]> Raphaël G. Git Repositories - youtubedl/blob - youtube_dl/extractor/youtube.py
3d8b31f9830b218e02021b41214010b4f61d29a9
[youtubedl] / 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_request,
23 compat_urlparse,
24 compat_str,
25 )
26 from ..utils import (
27 clean_html,
28 ExtractorError,
29 float_or_none,
30 get_element_by_attribute,
31 get_element_by_id,
32 int_or_none,
33 orderedSet,
34 str_to_int,
35 unescapeHTML,
36 unified_strdate,
37 uppercase_escape,
38 ISO3166Utils,
39 )
40
41
42 class YoutubeBaseInfoExtractor(InfoExtractor):
43 """Provide base functions for Youtube extractors"""
44 _LOGIN_URL = 'https://accounts.google.com/ServiceLogin'
45 _TWOFACTOR_URL = 'https://accounts.google.com/SecondFactor'
46 _NETRC_MACHINE = 'youtube'
47 # If True it will raise an error if no login info is provided
48 _LOGIN_REQUIRED = False
49
50 def _set_language(self):
51 self._set_cookie(
52 '.youtube.com', 'PREF', 'f1=50000000&hl=en',
53 # YouTube sets the expire time to about two months
54 expire_time=time.time() + 2 * 30 * 24 * 3600)
55
56 def _ids_to_results(self, ids):
57 return [
58 self.url_result(vid_id, 'Youtube', video_id=vid_id)
59 for vid_id in ids]
60
61 def _login(self):
62 """
63 Attempt to log in to YouTube.
64 True is returned if successful or skipped.
65 False is returned if login failed.
66
67 If _LOGIN_REQUIRED is set and no authentication was provided, an error is raised.
68 """
69 (username, password) = self._get_login_info()
70 # No authentication to be performed
71 if username is None:
72 if self._LOGIN_REQUIRED:
73 raise ExtractorError('No login info available, needed for using %s.' % self.IE_NAME, expected=True)
74 return True
75
76 login_page = self._download_webpage(
77 self._LOGIN_URL, None,
78 note='Downloading login page',
79 errnote='unable to fetch login page', fatal=False)
80 if login_page is False:
81 return
82
83 galx = self._search_regex(r'(?s)<input.+?name="GALX".+?value="(.+?)"',
84 login_page, 'Login GALX parameter')
85
86 # Log in
87 login_form_strs = {
88 'continue': 'https://www.youtube.com/signin?action_handle_signin=true&feature=sign_in_button&hl=en_US&nomobiletemp=1',
89 'Email': username,
90 'GALX': galx,
91 'Passwd': password,
92
93 'PersistentCookie': 'yes',
94 '_utf8': '霱',
95 'bgresponse': 'js_disabled',
96 'checkConnection': '',
97 'checkedDomains': 'youtube',
98 'dnConn': '',
99 'pstMsg': '0',
100 'rmShown': '1',
101 'secTok': '',
102 'signIn': 'Sign in',
103 'timeStmp': '',
104 'service': 'youtube',
105 'uilel': '3',
106 'hl': 'en_US',
107 }
108
109 # Convert to UTF-8 *before* urlencode because Python 2.x's urlencode
110 # chokes on unicode
111 login_form = dict((k.encode('utf-8'), v.encode('utf-8')) for k, v in login_form_strs.items())
112 login_data = compat_urllib_parse.urlencode(login_form).encode('ascii')
113
114 req = compat_urllib_request.Request(self._LOGIN_URL, login_data)
115 login_results = self._download_webpage(
116 req, None,
117 note='Logging in', errnote='unable to log in', fatal=False)
118 if login_results is False:
119 return False
120
121 if re.search(r'id="errormsg_0_Passwd"', login_results) is not None:
122 raise ExtractorError('Please use your account password and a two-factor code instead of an application-specific password.', expected=True)
123
124 # Two-Factor
125 # TODO add SMS and phone call support - these require making a request and then prompting the user
126
127 if re.search(r'(?i)<form[^>]* id="gaia_secondfactorform"', login_results) is not None:
128 tfa_code = self._get_tfa_info()
129
130 if tfa_code is None:
131 self._downloader.report_warning('Two-factor authentication required. Provide it with --twofactor <code>')
132 self._downloader.report_warning('(Note that only TOTP (Google Authenticator App) codes work at this time.)')
133 return False
134
135 # Unlike the first login form, secTok and timeStmp are both required for the TFA form
136
137 match = re.search(r'id="secTok"\n\s+value=\'(.+)\'/>', login_results, re.M | re.U)
138 if match is None:
139 self._downloader.report_warning('Failed to get secTok - did the page structure change?')
140 secTok = match.group(1)
141 match = re.search(r'id="timeStmp"\n\s+value=\'(.+)\'/>', login_results, re.M | re.U)
142 if match is None:
143 self._downloader.report_warning('Failed to get timeStmp - did the page structure change?')
144 timeStmp = match.group(1)
145
146 tfa_form_strs = {
147 'continue': 'https://www.youtube.com/signin?action_handle_signin=true&feature=sign_in_button&hl=en_US&nomobiletemp=1',
148 'smsToken': '',
149 'smsUserPin': tfa_code,
150 'smsVerifyPin': 'Verify',
151
152 'PersistentCookie': 'yes',
153 'checkConnection': '',
154 'checkedDomains': 'youtube',
155 'pstMsg': '1',
156 'secTok': secTok,
157 'timeStmp': timeStmp,
158 'service': 'youtube',
159 'hl': 'en_US',
160 }
161 tfa_form = dict((k.encode('utf-8'), v.encode('utf-8')) for k, v in tfa_form_strs.items())
162 tfa_data = compat_urllib_parse.urlencode(tfa_form).encode('ascii')
163
164 tfa_req = compat_urllib_request.Request(self._TWOFACTOR_URL, tfa_data)
165 tfa_results = self._download_webpage(
166 tfa_req, None,
167 note='Submitting TFA code', errnote='unable to submit tfa', fatal=False)
168
169 if tfa_results is False:
170 return False
171
172 if re.search(r'(?i)<form[^>]* id="gaia_secondfactorform"', tfa_results) is not None:
173 self._downloader.report_warning('Two-factor code expired. Please try again, or use a one-use backup code instead.')
174 return False
175 if re.search(r'(?i)<form[^>]* id="gaia_loginform"', tfa_results) is not None:
176 self._downloader.report_warning('unable to log in - did the page structure change?')
177 return False
178 if re.search(r'smsauth-interstitial-reviewsettings', tfa_results) is not None:
179 self._downloader.report_warning('Your Google account has a security notice. Please log in on your web browser, resolve the notice, and try again.')
180 return False
181
182 if re.search(r'(?i)<form[^>]* id="gaia_loginform"', login_results) is not None:
183 self._downloader.report_warning('unable to log in: bad username or password')
184 return False
185 return True
186
187 def _real_initialize(self):
188 if self._downloader is None:
189 return
190 self._set_language()
191 if not self._login():
192 return
193
194
195 class YoutubeIE(YoutubeBaseInfoExtractor):
196 IE_DESC = 'YouTube.com'
197 _VALID_URL = r"""(?x)^
198 (
199 (?:https?://|//) # http(s):// or protocol-independent URL
200 (?:(?:(?:(?:\w+\.)?[yY][oO][uU][tT][uU][bB][eE](?:-nocookie)?\.com/|
201 (?:www\.)?deturl\.com/www\.youtube\.com/|
202 (?:www\.)?pwnyoutube\.com/|
203 (?:www\.)?yourepeat\.com/|
204 tube\.majestyc\.net/|
205 youtube\.googleapis\.com/) # the various hostnames, with wildcard subdomains
206 (?:.*?\#/)? # handle anchor (#/) redirect urls
207 (?: # the various things that can precede the ID:
208 (?:(?:v|embed|e)/(?!videoseries)) # v/ or embed/ or e/
209 |(?: # or the v= param in all its forms
210 (?:(?:watch|movie)(?:_popup)?(?:\.php)?/?)? # preceding watch(_popup|.php) or nothing (like /?v=xxxx)
211 (?:\?|\#!?) # the params delimiter ? or # or #!
