1 from __future__ import unicode_literals
13 import xml.etree.ElementTree
15 from ..compat import (
20 compat_urllib_parse_urlparse,
41 class InfoExtractor(object):
42 """Information Extractor class.
44 Information extractors are the classes that, given a URL, extract
45 information about the video (or videos) the URL refers to. This
46 information includes the real video URL, the video title, author and
47 others. The information is stored in a dictionary which is then
48 passed to the YoutubeDL. The YoutubeDL processes this
49 information possibly downloading the video to the file system, among
50 other possible outcomes.
52 The type field determines the type of the result.
53 By far the most common value (and the default if _type is missing) is
54 "video", which indicates a single video.
56 For a video, the dictionaries must include the following fields:
59 title: Video title, unescaped.
61 Additionally, it must contain either a formats entry or a url one:
63 formats: A list of dictionaries for each format available, ordered
64 from worst to best quality.
67 * url Mandatory. The URL of the video file
68 * ext Will be calculated from URL if missing
69 * format A human-readable description of the format
70 ("mp4 container with h264/opus").
71 Calculated from the format_id, width, height.
72 and format_note fields if missing.
73 * format_id A short description of the format
74 ("mp4_h264_opus" or "19").
75 Technically optional, but strongly recommended.
76 * format_note Additional info about the format
77 ("3D" or "DASH video")
78 * width Width of the video, if known
79 * height Height of the video, if known
80 * resolution Textual description of width and height
81 * tbr Average bitrate of audio and video in KBit/s
82 * abr Average audio bitrate in KBit/s
83 * acodec Name of the audio codec in use
84 * asr Audio sampling rate in Hertz
85 * vbr Average video bitrate in KBit/s
87 * vcodec Name of the video codec in use
88 * container Name of the container format
89 * filesize The number of bytes, if known in advance
90 * filesize_approx An estimate for the number of bytes
91 * player_url SWF Player URL (used for rtmpdump).
92 * protocol The protocol that will be used for the actual
94 "http", "https", "rtsp", "rtmp", "rtmpe",
95 "m3u8", or "m3u8_native".
96 * preference Order number of this format. If this field is
97 present and not None, the formats get sorted
98 by this field, regardless of all other values.
99 -1 for default (order by other properties),
100 -2 or smaller for less than default.
101 < -1000 to hide the format (if there is
102 another one which is strictly better)
103 * language_preference Is this in the correct requested
105 10 if it's what the URL is about,
106 -1 for default (don't know),
107 -10 otherwise, other values reserved for now.
108 * quality Order number of the video quality of this
109 format, irrespective of the file format.
110 -1 for default (order by other properties),
111 -2 or smaller for less than default.
112 * source_preference Order number for this video source
113 (quality takes higher priority)
114 -1 for default (order by other properties),
115 -2 or smaller for less than default.
116 * http_headers A dictionary of additional HTTP headers
117 to add to the request.
118 * stretched_ratio If given and not 1, indicates that the
119 video's pixels are not square.
120 width : height ratio as float.
121 * no_resume The server does not support resuming the
122 (HTTP or RTMP) download. Boolean.
124 url: Final video URL.
125 ext: Video filename extension.
126 format: The video format, defaults to ext (used for --get-format)
127 player_url: SWF Player URL (used for rtmpdump).
129 The following fields are optional:
131 alt_title: A secondary title of the video.
132 display_id An alternative identifier for the video, not necessarily
133 unique, but available before title. Typically, id is
134 something like "4234987", title "Dancing naked mole rats",
135 and display_id "dancing-naked-mole-rats"
136 thumbnails: A list of dictionaries, with the following entries:
137 * "id" (optional, string) - Thumbnail format ID
139 * "preference" (optional, int) - quality of the image
140 * "width" (optional, int)
141 * "height" (optional, int)
142 * "resolution" (optional, string "{width}x{height"},
144 thumbnail: Full URL to a video thumbnail image.
145 description: Full video description.
146 uploader: Full name of the video uploader.
147 creator: The main artist who created the video.
148 timestamp: UNIX timestamp of the moment the video became available.
149 upload_date: Video upload date (YYYYMMDD).
150 If not explicitly set, calculated from timestamp.
151 uploader_id: Nickname or id of the video uploader.
152 location: Physical location where the video was filmed.
153 subtitles: The available subtitles as a dictionary in the format
154 {language: subformats}. "subformats" is a list sorted from
155 lower to higher preference, each element is a dictionary
156 with the "ext" entry and one of:
157 * "data": The subtitles file contents
158 * "url": A URL pointing to the subtitles file
159 automatic_captions: Like 'subtitles', used by the YoutubeIE for
160 automatically generated captions
161 duration: Length of the video in seconds, as an integer.
