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 creator of 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 if isinstance(url_or_request, compat_urllib_request.Request):
380 url_or_request = update_Request(
381 url_or_request, data=data, headers=headers, query=query)
384 url_or_request = update_url_query(url_or_request, query)
385 if data is not None or headers:
386 url_or_request = sanitized_Request(url_or_request, data, headers)
388 return self._downloader.urlopen(url_or_request)
389 except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
393 errnote = 'Unable to download webpage'
395 errmsg = '%s: %s' % (errnote, error_to_compat_str(err))
397 raise ExtractorError(errmsg, sys.exc_info()[2], cause=err)
399 self._downloader.report_warning(errmsg)
402 def _download_webpage_handle(self, url_or_request, video_id, note=None, errnote=None, fatal=True, encoding=None, data=None, headers={}, query={}):
403 """ Returns a tuple (page content as string, URL handle) """
404 # Strip hashes from the URL (#1038)
405 if isinstance(url_or_request, (compat_str, str)):
406 url_or_request = url_or_request.partition('#')[0]
408 urlh = self._request_webpage(url_or_request, video_id, note, errnote, fatal, data=data, headers=headers, query=query)
412 content = self._webpage_read_content(urlh, url_or_request, video_id, note, errnote, fatal, encoding=encoding)
413 return (content, urlh)
416 def _guess_encoding_from_content(content_type, webpage_bytes):
417 m = re.match(r'[a-zA-Z0-9_.-]+/[a-zA-Z0-9_.-]+\s*;\s*charset=(.+)', content_type)
419 encoding = m.group(1)
421 m = re.search(br'<meta[^>]+charset=[\'"]?([^\'")]+)[ /\'">]',
422 webpage_bytes[:1024])
424 encoding = m.group(1).decode('ascii')
425 elif webpage_bytes.startswith(b'\xff\xfe'):
432 def _webpage_read_content(self, urlh, url_or_request, video_id, note=None, errnote=None, fatal=True, prefix=None, encoding=None):
433 content_type = urlh.headers.get('Content-Type', '')
434 webpage_bytes = urlh.read()
435 if prefix is not None:
436 webpage_bytes = prefix + webpage_bytes
438 encoding = self._guess_encoding_from_content(content_type, webpage_bytes)
439 if self._downloader.params.get('dump_intermediate_pages', False):
441 url = url_or_request.get_full_url()
442 except AttributeError:
444 self.to_screen('Dumping request to ' + url)
445 dump = base64.b64encode(webpage_bytes).decode('ascii')
446 self._downloader.to_screen(dump)
447 if self._downloader.params.get('write_pages', False):
449 url = url_or_request.get_full_url()
450 except AttributeError:
452 basen = '%s_%s' % (video_id, url)
454 h = '___' + hashlib.md5(basen.encode('utf-8')).hexdigest()
455 basen = basen[:240 - len(h)] + h
456 raw_filename = basen + '.dump'
457 filename = sanitize_filename(raw_filename, restricted=True)
458 self.to_screen('Saving request to ' + filename)
459 # Working around MAX_PATH limitation on Windows (see
460 # http://msdn.microsoft.com/en-us/library/windows/desktop/aa365247(v=vs.85).aspx)
461 if compat_os_name == 'nt':
462 absfilepath = os.path.abspath(filename)
463 if len(absfilepath) > 259:
464 filename = '\\\\?\\' + absfilepath
465 with open(filename, 'wb') as outf:
466 outf.write(webpage_bytes)
469 content = webpage_bytes.decode(encoding, 'replace')
471 content = webpage_bytes.decode('utf-8', 'replace')
473 if ('<title>Access to this site is blocked</title>' in content and
474 'Websense' in content[:512]):
475 msg = 'Access to this webpage has been blocked by Websense filtering software in your network.'
476 blocked_iframe = self._html_search_regex(
477 r'<iframe src="([^"]+)"', content,
478 'Websense information URL', default=None)
480 msg += ' Visit %s for more details' % blocked_iframe
481 raise ExtractorError(msg, expected=True)
482 if '<title>The URL you requested has been blocked</title>' in content[:512]:
484 'Access to this webpage has been blocked by Indian censorship. '
485 'Use a VPN or proxy server (with --proxy) to route around it.')
486 block_msg = self._html_search_regex(
487 r'</h1><p>(.*?)</p>',
488 content, 'block message', default=None)
490 msg += ' (Message: "%s")' % block_msg.replace('\n', ' ')
491 raise ExtractorError(msg, expected=True)
495 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={}):
496 """ Returns the data of the page as a string """
499 while success is False:
501 res = self._download_webpage_handle(url_or_request, video_id, note, errnote, fatal, encoding=encoding, data=data, headers=headers, query=query)
503 except compat_http_client.IncompleteRead as e:
505 if try_count >= tries:
507 self._sleep(timeout, video_id)
514 def _download_xml(self, url_or_request, video_id,
515 note='Downloading XML', errnote='Unable to download XML',
516 transform_source=None, fatal=True, encoding=None, data=None, headers={}, query={}):
517 """Return the xml as an xml.etree.ElementTree.Element"""
518 xml_string = self._download_webpage(
519 url_or_request, video_id, note, errnote, fatal=fatal, encoding=encoding, data=data, headers=headers, query=query)
520 if xml_string is False:
523 xml_string = transform_source(xml_string)
524 return compat_etree_fromstring(xml_string.encode('utf-8'))
526 def _download_json(self, url_or_request, video_id,
527 note='Downloading JSON metadata',
528 errnote='Unable to download JSON metadata',
529 transform_source=None,
530 fatal=True, encoding=None, data=None, headers={}, query={}):
531 json_string = self._download_webpage(
532 url_or_request, video_id, note, errnote, fatal=fatal,
533 encoding=encoding, data=data, headers=headers, query=query)
534 if (not fatal) and json_string is False:
536 return self._parse_json(
537 json_string, video_id, transform_source=transform_source, fatal=fatal)
539 def _parse_json(self, json_string, video_id, transform_source=None, fatal=True):
541 json_string = transform_source(json_string)
543 return json.loads(json_string)
544 except ValueError as ve:
545 errmsg = '%s: Failed to parse JSON ' % video_id
547 raise ExtractorError(errmsg, cause=ve)
549 self.report_warning(errmsg + str(ve))
551 def report_warning(self, msg, video_id=None):
552 idstr = '' if video_id is None else '%s: ' % video_id
553 self._downloader.report_warning(
554 '[%s] %s%s' % (self.IE_NAME, idstr, msg))
556 def to_screen(self, msg):
557 """Print msg to screen, prefixing it with '[ie_name]'"""
558 self._downloader.to_screen('[%s] %s' % (self.IE_NAME, msg))
560 def report_extraction(self, id_or_name):
561 """Report information extraction."""
