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


10d7baef67f9ab7c832a150ba18a3f79c47c5f4e
[yt-dlp.git] / youtube_dl / extractor / youtube.py
1 # coding: utf-8
2
3 import itertools
4 import json
5 import os.path
6 import re
7 import traceback
8
9 from .common import InfoExtractor, SearchInfoExtractor
10 from .subtitles import SubtitlesInfoExtractor
11 from ..jsinterp import JSInterpreter
12 from ..swfinterp import SWFInterpreter
13 from ..utils import (
14 compat_chr,
15 compat_parse_qs,
16 compat_urllib_parse,
17 compat_urllib_request,
18 compat_urlparse,
19 compat_str,
20
21 clean_html,
22 get_element_by_id,
23 get_element_by_attribute,
24 ExtractorError,
25 int_or_none,
26 PagedList,
27 unescapeHTML,
28 unified_strdate,
29 orderedSet,
30 uppercase_escape,
31 )
32
33 class YoutubeBaseInfoExtractor(InfoExtractor):
34 """Provide base functions for Youtube extractors"""
35 _LOGIN_URL = 'https://accounts.google.com/ServiceLogin'
36 _TWOFACTOR_URL = 'https://accounts.google.com/SecondFactor'
37 _LANG_URL = r'https://www.youtube.com/?hl=en&persist_hl=1&gl=US&persist_gl=1&opt_out_ackd=1'
38 _AGE_URL = 'https://www.youtube.com/verify_age?next_url=/&gl=US&hl=en'
39 _NETRC_MACHINE = 'youtube'
40 # If True it will raise an error if no login info is provided
41 _LOGIN_REQUIRED = False
42
43 def _set_language(self):
44 return bool(self._download_webpage(
45 self._LANG_URL, None,
46 note=u'Setting language', errnote='unable to set language',
47 fatal=False))
48
49 def _login(self):
50 """
51 Attempt to log in to YouTube.
52 True is returned if successful or skipped.
53 False is returned if login failed.
54
55 If _LOGIN_REQUIRED is set and no authentication was provided, an error is raised.
56 """
57 (username, password) = self._get_login_info()
58 # No authentication to be performed
59 if username is None:
60 if self._LOGIN_REQUIRED:
61 raise ExtractorError(u'No login info available, needed for using %s.' % self.IE_NAME, expected=True)
62 return True
63
64 login_page = self._download_webpage(
65 self._LOGIN_URL, None,
66 note=u'Downloading login page',
67 errnote=u'unable to fetch login page', fatal=False)
68 if login_page is False:
69 return
70
71 galx = self._search_regex(r'(?s)<input.+?name="GALX".+?value="(.+?)"',
72 login_page, u'Login GALX parameter')
73
74 # Log in
75 login_form_strs = {
76 u'continue': u'https://www.youtube.com/signin?action_handle_signin=true&feature=sign_in_button&hl=en_US&nomobiletemp=1',
77 u'Email': username,
78 u'GALX': galx,
79 u'Passwd': password,
80
81 u'PersistentCookie': u'yes',
82 u'_utf8': u'霱',
83 u'bgresponse': u'js_disabled',
84 u'checkConnection': u'',
85 u'checkedDomains': u'youtube',
86 u'dnConn': u'',
87 u'pstMsg': u'0',
88 u'rmShown': u'1',
89 u'secTok': u'',
90 u'signIn': u'Sign in',
91 u'timeStmp': u'',
92 u'service': u'youtube',
93 u'uilel': u'3',
94 u'hl': u'en_US',
95 }
96
97 # Convert to UTF-8 *before* urlencode because Python 2.x's urlencode
98 # chokes on unicode
99 login_form = dict((k.encode('utf-8'), v.encode('utf-8')) for k,v in login_form_strs.items())
100 login_data = compat_urllib_parse.urlencode(login_form).encode('ascii')
101
102 req = compat_urllib_request.Request(self._LOGIN_URL, login_data)
103 login_results = self._download_webpage(
104 req, None,
105 note=u'Logging in', errnote=u'unable to log in', fatal=False)
106 if login_results is False:
107 return False
108
109 if re.search(r'id="errormsg_0_Passwd"', login_results) is not None:
110 raise ExtractorError(u'Please use your account password and a two-factor code instead of an application-specific password.', expected=True)
111
112 # Two-Factor
113 # TODO add SMS and phone call support - these require making a request and then prompting the user
114
115 if re.search(r'(?i)<form[^>]* id="gaia_secondfactorform"', login_results) is not None:
116 tfa_code = self._get_tfa_info()
117
118 if tfa_code is None:
119 self._downloader.report_warning(u'Two-factor authentication required. Provide it with --twofactor <code>')
120 self._downloader.report_warning(u'(Note that only TOTP (Google Authenticator App) codes work at this time.)')
121 return False
122
123 # Unlike the first login form, secTok and timeStmp are both required for the TFA form
124
125 match = re.search(r'id="secTok"\n\s+value=\'(.+)\'/>', login_results, re.M | re.U)
126 if match is None:
127 self._downloader.report_warning(u'Failed to get secTok - did the page structure change?')
128 secTok = match.group(1)
129 match = re.search(r'id="timeStmp"\n\s+value=\'(.+)\'/>', login_results, re.M | re.U)
130 if match is None:
131 self._downloader.report_warning(u'Failed to get timeStmp - did the page structure change?')
132 timeStmp = match.group(1)
133
134 tfa_form_strs = {
135 u'continue': u'https://www.youtube.com/signin?action_handle_signin=true&feature=sign_in_button&hl=en_US&nomobiletemp=1',
136 u'smsToken': u'',
137 u'smsUserPin': tfa_code,
138 u'smsVerifyPin': u'Verify',
139
140 u'PersistentCookie': u'yes',
141 u'checkConnection': u'',
142 u'checkedDomains': u'youtube',
143 u'pstMsg': u'1',
144 u'secTok': secTok,
145 u'timeStmp': timeStmp,
146 u'service': u'youtube',
147 u'hl': u'en_US',
148 }
149 tfa_form = dict((k.encode('utf-8'), v.encode('utf-8')) for k,v in tfa_form_strs.items())
150 tfa_data = compat_urllib_parse.urlencode(tfa_form).encode('ascii')
151
152 tfa_req = compat_urllib_request.Request(self._TWOFACTOR_URL, tfa_data)
153 tfa_results = self._download_webpage(
154 tfa_req, None,
155 note=u'Submitting TFA code', errnote=u'unable to submit tfa', fatal=False)
156
157 if tfa_results is False:
158 return False
159
160 if re.search(r'(?i)<form[^>]* id="gaia_secondfactorform"', tfa_results) is not None:
161 self._downloader.report_warning(u'Two-factor code expired. Please try again, or use a one-use backup code instead.')
162 return False
163 if re.search(r'(?i)<form[^>]* id="gaia_loginform"', tfa_results) is not None:
164 self._downloader.report_warning(u'unable to log in - did the page structure change?')
165 return False
166 if re.search(r'smsauth-interstitial-reviewsettings', tfa_results) is not None:
167 self._downloader.report_warning(u'Your Google account has a security notice. Please log in on your web browser, resolve the notice, and try again.')
168 return False
169
170 if re.search(r'(?i)<form[^>]* id="gaia_loginform"', login_results) is not None:
171 self._downloader.report_warning(u'unable to log in: bad username or password')
172 return False
173 return True
174
175 def _confirm_age(self):
176 age_form = {
177 'next_url': '/',
178 'action_confirm': 'Confirm',
179 }
180 req = compat_urllib_request.Request(self._AGE_URL,
181 compat_urllib_parse.urlencode(age_form).encode('ascii'))
182
183 self._download_webpage(
184 req, None,
185 note=u'Confirming age', errnote=u'Unable to confirm age')
186 return True
187
188 def _real_initialize(self):
189 if self._downloader is None:
190 return
191 if not self._set_language():
192 return
193 if not self._login():
194 return
195 self._confirm_age()
196
197
198 class YoutubeIE(YoutubeBaseInfoExtractor, SubtitlesInfoExtractor):
199 IE_DESC = u'YouTube.com'
200 _VALID_URL = r"""(?x)^
201 (
202 (?:https?://|//) # http(s):// or protocol-independent URL
203 (?:(?:(?:(?:\w+\.)?[yY][oO][uU][tT][uU][bB][eE](?:-nocookie)?\.com/|
204 (?:www\.)?deturl\.com/www\.youtube\.com/|
205 (?:www\.)?pwnyoutube\.com/|
206 (?:www\.)?yourepeat\.com/|
207 tube\.majestyc\.net/|
208 youtube\.googleapis\.com/) # the various hostnames, with wildcard subdomains
209 (?:.*?\#/)? # handle anchor (#/) redirect urls
210 (?: # the various things that can precede the ID:
211 (?:(?:v|embed|e)/) # v/ or embed/ or e/
212 |(?: # or the v= param in all its forms
213 (?:(?:watch|movie)(?:_popup)?(?:\.php)?/?)? # preceding watch(_popup|.php) or nothing (like /?v=xxxx)
214 (?:\?|\#!?) # the params delimiter ? or # or #!
