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_method HTTP method to use for the download.
112 * http_headers A dictionary of additional HTTP headers
113 to add to the request.
114 * http_post_data Additional data to send with a POST
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 url: Final video URL.
120 ext: Video filename extension.
121 format: The video format, defaults to ext (used for --get-format)
122 player_url: SWF Player URL (used for rtmpdump).
124 The following fields are optional:
126 alt_title: A secondary title of the video.
127 display_id An alternative identifier for the video, not necessarily
128 unique, but available before title. Typically, id is
129 something like "4234987", title "Dancing naked mole rats",
130 and display_id "dancing-naked-mole-rats"
131 thumbnails: A list of dictionaries, with the following entries:
133 * "width" (optional, int)
134 * "height" (optional, int)
135 * "resolution" (optional, string "{width}x{height"},
137 thumbnail: Full URL to a video thumbnail image.
138 description: Full video description.
139 uploader: Full name of the video uploader.
140 timestamp: UNIX timestamp of the moment the video became available.
141 upload_date: Video upload date (YYYYMMDD).
142 If not explicitly set, calculated from timestamp.
143 uploader_id: Nickname or id of the video uploader.
144 location: Physical location where the video was filmed.
145 subtitles: The subtitle file contents as a dictionary in the format
146 {language: subtitles}.
147 duration: Length of the video in seconds, as an integer.
148 view_count: How many users have watched the video on the platform.
149 like_count: Number of positive ratings of the video
150 dislike_count: Number of negative ratings of the video
151 comment_count: Number of comments on the video
152 comments: A list of comments, each with one or more of the following
153 properties (all but one of text or html optional):
154 * "author" - human-readable name of the comment author
155 * "author_id" - user ID of the comment author
157 * "html" - Comment as HTML
158 * "text" - Plain text of the comment
159 * "timestamp" - UNIX timestamp of comment
160 * "parent" - ID of the comment this one is replying to.
161 Set to "root" to indicate that this is a
162 comment to the original video.
163 age_limit: Age restriction for the video, as an integer (years)
164 webpage_url: The url to the video webpage, if given to youtube-dl it
165 should allow to get the same result again. (It will be set
166 by YoutubeDL if it's missing)
167 categories: A list of categories that the video falls in, for example
169 is_live: True, False, or None (=unknown). Whether this video is a
170 live stream that goes on instead of a fixed-length video.
172 Unless mentioned otherwise, the fields should be Unicode strings.
174 Unless mentioned otherwise, None is equivalent to absence of information.
177 _type "playlist" indicates multiple videos.
178 There must be a key "entries", which is a list, an iterable, or a PagedList
179 object, each element of which is a valid dictionary by this specification.
181 Additionally, playlists can have "title" and "id" attributes with the same
182 semantics as videos (see above).
185 _type "multi_video" indicates that there are multiple videos that
186 form a single show, for examples multiple acts of an opera or TV episode.
187 It must have an entries key like a playlist and contain all the keys
188 required for a video at the same time.
191 _type "url" indicates that the video must be extracted from another
192 location, possibly by a different extractor. Its only required key is:
193 "url" - the next URL to extract.
194 The key "ie_key" can be set to the class name (minus the trailing "IE",
195 e.g. "Youtube") if the extractor class is known in advance.
196 Additionally, the dictionary may have any properties of the resolved entity
197 known in advance, for example "title" if the title of the referred video is
201 _type "url_transparent" entities have the same specification as "url", but
202 indicate that the given additional information is more precise than the one
203 associated with the resolved URL.
204 This is useful when a site employs a video service that hosts the video and
205 its technical metadata, but that video service does not embed a useful
206 title, description etc.
209 Subclasses of this one should re-define the _real_initialize() and
210 _real_extract() methods and define a _VALID_URL regexp.
211 Probably, they should also be added to the list of extractors.
213 Finally, the _WORKING attribute should be set to False for broken IEs
214 in order to warn the users and skip the tests.
