]>
Commit | Line | Data |
---|---|---|
1 | # coding: utf-8 | |
2 | ||
3 | import collections | |
4 | import errno | |
5 | import io | |
6 | import itertools | |
7 | import json | |
8 | import os.path | |
9 | import re | |
10 | import string | |
11 | import struct | |
12 | import traceback | |
13 | import zlib | |
14 | ||
15 | from .common import InfoExtractor, SearchInfoExtractor | |
16 | from .subtitles import SubtitlesInfoExtractor | |
17 | from ..utils import ( | |
18 | compat_chr, | |
19 | compat_parse_qs, | |
20 | compat_urllib_parse, | |
21 | compat_urllib_request, | |
22 | compat_urlparse, | |
23 | compat_str, | |
24 | ||
25 | clean_html, | |
26 | get_cachedir, | |
27 | get_element_by_id, | |
28 | get_element_by_attribute, | |
29 | ExtractorError, | |
30 | int_or_none, | |
31 | PagedList, | |
32 | RegexNotFoundError, | |
33 | unescapeHTML, | |
34 | unified_strdate, | |
35 | orderedSet, | |
36 | write_json_file, | |
37 | ) | |
38 | ||
39 | class YoutubeBaseInfoExtractor(InfoExtractor): | |
40 | """Provide base functions for Youtube extractors""" | |
41 | _LOGIN_URL = 'https://accounts.google.com/ServiceLogin' | |
42 | _LANG_URL = r'https://www.youtube.com/?hl=en&persist_hl=1&gl=US&persist_gl=1&opt_out_ackd=1' | |
43 | _AGE_URL = 'http://www.youtube.com/verify_age?next_url=/&gl=US&hl=en' | |
44 | _NETRC_MACHINE = 'youtube' | |
45 | # If True it will raise an error if no login info is provided | |
46 | _LOGIN_REQUIRED = False | |
47 | ||
48 | def _set_language(self): | |
49 | return bool(self._download_webpage( | |
50 | self._LANG_URL, None, | |
51 | note=u'Setting language', errnote='unable to set language', | |
52 | fatal=False)) | |
53 | ||
54 | def _login(self): | |
55 | (username, password) = self._get_login_info() | |
56 | # No authentication to be performed | |
57 | if username is None: | |
58 | if self._LOGIN_REQUIRED: | |
59 | raise ExtractorError(u'No login info available, needed for using %s.' % self.IE_NAME, expected=True) | |
60 | return False | |
61 | ||
62 | login_page = self._download_webpage( | |
63 | self._LOGIN_URL, None, | |
64 | note=u'Downloading login page', | |
65 | errnote=u'unable to fetch login page', fatal=False) | |
66 | if login_page is False: | |
67 | return | |
68 | ||
69 | galx = self._search_regex(r'(?s)<input.+?name="GALX".+?value="(.+?)"', | |
70 | login_page, u'Login GALX parameter') | |
71 | ||
72 | # Log in | |
73 | login_form_strs = { | |
74 | u'continue': u'https://www.youtube.com/signin?action_handle_signin=true&feature=sign_in_button&hl=en_US&nomobiletemp=1', | |
75 | u'Email': username, | |
76 | u'GALX': galx, | |
77 | u'Passwd': password, | |
78 | u'PersistentCookie': u'yes', | |
79 | u'_utf8': u'霱', | |
80 | u'bgresponse': u'js_disabled', | |
81 | u'checkConnection': u'', | |
82 | u'checkedDomains': u'youtube', | |
83 | u'dnConn': u'', | |
84 | u'pstMsg': u'0', | |
85 | u'rmShown': u'1', | |
86 | u'secTok': u'', | |
87 | u'signIn': u'Sign in', | |
88 | u'timeStmp': u'', | |
89 | u'service': u'youtube', | |
90 | u'uilel': u'3', | |
91 | u'hl': u'en_US', | |
92 | } | |
93 | # Convert to UTF-8 *before* urlencode because Python 2.x's urlencode | |
94 | # chokes on unicode | |
95 | login_form = dict((k.encode('utf-8'), v.encode('utf-8')) for k,v in login_form_strs.items()) | |
96 | login_data = compat_urllib_parse.urlencode(login_form).encode('ascii') | |
97 | ||
98 | req = compat_urllib_request.Request(self._LOGIN_URL, login_data) | |
99 | login_results = self._download_webpage( | |
100 | req, None, | |
101 | note=u'Logging in', errnote=u'unable to log in', fatal=False) | |
102 | if login_results is False: | |
103 | return False | |
104 | if re.search(r'(?i)<form[^>]* id="gaia_loginform"', login_results) is not None: | |
105 | self._downloader.report_warning(u'unable to log in: bad username or password') | |
106 | return False | |
107 | return True | |
108 | ||
109 | def _confirm_age(self): | |
110 | age_form = { | |
111 | 'next_url': '/', | |
112 | 'action_confirm': 'Confirm', | |
113 | } | |
114 | req = compat_urllib_request.Request(self._AGE_URL, compat_urllib_parse.urlencode(age_form)) | |
115 | ||
116 | self._download_webpage( | |
117 | req, None, | |
118 | note=u'Confirming age', errnote=u'Unable to confirm age') | |
119 | return True | |
120 | ||
121 | def _real_initialize(self): | |
122 | if self._downloader is None: | |
123 | return | |
124 | if not self._set_language(): | |
125 | return | |
126 | if not self._login(): | |
127 | return | |
128 | self._confirm_age() | |
129 | ||
130 | ||
131 | class YoutubeIE(YoutubeBaseInfoExtractor, SubtitlesInfoExtractor): | |
132 | IE_DESC = u'YouTube.com' | |
133 | _VALID_URL = r"""(?x)^ | |
134 | ( | |
135 | (?:https?://|//)? # http(s):// or protocol-independent URL (optional) | |
136 | (?:(?:(?:(?:\w+\.)?[yY][oO][uU][tT][uU][bB][eE](?:-nocookie)?\.com/| | |
137 | (?:www\.)?deturl\.com/www\.youtube\.com/| | |
138 | (?:www\.)?pwnyoutube\.com| | |
139 | tube\.majestyc\.net/| | |
140 | youtube\.googleapis\.com/) # the various hostnames, with wildcard subdomains | |
141 | (?:.*?\#/)? # handle anchor (#/) redirect urls | |
142 | (?: # the various things that can precede the ID: | |
143 | (?:(?:v|embed|e)/) # v/ or embed/ or e/ | |
144 | |(?: # or the v= param in all its forms | |
145 | (?:(?:watch|movie)(?:_popup)?(?:\.php)?)? # preceding watch(_popup|.php) or nothing (like /?v=xxxx) | |
146 | (?:\?|\#!?) # the params delimiter ? or # or #! | |
147 | (?:.*?&)? # any other preceding param (like /?s=tuff&v=xxxx) | |
148 | v= | |
149 | ) | |
150 | )) | |
151 | |youtu\.be/ # just youtu.be/xxxx | |
152 | ) | |
153 | )? # all until now is optional -> you can pass the naked ID | |
154 | ([0-9A-Za-z_-]{11}) # here is it! the YouTube video ID | |
155 | (?(1).+)? # if we found the ID, everything can follow | |
156 | $""" | |
157 | _NEXT_URL_RE = r'[\?&]next_url=([^&]+)' | |
158 | _formats = { | |
159 | '5': {'ext': 'flv', 'width': 400, 'height': 240}, | |
160 | '6': {'ext': 'flv', 'width': 450, 'height': 270}, | |
161 | '13': {'ext': '3gp'}, | |
162 | '17': {'ext': '3gp', 'width': 176, 'height': 144}, | |
163 | '18': {'ext': 'mp4', 'width': 640, 'height': 360}, | |
164 | '22': {'ext': 'mp4', 'width': 1280, 'height': 720}, | |
165 | '34': {'ext': 'flv', 'width': 640, 'height': 360}, | |
166 | '35': {'ext': 'flv', 'width': 854, 'height': 480}, | |
167 | '36': {'ext': '3gp', 'width': 320, 'height': 240}, | |
168 | '37': {'ext': 'mp4', 'width': 1920, 'height': 1080}, | |
169 | '38': {'ext': 'mp4', 'width': 4096, 'height': 3072}, | |
170 | '43': {'ext': 'webm', 'width': 640, 'height': 360}, | |
171 | '44': {'ext': 'webm', 'width': 854, 'height': 480}, | |
172 | '45': {'ext': 'webm', 'width': 1280, 'height': 720}, | |
173 | '46': {'ext': 'webm', 'width': 1920, 'height': 1080}, | |
174 | ||
175 | ||
176 | # 3d videos | |
177 | '82': {'ext': 'mp4', 'height': 360, 'resolution': '360p', 'format_note': '3D', 'preference': -20}, | |
178 | '83': {'ext': 'mp4', 'height': 480, 'resolution': '480p', 'format_note': '3D', 'preference': -20}, | |
179 | '84': {'ext': 'mp4', 'height': 720, 'resolution': '720p', 'format_note': '3D', 'preference': -20}, | |
180 | '85': {'ext': 'mp4', 'height': 1080, 'resolution': '1080p', 'format_note': '3D', 'preference': -20}, | |
181 | '100': {'ext': 'webm', 'height': 360, 'resolution': '360p', 'format_note': '3D', 'preference': -20}, | |
182 | '101': {'ext': 'webm', 'height': 480, 'resolution': '480p', 'format_note': '3D', 'preference': -20}, | |
183 | '102': {'ext': 'webm', 'height': 720, 'resolution': '720p', 'format_note': '3D', 'preference': -20}, | |
184 | ||
185 | # Apple HTTP Live Streaming | |
186 | '92': {'ext': 'mp4', 'height': 240, 'resolution': '240p', 'format_note': 'HLS', 'preference': -10}, | |
187 | '93': {'ext': 'mp4', 'height': 360, 'resolution': '360p', 'format_note': 'HLS', 'preference': -10}, | |
188 | '94': {'ext': 'mp4', 'height': 480, 'resolution': '480p', 'format_note': 'HLS', 'preference': -10}, | |
189 | '95': {'ext': 'mp4', 'height': 720, 'resolution': '720p', 'format_note': 'HLS', 'preference': -10}, | |
190 | '96': {'ext': 'mp4', 'height': 1080, 'resolution': '1080p', 'format_note': 'HLS', 'preference': -10}, | |
191 | '132': {'ext': 'mp4', 'height': 240, 'resolution': '240p', 'format_note': 'HLS', 'preference': -10}, | |
192 | '151': {'ext': 'mp4', 'height': 72, 'resolution': '72p', 'format_note': 'HLS', 'preference': -10}, | |
193 | ||
194 | # DASH mp4 video | |
195 | '133': {'ext': 'mp4', 'height': 240, 'resolution': '240p', 'format_note': 'DASH video', 'preference': -40}, | |
196 | '134': {'ext': 'mp4', 'height': 360, 'resolution': '360p', 'format_note': 'DASH video', 'preference': -40}, | |
197 | '135': {'ext': 'mp4', 'height': 480, 'resolution': '480p', 'format_note': 'DASH video', 'preference': -40}, | |
198 | '136': {'ext': 'mp4', 'height': 720, 'resolution': '720p', 'format_note': 'DASH video', 'preference': -40}, | |
199 | '137': {'ext': 'mp4', 'height': 1080, 'resolution': '1080p', 'format_note': 'DASH video', 'preference': -40}, | |
200 | '138': {'ext': 'mp4', 'height': 1081, 'resolution': '>1080p', 'format_note': 'DASH video', 'preference': -40}, | |
201 | '160': {'ext': 'mp4', 'height': 192, 'resolution': '192p', 'format_note': 'DASH video', 'preference': -40}, | |
202 | '264': {'ext': 'mp4', 'height': 1080, 'resolution': '1080p', 'format_note': 'DASH video', 'preference': -40}, | |
203 | ||
204 | # Dash mp4 audio | |
205 | '139': {'ext': 'm4a', 'format_note': 'DASH audio', 'vcodec': 'none', 'abr': 48, 'preference': -50}, | |
206 | '140': {'ext': 'm4a', 'format_note': 'DASH audio', 'vcodec': 'none', 'abr': 128, 'preference': -50}, | |