212 (?:.*?&)? # any other preceding param (like /?s=tuff&v=xxxx)
213 v=
214 )
215 ))
216 |youtu\.be/ # just youtu.be/xxxx
217 |(?:www\.)?cleanvideosearch\.com/media/action/yt/watch\?videoId=
218 )
219 )? # all until now is optional -> you can pass the naked ID
220 ([0-9A-Za-z_-]{11}) # here is it! the YouTube video ID
221 (?!.*?&list=) # combined list/video URLs are handled by the playlist IE
222 (?(1).+)? # if we found the ID, everything can follow
223 $"""
224 _NEXT_URL_RE = r'[\?&]next_url=([^&]+)'
225 _formats = {
226 '5': {'ext': 'flv', 'width': 400, 'height': 240},
227 '6': {'ext': 'flv', 'width': 450, 'height': 270},
228 '13': {'ext': '3gp'},
229 '17': {'ext': '3gp', 'width': 176, 'height': 144},
230 '18': {'ext': 'mp4', 'width': 640, 'height': 360},
231 '22': {'ext': 'mp4', 'width': 1280, 'height': 720},
232 '34': {'ext': 'flv', 'width': 640, 'height': 360},
233 '35': {'ext': 'flv', 'width': 854, 'height': 480},
234 '36': {'ext': '3gp', 'width': 320, 'height': 240},
235 '37': {'ext': 'mp4', 'width': 1920, 'height': 1080},
236 '38': {'ext': 'mp4', 'width': 4096, 'height': 3072},
237 '43': {'ext': 'webm', 'width': 640, 'height': 360},
238 '44': {'ext': 'webm', 'width': 854, 'height': 480},
239 '45': {'ext': 'webm', 'width': 1280, 'height': 720},
240 '46': {'ext': 'webm', 'width': 1920, 'height': 1080},
241 '59': {'ext': 'mp4', 'width': 854, 'height': 480},
242 '78': {'ext': 'mp4', 'width': 854, 'height': 480},
243
244
245 # 3d videos
246 '82': {'ext': 'mp4', 'height': 360, 'format_note': '3D', 'preference': -20},
247 '83': {'ext': 'mp4', 'height': 480, 'format_note': '3D', 'preference': -20},
248 '84': {'ext': 'mp4', 'height': 720, 'format_note': '3D', 'preference': -20},
249 '85': {'ext': 'mp4', 'height': 1080, 'format_note': '3D', 'preference': -20},
250 '100': {'ext': 'webm', 'height': 360, 'format_note': '3D', 'preference': -20},
251 '101': {'ext': 'webm', 'height': 480, 'format_note': '3D', 'preference': -20},
252 '102': {'ext': 'webm', 'height': 720, 'format_note': '3D', 'preference': -20},
253
254 # Apple HTTP Live Streaming
255 '92': {'ext': 'mp4', 'height': 240, 'format_note': 'HLS', 'preference': -10},
256 '93': {'ext': 'mp4', 'height': 360, 'format_note': 'HLS', 'preference': -10},
257 '94': {'ext': 'mp4', 'height': 480, 'format_note': 'HLS', 'preference': -10},
258 '95': {'ext': 'mp4', 'height': 720, 'format_note': 'HLS', 'preference': -10},
259 '96': {'ext': 'mp4', 'height': 1080, 'format_note': 'HLS', 'preference': -10},
260 '132': {'ext': 'mp4', 'height': 240, 'format_note': 'HLS', 'preference': -10},
261 '151': {'ext': 'mp4', 'height': 72, 'format_note': 'HLS', 'preference': -10},
262
263 # DASH mp4 video
264 '133': {'ext': 'mp4', 'height': 240, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
265 '134': {'ext': 'mp4', 'height': 360, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
266 '135': {'ext': 'mp4', 'height': 480, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
267 '136': {'ext': 'mp4', 'height': 720, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
268 '137': {'ext': 'mp4', 'height': 1080, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
269 '138': {'ext': 'mp4', 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40}, # Height can vary (https://github.com/rg3/youtube-dl/issues/4559)
270 '160': {'ext': 'mp4', 'height': 144, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
271 '264': {'ext': 'mp4', 'height': 1440, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
272 '298': {'ext': 'mp4', 'height': 720, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40, 'fps': 60, 'vcodec': 'h264'},
273 '299': {'ext': 'mp4', 'height': 1080, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40, 'fps': 60, 'vcodec': 'h264'},
274 '266': {'ext': 'mp4', 'height': 2160, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40, 'vcodec': 'h264'},
275
276 # Dash mp4 audio
277 '139': {'ext': 'm4a', 'format_note': 'DASH audio', 'acodec': 'aac', 'vcodec': 'none', 'abr': 48, 'preference': -50, 'container': 'm4a_dash'},
278 '140': {'ext': 'm4a', 'format_note': 'DASH audio', 'acodec': 'aac', 'vcodec': 'none', 'abr': 128, 'preference': -50, 'container': 'm4a_dash'},
279 '141': {'ext': 'm4a', 'format_note': 'DASH audio', 'acodec': 'aac', 'vcodec': 'none', 'abr': 256, 'preference': -50, 'container': 'm4a_dash'},
280
281 # Dash webm
282 '167': {'ext': 'webm', 'height': 360, 'width': 640, 'format_note': 'DASH video', 'acodec': 'none', 'container': 'webm', 'vcodec': 'VP8', 'preference': -40},
283 '168': {'ext': 'webm', 'height': 480, 'width': 854, 'format_note': 'DASH video', 'acodec': 'none', 'container': 'webm', 'vcodec': 'VP8', 'preference': -40},
284 '169': {'ext': 'webm', 'height': 720, 'width': 1280, 'format_note': 'DASH video', 'acodec': 'none', 'container': 'webm', 'vcodec': 'VP8', 'preference': -40},
285 '170': {'ext': 'webm', 'height': 1080, 'width': 1920, 'format_note': 'DASH video', 'acodec': 'none', 'container': 'webm', 'vcodec': 'VP8', 'preference': -40},
286 '218': {'ext': 'webm', 'height': 480, 'width': 854, 'format_note': 'DASH video', 'acodec': 'none', 'container': 'webm', 'vcodec': 'VP8', 'preference': -40},
287 '219': {'ext': 'webm', 'height': 480, 'width': 854, 'format_note': 'DASH video', 'acodec': 'none', 'container': 'webm', 'vcodec': 'VP8', 'preference': -40},
288 '278': {'ext': 'webm', 'height': 144, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40, 'container': 'webm', 'vcodec': 'VP9'},
289 '242': {'ext': 'webm', 'height': 240, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
290 '243': {'ext': 'webm', 'height': 360, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
291 '244': {'ext': 'webm', 'height': 480, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
292 '245': {'ext': 'webm', 'height': 480, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
293 '246': {'ext': 'webm', 'height': 480, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
294 '247': {'ext': 'webm', 'height': 720, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
295 '248': {'ext': 'webm', 'height': 1080, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
296 '271': {'ext': 'webm', 'height': 1440, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
297 '272': {'ext': 'webm', 'height': 2160, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
298 '302': {'ext': 'webm', 'height': 720, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40, 'fps': 60, 'vcodec': 'VP9'},
299 '303': {'ext': 'webm', 'height': 1080, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40, 'fps': 60, 'vcodec': 'VP9'},
300 '308': {'ext': 'webm', 'height': 1440, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40, 'fps': 60, 'vcodec': 'VP9'},
301 '313': {'ext': 'webm', 'height': 2160, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40, 'vcodec': 'VP9'},
302 '315': {'ext': 'webm', 'height': 2160, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40, 'fps': 60, 'vcodec': 'VP9'},
303
304 # Dash webm audio
305 '171': {'ext': 'webm', 'vcodec': 'none', 'format_note': 'DASH audio', 'abr': 128, 'preference': -50},
306 '172': {'ext': 'webm', 'vcodec': 'none', 'format_note': 'DASH audio', 'abr': 256, 'preference': -50},
307
308 # Dash webm audio with opus inside
309 '249': {'ext': 'webm', 'vcodec': 'none', 'format_note': 'DASH audio', 'acodec': 'opus', 'abr': 50, 'preference': -50},
310 '250': {'ext': 'webm', 'vcodec': 'none', 'format_note': 'DASH audio', 'acodec': 'opus', 'abr': 70, 'preference': -50},
311 '251': {'ext': 'webm', 'vcodec': 'none', 'format_note': 'DASH audio', 'acodec': 'opus', 'abr': 160, 'preference': -50},
312
313 # RTMP (unnamed)
314 '_rtmp': {'protocol': 'rtmp'},
315 }
316
317 IE_NAME = 'youtube'
318 _TESTS = [
319 {
320 'url': 'http://www.youtube.com/watch?v=BaW_jenozKc',
321 'info_dict': {
322 'id': 'BaW_jenozKc',
323 'ext': 'mp4',
324 'title': 'youtube-dl test video "\'/\\ä↭𝕐',
325 'uploader': 'Philipp Hagemeister',
326 'uploader_id': 'phihag',
327 'upload_date': '20121002',
328 '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 .',
329 'categories': ['Science & Technology'],
330 'like_count': int,
331 'dislike_count': int,
332 }
333 },
334 {
335 'url': 'http://www.youtube.com/watch?v=UxxajLWwzqY',
336 'note': 'Test generic use_cipher_signature video (#897)',
337 'info_dict': {
338 'id': 'UxxajLWwzqY',
339 'ext': 'mp4',
340 'upload_date': '20120506',
341 'title': 'Icona Pop - I Love It (feat. Charli XCX) [OFFICIAL VIDEO]',
342 'description': 'md5:fea86fda2d5a5784273df5c7cc994d9f',
343 'uploader': 'Icona Pop',
344 'uploader_id': 'IconaPop',
345 }
346 },
347 {
348 'url': 'https://www.youtube.com/watch?v=07FYdnEawAQ',
349 'note': 'Test VEVO video with age protection (#956)',
350 'info_dict': {
351 'id': '07FYdnEawAQ',
352 'ext': 'mp4',
353 'upload_date': '20130703',
354 'title': 'Justin Timberlake - Tunnel Vision (Explicit)',
355 'description': 'md5:64249768eec3bc4276236606ea996373',
356 'uploader': 'justintimberlakeVEVO',
357 'uploader_id': 'justintimberlakeVEVO',
358 }
359 },
360 {
361 'url': '//www.YouTube.com/watch?v=yZIXLfi8CZQ',
362 'note': 'Embed-only video (#1746)',
363 'info_dict': {
364 'id': 'yZIXLfi8CZQ',
365 'ext': 'mp4',
366 'upload_date': '20120608',
367 'title': 'Principal Sexually Assaults A Teacher - Episode 117 - 8th June 2012',
368 'description': 'md5:09b78bd971f1e3e289601dfba15ca4f7',
