]>
Commit | Line | Data |
---|---|---|
c5e8d7af | 1 | # coding: utf-8 |
c5e8d7af | 2 | |
e0df6211 PH |
3 | import collections |
4 | import itertools | |
5 | import io | |
c5e8d7af PH |
6 | import json |
7 | import netrc | |
8 | import re | |
9 | import socket | |
e0df6211 PH |
10 | import string |
11 | import struct | |
12 | import traceback | |
13 | import zlib | |
c5e8d7af | 14 | |
b05654f0 | 15 | from .common import InfoExtractor, SearchInfoExtractor |
54d39d8b | 16 | from .subtitles import SubtitlesInfoExtractor |
c5e8d7af PH |
17 | from ..utils import ( |
18 | compat_http_client, | |
19 | compat_parse_qs, | |
20 | compat_urllib_error, | |
21 | compat_urllib_parse, | |
22 | compat_urllib_request, | |
23 | compat_str, | |
24 | ||
25 | clean_html, | |
26 | get_element_by_id, | |
27 | ExtractorError, | |
28 | unescapeHTML, | |
29 | unified_strdate, | |
04cc9617 | 30 | orderedSet, |
c5e8d7af PH |
31 | ) |
32 | ||
de7f3446 | 33 | class YoutubeBaseInfoExtractor(InfoExtractor): |
b2e8bc1b JMF |
34 | """Provide base functions for Youtube extractors""" |
35 | _LOGIN_URL = 'https://accounts.google.com/ServiceLogin' | |
36 | _LANG_URL = r'https://www.youtube.com/?hl=en&persist_hl=1&gl=US&persist_gl=1&opt_out_ackd=1' | |
37 | _AGE_URL = 'http://www.youtube.com/verify_age?next_url=/&gl=US&hl=en' | |
38 | _NETRC_MACHINE = 'youtube' | |
39 | # If True it will raise an error if no login info is provided | |
40 | _LOGIN_REQUIRED = False | |
41 | ||
42 | def report_lang(self): | |
43 | """Report attempt to set language.""" | |
44 | self.to_screen(u'Setting language') | |
45 | ||
46 | def _set_language(self): | |
47 | request = compat_urllib_request.Request(self._LANG_URL) | |
48 | try: | |
49 | self.report_lang() | |
50 | compat_urllib_request.urlopen(request).read() | |
51 | except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err: | |
52 | self._downloader.report_warning(u'unable to set language: %s' % compat_str(err)) | |
53 | return False | |
54 | return True | |
55 | ||
56 | def _login(self): | |
57 | (username, password) = self._get_login_info() | |
58 | # No authentication to be performed | |
59 | if username is None: | |
60 | if self._LOGIN_REQUIRED: | |
61 | raise ExtractorError(u'No login info available, needed for using %s.' % self.IE_NAME, expected=True) | |
62 | return False | |
63 | ||
64 | request = compat_urllib_request.Request(self._LOGIN_URL) | |
65 | try: | |
66 | login_page = compat_urllib_request.urlopen(request).read().decode('utf-8') | |
67 | except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err: | |
68 | self._downloader.report_warning(u'unable to fetch login page: %s' % compat_str(err)) | |
69 | return False | |
70 | ||
71 | galx = None | |
72 | dsh = None | |
73 | match = re.search(re.compile(r'<input.+?name="GALX".+?value="(.+?)"', re.DOTALL), login_page) | |
74 | if match: | |
75 | galx = match.group(1) | |
76 | match = re.search(re.compile(r'<input.+?name="dsh".+?value="(.+?)"', re.DOTALL), login_page) | |
77 | if match: | |
78 | dsh = match.group(1) | |
c5e8d7af | 79 | |
b2e8bc1b JMF |
80 | # Log in |
81 | login_form_strs = { | |
82 | u'continue': u'https://www.youtube.com/signin?action_handle_signin=true&feature=sign_in_button&hl=en_US&nomobiletemp=1', | |
83 | u'Email': username, | |
84 | u'GALX': galx, | |
85 | u'Passwd': password, | |
86 | u'PersistentCookie': u'yes', | |
87 | u'_utf8': u'霱', | |
88 | u'bgresponse': u'js_disabled', | |
89 | u'checkConnection': u'', | |
90 | u'checkedDomains': u'youtube', | |
91 | u'dnConn': u'', | |
92 | u'dsh': dsh, | |
93 | u'pstMsg': u'0', | |
94 | u'rmShown': u'1', | |
95 | u'secTok': u'', | |
96 | u'signIn': u'Sign in', | |
97 | u'timeStmp': u'', | |
98 | u'service': u'youtube', | |
99 | u'uilel': u'3', | |
100 | u'hl': u'en_US', | |
101 | } | |
102 | # Convert to UTF-8 *before* urlencode because Python 2.x's urlencode | |
103 | # chokes on unicode | |
104 | login_form = dict((k.encode('utf-8'), v.encode('utf-8')) for k,v in login_form_strs.items()) | |
105 | login_data = compat_urllib_parse.urlencode(login_form).encode('ascii') | |
106 | request = compat_urllib_request.Request(self._LOGIN_URL, login_data) | |
107 | try: | |
108 | self.report_login() | |
109 | login_results = compat_urllib_request.urlopen(request).read().decode('utf-8') | |
110 | if re.search(r'(?i)<form[^>]* id="gaia_loginform"', login_results) is not None: | |
111 | self._downloader.report_warning(u'unable to log in: bad username or password') | |
112 | return False | |
113 | except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err: | |
114 | self._downloader.report_warning(u'unable to log in: %s' % compat_str(err)) | |
115 | return False | |
116 | return True | |
117 | ||
118 | def _confirm_age(self): | |
119 | age_form = { | |
120 | 'next_url': '/', | |
121 | 'action_confirm': 'Confirm', | |
122 | } | |
123 | request = compat_urllib_request.Request(self._AGE_URL, compat_urllib_parse.urlencode(age_form)) | |
124 | try: | |
125 | self.report_age_confirmation() | |
126 | compat_urllib_request.urlopen(request).read().decode('utf-8') | |
127 | except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err: | |
128 | raise ExtractorError(u'Unable to confirm age: %s' % compat_str(err)) | |
129 | return True | |
130 | ||
131 | def _real_initialize(self): | |
132 | if self._downloader is None: | |
133 | return | |
134 | if not self._set_language(): | |
135 | return | |
136 | if not self._login(): | |
137 | return | |
138 | self._confirm_age() | |
c5e8d7af | 139 | |
8377574c | 140 | |
de7f3446 | 141 | class YoutubeIE(YoutubeBaseInfoExtractor, SubtitlesInfoExtractor): |
0f818663 | 142 | IE_DESC = u'YouTube.com' |
c5e8d7af PH |
143 | _VALID_URL = r"""^ |
144 | ( | |
145 | (?:https?://)? # http(s):// (optional) | |
f4b05232 | 146 | (?:(?:(?:(?:\w+\.)?youtube(?:-nocookie)?\.com/| |
e69ae5b9 JMF |
147 | tube\.majestyc\.net/| |
148 | youtube\.googleapis\.com/) # the various hostnames, with wildcard subdomains | |
c5e8d7af PH |
149 | (?:.*?\#/)? # handle anchor (#/) redirect urls |
150 | (?: # the various things that can precede the ID: | |
151 | (?:(?:v|embed|e)/) # v/ or embed/ or e/ | |
152 | |(?: # or the v= param in all its forms | |
d741e55a | 153 | (?:(?:watch|movie)(?:_popup)?(?:\.php)?)? # preceding watch(_popup|.php) or nothing (like /?v=xxxx) |
c5e8d7af PH |
154 | (?:\?|\#!?) # the params delimiter ? or # or #! |
155 | (?:.*?&)? # any other preceding param (like /?s=tuff&v=xxxx) | |
156 | v= | |
157 | ) | |
f4b05232 JMF |
158 | )) |
159 | |youtu\.be/ # just youtu.be/xxxx | |
160 | ) | |
c5e8d7af | 161 | )? # all until now is optional -> you can pass the naked ID |
8963d9c2 | 162 | ([0-9A-Za-z_-]{11}) # here is it! the YouTube video ID |
c5e8d7af PH |
163 | (?(1).+)? # if we found the ID, everything can follow |
164 | $""" | |
c5e8d7af | 165 | _NEXT_URL_RE = r'[\?&]next_url=([^&]+)' |
c5e8d7af | 166 | # Listed in order of quality |
bdc6b3fc | 167 | _available_formats = ['38', '37', '46', '22', '45', '35', '44', '34', '18', '43', '6', '5', '36', '17', '13', |
96fb5605 | 168 | # Apple HTTP Live Streaming |
bdc6b3fc | 169 | '96', '95', '94', '93', '92', '132', '151', |
939fbd26 JMF |
170 | # 3D |
171 | '85', '84', '102', '83', '101', '82', '100', | |
172 | # Dash video | |
173 | '138', '137', '248', '136', '247', '135', '246', | |
174 | '245', '244', '134', '243', '133', '242', '160', | |
175 | # Dash audio | |
176 | '141', '172', '140', '171', '139', | |
1d043b93 | 177 | ] |
bdc6b3fc | 178 | _available_formats_prefer_free = ['38', '46', '37', '45', '22', '44', '35', '43', '34', '18', '6', '5', '36', '17', '13', |
96fb5605 | 179 | # Apple HTTP Live Streaming |
bdc6b3fc AZ |
180 | '96', '95', '94', '93', '92', '132', '151', |
181 | # 3D | |
86fe61c8 | 182 | '85', '102', '84', '101', '83', '100', '82', |
939fbd26 JMF |
183 | # Dash video |
184 | '138', '248', '137', '247', '136', '246', '245', | |
185 | '244', '135', '243', '134', '242', '133', '160', | |
186 | # Dash audio | |
187 | '172', '141', '171', '140', '139', | |
1d043b93 | 188 | ] |
bdc6b3fc AZ |
189 | _video_formats_map = { |
190 | 'flv': ['35', '34', '6', '5'], | |
191 | '3gp': ['36', '17', '13'], | |
192 | 'mp4': ['38', '37', '22', '18'], | |
193 | 'webm': ['46', '45', '44', '43'], | |
194 | } | |
c5e8d7af PH |
195 | _video_extensions = { |
196 | '13': '3gp', | |
bdc6b3fc | 197 | '17': '3gp', |
c5e8d7af PH |
198 | '18': 'mp4', |
199 | '22': 'mp4', | |
bdc6b3fc | 200 | '36': '3gp', |
c5e8d7af | 201 | '37': 'mp4', |
d69cf69a | 202 | '38': 'mp4', |
c5e8d7af PH |
203 | '43': 'webm', |
204 | '44': 'webm', | |
205 | '45': 'webm', | |
206 | '46': 'webm', | |
1d043b93 | 207 | |
86fe61c8 AZ |
208 | # 3d videos |
209 | '82': 'mp4', | |
210 | '83': 'mp4', | |
211 | '84': 'mp4', | |
212 | '85': 'mp4', | |
213 | '100': 'webm', | |
214 | '101': 'webm', | |
215 | '102': 'webm', | |
836a086c | 216 | |
96fb5605 | 217 | # Apple HTTP Live Streaming |
1d043b93 JMF |
218 | '92': 'mp4', |
219 | '93': 'mp4', | |
220 | '94': 'mp4', | |
221 | '95': 'mp4', | |
222 | '96': 'mp4', | |
223 | '132': 'mp4', | |
224 | '151': 'mp4', | |
836a086c AZ |
225 | |
226 | # Dash mp4 | |
227 | '133': 'mp4', | |
228 | '134': 'mp4', | |
229 | '135': 'mp4', | |
230 | '136': 'mp4', | |
231 | '137': 'mp4', | |
232 | '138': 'mp4', | |
