]> Raphaël G. Git Repositories - youtubedl/blob - youtube_dl/extractor/youtube.py
Update changelog.
[youtubedl] / youtube_dl / extractor / youtube.py
1 # coding: utf-8
2
3 import json
4 import netrc
5 import re
6 import socket
7 import itertools
8
9 from .common import InfoExtractor, SearchInfoExtractor
10 from ..utils import (
11 compat_http_client,
12 compat_parse_qs,
13 compat_urllib_error,
14 compat_urllib_parse,
15 compat_urllib_request,
16 compat_str,
17
18 clean_html,
19 get_element_by_id,
20 ExtractorError,
21 unescapeHTML,
22 unified_strdate,
23 orderedSet,
24 )
25
26 class YoutubeBaseInfoExtractor(InfoExtractor):
27 """Provide base functions for Youtube extractors"""
28 _LOGIN_URL = 'https://accounts.google.com/ServiceLogin'
29 _LANG_URL = r'https://www.youtube.com/?hl=en&persist_hl=1&gl=US&persist_gl=1&opt_out_ackd=1'
30 _AGE_URL = 'http://www.youtube.com/verify_age?next_url=/&gl=US&hl=en'
31 _NETRC_MACHINE = 'youtube'
32 # If True it will raise an error if no login info is provided
33 _LOGIN_REQUIRED = False
34
35 def report_lang(self):
36 """Report attempt to set language."""
37 self.to_screen(u'Setting language')
38
39 def _set_language(self):
40 request = compat_urllib_request.Request(self._LANG_URL)
41 try:
42 self.report_lang()
43 compat_urllib_request.urlopen(request).read()
44 except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
45 self._downloader.report_warning(u'unable to set language: %s' % compat_str(err))
46 return False
47 return True
48
49 def _login(self):
50 (username, password) = self._get_login_info()
51 # No authentication to be performed
52 if username is None:
53 if self._LOGIN_REQUIRED:
54 raise ExtractorError(u'No login info available, needed for using %s.' % self.IE_NAME, expected=True)
55 return False
56
57 request = compat_urllib_request.Request(self._LOGIN_URL)
58 try:
59 login_page = compat_urllib_request.urlopen(request).read().decode('utf-8')
60 except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
61 self._downloader.report_warning(u'unable to fetch login page: %s' % compat_str(err))
62 return False
63
64 galx = None
65 dsh = None
66 match = re.search(re.compile(r'<input.+?name="GALX".+?value="(.+?)"', re.DOTALL), login_page)
67 if match:
68 galx = match.group(1)
69 match = re.search(re.compile(r'<input.+?name="dsh".+?value="(.+?)"', re.DOTALL), login_page)
70 if match:
71 dsh = match.group(1)
72
73 # Log in
74 login_form_strs = {
75 u'continue': u'https://www.youtube.com/signin?action_handle_signin=true&feature=sign_in_button&hl=en_US&nomobiletemp=1',
76 u'Email': username,
77 u'GALX': galx,
78 u'Passwd': password,
79 u'PersistentCookie': u'yes',
80 u'_utf8': u'霱',
81 u'bgresponse': u'js_disabled',
82 u'checkConnection': u'',
83 u'checkedDomains': u'youtube',
84 u'dnConn': u'',
85 u'dsh': dsh,
86 u'pstMsg': u'0',
87 u'rmShown': u'1',
88 u'secTok': u'',
89 u'signIn': u'Sign in',
90 u'timeStmp': u'',
91 u'service': u'youtube',
92 u'uilel': u'3',
93 u'hl': u'en_US',
94 }
95 # Convert to UTF-8 *before* urlencode because Python 2.x's urlencode
96 # chokes on unicode
97 login_form = dict((k.encode('utf-8'), v.encode('utf-8')) for k,v in login_form_strs.items())
98 login_data = compat_urllib_parse.urlencode(login_form).encode('ascii')
99 request = compat_urllib_request.Request(self._LOGIN_URL, login_data)
100 try:
101 self.report_login()
102 login_results = compat_urllib_request.urlopen(request).read().decode('utf-8')
103 if re.search(r'(?i)<form[^>]* id="gaia_loginform"', login_results) is not None:
104 self._downloader.report_warning(u'unable to log in: bad username or password')
105 return False
106 except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
107 self._downloader.report_warning(u'unable to log in: %s' % compat_str(err))
108 return False
109 return True
110
111 def _confirm_age(self):
112 age_form = {
113 'next_url': '/',
114 'action_confirm': 'Confirm',
115 }
116 request = compat_urllib_request.Request(self._AGE_URL, compat_urllib_parse.urlencode(age_form))
117 try:
118 self.report_age_confirmation()
119 compat_urllib_request.urlopen(request).read().decode('utf-8')
120 except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
121 raise ExtractorError(u'Unable to confirm age: %s' % compat_str(err))
122 return True
123
124 def _real_initialize(self):
125 if self._downloader is None:
126 return
127 if not self._set_language():
128 return
129 if not self._login():
130 return
131 self._confirm_age()
132
133 class YoutubeIE(YoutubeBaseInfoExtractor):
134 IE_DESC = u'YouTube.com'
135 _VALID_URL = r"""^
136 (
137 (?:https?://)? # http(s):// (optional)
138 (?:youtu\.be/|(?:\w+\.)?youtube(?:-nocookie)?\.com/|
139 tube\.majestyc\.net/) # the various hostnames, with wildcard subdomains
140 (?:.*?\#/)? # handle anchor (#/) redirect urls
141 (?: # the various things that can precede the ID:
142 (?:(?:v|embed|e)/) # v/ or embed/ or e/
143 |(?: # or the v= param in all its forms
144 (?:watch|movie(?:_popup)?(?:\.php)?)? # preceding watch(_popup|.php) or nothing (like /?v=xxxx)
145 (?:\?|\#!?) # the params delimiter ? or # or #!
