]> jfr.im git - yt-dlp.git/blame - youtube_dl/extractor/youtube.py
[youtube] add an extractor for the "Watch Later" list
[yt-dlp.git] / youtube_dl / extractor / youtube.py
CommitLineData
c5e8d7af 1# coding: utf-8
c5e8d7af
PH
2
3import json
4import netrc
5import re
6import socket
04cc9617 7import itertools
c5e8d7af 8
b05654f0 9from .common import InfoExtractor, SearchInfoExtractor
c5e8d7af
PH
10from ..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,
04cc9617 23 orderedSet,
c5e8d7af
PH
24)
25
b2e8bc1b
JMF
26class 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)
c5e8d7af 72
b2e8bc1b
JMF
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
133class YoutubeIE(YoutubeBaseInfoExtractor):
0f818663 134 IE_DESC = u'YouTube.com'
c5e8d7af
PH
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
bcd6e4bd 144 (?:watch|movie(?:_popup)?(?:\.php)?)? # preceding watch(_popup|.php) or nothing (like /?v=xxxx)
c5e8d7af
PH
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 $"""
c5e8d7af 154 _NEXT_URL_RE = r'[\?&]next_url=([^&]+)'
c5e8d7af
PH
155 # Listed in order of quality
156 _available_formats = ['38', '37', '46', '22', '45', '35', '44', '34', '18', '43', '6', '5', '17', '13']
157 _available_formats_prefer_free = ['38', '46', '37', '45', '22', '44', '35', '43', '34', '18', '6', '5', '17', '13']
158 _video_extensions = {
159 '13': '3gp',
160 '17': 'mp4',
161 '18': 'mp4',
162 '22': 'mp4',
163 '37': 'mp4',
d69cf69a 164 '38': 'mp4',
c5e8d7af
PH
165 '43': 'webm',
166 '44': 'webm',
167 '45': 'webm',
168 '46': 'webm',
169 }
170 _video_dimensions = {
171 '5': '240x400',
172 '6': '???',
173 '13': '???',
174 '17': '144x176',
175 '18': '360x640',
176 '22': '720x1280',
177 '34': '360x640',
178 '35': '480x854',
179 '37': '1080x1920',
180 '38': '3072x4096',
181 '43': '360x640',
182 '44': '480x854',
183 '45': '720x1280',
184 '46': '1080x1920',
185 }
186 IE_NAME = u'youtube'
2eb88d95
PH
187 _TESTS = [
188 {
0e853ca4
PH
189 u"url": u"http://www.youtube.com/watch?v=BaW_jenozKc",
190 u"file": u"BaW_jenozKc.mp4",
191 u"info_dict": {
192 u"title": u"youtube-dl test video \"'/\\ä↭𝕐",
193 u"uploader": u"Philipp Hagemeister",
194 u"uploader_id": u"phihag",
195 u"upload_date": u"20121002",
196 u"description": u"test chars: \"'/\\ä↭𝕐\n\nThis is a test video for youtube-dl.\n\nFor more information, contact phihag@phihag.de ."
2eb88d95 197 }
0e853ca4
PH
198 },
199 {
200 u"url": u"http://www.youtube.com/watch?v=1ltcDfZMA3U",
201 u"file": u"1ltcDfZMA3U.flv",
202 u"note": u"Test VEVO video (#897)",
203 u"info_dict": {
204 u"upload_date": u"20070518",
205 u"title": u"Maps - It Will Find You",
206 u"description": u"Music video by Maps performing It Will Find You.",
207 u"uploader": u"MuteUSA",
208 u"uploader_id": u"MuteUSA"
2eb88d95 209 }
0e853ca4
PH
210 },
211 {
212 u"url": u"http://www.youtube.com/watch?v=UxxajLWwzqY",
213 u"file": u"UxxajLWwzqY.mp4",
214 u"note": u"Test generic use_cipher_signature video (#897)",
215 u"info_dict": {
216 u"upload_date": u"20120506",
217 u"title": u"Icona Pop - I Love It (feat. Charli XCX) [OFFICIAL VIDEO]",
218 u"description": u"md5:b085c9804f5ab69f4adea963a2dceb3c",
219 u"uploader": u"IconaPop",
220 u"uploader_id": u"IconaPop"
2eb88d95 221 }
c108eb73
JMF
222 },
223 {
224 u"url": u"https://www.youtube.com/watch?v=07FYdnEawAQ",
225 u"file": u"07FYdnEawAQ.mp4",
226 u"note": u"Test VEVO video with age protection (#956)",
227 u"info_dict": {
228 u"upload_date": u"20130703",
229 u"title": u"Justin Timberlake - Tunnel Vision (Explicit)",
230 u"description": u"md5:64249768eec3bc4276236606ea996373",
231 u"uploader": u"justintimberlakeVEVO",
232 u"uploader_id": u"justintimberlakeVEVO"
233 }
234 },
2eb88d95
PH
235 ]
236
c5e8d7af
PH
237
238 @classmethod
239 def suitable(cls, url):
240 """Receives a URL and returns True if suitable for this IE."""
