1 from __future__ import unicode_literals
15 from ..compat import (
18 compat_etree_fromstring,
24 compat_urllib_parse_urlencode,
25 compat_urllib_request,
28 from ..downloader.f4m import remove_encrypted_media
58 class InfoExtractor(object):
59 """Information Extractor class.
61 Information extractors are the classes that, given a URL, extract
62 information about the video (or videos) the URL refers to. This
63 information includes the real video URL, the video title, author and
64 others. The information is stored in a dictionary which is then
65 passed to the YoutubeDL. The YoutubeDL processes this
66 information possibly downloading the video to the file system, among
67 other possible outcomes.
69 The type field determines the type of the result.
70 By far the most common value (and the default if _type is missing) is
71 "video", which indicates a single video.
73 For a video, the dictionaries must include the following fields:
76 title: Video title, unescaped.
78 Additionally, it must contain either a formats entry or a url one:
80 formats: A list of dictionaries for each format available, ordered
81 from worst to best quality.
84 * url Mandatory. The URL of the video file
85 * ext Will be calculated from URL if missing
86 * format A human-readable description of the format
87 ("mp4 container with h264/opus").
88 Calculated from the format_id, width, height.
89 and format_note fields if missing.
90 * format_id A short description of the format
91 ("mp4_h264_opus" or "19").
92 Technically optional, but strongly recommended.
93 * format_note Additional info about the format
94 ("3D" or "DASH video")
95 * width Width of the video, if known
96 * height Height of the video, if known
97 * resolution Textual description of width and height
98 * tbr Average bitrate of audio and video in KBit/s
99 * abr Average audio bitrate in KBit/s
100 * acodec Name of the audio codec in use
101 * asr Audio sampling rate in Hertz
102 * vbr Average video bitrate in KBit/s
104 * vcodec Name of the video codec in use
105 * container Name of the container format
106 * filesize The number of bytes, if known in advance
107 * filesize_approx An estimate for the number of bytes
108 * player_url SWF Player URL (used for rtmpdump).
109 * protocol The protocol that will be used for the actual
110 download, lower-case.
111 "http", "https", "rtsp", "rtmp", "rtmpe",
112 "m3u8", "m3u8_native" or "http_dash_segments".
113 * preference Order number of this format. If this field is
114 present and not None, the formats get sorted
115 by this field, regardless of all other values.
116 -1 for default (order by other properties),
117 -2 or smaller for less than default.
118 < -1000 to hide the format (if there is
119 another one which is strictly better)
120 * language Language code, e.g. "de" or "en-US".
121 * language_preference Is this in the language mentioned in
123 10 if it's what the URL is about,
124 -1 for default (don't know),
125 -10 otherwise, other values reserved for now.
126 * quality Order number of the video quality of this
127 format, irrespective of the file format.
128 -1 for default (order by other properties),
129 -2 or smaller for less than default.
130 * source_preference Order number for this video source
131 (quality takes higher priority)
132 -1 for default (order by other properties),
133 -2 or smaller for less than default.
134 * http_headers A dictionary of additional HTTP headers
135 to add to the request.
136 * stretched_ratio If given and not 1, indicates that the
137 video's pixels are not square.
138 width : height ratio as float.
139 * no_resume The server does not support resuming the
140 (HTTP or RTMP) download. Boolean.
142 url: Final video URL.
143 ext: Video filename extension.
144 format: The video format, defaults to ext (used for --get-format)
145 player_url: SWF Player URL (used for rtmpdump).
147 The following fields are optional:
149 alt_title: A secondary title of the video.
150 display_id An alternative identifier for the video, not necessarily
151 unique, but available before title. Typically, id is
152 something like "4234987", title "Dancing naked mole rats",
153 and display_id "dancing-naked-mole-rats"
154 thumbnails: A list of dictionaries, with the following entries:
155 * "id" (optional, string) - Thumbnail format ID
157 * "preference" (optional, int) - quality of the image
158 * "width" (optional, int)
159 * "height" (optional, int)
160 * "resolution" (optional, string "{width}x{height"},
162 thumbnail: Full URL to a video thumbnail image.
163 description: Full video description.
164 uploader: Full name of the video uploader.
165 license: License name the video is licensed under.
166 creator: The main artist who created the video.
167 release_date: The date (YYYYMMDD) when the video was released.
168 timestamp: UNIX timestamp of the moment the video became available.
169 upload_date: Video upload date (YYYYMMDD).
170 If not explicitly set, calculated from timestamp.
171 uploader_id: Nickname or id of the video uploader.
172 uploader_url: Full URL to a personal webpage of the video uploader.
173 location: Physical location where the video was filmed.
174 subtitles: The available subtitles as a dictionary in the format
175 {language: subformats}. "subformats" is a list sorted from
176 lower to higher preference, each element is a dictionary
177 with the "ext" entry and one of:
178 * "data": The subtitles file contents
179 * "url": A URL pointing to the subtitles file
180 "ext" will be calculated from URL if missing
181 automatic_captions: Like 'subtitles', used by the YoutubeIE for
182 automatically generated captions
183 duration: Length of the video in seconds, as an integer or float.
184 view_count: How many users have watched the video on the platform.
185 like_count: Number of positive ratings of the video
186 dislike_count: Number of negative ratings of the video
187 repost_count: Number of reposts of the video
188 average_rating: Average rating give by users, the scale used depends on the webpage
189 comment_count: Number of comments on the video
190 comments: A list of comments, each with one or more of the following
191 properties (all but one of text or html optional):
192 * "author" - human-readable name of the comment author
193 * "author_id" - user ID of the comment author
195 * "html" - Comment as HTML
196 * "text" - Plain text of the comment
197 * "timestamp" - UNIX timestamp of comment
198 * "parent" - ID of the comment this one is replying to.
199 Set to "root" to indicate that this is a
200 comment to the original video.
201 age_limit: Age restriction for the video, as an integer (years)
202 webpage_url: The URL to the video webpage, if given to youtube-dl it
203 should allow to get the same result again. (It will be set
204 by YoutubeDL if it's missing)
205 categories: A list of categories that the video falls in, for example
207 tags: A list of tags assigned to the video, e.g. ["sweden", "pop music"]
208 is_live: True, False, or None (=unknown). Whether this video is a
209 live stream that goes on instead of a fixed-length video.
210 start_time: Time in seconds where the reproduction should start, as
211 specified in the URL.
212 end_time: Time in seconds where the reproduction should end, as
213 specified in the URL.
215 The following fields should only be used when the video belongs to some logical
218 chapter: Name or title of the chapter the video belongs to.
219 chapter_number: Number of the chapter the video belongs to, as an integer.
220 chapter_id: Id of the chapter the video belongs to, as a unicode string.
222 The following fields should only be used when the video is an episode of some
225 series: Title of the series or programme the video episode belongs to.
226 season: Title of the season the video episode belongs to.
227 season_number: Number of the season the video episode belongs to, as an integer.
228 season_id: Id of the season the video episode belongs to, as a unicode string.
229 episode: Title of the video episode. Unlike mandatory video title field,
230 this field should denote the exact title of the video episode
231 without any kind of decoration.
232 episode_number: Number of the video episode within a season, as an integer.
233 episode_id: Id of the video episode, as a unicode string.
235 The following fields should only be used when the media is a track or a part of
238 track: Title of the track.
239 track_number: Number of the track within an album or a disc, as an integer.