562 self.to_screen('%s: Extracting information' % id_or_name)
564 def report_download_webpage(self, video_id):
565 """Report webpage download."""
566 self.to_screen('%s: Downloading webpage' % video_id)
568 def report_age_confirmation(self):
569 """Report attempt to confirm age."""
570 self.to_screen('Confirming age')
572 def report_login(self):
573 """Report attempt to log in."""
574 self.to_screen('Logging in')
577 def raise_login_required(msg='This video is only available for registered users'):
578 raise ExtractorError(
579 '%s. Use --username and --password or --netrc to provide account credentials.' % msg,
583 def raise_geo_restricted(msg='This video is not available from your location due to geo restriction'):
584 raise ExtractorError(
585 '%s. You might want to use --proxy to workaround.' % msg,
588 # Methods for following #608
590 def url_result(url, ie=None, video_id=None, video_title=None):
591 """Returns a URL that points to a page that should be processed"""
592 # TODO: ie should be the class used for getting the info
593 video_info = {'_type': 'url',
596 if video_id is not None:
597 video_info['id'] = video_id
598 if video_title is not None:
599 video_info['title'] = video_title
603 def playlist_result(entries, playlist_id=None, playlist_title=None, playlist_description=None):
604 """Returns a playlist"""
605 video_info = {'_type': 'playlist',
608 video_info['id'] = playlist_id
610 video_info['title'] = playlist_title
611 if playlist_description:
612 video_info['description'] = playlist_description
615 def _search_regex(self, pattern, string, name, default=NO_DEFAULT, fatal=True, flags=0, group=None):
617 Perform a regex search on the given string, using a single or a list of
618 patterns returning the first matching group.
619 In case of failure return a default value or raise a WARNING or a
620 RegexNotFoundError, depending on fatal, specifying the field name.
622 if isinstance(pattern, (str, compat_str, compiled_regex_type)):
623 mobj = re.search(pattern, string, flags)
626 mobj = re.search(p, string, flags)
630 if not self._downloader.params.get('no_color') and compat_os_name != 'nt' and sys.stderr.isatty():
631 _name = '\033[0;34m%s\033[0m' % name
637 # return the first matching group
638 return next(g for g in mobj.groups() if g is not None)
640 return mobj.group(group)
641 elif default is not NO_DEFAULT:
644 raise RegexNotFoundError('Unable to extract %s' % _name)
646 self._downloader.report_warning('unable to extract %s' % _name + bug_reports_message())
649 def _html_search_regex(self, pattern, string, name, default=NO_DEFAULT, fatal=True, flags=0, group=None):
651 Like _search_regex, but strips HTML tags and unescapes entities.
653 res = self._search_regex(pattern, string, name, default, fatal, flags, group)
655 return clean_html(res).strip()
659 def _get_login_info(self):
661 Get the login info as (username, password)
662 It will look in the netrc file using the _NETRC_MACHINE value
663 If there's no info available, return (None, None)
665 if self._downloader is None:
670 downloader_params = self._downloader.params
672 # Attempt to use provided username and password or .netrc data
673 if downloader_params.get('username') is not None:
674 username = downloader_params['username']
675 password = downloader_params['password']
676 elif downloader_params.get('usenetrc', False):
678 info = netrc.netrc().authenticators(self._NETRC_MACHINE)
683 raise netrc.NetrcParseError('No authenticators for %s' % self._NETRC_MACHINE)
684 except (IOError, netrc.NetrcParseError) as err:
685 self._downloader.report_warning('parsing .netrc: %s' % error_to_compat_str(err))
687 return (username, password)
689 def _get_tfa_info(self, note='two-factor verification code'):
691 Get the two-factor authentication info
692 TODO - asking the user will be required for sms/phone verify
693 currently just uses the command line option
694 If there's no info available, return None
696 if self._downloader is None:
698 downloader_params = self._downloader.params
700 if downloader_params.get('twofactor') is not None:
701 return downloader_params['twofactor']
703 return compat_getpass('Type %s and press [Return]: ' % note)
705 # Helper functions for extracting OpenGraph info
707 def _og_regexes(prop):
708 content_re = r'content=(?:"([^"]+?)"|\'([^\']+?)\'|\s*([^\s"\'=<>`]+?))'