221 def __init__(self, downloader=None):
222 """Constructor. Receives an optional downloader."""
224 self.set_downloader(downloader)
227 def suitable(cls, url):
228 """Receives a URL and returns True if suitable for this IE."""
230 # This does not use has/getattr intentionally - we want to know whether
231 # we have cached the regexp for *this* class, whereas getattr would also
232 # match the superclass
233 if '_VALID_URL_RE' not in cls.__dict__:
234 cls._VALID_URL_RE = re.compile(cls._VALID_URL)
235 return cls._VALID_URL_RE.match(url) is not None
238 def _match_id(cls, url):
239 if '_VALID_URL_RE' not in cls.__dict__:
240 cls._VALID_URL_RE = re.compile(cls._VALID_URL)
241 m = cls._VALID_URL_RE.match(url)
247 """Getter method for _WORKING."""
250 def initialize(self):
251 """Initializes an instance (authentication, etc)."""
253 self._real_initialize()
256 def extract(self, url):
257 """Extracts URL information and returns it in list of dicts."""
259 return self._real_extract(url)
261 def set_downloader(self, downloader):
262 """Sets the downloader for this IE."""
263 self._downloader = downloader
265 def _real_initialize(self):
266 """Real initialization process. Redefine in subclasses."""
269 def _real_extract(self, url):
270 """Real extraction process. Redefine in subclasses."""
275 """A string for getting the InfoExtractor with get_info_extractor"""
276 return cls.__name__[:-2]
280 return type(self).__name__[:-2]
282 def _request_webpage(self, url_or_request, video_id, note=None, errnote=None, fatal=True):
283 """ Returns the response handle """
285 self.report_download_webpage(video_id)
286 elif note is not False:
288 self.to_screen('%s' % (note,))
290 self.to_screen('%s: %s' % (video_id, note))
292 return self._downloader.urlopen(url_or_request)
293 except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
297 errnote = 'Unable to download webpage'
298 errmsg = '%s: %s' % (errnote, compat_str(err))
300 raise ExtractorError(errmsg, sys.exc_info()[2], cause=err)
302 self._downloader.report_warning(errmsg)
305 def _download_webpage_handle(self, url_or_request, video_id, note=None, errnote=None, fatal=True):
306 """ Returns a tuple (page content as string, URL handle) """
307 # Strip hashes from the URL (#1038)
308 if isinstance(url_or_request, (compat_str, str)):
309 url_or_request = url_or_request.partition('#')[0]
311 urlh = self._request_webpage(url_or_request, video_id, note, errnote, fatal)
315 content = self._webpage_read_content(urlh, url_or_request, video_id, note, errnote, fatal)
316 return (content, urlh)
318 def _webpage_read_content(self, urlh, url_or_request, video_id, note=None, errnote=None, fatal=True, prefix=None):
319 content_type = urlh.headers.get('Content-Type', '')
320 webpage_bytes = urlh.read()
321 if prefix is not None:
322 webpage_bytes = prefix + webpage_bytes
323 m = re.match(r'[a-zA-Z0-9_.-]+/[a-zA-Z0-9_.-]+\s*;\s*charset=(.+)', content_type)
325 encoding = m.group(1)
327 m = re.search(br'<meta[^>]+charset=[\'"]?([^\'")]+)[ /\'">]',
328 webpage_bytes[:1024])
330 encoding = m.group(1).decode('ascii')
331 elif webpage_bytes.startswith(b'\xff\xfe'):
335 if self._downloader.params.get('dump_intermediate_pages', False):
337 url = url_or_request.get_full_url()
338 except AttributeError:
340 self.to_screen('Dumping request to ' + url)
341 dump = base64.b64encode(webpage_bytes).decode('ascii')
342 self._downloader.to_screen(dump)
343 if self._downloader.params.get('write_pages', False):
345 url = url_or_request.get_full_url()
346 except AttributeError:
348 basen = '%s_%s' % (video_id, url)
350 h = '___' + hashlib.md5(basen.encode('utf-8')).hexdigest()
351 basen = basen[:240 - len(h)] + h
352 raw_filename = basen + '.dump'
353 filename = sanitize_filename(raw_filename, restricted=True)
354 self.to_screen('Saving request to ' + filename)
355 # Working around MAX_PATH limitation on Windows (see
356 # http://msdn.microsoft.com/en-us/library/windows/desktop/aa365247(v=vs.85).aspx)
358 absfilepath = os.path.abspath(filename)
359 if len(absfilepath) > 259:
360 filename = '\\\\?\\' + absfilepath
361 with open(filename, 'wb') as outf:
362 outf.write(webpage_bytes)
365 content = webpage_bytes.decode(encoding, 'replace')
367 content = webpage_bytes.decode('utf-8', 'replace')
369 if ('<title>Access to this site is blocked</title>' in content and
370 'Websense' in content[:512]):
371 msg = 'Access to this webpage has been blocked by Websense filtering software in your network.'
