]> jfr.im git - yt-dlp.git/blob - youtube_dl/extractor/youtube.py
projects / yt-dlp.git / blob
? search:


15e327ec8b5843fcd89e7943e4b6991dbcb4ed88
[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 str_to_int,
37 unescapeHTML,
38 unified_strdate,
39 uppercase_escape,
40 ISO3166Utils,
41 )
42
43
44 class YoutubeBaseInfoExtractor(InfoExtractor):
45 """Provide base functions for Youtube extractors"""
46 _LOGIN_URL = 'https://accounts.google.com/ServiceLogin'
47 _TWOFACTOR_URL = 'https://accounts.google.com/SecondFactor'
48 _NETRC_MACHINE = 'youtube'
49 # If True it will raise an error if no login info is provided
50 _LOGIN_REQUIRED = False
51
52 def _set_language(self):
53 self._set_cookie(
54 '.youtube.com', 'PREF', 'f1=50000000&hl=en',
55 # YouTube sets the expire time to about two months
56 expire_time=time.time() + 2 * 30 * 24 * 3600)
57
58 def _ids_to_results(self, ids):
59 return [
60 self.url_result(vid_id, 'Youtube', video_id=vid_id)
61 for vid_id in ids]
62
63 def _login(self):
64 """
65 Attempt to log in to YouTube.
66 True is returned if successful or skipped.
67 False is returned if login failed.
68
69 If _LOGIN_REQUIRED is set and no authentication was provided, an error is raised.
70 """
71 (username, password) = self._get_login_info()
72 # No authentication to be performed
73 if username is None:
74 if self._LOGIN_REQUIRED:
75 raise ExtractorError('No login info available, needed for using %s.' % self.IE_NAME, expected=True)
76 return True
77
78 login_page = self._download_webpage(
79 self._LOGIN_URL, None,
80 note='Downloading login page',
81 errnote='unable to fetch login page', fatal=False)
82 if login_page is False:
83 return
84
85 galx = self._search_regex(r'(?s)<input.+?name="GALX".+?value="(.+?)"',
86 login_page, 'Login GALX parameter')
87
88 # Log in
89 login_form_strs = {
90 'continue': 'https://www.youtube.com/signin?action_handle_signin=true&feature=sign_in_button&hl=en_US&nomobiletemp=1',
91 'Email': username,
92 'GALX': galx,
93 'Passwd': password,
94
95 'PersistentCookie': 'yes',
96 '_utf8': '霱',
97 'bgresponse': 'js_disabled',
98 'checkConnection': '',
99 'checkedDomains': 'youtube',
100 'dnConn': '',
101 'pstMsg': '0',
102 'rmShown': '1',
103 'secTok': '',
104 'signIn': 'Sign in',
105 'timeStmp': '',
106 'service': 'youtube',
107 'uilel': '3',
108 'hl': 'en_US',
109 }
110
111 # Convert to UTF-8 *before* urlencode because Python 2.x's urlencode
112 # chokes on unicode
113 login_form = dict((k.encode('utf-8'), v.encode('utf-8')) for k, v in login_form_strs.items())
114 login_data = compat_urllib_parse.urlencode(login_form).encode('ascii')
115
116 req = compat_urllib_request.Request(self._LOGIN_URL, login_data)
117 login_results = self._download_webpage(
118 req, None,
119 note='Logging in', errnote='unable to log in', fatal=False)
120 if login_results is False:
121 return False
122
123 if re.search(r'id="errormsg_0_Passwd"', login_results) is not None:
124 raise ExtractorError('Please use your account password and a two-factor code instead of an application-specific password.', expected=True)
125
126 # Two-Factor
127 # TODO add SMS and phone call support - these require making a request and then prompting the user
128
129 if re.search(r'(?i)<form[^>]* id="gaia_secondfactorform"', login_results) is not None:
130 tfa_code = self._get_tfa_info()
131
132 if tfa_code is None:
133 self._downloader.report_warning('Two-factor authentication required. Provide it with --twofactor <code>')
134 self._downloader.report_warning('(Note that only TOTP (Google Authenticator App) codes work at this time.)')
135 return False
136
137 # Unlike the first login form, secTok and timeStmp are both required for the TFA form
138
139 match = re.search(r'id="secTok"\n\s+value=\'(.+)\'/>', login_results, re.M | re.U)
140 if match is None:
141 self._downloader.report_warning('Failed to get secTok - did the page structure change?')
142 secTok = match.group(1)
143 match = re.search(r'id="timeStmp"\n\s+value=\'(.+)\'/>', login_results, re.M | re.U)
144 if match is None:
145 self._downloader.report_warning('Failed to get timeStmp - did the page structure change?')
146 timeStmp = match.group(1)
147
148 tfa_form_strs = {
149 'continue': 'https://www.youtube.com/signin?action_handle_signin=true&feature=sign_in_button&hl=en_US&nomobiletemp=1',
150 'smsToken': '',
151 'smsUserPin': tfa_code,
152 'smsVerifyPin': 'Verify',
153
154 'PersistentCookie': 'yes',
155 'checkConnection': '',
156 'checkedDomains': 'youtube',
157 'pstMsg': '1',
158 'secTok': secTok,
159 'timeStmp': timeStmp,
160 'service': 'youtube',
161 'hl': 'en_US',
162 }
163 tfa_form = dict((k.encode('utf-8'), v.encode('utf-8')) for k, v in tfa_form_strs.items())
164 tfa_data = compat_urllib_parse.urlencode(tfa_form).encode('ascii')
165
166 tfa_req = compat_urllib_request.Request(self._TWOFACTOR_URL, tfa_data)
167 tfa_results = self._download_webpage(
168 tfa_req, None,
169 note='Submitting TFA code', errnote='unable to submit tfa', fatal=False)
170
171 if tfa_results is False:
172 return False
173
174 if re.search(r'(?i)<form[^>]* id="gaia_secondfactorform"', tfa_results) is not None:
175 self._downloader.report_warning('Two-factor code expired. Please try again, or use a one-use backup code instead.')
176 return False
177 if re.search(r'(?i)<form[^>]* id="gaia_loginform"', tfa_results) is not None:
178 self._downloader.report_warning('unable to log in - did the page structure change?')
179 return False
180 if re.search(r'smsauth-interstitial-reviewsettings', tfa_results) is not None:
181 self._downloader.report_warning('Your Google account has a security notice. Please log in on your web browser, resolve the notice, and try again.')
182 return False
183
184 if re.search(r'(?i)<form[^>]* id="gaia_loginform"', login_results) is not None:
185 self._downloader.report_warning('unable to log in: bad username or password')
186 return False
187 return True
188
189 def _real_initialize(self):
190 if self._downloader is None:
191 return
192 self._set_language()
193 if not self._login():
194 return
195
196
197 class YoutubeIE(YoutubeBaseInfoExtractor):
198 IE_DESC = 'YouTube.com'
199 _VALID_URL = r"""(?x)^
200 (
201 (?:https?://|//) # http(s):// or protocol-independent URL
202 (?:(?:(?:(?:\w+\.)?[yY][oO][uU][tT][uU][bB][eE](?:-nocookie)?\.com/|
203 (?:www\.)?deturl\.com/www\.youtube\.com/|
204 (?:www\.)?pwnyoutube\.com/|
205 (?:www\.)?yourepeat\.com/|
206 tube\.majestyc\.net/|
207 youtube\.googleapis\.com/) # the various hostnames, with wildcard subdomains
208 (?:.*?\#/)? # handle anchor (#/) redirect urls
209 (?: # the various things that can precede the ID:
210 (?:(?:v|embed|e)/(?!videoseries)) # v/ or embed/ or e/
211 |(?: # or the v= param in all its forms
212 (?:(?:watch|movie)(?:_popup)?(?:\.php)?/?)? # preceding watch(_popup|.php) or nothing (like /?v=xxxx)
213 (?:\?|\#!?) # the params delimiter ? or # or #!
