15 from .common import InfoExtractor, SearchInfoExtractor
16 from .subtitles import SubtitlesInfoExtractor
21 compat_urllib_request,
28 get_element_by_attribute,
40 class YoutubeBaseInfoExtractor(InfoExtractor):
41 """Provide base functions for Youtube extractors"""
42 _LOGIN_URL = 'https://accounts.google.com/ServiceLogin'
43 _LANG_URL = r'https://www.youtube.com/?hl=en&persist_hl=1&gl=US&persist_gl=1&opt_out_ackd=1'
44 _AGE_URL = 'https://www.youtube.com/verify_age?next_url=/&gl=US&hl=en'
45 _NETRC_MACHINE = 'youtube'
46 # If True it will raise an error if no login info is provided
47 _LOGIN_REQUIRED = False
49 def _set_language(self):
50 return bool(self._download_webpage(
52 note=u'Setting language', errnote='unable to set language',
56 (username, password) = self._get_login_info()
57 # No authentication to be performed
59 if self._LOGIN_REQUIRED:
60 raise ExtractorError(u'No login info available, needed for using %s.' % self.IE_NAME, expected=True)
63 login_page = self._download_webpage(
64 self._LOGIN_URL, None,
65 note=u'Downloading login page',
66 errnote=u'unable to fetch login page', fatal=False)
67 if login_page is False:
70 galx = self._search_regex(r'(?s)<input.+?name="GALX".+?value="(.+?)"',
71 login_page, u'Login GALX parameter')
75 u'continue': u'https://www.youtube.com/signin?action_handle_signin=true&feature=sign_in_button&hl=en_US&nomobiletemp=1',
79 u'PersistentCookie': u'yes',
81 u'bgresponse': u'js_disabled',
82 u'checkConnection': u'',
83 u'checkedDomains': u'youtube',
88 u'signIn': u'Sign in',
90 u'service': u'youtube',
94 # Convert to UTF-8 *before* urlencode because Python 2.x's urlencode
96 login_form = dict((k.encode('utf-8'), v.encode('utf-8')) for k,v in login_form_strs.items())
97 login_data = compat_urllib_parse.urlencode(login_form).encode('ascii')
99 req = compat_urllib_request.Request(self._LOGIN_URL, login_data)
100 login_results = self._download_webpage(
102 note=u'Logging in', errnote=u'unable to log in', fatal=False)
103 if login_results is False:
105 if re.search(r'(?i)<form[^>]* id="gaia_loginform"', login_results) is not None:
106 self._downloader.report_warning(u'unable to log in: bad username or password')
110 def _confirm_age(self):
113 'action_confirm': 'Confirm',
115 req = compat_urllib_request.Request(self._AGE_URL,
116 compat_urllib_parse.urlencode(age_form).encode('ascii'))
118 self._download_webpage(
120 note=u'Confirming age', errnote=u'Unable to confirm age')
123 def _real_initialize(self):
124 if self._downloader is None:
126 if not self._set_language():
128 if not self._login():
133 class YoutubeIE(YoutubeBaseInfoExtractor, SubtitlesInfoExtractor):
134 IE_DESC = u'YouTube.com'
135 _VALID_URL = r"""(?x)^
137 (?:https?://|//)? # http(s):// or protocol-independent URL (optional)
138 (?:(?:(?:(?:\w+\.)?[yY][oO][uU][tT][uU][bB][eE](?:-nocookie)?\.com/|
139 (?:www\.)?deturl\.com/www\.youtube\.com/|
140 (?:www\.)?pwnyoutube\.com/|
141 (?:www\.)?yourepeat\.com/|
142 tube\.majestyc\.net/|
143 youtube\.googleapis\.com/) # the various hostnames, with wildcard subdomains
144 (?:.*?\#/)? # handle anchor (#/) redirect urls
145 (?: # the various things that can precede the ID:
146 (?:(?:v|embed|e)/) # v/ or embed/ or e/
147 |(?: # or the v= param in all its forms
148 (?:(?:watch|movie)(?:_popup)?(?:\.php)?/?)? # preceding watch(_popup|.php) or nothing (like /?v=xxxx)
149 (?:\?|\#!?) # the params delimiter ? or # or #!
150 (?:.*?&)? # any other preceding param (like /?s=tuff&v=xxxx)
154 |youtu\.be/ # just youtu.be/xxxx
156 )? # all until now is optional -> you can pass the naked ID
157 ([0-9A-Za-z_-]{11}) # here is it! the YouTube video ID
158 (?(1).+)? # if we found the ID, everything can follow
160 _NEXT_URL_RE = r'[\?&]next_url=([^&]+)'
162 '5': {'ext': 'flv', 'width': 400, 'height': 240},
163 '6': {'ext': 'flv', 'width': 450, 'height': 270},
164 '13': {'ext': '3gp'},
165 '17': {'ext': '3gp', 'width': 176, 'height': 144},
166 '18': {'ext': 'mp4', 'width': 640, 'height': 360},
167 '22': {'ext': 'mp4', 'width': 1280, 'height': 720},
168 '34': {'ext': 'flv', 'width': 640, 'height': 360},
169 '35': {'ext': 'flv', 'width': 854, 'height': 480},
170 '36': {'ext': '3gp', 'width': 320, 'height': 240},
171 '37': {'ext': 'mp4', 'width': 1920, 'height': 1080},
172 '38': {'ext': 'mp4', 'width': 4096, 'height': 3072},
173 '43': {'ext': 'webm', 'width': 640, 'height': 360},
174 '44': {'ext': 'webm', 'width': 854, 'height': 480},
175 '45': {'ext': 'webm', 'width': 1280, 'height': 720},
176 '46': {'ext': 'webm', 'width': 1920, 'height': 1080},
180 '82': {'ext': 'mp4', 'height': 360, 'resolution': '360p', 'format_note': '3D', 'preference': -20},
181 '83': {'ext': 'mp4', 'height': 480, 'resolution': '480p', 'format_note': '3D', 'preference': -20},
182 '84': {'ext': 'mp4', 'height': 720, 'resolution': '720p', 'format_note': '3D', 'preference': -20},
183 '85': {'ext': 'mp4', 'height': 1080, 'resolution': '1080p', 'format_note': '3D', 'preference': -20},
184 '100': {'ext': 'webm', 'height': 360, 'resolution': '360p', 'format_note': '3D', 'preference': -20},
185 '101': {'ext': 'webm', 'height': 480, 'resolution': '480p', 'format_note': '3D', 'preference': -20},
186 '102': {'ext': 'webm', 'height': 720, 'resolution': '720p', 'format_note': '3D', 'preference': -20},
188 # Apple HTTP Live Streaming
189 '92': {'ext': 'mp4', 'height': 240, 'resolution': '240p', 'format_note': 'HLS', 'preference': -10},
190 '93': {'ext': 'mp4', 'height': 360, 'resolution': '360p', 'format_note': 'HLS', 'preference': -10},
191 '94': {'ext': 'mp4', 'height': 480, 'resolution': '480p', 'format_note': 'HLS', 'preference': -10},
192 '95': {'ext': 'mp4', 'height': 720, 'resolution': '720p', 'format_note': 'HLS', 'preference': -10},
193 '96': {'ext': 'mp4', 'height': 1080, 'resolution': '1080p', 'format_note': 'HLS', 'preference': -10},
194 '132': {'ext': 'mp4', 'height': 240, 'resolution': '240p', 'format_note': 'HLS', 'preference': -10},
195 '151': {'ext': 'mp4', 'height': 72, 'resolution': '72p', 'format_note': 'HLS', 'preference': -10},
198 '133': {'ext': 'mp4', 'height': 240, 'resolution': '240p', 'format_note': 'DASH video', 'preference': -40},
199 '134': {'ext': 'mp4', 'height': 360, 'resolution': '360p', 'format_note': 'DASH video', 'preference': -40},
