1 from __future__ import unicode_literals
13 import xml.etree.ElementTree
15 from ..compat import (
19 compat_urllib_parse_urlparse,
34 _NO_DEFAULT = object()
37 class InfoExtractor(object):
38 """Information Extractor class.
40 Information extractors are the classes that, given a URL, extract
41 information about the video (or videos) the URL refers to. This
42 information includes the real video URL, the video title, author and
43 others. The information is stored in a dictionary which is then
44 passed to the YoutubeDL. The YoutubeDL processes this
45 information possibly downloading the video to the file system, among
46 other possible outcomes.
48 The type field determines the the type of the result.
49 By far the most common value (and the default if _type is missing) is
50 "video", which indicates a single video.
52 For a video, the dictionaries must include the following fields:
55 title: Video title, unescaped.
57 Additionally, it must contain either a formats entry or a url one:
59 formats: A list of dictionaries for each format available, ordered
60 from worst to best quality.
63 * url Mandatory. The URL of the video file
64 * ext Will be calculated from url if missing
65 * format A human-readable description of the format
66 ("mp4 container with h264/opus").
67 Calculated from the format_id, width, height.
68 and format_note fields if missing.
69 * format_id A short description of the format
70 ("mp4_h264_opus" or "19").
71 Technically optional, but strongly recommended.
72 * format_note Additional info about the format
73 ("3D" or "DASH video")
74 * width Width of the video, if known
75 * height Height of the video, if known
76 * resolution Textual description of width and height
77 * tbr Average bitrate of audio and video in KBit/s
78 * abr Average audio bitrate in KBit/s
79 * acodec Name of the audio codec in use
80 * asr Audio sampling rate in Hertz
81 * vbr Average video bitrate in KBit/s
83 * vcodec Name of the video codec in use
84 * container Name of the container format
85 * filesize The number of bytes, if known in advance
86 * filesize_approx An estimate for the number of bytes
87 * player_url SWF Player URL (used for rtmpdump).
88 * protocol The protocol that will be used for the actual
90 "http", "https", "rtsp", "rtmp", "m3u8" or so.
91 * preference Order number of this format. If this field is
92 present and not None, the formats get sorted
93 by this field, regardless of all other values.
94 -1 for default (order by other properties),
95 -2 or smaller for less than default.
96 < -1000 to hide the format (if there is
97 another one which is strictly better)
98 * language_preference Is this in the correct requested
100 10 if it's what the URL is about,
101 -1 for default (don't know),
102 -10 otherwise, other values reserved for now.
103 * quality Order number of the video quality of this
104 format, irrespective of the file format.
105 -1 for default (order by other properties),
106 -2 or smaller for less than default.
107 * source_preference Order number for this video source
108 (quality takes higher priority)
109 -1 for default (order by other properties),
110 -2 or smaller for less than default.
111 * http_referer HTTP Referer header value to set.
112 * http_method HTTP method to use for the download.
113 * http_headers A dictionary of additional HTTP headers
114 to add to the request.
115 * http_post_data Additional data to send with a POST
117 url: Final video URL.
118 ext: Video filename extension.
119 format: The video format, defaults to ext (used for --get-format)
120 player_url: SWF Player URL (used for rtmpdump).
122 The following fields are optional:
124 alt_title: A secondary title of the video.
125 display_id An alternative identifier for the video, not necessarily
126 unique, but available before title. Typically, id is
127 something like "4234987", title "Dancing naked mole rats",
128 and display_id "dancing-naked-mole-rats"
129 thumbnails: A list of dictionaries, with the following entries:
131 * "width" (optional, int)
132 * "height" (optional, int)
133 * "resolution" (optional, string "{width}x{height"},
135 thumbnail: Full URL to a video thumbnail image.
136 description: Full video description.
137 uploader: Full name of the video uploader.
138 timestamp: UNIX timestamp of the moment the video became available.
139 upload_date: Video upload date (YYYYMMDD).
140 If not explicitly set, calculated from timestamp.
141 uploader_id: Nickname or id of the video uploader.
142 location: Physical location where the video was filmed.
143 subtitles: The subtitle file contents as a dictionary in the format
144 {language: subtitles}.
145 duration: Length of the video in seconds, as an integer.
146 view_count: How many users have watched the video on the platform.