369 'uploader': 'SET India',
370 'uploader_id': 'setindia'
371 }
372 },
373 {
374 'url': 'http://www.youtube.com/watch?v=a9LDPn-MO4I',
375 'note': '256k DASH audio (format 141) via DASH manifest',
376 'info_dict': {
377 'id': 'a9LDPn-MO4I',
378 'ext': 'm4a',
379 'upload_date': '20121002',
380 'uploader_id': '8KVIDEO',
381 'description': '',
382 'uploader': '8KVIDEO',
383 'title': 'UHDTV TEST 8K VIDEO.mp4'
384 },
385 'params': {
386 'youtube_include_dash_manifest': True,
387 'format': '141',
388 },
389 },
390 # DASH manifest with encrypted signature
391 {
392 'url': 'https://www.youtube.com/watch?v=IB3lcPjvWLA',
393 'info_dict': {
394 'id': 'IB3lcPjvWLA',
395 'ext': 'm4a',
396 'title': 'Afrojack, Spree Wilson - The Spark ft. Spree Wilson',
397 'description': 'md5:12e7067fa6735a77bdcbb58cb1187d2d',
398 'uploader': 'AfrojackVEVO',
399 'uploader_id': 'AfrojackVEVO',
400 'upload_date': '20131011',
401 },
402 'params': {
403 'youtube_include_dash_manifest': True,
404 'format': '141',
405 },
406 },
407 # JS player signature function name containing $
408 {
409 'url': 'https://www.youtube.com/watch?v=nfWlot6h_JM',
410 'info_dict': {
411 'id': 'nfWlot6h_JM',
412 'ext': 'm4a',
413 'title': 'Taylor Swift - Shake It Off',
414 'description': 'md5:2acfda1b285bdd478ccec22f9918199d',
415 'uploader': 'TaylorSwiftVEVO',
416 'uploader_id': 'TaylorSwiftVEVO',
417 'upload_date': '20140818',
418 },
419 'params': {
420 'youtube_include_dash_manifest': True,
421 'format': '141',
422 },
423 },
424 # Controversy video
425 {
426 'url': 'https://www.youtube.com/watch?v=T4XJQO3qol8',
427 'info_dict': {
428 'id': 'T4XJQO3qol8',
429 'ext': 'mp4',
430 'upload_date': '20100909',
431 'uploader': 'The Amazing Atheist',
432 'uploader_id': 'TheAmazingAtheist',
433 'title': 'Burning Everyone\'s Koran',
434 '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',
435 }
436 },
437 # Normal age-gate video (No vevo, embed allowed)
438 {
439 'url': 'http://youtube.com/watch?v=HtVdAasjOgU',
440 'info_dict': {
441 'id': 'HtVdAasjOgU',
442 'ext': 'mp4',
443 'title': 'The Witcher 3: Wild Hunt - The Sword Of Destiny Trailer',
444 'description': 're:(?s).{100,}About the Game\n.*?The Witcher 3: Wild Hunt.{100,}',
445 'uploader': 'The Witcher',
446 'uploader_id': 'WitcherGame',
447 'upload_date': '20140605',
448 },
449 },
450 # Age-gate video with encrypted signature
451 {
452 'url': 'http://www.youtube.com/watch?v=6kLq3WMV1nU',
453 'info_dict': {
454 'id': '6kLq3WMV1nU',
455 'ext': 'mp4',
456 'title': 'Dedication To My Ex (Miss That) (Lyric Video)',
457 'description': 'md5:33765bb339e1b47e7e72b5490139bb41',
458 'uploader': 'LloydVEVO',
459 'uploader_id': 'LloydVEVO',
460 'upload_date': '20110629',
461 },
462 },
463 # video_info is None (https://github.com/rg3/youtube-dl/issues/4421)
464 {
465 'url': '__2ABJjxzNo',
466 'info_dict': {
467 'id': '__2ABJjxzNo',
468 'ext': 'mp4',
469 'upload_date': '20100430',
470 'uploader_id': 'deadmau5',
471 'description': 'md5:12c56784b8032162bb936a5f76d55360',
472 'uploader': 'deadmau5',
473 'title': 'Deadmau5 - Some Chords (HD)',
474 },
475 'expected_warnings': [
476 'DASH manifest missing',
477 ]
478 },
479 # Olympics (https://github.com/rg3/youtube-dl/issues/4431)
480 {
481 'url': 'lqQg6PlCWgI',
482 'info_dict': {
483 'id': 'lqQg6PlCWgI',
484 'ext': 'mp4',
485 'upload_date': '20120731',
486 'uploader_id': 'olympic',
487 'description': 'HO09 - Women - GER-AUS - Hockey - 31 July 2012 - London 2012 Olympic Games',
488 'uploader': 'Olympics',
489 'title': 'Hockey - Women - GER-AUS - London 2012 Olympic Games',
490 },
491 'params': {
492 'skip_download': 'requires avconv',
493 }
494 },
495 # Non-square pixels
496 {
497 'url': 'https://www.youtube.com/watch?v=_b-2C3KPAM0',
498 'info_dict': {
499 'id': '_b-2C3KPAM0',
500 'ext': 'mp4',
501 'stretched_ratio': 16 / 9.,
502 'upload_date': '20110310',
503 'uploader_id': 'AllenMeow',
504 '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',
505 'uploader': '孫艾倫',
506 'title': '[A-made] 變態妍字幕版 太妍 我就是這樣的人',
507 },
508 },
509 # url_encoded_fmt_stream_map is empty string
510 {
511 'url': 'qEJwOuvDf7I',
512 'info_dict': {
513 'id': 'qEJwOuvDf7I',
514 'ext': 'mp4',
515 'title': 'Обсуждение судебной практики по выборам 14 сентября 2014 года в Санкт-Петербурге',
516 'description': '',
517 'upload_date': '20150404',
518 'uploader_id': 'spbelect',
519 'uploader': 'Наблюдатели Петербурга',
520 },
521 'params': {
522 'skip_download': 'requires avconv',
523 }
524 },
525 # Extraction from multiple DASH manifests (https://github.com/rg3/youtube-dl/pull/6097)
526 {
527 'url': 'https://www.youtube.com/watch?v=FIl7x6_3R5Y',
528 'info_dict': {
529 'id': 'FIl7x6_3R5Y',
530 'ext': 'mp4',
531 'title': 'md5:7b81415841e02ecd4313668cde88737a',
532 'description': 'md5:116377fd2963b81ec4ce64b542173306',
533 'upload_date': '20150625',
534 'uploader_id': 'dorappi2000',
535 'uploader': 'dorappi2000',
536 'formats': 'mincount:33',
537 },
538 },
539 # DASH manifest with segment_list
540 {
541 'url': 'https://www.youtube.com/embed/CsmdDsKjzN8',
542 'md5': '8ce563a1d667b599d21064e982ab9e31',
543 'info_dict': {
544 'id': 'CsmdDsKjzN8',
545 'ext': 'mp4',
546 'upload_date': '20150501', # According to '<meta itemprop="datePublished"', but in other places it's 20150510
547 'uploader': 'Airtek',
548 'description': 'Retransmisión en directo de la XVIII media maratón de Zaragoza.',
549 'uploader_id': 'UCzTzUmjXxxacNnL8I3m4LnQ',
550 'title': 'Retransmisión XVIII Media maratón Zaragoza 2015',
551 },
552 'params': {
553 'youtube_include_dash_manifest': True,
554 'format': '135', # bestvideo
555 }
556 },
557 ]
558
559 def __init__(self, *args, **kwargs):
560 super(YoutubeIE, self).__init__(*args, **kwargs)
561 self._player_cache = {}
562
563 def report_video_info_webpage_download(self, video_id):
564 """Report attempt to download video info webpage."""
565 self.to_screen('%s: Downloading video info webpage' % video_id)
566
567 def report_information_extraction(self, video_id):
568 """Report attempt to extract video information."""
569 self.to_screen('%s: Extracting video information' % video_id)
570
571 def report_unavailable_format(self, video_id, format):
572 """Report extracted video URL."""
573 self.to_screen('%s: Format %s not available' % (video_id, format))
574
575 def report_rtmp_download(self):
576 """Indicate the download will use the RTMP protocol."""
577 self.to_screen('RTMP download detected')
578
579 def _signature_cache_id(self, example_sig):
580 """ Return a string representation of a signature """
581 return '.'.join(compat_str(len(part)) for part in example_sig.split('.'))
582
583 def _extract_signature_function(self, video_id, player_url, example_sig):
584 id_m = re.match(
585 r'.*?-(?P<id>[a-zA-Z0-9_-]+)(?:/watch_as3|/html5player)?\.(?P<ext>[a-z]+)$',
586 player_url)
587 if not id_m:
588 raise ExtractorError('Cannot identify player %r' % player_url)
589 player_type = id_m.group('ext')
590 player_id = id_m.group('id')
591
592 # Read from filesystem cache
593 func_id = '%s_%s_%s' % (
594 player_type, player_id, self._signature_cache_id(example_sig))
595 assert os.path.basename(func_id) == func_id
596
597 cache_spec = self._downloader.cache.load('youtube-sigfuncs', func_id)
598 if cache_spec is not None:
599 return lambda s: ''.join(s[i] for i in cache_spec)
600
601 download_note = (
602 'Downloading player %s' % player_url
603 if self._downloader.params.get('verbose') else
604 'Downloading %s player %s' % (player_type, player_id)
605 )
606 if player_type == 'js':
607 code = self._download_webpage(
608 player_url, video_id,
609 note=download_note,
610 errnote='Download of %s failed' % player_url)
611 res = self._parse_sig_js(code)
612 elif player_type == 'swf':
613 urlh = self._request_webpage(
614 player_url, video_id,
615 note=download_note,
616 errnote='Download of %s failed' % player_url)
617 code = urlh.read()
618 res = self._parse_sig_swf(code)
619 else:
620 assert False, 'Invalid player type %r' % player_type
621
622 test_string = ''.join(map(compat_chr, range(len(example_sig))))
623 cache_res = res(test_string)
624 cache_spec = [ord(c) for c in cache_res]
625
626 self._downloader.cache.store('youtube-sigfuncs', func_id, cache_spec)
627 return res
628
629 def _print_sig_code(self, func, example_sig):
630 def gen_sig_code(idxs):
631 def _genslice(start, end, step):
632 starts = '' if start == 0 else str(start)
633 ends = (':%d' % (end + step)) if end + step >= 0 else ':'
634 steps = '' if step == 1 else (':%d' % step)
635 return 's[%s%s%s]' % (starts, ends, steps)
636
637 step = None
638 # Quelch pyflakes warnings - start will be set when step is set
639 start = '(Never used)'
640 for i, prev in zip(idxs[1:], idxs[:-1]):
641 if step is not None:
642 if i - prev == step:
643 continue
644 yield _genslice(start, prev, step)
645 step = None
646 continue
647 if i - prev in [-1, 1]:
648 step = i - prev
649 start = prev
650 continue
651 else:
652 yield 's[%d]' % prev
653 if step is None:
654 yield 's[%d]' % i
655 else:
656 yield _genslice(start, i, step)
657
658 test_string = ''.join(map(compat_chr, range(len(example_sig))))
659 cache_res = func(test_string)
660 cache_spec = [ord(c) for c in cache_res]
661 expr_code = ' + '.join(gen_sig_code(cache_spec))
662 signature_id_tuple = '(%s)' % (
663 ', '.join(compat_str(len(p)) for p in example_sig.split('.')))