372 blocked_iframe = self._html_search_regex(
373 r'<iframe src="([^"]+)"', content,
374 'Websense information URL', default=None)
376 msg += ' Visit %s for more details' % blocked_iframe
377 raise ExtractorError(msg, expected=True)
381 def _download_webpage(self, url_or_request, video_id, note=None, errnote=None, fatal=True, tries=1, timeout=5):
382 """ Returns the data of the page as a string """
385 while success is False:
387 res = self._download_webpage_handle(url_or_request, video_id, note, errnote, fatal)
389 except compat_http_client.IncompleteRead as e:
391 if try_count >= tries:
393 self._sleep(timeout, video_id)
400 def _download_xml(self, url_or_request, video_id,
401 note='Downloading XML', errnote='Unable to download XML',
402 transform_source=None, fatal=True):
403 """Return the xml as an xml.etree.ElementTree.Element"""
404 xml_string = self._download_webpage(
405 url_or_request, video_id, note, errnote, fatal=fatal)
406 if xml_string is False:
409 xml_string = transform_source(xml_string)
410 return xml.etree.ElementTree.fromstring(xml_string.encode('utf-8'))
412 def _download_json(self, url_or_request, video_id,
413 note='Downloading JSON metadata',
414 errnote='Unable to download JSON metadata',
415 transform_source=None,
417 json_string = self._download_webpage(
418 url_or_request, video_id, note, errnote, fatal=fatal)
419 if (not fatal) and json_string is False:
421 return self._parse_json(
422 json_string, video_id, transform_source=transform_source, fatal=fatal)
424 def _parse_json(self, json_string, video_id, transform_source=None, fatal=True):
426 json_string = transform_source(json_string)
428 return json.loads(json_string)
429 except ValueError as ve:
430 errmsg = '%s: Failed to parse JSON ' % video_id
432 raise ExtractorError(errmsg, cause=ve)
434 self.report_warning(errmsg + str(ve))
436 def report_warning(self, msg, video_id=None):
437 idstr = '' if video_id is None else '%s: ' % video_id
438 self._downloader.report_warning(
439 '[%s] %s%s' % (self.IE_NAME, idstr, msg))
441 def to_screen(self, msg):
442 """Print msg to screen, prefixing it with '[ie_name]'"""
443 self._downloader.to_screen('[%s] %s' % (self.IE_NAME, msg))
445 def report_extraction(self, id_or_name):
446 """Report information extraction."""
447 self.to_screen('%s: Extracting information' % id_or_name)
449 def report_download_webpage(self, video_id):
450 """Report webpage download."""
451 self.to_screen('%s: Downloading webpage' % video_id)
453 def report_age_confirmation(self):
454 """Report attempt to confirm age."""
455 self.to_screen('Confirming age')
457 def report_login(self):
458 """Report attempt to log in."""