709 property_re = (r'(?:name|property)=(?:\'og:%(prop)s\'|"og:%(prop)s"|\s*og:%(prop)s\b)'
710 % {'prop': re.escape(prop)})
711 template = r'<meta[^>]+?%s[^>]+?%s'
713 template % (property_re, content_re),
714 template % (content_re, property_re),
718 def _meta_regex(prop):
719 return r'''(?isx)<meta
720 (?=[^>]+(?:itemprop|name|property|id|http-equiv)=(["\']?)%s\1)
721 [^>]+?content=(["\'])(?P<content>.*?)\2''' % re.escape(prop)
723 def _og_search_property(self, prop, html, name=None, **kargs):
725 name = 'OpenGraph %s' % prop
726 escaped = self._search_regex(self._og_regexes(prop), html, name, flags=re.DOTALL, **kargs)
729 return unescapeHTML(escaped)
731 def _og_search_thumbnail(self, html, **kargs):
732 return self._og_search_property('image', html, 'thumbnail URL', fatal=False, **kargs)
734 def _og_search_description(self, html, **kargs):
735 return self._og_search_property('description', html, fatal=False, **kargs)
737 def _og_search_title(self, html, **kargs):
738 return self._og_search_property('title', html, **kargs)
740 def _og_search_video_url(self, html, name='video url', secure=True, **kargs):
741 regexes = self._og_regexes('video') + self._og_regexes('video:url')
743 regexes = self._og_regexes('video:secure_url') + regexes
744 return self._html_search_regex(regexes, html, name, **kargs)
746 def _og_search_url(self, html, **kargs):
747 return self._og_search_property('url', html, **kargs)
749 def _html_search_meta(self, name, html, display_name=None, fatal=False, **kwargs):
750 if display_name is None:
752 return self._html_search_regex(
753 self._meta_regex(name),
754 html, display_name, fatal=fatal, group='content', **kwargs)
756 def _dc_search_uploader(self, html):
757 return self._html_search_meta('dc.creator', html, 'uploader')
759 def _rta_search(self, html):
760 # See http://www.rtalabel.org/index.php?content=howtofaq#single
761 if re.search(r'(?ix)<meta\s+name="rating"\s+'
762 r' content="RTA-5042-1996-1400-1577-RTA"',
767 def _media_rating_search(self, html):
768 # See http://www.tjg-designs.com/WP/metadata-code-examples-adding-metadata-to-your-web-pages/
769 rating = self._html_search_meta('rating', html)
781 return RATING_TABLE.get(rating.lower())
783 def _family_friendly_search(self, html):
784 # See http://schema.org/VideoObject
785 family_friendly = self._html_search_meta('isFamilyFriendly', html)
787 if not family_friendly:
796 return RATING_TABLE.get(family_friendly.lower())
798 def _twitter_search_player(self, html):
799 return self._html_search_meta('twitter:player', html,
800 'twitter card player')
802 def _search_json_ld(self, html, video_id, **kwargs):
803 json_ld = self._search_regex(
804 r'(?s)<script[^>]+type=(["\'])application/ld\+json\1[^>]*>(?P<json_ld>.+?)</script>',
805 html, 'JSON-LD', group='json_ld', **kwargs)
808 return self._json_ld(json_ld, video_id, fatal=kwargs.get('fatal', True))
810 def _json_ld(self, json_ld, video_id, fatal=True):
811 if isinstance(json_ld, compat_str):
812 json_ld = self._parse_json(json_ld, video_id, fatal=fatal)
816 if json_ld.get('@context') == 'http://schema.org':
817 item_type = json_ld.get('@type')
818 if item_type == 'TVEpisode':
820 'episode': unescapeHTML(json_ld.get('name')),
821 'episode_number': int_or_none(json_ld.get('episodeNumber')),
822 'description': unescapeHTML(json_ld.get('description')),
824 part_of_season = json_ld.get('partOfSeason')
825 if isinstance(part_of_season, dict) and part_of_season.get('@type') == 'TVSeason':
826 info['season_number'] = int_or_none(part_of_season.get('seasonNumber'))
827 part_of_series = json_ld.get('partOfSeries')
828 if isinstance(part_of_series, dict) and part_of_series.get('@type') == 'TVSeries':
829 info['series'] = unescapeHTML(part_of_series.get('name'))
830 elif item_type == 'Article':
832 'timestamp': parse_iso8601(json_ld.get('datePublished')),
833 'title': unescapeHTML(json_ld.get('headline')),
834 'description': unescapeHTML(json_ld.get('articleBody')),
836 return dict((k, v) for k, v in info.items() if v is not None)
839 def _hidden_inputs(html):
840 html = re.sub(r'<!--(?:(?!<!--).)*-->', '', html)
842 for input in re.findall(r'(?i)<input([^>]+)>', html):
843 if not re.search(r'type=(["\'])(?:hidden|submit)\1', input):
845 name = re.search(r'(?:name|id)=(["\'])(?P<value>.+?)\1', input)
848 value = re.search(r'value=(["\'])(?P<value>.*?)\1', input)
851 hidden_inputs[name.group('value')] = value.group('value')
854 def _form_hidden_inputs(self, form_id, html):
855 form = self._search_regex(
856 r'(?is)<form[^>]+?id=(["\'])%s\1[^>]*>(?P<form>.+?)</form>' % form_id,
857 html, '%s form' % form_id, group='form')
858 return self._hidden_inputs(form)
860 def _sort_formats(self, formats, field_preference=None):
862 raise ExtractorError('No video formats found')
865 # Automatically determine tbr when missing based on abr and vbr (improves
866 # formats sorting in some cases)
867 if 'tbr' not in f and f.get('abr') is not None and f.get('vbr') is not None:
868 f['tbr'] = f['abr'] + f['vbr']
871 # TODO remove the following workaround
872 from ..utils import determine_ext
873 if not f.get('ext') and 'url' in f:
874 f['ext'] = determine_ext(f['url'])
876 if isinstance(field_preference, (list, tuple)):
877 return tuple(f.get(field) if f.get(field) is not None else -1 for field in field_preference)
879 preference = f.get('preference')
880 if preference is None:
882 if f.get('ext') in ['f4f', 'f4m']: # Not yet supported
885 proto_preference = 0 if determine_protocol(f) in ['http', 'https'] else -0.1
887 if f.get('vcodec') == 'none': # audio only
889 if self._downloader.params.get('prefer_free_formats'):
890 ORDER = ['aac', 'mp3', 'm4a', 'webm', 'ogg', 'opus']
892 ORDER = ['webm', 'opus', 'ogg', 'mp3', 'aac', 'm4a']
895 audio_ext_preference = ORDER.index(f['ext'])
897 audio_ext_preference = -1
899 if f.get('acodec') == 'none': # video only
901 if self._downloader.params.get('prefer_free_formats'):
902 ORDER = ['flv', 'mp4', 'webm']
904 ORDER = ['webm', 'flv', 'mp4']