147 like_count: Number of positive ratings of the video
148 dislike_count: Number of negative ratings of the video
149 comment_count: Number of comments on the video
150 comments: A list of comments, each with one or more of the following
151 properties (all but one of text or html optional):
152 * "author" - human-readable name of the comment author
153 * "author_id" - user ID of the comment author
155 * "html" - Comment as HTML
156 * "text" - Plain text of the comment
157 * "timestamp" - UNIX timestamp of comment
158 * "parent" - ID of the comment this one is replying to.
159 Set to "root" to indicate that this is a
160 comment to the original video.
161 age_limit: Age restriction for the video, as an integer (years)
162 webpage_url: The url to the video webpage, if given to youtube-dl it
163 should allow to get the same result again. (It will be set
164 by YoutubeDL if it's missing)
165 categories: A list of categories that the video falls in, for example
167 is_live: True, False, or None (=unknown). Whether this video is a
168 live stream that goes on instead of a fixed-length video.
170 Unless mentioned otherwise, the fields should be Unicode strings.
172 Unless mentioned otherwise, None is equivalent to absence of information.
175 _type "playlist" indicates multiple videos.
176 There must be a key "entries", which is a list, an iterable, or a PagedList
177 object, each element of which is a valid dictionary by this specification.
179 Additionally, playlists can have "title" and "id" attributes with the same
180 semantics as videos (see above).
183 _type "multi_video" indicates that there are multiple videos that
184 form a single show, for examples multiple acts of an opera or TV episode.
185 It must have an entries key like a playlist and contain all the keys
186 required for a video at the same time.
189 _type "url" indicates that the video must be extracted from another
190 location, possibly by a different extractor. Its only required key is:
191 "url" - the next URL to extract.
192 The key "ie_key" can be set to the class name (minus the trailing "IE",
193 e.g. "Youtube") if the extractor class is known in advance.
194 Additionally, the dictionary may have any properties of the resolved entity
195 known in advance, for example "title" if the title of the referred video is
199 _type "url_transparent" entities have the same specification as "url", but
200 indicate that the given additional information is more precise than the one
201 associated with the resolved URL.
202 This is useful when a site employs a video service that hosts the video and
203 its technical metadata, but that video service does not embed a useful
204 title, description etc.
207 Subclasses of this one should re-define the _real_initialize() and
208 _real_extract() methods and define a _VALID_URL regexp.
209 Probably, they should also be added to the list of extractors.
211 Finally, the _WORKING attribute should be set to False for broken IEs
212 in order to warn the users and skip the tests.
219 def __init__(self, downloader=None):
220 """Constructor. Receives an optional downloader."""
222 self.set_downloader(downloader)
225 def suitable(cls, url):
226 """Receives a URL and returns True if suitable for this IE."""
228 # This does not use has/getattr intentionally - we want to know whether
229 # we have cached the regexp for *this* class, whereas getattr would also
230 # match the superclass
231 if '_VALID_URL_RE' not in cls.__dict__:
232 cls._VALID_URL_RE = re.compile(cls._VALID_URL)
233 return cls._VALID_URL_RE.match(url) is not None
236 def _match_id(cls, url):
237 if '_VALID_URL_RE' not in cls.__dict__:
238 cls._VALID_URL_RE = re.compile(cls._VALID_URL)
239 m = cls._VALID_URL_RE.match(url)
245 """Getter method for _WORKING."""
248 def initialize(self):
249 """Initializes an instance (authentication, etc)."""
251 self._real_initialize()
254 def extract(self, url):
255 """Extracts URL information and returns it in list of dicts."""
257 return self._real_extract(url)
259 def set_downloader(self, downloader):
260 """Sets the downloader for this IE."""
261 self._downloader = downloader
263 def _real_initialize(self):
264 """Real initialization process. Redefine in subclasses."""
267 def _real_extract(self, url):
268 """Real extraction process. Redefine in subclasses."""