215 (?:.*?&)? # any other preceding param (like /?s=tuff&v=xxxx)
216 v=
217 )
218 ))
219 |youtu\.be/ # just youtu.be/xxxx
220 |(?:www\.)?cleanvideosearch\.com/media/action/yt/watch\?videoId=
221 )
222 )? # all until now is optional -> you can pass the naked ID
223 ([0-9A-Za-z_-]{11}) # here is it! the YouTube video ID
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
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', 'height': 2160, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
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
273 # Dash mp4 audio
274 '139': {'ext': 'm4a', 'format_note': 'DASH audio', 'vcodec': 'none', 'abr': 48, 'preference': -50},
275 '140': {'ext': 'm4a', 'format_note': 'DASH audio', 'vcodec': 'none', 'abr': 128, 'preference': -50},
276 '141': {'ext': 'm4a', 'format_note': 'DASH audio', 'vcodec': 'none', 'abr': 256, 'preference': -50},
277
278 # Dash webm
279 '167': {'ext': 'webm', 'height': 360, 'width': 640, 'format_note': 'DASH video', 'acodec': 'none', 'container': 'webm', 'vcodec': 'VP8', 'preference': -40},
280 '168': {'ext': 'webm', 'height': 480, 'width': 854, 'format_note': 'DASH video', 'acodec': 'none', 'container': 'webm', 'vcodec': 'VP8', 'preference': -40},
281 '169': {'ext': 'webm', 'height': 720, 'width': 1280, 'format_note': 'DASH video', 'acodec': 'none', 'container': 'webm', 'vcodec': 'VP8', 'preference': -40},
282 '170': {'ext': 'webm', 'height': 1080, 'width': 1920, 'format_note': 'DASH video', 'acodec': 'none', 'container': 'webm', 'vcodec': 'VP8', 'preference': -40},
283 '218': {'ext': 'webm', 'height': 480, 'width': 854, 'format_note': 'DASH video', 'acodec': 'none', 'container': 'webm', 'vcodec': 'VP8', 'preference': -40},
284 '219': {'ext': 'webm', 'height': 480, 'width': 854, 'format_note': 'DASH video', 'acodec': 'none', 'container': 'webm', 'vcodec': 'VP8', 'preference': -40},
285 '242': {'ext': 'webm', 'height': 240, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
286 '243': {'ext': 'webm', 'height': 360, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
287 '244': {'ext': 'webm', 'height': 480, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
288 '245': {'ext': 'webm', 'height': 480, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
289 '246': {'ext': 'webm', 'height': 480, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
290 '247': {'ext': 'webm', 'height': 720, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
291 '248': {'ext': 'webm', 'height': 1080, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
292 '271': {'ext': 'webm', 'height': 1440, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
293 '272': {'ext': 'webm', 'height': 2160, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
294
295 # Dash webm audio
296 '171': {'ext': 'webm', 'vcodec': 'none', 'format_note': 'DASH audio', 'abr': 128, 'preference': -50},
297 '172': {'ext': 'webm', 'vcodec': 'none', 'format_note': 'DASH audio', 'abr': 256, 'preference': -50},
298
299 # RTMP (unnamed)
300 '_rtmp': {'protocol': 'rtmp'},
301 }
302
303 IE_NAME = u'youtube'
304 _TESTS = [
305 {
306 u"url": u"http://www.youtube.com/watch?v=BaW_jenozKc",
307 u"file": u"BaW_jenozKc.mp4",
308 u"info_dict": {
309 u"title": u"youtube-dl test video \"'/\\ä↭𝕐",
310 u"uploader": u"Philipp Hagemeister",
311 u"uploader_id": u"phihag",
312 u"upload_date": u"20121002",
313 u"description": u"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 .",
314 u"categories": [u'Science & Technology'],
315 'like_count': int,
316 'dislike_count': int,
317 }
318 },
319 {
320 u"url": u"http://www.youtube.com/watch?v=UxxajLWwzqY",
321 u"file": u"UxxajLWwzqY.mp4",
322 u"note": u"Test generic use_cipher_signature video (#897)",
323 u"info_dict": {
324 u"upload_date": u"20120506",
325 u"title": u"Icona Pop - I Love It (feat. Charli XCX) [OFFICIAL VIDEO]",
326 u"description": u"md5:fea86fda2d5a5784273df5c7cc994d9f",
327 u"uploader": u"Icona Pop",
328 u"uploader_id": u"IconaPop"
329 }
330 },
331 {
332 u"url": u"https://www.youtube.com/watch?v=07FYdnEawAQ",
333 u"file": u"07FYdnEawAQ.mp4",
334 u"note": u"Test VEVO video with age protection (#956)",
335 u"info_dict": {
336 u"upload_date": u"20130703",
337 u"title": u"Justin Timberlake - Tunnel Vision (Explicit)",
338 u"description": u"md5:64249768eec3bc4276236606ea996373",
339 u"uploader": u"justintimberlakeVEVO",
340 u"uploader_id": u"justintimberlakeVEVO"
341 }
342 },
343 {
344 u"url": u"//www.YouTube.com/watch?v=yZIXLfi8CZQ",
345 u"file": u"yZIXLfi8CZQ.mp4",
346 u"note": u"Embed-only video (#1746)",
347 u"info_dict": {
348 u"upload_date": u"20120608",
349 u"title": u"Principal Sexually Assaults A Teacher - Episode 117 - 8th June 2012",
350 u"description": u"md5:09b78bd971f1e3e289601dfba15ca4f7",
351 u"uploader": u"SET India",
352 u"uploader_id": u"setindia"
353 }
354 },
355 {
356 u"url": u"http://www.youtube.com/watch?v=a9LDPn-MO4I",
357 u"file": u"a9LDPn-MO4I.m4a",
358 u"note": u"256k DASH audio (format 141) via DASH manifest",
359 u"info_dict": {
360 u"upload_date": "20121002",
361 u"uploader_id": "8KVIDEO",
362 u"description": "No description available.",
363 u"uploader": "8KVIDEO",
364 u"title": "UHDTV TEST 8K VIDEO.mp4"
365 },
366 u"params": {
367 u"youtube_include_dash_manifest": True,
368 u"format": "141",
369 },
370 },
371 # DASH manifest with encrypted signature
372 {
373 u'url': u'https://www.youtube.com/watch?v=IB3lcPjvWLA',
374 u'info_dict': {
375 u'id': u'IB3lcPjvWLA',
376 u'ext': u'm4a',
377 u'title': u'Afrojack - The Spark ft. Spree Wilson',
378 u'description': u'md5:9717375db5a9a3992be4668bbf3bc0a8',
379 u'uploader': u'AfrojackVEVO',
380 u'uploader_id': u'AfrojackVEVO',
381 u'upload_date': u'20131011',
382 },
383 u"params": {
384 u'youtube_include_dash_manifest': True,
385 u'format': '141',
386 },
387 },
388 ]
389
390
391 @classmethod
392 def suitable(cls, url):
393 """Receives a URL and returns True if suitable for this IE."""
394 if YoutubePlaylistIE.suitable(url): return False
395 return re.match(cls._VALID_URL, url) is not None
396
397 def __init__(self, *args, **kwargs):
398 super(YoutubeIE, self).__init__(*args, **kwargs)
399 self._player_cache = {}
400
401 def report_video_info_webpage_download(self, video_id):
402 """Report attempt to download video info webpage."""