146 (?:.*?&)? # any other preceding param (like /?s=tuff&v=xxxx)
147 v=
148 )
149 )? # optional -> youtube.com/xxxx is OK
150 )? # all until now is optional -> you can pass the naked ID
151 ([0-9A-Za-z_-]+) # here is it! the YouTube video ID
152 (?(1).+)? # if we found the ID, everything can follow
153 $"""
154 _NEXT_URL_RE = r'[\?&]next_url=([^&]+)'
155 # Listed in order of quality
156 _available_formats = ['38', '37', '46', '22', '45', '35', '44', '34', '18', '43', '6', '5', '17', '13',
157 '95', '94', '93', '92', '132', '151',
158 '85', '84', '102', '83', '101', '82', '100',
159 ]
160 _available_formats_prefer_free = ['38', '46', '37', '45', '22', '44', '35', '43', '34', '18', '6', '5', '17', '13',
161 '95', '94', '93', '92', '132', '151',
162 '85', '102', '84', '101', '83', '100', '82',
163 ]
164 _video_extensions = {
165 '13': '3gp',
166 '17': 'mp4',
167 '18': 'mp4',
168 '22': 'mp4',
169 '37': 'mp4',
170 '38': 'mp4',
171 '43': 'webm',
172 '44': 'webm',
173 '45': 'webm',
174 '46': 'webm',
175
176 # 3d videos
177 '82': 'mp4',
178 '83': 'mp4',
179 '84': 'mp4',
180 '85': 'mp4',
181 '100': 'webm',
182 '101': 'webm',
183 '102': 'webm',
184
185 # videos that use m3u8
186 '92': 'mp4',
187 '93': 'mp4',
188 '94': 'mp4',
189 '95': 'mp4',
190 '96': 'mp4',
191 '132': 'mp4',
192 '151': 'mp4',
193 }
194 _video_dimensions = {
195 '5': '240x400',
196 '6': '???',
197 '13': '???',
198 '17': '144x176',
199 '18': '360x640',
200 '22': '720x1280',
201 '34': '360x640',
202 '35': '480x854',
203 '37': '1080x1920',
204 '38': '3072x4096',
205 '43': '360x640',
206 '44': '480x854',
207 '45': '720x1280',
208 '46': '1080x1920',
209 '82': '360p',
210 '83': '480p',
211 '84': '720p',
212 '85': '1080p',
213 '92': '240p',
214 '93': '360p',
215 '94': '480p',
216 '95': '720p',
217 '96': '1080p',
218 '100': '360p',
219 '101': '480p',
220 '102': '720p',
221 '132': '240p',
222 '151': '72p',
223 }
224 _3d_itags = ['85', '84', '102', '83', '101', '82', '100']
225 IE_NAME = u'youtube'
226 _TESTS = [
227 {
228 u"url": u"http://www.youtube.com/watch?v=BaW_jenozKc",
229 u"file": u"BaW_jenozKc.mp4",
230 u"info_dict": {
231 u"title": u"youtube-dl test video \"'/\\ä↭𝕐",
232 u"uploader": u"Philipp Hagemeister",
233 u"uploader_id": u"phihag",
234 u"upload_date": u"20121002",
235 u"description": u"test chars: \"'/\\ä↭𝕐\n\nThis is a test video for youtube-dl.\n\nFor more information, contact phihag@phihag.de ."
236 }
237 },
238 {
239 u"url": u"http://www.youtube.com/watch?v=1ltcDfZMA3U",
240 u"file": u"1ltcDfZMA3U.flv",
241 u"note": u"Test VEVO video (#897)",
242 u"info_dict": {
243 u"upload_date": u"20070518",
244 u"title": u"Maps - It Will Find You",
245 u"description": u"Music video by Maps performing It Will Find You.",
246 u"uploader": u"MuteUSA",
247 u"uploader_id": u"MuteUSA"
248 }
249 },
250 {
251 u"url": u"http://www.youtube.com/watch?v=UxxajLWwzqY",
252 u"file": u"UxxajLWwzqY.mp4",
253 u"note": u"Test generic use_cipher_signature video (#897)",
254 u"info_dict": {
255 u"upload_date": u"20120506",
256 u"title": u"Icona Pop - I Love It (feat. Charli XCX) [OFFICIAL VIDEO]",
257 u"description": u"md5:b085c9804f5ab69f4adea963a2dceb3c",
258 u"uploader": u"IconaPop",
259 u"uploader_id": u"IconaPop"
260 }
261 },
262 {
263 u"url": u"https://www.youtube.com/watch?v=07FYdnEawAQ",
264 u"file": u"07FYdnEawAQ.mp4",
265 u"note": u"Test VEVO video with age protection (#956)",
266 u"info_dict": {
267 u"upload_date": u"20130703",
268 u"title": u"Justin Timberlake - Tunnel Vision (Explicit)",
269 u"description": u"md5:64249768eec3bc4276236606ea996373",
270 u"uploader": u"justintimberlakeVEVO",
271 u"uploader_id": u"justintimberlakeVEVO"
272 }
273 },
274 {
275 u'url': u'https://www.youtube.com/watch?v=TGi3HqYrWHE',
276 u'file': u'TGi3HqYrWHE.mp4',
277 u'note': u'm3u8 video',
278 u'info_dict': {
279 u'title': u'Triathlon - Men - London 2012 Olympic Games',
280 u'description': u'- Men - TR02 - Triathlon - 07 August 2012 - London 2012 Olympic Games',
281 u'uploader': u'olympic',
282 u'upload_date': u'20120807',
283 u'uploader_id': u'olympic',
284 },
285 u'params': {
286 u'skip_download': True,
287 },
288 },
289 ]
290
291
292 @classmethod
293 def suitable(cls, url):
294 """Receives a URL and returns True if suitable for this IE."""
295 if YoutubePlaylistIE.suitable(url) or YoutubeSubscriptionsIE.suitable(url): return False
296 return re.match(cls._VALID_URL, url, re.VERBOSE) is not None
297
298 def report_video_webpage_download(self, video_id):
299 """Report attempt to download video webpage."""
300 self.to_screen(u'%s: Downloading video webpage' % video_id)
301
302 def report_video_info_webpage_download(self, video_id):
303 """Report attempt to download video info webpage."""
304 self.to_screen(u'%s: Downloading video info webpage' % video_id)
305
306 def report_video_subtitles_download(self, video_id):
307 """Report attempt to download video info webpage."""
