize_approx'},
1428 'id': {'convert': 'string', 'field': 'format_id'},
1429 'height
': {'convert': 'float_none'},
1430 'width
': {'convert': 'float_none'},
1431 'fps
': {'convert': 'float_none'},
1432 'tbr
': {'convert': 'float_none'},
1433 'vbr
': {'convert': 'float_none'},
1434 'abr
': {'convert': 'float_none'},
1435 'asr
': {'convert': 'float_none'},
1436 'source
': {'convert': 'ignore', 'field': 'source_preference'},
1438 'codec
': {'type': 'combined', 'field': ('vcodec', 'acodec')},
1439 'br
': {'type': 'combined', 'field': ('tbr', 'vbr', 'abr'), 'same_limit': True},
1440 'size
': {'type': 'combined', 'same_limit': True, 'field': ('filesize', 'fs_approx')},
1441 'ext
': {'type': 'combined', 'field': ('vext', 'aext')},
1442 'res
': {'type': 'multiple', 'field': ('height', 'width'), 'function': min},
1444 # Most of these exist only for compatibility reasons
1445 'dimension
': {'type': 'alias', 'field': 'res'},
1446 'resolution
': {'type': 'alias', 'field': 'res'},
1447 'extension
': {'type': 'alias', 'field': 'ext'},
1448 'bitrate
': {'type': 'alias', 'field': 'br'},
1449 'total_bitrate
': {'type': 'alias', 'field': 'tbr'},
1450 'video_bitrate
': {'type': 'alias', 'field': 'vbr'},
1451 'audio_bitrate
': {'type': 'alias', 'field': 'abr'},
1452 'framerate
': {'type': 'alias', 'field': 'fps'},
1453 'language_preference
': {'type': 'alias', 'field': 'lang'}, # not named as 'language
' because such a field exists
1454 'protocol
': {'type': 'alias', 'field': 'proto'},
1455 'source_preference
': {'type': 'alias', 'field': 'source'},
1456 'filesize_approx
': {'type': 'alias', 'field': 'fs_approx'},
1457 'filesize_estimate
': {'type': 'alias', 'field': 'size'},
1458 'samplerate
': {'type': 'alias', 'field': 'asr'},
1459 'video_ext
': {'type': 'alias', 'field': 'vext'},
1460 'audio_ext
': {'type': 'alias', 'field': 'aext'},
1461 'video_codec
': {'type': 'alias', 'field': 'vcodec'},
1462 'audio_codec
': {'type': 'alias', 'field': 'acodec'},
1463 'video
': {'type': 'alias', 'field': 'hasvid'},
1464 'has_video
': {'type': 'alias', 'field': 'hasvid'},
1465 'audio
': {'type': 'alias', 'field': 'hasaud'},
1466 'has_audio
': {'type': 'alias', 'field': 'hasaud'},
1467 'extractor
': {'type': 'alias', 'field': 'ie_pref'},
1468 'preference
': {'type': 'alias', 'field': 'ie_pref'},
1469 'extractor_preference
': {'type': 'alias', 'field': 'ie_pref'},
1470 'format_id
': {'type': 'alias', 'field': 'id'},
1475 def _get_field_setting(self, field, key):
1476 if field not in self.settings:
1477 self.settings[field] = {}
1478 propObj = self.settings[field]
1479 if key not in propObj:
1480 type = propObj.get('type')
1482 default = 'preference
' if type == 'extractor
' else (field,) if type in ('combined
', 'multiple
') else field
1483 elif key == 'convert
':
1484 default = 'order
' if type == 'ordered
' else 'float_string
' if field else 'ignore
'
1486 default = {'type': 'field', 'visible': True, 'order': [], 'not_in_list': (None,), 'function': max}.get(key, None)
1487 propObj[key] = default
1490 def _resolve_field_value(self, field, value, convertNone=False):
1495 value = value.lower()
1496 conversion = self._get_field_setting(field, 'convert
')
1497 if conversion == 'ignore
':
1499 if conversion == 'string
':
1501 elif conversion == 'float_none
':
1502 return float_or_none(value)
1503 elif conversion == 'bytes':
1504 return FileDownloader.parse_bytes(value)
1505 elif conversion == 'order
':
1506 order_list = (self._use_free_order and self._get_field_setting(field, 'order_free
')) or self._get_field_setting(field, 'order
')
1507 use_regex = self._get_field_setting(field, 'regex
')
1508 list_length = len(order_list)
1509 empty_pos = order_list.index('') if '' in order_list else list_length + 1
1510 if use_regex and value is not None:
1511 for i, regex in enumerate(order_list):
1512 if regex and re.match(regex, value):
1513 return list_length - i
1514 return list_length - empty_pos # not in list
1515 else: # not regex or value = None
1516 return list_length - (order_list.index(value) if value in order_list else empty_pos)
1518 if value.isnumeric():
1521 self.settings[field]['convert
'] = 'string
'
1524 def evaluate_params(self, params, sort_extractor):
1525 self._use_free_order = params.get('prefer_free_formats
', False)
1526 self._sort_user = params.get('format_sort
', [])
1527 self._sort_extractor = sort_extractor
1529 def add_item(field, reverse, closest, limit_text):
1530 field = field.lower()
1531 if field in self._order:
1533 self._order.append(field)
1534 limit = self._resolve_field_value(field, limit_text)
1537 'closest
': False if limit is None else closest,
1538 'limit_text
': limit_text,
1540 if field in self.settings:
1541 self.settings[field].update(data)
1543 self.settings[field] = data
1546 tuple(field for field in self.default if self._get_field_setting(field, 'forced
'))
1547 + (tuple() if params.get('format_sort_force
', False)
1548 else tuple(field for field in self.default if self._get_field_setting(field, 'priority
')))
1549 + tuple(self._sort_user) + tuple(sort_extractor) + self.default)
1551 for item in sort_list:
1552 match = re.match(self.regex, item)
1554 raise ExtractorError('Invalid format sort string
"%s" given by extractor
' % item)
1555 field = match.group('field
')
1558 if self._get_field_setting(field, 'type') == 'alias
':
1559 field = self._get_field_setting(field, 'field
')
1560 reverse = match.group('reverse
') is not None
1561 closest = match.group('separator
') == '~
'
1562 limit_text = match.group('limit
')
1564 has_limit = limit_text is not None
1565 has_multiple_fields = self._get_field_setting(field, 'type') == 'combined
'
1566 has_multiple_limits = has_limit and has_multiple_fields and not self._get_field_setting(field, 'same_limit
')
1568 fields = self._get_field_setting(field, 'field
') if has_multiple_fields else (field,)
1569 limits = limit_text.split(":") if has_multiple_limits else (limit_text,) if has_limit else tuple()
1570 limit_count = len(limits)
1571 for (i, f) in enumerate(fields):
1572 add_item(f, reverse, closest,
1573 limits[i] if i < limit_count
1574 else limits[0] if has_limit and not has_multiple_limits
1577 def print_verbose_info(self, to_screen):
1578 to_screen('[debug
] Sort order given by user
: %s' % ','.join(self._sort_user))
1579 if self._sort_extractor:
1580 to_screen('[debug
] Sort order given by extractor
: %s' % ', '.join(self._sort_extractor))
1581 to_screen('[debug
] Formats
sorted by
: %s' % ', '.join(['%s%s%s' % (
1582 '+' if self._get_field_setting(field, 'reverse
') else '', field,
1583 '%s%s(%s)' % ('~
' if self._get_field_setting(field, 'closest
') else ':',
1584 self._get_field_setting(field, 'limit_text
'),
1585 self._get_field_setting(field, 'limit
'))
1586 if self._get_field_setting(field, 'limit_text
') is not None else '')
1587 for field in self._order if self._get_field_setting(field, 'visible
')]))
1589 def _calculate_field_preference_from_value(self, format, field, type, value):
1590 reverse = self._get_field_setting(field, 'reverse
')
1591 closest = self._get_field_setting(field, 'closest
')
1592 limit = self._get_field_setting(field, 'limit
')
1594 if type == 'extractor
':
1595 maximum = self._get_field_setting(field, 'max')
1596 if value is None or (maximum is not None and value >= maximum):
1598 elif type == 'boolean
':
1599 in_list = self._get_field_setting(field, 'in_list
')
1600 not_in_list = self._get_field_setting(field, 'not_in_list
')
1601 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
1602 elif type == 'ordered
':
1603 value = self._resolve_field_value(field, value, True)
1605 # try to convert to number
1606 val_num = float_or_none(value)
1607 is_num = self._get_field_setting(field, 'convert
') != 'string
' and val_num is not None
1611 return ((-10, 0) if value is None
1612 else (1, value, 0) if not is_num # if a field has mixed strings and numbers, strings are sorted higher
1613 else (0, -abs(value - limit), value - limit if reverse else limit - value) if closest
1614 else (0, value, 0) if not reverse and (limit is None or value <= limit)
1615 else (0, -value, 0) if limit is None or (reverse and value == limit) or value > limit
1616 else (-1, value, 0))
1618 def _calculate_field_preference(self, format, field):
1619 type = self._get_field_setting(field, 'type') # extractor, boolean, ordered, field, multiple
1620 get_value = lambda f: format.get(self._get_field_setting(f, 'field
'))
1621 if type == 'multiple
':
1622 type = 'field
' # Only 'field
' is allowed in multiple for now
1623 actual_fields = self._get_field_setting(field, 'field
')
1625 def wrapped_function(values):
1626 values = tuple(filter(lambda x: x is not None, values))
1627 return (self._get_field_setting(field, 'function
')(*values) if len(values) > 1
1628 else values[0] if values
1631 value = wrapped_function((get_value(f) for f in actual_fields))
1633 value = get_value(field)
1634 return self._calculate_field_preference_from_value(format, field, type, value)
1636 def calculate_preference(self, format):
1637 # Determine missing protocol
1638 if not format.get('protocol
'):
1639 format['protocol
'] = determine_protocol(format)
1641 # Determine missing ext
1642 if not format.get('ext
') and 'url
' in format:
1643 format['ext
'] = determine_ext(format['url
'])
1644 if format.get('vcodec
') == 'none
':
1645 format['audio_ext
'] = format['ext
']
1646 format['video_ext
'] = 'none
'
1648 format['video_ext
'] = format['ext
']
1649 format['audio_ext
'] = 'none
'
1650 # if format.get('preference
') is None and format.get('ext
') in ('f4f
', 'f4m
'): # Not supported?