240 track_id: Id of the track (useful in case of custom indexing, e.g. 6.iii),
242 artist: Artist(s) of the track.
243 genre: Genre(s) of the track.
244 album: Title of the album the track belongs to.
245 album_type: Type of the album (e.g. "Demo", "Full-length", "Split", "Compilation", etc).
246 album_artist: List of all artists appeared on the album (e.g.
247 "Ash Borer / Fell Voices" or "Various Artists", useful for splits
249 disc_number: Number of the disc or other physical medium the track belongs to,
251 release_year: Year (YYYY) when the album was released.
253 Unless mentioned otherwise, the fields should be Unicode strings.
255 Unless mentioned otherwise, None is equivalent to absence of information.
258 _type "playlist" indicates multiple videos.
259 There must be a key "entries", which is a list, an iterable, or a PagedList
260 object, each element of which is a valid dictionary by this specification.
262 Additionally, playlists can have "title", "description" and "id" attributes
263 with the same semantics as videos (see above).
266 _type "multi_video" indicates that there are multiple videos that
267 form a single show, for examples multiple acts of an opera or TV episode.
268 It must have an entries key like a playlist and contain all the keys
269 required for a video at the same time.
272 _type "url" indicates that the video must be extracted from another
273 location, possibly by a different extractor. Its only required key is:
274 "url" - the next URL to extract.
275 The key "ie_key" can be set to the class name (minus the trailing "IE",
276 e.g. "Youtube") if the extractor class is known in advance.
277 Additionally, the dictionary may have any properties of the resolved entity
278 known in advance, for example "title" if the title of the referred video is
282 _type "url_transparent" entities have the same specification as "url", but
283 indicate that the given additional information is more precise than the one
284 associated with the resolved URL.
285 This is useful when a site employs a video service that hosts the video and
286 its technical metadata, but that video service does not embed a useful
287 title, description etc.
290 Subclasses of this one should re-define the _real_initialize() and
291 _real_extract() methods and define a _VALID_URL regexp.
292 Probably, they should also be added to the list of extractors.
294 Finally, the _WORKING attribute should be set to False for broken IEs
295 in order to warn the users and skip the tests.
302 def __init__(self, downloader=None):
303 """Constructor. Receives an optional downloader."""
305 self.set_downloader(downloader)
308 def suitable(cls, url):
309 """Receives a URL and returns True if suitable for this IE."""
311 # This does not use has/getattr intentionally - we want to know whether
312 # we have cached the regexp for *this* class, whereas getattr would also
313 # match the superclass
314 if '_VALID_URL_RE' not in cls.__dict__:
315 cls._VALID_URL_RE = re.compile(cls._VALID_URL)
316 return cls._VALID_URL_RE.match(url) is not None
319 def _match_id(cls, url):
320 if '_VALID_URL_RE' not in cls.__dict__:
321 cls._VALID_URL_RE = re.compile(cls._VALID_URL)
322 m = cls._VALID_URL_RE.match(url)
328 """Getter method for _WORKING."""
331 def initialize(self):
332 """Initializes an instance (authentication, etc)."""
334 self._real_initialize()
337 def extract(self, url):
338 """Extracts URL information and returns it in list of dicts."""
341 return self._real_extract(url)
342 except ExtractorError:
344 except compat_http_client.IncompleteRead as e:
345 raise ExtractorError('A network error has occurred.', cause=e, expected=True)
346 except (KeyError, StopIteration) as e:
347 raise ExtractorError('An extractor error has occurred.', cause=e)
349 def set_downloader(self, downloader):
350 """Sets the downloader for this IE."""
351 self._downloader = downloader
353 def _real_initialize(self):
354 """Real initialization process. Redefine in subclasses."""
357 def _real_extract(self, url):
358 """Real extraction process. Redefine in subclasses."""
363 """A string for getting the InfoExtractor with get_info_extractor"""
364 return compat_str(cls.__name__[:-2])
368 return compat_str(type(self).__name__[:-2])
370 def _request_webpage(self, url_or_request, video_id, note=None, errnote=None, fatal=True, data=None, headers={}, query={}):
371 """ Returns the response handle """
373 self.report_download_webpage(video_id)
374 elif note is not False:
376 self.to_screen('%s' % (note,))
378 self.to_screen('%s: %s' % (video_id, note))
379 # data, headers and query params will be ignored for `Request` objects
380 if isinstance(url_or_request, compat_urllib_request.Request):
381 url_or_request = update_Request(
382 url_or_request, data=data, headers=headers, query=query)
385 url_or_request = update_url_query(url_or_request, query)
387 url_or_request = sanitized_Request(url_or_request, data, headers)
389 return self._downloader.urlopen(url_or_request)
390 except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
394 errnote = 'Unable to download webpage'
396 errmsg = '%s: %s' % (errnote, error_to_compat_str(err))
398 raise ExtractorError(errmsg, sys.exc_info()[2], cause=err)
400 self._downloader.report_warning(errmsg)
403 def _download_webpage_handle(self, url_or_request, video_id, note=None, errnote=None, fatal=True, encoding=None, data=None, headers={}, query={}):
404 """ Returns a tuple (page content as string, URL handle) """
405 # Strip hashes from the URL (#1038)
406 if isinstance(url_or_request, (compat_str, str)):
407 url_or_request = url_or_request.partition('#')[0]
409 urlh = self._request_webpage(url_or_request, video_id, note, errnote, fatal, data=data, headers=headers, query=query)
413 content = self._webpage_read_content(urlh, url_or_request, video_id, note, errnote, fatal, encoding=encoding)
414 return (content, urlh)
417 def _guess_encoding_from_content(content_type, webpage_bytes):
418 m = re.match(r'[a-zA-Z0-9_.-]+/[a-zA-Z0-9_.-]+\s*;\s*charset=(.+)', content_type)
420 encoding = m.group(1)
422 m = re.search(br'<meta[^>]+charset=[\'"]?([^\'")]+)[ /\'">]',
423 webpage_bytes[:1024])
425 encoding = m.group(1).decode('ascii')
426 elif webpage_bytes.startswith(b'\xff\xfe'):
433 def _webpage_read_content(self, urlh, url_or_request, video_id, note=None, errnote=None, fatal=True, prefix=None, encoding=None):
434 content_type = urlh.headers.get('Content-Type', '')
435 webpage_bytes = urlh.read()
436 if prefix is not None:
437 webpage_bytes = prefix + webpage_bytes
439 encoding = self._guess_encoding_from_content(content_type, webpage_bytes)
440 if self._downloader.params.get('dump_intermediate_pages', False):
442 url = url_or_request.get_full_url()
443 except AttributeError:
445 self.to_screen('Dumping request to ' + url)
446 dump = base64.b64encode(webpage_bytes).decode('ascii')
447 self._downloader.to_screen(dump)
448 if self._downloader.params.get('write_pages', False):
450 url = url_or_request.get_full_url()
451 except AttributeError:
453 basen = '%s_%s' % (video_id, url)
455 h = '___' + hashlib.md5(basen.encode('utf-8')).hexdigest()
456 basen = basen[:240 - len(h)] + h
457 raw_filename = basen + '.dump'
458 filename = sanitize_filename(raw_filename, restricted=True)
459 self.to_screen('Saving request to ' + filename)
460 # Working around MAX_PATH limitation on Windows (see
461 # http://msdn.microsoft.com/en-us/library/windows/desktop/aa365247(v=vs.85).aspx)
462 if compat_os_name == 'nt':
463 absfilepath = os.path.abspath(filename)
464 if len(absfilepath) > 259:
465 filename = '\\\\?\\' + absfilepath
466 with open(filename, 'wb') as outf:
467 outf.write(webpage_bytes)
470 content = webpage_bytes.decode(encoding, 'replace')
472 content = webpage_bytes.decode('utf-8', 'replace')
474 if ('<title>Access to this site is blocked</title>' in content and
475 'Websense' in content[:512]):
476 msg = 'Access to this webpage has been blocked by Websense filtering software in your network.'
