]> Raphaƫl G. Git Repositories - youtubedl/blob - youtube_dl/extractor/common.py
debian/rules: Adapt to new upstream build system.
[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 math
14
15 from ..compat import (
16 compat_cookiejar,
17 compat_cookies,
18 compat_etree_fromstring,
19 compat_getpass,
20 compat_http_client,
21 compat_os_name,
22 compat_str,
23 compat_urllib_error,
24 compat_urllib_parse_urlencode,
25 compat_urllib_request,
26 compat_urlparse,
27 )
28 from ..downloader.f4m import remove_encrypted_media
29 from ..utils import (
30 NO_DEFAULT,
31 age_restricted,
32 bug_reports_message,
33 clean_html,
34 compiled_regex_type,
35 determine_ext,
36 error_to_compat_str,
37 ExtractorError,
38 fix_xml_ampersands,
39 float_or_none,
40 int_or_none,
41 parse_iso8601,
42 RegexNotFoundError,
43 sanitize_filename,
44 sanitized_Request,
45 unescapeHTML,
46 unified_strdate,
47 url_basename,
48 xpath_element,
49 xpath_text,
50 xpath_with_ns,
51 determine_protocol,
52 parse_duration,
53 mimetype2ext,
54 update_Request,
55 update_url_query,
56 parse_m3u8_attributes,
57 )
58
59
60 class InfoExtractor(object):
61 """Information Extractor class.
62
63 Information extractors are the classes that, given a URL, extract
64 information about the video (or videos) the URL refers to. This
65 information includes the real video URL, the video title, author and
66 others. The information is stored in a dictionary which is then
67 passed to the YoutubeDL. The YoutubeDL processes this
68 information possibly downloading the video to the file system, among
69 other possible outcomes.
70
71 The type field determines the type of the result.
72 By far the most common value (and the default if _type is missing) is
73 "video", which indicates a single video.
74
75 For a video, the dictionaries must include the following fields:
76
77 id: Video identifier.
78 title: Video title, unescaped.
79
80 Additionally, it must contain either a formats entry or a url one:
81
82 formats: A list of dictionaries for each format available, ordered
83 from worst to best quality.
84
85 Potential fields:
86 * url Mandatory. The URL of the video file
87 * ext Will be calculated from URL if missing
88 * format A human-readable description of the format
89 ("mp4 container with h264/opus").
90 Calculated from the format_id, width, height.
91 and format_note fields if missing.
92 * format_id A short description of the format
93 ("mp4_h264_opus" or "19").
94 Technically optional, but strongly recommended.
95 * format_note Additional info about the format
96 ("3D" or "DASH video")
97 * width Width of the video, if known
98 * height Height of the video, if known
99 * resolution Textual description of width and height
100 * tbr Average bitrate of audio and video in KBit/s
101 * abr Average audio bitrate in KBit/s
102 * acodec Name of the audio codec in use
103 * asr Audio sampling rate in Hertz
104 * vbr Average video bitrate in KBit/s
105 * fps Frame rate
106 * vcodec Name of the video codec in use
107 * container Name of the container format
108 * filesize The number of bytes, if known in advance
109 * filesize_approx An estimate for the number of bytes
110 * player_url SWF Player URL (used for rtmpdump).
111 * protocol The protocol that will be used for the actual
112 download, lower-case.
113 "http", "https", "rtsp", "rtmp", "rtmpe",
114 "m3u8", "m3u8_native" or "http_dash_segments".
115 * preference Order number of this format. If this field is
116 present and not None, the formats get sorted
117 by this field, regardless of all other values.
118 -1 for default (order by other properties),
119 -2 or smaller for less than default.
120 < -1000 to hide the format (if there is
121 another one which is strictly better)
122 * language Language code, e.g. "de" or "en-US".
123 * language_preference Is this in the language mentioned in
124 the URL?
125 10 if it's what the URL is about,
126 -1 for default (don't know),
127 -10 otherwise, other values reserved for now.
128 * quality Order number of the video quality of this
129 format, irrespective of the file format.
130 -1 for default (order by other properties),
131 -2 or smaller for less than default.
132 * source_preference Order number for this video source
133 (quality takes higher priority)
134 -1 for default (order by other properties),
135 -2 or smaller for less than default.
136 * http_headers A dictionary of additional HTTP headers
137 to add to the request.
138 * stretched_ratio If given and not 1, indicates that the
139 video's pixels are not square.
140 width : height ratio as float.
141 * no_resume The server does not support resuming the
142 (HTTP or RTMP) download. Boolean.
143
144 url: Final video URL.
145 ext: Video filename extension.
146 format: The video format, defaults to ext (used for --get-format)
147 player_url: SWF Player URL (used for rtmpdump).
148
149 The following fields are optional:
150
151 alt_title: A secondary title of the video.
152 display_id An alternative identifier for the video, not necessarily
153 unique, but available before title. Typically, id is
154 something like "4234987", title "Dancing naked mole rats",
155 and display_id "dancing-naked-mole-rats"
156 thumbnails: A list of dictionaries, with the following entries:
157 * "id" (optional, string) - Thumbnail format ID
158 * "url"
159 * "preference" (optional, int) - quality of the image
160 * "width" (optional, int)
161 * "height" (optional, int)
162 * "resolution" (optional, string "{width}x{height"},
163 deprecated)
164 thumbnail: Full URL to a video thumbnail image.
165 description: Full video description.
166 uploader: Full name of the video uploader.
167 license: License name the video is licensed under.
168 creator: The creator of the video.
169 release_date: The date (YYYYMMDD) when the video was released.
170 timestamp: UNIX timestamp of the moment the video became available.
171 upload_date: Video upload date (YYYYMMDD).
172 If not explicitly set, calculated from timestamp.
173 uploader_id: Nickname or id of the video uploader.
174 uploader_url: Full URL to a personal webpage of the video uploader.
175 location: Physical location where the video was filmed.
176 subtitles: The available subtitles as a dictionary in the format
177 {language: subformats}. "subformats" is a list sorted from
178 lower to higher preference, each element is a dictionary
179 with the "ext" entry and one of:
180 * "data": The subtitles file contents
181 * "url": A URL pointing to the subtitles file
182 "ext" will be calculated from URL if missing
183 automatic_captions: Like 'subtitles', used by the YoutubeIE for
184 automatically generated captions
185 duration: Length of the video in seconds, as an integer or float.
186 view_count: How many users have watched the video on the platform.
187 like_count: Number of positive ratings of the video
188 dislike_count: Number of negative ratings of the video
189 repost_count: Number of reposts of the video
190 average_rating: Average rating give by users, the scale used depends on the webpage
191 comment_count: Number of comments on the video
192 comments: A list of comments, each with one or more of the following
193 properties (all but one of text or html optional):
194 * "author" - human-readable name of the comment author
195 * "author_id" - user ID of the comment author
196 * "id" - Comment ID
197 * "html" - Comment as HTML
198 * "text" - Plain text of the comment
199 * "timestamp" - UNIX timestamp of comment
200 * "parent" - ID of the comment this one is replying to.
201 Set to "root" to indicate that this is a
202 comment to the original video.
203 age_limit: Age restriction for the video, as an integer (years)
204 webpage_url: The URL to the video webpage, if given to youtube-dl it
205 should allow to get the same result again. (It will be set
206 by YoutubeDL if it's missing)
207 categories: A list of categories that the video falls in, for example
208 ["Sports", "Berlin"]
209 tags: A list of tags assigned to the video, e.g. ["sweden", "pop music"]
210 is_live: True, False, or None (=unknown). Whether this video is a
211 live stream that goes on instead of a fixed-length video.
212 start_time: Time in seconds where the reproduction should start, as
213 specified in the URL.
214 end_time: Time in seconds where the reproduction should end, as
215 specified in the URL.
216
217 The following fields should only be used when the video belongs to some logical
218 chapter or section:
219
220 chapter: Name or title of the chapter the video belongs to.
221 chapter_number: Number of the chapter the video belongs to, as an integer.
222 chapter_id: Id of the chapter the video belongs to, as a unicode string.
223
224 The following fields should only be used when the video is an episode of some
225 series or programme:
226
227 series: Title of the series or programme the video episode belongs to.
228 season: Title of the season the video episode belongs to.
229 season_number: Number of the season the video episode belongs to, as an integer.
230 season_id: Id of the season the video episode belongs to, as a unicode string.
231 episode: Title of the video episode. Unlike mandatory video title field,
232 this field should denote the exact title of the video episode
233 without any kind of decoration.
234 episode_number: Number of the video episode within a season, as an integer.
235 episode_id: Id of the video episode, as a unicode string.
236
237 The following fields should only be used when the media is a track or a part of
238 a music album:
239
240 track: Title of the track.
241 track_number: Number of the track within an album or a disc, as an integer.
242 track_id: Id of the track (useful in case of custom indexing, e.g. 6.iii),
243 as a unicode string.
244 artist: Artist(s) of the track.
245 genre: Genre(s) of the track.
246 album: Title of the album the track belongs to.
247 album_type: Type of the album (e.g. "Demo", "Full-length", "Split", "Compilation", etc).