04cc9617 241 if YoutubePlaylistIE.suitable(url) or YoutubeSubscriptionsIE.suitable(url): return False
c5e8d7af
PH
242 return re.match(cls._VALID_URL, url, re.VERBOSE) is not None
243
c5e8d7af
PH
244 def report_video_webpage_download(self, video_id):
245 """Report attempt to download video webpage."""
246 self.to_screen(u'%s: Downloading video webpage' % video_id)
247
248 def report_video_info_webpage_download(self, video_id):
249 """Report attempt to download video info webpage."""
250 self.to_screen(u'%s: Downloading video info webpage' % video_id)
251
252 def report_video_subtitles_download(self, video_id):
253 """Report attempt to download video info webpage."""
254 self.to_screen(u'%s: Checking available subtitles' % video_id)
255
256 def report_video_subtitles_request(self, video_id, sub_lang, format):
257 """Report attempt to download video info webpage."""
258 self.to_screen(u'%s: Downloading video subtitles for %s.%s' % (video_id, sub_lang, format))
259
260 def report_video_subtitles_available(self, video_id, sub_lang_list):
261 """Report available subtitles."""
262 sub_lang = ",".join(list(sub_lang_list.keys()))
263 self.to_screen(u'%s: Available subtitles for video: %s' % (video_id, sub_lang))
264
265 def report_information_extraction(self, video_id):
266 """Report attempt to extract video information."""
267 self.to_screen(u'%s: Extracting video information' % video_id)
268
269 def report_unavailable_format(self, video_id, format):
270 """Report extracted video URL."""
271 self.to_screen(u'%s: Format %s not available' % (video_id, format))
272
273 def report_rtmp_download(self):
274 """Indicate the download will use the RTMP protocol."""
275 self.to_screen(u'RTMP download detected')
276
98bcd283 277 def _decrypt_signature(self, s):
257a2501 278 """Turn the encrypted s field into a working signature"""
6b37f0be 279
444b1165
JMF
280 if len(s) == 92:
281 return s[25] + s[3:25] + s[0] + s[26:42] + s[79] + s[43:79] + s[91] + s[80:83]
282 elif len(s) == 90:
283 return s[25] + s[3:25] + s[2] + s[26:40] + s[77] + s[41:77] + s[89] + s[78:81]
284 elif len(s) == 88:
be547e1d
PH
285 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]
286 elif len(s) == 87:
239e3e0c 287 return s[4:23] + s[86] + s[24:85]
be547e1d
PH
288 elif len(s) == 86:
289 return s[2:63] + s[82] + s[64:82] + s[63]
290 elif len(s) == 85:
a4358cba 291 return s[2:8] + s[0] + s[9:21] + s[65] + s[22:65] + s[84] + s[66:82] + s[21]
be547e1d
PH
292 elif len(s) == 84:
293 return s[83:36:-1] + s[2] + s[35:26:-1] + s[3] + s[25:3:-1] + s[26]
294 elif len(s) == 83:
295 return s[6] + s[3:6] + s[33] + s[7:24] + s[0] + s[25:33] + s[53] + s[34:53] + s[24] + s[54:]
296 elif len(s) == 82:
297 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]
298 elif len(s) == 81:
299 return s[6] + s[3:6] + s[33] + s[7:24] + s[0] + s[25:33] + s[2] + s[34:53] + s[24] + s[54:81]
300
301 else:
302 raise ExtractorError(u'Unable to decrypt signature, key length %d not supported; retrying might work' % (len(s)))
c5e8d7af
PH
303
304 def _get_available_subtitles(self, video_id):
305 self.report_video_subtitles_download(video_id)
306 request = compat_urllib_request.Request('http://video.google.com/timedtext?hl=en&type=list&v=%s' % video_id)
307 try:
308 sub_list = compat_urllib_request.urlopen(request).read().decode('utf-8')
309 except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
310 return (u'unable to download video subtitles: %s' % compat_str(err), None)
311 sub_lang_list = re.findall(r'name="([^"]*)"[^>]+lang_code="([\w\-]+)"', sub_list)
312 sub_lang_list = dict((l[1], l[0]) for l in sub_lang_list)
313 if not sub_lang_list:
314 return (u'video doesn\'t have subtitles', None)
315 return sub_lang_list
316
317 def _list_available_subtitles(self, video_id):
318 sub_lang_list = self._get_available_subtitles(video_id)
319 self.report_video_subtitles_available(video_id, sub_lang_list)
320
321 def _request_subtitle(self, sub_lang, sub_name, video_id, format):
322 """
323 Return tuple:
324 (error_message, sub_lang, sub)
325 """
326 self.report_video_subtitles_request(video_id, sub_lang, format)
327 params = compat_urllib_parse.urlencode({
328 'lang': sub_lang,
329 'name': sub_name,
330 'v': video_id,
331 'fmt': format,
332 })
333 url = 'http://www.youtube.com/api/timedtext?' + params
334 try:
335 sub = compat_urllib_request.urlopen(url).read().decode('utf-8')
336 except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
337 return (u'unable to download video subtitles: %s' % compat_str(err), None, None)
338 if not sub:
339 return (u'Did not fetch video subtitles', None, None)
340 return (None, sub_lang, sub)
341
342 def _request_automatic_caption(self, video_id, webpage):
343 """We need the webpage for getting the captions url, pass it as an
344 argument to speed up the process."""