477 blocked_iframe = self._html_search_regex(
478 r'<iframe src="([^"]+)"', content,
479 'Websense information URL', default=None)
481 msg += ' Visit %s for more details' % blocked_iframe
482 raise ExtractorError(msg, expected=True)
483 if '<title>The URL you requested has been blocked</title>' in content[:512]:
485 'Access to this webpage has been blocked by Indian censorship. '
486 'Use a VPN or proxy server (with --proxy) to route around it.')
487 block_msg = self._html_search_regex(
488 r'</h1><p>(.*?)</p>',
489 content, 'block message', default=None)
491 msg += ' (Message: "%s")' % block_msg.replace('\n', ' ')
492 raise ExtractorError(msg, expected=True)
496 def _download_webpage(self, url_or_request, video_id, note=None, errnote=None, fatal=True, tries=1, timeout=5, encoding=None, data=None, headers={}, query={}):
497 """ Returns the data of the page as a string """
500 while success is False:
502 res = self._download_webpage_handle(url_or_request, video_id, note, errnote, fatal, encoding=encoding, data=data, headers=headers, query=query)
504 except compat_http_client.IncompleteRead as e:
506 if try_count >= tries:
508 self._sleep(timeout, video_id)
515 def _download_xml(self, url_or_request, video_id,
516 note='Downloading XML', errnote='Unable to download XML',
517 transform_source=None, fatal=True, encoding=None, data=None, headers={}, query={}):
518 """Return the xml as an xml.etree.ElementTree.Element"""
519 xml_string = self._download_webpage(
520 url_or_request, video_id, note, errnote, fatal=fatal, encoding=encoding, data=data, headers=headers, query=query)
521 if xml_string is False:
524 xml_string = transform_source(xml_string)
525 return compat_etree_fromstring(xml_string.encode('utf-8'))
527 def _download_json(self, url_or_request, video_id,
528 note='Downloading JSON metadata',
529 errnote='Unable to download JSON metadata',
530 transform_source=None,
531 fatal=True, encoding=None, data=None, headers={}, query={}):
532 json_string = self._download_webpage(
533 url_or_request, video_id, note, errnote, fatal=fatal,
534 encoding=encoding, data=data, headers=headers, query=query)
535 if (not fatal) and json_string is False:
537 return self._parse_json(
538 json_string, video_id, transform_source=transform_source, fatal=fatal)
540 def _parse_json(self, json_string, video_id, transform_source=None, fatal=True):
542 json_string = transform_source(json_string)
544 return json.loads(json_string)
545 except ValueError as ve:
546 errmsg = '%s: Failed to parse JSON ' % video_id
548 raise ExtractorError(errmsg, cause=ve)
550 self.report_warning(errmsg + str(ve))
552 def report_warning(self, msg, video_id=None):
553 idstr = '' if video_id is None else '%s: ' % video_id
554 self._downloader.report_warning(
555 '[%s] %s%s' % (self.IE_NAME, idstr, msg))
557 def to_screen(self, msg):
558 """Print msg to screen, prefixing it with '[ie_name]'"""
559 self._downloader.to_screen('[%s] %s' % (self.IE_NAME, msg))
561 def report_extraction(self, id_or_name):
562 """Report information extraction."""
563 self.to_screen('%s: Extracting information' % id_or_name)
565 def report_download_webpage(self, video_id):
566 """Report webpage download."""
567 self.to_screen('%s: Downloading webpage' % video_id)
569 def report_age_confirmation(self):
570 """Report attempt to confirm age."""
571 self.to_screen('Confirming age')
573 def report_login(self):
574 """Report attempt to log in."""
575 self.to_screen('Logging in')
578 def raise_login_required(msg='This video is only available for registered users'):
579 raise ExtractorError(
580 '%s. Use --username and --password or --netrc to provide account credentials.' % msg,
584 def raise_geo_restricted(msg='This video is not available from your location due to geo restriction'):
585 raise ExtractorError(
586 '%s. You might want to use --proxy to workaround.' % msg,
589 # Methods for following #608
591 def url_result(url, ie=None, video_id=None, video_title=None):
592 """Returns a URL that points to a page that should be processed"""
593 # TODO: ie should be the class used for getting the info
594 video_info = {'_type': 'url',
597 if video_id is not None:
598 video_info['id'] = video_id
599 if video_title is not None:
600 video_info['title'] = video_title
604 def playlist_result(entries, playlist_id=None, playlist_title=None, playlist_description=None):
605 """Returns a playlist"""
606 video_info = {'_type': 'playlist',
609 video_info['id'] = playlist_id
611 video_info['title'] = playlist_title
612 if playlist_description:
613 video_info['description'] = playlist_description
616 def _search_regex(self, pattern, string, name, default=NO_DEFAULT, fatal=True, flags=0, group=None):
618 Perform a regex search on the given string, using a single or a list of
619 patterns returning the first matching group.
620 In case of failure return a default value or raise a WARNING or a
621 RegexNotFoundError, depending on fatal, specifying the field name.
623 if isinstance(pattern, (str, compat_str, compiled_regex_type)):
624 mobj = re.search(pattern, string, flags)
627 mobj = re.search(p, string, flags)
631 if not self._downloader.params.get('no_color') and compat_os_name != 'nt' and sys.stderr.isatty():
632 _name = '\033[0;34m%s\033[0m' % name
638 # return the first matching group
639 return next(g for g in mobj.groups() if g is not None)
641 return mobj.group(group)
642 elif default is not NO_DEFAULT:
645 raise RegexNotFoundError('Unable to extract %s' % _name)
647 self._downloader.report_warning('unable to extract %s' % _name + bug_reports_message())
650 def _html_search_regex(self, pattern, string, name, default=NO_DEFAULT, fatal=True, flags=0, group=None):
652 Like _search_regex, but strips HTML tags and unescapes entities.
654 res = self._search_regex(pattern, string, name, default, fatal, flags, group)
656 return clean_html(res).strip()
660 def _get_login_info(self):
662 Get the login info as (username, password)
663 It will look in the netrc file using the _NETRC_MACHINE value
664 If there's no info available, return (None, None)
666 if self._downloader is None:
671 downloader_params = self._downloader.params
673 # Attempt to use provided username and password or .netrc data
674 if downloader_params.get('username') is not None:
675 username = downloader_params['username']
676 password = downloader_params['password']
677 elif downloader_params.get('usenetrc', False):
679 info = netrc.netrc().authenticators(self._NETRC_MACHINE)
684 raise netrc.NetrcParseError('No authenticators for %s' % self._NETRC_MACHINE)
685 except (IOError, netrc.NetrcParseError) as err:
686 self._downloader.report_warning('parsing .netrc: %s' % error_to_compat_str(err))
688 return (username, password)
690 def _get_tfa_info(self, note='two-factor verification code'):
692 Get the two-factor authentication info
693 TODO - asking the user will be required for sms/phone verify
694 currently just uses the command line option
695 If there's no info available, return None
697 if self._downloader is None:
699 downloader_params = self._downloader.params
701 if downloader_params.get('twofactor') is not None:
702 return downloader_params['twofactor']
704 return compat_getpass('Type %s and press [Return]: ' % note)
706 # Helper functions for extracting OpenGraph info
708 def _og_regexes(prop):
709 content_re = r'content=(?:"([^"]+?)"|\'([^\']+?)\'|\s*([^\s"\'=<>`]+?))'