200 '135': {'ext': 'mp4', 'height': 480, 'resolution': '480p', 'format_note': 'DASH video', 'preference': -40},
201 '136': {'ext': 'mp4', 'height': 720, 'resolution': '720p', 'format_note': 'DASH video', 'preference': -40},
202 '137': {'ext': 'mp4', 'height': 1080, 'resolution': '1080p', 'format_note': 'DASH video', 'preference': -40},
203 '138': {'ext': 'mp4', 'height': 1081, 'resolution': '>1080p', 'format_note': 'DASH video', 'preference': -40},
204 '160': {'ext': 'mp4', 'height': 192, 'resolution': '192p', 'format_note': 'DASH video', 'preference': -40},
205 '264': {'ext': 'mp4', 'height': 1080, 'resolution': '1080p', 'format_note': 'DASH video', 'preference': -40},
208 '139': {'ext': 'm4a', 'format_note': 'DASH audio', 'vcodec': 'none', 'abr': 48, 'preference': -50},
209 '140': {'ext': 'm4a', 'format_note': 'DASH audio', 'vcodec': 'none', 'abr': 128, 'preference': -50},
210 '141': {'ext': 'm4a', 'format_note': 'DASH audio', 'vcodec': 'none', 'abr': 256, 'preference': -50},
213 '167': {'ext': 'webm', 'height': 360, 'width': 640, 'format_note': 'DASH video', 'container': 'webm', 'vcodec': 'VP8', 'acodec': 'none', 'preference': -40},
214 '168': {'ext': 'webm', 'height': 480, 'width': 854, 'format_note': 'DASH video', 'container': 'webm', 'vcodec': 'VP8', 'acodec': 'none', 'preference': -40},
215 '169': {'ext': 'webm', 'height': 720, 'width': 1280, 'format_note': 'DASH video', 'container': 'webm', 'vcodec': 'VP8', 'acodec': 'none', 'preference': -40},
216 '170': {'ext': 'webm', 'height': 1080, 'width': 1920, 'format_note': 'DASH video', 'container': 'webm', 'vcodec': 'VP8', 'acodec': 'none', 'preference': -40},
217 '218': {'ext': 'webm', 'height': 480, 'width': 854, 'format_note': 'DASH video', 'container': 'webm', 'vcodec': 'VP8', 'acodec': 'none', 'preference': -40},
218 '219': {'ext': 'webm', 'height': 480, 'width': 854, 'format_note': 'DASH video', 'container': 'webm', 'vcodec': 'VP8', 'acodec': 'none', 'preference': -40},
219 '242': {'ext': 'webm', 'height': 240, 'resolution': '240p', 'format_note': 'DASH webm', 'preference': -40},
220 '243': {'ext': 'webm', 'height': 360, 'resolution': '360p', 'format_note': 'DASH webm', 'preference': -40},
221 '244': {'ext': 'webm', 'height': 480, 'resolution': '480p', 'format_note': 'DASH webm', 'preference': -40},
222 '245': {'ext': 'webm', 'height': 480, 'resolution': '480p', 'format_note': 'DASH webm', 'preference': -40},
223 '246': {'ext': 'webm', 'height': 480, 'resolution': '480p', 'format_note': 'DASH webm', 'preference': -40},
224 '247': {'ext': 'webm', 'height': 720, 'resolution': '720p', 'format_note': 'DASH webm', 'preference': -40},
225 '248': {'ext': 'webm', 'height': 1080, 'resolution': '1080p', 'format_note': 'DASH webm', 'preference': -40},
228 '171': {'ext': 'webm', 'vcodec': 'none', 'format_note': 'DASH webm audio', 'abr': 48, 'preference': -50},
229 '172': {'ext': 'webm', 'vcodec': 'none', 'format_note': 'DASH webm audio', 'abr': 256, 'preference': -50},
232 '_rtmp': {'protocol': 'rtmp'},
238 u"url": u"http://www.youtube.com/watch?v=BaW_jenozKc",
239 u"file": u"BaW_jenozKc.mp4",
241 u"title": u"youtube-dl test video \"'/\\ä↭𝕐",
242 u"uploader": u"Philipp Hagemeister",
243 u"uploader_id": u"phihag",
244 u"upload_date": u"20121002",
245 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 ."
249 u"url": u"http://www.youtube.com/watch?v=UxxajLWwzqY",
250 u"file": u"UxxajLWwzqY.mp4",
251 u"note": u"Test generic use_cipher_signature video (#897)",
253 u"upload_date": u"20120506",
254 u"title": u"Icona Pop - I Love It (feat. Charli XCX) [OFFICIAL VIDEO]",
255 u"description": u"md5:5b292926389560516e384ac437c0ec07",
256 u"uploader": u"Icona Pop",
257 u"uploader_id": u"IconaPop"
261 u"url": u"https://www.youtube.com/watch?v=07FYdnEawAQ",
262 u"file": u"07FYdnEawAQ.mp4",
263 u"note": u"Test VEVO video with age protection (#956)",
265 u"upload_date": u"20130703",
266 u"title": u"Justin Timberlake - Tunnel Vision (Explicit)",
267 u"description": u"md5:64249768eec3bc4276236606ea996373",
268 u"uploader": u"justintimberlakeVEVO",
269 u"uploader_id": u"justintimberlakeVEVO"
273 u"url": u"//www.YouTube.com/watch?v=yZIXLfi8CZQ",
274 u"file": u"yZIXLfi8CZQ.mp4",
275 u"note": u"Embed-only video (#1746)",
277 u"upload_date": u"20120608",
278 u"title": u"Principal Sexually Assaults A Teacher - Episode 117 - 8th June 2012",
279 u"description": u"md5:09b78bd971f1e3e289601dfba15ca4f7",
280 u"uploader": u"SET India",
281 u"uploader_id": u"setindia"
285 u"url": u"http://www.youtube.com/watch?v=a9LDPn-MO4I",
286 u"file": u"a9LDPn-MO4I.m4a",
287 u"note": u"256k DASH audio (format 141) via DASH manifest",
289 u"upload_date": "20121002",
290 u"uploader_id": "8KVIDEO",
291 u"description": "No description available.",
292 u"uploader": "8KVIDEO",
293 u"title": "UHDTV TEST 8K VIDEO.mp4"
296 u"youtube_include_dash_manifest": True,
300 # DASH manifest with encrypted signature
302 u'url': u'https://www.youtube.com/watch?v=IB3lcPjvWLA',
304 u'id': u'IB3lcPjvWLA',
306 u'title': u'Afrojack - The Spark ft. Spree Wilson',
307 u'description': u'md5:3199ed45ee8836572865580804d7ac0f',
308 u'uploader': u'AfrojackVEVO',
309 u'uploader_id': u'AfrojackVEVO',
310 u'upload_date': u'20131011',
313 u'youtube_include_dash_manifest': True,
321 def suitable(cls, url):
322 """Receives a URL and returns True if suitable for this IE."""
323 if YoutubePlaylistIE.suitable(url): return False
324 return re.match(cls._VALID_URL, url) is not None
326 def __init__(self, *args, **kwargs):
327 super(YoutubeIE, self).__init__(*args, **kwargs)
328 self._player_cache = {}
330 def report_video_info_webpage_download(self, video_id):
331 """Report attempt to download video info webpage."""
332 self.to_screen(u'%s: Downloading video info webpage' % video_id)
334 def report_information_extraction(self, video_id):
335 """Report attempt to extract video information."""
336 self.to_screen(u'%s: Extracting video information' % video_id)
338 def report_unavailable_format(self, video_id, format):
339 """Report extracted video URL."""
340 self.to_screen(u'%s: Format %s not available' % (video_id, format))
342 def report_rtmp_download(self):
343 """Indicate the download will use the RTMP protocol."""