459 self.to_screen('Logging in')
461 # Methods for following #608
463 def url_result(url, ie=None, video_id=None):
464 """Returns a url that points to a page that should be processed"""
465 # TODO: ie should be the class used for getting the info
466 video_info = {'_type': 'url',
469 if video_id is not None:
470 video_info['id'] = video_id
474 def playlist_result(entries, playlist_id=None, playlist_title=None, playlist_description=None):
475 """Returns a playlist"""
476 video_info = {'_type': 'playlist',
479 video_info['id'] = playlist_id
481 video_info['title'] = playlist_title
482 if playlist_description:
483 video_info['description'] = playlist_description
486 def _search_regex(self, pattern, string, name, default=_NO_DEFAULT, fatal=True, flags=0, group=None):
488 Perform a regex search on the given string, using a single or a list of
489 patterns returning the first matching group.
490 In case of failure return a default value or raise a WARNING or a
491 RegexNotFoundError, depending on fatal, specifying the field name.
493 if isinstance(pattern, (str, compat_str, compiled_regex_type)):
494 mobj = re.search(pattern, string, flags)
497 mobj = re.search(p, string, flags)
501 if os.name != 'nt' and sys.stderr.isatty():
502 _name = '\033[0;34m%s\033[0m' % name
508 # return the first matching group
509 return next(g for g in mobj.groups() if g is not None)
511 return mobj.group(group)
512 elif default is not _NO_DEFAULT:
515 raise RegexNotFoundError('Unable to extract %s' % _name)
517 self._downloader.report_warning('unable to extract %s; '
518 'please report this issue on http://yt-dl.org/bug' % _name)
521 def _html_search_regex(self, pattern, string, name, default=_NO_DEFAULT, fatal=True, flags=0, group=None):
523 Like _search_regex, but strips HTML tags and unescapes entities.
525 res = self._search_regex(pattern, string, name, default, fatal, flags, group)
527 return clean_html(res).strip()
531 def _get_login_info(self):
533 Get the the login info as (username, password)
534 It will look in the netrc file using the _NETRC_MACHINE value
535 If there's no info available, return (None, None)
537 if self._downloader is None:
542 downloader_params = self._downloader.params
544 # Attempt to use provided username and password or .netrc data
545 if downloader_params.get('username', None) is not None:
546 username = downloader_params['username']
547 password = downloader_params['password']
548 elif downloader_params.get('usenetrc', False):
550 info = netrc.netrc().authenticators(self._NETRC_MACHINE)
555 raise netrc.NetrcParseError('No authenticators for %s' % self._NETRC_MACHINE)
556 except (IOError, netrc.NetrcParseError) as err:
557 self._downloader.report_warning('parsing .netrc: %s' % compat_str(err))
559 return (username, password)
561 def _get_tfa_info(self):
563 Get the two-factor authentication info
564 TODO - asking the user will be required for sms/phone verify
565 currently just uses the command line option
566 If there's no info available, return None
568 if self._downloader is None:
570 downloader_params = self._downloader.params
572 if downloader_params.get('twofactor', None) is not None:
573 return downloader_params['twofactor']
577 # Helper functions for extracting OpenGraph info
579 def _og_regexes(prop):
580 content_re = r'content=(?:"([^>]+?)"|\'([^>]+?)\')'
581 property_re = r'(?:name|property)=[\'"]og:%s[\'"]' % re.escape(prop)
582 template = r'<meta[^>]+?%s[^>]+?%s'
584 template % (property_re, content_re),
585 template % (content_re, property_re),