308 self.to_screen(u'%s: Checking available subtitles' % video_id)
309
310 def report_video_subtitles_request(self, video_id, sub_lang, format):
311 """Report attempt to download video info webpage."""
312 self.to_screen(u'%s: Downloading video subtitles for %s.%s' % (video_id, sub_lang, format))
313
314 def report_video_subtitles_available(self, video_id, sub_lang_list):
315 """Report available subtitles."""
316 sub_lang = ",".join(list(sub_lang_list.keys()))
317 self.to_screen(u'%s: Available subtitles for video: %s' % (video_id, sub_lang))
318
319 def report_information_extraction(self, video_id):
320 """Report attempt to extract video information."""
321 self.to_screen(u'%s: Extracting video information' % video_id)
322
323 def report_unavailable_format(self, video_id, format):
324 """Report extracted video URL."""
325 self.to_screen(u'%s: Format %s not available' % (video_id, format))
326
327 def report_rtmp_download(self):
328 """Indicate the download will use the RTMP protocol."""
329 self.to_screen(u'RTMP download detected')
330
331 def _decrypt_signature(self, s):
332 """Turn the encrypted s field into a working signature"""
333
334 if len(s) == 92:
335 return s[25] + s[3:25] + s[0] + s[26:42] + s[79] + s[43:79] + s[91] + s[80:83]
336 elif len(s) == 90:
337 return s[25] + s[3:25] + s[2] + s[26:40] + s[77] + s[41:77] + s[89] + s[78:81]
338 elif len(s) == 89:
339 return s[84:78:-1] + s[87] + s[77:60:-1] + s[0] + s[59:3:-1]
340 elif len(s) == 88:
341 return s[48] + s[81:67:-1] + s[82] + s[66:62:-1] + s[85] + s[61:48:-1] + s[67] + s[47:12:-1] + s[3] + s[11:3:-1] + s[2] + s[12]
342 elif len(s) == 87:
343 return s[6:27] + s[4] + s[28:39] + s[27] + s[40:59] + s[2] + s[60:]
344 elif len(s) == 86:
345 return s[5:20] + s[2] + s[21:]
346 elif len(s) == 85:
347 return s[83:34:-1] + s[0] + s[33:27:-1] + s[3] + s[26:19:-1] + s[34] + s[18:3:-1] + s[27]
348 elif len(s) == 84:
349 return s[83:27:-1] + s[0] + s[26:5:-1] + s[2:0:-1] + s[27]
350 elif len(s) == 83:
351 return s[81:64:-1] + s[82] + s[63:52:-1] + s[45] + s[51:45:-1] + s[1] + s[44:1:-1] + s[0]
352 elif len(s) == 82:
353 return s[36] + s[79:67:-1] + s[81] + s[66:40:-1] + s[33] + s[39:36:-1] + s[40] + s[35] + s[0] + s[67] + s[32:0:-1] + s[34]
354 elif len(s) == 81:
355 return s[56] + s[79:56:-1] + s[41] + s[55:41:-1] + s[80] + s[40:34:-1] + s[0] + s[33:29:-1] + s[34] + s[28:9:-1] + s[29] + s[8:0:-1] + s[9]
356 elif len(s) == 79:
357 return s[54] + s[77:54:-1] + s[39] + s[53:39:-1] + s[78] + s[38:34:-1] + s[0] + s[33:29:-1] + s[34] + s[28:9:-1] + s[29] + s[8:0:-1] + s[9]
358
359 else:
360 raise ExtractorError(u'Unable to decrypt signature, key length %d not supported; retrying might work' % (len(s)))
361
362 def _decrypt_signature_age_gate(self, s):
363 # The videos with age protection use another player, so the algorithms
364 # can be different.
365 if len(s) == 86:
366 return s[2:63] + s[82] + s[64:82] + s[63]
367 else:
368 # Fallback to the other algortihms
369 return self._decrypt_signature(s)
370
371
372 def _get_available_subtitles(self, video_id):
373 self.report_video_subtitles_download(video_id)
374 request = compat_urllib_request.Request('http://video.google.com/timedtext?hl=en&type=list&v=%s' % video_id)
375 try:
376 sub_list = compat_urllib_request.urlopen(request).read().decode('utf-8')
377 except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
378 return (u'unable to download video subtitles: %s' % compat_str(err), None)
379 sub_lang_list = re.findall(r'name="([^"]*)"[^>]+lang_code="([\w\-]+)"', sub_list)
380 sub_lang_list = dict((l[1], l[0]) for l in sub_lang_list)
381 if not sub_lang_list:
382 return (u'video doesn\'t have subtitles', None)
383 return sub_lang_list
384
385 def _list_available_subtitles(self, video_id):
386 sub_lang_list = self._get_available_subtitles(video_id)
387 self.report_video_subtitles_available(video_id, sub_lang_list)
388
389 def _request_subtitle(self, sub_lang, sub_name, video_id, format):
390 """
391 Return tuple:
392 (error_message, sub_lang, sub)
393 """
394 self.report_video_subtitles_request(video_id, sub_lang, format)
395 params = compat_urllib_parse.urlencode({
396 'lang': sub_lang,
397 'name': sub_name,
398 'v': video_id,
399 'fmt': format,
400 })
401 url = 'http://www.youtube.com/api/timedtext?' + params
402 try:
403 sub = compat_urllib_request.urlopen(url).read().decode('utf-8')
404 except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
405 return (u'unable to download video subtitles: %s' % compat_str(err), None, None)
406 if not sub:
407 return (u'Did not fetch video subtitles', None, None)
408 return (None, sub_lang, sub)
409
410 def _request_automatic_caption(self, video_id, webpage):
411 """We need the webpage for getting the captions url, pass it as an
412 argument to speed up the process."""
