14 from .common import InfoExtractor, SearchInfoExtractor
15 from .subtitles import SubtitlesInfoExtractor
16 from ..jsinterp import JSInterpreter
21 compat_urllib_request,
28 get_element_by_attribute,
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 = 'https://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
48 def _set_language(self):
49 return bool(self._download_webpage(
51 note=u'Setting language', errnote='unable to set language',
55 (username, password) = self._get_login_info()
56 # No authentication to be performed
58 if self._LOGIN_REQUIRED:
59 raise ExtractorError(u'No login info available, needed for using %s.' % self.IE_NAME, expected=True)
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:
69 galx = self._search_regex(r'(?s)<input.+?name="GALX".+?value="(.+?)"',
70 login_page, u'Login GALX parameter')
74 u'continue': u'https://www.youtube.com/signin?action_handle_signin=true&feature=sign_in_button&hl=en_US&nomobiletemp=1',
78 u'PersistentCookie': u'yes',
80 u'bgresponse': u'js_disabled',
81 u'checkConnection': u'',
82 u'checkedDomains': u'youtube',
87 u'signIn': u'Sign in',
89 u'service': u'youtube',
93 # Convert to UTF-8 *before* urlencode because Python 2.x's urlencode
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')
98 req = compat_urllib_request.Request(self._LOGIN_URL, login_data)
99 login_results = self._download_webpage(
101 note=u'Logging in', errnote=u'unable to log in', fatal=False)
102 if login_results is 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')
109 def _confirm_age(self):
112 'action_confirm': 'Confirm',
114 req = compat_urllib_request.Request(self._AGE_URL,
115 compat_urllib_parse.urlencode(age_form).encode('ascii'))
117 self._download_webpage(
119 note=u'Confirming age', errnote=u'Unable to confirm age')
122 def _real_initialize(self):
123 if self._downloader is None:
125 if not self._set_language():
127 if not self._login():
132 class YoutubeIE(YoutubeBaseInfoExtractor, SubtitlesInfoExtractor):
133 IE_DESC = u'YouTube.com'
134 _VALID_URL = r"""(?x)^
136 (?:https?://|//)? # http(s):// or protocol-independent URL (optional)
137 (?:(?:(?:(?:\w+\.)?[yY][oO][uU][tT][uU][bB][eE](?:-nocookie)?\.com/|
138 (?:www\.)?deturl\.com/www\.youtube\.com/|
139 (?:www\.)?pwnyoutube\.com/|
140 (?:www\.)?yourepeat\.com/|
141 tube\.majestyc\.net/|
142 youtube\.googleapis\.com/) # the various hostnames, with wildcard subdomains
143 (?:.*?\#/)? # handle anchor (#/) redirect urls
144 (?: # the various things that can precede the ID:
145 (?:(?:v|embed|e)/) # v/ or embed/ or e/
146 |(?: # or the v= param in all its forms
147 (?:(?:watch|movie)(?:_popup)?(?:\.php)?/?)? # preceding watch(_popup|.php) or nothing (like /?v=xxxx)
148 (?:\?|\#!?) # the params delimiter ? or # or #!
