1 from __future__ import unicode_literals
13 import xml.etree.ElementTree
15 from ..compat import (
20 compat_urllib_parse_urlparse,
39 class InfoExtractor(object):
40 """Information Extractor class.
42 Information extractors are the classes that, given a URL, extract
43 information about the video (or videos) the URL refers to. This
44 information includes the real video URL, the video title, author and
45 others. The information is stored in a dictionary which is then
46 passed to the YoutubeDL. The YoutubeDL processes this
47 information possibly downloading the video to the file system, among
48 other possible outcomes.
50 The type field determines the type of the result.
51 By far the most common value (and the default if _type is missing) is
52 "video", which indicates a single video.
54 For a video, the dictionaries must include the following fields:
57 title: Video title, unescaped.
59 Additionally, it must contain either a formats entry or a url one:
61 formats: A list of dictionaries for each format available, ordered
62 from worst to best quality.
65 * url Mandatory. The URL of the video file
66 * ext Will be calculated from url if missing
67 * format A human-readable description of the format
68 ("mp4 container with h264/opus").
69 Calculated from the format_id, width, height.
70 and format_note fields if missing.
71 * format_id A short description of the format
72 ("mp4_h264_opus" or "19").
73 Technically optional, but strongly recommended.
74 * format_note Additional info about the format
75 ("3D" or "DASH video")
76 * width Width of the video, if known
77 * height Height of the video, if known
78 * resolution Textual description of width and height
79 * tbr Average bitrate of audio and video in KBit/s
80 * abr Average audio bitrate in KBit/s
81 * acodec Name of the audio codec in use
82 * asr Audio sampling rate in Hertz
83 * vbr Average video bitrate in KBit/s
85 * vcodec Name of the video codec in use
86 * container Name of the container format
87 * filesize The number of bytes, if known in advance
88 * filesize_approx An estimate for the number of bytes
89 * player_url SWF Player URL (used for rtmpdump).
90 * protocol The protocol that will be used for the actual
92 "http", "https", "rtsp", "rtmp", "rtmpe",
93 "m3u8", or "m3u8_native".
94 * preference Order number of this format. If this field is
95 present and not None, the formats get sorted
96 by this field, regardless of all other values.
97 -1 for default (order by other properties),
98 -2 or smaller for less than default.
99 < -1000 to hide the format (if there is
100 another one which is strictly better)
101 * language_preference Is this in the correct requested
103 10 if it's what the URL is about,
104 -1 for default (don't know),
105 -10 otherwise, other values reserved for now.
106 * quality Order number of the video quality of this
107 format, irrespective of the file format.
108 -1 for default (order by other properties),
109 -2 or smaller for less than default.
110 * source_preference Order number for this video source
111 (quality takes higher priority)
112 -1 for default (order by other properties),
113 -2 or smaller for less than default.
114 * http_headers A dictionary of additional HTTP headers
115 to add to the request.
116 * stretched_ratio If given and not 1, indicates that the
117 video's pixels are not square.
118 width : height ratio as float.
119 * no_resume The server does not support resuming the
120 (HTTP or RTMP) download. Boolean.
122 url: Final video URL.
123 ext: Video filename extension.
124 format: The video format, defaults to ext (used for --get-format)
125 player_url: SWF Player URL (used for rtmpdump).
127 The following fields are optional:
129 alt_title: A secondary title of the video.
130 display_id An alternative identifier for the video, not necessarily
131 unique, but available before title. Typically, id is
132 something like "4234987", title "Dancing naked mole rats",
133 and display_id "dancing-naked-mole-rats"
134 thumbnails: A list of dictionaries, with the following entries:
135 * "id" (optional, string) - Thumbnail format ID
137 * "preference" (optional, int) - quality of the image
138 * "width" (optional, int)
139 * "height" (optional, int)
140 * "resolution" (optional, string "{width}x{height"},
142 thumbnail: Full URL to a video thumbnail image.
143 description: Full video description.
144 uploader: Full name of the video uploader.
145 creator: The main artist who created the video.
146 timestamp: UNIX timestamp of the moment the video became available.
147 upload_date: Video upload date (YYYYMMDD).
148 If not explicitly set, calculated from timestamp.
149 uploader_id: Nickname or id of the video uploader.
150 location: Physical location where the video was filmed.
151 subtitles: The available subtitles as a dictionary in the format
152 {language: subformats}. "subformats" is a list sorted from
153 lower to higher preference, each element is a dictionary
154 with the "ext" entry and one of:
155 * "data": The subtitles file contents
156 * "url": A url pointing to the subtitles file
157 automatic_captions: Like 'subtitles', used by the YoutubeIE for
158 automatically generated captions
159 duration: Length of the video in seconds, as an integer.