344 self.to_screen(u'RTMP download detected')
346 def _extract_signature_function(self, video_id, player_url, slen):
347 id_m = re.match(r'.*-(?P<id>[a-zA-Z0-9_-]+)\.(?P<ext>[a-z]+)$',
349 player_type = id_m.group('ext')
350 player_id = id_m.group('id')
352 # Read from filesystem cache
353 func_id = '%s_%s_%d' % (player_type, player_id, slen)
354 assert os.path.basename(func_id) == func_id
355 cache_dir = get_cachedir(self._downloader.params)
357 cache_enabled = cache_dir is not None
359 cache_fn = os.path.join(os.path.expanduser(cache_dir),
363 with io.open(cache_fn, 'r', encoding='utf-8') as cachef:
364 cache_spec = json.load(cachef)
365 return lambda s: u''.join(s[i] for i in cache_spec)
367 pass # No cache available
369 if player_type == 'js':
370 code = self._download_webpage(
371 player_url, video_id,
372 note=u'Downloading %s player %s' % (player_type, player_id),
373 errnote=u'Download of %s failed' % player_url)
374 res = self._parse_sig_js(code)
375 elif player_type == 'swf':
376 urlh = self._request_webpage(
377 player_url, video_id,
378 note=u'Downloading %s player %s' % (player_type, player_id),
379 errnote=u'Download of %s failed' % player_url)
381 res = self._parse_sig_swf(code)
383 assert False, 'Invalid player type %r' % player_type
387 test_string = u''.join(map(compat_chr, range(slen)))
388 cache_res = res(test_string)
389 cache_spec = [ord(c) for c in cache_res]
391 os.makedirs(os.path.dirname(cache_fn))
392 except OSError as ose:
393 if ose.errno != errno.EEXIST:
395 write_json_file(cache_spec, cache_fn)
397 tb = traceback.format_exc()
398 self._downloader.report_warning(
399 u'Writing cache to %r failed: %s' % (cache_fn, tb))
403 def _print_sig_code(self, func, slen):
404 def gen_sig_code(idxs):
405 def _genslice(start, end, step):
406 starts = u'' if start == 0 else str(start)
407 ends = (u':%d' % (end+step)) if end + step >= 0 else u':'
408 steps = u'' if step == 1 else (u':%d' % step)
409 return u's[%s%s%s]' % (starts, ends, steps)
412 start = '(Never used)' # Quelch pyflakes warnings - start will be
413 # set as soon as step is set
414 for i, prev in zip(idxs[1:], idxs[:-1]):
418 yield _genslice(start, prev, step)
421 if i - prev in [-1, 1]:
426 yield u's[%d]' % prev
430 yield _genslice(start, i, step)
432 test_string = u''.join(map(compat_chr, range(slen)))
433 cache_res = func(test_string)
434 cache_spec = [ord(c) for c in cache_res]
435 expr_code = u' + '.join(gen_sig_code(cache_spec))
436 code = u'if len(s) == %d:\n return %s\n' % (slen, expr_code)
437 self.to_screen(u'Extracted signature function:\n' + code)
439 def _parse_sig_js(self, jscode):
440 funcname = self._search_regex(
441 r'signature=([a-zA-Z]+)', jscode,
442 u'Initial JS player signature function name')
447 return string.lowercase.index(varname)
449 def interpret_statement(stmt, local_vars, allow_recursion=20):
450 if allow_recursion < 0:
451 raise ExtractorError(u'Recursion limit reached')
453 if stmt.startswith(u'var '):
454 stmt = stmt[len(u'var '):]
455 ass_m = re.match(r'^(?P<out>[a-z]+)(?:\[(?P<index>[^\]]+)\])?' +
456 r'=(?P<expr>.*)$', stmt)
458 if ass_m.groupdict().get('index'):
460 lvar = local_vars[ass_m.group('out')]
461 idx = interpret_expression(ass_m.group('index'),
462 local_vars, allow_recursion)
463 assert isinstance(idx, int)
466 expr = ass_m.group('expr')
469 local_vars[ass_m.group('out')] = val
471 expr = ass_m.group('expr')
472 elif stmt.startswith(u'return '):
474 expr = stmt[len(u'return '):]
476 raise ExtractorError(
477 u'Cannot determine left side of statement in %r' % stmt)
479 v = interpret_expression(expr, local_vars, allow_recursion)
482 def interpret_expression(expr, local_vars, allow_recursion):
487 return local_vars[expr]
489 m = re.match(r'^(?P<in>[a-z]+)\.(?P<member>.*)$', expr)
491 member = m.group('member')
492 val = local_vars[m.group('in')]
493 if member == 'split("")':
495 if member == 'join("")':
497 if member == 'length':
499 if member == 'reverse()':
501 slice_m = re.match(r'slice\((?P<idx>.*)\)', member)
503 idx = interpret_expression(
504 slice_m.group('idx'), local_vars, allow_recursion-1)
508 r'^(?P<in>[a-z]+)\[(?P<idx>.+)\]$', expr)
510 val = local_vars[m.group('in')]
511 idx = interpret_expression(m.group('idx'), local_vars,
515 m = re.match(r'^(?P<a>.+?)(?P<op>[%])(?P<b>.+?)$', expr)
517 a = interpret_expression(m.group('a'),
518 local_vars, allow_recursion)
519 b = interpret_expression(m.group('b'),
520 local_vars, allow_recursion)
524 r'^(?P<func>[a-zA-Z$]+)\((?P<args>[a-z0-9,]+)\)$', expr)
526 fname = m.group('func')
527 if fname not in functions:
528 functions[fname] = extract_function(fname)
529 argvals = [int(v) if v.isdigit() else local_vars[v]
530 for v in m.group('args').split(',')]
531 return functions[fname](argvals)
532 raise ExtractorError(u'Unsupported JS expression %r' % expr)
534 def extract_function(funcname):
536 r'function ' + re.escape(funcname) +
537 r'\((?P<args>[a-z,]+)\){(?P<code>[^}]+)}',
539 argnames = func_m.group('args').split(',')
542 local_vars = dict(zip(argnames, args))
543 for stmt in func_m.group('code').split(';'):
544 res = interpret_statement(stmt, local_vars)
548 initial_function = extract_function(funcname)
549 return lambda s: initial_function([s])
551 def _parse_sig_swf(self, file_contents):
552 if file_contents[1:3] != b'WS':
553 raise ExtractorError(
554 u'Not an SWF file; header is %r' % file_contents[:3])
555 if file_contents[:1] == b'C':
556 content = zlib.decompress(file_contents[8:])
558 raise NotImplementedError(u'Unsupported compression format %r' %
561 def extract_tags(content):
563 while pos < len(content):
564 header16 = struct.unpack('<H', content[pos:pos+2])[0]
566 tag_code = header16 >> 6
567 tag_len = header16 & 0x3f
569 tag_len = struct.unpack('<I', content[pos:pos+4])[0]
571 assert pos+tag_len <= len(content)
572 yield (tag_code, content[pos:pos+tag_len])
576 for tag_code, tag in extract_tags(content)
578 p = code_tag.index(b'\0', 4) + 1
579 code_reader = io.BytesIO(code_tag[p:])
581 # Parse ABC (AVM2 ByteCode)
582 def read_int(reader=None):
590 b = struct.unpack('<B', buf)[0]
591 res = res | ((b & 0x7f) << shift)
597 def u30(reader=None):
598 res = read_int(reader)
599 assert res & 0xf0000000 == 0
603 def s32(reader=None):
605 if v & 0x80000000 != 0:
606 v = - ((v ^ 0xffffffff) + 1)
609 def read_string(reader=None):
613 resb = reader.read(slen)
614 assert len(resb) == slen
615 return resb.decode('utf-8')
617 def read_bytes(count, reader=None):
620 resb = reader.read(count)
621 assert len(resb) == count
624 def read_byte(reader=None):
625 resb = read_bytes(1, reader=reader)
626 res = struct.unpack('<B', resb)[0]
629 # minor_version + major_version
634 for _c in range(1, int_count):
637 for _c in range(1, uint_count):
640 read_bytes((double_count-1) * 8)
642 constant_strings = [u'']
643 for _c in range(1, string_count):
645 constant_strings.append(s)
646 namespace_count = u30()
647 for _c in range(1, namespace_count):
651 for _c in range(1, ns_set_count):
653 for _c2 in range(count):
655 multiname_count = u30()
664 0x0e: 2, # MultinameA
665 0x1b: 1, # MultinameL
666 0x1c: 1, # MultinameLA
669 for _c in range(1, multiname_count):
671 assert kind in MULTINAME_SIZES, u'Invalid multiname kind %r' % kind
673 u30() # namespace_idx
675 multinames.append(constant_strings[name_idx])
677 multinames.append('[MULTINAME kind: %d]' % kind)
678 for _c2 in range(MULTINAME_SIZES[kind]):
683 MethodInfo = collections.namedtuple(
685 ['NEED_ARGUMENTS', 'NEED_REST'])
687 for method_id in range(method_count):
690 for _ in range(param_count):
692 u30() # name index (always 0 for youtube)
694 if flags & 0x08 != 0:
697 for c in range(option_count):
700 if flags & 0x80 != 0:
701 # Param names present
702 for _ in range(param_count):
704 mi = MethodInfo(flags & 0x01 != 0, flags & 0x04 != 0)
705 method_infos.append(mi)
708 metadata_count = u30()
709 for _c in range(metadata_count):
712 for _c2 in range(item_count):
716 def parse_traits_info():
717 trait_name_idx = u30()
718 kind_full = read_byte()
719 kind = kind_full & 0x0f
720 attrs = kind_full >> 4
722 if kind in [0x00, 0x06]: # Slot or Const
724 u30() # type_name_idx
728 elif kind in [0x01, 0x02, 0x03]: # Method / Getter / Setter
731 methods[multinames[trait_name_idx]] = method_idx
732 elif kind == 0x04: # Class
735 elif kind == 0x05: # Function
738 methods[function_idx] = multinames[trait_name_idx]
740 raise ExtractorError(u'Unsupported trait kind %d' % kind)
742 if attrs & 0x4 != 0: # Metadata present
743 metadata_count = u30()
744 for _c3 in range(metadata_count):
745 u30() # metadata index
750 TARGET_CLASSNAME = u'SignatureDecipher'
751 searched_idx = multinames.index(TARGET_CLASSNAME)
752 searched_class_id = None
754 for class_id in range(class_count):
756 if name_idx == searched_idx:
757 # We found the class we're looking for!