149 (?:.*?&)? # any other preceding param (like /?s=tuff&v=xxxx)
153 |youtu\.be/ # just youtu.be/xxxx
154 |https?://(?:www\.)?cleanvideosearch\.com/media/action/yt/watch\?videoId=
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, 'format_note': '3D', 'preference': -20},
181 '83': {'ext': 'mp4', 'height': 480, 'format_note': '3D', 'preference': -20},
182 '84': {'ext': 'mp4', 'height': 720, 'format_note': '3D', 'preference': -20},
183 '85': {'ext': 'mp4', 'height': 1080, 'format_note': '3D', 'preference': -20},
184 '100': {'ext': 'webm', 'height': 360, 'format_note': '3D', 'preference': -20},
185 '101': {'ext': 'webm', 'height': 480, 'format_note': '3D', 'preference': -20},
186 '102': {'ext': 'webm', 'height': 720, 'format_note': '3D', 'preference': -20},
188 # Apple HTTP Live Streaming
189 '92': {'ext': 'mp4', 'height': 240, 'format_note': 'HLS', 'preference': -10},
190 '93': {'ext': 'mp4', 'height': 360, 'format_note': 'HLS', 'preference': -10},
191 '94': {'ext': 'mp4', 'height': 480, 'format_note': 'HLS', 'preference': -10},
192 '95': {'ext': 'mp4', 'height': 720, 'format_note': 'HLS', 'preference': -10},
193 '96': {'ext': 'mp4', 'height': 1080, 'format_note': 'HLS', 'preference': -10},
194 '132': {'ext': 'mp4', 'height': 240, 'format_note': 'HLS', 'preference': -10},
195 '151': {'ext': 'mp4', 'height': 72, 'format_note': 'HLS', 'preference': -10},
198 '133': {'ext': 'mp4', 'height': 240, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
199 '134': {'ext': 'mp4', 'height': 360, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
200 '135': {'ext': 'mp4', 'height': 480, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
201 '136': {'ext': 'mp4', 'height': 720, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
202 '137': {'ext': 'mp4', 'height': 1080, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
203 '138': {'ext': 'mp4', 'height': 2160, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
204 '160': {'ext': 'mp4', 'height': 144, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
205 '264': {'ext': 'mp4', 'height': 1440, 'format_note': 'DASH video', 'acodec': 'none', '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', 'acodec': 'none', 'container': 'webm', 'vcodec': 'VP8', 'preference': -40},
214 '168': {'ext': 'webm', 'height': 480, 'width': 854, 'format_note': 'DASH video', 'acodec': 'none', 'container': 'webm', 'vcodec': 'VP8', 'preference': -40},
215 '169': {'ext': 'webm', 'height': 720, 'width': 1280, 'format_note': 'DASH video', 'acodec': 'none', 'container': 'webm', 'vcodec': 'VP8', 'preference': -40},
216 '170': {'ext': 'webm', 'height': 1080, 'width': 1920, 'format_note': 'DASH video', 'acodec': 'none', 'container': 'webm', 'vcodec': 'VP8', 'preference': -40},
217 '218': {'ext': 'webm', 'height': 480, 'width': 854, 'format_note': 'DASH video', 'acodec': 'none', 'container': 'webm', 'vcodec': 'VP8', 'preference': -40},
218 '219': {'ext': 'webm', 'height': 480, 'width': 854, 'format_note': 'DASH video', 'acodec': 'none', 'container': 'webm', 'vcodec': 'VP8', 'preference': -40},
219 '242': {'ext': 'webm', 'height': 240, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
220 '243': {'ext': 'webm', 'height': 360, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
221 '244': {'ext': 'webm', 'height': 480, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
222 '245': {'ext': 'webm', 'height': 480, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
223 '246': {'ext': 'webm', 'height': 480, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
224 '247': {'ext': 'webm', 'height': 720, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
225 '248': {'ext': 'webm', 'height': 1080, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
228 '171': {'ext': 'webm', 'vcodec': 'none', 'format_note': 'DASH audio', 'abr': 48, 'preference': -50},
229 '172': {'ext': 'webm', 'vcodec': 'none', 'format_note': 'DASH 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:fea86fda2d5a5784273df5c7cc994d9f",
