3 from __future__ import unicode_literals
12 from .common import InfoExtractor, SearchInfoExtractor
13 from .subtitles import SubtitlesInfoExtractor
14 from ..jsinterp import JSInterpreter
15 from ..swfinterp import SWFInterpreter
20 compat_urllib_request,
26 get_element_by_attribute,
36 class YoutubeBaseInfoExtractor(InfoExtractor):
37 """Provide base functions for Youtube extractors"""
38 _LOGIN_URL = 'https://accounts.google.com/ServiceLogin'
39 _TWOFACTOR_URL = 'https://accounts.google.com/SecondFactor'
40 _LANG_URL = r'https://www.youtube.com/?hl=en&persist_hl=1&gl=US&persist_gl=1&opt_out_ackd=1'
41 _AGE_URL = 'https://www.youtube.com/verify_age?next_url=/&gl=US&hl=en'
42 _NETRC_MACHINE = 'youtube'
43 # If True it will raise an error if no login info is provided
44 _LOGIN_REQUIRED = False
46 def _set_language(self):
47 return bool(self._download_webpage(
49 note=u'Setting language', errnote='unable to set language',
54 Attempt to log in to YouTube.
55 True is returned if successful or skipped.
56 False is returned if login failed.
58 If _LOGIN_REQUIRED is set and no authentication was provided, an error is raised.
60 (username, password) = self._get_login_info()
61 # No authentication to be performed
63 if self._LOGIN_REQUIRED:
64 raise ExtractorError(u'No login info available, needed for using %s.' % self.IE_NAME, expected=True)
67 login_page = self._download_webpage(
68 self._LOGIN_URL, None,
69 note=u'Downloading login page',
70 errnote=u'unable to fetch login page', fatal=False)
71 if login_page is False:
74 galx = self._search_regex(r'(?s)<input.+?name="GALX".+?value="(.+?)"',
75 login_page, 'Login GALX parameter')
79 'continue': 'https://www.youtube.com/signin?action_handle_signin=true&feature=sign_in_button&hl=en_US&nomobiletemp=1',
84 'PersistentCookie': 'yes',
86 'bgresponse': 'js_disabled',
87 'checkConnection': '',
88 'checkedDomains': 'youtube',
100 # Convert to UTF-8 *before* urlencode because Python 2.x's urlencode
102 login_form = dict((k.encode('utf-8'), v.encode('utf-8')) for k,v in login_form_strs.items())
103 login_data = compat_urllib_parse.urlencode(login_form).encode('ascii')
105 req = compat_urllib_request.Request(self._LOGIN_URL, login_data)
106 login_results = self._download_webpage(
108 note=u'Logging in', errnote=u'unable to log in', fatal=False)
109 if login_results is False:
112 if re.search(r'id="errormsg_0_Passwd"', login_results) is not None:
113 raise ExtractorError(u'Please use your account password and a two-factor code instead of an application-specific password.', expected=True)
116 # TODO add SMS and phone call support - these require making a request and then prompting the user
118 if re.search(r'(?i)<form[^>]* id="gaia_secondfactorform"', login_results) is not None:
119 tfa_code = self._get_tfa_info()
122 self._downloader.report_warning(u'Two-factor authentication required. Provide it with --twofactor <code>')
123 self._downloader.report_warning(u'(Note that only TOTP (Google Authenticator App) codes work at this time.)')
126 # Unlike the first login form, secTok and timeStmp are both required for the TFA form
128 match = re.search(r'id="secTok"\n\s+value=\'(.+)\'/>', login_results, re.M | re.U)
130 self._downloader.report_warning(u'Failed to get secTok - did the page structure change?')
131 secTok = match.group(1)
132 match = re.search(r'id="timeStmp"\n\s+value=\'(.+)\'/>', login_results, re.M | re.U)
134 self._downloader.report_warning(u'Failed to get timeStmp - did the page structure change?')
135 timeStmp = match.group(1)
138 'continue': 'https://www.youtube.com/signin?action_handle_signin=true&feature=sign_in_button&hl=en_US&nomobiletemp=1',
140 'smsUserPin': tfa_code,
141 'smsVerifyPin': 'Verify',
143 'PersistentCookie': 'yes',
144 'checkConnection': '',
145 'checkedDomains': 'youtube',
148 'timeStmp': timeStmp,
149 'service': 'youtube',
152 tfa_form = dict((k.encode('utf-8'), v.encode('utf-8')) for k,v in tfa_form_strs.items())
153 tfa_data = compat_urllib_parse.urlencode(tfa_form).encode('ascii')
155 tfa_req = compat_urllib_request.Request(self._TWOFACTOR_URL, tfa_data)
156 tfa_results = self._download_webpage(
158 note=u'Submitting TFA code', errnote=u'unable to submit tfa', fatal=False)
160 if tfa_results is False:
163 if re.search(r'(?i)<form[^>]* id="gaia_secondfactorform"', tfa_results) is not None:
164 self._downloader.report_warning(u'Two-factor code expired. Please try again, or use a one-use backup code instead.')
166 if re.search(r'(?i)<form[^>]* id="gaia_loginform"', tfa_results) is not None:
167 self._downloader.report_warning(u'unable to log in - did the page structure change?')
169 if re.search(r'smsauth-interstitial-reviewsettings', tfa_results) is not None:
170 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.')
173 if re.search(r'(?i)<form[^>]* id="gaia_loginform"', login_results) is not None:
174 self._downloader.report_warning(u'unable to log in: bad username or password')
178 def _confirm_age(self):
181 'action_confirm': 'Confirm',
183 req = compat_urllib_request.Request(self._AGE_URL,
184 compat_urllib_parse.urlencode(age_form).encode('ascii'))
186 self._download_webpage(
188 note=u'Confirming age', errnote=u'Unable to confirm age')
191 def _real_initialize(self):
192 if self._downloader is None:
194 if not self._set_language():
196 if not self._login():
201 class YoutubeIE(YoutubeBaseInfoExtractor, SubtitlesInfoExtractor):
202 IE_DESC = 'YouTube.com'
203 _VALID_URL = r"""(?x)^
205 (?:https?://|//) # http(s):// or protocol-independent URL
206 (?:(?:(?:(?:\w+\.)?[yY][oO][uU][tT][uU][bB][eE](?:-nocookie)?\.com/|
207 (?:www\.)?deturl\.com/www\.youtube\.com/|
208 (?:www\.)?pwnyoutube\.com/|
209 (?:www\.)?yourepeat\.com/|
210 tube\.majestyc\.net/|
211 youtube\.googleapis\.com/) # the various hostnames, with wildcard subdomains
212 (?:.*?\#/)? # handle anchor (#/) redirect urls
213 (?: # the various things that can precede the ID:
214 (?:(?:v|embed|e)/) # v/ or embed/ or e/
215 |(?: # or the v= param in all its forms
216 (?:(?:watch|movie)(?:_popup)?(?:\.php)?/?)? # preceding watch(_popup|.php) or nothing (like /?v=xxxx)
217 (?:\?|\#!?) # the params delimiter ? or # or #!