162 view_count: How many users have watched the video on the platform.
163 like_count: Number of positive ratings of the video
164 dislike_count: Number of negative ratings of the video
165 average_rating: Average rating give by users, the scale used depends on the webpage
166 comment_count: Number of comments on the video
167 comments: A list of comments, each with one or more of the following
168 properties (all but one of text or html optional):
169 * "author" - human-readable name of the comment author
170 * "author_id" - user ID of the comment author
172 * "html" - Comment as HTML
173 * "text" - Plain text of the comment
174 * "timestamp" - UNIX timestamp of comment
175 * "parent" - ID of the comment this one is replying to.
176 Set to "root" to indicate that this is a
177 comment to the original video.
178 age_limit: Age restriction for the video, as an integer (years)
179 webpage_url: The URL to the video webpage, if given to youtube-dl it
180 should allow to get the same result again. (It will be set
181 by YoutubeDL if it's missing)
182 categories: A list of categories that the video falls in, for example
184 tags: A list of tags assigned to the video, e.g. ["sweden", "pop music"]
185 is_live: True, False, or None (=unknown). Whether this video is a
186 live stream that goes on instead of a fixed-length video.
187 start_time: Time in seconds where the reproduction should start, as
188 specified in the URL.
189 end_time: Time in seconds where the reproduction should end, as
190 specified in the URL.
192 Unless mentioned otherwise, the fields should be Unicode strings.
194 Unless mentioned otherwise, None is equivalent to absence of information.
197 _type "playlist" indicates multiple videos.
198 There must be a key "entries", which is a list, an iterable, or a PagedList
199 object, each element of which is a valid dictionary by this specification.
201 Additionally, playlists can have "title" and "id" attributes with the same
202 semantics as videos (see above).
205 _type "multi_video" indicates that there are multiple videos that
206 form a single show, for examples multiple acts of an opera or TV episode.
207 It must have an entries key like a playlist and contain all the keys
208 required for a video at the same time.
211 _type "url" indicates that the video must be extracted from another
212 location, possibly by a different extractor. Its only required key is:
213 "url" - the next URL to extract.
214 The key "ie_key" can be set to the class name (minus the trailing "IE",
215 e.g. "Youtube") if the extractor class is known in advance.
216 Additionally, the dictionary may have any properties of the resolved entity
217 known in advance, for example "title" if the title of the referred video is
221 _type "url_transparent" entities have the same specification as "url", but
222 indicate that the given additional information is more precise than the one
223 associated with the resolved URL.
224 This is useful when a site employs a video service that hosts the video and
225 its technical metadata, but that video service does not embed a useful
226 title, description etc.
229 Subclasses of this one should re-define the _real_initialize() and
230 _real_extract() methods and define a _VALID_URL regexp.
231 Probably, they should also be added to the list of extractors.
233 Finally, the _WORKING attribute should be set to False for broken IEs
234 in order to warn the users and skip the tests.
241 def __init__(self, downloader=None):
242 """Constructor. Receives an optional downloader."""
244 self.set_downloader(downloader)
247 def suitable(cls, url):
248 """Receives a URL and returns True if suitable for this IE."""
250 # This does not use has/getattr intentionally - we want to know whether
251 # we have cached the regexp for *this* class, whereas getattr would also
252 # match the superclass
253 if '_VALID_URL_RE' not in cls.__dict__:
254 cls._VALID_URL_RE = re.compile(cls._VALID_URL)
255 return cls._VALID_URL_RE.match(url) is not None
258 def _match_id(cls, url):
259 if '_VALID_URL_RE' not in cls.__dict__:
260 cls._VALID_URL_RE = re.compile(cls._VALID_URL)
261 m = cls._VALID_URL_RE.match(url)
267 """Getter method for _WORKING."""
270 def initialize(self):
271 """Initializes an instance (authentication, etc)."""
273 self._real_initialize()
276 def extract(self, url):
277 """Extracts URL information and returns it in list of dicts."""
280 return self._real_extract(url)
281 except ExtractorError:
283 except compat_http_client.IncompleteRead as e:
284 raise ExtractorError('A network error has occured.', cause=e, expected=True)
285 except (KeyError, StopIteration) as e:
286 raise ExtractorError('An extractor error has occured.', cause=e)
288 def set_downloader(self, downloader):
289 """Sets the downloader for this IE."""
290 self._downloader = downloader
292 def _real_initialize(self):
293 """Real initialization process. Redefine in subclasses."""
296 def _real_extract(self, url):
297 """Real extraction process. Redefine in subclasses."""
