]> jfr.im git - yt-dlp.git/blob - youtube_dl/extractor/youtube.py
projects / yt-dlp.git / blob
? search:
summary | shortlog | log | commit | commitdiff | tree
blame | history | raw | HEAD
[youtube] Fix missing format details for 60fps DASH formats
[yt-dlp.git] / youtube_dl / extractor / youtube.py
1 # coding: utf-8
2
3 from __future__ import unicode_literals
4
5
6 import itertools
7 import json
8 import os.path
9 import re
10 import time
11 import traceback
12
13 from .common import InfoExtractor, SearchInfoExtractor
14 from ..jsinterp import JSInterpreter
15 from ..swfinterp import SWFInterpreter
16 from ..compat import (
17 compat_chr,
18 compat_parse_qs,
19 compat_urllib_parse,
20 compat_urllib_parse_unquote,
21 compat_urllib_parse_unquote_plus,
22 compat_urllib_parse_urlparse,
23 compat_urllib_request,
24 compat_urlparse,
25 compat_str,
26 )
27 from ..utils import (
28 clean_html,
29 ExtractorError,
30 float_or_none,
31 get_element_by_attribute,
32 get_element_by_id,
33 int_or_none,
34 orderedSet,
35 parse_duration,
36 remove_start,
37 smuggle_url,
38 str_to_int,
39 unescapeHTML,
40 unified_strdate,
41 unsmuggle_url,
42 uppercase_escape,
43 ISO3166Utils,
44 )
45
46
47 class YoutubeBaseInfoExtractor(InfoExtractor):
48 """Provide base functions for Youtube extractors"""
49 _LOGIN_URL = 'https://accounts.google.com/ServiceLogin'
50 _TWOFACTOR_URL = 'https://accounts.google.com/signin/challenge'
51 _NETRC_MACHINE = 'youtube'
52 # If True it will raise an error if no login info is provided
53 _LOGIN_REQUIRED = False
54
55 def _set_language(self):
56 self._set_cookie(
57 '.youtube.com', 'PREF', 'f1=50000000&hl=en',
58 # YouTube sets the expire time to about two months
59 expire_time=time.time() + 2 * 30 * 24 * 3600)
60
61 def _ids_to_results(self, ids):
62 return [
63 self.url_result(vid_id, 'Youtube', video_id=vid_id)
64 for vid_id in ids]
65
66 def _login(self):
67 """
68 Attempt to log in to YouTube.
69 True is returned if successful or skipped.
70 False is returned if login failed.
71
72 If _LOGIN_REQUIRED is set and no authentication was provided, an error is raised.
73 """
74 (username, password) = self._get_login_info()
75 # No authentication to be performed
76 if username is None:
77 if self._LOGIN_REQUIRED:
78 raise ExtractorError('No login info available, needed for using %s.' % self.IE_NAME, expected=True)
79 return True
80
81 login_page = self._download_webpage(
82 self._LOGIN_URL, None,
83 note='Downloading login page',
84 errnote='unable to fetch login page', fatal=False)
85 if login_page is False:
86 return
87
88 galx = self._search_regex(r'(?s)<input.+?name="GALX".+?value="(.+?)"',
89 login_page, 'Login GALX parameter')
90
91 # Log in
92 login_form_strs = {
93 'continue': 'https://www.youtube.com/signin?action_handle_signin=true&feature=sign_in_button&hl=en_US&nomobiletemp=1',
94 'Email': username,
95 'GALX': galx,
96 'Passwd': password,
97
98 'PersistentCookie': 'yes',
99 '_utf8': '霱',
100 'bgresponse': 'js_disabled',
101 'checkConnection': '',
102 'checkedDomains': 'youtube',
103 'dnConn': '',
104 'pstMsg': '0',
105 'rmShown': '1',
106 'secTok': '',
107 'signIn': 'Sign in',
108 'timeStmp': '',
109 'service': 'youtube',
110 'uilel': '3',
111 'hl': 'en_US',
112 }
113
114 # Convert to UTF-8 *before* urlencode because Python 2.x's urlencode
115 # chokes on unicode
116 login_form = dict((k.encode('utf-8'), v.encode('utf-8')) for k, v in login_form_strs.items())
117 login_data = compat_urllib_parse.urlencode(login_form).encode('ascii')
118
119 req = compat_urllib_request.Request(self._LOGIN_URL, login_data)
120 login_results = self._download_webpage(
121 req, None,
122 note='Logging in', errnote='unable to log in', fatal=False)
123 if login_results is False:
124 return False
125
126 if re.search(r'id="errormsg_0_Passwd"', login_results) is not None:
127 raise ExtractorError('Please use your account password and a two-factor code instead of an application-specific password.', expected=True)
128
129 # Two-Factor
130 # TODO add SMS and phone call support - these require making a request and then prompting the user
131
132 if re.search(r'(?i)<form[^>]* id="challenge"', login_results) is not None:
133 tfa_code = self._get_tfa_info('2-step verification code')
134
135 if not tfa_code:
136 self._downloader.report_warning(
137 'Two-factor authentication required. Provide it either interactively or with --twofactor <code>'
138 '(Note that only TOTP (Google Authenticator App) codes work at this time.)')
139 return False
140
141 tfa_code = remove_start(tfa_code, 'G-')
142
143 tfa_form_strs = self._form_hidden_inputs('challenge', login_results)
144
145 tfa_form_strs.update({
146 'Pin': tfa_code,
147 'TrustDevice': 'on',
148 })
149
150 tfa_form = dict((k.encode('utf-8'), v.encode('utf-8')) for k, v in tfa_form_strs.items())
151 tfa_data = compat_urllib_parse.urlencode(tfa_form).encode('ascii')
152
153 tfa_req = compat_urllib_request.Request(self._TWOFACTOR_URL, tfa_data)
154 tfa_results = self._download_webpage(
155 tfa_req, None,
156 note='Submitting TFA code', errnote='unable to submit tfa', fatal=False)
157
158 if tfa_results is False:
159 return False
160
161 if re.search(r'(?i)<form[^>]* id="challenge"', tfa_results) is not None:
162 self._downloader.report_warning('Two-factor code expired or invalid. Please try again, or use a one-use backup code instead.')
163 return False
164 if re.search(r'(?i)<form[^>]* id="gaia_loginform"', tfa_results) is not None:
165 self._downloader.report_warning('unable to log in - did the page structure change?')
166 return False
167 if re.search(r'smsauth-interstitial-reviewsettings', tfa_results) is not None:
168 self._downloader.report_warning('Your Google account has a security notice. Please log in on your web browser, resolve the notice, and try again.')
169 return False
170
171 if re.search(r'(?i)<form[^>]* id="gaia_loginform"', login_results) is not None:
172 self._downloader.report_warning('unable to log in: bad username or password')
173 return False
174 return True
175
176 def _real_initialize(self):
177 if self._downloader is None:
178 return
179 self._set_language()
180 if not self._login():
181 return
182
183
184 class YoutubeIE(YoutubeBaseInfoExtractor):
185 IE_DESC = 'YouTube.com'
186 _VALID_URL = r"""(?x)^
187 (
188 (?:https?://|//) # http(s):// or protocol-independent URL
189 (?:(?:(?:(?:\w+\.)?[yY][oO][uU][tT][uU][bB][eE](?:-nocookie)?\.com/|
190 (?:www\.)?deturl\.com/www\.youtube\.com/|
191 (?:www\.)?pwnyoutube\.com/|
192 (?:www\.)?yourepeat\.com/|
193 tube\.majestyc\.net/|
194 youtube\.googleapis\.com/) # the various hostnames, with wildcard subdomains
195 (?:.*?\#/)? # handle anchor (#/) redirect urls
196 (?: # the various things that can precede the ID:
197 (?:(?:v|embed|e)/(?!videoseries)) # v/ or embed/ or e/
198 |(?: # or the v= param in all its forms
199 (?:(?:watch|movie)(?:_popup)?(?:\.php)?/?)? # preceding watch(_popup|.php) or nothing (like /?v=xxxx)
200 (?:\?|\#!?) # the params delimiter ? or # or #!