403 self.to_screen(u'%s: Downloading video info webpage' % video_id)
404
405 def report_information_extraction(self, video_id):
406 """Report attempt to extract video information."""
407 self.to_screen(u'%s: Extracting video information' % video_id)
408
409 def report_unavailable_format(self, video_id, format):
410 """Report extracted video URL."""
411 self.to_screen(u'%s: Format %s not available' % (video_id, format))
412
413 def report_rtmp_download(self):
414 """Indicate the download will use the RTMP protocol."""
415 self.to_screen(u'RTMP download detected')
416
417 def _signature_cache_id(self, example_sig):
418 """ Return a string representation of a signature """
419 return u'.'.join(compat_str(len(part)) for part in example_sig.split('.'))
420
421 def _extract_signature_function(self, video_id, player_url, example_sig):
422 id_m = re.match(
423 r'.*-(?P<id>[a-zA-Z0-9_-]+)(?:/watch_as3|/html5player)?\.(?P<ext>[a-z]+)$',
424 player_url)
425 if not id_m:
426 raise ExtractorError('Cannot identify player %r' % player_url)
427 player_type = id_m.group('ext')
428 player_id = id_m.group('id')
429
430 # Read from filesystem cache
431 func_id = '%s_%s_%s' % (
432 player_type, player_id, self._signature_cache_id(example_sig))
433 assert os.path.basename(func_id) == func_id
434
435 cache_spec = self._downloader.cache.load(u'youtube-sigfuncs', func_id)
436 if cache_spec is not None:
437 return lambda s: u''.join(s[i] for i in cache_spec)
438
439 if player_type == 'js':
440 code = self._download_webpage(
441 player_url, video_id,
442 note=u'Downloading %s player %s' % (player_type, player_id),
443 errnote=u'Download of %s failed' % player_url)
444 res = self._parse_sig_js(code)
445 elif player_type == 'swf':
446 urlh = self._request_webpage(
447 player_url, video_id,
448 note=u'Downloading %s player %s' % (player_type, player_id),
449 errnote=u'Download of %s failed' % player_url)
450 code = urlh.read()
451 res = self._parse_sig_swf(code)
452 else:
453 assert False, 'Invalid player type %r' % player_type
454
455 if cache_spec is None:
456 test_string = u''.join(map(compat_chr, range(len(example_sig))))
457 cache_res = res(test_string)
458 cache_spec = [ord(c) for c in cache_res]
459
460 self._downloader.cache.store(u'youtube-sigfuncs', func_id, cache_spec)
461 return res
462
463 def _print_sig_code(self, func, example_sig):
464 def gen_sig_code(idxs):
465 def _genslice(start, end, step):
466 starts = u'' if start == 0 else str(start)
467 ends = (u':%d' % (end+step)) if end + step >= 0 else u':'
468 steps = u'' if step == 1 else (u':%d' % step)
469 return u's[%s%s%s]' % (starts, ends, steps)
470
471 step = None
472 start = '(Never used)' # Quelch pyflakes warnings - start will be
473 # set as soon as step is set
474 for i, prev in zip(idxs[1:], idxs[:-1]):
475 if step is not None:
476 if i - prev == step:
477 continue
478 yield _genslice(start, prev, step)
479 step = None
480 continue
481 if i - prev in [-1, 1]:
482 step = i - prev
483 start = prev
484 continue
485 else:
486 yield u's[%d]' % prev
487 if step is None:
488 yield u's[%d]' % i
489 else:
490 yield _genslice(start, i, step)
491
492 test_string = u''.join(map(compat_chr, range(len(example_sig))))
493 cache_res = func(test_string)
494 cache_spec = [ord(c) for c in cache_res]
495 expr_code = u' + '.join(gen_sig_code(cache_spec))
496 signature_id_tuple = '(%s)' % (
497 ', '.join(compat_str(len(p)) for p in example_sig.split('.')))
498 code = (u'if tuple(len(p) for p in s.split(\'.\')) == %s:\n'
499 u' return %s\n') % (signature_id_tuple, expr_code)
500 self.to_screen(u'Extracted signature function:\n' + code)
501
502 def _parse_sig_js(self, jscode):
503 funcname = self._search_regex(
504 r'signature=([$a-zA-Z]+)', jscode,
505 u'Initial JS player signature function name')
506
507 jsi = JSInterpreter(jscode)
508 initial_function = jsi.extract_function(funcname)
509 return lambda s: initial_function([s])
510
511 def _parse_sig_swf(self, file_contents):
512 swfi = SWFInterpreter(file_contents)
513 TARGET_CLASSNAME = u'SignatureDecipher'
514 searched_class = swfi.extract_class(TARGET_CLASSNAME)
515 initial_function = swfi.extract_function(searched_class, u'decipher')
516 return lambda s: initial_function([s])
517
518 def _decrypt_signature(self, s, video_id, player_url, age_gate=False):
519 """Turn the encrypted s field into a working signature"""
520
521 if player_url is None:
522 raise ExtractorError(u'Cannot decrypt signature without player_url')
523
524 if player_url.startswith(u'//'):
525 player_url = u'https:' + player_url
526 try:
527 player_id = (player_url, self._signature_cache_id(s))
528 if player_id not in self._player_cache:
529 func = self._extract_signature_function(
530 video_id, player_url, s
531 )
532 self._player_cache[player_id] = func
533 func = self._player_cache[player_id]
534 if self._downloader.params.get('youtube_print_sig_code'):
535 self._print_sig_code(func, s)
536 return func(s)
537 except Exception as e:
538 tb = traceback.format_exc()
539 raise ExtractorError(
540 u'Signature extraction failed: ' + tb, cause=e)
541
542 def _get_available_subtitles(self, video_id, webpage):
543 try:
544 sub_list = self._download_webpage(
545 'https://video.google.com/timedtext?hl=en&type=list&v=%s' % video_id,
546 video_id, note=False)
547 except ExtractorError as err:
548 self._downloader.report_warning(u'unable to download video subtitles: %s' % compat_str(err))
549 return {}
550 lang_list = re.findall(r'name="([^"]*)"[^>]+lang_code="([\w\-]+)"', sub_list)
551
552 sub_lang_list = {}
553 for l in lang_list:
554 lang = l[1]
555 if lang in sub_lang_list:
556 continue
557 params = compat_urllib_parse.urlencode({
558 'lang': lang,
559 'v': video_id,
560 'fmt': self._downloader.params.get('subtitlesformat', 'srt'),
561 'name': unescapeHTML(l[0]).encode('utf-8'),
562 })
563 url = u'https://www.youtube.com/api/timedtext?' + params
564 sub_lang_list[lang] = url
565 if not sub_lang_list:
566 self._downloader.report_warning(u'video doesn\'t have subtitles')
567 return {}
568 return sub_lang_list
569
570 def _get_available_automatic_caption(self, video_id, webpage):
571 """We need the webpage for getting the captions url, pass it as an
572 argument to speed up the process."""