207 | '141': {'ext': 'm4a', 'format_note': 'DASH audio', 'vcodec': 'none', 'abr': 256, 'preference': -50}, | |
208 | ||
209 | # Dash webm | |
210 | '167': {'ext': 'webm', 'height': 360, 'width': 640, 'format_note': 'DASH video', 'container': 'webm', 'vcodec': 'VP8', 'acodec': 'none', 'preference': -40}, | |
211 | '168': {'ext': 'webm', 'height': 480, 'width': 854, 'format_note': 'DASH video', 'container': 'webm', 'vcodec': 'VP8', 'acodec': 'none', 'preference': -40}, | |
212 | '169': {'ext': 'webm', 'height': 720, 'width': 1280, 'format_note': 'DASH video', 'container': 'webm', 'vcodec': 'VP8', 'acodec': 'none', 'preference': -40}, | |
213 | '170': {'ext': 'webm', 'height': 1080, 'width': 1920, 'format_note': 'DASH video', 'container': 'webm', 'vcodec': 'VP8', 'acodec': 'none', 'preference': -40}, | |
214 | '218': {'ext': 'webm', 'height': 480, 'width': 854, 'format_note': 'DASH video', 'container': 'webm', 'vcodec': 'VP8', 'acodec': 'none', 'preference': -40}, | |
215 | '219': {'ext': 'webm', 'height': 480, 'width': 854, 'format_note': 'DASH video', 'container': 'webm', 'vcodec': 'VP8', 'acodec': 'none', 'preference': -40}, | |
216 | '242': {'ext': 'webm', 'height': 240, 'resolution': '240p', 'format_note': 'DASH webm', 'preference': -40}, | |
217 | '243': {'ext': 'webm', 'height': 360, 'resolution': '360p', 'format_note': 'DASH webm', 'preference': -40}, | |
218 | '244': {'ext': 'webm', 'height': 480, 'resolution': '480p', 'format_note': 'DASH webm', 'preference': -40}, | |
219 | '245': {'ext': 'webm', 'height': 480, 'resolution': '480p', 'format_note': 'DASH webm', 'preference': -40}, | |
220 | '246': {'ext': 'webm', 'height': 480, 'resolution': '480p', 'format_note': 'DASH webm', 'preference': -40}, | |
221 | '247': {'ext': 'webm', 'height': 720, 'resolution': '720p', 'format_note': 'DASH webm', 'preference': -40}, | |
222 | '248': {'ext': 'webm', 'height': 1080, 'resolution': '1080p', 'format_note': 'DASH webm', 'preference': -40}, | |
223 | ||
224 | # Dash webm audio | |
225 | '171': {'ext': 'webm', 'vcodec': 'none', 'format_note': 'DASH webm audio', 'abr': 48, 'preference': -50}, | |
226 | '172': {'ext': 'webm', 'vcodec': 'none', 'format_note': 'DASH webm audio', 'abr': 256, 'preference': -50}, | |
227 | ||
228 | # RTMP (unnamed) | |
229 | '_rtmp': {'protocol': 'rtmp'}, | |
230 | } | |
231 | ||
232 | IE_NAME = u'youtube' | |
233 | _TESTS = [ | |
234 | { | |
235 | u"url": u"http://www.youtube.com/watch?v=BaW_jenozKc", | |
236 | u"file": u"BaW_jenozKc.mp4", | |
237 | u"info_dict": { | |
238 | u"title": u"youtube-dl test video \"'/\\ä↭𝕐", | |
239 | u"uploader": u"Philipp Hagemeister", | |
240 | u"uploader_id": u"phihag", | |
241 | u"upload_date": u"20121002", | |
242 | u"description": u"test chars: \"'/\\ä↭𝕐\ntest URL: https://github.com/rg3/youtube-dl/issues/1892\n\nThis is a test video for youtube-dl.\n\nFor more information, contact phihag@phihag.de ." | |
243 | } | |
244 | }, | |
245 | { | |
246 | u"url": u"http://www.youtube.com/watch?v=UxxajLWwzqY", | |
247 | u"file": u"UxxajLWwzqY.mp4", | |
248 | u"note": u"Test generic use_cipher_signature video (#897)", | |
249 | u"info_dict": { | |
250 | u"upload_date": u"20120506", | |
251 | u"title": u"Icona Pop - I Love It (feat. Charli XCX) [OFFICIAL VIDEO]", | |
252 | u"description": u"md5:5b292926389560516e384ac437c0ec07", | |
253 | u"uploader": u"Icona Pop", | |
254 | u"uploader_id": u"IconaPop" | |
255 | } | |
256 | }, | |
257 | { | |
258 | u"url": u"https://www.youtube.com/watch?v=07FYdnEawAQ", | |
259 | u"file": u"07FYdnEawAQ.mp4", | |
260 | u"note": u"Test VEVO video with age protection (#956)", | |
261 | u"info_dict": { | |
262 | u"upload_date": u"20130703", | |
263 | u"title": u"Justin Timberlake - Tunnel Vision (Explicit)", | |
264 | u"description": u"md5:64249768eec3bc4276236606ea996373", | |
265 | u"uploader": u"justintimberlakeVEVO", | |
266 | u"uploader_id": u"justintimberlakeVEVO" | |
267 | } | |
268 | }, | |
269 | { | |
270 | u"url": u"//www.YouTube.com/watch?v=yZIXLfi8CZQ", | |
271 | u"file": u"yZIXLfi8CZQ.mp4", | |
272 | u"note": u"Embed-only video (#1746)", | |
273 | u"info_dict": { | |
274 | u"upload_date": u"20120608", | |
275 | u"title": u"Principal Sexually Assaults A Teacher - Episode 117 - 8th June 2012", | |
276 | u"description": u"md5:09b78bd971f1e3e289601dfba15ca4f7", | |
277 | u"uploader": u"SET India", | |
278 | u"uploader_id": u"setindia" | |
279 | } | |
280 | }, | |
281 | { | |
282 | u"url": u"http://www.youtube.com/watch?v=a9LDPn-MO4I", | |
283 | u"file": u"a9LDPn-MO4I.m4a", | |
284 | u"note": u"256k DASH audio (format 141) via DASH manifest", | |
285 | u"info_dict": { | |
286 | u"upload_date": "20121002", | |
287 | u"uploader_id": "8KVIDEO", | |
288 | u"description": "No description available.", | |
289 | u"uploader": "8KVIDEO", | |
290 | u"title": "UHDTV TEST 8K VIDEO.mp4" | |
291 | }, | |
292 | u"params": { | |
293 | u"youtube_include_dash_manifest": True, | |
294 | u"format": "141", | |
295 | }, | |
296 | }, | |
297 | ] | |
298 | ||
299 | ||
300 | @classmethod | |
301 | def suitable(cls, url): | |
302 | """Receives a URL and returns True if suitable for this IE.""" | |
303 | if YoutubePlaylistIE.suitable(url): return False | |
304 | return re.match(cls._VALID_URL, url) is not None | |
305 | ||
306 | def __init__(self, *args, **kwargs): | |
307 | super(YoutubeIE, self).__init__(*args, **kwargs) | |
308 | self._player_cache = {} | |
309 | ||
310 | def report_video_info_webpage_download(self, video_id): | |
311 | """Report attempt to download video info webpage.""" | |
312 | self.to_screen(u'%s: Downloading video info webpage' % video_id) | |
313 | ||
314 | def report_information_extraction(self, video_id): | |
315 | """Report attempt to extract video information.""" | |
316 | self.to_screen(u'%s: Extracting video information' % video_id) | |
317 | ||
318 | def report_unavailable_format(self, video_id, format): | |
319 | """Report extracted video URL.""" | |
320 | self.to_screen(u'%s: Format %s not available' % (video_id, format)) | |
321 | ||
322 | def report_rtmp_download(self): | |
323 | """Indicate the download will use the RTMP protocol.""" | |
324 | self.to_screen(u'RTMP download detected') | |
325 | ||
326 | def _extract_signature_function(self, video_id, player_url, slen): | |
327 | id_m = re.match(r'.*-(?P<id>[a-zA-Z0-9_-]+)\.(?P<ext>[a-z]+)$', | |
328 | player_url) | |
329 | player_type = id_m.group('ext') | |
330 | player_id = id_m.group('id') | |
331 | ||
332 | # Read from filesystem cache | |
333 | func_id = '%s_%s_%d' % (player_type, player_id, slen) | |
334 | assert os.path.basename(func_id) == func_id | |
335 | cache_dir = get_cachedir(self._downloader.params) | |
336 | ||
337 | cache_enabled = cache_dir is not None | |
338 | if cache_enabled: | |
339 | cache_fn = os.path.join(os.path.expanduser(cache_dir), | |
340 | u'youtube-sigfuncs', | |
341 | func_id + '.json') | |
342 | try: | |
343 | with io.open(cache_fn, 'r', encoding='utf-8') as cachef: | |
344 | cache_spec = json.load(cachef) | |
345 | return lambda s: u''.join(s[i] for i in cache_spec) | |
346 | except IOError: | |
347 | pass # No cache available | |
348 | ||
349 | if player_type == 'js': | |
350 | code = self._download_webpage( | |
351 | player_url, video_id, | |
352 | note=u'Downloading %s player %s' % (player_type, player_id), | |
353 | errnote=u'Download of %s failed' % player_url) | |
354 | res = self._parse_sig_js(code) | |
355 | elif player_type == 'swf': | |
356 | urlh = self._request_webpage( | |
357 | player_url, video_id, | |
358 | note=u'Downloading %s player %s' % (player_type, player_id), | |
359 | errnote=u'Download of %s failed' % player_url) | |
360 | code = urlh.read() | |
361 | res = self._parse_sig_swf(code) | |
362 | else: | |
363 | assert False, 'Invalid player type %r' % player_type | |
364 | ||
365 | if cache_enabled: | |
366 | try: | |
367 | test_string = u''.join(map(compat_chr, range(slen))) | |
368 | cache_res = res(test_string) | |
369 | cache_spec = [ord(c) for c in cache_res] | |
370 | try: | |
371 | os.makedirs(os.path.dirname(cache_fn)) | |
372 | except OSError as ose: | |
373 | if ose.errno != errno.EEXIST: | |
374 | raise | |
375 | write_json_file(cache_spec, cache_fn) | |
376 | except Exception: | |
377 | tb = traceback.format_exc() | |
378 | self._downloader.report_warning( | |
379 | u'Writing cache to %r failed: %s' % (cache_fn, tb)) | |
380 | ||
381 | return res | |
382 | ||
383 | def _print_sig_code(self, func, slen): | |
384 | def gen_sig_code(idxs): | |
385 | def _genslice(start, end, step): | |
386 | starts = u'' if start == 0 else str(start) | |
387 | ends = (u':%d' % (end+step)) if end + step >= 0 else u':' | |
388 | steps = u'' if step == 1 else (u':%d' % step) | |
389 | return u's[%s%s%s]' % (starts, ends, steps) | |
390 | ||
391 | step = None | |
392 | start = '(Never used)' # Quelch pyflakes warnings - start will be | |
393 | # set as soon as step is set | |
394 | for i, prev in zip(idxs[1:], idxs[:-1]): | |
395 | if step is not None: | |
396 | if i - prev == step: | |
397 | continue | |
398 | yield _genslice(start, prev, step) | |
399 | step = None | |
400 | continue | |
401 | if i - prev in [-1, 1]: | |
402 | step = i - prev | |
403 | start = prev | |
404 | continue | |
405 | else: | |
406 | yield u's[%d]' % prev | |
407 | if step is None: | |
408 | yield u's[%d]' % i | |
409 | else: | |
410 | yield _genslice(start, i, step) | |
411 | ||
412 | test_string = u''.join(map(compat_chr, range(slen))) | |
413 | cache_res = func(test_string) | |
414 | cache_spec = [ord(c) for c in cache_res] | |
415 | expr_code = u' + '.join(gen_sig_code(cache_spec)) | |
