9 import xml.etree.ElementTree
14 compat_urllib_parse_urlparse,
24 _NO_DEFAULT = object()
27 class InfoExtractor(object):
28 """Information Extractor class.
30 Information extractors are the classes that, given a URL, extract
31 information about the video (or videos) the URL refers to. This
32 information includes the real video URL, the video title, author and
33 others. The information is stored in a dictionary which is then
34 passed to the FileDownloader. The FileDownloader processes this
35 information possibly downloading the video to the file system, among
36 other possible outcomes.
38 The dictionaries must include the following fields:
41 title: Video title, unescaped.
43 Additionally, it must contain either a formats entry or a url one:
45 formats: A list of dictionaries for each format available, ordered
46 from worst to best quality.
49 * url Mandatory. The URL of the video file
50 * ext Will be calculated from url if missing
51 * format A human-readable description of the format
52 ("mp4 container with h264/opus").
53 Calculated from the format_id, width, height.
54 and format_note fields if missing.
55 * format_id A short description of the format
56 ("mp4_h264_opus" or "19").
57 Technically optional, but strongly recommended.
58 * format_note Additional info about the format
59 ("3D" or "DASH video")
60 * width Width of the video, if known
61 * height Height of the video, if known
62 * resolution Textual description of width and height
63 * tbr Average bitrate of audio and video in KBit/s
64 * abr Average audio bitrate in KBit/s
65 * acodec Name of the audio codec in use
66 * asr Audio sampling rate in Hertz
67 * vbr Average video bitrate in KBit/s
68 * vcodec Name of the video codec in use
69 * container Name of the container format
70 * filesize The number of bytes, if known in advance
71 * player_url SWF Player URL (used for rtmpdump).
72 * protocol The protocol that will be used for the actual
74 "http", "https", "rtsp", "rtmp", "m3u8" or so.
75 * preference Order number of this format. If this field is
76 present and not None, the formats get sorted
77 by this field, regardless of all other values.
78 -1 for default (order by other properties),
79 -2 or smaller for less than default.
80 * quality Order number of the video quality of this
81 format, irrespective of the file format.
82 -1 for default (order by other properties),
83 -2 or smaller for less than default.
85 ext: Video filename extension.
86 format: The video format, defaults to ext (used for --get-format)
87 player_url: SWF Player URL (used for rtmpdump).
89 The following fields are optional:
91 display_id An alternative identifier for the video, not necessarily
92 unique, but available before title. Typically, id is
93 something like "4234987", title "Dancing naked mole rats",
94 and display_id "dancing-naked-mole-rats"
95 thumbnails: A list of dictionaries (with the entries "resolution" and
96 "url") for the varying thumbnails
97 thumbnail: Full URL to a video thumbnail image.
98 description: One-line video description.
99 uploader: Full name of the video uploader.
100 timestamp: UNIX timestamp of the moment the video became available.
101 upload_date: Video upload date (YYYYMMDD).
102 If not explicitly set, calculated from timestamp.
103 uploader_id: Nickname or id of the video uploader.
104 location: Physical location of the video.
105 subtitles: The subtitle file contents as a dictionary in the format
106 {language: subtitles}.
107 duration: Length of the video in seconds, as an integer.
108 view_count: How many users have watched the video on the platform.
109 like_count: Number of positive ratings of the video
110 dislike_count: Number of negative ratings of the video
111 comment_count: Number of comments on the video
112 age_limit: Age restriction for the video, as an integer (years)
113 webpage_url: The url to the video webpage, if given to youtube-dl it
114 should allow to get the same result again. (It will be set
115 by YoutubeDL if it's missing)
117 Unless mentioned otherwise, the fields should be Unicode strings.
119 Subclasses of this one should re-define the _real_initialize() and
120 _real_extract() methods and define a _VALID_URL regexp.
121 Probably, they should also be added to the list of extractors.
123 Finally, the _WORKING attribute should be set to False for broken IEs
124 in order to warn the users and skip the tests.
131 def __init__(self, downloader=None):
132 """Constructor. Receives an optional downloader."""
134 self.set_downloader(downloader)
137 def suitable(cls, url):
138 """Receives a URL and returns True if suitable for this IE."""
140 # This does not use has/getattr intentionally - we want to know whether
141 # we have cached the regexp for *this* class, whereas getattr would also
142 # match the superclass
143 if '_VALID_URL_RE' not in cls.__dict__:
144 cls._VALID_URL_RE = re.compile(cls._VALID_URL)
145 return cls._VALID_URL_RE.match(url) is not None
149 """Getter method for _WORKING."""
152 def initialize(self):
153 """Initializes an instance (authentication, etc)."""