233 | '139': 'mp4', | |
234 | '140': 'mp4', | |
235 | '141': 'mp4', | |
236 | '160': 'mp4', | |
237 | ||
238 | # Dash webm | |
239 | '171': 'webm', | |
240 | '172': 'webm', | |
241 | '242': 'webm', | |
242 | '243': 'webm', | |
243 | '244': 'webm', | |
244 | '245': 'webm', | |
245 | '246': 'webm', | |
246 | '247': 'webm', | |
247 | '248': 'webm', | |
c5e8d7af PH |
248 | } |
249 | _video_dimensions = { | |
250 | '5': '240x400', | |
251 | '6': '???', | |
252 | '13': '???', | |
253 | '17': '144x176', | |
254 | '18': '360x640', | |
255 | '22': '720x1280', | |
256 | '34': '360x640', | |
257 | '35': '480x854', | |
bdc6b3fc | 258 | '36': '240x320', |
c5e8d7af PH |
259 | '37': '1080x1920', |
260 | '38': '3072x4096', | |
261 | '43': '360x640', | |
262 | '44': '480x854', | |
263 | '45': '720x1280', | |
264 | '46': '1080x1920', | |
86fe61c8 AZ |
265 | '82': '360p', |
266 | '83': '480p', | |
267 | '84': '720p', | |
268 | '85': '1080p', | |
1d043b93 JMF |
269 | '92': '240p', |
270 | '93': '360p', | |
271 | '94': '480p', | |
272 | '95': '720p', | |
273 | '96': '1080p', | |
86fe61c8 AZ |
274 | '100': '360p', |
275 | '101': '480p', | |
836a086c | 276 | '102': '720p', |
1d043b93 JMF |
277 | '132': '240p', |
278 | '151': '72p', | |
836a086c AZ |
279 | '133': '240p', |
280 | '134': '360p', | |
281 | '135': '480p', | |
282 | '136': '720p', | |
283 | '137': '1080p', | |
284 | '138': '>1080p', | |
285 | '139': '48k', | |
286 | '140': '128k', | |
287 | '141': '256k', | |
288 | '160': '192p', | |
289 | '171': '128k', | |
290 | '172': '256k', | |
291 | '242': '240p', | |
292 | '243': '360p', | |
293 | '244': '480p', | |
294 | '245': '480p', | |
295 | '246': '480p', | |
296 | '247': '720p', | |
297 | '248': '1080p', | |
c5e8d7af | 298 | } |
836a086c AZ |
299 | _special_itags = { |
300 | '82': '3D', | |
301 | '83': '3D', | |
302 | '84': '3D', | |
303 | '85': '3D', | |
304 | '100': '3D', | |
305 | '101': '3D', | |
306 | '102': '3D', | |
307 | '133': 'DASH Video', | |
308 | '134': 'DASH Video', | |
309 | '135': 'DASH Video', | |
310 | '136': 'DASH Video', | |
311 | '137': 'DASH Video', | |
312 | '138': 'DASH Video', | |
313 | '139': 'DASH Audio', | |
314 | '140': 'DASH Audio', | |
315 | '141': 'DASH Audio', | |
316 | '160': 'DASH Video', | |
317 | '171': 'DASH Audio', | |
318 | '172': 'DASH Audio', | |
319 | '242': 'DASH Video', | |
320 | '243': 'DASH Video', | |
321 | '244': 'DASH Video', | |
322 | '245': 'DASH Video', | |
323 | '246': 'DASH Video', | |
324 | '247': 'DASH Video', | |
325 | '248': 'DASH Video', | |
c5e8d7af | 326 | } |
836a086c | 327 | |
c5e8d7af | 328 | IE_NAME = u'youtube' |
2eb88d95 PH |
329 | _TESTS = [ |
330 | { | |
0e853ca4 PH |
331 | u"url": u"http://www.youtube.com/watch?v=BaW_jenozKc", |
332 | u"file": u"BaW_jenozKc.mp4", | |
333 | u"info_dict": { | |
334 | u"title": u"youtube-dl test video \"'/\\ä↭𝕐", | |
335 | u"uploader": u"Philipp Hagemeister", | |
336 | u"uploader_id": u"phihag", | |
337 | u"upload_date": u"20121002", | |
338 | 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 | 339 | } |
0e853ca4 PH |
340 | }, |
341 | { | |
342 | u"url": u"http://www.youtube.com/watch?v=1ltcDfZMA3U", | |
343 | u"file": u"1ltcDfZMA3U.flv", | |
344 | u"note": u"Test VEVO video (#897)", | |
345 | u"info_dict": { | |
346 | u"upload_date": u"20070518", | |
347 | u"title": u"Maps - It Will Find You", | |
348 | u"description": u"Music video by Maps performing It Will Find You.", | |
349 | u"uploader": u"MuteUSA", | |
350 | u"uploader_id": u"MuteUSA" | |
2eb88d95 | 351 | } |
0e853ca4 PH |
352 | }, |
353 | { | |
354 | u"url": u"http://www.youtube.com/watch?v=UxxajLWwzqY", | |
355 | u"file": u"UxxajLWwzqY.mp4", | |
356 | u"note": u"Test generic use_cipher_signature video (#897)", | |
357 | u"info_dict": { | |
358 | u"upload_date": u"20120506", | |
359 | u"title": u"Icona Pop - I Love It (feat. Charli XCX) [OFFICIAL VIDEO]", | |
c7bf7366 | 360 | u"description": u"md5:3e2666e0a55044490499ea45fe9037b7", |
45ed795c | 361 | u"uploader": u"Icona Pop", |
0e853ca4 | 362 | u"uploader_id": u"IconaPop" |
2eb88d95 | 363 | } |
c108eb73 JMF |
364 | }, |
365 | { | |
366 | u"url": u"https://www.youtube.com/watch?v=07FYdnEawAQ", | |
367 | u"file": u"07FYdnEawAQ.mp4", | |
368 | u"note": u"Test VEVO video with age protection (#956)", | |
369 | u"info_dict": { | |
370 | u"upload_date": u"20130703", | |
371 | u"title": u"Justin Timberlake - Tunnel Vision (Explicit)", | |
372 | u"description": u"md5:64249768eec3bc4276236606ea996373", | |
373 | u"uploader": u"justintimberlakeVEVO", | |
374 | u"uploader_id": u"justintimberlakeVEVO" | |
375 | } | |
376 | }, | |
1d043b93 JMF |
377 | { |
378 | u'url': u'https://www.youtube.com/watch?v=TGi3HqYrWHE', | |
379 | u'file': u'TGi3HqYrWHE.mp4', | |
380 | u'note': u'm3u8 video', | |
381 | u'info_dict': { | |
382 | u'title': u'Triathlon - Men - London 2012 Olympic Games', | |
383 | u'description': u'- Men - TR02 - Triathlon - 07 August 2012 - London 2012 Olympic Games', | |
384 | u'uploader': u'olympic', | |
385 | u'upload_date': u'20120807', | |
386 | u'uploader_id': u'olympic', | |
387 | }, | |
388 | u'params': { | |
389 | u'skip_download': True, | |
390 | }, | |
391 | }, | |
2eb88d95 PH |
392 | ] |
393 | ||
c5e8d7af PH |
394 | |
395 | @classmethod | |
396 | def suitable(cls, url): | |
397 | """Receives a URL and returns True if suitable for this IE.""" | |
e3ea4790 | 398 | if YoutubePlaylistIE.suitable(url): return False |
c5e8d7af PH |
399 | return re.match(cls._VALID_URL, url, re.VERBOSE) is not None |
400 | ||
e0df6211 PH |
401 | def __init__(self, *args, **kwargs): |
402 | super(YoutubeIE, self).__init__(*args, **kwargs) | |
403 | self._jsplayer_cache = {} | |
404 | ||
c5e8d7af PH |
405 | def report_video_webpage_download(self, video_id): |
406 | """Report attempt to download video webpage.""" | |
407 | self.to_screen(u'%s: Downloading video webpage' % video_id) | |
408 | ||
409 | def report_video_info_webpage_download(self, video_id): | |
410 | """Report attempt to download video info webpage.""" | |
411 | self.to_screen(u'%s: Downloading video info webpage' % video_id) | |
412 | ||
c5e8d7af PH |
413 | def report_information_extraction(self, video_id): |
414 | """Report attempt to extract video information.""" | |
415 | self.to_screen(u'%s: Extracting video information' % video_id) | |
416 | ||
417 | def report_unavailable_format(self, video_id, format): | |
418 | """Report extracted video URL.""" | |
419 | self.to_screen(u'%s: Format %s not available' % (video_id, format)) | |
420 | ||
421 | def report_rtmp_download(self): | |
422 | """Indicate the download will use the RTMP protocol.""" | |
423 | self.to_screen(u'RTMP download detected') | |
424 | ||
e0df6211 PH |
425 | def _extract_signature_function(self, video_id, player_url): |
426 | id_m = re.match(r'.*-(?P<id>[^.]+)\.(?P<ext>[^.]+)$', player_url) | |
427 | player_type = id_m.group('ext') | |
428 | player_id = id_m.group('id') | |
429 | ||
430 | if player_type == 'js': | |
431 | code = self._download_webpage( | |
432 | player_url, video_id, | |
433 | note=u'Downloading %s player %s' % (player_type, jsplayer_id), | |
434 | errnote=u'Download of %s failed' % player_url) | |
435 | return self._parse_sig_js(code) | |
436 | elif player_tpye == 'swf': | |
437 | urlh = self._request_webpage( | |
438 | player_url, video_id, | |
439 | note=u'Downloading %s player %s' % (player_type, jsplayer_id), | |
440 | errnote=u'Download of %s failed' % player_url) | |
441 | code = urlh.read() | |
442 | return self._parse_sig_swf(code) | |
443 | else: | |
444 | assert False, 'Invalid player type %r' % player_type | |
445 | ||
446 | def _parse_sig_js(self, jscode): | |
447 | funcname = self._search_regex( | |
448 | r'signature=([a-zA-Z]+)', jscode, | |
449 | u'Initial JS player signature function name') | |
450 | ||
451 | functions = {} | |
452 | ||
453 | def argidx(varname): | |
454 | return string.lowercase.index(varname) | |
455 | ||
456 | def interpret_statement(stmt, local_vars, allow_recursion=20): | |
457 | if allow_recursion < 0: | |
458 | raise ExctractorError(u'Recursion limit reached') | |
459 | ||
460 | if stmt.startswith(u'var '): | |
461 | stmt = stmt[len(u'var '):] | |
462 | ass_m = re.match(r'^(?P<out>[a-z]+)(?:\[(?P<index>[^\]]+)\])?' + | |
463 | r'=(?P<expr>.*)$', stmt) | |
464 | if ass_m: | |
465 | if ass_m.groupdict().get('index'): | |
466 | def assign(val): | |
467 | lvar = local_vars[ass_m.group('out')] | |
468 | idx = interpret_expression(ass_m.group('index'), | |
469 | local_vars, allow_recursion) | |
470 | assert isinstance(idx, int) | |
471 | lvar[idx] = val | |
472 | return val | |
473 | expr = ass_m.group('expr') | |
474 | else: | |
475 | def assign(val): | |
476 | local_vars[ass_m.group('out')] = val | |
477 | return val | |
478 | expr = ass_m.group('expr') | |
479 | elif stmt.startswith(u'return '): | |
480 | assign = lambda v: v | |
481 | expr = stmt[len(u'return '):] | |
482 | else: | |
483 | raise ExtractorError( | |
484 | u'Cannot determine left side of statement in %r' % stmt) | |
485 | ||
486 | v = interpret_expression(expr, local_vars, allow_recursion) | |
487 | return assign(v) | |
488 | ||
489 | def interpret_expression(expr, local_vars, allow_recursion): | |
490 | if expr.isdigit(): | |
491 | return int(expr) | |