416 | code = u'if len(s) == %d:\n return %s\n' % (slen, expr_code) | |
417 | self.to_screen(u'Extracted signature function:\n' + code) | |
418 | ||
419 | def _parse_sig_js(self, jscode): | |
420 | funcname = self._search_regex( | |
421 | r'signature=([a-zA-Z]+)', jscode, | |
422 | u'Initial JS player signature function name') | |
423 | ||
424 | functions = {} | |
425 | ||
426 | def argidx(varname): | |
427 | return string.lowercase.index(varname) | |
428 | ||
429 | def interpret_statement(stmt, local_vars, allow_recursion=20): | |
430 | if allow_recursion < 0: | |
431 | raise ExtractorError(u'Recursion limit reached') | |
432 | ||
433 | if stmt.startswith(u'var '): | |
434 | stmt = stmt[len(u'var '):] | |
435 | ass_m = re.match(r'^(?P<out>[a-z]+)(?:\[(?P<index>[^\]]+)\])?' + | |
436 | r'=(?P<expr>.*)$', stmt) | |
437 | if ass_m: | |
438 | if ass_m.groupdict().get('index'): | |
439 | def assign(val): | |
440 | lvar = local_vars[ass_m.group('out')] | |
441 | idx = interpret_expression(ass_m.group('index'), | |
442 | local_vars, allow_recursion) | |
443 | assert isinstance(idx, int) | |
444 | lvar[idx] = val | |
445 | return val | |
446 | expr = ass_m.group('expr') | |
447 | else: | |
448 | def assign(val): | |
449 | local_vars[ass_m.group('out')] = val | |
450 | return val | |
451 | expr = ass_m.group('expr') | |
452 | elif stmt.startswith(u'return '): | |
453 | assign = lambda v: v | |
454 | expr = stmt[len(u'return '):] | |
455 | else: | |
456 | raise ExtractorError( | |
457 | u'Cannot determine left side of statement in %r' % stmt) | |
458 | ||
459 | v = interpret_expression(expr, local_vars, allow_recursion) | |
460 | return assign(v) | |
461 | ||
462 | def interpret_expression(expr, local_vars, allow_recursion): | |
463 | if expr.isdigit(): | |
464 | return int(expr) | |
465 | ||
466 | if expr.isalpha(): | |
467 | return local_vars[expr] | |
468 | ||
469 | m = re.match(r'^(?P<in>[a-z]+)\.(?P<member>.*)$', expr) | |
470 | if m: | |
471 | member = m.group('member') | |
472 | val = local_vars[m.group('in')] | |
473 | if member == 'split("")': | |
474 | return list(val) | |
475 | if member == 'join("")': | |
476 | return u''.join(val) | |
477 | if member == 'length': | |
478 | return len(val) | |
479 | if member == 'reverse()': | |
480 | return val[::-1] | |
481 | slice_m = re.match(r'slice\((?P<idx>.*)\)', member) | |
482 | if slice_m: | |
483 | idx = interpret_expression( | |
484 | slice_m.group('idx'), local_vars, allow_recursion-1) | |
485 | return val[idx:] | |
486 | ||
487 | m = re.match( | |
488 | r'^(?P<in>[a-z]+)\[(?P<idx>.+)\]$', expr) | |
489 | if m: | |
490 | val = local_vars[m.group('in')] | |
491 | idx = interpret_expression(m.group('idx'), local_vars, | |
492 | allow_recursion-1) | |
493 | return val[idx] | |
494 | ||
495 | m = re.match(r'^(?P<a>.+?)(?P<op>[%])(?P<b>.+?)$', expr) | |
496 | if m: | |
497 | a = interpret_expression(m.group('a'), | |
498 | local_vars, allow_recursion) | |
499 | b = interpret_expression(m.group('b'), | |
500 | local_vars, allow_recursion) | |
501 | return a % b | |
502 | ||
503 | m = re.match( | |
504 | r'^(?P<func>[a-zA-Z]+)\((?P<args>[a-z0-9,]+)\)$', expr) | |
505 | if m: | |
506 | fname = m.group('func') | |
507 | if fname not in functions: | |
508 | functions[fname] = extract_function(fname) | |
509 | argvals = [int(v) if v.isdigit() else local_vars[v] | |
510 | for v in m.group('args').split(',')] | |
511 | return functions[fname](argvals) | |
512 | raise ExtractorError(u'Unsupported JS expression %r' % expr) | |
513 | ||
514 | def extract_function(funcname): | |
515 | func_m = re.search( | |
516 | r'function ' + re.escape(funcname) + | |
517 | r'\((?P<args>[a-z,]+)\){(?P<code>[^}]+)}', | |
518 | jscode) | |
519 | argnames = func_m.group('args').split(',') | |
520 | ||
521 | def resf(args): | |
522 | local_vars = dict(zip(argnames, args)) | |
523 | for stmt in func_m.group('code').split(';'): | |
524 | res = interpret_statement(stmt, local_vars) | |
525 | return res | |
526 | return resf | |
527 | ||
528 | initial_function = extract_function(funcname) | |
529 | return lambda s: initial_function([s]) | |
530 | ||
531 | def _parse_sig_swf(self, file_contents): | |
532 | if file_contents[1:3] != b'WS': | |
533 | raise ExtractorError( | |
534 | u'Not an SWF file; header is %r' % file_contents[:3]) | |
535 | if file_contents[:1] == b'C': | |
536 | content = zlib.decompress(file_contents[8:]) | |
537 | else: | |
538 | raise NotImplementedError(u'Unsupported compression format %r' % | |
539 | file_contents[:1]) | |
540 | ||
541 | def extract_tags(content): | |
542 | pos = 0 | |
543 | while pos < len(content): | |
544 | header16 = struct.unpack('<H', content[pos:pos+2])[0] | |
545 | pos += 2 | |
546 | tag_code = header16 >> 6 | |
547 | tag_len = header16 & 0x3f | |
548 | if tag_len == 0x3f: | |
549 | tag_len = struct.unpack('<I', content[pos:pos+4])[0] | |
550 | pos += 4 | |
551 | assert pos+tag_len <= len(content) | |
552 | yield (tag_code, content[pos:pos+tag_len]) | |
553 | pos += tag_len | |
554 | ||
555 | code_tag = next(tag | |
556 | for tag_code, tag in extract_tags(content) | |
557 | if tag_code == 82) | |
558 | p = code_tag.index(b'\0', 4) + 1 | |
559 | code_reader = io.BytesIO(code_tag[p:]) | |
560 | ||
561 | # Parse ABC (AVM2 ByteCode) | |
562 | def read_int(reader=None): | |
563 | if reader is None: | |
564 | reader = code_reader | |
565 | res = 0 | |
566 | shift = 0 | |
567 | for _ in range(5): | |
568 | buf = reader.read(1) | |
569 | assert len(buf) == 1 | |
570 | b = struct.unpack('<B', buf)[0] | |
571 | res = res | ((b & 0x7f) << shift) | |
572 | if b & 0x80 == 0: | |
573 | break | |
574 | shift += 7 | |
575 | return res | |
576 | ||
577 | def u30(reader=None): | |
578 | res = read_int(reader) | |
579 | assert res & 0xf0000000 == 0 | |
580 | return res | |
581 | u32 = read_int | |
582 | ||
583 | def s32(reader=None): | |
584 | v = read_int(reader) | |
585 | if v & 0x80000000 != 0: | |
586 | v = - ((v ^ 0xffffffff) + 1) | |
587 | return v | |
588 | ||
589 | def read_string(reader=None): | |
590 | if reader is None: | |
591 | reader = code_reader | |
592 | slen = u30(reader) | |
593 | resb = reader.read(slen) | |
594 | assert len(resb) == slen | |
595 | return resb.decode('utf-8') | |
596 | ||
597 | def read_bytes(count, reader=None): | |
598 | if reader is None: | |
599 | reader = code_reader | |
600 | resb = reader.read(count) | |
601 | assert len(resb) == count | |
602 | return resb | |
603 | ||
604 | def read_byte(reader=None): | |
605 | resb = read_bytes(1, reader=reader) | |
606 | res = struct.unpack('<B', resb)[0] | |
607 | return res | |
608 | ||
609 | # minor_version + major_version | |
610 | read_bytes(2 + 2) | |
611 | ||
612 | # Constant pool | |
613 | int_count = u30() | |
614 | for _c in range(1, int_count): | |
615 | s32() | |
616 | uint_count = u30() | |
617 | for _c in range(1, uint_count): | |
618 | u32() | |
619 | double_count = u30() | |
620 | read_bytes((double_count-1) * 8) | |
621 | string_count = u30() | |
622 | constant_strings = [u''] | |
623 | for _c in range(1, string_count): | |
624 | s = read_string() | |
625 | constant_strings.append(s) | |
626 | namespace_count = u30() | |
627 | for _c in range(1, namespace_count): | |
628 | read_bytes(1) # kind | |
629 | u30() # name | |
630 | ns_set_count = u30() | |
631 | for _c in range(1, ns_set_count): | |
632 | count = u30() | |
633 | for _c2 in range(count): | |
634 | u30() | |
635 | multiname_count = u30() | |
636 | MULTINAME_SIZES = { | |
637 | 0x07: 2, # QName | |
638 | 0x0d: 2, # QNameA | |
639 | 0x0f: 1, # RTQName | |
640 | 0x10: 1, # RTQNameA | |
641 | 0x11: 0, # RTQNameL | |
642 | 0x12: 0, # RTQNameLA | |
643 | 0x09: 2, # Multiname | |
644 | 0x0e: 2, # MultinameA | |
645 | 0x1b: 1, # MultinameL | |
646 | 0x1c: 1, # MultinameLA | |
647 | } | |
648 | multinames = [u''] | |
649 | for _c in range(1, multiname_count): | |
650 | kind = u30() | |
651 | assert kind in MULTINAME_SIZES, u'Invalid multiname kind %r' % kind | |
652 | if kind == 0x07: | |
653 | u30() # namespace_idx | |
654 | name_idx = u30() | |
655 | multinames.append(constant_strings[name_idx]) | |
656 | else: | |
657 | multinames.append('[MULTINAME kind: %d]' % kind) | |
658 | for _c2 in range(MULTINAME_SIZES[kind]): | |
659 | u30() | |
660 | ||
661 | # Methods | |
662 | method_count = u30() | |
663 | MethodInfo = collections.namedtuple( | |
664 | 'MethodInfo', | |
665 | ['NEED_ARGUMENTS', 'NEED_REST']) | |
666 | method_infos = [] | |
667 | for method_id in range(method_count): | |
668 | param_count = u30() | |
669 | u30() # return type | |
670 | for _ in range(param_count): | |
671 | u30() # param type | |
672 | u30() # name index (always 0 for youtube) | |
673 | flags = read_byte() | |
674 | if flags & 0x08 != 0: | |
675 | # Options present | |
676 | option_count = u30() | |
677 | for c in range(option_count): | |
678 | u30() # val | |
679 | read_bytes(1) # kind | |
680 | if flags & 0x80 != 0: | |
681 | # Param names present | |
682 | for _ in range(param_count): | |
683 | u30() # param name | |
684 | mi = MethodInfo(flags & 0x01 != 0, flags & 0x04 != 0) | |
685 | method_infos.append(mi) | |
686 | ||
687 | # Metadata | |
688 | metadata_count = u30() | |
689 | for _c in range(metadata_count): | |
690 | u30() # name | |
691 | item_count = u30() | |
692 | for _c2 in range(item_count): | |
693 | u30() # key | |
694 | u30() # value | |
695 | ||
696 | def parse_traits_info(): | |
697 | trait_name_idx = u30() | |
698 | kind_full = read_byte() | |
699 | kind = kind_full & 0x0f | |
700 | attrs = kind_full >> 4 | |
701 | methods = {} | |
702 | if kind in [0x00, 0x06]: # Slot or Const | |