273 """A string for getting the InfoExtractor with get_info_extractor"""
274 return cls.__name__[:-2]
278 return type(self).__name__[:-2]
280 def _request_webpage(self, url_or_request, video_id, note=None, errnote=None, fatal=True):
281 """ Returns the response handle """
283 self.report_download_webpage(video_id)
284 elif note is not False:
286 self.to_screen('%s' % (note,))
288 self.to_screen('%s: %s' % (video_id, note))
290 return self._downloader.urlopen(url_or_request)
291 except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
295 errnote = 'Unable to download webpage'
296 errmsg = '%s: %s' % (errnote, compat_str(err))
298 raise ExtractorError(errmsg, sys.exc_info()[2], cause=err)
300 self._downloader.report_warning(errmsg)
303 def _download_webpage_handle(self, url_or_request, video_id, note=None, errnote=None, fatal=True):
304 """ Returns a tuple (page content as string, URL handle) """
305 # Strip hashes from the URL (#1038)
306 if isinstance(url_or_request, (compat_str, str)):
307 url_or_request = url_or_request.partition('#')[0]
309 urlh = self._request_webpage(url_or_request, video_id, note, errnote, fatal)
313 content = self._webpage_read_content(urlh, url_or_request, video_id, note, errnote, fatal)
314 return (content, urlh)
316 def _webpage_read_content(self, urlh, url_or_request, video_id, note=None, errnote=None, fatal=True, prefix=None):
317 content_type = urlh.headers.get('Content-Type', '')
318 webpage_bytes = urlh.read()
319 if prefix is not None:
320 webpage_bytes = prefix + webpage_bytes
321 m = re.match(r'[a-zA-Z0-9_.-]+/[a-zA-Z0-9_.-]+\s*;\s*charset=(.+)', content_type)
323 encoding = m.group(1)
325 m = re.search(br'<meta[^>]+charset=[\'"]?([^\'")]+)[ /\'">]',
326 webpage_bytes[:1024])
328 encoding = m.group(1).decode('ascii')
329 elif webpage_bytes.startswith(b'\xff\xfe'):
333 if self._downloader.params.get('dump_intermediate_pages', False):
335 url = url_or_request.get_full_url()
336 except AttributeError:
338 self.to_screen('Dumping request to ' + url)
339 dump = base64.b64encode(webpage_bytes).decode('ascii')
340 self._downloader.to_screen(dump)
341 if self._downloader.params.get('write_pages', False):
343 url = url_or_request.get_full_url()
344 except AttributeError:
346 basen = '%s_%s' % (video_id, url)
348 h = '___' + hashlib.md5(basen.encode('utf-8')).hexdigest()
349 basen = basen[:240 - len(h)] + h
350 raw_filename = basen + '.dump'
351 filename = sanitize_filename(raw_filename, restricted=True)
352 self.to_screen('Saving request to ' + filename)
353 # Working around MAX_PATH limitation on Windows (see
354 # http://msdn.microsoft.com/en-us/library/windows/desktop/aa365247(v=vs.85).aspx)
356 absfilepath = os.path.abspath(filename)
357 if len(absfilepath) > 259:
358 filename = '\\\\?\\' + absfilepath
359 with open(filename, 'wb') as outf:
360 outf.write(webpage_bytes)
363 content = webpage_bytes.decode(encoding, 'replace')
365 content = webpage_bytes.decode('utf-8', 'replace')
367 if ('<title>Access to this site is blocked</title>' in content and
368 'Websense' in content[:512]):
369 msg = 'Access to this webpage has been blocked by Websense filtering software in your network.'