758 searched_class_id = class_id
759 u30() # super_name idx
761 if flags & 0x08 != 0: # Protected namespace is present
762 u30() # protected_ns_idx
764 for _c2 in range(intrf_count):
768 for _c2 in range(trait_count):
771 if searched_class_id is None:
772 raise ExtractorError(u'Target class %r not found' %
777 for class_id in range(class_count):
780 for _c2 in range(trait_count):
781 trait_methods = parse_traits_info()
782 if class_id == searched_class_id:
783 method_names.update(trait_methods.items())
784 method_idxs.update(dict(
786 for name, idx in trait_methods.items()))
790 for _c in range(script_count):
793 for _c2 in range(trait_count):
797 method_body_count = u30()
798 Method = collections.namedtuple('Method', ['code', 'local_count'])
800 for _c in range(method_body_count):
804 u30() # init_scope_depth
805 u30() # max_scope_depth
807 code = read_bytes(code_length)
808 if method_idx in method_idxs:
809 m = Method(code, local_count)
810 methods[method_idxs[method_idx]] = m
811 exception_count = u30()
812 for _c2 in range(exception_count):
819 for _c2 in range(trait_count):
822 assert p + code_reader.tell() == len(code_tag)
823 assert len(methods) == len(method_idxs)
825 method_pyfunctions = {}
827 def extract_function(func_name):
828 if func_name in method_pyfunctions:
829 return method_pyfunctions[func_name]
830 if func_name not in methods:
831 raise ExtractorError(u'Cannot find function %r' % func_name)
832 m = methods[func_name]
835 registers = ['(this)'] + list(args) + [None] * m.local_count
837 coder = io.BytesIO(m.code)
839 opcode = struct.unpack('!B', coder.read(1))[0]
840 if opcode == 36: # pushbyte
841 v = struct.unpack('!B', coder.read(1))[0]
843 elif opcode == 44: # pushstring
845 stack.append(constant_strings[idx])
846 elif opcode == 48: # pushscope
847 # We don't implement the scope register, so we'll just
848 # ignore the popped value
850 elif opcode == 70: # callproperty
852 mname = multinames[index]
853 arg_count = u30(coder)
854 args = list(reversed(
855 [stack.pop() for _ in range(arg_count)]))
857 if mname == u'split':
858 assert len(args) == 1
859 assert isinstance(args[0], compat_str)
860 assert isinstance(obj, compat_str)
864 res = obj.split(args[0])
866 elif mname == u'slice':
867 assert len(args) == 1
868 assert isinstance(args[0], int)
869 assert isinstance(obj, list)
872 elif mname == u'join':
873 assert len(args) == 1
874 assert isinstance(args[0], compat_str)
875 assert isinstance(obj, list)
876 res = args[0].join(obj)
878 elif mname in method_pyfunctions:
879 stack.append(method_pyfunctions[mname](args))
881 raise NotImplementedError(
882 u'Unsupported property %r on %r'
884 elif opcode == 72: # returnvalue
887 elif opcode == 79: # callpropvoid
889 mname = multinames[index]
890 arg_count = u30(coder)
891 args = list(reversed(
892 [stack.pop() for _ in range(arg_count)]))
894 if mname == u'reverse':
895 assert isinstance(obj, list)
898 raise NotImplementedError(
899 u'Unsupported (void) property %r on %r'
901 elif opcode == 93: # findpropstrict
903 mname = multinames[index]
904 res = extract_function(mname)
906 elif opcode == 97: # setproperty
911 assert isinstance(obj, list)
912 assert isinstance(idx, int)
914 elif opcode == 98: # getlocal
916 stack.append(registers[index])
917 elif opcode == 99: # setlocal
920 registers[index] = value
921 elif opcode == 102: # getproperty
923 pname = multinames[index]
924 if pname == u'length':
926 assert isinstance(obj, list)
927 stack.append(len(obj))
928 else: # Assume attribute access
930 assert isinstance(idx, int)
932 assert isinstance(obj, list)
933 stack.append(obj[idx])
934 elif opcode == 128: # coerce
936 elif opcode == 133: # coerce_s
937 assert isinstance(stack[-1], (type(None), compat_str))
938 elif opcode == 164: # modulo
941 res = value1 % value2
943 elif opcode == 208: # getlocal_0
944 stack.append(registers[0])
945 elif opcode == 209: # getlocal_1
946 stack.append(registers[1])
947 elif opcode == 210: # getlocal_2
948 stack.append(registers[2])
949 elif opcode == 211: # getlocal_3
950 stack.append(registers[3])
951 elif opcode == 214: # setlocal_2
952 registers[2] = stack.pop()
953 elif opcode == 215: # setlocal_3
954 registers[3] = stack.pop()
956 raise NotImplementedError(
957 u'Unsupported opcode %d' % opcode)
959 method_pyfunctions[func_name] = resfunc
962 initial_function = extract_function(u'decipher')
963 return lambda s: initial_function([s])
965 def _decrypt_signature(self, s, video_id, player_url, age_gate=False):
966 """Turn the encrypted s field into a working signature"""
968 if player_url is not None:
969 if player_url.startswith(u'//'):
970 player_url = u'https:' + player_url
972 player_id = (player_url, len(s))
973 if player_id not in self._player_cache:
974 func = self._extract_signature_function(
975 video_id, player_url, len(s)
977 self._player_cache[player_id] = func
978 func = self._player_cache[player_id]
979 if self._downloader.params.get('youtube_print_sig_code'):
980 self._print_sig_code(func, len(s))
983 tb = traceback.format_exc()
984 self._downloader.report_warning(
985 u'Automatic signature extraction failed: ' + tb)
987 self._downloader.report_warning(
988 u'Warning: Falling back to static signature algorithm')
990 return self._static_decrypt_signature(
991 s, video_id, player_url, age_gate)
993 def _static_decrypt_signature(self, s, video_id, player_url, age_gate):
995 # The videos with age protection use another player, so the
996 # algorithms can be different.
998 return s[2:63] + s[82] + s[64:82] + s[63]
1001 return s[86:29:-1] + s[88] + s[28:5:-1]
1003 return s[25] + s[3:25] + s[0] + s[26:42] + s[79] + s[43:79] + s[91] + s[80:83]
1005 return s[84:27:-1] + s[86] + s[26:5:-1]
1007 return s[25] + s[3:25] + s[2] + s[26:40] + s[77] + s[41:77] + s[89] + s[78:81]
1009 return s[84:78:-1] + s[87] + s[77:60:-1] + s[0] + s[59:3:-1]
1011 return s[7:28] + s[87] + s[29:45] + s[55] + s[46:55] + s[2] + s[56:87] + s[28]
1013 return s[6:27] + s[4] + s[28:39] + s[27] + s[40:59] + s[2] + s[60:]
1015 return s[80:72:-1] + s[16] + s[71:39:-1] + s[72] + s[38:16:-1] + s[82] + s[15::-1]
1017 return s[3:11] + s[0] + s[12:55] + s[84] + s[56:84]
1019 return s[78:70:-1] + s[14] + s[69:37:-1] + s[70] + s[36:14:-1] + s[80] + s[:14][::-1]
1021 return s[80:63:-1] + s[0] + s[62:0:-1] + s[63]
1023 return s[80:37:-1] + s[7] + s[36:7:-1] + s[0] + s[6:0:-1] + s[37]
1025 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]
1027 return s[1:19] + s[0] + s[20:68] + s[19] + s[69:80]
1029 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]
1032 raise ExtractorError(u'Unable to decrypt signature, key length %d not supported; retrying might work' % (len(s)))
1034 def _get_available_subtitles(self, video_id, webpage):
1036 sub_list = self._download_webpage(
1037 'https://video.google.com/timedtext?hl=en&type=list&v=%s' % video_id,
1038 video_id, note=False)
1039 except ExtractorError as err:
1040 self._downloader.report_warning(u'unable to download video subtitles: %s' % compat_str(err))
1042 lang_list = re.findall(r'name="([^"]*)"[^>]+lang_code="([\w\-]+)"', sub_list)
1047 params = compat_urllib_parse.urlencode({
1050 'fmt': self._downloader.params.get('subtitlesformat', 'srt'),
1051 'name': unescapeHTML(l[0]).encode('utf-8'),
1053 url = u'https://www.youtube.com/api/timedtext?' + params
1054 sub_lang_list[lang] = url
1055 if not sub_lang_list:
1056 self._downloader.report_warning(u'video doesn\'t have subtitles')
1058 return sub_lang_list
1060 def _get_available_automatic_caption(self, video_id, webpage):
1061 """We need the webpage for getting the captions url, pass it as an
1062 argument to speed up the process."""