248 album_artist: List of all artists appeared on the album (e.g.
249 "Ash Borer / Fell Voices" or "Various Artists", useful for splits
250 and compilations).
251 disc_number: Number of the disc or other physical medium the track belongs to,
252 as an integer.
253 release_year: Year (YYYY) when the album was released.
254
255 Unless mentioned otherwise, the fields should be Unicode strings.
256
257 Unless mentioned otherwise, None is equivalent to absence of information.
258
259
260 _type "playlist" indicates multiple videos.
261 There must be a key "entries", which is a list, an iterable, or a PagedList
262 object, each element of which is a valid dictionary by this specification.
263
264 Additionally, playlists can have "title", "description" and "id" attributes
265 with the same semantics as videos (see above).
266
267
268 _type "multi_video" indicates that there are multiple videos that
269 form a single show, for examples multiple acts of an opera or TV episode.
270 It must have an entries key like a playlist and contain all the keys
271 required for a video at the same time.
272
273
274 _type "url" indicates that the video must be extracted from another
275 location, possibly by a different extractor. Its only required key is:
276 "url" - the next URL to extract.
277 The key "ie_key" can be set to the class name (minus the trailing "IE",
278 e.g. "Youtube") if the extractor class is known in advance.
279 Additionally, the dictionary may have any properties of the resolved entity
280 known in advance, for example "title" if the title of the referred video is
281 known ahead of time.
282
283
284 _type "url_transparent" entities have the same specification as "url", but
285 indicate that the given additional information is more precise than the one
286 associated with the resolved URL.
287 This is useful when a site employs a video service that hosts the video and
288 its technical metadata, but that video service does not embed a useful
289 title, description etc.
290
291
292 Subclasses of this one should re-define the _real_initialize() and
293 _real_extract() methods and define a _VALID_URL regexp.
294 Probably, they should also be added to the list of extractors.
295
296 Finally, the _WORKING attribute should be set to False for broken IEs
297 in order to warn the users and skip the tests.
298 """
299
300 _ready = False
301 _downloader = None
302 _WORKING = True
303
304 def __init__(self, downloader=None):
305 """Constructor. Receives an optional downloader."""
306 self._ready = False
307 self.set_downloader(downloader)
308
309 @classmethod
310 def suitable(cls, url):
311 """Receives a URL and returns True if suitable for this IE."""
312
313 # This does not use has/getattr intentionally - we want to know whether
314 # we have cached the regexp for *this* class, whereas getattr would also
315 # match the superclass
316 if '_VALID_URL_RE' not in cls.__dict__:
317 cls._VALID_URL_RE = re.compile(cls._VALID_URL)
318 return cls._VALID_URL_RE.match(url) is not None
319
320 @classmethod
321 def _match_id(cls, url):
322 if '_VALID_URL_RE' not in cls.__dict__:
323 cls._VALID_URL_RE = re.compile(cls._VALID_URL)
324 m = cls._VALID_URL_RE.match(url)
325 assert m
326 return m.group('id')
327
328 @classmethod
329 def working(cls):
330 """Getter method for _WORKING."""
331 return cls._WORKING
332
333 def initialize(self):
334 """Initializes an instance (authentication, etc)."""
335 if not self._ready:
336 self._real_initialize()
337 self._ready = True
338
339 def extract(self, url):
340 """Extracts URL information and returns it in list of dicts."""
341 try:
342 self.initialize()
343 return self._real_extract(url)
344 except ExtractorError:
345 raise
346 except compat_http_client.IncompleteRead as e:
347 raise ExtractorError('A network error has occurred.', cause=e, expected=True)
348 except (KeyError, StopIteration) as e:
349 raise ExtractorError('An extractor error has occurred.', cause=e)
350
351 def set_downloader(self, downloader):
352 """Sets the downloader for this IE."""
353 self._downloader = downloader
354
355 def _real_initialize(self):
356 """Real initialization process. Redefine in subclasses."""
357 pass
358
359 def _real_extract(self, url):
360 """Real extraction process. Redefine in subclasses."""
361 pass
362
363 @classmethod
364 def ie_key(cls):
365 """A string for getting the InfoExtractor with get_info_extractor"""
366 return compat_str(cls.__name__[:-2])
367
368 @property
369 def IE_NAME(self):
370 return compat_str(type(self).__name__[:-2])
371
372 def _request_webpage(self, url_or_request, video_id, note=None, errnote=None, fatal=True, data=None, headers={}, query={}):
373 """ Returns the response handle """
374 if note is None:
375 self.report_download_webpage(video_id)
376 elif note is not False:
377 if video_id is None:
378 self.to_screen('%s' % (note,))
379 else:
380 self.to_screen('%s: %s' % (video_id, note))
381 if isinstance(url_or_request, compat_urllib_request.Request):
382 url_or_request = update_Request(
383 url_or_request, data=data, headers=headers, query=query)
384 else:
385 if query:
386 url_or_request = update_url_query(url_or_request, query)
387 if data is not None or headers:
388 url_or_request = sanitized_Request(url_or_request, data, headers)
389 try:
390 return self._downloader.urlopen(url_or_request)
391 except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
392 if errnote is False:
393 return False
394 if errnote is None:
395 errnote = 'Unable to download webpage'
396
397 errmsg = '%s: %s' % (errnote, error_to_compat_str(err))
398 if fatal:
399 raise ExtractorError(errmsg, sys.exc_info()[2], cause=err)
400 else:
401 self._downloader.report_warning(errmsg)
402 return False
403
404 def _download_webpage_handle(self, url_or_request, video_id, note=None, errnote=None, fatal=True, encoding=None, data=None, headers={}, query={}):
405 """ Returns a tuple (page content as string, URL handle) """
406 # Strip hashes from the URL (#1038)
407 if isinstance(url_or_request, (compat_str, str)):
408 url_or_request = url_or_request.partition('#')[0]
409
410 urlh = self._request_webpage(url_or_request, video_id, note, errnote, fatal, data=data, headers=headers, query=query)
411 if urlh is False:
412 assert not fatal
413 return False
414 content = self._webpage_read_content(urlh, url_or_request, video_id, note, errnote, fatal, encoding=encoding)
415 return (content, urlh)
416
417 @staticmethod
418 def _guess_encoding_from_content(content_type, webpage_bytes):
419 m = re.match(r'[a-zA-Z0-9_.-]+/[a-zA-Z0-9_.-]+\s*;\s*charset=(.+)', content_type)
420 if m:
421 encoding = m.group(1)
422 else:
423 m = re.search(br'<meta[^>]+charset=[\'"]?([^\'")]+)[ /\'">]',
424 webpage_bytes[:1024])
425 if m:
426 encoding = m.group(1).decode('ascii')
427 elif webpage_bytes.startswith(b'\xff\xfe'):
428 encoding = 'utf-16'
429 else:
430 encoding = 'utf-8'
431
432 return encoding
433
434 def _webpage_read_content(self, urlh, url_or_request, video_id, note=None, errnote=None, fatal=True, prefix=None, encoding=None):
435 content_type = urlh.headers.get('Content-Type', '')
436 webpage_bytes = urlh.read()
437 if prefix is not None:
438 webpage_bytes = prefix + webpage_bytes
439 if not encoding:
440 encoding = self._guess_encoding_from_content(content_type, webpage_bytes)
441 if self._downloader.params.get('dump_intermediate_pages', False):
442 try:
443 url = url_or_request.get_full_url()
444 except AttributeError:
445 url = url_or_request
446 self.to_screen('Dumping request to ' + url)
447 dump = base64.b64encode(webpage_bytes).decode('ascii')
448 self._downloader.to_screen(dump)
449 if self._downloader.params.get('write_pages', False):
450 try:
451 url = url_or_request.get_full_url()
452 except AttributeError:
453 url = url_or_request
454 basen = '%s_%s' % (video_id, url)
455 if len(basen) > 240:
456 h = '___' + hashlib.md5(basen.encode('utf-8')).hexdigest()
457 basen = basen[:240 - len(h)] + h
458 raw_filename = basen + '.dump'
459 filename = sanitize_filename(raw_filename, restricted=True)
460 self.to_screen('Saving request to ' + filename)
461 # Working around MAX_PATH limitation on Windows (see
462 # http://msdn.microsoft.com/en-us/library/windows/desktop/aa365247(v=vs.85).aspx)
463 if compat_os_name == 'nt':
464 absfilepath = os.path.abspath(filename)
465 if len(absfilepath) > 259:
466 filename = '\\\\?\\' + absfilepath
467 with open(filename, 'wb') as outf:
468 outf.write(webpage_bytes)
469
470 try:
471 content = webpage_bytes.decode(encoding, 'replace')
472 except LookupError:
473 content = webpage_bytes.decode('utf-8', 'replace')
474
475 if ('<title>Access to this site is blocked</title>' in content and
476 'Websense' in content[:512]):
477 msg = 'Access to this webpage has been blocked by Websense filtering software in your network.'