664 code = ('if tuple(len(p) for p in s.split(\'.\')) == %s:\n'
665 ' return %s\n') % (signature_id_tuple, expr_code)
666 self.to_screen('Extracted signature function:\n' + code)
667
668 def _parse_sig_js(self, jscode):
669 funcname = self._search_regex(
670 r'\.sig\|\|([a-zA-Z0-9$]+)\(', jscode,
671 'Initial JS player signature function name')
672
673 jsi = JSInterpreter(jscode)
674 initial_function = jsi.extract_function(funcname)
675 return lambda s: initial_function([s])
676
677 def _parse_sig_swf(self, file_contents):
678 swfi = SWFInterpreter(file_contents)
679 TARGET_CLASSNAME = 'SignatureDecipher'
680 searched_class = swfi.extract_class(TARGET_CLASSNAME)
681 initial_function = swfi.extract_function(searched_class, 'decipher')
682 return lambda s: initial_function([s])
683
684 def _decrypt_signature(self, s, video_id, player_url, age_gate=False):
685 """Turn the encrypted s field into a working signature"""
686
687 if player_url is None:
688 raise ExtractorError('Cannot decrypt signature without player_url')
689
690 if player_url.startswith('//'):
691 player_url = 'https:' + player_url
692 try:
693 player_id = (player_url, self._signature_cache_id(s))
694 if player_id not in self._player_cache:
695 func = self._extract_signature_function(
696 video_id, player_url, s
697 )
698 self._player_cache[player_id] = func
699 func = self._player_cache[player_id]
700 if self._downloader.params.get('youtube_print_sig_code'):
701 self._print_sig_code(func, s)
702 return func(s)
703 except Exception as e:
704 tb = traceback.format_exc()
705 raise ExtractorError(
706 'Signature extraction failed: ' + tb, cause=e)
707
708 def _get_subtitles(self, video_id, webpage):
709 try:
710 subs_doc = self._download_xml(
711 'https://video.google.com/timedtext?hl=en&type=list&v=%s' % video_id,
712 video_id, note=False)
713 except ExtractorError as err:
714 self._downloader.report_warning('unable to download video subtitles: %s' % compat_str(err))
715 return {}
716
717 sub_lang_list = {}
718 for track in subs_doc.findall('track'):
719 lang = track.attrib['lang_code']
720 if lang in sub_lang_list:
721 continue
722 sub_formats = []
723 for ext in ['sbv', 'vtt', 'srt']:
724 params = compat_urllib_parse.urlencode({
725 'lang': lang,
726 'v': video_id,
727 'fmt': ext,
728 'name': track.attrib['name'].encode('utf-8'),
729 })
730 sub_formats.append({
731 'url': 'https://www.youtube.com/api/timedtext?' + params,
732 'ext': ext,
733 })
734 sub_lang_list[lang] = sub_formats
735 if not sub_lang_list:
736 self._downloader.report_warning('video doesn\'t have subtitles')
737 return {}
738 return sub_lang_list
739
740 def _get_automatic_captions(self, video_id, webpage):
741 """We need the webpage for getting the captions url, pass it as an
742 argument to speed up the process."""
743 self.to_screen('%s: Looking for automatic captions' % video_id)
744 mobj = re.search(r';ytplayer.config = ({.*?});', webpage)
745 err_msg = 'Couldn\'t find automatic captions for %s' % video_id
746 if mobj is None:
747 self._downloader.report_warning(err_msg)
748 return {}
749 player_config = json.loads(mobj.group(1))
750 try:
751 args = player_config['args']
752 caption_url = args['ttsurl']
753 timestamp = args['timestamp']
754 # We get the available subtitles
755 list_params = compat_urllib_parse.urlencode({
756 'type': 'list',
757 'tlangs': 1,
758 'asrs': 1,
759 })
760 list_url = caption_url + '&' + list_params
761 caption_list = self._download_xml(list_url, video_id)
762 original_lang_node = caption_list.find('track')
763 if original_lang_node is None:
764 self._downloader.report_warning('Video doesn\'t have automatic captions')
765 return {}
766 original_lang = original_lang_node.attrib['lang_code']
767 caption_kind = original_lang_node.attrib.get('kind', '')
768
769 sub_lang_list = {}
770 for lang_node in caption_list.findall('target'):
771 sub_lang = lang_node.attrib['lang_code']
772 sub_formats = []
773 for ext in ['sbv', 'vtt', 'srt']:
774 params = compat_urllib_parse.urlencode({
775 'lang': original_lang,
776 'tlang': sub_lang,
777 'fmt': ext,
778 'ts': timestamp,
779 'kind': caption_kind,
780 })
781 sub_formats.append({
782 'url': caption_url + '&' + params,
783 'ext': ext,
784 })
785 sub_lang_list[sub_lang] = sub_formats
786 return sub_lang_list
787 # An extractor error can be raise by the download process if there are
788 # no automatic captions but there are subtitles
789 except (KeyError, ExtractorError):
790 self._downloader.report_warning(err_msg)
791 return {}
792
793 @classmethod
794 def extract_id(cls, url):
795 mobj = re.match(cls._VALID_URL, url, re.VERBOSE)
796 if mobj is None:
797 raise ExtractorError('Invalid URL: %s' % url)
798 video_id = mobj.group(2)
799 return video_id
800
801 def _extract_from_m3u8(self, manifest_url, video_id):
802 url_map = {}
803
804 def _get_urls(_manifest):
805 lines = _manifest.split('\n')
806 urls = filter(lambda l: l and not l.startswith('#'),
807 lines)
808 return urls
809 manifest = self._download_webpage(manifest_url, video_id, 'Downloading formats manifest')
810 formats_urls = _get_urls(manifest)
811 for format_url in formats_urls:
812 itag = self._search_regex(r'itag/(\d+?)/', format_url, 'itag')
813 url_map[itag] = format_url
814 return url_map
815
816 def _extract_annotations(self, video_id):
817 url = 'https://www.youtube.com/annotations_invideo?features=1&legacy=1&video_id=%s' % video_id
818 return self._download_webpage(url, video_id, note='Searching for annotations.', errnote='Unable to download video annotations.')
819
820 def _parse_dash_manifest(
821 self, video_id, dash_manifest_url, player_url, age_gate, fatal=True):
822 def decrypt_sig(mobj):
823 s = mobj.group(1)
824 dec_s = self._decrypt_signature(s, video_id, player_url, age_gate)
825 return '/signature/%s' % dec_s
826 dash_manifest_url = re.sub(r'/s/([a-fA-F0-9\.]+)', decrypt_sig, dash_manifest_url)
827 dash_doc = self._download_xml(
828 dash_manifest_url, video_id,
829 note='Downloading DASH manifest',
830 errnote='Could not download DASH manifest',
831 fatal=fatal)
832
833 if dash_doc is False:
834 return []
835
836 formats = []
837 for a in dash_doc.findall('.//{urn:mpeg:DASH:schema:MPD:2011}AdaptationSet'):
838 mime_type = a.attrib.get('mimeType')
839 for r in a.findall('{urn:mpeg:DASH:schema:MPD:2011}Representation'):
840 url_el = r.find('{urn:mpeg:DASH:schema:MPD:2011}BaseURL')
841 if url_el is None:
842 continue
843 if mime_type == 'text/vtt':
844 # TODO implement WebVTT downloading
845 pass
846 elif mime_type.startswith('audio/') or mime_type.startswith('video/'):
847 segment_list = r.find('{urn:mpeg:DASH:schema:MPD:2011}SegmentList')
848 format_id = r.attrib['id']
849 video_url = url_el.text
850 filesize = int_or_none(url_el.attrib.get('{http://youtube.com/yt/2012/10/10}contentLength'))
851 f = {
852 'format_id': format_id,
853 'url': video_url,
854 'width': int_or_none(r.attrib.get('width')),
855 'height': int_or_none(r.attrib.get('height')),
856 'tbr': int_or_none(r.attrib.get('bandwidth'), 1000),
857 'asr': int_or_none(r.attrib.get('audioSamplingRate')),
858 'filesize': filesize,
859 'fps': int_or_none(r.attrib.get('frameRate')),
860 }
861 if segment_list is not None:
862 f.update({
863 'initialization_url': segment_list.find('{urn:mpeg:DASH:schema:MPD:2011}Initialization').attrib['sourceURL'],
864 'segment_urls': [segment.attrib.get('media') for segment in segment_list.findall('{urn:mpeg:DASH:schema:MPD:2011}SegmentURL')],
865 'protocol': 'http_dash_segments',
866 })
867 try:
868 existing_format = next(
869 fo for fo in formats
870 if fo['format_id'] == format_id)
871 except StopIteration:
872 full_info = self._formats.get(format_id, {}).copy()
873 full_info.update(f)
874 codecs = r.attrib.get('codecs')
875 if codecs:
876 if full_info.get('acodec') == 'none' and 'vcodec' not in full_info:
877 full_info['vcodec'] = codecs
878 elif full_info.get('vcodec') == 'none' and 'acodec' not in full_info:
879 full_info['acodec'] = codecs
880 formats.append(full_info)
881 else:
882 existing_format.update(f)
883 else:
884 self.report_warning('Unknown MIME type %s in DASH manifest' % mime_type)
885 return formats
886
887 def _real_extract(self, url):
888 proto = (
889 'http' if self._downloader.params.get('prefer_insecure', False)
890 else 'https')
891
892 # Extract original video URL from URL with redirection, like age verification, using next_url parameter
893 mobj = re.search(self._NEXT_URL_RE, url)
894 if mobj:
895 url = proto + '://www.youtube.com/' + compat_urllib_parse_unquote(mobj.group(1)).lstrip('/')
896 video_id = self.extract_id(url)
897
898 # Get video webpage
899 url = proto + '://www.youtube.com/watch?v=%s&gl=US&hl=en&has_verified=1&bpctr=9999999999' % video_id
900 video_webpage = self._download_webpage(url, video_id)
901
902 # Attempt to extract SWF player URL
903 mobj = re.search(r'swfConfig.*?"(https?:\\/\\/.*?watch.*?-.*?\.swf)"', video_webpage)
904 if mobj is not None:
905 player_url = re.sub(r'\\(.)', r'\1', mobj.group(1))
906 else:
907 player_url = None
908
909 dash_mpds = []
910
911 def add_dash_mpd(video_info):
912 dash_mpd = video_info.get('dashmpd')
913 if dash_mpd and dash_mpd[0] not in dash_mpds:
914 dash_mpds.append(dash_mpd[0])
915
916 # Get video info
917 embed_webpage = None
918 is_live = None
919 if re.search(r'player-age-gate-content">', video_webpage) is not None:
920 age_gate = True
921 # We simulate the access to the video from www.youtube.com/v/{video_id}
922 # this can be viewed without login into Youtube
923 url = proto + '://www.youtube.com/embed/%s' % video_id
924 embed_webpage = self._download_webpage(url, video_id, 'Downloading embed webpage')
925 data = compat_urllib_parse.urlencode({
926 'video_id': video_id,
927 'eurl': 'https://youtube.googleapis.com/v/' + video_id,
928 'sts': self._search_regex(
929 r'"sts"\s*:\s*(\d+)', embed_webpage, 'sts', default=''),
930 })
931 video_info_url = proto + '://www.youtube.com/get_video_info?' + data
932 video_info_webpage = self._download_webpage(
933 video_info_url, video_id,
934 note='Refetching age-gated info webpage',
935 errnote='unable to download video info webpage')
936 video_info = compat_parse_qs(video_info_webpage)
937 add_dash_mpd(video_info)
938 else:
939 age_gate = False
940 video_info = None
941 # Try looking directly into the video webpage
942 mobj = re.search(r';ytplayer\.config\s*=\s*({.*?});', video_webpage)
943 if mobj:
944 json_code = uppercase_escape(mobj.group(1))
945 ytplayer_config = json.loads(json_code)
946 args = ytplayer_config['args']
947 if args.get('url_encoded_fmt_stream_map'):
948 # Convert to the same format returned by compat_parse_qs
949 video_info = dict((k, [v]) for k, v in args.items())
950 add_dash_mpd(video_info)
951 if args.get('livestream') == '1' or args.get('live_playback') == 1:
952 is_live = True
953 if not video_info or self._downloader.params.get('youtube_include_dash_manifest', True):
954 # We also try looking in get_video_info since it may contain different dashmpd
955 # URL that points to a DASH manifest with possibly different itag set (some itags
956 # are missing from DASH manifest pointed by webpage's dashmpd, some - from DASH
957 # manifest pointed by get_video_info's dashmpd).
