1 from __future__ import unicode_literals
13 import xml.etree.ElementTree
15 from ..compat import (
18 compat_urllib_parse_urlparse,
32 _NO_DEFAULT = object()
35 class InfoExtractor(object):
36 """Information Extractor class.
38 Information extractors are the classes that, given a URL, extract
39 information about the video (or videos) the URL refers to. This
40 information includes the real video URL, the video title, author and
41 others. The information is stored in a dictionary which is then
42 passed to the FileDownloader. The FileDownloader processes this
43 information possibly downloading the video to the file system, among
44 other possible outcomes.
46 The type field determines the the type of the result.
47 By far the most common value (and the default if _type is missing) is
48 "video", which indicates a single video.
50 For a video, the dictionaries must include the following fields:
53 title: Video title, unescaped.
55 Additionally, it must contain either a formats entry or a url one:
57 formats: A list of dictionaries for each format available, ordered
58 from worst to best quality.
61 * url Mandatory. The URL of the video file
62 * ext Will be calculated from url if missing
63 * format A human-readable description of the format
64 ("mp4 container with h264/opus").
65 Calculated from the format_id, width, height.
66 and format_note fields if missing.
67 * format_id A short description of the format
68 ("mp4_h264_opus" or "19").
69 Technically optional, but strongly recommended.
70 * format_note Additional info about the format
71 ("3D" or "DASH video")
72 * width Width of the video, if known
73 * height Height of the video, if known
74 * resolution Textual description of width and height
75 * tbr Average bitrate of audio and video in KBit/s
76 * abr Average audio bitrate in KBit/s
77 * acodec Name of the audio codec in use
78 * asr Audio sampling rate in Hertz
79 * vbr Average video bitrate in KBit/s
81 * vcodec Name of the video codec in use
82 * container Name of the container format
83 * filesize The number of bytes, if known in advance
84 * filesize_approx An estimate for the number of bytes
85 * player_url SWF Player URL (used for rtmpdump).
86 * protocol The protocol that will be used for the actual
88 "http", "https", "rtsp", "rtmp", "m3u8" or so.
89 * preference Order number of this format. If this field is
90 present and not None, the formats get sorted
91 by this field, regardless of all other values.
92 -1 for default (order by other properties),
93 -2 or smaller for less than default.
94 * language_preference Is this in the correct requested
96 10 if it's what the URL is about,
97 -1 for default (don't know),
98 -10 otherwise, other values reserved for now.
99 * quality Order number of the video quality of this
100 format, irrespective of the file format.
101 -1 for default (order by other properties),
102 -2 or smaller for less than default.
103 * source_preference Order number for this video source
104 (quality takes higher priority)
105 -1 for default (order by other properties),
106 -2 or smaller for less than default.
107 * http_referer HTTP Referer header value to set.
108 * http_method HTTP method to use for the download.
109 * http_headers A dictionary of additional HTTP headers
110 to add to the request.
111 * http_post_data Additional data to send with a POST
113 url: Final video URL.
114 ext: Video filename extension.
115 format: The video format, defaults to ext (used for --get-format)
116 player_url: SWF Player URL (used for rtmpdump).
118 The following fields are optional:
120 display_id An alternative identifier for the video, not necessarily
121 unique, but available before title. Typically, id is
122 something like "4234987", title "Dancing naked mole rats",
123 and display_id "dancing-naked-mole-rats"
124 thumbnails: A list of dictionaries, with the following entries:
126 * "width" (optional, int)
127 * "height" (optional, int)
128 * "resolution" (optional, string "{width}x{height"},
130 thumbnail: Full URL to a video thumbnail image.
131 description: One-line video description.
132 uploader: Full name of the video uploader.
133 timestamp: UNIX timestamp of the moment the video became available.
134 upload_date: Video upload date (YYYYMMDD).
135 If not explicitly set, calculated from timestamp.
136 uploader_id: Nickname or id of the video uploader.
137 location: Physical location where the video was filmed.
138 subtitles: The subtitle file contents as a dictionary in the format
139 {language: subtitles}.
140 duration: Length of the video in seconds, as an integer.
141 view_count: How many users have watched the video on the platform.
142 like_count: Number of positive ratings of the video
143 dislike_count: Number of negative ratings of the video
144 comment_count: Number of comments on the video
145 age_limit: Age restriction for the video, as an integer (years)
146 webpage_url: The url to the video webpage, if given to youtube-dl it
147 should allow to get the same result again. (It will be set
148 by YoutubeDL if it's missing)
149 categories: A list of categories that the video falls in, for example
151 is_live: True, False, or None (=unknown). Whether this video is a
152 live stream that goes on instead of a fixed-length video.