492 | ||
493 | if expr.isalpha(): | |
494 | return local_vars[expr] | |
495 | ||
496 | m = re.match(r'^(?P<in>[a-z]+)\.(?P<member>.*)$', expr) | |
497 | if m: | |
498 | member = m.group('member') | |
499 | val = local_vars[m.group('in')] | |
500 | if member == 'split("")': | |
501 | return list(val) | |
502 | if member == 'join("")': | |
503 | return u''.join(val) | |
504 | if member == 'length': | |
505 | return len(val) | |
506 | if member == 'reverse()': | |
507 | return val[::-1] | |
508 | slice_m = re.match(r'slice\((?P<idx>.*)\)', member) | |
509 | if slice_m: | |
510 | idx = interpret_expression( | |
511 | slice_m.group('idx'), local_vars, allow_recursion-1) | |
512 | return val[idx:] | |
513 | ||
514 | m = re.match( | |
515 | r'^(?P<in>[a-z]+)\[(?P<idx>.+)\]$', expr) | |
516 | if m: | |
517 | val = local_vars[m.group('in')] | |
518 | idx = interpret_expression(m.group('idx'), local_vars, | |
519 | allow_recursion-1) | |
520 | return val[idx] | |
521 | ||
522 | m = re.match(r'^(?P<a>.+?)(?P<op>[%])(?P<b>.+?)$', expr) | |
523 | if m: | |
524 | a = interpret_expression(m.group('a'), | |
525 | local_vars, allow_recursion) | |
526 | b = interpret_expression(m.group('b'), | |
527 | local_vars, allow_recursion) | |
528 | return a % b | |
529 | ||
530 | m = re.match( | |
531 | r'^(?P<func>[a-zA-Z]+)\((?P<args>[a-z0-9,]+)\)$', expr) | |
532 | if m: | |
533 | fname = m.group('func') | |
534 | if fname not in functions: | |
535 | functions[fname] = extract_function(fname) | |
536 | argvals = [int(v) if v.isdigit() else local_vars[v] | |
537 | for v in m.group('args').split(',')] | |
538 | return functions[fname](argvals) | |
539 | raise ExtractorError(u'Unsupported JS expression %r' % expr) | |
540 | ||
541 | def extract_function(funcname): | |
542 | func_m = re.search( | |
543 | r'function ' + re.escape(funcname) + | |
544 | r'\((?P<args>[a-z,]+)\){(?P<code>[^}]+)}', | |
545 | jscode) | |
546 | argnames = func_m.group('args').split(',') | |
547 | ||
548 | def resf(args): | |
549 | local_vars = dict(zip(argnames, args)) | |
550 | for stmt in func_m.group('code').split(';'): | |
551 | res = interpret_statement(stmt, local_vars) | |
552 | return res | |
553 | return resf | |
554 | ||
555 | initial_function = extract_function(funcname) | |
556 | return lambda s: initial_function([s]) | |
557 | ||
558 | def _parse_sig_swf(self, file_contents): | |
559 | if file_contents[1:3] != b'WS': | |
560 | raise ExtractorError( | |
561 | u'Not an SWF file; header is %r' % file_contents[:3]) | |
562 | if file_contents[:1] == b'C': | |
563 | content = zlib.decompress(file_contents[8:]) | |
564 | else: | |
565 | raise NotImplementedError(u'Unsupported compression format %r' % | |
566 | file_contents[:1]) | |
567 | ||
568 | def extract_tags(content): | |
569 | pos = 0 | |
570 | while pos < len(content): | |
571 | header16 = struct.unpack('<H', content[pos:pos+2])[0] | |
572 | pos += 2 | |
573 | tag_code = header16 >> 6 | |
574 | tag_len = header16 & 0x3f | |
575 | if tag_len == 0x3f: | |
576 | tag_len = struct.unpack('<I', content[pos:pos+4])[0] | |
577 | pos += 4 | |
578 | assert pos+tag_len <= len(content) | |
579 | yield (tag_code, content[pos:pos+tag_len]) | |
580 | pos += tag_len | |
581 | ||
582 | code_tag = next(tag | |
583 | for tag_code, tag in extract_tags(content) | |
584 | if tag_code == 82) | |
585 | p = code_tag.index(b'\0', 4) + 1 | |
586 | ||
587 | # Parse ABC (AVM2 ByteCode) | |
588 | def read_int(data=None, pos=None): | |
589 | if hasattr(data, 'read'): | |
590 | assert pos is None | |
591 | ||
592 | res = 0 | |
593 | shift = 0 | |
594 | for _ in range(5): | |
595 | buf = data.read(1) | |
596 | assert len(buf) == 1 | |
597 | b = struct.unpack('<B', buf)[0] | |
598 | res = res | ((b & 0x7f) << shift) | |
599 | if b & 0x80 == 0: | |
600 | break | |
601 | shift += 7 | |
602 | return res | |
603 | ||
604 | if data is None: | |
605 | data = code_tag | |
606 | if pos is None: | |
607 | pos = p | |
608 | res = 0 | |
609 | shift = 0 | |
610 | for _ in range(5): | |
611 | b = struct.unpack('<B', data[pos:pos+1])[0] | |
612 | pos += 1 | |
613 | res = res | ((b & 0x7f) << shift) | |
614 | if b & 0x80 == 0: | |
615 | break | |
616 | shift += 7 | |
617 | return (res, pos) | |
618 | assert read_int(b'\x00', 0) == (0, 1) | |
619 | assert read_int(b'\x10', 0) == (16, 1) | |
620 | assert read_int(b'\x34', 0) == (0x34, 1) | |
621 | assert read_int(b'\xb4\x12', 0) == (0x12 * 0x80 + 0x34, 2) | |
622 | assert read_int(b'\xff\xff\xff\x00', 0) == (0x1fffff, 4) | |
623 | ||
624 | def u30(*args, **kwargs): | |
625 | res = read_int(*args, **kwargs) | |
626 | if isinstance(res, tuple): | |
627 | assert res[0] & 0xf0000000 == 0 | |
628 | else: | |
629 | assert res & 0xf0000000 == 0 | |
630 | return res | |
631 | u32 = read_int | |
632 | ||
633 | def s32(data=None, pos=None): | |
634 | v, pos = read_int(data, pos) | |
635 | if v & 0x80000000 != 0: | |
636 | v = - ((v ^ 0xffffffff) + 1) | |
637 | return (v, pos) | |
638 | assert s32(b'\xff\xff\xff\xff\x0f', 0) == (-1, 5) | |
639 | ||
640 | def string(): | |
641 | slen, p = u30() | |
642 | return (code_tag[p:p+slen].decode('utf-8'), p + slen) | |
643 | ||
644 | def read_byte(data=None, pos=None): | |
645 | if data is None: | |
646 | data = code_tag | |
647 | if pos is None: | |
648 | pos = p | |
649 | res = struct.unpack('<B', data[pos:pos+1])[0] | |
650 | return (res, pos + 1) | |
651 | ||
652 | # minor_version + major_version | |
653 | p += 2 + 2 | |
654 | ||
655 | # Constant pool | |
656 | int_count, p = u30() | |
657 | for _c in range(1, int_count): | |
658 | _, p = s32() | |
659 | uint_count, p = u30() | |
660 | for _c in range(1, uint_count): | |
661 | _, p = u32() | |
662 | double_count, p = u30() | |
663 | p += (double_count-1) * 8 | |
664 | string_count, p = u30() | |
665 | constant_strings = [u''] | |
666 | for _c in range(1, string_count): | |
667 | s, p = string() | |
668 | constant_strings.append(s) | |
669 | namespace_count, p = u30() | |
670 | for _c in range(1, namespace_count): | |
671 | p += 1 # kind | |
672 | _, p = u30() # name | |
673 | ns_set_count, p = u30() | |
674 | for _c in range(1, ns_set_count): | |
675 | count, p = u30() | |
676 | for _c2 in range(count): | |
677 | _, p = u30() | |
678 | multiname_count, p = u30() | |
679 | MULTINAME_SIZES = { | |
680 | 0x07: 2, # QName | |
681 | 0x0d: 2, # QNameA | |
682 | 0x0f: 1, # RTQName | |
683 | 0x10: 1, # RTQNameA | |
684 | 0x11: 0, # RTQNameL | |
685 | 0x12: 0, # RTQNameLA | |
686 | 0x09: 2, # Multiname | |
687 | 0x0e: 2, # MultinameA | |
688 | 0x1b: 1, # MultinameL | |
689 | 0x1c: 1, # MultinameLA | |
690 | } | |
691 | multinames = [u''] | |
692 | for _c in range(1, multiname_count): | |
693 | kind, p = u30() | |
694 | assert kind in MULTINAME_SIZES, u'Invalid multiname kind %r' % kind | |
695 | if kind == 0x07: | |
696 | namespace_idx, p = u30() | |
697 | name_idx, p = u30() | |
698 | multinames.append(constant_strings[name_idx]) | |
699 | else: | |
700 | multinames.append('[MULTINAME kind: %d]' % kind) | |
701 | for _c2 in range(MULTINAME_SIZES[kind]): | |
702 | _, p = u30() | |
703 | ||
704 | # Methods | |
705 | method_count, p = u30() | |
706 | MethodInfo = collections.namedtuple( | |
707 | 'MethodInfo', | |
708 | ['NEED_ARGUMENTS', 'NEED_REST']) | |
709 | method_infos = [] | |
710 | for method_id in range(method_count): | |
711 | param_count, p = u30() | |
712 | _, p = u30() # return type | |
713 | for _ in range(param_count): | |
714 | _, p = u30() # param type | |
715 | _, p = u30() # name index (always 0 for youtube) | |
716 | flags, p = read_byte() | |
717 | if flags & 0x08 != 0: | |
718 | # Options present | |
719 | option_count, p = u30() | |
720 | for c in range(option_count): | |
721 | _, p = u30() # val | |
722 | p += 1 # kind | |
723 | if flags & 0x80 != 0: | |
724 | # Param names present | |
725 | for _ in range(param_count): | |
726 | _, p = u30() # param name | |
727 | mi = MethodInfo(flags & 0x01 != 0, flags & 0x04 != 0) | |
728 | method_infos.append(mi) | |
729 | ||
730 | # Metadata | |
731 | metadata_count, p = u30() | |
732 | for _c in range(metadata_count): | |
733 | _, p = u30() # name | |
734 | item_count, p = u30() | |
735 | for _c2 in range(item_count): | |
736 | _, p = u30() # key | |
737 | _, p = u30() # value | |
738 | ||
739 | def parse_traits_info(pos=None): | |
740 | if pos is None: | |
741 | pos = p | |
742 | trait_name_idx, pos = u30(pos=pos) | |
743 | kind_full, pos = read_byte(pos=pos) | |
744 | kind = kind_full & 0x0f | |
745 | attrs = kind_full >> 4 | |
746 | methods = {} | |
747 | if kind in [0x00, 0x06]: # Slot or Const | |
748 | _, pos = u30(pos=pos) # Slot id | |
749 | type_name_idx, pos = u30(pos=pos) | |
750 | vindex, pos = u30(pos=pos) | |
751 | if vindex != 0: | |
752 | _, pos = read_byte(pos=pos) # vkind | |
753 | elif kind in [0x01, 0x02, 0x03]: # Method / Getter / Setter | |
754 | _, pos = u30(pos=pos) # disp_id | |
755 | method_idx, pos = u30(pos=pos) | |
756 | methods[multinames[trait_name_idx]] = method_idx | |
757 | elif kind == 0x04: # Class | |
758 | _, pos = u30(pos=pos) # slot_id | |
759 | _, pos = u30(pos=pos) # classi | |
760 | elif kind == 0x05: # Function | |
761 | _, pos = u30(pos=pos) # slot_id | |
762 | function_idx, pos = u30(pos=pos) | |