155 self._real_initialize()
158 def extract(self, url):
159 """Extracts URL information and returns it in list of dicts."""
161 return self._real_extract(url)
163 def set_downloader(self, downloader):
164 """Sets the downloader for this IE."""
165 self._downloader = downloader
167 def _real_initialize(self):
168 """Real initialization process. Redefine in subclasses."""
171 def _real_extract(self, url):
172 """Real extraction process. Redefine in subclasses."""
177 """A string for getting the InfoExtractor with get_info_extractor"""
178 return cls.__name__[:-2]
182 return type(self).__name__[:-2]
184 def _request_webpage(self, url_or_request, video_id, note=None, errnote=None, fatal=True):
185 """ Returns the response handle """
187 self.report_download_webpage(video_id)
188 elif note is not False:
190 self.to_screen(u'%s' % (note,))
192 self.to_screen(u'%s: %s' % (video_id, note))
194 return self._downloader.urlopen(url_or_request)
195 except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
199 errnote = u'Unable to download webpage'
200 errmsg = u'%s: %s' % (errnote, compat_str(err))
202 raise ExtractorError(errmsg, sys.exc_info()[2], cause=err)
204 self._downloader.report_warning(errmsg)
207 def _download_webpage_handle(self, url_or_request, video_id, note=None, errnote=None, fatal=True):
208 """ Returns a tuple (page content as string, URL handle) """
210 # Strip hashes from the URL (#1038)
211 if isinstance(url_or_request, (compat_str, str)):
212 url_or_request = url_or_request.partition('#')[0]
214 urlh = self._request_webpage(url_or_request, video_id, note, errnote, fatal)
218 content_type = urlh.headers.get('Content-Type', '')
219 webpage_bytes = urlh.read()
220 m = re.match(r'[a-zA-Z0-9_.-]+/[a-zA-Z0-9_.-]+\s*;\s*charset=(.+)', content_type)
222 encoding = m.group(1)
224 m = re.search(br'<meta[^>]+charset=[\'"]?([^\'")]+)[ /\'">]',
225 webpage_bytes[:1024])
227 encoding = m.group(1).decode('ascii')
228 elif webpage_bytes.startswith(b'\xff\xfe'):
232 if self._downloader.params.get('dump_intermediate_pages', False):
234 url = url_or_request.get_full_url()
235 except AttributeError:
237 self.to_screen(u'Dumping request to ' + url)
238 dump = base64.b64encode(webpage_bytes).decode('ascii')
239 self._downloader.to_screen(dump)
240 if self._downloader.params.get('write_pages', False):
242 url = url_or_request.get_full_url()
243 except AttributeError:
246 h = u'___' + hashlib.md5(url.encode('utf-8')).hexdigest()
247 url = url[:200 - len(h)] + h
248 raw_filename = ('%s_%s.dump' % (video_id, url))
249 filename = sanitize_filename(raw_filename, restricted=True)
250 self.to_screen(u'Saving request to ' + filename)
251 with open(filename, 'wb') as outf:
252 outf.write(webpage_bytes)
254 content = webpage_bytes.decode(encoding, 'replace')
256 if (u'<title>Access to this site is blocked</title>' in content and
257 u'Websense' in content[:512]):
258 msg = u'Access to URL %s has been blocked by Websense filtering software in your network.' % urlh.geturl()
259 blocked_iframe = self._html_search_regex(
260 r'<iframe src="([^"]+)"', content,
261 u'Websense information URL', default=None)
263 msg += u' Visit %s for more details' % blocked_iframe
264 raise ExtractorError(msg, expected=True)
266 return (content, urlh)
268 def _download_webpage(self, url_or_request, video_id, note=None, errnote=None, fatal=True):
269 """ Returns the data of the page as a string """
270 res = self._download_webpage_handle(url_or_request, video_id, note, errnote, fatal)
277 def _download_xml(self, url_or_request, video_id,
278 note=u'Downloading XML', errnote=u'Unable to download XML',
279 transform_source=None):
280 """Return the xml as an xml.etree.ElementTree.Element"""
281 xml_string = self._download_webpage(url_or_request, video_id, note, errnote)
283 xml_string = transform_source(xml_string)
284 return xml.etree.ElementTree.fromstring(xml_string.encode('utf-8'))
286 def _download_json(self, url_or_request, video_id,
287 note=u'Downloading JSON metadata',
288 errnote=u'Unable to download JSON metadata',
289 transform_source=None):
290 json_string = self._download_webpage(url_or_request, video_id, note, errnote)
292 json_string = transform_source(json_string)
294 return json.loads(json_string)
295 except ValueError as ve:
296 raise ExtractorError('Failed to download JSON', cause=ve)
298 def report_warning(self, msg, video_id=None):
299 idstr = u'' if video_id is None else u'%s: ' % video_id
300 self._downloader.report_warning(
301 u'[%s] %s%s' % (self.IE_NAME, idstr, msg))
303 def to_screen(self, msg):
304 """Print msg to screen, prefixing it with '[ie_name]'"""
305 self._downloader.to_screen(u'[%s] %s' % (self.IE_NAME, msg))
307 def report_extraction(self, id_or_name):
308 """Report information extraction."""
