1 from __future__ import unicode_literals
15 from ..compat import (
18 compat_etree_fromstring,
24 compat_urllib_parse_urlencode,
27 from ..downloader.f4m import remove_encrypted_media
56 class InfoExtractor(object):
57 """Information Extractor class.
59 Information extractors are the classes that, given a URL, extract
60 information about the video (or videos) the URL refers to. This
61 information includes the real video URL, the video title, author and
62 others. The information is stored in a dictionary which is then
63 passed to the YoutubeDL. The YoutubeDL processes this
64 information possibly downloading the video to the file system, among
65 other possible outcomes.
67 The type field determines the type of the result.
68 By far the most common value (and the default if _type is missing) is
69 "video", which indicates a single video.
71 For a video, the dictionaries must include the following fields:
74 title: Video title, unescaped.
76 Additionally, it must contain either a formats entry or a url one:
78 formats: A list of dictionaries for each format available, ordered
79 from worst to best quality.
82 * url Mandatory. The URL of the video file
83 * ext Will be calculated from URL if missing
84 * format A human-readable description of the format
85 ("mp4 container with h264/opus").
86 Calculated from the format_id, width, height.
87 and format_note fields if missing.
88 * format_id A short description of the format
89 ("mp4_h264_opus" or "19").
90 Technically optional, but strongly recommended.
91 * format_note Additional info about the format
92 ("3D" or "DASH video")
93 * width Width of the video, if known
94 * height Height of the video, if known
95 * resolution Textual description of width and height
96 * tbr Average bitrate of audio and video in KBit/s
97 * abr Average audio bitrate in KBit/s
98 * acodec Name of the audio codec in use
99 * asr Audio sampling rate in Hertz
100 * vbr Average video bitrate in KBit/s
102 * vcodec Name of the video codec in use
103 * container Name of the container format
104 * filesize The number of bytes, if known in advance
105 * filesize_approx An estimate for the number of bytes
106 * player_url SWF Player URL (used for rtmpdump).
107 * protocol The protocol that will be used for the actual
108 download, lower-case.
109 "http", "https", "rtsp", "rtmp", "rtmpe",
110 "m3u8", "m3u8_native" or "http_dash_segments".
111 * preference Order number of this format. If this field is
112 present and not None, the formats get sorted
113 by this field, regardless of all other values.
114 -1 for default (order by other properties),
115 -2 or smaller for less than default.
116 < -1000 to hide the format (if there is
117 another one which is strictly better)
118 * language Language code, e.g. "de" or "en-US".
119 * language_preference Is this in the language mentioned in
121 10 if it's what the URL is about,
122 -1 for default (don't know),
123 -10 otherwise, other values reserved for now.
124 * quality Order number of the video quality of this
125 format, irrespective of the file format.
126 -1 for default (order by other properties),
127 -2 or smaller for less than default.
128 * source_preference Order number for this video source
129 (quality takes higher priority)
130 -1 for default (order by other properties),
131 -2 or smaller for less than default.
132 * http_headers A dictionary of additional HTTP headers
133 to add to the request.
134 * stretched_ratio If given and not 1, indicates that the
135 video's pixels are not square.
136 width : height ratio as float.
137 * no_resume The server does not support resuming the
138 (HTTP or RTMP) download. Boolean.
140 url: Final video URL.
141 ext: Video filename extension.
142 format: The video format, defaults to ext (used for --get-format)
143 player_url: SWF Player URL (used for rtmpdump).
145 The following fields are optional:
147 alt_title: A secondary title of the video.
148 display_id An alternative identifier for the video, not necessarily
149 unique, but available before title. Typically, id is
150 something like "4234987", title "Dancing naked mole rats",
151 and display_id "dancing-naked-mole-rats"
152 thumbnails: A list of dictionaries, with the following entries:
153 * "id" (optional, string) - Thumbnail format ID
155 * "preference" (optional, int) - quality of the image
156 * "width" (optional, int)
157 * "height" (optional, int)
158 * "resolution" (optional, string "{width}x{height"},
160 thumbnail: Full URL to a video thumbnail image.
161 description: Full video description.
162 uploader: Full name of the video uploader.
163 license: License name the video is licensed under.
164 creator: The main artist who created the video.
165 release_date: The date (YYYYMMDD) when the video was released.
166 timestamp: UNIX timestamp of the moment the video became available.
167 upload_date: Video upload date (YYYYMMDD).
168 If not explicitly set, calculated from timestamp.
169 uploader_id: Nickname or id of the video uploader.
170 uploader_url: Full URL to a personal webpage of the video uploader.
171 location: Physical location where the video was filmed.
172 subtitles: The available subtitles as a dictionary in the format
173 {language: subformats}. "subformats" is a list sorted from
174 lower to higher preference, each element is a dictionary
175 with the "ext" entry and one of:
176 * "data": The subtitles file contents
177 * "url": A URL pointing to the subtitles file
178 "ext" will be calculated from URL if missing
179 automatic_captions: Like 'subtitles', used by the YoutubeIE for
180 automatically generated captions
181 duration: Length of the video in seconds, as an integer or float.
182 view_count: How many users have watched the video on the platform.
183 like_count: Number of positive ratings of the video
184 dislike_count: Number of negative ratings of the video
185 repost_count: Number of reposts of the video
186 average_rating: Average rating give by users, the scale used depends on the webpage
187 comment_count: Number of comments on the video
188 comments: A list of comments, each with one or more of the following
189 properties (all but one of text or html optional):
190 * "author" - human-readable name of the comment author
191 * "author_id" - user ID of the comment author
193 * "html" - Comment as HTML
194 * "text" - Plain text of the comment
195 * "timestamp" - UNIX timestamp of comment
196 * "parent" - ID of the comment this one is replying to.
197 Set to "root" to indicate that this is a
198 comment to the original video.
199 age_limit: Age restriction for the video, as an integer (years)
200 webpage_url: The URL to the video webpage, if given to youtube-dl it
201 should allow to get the same result again. (It will be set
202 by YoutubeDL if it's missing)
203 categories: A list of categories that the video falls in, for example
205 tags: A list of tags assigned to the video, e.g. ["sweden", "pop music"]
206 is_live: True, False, or None (=unknown). Whether this video is a
207 live stream that goes on instead of a fixed-length video.
208 start_time: Time in seconds where the reproduction should start, as
209 specified in the URL.
210 end_time: Time in seconds where the reproduction should end, as
211 specified in the URL.
213 The following fields should only be used when the video belongs to some logical
216 chapter: Name or title of the chapter the video belongs to.
217 chapter_number: Number of the chapter the video belongs to, as an integer.
218 chapter_id: Id of the chapter the video belongs to, as a unicode string.
220 The following fields should only be used when the video is an episode of some
223 series: Title of the series or programme the video episode belongs to.
224 season: Title of the season the video episode belongs to.
225 season_number: Number of the season the video episode belongs to, as an integer.
226 season_id: Id of the season the video episode belongs to, as a unicode string.
227 episode: Title of the video episode. Unlike mandatory video title field,
228 this field should denote the exact title of the video episode
229 without any kind of decoration.
230 episode_number: Number of the video episode within a season, as an integer.
231 episode_id: Id of the video episode, as a unicode string.
233 Unless mentioned otherwise, the fields should be Unicode strings.
235 Unless mentioned otherwise, None is equivalent to absence of information.
238 _type "playlist" indicates multiple videos.