214 (?:.*?&)? # any other preceding param (like /?s=tuff&v=xxxx)
215 v=
216 )
217 ))
218 |youtu\.be/ # just youtu.be/xxxx
219 |(?:www\.)?cleanvideosearch\.com/media/action/yt/watch\?videoId=
220 )
221 )? # all until now is optional -> you can pass the naked ID
222 ([0-9A-Za-z_-]{11}) # here is it! the YouTube video ID
223 (?!.*?&list=) # combined list/video URLs are handled by the playlist IE
224 (?(1).+)? # if we found the ID, everything can follow
225 $"""
226 _NEXT_URL_RE = r'[\?&]next_url=([^&]+)'
227 _formats = {
228 '5': {'ext': 'flv', 'width': 400, 'height': 240},
229 '6': {'ext': 'flv', 'width': 450, 'height': 270},
230 '13': {'ext': '3gp'},
231 '17': {'ext': '3gp', 'width': 176, 'height': 144},
232 '18': {'ext': 'mp4', 'width': 640, 'height': 360},
233 '22': {'ext': 'mp4', 'width': 1280, 'height': 720},
234 '34': {'ext': 'flv', 'width': 640, 'height': 360},
235 '35': {'ext': 'flv', 'width': 854, 'height': 480},
236 '36': {'ext': '3gp', 'width': 320, 'height': 240},
237 '37': {'ext': 'mp4', 'width': 1920, 'height': 1080},
238 '38': {'ext': 'mp4', 'width': 4096, 'height': 3072},
239 '43': {'ext': 'webm', 'width': 640, 'height': 360},
240 '44': {'ext': 'webm', 'width': 854, 'height': 480},
241 '45': {'ext': 'webm', 'width': 1280, 'height': 720},
242 '46': {'ext': 'webm', 'width': 1920, 'height': 1080},
243 '59': {'ext': 'mp4', 'width': 854, 'height': 480},
244 '78': {'ext': 'mp4', 'width': 854, 'height': 480},
245
246
247 # 3d videos
248 '82': {'ext': 'mp4', 'height': 360, 'format_note': '3D', 'preference': -20},
249 '83': {'ext': 'mp4', 'height': 480, 'format_note': '3D', 'preference': -20},
250 '84': {'ext': 'mp4', 'height': 720, 'format_note': '3D', 'preference': -20},
251 '85': {'ext': 'mp4', 'height': 1080, 'format_note': '3D', 'preference': -20},
252 '100': {'ext': 'webm', 'height': 360, 'format_note': '3D', 'preference': -20},
253 '101': {'ext': 'webm', 'height': 480, 'format_note': '3D', 'preference': -20},
254 '102': {'ext': 'webm', 'height': 720, 'format_note': '3D', 'preference': -20},
255
256 # Apple HTTP Live Streaming
257 '92': {'ext': 'mp4', 'height': 240, 'format_note': 'HLS', 'preference': -10},
258 '93': {'ext': 'mp4', 'height': 360, 'format_note': 'HLS', 'preference': -10},
259 '94': {'ext': 'mp4', 'height': 480, 'format_note': 'HLS', 'preference': -10},
260 '95': {'ext': 'mp4', 'height': 720, 'format_note': 'HLS', 'preference': -10},
261 '96': {'ext': 'mp4', 'height': 1080, 'format_note': 'HLS', 'preference': -10},
262 '132': {'ext': 'mp4', 'height': 240, 'format_note': 'HLS', 'preference': -10},
263 '151': {'ext': 'mp4', 'height': 72, 'format_note': 'HLS', 'preference': -10},
264
265 # DASH mp4 video
266 '133': {'ext': 'mp4', 'height': 240, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
267 '134': {'ext': 'mp4', 'height': 360, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
268 '135': {'ext': 'mp4', 'height': 480, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
269 '136': {'ext': 'mp4', 'height': 720, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
270 '137': {'ext': 'mp4', 'height': 1080, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
271 '138': {'ext': 'mp4', 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40}, # Height can vary (https://github.com/rg3/youtube-dl/issues/4559)
272 '160': {'ext': 'mp4', 'height': 144, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
273 '264': {'ext': 'mp4', 'height': 1440, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
274 '298': {'ext': 'mp4', 'height': 720, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40, 'fps': 60, 'vcodec': 'h264'},
275 '299': {'ext': 'mp4', 'height': 1080, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40, 'fps': 60, 'vcodec': 'h264'},
276 '266': {'ext': 'mp4', 'height': 2160, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40, 'vcodec': 'h264'},
277
278 # Dash mp4 audio
279 '139': {'ext': 'm4a', 'format_note': 'DASH audio', 'acodec': 'aac', 'vcodec': 'none', 'abr': 48, 'preference': -50, 'container': 'm4a_dash'},
280 '140': {'ext': 'm4a', 'format_note': 'DASH audio', 'acodec': 'aac', 'vcodec': 'none', 'abr': 128, 'preference': -50, 'container': 'm4a_dash'},
281 '141': {'ext': 'm4a', 'format_note': 'DASH audio', 'acodec': 'aac', 'vcodec': 'none', 'abr': 256, 'preference': -50, 'container': 'm4a_dash'},
282
283 # Dash webm
284 '167': {'ext': 'webm', 'height': 360, 'width': 640, 'format_note': 'DASH video', 'acodec': 'none', 'container': 'webm', 'vcodec': 'vp8', 'preference': -40},
285 '168': {'ext': 'webm', 'height': 480, 'width': 854, 'format_note': 'DASH video', 'acodec': 'none', 'container': 'webm', 'vcodec': 'vp8', 'preference': -40},
286 '169': {'ext': 'webm', 'height': 720, 'width': 1280, 'format_note': 'DASH video', 'acodec': 'none', 'container': 'webm', 'vcodec': 'vp8', 'preference': -40},
287 '170': {'ext': 'webm', 'height': 1080, 'width': 1920, 'format_note': 'DASH video', 'acodec': 'none', 'container': 'webm', 'vcodec': 'vp8', 'preference': -40},
288 '218': {'ext': 'webm', 'height': 480, 'width': 854, 'format_note': 'DASH video', 'acodec': 'none', 'container': 'webm', 'vcodec': 'vp8', 'preference': -40},
289 '219': {'ext': 'webm', 'height': 480, 'width': 854, 'format_note': 'DASH video', 'acodec': 'none', 'container': 'webm', 'vcodec': 'vp8', 'preference': -40},
290 '278': {'ext': 'webm', 'height': 144, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40, 'container': 'webm', 'vcodec': 'vp9'},
291 '242': {'ext': 'webm', 'height': 240, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
292 '243': {'ext': 'webm', 'height': 360, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
293 '244': {'ext': 'webm', 'height': 480, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
294 '245': {'ext': 'webm', 'height': 480, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
295 '246': {'ext': 'webm', 'height': 480, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
296 '247': {'ext': 'webm', 'height': 720, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
297 '248': {'ext': 'webm', 'height': 1080, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
298 '271': {'ext': 'webm', 'height': 1440, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
299 '272': {'ext': 'webm', 'height': 2160, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
300 '302': {'ext': 'webm', 'height': 720, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40, 'fps': 60, 'vcodec': 'vp9'},
301 '303': {'ext': 'webm', 'height': 1080, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40, 'fps': 60, 'vcodec': 'vp9'},
302 '308': {'ext': 'webm', 'height': 1440, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40, 'fps': 60, 'vcodec': 'vp9'},
303 '313': {'ext': 'webm', 'height': 2160, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40, 'vcodec': 'vp9'},
304 '315': {'ext': 'webm', 'height': 2160, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40, 'fps': 60, 'vcodec': 'vp9'},
305
306 # Dash webm audio
307 '171': {'ext': 'webm', 'vcodec': 'none', 'format_note': 'DASH audio', 'abr': 128, 'preference': -50},
308 '172': {'ext': 'webm', 'vcodec': 'none', 'format_note': 'DASH audio', 'abr': 256, 'preference': -50},
309
310 # Dash webm audio with opus inside
311 '249': {'ext': 'webm', 'vcodec': 'none', 'format_note': 'DASH audio', 'acodec': 'opus', 'abr': 50, 'preference': -50},
312 '250': {'ext': 'webm', 'vcodec': 'none', 'format_note': 'DASH audio', 'acodec': 'opus', 'abr': 70, 'preference': -50},
313 '251': {'ext': 'webm', 'vcodec': 'none', 'format_note': 'DASH audio', 'acodec': 'opus', 'abr': 160, 'preference': -50},
314
315 # RTMP (unnamed)
316 '_rtmp': {'protocol': 'rtmp'},
317 }
318
319 IE_NAME = 'youtube'
320 _TESTS = [
321 {
322 'url': 'http://www.youtube.com/watch?v=BaW_jenozKcj&t=1s&end=9',
323 'info_dict': {
324 'id': 'BaW_jenozKc',
325 'ext': 'mp4',
326 'title': 'youtube-dl test video "\'/\\ä↭𝕐',
327 'uploader': 'Philipp Hagemeister',
328 'uploader_id': 'phihag',
329 'upload_date': '20121002',
330 '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 .',
331 'categories': ['Science & Technology'],
332 'like_count': int,
333 'dislike_count': int,
334 'start_time': 1,
335 'end_time': 9,
336 }
337 },
338 {
339 'url': 'http://www.youtube.com/watch?v=UxxajLWwzqY',
340 'note': 'Test generic use_cipher_signature video (#897)',
341 'info_dict': {
342 'id': 'UxxajLWwzqY',
343 'ext': 'mp4',
344 'upload_date': '20120506',
345 'title': 'Icona Pop - I Love It (feat. Charli XCX) [OFFICIAL VIDEO]',
346 'description': 'md5:fea86fda2d5a5784273df5c7cc994d9f',
347 'uploader': 'Icona Pop',
348 'uploader_id': 'IconaPop',
349 }
350 },
351 {
352 'url': 'https://www.youtube.com/watch?v=07FYdnEawAQ',
353 'note': 'Test VEVO video with age protection (#956)',
354 'info_dict': {
355 'id': '07FYdnEawAQ',
356 'ext': 'mp4',
357 'upload_date': '20130703',
358 'title': 'Justin Timberlake - Tunnel Vision (Explicit)',
359 'description': 'md5:64249768eec3bc4276236606ea996373',
360 'uploader': 'justintimberlakeVEVO',
361 'uploader_id': 'justintimberlakeVEVO',
362 }
363 },
364 {
365 'url': '//www.YouTube.com/watch?v=yZIXLfi8CZQ',
366 'note': 'Embed-only video (#1746)',
367 'info_dict': {
368 'id': 'yZIXLfi8CZQ',
369 'ext': 'mp4',
370 'upload_date': '20120608',
371 'title': 'Principal Sexually Assaults A Teacher - Episode 117 - 8th June 2012',
372 'description': 'md5:09b78bd971f1e3e289601dfba15ca4f7',