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'md5:9717375db5a9a3992be4668bbf3bc0a8',
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')
444 jsi = JSInterpreter(jscode)
445 initial_function = jsi.extract_function(funcname)
446 return lambda s: initial_function([s])
448 def _parse_sig_swf(self, file_contents):
449 if file_contents[1:3] != b'WS':
450 raise ExtractorError(
451 u'Not an SWF file; header is %r' % file_contents[:3])
452 if file_contents[:1] == b'C':
453 content = zlib.decompress(file_contents[8:])
455 raise NotImplementedError(u'Unsupported compression format %r' %
458 def extract_tags(content):
460 while pos < len(content):
461 header16 = struct.unpack('<H', content[pos:pos+2])[0]
463 tag_code = header16 >> 6
464 tag_len = header16 & 0x3f
466 tag_len = struct.unpack('<I', content[pos:pos+4])[0]
468 assert pos+tag_len <= len(content)
469 yield (tag_code, content[pos:pos+tag_len])
473 for tag_code, tag in extract_tags(content)
475 p = code_tag.index(b'\0', 4) + 1
476 code_reader = io.BytesIO(code_tag[p:])
478 # Parse ABC (AVM2 ByteCode)
479 def read_int(reader=None):
487 b = struct.unpack('<B', buf)[0]
488 res = res | ((b & 0x7f) << shift)
494 def u30(reader=None):
495 res = read_int(reader)
496 assert res & 0xf0000000 == 0
500 def s32(reader=None):
502 if v & 0x80000000 != 0:
503 v = - ((v ^ 0xffffffff) + 1)
506 def read_string(reader=None):
510 resb = reader.read(slen)
511 assert len(resb) == slen
512 return resb.decode('utf-8')
514 def read_bytes(count, reader=None):
517 resb = reader.read(count)
518 assert len(resb) == count
521 def read_byte(reader=None):
522 resb = read_bytes(1, reader=reader)
523 res = struct.unpack('<B', resb)[0]
526 # minor_version + major_version
531 for _c in range(1, int_count):
534 for _c in range(1, uint_count):
537 read_bytes((double_count-1) * 8)
539 constant_strings = [u'']
540 for _c in range(1, string_count):
542 constant_strings.append(s)
543 namespace_count = u30()
544 for _c in range(1, namespace_count):
548 for _c in range(1, ns_set_count):
550 for _c2 in range(count):
552 multiname_count = u30()
561 0x0e: 2, # MultinameA
562 0x1b: 1, # MultinameL
563 0x1c: 1, # MultinameLA
566 for _c in range(1, multiname_count):
568 assert kind in MULTINAME_SIZES, u'Invalid multiname kind %r' % kind
570 u30() # namespace_idx
572 multinames.append(constant_strings[name_idx])
574 multinames.append('[MULTINAME kind: %d]' % kind)
575 for _c2 in range(MULTINAME_SIZES[kind]):
580 MethodInfo = collections.namedtuple(
582 ['NEED_ARGUMENTS', 'NEED_REST'])
584 for method_id in range(method_count):
587 for _ in range(param_count):
589 u30() # name index (always 0 for youtube)
591 if flags & 0x08 != 0:
594 for c in range(option_count):
597 if flags & 0x80 != 0:
598 # Param names present
599 for _ in range(param_count):
601 mi = MethodInfo(flags & 0x01 != 0, flags & 0x04 != 0)
602 method_infos.append(mi)
605 metadata_count = u30()
606 for _c in range(metadata_count):
609 for _c2 in range(item_count):
613 def parse_traits_info():
614 trait_name_idx = u30()
615 kind_full = read_byte()
616 kind = kind_full & 0x0f
617 attrs = kind_full >> 4
619 if kind in [0x00, 0x06]: # Slot or Const
621 u30() # type_name_idx
625 elif kind in [0x01, 0x02, 0x03]: # Method / Getter / Setter
628 methods[multinames[trait_name_idx]] = method_idx
629 elif kind == 0x04: # Class
632 elif kind == 0x05: # Function
635 methods[function_idx] = multinames[trait_name_idx]
637 raise ExtractorError(u'Unsupported trait kind %d' % kind)
639 if attrs & 0x4 != 0: # Metadata present
640 metadata_count = u30()
641 for _c3 in range(metadata_count):
642 u30() # metadata index
647 TARGET_CLASSNAME = u'SignatureDecipher'
648 searched_idx = multinames.index(TARGET_CLASSNAME)
649 searched_class_id = None
651 for class_id in range(class_count):
653 if name_idx == searched_idx:
654 # We found the class we're looking for!