906 ext_preference = ORDER.index(f['ext'])
909 audio_ext_preference = 0
913 f.get('language_preference') if f.get('language_preference') is not None else -1,
914 f.get('quality') if f.get('quality') is not None else -1,
915 f.get('tbr') if f.get('tbr') is not None else -1,
916 f.get('filesize') if f.get('filesize') is not None else -1,
917 f.get('vbr') if f.get('vbr') is not None else -1,
918 f.get('height') if f.get('height') is not None else -1,
919 f.get('width') if f.get('width') is not None else -1,
922 f.get('abr') if f.get('abr') is not None else -1,
923 audio_ext_preference,
924 f.get('fps') if f.get('fps') is not None else -1,
925 f.get('filesize_approx') if f.get('filesize_approx') is not None else -1,
926 f.get('source_preference') if f.get('source_preference') is not None else -1,
927 f.get('format_id') if f.get('format_id') is not None else '',
929 formats.sort(key=_formats_key)
931 def _check_formats(self, formats, video_id):
934 lambda f: self._is_valid_url(
936 item='%s video format' % f.get('format_id') if f.get('format_id') else 'video'),
940 def _remove_duplicate_formats(formats):
944 if f['url'] not in format_urls:
945 format_urls.add(f['url'])
946 unique_formats.append(f)
947 formats[:] = unique_formats
949 def _is_valid_url(self, url, video_id, item='video'):
950 url = self._proto_relative_url(url, scheme='http:')
951 # For now assume non HTTP(S) URLs always valid
952 if not (url.startswith('http://') or url.startswith('https://')):
955 self._request_webpage(url, video_id, 'Checking %s URL' % item)
957 except ExtractorError as e:
958 if isinstance(e.cause, compat_urllib_error.URLError):
960 '%s: %s URL is invalid, skipping' % (video_id, item))
964 def http_scheme(self):
965 """ Either "http:" or "https:", depending on the user's preferences """
968 if self._downloader.params.get('prefer_insecure', False)
971 def _proto_relative_url(self, url, scheme=None):
974 if url.startswith('//'):
976 scheme = self.http_scheme()
981 def _sleep(self, timeout, video_id, msg_template=None):
982 if msg_template is None:
983 msg_template = '%(video_id)s: Waiting for %(timeout)s seconds'
984 msg = msg_template % {'video_id': video_id, 'timeout': timeout}
988 def _extract_f4m_formats(self, manifest_url, video_id, preference=None, f4m_id=None,
989 transform_source=lambda s: fix_xml_ampersands(s).strip(),
991 manifest = self._download_xml(
992 manifest_url, video_id, 'Downloading f4m manifest',
993 'Unable to download f4m manifest',
994 # Some manifests may be malformed, e.g. prosiebensat1 generated manifests
995 # (see https://github.com/rg3/youtube-dl/issues/6215#issuecomment-121704244)
996 transform_source=transform_source,
999 if manifest is False:
1002 return self._parse_f4m_formats(
1003 manifest, manifest_url, video_id, preference=preference, f4m_id=f4m_id,
1004 transform_source=transform_source, fatal=fatal)
1006 def _parse_f4m_formats(self, manifest, manifest_url, video_id, preference=None, f4m_id=None,
1007 transform_source=lambda s: fix_xml_ampersands(s).strip(),
1009 # currently youtube-dl cannot decode the playerVerificationChallenge as Akamai uses Adobe Alchemy
1010 akamai_pv = manifest.find('{http://ns.adobe.com/f4m/1.0}pv-2.0')
1011 if akamai_pv is not None and ';' in akamai_pv.text:
1012 playerVerificationChallenge = akamai_pv.text.split(';')[0]
1013 if playerVerificationChallenge.strip() != '':
1017 manifest_version = '1.0'
1018 media_nodes = manifest.findall('{http://ns.adobe.com/f4m/1.0}media')
1020 manifest_version = '2.0'
1021 media_nodes = manifest.findall('{http://ns.adobe.com/f4m/2.0}media')
1022 # Remove unsupported DRM protected media from final formats
1023 # rendition (see https://github.com/rg3/youtube-dl/issues/8573).
1024 media_nodes = remove_encrypted_media(media_nodes)
1027 base_url = xpath_text(
1028 manifest, ['{http://ns.adobe.com/f4m/1.0}baseURL', '{http://ns.adobe.com/f4m/2.0}baseURL'],
1029 'base URL', default=None)
1031 base_url = base_url.strip()
1032 for i, media_el in enumerate(media_nodes):
1033 if manifest_version == '2.0':
1034 media_url = media_el.attrib.get('href') or media_el.attrib.get('url')
1038 media_url if media_url.startswith('http://') or media_url.startswith('https://')
1039 else ((base_url or '/'.join(manifest_url.split('/')[:-1])) + '/' + media_url))
1040 # If media_url is itself a f4m manifest do the recursive extraction
1041 # since bitrates in parent manifest (this one) and media_url manifest
1042 # may differ leading to inability to resolve the format by requested
1043 # bitrate in f4m downloader
1044 if determine_ext(manifest_url) == 'f4m':
1045 formats.extend(self._extract_f4m_formats(
1046 manifest_url, video_id, preference=preference, f4m_id=f4m_id,
1047 transform_source=transform_source, fatal=fatal))
1049 tbr = int_or_none(media_el.attrib.get('bitrate'))
1051 'format_id': '-'.join(filter(None, [f4m_id, compat_str(i if tbr is None else tbr)])),
1052 'url': manifest_url,
1055 'width': int_or_none(media_el.attrib.get('width')),
1056 'height': int_or_none(media_el.attrib.get('height')),
1057 'preference': preference,
1061 def _m3u8_meta_format(self, m3u8_url, ext=None, preference=None, m3u8_id=None):
1063 'format_id': '-'.join(filter(None, [m3u8_id, 'meta'])),
1067 'preference': preference - 1 if preference else -1,
1068 'resolution': 'multiple',
1069 'format_note': 'Quality selection URL',
1072 def _extract_m3u8_formats(self, m3u8_url, video_id, ext=None,
1073 entry_protocol='m3u8', preference=None,
1074 m3u8_id=None, note=None, errnote=None,
1075 fatal=True, live=False):
1077 formats = [self._m3u8_meta_format(m3u8_url, ext, preference, m3u8_id)]
1079 format_url = lambda u: (
1081 if re.match(r'^https?://', u)
1082 else compat_urlparse.urljoin(m3u8_url, u))
1084 res = self._download_webpage_handle(
1086 note=note or 'Downloading m3u8 information',
1087 errnote=errnote or 'Failed to download m3u8 information',
1091 m3u8_doc, urlh = res
1092 m3u8_url = urlh.geturl()
1094 # We should try extracting formats only from master playlists [1], i.e.
