]> jfr.im git - yt-dlp.git/blob - youtube_dl/extractor/youtube.py
projects / yt-dlp.git / blob
? search:
summary | shortlog | log | commit | commitdiff | tree
blame | history | raw | HEAD
Merge branch 'peugeot-sunporno'
[yt-dlp.git] / youtube_dl / extractor / youtube.py
1 # coding: utf-8
2
3 import errno
4 import io
5 import itertools
6 import json
7 import os.path
8 import re
9 import traceback
10
11 from .common import InfoExtractor, SearchInfoExtractor
12 from .subtitles import SubtitlesInfoExtractor
13 from ..jsinterp import JSInterpreter
14 from ..swfinterp import SWFInterpreter
15 from ..utils import (
16 compat_chr,
17 compat_parse_qs,
18 compat_urllib_parse,
19 compat_urllib_request,
20 compat_urlparse,
21 compat_str,
22
23 clean_html,
24 get_cachedir,
25 get_element_by_id,
26 get_element_by_attribute,
27 ExtractorError,
28 int_or_none,
29 PagedList,
30 unescapeHTML,
31 unified_strdate,
32 orderedSet,
33 write_json_file,
34 uppercase_escape,
35 )
36
37 class YoutubeBaseInfoExtractor(InfoExtractor):
38 """Provide base functions for Youtube extractors"""
39 _LOGIN_URL = 'https://accounts.google.com/ServiceLogin'
40 _TWOFACTOR_URL = 'https://accounts.google.com/SecondFactor'
41 _LANG_URL = r'https://www.youtube.com/?hl=en&persist_hl=1&gl=US&persist_gl=1&opt_out_ackd=1'
42 _AGE_URL = 'https://www.youtube.com/verify_age?next_url=/&gl=US&hl=en'
43 _NETRC_MACHINE = 'youtube'
44 # If True it will raise an error if no login info is provided
45 _LOGIN_REQUIRED = False
46
47 def _set_language(self):
48 return bool(self._download_webpage(
49 self._LANG_URL, None,
50 note=u'Setting language', errnote='unable to set language',
51 fatal=False))
52
53 def _login(self):
54 """
55 Attempt to log in to YouTube.
56 True is returned if successful or skipped.
57 False is returned if login failed.
58
59 If _LOGIN_REQUIRED is set and no authentication was provided, an error is raised.
60 """
61 (username, password) = self._get_login_info()
62 # No authentication to be performed
63 if username is None:
64 if self._LOGIN_REQUIRED:
65 raise ExtractorError(u'No login info available, needed for using %s.' % self.IE_NAME, expected=True)
66 return True
67
68 login_page = self._download_webpage(
69 self._LOGIN_URL, None,
70 note=u'Downloading login page',
71 errnote=u'unable to fetch login page', fatal=False)
72 if login_page is False:
73 return
74
75 galx = self._search_regex(r'(?s)<input.+?name="GALX".+?value="(.+?)"',
76 login_page, u'Login GALX parameter')
77
78 # Log in
79 login_form_strs = {
80 u'continue': u'https://www.youtube.com/signin?action_handle_signin=true&feature=sign_in_button&hl=en_US&nomobiletemp=1',
81 u'Email': username,
82 u'GALX': galx,
83 u'Passwd': password,
84
85 u'PersistentCookie': u'yes',
86 u'_utf8': u'霱',
87 u'bgresponse': u'js_disabled',
88 u'checkConnection': u'',
89 u'checkedDomains': u'youtube',
90 u'dnConn': u'',
91 u'pstMsg': u'0',
92 u'rmShown': u'1',
93 u'secTok': u'',
94 u'signIn': u'Sign in',
95 u'timeStmp': u'',
96 u'service': u'youtube',
97 u'uilel': u'3',
98 u'hl': u'en_US',
99 }
100
101 # Convert to UTF-8 *before* urlencode because Python 2.x's urlencode
102 # chokes on unicode
103 login_form = dict((k.encode('utf-8'), v.encode('utf-8')) for k,v in login_form_strs.items())
104 login_data = compat_urllib_parse.urlencode(login_form).encode('ascii')
105
106 req = compat_urllib_request.Request(self._LOGIN_URL, login_data)
107 login_results = self._download_webpage(
108 req, None,
109 note=u'Logging in', errnote=u'unable to log in', fatal=False)
110 if login_results is False:
111 return False
112
113 if re.search(r'id="errormsg_0_Passwd"', login_results) is not None:
114 raise ExtractorError(u'Please use your account password and a two-factor code instead of an application-specific password.', expected=True)
115
116 # Two-Factor
117 # TODO add SMS and phone call support - these require making a request and then prompting the user
118
119 if re.search(r'(?i)<form[^>]* id="gaia_secondfactorform"', login_results) is not None:
120 tfa_code = self._get_tfa_info()
121
122 if tfa_code is None:
123 self._downloader.report_warning(u'Two-factor authentication required. Provide it with --twofactor <code>')
124 self._downloader.report_warning(u'(Note that only TOTP (Google Authenticator App) codes work at this time.)')
125 return False
126
127 # Unlike the first login form, secTok and timeStmp are both required for the TFA form
128
129 match = re.search(r'id="secTok"\n\s+value=\'(.+)\'/>', login_results, re.M | re.U)
130 if match is None:
131 self._downloader.report_warning(u'Failed to get secTok - did the page structure change?')
132 secTok = match.group(1)
133 match = re.search(r'id="timeStmp"\n\s+value=\'(.+)\'/>', login_results, re.M | re.U)
134 if match is None:
135 self._downloader.report_warning(u'Failed to get timeStmp - did the page structure change?')
136 timeStmp = match.group(1)
137
138 tfa_form_strs = {
139 u'continue': u'https://www.youtube.com/signin?action_handle_signin=true&feature=sign_in_button&hl=en_US&nomobiletemp=1',
140 u'smsToken': u'',
141 u'smsUserPin': tfa_code,
142 u'smsVerifyPin': u'Verify',
143
144 u'PersistentCookie': u'yes',
145 u'checkConnection': u'',
146 u'checkedDomains': u'youtube',
147 u'pstMsg': u'1',
148 u'secTok': secTok,
149 u'timeStmp': timeStmp,
150 u'service': u'youtube',
151 u'hl': u'en_US',
152 }
153 tfa_form = dict((k.encode('utf-8'), v.encode('utf-8')) for k,v in tfa_form_strs.items())
154 tfa_data = compat_urllib_parse.urlencode(tfa_form).encode('ascii')
155
156 tfa_req = compat_urllib_request.Request(self._TWOFACTOR_URL, tfa_data)
157 tfa_results = self._download_webpage(
158 tfa_req, None,
159 note=u'Submitting TFA code', errnote=u'unable to submit tfa', fatal=False)
160
161 if tfa_results is False:
162 return False
163
164 if re.search(r'(?i)<form[^>]* id="gaia_secondfactorform"', tfa_results) is not None:
165 self._downloader.report_warning(u'Two-factor code expired. Please try again, or use a one-use backup code instead.')
166 return False
167 if re.search(r'(?i)<form[^>]* id="gaia_loginform"', tfa_results) is not None:
168 self._downloader.report_warning(u'unable to log in - did the page structure change?')
169 return False
170 if re.search(r'smsauth-interstitial-reviewsettings', tfa_results) is not None:
171 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.')
172 return False
173
174 if re.search(r'(?i)<form[^>]* id="gaia_loginform"', login_results) is not None:
175 self._downloader.report_warning(u'unable to log in: bad username or password')
176 return False
177 return True
178
179 def _confirm_age(self):
180 age_form = {
181 'next_url': '/',
182 'action_confirm': 'Confirm',
183 }
184 req = compat_urllib_request.Request(self._AGE_URL,
185 compat_urllib_parse.urlencode(age_form).encode('ascii'))
186
187 self._download_webpage(
188 req, None,
189 note=u'Confirming age', errnote=u'Unable to confirm age')
190 return True
191
192 def _real_initialize(self):
193 if self._downloader is None:
194 return
195 if not self._set_language():
196 return
197 if not self._login():
198 return
199 self._confirm_age()
200
201
202 class YoutubeIE(YoutubeBaseInfoExtractor, SubtitlesInfoExtractor):
203 IE_DESC = u'YouTube.com'
204 _VALID_URL = r"""(?x)^
205 (
206 (?:https?://|//)? # http(s):// or protocol-independent URL (optional)
207 (?:(?:(?:(?:\w+\.)?[yY][oO][uU][tT][uU][bB][eE](?:-nocookie)?\.com/|
208 (?:www\.)?deturl\.com/www\.youtube\.com/|
209 (?:www\.)?pwnyoutube\.com/|
210 (?:www\.)?yourepeat\.com/|
211 tube\.majestyc\.net/|
212 youtube\.googleapis\.com/) # the various hostnames, with wildcard subdomains
213 (?:.*?\#/)? # handle anchor (#/) redirect urls
214 (?: # the various things that can precede the ID:
215 (?:(?:v|embed|e)/) # v/ or embed/ or e/
216 |(?: # or the v= param in all its forms
217 (?:(?:watch|movie)(?:_popup)?(?:\.php)?/?)? # preceding watch(_popup|.php) or nothing (like /?v=xxxx)
218 (?:\?|\#!?) # the params delimiter ? or # or #!