370 blocked_iframe = self._html_search_regex(
371 r'<iframe src="([^"]+)"', content,
372 'Websense information URL', default=None)
374 msg += ' Visit %s for more details' % blocked_iframe
375 raise ExtractorError(msg, expected=True)
379 def _download_webpage(self, url_or_request, video_id, note=None, errnote=None, fatal=True, tries=1, timeout=5):
380 """ Returns the data of the page as a string """
383 while success is False:
385 res = self._download_webpage_handle(url_or_request, video_id, note, errnote, fatal)
387 except compat_http_client.IncompleteRead as e:
389 if try_count >= tries:
391 self._sleep(timeout, video_id)
398 def _download_xml(self, url_or_request, video_id,
399 note='Downloading XML', errnote='Unable to download XML',
400 transform_source=None, fatal=True):
401 """Return the xml as an xml.etree.ElementTree.Element"""
402 xml_string = self._download_webpage(
403 url_or_request, video_id, note, errnote, fatal=fatal)
404 if xml_string is False:
407 xml_string = transform_source(xml_string)
408 return xml.etree.ElementTree.fromstring(xml_string.encode('utf-8'))
410 def _download_json(self, url_or_request, video_id,
411 note='Downloading JSON metadata',
412 errnote='Unable to download JSON metadata',
413 transform_source=None,
415 json_string = self._download_webpage(
416 url_or_request, video_id, note, errnote, fatal=fatal)
417 if (not fatal) and json_string is False:
419 return self._parse_json(
420 json_string, video_id, transform_source=transform_source, fatal=fatal)
422 def _parse_json(self, json_string, video_id, transform_source=None, fatal=True):
424 json_string = transform_source(json_string)
426 return json.loads(json_string)
427 except ValueError as ve:
428 errmsg = '%s: Failed to parse JSON ' % video_id
430 raise ExtractorError(errmsg, cause=ve)
432 self.report_warning(errmsg + str(ve))
434 def report_warning(self, msg, video_id=None):
435 idstr = '' if video_id is None else '%s: ' % video_id
436 self._downloader.report_warning(
437 '[%s] %s%s' % (self.IE_NAME, idstr, msg))
439 def to_screen(self, msg):
440 """Print msg to screen, prefixing it with '[ie_name]'"""
441 self._downloader.to_screen('[%s] %s' % (self.IE_NAME, msg))
443 def report_extraction(self, id_or_name):
444 """Report information extraction."""
445 self.to_screen('%s: Extracting information' % id_or_name)
447 def report_download_webpage(self, video_id):
448 """Report webpage download."""
449 self.to_screen('%s: Downloading webpage' % video_id)
451 def report_age_confirmation(self):
452 """Report attempt to confirm age."""
453 self.to_screen('Confirming age')
455 def report_login(self):
456 """Report attempt to log in."""
457 self.to_screen('Logging in')
459 # Methods for following #608
461 def url_result(url, ie=None, video_id=None):
462 """Returns a url that points to a page that should be processed"""
463 # TODO: ie should be the class used for getting the info
464 video_info = {'_type': 'url',
467 if video_id is not None:
468 video_info['id'] = video_id
472 def playlist_result(entries, playlist_id=None, playlist_title=None, playlist_description=None):
473 """Returns a playlist"""
474 video_info = {'_type': 'playlist',
477 video_info['id'] = playlist_id
479 video_info['title'] = playlist_title
480 if playlist_description:
481 video_info['description'] = playlist_description
484 def _search_regex(self, pattern, string, name, default=_NO_DEFAULT, fatal=True, flags=0, group=None):
486 Perform a regex search on the given string, using a single or a list of
487 patterns returning the first matching group.
488 In case of failure return a default value or raise a WARNING or a
489 RegexNotFoundError, depending on fatal, specifying the field name.
491 if isinstance(pattern, (str, compat_str, compiled_regex_type)):
492 mobj = re.search(pattern, string, flags)
495 mobj = re.search(p, string, flags)
499 if os.name != 'nt' and sys.stderr.isatty():
500 _name = '\033[0;34m%s\033[0m' % name
506 # return the first matching group
507 return next(g for g in mobj.groups() if g is not None)
509 return mobj.group(group)
510 elif default is not _NO_DEFAULT:
513 raise RegexNotFoundError('Unable to extract %s' % _name)
515 self._downloader.report_warning('unable to extract %s; '
516 'please report this issue on http://yt-dl.org/bug' % _name)
519 def _html_search_regex(self, pattern, string, name, default=_NO_DEFAULT, fatal=True, flags=0, group=None):
