]> Raphaƫl G. Git Repositories - youtubedl/blob - youtube_dl/extractor/common.py
cecf917ffb67040739bbacf573297ba891ec9ea1
[youtubedl] / youtube_dl / extractor / common.py
1 from __future__ import unicode_literals
2
3 import base64
4 import datetime
5 import hashlib
6 import json
7 import netrc
8 import os
9 import re
10 import socket
11 import sys
12 import time
13 import xml.etree.ElementTree
14
15 from ..compat import (
16 compat_cookiejar,
17 compat_HTTPError,
18 compat_http_client,
19 compat_urllib_error,
20 compat_urllib_parse_urlparse,
21 compat_urlparse,
22 compat_str,
23 )
24 from ..utils import (
25 age_restricted,
26 bug_reports_message,
27 clean_html,
28 compiled_regex_type,
29 ExtractorError,
30 float_or_none,
31 int_or_none,
32 RegexNotFoundError,
33 sanitize_filename,
34 unescapeHTML,
35 )
36 _NO_DEFAULT = object()
37
38
39 class InfoExtractor(object):
40 """Information Extractor class.
41
42 Information extractors are the classes that, given a URL, extract
43 information about the video (or videos) the URL refers to. This
44 information includes the real video URL, the video title, author and
45 others. The information is stored in a dictionary which is then
46 passed to the YoutubeDL. The YoutubeDL processes this
47 information possibly downloading the video to the file system, among
48 other possible outcomes.
49
50 The type field determines the type of the result.
51 By far the most common value (and the default if _type is missing) is
52 "video", which indicates a single video.
53
54 For a video, the dictionaries must include the following fields:
55
56 id: Video identifier.
57 title: Video title, unescaped.
58
59 Additionally, it must contain either a formats entry or a url one:
60
61 formats: A list of dictionaries for each format available, ordered
62 from worst to best quality.
63
64 Potential fields:
65 * url Mandatory. The URL of the video file
66 * ext Will be calculated from url if missing
67 * format A human-readable description of the format
68 ("mp4 container with h264/opus").
69 Calculated from the format_id, width, height.
70 and format_note fields if missing.
71 * format_id A short description of the format
72 ("mp4_h264_opus" or "19").
73 Technically optional, but strongly recommended.
74 * format_note Additional info about the format
75 ("3D" or "DASH video")
76 * width Width of the video, if known
77 * height Height of the video, if known
78 * resolution Textual description of width and height
79 * tbr Average bitrate of audio and video in KBit/s
80 * abr Average audio bitrate in KBit/s
81 * acodec Name of the audio codec in use
82 * asr Audio sampling rate in Hertz
83 * vbr Average video bitrate in KBit/s
84 * fps Frame rate
85 * vcodec Name of the video codec in use
86 * container Name of the container format
87 * filesize The number of bytes, if known in advance
88 * filesize_approx An estimate for the number of bytes
89 * player_url SWF Player URL (used for rtmpdump).
90 * protocol The protocol that will be used for the actual
91 download, lower-case.
92 "http", "https", "rtsp", "rtmp", "rtmpe",
93 "m3u8", or "m3u8_native".
94 * preference Order number of this format. If this field is
95 present and not None, the formats get sorted
96 by this field, regardless of all other values.
97 -1 for default (order by other properties),
98 -2 or smaller for less than default.
99 < -1000 to hide the format (if there is
100 another one which is strictly better)
101 * language_preference Is this in the correct requested
102 language?
103 10 if it's what the URL is about,
104 -1 for default (don't know),
105 -10 otherwise, other values reserved for now.
106 * quality Order number of the video quality of this
107 format, irrespective of the file format.
108 -1 for default (order by other properties),
109 -2 or smaller for less than default.
110 * source_preference Order number for this video source
111 (quality takes higher priority)
112 -1 for default (order by other properties),
113 -2 or smaller for less than default.
114 * http_headers A dictionary of additional HTTP headers
115 to add to the request.
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 * no_resume The server does not support resuming the
120 (HTTP or RTMP) download. Boolean.
121
122 url: Final video URL.
123 ext: Video filename extension.
124 format: The video format, defaults to ext (used for --get-format)
125 player_url: SWF Player URL (used for rtmpdump).
126
127 The following fields are optional:
128
129 alt_title: A secondary title of the video.
130 display_id An alternative identifier for the video, not necessarily
131 unique, but available before title. Typically, id is
132 something like "4234987", title "Dancing naked mole rats",
133 and display_id "dancing-naked-mole-rats"
134 thumbnails: A list of dictionaries, with the following entries:
135 * "id" (optional, string) - Thumbnail format ID
136 * "url"
137 * "preference" (optional, int) - quality of the image
138 * "width" (optional, int)
139 * "height" (optional, int)
140 * "resolution" (optional, string "{width}x{height"},
141 deprecated)
142 thumbnail: Full URL to a video thumbnail image.
143 description: Full video description.
144 uploader: Full name of the video uploader.
145 creator: The main artist who created the video.
146 timestamp: UNIX timestamp of the moment the video became available.
147 upload_date: Video upload date (YYYYMMDD).
148 If not explicitly set, calculated from timestamp.
149 uploader_id: Nickname or id of the video uploader.
150 location: Physical location where the video was filmed.