703 | u30() # Slot id | |
704 | u30() # type_name_idx | |
705 | vindex = u30() | |
706 | if vindex != 0: | |
707 | read_byte() # vkind | |
708 | elif kind in [0x01, 0x02, 0x03]: # Method / Getter / Setter | |
709 | u30() # disp_id | |
710 | method_idx = u30() | |
711 | methods[multinames[trait_name_idx]] = method_idx | |
712 | elif kind == 0x04: # Class | |
713 | u30() # slot_id | |
714 | u30() # classi | |
715 | elif kind == 0x05: # Function | |
716 | u30() # slot_id | |
717 | function_idx = u30() | |
718 | methods[function_idx] = multinames[trait_name_idx] | |
719 | else: | |
720 | raise ExtractorError(u'Unsupported trait kind %d' % kind) | |
721 | ||
722 | if attrs & 0x4 != 0: # Metadata present | |
723 | metadata_count = u30() | |
724 | for _c3 in range(metadata_count): | |
725 | u30() # metadata index | |
726 | ||
727 | return methods | |
728 | ||
729 | # Classes | |
730 | TARGET_CLASSNAME = u'SignatureDecipher' | |
731 | searched_idx = multinames.index(TARGET_CLASSNAME) | |
732 | searched_class_id = None | |
733 | class_count = u30() | |
734 | for class_id in range(class_count): | |
735 | name_idx = u30() | |
736 | if name_idx == searched_idx: | |
737 | # We found the class we're looking for! | |
738 | searched_class_id = class_id | |
739 | u30() # super_name idx | |
740 | flags = read_byte() | |
741 | if flags & 0x08 != 0: # Protected namespace is present | |
742 | u30() # protected_ns_idx | |
743 | intrf_count = u30() | |
744 | for _c2 in range(intrf_count): | |
745 | u30() | |
746 | u30() # iinit | |
747 | trait_count = u30() | |
748 | for _c2 in range(trait_count): | |
749 | parse_traits_info() | |
750 | ||
751 | if searched_class_id is None: | |
752 | raise ExtractorError(u'Target class %r not found' % | |
753 | TARGET_CLASSNAME) | |
754 | ||
755 | method_names = {} | |
756 | method_idxs = {} | |
757 | for class_id in range(class_count): | |
758 | u30() # cinit | |
759 | trait_count = u30() | |
760 | for _c2 in range(trait_count): | |
761 | trait_methods = parse_traits_info() | |
762 | if class_id == searched_class_id: | |
763 | method_names.update(trait_methods.items()) | |
764 | method_idxs.update(dict( | |
765 | (idx, name) | |
766 | for name, idx in trait_methods.items())) | |
767 | ||
768 | # Scripts | |
769 | script_count = u30() | |
770 | for _c in range(script_count): | |
771 | u30() # init | |
772 | trait_count = u30() | |
773 | for _c2 in range(trait_count): | |
774 | parse_traits_info() | |
775 | ||
776 | # Method bodies | |
777 | method_body_count = u30() | |
778 | Method = collections.namedtuple('Method', ['code', 'local_count']) | |
779 | methods = {} | |
780 | for _c in range(method_body_count): | |
781 | method_idx = u30() | |
782 | u30() # max_stack | |
783 | local_count = u30() | |
784 | u30() # init_scope_depth | |
785 | u30() # max_scope_depth | |
786 | code_length = u30() | |
787 | code = read_bytes(code_length) | |
788 | if method_idx in method_idxs: | |
789 | m = Method(code, local_count) | |
790 | methods[method_idxs[method_idx]] = m | |
791 | exception_count = u30() | |
792 | for _c2 in range(exception_count): | |
793 | u30() # from | |
794 | u30() # to | |
795 | u30() # target | |
796 | u30() # exc_type | |
797 | u30() # var_name | |
798 | trait_count = u30() | |
799 | for _c2 in range(trait_count): | |
800 | parse_traits_info() | |
801 | ||
802 | assert p + code_reader.tell() == len(code_tag) | |
803 | assert len(methods) == len(method_idxs) | |
804 | ||
805 | method_pyfunctions = {} | |
806 | ||
807 | def extract_function(func_name): | |
808 | if func_name in method_pyfunctions: | |
809 | return method_pyfunctions[func_name] | |
810 | if func_name not in methods: | |
811 | raise ExtractorError(u'Cannot find function %r' % func_name) | |
812 | m = methods[func_name] | |
813 | ||
814 | def resfunc(args): | |
815 | registers = ['(this)'] + list(args) + [None] * m.local_count | |
816 | stack = [] | |
817 | coder = io.BytesIO(m.code) | |
818 | while True: | |
819 | opcode = struct.unpack('!B', coder.read(1))[0] | |
820 | if opcode == 36: # pushbyte | |
821 | v = struct.unpack('!B', coder.read(1))[0] | |
822 | stack.append(v) | |
823 | elif opcode == 44: # pushstring | |
824 | idx = u30(coder) | |
825 | stack.append(constant_strings[idx]) | |
826 | elif opcode == 48: # pushscope | |
827 | # We don't implement the scope register, so we'll just | |
828 | # ignore the popped value | |
829 | stack.pop() | |
830 | elif opcode == 70: # callproperty | |
831 | index = u30(coder) | |
832 | mname = multinames[index] | |
833 | arg_count = u30(coder) | |
834 | args = list(reversed( | |
835 | [stack.pop() for _ in range(arg_count)])) | |
836 | obj = stack.pop() | |
837 | if mname == u'split': | |
838 | assert len(args) == 1 | |
839 | assert isinstance(args[0], compat_str) | |
840 | assert isinstance(obj, compat_str) | |
841 | if args[0] == u'': | |
842 | res = list(obj) | |
843 | else: | |
844 | res = obj.split(args[0]) | |
845 | stack.append(res) | |
846 | elif mname == u'slice': | |
847 | assert len(args) == 1 | |
848 | assert isinstance(args[0], int) | |
849 | assert isinstance(obj, list) | |
850 | res = obj[args[0]:] | |
851 | stack.append(res) | |
852 | elif mname == u'join': | |
853 | assert len(args) == 1 | |
854 | assert isinstance(args[0], compat_str) | |
855 | assert isinstance(obj, list) | |
856 | res = args[0].join(obj) | |
857 | stack.append(res) | |
858 | elif mname in method_pyfunctions: | |
859 | stack.append(method_pyfunctions[mname](args)) | |
860 | else: | |
861 | raise NotImplementedError( | |
862 | u'Unsupported property %r on %r' | |
863 | % (mname, obj)) | |
864 | elif opcode == 72: # returnvalue | |
865 | res = stack.pop() | |
866 | return res | |
867 | elif opcode == 79: # callpropvoid | |
868 | index = u30(coder) | |
869 | mname = multinames[index] | |
870 | arg_count = u30(coder) | |
871 | args = list(reversed( | |
872 | [stack.pop() for _ in range(arg_count)])) | |
873 | obj = stack.pop() | |
874 | if mname == u'reverse': | |
875 | assert isinstance(obj, list) | |
876 | obj.reverse() | |
877 | else: | |
878 | raise NotImplementedError( | |
879 | u'Unsupported (void) property %r on %r' | |
880 | % (mname, obj)) | |
881 | elif opcode == 93: # findpropstrict | |
882 | index = u30(coder) | |
883 | mname = multinames[index] | |
884 | res = extract_function(mname) | |
885 | stack.append(res) | |
886 | elif opcode == 97: # setproperty | |
887 | index = u30(coder) | |
888 | value = stack.pop() | |
889 | idx = stack.pop() | |
890 | obj = stack.pop() | |
891 | assert isinstance(obj, list) | |
892 | assert isinstance(idx, int) | |
893 | obj[idx] = value | |
894 | elif opcode == 98: # getlocal | |
895 | index = u30(coder) | |
896 | stack.append(registers[index]) | |
897 | elif opcode == 99: # setlocal | |
898 | index = u30(coder) | |
899 | value = stack.pop() | |
900 | registers[index] = value | |
901 | elif opcode == 102: # getproperty | |
902 | index = u30(coder) | |
903 | pname = multinames[index] | |
904 | if pname == u'length': | |
905 | obj = stack.pop() | |
906 | assert isinstance(obj, list) | |
907 | stack.append(len(obj)) | |
908 | else: # Assume attribute access | |
909 | idx = stack.pop() | |
910 | assert isinstance(idx, int) | |
911 | obj = stack.pop() | |
912 | assert isinstance(obj, list) | |
913 | stack.append(obj[idx]) | |
914 | elif opcode == 128: # coerce | |
915 | u30(coder) | |
916 | elif opcode == 133: # coerce_s | |
917 | assert isinstance(stack[-1], (type(None), compat_str)) | |
918 | elif opcode == 164: # modulo | |
919 | value2 = stack.pop() | |
920 | value1 = stack.pop() | |
921 | res = value1 % value2 | |
922 | stack.append(res) | |
923 | elif opcode == 208: # getlocal_0 | |
924 | stack.append(registers[0]) | |
925 | elif opcode == 209: # getlocal_1 | |
926 | stack.append(registers[1]) | |
927 | elif opcode == 210: # getlocal_2 | |
928 | stack.append(registers[2]) | |
929 | elif opcode == 211: # getlocal_3 | |
930 | stack.append(registers[3]) | |
931 | elif opcode == 214: # setlocal_2 | |
932 | registers[2] = stack.pop() | |
933 | elif opcode == 215: # setlocal_3 | |
934 | registers[3] = stack.pop() | |
935 | else: | |
936 | raise NotImplementedError( | |
937 | u'Unsupported opcode %d' % opcode) | |
938 | ||
939 | method_pyfunctions[func_name] = resfunc | |
940 | return resfunc | |
941 | ||
942 | initial_function = extract_function(u'decipher') | |
943 | return lambda s: initial_function([s]) | |
944 | ||
945 | def _decrypt_signature(self, s, video_id, player_url, age_gate=False): | |
946 | """Turn the encrypted s field into a working signature""" | |
947 | ||
948 | if player_url is not None: | |
949 | if player_url.startswith(u'//'): | |
950 | player_url = u'https:' + player_url | |
951 | try: | |
952 | player_id = (player_url, len(s)) | |
953 | if player_id not in self._player_cache: | |
954 | func = self._extract_signature_function( | |
955 | video_id, player_url, len(s) | |
956 | ) | |
957 | self._player_cache[player_id] = func | |
958 | func = self._player_cache[player_id] | |
959 | if self._downloader.params.get('youtube_print_sig_code'): | |
960 | self._print_sig_code(func, len(s)) | |
961 | return func(s) | |
962 | except Exception: | |
963 | tb = traceback.format_exc() | |
964 | self._downloader.report_warning( | |
965 | u'Automatic signature extraction failed: ' + tb) | |
966 | ||
967 | self._downloader.report_warning( | |
968 | u'Warning: Falling back to static signature algorithm') | |
969 | ||
970 | return self._static_decrypt_signature( | |
971 | s, video_id, player_url, age_gate) | |
972 | ||
973 | def _static_decrypt_signature(self, s, video_id, player_url, age_gate): | |
