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