373 'uploader': 'SET India',
374 'uploader_id': 'setindia'
375 }
376 },
377 {
378 'url': 'http://www.youtube.com/watch?v=a9LDPn-MO4I',
379 'note': '256k DASH audio (format 141) via DASH manifest',
380 'info_dict': {
381 'id': 'a9LDPn-MO4I',
382 'ext': 'm4a',
383 'upload_date': '20121002',
384 'uploader_id': '8KVIDEO',
385 'description': '',
386 'uploader': '8KVIDEO',
387 'title': 'UHDTV TEST 8K VIDEO.mp4'
388 },
389 'params': {
390 'youtube_include_dash_manifest': True,
391 'format': '141',
392 },
393 },
394 # DASH manifest with encrypted signature
395 {
396 'url': 'https://www.youtube.com/watch?v=IB3lcPjvWLA',
397 'info_dict': {
398 'id': 'IB3lcPjvWLA',
399 'ext': 'm4a',
400 'title': 'Afrojack, Spree Wilson - The Spark ft. Spree Wilson',
401 'description': 'md5:12e7067fa6735a77bdcbb58cb1187d2d',
402 'uploader': 'AfrojackVEVO',
403 'uploader_id': 'AfrojackVEVO',
404 'upload_date': '20131011',
405 },
406 'params': {
407 'youtube_include_dash_manifest': True,
408 'format': '141',
409 },
410 },
411 # JS player signature function name containing $
412 {
413 'url': 'https://www.youtube.com/watch?v=nfWlot6h_JM',
414 'info_dict': {
415 'id': 'nfWlot6h_JM',
416 'ext': 'm4a',
417 'title': 'Taylor Swift - Shake It Off',
418 'description': 'md5:2acfda1b285bdd478ccec22f9918199d',
419 'uploader': 'TaylorSwiftVEVO',
420 'uploader_id': 'TaylorSwiftVEVO',
421 'upload_date': '20140818',
422 },
423 'params': {
424 'youtube_include_dash_manifest': True,
425 'format': '141',
426 },
427 },
428 # Controversy video
429 {
430 'url': 'https://www.youtube.com/watch?v=T4XJQO3qol8',
431 'info_dict': {
432 'id': 'T4XJQO3qol8',
433 'ext': 'mp4',
434 'upload_date': '20100909',
435 'uploader': 'The Amazing Atheist',
436 'uploader_id': 'TheAmazingAtheist',
437 'title': 'Burning Everyone\'s Koran',
438 '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',
439 }
440 },
441 # Normal age-gate video (No vevo, embed allowed)
442 {
443 'url': 'http://youtube.com/watch?v=HtVdAasjOgU',
444 'info_dict': {
445 'id': 'HtVdAasjOgU',
446 'ext': 'mp4',
447 'title': 'The Witcher 3: Wild Hunt - The Sword Of Destiny Trailer',
448 'description': 're:(?s).{100,}About the Game\n.*?The Witcher 3: Wild Hunt.{100,}',
449 'uploader': 'The Witcher',
450 'uploader_id': 'WitcherGame',
451 'upload_date': '20140605',
452 },
453 },
454 # Age-gate video with encrypted signature
455 {
456 'url': 'http://www.youtube.com/watch?v=6kLq3WMV1nU',
457 'info_dict': {
458 'id': '6kLq3WMV1nU',
459 'ext': 'mp4',
460 'title': 'Dedication To My Ex (Miss That) (Lyric Video)',
461 'description': 'md5:33765bb339e1b47e7e72b5490139bb41',
462 'uploader': 'LloydVEVO',
463 'uploader_id': 'LloydVEVO',
464 'upload_date': '20110629',
465 },
466 },
467 # video_info is None (https://github.com/rg3/youtube-dl/issues/4421)
468 {
469 'url': '__2ABJjxzNo',
470 'info_dict': {
471 'id': '__2ABJjxzNo',
472 'ext': 'mp4',
473 'upload_date': '20100430',
474 'uploader_id': 'deadmau5',
475 'description': 'md5:12c56784b8032162bb936a5f76d55360',
476 'uploader': 'deadmau5',
477 'title': 'Deadmau5 - Some Chords (HD)',
478 },
479 'expected_warnings': [
480 'DASH manifest missing',
481 ]
482 },
483 # Olympics (https://github.com/rg3/youtube-dl/issues/4431)
484 {
485 'url': 'lqQg6PlCWgI',
486 'info_dict': {
487 'id': 'lqQg6PlCWgI',
488 'ext': 'mp4',
489 'upload_date': '20120731',
490 'uploader_id': 'olympic',
491 'description': 'HO09 - Women - GER-AUS - Hockey - 31 July 2012 - London 2012 Olympic Games',
492 'uploader': 'Olympics',
493 'title': 'Hockey - Women - GER-AUS - London 2012 Olympic Games',
494 },
495 'params': {
496 'skip_download': 'requires avconv',
497 }
498 },
499 # Non-square pixels
500 {
501 'url': 'https://www.youtube.com/watch?v=_b-2C3KPAM0',
502 'info_dict': {
503 'id': '_b-2C3KPAM0',
504 'ext': 'mp4',
505 'stretched_ratio': 16 / 9.,
506 'upload_date': '20110310',
507 'uploader_id': 'AllenMeow',
508 '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',
509 'uploader': '孫艾倫',
510 'title': '[A-made] 變態妍字幕版 太妍 我就是這樣的人',
511 },
512 },
513 # url_encoded_fmt_stream_map is empty string
514 {
515 'url': 'qEJwOuvDf7I',
516 'info_dict': {
517 'id': 'qEJwOuvDf7I',
518 'ext': 'mp4',
519 'title': 'Обсуждение судебной практики по выборам 14 сентября 2014 года в Санкт-Петербурге',
520 'description': '',
521 'upload_date': '20150404',
522 'uploader_id': 'spbelect',
523 'uploader': 'Наблюдатели Петербурга',
524 },
525 'params': {
526 'skip_download': 'requires avconv',
527 }
528 },
529 # Extraction from multiple DASH manifests (https://github.com/rg3/youtube-dl/pull/6097)
530 {
531 'url': 'https://www.youtube.com/watch?v=FIl7x6_3R5Y',
532 'info_dict': {
533 'id': 'FIl7x6_3R5Y',
534 'ext': 'mp4',
535 'title': 'md5:7b81415841e02ecd4313668cde88737a',
536 'description': 'md5:116377fd2963b81ec4ce64b542173306',
537 'upload_date': '20150625',
538 'uploader_id': 'dorappi2000',
539 'uploader': 'dorappi2000',
540 'formats': 'mincount:33',
541 },
542 },
543 # DASH manifest with segment_list
544 {
545 'url': 'https://www.youtube.com/embed/CsmdDsKjzN8',
546 'md5': '8ce563a1d667b599d21064e982ab9e31',
547 'info_dict': {
548 'id': 'CsmdDsKjzN8',
549 'ext': 'mp4',
550 'upload_date': '20150501', # According to '<meta itemprop="datePublished"', but in other places it's 20150510
551 'uploader': 'Airtek',
552 'description': 'Retransmisión en directo de la XVIII media maratón de Zaragoza.',
553 'uploader_id': 'UCzTzUmjXxxacNnL8I3m4LnQ',
554 'title': 'Retransmisión XVIII Media maratón Zaragoza 2015',
555 },
556 'params': {
557 'youtube_include_dash_manifest': True,
558 'format': '135', # bestvideo
559 }
560 },
561 ]
562
563 def __init__(self, *args, **kwargs):
564 super(YoutubeIE, self).__init__(*args, **kwargs)
565 self._player_cache = {}
566
567 def report_video_info_webpage_download(self, video_id):
568 """Report attempt to download video info webpage."""
569 self.to_screen('%s: Downloading video info webpage' % video_id)
570
571 def report_information_extraction(self, video_id):
572 """Report attempt to extract video information."""
573 self.to_screen('%s: Extracting video information' % video_id)
574
575 def report_unavailable_format(self, video_id, format):
576 """Report extracted video URL."""
577 self.to_screen('%s: Format %s not available' % (video_id, format))
578
579 def report_rtmp_download(self):
580 """Indicate the download will use the RTMP protocol."""
581 self.to_screen('RTMP download detected')
582
583 def _signature_cache_id(self, example_sig):
584 """ Return a string representation of a signature """
585 return '.'.join(compat_str(len(part)) for part in example_sig.split('.'))
586
587 def _extract_signature_function(self, video_id, player_url, example_sig):
588 id_m = re.match(
589 r'.*?-(?P<id>[a-zA-Z0-9_-]+)(?:/watch_as3|/html5player)?\.(?P<ext>[a-z]+)$',
590 player_url)
591 if not id_m:
592 raise ExtractorError('Cannot identify player %r' % player_url)
593 player_type = id_m.group('ext')
594 player_id = id_m.group('id')
595
596 # Read from filesystem cache
597 func_id = '%s_%s_%s' % (
598 player_type, player_id, self._signature_cache_id(example_sig))
599 assert os.path.basename(func_id) == func_id
600
601 cache_spec = self._downloader.cache.load('youtube-sigfuncs', func_id)
602 if cache_spec is not None:
603 return lambda s: ''.join(s[i] for i in cache_spec)
604
605 download_note = (
606 'Downloading player %s' % player_url
607 if self._downloader.params.get('verbose') else
608 'Downloading %s player %s' % (player_type, player_id)
609 )
610 if player_type == 'js':
611 code = self._download_webpage(
612 player_url, video_id,
613 note=download_note,
614 errnote='Download of %s failed' % player_url)
615 res = self._parse_sig_js(code)
616 elif player_type == 'swf':
617 urlh = self._request_webpage(
618 player_url, video_id,
619 note=download_note,
620 errnote='Download of %s failed' % player_url)
621 code = urlh.read()
622 res = self._parse_sig_swf(code)
623 else:
624 assert False, 'Invalid player type %r' % player_type
625
626 test_string = ''.join(map(compat_chr, range(len(example_sig))))
627 cache_res = res(test_string)
628 cache_spec = [ord(c) for c in cache_res]
629
630 self._downloader.cache.store('youtube-sigfuncs', func_id, cache_spec)
631 return res
632
633 def _print_sig_code(self, func, example_sig):
634 def gen_sig_code(idxs):
635 def _genslice(start, end, step):
636 starts = '' if start == 0 else str(start)
637 ends = (':%d' % (end + step)) if end + step >= 0 else ':'
638 steps = '' if step == 1 else (':%d' % step)
639 return 's[%s%s%s]' % (starts, ends, steps)
640
641 step = None
642 # Quelch pyflakes warnings - start will be set when step is set
643 start = '(Never used)'
644 for i, prev in zip(idxs[1:], idxs[:-1]):
645 if step is not None:
646 if i - prev == step:
647 continue
648 yield _genslice(start, prev, step)
649 step = None
650 continue
651 if i - prev in [-1, 1]:
652 step = i - prev
653 start = prev
654 continue
655 else:
656 yield 's[%d]' % prev
657 if step is None:
658 yield 's[%d]' % i
659 else:
660 yield _genslice(start, i, step)
661
662 test_string = ''.join(map(compat_chr, range(len(example_sig))))
663 cache_res = func(test_string)
664 cache_spec = [ord(c) for c in cache_res]
665 expr_code = ' + '.join(gen_sig_code(cache_spec))
666 signature_id_tuple = '(%s)' % (
667 ', '.join(compat_str(len(p)) for p in example_sig.split('.')))