151 subtitles: The available subtitles as a dictionary in the format
152 {language: subformats}. "subformats" is a list sorted from
153 lower to higher preference, each element is a dictionary
154 with the "ext" entry and one of:
155 * "data": The subtitles file contents
156 * "url": A url pointing to the subtitles file
157 automatic_captions: Like 'subtitles', used by the YoutubeIE for
158 automatically generated captions
159 duration: Length of the video in seconds, as an integer.
160 view_count: How many users have watched the video on the platform.
161 like_count: Number of positive ratings of the video
162 dislike_count: Number of negative ratings of the video
163 average_rating: Average rating give by users, the scale used depends on the webpage
164 comment_count: Number of comments on the video
165 comments: A list of comments, each with one or more of the following
166 properties (all but one of text or html optional):
167 * "author" - human-readable name of the comment author
168 * "author_id" - user ID of the comment author
169 * "id" - Comment ID
170 * "html" - Comment as HTML
171 * "text" - Plain text of the comment
172 * "timestamp" - UNIX timestamp of comment
173 * "parent" - ID of the comment this one is replying to.
174 Set to "root" to indicate that this is a
175 comment to the original video.
176 age_limit: Age restriction for the video, as an integer (years)
177 webpage_url: The url to the video webpage, if given to youtube-dl it
178 should allow to get the same result again. (It will be set
179 by YoutubeDL if it's missing)
180 categories: A list of categories that the video falls in, for example
181 ["Sports", "Berlin"]
182 is_live: True, False, or None (=unknown). Whether this video is a
183 live stream that goes on instead of a fixed-length video.
184
185 Unless mentioned otherwise, the fields should be Unicode strings.
186
187 Unless mentioned otherwise, None is equivalent to absence of information.
188
189
190 _type "playlist" indicates multiple videos.
191 There must be a key "entries", which is a list, an iterable, or a PagedList
192 object, each element of which is a valid dictionary by this specification.
193
194 Additionally, playlists can have "title" and "id" attributes with the same
195 semantics as videos (see above).
196
197
198 _type "multi_video" indicates that there are multiple videos that
199 form a single show, for examples multiple acts of an opera or TV episode.
200 It must have an entries key like a playlist and contain all the keys
201 required for a video at the same time.
202
203
204 _type "url" indicates that the video must be extracted from another
205 location, possibly by a different extractor. Its only required key is:
206 "url" - the next URL to extract.
207 The key "ie_key" can be set to the class name (minus the trailing "IE",
208 e.g. "Youtube") if the extractor class is known in advance.
209 Additionally, the dictionary may have any properties of the resolved entity
210 known in advance, for example "title" if the title of the referred video is
211 known ahead of time.
212
213
214 _type "url_transparent" entities have the same specification as "url", but
215 indicate that the given additional information is more precise than the one
216 associated with the resolved URL.
217 This is useful when a site employs a video service that hosts the video and
218 its technical metadata, but that video service does not embed a useful
219 title, description etc.
220
221
222 Subclasses of this one should re-define the _real_initialize() and
223 _real_extract() methods and define a _VALID_URL regexp.
224 Probably, they should also be added to the list of extractors.
225
226 Finally, the _WORKING attribute should be set to False for broken IEs
227 in order to warn the users and skip the tests.
228 """
229
230 _ready = False
231 _downloader = None
232 _WORKING = True
233
234 def __init__(self, downloader=None):
235 """Constructor. Receives an optional downloader."""
236 self._ready = False
237 self.set_downloader(downloader)
238
239 @classmethod
240 def suitable(cls, url):
241 """Receives a URL and returns True if suitable for this IE."""
242
243 # This does not use has/getattr intentionally - we want to know whether
244 # we have cached the regexp for *this* class, whereas getattr would also
245 # match the superclass
246 if '_VALID_URL_RE' not in cls.__dict__:
247 cls._VALID_URL_RE = re.compile(cls._VALID_URL)
248 return cls._VALID_URL_RE.match(url) is not None
249
250 @classmethod
251 def _match_id(cls, url):
252 if '_VALID_URL_RE' not in cls.__dict__:
253 cls._VALID_URL_RE = re.compile(cls._VALID_URL)
254 m = cls._VALID_URL_RE.match(url)
255 assert m
256 return m.group('id')
257
258 @classmethod
259 def working(cls):
260 """Getter method for _WORKING."""
261 return cls._WORKING
262
263 def initialize(self):
264 """Initializes an instance (authentication, etc)."""
265 if not self._ready:
266 self._real_initialize()
267 self._ready = True
268
269 def extract(self, url):
270 """Extracts URL information and returns it in list of dicts."""
271 try:
272 self.initialize()
273 return self._real_extract(url)
274 except ExtractorError:
275 raise
276 except compat_http_client.IncompleteRead as e:
277 raise ExtractorError('A network error has occured.', cause=e, expected=True)
278 except (KeyError, StopIteration) as e:
279 raise ExtractorError('An extractor error has occured.', cause=e)
280
281 def set_downloader(self, downloader):
282 """Sets the downloader for this IE."""
283 self._downloader = downloader
284
285 def _real_initialize(self):
286 """Real initialization process. Redefine in subclasses."""
287 pass
288
289 def _real_extract(self, url):
290 """Real extraction process. Redefine in subclasses."""
