1 from __future__ import unicode_literals
15 from ..compat import (
18 compat_etree_fromstring,
24 compat_urllib_parse_unquote,
25 compat_urllib_parse_urlencode,
26 compat_urllib_request,
29 from ..downloader.f4m import remove_encrypted_media
59 parse_m3u8_attributes,
65 class InfoExtractor(object):
66 """Information Extractor class.
68 Information extractors are the classes that, given a URL, extract
69 information about the video (or videos) the URL refers to. This
70 information includes the real video URL, the video title, author and
71 others. The information is stored in a dictionary which is then
72 passed to the YoutubeDL. The YoutubeDL processes this
73 information possibly downloading the video to the file system, among
74 other possible outcomes.
76 The type field determines the type of the result.
77 By far the most common value (and the default if _type is missing) is
78 "video", which indicates a single video.
80 For a video, the dictionaries must include the following fields:
83 title: Video title, unescaped.
85 Additionally, it must contain either a formats entry or a url one:
87 formats: A list of dictionaries for each format available, ordered
88 from worst to best quality.
91 * url Mandatory. The URL of the video file
93 The URL of the manifest file in case of
94 fragmented media (DASH, hls, hds)
95 * ext Will be calculated from URL if missing
96 * format A human-readable description of the format
97 ("mp4 container with h264/opus").
98 Calculated from the format_id, width, height.
99 and format_note fields if missing.
100 * format_id A short description of the format
101 ("mp4_h264_opus" or "19").
102 Technically optional, but strongly recommended.
103 * format_note Additional info about the format
104 ("3D" or "DASH video")
105 * width Width of the video, if known
106 * height Height of the video, if known
107 * resolution Textual description of width and height
108 * tbr Average bitrate of audio and video in KBit/s
109 * abr Average audio bitrate in KBit/s
110 * acodec Name of the audio codec in use
111 * asr Audio sampling rate in Hertz
112 * vbr Average video bitrate in KBit/s
114 * vcodec Name of the video codec in use
115 * container Name of the container format
116 * filesize The number of bytes, if known in advance
117 * filesize_approx An estimate for the number of bytes
118 * player_url SWF Player URL (used for rtmpdump).
119 * protocol The protocol that will be used for the actual
120 download, lower-case.
121 "http", "https", "rtsp", "rtmp", "rtmpe",
122 "m3u8", "m3u8_native" or "http_dash_segments".
123 * fragments A list of fragments of the fragmented media,
124 with the following entries:
125 * "url" (mandatory) - fragment's URL
126 * "duration" (optional, int or float)
127 * "filesize" (optional, int)
128 * preference Order number of this format. If this field is
129 present and not None, the formats get sorted
130 by this field, regardless of all other values.
131 -1 for default (order by other properties),
132 -2 or smaller for less than default.
133 < -1000 to hide the format (if there is
134 another one which is strictly better)
135 * language Language code, e.g. "de" or "en-US".
136 * language_preference Is this in the language mentioned in
138 10 if it's what the URL is about,
139 -1 for default (don't know),
140 -10 otherwise, other values reserved for now.
141 * quality Order number of the video quality of this
142 format, irrespective of the file format.
143 -1 for default (order by other properties),
144 -2 or smaller for less than default.
145 * source_preference Order number for this video source
146 (quality takes higher priority)
147 -1 for default (order by other properties),
148 -2 or smaller for less than default.
149 * http_headers A dictionary of additional HTTP headers
150 to add to the request.
151 * stretched_ratio If given and not 1, indicates that the
152 video's pixels are not square.
153 width : height ratio as float.
154 * no_resume The server does not support resuming the
155 (HTTP or RTMP) download. Boolean.
157 url: Final video URL.
158 ext: Video filename extension.
159 format: The video format, defaults to ext (used for --get-format)
160 player_url: SWF Player URL (used for rtmpdump).
162 The following fields are optional:
164 alt_title: A secondary title of the video.
165 display_id An alternative identifier for the video, not necessarily
166 unique, but available before title. Typically, id is
167 something like "4234987", title "Dancing naked mole rats",
168 and display_id "dancing-naked-mole-rats"
169 thumbnails: A list of dictionaries, with the following entries:
170 * "id" (optional, string) - Thumbnail format ID
172 * "preference" (optional, int) - quality of the image
173 * "width" (optional, int)
174 * "height" (optional, int)
175 * "resolution" (optional, string "{width}x{height"},
177 * "filesize" (optional, int)
178 thumbnail: Full URL to a video thumbnail image.
179 description: Full video description.
180 uploader: Full name of the video uploader.
181 license: License name the video is licensed under.
182 creator: The creator of the video.
183 release_date: The date (YYYYMMDD) when the video was released.
184 timestamp: UNIX timestamp of the moment the video became available.
185 upload_date: Video upload date (YYYYMMDD).
186 If not explicitly set, calculated from timestamp.
187 uploader_id: Nickname or id of the video uploader.
188 uploader_url: Full URL to a personal webpage of the video uploader.
189 location: Physical location where the video was filmed.
190 subtitles: The available subtitles as a dictionary in the format
191 {language: subformats}. "subformats" is a list sorted from
192 lower to higher preference, each element is a dictionary
193 with the "ext" entry and one of:
194 * "data": The subtitles file contents
195 * "url": A URL pointing to the subtitles file
196 "ext" will be calculated from URL if missing
197 automatic_captions: Like 'subtitles', used by the YoutubeIE for
198 automatically generated captions
199 duration: Length of the video in seconds, as an integer or float.
200 view_count: How many users have watched the video on the platform.
201 like_count: Number of positive ratings of the video
202 dislike_count: Number of negative ratings of the video
203 repost_count: Number of reposts of the video
204 average_rating: Average rating give by users, the scale used depends on the webpage
205 comment_count: Number of comments on the video
206 comments: A list of comments, each with one or more of the following
207 properties (all but one of text or html optional):
208 * "author" - human-readable name of the comment author
209 * "author_id" - user ID of the comment author
211 * "html" - Comment as HTML
212 * "text" - Plain text of the comment
213 * "timestamp" - UNIX timestamp of comment
214 * "parent" - ID of the comment this one is replying to.
215 Set to "root" to indicate that this is a
216 comment to the original video.
217 age_limit: Age restriction for the video, as an integer (years)
218 webpage_url: The URL to the video webpage, if given to youtube-dl it
219 should allow to get the same result again. (It will be set
220 by YoutubeDL if it's missing)
221 categories: A list of categories that the video falls in, for example
223 tags: A list of tags assigned to the video, e.g. ["sweden", "pop music"]
224 is_live: True, False, or None (=unknown). Whether this video is a
225 live stream that goes on instead of a fixed-length video.
226 start_time: Time in seconds where the reproduction should start, as
227 specified in the URL.
228 end_time: Time in seconds where the reproduction should end, as
229 specified in the URL.
231 The following fields should only be used when the video belongs to some logical
234 chapter: Name or title of the chapter the video belongs to.
235 chapter_number: Number of the chapter the video belongs to, as an integer.
236 chapter_id: Id of the chapter the video belongs to, as a unicode string.
238 The following fields should only be used when the video is an episode of some
239 series, programme or podcast:
241 series: Title of the series or programme the video episode belongs to.
242 season: Title of the season the video episode belongs to.
243 season_number: Number of the season the video episode belongs to, as an integer.
244 season_id: Id of the season the video episode belongs to, as a unicode string.
245 episode: Title of the video episode. Unlike mandatory video title field,
246 this field should denote the exact title of the video episode
247 without any kind of decoration.
248 episode_number: Number of the video episode within a season, as an integer.
249 episode_id: Id of the video episode, as a unicode string.
251 The following fields should only be used when the media is a track or a part of
254 track: Title of the track.
255 track_number: Number of the track within an album or a disc, as an integer.
256 track_id: Id of the track (useful in case of custom indexing, e.g. 6.iii),
258 artist: Artist(s) of the track.
259 genre: Genre(s) of the track.
260 album: Title of the album the track belongs to.
261 album_type: Type of the album (e.g. "Demo", "Full-length", "Split", "Compilation", etc).
262 album_artist: List of all artists appeared on the album (e.g.
263 "Ash Borer / Fell Voices" or "Various Artists", useful for splits
265 disc_number: Number of the disc or other physical medium the track belongs to,
267 release_year: Year (YYYY) when the album was released.
269 Unless mentioned otherwise, the fields should be Unicode strings.
271 Unless mentioned otherwise, None is equivalent to absence of information.
274 _type "playlist" indicates multiple videos.
275 There must be a key "entries", which is a list, an iterable, or a PagedList
276 object, each element of which is a valid dictionary by this specification.
278 Additionally, playlists can have "title", "description" and "id" attributes
279 with the same semantics as videos (see above).
282 _type "multi_video" indicates that there are multiple videos that
283 form a single show, for examples multiple acts of an opera or TV episode.
284 It must have an entries key like a playlist and contain all the keys
285 required for a video at the same time.
288 _type "url" indicates that the video must be extracted from another
289 location, possibly by a different extractor. Its only required key is:
290 "url" - the next URL to extract.
291 The key "ie_key" can be set to the class name (minus the trailing "IE",
292 e.g. "Youtube") if the extractor class is known in advance.
293 Additionally, the dictionary may have any properties of the resolved entity
294 known in advance, for example "title" if the title of the referred video is
298 _type "url_transparent" entities have the same specification as "url", but
299 indicate that the given additional information is more precise than the one
300 associated with the resolved URL.
301 This is useful when a site employs a video service that hosts the video and
302 its technical metadata, but that video service does not embed a useful
303 title, description etc.
306 Subclasses of this one should re-define the _real_initialize() and
307 _real_extract() methods and define a _VALID_URL regexp.