1651 # format['preference
'] = -1000
1653 # Determine missing bitrates
1654 if format.get('tbr
') is None:
1655 if format.get('vbr
') is not None and format.get('abr
') is not None:
1656 format['tbr
'] = format.get('vbr
', 0) + format.get('abr
', 0)
1658 if format.get('vcodec
') != "none" and format.get('vbr
') is None:
1659 format['vbr
'] = format.get('tbr
') - format.get('abr
', 0)
1660 if format.get('acodec
') != "none" and format.get('abr
') is None:
1661 format['abr
'] = format.get('tbr
') - format.get('vbr
', 0)
1663 return tuple(self._calculate_field_preference(format, field) for field in self._order)
1665 def _sort_formats(self, formats, field_preference=[]):
1667 raise ExtractorError('No video formats found
')
1668 format_sort = self.FormatSort() # params and to_screen are taken from the downloader
1669 format_sort.evaluate_params(self._downloader.params, field_preference)
1670 if self._downloader.params.get('verbose
', False):
1671 format_sort.print_verbose_info(self._downloader.to_screen)
1672 formats.sort(key=lambda f: format_sort.calculate_preference(f))
1674 def _check_formats(self, formats, video_id):
1676 formats[:] = filter(
1677 lambda f: self._is_valid_url(
1679 item='%s video format
' % f.get('format_id
') if f.get('format_id
') else 'video
'),
1683 def _remove_duplicate_formats(formats):
1687 if f['url
'] not in format_urls:
1688 format_urls.add(f['url
'])
1689 unique_formats.append(f)
1690 formats[:] = unique_formats
1692 def _is_valid_url(self, url, video_id, item='video
', headers={}):
1693 url = self._proto_relative_url(url, scheme='http
:')
1694 # For now assume non HTTP(S) URLs always valid
1695 if not (url.startswith('http
://') or url.startswith('https
://')):
1698 self._request_webpage(url, video_id, 'Checking
%s URL
' % item, headers=headers)
1700 except ExtractorError as e:
1702 '%s: %s URL
is invalid
, skipping
: %s'
1703 % (video_id, item, error_to_compat_str(e.cause)))
1706 def http_scheme(self):
1707 """ Either "http:" or "https:", depending on the user's preferences
"""
1710 if self._downloader.params.get('prefer_insecure', False)
1713 def _proto_relative_url(self, url, scheme=None):
1716 if url.startswith('//'):
1718 scheme = self.http_scheme()
1723 def _sleep(self, timeout, video_id, msg_template=None):
1724 if msg_template is None:
1725 msg_template = '%(video_id)s: Waiting for %(timeout)s seconds'
1726 msg = msg_template % {'video_id': video_id, 'timeout': timeout}
1730 def _extract_f4m_formats(self, manifest_url, video_id, preference=None, quality=None, f4m_id=None,
1731 transform_source=lambda s: fix_xml_ampersands(s).strip(),
1732 fatal=True, m3u8_id=None, data=None, headers={}, query={}):
1733 manifest = self._download_xml(
1734 manifest_url, video_id, 'Downloading f4m manifest',
1735 'Unable to download f4m manifest',
1736 # Some manifests may be malformed, e.g. prosiebensat1 generated manifests
1737 # (see https://github.com/ytdl-org/youtube-dl/issues/6215#issuecomment-121704244)
1738 transform_source=transform_source,
1739 fatal=fatal, data=data, headers=headers, query=query)
1741 if manifest is False:
1744 return self._parse_f4m_formats(
1745 manifest, manifest_url, video_id, preference=preference, quality=quality, f4m_id=f4m_id,
1746 transform_source=transform_source, fatal=fatal, m3u8_id=m3u8_id)
1748 def _parse_f4m_formats(self, manifest, manifest_url, video_id, preference=None, quality=None, f4m_id=None,
1749 transform_source=lambda s: fix_xml_ampersands(s).strip(),
1750 fatal=True, m3u8_id=None):
1751 if not isinstance(manifest, compat_etree_Element) and not fatal:
1754 # currently yt-dlp cannot decode the playerVerificationChallenge as Akamai uses Adobe Alchemy
1755 akamai_pv = manifest.find('{http://ns.adobe.com/f4m/1.0}pv-2.0')
1756 if akamai_pv is not None and ';' in akamai_pv.text:
1757 playerVerificationChallenge = akamai_pv.text.split(';')[0]
1758 if playerVerificationChallenge.strip() != '':
1762 manifest_version = '1.0'
1763 media_nodes = manifest.findall('{http://ns.adobe.com/f4m/1.0}media')
1765 manifest_version = '2.0'
1766 media_nodes = manifest.findall('{http://ns.adobe.com/f4m/2.0}media')
1767 # Remove unsupported DRM protected media from final formats
1768 # rendition (see https://github.com/ytdl-org/youtube-dl/issues/8573).
1769 media_nodes = remove_encrypted_media(media_nodes)
1773 manifest_base_url = get_base_url(manifest)
1775 bootstrap_info = xpath_element(
1776 manifest, ['{http://ns.adobe.com/f4m/1.0}bootstrapInfo', '{http://ns.adobe.com/f4m/2.0}bootstrapInfo'],
1777 'bootstrap info', default=None)
1780 mime_type = xpath_text(
1781 manifest, ['{http://ns.adobe.com/f4m/1.0}mimeType', '{http://ns.adobe.com/f4m/2.0}mimeType'],
1782 'base URL', default=None)
1783 if mime_type and mime_type.startswith('audio/'):
1786 for i, media_el in enumerate(media_nodes):
1787 tbr = int_or_none(media_el.attrib.get('bitrate'))
1788 width = int_or_none(media_el.attrib.get('width'))
1789 height = int_or_none(media_el.attrib.get('height'))
1790 format_id = '-'.join(filter(None, [f4m_id, compat_str(i if tbr is None else tbr)]))
1791 # If <bootstrapInfo> is present, the specified f4m is a
1792 # stream-level manifest, and only set-level manifests may refer to
1793 # external resources. See section 11.4 and section 4 of F4M spec
1794 if bootstrap_info is None:
1796 # @href is introduced in 2.0, see section 11.6 of F4M spec
1797 if manifest_version == '2.0':
1798 media_url = media_el.attrib.get('href')
1799 if media_url is None:
1800 media_url = media_el.attrib.get('url')
1804 media_url if media_url.startswith('http://') or media_url.startswith('https://')
1805 else ((manifest_base_url or '/'.join(manifest_url.split('/')[:-1])) + '/' + media_url))
1806 # If media_url is itself a f4m manifest do the recursive extraction
1807 # since bitrates in parent manifest (this one) and media_url manifest
1808 # may differ leading to inability to resolve the format by requested
1809 # bitrate in f4m downloader
1810 ext = determine_ext(manifest_url)
1812 f4m_formats = self._extract_f4m_formats(
1813 manifest_url, video_id, preference=preference, quality=quality, f4m_id=f4m_id,
1814 transform_source=transform_source, fatal=fatal)
1815 # Sometimes stream-level manifest contains single media entry that
1816 # does not contain any quality metadata (e.g. http://matchtv.ru/#live-player).
1817 # At the same time parent's media entry in set-level manifest may
1818 # contain it. We will copy it from parent in such cases.
1819 if len(f4m_formats) == 1:
1822 'tbr': f.get('tbr') or tbr,
1823 'width': f.get('width') or width,
1824 'height': f.get('height') or height,
1825 'format_id': f.get('format_id') if not tbr else format_id,
1828 formats.extend(f4m_formats)
1831 formats.extend(self._extract_m3u8_formats(
1832 manifest_url, video_id, 'mp4', preference=preference,
1833 quality=quality, m3u8_id=m3u8_id, fatal=fatal))
1836 'format_id': format_id,
1837 'url': manifest_url,
1838 'manifest_url': manifest_url,
1839 'ext': 'flv' if bootstrap_info is not None else None,
1845 'preference': preference,
1850 def _m3u8_meta_format(self, m3u8_url, ext=None, preference=None, quality=None, m3u8_id=None):
1852 'format_id': '-'.join(filter(None, [m3u8_id, 'meta'])),
1856 'preference': preference - 100 if preference else -100,
1858 'resolution': 'multiple',
1859 'format_note': 'Quality selection URL',
1862 def _extract_m3u8_formats(self, m3u8_url, video_id, ext=None,
1863 entry_protocol='m3u8', preference=None, quality=None,
1864 m3u8_id=None, note=None, errnote=None,
1865 fatal=True, live=False, data=None, headers={},
1867 res = self._download_webpage_handle(
1869 note=note or 'Downloading m3u8 information',
1870 errnote=errnote or 'Failed to download m3u8 information',
1871 fatal=fatal, data=data, headers=headers, query=query)
1876 m3u8_doc, urlh = res
1877 m3u8_url = urlh.geturl()
1879 return self._parse_m3u8_formats(
1880 m3u8_doc, m3u8_url, ext=ext, entry_protocol=entry_protocol,
1881 preference=preference, quality=quality, m3u8_id=m3u8_id,
1882 note=note, errnote=errnote, fatal=fatal, live=live, data=data,
1883 headers=headers, query=query, video_id=video_id)
1885 def _parse_m3u8_formats(self, m3u8_doc, m3u8_url, ext=None,
1886 entry_protocol='m3u8', preference=None, quality=None,
1887 m3u8_id=None, live=False, note=None, errnote=None,
1888 fatal=True, data=None, headers={}, query={}, video_id=None):
1889 if '#EXT-X-FAXS-CM:' in m3u8_doc: # Adobe Flash Access
1892 if re.search(r'#EXT-X-SESSION-KEY:.*?URI="skd://', m3u8_doc): # Apple FairPlay
1897 format_url = lambda u: (
1899 if re.match(r'^https?://', u)
1900 else compat_urlparse.urljoin(m3u8_url, u))
1902 split_discontinuity = self._downloader.params.get('hls_split_discontinuity', False)
1905 # 1. https://tools.ietf.org/html/draft-pantos-http-live-streaming-21
1906 # 2. https://github.com/ytdl-org/youtube-dl/issues/12211
1907 # 3. https://github.com/ytdl-org/youtube-dl/issues/18923
1909 # We should try extracting formats only from master playlists [1, 4.3.4],
1910 # i.e. playlists that describe available qualities. On the other hand
1911 # media playlists [1, 4.3.3] should be returned as is since they contain
1912 # just the media without qualities renditions.