668 code = ('if tuple(len(p) for p in s.split(\'.\')) == %s:\n'
669 ' return %s\n') % (signature_id_tuple, expr_code)
670 self.to_screen('Extracted signature function:\n' + code)
671
672 def _parse_sig_js(self, jscode):
673 funcname = self._search_regex(
674 r'\.sig\|\|([a-zA-Z0-9$]+)\(', jscode,
675 'Initial JS player signature function name')
676
677 jsi = JSInterpreter(jscode)
678 initial_function = jsi.extract_function(funcname)
679 return lambda s: initial_function([s])
680
681 def _parse_sig_swf(self, file_contents):
682 swfi = SWFInterpreter(file_contents)
683 TARGET_CLASSNAME = 'SignatureDecipher'
684 searched_class = swfi.extract_class(TARGET_CLASSNAME)
685 initial_function = swfi.extract_function(searched_class, 'decipher')
686 return lambda s: initial_function([s])
687
688 def _decrypt_signature(self, s, video_id, player_url, age_gate=False):
689 """Turn the encrypted s field into a working signature"""
690
691 if player_url is None:
692 raise ExtractorError('Cannot decrypt signature without player_url')
693
694 if player_url.startswith('//'):
695 player_url = 'https:' + player_url
696 try:
697 player_id = (player_url, self._signature_cache_id(s))
698 if player_id not in self._player_cache:
699 func = self._extract_signature_function(
700 video_id, player_url, s
701 )
702 self._player_cache[player_id] = func
703 func = self._player_cache[player_id]
704 if self._downloader.params.get('youtube_print_sig_code'):
705 self._print_sig_code(func, s)
706 return func(s)
707 except Exception as e:
708 tb = traceback.format_exc()
709 raise ExtractorError(
710 'Signature extraction failed: ' + tb, cause=e)
711
712 def _get_subtitles(self, video_id, webpage):
713 try:
714 subs_doc = self._download_xml(
715 'https://video.google.com/timedtext?hl=en&type=list&v=%s' % video_id,
716 video_id, note=False)
717 except ExtractorError as err:
718 self._downloader.report_warning('unable to download video subtitles: %s' % compat_str(err))
719 return {}
720
721 sub_lang_list = {}
722 for track in subs_doc.findall('track'):
723 lang = track.attrib['lang_code']
724 if lang in sub_lang_list:
725 continue
726 sub_formats = []
727 for ext in ['sbv', 'vtt', 'srt']:
728 params = compat_urllib_parse.urlencode({
729 'lang': lang,
730 'v': video_id,
731 'fmt': ext,
732 'name': track.attrib['name'].encode('utf-8'),
733 })
734 sub_formats.append({
735 'url': 'https://www.youtube.com/api/timedtext?' + params,
736 'ext': ext,
737 })
738 sub_lang_list[lang] = sub_formats
739 if not sub_lang_list:
740 self._downloader.report_warning('video doesn\'t have subtitles')
741 return {}
742 return sub_lang_list
743
744 def _get_automatic_captions(self, video_id, webpage):
745 """We need the webpage for getting the captions url, pass it as an
746 argument to speed up the process."""
747 self.to_screen('%s: Looking for automatic captions' % video_id)
748 mobj = re.search(r';ytplayer.config = ({.*?});', webpage)
749 err_msg = 'Couldn\'t find automatic captions for %s' % video_id
750 if mobj is None:
751 self._downloader.report_warning(err_msg)
752 return {}
753 player_config = json.loads(mobj.group(1))
754 try:
755 args = player_config['args']
756 caption_url = args['ttsurl']
757 timestamp = args['timestamp']
758 # We get the available subtitles
759 list_params = compat_urllib_parse.urlencode({
760 'type': 'list',
761 'tlangs': 1,
762 'asrs': 1,
763 })
764 list_url = caption_url + '&' + list_params
765 caption_list = self._download_xml(list_url, video_id)
766 original_lang_node = caption_list.find('track')
767 if original_lang_node is None:
768 self._downloader.report_warning('Video doesn\'t have automatic captions')
769 return {}
770 original_lang = original_lang_node.attrib['lang_code']
771 caption_kind = original_lang_node.attrib.get('kind', '')
772
773 sub_lang_list = {}
774 for lang_node in caption_list.findall('target'):
775 sub_lang = lang_node.attrib['lang_code']
776 sub_formats = []
777 for ext in ['sbv', 'vtt', 'srt']:
778 params = compat_urllib_parse.urlencode({
779 'lang': original_lang,
780 'tlang': sub_lang,
781 'fmt': ext,
782 'ts': timestamp,
783 'kind': caption_kind,
784 })
785 sub_formats.append({
786 'url': caption_url + '&' + params,
787 'ext': ext,
788 })
789 sub_lang_list[sub_lang] = sub_formats
790 return sub_lang_list
791 # An extractor error can be raise by the download process if there are
792 # no automatic captions but there are subtitles
793 except (KeyError, ExtractorError):
794 self._downloader.report_warning(err_msg)
795 return {}
796
797 @classmethod
798 def extract_id(cls, url):
799 mobj = re.match(cls._VALID_URL, url, re.VERBOSE)
800 if mobj is None:
801 raise ExtractorError('Invalid URL: %s' % url)
802 video_id = mobj.group(2)
803 return video_id
804
805 def _extract_from_m3u8(self, manifest_url, video_id):
806 url_map = {}
807
808 def _get_urls(_manifest):
809 lines = _manifest.split('\n')
810 urls = filter(lambda l: l and not l.startswith('#'),
811 lines)
812 return urls
813 manifest = self._download_webpage(manifest_url, video_id, 'Downloading formats manifest')
814 formats_urls = _get_urls(manifest)
815 for format_url in formats_urls:
816 itag = self._search_regex(r'itag/(\d+?)/', format_url, 'itag')
817 url_map[itag] = format_url
818 return url_map
819
820 def _extract_annotations(self, video_id):
821 url = 'https://www.youtube.com/annotations_invideo?features=1&legacy=1&video_id=%s' % video_id
822 return self._download_webpage(url, video_id, note='Searching for annotations.', errnote='Unable to download video annotations.')
823
824 def _parse_dash_manifest(
825 self, video_id, dash_manifest_url, player_url, age_gate, fatal=True):
826 def decrypt_sig(mobj):
827 s = mobj.group(1)
828 dec_s = self._decrypt_signature(s, video_id, player_url, age_gate)
829 return '/signature/%s' % dec_s
830 dash_manifest_url = re.sub(r'/s/([a-fA-F0-9\.]+)', decrypt_sig, dash_manifest_url)
831 dash_doc = self._download_xml(
832 dash_manifest_url, video_id,
833 note='Downloading DASH manifest',
834 errnote='Could not download DASH manifest',
835 fatal=fatal)
836
837 if dash_doc is False:
838 return []
839
840 formats = []
841 for a in dash_doc.findall('.//{urn:mpeg:DASH:schema:MPD:2011}AdaptationSet'):
842 mime_type = a.attrib.get('mimeType')
843 for r in a.findall('{urn:mpeg:DASH:schema:MPD:2011}Representation'):
844 url_el = r.find('{urn:mpeg:DASH:schema:MPD:2011}BaseURL')
845 if url_el is None:
846 continue
847 if mime_type == 'text/vtt':
848 # TODO implement WebVTT downloading
849 pass
850 elif mime_type.startswith('audio/') or mime_type.startswith('video/'):
851 segment_list = r.find('{urn:mpeg:DASH:schema:MPD:2011}SegmentList')
852 format_id = r.attrib['id']
853 video_url = url_el.text
854 filesize = int_or_none(url_el.attrib.get('{http://youtube.com/yt/2012/10/10}contentLength'))
855 f = {
856 'format_id': format_id,
857 'url': video_url,
858 'width': int_or_none(r.attrib.get('width')),
859 'height': int_or_none(r.attrib.get('height')),
860 'tbr': int_or_none(r.attrib.get('bandwidth'), 1000),
861 'asr': int_or_none(r.attrib.get('audioSamplingRate')),
862 'filesize': filesize,
863 'fps': int_or_none(r.attrib.get('frameRate')),
864 }
865 if segment_list is not None:
866 f.update({
867 'initialization_url': segment_list.find('{urn:mpeg:DASH:schema:MPD:2011}Initialization').attrib['sourceURL'],
868 'segment_urls': [segment.attrib.get('media') for segment in segment_list.findall('{urn:mpeg:DASH:schema:MPD:2011}SegmentURL')],
869 'protocol': 'http_dash_segments',
870 })
871 try:
872 existing_format = next(
873 fo for fo in formats
874 if fo['format_id'] == format_id)
875 except StopIteration:
876 full_info = self._formats.get(format_id, {}).copy()
877 full_info.update(f)
878 codecs = r.attrib.get('codecs')
879 if codecs:
880 if full_info.get('acodec') == 'none' and 'vcodec' not in full_info:
881 full_info['vcodec'] = codecs
882 elif full_info.get('vcodec') == 'none' and 'acodec' not in full_info:
883 full_info['acodec'] = codecs
884 formats.append(full_info)
885 else:
886 existing_format.update(f)
887 else:
888 self.report_warning('Unknown MIME type %s in DASH manifest' % mime_type)
889 return formats
890
891 def _real_extract(self, url):
892 proto = (
893 'http' if self._downloader.params.get('prefer_insecure', False)
894 else 'https')
895
896 start_time = None
897 end_time = None
898 parsed_url = compat_urllib_parse_urlparse(url)
899 for component in [parsed_url.fragment, parsed_url.query]:
900 query = compat_parse_qs(component)
901 if start_time is None and 't' in query:
902 start_time = parse_duration(query['t'][0])
903 if start_time is None and 'start' in query:
904 start_time = parse_duration(query['start'][0])
905 if end_time is None and 'end' in query:
906 end_time = parse_duration(query['end'][0])
907
908 # Extract original video URL from URL with redirection, like age verification, using next_url parameter
909 mobj = re.search(self._NEXT_URL_RE, url)
910 if mobj:
911 url = proto + '://www.youtube.com/' + compat_urllib_parse_unquote(mobj.group(1)).lstrip('/')
912 video_id = self.extract_id(url)
913
914 # Get video webpage
915 url = proto + '://www.youtube.com/watch?v=%s&gl=US&hl=en&has_verified=1&bpctr=9999999999' % video_id
916 video_webpage = self._download_webpage(url, video_id)
917
918 # Attempt to extract SWF player URL
919 mobj = re.search(r'swfConfig.*?"(https?:\\/\\/.*?watch.*?-.*?\.swf)"', video_webpage)
920 if mobj is not None:
921 player_url = re.sub(r'\\(.)', r'\1', mobj.group(1))
922 else:
923 player_url = None
924
925 dash_mpds = []
926
927 def add_dash_mpd(video_info):
928 dash_mpd = video_info.get('dashmpd')
929 if dash_mpd and dash_mpd[0] not in dash_mpds:
930 dash_mpds.append(dash_mpd[0])
931
932 # Get video info
933 embed_webpage = None
934 is_live = None
935 if re.search(r'player-age-gate-content">', video_webpage) is not None:
936 age_gate = True
937 # We simulate the access to the video from www.youtube.com/v/{video_id}
938 # this can be viewed without login into Youtube
939 url = proto + '://www.youtube.com/embed/%s' % video_id
940 embed_webpage = self._download_webpage(url, video_id, 'Downloading embed webpage')
941 data = compat_urllib_parse.urlencode({
942 'video_id': video_id,
943 'eurl': 'https://youtube.googleapis.com/v/' + video_id,
944 'sts': self._search_regex(
945 r'"sts"\s*:\s*(\d+)', embed_webpage, 'sts', default=''),
946 })
947 video_info_url = proto + '://www.youtube.com/get_video_info?' + data
948 video_info_webpage = self._download_webpage(
949 video_info_url, video_id,
950 note='Refetching age-gated info webpage',
951 errnote='unable to download video info webpage')
952 video_info = compat_parse_qs(video_info_webpage)
953 add_dash_mpd(video_info)
954 else:
955 age_gate = False
956 video_info = None
957 # Try looking directly into the video webpage
958 mobj = re.search(r';ytplayer\.config\s*=\s*({.*?});', video_webpage)
959 if mobj:
960 json_code = uppercase_escape(mobj.group(1))
961 ytplayer_config = json.loads(json_code)
962 args = ytplayer_config['args']
963 if args.get('url_encoded_fmt_stream_map'):
964 # Convert to the same format returned by compat_parse_qs
965 video_info = dict((k, [v]) for k, v in args.items())
966 add_dash_mpd(video_info)
967 if args.get('livestream') == '1' or args.get('live_playback') == 1:
968 is_live = True
969 if not video_info or self._downloader.params.get('youtube_include_dash_manifest', True):
970 # We also try looking in get_video_info since it may contain different dashmpd
971 # URL that points to a DASH manifest with possibly different itag set (some itags
972 # are missing from DASH manifest pointed by webpage's dashmpd, some - from DASH
973 # manifest pointed by get_video_info's dashmpd).