958 # The general idea is to take a union of itags of both DASH manifests (for example
959 # video with such 'manifest behavior' see https://github.com/rg3/youtube-dl/issues/6093)
960 self.report_video_info_webpage_download(video_id)
961 for el_type in ['&el=info', '&el=embedded', '&el=detailpage', '&el=vevo', '']:
962 video_info_url = (
963 '%s://www.youtube.com/get_video_info?&video_id=%s%s&ps=default&eurl=&gl=US&hl=en'
964 % (proto, video_id, el_type))
965 video_info_webpage = self._download_webpage(
966 video_info_url,
967 video_id, note=False,
968 errnote='unable to download video info webpage')
969 get_video_info = compat_parse_qs(video_info_webpage)
970 add_dash_mpd(get_video_info)
971 if not video_info:
972 video_info = get_video_info
973 if 'token' in get_video_info:
974 break
975 if 'token' not in video_info:
976 if 'reason' in video_info:
977 if 'The uploader has not made this video available in your country.' in video_info['reason']:
978 regions_allowed = self._html_search_meta('regionsAllowed', video_webpage, default=None)
979 if regions_allowed is not None:
980 raise ExtractorError('YouTube said: This video is available in %s only' % (
981 ', '.join(map(ISO3166Utils.short2full, regions_allowed.split(',')))),
982 expected=True)
983 raise ExtractorError(
984 'YouTube said: %s' % video_info['reason'][0],
985 expected=True, video_id=video_id)
986 else:
987 raise ExtractorError(
988 '"token" parameter not in video info for unknown reason',
989 video_id=video_id)
990
991 if 'view_count' in video_info:
992 view_count = int(video_info['view_count'][0])
993 else:
994 view_count = None
995
996 # Check for "rental" videos
997 if 'ypc_video_rental_bar_text' in video_info and 'author' not in video_info:
998 raise ExtractorError('"rental" videos not supported')
999
1000 # Start extracting information
1001 self.report_information_extraction(video_id)
1002
1003 # uploader
1004 if 'author' not in video_info:
1005 raise ExtractorError('Unable to extract uploader name')
1006 video_uploader = compat_urllib_parse_unquote_plus(video_info['author'][0])
1007
1008 # uploader_id
1009 video_uploader_id = None
1010 mobj = re.search(r'<link itemprop="url" href="http://www.youtube.com/(?:user|channel)/([^"]+)">', video_webpage)
1011 if mobj is not None:
1012 video_uploader_id = mobj.group(1)
1013 else:
1014 self._downloader.report_warning('unable to extract uploader nickname')
1015
1016 # title
1017 if 'title' in video_info:
1018 video_title = video_info['title'][0]
1019 else:
1020 self._downloader.report_warning('Unable to extract video title')
1021 video_title = '_'
1022
1023 # thumbnail image
1024 # We try first to get a high quality image:
1025 m_thumb = re.search(r'<span itemprop="thumbnail".*?href="(.*?)">',
1026 video_webpage, re.DOTALL)
1027 if m_thumb is not None:
1028 video_thumbnail = m_thumb.group(1)
1029 elif 'thumbnail_url' not in video_info:
1030 self._downloader.report_warning('unable to extract video thumbnail')
1031 video_thumbnail = None
1032 else: # don't panic if we can't find it
1033 video_thumbnail = compat_urllib_parse_unquote_plus(video_info['thumbnail_url'][0])
1034
1035 # upload date
1036 upload_date = self._html_search_meta(
1037 'datePublished', video_webpage, 'upload date', default=None)
1038 if not upload_date:
1039 upload_date = self._search_regex(
1040 [r'(?s)id="eow-date.*?>(.*?)</span>',
1041 r'id="watch-uploader-info".*?>.*?(?:Published|Uploaded|Streamed live|Started) on (.+?)</strong>'],
1042 video_webpage, 'upload date', default=None)
1043 if upload_date:
1044 upload_date = ' '.join(re.sub(r'[/,-]', r' ', mobj.group(1)).split())
1045 upload_date = unified_strdate(upload_date)
1046
1047 m_cat_container = self._search_regex(
1048 r'(?s)<h4[^>]*>\s*Category\s*</h4>\s*<ul[^>]*>(.*?)</ul>',
1049 video_webpage, 'categories', default=None)
1050 if m_cat_container:
1051 category = self._html_search_regex(
1052 r'(?s)<a[^<]+>(.*?)</a>', m_cat_container, 'category',
1053 default=None)
1054 video_categories = None if category is None else [category]
1055 else:
1056 video_categories = None
1057
1058 # description
1059 video_description = get_element_by_id("eow-description", video_webpage)
1060 if video_description:
1061 video_description = re.sub(r'''(?x)
1062 <a\s+
1063 (?:[a-zA-Z-]+="[^"]+"\s+)*?
1064 title="([^"]+)"\s+
1065 (?:[a-zA-Z-]+="[^"]+"\s+)*?
1066 class="yt-uix-redirect-link"\s*>
1067 [^<]+
1068 </a>
1069 ''', r'\1', video_description)
1070 video_description = clean_html(video_description)
1071 else:
1072 fd_mobj = re.search(r'<meta name="description" content="([^"]+)"', video_webpage)
1073 if fd_mobj:
1074 video_description = unescapeHTML(fd_mobj.group(1))
1075 else:
1076 video_description = ''
1077
1078 def _extract_count(count_name):
1079 return str_to_int(self._search_regex(
1080 r'-%s-button[^>]+><span[^>]+class="yt-uix-button-content"[^>]*>([\d,]+)</span>'
1081 % re.escape(count_name),
1082 video_webpage, count_name, default=None))
1083
1084 like_count = _extract_count('like')
1085 dislike_count = _extract_count('dislike')
1086
1087 # subtitles
1088 video_subtitles = self.extract_subtitles(video_id, video_webpage)
1089 automatic_captions = self.extract_automatic_captions(video_id, video_webpage)
1090
1091 if 'length_seconds' not in video_info:
1092 self._downloader.report_warning('unable to extract video duration')
1093 video_duration = None
1094 else:
1095 video_duration = int(compat_urllib_parse_unquote_plus(video_info['length_seconds'][0]))
1096
1097 # annotations
1098 video_annotations = None
1099 if self._downloader.params.get('writeannotations', False):
1100 video_annotations = self._extract_annotations(video_id)
1101
1102 def _map_to_format_list(urlmap):
1103 formats = []
1104 for itag, video_real_url in urlmap.items():
1105 dct = {
1106 'format_id': itag,
1107 'url': video_real_url,
1108 'player_url': player_url,
1109 }
1110 if itag in self._formats:
1111 dct.update(self._formats[itag])
1112 formats.append(dct)
1113 return formats
1114
1115 if 'conn' in video_info and video_info['conn'][0].startswith('rtmp'):
1116 self.report_rtmp_download()
1117 formats = [{
1118 'format_id': '_rtmp',
1119 'protocol': 'rtmp',
1120 'url': video_info['conn'][0],
1121 'player_url': player_url,
1122 }]
1123 elif len(video_info.get('url_encoded_fmt_stream_map', [''])[0]) >= 1 or len(video_info.get('adaptive_fmts', [''])[0]) >= 1:
1124 encoded_url_map = video_info.get('url_encoded_fmt_stream_map', [''])[0] + ',' + video_info.get('adaptive_fmts', [''])[0]
1125 if 'rtmpe%3Dyes' in encoded_url_map:
1126 raise ExtractorError('rtmpe downloads are not supported, see https://github.com/rg3/youtube-dl/issues/343 for more information.', expected=True)
1127 url_map = {}
1128 for url_data_str in encoded_url_map.split(','):
1129 url_data = compat_parse_qs(url_data_str)
1130 if 'itag' not in url_data or 'url' not in url_data:
1131 continue
1132 format_id = url_data['itag'][0]
1133 url = url_data['url'][0]
1134
1135 if 'sig' in url_data:
1136 url += '&signature=' + url_data['sig'][0]
1137 elif 's' in url_data:
1138 encrypted_sig = url_data['s'][0]
1139 ASSETS_RE = r'"assets":.+?"js":\s*("[^"]+")'
1140
1141 jsplayer_url_json = self._search_regex(
1142 ASSETS_RE,
1143 embed_webpage if age_gate else video_webpage,
1144 'JS player URL (1)', default=None)
1145 if not jsplayer_url_json and not age_gate:
1146 # We need the embed website after all
1147 if embed_webpage is None:
1148 embed_url = proto + '://www.youtube.com/embed/%s' % video_id
1149 embed_webpage = self._download_webpage(
1150 embed_url, video_id, 'Downloading embed webpage')
1151 jsplayer_url_json = self._search_regex(
1152 ASSETS_RE, embed_webpage, 'JS player URL')
1153
1154 player_url = json.loads(jsplayer_url_json)
1155 if player_url is None:
1156 player_url_json = self._search_regex(
1157 r'ytplayer\.config.*?"url"\s*:\s*("[^"]+")',
1158 video_webpage, 'age gate player URL')
1159 player_url = json.loads(player_url_json)
1160
1161 if self._downloader.params.get('verbose'):
1162 if player_url is None:
1163 player_version = 'unknown'
1164 player_desc = 'unknown'
1165 else:
1166 if player_url.endswith('swf'):
1167 player_version = self._search_regex(
1168 r'-(.+?)(?:/watch_as3)?\.swf$', player_url,
1169 'flash player', fatal=False)
1170 player_desc = 'flash player %s' % player_version
1171 else:
1172 player_version = self._search_regex(
1173 r'html5player-([^/]+?)(?:/html5player)?\.js',
1174 player_url,
1175 'html5 player', fatal=False)
1176 player_desc = 'html5 player %s' % player_version
1177
1178 parts_sizes = self._signature_cache_id(encrypted_sig)
1179 self.to_screen('{%s} signature length %s, %s' %
1180 (format_id, parts_sizes, player_desc))
1181
1182 signature = self._decrypt_signature(
1183 encrypted_sig, video_id, player_url, age_gate)
1184 url += '&signature=' + signature
1185 if 'ratebypass' not in url:
1186 url += '&ratebypass=yes'
1187 url_map[format_id] = url
1188 formats = _map_to_format_list(url_map)
1189 elif video_info.get('hlsvp'):
1190 manifest_url = video_info['hlsvp'][0]
1191 url_map = self._extract_from_m3u8(manifest_url, video_id)
1192 formats = _map_to_format_list(url_map)
1193 else:
1194 raise ExtractorError('no conn, hlsvp or url_encoded_fmt_stream_map information found in video info')
1195
1196 # Look for the DASH manifest
1197 if self._downloader.params.get('youtube_include_dash_manifest', True):
1198 dash_mpd_fatal = True
1199 for dash_manifest_url in dash_mpds:
1200 dash_formats = {}
1201 try:
1202 for df in self._parse_dash_manifest(
1203 video_id, dash_manifest_url, player_url, age_gate, dash_mpd_fatal):
1204 # Do not overwrite DASH format found in some previous DASH manifest
1205 if df['format_id'] not in dash_formats:
1206 dash_formats[df['format_id']] = df
1207 # Additional DASH manifests may end up in HTTP Error 403 therefore
1208 # allow them to fail without bug report message if we already have
1209 # some DASH manifest succeeded. This is temporary workaround to reduce
1210 # burst of bug reports until we figure out the reason and whether it
1211 # can be fixed at all.