655 searched_class_id = class_id
656 u30() # super_name idx
658 if flags & 0x08 != 0: # Protected namespace is present
659 u30() # protected_ns_idx
661 for _c2 in range(intrf_count):
665 for _c2 in range(trait_count):
668 if searched_class_id is None:
669 raise ExtractorError(u'Target class %r not found' %
674 for class_id in range(class_count):
677 for _c2 in range(trait_count):
678 trait_methods = parse_traits_info()
679 if class_id == searched_class_id:
680 method_names.update(trait_methods.items())
681 method_idxs.update(dict(
683 for name, idx in trait_methods.items()))
687 for _c in range(script_count):
690 for _c2 in range(trait_count):
694 method_body_count = u30()
695 Method = collections.namedtuple('Method', ['code', 'local_count'])
697 for _c in range(method_body_count):
701 u30() # init_scope_depth
702 u30() # max_scope_depth
704 code = read_bytes(code_length)
705 if method_idx in method_idxs:
706 m = Method(code, local_count)
707 methods[method_idxs[method_idx]] = m
708 exception_count = u30()
709 for _c2 in range(exception_count):
716 for _c2 in range(trait_count):
719 assert p + code_reader.tell() == len(code_tag)
720 assert len(methods) == len(method_idxs)
722 method_pyfunctions = {}
724 def extract_function(func_name):
725 if func_name in method_pyfunctions:
726 return method_pyfunctions[func_name]
727 if func_name not in methods:
728 raise ExtractorError(u'Cannot find function %r' % func_name)
729 m = methods[func_name]
732 registers = ['(this)'] + list(args) + [None] * m.local_count
734 coder = io.BytesIO(m.code)
736 opcode = struct.unpack('!B', coder.read(1))[0]
737 if opcode == 36: # pushbyte
738 v = struct.unpack('!B', coder.read(1))[0]
740 elif opcode == 44: # pushstring
742 stack.append(constant_strings[idx])
743 elif opcode == 48: # pushscope
744 # We don't implement the scope register, so we'll just
745 # ignore the popped value
747 elif opcode == 70: # callproperty
749 mname = multinames[index]
750 arg_count = u30(coder)
751 args = list(reversed(
752 [stack.pop() for _ in range(arg_count)]))
754 if mname == u'split':
755 assert len(args) == 1
756 assert isinstance(args[0], compat_str)
757 assert isinstance(obj, compat_str)
761 res = obj.split(args[0])
763 elif mname == u'slice':
764 assert len(args) == 1
765 assert isinstance(args[0], int)
766 assert isinstance(obj, list)
769 elif mname == u'join':
770 assert len(args) == 1
771 assert isinstance(args[0], compat_str)
772 assert isinstance(obj, list)
773 res = args[0].join(obj)
775 elif mname in method_pyfunctions:
776 stack.append(method_pyfunctions[mname](args))
778 raise NotImplementedError(
779 u'Unsupported property %r on %r'
781 elif opcode == 72: # returnvalue
784 elif opcode == 79: # callpropvoid
786 mname = multinames[index]
787 arg_count = u30(coder)
788 args = list(reversed(
789 [stack.pop() for _ in range(arg_count)]))
791 if mname == u'reverse':
792 assert isinstance(obj, list)
795 raise NotImplementedError(
796 u'Unsupported (void) property %r on %r'
798 elif opcode == 93: # findpropstrict
800 mname = multinames[index]
801 res = extract_function(mname)
803 elif opcode == 97: # setproperty
808 assert isinstance(obj, list)
809 assert isinstance(idx, int)
811 elif opcode == 98: # getlocal
813 stack.append(registers[index])
814 elif opcode == 99: # setlocal
817 registers[index] = value
818 elif opcode == 102: # getproperty
820 pname = multinames[index]
821 if pname == u'length':
823 assert isinstance(obj, list)
824 stack.append(len(obj))
825 else: # Assume attribute access
827 assert isinstance(idx, int)
829 assert isinstance(obj, list)
830 stack.append(obj[idx])
831 elif opcode == 128: # coerce
833 elif opcode == 133: # coerce_s
834 assert isinstance(stack[-1], (type(None), compat_str))
835 elif opcode == 164: # modulo
838 res = value1 % value2
840 elif opcode == 208: # getlocal_0
841 stack.append(registers[0])
842 elif opcode == 209: # getlocal_1
843 stack.append(registers[1])
844 elif opcode == 210: # getlocal_2
845 stack.append(registers[2])
846 elif opcode == 211: # getlocal_3
847 stack.append(registers[3])
848 elif opcode == 214: # setlocal_2
849 registers[2] = stack.pop()
850 elif opcode == 215: # setlocal_3
851 registers[3] = stack.pop()
853 raise NotImplementedError(
854 u'Unsupported opcode %d' % opcode)
856 method_pyfunctions[func_name] = resfunc
859 initial_function = extract_function(u'decipher')
860 return lambda s: initial_function([s])
862 def _decrypt_signature(self, s, video_id, player_url, age_gate=False):
863 """Turn the encrypted s field into a working signature"""
865 if player_url is not None:
866 if player_url.startswith(u'//'):
867 player_url = u'https:' + player_url
869 player_id = (player_url, len(s))
870 if player_id not in self._player_cache:
871 func = self._extract_signature_function(
872 video_id, player_url, len(s)
874 self._player_cache[player_id] = func
875 func = self._player_cache[player_id]
876 if self._downloader.params.get('youtube_print_sig_code'):
877 self._print_sig_code(func, len(s))
880 tb = traceback.format_exc()
881 self._downloader.report_warning(
882 u'Automatic signature extraction failed: ' + tb)
884 self._downloader.report_warning(
885 u'Warning: Falling back to static signature algorithm')
887 return self._static_decrypt_signature(
888 s, video_id, player_url, age_gate)
890 def _static_decrypt_signature(self, s, video_id, player_url, age_gate):
892 # The videos with age protection use another player, so the
893 # algorithms can be different.
