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:
132 * "id" (optional, string) - Thumbnail format ID
134 * "preference" (optional, int) - quality of the image
135 * "width" (optional, int)
136 * "height" (optional, int)
137 * "resolution" (optional, string "{width}x{height"},
139 thumbnail: Full URL to a video thumbnail image.
140 description: Full video description.
141 uploader: Full name of the video uploader.
142 timestamp: UNIX timestamp of the moment the video became available.
143 upload_date: Video upload date (YYYYMMDD).
144 If not explicitly set, calculated from timestamp.
145 uploader_id: Nickname or id of the video uploader.
146 location: Physical location where the video was filmed.
147 subtitles: The subtitle file contents as a dictionary in the format
148 {language: subtitles}.
149 duration: Length of the video in seconds, as an integer.
150 view_count: How many users have watched the video on the platform.
151 like_count: Number of positive ratings of the video
152 dislike_count: Number of negative ratings of the video
153 comment_count: Number of comments on the video
154 comments: A list of comments, each with one or more of the following
155 properties (all but one of text or html optional):
156 * "author" - human-readable name of the comment author
157 * "author_id" - user ID of the comment author
159 * "html" - Comment as HTML
160 * "text" - Plain text of the comment
161 * "timestamp" - UNIX timestamp of comment
162 * "parent" - ID of the comment this one is replying to.
163 Set to "root" to indicate that this is a
164 comment to the original video.
165 age_limit: Age restriction for the video, as an integer (years)
166 webpage_url: The url to the video webpage, if given to youtube-dl it
167 should allow to get the same result again. (It will be set
168 by YoutubeDL if it's missing)
169 categories: A list of categories that the video falls in, for example
171 is_live: True, False, or None (=unknown). Whether this video is a
172 live stream that goes on instead of a fixed-length video.
174 Unless mentioned otherwise, the fields should be Unicode strings.
176 Unless mentioned otherwise, None is equivalent to absence of information.
179 _type "playlist" indicates multiple videos.
180 There must be a key "entries", which is a list, an iterable, or a PagedList
181 object, each element of which is a valid dictionary by this specification.
183 Additionally, playlists can have "title" and "id" attributes with the same
184 semantics as videos (see above).
187 _type "multi_video" indicates that there are multiple videos that
188 form a single show, for examples multiple acts of an opera or TV episode.
189 It must have an entries key like a playlist and contain all the keys
190 required for a video at the same time.
193 _type "url" indicates that the video must be extracted from another
194 location, possibly by a different extractor. Its only required key is:
195 "url" - the next URL to extract.
196 The key "ie_key" can be set to the class name (minus the trailing "IE",
197 e.g. "Youtube") if the extractor class is known in advance.
198 Additionally, the dictionary may have any properties of the resolved entity
199 known in advance, for example "title" if the title of the referred video is
203 _type "url_transparent" entities have the same specification as "url", but
204 indicate that the given additional information is more precise than the one
205 associated with the resolved URL.
206 This is useful when a site employs a video service that hosts the video and
207 its technical metadata, but that video service does not embed a useful
208 title, description etc.
211 Subclasses of this one should re-define the _real_initialize() and
212 _real_extract() methods and define a _VALID_URL regexp.
213 Probably, they should also be added to the list of extractors.
215 Finally, the _WORKING attribute should be set to False for broken IEs
216 in order to warn the users and skip the tests.
223 def __init__(self, downloader=None):
224 """Constructor. Receives an optional downloader."""
226 self.set_downloader(downloader)
229 def suitable(cls, url):
230 """Receives a URL and returns True if suitable for this IE."""
232 # This does not use has/getattr intentionally - we want to know whether
233 # we have cached the regexp for *this* class, whereas getattr would also
234 # match the superclass
235 if '_VALID_URL_RE' not in cls.__dict__:
236 cls._VALID_URL_RE = re.compile(cls._VALID_URL)
237 return cls._VALID_URL_RE.match(url) is not None
240 def _match_id(cls, url):
241 if '_VALID_URL_RE' not in cls.__dict__:
242 cls._VALID_URL_RE = re.compile(cls._VALID_URL)
243 m = cls._VALID_URL_RE.match(url)
249 """Getter method for _WORKING."""
252 def initialize(self):
253 """Initializes an instance (authentication, etc)."""
255 self._real_initialize()
258 def extract(self, url):
259 """Extracts URL information and returns it in list of dicts."""