308 Probably, they should also be added to the list of extractors.
310 Finally, the _WORKING attribute should be set to False for broken IEs
311 in order to warn the users and skip the tests.
318 def __init__(self, downloader=None):
319 """Constructor. Receives an optional downloader."""
321 self.set_downloader(downloader)
324 def suitable(cls, url):
325 """Receives a URL and returns True if suitable for this IE."""
327 # This does not use has/getattr intentionally - we want to know whether
328 # we have cached the regexp for *this* class, whereas getattr would also
329 # match the superclass
330 if '_VALID_URL_RE' not in cls.__dict__:
331 cls._VALID_URL_RE = re.compile(cls._VALID_URL)
332 return cls._VALID_URL_RE.match(url) is not None
335 def _match_id(cls, url):
336 if '_VALID_URL_RE' not in cls.__dict__:
337 cls._VALID_URL_RE = re.compile(cls._VALID_URL)
338 m = cls._VALID_URL_RE.match(url)
344 """Getter method for _WORKING."""
347 def initialize(self):
348 """Initializes an instance (authentication, etc)."""
350 self._real_initialize()
353 def extract(self, url):
354 """Extracts URL information and returns it in list of dicts."""
357 return self._real_extract(url)
358 except ExtractorError:
360 except compat_http_client.IncompleteRead as e:
361 raise ExtractorError('A network error has occurred.', cause=e, expected=True)
362 except (KeyError, StopIteration) as e:
363 raise ExtractorError('An extractor error has occurred.', cause=e)
365 def set_downloader(self, downloader):
366 """Sets the downloader for this IE."""
367 self._downloader = downloader
369 def _real_initialize(self):
370 """Real initialization process. Redefine in subclasses."""
373 def _real_extract(self, url):
374 """Real extraction process. Redefine in subclasses."""
379 """A string for getting the InfoExtractor with get_info_extractor"""
380 return compat_str(cls.__name__[:-2])
384 return compat_str(type(self).__name__[:-2])
386 def _request_webpage(self, url_or_request, video_id, note=None, errnote=None, fatal=True, data=None, headers={}, query={}):
387 """ Returns the response handle """
389 self.report_download_webpage(video_id)
390 elif note is not False:
392 self.to_screen('%s' % (note,))
394 self.to_screen('%s: %s' % (video_id, note))
395 if isinstance(url_or_request, compat_urllib_request.Request):
396 url_or_request = update_Request(
397 url_or_request, data=data, headers=headers, query=query)
400 url_or_request = update_url_query(url_or_request, query)
401 if data is not None or headers:
402 url_or_request = sanitized_Request(url_or_request, data, headers)
404 return self._downloader.urlopen(url_or_request)
405 except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
409 errnote = 'Unable to download webpage'
411 errmsg = '%s: %s' % (errnote, error_to_compat_str(err))
413 raise ExtractorError(errmsg, sys.exc_info()[2], cause=err)
415 self._downloader.report_warning(errmsg)
418 def _download_webpage_handle(self, url_or_request, video_id, note=None, errnote=None, fatal=True, encoding=None, data=None, headers={}, query={}):
419 """ Returns a tuple (page content as string, URL handle) """
420 # Strip hashes from the URL (#1038)
421 if isinstance(url_or_request, (compat_str, str)):
422 url_or_request = url_or_request.partition('#')[0]
424 urlh = self._request_webpage(url_or_request, video_id, note, errnote, fatal, data=data, headers=headers, query=query)
428 content = self._webpage_read_content(urlh, url_or_request, video_id, note, errnote, fatal, encoding=encoding)
429 return (content, urlh)
432 def _guess_encoding_from_content(content_type, webpage_bytes):
433 m = re.match(r'[a-zA-Z0-9_.-]+/[a-zA-Z0-9_.-]+\s*;\s*charset=(.+)', content_type)
435 encoding = m.group(1)
437 m = re.search(br'<meta[^>]+charset=[\'"]?([^\'")]+)[ /\'">]',
438 webpage_bytes[:1024])
440 encoding = m.group(1).decode('ascii')
441 elif webpage_bytes.startswith(b'\xff\xfe'):
448 def _webpage_read_content(self, urlh, url_or_request, video_id, note=None, errnote=None, fatal=True, prefix=None, encoding=None):
449 content_type = urlh.headers.get('Content-Type', '')
450 webpage_bytes = urlh.read()
451 if prefix is not None:
452 webpage_bytes = prefix + webpage_bytes
454 encoding = self._guess_encoding_from_content(content_type, webpage_bytes)
455 if self._downloader.params.get('dump_intermediate_pages', False):
457 url = url_or_request.get_full_url()
458 except AttributeError:
460 self.to_screen('Dumping request to ' + url)
461 dump = base64.b64encode(webpage_bytes).decode('ascii')
462 self._downloader.to_screen(dump)
463 if self._downloader.params.get('write_pages', False):
465 url = url_or_request.get_full_url()
466 except AttributeError:
468 basen = '%s_%s' % (video_id, url)
470 h = '___' + hashlib.md5(basen.encode('utf-8')).hexdigest()
471 basen = basen[:240 - len(h)] + h
472 raw_filename = basen + '.dump'
473 filename = sanitize_filename(raw_filename, restricted=True)
474 self.to_screen('Saving request to ' + filename)
475 # Working around MAX_PATH limitation on Windows (see
476 # http://msdn.microsoft.com/en-us/library/windows/desktop/aa365247(v=vs.85).aspx)
477 if compat_os_name == 'nt':
478 absfilepath = os.path.abspath(filename)
479 if len(absfilepath) > 259:
480 filename = '\\\\?\\' + absfilepath
481 with open(filename, 'wb') as outf:
482 outf.write(webpage_bytes)
485 content = webpage_bytes.decode(encoding, 'replace')
487 content = webpage_bytes.decode('utf-8', 'replace')
489 if ('<title>Access to this site is blocked</title>' in content and
490 'Websense' in content[:512]):
491 msg = 'Access to this webpage has been blocked by Websense filtering software in your network.'
492 blocked_iframe = self._html_search_regex(
493 r'<iframe src="([^"]+)"', content,
494 'Websense information URL', default=None)
496 msg += ' Visit %s for more details' % blocked_iframe
497 raise ExtractorError(msg, expected=True)
498 if '<title>The URL you requested has been blocked</title>' in content[:512]:
500 'Access to this webpage has been blocked by Indian censorship. '
501 'Use a VPN or proxy server (with --proxy) to route around it.')
502 block_msg = self._html_search_regex(
503 r'</h1><p>(.*?)</p>',
504 content, 'block message', default=None)
506 msg += ' (Message: "%s")' % block_msg.replace('\n', ' ')
507 raise ExtractorError(msg, expected=True)
511 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={}):
512 """ Returns the data of the page as a string """
515 while success is False:
517 res = self._download_webpage_handle(url_or_request, video_id, note, errnote, fatal, encoding=encoding, data=data, headers=headers, query=query)
519 except compat_http_client.IncompleteRead as e:
521 if try_count >= tries:
523 self._sleep(timeout, video_id)
530 def _download_xml(self, url_or_request, video_id,
531 note='Downloading XML', errnote='Unable to download XML',
532 transform_source=None, fatal=True, encoding=None, data=None, headers={}, query={}):
533 """Return the xml as an xml.etree.ElementTree.Element"""
534 xml_string = self._download_webpage(
535 url_or_request, video_id, note, errnote, fatal=fatal, encoding=encoding, data=data, headers=headers, query=query)
536 if xml_string is False:
539 xml_string = transform_source(xml_string)
540 return compat_etree_fromstring(xml_string.encode('utf-8'))
542 def _download_json(self, url_or_request, video_id,
543 note='Downloading JSON metadata',
544 errnote='Unable to download JSON metadata',
545 transform_source=None,
546 fatal=True, encoding=None, data=None, headers={}, query={}):
547 json_string = self._download_webpage(
548 url_or_request, video_id, note, errnote, fatal=fatal,
549 encoding=encoding, data=data, headers=headers, query=query)
550 if (not fatal) and json_string is False:
552 return self._parse_json(
553 json_string, video_id, transform_source=transform_source, fatal=fatal)
555 def _parse_json(self, json_string, video_id, transform_source=None, fatal=True):
557 json_string = transform_source(json_string)
559 return json.loads(json_string)
560 except ValueError as ve:
561 errmsg = '%s: Failed to parse JSON ' % video_id
563 raise ExtractorError(errmsg, cause=ve)
565 self.report_warning(errmsg + str(ve))
567 def report_warning(self, msg, video_id=None):
568 idstr = '' if video_id is None else '%s: ' % video_id
569 self._downloader.report_warning(
570 '[%s] %s%s' % (self.IE_NAME, idstr, msg))
572 def to_screen(self, msg):
573 """Print msg to screen, prefixing it with '[ie_name]'"""
574 self._downloader.to_screen('[%s] %s' % (self.IE_NAME, msg))
576 def report_extraction(self, id_or_name):
577 """Report information extraction."""
578 self.to_screen('%s: Extracting information' % id_or_name)
580 def report_download_webpage(self, video_id):
581 """Report webpage download."""
582 self.to_screen('%s: Downloading webpage' % video_id)
584 def report_age_confirmation(self):
585 """Report attempt to confirm age."""
586 self.to_screen('Confirming age')
588 def report_login(self):
589 """Report attempt to log in."""