974 | if age_gate: | |
975 | # The videos with age protection use another player, so the | |
976 | # algorithms can be different. | |
977 | if len(s) == 86: | |
978 | return s[2:63] + s[82] + s[64:82] + s[63] | |
979 | ||
980 | if len(s) == 93: | |
981 | return s[86:29:-1] + s[88] + s[28:5:-1] | |
982 | elif len(s) == 92: | |
983 | return s[25] + s[3:25] + s[0] + s[26:42] + s[79] + s[43:79] + s[91] + s[80:83] | |
984 | elif len(s) == 91: | |
985 | return s[84:27:-1] + s[86] + s[26:5:-1] | |
986 | elif len(s) == 90: | |
987 | return s[25] + s[3:25] + s[2] + s[26:40] + s[77] + s[41:77] + s[89] + s[78:81] | |
988 | elif len(s) == 89: | |
989 | return s[84:78:-1] + s[87] + s[77:60:-1] + s[0] + s[59:3:-1] | |
990 | elif len(s) == 88: | |
991 | return s[7:28] + s[87] + s[29:45] + s[55] + s[46:55] + s[2] + s[56:87] + s[28] | |
992 | elif len(s) == 87: | |
993 | return s[6:27] + s[4] + s[28:39] + s[27] + s[40:59] + s[2] + s[60:] | |
994 | elif len(s) == 86: | |
995 | return s[80:72:-1] + s[16] + s[71:39:-1] + s[72] + s[38:16:-1] + s[82] + s[15::-1] | |
996 | elif len(s) == 85: | |
997 | return s[3:11] + s[0] + s[12:55] + s[84] + s[56:84] | |
998 | elif len(s) == 84: | |
999 | return s[78:70:-1] + s[14] + s[69:37:-1] + s[70] + s[36:14:-1] + s[80] + s[:14][::-1] | |
1000 | elif len(s) == 83: | |
1001 | return s[80:63:-1] + s[0] + s[62:0:-1] + s[63] | |
1002 | elif len(s) == 82: | |
1003 | return s[80:37:-1] + s[7] + s[36:7:-1] + s[0] + s[6:0:-1] + s[37] | |
1004 | elif len(s) == 81: | |
1005 | 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] | |
1006 | elif len(s) == 80: | |
1007 | return s[1:19] + s[0] + s[20:68] + s[19] + s[69:80] | |
1008 | elif len(s) == 79: | |
1009 | 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] | |
1010 | ||
1011 | else: | |
1012 | raise ExtractorError(u'Unable to decrypt signature, key length %d not supported; retrying might work' % (len(s))) | |
1013 | ||
1014 | def _get_available_subtitles(self, video_id, webpage): | |
1015 | try: | |
1016 | sub_list = self._download_webpage( | |
1017 | 'http://video.google.com/timedtext?hl=en&type=list&v=%s' % video_id, | |
1018 | video_id, note=False) | |
1019 | except ExtractorError as err: | |
1020 | self._downloader.report_warning(u'unable to download video subtitles: %s' % compat_str(err)) | |
1021 | return {} | |
1022 | lang_list = re.findall(r'name="([^"]*)"[^>]+lang_code="([\w\-]+)"', sub_list) | |
1023 | ||
1024 | sub_lang_list = {} | |
1025 | for l in lang_list: | |
1026 | lang = l[1] | |
1027 | params = compat_urllib_parse.urlencode({ | |
1028 | 'lang': lang, | |
1029 | 'v': video_id, | |
1030 | 'fmt': self._downloader.params.get('subtitlesformat', 'srt'), | |
1031 | 'name': unescapeHTML(l[0]).encode('utf-8'), | |
1032 | }) | |
1033 | url = u'http://www.youtube.com/api/timedtext?' + params | |
1034 | sub_lang_list[lang] = url | |
1035 | if not sub_lang_list: | |
1036 | self._downloader.report_warning(u'video doesn\'t have subtitles') | |
1037 | return {} | |
1038 | return sub_lang_list | |
1039 | ||
1040 | def _get_available_automatic_caption(self, video_id, webpage): | |
1041 | """We need the webpage for getting the captions url, pass it as an | |
1042 | argument to speed up the process.""" | |
1043 | sub_format = self._downloader.params.get('subtitlesformat', 'srt') | |
1044 | self.to_screen(u'%s: Looking for automatic captions' % video_id) | |
1045 | mobj = re.search(r';ytplayer.config = ({.*?});', webpage) | |
1046 | err_msg = u'Couldn\'t find automatic captions for %s' % video_id | |
1047 | if mobj is None: | |
1048 | self._downloader.report_warning(err_msg) | |
1049 | return {} | |
1050 | player_config = json.loads(mobj.group(1)) | |
1051 | try: | |
1052 | args = player_config[u'args'] | |
1053 | caption_url = args[u'ttsurl'] | |
1054 | timestamp = args[u'timestamp'] | |
1055 | # We get the available subtitles | |
1056 | list_params = compat_urllib_parse.urlencode({ | |
1057 | 'type': 'list', | |
1058 | 'tlangs': 1, | |
1059 | 'asrs': 1, | |
1060 | }) | |
1061 | list_url = caption_url + '&' + list_params | |
1062 | caption_list = self._download_xml(list_url, video_id) | |
1063 | original_lang_node = caption_list.find('track') | |
1064 | if original_lang_node is None or original_lang_node.attrib.get('kind') != 'asr' : | |
1065 | self._downloader.report_warning(u'Video doesn\'t have automatic captions') | |
1066 | return {} | |
1067 | original_lang = original_lang_node.attrib['lang_code'] | |
1068 | ||
1069 | sub_lang_list = {} | |
1070 | for lang_node in caption_list.findall('target'): | |
1071 | sub_lang = lang_node.attrib['lang_code'] | |
1072 | params = compat_urllib_parse.urlencode({ | |
1073 | 'lang': original_lang, | |
1074 | 'tlang': sub_lang, | |
1075 | 'fmt': sub_format, | |
1076 | 'ts': timestamp, | |
1077 | 'kind': 'asr', | |
1078 | }) | |
1079 | sub_lang_list[sub_lang] = caption_url + '&' + params | |
1080 | return sub_lang_list | |
1081 | # An extractor error can be raise by the download process if there are | |
1082 | # no automatic captions but there are subtitles | |
1083 | except (KeyError, ExtractorError): | |
1084 | self._downloader.report_warning(err_msg) | |
1085 | return {} | |
1086 | ||
1087 | def _extract_id(self, url): | |
1088 | mobj = re.match(self._VALID_URL, url, re.VERBOSE) | |
1089 | if mobj is None: | |
1090 | raise ExtractorError(u'Invalid URL: %s' % url) | |
1091 | video_id = mobj.group(2) | |
1092 | return video_id | |
1093 | ||
1094 | def _extract_from_m3u8(self, manifest_url, video_id): | |
1095 | url_map = {} | |
1096 | def _get_urls(_manifest): | |
1097 | lines = _manifest.split('\n') | |
1098 | urls = filter(lambda l: l and not l.startswith('#'), | |
1099 | lines) | |
1100 | return urls | |
1101 | manifest = self._download_webpage(manifest_url, video_id, u'Downloading formats manifest') | |
1102 | formats_urls = _get_urls(manifest) | |
1103 | for format_url in formats_urls: | |
1104 | itag = self._search_regex(r'itag/(\d+?)/', format_url, 'itag') | |
1105 | url_map[itag] = format_url | |
1106 | return url_map | |
1107 | ||
1108 | def _extract_annotations(self, video_id): | |
1109 | url = 'https://www.youtube.com/annotations_invideo?features=1&legacy=1&video_id=%s' % video_id | |
1110 | return self._download_webpage(url, video_id, note=u'Searching for annotations.', errnote=u'Unable to download video annotations.') | |
1111 | ||
1112 | def _real_extract(self, url): | |
1113 | # Extract original video URL from URL with redirection, like age verification, using next_url parameter | |
1114 | mobj = re.search(self._NEXT_URL_RE, url) | |
1115 | if mobj: | |
1116 | url = 'https://www.youtube.com/' + compat_urllib_parse.unquote(mobj.group(1)).lstrip('/') | |
1117 | video_id = self._extract_id(url) | |
1118 | ||
1119 | # Get video webpage | |
1120 | url = 'https://www.youtube.com/watch?v=%s&gl=US&hl=en&has_verified=1' % video_id | |
1121 | video_webpage = self._download_webpage(url, video_id) | |
1122 | ||
1123 | # Attempt to extract SWF player URL | |
1124 | mobj = re.search(r'swfConfig.*?"(https?:\\/\\/.*?watch.*?-.*?\.swf)"', video_webpage) | |
1125 | if mobj is not None: | |
1126 | player_url = re.sub(r'\\(.)', r'\1', mobj.group(1)) | |
1127 | else: | |
1128 | player_url = None | |
1129 | ||
1130 | # Get video info | |
1131 | self.report_video_info_webpage_download(video_id) | |
1132 | if re.search(r'player-age-gate-content">', video_webpage) is not None: | |
1133 | self.report_age_confirmation() | |
1134 | age_gate = True | |
1135 | # We simulate the access to the video from www.youtube.com/v/{video_id} | |
1136 | # this can be viewed without login into Youtube | |
1137 | data = compat_urllib_parse.urlencode({'video_id': video_id, | |
1138 | 'el': 'player_embedded', | |
1139 | 'gl': 'US', | |
1140 | 'hl': 'en', | |
1141 | 'eurl': 'https://youtube.googleapis.com/v/' + video_id, | |
1142 | 'asv': 3, | |
1143 | 'sts':'1588', | |
1144 | }) | |
1145 | video_info_url = 'https://www.youtube.com/get_video_info?' + data | |
1146 | video_info_webpage = self._download_webpage(video_info_url, video_id, | |
1147 | note=False, | |
1148 | errnote='unable to download video info webpage') | |
1149 | video_info = compat_parse_qs(video_info_webpage) | |
1150 | else: | |
1151 | age_gate = False | |
1152 | for el_type in ['&el=embedded', '&el=detailpage', '&el=vevo', '']: | |
1153 | video_info_url = ('https://www.youtube.com/get_video_info?&video_id=%s%s&ps=default&eurl=&gl=US&hl=en' | |
1154 | % (video_id, el_type)) | |
1155 | video_info_webpage = self._download_webpage(video_info_url, video_id, | |
1156 | note=False, | |
1157 | errnote='unable to download video info webpage') | |
1158 | video_info = compat_parse_qs(video_info_webpage) | |
1159 | if 'token' in video_info: | |
1160 | break | |
1161 | if 'token' not in video_info: | |
1162 | if 'reason' in video_info: | |
1163 | raise ExtractorError(u'YouTube said: %s' % video_info['reason'][0], expected=True) | |
1164 | else: | |
1165 | raise ExtractorError(u'"token" parameter not in video info for unknown reason') | |
1166 | ||
1167 | if 'view_count' in video_info: | |
1168 | view_count = int(video_info['view_count'][0]) | |
1169 | else: | |
1170 | view_count = None | |
1171 | ||
1172 | # Check for "rental" videos | |
1173 | if 'ypc_video_rental_bar_text' in video_info and 'author' not in video_info: | |
1174 | raise ExtractorError(u'"rental" videos not supported') | |
1175 | ||
1176 | # Start extracting information | |
1177 | self.report_information_extraction(video_id) | |
1178 | ||
1179 | # uploader | |
1180 | if 'author' not in video_info: | |
1181 | raise ExtractorError(u'Unable to extract uploader name') | |
