1 from __future__ import unicode_literals
13 import xml.etree.ElementTree
15 from ..compat import (
20 compat_urllib_parse_urlparse,
36 _NO_DEFAULT = object()
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 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_method HTTP method to use for the download.
115 * http_headers A dictionary of additional HTTP headers
116 to add to the request.
117 * http_post_data Additional data to send with a POST
119 * stretched_ratio If given and not 1, indicates that the
120 video's pixels are not square.
121 width : height ratio as float.
122 * no_resume The server does not support resuming the
123 (HTTP or RTMP) download. Boolean.
125 url: Final video URL.
126 ext: Video filename extension.
127 format: The video format, defaults to ext (used for --get-format)
128 player_url: SWF Player URL (used for rtmpdump).
130 The following fields are optional:
132 alt_title: A secondary title of the video.
133 display_id An alternative identifier for the video, not necessarily
134 unique, but available before title. Typically, id is
135 something like "4234987", title "Dancing naked mole rats",
136 and display_id "dancing-naked-mole-rats"
137 thumbnails: A list of dictionaries, with the following entries:
138 * "id" (optional, string) - Thumbnail format ID
140 * "preference" (optional, int) - quality of the image
141 * "width" (optional, int)
142 * "height" (optional, int)
143 * "resolution" (optional, string "{width}x{height"},
145 thumbnail: Full URL to a video thumbnail image.
146 description: Full video description.
147 uploader: Full name of the video uploader.
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 subtitle file contents as a dictionary in the format
154 {language: subtitles}.
155 duration: Length of the video in seconds, as an integer.
156 view_count: How many users have watched the video on the platform.
157 like_count: Number of positive ratings of the video
158 dislike_count: Number of negative ratings of the video
159 comment_count: Number of comments on the video
160 comments: A list of comments, each with one or more of the following
161 properties (all but one of text or html optional):
162 * "author" - human-readable name of the comment author
163 * "author_id" - user ID of the comment author
165 * "html" - Comment as HTML
166 * "text" - Plain text of the comment
167 * "timestamp" - UNIX timestamp of comment
168 * "parent" - ID of the comment this one is replying to.
169 Set to "root" to indicate that this is a
170 comment to the original video.
171 age_limit: Age restriction for the video, as an integer (years)
172 webpage_url: The url to the video webpage, if given to youtube-dl it
173 should allow to get the same result again. (It will be set
174 by YoutubeDL if it's missing)
175 categories: A list of categories that the video falls in, for example
177 is_live: True, False, or None (=unknown). Whether this video is a
178 live stream that goes on instead of a fixed-length video.
180 Unless mentioned otherwise, the fields should be Unicode strings.
182 Unless mentioned otherwise, None is equivalent to absence of information.
185 _type "playlist" indicates multiple videos.
186 There must be a key "entries", which is a list, an iterable, or a PagedList
187 object, each element of which is a valid dictionary by this specification.
189 Additionally, playlists can have "title" and "id" attributes with the same
190 semantics as videos (see above).
193 _type "multi_video" indicates that there are multiple videos that
194 form a single show, for examples multiple acts of an opera or TV episode.
195 It must have an entries key like a playlist and contain all the keys
196 required for a video at the same time.
199 _type "url" indicates that the video must be extracted from another
200 location, possibly by a different extractor. Its only required key is:
201 "url" - the next URL to extract.
202 The key "ie_key" can be set to the class name (minus the trailing "IE",
203 e.g. "Youtube") if the extractor class is known in advance.
204 Additionally, the dictionary may have any properties of the resolved entity
205 known in advance, for example "title" if the title of the referred video is
209 _type "url_transparent" entities have the same specification as "url", but
210 indicate that the given additional information is more precise than the one
211 associated with the resolved URL.
212 This is useful when a site employs a video service that hosts the video and
213 its technical metadata, but that video service does not embed a useful
214 title, description etc.
217 Subclasses of this one should re-define the _real_initialize() and
218 _real_extract() methods and define a _VALID_URL regexp.