590 self.to_screen('Logging in')
593 def raise_login_required(msg='This video is only available for registered users'):
594 raise ExtractorError(
595 '%s. Use --username and --password or --netrc to provide account credentials.' % msg,
599 def raise_geo_restricted(msg='This video is not available from your location due to geo restriction'):
600 raise ExtractorError(
601 '%s. You might want to use --proxy to workaround.' % msg,
604 # Methods for following #608
606 def url_result(url, ie=None, video_id=None, video_title=None):
607 """Returns a URL that points to a page that should be processed"""
608 # TODO: ie should be the class used for getting the info
609 video_info = {'_type': 'url',
612 if video_id is not None:
613 video_info['id'] = video_id
614 if video_title is not None:
615 video_info['title'] = video_title
619 def playlist_result(entries, playlist_id=None, playlist_title=None, playlist_description=None):
620 """Returns a playlist"""
621 video_info = {'_type': 'playlist',
624 video_info['id'] = playlist_id
626 video_info['title'] = playlist_title
627 if playlist_description:
628 video_info['description'] = playlist_description
631 def _search_regex(self, pattern, string, name, default=NO_DEFAULT, fatal=True, flags=0, group=None):
633 Perform a regex search on the given string, using a single or a list of
634 patterns returning the first matching group.
635 In case of failure return a default value or raise a WARNING or a
636 RegexNotFoundError, depending on fatal, specifying the field name.
638 if isinstance(pattern, (str, compat_str, compiled_regex_type)):
639 mobj = re.search(pattern, string, flags)
642 mobj = re.search(p, string, flags)
646 if not self._downloader.params.get('no_color') and compat_os_name != 'nt' and sys.stderr.isatty():
647 _name = '\033[0;34m%s\033[0m' % name
653 # return the first matching group
654 return next(g for g in mobj.groups() if g is not None)
656 return mobj.group(group)
657 elif default is not NO_DEFAULT:
660 raise RegexNotFoundError('Unable to extract %s' % _name)
662 self._downloader.report_warning('unable to extract %s' % _name + bug_reports_message())
665 def _html_search_regex(self, pattern, string, name, default=NO_DEFAULT, fatal=True, flags=0, group=None):
667 Like _search_regex, but strips HTML tags and unescapes entities.
669 res = self._search_regex(pattern, string, name, default, fatal, flags, group)
671 return clean_html(res).strip()
675 def _get_netrc_login_info(self, netrc_machine=None):
678 netrc_machine = netrc_machine or self._NETRC_MACHINE
680 if self._downloader.params.get('usenetrc', False):
682 info = netrc.netrc().authenticators(netrc_machine)
687 raise netrc.NetrcParseError(
688 'No authenticators for %s' % netrc_machine)
689 except (IOError, netrc.NetrcParseError) as err:
690 self._downloader.report_warning(
691 'parsing .netrc: %s' % error_to_compat_str(err))
693 return username, password
695 def _get_login_info(self, username_option='username', password_option='password', netrc_machine=None):
697 Get the login info as (username, password)
698 First look for the manually specified credentials using username_option
699 and password_option as keys in params dictionary. If no such credentials
700 available look in the netrc file using the netrc_machine or _NETRC_MACHINE
702 If there's no info available, return (None, None)
704 if self._downloader is None:
707 downloader_params = self._downloader.params
709 # Attempt to use provided username and password or .netrc data
710 if downloader_params.get(username_option) is not None:
711 username = downloader_params[username_option]
712 password = downloader_params[password_option]
714 username, password = self._get_netrc_login_info(netrc_machine)
716 return username, password
718 def _get_tfa_info(self, note='two-factor verification code'):
720 Get the two-factor authentication info
721 TODO - asking the user will be required for sms/phone verify
722 currently just uses the command line option
723 If there's no info available, return None
725 if self._downloader is None:
727 downloader_params = self._downloader.params
729 if downloader_params.get('twofactor') is not None:
730 return downloader_params['twofactor']
732 return compat_getpass('Type %s and press [Return]: ' % note)
734 # Helper functions for extracting OpenGraph info
736 def _og_regexes(prop):
737 content_re = r'content=(?:"([^"]+?)"|\'([^\']+?)\'|\s*([^\s"\'=<>`]+?))'
738 property_re = (r'(?:name|property)=(?:\'og:%(prop)s\'|"og:%(prop)s"|\s*og:%(prop)s\b)'
739 % {'prop': re.escape(prop)})
740 template = r'<meta[^>]+?%s[^>]+?%s'
742 template % (property_re, content_re),
743 template % (content_re, property_re),
747 def _meta_regex(prop):
748 return r'''(?isx)<meta
749 (?=[^>]+(?:itemprop|name|property|id|http-equiv)=(["\']?)%s\1)
750 [^>]+?content=(["\'])(?P<content>.*?)\2''' % re.escape(prop)
752 def _og_search_property(self, prop, html, name=None, **kargs):
753 if not isinstance(prop, (list, tuple)):
756 name = 'OpenGraph %s' % prop[0]
759 og_regexes.extend(self._og_regexes(p))
760 escaped = self._search_regex(og_regexes, html, name, flags=re.DOTALL, **kargs)
763 return unescapeHTML(escaped)
765 def _og_search_thumbnail(self, html, **kargs):
766 return self._og_search_property('image', html, 'thumbnail URL', fatal=False, **kargs)
768 def _og_search_description(self, html, **kargs):
769 return self._og_search_property('description', html, fatal=False, **kargs)
771 def _og_search_title(self, html, **kargs):
772 return self._og_search_property('title', html, **kargs)
774 def _og_search_video_url(self, html, name='video url', secure=True, **kargs):
775 regexes = self._og_regexes('video') + self._og_regexes('video:url')
777 regexes = self._og_regexes('video:secure_url') + regexes
778 return self._html_search_regex(regexes, html, name, **kargs)
780 def _og_search_url(self, html, **kargs):
781 return self._og_search_property('url', html, **kargs)
783 def _html_search_meta(self, name, html, display_name=None, fatal=False, **kwargs):
784 if not isinstance(name, (list, tuple)):
786 if display_name is None:
787 display_name = name[0]
788 return self._html_search_regex(
789 [self._meta_regex(n) for n in name],
790 html, display_name, fatal=fatal, group='content', **kwargs)
792 def _dc_search_uploader(self, html):
793 return self._html_search_meta('dc.creator', html, 'uploader')
795 def _rta_search(self, html):
796 # See http://www.rtalabel.org/index.php?content=howtofaq#single
797 if re.search(r'(?ix)<meta\s+name="rating"\s+'
798 r' content="RTA-5042-1996-1400-1577-RTA"',
803 def _media_rating_search(self, html):
804 # See http://www.tjg-designs.com/WP/metadata-code-examples-adding-metadata-to-your-web-pages/
805 rating = self._html_search_meta('rating', html)
817 return RATING_TABLE.get(rating.lower())
819 def _family_friendly_search(self, html):
820 # See http://schema.org/VideoObject
821 family_friendly = self._html_search_meta('isFamilyFriendly', html)
823 if not family_friendly:
832 return RATING_TABLE.get(family_friendly.lower())
834 def _twitter_search_player(self, html):
835 return self._html_search_meta('twitter:player', html,
836 'twitter card player')
838 def _search_json_ld(self, html, video_id, expected_type=None, **kwargs):
839 json_ld = self._search_regex(
840 r'(?s)<script[^>]+type=(["\'])application/ld\+json\1[^>]*>(?P<json_ld>.+?)</script>',
841 html, 'JSON-LD', group='json_ld', **kwargs)
842 default = kwargs.get('default', NO_DEFAULT)
844 return default if default is not NO_DEFAULT else {}
845 # JSON-LD may be malformed and thus `fatal` should be respected.
846 # At the same time `default` may be passed that assumes `fatal=False`
847 # for _search_regex. Let's simulate the same behavior here as well.