160 view_count: How many users have watched the video on the platform.
161 like_count: Number of positive ratings of the video
162 dislike_count: Number of negative ratings of the video
163 average_rating: Average rating give by users, the scale used depends on the webpage
164 comment_count: Number of comments on the video
165 comments: A list of comments, each with one or more of the following
166 properties (all but one of text or html optional):
167 * "author" - human-readable name of the comment author
168 * "author_id" - user ID of the comment author
170 * "html" - Comment as HTML
171 * "text" - Plain text of the comment
172 * "timestamp" - UNIX timestamp of comment
173 * "parent" - ID of the comment this one is replying to.
174 Set to "root" to indicate that this is a
175 comment to the original video.
176 age_limit: Age restriction for the video, as an integer (years)
177 webpage_url: The url to the video webpage, if given to youtube-dl it
178 should allow to get the same result again. (It will be set
179 by YoutubeDL if it's missing)
180 categories: A list of categories that the video falls in, for example
182 is_live: True, False, or None (=unknown). Whether this video is a
183 live stream that goes on instead of a fixed-length video.
185 Unless mentioned otherwise, the fields should be Unicode strings.
187 Unless mentioned otherwise, None is equivalent to absence of information.
190 _type "playlist" indicates multiple videos.
191 There must be a key "entries", which is a list, an iterable, or a PagedList
192 object, each element of which is a valid dictionary by this specification.
194 Additionally, playlists can have "title" and "id" attributes with the same
195 semantics as videos (see above).
198 _type "multi_video" indicates that there are multiple videos that
199 form a single show, for examples multiple acts of an opera or TV episode.
200 It must have an entries key like a playlist and contain all the keys
201 required for a video at the same time.
204 _type "url" indicates that the video must be extracted from another
205 location, possibly by a different extractor. Its only required key is:
206 "url" - the next URL to extract.
207 The key "ie_key" can be set to the class name (minus the trailing "IE",
208 e.g. "Youtube") if the extractor class is known in advance.
209 Additionally, the dictionary may have any properties of the resolved entity
210 known in advance, for example "title" if the title of the referred video is
214 _type "url_transparent" entities have the same specification as "url", but
215 indicate that the given additional information is more precise than the one
216 associated with the resolved URL.
217 This is useful when a site employs a video service that hosts the video and
218 its technical metadata, but that video service does not embed a useful
219 title, description etc.
222 Subclasses of this one should re-define the _real_initialize() and
223 _real_extract() methods and define a _VALID_URL regexp.
224 Probably, they should also be added to the list of extractors.
226 Finally, the _WORKING attribute should be set to False for broken IEs
227 in order to warn the users and skip the tests.
234 def __init__(self, downloader=None):
235 """Constructor. Receives an optional downloader."""
237 self.set_downloader(downloader)
240 def suitable(cls, url):
241 """Receives a URL and returns True if suitable for this IE."""
243 # This does not use has/getattr intentionally - we want to know whether
244 # we have cached the regexp for *this* class, whereas getattr would also
245 # match the superclass
246 if '_VALID_URL_RE' not in cls.__dict__:
247 cls._VALID_URL_RE = re.compile(cls._VALID_URL)
248 return cls._VALID_URL_RE.match(url) is not None
251 def _match_id(cls, url):
252 if '_VALID_URL_RE' not in cls.__dict__:
253 cls._VALID_URL_RE = re.compile(cls._VALID_URL)
254 m = cls._VALID_URL_RE.match(url)
260 """Getter method for _WORKING."""
263 def initialize(self):
264 """Initializes an instance (authentication, etc)."""
266 self._real_initialize()
269 def extract(self, url):
270 """Extracts URL information and returns it in list of dicts."""
273 return self._real_extract(url)
274 except ExtractorError:
276 except compat_http_client.IncompleteRead as e:
277 raise ExtractorError('A network error has occured.', cause=e, expected=True)
278 except (KeyError, StopIteration) as e:
279 raise ExtractorError('An extractor error has occured.', cause=e)
281 def set_downloader(self, downloader):
282 """Sets the downloader for this IE."""
283 self._downloader = downloader
285 def _real_initialize(self):
286 """Real initialization process. Redefine in subclasses."""
289 def _real_extract(self, url):
290 """Real extraction process. Redefine in subclasses."""