1063 sub_format = self._downloader.params.get('subtitlesformat', 'srt')
1064 self.to_screen(u'%s: Looking for automatic captions' % video_id)
1065 mobj = re.search(r';ytplayer.config = ({.*?});', webpage)
1066 err_msg = u'Couldn\'t find automatic captions for %s' % video_id
1068 self._downloader.report_warning(err_msg)
1070 player_config = json.loads(mobj.group(1))
1072 args = player_config[u'args']
1073 caption_url = args[u'ttsurl']
1074 timestamp = args[u'timestamp']
1075 # We get the available subtitles
1076 list_params = compat_urllib_parse.urlencode({
1081 list_url = caption_url + '&' + list_params
1082 caption_list = self._download_xml(list_url, video_id)
1083 original_lang_node = caption_list.find('track')
1084 if original_lang_node is None or original_lang_node.attrib.get('kind') != 'asr' :
1085 self._downloader.report_warning(u'Video doesn\'t have automatic captions')
1087 original_lang = original_lang_node.attrib['lang_code']
1090 for lang_node in caption_list.findall('target'):
1091 sub_lang = lang_node.attrib['lang_code']
1092 params = compat_urllib_parse.urlencode({
1093 'lang': original_lang,
1099 sub_lang_list[sub_lang] = caption_url + '&' + params
1100 return sub_lang_list
1101 # An extractor error can be raise by the download process if there are
1102 # no automatic captions but there are subtitles
1103 except (KeyError, ExtractorError):
1104 self._downloader.report_warning(err_msg)
1108 def extract_id(cls, url):
1109 mobj = re.match(cls._VALID_URL, url, re.VERBOSE)
1111 raise ExtractorError(u'Invalid URL: %s' % url)
1112 video_id = mobj.group(2)
1115 def _extract_from_m3u8(self, manifest_url, video_id):
1117 def _get_urls(_manifest):
1118 lines = _manifest.split('\n')
1119 urls = filter(lambda l: l and not l.startswith('#'),
1122 manifest = self._download_webpage(manifest_url, video_id, u'Downloading formats manifest')
1123 formats_urls = _get_urls(manifest)
1124 for format_url in formats_urls:
1125 itag = self._search_regex(r'itag/(\d+?)/', format_url, 'itag')
1126 url_map[itag] = format_url
1129 def _extract_annotations(self, video_id):
1130 url = 'https://www.youtube.com/annotations_invideo?features=1&legacy=1&video_id=%s' % video_id
1131 return self._download_webpage(url, video_id, note=u'Searching for annotations.', errnote=u'Unable to download video annotations.')
1133 def _real_extract(self, url):
1134 # Extract original video URL from URL with redirection, like age verification, using next_url parameter
1135 mobj = re.search(self._NEXT_URL_RE, url)
1137 url = 'https://www.youtube.com/' + compat_urllib_parse.unquote(mobj.group(1)).lstrip('/')
1138 video_id = self.extract_id(url)
1141 url = 'https://www.youtube.com/watch?v=%s&gl=US&hl=en&has_verified=1' % video_id
1142 video_webpage = self._download_webpage(url, video_id)
1144 # Attempt to extract SWF player URL
1145 mobj = re.search(r'swfConfig.*?"(https?:\\/\\/.*?watch.*?-.*?\.swf)"', video_webpage)
1146 if mobj is not None:
1147 player_url = re.sub(r'\\(.)', r'\1', mobj.group(1))
1152 self.report_video_info_webpage_download(video_id)
1153 if re.search(r'player-age-gate-content">', video_webpage) is not None:
1154 self.report_age_confirmation()
1156 # We simulate the access to the video from www.youtube.com/v/{video_id}
1157 # this can be viewed without login into Youtube
1158 data = compat_urllib_parse.urlencode({'video_id': video_id,
1159 'el': 'player_embedded',
1162 'eurl': 'https://youtube.googleapis.com/v/' + video_id,
1166 video_info_url = 'https://www.youtube.com/get_video_info?' + data
1167 video_info_webpage = self._download_webpage(video_info_url, video_id,
1169 errnote='unable to download video info webpage')
1170 video_info = compat_parse_qs(video_info_webpage)
1173 for el_type in ['&el=embedded', '&el=detailpage', '&el=vevo', '']:
1174 video_info_url = ('https://www.youtube.com/get_video_info?&video_id=%s%s&ps=default&eurl=&gl=US&hl=en'
1175 % (video_id, el_type))
1176 video_info_webpage = self._download_webpage(video_info_url, video_id,
1178 errnote='unable to download video info webpage')
1179 video_info = compat_parse_qs(video_info_webpage)
1180 if 'token' in video_info:
1182 if 'token' not in video_info:
1183 if 'reason' in video_info:
1184 raise ExtractorError(u'YouTube said: %s' % video_info['reason'][0], expected=True)
1186 raise ExtractorError(u'"token" parameter not in video info for unknown reason')
1188 if 'view_count' in video_info:
1189 view_count = int(video_info['view_count'][0])
1193 # Check for "rental" videos
1194 if 'ypc_video_rental_bar_text' in video_info and 'author' not in video_info:
1195 raise ExtractorError(u'"rental" videos not supported')
1197 # Start extracting information
1198 self.report_information_extraction(video_id)
1201 if 'author' not in video_info:
1202 raise ExtractorError(u'Unable to extract uploader name')
1203 video_uploader = compat_urllib_parse.unquote_plus(video_info['author'][0])
1206 video_uploader_id = None
1207 mobj = re.search(r'<link itemprop="url" href="http://www.youtube.com/(?:user|channel)/([^"]+)">', video_webpage)
1208 if mobj is not None:
1209 video_uploader_id = mobj.group(1)
1211 self._downloader.report_warning(u'unable to extract uploader nickname')
1214 if 'title' in video_info:
1215 video_title = compat_urllib_parse.unquote_plus(video_info['title'][0])
1217 self._downloader.report_warning(u'Unable to extract video title')
1221 # We try first to get a high quality image:
1222 m_thumb = re.search(r'<span itemprop="thumbnail".*?href="(.*?)">',
1223 video_webpage, re.DOTALL)
1224 if m_thumb is not None:
1225 video_thumbnail = m_thumb.group(1)
1226 elif 'thumbnail_url' not in video_info:
1227 self._downloader.report_warning(u'unable to extract video thumbnail')
1228 video_thumbnail = None
1229 else: # don't panic if we can't find it
1230 video_thumbnail = compat_urllib_parse.unquote_plus(video_info['thumbnail_url'][0])
1234 mobj = re.search(r'id="eow-date.*?>(.*?)</span>', video_webpage, re.DOTALL)
1235 if mobj is not None:
1236 upload_date = ' '.join(re.sub(r'[/,-]', r' ', mobj.group(1)).split())
1237 upload_date = unified_strdate(upload_date)
1240 video_description = get_element_by_id("eow-description", video_webpage)
1241 if video_description:
1242 video_description = re.sub(r'''(?x)
1244 (?:[a-zA-Z-]+="[^"]+"\s+)*?
1246 (?:[a-zA-Z-]+="[^"]+"\s+)*?