345 sub_lang = self._downloader.params.get('subtitleslang') or 'en'
346 sub_format = self._downloader.params.get('subtitlesformat')
347 self.to_screen(u'%s: Looking for automatic captions' % video_id)
348 mobj = re.search(r';ytplayer.config = ({.*?});', webpage)
349 err_msg = u'Couldn\'t find automatic captions for "%s"' % sub_lang
350 if mobj is None:
351 return [(err_msg, None, None)]
352 player_config = json.loads(mobj.group(1))
353 try:
354 args = player_config[u'args']
355 caption_url = args[u'ttsurl']
356 timestamp = args[u'timestamp']
357 params = compat_urllib_parse.urlencode({
358 'lang': 'en',
359 'tlang': sub_lang,
360 'fmt': sub_format,
361 'ts': timestamp,
362 'kind': 'asr',
363 })
364 subtitles_url = caption_url + '&' + params
365 sub = self._download_webpage(subtitles_url, video_id, u'Downloading automatic captions')
366 return [(None, sub_lang, sub)]
367 except KeyError:
368 return [(err_msg, None, None)]
369
370 def _extract_subtitle(self, video_id):
371 """
372 Return a list with a tuple:
373 [(error_message, sub_lang, sub)]
374 """
375 sub_lang_list = self._get_available_subtitles(video_id)
376 sub_format = self._downloader.params.get('subtitlesformat')
377 if isinstance(sub_lang_list,tuple): #There was some error, it didn't get the available subtitles
378 return [(sub_lang_list[0], None, None)]
379 if self._downloader.params.get('subtitleslang', False):
380 sub_lang = self._downloader.params.get('subtitleslang')
381 elif 'en' in sub_lang_list:
382 sub_lang = 'en'
383 else:
384 sub_lang = list(sub_lang_list.keys())[0]
385 if not sub_lang in sub_lang_list:
386 return [(u'no closed captions found in the specified language "%s"' % sub_lang, None, None)]
387
388 subtitle = self._request_subtitle(sub_lang, sub_lang_list[sub_lang].encode('utf-8'), video_id, sub_format)
389 return [subtitle]
390
391 def _extract_all_subtitles(self, video_id):
392 sub_lang_list = self._get_available_subtitles(video_id)
393 sub_format = self._downloader.params.get('subtitlesformat')
394 if isinstance(sub_lang_list,tuple): #There was some error, it didn't get the available subtitles
395 return [(sub_lang_list[0], None, None)]
396 subtitles = []
397 for sub_lang in sub_lang_list:
398 subtitle = self._request_subtitle(sub_lang, sub_lang_list[sub_lang].encode('utf-8'), video_id, sub_format)
399 subtitles.append(subtitle)
400 return subtitles
401
402 def _print_formats(self, formats):
403 print('Available formats:')
404 for x in formats:
405 print('%s\t:\t%s\t[%s]' %(x, self._video_extensions.get(x, 'flv'), self._video_dimensions.get(x, '???')))
406
c5e8d7af
PH
407 def _extract_id(self, url):
408 mobj = re.match(self._VALID_URL, url, re.VERBOSE)
409 if mobj is None:
410 raise ExtractorError(u'Invalid URL: %s' % url)
411 video_id = mobj.group(2)
412 return video_id
413
414 def _real_extract(self, url):
d7f44b5b
PH
415 if re.match(r'(?:https?://)?[^/]+/watch\?feature=[a-z_]+$', url):
416 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 ).')
417
c5e8d7af
PH
418 # Extract original video URL from URL with redirection, like age verification, using next_url parameter
419 mobj = re.search(self._NEXT_URL_RE, url)
420 if mobj:
421 url = 'https://www.youtube.com/' + compat_urllib_parse.unquote(mobj.group(1)).lstrip('/')
422 video_id = self._extract_id(url)
423
424 # Get video webpage
425 self.report_video_webpage_download(video_id)
426 url = 'https://www.youtube.com/watch?v=%s&gl=US&hl=en&has_verified=1' % video_id
427 request = compat_urllib_request.Request(url)
428 try:
429 video_webpage_bytes = compat_urllib_request.urlopen(request).read()
430 except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
431 raise ExtractorError(u'Unable to download video webpage: %s' % compat_str(err))
432
433 video_webpage = video_webpage_bytes.decode('utf-8', 'ignore')
434
435 # Attempt to extract SWF player URL
436 mobj = re.search(r'swfConfig.*?"(http:\\/\\/.*?watch.*?-.*?\.swf)"', video_webpage)
437 if mobj is not None:
438 player_url = re.sub(r'\\(.)', r'\1', mobj.group(1))
439 else:
440 player_url = None
441
442 # Get video info
443 self.report_video_info_webpage_download(video_id)
c108eb73
JMF
444 if re.search(r'player-age-gate-content">', video_webpage) is not None:
445 self.report_age_confirmation()
446 age_gate = True
447 # We simulate the access to the video from www.youtube.com/v/{video_id}
448 # this can be viewed without login into Youtube
449 data = compat_urllib_parse.urlencode({'video_id': video_id,
450 'el': 'embedded',
451 'gl': 'US',
452 'hl': 'en',
453 'eurl': 'https://youtube.googleapis.com/v/' + video_id,
454 'asv': 3,
455 'sts':'1588',
456 })
457 video_info_url = 'https://www.youtube.com/get_video_info?' + data