154 Unless mentioned otherwise, the fields should be Unicode strings.
156 Unless mentioned otherwise, None is equivalent to absence of information.
159 _type "playlist" indicates multiple videos.
160 There must be a key "entries", which is a list or a PagedList object, each
161 element of which is a valid dictionary under this specfication.
163 Additionally, playlists can have "title" and "id" attributes with the same
164 semantics as videos (see above).
167 _type "multi_video" indicates that there are multiple videos that
168 form a single show, for examples multiple acts of an opera or TV episode.
169 It must have an entries key like a playlist and contain all the keys
170 required for a video at the same time.
173 _type "url" indicates that the video must be extracted from another
174 location, possibly by a different extractor. Its only required key is:
175 "url" - the next URL to extract.
177 Additionally, it may have properties believed to be identical to the
178 resolved entity, for example "title" if the title of the referred video is
182 _type "url_transparent" entities have the same specification as "url", but
183 indicate that the given additional information is more precise than the one
184 associated with the resolved URL.
185 This is useful when a site employs a video service that hosts the video and
186 its technical metadata, but that video service does not embed a useful
187 title, description etc.
190 Subclasses of this one should re-define the _real_initialize() and
191 _real_extract() methods and define a _VALID_URL regexp.
192 Probably, they should also be added to the list of extractors.
194 Finally, the _WORKING attribute should be set to False for broken IEs
195 in order to warn the users and skip the tests.
202 def __init__(self, downloader=None):
203 """Constructor. Receives an optional downloader."""
205 self.set_downloader(downloader)
208 def suitable(cls, url):
209 """Receives a URL and returns True if suitable for this IE."""
211 # This does not use has/getattr intentionally - we want to know whether
212 # we have cached the regexp for *this* class, whereas getattr would also
213 # match the superclass
214 if '_VALID_URL_RE' not in cls.__dict__:
215 cls._VALID_URL_RE = re.compile(cls._VALID_URL)
216 return cls._VALID_URL_RE.match(url) is not None
219 def _match_id(cls, url):
220 if '_VALID_URL_RE' not in cls.__dict__:
221 cls._VALID_URL_RE = re.compile(cls._VALID_URL)
222 m = cls._VALID_URL_RE.match(url)
228 """Getter method for _WORKING."""
231 def initialize(self):
232 """Initializes an instance (authentication, etc)."""
234 self._real_initialize()
237 def extract(self, url):
238 """Extracts URL information and returns it in list of dicts."""
240 return self._real_extract(url)
242 def set_downloader(self, downloader):
243 """Sets the downloader for this IE."""
244 self._downloader = downloader
246 def _real_initialize(self):
247 """Real initialization process. Redefine in subclasses."""
250 def _real_extract(self, url):
251 """Real extraction process. Redefine in subclasses."""
256 """A string for getting the InfoExtractor with get_info_extractor"""
257 return cls.__name__[:-2]
261 return type(self).__name__[:-2]
263 def _request_webpage(self, url_or_request, video_id, note=None, errnote=None, fatal=True):
264 """ Returns the response handle """
266 self.report_download_webpage(video_id)
267 elif note is not False:
269 self.to_screen('%s' % (note,))
271 self.to_screen('%s: %s' % (video_id, note))
273 return self._downloader.urlopen(url_or_request)
274 except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
278 errnote = 'Unable to download webpage'
279 errmsg = '%s: %s' % (errnote, compat_str(err))
281 raise ExtractorError(errmsg, sys.exc_info()[2], cause=err)
283 self._downloader.report_warning(errmsg)
286 def _download_webpage_handle(self, url_or_request, video_id, note=None, errnote=None, fatal=True):
287 """ Returns a tuple (page content as string, URL handle) """
288 # Strip hashes from the URL (#1038)
289 if isinstance(url_or_request, (compat_str, str)):
290 url_or_request = url_or_request.partition('#')[0]
292 urlh = self._request_webpage(url_or_request, video_id, note, errnote, fatal)