295 """A string for getting the InfoExtractor with get_info_extractor"""
296 return cls.__name__[:-2]
300 return type(self).__name__[:-2]
302 def _request_webpage(self, url_or_request, video_id, note=None, errnote=None, fatal=True):
303 """ Returns the response handle """
305 self.report_download_webpage(video_id)
306 elif note is not False:
308 self.to_screen('%s' % (note,))
310 self.to_screen('%s: %s' % (video_id, note))
312 return self._downloader.urlopen(url_or_request)
313 except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
317 errnote = 'Unable to download webpage'
318 errmsg = '%s: %s' % (errnote, compat_str(err))
320 raise ExtractorError(errmsg, sys.exc_info()[2], cause=err)
322 self._downloader.report_warning(errmsg)
325 def _download_webpage_handle(self, url_or_request, video_id, note=None, errnote=None, fatal=True, encoding=None):
326 """ Returns a tuple (page content as string, URL handle) """
327 # Strip hashes from the URL (#1038)
328 if isinstance(url_or_request, (compat_str, str)):
329 url_or_request = url_or_request.partition('#')[0]
331 urlh = self._request_webpage(url_or_request, video_id, note, errnote, fatal)
335 content = self._webpage_read_content(urlh, url_or_request, video_id, note, errnote, fatal, encoding=encoding)
336 return (content, urlh)
339 def _guess_encoding_from_content(content_type, webpage_bytes):
340 m = re.match(r'[a-zA-Z0-9_.-]+/[a-zA-Z0-9_.-]+\s*;\s*charset=(.+)', content_type)
342 encoding = m.group(1)
344 m = re.search(br'<meta[^>]+charset=[\'"]?([^\'")]+)[ /\'">]',
345 webpage_bytes[:1024])
347 encoding = m.group(1).decode('ascii')
348 elif webpage_bytes.startswith(b'\xff\xfe'):
355 def _webpage_read_content(self, urlh, url_or_request, video_id, note=None, errnote=None, fatal=True, prefix=None, encoding=None):
356 content_type = urlh.headers.get('Content-Type', '')
357 webpage_bytes = urlh.read()
358 if prefix is not None:
359 webpage_bytes = prefix + webpage_bytes
361 encoding = self._guess_encoding_from_content(content_type, webpage_bytes)
362 if self._downloader.params.get('dump_intermediate_pages', False):
364 url = url_or_request.get_full_url()
365 except AttributeError:
367 self.to_screen('Dumping request to ' + url)
368 dump = base64.b64encode(webpage_bytes).decode('ascii')
369 self._downloader.to_screen(dump)
370 if self._downloader.params.get('write_pages', False):
372 url = url_or_request.get_full_url()
373 except AttributeError:
375 basen = '%s_%s' % (video_id, url)
377 h = '___' + hashlib.md5(basen.encode('utf-8')).hexdigest()
378 basen = basen[:240 - len(h)] + h
379 raw_filename = basen + '.dump'
380 filename = sanitize_filename(raw_filename, restricted=True)
381 self.to_screen('Saving request to ' + filename)
382 # Working around MAX_PATH limitation on Windows (see
383 # http://msdn.microsoft.com/en-us/library/windows/desktop/aa365247(v=vs.85).aspx)
385 absfilepath = os.path.abspath(filename)
386 if len(absfilepath) > 259:
387 filename = '\\\\?\\' + absfilepath
388 with open(filename, 'wb') as outf:
389 outf.write(webpage_bytes)
392 content = webpage_bytes.decode(encoding, 'replace')
394 content = webpage_bytes.decode('utf-8', 'replace')
396 if ('<title>Access to this site is blocked</title>' in content and
397 'Websense' in content[:512]):
398 msg = 'Access to this webpage has been blocked by Websense filtering software in your network.'
399 blocked_iframe = self._html_search_regex(
400 r'<iframe src="([^"]+)"', content,
401 'Websense information URL', default=None)
403 msg += ' Visit %s for more details' % blocked_iframe
404 raise ExtractorError(msg, expected=True)
405 if '<title>The URL you requested has been blocked</title>' in content[:512]:
407 'Access to this webpage has been blocked by Indian censorship. '
408 'Use a VPN or proxy server (with --proxy) to route around it.')