218 (?:.*?&)? # any other preceding param (like /?s=tuff&v=xxxx)
222 |youtu\.be/ # just youtu.be/xxxx
223 |(?:www\.)?cleanvideosearch\.com/media/action/yt/watch\?videoId=
225 )? # all until now is optional -> you can pass the naked ID
226 ([0-9A-Za-z_-]{11}) # here is it! the YouTube video ID
227 (?!.*?&list=) # combined list/video URLs are handled by the playlist IE
228 (?(1).+)? # if we found the ID, everything can follow
230 _NEXT_URL_RE = r'[\?&]next_url=([^&]+)'
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},
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},
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},
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},
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},
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},
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},
304 '_rtmp': {'protocol': 'rtmp'},
310 u"url": u"http://www.youtube.com/watch?v=BaW_jenozKc",
311 u"file": u"BaW_jenozKc.mp4",
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'],
320 'dislike_count': int,
324 u"url": u"http://www.youtube.com/watch?v=UxxajLWwzqY",
325 u"file": u"UxxajLWwzqY.mp4",
326 u"note": u"Test generic use_cipher_signature video (#897)",
328 u"upload_date": u"20120506",
329 u"title": u"Icona Pop - I Love It (feat. Charli XCX) [OFFICIAL VIDEO]",
330 u"description": u"md5:fea86fda2d5a5784273df5c7cc994d9f",
331 u"uploader": u"Icona Pop",
332 u"uploader_id": u"IconaPop"
336 u"url": u"https://www.youtube.com/watch?v=07FYdnEawAQ",
337 u"file": u"07FYdnEawAQ.mp4",
338 u"note": u"Test VEVO video with age protection (#956)",
340 u"upload_date": u"20130703",
341 u"title": u"Justin Timberlake - Tunnel Vision (Explicit)",
342 u"description": u"md5:64249768eec3bc4276236606ea996373",
343 u"uploader": u"justintimberlakeVEVO",
344 u"uploader_id": u"justintimberlakeVEVO"
348 u"url": u"//www.YouTube.com/watch?v=yZIXLfi8CZQ",
349 u"file": u"yZIXLfi8CZQ.mp4",
350 u"note": u"Embed-only video (#1746)",
352 u"upload_date": u"20120608",
353 u"title": u"Principal Sexually Assaults A Teacher - Episode 117 - 8th June 2012",
354 u"description": u"md5:09b78bd971f1e3e289601dfba15ca4f7",
355 u"uploader": u"SET India",
356 u"uploader_id": u"setindia"
360 u"url": u"http://www.youtube.com/watch?v=a9LDPn-MO4I",
361 u"file": u"a9LDPn-MO4I.m4a",
362 u"note": u"256k DASH audio (format 141) via DASH manifest",
364 u"upload_date": "20121002",
365 u"uploader_id": "8KVIDEO",
367 u"uploader": "8KVIDEO",
368 u"title": "UHDTV TEST 8K VIDEO.mp4"
371 u"youtube_include_dash_manifest": True,
375 # DASH manifest with encrypted signature
377 'url': 'https://www.youtube.com/watch?v=IB3lcPjvWLA',
381 'title': 'Afrojack - The Spark ft. Spree Wilson',
382 'description': 'md5:9717375db5a9a3992be4668bbf3bc0a8',
383 'uploader': 'AfrojackVEVO',
384 'uploader_id': 'AfrojackVEVO',
385 'upload_date': '20131011',
388 'youtube_include_dash_manifest': True,
394 def __init__(self, *args, **kwargs):
395 super(YoutubeIE, self).__init__(*args, **kwargs)
396 self._player_cache = {}
398 def report_video_info_webpage_download(self, video_id):
399 """Report attempt to download video info webpage."""
400 self.to_screen(u'%s: Downloading video info webpage' % video_id)
402 def report_information_extraction(self, video_id):
403 """Report attempt to extract video information."""
404 self.to_screen(u'%s: Extracting video information' % video_id)
406 def report_unavailable_format(self, video_id, format):
407 """Report extracted video URL."""
408 self.to_screen(u'%s: Format %s not available' % (video_id, format))
410 def report_rtmp_download(self):
411 """Indicate the download will use the RTMP protocol."""
412 self.to_screen(u'RTMP download detected')
414 def _signature_cache_id(self, example_sig):
415 """ Return a string representation of a signature """
416 return '.'.join(compat_str(len(part)) for part in example_sig.split('.'))
418 def _extract_signature_function(self, video_id, player_url, example_sig):
420 r'.*-(?P<id>[a-zA-Z0-9_-]+)(?:/watch_as3|/html5player)?\.(?P<ext>[a-z]+)$',
423 raise ExtractorError('Cannot identify player %r' % player_url)
424 player_type = id_m.group('ext')
425 player_id = id_m.group('id')
427 # Read from filesystem cache
428 func_id = '%s_%s_%s' % (
429 player_type, player_id, self._signature_cache_id(example_sig))
430 assert os.path.basename(func_id) == func_id
432 cache_spec = self._downloader.cache.load(u'youtube-sigfuncs', func_id)
433 if cache_spec is not None:
434 return lambda s: ''.join(s[i] for i in cache_spec)
436 if player_type == 'js':
437 code = self._download_webpage(
438 player_url, video_id,
439 note=u'Downloading %s player %s' % (player_type, player_id),
440 errnote=u'Download of %s failed' % player_url)
441 res = self._parse_sig_js(code)
442 elif player_type == 'swf':
443 urlh = self._request_webpage(
444 player_url, video_id,
445 note=u'Downloading %s player %s' % (player_type, player_id),
446 errnote=u'Download of %s failed' % player_url)
448 res = self._parse_sig_swf(code)
450 assert False, 'Invalid player type %r' % player_type
452 if cache_spec is None:
453 test_string = ''.join(map(compat_chr, range(len(example_sig))))
454 cache_res = res(test_string)
455 cache_spec = [ord(c) for c in cache_res]
457 self._downloader.cache.store(u'youtube-sigfuncs', func_id, cache_spec)
460 def _print_sig_code(self, func, example_sig):
461 def gen_sig_code(idxs):
462 def _genslice(start, end, step):
463 starts = '' if start == 0 else str(start)
464 ends = (u':%d' % (end+step)) if end + step >= 0 else ':'
465 steps = '' if step == 1 else (u':%d' % step)
466 return 's[%s%s%s]' % (starts, ends, steps)
469 start = '(Never used)' # Quelch pyflakes warnings - start will be
470 # set as soon as step is set
471 for i, prev in zip(idxs[1:], idxs[:-1]):
475 yield _genslice(start, prev, step)
478 if i - prev in [-1, 1]:
487 yield _genslice(start, i, step)
489 test_string = ''.join(map(compat_chr, range(len(example_sig))))
490 cache_res = func(test_string)
491 cache_spec = [ord(c) for c in cache_res]
492 expr_code = ' + '.join(gen_sig_code(cache_spec))
493 signature_id_tuple = '(%s)' % (
494 ', '.join(compat_str(len(p)) for p in example_sig.split('.')))