1913 # Fortunately, master playlist can be easily distinguished from media
1914 # playlist based on particular tags availability. As of [1, 4.3.3, 4.3.4]
1915 # master playlist tags MUST NOT appear in a media playlist and vice versa.
1916 # As of [1, 4.3.3.1] #EXT-X-TARGETDURATION tag is REQUIRED for every
1917 # media playlist and MUST NOT appear in master playlist thus we can
1918 # clearly detect media playlist with this criterion.
1920 def _extract_m3u8_playlist_formats(format_url=None, m3u8_doc=None, video_id=None,
1921 fatal=True, data=None, headers={}):
1925 res = self._download_webpage_handle(
1926 format_url, video_id,
1928 errnote='Failed to download m3u8 playlist information',
1929 fatal=fatal, data=data, headers=headers)
1934 m3u8_doc, urlh = res
1935 format_url = urlh.geturl()
1937 playlist_formats = []
1940 if split_discontinuity
1947 for line in m3u8_doc.splitlines():
1948 if not line.startswith('#'):
1949 format_info['files'].append(line)
1950 elif split_discontinuity and line.startswith('#EXT-X-DISCONTINUITY'):
1952 playlist_formats.append(format_info)
1958 playlist_formats.append(format_info)
1959 return playlist_formats
1961 if '#EXT-X-TARGETDURATION' in m3u8_doc: # media playlist, return as is
1963 playlist_formats = _extract_m3u8_playlist_formats(m3u8_doc=m3u8_doc)
1965 for format in playlist_formats:
1968 format_id.append(m3u8_id)
1969 format_index = format.get('index')
1971 format_id.append(str(format_index))
1973 'format_id': '-'.join(format_id),
1974 'format_index': format_index,
1977 'protocol': entry_protocol,
1978 'preference': preference,
1986 last_stream_inf = {}
1988 def extract_media(x_media_line):
1989 media = parse_m3u8_attributes(x_media_line)
1990 # As per [1, 4.3.4.1] TYPE, GROUP-ID and NAME are REQUIRED
1991 media_type, group_id, name = media.get('TYPE'), media.get('GROUP-ID'), media.get('NAME')
1992 if not (media_type and group_id and name):
1994 groups.setdefault(group_id, []).append(media)
1995 if media_type not in ('VIDEO', 'AUDIO'):
1997 media_url = media.get('URI')
1999 manifest_url = format_url(media_url)
2001 playlist_formats = _extract_m3u8_playlist_formats(manifest_url, video_id=video_id,
2002 fatal=fatal, data=data, headers=headers)
2004 for format in playlist_formats:
2005 format_index = format.get('index')
2006 for v in (m3u8_id, group_id, name):
2010 format_id.append(str(format_index))
2012 'format_id': '-'.join(format_id),
2013 'format_index': format_index,
2014 'url': manifest_url,
2015 'manifest_url': m3u8_url,
2016 'language': media.get('LANGUAGE'),
2018 'protocol': entry_protocol,
2019 'preference': preference,
2022 if media_type == 'AUDIO':
2023 f['vcodec'] = 'none'
2026 def build_stream_name():
2027 # Despite specification does not mention NAME attribute for
2028 # EXT-X-STREAM-INF tag it still sometimes may be present (see [1]
2029 # or vidio test in TestInfoExtractor.test_parse_m3u8_formats)
2030 # 1. http://www.vidio.com/watch/165683-dj_ambred-booyah-live-2015
2031 stream_name = last_stream_inf.get('NAME')
2034 # If there is no NAME in EXT-X-STREAM-INF it will be obtained
2035 # from corresponding rendition group
2036 stream_group_id = last_stream_inf.get('VIDEO')
2037 if not stream_group_id:
2039 stream_group = groups.get(stream_group_id)
2040 if not stream_group:
2041 return stream_group_id
2042 rendition = stream_group[0]
2043 return rendition.get('NAME') or stream_group_id
2045 # parse EXT-X-MEDIA tags before EXT-X-STREAM-INF in order to have the
2046 # chance to detect video only formats when EXT-X-STREAM-INF tags
2047 # precede EXT-X-MEDIA tags in HLS manifest such as [3].
2048 for line in m3u8_doc.splitlines():
2049 if line.startswith('#EXT-X-MEDIA:'):
2052 for line in m3u8_doc.splitlines():
2053 if line.startswith('#EXT-X-STREAM-INF:'):
2054 last_stream_inf = parse_m3u8_attributes(line)
2055 elif line.startswith('#') or not line.strip():
2058 tbr = float_or_none(
2059 last_stream_inf.get('AVERAGE-BANDWIDTH')
2060 or last_stream_inf.get('BANDWIDTH'), scale=1000)
2061 manifest_url = format_url(line.strip())
2063 playlist_formats = _extract_m3u8_playlist_formats(manifest_url, video_id=video_id,
2064 fatal=fatal, data=data, headers=headers)
2066 for frmt in playlist_formats:
2069 format_id.append(m3u8_id)
2070 format_index = frmt.get('index')
2071 stream_name = build_stream_name()
2072 # Bandwidth of live streams may differ over time thus making
2073 # format_id unpredictable. So it's better to keep provided
2076 format_id.append(stream_name if stream_name else '%d' % (tbr if tbr else len(formats)))
2078 format_id.append(str(format_index))
2080 'format_id': '-'.join(format_id),
2081 'format_index': format_index,
2082 'url': manifest_url,
2083 'manifest_url': m3u8_url,
2086 'fps': float_or_none(last_stream_inf.get('FRAME-RATE')),
2087 'protocol': entry_protocol,
2088 'preference': preference,
2091 resolution = last_stream_inf.get('RESOLUTION')
2093 mobj = re.search(r'(?P<width>\d+)[xX](?P<height>\d+)', resolution)
2095 f['width'] = int(mobj.group('width'))
2096 f['height'] = int(mobj.group('height'))
2097 # Unified Streaming Platform
2099 r'audio.*?(?:%3D|=)(\d+)(?:-video.*?(?:%3D|=)(\d+))?', f['url'])
2101 abr, vbr = mobj.groups()
2102 abr, vbr = float_or_none(abr, 1000), float_or_none(vbr, 1000)
2107 codecs = parse_codecs(last_stream_inf.get('CODECS'))
2109 audio_group_id = last_stream_inf.get('AUDIO')
2110 # As per [1, 4.3.4.1.1] any EXT-X-STREAM-INF tag which
2111 # references a rendition group MUST have a CODECS attribute.
2112 # However, this is not always respected, for example, [2]
2113 # contains EXT-X-STREAM-INF tag which references AUDIO
2114 # rendition group but does not have CODECS and despite
2115 # referencing an audio group it represents a complete
2116 # (with audio and video) format. So, for such cases we will
2117 # ignore references to rendition groups and treat them
2118 # as complete formats.