409 block_msg = self._html_search_regex(
410 r'</h1><p>(.*?)</p>',
411 content, 'block message', default=None)
413 msg += ' (Message: "%s")' % block_msg.replace('\n', ' ')
414 raise ExtractorError(msg, expected=True)
418 def _download_webpage(self, url_or_request, video_id, note=None, errnote=None, fatal=True, tries=1, timeout=5, encoding=None):
419 """ Returns the data of the page as a string """
422 while success is False:
424 res = self._download_webpage_handle(url_or_request, video_id, note, errnote, fatal, encoding=encoding)
426 except compat_http_client.IncompleteRead as e:
428 if try_count >= tries:
430 self._sleep(timeout, video_id)
437 def _download_xml(self, url_or_request, video_id,
438 note='Downloading XML', errnote='Unable to download XML',
439 transform_source=None, fatal=True, encoding=None):
440 """Return the xml as an xml.etree.ElementTree.Element"""
441 xml_string = self._download_webpage(
442 url_or_request, video_id, note, errnote, fatal=fatal, encoding=encoding)
443 if xml_string is False:
446 xml_string = transform_source(xml_string)
447 return xml.etree.ElementTree.fromstring(xml_string.encode('utf-8'))
449 def _download_json(self, url_or_request, video_id,
450 note='Downloading JSON metadata',
451 errnote='Unable to download JSON metadata',
452 transform_source=None,
453 fatal=True, encoding=None):
454 json_string = self._download_webpage(
455 url_or_request, video_id, note, errnote, fatal=fatal,
457 if (not fatal) and json_string is False:
459 return self._parse_json(
460 json_string, video_id, transform_source=transform_source, fatal=fatal)
462 def _parse_json(self, json_string, video_id, transform_source=None, fatal=True):
464 json_string = transform_source(json_string)
466 return json.loads(json_string)
467 except ValueError as ve:
468 errmsg = '%s: Failed to parse JSON ' % video_id
470 raise ExtractorError(errmsg, cause=ve)
472 self.report_warning(errmsg + str(ve))
474 def report_warning(self, msg, video_id=None):
475 idstr = '' if video_id is None else '%s: ' % video_id
476 self._downloader.report_warning(
477 '[%s] %s%s' % (self.IE_NAME, idstr, msg))
479 def to_screen(self, msg):
480 """Print msg to screen, prefixing it with '[ie_name]'"""
481 self._downloader.to_screen('[%s] %s' % (self.IE_NAME, msg))
483 def report_extraction(self, id_or_name):
484 """Report information extraction."""
485 self.to_screen('%s: Extracting information' % id_or_name)
487 def report_download_webpage(self, video_id):
488 """Report webpage download."""
489 self.to_screen('%s: Downloading webpage' % video_id)
491 def report_age_confirmation(self):
492 """Report attempt to confirm age."""
493 self.to_screen('Confirming age')
495 def report_login(self):
496 """Report attempt to log in."""
497 self.to_screen('Logging in')
499 # Methods for following #608
501 def url_result(url, ie=None, video_id=None, video_title=None):
502 """Returns a url that points to a page that should be processed"""
503 # TODO: ie should be the class used for getting the info
504 video_info = {'_type': 'url',
507 if video_id is not None:
508 video_info['id'] = video_id
509 if video_title is not None:
510 video_info['title'] = video_title
514 def playlist_result(entries, playlist_id=None, playlist_title=None, playlist_description=None):
515 """Returns a playlist"""
516 video_info = {'_type': 'playlist',
519 video_info['id'] = playlist_id
521 video_info['title'] = playlist_title
522 if playlist_description:
523 video_info['description'] = playlist_description
526 def _search_regex(self, pattern, string, name, default=NO_DEFAULT, fatal=True, flags=0, group=None):
528 Perform a regex search on the given string, using a single or a list of
529 patterns returning the first matching group.
530 In case of failure return a default value or raise a WARNING or a
531 RegexNotFoundError, depending on fatal, specifying the field name.
533 if isinstance(pattern, (str, compat_str, compiled_regex_type)):
534 mobj = re.search(pattern, string, flags)
537 mobj = re.search(p, string, flags)
541 if not self._downloader.params.get('no_color') and os.name != 'nt' and sys.stderr.isatty():
542 _name = '\033[0;34m%s\033[0m' % name
548 # return the first matching group
549 return next(g for g in mobj.groups() if g is not None)
551 return mobj.group(group)
552 elif default is not NO_DEFAULT:
555 raise RegexNotFoundError('Unable to extract %s' % _name)
557 self._downloader.report_warning('unable to extract %s' % _name + bug_reports_message())
560 def _html_search_regex(self, pattern, string, name, default=NO_DEFAULT, fatal=True, flags=0, group=None):