1095 # playlists that describe available qualities. On the other hand media
1096 # playlists [2] should be returned as is since they contain just the media
1097 # without qualities renditions.
1098 # Fortunately, master playlist can be easily distinguished from media
1099 # playlist based on particular tags availability. As of [1, 2] master
1100 # playlist tags MUST NOT appear in a media playist and vice versa.
1101 # As of [3] #EXT-X-TARGETDURATION tag is REQUIRED for every media playlist
1102 # and MUST NOT appear in master playlist thus we can clearly detect media
1103 # playlist with this criterion.
1104 # 1. https://tools.ietf.org/html/draft-pantos-http-live-streaming-17#section-4.3.4
1105 # 2. https://tools.ietf.org/html/draft-pantos-http-live-streaming-17#section-4.3.3
1106 # 3. https://tools.ietf.org/html/draft-pantos-http-live-streaming-17#section-4.3.3.1
1107 if '#EXT-X-TARGETDURATION' in m3u8_doc: # media playlist, return as is
1110 'format_id': m3u8_id,
1112 'protocol': entry_protocol,
1113 'preference': preference,
1117 kv_rex = re.compile(
1118 r'(?P<key>[a-zA-Z_-]+)=(?P<val>"[^"]+"|[^",]+)(?:,|$)')
1119 for line in m3u8_doc.splitlines():
1120 if line.startswith('#EXT-X-STREAM-INF:'):
1122 for m in kv_rex.finditer(line):
1124 if v.startswith('"'):
1126 last_info[m.group('key')] = v
1127 elif line.startswith('#EXT-X-MEDIA:'):
1129 for m in kv_rex.finditer(line):
1131 if v.startswith('"'):
1133 last_media[m.group('key')] = v
1134 elif line.startswith('#') or not line.strip():
1137 if last_info is None:
1138 formats.append({'url': format_url(line)})
1140 tbr = int_or_none(last_info.get('BANDWIDTH'), scale=1000)
1143 format_id.append(m3u8_id)
1144 last_media_name = last_media.get('NAME') if last_media and last_media.get('TYPE') not in ('SUBTITLES', 'CLOSED-CAPTIONS') else None
1145 # Despite specification does not mention NAME attribute for
1146 # EXT-X-STREAM-INF it still sometimes may be present
1147 stream_name = last_info.get('NAME') or last_media_name
1148 # Bandwidth of live streams may differ over time thus making
1149 # format_id unpredictable. So it's better to keep provided
1152 format_id.append(stream_name if stream_name else '%d' % (tbr if tbr else len(formats)))
1154 'format_id': '-'.join(format_id),
1155 'url': format_url(line.strip()),
1158 'protocol': entry_protocol,
1159 'preference': preference,
1161 resolution = last_info.get('RESOLUTION')
1163 width_str, height_str = resolution.split('x')
1164 f['width'] = int(width_str)
1165 f['height'] = int(height_str)
1166 codecs = last_info.get('CODECS')
1168 vcodec, acodec = [None] * 2
1169 va_codecs = codecs.split(',')
1170 if len(va_codecs) == 1:
1171 # Audio only entries usually come with single codec and
1172 # no resolution. For more robustness we also check it to
1174 if not resolution and va_codecs[0].startswith('mp4a'):
1175 vcodec, acodec = 'none', va_codecs[0]
1177 vcodec = va_codecs[0]
1179 vcodec, acodec = va_codecs[:2]
1184 if last_media is not None:
1185 f['m3u8_media'] = last_media
1192 def _xpath_ns(path, namespace=None):
1196 for c in path.split('/'):
1197 if not c or c == '.':
1200 out.append('{%s}%s' % (namespace, c))
1201 return '/'.join(out)
1203 def _extract_smil_formats(self, smil_url, video_id, fatal=True, f4m_params=None, transform_source=None):
1204 smil = self._download_smil(smil_url, video_id, fatal=fatal, transform_source=transform_source)
1210 namespace = self._parse_smil_namespace(smil)
1212 return self._parse_smil_formats(
1213 smil, smil_url, video_id, namespace=namespace, f4m_params=f4m_params)
1215 def _extract_smil_info(self, smil_url, video_id, fatal=True, f4m_params=None):
1216 smil = self._download_smil(smil_url, video_id, fatal=fatal)
1219 return self._parse_smil(smil, smil_url, video_id, f4m_params=f4m_params)
1221 def _download_smil(self, smil_url, video_id, fatal=True, transform_source=None):
1222 return self._download_xml(
1223 smil_url, video_id, 'Downloading SMIL file',
1224 'Unable to download SMIL file', fatal=fatal, transform_source=transform_source)
1226 def _parse_smil(self, smil, smil_url, video_id, f4m_params=None):
1227 namespace = self._parse_smil_namespace(smil)
1229 formats = self._parse_smil_formats(
1230 smil, smil_url, video_id, namespace=namespace, f4m_params=f4m_params)
1231 subtitles = self._parse_smil_subtitles(smil, namespace=namespace)
1233 video_id = os.path.splitext(url_basename(smil_url))[0]
1237 for meta in smil.findall(self._xpath_ns('./head/meta', namespace)):
1238 name = meta.attrib.get('name')
1239 content = meta.attrib.get('content')
1240 if not name or not content:
1242 if not title and name == 'title':
1244 elif not description and name in ('description', 'abstract'):
1245 description = content
1246 elif not upload_date and name == 'date':
1247 upload_date = unified_strdate(content)
1250 'id': image.get('type'),
1251 'url': image.get('src'),
1252 'width': int_or_none(image.get('width')),
1253 'height': int_or_none(image.get('height')),
1254 } for image in smil.findall(self._xpath_ns('.//image', namespace)) if image.get('src')]
1258 'title': title or video_id,
1259 'description': description,
1260 'upload_date': upload_date,
1261 'thumbnails': thumbnails,
1263 'subtitles': subtitles,
1266 def _parse_smil_namespace(self, smil):
1267 return self._search_regex(
1268 r'(?i)^{([^}]+)?}smil$', smil.tag, 'namespace', default=None)