201 (?:.*?&)?? # any other preceding param (like /?s=tuff&v=xxxx)
202 v=
203 )
204 ))
205 |(?:
206 youtu\.be| # just youtu.be/xxxx
207 vid\.plus # or vid.plus/xxxx
208 )/
209 |(?:www\.)?cleanvideosearch\.com/media/action/yt/watch\?videoId=
210 )
211 )? # all until now is optional -> you can pass the naked ID
212 ([0-9A-Za-z_-]{11}) # here is it! the YouTube video ID
213 (?!.*?&list=) # combined list/video URLs are handled by the playlist IE
214 (?(1).+)? # if we found the ID, everything can follow
215 $"""
216 _NEXT_URL_RE = r'[\?&]next_url=([^&]+)'
217 _formats = {
218 '5': {'ext': 'flv', 'width': 400, 'height': 240},
219 '6': {'ext': 'flv', 'width': 450, 'height': 270},
220 '13': {'ext': '3gp'},
221 '17': {'ext': '3gp', 'width': 176, 'height': 144},
222 '18': {'ext': 'mp4', 'width': 640, 'height': 360},
223 '22': {'ext': 'mp4', 'width': 1280, 'height': 720},
224 '34': {'ext': 'flv', 'width': 640, 'height': 360},
225 '35': {'ext': 'flv', 'width': 854, 'height': 480},
226 '36': {'ext': '3gp', 'width': 320, 'height': 240},
227 '37': {'ext': 'mp4', 'width': 1920, 'height': 1080},
228 '38': {'ext': 'mp4', 'width': 4096, 'height': 3072},
229 '43': {'ext': 'webm', 'width': 640, 'height': 360},
230 '44': {'ext': 'webm', 'width': 854, 'height': 480},
231 '45': {'ext': 'webm', 'width': 1280, 'height': 720},
232 '46': {'ext': 'webm', 'width': 1920, 'height': 1080},
233 '59': {'ext': 'mp4', 'width': 854, 'height': 480},
234 '78': {'ext': 'mp4', 'width': 854, 'height': 480},
235
236
237 # 3d videos
238 '82': {'ext': 'mp4', 'height': 360, 'format_note': '3D', 'preference': -20},
239 '83': {'ext': 'mp4', 'height': 480, 'format_note': '3D', 'preference': -20},
240 '84': {'ext': 'mp4', 'height': 720, 'format_note': '3D', 'preference': -20},
241 '85': {'ext': 'mp4', 'height': 1080, 'format_note': '3D', 'preference': -20},
242 '100': {'ext': 'webm', 'height': 360, 'format_note': '3D', 'preference': -20},
243 '101': {'ext': 'webm', 'height': 480, 'format_note': '3D', 'preference': -20},
244 '102': {'ext': 'webm', 'height': 720, 'format_note': '3D', 'preference': -20},
245
246 # Apple HTTP Live Streaming
247 '92': {'ext': 'mp4', 'height': 240, 'format_note': 'HLS', 'preference': -10},
248 '93': {'ext': 'mp4', 'height': 360, 'format_note': 'HLS', 'preference': -10},
249 '94': {'ext': 'mp4', 'height': 480, 'format_note': 'HLS', 'preference': -10},
250 '95': {'ext': 'mp4', 'height': 720, 'format_note': 'HLS', 'preference': -10},
251 '96': {'ext': 'mp4', 'height': 1080, 'format_note': 'HLS', 'preference': -10},
252 '132': {'ext': 'mp4', 'height': 240, 'format_note': 'HLS', 'preference': -10},
253 '151': {'ext': 'mp4', 'height': 72, 'format_note': 'HLS', 'preference': -10},
254
255 # DASH mp4 video
256 '133': {'ext': 'mp4', 'height': 240, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
257 '134': {'ext': 'mp4', 'height': 360, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
258 '135': {'ext': 'mp4', 'height': 480, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
259 '136': {'ext': 'mp4', 'height': 720, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
260 '137': {'ext': 'mp4', 'height': 1080, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
261 '138': {'ext': 'mp4', 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40}, # Height can vary (https://github.com/rg3/youtube-dl/issues/4559)
262 '160': {'ext': 'mp4', 'height': 144, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
263 '264': {'ext': 'mp4', 'height': 1440, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
264 '298': {'ext': 'mp4', 'height': 720, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40, 'fps': 60, 'vcodec': 'h264'},
265 '299': {'ext': 'mp4', 'height': 1080, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40, 'fps': 60, 'vcodec': 'h264'},
266 '266': {'ext': 'mp4', 'height': 2160, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40, 'vcodec': 'h264'},
267
268 # Dash mp4 audio
269 '139': {'ext': 'm4a', 'format_note': 'DASH audio', 'acodec': 'aac', 'vcodec': 'none', 'abr': 48, 'preference': -50, 'container': 'm4a_dash'},
270 '140': {'ext': 'm4a', 'format_note': 'DASH audio', 'acodec': 'aac', 'vcodec': 'none', 'abr': 128, 'preference': -50, 'container': 'm4a_dash'},
271 '141': {'ext': 'm4a', 'format_note': 'DASH audio', 'acodec': 'aac', 'vcodec': 'none', 'abr': 256, 'preference': -50, 'container': 'm4a_dash'},
272
273 # Dash webm
274 '167': {'ext': 'webm', 'height': 360, 'width': 640, 'format_note': 'DASH video', 'acodec': 'none', 'container': 'webm', 'vcodec': 'vp8', 'preference': -40},
275 '168': {'ext': 'webm', 'height': 480, 'width': 854, 'format_note': 'DASH video', 'acodec': 'none', 'container': 'webm', 'vcodec': 'vp8', 'preference': -40},
276 '169': {'ext': 'webm', 'height': 720, 'width': 1280, 'format_note': 'DASH video', 'acodec': 'none', 'container': 'webm', 'vcodec': 'vp8', 'preference': -40},
277 '170': {'ext': 'webm', 'height': 1080, 'width': 1920, 'format_note': 'DASH video', 'acodec': 'none', 'container': 'webm', 'vcodec': 'vp8', 'preference': -40},
278 '218': {'ext': 'webm', 'height': 480, 'width': 854, 'format_note': 'DASH video', 'acodec': 'none', 'container': 'webm', 'vcodec': 'vp8', 'preference': -40},
279 '219': {'ext': 'webm', 'height': 480, 'width': 854, 'format_note': 'DASH video', 'acodec': 'none', 'container': 'webm', 'vcodec': 'vp8', 'preference': -40},
280 '278': {'ext': 'webm', 'height': 144, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40, 'container': 'webm', 'vcodec': 'vp9'},
281 '242': {'ext': 'webm', 'height': 240, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
282 '243': {'ext': 'webm', 'height': 360, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
283 '244': {'ext': 'webm', 'height': 480, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
284 '245': {'ext': 'webm', 'height': 480, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
285 '246': {'ext': 'webm', 'height': 480, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
286 '247': {'ext': 'webm', 'height': 720, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
287 '248': {'ext': 'webm', 'height': 1080, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
288 '271': {'ext': 'webm', 'height': 1440, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
289 '272': {'ext': 'webm', 'height': 2160, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
290 '302': {'ext': 'webm', 'height': 720, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40, 'fps': 60, 'vcodec': 'vp9'},
291 '303': {'ext': 'webm', 'height': 1080, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40, 'fps': 60, 'vcodec': 'vp9'},
292 '308': {'ext': 'webm', 'height': 1440, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40, 'fps': 60, 'vcodec': 'vp9'},
293 '313': {'ext': 'webm', 'height': 2160, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40, 'vcodec': 'vp9'},
294 '315': {'ext': 'webm', 'height': 2160, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40, 'fps': 60, 'vcodec': 'vp9'},
295
296 # Dash webm audio
297 '171': {'ext': 'webm', 'vcodec': 'none', 'format_note': 'DASH audio', 'abr': 128, 'preference': -50},
298 '172': {'ext': 'webm', 'vcodec': 'none', 'format_note': 'DASH audio', 'abr': 256, 'preference': -50},
299
300 # Dash webm audio with opus inside
301 '249': {'ext': 'webm', 'vcodec': 'none', 'format_note': 'DASH audio', 'acodec': 'opus', 'abr': 50, 'preference': -50},
302 '250': {'ext': 'webm', 'vcodec': 'none', 'format_note': 'DASH audio', 'acodec': 'opus', 'abr': 70, 'preference': -50},
303 '251': {'ext': 'webm', 'vcodec': 'none', 'format_note': 'DASH audio', 'acodec': 'opus', 'abr': 160, 'preference': -50},
304
305 # RTMP (unnamed)
306 '_rtmp': {'protocol': 'rtmp'},
307 }
308
309 IE_NAME = 'youtube'
310 _TESTS = [
311 {
312 'url': 'http://www.youtube.com/watch?v=BaW_jenozKcj&t=1s&end=9',
313 'info_dict': {
314 'id': 'BaW_jenozKc',
315 'ext': 'mp4',
316 'title': 'youtube-dl test video "\'/\\ä↭𝕐',
317 'uploader': 'Philipp Hagemeister',
318 'uploader_id': 'phihag',
319 'upload_date': '20121002',
320 'description': 'test chars: "\'/\\ä↭𝕐\ntest URL: https://github.com/rg3/youtube-dl/issues/1892\n\nThis is a test video for youtube-dl.\n\nFor more information, contact phihag@phihag.de .',
321 'categories': ['Science & Technology'],
322 'tags': ['youtube-dl'],
323 'like_count': int,
324 'dislike_count': int,
325 'start_time': 1,
326 'end_time': 9,
327 }
328 },
329 {
330 'url': 'http://www.youtube.com/watch?v=UxxajLWwzqY',
331 'note': 'Test generic use_cipher_signature video (#897)',
332 'info_dict': {
333 'id': 'UxxajLWwzqY',
334 'ext': 'mp4',
335 'upload_date': '20120506',
336 'title': 'Icona Pop - I Love It (feat. Charli XCX) [OFFICIAL VIDEO]',
337 'description': 'md5:782e8651347686cba06e58f71ab51773',
338 'tags': ['Icona Pop i love it', 'sweden', 'pop music', 'big beat records', 'big beat', 'charli',
339 'xcx', 'charli xcx', 'girls', 'hbo', 'i love it', "i don't care", 'icona', 'pop',
340 'iconic ep', 'iconic', 'love', 'it'],
341 'uploader': 'Icona Pop',
342 'uploader_id': 'IconaPop',
343 }
344 },
345 {
346 'url': 'https://www.youtube.com/watch?v=07FYdnEawAQ',
347 'note': 'Test VEVO video with age protection (#956)',
348 'info_dict': {
349 'id': '07FYdnEawAQ',
350 'ext': 'mp4',
351 'upload_date': '20130703',
352 'title': 'Justin Timberlake - Tunnel Vision (Explicit)',
353 'description': 'md5:64249768eec3bc4276236606ea996373',
354 'uploader': 'justintimberlakeVEVO',
355 'uploader_id': 'justintimberlakeVEVO',
356 'age_limit': 18,
357 }
358 },
359 {
360 'url': '//www.YouTube.com/watch?v=yZIXLfi8CZQ',
361 'note': 'Embed-only video (#1746)',
362 'info_dict': {
363 'id': 'yZIXLfi8CZQ',
364 'ext': 'mp4',
365 'upload_date': '20120608',
366 'title': 'Principal Sexually Assaults A Teacher - Episode 117 - 8th June 2012',
367 'description': 'md5:09b78bd971f1e3e289601dfba15ca4f7',
368 'uploader': 'SET India',
369 'uploader_id': 'setindia'
370 }
371 },
372 {
373 'url': 'http://www.youtube.com/watch?v=BaW_jenozKcj&v=UxxajLWwzqY',
374 'note': 'Use the first video ID in the URL',
375 'info_dict': {
376 'id': 'BaW_jenozKc',
377 'ext': 'mp4',
378 'title': 'youtube-dl test video "\'/\\ä↭𝕐',
379 'uploader': 'Philipp Hagemeister',
380 'uploader_id': 'phihag',
381 'upload_date': '20121002',
382 'description': 'test chars: "\'/\\ä↭𝕐\ntest URL: https://github.com/rg3/youtube-dl/issues/1892\n\nThis is a test video for youtube-dl.\n\nFor more information, contact phihag@phihag.de .',
383 'categories': ['Science & Technology'],
384 'tags': ['youtube-dl'],
385 'like_count': int,
386 'dislike_count': int,
387 },
388 'params': {
389 'skip_download': True,
390 },
391 },
392 {
393 'url': 'http://www.youtube.com/watch?v=a9LDPn-MO4I',
394 'note': '256k DASH audio (format 141) via DASH manifest',
395 'info_dict': {
396 'id': 'a9LDPn-MO4I',
397 'ext': 'm4a',
398 'upload_date': '20121002',
399 'uploader_id': '8KVIDEO',
400 'description': '',