219 (?:.*?&)? # any other preceding param (like /?s=tuff&v=xxxx)
220 v=
221 )
222 ))
223 |youtu\.be/ # just youtu.be/xxxx
224 |https?://(?:www\.)?cleanvideosearch\.com/media/action/yt/watch\?videoId=
225 )
226 )? # all until now is optional -> you can pass the naked ID
227 ([0-9A-Za-z_-]{11}) # here is it! the YouTube video ID
228 (?(1).+)? # if we found the ID, everything can follow
229 $"""
230 _NEXT_URL_RE = r'[\?&]next_url=([^&]+)'
231 _formats = {
232 '5': {'ext': 'flv', 'width': 400, 'height': 240},
233 '6': {'ext': 'flv', 'width': 450, 'height': 270},
234 '13': {'ext': '3gp'},
235 '17': {'ext': '3gp', 'width': 176, 'height': 144},
236 '18': {'ext': 'mp4', 'width': 640, 'height': 360},
237 '22': {'ext': 'mp4', 'width': 1280, 'height': 720},
238 '34': {'ext': 'flv', 'width': 640, 'height': 360},
239 '35': {'ext': 'flv', 'width': 854, 'height': 480},
240 '36': {'ext': '3gp', 'width': 320, 'height': 240},
241 '37': {'ext': 'mp4', 'width': 1920, 'height': 1080},
242 '38': {'ext': 'mp4', 'width': 4096, 'height': 3072},
243 '43': {'ext': 'webm', 'width': 640, 'height': 360},
244 '44': {'ext': 'webm', 'width': 854, 'height': 480},
245 '45': {'ext': 'webm', 'width': 1280, 'height': 720},
246 '46': {'ext': 'webm', 'width': 1920, 'height': 1080},
247
248
249 # 3d videos
250 '82': {'ext': 'mp4', 'height': 360, 'format_note': '3D', 'preference': -20},
251 '83': {'ext': 'mp4', 'height': 480, 'format_note': '3D', 'preference': -20},
252 '84': {'ext': 'mp4', 'height': 720, 'format_note': '3D', 'preference': -20},
253 '85': {'ext': 'mp4', 'height': 1080, 'format_note': '3D', 'preference': -20},
254 '100': {'ext': 'webm', 'height': 360, 'format_note': '3D', 'preference': -20},
255 '101': {'ext': 'webm', 'height': 480, 'format_note': '3D', 'preference': -20},
256 '102': {'ext': 'webm', 'height': 720, 'format_note': '3D', 'preference': -20},
257
258 # Apple HTTP Live Streaming
259 '92': {'ext': 'mp4', 'height': 240, 'format_note': 'HLS', 'preference': -10},
260 '93': {'ext': 'mp4', 'height': 360, 'format_note': 'HLS', 'preference': -10},
261 '94': {'ext': 'mp4', 'height': 480, 'format_note': 'HLS', 'preference': -10},
262 '95': {'ext': 'mp4', 'height': 720, 'format_note': 'HLS', 'preference': -10},
263 '96': {'ext': 'mp4', 'height': 1080, 'format_note': 'HLS', 'preference': -10},
264 '132': {'ext': 'mp4', 'height': 240, 'format_note': 'HLS', 'preference': -10},
265 '151': {'ext': 'mp4', 'height': 72, 'format_note': 'HLS', 'preference': -10},
266
267 # DASH mp4 video
268 '133': {'ext': 'mp4', 'height': 240, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
269 '134': {'ext': 'mp4', 'height': 360, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
270 '135': {'ext': 'mp4', 'height': 480, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
271 '136': {'ext': 'mp4', 'height': 720, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
272 '137': {'ext': 'mp4', 'height': 1080, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
273 '138': {'ext': 'mp4', 'height': 2160, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
274 '160': {'ext': 'mp4', 'height': 144, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
275 '264': {'ext': 'mp4', 'height': 1440, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
276
277 # Dash mp4 audio
278 '139': {'ext': 'm4a', 'format_note': 'DASH audio', 'vcodec': 'none', 'abr': 48, 'preference': -50},
279 '140': {'ext': 'm4a', 'format_note': 'DASH audio', 'vcodec': 'none', 'abr': 128, 'preference': -50},
280 '141': {'ext': 'm4a', 'format_note': 'DASH audio', 'vcodec': 'none', 'abr': 256, 'preference': -50},
281
282 # Dash webm
283 '167': {'ext': 'webm', 'height': 360, 'width': 640, 'format_note': 'DASH video', 'acodec': 'none', 'container': 'webm', 'vcodec': 'VP8', 'preference': -40},
284 '168': {'ext': 'webm', 'height': 480, 'width': 854, 'format_note': 'DASH video', 'acodec': 'none', 'container': 'webm', 'vcodec': 'VP8', 'preference': -40},
285 '169': {'ext': 'webm', 'height': 720, 'width': 1280, 'format_note': 'DASH video', 'acodec': 'none', 'container': 'webm', 'vcodec': 'VP8', 'preference': -40},
286 '170': {'ext': 'webm', 'height': 1080, 'width': 1920, 'format_note': 'DASH video', 'acodec': 'none', 'container': 'webm', 'vcodec': 'VP8', 'preference': -40},
287 '218': {'ext': 'webm', 'height': 480, 'width': 854, 'format_note': 'DASH video', 'acodec': 'none', 'container': 'webm', 'vcodec': 'VP8', 'preference': -40},
288 '219': {'ext': 'webm', 'height': 480, 'width': 854, 'format_note': 'DASH video', 'acodec': 'none', 'container': 'webm', 'vcodec': 'VP8', 'preference': -40},
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
299 # Dash webm audio
300 '171': {'ext': 'webm', 'vcodec': 'none', 'format_note': 'DASH audio', 'abr': 128, 'preference': -50},
301 '172': {'ext': 'webm', 'vcodec': 'none', 'format_note': 'DASH audio', 'abr': 256, 'preference': -50},
302
303 # RTMP (unnamed)
304 '_rtmp': {'protocol': 'rtmp'},
305 }
306
307 IE_NAME = u'youtube'
308 _TESTS = [
309 {
310 u"url": u"http://www.youtube.com/watch?v=BaW_jenozKc",
311 u"file": u"BaW_jenozKc.mp4",
312 u"info_dict": {
313 u"title": u"youtube-dl test video \"'/\\ä↭𝕐",
314 u"uploader": u"Philipp Hagemeister",
315 u"uploader_id": u"phihag",
316 u"upload_date": u"20121002",
317 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 .",
318 u"categories": [u'Science & Technology'],
319 }
320 },
321 {
322 u"url": u"http://www.youtube.com/watch?v=UxxajLWwzqY",
323 u"file": u"UxxajLWwzqY.mp4",
324 u"note": u"Test generic use_cipher_signature video (#897)",
325 u"info_dict": {
326 u"upload_date": u"20120506",
327 u"title": u"Icona Pop - I Love It (feat. Charli XCX) [OFFICIAL VIDEO]",
328 u"description": u"md5:fea86fda2d5a5784273df5c7cc994d9f",
329 u"uploader": u"Icona Pop",
330 u"uploader_id": u"IconaPop"
331 }
332 },
333 {
334 u"url": u"https://www.youtube.com/watch?v=07FYdnEawAQ",
335 u"file": u"07FYdnEawAQ.mp4",
336 u"note": u"Test VEVO video with age protection (#956)",
337 u"info_dict": {
338 u"upload_date": u"20130703",
339 u"title": u"Justin Timberlake - Tunnel Vision (Explicit)",
340 u"description": u"md5:64249768eec3bc4276236606ea996373",
341 u"uploader": u"justintimberlakeVEVO",
342 u"uploader_id": u"justintimberlakeVEVO"
343 }
344 },
345 {
346 u"url": u"//www.YouTube.com/watch?v=yZIXLfi8CZQ",
347 u"file": u"yZIXLfi8CZQ.mp4",
348 u"note": u"Embed-only video (#1746)",
349 u"info_dict": {
350 u"upload_date": u"20120608",
351 u"title": u"Principal Sexually Assaults A Teacher - Episode 117 - 8th June 2012",
352 u"description": u"md5:09b78bd971f1e3e289601dfba15ca4f7",
353 u"uploader": u"SET India",
354 u"uploader_id": u"setindia"
355 }
356 },
357 {
358 u"url": u"http://www.youtube.com/watch?v=a9LDPn-MO4I",
359 u"file": u"a9LDPn-MO4I.m4a",
360 u"note": u"256k DASH audio (format 141) via DASH manifest",
361 u"info_dict": {
362 u"upload_date": "20121002",
363 u"uploader_id": "8KVIDEO",
364 u"description": "No description available.",
365 u"uploader": "8KVIDEO",
366 u"title": "UHDTV TEST 8K VIDEO.mp4"
367 },
368 u"params": {
369 u"youtube_include_dash_manifest": True,
370 u"format": "141",
371 },
372 },
373 # DASH manifest with encrypted signature
374 {
375 u'url': u'https://www.youtube.com/watch?v=IB3lcPjvWLA',
376 u'info_dict': {
377 u'id': u'IB3lcPjvWLA',
378 u'ext': u'm4a',
379 u'title': u'Afrojack - The Spark ft. Spree Wilson',
380 u'description': u'md5:9717375db5a9a3992be4668bbf3bc0a8',
381 u'uploader': u'AfrojackVEVO',
382 u'uploader_id': u'AfrojackVEVO',
383 u'upload_date': u'20131011',
384 },
385 u"params": {
386 u'youtube_include_dash_manifest': True,
387 u'format': '141',
388 },
389 },
390 ]
391
392
393 @classmethod
394 def suitable(cls, url):
395 """Receives a URL and returns True if suitable for this IE."""