291 pass
292
293 @classmethod
294 def ie_key(cls):
295 """A string for getting the InfoExtractor with get_info_extractor"""
296 return cls.__name__[:-2]
297
298 @property
299 def IE_NAME(self):
300 return type(self).__name__[:-2]
301
302 def _request_webpage(self, url_or_request, video_id, note=None, errnote=None, fatal=True):
303 """ Returns the response handle """
304 if note is None:
305 self.report_download_webpage(video_id)
306 elif note is not False:
307 if video_id is None:
308 self.to_screen('%s' % (note,))
309 else:
310 self.to_screen('%s: %s' % (video_id, note))
311 try:
312 return self._downloader.urlopen(url_or_request)
313 except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
314 if errnote is False:
315 return False
316 if errnote is None:
317 errnote = 'Unable to download webpage'
318 errmsg = '%s: %s' % (errnote, compat_str(err))
319 if fatal:
320 raise ExtractorError(errmsg, sys.exc_info()[2], cause=err)
321 else:
322 self._downloader.report_warning(errmsg)
323 return False
324
325 def _download_webpage_handle(self, url_or_request, video_id, note=None, errnote=None, fatal=True, encoding=None):
326 """ Returns a tuple (page content as string, URL handle) """
327 # Strip hashes from the URL (#1038)
328 if isinstance(url_or_request, (compat_str, str)):
329 url_or_request = url_or_request.partition('#')[0]
330
331 urlh = self._request_webpage(url_or_request, video_id, note, errnote, fatal)
332 if urlh is False:
333 assert not fatal
334 return False
335 content = self._webpage_read_content(urlh, url_or_request, video_id, note, errnote, fatal, encoding=encoding)
336 return (content, urlh)
337
338 @staticmethod
339 def _guess_encoding_from_content(content_type, webpage_bytes):
340 m = re.match(r'[a-zA-Z0-9_.-]+/[a-zA-Z0-9_.-]+\s*;\s*charset=(.+)', content_type)
341 if m:
342 encoding = m.group(1)
343 else:
344 m = re.search(br'<meta[^>]+charset=[\'"]?([^\'")]+)[ /\'">]',
345 webpage_bytes[:1024])
346 if m:
347 encoding = m.group(1).decode('ascii')
348 elif webpage_bytes.startswith(b'\xff\xfe'):
349 encoding = 'utf-16'
350 else:
351 encoding = 'utf-8'
352
353 return encoding
354
355 def _webpage_read_content(self, urlh, url_or_request, video_id, note=None, errnote=None, fatal=True, prefix=None, encoding=None):
356 content_type = urlh.headers.get('Content-Type', '')
357 webpage_bytes = urlh.read()
358 if prefix is not None:
359 webpage_bytes = prefix + webpage_bytes
360 if not encoding:
361 encoding = self._guess_encoding_from_content(content_type, webpage_bytes)
362 if self._downloader.params.get('dump_intermediate_pages', False):
363 try:
364 url = url_or_request.get_full_url()
365 except AttributeError:
366 url = url_or_request
367 self.to_screen('Dumping request to ' + url)
368 dump = base64.b64encode(webpage_bytes).decode('ascii')
369 self._downloader.to_screen(dump)
370 if self._downloader.params.get('write_pages', False):
371 try:
372 url = url_or_request.get_full_url()
373 except AttributeError:
374 url = url_or_request
375 basen = '%s_%s' % (video_id, url)
376 if len(basen) > 240:
377 h = '___' + hashlib.md5(basen.encode('utf-8')).hexdigest()
378 basen = basen[:240 - len(h)] + h
379 raw_filename = basen + '.dump'
380 filename = sanitize_filename(raw_filename, restricted=True)
381 self.to_screen('Saving request to ' + filename)
382 # Working around MAX_PATH limitation on Windows (see
383 # http://msdn.microsoft.com/en-us/library/windows/desktop/aa365247(v=vs.85).aspx)
384 if os.name == 'nt':
385 absfilepath = os.path.abspath(filename)
386 if len(absfilepath) > 259:
387 filename = '\\\\?\\' + absfilepath
388 with open(filename, 'wb') as outf:
389 outf.write(webpage_bytes)
390
391 try:
392 content = webpage_bytes.decode(encoding, 'replace')
393 except LookupError:
394 content = webpage_bytes.decode('utf-8', 'replace')
395
396 if ('<title>Access to this site is blocked</title>' in content and
397 'Websense' in content[:512]):
398 msg = 'Access to this webpage has been blocked by Websense filtering software in your network.'
399 blocked_iframe = self._html_search_regex(
400 r'<iframe src="([^"]+)"', content,
401 'Websense information URL', default=None)
402 if blocked_iframe:
403 msg += ' Visit %s for more details' % blocked_iframe
404 raise ExtractorError(msg, expected=True)
405 if '<title>The URL you requested has been blocked</title>' in content[:512]:
406 msg = (
407 'Access to this webpage has been blocked by Indian censorship. '
408 'Use a VPN or proxy server (with --proxy) to route around it.')