302 """A string for getting the InfoExtractor with get_info_extractor"""
303 return cls.__name__[:-2]
307 return type(self).__name__[:-2]
309 def _request_webpage(self, url_or_request, video_id, note=None, errnote=None, fatal=True):
310 """ Returns the response handle """
312 self.report_download_webpage(video_id)
313 elif note is not False:
315 self.to_screen('%s' % (note,))
317 self.to_screen('%s: %s' % (video_id, note))
319 return self._downloader.urlopen(url_or_request)
320 except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
324 errnote = 'Unable to download webpage'
325 errmsg = '%s: %s' % (errnote, compat_str(err))
327 raise ExtractorError(errmsg, sys.exc_info()[2], cause=err)
329 self._downloader.report_warning(errmsg)
332 def _download_webpage_handle(self, url_or_request, video_id, note=None, errnote=None, fatal=True, encoding=None):
333 """ Returns a tuple (page content as string, URL handle) """
334 # Strip hashes from the URL (#1038)
335 if isinstance(url_or_request, (compat_str, str)):
336 url_or_request = url_or_request.partition('#')[0]
338 urlh = self._request_webpage(url_or_request, video_id, note, errnote, fatal)
342 content = self._webpage_read_content(urlh, url_or_request, video_id, note, errnote, fatal, encoding=encoding)
343 return (content, urlh)
346 def _guess_encoding_from_content(content_type, webpage_bytes):
347 m = re.match(r'[a-zA-Z0-9_.-]+/[a-zA-Z0-9_.-]+\s*;\s*charset=(.+)', content_type)
349 encoding = m.group(1)
351 m = re.search(br'<meta[^>]+charset=[\'"]?([^\'")]+)[ /\'">]',
352 webpage_bytes[:1024])
354 encoding = m.group(1).decode('ascii')
355 elif webpage_bytes.startswith(b'\xff\xfe'):
362 def _webpage_read_content(self, urlh, url_or_request, video_id, note=None, errnote=None, fatal=True, prefix=None, encoding=None):
363 content_type = urlh.headers.get('Content-Type', '')
364 webpage_bytes = urlh.read()
365 if prefix is not None:
366 webpage_bytes = prefix + webpage_bytes
368 encoding = self._guess_encoding_from_content(content_type, webpage_bytes)
369 if self._downloader.params.get('dump_intermediate_pages', False):
371 url = url_or_request.get_full_url()
372 except AttributeError:
374 self.to_screen('Dumping request to ' + url)
375 dump = base64.b64encode(webpage_bytes).decode('ascii')
376 self._downloader.to_screen(dump)
377 if self._downloader.params.get('write_pages', False):
379 url = url_or_request.get_full_url()
380 except AttributeError:
382 basen = '%s_%s' % (video_id, url)
384 h = '___' + hashlib.md5(basen.encode('utf-8')).hexdigest()
385 basen = basen[:240 - len(h)] + h
386 raw_filename = basen + '.dump'
387 filename = sanitize_filename(raw_filename, restricted=True)
388 self.to_screen('Saving request to ' + filename)
389 # Working around MAX_PATH limitation on Windows (see
390 # http://msdn.microsoft.com/en-us/library/windows/desktop/aa365247(v=vs.85).aspx)
392 absfilepath = os.path.abspath(filename)
393 if len(absfilepath) > 259:
394 filename = '\\\\?\\' + absfilepath
395 with open(filename, 'wb') as outf:
396 outf.write(webpage_bytes)
399 content = webpage_bytes.decode(encoding, 'replace')
401 content = webpage_bytes.decode('utf-8', 'replace')
403 if ('<title>Access to this site is blocked</title>' in content and
404 'Websense' in content[:512]):
405 msg = 'Access to this webpage has been blocked by Websense filtering software in your network.'
406 blocked_iframe = self._html_search_regex(
407 r'<iframe src="([^"]+)"', content,
408 'Websense information URL', default=None)
410 msg += ' Visit %s for more details' % blocked_iframe
411 raise ExtractorError(msg, expected=True)
412 if '<title>The URL you requested has been blocked</title>' in content[:512]:
414 'Access to this webpage has been blocked by Indian censorship. '
415 'Use a VPN or proxy server (with --proxy) to route around it.')