478 blocked_iframe = self._html_search_regex(
479 r'<iframe src="([^"]+)"', content,
480 'Websense information URL', default=None)
481 if blocked_iframe:
482 msg += ' Visit %s for more details' % blocked_iframe
483 raise ExtractorError(msg, expected=True)
484 if '<title>The URL you requested has been blocked</title>' in content[:512]:
485 msg = (
486 'Access to this webpage has been blocked by Indian censorship. '
487 'Use a VPN or proxy server (with --proxy) to route around it.')
488 block_msg = self._html_search_regex(
489 r'</h1><p>(.*?)</p>',
490 content, 'block message', default=None)
491 if block_msg:
492 msg += ' (Message: "%s")' % block_msg.replace('\n', ' ')
493 raise ExtractorError(msg, expected=True)
494
495 return content
496
497 def _download_webpage(self, url_or_request, video_id, note=None, errnote=None, fatal=True, tries=1, timeout=5, encoding=None, data=None, headers={}, query={}):
498 """ Returns the data of the page as a string """
499 success = False
500 try_count = 0
501 while success is False:
502 try:
503 res = self._download_webpage_handle(url_or_request, video_id, note, errnote, fatal, encoding=encoding, data=data, headers=headers, query=query)
504 success = True
505 except compat_http_client.IncompleteRead as e:
506 try_count += 1
507 if try_count >= tries:
508 raise e
509 self._sleep(timeout, video_id)
510 if res is False:
511 return res
512 else:
513 content, _ = res
514 return content
515
516 def _download_xml(self, url_or_request, video_id,
517 note='Downloading XML', errnote='Unable to download XML',
518 transform_source=None, fatal=True, encoding=None, data=None, headers={}, query={}):
519 """Return the xml as an xml.etree.ElementTree.Element"""
520 xml_string = self._download_webpage(
521 url_or_request, video_id, note, errnote, fatal=fatal, encoding=encoding, data=data, headers=headers, query=query)
522 if xml_string is False:
523 return xml_string
524 if transform_source:
525 xml_string = transform_source(xml_string)
526 return compat_etree_fromstring(xml_string.encode('utf-8'))
527
528 def _download_json(self, url_or_request, video_id,
529 note='Downloading JSON metadata',
530 errnote='Unable to download JSON metadata',
531 transform_source=None,
532 fatal=True, encoding=None, data=None, headers={}, query={}):
533 json_string = self._download_webpage(
534 url_or_request, video_id, note, errnote, fatal=fatal,
535 encoding=encoding, data=data, headers=headers, query=query)
536 if (not fatal) and json_string is False:
537 return None
538 return self._parse_json(
539 json_string, video_id, transform_source=transform_source, fatal=fatal)
540
541 def _parse_json(self, json_string, video_id, transform_source=None, fatal=True):
542 if transform_source:
543 json_string = transform_source(json_string)
544 try:
545 return json.loads(json_string)
546 except ValueError as ve:
547 errmsg = '%s: Failed to parse JSON ' % video_id
548 if fatal:
549 raise ExtractorError(errmsg, cause=ve)
550 else:
551 self.report_warning(errmsg + str(ve))
552
553 def report_warning(self, msg, video_id=None):
554 idstr = '' if video_id is None else '%s: ' % video_id
555 self._downloader.report_warning(
556 '[%s] %s%s' % (self.IE_NAME, idstr, msg))
557
558 def to_screen(self, msg):
559 """Print msg to screen, prefixing it with '[ie_name]'"""
560 self._downloader.to_screen('[%s] %s' % (self.IE_NAME, msg))
561
562 def report_extraction(self, id_or_name):
563 """Report information extraction."""
564 self.to_screen('%s: Extracting information' % id_or_name)
565
566 def report_download_webpage(self, video_id):
567 """Report webpage download."""
568 self.to_screen('%s: Downloading webpage' % video_id)
569
570 def report_age_confirmation(self):
571 """Report attempt to confirm age."""
572 self.to_screen('Confirming age')
573
574 def report_login(self):
575 """Report attempt to log in."""
576 self.to_screen('Logging in')
577
578 @staticmethod
579 def raise_login_required(msg='This video is only available for registered users'):
580 raise ExtractorError(
581 '%s. Use --username and --password or --netrc to provide account credentials.' % msg,
582 expected=True)
583
584 @staticmethod
585 def raise_geo_restricted(msg='This video is not available from your location due to geo restriction'):
586 raise ExtractorError(
587 '%s. You might want to use --proxy to workaround.' % msg,
588 expected=True)
589
590 # Methods for following #608
591 @staticmethod
592 def url_result(url, ie=None, video_id=None, video_title=None):
593 """Returns a URL that points to a page that should be processed"""
594 # TODO: ie should be the class used for getting the info
595 video_info = {'_type': 'url',
596 'url': url,
597 'ie_key': ie}
598 if video_id is not None:
599 video_info['id'] = video_id
600 if video_title is not None:
601 video_info['title'] = video_title
602 return video_info
603
604 @staticmethod
605 def playlist_result(entries, playlist_id=None, playlist_title=None, playlist_description=None):
606 """Returns a playlist"""
607 video_info = {'_type': 'playlist',
608 'entries': entries}
609 if playlist_id:
610 video_info['id'] = playlist_id
611 if playlist_title:
612 video_info['title'] = playlist_title
613 if playlist_description:
614 video_info['description'] = playlist_description
615 return video_info
616
617 def _search_regex(self, pattern, string, name, default=NO_DEFAULT, fatal=True, flags=0, group=None):
618 """
619 Perform a regex search on the given string, using a single or a list of
620 patterns returning the first matching group.
621 In case of failure return a default value or raise a WARNING or a
622 RegexNotFoundError, depending on fatal, specifying the field name.
623 """
624 if isinstance(pattern, (str, compat_str, compiled_regex_type)):
625 mobj = re.search(pattern, string, flags)
626 else:
627 for p in pattern:
628 mobj = re.search(p, string, flags)
629 if mobj:
630 break
631
632 if not self._downloader.params.get('no_color') and compat_os_name != 'nt' and sys.stderr.isatty():
633 _name = '\033[0;34m%s\033[0m' % name
634 else:
635 _name = name
636
637 if mobj:
638 if group is None:
639 # return the first matching group
640 return next(g for g in mobj.groups() if g is not None)
641 else:
642 return mobj.group(group)
643 elif default is not NO_DEFAULT:
644 return default
645 elif fatal:
646 raise RegexNotFoundError('Unable to extract %s' % _name)
647 else:
648 self._downloader.report_warning('unable to extract %s' % _name + bug_reports_message())
649 return None
650
651 def _html_search_regex(self, pattern, string, name, default=NO_DEFAULT, fatal=True, flags=0, group=None):
652 """
653 Like _search_regex, but strips HTML tags and unescapes entities.
654 """
655 res = self._search_regex(pattern, string, name, default, fatal, flags, group)
656 if res:
657 return clean_html(res).strip()
658 else:
659 return res
660
661 def _get_login_info(self):
662 """
663 Get the login info as (username, password)
664 It will look in the netrc file using the _NETRC_MACHINE value
665 If there's no info available, return (None, None)
666 """
667 if self._downloader is None:
668 return (None, None)
669
670 username = None
671 password = None
672 downloader_params = self._downloader.params
673
674 # Attempt to use provided username and password or .netrc data
675 if downloader_params.get('username') is not None:
676 username = downloader_params['username']
677 password = downloader_params['password']
678 elif downloader_params.get('usenetrc', False):
679 try:
680 info = netrc.netrc().authenticators(self._NETRC_MACHINE)
681 if info is not None:
682 username = info[0]
683 password = info[2]
684 else:
685 raise netrc.NetrcParseError('No authenticators for %s' % self._NETRC_MACHINE)
686 except (IOError, netrc.NetrcParseError) as err:
687 self._downloader.report_warning('parsing .netrc: %s' % error_to_compat_str(err))
688
689 return (username, password)
690
691 def _get_tfa_info(self, note='two-factor verification code'):
692 """
693 Get the two-factor authentication info
694 TODO - asking the user will be required for sms/phone verify
695 currently just uses the command line option
696 If there's no info available, return None
697 """
698 if self._downloader is None:
699 return None
700 downloader_params = self._downloader.params
701
702 if downloader_params.get('twofactor') is not None:
703 return downloader_params['twofactor']
704
705 return compat_getpass('Type %s and press [Return]: ' % note)
706
707 # Helper functions for extracting OpenGraph info
708 @staticmethod
709 def _og_regexes(prop):
710 content_re = r'content=(?:"([^"]+?)"|\'([^\']+?)\'|\s*([^\s"\'=<>`]+?))'