1270 def _parse_smil_formats(self, smil, smil_url, video_id, namespace=None, f4m_params=None, transform_rtmp_url=None):
1272 for meta in smil.findall(self._xpath_ns('./head/meta', namespace)):
1273 b = meta.get('base') or meta.get('httpBase')
1284 media = smil.findall(self._xpath_ns('.//video', namespace)) + smil.findall(self._xpath_ns('.//audio', namespace))
1285 for medium in media:
1286 src = medium.get('src')
1287 if not src or src in srcs:
1291 bitrate = float_or_none(medium.get('system-bitrate') or medium.get('systemBitrate'), 1000)
1292 filesize = int_or_none(medium.get('size') or medium.get('fileSize'))
1293 width = int_or_none(medium.get('width'))
1294 height = int_or_none(medium.get('height'))
1295 proto = medium.get('proto')
1296 ext = medium.get('ext')
1297 src_ext = determine_ext(src)
1298 streamer = medium.get('streamer') or base
1300 if proto == 'rtmp' or streamer.startswith('rtmp'):
1306 'format_id': 'rtmp-%d' % (rtmp_count if bitrate is None else bitrate),
1308 'filesize': filesize,
1312 if transform_rtmp_url:
1313 streamer, src = transform_rtmp_url(streamer, src)
1314 formats[-1].update({
1320 src_url = src if src.startswith('http') else compat_urlparse.urljoin(base, src)
1321 src_url = src_url.strip()
1323 if proto == 'm3u8' or src_ext == 'm3u8':
1324 m3u8_formats = self._extract_m3u8_formats(
1325 src_url, video_id, ext or 'mp4', m3u8_id='hls', fatal=False)
1326 if len(m3u8_formats) == 1:
1328 m3u8_formats[0].update({
1329 'format_id': 'hls-%d' % (m3u8_count if bitrate is None else bitrate),
1334 formats.extend(m3u8_formats)
1337 if src_ext == 'f4m':
1342 'plugin': 'flowplayer-3.2.0.1',
1344 f4m_url += '&' if '?' in f4m_url else '?'
1345 f4m_url += compat_urllib_parse_urlencode(f4m_params)
1346 formats.extend(self._extract_f4m_formats(f4m_url, video_id, f4m_id='hds', fatal=False))
1349 if src_url.startswith('http') and self._is_valid_url(src, video_id):
1353 'ext': ext or src_ext or 'flv',
1354 'format_id': 'http-%d' % (bitrate or http_count),
1356 'filesize': filesize,
1364 def _parse_smil_subtitles(self, smil, namespace=None, subtitles_lang='en'):
1367 for num, textstream in enumerate(smil.findall(self._xpath_ns('.//textstream', namespace))):
1368 src = textstream.get('src')
1369 if not src or src in urls:
1372 ext = textstream.get('ext') or mimetype2ext(textstream.get('type')) or determine_ext(src)
1373 lang = textstream.get('systemLanguage') or textstream.get('systemLanguageName') or textstream.get('lang') or subtitles_lang
1374 subtitles.setdefault(lang, []).append({
1380 def _extract_xspf_playlist(self, playlist_url, playlist_id, fatal=True):
1381 xspf = self._download_xml(
1382 playlist_url, playlist_id, 'Downloading xpsf playlist',
1383 'Unable to download xspf manifest', fatal=fatal)
1386 return self._parse_xspf(xspf, playlist_id)
1388 def _parse_xspf(self, playlist, playlist_id):
1390 'xspf': 'http://xspf.org/ns/0/',
1391 's1': 'http://static.streamone.nl/player/ns/0',
1395 for track in playlist.findall(xpath_with_ns('./xspf:trackList/xspf:track', NS_MAP)):
1397 track, xpath_with_ns('./xspf:title', NS_MAP), 'title', default=playlist_id)
1398 description = xpath_text(
1399 track, xpath_with_ns('./xspf:annotation', NS_MAP), 'description')
1400 thumbnail = xpath_text(
1401 track, xpath_with_ns('./xspf:image', NS_MAP), 'thumbnail')
1402 duration = float_or_none(
1403 xpath_text(track, xpath_with_ns('./xspf:duration', NS_MAP), 'duration'), 1000)
1406 'url': location.text,
1407 'format_id': location.get(xpath_with_ns('s1:label', NS_MAP)),
1408 'width': int_or_none(location.get(xpath_with_ns('s1:width', NS_MAP))),
1409 'height': int_or_none(location.get(xpath_with_ns('s1:height', NS_MAP))),
1410 } for location in track.findall(xpath_with_ns('./xspf:location', NS_MAP))]
1411 self._sort_formats(formats)
1416 'description': description,
1417 'thumbnail': thumbnail,
1418 'duration': duration,
1423 def _extract_mpd_formats(self, mpd_url, video_id, mpd_id=None, note=None, errnote=None, fatal=True, formats_dict={}):
1424 res = self._download_webpage_handle(
1426 note=note or 'Downloading MPD manifest',
1427 errnote=errnote or 'Failed to download MPD manifest',
1432 mpd_base_url = re.match(r'https?://.+/', urlh.geturl()).group()
1434 return self._parse_mpd_formats(
1435 compat_etree_fromstring(mpd.encode('utf-8')), mpd_id, mpd_base_url, formats_dict=formats_dict)
1437 def _parse_mpd_formats(self, mpd_doc, mpd_id=None, mpd_base_url='', formats_dict={}):
1438 if mpd_doc.get('type') == 'dynamic':
1441 namespace = self._search_regex(r'(?i)^{([^}]+)?}MPD$', mpd_doc.tag, 'namespace', default=None)
1444 return self._xpath_ns(path, namespace)
1446 def is_drm_protected(element):
1447 return element.find(_add_ns('ContentProtection')) is not None
1449 def extract_multisegment_info(element, ms_parent_info):
1450 ms_info = ms_parent_info.copy()
1451 segment_list = element.find(_add_ns('SegmentList'))
1452 if segment_list is not None:
1453 segment_urls_e = segment_list.findall(_add_ns('SegmentURL'))
1455 ms_info['segment_urls'] = [segment.attrib['media'] for segment in segment_urls_e]
1456 initialization = segment_list.find(_add_ns('Initialization'))
1457 if initialization is not None:
1458 ms_info['initialization_url'] = initialization.attrib['sourceURL']
1460 segment_template = element.find(_add_ns('SegmentTemplate'))