396 if YoutubePlaylistIE.suitable(url): return False
397 return re.match(cls._VALID_URL, url) is not None
398
399 def __init__(self, *args, **kwargs):
400 super(YoutubeIE, self).__init__(*args, **kwargs)
401 self._player_cache = {}
402
403 def report_video_info_webpage_download(self, video_id):
404 """Report attempt to download video info webpage."""
405 self.to_screen(u'%s: Downloading video info webpage' % video_id)
406
407 def report_information_extraction(self, video_id):
408 """Report attempt to extract video information."""
409 self.to_screen(u'%s: Extracting video information' % video_id)
410
411 def report_unavailable_format(self, video_id, format):
412 """Report extracted video URL."""
413 self.to_screen(u'%s: Format %s not available' % (video_id, format))
414
415 def report_rtmp_download(self):
416 """Indicate the download will use the RTMP protocol."""
417 self.to_screen(u'RTMP download detected')
418
419 def _signature_cache_id(self, example_sig):
420 """ Return a string representation of a signature """
421 return u'.'.join(compat_str(len(part)) for part in example_sig.split('.'))
422
423 def _extract_signature_function(self, video_id, player_url, example_sig):
424 id_m = re.match(
425 r'.*-(?P<id>[a-zA-Z0-9_-]+)(?:/watch_as3|/html5player)?\.(?P<ext>[a-z]+)$',
426 player_url)
427 if not id_m:
428 raise ExtractorError('Cannot identify player %r' % player_url)
429 player_type = id_m.group('ext')
430 player_id = id_m.group('id')
431
432 # Read from filesystem cache
433 func_id = '%s_%s_%s' % (
434 player_type, player_id, self._signature_cache_id(example_sig))
435 assert os.path.basename(func_id) == func_id
436 cache_dir = get_cachedir(self._downloader.params)
437
438 cache_enabled = cache_dir is not None
439 if cache_enabled:
440 cache_fn = os.path.join(os.path.expanduser(cache_dir),
441 u'youtube-sigfuncs',
442 func_id + '.json')
443 try:
444 with io.open(cache_fn, 'r', encoding='utf-8') as cachef:
445 cache_spec = json.load(cachef)
446 return lambda s: u''.join(s[i] for i in cache_spec)
447 except IOError:
448 pass # No cache available
449 except ValueError:
450 try:
451 file_size = os.path.getsize(cache_fn)
452 except (OSError, IOError) as oe:
453 file_size = str(oe)
454 self._downloader.report_warning(
455 u'Cache %s failed (%s)' % (cache_fn, file_size))
456
457 if player_type == 'js':
458 code = self._download_webpage(
459 player_url, video_id,
460 note=u'Downloading %s player %s' % (player_type, player_id),
461 errnote=u'Download of %s failed' % player_url)
462 res = self._parse_sig_js(code)
463 elif player_type == 'swf':
464 urlh = self._request_webpage(
465 player_url, video_id,
466 note=u'Downloading %s player %s' % (player_type, player_id),
467 errnote=u'Download of %s failed' % player_url)
468 code = urlh.read()
469 res = self._parse_sig_swf(code)
470 else:
471 assert False, 'Invalid player type %r' % player_type
472
473 if cache_enabled:
474 try:
475 test_string = u''.join(map(compat_chr, range(len(example_sig))))
476 cache_res = res(test_string)
477 cache_spec = [ord(c) for c in cache_res]
478 try:
479 os.makedirs(os.path.dirname(cache_fn))
480 except OSError as ose:
481 if ose.errno != errno.EEXIST:
482 raise
483 write_json_file(cache_spec, cache_fn)
484 except Exception:
485 tb = traceback.format_exc()
486 self._downloader.report_warning(
487 u'Writing cache to %r failed: %s' % (cache_fn, tb))
488
489 return res
490
491 def _print_sig_code(self, func, example_sig):
492 def gen_sig_code(idxs):
493 def _genslice(start, end, step):
494 starts = u'' if start == 0 else str(start)
495 ends = (u':%d' % (end+step)) if end + step >= 0 else u':'
496 steps = u'' if step == 1 else (u':%d' % step)
497 return u's[%s%s%s]' % (starts, ends, steps)
498
499 step = None
500 start = '(Never used)' # Quelch pyflakes warnings - start will be
501 # set as soon as step is set
502 for i, prev in zip(idxs[1:], idxs[:-1]):
503 if step is not None:
504 if i - prev == step:
505 continue
506 yield _genslice(start, prev, step)
507 step = None
508 continue
509 if i - prev in [-1, 1]:
510 step = i - prev
511 start = prev
512 continue
513 else:
514 yield u's[%d]' % prev
515 if step is None:
516 yield u's[%d]' % i
517 else:
518 yield _genslice(start, i, step)
519
520 test_string = u''.join(map(compat_chr, range(len(example_sig))))
521 cache_res = func(test_string)
522 cache_spec = [ord(c) for c in cache_res]
523 expr_code = u' + '.join(gen_sig_code(cache_spec))
524 signature_id_tuple = '(%s)' % (
525 ', '.join(compat_str(len(p)) for p in example_sig.split('.')))