c5e8d7af
PH
458 video_info_webpage = self._download_webpage(video_info_url, video_id,
459 note=False,
460 errnote='unable to download video info webpage')
461 video_info = compat_parse_qs(video_info_webpage)
c108eb73
JMF
462 else:
463 age_gate = False
464 for el_type in ['&el=embedded', '&el=detailpage', '&el=vevo', '']:
465 video_info_url = ('https://www.youtube.com/get_video_info?&video_id=%s%s&ps=default&eurl=&gl=US&hl=en'
466 % (video_id, el_type))
467 video_info_webpage = self._download_webpage(video_info_url, video_id,
468 note=False,
469 errnote='unable to download video info webpage')
470 video_info = compat_parse_qs(video_info_webpage)
471 if 'token' in video_info:
472 break
c5e8d7af
PH
473 if 'token' not in video_info:
474 if 'reason' in video_info:
9a82b238 475 raise ExtractorError(u'YouTube said: %s' % video_info['reason'][0], expected=True)
c5e8d7af
PH
476 else:
477 raise ExtractorError(u'"token" parameter not in video info for unknown reason')
478
479 # Check for "rental" videos
480 if 'ypc_video_rental_bar_text' in video_info and 'author' not in video_info:
481 raise ExtractorError(u'"rental" videos not supported')
482
483 # Start extracting information
484 self.report_information_extraction(video_id)
485
486 # uploader
487 if 'author' not in video_info:
488 raise ExtractorError(u'Unable to extract uploader name')
489 video_uploader = compat_urllib_parse.unquote_plus(video_info['author'][0])
490
491 # uploader_id
492 video_uploader_id = None
493 mobj = re.search(r'<link itemprop="url" href="http://www.youtube.com/(?:user|channel)/([^"]+)">', video_webpage)
494 if mobj is not None:
495 video_uploader_id = mobj.group(1)
496 else:
497 self._downloader.report_warning(u'unable to extract uploader nickname')
498
499 # title
500 if 'title' not in video_info:
501 raise ExtractorError(u'Unable to extract video title')
502 video_title = compat_urllib_parse.unquote_plus(video_info['title'][0])
503
504 # thumbnail image
7763b04e
JMF
505 # We try first to get a high quality image:
506 m_thumb = re.search(r'<span itemprop="thumbnail".*?href="(.*?)">',
507 video_webpage, re.DOTALL)
508 if m_thumb is not None:
509 video_thumbnail = m_thumb.group(1)
510 elif 'thumbnail_url' not in video_info:
c5e8d7af
PH
511 self._downloader.report_warning(u'unable to extract video thumbnail')
512 video_thumbnail = ''
513 else: # don't panic if we can't find it
514 video_thumbnail = compat_urllib_parse.unquote_plus(video_info['thumbnail_url'][0])
515
516 # upload date
517 upload_date = None
518 mobj = re.search(r'id="eow-date.*?>(.*?)</span>', video_webpage, re.DOTALL)
519 if mobj is not None:
520 upload_date = ' '.join(re.sub(r'[/,-]', r' ', mobj.group(1)).split())
521 upload_date = unified_strdate(upload_date)
522
523 # description
524 video_description = get_element_by_id("eow-description", video_webpage)
525 if video_description:
526 video_description = clean_html(video_description)
527 else:
528 fd_mobj = re.search(r'<meta name="description" content="([^"]+)"', video_webpage)
529 if fd_mobj:
530 video_description = unescapeHTML(fd_mobj.group(1))
531 else:
532 video_description = u''
533
534 # subtitles
535 video_subtitles = None
536
537 if self._downloader.params.get('writesubtitles', False):
538 video_subtitles = self._extract_subtitle(video_id)
539 if video_subtitles:
540 (sub_error, sub_lang, sub) = video_subtitles[0]
541 if sub_error:
b004821f
JMF
542 self._downloader.report_warning(sub_error)
543
544 if self._downloader.params.get('writeautomaticsub', False):
545 video_subtitles = self._request_automatic_caption(video_id, video_webpage)
546 (sub_error, sub_lang, sub) = video_subtitles[0]
547 if sub_error:
548 self._downloader.report_warning(sub_error)
c5e8d7af
PH
549
550 if self._downloader.params.get('allsubtitles', False):
551 video_subtitles = self._extract_all_subtitles(video_id)
552 for video_subtitle in video_subtitles:
553 (sub_error, sub_lang, sub) = video_subtitle
554 if sub_error:
555 self._downloader.report_warning(sub_error)
556
557 if self._downloader.params.get('listsubtitles', False):
93d3a642 558 self._list_available_subtitles(video_id)
c5e8d7af
PH
559 return
560
561 if 'length_seconds' not in video_info:
562 self._downloader.report_warning(u'unable to extract video duration')
563 video_duration = ''
564 else:
565 video_duration = compat_urllib_parse.unquote_plus(video_info['length_seconds'][0])
566
c5e8d7af
PH
567 # Decide which formats to download
568 req_format = self._downloader.params.get('format', None)
569
570 try:
571 mobj = re.search(r';ytplayer.config = ({.*?});', video_webpage)
50be92c1
PH
572 if not mobj:
573 raise ValueError('Could not find vevo ID')
c5e8d7af
PH