588 def _og_search_property(self, prop, html, name=None, **kargs):
590 name = 'OpenGraph %s' % prop
591 escaped = self._search_regex(self._og_regexes(prop), html, name, flags=re.DOTALL, **kargs)
594 return unescapeHTML(escaped)
596 def _og_search_thumbnail(self, html, **kargs):
597 return self._og_search_property('image', html, 'thumbnail url', fatal=False, **kargs)
599 def _og_search_description(self, html, **kargs):
600 return self._og_search_property('description', html, fatal=False, **kargs)
602 def _og_search_title(self, html, **kargs):
603 return self._og_search_property('title', html, **kargs)
605 def _og_search_video_url(self, html, name='video url', secure=True, **kargs):
606 regexes = self._og_regexes('video') + self._og_regexes('video:url')
608 regexes = self._og_regexes('video:secure_url') + regexes
609 return self._html_search_regex(regexes, html, name, **kargs)
611 def _og_search_url(self, html, **kargs):
612 return self._og_search_property('url', html, **kargs)
614 def _html_search_meta(self, name, html, display_name=None, fatal=False, **kwargs):
615 if display_name is None:
617 return self._html_search_regex(
619 (?=[^>]+(?:itemprop|name|property)=(["\']?)%s\1)
620 [^>]+?content=(["\'])(?P<content>.*?)\2''' % re.escape(name),
621 html, display_name, fatal=fatal, group='content', **kwargs)
623 def _dc_search_uploader(self, html):
624 return self._html_search_meta('dc.creator', html, 'uploader')
626 def _rta_search(self, html):
627 # See http://www.rtalabel.org/index.php?content=howtofaq#single
628 if re.search(r'(?ix)<meta\s+name="rating"\s+'
629 r' content="RTA-5042-1996-1400-1577-RTA"',
634 def _media_rating_search(self, html):
635 # See http://www.tjg-designs.com/WP/metadata-code-examples-adding-metadata-to-your-web-pages/
636 rating = self._html_search_meta('rating', html)
648 return RATING_TABLE.get(rating.lower(), None)
650 def _twitter_search_player(self, html):
651 return self._html_search_meta('twitter:player', html,
652 'twitter card player')
654 def _sort_formats(self, formats):
656 raise ExtractorError('No video formats found')
659 # TODO remove the following workaround
660 from ..utils import determine_ext
661 if not f.get('ext') and 'url' in f:
662 f['ext'] = determine_ext(f['url'])
664 preference = f.get('preference')
665 if preference is None:
666 proto = f.get('protocol')
668 proto = compat_urllib_parse_urlparse(f.get('url', '')).scheme
670 preference = 0 if proto in ['http', 'https'] else -0.1
671 if f.get('ext') in ['f4f', 'f4m']: # Not yet supported
674 if f.get('vcodec') == 'none': # audio only
675 if self._downloader.params.get('prefer_free_formats'):
676 ORDER = ['aac', 'mp3', 'm4a', 'webm', 'ogg', 'opus']
678 ORDER = ['webm', 'opus', 'ogg', 'mp3', 'aac', 'm4a']
681 audio_ext_preference = ORDER.index(f['ext'])
683 audio_ext_preference = -1
685 if self._downloader.params.get('prefer_free_formats'):
686 ORDER = ['flv', 'mp4', 'webm']
688 ORDER = ['webm', 'flv', 'mp4']
690 ext_preference = ORDER.index(f['ext'])
693 audio_ext_preference = 0
697 f.get('language_preference') if f.get('language_preference') is not None else -1,
698 f.get('quality') if f.get('quality') is not None else -1,
699 f.get('height') if f.get('height') is not None else -1,
700 f.get('width') if f.get('width') is not None else -1,
702 f.get('tbr') if f.get('tbr') is not None else -1,
703 f.get('vbr') if f.get('vbr') is not None else -1,
704 f.get('abr') if f.get('abr') is not None else -1,
705 audio_ext_preference,
706 f.get('fps') if f.get('fps') is not None else -1,
707 f.get('filesize') if f.get('filesize') is not None else -1,