974 # The general idea is to take a union of itags of both DASH manifests (for example
975 # video with such 'manifest behavior' see https://github.com/rg3/youtube-dl/issues/6093)
976 self.report_video_info_webpage_download(video_id)
977 for el_type in ['&el=info', '&el=embedded', '&el=detailpage', '&el=vevo', '']:
978 video_info_url = (
979 '%s://www.youtube.com/get_video_info?&video_id=%s%s&ps=default&eurl=&gl=US&hl=en'
980 % (proto, video_id, el_type))
981 video_info_webpage = self._download_webpage(
982 video_info_url,
983 video_id, note=False,
984 errnote='unable to download video info webpage')
985 get_video_info = compat_parse_qs(video_info_webpage)
986 if get_video_info.get('use_cipher_signature') != ['True']:
987 add_dash_mpd(get_video_info)
988 if not video_info:
989 video_info = get_video_info
990 if 'token' in get_video_info:
991 break
992 if 'token' not in video_info:
993 if 'reason' in video_info:
994 if 'The uploader has not made this video available in your country.' in video_info['reason']:
995 regions_allowed = self._html_search_meta('regionsAllowed', video_webpage, default=None)
996 if regions_allowed:
997 raise ExtractorError('YouTube said: This video is available in %s only' % (
998 ', '.join(map(ISO3166Utils.short2full, regions_allowed.split(',')))),
999 expected=True)
1000 raise ExtractorError(
1001 'YouTube said: %s' % video_info['reason'][0],
1002 expected=True, video_id=video_id)
1003 else:
1004 raise ExtractorError(
1005 '"token" parameter not in video info for unknown reason',
1006 video_id=video_id)
1007
1008 if 'view_count' in video_info:
1009 view_count = int(video_info['view_count'][0])
1010 else:
1011 view_count = None
1012
1013 # Check for "rental" videos
1014 if 'ypc_video_rental_bar_text' in video_info and 'author' not in video_info:
1015 raise ExtractorError('"rental" videos not supported')
1016
1017 # Start extracting information
1018 self.report_information_extraction(video_id)
1019
1020 # uploader
1021 if 'author' not in video_info:
1022 raise ExtractorError('Unable to extract uploader name')
1023 video_uploader = compat_urllib_parse_unquote_plus(video_info['author'][0])
1024
1025 # uploader_id
1026 video_uploader_id = None
1027 mobj = re.search(r'<link itemprop="url" href="http://www.youtube.com/(?:user|channel)/([^"]+)">', video_webpage)
1028 if mobj is not None:
1029 video_uploader_id = mobj.group(1)
1030 else:
1031 self._downloader.report_warning('unable to extract uploader nickname')
1032
1033 # title
1034 if 'title' in video_info:
1035 video_title = video_info['title'][0]
1036 else:
1037 self._downloader.report_warning('Unable to extract video title')
1038 video_title = '_'
1039
1040 # thumbnail image
1041 # We try first to get a high quality image:
1042 m_thumb = re.search(r'<span itemprop="thumbnail".*?href="(.*?)">',
1043 video_webpage, re.DOTALL)
1044 if m_thumb is not None:
1045 video_thumbnail = m_thumb.group(1)
1046 elif 'thumbnail_url' not in video_info:
1047 self._downloader.report_warning('unable to extract video thumbnail')
1048 video_thumbnail = None
1049 else: # don't panic if we can't find it
1050 video_thumbnail = compat_urllib_parse_unquote_plus(video_info['thumbnail_url'][0])
1051
1052 # upload date
1053 upload_date = self._html_search_meta(
1054 'datePublished', video_webpage, 'upload date', default=None)
1055 if not upload_date:
1056 upload_date = self._search_regex(
1057 [r'(?s)id="eow-date.*?>(.*?)</span>',
1058 r'id="watch-uploader-info".*?>.*?(?:Published|Uploaded|Streamed live|Started) on (.+?)</strong>'],
1059 video_webpage, 'upload date', default=None)
1060 if upload_date:
1061 upload_date = ' '.join(re.sub(r'[/,-]', r' ', mobj.group(1)).split())
1062 upload_date = unified_strdate(upload_date)
1063
1064 m_cat_container = self._search_regex(
1065 r'(?s)<h4[^>]*>\s*Category\s*</h4>\s*<ul[^>]*>(.*?)</ul>',
1066 video_webpage, 'categories', default=None)
1067 if m_cat_container:
1068 category = self._html_search_regex(
1069 r'(?s)<a[^<]+>(.*?)</a>', m_cat_container, 'category',
1070 default=None)
1071 video_categories = None if category is None else [category]
1072 else:
1073 video_categories = None
1074
1075 m = re.findall(r'''<meta(?:\s+[a-zA-Z0-9:._-]+(?:=[a-zA-Z0-9:._-]+|="[^"]+"|='[^']+'))*?\s+property=['"]?og:video:tag['"]?(?:\s+[a-zA-Z0-9:._-]+(?:=[a-zA-Z0-9:._-]+|="[^"]+"|='[^']+'))*?\s+content=['"]?([^>'"]+?)['"]?\s*>'''
1076 , video_webpage, re.DOTALL | re.IGNORECASE);
1077 video_tags = ", ".join(m)
1078 # description
1079 video_description = get_element_by_id("eow-description", video_webpage)
1080 if video_description:
1081 video_description = re.sub(r'''(?x)
1082 <a\s+
1083 (?:[a-zA-Z-]+="[^"]+"\s+)*?
1084 title="([^"]+)"\s+
1085 (?:[a-zA-Z-]+="[^"]+"\s+)*?
1086 class="yt-uix-redirect-link"\s*>
1087 [^<]+
1088 </a>
1089 ''', r'\1', video_description)
1090 video_description = clean_html(video_description)
1091 else:
1092 fd_mobj = re.search(r'<meta name="description" content="([^"]+)"', video_webpage)
1093 if fd_mobj:
1094 video_description = unescapeHTML(fd_mobj.group(1))
1095 else:
1096 video_description = ''
1097
1098 def _extract_count(count_name):
1099 return str_to_int(self._search_regex(
1100 r'-%s-button[^>]+><span[^>]+class="yt-uix-button-content"[^>]*>([\d,]+)</span>'
1101 % re.escape(count_name),
1102 video_webpage, count_name, default=None))
1103
1104 like_count = _extract_count('like')
1105 dislike_count = _extract_count('dislike')
1106
1107 # subtitles
1108 video_subtitles = self.extract_subtitles(video_id, video_webpage)
1109 automatic_captions = self.extract_automatic_captions(video_id, video_webpage)
1110
1111 if 'length_seconds' not in video_info:
1112 self._downloader.report_warning('unable to extract video duration')
1113 video_duration = None
1114 else:
1115 video_duration = int(compat_urllib_parse_unquote_plus(video_info['length_seconds'][0]))
1116
1117 # annotations
1118 video_annotations = None
1119 if self._downloader.params.get('writeannotations', False):
1120 video_annotations = self._extract_annotations(video_id)
1121
1122 def _map_to_format_list(urlmap):
1123 formats = []
1124 for itag, video_real_url in urlmap.items():
1125 dct = {
1126 'format_id': itag,
1127 'url': video_real_url,
1128 'player_url': player_url,
1129 }
1130 if itag in self._formats:
1131 dct.update(self._formats[itag])
1132 formats.append(dct)
1133 return formats
1134
1135 if 'conn' in video_info and video_info['conn'][0].startswith('rtmp'):
1136 self.report_rtmp_download()
1137 formats = [{
1138 'format_id': '_rtmp',
1139 'protocol': 'rtmp',
1140 'url': video_info['conn'][0],
1141 'player_url': player_url,
1142 }]
1143 elif len(video_info.get('url_encoded_fmt_stream_map', [''])[0]) >= 1 or len(video_info.get('adaptive_fmts', [''])[0]) >= 1:
1144 encoded_url_map = video_info.get('url_encoded_fmt_stream_map', [''])[0] + ',' + video_info.get('adaptive_fmts', [''])[0]
1145 if 'rtmpe%3Dyes' in encoded_url_map:
1146 raise ExtractorError('rtmpe downloads are not supported, see https://github.com/rg3/youtube-dl/issues/343 for more information.', expected=True)
1147 url_map = {}
1148 for url_data_str in encoded_url_map.split(','):
1149 url_data = compat_parse_qs(url_data_str)
1150 if 'itag' not in url_data or 'url' not in url_data:
1151 continue
1152 format_id = url_data['itag'][0]
1153 url = url_data['url'][0]
1154
1155 if 'sig' in url_data:
1156 url += '&signature=' + url_data['sig'][0]
1157 elif 's' in url_data:
1158 encrypted_sig = url_data['s'][0]
1159 ASSETS_RE = r'"assets":.+?"js":\s*("[^"]+")'
1160
1161 jsplayer_url_json = self._search_regex(
1162 ASSETS_RE,
1163 embed_webpage if age_gate else video_webpage,
1164 'JS player URL (1)', default=None)
1165 if not jsplayer_url_json and not age_gate:
1166 # We need the embed website after all
1167 if embed_webpage is None:
1168 embed_url = proto + '://www.youtube.com/embed/%s' % video_id
1169 embed_webpage = self._download_webpage(
1170 embed_url, video_id, 'Downloading embed webpage')
1171 jsplayer_url_json = self._search_regex(
1172 ASSETS_RE, embed_webpage, 'JS player URL')
1173
1174 player_url = json.loads(jsplayer_url_json)
1175 if player_url is None:
1176 player_url_json = self._search_regex(
1177 r'ytplayer\.config.*?"url"\s*:\s*("[^"]+")',
1178 video_webpage, 'age gate player URL')
1179 player_url = json.loads(player_url_json)
1180
1181 if self._downloader.params.get('verbose'):
1182 if player_url is None:
1183 player_version = 'unknown'
1184 player_desc = 'unknown'
1185 else:
1186 if player_url.endswith('swf'):
1187 player_version = self._search_regex(
1188 r'-(.+?)(?:/watch_as3)?\.swf$', player_url,
1189 'flash player', fatal=False)
1190 player_desc = 'flash player %s' % player_version
1191 else:
1192 player_version = self._search_regex(
1193 r'html5player-([^/]+?)(?:/html5player)?\.js',
1194 player_url,
1195 'html5 player', fatal=False)
1196 player_desc = 'html5 player %s' % player_version
1197
1198 parts_sizes = self._signature_cache_id(encrypted_sig)
1199 self.to_screen('{%s} signature length %s, %s' %
1200 (format_id, parts_sizes, player_desc))
1201
1202 signature = self._decrypt_signature(
1203 encrypted_sig, video_id, player_url, age_gate)
1204 url += '&signature=' + signature
1205 if 'ratebypass' not in url:
1206 url += '&ratebypass=yes'
1207 url_map[format_id] = url
1208 formats = _map_to_format_list(url_map)
1209 elif video_info.get('hlsvp'):
1210 manifest_url = video_info['hlsvp'][0]
1211 url_map = self._extract_from_m3u8(manifest_url, video_id)
1212 formats = _map_to_format_list(url_map)
1213 else:
1214 raise ExtractorError('no conn, hlsvp or url_encoded_fmt_stream_map information found in video info')
1215
1216 # Look for the DASH manifest
1217 if self._downloader.params.get('youtube_include_dash_manifest', True):
1218 dash_mpd_fatal = True
1219 for dash_manifest_url in dash_mpds:
1220 dash_formats = {}
1221 try:
1222 for df in self._parse_dash_manifest(
1223 video_id, dash_manifest_url, player_url, age_gate, dash_mpd_fatal):
1224 # Do not overwrite DASH format found in some previous DASH manifest
1225 if df['format_id'] not in dash_formats:
1226 dash_formats[df['format_id']] = df
1227 # Additional DASH manifests may end up in HTTP Error 403 therefore
1228 # allow them to fail without bug report message if we already have
1229 # some DASH manifest succeeded. This is temporary workaround to reduce
1230 # burst of bug reports until we figure out the reason and whether it
1231 # can be fixed at all.