401 'uploader': '8KVIDEO',
402 'title': 'UHDTV TEST 8K VIDEO.mp4'
403 },
404 'params': {
405 'youtube_include_dash_manifest': True,
406 'format': '141',
407 },
408 },
409 # DASH manifest with encrypted signature
410 {
411 'url': 'https://www.youtube.com/watch?v=IB3lcPjvWLA',
412 'info_dict': {
413 'id': 'IB3lcPjvWLA',
414 'ext': 'm4a',
415 'title': 'Afrojack, Spree Wilson - The Spark ft. Spree Wilson',
416 'description': 'md5:12e7067fa6735a77bdcbb58cb1187d2d',
417 'uploader': 'AfrojackVEVO',
418 'uploader_id': 'AfrojackVEVO',
419 'upload_date': '20131011',
420 },
421 'params': {
422 'youtube_include_dash_manifest': True,
423 'format': '141',
424 },
425 },
426 # JS player signature function name containing $
427 {
428 'url': 'https://www.youtube.com/watch?v=nfWlot6h_JM',
429 'info_dict': {
430 'id': 'nfWlot6h_JM',
431 'ext': 'm4a',
432 'title': 'Taylor Swift - Shake It Off',
433 'description': 'md5:95f66187cd7c8b2c13eb78e1223b63c3',
434 'uploader': 'TaylorSwiftVEVO',
435 'uploader_id': 'TaylorSwiftVEVO',
436 'upload_date': '20140818',
437 },
438 'params': {
439 'youtube_include_dash_manifest': True,
440 'format': '141',
441 },
442 },
443 # Controversy video
444 {
445 'url': 'https://www.youtube.com/watch?v=T4XJQO3qol8',
446 'info_dict': {
447 'id': 'T4XJQO3qol8',
448 'ext': 'mp4',
449 'upload_date': '20100909',
450 'uploader': 'The Amazing Atheist',
451 'uploader_id': 'TheAmazingAtheist',
452 'title': 'Burning Everyone\'s Koran',
453 'description': 'SUBSCRIBE: http://www.youtube.com/saturninefilms\n\nEven Obama has taken a stand against freedom on this issue: http://www.huffingtonpost.com/2010/09/09/obama-gma-interview-quran_n_710282.html',
454 }
455 },
456 # Normal age-gate video (No vevo, embed allowed)
457 {
458 'url': 'http://youtube.com/watch?v=HtVdAasjOgU',
459 'info_dict': {
460 'id': 'HtVdAasjOgU',
461 'ext': 'mp4',
462 'title': 'The Witcher 3: Wild Hunt - The Sword Of Destiny Trailer',
463 'description': 're:(?s).{100,}About the Game\n.*?The Witcher 3: Wild Hunt.{100,}',
464 'uploader': 'The Witcher',
465 'uploader_id': 'WitcherGame',
466 'upload_date': '20140605',
467 'age_limit': 18,
468 },
469 },
470 # Age-gate video with encrypted signature
471 {
472 'url': 'http://www.youtube.com/watch?v=6kLq3WMV1nU',
473 'info_dict': {
474 'id': '6kLq3WMV1nU',
475 'ext': 'mp4',
476 'title': 'Dedication To My Ex (Miss That) (Lyric Video)',
477 'description': 'md5:33765bb339e1b47e7e72b5490139bb41',
478 'uploader': 'LloydVEVO',
479 'uploader_id': 'LloydVEVO',
480 'upload_date': '20110629',
481 'age_limit': 18,
482 },
483 },
484 # video_info is None (https://github.com/rg3/youtube-dl/issues/4421)
485 {
486 'url': '__2ABJjxzNo',
487 'info_dict': {
488 'id': '__2ABJjxzNo',
489 'ext': 'mp4',
490 'upload_date': '20100430',
491 'uploader_id': 'deadmau5',
492 'description': 'md5:12c56784b8032162bb936a5f76d55360',
493 'uploader': 'deadmau5',
494 'title': 'Deadmau5 - Some Chords (HD)',
495 },
496 'expected_warnings': [
497 'DASH manifest missing',
498 ]
499 },
500 # Olympics (https://github.com/rg3/youtube-dl/issues/4431)
501 {
502 'url': 'lqQg6PlCWgI',
503 'info_dict': {
504 'id': 'lqQg6PlCWgI',
505 'ext': 'mp4',
506 'upload_date': '20120724',
507 'uploader_id': 'olympic',
508 'description': 'HO09 - Women - GER-AUS - Hockey - 31 July 2012 - London 2012 Olympic Games',
509 'uploader': 'Olympics',
510 'title': 'Hockey - Women - GER-AUS - London 2012 Olympic Games',
511 },
512 'params': {
513 'skip_download': 'requires avconv',
514 }
515 },
516 # Non-square pixels
517 {
518 'url': 'https://www.youtube.com/watch?v=_b-2C3KPAM0',
519 'info_dict': {
520 'id': '_b-2C3KPAM0',
521 'ext': 'mp4',
522 'stretched_ratio': 16 / 9.,
523 'upload_date': '20110310',
524 'uploader_id': 'AllenMeow',
525 'description': 'made by Wacom from Korea | 字幕&加油添醋 by TY\'s Allen | 感謝heylisa00cavey1001同學熱情提供梗及翻譯',
526 'uploader': '孫艾倫',
527 'title': '[A-made] 變態妍字幕版 太妍 我就是這樣的人',
528 },
529 },
530 # url_encoded_fmt_stream_map is empty string
531 {
532 'url': 'qEJwOuvDf7I',
533 'info_dict': {
534 'id': 'qEJwOuvDf7I',
535 'ext': 'webm',
536 'title': 'Обсуждение судебной практики по выборам 14 сентября 2014 года в Санкт-Петербурге',
537 'description': '',
538 'upload_date': '20150404',
539 'uploader_id': 'spbelect',
540 'uploader': 'Наблюдатели Петербурга',
541 },
542 'params': {
543 'skip_download': 'requires avconv',
544 }
545 },
546 # Extraction from multiple DASH manifests (https://github.com/rg3/youtube-dl/pull/6097)
547 {
548 'url': 'https://www.youtube.com/watch?v=FIl7x6_3R5Y',
549 'info_dict': {
550 'id': 'FIl7x6_3R5Y',
551 'ext': 'mp4',
552 'title': 'md5:7b81415841e02ecd4313668cde88737a',
553 'description': 'md5:116377fd2963b81ec4ce64b542173306',
554 'upload_date': '20150625',
555 'uploader_id': 'dorappi2000',
556 'uploader': 'dorappi2000',
557 'formats': 'mincount:33',
558 },
559 },
560 # DASH manifest with segment_list
561 {
562 'url': 'https://www.youtube.com/embed/CsmdDsKjzN8',
563 'md5': '8ce563a1d667b599d21064e982ab9e31',
564 'info_dict': {
565 'id': 'CsmdDsKjzN8',
566 'ext': 'mp4',
567 'upload_date': '20150501', # According to '<meta itemprop="datePublished"', but in other places it's 20150510
568 'uploader': 'Airtek',
569 'description': 'Retransmisión en directo de la XVIII media maratón de Zaragoza.',
570 'uploader_id': 'UCzTzUmjXxxacNnL8I3m4LnQ',
571 'title': 'Retransmisión XVIII Media maratón Zaragoza 2015',
572 },
573 'params': {
574 'youtube_include_dash_manifest': True,
575 'format': '135', # bestvideo
576 }
577 },
578 {
579 # Multifeed videos (multiple cameras), URL is for Main Camera
580 'url': 'https://www.youtube.com/watch?v=jqWvoWXjCVs',
581 'info_dict': {
582 'id': 'jqWvoWXjCVs',
583 'title': 'teamPGP: Rocket League Noob Stream',
584 'description': 'md5:dc7872fb300e143831327f1bae3af010',
585 },
586 'playlist': [{
587 'info_dict': {
588 'id': 'jqWvoWXjCVs',
589 'ext': 'mp4',
590 'title': 'teamPGP: Rocket League Noob Stream (Main Camera)',
591 'description': 'md5:dc7872fb300e143831327f1bae3af010',
592 'upload_date': '20150721',
593 'uploader': 'Beer Games Beer',
594 'uploader_id': 'beergamesbeer',
595 },
596 }, {
597 'info_dict': {
598 'id': '6h8e8xoXJzg',
599 'ext': 'mp4',
600 'title': 'teamPGP: Rocket League Noob Stream (kreestuh)',
601 'description': 'md5:dc7872fb300e143831327f1bae3af010',
602 'upload_date': '20150721',
603 'uploader': 'Beer Games Beer',
604 'uploader_id': 'beergamesbeer',
605 },
606 }, {
607 'info_dict': {
608 'id': 'PUOgX5z9xZw',
609 'ext': 'mp4',
610 'title': 'teamPGP: Rocket League Noob Stream (grizzle)',
611 'description': 'md5:dc7872fb300e143831327f1bae3af010',
612 'upload_date': '20150721',
613 'uploader': 'Beer Games Beer',
614 'uploader_id': 'beergamesbeer',
615 },
616 }, {
617 'info_dict': {
618 'id': 'teuwxikvS5k',
619 'ext': 'mp4',
620 'title': 'teamPGP: Rocket League Noob Stream (zim)',
621 'description': 'md5:dc7872fb300e143831327f1bae3af010',
622 'upload_date': '20150721',
623 'uploader': 'Beer Games Beer',
624 'uploader_id': 'beergamesbeer',
625 },
626 }],
627 'params': {
628 'skip_download': True,
629 },
630 },
631 {
632 'url': 'http://vid.plus/FlRa-iH7PGw',
633 'only_matching': True,
634 }
635 ]
636
637 def __init__(self, *args, **kwargs):
638 super(YoutubeIE, self).__init__(*args, **kwargs)
639 self._player_cache = {}
640
641 def report_video_info_webpage_download(self, video_id):
642 """Report attempt to download video info webpage."""
643 self.to_screen('%s: Downloading video info webpage' % video_id)
644
645 def report_information_extraction(self, video_id):
646 """Report attempt to extract video information."""
647 self.to_screen('%s: Extracting video information' % video_id)
648
649 def report_unavailable_format(self, video_id, format):
650 """Report extracted video URL."""
651 self.to_screen('%s: Format %s not available' % (video_id, format))
652
653 def report_rtmp_download(self):
654 """Indicate the download will use the RTMP protocol."""
655 self.to_screen('RTMP download detected')
656
657 def _signature_cache_id(self, example_sig):
658 """ Return a string representation of a signature """
659 return '.'.join(compat_str(len(part)) for part in example_sig.split('.'))
660
661 def _extract_signature_function(self, video_id, player_url, example_sig):
662 id_m = re.match(
663 r'.*?-(?P<id>[a-zA-Z0-9_-]+)(?:/watch_as3|/html5player(?:-new)?)?\.(?P<ext>[a-z]+)$',
664 player_url)
665 if not id_m:
666 raise ExtractorError('Cannot identify player %r' % player_url)
667 player_type = id_m.group('ext')
668 player_id = id_m.group('id')
669
670 # Read from filesystem cache
671 func_id = '%s_%s_%s' % (
672 player_type, player_id, self._signature_cache_id(example_sig))
673 assert os.path.basename(func_id) == func_id
674
675 cache_spec = self._downloader.cache.load('youtube-sigfuncs', func_id)
676 if cache_spec is not None:
677 return lambda s: ''.join(s[i] for i in cache_spec)
678
679 download_note = (
680 'Downloading player %s' % player_url
681 if self._downloader.params.get('verbose') else
682 'Downloading %s player %s' % (player_type, player_id)
683 )
684 if player_type == 'js':
685 code = self._download_webpage(
686 player_url, video_id,
687 note=download_note,
688 errnote='Download of %s failed' % player_url)
689 res = self._parse_sig_js(code)
690 elif player_type == 'swf':
691 urlh = self._request_webpage(
692 player_url, video_id,
693 note=download_note,
694 errnote='Download of %s failed' % player_url)
695 code = urlh.read()
696 res = self._parse_sig_swf(code)
697 else:
698 assert False, 'Invalid player type %r' % player_type
699
700 test_string = ''.join(map(compat_chr, range(len(example_sig))))
701 cache_res = res(test_string)
702 cache_spec = [ord(c) for c in cache_res]
703
704 self._downloader.cache.store('youtube-sigfuncs', func_id, cache_spec)
705 return res
706
707 def _print_sig_code(self, func, example_sig):
708 def gen_sig_code(idxs):
709 def _genslice(start, end, step):
710 starts = '' if start == 0 else str(start)
711 ends = (':%d' % (end + step)) if end + step >= 0 else ':'
712 steps = '' if step == 1 else (':%d' % step)
713 return 's[%s%s%s]' % (starts, ends, steps)
714
715 step = None
716 # Quelch pyflakes warnings - start will be set when step is set
717 start = '(Never used)'
718 for i, prev in zip(idxs[1:], idxs[:-1]):
719 if step is not None:
720 if i - prev == step:
721 continue
722 yield _genslice(start, prev, step)
723 step = None
724 continue
725 if i - prev in [-1, 1]:
726 step = i - prev
727 start = prev
728 continue
729 else:
730 yield 's[%d]' % prev
731 if step is None:
732 yield 's[%d]' % i
733 else:
734 yield _genslice(start, i, step)
735
736 test_string = ''.join(map(compat_chr, range(len(example_sig))))
737 cache_res = func(test_string)
738 cache_spec = [ord(c) for c in cache_res]
739 expr_code = ' + '.join(gen_sig_code(cache_spec))
740 signature_id_tuple = '(%s)' % (
741 ', '.join(compat_str(len(p)) for p in example_sig.split('.')))