309 self.to_screen(u'%s: Extracting information' % id_or_name)
311 def report_download_webpage(self, video_id):
312 """Report webpage download."""
313 self.to_screen(u'%s: Downloading webpage' % video_id)
315 def report_age_confirmation(self):
316 """Report attempt to confirm age."""
317 self.to_screen(u'Confirming age')
319 def report_login(self):
320 """Report attempt to log in."""
321 self.to_screen(u'Logging in')
323 #Methods for following #608
325 def url_result(url, ie=None, video_id=None):
326 """Returns a url that points to a page that should be processed"""
327 #TODO: ie should be the class used for getting the info
328 video_info = {'_type': 'url',
331 if video_id is not None:
332 video_info['id'] = video_id
335 def playlist_result(entries, playlist_id=None, playlist_title=None):
336 """Returns a playlist"""
337 video_info = {'_type': 'playlist',
340 video_info['id'] = playlist_id
342 video_info['title'] = playlist_title
345 def _search_regex(self, pattern, string, name, default=_NO_DEFAULT, fatal=True, flags=0):
347 Perform a regex search on the given string, using a single or a list of
348 patterns returning the first matching group.
349 In case of failure return a default value or raise a WARNING or a
350 RegexNotFoundError, depending on fatal, specifying the field name.
352 if isinstance(pattern, (str, compat_str, compiled_regex_type)):
353 mobj = re.search(pattern, string, flags)
356 mobj = re.search(p, string, flags)
359 if os.name != 'nt' and sys.stderr.isatty():
360 _name = u'\033[0;34m%s\033[0m' % name
365 # return the first matching group
366 return next(g for g in mobj.groups() if g is not None)
367 elif default is not _NO_DEFAULT:
370 raise RegexNotFoundError(u'Unable to extract %s' % _name)
372 self._downloader.report_warning(u'unable to extract %s; '
373 u'please report this issue on http://yt-dl.org/bug' % _name)
376 def _html_search_regex(self, pattern, string, name, default=_NO_DEFAULT, fatal=True, flags=0):
378 Like _search_regex, but strips HTML tags and unescapes entities.
380 res = self._search_regex(pattern, string, name, default, fatal, flags)
382 return clean_html(res).strip()
386 def _get_login_info(self):
388 Get the the login info as (username, password)
389 It will look in the netrc file using the _NETRC_MACHINE value
390 If there's no info available, return (None, None)
392 if self._downloader is None:
397 downloader_params = self._downloader.params
399 # Attempt to use provided username and password or .netrc data
400 if downloader_params.get('username', None) is not None:
401 username = downloader_params['username']
402 password = downloader_params['password']
403 elif downloader_params.get('usenetrc', False):
405 info = netrc.netrc().authenticators(self._NETRC_MACHINE)
410 raise netrc.NetrcParseError('No authenticators for %s' % self._NETRC_MACHINE)
411 except (IOError, netrc.NetrcParseError) as err:
412 self._downloader.report_warning(u'parsing .netrc: %s' % compat_str(err))
414 return (username, password)
416 # Helper functions for extracting OpenGraph info
418 def _og_regexes(prop):
419 content_re = r'content=(?:"([^>]+?)"|\'([^>]+?)\')'
420 property_re = r'(?:name|property)=[\'"]og:%s[\'"]' % re.escape(prop)
421 template = r'<meta[^>]+?%s[^>]+?%s'
423 template % (property_re, content_re),
424 template % (content_re, property_re),
427 def _og_search_property(self, prop, html, name=None, **kargs):
429 name = 'OpenGraph %s' % prop
430 escaped = self._search_regex(self._og_regexes(prop), html, name, flags=re.DOTALL, **kargs)
433 return unescapeHTML(escaped)
435 def _og_search_thumbnail(self, html, **kargs):
436 return self._og_search_property('image', html, u'thumbnail url', fatal=False, **kargs)
438 def _og_search_description(self, html, **kargs):
439 return self._og_search_property('description', html, fatal=False, **kargs)
441 def _og_search_title(self, html, **kargs):
442 return self._og_search_property('title', html, **kargs)
444 def _og_search_video_url(self, html, name='video url', secure=True, **kargs):