574 info = json.loads(mobj.group(1))
575 args = info['args']
7ce7e394
JMF
576 # Easy way to know if the 's' value is in url_encoded_fmt_stream_map
577 # this signatures are encrypted
578 m_s = re.search(r'[&,]s=', args['url_encoded_fmt_stream_map'])
579 if m_s is not None:
580 self.to_screen(u'%s: Encrypted signatures detected.' % video_id)
c5e8d7af
PH
581 video_info['url_encoded_fmt_stream_map'] = [args['url_encoded_fmt_stream_map']]
582 except ValueError:
583 pass
584
585 if 'conn' in video_info and video_info['conn'][0].startswith('rtmp'):
586 self.report_rtmp_download()
587 video_url_list = [(None, video_info['conn'][0])]
588 elif 'url_encoded_fmt_stream_map' in video_info and len(video_info['url_encoded_fmt_stream_map']) >= 1:
a7055eb9
JMF
589 if 'rtmpe%3Dyes' in video_info['url_encoded_fmt_stream_map'][0]:
590 raise ExtractorError('rtmpe downloads are not supported, see https://github.com/rg3/youtube-dl/issues/343 for more information.', expected=True)
c5e8d7af
PH
591 url_map = {}
592 for url_data_str in video_info['url_encoded_fmt_stream_map'][0].split(','):
593 url_data = compat_parse_qs(url_data_str)
594 if 'itag' in url_data and 'url' in url_data:
595 url = url_data['url'][0]
596 if 'sig' in url_data:
597 url += '&signature=' + url_data['sig'][0]
598 elif 's' in url_data:
769fda3c
FV
599 if self._downloader.params.get('verbose'):
600 s = url_data['s'][0]
c108eb73
JMF
601 if age_gate:
602 player_version = self._search_regex(r'ad3-(.+?)\.swf',
603 video_info['ad3_module'][0], 'flash player',
604 fatal=False)
605 player = 'flash player %s' % player_version
606 else:
607 player = u'html5 player %s' % self._search_regex(r'html5player-(.+?)\.js', video_webpage,
608 'html5 player', fatal=False)
5a76c651
JMF
609 parts_sizes = u'.'.join(compat_str(len(part)) for part in s.split('.'))
610 self.to_screen(u'encrypted signature length %d (%s), itag %s, %s' %
611 (len(s), parts_sizes, url_data['itag'][0], player))
c5e8d7af
PH
612 signature = self._decrypt_signature(url_data['s'][0])
613 url += '&signature=' + signature
614 if 'ratebypass' not in url:
615 url += '&ratebypass=yes'
616 url_map[url_data['itag'][0]] = url
617
618 format_limit = self._downloader.params.get('format_limit', None)
619 available_formats = self._available_formats_prefer_free if self._downloader.params.get('prefer_free_formats', False) else self._available_formats
620 if format_limit is not None and format_limit in available_formats:
621 format_list = available_formats[available_formats.index(format_limit):]
622 else:
623 format_list = available_formats
624 existing_formats = [x for x in format_list if x in url_map]
625 if len(existing_formats) == 0:
626 raise ExtractorError(u'no known formats available for video')
627 if self._downloader.params.get('listformats', None):
628 self._print_formats(existing_formats)
629 return
630 if req_format is None or req_format == 'best':
631 video_url_list = [(existing_formats[0], url_map[existing_formats[0]])] # Best quality
632 elif req_format == 'worst':
d828f3a5 633 video_url_list = [(existing_formats[-1], url_map[existing_formats[-1]])] # worst quality
c5e8d7af
PH
634 elif req_format in ('-1', 'all'):
635 video_url_list = [(f, url_map[f]) for f in existing_formats] # All formats
636 else:
637 # Specific formats. We pick the first in a slash-delimeted sequence.
638 # For example, if '1/2/3/4' is requested and '2' and '4' are available, we pick '2'.
639 req_formats = req_format.split('/')
640 video_url_list = None
641 for rf in req_formats:
642 if rf in url_map:
643 video_url_list = [(rf, url_map[rf])]
644 break
645 if video_url_list is None:
646 raise ExtractorError(u'requested format not available')
647 else:
648 raise ExtractorError(u'no conn or url_encoded_fmt_stream_map information found in video info')
649
650 results = []
651 for format_param, video_real_url in video_url_list:
652 # Extension
653 video_extension = self._video_extensions.get(format_param, 'flv')
654
655 video_format = '{0} - {1}'.format(format_param if format_param else video_extension,
656 self._video_dimensions.get(format_param, '???'))
657
658 results.append({
659 'id': video_id,
660 'url': video_real_url,
661 'uploader': video_uploader,
662 'uploader_id': video_uploader_id,
663 'upload_date': upload_date,
664 'title': video_title,
665 'ext': video_extension,
666 'format': video_format,
667 'thumbnail': video_thumbnail,
668 'description': video_description,
669 'player_url': player_url,
670 'subtitles': video_subtitles,
671 'duration': video_duration
672 })
673 return results
674
675class YoutubePlaylistIE(InfoExtractor):
0f818663 676 IE_DESC = u'YouTube.com playlists'
c5e8d7af
PH
677 _VALID_URL = r"""(?:
678 (?:https?://)?