219 Probably, they should also be added to the list of extractors.
221 Finally, the _WORKING attribute should be set to False for broken IEs
222 in order to warn the users and skip the tests.
229 def __init__(self, downloader=None):
230 """Constructor. Receives an optional downloader."""
232 self.set_downloader(downloader)
235 def suitable(cls, url):
236 """Receives a URL and returns True if suitable for this IE."""
238 # This does not use has/getattr intentionally - we want to know whether
239 # we have cached the regexp for *this* class, whereas getattr would also
240 # match the superclass
241 if '_VALID_URL_RE' not in cls.__dict__:
242 cls._VALID_URL_RE = re.compile(cls._VALID_URL)
243 return cls._VALID_URL_RE.match(url) is not None
246 def _match_id(cls, url):
247 if '_VALID_URL_RE' not in cls.__dict__:
248 cls._VALID_URL_RE = re.compile(cls._VALID_URL)
249 m = cls._VALID_URL_RE.match(url)
255 """Getter method for _WORKING."""
258 def initialize(self):
259 """Initializes an instance (authentication, etc)."""
261 self._real_initialize()
264 def extract(self, url):
265 """Extracts URL information and returns it in list of dicts."""
267 return self._real_extract(url)
269 def set_downloader(self, downloader):
270 """Sets the downloader for this IE."""
271 self._downloader = downloader
273 def _real_initialize(self):
274 """Real initialization process. Redefine in subclasses."""
277 def _real_extract(self, url):
278 """Real extraction process. Redefine in subclasses."""
283 """A string for getting the InfoExtractor with get_info_extractor"""
284 return cls.__name__[:-2]
288 return type(self).__name__[:-2]
290 def _request_webpage(self, url_or_request, video_id, note=None, errnote=None, fatal=True):
291 """ Returns the response handle """
293 self.report_download_webpage(video_id)
294 elif note is not False:
296 self.to_screen('%s' % (note,))
298 self.to_screen('%s: %s' % (video_id, note))
300 return self._downloader.urlopen(url_or_request)
301 except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
305 errnote = 'Unable to download webpage'
306 errmsg = '%s: %s' % (errnote, compat_str(err))
308 raise ExtractorError(errmsg, sys.exc_info()[2], cause=err)
310 self._downloader.report_warning(errmsg)
313 def _download_webpage_handle(self, url_or_request, video_id, note=None, errnote=None, fatal=True):
314 """ Returns a tuple (page content as string, URL handle) """
315 # Strip hashes from the URL (#1038)
316 if isinstance(url_or_request, (compat_str, str)):
317 url_or_request = url_or_request.partition('#')[0]
319 urlh = self._request_webpage(url_or_request, video_id, note, errnote, fatal)
323 content = self._webpage_read_content(urlh, url_or_request, video_id, note, errnote, fatal)
324 return (content, urlh)
326 def _webpage_read_content(self, urlh, url_or_request, video_id, note=None, errnote=None, fatal=True, prefix=None):
327 content_type = urlh.headers.get('Content-Type', '')
328 webpage_bytes = urlh.read()
329 if prefix is not None:
330 webpage_bytes = prefix + webpage_bytes
331 m = re.match(r'[a-zA-Z0-9_.-]+/[a-zA-Z0-9_.-]+\s*;\s*charset=(.+)', content_type)
333 encoding = m.group(1)
335 m = re.search(br'<meta[^>]+charset=[\'"]?([^\'")]+)[ /\'">]',
336 webpage_bytes[:1024])
338 encoding = m.group(1).decode('ascii')
339 elif webpage_bytes.startswith(b'\xff\xfe'):
343 if self._downloader.params.get('dump_intermediate_pages', False):
345 url = url_or_request.get_full_url()
346 except AttributeError:
348 self.to_screen('Dumping request to ' + url)
349 dump = base64.b64encode(webpage_bytes).decode('ascii')
350 self._downloader.to_screen(dump)
351 if self._downloader.params.get('write_pages', False):
353 url = url_or_request.get_full_url()
354 except AttributeError:
356 basen = '%s_%s' % (video_id, url)
358 h = '___' + hashlib.md5(basen.encode('utf-8')).hexdigest()
359 basen = basen[:240 - len(h)] + h
360 raw_filename = basen + '.dump'
361 filename = sanitize_filename(raw_filename, restricted=True)
362 self.to_screen('Saving request to ' + filename)
363 # Working around MAX_PATH limitation on Windows (see
364 # http://msdn.microsoft.com/en-us/library/windows/desktop/aa365247(v=vs.85).aspx)
366 absfilepath = os.path.abspath(filename)
367 if len(absfilepath) > 259:
368 filename = '\\\\?\\' + absfilepath
369 with open(filename, 'wb') as outf:
370 outf.write(webpage_bytes)
373 content = webpage_bytes.decode(encoding, 'replace')
375 content = webpage_bytes.decode('utf-8', 'replace')
377 if ('<title>Access to this site is blocked</title>' in content and
378 'Websense' in content[:512]):
379 msg = 'Access to this webpage has been blocked by Websense filtering software in your network.'