526 code = (u'if tuple(len(p) for p in s.split(\'.\')) == %s:\n'
527 u' return %s\n') % (signature_id_tuple, expr_code)
528 self.to_screen(u'Extracted signature function:\n' + code)
529
530 def _parse_sig_js(self, jscode):
531 funcname = self._search_regex(
532 r'signature=([$a-zA-Z]+)', jscode,
533 u'Initial JS player signature function name')
534
535 jsi = JSInterpreter(jscode)
536 initial_function = jsi.extract_function(funcname)
537 return lambda s: initial_function([s])
538
539 def _parse_sig_swf(self, file_contents):
540 swfi = SWFInterpreter(file_contents)
541 TARGET_CLASSNAME = u'SignatureDecipher'
542 searched_class = swfi.extract_class(TARGET_CLASSNAME)
543 initial_function = swfi.extract_function(searched_class, u'decipher')
544 return lambda s: initial_function([s])
545
546 def _decrypt_signature(self, s, video_id, player_url, age_gate=False):
547 """Turn the encrypted s field into a working signature"""
548
549 if player_url is None:
550 raise ExtractorError(u'Cannot decrypt signature without player_url')
551
552 if player_url.startswith(u'//'):
553 player_url = u'https:' + player_url
554 try:
555 player_id = (player_url, self._signature_cache_id(s))
556 if player_id not in self._player_cache:
557 func = self._extract_signature_function(
558 video_id, player_url, s
559 )
560 self._player_cache[player_id] = func
561 func = self._player_cache[player_id]
562 if self._downloader.params.get('youtube_print_sig_code'):
563 self._print_sig_code(func, s)
564 return func(s)
565 except Exception as e:
566 tb = traceback.format_exc()
567 raise ExtractorError(
568 u'Signature extraction failed: ' + tb, cause=e)
569
570 def _get_available_subtitles(self, video_id, webpage):
571 try:
572 sub_list = self._download_webpage(
573 'https://video.google.com/timedtext?hl=en&type=list&v=%s' % video_id,
574 video_id, note=False)
575 except ExtractorError as err:
576 self._downloader.report_warning(u'unable to download video subtitles: %s' % compat_str(err))
577 return {}
578 lang_list = re.findall(r'name="([^"]*)"[^>]+lang_code="([\w\-]+)"', sub_list)
579
580 sub_lang_list = {}
581 for l in lang_list:
582 lang = l[1]
583 if lang in sub_lang_list:
584 continue
585 params = compat_urllib_parse.urlencode({
586 'lang': lang,
587 'v': video_id,
588 'fmt': self._downloader.params.get('subtitlesformat', 'srt'),
589 'name': unescapeHTML(l[0]).encode('utf-8'),
590 })
591 url = u'https://www.youtube.com/api/timedtext?' + params
592 sub_lang_list[lang] = url
593 if not sub_lang_list:
594 self._downloader.report_warning(u'video doesn\'t have subtitles')
595 return {}
596 return sub_lang_list
597
598 def _get_available_automatic_caption(self, video_id, webpage):
599 """We need the webpage for getting the captions url, pass it as an
600 argument to speed up the process."""
601 sub_format = self._downloader.params.get('subtitlesformat', 'srt')
602 self.to_screen(u'%s: Looking for automatic captions' % video_id)
603 mobj = re.search(r';ytplayer.config = ({.*?});', webpage)
604 err_msg = u'Couldn\'t find automatic captions for %s' % video_id
605 if mobj is None:
606 self._downloader.report_warning(err_msg)
607 return {}
608 player_config = json.loads(mobj.group(1))
609 try:
610 args = player_config[u'args']
611 caption_url = args[u'ttsurl']
612 timestamp = args[u'timestamp']
613 # We get the available subtitles
614 list_params = compat_urllib_parse.urlencode({
615 'type': 'list',
616 'tlangs': 1,
617 'asrs': 1,
618 })
619 list_url = caption_url + '&' + list_params
620 caption_list = self._download_xml(list_url, video_id)
621 original_lang_node = caption_list.find('track')
622 if original_lang_node is None or original_lang_node.attrib.get('kind') != 'asr' :
623 self._downloader.report_warning(u'Video doesn\'t have automatic captions')
624 return {}
625 original_lang = original_lang_node.attrib['lang_code']
626
627 sub_lang_list = {}
628 for lang_node in caption_list.findall('target'):
629 sub_lang = lang_node.attrib['lang_code']
630 params = compat_urllib_parse.urlencode({
631 'lang': original_lang,
632 'tlang': sub_lang,
633 'fmt': sub_format,
634 'ts': timestamp,
635 'kind': 'asr',
636 })
637 sub_lang_list[sub_lang] = caption_url + '&' + params
638 return sub_lang_list
639 # An extractor error can be raise by the download process if there are
640 # no automatic captions but there are subtitles
641 except (KeyError, ExtractorError):
642 self._downloader.report_warning(err_msg)
643 return {}
644
645 @classmethod
646 def extract_id(cls, url):
647 mobj = re.match(cls._VALID_URL, url, re.VERBOSE)
648 if mobj is None:
649 raise ExtractorError(u'Invalid URL: %s' % url)
650 video_id = mobj.group(2)
651 return video_id
652
653 def _extract_from_m3u8(self, manifest_url, video_id):
654 url_map = {}
655 def _get_urls(_manifest):
656 lines = _manifest.split('\n')
657 urls = filter(lambda l: l and not l.startswith('#'),
658 lines)
659 return urls
660 manifest = self._download_webpage(manifest_url, video_id, u'Downloading formats manifest')
661 formats_urls = _get_urls(manifest)
662 for format_url in formats_urls:
663 itag = self._search_regex(r'itag/(\d+?)/', format_url, 'itag')
664 url_map[itag] = format_url
665 return url_map
666
667 def _extract_annotations(self, video_id):
668 url = 'https://www.youtube.com/annotations_invideo?features=1&legacy=1&video_id=%s' % video_id
669 return self._download_webpage(url, video_id, note=u'Searching for annotations.', errnote=u'Unable to download video annotations.')
670
671 def _real_extract(self, url):
672 proto = (
673 u'http' if self._downloader.params.get('prefer_insecure', False)
674 else u'https')
675
676 # Extract original video URL from URL with redirection, like age verification, using next_url parameter
677 mobj = re.search(self._NEXT_URL_RE, url)
678 if mobj:
679 url = proto + '://www.youtube.com/' + compat_urllib_parse.unquote(mobj.group(1)).lstrip('/')
680 video_id = self.extract_id(url)
681
682 # Get video webpage
683 url = proto + '://www.youtube.com/watch?v=%s&gl=US&hl=en&has_verified=1' % video_id
684 video_webpage = self._download_webpage(url, video_id)
685
686 # Attempt to extract SWF player URL
687 mobj = re.search(r'swfConfig.*?"(https?:\\/\\/.*?watch.*?-.*?\.swf)"', video_webpage)
688 if mobj is not None:
689 player_url = re.sub(r'\\(.)', r'\1', mobj.group(1))
690 else:
691 player_url = None
692
693 # Get video info
694 self.report_video_info_webpage_download(video_id)
695 if re.search(r'player-age-gate-content">', video_webpage) is not None:
696 self.report_age_confirmation()
697 age_gate = True
698 # We simulate the access to the video from www.youtube.com/v/{video_id}
699 # this can be viewed without login into Youtube
700 data = compat_urllib_parse.urlencode({
701 'video_id': video_id,
702 'eurl': 'https://youtube.googleapis.com/v/' + video_id,
703 'sts': self._search_regex(
704 r'"sts"\s*:\s*(\d+)', video_webpage, 'sts'),
705 })
706 video_info_url = proto + '://www.youtube.com/get_video_info?' + data
707 video_info_webpage = self._download_webpage(video_info_url, video_id,
708 note=False,
709 errnote='unable to download video info webpage')
710 video_info = compat_parse_qs(video_info_webpage)
711 else:
712 age_gate = False
713 for el_type in ['&el=embedded', '&el=detailpage', '&el=vevo', '']:
714 video_info_url = (proto + '://www.youtube.com/get_video_info?&video_id=%s%s&ps=default&eurl=&gl=US&hl=en'
715 % (video_id, el_type))
716 video_info_webpage = self._download_webpage(video_info_url, video_id,
717 note=False,
718 errnote='unable to download video info webpage')
719 video_info = compat_parse_qs(video_info_webpage)
720 if 'token' in video_info:
721 break
722 if 'token' not in video_info:
723 if 'reason' in video_info:
724 raise ExtractorError(
725 u'YouTube said: %s' % video_info['reason'][0],
726 expected=True, video_id=video_id)
727 else:
728 raise ExtractorError(
729 u'"token" parameter not in video info for unknown reason',
730 video_id=video_id)
731
732 if 'view_count' in video_info:
733 view_count = int(video_info['view_count'][0])
734 else:
735 view_count = None
736
737 # Check for "rental" videos
738 if 'ypc_video_rental_bar_text' in video_info and 'author' not in video_info:
739 raise ExtractorError(u'"rental" videos not supported')
740
741 # Start extracting information
742 self.report_information_extraction(video_id)
743
744 # uploader
745 if 'author' not in video_info:
746 raise ExtractorError(u'Unable to extract uploader name')
747 video_uploader = compat_urllib_parse.unquote_plus(video_info['author'][0])
748
749 # uploader_id
750 video_uploader_id = None
751 mobj = re.search(r'<link itemprop="url" href="http://www.youtube.com/(?:user|channel)/([^"]+)">', video_webpage)
752 if mobj is not None:
753 video_uploader_id = mobj.group(1)
754 else:
755 self._downloader.report_warning(u'unable to extract uploader nickname')
756
757 # title
758 if 'title' in video_info:
759 video_title = video_info['title'][0]
760 else:
761 self._downloader.report_warning(u'Unable to extract video title')
762 video_title = u'_'
763
764 # thumbnail image
765 # We try first to get a high quality image:
766 m_thumb = re.search(r'<span itemprop="thumbnail".*?href="(.*?)">',
767 video_webpage, re.DOTALL)
768 if m_thumb is not None:
769 video_thumbnail = m_thumb.group(1)
770 elif 'thumbnail_url' not in video_info:
771 self._downloader.report_warning(u'unable to extract video thumbnail')
772 video_thumbnail = None
773 else: # don't panic if we can't find it
774 video_thumbnail = compat_urllib_parse.unquote_plus(video_info['thumbnail_url'][0])
775
776 # upload date
777 upload_date = None
778 mobj = re.search(r'(?s)id="eow-date.*?>(.*?)</span>', video_webpage)
779 if mobj is None:
780 mobj = re.search(
781 r'(?s)id="watch-uploader-info".*?>.*?(?:Published|Uploaded|Streamed live) on (.*?)</strong>',
782 video_webpage)
783 if mobj is not None:
784 upload_date = ' '.join(re.sub(r'[/,-]', r' ', mobj.group(1)).split())
785 upload_date = unified_strdate(upload_date)
786
787 m_cat_container = get_element_by_id("eow-category", video_webpage)
788 if m_cat_container:
789 category = self._html_search_regex(
790 r'(?s)<a[^<]+>(.*?)</a>', m_cat_container, 'category',
791 default=None)
792 video_categories = None if category is None else [category]
793 else:
794 video_categories = None
795
796 # description
797 video_description = get_element_by_id("eow-description", video_webpage)
798 if video_description:
799 video_description = re.sub(r'''(?x)
800 <a\s+
801 (?:[a-zA-Z-]+="[^"]+"\s+)*?