521 Like _search_regex, but strips HTML tags and unescapes entities.
523 res = self._search_regex(pattern, string, name, default, fatal, flags, group)
525 return clean_html(res).strip()
529 def _get_login_info(self):
531 Get the the login info as (username, password)
532 It will look in the netrc file using the _NETRC_MACHINE value
533 If there's no info available, return (None, None)
535 if self._downloader is None:
540 downloader_params = self._downloader.params
542 # Attempt to use provided username and password or .netrc data
543 if downloader_params.get('username', None) is not None:
544 username = downloader_params['username']
545 password = downloader_params['password']
546 elif downloader_params.get('usenetrc', False):
548 info = netrc.netrc().authenticators(self._NETRC_MACHINE)
553 raise netrc.NetrcParseError('No authenticators for %s' % self._NETRC_MACHINE)
554 except (IOError, netrc.NetrcParseError) as err:
555 self._downloader.report_warning('parsing .netrc: %s' % compat_str(err))
557 return (username, password)
559 def _get_tfa_info(self):
561 Get the two-factor authentication info
562 TODO - asking the user will be required for sms/phone verify
563 currently just uses the command line option
564 If there's no info available, return None
566 if self._downloader is None:
568 downloader_params = self._downloader.params
570 if downloader_params.get('twofactor', None) is not None:
571 return downloader_params['twofactor']
575 # Helper functions for extracting OpenGraph info
577 def _og_regexes(prop):
578 content_re = r'content=(?:"([^>]+?)"|\'([^>]+?)\')'
579 property_re = r'(?:name|property)=[\'"]og:%s[\'"]' % re.escape(prop)
580 template = r'<meta[^>]+?%s[^>]+?%s'
582 template % (property_re, content_re),
583 template % (content_re, property_re),
586 def _og_search_property(self, prop, html, name=None, **kargs):
588 name = 'OpenGraph %s' % prop
589 escaped = self._search_regex(self._og_regexes(prop), html, name, flags=re.DOTALL, **kargs)
592 return unescapeHTML(escaped)
594 def _og_search_thumbnail(self, html, **kargs):
595 return self._og_search_property('image', html, 'thumbnail url', fatal=False, **kargs)
597 def _og_search_description(self, html, **kargs):
598 return self._og_search_property('description', html, fatal=False, **kargs)
600 def _og_search_title(self, html, **kargs):
601 return self._og_search_property('title', html, **kargs)
603 def _og_search_video_url(self, html, name='video url', secure=True, **kargs):
604 regexes = self._og_regexes('video') + self._og_regexes('video:url')
606 regexes = self._og_regexes('video:secure_url') + regexes
607 return self._html_search_regex(regexes, html, name, **kargs)
609 def _og_search_url(self, html, **kargs):
610 return self._og_search_property('url', html, **kargs)
612 def _html_search_meta(self, name, html, display_name=None, fatal=False, **kwargs):
613 if display_name is None:
615 return self._html_search_regex(
617 (?=[^>]+(?:itemprop|name|property)=(["\']?)%s\1)
618 [^>]+?content=(["\'])(?P<content>.*?)\2''' % re.escape(name),
619 html, display_name, fatal=fatal, group='content', **kwargs)
621 def _dc_search_uploader(self, html):
622 return self._html_search_meta('dc.creator', html, 'uploader')
624 def _rta_search(self, html):
625 # See http://www.rtalabel.org/index.php?content=howtofaq#single
626 if re.search(r'(?ix)<meta\s+name="rating"\s+'
627 r' content="RTA-5042-1996-1400-1577-RTA"',
632 def _media_rating_search(self, html):
633 # See http://www.tjg-designs.com/WP/metadata-code-examples-adding-metadata-to-your-web-pages/
634 rating = self._html_search_meta('rating', html)
646 return RATING_TABLE.get(rating.lower(), None)
648 def _twitter_search_player(self, html):
649 return self._html_search_meta('twitter:player', html,
650 'twitter card player')
652 def _sort_formats(self, formats):
654 raise ExtractorError('No video formats found')
657 # TODO remove the following workaround
658 from ..utils import determine_ext
659 if not f.get('ext') and 'url' in f:
660 f['ext'] = determine_ext(f['url'])
662 preference = f.get('preference')