416 block_msg = self._html_search_regex(
417 r'</h1><p>(.*?)</p>',
418 content, 'block message', default=None)
420 msg += ' (Message: "%s")' % block_msg.replace('\n', ' ')
421 raise ExtractorError(msg, expected=True)
425 def _download_webpage(self, url_or_request, video_id, note=None, errnote=None, fatal=True, tries=1, timeout=5, encoding=None):
426 """ Returns the data of the page as a string """
429 while success is False:
431 res = self._download_webpage_handle(url_or_request, video_id, note, errnote, fatal, encoding=encoding)
433 except compat_http_client.IncompleteRead as e:
435 if try_count >= tries:
437 self._sleep(timeout, video_id)
444 def _download_xml(self, url_or_request, video_id,
445 note='Downloading XML', errnote='Unable to download XML',
446 transform_source=None, fatal=True, encoding=None):
447 """Return the xml as an xml.etree.ElementTree.Element"""
448 xml_string = self._download_webpage(
449 url_or_request, video_id, note, errnote, fatal=fatal, encoding=encoding)
450 if xml_string is False:
453 xml_string = transform_source(xml_string)
454 return xml.etree.ElementTree.fromstring(xml_string.encode('utf-8'))
456 def _download_json(self, url_or_request, video_id,
457 note='Downloading JSON metadata',
458 errnote='Unable to download JSON metadata',
459 transform_source=None,
460 fatal=True, encoding=None):
461 json_string = self._download_webpage(
462 url_or_request, video_id, note, errnote, fatal=fatal,
464 if (not fatal) and json_string is False:
466 return self._parse_json(
467 json_string, video_id, transform_source=transform_source, fatal=fatal)
469 def _parse_json(self, json_string, video_id, transform_source=None, fatal=True):
471 json_string = transform_source(json_string)
473 return json.loads(json_string)
474 except ValueError as ve:
475 errmsg = '%s: Failed to parse JSON ' % video_id
477 raise ExtractorError(errmsg, cause=ve)
479 self.report_warning(errmsg + str(ve))
481 def report_warning(self, msg, video_id=None):
482 idstr = '' if video_id is None else '%s: ' % video_id
483 self._downloader.report_warning(
484 '[%s] %s%s' % (self.IE_NAME, idstr, msg))
486 def to_screen(self, msg):
487 """Print msg to screen, prefixing it with '[ie_name]'"""
488 self._downloader.to_screen('[%s] %s' % (self.IE_NAME, msg))
490 def report_extraction(self, id_or_name):
491 """Report information extraction."""
492 self.to_screen('%s: Extracting information' % id_or_name)
494 def report_download_webpage(self, video_id):
495 """Report webpage download."""
496 self.to_screen('%s: Downloading webpage' % video_id)
498 def report_age_confirmation(self):
499 """Report attempt to confirm age."""
500 self.to_screen('Confirming age')
502 def report_login(self):
503 """Report attempt to log in."""
504 self.to_screen('Logging in')
506 # Methods for following #608
508 def url_result(url, ie=None, video_id=None, video_title=None):
509 """Returns a URL that points to a page that should be processed"""
510 # TODO: ie should be the class used for getting the info
511 video_info = {'_type': 'url',
514 if video_id is not None:
515 video_info['id'] = video_id
516 if video_title is not None:
517 video_info['title'] = video_title
521 def playlist_result(entries, playlist_id=None, playlist_title=None, playlist_description=None):
522 """Returns a playlist"""
523 video_info = {'_type': 'playlist',
526 video_info['id'] = playlist_id
528 video_info['title'] = playlist_title
529 if playlist_description:
530 video_info['description'] = playlist_description
533 def _search_regex(self, pattern, string, name, default=NO_DEFAULT, fatal=True, flags=0, group=None):
535 Perform a regex search on the given string, using a single or a list of
536 patterns returning the first matching group.
537 In case of failure return a default value or raise a WARNING or a
538 RegexNotFoundError, depending on fatal, specifying the field name.
540 if isinstance(pattern, (str, compat_str, compiled_regex_type)):
541 mobj = re.search(pattern, string, flags)
544 mobj = re.search(p, string, flags)
548 if not self._downloader.params.get('no_color') and os.name != 'nt' and sys.stderr.isatty():
549 _name = '\033[0;34m%s\033[0m' % name
555 # return the first matching group
556 return next(g for g in mobj.groups() if g is not None)
558 return mobj.group(group)
559 elif default is not NO_DEFAULT:
562 raise RegexNotFoundError('Unable to extract %s' % _name)
564 self._downloader.report_warning('unable to extract %s' % _name + bug_reports_message())
567 def _html_search_regex(self, pattern, string, name, default=NO_DEFAULT, fatal=True, flags=0, group=None):