296 content = self._webpage_read_content(urlh, url_or_request, video_id, note, errnote, fatal)
297 return (content, urlh)
299 def _webpage_read_content(self, urlh, url_or_request, video_id, note=None, errnote=None, fatal=True, prefix=None):
300 content_type = urlh.headers.get('Content-Type', '')
301 webpage_bytes = urlh.read()
302 if prefix is not None:
303 webpage_bytes = prefix + webpage_bytes
304 m = re.match(r'[a-zA-Z0-9_.-]+/[a-zA-Z0-9_.-]+\s*;\s*charset=(.+)', content_type)
306 encoding = m.group(1)
308 m = re.search(br'<meta[^>]+charset=[\'"]?([^\'")]+)[ /\'">]',
309 webpage_bytes[:1024])
311 encoding = m.group(1).decode('ascii')
312 elif webpage_bytes.startswith(b'\xff\xfe'):
316 if self._downloader.params.get('dump_intermediate_pages', False):
318 url = url_or_request.get_full_url()
319 except AttributeError:
321 self.to_screen('Dumping request to ' + url)
322 dump = base64.b64encode(webpage_bytes).decode('ascii')
323 self._downloader.to_screen(dump)
324 if self._downloader.params.get('write_pages', False):
326 url = url_or_request.get_full_url()
327 except AttributeError:
329 basen = '%s_%s' % (video_id, url)
331 h = '___' + hashlib.md5(basen.encode('utf-8')).hexdigest()
332 basen = basen[:240 - len(h)] + h
333 raw_filename = basen + '.dump'
334 filename = sanitize_filename(raw_filename, restricted=True)
335 self.to_screen('Saving request to ' + filename)
336 # Working around MAX_PATH limitation on Windows (see
337 # http://msdn.microsoft.com/en-us/library/windows/desktop/aa365247(v=vs.85).aspx)
339 absfilepath = os.path.abspath(filename)
340 if len(absfilepath) > 259:
341 filename = '\\\\?\\' + absfilepath
342 with open(filename, 'wb') as outf:
343 outf.write(webpage_bytes)
346 content = webpage_bytes.decode(encoding, 'replace')
348 content = webpage_bytes.decode('utf-8', 'replace')
350 if ('<title>Access to this site is blocked</title>' in content and
351 'Websense' in content[:512]):
352 msg = 'Access to this webpage has been blocked by Websense filtering software in your network.'
353 blocked_iframe = self._html_search_regex(
354 r'<iframe src="([^"]+)"', content,
355 'Websense information URL', default=None)
357 msg += ' Visit %s for more details' % blocked_iframe
358 raise ExtractorError(msg, expected=True)
362 def _download_webpage(self, url_or_request, video_id, note=None, errnote=None, fatal=True):
363 """ Returns the data of the page as a string """
364 res = self._download_webpage_handle(url_or_request, video_id, note, errnote, fatal)
371 def _download_xml(self, url_or_request, video_id,
372 note='Downloading XML', errnote='Unable to download XML',
373 transform_source=None, fatal=True):
374 """Return the xml as an xml.etree.ElementTree.Element"""
375 xml_string = self._download_webpage(
376 url_or_request, video_id, note, errnote, fatal=fatal)
377 if xml_string is False:
380 xml_string = transform_source(xml_string)
381 return xml.etree.ElementTree.fromstring(xml_string.encode('utf-8'))
383 def _download_json(self, url_or_request, video_id,
384 note='Downloading JSON metadata',
385 errnote='Unable to download JSON metadata',
386 transform_source=None,
388 json_string = self._download_webpage(
389 url_or_request, video_id, note, errnote, fatal=fatal)
390 if (not fatal) and json_string is False:
393 json_string = transform_source(json_string)
395 return json.loads(json_string)
396 except ValueError as ve:
397 errmsg = '%s: Failed to parse JSON ' % video_id
399 raise ExtractorError(errmsg, cause=ve)
401 self.report_warning(errmsg + str(ve))
403 def report_warning(self, msg, video_id=None):
404 idstr = '' if video_id is None else '%s: ' % video_id
405 self._downloader.report_warning(
406 '[%s] %s%s' % (self.IE_NAME, idstr, msg))
408 def to_screen(self, msg):
409 """Print msg to screen, prefixing it with '[ie_name]'"""
410 self._downloader.to_screen('[%s] %s' % (self.IE_NAME, msg))
412 def report_extraction(self, id_or_name):
413 """Report information extraction."""
414 self.to_screen('%s: Extracting information' % id_or_name)
416 def report_download_webpage(self, video_id):
417 """Report webpage download."""
418 self.to_screen('%s: Downloading webpage' % video_id)
420 def report_age_confirmation(self):
421 """Report attempt to confirm age."""