573 sub_format = self._downloader.params.get('subtitlesformat', 'srt')
574 self.to_screen(u'%s: Looking for automatic captions' % video_id)
575 mobj = re.search(r';ytplayer.config = ({.*?});', webpage)
576 err_msg = u'Couldn\'t find automatic captions for %s' % video_id
577 if mobj is None:
578 self._downloader.report_warning(err_msg)
579 return {}
580 player_config = json.loads(mobj.group(1))
581 try:
582 args = player_config[u'args']
583 caption_url = args[u'ttsurl']
584 timestamp = args[u'timestamp']
585 # We get the available subtitles
586 list_params = compat_urllib_parse.urlencode({
587 'type': 'list',
588 'tlangs': 1,
589 'asrs': 1,
590 })
591 list_url = caption_url + '&' + list_params
592 caption_list = self._download_xml(list_url, video_id)
593 original_lang_node = caption_list.find('track')
594 if original_lang_node is None or original_lang_node.attrib.get('kind') != 'asr' :
595 self._downloader.report_warning(u'Video doesn\'t have automatic captions')
596 return {}
597 original_lang = original_lang_node.attrib['lang_code']
598
599 sub_lang_list = {}
600 for lang_node in caption_list.findall('target'):
601 sub_lang = lang_node.attrib['lang_code']
602 params = compat_urllib_parse.urlencode({
603 'lang': original_lang,
604 'tlang': sub_lang,
605 'fmt': sub_format,
606 'ts': timestamp,
607 'kind': 'asr',
608 })
609 sub_lang_list[sub_lang] = caption_url + '&' + params
610 return sub_lang_list
611 # An extractor error can be raise by the download process if there are
612 # no automatic captions but there are subtitles
613 except (KeyError, ExtractorError):
614 self._downloader.report_warning(err_msg)
615 return {}
616
617 @classmethod
618 def extract_id(cls, url):
619 mobj = re.match(cls._VALID_URL, url, re.VERBOSE)
620 if mobj is None:
621 raise ExtractorError(u'Invalid URL: %s' % url)
622 video_id = mobj.group(2)
623 return video_id
624
625 def _extract_from_m3u8(self, manifest_url, video_id):
626 url_map = {}
627 def _get_urls(_manifest):
628 lines = _manifest.split('\n')
629 urls = filter(lambda l: l and not l.startswith('#'),
630 lines)
631 return urls
632 manifest = self._download_webpage(manifest_url, video_id, u'Downloading formats manifest')
633 formats_urls = _get_urls(manifest)
634 for format_url in formats_urls:
635 itag = self._search_regex(r'itag/(\d+?)/', format_url, 'itag')
636 url_map[itag] = format_url
637 return url_map
638
639 def _extract_annotations(self, video_id):
640 url = 'https://www.youtube.com/annotations_invideo?features=1&legacy=1&video_id=%s' % video_id
641 return self._download_webpage(url, video_id, note=u'Searching for annotations.', errnote=u'Unable to download video annotations.')
642
643 def _real_extract(self, url):
644 proto = (
645 u'http' if self._downloader.params.get('prefer_insecure', False)
646 else u'https')
647
648 # Extract original video URL from URL with redirection, like age verification, using next_url parameter
649 mobj = re.search(self._NEXT_URL_RE, url)
650 if mobj:
651 url = proto + '://www.youtube.com/' + compat_urllib_parse.unquote(mobj.group(1)).lstrip('/')
652 video_id = self.extract_id(url)
653
654 # Get video webpage
655 url = proto + '://www.youtube.com/watch?v=%s&gl=US&hl=en&has_verified=1' % video_id
656 video_webpage = self._download_webpage(url, video_id)
657
658 # Attempt to extract SWF player URL
659 mobj = re.search(r'swfConfig.*?"(https?:\\/\\/.*?watch.*?-.*?\.swf)"', video_webpage)
660 if mobj is not None:
661 player_url = re.sub(r'\\(.)', r'\1', mobj.group(1))
662 else:
663 player_url = None
664
665 # Get video info
666 self.report_video_info_webpage_download(video_id)
667 if re.search(r'player-age-gate-content">', video_webpage) is not None:
668 self.report_age_confirmation()
669 age_gate = True
670 # We simulate the access to the video from www.youtube.com/v/{video_id}
671 # this can be viewed without login into Youtube
672 data = compat_urllib_parse.urlencode({
673 'video_id': video_id,
674 'eurl': 'https://youtube.googleapis.com/v/' + video_id,
675 'sts': self._search_regex(
676 r'"sts"\s*:\s*(\d+)', video_webpage, 'sts'),
677 })
678 video_info_url = proto + '://www.youtube.com/get_video_info?' + data
679 video_info_webpage = self._download_webpage(video_info_url, video_id,
680 note=False,
681 errnote='unable to download video info webpage')
682 video_info = compat_parse_qs(video_info_webpage)
683 else:
684 age_gate = False
685 for el_type in ['&el=embedded', '&el=detailpage', '&el=vevo', '']:
686 video_info_url = (proto + '://www.youtube.com/get_video_info?&video_id=%s%s&ps=default&eurl=&gl=US&hl=en'
687 % (video_id, el_type))
688 video_info_webpage = self._download_webpage(video_info_url, video_id,
689 note=False,
690 errnote='unable to download video info webpage')
691 video_info = compat_parse_qs(video_info_webpage)
692 if 'token' in video_info:
693 break
694 if 'token' not in video_info:
695 if 'reason' in video_info:
696 raise ExtractorError(
697 u'YouTube said: %s' % video_info['reason'][0],
698 expected=True, video_id=video_id)
699 else:
700 raise ExtractorError(
701 u'"token" parameter not in video info for unknown reason',
702 video_id=video_id)
703
704 if 'view_count' in video_info:
705 view_count = int(video_info['view_count'][0])
706 else:
707 view_count = None
708
709 # Check for "rental" videos
710 if 'ypc_video_rental_bar_text' in video_info and 'author' not in video_info:
711 raise ExtractorError(u'"rental" videos not supported')
712
713 # Start extracting information
714 self.report_information_extraction(video_id)
715
716 # uploader
717 if 'author' not in video_info:
718 raise ExtractorError(u'Unable to extract uploader name')
719 video_uploader = compat_urllib_parse.unquote_plus(video_info['author'][0])
720
721 # uploader_id
722 video_uploader_id = None
723 mobj = re.search(r'<link itemprop="url" href="http://www.youtube.com/(?:user|channel)/([^"]+)">', video_webpage)
724 if mobj is not None:
725 video_uploader_id = mobj.group(1)
726 else:
727 self._downloader.report_warning(u'unable to extract uploader nickname')
728
729 # title
730 if 'title' in video_info:
731 video_title = video_info['title'][0]
732 else:
733 self._downloader.report_warning(u'Unable to extract video title')
734 video_title = u'_'
735
736 # thumbnail image
737 # We try first to get a high quality image:
738 m_thumb = re.search(r'<span itemprop="thumbnail".*?href="(.*?)">',
739 video_webpage, re.DOTALL)
740 if m_thumb is not None:
741 video_thumbnail = m_thumb.group(1)
742 elif 'thumbnail_url' not in video_info:
743 self._downloader.report_warning(u'unable to extract video thumbnail')
744 video_thumbnail = None
745 else: # don't panic if we can't find it
746 video_thumbnail = compat_urllib_parse.unquote_plus(video_info['thumbnail_url'][0])
747
748 # upload date
749 upload_date = None
750 mobj = re.search(r'(?s)id="eow-date.*?>(.*?)</span>', video_webpage)
751 if mobj is None:
752 mobj = re.search(
753 r'(?s)id="watch-uploader-info".*?>.*?(?:Published|Uploaded|Streamed live) on (.*?)</strong>',
754 video_webpage)
755 if mobj is not None:
756 upload_date = ' '.join(re.sub(r'[/,-]', r' ', mobj.group(1)).split())
757 upload_date = unified_strdate(upload_date)
758
759 m_cat_container = self._search_regex(
760 r'(?s)<h4[^>]*>\s*Category\s*</h4>\s*<ul[^>]*>(.*?)</ul>',
761 video_webpage, 'categories', fatal=False)
762 if m_cat_container:
763 category = self._html_search_regex(
764 r'(?s)<a[^<]+>(.*?)</a>', m_cat_container, 'category',
765 default=None)
766 video_categories = None if category is None else [category]
767 else:
768 video_categories = None
769
770 # description
771 video_description = get_element_by_id("eow-description", video_webpage)
772 if video_description:
773 video_description = re.sub(r'''(?x)
774 <a\s+
775 (?:[a-zA-Z-]+="[^"]+"\s+)*?
776 title="([^"]+)"\s+
777 (?:[a-zA-Z-]+="[^"]+"\s+)*?