413 sub_lang = self._downloader.params.get('subtitleslang') or 'en'
414 sub_format = self._downloader.params.get('subtitlesformat')
415 self.to_screen(u'%s: Looking for automatic captions' % video_id)
416 mobj = re.search(r';ytplayer.config = ({.*?});', webpage)
417 err_msg = u'Couldn\'t find automatic captions for "%s"' % sub_lang
418 if mobj is None:
419 return [(err_msg, None, None)]
420 player_config = json.loads(mobj.group(1))
421 try:
422 args = player_config[u'args']
423 caption_url = args[u'ttsurl']
424 timestamp = args[u'timestamp']
425 params = compat_urllib_parse.urlencode({
426 'lang': 'en',
427 'tlang': sub_lang,
428 'fmt': sub_format,
429 'ts': timestamp,
430 'kind': 'asr',
431 })
432 subtitles_url = caption_url + '&' + params
433 sub = self._download_webpage(subtitles_url, video_id, u'Downloading automatic captions')
434 return [(None, sub_lang, sub)]
435 except KeyError:
436 return [(err_msg, None, None)]
437
438 def _extract_subtitle(self, video_id):
439 """
440 Return a list with a tuple:
441 [(error_message, sub_lang, sub)]
442 """
443 sub_lang_list = self._get_available_subtitles(video_id)
444 sub_format = self._downloader.params.get('subtitlesformat')
445 if isinstance(sub_lang_list,tuple): #There was some error, it didn't get the available subtitles
446 return [(sub_lang_list[0], None, None)]
447 if self._downloader.params.get('subtitleslang', False):
448 sub_lang = self._downloader.params.get('subtitleslang')
449 elif 'en' in sub_lang_list:
450 sub_lang = 'en'
451 else:
452 sub_lang = list(sub_lang_list.keys())[0]
453 if not sub_lang in sub_lang_list:
454 return [(u'no closed captions found in the specified language "%s"' % sub_lang, None, None)]
455
456 subtitle = self._request_subtitle(sub_lang, sub_lang_list[sub_lang].encode('utf-8'), video_id, sub_format)
457 return [subtitle]
458
459 def _extract_all_subtitles(self, video_id):
460 sub_lang_list = self._get_available_subtitles(video_id)
461 sub_format = self._downloader.params.get('subtitlesformat')
462 if isinstance(sub_lang_list,tuple): #There was some error, it didn't get the available subtitles
463 return [(sub_lang_list[0], None, None)]
464 subtitles = []
465 for sub_lang in sub_lang_list:
466 subtitle = self._request_subtitle(sub_lang, sub_lang_list[sub_lang].encode('utf-8'), video_id, sub_format)
467 subtitles.append(subtitle)
468 return subtitles
469
470 def _print_formats(self, formats):
471 print('Available formats:')
472 for x in formats:
473 print('%s\t:\t%s\t[%s]%s' %(x, self._video_extensions.get(x, 'flv'),
474 self._video_dimensions.get(x, '???'),
475 ' (3D)' if x in self._3d_itags else ''))
476
477 def _extract_id(self, url):
478 mobj = re.match(self._VALID_URL, url, re.VERBOSE)
479 if mobj is None:
480 raise ExtractorError(u'Invalid URL: %s' % url)
481 video_id = mobj.group(2)
482 return video_id
483
484 def _get_video_url_list(self, url_map):
485 """
486 Transform a dictionary in the format {itag:url} to a list of (itag, url)
487 with the requested formats.
488 """
489 req_format = self._downloader.params.get('format', None)
490 format_limit = self._downloader.params.get('format_limit', None)
491 available_formats = self._available_formats_prefer_free if self._downloader.params.get('prefer_free_formats', False) else self._available_formats
492 if format_limit is not None and format_limit in available_formats:
493 format_list = available_formats[available_formats.index(format_limit):]
494 else:
495 format_list = available_formats
496 existing_formats = [x for x in format_list if x in url_map]
497 if len(existing_formats) == 0:
498 raise ExtractorError(u'no known formats available for video')
499 if self._downloader.params.get('listformats', None):
500 self._print_formats(existing_formats)
501 return
502 if req_format is None or req_format == 'best':
503 video_url_list = [(existing_formats[0], url_map[existing_formats[0]])] # Best quality
504 elif req_format == 'worst':
505 video_url_list = [(existing_formats[-1], url_map[existing_formats[-1]])] # worst quality
506 elif req_format in ('-1', 'all'):
507 video_url_list = [(f, url_map[f]) for f in existing_formats] # All formats
508 else:
509 # Specific formats. We pick the first in a slash-delimeted sequence.
510 # For example, if '1/2/3/4' is requested and '2' and '4' are available, we pick '2'.
511 req_formats = req_format.split('/')
512 video_url_list = None
513 for rf in req_formats:
514 if rf in url_map:
515 video_url_list = [(rf, url_map[rf])]
516 break
517 if video_url_list is None:
518 raise ExtractorError(u'requested format not available')
519 return video_url_list
520
521 def _extract_from_m3u8(self, manifest_url, video_id):
522 url_map = {}
523 def _get_urls(_manifest):
524 lines = _manifest.split('\n')
525 urls = filter(lambda l: l and not l.startswith('#'),
526 lines)
527 return urls
528 manifest = self._download_webpage(manifest_url, video_id, u'Downloading formats manifest')
529 formats_urls = _get_urls(manifest)
530 for format_url in formats_urls:
531 itag = self._search_regex(r'itag/(\d+?)/', format_url, 'itag')
532 url_map[itag] = format_url
533 return url_map
534
535 def _real_extract(self, url):
536 if re.match(r'(?:https?://)?[^/]+/watch\?feature=[a-z_]+$', url):
537 self._downloader.report_warning(u'Did you forget to quote the URL? Remember that & is a meta-character in most shells, so you want to put the URL in quotes, like youtube-dl \'http://www.youtube.com/watch?feature=foo&v=BaW_jenozKc\' (or simply youtube-dl BaW_jenozKc ).')