1247 class="yt-uix-redirect-link"\s*>
1250 ''', r'\1', video_description)
1251 video_description = clean_html(video_description)
1253 fd_mobj = re.search(r'<meta name="description" content="([^"]+)"', video_webpage)
1255 video_description = unescapeHTML(fd_mobj.group(1))
1257 video_description = u''
1259 def _extract_count(klass):
1260 count = self._search_regex(
1261 r'class="%s">([\d,]+)</span>' % re.escape(klass),
1262 video_webpage, klass, default=None)
1263 if count is not None:
1264 return int(count.replace(',', ''))
1266 like_count = _extract_count(u'likes-count')
1267 dislike_count = _extract_count(u'dislikes-count')
1270 video_subtitles = self.extract_subtitles(video_id, video_webpage)
1272 if self._downloader.params.get('listsubtitles', False):
1273 self._list_available_subtitles(video_id, video_webpage)
1276 if 'length_seconds' not in video_info:
1277 self._downloader.report_warning(u'unable to extract video duration')
1278 video_duration = None
1280 video_duration = int(compat_urllib_parse.unquote_plus(video_info['length_seconds'][0]))
1283 video_annotations = None
1284 if self._downloader.params.get('writeannotations', False):
1285 video_annotations = self._extract_annotations(video_id)
1287 # Decide which formats to download
1289 mobj = re.search(r';ytplayer.config = ({.*?});', video_webpage)
1291 raise ValueError('Could not find vevo ID')
1292 ytplayer_config = json.loads(mobj.group(1))
1293 args = ytplayer_config['args']
1294 # Easy way to know if the 's' value is in url_encoded_fmt_stream_map
1295 # this signatures are encrypted
1296 if 'url_encoded_fmt_stream_map' not in args:
1297 raise ValueError(u'No stream_map present') # caught below
1298 re_signature = re.compile(r'[&,]s=')
1299 m_s = re_signature.search(args['url_encoded_fmt_stream_map'])
1301 self.to_screen(u'%s: Encrypted signatures detected.' % video_id)
1302 video_info['url_encoded_fmt_stream_map'] = [args['url_encoded_fmt_stream_map']]
1303 m_s = re_signature.search(args.get('adaptive_fmts', u''))
1305 if 'adaptive_fmts' in video_info:
1306 video_info['adaptive_fmts'][0] += ',' + args['adaptive_fmts']
1308 video_info['adaptive_fmts'] = [args['adaptive_fmts']]
1312 def _map_to_format_list(urlmap):
1314 for itag, video_real_url in urlmap.items():
1317 'url': video_real_url,
1318 'player_url': player_url,
1320 if itag in self._formats:
1321 dct.update(self._formats[itag])
1325 if 'conn' in video_info and video_info['conn'][0].startswith('rtmp'):
1326 self.report_rtmp_download()
1328 'format_id': '_rtmp',
1330 'url': video_info['conn'][0],
1331 'player_url': player_url,
1333 elif len(video_info.get('url_encoded_fmt_stream_map', [])) >= 1 or len(video_info.get('adaptive_fmts', [])) >= 1:
1334 encoded_url_map = video_info.get('url_encoded_fmt_stream_map', [''])[0] + ',' + video_info.get('adaptive_fmts',[''])[0]
1335 if 'rtmpe%3Dyes' in encoded_url_map:
1336 raise ExtractorError('rtmpe downloads are not supported, see https://github.com/rg3/youtube-dl/issues/343 for more information.', expected=True)
1338 for url_data_str in encoded_url_map.split(','):
1339 url_data = compat_parse_qs(url_data_str)
1340 if 'itag' in url_data and 'url' in url_data:
1341 url = url_data['url'][0]
1342 if 'sig' in url_data:
1343 url += '&signature=' + url_data['sig'][0]
1344 elif 's' in url_data:
1345 encrypted_sig = url_data['s'][0]
1346 if self._downloader.params.get('verbose'):
1348 if player_url is None:
1349 player_version = 'unknown'
1351 player_version = self._search_regex(
1352 r'-(.+)\.swf$', player_url,
1353 u'flash player', fatal=False)
1354 player_desc = 'flash player %s' % player_version
1356 player_version = self._search_regex(
1357 r'html5player-(.+?)\.js', video_webpage,
1358 'html5 player', fatal=False)
1359 player_desc = u'html5 player %s' % player_version
1361 parts_sizes = u'.'.join(compat_str(len(part)) for part in encrypted_sig.split('.'))
1362 self.to_screen(u'encrypted signature length %d (%s), itag %s, %s' %
1363 (len(encrypted_sig), parts_sizes, url_data['itag'][0], player_desc))
1366 jsplayer_url_json = self._search_regex(
1367 r'"assets":.+?"js":\s*("[^"]+")',
1368 video_webpage, u'JS player URL')
1369 player_url = json.loads(jsplayer_url_json)
1371 signature = self._decrypt_signature(
1372 encrypted_sig, video_id, player_url, age_gate)
1373 url += '&signature=' + signature
1374 if 'ratebypass' not in url:
1375 url += '&ratebypass=yes'
1376 url_map[url_data['itag'][0]] = url
1377 formats = _map_to_format_list(url_map)
1378 elif video_info.get('hlsvp'):
1379 manifest_url = video_info['hlsvp'][0]
1380 url_map = self._extract_from_m3u8(manifest_url, video_id)
1381 formats = _map_to_format_list(url_map)
1383 raise ExtractorError(u'no conn, hlsvp or url_encoded_fmt_stream_map information found in video info')
1385 # Look for the DASH manifest
1386 if (self._downloader.params.get('youtube_include_dash_manifest', False)):
1388 # The DASH manifest used needs to be the one from the original video_webpage.
1389 # The one found in get_video_info seems to be using different signatures.
1390 # However, in the case of an age restriction there won't be any embedded dashmpd in the video_webpage.
1391 # Luckily, it seems, this case uses some kind of default signature (len == 86), so the
1392 # combination of get_video_info and the _static_decrypt_signature() decryption fallback will work here.
1394 dash_manifest_url = video_info.get('dashmpd')[0]
1396 dash_manifest_url = ytplayer_config['args']['dashmpd']
1397 def decrypt_sig(mobj):
1399 dec_s = self._decrypt_signature(s, video_id, player_url, age_gate)
1400 return '/signature/%s' % dec_s
1401 dash_manifest_url = re.sub(r'/s/([\w\.]+)', decrypt_sig, dash_manifest_url)
1402 dash_doc = self._download_xml(
1403 dash_manifest_url, video_id,
1404 note=u'Downloading DASH manifest',
1405 errnote=u'Could not download DASH manifest')
1406 for r in dash_doc.findall(u'.//{urn:mpeg:DASH:schema:MPD:2011}Representation'):
1407 url_el = r.find('{urn:mpeg:DASH:schema:MPD:2011}BaseURL')
1410 format_id = r.attrib['id']
1411 video_url = url_el.text
1412 filesize = int_or_none(url_el.attrib.get('{http://youtube.com/yt/2012/10/10}contentLength'))
1414 'format_id': format_id,
1416 'width': int_or_none(r.attrib.get('width')),
1417 'tbr': int_or_none(r.attrib.get('bandwidth'), 1000),
1418 'asr': int_or_none(r.attrib.get('audioSamplingRate')),
1419 'filesize': filesize,
1422 existing_format = next(
1423 fo for fo in formats
1424 if fo['format_id'] == format_id)
1425 except StopIteration:
1426 f.update(self._formats.get(format_id, {}))
1429 existing_format.update(f)
1431 except (ExtractorError, KeyError) as e:
1432 self.report_warning(u'Skipping DASH manifest: %s' % e, video_id)
1434 self._sort_formats(formats)
1438 'uploader': video_uploader,
1439 'uploader_id': video_uploader_id,
1440 'upload_date': upload_date,
1441 'title': video_title,
1442 'thumbnail': video_thumbnail,
1443 'description': video_description,
1444 'subtitles': video_subtitles,
1445 'duration': video_duration,
1446 'age_limit': 18 if age_gate else 0,
1447 'annotations': video_annotations,
1448 'webpage_url': 'https://www.youtube.com/watch?v=%s' % video_id,
1449 'view_count': view_count,
1450 'like_count': like_count,
1451 'dislike_count': dislike_count,
1455 class YoutubePlaylistIE(YoutubeBaseInfoExtractor):
1456 IE_DESC = u'YouTube.com playlists'
1457 _VALID_URL = r"""(?x)(?:
1462 (?:course|view_play_list|my_playlists|artist|playlist|watch)