763 | methods[function_idx] = multinames[trait_name_idx] | |
764 | else: | |
765 | raise ExtractorError(u'Unsupported trait kind %d' % kind) | |
766 | ||
767 | if attrs & 0x4 != 0: # Metadata present | |
768 | metadata_count, pos = u30(pos=pos) | |
769 | for _c3 in range(metadata_count): | |
770 | _, pos = u30(pos=pos) | |
771 | ||
772 | return (methods, pos) | |
773 | ||
774 | # Classes | |
775 | TARGET_CLASSNAME = u'SignatureDecipher' | |
776 | searched_idx = multinames.index(TARGET_CLASSNAME) | |
777 | searched_class_id = None | |
778 | class_count, p = u30() | |
779 | for class_id in range(class_count): | |
780 | name_idx, p = u30() | |
781 | if name_idx == searched_idx: | |
782 | # We found the class we're looking for! | |
783 | searched_class_id = class_id | |
784 | _, p = u30() # super_name idx | |
785 | flags, p = read_byte() | |
786 | if flags & 0x08 != 0: # Protected namespace is present | |
787 | protected_ns_idx, p = u30() | |
788 | intrf_count, p = u30() | |
789 | for _c2 in range(intrf_count): | |
790 | _, p = u30() | |
791 | _, p = u30() # iinit | |
792 | trait_count, p = u30() | |
793 | for _c2 in range(trait_count): | |
794 | _, p = parse_traits_info() | |
795 | ||
796 | if searched_class_id is None: | |
797 | raise ExtractorError(u'Target class %r not found' % | |
798 | TARGET_CLASSNAME) | |
799 | ||
800 | method_names = {} | |
801 | method_idxs = {} | |
802 | for class_id in range(class_count): | |
803 | _, p = u30() # cinit | |
804 | trait_count, p = u30() | |
805 | for _c2 in range(trait_count): | |
806 | trait_methods, p = parse_traits_info() | |
807 | if class_id == searched_class_id: | |
808 | method_names.update(trait_methods.items()) | |
809 | method_idxs.update(dict( | |
810 | (idx, name) | |
811 | for name, idx in trait_methods.items())) | |
812 | ||
813 | # Scripts | |
814 | script_count, p = u30() | |
815 | for _c in range(script_count): | |
816 | _, p = u30() # init | |
817 | trait_count, p = u30() | |
818 | for _c2 in range(trait_count): | |
819 | _, p = parse_traits_info() | |
820 | ||
821 | # Method bodies | |
822 | method_body_count, p = u30() | |
823 | Method = collections.namedtuple('Method', ['code', 'local_count']) | |
824 | methods = {} | |
825 | for _c in range(method_body_count): | |
826 | method_idx, p = u30() | |
827 | max_stack, p = u30() | |
828 | local_count, p = u30() | |
829 | init_scope_depth, p = u30() | |
830 | max_scope_depth, p = u30() | |
831 | code_length, p = u30() | |
832 | if method_idx in method_idxs: | |
833 | m = Method(code_tag[p:p+code_length], local_count) | |
834 | methods[method_idxs[method_idx]] = m | |
835 | p += code_length | |
836 | exception_count, p = u30() | |
837 | for _c2 in range(exception_count): | |
838 | _, p = u30() # from | |
839 | _, p = u30() # to | |
840 | _, p = u30() # target | |
841 | _, p = u30() # exc_type | |
842 | _, p = u30() # var_name | |
843 | trait_count, p = u30() | |
844 | for _c2 in range(trait_count): | |
845 | _, p = parse_traits_info() | |
846 | ||
847 | assert p == len(code_tag) | |
848 | assert len(methods) == len(method_idxs) | |
849 | ||
850 | method_pyfunctions = {} | |
851 | ||
852 | def extract_function(func_name): | |
853 | if func_name in method_pyfunctions: | |
854 | return method_pyfunctions[func_name] | |
855 | if func_name not in methods: | |
856 | raise ExtractorError(u'Cannot find function %r' % func_name) | |
857 | m = methods[func_name] | |
858 | ||
859 | def resfunc(args): | |
e0df6211 PH |
860 | registers = ['(this)'] + list(args) + [None] * m.local_count |
861 | stack = [] | |
862 | coder = io.BytesIO(m.code) | |
863 | while True: | |
864 | opcode = struct.unpack('!B', coder.read(1))[0] | |
a7177865 | 865 | if opcode == 36: # pushbyte |
e0df6211 PH |
866 | v = struct.unpack('!B', coder.read(1))[0] |
867 | stack.append(v) | |
868 | elif opcode == 44: # pushstring | |
869 | idx = u30(coder) | |
870 | stack.append(constant_strings[idx]) | |
871 | elif opcode == 48: # pushscope | |
872 | # We don't implement the scope register, so we'll just | |
873 | # ignore the popped value | |
874 | stack.pop() | |
875 | elif opcode == 70: # callproperty | |
876 | index = u30(coder) | |
877 | mname = multinames[index] | |
878 | arg_count = u30(coder) | |
879 | args = list(reversed( | |
880 | [stack.pop() for _ in range(arg_count)])) | |
881 | obj = stack.pop() | |
882 | if mname == u'split': | |
883 | assert len(args) == 1 | |
884 | assert isinstance(args[0], compat_str) | |
885 | assert isinstance(obj, compat_str) | |
886 | if args[0] == u'': | |
887 | res = list(obj) | |
888 | else: | |
889 | res = obj.split(args[0]) | |
890 | stack.append(res) | |
a7177865 PH |
891 | elif mname == u'slice': |
892 | assert len(args) == 1 | |
893 | assert isinstance(args[0], int) | |
894 | assert isinstance(obj, list) | |
895 | res = obj[args[0]:] | |
896 | stack.append(res) | |
897 | elif mname == u'join': | |
898 | assert len(args) == 1 | |
899 | assert isinstance(args[0], compat_str) | |
900 | assert isinstance(obj, list) | |
901 | res = args[0].join(obj) | |
902 | stack.append(res) | |
e0df6211 PH |
903 | elif mname in method_pyfunctions: |
904 | stack.append(method_pyfunctions[mname](args)) | |
905 | else: | |
906 | raise NotImplementedError( | |
907 | u'Unsupported property %r on %r' | |
908 | % (mname, obj)) | |
a7177865 PH |
909 | elif opcode == 72: # returnvalue |
910 | res = stack.pop() | |
911 | return res | |
912 | elif opcode == 79: # callpropvoid | |
913 | index = u30(coder) | |
914 | mname = multinames[index] | |
915 | arg_count = u30(coder) | |
916 | args = list(reversed( | |
917 | [stack.pop() for _ in range(arg_count)])) | |
918 | obj = stack.pop() | |
919 | if mname == u'reverse': | |
920 | assert isinstance(obj, list) | |
921 | obj.reverse() | |
922 | else: | |
923 | raise NotImplementedError( | |
924 | u'Unsupported (void) property %r on %r' | |
925 | % (mname, obj)) | |
e0df6211 PH |
926 | elif opcode == 93: # findpropstrict |
927 | index = u30(coder) | |
928 | mname = multinames[index] | |
929 | res = extract_function(mname) | |
930 | stack.append(res) | |
931 | elif opcode == 97: # setproperty | |
932 | index = u30(coder) | |
933 | value = stack.pop() | |
934 | idx = stack.pop() | |
935 | obj = stack.pop() | |
936 | assert isinstance(obj, list) | |
937 | assert isinstance(idx, int) | |
938 | obj[idx] = value | |
939 | elif opcode == 98: # getlocal | |
940 | index = u30(coder) | |
941 | stack.append(registers[index]) | |
942 | elif opcode == 99: # setlocal | |
943 | index = u30(coder) | |
944 | value = stack.pop() | |
945 | registers[index] = value | |
946 | elif opcode == 102: # getproperty | |
947 | index = u30(coder) | |
948 | pname = multinames[index] | |
949 | if pname == u'length': | |
950 | obj = stack.pop() | |
951 | assert isinstance(obj, list) | |
952 | stack.append(len(obj)) | |
953 | else: # Assume attribute access | |
954 | idx = stack.pop() | |
955 | assert isinstance(idx, int) | |
956 | obj = stack.pop() | |
957 | assert isinstance(obj, list) | |
958 | stack.append(obj[idx]) | |
959 | elif opcode == 128: # coerce | |
960 | _ = u30(coder) | |
961 | elif opcode == 133: # coerce_s | |
962 | assert isinstance(stack[-1], (type(None), compat_str)) | |
963 | elif opcode == 164: # modulo | |
964 | value2 = stack.pop() | |
965 | value1 = stack.pop() | |
966 | res = value1 % value2 | |
967 | stack.append(res) | |
a7177865 PH |
968 | elif opcode == 208: # getlocal_0 |
969 | stack.append(registers[0]) | |
970 | elif opcode == 209: # getlocal_1 | |
971 | stack.append(registers[1]) | |
972 | elif opcode == 210: # getlocal_2 | |
973 | stack.append(registers[2]) | |
974 | elif opcode == 211: # getlocal_3 | |
975 | stack.append(registers[3]) | |
e0df6211 PH |
976 | elif opcode == 214: # setlocal_2 |
977 | registers[2] = stack.pop() | |
978 | elif opcode == 215: # setlocal_3 | |
979 | registers[3] = stack.pop() | |
980 | else: | |
981 | raise NotImplementedError( | |
982 | u'Unsupported opcode %d' % opcode) | |
983 | ||
984 | method_pyfunctions[func_name] = resfunc | |
985 | return resfunc | |
986 | ||
987 | initial_function = extract_function(u'decipher') | |
988 | return lambda s: initial_function([s]) | |
989 | ||
990 | def _decrypt_signature(self, s, video_id, jsplayer_url, age_gate=False): | |
257a2501 | 991 | """Turn the encrypted s field into a working signature""" |
6b37f0be | 992 | |
e0df6211 PH |
993 | if jsplayer_url is not None: |
994 | try: | |
995 | if jsplayer_url not in self._jsplayer_cache: | |
996 | self._jsplayer_cache[jsplayer_url] = self._extract_signature_function( | |
997 | video_id, jsplayer_url | |
998 | ) | |
999 | return self._jsplayer_cache[jsplayer_url]([s]) | |
1000 | except Exception as e: | |
1001 | tb = traceback.format_exc() | |
1002 | self._downloader.report_warning(u'Automatic signature extraction failed: ' + tb) | |
1003 | ||
1004 | self._downloader.report_warning(u'Warning: Falling back to static signature algorithm') | |
1005 | ||
1006 | if age_gate: | |
1007 | # The videos with age protection use another player, so the | |
1008 | # algorithms can be different. | |
1009 | if len(s) == 86: | |
1010 | return s[2:63] + s[82] + s[64:82] + s[63] | |
1011 | ||
1012 | if len(s) == 92: | |
444b1165 JMF |