380 blocked_iframe = self._html_search_regex(
381 r'<iframe src="([^"]+)"', content,
382 'Websense information URL', default=None)
384 msg += ' Visit %s for more details' % blocked_iframe
385 raise ExtractorError(msg, expected=True)
389 def _download_webpage(self, url_or_request, video_id, note=None, errnote=None, fatal=True, tries=1, timeout=5):
390 """ Returns the data of the page as a string """
393 while success is False:
395 res = self._download_webpage_handle(url_or_request, video_id, note, errnote, fatal)
397 except compat_http_client.IncompleteRead as e:
399 if try_count >= tries:
401 self._sleep(timeout, video_id)
408 def _download_xml(self, url_or_request, video_id,
409 note='Downloading XML', errnote='Unable to download XML',
410 transform_source=None, fatal=True):
411 """Return the xml as an xml.etree.ElementTree.Element"""
412 xml_string = self._download_webpage(
413 url_or_request, video_id, note, errnote, fatal=fatal)
414 if xml_string is False:
417 xml_string = transform_source(xml_string)
418 return xml.etree.ElementTree.fromstring(xml_string.encode('utf-8'))
420 def _download_json(self, url_or_request, video_id,
421 note='Downloading JSON metadata',
422 errnote='Unable to download JSON metadata',
423 transform_source=None,
425 json_string = self._download_webpage(
426 url_or_request, video_id, note, errnote, fatal=fatal)
427 if (not fatal) and json_string is False:
429 return self._parse_json(
430 json_string, video_id, transform_source=transform_source, fatal=fatal)
432 def _parse_json(self, json_string, video_id, transform_source=None, fatal=True):
434 json_string = transform_source(json_string)
436 return json.loads(json_string)
437 except ValueError as ve:
438 errmsg = '%s: Failed to parse JSON ' % video_id
440 raise ExtractorError(errmsg, cause=ve)
442 self.report_warning(errmsg + str(ve))
444 def report_warning(self, msg, video_id=None):
445 idstr = '' if video_id is None else '%s: ' % video_id
446 self._downloader.report_warning(
447 '[%s] %s%s' % (self.IE_NAME, idstr, msg))
449 def to_screen(self, msg):
450 """Print msg to screen, prefixing it with '[ie_name]'"""
451 self._downloader.to_screen('[%s] %s' % (self.IE_NAME, msg))
453 def report_extraction(self, id_or_name):
454 """Report information extraction."""
455 self.to_screen('%s: Extracting information' % id_or_name)
457 def report_download_webpage(self, video_id):
458 """Report webpage download."""
459 self.to_screen('%s: Downloading webpage' % video_id)
461 def report_age_confirmation(self):
462 """Report attempt to confirm age."""
463 self.to_screen('Confirming age')
465 def report_login(self):
466 """Report attempt to log in."""