1212 dash_mpd_fatal = False
1213 except (ExtractorError, KeyError) as e:
1214 self.report_warning(
1215 'Skipping DASH manifest: %r' % e, video_id)
1216 if dash_formats:
1217 # Remove the formats we found through non-DASH, they
1218 # contain less info and it can be wrong, because we use
1219 # fixed values (for example the resolution). See
1220 # https://github.com/rg3/youtube-dl/issues/5774 for an
1221 # example.
1222 formats = [f for f in formats if f['format_id'] not in dash_formats.keys()]
1223 formats.extend(dash_formats.values())
1224
1225 # Check for malformed aspect ratio
1226 stretched_m = re.search(
1227 r'<meta\s+property="og:video:tag".*?content="yt:stretch=(?P<w>[0-9]+):(?P<h>[0-9]+)">',
1228 video_webpage)
1229 if stretched_m:
1230 ratio = float(stretched_m.group('w')) / float(stretched_m.group('h'))
1231 for f in formats:
1232 if f.get('vcodec') != 'none':
1233 f['stretched_ratio'] = ratio
1234
1235 self._sort_formats(formats)
1236
1237 return {
1238 'id': video_id,
1239 'uploader': video_uploader,
1240 'uploader_id': video_uploader_id,
1241 'upload_date': upload_date,
1242 'title': video_title,
1243 'thumbnail': video_thumbnail,
1244 'description': video_description,
1245 'categories': video_categories,
1246 'subtitles': video_subtitles,
1247 'automatic_captions': automatic_captions,
1248 'duration': video_duration,
1249 'age_limit': 18 if age_gate else 0,
1250 'annotations': video_annotations,
1251 'webpage_url': proto + '://www.youtube.com/watch?v=%s' % video_id,
1252 'view_count': view_count,
1253 'like_count': like_count,
1254 'dislike_count': dislike_count,
1255 'average_rating': float_or_none(video_info.get('avg_rating', [None])[0]),
1256 'formats': formats,
1257 'is_live': is_live,
1258 }
1259
1260
1261 class YoutubePlaylistIE(YoutubeBaseInfoExtractor):
1262 IE_DESC = 'YouTube.com playlists'
1263 _VALID_URL = r"""(?x)(?:
1264 (?:https?://)?
1265 (?:\w+\.)?
1266 youtube\.com/
1267 (?:
1268 (?:course|view_play_list|my_playlists|artist|playlist|watch|embed/videoseries)
1269 \? (?:.*?&)*? (?:p|a|list)=
1270 | p/
1271 )
1272 (
1273 (?:PL|LL|EC|UU|FL|RD|UL)?[0-9A-Za-z-_]{10,}
1274 # Top tracks, they can also include dots
1275 |(?:MC)[\w\.]*
1276 )
1277 .*
1278 |
1279 ((?:PL|LL|EC|UU|FL|RD|UL)[0-9A-Za-z-_]{10,})
1280 )"""
1281 _TEMPLATE_URL = 'https://www.youtube.com/playlist?list=%s'
1282 _VIDEO_RE = r'href="\s*/watch\?v=(?P<id>[0-9A-Za-z_-]{11})&amp;[^"]*?index=(?P<index>\d+)'
1283 IE_NAME = 'youtube:playlist'
1284 _TESTS = [{
1285 'url': 'https://www.youtube.com/playlist?list=PLwiyx1dc3P2JR9N8gQaQN_BCvlSlap7re',
1286 'info_dict': {
1287 'title': 'ytdl test PL',
1288 'id': 'PLwiyx1dc3P2JR9N8gQaQN_BCvlSlap7re',
1289 },
1290 'playlist_count': 3,
1291 }, {
1292 'url': 'https://www.youtube.com/playlist?list=PLtPgu7CB4gbZDA7i_euNxn75ISqxwZPYx',
1293 'info_dict': {
1294 'id': 'PLtPgu7CB4gbZDA7i_euNxn75ISqxwZPYx',
1295 'title': 'YDL_Empty_List',
1296 },
1297 'playlist_count': 0,
1298 }, {
1299 'note': 'Playlist with deleted videos (#651). As a bonus, the video #51 is also twice in this list.',
1300 'url': 'https://www.youtube.com/playlist?list=PLwP_SiAcdui0KVebT0mU9Apz359a4ubsC',
1301 'info_dict': {
1302 'title': '29C3: Not my department',
1303 'id': 'PLwP_SiAcdui0KVebT0mU9Apz359a4ubsC',
1304 },
1305 'playlist_count': 95,
1306 }, {
1307 'note': 'issue #673',
1308 'url': 'PLBB231211A4F62143',
1309 'info_dict': {
1310 'title': '[OLD]Team Fortress 2 (Class-based LP)',
1311 'id': 'PLBB231211A4F62143',
1312 },
1313 'playlist_mincount': 26,
1314 }, {
1315 'note': 'Large playlist',
1316 'url': 'https://www.youtube.com/playlist?list=UUBABnxM4Ar9ten8Mdjj1j0Q',
1317 'info_dict': {
1318 'title': 'Uploads from Cauchemar',
1319 'id': 'UUBABnxM4Ar9ten8Mdjj1j0Q',
1320 },
1321 'playlist_mincount': 799,
1322 }, {
1323 'url': 'PLtPgu7CB4gbY9oDN3drwC3cMbJggS7dKl',
1324 'info_dict': {
1325 'title': 'YDL_safe_search',
1326 'id': 'PLtPgu7CB4gbY9oDN3drwC3cMbJggS7dKl',
1327 },
1328 'playlist_count': 2,
1329 }, {
1330 'note': 'embedded',
1331 'url': 'http://www.youtube.com/embed/videoseries?list=PL6IaIsEjSbf96XFRuNccS_RuEXwNdsoEu',
1332 'playlist_count': 4,
1333 'info_dict': {
1334 'title': 'JODA15',
1335 'id': 'PL6IaIsEjSbf96XFRuNccS_RuEXwNdsoEu',
1336 }
1337 }, {
1338 'note': 'Embedded SWF player',
1339 'url': 'http://www.youtube.com/p/YN5VISEtHet5D4NEvfTd0zcgFk84NqFZ?hl=en_US&fs=1&rel=0',
1340 'playlist_count': 4,
1341 'info_dict': {
1342 'title': 'JODA7',
1343 'id': 'YN5VISEtHet5D4NEvfTd0zcgFk84NqFZ',
1344 }
1345 }, {
1346 'note': 'Buggy playlist: the webpage has a "Load more" button but it doesn\'t have more videos',
1347 'url': 'https://www.youtube.com/playlist?list=UUXw-G3eDE9trcvY2sBMM_aA',
1348 'info_dict': {
1349 'title': 'Uploads from Interstellar Movie',
1350 'id': 'UUXw-G3eDE9trcvY2sBMM_aA',
1351 },
1352 'playlist_mincout': 21,
1353 }]
1354
1355 def _real_initialize(self):
1356 self._login()
1357
1358 def _extract_mix(self, playlist_id):
1359 # The mixes are generated from a single video
1360 # the id of the playlist is just 'RD' + video_id
1361 url = 'https://youtube.com/watch?v=%s&list=%s' % (playlist_id[-11:], playlist_id)
1362 webpage = self._download_webpage(
1363 url, playlist_id, 'Downloading Youtube mix')
1364 search_title = lambda class_name: get_element_by_attribute('class', class_name, webpage)
1365 title_span = (
1366 search_title('playlist-title') or
1367 search_title('title long-title') or
1368 search_title('title'))
1369 title = clean_html(title_span)
1370 ids = orderedSet(re.findall(
1371 r'''(?xs)data-video-username=".*?".*?