495 code = (u'if tuple(len(p) for p in s.split(\'.\')) == %s:\n'
496 ' return %s\n') % (signature_id_tuple, expr_code)
497 self.to_screen(u'Extracted signature function:\n' + code)
499 def _parse_sig_js(self, jscode):
500 funcname = self._search_regex(
501 r'signature=([$a-zA-Z]+)', jscode,
502 'Initial JS player signature function name')
504 jsi = JSInterpreter(jscode)
505 initial_function = jsi.extract_function(funcname)
506 return lambda s: initial_function([s])
508 def _parse_sig_swf(self, file_contents):
509 swfi = SWFInterpreter(file_contents)
510 TARGET_CLASSNAME = 'SignatureDecipher'
511 searched_class = swfi.extract_class(TARGET_CLASSNAME)
512 initial_function = swfi.extract_function(searched_class, 'decipher')
513 return lambda s: initial_function([s])
515 def _decrypt_signature(self, s, video_id, player_url, age_gate=False):
516 """Turn the encrypted s field into a working signature"""
518 if player_url is None:
519 raise ExtractorError(u'Cannot decrypt signature without player_url')
521 if player_url.startswith(u'//'):
522 player_url = 'https:' + player_url
524 player_id = (player_url, self._signature_cache_id(s))
525 if player_id not in self._player_cache:
526 func = self._extract_signature_function(
527 video_id, player_url, s
529 self._player_cache[player_id] = func
530 func = self._player_cache[player_id]
531 if self._downloader.params.get('youtube_print_sig_code'):
532 self._print_sig_code(func, s)
534 except Exception as e:
535 tb = traceback.format_exc()
536 raise ExtractorError(
537 'Signature extraction failed: ' + tb, cause=e)
539 def _get_available_subtitles(self, video_id, webpage):
541 sub_list = self._download_webpage(
542 'https://video.google.com/timedtext?hl=en&type=list&v=%s' % video_id,
543 video_id, note=False)
544 except ExtractorError as err:
545 self._downloader.report_warning(u'unable to download video subtitles: %s' % compat_str(err))
547 lang_list = re.findall(r'name="([^"]*)"[^>]+lang_code="([\w\-]+)"', sub_list)
552 if lang in sub_lang_list:
554 params = compat_urllib_parse.urlencode({
557 'fmt': self._downloader.params.get('subtitlesformat', 'srt'),
558 'name': unescapeHTML(l[0]).encode('utf-8'),
560 url = 'https://www.youtube.com/api/timedtext?' + params
561 sub_lang_list[lang] = url
562 if not sub_lang_list:
563 self._downloader.report_warning(u'video doesn\'t have subtitles')
567 def _get_available_automatic_caption(self, video_id, webpage):
568 """We need the webpage for getting the captions url, pass it as an
569 argument to speed up the process."""
570 sub_format = self._downloader.params.get('subtitlesformat', 'srt')
571 self.to_screen(u'%s: Looking for automatic captions' % video_id)
572 mobj = re.search(r';ytplayer.config = ({.*?});', webpage)
573 err_msg = 'Couldn\'t find automatic captions for %s' % video_id
575 self._downloader.report_warning(err_msg)
577 player_config = json.loads(mobj.group(1))
579 args = player_config[u'args']
580 caption_url = args[u'ttsurl']
581 timestamp = args[u'timestamp']
582 # We get the available subtitles
583 list_params = compat_urllib_parse.urlencode({
588 list_url = caption_url + '&' + list_params
589 caption_list = self._download_xml(list_url, video_id)
590 original_lang_node = caption_list.find('track')
591 if original_lang_node is None or original_lang_node.attrib.get('kind') != 'asr' :
592 self._downloader.report_warning(u'Video doesn\'t have automatic captions')
594 original_lang = original_lang_node.attrib['lang_code']
597 for lang_node in caption_list.findall('target'):
598 sub_lang = lang_node.attrib['lang_code']
599 params = compat_urllib_parse.urlencode({
600 'lang': original_lang,
606 sub_lang_list[sub_lang] = caption_url + '&' + params
608 # An extractor error can be raise by the download process if there are
609 # no automatic captions but there are subtitles
610 except (KeyError, ExtractorError):
611 self._downloader.report_warning(err_msg)
615 def extract_id(cls, url):
616 mobj = re.match(cls._VALID_URL, url, re.VERBOSE)
618 raise ExtractorError(u'Invalid URL: %s' % url)
619 video_id = mobj.group(2)
622 def _extract_from_m3u8(self, manifest_url, video_id):
624 def _get_urls(_manifest):
625 lines = _manifest.split('\n')
626 urls = filter(lambda l: l and not l.startswith('#'),
629 manifest = self._download_webpage(manifest_url, video_id, 'Downloading formats manifest')
630 formats_urls = _get_urls(manifest)
631 for format_url in formats_urls:
632 itag = self._search_regex(r'itag/(\d+?)/', format_url, 'itag')
633 url_map[itag] = format_url
636 def _extract_annotations(self, video_id):
637 url = 'https://www.youtube.com/annotations_invideo?features=1&legacy=1&video_id=%s' % video_id
638 return self._download_webpage(url, video_id, note=u'Searching for annotations.', errnote=u'Unable to download video annotations.')