467 self.to_screen('Logging in')
469 # Methods for following #608
471 def url_result(url, ie=None, video_id=None):
472 """Returns a url that points to a page that should be processed"""
473 # TODO: ie should be the class used for getting the info
474 video_info = {'_type': 'url',
477 if video_id is not None:
478 video_info['id'] = video_id
482 def playlist_result(entries, playlist_id=None, playlist_title=None, playlist_description=None):
483 """Returns a playlist"""
484 video_info = {'_type': 'playlist',
487 video_info['id'] = playlist_id
489 video_info['title'] = playlist_title
490 if playlist_description:
491 video_info['description'] = playlist_description
494 def _search_regex(self, pattern, string, name, default=_NO_DEFAULT, fatal=True, flags=0, group=None):
496 Perform a regex search on the given string, using a single or a list of
497 patterns returning the first matching group.
498 In case of failure return a default value or raise a WARNING or a
499 RegexNotFoundError, depending on fatal, specifying the field name.
501 if isinstance(pattern, (str, compat_str, compiled_regex_type)):
502 mobj = re.search(pattern, string, flags)
505 mobj = re.search(p, string, flags)
509 if os.name != 'nt' and sys.stderr.isatty():
510 _name = '\033[0;34m%s\033[0m' % name
516 # return the first matching group
517 return next(g for g in mobj.groups() if g is not None)
519 return mobj.group(group)
520 elif default is not _NO_DEFAULT:
523 raise RegexNotFoundError('Unable to extract %s' % _name)
525 self._downloader.report_warning('unable to extract %s; '
526 'please report this issue on http://yt-dl.org/bug' % _name)
529 def _html_search_regex(self, pattern, string, name, default=_NO_DEFAULT, fatal=True, flags=0, group=None):
531 Like _search_regex, but strips HTML tags and unescapes entities.
533 res = self._search_regex(pattern, string, name, default, fatal, flags, group)
535 return clean_html(res).strip()
539 def _get_login_info(self):
541 Get the the login info as (username, password)
542 It will look in the netrc file using the _NETRC_MACHINE value
543 If there's no info available, return (None, None)
545 if self._downloader is None:
550 downloader_params = self._downloader.params
552 # Attempt to use provided username and password or .netrc data
553 if downloader_params.get('username', None) is not None:
554 username = downloader_params['username']
555 password = downloader_params['password']
556 elif downloader_params.get('usenetrc', False):
558 info = netrc.netrc().authenticators(self._NETRC_MACHINE)
563 raise netrc.NetrcParseError('No authenticators for %s' % self._NETRC_MACHINE)
564 except (IOError, netrc.NetrcParseError) as err:
565 self._downloader.report_warning('parsing .netrc: %s' % compat_str(err))
567 return (username, password)
569 def _get_tfa_info(self):
571 Get the two-factor authentication info
572 TODO - asking the user will be required for sms/phone verify
573 currently just uses the command line option
574 If there's no info available, return None
576 if self._downloader is None:
578 downloader_params = self._downloader.params
580 if downloader_params.get('twofactor', None) is not None:
581 return downloader_params['twofactor']
585 # Helper functions for extracting OpenGraph info
587 def _og_regexes(prop):
588 content_re = r'content=(?:"([^>]+?)"|\'([^>]+?)\')'
589 property_re = r'(?:name|property)=[\'"]og:%s[\'"]' % re.escape(prop)
590 template = r'<meta[^>]+?%s[^>]+?%s'
592 template % (property_re, content_re),
593 template % (content_re, property_re),
596 def _og_search_property(self, prop, html, name=None, **kargs):
598 name = 'OpenGraph %s' % prop