802 title="([^"]+)"\s+
803 (?:[a-zA-Z-]+="[^"]+"\s+)*?
804 class="yt-uix-redirect-link"\s*>
805 [^<]+
806 </a>
807 ''', r'\1', video_description)
808 video_description = clean_html(video_description)
809 else:
810 fd_mobj = re.search(r'<meta name="description" content="([^"]+)"', video_webpage)
811 if fd_mobj:
812 video_description = unescapeHTML(fd_mobj.group(1))
813 else:
814 video_description = u''
815
816 def _extract_count(count_name):
817 count = self._search_regex(
818 r'id="watch-%s"[^>]*>.*?([\d,]+)\s*</span>' % re.escape(count_name),
819 video_webpage, count_name, default=None)
820 if count is not None:
821 return int(count.replace(',', ''))
822 return None
823 like_count = _extract_count(u'like')
824 dislike_count = _extract_count(u'dislike')
825
826 # subtitles
827 video_subtitles = self.extract_subtitles(video_id, video_webpage)
828
829 if self._downloader.params.get('listsubtitles', False):
830 self._list_available_subtitles(video_id, video_webpage)
831 return
832
833 if 'length_seconds' not in video_info:
834 self._downloader.report_warning(u'unable to extract video duration')
835 video_duration = None
836 else:
837 video_duration = int(compat_urllib_parse.unquote_plus(video_info['length_seconds'][0]))
838
839 # annotations
840 video_annotations = None
841 if self._downloader.params.get('writeannotations', False):
842 video_annotations = self._extract_annotations(video_id)
843
844 # Decide which formats to download
845 try:
846 mobj = re.search(r';ytplayer\.config\s*=\s*({.*?});', video_webpage)
847 if not mobj:
848 raise ValueError('Could not find vevo ID')
849 json_code = uppercase_escape(mobj.group(1))
850 ytplayer_config = json.loads(json_code)
851 args = ytplayer_config['args']
852 # Easy way to know if the 's' value is in url_encoded_fmt_stream_map
853 # this signatures are encrypted
854 if 'url_encoded_fmt_stream_map' not in args:
855 raise ValueError(u'No stream_map present') # caught below
856 re_signature = re.compile(r'[&,]s=')
857 m_s = re_signature.search(args['url_encoded_fmt_stream_map'])
858 if m_s is not None:
859 self.to_screen(u'%s: Encrypted signatures detected.' % video_id)
860 video_info['url_encoded_fmt_stream_map'] = [args['url_encoded_fmt_stream_map']]
861 m_s = re_signature.search(args.get('adaptive_fmts', u''))
862 if m_s is not None:
863 if 'adaptive_fmts' in video_info:
864 video_info['adaptive_fmts'][0] += ',' + args['adaptive_fmts']
865 else:
866 video_info['adaptive_fmts'] = [args['adaptive_fmts']]
867 except ValueError:
868 pass
869
870 def _map_to_format_list(urlmap):
871 formats = []
872 for itag, video_real_url in urlmap.items():
873 dct = {
874 'format_id': itag,
875 'url': video_real_url,
876 'player_url': player_url,
877 }
878 if itag in self._formats:
879 dct.update(self._formats[itag])
880 formats.append(dct)
881 return formats
882
883 if 'conn' in video_info and video_info['conn'][0].startswith('rtmp'):
884 self.report_rtmp_download()
885 formats = [{
886 'format_id': '_rtmp',
887 'protocol': 'rtmp',
888 'url': video_info['conn'][0],
889 'player_url': player_url,
890 }]
891 elif len(video_info.get('url_encoded_fmt_stream_map', [])) >= 1 or len(video_info.get('adaptive_fmts', [])) >= 1:
892 encoded_url_map = video_info.get('url_encoded_fmt_stream_map', [''])[0] + ',' + video_info.get('adaptive_fmts',[''])[0]
893 if 'rtmpe%3Dyes' in encoded_url_map:
894 raise ExtractorError('rtmpe downloads are not supported, see https://github.com/rg3/youtube-dl/issues/343 for more information.', expected=True)
895 url_map = {}
896 for url_data_str in encoded_url_map.split(','):
897 url_data = compat_parse_qs(url_data_str)
898 if 'itag' not in url_data or 'url' not in url_data:
899 continue
900 format_id = url_data['itag'][0]
901 url = url_data['url'][0]
902
903 if 'sig' in url_data:
904 url += '&signature=' + url_data['sig'][0]
905 elif 's' in url_data:
906 encrypted_sig = url_data['s'][0]
907
908 if not age_gate:
909 jsplayer_url_json = self._search_regex(
910 r'"assets":.+?"js":\s*("[^"]+")',
911 video_webpage, u'JS player URL')
912 player_url = json.loads(jsplayer_url_json)
913 if player_url is None:
914 player_url_json = self._search_regex(
915 r'ytplayer\.config.*?"url"\s*:\s*("[^"]+")',
916 video_webpage, u'age gate player URL')
917 player_url = json.loads(player_url_json)
918
919 if self._downloader.params.get('verbose'):
920 if player_url is None:
921 player_version = 'unknown'
922 player_desc = 'unknown'
923 else:
924 if player_url.endswith('swf'):
925 player_version = self._search_regex(
926 r'-(.+?)(?:/watch_as3)?\.swf$', player_url,
927 u'flash player', fatal=False)
928 player_desc = 'flash player %s' % player_version
929 else:
930 player_version = self._search_regex(
931 r'html5player-([^/]+?)(?:/html5player)?\.js',
932 player_url,
933 'html5 player', fatal=False)
934 player_desc = u'html5 player %s' % player_version
935
936 parts_sizes = self._signature_cache_id(encrypted_sig)
937 self.to_screen(u'{%s} signature length %s, %s' %
938 (format_id, parts_sizes, player_desc))
939
940 signature = self._decrypt_signature(
941 encrypted_sig, video_id, player_url, age_gate)
942 url += '&signature=' + signature
943 if 'ratebypass' not in url:
944 url += '&ratebypass=yes'
945 url_map[format_id] = url
946 formats = _map_to_format_list(url_map)
947 elif video_info.get('hlsvp'):
948 manifest_url = video_info['hlsvp'][0]
949 url_map = self._extract_from_m3u8(manifest_url, video_id)
950 formats = _map_to_format_list(url_map)
951 else:
952 raise ExtractorError(u'no conn, hlsvp or url_encoded_fmt_stream_map information found in video info')
953
954 # Look for the DASH manifest
955 if (self._downloader.params.get('youtube_include_dash_manifest', False)):
956 try:
957 # The DASH manifest used needs to be the one from the original video_webpage.
958 # The one found in get_video_info seems to be using different signatures.
959 # However, in the case of an age restriction there won't be any embedded dashmpd in the video_webpage.
960 # Luckily, it seems, this case uses some kind of default signature (len == 86), so the
961 # combination of get_video_info and the _static_decrypt_signature() decryption fallback will work here.