640 def _real_extract(self, url):
642 'http' if self._downloader.params.get('prefer_insecure', False)
645 # Extract original video URL from URL with redirection, like age verification, using next_url parameter
646 mobj = re.search(self._NEXT_URL_RE, url)
648 url = proto + '://www.youtube.com/' + compat_urllib_parse.unquote(mobj.group(1)).lstrip('/')
649 video_id = self.extract_id(url)
652 url = proto + '://www.youtube.com/watch?v=%s&gl=US&hl=en&has_verified=1' % video_id
653 video_webpage = self._download_webpage(url, video_id)
655 # Attempt to extract SWF player URL
656 mobj = re.search(r'swfConfig.*?"(https?:\\/\\/.*?watch.*?-.*?\.swf)"', video_webpage)
658 player_url = re.sub(r'\\(.)', r'\1', mobj.group(1))
663 self.report_video_info_webpage_download(video_id)
664 if re.search(r'player-age-gate-content">', video_webpage) is not None:
665 self.report_age_confirmation()
667 # We simulate the access to the video from www.youtube.com/v/{video_id}
668 # this can be viewed without login into Youtube
669 data = compat_urllib_parse.urlencode({
670 'video_id': video_id,
671 'eurl': 'https://youtube.googleapis.com/v/' + video_id,
672 'sts': self._search_regex(
673 r'"sts"\s*:\s*(\d+)', video_webpage, 'sts'),
675 video_info_url = proto + '://www.youtube.com/get_video_info?' + data
676 video_info_webpage = self._download_webpage(video_info_url, video_id,
678 errnote='unable to download video info webpage')
679 video_info = compat_parse_qs(video_info_webpage)
682 for el_type in ['&el=embedded', '&el=detailpage', '&el=vevo', '']:
683 video_info_url = (proto + '://www.youtube.com/get_video_info?&video_id=%s%s&ps=default&eurl=&gl=US&hl=en'
684 % (video_id, el_type))
685 video_info_webpage = self._download_webpage(video_info_url, video_id,
687 errnote='unable to download video info webpage')
688 video_info = compat_parse_qs(video_info_webpage)
689 if 'token' in video_info:
691 if 'token' not in video_info:
692 if 'reason' in video_info:
693 raise ExtractorError(
694 'YouTube said: %s' % video_info['reason'][0],
695 expected=True, video_id=video_id)
697 raise ExtractorError(
698 '"token" parameter not in video info for unknown reason',
701 if 'view_count' in video_info:
702 view_count = int(video_info['view_count'][0])
706 # Check for "rental" videos
707 if 'ypc_video_rental_bar_text' in video_info and 'author' not in video_info:
708 raise ExtractorError(u'"rental" videos not supported')
710 # Start extracting information
711 self.report_information_extraction(video_id)
714 if 'author' not in video_info:
715 raise ExtractorError(u'Unable to extract uploader name')
716 video_uploader = compat_urllib_parse.unquote_plus(video_info['author'][0])
719 video_uploader_id = None
720 mobj = re.search(r'<link itemprop="url" href="http://www.youtube.com/(?:user|channel)/([^"]+)">', video_webpage)
722 video_uploader_id = mobj.group(1)
724 self._downloader.report_warning(u'unable to extract uploader nickname')
727 if 'title' in video_info:
728 video_title = video_info['title'][0]
730 self._downloader.report_warning(u'Unable to extract video title')
734 # We try first to get a high quality image:
735 m_thumb = re.search(r'<span itemprop="thumbnail".*?href="(.*?)">',
736 video_webpage, re.DOTALL)
737 if m_thumb is not None:
738 video_thumbnail = m_thumb.group(1)
739 elif 'thumbnail_url' not in video_info:
740 self._downloader.report_warning(u'unable to extract video thumbnail')
741 video_thumbnail = None
742 else: # don't panic if we can't find it
743 video_thumbnail = compat_urllib_parse.unquote_plus(video_info['thumbnail_url'][0])
747 mobj = re.search(r'(?s)id="eow-date.*?>(.*?)</span>', video_webpage)
750 r'(?s)id="watch-uploader-info".*?>.*?(?:Published|Uploaded|Streamed live) on (.*?)</strong>',
753 upload_date = ' '.join(re.sub(r'[/,-]', r' ', mobj.group(1)).split())
754 upload_date = unified_strdate(upload_date)
756 m_cat_container = self._search_regex(
757 r'(?s)<h4[^>]*>\s*Category\s*</h4>\s*<ul[^>]*>(.*?)</ul>',
758 video_webpage, 'categories', fatal=False)
760 category = self._html_search_regex(
761 r'(?s)<a[^<]+>(.*?)</a>', m_cat_container, 'category',
763 video_categories = None if category is None else [category]
765 video_categories = None
768 video_description = get_element_by_id("eow-description", video_webpage)
769 if video_description:
770 video_description = re.sub(r'''(?x)
772 (?:[a-zA-Z-]+="[^"]+"\s+)*?
774 (?:[a-zA-Z-]+="[^"]+"\s+)*?
775 class="yt-uix-redirect-link"\s*>
778 ''', r'\1', video_description)
779 video_description = clean_html(video_description)
781 fd_mobj = re.search(r'<meta name="description" content="([^"]+)"', video_webpage)
783 video_description = unescapeHTML(fd_mobj.group(1))
785 video_description = ''
787 def _extract_count(count_name):
788 count = self._search_regex(
789 r'id="watch-%s"[^>]*>.*?([\d,]+)\s*</span>' % re.escape(count_name),
790 video_webpage, count_name, default=None)
791 if count is not None:
792 return int(count.replace(',', ''))
794 like_count = _extract_count(u'like')
795 dislike_count = _extract_count(u'dislike')
798 video_subtitles = self.extract_subtitles(video_id, video_webpage)
800 if self._downloader.params.get('listsubtitles', False):
801 self._list_available_subtitles(video_id, video_webpage)
804 if 'length_seconds' not in video_info:
805 self._downloader.report_warning(u'unable to extract video duration')
806 video_duration = None
808 video_duration = int(compat_urllib_parse.unquote_plus(video_info['length_seconds'][0]))
811 video_annotations = None
812 if self._downloader.params.get('writeannotations', False):
813 video_annotations = self._extract_annotations(video_id)
815 # Decide which formats to download
817 mobj = re.search(r';ytplayer\.config\s*=\s*({.*?});', video_webpage)
819 raise ValueError('Could not find vevo ID')
820 json_code = uppercase_escape(mobj.group(1))
821 ytplayer_config = json.loads(json_code)
822 args = ytplayer_config['args']
823 # Easy way to know if the 's' value is in url_encoded_fmt_stream_map
824 # this signatures are encrypted
825 if 'url_encoded_fmt_stream_map' not in args:
826 raise ValueError(u'No stream_map present') # caught below
827 re_signature = re.compile(r'[&,]s=')
828 m_s = re_signature.search(args['url_encoded_fmt_stream_map'])
830 self.to_screen(u'%s: Encrypted signatures detected.' % video_id)
831 video_info['url_encoded_fmt_stream_map'] = [args['url_encoded_fmt_stream_map']]
832 m_s = re_signature.search(args.get('adaptive_fmts', ''))
834 if 'adaptive_fmts' in video_info:
835 video_info['adaptive_fmts'][0] += ',' + args['adaptive_fmts']
837 video_info['adaptive_fmts'] = [args['adaptive_fmts']]
841 def _map_to_format_list(urlmap):
843 for itag, video_real_url in urlmap.items():
846 'url': video_real_url,
847 'player_url': player_url,
849 if itag in self._formats:
850 dct.update(self._formats[itag])
854 if 'conn' in video_info and video_info['conn'][0].startswith('rtmp'):
855 self.report_rtmp_download()
857 'format_id': '_rtmp',
859 'url': video_info['conn'][0],
860 'player_url': player_url,
862 elif len(video_info.get('url_encoded_fmt_stream_map', [])) >= 1 or len(video_info.get('adaptive_fmts', [])) >= 1:
863 encoded_url_map = video_info.get('url_encoded_fmt_stream_map', [''])[0] + ',' + video_info.get('adaptive_fmts',[''])[0]
864 if 'rtmpe%3Dyes' in encoded_url_map:
865 raise ExtractorError('rtmpe downloads are not supported, see https://github.com/rg3/youtube-dl/issues/343 for more information.', expected=True)
867 for url_data_str in encoded_url_map.split(','):
868 url_data = compat_parse_qs(url_data_str)
869 if 'itag' not in url_data or 'url' not in url_data:
871 format_id = url_data['itag'][0]
872 url = url_data['url'][0]
874 if 'sig' in url_data:
875 url += '&signature=' + url_data['sig'][0]
876 elif 's' in url_data:
877 encrypted_sig = url_data['s'][0]
880 jsplayer_url_json = self._search_regex(
881 r'"assets":.+?"js":\s*("[^"]+")',
882 video_webpage, 'JS player URL')
883 player_url = json.loads(jsplayer_url_json)
884 if player_url is None:
885 player_url_json = self._search_regex(
886 r'ytplayer\.config.*?"url"\s*:\s*("[^"]+")',
887 video_webpage, 'age gate player URL')
888 player_url = json.loads(player_url_json)
890 if self._downloader.params.get('verbose'):
891 if player_url is None:
892 player_version = 'unknown'
893 player_desc = 'unknown'
895 if player_url.endswith('swf'):
896 player_version = self._search_regex(
897 r'-(.+?)(?:/watch_as3)?\.swf$', player_url,
898 'flash player', fatal=False)
899 player_desc = 'flash player %s' % player_version
901 player_version = self._search_regex(
902 r'html5player-([^/]+?)(?:/html5player)?\.js',
904 'html5 player', fatal=False)
905 player_desc = 'html5 player %s' % player_version
907 parts_sizes = self._signature_cache_id(encrypted_sig)
908 self.to_screen(u'{%s} signature length %s, %s' %
909 (format_id, parts_sizes, player_desc))
911 signature = self._decrypt_signature(
912 encrypted_sig, video_id, player_url, age_gate)
913 url += '&signature=' + signature
914 if 'ratebypass' not in url:
915 url += '&ratebypass=yes'
916 url_map[format_id] = url
917 formats = _map_to_format_list(url_map)
918 elif video_info.get('hlsvp'):
919 manifest_url = video_info['hlsvp'][0]
920 url_map = self._extract_from_m3u8(manifest_url, video_id)
921 formats = _map_to_format_list(url_map)
923 raise ExtractorError(u'no conn, hlsvp or url_encoded_fmt_stream_map information found in video info')
925 # Look for the DASH manifest
926 if (self._downloader.params.get('youtube_include_dash_manifest', False)):
928 # The DASH manifest used needs to be the one from the original video_webpage.
929 # The one found in get_video_info seems to be using different signatures.
930 # However, in the case of an age restriction there won't be any embedded dashmpd in the video_webpage.
931 # Luckily, it seems, this case uses some kind of default signature (len == 86), so the
932 # combination of get_video_info and the _static_decrypt_signature() decryption fallback will work here.
934 dash_manifest_url = video_info.get('dashmpd')[0]
936 dash_manifest_url = ytplayer_config['args']['dashmpd']
937 def decrypt_sig(mobj):
939 dec_s = self._decrypt_signature(s, video_id, player_url, age_gate)
940 return '/signature/%s' % dec_s
941 dash_manifest_url = re.sub(r'/s/([\w\.]+)', decrypt_sig, dash_manifest_url)
942 dash_doc = self._download_xml(
943 dash_manifest_url, video_id,
944 note=u'Downloading DASH manifest',
945 errnote=u'Could not download DASH manifest')
946 for r in dash_doc.findall(u'.//{urn:mpeg:DASH:schema:MPD:2011}Representation'):
947 url_el = r.find('{urn:mpeg:DASH:schema:MPD:2011}BaseURL')
950 format_id = r.attrib['id']
951 video_url = url_el.text
952 filesize = int_or_none(url_el.attrib.get('{http://youtube.com/yt/2012/10/10}contentLength'))
954 'format_id': format_id,
956 'width': int_or_none(r.attrib.get('width')),
957 'tbr': int_or_none(r.attrib.get('bandwidth'), 1000),
958 'asr': int_or_none(r.attrib.get('audioSamplingRate')),
959 'filesize': filesize,
962 existing_format = next(
964 if fo['format_id'] == format_id)
965 except StopIteration:
966 f.update(self._formats.get(format_id, {}))
969 existing_format.update(f)
971 except (ExtractorError, KeyError) as e:
972 self.report_warning(u'Skipping DASH manifest: %s' % e, video_id)
974 self._sort_formats(formats)
978 'uploader': video_uploader,
979 'uploader_id': video_uploader_id,
980 'upload_date': upload_date,
981 'title': video_title,
982 'thumbnail': video_thumbnail,
983 'description': video_description,
984 'categories': video_categories,
985 'subtitles': video_subtitles,
986 'duration': video_duration,
987 'age_limit': 18 if age_gate else 0,
988 'annotations': video_annotations,
989 'webpage_url': proto + '://www.youtube.com/watch?v=%s' % video_id,
990 'view_count': view_count,
991 'like_count': like_count,
992 'dislike_count': dislike_count,
996 class YoutubePlaylistIE(YoutubeBaseInfoExtractor):
997 IE_DESC = 'YouTube.com playlists'
998 _VALID_URL = r"""(?x)(?:
1003 (?:course|view_play_list|my_playlists|artist|playlist|watch)
1004 \? (?:.*?&)*? (?:p|a|list)=
1008 (?:PL|LL|EC|UU|FL|RD)?[0-9A-Za-z-_]{10,}
1009 # Top tracks, they can also include dots
1014 ((?:PL|LL|EC|UU|FL|RD)[0-9A-Za-z-_]{10,})
1016 _TEMPLATE_URL = 'https://www.youtube.com/playlist?list=%s'
1017 _MORE_PAGES_INDICATOR = r'data-link-type="next"'
1018 _VIDEO_RE = r'href="\s*/watch\?v=(?P<id>[0-9A-Za-z_-]{11})&[^"]*?index=(?P<index>\d+)'
1019 IE_NAME = 'youtube:playlist'
1021 'url': 'https://www.youtube.com/playlist?list=PLwiyx1dc3P2JR9N8gQaQN_BCvlSlap7re',
1023 'title': 'ytdl test PL',
1025 'playlist_count': 3,
1027 'url': 'https://www.youtube.com/playlist?list=PLtPgu7CB4gbZDA7i_euNxn75ISqxwZPYx',
1029 'title': 'YDL_Empty_List',
1031 'playlist_count': 0,
1033 'note': 'Playlist with deleted videos (#651). As a bonus, the video #51 is also twice in this list.',
1034 'url': 'https://www.youtube.com/playlist?list=PLwP_SiAcdui0KVebT0mU9Apz359a4ubsC',
1036 'title': '29C3: Not my department',
1038 'playlist_count': 95,
1040 'note': 'issue #673',
1041 'url': 'PLBB231211A4F62143',
1043 'title': 'Team Fortress 2 (Class-based LP)',
1045 'playlist_mincount': 26,
1047 'note': 'Large playlist',
1048 'url': 'https://www.youtube.com/playlist?list=UUBABnxM4Ar9ten8Mdjj1j0Q',
1050 'title': 'Uploads from Cauchemar',
1052 'playlist_mincount': 799,
1054 'url': 'PLtPgu7CB4gbY9oDN3drwC3cMbJggS7dKl',
1056 'title': 'YDL_safe_search',
1058 'playlist_count': 2,
1061 def _real_initialize(self):
1064 def _ids_to_results(self, ids):
1066 self.url_result(vid_id, 'Youtube', video_id=vid_id)
1069 def _extract_mix(self, playlist_id):
1070 # The mixes are generated from a a single video
1071 # the id of the playlist is just 'RD' + video_id
1072 url = 'https://youtube.com/watch?v=%s&list=%s' % (playlist_id[-11:], playlist_id)
1073 webpage = self._download_webpage(
1074 url, playlist_id, 'Downloading Youtube mix')
1075 search_title = lambda class_name: get_element_by_attribute('class', class_name, webpage)
1077 search_title('playlist-title') or
1078 search_title('title long-title') or
1079 search_title('title'))
1080 title = clean_html(title_span)
1081 ids = orderedSet(re.findall(
1082 r'''(?xs)data-video-username=".*?".*?