1372 href="/watch\?v=([0-9A-Za-z_-]{11})&amp;[^"]*?list=%s''' % re.escape(playlist_id),
1373 webpage))
1374 url_results = self._ids_to_results(ids)
1375
1376 return self.playlist_result(url_results, playlist_id, title)
1377
1378 def _extract_playlist(self, playlist_id):
1379 url = self._TEMPLATE_URL % playlist_id
1380 page = self._download_webpage(url, playlist_id)
1381
1382 for match in re.findall(r'<div class="yt-alert-message">([^<]+)</div>', page):
1383 match = match.strip()
1384 # Check if the playlist exists or is private
1385 if re.match(r'[^<]*(The|This) playlist (does not exist|is private)[^<]*', match):
1386 raise ExtractorError(
1387 'The playlist doesn\'t exist or is private, use --username or '
1388 '--netrc to access it.',
1389 expected=True)
1390 elif re.match(r'[^<]*Invalid parameters[^<]*', match):
1391 raise ExtractorError(
1392 'Invalid parameters. Maybe URL is incorrect.',
1393 expected=True)
1394 elif re.match(r'[^<]*Choose your language[^<]*', match):
1395 continue
1396 else:
1397 self.report_warning('Youtube gives an alert message: ' + match)
1398
1399 # Extract the video ids from the playlist pages
1400 def _entries():
1401 more_widget_html = content_html = page
1402 for page_num in itertools.count(1):
1403 matches = re.finditer(self._VIDEO_RE, content_html)
1404 # We remove the duplicates and the link with index 0
1405 # (it's not the first video of the playlist)
1406 new_ids = orderedSet(m.group('id') for m in matches if m.group('index') != '0')
1407 for vid_id in new_ids:
1408 yield self.url_result(vid_id, 'Youtube', video_id=vid_id)
1409
1410 mobj = re.search(r'data-uix-load-more-href="/?(?P<more>[^"]+)"', more_widget_html)
1411 if not mobj:
1412 break
1413
1414 more = self._download_json(
1415 'https://youtube.com/%s' % mobj.group('more'), playlist_id,
1416 'Downloading page #%s' % page_num,
1417 transform_source=uppercase_escape)
1418 content_html = more['content_html']
1419 if not content_html.strip():
1420 # Some webpages show a "Load more" button but they don't
1421 # have more videos
1422 break
1423 more_widget_html = more['load_more_widget_html']
1424
1425 playlist_title = self._html_search_regex(
1426 r'(?s)<h1 class="pl-header-title[^"]*">\s*(.*?)\s*</h1>',
1427 page, 'title')
1428
1429 return self.playlist_result(_entries(), playlist_id, playlist_title)
1430
1431 def _real_extract(self, url):
1432 # Extract playlist id
1433 mobj = re.match(self._VALID_URL, url)
1434 if mobj is None:
1435 raise ExtractorError('Invalid URL: %s' % url)
1436 playlist_id = mobj.group(1) or mobj.group(2)
1437
1438 # Check if it's a video-specific URL
1439 query_dict = compat_urlparse.parse_qs(compat_urlparse.urlparse(url).query)
1440 if 'v' in query_dict:
1441 video_id = query_dict['v'][0]
1442 if self._downloader.params.get('noplaylist'):
1443 self.to_screen('Downloading just video %s because of --no-playlist' % video_id)
1444 return self.url_result(video_id, 'Youtube', video_id=video_id)
1445 else:
1446 self.to_screen('Downloading playlist %s - add --no-playlist to just download video %s' % (playlist_id, video_id))
1447
1448 if playlist_id.startswith('RD') or playlist_id.startswith('UL'):
1449 # Mixes require a custom extraction process
1450 return self._extract_mix(playlist_id)
1451
1452 return self._extract_playlist(playlist_id)
1453
1454
1455 class YoutubeChannelIE(InfoExtractor):
1456 IE_DESC = 'YouTube.com channels'
1457 _VALID_URL = r'https?://(?:youtu\.be|(?:\w+\.)?youtube(?:-nocookie)?\.com)/channel/(?P<id>[0-9A-Za-z_-]+)'
1458 _TEMPLATE_URL = 'https://www.youtube.com/channel/%s/videos'
1459 IE_NAME = 'youtube:channel'
1460 _TESTS = [{
1461 'note': 'paginated channel',
1462 'url': 'https://www.youtube.com/channel/UCKfVa3S1e4PHvxWcwyMMg8w',
1463 'playlist_mincount': 91,
1464 'info_dict': {
1465 'id': 'UCKfVa3S1e4PHvxWcwyMMg8w',
1466 }
1467 }]
1468
1469 @staticmethod
1470 def extract_videos_from_page(page):
1471 ids_in_page = []
1472 titles_in_page = []
1473 for mobj in re.finditer(r'(?:title="(?P<title>[^"]+)"[^>]+)?href="/watch\?v=(?P<id>[0-9A-Za-z_-]+)&?', page):
1474 video_id = mobj.group('id')
1475 video_title = unescapeHTML(mobj.group('title'))
1476 try:
1477 idx = ids_in_page.index(video_id)
1478 if video_title and not titles_in_page[idx]:
1479 titles_in_page[idx] = video_title
1480 except ValueError:
1481 ids_in_page.append(video_id)
1482 titles_in_page.append(video_title)
1483 return zip(ids_in_page, titles_in_page)
1484
1485 def _real_extract(self, url):
1486 channel_id = self._match_id(url)
1487
1488 url = self._TEMPLATE_URL % channel_id
1489
1490 # Channel by page listing is restricted to 35 pages of 30 items, i.e. 1050 videos total (see #5778)
1491 # Workaround by extracting as a playlist if managed to obtain channel playlist URL
1492 # otherwise fallback on channel by page extraction
1493 channel_page = self._download_webpage(
1494 url + '?view=57', channel_id,
1495 'Downloading channel page', fatal=False)
1496 channel_playlist_id = self._html_search_meta(
1497 'channelId', channel_page, 'channel id', default=None)
1498 if not channel_playlist_id:
1499 channel_playlist_id = self._search_regex(
1500 r'data-channel-external-id="([^"]+)"',
1501 channel_page, 'channel id', default=None)
1502 if channel_playlist_id and channel_playlist_id.startswith('UC'):
1503 playlist_id = 'UU' + channel_playlist_id[2:]
1504 return self.url_result(
1505 compat_urlparse.urljoin(url, '/playlist?list=%s' % playlist_id), 'YoutubePlaylist')
1506
1507 channel_page = self._download_webpage(url, channel_id, 'Downloading page #1')
1508 autogenerated = re.search(r'''(?x)
1509 class="[^"]*?(?:
1510 channel-header-autogenerated-label|
1511 yt-channel-title-autogenerated
1512 )[^"]*"''', channel_page) is not None
1513
1514 if autogenerated:
1515 # The videos are contained in a single page
1516 # the ajax pages can't be used, they are empty
1517 entries = [
1518 self.url_result(
1519 video_id, 'Youtube', video_id=video_id,
1520 video_title=video_title)
1521 for video_id, video_title in self.extract_videos_from_page(channel_page)]
1522 return self.playlist_result(entries, channel_id)
1523
1524 def _entries():
1525 more_widget_html = content_html = channel_page
1526 for pagenum in itertools.count(1):
1527
1528 for video_id, video_title in self.extract_videos_from_page(content_html):
1529 yield self.url_result(
1530 video_id, 'Youtube', video_id=video_id,
1531 video_title=video_title)
1532
1533 mobj = re.search(
1534 r'data-uix-load-more-href="/?(?P<more>[^"]+)"',
1535 more_widget_html)
1536 if not mobj:
1537 break
1538
1539 more = self._download_json(
1540 'https://youtube.com/%s' % mobj.group('more'), channel_id,
1541 'Downloading page #%s' % (pagenum + 1),
1542 transform_source=uppercase_escape)
1543 content_html = more['content_html']
1544 more_widget_html = more['load_more_widget_html']
1545
1546 return self.playlist_result(_entries(), channel_id)
1547
1548
1549 class YoutubeUserIE(YoutubeChannelIE):
1550 IE_DESC = 'YouTube.com user videos (URL or "ytuser" keyword)'
1551 _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_-]+)'
1552 _TEMPLATE_URL = 'https://www.youtube.com/user/%s/videos'
1553 IE_NAME = 'youtube:user'
1554
1555 _TESTS = [{
1556 'url': 'https://www.youtube.com/user/TheLinuxFoundation',
1557 'playlist_mincount': 320,
1558 'info_dict': {
1559 'title': 'TheLinuxFoundation',
1560 }
1561 }, {
1562 'url': 'ytuser:phihag',
1563 'only_matching': True,
1564 }]
1565
1566 @classmethod
1567 def suitable(cls, url):
1568 # Don't return True if the url can be extracted with other youtube
1569 # extractor, the regex would is too permissive and it would match.