962 if age_gate:
963 dash_manifest_url = video_info.get('dashmpd')[0]
964 else:
965 dash_manifest_url = ytplayer_config['args']['dashmpd']
966 def decrypt_sig(mobj):
967 s = mobj.group(1)
968 dec_s = self._decrypt_signature(s, video_id, player_url, age_gate)
969 return '/signature/%s' % dec_s
970 dash_manifest_url = re.sub(r'/s/([\w\.]+)', decrypt_sig, dash_manifest_url)
971 dash_doc = self._download_xml(
972 dash_manifest_url, video_id,
973 note=u'Downloading DASH manifest',
974 errnote=u'Could not download DASH manifest')
975 for r in dash_doc.findall(u'.//{urn:mpeg:DASH:schema:MPD:2011}Representation'):
976 url_el = r.find('{urn:mpeg:DASH:schema:MPD:2011}BaseURL')
977 if url_el is None:
978 continue
979 format_id = r.attrib['id']
980 video_url = url_el.text
981 filesize = int_or_none(url_el.attrib.get('{http://youtube.com/yt/2012/10/10}contentLength'))
982 f = {
983 'format_id': format_id,
984 'url': video_url,
985 'width': int_or_none(r.attrib.get('width')),
986 'tbr': int_or_none(r.attrib.get('bandwidth'), 1000),
987 'asr': int_or_none(r.attrib.get('audioSamplingRate')),
988 'filesize': filesize,
989 }
990 try:
991 existing_format = next(
992 fo for fo in formats
993 if fo['format_id'] == format_id)
994 except StopIteration:
995 f.update(self._formats.get(format_id, {}))
996 formats.append(f)
997 else:
998 existing_format.update(f)
999
1000 except (ExtractorError, KeyError) as e:
1001 self.report_warning(u'Skipping DASH manifest: %s' % e, video_id)
1002
1003 self._sort_formats(formats)
1004
1005 return {
1006 'id': video_id,
1007 'uploader': video_uploader,
1008 'uploader_id': video_uploader_id,
1009 'upload_date': upload_date,
1010 'title': video_title,
1011 'thumbnail': video_thumbnail,
1012 'description': video_description,
1013 'categories': video_categories,
1014 'subtitles': video_subtitles,
1015 'duration': video_duration,
1016 'age_limit': 18 if age_gate else 0,
1017 'annotations': video_annotations,
1018 'webpage_url': proto + '://www.youtube.com/watch?v=%s' % video_id,
1019 'view_count': view_count,
1020 'like_count': like_count,
1021 'dislike_count': dislike_count,
1022 'formats': formats,
1023 }
1024
1025 class YoutubePlaylistIE(YoutubeBaseInfoExtractor):
1026 IE_DESC = u'YouTube.com playlists'
1027 _VALID_URL = r"""(?x)(?:
1028 (?:https?://)?
1029 (?:\w+\.)?
1030 youtube\.com/
1031 (?:
1032 (?:course|view_play_list|my_playlists|artist|playlist|watch)
1033 \? (?:.*?&)*? (?:p|a|list)=
1034 | p/
1035 )
1036 (
1037 (?:PL|LL|EC|UU|FL|RD)?[0-9A-Za-z-_]{10,}
1038 # Top tracks, they can also include dots
1039 |(?:MC)[\w\.]*
1040 )
1041 .*
1042 |
1043 ((?:PL|LL|EC|UU|FL|RD)[0-9A-Za-z-_]{10,})
1044 )"""
1045 _TEMPLATE_URL = 'https://www.youtube.com/playlist?list=%s'
1046 _MORE_PAGES_INDICATOR = r'data-link-type="next"'
1047 _VIDEO_RE = r'href="\s*/watch\?v=(?P<id>[0-9A-Za-z_-]{11})&amp;[^"]*?index=(?P<index>\d+)'
1048 IE_NAME = u'youtube:playlist'
1049
1050 def _real_initialize(self):
1051 self._login()
1052
1053 def _ids_to_results(self, ids):
1054 return [self.url_result(vid_id, 'Youtube', video_id=vid_id)
1055 for vid_id in ids]
1056
1057 def _extract_mix(self, playlist_id):
1058 # The mixes are generated from a a single video
1059 # the id of the playlist is just 'RD' + video_id
1060 url = 'https://youtube.com/watch?v=%s&list=%s' % (playlist_id[-11:], playlist_id)
1061 webpage = self._download_webpage(url, playlist_id, u'Downloading Youtube mix')
1062 search_title = lambda class_name: get_element_by_attribute('class', class_name, webpage)
1063 title_span = (search_title('playlist-title') or
1064 search_title('title long-title') or search_title('title'))
1065 title = clean_html(title_span)
1066 video_re = r'''(?x)data-video-username=".*?".*?
1067 href="/watch\?v=([0-9A-Za-z_-]{11})&amp;[^"]*?list=%s''' % re.escape(playlist_id)
1068 ids = orderedSet(re.findall(video_re, webpage, flags=re.DOTALL))
1069 url_results = self._ids_to_results(ids)
1070
1071 return self.playlist_result(url_results, playlist_id, title)
1072
1073 def _real_extract(self, url):
1074 # Extract playlist id
1075 mobj = re.match(self._VALID_URL, url)
1076 if mobj is None:
1077 raise ExtractorError(u'Invalid URL: %s' % url)
1078 playlist_id = mobj.group(1) or mobj.group(2)
1079
1080 # Check if it's a video-specific URL
1081 query_dict = compat_urlparse.parse_qs(compat_urlparse.urlparse(url).query)
1082 if 'v' in query_dict:
1083 video_id = query_dict['v'][0]
1084 if self._downloader.params.get('noplaylist'):
1085 self.to_screen(u'Downloading just video %s because of --no-playlist' % video_id)
1086 return self.url_result(video_id, 'Youtube', video_id=video_id)
1087 else:
1088 self.to_screen(u'Downloading playlist %s - add --no-playlist to just download video %s' % (playlist_id, video_id))
1089
1090 if playlist_id.startswith('RD'):
1091 # Mixes require a custom extraction process
1092 return self._extract_mix(playlist_id)
1093 if playlist_id.startswith('TL'):
1094 raise ExtractorError(u'For downloading YouTube.com top lists, use '
1095 u'the "yttoplist" keyword, for example "youtube-dl \'yttoplist:music:Top Tracks\'"', expected=True)
1096
1097 url = self._TEMPLATE_URL % playlist_id
1098 page = self._download_webpage(url, playlist_id)
1099 more_widget_html = content_html = page
1100
1101 # Check if the playlist exists or is private
1102 if re.search(r'<div class="yt-alert-message">[^<]*?(The|This) playlist (does not exist|is private)[^<]*?</div>', page) is not None:
1103 raise ExtractorError(
1104 u'The playlist doesn\'t exist or is private, use --username or '
1105 '--netrc to access it.',
1106 expected=True)
1107
1108 # Extract the video ids from the playlist pages
1109 ids = []
1110
1111 for page_num in itertools.count(1):
1112 matches = re.finditer(self._VIDEO_RE, content_html)
1113 # We remove the duplicates and the link with index 0
1114 # (it's not the first video of the playlist)
1115 new_ids = orderedSet(m.group('id') for m in matches if m.group('index') != '0')
1116 ids.extend(new_ids)
1117
1118 mobj = re.search(r'data-uix-load-more-href="/?(?P<more>[^"]+)"', more_widget_html)
1119 if not mobj:
1120 break
1121
1122 more = self._download_json(
1123 'https://youtube.com/%s' % mobj.group('more'), playlist_id,
1124 'Downloading page #%s' % page_num,
1125 transform_source=uppercase_escape)
1126 content_html = more['content_html']
1127 more_widget_html = more['load_more_widget_html']
1128
1129 playlist_title = self._html_search_regex(
1130 r'(?s)<h1 class="pl-header-title[^"]*">\s*(.*?)\s*</h1>',
1131 page, u'title')
1132
1133 url_results = self._ids_to_results(ids)
1134 return self.playlist_result(url_results, playlist_id, playlist_title)
1135
1136
1137 class YoutubeTopListIE(YoutubePlaylistIE):
1138 IE_NAME = u'youtube:toplist'
1139 IE_DESC = (u'YouTube.com top lists, "yttoplist:{channel}:{list title}"'
1140 u' (Example: "yttoplist:music:Top Tracks")')
1141 _VALID_URL = r'yttoplist:(?P<chann>.*?):(?P<title>.*?)$'