778 class="yt-uix-redirect-link"\s*>
779 [^<]+
780 </a>
781 ''', r'\1', video_description)
782 video_description = clean_html(video_description)
783 else:
784 fd_mobj = re.search(r'<meta name="description" content="([^"]+)"', video_webpage)
785 if fd_mobj:
786 video_description = unescapeHTML(fd_mobj.group(1))
787 else:
788 video_description = u''
789
790 def _extract_count(count_name):
791 count = self._search_regex(
792 r'id="watch-%s"[^>]*>.*?([\d,]+)\s*</span>' % re.escape(count_name),
793 video_webpage, count_name, default=None)
794 if count is not None:
795 return int(count.replace(',', ''))
796 return None
797 like_count = _extract_count(u'like')
798 dislike_count = _extract_count(u'dislike')
799
800 # subtitles
801 video_subtitles = self.extract_subtitles(video_id, video_webpage)
802
803 if self._downloader.params.get('listsubtitles', False):
804 self._list_available_subtitles(video_id, video_webpage)
805 return
806
807 if 'length_seconds' not in video_info:
808 self._downloader.report_warning(u'unable to extract video duration')
809 video_duration = None
810 else:
811 video_duration = int(compat_urllib_parse.unquote_plus(video_info['length_seconds'][0]))
812
813 # annotations
814 video_annotations = None
815 if self._downloader.params.get('writeannotations', False):
816 video_annotations = self._extract_annotations(video_id)
817
818 # Decide which formats to download
819 try:
820 mobj = re.search(r';ytplayer\.config\s*=\s*({.*?});', video_webpage)
821 if not mobj:
822 raise ValueError('Could not find vevo ID')
823 json_code = uppercase_escape(mobj.group(1))
824 ytplayer_config = json.loads(json_code)
825 args = ytplayer_config['args']
826 # Easy way to know if the 's' value is in url_encoded_fmt_stream_map
827 # this signatures are encrypted
828 if 'url_encoded_fmt_stream_map' not in args:
829 raise ValueError(u'No stream_map present') # caught below
830 re_signature = re.compile(r'[&,]s=')
831 m_s = re_signature.search(args['url_encoded_fmt_stream_map'])
832 if m_s is not None:
833 self.to_screen(u'%s: Encrypted signatures detected.' % video_id)
834 video_info['url_encoded_fmt_stream_map'] = [args['url_encoded_fmt_stream_map']]
835 m_s = re_signature.search(args.get('adaptive_fmts', u''))
836 if m_s is not None:
837 if 'adaptive_fmts' in video_info:
838 video_info['adaptive_fmts'][0] += ',' + args['adaptive_fmts']
839 else:
840 video_info['adaptive_fmts'] = [args['adaptive_fmts']]
841 except ValueError:
842 pass
843
844 def _map_to_format_list(urlmap):
845 formats = []
846 for itag, video_real_url in urlmap.items():
847 dct = {
848 'format_id': itag,
849 'url': video_real_url,
850 'player_url': player_url,
851 }
852 if itag in self._formats:
853 dct.update(self._formats[itag])
854 formats.append(dct)
855 return formats
856
857 if 'conn' in video_info and video_info['conn'][0].startswith('rtmp'):
858 self.report_rtmp_download()
859 formats = [{
860 'format_id': '_rtmp',
861 'protocol': 'rtmp',
862 'url': video_info['conn'][0],
863 'player_url': player_url,
864 }]
865 elif len(video_info.get('url_encoded_fmt_stream_map', [])) >= 1 or len(video_info.get('adaptive_fmts', [])) >= 1:
866 encoded_url_map = video_info.get('url_encoded_fmt_stream_map', [''])[0] + ',' + video_info.get('adaptive_fmts',[''])[0]
867 if 'rtmpe%3Dyes' in encoded_url_map:
868 raise ExtractorError('rtmpe downloads are not supported, see https://github.com/rg3/youtube-dl/issues/343 for more information.', expected=True)
869 url_map = {}
870 for url_data_str in encoded_url_map.split(','):
871 url_data = compat_parse_qs(url_data_str)
872 if 'itag' not in url_data or 'url' not in url_data:
873 continue
874 format_id = url_data['itag'][0]
875 url = url_data['url'][0]
876
877 if 'sig' in url_data:
878 url += '&signature=' + url_data['sig'][0]
879 elif 's' in url_data:
880 encrypted_sig = url_data['s'][0]
881
882 if not age_gate:
883 jsplayer_url_json = self._search_regex(
884 r'"assets":.+?"js":\s*("[^"]+")',
885 video_webpage, u'JS player URL')
886 player_url = json.loads(jsplayer_url_json)
887 if player_url is None:
888 player_url_json = self._search_regex(
889 r'ytplayer\.config.*?"url"\s*:\s*("[^"]+")',
890 video_webpage, u'age gate player URL')
891 player_url = json.loads(player_url_json)
892
893 if self._downloader.params.get('verbose'):
894 if player_url is None:
895 player_version = 'unknown'
896 player_desc = 'unknown'
897 else:
898 if player_url.endswith('swf'):
899 player_version = self._search_regex(
900 r'-(.+?)(?:/watch_as3)?\.swf$', player_url,
901 u'flash player', fatal=False)
902 player_desc = 'flash player %s' % player_version
903 else:
904 player_version = self._search_regex(
905 r'html5player-([^/]+?)(?:/html5player)?\.js',
906 player_url,
907 'html5 player', fatal=False)
908 player_desc = u'html5 player %s' % player_version
909
910 parts_sizes = self._signature_cache_id(encrypted_sig)
911 self.to_screen(u'{%s} signature length %s, %s' %
912 (format_id, parts_sizes, player_desc))
913
914 signature = self._decrypt_signature(
915 encrypted_sig, video_id, player_url, age_gate)
916 url += '&signature=' + signature
917 if 'ratebypass' not in url:
918 url += '&ratebypass=yes'
919 url_map[format_id] = url
920 formats = _map_to_format_list(url_map)
921 elif video_info.get('hlsvp'):
922 manifest_url = video_info['hlsvp'][0]
923 url_map = self._extract_from_m3u8(manifest_url, video_id)
924 formats = _map_to_format_list(url_map)
925 else:
926 raise ExtractorError(u'no conn, hlsvp or url_encoded_fmt_stream_map information found in video info')
927
928 # Look for the DASH manifest
929 if (self._downloader.params.get('youtube_include_dash_manifest', False)):
930 try:
931 # The DASH manifest used needs to be the one from the original video_webpage.
932 # The one found in get_video_info seems to be using different signatures.
933 # However, in the case of an age restriction there won't be any embedded dashmpd in the video_webpage.
934 # Luckily, it seems, this case uses some kind of default signature (len == 86), so the
935 # combination of get_video_info and the _static_decrypt_signature() decryption fallback will work here.
936 if age_gate:
937 dash_manifest_url = video_info.get('dashmpd')[0]
938 else:
939 dash_manifest_url = ytplayer_config['args']['dashmpd']
940 def decrypt_sig(mobj):
941 s = mobj.group(1)
942 dec_s = self._decrypt_signature(s, video_id, player_url, age_gate)
943 return '/signature/%s' % dec_s
944 dash_manifest_url = re.sub(r'/s/([\w\.]+)', decrypt_sig, dash_manifest_url)
945 dash_doc = self._download_xml(
946 dash_manifest_url, video_id,
947 note=u'Downloading DASH manifest',
948 errnote=u'Could not download DASH manifest')
949 for r in dash_doc.findall(u'.//{urn:mpeg:DASH:schema:MPD:2011}Representation'):
950 url_el = r.find('{urn:mpeg:DASH:schema:MPD:2011}BaseURL')
951 if url_el is None:
952 continue
953 format_id = r.attrib['id']
954 video_url = url_el.text
955 filesize = int_or_none(url_el.attrib.get('{http://youtube.com/yt/2012/10/10}contentLength'))
956 f = {
957 'format_id': format_id,
958 'url': video_url,
959 'width': int_or_none(r.attrib.get('width')),
960 'tbr': int_or_none(r.attrib.get('bandwidth'), 1000),
961 'asr': int_or_none(r.attrib.get('audioSamplingRate')),
962 'filesize': filesize,
963 }
964 try:
965 existing_format = next(
966 fo for fo in formats
967 if fo['format_id'] == format_id)
968 except StopIteration:
969 f.update(self._formats.get(format_id, {}))
970 formats.append(f)
971 else:
972 existing_format.update(f)
973
974 except (ExtractorError, KeyError) as e:
975 self.report_warning(u'Skipping DASH manifest: %s' % e, video_id)
976
977 self._sort_formats(formats)
978
979 return {
980 'id': video_id,
981 'uploader': video_uploader,
982 'uploader_id': video_uploader_id,
983 'upload_date': upload_date,
984 'title': video_title,
985 'thumbnail': video_thumbnail,
986 'description': video_description,
987 'categories': video_categories,
988 'subtitles': video_subtitles,
989 'duration': video_duration,
990 'age_limit': 18 if age_gate else 0,
991 'annotations': video_annotations,
992 'webpage_url': proto + '://www.youtube.com/watch?v=%s' % video_id,
993 'view_count': view_count,
994 'like_count': like_count,
995 'dislike_count': dislike_count,
996 'formats': formats,
997 }
998
999 class YoutubePlaylistIE(YoutubeBaseInfoExtractor):
1000 IE_DESC = u'YouTube.com playlists'
1001 _VALID_URL = r"""(?x)(?:
1002 (?:https?://)?