599 escaped = self._search_regex(self._og_regexes(prop), html, name, flags=re.DOTALL, **kargs)
602 return unescapeHTML(escaped)
604 def _og_search_thumbnail(self, html, **kargs):
605 return self._og_search_property('image', html, 'thumbnail url', fatal=False, **kargs)
607 def _og_search_description(self, html, **kargs):
608 return self._og_search_property('description', html, fatal=False, **kargs)
610 def _og_search_title(self, html, **kargs):
611 return self._og_search_property('title', html, **kargs)
613 def _og_search_video_url(self, html, name='video url', secure=True, **kargs):
614 regexes = self._og_regexes('video') + self._og_regexes('video:url')
616 regexes = self._og_regexes('video:secure_url') + regexes
617 return self._html_search_regex(regexes, html, name, **kargs)
619 def _og_search_url(self, html, **kargs):
620 return self._og_search_property('url', html, **kargs)
622 def _html_search_meta(self, name, html, display_name=None, fatal=False, **kwargs):
623 if display_name is None:
625 return self._html_search_regex(
627 (?=[^>]+(?:itemprop|name|property)=(["\']?)%s\1)
628 [^>]+?content=(["\'])(?P<content>.*?)\2''' % re.escape(name),
629 html, display_name, fatal=fatal, group='content', **kwargs)
631 def _dc_search_uploader(self, html):
632 return self._html_search_meta('dc.creator', html, 'uploader')
634 def _rta_search(self, html):
635 # See http://www.rtalabel.org/index.php?content=howtofaq#single
636 if re.search(r'(?ix)<meta\s+name="rating"\s+'
637 r' content="RTA-5042-1996-1400-1577-RTA"',
642 def _media_rating_search(self, html):
643 # See http://www.tjg-designs.com/WP/metadata-code-examples-adding-metadata-to-your-web-pages/
644 rating = self._html_search_meta('rating', html)
656 return RATING_TABLE.get(rating.lower(), None)
658 def _twitter_search_player(self, html):
659 return self._html_search_meta('twitter:player', html,
660 'twitter card player')
662 def _sort_formats(self, formats):
664 raise ExtractorError('No video formats found')
667 # TODO remove the following workaround
668 from ..utils import determine_ext
669 if not f.get('ext') and 'url' in f:
670 f['ext'] = determine_ext(f['url'])
672 preference = f.get('preference')
673 if preference is None:
674 proto = f.get('protocol')
676 proto = compat_urllib_parse_urlparse(f.get('url', '')).scheme
678 preference = 0 if proto in ['http', 'https'] else -0.1
679 if f.get('ext') in ['f4f', 'f4m']: # Not yet supported
682 if f.get('vcodec') == 'none': # audio only
683 if self._downloader.params.get('prefer_free_formats'):
684 ORDER = ['aac', 'mp3', 'm4a', 'webm', 'ogg', 'opus']
686 ORDER = ['webm', 'opus', 'ogg', 'mp3', 'aac', 'm4a']
689 audio_ext_preference = ORDER.index(f['ext'])
691 audio_ext_preference = -1
693 if self._downloader.params.get('prefer_free_formats'):
694 ORDER = ['flv', 'mp4', 'webm']
696 ORDER = ['webm', 'flv', 'mp4']
698 ext_preference = ORDER.index(f['ext'])
701 audio_ext_preference = 0
705 f.get('language_preference') if f.get('language_preference') is not None else -1,
706 f.get('quality') if f.get('quality') is not None else -1,
707 f.get('height') if f.get('height') is not None else -1,
708 f.get('width') if f.get('width') is not None else -1,
710 f.get('tbr') if f.get('tbr') is not None else -1,
711 f.get('vbr') if f.get('vbr') is not None else -1,
712 f.get('abr') if f.get('abr') is not None else -1,
713 audio_ext_preference,
714 f.get('fps') if f.get('fps') is not None else -1,
715 f.get('filesize') if f.get('filesize') is not None else -1,
716 f.get('filesize_approx') if f.get('filesize_approx') is not None else -1,
717 f.get('source_preference') if f.get('source_preference') is not None else -1,
720 formats.sort(key=_formats_key)