239 There must be a key "entries", which is a list, an iterable, or a PagedList
240 object, each element of which is a valid dictionary by this specification.
242 Additionally, playlists can have "title", "description" and "id" attributes
243 with the same semantics as videos (see above).
246 _type "multi_video" indicates that there are multiple videos that
247 form a single show, for examples multiple acts of an opera or TV episode.
248 It must have an entries key like a playlist and contain all the keys
249 required for a video at the same time.
252 _type "url" indicates that the video must be extracted from another
253 location, possibly by a different extractor. Its only required key is:
254 "url" - the next URL to extract.
255 The key "ie_key" can be set to the class name (minus the trailing "IE",
256 e.g. "Youtube") if the extractor class is known in advance.
257 Additionally, the dictionary may have any properties of the resolved entity
258 known in advance, for example "title" if the title of the referred video is
262 _type "url_transparent" entities have the same specification as "url", but
263 indicate that the given additional information is more precise than the one
264 associated with the resolved URL.
265 This is useful when a site employs a video service that hosts the video and
266 its technical metadata, but that video service does not embed a useful
267 title, description etc.
270 Subclasses of this one should re-define the _real_initialize() and
271 _real_extract() methods and define a _VALID_URL regexp.
272 Probably, they should also be added to the list of extractors.
274 Finally, the _WORKING attribute should be set to False for broken IEs
275 in order to warn the users and skip the tests.
282 def __init__(self, downloader=None):
283 """Constructor. Receives an optional downloader."""
285 self.set_downloader(downloader)
288 def suitable(cls, url):
289 """Receives a URL and returns True if suitable for this IE."""
291 # This does not use has/getattr intentionally - we want to know whether
292 # we have cached the regexp for *this* class, whereas getattr would also
293 # match the superclass
294 if '_VALID_URL_RE' not in cls.__dict__:
295 cls._VALID_URL_RE = re.compile(cls._VALID_URL)
296 return cls._VALID_URL_RE.match(url) is not None
299 def _match_id(cls, url):
300 if '_VALID_URL_RE' not in cls.__dict__:
301 cls._VALID_URL_RE = re.compile(cls._VALID_URL)
302 m = cls._VALID_URL_RE.match(url)
308 """Getter method for _WORKING."""
311 def initialize(self):
312 """Initializes an instance (authentication, etc)."""
314 self._real_initialize()
317 def extract(self, url):
318 """Extracts URL information and returns it in list of dicts."""
321 return self._real_extract(url)
322 except ExtractorError:
324 except compat_http_client.IncompleteRead as e:
325 raise ExtractorError('A network error has occurred.', cause=e, expected=True)
326 except (KeyError, StopIteration) as e:
327 raise ExtractorError('An extractor error has occurred.', cause=e)
329 def set_downloader(self, downloader):
330 """Sets the downloader for this IE."""
331 self._downloader = downloader
333 def _real_initialize(self):
334 """Real initialization process. Redefine in subclasses."""
337 def _real_extract(self, url):
338 """Real extraction process. Redefine in subclasses."""
343 """A string for getting the InfoExtractor with get_info_extractor"""
344 return compat_str(cls.__name__[:-2])
348 return compat_str(type(self).__name__[:-2])
350 def _request_webpage(self, url_or_request, video_id, note=None, errnote=None, fatal=True, data=None, headers=None, query=None):
351 """ Returns the response handle """
353 self.report_download_webpage(video_id)
354 elif note is not False:
356 self.to_screen('%s' % (note,))
358 self.to_screen('%s: %s' % (video_id, note))
359 # data, headers and query params will be ignored for `Request` objects
360 if isinstance(url_or_request, compat_str):
362 url_or_request = update_url_query(url_or_request, query)
364 url_or_request = sanitized_Request(url_or_request, data, headers or {})
366 return self._downloader.urlopen(url_or_request)
367 except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
371 errnote = 'Unable to download webpage'
373 errmsg = '%s: %s' % (errnote, error_to_compat_str(err))
375 raise ExtractorError(errmsg, sys.exc_info()[2], cause=err)
377 self._downloader.report_warning(errmsg)
380 def _download_webpage_handle(self, url_or_request, video_id, note=None, errnote=None, fatal=True, encoding=None, data=None, headers=None, query=None):
381 """ Returns a tuple (page content as string, URL handle) """
382 # Strip hashes from the URL (#1038)
383 if isinstance(url_or_request, (compat_str, str)):
384 url_or_request = url_or_request.partition('#')[0]
386 urlh = self._request_webpage(url_or_request, video_id, note, errnote, fatal, data=data, headers=headers, query=query)
390 content = self._webpage_read_content(urlh, url_or_request, video_id, note, errnote, fatal, encoding=encoding)
391 return (content, urlh)
394 def _guess_encoding_from_content(content_type, webpage_bytes):
395 m = re.match(r'[a-zA-Z0-9_.-]+/[a-zA-Z0-9_.-]+\s*;\s*charset=(.+)', content_type)
397 encoding = m.group(1)
399 m = re.search(br'<meta[^>]+charset=[\'"]?([^\'")]+)[ /\'">]',
400 webpage_bytes[:1024])
402 encoding = m.group(1).decode('ascii')
403 elif webpage_bytes.startswith(b'\xff\xfe'):
410 def _webpage_read_content(self, urlh, url_or_request, video_id, note=None, errnote=None, fatal=True, prefix=None, encoding=None):
411 content_type = urlh.headers.get('Content-Type', '')
412 webpage_bytes = urlh.read()
413 if prefix is not None:
414 webpage_bytes = prefix + webpage_bytes
416 encoding = self._guess_encoding_from_content(content_type, webpage_bytes)
417 if self._downloader.params.get('dump_intermediate_pages', False):
419 url = url_or_request.get_full_url()
420 except AttributeError:
422 self.to_screen('Dumping request to ' + url)
423 dump = base64.b64encode(webpage_bytes).decode('ascii')
424 self._downloader.to_screen(dump)
425 if self._downloader.params.get('write_pages', False):
427 url = url_or_request.get_full_url()
428 except AttributeError:
430 basen = '%s_%s' % (video_id, url)
432 h = '___' + hashlib.md5(basen.encode('utf-8')).hexdigest()
433 basen = basen[:240 - len(h)] + h
434 raw_filename = basen + '.dump'
435 filename = sanitize_filename(raw_filename, restricted=True)
436 self.to_screen('Saving request to ' + filename)
437 # Working around MAX_PATH limitation on Windows (see
438 # http://msdn.microsoft.com/en-us/library/windows/desktop/aa365247(v=vs.85).aspx)
439 if compat_os_name == 'nt':
440 absfilepath = os.path.abspath(filename)
441 if len(absfilepath) > 259:
442 filename = '\\\\?\\' + absfilepath
443 with open(filename, 'wb') as outf:
444 outf.write(webpage_bytes)
447 content = webpage_bytes.decode(encoding, 'replace')
449 content = webpage_bytes.decode('utf-8', 'replace')
451 if ('<title>Access to this site is blocked</title>' in content and
452 'Websense' in content[:512]):
453 msg = 'Access to this webpage has been blocked by Websense filtering software in your network.'