569 Like _search_regex, but strips HTML tags and unescapes entities.
571 res = self._search_regex(pattern, string, name, default, fatal, flags, group)
573 return clean_html(res).strip()
577 def _get_login_info(self):
579 Get the login info as (username, password)
580 It will look in the netrc file using the _NETRC_MACHINE value
581 If there's no info available, return (None, None)
583 if self._downloader is None:
588 downloader_params = self._downloader.params
590 # Attempt to use provided username and password or .netrc data
591 if downloader_params.get('username', None) is not None:
592 username = downloader_params['username']
593 password = downloader_params['password']
594 elif downloader_params.get('usenetrc', False):
596 info = netrc.netrc().authenticators(self._NETRC_MACHINE)
601 raise netrc.NetrcParseError('No authenticators for %s' % self._NETRC_MACHINE)
602 except (IOError, netrc.NetrcParseError) as err:
603 self._downloader.report_warning('parsing .netrc: %s' % compat_str(err))
605 return (username, password)
607 def _get_tfa_info(self):
609 Get the two-factor authentication info
610 TODO - asking the user will be required for sms/phone verify
611 currently just uses the command line option
612 If there's no info available, return None
614 if self._downloader is None:
616 downloader_params = self._downloader.params
618 if downloader_params.get('twofactor', None) is not None:
619 return downloader_params['twofactor']
623 # Helper functions for extracting OpenGraph info
625 def _og_regexes(prop):
626 content_re = r'content=(?:"([^>]+?)"|\'([^>]+?)\')'
627 property_re = r'(?:name|property)=[\'"]og:%s[\'"]' % re.escape(prop)
628 template = r'<meta[^>]+?%s[^>]+?%s'
630 template % (property_re, content_re),
631 template % (content_re, property_re),
635 def _meta_regex(prop):
636 return r'''(?isx)<meta
637 (?=[^>]+(?:itemprop|name|property)=(["\']?)%s\1)
638 [^>]+?content=(["\'])(?P<content>.*?)\2''' % re.escape(prop)
640 def _og_search_property(self, prop, html, name=None, **kargs):
642 name = 'OpenGraph %s' % prop
643 escaped = self._search_regex(self._og_regexes(prop), html, name, flags=re.DOTALL, **kargs)
646 return unescapeHTML(escaped)
648 def _og_search_thumbnail(self, html, **kargs):
649 return self._og_search_property('image', html, 'thumbnail URL', fatal=False, **kargs)
651 def _og_search_description(self, html, **kargs):
652 return self._og_search_property('description', html, fatal=False, **kargs)
654 def _og_search_title(self, html, **kargs):
655 return self._og_search_property('title', html, **kargs)
657 def _og_search_video_url(self, html, name='video url', secure=True, **kargs):
658 regexes = self._og_regexes('video') + self._og_regexes('video:url')
660 regexes = self._og_regexes('video:secure_url') + regexes
661 return self._html_search_regex(regexes, html, name, **kargs)
663 def _og_search_url(self, html, **kargs):
664 return self._og_search_property('url', html, **kargs)
666 def _html_search_meta(self, name, html, display_name=None, fatal=False, **kwargs):
667 if display_name is None:
669 return self._html_search_regex(
670 self._meta_regex(name),
671 html, display_name, fatal=fatal, group='content', **kwargs)
673 def _dc_search_uploader(self, html):
674 return self._html_search_meta('dc.creator', html, 'uploader')
676 def _rta_search(self, html):
677 # See http://www.rtalabel.org/index.php?content=howtofaq#single
678 if re.search(r'(?ix)<meta\s+name="rating"\s+'
679 r' content="RTA-5042-1996-1400-1577-RTA"',
684 def _media_rating_search(self, html):
685 # See http://www.tjg-designs.com/WP/metadata-code-examples-adding-metadata-to-your-web-pages/
686 rating = self._html_search_meta('rating', html)
698 return RATING_TABLE.get(rating.lower(), None)
700 def _family_friendly_search(self, html):
701 # See http://schema.org/VideoObject
702 family_friendly = self._html_search_meta('isFamilyFriendly', html)
704 if not family_friendly:
713 return RATING_TABLE.get(family_friendly.lower(), None)
715 def _twitter_search_player(self, html):
716 return self._html_search_meta('twitter:player', html,
717 'twitter card player')
720 def _hidden_inputs(html):
722 (input.group('name'), input.group('value')) for input in re.finditer(
725 type=(?P<q_hidden>["\'])hidden(?P=q_hidden)\s+
726 name=(?P<q_name>["\'])(?P<name>.+?)(?P=q_name)\s+
727 (?:id=(?P<q_id>["\']).+?(?P=q_id)\s+)?