409 block_msg = self._html_search_regex(
410 r'</h1><p>(.*?)</p>',
411 content, 'block message', default=None)
412 if block_msg:
413 msg += ' (Message: "%s")' % block_msg.replace('\n', ' ')
414 raise ExtractorError(msg, expected=True)
415
416 return content
417
418 def _download_webpage(self, url_or_request, video_id, note=None, errnote=None, fatal=True, tries=1, timeout=5, encoding=None):
419 """ Returns the data of the page as a string """
420 success = False
421 try_count = 0
422 while success is False:
423 try:
424 res = self._download_webpage_handle(url_or_request, video_id, note, errnote, fatal, encoding=encoding)
425 success = True
426 except compat_http_client.IncompleteRead as e:
427 try_count += 1
428 if try_count >= tries:
429 raise e
430 self._sleep(timeout, video_id)
431 if res is False:
432 return res
433 else:
434 content, _ = res
435 return content
436
437 def _download_xml(self, url_or_request, video_id,
438 note='Downloading XML', errnote='Unable to download XML',
439 transform_source=None, fatal=True, encoding=None):
440 """Return the xml as an xml.etree.ElementTree.Element"""
441 xml_string = self._download_webpage(
442 url_or_request, video_id, note, errnote, fatal=fatal, encoding=encoding)
443 if xml_string is False:
444 return xml_string
445 if transform_source:
446 xml_string = transform_source(xml_string)
447 return xml.etree.ElementTree.fromstring(xml_string.encode('utf-8'))
448
449 def _download_json(self, url_or_request, video_id,
450 note='Downloading JSON metadata',
451 errnote='Unable to download JSON metadata',
452 transform_source=None,
453 fatal=True, encoding=None):
454 json_string = self._download_webpage(
455 url_or_request, video_id, note, errnote, fatal=fatal,
456 encoding=encoding)
457 if (not fatal) and json_string is False:
458 return None
459 return self._parse_json(
460 json_string, video_id, transform_source=transform_source, fatal=fatal)
461
462 def _parse_json(self, json_string, video_id, transform_source=None, fatal=True):
463 if transform_source:
464 json_string = transform_source(json_string)
465 try:
466 return json.loads(json_string)
467 except ValueError as ve:
468 errmsg = '%s: Failed to parse JSON ' % video_id
469 if fatal:
470 raise ExtractorError(errmsg, cause=ve)
471 else:
472 self.report_warning(errmsg + str(ve))
473
474 def report_warning(self, msg, video_id=None):
475 idstr = '' if video_id is None else '%s: ' % video_id
476 self._downloader.report_warning(
477 '[%s] %s%s' % (self.IE_NAME, idstr, msg))
478
479 def to_screen(self, msg):
480 """Print msg to screen, prefixing it with '[ie_name]'"""
481 self._downloader.to_screen('[%s] %s' % (self.IE_NAME, msg))
482
483 def report_extraction(self, id_or_name):
484 """Report information extraction."""
485 self.to_screen('%s: Extracting information' % id_or_name)
486
487 def report_download_webpage(self, video_id):
488 """Report webpage download."""
489 self.to_screen('%s: Downloading webpage' % video_id)
490
491 def report_age_confirmation(self):
492 """Report attempt to confirm age."""
493 self.to_screen('Confirming age')
494
495 def report_login(self):
496 """Report attempt to log in."""
497 self.to_screen('Logging in')
498
499 # Methods for following #608
500 @staticmethod
501 def url_result(url, ie=None, video_id=None, video_title=None):
502 """Returns a url that points to a page that should be processed"""
503 # TODO: ie should be the class used for getting the info
504 video_info = {'_type': 'url',
505 'url': url,
506 'ie_key': ie}
507 if video_id is not None:
508 video_info['id'] = video_id
509 if video_title is not None:
510 video_info['title'] = video_title
511 return video_info
512
513 @staticmethod
514 def playlist_result(entries, playlist_id=None, playlist_title=None, playlist_description=None):
515 """Returns a playlist"""
516 video_info = {'_type': 'playlist',
517 'entries': entries}
518 if playlist_id:
519 video_info['id'] = playlist_id
520 if playlist_title:
521 video_info['title'] = playlist_title
522 if playlist_description:
523 video_info['description'] = playlist_description
524 return video_info
525
526 def _search_regex(self, pattern, string, name, default=_NO_DEFAULT, fatal=True, flags=0, group=None):
527 """
528 Perform a regex search on the given string, using a single or a list of
529 patterns returning the first matching group.
530 In case of failure return a default value or raise a WARNING or a
531 RegexNotFoundError, depending on fatal, specifying the field name.
532 """
533 if isinstance(pattern, (str, compat_str, compiled_regex_type)):
534 mobj = re.search(pattern, string, flags)
535 else:
536 for p in pattern:
537 mobj = re.search(p, string, flags)
538 if mobj:
539 break
540
541 if not self._downloader.params.get('no_color') and os.name != 'nt' and sys.stderr.isatty():
542 _name = '\033[0;34m%s\033[0m' % name
543 else:
544 _name = name
545
546 if mobj:
547 if group is None:
548 # return the first matching group
549 return next(g for g in mobj.groups() if g is not None)
550 else:
551 return mobj.group(group)
552 elif default is not _NO_DEFAULT:
553 return default
554 elif fatal:
555 raise RegexNotFoundError('Unable to extract %s' % _name)
556 else:
557 self._downloader.report_warning('unable to extract %s' % _name + bug_reports_message())
558 return None
559
560 def _html_search_regex(self, pattern, string, name, default=_NO_DEFAULT, fatal=True, flags=0, group=None):
561 """