1013 | return s[25] + s[3:25] + s[0] + s[26:42] + s[79] + s[43:79] + s[91] + s[80:83] |
1014 | elif len(s) == 90: | |
1015 | return s[25] + s[3:25] + s[2] + s[26:40] + s[77] + s[41:77] + s[89] + s[78:81] | |
8a9d86a2 | 1016 | elif len(s) == 89: |
1017 | return s[84:78:-1] + s[87] + s[77:60:-1] + s[0] + s[59:3:-1] | |
444b1165 | 1018 | elif len(s) == 88: |
3e223834 | 1019 | return s[7:28] + s[87] + s[29:45] + s[55] + s[46:55] + s[2] + s[56:87] + s[28] |
be547e1d | 1020 | elif len(s) == 87: |
3a725669 | 1021 | return s[6:27] + s[4] + s[28:39] + s[27] + s[40:59] + s[2] + s[60:] |
be547e1d | 1022 | elif len(s) == 86: |
1cf911bc | 1023 | return s[5:34] + s[0] + s[35:38] + s[3] + s[39:45] + s[38] + s[46:53] + s[73] + s[54:73] + s[85] + s[74:85] + s[53] |
be547e1d | 1024 | elif len(s) == 85: |
6ae8ee3f | 1025 | return s[3:11] + s[0] + s[12:55] + s[84] + s[56:84] |
be547e1d | 1026 | elif len(s) == 84: |
23b00bc0 | 1027 | return s[81:36:-1] + s[0] + s[35:2:-1] |
be547e1d | 1028 | elif len(s) == 83: |
e1842025 | 1029 | 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] |
be547e1d | 1030 | elif len(s) == 82: |
ce85f022 | 1031 | return s[80:73:-1] + s[81] + s[72:54:-1] + s[2] + s[53:43:-1] + s[0] + s[42:2:-1] + s[43] + s[1] + s[54] |
be547e1d | 1032 | elif len(s) == 81: |
aedd6bb9 | 1033 | 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] |
066090dd JMF |
1034 | elif len(s) == 80: |
1035 | return s[1:19] + s[0] + s[20:68] + s[19] + s[69:80] | |
5c468ca8 JMF |
1036 | elif len(s) == 79: |
1037 | 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] | |
be547e1d PH |
1038 | |
1039 | else: | |
1040 | raise ExtractorError(u'Unable to decrypt signature, key length %d not supported; retrying might work' % (len(s))) | |
c5e8d7af | 1041 | |
75952c6e JMF |
1042 | def _decrypt_signature_age_gate(self, s): |
1043 | # The videos with age protection use another player, so the algorithms | |
1044 | # can be different. | |
1045 | if len(s) == 86: | |
1046 | return s[2:63] + s[82] + s[64:82] + s[63] | |
1047 | else: | |
1048 | # Fallback to the other algortihms | |
b072a9de | 1049 | return self._decrypt_signature(s) |
c5e8d7af | 1050 | |
de7f3446 | 1051 | def _get_available_subtitles(self, video_id): |
de7f3446 | 1052 | try: |
7fad1c63 JMF |
1053 | sub_list = self._download_webpage( |
1054 | 'http://video.google.com/timedtext?hl=en&type=list&v=%s' % video_id, | |
1055 | video_id, note=False) | |
1056 | except ExtractorError as err: | |
de7f3446 JMF |
1057 | self._downloader.report_warning(u'unable to download video subtitles: %s' % compat_str(err)) |
1058 | return {} | |
1059 | lang_list = re.findall(r'name="([^"]*)"[^>]+lang_code="([\w\-]+)"', sub_list) | |
1060 | ||
1061 | sub_lang_list = {} | |
1062 | for l in lang_list: | |
1063 | lang = l[1] | |
1064 | params = compat_urllib_parse.urlencode({ | |
1065 | 'lang': lang, | |
1066 | 'v': video_id, | |
1067 | 'fmt': self._downloader.params.get('subtitlesformat'), | |
1068 | }) | |
1069 | url = u'http://www.youtube.com/api/timedtext?' + params | |
1070 | sub_lang_list[lang] = url | |
1071 | if not sub_lang_list: | |
1072 | self._downloader.report_warning(u'video doesn\'t have subtitles') | |
1073 | return {} | |
1074 | return sub_lang_list | |
1075 | ||
055e6f36 | 1076 | def _get_available_automatic_caption(self, video_id, webpage): |
de7f3446 JMF |
1077 | """We need the webpage for getting the captions url, pass it as an |
1078 | argument to speed up the process.""" | |
de7f3446 JMF |
1079 | sub_format = self._downloader.params.get('subtitlesformat') |
1080 | self.to_screen(u'%s: Looking for automatic captions' % video_id) | |
1081 | mobj = re.search(r';ytplayer.config = ({.*?});', webpage) | |
055e6f36 | 1082 | err_msg = u'Couldn\'t find automatic captions for %s' % video_id |
de7f3446 JMF |
1083 | if mobj is None: |
1084 | self._downloader.report_warning(err_msg) | |
1085 | return {} | |
1086 | player_config = json.loads(mobj.group(1)) | |
1087 | try: | |
1088 | args = player_config[u'args'] | |
1089 | caption_url = args[u'ttsurl'] | |
1090 | timestamp = args[u'timestamp'] | |
055e6f36 JMF |
1091 | # We get the available subtitles |
1092 | list_params = compat_urllib_parse.urlencode({ | |
1093 | 'type': 'list', | |
1094 | 'tlangs': 1, | |
1095 | 'asrs': 1, | |
de7f3446 | 1096 | }) |
055e6f36 JMF |
1097 | list_url = caption_url + '&' + list_params |
1098 | list_page = self._download_webpage(list_url, video_id) | |
1099 | caption_list = xml.etree.ElementTree.fromstring(list_page.encode('utf-8')) | |
e3dc22ca JMF |
1100 | original_lang_node = caption_list.find('track') |
1101 | if original_lang_node.attrib.get('kind') != 'asr' : | |
1102 | self._downloader.report_warning(u'Video doesn\'t have automatic captions') | |
1103 | return {} | |
1104 | original_lang = original_lang_node.attrib['lang_code'] | |
055e6f36 JMF |
1105 | |
1106 | sub_lang_list = {} | |
1107 | for lang_node in caption_list.findall('target'): | |
1108 | sub_lang = lang_node.attrib['lang_code'] | |
1109 | params = compat_urllib_parse.urlencode({ | |
1110 | 'lang': original_lang, | |
1111 | 'tlang': sub_lang, | |
1112 | 'fmt': sub_format, | |
1113 | 'ts': timestamp, | |
1114 | 'kind': 'asr', | |
1115 | }) | |
1116 | sub_lang_list[sub_lang] = caption_url + '&' + params | |
1117 | return sub_lang_list | |
de7f3446 JMF |
1118 | # An extractor error can be raise by the download process if there are |
1119 | # no automatic captions but there are subtitles | |
1120 | except (KeyError, ExtractorError): | |
1121 | self._downloader.report_warning(err_msg) | |
1122 | return {} | |
1123 | ||
c5e8d7af PH |
1124 | def _print_formats(self, formats): |
1125 | print('Available formats:') | |
1126 | for x in formats: | |
03cc7c20 JMF |
1127 | print('%s\t:\t%s\t[%s]%s' %(x, self._video_extensions.get(x, 'flv'), |
1128 | self._video_dimensions.get(x, '???'), | |
836a086c | 1129 | ' ('+self._special_itags[x]+')' if x in self._special_itags else '')) |
c5e8d7af PH |
1130 | |
1131 | def _extract_id(self, url): | |
1132 | mobj = re.match(self._VALID_URL, url, re.VERBOSE) | |
1133 | if mobj is None: | |
1134 | raise ExtractorError(u'Invalid URL: %s' % url) | |
1135 | video_id = mobj.group(2) | |
1136 | return video_id | |
1137 | ||
1d043b93 JMF |
1138 | def _get_video_url_list(self, url_map): |
1139 | """ | |
1140 | Transform a dictionary in the format {itag:url} to a list of (itag, url) | |
1141 | with the requested formats. | |
1142 | """ | |
1143 | req_format = self._downloader.params.get('format', None) | |
1144 | format_limit = self._downloader.params.get('format_limit', None) | |
1145 | available_formats = self._available_formats_prefer_free if self._downloader.params.get('prefer_free_formats', False) else self._available_formats | |
1146 | if format_limit is not None and format_limit in available_formats: | |
1147 | format_list = available_formats[available_formats.index(format_limit):] | |
1148 | else: | |
1149 | format_list = available_formats | |
1150 | existing_formats = [x for x in format_list if x in url_map] | |
1151 | if len(existing_formats) == 0: | |
1152 | raise ExtractorError(u'no known formats available for video') | |
1153 | if self._downloader.params.get('listformats', None): | |
1154 | self._print_formats(existing_formats) | |
1155 | return | |
1156 | if req_format is None or req_format == 'best': | |
1157 | video_url_list = [(existing_formats[0], url_map[existing_formats[0]])] # Best quality | |
1158 | elif req_format == 'worst': | |
1159 | video_url_list = [(existing_formats[-1], url_map[existing_formats[-1]])] # worst quality | |
1160 | elif req_format in ('-1', 'all'): | |
1161 | video_url_list = [(f, url_map[f]) for f in existing_formats] # All formats | |
1162 | else: | |
1163 | # Specific formats. We pick the first in a slash-delimeted sequence. | |
bdc6b3fc AZ |
1164 | # Format can be specified as itag or 'mp4' or 'flv' etc. We pick the highest quality |
1165 | # available in the specified format. For example, | |
1166 | # if '1/2/3/4' is requested and '2' and '4' are available, we pick '2'. | |
1167 | # if '1/mp4/3/4' is requested and '1' and '5' (is a mp4) are available, we pick '1'. | |
1168 | # if '1/mp4/3/4' is requested and '4' and '5' (is a mp4) are available, we pick '5'. | |
1d043b93 JMF |
1169 | req_formats = req_format.split('/') |
1170 | video_url_list = None | |
1171 | for rf in req_formats: | |
1172 | if rf in url_map: | |
1173 | video_url_list = [(rf, url_map[rf])] | |
1174 | break | |
bdc6b3fc AZ |
1175 | if rf in self._video_formats_map: |
1176 | for srf in self._video_formats_map[rf]: | |
1177 | if srf in url_map: | |
1178 | video_url_list = [(srf, url_map[srf])] | |
1179 | break | |
1180 | else: | |
1181 | continue | |
1182 | break | |
1d043b93 JMF |
1183 | if video_url_list is None: |
1184 | raise ExtractorError(u'requested format not available') | |
1185 | return video_url_list | |
1186 | ||
1187 | def _extract_from_m3u8(self, manifest_url, video_id): | |
1188 | url_map = {} | |
1189 | def _get_urls(_manifest): | |
1190 | lines = _manifest.split('\n') | |
1191 | urls = filter(lambda l: l and not l.startswith('#'), | |
1192 | lines) | |
1193 | return urls | |
1194 | manifest = self._download_webpage(manifest_url, video_id, u'Downloading formats manifest') | |
1195 | formats_urls = _get_urls(manifest) | |