710 property_re = (r'(?:name|property)=(?:\'og:%(prop)s\'|"og:%(prop)s"|\s*og:%(prop)s\b)'
711 % {'prop': re.escape(prop)})
712 template = r'<meta[^>]+?%s[^>]+?%s'
714 template % (property_re, content_re),
715 template % (content_re, property_re),
719 def _meta_regex(prop):
720 return r'''(?isx)<meta
721 (?=[^>]+(?:itemprop|name|property|id|http-equiv)=(["\']?)%s\1)
722 [^>]+?content=(["\'])(?P<content>.*?)\2''' % re.escape(prop)
724 def _og_search_property(self, prop, html, name=None, **kargs):
726 name = 'OpenGraph %s' % prop
727 escaped = self._search_regex(self._og_regexes(prop), html, name, flags=re.DOTALL, **kargs)
730 return unescapeHTML(escaped)
732 def _og_search_thumbnail(self, html, **kargs):
733 return self._og_search_property('image', html, 'thumbnail URL', fatal=False, **kargs)
735 def _og_search_description(self, html, **kargs):
736 return self._og_search_property('description', html, fatal=False, **kargs)
738 def _og_search_title(self, html, **kargs):
739 return self._og_search_property('title', html, **kargs)
741 def _og_search_video_url(self, html, name='video url', secure=True, **kargs):
742 regexes = self._og_regexes('video') + self._og_regexes('video:url')
744 regexes = self._og_regexes('video:secure_url') + regexes
745 return self._html_search_regex(regexes, html, name, **kargs)
747 def _og_search_url(self, html, **kargs):
748 return self._og_search_property('url', html, **kargs)
750 def _html_search_meta(self, name, html, display_name=None, fatal=False, **kwargs):
751 if display_name is None:
753 return self._html_search_regex(
754 self._meta_regex(name),
755 html, display_name, fatal=fatal, group='content', **kwargs)
757 def _dc_search_uploader(self, html):
758 return self._html_search_meta('dc.creator', html, 'uploader')
760 def _rta_search(self, html):
761 # See http://www.rtalabel.org/index.php?content=howtofaq#single
762 if re.search(r'(?ix)<meta\s+name="rating"\s+'
763 r' content="RTA-5042-1996-1400-1577-RTA"',
768 def _media_rating_search(self, html):
769 # See http://www.tjg-designs.com/WP/metadata-code-examples-adding-metadata-to-your-web-pages/
770 rating = self._html_search_meta('rating', html)
782 return RATING_TABLE.get(rating.lower())
784 def _family_friendly_search(self, html):
785 # See http://schema.org/VideoObject
786 family_friendly = self._html_search_meta('isFamilyFriendly', html)
788 if not family_friendly:
797 return RATING_TABLE.get(family_friendly.lower())
799 def _twitter_search_player(self, html):
800 return self._html_search_meta('twitter:player', html,
801 'twitter card player')
803 def _search_json_ld(self, html, video_id, **kwargs):
804 json_ld = self._search_regex(
805 r'(?s)<script[^>]+type=(["\'])application/ld\+json\1[^>]*>(?P<json_ld>.+?)</script>',
806 html, 'JSON-LD', group='json_ld', **kwargs)
809 return self._json_ld(json_ld, video_id, fatal=kwargs.get('fatal', True))
811 def _json_ld(self, json_ld, video_id, fatal=True):
812 if isinstance(json_ld, compat_str):
813 json_ld = self._parse_json(json_ld, video_id, fatal=fatal)
817 if json_ld.get('@context') == 'http://schema.org':
818 item_type = json_ld.get('@type')
819 if item_type == 'TVEpisode':
821 'episode': unescapeHTML(json_ld.get('name')),
822 'episode_number': int_or_none(json_ld.get('episodeNumber')),
823 'description': unescapeHTML(json_ld.get('description')),
825 part_of_season = json_ld.get('partOfSeason')
826 if isinstance(part_of_season, dict) and part_of_season.get('@type') == 'TVSeason':
827 info['season_number'] = int_or_none(part_of_season.get('seasonNumber'))
828 part_of_series = json_ld.get('partOfSeries')
829 if isinstance(part_of_series, dict) and part_of_series.get('@type') == 'TVSeries':
830 info['series'] = unescapeHTML(part_of_series.get('name'))
831 elif item_type == 'Article':
833 'timestamp': parse_iso8601(json_ld.get('datePublished')),
834 'title': unescapeHTML(json_ld.get('headline')),
835 'description': unescapeHTML(json_ld.get('articleBody')),
837 return dict((k, v) for k, v in info.items() if v is not None)
840 def _hidden_inputs(html):
841 html = re.sub(r'<!--(?:(?!<!--).)*-->', '', html)
843 for input in re.findall(r'(?i)<input([^>]+)>', html):
844 if not re.search(r'type=(["\'])(?:hidden|submit)\1', input):
846 name = re.search(r'name=(["\'])(?P<value>.+?)\1', input)
849 value = re.search(r'value=(["\'])(?P<value>.*?)\1', input)
852 hidden_inputs[name.group('value')] = value.group('value')
855 def _form_hidden_inputs(self, form_id, html):
856 form = self._search_regex(
857 r'(?is)<form[^>]+?id=(["\'])%s\1[^>]*>(?P<form>.+?)</form>' % form_id,
858 html, '%s form' % form_id, group='form')
859 return self._hidden_inputs(form)
861 def _sort_formats(self, formats, field_preference=None):
863 raise ExtractorError('No video formats found')
866 # Automatically determine tbr when missing based on abr and vbr (improves
867 # formats sorting in some cases)
868 if 'tbr' not in f and f.get('abr') is not None and f.get('vbr') is not None:
869 f['tbr'] = f['abr'] + f['vbr']
872 # TODO remove the following workaround
873 from ..utils import determine_ext
874 if not f.get('ext') and 'url' in f:
875 f['ext'] = determine_ext(f['url'])
877 if isinstance(field_preference, (list, tuple)):
878 return tuple(f.get(field) if f.get(field) is not None else -1 for field in field_preference)
880 preference = f.get('preference')
881 if preference is None:
883 if f.get('ext') in ['f4f', 'f4m']: # Not yet supported
886 proto_preference = 0 if determine_protocol(f) in ['http', 'https'] else -0.1
888 if f.get('vcodec') == 'none': # audio only
890 if self._downloader.params.get('prefer_free_formats'):
891 ORDER = ['aac', 'mp3', 'm4a', 'webm', 'ogg', 'opus']
893 ORDER = ['webm', 'opus', 'ogg', 'mp3', 'aac', 'm4a']
896 audio_ext_preference = ORDER.index(f['ext'])
898 audio_ext_preference = -1
900 if f.get('acodec') == 'none': # video only
902 if self._downloader.params.get('prefer_free_formats'):
903 ORDER = ['flv', 'mp4', 'webm']
905 ORDER = ['webm', 'flv', 'mp4']
907 ext_preference = ORDER.index(f['ext'])