562 Like _search_regex, but strips HTML tags and unescapes entities.
563 """
564 res = self._search_regex(pattern, string, name, default, fatal, flags, group)
565 if res:
566 return clean_html(res).strip()
567 else:
568 return res
569
570 def _get_login_info(self):
571 """
572 Get the login info as (username, password)
573 It will look in the netrc file using the _NETRC_MACHINE value
574 If there's no info available, return (None, None)
575 """
576 if self._downloader is None:
577 return (None, None)
578
579 username = None
580 password = None
581 downloader_params = self._downloader.params
582
583 # Attempt to use provided username and password or .netrc data
584 if downloader_params.get('username', None) is not None:
585 username = downloader_params['username']
586 password = downloader_params['password']
587 elif downloader_params.get('usenetrc', False):
588 try:
589 info = netrc.netrc().authenticators(self._NETRC_MACHINE)
590 if info is not None:
591 username = info[0]
592 password = info[2]
593 else:
594 raise netrc.NetrcParseError('No authenticators for %s' % self._NETRC_MACHINE)
595 except (IOError, netrc.NetrcParseError) as err:
596 self._downloader.report_warning('parsing .netrc: %s' % compat_str(err))
597
598 return (username, password)
599
600 def _get_tfa_info(self):
601 """
602 Get the two-factor authentication info
603 TODO - asking the user will be required for sms/phone verify
604 currently just uses the command line option
605 If there's no info available, return None
606 """
607 if self._downloader is None:
608 return None
609 downloader_params = self._downloader.params
610
611 if downloader_params.get('twofactor', None) is not None:
612 return downloader_params['twofactor']
613
614 return None
615
616 # Helper functions for extracting OpenGraph info
617 @staticmethod
618 def _og_regexes(prop):
619 content_re = r'content=(?:"([^>]+?)"|\'([^>]+?)\')'
620 property_re = r'(?:name|property)=[\'"]og:%s[\'"]' % re.escape(prop)
621 template = r'<meta[^>]+?%s[^>]+?%s'
622 return [
623 template % (property_re, content_re),
624 template % (content_re, property_re),
625 ]
626
627 def _og_search_property(self, prop, html, name=None, **kargs):
628 if name is None:
629 name = 'OpenGraph %s' % prop
630 escaped = self._search_regex(self._og_regexes(prop), html, name, flags=re.DOTALL, **kargs)
631 if escaped is None:
632 return None
633 return unescapeHTML(escaped)
634
635 def _og_search_thumbnail(self, html, **kargs):
636 return self._og_search_property('image', html, 'thumbnail url', fatal=False, **kargs)
637
638 def _og_search_description(self, html, **kargs):
639 return self._og_search_property('description', html, fatal=False, **kargs)
640
641 def _og_search_title(self, html, **kargs):
642 return self._og_search_property('title', html, **kargs)
643
644 def _og_search_video_url(self, html, name='video url', secure=True, **kargs):
645 regexes = self._og_regexes('video') + self._og_regexes('video:url')
646 if secure:
647 regexes = self._og_regexes('video:secure_url') + regexes
648 return self._html_search_regex(regexes, html, name, **kargs)
649
650 def _og_search_url(self, html, **kargs):
651 return self._og_search_property('url', html, **kargs)
652
653 def _html_search_meta(self, name, html, display_name=None, fatal=False, **kwargs):
654 if display_name is None:
655 display_name = name
656 return self._html_search_regex(
657 r'''(?isx)<meta
658 (?=[^>]+(?:itemprop|name|property)=(["\']?)%s\1)
659 [^>]+?content=(["\'])(?P<content>.*?)\2''' % re.escape(name),
660 html, display_name, fatal=fatal, group='content', **kwargs)
661
662 def _dc_search_uploader(self, html):
663 return self._html_search_meta('dc.creator', html, 'uploader')
664
665 def _rta_search(self, html):
666 # See http://www.rtalabel.org/index.php?content=howtofaq#single
667 if re.search(r'(?ix)<meta\s+name="rating"\s+'
668 r' content="RTA-5042-1996-1400-1577-RTA"',
669 html):
670 return 18
671 return 0
672
673 def _media_rating_search(self, html):
674 # See http://www.tjg-designs.com/WP/metadata-code-examples-adding-metadata-to-your-web-pages/
675 rating = self._html_search_meta('rating', html)
676
677 if not rating:
678 return None
679
680 RATING_TABLE = {
681 'safe for kids': 0,
682 'general': 8,
683 '14 years': 14,
684 'mature': 17,
685 'restricted': 19,
686 }
687 return RATING_TABLE.get(rating.lower(), None)
688
689 def _family_friendly_search(self, html):
690 # See http://schema.org/VideoObject
691 family_friendly = self._html_search_meta('isFamilyFriendly', html)
692
693 if not family_friendly:
694 return None
695
696 RATING_TABLE = {
697 '1': 0,
698 'true': 0,
699 '0': 18,
700 'false': 18,
701 }
702 return RATING_TABLE.get(family_friendly.lower(), None)