848 fatal = kwargs.get('fatal', True) if default == NO_DEFAULT else False
849 return self._json_ld(json_ld, video_id, fatal=fatal, expected_type=expected_type)
851 def _json_ld(self, json_ld, video_id, fatal=True, expected_type=None):
852 if isinstance(json_ld, compat_str):
853 json_ld = self._parse_json(json_ld, video_id, fatal=fatal)
857 if not isinstance(json_ld, (list, tuple, dict)):
859 if isinstance(json_ld, dict):
862 if e.get('@context') == 'http://schema.org':
863 item_type = e.get('@type')
864 if expected_type is not None and expected_type != item_type:
866 if item_type == 'TVEpisode':
868 'episode': unescapeHTML(e.get('name')),
869 'episode_number': int_or_none(e.get('episodeNumber')),
870 'description': unescapeHTML(e.get('description')),
872 part_of_season = e.get('partOfSeason')
873 if isinstance(part_of_season, dict) and part_of_season.get('@type') == 'TVSeason':
874 info['season_number'] = int_or_none(part_of_season.get('seasonNumber'))
875 part_of_series = e.get('partOfSeries') or e.get('partOfTVSeries')
876 if isinstance(part_of_series, dict) and part_of_series.get('@type') == 'TVSeries':
877 info['series'] = unescapeHTML(part_of_series.get('name'))
878 elif item_type == 'Article':
880 'timestamp': parse_iso8601(e.get('datePublished')),
881 'title': unescapeHTML(e.get('headline')),
882 'description': unescapeHTML(e.get('articleBody')),
884 elif item_type == 'VideoObject':
886 'url': e.get('contentUrl'),
887 'title': unescapeHTML(e.get('name')),
888 'description': unescapeHTML(e.get('description')),
889 'thumbnail': e.get('thumbnailUrl'),
890 'duration': parse_duration(e.get('duration')),
891 'timestamp': unified_timestamp(e.get('uploadDate')),
892 'filesize': float_or_none(e.get('contentSize')),
893 'tbr': int_or_none(e.get('bitrate')),
894 'width': int_or_none(e.get('width')),
895 'height': int_or_none(e.get('height')),
898 return dict((k, v) for k, v in info.items() if v is not None)
901 def _hidden_inputs(html):
902 html = re.sub(r'<!--(?:(?!<!--).)*-->', '', html)
904 for input in re.findall(r'(?i)(<input[^>]+>)', html):
905 attrs = extract_attributes(input)
908 if attrs.get('type') not in ('hidden', 'submit'):
910 name = attrs.get('name') or attrs.get('id')
911 value = attrs.get('value')
912 if name and value is not None:
913 hidden_inputs[name] = value
916 def _form_hidden_inputs(self, form_id, html):
917 form = self._search_regex(
918 r'(?is)<form[^>]+?id=(["\'])%s\1[^>]*>(?P<form>.+?)</form>' % form_id,
919 html, '%s form' % form_id, group='form')
920 return self._hidden_inputs(form)
922 def _sort_formats(self, formats, field_preference=None):
924 raise ExtractorError('No video formats found')
927 # Automatically determine tbr when missing based on abr and vbr (improves
928 # formats sorting in some cases)
929 if 'tbr' not in f and f.get('abr') is not None and f.get('vbr') is not None:
930 f['tbr'] = f['abr'] + f['vbr']
933 # TODO remove the following workaround
934 from ..utils import determine_ext
935 if not f.get('ext') and 'url' in f:
936 f['ext'] = determine_ext(f['url'])
938 if isinstance(field_preference, (list, tuple)):
941 if f.get(field) is not None
942 else ('' if field == 'format_id' else -1)
943 for field in field_preference)
945 preference = f.get('preference')
946 if preference is None:
948 if f.get('ext') in ['f4f', 'f4m']: # Not yet supported
951 protocol = f.get('protocol') or determine_protocol(f)
952 proto_preference = 0 if protocol in ['http', 'https'] else (-0.5 if protocol == 'rtsp' else -0.1)
954 if f.get('vcodec') == 'none': # audio only
956 if self._downloader.params.get('prefer_free_formats'):
957 ORDER = ['aac', 'mp3', 'm4a', 'webm', 'ogg', 'opus']
959 ORDER = ['webm', 'opus', 'ogg', 'mp3', 'aac', 'm4a']
962 audio_ext_preference = ORDER.index(f['ext'])
964 audio_ext_preference = -1
966 if f.get('acodec') == 'none': # video only
968 if self._downloader.params.get('prefer_free_formats'):
969 ORDER = ['flv', 'mp4', 'webm']
971 ORDER = ['webm', 'flv', 'mp4']
973 ext_preference = ORDER.index(f['ext'])
976 audio_ext_preference = 0
980 f.get('language_preference') if f.get('language_preference') is not None else -1,
981 f.get('quality') if f.get('quality') is not None else -1,
982 f.get('tbr') if f.get('tbr') is not None else -1,
983 f.get('filesize') if f.get('filesize') is not None else -1,
984 f.get('vbr') if f.get('vbr') is not None else -1,
985 f.get('height') if f.get('height') is not None else -1,
986 f.get('width') if f.get('width') is not None else -1,
989 f.get('abr') if f.get('abr') is not None else -1,
990 audio_ext_preference,
991 f.get('fps') if f.get('fps') is not None else -1,
992 f.get('filesize_approx') if f.get('filesize_approx') is not None else -1,
993 f.get('source_preference') if f.get('source_preference') is not None else -1,
994 f.get('format_id') if f.get('format_id') is not None else '',
996 formats.sort(key=_formats_key)
998 def _check_formats(self, formats, video_id):
1000 formats[:] = filter(
1001 lambda f: self._is_valid_url(
1003 item='%s video format' % f.get('format_id') if f.get('format_id') else 'video'),
1007 def _remove_duplicate_formats(formats):
1011 if f['url'] not in format_urls:
1012 format_urls.add(f['url'])
1013 unique_formats.append(f)
1014 formats[:] = unique_formats
1016 def _is_valid_url(self, url, video_id, item='video'):
1017 url = self._proto_relative_url(url, scheme='http:')
1018 # For now assume non HTTP(S) URLs always valid
1019 if not (url.startswith('http://') or url.startswith('https://')):
1022 self._request_webpage(url, video_id, 'Checking %s URL' % item)
1024 except ExtractorError as e:
1025 if isinstance(e.cause, compat_urllib_error.URLError):
1027 '%s: %s URL is invalid, skipping' % (video_id, item))
1031 def http_scheme(self):
1032 """ Either "http:" or "https:", depending on the user's preferences """
1035 if self._downloader.params.get('prefer_insecure', False)
1038 def _proto_relative_url(self, url, scheme=None):
1041 if url.startswith('//'):
1043 scheme = self.http_scheme()
1048 def _sleep(self, timeout, video_id, msg_template=None):
1049 if msg_template is None:
1050 msg_template = '%(video_id)s: Waiting for %(timeout)s seconds'
1051 msg = msg_template % {'video_id': video_id, 'timeout': timeout}
1055 def _extract_f4m_formats(self, manifest_url, video_id, preference=None, f4m_id=None,
1056 transform_source=lambda s: fix_xml_ampersands(s).strip(),
1057 fatal=True, m3u8_id=None):
1058 manifest = self._download_xml(
1059 manifest_url, video_id, 'Downloading f4m manifest',
1060 'Unable to download f4m manifest',
1061 # Some manifests may be malformed, e.g. prosiebensat1 generated manifests
1062 # (see https://github.com/rg3/youtube-dl/issues/6215#issuecomment-121704244)
1063 transform_source=transform_source,
1066 if manifest is False:
1069 return self._parse_f4m_formats(
1070 manifest, manifest_url, video_id, preference=preference, f4m_id=f4m_id,
1071 transform_source=transform_source, fatal=fatal, m3u8_id=m3u8_id)
1073 def _parse_f4m_formats(self, manifest, manifest_url, video_id, preference=None, f4m_id=None,
1074 transform_source=lambda s: fix_xml_ampersands(s).strip(),
1075 fatal=True, m3u8_id=None):
1076 # currently youtube-dl cannot decode the playerVerificationChallenge as Akamai uses Adobe Alchemy
1077 akamai_pv = manifest.find('{http://ns.adobe.com/f4m/1.0}pv-2.0')
1078 if akamai_pv is not None and ';' in akamai_pv.text:
1079 playerVerificationChallenge = akamai_pv.text.split(';')[0]
1080 if playerVerificationChallenge.strip() != '':
1084 manifest_version = '1.0'
1085 media_nodes = manifest.findall('{http://ns.adobe.com/f4m/1.0}media')
1087 manifest_version = '2.0'
1088 media_nodes = manifest.findall('{http://ns.adobe.com/f4m/2.0}media')
1089 # Remove unsupported DRM protected media from final formats
1090 # rendition (see https://github.com/rg3/youtube-dl/issues/8573).
1091 media_nodes = remove_encrypted_media(media_nodes)
1094 base_url = xpath_text(
1095 manifest, ['{http://ns.adobe.com/f4m/1.0}baseURL', '{http://ns.adobe.com/f4m/2.0}baseURL'],
1096 'base URL', default=None)
1098 base_url = base_url.strip()
1100 bootstrap_info = xpath_element(
1101 manifest, ['{http://ns.adobe.com/f4m/1.0}bootstrapInfo', '{http://ns.adobe.com/f4m/2.0}bootstrapInfo'],
1102 'bootstrap info', default=None)
1105 mime_type = xpath_text(
1106 manifest, ['{http://ns.adobe.com/f4m/1.0}mimeType', '{http://ns.adobe.com/f4m/2.0}mimeType'],
1107 'base URL', default=None)
1108 if mime_type and mime_type.startswith('audio/'):
1111 for i, media_el in enumerate(media_nodes):
1112 tbr = int_or_none(media_el.attrib.get('bitrate'))
1113 width = int_or_none(media_el.attrib.get('width'))
1114 height = int_or_none(media_el.attrib.get('height'))
1115 format_id = '-'.join(filter(None, [f4m_id, compat_str(i if tbr is None else tbr)]))
1116 # If <bootstrapInfo> is present, the specified f4m is a
1117 # stream-level manifest, and only set-level manifests may refer to
1118 # external resources. See section 11.4 and section 4 of F4M spec
1119 if bootstrap_info is None:
1121 # @href is introduced in 2.0, see section 11.6 of F4M spec
1122 if manifest_version == '2.0':
1123 media_url = media_el.attrib.get('href')
1124 if media_url is None:
1125 media_url = media_el.attrib.get('url')
1129 media_url if media_url.startswith('http://') or media_url.startswith('https://')
1130 else ((base_url or '/'.join(manifest_url.split('/')[:-1])) + '/' + media_url))
1131 # If media_url is itself a f4m manifest do the recursive extraction
1132 # since bitrates in parent manifest (this one) and media_url manifest
1133 # may differ leading to inability to resolve the format by requested
1134 # bitrate in f4m downloader
1135 ext = determine_ext(manifest_url)
1137 f4m_formats = self._extract_f4m_formats(
1138 manifest_url, video_id, preference=preference, f4m_id=f4m_id,
1139 transform_source=transform_source, fatal=fatal)
1140 # Sometimes stream-level manifest contains single media entry that
1141 # does not contain any quality metadata (e.g. http://matchtv.ru/#live-player).
1142 # At the same time parent's media entry in set-level manifest may
1143 # contain it. We will copy it from parent in such cases.