728 value=(?P<q_value>["\'])(?P<value>.*?)(?P=q_value)
732 def _form_hidden_inputs(self, form_id, html):
733 form = self._search_regex(
734 r'(?s)<form[^>]+?id=(["\'])%s\1[^>]*>(?P<form>.+?)</form>' % form_id,
735 html, '%s form' % form_id, group='form')
736 return self._hidden_inputs(form)
738 def _sort_formats(self, formats, field_preference=None):
740 raise ExtractorError('No video formats found')
743 # TODO remove the following workaround
744 from ..utils import determine_ext
745 if not f.get('ext') and 'url' in f:
746 f['ext'] = determine_ext(f['url'])
748 if isinstance(field_preference, (list, tuple)):
749 return tuple(f.get(field) if f.get(field) is not None else -1 for field in field_preference)
751 preference = f.get('preference')
752 if preference is None:
753 proto = f.get('protocol')
755 proto = compat_urllib_parse_urlparse(f.get('url', '')).scheme
757 preference = 0 if proto in ['http', 'https'] else -0.1
758 if f.get('ext') in ['f4f', 'f4m']: # Not yet supported
761 if f.get('vcodec') == 'none': # audio only
762 if self._downloader.params.get('prefer_free_formats'):
763 ORDER = ['aac', 'mp3', 'm4a', 'webm', 'ogg', 'opus']
765 ORDER = ['webm', 'opus', 'ogg', 'mp3', 'aac', 'm4a']
768 audio_ext_preference = ORDER.index(f['ext'])
770 audio_ext_preference = -1
772 if self._downloader.params.get('prefer_free_formats'):
773 ORDER = ['flv', 'mp4', 'webm']
775 ORDER = ['webm', 'flv', 'mp4']
777 ext_preference = ORDER.index(f['ext'])
780 audio_ext_preference = 0
784 f.get('language_preference') if f.get('language_preference') is not None else -1,
785 f.get('quality') if f.get('quality') is not None else -1,
786 f.get('tbr') if f.get('tbr') is not None else -1,
787 f.get('filesize') if f.get('filesize') is not None else -1,
788 f.get('vbr') if f.get('vbr') is not None else -1,
789 f.get('height') if f.get('height') is not None else -1,
790 f.get('width') if f.get('width') is not None else -1,
792 f.get('abr') if f.get('abr') is not None else -1,
793 audio_ext_preference,
794 f.get('fps') if f.get('fps') is not None else -1,
795 f.get('filesize_approx') if f.get('filesize_approx') is not None else -1,
796 f.get('source_preference') if f.get('source_preference') is not None else -1,
797 f.get('format_id') if f.get('format_id') is not None else '',
799 formats.sort(key=_formats_key)
801 def _check_formats(self, formats, video_id):
804 lambda f: self._is_valid_url(
806 item='%s video format' % f.get('format_id') if f.get('format_id') else 'video'),
809 def _is_valid_url(self, url, video_id, item='video'):
810 url = self._proto_relative_url(url, scheme='http:')
811 # For now assume non HTTP(S) URLs always valid
812 if not (url.startswith('http://') or url.startswith('https://')):
815 self._request_webpage(url, video_id, 'Checking %s URL' % item)
817 except ExtractorError as e:
818 if isinstance(e.cause, compat_HTTPError):
820 '%s: %s URL is invalid, skipping' % (video_id, item))
824 def http_scheme(self):
825 """ Either "http:" or "https:", depending on the user's preferences """
828 if self._downloader.params.get('prefer_insecure', False)
831 def _proto_relative_url(self, url, scheme=None):
834 if url.startswith('//'):
836 scheme = self.http_scheme()
841 def _sleep(self, timeout, video_id, msg_template=None):
842 if msg_template is None:
843 msg_template = '%(video_id)s: Waiting for %(timeout)s seconds'
844 msg = msg_template % {'video_id': video_id, 'timeout': timeout}
848 def _extract_f4m_formats(self, manifest_url, video_id, preference=None, f4m_id=None,
849 transform_source=lambda s: fix_xml_ampersands(s).strip()):
850 manifest = self._download_xml(
851 manifest_url, video_id, 'Downloading f4m manifest',
852 'Unable to download f4m manifest',
853 # Some manifests may be malformed, e.g. prosiebensat1 generated manifests
854 # (see https://github.com/rg3/youtube-dl/issues/6215#issuecomment-121704244)
855 transform_source=transform_source)
858 manifest_version = '1.0'
859 media_nodes = manifest.findall('{http://ns.adobe.com/f4m/1.0}media')
861 manifest_version = '2.0'
862 media_nodes = manifest.findall('{http://ns.adobe.com/f4m/2.0}media')
863 for i, media_el in enumerate(media_nodes):