703
704 def _twitter_search_player(self, html):
705 return self._html_search_meta('twitter:player', html,
706 'twitter card player')
707
708 def _sort_formats(self, formats, field_preference=None):
709 if not formats:
710 raise ExtractorError('No video formats found')
711
712 def _formats_key(f):
713 # TODO remove the following workaround
714 from ..utils import determine_ext
715 if not f.get('ext') and 'url' in f:
716 f['ext'] = determine_ext(f['url'])
717
718 if isinstance(field_preference, (list, tuple)):
719 return tuple(f.get(field) if f.get(field) is not None else -1 for field in field_preference)
720
721 preference = f.get('preference')
722 if preference is None:
723 proto = f.get('protocol')
724 if proto is None:
725 proto = compat_urllib_parse_urlparse(f.get('url', '')).scheme
726
727 preference = 0 if proto in ['http', 'https'] else -0.1
728 if f.get('ext') in ['f4f', 'f4m']: # Not yet supported
729 preference -= 0.5
730
731 if f.get('vcodec') == 'none': # audio only
732 if self._downloader.params.get('prefer_free_formats'):
733 ORDER = ['aac', 'mp3', 'm4a', 'webm', 'ogg', 'opus']
734 else:
735 ORDER = ['webm', 'opus', 'ogg', 'mp3', 'aac', 'm4a']
736 ext_preference = 0
737 try:
738 audio_ext_preference = ORDER.index(f['ext'])
739 except ValueError:
740 audio_ext_preference = -1
741 else:
742 if self._downloader.params.get('prefer_free_formats'):
743 ORDER = ['flv', 'mp4', 'webm']
744 else:
745 ORDER = ['webm', 'flv', 'mp4']
746 try:
747 ext_preference = ORDER.index(f['ext'])
748 except ValueError:
749 ext_preference = -1
750 audio_ext_preference = 0
751
752 return (
753 preference,
754 f.get('language_preference') if f.get('language_preference') is not None else -1,
755 f.get('quality') if f.get('quality') is not None else -1,
756 f.get('tbr') if f.get('tbr') is not None else -1,
757 f.get('filesize') if f.get('filesize') is not None else -1,
758 f.get('vbr') if f.get('vbr') is not None else -1,
759 f.get('height') if f.get('height') is not None else -1,
760 f.get('width') if f.get('width') is not None else -1,
761 ext_preference,
762 f.get('abr') if f.get('abr') is not None else -1,
763 audio_ext_preference,
764 f.get('fps') if f.get('fps') is not None else -1,
765 f.get('filesize_approx') if f.get('filesize_approx') is not None else -1,
766 f.get('source_preference') if f.get('source_preference') is not None else -1,
767 f.get('format_id') if f.get('format_id') is not None else '',
768 )
769 formats.sort(key=_formats_key)
770
771 def _check_formats(self, formats, video_id):
772 if formats:
773 formats[:] = filter(
774 lambda f: self._is_valid_url(
775 f['url'], video_id,
776 item='%s video format' % f.get('format_id') if f.get('format_id') else 'video'),
777 formats)
778
779 def _is_valid_url(self, url, video_id, item='video'):
780 url = self._proto_relative_url(url, scheme='http:')
781 # For now assume non HTTP(S) URLs always valid
782 if not (url.startswith('http://') or url.startswith('https://')):
783 return True
784 try:
785 self._request_webpage(url, video_id, 'Checking %s URL' % item)
786 return True
787 except ExtractorError as e:
788 if isinstance(e.cause, compat_HTTPError):
789 self.to_screen(
790 '%s: %s URL is invalid, skipping' % (video_id, item))
791 return False
792 raise
793
794 def http_scheme(self):
795 """ Either "http:" or "https:", depending on the user's preferences """
796 return (
797 'http:'
798 if self._downloader.params.get('prefer_insecure', False)
799 else 'https:')
800
801 def _proto_relative_url(self, url, scheme=None):
802 if url is None:
803 return url
804 if url.startswith('//'):
805 if scheme is None:
806 scheme = self.http_scheme()
807 return scheme + url
808 else:
809 return url
810
811 def _sleep(self, timeout, video_id, msg_template=None):
812 if msg_template is None:
813 msg_template = '%(video_id)s: Waiting for %(timeout)s seconds'
814 msg = msg_template % {'video_id': video_id, 'timeout': timeout}
815 self.to_screen(msg)
816 time.sleep(timeout)
817
818 def _extract_f4m_formats(self, manifest_url, video_id, preference=None, f4m_id=None):
819 manifest = self._download_xml(
820 manifest_url, video_id, 'Downloading f4m manifest',
821 'Unable to download f4m manifest')
822
823 formats = []
824 manifest_version = '1.0'
825 media_nodes = manifest.findall('{http://ns.adobe.com/f4m/1.0}media')
826 if not media_nodes:
827 manifest_version = '2.0'
828 media_nodes = manifest.findall('{http://ns.adobe.com/f4m/2.0}media')
829 for i, media_el in enumerate(media_nodes):
830 if manifest_version == '2.0':
831 manifest_url = ('/'.join(manifest_url.split('/')[:-1]) + '/' +
832 (media_el.attrib.get('href') or media_el.attrib.get('url')))
833 tbr = int_or_none(media_el.attrib.get('bitrate'))