562 Like _search_regex, but strips HTML tags and unescapes entities.
564 res = self._search_regex(pattern, string, name, default, fatal, flags, group)
566 return clean_html(res).strip()
570 def _get_login_info(self):
572 Get the login info as (username, password)
573 It will look in the netrc file using the _NETRC_MACHINE value
574 If there's no info available, return (None, None)
576 if self._downloader is None:
581 downloader_params = self._downloader.params
583 # Attempt to use provided username and password or .netrc data
584 if downloader_params.get('username', None) is not None:
585 username = downloader_params['username']
586 password = downloader_params['password']
587 elif downloader_params.get('usenetrc', False):
589 info = netrc.netrc().authenticators(self._NETRC_MACHINE)
594 raise netrc.NetrcParseError('No authenticators for %s' % self._NETRC_MACHINE)
595 except (IOError, netrc.NetrcParseError) as err:
596 self._downloader.report_warning('parsing .netrc: %s' % compat_str(err))
598 return (username, password)
600 def _get_tfa_info(self):
602 Get the two-factor authentication info
603 TODO - asking the user will be required for sms/phone verify
604 currently just uses the command line option
605 If there's no info available, return None
607 if self._downloader is None:
609 downloader_params = self._downloader.params
611 if downloader_params.get('twofactor', None) is not None:
612 return downloader_params['twofactor']
616 # Helper functions for extracting OpenGraph info
618 def _og_regexes(prop):
619 content_re = r'content=(?:"([^>]+?)"|\'([^>]+?)\')'
620 property_re = r'(?:name|property)=[\'"]og:%s[\'"]' % re.escape(prop)
621 template = r'<meta[^>]+?%s[^>]+?%s'
623 template % (property_re, content_re),
624 template % (content_re, property_re),
627 def _og_search_property(self, prop, html, name=None, **kargs):
629 name = 'OpenGraph %s' % prop
630 escaped = self._search_regex(self._og_regexes(prop), html, name, flags=re.DOTALL, **kargs)
633 return unescapeHTML(escaped)
635 def _og_search_thumbnail(self, html, **kargs):
636 return self._og_search_property('image', html, 'thumbnail url', fatal=False, **kargs)
638 def _og_search_description(self, html, **kargs):
639 return self._og_search_property('description', html, fatal=False, **kargs)
641 def _og_search_title(self, html, **kargs):
642 return self._og_search_property('title', html, **kargs)
644 def _og_search_video_url(self, html, name='video url', secure=True, **kargs):
645 regexes = self._og_regexes('video') + self._og_regexes('video:url')
647 regexes = self._og_regexes('video:secure_url') + regexes
648 return self._html_search_regex(regexes, html, name, **kargs)
650 def _og_search_url(self, html, **kargs):
651 return self._og_search_property('url', html, **kargs)
653 def _html_search_meta(self, name, html, display_name=None, fatal=False, **kwargs):
654 if display_name is None:
656 return self._html_search_regex(
658 (?=[^>]+(?:itemprop|name|property)=(["\']?)%s\1)
659 [^>]+?content=(["\'])(?P<content>.*?)\2''' % re.escape(name),
660 html, display_name, fatal=fatal, group='content', **kwargs)
662 def _dc_search_uploader(self, html):
663 return self._html_search_meta('dc.creator', html, 'uploader')
665 def _rta_search(self, html):
666 # See http://www.rtalabel.org/index.php?content=howtofaq#single
667 if re.search(r'(?ix)<meta\s+name="rating"\s+'
668 r' content="RTA-5042-1996-1400-1577-RTA"',
673 def _media_rating_search(self, html):
674 # See http://www.tjg-designs.com/WP/metadata-code-examples-adding-metadata-to-your-web-pages/
675 rating = self._html_search_meta('rating', html)
687 return RATING_TABLE.get(rating.lower(), None)
689 def _family_friendly_search(self, html):
690 # See http://schema.org/VideoObject
691 family_friendly = self._html_search_meta('isFamilyFriendly', html)
693 if not family_friendly:
702 return RATING_TABLE.get(family_friendly.lower(), None)
704 def _twitter_search_player(self, html):
705 return self._html_search_meta('twitter:player', html,
706 'twitter card player')
709 def _hidden_inputs(html):
711 (input.group('name'), input.group('value')) for input in re.finditer(
714 type=(?P<q_hidden>["\'])hidden(?P=q_hidden)\s+
715 name=(?P<q_name>["\'])(?P<name>.+?)(?P=q_name)\s+
716 (?:id=(?P<q_id>["\']).+?(?P=q_id)\s+)?