1144 if len(f4m_formats) == 1:
1147 'tbr': f.get('tbr') or tbr,
1148 'width': f.get('width') or width,
1149 'height': f.get('height') or height,
1150 'format_id': f.get('format_id') if not tbr else format_id,
1153 formats.extend(f4m_formats)
1156 formats.extend(self._extract_m3u8_formats(
1157 manifest_url, video_id, 'mp4', preference=preference,
1158 m3u8_id=m3u8_id, fatal=fatal))
1161 'format_id': format_id,
1162 'url': manifest_url,
1163 'manifest_url': manifest_url,
1164 'ext': 'flv' if bootstrap_info is not None else None,
1169 'preference': preference,
1173 def _m3u8_meta_format(self, m3u8_url, ext=None, preference=None, m3u8_id=None):
1175 'format_id': '-'.join(filter(None, [m3u8_id, 'meta'])),
1179 'preference': preference - 100 if preference else -100,
1180 'resolution': 'multiple',
1181 'format_note': 'Quality selection URL',
1184 def _extract_m3u8_formats(self, m3u8_url, video_id, ext=None,
1185 entry_protocol='m3u8', preference=None,
1186 m3u8_id=None, note=None, errnote=None,
1187 fatal=True, live=False):
1189 res = self._download_webpage_handle(
1191 note=note or 'Downloading m3u8 information',
1192 errnote=errnote or 'Failed to download m3u8 information',
1196 m3u8_doc, urlh = res
1197 m3u8_url = urlh.geturl()
1199 formats = [self._m3u8_meta_format(m3u8_url, ext, preference, m3u8_id)]
1201 format_url = lambda u: (
1203 if re.match(r'^https?://', u)
1204 else compat_urlparse.urljoin(m3u8_url, u))
1206 # We should try extracting formats only from master playlists [1], i.e.
1207 # playlists that describe available qualities. On the other hand media
1208 # playlists [2] should be returned as is since they contain just the media
1209 # without qualities renditions.
1210 # Fortunately, master playlist can be easily distinguished from media
1211 # playlist based on particular tags availability. As of [1, 2] master
1212 # playlist tags MUST NOT appear in a media playist and vice versa.
1213 # As of [3] #EXT-X-TARGETDURATION tag is REQUIRED for every media playlist
1214 # and MUST NOT appear in master playlist thus we can clearly detect media
1215 # playlist with this criterion.
1216 # 1. https://tools.ietf.org/html/draft-pantos-http-live-streaming-17#section-4.3.4
1217 # 2. https://tools.ietf.org/html/draft-pantos-http-live-streaming-17#section-4.3.3
1218 # 3. https://tools.ietf.org/html/draft-pantos-http-live-streaming-17#section-4.3.3.1
1219 if '#EXT-X-TARGETDURATION' in m3u8_doc: # media playlist, return as is
1222 'format_id': m3u8_id,
1224 'protocol': entry_protocol,
1225 'preference': preference,
1229 for line in m3u8_doc.splitlines():
1230 if line.startswith('#EXT-X-STREAM-INF:'):
1231 last_info = parse_m3u8_attributes(line)
1232 elif line.startswith('#EXT-X-MEDIA:'):
1233 media = parse_m3u8_attributes(line)
1234 media_type = media.get('TYPE')
1235 if media_type in ('VIDEO', 'AUDIO'):
1236 media_url = media.get('URI')
1239 for v in (media.get('GROUP-ID'), media.get('NAME')):
1243 'format_id': '-'.join(format_id),
1244 'url': format_url(media_url),
1245 'language': media.get('LANGUAGE'),
1246 'vcodec': 'none' if media_type == 'AUDIO' else None,
1248 'protocol': entry_protocol,
1249 'preference': preference,
1252 # When there is no URI in EXT-X-MEDIA let this tag's
1253 # data be used by regular URI lines below
1255 elif line.startswith('#') or not line.strip():
1258 tbr = int_or_none(last_info.get('AVERAGE-BANDWIDTH') or last_info.get('BANDWIDTH'), scale=1000)
1261 format_id.append(m3u8_id)
1262 # Despite specification does not mention NAME attribute for
1263 # EXT-X-STREAM-INF it still sometimes may be present
1264 stream_name = last_info.get('NAME') or last_media.get('NAME')
1265 # Bandwidth of live streams may differ over time thus making
1266 # format_id unpredictable. So it's better to keep provided
1269 format_id.append(stream_name if stream_name else '%d' % (tbr if tbr else len(formats)))
1270 manifest_url = format_url(line.strip())
1272 'format_id': '-'.join(format_id),
1273 'url': manifest_url,
1274 'manifest_url': manifest_url,
1277 'fps': float_or_none(last_info.get('FRAME-RATE')),
1278 'protocol': entry_protocol,
1279 'preference': preference,
1281 resolution = last_info.get('RESOLUTION')
1283 width_str, height_str = resolution.split('x')
1284 f['width'] = int(width_str)
1285 f['height'] = int(height_str)
1286 # Unified Streaming Platform
1288 r'audio.*?(?:%3D|=)(\d+)(?:-video.*?(?:%3D|=)(\d+))?', f['url'])
1290 abr, vbr = mobj.groups()
1291 abr, vbr = float_or_none(abr, 1000), float_or_none(vbr, 1000)
1296 f.update(parse_codecs(last_info.get('CODECS')))
1303 def _xpath_ns(path, namespace=None):
1307 for c in path.split('/'):
1308 if not c or c == '.':
1311 out.append('{%s}%s' % (namespace, c))
1312 return '/'.join(out)
1314 def _extract_smil_formats(self, smil_url, video_id, fatal=True, f4m_params=None, transform_source=None):
1315 smil = self._download_smil(smil_url, video_id, fatal=fatal, transform_source=transform_source)
1321 namespace = self._parse_smil_namespace(smil)
1323 return self._parse_smil_formats(
1324 smil, smil_url, video_id, namespace=namespace, f4m_params=f4m_params)
1326 def _extract_smil_info(self, smil_url, video_id, fatal=True, f4m_params=None):
1327 smil = self._download_smil(smil_url, video_id, fatal=fatal)
1330 return self._parse_smil(smil, smil_url, video_id, f4m_params=f4m_params)
1332 def _download_smil(self, smil_url, video_id, fatal=True, transform_source=None):
1333 return self._download_xml(
1334 smil_url, video_id, 'Downloading SMIL file',
1335 'Unable to download SMIL file', fatal=fatal, transform_source=transform_source)
1337 def _parse_smil(self, smil, smil_url, video_id, f4m_params=None):
1338 namespace = self._parse_smil_namespace(smil)
1340 formats = self._parse_smil_formats(
1341 smil, smil_url, video_id, namespace=namespace, f4m_params=f4m_params)
1342 subtitles = self._parse_smil_subtitles(smil, namespace=namespace)
1344 video_id = os.path.splitext(url_basename(smil_url))[0]
1348 for meta in smil.findall(self._xpath_ns('./head/meta', namespace)):
1349 name = meta.attrib.get('name')
1350 content = meta.attrib.get('content')
1351 if not name or not content:
1353 if not title and name == 'title':
1355 elif not description and name in ('description', 'abstract'):
1356 description = content
1357 elif not upload_date and name == 'date':
1358 upload_date = unified_strdate(content)
1361 'id': image.get('type'),
1362 'url': image.get('src'),
1363 'width': int_or_none(image.get('width')),
1364 'height': int_or_none(image.get('height')),
1365 } for image in smil.findall(self._xpath_ns('.//image', namespace)) if image.get('src')]
1369 'title': title or video_id,
1370 'description': description,
1371 'upload_date': upload_date,
1372 'thumbnails': thumbnails,
1374 'subtitles': subtitles,
1377 def _parse_smil_namespace(self, smil):
1378 return self._search_regex(
1379 r'(?i)^{([^}]+)?}smil$', smil.tag, 'namespace', default=None)
1381 def _parse_smil_formats(self, smil, smil_url, video_id, namespace=None, f4m_params=None, transform_rtmp_url=None):
1383 for meta in smil.findall(self._xpath_ns('./head/meta', namespace)):
1384 b = meta.get('base') or meta.get('httpBase')
1395 media = smil.findall(self._xpath_ns('.//video', namespace)) + smil.findall(self._xpath_ns('.//audio', namespace))
1396 for medium in media:
1397 src = medium.get('src')
1398 if not src or src in srcs:
1402 bitrate = float_or_none(medium.get('system-bitrate') or medium.get('systemBitrate'), 1000)
1403 filesize = int_or_none(medium.get('size') or medium.get('fileSize'))
1404 width = int_or_none(medium.get('width'))
1405 height = int_or_none(medium.get('height'))
1406 proto = medium.get('proto')
1407 ext = medium.get('ext')
1408 src_ext = determine_ext(src)
1409 streamer = medium.get('streamer') or base
1411 if proto == 'rtmp' or streamer.startswith('rtmp'):
1417 'format_id': 'rtmp-%d' % (rtmp_count if bitrate is None else bitrate),
1419 'filesize': filesize,
1423 if transform_rtmp_url:
1424 streamer, src = transform_rtmp_url(streamer, src)
1425 formats[-1].update({
1431 src_url = src if src.startswith('http') else compat_urlparse.urljoin(base, src)
1432 src_url = src_url.strip()
1434 if proto == 'm3u8' or src_ext == 'm3u8':
1435 m3u8_formats = self._extract_m3u8_formats(
1436 src_url, video_id, ext or 'mp4', m3u8_id='hls', fatal=False)
1437 if len(m3u8_formats) == 1:
1439 m3u8_formats[0].update({
1440 'format_id': 'hls-%d' % (m3u8_count if bitrate is None else bitrate),
1445 formats.extend(m3u8_formats)
1448 if src_ext == 'f4m':
1453 'plugin': 'flowplayer-3.2.0.1',
1455 f4m_url += '&' if '?' in f4m_url else '?'