422 self.to_screen('Confirming age')
424 def report_login(self):
425 """Report attempt to log in."""
426 self.to_screen('Logging in')
428 # Methods for following #608
430 def url_result(url, ie=None, video_id=None):
431 """Returns a url that points to a page that should be processed"""
432 # TODO: ie should be the class used for getting the info
433 video_info = {'_type': 'url',
436 if video_id is not None:
437 video_info['id'] = video_id
441 def playlist_result(entries, playlist_id=None, playlist_title=None):
442 """Returns a playlist"""
443 video_info = {'_type': 'playlist',
446 video_info['id'] = playlist_id
448 video_info['title'] = playlist_title
451 def _search_regex(self, pattern, string, name, default=_NO_DEFAULT, fatal=True, flags=0, group=None):
453 Perform a regex search on the given string, using a single or a list of
454 patterns returning the first matching group.
455 In case of failure return a default value or raise a WARNING or a
456 RegexNotFoundError, depending on fatal, specifying the field name.
458 if isinstance(pattern, (str, compat_str, compiled_regex_type)):
459 mobj = re.search(pattern, string, flags)
462 mobj = re.search(p, string, flags)
466 if os.name != 'nt' and sys.stderr.isatty():
467 _name = '\033[0;34m%s\033[0m' % name
473 # return the first matching group
474 return next(g for g in mobj.groups() if g is not None)
476 return mobj.group(group)
477 elif default is not _NO_DEFAULT:
480 raise RegexNotFoundError('Unable to extract %s' % _name)
482 self._downloader.report_warning('unable to extract %s; '
483 'please report this issue on http://yt-dl.org/bug' % _name)
486 def _html_search_regex(self, pattern, string, name, default=_NO_DEFAULT, fatal=True, flags=0, group=None):
488 Like _search_regex, but strips HTML tags and unescapes entities.
490 res = self._search_regex(pattern, string, name, default, fatal, flags, group)
492 return clean_html(res).strip()
496 def _get_login_info(self):
498 Get the the login info as (username, password)
499 It will look in the netrc file using the _NETRC_MACHINE value
500 If there's no info available, return (None, None)
502 if self._downloader is None:
507 downloader_params = self._downloader.params
509 # Attempt to use provided username and password or .netrc data
510 if downloader_params.get('username', None) is not None:
511 username = downloader_params['username']
512 password = downloader_params['password']
513 elif downloader_params.get('usenetrc', False):
515 info = netrc.netrc().authenticators(self._NETRC_MACHINE)
520 raise netrc.NetrcParseError('No authenticators for %s' % self._NETRC_MACHINE)
521 except (IOError, netrc.NetrcParseError) as err:
522 self._downloader.report_warning('parsing .netrc: %s' % compat_str(err))
524 return (username, password)
526 def _get_tfa_info(self):
528 Get the two-factor authentication info
529 TODO - asking the user will be required for sms/phone verify
530 currently just uses the command line option
531 If there's no info available, return None
533 if self._downloader is None:
535 downloader_params = self._downloader.params
537 if downloader_params.get('twofactor', None) is not None:
538 return downloader_params['twofactor']
542 # Helper functions for extracting OpenGraph info
544 def _og_regexes(prop):
545 content_re = r'content=(?:"([^>]+?)"|\'([^>]+?)\')'
546 property_re = r'(?:name|property)=[\'"]og:%s[\'"]' % re.escape(prop)
547 template = r'<meta[^>]+?%s[^>]+?%s'
549 template % (property_re, content_re),
550 template % (content_re, property_re),
553 def _og_search_property(self, prop, html, name=None, **kargs):
555 name = 'OpenGraph %s' % prop
556 escaped = self._search_regex(self._og_regexes(prop), html, name, flags=re.DOTALL, **kargs)
559 return unescapeHTML(escaped)
561 def _og_search_thumbnail(self, html, **kargs):
562 return self._og_search_property('image', html, 'thumbnail url', fatal=False, **kargs)
564 def _og_search_description(self, html, **kargs):
565 return self._og_search_property('description', html, fatal=False, **kargs)