538
539 # Extract original video URL from URL with redirection, like age verification, using next_url parameter
540 mobj = re.search(self._NEXT_URL_RE, url)
541 if mobj:
542 url = 'https://www.youtube.com/' + compat_urllib_parse.unquote(mobj.group(1)).lstrip('/')
543 video_id = self._extract_id(url)
544
545 # Get video webpage
546 self.report_video_webpage_download(video_id)
547 url = 'https://www.youtube.com/watch?v=%s&gl=US&hl=en&has_verified=1' % video_id
548 request = compat_urllib_request.Request(url)
549 try:
550 video_webpage_bytes = compat_urllib_request.urlopen(request).read()
551 except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
552 raise ExtractorError(u'Unable to download video webpage: %s' % compat_str(err))
553
554 video_webpage = video_webpage_bytes.decode('utf-8', 'ignore')
555
556 # Attempt to extract SWF player URL
557 mobj = re.search(r'swfConfig.*?"(http:\\/\\/.*?watch.*?-.*?\.swf)"', video_webpage)
558 if mobj is not None:
559 player_url = re.sub(r'\\(.)', r'\1', mobj.group(1))
560 else:
561 player_url = None
562
563 # Get video info
564 self.report_video_info_webpage_download(video_id)
565 if re.search(r'player-age-gate-content">', video_webpage) is not None:
566 self.report_age_confirmation()
567 age_gate = True
568 # We simulate the access to the video from www.youtube.com/v/{video_id}
569 # this can be viewed without login into Youtube
570 data = compat_urllib_parse.urlencode({'video_id': video_id,
571 'el': 'embedded',
572 'gl': 'US',
573 'hl': 'en',
574 'eurl': 'https://youtube.googleapis.com/v/' + video_id,
575 'asv': 3,
576 'sts':'1588',
577 })
578 video_info_url = 'https://www.youtube.com/get_video_info?' + data
579 video_info_webpage = self._download_webpage(video_info_url, video_id,
580 note=False,
581 errnote='unable to download video info webpage')
582 video_info = compat_parse_qs(video_info_webpage)
583 else:
584 age_gate = False
585 for el_type in ['&el=embedded', '&el=detailpage', '&el=vevo', '']:
586 video_info_url = ('https://www.youtube.com/get_video_info?&video_id=%s%s&ps=default&eurl=&gl=US&hl=en'
587 % (video_id, el_type))
588 video_info_webpage = self._download_webpage(video_info_url, video_id,
589 note=False,
590 errnote='unable to download video info webpage')
591 video_info = compat_parse_qs(video_info_webpage)
592 if 'token' in video_info:
593 break
594 if 'token' not in video_info:
595 if 'reason' in video_info:
596 raise ExtractorError(u'YouTube said: %s' % video_info['reason'][0], expected=True)
597 else:
598 raise ExtractorError(u'"token" parameter not in video info for unknown reason')
599
600 # Check for "rental" videos
601 if 'ypc_video_rental_bar_text' in video_info and 'author' not in video_info:
602 raise ExtractorError(u'"rental" videos not supported')
603
604 # Start extracting information
605 self.report_information_extraction(video_id)
606
607 # uploader
608 if 'author' not in video_info:
609 raise ExtractorError(u'Unable to extract uploader name')
610 video_uploader = compat_urllib_parse.unquote_plus(video_info['author'][0])
611
612 # uploader_id
613 video_uploader_id = None
614 mobj = re.search(r'<link itemprop="url" href="http://www.youtube.com/(?:user|channel)/([^"]+)">', video_webpage)
615 if mobj is not None:
616 video_uploader_id = mobj.group(1)
617 else:
618 self._downloader.report_warning(u'unable to extract uploader nickname')
619
620 # title
621 if 'title' not in video_info:
622 raise ExtractorError(u'Unable to extract video title')
623 video_title = compat_urllib_parse.unquote_plus(video_info['title'][0])
624
625 # thumbnail image
626 # We try first to get a high quality image:
627 m_thumb = re.search(r'<span itemprop="thumbnail".*?href="(.*?)">',
628 video_webpage, re.DOTALL)
629 if m_thumb is not None:
630 video_thumbnail = m_thumb.group(1)
631 elif 'thumbnail_url' not in video_info:
632 self._downloader.report_warning(u'unable to extract video thumbnail')
633 video_thumbnail = ''
634 else: # don't panic if we can't find it
635 video_thumbnail = compat_urllib_parse.unquote_plus(video_info['thumbnail_url'][0])
636
637 # upload date
638 upload_date = None
639 mobj = re.search(r'id="eow-date.*?>(.*?)</span>', video_webpage, re.DOTALL)
640 if mobj is not None:
641 upload_date = ' '.join(re.sub(r'[/,-]', r' ', mobj.group(1)).split())
642 upload_date = unified_strdate(upload_date)
643
644 # description
645 video_description = get_element_by_id("eow-description", video_webpage)
646 if video_description:
647 video_description = clean_html(video_description)
648 else:
649 fd_mobj = re.search(r'<meta name="description" content="([^"]+)"', video_webpage)
650 if fd_mobj:
651 video_description = unescapeHTML(fd_mobj.group(1))
652 else:
653 video_description = u''
654
655 # subtitles
656 video_subtitles = None
657
658 if self._downloader.params.get('writesubtitles', False):
659 video_subtitles = self._extract_subtitle(video_id)
660 if video_subtitles:
661 (sub_error, sub_lang, sub) = video_subtitles[0]
662 if sub_error:
663 self._downloader.report_warning(sub_error)
664
665 if self._downloader.params.get('writeautomaticsub', False):
666 video_subtitles = self._request_automatic_caption(video_id, video_webpage)