2119 if audio_group_id and codecs and f.get('vcodec') != 'none':
2120 audio_group = groups.get(audio_group_id)
2121 if audio_group and audio_group[0].get('URI'):
2122 # TODO: update acodec for audio only formats with
2124 f['acodec'] = 'none'
2125 if not f.get('ext'):
2126 f['ext'] = 'm4a' if f.get('vcodec') == 'none' else 'mp4'
2130 progressive_uri = last_stream_inf.get('PROGRESSIVE-URI')
2133 del http_f['manifest_url']
2135 'format_id': f['format_id'].replace('hls-', 'http-'),
2137 'url': progressive_uri,
2139 formats.append(http_f)
2141 last_stream_inf = {}
2145 def _xpath_ns(path, namespace=None):
2149 for c in path.split('/'):
2150 if not c or c == '.':
2153 out.append('{%s}%s' % (namespace, c))
2154 return '/'.join(out)
2156 def _extract_smil_formats(self, smil_url, video_id, fatal=True, f4m_params=None, transform_source=None):
2157 smil = self._download_smil(smil_url, video_id, fatal=fatal, transform_source=transform_source)
2163 namespace = self._parse_smil_namespace(smil)
2165 return self._parse_smil_formats(
2166 smil, smil_url, video_id, namespace=namespace, f4m_params=f4m_params)
2168 def _extract_smil_info(self, smil_url, video_id, fatal=True, f4m_params=None):
2169 smil = self._download_smil(smil_url, video_id, fatal=fatal)
2172 return self._parse_smil(smil, smil_url, video_id, f4m_params=f4m_params)
2174 def _download_smil(self, smil_url, video_id, fatal=True, transform_source=None):
2175 return self._download_xml(
2176 smil_url, video_id, 'Downloading SMIL file',
2177 'Unable to download SMIL file', fatal=fatal, transform_source=transform_source)
2179 def _parse_smil(self, smil, smil_url, video_id, f4m_params=None):
2180 namespace = self._parse_smil_namespace(smil)
2182 formats = self._parse_smil_formats(
2183 smil, smil_url, video_id, namespace=namespace, f4m_params=f4m_params)
2184 subtitles = self._parse_smil_subtitles(smil, namespace=namespace)
2186 video_id = os.path.splitext(url_basename(smil_url))[0]
2190 for meta in smil.findall(self._xpath_ns('./head/meta', namespace)):
2191 name = meta.attrib.get('name')
2192 content = meta.attrib.get('content')
2193 if not name or not content:
2195 if not title and name == 'title':
2197 elif not description and name in ('description', 'abstract'):
2198 description = content
2199 elif not upload_date and name == 'date':
2200 upload_date = unified_strdate(content)
2203 'id': image.get('type'),
2204 'url': image.get('src'),
2205 'width': int_or_none(image.get('width')),
2206 'height': int_or_none(image.get('height')),
2207 } for image in smil.findall(self._xpath_ns('.//image', namespace)) if image.get('src')]
2211 'title': title or video_id,
2212 'description': description,
2213 'upload_date': upload_date,
2214 'thumbnails': thumbnails,
2216 'subtitles': subtitles,
2219 def _parse_smil_namespace(self, smil):
2220 return self._search_regex(
2221 r'(?i)^{([^}]+)?}smil$', smil.tag, 'namespace', default=None)
2223 def _parse_smil_formats(self, smil, smil_url, video_id, namespace=None, f4m_params=None, transform_rtmp_url=None):
2225 for meta in smil.findall(self._xpath_ns('./head/meta', namespace)):
2226 b = meta.get('base') or meta.get('httpBase')
2237 media = smil.findall(self._xpath_ns('.//video', namespace)) + smil.findall(self._xpath_ns('.//audio', namespace))
2238 for medium in media:
2239 src = medium.get('src')
2240 if not src or src in srcs:
2244 bitrate = float_or_none(medium.get('system-bitrate') or medium.get('systemBitrate'), 1000)
2245 filesize = int_or_none(medium.get('size') or medium.get('fileSize'))
2246 width = int_or_none(medium.get('width'))
2247 height = int_or_none(medium.get('height'))
2248 proto = medium.get('proto')
2249 ext = medium.get('ext')
2250 src_ext = determine_ext(src)
2251 streamer = medium.get('streamer') or base
2253 if proto == 'rtmp' or streamer.startswith('rtmp'):
2259 'format_id': 'rtmp-%d' % (rtmp_count if bitrate is None else bitrate),
2261 'filesize': filesize,
2265 if transform_rtmp_url:
2266 streamer, src = transform_rtmp_url(streamer, src)
2267 formats[-1].update({
2273 src_url = src if src.startswith('http') else compat_urlparse.urljoin(base, src)
2274 src_url = src_url.strip()
2276 if proto == 'm3u8' or src_ext == 'm3u8':
2277 m3u8_formats = self._extract_m3u8_formats(
2278 src_url, video_id, ext or 'mp4', m3u8_id='hls', fatal=False)
2279 if len(m3u8_formats) == 1:
2281 m3u8_formats[0].update({
2282 'format_id': 'hls-%d' % (m3u8_count if bitrate is None else bitrate),
2287 formats.extend(m3u8_formats)
2288 elif src_ext == 'f4m':
2293 'plugin': 'flowplayer-3.2.0.1',
2295 f4m_url += '&' if '?' in f4m_url else '?'
2296 f4m_url += compat_urllib_parse_urlencode(f4m_params)
2297 formats.extend(self._extract_f4m_formats(f4m_url, video_id, f4m_id='hds', fatal=False))
2298 elif src_ext == 'mpd':
2299 formats.extend(self._extract_mpd_formats(
2300 src_url, video_id, mpd_id='dash', fatal=False))
2301 elif re.search(r'\.ism/[Mm]anifest', src_url):
2302 formats.extend(self._extract_ism_formats(
2303 src_url, video_id, ism_id='mss', fatal=False))
2304 elif src_url.startswith('http') and self._is_valid_url(src, video_id):
2308 'ext': ext or src_ext or 'flv',
2309 'format_id': 'http-%d' % (bitrate or http_count),
2311 'filesize': filesize,
2318 def _parse_smil_subtitles(self, smil, namespace=None, subtitles_lang='en'):
2321 for num, textstream in enumerate(smil.findall(self._xpath_ns('.//textstream', namespace))):
2322 src = textstream.get('src')
2323 if not src or src in urls:
2326 ext = textstream.get('ext') or mimetype2ext(textstream.get('type')) or determine_ext(src)
2327 lang = textstream.get('systemLanguage') or textstream.get('systemLanguageName') or textstream.get('lang') or subtitles_lang
2328 subtitles.setdefault(lang, []).append({
2334 def _extract_xspf_playlist(self, xspf_url, playlist_id, fatal=True):
2335 xspf = self._download_xml(
2336 xspf_url, playlist_id, 'Downloading xpsf playlist',
2337 'Unable to download xspf manifest', fatal=fatal)
2340 return self._parse_xspf(
2341 xspf, playlist_id, xspf_url=xspf_url,
2342 xspf_base_url=base_url(xspf_url))
2344 def _parse_xspf(self, xspf_doc, playlist_id, xspf_url=None, xspf_base_url=None):
2346 'xspf': 'http://xspf.org/ns/0/',
2347 's1': 'http://static.streamone.nl/player/ns/0',
2351 for track in xspf_doc.findall(xpath_with_ns('./xspf:trackList/xspf:track', NS_MAP)):
2353 track, xpath_with_ns('./xspf:title', NS_MAP), 'title', default=playlist_id)
2354 description = xpath_text(
2355 track, xpath_with_ns('./xspf:annotation', NS_MAP), 'description')
2356 thumbnail = xpath_text(
2357 track, xpath_with_ns('./xspf:image', NS_MAP), 'thumbnail')
2358 duration = float_or_none(
2359 xpath_text(track, xpath_with_ns('./xspf:duration', NS_MAP), 'duration'), 1000)
2362 for location in track.findall(xpath_with_ns('./xspf:location', NS_MAP)):
2363 format_url = urljoin(xspf_base_url, location.text)
2368 'manifest_url': xspf_url,
2369 'format_id': location.get(xpath_with_ns('s1:label', NS_MAP)),
2370 'width': int_or_none(location.get(xpath_with_ns('s1:width', NS_MAP))),
2371 'height': int_or_none(location.get(xpath_with_ns('s1:height', NS_MAP))),
2373 self._sort_formats(formats)
2378 'description': description,
2379 'thumbnail': thumbnail,
2380 'duration': duration,
2385 def _extract_mpd_formats(self, mpd_url, video_id, mpd_id=None, note=None, errnote=None, fatal=True, data=None, headers={}, query={}):
2386 res = self._download_xml_handle(
2388 note=note or 'Downloading MPD manifest',
2389 errnote=errnote or 'Failed to download MPD manifest',
2390 fatal=fatal, data=data, headers=headers, query=query)
2396 mpd_base_url = base_url(urlh.geturl())
2398 return self._parse_mpd_formats(
2399 mpd_doc, mpd_id, mpd_base_url, mpd_url)
2401 def _parse_mpd_formats(self, mpd_doc, mpd_id=None, mpd_base_url='', mpd_url=None):
2403 Parse formats
from MPD manifest
.