454 blocked_iframe = self._html_search_regex(
455 r'<iframe src="([^"]+)"', content,
456 'Websense information URL', default=None)
458 msg += ' Visit %s for more details' % blocked_iframe
459 raise ExtractorError(msg, expected=True)
460 if '<title>The URL you requested has been blocked</title>' in content[:512]:
462 'Access to this webpage has been blocked by Indian censorship. '
463 'Use a VPN or proxy server (with --proxy) to route around it.')
464 block_msg = self._html_search_regex(
465 r'</h1><p>(.*?)</p>',
466 content, 'block message', default=None)
468 msg += ' (Message: "%s")' % block_msg.replace('\n', ' ')
469 raise ExtractorError(msg, expected=True)
473 def _download_webpage(self, url_or_request, video_id, note=None, errnote=None, fatal=True, tries=1, timeout=5, encoding=None, data=None, headers=None, query=None):
474 """ Returns the data of the page as a string """
477 while success is False:
479 res = self._download_webpage_handle(url_or_request, video_id, note, errnote, fatal, encoding=encoding, data=data, headers=headers, query=query)
481 except compat_http_client.IncompleteRead as e:
483 if try_count >= tries:
485 self._sleep(timeout, video_id)
492 def _download_xml(self, url_or_request, video_id,
493 note='Downloading XML', errnote='Unable to download XML',
494 transform_source=None, fatal=True, encoding=None, data=None, headers=None, query=None):
495 """Return the xml as an xml.etree.ElementTree.Element"""
496 xml_string = self._download_webpage(
497 url_or_request, video_id, note, errnote, fatal=fatal, encoding=encoding, data=data, headers=headers, query=query)
498 if xml_string is False:
501 xml_string = transform_source(xml_string)
502 return compat_etree_fromstring(xml_string.encode('utf-8'))
504 def _download_json(self, url_or_request, video_id,
505 note='Downloading JSON metadata',
506 errnote='Unable to download JSON metadata',
507 transform_source=None,
508 fatal=True, encoding=None, data=None, headers=None, query=None):
509 json_string = self._download_webpage(
510 url_or_request, video_id, note, errnote, fatal=fatal,
511 encoding=encoding, data=data, headers=headers, query=query)
512 if (not fatal) and json_string is False:
514 return self._parse_json(
515 json_string, video_id, transform_source=transform_source, fatal=fatal)
517 def _parse_json(self, json_string, video_id, transform_source=None, fatal=True):
519 json_string = transform_source(json_string)
521 return json.loads(json_string)
522 except ValueError as ve:
523 errmsg = '%s: Failed to parse JSON ' % video_id
525 raise ExtractorError(errmsg, cause=ve)
527 self.report_warning(errmsg + str(ve))
529 def report_warning(self, msg, video_id=None):
530 idstr = '' if video_id is None else '%s: ' % video_id
531 self._downloader.report_warning(
532 '[%s] %s%s' % (self.IE_NAME, idstr, msg))
534 def to_screen(self, msg):
535 """Print msg to screen, prefixing it with '[ie_name]'"""
536 self._downloader.to_screen('[%s] %s' % (self.IE_NAME, msg))
538 def report_extraction(self, id_or_name):
539 """Report information extraction."""
540 self.to_screen('%s: Extracting information' % id_or_name)
542 def report_download_webpage(self, video_id):
543 """Report webpage download."""
544 self.to_screen('%s: Downloading webpage' % video_id)
546 def report_age_confirmation(self):
547 """Report attempt to confirm age."""
548 self.to_screen('Confirming age')
550 def report_login(self):
551 """Report attempt to log in."""
552 self.to_screen('Logging in')
555 def raise_login_required(msg='This video is only available for registered users'):
556 raise ExtractorError(
557 '%s. Use --username and --password or --netrc to provide account credentials.' % msg,
561 def raise_geo_restricted(msg='This video is not available from your location due to geo restriction'):
562 raise ExtractorError(
563 '%s. You might want to use --proxy to workaround.' % msg,
566 # Methods for following #608
568 def url_result(url, ie=None, video_id=None, video_title=None):
569 """Returns a URL that points to a page that should be processed"""
570 # TODO: ie should be the class used for getting the info
571 video_info = {'_type': 'url',
574 if video_id is not None:
575 video_info['id'] = video_id
576 if video_title is not None:
577 video_info['title'] = video_title
581 def playlist_result(entries, playlist_id=None, playlist_title=None, playlist_description=None):
582 """Returns a playlist"""
583 video_info = {'_type': 'playlist',
586 video_info['id'] = playlist_id
588 video_info['title'] = playlist_title
589 if playlist_description:
590 video_info['description'] = playlist_description
593 def _search_regex(self, pattern, string, name, default=NO_DEFAULT, fatal=True, flags=0, group=None):
595 Perform a regex search on the given string, using a single or a list of
596 patterns returning the first matching group.
597 In case of failure return a default value or raise a WARNING or a
598 RegexNotFoundError, depending on fatal, specifying the field name.
600 if isinstance(pattern, (str, compat_str, compiled_regex_type)):
601 mobj = re.search(pattern, string, flags)
604 mobj = re.search(p, string, flags)
608 if not self._downloader.params.get('no_color') and compat_os_name != 'nt' and sys.stderr.isatty():
609 _name = '\033[0;34m%s\033[0m' % name
615 # return the first matching group
616 return next(g for g in mobj.groups() if g is not None)
618 return mobj.group(group)
619 elif default is not NO_DEFAULT:
622 raise RegexNotFoundError('Unable to extract %s' % _name)
624 self._downloader.report_warning('unable to extract %s' % _name + bug_reports_message())
627 def _html_search_regex(self, pattern, string, name, default=NO_DEFAULT, fatal=True, flags=0, group=None):
629 Like _search_regex, but strips HTML tags and unescapes entities.
631 res = self._search_regex(pattern, string, name, default, fatal, flags, group)
633 return clean_html(res).strip()
637 def _get_login_info(self):
639 Get the login info as (username, password)
640 It will look in the netrc file using the _NETRC_MACHINE value
641 If there's no info available, return (None, None)
643 if self._downloader is None:
648 downloader_params = self._downloader.params
650 # Attempt to use provided username and password or .netrc data
651 if downloader_params.get('username') is not None:
652 username = downloader_params['username']
653 password = downloader_params['password']
654 elif downloader_params.get('usenetrc', False):
656 info = netrc.netrc().authenticators(self._NETRC_MACHINE)
661 raise netrc.NetrcParseError('No authenticators for %s' % self._NETRC_MACHINE)
662 except (IOError, netrc.NetrcParseError) as err:
663 self._downloader.report_warning('parsing .netrc: %s' % error_to_compat_str(err))
665 return (username, password)
667 def _get_tfa_info(self, note='two-factor verification code'):
669 Get the two-factor authentication info
670 TODO - asking the user will be required for sms/phone verify
671 currently just uses the command line option
672 If there's no info available, return None
674 if self._downloader is None:
676 downloader_params = self._downloader.params
678 if downloader_params.get('twofactor') is not None:
679 return downloader_params['twofactor']
681 return compat_getpass('Type %s and press [Return]: ' % note)
683 # Helper functions for extracting OpenGraph info
685 def _og_regexes(prop):
686 content_re = r'content=(?:"([^"]+?)"|\'([^\']+?)\'|\s*([^\s"\'=<>`]+?))'