1232 dash_mpd_fatal = False
1233 except (ExtractorError, KeyError) as e:
1234 self.report_warning(
1235 'Skipping DASH manifest: %r' % e, video_id)
1236 if dash_formats:
1237 # Remove the formats we found through non-DASH, they
1238 # contain less info and it can be wrong, because we use
1239 # fixed values (for example the resolution). See
1240 # https://github.com/rg3/youtube-dl/issues/5774 for an
1241 # example.
1242 formats = [f for f in formats if f['format_id'] not in dash_formats.keys()]
1243 formats.extend(dash_formats.values())
1244
1245 # Check for malformed aspect ratio
1246 stretched_m = re.search(
1247 r'<meta\s+property="og:video:tag".*?content="yt:stretch=(?P<w>[0-9]+):(?P<h>[0-9]+)">',
1248 video_webpage)
1249 if stretched_m:
1250 ratio = float(stretched_m.group('w')) / float(stretched_m.group('h'))
1251 for f in formats:
1252 if f.get('vcodec') != 'none':
1253 f['stretched_ratio'] = ratio
1254
1255 self._sort_formats(formats)
1256
1257 return {
1258 'id': video_id,
1259 'uploader': video_uploader,
1260 'uploader_id': video_uploader_id,
1261 'upload_date': upload_date,
1262 'title': video_title,
1263 'thumbnail': video_thumbnail,
1264 'description': video_description,
1265 'tags' : video_tags,
1266 'categories': video_categories,
1267 'subtitles': video_subtitles,
1268 'automatic_captions': automatic_captions,
1269 'duration': video_duration,
1270 'age_limit': 18 if age_gate else 0,
1271 'annotations': video_annotations,
1272 'webpage_url': proto + '://www.youtube.com/watch?v=%s' % video_id,
1273 'view_count': view_count,
1274 'like_count': like_count,
1275 'dislike_count': dislike_count,
1276 'average_rating': float_or_none(video_info.get('avg_rating', [None])[0]),
1277 'formats': formats,
1278 'is_live': is_live,
1279 'start_time': start_time,
1280 'end_time': end_time,
1281 }
1282
1283
1284 class YoutubePlaylistIE(YoutubeBaseInfoExtractor):
1285 IE_DESC = 'YouTube.com playlists'
1286 _VALID_URL = r"""(?x)(?:
1287 (?:https?://)?
1288 (?:\w+\.)?
1289 youtube\.com/
1290 (?:
1291 (?:course|view_play_list|my_playlists|artist|playlist|watch|embed/videoseries)
1292 \? (?:.*?&)*? (?:p|a|list)=
1293 | p/
1294 )
1295 (
1296 (?:PL|LL|EC|UU|FL|RD|UL)?[0-9A-Za-z-_]{10,}
1297 # Top tracks, they can also include dots
1298 |(?:MC)[\w\.]*
1299 )
1300 .*
1301 |
1302 ((?:PL|LL|EC|UU|FL|RD|UL)[0-9A-Za-z-_]{10,})
1303 )"""
1304 _TEMPLATE_URL = 'https://www.youtube.com/playlist?list=%s'
1305 _VIDEO_RE = r'href="\s*/watch\?v=(?P<id>[0-9A-Za-z_-]{11})&amp;[^"]*?index=(?P<index>\d+)'
1306 IE_NAME = 'youtube:playlist'
1307 _TESTS = [{
1308 'url': 'https://www.youtube.com/playlist?list=PLwiyx1dc3P2JR9N8gQaQN_BCvlSlap7re',
1309 'info_dict': {
1310 'title': 'ytdl test PL',
1311 'id': 'PLwiyx1dc3P2JR9N8gQaQN_BCvlSlap7re',
1312 },
1313 'playlist_count': 3,
1314 }, {
1315 'url': 'https://www.youtube.com/playlist?list=PLtPgu7CB4gbZDA7i_euNxn75ISqxwZPYx',
1316 'info_dict': {
1317 'id': 'PLtPgu7CB4gbZDA7i_euNxn75ISqxwZPYx',
1318 'title': 'YDL_Empty_List',
1319 },
1320 'playlist_count': 0,
1321 }, {
1322 'note': 'Playlist with deleted videos (#651). As a bonus, the video #51 is also twice in this list.',
1323 'url': 'https://www.youtube.com/playlist?list=PLwP_SiAcdui0KVebT0mU9Apz359a4ubsC',
1324 'info_dict': {
1325 'title': '29C3: Not my department',
1326 'id': 'PLwP_SiAcdui0KVebT0mU9Apz359a4ubsC',
1327 },
1328 'playlist_count': 95,
1329 }, {
1330 'note': 'issue #673',
1331 'url': 'PLBB231211A4F62143',
1332 'info_dict': {
1333 'title': '[OLD]Team Fortress 2 (Class-based LP)',
1334 'id': 'PLBB231211A4F62143',
1335 },
1336 'playlist_mincount': 26,
1337 }, {
1338 'note': 'Large playlist',
1339 'url': 'https://www.youtube.com/playlist?list=UUBABnxM4Ar9ten8Mdjj1j0Q',
1340 'info_dict': {
1341 'title': 'Uploads from Cauchemar',
1342 'id': 'UUBABnxM4Ar9ten8Mdjj1j0Q',
1343 },
1344 'playlist_mincount': 799,
1345 }, {
1346 'url': 'PLtPgu7CB4gbY9oDN3drwC3cMbJggS7dKl',
1347 'info_dict': {
1348 'title': 'YDL_safe_search',
1349 'id': 'PLtPgu7CB4gbY9oDN3drwC3cMbJggS7dKl',
1350 },
1351 'playlist_count': 2,
1352 }, {
1353 'note': 'embedded',
1354 'url': 'http://www.youtube.com/embed/videoseries?list=PL6IaIsEjSbf96XFRuNccS_RuEXwNdsoEu',
1355 'playlist_count': 4,
1356 'info_dict': {
1357 'title': 'JODA15',
1358 'id': 'PL6IaIsEjSbf96XFRuNccS_RuEXwNdsoEu',
1359 }
1360 }, {
1361 'note': 'Embedded SWF player',
1362 'url': 'http://www.youtube.com/p/YN5VISEtHet5D4NEvfTd0zcgFk84NqFZ?hl=en_US&fs=1&rel=0',
1363 'playlist_count': 4,
1364 'info_dict': {
1365 'title': 'JODA7',
1366 'id': 'YN5VISEtHet5D4NEvfTd0zcgFk84NqFZ',
1367 }
1368 }, {
1369 'note': 'Buggy playlist: the webpage has a "Load more" button but it doesn\'t have more videos',
1370 'url': 'https://www.youtube.com/playlist?list=UUXw-G3eDE9trcvY2sBMM_aA',
1371 'info_dict': {
1372 'title': 'Uploads from Interstellar Movie',
1373 'id': 'UUXw-G3eDE9trcvY2sBMM_aA',
1374 },
1375 'playlist_mincout': 21,
1376 }]
1377
1378 def _real_initialize(self):
1379 self._login()
1380
1381 def _extract_mix(self, playlist_id):
1382 # The mixes are generated from a single video
1383 # the id of the playlist is just 'RD' + video_id
1384 url = 'https://youtube.com/watch?v=%s&list=%s' % (playlist_id[-11:], playlist_id)
1385 webpage = self._download_webpage(
1386 url, playlist_id, 'Downloading Youtube mix')
1387 search_title = lambda class_name: get_element_by_attribute('class', class_name, webpage)
1388 title_span = (
1389 search_title('playlist-title') or
1390 search_title('title long-title') or
1391 search_title('title'))
1392 title = clean_html(title_span)
1393 ids = orderedSet(re.findall(
1394 r'''(?xs)data-video-username=".*?".*?