2405 1. MPEG
-DASH Standard
, ISO
/IEC
23009-1:2014(E
),
2406 http
://standards
.iso
.org
/ittf
/PubliclyAvailableStandards
/c065274_ISO_IEC_23009
-1_2014.zip
2407 2. https
://en
.wikipedia
.org
/wiki
/Dynamic_Adaptive_Streaming_over_HTTP
2409 if not self._downloader.params.get('dynamic_mpd'):
2410 if mpd_doc.get('type') == 'dynamic':
2413 namespace = self._search_regex(r'(?i)^{([^}]+)?}MPD$', mpd_doc.tag, 'namespace', default=None)
2416 return self._xpath_ns(path, namespace)
2418 def is_drm_protected(element):
2419 return element.find(_add_ns('ContentProtection')) is not None
2421 def extract_multisegment_info(element, ms_parent_info):
2422 ms_info = ms_parent_info.copy()
2424 # As per [1, 5.3.9.2.2] SegmentList and SegmentTemplate share some
2425 # common attributes and elements. We will only extract relevant
2427 def extract_common(source):
2428 segment_timeline = source.find(_add_ns('SegmentTimeline'))
2429 if segment_timeline is not None:
2430 s_e = segment_timeline.findall(_add_ns('S'))
2432 ms_info['total_number'] = 0
2435 r = int(s.get('r', 0))
2436 ms_info['total_number'] += 1 + r
2437 ms_info['s'].append({
2438 't': int(s.get('t', 0)),
2439 # @d is mandatory (see [1, 5.3.9.6.2, Table 17, page 60])
2440 'd': int(s.attrib['d']),
2443 start_number = source.get('startNumber')
2445 ms_info['start_number'] = int(start_number)
2446 timescale = source.get('timescale')
2448 ms_info['timescale'] = int(timescale)
2449 segment_duration = source.get('duration')
2450 if segment_duration:
2451 ms_info['segment_duration'] = float(segment_duration)
2453 def extract_Initialization(source):
2454 initialization = source.find(_add_ns('Initialization'))
2455 if initialization is not None:
2456 ms_info['initialization_url'] = initialization.attrib['sourceURL']
2458 segment_list = element.find(_add_ns('SegmentList'))
2459 if segment_list is not None:
2460 extract_common(segment_list)
2461 extract_Initialization(segment_list)
2462 segment_urls_e = segment_list.findall(_add_ns('SegmentURL'))
2464 ms_info['segment_urls'] = [segment.attrib['media'] for segment in segment_urls_e]
2466 segment_template = element.find(_add_ns('SegmentTemplate'))
2467 if segment_template is not None:
2468 extract_common(segment_template)
2469 media = segment_template.get('media')
2471 ms_info['media'] = media
2472 initialization = segment_template.get('initialization')
2474 ms_info['initialization'] = initialization
2476 extract_Initialization(segment_template)
2479 skip_unplayable = not self._downloader.params.get('allow_unplayable_formats')
2481 mpd_duration = parse_duration(mpd_doc.get('mediaPresentationDuration'))
2483 for period in mpd_doc.findall(_add_ns('Period')):
2484 period_duration = parse_duration(period.get('duration')) or mpd_duration
2485 period_ms_info = extract_multisegment_info(period, {
2489 for adaptation_set in period.findall(_add_ns('AdaptationSet')):
2490 if skip_unplayable and is_drm_protected(adaptation_set):
2492 adaption_set_ms_info = extract_multisegment_info(adaptation_set, period_ms_info)
2493 for representation in adaptation_set.findall(_add_ns('Representation')):
2494 if skip_unplayable and is_drm_protected(representation):
2496 representation_attrib = adaptation_set.attrib.copy()
2497 representation_attrib.update(representation.attrib)
2498 # According to [1, 5.3.7.2, Table 9, page 41], @mimeType is mandatory
2499 mime_type = representation_attrib['mimeType']
2500 content_type = mime_type.split('/')[0]
2501 if content_type == 'text':
2502 # TODO implement WebVTT downloading
2504 elif content_type in ('video', 'audio'):
2506 for element in (representation, adaptation_set, period, mpd_doc):
2507 base_url_e = element.find(_add_ns('BaseURL'))
2508 if base_url_e is not None:
2509 base_url = base_url_e.text + base_url
2510 if re.match(r'^https?://', base_url):
2512 if mpd_base_url and not re.match(r'^https?://', base_url):
2513 if not mpd_base_url.endswith('/') and not base_url.startswith('/'):
2515 base_url = mpd_base_url + base_url
2516 representation_id = representation_attrib.get('id')
2517 lang = representation_attrib.get('lang')
2518 url_el = representation.find(_add_ns('BaseURL'))
2519 filesize = int_or_none(url_el.attrib.get('{http://youtube.com/yt/2012/10/10}contentLength') if url_el is not None else None)
2520 bandwidth = int_or_none(representation_attrib.get('bandwidth'))
2522 'format_id': '%s-%s' % (mpd_id, representation_id) if mpd_id else representation_id,
2523 'manifest_url': mpd_url,
2524 'ext': mimetype2ext(mime_type),
2525 'width': int_or_none(representation_attrib.get('width')),
2526 'height': int_or_none(representation_attrib.get('height')),
2527 'tbr': float_or_none(bandwidth, 1000),
2528 'asr': int_or_none(representation_attrib.get('audioSamplingRate')),
2529 'fps': int_or_none(representation_attrib.get('frameRate')),
2530 'language': lang if lang not in ('mul', 'und', 'zxx', 'mis') else None,
2531 'format_note': 'DASH %s' % content_type,
2532 'filesize': filesize,
2533 'container': mimetype2ext(mime_type) + '_dash',
2535 f.update(parse_codecs(representation_attrib.get('codecs')))
2536 representation_ms_info = extract_multisegment_info(representation, adaption_set_ms_info)
2538 def prepare_template(template_name, identifiers):
2539 tmpl = representation_ms_info[template_name]
2540 # First of, % characters outside $...$ templates
2541 # must be escaped by doubling for proper processing
2542 # by % operator string formatting used further (see
2543 # https://github.com/ytdl-org/youtube-dl/issues/16867).
2549 in_template = not in_template
2550 elif c == '%' and not in_template:
2552 # Next, $...$ templates are translated to their
2553 # %(...) counterparts to be used with % operator
2554 t = t.replace('$RepresentationID$', representation_id)
2555 t = re.sub(r'\$(%s)\$' % '|'.join(identifiers), r'%(\1)d', t)
2556 t = re.sub(r'\$(%s)%%([^$]+)\$' % '|'.join(identifiers), r'%(\1)\2', t)
2557 t.replace('$$', '$')
2560 # @initialization is a regular template like @media one
2561 # so it should be handled just the same way (see
2562 # https://github.com/ytdl-org/youtube-dl/issues/11605)
2563 if 'initialization' in representation_ms_info:
2564 initialization_template = prepare_template(
2566 # As per [1, 5.3.9.4.2, Table 15, page 54] $Number$ and
2567 # $Time$ shall not be included for @initialization thus
2568 # only $Bandwidth$ remains
2570 representation_ms_info['initialization_url'] = initialization_template % {
2571 'Bandwidth': bandwidth,
2574 def location_key(location):
2575 return 'url' if re.match(r'^https?://', location) else 'path'
2577 if 'segment_urls' not in representation_ms_info and 'media' in representation_ms_info:
2579 media_template = prepare_template('media', ('Number', 'Bandwidth', 'Time'))
2580 media_location_key = location_key(media_template)
2582 # As per [1, 5.3.9.4.4, Table 16, page 55] $Number$ and $Time$
2583 # can't be used at the same time
2584 if '%(Number' in media_template and 's' not in representation_ms_info:
2585 segment_duration = None
2586 if 'total_number' not in representation_ms_info and 'segment_duration' in representation_ms_info:
2587 segment_duration = float_or_none(representation_ms_info['segment_duration'], representation_ms_info['timescale'])
2588 representation_ms_info['total_number'] = int(math.ceil(float(period_duration) / segment_duration))
2589 representation_ms_info['fragments'] = [{
2590 media_location_key: media_template % {
2591 'Number': segment_number,
2592 'Bandwidth': bandwidth,
2594 'duration': segment_duration,
2595 } for segment_number in range(
2596 representation_ms_info['start_number'],
2597 representation_ms_info['total_number'] + representation_ms_info['start_number'])]
2599 # $Number*$ or $Time$ in media template with S list available
2600 # Example $Number*$: http://www.svtplay.se/klipp/9023742/stopptid-om-bjorn-borg
2601 # Example $Time$: https://play.arkena.com/embed/avp/v2/player/media/b41dda37-d8e7-4d3f-b1b5-9a9db578bdfe/1/129411
2602 representation_ms_info['fragments'] = []
2605 segment_number = representation_ms_info['start_number']
2607 def add_segment_url():
2608 segment_url = media_template % {
2609 'Time': segment_time,
2610 'Bandwidth': bandwidth,
2611 'Number': segment_number,
2613 representation_ms_info['fragments'].append({
2614 media_location_key: segment_url,
2615 'duration': float_or_none(segment_d, representation_ms_info['timescale']),
2618 for num, s in enumerate(representation_ms_info['s']):
2619 segment_time = s.get('t') or segment_time
2623 for r in range(s.get('r', 0)):
2624 segment_time += segment_d
2627 segment_time += segment_d
2628 elif 'segment_urls' in representation_ms_info and 's' in representation_ms_info:
2630 # Example: https://www.youtube.com/watch?v=iXZV5uAYMJI
2631 # or any YouTube dashsegments video
2634 timescale = representation_ms_info['timescale']
2635 for s in representation_ms_info['s']:
2636 duration = float_or_none(s['d'], timescale)
2637 for r in range(s.get('r', 0) + 1):
2638 segment_uri = representation_ms_info['segment_urls'][segment_index]
2640 location_key(segment_uri): segment_uri,
2641 'duration': duration,
2644 representation_ms_info['fragments'] = fragments
2645 elif 'segment_urls' in representation_ms_info:
2646 # Segment URLs with no SegmentTimeline
2647 # Example: https://www.seznam.cz/zpravy/clanek/cesko-zasahne-vitr-o-sile-vichrice-muze-byt-i-zivotu-nebezpecny-39091
2648 # https://github.com/ytdl-org/youtube-dl/pull/14844
2650 segment_duration = float_or_none(
2651 representation_ms_info['segment_duration'],
2652 representation_ms_info['timescale']) if 'segment_duration' in representation_ms_info else None
2653 for segment_url in representation_ms_info['segment_urls']:
2655 location_key(segment_url): segment_url,
2657 if segment_duration:
2658 fragment['duration'] = segment_duration
2659 fragments.append(fragment)
2660 representation_ms_info['fragments'] = fragments
2661 # If there is a fragments key available then we correctly recognized fragmented media.
2662 # Otherwise we will assume unfragmented media with direct access. Technically, such
2663 # assumption is not necessarily correct since we may simply have no support for
2664 # some forms of fragmented media renditions yet, but for now we'll use this fallback.
2665 if 'fragments' in representation_ms_info:
2667 # NB: mpd_url may be empty when MPD manifest is parsed from a string
2668 'url': mpd_url or base_url,
2669 'fragment_base_url': base_url,
2671 'protocol': 'http_dash_segments',
2673 if 'initialization_url' in representation_ms_info:
2674 initialization_url = representation_ms_info['initialization_url']
2675 if not f.get('url'):
2676 f['url'] = initialization_url
2677 f['fragments'].append({location_key(initialization_url): initialization_url})
2678 f['fragments'].extend(representation_ms_info['fragments'])
2680 # Assuming direct URL to unfragmented media.