711 property_re = (r'(?:name|property)=(?:\'og:%(prop)s\'|"og:%(prop)s"|\s*og:%(prop)s\b)'
712 % {'prop': re.escape(prop)})
713 template = r'<meta[^>]+?%s[^>]+?%s'
714 return [
715 template % (property_re, content_re),
716 template % (content_re, property_re),
717 ]
718
719 @staticmethod
720 def _meta_regex(prop):
721 return r'''(?isx)<meta
722 (?=[^>]+(?:itemprop|name|property|id|http-equiv)=(["\']?)%s\1)
723 [^>]+?content=(["\'])(?P<content>.*?)\2''' % re.escape(prop)
724
725 def _og_search_property(self, prop, html, name=None, **kargs):
726 if name is None:
727 name = 'OpenGraph %s' % prop
728 escaped = self._search_regex(self._og_regexes(prop), html, name, flags=re.DOTALL, **kargs)
729 if escaped is None:
730 return None
731 return unescapeHTML(escaped)
732
733 def _og_search_thumbnail(self, html, **kargs):
734 return self._og_search_property('image', html, 'thumbnail URL', fatal=False, **kargs)
735
736 def _og_search_description(self, html, **kargs):
737 return self._og_search_property('description', html, fatal=False, **kargs)
738
739 def _og_search_title(self, html, **kargs):
740 return self._og_search_property('title', html, **kargs)
741
742 def _og_search_video_url(self, html, name='video url', secure=True, **kargs):
743 regexes = self._og_regexes('video') + self._og_regexes('video:url')
744 if secure:
745 regexes = self._og_regexes('video:secure_url') + regexes
746 return self._html_search_regex(regexes, html, name, **kargs)
747
748 def _og_search_url(self, html, **kargs):
749 return self._og_search_property('url', html, **kargs)
750
751 def _html_search_meta(self, name, html, display_name=None, fatal=False, **kwargs):
752 if display_name is None:
753 display_name = name
754 return self._html_search_regex(
755 self._meta_regex(name),
756 html, display_name, fatal=fatal, group='content', **kwargs)
757
758 def _dc_search_uploader(self, html):
759 return self._html_search_meta('dc.creator', html, 'uploader')
760
761 def _rta_search(self, html):
762 # See http://www.rtalabel.org/index.php?content=howtofaq#single
763 if re.search(r'(?ix)<meta\s+name="rating"\s+'
764 r' content="RTA-5042-1996-1400-1577-RTA"',
765 html):
766 return 18
767 return 0
768
769 def _media_rating_search(self, html):
770 # See http://www.tjg-designs.com/WP/metadata-code-examples-adding-metadata-to-your-web-pages/
771 rating = self._html_search_meta('rating', html)
772
773 if not rating:
774 return None
775
776 RATING_TABLE = {
777 'safe for kids': 0,
778 'general': 8,
779 '14 years': 14,
780 'mature': 17,
781 'restricted': 19,
782 }
783 return RATING_TABLE.get(rating.lower())
784
785 def _family_friendly_search(self, html):
786 # See http://schema.org/VideoObject
787 family_friendly = self._html_search_meta('isFamilyFriendly', html)
788
789 if not family_friendly:
790 return None
791
792 RATING_TABLE = {
793 '1': 0,
794 'true': 0,
795 '0': 18,
796 'false': 18,
797 }
798 return RATING_TABLE.get(family_friendly.lower())
799
800 def _twitter_search_player(self, html):
801 return self._html_search_meta('twitter:player', html,
802 'twitter card player')
803
804 def _search_json_ld(self, html, video_id, **kwargs):
805 json_ld = self._search_regex(
806 r'(?s)<script[^>]+type=(["\'])application/ld\+json\1[^>]*>(?P<json_ld>.+?)</script>',
807 html, 'JSON-LD', group='json_ld', **kwargs)
808 if not json_ld:
809 return {}
810 return self._json_ld(json_ld, video_id, fatal=kwargs.get('fatal', True))
811
812 def _json_ld(self, json_ld, video_id, fatal=True):
813 if isinstance(json_ld, compat_str):
814 json_ld = self._parse_json(json_ld, video_id, fatal=fatal)
815 if not json_ld:
816 return {}
817 info = {}
818 if json_ld.get('@context') == 'http://schema.org':
819 item_type = json_ld.get('@type')
820 if item_type == 'TVEpisode':
821 info.update({
822 'episode': unescapeHTML(json_ld.get('name')),
823 'episode_number': int_or_none(json_ld.get('episodeNumber')),
824 'description': unescapeHTML(json_ld.get('description')),
825 })
826 part_of_season = json_ld.get('partOfSeason')
827 if isinstance(part_of_season, dict) and part_of_season.get('@type') == 'TVSeason':
828 info['season_number'] = int_or_none(part_of_season.get('seasonNumber'))
829 part_of_series = json_ld.get('partOfSeries')
830 if isinstance(part_of_series, dict) and part_of_series.get('@type') == 'TVSeries':
831 info['series'] = unescapeHTML(part_of_series.get('name'))
832 elif item_type == 'Article':
833 info.update({
834 'timestamp': parse_iso8601(json_ld.get('datePublished')),
835 'title': unescapeHTML(json_ld.get('headline')),
836 'description': unescapeHTML(json_ld.get('articleBody')),
837 })
838 return dict((k, v) for k, v in info.items() if v is not None)
839
840 @staticmethod
841 def _hidden_inputs(html):
842 html = re.sub(r'<!--(?:(?!<!--).)*-->', '', html)
843 hidden_inputs = {}
844 for input in re.findall(r'(?i)<input([^>]+)>', html):
845 if not re.search(r'type=(["\'])(?:hidden|submit)\1', input):
846 continue
847 name = re.search(r'(?:name|id)=(["\'])(?P<value>.+?)\1', input)
848 if not name:
849 continue
850 value = re.search(r'value=(["\'])(?P<value>.*?)\1', input)
851 if not value:
852 continue
853 hidden_inputs[name.group('value')] = value.group('value')
854 return hidden_inputs
855
856 def _form_hidden_inputs(self, form_id, html):
857 form = self._search_regex(
858 r'(?is)<form[^>]+?id=(["\'])%s\1[^>]*>(?P<form>.+?)</form>' % form_id,
859 html, '%s form' % form_id, group='form')
860 return self._hidden_inputs(form)
861
862 def _sort_formats(self, formats, field_preference=None):
863 if not formats:
864 raise ExtractorError('No video formats found')
865
866 for f in formats:
867 # Automatically determine tbr when missing based on abr and vbr (improves
868 # formats sorting in some cases)
869 if 'tbr' not in f and f.get('abr') is not None and f.get('vbr') is not None:
870 f['tbr'] = f['abr'] + f['vbr']
871
872 def _formats_key(f):
873 # TODO remove the following workaround
874 from ..utils import determine_ext
875 if not f.get('ext') and 'url' in f:
876 f['ext'] = determine_ext(f['url'])
877
878 if isinstance(field_preference, (list, tuple)):
879 return tuple(f.get(field) if f.get(field) is not None else -1 for field in field_preference)
880
881 preference = f.get('preference')
882 if preference is None:
883 preference = 0
884 if f.get('ext') in ['f4f', 'f4m']: # Not yet supported
885 preference -= 0.5
886
887 proto_preference = 0 if determine_protocol(f) in ['http', 'https'] else -0.1
888
889 if f.get('vcodec') == 'none': # audio only
890 preference -= 50
891 if self._downloader.params.get('prefer_free_formats'):
892 ORDER = ['aac', 'mp3', 'm4a', 'webm', 'ogg', 'opus']
893 else:
894 ORDER = ['webm', 'opus', 'ogg', 'mp3', 'aac', 'm4a']
895 ext_preference = 0
896 try:
897 audio_ext_preference = ORDER.index(f['ext'])
898 except ValueError:
899 audio_ext_preference = -1
900 else:
901 if f.get('acodec') == 'none': # video only
902 preference -= 40
903 if self._downloader.params.get('prefer_free_formats'):
904 ORDER = ['flv', 'mp4', 'webm']
905 else:
906 ORDER = ['webm', 'flv', 'mp4']
907 try:
908 ext_preference = ORDER.index(f['ext'])
909 except ValueError:
910 ext_preference = -1
911 audio_ext_preference = 0
912
913 return (
914 preference,
915 f.get('language_preference') if f.get('language_preference') is not None else -1,
916 f.get('quality') if f.get('quality') is not None else -1,
917 f.get('tbr') if f.get('tbr') is not None else -1,
918 f.get('filesize') if f.get('filesize') is not None else -1,
919 f.get('vbr') if f.get('vbr') is not None else -1,
920 f.get('height') if f.get('height') is not None else -1,
921 f.get('width') if f.get('width') is not None else -1,
922 proto_preference,
923 ext_preference,
924 f.get('abr') if f.get('abr') is not None else -1,
925 audio_ext_preference,
926 f.get('fps') if f.get('fps') is not None else -1,
927 f.get('filesize_approx') if f.get('filesize_approx') is not None else -1,
928 f.get('source_preference') if f.get('source_preference') is not None else -1,
929 f.get('format_id') if f.get('format_id') is not None else '',
930 )
931 formats.sort(key=_formats_key)
932
933 def _check_formats(self, formats, video_id):
934 if formats:
935 formats[:] = filter(
936 lambda f: self._is_valid_url(
937 f['url'], video_id,
938 item='%s video format' % f.get('format_id') if f.get('format_id') else 'video'),
939 formats)
940
941 @staticmethod
942 def _remove_duplicate_formats(formats):
943 format_urls = set()
944 unique_formats = []
945 for f in formats:
946 if f['url'] not in format_urls:
947 format_urls.add(f['url'])
948 unique_formats.append(f)
949 formats[:] = unique_formats
950
951 def _is_valid_url(self, url, video_id, item='video'):
952 url = self._proto_relative_url(url, scheme='http:')
953 # For now assume non HTTP(S) URLs always valid
954 if not (url.startswith('http://') or url.startswith('https://')):
955 return True
956 try:
957 self._request_webpage(url, video_id, 'Checking %s URL' % item)
958 return True
959 except ExtractorError as e:
960 if isinstance(e.cause, compat_urllib_error.URLError):
961 self.to_screen(
962 '%s: %s URL is invalid, skipping' % (video_id, item))
963 return False
964 raise
965
966 def http_scheme(self):
967 """ Either "http:" or "https:", depending on the user's preferences """
968 return (
969 'http:'
970 if self._downloader.params.get('prefer_insecure', False)
971 else 'https:')
972
973 def _proto_relative_url(self, url, scheme=None):
974 if url is None:
975 return url
976 if url.startswith('//'):
977 if scheme is None:
978 scheme = self.http_scheme()
979 return scheme + url
980 else:
981 return url
982
983 def _sleep(self, timeout, video_id, msg_template=None):
984 if msg_template is None:
985 msg_template = '%(video_id)s: Waiting for %(timeout)s seconds'
986 msg = msg_template % {'video_id': video_id, 'timeout': timeout}
987 self.to_screen(msg)
988 time.sleep(timeout)
989
990 def _extract_f4m_formats(self, manifest_url, video_id, preference=None, f4m_id=None,
991 transform_source=lambda s: fix_xml_ampersands(s).strip(),
992 fatal=True, m3u8_id=None):
993 manifest = self._download_xml(
994 manifest_url, video_id, 'Downloading f4m manifest',
995 'Unable to download f4m manifest',
996 # Some manifests may be malformed, e.g. prosiebensat1 generated manifests
997 # (see https://github.com/rg3/youtube-dl/issues/6215#issuecomment-121704244)
998 transform_source=transform_source,
999 fatal=fatal)
1000
1001 if manifest is False:
1002 return []
1003
1004 return self._parse_f4m_formats(
1005 manifest, manifest_url, video_id, preference=preference, f4m_id=f4m_id,
1006 transform_source=transform_source, fatal=fatal, m3u8_id=m3u8_id)
1007
1008 def _parse_f4m_formats(self, manifest, manifest_url, video_id, preference=None, f4m_id=None,
1009 transform_source=lambda s: fix_xml_ampersands(s).strip(),
1010 fatal=True, m3u8_id=None):
1011 # currently youtube-dl cannot decode the playerVerificationChallenge as Akamai uses Adobe Alchemy
1012 akamai_pv = manifest.find('{http://ns.adobe.com/f4m/1.0}pv-2.0')
1013 if akamai_pv is not None and ';' in akamai_pv.text:
1014 playerVerificationChallenge = akamai_pv.text.split(';')[0]
1015 if playerVerificationChallenge.strip() != '':
1016 return []
1017
1018 formats = []
1019 manifest_version = '1.0'
1020 media_nodes = manifest.findall('{http://ns.adobe.com/f4m/1.0}media')
1021 if not media_nodes:
1022 manifest_version = '2.0'
1023 media_nodes = manifest.findall('{http://ns.adobe.com/f4m/2.0}media')
1024 # Remove unsupported DRM protected media from final formats
1025 # rendition (see https://github.com/rg3/youtube-dl/issues/8573).
1026 media_nodes = remove_encrypted_media(media_nodes)
1027 if not media_nodes:
1028 return formats
1029 base_url = xpath_text(
1030 manifest, ['{http://ns.adobe.com/f4m/1.0}baseURL', '{http://ns.adobe.com/f4m/2.0}baseURL'],
1031 'base URL', default=None)
1032 if base_url:
1033 base_url = base_url.strip()
1034
1035 bootstrap_info = xpath_element(
1036 manifest, ['{http://ns.adobe.com/f4m/1.0}bootstrapInfo', '{http://ns.adobe.com/f4m/2.0}bootstrapInfo'],
1037 'bootstrap info', default=None)
1038
1039 for i, media_el in enumerate(media_nodes):
1040 tbr = int_or_none(media_el.attrib.get('bitrate'))
1041 width = int_or_none(media_el.attrib.get('width'))
1042 height = int_or_none(media_el.attrib.get('height'))
1043 format_id = '-'.join(filter(None, [f4m_id, compat_str(i if tbr is None else tbr)]))
1044 # If <bootstrapInfo> is present, the specified f4m is a
1045 # stream-level manifest, and only set-level manifests may refer to
1046 # external resources. See section 11.4 and section 4 of F4M spec
1047 if bootstrap_info is None:
1048 media_url = None
1049 # @href is introduced in 2.0, see section 11.6 of F4M spec
1050 if manifest_version == '2.0':
1051 media_url = media_el.attrib.get('href')
1052 if media_url is None:
1053 media_url = media_el.attrib.get('url')
1054 if not media_url:
1055 continue
1056 manifest_url = (
1057 media_url if media_url.startswith('http://') or media_url.startswith('https://')
1058 else ((base_url or '/'.join(manifest_url.split('/')[:-1])) + '/' + media_url))
1059 # If media_url is itself a f4m manifest do the recursive extraction
1060 # since bitrates in parent manifest (this one) and media_url manifest
1061 # may differ leading to inability to resolve the format by requested
1062 # bitrate in f4m downloader
1063 ext = determine_ext(manifest_url)
1064 if ext == 'f4m':
1065 f4m_formats = self._extract_f4m_formats(
1066 manifest_url, video_id, preference=preference, f4m_id=f4m_id,
1067 transform_source=transform_source, fatal=fatal)
1068 # Sometimes stream-level manifest contains single media entry that
1069 # does not contain any quality metadata (e.g. http://matchtv.ru/#live-player).
1070 # At the same time parent's media entry in set-level manifest may
1071 # contain it. We will copy it from parent in such cases.
1072 if len(f4m_formats) == 1:
1073 f = f4m_formats[0]
1074 f.update({
1075 'tbr': f.get('tbr') or tbr,
1076 'width': f.get('width') or width,
1077 'height': f.get('height') or height,
1078 'format_id': f.get('format_id') if not tbr else format_id,
1079 })
1080 formats.extend(f4m_formats)
1081 continue
1082 elif ext == 'm3u8':
1083 formats.extend(self._extract_m3u8_formats(
1084 manifest_url, video_id, 'mp4', preference=preference,
1085 m3u8_id=m3u8_id, fatal=fatal))
1086 continue
1087 formats.append({
1088 'format_id': format_id,
1089 'url': manifest_url,
1090 'ext': 'flv' if bootstrap_info is not None else None,
1091 'tbr': tbr,
1092 'width': width,
1093 'height': height,
1094 'preference': preference,
1095 })
1096 return formats
1097
1098 def _m3u8_meta_format(self, m3u8_url, ext=None, preference=None, m3u8_id=None):
1099 return {
1100 'format_id': '-'.join(filter(None, [m3u8_id, 'meta'])),
1101 'url': m3u8_url,
1102 'ext': ext,
1103 'protocol': 'm3u8',
1104 'preference': preference - 1 if preference else -1,
1105 'resolution': 'multiple',
1106 'format_note': 'Quality selection URL',
1107 }
1108
1109 def _extract_m3u8_formats(self, m3u8_url, video_id, ext=None,
1110 entry_protocol='m3u8', preference=None,
1111 m3u8_id=None, note=None, errnote=None,
1112 fatal=True, live=False):
1113
1114 formats = [self._m3u8_meta_format(m3u8_url, ext, preference, m3u8_id)]
1115
1116 format_url = lambda u: (
1117 u
1118 if re.match(r'^https?://', u)
1119 else compat_urlparse.urljoin(m3u8_url, u))
1120
1121 res = self._download_webpage_handle(
1122 m3u8_url, video_id,
1123 note=note or 'Downloading m3u8 information',
1124 errnote=errnote or 'Failed to download m3u8 information',
1125 fatal=fatal)
1126 if res is False:
1127 return []
1128 m3u8_doc, urlh = res
1129 m3u8_url = urlh.geturl()
1130
1131 # We should try extracting formats only from master playlists [1], i.e.
1132 # playlists that describe available qualities. On the other hand media
1133 # playlists [2] should be returned as is since they contain just the media
1134 # without qualities renditions.
1135 # Fortunately, master playlist can be easily distinguished from media
1136 # playlist based on particular tags availability. As of [1, 2] master
1137 # playlist tags MUST NOT appear in a media playist and vice versa.
1138 # As of [3] #EXT-X-TARGETDURATION tag is REQUIRED for every media playlist
1139 # and MUST NOT appear in master playlist thus we can clearly detect media
1140 # playlist with this criterion.