1196 | for format_url in formats_urls: | |
890f62e8 | 1197 | itag = self._search_regex(r'itag/(\d+?)/', format_url, 'itag') |
1d043b93 JMF |
1198 | url_map[itag] = format_url |
1199 | return url_map | |
1200 | ||
c5e8d7af | 1201 | def _real_extract(self, url): |
d7f44b5b PH |
1202 | if re.match(r'(?:https?://)?[^/]+/watch\?feature=[a-z_]+$', url): |
1203 | 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 ).') | |
1204 | ||
c5e8d7af PH |
1205 | # Extract original video URL from URL with redirection, like age verification, using next_url parameter |
1206 | mobj = re.search(self._NEXT_URL_RE, url) | |
1207 | if mobj: | |
1208 | url = 'https://www.youtube.com/' + compat_urllib_parse.unquote(mobj.group(1)).lstrip('/') | |
1209 | video_id = self._extract_id(url) | |
1210 | ||
1211 | # Get video webpage | |
1212 | self.report_video_webpage_download(video_id) | |
1213 | url = 'https://www.youtube.com/watch?v=%s&gl=US&hl=en&has_verified=1' % video_id | |
1214 | request = compat_urllib_request.Request(url) | |
1215 | try: | |
1216 | video_webpage_bytes = compat_urllib_request.urlopen(request).read() | |
1217 | except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err: | |
1218 | raise ExtractorError(u'Unable to download video webpage: %s' % compat_str(err)) | |
1219 | ||
1220 | video_webpage = video_webpage_bytes.decode('utf-8', 'ignore') | |
1221 | ||
1222 | # Attempt to extract SWF player URL | |
e0df6211 | 1223 | mobj = re.search(r'swfConfig.*?"(https?:\\/\\/.*?watch.*?-.*?\.swf)"', video_webpage) |
c5e8d7af PH |
1224 | if mobj is not None: |
1225 | player_url = re.sub(r'\\(.)', r'\1', mobj.group(1)) | |
1226 | else: | |
1227 | player_url = None | |
1228 | ||
1229 | # Get video info | |
1230 | self.report_video_info_webpage_download(video_id) | |
c108eb73 JMF |
1231 | if re.search(r'player-age-gate-content">', video_webpage) is not None: |
1232 | self.report_age_confirmation() | |
1233 | age_gate = True | |
1234 | # We simulate the access to the video from www.youtube.com/v/{video_id} | |
1235 | # this can be viewed without login into Youtube | |
1236 | data = compat_urllib_parse.urlencode({'video_id': video_id, | |
1237 | 'el': 'embedded', | |
1238 | 'gl': 'US', | |
1239 | 'hl': 'en', | |
1240 | 'eurl': 'https://youtube.googleapis.com/v/' + video_id, | |
1241 | 'asv': 3, | |
1242 | 'sts':'1588', | |
1243 | }) | |
1244 | video_info_url = 'https://www.youtube.com/get_video_info?' + data | |
c5e8d7af PH |
1245 | video_info_webpage = self._download_webpage(video_info_url, video_id, |
1246 | note=False, | |
1247 | errnote='unable to download video info webpage') | |
1248 | video_info = compat_parse_qs(video_info_webpage) | |
c108eb73 JMF |
1249 | else: |
1250 | age_gate = False | |
1251 | for el_type in ['&el=embedded', '&el=detailpage', '&el=vevo', '']: | |
1252 | video_info_url = ('https://www.youtube.com/get_video_info?&video_id=%s%s&ps=default&eurl=&gl=US&hl=en' | |
1253 | % (video_id, el_type)) | |
1254 | video_info_webpage = self._download_webpage(video_info_url, video_id, | |
1255 | note=False, | |
1256 | errnote='unable to download video info webpage') | |
1257 | video_info = compat_parse_qs(video_info_webpage) | |
1258 | if 'token' in video_info: | |
1259 | break | |
c5e8d7af PH |
1260 | if 'token' not in video_info: |
1261 | if 'reason' in video_info: | |
9a82b238 | 1262 | raise ExtractorError(u'YouTube said: %s' % video_info['reason'][0], expected=True) |
c5e8d7af PH |
1263 | else: |
1264 | raise ExtractorError(u'"token" parameter not in video info for unknown reason') | |
1265 | ||
1266 | # Check for "rental" videos | |
1267 | if 'ypc_video_rental_bar_text' in video_info and 'author' not in video_info: | |
1268 | raise ExtractorError(u'"rental" videos not supported') | |
1269 | ||
1270 | # Start extracting information | |
1271 | self.report_information_extraction(video_id) | |
1272 | ||
1273 | # uploader | |
1274 | if 'author' not in video_info: | |
1275 | raise ExtractorError(u'Unable to extract uploader name') | |
1276 | video_uploader = compat_urllib_parse.unquote_plus(video_info['author'][0]) | |
1277 | ||
1278 | # uploader_id | |
1279 | video_uploader_id = None | |
1280 | mobj = re.search(r'<link itemprop="url" href="http://www.youtube.com/(?:user|channel)/([^"]+)">', video_webpage) | |
1281 | if mobj is not None: | |
1282 | video_uploader_id = mobj.group(1) | |
1283 | else: | |
1284 | self._downloader.report_warning(u'unable to extract uploader nickname') | |
1285 | ||
1286 | # title | |
1287 | if 'title' not in video_info: | |
1288 | raise ExtractorError(u'Unable to extract video title') | |
1289 | video_title = compat_urllib_parse.unquote_plus(video_info['title'][0]) | |
1290 | ||
1291 | # thumbnail image | |
7763b04e JMF |
1292 | # We try first to get a high quality image: |
1293 | m_thumb = re.search(r'<span itemprop="thumbnail".*?href="(.*?)">', | |
1294 | video_webpage, re.DOTALL) | |
1295 | if m_thumb is not None: | |
1296 | video_thumbnail = m_thumb.group(1) | |
1297 | elif 'thumbnail_url' not in video_info: | |
c5e8d7af PH |
1298 | self._downloader.report_warning(u'unable to extract video thumbnail') |
1299 | video_thumbnail = '' | |
1300 | else: # don't panic if we can't find it | |
1301 | video_thumbnail = compat_urllib_parse.unquote_plus(video_info['thumbnail_url'][0]) | |
1302 | ||
1303 | # upload date | |
1304 | upload_date = None | |
1305 | mobj = re.search(r'id="eow-date.*?>(.*?)</span>', video_webpage, re.DOTALL) | |
1306 | if mobj is not None: | |
1307 | upload_date = ' '.join(re.sub(r'[/,-]', r' ', mobj.group(1)).split()) | |
1308 | upload_date = unified_strdate(upload_date) | |
1309 | ||
1310 | # description | |
1311 | video_description = get_element_by_id("eow-description", video_webpage) | |
1312 | if video_description: | |
1313 | video_description = clean_html(video_description) | |
1314 | else: | |
1315 | fd_mobj = re.search(r'<meta name="description" content="([^"]+)"', video_webpage) | |
1316 | if fd_mobj: | |
1317 | video_description = unescapeHTML(fd_mobj.group(1)) | |
1318 | else: | |
1319 | video_description = u'' | |
1320 | ||
1321 | # subtitles | |
d82134c3 | 1322 | video_subtitles = self.extract_subtitles(video_id, video_webpage) |
c5e8d7af | 1323 | |
c5e8d7af | 1324 | if self._downloader.params.get('listsubtitles', False): |
d665f8d3 | 1325 | self._list_available_subtitles(video_id, video_webpage) |
c5e8d7af PH |
1326 | return |
1327 | ||
1328 | if 'length_seconds' not in video_info: | |
1329 | self._downloader.report_warning(u'unable to extract video duration') | |
1330 | video_duration = '' | |
1331 | else: | |
1332 | video_duration = compat_urllib_parse.unquote_plus(video_info['length_seconds'][0]) | |
1333 | ||
c5e8d7af | 1334 | # Decide which formats to download |
c5e8d7af PH |
1335 | |
1336 | try: | |
1337 | mobj = re.search(r';ytplayer.config = ({.*?});', video_webpage) | |
50be92c1 PH |
1338 | if not mobj: |
1339 | raise ValueError('Could not find vevo ID') | |
c5e8d7af PH |
1340 | info = json.loads(mobj.group(1)) |
1341 | args = info['args'] | |
7ce7e394 JMF |
1342 | # Easy way to know if the 's' value is in url_encoded_fmt_stream_map |
1343 | # this signatures are encrypted | |
1344 | m_s = re.search(r'[&,]s=', args['url_encoded_fmt_stream_map']) | |
1345 | if m_s is not None: | |
1346 | self.to_screen(u'%s: Encrypted signatures detected.' % video_id) | |
c5e8d7af | 1347 | video_info['url_encoded_fmt_stream_map'] = [args['url_encoded_fmt_stream_map']] |
cde846b3 | 1348 | m_s = re.search(r'[&,]s=', args.get('adaptive_fmts', u'')) |
b7a68384 | 1349 | if m_s is not None: |
37b6d5f6 AZ |
1350 | if 'url_encoded_fmt_stream_map' in video_info: |
1351 | video_info['url_encoded_fmt_stream_map'][0] += ',' + args['adaptive_fmts'] | |
1352 | else: | |
1353 | video_info['url_encoded_fmt_stream_map'] = [args['adaptive_fmts']] | |
211fbc13 | 1354 | elif 'adaptive_fmts' in video_info: |
37b6d5f6 AZ |
1355 | if 'url_encoded_fmt_stream_map' in video_info: |
1356 | video_info['url_encoded_fmt_stream_map'][0] += ',' + video_info['adaptive_fmts'][0] | |
1357 | else: | |
1358 | video_info['url_encoded_fmt_stream_map'] = video_info['adaptive_fmts'] | |
c5e8d7af PH |
1359 | except ValueError: |
1360 | pass | |
1361 | ||
1362 | if 'conn' in video_info and video_info['conn'][0].startswith('rtmp'): | |
1363 | self.report_rtmp_download() | |
1364 | video_url_list = [(None, video_info['conn'][0])] | |
1365 | elif 'url_encoded_fmt_stream_map' in video_info and len(video_info['url_encoded_fmt_stream_map']) >= 1: | |
a7055eb9 JMF |
1366 | if 'rtmpe%3Dyes' in video_info['url_encoded_fmt_stream_map'][0]: |
1367 | raise ExtractorError('rtmpe downloads are not supported, see https://github.com/rg3/youtube-dl/issues/343 for more information.', expected=True) | |
c5e8d7af PH |
1368 | url_map = {} |
1369 | for url_data_str in video_info['url_encoded_fmt_stream_map'][0].split(','): | |
1370 | url_data = compat_parse_qs(url_data_str) | |
1371 | if 'itag' in url_data and 'url' in url_data: | |
1372 | url = url_data['url'][0] | |
1373 | if 'sig' in url_data: | |
1374 | url += '&signature=' + url_data['sig'][0] | |
1375 | elif 's' in url_data: | |
e0df6211 | 1376 | encrypted_sig = url_data['s'][0] |
769fda3c | 1377 | if self._downloader.params.get('verbose'): |