708 f.get('filesize_approx') if f.get('filesize_approx') is not None else -1,
709 f.get('source_preference') if f.get('source_preference') is not None else -1,
712 formats.sort(key=_formats_key)
714 def http_scheme(self):
715 """ Either "http:" or "https:", depending on the user's preferences """
718 if self._downloader.params.get('prefer_insecure', False)
721 def _proto_relative_url(self, url, scheme=None):
724 if url.startswith('//'):
726 scheme = self.http_scheme()
731 def _sleep(self, timeout, video_id, msg_template=None):
732 if msg_template is None:
733 msg_template = '%(video_id)s: Waiting for %(timeout)s seconds'
734 msg = msg_template % {'video_id': video_id, 'timeout': timeout}
738 def _extract_f4m_formats(self, manifest_url, video_id):
739 manifest = self._download_xml(
740 manifest_url, video_id, 'Downloading f4m manifest',
741 'Unable to download f4m manifest')
744 manifest_version = '1.0'
745 media_nodes = manifest.findall('{http://ns.adobe.com/f4m/1.0}media')
747 manifest_version = '2.0'
748 media_nodes = manifest.findall('{http://ns.adobe.com/f4m/2.0}media')
749 for i, media_el in enumerate(media_nodes):
750 if manifest_version == '2.0':
751 manifest_url = '/'.join(manifest_url.split('/')[:-1]) + '/' + media_el.attrib.get('href')
752 tbr = int_or_none(media_el.attrib.get('bitrate'))
753 format_id = 'f4m-%d' % (i if tbr is None else tbr)
755 'format_id': format_id,
759 'width': int_or_none(media_el.attrib.get('width')),
760 'height': int_or_none(media_el.attrib.get('height')),
762 self._sort_formats(formats)
766 def _extract_m3u8_formats(self, m3u8_url, video_id, ext=None,
767 entry_protocol='m3u8', preference=None):
770 'format_id': 'm3u8-meta',
775 'resolution': 'multiple',
776 'format_note': 'Quality selection URL',
779 format_url = lambda u: (
781 if re.match(r'^https?://', u)
782 else compat_urlparse.urljoin(m3u8_url, u))
784 m3u8_doc = self._download_webpage(
786 note='Downloading m3u8 information',
787 errnote='Failed to download m3u8 information')
790 r'(?P<key>[a-zA-Z_-]+)=(?P<val>"[^"]+"|[^",]+)(?:,|$)')
791 for line in m3u8_doc.splitlines():
792 if line.startswith('#EXT-X-STREAM-INF:'):
794 for m in kv_rex.finditer(line):
796 if v.startswith('"'):
798 last_info[m.group('key')] = v
799 elif line.startswith('#') or not line.strip():
802 if last_info is None:
803 formats.append({'url': format_url(line)})
805 tbr = int_or_none(last_info.get('BANDWIDTH'), scale=1000)
808 'format_id': 'm3u8-%d' % (tbr if tbr else len(formats)),
809 'url': format_url(line.strip()),
812 'protocol': entry_protocol,
813 'preference': preference,
815 codecs = last_info.get('CODECS')
817 # TODO: looks like video codec is not always necessarily goes first
818 va_codecs = codecs.split(',')
820 f['vcodec'] = va_codecs[0].partition('.')[0]
821 if len(va_codecs) > 1 and va_codecs[1]:
822 f['acodec'] = va_codecs[1].partition('.')[0]
823 resolution = last_info.get('RESOLUTION')
825 width_str, height_str = resolution.split('x')
826 f['width'] = int(width_str)
827 f['height'] = int(height_str)
830 self._sort_formats(formats)
833 # TODO: improve extraction
834 def _extract_smil_formats(self, smil_url, video_id):
835 smil = self._download_xml(
836 smil_url, video_id, 'Downloading SMIL file',
837 'Unable to download SMIL file')
839 base = smil.find('./head/meta').get('base')
843 for video in smil.findall('./body/switch/video'):
844 src = video.get('src')
847 bitrate = int_or_none(video.get('system-bitrate') or video.get('systemBitrate'), 1000)
848 width = int_or_none(video.get('width'))