1003 (?:\w+\.)?
1004 youtube\.com/
1005 (?:
1006 (?:course|view_play_list|my_playlists|artist|playlist|watch)
1007 \? (?:.*?&)*? (?:p|a|list)=
1008 | p/
1009 )
1010 (
1011 (?:PL|LL|EC|UU|FL|RD)?[0-9A-Za-z-_]{10,}
1012 # Top tracks, they can also include dots
1013 |(?:MC)[\w\.]*
1014 )
1015 .*
1016 |
1017 ((?:PL|LL|EC|UU|FL|RD)[0-9A-Za-z-_]{10,})
1018 )"""
1019 _TEMPLATE_URL = 'https://www.youtube.com/playlist?list=%s'
1020 _MORE_PAGES_INDICATOR = r'data-link-type="next"'
1021 _VIDEO_RE = r'href="\s*/watch\?v=(?P<id>[0-9A-Za-z_-]{11})&amp;[^"]*?index=(?P<index>\d+)'
1022 IE_NAME = u'youtube:playlist'
1023 _TESTS = [{
1024 'url': 'https://www.youtube.com/playlist?list=PLwiyx1dc3P2JR9N8gQaQN_BCvlSlap7re',
1025 'info_dict': {
1026 'title': 'ytdl test PL',
1027 },
1028 'playlist_count': 3,
1029 }]
1030
1031 def _real_initialize(self):
1032 self._login()
1033
1034 def _ids_to_results(self, ids):
1035 return [
1036 self.url_result(vid_id, 'Youtube', video_id=vid_id)
1037 for vid_id in ids]
1038
1039 def _extract_mix(self, playlist_id):
1040 # The mixes are generated from a a single video
1041 # the id of the playlist is just 'RD' + video_id
1042 url = 'https://youtube.com/watch?v=%s&list=%s' % (playlist_id[-11:], playlist_id)
1043 webpage = self._download_webpage(
1044 url, playlist_id, u'Downloading Youtube mix')
1045 search_title = lambda class_name: get_element_by_attribute('class', class_name, webpage)
1046 title_span = (
1047 search_title('playlist-title') or
1048 search_title('title long-title') or
1049 search_title('title'))
1050 title = clean_html(title_span)
1051 ids = orderedSet(re.findall(
1052 r'''(?xs)data-video-username=".*?".*?
1053 href="/watch\?v=([0-9A-Za-z_-]{11})&amp;[^"]*?list=%s''' % re.escape(playlist_id),
1054 webpage))
1055 url_results = self._ids_to_results(ids)
1056
1057 return self.playlist_result(url_results, playlist_id, title)
1058
1059 def _real_extract(self, url):
1060 # Extract playlist id
1061 mobj = re.match(self._VALID_URL, url)
1062 if mobj is None:
1063 raise ExtractorError(u'Invalid URL: %s' % url)
1064 playlist_id = mobj.group(1) or mobj.group(2)
1065
1066 # Check if it's a video-specific URL
1067 query_dict = compat_urlparse.parse_qs(compat_urlparse.urlparse(url).query)
1068 if 'v' in query_dict:
1069 video_id = query_dict['v'][0]
1070 if self._downloader.params.get('noplaylist'):
1071 self.to_screen(u'Downloading just video %s because of --no-playlist' % video_id)
1072 return self.url_result(video_id, 'Youtube', video_id=video_id)
1073 else:
1074 self.to_screen(u'Downloading playlist %s - add --no-playlist to just download video %s' % (playlist_id, video_id))
1075
1076 if playlist_id.startswith('RD'):
1077 # Mixes require a custom extraction process
1078 return self._extract_mix(playlist_id)
1079 if playlist_id.startswith('TL'):
1080 raise ExtractorError(u'For downloading YouTube.com top lists, use '
1081 u'the "yttoplist" keyword, for example "youtube-dl \'yttoplist:music:Top Tracks\'"', expected=True)
1082
1083 url = self._TEMPLATE_URL % playlist_id
1084 page = self._download_webpage(url, playlist_id)
1085 more_widget_html = content_html = page
1086
1087 # Check if the playlist exists or is private
1088 if re.search(r'<div class="yt-alert-message">[^<]*?(The|This) playlist (does not exist|is private)[^<]*?</div>', page) is not None:
1089 raise ExtractorError(
1090 u'The playlist doesn\'t exist or is private, use --username or '
1091 '--netrc to access it.',
1092 expected=True)
1093
1094 # Extract the video ids from the playlist pages
1095 ids = []
1096
1097 for page_num in itertools.count(1):
1098 matches = re.finditer(self._VIDEO_RE, content_html)
1099 # We remove the duplicates and the link with index 0
1100 # (it's not the first video of the playlist)
1101 new_ids = orderedSet(m.group('id') for m in matches if m.group('index') != '0')
1102 ids.extend(new_ids)
1103
1104 mobj = re.search(r'data-uix-load-more-href="/?(?P<more>[^"]+)"', more_widget_html)
1105 if not mobj:
1106 break
1107
1108 more = self._download_json(
1109 'https://youtube.com/%s' % mobj.group('more'), playlist_id,
1110 'Downloading page #%s' % page_num,
1111 transform_source=uppercase_escape)
1112 content_html = more['content_html']
1113 more_widget_html = more['load_more_widget_html']
1114
1115 playlist_title = self._html_search_regex(
1116 r'(?s)<h1 class="pl-header-title[^"]*">\s*(.*?)\s*</h1>',
1117 page, u'title')
1118
1119 url_results = self._ids_to_results(ids)
1120 return self.playlist_result(url_results, playlist_id, playlist_title)
1121
1122
1123 class YoutubeTopListIE(YoutubePlaylistIE):
1124 IE_NAME = u'youtube:toplist'
1125 IE_DESC = (u'YouTube.com top lists, "yttoplist:{channel}:{list title}"'
1126 u' (Example: "yttoplist:music:Top Tracks")')
1127 _VALID_URL = r'yttoplist:(?P<chann>.*?):(?P<title>.*?)$'
1128 _TESTS = []
1129
1130 def _real_extract(self, url):
1131 mobj = re.match(self._VALID_URL, url)
1132 channel = mobj.group('chann')
1133 title = mobj.group('title')
1134 query = compat_urllib_parse.urlencode({'title': title})
1135 playlist_re = 'href="([^"]+?%s.*?)"' % re.escape(query)
1136 channel_page = self._download_webpage('https://www.youtube.com/%s' % channel, title)
1137 link = self._html_search_regex(playlist_re, channel_page, u'list')
1138 url = compat_urlparse.urljoin('https://www.youtube.com/', link)
1139
1140 video_re = r'data-index="\d+".*?data-video-id="([0-9A-Za-z_-]{11})"'
1141 ids = []
1142 # sometimes the webpage doesn't contain the videos
1143 # retry until we get them
1144 for i in itertools.count(0):
1145 msg = u'Downloading Youtube mix'
1146 if i > 0:
1147 msg += ', retry #%d' % i
1148
1149 webpage = self._download_webpage(url, title, msg)
1150 ids = orderedSet(re.findall(video_re, webpage))
1151 if ids:
1152 break
1153 url_results = self._ids_to_results(ids)
1154 return self.playlist_result(url_results, playlist_title=title)
1155
1156
1157 class YoutubeChannelIE(InfoExtractor):
1158 IE_DESC = u'YouTube.com channels'
1159 _VALID_URL = r"^(?:https?://)?(?:youtu\.be|(?:\w+\.)?youtube(?:-nocookie)?\.com)/channel/([0-9A-Za-z_-]+)"
1160 _MORE_PAGES_INDICATOR = 'yt-uix-load-more'
1161 _MORE_PAGES_URL = 'https://www.youtube.com/c4_browse_ajax?action_load_more_videos=1&flow=list&paging=%s&view=0&sort=da&channel_id=%s'
1162 IE_NAME = u'youtube:channel'
1163
1164 def extract_videos_from_page(self, page):
1165 ids_in_page = []
1166 for mobj in re.finditer(r'href="/watch\?v=([0-9A-Za-z_-]+)&?', page):
1167 if mobj.group(1) not in ids_in_page:
1168 ids_in_page.append(mobj.group(1))
1169 return ids_in_page
1170
1171 def _real_extract(self, url):
1172 # Extract channel id
1173 mobj = re.match(self._VALID_URL, url)
1174 if mobj is None:
1175 raise ExtractorError(u'Invalid URL: %s' % url)
1176
1177 # Download channel page
1178 channel_id = mobj.group(1)
1179 video_ids = []
1180 url = 'https://www.youtube.com/channel/%s/videos' % channel_id
1181 channel_page = self._download_webpage(url, channel_id)
1182 autogenerated = re.search(r'''(?x)
1183 class="[^"]*?(?:
1184 channel-header-autogenerated-label|
1185 yt-channel-title-autogenerated
1186 )[^"]*"''', channel_page) is not None
1187