1141 # 1. https://tools.ietf.org/html/draft-pantos-http-live-streaming-17#section-4.3.4
1142 # 2. https://tools.ietf.org/html/draft-pantos-http-live-streaming-17#section-4.3.3
1143 # 3. https://tools.ietf.org/html/draft-pantos-http-live-streaming-17#section-4.3.3.1
1144 if '#EXT-X-TARGETDURATION' in m3u8_doc: # media playlist, return as is
1145 return [{
1146 'url': m3u8_url,
1147 'format_id': m3u8_id,
1148 'ext': ext,
1149 'protocol': entry_protocol,
1150 'preference': preference,
1151 }]
1152 last_info = None
1153 last_media = None
1154 for line in m3u8_doc.splitlines():
1155 if line.startswith('#EXT-X-STREAM-INF:'):
1156 last_info = parse_m3u8_attributes(line)
1157 elif line.startswith('#EXT-X-MEDIA:'):
1158 last_media = parse_m3u8_attributes(line)
1159 elif line.startswith('#') or not line.strip():
1160 continue
1161 else:
1162 if last_info is None:
1163 formats.append({'url': format_url(line)})
1164 continue
1165 tbr = int_or_none(last_info.get('BANDWIDTH'), scale=1000)
1166 format_id = []
1167 if m3u8_id:
1168 format_id.append(m3u8_id)
1169 last_media_name = last_media.get('NAME') if last_media and last_media.get('TYPE') not in ('SUBTITLES', 'CLOSED-CAPTIONS') else None
1170 # Despite specification does not mention NAME attribute for
1171 # EXT-X-STREAM-INF it still sometimes may be present
1172 stream_name = last_info.get('NAME') or last_media_name
1173 # Bandwidth of live streams may differ over time thus making
1174 # format_id unpredictable. So it's better to keep provided
1175 # format_id intact.
1176 if not live:
1177 format_id.append(stream_name if stream_name else '%d' % (tbr if tbr else len(formats)))
1178 f = {
1179 'format_id': '-'.join(format_id),
1180 'url': format_url(line.strip()),
1181 'tbr': tbr,
1182 'ext': ext,
1183 'protocol': entry_protocol,
1184 'preference': preference,
1185 }
1186 resolution = last_info.get('RESOLUTION')
1187 if resolution:
1188 width_str, height_str = resolution.split('x')
1189 f['width'] = int(width_str)
1190 f['height'] = int(height_str)
1191 codecs = last_info.get('CODECS')
1192 if codecs:
1193 vcodec, acodec = [None] * 2
1194 va_codecs = codecs.split(',')
1195 if len(va_codecs) == 1:
1196 # Audio only entries usually come with single codec and
1197 # no resolution. For more robustness we also check it to
1198 # be mp4 audio.
1199 if not resolution and va_codecs[0].startswith('mp4a'):
1200 vcodec, acodec = 'none', va_codecs[0]
1201 else:
1202 vcodec = va_codecs[0]
1203 else:
1204 vcodec, acodec = va_codecs[:2]
1205 f.update({
1206 'acodec': acodec,
1207 'vcodec': vcodec,
1208 })
1209 if last_media is not None:
1210 f['m3u8_media'] = last_media
1211 last_media = None
1212 formats.append(f)
1213 last_info = {}
1214 return formats
1215
1216 @staticmethod
1217 def _xpath_ns(path, namespace=None):
1218 if not namespace:
1219 return path
1220 out = []
1221 for c in path.split('/'):
1222 if not c or c == '.':
1223 out.append(c)
1224 else:
1225 out.append('{%s}%s' % (namespace, c))
1226 return '/'.join(out)
1227
1228 def _extract_smil_formats(self, smil_url, video_id, fatal=True, f4m_params=None, transform_source=None):
1229 smil = self._download_smil(smil_url, video_id, fatal=fatal, transform_source=transform_source)
1230
1231 if smil is False:
1232 assert not fatal
1233 return []
1234
1235 namespace = self._parse_smil_namespace(smil)
1236
1237 return self._parse_smil_formats(
1238 smil, smil_url, video_id, namespace=namespace, f4m_params=f4m_params)
1239
1240 def _extract_smil_info(self, smil_url, video_id, fatal=True, f4m_params=None):
1241 smil = self._download_smil(smil_url, video_id, fatal=fatal)
1242 if smil is False:
1243 return {}
1244 return self._parse_smil(smil, smil_url, video_id, f4m_params=f4m_params)
1245
1246 def _download_smil(self, smil_url, video_id, fatal=True, transform_source=None):
1247 return self._download_xml(
1248 smil_url, video_id, 'Downloading SMIL file',
1249 'Unable to download SMIL file', fatal=fatal, transform_source=transform_source)
1250
1251 def _parse_smil(self, smil, smil_url, video_id, f4m_params=None):
1252 namespace = self._parse_smil_namespace(smil)
1253
1254 formats = self._parse_smil_formats(
1255 smil, smil_url, video_id, namespace=namespace, f4m_params=f4m_params)
1256 subtitles = self._parse_smil_subtitles(smil, namespace=namespace)
1257
1258 video_id = os.path.splitext(url_basename(smil_url))[0]
1259 title = None
1260 description = None
1261 upload_date = None
1262 for meta in smil.findall(self._xpath_ns('./head/meta', namespace)):
1263 name = meta.attrib.get('name')
1264 content = meta.attrib.get('content')
1265 if not name or not content:
1266 continue
1267 if not title and name == 'title':
1268 title = content
1269 elif not description and name in ('description', 'abstract'):
1270 description = content
1271 elif not upload_date and name == 'date':
1272 upload_date = unified_strdate(content)
1273
1274 thumbnails = [{
1275 'id': image.get('type'),
1276 'url': image.get('src'),
1277 'width': int_or_none(image.get('width')),
1278 'height': int_or_none(image.get('height')),
1279 } for image in smil.findall(self._xpath_ns('.//image', namespace)) if image.get('src')]
1280
1281 return {
1282 'id': video_id,
1283 'title': title or video_id,
1284 'description': description,
1285 'upload_date': upload_date,
1286 'thumbnails': thumbnails,
1287 'formats': formats,
1288 'subtitles': subtitles,
1289 }
1290
1291 def _parse_smil_namespace(self, smil):
1292 return self._search_regex(
1293 r'(?i)^{([^}]+)?}smil$', smil.tag, 'namespace', default=None)
1294
1295 def _parse_smil_formats(self, smil, smil_url, video_id, namespace=None, f4m_params=None, transform_rtmp_url=None):
1296 base = smil_url
1297 for meta in smil.findall(self._xpath_ns('./head/meta', namespace)):
1298 b = meta.get('base') or meta.get('httpBase')
1299 if b:
1300 base = b
1301 break
1302
1303 formats = []
1304 rtmp_count = 0
1305 http_count = 0
1306 m3u8_count = 0
1307
1308 srcs = []
1309 media = smil.findall(self._xpath_ns('.//video', namespace)) + smil.findall(self._xpath_ns('.//audio', namespace))
1310 for medium in media:
1311 src = medium.get('src')
1312 if not src or src in srcs:
1313 continue
1314 srcs.append(src)
1315
1316 bitrate = float_or_none(medium.get('system-bitrate') or medium.get('systemBitrate'), 1000)
1317 filesize = int_or_none(medium.get('size') or medium.get('fileSize'))
1318 width = int_or_none(medium.get('width'))
1319 height = int_or_none(medium.get('height'))
1320 proto = medium.get('proto')
1321 ext = medium.get('ext')
1322 src_ext = determine_ext(src)
1323 streamer = medium.get('streamer') or base
1324
1325 if proto == 'rtmp' or streamer.startswith('rtmp'):
1326 rtmp_count += 1
1327 formats.append({
1328 'url': streamer,
1329 'play_path': src,
1330 'ext': 'flv',
1331 'format_id': 'rtmp-%d' % (rtmp_count if bitrate is None else bitrate),
1332 'tbr': bitrate,
1333 'filesize': filesize,
1334 'width': width,
1335 'height': height,
1336 })
1337 if transform_rtmp_url:
1338 streamer, src = transform_rtmp_url(streamer, src)
1339 formats[-1].update({
1340 'url': streamer,
1341 'play_path': src,
1342 })
1343 continue
1344
1345 src_url = src if src.startswith('http') else compat_urlparse.urljoin(base, src)
1346 src_url = src_url.strip()
1347
1348 if proto == 'm3u8' or src_ext == 'm3u8':
1349 m3u8_formats = self._extract_m3u8_formats(
1350 src_url, video_id, ext or 'mp4', m3u8_id='hls', fatal=False)
1351 if len(m3u8_formats) == 1:
1352 m3u8_count += 1
1353 m3u8_formats[0].update({
1354 'format_id': 'hls-%d' % (m3u8_count if bitrate is None else bitrate),
1355 'tbr': bitrate,
1356 'width': width,
1357 'height': height,
1358 })
1359 formats.extend(m3u8_formats)
1360 continue
1361
1362 if src_ext == 'f4m':
1363 f4m_url = src_url
1364 if not f4m_params:
1365 f4m_params = {
1366 'hdcore': '3.2.0',
1367 'plugin': 'flowplayer-3.2.0.1',
1368 }
1369 f4m_url += '&' if '?' in f4m_url else '?'