834 formats.append({
835 'format_id': '-'.join(filter(None, [f4m_id, compat_str(i if tbr is None else tbr)])),
836 'url': manifest_url,
837 'ext': 'flv',
838 'tbr': tbr,
839 'width': int_or_none(media_el.attrib.get('width')),
840 'height': int_or_none(media_el.attrib.get('height')),
841 'preference': preference,
842 })
843 self._sort_formats(formats)
844
845 return formats
846
847 def _extract_m3u8_formats(self, m3u8_url, video_id, ext=None,
848 entry_protocol='m3u8', preference=None,
849 m3u8_id=None):
850
851 formats = [{
852 'format_id': '-'.join(filter(None, [m3u8_id, 'meta'])),
853 'url': m3u8_url,
854 'ext': ext,
855 'protocol': 'm3u8',
856 'preference': preference - 1 if preference else -1,
857 'resolution': 'multiple',
858 'format_note': 'Quality selection URL',
859 }]
860
861 format_url = lambda u: (
862 u
863 if re.match(r'^https?://', u)
864 else compat_urlparse.urljoin(m3u8_url, u))
865
866 m3u8_doc = self._download_webpage(
867 m3u8_url, video_id,
868 note='Downloading m3u8 information',
869 errnote='Failed to download m3u8 information')
870 last_info = None
871 last_media = None
872 kv_rex = re.compile(
873 r'(?P<key>[a-zA-Z_-]+)=(?P<val>"[^"]+"|[^",]+)(?:,|$)')
874 for line in m3u8_doc.splitlines():
875 if line.startswith('#EXT-X-STREAM-INF:'):
876 last_info = {}
877 for m in kv_rex.finditer(line):
878 v = m.group('val')
879 if v.startswith('"'):
880 v = v[1:-1]
881 last_info[m.group('key')] = v
882 elif line.startswith('#EXT-X-MEDIA:'):
883 last_media = {}
884 for m in kv_rex.finditer(line):
885 v = m.group('val')
886 if v.startswith('"'):
887 v = v[1:-1]
888 last_media[m.group('key')] = v
889 elif line.startswith('#') or not line.strip():
890 continue
891 else:
892 if last_info is None:
893 formats.append({'url': format_url(line)})
894 continue
895 tbr = int_or_none(last_info.get('BANDWIDTH'), scale=1000)
896 format_id = []
897 if m3u8_id:
898 format_id.append(m3u8_id)
899 last_media_name = last_media.get('NAME') if last_media and last_media.get('TYPE') != 'SUBTITLES' else None
900 format_id.append(last_media_name if last_media_name else '%d' % (tbr if tbr else len(formats)))
901 f = {
902 'format_id': '-'.join(format_id),
903 'url': format_url(line.strip()),
904 'tbr': tbr,
905 'ext': ext,
906 'protocol': entry_protocol,
907 'preference': preference,
908 }
909 codecs = last_info.get('CODECS')
910 if codecs:
911 # TODO: looks like video codec is not always necessarily goes first
912 va_codecs = codecs.split(',')
913 if va_codecs[0]:
914 f['vcodec'] = va_codecs[0].partition('.')[0]
915 if len(va_codecs) > 1 and va_codecs[1]:
916 f['acodec'] = va_codecs[1].partition('.')[0]
917 resolution = last_info.get('RESOLUTION')
918 if resolution:
919 width_str, height_str = resolution.split('x')
920 f['width'] = int(width_str)
921 f['height'] = int(height_str)
922 if last_media is not None:
923 f['m3u8_media'] = last_media
924 last_media = None
925 formats.append(f)
926 last_info = {}
927 self._sort_formats(formats)
928 return formats
929
930 # TODO: improve extraction
931 def _extract_smil_formats(self, smil_url, video_id, fatal=True):
932 smil = self._download_xml(
933 smil_url, video_id, 'Downloading SMIL file',
934 'Unable to download SMIL file', fatal=fatal)
935 if smil is False:
936 assert not fatal
937 return []
938
939 base = smil.find('./head/meta').get('base')
940
941 formats = []
942 rtmp_count = 0
943 if smil.findall('./body/seq/video'):
944 video = smil.findall('./body/seq/video')[0]
945 fmts, rtmp_count = self._parse_smil_video(video, video_id, base, rtmp_count)
946 formats.extend(fmts)
947 else:
948 for video in smil.findall('./body/switch/video'):
949 fmts, rtmp_count = self._parse_smil_video(video, video_id, base, rtmp_count)
950 formats.extend(fmts)
951
952 self._sort_formats(formats)
953
954 return formats
955
956 def _parse_smil_video(self, video, video_id, base, rtmp_count):
957 src = video.get('src')
958 if not src:
959 return ([], rtmp_count)
960 bitrate = int_or_none(video.get('system-bitrate') or video.get('systemBitrate'), 1000)
961 width = int_or_none(video.get('width'))
962 height = int_or_none(video.get('height'))
963 proto = video.get('proto')
964 if not proto:
965 if base:
966 if base.startswith('rtmp'):
967 proto = 'rtmp'
968 elif base.startswith('http'):
969 proto = 'http'
970 ext = video.get('ext')
971 if proto == 'm3u8':
972 return (self._extract_m3u8_formats(src, video_id, ext), rtmp_count)
973 elif proto == 'rtmp':
974 rtmp_count += 1
975 streamer = video.get('streamer') or base
976 return ([{
977 'url': streamer,
978 'play_path': src,
979 'ext': 'flv',
980 'format_id': 'rtmp-%d' % (rtmp_count if bitrate is None else bitrate),
981 'tbr': bitrate,