1188 if autogenerated:
1189 # The videos are contained in a single page
1190 # the ajax pages can't be used, they are empty
1191 video_ids = self.extract_videos_from_page(channel_page)
1192 else:
1193 # Download all channel pages using the json-based channel_ajax query
1194 for pagenum in itertools.count(1):
1195 url = self._MORE_PAGES_URL % (pagenum, channel_id)
1196 page = self._download_json(
1197 url, channel_id, note=u'Downloading page #%s' % pagenum,
1198 transform_source=uppercase_escape)
1199
1200 ids_in_page = self.extract_videos_from_page(page['content_html'])
1201 video_ids.extend(ids_in_page)
1202
1203 if self._MORE_PAGES_INDICATOR not in page['load_more_widget_html']:
1204 break
1205
1206 self._downloader.to_screen(u'[youtube] Channel %s: Found %i videos' % (channel_id, len(video_ids)))
1207
1208 url_entries = [self.url_result(video_id, 'Youtube', video_id=video_id)
1209 for video_id in video_ids]
1210 return self.playlist_result(url_entries, channel_id)
1211
1212
1213 class YoutubeUserIE(InfoExtractor):
1214 IE_DESC = u'YouTube.com user videos (URL or "ytuser" keyword)'
1215 _VALID_URL = r'(?:(?:(?:https?://)?(?:\w+\.)?youtube\.com/(?:user/)?(?!(?:attribution_link|watch|results)(?:$|[^a-z_A-Z0-9-])))|ytuser:)(?!feed/)([A-Za-z0-9_-]+)'
1216 _TEMPLATE_URL = 'https://gdata.youtube.com/feeds/api/users/%s'
1217 _GDATA_PAGE_SIZE = 50
1218 _GDATA_URL = 'https://gdata.youtube.com/feeds/api/users/%s/uploads?max-results=%d&start-index=%d&alt=json'
1219 IE_NAME = u'youtube:user'
1220
1221 @classmethod
1222 def suitable(cls, url):
1223 # Don't return True if the url can be extracted with other youtube
1224 # extractor, the regex would is too permissive and it would match.
1225 other_ies = iter(klass for (name, klass) in globals().items() if name.endswith('IE') and klass is not cls)
1226 if any(ie.suitable(url) for ie in other_ies): return False
1227 else: return super(YoutubeUserIE, cls).suitable(url)
1228
1229 def _real_extract(self, url):
1230 # Extract username
1231 mobj = re.match(self._VALID_URL, url)
1232 if mobj is None:
1233 raise ExtractorError(u'Invalid URL: %s' % url)
1234
1235 username = mobj.group(1)
1236
1237 # Download video ids using YouTube Data API. Result size per
1238 # query is limited (currently to 50 videos) so we need to query
1239 # page by page until there are no video ids - it means we got
1240 # all of them.
1241
1242 def download_page(pagenum):
1243 start_index = pagenum * self._GDATA_PAGE_SIZE + 1
1244
1245 gdata_url = self._GDATA_URL % (username, self._GDATA_PAGE_SIZE, start_index)
1246 page = self._download_webpage(
1247 gdata_url, username,
1248 u'Downloading video ids from %d to %d' % (
1249 start_index, start_index + self._GDATA_PAGE_SIZE))
1250
1251 try:
1252 response = json.loads(page)
1253 except ValueError as err:
1254 raise ExtractorError(u'Invalid JSON in API response: ' + compat_str(err))
1255 if 'entry' not in response['feed']:
1256 return
1257
1258 # Extract video identifiers
1259 entries = response['feed']['entry']
1260 for entry in entries:
1261 title = entry['title']['$t']
1262 video_id = entry['id']['$t'].split('/')[-1]
1263 yield {
1264 '_type': 'url',
1265 'url': video_id,
1266 'ie_key': 'Youtube',
1267 'id': video_id,
1268 'title': title,
1269 }
1270 url_results = PagedList(download_page, self._GDATA_PAGE_SIZE)
1271
1272 return self.playlist_result(url_results, playlist_title=username)
1273
1274
1275 class YoutubeSearchIE(SearchInfoExtractor):
1276 IE_DESC = u'YouTube.com searches'
1277 _API_URL = u'https://gdata.youtube.com/feeds/api/videos?q=%s&start-index=%i&max-results=50&v=2&alt=jsonc'
1278 _MAX_RESULTS = 1000
1279 IE_NAME = u'youtube:search'
1280 _SEARCH_KEY = 'ytsearch'
1281
1282 def _get_n_results(self, query, n):
1283 """Get a specified number of results for a query"""
1284
1285 video_ids = []
1286 pagenum = 0
1287 limit = n
1288 PAGE_SIZE = 50
1289
1290 while (PAGE_SIZE * pagenum) < limit:
1291 result_url = self._API_URL % (
1292 compat_urllib_parse.quote_plus(query.encode('utf-8')),
1293 (PAGE_SIZE * pagenum) + 1)
1294 data_json = self._download_webpage(
1295 result_url, video_id=u'query "%s"' % query,
1296 note=u'Downloading page %s' % (pagenum + 1),
1297 errnote=u'Unable to download API page')
1298 data = json.loads(data_json)
1299 api_response = data['data']
1300
1301 if 'items' not in api_response:
1302 raise ExtractorError(
1303 u'[youtube] No video results', expected=True)
1304
1305 new_ids = list(video['id'] for video in api_response['items'])
1306 video_ids += new_ids
1307
1308 limit = min(n, api_response['totalItems'])
1309 pagenum += 1
1310
1311 if len(video_ids) > n:
1312 video_ids = video_ids[:n]
1313 videos = [self.url_result(video_id, 'Youtube', video_id=video_id)
1314 for video_id in video_ids]
1315 return self.playlist_result(videos, query)
1316
1317
1318 class YoutubeSearchDateIE(YoutubeSearchIE):
1319 IE_NAME = YoutubeSearchIE.IE_NAME + ':date'
1320 _API_URL = 'https://gdata.youtube.com/feeds/api/videos?q=%s&start-index=%i&max-results=50&v=2&alt=jsonc&orderby=published'
1321 _SEARCH_KEY = 'ytsearchdate'
1322 IE_DESC = u'YouTube.com searches, newest videos first'
1323
1324
1325 class YoutubeSearchURLIE(InfoExtractor):
1326 IE_DESC = u'YouTube.com search URLs'
1327 IE_NAME = u'youtube:search_url'
1328 _VALID_URL = r'https?://(?:www\.)?youtube\.com/results\?(.*?&)?search_query=(?P<query>[^&]+)(?:[&]|$)'
1329
1330 def _real_extract(self, url):
1331 mobj = re.match(self._VALID_URL, url)
1332 query = compat_urllib_parse.unquote_plus(mobj.group('query'))
1333
1334 webpage = self._download_webpage(url, query)
1335 result_code = self._search_regex(
1336 r'(?s)<ol class="item-section"(.*?)</ol>', webpage, u'result HTML')
1337
1338 part_codes = re.findall(
1339 r'(?s)<h3 class="yt-lockup-title">(.*?)</h3>', result_code)
1340 entries = []
1341 for part_code in part_codes:
1342 part_title = self._html_search_regex(
1343 [r'(?s)title="([^"]+)"', r'>([^<]+)</a>'], part_code, 'item title', fatal=False)
1344 part_url_snippet = self._html_search_regex(
1345 r'(?s)href="([^"]+)"', part_code, 'item URL')
1346 part_url = compat_urlparse.urljoin(
1347 'https://www.youtube.com/', part_url_snippet)
1348 entries.append({
1349 '_type': 'url',
1350 'url': part_url,
1351 'title': part_title,
1352 })
1353
1354 return {
1355 '_type': 'playlist',
1356 'entries': entries,
1357 'title': query,
1358 }
1359
1360
1361 class YoutubeShowIE(InfoExtractor):
1362 IE_DESC = u'YouTube.com (multi-season) shows'
1363 _VALID_URL = r'https?://www\.youtube\.com/show/(.*)'
1364 IE_NAME = u'youtube:show'
1365
1366 def _real_extract(self, url):
1367 mobj = re.match(self._VALID_URL, url)
1368 show_name = mobj.group(1)
1369 webpage = self._download_webpage(url, show_name, u'Downloading show webpage')
1370 # There's one playlist for each season of the show
1371 m_seasons = list(re.finditer(r'href="(/playlist\?list=.*?)"', webpage))
1372 self.to_screen(u'%s: Found %s seasons' % (show_name, len(m_seasons)))
1373 return [self.url_result('https://www.youtube.com' + season.group(1), 'YoutubePlaylist') for season in m_seasons]
1374
1375
1376 class YoutubeFeedsInfoExtractor(YoutubeBaseInfoExtractor):
1377 """