687 property_re = (r'(?:name|property)=(?:\'og:%(prop)s\'|"og:%(prop)s"|\s*og:%(prop)s\b)'
688 % {'prop': re.escape(prop)})
689 template = r'<meta[^>]+?%s[^>]+?%s'
691 template % (property_re, content_re),
692 template % (content_re, property_re),
696 def _meta_regex(prop):
697 return r'''(?isx)<meta
698 (?=[^>]+(?:itemprop|name|property|id|http-equiv)=(["\']?)%s\1)
699 [^>]+?content=(["\'])(?P<content>.*?)\2''' % re.escape(prop)
701 def _og_search_property(self, prop, html, name=None, **kargs):
703 name = 'OpenGraph %s' % prop
704 escaped = self._search_regex(self._og_regexes(prop), html, name, flags=re.DOTALL, **kargs)
707 return unescapeHTML(escaped)
709 def _og_search_thumbnail(self, html, **kargs):
710 return self._og_search_property('image', html, 'thumbnail URL', fatal=False, **kargs)
712 def _og_search_description(self, html, **kargs):
713 return self._og_search_property('description', html, fatal=False, **kargs)
715 def _og_search_title(self, html, **kargs):
716 return self._og_search_property('title', html, **kargs)
718 def _og_search_video_url(self, html, name='video url', secure=True, **kargs):
719 regexes = self._og_regexes('video') + self._og_regexes('video:url')
721 regexes = self._og_regexes('video:secure_url') + regexes
722 return self._html_search_regex(regexes, html, name, **kargs)
724 def _og_search_url(self, html, **kargs):
725 return self._og_search_property('url', html, **kargs)
727 def _html_search_meta(self, name, html, display_name=None, fatal=False, **kwargs):
728 if display_name is None:
730 return self._html_search_regex(
731 self._meta_regex(name),
732 html, display_name, fatal=fatal, group='content', **kwargs)
734 def _dc_search_uploader(self, html):
735 return self._html_search_meta('dc.creator', html, 'uploader')
737 def _rta_search(self, html):
738 # See http://www.rtalabel.org/index.php?content=howtofaq#single
739 if re.search(r'(?ix)<meta\s+name="rating"\s+'
740 r' content="RTA-5042-1996-1400-1577-RTA"',
745 def _media_rating_search(self, html):
746 # See http://www.tjg-designs.com/WP/metadata-code-examples-adding-metadata-to-your-web-pages/
747 rating = self._html_search_meta('rating', html)
759 return RATING_TABLE.get(rating.lower())
761 def _family_friendly_search(self, html):
762 # See http://schema.org/VideoObject
763 family_friendly = self._html_search_meta('isFamilyFriendly', html)
765 if not family_friendly:
774 return RATING_TABLE.get(family_friendly.lower())
776 def _twitter_search_player(self, html):
777 return self._html_search_meta('twitter:player', html,
778 'twitter card player')
780 def _search_json_ld(self, html, video_id, **kwargs):
781 json_ld = self._search_regex(
782 r'(?s)<script[^>]+type=(["\'])application/ld\+json\1[^>]*>(?P<json_ld>.+?)</script>',
783 html, 'JSON-LD', group='json_ld', **kwargs)
786 return self._json_ld(json_ld, video_id, fatal=kwargs.get('fatal', True))
788 def _json_ld(self, json_ld, video_id, fatal=True):
789 if isinstance(json_ld, compat_str):
790 json_ld = self._parse_json(json_ld, video_id, fatal=fatal)
794 if json_ld.get('@context') == 'http://schema.org':
795 item_type = json_ld.get('@type')
796 if item_type == 'TVEpisode':
798 'episode': unescapeHTML(json_ld.get('name')),
799 'episode_number': int_or_none(json_ld.get('episodeNumber')),
800 'description': unescapeHTML(json_ld.get('description')),
802 part_of_season = json_ld.get('partOfSeason')
803 if isinstance(part_of_season, dict) and part_of_season.get('@type') == 'TVSeason':
804 info['season_number'] = int_or_none(part_of_season.get('seasonNumber'))
805 part_of_series = json_ld.get('partOfSeries')
806 if isinstance(part_of_series, dict) and part_of_series.get('@type') == 'TVSeries':
807 info['series'] = unescapeHTML(part_of_series.get('name'))
808 elif item_type == 'Article':
810 'timestamp': parse_iso8601(json_ld.get('datePublished')),
811 'title': unescapeHTML(json_ld.get('headline')),
812 'description': unescapeHTML(json_ld.get('articleBody')),
814 return dict((k, v) for k, v in info.items() if v is not None)
817 def _hidden_inputs(html):
818 html = re.sub(r'<!--(?:(?!<!--).)*-->', '', html)
820 for input in re.findall(r'(?i)<input([^>]+)>', html):
821 if not re.search(r'type=(["\'])(?:hidden|submit)\1', input):
823 name = re.search(r'name=(["\'])(?P<value>.+?)\1', input)
826 value = re.search(r'value=(["\'])(?P<value>.*?)\1', input)
829 hidden_inputs[name.group('value')] = value.group('value')
832 def _form_hidden_inputs(self, form_id, html):
833 form = self._search_regex(
834 r'(?is)<form[^>]+?id=(["\'])%s\1[^>]*>(?P<form>.+?)</form>' % form_id,
835 html, '%s form' % form_id, group='form')
836 return self._hidden_inputs(form)
838 def _sort_formats(self, formats, field_preference=None):
840 raise ExtractorError('No video formats found')
843 # Automatically determine tbr when missing based on abr and vbr (improves
844 # formats sorting in some cases)
845 if 'tbr' not in f and f.get('abr') is not None and f.get('vbr') is not None:
846 f['tbr'] = f['abr'] + f['vbr']
849 # TODO remove the following workaround
850 from ..utils import determine_ext
851 if not f.get('ext') and 'url' in f:
852 f['ext'] = determine_ext(f['url'])
854 if isinstance(field_preference, (list, tuple)):
855 return tuple(f.get(field) if f.get(field) is not None else -1 for field in field_preference)
857 preference = f.get('preference')
858 if preference is None:
860 if f.get('ext') in ['f4f', 'f4m']: # Not yet supported
863 proto_preference = 0 if determine_protocol(f) in ['http', 'https'] else -0.1
865 if f.get('vcodec') == 'none': # audio only
867 if self._downloader.params.get('prefer_free_formats'):
868 ORDER = ['aac', 'mp3', 'm4a', 'webm', 'ogg', 'opus']
870 ORDER = ['webm', 'opus', 'ogg', 'mp3', 'aac', 'm4a']
873 audio_ext_preference = ORDER.index(f['ext'])
875 audio_ext_preference = -1
877 if f.get('acodec') == 'none': # video only
879 if self._downloader.params.get('prefer_free_formats'):
880 ORDER = ['flv', 'mp4', 'webm']
882 ORDER = ['webm', 'flv', 'mp4']
884 ext_preference = ORDER.index(f['ext'])