1142
1143 def _real_extract(self, url):
1144 mobj = re.match(self._VALID_URL, url)
1145 channel = mobj.group('chann')
1146 title = mobj.group('title')
1147 query = compat_urllib_parse.urlencode({'title': title})
1148 playlist_re = 'href="([^"]+?%s.*?)"' % re.escape(query)
1149 channel_page = self._download_webpage('https://www.youtube.com/%s' % channel, title)
1150 link = self._html_search_regex(playlist_re, channel_page, u'list')
1151 url = compat_urlparse.urljoin('https://www.youtube.com/', link)
1152
1153 video_re = r'data-index="\d+".*?data-video-id="([0-9A-Za-z_-]{11})"'
1154 ids = []
1155 # sometimes the webpage doesn't contain the videos
1156 # retry until we get them
1157 for i in itertools.count(0):
1158 msg = u'Downloading Youtube mix'
1159 if i > 0:
1160 msg += ', retry #%d' % i
1161 webpage = self._download_webpage(url, title, msg)
1162 ids = orderedSet(re.findall(video_re, webpage))
1163 if ids:
1164 break
1165 url_results = self._ids_to_results(ids)
1166 return self.playlist_result(url_results, playlist_title=title)
1167
1168
1169 class YoutubeChannelIE(InfoExtractor):
1170 IE_DESC = u'YouTube.com channels'
1171 _VALID_URL = r"^(?:https?://)?(?:youtu\.be|(?:\w+\.)?youtube(?:-nocookie)?\.com)/channel/([0-9A-Za-z_-]+)"
1172 _MORE_PAGES_INDICATOR = 'yt-uix-load-more'
1173 _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'
1174 IE_NAME = u'youtube:channel'
1175
1176 def extract_videos_from_page(self, page):
1177 ids_in_page = []
1178 for mobj in re.finditer(r'href="/watch\?v=([0-9A-Za-z_-]+)&?', page):
1179 if mobj.group(1) not in ids_in_page:
1180 ids_in_page.append(mobj.group(1))
1181 return ids_in_page
1182
1183 def _real_extract(self, url):
1184 # Extract channel id
1185 mobj = re.match(self._VALID_URL, url)
1186 if mobj is None:
1187 raise ExtractorError(u'Invalid URL: %s' % url)
1188
1189 # Download channel page
1190 channel_id = mobj.group(1)
1191 video_ids = []
1192 url = 'https://www.youtube.com/channel/%s/videos' % channel_id
1193 channel_page = self._download_webpage(url, channel_id)
1194 autogenerated = re.search(r'''(?x)
1195 class="[^"]*?(?:
1196 channel-header-autogenerated-label|
1197 yt-channel-title-autogenerated
1198 )[^"]*"''', channel_page) is not None
1199
1200 if autogenerated:
1201 # The videos are contained in a single page
1202 # the ajax pages can't be used, they are empty
1203 video_ids = self.extract_videos_from_page(channel_page)
1204 else:
1205 # Download all channel pages using the json-based channel_ajax query
1206 for pagenum in itertools.count(1):
1207 url = self._MORE_PAGES_URL % (pagenum, channel_id)
1208 page = self._download_json(
1209 url, channel_id, note=u'Downloading page #%s' % pagenum,
1210 transform_source=uppercase_escape)
1211
1212 ids_in_page = self.extract_videos_from_page(page['content_html'])
1213 video_ids.extend(ids_in_page)
1214
1215 if self._MORE_PAGES_INDICATOR not in page['load_more_widget_html']:
1216 break
1217
1218 self._downloader.to_screen(u'[youtube] Channel %s: Found %i videos' % (channel_id, len(video_ids)))
1219
1220 url_entries = [self.url_result(video_id, 'Youtube', video_id=video_id)
1221 for video_id in video_ids]
1222 return self.playlist_result(url_entries, channel_id)
1223
1224
1225 class YoutubeUserIE(InfoExtractor):
1226 IE_DESC = u'YouTube.com user videos (URL or "ytuser" keyword)'
1227 _VALID_URL = r'(?:(?:(?:https?://)?(?:\w+\.)?youtube\.com/(?:user/)?(?!(?:attribution_link|watch|results)(?:$|[^a-z_A-Z0-9-])))|ytuser:)(?!feed/)([A-Za-z0-9_-]+)'
1228 _TEMPLATE_URL = 'https://gdata.youtube.com/feeds/api/users/%s'
1229 _GDATA_PAGE_SIZE = 50
1230 _GDATA_URL = 'https://gdata.youtube.com/feeds/api/users/%s/uploads?max-results=%d&start-index=%d&alt=json'
1231 IE_NAME = u'youtube:user'
1232
1233 @classmethod
1234 def suitable(cls, url):
1235 # Don't return True if the url can be extracted with other youtube
1236 # extractor, the regex would is too permissive and it would match.
1237 other_ies = iter(klass for (name, klass) in globals().items() if name.endswith('IE') and klass is not cls)
1238 if any(ie.suitable(url) for ie in other_ies): return False
1239 else: return super(YoutubeUserIE, cls).suitable(url)
1240
1241 def _real_extract(self, url):
1242 # Extract username
1243 mobj = re.match(self._VALID_URL, url)
1244 if mobj is None:
1245 raise ExtractorError(u'Invalid URL: %s' % url)
1246
1247 username = mobj.group(1)
1248
1249 # Download video ids using YouTube Data API. Result size per
1250 # query is limited (currently to 50 videos) so we need to query
1251 # page by page until there are no video ids - it means we got
1252 # all of them.
1253
1254 def download_page(pagenum):
1255 start_index = pagenum * self._GDATA_PAGE_SIZE + 1
1256
1257 gdata_url = self._GDATA_URL % (username, self._GDATA_PAGE_SIZE, start_index)
1258 page = self._download_webpage(
1259 gdata_url, username,
1260 u'Downloading video ids from %d to %d' % (
1261 start_index, start_index + self._GDATA_PAGE_SIZE))
1262
1263 try:
1264 response = json.loads(page)
1265 except ValueError as err:
1266 raise ExtractorError(u'Invalid JSON in API response: ' + compat_str(err))
1267 if 'entry' not in response['feed']:
1268 return
1269
1270 # Extract video identifiers
1271 entries = response['feed']['entry']
1272 for entry in entries:
1273 title = entry['title']['$t']
1274 video_id = entry['id']['$t'].split('/')[-1]
1275 yield {
1276 '_type': 'url',
1277 'url': video_id,
1278 'ie_key': 'Youtube',
1279 'id': video_id,
1280 'title': title,
1281 }
1282 url_results = PagedList(download_page, self._GDATA_PAGE_SIZE)
1283
1284 return self.playlist_result(url_results, playlist_title=username)
1285
1286
1287 class YoutubeSearchIE(SearchInfoExtractor):
1288 IE_DESC = u'YouTube.com searches'
1289 _API_URL = u'https://gdata.youtube.com/feeds/api/videos?q=%s&start-index=%i&max-results=50&v=2&alt=jsonc'
1290 _MAX_RESULTS = 1000
1291 IE_NAME = u'youtube:search'
1292 _SEARCH_KEY = 'ytsearch'
1293
1294 def _get_n_results(self, query, n):
1295 """Get a specified number of results for a query"""
1296
1297 video_ids = []
1298 pagenum = 0
1299 limit = n
1300 PAGE_SIZE = 50
1301
1302 while (PAGE_SIZE * pagenum) < limit:
1303 result_url = self._API_URL % (
1304 compat_urllib_parse.quote_plus(query.encode('utf-8')),
1305 (PAGE_SIZE * pagenum) + 1)
1306 data_json = self._download_webpage(
1307 result_url, video_id=u'query "%s"' % query,
1308 note=u'Downloading page %s' % (pagenum + 1),
1309 errnote=u'Unable to download API page')
1310 data = json.loads(data_json)
1311 api_response = data['data']
1312
1313 if 'items' not in api_response:
1314 raise ExtractorError(
1315 u'[youtube] No video results', expected=True)
1316
1317 new_ids = list(video['id'] for video in api_response['items'])
1318 video_ids += new_ids
1319
1320 limit = min(n, api_response['totalItems'])
1321 pagenum += 1
1322
1323 if len(video_ids) > n:
1324 video_ids = video_ids[:n]