849 height = int_or_none(video.get('height'))
850 proto = video.get('proto')
853 if base.startswith('rtmp'):
855 elif base.startswith('http'):
857 ext = video.get('ext')
859 formats.extend(self._extract_m3u8_formats(src, video_id, ext))
860 elif proto == 'rtmp':
862 streamer = video.get('streamer') or base
867 'format_id': 'rtmp-%d' % (rtmp_count if bitrate is None else bitrate),
872 self._sort_formats(formats)
876 def _live_title(self, name):
877 """ Generate the title for a live video """
878 now = datetime.datetime.now()
879 now_str = now.strftime("%Y-%m-%d %H:%M")
880 return name + ' ' + now_str
882 def _int(self, v, name, fatal=False, **kwargs):
883 res = int_or_none(v, **kwargs)
884 if 'get_attr' in kwargs:
885 print(getattr(v, kwargs['get_attr']))
887 msg = 'Failed to extract %s: Could not parse value %r' % (name, v)
889 raise ExtractorError(msg)
891 self._downloader.report_warning(msg)
894 def _float(self, v, name, fatal=False, **kwargs):
895 res = float_or_none(v, **kwargs)
897 msg = 'Failed to extract %s: Could not parse value %r' % (name, v)
899 raise ExtractorError(msg)
901 self._downloader.report_warning(msg)
904 def _set_cookie(self, domain, name, value, expire_time=None):
905 cookie = compat_cookiejar.Cookie(
906 0, name, value, None, None, domain, None,
907 None, '/', True, False, expire_time, '', None, None, None)
908 self._downloader.cookiejar.set_cookie(cookie)
910 def get_testcases(self, include_onlymatching=False):
911 t = getattr(self, '_TEST', None)
913 assert not hasattr(self, '_TESTS'), \
914 '%s has _TEST and _TESTS' % type(self).__name__
917 tests = getattr(self, '_TESTS', [])
919 if not include_onlymatching and t.get('only_matching', False):
921 t['name'] = type(self).__name__[:-len('IE')]
924 def is_suitable(self, age_limit):
925 """ Test whether the extractor is generally suitable for the given
926 age limit (i.e. pornographic sites are not, all others usually are) """
928 any_restricted = False
929 for tc in self.get_testcases(include_onlymatching=False):
931 tc = tc['playlist'][0]
932 is_restricted = age_restricted(
933 tc.get('info_dict', {}).get('age_limit'), age_limit)
934 if not is_restricted:
936 any_restricted = any_restricted or is_restricted
937 return not any_restricted
940 class SearchInfoExtractor(InfoExtractor):
942 Base class for paged search queries extractors.
943 They accept urls in the format _SEARCH_KEY(|all|[0-9]):{query}
944 Instances should define _SEARCH_KEY and _MAX_RESULTS.
948 def _make_valid_url(cls):
949 return r'%s(?P<prefix>|[1-9][0-9]*|all):(?P<query>[\s\S]+)' % cls._SEARCH_KEY
952 def suitable(cls, url):
953 return re.match(cls._make_valid_url(), url) is not None
955 def _real_extract(self, query):
956 mobj = re.match(self._make_valid_url(), query)
958 raise ExtractorError('Invalid search query "%s"' % query)
960 prefix = mobj.group('prefix')
961 query = mobj.group('query')
963 return self._get_n_results(query, 1)
964 elif prefix == 'all':
965 return self._get_n_results(query, self._MAX_RESULTS)
969 raise ExtractorError('invalid download number %s for query "%s"' % (n, query))
970 elif n > self._MAX_RESULTS:
971 self._downloader.report_warning('%s returns max %i results (you requested %i)' % (self._SEARCH_KEY, self._MAX_RESULTS, n))
972 n = self._MAX_RESULTS
973 return self._get_n_results(query, n)
975 def _get_n_results(self, query, n):
976 """Get a specified number of results for a query"""
977 raise NotImplementedError("This method must be implemented by subclasses")
980 def SEARCH_KEY(self):
981 return self._SEARCH_KEY