679 (?:\w+\.)?
680 youtube\.com/
681 (?:
682 (?:course|view_play_list|my_playlists|artist|playlist|watch)
683 \? (?:.*?&)*? (?:p|a|list)=
684 | p/
685 )
c626a3d9 686 ((?:PL|EC|UU|FL)?[0-9A-Za-z-_]{10,})
c5e8d7af
PH
687 .*
688 |
c626a3d9 689 ((?:PL|EC|UU|FL)[0-9A-Za-z-_]{10,})
c5e8d7af
PH
690 )"""
691 _TEMPLATE_URL = 'https://gdata.youtube.com/feeds/api/playlists/%s?max-results=%i&start-index=%i&v=2&alt=json&safeSearch=none'
692 _MAX_RESULTS = 50
693 IE_NAME = u'youtube:playlist'
694
695 @classmethod
696 def suitable(cls, url):
697 """Receives a URL and returns True if suitable for this IE."""
698 return re.match(cls._VALID_URL, url, re.VERBOSE) is not None
699
700 def _real_extract(self, url):
701 # Extract playlist id
702 mobj = re.match(self._VALID_URL, url, re.VERBOSE)
703 if mobj is None:
704 raise ExtractorError(u'Invalid URL: %s' % url)
705
706 # Download playlist videos from API
707 playlist_id = mobj.group(1) or mobj.group(2)
708 page_num = 1
709 videos = []
710
711 while True:
771822eb
JMF
712 start_index = self._MAX_RESULTS * (page_num - 1) + 1
713 if start_index >= 1000:
714 self._downloader.report_warning(u'Max number of results reached')
715 break
716 url = self._TEMPLATE_URL % (playlist_id, self._MAX_RESULTS, start_index)
c5e8d7af
PH
717 page = self._download_webpage(url, playlist_id, u'Downloading page #%s' % page_num)
718
719 try:
720 response = json.loads(page)
721 except ValueError as err:
722 raise ExtractorError(u'Invalid JSON in API response: ' + compat_str(err))
723
724 if 'feed' not in response:
725 raise ExtractorError(u'Got a malformed response from YouTube API')
726 playlist_title = response['feed']['title']['$t']
727 if 'entry' not in response['feed']:
728 # Number of videos is a multiple of self._MAX_RESULTS
729 break
730
731 for entry in response['feed']['entry']:
732 index = entry['yt$position']['$t']
733 if 'media$group' in entry and 'media$player' in entry['media$group']:
734 videos.append((index, entry['media$group']['media$player']['url']))
c5e8d7af
PH
735 page_num += 1
736
737 videos = [v[1] for v in sorted(videos)]
738
20c3893f 739 url_results = [self.url_result(vurl, 'Youtube') for vurl in videos]
c5e8d7af
PH
740 return [self.playlist_result(url_results, playlist_id, playlist_title)]
741
742
743class YoutubeChannelIE(InfoExtractor):
0f818663 744 IE_DESC = u'YouTube.com channels'
c5e8d7af
PH
745 _VALID_URL = r"^(?:https?://)?(?:youtu\.be|(?:\w+\.)?youtube(?:-nocookie)?\.com)/channel/([0-9A-Za-z_-]+)"
746 _TEMPLATE_URL = 'http://www.youtube.com/channel/%s/videos?sort=da&flow=list&view=0&page=%s&gl=US&hl=en'
747 _MORE_PAGES_INDICATOR = 'yt-uix-load-more'
252580c5 748 _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'
c5e8d7af
PH
749 IE_NAME = u'youtube:channel'
750
751 def extract_videos_from_page(self, page):
752 ids_in_page = []
753 for mobj in re.finditer(r'href="/watch\?v=([0-9A-Za-z_-]+)&?', page):
754 if mobj.group(1) not in ids_in_page:
755 ids_in_page.append(mobj.group(1))
756 return ids_in_page
757
758 def _real_extract(self, url):
759 # Extract channel id
760 mobj = re.match(self._VALID_URL, url)
761 if mobj is None:
762 raise ExtractorError(u'Invalid URL: %s' % url)
763
764 # Download channel page
765 channel_id = mobj.group(1)
766 video_ids = []
767 pagenum = 1
768
769 url = self._TEMPLATE_URL % (channel_id, pagenum)
770 page = self._download_webpage(url, channel_id,
771 u'Downloading page #%s' % pagenum)
772
773 # Extract video identifiers
774 ids_in_page = self.extract_videos_from_page(page)
775 video_ids.extend(ids_in_page)
776
777 # Download any subsequent channel pages using the json-based channel_ajax query
778 if self._MORE_PAGES_INDICATOR in page:
779 while True:
780 pagenum = pagenum + 1
781
782 url = self._MORE_PAGES_URL % (pagenum, channel_id)
783 page = self._download_webpage(url, channel_id,
784 u'Downloading page #%s' % pagenum)
785
786 page = json.loads(page)
787
788 ids_in_page = self.extract_videos_from_page(page['content_html'])
789 video_ids.extend(ids_in_page)
790
791 if self._MORE_PAGES_INDICATOR not in page['load_more_widget_html']:
792 break
793
794 self._downloader.to_screen(u'[youtube] Channel %s: Found %i videos' % (channel_id, len(video_ids)))
795
796 urls = ['http://www.youtube.com/watch?v=%s' % id for id in video_ids]
20c3893f 797 url_entries = [self.url_result(eurl, 'Youtube') for eurl in urls]
c5e8d7af
PH
798 return [self.playlist_result(url_entries, channel_id)]
799
800
801class YoutubeUserIE(InfoExtractor):
0f818663 802 IE_DESC = u'YouTube.com user videos (URL or "ytuser" keyword)'