667 (sub_error, sub_lang, sub) = video_subtitles[0]
668 if sub_error:
669 self._downloader.report_warning(sub_error)
670
671 if self._downloader.params.get('allsubtitles', False):
672 video_subtitles = self._extract_all_subtitles(video_id)
673 for video_subtitle in video_subtitles:
674 (sub_error, sub_lang, sub) = video_subtitle
675 if sub_error:
676 self._downloader.report_warning(sub_error)
677
678 if self._downloader.params.get('listsubtitles', False):
679 self._list_available_subtitles(video_id)
680 return
681
682 if 'length_seconds' not in video_info:
683 self._downloader.report_warning(u'unable to extract video duration')
684 video_duration = ''
685 else:
686 video_duration = compat_urllib_parse.unquote_plus(video_info['length_seconds'][0])
687
688 # Decide which formats to download
689
690 try:
691 mobj = re.search(r';ytplayer.config = ({.*?});', video_webpage)
692 if not mobj:
693 raise ValueError('Could not find vevo ID')
694 info = json.loads(mobj.group(1))
695 args = info['args']
696 # Easy way to know if the 's' value is in url_encoded_fmt_stream_map
697 # this signatures are encrypted
698 m_s = re.search(r'[&,]s=', args['url_encoded_fmt_stream_map'])
699 if m_s is not None:
700 self.to_screen(u'%s: Encrypted signatures detected.' % video_id)
701 video_info['url_encoded_fmt_stream_map'] = [args['url_encoded_fmt_stream_map']]
702 except ValueError:
703 pass
704
705 if 'conn' in video_info and video_info['conn'][0].startswith('rtmp'):
706 self.report_rtmp_download()
707 video_url_list = [(None, video_info['conn'][0])]
708 elif 'url_encoded_fmt_stream_map' in video_info and len(video_info['url_encoded_fmt_stream_map']) >= 1:
709 if 'rtmpe%3Dyes' in video_info['url_encoded_fmt_stream_map'][0]:
710 raise ExtractorError('rtmpe downloads are not supported, see https://github.com/rg3/youtube-dl/issues/343 for more information.', expected=True)
711 url_map = {}
712 for url_data_str in video_info['url_encoded_fmt_stream_map'][0].split(','):
713 url_data = compat_parse_qs(url_data_str)
714 if 'itag' in url_data and 'url' in url_data:
715 url = url_data['url'][0]
716 if 'sig' in url_data:
717 url += '&signature=' + url_data['sig'][0]
718 elif 's' in url_data:
719 if self._downloader.params.get('verbose'):
720 s = url_data['s'][0]
721 if age_gate:
722 player_version = self._search_regex(r'ad3-(.+?)\.swf',
723 video_info['ad3_module'][0] if 'ad3_module' in video_info else 'NOT FOUND',
724 'flash player', fatal=False)
725 player = 'flash player %s' % player_version
726 else:
727 player = u'html5 player %s' % self._search_regex(r'html5player-(.+?)\.js', video_webpage,
728 'html5 player', fatal=False)
729 parts_sizes = u'.'.join(compat_str(len(part)) for part in s.split('.'))
730 self.to_screen(u'encrypted signature length %d (%s), itag %s, %s' %
731 (len(s), parts_sizes, url_data['itag'][0], player))
732 encrypted_sig = url_data['s'][0]
733 if age_gate:
734 signature = self._decrypt_signature_age_gate(encrypted_sig)
735 else:
736 signature = self._decrypt_signature(encrypted_sig)
737 url += '&signature=' + signature
738 if 'ratebypass' not in url:
739 url += '&ratebypass=yes'
740 url_map[url_data['itag'][0]] = url
741 video_url_list = self._get_video_url_list(url_map)
742 if not video_url_list:
743 return
744 elif video_info.get('hlsvp'):
745 manifest_url = video_info['hlsvp'][0]
746 url_map = self._extract_from_m3u8(manifest_url, video_id)
747 video_url_list = self._get_video_url_list(url_map)
748 if not video_url_list:
749 return
750
751 else:
752 raise ExtractorError(u'no conn or url_encoded_fmt_stream_map information found in video info')
753
754 results = []
755 for format_param, video_real_url in video_url_list:
756 # Extension
757 video_extension = self._video_extensions.get(format_param, 'flv')
758
759 video_format = '{0} - {1}{2}'.format(format_param if format_param else video_extension,
760 self._video_dimensions.get(format_param, '???'),
761 ' (3D)' if format_param in self._3d_itags else '')
762
763 results.append({
764 'id': video_id,
765 'url': video_real_url,
766 'uploader': video_uploader,
767 'uploader_id': video_uploader_id,
768 'upload_date': upload_date,
769 'title': video_title,
770 'ext': video_extension,
771 'format': video_format,
772 'thumbnail': video_thumbnail,
773 'description': video_description,
774 'player_url': player_url,
775 'subtitles': video_subtitles,
776 'duration': video_duration
777 })
778 return results
779
780 class YoutubePlaylistIE(InfoExtractor):
781 IE_DESC = u'YouTube.com playlists'
782 _VALID_URL = r"""(?:
783 (?:https?://)?
784 (?:\w+\.)?
785 youtube\.com/
786 (?:
787 (?:course|view_play_list|my_playlists|artist|playlist|watch)
788 \? (?:.*?&)*? (?:p|a|list)=
789 | p/
790 )
791 ((?:PL|EC|UU|FL)?[0-9A-Za-z-_]{10,})
792 .*
793 |
794 ((?:PL|EC|UU|FL)[0-9A-Za-z-_]{10,})
795 )"""
796 _TEMPLATE_URL = 'https://gdata.youtube.com/feeds/api/playlists/%s?max-results=%i&start-index=%i&v=2&alt=json&safeSearch=none'
797 _MAX_RESULTS = 50
798 IE_NAME = u'youtube:playlist'
799
800 @classmethod
801 def suitable(cls, url):
802 """Receives a URL and returns True if suitable for this IE."""