1370 f4m_url += compat_urllib_parse_urlencode(f4m_params)
1371 formats.extend(self._extract_f4m_formats(f4m_url, video_id, f4m_id='hds', fatal=False))
1372 continue
1373
1374 if src_url.startswith('http') and self._is_valid_url(src, video_id):
1375 http_count += 1
1376 formats.append({
1377 'url': src_url,
1378 'ext': ext or src_ext or 'flv',
1379 'format_id': 'http-%d' % (bitrate or http_count),
1380 'tbr': bitrate,
1381 'filesize': filesize,
1382 'width': width,
1383 'height': height,
1384 })
1385 continue
1386
1387 return formats
1388
1389 def _parse_smil_subtitles(self, smil, namespace=None, subtitles_lang='en'):
1390 urls = []
1391 subtitles = {}
1392 for num, textstream in enumerate(smil.findall(self._xpath_ns('.//textstream', namespace))):
1393 src = textstream.get('src')
1394 if not src or src in urls:
1395 continue
1396 urls.append(src)
1397 ext = textstream.get('ext') or mimetype2ext(textstream.get('type')) or determine_ext(src)
1398 lang = textstream.get('systemLanguage') or textstream.get('systemLanguageName') or textstream.get('lang') or subtitles_lang
1399 subtitles.setdefault(lang, []).append({
1400 'url': src,
1401 'ext': ext,
1402 })
1403 return subtitles
1404
1405 def _extract_xspf_playlist(self, playlist_url, playlist_id, fatal=True):
1406 xspf = self._download_xml(
1407 playlist_url, playlist_id, 'Downloading xpsf playlist',
1408 'Unable to download xspf manifest', fatal=fatal)
1409 if xspf is False:
1410 return []
1411 return self._parse_xspf(xspf, playlist_id)
1412
1413 def _parse_xspf(self, playlist, playlist_id):
1414 NS_MAP = {
1415 'xspf': 'http://xspf.org/ns/0/',
1416 's1': 'http://static.streamone.nl/player/ns/0',
1417 }
1418
1419 entries = []
1420 for track in playlist.findall(xpath_with_ns('./xspf:trackList/xspf:track', NS_MAP)):
1421 title = xpath_text(
1422 track, xpath_with_ns('./xspf:title', NS_MAP), 'title', default=playlist_id)
1423 description = xpath_text(
1424 track, xpath_with_ns('./xspf:annotation', NS_MAP), 'description')
1425 thumbnail = xpath_text(
1426 track, xpath_with_ns('./xspf:image', NS_MAP), 'thumbnail')
1427 duration = float_or_none(
1428 xpath_text(track, xpath_with_ns('./xspf:duration', NS_MAP), 'duration'), 1000)
1429
1430 formats = [{
1431 'url': location.text,
1432 'format_id': location.get(xpath_with_ns('s1:label', NS_MAP)),
1433 'width': int_or_none(location.get(xpath_with_ns('s1:width', NS_MAP))),
1434 'height': int_or_none(location.get(xpath_with_ns('s1:height', NS_MAP))),
1435 } for location in track.findall(xpath_with_ns('./xspf:location', NS_MAP))]
1436 self._sort_formats(formats)
1437
1438 entries.append({
1439 'id': playlist_id,
1440 'title': title,
1441 'description': description,
1442 'thumbnail': thumbnail,
1443 'duration': duration,
1444 'formats': formats,
1445 })
1446 return entries
1447
1448 def _extract_mpd_formats(self, mpd_url, video_id, mpd_id=None, note=None, errnote=None, fatal=True, formats_dict={}):
1449 res = self._download_webpage_handle(
1450 mpd_url, video_id,
1451 note=note or 'Downloading MPD manifest',
1452 errnote=errnote or 'Failed to download MPD manifest',
1453 fatal=fatal)
1454 if res is False:
1455 return []
1456 mpd, urlh = res
1457 mpd_base_url = re.match(r'https?://.+/', urlh.geturl()).group()
1458
1459 return self._parse_mpd_formats(
1460 compat_etree_fromstring(mpd.encode('utf-8')), mpd_id, mpd_base_url, formats_dict=formats_dict)
1461
1462 def _parse_mpd_formats(self, mpd_doc, mpd_id=None, mpd_base_url='', formats_dict={}):
1463 if mpd_doc.get('type') == 'dynamic':
1464 return []
1465
1466 namespace = self._search_regex(r'(?i)^{([^}]+)?}MPD$', mpd_doc.tag, 'namespace', default=None)
1467
1468 def _add_ns(path):
1469 return self._xpath_ns(path, namespace)
1470
1471 def is_drm_protected(element):
1472 return element.find(_add_ns('ContentProtection')) is not None
1473
1474 def extract_multisegment_info(element, ms_parent_info):
1475 ms_info = ms_parent_info.copy()
1476 segment_list = element.find(_add_ns('SegmentList'))
1477 if segment_list is not None:
1478 segment_urls_e = segment_list.findall(_add_ns('SegmentURL'))
1479 if segment_urls_e:
1480 ms_info['segment_urls'] = [segment.attrib['media'] for segment in segment_urls_e]
1481 initialization = segment_list.find(_add_ns('Initialization'))
1482 if initialization is not None:
1483 ms_info['initialization_url'] = initialization.attrib['sourceURL']
1484 else:
1485 segment_template = element.find(_add_ns('SegmentTemplate'))
1486 if segment_template is not None:
1487 start_number = segment_template.get('startNumber')
1488 if start_number:
1489 ms_info['start_number'] = int(start_number)
1490 segment_timeline = segment_template.find(_add_ns('SegmentTimeline'))
1491 if segment_timeline is not None:
1492 s_e = segment_timeline.findall(_add_ns('S'))
1493 if s_e:
1494 ms_info['total_number'] = 0
1495 for s in s_e:
1496 ms_info['total_number'] += 1 + int(s.get('r', '0'))
1497 else:
1498 timescale = segment_template.get('timescale')
1499 if timescale:
1500 ms_info['timescale'] = int(timescale)
1501 segment_duration = segment_template.get('duration')
1502 if segment_duration:
1503 ms_info['segment_duration'] = int(segment_duration)
1504 media_template = segment_template.get('media')
1505 if media_template:
1506 ms_info['media_template'] = media_template
1507 initialization = segment_template.get('initialization')
1508 if initialization:
1509 ms_info['initialization_url'] = initialization
1510 else:
1511 initialization = segment_template.find(_add_ns('Initialization'))
1512 if initialization is not None:
1513 ms_info['initialization_url'] = initialization.attrib['sourceURL']
1514 return ms_info
1515
1516 mpd_duration = parse_duration(mpd_doc.get('mediaPresentationDuration'))
1517 formats = []
1518 for period in mpd_doc.findall(_add_ns('Period')):
1519 period_duration = parse_duration(period.get('duration')) or mpd_duration
1520 period_ms_info = extract_multisegment_info(period, {
1521 'start_number': 1,
1522 'timescale': 1,
1523 })
1524 for adaptation_set in period.findall(_add_ns('AdaptationSet')):
1525 if is_drm_protected(adaptation_set):
1526 continue
1527 adaption_set_ms_info = extract_multisegment_info(adaptation_set, period_ms_info)
1528 for representation in adaptation_set.findall(_add_ns('Representation')):
1529 if is_drm_protected(representation):
1530 continue
1531 representation_attrib = adaptation_set.attrib.copy()
1532 representation_attrib.update(representation.attrib)
1533 # According to page 41 of ISO/IEC 29001-1:2014, @mimeType is mandatory
1534 mime_type = representation_attrib['mimeType']
1535 content_type = mime_type.split('/')[0]
1536 if content_type == 'text':
1537 # TODO implement WebVTT downloading
1538 pass
1539 elif content_type == 'video' or content_type == 'audio':
1540 base_url = ''
1541 for element in (representation, adaptation_set, period, mpd_doc):
1542 base_url_e = element.find(_add_ns('BaseURL'))
1543 if base_url_e is not None:
1544 base_url = base_url_e.text + base_url
1545 if re.match(r'^https?://', base_url):
1546 break
1547 if mpd_base_url and not re.match(r'^https?://', base_url):
1548 if not mpd_base_url.endswith('/') and not base_url.startswith('/'):
1549 mpd_base_url += '/'
1550 base_url = mpd_base_url + base_url
1551 representation_id = representation_attrib.get('id')
1552 lang = representation_attrib.get('lang')
1553 url_el = representation.find(_add_ns('BaseURL'))
1554 filesize = int_or_none(url_el.attrib.get('{http://youtube.com/yt/2012/10/10}contentLength') if url_el is not None else None)
1555 f = {
1556 'format_id': '%s-%s' % (mpd_id, representation_id) if mpd_id else representation_id,
1557 'url': base_url,
1558 'ext': mimetype2ext(mime_type),
1559 'width': int_or_none(representation_attrib.get('width')),
1560 'height': int_or_none(representation_attrib.get('height')),
1561 'tbr': int_or_none(representation_attrib.get('bandwidth'), 1000),
1562 'asr': int_or_none(representation_attrib.get('audioSamplingRate')),
1563 'fps': int_or_none(representation_attrib.get('frameRate')),
1564 'vcodec': 'none' if content_type == 'audio' else representation_attrib.get('codecs'),