2684 self.report_warning('Unknown MIME type %s in DASH manifest' % mime_type)
2687 def _extract_ism_formats(self, ism_url, video_id, ism_id=None, note=None, errnote=None, fatal=True, data=None, headers={}, query={}):
2688 res = self._download_xml_handle(
2690 note=note or 'Downloading ISM manifest',
2691 errnote=errnote or 'Failed to download ISM manifest',
2692 fatal=fatal, data=data, headers=headers, query=query)
2699 return self._parse_ism_formats(ism_doc, urlh.geturl(), ism_id)
2701 def _parse_ism_formats(self, ism_doc, ism_url, ism_id=None):
2703 Parse formats
from ISM manifest
.
2705 1. [MS
-SSTR
]: Smooth Streaming Protocol
,
2706 https
://msdn
.microsoft
.com
/en
-us
/library
/ff469518
.aspx
2708 if ism_doc.get('IsLive') == 'TRUE':
2710 if (not self._downloader.params.get('allow_unplayable_formats')
2711 and ism_doc.find('Protection') is not None):
2714 duration = int(ism_doc.attrib['Duration'])
2715 timescale = int_or_none(ism_doc.get('TimeScale')) or 10000000
2718 for stream in ism_doc.findall('StreamIndex'):
2719 stream_type = stream.get('Type')
2720 if stream_type not in ('video', 'audio'):
2722 url_pattern = stream.attrib['Url']
2723 stream_timescale = int_or_none(stream.get('TimeScale')) or timescale
2724 stream_name = stream.get('Name')
2725 for track in stream.findall('QualityLevel'):
2726 fourcc = track.get('FourCC', 'AACL' if track.get('AudioTag') == '255' else None)
2727 # TODO: add support for WVC1 and WMAP
2728 if fourcc not in ('H264', 'AVC1', 'AACL'):
2729 self.report_warning('%s is not a supported codec' % fourcc)
2731 tbr = int(track.attrib['Bitrate']) // 1000
2732 # [1] does not mention Width and Height attributes. However,
2733 # they're often present while MaxWidth and MaxHeight are
2734 # missing, so should be used as fallbacks
2735 width = int_or_none(track.get('MaxWidth') or track.get('Width'))
2736 height = int_or_none(track.get('MaxHeight') or track.get('Height'))
2737 sampling_rate = int_or_none(track.get('SamplingRate'))
2739 track_url_pattern = re.sub(r'{[Bb]itrate}', track.attrib['Bitrate'], url_pattern)
2740 track_url_pattern = compat_urlparse.urljoin(ism_url, track_url_pattern)
2746 stream_fragments = stream.findall('c')
2747 for stream_fragment_index, stream_fragment in enumerate(stream_fragments):
2748 fragment_ctx['time'] = int_or_none(stream_fragment.get('t')) or fragment_ctx['time']
2749 fragment_repeat = int_or_none(stream_fragment.get('r')) or 1
2750 fragment_ctx['duration'] = int_or_none(stream_fragment.get('d'))
2751 if not fragment_ctx['duration']:
2753 next_fragment_time = int(stream_fragment[stream_fragment_index + 1].attrib['t'])
2755 next_fragment_time = duration
2756 fragment_ctx['duration'] = (next_fragment_time - fragment_ctx['time']) / fragment_repeat
2757 for _ in range(fragment_repeat):
2759 'url': re.sub(r'{start[ _]time}', compat_str(fragment_ctx['time']), track_url_pattern),
2760 'duration': fragment_ctx['duration'] / stream_timescale,
2762 fragment_ctx['time'] += fragment_ctx['duration']
2766 format_id.append(ism_id)
2768 format_id.append(stream_name)
2769 format_id.append(compat_str(tbr))
2772 'format_id': '-'.join(format_id),
2774 'manifest_url': ism_url,
2775 'ext': 'ismv' if stream_type == 'video' else 'isma',
2779 'asr': sampling_rate,
2780 'vcodec': 'none' if stream_type == 'audio' else fourcc,
2781 'acodec': 'none' if stream_type == 'video' else fourcc,
2783 'fragments': fragments,
2784 '_download_params': {
2785 'duration': duration,
2786 'timescale': stream_timescale,
2787 'width': width or 0,
2788 'height': height or 0,
2790 'codec_private_data': track.get('CodecPrivateData'),
2791 'sampling_rate': sampling_rate,
2792 'channels': int_or_none(track.get('Channels', 2)),
2793 'bits_per_sample': int_or_none(track.get('BitsPerSample', 16)),
2794 'nal_unit_length_field': int_or_none(track.get('NALUnitLengthField', 4)),
2799 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):
2800 def absolute_url(item_url):
2801 return urljoin(base_url, item_url)
2803 def parse_content_type(content_type):
2804 if not content_type:
2806 ctr = re.search(r'(?P<mimetype>[^/]+/[^;]+)(?:;\s*codecs="?(?P<codecs>[^"]+))?', content_type)
2808 mimetype, codecs = ctr.groups()
2809 f = parse_codecs(codecs)
2810 f['ext'] = mimetype2ext(mimetype)
2814 def _media_formats(src, cur_media_type, type_info={}):
2815 full_url = absolute_url(src)
2816 ext = type_info.get('ext') or determine_ext(full_url)
2818 is_plain_url = False
2819 formats = self._extract_m3u8_formats(
2820 full_url, video_id, ext='mp4',
2821 entry_protocol=m3u8_entry_protocol, m3u8_id=m3u8_id,
2822 preference=preference, quality=quality, fatal=False)
2824 is_plain_url = False
2825 formats = self._extract_mpd_formats(
2826 full_url, video_id, mpd_id=mpd_id, fatal=False)
2831 'vcodec': 'none' if cur_media_type == 'audio' else None,
2833 return is_plain_url, formats
2836 # amp-video and amp-audio are very similar to their HTML5 counterparts
2837 # so we wll include them right here (see
2838 # https://www.ampproject.org/docs/reference/components/amp-video)
2839 # For dl8-* tags see https://delight-vr.com/documentation/dl8-video/
2840 _MEDIA_TAG_NAME_RE = r'(?:(?:amp|dl8(?:-live)?)-)?(video|audio)'
2841 media_tags = [(media_tag, media_tag_name, media_type, '')
2842 for media_tag, media_tag_name, media_type
2843 in re.findall(r'(?s)(<(%s)[^>]*/>)' % _MEDIA_TAG_NAME_RE, webpage)]
2844 media_tags.extend(re.findall(
2845 # We only allow video|audio followed by a whitespace or '>'.
2846 # Allowing more characters may end up in significant slow down (see
2847 # https://github.com/ytdl-org/youtube-dl/issues/11979, example URL:
2848 # http://www.porntrex.com/maps/videositemap.xml).
2849 r'(?s)(<(?P<tag>%s)(?:\s+[^>]*)?>)(.*?)</(?P=tag)>' % _MEDIA_TAG_NAME_RE, webpage))
2850 for media_tag, _, media_type, media_content in media_tags:
2855 media_attributes = extract_attributes(media_tag)
2856 src = strip_or_none(media_attributes.get('src'))
2858 _, formats = _media_formats(src, media_type)
2859 media_info['formats'].extend(formats)
2860 media_info['thumbnail'] = absolute_url(media_attributes.get('poster'))
2862 for source_tag in re.findall(r'<source[^>]+>', media_content):
2863 s_attr = extract_attributes(source_tag)
2864 # data-video-src and data-src are non standard but seen
2865 # several times in the wild
2866 src = strip_or_none(dict_get(s_attr, ('src', 'data-video-src', 'data-src')))
2869 f = parse_content_type(s_attr.get('type'))
2870 is_plain_url, formats = _media_formats(src, media_type, f)
2872 # width, height, res, label and title attributes are
2873 # all not standard but seen several times in the wild
2876 for lbl in ('label', 'title')
2877 if str_or_none(s_attr.get(lbl))
2879 width = int_or_none(s_attr.get('width'))
2880 height = (int_or_none(s_attr.get('height'))
2881 or int_or_none(s_attr.get('res')))
2882 if not width or not height:
2884 resolution = parse_resolution(lbl)
2887 width = width or resolution.get('width')
2888 height = height or resolution.get('height')
2890 tbr = parse_bitrate(lbl)
2899 'format_id': s_attr.get('label') or s_attr.get('title'),
2901 f.update(formats[0])
2902 media_info['formats'].append(f)
2904 media_info['formats'].extend(formats)
2905 for track_tag in re.findall(r'<track[^>]+>', media_content):
2906 track_attributes = extract_attributes(track_tag)
2907 kind = track_attributes.get('kind')
2908 if not kind or kind in ('subtitles', 'captions'):
2909 src = strip_or_none(track_attributes.get('src'))
2912 lang = track_attributes.get('srclang') or track_attributes.get('lang') or track_attributes.get('label')
2913 media_info['subtitles'].setdefault(lang, []).append({
2914 'url': absolute_url(src),
2916 for f in media_info['formats']:
2917 f.setdefault('http_headers', {})['Referer'] = base_url
2918 if media_info['formats'] or media_info['subtitles']:
2919 entries.append(media_info)
2922 def _extract_akamai_formats(self, manifest_url, video_id, hosts={}):
2923 signed = 'hdnea=' in manifest_url
2925 # https://learn.akamai.com/en-us/webhelp/media-services-on-demand/stream-packaging-user-guide/GUID-BE6C0F73-1E06-483B-B0EA-57984B91B7F9.html
2926 manifest_url = re.sub(
2927 r'(?:b=[\d,-]+|(?:__a__|attributes)=off|__b__=\d+)&?',
2928 '', manifest_url).strip('?')