803 return re.match(cls._VALID_URL, url, re.VERBOSE) is not None
804
805 def _real_extract(self, url):
806 # Extract playlist id
807 mobj = re.match(self._VALID_URL, url, re.VERBOSE)
808 if mobj is None:
809 raise ExtractorError(u'Invalid URL: %s' % url)
810
811 # Download playlist videos from API
812 playlist_id = mobj.group(1) or mobj.group(2)
813 videos = []
814
815 for page_num in itertools.count(1):
816 start_index = self._MAX_RESULTS * (page_num - 1) + 1
817 if start_index >= 1000:
818 self._downloader.report_warning(u'Max number of results reached')
819 break
820 url = self._TEMPLATE_URL % (playlist_id, self._MAX_RESULTS, start_index)
821 page = self._download_webpage(url, playlist_id, u'Downloading page #%s' % page_num)
822
823 try:
824 response = json.loads(page)
825 except ValueError as err:
826 raise ExtractorError(u'Invalid JSON in API response: ' + compat_str(err))
827
828 if 'feed' not in response:
829 raise ExtractorError(u'Got a malformed response from YouTube API')
830 playlist_title = response['feed']['title']['$t']
831 if 'entry' not in response['feed']:
832 # Number of videos is a multiple of self._MAX_RESULTS
833 break
834
835 for entry in response['feed']['entry']:
836 index = entry['yt$position']['$t']
837 if 'media$group' in entry and 'media$player' in entry['media$group']:
838 videos.append((index, entry['media$group']['media$player']['url']))
839
840 videos = [v[1] for v in sorted(videos)]
841
842 url_results = [self.url_result(vurl, 'Youtube') for vurl in videos]
843 return [self.playlist_result(url_results, playlist_id, playlist_title)]
844
845
846 class YoutubeChannelIE(InfoExtractor):
847 IE_DESC = u'YouTube.com channels'
848 _VALID_URL = r"^(?:https?://)?(?:youtu\.be|(?:\w+\.)?youtube(?:-nocookie)?\.com)/channel/([0-9A-Za-z_-]+)"
849 _TEMPLATE_URL = 'http://www.youtube.com/channel/%s/videos?sort=da&flow=list&view=0&page=%s&gl=US&hl=en'
850 _MORE_PAGES_INDICATOR = 'yt-uix-load-more'
851 _MORE_PAGES_URL = 'http://www.youtube.com/c4_browse_ajax?action_load_more_videos=1&flow=list&paging=%s&view=0&sort=da&channel_id=%s'
852 IE_NAME = u'youtube:channel'
853
854 def extract_videos_from_page(self, page):
855 ids_in_page = []
856 for mobj in re.finditer(r'href="/watch\?v=([0-9A-Za-z_-]+)&?', page):
857 if mobj.group(1) not in ids_in_page:
858 ids_in_page.append(mobj.group(1))
859 return ids_in_page
860
861 def _real_extract(self, url):
862 # Extract channel id
863 mobj = re.match(self._VALID_URL, url)
864 if mobj is None:
865 raise ExtractorError(u'Invalid URL: %s' % url)
866
867 # Download channel page
868 channel_id = mobj.group(1)
869 video_ids = []
870 pagenum = 1
871
872 url = self._TEMPLATE_URL % (channel_id, pagenum)
873 page = self._download_webpage(url, channel_id,
874 u'Downloading page #%s' % pagenum)
875
876 # Extract video identifiers
877 ids_in_page = self.extract_videos_from_page(page)
878 video_ids.extend(ids_in_page)
879
880 # Download any subsequent channel pages using the json-based channel_ajax query
881 if self._MORE_PAGES_INDICATOR in page:
882 for pagenum in itertools.count(1):
883 url = self._MORE_PAGES_URL % (pagenum, channel_id)
884 page = self._download_webpage(url, channel_id,
885 u'Downloading page #%s' % pagenum)
886
887 page = json.loads(page)
888
889 ids_in_page = self.extract_videos_from_page(page['content_html'])
890 video_ids.extend(ids_in_page)
891
892 if self._MORE_PAGES_INDICATOR not in page['load_more_widget_html']:
893 break
894
895 self._downloader.to_screen(u'[youtube] Channel %s: Found %i videos' % (channel_id, len(video_ids)))
896
897 urls = ['http://www.youtube.com/watch?v=%s' % id for id in video_ids]
898 url_entries = [self.url_result(eurl, 'Youtube') for eurl in urls]
899 return [self.playlist_result(url_entries, channel_id)]
900
901
902 class YoutubeUserIE(InfoExtractor):
903 IE_DESC = u'YouTube.com user videos (URL or "ytuser" keyword)'
904 _VALID_URL = r'(?:(?:(?:https?://)?(?:\w+\.)?youtube\.com/user/)|ytuser:)([A-Za-z0-9_-]+)'
905 _TEMPLATE_URL = 'http://gdata.youtube.com/feeds/api/users/%s'
906 _GDATA_PAGE_SIZE = 50
907 _GDATA_URL = 'http://gdata.youtube.com/feeds/api/users/%s/uploads?max-results=%d&start-index=%d'
908 _VIDEO_INDICATOR = r'/watch\?v=(.+?)[\<&]'
909 IE_NAME = u'youtube:user'
910
911 def _real_extract(self, url):
912 # Extract username
913 mobj = re.match(self._VALID_URL, url)
914 if mobj is None:
915 raise ExtractorError(u'Invalid URL: %s' % url)
916
917 username = mobj.group(1)
918
919 # Download video ids using YouTube Data API. Result size per
920 # query is limited (currently to 50 videos) so we need to query
921 # page by page until there are no video ids - it means we got
922 # all of them.
923
924 video_ids = []
925
926 for pagenum in itertools.count(0):
927 start_index = pagenum * self._GDATA_PAGE_SIZE + 1
928
929 gdata_url = self._GDATA_URL % (username, self._GDATA_PAGE_SIZE, start_index)
930 page = self._download_webpage(gdata_url, username,
931 u'Downloading video ids from %d to %d' % (start_index, start_index + self._GDATA_PAGE_SIZE))
932
933 # Extract video identifiers
934 ids_in_page = []
935
936 for mobj in re.finditer(self._VIDEO_INDICATOR, page):
937 if mobj.group(1) not in ids_in_page:
938 ids_in_page.append(mobj.group(1))
939
940 video_ids.extend(ids_in_page)
941
942 # A little optimization - if current page is not
943 # "full", ie. does not contain PAGE_SIZE video ids then
944 # we can assume that this page is the last one - there
945 # are no more ids on further pages - no need to query
946 # again.
947
948 if len(ids_in_page) < self._GDATA_PAGE_SIZE:
949 break
950
951 urls = ['http://www.youtube.com/watch?v=%s' % video_id for video_id in video_ids]
952 url_results = [self.url_result(rurl, 'Youtube') for rurl in urls]
953 return [self.playlist_result(url_results, playlist_title = username)]
954
955 class YoutubeSearchIE(SearchInfoExtractor):
956 IE_DESC = u'YouTube.com searches'
957 _API_URL = 'https://gdata.youtube.com/feeds/api/videos?q=%s&start-index=%i&max-results=50&v=2&alt=jsonc'
958 _MAX_RESULTS = 1000
959 IE_NAME = u'youtube:search'
960 _SEARCH_KEY = 'ytsearch'
961
962 def report_download_page(self, query, pagenum):
963 """Report attempt to download search page with given number."""