261 return self._real_extract(url)
263 def set_downloader(self, downloader):
264 """Sets the downloader for this IE."""
265 self._downloader = downloader
267 def _real_initialize(self):
268 """Real initialization process. Redefine in subclasses."""
271 def _real_extract(self, url):
272 """Real extraction process. Redefine in subclasses."""
277 """A string for getting the InfoExtractor with get_info_extractor"""
278 return cls.__name__[:-2]
282 return type(self).__name__[:-2]
284 def _request_webpage(self, url_or_request, video_id, note=None, errnote=None, fatal=True):
285 """ Returns the response handle """
287 self.report_download_webpage(video_id)
288 elif note is not False:
290 self.to_screen('%s' % (note,))
292 self.to_screen('%s: %s' % (video_id, note))
294 return self._downloader.urlopen(url_or_request)
295 except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
299 errnote = 'Unable to download webpage'
300 errmsg = '%s: %s' % (errnote, compat_str(err))
302 raise ExtractorError(errmsg, sys.exc_info()[2], cause=err)
304 self._downloader.report_warning(errmsg)
307 def _download_webpage_handle(self, url_or_request, video_id, note=None, errnote=None, fatal=True):
308 """ Returns a tuple (page content as string, URL handle) """
309 # Strip hashes from the URL (#1038)
310 if isinstance(url_or_request, (compat_str, str)):
311 url_or_request = url_or_request.partition('#')[0]
313 urlh = self._request_webpage(url_or_request, video_id, note, errnote, fatal)
317 content = self._webpage_read_content(urlh, url_or_request, video_id, note, errnote, fatal)
318 return (content, urlh)
320 def _webpage_read_content(self, urlh, url_or_request, video_id, note=None, errnote=None, fatal=True, prefix=None):
321 content_type = urlh.headers.get('Content-Type', '')
322 webpage_bytes = urlh.read()
323 if prefix is not None:
324 webpage_bytes = prefix + webpage_bytes
325 m = re.match(r'[a-zA-Z0-9_.-]+/[a-zA-Z0-9_.-]+\s*;\s*charset=(.+)', content_type)
327 encoding = m.group(1)
329 m = re.search(br'<meta[^>]+charset=[\'"]?([^\'")]+)[ /\'">]',
330 webpage_bytes[:1024])
332 encoding = m.group(1).decode('ascii')
333 elif webpage_bytes.startswith(b'\xff\xfe'):
337 if self._downloader.params.get('dump_intermediate_pages', False):
339 url = url_or_request.get_full_url()
340 except AttributeError:
342 self.to_screen('Dumping request to ' + url)
343 dump = base64.b64encode(webpage_bytes).decode('ascii')
344 self._downloader.to_screen(dump)
345 if self._downloader.params.get('write_pages', False):
347 url = url_or_request.get_full_url()
348 except AttributeError:
350 basen = '%s_%s' % (video_id, url)
352 h = '___' + hashlib.md5(basen.encode('utf-8')).hexdigest()
353 basen = basen[:240 - len(h)] + h
354 raw_filename = basen + '.dump'
355 filename = sanitize_filename(raw_filename, restricted=True)
356 self.to_screen('Saving request to ' + filename)
357 # Working around MAX_PATH limitation on Windows (see
358 # http://msdn.microsoft.com/en-us/library/windows/desktop/aa365247(v=vs.85).aspx)
360 absfilepath = os.path.abspath(filename)
361 if len(absfilepath) > 259:
362 filename = '\\\\?\\' + absfilepath
363 with open(filename, 'wb') as outf:
364 outf.write(webpage_bytes)
367 content = webpage_bytes.decode(encoding, 'replace')
369 content = webpage_bytes.decode('utf-8', 'replace')
371 if ('<title>Access to this site is blocked</title>' in content and
372 'Websense' in content[:512]):
373 msg = 'Access to this webpage has been blocked by Websense filtering software in your network.'