742 code = ('if tuple(len(p) for p in s.split(\'.\')) == %s:\n'
743 ' return %s\n') % (signature_id_tuple, expr_code)
744 self.to_screen('Extracted signature function:\n' + code)
745
746 def _parse_sig_js(self, jscode):
747 funcname = self._search_regex(
748 r'\.sig\|\|([a-zA-Z0-9$]+)\(', jscode,
749 'Initial JS player signature function name')
750
751 jsi = JSInterpreter(jscode)
752 initial_function = jsi.extract_function(funcname)
753 return lambda s: initial_function([s])
754
755 def _parse_sig_swf(self, file_contents):
756 swfi = SWFInterpreter(file_contents)
757 TARGET_CLASSNAME = 'SignatureDecipher'
758 searched_class = swfi.extract_class(TARGET_CLASSNAME)
759 initial_function = swfi.extract_function(searched_class, 'decipher')
760 return lambda s: initial_function([s])
761
762 def _decrypt_signature(self, s, video_id, player_url, age_gate=False):
763 """Turn the encrypted s field into a working signature"""
764
765 if player_url is None:
766 raise ExtractorError('Cannot decrypt signature without player_url')
767
768 if player_url.startswith('//'):
769 player_url = 'https:' + player_url
770 try:
771 player_id = (player_url, self._signature_cache_id(s))
772 if player_id not in self._player_cache:
773 func = self._extract_signature_function(
774 video_id, player_url, s
775 )
776 self._player_cache[player_id] = func
777 func = self._player_cache[player_id]
778 if self._downloader.params.get('youtube_print_sig_code'):
779 self._print_sig_code(func, s)
780 return func(s)
781 except Exception as e:
782 tb = traceback.format_exc()
783 raise ExtractorError(
784 'Signature extraction failed: ' + tb, cause=e)
785
786 def _get_subtitles(self, video_id, webpage):
787 try:
788 subs_doc = self._download_xml(
789 'https://video.google.com/timedtext?hl=en&type=list&v=%s' % video_id,
790 video_id, note=False)
791 except ExtractorError as err:
792 self._downloader.report_warning('unable to download video subtitles: %s' % compat_str(err))
793 return {}
794
795 sub_lang_list = {}
796 for track in subs_doc.findall('track'):
797 lang = track.attrib['lang_code']
798 if lang in sub_lang_list:
799 continue
800 sub_formats = []
801 for ext in ['sbv', 'vtt', 'srt']:
802 params = compat_urllib_parse.urlencode({
803 'lang': lang,
804 'v': video_id,
805 'fmt': ext,
806 'name': track.attrib['name'].encode('utf-8'),
807 })
808 sub_formats.append({
809 'url': 'https://www.youtube.com/api/timedtext?' + params,
810 'ext': ext,
811 })
812 sub_lang_list[lang] = sub_formats
813 if not sub_lang_list:
814 self._downloader.report_warning('video doesn\'t have subtitles')
815 return {}
816 return sub_lang_list
817
818 def _get_automatic_captions(self, video_id, webpage):
819 """We need the webpage for getting the captions url, pass it as an
820 argument to speed up the process."""
821 self.to_screen('%s: Looking for automatic captions' % video_id)
822 mobj = re.search(r';ytplayer.config = ({.*?});', webpage)
823 err_msg = 'Couldn\'t find automatic captions for %s' % video_id
824 if mobj is None:
825 self._downloader.report_warning(err_msg)
826 return {}
827 player_config = json.loads(mobj.group(1))
828 try:
829 args = player_config['args']
830 caption_url = args['ttsurl']
831 timestamp = args['timestamp']
832 # We get the available subtitles
833 list_params = compat_urllib_parse.urlencode({
834 'type': 'list',
835 'tlangs': 1,
836 'asrs': 1,
837 })
838 list_url = caption_url + '&' + list_params
839 caption_list = self._download_xml(list_url, video_id)
840 original_lang_node = caption_list.find('track')
841 if original_lang_node is None:
842 self._downloader.report_warning('Video doesn\'t have automatic captions')
843 return {}
844 original_lang = original_lang_node.attrib['lang_code']
845 caption_kind = original_lang_node.attrib.get('kind', '')
846
847 sub_lang_list = {}
848 for lang_node in caption_list.findall('target'):
849 sub_lang = lang_node.attrib['lang_code']
850 sub_formats = []
851 for ext in ['sbv', 'vtt', 'srt']:
852 params = compat_urllib_parse.urlencode({
853 'lang': original_lang,
854 'tlang': sub_lang,
855 'fmt': ext,
856 'ts': timestamp,
857 'kind': caption_kind,
858 })
859 sub_formats.append({
860 'url': caption_url + '&' + params,
861 'ext': ext,
862 })
863 sub_lang_list[sub_lang] = sub_formats
864 return sub_lang_list
865 # An extractor error can be raise by the download process if there are
866 # no automatic captions but there are subtitles
867 except (KeyError, ExtractorError):
868 self._downloader.report_warning(err_msg)
869 return {}
870
871 @classmethod
872 def extract_id(cls, url):
873 mobj = re.match(cls._VALID_URL, url, re.VERBOSE)
874 if mobj is None:
875 raise ExtractorError('Invalid URL: %s' % url)
876 video_id = mobj.group(2)
877 return video_id
878
879 def _extract_from_m3u8(self, manifest_url, video_id):
880 url_map = {}
881
882 def _get_urls(_manifest):
883 lines = _manifest.split('\n')
884 urls = filter(lambda l: l and not l.startswith('#'),
885 lines)
886 return urls
887 manifest = self._download_webpage(manifest_url, video_id, 'Downloading formats manifest')
888 formats_urls = _get_urls(manifest)
889 for format_url in formats_urls:
890 itag = self._search_regex(r'itag/(\d+?)/', format_url, 'itag')
891 url_map[itag] = format_url
892 return url_map
893
894 def _extract_annotations(self, video_id):
895 url = 'https://www.youtube.com/annotations_invideo?features=1&legacy=1&video_id=%s' % video_id
896 return self._download_webpage(url, video_id, note='Searching for annotations.', errnote='Unable to download video annotations.')
897
898 def _parse_dash_manifest(
899 self, video_id, dash_manifest_url, player_url, age_gate, fatal=True):
900 def decrypt_sig(mobj):
901 s = mobj.group(1)
902 dec_s = self._decrypt_signature(s, video_id, player_url, age_gate)
903 return '/signature/%s' % dec_s
904 dash_manifest_url = re.sub(r'/s/([a-fA-F0-9\.]+)', decrypt_sig, dash_manifest_url)
905 dash_doc = self._download_xml(
906 dash_manifest_url, video_id,
907 note='Downloading DASH manifest',
908 errnote='Could not download DASH manifest',
909 fatal=fatal)
910
911 if dash_doc is False:
912 return []
913
914 formats = []
915 for a in dash_doc.findall('.//{urn:mpeg:DASH:schema:MPD:2011}AdaptationSet'):
916 mime_type = a.attrib.get('mimeType')
917 for r in a.findall('{urn:mpeg:DASH:schema:MPD:2011}Representation'):
918 url_el = r.find('{urn:mpeg:DASH:schema:MPD:2011}BaseURL')
919 if url_el is None:
920 continue
921 if mime_type == 'text/vtt':
922 # TODO implement WebVTT downloading
923 pass
924 elif mime_type.startswith('audio/') or mime_type.startswith('video/'):
925 segment_list = r.find('{urn:mpeg:DASH:schema:MPD:2011}SegmentList')
926 format_id = r.attrib['id']
927 video_url = url_el.text
928 filesize = int_or_none(url_el.attrib.get('{http://youtube.com/yt/2012/10/10}contentLength'))
929 f = {
930 'format_id': format_id,
931 'url': video_url,
932 'width': int_or_none(r.attrib.get('width')),
933 'height': int_or_none(r.attrib.get('height')),
934 'tbr': int_or_none(r.attrib.get('bandwidth'), 1000),
935 'asr': int_or_none(r.attrib.get('audioSamplingRate')),
936 'filesize': filesize,
937 'fps': int_or_none(r.attrib.get('frameRate')),
938 }
939 if segment_list is not None:
940 f.update({
941 'initialization_url': segment_list.find('{urn:mpeg:DASH:schema:MPD:2011}Initialization').attrib['sourceURL'],
942 'segment_urls': [segment.attrib.get('media') for segment in segment_list.findall('{urn:mpeg:DASH:schema:MPD:2011}SegmentURL')],
943 'protocol': 'http_dash_segments',
944 })
945 try:
946 existing_format = next(
947 fo for fo in formats
948 if fo['format_id'] == format_id)
949 except StopIteration:
950 full_info = self._formats.get(format_id, {}).copy()
951 full_info.update(f)
952 codecs = r.attrib.get('codecs')
953 if codecs:
954 if full_info.get('acodec') == 'none' and 'vcodec' not in full_info:
955 full_info['vcodec'] = codecs
956 elif full_info.get('vcodec') == 'none' and 'acodec' not in full_info:
957 full_info['acodec'] = codecs
958 formats.append(full_info)
959 else:
960 existing_format.update(f)
961 else:
962 self.report_warning('Unknown MIME type %s in DASH manifest' % mime_type)
963 return formats
964
965 def _real_extract(self, url):
966 url, smuggled_data = unsmuggle_url(url, {})
967
968 proto = (
969 'http' if self._downloader.params.get('prefer_insecure', False)
970 else 'https')
971
972 start_time = None
973 end_time = None
974 parsed_url = compat_urllib_parse_urlparse(url)
975 for component in [parsed_url.fragment, parsed_url.query]:
976 query = compat_parse_qs(component)
977 if start_time is None and 't' in query:
978 start_time = parse_duration(query['t'][0])
979 if start_time is None and 'start' in query:
980 start_time = parse_duration(query['start'][0])
981 if end_time is None and 'end' in query:
982 end_time = parse_duration(query['end'][0])
983
984 # Extract original video URL from URL with redirection, like age verification, using next_url parameter
985 mobj = re.search(self._NEXT_URL_RE, url)
986 if mobj:
987 url = proto + '://www.youtube.com/' + compat_urllib_parse_unquote(mobj.group(1)).lstrip('/')
988 video_id = self.extract_id(url)
989
990 # Get video webpage
991 url = proto + '://www.youtube.com/watch?v=%s&gl=US&hl=en&has_verified=1&bpctr=9999999999' % video_id
992 video_webpage = self._download_webpage(url, video_id)
993
994 # Attempt to extract SWF player URL
995 mobj = re.search(r'swfConfig.*?"(https?:\\/\\/.*?watch.*?-.*?\.swf)"', video_webpage)
996 if mobj is not None:
997 player_url = re.sub(r'\\(.)', r'\1', mobj.group(1))
998 else:
999 player_url = None
1000
1001 dash_mpds = []
1002
1003 def add_dash_mpd(video_info):
1004 dash_mpd = video_info.get('dashmpd')
1005 if dash_mpd and dash_mpd[0] not in dash_mpds:
1006 dash_mpds.append(dash_mpd[0])
1007
1008 # Get video info
1009 embed_webpage = None
1010 is_live = None
1011 if re.search(r'player-age-gate-content">', video_webpage) is not None:
1012 age_gate = True
1013 # We simulate the access to the video from www.youtube.com/v/{video_id}
1014 # this can be viewed without login into Youtube
1015 url = proto + '://www.youtube.com/embed/%s' % video_id
1016 embed_webpage = self._download_webpage(url, video_id, 'Downloading embed webpage')
1017 data = compat_urllib_parse.urlencode({
1018 'video_id': video_id,
1019 'eurl': 'https://youtube.googleapis.com/v/' + video_id,
1020 'sts': self._search_regex(
1021 r'"sts"\s*:\s*(\d+)', embed_webpage, 'sts', default=''),
1022 })
1023 video_info_url = proto + '://www.youtube.com/get_video_info?' + data
1024 video_info_webpage = self._download_webpage(
1025 video_info_url, video_id,
1026 note='Refetching age-gated info webpage',
1027 errnote='unable to download video info webpage')
1028 video_info = compat_parse_qs(video_info_webpage)
1029 add_dash_mpd(video_info)
1030 else:
1031 age_gate = False
1032 video_info = None
1033 # Try looking directly into the video webpage
1034 mobj = re.search(r';ytplayer\.config\s*=\s*({.*?});', video_webpage)
1035 if mobj:
1036 json_code = uppercase_escape(mobj.group(1))
1037 ytplayer_config = json.loads(json_code)
1038 args = ytplayer_config['args']
1039 if args.get('url_encoded_fmt_stream_map'):
1040 # Convert to the same format returned by compat_parse_qs
1041 video_info = dict((k, [v]) for k, v in args.items())
1042 add_dash_mpd(video_info)
1043 if args.get('livestream') == '1' or args.get('live_playback') == 1:
1044 is_live = True
1045 if not video_info or self._downloader.params.get('youtube_include_dash_manifest', True):
1046 # We also try looking in get_video_info since it may contain different dashmpd
1047 # URL that points to a DASH manifest with possibly different itag set (some itags
1048 # are missing from DASH manifest pointed by webpage's dashmpd, some - from DASH
1049 # manifest pointed by get_video_info's dashmpd).