1182 | video_uploader = compat_urllib_parse.unquote_plus(video_info['author'][0]) | |
1183 | ||
1184 | # uploader_id | |
1185 | video_uploader_id = None | |
1186 | mobj = re.search(r'<link itemprop="url" href="http://www.youtube.com/(?:user|channel)/([^"]+)">', video_webpage) | |
1187 | if mobj is not None: | |
1188 | video_uploader_id = mobj.group(1) | |
1189 | else: | |
1190 | self._downloader.report_warning(u'unable to extract uploader nickname') | |
1191 | ||
1192 | # title | |
1193 | if 'title' in video_info: | |
1194 | video_title = compat_urllib_parse.unquote_plus(video_info['title'][0]) | |
1195 | else: | |
1196 | self._downloader.report_warning(u'Unable to extract video title') | |
1197 | video_title = u'_' | |
1198 | ||
1199 | # thumbnail image | |
1200 | # We try first to get a high quality image: | |
1201 | m_thumb = re.search(r'<span itemprop="thumbnail".*?href="(.*?)">', | |
1202 | video_webpage, re.DOTALL) | |
1203 | if m_thumb is not None: | |
1204 | video_thumbnail = m_thumb.group(1) | |
1205 | elif 'thumbnail_url' not in video_info: | |
1206 | self._downloader.report_warning(u'unable to extract video thumbnail') | |
1207 | video_thumbnail = None | |
1208 | else: # don't panic if we can't find it | |
1209 | video_thumbnail = compat_urllib_parse.unquote_plus(video_info['thumbnail_url'][0]) | |
1210 | ||
1211 | # upload date | |
1212 | upload_date = None | |
1213 | mobj = re.search(r'id="eow-date.*?>(.*?)</span>', video_webpage, re.DOTALL) | |
1214 | if mobj is not None: | |
1215 | upload_date = ' '.join(re.sub(r'[/,-]', r' ', mobj.group(1)).split()) | |
1216 | upload_date = unified_strdate(upload_date) | |
1217 | ||
1218 | # description | |
1219 | video_description = get_element_by_id("eow-description", video_webpage) | |
1220 | if video_description: | |
1221 | video_description = re.sub(r'''(?x) | |
1222 | <a\s+ | |
1223 | (?:[a-zA-Z-]+="[^"]+"\s+)*? | |
1224 | title="([^"]+)"\s+ | |
1225 | (?:[a-zA-Z-]+="[^"]+"\s+)*? | |
1226 | class="yt-uix-redirect-link"\s*> | |
1227 | [^<]+ | |
1228 | </a> | |
1229 | ''', r'\1', video_description) | |
1230 | video_description = clean_html(video_description) | |
1231 | else: | |
1232 | fd_mobj = re.search(r'<meta name="description" content="([^"]+)"', video_webpage) | |
1233 | if fd_mobj: | |
1234 | video_description = unescapeHTML(fd_mobj.group(1)) | |
1235 | else: | |
1236 | video_description = u'' | |
1237 | ||
1238 | def _extract_count(klass): | |
1239 | count = self._search_regex( | |
1240 | r'class="%s">([\d,]+)</span>' % re.escape(klass), | |
1241 | video_webpage, klass, default=None) | |
1242 | if count is not None: | |
1243 | return int(count.replace(',', '')) | |
1244 | return None | |
1245 | like_count = _extract_count(u'likes-count') | |
1246 | dislike_count = _extract_count(u'dislikes-count') | |
1247 | ||
1248 | # subtitles | |
1249 | video_subtitles = self.extract_subtitles(video_id, video_webpage) | |
1250 | ||
1251 | if self._downloader.params.get('listsubtitles', False): | |
1252 | self._list_available_subtitles(video_id, video_webpage) | |
1253 | return | |
1254 | ||
1255 | if 'length_seconds' not in video_info: | |
1256 | self._downloader.report_warning(u'unable to extract video duration') | |
1257 | video_duration = None | |
1258 | else: | |
1259 | video_duration = int(compat_urllib_parse.unquote_plus(video_info['length_seconds'][0])) | |
1260 | ||
1261 | # annotations | |
1262 | video_annotations = None | |
1263 | if self._downloader.params.get('writeannotations', False): | |
1264 | video_annotations = self._extract_annotations(video_id) | |
1265 | ||
1266 | # Decide which formats to download | |
1267 | try: | |
1268 | mobj = re.search(r';ytplayer.config = ({.*?});', video_webpage) | |
1269 | if not mobj: | |
1270 | raise ValueError('Could not find vevo ID') | |
1271 | info = json.loads(mobj.group(1)) | |
1272 | args = info['args'] | |
1273 | # Easy way to know if the 's' value is in url_encoded_fmt_stream_map | |
1274 | # this signatures are encrypted | |
1275 | if 'url_encoded_fmt_stream_map' not in args: | |
1276 | raise ValueError(u'No stream_map present') # caught below | |
1277 | re_signature = re.compile(r'[&,]s=') | |
1278 | m_s = re_signature.search(args['url_encoded_fmt_stream_map']) | |
1279 | if m_s is not None: | |
1280 | self.to_screen(u'%s: Encrypted signatures detected.' % video_id) | |
1281 | video_info['url_encoded_fmt_stream_map'] = [args['url_encoded_fmt_stream_map']] | |
1282 | m_s = re_signature.search(args.get('adaptive_fmts', u'')) | |
1283 | if m_s is not None: | |
1284 | if 'adaptive_fmts' in video_info: | |
1285 | video_info['adaptive_fmts'][0] += ',' + args['adaptive_fmts'] | |
1286 | else: | |
1287 | video_info['adaptive_fmts'] = [args['adaptive_fmts']] | |
1288 | except ValueError: | |
1289 | pass | |
1290 | ||
1291 | def _map_to_format_list(urlmap): | |
1292 | formats = [] | |
1293 | for itag, video_real_url in urlmap.items(): | |
1294 | dct = { | |
1295 | 'format_id': itag, | |
1296 | 'url': video_real_url, | |
1297 | 'player_url': player_url, | |
1298 | } | |
1299 | if itag in self._formats: | |
1300 | dct.update(self._formats[itag]) | |
1301 | formats.append(dct) | |
1302 | return formats | |
1303 | ||
1304 | if 'conn' in video_info and video_info['conn'][0].startswith('rtmp'): | |
1305 | self.report_rtmp_download() | |
1306 | formats = [{ | |
1307 | 'format_id': '_rtmp', | |
1308 | 'protocol': 'rtmp', | |
1309 | 'url': video_info['conn'][0], | |
1310 | 'player_url': player_url, | |
1311 | }] | |
1312 | elif len(video_info.get('url_encoded_fmt_stream_map', [])) >= 1 or len(video_info.get('adaptive_fmts', [])) >= 1: | |
1313 | encoded_url_map = video_info.get('url_encoded_fmt_stream_map', [''])[0] + ',' + video_info.get('adaptive_fmts',[''])[0] | |
1314 | if 'rtmpe%3Dyes' in encoded_url_map: | |
1315 | raise ExtractorError('rtmpe downloads are not supported, see https://github.com/rg3/youtube-dl/issues/343 for more information.', expected=True) | |
1316 | url_map = {} | |
1317 | for url_data_str in encoded_url_map.split(','): | |
1318 | url_data = compat_parse_qs(url_data_str) | |
1319 | if 'itag' in url_data and 'url' in url_data: | |
1320 | url = url_data['url'][0] | |
1321 | if 'sig' in url_data: | |
1322 | url += '&signature=' + url_data['sig'][0] | |
1323 | elif 's' in url_data: | |
1324 | encrypted_sig = url_data['s'][0] | |
1325 | if self._downloader.params.get('verbose'): | |
1326 | if age_gate: | |
1327 | if player_url is None: | |
1328 | player_version = 'unknown' | |
1329 | else: | |
1330 | player_version = self._search_regex( | |
1331 | r'-(.+)\.swf$', player_url, | |
1332 | u'flash player', fatal=False) | |
1333 | player_desc = 'flash player %s' % player_version | |
1334 | else: | |
1335 | player_version = self._search_regex( | |
1336 | r'html5player-(.+?)\.js', video_webpage, | |
1337 | 'html5 player', fatal=False) | |
1338 | player_desc = u'html5 player %s' % player_version | |
1339 | ||
1340 | parts_sizes = u'.'.join(compat_str(len(part)) for part in encrypted_sig.split('.')) | |
1341 | self.to_screen(u'encrypted signature length %d (%s), itag %s, %s' % | |
1342 | (len(encrypted_sig), parts_sizes, url_data['itag'][0], player_desc)) | |
1343 | ||
1344 | if not age_gate: | |
1345 | jsplayer_url_json = self._search_regex( | |
1346 | r'"assets":.+?"js":\s*("[^"]+")', | |
1347 | video_webpage, u'JS player URL') | |
1348 | player_url = json.loads(jsplayer_url_json) | |
1349 | ||
1350 | signature = self._decrypt_signature( | |
1351 | encrypted_sig, video_id, player_url, age_gate) | |
1352 | url += '&signature=' + signature | |
1353 | if 'ratebypass' not in url: | |
1354 | url += '&ratebypass=yes' | |
1355 | url_map[url_data['itag'][0]] = url | |
1356 | formats = _map_to_format_list(url_map) | |
1357 | elif video_info.get('hlsvp'): | |
1358 | manifest_url = video_info['hlsvp'][0] | |
1359 | url_map = self._extract_from_m3u8(manifest_url, video_id) | |
1360 | formats = _map_to_format_list(url_map) | |
1361 | else: | |
1362 | raise ExtractorError(u'no conn, hlsvp or url_encoded_fmt_stream_map information found in video info') | |
1363 | ||
1364 | # Look for the DASH manifest | |
1365 | dash_manifest_url_lst = video_info.get('dashmpd') | |
1366 | if (dash_manifest_url_lst and dash_manifest_url_lst[0] and | |
1367 | self._downloader.params.get('youtube_include_dash_manifest', False)): | |
1368 | try: | |
1369 | dash_doc = self._download_xml( | |
1370 | dash_manifest_url_lst[0], video_id, | |
1371 | note=u'Downloading DASH manifest', | |
1372 | errnote=u'Could not download DASH manifest') | |
1373 | for r in dash_doc.findall(u'.//{urn:mpeg:DASH:schema:MPD:2011}Representation'): | |
1374 | url_el = r.find('{urn:mpeg:DASH:schema:MPD:2011}BaseURL') | |
1375 | if url_el is None: | |
1376 | continue | |
1377 | format_id = r.attrib['id'] | |
1378 | video_url = url_el.text | |
1379 | filesize = int_or_none(url_el.attrib.get('{http://youtube.com/yt/2012/10/10}contentLength')) | |
1380 | f = { | |
1381 | 'format_id': format_id, | |
1382 | 'url': video_url, | |
1383 | 'width': int_or_none(r.attrib.get('width')), | |
1384 | 'tbr': int_or_none(r.attrib.get('bandwidth'), 1000), | |
1385 | 'asr': int_or_none(r.attrib.get('audioSamplingRate')), | |
1386 | 'filesize': filesize, | |
1387 | } | |
1388 | try: | |
1389 | existing_format = next( | |
1390 | fo for fo in formats | |
1391 | if fo['format_id'] == format_id) | |
1392 | except StopIteration: | |
1393 | f.update(self._formats.get(format_id, {})) | |
1394 | formats.append(f) | |
1395 | else: | |
1396 | existing_format.update(f) | |
1397 | ||
1398 | except (ExtractorError, KeyError) as e: | |
1399 | self.report_warning(u'Skipping DASH manifest: %s' % e, video_id) | |