374 blocked_iframe = self._html_search_regex(
375 r'<iframe src="([^"]+)"', content,
376 'Websense information URL', default=None)
378 msg += ' Visit %s for more details' % blocked_iframe
379 raise ExtractorError(msg, expected=True)
383 def _download_webpage(self, url_or_request, video_id, note=None, errnote=None, fatal=True, tries=1, timeout=5):
384 """ Returns the data of the page as a string """
387 while success is False:
389 res = self._download_webpage_handle(url_or_request, video_id, note, errnote, fatal)
391 except compat_http_client.IncompleteRead as e:
393 if try_count >= tries:
395 self._sleep(timeout, video_id)
402 def _download_xml(self, url_or_request, video_id,
403 note='Downloading XML', errnote='Unable to download XML',
404 transform_source=None, fatal=True):
405 """Return the xml as an xml.etree.ElementTree.Element"""
406 xml_string = self._download_webpage(
407 url_or_request, video_id, note, errnote, fatal=fatal)
408 if xml_string is False:
411 xml_string = transform_source(xml_string)
412 return xml.etree.ElementTree.fromstring(xml_string.encode('utf-8'))
414 def _download_json(self, url_or_request, video_id,
415 note='Downloading JSON metadata',
416 errnote='Unable to download JSON metadata',
417 transform_source=None,
419 json_string = self._download_webpage(
420 url_or_request, video_id, note, errnote, fatal=fatal)
421 if (not fatal) and json_string is False:
423 return self._parse_json(
424 json_string, video_id, transform_source=transform_source, fatal=fatal)
426 def _parse_json(self, json_string, video_id, transform_source=None, fatal=True):
428 json_string = transform_source(json_string)
430 return json.loads(json_string)
431 except ValueError as ve:
432 errmsg = '%s: Failed to parse JSON ' % video_id
434 raise ExtractorError(errmsg, cause=ve)
436 self.report_warning(errmsg + str(ve))
438 def report_warning(self, msg, video_id=None):
439 idstr = '' if video_id is None else '%s: ' % video_id
440 self._downloader.report_warning(
441 '[%s] %s%s' % (self.IE_NAME, idstr, msg))
443 def to_screen(self, msg):
444 """Print msg to screen, prefixing it with '[ie_name]'"""
445 self._downloader.to_screen('[%s] %s' % (self.IE_NAME, msg))
447 def report_extraction(self, id_or_name):
448 """Report information extraction."""
449 self.to_screen('%s: Extracting information' % id_or_name)
451 def report_download_webpage(self, video_id):
452 """Report webpage download."""
453 self.to_screen('%s: Downloading webpage' % video_id)
455 def report_age_confirmation(self):
456 """Report attempt to confirm age."""
457 self.to_screen('Confirming age')
459 def report_login(self):
460 """Report attempt to log in."""
461 self.to_screen('Logging in')
463 # Methods for following #608
465 def url_result(url, ie=None, video_id=None):
466 """Returns a url that points to a page that should be processed"""
467 # TODO: ie should be the class used for getting the info
468 video_info = {'_type': 'url',
471 if video_id is not None:
472 video_info['id'] = video_id
476 def playlist_result(entries, playlist_id=None, playlist_title=None, playlist_description=None):
477 """Returns a playlist"""
478 video_info = {'_type': 'playlist',
481 video_info['id'] = playlist_id
483 video_info['title'] = playlist_title
484 if playlist_description:
485 video_info['description'] = playlist_description
488 def _search_regex(self, pattern, string, name, default=_NO_DEFAULT, fatal=True, flags=0, group=None):
490 Perform a regex search on the given string, using a single or a list of
491 patterns returning the first matching group.
492 In case of failure return a default value or raise a WARNING or a
493 RegexNotFoundError, depending on fatal, specifying the field name.
495 if isinstance(pattern, (str, compat_str, compiled_regex_type)):
496 mobj = re.search(pattern, string, flags)
499 mobj = re.search(p, string, flags)
503 if os.name != 'nt' and sys.stderr.isatty():
504 _name = '\033[0;34m%s\033[0m' % name
510 # return the first matching group
511 return next(g for g in mobj.groups() if g is not None)
513 return mobj.group(group)
514 elif default is not _NO_DEFAULT:
517 raise RegexNotFoundError('Unable to extract %s' % _name)
519 self._downloader.report_warning('unable to extract %s; '
520 'please report this issue on http://yt-dl.org/bug' % _name)
523 def _html_search_regex(self, pattern, string, name, default=_NO_DEFAULT, fatal=True, flags=0, group=None):