2932 hdcore_sign = 'hdcore=3.7.0'
2933 f4m_url = re.sub(r'(https?://[^/]+)/i/', r'\1/z/', manifest_url).replace('/master.m3u8', '/manifest.f4m')
2934 hds_host = hosts.get('hds')
2936 f4m_url = re.sub(r'(https?://)[^/]+', r'\1' + hds_host, f4m_url)
2937 if 'hdcore=' not in f4m_url:
2938 f4m_url += ('&' if '?' in f4m_url else '?') + hdcore_sign
2939 f4m_formats = self._extract_f4m_formats(
2940 f4m_url, video_id, f4m_id='hds', fatal=False)
2941 for entry in f4m_formats:
2942 entry.update({'extra_param_to_segment_url': hdcore_sign})
2943 formats.extend(f4m_formats)
2945 m3u8_url = re.sub(r'(https?://[^/]+)/z/', r'\1/i/', manifest_url).replace('/manifest.f4m', '/master.m3u8')
2946 hls_host = hosts.get('hls')
2948 m3u8_url = re.sub(r'(https?://)[^/]+', r'\1' + hls_host, m3u8_url)
2949 m3u8_formats = self._extract_m3u8_formats(
2950 m3u8_url, video_id, 'mp4', 'm3u8_native',
2951 m3u8_id='hls', fatal=False)
2952 formats.extend(m3u8_formats)
2954 http_host = hosts.get('http')
2955 if http_host and m3u8_formats and not signed:
2956 REPL_REGEX = r'https?://[^/]+/i/([^,]+),([^/]+),([^/]+)\.csmil/.+'
2957 qualities = re.match(REPL_REGEX, m3u8_url).group(2).split(',')
2958 qualities_length = len(qualities)
2959 if len(m3u8_formats) in (qualities_length, qualities_length + 1):
2961 for f in m3u8_formats:
2962 if f['vcodec'] != 'none':
2963 for protocol in ('http', 'https'):
2965 del http_f['manifest_url']
2967 REPL_REGEX, protocol + r'://%s/\g<1>%s\3' % (http_host, qualities[i]), f['url'])
2969 'format_id': http_f['format_id'].replace('hls-', protocol + '-'),
2971 'protocol': protocol,
2973 formats.append(http_f)
2978 def _extract_wowza_formats(self, url, video_id, m3u8_entry_protocol='m3u8_native', skip_protocols=[]):
2979 query = compat_urlparse.urlparse(url).query
2980 url = re.sub(r'/(?:manifest|playlist|jwplayer)\.(?:m3u8|f4m|mpd|smil)', '', url)
2982 r'(?:(?:http|rtmp|rtsp)(?P<s>s)?:)?(?P<url>//[^?]+)', url)
2983 url_base = mobj.group('url')
2984 http_base_url = '%s%s:%s' % ('http', mobj.group('s') or '', url_base)
2987 def manifest_url(manifest):
2988 m_url = '%s/%s' % (http_base_url, manifest)
2990 m_url += '?%s' % query
2993 if 'm3u8' not in skip_protocols:
2994 formats.extend(self._extract_m3u8_formats(
2995 manifest_url('playlist.m3u8'), video_id, 'mp4',
2996 m3u8_entry_protocol, m3u8_id='hls', fatal=False))
2997 if 'f4m' not in skip_protocols:
2998 formats.extend(self._extract_f4m_formats(
2999 manifest_url('manifest.f4m'),
3000 video_id, f4m_id='hds', fatal=False))
3001 if 'dash' not in skip_protocols:
3002 formats.extend(self._extract_mpd_formats(
3003 manifest_url('manifest.mpd'),
3004 video_id, mpd_id='dash', fatal=False))
3005 if re.search(r'(?:/smil:|\.smil)', url_base):
3006 if 'smil' not in skip_protocols:
3007 rtmp_formats = self._extract_smil_formats(
3008 manifest_url('jwplayer.smil'),
3009 video_id, fatal=False)
3010 for rtmp_format in rtmp_formats:
3011 rtsp_format = rtmp_format.copy()
3012 rtsp_format['url'] = '%s/%s' % (rtmp_format['url'], rtmp_format['play_path'])
3013 del rtsp_format['play_path']
3014 del rtsp_format['ext']
3015 rtsp_format.update({
3016 'url': rtsp_format['url'].replace('rtmp://', 'rtsp://'),
3017 'format_id': rtmp_format['format_id'].replace('rtmp', 'rtsp'),
3020 formats.extend([rtmp_format, rtsp_format])
3022 for protocol in ('rtmp', 'rtsp'):
3023 if protocol not in skip_protocols:
3025 'url': '%s:%s' % (protocol, url_base),
3026 'format_id': protocol,
3027 'protocol': protocol,
3031 def _find_jwplayer_data(self, webpage, video_id=None, transform_source=js_to_json):
3033 r'(?s)jwplayer\((?P<quote>[\'"])[^\'" ]+(?P=quote)\)(?!</script>).*?\.setup\s*\((?P<options>[^)]+)\)',
3037 jwplayer_data = self._parse_json(mobj.group('options'),
3039 transform_source=transform_source)
3040 except ExtractorError:
3043 if isinstance(jwplayer_data, dict):
3044 return jwplayer_data
3046 def _extract_jwplayer_data(self, webpage, video_id, *args, **kwargs):
3047 jwplayer_data = self._find_jwplayer_data(
3048 webpage, video_id, transform_source=js_to_json)
3049 return self._parse_jwplayer_data(
3050 jwplayer_data, video_id, *args, **kwargs)
3052 def _parse_jwplayer_data(self, jwplayer_data, video_id=None, require_title=True,
3053 m3u8_id=None, mpd_id=None, rtmp_params=None, base_url=None):
3054 # JWPlayer backward compatibility: flattened playlists
3055 # https://github.com/jwplayer/jwplayer/blob/v7.4.3/src/js/api/config.js#L81-L96
3056 if 'playlist' not in jwplayer_data:
3057 jwplayer_data = {'playlist': [jwplayer_data]}
3061 # JWPlayer backward compatibility: single playlist item
3062 # https://github.com/jwplayer/jwplayer/blob/v7.7.0/src/js/playlist/playlist.js#L10
3063 if not isinstance(jwplayer_data['playlist'], list):
3064 jwplayer_data['playlist'] = [jwplayer_data['playlist']]
3066 for video_data in jwplayer_data['playlist']:
3067 # JWPlayer backward compatibility: flattened sources
3068 # https://github.com/jwplayer/jwplayer/blob/v7.4.3/src/js/playlist/item.js#L29-L35
3069 if 'sources' not in video_data:
3070 video_data['sources'] = [video_data]
3072 this_video_id = video_id or video_data['mediaid']
3074 formats = self._parse_jwplayer_formats(
3075 video_data['sources'], video_id=this_video_id, m3u8_id=m3u8_id,
3076 mpd_id=mpd_id, rtmp_params=rtmp_params, base_url=base_url)
3079 tracks = video_data.get('tracks')
3080 if tracks and isinstance(tracks, list):
3081 for track in tracks:
3082 if not isinstance(track, dict):
3084 track_kind = track.get('kind')
3085 if not track_kind or not isinstance(track_kind, compat_str):
3087 if track_kind.lower() not in ('captions', 'subtitles'):
3089 track_url = urljoin(base_url, track.get('file'))
3092 subtitles.setdefault(track.get('label') or 'en', []).append({
3093 'url': self._proto_relative_url(track_url)
3097 'id': this_video_id,
3098 'title': unescapeHTML(video_data['title'] if require_title else video_data.get('title')),
3099 'description': clean_html(video_data.get('description')),
3100 'thumbnail': urljoin(base_url, self._proto_relative_url(video_data.get('image'))),
3101 'timestamp': int_or_none(video_data.get('pubdate')),
3102 'duration': float_or_none(jwplayer_data.get('duration') or video_data.get('duration')),
3103 'subtitles': subtitles,
3105 # https://github.com/jwplayer/jwplayer/blob/master/src/js/utils/validator.js#L32
3106 if len(formats) == 1 and re.search(r'^(?:http|//).*(?:youtube\.com|youtu\.be)/.+', formats[0]['url']):
3108 '_type': 'url_transparent',
3109 'url': formats[0]['url'],
3112 self._sort_formats(formats)
3113 entry['formats'] = formats
3114 entries.append(entry)
3115 if len(entries) == 1:
3118 return self.playlist_result(entries)
3120 def _parse_jwplayer_formats(self, jwplayer_sources_data, video_id=None,
3121 m3u8_id=None, mpd_id=None, rtmp_params=None, base_url=None):
3124 for source in jwplayer_sources_data:
3125 if not isinstance(source, dict):
3127 source_url = urljoin(
3128 base_url, self._proto_relative_url(source.get('file')))
3129 if not source_url or source_url in urls:
3131 urls.append(source_url)
3132 source_type = source.get('type') or ''
3133 ext = mimetype2ext(source_type) or determine_ext(source_url)
3134 if source_type == 'hls' or ext == 'm3u8':
3135 formats.extend(self._extract_m3u8_formats(
3136 source_url, video_id, 'mp4', entry_protocol='m3u8_native',
3137 m3u8_id=m3u8_id, fatal=False))
3138 elif source_type == 'dash' or ext == 'mpd':
3139 formats.extend(self._extract_mpd_formats(
3140 source_url, video_id, mpd_id=mpd_id, fatal=False))
3142 formats.extend(self._extract_smil_formats(
3143 source_url, video_id, fatal=False))
3144 # https://github.com/jwplayer/jwplayer/blob/master/src/js/providers/default.js#L67
3145 elif source_type.startswith('audio') or ext in (
3146 'oga', 'aac', 'mp3', 'mpeg', 'vorbis'):
3153 height = int_or_none(source.get('height'))
3155 # Often no height is provided but there is a label in
3156 # format like "1080p", "720p SD", or 1080.