1570 other_ies = iter(klass for (name, klass) in globals().items() if name.endswith('IE') and klass is not cls)
1571 if any(ie.suitable(url) for ie in other_ies):
1572 return False
1573 else:
1574 return super(YoutubeUserIE, cls).suitable(url)
1575
1576
1577 class YoutubeSearchIE(SearchInfoExtractor, YoutubePlaylistIE):
1578 IE_DESC = 'YouTube.com searches'
1579 # there doesn't appear to be a real limit, for example if you search for
1580 # 'python' you get more than 8.000.000 results
1581 _MAX_RESULTS = float('inf')
1582 IE_NAME = 'youtube:search'
1583 _SEARCH_KEY = 'ytsearch'
1584 _EXTRA_QUERY_ARGS = {}
1585 _TESTS = []
1586
1587 def _get_n_results(self, query, n):
1588 """Get a specified number of results for a query"""
1589
1590 videos = []
1591 limit = n
1592
1593 for pagenum in itertools.count(1):
1594 url_query = {
1595 'search_query': query.encode('utf-8'),
1596 'page': pagenum,
1597 'spf': 'navigate',
1598 }
1599 url_query.update(self._EXTRA_QUERY_ARGS)
1600 result_url = 'https://www.youtube.com/results?' + compat_urllib_parse.urlencode(url_query)
1601 data = self._download_json(
1602 result_url, video_id='query "%s"' % query,
1603 note='Downloading page %s' % pagenum,
1604 errnote='Unable to download API page')
1605 html_content = data[1]['body']['content']
1606
1607 if 'class="search-message' in html_content:
1608 raise ExtractorError(
1609 '[youtube] No video results', expected=True)
1610
1611 new_videos = self._ids_to_results(orderedSet(re.findall(
1612 r'href="/watch\?v=(.{11})', html_content)))
1613 videos += new_videos
1614 if not new_videos or len(videos) > limit:
1615 break
1616
1617 if len(videos) > n:
1618 videos = videos[:n]
1619 return self.playlist_result(videos, query)
1620
1621
1622 class YoutubeSearchDateIE(YoutubeSearchIE):
1623 IE_NAME = YoutubeSearchIE.IE_NAME + ':date'
1624 _SEARCH_KEY = 'ytsearchdate'
1625 IE_DESC = 'YouTube.com searches, newest videos first'
1626 _EXTRA_QUERY_ARGS = {'search_sort': 'video_date_uploaded'}
1627
1628
1629 class YoutubeSearchURLIE(InfoExtractor):
1630 IE_DESC = 'YouTube.com search URLs'
1631 IE_NAME = 'youtube:search_url'
1632 _VALID_URL = r'https?://(?:www\.)?youtube\.com/results\?(.*?&)?search_query=(?P<query>[^&]+)(?:[&]|$)'
1633 _TESTS = [{
1634 'url': 'https://www.youtube.com/results?baz=bar&search_query=youtube-dl+test+video&filters=video&lclk=video',
1635 'playlist_mincount': 5,
1636 'info_dict': {
1637 'title': 'youtube-dl test video',
1638 }
1639 }]
1640
1641 def _real_extract(self, url):
1642 mobj = re.match(self._VALID_URL, url)
1643 query = compat_urllib_parse_unquote_plus(mobj.group('query'))
1644
1645 webpage = self._download_webpage(url, query)
1646 result_code = self._search_regex(
1647 r'(?s)<ol[^>]+class="item-section"(.*?)</ol>', webpage, 'result HTML')
1648
1649 part_codes = re.findall(
1650 r'(?s)<h3 class="yt-lockup-title">(.*?)</h3>', result_code)
1651 entries = []
1652 for part_code in part_codes:
1653 part_title = self._html_search_regex(
1654 [r'(?s)title="([^"]+)"', r'>([^<]+)</a>'], part_code, 'item title', fatal=False)
1655 part_url_snippet = self._html_search_regex(
1656 r'(?s)href="([^"]+)"', part_code, 'item URL')
1657 part_url = compat_urlparse.urljoin(
1658 'https://www.youtube.com/', part_url_snippet)
1659 entries.append({
1660 '_type': 'url',
1661 'url': part_url,
1662 'title': part_title,
1663 })
1664
1665 return {
1666 '_type': 'playlist',
1667 'entries': entries,
1668 'title': query,
1669 }
1670
1671
1672 class YoutubeShowIE(InfoExtractor):
1673 IE_DESC = 'YouTube.com (multi-season) shows'
1674 _VALID_URL = r'https?://www\.youtube\.com/show/(?P<id>[^?#]*)'
1675 IE_NAME = 'youtube:show'
1676 _TESTS = [{
1677 'url': 'http://www.youtube.com/show/airdisasters',
1678 'playlist_mincount': 3,
1679 'info_dict': {
1680 'id': 'airdisasters',
1681 'title': 'Air Disasters',
1682 }
1683 }]
1684
1685 def _real_extract(self, url):
1686 mobj = re.match(self._VALID_URL, url)
1687 playlist_id = mobj.group('id')
1688 webpage = self._download_webpage(
1689 url, playlist_id, 'Downloading show webpage')
1690 # There's one playlist for each season of the show
1691 m_seasons = list(re.finditer(r'href="(/playlist\?list=.*?)"', webpage))
1692 self.to_screen('%s: Found %s seasons' % (playlist_id, len(m_seasons)))
1693 entries = [
1694 self.url_result(
1695 'https://www.youtube.com' + season.group(1), 'YoutubePlaylist')
1696 for season in m_seasons
1697 ]
1698 title = self._og_search_title(webpage, fatal=False)
1699
1700 return {
1701 '_type': 'playlist',
1702 'id': playlist_id,
1703 'title': title,
1704 'entries': entries,
1705 }
1706
1707
1708 class YoutubeFeedsInfoExtractor(YoutubeBaseInfoExtractor):
1709 """
1710 Base class for feed extractors
1711 Subclasses must define the _FEED_NAME and _PLAYLIST_TITLE properties.
1712 """
1713 _LOGIN_REQUIRED = True
1714
1715 @property
1716 def IE_NAME(self):
1717 return 'youtube:%s' % self._FEED_NAME
1718
1719 def _real_initialize(self):
1720 self._login()
1721
1722 def _real_extract(self, url):
1723 page = self._download_webpage(
1724 'https://www.youtube.com/feed/%s' % self._FEED_NAME, self._PLAYLIST_TITLE)
1725
1726 # The extraction process is the same as for playlists, but the regex
1727 # for the video ids doesn't contain an index
1728 ids = []
1729 more_widget_html = content_html = page
1730 for page_num in itertools.count(1):
1731 matches = re.findall(r'href="\s*/watch\?v=([0-9A-Za-z_-]{11})', content_html)
1732
1733 # 'recommended' feed has infinite 'load more' and each new portion spins
1734 # the same videos in (sometimes) slightly different order, so we'll check
1735 # for unicity and break when portion has no new videos
1736 new_ids = filter(lambda video_id: video_id not in ids, orderedSet(matches))
1737 if not new_ids:
1738 break
1739
1740 ids.extend(new_ids)
1741
1742 mobj = re.search(r'data-uix-load-more-href="/?(?P<more>[^"]+)"', more_widget_html)
1743 if not mobj:
1744 break
1745
1746 more = self._download_json(
1747 'https://youtube.com/%s' % mobj.group('more'), self._PLAYLIST_TITLE,
1748 'Downloading page #%s' % page_num,
1749 transform_source=uppercase_escape)
1750 content_html = more['content_html']
1751 more_widget_html = more['load_more_widget_html']
1752
1753 return self.playlist_result(
1754 self._ids_to_results(ids), playlist_title=self._PLAYLIST_TITLE)
1755
1756
1757 class YoutubeWatchLaterIE(YoutubePlaylistIE):
1758 IE_NAME = 'youtube:watchlater'
1759 IE_DESC = 'Youtube watch later list, ":ytwatchlater" for short (requires authentication)'
1760 _VALID_URL = r'https?://www\.youtube\.com/(?:feed/watch_later|playlist\?list=WL)|:ytwatchlater'
1761
1762 _TESTS = [] # override PlaylistIE tests
1763
1764 def _real_extract(self, url):
1765 return self._extract_playlist('WL')
1766
1767
1768 class YoutubeFavouritesIE(YoutubeBaseInfoExtractor):
1769 IE_NAME = 'youtube:favorites'
1770 IE_DESC = 'YouTube.com favourite videos, ":ytfav" for short (requires authentication)'
1771 _VALID_URL = r'https?://www\.youtube\.com/my_favorites|:ytfav(?:ou?rites)?'
1772 _LOGIN_REQUIRED = True
1773
1774 def _real_extract(self, url):
1775 webpage = self._download_webpage('https://www.youtube.com/my_favorites', 'Youtube Favourites videos')
1776 playlist_id = self._search_regex(r'list=(.+?)["&]', webpage, 'favourites playlist id')
1777 return self.url_result(playlist_id, 'YoutubePlaylist')
1778
1779
1780 class YoutubeRecommendedIE(YoutubeFeedsInfoExtractor):
1781 IE_DESC = 'YouTube.com recommended videos, ":ytrec" for short (requires authentication)'
1782 _VALID_URL = r'https?://www\.youtube\.com/feed/recommended|:ytrec(?:ommended)?'
1783 _FEED_NAME = 'recommended'
1784 _PLAYLIST_TITLE = 'Youtube Recommended videos'
1785
1786
1787 class YoutubeSubscriptionsIE(YoutubeFeedsInfoExtractor):
1788 IE_DESC = 'YouTube.com subscriptions feed, "ytsubs" keyword (requires authentication)'
1789 _VALID_URL = r'https?://www\.youtube\.com/feed/subscriptions|:ytsubs(?:criptions)?'
1790 _FEED_NAME = 'subscriptions'
1791 _PLAYLIST_TITLE = 'Youtube Subscriptions'
1792
1793
1794 class YoutubeHistoryIE(YoutubeFeedsInfoExtractor):
1795 IE_DESC = 'Youtube watch history, ":ythistory" for short (requires authentication)'
1796 _VALID_URL = 'https?://www\.youtube\.com/feed/history|:ythistory'
1797 _FEED_NAME = 'history'
1798 _PLAYLIST_TITLE = 'Youtube History'
1799
1800
1801 class YoutubeTruncatedURLIE(InfoExtractor):
1802 IE_NAME = 'youtube:truncated_url'
1803 IE_DESC = False # Do not list
1804 _VALID_URL = r'''(?x)
1805 (?:https?://)?
1806 (?:\w+\.)?[yY][oO][uU][tT][uU][bB][eE](?:-nocookie)?\.com/
1807 (?:watch\?(?:
1808 feature=[a-z_]+|
1809 annotation_id=annotation_[^&]+|
1810 x-yt-cl=[0-9]+|
1811 hl=[^&]*|
1812 )?
1813 |
1814 attribution_link\?a=[^&]+
1815 )
1816 $
1817 '''
1818
1819 _TESTS = [{
1820 'url': 'http://www.youtube.com/watch?annotation_id=annotation_3951667041',
1821 'only_matching': True,
1822 }, {
1823 'url': 'http://www.youtube.com/watch?',
1824 'only_matching': True,
1825 }, {
1826 'url': 'https://www.youtube.com/watch?x-yt-cl=84503534',
1827 'only_matching': True,
1828 }, {
1829 'url': 'https://www.youtube.com/watch?feature=foo',
1830 'only_matching': True,
1831 }, {
1832 'url': 'https://www.youtube.com/watch?hl=en-GB',
1833 'only_matching': True,
1834 }]
1835
1836 def _real_extract(self, url):
1837 raise ExtractorError(
1838 'Did you forget to quote the URL? Remember that & is a meta '
1839 'character in most shells, so you want to put the URL in quotes, '
1840 'like youtube-dl '
1841 '"http://www.youtube.com/watch?feature=foo&v=BaW_jenozKc" '
1842 ' or simply youtube-dl BaW_jenozKc .',
1843 expected=True)
1844
1845
1846 class YoutubeTruncatedIDIE(InfoExtractor):
1847 IE_NAME = 'youtube:truncated_id'
1848 IE_DESC = False # Do not list
1849 _VALID_URL = r'https?://(?:www\.)?youtube\.com/watch\?v=(?P<id>[0-9A-Za-z_-]{1,10})$'
1850
1851 _TESTS = [{
1852 'url': 'https://www.youtube.com/watch?v=N_708QY7Ob',
1853 'only_matching': True,
1854 }]
1855
1856 def _real_extract(self, url):
1857 video_id = self._match_id(url)
1858 raise ExtractorError(
1859 'Incomplete YouTube ID %s. URL %s looks truncated.' % (video_id, url),
1860 expected=True)