910 audio_ext_preference = 0
914 f.get('language_preference') if f.get('language_preference') is not None else -1,
915 f.get('quality') if f.get('quality') is not None else -1,
916 f.get('tbr') if f.get('tbr') is not None else -1,
917 f.get('filesize') if f.get('filesize') is not None else -1,
918 f.get('vbr') if f.get('vbr') is not None else -1,
919 f.get('height') if f.get('height') is not None else -1,
920 f.get('width') if f.get('width') is not None else -1,
923 f.get('abr') if f.get('abr') is not None else -1,
924 audio_ext_preference,
925 f.get('fps') if f.get('fps') is not None else -1,
926 f.get('filesize_approx') if f.get('filesize_approx') is not None else -1,
927 f.get('source_preference') if f.get('source_preference') is not None else -1,
928 f.get('format_id') if f.get('format_id') is not None else '',
930 formats.sort(key=_formats_key)
932 def _check_formats(self, formats, video_id):
935 lambda f: self._is_valid_url(
937 item='%s video format' % f.get('format_id') if f.get('format_id') else 'video'),
941 def _remove_duplicate_formats(formats):
945 if f['url'] not in format_urls:
946 format_urls.add(f['url'])
947 unique_formats.append(f)
948 formats[:] = unique_formats
950 def _is_valid_url(self, url, video_id, item='video'):
951 url = self._proto_relative_url(url, scheme='http:')
952 # For now assume non HTTP(S) URLs always valid
953 if not (url.startswith('http://') or url.startswith('https://')):
956 self._request_webpage(url, video_id, 'Checking %s URL' % item)
958 except ExtractorError as e:
959 if isinstance(e.cause, compat_urllib_error.URLError):
961 '%s: %s URL is invalid, skipping' % (video_id, item))
965 def http_scheme(self):
966 """ Either "http:" or "https:", depending on the user's preferences """
969 if self._downloader.params.get('prefer_insecure', False)
972 def _proto_relative_url(self, url, scheme=None):
975 if url.startswith('//'):
977 scheme = self.http_scheme()
982 def _sleep(self, timeout, video_id, msg_template=None):
983 if msg_template is None:
984 msg_template = '%(video_id)s: Waiting for %(timeout)s seconds'
985 msg = msg_template % {'video_id': video_id, 'timeout': timeout}
989 def _extract_f4m_formats(self, manifest_url, video_id, preference=None, f4m_id=None,
990 transform_source=lambda s: fix_xml_ampersands(s).strip(),
992 manifest = self._download_xml(
993 manifest_url, video_id, 'Downloading f4m manifest',
994 'Unable to download f4m manifest',
995 # Some manifests may be malformed, e.g. prosiebensat1 generated manifests
996 # (see https://github.com/rg3/youtube-dl/issues/6215#issuecomment-121704244)
997 transform_source=transform_source,
1000 if manifest is False:
1003 return self._parse_f4m_formats(
1004 manifest, manifest_url, video_id, preference=preference, f4m_id=f4m_id,
1005 transform_source=transform_source, fatal=fatal)
1007 def _parse_f4m_formats(self, manifest, manifest_url, video_id, preference=None, f4m_id=None,
1008 transform_source=lambda s: fix_xml_ampersands(s).strip(),
1011 manifest_version = '1.0'
1012 media_nodes = manifest.findall('{http://ns.adobe.com/f4m/1.0}media')
1014 manifest_version = '2.0'
1015 media_nodes = manifest.findall('{http://ns.adobe.com/f4m/2.0}media')
1016 # Remove unsupported DRM protected media from final formats
1017 # rendition (see https://github.com/rg3/youtube-dl/issues/8573).
1018 media_nodes = remove_encrypted_media(media_nodes)
1021 base_url = xpath_text(
1022 manifest, ['{http://ns.adobe.com/f4m/1.0}baseURL', '{http://ns.adobe.com/f4m/2.0}baseURL'],
1023 'base URL', default=None)
1025 base_url = base_url.strip()
1026 for i, media_el in enumerate(media_nodes):
1027 if manifest_version == '2.0':
1028 media_url = media_el.attrib.get('href') or media_el.attrib.get('url')
1032 media_url if media_url.startswith('http://') or media_url.startswith('https://')
1033 else ((base_url or '/'.join(manifest_url.split('/')[:-1])) + '/' + media_url))
1034 # If media_url is itself a f4m manifest do the recursive extraction
1035 # since bitrates in parent manifest (this one) and media_url manifest
1036 # may differ leading to inability to resolve the format by requested
1037 # bitrate in f4m downloader
1038 if determine_ext(manifest_url) == 'f4m':
1039 formats.extend(self._extract_f4m_formats(
1040 manifest_url, video_id, preference=preference, f4m_id=f4m_id,
1041 transform_source=transform_source, fatal=fatal))
1043 tbr = int_or_none(media_el.attrib.get('bitrate'))
1045 'format_id': '-'.join(filter(None, [f4m_id, compat_str(i if tbr is None else tbr)])),
1046 'url': manifest_url,
1049 'width': int_or_none(media_el.attrib.get('width')),
1050 'height': int_or_none(media_el.attrib.get('height')),
1051 'preference': preference,
1055 def _extract_m3u8_formats(self, m3u8_url, video_id, ext=None,
1056 entry_protocol='m3u8', preference=None,
1057 m3u8_id=None, note=None, errnote=None,
1061 'format_id': '-'.join(filter(None, [m3u8_id, 'meta'])),
1065 'preference': preference - 1 if preference else -1,
1066 'resolution': 'multiple',
1067 'format_note': 'Quality selection URL',
1070 format_url = lambda u: (
1072 if re.match(r'^https?://', u)
1073 else compat_urlparse.urljoin(m3u8_url, u))
1075 res = self._download_webpage_handle(
1077 note=note or 'Downloading m3u8 information',
1078 errnote=errnote or 'Failed to download m3u8 information',
1082 m3u8_doc, urlh = res
1083 m3u8_url = urlh.geturl()
1085 # We should try extracting formats only from master playlists [1], i.e.
1086 # playlists that describe available qualities. On the other hand media
1087 # playlists [2] should be returned as is since they contain just the media
1088 # without qualities renditions.
1089 # Fortunately, master playlist can be easily distinguished from media
1090 # playlist based on particular tags availability. As of [1, 2] master
1091 # playlist tags MUST NOT appear in a media playist and vice versa.
1092 # As of [3] #EXT-X-TARGETDURATION tag is REQUIRED for every media playlist
1093 # and MUST NOT appear in master playlist thus we can clearly detect media
1094 # playlist with this criterion.