3157 height = int_or_none(self._search_regex(
3158 r'^(\d{3,4})[pP]?(?:\b|$)', compat_str(source.get('label') or ''),
3159 'height', default=None))
3162 'width': int_or_none(source.get('width')),
3164 'tbr': int_or_none(source.get('bitrate')),
3167 if source_url.startswith('rtmp'):
3168 a_format['ext'] = 'flv'
3169 # See com/longtailvideo/jwplayer/media/RTMPMediaProvider.as
3170 # of jwplayer.flash.swf
3171 rtmp_url_parts = re.split(
3172 r'((?:mp4|mp3|flv):)', source_url, 1)
3173 if len(rtmp_url_parts) == 3:
3174 rtmp_url, prefix, play_path = rtmp_url_parts
3177 'play_path': prefix + play_path,
3180 a_format.update(rtmp_params)
3181 formats.append(a_format)
3184 def _live_title(self, name):
3185 """ Generate the title
for a live video
"""
3186 now = datetime.datetime.now()
3187 now_str = now.strftime('%Y-%m-%d %H:%M')
3188 return name + ' ' + now_str
3190 def _int(self, v, name, fatal=False, **kwargs):
3191 res = int_or_none(v, **kwargs)
3192 if 'get_attr' in kwargs:
3193 print(getattr(v, kwargs['get_attr']))
3195 msg = 'Failed to extract %s: Could not parse value %r' % (name, v)
3197 raise ExtractorError(msg)
3199 self._downloader.report_warning(msg)
3202 def _float(self, v, name, fatal=False, **kwargs):
3203 res = float_or_none(v, **kwargs)
3205 msg = 'Failed to extract %s: Could not parse value %r' % (name, v)
3207 raise ExtractorError(msg)
3209 self._downloader.report_warning(msg)
3212 def _set_cookie(self, domain, name, value, expire_time=None, port=None,
3213 path='/', secure=False, discard=False, rest={}, **kwargs):
3214 cookie = compat_cookiejar_Cookie(
3215 0, name, value, port, port is not None, domain, True,
3216 domain.startswith('.'), path, True, secure, expire_time,
3217 discard, None, None, rest)
3218 self._downloader.cookiejar.set_cookie(cookie)
3220 def _get_cookies(self, url):
3221 """ Return a compat_cookies
.SimpleCookie
with the cookies
for the url
"""
3222 req = sanitized_Request(url)
3223 self._downloader.cookiejar.add_cookie_header(req)
3224 return compat_cookies.SimpleCookie(req.get_header('Cookie'))
3226 def _apply_first_set_cookie_header(self, url_handle, cookie):
3228 Apply first Set
-Cookie header instead of the last
. Experimental
.
3230 Some
sites (e
.g
. [1-3]) may serve two cookies under the same name
3231 in Set
-Cookie header
and expect the
first (old
) one to be
set rather
3232 than
second (new
). However
, as of RFC6265 the newer one cookie
3233 should be
set into cookie store what actually happens
.
3234 We will workaround this issue by resetting the cookie to
3235 the first one manually
.
3236 1. https
://new
.vk
.com
/
3237 2. https
://github
.com
/ytdl
-org
/youtube
-dl
/issues
/9841#issuecomment-227871201
3238 3. https
://learning
.oreilly
.com
/
3240 for header, cookies in url_handle.headers.items():
3241 if header.lower() != 'set-cookie':
3243 if sys.version_info[0] >= 3:
3244 cookies = cookies.encode('iso-8859-1')
3245 cookies = cookies.decode('utf-8')
3246 cookie_value = re.search(
3247 r'%s=(.+?);.*?\b[Dd]omain=(.+?)(?:[,;]|$)' % cookie, cookies)
3249 value, domain = cookie_value.groups()
3250 self._set_cookie(domain, cookie, value)
3253 def get_testcases(self, include_onlymatching=False):
3254 t = getattr(self, '_TEST', None)
3256 assert not hasattr(self, '_TESTS'), \
3257 '%s has _TEST and _TESTS' % type(self).__name__
3260 tests = getattr(self, '_TESTS', [])
3262 if not include_onlymatching and t.get('only_matching', False):
3264 t['name'] = type(self).__name__[:-len('IE')]
3267 def is_suitable(self, age_limit):
3268 """ Test whether the extractor
is generally suitable
for the given
3269 age
limit (i
.e
. pornographic sites are
not, all others usually are
) """
3271 any_restricted = False
3272 for tc in self.get_testcases(include_onlymatching=False):
3273 if tc.get('playlist', []):
3274 tc = tc['playlist'][0]
3275 is_restricted = age_restricted(
3276 tc.get('info_dict', {}).get('age_limit'), age_limit)
3277 if not is_restricted:
3279 any_restricted = any_restricted or is_restricted
3280 return not any_restricted
3282 def extract_subtitles(self, *args, **kwargs):
3283 if (self._downloader.params.get('writesubtitles', False)
3284 or self._downloader.params.get('listsubtitles')):
3285 return self._get_subtitles(*args, **kwargs)
3288 def _get_subtitles(self, *args, **kwargs):
3289 raise NotImplementedError('This method must be implemented by subclasses')
3292 def _merge_subtitle_items(subtitle_list1, subtitle_list2):
3293 """ Merge subtitle items
for one language
. Items
with duplicated URLs
3294 will be dropped
. """
3295 list1_urls = set([item['url'] for item in subtitle_list1])
3296 ret = list(subtitle_list1)
3297 ret.extend([item for item in subtitle_list2 if item['url'] not in list1_urls])
3301 def _merge_subtitles(cls, subtitle_dict1, subtitle_dict2):
3302 """ Merge two subtitle dictionaries
, language by language
. """
3303 ret = dict(subtitle_dict1)
3304 for lang in subtitle_dict2:
3305 ret[lang] = cls._merge_subtitle_items(subtitle_dict1.get(lang, []), subtitle_dict2[lang])
3308 def extract_automatic_captions(self, *args, **kwargs):
3309 if (self._downloader.params.get('writeautomaticsub', False)
3310 or self._downloader.params.get('listsubtitles')):
3311 return self._get_automatic_captions(*args, **kwargs)
3314 def _get_automatic_captions(self, *args, **kwargs):
3315 raise NotImplementedError('This method must be implemented by subclasses')
3317 def mark_watched(self, *args, **kwargs):
3318 if (self._downloader.params.get('mark_watched', False)
3319 and (self._get_login_info()[0] is not None
3320 or self._downloader.params.get('cookiefile') is not None)):
3321 self._mark_watched(*args, **kwargs)
3323 def _mark_watched(self, *args, **kwargs):
3324 raise NotImplementedError('This method must be implemented by subclasses')
3326 def geo_verification_headers(self):
3328 geo_verification_proxy = self._downloader.params.get('geo_verification_proxy')
3329 if geo_verification_proxy:
3330 headers['Ytdl-request-proxy'] = geo_verification_proxy
3333 def _generic_id(self, url):
3334 return compat_urllib_parse_unquote(os.path.splitext(url.rstrip('/').split('/')[-1])[0])
3336 def _generic_title(self, url):
3337 return compat_urllib_parse_unquote(os.path.splitext(url_basename(url))[0])
3340 def _availability(is_private, needs_premium, needs_subscription, needs_auth, is_unlisted):
3341 all_known = all(map(
3342 lambda x: x is not None,
3343 (is_private, needs_premium, needs_subscription, needs_auth, is_unlisted)))
3345 'private' if is_private
3346 else 'premium_only' if needs_premium
3347 else 'subscriber_only' if needs_subscription
3348 else 'needs_auth' if needs_auth
3349 else 'unlisted' if is_unlisted
3350 else 'public' if all_known
3354 class SearchInfoExtractor(InfoExtractor):
3356 Base
class for paged search queries extractors
.
3357 They accept URLs
in the format
_SEARCH_KEY(|all|
[0-9]):{query}
3358 Instances should define _SEARCH_KEY
and _MAX_RESULTS
.
3362 def _make_valid_url(cls):
3363 return r'%s(?P<prefix>|[1-9][0-9]*|all):(?P<query>[\s\S]+)' % cls._SEARCH_KEY
3366 def suitable(cls, url):
3367 return re.match(cls._make_valid_url(), url) is not None
3369 def _real_extract(self, query):
3370 mobj = re.match(self._make_valid_url(), query)
3372 raise ExtractorError('Invalid search query "%s"' % query)
3374 prefix = mobj.group('prefix')
3375 query = mobj.group('query')
3377 return self._get_n_results(query, 1)
3378 elif prefix == 'all':
3379 return self._get_n_results(query, self._MAX_RESULTS)
3383 raise ExtractorError('invalid download number %s for query "%s"' % (n, query))
3384 elif n > self._MAX_RESULTS:
3385 self._downloader.report_warning('%s returns max %i results (you requested %i)' % (self._SEARCH_KEY, self._MAX_RESULTS, n))
3386 n = self._MAX_RESULTS
3387 return self._get_n_results(query, n)
3389 def _get_n_results(self, query, n):
3390 """Get a specified number of results
for a query
"""
3391 raise NotImplementedError('This method must be implemented by subclasses')
3394 def SEARCH_KEY(self):
3395 return self._SEARCH_KEY