1565 'acodec': 'none' if content_type == 'video' else representation_attrib.get('codecs'),
1566 'language': lang if lang not in ('mul', 'und', 'zxx', 'mis') else None,
1567 'format_note': 'DASH %s' % content_type,
1568 'filesize': filesize,
1569 }
1570 representation_ms_info = extract_multisegment_info(representation, adaption_set_ms_info)
1571 if 'segment_urls' not in representation_ms_info and 'media_template' in representation_ms_info:
1572 if 'total_number' not in representation_ms_info and 'segment_duration':
1573 segment_duration = float(representation_ms_info['segment_duration']) / float(representation_ms_info['timescale'])
1574 representation_ms_info['total_number'] = int(math.ceil(float(period_duration) / segment_duration))
1575 media_template = representation_ms_info['media_template']
1576 media_template = media_template.replace('$RepresentationID$', representation_id)
1577 media_template = re.sub(r'\$(Number|Bandwidth)\$', r'%(\1)d', media_template)
1578 media_template = re.sub(r'\$(Number|Bandwidth)%([^$]+)\$', r'%(\1)\2', media_template)
1579 media_template.replace('$$', '$')
1580 representation_ms_info['segment_urls'] = [
1581 media_template % {
1582 'Number': segment_number,
1583 'Bandwidth': representation_attrib.get('bandwidth')}
1584 for segment_number in range(
1585 representation_ms_info['start_number'],
1586 representation_ms_info['total_number'] + representation_ms_info['start_number'])]
1587 if 'segment_urls' in representation_ms_info:
1588 f.update({
1589 'segment_urls': representation_ms_info['segment_urls'],
1590 'protocol': 'http_dash_segments',
1591 })
1592 if 'initialization_url' in representation_ms_info:
1593 initialization_url = representation_ms_info['initialization_url'].replace('$RepresentationID$', representation_id)
1594 f.update({
1595 'initialization_url': initialization_url,
1596 })
1597 if not f.get('url'):
1598 f['url'] = initialization_url
1599 try:
1600 existing_format = next(
1601 fo for fo in formats
1602 if fo['format_id'] == representation_id)
1603 except StopIteration:
1604 full_info = formats_dict.get(representation_id, {}).copy()
1605 full_info.update(f)
1606 formats.append(full_info)
1607 else:
1608 existing_format.update(f)
1609 else:
1610 self.report_warning('Unknown MIME type %s in DASH manifest' % mime_type)
1611 return formats
1612
1613 def _live_title(self, name):
1614 """ Generate the title for a live video """
1615 now = datetime.datetime.now()
1616 now_str = now.strftime('%Y-%m-%d %H:%M')
1617 return name + ' ' + now_str
1618
1619 def _int(self, v, name, fatal=False, **kwargs):
1620 res = int_or_none(v, **kwargs)
1621 if 'get_attr' in kwargs:
1622 print(getattr(v, kwargs['get_attr']))
1623 if res is None:
1624 msg = 'Failed to extract %s: Could not parse value %r' % (name, v)
1625 if fatal:
1626 raise ExtractorError(msg)
1627 else:
1628 self._downloader.report_warning(msg)
1629 return res
1630
1631 def _float(self, v, name, fatal=False, **kwargs):
1632 res = float_or_none(v, **kwargs)
1633 if res is None:
1634 msg = 'Failed to extract %s: Could not parse value %r' % (name, v)
1635 if fatal:
1636 raise ExtractorError(msg)
1637 else:
1638 self._downloader.report_warning(msg)
1639 return res
1640
1641 def _set_cookie(self, domain, name, value, expire_time=None):
1642 cookie = compat_cookiejar.Cookie(
1643 0, name, value, None, None, domain, None,
1644 None, '/', True, False, expire_time, '', None, None, None)
1645 self._downloader.cookiejar.set_cookie(cookie)
1646
1647 def _get_cookies(self, url):
1648 """ Return a compat_cookies.SimpleCookie with the cookies for the url """
1649 req = sanitized_Request(url)
1650 self._downloader.cookiejar.add_cookie_header(req)
1651 return compat_cookies.SimpleCookie(req.get_header('Cookie'))
1652
1653 def get_testcases(self, include_onlymatching=False):
1654 t = getattr(self, '_TEST', None)
1655 if t:
1656 assert not hasattr(self, '_TESTS'), \
1657 '%s has _TEST and _TESTS' % type(self).__name__
1658 tests = [t]
1659 else:
1660 tests = getattr(self, '_TESTS', [])
1661 for t in tests:
1662 if not include_onlymatching and t.get('only_matching', False):
1663 continue
1664 t['name'] = type(self).__name__[:-len('IE')]
1665 yield t
1666
1667 def is_suitable(self, age_limit):
1668 """ Test whether the extractor is generally suitable for the given
1669 age limit (i.e. pornographic sites are not, all others usually are) """
1670
1671 any_restricted = False
1672 for tc in self.get_testcases(include_onlymatching=False):
1673 if 'playlist' in tc:
1674 tc = tc['playlist'][0]
1675 is_restricted = age_restricted(
1676 tc.get('info_dict', {}).get('age_limit'), age_limit)
1677 if not is_restricted:
1678 return True
1679 any_restricted = any_restricted or is_restricted
1680 return not any_restricted
1681
1682 def extract_subtitles(self, *args, **kwargs):
1683 if (self._downloader.params.get('writesubtitles', False) or
1684 self._downloader.params.get('listsubtitles')):
1685 return self._get_subtitles(*args, **kwargs)
1686 return {}
1687
1688 def _get_subtitles(self, *args, **kwargs):
1689 raise NotImplementedError('This method must be implemented by subclasses')
1690
1691 @staticmethod
1692 def _merge_subtitle_items(subtitle_list1, subtitle_list2):
1693 """ Merge subtitle items for one language. Items with duplicated URLs
1694 will be dropped. """
1695 list1_urls = set([item['url'] for item in subtitle_list1])
1696 ret = list(subtitle_list1)
1697 ret.extend([item for item in subtitle_list2 if item['url'] not in list1_urls])
1698 return ret
1699
1700 @classmethod
1701 def _merge_subtitles(cls, subtitle_dict1, subtitle_dict2):
1702 """ Merge two subtitle dictionaries, language by language. """
1703 ret = dict(subtitle_dict1)
1704 for lang in subtitle_dict2:
1705 ret[lang] = cls._merge_subtitle_items(subtitle_dict1.get(lang, []), subtitle_dict2[lang])
1706 return ret
1707
1708 def extract_automatic_captions(self, *args, **kwargs):
1709 if (self._downloader.params.get('writeautomaticsub', False) or
1710 self._downloader.params.get('listsubtitles')):
1711 return self._get_automatic_captions(*args, **kwargs)
1712 return {}
1713
1714 def _get_automatic_captions(self, *args, **kwargs):
1715 raise NotImplementedError('This method must be implemented by subclasses')
1716
1717 def mark_watched(self, *args, **kwargs):
1718 if (self._downloader.params.get('mark_watched', False) and
1719 (self._get_login_info()[0] is not None or
1720 self._downloader.params.get('cookiefile') is not None)):
1721 self._mark_watched(*args, **kwargs)
1722
1723 def _mark_watched(self, *args, **kwargs):
1724 raise NotImplementedError('This method must be implemented by subclasses')
1725
1726
1727 class SearchInfoExtractor(InfoExtractor):
1728 """
1729 Base class for paged search queries extractors.
1730 They accept URLs in the format _SEARCH_KEY(|all|[0-9]):{query}
1731 Instances should define _SEARCH_KEY and _MAX_RESULTS.
1732 """
1733
1734 @classmethod
1735 def _make_valid_url(cls):
1736 return r'%s(?P<prefix>|[1-9][0-9]*|all):(?P<query>[\s\S]+)' % cls._SEARCH_KEY
1737
1738 @classmethod
1739 def suitable(cls, url):
1740 return re.match(cls._make_valid_url(), url) is not None
1741
1742 def _real_extract(self, query):
1743 mobj = re.match(self._make_valid_url(), query)
1744 if mobj is None:
1745 raise ExtractorError('Invalid search query "%s"' % query)
1746
1747 prefix = mobj.group('prefix')
1748 query = mobj.group('query')
1749 if prefix == '':
1750 return self._get_n_results(query, 1)
1751 elif prefix == 'all':
1752 return self._get_n_results(query, self._MAX_RESULTS)
1753 else:
1754 n = int(prefix)
1755 if n <= 0:
1756 raise ExtractorError('invalid download number %s for query "%s"' % (n, query))
1757 elif n > self._MAX_RESULTS:
1758 self._downloader.report_warning('%s returns max %i results (you requested %i)' % (self._SEARCH_KEY, self._MAX_RESULTS, n))
1759 n = self._MAX_RESULTS
1760 return self._get_n_results(query, n)
1761
1762 def _get_n_results(self, query, n):
1763 """Get a specified number of results for a query"""
1764 raise NotImplementedError('This method must be implemented by subclasses')
1765
1766 @property
1767 def SEARCH_KEY(self):
1768 return self._SEARCH_KEY