864 if manifest_version == '2.0':
865 media_url = media_el.attrib.get('href') or media_el.attrib.get('url')
869 media_url if media_url.startswith('http://') or media_url.startswith('https://')
870 else ('/'.join(manifest_url.split('/')[:-1]) + '/' + media_url))
871 # If media_url is itself a f4m manifest do the recursive extraction
872 # since bitrates in parent manifest (this one) and media_url manifest
873 # may differ leading to inability to resolve the format by requested
874 # bitrate in f4m downloader
875 if determine_ext(manifest_url) == 'f4m':
876 formats.extend(self._extract_f4m_formats(manifest_url, video_id, preference, f4m_id))
878 tbr = int_or_none(media_el.attrib.get('bitrate'))
880 'format_id': '-'.join(filter(None, [f4m_id, compat_str(i if tbr is None else tbr)])),
884 'width': int_or_none(media_el.attrib.get('width')),
885 'height': int_or_none(media_el.attrib.get('height')),
886 'preference': preference,
888 self._sort_formats(formats)
892 def _extract_m3u8_formats(self, m3u8_url, video_id, ext=None,
893 entry_protocol='m3u8', preference=None,
894 m3u8_id=None, note=None, errnote=None,
898 'format_id': '-'.join(filter(None, [m3u8_id, 'meta'])),
902 'preference': preference - 1 if preference else -1,
903 'resolution': 'multiple',
904 'format_note': 'Quality selection URL',
907 format_url = lambda u: (
909 if re.match(r'^https?://', u)
910 else compat_urlparse.urljoin(m3u8_url, u))
912 m3u8_doc = self._download_webpage(
914 note=note or 'Downloading m3u8 information',
915 errnote=errnote or 'Failed to download m3u8 information',
917 if m3u8_doc is False:
922 r'(?P<key>[a-zA-Z_-]+)=(?P<val>"[^"]+"|[^",]+)(?:,|$)')
923 for line in m3u8_doc.splitlines():
924 if line.startswith('#EXT-X-STREAM-INF:'):
926 for m in kv_rex.finditer(line):
928 if v.startswith('"'):
930 last_info[m.group('key')] = v
931 elif line.startswith('#EXT-X-MEDIA:'):
933 for m in kv_rex.finditer(line):
935 if v.startswith('"'):
937 last_media[m.group('key')] = v
938 elif line.startswith('#') or not line.strip():
941 if last_info is None:
942 formats.append({'url': format_url(line)})
944 tbr = int_or_none(last_info.get('BANDWIDTH'), scale=1000)
947 format_id.append(m3u8_id)
948 last_media_name = last_media.get('NAME') if last_media and last_media.get('TYPE') != 'SUBTITLES' else None
949 format_id.append(last_media_name if last_media_name else '%d' % (tbr if tbr else len(formats)))
951 'format_id': '-'.join(format_id),
952 'url': format_url(line.strip()),
955 'protocol': entry_protocol,
956 'preference': preference,
958 codecs = last_info.get('CODECS')
960 # TODO: looks like video codec is not always necessarily goes first
961 va_codecs = codecs.split(',')
963 f['vcodec'] = va_codecs[0].partition('.')[0]
964 if len(va_codecs) > 1 and va_codecs[1]:
965 f['acodec'] = va_codecs[1].partition('.')[0]
966 resolution = last_info.get('RESOLUTION')
968 width_str, height_str = resolution.split('x')
969 f['width'] = int(width_str)
970 f['height'] = int(height_str)
971 if last_media is not None:
972 f['m3u8_media'] = last_media
976 self._sort_formats(formats)
979 # TODO: improve extraction
980 def _extract_smil_formats(self, smil_url, video_id, fatal=True):
981 smil = self._download_xml(
982 smil_url, video_id, 'Downloading SMIL file',
983 'Unable to download SMIL file', fatal=fatal)
988 base = smil.find('./head/meta').get('base')
992 if smil.findall('./body/seq/video'):
993 video = smil.findall('./body/seq/video')[0]
994 fmts, rtmp_count = self._parse_smil_video(video, video_id, base, rtmp_count)
997 for video in smil.findall('./body/switch/video'):
998 fmts, rtmp_count = self._parse_smil_video(video, video_id, base, rtmp_count)
1001 self._sort_formats(formats)
1005 def _parse_smil_video(self, video, video_id, base, rtmp_count):
1006 src = video.get('src')
1008 return [], rtmp_count
1009 bitrate = int_or_none(video.get('system-bitrate') or video.get('systemBitrate'), 1000)
1010 width = int_or_none(video.get('width'))
1011 height = int_or_none(video.get('height'))
1012 proto = video.get('proto')
1015 if base.startswith('rtmp'):
1017 elif base.startswith('http'):
1019 ext = video.get('ext')
1021 return self._extract_m3u8_formats(src, video_id, ext), rtmp_count