1463 \? (?:.*?&)*? (?:p|a|list)=
1467 (?:PL|EC|UU|FL|RD)?[0-9A-Za-z-_]{10,}
1468 # Top tracks, they can also include dots
1473 ((?:PL|EC|UU|FL|RD)[0-9A-Za-z-_]{10,})
1475 _TEMPLATE_URL = 'https://www.youtube.com/playlist?list=%s'
1476 _MORE_PAGES_INDICATOR = r'data-link-type="next"'
1477 _VIDEO_RE = r'href="\s*/watch\?v=(?P<id>[0-9A-Za-z_-]{11})&[^"]*?index=(?P<index>\d+)'
1478 IE_NAME = u'youtube:playlist'
1480 def _real_initialize(self):
1483 def _ids_to_results(self, ids):
1484 return [self.url_result(vid_id, 'Youtube', video_id=vid_id)
1487 def _extract_mix(self, playlist_id):
1488 # The mixes are generated from a a single video
1489 # the id of the playlist is just 'RD' + video_id
1490 url = 'https://youtube.com/watch?v=%s&list=%s' % (playlist_id[-11:], playlist_id)
1491 webpage = self._download_webpage(url, playlist_id, u'Downloading Youtube mix')
1492 title_span = (get_element_by_attribute('class', 'title long-title', webpage) or
1493 get_element_by_attribute('class', 'title ', webpage))
1494 title = clean_html(title_span)
1495 video_re = r'data-index="\d+".*?href="/watch\?v=([0-9A-Za-z_-]{11})&[^"]*?list=%s' % re.escape(playlist_id)
1496 ids = orderedSet(re.findall(video_re, webpage))
1497 url_results = self._ids_to_results(ids)
1499 return self.playlist_result(url_results, playlist_id, title)
1501 def _real_extract(self, url):
1502 # Extract playlist id
1503 mobj = re.match(self._VALID_URL, url)
1505 raise ExtractorError(u'Invalid URL: %s' % url)
1506 playlist_id = mobj.group(1) or mobj.group(2)
1508 # Check if it's a video-specific URL
1509 query_dict = compat_urlparse.parse_qs(compat_urlparse.urlparse(url).query)
1510 if 'v' in query_dict:
1511 video_id = query_dict['v'][0]
1512 if self._downloader.params.get('noplaylist'):
1513 self.to_screen(u'Downloading just video %s because of --no-playlist' % video_id)
1514 return self.url_result(video_id, 'Youtube', video_id=video_id)
1516 self.to_screen(u'Downloading playlist PL%s - add --no-playlist to just download video %s' % (playlist_id, video_id))
1518 if playlist_id.startswith('RD'):
1519 # Mixes require a custom extraction process
1520 return self._extract_mix(playlist_id)
1521 if playlist_id.startswith('TL'):
1522 raise ExtractorError(u'For downloading YouTube.com top lists, use '
1523 u'the "yttoplist" keyword, for example "youtube-dl \'yttoplist:music:Top Tracks\'"', expected=True)
1525 url = self._TEMPLATE_URL % playlist_id
1526 page = self._download_webpage(url, playlist_id)
1527 more_widget_html = content_html = page
1529 # Extract the video ids from the playlist pages
1532 for page_num in itertools.count(1):
1533 matches = re.finditer(self._VIDEO_RE, content_html)
1534 # We remove the duplicates and the link with index 0
1535 # (it's not the first video of the playlist)
1536 new_ids = orderedSet(m.group('id') for m in matches if m.group('index') != '0')
1539 mobj = re.search(r'data-uix-load-more-href="/?(?P<more>[^"]+)"', more_widget_html)
1543 more = self._download_json(
1544 'https://youtube.com/%s' % mobj.group('more'), playlist_id, 'Downloading page #%s' % page_num)
1545 content_html = more['content_html']
1546 more_widget_html = more['load_more_widget_html']
1548 playlist_title = self._html_search_regex(
1549 r'<h1 class="pl-header-title">\s*(.*?)\s*</h1>', page, u'title')
1551 url_results = self._ids_to_results(ids)
1552 return self.playlist_result(url_results, playlist_id, playlist_title)
1555 class YoutubeTopListIE(YoutubePlaylistIE):
1556 IE_NAME = u'youtube:toplist'
1557 IE_DESC = (u'YouTube.com top lists, "yttoplist:{channel}:{list title}"'
1558 u' (Example: "yttoplist:music:Top Tracks")')
1559 _VALID_URL = r'yttoplist:(?P<chann>.*?):(?P<title>.*?)$'
1561 def _real_extract(self, url):
1562 mobj = re.match(self._VALID_URL, url)
1563 channel = mobj.group('chann')
1564 title = mobj.group('title')
1565 query = compat_urllib_parse.urlencode({'title': title})
1566 playlist_re = 'href="([^"]+?%s.*?)"' % re.escape(query)
1567 channel_page = self._download_webpage('https://www.youtube.com/%s' % channel, title)
1568 link = self._html_search_regex(playlist_re, channel_page, u'list')
1569 url = compat_urlparse.urljoin('https://www.youtube.com/', link)
1571 video_re = r'data-index="\d+".*?data-video-id="([0-9A-Za-z_-]{11})"'
1573 # sometimes the webpage doesn't contain the videos
1574 # retry until we get them
1575 for i in itertools.count(0):
1576 msg = u'Downloading Youtube mix'
1578 msg += ', retry #%d' % i
1579 webpage = self._download_webpage(url, title, msg)
1580 ids = orderedSet(re.findall(video_re, webpage))
1583 url_results = self._ids_to_results(ids)
1584 return self.playlist_result(url_results, playlist_title=title)
1587 class YoutubeChannelIE(InfoExtractor):
1588 IE_DESC = u'YouTube.com channels'
1589 _VALID_URL = r"^(?:https?://)?(?:youtu\.be|(?:\w+\.)?youtube(?:-nocookie)?\.com)/channel/([0-9A-Za-z_-]+)"
1590 _MORE_PAGES_INDICATOR = 'yt-uix-load-more'
1591 _MORE_PAGES_URL = 'https://www.youtube.com/c4_browse_ajax?action_load_more_videos=1&flow=list&paging=%s&view=0&sort=da&channel_id=%s'
1592 IE_NAME = u'youtube:channel'
1594 def extract_videos_from_page(self, page):
1596 for mobj in re.finditer(r'href="/watch\?v=([0-9A-Za-z_-]+)&?', page):
1597 if mobj.group(1) not in ids_in_page:
1598 ids_in_page.append(mobj.group(1))
1601 def _real_extract(self, url):
1602 # Extract channel id
1603 mobj = re.match(self._VALID_URL, url)
1605 raise ExtractorError(u'Invalid URL: %s' % url)
1607 # Download channel page
1608 channel_id = mobj.group(1)
1610 url = 'https://www.youtube.com/channel/%s/videos' % channel_id
1611 channel_page = self._download_webpage(url, channel_id)
1612 autogenerated = re.search(r'''(?x)
1614 channel-header-autogenerated-label|
1615 yt-channel-title-autogenerated
1616 )[^"]*"''', channel_page) is not None
1619 # The videos are contained in a single page
1620 # the ajax pages can't be used, they are empty
1621 video_ids = self.extract_videos_from_page(channel_page)
1623 # Download all channel pages using the json-based channel_ajax query
1624 for pagenum in itertools.count(1):
1625 url = self._MORE_PAGES_URL % (pagenum, channel_id)
1626 page = self._download_json(
1627 url, channel_id, note=u'Downloading page #%s' % pagenum,
1628 transform_source=uppercase_escape)
1630 ids_in_page = self.extract_videos_from_page(page['content_html'])
1631 video_ids.extend(ids_in_page)
1633 if self._MORE_PAGES_INDICATOR not in page['load_more_widget_html']:
1636 self._downloader.to_screen(u'[youtube] Channel %s: Found %i videos' % (channel_id, len(video_ids)))
1638 url_entries = [self.url_result(video_id, 'Youtube', video_id=video_id)
1639 for video_id in video_ids]
1640 return self.playlist_result(url_entries, channel_id)
1643 class YoutubeUserIE(InfoExtractor):
1644 IE_DESC = u'YouTube.com user videos (URL or "ytuser" keyword)'
1645 _VALID_URL = r'(?:(?:(?:https?://)?(?:\w+\.)?youtube\.com/(?:user/)?(?!(?:attribution_link|watch)(?:$|[^a-z_A-Z0-9-])))|ytuser:)(?!feed/)([A-Za-z0-9_-]+)'
1646 _TEMPLATE_URL = 'https://gdata.youtube.com/feeds/api/users/%s'
1647 _GDATA_PAGE_SIZE = 50
1648 _GDATA_URL = 'https://gdata.youtube.com/feeds/api/users/%s/uploads?max-results=%d&start-index=%d&alt=json'
1649 IE_NAME = u'youtube:user'
1652 def suitable(cls, url):
1653 # Don't return True if the url can be extracted with other youtube
1654 # extractor, the regex would is too permissive and it would match.