1461 if segment_template is not None:
1462 start_number = segment_template.get('startNumber')
1464 ms_info['start_number'] = int(start_number)
1465 segment_timeline = segment_template.find(_add_ns('SegmentTimeline'))
1466 if segment_timeline is not None:
1467 s_e = segment_timeline.findall(_add_ns('S'))
1469 ms_info['total_number'] = 0
1471 ms_info['total_number'] += 1 + int(s.get('r', '0'))
1473 timescale = segment_template.get('timescale')
1475 ms_info['timescale'] = int(timescale)
1476 segment_duration = segment_template.get('duration')
1477 if segment_duration:
1478 ms_info['segment_duration'] = int(segment_duration)
1479 media_template = segment_template.get('media')
1481 ms_info['media_template'] = media_template
1482 initialization = segment_template.get('initialization')
1484 ms_info['initialization_url'] = initialization
1486 initialization = segment_template.find(_add_ns('Initialization'))
1487 if initialization is not None:
1488 ms_info['initialization_url'] = initialization.attrib['sourceURL']
1491 mpd_duration = parse_duration(mpd_doc.get('mediaPresentationDuration'))
1493 for period in mpd_doc.findall(_add_ns('Period')):
1494 period_duration = parse_duration(period.get('duration')) or mpd_duration
1495 period_ms_info = extract_multisegment_info(period, {
1499 for adaptation_set in period.findall(_add_ns('AdaptationSet')):
1500 if is_drm_protected(adaptation_set):
1502 adaption_set_ms_info = extract_multisegment_info(adaptation_set, period_ms_info)
1503 for representation in adaptation_set.findall(_add_ns('Representation')):
1504 if is_drm_protected(representation):
1506 representation_attrib = adaptation_set.attrib.copy()
1507 representation_attrib.update(representation.attrib)
1508 # According to page 41 of ISO/IEC 29001-1:2014, @mimeType is mandatory
1509 mime_type = representation_attrib['mimeType']
1510 content_type = mime_type.split('/')[0]
1511 if content_type == 'text':
1512 # TODO implement WebVTT downloading
1514 elif content_type == 'video' or content_type == 'audio':
1516 for element in (representation, adaptation_set, period, mpd_doc):
1517 base_url_e = element.find(_add_ns('BaseURL'))
1518 if base_url_e is not None:
1519 base_url = base_url_e.text + base_url
1520 if re.match(r'^https?://', base_url):
1522 if mpd_base_url and not re.match(r'^https?://', base_url):
1523 if not mpd_base_url.endswith('/') and not base_url.startswith('/'):
1525 base_url = mpd_base_url + base_url
1526 representation_id = representation_attrib.get('id')
1527 lang = representation_attrib.get('lang')
1528 url_el = representation.find(_add_ns('BaseURL'))
1529 filesize = int_or_none(url_el.attrib.get('{http://youtube.com/yt/2012/10/10}contentLength') if url_el is not None else None)
1531 'format_id': '%s-%s' % (mpd_id, representation_id) if mpd_id else representation_id,
1533 'ext': mimetype2ext(mime_type),
1534 'width': int_or_none(representation_attrib.get('width')),
1535 'height': int_or_none(representation_attrib.get('height')),
1536 'tbr': int_or_none(representation_attrib.get('bandwidth'), 1000),
1537 'asr': int_or_none(representation_attrib.get('audioSamplingRate')),
1538 'fps': int_or_none(representation_attrib.get('frameRate')),
1539 'vcodec': 'none' if content_type == 'audio' else representation_attrib.get('codecs'),
1540 'acodec': 'none' if content_type == 'video' else representation_attrib.get('codecs'),
1541 'language': lang if lang not in ('mul', 'und', 'zxx', 'mis') else None,
1542 'format_note': 'DASH %s' % content_type,
1543 'filesize': filesize,
1545 representation_ms_info = extract_multisegment_info(representation, adaption_set_ms_info)
1546 if 'segment_urls' not in representation_ms_info and 'media_template' in representation_ms_info:
1547 if 'total_number' not in representation_ms_info and 'segment_duration':
1548 segment_duration = float(representation_ms_info['segment_duration']) / float(representation_ms_info['timescale'])
1549 representation_ms_info['total_number'] = int(math.ceil(float(period_duration) / segment_duration))
1550 media_template = representation_ms_info['media_template']
1551 media_template = media_template.replace('$RepresentationID$', representation_id)
1552 media_template = re.sub(r'\$(Number|Bandwidth)\$', r'%(\1)d', media_template)
1553 media_template = re.sub(r'\$(Number|Bandwidth)%([^$]+)\$', r'%(\1)\2', media_template)
1554 media_template.replace('$$', '$')
1555 representation_ms_info['segment_urls'] = [
1557 'Number': segment_number,
1558 'Bandwidth': representation_attrib.get('bandwidth')}
1559 for segment_number in range(
1560 representation_ms_info['start_number'],
1561 representation_ms_info['total_number'] + representation_ms_info['start_number'])]
1562 if 'segment_urls' in representation_ms_info:
1564 'segment_urls': representation_ms_info['segment_urls'],
1565 'protocol': 'http_dash_segments',
1567 if 'initialization_url' in representation_ms_info:
1568 initialization_url = representation_ms_info['initialization_url'].replace('$RepresentationID$', representation_id)
1570 'initialization_url': initialization_url,
1572 if not f.get('url'):
1573 f['url'] = initialization_url
1575 existing_format = next(
1576 fo for fo in formats
1577 if fo['format_id'] == representation_id)
1578 except StopIteration:
1579 full_info = formats_dict.get(representation_id, {}).copy()
1581 formats.append(full_info)