722 def _check_formats(self, formats, video_id):
725 lambda f: self._is_valid_url(
727 item='%s video format' % f.get('format_id') if f.get('format_id') else 'video'),
730 def _is_valid_url(self, url, video_id, item='video'):
732 self._request_webpage(
733 HEADRequest(url), video_id,
734 'Checking %s URL' % item)
736 except ExtractorError as e:
737 if isinstance(e.cause, compat_HTTPError):
739 '%s URL is invalid, skipping' % item, video_id)
743 def http_scheme(self):
744 """ Either "http:" or "https:", depending on the user's preferences """
747 if self._downloader.params.get('prefer_insecure', False)
750 def _proto_relative_url(self, url, scheme=None):
753 if url.startswith('//'):
755 scheme = self.http_scheme()
760 def _sleep(self, timeout, video_id, msg_template=None):
761 if msg_template is None:
762 msg_template = '%(video_id)s: Waiting for %(timeout)s seconds'
763 msg = msg_template % {'video_id': video_id, 'timeout': timeout}
767 def _extract_f4m_formats(self, manifest_url, video_id):
768 manifest = self._download_xml(
769 manifest_url, video_id, 'Downloading f4m manifest',
770 'Unable to download f4m manifest')
773 manifest_version = '1.0'
774 media_nodes = manifest.findall('{http://ns.adobe.com/f4m/1.0}media')
776 manifest_version = '2.0'
777 media_nodes = manifest.findall('{http://ns.adobe.com/f4m/2.0}media')
778 for i, media_el in enumerate(media_nodes):
779 if manifest_version == '2.0':
780 manifest_url = '/'.join(manifest_url.split('/')[:-1]) + '/' + media_el.attrib.get('href')
781 tbr = int_or_none(media_el.attrib.get('bitrate'))
782 format_id = 'f4m-%d' % (i if tbr is None else tbr)
784 'format_id': format_id,
788 'width': int_or_none(media_el.attrib.get('width')),
789 'height': int_or_none(media_el.attrib.get('height')),
791 self._sort_formats(formats)
795 def _extract_m3u8_formats(self, m3u8_url, video_id, ext=None,
796 entry_protocol='m3u8', preference=None):
799 'format_id': 'm3u8-meta',
804 'resolution': 'multiple',
805 'format_note': 'Quality selection URL',
808 format_url = lambda u: (
810 if re.match(r'^https?://', u)
811 else compat_urlparse.urljoin(m3u8_url, u))
813 m3u8_doc = self._download_webpage(
815 note='Downloading m3u8 information',
816 errnote='Failed to download m3u8 information')
819 r'(?P<key>[a-zA-Z_-]+)=(?P<val>"[^"]+"|[^",]+)(?:,|$)')
820 for line in m3u8_doc.splitlines():
821 if line.startswith('#EXT-X-STREAM-INF:'):
823 for m in kv_rex.finditer(line):
825 if v.startswith('"'):
827 last_info[m.group('key')] = v
828 elif line.startswith('#') or not line.strip():
831 if last_info is None:
832 formats.append({'url': format_url(line)})
834 tbr = int_or_none(last_info.get('BANDWIDTH'), scale=1000)
837 'format_id': 'm3u8-%d' % (tbr if tbr else len(formats)),
838 'url': format_url(line.strip()),
841 'protocol': entry_protocol,
842 'preference': preference,
844 codecs = last_info.get('CODECS')
846 # TODO: looks like video codec is not always necessarily goes first
847 va_codecs = codecs.split(',')
849 f['vcodec'] = va_codecs[0].partition('.')[0]
850 if len(va_codecs) > 1 and va_codecs[1]:
851 f['acodec'] = va_codecs[1].partition('.')[0]
852 resolution = last_info.get('RESOLUTION')
854 width_str, height_str = resolution.split('x')
855 f['width'] = int(width_str)
856 f['height'] = int(height_str)
859 self._sort_formats(formats)
862 # TODO: improve extraction
863 def _extract_smil_formats(self, smil_url, video_id):
864 smil = self._download_xml(
865 smil_url, video_id, 'Downloading SMIL file',
866 'Unable to download SMIL file')
868 base = smil.find('./head/meta').get('base')
872 for video in smil.findall('./body/switch/video'):
873 src = video.get('src')