982 'width': width,
983 'height': height,
984 }], rtmp_count)
985 elif proto.startswith('http'):
986 return ([{
987 'url': base + src,
988 'ext': ext or 'flv',
989 'tbr': bitrate,
990 'width': width,
991 'height': height,
992 }], rtmp_count)
993
994 def _live_title(self, name):
995 """ Generate the title for a live video """
996 now = datetime.datetime.now()
997 now_str = now.strftime("%Y-%m-%d %H:%M")
998 return name + ' ' + now_str
999
1000 def _int(self, v, name, fatal=False, **kwargs):
1001 res = int_or_none(v, **kwargs)
1002 if 'get_attr' in kwargs:
1003 print(getattr(v, kwargs['get_attr']))
1004 if res is None:
1005 msg = 'Failed to extract %s: Could not parse value %r' % (name, v)
1006 if fatal:
1007 raise ExtractorError(msg)
1008 else:
1009 self._downloader.report_warning(msg)
1010 return res
1011
1012 def _float(self, v, name, fatal=False, **kwargs):
1013 res = float_or_none(v, **kwargs)
1014 if res is None:
1015 msg = 'Failed to extract %s: Could not parse value %r' % (name, v)
1016 if fatal:
1017 raise ExtractorError(msg)
1018 else:
1019 self._downloader.report_warning(msg)
1020 return res
1021
1022 def _set_cookie(self, domain, name, value, expire_time=None):
1023 cookie = compat_cookiejar.Cookie(
1024 0, name, value, None, None, domain, None,
1025 None, '/', True, False, expire_time, '', None, None, None)
1026 self._downloader.cookiejar.set_cookie(cookie)
1027
1028 def get_testcases(self, include_onlymatching=False):
1029 t = getattr(self, '_TEST', None)
1030 if t:
1031 assert not hasattr(self, '_TESTS'), \
1032 '%s has _TEST and _TESTS' % type(self).__name__
1033 tests = [t]
1034 else:
1035 tests = getattr(self, '_TESTS', [])
1036 for t in tests:
1037 if not include_onlymatching and t.get('only_matching', False):
1038 continue
1039 t['name'] = type(self).__name__[:-len('IE')]
1040 yield t
1041
1042 def is_suitable(self, age_limit):
1043 """ Test whether the extractor is generally suitable for the given
1044 age limit (i.e. pornographic sites are not, all others usually are) """
1045
1046 any_restricted = False
1047 for tc in self.get_testcases(include_onlymatching=False):
1048 if 'playlist' in tc:
1049 tc = tc['playlist'][0]
1050 is_restricted = age_restricted(
1051 tc.get('info_dict', {}).get('age_limit'), age_limit)
1052 if not is_restricted:
1053 return True
1054 any_restricted = any_restricted or is_restricted
1055 return not any_restricted
1056
1057 def extract_subtitles(self, *args, **kwargs):
1058 if (self._downloader.params.get('writesubtitles', False) or
1059 self._downloader.params.get('listsubtitles')):
1060 return self._get_subtitles(*args, **kwargs)
1061 return {}
1062
1063 def _get_subtitles(self, *args, **kwargs):
1064 raise NotImplementedError("This method must be implemented by subclasses")
1065
1066 def extract_automatic_captions(self, *args, **kwargs):
1067 if (self._downloader.params.get('writeautomaticsub', False) or
1068 self._downloader.params.get('listsubtitles')):
1069 return self._get_automatic_captions(*args, **kwargs)
1070 return {}
1071
1072 def _get_automatic_captions(self, *args, **kwargs):
1073 raise NotImplementedError("This method must be implemented by subclasses")
1074
1075
1076 class SearchInfoExtractor(InfoExtractor):
1077 """
1078 Base class for paged search queries extractors.
1079 They accept urls in the format _SEARCH_KEY(|all|[0-9]):{query}
1080 Instances should define _SEARCH_KEY and _MAX_RESULTS.
1081 """
1082
1083 @classmethod
1084 def _make_valid_url(cls):
1085 return r'%s(?P<prefix>|[1-9][0-9]*|all):(?P<query>[\s\S]+)' % cls._SEARCH_KEY
1086
1087 @classmethod
1088 def suitable(cls, url):
1089 return re.match(cls._make_valid_url(), url) is not None
1090
1091 def _real_extract(self, query):
1092 mobj = re.match(self._make_valid_url(), query)
1093 if mobj is None:
1094 raise ExtractorError('Invalid search query "%s"' % query)
1095
1096 prefix = mobj.group('prefix')
1097 query = mobj.group('query')
1098 if prefix == '':
1099 return self._get_n_results(query, 1)
1100 elif prefix == 'all':
1101 return self._get_n_results(query, self._MAX_RESULTS)
1102 else:
1103 n = int(prefix)
1104 if n <= 0:
1105 raise ExtractorError('invalid download number %s for query "%s"' % (n, query))
1106 elif n > self._MAX_RESULTS:
1107 self._downloader.report_warning('%s returns max %i results (you requested %i)' % (self._SEARCH_KEY, self._MAX_RESULTS, n))
1108 n = self._MAX_RESULTS
1109 return self._get_n_results(query, n)
1110
1111 def _get_n_results(self, query, n):
1112 """Get a specified number of results for a query"""
1113 raise NotImplementedError("This method must be implemented by subclasses")
1114
1115 @property
1116 def SEARCH_KEY(self):
1117 return self._SEARCH_KEY