1050 # The general idea is to take a union of itags of both DASH manifests (for example
1051 # video with such 'manifest behavior' see https://github.com/rg3/youtube-dl/issues/6093)
1052 self.report_video_info_webpage_download(video_id)
1053 for el_type in ['&el=info', '&el=embedded', '&el=detailpage', '&el=vevo', '']:
1054 video_info_url = (
1055 '%s://www.youtube.com/get_video_info?&video_id=%s%s&ps=default&eurl=&gl=US&hl=en'
1056 % (proto, video_id, el_type))
1057 video_info_webpage = self._download_webpage(
1058 video_info_url,
1059 video_id, note=False,
1060 errnote='unable to download video info webpage')
1061 get_video_info = compat_parse_qs(video_info_webpage)
1062 if get_video_info.get('use_cipher_signature') != ['True']:
1063 add_dash_mpd(get_video_info)
1064 if not video_info:
1065 video_info = get_video_info
1066 if 'token' in get_video_info:
1067 break
1068 if 'token' not in video_info:
1069 if 'reason' in video_info:
1070 if 'The uploader has not made this video available in your country.' in video_info['reason']:
1071 regions_allowed = self._html_search_meta('regionsAllowed', video_webpage, default=None)
1072 if regions_allowed:
1073 raise ExtractorError('YouTube said: This video is available in %s only' % (
1074 ', '.join(map(ISO3166Utils.short2full, regions_allowed.split(',')))),
1075 expected=True)
1076 raise ExtractorError(
1077 'YouTube said: %s' % video_info['reason'][0],
1078 expected=True, video_id=video_id)
1079 else:
1080 raise ExtractorError(
1081 '"token" parameter not in video info for unknown reason',
1082 video_id=video_id)
1083
1084 # title
1085 if 'title' in video_info:
1086 video_title = video_info['title'][0]
1087 else:
1088 self._downloader.report_warning('Unable to extract video title')
1089 video_title = '_'
1090
1091 # description
1092 video_description = get_element_by_id("eow-description", video_webpage)
1093 if video_description:
1094 video_description = re.sub(r'''(?x)
1095 <a\s+
1096 (?:[a-zA-Z-]+="[^"]+"\s+)*?
1097 title="([^"]+)"\s+
1098 (?:[a-zA-Z-]+="[^"]+"\s+)*?
1099 class="yt-uix-redirect-link"\s*>
1100 [^<]+
1101 </a>
1102 ''', r'\1', video_description)
1103 video_description = clean_html(video_description)
1104 else:
1105 fd_mobj = re.search(r'<meta name="description" content="([^"]+)"', video_webpage)
1106 if fd_mobj:
1107 video_description = unescapeHTML(fd_mobj.group(1))
1108 else:
1109 video_description = ''
1110
1111 if 'multifeed_metadata_list' in video_info and not smuggled_data.get('force_singlefeed', False):
1112 if not self._downloader.params.get('noplaylist'):
1113 entries = []
1114 feed_ids = []
1115 multifeed_metadata_list = compat_urllib_parse_unquote_plus(video_info['multifeed_metadata_list'][0])
1116 for feed in multifeed_metadata_list.split(','):
1117 feed_data = compat_parse_qs(feed)
1118 entries.append({
1119 '_type': 'url_transparent',
1120 'ie_key': 'Youtube',
1121 'url': smuggle_url(
1122 '%s://www.youtube.com/watch?v=%s' % (proto, feed_data['id'][0]),
1123 {'force_singlefeed': True}),
1124 'title': '%s (%s)' % (video_title, feed_data['title'][0]),
1125 })
1126 feed_ids.append(feed_data['id'][0])
1127 self.to_screen(
1128 'Downloading multifeed video (%s) - add --no-playlist to just download video %s'
1129 % (', '.join(feed_ids), video_id))
1130 return self.playlist_result(entries, video_id, video_title, video_description)
1131 self.to_screen('Downloading just video %s because of --no-playlist' % video_id)
1132
1133 if 'view_count' in video_info:
1134 view_count = int(video_info['view_count'][0])
1135 else:
1136 view_count = None
1137
1138 # Check for "rental" videos
1139 if 'ypc_video_rental_bar_text' in video_info and 'author' not in video_info:
1140 raise ExtractorError('"rental" videos not supported')
1141
1142 # Start extracting information
1143 self.report_information_extraction(video_id)
1144
1145 # uploader
1146 if 'author' not in video_info:
1147 raise ExtractorError('Unable to extract uploader name')
1148 video_uploader = compat_urllib_parse_unquote_plus(video_info['author'][0])
1149
1150 # uploader_id
1151 video_uploader_id = None
1152 mobj = re.search(r'<link itemprop="url" href="http://www.youtube.com/(?:user|channel)/([^"]+)">', video_webpage)
1153 if mobj is not None:
1154 video_uploader_id = mobj.group(1)
1155 else:
1156 self._downloader.report_warning('unable to extract uploader nickname')
1157
1158 # thumbnail image
1159 # We try first to get a high quality image:
1160 m_thumb = re.search(r'<span itemprop="thumbnail".*?href="(.*?)">',
1161 video_webpage, re.DOTALL)
1162 if m_thumb is not None:
1163 video_thumbnail = m_thumb.group(1)
1164 elif 'thumbnail_url' not in video_info:
1165 self._downloader.report_warning('unable to extract video thumbnail')
1166 video_thumbnail = None
1167 else: # don't panic if we can't find it
1168 video_thumbnail = compat_urllib_parse_unquote_plus(video_info['thumbnail_url'][0])
1169
1170 # upload date
1171 upload_date = self._html_search_meta(
1172 'datePublished', video_webpage, 'upload date', default=None)
1173 if not upload_date:
1174 upload_date = self._search_regex(
1175 [r'(?s)id="eow-date.*?>(.*?)</span>',
1176 r'id="watch-uploader-info".*?>.*?(?:Published|Uploaded|Streamed live|Started) on (.+?)</strong>'],
1177 video_webpage, 'upload date', default=None)
1178 if upload_date:
1179 upload_date = ' '.join(re.sub(r'[/,-]', r' ', mobj.group(1)).split())
1180 upload_date = unified_strdate(upload_date)
1181
1182 m_cat_container = self._search_regex(
1183 r'(?s)<h4[^>]*>\s*Category\s*</h4>\s*<ul[^>]*>(.*?)</ul>',
1184 video_webpage, 'categories', default=None)
1185 if m_cat_container:
1186 category = self._html_search_regex(
1187 r'(?s)<a[^<]+>(.*?)</a>', m_cat_container, 'category',
1188 default=None)
1189 video_categories = None if category is None else [category]
1190 else:
1191 video_categories = None
1192
1193 video_tags = [
1194 unescapeHTML(m.group('content'))
1195 for m in re.finditer(self._meta_regex('og:video:tag'), video_webpage)]
1196
1197 def _extract_count(count_name):
1198 return str_to_int(self._search_regex(
1199 r'-%s-button[^>]+><span[^>]+class="yt-uix-button-content"[^>]*>([\d,]+)</span>'
1200 % re.escape(count_name),
1201 video_webpage, count_name, default=None))
1202
1203 like_count = _extract_count('like')
1204 dislike_count = _extract_count('dislike')
1205
1206 # subtitles
1207 video_subtitles = self.extract_subtitles(video_id, video_webpage)
1208 automatic_captions = self.extract_automatic_captions(video_id, video_webpage)
1209
1210 if 'length_seconds' not in video_info:
1211 self._downloader.report_warning('unable to extract video duration')
1212 video_duration = None
1213 else:
1214 video_duration = int(compat_urllib_parse_unquote_plus(video_info['length_seconds'][0]))
1215
1216 # annotations
1217 video_annotations = None
1218 if self._downloader.params.get('writeannotations', False):
1219 video_annotations = self._extract_annotations(video_id)
1220
1221 def _map_to_format_list(urlmap):
1222 formats = []
1223 for itag, video_real_url in urlmap.items():
1224 dct = {
1225 'format_id': itag,
1226 'url': video_real_url,
1227 'player_url': player_url,
1228 }
1229 if itag in self._formats:
1230 dct.update(self._formats[itag])
1231 formats.append(dct)
1232 return formats
1233
1234 if 'conn' in video_info and video_info['conn'][0].startswith('rtmp'):
1235 self.report_rtmp_download()
1236 formats = [{
1237 'format_id': '_rtmp',
1238 'protocol': 'rtmp',
1239 'url': video_info['conn'][0],
1240 'player_url': player_url,
1241 }]
1242 elif len(video_info.get('url_encoded_fmt_stream_map', [''])[0]) >= 1 or len(video_info.get('adaptive_fmts', [''])[0]) >= 1:
1243 encoded_url_map = video_info.get('url_encoded_fmt_stream_map', [''])[0] + ',' + video_info.get('adaptive_fmts', [''])[0]
1244 if 'rtmpe%3Dyes' in encoded_url_map:
1245 raise ExtractorError('rtmpe downloads are not supported, see https://github.com/rg3/youtube-dl/issues/343 for more information.', expected=True)
1246 formats = []
1247 for url_data_str in encoded_url_map.split(','):
1248 url_data = compat_parse_qs(url_data_str)
1249 if 'itag' not in url_data or 'url' not in url_data:
1250 continue
1251 format_id = url_data['itag'][0]
1252 url = url_data['url'][0]
1253
1254 if 'sig' in url_data:
1255 url += '&signature=' + url_data['sig'][0]
1256 elif 's' in url_data:
1257 encrypted_sig = url_data['s'][0]
1258 ASSETS_RE = r'"assets":.+?"js":\s*("[^"]+")'
1259
1260 jsplayer_url_json = self._search_regex(
1261 ASSETS_RE,
1262 embed_webpage if age_gate else video_webpage,
1263 'JS player URL (1)', default=None)
1264 if not jsplayer_url_json and not age_gate:
1265 # We need the embed website after all
1266 if embed_webpage is None:
1267 embed_url = proto + '://www.youtube.com/embed/%s' % video_id
1268 embed_webpage = self._download_webpage(
1269 embed_url, video_id, 'Downloading embed webpage')
1270 jsplayer_url_json = self._search_regex(