1456 f4m_url += compat_urllib_parse_urlencode(f4m_params)
1457 formats.extend(self._extract_f4m_formats(f4m_url, video_id, f4m_id='hds', fatal=False))
1460 if src_url.startswith('http') and self._is_valid_url(src, video_id):
1464 'ext': ext or src_ext or 'flv',
1465 'format_id': 'http-%d' % (bitrate or http_count),
1467 'filesize': filesize,
1475 def _parse_smil_subtitles(self, smil, namespace=None, subtitles_lang='en'):
1478 for num, textstream in enumerate(smil.findall(self._xpath_ns('.//textstream', namespace))):
1479 src = textstream.get('src')
1480 if not src or src in urls:
1483 ext = textstream.get('ext') or mimetype2ext(textstream.get('type')) or determine_ext(src)
1484 lang = textstream.get('systemLanguage') or textstream.get('systemLanguageName') or textstream.get('lang') or subtitles_lang
1485 subtitles.setdefault(lang, []).append({
1491 def _extract_xspf_playlist(self, playlist_url, playlist_id, fatal=True):
1492 xspf = self._download_xml(
1493 playlist_url, playlist_id, 'Downloading xpsf playlist',
1494 'Unable to download xspf manifest', fatal=fatal)
1497 return self._parse_xspf(xspf, playlist_id)
1499 def _parse_xspf(self, playlist, playlist_id):
1501 'xspf': 'http://xspf.org/ns/0/',
1502 's1': 'http://static.streamone.nl/player/ns/0',
1506 for track in playlist.findall(xpath_with_ns('./xspf:trackList/xspf:track', NS_MAP)):
1508 track, xpath_with_ns('./xspf:title', NS_MAP), 'title', default=playlist_id)
1509 description = xpath_text(
1510 track, xpath_with_ns('./xspf:annotation', NS_MAP), 'description')
1511 thumbnail = xpath_text(
1512 track, xpath_with_ns('./xspf:image', NS_MAP), 'thumbnail')
1513 duration = float_or_none(
1514 xpath_text(track, xpath_with_ns('./xspf:duration', NS_MAP), 'duration'), 1000)
1517 'url': location.text,
1518 'format_id': location.get(xpath_with_ns('s1:label', NS_MAP)),
1519 'width': int_or_none(location.get(xpath_with_ns('s1:width', NS_MAP))),
1520 'height': int_or_none(location.get(xpath_with_ns('s1:height', NS_MAP))),
1521 } for location in track.findall(xpath_with_ns('./xspf:location', NS_MAP))]
1522 self._sort_formats(formats)
1527 'description': description,
1528 'thumbnail': thumbnail,
1529 'duration': duration,
1534 def _extract_mpd_formats(self, mpd_url, video_id, mpd_id=None, note=None, errnote=None, fatal=True, formats_dict={}):
1535 res = self._download_webpage_handle(
1537 note=note or 'Downloading MPD manifest',
1538 errnote=errnote or 'Failed to download MPD manifest',
1543 mpd_base_url = base_url(urlh.geturl())
1545 return self._parse_mpd_formats(
1546 compat_etree_fromstring(mpd.encode('utf-8')), mpd_id, mpd_base_url,
1547 formats_dict=formats_dict, mpd_url=mpd_url)
1549 def _parse_mpd_formats(self, mpd_doc, mpd_id=None, mpd_base_url='', formats_dict={}, mpd_url=None):
1551 Parse formats from MPD manifest.
1553 1. MPEG-DASH Standard, ISO/IEC 23009-1:2014(E),
1554 http://standards.iso.org/ittf/PubliclyAvailableStandards/c065274_ISO_IEC_23009-1_2014.zip
1555 2. https://en.wikipedia.org/wiki/Dynamic_Adaptive_Streaming_over_HTTP
1557 if mpd_doc.get('type') == 'dynamic':
1560 namespace = self._search_regex(r'(?i)^{([^}]+)?}MPD$', mpd_doc.tag, 'namespace', default=None)
1563 return self._xpath_ns(path, namespace)
1565 def is_drm_protected(element):
1566 return element.find(_add_ns('ContentProtection')) is not None
1568 def extract_multisegment_info(element, ms_parent_info):
1569 ms_info = ms_parent_info.copy()
1571 # As per [1, 5.3.9.2.2] SegmentList and SegmentTemplate share some
1572 # common attributes and elements. We will only extract relevant
1574 def extract_common(source):
1575 segment_timeline = source.find(_add_ns('SegmentTimeline'))
1576 if segment_timeline is not None:
1577 s_e = segment_timeline.findall(_add_ns('S'))
1579 ms_info['total_number'] = 0
1582 r = int(s.get('r', 0))
1583 ms_info['total_number'] += 1 + r
1584 ms_info['s'].append({
1585 't': int(s.get('t', 0)),
1586 # @d is mandatory (see [1, 5.3.9.6.2, Table 17, page 60])
1587 'd': int(s.attrib['d']),
1590 start_number = source.get('startNumber')
1592 ms_info['start_number'] = int(start_number)
1593 timescale = source.get('timescale')
1595 ms_info['timescale'] = int(timescale)
1596 segment_duration = source.get('duration')
1597 if segment_duration:
1598 ms_info['segment_duration'] = int(segment_duration)
1600 def extract_Initialization(source):
1601 initialization = source.find(_add_ns('Initialization'))
1602 if initialization is not None:
1603 ms_info['initialization_url'] = initialization.attrib['sourceURL']
1605 segment_list = element.find(_add_ns('SegmentList'))
1606 if segment_list is not None:
1607 extract_common(segment_list)
1608 extract_Initialization(segment_list)
1609 segment_urls_e = segment_list.findall(_add_ns('SegmentURL'))
1611 ms_info['segment_urls'] = [segment.attrib['media'] for segment in segment_urls_e]
1613 segment_template = element.find(_add_ns('SegmentTemplate'))
1614 if segment_template is not None:
1615 extract_common(segment_template)
1616 media_template = segment_template.get('media')
1618 ms_info['media_template'] = media_template
1619 initialization = segment_template.get('initialization')
1621 ms_info['initialization_url'] = initialization
1623 extract_Initialization(segment_template)
1626 def combine_url(base_url, target_url):
1627 if re.match(r'^https?://', target_url):
1629 return '%s%s%s' % (base_url, '' if base_url.endswith('/') else '/', target_url)
1631 mpd_duration = parse_duration(mpd_doc.get('mediaPresentationDuration'))
1633 for period in mpd_doc.findall(_add_ns('Period')):
1634 period_duration = parse_duration(period.get('duration')) or mpd_duration
1635 period_ms_info = extract_multisegment_info(period, {
1639 for adaptation_set in period.findall(_add_ns('AdaptationSet')):
1640 if is_drm_protected(adaptation_set):
1642 adaption_set_ms_info = extract_multisegment_info(adaptation_set, period_ms_info)
1643 for representation in adaptation_set.findall(_add_ns('Representation')):
1644 if is_drm_protected(representation):
1646 representation_attrib = adaptation_set.attrib.copy()
1647 representation_attrib.update(representation.attrib)
1648 # According to [1, 5.3.7.2, Table 9, page 41], @mimeType is mandatory
1649 mime_type = representation_attrib['mimeType']
1650 content_type = mime_type.split('/')[0]
1651 if content_type == 'text':
1652 # TODO implement WebVTT downloading
1654 elif content_type == 'video' or content_type == 'audio':
1656 for element in (representation, adaptation_set, period, mpd_doc):
1657 base_url_e = element.find(_add_ns('BaseURL'))
1658 if base_url_e is not None:
1659 base_url = base_url_e.text + base_url
1660 if re.match(r'^https?://', base_url):
1662 if mpd_base_url and not re.match(r'^https?://', base_url):
1663 if not mpd_base_url.endswith('/') and not base_url.startswith('/'):
1665 base_url = mpd_base_url + base_url
1666 representation_id = representation_attrib.get('id')
1667 lang = representation_attrib.get('lang')
1668 url_el = representation.find(_add_ns('BaseURL'))
1669 filesize = int_or_none(url_el.attrib.get('{http://youtube.com/yt/2012/10/10}contentLength') if url_el is not None else None)
1671 'format_id': '%s-%s' % (mpd_id, representation_id) if mpd_id else representation_id,
1673 'manifest_url': mpd_url,
1674 'ext': mimetype2ext(mime_type),
1675 'width': int_or_none(representation_attrib.get('width')),
1676 'height': int_or_none(representation_attrib.get('height')),
1677 'tbr': int_or_none(representation_attrib.get('bandwidth'), 1000),
1678 'asr': int_or_none(representation_attrib.get('audioSamplingRate')),
1679 'fps': int_or_none(representation_attrib.get('frameRate')),
1680 'vcodec': 'none' if content_type == 'audio' else representation_attrib.get('codecs'),
1681 'acodec': 'none' if content_type == 'video' else representation_attrib.get('codecs'),
1682 'language': lang if lang not in ('mul', 'und', 'zxx', 'mis') else None,
1683 'format_note': 'DASH %s' % content_type,
1684 'filesize': filesize,
1686 representation_ms_info = extract_multisegment_info(representation, adaption_set_ms_info)
1687 if 'segment_urls' not in representation_ms_info and 'media_template' in representation_ms_info:
1689 media_template = representation_ms_info['media_template']
1690 media_template = media_template.replace('$RepresentationID$', representation_id)
1691 media_template = re.sub(r'\$(Number|Bandwidth|Time)\$', r'%(\1)d', media_template)
1692 media_template = re.sub(r'\$(Number|Bandwidth|Time)%([^$]+)\$', r'%(\1)\2', media_template)
1693 media_template.replace('$$', '$')
1695 # As per [1, 5.3.9.4.4, Table 16, page 55] $Number$ and $Time$
1696 # can't be used at the same time
1697 if '%(Number' in media_template and 's' not in representation_ms_info:
1698 segment_duration = None
1699 if 'total_number' not in representation_ms_info and 'segment_duration':
1700 segment_duration = float_or_none(representation_ms_info['segment_duration'], representation_ms_info['timescale'])
1701 representation_ms_info['total_number'] = int(math.ceil(float(period_duration) / segment_duration))
1702 representation_ms_info['fragments'] = [{
1703 'url': media_template % {
1704 'Number': segment_number,
1705 'Bandwidth': representation_attrib.get('bandwidth'),
1707 'duration': segment_duration,
1708 } for segment_number in range(
1709 representation_ms_info['start_number'],
1710 representation_ms_info['total_number'] + representation_ms_info['start_number'])]
1712 # $Number*$ or $Time$ in media template with S list available
1713 # Example $Number*$: http://www.svtplay.se/klipp/9023742/stopptid-om-bjorn-borg
1714 # Example $Time$: https://play.arkena.com/embed/avp/v2/player/media/b41dda37-d8e7-4d3f-b1b5-9a9db578bdfe/1/129411
1715 representation_ms_info['fragments'] = []
1718 segment_number = representation_ms_info['start_number']
1720 def add_segment_url():
1721 segment_url = media_template % {
1722 'Time': segment_time,
1723 'Bandwidth': representation_attrib.get('bandwidth'),
1724 'Number': segment_number,
1726 representation_ms_info['fragments'].append({
1728 'duration': float_or_none(segment_d, representation_ms_info['timescale']),
1731 for num, s in enumerate(representation_ms_info['s']):
1732 segment_time = s.get('t') or segment_time
1736 for r in range(s.get('r', 0)):
1737 segment_time += segment_d
1740 segment_time += segment_d
1741 elif 'segment_urls' in representation_ms_info and 's' in representation_ms_info:
1743 # Example: https://www.youtube.com/watch?v=iXZV5uAYMJI
1744 # or any YouTube dashsegments video
1747 for segment_url in representation_ms_info['segment_urls']:
1748 s = representation_ms_info['s'][s_num]
1749 for r in range(s.get('r', 0) + 1):
1752 'duration': float_or_none(s['d'], representation_ms_info['timescale']),
1754 representation_ms_info['fragments'] = fragments
1755 # NB: MPD manifest may contain direct URLs to unfragmented media.