1083 href="/watch\?v=([0-9A-Za-z_-]{11})&[^"]*?list=%s''' % re.escape(playlist_id),
1085 url_results = self._ids_to_results(ids)
1087 return self.playlist_result(url_results, playlist_id, title)
1089 def _real_extract(self, url):
1090 # Extract playlist id
1091 mobj = re.match(self._VALID_URL, url)
1093 raise ExtractorError(u'Invalid URL: %s' % url)
1094 playlist_id = mobj.group(1) or mobj.group(2)
1096 # Check if it's a video-specific URL
1097 query_dict = compat_urlparse.parse_qs(compat_urlparse.urlparse(url).query)
1098 if 'v' in query_dict:
1099 video_id = query_dict['v'][0]
1100 if self._downloader.params.get('noplaylist'):
1101 self.to_screen(u'Downloading just video %s because of --no-playlist' % video_id)
1102 return self.url_result(video_id, 'Youtube', video_id=video_id)
1104 self.to_screen(u'Downloading playlist %s - add --no-playlist to just download video %s' % (playlist_id, video_id))
1106 if playlist_id.startswith('RD'):
1107 # Mixes require a custom extraction process
1108 return self._extract_mix(playlist_id)
1109 if playlist_id.startswith('TL'):
1110 raise ExtractorError(u'For downloading YouTube.com top lists, use '
1111 'the "yttoplist" keyword, for example "youtube-dl \'yttoplist:music:Top Tracks\'"', expected=True)
1113 url = self._TEMPLATE_URL % playlist_id
1114 page = self._download_webpage(url, playlist_id)
1115 more_widget_html = content_html = page
1117 # Check if the playlist exists or is private
1118 if re.search(r'<div class="yt-alert-message">[^<]*?(The|This) playlist (does not exist|is private)[^<]*?</div>', page) is not None:
1119 raise ExtractorError(
1120 'The playlist doesn\'t exist or is private, use --username or '
1121 '--netrc to access it.',
1124 # Extract the video ids from the playlist pages
1127 for page_num in itertools.count(1):
1128 matches = re.finditer(self._VIDEO_RE, content_html)
1129 # We remove the duplicates and the link with index 0
1130 # (it's not the first video of the playlist)
1131 new_ids = orderedSet(m.group('id') for m in matches if m.group('index') != '0')
1134 mobj = re.search(r'data-uix-load-more-href="/?(?P<more>[^"]+)"', more_widget_html)
1138 more = self._download_json(
1139 'https://youtube.com/%s' % mobj.group('more'), playlist_id,
1140 'Downloading page #%s' % page_num,
1141 transform_source=uppercase_escape)
1142 content_html = more['content_html']
1143 more_widget_html = more['load_more_widget_html']
1145 playlist_title = self._html_search_regex(
1146 r'(?s)<h1 class="pl-header-title[^"]*">\s*(.*?)\s*</h1>',
1149 url_results = self._ids_to_results(ids)
1150 return self.playlist_result(url_results, playlist_id, playlist_title)
1153 class YoutubeTopListIE(YoutubePlaylistIE):
1154 IE_NAME = 'youtube:toplist'
1155 IE_DESC = (u'YouTube.com top lists, "yttoplist:{channel}:{list title}"'
1156 ' (Example: "yttoplist:music:Top Tracks")')
1157 _VALID_URL = r'yttoplist:(?P<chann>.*?):(?P<title>.*?)$'
1160 def _real_extract(self, url):
1161 mobj = re.match(self._VALID_URL, url)
1162 channel = mobj.group('chann')
1163 title = mobj.group('title')
1164 query = compat_urllib_parse.urlencode({'title': title})
1165 playlist_re = 'href="([^"]+?%s.*?)"' % re.escape(query)
1166 channel_page = self._download_webpage('https://www.youtube.com/%s' % channel, title)
1167 link = self._html_search_regex(playlist_re, channel_page, 'list')
1168 url = compat_urlparse.urljoin('https://www.youtube.com/', link)
1170 video_re = r'data-index="\d+".*?data-video-id="([0-9A-Za-z_-]{11})"'
1172 # sometimes the webpage doesn't contain the videos
1173 # retry until we get them
1174 for i in itertools.count(0):
1175 msg = 'Downloading Youtube mix'
1177 msg += ', retry #%d' % i
1179 webpage = self._download_webpage(url, title, msg)
1180 ids = orderedSet(re.findall(video_re, webpage))
1183 url_results = self._ids_to_results(ids)
1184 return self.playlist_result(url_results, playlist_title=title)
1187 class YoutubeChannelIE(InfoExtractor):
1188 IE_DESC = 'YouTube.com channels'
1189 _VALID_URL = r"^(?:https?://)?(?:youtu\.be|(?:\w+\.)?youtube(?:-nocookie)?\.com)/channel/([0-9A-Za-z_-]+)"
1190 _MORE_PAGES_INDICATOR = 'yt-uix-load-more'
1191 _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'
1192 IE_NAME = 'youtube:channel'
1194 def extract_videos_from_page(self, page):
1196 for mobj in re.finditer(r'href="/watch\?v=([0-9A-Za-z_-]+)&?', page):
1197 if mobj.group(1) not in ids_in_page:
1198 ids_in_page.append(mobj.group(1))
1201 def _real_extract(self, url):
1202 # Extract channel id
1203 mobj = re.match(self._VALID_URL, url)
1205 raise ExtractorError(u'Invalid URL: %s' % url)
1207 # Download channel page
1208 channel_id = mobj.group(1)
1210 url = 'https://www.youtube.com/channel/%s/videos' % channel_id
1211 channel_page = self._download_webpage(url, channel_id)
1212 autogenerated = re.search(r'''(?x)
1214 channel-header-autogenerated-label|
1215 yt-channel-title-autogenerated
1216 )[^"]*"''', channel_page) is not None
1219 # The videos are contained in a single page
1220 # the ajax pages can't be used, they are empty
1221 video_ids = self.extract_videos_from_page(channel_page)
1223 # Download all channel pages using the json-based channel_ajax query
1224 for pagenum in itertools.count(1):
1225 url = self._MORE_PAGES_URL % (pagenum, channel_id)
1226 page = self._download_json(
1227 url, channel_id, note=u'Downloading page #%s' % pagenum,
1228 transform_source=uppercase_escape)
1230 ids_in_page = self.extract_videos_from_page(page['content_html'])
1231 video_ids.extend(ids_in_page)
1233 if self._MORE_PAGES_INDICATOR not in page['load_more_widget_html']:
1236 self._downloader.to_screen(u'[youtube] Channel %s: Found %i videos' % (channel_id, len(video_ids)))
1238 url_entries = [self.url_result(video_id, 'Youtube', video_id=video_id)
1239 for video_id in video_ids]
1240 return self.playlist_result(url_entries, channel_id)
1243 class YoutubeUserIE(InfoExtractor):
1244 IE_DESC = 'YouTube.com user videos (URL or "ytuser" keyword)'
1245 _VALID_URL = r'(?:(?:(?:https?://)?(?:\w+\.)?youtube\.com/(?:user/)?(?!(?:attribution_link|watch|results)(?:$|[^a-z_A-Z0-9-])))|ytuser:)(?!feed/)([A-Za-z0-9_-]+)'
1246 _TEMPLATE_URL = 'https://gdata.youtube.com/feeds/api/users/%s'
1247 _GDATA_PAGE_SIZE = 50
1248 _GDATA_URL = 'https://gdata.youtube.com/feeds/api/users/%s/uploads?max-results=%d&start-index=%d&alt=json'
1249 IE_NAME = 'youtube:user'
1252 def suitable(cls, url):
1253 # Don't return True if the url can be extracted with other youtube
1254 # extractor, the regex would is too permissive and it would match.