c5e8d7af
PH
803 _VALID_URL = r'(?:(?:(?:https?://)?(?:\w+\.)?youtube\.com/user/)|ytuser:)([A-Za-z0-9_-]+)'
804 _TEMPLATE_URL = 'http://gdata.youtube.com/feeds/api/users/%s'
805 _GDATA_PAGE_SIZE = 50
806 _GDATA_URL = 'http://gdata.youtube.com/feeds/api/users/%s/uploads?max-results=%d&start-index=%d'
807 _VIDEO_INDICATOR = r'/watch\?v=(.+?)[\<&]'
808 IE_NAME = u'youtube:user'
809
810 def _real_extract(self, url):
811 # Extract username
812 mobj = re.match(self._VALID_URL, url)
813 if mobj is None:
814 raise ExtractorError(u'Invalid URL: %s' % url)
815
816 username = mobj.group(1)
817
818 # Download video ids using YouTube Data API. Result size per
819 # query is limited (currently to 50 videos) so we need to query
820 # page by page until there are no video ids - it means we got
821 # all of them.
822
823 video_ids = []
824 pagenum = 0
825
826 while True:
827 start_index = pagenum * self._GDATA_PAGE_SIZE + 1
828
829 gdata_url = self._GDATA_URL % (username, self._GDATA_PAGE_SIZE, start_index)
830 page = self._download_webpage(gdata_url, username,
831 u'Downloading video ids from %d to %d' % (start_index, start_index + self._GDATA_PAGE_SIZE))
832
833 # Extract video identifiers
834 ids_in_page = []
835
836 for mobj in re.finditer(self._VIDEO_INDICATOR, page):
837 if mobj.group(1) not in ids_in_page:
838 ids_in_page.append(mobj.group(1))
839
840 video_ids.extend(ids_in_page)
841
842 # A little optimization - if current page is not
843 # "full", ie. does not contain PAGE_SIZE video ids then
844 # we can assume that this page is the last one - there
845 # are no more ids on further pages - no need to query
846 # again.
847
848 if len(ids_in_page) < self._GDATA_PAGE_SIZE:
849 break
850
851 pagenum += 1
852
853 urls = ['http://www.youtube.com/watch?v=%s' % video_id for video_id in video_ids]
20c3893f 854 url_results = [self.url_result(rurl, 'Youtube') for rurl in urls]
c5e8d7af 855 return [self.playlist_result(url_results, playlist_title = username)]
b05654f0
PH
856
857class YoutubeSearchIE(SearchInfoExtractor):
0f818663 858 IE_DESC = u'YouTube.com searches'
b05654f0
PH
859 _API_URL = 'https://gdata.youtube.com/feeds/api/videos?q=%s&start-index=%i&max-results=50&v=2&alt=jsonc'
860 _MAX_RESULTS = 1000
861 IE_NAME = u'youtube:search'
862 _SEARCH_KEY = 'ytsearch'
863
864 def report_download_page(self, query, pagenum):
865 """Report attempt to download search page with given number."""
866 self._downloader.to_screen(u'[youtube] query "%s": Downloading page %s' % (query, pagenum))
867
868 def _get_n_results(self, query, n):
869 """Get a specified number of results for a query"""
870
871 video_ids = []
872 pagenum = 0
873 limit = n
874
875 while (50 * pagenum) < limit:
876 self.report_download_page(query, pagenum+1)
877 result_url = self._API_URL % (compat_urllib_parse.quote_plus(query), (50*pagenum)+1)
878 request = compat_urllib_request.Request(result_url)
879 try:
880 data = compat_urllib_request.urlopen(request).read().decode('utf-8')
881 except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
882 raise ExtractorError(u'Unable to download API page: %s' % compat_str(err))
883 api_response = json.loads(data)['data']
884
885 if not 'items' in api_response:
886 raise ExtractorError(u'[youtube] No video results')
887
888 new_ids = list(video['id'] for video in api_response['items'])
889 video_ids += new_ids
890
891 limit = min(n, api_response['totalItems'])
892 pagenum += 1
893
894 if len(video_ids) > n:
895 video_ids = video_ids[:n]
896 videos = [self.url_result('http://www.youtube.com/watch?v=%s' % id, 'Youtube') for id in video_ids]
897 return self.playlist_result(videos, query)
75dff0ee
JMF
898
899
900class YoutubeShowIE(InfoExtractor):
0f818663 901 IE_DESC = u'YouTube.com (multi-season) shows'
75dff0ee
JMF
902 _VALID_URL = r'https?://www\.youtube\.com/show/(.*)'
903 IE_NAME = u'youtube:show'
904
905 def _real_extract(self, url):
906 mobj = re.match(self._VALID_URL, url)
907 show_name = mobj.group(1)
908 webpage = self._download_webpage(url, show_name, u'Downloading show webpage')
909 # There's one playlist for each season of the show
910 m_seasons = list(re.finditer(r'href="(/playlist\?list=.*?)"', webpage))
911 self.to_screen(u'%s: Found %s seasons' % (show_name, len(m_seasons)))
912 return [self.url_result('https://www.youtube.com' + season.group(1), 'YoutubePlaylist') for season in m_seasons]
04cc9617
JMF
913
914
b2e8bc1b 915class YoutubeFeedsInfoExtractor(YoutubeBaseInfoExtractor):
d7ae0639
JMF
916 """