663 if preference is None:
664 proto = f.get('protocol')
666 proto = compat_urllib_parse_urlparse(f.get('url', '')).scheme
668 preference = 0 if proto in ['http', 'https'] else -0.1
669 if f.get('ext') in ['f4f', 'f4m']: # Not yet supported
672 if f.get('vcodec') == 'none': # audio only
673 if self._downloader.params.get('prefer_free_formats'):
674 ORDER = ['aac', 'mp3', 'm4a', 'webm', 'ogg', 'opus']
676 ORDER = ['webm', 'opus', 'ogg', 'mp3', 'aac', 'm4a']
679 audio_ext_preference = ORDER.index(f['ext'])
681 audio_ext_preference = -1
683 if self._downloader.params.get('prefer_free_formats'):
684 ORDER = ['flv', 'mp4', 'webm']
686 ORDER = ['webm', 'flv', 'mp4']
688 ext_preference = ORDER.index(f['ext'])
691 audio_ext_preference = 0
695 f.get('language_preference') if f.get('language_preference') is not None else -1,
696 f.get('quality') if f.get('quality') is not None else -1,
697 f.get('height') if f.get('height') is not None else -1,
698 f.get('width') if f.get('width') is not None else -1,
700 f.get('tbr') if f.get('tbr') is not None else -1,
701 f.get('vbr') if f.get('vbr') is not None else -1,
702 f.get('abr') if f.get('abr') is not None else -1,
703 audio_ext_preference,
704 f.get('fps') if f.get('fps') is not None else -1,
705 f.get('filesize') if f.get('filesize') is not None else -1,
706 f.get('filesize_approx') if f.get('filesize_approx') is not None else -1,
707 f.get('source_preference') if f.get('source_preference') is not None else -1,
710 formats.sort(key=_formats_key)
712 def http_scheme(self):
713 """ Either "http:" or "https:", depending on the user's preferences """
716 if self._downloader.params.get('prefer_insecure', False)
719 def _proto_relative_url(self, url, scheme=None):
722 if url.startswith('//'):
724 scheme = self.http_scheme()
729 def _sleep(self, timeout, video_id, msg_template=None):
730 if msg_template is None:
731 msg_template = '%(video_id)s: Waiting for %(timeout)s seconds'
732 msg = msg_template % {'video_id': video_id, 'timeout': timeout}
736 def _extract_f4m_formats(self, manifest_url, video_id):
737 manifest = self._download_xml(
738 manifest_url, video_id, 'Downloading f4m manifest',
739 'Unable to download f4m manifest')
742 media_nodes = manifest.findall('{http://ns.adobe.com/f4m/1.0}media')
743 for i, media_el in enumerate(media_nodes):
744 tbr = int_or_none(media_el.attrib.get('bitrate'))
745 format_id = 'f4m-%d' % (i if tbr is None else tbr)
747 'format_id': format_id,
751 'width': int_or_none(media_el.attrib.get('width')),
752 'height': int_or_none(media_el.attrib.get('height')),
754 self._sort_formats(formats)
758 def _extract_m3u8_formats(self, m3u8_url, video_id, ext=None,
759 entry_protocol='m3u8', preference=None):
762 'format_id': 'm3u8-meta',
767 'resolution': 'multiple',
768 'format_note': 'Quality selection URL',
771 format_url = lambda u: (
773 if re.match(r'^https?://', u)
774 else compat_urlparse.urljoin(m3u8_url, u))
776 m3u8_doc = self._download_webpage(
778 note='Downloading m3u8 information',
779 errnote='Failed to download m3u8 information')
782 r'(?P<key>[a-zA-Z_-]+)=(?P<val>"[^"]+"|[^",]+)(?:,|$)')
783 for line in m3u8_doc.splitlines():
784 if line.startswith('#EXT-X-STREAM-INF:'):
786 for m in kv_rex.finditer(line):
788 if v.startswith('"'):
790 last_info[m.group('key')] = v
791 elif line.startswith('#') or not line.strip():
794 if last_info is None:
795 formats.append({'url': format_url(line)})
797 tbr = int_or_none(last_info.get('BANDWIDTH'), scale=1000)
800 'format_id': 'm3u8-%d' % (tbr if tbr else len(formats)),
801 'url': format_url(line.strip()),
804 'protocol': entry_protocol,
805 'preference': preference,
807 codecs = last_info.get('CODECS')
809 # TODO: looks like video codec is not always necessarily goes first
810 va_codecs = codecs.split(',')
812 f['vcodec'] = va_codecs[0].partition('.')[0]
813 if len(va_codecs) > 1 and va_codecs[1]:
814 f['acodec'] = va_codecs[1].partition('.')[0]
815 resolution = last_info.get('RESOLUTION')
817 width_str, height_str = resolution.split('x')
818 f['width'] = int(width_str)