1271 ASSETS_RE, embed_webpage, 'JS player URL')
1272
1273 player_url = json.loads(jsplayer_url_json)
1274 if player_url is None:
1275 player_url_json = self._search_regex(
1276 r'ytplayer\.config.*?"url"\s*:\s*("[^"]+")',
1277 video_webpage, 'age gate player URL')
1278 player_url = json.loads(player_url_json)
1279
1280 if self._downloader.params.get('verbose'):
1281 if player_url is None:
1282 player_version = 'unknown'
1283 player_desc = 'unknown'
1284 else:
1285 if player_url.endswith('swf'):
1286 player_version = self._search_regex(
1287 r'-(.+?)(?:/watch_as3)?\.swf$', player_url,
1288 'flash player', fatal=False)
1289 player_desc = 'flash player %s' % player_version
1290 else:
1291 player_version = self._search_regex(
1292 r'html5player-([^/]+?)(?:/html5player(?:-new)?)?\.js',
1293 player_url,
1294 'html5 player', fatal=False)
1295 player_desc = 'html5 player %s' % player_version
1296
1297 parts_sizes = self._signature_cache_id(encrypted_sig)
1298 self.to_screen('{%s} signature length %s, %s' %
1299 (format_id, parts_sizes, player_desc))
1300
1301 signature = self._decrypt_signature(
1302 encrypted_sig, video_id, player_url, age_gate)
1303 url += '&signature=' + signature
1304 if 'ratebypass' not in url:
1305 url += '&ratebypass=yes'
1306
1307 width = None
1308 height = None
1309 size_str = url_data.get('size', [''])[0]
1310 if size_str.count('x') == 1:
1311 width, height = [int_or_none(x) for x in size_str.split('x')]
1312
1313 format_url = {
1314 'format_id': format_id,
1315 'url': url,
1316 'player_url': player_url,
1317 # As of this writing these are only defined for DASH formats:
1318 'filesize': int_or_none(url_data.get('clen', [None])[0]),
1319 'tbr': float_or_none(url_data.get('bitrate', [None])[0], scale=1024),
1320 'width': width,
1321 'height': height,
1322 'fps': int_or_none(url_data.get('fps', [None])[0]),
1323 }
1324
1325 # drop Nones so they do not overwrite the defaults from self._formats
1326 format_url = dict((k, v) for k, v in format_url.items() if v is not None)
1327
1328 format_full = self._formats.get(format_id, {}).copy()
1329 format_full.update(format_url)
1330
1331 formats.append(format_full)
1332
1333 elif video_info.get('hlsvp'):
1334 manifest_url = video_info['hlsvp'][0]
1335 url_map = self._extract_from_m3u8(manifest_url, video_id)
1336 formats = _map_to_format_list(url_map)
1337 else:
1338 raise ExtractorError('no conn, hlsvp or url_encoded_fmt_stream_map information found in video info')
1339
1340 # Look for the DASH manifest
1341 if self._downloader.params.get('youtube_include_dash_manifest', True):
1342 dash_mpd_fatal = True
1343 for dash_manifest_url in dash_mpds:
1344 dash_formats = {}
1345 try:
1346 for df in self._parse_dash_manifest(
1347 video_id, dash_manifest_url, player_url, age_gate, dash_mpd_fatal):
1348 # Do not overwrite DASH format found in some previous DASH manifest
1349 if df['format_id'] not in dash_formats:
1350 dash_formats[df['format_id']] = df
1351 # Additional DASH manifests may end up in HTTP Error 403 therefore
1352 # allow them to fail without bug report message if we already have
1353 # some DASH manifest succeeded. This is temporary workaround to reduce
1354 # burst of bug reports until we figure out the reason and whether it
1355 # can be fixed at all.
1356 dash_mpd_fatal = False
1357 except (ExtractorError, KeyError) as e:
1358 self.report_warning(
1359 'Skipping DASH manifest: %r' % e, video_id)
1360 if dash_formats:
1361 # Remove the formats we found through non-DASH, they
1362 # contain less info and it can be wrong, because we use
1363 # fixed values (for example the resolution). See
1364 # https://github.com/rg3/youtube-dl/issues/5774 for an
1365 # example.
1366 formats = [f for f in formats if f['format_id'] not in dash_formats.keys()]
1367 formats.extend(dash_formats.values())
1368
1369 # Check for malformed aspect ratio
1370 stretched_m = re.search(
1371 r'<meta\s+property="og:video:tag".*?content="yt:stretch=(?P<w>[0-9]+):(?P<h>[0-9]+)">',
1372 video_webpage)
1373 if stretched_m:
1374 ratio = float(stretched_m.group('w')) / float(stretched_m.group('h'))
1375 for f in formats:
1376 if f.get('vcodec') != 'none':
1377 f['stretched_ratio'] = ratio
1378
1379 self._sort_formats(formats)
1380
1381 return {
1382 'id': video_id,
1383 'uploader': video_uploader,
1384 'uploader_id': video_uploader_id,
1385 'upload_date': upload_date,
1386 'title': video_title,
1387 'thumbnail': video_thumbnail,
1388 'description': video_description,
1389 'categories': video_categories,
1390 'tags': video_tags,
1391 'subtitles': video_subtitles,
1392 'automatic_captions': automatic_captions,
1393 'duration': video_duration,
1394 'age_limit': 18 if age_gate else 0,
1395 'annotations': video_annotations,
1396 'webpage_url': proto + '://www.youtube.com/watch?v=%s' % video_id,
1397 'view_count': view_count,
1398 'like_count': like_count,
1399 'dislike_count': dislike_count,
1400 'average_rating': float_or_none(video_info.get('avg_rating', [None])[0]),
1401 'formats': formats,
1402 'is_live': is_live,
1403 'start_time': start_time,
1404 'end_time': end_time,
1405 }
1406
1407
1408 class YoutubePlaylistIE(YoutubeBaseInfoExtractor):
1409 IE_DESC = 'YouTube.com playlists'
1410 _VALID_URL = r"""(?x)(?:
1411 (?:https?://)?
1412 (?:\w+\.)?
1413 youtube\.com/
1414 (?:
1415 (?:course|view_play_list|my_playlists|artist|playlist|watch|embed/videoseries)
1416 \? (?:.*?&)*? (?:p|a|list)=
1417 | p/
1418 )
1419 (
1420 (?:PL|LL|EC|UU|FL|RD|UL)?[0-9A-Za-z-_]{10,}
1421 # Top tracks, they can also include dots
1422 |(?:MC)[\w\.]*
1423 )
1424 .*
1425 |
1426 ((?:PL|LL|EC|UU|FL|RD|UL)[0-9A-Za-z-_]{10,})
1427 )"""
1428 _TEMPLATE_URL = 'https://www.youtube.com/playlist?list=%s'
1429 _VIDEO_RE = r'href="\s*/watch\?v=(?P<id>[0-9A-Za-z_-]{11})&amp;[^"]*?index=(?P<index>\d+)'
1430 IE_NAME = 'youtube:playlist'
1431 _TESTS = [{
1432 'url': 'https://www.youtube.com/playlist?list=PLwiyx1dc3P2JR9N8gQaQN_BCvlSlap7re',
1433 'info_dict': {
1434 'title': 'ytdl test PL',
1435 'id': 'PLwiyx1dc3P2JR9N8gQaQN_BCvlSlap7re',
1436 },
1437 'playlist_count': 3,
1438 }, {
1439 'url': 'https://www.youtube.com/playlist?list=PLtPgu7CB4gbZDA7i_euNxn75ISqxwZPYx',
1440 'info_dict': {
1441 'id': 'PLtPgu7CB4gbZDA7i_euNxn75ISqxwZPYx',
1442 'title': 'YDL_Empty_List',
1443 },
1444 'playlist_count': 0,
1445 }, {
1446 'note': 'Playlist with deleted videos (#651). As a bonus, the video #51 is also twice in this list.',
1447 'url': 'https://www.youtube.com/playlist?list=PLwP_SiAcdui0KVebT0mU9Apz359a4ubsC',
1448 'info_dict': {
1449 'title': '29C3: Not my department',
1450 'id': 'PLwP_SiAcdui0KVebT0mU9Apz359a4ubsC',
1451 },
1452 'playlist_count': 95,
1453 }, {
1454 'note': 'issue #673',
1455 'url': 'PLBB231211A4F62143',
1456 'info_dict': {
1457 'title': '[OLD]Team Fortress 2 (Class-based LP)',
1458 'id': 'PLBB231211A4F62143',
1459 },
1460 'playlist_mincount': 26,
1461 }, {
1462 'note': 'Large playlist',
1463 'url': 'https://www.youtube.com/playlist?list=UUBABnxM4Ar9ten8Mdjj1j0Q',
1464 'info_dict': {
1465 'title': 'Uploads from Cauchemar',
1466 'id': 'UUBABnxM4Ar9ten8Mdjj1j0Q',
1467 },
1468 'playlist_mincount': 799,
1469 }, {
1470 'url': 'PLtPgu7CB4gbY9oDN3drwC3cMbJggS7dKl',
1471 'info_dict': {
1472 'title': 'YDL_safe_search',
1473 'id': 'PLtPgu7CB4gbY9oDN3drwC3cMbJggS7dKl',
1474 },
1475 'playlist_count': 2,
1476 }, {
1477 'note': 'embedded',
1478 'url': 'http://www.youtube.com/embed/videoseries?list=PL6IaIsEjSbf96XFRuNccS_RuEXwNdsoEu',
1479 'playlist_count': 4,
1480 'info_dict': {
1481 'title': 'JODA15',
1482 'id': 'PL6IaIsEjSbf96XFRuNccS_RuEXwNdsoEu',
1483 }
1484 }, {
1485 'note': 'Embedded SWF player',
1486 'url': 'http://www.youtube.com/p/YN5VISEtHet5D4NEvfTd0zcgFk84NqFZ?hl=en_US&fs=1&rel=0',
1487 'playlist_count': 4,
1488 'info_dict': {
1489 'title': 'JODA7',
1490 'id': 'YN5VISEtHet5D4NEvfTd0zcgFk84NqFZ',
1491 }
1492 }, {
1493 'note': 'Buggy playlist: the webpage has a "Load more" button but it doesn\'t have more videos',
1494 'url': 'https://www.youtube.com/playlist?list=UUXw-G3eDE9trcvY2sBMM_aA',
1495 'info_dict': {
1496 'title': 'Uploads from Interstellar Movie',
1497 'id': 'UUXw-G3eDE9trcvY2sBMM_aA',
1498 },
1499 'playlist_mincout': 21,
1500 }]
1501
1502 def _real_initialize(self):
1503 self._login()
1504
1505 def _extract_mix(self, playlist_id):
1506 # The mixes are generated from a single video
1507 # the id of the playlist is just 'RD' + video_id
1508 url = 'https://youtube.com/watch?v=%s&list=%s' % (playlist_id[-11:], playlist_id)
1509 webpage = self._download_webpage(
1510 url, playlist_id, 'Downloading Youtube mix')
1511 search_title = lambda class_name: get_element_by_attribute('class', class_name, webpage)
1512 title_span = (
1513 search_title('playlist-title') or
1514 search_title('title long-title') or
1515 search_title('title'))
1516 title = clean_html(title_span)
1517 ids = orderedSet(re.findall(
1518 r'''(?xs)data-video-username=".*?".*?