445 regexes = self._og_regexes('video')
446 if secure: regexes = self._og_regexes('video:secure_url') + regexes
447 return self._html_search_regex(regexes, html, name, **kargs)
449 def _html_search_meta(self, name, html, display_name=None, fatal=False):
450 if display_name is None:
452 return self._html_search_regex(
454 (?=[^>]+(?:itemprop|name|property)=["\']%s["\'])
455 [^>]+content=["\']([^"\']+)["\']''' % re.escape(name),
456 html, display_name, fatal=fatal)
458 def _dc_search_uploader(self, html):
459 return self._html_search_meta('dc.creator', html, 'uploader')
461 def _rta_search(self, html):
462 # See http://www.rtalabel.org/index.php?content=howtofaq#single
463 if re.search(r'(?ix)<meta\s+name="rating"\s+'
464 r' content="RTA-5042-1996-1400-1577-RTA"',
469 def _media_rating_search(self, html):
470 # See http://www.tjg-designs.com/WP/metadata-code-examples-adding-metadata-to-your-web-pages/
471 rating = self._html_search_meta('rating', html)
483 return RATING_TABLE.get(rating.lower(), None)
485 def _twitter_search_player(self, html):
486 return self._html_search_meta('twitter:player', html,
487 'twitter card player')
489 def _sort_formats(self, formats):
491 raise ExtractorError(u'No video formats found')
494 # TODO remove the following workaround
495 from ..utils import determine_ext
496 if not f.get('ext') and 'url' in f:
497 f['ext'] = determine_ext(f['url'])
499 preference = f.get('preference')
500 if preference is None:
501 proto = f.get('protocol')
503 proto = compat_urllib_parse_urlparse(f.get('url', '')).scheme
505 preference = 0 if proto in ['http', 'https'] else -0.1
506 if f.get('ext') in ['f4f', 'f4m']: # Not yet supported
509 if f.get('vcodec') == 'none': # audio only
510 if self._downloader.params.get('prefer_free_formats'):
511 ORDER = [u'aac', u'mp3', u'm4a', u'webm', u'ogg', u'opus']
513 ORDER = [u'webm', u'opus', u'ogg', u'mp3', u'aac', u'm4a']
516 audio_ext_preference = ORDER.index(f['ext'])
518 audio_ext_preference = -1
520 if self._downloader.params.get('prefer_free_formats'):
521 ORDER = [u'flv', u'mp4', u'webm']
523 ORDER = [u'webm', u'flv', u'mp4']
525 ext_preference = ORDER.index(f['ext'])
528 audio_ext_preference = 0
532 f.get('quality') if f.get('quality') is not None else -1,
533 f.get('height') if f.get('height') is not None else -1,
534 f.get('width') if f.get('width') is not None else -1,
536 f.get('tbr') if f.get('tbr') is not None else -1,
537 f.get('vbr') if f.get('vbr') is not None else -1,
538 f.get('abr') if f.get('abr') is not None else -1,
539 audio_ext_preference,
540 f.get('filesize') if f.get('filesize') is not None else -1,
543 formats.sort(key=_formats_key)
546 class SearchInfoExtractor(InfoExtractor):
548 Base class for paged search queries extractors.
549 They accept urls in the format _SEARCH_KEY(|all|[0-9]):{query}
550 Instances should define _SEARCH_KEY and _MAX_RESULTS.
554 def _make_valid_url(cls):
555 return r'%s(?P<prefix>|[1-9][0-9]*|all):(?P<query>[\s\S]+)' % cls._SEARCH_KEY
558 def suitable(cls, url):
559 return re.match(cls._make_valid_url(), url) is not None
561 def _real_extract(self, query):
562 mobj = re.match(self._make_valid_url(), query)
564 raise ExtractorError(u'Invalid search query "%s"' % query)
566 prefix = mobj.group('prefix')
567 query = mobj.group('query')
569 return self._get_n_results(query, 1)
570 elif prefix == 'all':
571 return self._get_n_results(query, self._MAX_RESULTS)
575 raise ExtractorError(u'invalid download number %s for query "%s"' % (n, query))
576 elif n > self._MAX_RESULTS:
577 self._downloader.report_warning(u'%s returns max %i results (you requested %i)' % (self._SEARCH_KEY, self._MAX_RESULTS, n))
578 n = self._MAX_RESULTS
579 return self._get_n_results(query, n)
581 def _get_n_results(self, query, n):
582 """Get a specified number of results for a query"""
583 raise NotImplementedError("This method must be implemented by subclasses")
586 def SEARCH_KEY(self):
587 return self._SEARCH_KEY