1400 | ||
1401 | self._sort_formats(formats) | |
1402 | ||
1403 | return { | |
1404 | 'id': video_id, | |
1405 | 'uploader': video_uploader, | |
1406 | 'uploader_id': video_uploader_id, | |
1407 | 'upload_date': upload_date, | |
1408 | 'title': video_title, | |
1409 | 'thumbnail': video_thumbnail, | |
1410 | 'description': video_description, | |
1411 | 'subtitles': video_subtitles, | |
1412 | 'duration': video_duration, | |
1413 | 'age_limit': 18 if age_gate else 0, | |
1414 | 'annotations': video_annotations, | |
1415 | 'webpage_url': 'https://www.youtube.com/watch?v=%s' % video_id, | |
1416 | 'view_count': view_count, | |
1417 | 'like_count': like_count, | |
1418 | 'dislike_count': dislike_count, | |
1419 | 'formats': formats, | |
1420 | } | |
1421 | ||
1422 | class YoutubePlaylistIE(YoutubeBaseInfoExtractor): | |
1423 | IE_DESC = u'YouTube.com playlists' | |
1424 | _VALID_URL = r"""(?: | |
1425 | (?:https?://)? | |
1426 | (?:\w+\.)? | |
1427 | youtube\.com/ | |
1428 | (?: | |
1429 | (?:course|view_play_list|my_playlists|artist|playlist|watch) | |
1430 | \? (?:.*?&)*? (?:p|a|list)= | |
1431 | | p/ | |
1432 | ) | |
1433 | ((?:PL|EC|UU|FL|RD)?[0-9A-Za-z-_]{10,}) | |
1434 | .* | |
1435 | | | |
1436 | ((?:PL|EC|UU|FL|RD)[0-9A-Za-z-_]{10,}) | |
1437 | )""" | |
1438 | _TEMPLATE_URL = 'https://www.youtube.com/playlist?list=%s&page=%s' | |
1439 | _MORE_PAGES_INDICATOR = r'data-link-type="next"' | |
1440 | _VIDEO_RE = r'href="/watch\?v=(?P<id>[0-9A-Za-z_-]{11})&[^"]*?index=(?P<index>\d+)' | |
1441 | IE_NAME = u'youtube:playlist' | |
1442 | ||
1443 | @classmethod | |
1444 | def suitable(cls, url): | |
1445 | """Receives a URL and returns True if suitable for this IE.""" | |
1446 | return re.match(cls._VALID_URL, url, re.VERBOSE) is not None | |
1447 | ||
1448 | def _real_initialize(self): | |
1449 | self._login() | |
1450 | ||
1451 | def _ids_to_results(self, ids): | |
1452 | return [self.url_result(vid_id, 'Youtube', video_id=vid_id) | |
1453 | for vid_id in ids] | |
1454 | ||
1455 | def _extract_mix(self, playlist_id): | |
1456 | # The mixes are generated from a a single video | |
1457 | # the id of the playlist is just 'RD' + video_id | |
1458 | url = 'https://youtube.com/watch?v=%s&list=%s' % (playlist_id[-11:], playlist_id) | |
1459 | webpage = self._download_webpage(url, playlist_id, u'Downloading Youtube mix') | |
1460 | title_span = (get_element_by_attribute('class', 'title long-title', webpage) or | |
1461 | get_element_by_attribute('class', 'title ', webpage)) | |
1462 | title = clean_html(title_span) | |
1463 | video_re = r'data-index="\d+".*?href="/watch\?v=([0-9A-Za-z_-]{11})&[^"]*?list=%s' % re.escape(playlist_id) | |
1464 | ids = orderedSet(re.findall(video_re, webpage)) | |
1465 | url_results = self._ids_to_results(ids) | |
1466 | ||
1467 | return self.playlist_result(url_results, playlist_id, title) | |
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 | playlist_id = mobj.group(1) or mobj.group(2) | |
1475 | ||
1476 | # Check if it's a video-specific URL | |
1477 | query_dict = compat_urlparse.parse_qs(compat_urlparse.urlparse(url).query) | |
1478 | if 'v' in query_dict: | |
1479 | video_id = query_dict['v'][0] | |
1480 | if self._downloader.params.get('noplaylist'): | |
1481 | self.to_screen(u'Downloading just video %s because of --no-playlist' % video_id) | |
1482 | return self.url_result(video_id, 'Youtube', video_id=video_id) | |
1483 | else: | |
1484 | self.to_screen(u'Downloading playlist PL%s - add --no-playlist to just download video %s' % (playlist_id, video_id)) | |
1485 | ||
1486 | if playlist_id.startswith('RD'): | |
1487 | # Mixes require a custom extraction process | |
1488 | return self._extract_mix(playlist_id) | |
1489 | if playlist_id.startswith('TL'): | |
1490 | raise ExtractorError(u'For downloading YouTube.com top lists, use ' | |
1491 | u'the "yttoplist" keyword, for example "youtube-dl \'yttoplist:music:Top Tracks\'"', expected=True) | |
1492 | ||
1493 | # Extract the video ids from the playlist pages | |
1494 | ids = [] | |
1495 | ||
1496 | for page_num in itertools.count(1): | |
1497 | url = self._TEMPLATE_URL % (playlist_id, page_num) | |
1498 | page = self._download_webpage(url, playlist_id, u'Downloading page #%s' % page_num) | |
1499 | matches = re.finditer(self._VIDEO_RE, page) | |
1500 | # We remove the duplicates and the link with index 0 | |
1501 | # (it's not the first video of the playlist) | |
1502 | new_ids = orderedSet(m.group('id') for m in matches if m.group('index') != '0') | |
1503 | ids.extend(new_ids) | |
1504 | ||
1505 | if re.search(self._MORE_PAGES_INDICATOR, page) is None: | |
1506 | break | |
1507 | ||
1508 | try: | |
1509 | playlist_title = self._og_search_title(page) | |
1510 | except RegexNotFoundError: | |
1511 | self.report_warning( | |
1512 | u'Playlist page is missing OpenGraph title, falling back ...', | |
1513 | playlist_id) | |
1514 | playlist_title = self._html_search_regex( | |
1515 | r'<h1 class="pl-header-title">(.*?)</h1>', page, u'title') | |
1516 | ||
1517 | url_results = self._ids_to_results(ids) | |
1518 | return self.playlist_result(url_results, playlist_id, playlist_title) | |
1519 | ||
1520 | ||
1521 | class YoutubeTopListIE(YoutubePlaylistIE): | |
1522 | IE_NAME = u'youtube:toplist' | |
1523 | IE_DESC = (u'YouTube.com top lists, "yttoplist:{channel}:{list title}"' | |
1524 | u' (Example: "yttoplist:music:Top Tracks")') | |
1525 | _VALID_URL = r'yttoplist:(?P<chann>.*?):(?P<title>.*?)$' | |
1526 | ||
1527 | def _real_extract(self, url): | |
1528 | mobj = re.match(self._VALID_URL, url) | |
1529 | channel = mobj.group('chann') | |
1530 | title = mobj.group('title') | |
1531 | query = compat_urllib_parse.urlencode({'title': title}) | |
1532 | playlist_re = 'href="([^"]+?%s[^"]+?)"' % re.escape(query) | |
1533 | channel_page = self._download_webpage('https://www.youtube.com/%s' % channel, title) | |
1534 | link = self._html_search_regex(playlist_re, channel_page, u'list') | |
1535 | url = compat_urlparse.urljoin('https://www.youtube.com/', link) | |
1536 | ||
1537 | video_re = r'data-index="\d+".*?data-video-id="([0-9A-Za-z_-]{11})"' | |
1538 | ids = [] | |
1539 | # sometimes the webpage doesn't contain the videos | |
1540 | # retry until we get them | |
1541 | for i in itertools.count(0): | |
1542 | msg = u'Downloading Youtube mix' | |
1543 | if i > 0: | |
1544 | msg += ', retry #%d' % i | |
1545 | webpage = self._download_webpage(url, title, msg) | |
1546 | ids = orderedSet(re.findall(video_re, webpage)) | |
1547 | if ids: | |
1548 | break | |
1549 | url_results = self._ids_to_results(ids) | |
1550 | return self.playlist_result(url_results, playlist_title=title) | |
1551 | ||
1552 | ||
1553 | class YoutubeChannelIE(InfoExtractor): | |
1554 | IE_DESC = u'YouTube.com channels' | |
1555 | _VALID_URL = r"^(?:https?://)?(?:youtu\.be|(?:\w+\.)?youtube(?:-nocookie)?\.com)/channel/([0-9A-Za-z_-]+)" | |
1556 | _MORE_PAGES_INDICATOR = 'yt-uix-load-more' | |
1557 | _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' | |
1558 | IE_NAME = u'youtube:channel' | |
1559 | ||
1560 | def extract_videos_from_page(self, page): | |
1561 | ids_in_page = [] | |
1562 | for mobj in re.finditer(r'href="/watch\?v=([0-9A-Za-z_-]+)&?', page): | |
1563 | if mobj.group(1) not in ids_in_page: | |
1564 | ids_in_page.append(mobj.group(1)) | |
1565 | return ids_in_page | |
1566 | ||
1567 | def _real_extract(self, url): | |
1568 | # Extract channel id | |
1569 | mobj = re.match(self._VALID_URL, url) | |
1570 | if mobj is None: | |
1571 | raise ExtractorError(u'Invalid URL: %s' % url) | |
1572 | ||
1573 | # Download channel page | |
1574 | channel_id = mobj.group(1) | |
1575 | video_ids = [] | |
1576 | url = 'https://www.youtube.com/channel/%s/videos' % channel_id | |
1577 | channel_page = self._download_webpage(url, channel_id) | |
1578 | autogenerated = re.search(r'''(?x) | |
1579 | class="[^"]*?(?: | |
1580 | channel-header-autogenerated-label| | |
1581 | yt-channel-title-autogenerated | |
1582 | )[^"]*"''', channel_page) is not None | |
1583 | ||
1584 | if autogenerated: | |
1585 | # The videos are contained in a single page | |
1586 | # the ajax pages can't be used, they are empty | |
1587 | video_ids = self.extract_videos_from_page(channel_page) | |
1588 | else: | |
1589 | # Download all channel pages using the json-based channel_ajax query | |
1590 | for pagenum in itertools.count(1): | |
1591 | url = self._MORE_PAGES_URL % (pagenum, channel_id) | |
1592 | page = self._download_webpage(url, channel_id, | |
1593 | u'Downloading page #%s' % pagenum) | |
1594 | ||
1595 | page = json.loads(page) | |
1596 | ||
1597 | ids_in_page = self.extract_videos_from_page(page['content_html']) | |
1598 | video_ids.extend(ids_in_page) | |
1599 | ||
1600 | if self._MORE_PAGES_INDICATOR not in page['load_more_widget_html']: | |
1601 | break | |
1602 | ||
1603 | self._downloader.to_screen(u'[youtube] Channel %s: Found %i videos' % (channel_id, len(video_ids))) | |
1604 | ||
1605 | url_entries = [self.url_result(video_id, 'Youtube', video_id=video_id) | |
1606 | for video_id in video_ids] | |
1607 | return self.playlist_result(url_entries, channel_id) | |
1608 | ||
1609 | ||
1610 | class YoutubeUserIE(InfoExtractor): | |
1611 | IE_DESC = u'YouTube.com user videos (URL or "ytuser" keyword)' | |
1612 | _VALID_URL = r'(?:(?:(?:https?://)?(?:\w+\.)?youtube\.com/(?:user/)?(?!(?:attribution_link|watch)(?:$|[^a-z_A-Z0-9-])))|ytuser:)(?!feed/)([A-Za-z0-9_-]+)' | |
1613 | _TEMPLATE_URL = 'http://gdata.youtube.com/feeds/api/users/%s' | |
1614 | _GDATA_PAGE_SIZE = 50 | |