c108eb73 | 1378 | if age_gate: |
e0df6211 PH |
1379 | player_version = self._search_regex(r'-(.+)\.swf$', |
1380 | player_url if player_url else 'NOT FOUND', | |
1381 | 'flash player', fatal=False) | |
1382 | player_desc = 'flash player %s' % player_version | |
c108eb73 | 1383 | else: |
e0df6211 | 1384 | player_version = self._search_regex(r'html5player-(.+?)\.js', video_webpage, |
c108eb73 | 1385 | 'html5 player', fatal=False) |
e0df6211 PH |
1386 | player_desc = u'html5 player %s' % player_version |
1387 | ||
1388 | parts_sizes = u'.'.join(compat_str(len(part)) for part in encrypted_sig.split('.')) | |
5a76c651 | 1389 | self.to_screen(u'encrypted signature length %d (%s), itag %s, %s' % |
e0df6211 PH |
1390 | (len(encrypted_sig), parts_sizes, url_data['itag'][0], player_desc)) |
1391 | ||
75952c6e | 1392 | if age_gate: |
e0df6211 | 1393 | jsplayer_url = None |
75952c6e | 1394 | else: |
e0df6211 PH |
1395 | jsplayer_url_json = self._search_regex( |
1396 | r'"assets":.+?"js":\s*("[^"]+")', | |
1397 | video_webpage, u'JS player URL') | |
1398 | jsplayer_url = json.loads(jsplayer_url_json) | |
1399 | ||
1400 | signature = self._decrypt_signature(encrypted_sig, video_id, jsplayer_url, age_gate) | |
c5e8d7af PH |
1401 | url += '&signature=' + signature |
1402 | if 'ratebypass' not in url: | |
1403 | url += '&ratebypass=yes' | |
1404 | url_map[url_data['itag'][0]] = url | |
1d043b93 JMF |
1405 | video_url_list = self._get_video_url_list(url_map) |
1406 | if not video_url_list: | |
c5e8d7af | 1407 | return |
1d043b93 JMF |
1408 | elif video_info.get('hlsvp'): |
1409 | manifest_url = video_info['hlsvp'][0] | |
1410 | url_map = self._extract_from_m3u8(manifest_url, video_id) | |
1411 | video_url_list = self._get_video_url_list(url_map) | |
1412 | if not video_url_list: | |
1413 | return | |
1414 | ||
c5e8d7af PH |
1415 | else: |
1416 | raise ExtractorError(u'no conn or url_encoded_fmt_stream_map information found in video info') | |
1417 | ||
1418 | results = [] | |
1419 | for format_param, video_real_url in video_url_list: | |
1420 | # Extension | |
1421 | video_extension = self._video_extensions.get(format_param, 'flv') | |
1422 | ||
03cc7c20 JMF |
1423 | video_format = '{0} - {1}{2}'.format(format_param if format_param else video_extension, |
1424 | self._video_dimensions.get(format_param, '???'), | |
836a086c | 1425 | ' ('+self._special_itags[format_param]+')' if format_param in self._special_itags else '') |
c5e8d7af PH |
1426 | |
1427 | results.append({ | |
1428 | 'id': video_id, | |
1429 | 'url': video_real_url, | |
1430 | 'uploader': video_uploader, | |
1431 | 'uploader_id': video_uploader_id, | |
1432 | 'upload_date': upload_date, | |
1433 | 'title': video_title, | |
1434 | 'ext': video_extension, | |
1435 | 'format': video_format, | |
1436 | 'thumbnail': video_thumbnail, | |
1437 | 'description': video_description, | |
1438 | 'player_url': player_url, | |
1439 | 'subtitles': video_subtitles, | |
1440 | 'duration': video_duration | |
1441 | }) | |
1442 | return results | |
1443 | ||
1444 | class YoutubePlaylistIE(InfoExtractor): | |
0f818663 | 1445 | IE_DESC = u'YouTube.com playlists' |
c5e8d7af PH |
1446 | _VALID_URL = r"""(?: |
1447 | (?:https?://)? | |
1448 | (?:\w+\.)? | |
1449 | youtube\.com/ | |
1450 | (?: | |
1451 | (?:course|view_play_list|my_playlists|artist|playlist|watch) | |
1452 | \? (?:.*?&)*? (?:p|a|list)= | |
1453 | | p/ | |
1454 | ) | |
c626a3d9 | 1455 | ((?:PL|EC|UU|FL)?[0-9A-Za-z-_]{10,}) |
c5e8d7af PH |
1456 | .* |
1457 | | | |
c626a3d9 | 1458 | ((?:PL|EC|UU|FL)[0-9A-Za-z-_]{10,}) |
c5e8d7af PH |
1459 | )""" |
1460 | _TEMPLATE_URL = 'https://gdata.youtube.com/feeds/api/playlists/%s?max-results=%i&start-index=%i&v=2&alt=json&safeSearch=none' | |
1461 | _MAX_RESULTS = 50 | |
1462 | IE_NAME = u'youtube:playlist' | |
1463 | ||
1464 | @classmethod | |
1465 | def suitable(cls, url): | |
1466 | """Receives a URL and returns True if suitable for this IE.""" | |
1467 | return re.match(cls._VALID_URL, url, re.VERBOSE) is not None | |
1468 | ||
1469 | def _real_extract(self, url): | |
1470 | # Extract playlist id | |
1471 | mobj = re.match(self._VALID_URL, url, re.VERBOSE) | |
1472 | if mobj is None: | |
1473 | raise ExtractorError(u'Invalid URL: %s' % url) | |
1474 | ||
1475 | # Download playlist videos from API | |
1476 | playlist_id = mobj.group(1) or mobj.group(2) | |
c5e8d7af PH |
1477 | videos = [] |
1478 | ||
755eb032 | 1479 | for page_num in itertools.count(1): |
771822eb JMF |
1480 | start_index = self._MAX_RESULTS * (page_num - 1) + 1 |
1481 | if start_index >= 1000: | |
1482 | self._downloader.report_warning(u'Max number of results reached') | |
1483 | break | |
1484 | url = self._TEMPLATE_URL % (playlist_id, self._MAX_RESULTS, start_index) | |
c5e8d7af PH |
1485 | page = self._download_webpage(url, playlist_id, u'Downloading page #%s' % page_num) |
1486 | ||
1487 | try: | |
1488 | response = json.loads(page) | |
1489 | except ValueError as err: | |
1490 | raise ExtractorError(u'Invalid JSON in API response: ' + compat_str(err)) | |
1491 | ||
1492 | if 'feed' not in response: | |
1493 | raise ExtractorError(u'Got a malformed response from YouTube API') | |
1494 | playlist_title = response['feed']['title']['$t'] | |
1495 | if 'entry' not in response['feed']: | |
1496 | # Number of videos is a multiple of self._MAX_RESULTS | |
1497 | break | |
1498 | ||
1499 | for entry in response['feed']['entry']: | |
1500 | index = entry['yt$position']['$t'] | |
c215217e JMF |
1501 | if 'media$group' in entry and 'yt$videoid' in entry['media$group']: |
1502 | videos.append(( | |
1503 | index, | |
1504 | 'https://www.youtube.com/watch?v=' + entry['media$group']['yt$videoid']['$t'] | |
1505 | )) | |
c5e8d7af | 1506 | |
c5e8d7af PH |
1507 | videos = [v[1] for v in sorted(videos)] |
1508 | ||
20c3893f | 1509 | url_results = [self.url_result(vurl, 'Youtube') for vurl in videos] |
c5e8d7af PH |
1510 | return [self.playlist_result(url_results, playlist_id, playlist_title)] |
1511 | ||
1512 | ||
1513 | class YoutubeChannelIE(InfoExtractor): | |
0f818663 | 1514 | IE_DESC = u'YouTube.com channels' |
c5e8d7af PH |
1515 | _VALID_URL = r"^(?:https?://)?(?:youtu\.be|(?:\w+\.)?youtube(?:-nocookie)?\.com)/channel/([0-9A-Za-z_-]+)" |
1516 | _TEMPLATE_URL = 'http://www.youtube.com/channel/%s/videos?sort=da&flow=list&view=0&page=%s&gl=US&hl=en' | |
1517 | _MORE_PAGES_INDICATOR = 'yt-uix-load-more' | |
252580c5 | 1518 | _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 |
1519 | IE_NAME = u'youtube:channel' |
1520 | ||
1521 | def extract_videos_from_page(self, page): | |
1522 | ids_in_page = [] | |
1523 | for mobj in re.finditer(r'href="/watch\?v=([0-9A-Za-z_-]+)&?', page): | |
1524 | if mobj.group(1) not in ids_in_page: | |
1525 | ids_in_page.append(mobj.group(1)) | |
1526 | return ids_in_page | |
1527 | ||
1528 | def _real_extract(self, url): | |
1529 | # Extract channel id | |
1530 | mobj = re.match(self._VALID_URL, url) | |
1531 | if mobj is None: | |
1532 | raise ExtractorError(u'Invalid URL: %s' % url) | |
1533 | ||
1534 | # Download channel page | |
1535 | channel_id = mobj.group(1) | |
1536 | video_ids = [] | |
1537 | pagenum = 1 | |
1538 | ||
1539 | url = self._TEMPLATE_URL % (channel_id, pagenum) | |
1540 | page = self._download_webpage(url, channel_id, | |
1541 | u'Downloading page #%s' % pagenum) | |
1542 | ||
1543 | # Extract video identifiers | |
1544 | ids_in_page = self.extract_videos_from_page(page) | |
1545 | video_ids.extend(ids_in_page) | |
1546 | ||
1547 | # Download any subsequent channel pages using the json-based channel_ajax query | |
1548 | if self._MORE_PAGES_INDICATOR in page: | |
755eb032 | 1549 | for pagenum in itertools.count(1): |
c5e8d7af PH |
1550 | url = self._MORE_PAGES_URL % (pagenum, channel_id) |
1551 | page = self._download_webpage(url, channel_id, | |
1552 | u'Downloading page #%s' % pagenum) | |
1553 | ||
1554 | page = json.loads(page) | |
1555 | ||
1556 | ids_in_page = self.extract_videos_from_page(page['content_html']) | |
1557 | video_ids.extend(ids_in_page) | |
1558 | ||
1559 | if self._MORE_PAGES_INDICATOR not in page['load_more_widget_html']: | |
1560 | break | |
1561 | ||
1562 | self._downloader.to_screen(u'[youtube] Channel %s: Found %i videos' % (channel_id, len(video_ids))) | |
1563 | ||
1564 | urls = ['http://www.youtube.com/watch?v=%s' % id for id in video_ids] | |
20c3893f | 1565 | url_entries = [self.url_result(eurl, 'Youtube') for eurl in urls] |
c5e8d7af PH |
1566 | return [self.playlist_result(url_entries, channel_id)] |
1567 | ||
1568 | ||
1569 | class YoutubeUserIE(InfoExtractor): | |
0f818663 | 1570 | IE_DESC = u'YouTube.com user videos (URL or "ytuser" keyword)' |
faab1d38 | 1571 | _VALID_URL = r'(?:(?:(?:https?://)?(?:\w+\.)?youtube\.com/(?:user/)?)|ytuser:)(?!feed/)([A-Za-z0-9_-]+)' |
c5e8d7af PH |
1572 | _TEMPLATE_URL = 'http://gdata.youtube.com/feeds/api/users/%s' |
1573 | _GDATA_PAGE_SIZE = 50 | |
fd9cf738 | 1574 | _GDATA_URL = 'http://gdata.youtube.com/feeds/api/users/%s/uploads?max-results=%d&start-index=%d&alt=json' |
c5e8d7af PH |
1575 | IE_NAME = u'youtube:user' |
1576 | ||
e3ea4790 | 1577 | @classmethod |
f4b05232 | 1578 | def suitable(cls, url): |
e3ea4790 JMF |
1579 | # Don't return True if the url can be extracted with other youtube |