1655 other_ies = iter(klass for (name, klass) in globals().items() if name.endswith('IE') and klass is not cls)
1656 if any(ie.suitable(url) for ie in other_ies): return False
1657 else: return super(YoutubeUserIE, cls).suitable(url)
1659 def _real_extract(self, url):
1661 mobj = re.match(self._VALID_URL, url)
1663 raise ExtractorError(u'Invalid URL: %s' % url)
1665 username = mobj.group(1)
1667 # Download video ids using YouTube Data API. Result size per
1668 # query is limited (currently to 50 videos) so we need to query
1669 # page by page until there are no video ids - it means we got
1672 def download_page(pagenum):
1673 start_index = pagenum * self._GDATA_PAGE_SIZE + 1
1675 gdata_url = self._GDATA_URL % (username, self._GDATA_PAGE_SIZE, start_index)
1676 page = self._download_webpage(
1677 gdata_url, username,
1678 u'Downloading video ids from %d to %d' % (
1679 start_index, start_index + self._GDATA_PAGE_SIZE))
1682 response = json.loads(page)
1683 except ValueError as err:
1684 raise ExtractorError(u'Invalid JSON in API response: ' + compat_str(err))
1685 if 'entry' not in response['feed']:
1688 # Extract video identifiers
1689 entries = response['feed']['entry']
1690 for entry in entries:
1691 title = entry['title']['$t']
1692 video_id = entry['id']['$t'].split('/')[-1]
1696 'ie_key': 'Youtube',
1700 url_results = PagedList(download_page, self._GDATA_PAGE_SIZE)
1702 return self.playlist_result(url_results, playlist_title=username)
1705 class YoutubeSearchIE(SearchInfoExtractor):
1706 IE_DESC = u'YouTube.com searches'
1707 _API_URL = 'https://gdata.youtube.com/feeds/api/videos?q=%s&start-index=%i&max-results=50&v=2&alt=jsonc'
1709 IE_NAME = u'youtube:search'
1710 _SEARCH_KEY = 'ytsearch'
1712 def _get_n_results(self, query, n):
1713 """Get a specified number of results for a query"""
1719 while (50 * pagenum) < limit:
1720 result_url = self._API_URL % (compat_urllib_parse.quote_plus(query), (50*pagenum)+1)
1721 data_json = self._download_webpage(
1722 result_url, video_id=u'query "%s"' % query,
1723 note=u'Downloading page %s' % (pagenum + 1),
1724 errnote=u'Unable to download API page')
1725 data = json.loads(data_json)
1726 api_response = data['data']
1728 if 'items' not in api_response:
1729 raise ExtractorError(
1730 u'[youtube] No video results', expected=True)
1732 new_ids = list(video['id'] for video in api_response['items'])
1733 video_ids += new_ids
1735 limit = min(n, api_response['totalItems'])
1738 if len(video_ids) > n:
1739 video_ids = video_ids[:n]
1740 videos = [self.url_result(video_id, 'Youtube', video_id=video_id)
1741 for video_id in video_ids]
1742 return self.playlist_result(videos, query)
1744 class YoutubeSearchDateIE(YoutubeSearchIE):
1745 IE_NAME = YoutubeSearchIE.IE_NAME + ':date'
1746 _API_URL = 'https://gdata.youtube.com/feeds/api/videos?q=%s&start-index=%i&max-results=50&v=2&alt=jsonc&orderby=published'
1747 _SEARCH_KEY = 'ytsearchdate'
1748 IE_DESC = u'YouTube.com searches, newest videos first'
1750 class YoutubeShowIE(InfoExtractor):
1751 IE_DESC = u'YouTube.com (multi-season) shows'
1752 _VALID_URL = r'https?://www\.youtube\.com/show/(.*)'
1753 IE_NAME = u'youtube:show'
1755 def _real_extract(self, url):
1756 mobj = re.match(self._VALID_URL, url)
1757 show_name = mobj.group(1)
1758 webpage = self._download_webpage(url, show_name, u'Downloading show webpage')
1759 # There's one playlist for each season of the show
1760 m_seasons = list(re.finditer(r'href="(/playlist\?list=.*?)"', webpage))
1761 self.to_screen(u'%s: Found %s seasons' % (show_name, len(m_seasons)))
1762 return [self.url_result('https://www.youtube.com' + season.group(1), 'YoutubePlaylist') for season in m_seasons]
1765 class YoutubeFeedsInfoExtractor(YoutubeBaseInfoExtractor):
1767 Base class for extractors that fetch info from
1768 http://www.youtube.com/feed_ajax
1769 Subclasses must define the _FEED_NAME and _PLAYLIST_TITLE properties.
1771 _LOGIN_REQUIRED = True
1772 # use action_load_personal_feed instead of action_load_system_feed
1773 _PERSONAL_FEED = False
1776 def _FEED_TEMPLATE(self):
1777 action = 'action_load_system_feed'
1778 if self._PERSONAL_FEED:
1779 action = 'action_load_personal_feed'
1780 return 'https://www.youtube.com/feed_ajax?%s=1&feed_name=%s&paging=%%s' % (action, self._FEED_NAME)
1784 return u'youtube:%s' % self._FEED_NAME
1786 def _real_initialize(self):
1789 def _real_extract(self, url):
1792 for i in itertools.count(1):
1793 info = self._download_webpage(self._FEED_TEMPLATE % paging,
1794 u'%s feed' % self._FEED_NAME,
1795 u'Downloading page %s' % i)
1796 info = json.loads(info)
1797 feed_html = info['feed_html']
1798 m_ids = re.finditer(r'"/watch\?v=(.*?)["&]', feed_html)
1799 ids = orderedSet(m.group(1) for m in m_ids)
1800 feed_entries.extend(
1801 self.url_result(video_id, 'Youtube', video_id=video_id)
1802 for video_id in ids)
1803 if info['paging'] is None:
1805 paging = info['paging']
1806 return self.playlist_result(feed_entries, playlist_title=self._PLAYLIST_TITLE)
1808 class YoutubeSubscriptionsIE(YoutubeFeedsInfoExtractor):
1809 IE_DESC = u'YouTube.com subscriptions feed, "ytsubs" keyword(requires authentication)'
1810 _VALID_URL = r'https?://www\.youtube\.com/feed/subscriptions|:ytsubs(?:criptions)?'
1811 _FEED_NAME = 'subscriptions'
1812 _PLAYLIST_TITLE = u'Youtube Subscriptions'
1814 class YoutubeRecommendedIE(YoutubeFeedsInfoExtractor):
1815 IE_DESC = u'YouTube.com recommended videos, "ytrec" keyword (requires authentication)'
1816 _VALID_URL = r'https?://www\.youtube\.com/feed/recommended|:ytrec(?:ommended)?'
1817 _FEED_NAME = 'recommended'
1818 _PLAYLIST_TITLE = u'Youtube Recommended videos'
1820 class YoutubeWatchLaterIE(YoutubeFeedsInfoExtractor):
1821 IE_DESC = u'Youtube watch later list, "ytwatchlater" keyword (requires authentication)'
1822 _VALID_URL = r'https?://www\.youtube\.com/feed/watch_later|:ytwatchlater'
1823 _FEED_NAME = 'watch_later'
1824 _PLAYLIST_TITLE = u'Youtube Watch Later'
1825 _PERSONAL_FEED = True
1827 class YoutubeHistoryIE(YoutubeFeedsInfoExtractor):
1828 IE_DESC = u'Youtube watch history, "ythistory" keyword (requires authentication)'
1829 _VALID_URL = u'https?://www\.youtube\.com/feed/history|:ythistory'
1830 _FEED_NAME = 'history'
1831 _PERSONAL_FEED = True
1832 _PLAYLIST_TITLE = u'Youtube Watch History'
1834 class YoutubeFavouritesIE(YoutubeBaseInfoExtractor):
1835 IE_NAME = u'youtube:favorites'
1836 IE_DESC = u'YouTube.com favourite videos, "ytfav" keyword (requires authentication)'
1837 _VALID_URL = r'https?://www\.youtube\.com/my_favorites|:ytfav(?:ou?rites)?'
1838 _LOGIN_REQUIRED = True
1840 def _real_extract(self, url):
1841 webpage = self._download_webpage('https://www.youtube.com/my_favorites', 'Youtube Favourites videos')
1842 playlist_id = self._search_regex(r'list=(.+?)["&]', webpage, u'favourites playlist id')
1843 return self.url_result(playlist_id, 'YoutubePlaylist')
1846 class YoutubeTruncatedURLIE(InfoExtractor):
1847 IE_NAME = 'youtube:truncated_url'
1848 IE_DESC = False # Do not list
1849 _VALID_URL = r'''(?x)
1850 (?:https?://)?[^/]+/watch\?(?:feature=[a-z_]+)?$|
1851 (?:https?://)?(?:www\.)?youtube\.com/attribution_link\?a=[^&]+$
1854 def _real_extract(self, url):
1855 raise ExtractorError(
1856 u'Did you forget to quote the URL? Remember that & is a meta '
1857 u'character in most shells, so you want to put the URL in quotes, '
1859 u'"http://www.youtube.com/watch?feature=foo&v=BaW_jenozKc" '
1860 u' or simply youtube-dl BaW_jenozKc .',