1255 other_ies = iter(klass for (name, klass) in globals().items() if name.endswith('IE') and klass is not cls)
1256 if any(ie.suitable(url) for ie in other_ies): return False
1257 else: return super(YoutubeUserIE, cls).suitable(url)
1259 def _real_extract(self, url):
1261 mobj = re.match(self._VALID_URL, url)
1263 raise ExtractorError(u'Invalid URL: %s' % url)
1265 username = mobj.group(1)
1267 # Download video ids using YouTube Data API. Result size per
1268 # query is limited (currently to 50 videos) so we need to query
1269 # page by page until there are no video ids - it means we got
1272 def download_page(pagenum):
1273 start_index = pagenum * self._GDATA_PAGE_SIZE + 1
1275 gdata_url = self._GDATA_URL % (username, self._GDATA_PAGE_SIZE, start_index)
1276 page = self._download_webpage(
1277 gdata_url, username,
1278 'Downloading video ids from %d to %d' % (
1279 start_index, start_index + self._GDATA_PAGE_SIZE))
1282 response = json.loads(page)
1283 except ValueError as err:
1284 raise ExtractorError(u'Invalid JSON in API response: ' + compat_str(err))
1285 if 'entry' not in response['feed']:
1288 # Extract video identifiers
1289 entries = response['feed']['entry']
1290 for entry in entries:
1291 title = entry['title']['$t']
1292 video_id = entry['id']['$t'].split('/')[-1]
1296 'ie_key': 'Youtube',
1300 url_results = PagedList(download_page, self._GDATA_PAGE_SIZE)
1302 return self.playlist_result(url_results, playlist_title=username)
1305 class YoutubeSearchIE(SearchInfoExtractor):
1306 IE_DESC = 'YouTube.com searches'
1307 _API_URL = 'https://gdata.youtube.com/feeds/api/videos?q=%s&start-index=%i&max-results=50&v=2&alt=jsonc'
1309 IE_NAME = 'youtube:search'
1310 _SEARCH_KEY = 'ytsearch'
1312 def _get_n_results(self, query, n):
1313 """Get a specified number of results for a query"""
1320 while (PAGE_SIZE * pagenum) < limit:
1321 result_url = self._API_URL % (
1322 compat_urllib_parse.quote_plus(query.encode('utf-8')),
1323 (PAGE_SIZE * pagenum) + 1)
1324 data_json = self._download_webpage(
1325 result_url, video_id=u'query "%s"' % query,
1326 note=u'Downloading page %s' % (pagenum + 1),
1327 errnote=u'Unable to download API page')
1328 data = json.loads(data_json)
1329 api_response = data['data']
1331 if 'items' not in api_response:
1332 raise ExtractorError(
1333 '[youtube] No video results', expected=True)
1335 new_ids = list(video['id'] for video in api_response['items'])
1336 video_ids += new_ids
1338 limit = min(n, api_response['totalItems'])
1341 if len(video_ids) > n:
1342 video_ids = video_ids[:n]
1343 videos = [self.url_result(video_id, 'Youtube', video_id=video_id)
1344 for video_id in video_ids]
1345 return self.playlist_result(videos, query)
1348 class YoutubeSearchDateIE(YoutubeSearchIE):
1349 IE_NAME = YoutubeSearchIE.IE_NAME + ':date'
1350 _API_URL = 'https://gdata.youtube.com/feeds/api/videos?q=%s&start-index=%i&max-results=50&v=2&alt=jsonc&orderby=published'
1351 _SEARCH_KEY = 'ytsearchdate'
1352 IE_DESC = 'YouTube.com searches, newest videos first'
1355 class YoutubeSearchURLIE(InfoExtractor):
1356 IE_DESC = 'YouTube.com search URLs'
1357 IE_NAME = 'youtube:search_url'
1358 _VALID_URL = r'https?://(?:www\.)?youtube\.com/results\?(.*?&)?search_query=(?P<query>[^&]+)(?:[&]|$)'
1360 def _real_extract(self, url):
1361 mobj = re.match(self._VALID_URL, url)
1362 query = compat_urllib_parse.unquote_plus(mobj.group('query'))
1364 webpage = self._download_webpage(url, query)
1365 result_code = self._search_regex(
1366 r'(?s)<ol class="item-section"(.*?)</ol>', webpage, 'result HTML')
1368 part_codes = re.findall(
1369 r'(?s)<h3 class="yt-lockup-title">(.*?)</h3>', result_code)
1371 for part_code in part_codes:
1372 part_title = self._html_search_regex(
1373 [r'(?s)title="([^"]+)"', r'>([^<]+)</a>'], part_code, 'item title', fatal=False)
1374 part_url_snippet = self._html_search_regex(
1375 r'(?s)href="([^"]+)"', part_code, 'item URL')
1376 part_url = compat_urlparse.urljoin(
1377 'https://www.youtube.com/', part_url_snippet)
1381 'title': part_title,
1385 '_type': 'playlist',
1391 class YoutubeShowIE(InfoExtractor):
1392 IE_DESC = 'YouTube.com (multi-season) shows'
1393 _VALID_URL = r'https?://www\.youtube\.com/show/(.*)'
1394 IE_NAME = 'youtube:show'
1396 def _real_extract(self, url):
1397 mobj = re.match(self._VALID_URL, url)
1398 show_name = mobj.group(1)
1399 webpage = self._download_webpage(url, show_name, 'Downloading show webpage')
1400 # There's one playlist for each season of the show
1401 m_seasons = list(re.finditer(r'href="(/playlist\?list=.*?)"', webpage))
1402 self.to_screen(u'%s: Found %s seasons' % (show_name, len(m_seasons)))
1403 return [self.url_result('https://www.youtube.com' + season.group(1), 'YoutubePlaylist') for season in m_seasons]
1406 class YoutubeFeedsInfoExtractor(YoutubeBaseInfoExtractor):