1583 existing_format.update(f)
1585 self.report_warning('Unknown MIME type %s in DASH manifest' % mime_type)
1588 def _live_title(self, name):
1589 """ Generate the title for a live video """
1590 now = datetime.datetime.now()
1591 now_str = now.strftime('%Y-%m-%d %H:%M')
1592 return name + ' ' + now_str
1594 def _int(self, v, name, fatal=False, **kwargs):
1595 res = int_or_none(v, **kwargs)
1596 if 'get_attr' in kwargs:
1597 print(getattr(v, kwargs['get_attr']))
1599 msg = 'Failed to extract %s: Could not parse value %r' % (name, v)
1601 raise ExtractorError(msg)
1603 self._downloader.report_warning(msg)
1606 def _float(self, v, name, fatal=False, **kwargs):
1607 res = float_or_none(v, **kwargs)
1609 msg = 'Failed to extract %s: Could not parse value %r' % (name, v)
1611 raise ExtractorError(msg)
1613 self._downloader.report_warning(msg)
1616 def _set_cookie(self, domain, name, value, expire_time=None):
1617 cookie = compat_cookiejar.Cookie(
1618 0, name, value, None, None, domain, None,
1619 None, '/', True, False, expire_time, '', None, None, None)
1620 self._downloader.cookiejar.set_cookie(cookie)
1622 def _get_cookies(self, url):
1623 """ Return a compat_cookies.SimpleCookie with the cookies for the url """
1624 req = sanitized_Request(url)
1625 self._downloader.cookiejar.add_cookie_header(req)
1626 return compat_cookies.SimpleCookie(req.get_header('Cookie'))
1628 def get_testcases(self, include_onlymatching=False):
1629 t = getattr(self, '_TEST', None)
1631 assert not hasattr(self, '_TESTS'), \
1632 '%s has _TEST and _TESTS' % type(self).__name__
1635 tests = getattr(self, '_TESTS', [])
1637 if not include_onlymatching and t.get('only_matching', False):
1639 t['name'] = type(self).__name__[:-len('IE')]
1642 def is_suitable(self, age_limit):
1643 """ Test whether the extractor is generally suitable for the given
1644 age limit (i.e. pornographic sites are not, all others usually are) """
1646 any_restricted = False
1647 for tc in self.get_testcases(include_onlymatching=False):
1648 if 'playlist' in tc:
1649 tc = tc['playlist'][0]
1650 is_restricted = age_restricted(
1651 tc.get('info_dict', {}).get('age_limit'), age_limit)
1652 if not is_restricted:
1654 any_restricted = any_restricted or is_restricted
1655 return not any_restricted
1657 def extract_subtitles(self, *args, **kwargs):
1658 if (self._downloader.params.get('writesubtitles', False) or
1659 self._downloader.params.get('listsubtitles')):
1660 return self._get_subtitles(*args, **kwargs)
1663 def _get_subtitles(self, *args, **kwargs):
1664 raise NotImplementedError('This method must be implemented by subclasses')
1667 def _merge_subtitle_items(subtitle_list1, subtitle_list2):
1668 """ Merge subtitle items for one language. Items with duplicated URLs
1669 will be dropped. """
1670 list1_urls = set([item['url'] for item in subtitle_list1])
1671 ret = list(subtitle_list1)
1672 ret.extend([item for item in subtitle_list2 if item['url'] not in list1_urls])
1676 def _merge_subtitles(cls, subtitle_dict1, subtitle_dict2):
1677 """ Merge two subtitle dictionaries, language by language. """
1678 ret = dict(subtitle_dict1)
1679 for lang in subtitle_dict2:
1680 ret[lang] = cls._merge_subtitle_items(subtitle_dict1.get(lang, []), subtitle_dict2[lang])
1683 def extract_automatic_captions(self, *args, **kwargs):
1684 if (self._downloader.params.get('writeautomaticsub', False) or
1685 self._downloader.params.get('listsubtitles')):
1686 return self._get_automatic_captions(*args, **kwargs)
1689 def _get_automatic_captions(self, *args, **kwargs):
1690 raise NotImplementedError('This method must be implemented by subclasses')
1692 def mark_watched(self, *args, **kwargs):
1693 if (self._downloader.params.get('mark_watched', False) and
1694 (self._get_login_info()[0] is not None or
1695 self._downloader.params.get('cookiefile') is not None)):
1696 self._mark_watched(*args, **kwargs)
1698 def _mark_watched(self, *args, **kwargs):
1699 raise NotImplementedError('This method must be implemented by subclasses')
1702 class SearchInfoExtractor(InfoExtractor):
1704 Base class for paged search queries extractors.
1705 They accept URLs in the format _SEARCH_KEY(|all|[0-9]):{query}
1706 Instances should define _SEARCH_KEY and _MAX_RESULTS.
1710 def _make_valid_url(cls):
1711 return r'%s(?P<prefix>|[1-9][0-9]*|all):(?P<query>[\s\S]+)' % cls._SEARCH_KEY
1714 def suitable(cls, url):
1715 return re.match(cls._make_valid_url(), url) is not None
1717 def _real_extract(self, query):
1718 mobj = re.match(self._make_valid_url(), query)
1720 raise ExtractorError('Invalid search query "%s"' % query)
1722 prefix = mobj.group('prefix')
1723 query = mobj.group('query')
1725 return self._get_n_results(query, 1)
1726 elif prefix == 'all':
1727 return self._get_n_results(query, self._MAX_RESULTS)
1731 raise ExtractorError('invalid download number %s for query "%s"' % (n, query))
1732 elif n > self._MAX_RESULTS:
1733 self._downloader.report_warning('%s returns max %i results (you requested %i)' % (self._SEARCH_KEY, self._MAX_RESULTS, n))
1734 n = self._MAX_RESULTS
1735 return self._get_n_results(query, n)
1737 def _get_n_results(self, query, n):
1738 """Get a specified number of results for a query"""
1739 raise NotImplementedError('This method must be implemented by subclasses')
1742 def SEARCH_KEY(self):
1743 return self._SEARCH_KEY