1519 href="/watch\?v=([0-9A-Za-z_-]{11})&amp;[^"]*?list=%s''' % re.escape(playlist_id),
1520 webpage))
1521 url_results = self._ids_to_results(ids)
1522
1523 return self.playlist_result(url_results, playlist_id, title)
1524
1525 def _extract_playlist(self, playlist_id):
1526 url = self._TEMPLATE_URL % playlist_id
1527 page = self._download_webpage(url, playlist_id)
1528
1529 for match in re.findall(r'<div class="yt-alert-message">([^<]+)</div>', page):
1530 match = match.strip()
1531 # Check if the playlist exists or is private
1532 if re.match(r'[^<]*(The|This) playlist (does not exist|is private)[^<]*', match):
1533 raise ExtractorError(
1534 'The playlist doesn\'t exist or is private, use --username or '
1535 '--netrc to access it.',
1536 expected=True)
1537 elif re.match(r'[^<]*Invalid parameters[^<]*', match):
1538 raise ExtractorError(
1539 'Invalid parameters. Maybe URL is incorrect.',
1540 expected=True)
1541 elif re.match(r'[^<]*Choose your language[^<]*', match):
1542 continue
1543 else:
1544 self.report_warning('Youtube gives an alert message: ' + match)
1545
1546 # Extract the video ids from the playlist pages
1547 def _entries():
1548 more_widget_html = content_html = page
1549 for page_num in itertools.count(1):
1550 matches = re.finditer(self._VIDEO_RE, content_html)
1551 # We remove the duplicates and the link with index 0
1552 # (it's not the first video of the playlist)
1553 new_ids = orderedSet(m.group('id') for m in matches if m.group('index') != '0')
1554 for vid_id in new_ids:
1555 yield self.url_result(vid_id, 'Youtube', video_id=vid_id)
1556
1557 mobj = re.search(r'data-uix-load-more-href="/?(?P<more>[^"]+)"', more_widget_html)
1558 if not mobj:
1559 break
1560
1561 more = self._download_json(
1562 'https://youtube.com/%s' % mobj.group('more'), playlist_id,
1563 'Downloading page #%s' % page_num,
1564 transform_source=uppercase_escape)
1565 content_html = more['content_html']
1566 if not content_html.strip():
1567 # Some webpages show a "Load more" button but they don't
1568 # have more videos
1569 break
1570 more_widget_html = more['load_more_widget_html']
1571
1572 playlist_title = self._html_search_regex(
1573 r'(?s)<h1 class="pl-header-title[^"]*">\s*(.*?)\s*</h1>',
1574 page, 'title')
1575
1576 return self.playlist_result(_entries(), playlist_id, playlist_title)
1577
1578 def _real_extract(self, url):
1579 # Extract playlist id
1580 mobj = re.match(self._VALID_URL, url)
1581 if mobj is None:
1582 raise ExtractorError('Invalid URL: %s' % url)
1583 playlist_id = mobj.group(1) or mobj.group(2)
1584
1585 # Check if it's a video-specific URL
1586 query_dict = compat_urlparse.parse_qs(compat_urlparse.urlparse(url).query)
1587 if 'v' in query_dict:
1588 video_id = query_dict['v'][0]
1589 if self._downloader.params.get('noplaylist'):
1590 self.to_screen('Downloading just video %s because of --no-playlist' % video_id)
1591 return self.url_result(video_id, 'Youtube', video_id=video_id)
1592 else:
1593 self.to_screen('Downloading playlist %s - add --no-playlist to just download video %s' % (playlist_id, video_id))
1594
1595 if playlist_id.startswith('RD') or playlist_id.startswith('UL'):
1596 # Mixes require a custom extraction process
1597 return self._extract_mix(playlist_id)
1598
1599 return self._extract_playlist(playlist_id)
1600
1601
1602 class YoutubeChannelIE(InfoExtractor):
1603 IE_DESC = 'YouTube.com channels'
1604 _VALID_URL = r'https?://(?:youtu\.be|(?:\w+\.)?youtube(?:-nocookie)?\.com)/channel/(?P<id>[0-9A-Za-z_-]+)'
1605 _TEMPLATE_URL = 'https://www.youtube.com/channel/%s/videos'
1606 IE_NAME = 'youtube:channel'
1607 _TESTS = [{
1608 'note': 'paginated channel',
1609 'url': 'https://www.youtube.com/channel/UCKfVa3S1e4PHvxWcwyMMg8w',
1610 'playlist_mincount': 91,
1611 'info_dict': {
1612 'id': 'UCKfVa3S1e4PHvxWcwyMMg8w',
1613 }
1614 }]
1615
1616 @staticmethod
1617 def extract_videos_from_page(page):
1618 ids_in_page = []
1619 titles_in_page = []
1620 for mobj in re.finditer(r'(?:title="(?P<title>[^"]+)"[^>]+)?href="/watch\?v=(?P<id>[0-9A-Za-z_-]+)&?', page):
1621 video_id = mobj.group('id')
1622 video_title = unescapeHTML(mobj.group('title'))
1623 try:
1624 idx = ids_in_page.index(video_id)
1625 if video_title and not titles_in_page[idx]:
1626 titles_in_page[idx] = video_title
1627 except ValueError:
1628 ids_in_page.append(video_id)
1629 titles_in_page.append(video_title)
1630 return zip(ids_in_page, titles_in_page)
1631
1632 def _real_extract(self, url):
1633 channel_id = self._match_id(url)
1634
1635 url = self._TEMPLATE_URL % channel_id
1636
1637 # Channel by page listing is restricted to 35 pages of 30 items, i.e. 1050 videos total (see #5778)
1638 # Workaround by extracting as a playlist if managed to obtain channel playlist URL
1639 # otherwise fallback on channel by page extraction
1640 channel_page = self._download_webpage(
1641 url + '?view=57', channel_id,
1642 'Downloading channel page', fatal=False)
1643 channel_playlist_id = self._html_search_meta(
1644 'channelId', channel_page, 'channel id', default=None)
1645 if not channel_playlist_id:
1646 channel_playlist_id = self._search_regex(
1647 r'data-channel-external-id="([^"]+)"',
1648 channel_page, 'channel id', default=None)
1649 if channel_playlist_id and channel_playlist_id.startswith('UC'):
1650 playlist_id = 'UU' + channel_playlist_id[2:]
1651 return self.url_result(
1652 compat_urlparse.urljoin(url, '/playlist?list=%s' % playlist_id), 'YoutubePlaylist')
1653
1654 channel_page = self._download_webpage(url, channel_id, 'Downloading page #1')
1655 autogenerated = re.search(r'''(?x)
1656 class="[^"]*?(?:
1657 channel-header-autogenerated-label|
1658 yt-channel-title-autogenerated
1659 )[^"]*"''', channel_page) is not None
1660
1661 if autogenerated:
1662 # The videos are contained in a single page
1663 # the ajax pages can't be used, they are empty
1664 entries = [
1665 self.url_result(
1666 video_id, 'Youtube', video_id=video_id,
1667 video_title=video_title)
1668 for video_id, video_title in self.extract_videos_from_page(channel_page)]
1669 return self.playlist_result(entries, channel_id)
1670
1671 def _entries():
1672 more_widget_html = content_html = channel_page
1673 for pagenum in itertools.count(1):
1674
1675 for video_id, video_title in self.extract_videos_from_page(content_html):
1676 yield self.url_result(
1677 video_id, 'Youtube', video_id=video_id,
1678 video_title=video_title)
1679
1680 mobj = re.search(
1681 r'data-uix-load-more-href="/?(?P<more>[^"]+)"',
1682 more_widget_html)
1683 if not mobj:
1684 break
1685
1686 more = self._download_json(
1687 'https://youtube.com/%s' % mobj.group('more'), channel_id,
1688 'Downloading page #%s' % (pagenum + 1),
1689 transform_source=uppercase_escape)
1690 content_html = more['content_html']
1691 more_widget_html = more['load_more_widget_html']
1692
1693 return self.playlist_result(_entries(), channel_id)
1694
1695
1696 class YoutubeUserIE(YoutubeChannelIE):
1697 IE_DESC = 'YouTube.com user videos (URL or "ytuser" keyword)'
1698 _VALID_URL = r'(?:(?:(?:https?://)?(?:\w+\.)?youtube\.com/(?:user/)?(?!(?:attribution_link|watch|results)(?:$|[^a-z_A-Z0-9-])))|ytuser:)(?!feed/)(?P<id>[A-Za-z0-9_-]+)'
1699 _TEMPLATE_URL = 'https://www.youtube.com/user/%s/videos'
1700 IE_NAME = 'youtube:user'
1701
1702 _TESTS = [{
1703 'url': 'https://www.youtube.com/user/TheLinuxFoundation',
1704 'playlist_mincount': 320,
1705 'info_dict': {
1706 'title': 'TheLinuxFoundation',
1707 }
1708 }, {
1709 'url': 'ytuser:phihag',
1710 'only_matching': True,
1711 }]
1712
1713 @classmethod
1714 def suitable(cls, url):
1715 # Don't return True if the url can be extracted with other youtube
1716 # extractor, the regex would is too permissive and it would match.