1756 # No fragments key is present in this case.
1757 if 'fragments' in representation_ms_info:
1760 'protocol': 'http_dash_segments',
1762 if 'initialization_url' in representation_ms_info:
1763 initialization_url = representation_ms_info['initialization_url'].replace('$RepresentationID$', representation_id)
1764 if not f.get('url'):
1765 f['url'] = initialization_url
1766 f['fragments'].append({'url': initialization_url})
1767 f['fragments'].extend(representation_ms_info['fragments'])
1768 for fragment in f['fragments']:
1769 fragment['url'] = combine_url(base_url, fragment['url'])
1771 existing_format = next(
1772 fo for fo in formats
1773 if fo['format_id'] == representation_id)
1774 except StopIteration:
1775 full_info = formats_dict.get(representation_id, {}).copy()
1777 formats.append(full_info)
1779 existing_format.update(f)
1781 self.report_warning('Unknown MIME type %s in DASH manifest' % mime_type)
1784 def _extract_ism_formats(self, ism_url, video_id, ism_id=None, note=None, errnote=None, fatal=True):
1785 res = self._download_webpage_handle(
1787 note=note or 'Downloading ISM manifest',
1788 errnote=errnote or 'Failed to download ISM manifest',
1794 return self._parse_ism_formats(
1795 compat_etree_fromstring(ism.encode('utf-8')), urlh.geturl(), ism_id)
1797 def _parse_ism_formats(self, ism_doc, ism_url, ism_id=None):
1798 if ism_doc.get('IsLive') == 'TRUE' or ism_doc.find('Protection') is not None:
1801 duration = int(ism_doc.attrib['Duration'])
1802 timescale = int_or_none(ism_doc.get('TimeScale')) or 10000000
1805 for stream in ism_doc.findall('StreamIndex'):
1806 stream_type = stream.get('Type')
1807 if stream_type not in ('video', 'audio'):
1809 url_pattern = stream.attrib['Url']
1810 stream_timescale = int_or_none(stream.get('TimeScale')) or timescale
1811 stream_name = stream.get('Name')
1812 for track in stream.findall('QualityLevel'):
1813 fourcc = track.get('FourCC')
1814 # TODO: add support for WVC1 and WMAP
1815 if fourcc not in ('H264', 'AVC1', 'AACL'):
1816 self.report_warning('%s is not a supported codec' % fourcc)
1818 tbr = int(track.attrib['Bitrate']) // 1000
1819 width = int_or_none(track.get('MaxWidth'))
1820 height = int_or_none(track.get('MaxHeight'))
1821 sampling_rate = int_or_none(track.get('SamplingRate'))
1823 track_url_pattern = re.sub(r'{[Bb]itrate}', track.attrib['Bitrate'], url_pattern)
1824 track_url_pattern = compat_urlparse.urljoin(ism_url, track_url_pattern)
1830 stream_fragments = stream.findall('c')
1831 for stream_fragment_index, stream_fragment in enumerate(stream_fragments):
1832 fragment_ctx['time'] = int_or_none(stream_fragment.get('t')) or fragment_ctx['time']
1833 fragment_repeat = int_or_none(stream_fragment.get('r')) or 1
1834 fragment_ctx['duration'] = int_or_none(stream_fragment.get('d'))
1835 if not fragment_ctx['duration']:
1837 next_fragment_time = int(stream_fragment[stream_fragment_index + 1].attrib['t'])
1839 next_fragment_time = duration
1840 fragment_ctx['duration'] = (next_fragment_time - fragment_ctx['time']) / fragment_repeat
1841 for _ in range(fragment_repeat):
1843 'url': re.sub(r'{start[ _]time}', compat_str(fragment_ctx['time']), track_url_pattern),
1844 'duration': fragment_ctx['duration'] / stream_timescale,
1846 fragment_ctx['time'] += fragment_ctx['duration']
1850 format_id.append(ism_id)
1852 format_id.append(stream_name)
1853 format_id.append(compat_str(tbr))
1856 'format_id': '-'.join(format_id),
1858 'manifest_url': ism_url,
1859 'ext': 'ismv' if stream_type == 'video' else 'isma',
1863 'asr': sampling_rate,
1864 'vcodec': 'none' if stream_type == 'audio' else fourcc,
1865 'acodec': 'none' if stream_type == 'video' else fourcc,
1867 'fragments': fragments,
1868 '_download_params': {
1869 'duration': duration,
1870 'timescale': stream_timescale,
1871 'width': width or 0,
1872 'height': height or 0,
1874 'codec_private_data': track.get('CodecPrivateData'),
1875 'sampling_rate': sampling_rate,
1876 'channels': int_or_none(track.get('Channels', 2)),
1877 'bits_per_sample': int_or_none(track.get('BitsPerSample', 16)),
1878 'nal_unit_length_field': int_or_none(track.get('NALUnitLengthField', 4)),
1883 def _parse_html5_media_entries(self, base_url, webpage, video_id, m3u8_id=None, m3u8_entry_protocol='m3u8'):
1884 def absolute_url(video_url):
1885 return compat_urlparse.urljoin(base_url, video_url)
1887 def parse_content_type(content_type):
1888 if not content_type:
1890 ctr = re.search(r'(?P<mimetype>[^/]+/[^;]+)(?:;\s*codecs="?(?P<codecs>[^"]+))?', content_type)
1892 mimetype, codecs = ctr.groups()
1893 f = parse_codecs(codecs)
1894 f['ext'] = mimetype2ext(mimetype)
1898 def _media_formats(src, cur_media_type):
1899 full_url = absolute_url(src)
1900 if determine_ext(full_url) == 'm3u8':
1901 is_plain_url = False
1902 formats = self._extract_m3u8_formats(
1903 full_url, video_id, ext='mp4',
1904 entry_protocol=m3u8_entry_protocol, m3u8_id=m3u8_id)
1909 'vcodec': 'none' if cur_media_type == 'audio' else None,
1911 return is_plain_url, formats
1914 media_tags = [(media_tag, media_type, '')
1915 for media_tag, media_type
1916 in re.findall(r'(?s)(<(video|audio)[^>]*/>)', webpage)]
1917 media_tags.extend(re.findall(r'(?s)(<(?P<tag>video|audio)[^>]*>)(.*?)</(?P=tag)>', webpage))
1918 for media_tag, media_type, media_content in media_tags:
1923 media_attributes = extract_attributes(media_tag)
1924 src = media_attributes.get('src')
1926 _, formats = _media_formats(src, media_type)
1927 media_info['formats'].extend(formats)
1928 media_info['thumbnail'] = media_attributes.get('poster')
1930 for source_tag in re.findall(r'<source[^>]+>', media_content):
1931 source_attributes = extract_attributes(source_tag)
1932 src = source_attributes.get('src')
1935 is_plain_url, formats = _media_formats(src, media_type)
1937 f = parse_content_type(source_attributes.get('type'))
1938 f.update(formats[0])
1939 media_info['formats'].append(f)
1941 media_info['formats'].extend(formats)
1942 for track_tag in re.findall(r'<track[^>]+>', media_content):
1943 track_attributes = extract_attributes(track_tag)
1944 kind = track_attributes.get('kind')
1945 if not kind or kind in ('subtitles', 'captions'):
1946 src = track_attributes.get('src')
1949 lang = track_attributes.get('srclang') or track_attributes.get('lang') or track_attributes.get('label')
1950 media_info['subtitles'].setdefault(lang, []).append({
1951 'url': absolute_url(src),
1953 if media_info['formats'] or media_info['subtitles']:
1954 entries.append(media_info)
1957 def _extract_akamai_formats(self, manifest_url, video_id):
1959 hdcore_sign = 'hdcore=3.7.0'
1960 f4m_url = re.sub(r'(https?://.+?)/i/', r'\1/z/', manifest_url).replace('/master.m3u8', '/manifest.f4m')
1961 if 'hdcore=' not in f4m_url:
1962 f4m_url += ('&' if '?' in f4m_url else '?') + hdcore_sign
1963 f4m_formats = self._extract_f4m_formats(
1964 f4m_url, video_id, f4m_id='hds', fatal=False)
1965 for entry in f4m_formats:
1966 entry.update({'extra_param_to_segment_url': hdcore_sign})
1967 formats.extend(f4m_formats)
1968 m3u8_url = re.sub(r'(https?://.+?)/z/', r'\1/i/', manifest_url).replace('/manifest.f4m', '/master.m3u8')
1969 formats.extend(self._extract_m3u8_formats(
1970 m3u8_url, video_id, 'mp4', 'm3u8_native',
1971 m3u8_id='hls', fatal=False))