1378 Base class for extractors that fetch info from
1379 http://www.youtube.com/feed_ajax
1380 Subclasses must define the _FEED_NAME and _PLAYLIST_TITLE properties.
1381 """
1382 _LOGIN_REQUIRED = True
1383 # use action_load_personal_feed instead of action_load_system_feed
1384 _PERSONAL_FEED = False
1385
1386 @property
1387 def _FEED_TEMPLATE(self):
1388 action = 'action_load_system_feed'
1389 if self._PERSONAL_FEED:
1390 action = 'action_load_personal_feed'
1391 return 'https://www.youtube.com/feed_ajax?%s=1&feed_name=%s&paging=%%s' % (action, self._FEED_NAME)
1392
1393 @property
1394 def IE_NAME(self):
1395 return u'youtube:%s' % self._FEED_NAME
1396
1397 def _real_initialize(self):
1398 self._login()
1399
1400 def _real_extract(self, url):
1401 feed_entries = []
1402 paging = 0
1403 for i in itertools.count(1):
1404 info = self._download_json(self._FEED_TEMPLATE % paging,
1405 u'%s feed' % self._FEED_NAME,
1406 u'Downloading page %s' % i)
1407 feed_html = info.get('feed_html') or info.get('content_html')
1408 load_more_widget_html = info.get('load_more_widget_html') or feed_html
1409 m_ids = re.finditer(r'"/watch\?v=(.*?)["&]', feed_html)
1410 ids = orderedSet(m.group(1) for m in m_ids)
1411 feed_entries.extend(
1412 self.url_result(video_id, 'Youtube', video_id=video_id)
1413 for video_id in ids)
1414 mobj = re.search(
1415 r'data-uix-load-more-href="/?[^"]+paging=(?P<paging>\d+)',
1416 load_more_widget_html)
1417 if mobj is None:
1418 break
1419 paging = mobj.group('paging')
1420 return self.playlist_result(feed_entries, playlist_title=self._PLAYLIST_TITLE)
1421
1422 class YoutubeRecommendedIE(YoutubeFeedsInfoExtractor):
1423 IE_DESC = u'YouTube.com recommended videos, "ytrec" keyword (requires authentication)'
1424 _VALID_URL = r'https?://www\.youtube\.com/feed/recommended|:ytrec(?:ommended)?'
1425 _FEED_NAME = 'recommended'
1426 _PLAYLIST_TITLE = u'Youtube Recommended videos'
1427
1428 class YoutubeWatchLaterIE(YoutubeFeedsInfoExtractor):
1429 IE_DESC = u'Youtube watch later list, "ytwatchlater" keyword (requires authentication)'
1430 _VALID_URL = r'https?://www\.youtube\.com/feed/watch_later|:ytwatchlater'
1431 _FEED_NAME = 'watch_later'
1432 _PLAYLIST_TITLE = u'Youtube Watch Later'
1433 _PERSONAL_FEED = True
1434
1435 class YoutubeHistoryIE(YoutubeFeedsInfoExtractor):
1436 IE_DESC = u'Youtube watch history, "ythistory" keyword (requires authentication)'
1437 _VALID_URL = u'https?://www\.youtube\.com/feed/history|:ythistory'
1438 _FEED_NAME = 'history'
1439 _PERSONAL_FEED = True
1440 _PLAYLIST_TITLE = u'Youtube Watch History'
1441
1442 class YoutubeFavouritesIE(YoutubeBaseInfoExtractor):
1443 IE_NAME = u'youtube:favorites'
1444 IE_DESC = u'YouTube.com favourite videos, "ytfav" keyword (requires authentication)'
1445 _VALID_URL = r'https?://www\.youtube\.com/my_favorites|:ytfav(?:ou?rites)?'
1446 _LOGIN_REQUIRED = True
1447
1448 def _real_extract(self, url):
1449 webpage = self._download_webpage('https://www.youtube.com/my_favorites', 'Youtube Favourites videos')
1450 playlist_id = self._search_regex(r'list=(.+?)["&]', webpage, u'favourites playlist id')
1451 return self.url_result(playlist_id, 'YoutubePlaylist')
1452
1453
1454 class YoutubeSubscriptionsIE(YoutubePlaylistIE):
1455 IE_NAME = u'youtube:subscriptions'
1456 IE_DESC = u'YouTube.com subscriptions feed, "ytsubs" keyword (requires authentication)'
1457 _VALID_URL = r'https?://www\.youtube\.com/feed/subscriptions|:ytsubs(?:criptions)?'
1458 _TESTS = []
1459
1460 def _real_extract(self, url):
1461 title = u'Youtube Subscriptions'
1462 page = self._download_webpage('https://www.youtube.com/feed/subscriptions', title)
1463
1464 # The extraction process is the same as for playlists, but the regex
1465 # for the video ids doesn't contain an index
1466 ids = []
1467 more_widget_html = content_html = page
1468
1469 for page_num in itertools.count(1):
1470 matches = re.findall(r'href="\s*/watch\?v=([0-9A-Za-z_-]{11})', content_html)
1471 new_ids = orderedSet(matches)
1472 ids.extend(new_ids)
1473
1474 mobj = re.search(r'data-uix-load-more-href="/?(?P<more>[^"]+)"', more_widget_html)
1475 if not mobj:
1476 break
1477
1478 more = self._download_json(
1479 'https://youtube.com/%s' % mobj.group('more'), title,
1480 'Downloading page #%s' % page_num,
1481 transform_source=uppercase_escape)
1482 content_html = more['content_html']
1483 more_widget_html = more['load_more_widget_html']
1484
1485 return {
1486 '_type': 'playlist',
1487 'title': title,
1488 'entries': self._ids_to_results(ids),
1489 }
1490
1491
1492 class YoutubeTruncatedURLIE(InfoExtractor):
1493 IE_NAME = 'youtube:truncated_url'
1494 IE_DESC = False # Do not list
1495 _VALID_URL = r'''(?x)
1496 (?:https?://)?[^/]+/watch\?(?:
1497 feature=[a-z_]+|
1498 annotation_id=annotation_[^&]+
1499 )?$|
1500 (?:https?://)?(?:www\.)?youtube\.com/attribution_link\?a=[^&]+$
1501 '''
1502
1503 _TESTS = [{
1504 'url': 'http://www.youtube.com/watch?annotation_id=annotation_3951667041',
1505 'only_matching': True,
1506 }, {
1507 'url': 'http://www.youtube.com/watch?',
1508 'only_matching': True,
1509 }]
1510
1511 def _real_extract(self, url):
1512 raise ExtractorError(
1513 u'Did you forget to quote the URL? Remember that & is a meta '
1514 u'character in most shells, so you want to put the URL in quotes, '
1515 u'like youtube-dl '
1516 u'"http://www.youtube.com/watch?feature=foo&v=BaW_jenozKc" '
1517 u' or simply youtube-dl BaW_jenozKc .',
1518 expected=True)
fork of https://github.com/yt-dlp/yt-dlp/