887 audio_ext_preference = 0
891 f.get('language_preference') if f.get('language_preference') is not None else -1,
892 f.get('quality') if f.get('quality') is not None else -1,
893 f.get('tbr') if f.get('tbr') is not None else -1,
894 f.get('filesize') if f.get('filesize') is not None else -1,
895 f.get('vbr') if f.get('vbr') is not None else -1,
896 f.get('height') if f.get('height') is not None else -1,
897 f.get('width') if f.get('width') is not None else -1,
900 f.get('abr') if f.get('abr') is not None else -1,
901 audio_ext_preference,
902 f.get('fps') if f.get('fps') is not None else -1,
903 f.get('filesize_approx') if f.get('filesize_approx') is not None else -1,
904 f.get('source_preference') if f.get('source_preference') is not None else -1,
905 f.get('format_id') if f.get('format_id') is not None else '',
907 formats.sort(key=_formats_key)
909 def _check_formats(self, formats, video_id):
912 lambda f: self._is_valid_url(
914 item='%s video format' % f.get('format_id') if f.get('format_id') else 'video'),
918 def _remove_duplicate_formats(formats):
922 if f['url'] not in format_urls:
923 format_urls.add(f['url'])
924 unique_formats.append(f)
925 formats[:] = unique_formats
927 def _is_valid_url(self, url, video_id, item='video'):
928 url = self._proto_relative_url(url, scheme='http:')
929 # For now assume non HTTP(S) URLs always valid
930 if not (url.startswith('http://') or url.startswith('https://')):
933 self._request_webpage(url, video_id, 'Checking %s URL' % item)
935 except ExtractorError as e:
936 if isinstance(e.cause, compat_urllib_error.URLError):
938 '%s: %s URL is invalid, skipping' % (video_id, item))
942 def http_scheme(self):
943 """ Either "http:" or "https:", depending on the user's preferences """
946 if self._downloader.params.get('prefer_insecure', False)
949 def _proto_relative_url(self, url, scheme=None):
952 if url.startswith('//'):
954 scheme = self.http_scheme()
959 def _sleep(self, timeout, video_id, msg_template=None):
960 if msg_template is None:
961 msg_template = '%(video_id)s: Waiting for %(timeout)s seconds'
962 msg = msg_template % {'video_id': video_id, 'timeout': timeout}
966 def _extract_f4m_formats(self, manifest_url, video_id, preference=None, f4m_id=None,
967 transform_source=lambda s: fix_xml_ampersands(s).strip(),
969 manifest = self._download_xml(
970 manifest_url, video_id, 'Downloading f4m manifest',
971 'Unable to download f4m manifest',
972 # Some manifests may be malformed, e.g. prosiebensat1 generated manifests
973 # (see https://github.com/rg3/youtube-dl/issues/6215#issuecomment-121704244)
974 transform_source=transform_source,
977 if manifest is False:
980 return self._parse_f4m_formats(
981 manifest, manifest_url, video_id, preference=preference, f4m_id=f4m_id,
982 transform_source=transform_source, fatal=fatal)
984 def _parse_f4m_formats(self, manifest, manifest_url, video_id, preference=None, f4m_id=None,
985 transform_source=lambda s: fix_xml_ampersands(s).strip(),
988 manifest_version = '1.0'
989 media_nodes = manifest.findall('{http://ns.adobe.com/f4m/1.0}media')
991 manifest_version = '2.0'
992 media_nodes = manifest.findall('{http://ns.adobe.com/f4m/2.0}media')
993 # Remove unsupported DRM protected media from final formats
994 # rendition (see https://github.com/rg3/youtube-dl/issues/8573).
995 media_nodes = remove_encrypted_media(media_nodes)
998 base_url = xpath_text(
999 manifest, ['{http://ns.adobe.com/f4m/1.0}baseURL', '{http://ns.adobe.com/f4m/2.0}baseURL'],
1000 'base URL', default=None)
1002 base_url = base_url.strip()
1003 for i, media_el in enumerate(media_nodes):
1004 if manifest_version == '2.0':
1005 media_url = media_el.attrib.get('href') or media_el.attrib.get('url')
1009 media_url if media_url.startswith('http://') or media_url.startswith('https://')
1010 else ((base_url or '/'.join(manifest_url.split('/')[:-1])) + '/' + media_url))
1011 # If media_url is itself a f4m manifest do the recursive extraction
1012 # since bitrates in parent manifest (this one) and media_url manifest
1013 # may differ leading to inability to resolve the format by requested
1014 # bitrate in f4m downloader
1015 if determine_ext(manifest_url) == 'f4m':
1016 formats.extend(self._extract_f4m_formats(
1017 manifest_url, video_id, preference=preference, f4m_id=f4m_id,
1018 transform_source=transform_source, fatal=fatal))
1020 tbr = int_or_none(media_el.attrib.get('bitrate'))
1022 'format_id': '-'.join(filter(None, [f4m_id, compat_str(i if tbr is None else tbr)])),
1023 'url': manifest_url,
1026 'width': int_or_none(media_el.attrib.get('width')),
1027 'height': int_or_none(media_el.attrib.get('height')),
1028 'preference': preference,
1032 def _extract_m3u8_formats(self, m3u8_url, video_id, ext=None,
1033 entry_protocol='m3u8', preference=None,
1034 m3u8_id=None, note=None, errnote=None,
1038 'format_id': '-'.join(filter(None, [m3u8_id, 'meta'])),
1042 'preference': preference - 1 if preference else -1,
1043 'resolution': 'multiple',
1044 'format_note': 'Quality selection URL',
1047 format_url = lambda u: (
1049 if re.match(r'^https?://', u)
1050 else compat_urlparse.urljoin(m3u8_url, u))
1052 res = self._download_webpage_handle(
1054 note=note or 'Downloading m3u8 information',
1055 errnote=errnote or 'Failed to download m3u8 information',
1059 m3u8_doc, urlh = res
1060 m3u8_url = urlh.geturl()
1062 # We should try extracting formats only from master playlists [1], i.e.
1063 # playlists that describe available qualities. On the other hand media
1064 # playlists [2] should be returned as is since they contain just the media
1065 # without qualities renditions.
1066 # Fortunately, master playlist can be easily distinguished from media
1067 # playlist based on particular tags availability. As of [1, 2] master
1068 # playlist tags MUST NOT appear in a media playist and vice versa.
1069 # As of [3] #EXT-X-TARGETDURATION tag is REQUIRED for every media playlist
1070 # and MUST NOT appear in master playlist thus we can clearly detect media
1071 # playlist with this criterion.