525 Like _search_regex, but strips HTML tags and unescapes entities.
527 res = self._search_regex(pattern, string, name, default, fatal, flags, group)
529 return clean_html(res).strip()
533 def _get_login_info(self):
535 Get the the login info as (username, password)
536 It will look in the netrc file using the _NETRC_MACHINE value
537 If there's no info available, return (None, None)
539 if self._downloader is None:
544 downloader_params = self._downloader.params
546 # Attempt to use provided username and password or .netrc data
547 if downloader_params.get('username', None) is not None:
548 username = downloader_params['username']
549 password = downloader_params['password']
550 elif downloader_params.get('usenetrc', False):
552 info = netrc.netrc().authenticators(self._NETRC_MACHINE)
557 raise netrc.NetrcParseError('No authenticators for %s' % self._NETRC_MACHINE)
558 except (IOError, netrc.NetrcParseError) as err:
559 self._downloader.report_warning('parsing .netrc: %s' % compat_str(err))
561 return (username, password)
563 def _get_tfa_info(self):
565 Get the two-factor authentication info
566 TODO - asking the user will be required for sms/phone verify
567 currently just uses the command line option
568 If there's no info available, return None
570 if self._downloader is None:
572 downloader_params = self._downloader.params
574 if downloader_params.get('twofactor', None) is not None:
575 return downloader_params['twofactor']
579 # Helper functions for extracting OpenGraph info
581 def _og_regexes(prop):
582 content_re = r'content=(?:"([^>]+?)"|\'([^>]+?)\')'
583 property_re = r'(?:name|property)=[\'"]og:%s[\'"]' % re.escape(prop)
584 template = r'<meta[^>]+?%s[^>]+?%s'
586 template % (property_re, content_re),
587 template % (content_re, property_re),
590 def _og_search_property(self, prop, html, name=None, **kargs):
592 name = 'OpenGraph %s' % prop
593 escaped = self._search_regex(self._og_regexes(prop), html, name, flags=re.DOTALL, **kargs)
596 return unescapeHTML(escaped)
598 def _og_search_thumbnail(self, html, **kargs):
599 return self._og_search_property('image', html, 'thumbnail url', fatal=False, **kargs)
601 def _og_search_description(self, html, **kargs):
602 return self._og_search_property('description', html, fatal=False, **kargs)
604 def _og_search_title(self, html, **kargs):
605 return self._og_search_property('title', html, **kargs)
607 def _og_search_video_url(self, html, name='video url', secure=True, **kargs):
608 regexes = self._og_regexes('video') + self._og_regexes('video:url')
610 regexes = self._og_regexes('video:secure_url') + regexes
611 return self._html_search_regex(regexes, html, name, **kargs)
613 def _og_search_url(self, html, **kargs):
614 return self._og_search_property('url', html, **kargs)
616 def _html_search_meta(self, name, html, display_name=None, fatal=False, **kwargs):
617 if display_name is None:
619 return self._html_search_regex(
621 (?=[^>]+(?:itemprop|name|property)=(["\']?)%s\1)
622 [^>]+?content=(["\'])(?P<content>.*?)\2''' % re.escape(name),
623 html, display_name, fatal=fatal, group='content', **kwargs)
625 def _dc_search_uploader(self, html):
626 return self._html_search_meta('dc.creator', html, 'uploader')
628 def _rta_search(self, html):
629 # See http://www.rtalabel.org/index.php?content=howtofaq#single
630 if re.search(r'(?ix)<meta\s+name="rating"\s+'
631 r' content="RTA-5042-1996-1400-1577-RTA"',
636 def _media_rating_search(self, html):
637 # See http://www.tjg-designs.com/WP/metadata-code-examples-adding-metadata-to-your-web-pages/
638 rating = self._html_search_meta('rating', html)
650 return RATING_TABLE.get(rating.lower(), None)
652 def _twitter_search_player(self, html):
653 return self._html_search_meta('twitter:player', html,
654 'twitter card player')
656 def _sort_formats(self, formats):
658 raise ExtractorError('No video formats found')
661 # TODO remove the following workaround
662 from ..utils import determine_ext
663 if not f.get('ext') and 'url' in f:
664 f['ext'] = determine_ext(f['url'])
666 preference = f.get('preference')
667 if preference is None:
668 proto = f.get('protocol')