1974 def _extract_wowza_formats(self, url, video_id, m3u8_entry_protocol='m3u8_native', skip_protocols=[]):
1975 url = re.sub(r'/(?:manifest|playlist|jwplayer)\.(?:m3u8|f4m|mpd|smil)', '', url)
1976 url_base = self._search_regex(r'(?:https?|rtmp|rtsp)(://[^?]+)', url, 'format url')
1977 http_base_url = 'http' + url_base
1979 if 'm3u8' not in skip_protocols:
1980 formats.extend(self._extract_m3u8_formats(
1981 http_base_url + '/playlist.m3u8', video_id, 'mp4',
1982 m3u8_entry_protocol, m3u8_id='hls', fatal=False))
1983 if 'f4m' not in skip_protocols:
1984 formats.extend(self._extract_f4m_formats(
1985 http_base_url + '/manifest.f4m',
1986 video_id, f4m_id='hds', fatal=False))
1987 if 'dash' not in skip_protocols:
1988 formats.extend(self._extract_mpd_formats(
1989 http_base_url + '/manifest.mpd',
1990 video_id, mpd_id='dash', fatal=False))
1991 if re.search(r'(?:/smil:|\.smil)', url_base):
1992 if 'smil' not in skip_protocols:
1993 rtmp_formats = self._extract_smil_formats(
1994 http_base_url + '/jwplayer.smil',
1995 video_id, fatal=False)
1996 for rtmp_format in rtmp_formats:
1997 rtsp_format = rtmp_format.copy()
1998 rtsp_format['url'] = '%s/%s' % (rtmp_format['url'], rtmp_format['play_path'])
1999 del rtsp_format['play_path']
2000 del rtsp_format['ext']
2001 rtsp_format.update({
2002 'url': rtsp_format['url'].replace('rtmp://', 'rtsp://'),
2003 'format_id': rtmp_format['format_id'].replace('rtmp', 'rtsp'),
2006 formats.extend([rtmp_format, rtsp_format])
2008 for protocol in ('rtmp', 'rtsp'):
2009 if protocol not in skip_protocols:
2011 'url': protocol + url_base,
2012 'format_id': protocol,
2013 'protocol': protocol,
2017 def _live_title(self, name):
2018 """ Generate the title for a live video """
2019 now = datetime.datetime.now()
2020 now_str = now.strftime('%Y-%m-%d %H:%M')
2021 return name + ' ' + now_str
2023 def _int(self, v, name, fatal=False, **kwargs):
2024 res = int_or_none(v, **kwargs)
2025 if 'get_attr' in kwargs:
2026 print(getattr(v, kwargs['get_attr']))
2028 msg = 'Failed to extract %s: Could not parse value %r' % (name, v)
2030 raise ExtractorError(msg)
2032 self._downloader.report_warning(msg)
2035 def _float(self, v, name, fatal=False, **kwargs):
2036 res = float_or_none(v, **kwargs)
2038 msg = 'Failed to extract %s: Could not parse value %r' % (name, v)
2040 raise ExtractorError(msg)
2042 self._downloader.report_warning(msg)
2045 def _set_cookie(self, domain, name, value, expire_time=None):
2046 cookie = compat_cookiejar.Cookie(
2047 0, name, value, None, None, domain, None,
2048 None, '/', True, False, expire_time, '', None, None, None)
2049 self._downloader.cookiejar.set_cookie(cookie)
2051 def _get_cookies(self, url):
2052 """ Return a compat_cookies.SimpleCookie with the cookies for the url """
2053 req = sanitized_Request(url)
2054 self._downloader.cookiejar.add_cookie_header(req)
2055 return compat_cookies.SimpleCookie(req.get_header('Cookie'))
2057 def get_testcases(self, include_onlymatching=False):
2058 t = getattr(self, '_TEST', None)
2060 assert not hasattr(self, '_TESTS'), \
2061 '%s has _TEST and _TESTS' % type(self).__name__
2064 tests = getattr(self, '_TESTS', [])
2066 if not include_onlymatching and t.get('only_matching', False):
2068 t['name'] = type(self).__name__[:-len('IE')]
2071 def is_suitable(self, age_limit):
2072 """ Test whether the extractor is generally suitable for the given
2073 age limit (i.e. pornographic sites are not, all others usually are) """
2075 any_restricted = False
2076 for tc in self.get_testcases(include_onlymatching=False):
2077 if tc.get('playlist', []):
2078 tc = tc['playlist'][0]
2079 is_restricted = age_restricted(
2080 tc.get('info_dict', {}).get('age_limit'), age_limit)
2081 if not is_restricted:
2083 any_restricted = any_restricted or is_restricted
2084 return not any_restricted
2086 def extract_subtitles(self, *args, **kwargs):
2087 if (self._downloader.params.get('writesubtitles', False) or
2088 self._downloader.params.get('listsubtitles')):
2089 return self._get_subtitles(*args, **kwargs)
2092 def _get_subtitles(self, *args, **kwargs):
2093 raise NotImplementedError('This method must be implemented by subclasses')
2096 def _merge_subtitle_items(subtitle_list1, subtitle_list2):
2097 """ Merge subtitle items for one language. Items with duplicated URLs
2098 will be dropped. """
2099 list1_urls = set([item['url'] for item in subtitle_list1])
2100 ret = list(subtitle_list1)
2101 ret.extend([item for item in subtitle_list2 if item['url'] not in list1_urls])
2105 def _merge_subtitles(cls, subtitle_dict1, subtitle_dict2):
2106 """ Merge two subtitle dictionaries, language by language. """
2107 ret = dict(subtitle_dict1)
2108 for lang in subtitle_dict2:
2109 ret[lang] = cls._merge_subtitle_items(subtitle_dict1.get(lang, []), subtitle_dict2[lang])
2112 def extract_automatic_captions(self, *args, **kwargs):
2113 if (self._downloader.params.get('writeautomaticsub', False) or
2114 self._downloader.params.get('listsubtitles')):
2115 return self._get_automatic_captions(*args, **kwargs)
2118 def _get_automatic_captions(self, *args, **kwargs):
2119 raise NotImplementedError('This method must be implemented by subclasses')
2121 def mark_watched(self, *args, **kwargs):
2122 if (self._downloader.params.get('mark_watched', False) and
2123 (self._get_login_info()[0] is not None or
2124 self._downloader.params.get('cookiefile') is not None)):
2125 self._mark_watched(*args, **kwargs)
2127 def _mark_watched(self, *args, **kwargs):
2128 raise NotImplementedError('This method must be implemented by subclasses')
2130 def geo_verification_headers(self):
2132 geo_verification_proxy = self._downloader.params.get('geo_verification_proxy')
2133 if geo_verification_proxy:
2134 headers['Ytdl-request-proxy'] = geo_verification_proxy
2137 def _generic_id(self, url):
2138 return compat_urllib_parse_unquote(os.path.splitext(url.rstrip('/').split('/')[-1])[0])
2140 def _generic_title(self, url):
2141 return compat_urllib_parse_unquote(os.path.splitext(url_basename(url))[0])
2144 class SearchInfoExtractor(InfoExtractor):
2146 Base class for paged search queries extractors.
2147 They accept URLs in the format _SEARCH_KEY(|all|[0-9]):{query}
2148 Instances should define _SEARCH_KEY and _MAX_RESULTS.
2152 def _make_valid_url(cls):
2153 return r'%s(?P<prefix>|[1-9][0-9]*|all):(?P<query>[\s\S]+)' % cls._SEARCH_KEY
2156 def suitable(cls, url):
2157 return re.match(cls._make_valid_url(), url) is not None
2159 def _real_extract(self, query):
2160 mobj = re.match(self._make_valid_url(), query)
2162 raise ExtractorError('Invalid search query "%s"' % query)
2164 prefix = mobj.group('prefix')
2165 query = mobj.group('query')
2167 return self._get_n_results(query, 1)
2168 elif prefix == 'all':
2169 return self._get_n_results(query, self._MAX_RESULTS)
2173 raise ExtractorError('invalid download number %s for query "%s"' % (n, query))
2174 elif n > self._MAX_RESULTS:
2175 self._downloader.report_warning('%s returns max %i results (you requested %i)' % (self._SEARCH_KEY, self._MAX_RESULTS, n))
2176 n = self._MAX_RESULTS
2177 return self._get_n_results(query, n)
2179 def _get_n_results(self, query, n):
2180 """Get a specified number of results for a query"""
2181 raise NotImplementedError('This method must be implemented by subclasses')
2184 def SEARCH_KEY(self):
2185 return self._SEARCH_KEY