1072 # 1. https://tools.ietf.org/html/draft-pantos-http-live-streaming-17#section-4.3.4
1073 # 2. https://tools.ietf.org/html/draft-pantos-http-live-streaming-17#section-4.3.3
1074 # 3. https://tools.ietf.org/html/draft-pantos-http-live-streaming-17#section-4.3.3.1
1075 if '#EXT-X-TARGETDURATION' in m3u8_doc: # media playlist, return as is
1078 'format_id': m3u8_id,
1080 'protocol': entry_protocol,
1081 'preference': preference,
1085 kv_rex = re.compile(
1086 r'(?P<key>[a-zA-Z_-]+)=(?P<val>"[^"]+"|[^",]+)(?:,|$)')
1087 for line in m3u8_doc.splitlines():
1088 if line.startswith('#EXT-X-STREAM-INF:'):
1090 for m in kv_rex.finditer(line):
1092 if v.startswith('"'):
1094 last_info[m.group('key')] = v
1095 elif line.startswith('#EXT-X-MEDIA:'):
1097 for m in kv_rex.finditer(line):
1099 if v.startswith('"'):
1101 last_media[m.group('key')] = v
1102 elif line.startswith('#') or not line.strip():
1105 if last_info is None:
1106 formats.append({'url': format_url(line)})
1108 tbr = int_or_none(last_info.get('BANDWIDTH'), scale=1000)
1111 format_id.append(m3u8_id)
1112 last_media_name = last_media.get('NAME') if last_media and last_media.get('TYPE') != 'SUBTITLES' else None
1113 format_id.append(last_media_name if last_media_name else '%d' % (tbr if tbr else len(formats)))
1115 'format_id': '-'.join(format_id),
1116 'url': format_url(line.strip()),
1119 'protocol': entry_protocol,
1120 'preference': preference,
1122 resolution = last_info.get('RESOLUTION')
1124 width_str, height_str = resolution.split('x')
1125 f['width'] = int(width_str)
1126 f['height'] = int(height_str)
1127 codecs = last_info.get('CODECS')
1129 vcodec, acodec = [None] * 2
1130 va_codecs = codecs.split(',')
1131 if len(va_codecs) == 1:
1132 # Audio only entries usually come with single codec and
1133 # no resolution. For more robustness we also check it to
1135 if not resolution and va_codecs[0].startswith('mp4a'):
1136 vcodec, acodec = 'none', va_codecs[0]
1138 vcodec = va_codecs[0]
1140 vcodec, acodec = va_codecs[:2]
1145 if last_media is not None:
1146 f['m3u8_media'] = last_media
1153 def _xpath_ns(path, namespace=None):
1157 for c in path.split('/'):
1158 if not c or c == '.':
1161 out.append('{%s}%s' % (namespace, c))
1162 return '/'.join(out)
1164 def _extract_smil_formats(self, smil_url, video_id, fatal=True, f4m_params=None, transform_source=None):
1165 smil = self._download_smil(smil_url, video_id, fatal=fatal, transform_source=transform_source)
1171 namespace = self._parse_smil_namespace(smil)
1173 return self._parse_smil_formats(
1174 smil, smil_url, video_id, namespace=namespace, f4m_params=f4m_params)
1176 def _extract_smil_info(self, smil_url, video_id, fatal=True, f4m_params=None):
1177 smil = self._download_smil(smil_url, video_id, fatal=fatal)
1180 return self._parse_smil(smil, smil_url, video_id, f4m_params=f4m_params)
1182 def _download_smil(self, smil_url, video_id, fatal=True, transform_source=None):
1183 return self._download_xml(
1184 smil_url, video_id, 'Downloading SMIL file',
1185 'Unable to download SMIL file', fatal=fatal, transform_source=transform_source)
1187 def _parse_smil(self, smil, smil_url, video_id, f4m_params=None):
1188 namespace = self._parse_smil_namespace(smil)
1190 formats = self._parse_smil_formats(
1191 smil, smil_url, video_id, namespace=namespace, f4m_params=f4m_params)
1192 subtitles = self._parse_smil_subtitles(smil, namespace=namespace)
1194 video_id = os.path.splitext(url_basename(smil_url))[0]
1198 for meta in smil.findall(self._xpath_ns('./head/meta', namespace)):
1199 name = meta.attrib.get('name')
1200 content = meta.attrib.get('content')
1201 if not name or not content:
1203 if not title and name == 'title':
1205 elif not description and name in ('description', 'abstract'):
1206 description = content
1207 elif not upload_date and name == 'date':
1208 upload_date = unified_strdate(content)
1211 'id': image.get('type'),
1212 'url': image.get('src'),
1213 'width': int_or_none(image.get('width')),
1214 'height': int_or_none(image.get('height')),
1215 } for image in smil.findall(self._xpath_ns('.//image', namespace)) if image.get('src')]
1219 'title': title or video_id,
1220 'description': description,
1221 'upload_date': upload_date,
1222 'thumbnails': thumbnails,
1224 'subtitles': subtitles,
1227 def _parse_smil_namespace(self, smil):
1228 return self._search_regex(
1229 r'(?i)^{([^}]+)?}smil$', smil.tag, 'namespace', default=None)
1231 def _parse_smil_formats(self, smil, smil_url, video_id, namespace=None, f4m_params=None, transform_rtmp_url=None):
1233 for meta in smil.findall(self._xpath_ns('./head/meta', namespace)):
1234 b = meta.get('base') or meta.get('httpBase')
1245 videos = smil.findall(self._xpath_ns('.//video', namespace))
1246 for video in videos:
1247 src = video.get('src')
1248 if not src or src in srcs:
1252 bitrate = float_or_none(video.get('system-bitrate') or video.get('systemBitrate'), 1000)
1253 filesize = int_or_none(video.get('size') or video.get('fileSize'))
1254 width = int_or_none(video.get('width'))
1255 height = int_or_none(video.get('height'))
1256 proto = video.get('proto')
1257 ext = video.get('ext')
1258 src_ext = determine_ext(src)
1259 streamer = video.get('streamer') or base
1261 if proto == 'rtmp' or streamer.startswith('rtmp'):
1267 'format_id': 'rtmp-%d' % (rtmp_count if bitrate is None else bitrate),
1269 'filesize': filesize,
1273 if transform_rtmp_url:
1274 streamer, src = transform_rtmp_url(streamer, src)
1275 formats[-1].update({
1281 src_url = src if src.startswith('http') else compat_urlparse.urljoin(base, src)
1282 src_url = src_url.strip()
1284 if proto == 'm3u8' or src_ext == 'm3u8':
1285 m3u8_formats = self._extract_m3u8_formats(
1286 src_url, video_id, ext or 'mp4', m3u8_id='hls', fatal=False)
1287 if len(m3u8_formats) == 1:
1289 m3u8_formats[0].update({
1290 'format_id': 'hls-%d' % (m3u8_count if bitrate is None else bitrate),
1295 formats.extend(m3u8_formats)
1298 if src_ext == 'f4m':
1303 'plugin': 'flowplayer-3.2.0.1',
1305 f4m_url += '&' if '?' in f4m_url else '?'