670 proto = compat_urllib_parse_urlparse(f.get('url', '')).scheme
672 preference = 0 if proto in ['http', 'https'] else -0.1
673 if f.get('ext') in ['f4f', 'f4m']: # Not yet supported
676 if f.get('vcodec') == 'none': # audio only
677 if self._downloader.params.get('prefer_free_formats'):
678 ORDER = ['aac', 'mp3', 'm4a', 'webm', 'ogg', 'opus']
680 ORDER = ['webm', 'opus', 'ogg', 'mp3', 'aac', 'm4a']
683 audio_ext_preference = ORDER.index(f['ext'])
685 audio_ext_preference = -1
687 if self._downloader.params.get('prefer_free_formats'):
688 ORDER = ['flv', 'mp4', 'webm']
690 ORDER = ['webm', 'flv', 'mp4']
692 ext_preference = ORDER.index(f['ext'])
695 audio_ext_preference = 0
699 f.get('language_preference') if f.get('language_preference') is not None else -1,
700 f.get('quality') if f.get('quality') is not None else -1,
701 f.get('height') if f.get('height') is not None else -1,
702 f.get('width') if f.get('width') is not None else -1,
704 f.get('tbr') if f.get('tbr') is not None else -1,
705 f.get('vbr') if f.get('vbr') is not None else -1,
706 f.get('abr') if f.get('abr') is not None else -1,
707 audio_ext_preference,
708 f.get('fps') if f.get('fps') is not None else -1,
709 f.get('filesize') if f.get('filesize') is not None else -1,
710 f.get('filesize_approx') if f.get('filesize_approx') is not None else -1,
711 f.get('source_preference') if f.get('source_preference') is not None else -1,
714 formats.sort(key=_formats_key)
716 def http_scheme(self):
717 """ Either "http:" or "https:", depending on the user's preferences """
720 if self._downloader.params.get('prefer_insecure', False)
723 def _proto_relative_url(self, url, scheme=None):
726 if url.startswith('//'):
728 scheme = self.http_scheme()
733 def _sleep(self, timeout, video_id, msg_template=None):
734 if msg_template is None:
735 msg_template = '%(video_id)s: Waiting for %(timeout)s seconds'
736 msg = msg_template % {'video_id': video_id, 'timeout': timeout}
740 def _extract_f4m_formats(self, manifest_url, video_id):
741 manifest = self._download_xml(
742 manifest_url, video_id, 'Downloading f4m manifest',
743 'Unable to download f4m manifest')
746 manifest_version = '1.0'
747 media_nodes = manifest.findall('{http://ns.adobe.com/f4m/1.0}media')
749 manifest_version = '2.0'
750 media_nodes = manifest.findall('{http://ns.adobe.com/f4m/2.0}media')
751 for i, media_el in enumerate(media_nodes):
752 if manifest_version == '2.0':
753 manifest_url = '/'.join(manifest_url.split('/')[:-1]) + '/' + media_el.attrib.get('href')
754 tbr = int_or_none(media_el.attrib.get('bitrate'))
755 format_id = 'f4m-%d' % (i if tbr is None else tbr)
757 'format_id': format_id,
761 'width': int_or_none(media_el.attrib.get('width')),
762 'height': int_or_none(media_el.attrib.get('height')),
764 self._sort_formats(formats)
768 def _extract_m3u8_formats(self, m3u8_url, video_id, ext=None,
769 entry_protocol='m3u8', preference=None):
772 'format_id': 'm3u8-meta',
777 'resolution': 'multiple',
778 'format_note': 'Quality selection URL',
781 format_url = lambda u: (
783 if re.match(r'^https?://', u)
784 else compat_urlparse.urljoin(m3u8_url, u))
786 m3u8_doc = self._download_webpage(
788 note='Downloading m3u8 information',
789 errnote='Failed to download m3u8 information')
792 r'(?P<key>[a-zA-Z_-]+)=(?P<val>"[^"]+"|[^",]+)(?:,|$)')
793 for line in m3u8_doc.splitlines():
794 if line.startswith('#EXT-X-STREAM-INF:'):
796 for m in kv_rex.finditer(line):
798 if v.startswith('"'):
800 last_info[m.group('key')] = v
801 elif line.startswith('#') or not line.strip():
804 if last_info is None:
805 formats.append({'url': format_url(line)})
807 tbr = int_or_none(last_info.get('BANDWIDTH'), scale=1000)
810 'format_id': 'm3u8-%d' % (tbr if tbr else len(formats)),
811 'url': format_url(line.strip()),
814 'protocol': entry_protocol,
815 'preference': preference,
817 codecs = last_info.get('CODECS')
819 # TODO: looks like video codec is not always necessarily goes first
820 va_codecs = codecs.split(',')
822 f['vcodec'] = va_codecs[0].partition('.')[0]
823 if len(va_codecs) > 1 and va_codecs[1]:
824 f['acodec'] = va_codecs[1].partition('.')[0]