1395 href="/watch\?v=([0-9A-Za-z_-]{11})&amp;[^"]*?list=%s''' % re.escape(playlist_id),
1396 webpage))
1397 url_results = self._ids_to_results(ids)
1398
1399 return self.playlist_result(url_results, playlist_id, title)
1400
1401 def _extract_playlist(self, playlist_id):
1402 url = self._TEMPLATE_URL % playlist_id
1403 page = self._download_webpage(url, playlist_id)
1404
1405 for match in re.findall(r'<div class="yt-alert-message">([^<]+)</div>', page):
1406 match = match.strip()
1407 # Check if the playlist exists or is private
1408 if re.match(r'[^<]*(The|This) playlist (does not exist|is private)[^<]*', match):
1409 raise ExtractorError(
1410 'The playlist doesn\'t exist or is private, use --username or '
1411 '--netrc to access it.',
1412 expected=True)
1413 elif re.match(r'[^<]*Invalid parameters[^<]*', match):
1414 raise ExtractorError(
1415 'Invalid parameters. Maybe URL is incorrect.',
1416 expected=True)
1417 elif re.match(r'[^<]*Choose your language[^<]*', match):
1418 continue
1419 else:
1420 self.report_warning('Youtube gives an alert message: ' + match)
1421
1422 # Extract the video ids from the playlist pages
1423 def _entries():
1424 more_widget_html = content_html = page
1425 for page_num in itertools.count(1):
1426 matches = re.finditer(self._VIDEO_RE, content_html)
1427 # We remove the duplicates and the link with index 0
1428 # (it's not the first video of the playlist)
1429 new_ids = orderedSet(m.group('id') for m in matches if m.group('index') != '0')
1430 for vid_id in new_ids:
1431 yield self.url_result(vid_id, 'Youtube', video_id=vid_id)
1432
1433 mobj = re.search(r'data-uix-load-more-href="/?(?P<more>[^"]+)"', more_widget_html)
1434 if not mobj:
1435 break
1436
1437 more = self._download_json(
1438 'https://youtube.com/%s' % mobj.group('more'), playlist_id,
1439 'Downloading page #%s' % page_num,
1440 transform_source=uppercase_escape)
1441 content_html = more['content_html']
1442 if not content_html.strip():
1443 # Some webpages show a "Load more" button but they don't
1444 # have more videos
1445 break
1446 more_widget_html = more['load_more_widget_html']
1447
1448 playlist_title = self._html_search_regex(
1449 r'(?s)<h1 class="pl-header-title[^"]*">\s*(.*?)\s*</h1>',
1450 page, 'title')
1451
1452 return self.playlist_result(_entries(), playlist_id, playlist_title)
1453
1454 def _real_extract(self, url):
1455 # Extract playlist id
1456 mobj = re.match(self._VALID_URL, url)
1457 if mobj is None:
1458 raise ExtractorError('Invalid URL: %s' % url)
1459 playlist_id = mobj.group(1) or mobj.group(2)
1460
1461 # Check if it's a video-specific URL
1462 query_dict = compat_urlparse.parse_qs(compat_urlparse.urlparse(url).query)
1463 if 'v' in query_dict:
1464 video_id = query_dict['v'][0]
1465 if self._downloader.params.get('noplaylist'):
1466 self.to_screen('Downloading just video %s because of --no-playlist' % video_id)
1467 return self.url_result(video_id, 'Youtube', video_id=video_id)
1468 else:
1469 self.to_screen('Downloading playlist %s - add --no-playlist to just download video %s' % (playlist_id, video_id))
1470
1471 if playlist_id.startswith('RD') or playlist_id.startswith('UL'):
1472 # Mixes require a custom extraction process
1473 return self._extract_mix(playlist_id)
1474
1475 return self._extract_playlist(playlist_id)
1476
1477
1478 class YoutubeChannelIE(InfoExtractor):
1479 IE_DESC = 'YouTube.com channels'
1480 _VALID_URL = r'https?://(?:youtu\.be|(?:\w+\.)?youtube(?:-nocookie)?\.com)/channel/(?P<id>[0-9A-Za-z_-]+)'
1481 _TEMPLATE_URL = 'https://www.youtube.com/channel/%s/videos'
1482 IE_NAME = 'youtube:channel'
1483 _TESTS = [{
1484 'note': 'paginated channel',
1485 'url': 'https://www.youtube.com/channel/UCKfVa3S1e4PHvxWcwyMMg8w',
1486 'playlist_mincount': 91,
1487 'info_dict': {
1488 'id': 'UCKfVa3S1e4PHvxWcwyMMg8w',
1489 }
1490 }]
1491
1492 @staticmethod
1493 def extract_videos_from_page(page):
1494 ids_in_page = []
1495 titles_in_page = []
1496 for mobj in re.finditer(r'(?:title="(?P<title>[^"]+)"[^>]+)?href="/watch\?v=(?P<id>[0-9A-Za-z_-]+)&?', page):
1497 video_id = mobj.group('id')
1498 video_title = unescapeHTML(mobj.group('title'))
1499 try:
1500 idx = ids_in_page.index(video_id)
1501 if video_title and not titles_in_page[idx]:
1502 titles_in_page[idx] = video_title
1503 except ValueError:
1504 ids_in_page.append(video_id)
1505 titles_in_page.append(video_title)
1506 return zip(ids_in_page, titles_in_page)
1507
1508 def _real_extract(self, url):
1509 channel_id = self._match_id(url)
1510
1511 url = self._TEMPLATE_URL % channel_id
1512
1513 # Channel by page listing is restricted to 35 pages of 30 items, i.e. 1050 videos total (see #5778)
1514 # Workaround by extracting as a playlist if managed to obtain channel playlist URL
1515 # otherwise fallback on channel by page extraction
1516 channel_page = self._download_webpage(
1517 url + '?view=57', channel_id,
1518 'Downloading channel page', fatal=False)
1519 channel_playlist_id = self._html_search_meta(
1520 'channelId', channel_page, 'channel id', default=None)
1521 if not channel_playlist_id:
1522 channel_playlist_id = self._search_regex(
1523 r'data-channel-external-id="([^"]+)"',
1524 channel_page, 'channel id', default=None)
1525 if channel_playlist_id and channel_playlist_id.startswith('UC'):
1526 playlist_id = 'UU' + channel_playlist_id[2:]
1527 return self.url_result(
1528 compat_urlparse.urljoin(url, '/playlist?list=%s' % playlist_id), 'YoutubePlaylist')
1529
1530 channel_page = self._download_webpage(url, channel_id, 'Downloading page #1')
1531 autogenerated = re.search(r'''(?x)
1532 class="[^"]*?(?:
1533 channel-header-autogenerated-label|
1534 yt-channel-title-autogenerated
1535 )[^"]*"''', channel_page) is not None
1536
1537 if autogenerated:
1538 # The videos are contained in a single page
1539 # the ajax pages can't be used, they are empty
1540 entries = [
1541 self.url_result(
1542 video_id, 'Youtube', video_id=video_id,
1543 video_title=video_title)
1544 for video_id, video_title in self.extract_videos_from_page(channel_page)]
1545 return self.playlist_result(entries, channel_id)
1546
1547 def _entries():
1548 more_widget_html = content_html = channel_page
1549 for pagenum in itertools.count(1):
1550
1551 for video_id, video_title in self.extract_videos_from_page(content_html):
1552 yield self.url_result(
1553 video_id, 'Youtube', video_id=video_id,
1554 video_title=video_title)
1555
1556 mobj = re.search(
1557 r'data-uix-load-more-href="/?(?P<more>[^"]+)"',
1558 more_widget_html)
1559 if not mobj:
1560 break
1561
1562 more = self._download_json(
1563 'https://youtube.com/%s' % mobj.group('more'), channel_id,
1564 'Downloading page #%s' % (pagenum + 1),
1565 transform_source=uppercase_escape)
1566 content_html = more['content_html']
1567 more_widget_html = more['load_more_widget_html']
1568
1569 return self.playlist_result(_entries(), channel_id)
1570
1571
1572 class YoutubeUserIE(YoutubeChannelIE):
1573 IE_DESC = 'YouTube.com user videos (URL or "ytuser" keyword)'
1574 _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_-]+)'
1575 _TEMPLATE_URL = 'https://www.youtube.com/user/%s/videos'
1576 IE_NAME = 'youtube:user'
1577
1578 _TESTS = [{
1579 'url': 'https://www.youtube.com/user/TheLinuxFoundation',
1580 'playlist_mincount': 320,
1581 'info_dict': {
1582 'title': 'TheLinuxFoundation',
1583 }
1584 }, {
1585 'url': 'ytuser:phihag',
1586 'only_matching': True,
1587 }]
1588
1589 @classmethod
1590 def suitable(cls, url):
1591 # Don't return True if the url can be extracted with other youtube
1592 # extractor, the regex would is too permissive and it would match.