1306 f4m_url += compat_urllib_parse_urlencode(f4m_params)
1307 formats.extend(self._extract_f4m_formats(f4m_url, video_id, f4m_id='hds', fatal=False))
1310 if src_url.startswith('http') and self._is_valid_url(src, video_id):
1314 'ext': ext or src_ext or 'flv',
1315 'format_id': 'http-%d' % (bitrate or http_count),
1317 'filesize': filesize,
1325 def _parse_smil_subtitles(self, smil, namespace=None, subtitles_lang='en'):
1328 for num, textstream in enumerate(smil.findall(self._xpath_ns('.//textstream', namespace))):
1329 src = textstream.get('src')
1330 if not src or src in urls:
1333 ext = textstream.get('ext') or determine_ext(src) or mimetype2ext(textstream.get('type'))
1334 lang = textstream.get('systemLanguage') or textstream.get('systemLanguageName') or textstream.get('lang') or subtitles_lang
1335 subtitles.setdefault(lang, []).append({
1341 def _extract_xspf_playlist(self, playlist_url, playlist_id, fatal=True):
1342 xspf = self._download_xml(
1343 playlist_url, playlist_id, 'Downloading xpsf playlist',
1344 'Unable to download xspf manifest', fatal=fatal)
1347 return self._parse_xspf(xspf, playlist_id)
1349 def _parse_xspf(self, playlist, playlist_id):
1351 'xspf': 'http://xspf.org/ns/0/',
1352 's1': 'http://static.streamone.nl/player/ns/0',
1356 for track in playlist.findall(xpath_with_ns('./xspf:trackList/xspf:track', NS_MAP)):
1358 track, xpath_with_ns('./xspf:title', NS_MAP), 'title', default=playlist_id)
1359 description = xpath_text(
1360 track, xpath_with_ns('./xspf:annotation', NS_MAP), 'description')
1361 thumbnail = xpath_text(
1362 track, xpath_with_ns('./xspf:image', NS_MAP), 'thumbnail')
1363 duration = float_or_none(
1364 xpath_text(track, xpath_with_ns('./xspf:duration', NS_MAP), 'duration'), 1000)
1367 'url': location.text,
1368 'format_id': location.get(xpath_with_ns('s1:label', NS_MAP)),
1369 'width': int_or_none(location.get(xpath_with_ns('s1:width', NS_MAP))),
1370 'height': int_or_none(location.get(xpath_with_ns('s1:height', NS_MAP))),
1371 } for location in track.findall(xpath_with_ns('./xspf:location', NS_MAP))]
1372 self._sort_formats(formats)
1377 'description': description,
1378 'thumbnail': thumbnail,
1379 'duration': duration,
1384 def _extract_mpd_formats(self, mpd_url, video_id, mpd_id=None, note=None, errnote=None, fatal=True, formats_dict={}):
1385 res = self._download_webpage_handle(
1387 note=note or 'Downloading MPD manifest',
1388 errnote=errnote or 'Failed to download MPD manifest',
1393 mpd_base_url = re.match(r'https?://.+/', urlh.geturl()).group()
1395 return self._parse_mpd_formats(
1396 compat_etree_fromstring(mpd.encode('utf-8')), mpd_id, mpd_base_url, formats_dict=formats_dict)
1398 def _parse_mpd_formats(self, mpd_doc, mpd_id=None, mpd_base_url='', formats_dict={}):
1399 if mpd_doc.get('type') == 'dynamic':
1402 namespace = self._search_regex(r'(?i)^{([^}]+)?}MPD$', mpd_doc.tag, 'namespace', default=None)
1405 return self._xpath_ns(path, namespace)
1407 def is_drm_protected(element):
1408 return element.find(_add_ns('ContentProtection')) is not None
1410 def extract_multisegment_info(element, ms_parent_info):
1411 ms_info = ms_parent_info.copy()
1412 segment_list = element.find(_add_ns('SegmentList'))
1413 if segment_list is not None:
1414 segment_urls_e = segment_list.findall(_add_ns('SegmentURL'))
1416 ms_info['segment_urls'] = [segment.attrib['media'] for segment in segment_urls_e]
1417 initialization = segment_list.find(_add_ns('Initialization'))
1418 if initialization is not None:
1419 ms_info['initialization_url'] = initialization.attrib['sourceURL']
1421 segment_template = element.find(_add_ns('SegmentTemplate'))
1422 if segment_template is not None:
1423 start_number = segment_template.get('startNumber')
1425 ms_info['start_number'] = int(start_number)
1426 segment_timeline = segment_template.find(_add_ns('SegmentTimeline'))
1427 if segment_timeline is not None:
1428 s_e = segment_timeline.findall(_add_ns('S'))
1430 ms_info['total_number'] = 0
1432 ms_info['total_number'] += 1 + int(s.get('r', '0'))
1434 timescale = segment_template.get('timescale')
1436 ms_info['timescale'] = int(timescale)
1437 segment_duration = segment_template.get('duration')
1438 if segment_duration:
1439 ms_info['segment_duration'] = int(segment_duration)
1440 media_template = segment_template.get('media')
1442 ms_info['media_template'] = media_template
1443 initialization = segment_template.get('initialization')
1445 ms_info['initialization_url'] = initialization
1447 initialization = segment_template.find(_add_ns('Initialization'))
1448 if initialization is not None:
1449 ms_info['initialization_url'] = initialization.attrib['sourceURL']
1452 mpd_duration = parse_duration(mpd_doc.get('mediaPresentationDuration'))
1454 for period in mpd_doc.findall(_add_ns('Period')):
1455 period_duration = parse_duration(period.get('duration')) or mpd_duration
1456 period_ms_info = extract_multisegment_info(period, {
1460 for adaptation_set in period.findall(_add_ns('AdaptationSet')):
1461 if is_drm_protected(adaptation_set):
1463 adaption_set_ms_info = extract_multisegment_info(adaptation_set, period_ms_info)
1464 for representation in adaptation_set.findall(_add_ns('Representation')):
1465 if is_drm_protected(representation):
1467 representation_attrib = adaptation_set.attrib.copy()
1468 representation_attrib.update(representation.attrib)
1469 # According to page 41 of ISO/IEC 29001-1:2014, @mimeType is mandatory
1470 mime_type = representation_attrib['mimeType']
1471 content_type = mime_type.split('/')[0]
1472 if content_type == 'text':
1473 # TODO implement WebVTT downloading
1475 elif content_type == 'video' or content_type == 'audio':
1477 for element in (representation, adaptation_set, period, mpd_doc):
1478 base_url_e = element.find(_add_ns('BaseURL'))
1479 if base_url_e is not None:
1480 base_url = base_url_e.text + base_url
1481 if re.match(r'^https?://', base_url):
1483 if mpd_base_url and not re.match(r'^https?://', base_url):
1484 if not mpd_base_url.endswith('/') and not base_url.startswith('/'):
1486 base_url = mpd_base_url + base_url
1487 representation_id = representation_attrib.get('id')
1488 lang = representation_attrib.get('lang')
1489 url_el = representation.find(_add_ns('BaseURL'))
1490 filesize = int_or_none(url_el.attrib.get('{http://youtube.com/yt/2012/10/10}contentLength') if url_el is not None else None)
1492 'format_id': '%s-%s' % (mpd_id, representation_id) if mpd_id else representation_id,
1494 'ext': mimetype2ext(mime_type),
1495 'width': int_or_none(representation_attrib.get('width')),
1496 'height': int_or_none(representation_attrib.get('height')),
1497 'tbr': int_or_none(representation_attrib.get('bandwidth'), 1000),
1498 'asr': int_or_none(representation_attrib.get('audioSamplingRate')),
1499 'fps': int_or_none(representation_attrib.get('frameRate')),
1500 'vcodec': 'none' if content_type == 'audio' else representation_attrib.get('codecs'),
1501 'acodec': 'none' if content_type == 'video' else representation_attrib.get('codecs'),
1502 'language': lang if lang not in ('mul', 'und', 'zxx', 'mis') else None,
1503 'format_note': 'DASH %s' % content_type,