1325 videos = [self.url_result(video_id, 'Youtube', video_id=video_id)
1326 for video_id in video_ids]
1327 return self.playlist_result(videos, query)
1328
1329
1330 class YoutubeSearchDateIE(YoutubeSearchIE):
1331 IE_NAME = YoutubeSearchIE.IE_NAME + ':date'
1332 _API_URL = 'https://gdata.youtube.com/feeds/api/videos?q=%s&start-index=%i&max-results=50&v=2&alt=jsonc&orderby=published'
1333 _SEARCH_KEY = 'ytsearchdate'
1334 IE_DESC = u'YouTube.com searches, newest videos first'
1335
1336
1337 class YoutubeSearchURLIE(InfoExtractor):
1338 IE_DESC = u'YouTube.com search URLs'
1339 IE_NAME = u'youtube:search_url'
1340 _VALID_URL = r'https?://(?:www\.)?youtube\.com/results\?(.*?&)?search_query=(?P<query>[^&]+)(?:[&]|$)'
1341
1342 def _real_extract(self, url):
1343 mobj = re.match(self._VALID_URL, url)
1344 query = compat_urllib_parse.unquote_plus(mobj.group('query'))
1345
1346 webpage = self._download_webpage(url, query)
1347 result_code = self._search_regex(
1348 r'(?s)<ol class="item-section"(.*?)</ol>', webpage, u'result HTML')
1349
1350 part_codes = re.findall(
1351 r'(?s)<h3 class="yt-lockup-title">(.*?)</h3>', result_code)
1352 entries = []
1353 for part_code in part_codes:
1354 part_title = self._html_search_regex(
1355 [r'(?s)title="([^"]+)"', r'>([^<]+)</a>'], part_code, 'item title', fatal=False)
1356 part_url_snippet = self._html_search_regex(
1357 r'(?s)href="([^"]+)"', part_code, 'item URL')
1358 part_url = compat_urlparse.urljoin(
1359 'https://www.youtube.com/', part_url_snippet)
1360 entries.append({
1361 '_type': 'url',
1362 'url': part_url,
1363 'title': part_title,
1364 })
1365
1366 return {
1367 '_type': 'playlist',
1368 'entries': entries,
1369 'title': query,
1370 }
1371
1372
1373 class YoutubeShowIE(InfoExtractor):
1374 IE_DESC = u'YouTube.com (multi-season) shows'
1375 _VALID_URL = r'https?://www\.youtube\.com/show/(.*)'
1376 IE_NAME = u'youtube:show'
1377
1378 def _real_extract(self, url):
1379 mobj = re.match(self._VALID_URL, url)
1380 show_name = mobj.group(1)
1381 webpage = self._download_webpage(url, show_name, u'Downloading show webpage')
1382 # There's one playlist for each season of the show
1383 m_seasons = list(re.finditer(r'href="(/playlist\?list=.*?)"', webpage))
1384 self.to_screen(u'%s: Found %s seasons' % (show_name, len(m_seasons)))
1385 return [self.url_result('https://www.youtube.com' + season.group(1), 'YoutubePlaylist') for season in m_seasons]
1386
1387
1388 class YoutubeFeedsInfoExtractor(YoutubeBaseInfoExtractor):
1389 """
1390 Base class for extractors that fetch info from
1391 http://www.youtube.com/feed_ajax
1392 Subclasses must define the _FEED_NAME and _PLAYLIST_TITLE properties.
1393 """
1394 _LOGIN_REQUIRED = True
1395 # use action_load_personal_feed instead of action_load_system_feed
1396 _PERSONAL_FEED = False
1397
1398 @property
1399 def _FEED_TEMPLATE(self):
1400 action = 'action_load_system_feed'
1401 if self._PERSONAL_FEED:
1402 action = 'action_load_personal_feed'
1403 return 'https://www.youtube.com/feed_ajax?%s=1&feed_name=%s&paging=%%s' % (action, self._FEED_NAME)
1404
1405 @property
1406 def IE_NAME(self):
1407 return u'youtube:%s' % self._FEED_NAME
1408
1409 def _real_initialize(self):
1410 self._login()
1411
1412 def _real_extract(self, url):
1413 feed_entries = []
1414 paging = 0
1415 for i in itertools.count(1):
1416 info = self._download_json(self._FEED_TEMPLATE % paging,
1417 u'%s feed' % self._FEED_NAME,
1418 u'Downloading page %s' % i)
1419 feed_html = info.get('feed_html') or info.get('content_html')
1420 m_ids = re.finditer(r'"/watch\?v=(.*?)["&]', feed_html)
1421 ids = orderedSet(m.group(1) for m in m_ids)
1422 feed_entries.extend(
1423 self.url_result(video_id, 'Youtube', video_id=video_id)
1424 for video_id in ids)
1425 mobj = re.search(
1426 r'data-uix-load-more-href="/?[^"]+paging=(?P<paging>\d+)',
1427 feed_html)
1428 if mobj is None:
1429 break
1430 paging = mobj.group('paging')
1431 return self.playlist_result(feed_entries, playlist_title=self._PLAYLIST_TITLE)
1432
1433 class YoutubeSubscriptionsIE(YoutubeFeedsInfoExtractor):
1434 IE_DESC = u'YouTube.com subscriptions feed, "ytsubs" keyword (requires authentication)'
1435 _VALID_URL = r'https?://www\.youtube\.com/feed/subscriptions|:ytsubs(?:criptions)?'
1436 _FEED_NAME = 'subscriptions'
1437 _PLAYLIST_TITLE = u'Youtube Subscriptions'
1438
1439 class YoutubeRecommendedIE(YoutubeFeedsInfoExtractor):
1440 IE_DESC = u'YouTube.com recommended videos, "ytrec" keyword (requires authentication)'
1441 _VALID_URL = r'https?://www\.youtube\.com/feed/recommended|:ytrec(?:ommended)?'
1442 _FEED_NAME = 'recommended'
1443 _PLAYLIST_TITLE = u'Youtube Recommended videos'
1444
1445 class YoutubeWatchLaterIE(YoutubeFeedsInfoExtractor):
1446 IE_DESC = u'Youtube watch later list, "ytwatchlater" keyword (requires authentication)'
1447 _VALID_URL = r'https?://www\.youtube\.com/feed/watch_later|:ytwatchlater'
1448 _FEED_NAME = 'watch_later'
1449 _PLAYLIST_TITLE = u'Youtube Watch Later'
1450 _PERSONAL_FEED = True
1451
1452 class YoutubeHistoryIE(YoutubeFeedsInfoExtractor):
1453 IE_DESC = u'Youtube watch history, "ythistory" keyword (requires authentication)'
1454 _VALID_URL = u'https?://www\.youtube\.com/feed/history|:ythistory'
1455 _FEED_NAME = 'history'
1456 _PERSONAL_FEED = True
1457 _PLAYLIST_TITLE = u'Youtube Watch History'
1458
1459 class YoutubeFavouritesIE(YoutubeBaseInfoExtractor):
1460 IE_NAME = u'youtube:favorites'
1461 IE_DESC = u'YouTube.com favourite videos, "ytfav" keyword (requires authentication)'
1462 _VALID_URL = r'https?://www\.youtube\.com/my_favorites|:ytfav(?:ou?rites)?'
1463 _LOGIN_REQUIRED = True
1464
1465 def _real_extract(self, url):
1466 webpage = self._download_webpage('https://www.youtube.com/my_favorites', 'Youtube Favourites videos')
1467 playlist_id = self._search_regex(r'list=(.+?)["&]', webpage, u'favourites playlist id')
1468 return self.url_result(playlist_id, 'YoutubePlaylist')
1469
1470
1471 class YoutubeTruncatedURLIE(InfoExtractor):
1472 IE_NAME = 'youtube:truncated_url'
1473 IE_DESC = False # Do not list
1474 _VALID_URL = r'''(?x)
1475 (?:https?://)?[^/]+/watch\?(?:
1476 feature=[a-z_]+|
1477 annotation_id=annotation_[^&]+
1478 )?$|
1479 (?:https?://)?(?:www\.)?youtube\.com/attribution_link\?a=[^&]+$
1480 '''
1481
1482 _TESTS = [{
1483 'url': 'http://www.youtube.com/watch?annotation_id=annotation_3951667041',
1484 'only_matching': True,
1485 }, {
1486 'url': 'http://www.youtube.com/watch?',
1487 'only_matching': True,
1488 }]
1489
1490 def _real_extract(self, url):
1491 raise ExtractorError(
1492 u'Did you forget to quote the URL? Remember that & is a meta '
1493 u'character in most shells, so you want to put the URL in quotes, '
1494 u'like youtube-dl '
1495 u'"http://www.youtube.com/watch?feature=foo&v=BaW_jenozKc" '
1496 u' or simply youtube-dl BaW_jenozKc .',
1497 expected=True)
fork of https://github.com/yt-dlp/yt-dlp/