1095 # 1. https://tools.ietf.org/html/draft-pantos-http-live-streaming-17#section-4.3.4
1096 # 2. https://tools.ietf.org/html/draft-pantos-http-live-streaming-17#section-4.3.3
1097 # 3. https://tools.ietf.org/html/draft-pantos-http-live-streaming-17#section-4.3.3.1
1098 if '#EXT-X-TARGETDURATION' in m3u8_doc: # media playlist, return as is
1101 'format_id': m3u8_id,
1103 'protocol': entry_protocol,
1104 'preference': preference,
1108 kv_rex = re.compile(
1109 r'(?P<key>[a-zA-Z_-]+)=(?P<val>"[^"]+"|[^",]+)(?:,|$)')
1110 for line in m3u8_doc.splitlines():
1111 if line.startswith('#EXT-X-STREAM-INF:'):
1113 for m in kv_rex.finditer(line):
1115 if v.startswith('"'):
1117 last_info[m.group('key')] = v
1118 elif line.startswith('#EXT-X-MEDIA:'):
1120 for m in kv_rex.finditer(line):
1122 if v.startswith('"'):
1124 last_media[m.group('key')] = v
1125 elif line.startswith('#') or not line.strip():
1128 if last_info is None:
1129 formats.append({'url': format_url(line)})
1131 tbr = int_or_none(last_info.get('BANDWIDTH'), scale=1000)
1134 format_id.append(m3u8_id)
1135 last_media_name = last_media.get('NAME') if last_media and last_media.get('TYPE') != 'SUBTITLES' else None
1136 format_id.append(last_media_name if last_media_name else '%d' % (tbr if tbr else len(formats)))
1138 'format_id': '-'.join(format_id),
1139 'url': format_url(line.strip()),
1142 'protocol': entry_protocol,
1143 'preference': preference,
1145 resolution = last_info.get('RESOLUTION')
1147 width_str, height_str = resolution.split('x')
1148 f['width'] = int(width_str)
1149 f['height'] = int(height_str)
1150 codecs = last_info.get('CODECS')
1152 vcodec, acodec = [None] * 2
1153 va_codecs = codecs.split(',')
1154 if len(va_codecs) == 1:
1155 # Audio only entries usually come with single codec and
1156 # no resolution. For more robustness we also check it to
1158 if not resolution and va_codecs[0].startswith('mp4a'):
1159 vcodec, acodec = 'none', va_codecs[0]
1161 vcodec = va_codecs[0]
1163 vcodec, acodec = va_codecs[:2]
1168 if last_media is not None:
1169 f['m3u8_media'] = last_media
1176 def _xpath_ns(path, namespace=None):
1180 for c in path.split('/'):
1181 if not c or c == '.':
1184 out.append('{%s}%s' % (namespace, c))
1185 return '/'.join(out)
1187 def _extract_smil_formats(self, smil_url, video_id, fatal=True, f4m_params=None, transform_source=None):
1188 smil = self._download_smil(smil_url, video_id, fatal=fatal, transform_source=transform_source)
1194 namespace = self._parse_smil_namespace(smil)
1196 return self._parse_smil_formats(
1197 smil, smil_url, video_id, namespace=namespace, f4m_params=f4m_params)
1199 def _extract_smil_info(self, smil_url, video_id, fatal=True, f4m_params=None):
1200 smil = self._download_smil(smil_url, video_id, fatal=fatal)
1203 return self._parse_smil(smil, smil_url, video_id, f4m_params=f4m_params)
1205 def _download_smil(self, smil_url, video_id, fatal=True, transform_source=None):
1206 return self._download_xml(
1207 smil_url, video_id, 'Downloading SMIL file',
1208 'Unable to download SMIL file', fatal=fatal, transform_source=transform_source)
1210 def _parse_smil(self, smil, smil_url, video_id, f4m_params=None):
1211 namespace = self._parse_smil_namespace(smil)
1213 formats = self._parse_smil_formats(
1214 smil, smil_url, video_id, namespace=namespace, f4m_params=f4m_params)
1215 subtitles = self._parse_smil_subtitles(smil, namespace=namespace)
1217 video_id = os.path.splitext(url_basename(smil_url))[0]
1221 for meta in smil.findall(self._xpath_ns('./head/meta', namespace)):
1222 name = meta.attrib.get('name')
1223 content = meta.attrib.get('content')
1224 if not name or not content:
1226 if not title and name == 'title':
1228 elif not description and name in ('description', 'abstract'):
1229 description = content
1230 elif not upload_date and name == 'date':
1231 upload_date = unified_strdate(content)
1234 'id': image.get('type'),
1235 'url': image.get('src'),
1236 'width': int_or_none(image.get('width')),
1237 'height': int_or_none(image.get('height')),
1238 } for image in smil.findall(self._xpath_ns('.//image', namespace)) if image.get('src')]
1242 'title': title or video_id,
1243 'description': description,
1244 'upload_date': upload_date,
1245 'thumbnails': thumbnails,
1247 'subtitles': subtitles,
1250 def _parse_smil_namespace(self, smil):
1251 return self._search_regex(
1252 r'(?i)^{([^}]+)?}smil$', smil.tag, 'namespace', default=None)
1254 def _parse_smil_formats(self, smil, smil_url, video_id, namespace=None, f4m_params=None, transform_rtmp_url=None):
1256 for meta in smil.findall(self._xpath_ns('./head/meta', namespace)):
1257 b = meta.get('base') or meta.get('httpBase')
1268 videos = smil.findall(self._xpath_ns('.//video', namespace))
1269 for video in videos:
1270 src = video.get('src')
1271 if not src or src in srcs:
1275 bitrate = float_or_none(video.get('system-bitrate') or video.get('systemBitrate'), 1000)
1276 filesize = int_or_none(video.get('size') or video.get('fileSize'))
1277 width = int_or_none(video.get('width'))
1278 height = int_or_none(video.get('height'))
1279 proto = video.get('proto')
1280 ext = video.get('ext')
1281 src_ext = determine_ext(src)
1282 streamer = video.get('streamer') or base
1284 if proto == 'rtmp' or streamer.startswith('rtmp'):
1290 'format_id': 'rtmp-%d' % (rtmp_count if bitrate is None else bitrate),
1292 'filesize': filesize,
1296 if transform_rtmp_url:
1297 streamer, src = transform_rtmp_url(streamer, src)
1298 formats[-1].update({
1304 src_url = src if src.startswith('http') else compat_urlparse.urljoin(base, src)
1305 src_url = src_url.strip()
1307 if proto == 'm3u8' or src_ext == 'm3u8':
1308 m3u8_formats = self._extract_m3u8_formats(
1309 src_url, video_id, ext or 'mp4', m3u8_id='hls', fatal=False)
1310 if len(m3u8_formats) == 1:
1312 m3u8_formats[0].update({
1313 'format_id': 'hls-%d' % (m3u8_count if bitrate is None else bitrate),
1318 formats.extend(m3u8_formats)
1321 if src_ext == 'f4m':
1326 'plugin': 'flowplayer-3.2.0.1',
1328 f4m_url += '&' if '?' in f4m_url else '?'