1615 | _GDATA_URL = 'http://gdata.youtube.com/feeds/api/users/%s/uploads?max-results=%d&start-index=%d&alt=json' | |
1616 | IE_NAME = u'youtube:user' | |
1617 | ||
1618 | @classmethod | |
1619 | def suitable(cls, url): | |
1620 | # Don't return True if the url can be extracted with other youtube | |
1621 | # extractor, the regex would is too permissive and it would match. | |
1622 | other_ies = iter(klass for (name, klass) in globals().items() if name.endswith('IE') and klass is not cls) | |
1623 | if any(ie.suitable(url) for ie in other_ies): return False | |
1624 | else: return super(YoutubeUserIE, cls).suitable(url) | |
1625 | ||
1626 | def _real_extract(self, url): | |
1627 | # Extract username | |
1628 | mobj = re.match(self._VALID_URL, url) | |
1629 | if mobj is None: | |
1630 | raise ExtractorError(u'Invalid URL: %s' % url) | |
1631 | ||
1632 | username = mobj.group(1) | |
1633 | ||
1634 | # Download video ids using YouTube Data API. Result size per | |
1635 | # query is limited (currently to 50 videos) so we need to query | |
1636 | # page by page until there are no video ids - it means we got | |
1637 | # all of them. | |
1638 | ||
1639 | def download_page(pagenum): | |
1640 | start_index = pagenum * self._GDATA_PAGE_SIZE + 1 | |
1641 | ||
1642 | gdata_url = self._GDATA_URL % (username, self._GDATA_PAGE_SIZE, start_index) | |
1643 | page = self._download_webpage( | |
1644 | gdata_url, username, | |
1645 | u'Downloading video ids from %d to %d' % ( | |
1646 | start_index, start_index + self._GDATA_PAGE_SIZE)) | |
1647 | ||
1648 | try: | |
1649 | response = json.loads(page) | |
1650 | except ValueError as err: | |
1651 | raise ExtractorError(u'Invalid JSON in API response: ' + compat_str(err)) | |
1652 | if 'entry' not in response['feed']: | |
1653 | return | |
1654 | ||
1655 | # Extract video identifiers | |
1656 | entries = response['feed']['entry'] | |
1657 | for entry in entries: | |
1658 | title = entry['title']['$t'] | |
1659 | video_id = entry['id']['$t'].split('/')[-1] | |
1660 | yield { | |
1661 | '_type': 'url', | |
1662 | 'url': video_id, | |
1663 | 'ie_key': 'Youtube', | |
1664 | 'id': 'video_id', | |
1665 | 'title': title, | |
1666 | } | |
1667 | url_results = PagedList(download_page, self._GDATA_PAGE_SIZE) | |
1668 | ||
1669 | return self.playlist_result(url_results, playlist_title=username) | |
1670 | ||
1671 | ||
1672 | class YoutubeSearchIE(SearchInfoExtractor): | |
1673 | IE_DESC = u'YouTube.com searches' | |
1674 | _API_URL = 'https://gdata.youtube.com/feeds/api/videos?q=%s&start-index=%i&max-results=50&v=2&alt=jsonc' | |
1675 | _MAX_RESULTS = 1000 | |
1676 | IE_NAME = u'youtube:search' | |
1677 | _SEARCH_KEY = 'ytsearch' | |
1678 | ||
1679 | def _get_n_results(self, query, n): | |
1680 | """Get a specified number of results for a query""" | |
1681 | ||
1682 | video_ids = [] | |
1683 | pagenum = 0 | |
1684 | limit = n | |
1685 | ||
1686 | while (50 * pagenum) < limit: | |
1687 | result_url = self._API_URL % (compat_urllib_parse.quote_plus(query), (50*pagenum)+1) | |
1688 | data_json = self._download_webpage( | |
1689 | result_url, video_id=u'query "%s"' % query, | |
1690 | note=u'Downloading page %s' % (pagenum + 1), | |
1691 | errnote=u'Unable to download API page') | |
1692 | data = json.loads(data_json) | |
1693 | api_response = data['data'] | |
1694 | ||
1695 | if 'items' not in api_response: | |
1696 | raise ExtractorError(u'[youtube] No video results') | |
1697 | ||
1698 | new_ids = list(video['id'] for video in api_response['items']) | |
1699 | video_ids += new_ids | |
1700 | ||
1701 | limit = min(n, api_response['totalItems']) | |
1702 | pagenum += 1 | |
1703 | ||
1704 | if len(video_ids) > n: | |
1705 | video_ids = video_ids[:n] | |
1706 | videos = [self.url_result(video_id, 'Youtube', video_id=video_id) | |
1707 | for video_id in video_ids] | |
1708 | return self.playlist_result(videos, query) | |
1709 | ||
1710 | class YoutubeSearchDateIE(YoutubeSearchIE): | |
1711 | IE_NAME = YoutubeSearchIE.IE_NAME + ':date' | |
1712 | _API_URL = 'https://gdata.youtube.com/feeds/api/videos?q=%s&start-index=%i&max-results=50&v=2&alt=jsonc&orderby=published' | |
1713 | _SEARCH_KEY = 'ytsearchdate' | |
1714 | IE_DESC = u'YouTube.com searches, newest videos first' | |
1715 | ||
1716 | class YoutubeShowIE(InfoExtractor): | |
1717 | IE_DESC = u'YouTube.com (multi-season) shows' | |
1718 | _VALID_URL = r'https?://www\.youtube\.com/show/(.*)' | |
1719 | IE_NAME = u'youtube:show' | |
1720 | ||
1721 | def _real_extract(self, url): | |
1722 | mobj = re.match(self._VALID_URL, url) | |
1723 | show_name = mobj.group(1) | |
1724 | webpage = self._download_webpage(url, show_name, u'Downloading show webpage') | |
1725 | # There's one playlist for each season of the show | |
1726 | m_seasons = list(re.finditer(r'href="(/playlist\?list=.*?)"', webpage)) | |
1727 | self.to_screen(u'%s: Found %s seasons' % (show_name, len(m_seasons))) | |
1728 | return [self.url_result('https://www.youtube.com' + season.group(1), 'YoutubePlaylist') for season in m_seasons] | |
1729 | ||
1730 | ||
1731 | class YoutubeFeedsInfoExtractor(YoutubeBaseInfoExtractor): | |
1732 | """ | |
1733 | Base class for extractors that fetch info from | |
1734 | http://www.youtube.com/feed_ajax | |
1735 | Subclasses must define the _FEED_NAME and _PLAYLIST_TITLE properties. | |
1736 | """ | |
1737 | _LOGIN_REQUIRED = True | |
1738 | # use action_load_personal_feed instead of action_load_system_feed | |
1739 | _PERSONAL_FEED = False | |
1740 | ||
1741 | @property | |
1742 | def _FEED_TEMPLATE(self): | |
1743 | action = 'action_load_system_feed' | |
1744 | if self._PERSONAL_FEED: | |
1745 | action = 'action_load_personal_feed' | |
1746 | return 'http://www.youtube.com/feed_ajax?%s=1&feed_name=%s&paging=%%s' % (action, self._FEED_NAME) | |
1747 | ||
1748 | @property | |
1749 | def IE_NAME(self): | |
1750 | return u'youtube:%s' % self._FEED_NAME | |
1751 | ||
1752 | def _real_initialize(self): | |
1753 | self._login() | |
1754 | ||
1755 | def _real_extract(self, url): | |
1756 | feed_entries = [] | |
1757 | paging = 0 | |
1758 | for i in itertools.count(1): | |
1759 | info = self._download_webpage(self._FEED_TEMPLATE % paging, | |
1760 | u'%s feed' % self._FEED_NAME, | |
1761 | u'Downloading page %s' % i) | |
1762 | info = json.loads(info) | |
1763 | feed_html = info['feed_html'] | |
1764 | m_ids = re.finditer(r'"/watch\?v=(.*?)["&]', feed_html) | |
1765 | ids = orderedSet(m.group(1) for m in m_ids) | |
1766 | feed_entries.extend( | |
1767 | self.url_result(video_id, 'Youtube', video_id=video_id) | |
1768 | for video_id in ids) | |
1769 | if info['paging'] is None: | |
1770 | break | |
1771 | paging = info['paging'] | |
1772 | return self.playlist_result(feed_entries, playlist_title=self._PLAYLIST_TITLE) | |
1773 | ||
1774 | class YoutubeSubscriptionsIE(YoutubeFeedsInfoExtractor): | |
1775 | IE_DESC = u'YouTube.com subscriptions feed, "ytsubs" keyword(requires authentication)' | |
1776 | _VALID_URL = r'https?://www\.youtube\.com/feed/subscriptions|:ytsubs(?:criptions)?' | |
1777 | _FEED_NAME = 'subscriptions' | |
1778 | _PLAYLIST_TITLE = u'Youtube Subscriptions' | |
1779 | ||
1780 | class YoutubeRecommendedIE(YoutubeFeedsInfoExtractor): | |
1781 | IE_DESC = u'YouTube.com recommended videos, "ytrec" keyword (requires authentication)' | |
1782 | _VALID_URL = r'https?://www\.youtube\.com/feed/recommended|:ytrec(?:ommended)?' | |
1783 | _FEED_NAME = 'recommended' | |
1784 | _PLAYLIST_TITLE = u'Youtube Recommended videos' | |
1785 | ||
1786 | class YoutubeWatchLaterIE(YoutubeFeedsInfoExtractor): | |
1787 | IE_DESC = u'Youtube watch later list, "ytwatchlater" keyword (requires authentication)' | |
1788 | _VALID_URL = r'https?://www\.youtube\.com/feed/watch_later|:ytwatchlater' | |
1789 | _FEED_NAME = 'watch_later' | |
1790 | _PLAYLIST_TITLE = u'Youtube Watch Later' | |
1791 | _PERSONAL_FEED = True | |
1792 | ||
1793 | class YoutubeHistoryIE(YoutubeFeedsInfoExtractor): | |
1794 | IE_DESC = u'Youtube watch history, "ythistory" keyword (requires authentication)' | |
1795 | _VALID_URL = u'https?://www\.youtube\.com/feed/history|:ythistory' | |
1796 | _FEED_NAME = 'history' | |
1797 | _PERSONAL_FEED = True | |
1798 | _PLAYLIST_TITLE = u'Youtube Watch History' | |
1799 | ||
1800 | class YoutubeFavouritesIE(YoutubeBaseInfoExtractor): | |
1801 | IE_NAME = u'youtube:favorites' | |
1802 | IE_DESC = u'YouTube.com favourite videos, "ytfav" keyword (requires authentication)' | |
1803 | _VALID_URL = r'https?://www\.youtube\.com/my_favorites|:ytfav(?:ou?rites)?' | |
1804 | _LOGIN_REQUIRED = True | |
1805 | ||
1806 | def _real_extract(self, url): | |
1807 | webpage = self._download_webpage('https://www.youtube.com/my_favorites', 'Youtube Favourites videos') | |
1808 | playlist_id = self._search_regex(r'list=(.+?)["&]', webpage, u'favourites playlist id') | |
1809 | return self.url_result(playlist_id, 'YoutubePlaylist') | |
1810 | ||
1811 | ||
1812 | class YoutubeTruncatedURLIE(InfoExtractor): | |
1813 | IE_NAME = 'youtube:truncated_url' | |
1814 | IE_DESC = False # Do not list | |
1815 | _VALID_URL = r'''(?x) | |
1816 | (?:https?://)?[^/]+/watch\?feature=[a-z_]+$| | |
1817 | (?:https?://)?(?:www\.)?youtube\.com/attribution_link\?a=[^&]+$ | |
1818 | ''' | |
1819 | ||
1820 | def _real_extract(self, url): | |
1821 | raise ExtractorError( | |
1822 | u'Did you forget to quote the URL? Remember that & is a meta ' | |
1823 | u'character in most shells, so you want to put the URL in quotes, ' | |
1824 | u'like youtube-dl ' | |
1825 | u'"http://www.youtube.com/watch?feature=foo&v=BaW_jenozKc" ' | |
1826 | u' or simply youtube-dl BaW_jenozKc .', | |
1827 | expected=True) |