567 def _og_search_title(self, html, **kargs):
568 return self._og_search_property('title', html, **kargs)
570 def _og_search_video_url(self, html, name='video url', secure=True, **kargs):
571 regexes = self._og_regexes('video') + self._og_regexes('video:url')
573 regexes = self._og_regexes('video:secure_url') + regexes
574 return self._html_search_regex(regexes, html, name, **kargs)
576 def _og_search_url(self, html, **kargs):
577 return self._og_search_property('url', html, **kargs)
579 def _html_search_meta(self, name, html, display_name=None, fatal=False, **kwargs):
580 if display_name is None:
582 return self._html_search_regex(
584 (?=[^>]+(?:itemprop|name|property)=(["\']?)%s\1)
585 [^>]+content=(["\'])(?P<content>.*?)\1''' % re.escape(name),
586 html, display_name, fatal=fatal, group='content', **kwargs)
588 def _dc_search_uploader(self, html):
589 return self._html_search_meta('dc.creator', html, 'uploader')
591 def _rta_search(self, html):
592 # See http://www.rtalabel.org/index.php?content=howtofaq#single
593 if re.search(r'(?ix)<meta\s+name="rating"\s+'
594 r' content="RTA-5042-1996-1400-1577-RTA"',
599 def _media_rating_search(self, html):
600 # See http://www.tjg-designs.com/WP/metadata-code-examples-adding-metadata-to-your-web-pages/
601 rating = self._html_search_meta('rating', html)
613 return RATING_TABLE.get(rating.lower(), None)
615 def _twitter_search_player(self, html):
616 return self._html_search_meta('twitter:player', html,
617 'twitter card player')
619 def _sort_formats(self, formats):
621 raise ExtractorError('No video formats found')
624 # TODO remove the following workaround
625 from ..utils import determine_ext
626 if not f.get('ext') and 'url' in f:
627 f['ext'] = determine_ext(f['url'])
629 preference = f.get('preference')
630 if preference is None:
631 proto = f.get('protocol')
633 proto = compat_urllib_parse_urlparse(f.get('url', '')).scheme
635 preference = 0 if proto in ['http', 'https'] else -0.1
636 if f.get('ext') in ['f4f', 'f4m']: # Not yet supported
639 if f.get('vcodec') == 'none': # audio only
640 if self._downloader.params.get('prefer_free_formats'):
641 ORDER = ['aac', 'mp3', 'm4a', 'webm', 'ogg', 'opus']
643 ORDER = ['webm', 'opus', 'ogg', 'mp3', 'aac', 'm4a']
646 audio_ext_preference = ORDER.index(f['ext'])
648 audio_ext_preference = -1
650 if self._downloader.params.get('prefer_free_formats'):
651 ORDER = ['flv', 'mp4', 'webm']
653 ORDER = ['webm', 'flv', 'mp4']
655 ext_preference = ORDER.index(f['ext'])
658 audio_ext_preference = 0
662 f.get('language_preference') if f.get('language_preference') is not None else -1,
663 f.get('quality') if f.get('quality') is not None else -1,
664 f.get('height') if f.get('height') is not None else -1,
665 f.get('width') if f.get('width') is not None else -1,
667 f.get('tbr') if f.get('tbr') is not None else -1,
668 f.get('vbr') if f.get('vbr') is not None else -1,
669 f.get('abr') if f.get('abr') is not None else -1,
670 audio_ext_preference,
671 f.get('fps') if f.get('fps') is not None else -1,
672 f.get('filesize') if f.get('filesize') is not None else -1,
673 f.get('filesize_approx') if f.get('filesize_approx') is not None else -1,
674 f.get('source_preference') if f.get('source_preference') is not None else -1,
677 formats.sort(key=_formats_key)
679 def http_scheme(self):
680 """ Either "http:" or "https:", depending on the user's preferences """
683 if self._downloader.params.get('prefer_insecure', False)
686 def _proto_relative_url(self, url, scheme=None):
689 if url.startswith('//'):
691 scheme = self.http_scheme()
696 def _sleep(self, timeout, video_id, msg_template=None):
697 if msg_template is None:
698 msg_template = '%(video_id)s: Waiting for %(timeout)s seconds'
699 msg = msg_template % {'video_id': video_id, 'timeout': timeout}
703 def _extract_f4m_formats(self, manifest_url, video_id):
704 manifest = self._download_xml(
705 manifest_url, video_id, 'Downloading f4m manifest',