876 bitrate = int_or_none(video.get('system-bitrate') or video.get('systemBitrate'), 1000)
877 width = int_or_none(video.get('width'))
878 height = int_or_none(video.get('height'))
879 proto = video.get('proto')
882 if base.startswith('rtmp'):
884 elif base.startswith('http'):
886 ext = video.get('ext')
888 formats.extend(self._extract_m3u8_formats(src, video_id, ext))
889 elif proto == 'rtmp':
891 streamer = video.get('streamer') or base
896 'format_id': 'rtmp-%d' % (rtmp_count if bitrate is None else bitrate),
901 self._sort_formats(formats)
905 def _live_title(self, name):
906 """ Generate the title for a live video """
907 now = datetime.datetime.now()
908 now_str = now.strftime("%Y-%m-%d %H:%M")
909 return name + ' ' + now_str
911 def _int(self, v, name, fatal=False, **kwargs):
912 res = int_or_none(v, **kwargs)
913 if 'get_attr' in kwargs:
914 print(getattr(v, kwargs['get_attr']))
916 msg = 'Failed to extract %s: Could not parse value %r' % (name, v)
918 raise ExtractorError(msg)
920 self._downloader.report_warning(msg)
923 def _float(self, v, name, fatal=False, **kwargs):
924 res = float_or_none(v, **kwargs)
926 msg = 'Failed to extract %s: Could not parse value %r' % (name, v)
928 raise ExtractorError(msg)
930 self._downloader.report_warning(msg)
933 def _set_cookie(self, domain, name, value, expire_time=None):
934 cookie = compat_cookiejar.Cookie(
935 0, name, value, None, None, domain, None,
936 None, '/', True, False, expire_time, '', None, None, None)
937 self._downloader.cookiejar.set_cookie(cookie)
939 def get_testcases(self, include_onlymatching=False):
940 t = getattr(self, '_TEST', None)
942 assert not hasattr(self, '_TESTS'), \
943 '%s has _TEST and _TESTS' % type(self).__name__
946 tests = getattr(self, '_TESTS', [])
948 if not include_onlymatching and t.get('only_matching', False):
950 t['name'] = type(self).__name__[:-len('IE')]
953 def is_suitable(self, age_limit):
954 """ Test whether the extractor is generally suitable for the given
955 age limit (i.e. pornographic sites are not, all others usually are) """
957 any_restricted = False
958 for tc in self.get_testcases(include_onlymatching=False):
960 tc = tc['playlist'][0]
961 is_restricted = age_restricted(
962 tc.get('info_dict', {}).get('age_limit'), age_limit)
963 if not is_restricted:
965 any_restricted = any_restricted or is_restricted
966 return not any_restricted
969 class SearchInfoExtractor(InfoExtractor):
971 Base class for paged search queries extractors.
972 They accept urls in the format _SEARCH_KEY(|all|[0-9]):{query}
973 Instances should define _SEARCH_KEY and _MAX_RESULTS.
977 def _make_valid_url(cls):
978 return r'%s(?P<prefix>|[1-9][0-9]*|all):(?P<query>[\s\S]+)' % cls._SEARCH_KEY
981 def suitable(cls, url):
982 return re.match(cls._make_valid_url(), url) is not None
984 def _real_extract(self, query):
985 mobj = re.match(self._make_valid_url(), query)
987 raise ExtractorError('Invalid search query "%s"' % query)
989 prefix = mobj.group('prefix')
990 query = mobj.group('query')
992 return self._get_n_results(query, 1)
993 elif prefix == 'all':
994 return self._get_n_results(query, self._MAX_RESULTS)
998 raise ExtractorError('invalid download number %s for query "%s"' % (n, query))
999 elif n > self._MAX_RESULTS:
1000 self._downloader.report_warning('%s returns max %i results (you requested %i)' % (self._SEARCH_KEY, self._MAX_RESULTS, n))
1001 n = self._MAX_RESULTS
1002 return self._get_n_results(query, n)
1004 def _get_n_results(self, query, n):
1005 """Get a specified number of results for a query"""
1006 raise NotImplementedError("This method must be implemented by subclasses")
1009 def SEARCH_KEY(self):
1010 return self._SEARCH_KEY