1504 'filesize': filesize,
1506 representation_ms_info = extract_multisegment_info(representation, adaption_set_ms_info)
1507 if 'segment_urls' not in representation_ms_info and 'media_template' in representation_ms_info:
1508 if 'total_number' not in representation_ms_info and 'segment_duration':
1509 segment_duration = float(representation_ms_info['segment_duration']) / float(representation_ms_info['timescale'])
1510 representation_ms_info['total_number'] = int(math.ceil(float(period_duration) / segment_duration))
1511 media_template = representation_ms_info['media_template']
1512 media_template = media_template.replace('$RepresentationID$', representation_id)
1513 media_template = re.sub(r'\$(Number|Bandwidth)(?:%(0\d+)d)?\$', r'%(\1)\2d', media_template)
1514 media_template.replace('$$', '$')
1515 representation_ms_info['segment_urls'] = [media_template % {'Number': segment_number, 'Bandwidth': representation_attrib.get('bandwidth')} for segment_number in range(representation_ms_info['start_number'], representation_ms_info['total_number'] + representation_ms_info['start_number'])]
1516 if 'segment_urls' in representation_ms_info:
1518 'segment_urls': representation_ms_info['segment_urls'],
1519 'protocol': 'http_dash_segments',
1521 if 'initialization_url' in representation_ms_info:
1522 initialization_url = representation_ms_info['initialization_url'].replace('$RepresentationID$', representation_id)
1524 'initialization_url': initialization_url,
1526 if not f.get('url'):
1527 f['url'] = initialization_url
1529 existing_format = next(
1530 fo for fo in formats
1531 if fo['format_id'] == representation_id)
1532 except StopIteration:
1533 full_info = formats_dict.get(representation_id, {}).copy()
1535 formats.append(full_info)
1537 existing_format.update(f)
1539 self.report_warning('Unknown MIME type %s in DASH manifest' % mime_type)
1542 def _live_title(self, name):
1543 """ Generate the title for a live video """
1544 now = datetime.datetime.now()
1545 now_str = now.strftime('%Y-%m-%d %H:%M')
1546 return name + ' ' + now_str
1548 def _int(self, v, name, fatal=False, **kwargs):
1549 res = int_or_none(v, **kwargs)
1550 if 'get_attr' in kwargs:
1551 print(getattr(v, kwargs['get_attr']))
1553 msg = 'Failed to extract %s: Could not parse value %r' % (name, v)
1555 raise ExtractorError(msg)
1557 self._downloader.report_warning(msg)
1560 def _float(self, v, name, fatal=False, **kwargs):
1561 res = float_or_none(v, **kwargs)
1563 msg = 'Failed to extract %s: Could not parse value %r' % (name, v)
1565 raise ExtractorError(msg)
1567 self._downloader.report_warning(msg)
1570 def _set_cookie(self, domain, name, value, expire_time=None):
1571 cookie = compat_cookiejar.Cookie(
1572 0, name, value, None, None, domain, None,
1573 None, '/', True, False, expire_time, '', None, None, None)
1574 self._downloader.cookiejar.set_cookie(cookie)
1576 def _get_cookies(self, url):
1577 """ Return a compat_cookies.SimpleCookie with the cookies for the url """
1578 req = sanitized_Request(url)
1579 self._downloader.cookiejar.add_cookie_header(req)
1580 return compat_cookies.SimpleCookie(req.get_header('Cookie'))
1582 def get_testcases(self, include_onlymatching=False):
1583 t = getattr(self, '_TEST', None)
1585 assert not hasattr(self, '_TESTS'), \
1586 '%s has _TEST and _TESTS' % type(self).__name__
1589 tests = getattr(self, '_TESTS', [])
1591 if not include_onlymatching and t.get('only_matching', False):
1593 t['name'] = type(self).__name__[:-len('IE')]
1596 def is_suitable(self, age_limit):
1597 """ Test whether the extractor is generally suitable for the given
1598 age limit (i.e. pornographic sites are not, all others usually are) """
1600 any_restricted = False
1601 for tc in self.get_testcases(include_onlymatching=False):
1602 if 'playlist' in tc:
1603 tc = tc['playlist'][0]
1604 is_restricted = age_restricted(
1605 tc.get('info_dict', {}).get('age_limit'), age_limit)
1606 if not is_restricted:
1608 any_restricted = any_restricted or is_restricted
1609 return not any_restricted
1611 def extract_subtitles(self, *args, **kwargs):
1612 if (self._downloader.params.get('writesubtitles', False) or
1613 self._downloader.params.get('listsubtitles')):
1614 return self._get_subtitles(*args, **kwargs)
1617 def _get_subtitles(self, *args, **kwargs):
1618 raise NotImplementedError('This method must be implemented by subclasses')
1621 def _merge_subtitle_items(subtitle_list1, subtitle_list2):
1622 """ Merge subtitle items for one language. Items with duplicated URLs
1623 will be dropped. """
1624 list1_urls = set([item['url'] for item in subtitle_list1])
1625 ret = list(subtitle_list1)
1626 ret.extend([item for item in subtitle_list2 if item['url'] not in list1_urls])
1630 def _merge_subtitles(cls, subtitle_dict1, subtitle_dict2):
1631 """ Merge two subtitle dictionaries, language by language. """
1632 ret = dict(subtitle_dict1)
1633 for lang in subtitle_dict2:
1634 ret[lang] = cls._merge_subtitle_items(subtitle_dict1.get(lang, []), subtitle_dict2[lang])
1637 def extract_automatic_captions(self, *args, **kwargs):
1638 if (self._downloader.params.get('writeautomaticsub', False) or
1639 self._downloader.params.get('listsubtitles')):
1640 return self._get_automatic_captions(*args, **kwargs)
1643 def _get_automatic_captions(self, *args, **kwargs):
1644 raise NotImplementedError('This method must be implemented by subclasses')
1646 def mark_watched(self, *args, **kwargs):
1647 if (self._downloader.params.get('mark_watched', False) and
1648 (self._get_login_info()[0] is not None or
1649 self._downloader.params.get('cookiefile') is not None)):
1650 self._mark_watched(*args, **kwargs)
1652 def _mark_watched(self, *args, **kwargs):
1653 raise NotImplementedError('This method must be implemented by subclasses')
1656 class SearchInfoExtractor(InfoExtractor):
1658 Base class for paged search queries extractors.
1659 They accept URLs in the format _SEARCH_KEY(|all|[0-9]):{query}
1660 Instances should define _SEARCH_KEY and _MAX_RESULTS.
1664 def _make_valid_url(cls):
1665 return r'%s(?P<prefix>|[1-9][0-9]*|all):(?P<query>[\s\S]+)' % cls._SEARCH_KEY
1668 def suitable(cls, url):
1669 return re.match(cls._make_valid_url(), url) is not None
1671 def _real_extract(self, query):
1672 mobj = re.match(self._make_valid_url(), query)
1674 raise ExtractorError('Invalid search query "%s"' % query)
1676 prefix = mobj.group('prefix')
1677 query = mobj.group('query')
1679 return self._get_n_results(query, 1)
1680 elif prefix == 'all':
1681 return self._get_n_results(query, self._MAX_RESULTS)
1685 raise ExtractorError('invalid download number %s for query "%s"' % (n, query))
1686 elif n > self._MAX_RESULTS:
1687 self._downloader.report_warning('%s returns max %i results (you requested %i)' % (self._SEARCH_KEY, self._MAX_RESULTS, n))
1688 n = self._MAX_RESULTS
1689 return self._get_n_results(query, n)
1691 def _get_n_results(self, query, n):
1692 """Get a specified number of results for a query"""
1693 raise NotImplementedError('This method must be implemented by subclasses')
1696 def SEARCH_KEY(self):
1697 return self._SEARCH_KEY