717 value=(?P<q_value>["\'])(?P<value>.*?)(?P=q_value)
721 def _form_hidden_inputs(self, form_id, html):
722 form = self._search_regex(
723 r'(?s)<form[^>]+?id=(["\'])%s\1[^>]*>(?P<form>.+?)</form>' % form_id,
724 html, '%s form' % form_id, group='form')
725 return self._hidden_inputs(form)
727 def _sort_formats(self, formats, field_preference=None):
729 raise ExtractorError('No video formats found')
732 # TODO remove the following workaround
733 from ..utils import determine_ext
734 if not f.get('ext') and 'url' in f:
735 f['ext'] = determine_ext(f['url'])
737 if isinstance(field_preference, (list, tuple)):
738 return tuple(f.get(field) if f.get(field) is not None else -1 for field in field_preference)
740 preference = f.get('preference')
741 if preference is None:
742 proto = f.get('protocol')
744 proto = compat_urllib_parse_urlparse(f.get('url', '')).scheme
746 preference = 0 if proto in ['http', 'https'] else -0.1
747 if f.get('ext') in ['f4f', 'f4m']: # Not yet supported
750 if f.get('vcodec') == 'none': # audio only
751 if self._downloader.params.get('prefer_free_formats'):
752 ORDER = ['aac', 'mp3', 'm4a', 'webm', 'ogg', 'opus']
754 ORDER = ['webm', 'opus', 'ogg', 'mp3', 'aac', 'm4a']
757 audio_ext_preference = ORDER.index(f['ext'])
759 audio_ext_preference = -1
761 if self._downloader.params.get('prefer_free_formats'):
762 ORDER = ['flv', 'mp4', 'webm']
764 ORDER = ['webm', 'flv', 'mp4']
766 ext_preference = ORDER.index(f['ext'])
769 audio_ext_preference = 0
773 f.get('language_preference') if f.get('language_preference') is not None else -1,
774 f.get('quality') if f.get('quality') is not None else -1,
775 f.get('tbr') if f.get('tbr') is not None else -1,
776 f.get('filesize') if f.get('filesize') is not None else -1,
777 f.get('vbr') if f.get('vbr') is not None else -1,
778 f.get('height') if f.get('height') is not None else -1,
779 f.get('width') if f.get('width') is not None else -1,
781 f.get('abr') if f.get('abr') is not None else -1,
782 audio_ext_preference,
783 f.get('fps') if f.get('fps') is not None else -1,
784 f.get('filesize_approx') if f.get('filesize_approx') is not None else -1,
785 f.get('source_preference') if f.get('source_preference') is not None else -1,
786 f.get('format_id') if f.get('format_id') is not None else '',
788 formats.sort(key=_formats_key)
790 def _check_formats(self, formats, video_id):
793 lambda f: self._is_valid_url(
795 item='%s video format' % f.get('format_id') if f.get('format_id') else 'video'),
798 def _is_valid_url(self, url, video_id, item='video'):
799 url = self._proto_relative_url(url, scheme='http:')
800 # For now assume non HTTP(S) URLs always valid
801 if not (url.startswith('http://') or url.startswith('https://')):
804 self._request_webpage(url, video_id, 'Checking %s URL' % item)
806 except ExtractorError as e:
807 if isinstance(e.cause, compat_HTTPError):
809 '%s: %s URL is invalid, skipping' % (video_id, item))
813 def http_scheme(self):
814 """ Either "http:" or "https:", depending on the user's preferences """
817 if self._downloader.params.get('prefer_insecure', False)
820 def _proto_relative_url(self, url, scheme=None):
823 if url.startswith('//'):
825 scheme = self.http_scheme()
830 def _sleep(self, timeout, video_id, msg_template=None):
831 if msg_template is None:
832 msg_template = '%(video_id)s: Waiting for %(timeout)s seconds'
833 msg = msg_template % {'video_id': video_id, 'timeout': timeout}
837 def _extract_f4m_formats(self, manifest_url, video_id, preference=None, f4m_id=None):
838 manifest = self._download_xml(
839 manifest_url, video_id, 'Downloading f4m manifest',
840 'Unable to download f4m manifest')
843 manifest_version = '1.0'
844 media_nodes = manifest.findall('{http://ns.adobe.com/f4m/1.0}media')
846 manifest_version = '2.0'
847 media_nodes = manifest.findall('{http://ns.adobe.com/f4m/2.0}media')
848 for i, media_el in enumerate(media_nodes):
849 if manifest_version == '2.0':