706 'Unable to download f4m manifest')
709 media_nodes = manifest.findall('{http://ns.adobe.com/f4m/1.0}media')
710 for i, media_el in enumerate(media_nodes):
711 tbr = int_or_none(media_el.attrib.get('bitrate'))
712 format_id = 'f4m-%d' % (i if tbr is None else tbr)
714 'format_id': format_id,
718 'width': int_or_none(media_el.attrib.get('width')),
719 'height': int_or_none(media_el.attrib.get('height')),
721 self._sort_formats(formats)
725 def _extract_m3u8_formats(self, m3u8_url, video_id, ext=None,
726 entry_protocol='m3u8', preference=None):
729 'format_id': 'm3u8-meta',
734 'resolution': 'multiple',
735 'format_note': 'Quality selection URL',
738 format_url = lambda u: (
740 if re.match(r'^https?://', u)
741 else compat_urlparse.urljoin(m3u8_url, u))
743 m3u8_doc = self._download_webpage(
745 note='Downloading m3u8 information',
746 errnote='Failed to download m3u8 information')
749 r'(?P<key>[a-zA-Z_-]+)=(?P<val>"[^"]+"|[^",]+)(?:,|$)')
750 for line in m3u8_doc.splitlines():
751 if line.startswith('#EXT-X-STREAM-INF:'):
753 for m in kv_rex.finditer(line):
755 if v.startswith('"'):
757 last_info[m.group('key')] = v
758 elif line.startswith('#') or not line.strip():
761 if last_info is None:
762 formats.append({'url': format_url(line)})
764 tbr = int_or_none(last_info.get('BANDWIDTH'), scale=1000)
767 'format_id': 'm3u8-%d' % (tbr if tbr else len(formats)),
768 'url': format_url(line.strip()),
771 'protocol': entry_protocol,
772 'preference': preference,
774 codecs = last_info.get('CODECS')
776 # TODO: looks like video codec is not always necessarily goes first
777 va_codecs = codecs.split(',')
779 f['vcodec'] = va_codecs[0].partition('.')[0]
780 if len(va_codecs) > 1 and va_codecs[1]:
781 f['acodec'] = va_codecs[1].partition('.')[0]
782 resolution = last_info.get('RESOLUTION')
784 width_str, height_str = resolution.split('x')
785 f['width'] = int(width_str)
786 f['height'] = int(height_str)
789 self._sort_formats(formats)
792 def _live_title(self, name):
793 """ Generate the title for a live video """
794 now = datetime.datetime.now()
795 now_str = now.strftime("%Y-%m-%d %H:%M")
796 return name + ' ' + now_str
798 def _int(self, v, name, fatal=False, **kwargs):
799 res = int_or_none(v, **kwargs)
800 if 'get_attr' in kwargs:
801 print(getattr(v, kwargs['get_attr']))
803 msg = 'Failed to extract %s: Could not parse value %r' % (name, v)
805 raise ExtractorError(msg)
807 self._downloader.report_warning(msg)
810 def _float(self, v, name, fatal=False, **kwargs):
811 res = float_or_none(v, **kwargs)
813 msg = 'Failed to extract %s: Could not parse value %r' % (name, v)
815 raise ExtractorError(msg)
817 self._downloader.report_warning(msg)
821 class SearchInfoExtractor(InfoExtractor):
823 Base class for paged search queries extractors.
824 They accept urls in the format _SEARCH_KEY(|all|[0-9]):{query}
825 Instances should define _SEARCH_KEY and _MAX_RESULTS.
829 def _make_valid_url(cls):
830 return r'%s(?P<prefix>|[1-9][0-9]*|all):(?P<query>[\s\S]+)' % cls._SEARCH_KEY
833 def suitable(cls, url):
834 return re.match(cls._make_valid_url(), url) is not None
836 def _real_extract(self, query):
837 mobj = re.match(self._make_valid_url(), query)
839 raise ExtractorError('Invalid search query "%s"' % query)
841 prefix = mobj.group('prefix')
842 query = mobj.group('query')
844 return self._get_n_results(query, 1)
845 elif prefix == 'all':
846 return self._get_n_results(query, self._MAX_RESULTS)
850 raise ExtractorError('invalid download number %s for query "%s"' % (n, query))
851 elif n > self._MAX_RESULTS:
852 self._downloader.report_warning('%s returns max %i results (you requested %i)' % (self._SEARCH_KEY, self._MAX_RESULTS, n))
853 n = self._MAX_RESULTS
854 return self._get_n_results(query, n)
856 def _get_n_results(self, query, n):
857 """Get a specified number of results for a query"""
858 raise NotImplementedError("This method must be implemented by subclasses")
861 def SEARCH_KEY(self):
862 return self._SEARCH_KEY