1329 f4m_url += compat_urllib_parse_urlencode(f4m_params)
1330 formats.extend(self._extract_f4m_formats(f4m_url, video_id, f4m_id='hds', fatal=False))
1333 if src_url.startswith('http') and self._is_valid_url(src, video_id):
1337 'ext': ext or src_ext or 'flv',
1338 'format_id': 'http-%d' % (bitrate or http_count),
1340 'filesize': filesize,
1348 def _parse_smil_subtitles(self, smil, namespace=None, subtitles_lang='en'):
1351 for num, textstream in enumerate(smil.findall(self._xpath_ns('.//textstream', namespace))):
1352 src = textstream.get('src')
1353 if not src or src in urls:
1356 ext = textstream.get('ext') or mimetype2ext(textstream.get('type')) or determine_ext(src)
1357 lang = textstream.get('systemLanguage') or textstream.get('systemLanguageName') or textstream.get('lang') or subtitles_lang
1358 subtitles.setdefault(lang, []).append({
1364 def _extract_xspf_playlist(self, playlist_url, playlist_id, fatal=True):
1365 xspf = self._download_xml(
1366 playlist_url, playlist_id, 'Downloading xpsf playlist',
1367 'Unable to download xspf manifest', fatal=fatal)
1370 return self._parse_xspf(xspf, playlist_id)
1372 def _parse_xspf(self, playlist, playlist_id):
1374 'xspf': 'http://xspf.org/ns/0/',
1375 's1': 'http://static.streamone.nl/player/ns/0',
1379 for track in playlist.findall(xpath_with_ns('./xspf:trackList/xspf:track', NS_MAP)):
1381 track, xpath_with_ns('./xspf:title', NS_MAP), 'title', default=playlist_id)
1382 description = xpath_text(
1383 track, xpath_with_ns('./xspf:annotation', NS_MAP), 'description')
1384 thumbnail = xpath_text(
1385 track, xpath_with_ns('./xspf:image', NS_MAP), 'thumbnail')
1386 duration = float_or_none(
1387 xpath_text(track, xpath_with_ns('./xspf:duration', NS_MAP), 'duration'), 1000)
1390 'url': location.text,
1391 'format_id': location.get(xpath_with_ns('s1:label', NS_MAP)),
1392 'width': int_or_none(location.get(xpath_with_ns('s1:width', NS_MAP))),
1393 'height': int_or_none(location.get(xpath_with_ns('s1:height', NS_MAP))),
1394 } for location in track.findall(xpath_with_ns('./xspf:location', NS_MAP))]
1395 self._sort_formats(formats)
1400 'description': description,
1401 'thumbnail': thumbnail,
1402 'duration': duration,
1407 def _extract_mpd_formats(self, mpd_url, video_id, mpd_id=None, note=None, errnote=None, fatal=True, formats_dict={}):
1408 res = self._download_webpage_handle(
1410 note=note or 'Downloading MPD manifest',
1411 errnote=errnote or 'Failed to download MPD manifest',
1416 mpd_base_url = re.match(r'https?://.+/', urlh.geturl()).group()
1418 return self._parse_mpd_formats(
1419 compat_etree_fromstring(mpd.encode('utf-8')), mpd_id, mpd_base_url, formats_dict=formats_dict)
1421 def _parse_mpd_formats(self, mpd_doc, mpd_id=None, mpd_base_url='', formats_dict={}):
1422 if mpd_doc.get('type') == 'dynamic':
1425 namespace = self._search_regex(r'(?i)^{([^}]+)?}MPD$', mpd_doc.tag, 'namespace', default=None)
1428 return self._xpath_ns(path, namespace)
1430 def is_drm_protected(element):
1431 return element.find(_add_ns('ContentProtection')) is not None
1433 def extract_multisegment_info(element, ms_parent_info):
1434 ms_info = ms_parent_info.copy()
1435 segment_list = element.find(_add_ns('SegmentList'))
1436 if segment_list is not None:
1437 segment_urls_e = segment_list.findall(_add_ns('SegmentURL'))
1439 ms_info['segment_urls'] = [segment.attrib['media'] for segment in segment_urls_e]
1440 initialization = segment_list.find(_add_ns('Initialization'))
1441 if initialization is not None:
1442 ms_info['initialization_url'] = initialization.attrib['sourceURL']
1444 segment_template = element.find(_add_ns('SegmentTemplate'))
1445 if segment_template is not None:
1446 start_number = segment_template.get('startNumber')
1448 ms_info['start_number'] = int(start_number)
1449 segment_timeline = segment_template.find(_add_ns('SegmentTimeline'))
1450 if segment_timeline is not None:
1451 s_e = segment_timeline.findall(_add_ns('S'))
1453 ms_info['total_number'] = 0
1455 ms_info['total_number'] += 1 + int(s.get('r', '0'))
1457 timescale = segment_template.get('timescale')
1459 ms_info['timescale'] = int(timescale)
1460 segment_duration = segment_template.get('duration')
1461 if segment_duration:
1462 ms_info['segment_duration'] = int(segment_duration)
1463 media_template = segment_template.get('media')
1465 ms_info['media_template'] = media_template
1466 initialization = segment_template.get('initialization')
1468 ms_info['initialization_url'] = initialization
1470 initialization = segment_template.find(_add_ns('Initialization'))
1471 if initialization is not None:
1472 ms_info['initialization_url'] = initialization.attrib['sourceURL']
1475 mpd_duration = parse_duration(mpd_doc.get('mediaPresentationDuration'))
1477 for period in mpd_doc.findall(_add_ns('Period')):
1478 period_duration = parse_duration(period.get('duration')) or mpd_duration
1479 period_ms_info = extract_multisegment_info(period, {
1483 for adaptation_set in period.findall(_add_ns('AdaptationSet')):
1484 if is_drm_protected(adaptation_set):
1486 adaption_set_ms_info = extract_multisegment_info(adaptation_set, period_ms_info)
1487 for representation in adaptation_set.findall(_add_ns('Representation')):
1488 if is_drm_protected(representation):
1490 representation_attrib = adaptation_set.attrib.copy()
1491 representation_attrib.update(representation.attrib)
1492 # According to page 41 of ISO/IEC 29001-1:2014, @mimeType is mandatory
1493 mime_type = representation_attrib['mimeType']
1494 content_type = mime_type.split('/')[0]
1495 if content_type == 'text':
1496 # TODO implement WebVTT downloading
1498 elif content_type == 'video' or content_type == 'audio':
1500 for element in (representation, adaptation_set, period, mpd_doc):
1501 base_url_e = element.find(_add_ns('BaseURL'))
1502 if base_url_e is not None:
1503 base_url = base_url_e.text + base_url
1504 if re.match(r'^https?://', base_url):
1506 if mpd_base_url and not re.match(r'^https?://', base_url):
1507 if not mpd_base_url.endswith('/') and not base_url.startswith('/'):
1509 base_url = mpd_base_url + base_url
1510 representation_id = representation_attrib.get('id')
1511 lang = representation_attrib.get('lang')
1512 url_el = representation.find(_add_ns('BaseURL'))
1513 filesize = int_or_none(url_el.attrib.get('{http://youtube.com/yt/2012/10/10}contentLength') if url_el is not None else None)
1515 'format_id': '%s-%s' % (mpd_id, representation_id) if mpd_id else representation_id,
1517 'ext': mimetype2ext(mime_type),
1518 'width': int_or_none(representation_attrib.get('width')),
1519 'height': int_or_none(representation_attrib.get('height')),
1520 'tbr': int_or_none(representation_attrib.get('bandwidth'), 1000),
1521 'asr': int_or_none(representation_attrib.get('audioSamplingRate')),
1522 'fps': int_or_none(representation_attrib.get('frameRate')),
1523 'vcodec': 'none' if content_type == 'audio' else representation_attrib.get('codecs'),
1524 'acodec': 'none' if content_type == 'video' else representation_attrib.get('codecs'),
1525 'language': lang if lang not in ('mul', 'und', 'zxx', 'mis') else None,
1526 'format_note': 'DASH %s' % content_type,
1527 'filesize': filesize,