1022 elif proto == 'rtmp':
1024 streamer = video.get('streamer') or base
1029 'format_id': 'rtmp-%d' % (rtmp_count if bitrate is None else bitrate),
1034 elif proto.startswith('http'):
1037 'ext': ext or 'flv',
1043 def _live_title(self, name):
1044 """ Generate the title for a live video """
1045 now = datetime.datetime.now()
1046 now_str = now.strftime("%Y-%m-%d %H:%M")
1047 return name + ' ' + now_str
1049 def _int(self, v, name, fatal=False, **kwargs):
1050 res = int_or_none(v, **kwargs)
1051 if 'get_attr' in kwargs:
1052 print(getattr(v, kwargs['get_attr']))
1054 msg = 'Failed to extract %s: Could not parse value %r' % (name, v)
1056 raise ExtractorError(msg)
1058 self._downloader.report_warning(msg)
1061 def _float(self, v, name, fatal=False, **kwargs):
1062 res = float_or_none(v, **kwargs)
1064 msg = 'Failed to extract %s: Could not parse value %r' % (name, v)
1066 raise ExtractorError(msg)
1068 self._downloader.report_warning(msg)
1071 def _set_cookie(self, domain, name, value, expire_time=None):
1072 cookie = compat_cookiejar.Cookie(
1073 0, name, value, None, None, domain, None,
1074 None, '/', True, False, expire_time, '', None, None, None)
1075 self._downloader.cookiejar.set_cookie(cookie)
1077 def get_testcases(self, include_onlymatching=False):
1078 t = getattr(self, '_TEST', None)
1080 assert not hasattr(self, '_TESTS'), \
1081 '%s has _TEST and _TESTS' % type(self).__name__
1084 tests = getattr(self, '_TESTS', [])
1086 if not include_onlymatching and t.get('only_matching', False):
1088 t['name'] = type(self).__name__[:-len('IE')]
1091 def is_suitable(self, age_limit):
1092 """ Test whether the extractor is generally suitable for the given
1093 age limit (i.e. pornographic sites are not, all others usually are) """
1095 any_restricted = False
1096 for tc in self.get_testcases(include_onlymatching=False):
1097 if 'playlist' in tc:
1098 tc = tc['playlist'][0]
1099 is_restricted = age_restricted(
1100 tc.get('info_dict', {}).get('age_limit'), age_limit)
1101 if not is_restricted:
1103 any_restricted = any_restricted or is_restricted
1104 return not any_restricted
1106 def extract_subtitles(self, *args, **kwargs):
1107 if (self._downloader.params.get('writesubtitles', False) or
1108 self._downloader.params.get('listsubtitles')):
1109 return self._get_subtitles(*args, **kwargs)
1112 def _get_subtitles(self, *args, **kwargs):
1113 raise NotImplementedError("This method must be implemented by subclasses")
1115 def extract_automatic_captions(self, *args, **kwargs):
1116 if (self._downloader.params.get('writeautomaticsub', False) or
1117 self._downloader.params.get('listsubtitles')):
1118 return self._get_automatic_captions(*args, **kwargs)
1121 def _get_automatic_captions(self, *args, **kwargs):
1122 raise NotImplementedError("This method must be implemented by subclasses")
1125 class SearchInfoExtractor(InfoExtractor):
1127 Base class for paged search queries extractors.
1128 They accept URLs in the format _SEARCH_KEY(|all|[0-9]):{query}
1129 Instances should define _SEARCH_KEY and _MAX_RESULTS.
1133 def _make_valid_url(cls):
1134 return r'%s(?P<prefix>|[1-9][0-9]*|all):(?P<query>[\s\S]+)' % cls._SEARCH_KEY
1137 def suitable(cls, url):
1138 return re.match(cls._make_valid_url(), url) is not None
1140 def _real_extract(self, query):
1141 mobj = re.match(self._make_valid_url(), query)
1143 raise ExtractorError('Invalid search query "%s"' % query)
1145 prefix = mobj.group('prefix')
1146 query = mobj.group('query')
1148 return self._get_n_results(query, 1)
1149 elif prefix == 'all':
1150 return self._get_n_results(query, self._MAX_RESULTS)
1154 raise ExtractorError('invalid download number %s for query "%s"' % (n, query))
1155 elif n > self._MAX_RESULTS:
1156 self._downloader.report_warning('%s returns max %i results (you requested %i)' % (self._SEARCH_KEY, self._MAX_RESULTS, n))
1157 n = self._MAX_RESULTS
1158 return self._get_n_results(query, n)
1160 def _get_n_results(self, query, n):
1161 """Get a specified number of results for a query"""
1162 raise NotImplementedError("This method must be implemented by subclasses")
1165 def SEARCH_KEY(self):
1166 return self._SEARCH_KEY