1408 Base class for extractors that fetch info from
1409 http://www.youtube.com/feed_ajax
1410 Subclasses must define the _FEED_NAME and _PLAYLIST_TITLE properties.
1412 _LOGIN_REQUIRED = True
1413 # use action_load_personal_feed instead of action_load_system_feed
1414 _PERSONAL_FEED = False
1417 def _FEED_TEMPLATE(self):
1418 action = 'action_load_system_feed'
1419 if self._PERSONAL_FEED:
1420 action = 'action_load_personal_feed'
1421 return 'https://www.youtube.com/feed_ajax?%s=1&feed_name=%s&paging=%%s' % (action, self._FEED_NAME)
1425 return 'youtube:%s' % self._FEED_NAME
1427 def _real_initialize(self):
1430 def _real_extract(self, url):
1433 for i in itertools.count(1):
1434 info = self._download_json(self._FEED_TEMPLATE % paging,
1435 '%s feed' % self._FEED_NAME,
1436 'Downloading page %s' % i)
1437 feed_html = info.get('feed_html') or info.get('content_html')
1438 load_more_widget_html = info.get('load_more_widget_html') or feed_html
1439 m_ids = re.finditer(r'"/watch\?v=(.*?)["&]', feed_html)
1440 ids = orderedSet(m.group(1) for m in m_ids)
1441 feed_entries.extend(
1442 self.url_result(video_id, 'Youtube', video_id=video_id)
1443 for video_id in ids)
1445 r'data-uix-load-more-href="/?[^"]+paging=(?P<paging>\d+)',
1446 load_more_widget_html)
1449 paging = mobj.group('paging')
1450 return self.playlist_result(feed_entries, playlist_title=self._PLAYLIST_TITLE)
1452 class YoutubeRecommendedIE(YoutubeFeedsInfoExtractor):
1453 IE_DESC = 'YouTube.com recommended videos, "ytrec" keyword (requires authentication)'
1454 _VALID_URL = r'https?://www\.youtube\.com/feed/recommended|:ytrec(?:ommended)?'
1455 _FEED_NAME = 'recommended'
1456 _PLAYLIST_TITLE = 'Youtube Recommended videos'
1458 class YoutubeWatchLaterIE(YoutubeFeedsInfoExtractor):
1459 IE_DESC = 'Youtube watch later list, "ytwatchlater" keyword (requires authentication)'
1460 _VALID_URL = r'https?://www\.youtube\.com/feed/watch_later|:ytwatchlater'
1461 _FEED_NAME = 'watch_later'
1462 _PLAYLIST_TITLE = 'Youtube Watch Later'
1463 _PERSONAL_FEED = True
1465 class YoutubeHistoryIE(YoutubeFeedsInfoExtractor):
1466 IE_DESC = 'Youtube watch history, "ythistory" keyword (requires authentication)'
1467 _VALID_URL = 'https?://www\.youtube\.com/feed/history|:ythistory'
1468 _FEED_NAME = 'history'
1469 _PERSONAL_FEED = True
1470 _PLAYLIST_TITLE = 'Youtube Watch History'
1472 class YoutubeFavouritesIE(YoutubeBaseInfoExtractor):
1473 IE_NAME = 'youtube:favorites'
1474 IE_DESC = 'YouTube.com favourite videos, "ytfav" keyword (requires authentication)'
1475 _VALID_URL = r'https?://www\.youtube\.com/my_favorites|:ytfav(?:ou?rites)?'
1476 _LOGIN_REQUIRED = True
1478 def _real_extract(self, url):
1479 webpage = self._download_webpage('https://www.youtube.com/my_favorites', 'Youtube Favourites videos')
1480 playlist_id = self._search_regex(r'list=(.+?)["&]', webpage, 'favourites playlist id')
1481 return self.url_result(playlist_id, 'YoutubePlaylist')
1484 class YoutubeSubscriptionsIE(YoutubePlaylistIE):
1485 IE_NAME = 'youtube:subscriptions'
1486 IE_DESC = 'YouTube.com subscriptions feed, "ytsubs" keyword (requires authentication)'
1487 _VALID_URL = r'https?://www\.youtube\.com/feed/subscriptions|:ytsubs(?:criptions)?'
1490 def _real_extract(self, url):
1491 title = 'Youtube Subscriptions'
1492 page = self._download_webpage('https://www.youtube.com/feed/subscriptions', title)
1494 # The extraction process is the same as for playlists, but the regex
1495 # for the video ids doesn't contain an index
1497 more_widget_html = content_html = page
1499 for page_num in itertools.count(1):
1500 matches = re.findall(r'href="\s*/watch\?v=([0-9A-Za-z_-]{11})', content_html)
1501 new_ids = orderedSet(matches)
1504 mobj = re.search(r'data-uix-load-more-href="/?(?P<more>[^"]+)"', more_widget_html)
1508 more = self._download_json(
1509 'https://youtube.com/%s' % mobj.group('more'), title,
1510 'Downloading page #%s' % page_num,
1511 transform_source=uppercase_escape)
1512 content_html = more['content_html']
1513 more_widget_html = more['load_more_widget_html']
1516 '_type': 'playlist',
1518 'entries': self._ids_to_results(ids),
1522 class YoutubeTruncatedURLIE(InfoExtractor):
1523 IE_NAME = 'youtube:truncated_url'
1524 IE_DESC = False # Do not list
1525 _VALID_URL = r'''(?x)
1526 (?:https?://)?[^/]+/watch\?(?:
1528 annotation_id=annotation_[^&]+
1530 (?:https?://)?(?:www\.)?youtube\.com/attribution_link\?a=[^&]+$
1534 'url': 'http://www.youtube.com/watch?annotation_id=annotation_3951667041',
1535 'only_matching': True,
1537 'url': 'http://www.youtube.com/watch?',
1538 'only_matching': True,
1541 def _real_extract(self, url):
1542 raise ExtractorError(
1543 'Did you forget to quote the URL? Remember that & is a meta '
1544 'character in most shells, so you want to put the URL in quotes, '
1546 '"http://www.youtube.com/watch?feature=foo&v=BaW_jenozKc" '
1547 ' or simply youtube-dl BaW_jenozKc .',