1529 representation_ms_info = extract_multisegment_info(representation, adaption_set_ms_info)
1530 if 'segment_urls' not in representation_ms_info and 'media_template' in representation_ms_info:
1531 if 'total_number' not in representation_ms_info and 'segment_duration':
1532 segment_duration = float(representation_ms_info['segment_duration']) / float(representation_ms_info['timescale'])
1533 representation_ms_info['total_number'] = int(math.ceil(float(period_duration) / segment_duration))
1534 media_template = representation_ms_info['media_template']
1535 media_template = media_template.replace('$RepresentationID$', representation_id)
1536 media_template = re.sub(r'\$(Number|Bandwidth)\$', r'%(\1)d', media_template)
1537 media_template = re.sub(r'\$(Number|Bandwidth)%([^$]+)\$', r'%(\1)\2', media_template)
1538 media_template.replace('$$', '$')
1539 representation_ms_info['segment_urls'] = [
1541 'Number': segment_number,
1542 'Bandwidth': representation_attrib.get('bandwidth')}
1543 for segment_number in range(
1544 representation_ms_info['start_number'],
1545 representation_ms_info['total_number'] + representation_ms_info['start_number'])]
1546 if 'segment_urls' in representation_ms_info:
1548 'segment_urls': representation_ms_info['segment_urls'],
1549 'protocol': 'http_dash_segments',
1551 if 'initialization_url' in representation_ms_info:
1552 initialization_url = representation_ms_info['initialization_url'].replace('$RepresentationID$', representation_id)
1554 'initialization_url': initialization_url,
1556 if not f.get('url'):
1557 f['url'] = initialization_url
1559 existing_format = next(
1560 fo for fo in formats
1561 if fo['format_id'] == representation_id)
1562 except StopIteration:
1563 full_info = formats_dict.get(representation_id, {}).copy()
1565 formats.append(full_info)
1567 existing_format.update(f)
1569 self.report_warning('Unknown MIME type %s in DASH manifest' % mime_type)
1572 def _live_title(self, name):
1573 """ Generate the title for a live video """
1574 now = datetime.datetime.now()
1575 now_str = now.strftime('%Y-%m-%d %H:%M')
1576 return name + ' ' + now_str
1578 def _int(self, v, name, fatal=False, **kwargs):
1579 res = int_or_none(v, **kwargs)
1580 if 'get_attr' in kwargs:
1581 print(getattr(v, kwargs['get_attr']))
1583 msg = 'Failed to extract %s: Could not parse value %r' % (name, v)
1585 raise ExtractorError(msg)
1587 self._downloader.report_warning(msg)
1590 def _float(self, v, name, fatal=False, **kwargs):
1591 res = float_or_none(v, **kwargs)
1593 msg = 'Failed to extract %s: Could not parse value %r' % (name, v)
1595 raise ExtractorError(msg)
1597 self._downloader.report_warning(msg)
1600 def _set_cookie(self, domain, name, value, expire_time=None):
1601 cookie = compat_cookiejar.Cookie(
1602 0, name, value, None, None, domain, None,
1603 None, '/', True, False, expire_time, '', None, None, None)
1604 self._downloader.cookiejar.set_cookie(cookie)
1606 def _get_cookies(self, url):
1607 """ Return a compat_cookies.SimpleCookie with the cookies for the url """
1608 req = sanitized_Request(url)
1609 self._downloader.cookiejar.add_cookie_header(req)
1610 return compat_cookies.SimpleCookie(req.get_header('Cookie'))
1612 def get_testcases(self, include_onlymatching=False):
1613 t = getattr(self, '_TEST', None)
1615 assert not hasattr(self, '_TESTS'), \
1616 '%s has _TEST and _TESTS' % type(self).__name__
1619 tests = getattr(self, '_TESTS', [])
1621 if not include_onlymatching and t.get('only_matching', False):
1623 t['name'] = type(self).__name__[:-len('IE')]
1626 def is_suitable(self, age_limit):
1627 """ Test whether the extractor is generally suitable for the given
1628 age limit (i.e. pornographic sites are not, all others usually are) """
1630 any_restricted = False
1631 for tc in self.get_testcases(include_onlymatching=False):
1632 if 'playlist' in tc:
1633 tc = tc['playlist'][0]
1634 is_restricted = age_restricted(
1635 tc.get('info_dict', {}).get('age_limit'), age_limit)
1636 if not is_restricted:
1638 any_restricted = any_restricted or is_restricted
1639 return not any_restricted
1641 def extract_subtitles(self, *args, **kwargs):
1642 if (self._downloader.params.get('writesubtitles', False) or
1643 self._downloader.params.get('listsubtitles')):
1644 return self._get_subtitles(*args, **kwargs)
1647 def _get_subtitles(self, *args, **kwargs):
1648 raise NotImplementedError('This method must be implemented by subclasses')
1651 def _merge_subtitle_items(subtitle_list1, subtitle_list2):
1652 """ Merge subtitle items for one language. Items with duplicated URLs
1653 will be dropped. """
1654 list1_urls = set([item['url'] for item in subtitle_list1])
1655 ret = list(subtitle_list1)
1656 ret.extend([item for item in subtitle_list2 if item['url'] not in list1_urls])
1660 def _merge_subtitles(cls, subtitle_dict1, subtitle_dict2):
1661 """ Merge two subtitle dictionaries, language by language. """
1662 ret = dict(subtitle_dict1)
1663 for lang in subtitle_dict2:
1664 ret[lang] = cls._merge_subtitle_items(subtitle_dict1.get(lang, []), subtitle_dict2[lang])
1667 def extract_automatic_captions(self, *args, **kwargs):
1668 if (self._downloader.params.get('writeautomaticsub', False) or
1669 self._downloader.params.get('listsubtitles')):
1670 return self._get_automatic_captions(*args, **kwargs)
1673 def _get_automatic_captions(self, *args, **kwargs):
1674 raise NotImplementedError('This method must be implemented by subclasses')
1676 def mark_watched(self, *args, **kwargs):
1677 if (self._downloader.params.get('mark_watched', False) and
1678 (self._get_login_info()[0] is not None or
1679 self._downloader.params.get('cookiefile') is not None)):
1680 self._mark_watched(*args, **kwargs)
1682 def _mark_watched(self, *args, **kwargs):
1683 raise NotImplementedError('This method must be implemented by subclasses')
1686 class SearchInfoExtractor(InfoExtractor):
1688 Base class for paged search queries extractors.
1689 They accept URLs in the format _SEARCH_KEY(|all|[0-9]):{query}
1690 Instances should define _SEARCH_KEY and _MAX_RESULTS.
1694 def _make_valid_url(cls):
1695 return r'%s(?P<prefix>|[1-9][0-9]*|all):(?P<query>[\s\S]+)' % cls._SEARCH_KEY
1698 def suitable(cls, url):
1699 return re.match(cls._make_valid_url(), url) is not None
1701 def _real_extract(self, query):
1702 mobj = re.match(self._make_valid_url(), query)
1704 raise ExtractorError('Invalid search query "%s"' % query)
1706 prefix = mobj.group('prefix')
1707 query = mobj.group('query')
1709 return self._get_n_results(query, 1)
1710 elif prefix == 'all':
1711 return self._get_n_results(query, self._MAX_RESULTS)
1715 raise ExtractorError('invalid download number %s for query "%s"' % (n, query))
1716 elif n > self._MAX_RESULTS:
1717 self._downloader.report_warning('%s returns max %i results (you requested %i)' % (self._SEARCH_KEY, self._MAX_RESULTS, n))
1718 n = self._MAX_RESULTS
1719 return self._get_n_results(query, n)
1721 def _get_n_results(self, query, n):
1722 """Get a specified number of results for a query"""
1723 raise NotImplementedError('This method must be implemented by subclasses')
1726 def SEARCH_KEY(self):
1727 return self._SEARCH_KEY