1580 | # extractor, the regex would is too permissive and it would match. | |
1581 | other_ies = iter(klass for (name, klass) in globals().items() if name.endswith('IE') and klass is not cls) | |
1582 | if any(ie.suitable(url) for ie in other_ies): return False | |
f4b05232 JMF |
1583 | else: return super(YoutubeUserIE, cls).suitable(url) |
1584 | ||
c5e8d7af PH |
1585 | def _real_extract(self, url): |
1586 | # Extract username | |
1587 | mobj = re.match(self._VALID_URL, url) | |
1588 | if mobj is None: | |
1589 | raise ExtractorError(u'Invalid URL: %s' % url) | |
1590 | ||
1591 | username = mobj.group(1) | |
1592 | ||
1593 | # Download video ids using YouTube Data API. Result size per | |
1594 | # query is limited (currently to 50 videos) so we need to query | |
1595 | # page by page until there are no video ids - it means we got | |
1596 | # all of them. | |
1597 | ||
1598 | video_ids = [] | |
c5e8d7af | 1599 | |
755eb032 | 1600 | for pagenum in itertools.count(0): |
c5e8d7af PH |
1601 | start_index = pagenum * self._GDATA_PAGE_SIZE + 1 |
1602 | ||
1603 | gdata_url = self._GDATA_URL % (username, self._GDATA_PAGE_SIZE, start_index) | |
1604 | page = self._download_webpage(gdata_url, username, | |
1605 | u'Downloading video ids from %d to %d' % (start_index, start_index + self._GDATA_PAGE_SIZE)) | |
1606 | ||
fd9cf738 JMF |
1607 | try: |
1608 | response = json.loads(page) | |
1609 | except ValueError as err: | |
1610 | raise ExtractorError(u'Invalid JSON in API response: ' + compat_str(err)) | |
71c82637 JMF |
1611 | if 'entry' not in response['feed']: |
1612 | # Number of videos is a multiple of self._MAX_RESULTS | |
1613 | break | |
fd9cf738 | 1614 | |
c5e8d7af PH |
1615 | # Extract video identifiers |
1616 | ids_in_page = [] | |
fd9cf738 JMF |
1617 | for entry in response['feed']['entry']: |
1618 | ids_in_page.append(entry['id']['$t'].split('/')[-1]) | |
c5e8d7af PH |
1619 | video_ids.extend(ids_in_page) |
1620 | ||
1621 | # A little optimization - if current page is not | |
1622 | # "full", ie. does not contain PAGE_SIZE video ids then | |
1623 | # we can assume that this page is the last one - there | |
1624 | # are no more ids on further pages - no need to query | |
1625 | # again. | |
1626 | ||
1627 | if len(ids_in_page) < self._GDATA_PAGE_SIZE: | |
1628 | break | |
1629 | ||
c5e8d7af | 1630 | urls = ['http://www.youtube.com/watch?v=%s' % video_id for video_id in video_ids] |
20c3893f | 1631 | url_results = [self.url_result(rurl, 'Youtube') for rurl in urls] |
c5e8d7af | 1632 | return [self.playlist_result(url_results, playlist_title = username)] |
b05654f0 PH |
1633 | |
1634 | class YoutubeSearchIE(SearchInfoExtractor): | |
0f818663 | 1635 | IE_DESC = u'YouTube.com searches' |
b05654f0 PH |
1636 | _API_URL = 'https://gdata.youtube.com/feeds/api/videos?q=%s&start-index=%i&max-results=50&v=2&alt=jsonc' |
1637 | _MAX_RESULTS = 1000 | |
1638 | IE_NAME = u'youtube:search' | |
1639 | _SEARCH_KEY = 'ytsearch' | |
1640 | ||
1641 | def report_download_page(self, query, pagenum): | |
1642 | """Report attempt to download search page with given number.""" | |
1643 | self._downloader.to_screen(u'[youtube] query "%s": Downloading page %s' % (query, pagenum)) | |
1644 | ||
1645 | def _get_n_results(self, query, n): | |
1646 | """Get a specified number of results for a query""" | |
1647 | ||
1648 | video_ids = [] | |
1649 | pagenum = 0 | |
1650 | limit = n | |
1651 | ||
1652 | while (50 * pagenum) < limit: | |
1653 | self.report_download_page(query, pagenum+1) | |
1654 | result_url = self._API_URL % (compat_urllib_parse.quote_plus(query), (50*pagenum)+1) | |
1655 | request = compat_urllib_request.Request(result_url) | |
1656 | try: | |
1657 | data = compat_urllib_request.urlopen(request).read().decode('utf-8') | |
1658 | except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err: | |
1659 | raise ExtractorError(u'Unable to download API page: %s' % compat_str(err)) | |
1660 | api_response = json.loads(data)['data'] | |
1661 | ||
1662 | if not 'items' in api_response: | |
1663 | raise ExtractorError(u'[youtube] No video results') | |
1664 | ||
1665 | new_ids = list(video['id'] for video in api_response['items']) | |
1666 | video_ids += new_ids | |
1667 | ||
1668 | limit = min(n, api_response['totalItems']) | |
1669 | pagenum += 1 | |
1670 | ||
1671 | if len(video_ids) > n: | |
1672 | video_ids = video_ids[:n] | |
1673 | videos = [self.url_result('http://www.youtube.com/watch?v=%s' % id, 'Youtube') for id in video_ids] | |
1674 | return self.playlist_result(videos, query) | |
75dff0ee JMF |
1675 | |
1676 | ||
1677 | class YoutubeShowIE(InfoExtractor): | |
0f818663 | 1678 | IE_DESC = u'YouTube.com (multi-season) shows' |
75dff0ee JMF |
1679 | _VALID_URL = r'https?://www\.youtube\.com/show/(.*)' |
1680 | IE_NAME = u'youtube:show' | |
1681 | ||
1682 | def _real_extract(self, url): | |
1683 | mobj = re.match(self._VALID_URL, url) | |
1684 | show_name = mobj.group(1) | |
1685 | webpage = self._download_webpage(url, show_name, u'Downloading show webpage') | |
1686 | # There's one playlist for each season of the show | |
1687 | m_seasons = list(re.finditer(r'href="(/playlist\?list=.*?)"', webpage)) | |
1688 | self.to_screen(u'%s: Found %s seasons' % (show_name, len(m_seasons))) | |
1689 | return [self.url_result('https://www.youtube.com' + season.group(1), 'YoutubePlaylist') for season in m_seasons] | |
04cc9617 JMF |
1690 | |
1691 | ||
b2e8bc1b | 1692 | class YoutubeFeedsInfoExtractor(YoutubeBaseInfoExtractor): |
d7ae0639 JMF |
1693 | """ |
1694 | Base class for extractors that fetch info from | |
1695 | http://www.youtube.com/feed_ajax | |
1696 | Subclasses must define the _FEED_NAME and _PLAYLIST_TITLE properties. | |
1697 | """ | |
b2e8bc1b | 1698 | _LOGIN_REQUIRED = True |
04cc9617 | 1699 | _PAGING_STEP = 30 |
43ba5456 JMF |
1700 | # use action_load_personal_feed instead of action_load_system_feed |
1701 | _PERSONAL_FEED = False | |
04cc9617 | 1702 | |
d7ae0639 JMF |
1703 | @property |
1704 | def _FEED_TEMPLATE(self): | |
43ba5456 JMF |
1705 | action = 'action_load_system_feed' |
1706 | if self._PERSONAL_FEED: | |
1707 | action = 'action_load_personal_feed' | |
1708 | return 'http://www.youtube.com/feed_ajax?%s=1&feed_name=%s&paging=%%s' % (action, self._FEED_NAME) | |
d7ae0639 JMF |
1709 | |
1710 | @property | |
1711 | def IE_NAME(self): | |
1712 | return u'youtube:%s' % self._FEED_NAME | |
04cc9617 | 1713 | |
81f0259b | 1714 | def _real_initialize(self): |
b2e8bc1b | 1715 | self._login() |
81f0259b | 1716 | |
04cc9617 JMF |
1717 | def _real_extract(self, url): |
1718 | feed_entries = [] | |
1719 | # The step argument is available only in 2.7 or higher | |
1720 | for i in itertools.count(0): | |
1721 | paging = i*self._PAGING_STEP | |
d7ae0639 JMF |
1722 | info = self._download_webpage(self._FEED_TEMPLATE % paging, |
1723 | u'%s feed' % self._FEED_NAME, | |
04cc9617 JMF |
1724 | u'Downloading page %s' % i) |
1725 | info = json.loads(info) | |
1726 | feed_html = info['feed_html'] | |
43ba5456 | 1727 | m_ids = re.finditer(r'"/watch\?v=(.*?)["&]', feed_html) |
04cc9617 JMF |
1728 | ids = orderedSet(m.group(1) for m in m_ids) |
1729 | feed_entries.extend(self.url_result(id, 'Youtube') for id in ids) | |
1730 | if info['paging'] is None: | |
1731 | break | |
d7ae0639 JMF |
1732 | return self.playlist_result(feed_entries, playlist_title=self._PLAYLIST_TITLE) |
1733 | ||
1734 | class YoutubeSubscriptionsIE(YoutubeFeedsInfoExtractor): | |
1735 | IE_DESC = u'YouTube.com subscriptions feed, "ytsubs" keyword(requires authentication)' | |
1736 | _VALID_URL = r'https?://www\.youtube\.com/feed/subscriptions|:ytsubs(?:criptions)?' | |
1737 | _FEED_NAME = 'subscriptions' | |
1738 | _PLAYLIST_TITLE = u'Youtube Subscriptions' | |
1739 | ||
1740 | class YoutubeRecommendedIE(YoutubeFeedsInfoExtractor): | |
1741 | IE_DESC = u'YouTube.com recommended videos, "ytrec" keyword (requires authentication)' | |
1742 | _VALID_URL = r'https?://www\.youtube\.com/feed/recommended|:ytrec(?:ommended)?' | |
1743 | _FEED_NAME = 'recommended' | |
1744 | _PLAYLIST_TITLE = u'Youtube Recommended videos' | |
c626a3d9 | 1745 | |
43ba5456 JMF |
1746 | class YoutubeWatchLaterIE(YoutubeFeedsInfoExtractor): |
1747 | IE_DESC = u'Youtube watch later list, "ytwatchlater" keyword (requires authentication)' | |
1748 | _VALID_URL = r'https?://www\.youtube\.com/feed/watch_later|:ytwatchlater' | |
1749 | _FEED_NAME = 'watch_later' | |
1750 | _PLAYLIST_TITLE = u'Youtube Watch Later' | |
1751 | _PAGING_STEP = 100 | |
1752 | _PERSONAL_FEED = True | |
c626a3d9 JMF |
1753 | |
1754 | class YoutubeFavouritesIE(YoutubeBaseInfoExtractor): | |
1755 | IE_NAME = u'youtube:favorites' | |
1756 | IE_DESC = u'YouTube.com favourite videos, "ytfav" keyword (requires authentication)' | |
c7a7750d | 1757 | _VALID_URL = r'https?://www\.youtube\.com/my_favorites|:ytfav(?:ou?rites)?' |
c626a3d9 JMF |
1758 | _LOGIN_REQUIRED = True |
1759 | ||
1760 | def _real_extract(self, url): | |
1761 | webpage = self._download_webpage('https://www.youtube.com/my_favorites', 'Youtube Favourites videos') | |
1762 | playlist_id = self._search_regex(r'list=(.+?)["&]', webpage, u'favourites playlist id') | |
1763 | return self.url_result(playlist_id, 'YoutubePlaylist') |