964 self._downloader.to_screen(u'[youtube] query "%s": Downloading page %s' % (query, pagenum))
965
966 def _get_n_results(self, query, n):
967 """Get a specified number of results for a query"""
968
969 video_ids = []
970 pagenum = 0
971 limit = n
972
973 while (50 * pagenum) < limit:
974 self.report_download_page(query, pagenum+1)
975 result_url = self._API_URL % (compat_urllib_parse.quote_plus(query), (50*pagenum)+1)
976 request = compat_urllib_request.Request(result_url)
977 try:
978 data = compat_urllib_request.urlopen(request).read().decode('utf-8')
979 except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
980 raise ExtractorError(u'Unable to download API page: %s' % compat_str(err))
981 api_response = json.loads(data)['data']
982
983 if not 'items' in api_response:
984 raise ExtractorError(u'[youtube] No video results')
985
986 new_ids = list(video['id'] for video in api_response['items'])
987 video_ids += new_ids
988
989 limit = min(n, api_response['totalItems'])
990 pagenum += 1
991
992 if len(video_ids) > n:
993 video_ids = video_ids[:n]
994 videos = [self.url_result('http://www.youtube.com/watch?v=%s' % id, 'Youtube') for id in video_ids]
995 return self.playlist_result(videos, query)
996
997
998 class YoutubeShowIE(InfoExtractor):
999 IE_DESC = u'YouTube.com (multi-season) shows'
1000 _VALID_URL = r'https?://www\.youtube\.com/show/(.*)'
1001 IE_NAME = u'youtube:show'
1002
1003 def _real_extract(self, url):
1004 mobj = re.match(self._VALID_URL, url)
1005 show_name = mobj.group(1)
1006 webpage = self._download_webpage(url, show_name, u'Downloading show webpage')
1007 # There's one playlist for each season of the show
1008 m_seasons = list(re.finditer(r'href="(/playlist\?list=.*?)"', webpage))
1009 self.to_screen(u'%s: Found %s seasons' % (show_name, len(m_seasons)))
1010 return [self.url_result('https://www.youtube.com' + season.group(1), 'YoutubePlaylist') for season in m_seasons]
1011
1012
1013 class YoutubeFeedsInfoExtractor(YoutubeBaseInfoExtractor):
1014 """
1015 Base class for extractors that fetch info from
1016 http://www.youtube.com/feed_ajax
1017 Subclasses must define the _FEED_NAME and _PLAYLIST_TITLE properties.
1018 """
1019 _LOGIN_REQUIRED = True
1020 _PAGING_STEP = 30
1021 # use action_load_personal_feed instead of action_load_system_feed
1022 _PERSONAL_FEED = False
1023
1024 @property
1025 def _FEED_TEMPLATE(self):
1026 action = 'action_load_system_feed'
1027 if self._PERSONAL_FEED:
1028 action = 'action_load_personal_feed'
1029 return 'http://www.youtube.com/feed_ajax?%s=1&feed_name=%s&paging=%%s' % (action, self._FEED_NAME)
1030
1031 @property
1032 def IE_NAME(self):
1033 return u'youtube:%s' % self._FEED_NAME
1034
1035 def _real_initialize(self):
1036 self._login()
1037
1038 def _real_extract(self, url):
1039 feed_entries = []
1040 # The step argument is available only in 2.7 or higher
1041 for i in itertools.count(0):
1042 paging = i*self._PAGING_STEP
1043 info = self._download_webpage(self._FEED_TEMPLATE % paging,
1044 u'%s feed' % self._FEED_NAME,
1045 u'Downloading page %s' % i)
1046 info = json.loads(info)
1047 feed_html = info['feed_html']
1048 m_ids = re.finditer(r'"/watch\?v=(.*?)["&]', feed_html)
1049 ids = orderedSet(m.group(1) for m in m_ids)
1050 feed_entries.extend(self.url_result(id, 'Youtube') for id in ids)
1051 if info['paging'] is None:
1052 break
1053 return self.playlist_result(feed_entries, playlist_title=self._PLAYLIST_TITLE)
1054
1055 class YoutubeSubscriptionsIE(YoutubeFeedsInfoExtractor):
1056 IE_DESC = u'YouTube.com subscriptions feed, "ytsubs" keyword(requires authentication)'
1057 _VALID_URL = r'https?://www\.youtube\.com/feed/subscriptions|:ytsubs(?:criptions)?'
1058 _FEED_NAME = 'subscriptions'
1059 _PLAYLIST_TITLE = u'Youtube Subscriptions'
1060
1061 class YoutubeRecommendedIE(YoutubeFeedsInfoExtractor):
1062 IE_DESC = u'YouTube.com recommended videos, "ytrec" keyword (requires authentication)'
1063 _VALID_URL = r'https?://www\.youtube\.com/feed/recommended|:ytrec(?:ommended)?'
1064 _FEED_NAME = 'recommended'
1065 _PLAYLIST_TITLE = u'Youtube Recommended videos'
1066
1067 class YoutubeWatchLaterIE(YoutubeFeedsInfoExtractor):
1068 IE_DESC = u'Youtube watch later list, "ytwatchlater" keyword (requires authentication)'
1069 _VALID_URL = r'https?://www\.youtube\.com/feed/watch_later|:ytwatchlater'
1070 _FEED_NAME = 'watch_later'
1071 _PLAYLIST_TITLE = u'Youtube Watch Later'
1072 _PAGING_STEP = 100
1073 _PERSONAL_FEED = True
1074
1075 class YoutubeFavouritesIE(YoutubeBaseInfoExtractor):
1076 IE_NAME = u'youtube:favorites'
1077 IE_DESC = u'YouTube.com favourite videos, "ytfav" keyword (requires authentication)'
1078 _VALID_URL = r'https?://www\.youtube\.com/my_favorites|:ytfav(?:o?rites)?'
1079 _LOGIN_REQUIRED = True
1080
1081 def _real_extract(self, url):
1082 webpage = self._download_webpage('https://www.youtube.com/my_favorites', 'Youtube Favourites videos')
1083 playlist_id = self._search_regex(r'list=(.+?)["&]', webpage, u'favourites playlist id')
1084 return self.url_result(playlist_id, 'YoutubePlaylist')