850 manifest_url = ('/'.join(manifest_url.split('/')[:-1]) + '/' +
851 (media_el.attrib.get('href') or media_el.attrib.get('url')))
852 tbr = int_or_none(media_el.attrib.get('bitrate'))
854 'format_id': '-'.join(filter(None, [f4m_id, compat_str(i if tbr is None else tbr)])),
858 'width': int_or_none(media_el.attrib.get('width')),
859 'height': int_or_none(media_el.attrib.get('height')),
860 'preference': preference,
862 self._sort_formats(formats)
866 def _extract_m3u8_formats(self, m3u8_url, video_id, ext=None,
867 entry_protocol='m3u8', preference=None,
868 m3u8_id=None, note=None, errnote=None,
872 'format_id': '-'.join(filter(None, [m3u8_id, 'meta'])),
876 'preference': preference - 1 if preference else -1,
877 'resolution': 'multiple',
878 'format_note': 'Quality selection URL',
881 format_url = lambda u: (
883 if re.match(r'^https?://', u)
884 else compat_urlparse.urljoin(m3u8_url, u))
886 m3u8_doc = self._download_webpage(
888 note=note or 'Downloading m3u8 information',
889 errnote=errnote or 'Failed to download m3u8 information',
891 if m3u8_doc is False:
896 r'(?P<key>[a-zA-Z_-]+)=(?P<val>"[^"]+"|[^",]+)(?:,|$)')
897 for line in m3u8_doc.splitlines():
898 if line.startswith('#EXT-X-STREAM-INF:'):
900 for m in kv_rex.finditer(line):
902 if v.startswith('"'):
904 last_info[m.group('key')] = v
905 elif line.startswith('#EXT-X-MEDIA:'):
907 for m in kv_rex.finditer(line):
909 if v.startswith('"'):
911 last_media[m.group('key')] = v
912 elif line.startswith('#') or not line.strip():
915 if last_info is None:
916 formats.append({'url': format_url(line)})
918 tbr = int_or_none(last_info.get('BANDWIDTH'), scale=1000)
921 format_id.append(m3u8_id)
922 last_media_name = last_media.get('NAME') if last_media and last_media.get('TYPE') != 'SUBTITLES' else None
923 format_id.append(last_media_name if last_media_name else '%d' % (tbr if tbr else len(formats)))
925 'format_id': '-'.join(format_id),
926 'url': format_url(line.strip()),
929 'protocol': entry_protocol,
930 'preference': preference,
932 codecs = last_info.get('CODECS')
934 # TODO: looks like video codec is not always necessarily goes first
935 va_codecs = codecs.split(',')
937 f['vcodec'] = va_codecs[0].partition('.')[0]
938 if len(va_codecs) > 1 and va_codecs[1]:
939 f['acodec'] = va_codecs[1].partition('.')[0]
940 resolution = last_info.get('RESOLUTION')
942 width_str, height_str = resolution.split('x')
943 f['width'] = int(width_str)
944 f['height'] = int(height_str)
945 if last_media is not None:
946 f['m3u8_media'] = last_media
950 self._sort_formats(formats)
953 # TODO: improve extraction
954 def _extract_smil_formats(self, smil_url, video_id, fatal=True):
955 smil = self._download_xml(
956 smil_url, video_id, 'Downloading SMIL file',
957 'Unable to download SMIL file', fatal=fatal)
962 base = smil.find('./head/meta').get('base')
966 if smil.findall('./body/seq/video'):
967 video = smil.findall('./body/seq/video')[0]
968 fmts, rtmp_count = self._parse_smil_video(video, video_id, base, rtmp_count)
971 for video in smil.findall('./body/switch/video'):
972 fmts, rtmp_count = self._parse_smil_video(video, video_id, base, rtmp_count)
975 self._sort_formats(formats)
979 def _parse_smil_video(self, video, video_id, base, rtmp_count):
980 src = video.get('src')
982 return ([], rtmp_count)
983 bitrate = int_or_none(video.get('system-bitrate') or video.get('systemBitrate'), 1000)
984 width = int_or_none(video.get('width'))
985 height = int_or_none(video.get('height'))
986 proto = video.get('proto')
989 if base.startswith('rtmp'):
991 elif base.startswith('http'):
993 ext = video.get('ext')
995 return (self._extract_m3u8_formats(src, video_id, ext), rtmp_count)
996 elif proto == 'rtmp':
998 streamer = video.get('streamer') or base
1003 'format_id': 'rtmp-%d' % (rtmp_count if bitrate is None else bitrate),
1008 elif proto.startswith('http'):
1011 'ext': ext or 'flv',
1017 def _live_title(self, name):