825 resolution = last_info.get('RESOLUTION')
827 width_str, height_str = resolution.split('x')
828 f['width'] = int(width_str)
829 f['height'] = int(height_str)
832 self._sort_formats(formats)
835 # TODO: improve extraction
836 def _extract_smil_formats(self, smil_url, video_id):
837 smil = self._download_xml(
838 smil_url, video_id, 'Downloading SMIL file',
839 'Unable to download SMIL file')
841 base = smil.find('./head/meta').get('base')
845 for video in smil.findall('./body/switch/video'):
846 src = video.get('src')
849 bitrate = int_or_none(video.get('system-bitrate') or video.get('systemBitrate'), 1000)
850 width = int_or_none(video.get('width'))
851 height = int_or_none(video.get('height'))
852 proto = video.get('proto')
855 if base.startswith('rtmp'):
857 elif base.startswith('http'):
859 ext = video.get('ext')
861 formats.extend(self._extract_m3u8_formats(src, video_id, ext))
862 elif proto == 'rtmp':
864 streamer = video.get('streamer') or base
869 'format_id': 'rtmp-%d' % (rtmp_count if bitrate is None else bitrate),
874 self._sort_formats(formats)
878 def _live_title(self, name):
879 """ Generate the title for a live video """
880 now = datetime.datetime.now()
881 now_str = now.strftime("%Y-%m-%d %H:%M")
882 return name + ' ' + now_str
884 def _int(self, v, name, fatal=False, **kwargs):
885 res = int_or_none(v, **kwargs)
886 if 'get_attr' in kwargs:
887 print(getattr(v, kwargs['get_attr']))
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 _float(self, v, name, fatal=False, **kwargs):
897 res = float_or_none(v, **kwargs)
899 msg = 'Failed to extract %s: Could not parse value %r' % (name, v)
901 raise ExtractorError(msg)
903 self._downloader.report_warning(msg)
906 def _set_cookie(self, domain, name, value, expire_time=None):
907 cookie = compat_cookiejar.Cookie(
908 0, name, value, None, None, domain, None,
909 None, '/', True, False, expire_time, '', None, None, None)
910 self._downloader.cookiejar.set_cookie(cookie)
912 def get_testcases(self, include_onlymatching=False):
913 t = getattr(self, '_TEST', None)
915 assert not hasattr(self, '_TESTS'), \
916 '%s has _TEST and _TESTS' % type(self).__name__
919 tests = getattr(self, '_TESTS', [])
921 if not include_onlymatching and t.get('only_matching', False):
923 t['name'] = type(self).__name__[:-len('IE')]
926 def is_suitable(self, age_limit):
927 """ Test whether the extractor is generally suitable for the given
928 age limit (i.e. pornographic sites are not, all others usually are) """
930 any_restricted = False
931 for tc in self.get_testcases(include_onlymatching=False):
933 tc = tc['playlist'][0]
934 is_restricted = age_restricted(
935 tc.get('info_dict', {}).get('age_limit'), age_limit)
936 if not is_restricted:
938 any_restricted = any_restricted or is_restricted
939 return not any_restricted
942 class SearchInfoExtractor(InfoExtractor):
944 Base class for paged search queries extractors.
945 They accept urls in the format _SEARCH_KEY(|all|[0-9]):{query}
946 Instances should define _SEARCH_KEY and _MAX_RESULTS.
950 def _make_valid_url(cls):
951 return r'%s(?P<prefix>|[1-9][0-9]*|all):(?P<query>[\s\S]+)' % cls._SEARCH_KEY
954 def suitable(cls, url):
955 return re.match(cls._make_valid_url(), url) is not None
957 def _real_extract(self, query):
958 mobj = re.match(self._make_valid_url(), query)
960 raise ExtractorError('Invalid search query "%s"' % query)
962 prefix = mobj.group('prefix')
963 query = mobj.group('query')
965 return self._get_n_results(query, 1)
966 elif prefix == 'all':
967 return self._get_n_results(query, self._MAX_RESULTS)
971 raise ExtractorError('invalid download number %s for query "%s"' % (n, query))
972 elif n > self._MAX_RESULTS:
973 self._downloader.report_warning('%s returns max %i results (you requested %i)' % (self._SEARCH_KEY, self._MAX_RESULTS, n))
974 n = self._MAX_RESULTS
975 return self._get_n_results(query, n)
977 def _get_n_results(self, query, n):
978 """Get a specified number of results for a query"""
979 raise NotImplementedError("This method must be implemented by subclasses")
982 def SEARCH_KEY(self):
983 return self._SEARCH_KEY