917 Base class for extractors that fetch info from
918 http://www.youtube.com/feed_ajax
919 Subclasses must define the _FEED_NAME and _PLAYLIST_TITLE properties.
920 """
b2e8bc1b 921 _LOGIN_REQUIRED = True
04cc9617 922 _PAGING_STEP = 30
43ba5456
JMF
923 # use action_load_personal_feed instead of action_load_system_feed
924 _PERSONAL_FEED = False
04cc9617 925
d7ae0639
JMF
926 @property
927 def _FEED_TEMPLATE(self):
43ba5456
JMF
928 action = 'action_load_system_feed'
929 if self._PERSONAL_FEED:
930 action = 'action_load_personal_feed'
931 return 'http://www.youtube.com/feed_ajax?%s=1&feed_name=%s&paging=%%s' % (action, self._FEED_NAME)
d7ae0639
JMF
932
933 @property
934 def IE_NAME(self):
935 return u'youtube:%s' % self._FEED_NAME
936
81f0259b 937 def _real_initialize(self):
b2e8bc1b 938 self._login()
81f0259b 939
04cc9617
JMF
940 def _real_extract(self, url):
941 feed_entries = []
942 # The step argument is available only in 2.7 or higher
943 for i in itertools.count(0):
944 paging = i*self._PAGING_STEP
d7ae0639
JMF
945 info = self._download_webpage(self._FEED_TEMPLATE % paging,
946 u'%s feed' % self._FEED_NAME,
04cc9617
JMF
947 u'Downloading page %s' % i)
948 info = json.loads(info)
949 feed_html = info['feed_html']
43ba5456 950 m_ids = re.finditer(r'"/watch\?v=(.*?)["&]', feed_html)
04cc9617
JMF
951 ids = orderedSet(m.group(1) for m in m_ids)
952 feed_entries.extend(self.url_result(id, 'Youtube') for id in ids)
953 if info['paging'] is None:
954 break
d7ae0639
JMF
955 return self.playlist_result(feed_entries, playlist_title=self._PLAYLIST_TITLE)
956
957class YoutubeSubscriptionsIE(YoutubeFeedsInfoExtractor):
958 IE_DESC = u'YouTube.com subscriptions feed, "ytsubs" keyword(requires authentication)'
959 _VALID_URL = r'https?://www\.youtube\.com/feed/subscriptions|:ytsubs(?:criptions)?'
960 _FEED_NAME = 'subscriptions'
961 _PLAYLIST_TITLE = u'Youtube Subscriptions'
962
963class YoutubeRecommendedIE(YoutubeFeedsInfoExtractor):
964 IE_DESC = u'YouTube.com recommended videos, "ytrec" keyword (requires authentication)'
965 _VALID_URL = r'https?://www\.youtube\.com/feed/recommended|:ytrec(?:ommended)?'
966 _FEED_NAME = 'recommended'
967 _PLAYLIST_TITLE = u'Youtube Recommended videos'
c626a3d9 968
43ba5456
JMF
969class YoutubeWatchLaterIE(YoutubeFeedsInfoExtractor):
970 IE_DESC = u'Youtube watch later list, "ytwatchlater" keyword (requires authentication)'
971 _VALID_URL = r'https?://www\.youtube\.com/feed/watch_later|:ytwatchlater'
972 _FEED_NAME = 'watch_later'
973 _PLAYLIST_TITLE = u'Youtube Watch Later'
974 _PAGING_STEP = 100
975 _PERSONAL_FEED = True
c626a3d9
JMF
976
977class YoutubeFavouritesIE(YoutubeBaseInfoExtractor):
978 IE_NAME = u'youtube:favorites'
979 IE_DESC = u'YouTube.com favourite videos, "ytfav" keyword (requires authentication)'
980 _VALID_URL = r'https?://www\.youtube\.com/my_favorites|:ytfav(?:o?rites)?'
981 _LOGIN_REQUIRED = True
982
983 def _real_extract(self, url):
984 webpage = self._download_webpage('https://www.youtube.com/my_favorites', 'Youtube Favourites videos')
985 playlist_id = self._search_regex(r'list=(.+?)["&]', webpage, u'favourites playlist id')
986 return self.url_result(playlist_id, 'YoutubePlaylist')