1717 other_ies = iter(klass for (name, klass) in globals().items() if name.endswith('IE') and klass is not cls)
1718 if any(ie.suitable(url) for ie in other_ies):
1719 return False
1720 else:
1721 return super(YoutubeUserIE, cls).suitable(url)
1722
1723
1724 class YoutubeSearchIE(SearchInfoExtractor, YoutubePlaylistIE):
1725 IE_DESC = 'YouTube.com searches'
1726 # there doesn't appear to be a real limit, for example if you search for
1727 # 'python' you get more than 8.000.000 results
1728 _MAX_RESULTS = float('inf')
1729 IE_NAME = 'youtube:search'
1730 _SEARCH_KEY = 'ytsearch'
1731 _EXTRA_QUERY_ARGS = {}
1732 _TESTS = []
1733
1734 def _get_n_results(self, query, n):
1735 """Get a specified number of results for a query"""
1736
1737 videos = []
1738 limit = n
1739
1740 for pagenum in itertools.count(1):
1741 url_query = {
1742 'search_query': query.encode('utf-8'),
1743 'page': pagenum,
1744 'spf': 'navigate',
1745 }
1746 url_query.update(self._EXTRA_QUERY_ARGS)
1747 result_url = 'https://www.youtube.com/results?' + compat_urllib_parse.urlencode(url_query)
1748 data = self._download_json(
1749 result_url, video_id='query "%s"' % query,
1750 note='Downloading page %s' % pagenum,
1751 errnote='Unable to download API page')
1752 html_content = data[1]['body']['content']
1753
1754 if 'class="search-message' in html_content:
1755 raise ExtractorError(
1756 '[youtube] No video results', expected=True)
1757
1758 new_videos = self._ids_to_results(orderedSet(re.findall(
1759 r'href="/watch\?v=(.{11})', html_content)))
1760 videos += new_videos
1761 if not new_videos or len(videos) > limit:
1762 break
1763
1764 if len(videos) > n:
1765 videos = videos[:n]
1766 return self.playlist_result(videos, query)
1767
1768
1769 class YoutubeSearchDateIE(YoutubeSearchIE):
1770 IE_NAME = YoutubeSearchIE.IE_NAME + ':date'
1771 _SEARCH_KEY = 'ytsearchdate'
1772 IE_DESC = 'YouTube.com searches, newest videos first'
1773 _EXTRA_QUERY_ARGS = {'search_sort': 'video_date_uploaded'}
1774
1775
1776 class YoutubeSearchURLIE(InfoExtractor):
1777 IE_DESC = 'YouTube.com search URLs'
1778 IE_NAME = 'youtube:search_url'
1779 _VALID_URL = r'https?://(?:www\.)?youtube\.com/results\?(.*?&)?search_query=(?P<query>[^&]+)(?:[&]|$)'
1780 _TESTS = [{
1781 'url': 'https://www.youtube.com/results?baz=bar&search_query=youtube-dl+test+video&filters=video&lclk=video',
1782 'playlist_mincount': 5,
1783 'info_dict': {
1784 'title': 'youtube-dl test video',
1785 }
1786 }]
1787
1788 def _real_extract(self, url):
1789 mobj = re.match(self._VALID_URL, url)
1790 query = compat_urllib_parse_unquote_plus(mobj.group('query'))
1791
1792 webpage = self._download_webpage(url, query)
1793 result_code = self._search_regex(
1794 r'(?s)<ol[^>]+class="item-section"(.*?)</ol>', webpage, 'result HTML')
1795
1796 part_codes = re.findall(
1797 r'(?s)<h3[^>]+class="[^"]*yt-lockup-title[^"]*"[^>]*>(.*?)</h3>', result_code)
1798 entries = []
1799 for part_code in part_codes:
1800 part_title = self._html_search_regex(
1801 [r'(?s)title="([^"]+)"', r'>([^<]+)</a>'], part_code, 'item title', fatal=False)
1802 part_url_snippet = self._html_search_regex(
1803 r'(?s)href="([^"]+)"', part_code, 'item URL')
1804 part_url = compat_urlparse.urljoin(
1805 'https://www.youtube.com/', part_url_snippet)
1806 entries.append({
1807 '_type': 'url',
1808 'url': part_url,
1809 'title': part_title,
1810 })
1811
1812 return {
1813 '_type': 'playlist',
1814 'entries': entries,
1815 'title': query,
1816 }
1817
1818
1819 class YoutubeShowIE(InfoExtractor):
1820 IE_DESC = 'YouTube.com (multi-season) shows'
1821 _VALID_URL = r'https?://www\.youtube\.com/show/(?P<id>[^?#]*)'
1822 IE_NAME = 'youtube:show'
1823 _TESTS = [{
1824 'url': 'http://www.youtube.com/show/airdisasters',
1825 'playlist_mincount': 3,
1826 'info_dict': {
1827 'id': 'airdisasters',
1828 'title': 'Air Disasters',
1829 }
1830 }]
1831
1832 def _real_extract(self, url):
1833 mobj = re.match(self._VALID_URL, url)
1834 playlist_id = mobj.group('id')
1835 webpage = self._download_webpage(
1836 url, playlist_id, 'Downloading show webpage')
1837 # There's one playlist for each season of the show
1838 m_seasons = list(re.finditer(r'href="(/playlist\?list=.*?)"', webpage))
1839 self.to_screen('%s: Found %s seasons' % (playlist_id, len(m_seasons)))
1840 entries = [
1841 self.url_result(
1842 'https://www.youtube.com' + season.group(1), 'YoutubePlaylist')
1843 for season in m_seasons
1844 ]
1845 title = self._og_search_title(webpage, fatal=False)
1846
1847 return {
1848 '_type': 'playlist',
1849 'id': playlist_id,
1850 'title': title,
1851 'entries': entries,
1852 }
1853
1854
1855 class YoutubeFeedsInfoExtractor(YoutubeBaseInfoExtractor):
1856 """
1857 Base class for feed extractors
1858 Subclasses must define the _FEED_NAME and _PLAYLIST_TITLE properties.
1859 """
1860 _LOGIN_REQUIRED = True
1861
1862 @property
1863 def IE_NAME(self):
1864 return 'youtube:%s' % self._FEED_NAME
1865
1866 def _real_initialize(self):
1867 self._login()
1868
1869 def _real_extract(self, url):
1870 page = self._download_webpage(
1871 'https://www.youtube.com/feed/%s' % self._FEED_NAME, self._PLAYLIST_TITLE)
1872
1873 # The extraction process is the same as for playlists, but the regex
1874 # for the video ids doesn't contain an index
1875 ids = []
1876 more_widget_html = content_html = page
1877 for page_num in itertools.count(1):
1878 matches = re.findall(r'href="\s*/watch\?v=([0-9A-Za-z_-]{11})', content_html)
1879
1880 # 'recommended' feed has infinite 'load more' and each new portion spins
1881 # the same videos in (sometimes) slightly different order, so we'll check
1882 # for unicity and break when portion has no new videos
1883 new_ids = filter(lambda video_id: video_id not in ids, orderedSet(matches))
1884 if not new_ids:
1885 break
1886
1887 ids.extend(new_ids)
1888
1889 mobj = re.search(r'data-uix-load-more-href="/?(?P<more>[^"]+)"', more_widget_html)
1890 if not mobj:
1891 break
1892
1893 more = self._download_json(
1894 'https://youtube.com/%s' % mobj.group('more'), self._PLAYLIST_TITLE,
1895 'Downloading page #%s' % page_num,
1896 transform_source=uppercase_escape)
1897 content_html = more['content_html']
1898 more_widget_html = more['load_more_widget_html']
1899
1900 return self.playlist_result(
1901 self._ids_to_results(ids), playlist_title=self._PLAYLIST_TITLE)
1902
1903
1904 class YoutubeWatchLaterIE(YoutubePlaylistIE):
1905 IE_NAME = 'youtube:watchlater'
1906 IE_DESC = 'Youtube watch later list, ":ytwatchlater" for short (requires authentication)'
1907 _VALID_URL = r'https?://www\.youtube\.com/(?:feed/watch_later|playlist\?list=WL)|:ytwatchlater'
1908
1909 _TESTS = [] # override PlaylistIE tests
1910
1911 def _real_extract(self, url):
1912 return self._extract_playlist('WL')
1913
1914
1915 class YoutubeFavouritesIE(YoutubeBaseInfoExtractor):
1916 IE_NAME = 'youtube:favorites'
1917 IE_DESC = 'YouTube.com favourite videos, ":ytfav" for short (requires authentication)'
1918 _VALID_URL = r'https?://www\.youtube\.com/my_favorites|:ytfav(?:ou?rites)?'
1919 _LOGIN_REQUIRED = True
1920
1921 def _real_extract(self, url):
1922 webpage = self._download_webpage('https://www.youtube.com/my_favorites', 'Youtube Favourites videos')
1923 playlist_id = self._search_regex(r'list=(.+?)["&]', webpage, 'favourites playlist id')
1924 return self.url_result(playlist_id, 'YoutubePlaylist')
1925
1926
1927 class YoutubeRecommendedIE(YoutubeFeedsInfoExtractor):
1928 IE_DESC = 'YouTube.com recommended videos, ":ytrec" for short (requires authentication)'
1929 _VALID_URL = r'https?://www\.youtube\.com/feed/recommended|:ytrec(?:ommended)?'
1930 _FEED_NAME = 'recommended'
1931 _PLAYLIST_TITLE = 'Youtube Recommended videos'
1932
1933
1934 class YoutubeSubscriptionsIE(YoutubeFeedsInfoExtractor):
1935 IE_DESC = 'YouTube.com subscriptions feed, "ytsubs" keyword (requires authentication)'
1936 _VALID_URL = r'https?://www\.youtube\.com/feed/subscriptions|:ytsubs(?:criptions)?'
1937 _FEED_NAME = 'subscriptions'
1938 _PLAYLIST_TITLE = 'Youtube Subscriptions'
1939
1940
1941 class YoutubeHistoryIE(YoutubeFeedsInfoExtractor):
1942 IE_DESC = 'Youtube watch history, ":ythistory" for short (requires authentication)'
1943 _VALID_URL = 'https?://www\.youtube\.com/feed/history|:ythistory'
1944 _FEED_NAME = 'history'
1945 _PLAYLIST_TITLE = 'Youtube History'
1946
1947
1948 class YoutubeTruncatedURLIE(InfoExtractor):
1949 IE_NAME = 'youtube:truncated_url'
1950 IE_DESC = False # Do not list
1951 _VALID_URL = r'''(?x)
1952 (?:https?://)?
1953 (?:\w+\.)?[yY][oO][uU][tT][uU][bB][eE](?:-nocookie)?\.com/
1954 (?:watch\?(?:
1955 feature=[a-z_]+|
1956 annotation_id=annotation_[^&]+|
1957 x-yt-cl=[0-9]+|
1958 hl=[^&]*|
1959 )?
1960 |
1961 attribution_link\?a=[^&]+
1962 )
1963 $
1964 '''
1965
1966 _TESTS = [{
1967 'url': 'http://www.youtube.com/watch?annotation_id=annotation_3951667041',
1968 'only_matching': True,
1969 }, {
1970 'url': 'http://www.youtube.com/watch?',
1971 'only_matching': True,
1972 }, {
1973 'url': 'https://www.youtube.com/watch?x-yt-cl=84503534',
1974 'only_matching': True,
1975 }, {
1976 'url': 'https://www.youtube.com/watch?feature=foo',
1977 'only_matching': True,
1978 }, {
1979 'url': 'https://www.youtube.com/watch?hl=en-GB',
1980 'only_matching': True,
1981 }]
1982
1983 def _real_extract(self, url):
1984 raise ExtractorError(
1985 'Did you forget to quote the URL? Remember that & is a meta '
1986 'character in most shells, so you want to put the URL in quotes, '
1987 'like youtube-dl '
1988 '"http://www.youtube.com/watch?feature=foo&v=BaW_jenozKc" '
1989 ' or simply youtube-dl BaW_jenozKc .',
1990 expected=True)
1991
1992
1993 class YoutubeTruncatedIDIE(InfoExtractor):
1994 IE_NAME = 'youtube:truncated_id'
1995 IE_DESC = False # Do not list
1996 _VALID_URL = r'https?://(?:www\.)?youtube\.com/watch\?v=(?P<id>[0-9A-Za-z_-]{1,10})$'
1997
1998 _TESTS = [{
1999 'url': 'https://www.youtube.com/watch?v=N_708QY7Ob',
2000 'only_matching': True,
2001 }]
2002
2003 def _real_extract(self, url):
2004 video_id = self._match_id(url)
2005 raise ExtractorError(
2006 'Incomplete YouTube ID %s. URL %s looks truncated.' % (video_id, url),
2007 expected=True)
fork of https://github.com/yt-dlp/yt-dlp/