1018 """ Generate the title for a live video """
1019 now = datetime.datetime.now()
1020 now_str = now.strftime("%Y-%m-%d %H:%M")
1021 return name + ' ' + now_str
1023 def _int(self, v, name, fatal=False, **kwargs):
1024 res = int_or_none(v, **kwargs)
1025 if 'get_attr' in kwargs:
1026 print(getattr(v, kwargs['get_attr']))
1028 msg = 'Failed to extract %s: Could not parse value %r' % (name, v)
1030 raise ExtractorError(msg)
1032 self._downloader.report_warning(msg)
1035 def _float(self, v, name, fatal=False, **kwargs):
1036 res = float_or_none(v, **kwargs)
1038 msg = 'Failed to extract %s: Could not parse value %r' % (name, v)
1040 raise ExtractorError(msg)
1042 self._downloader.report_warning(msg)
1045 def _set_cookie(self, domain, name, value, expire_time=None):
1046 cookie = compat_cookiejar.Cookie(
1047 0, name, value, None, None, domain, None,
1048 None, '/', True, False, expire_time, '', None, None, None)
1049 self._downloader.cookiejar.set_cookie(cookie)
1051 def get_testcases(self, include_onlymatching=False):
1052 t = getattr(self, '_TEST', None)
1054 assert not hasattr(self, '_TESTS'), \
1055 '%s has _TEST and _TESTS' % type(self).__name__
1058 tests = getattr(self, '_TESTS', [])
1060 if not include_onlymatching and t.get('only_matching', False):
1062 t['name'] = type(self).__name__[:-len('IE')]
1065 def is_suitable(self, age_limit):
1066 """ Test whether the extractor is generally suitable for the given
1067 age limit (i.e. pornographic sites are not, all others usually are) """
1069 any_restricted = False
1070 for tc in self.get_testcases(include_onlymatching=False):
1071 if 'playlist' in tc:
1072 tc = tc['playlist'][0]
1073 is_restricted = age_restricted(
1074 tc.get('info_dict', {}).get('age_limit'), age_limit)
1075 if not is_restricted:
1077 any_restricted = any_restricted or is_restricted
1078 return not any_restricted
1080 def extract_subtitles(self, *args, **kwargs):
1081 if (self._downloader.params.get('writesubtitles', False) or
1082 self._downloader.params.get('listsubtitles')):
1083 return self._get_subtitles(*args, **kwargs)
1086 def _get_subtitles(self, *args, **kwargs):
1087 raise NotImplementedError("This method must be implemented by subclasses")
1089 def extract_automatic_captions(self, *args, **kwargs):
1090 if (self._downloader.params.get('writeautomaticsub', False) or
1091 self._downloader.params.get('listsubtitles')):
1092 return self._get_automatic_captions(*args, **kwargs)
1095 def _get_automatic_captions(self, *args, **kwargs):
1096 raise NotImplementedError("This method must be implemented by subclasses")
1099 class SearchInfoExtractor(InfoExtractor):
1101 Base class for paged search queries extractors.
1102 They accept urls in the format _SEARCH_KEY(|all|[0-9]):{query}
1103 Instances should define _SEARCH_KEY and _MAX_RESULTS.
1107 def _make_valid_url(cls):
1108 return r'%s(?P<prefix>|[1-9][0-9]*|all):(?P<query>[\s\S]+)' % cls._SEARCH_KEY
1111 def suitable(cls, url):
1112 return re.match(cls._make_valid_url(), url) is not None
1114 def _real_extract(self, query):
1115 mobj = re.match(self._make_valid_url(), query)
1117 raise ExtractorError('Invalid search query "%s"' % query)
1119 prefix = mobj.group('prefix')
1120 query = mobj.group('query')
1122 return self._get_n_results(query, 1)
1123 elif prefix == 'all':
1124 return self._get_n_results(query, self._MAX_RESULTS)
1128 raise ExtractorError('invalid download number %s for query "%s"' % (n, query))
1129 elif n > self._MAX_RESULTS:
1130 self._downloader.report_warning('%s returns max %i results (you requested %i)' % (self._SEARCH_KEY, self._MAX_RESULTS, n))
1131 n = self._MAX_RESULTS
1132 return self._get_n_results(query, n)
1134 def _get_n_results(self, query, n):
1135 """Get a specified number of results for a query"""
1136 raise NotImplementedError("This method must be implemented by subclasses")
1139 def SEARCH_KEY(self):
1140 return self._SEARCH_KEY