1593 other_ies = iter(klass for (name, klass) in globals().items() if name.endswith('IE') and klass is not cls)
1594 if any(ie.suitable(url) for ie in other_ies):
1595 return False
1596 else:
1597 return super(YoutubeUserIE, cls).suitable(url)
1598
1599
1600 class YoutubeSearchIE(SearchInfoExtractor, YoutubePlaylistIE):
1601 IE_DESC = 'YouTube.com searches'
1602 # there doesn't appear to be a real limit, for example if you search for
1603 # 'python' you get more than 8.000.000 results
1604 _MAX_RESULTS = float('inf')
1605 IE_NAME = 'youtube:search'
1606 _SEARCH_KEY = 'ytsearch'
1607 _EXTRA_QUERY_ARGS = {}
1608 _TESTS = []
1609
1610 def _get_n_results(self, query, n):
1611 """Get a specified number of results for a query"""
1612
1613 videos = []
1614 limit = n
1615
1616 for pagenum in itertools.count(1):
1617 url_query = {
1618 'search_query': query.encode('utf-8'),
1619 'page': pagenum,
1620 'spf': 'navigate',
1621 }
1622 url_query.update(self._EXTRA_QUERY_ARGS)
1623 result_url = 'https://www.youtube.com/results?' + compat_urllib_parse.urlencode(url_query)
1624 data = self._download_json(
1625 result_url, video_id='query "%s"' % query,
1626 note='Downloading page %s' % pagenum,
1627 errnote='Unable to download API page')
1628 html_content = data[1]['body']['content']
1629
1630 if 'class="search-message' in html_content:
1631 raise ExtractorError(
1632 '[youtube] No video results', expected=True)
1633
1634 new_videos = self._ids_to_results(orderedSet(re.findall(
1635 r'href="/watch\?v=(.{11})', html_content)))
1636 videos += new_videos
1637 if not new_videos or len(videos) > limit:
1638 break
1639
1640 if len(videos) > n:
1641 videos = videos[:n]
1642 return self.playlist_result(videos, query)
1643
1644
1645 class YoutubeSearchDateIE(YoutubeSearchIE):
1646 IE_NAME = YoutubeSearchIE.IE_NAME + ':date'
1647 _SEARCH_KEY = 'ytsearchdate'
1648 IE_DESC = 'YouTube.com searches, newest videos first'
1649 _EXTRA_QUERY_ARGS = {'search_sort': 'video_date_uploaded'}
1650
1651
1652 class YoutubeSearchURLIE(InfoExtractor):
1653 IE_DESC = 'YouTube.com search URLs'
1654 IE_NAME = 'youtube:search_url'
1655 _VALID_URL = r'https?://(?:www\.)?youtube\.com/results\?(.*?&)?search_query=(?P<query>[^&]+)(?:[&]|$)'
1656 _TESTS = [{
1657 'url': 'https://www.youtube.com/results?baz=bar&search_query=youtube-dl+test+video&filters=video&lclk=video',
1658 'playlist_mincount': 5,
1659 'info_dict': {
1660 'title': 'youtube-dl test video',
1661 }
1662 }]
1663
1664 def _real_extract(self, url):
1665 mobj = re.match(self._VALID_URL, url)
1666 query = compat_urllib_parse_unquote_plus(mobj.group('query'))
1667
1668 webpage = self._download_webpage(url, query)
1669 result_code = self._search_regex(
1670 r'(?s)<ol[^>]+class="item-section"(.*?)</ol>', webpage, 'result HTML')
1671
1672 part_codes = re.findall(
1673 r'(?s)<h3 class="yt-lockup-title">(.*?)</h3>', result_code)
1674 entries = []
1675 for part_code in part_codes:
1676 part_title = self._html_search_regex(
1677 [r'(?s)title="([^"]+)"', r'>([^<]+)</a>'], part_code, 'item title', fatal=False)
1678 part_url_snippet = self._html_search_regex(
1679 r'(?s)href="([^"]+)"', part_code, 'item URL')
1680 part_url = compat_urlparse.urljoin(
1681 'https://www.youtube.com/', part_url_snippet)
1682 entries.append({
1683 '_type': 'url',
1684 'url': part_url,
1685 'title': part_title,
1686 })
1687
1688 return {
1689 '_type': 'playlist',
1690 'entries': entries,
1691 'title': query,
1692 }
1693
1694
1695 class YoutubeShowIE(InfoExtractor):
1696 IE_DESC = 'YouTube.com (multi-season) shows'
1697 _VALID_URL = r'https?://www\.youtube\.com/show/(?P<id>[^?#]*)'
1698 IE_NAME = 'youtube:show'
1699 _TESTS = [{
1700 'url': 'http://www.youtube.com/show/airdisasters',
1701 'playlist_mincount': 3,
1702 'info_dict': {
1703 'id': 'airdisasters',
1704 'title': 'Air Disasters',
1705 }
1706 }]
1707
1708 def _real_extract(self, url):
1709 mobj = re.match(self._VALID_URL, url)
1710 playlist_id = mobj.group('id')
1711 webpage = self._download_webpage(
1712 url, playlist_id, 'Downloading show webpage')
1713 # There's one playlist for each season of the show
1714 m_seasons = list(re.finditer(r'href="(/playlist\?list=.*?)"', webpage))
1715 self.to_screen('%s: Found %s seasons' % (playlist_id, len(m_seasons)))
1716 entries = [
1717 self.url_result(
1718 'https://www.youtube.com' + season.group(1), 'YoutubePlaylist')
1719 for season in m_seasons
1720 ]
1721 title = self._og_search_title(webpage, fatal=False)
1722
1723 return {
1724 '_type': 'playlist',
1725 'id': playlist_id,
1726 'title': title,
1727 'entries': entries,
1728 }
1729
1730
1731 class YoutubeFeedsInfoExtractor(YoutubeBaseInfoExtractor):
1732 """
1733 Base class for feed extractors
1734 Subclasses must define the _FEED_NAME and _PLAYLIST_TITLE properties.
1735 """
1736 _LOGIN_REQUIRED = True
1737
1738 @property
1739 def IE_NAME(self):
1740 return 'youtube:%s' % self._FEED_NAME
1741
1742 def _real_initialize(self):
1743 self._login()
1744
1745 def _real_extract(self, url):
1746 page = self._download_webpage(
1747 'https://www.youtube.com/feed/%s' % self._FEED_NAME, self._PLAYLIST_TITLE)
1748
1749 # The extraction process is the same as for playlists, but the regex
1750 # for the video ids doesn't contain an index
1751 ids = []
1752 more_widget_html = content_html = page
1753 for page_num in itertools.count(1):
1754 matches = re.findall(r'href="\s*/watch\?v=([0-9A-Za-z_-]{11})', content_html)
1755
1756 # 'recommended' feed has infinite 'load more' and each new portion spins
1757 # the same videos in (sometimes) slightly different order, so we'll check
1758 # for unicity and break when portion has no new videos
1759 new_ids = filter(lambda video_id: video_id not in ids, orderedSet(matches))
1760 if not new_ids:
1761 break
1762
1763 ids.extend(new_ids)
1764
1765 mobj = re.search(r'data-uix-load-more-href="/?(?P<more>[^"]+)"', more_widget_html)
1766 if not mobj:
1767 break
1768
1769 more = self._download_json(
1770 'https://youtube.com/%s' % mobj.group('more'), self._PLAYLIST_TITLE,
1771 'Downloading page #%s' % page_num,
1772 transform_source=uppercase_escape)
1773 content_html = more['content_html']
1774 more_widget_html = more['load_more_widget_html']
1775
1776 return self.playlist_result(
1777 self._ids_to_results(ids), playlist_title=self._PLAYLIST_TITLE)
1778
1779
1780 class YoutubeWatchLaterIE(YoutubePlaylistIE):
1781 IE_NAME = 'youtube:watchlater'
1782 IE_DESC = 'Youtube watch later list, ":ytwatchlater" for short (requires authentication)'
1783 _VALID_URL = r'https?://www\.youtube\.com/(?:feed/watch_later|playlist\?list=WL)|:ytwatchlater'
1784
1785 _TESTS = [] # override PlaylistIE tests
1786
1787 def _real_extract(self, url):
1788 return self._extract_playlist('WL')
1789
1790
1791 class YoutubeFavouritesIE(YoutubeBaseInfoExtractor):
1792 IE_NAME = 'youtube:favorites'
1793 IE_DESC = 'YouTube.com favourite videos, ":ytfav" for short (requires authentication)'
1794 _VALID_URL = r'https?://www\.youtube\.com/my_favorites|:ytfav(?:ou?rites)?'
1795 _LOGIN_REQUIRED = True
1796
1797 def _real_extract(self, url):
1798 webpage = self._download_webpage('https://www.youtube.com/my_favorites', 'Youtube Favourites videos')
1799 playlist_id = self._search_regex(r'list=(.+?)["&]', webpage, 'favourites playlist id')
1800 return self.url_result(playlist_id, 'YoutubePlaylist')
1801
1802
1803 class YoutubeRecommendedIE(YoutubeFeedsInfoExtractor):
1804 IE_DESC = 'YouTube.com recommended videos, ":ytrec" for short (requires authentication)'
1805 _VALID_URL = r'https?://www\.youtube\.com/feed/recommended|:ytrec(?:ommended)?'
1806 _FEED_NAME = 'recommended'
1807 _PLAYLIST_TITLE = 'Youtube Recommended videos'
1808
1809
1810 class YoutubeSubscriptionsIE(YoutubeFeedsInfoExtractor):
1811 IE_DESC = 'YouTube.com subscriptions feed, "ytsubs" keyword (requires authentication)'
1812 _VALID_URL = r'https?://www\.youtube\.com/feed/subscriptions|:ytsubs(?:criptions)?'
1813 _FEED_NAME = 'subscriptions'
1814 _PLAYLIST_TITLE = 'Youtube Subscriptions'
1815
1816
1817 class YoutubeHistoryIE(YoutubeFeedsInfoExtractor):
1818 IE_DESC = 'Youtube watch history, ":ythistory" for short (requires authentication)'
1819 _VALID_URL = 'https?://www\.youtube\.com/feed/history|:ythistory'
1820 _FEED_NAME = 'history'
1821 _PLAYLIST_TITLE = 'Youtube History'
1822
1823
1824 class YoutubeTruncatedURLIE(InfoExtractor):
1825 IE_NAME = 'youtube:truncated_url'
1826 IE_DESC = False # Do not list
1827 _VALID_URL = r'''(?x)
1828 (?:https?://)?
1829 (?:\w+\.)?[yY][oO][uU][tT][uU][bB][eE](?:-nocookie)?\.com/
1830 (?:watch\?(?:
1831 feature=[a-z_]+|
1832 annotation_id=annotation_[^&]+|
1833 x-yt-cl=[0-9]+|
1834 hl=[^&]*|
1835 )?
1836 |
1837 attribution_link\?a=[^&]+
1838 )
1839 $
1840 '''
1841
1842 _TESTS = [{
1843 'url': 'http://www.youtube.com/watch?annotation_id=annotation_3951667041',
1844 'only_matching': True,
1845 }, {
1846 'url': 'http://www.youtube.com/watch?',
1847 'only_matching': True,
1848 }, {
1849 'url': 'https://www.youtube.com/watch?x-yt-cl=84503534',
1850 'only_matching': True,
1851 }, {
1852 'url': 'https://www.youtube.com/watch?feature=foo',
1853 'only_matching': True,
1854 }, {
1855 'url': 'https://www.youtube.com/watch?hl=en-GB',
1856 'only_matching': True,
1857 }]
1858
1859 def _real_extract(self, url):
1860 raise ExtractorError(
1861 'Did you forget to quote the URL? Remember that & is a meta '
1862 'character in most shells, so you want to put the URL in quotes, '
1863 'like youtube-dl '
1864 '"http://www.youtube.com/watch?feature=foo&v=BaW_jenozKc" '
1865 ' or simply youtube-dl BaW_jenozKc .',
1866 expected=True)
1867
1868
1869 class YoutubeTruncatedIDIE(InfoExtractor):
1870 IE_NAME = 'youtube:truncated_id'
1871 IE_DESC = False # Do not list
1872 _VALID_URL = r'https?://(?:www\.)?youtube\.com/watch\?v=(?P<id>[0-9A-Za-z_-]{1,10})$'
1873
1874 _TESTS = [{
1875 'url': 'https://www.youtube.com/watch?v=N_708QY7Ob',
1876 'only_matching': True,
1877 }]
1878
1879 def _real_extract(self, url):
1880 video_id = self._match_id(url)
1881 raise ExtractorError(
1882 'Incomplete YouTube ID %s. URL %s looks truncated.' % (video_id, url),
1883 expected=True)
fork of https://github.com/yt-dlp/yt-dlp/