819 f['height'] = int(height_str)
822 self._sort_formats(formats)
825 # TODO: improve extraction
826 def _extract_smil_formats(self, smil_url, video_id):
827 smil = self._download_xml(
828 smil_url, video_id, 'Downloading SMIL file',
829 'Unable to download SMIL file')
831 base = smil.find('./head/meta').get('base')
835 for video in smil.findall('./body/switch/video'):
836 src = video.get('src')
839 bitrate = int_or_none(video.get('system-bitrate') or video.get('systemBitrate'), 1000)
840 width = int_or_none(video.get('width'))
841 height = int_or_none(video.get('height'))
842 proto = video.get('proto')
845 if base.startswith('rtmp'):
847 elif base.startswith('http'):
849 ext = video.get('ext')
851 formats.extend(self._extract_m3u8_formats(src, video_id, ext))
852 elif proto == 'rtmp':
854 streamer = video.get('streamer') or base
859 'format_id': 'rtmp-%d' % (rtmp_count if bitrate is None else bitrate),
864 self._sort_formats(formats)
868 def _live_title(self, name):
869 """ Generate the title for a live video """
870 now = datetime.datetime.now()
871 now_str = now.strftime("%Y-%m-%d %H:%M")
872 return name + ' ' + now_str
874 def _int(self, v, name, fatal=False, **kwargs):
875 res = int_or_none(v, **kwargs)
876 if 'get_attr' in kwargs:
877 print(getattr(v, kwargs['get_attr']))
879 msg = 'Failed to extract %s: Could not parse value %r' % (name, v)
881 raise ExtractorError(msg)
883 self._downloader.report_warning(msg)
886 def _float(self, v, name, fatal=False, **kwargs):
887 res = float_or_none(v, **kwargs)
889 msg = 'Failed to extract %s: Could not parse value %r' % (name, v)
891 raise ExtractorError(msg)
893 self._downloader.report_warning(msg)
896 def _set_cookie(self, domain, name, value, expire_time=None):
897 cookie = compat_cookiejar.Cookie(
898 0, name, value, None, None, domain, None,
899 None, '/', True, False, expire_time, '', None, None, None)
900 self._downloader.cookiejar.set_cookie(cookie)
902 def get_testcases(self, include_onlymatching=False):
903 t = getattr(self, '_TEST', None)
905 assert not hasattr(self, '_TESTS'), \
906 '%s has _TEST and _TESTS' % type(self).__name__
909 tests = getattr(self, '_TESTS', [])
911 if not include_onlymatching and t.get('only_matching', False):
913 t['name'] = type(self).__name__[:-len('IE')]
916 def is_suitable(self, age_limit):
917 """ Test whether the extractor is generally suitable for the given
918 age limit (i.e. pornographic sites are not, all others usually are) """
920 any_restricted = False
921 for tc in self.get_testcases(include_onlymatching=False):
923 tc = tc['playlist'][0]
924 is_restricted = age_restricted(
925 tc.get('info_dict', {}).get('age_limit'), age_limit)
926 if not is_restricted:
928 any_restricted = any_restricted or is_restricted
929 return not any_restricted
932 class SearchInfoExtractor(InfoExtractor):
934 Base class for paged search queries extractors.
935 They accept urls in the format _SEARCH_KEY(|all|[0-9]):{query}
936 Instances should define _SEARCH_KEY and _MAX_RESULTS.
940 def _make_valid_url(cls):
941 return r'%s(?P<prefix>|[1-9][0-9]*|all):(?P<query>[\s\S]+)' % cls._SEARCH_KEY
944 def suitable(cls, url):
945 return re.match(cls._make_valid_url(), url) is not None
947 def _real_extract(self, query):
948 mobj = re.match(self._make_valid_url(), query)
950 raise ExtractorError('Invalid search query "%s"' % query)
952 prefix = mobj.group('prefix')
953 query = mobj.group('query')
955 return self._get_n_results(query, 1)
956 elif prefix == 'all':
957 return self._get_n_results(query, self._MAX_RESULTS)
961 raise ExtractorError('invalid download number %s for query "%s"' % (n, query))
962 elif n > self._MAX_RESULTS:
963 self._downloader.report_warning('%s returns max %i results (you requested %i)' % (self._SEARCH_KEY, self._MAX_RESULTS, n))
964 n = self._MAX_RESULTS
965 return self._get_n_results(query, n)
967 def _get_n_results(self, query, n):
968 """Get a specified number of results for a query"""
969 raise NotImplementedError("This method must be implemented by subclasses")
972 def SEARCH_KEY(self):
973 return self._SEARCH_KEY