895 return s[2:63] + s[82] + s[64:82] + s[63]
898 return s[86:29:-1] + s[88] + s[28:5:-1]
900 return s[25] + s[3:25] + s[0] + s[26:42] + s[79] + s[43:79] + s[91] + s[80:83]
902 return s[84:27:-1] + s[86] + s[26:5:-1]
904 return s[25] + s[3:25] + s[2] + s[26:40] + s[77] + s[41:77] + s[89] + s[78:81]
906 return s[84:78:-1] + s[87] + s[77:60:-1] + s[0] + s[59:3:-1]
908 return s[7:28] + s[87] + s[29:45] + s[55] + s[46:55] + s[2] + s[56:87] + s[28]
910 return s[6:27] + s[4] + s[28:39] + s[27] + s[40:59] + s[2] + s[60:]
912 return s[80:72:-1] + s[16] + s[71:39:-1] + s[72] + s[38:16:-1] + s[82] + s[15::-1]
914 return s[3:11] + s[0] + s[12:55] + s[84] + s[56:84]
916 return s[78:70:-1] + s[14] + s[69:37:-1] + s[70] + s[36:14:-1] + s[80] + s[:14][::-1]
918 return s[80:63:-1] + s[0] + s[62:0:-1] + s[63]
920 return s[80:37:-1] + s[7] + s[36:7:-1] + s[0] + s[6:0:-1] + s[37]
922 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]
924 return s[1:19] + s[0] + s[20:68] + s[19] + s[69:80]
926 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]
929 raise ExtractorError(u'Unable to decrypt signature, key length %d not supported; retrying might work' % (len(s)))
931 def _get_available_subtitles(self, video_id, webpage):
933 sub_list = self._download_webpage(
934 'https://video.google.com/timedtext?hl=en&type=list&v=%s' % video_id,
935 video_id, note=False)
936 except ExtractorError as err:
937 self._downloader.report_warning(u'unable to download video subtitles: %s' % compat_str(err))
939 lang_list = re.findall(r'name="([^"]*)"[^>]+lang_code="([\w\-]+)"', sub_list)
944 params = compat_urllib_parse.urlencode({
947 'fmt': self._downloader.params.get('subtitlesformat', 'srt'),
948 'name': unescapeHTML(l[0]).encode('utf-8'),
950 url = u'https://www.youtube.com/api/timedtext?' + params
951 sub_lang_list[lang] = url
952 if not sub_lang_list:
953 self._downloader.report_warning(u'video doesn\'t have subtitles')
957 def _get_available_automatic_caption(self, video_id, webpage):
958 """We need the webpage for getting the captions url, pass it as an
959 argument to speed up the process."""
960 sub_format = self._downloader.params.get('subtitlesformat', 'srt')
961 self.to_screen(u'%s: Looking for automatic captions' % video_id)
962 mobj = re.search(r';ytplayer.config = ({.*?});', webpage)
963 err_msg = u'Couldn\'t find automatic captions for %s' % video_id
965 self._downloader.report_warning(err_msg)
967 player_config = json.loads(mobj.group(1))
969 args = player_config[u'args']
970 caption_url = args[u'ttsurl']
971 timestamp = args[u'timestamp']
972 # We get the available subtitles
973 list_params = compat_urllib_parse.urlencode({
978 list_url = caption_url + '&' + list_params
979 caption_list = self._download_xml(list_url, video_id)
980 original_lang_node = caption_list.find('track')
981 if original_lang_node is None or original_lang_node.attrib.get('kind') != 'asr' :
982 self._downloader.report_warning(u'Video doesn\'t have automatic captions')
984 original_lang = original_lang_node.attrib['lang_code']
987 for lang_node in caption_list.findall('target'):
988 sub_lang = lang_node.attrib['lang_code']
989 params = compat_urllib_parse.urlencode({
990 'lang': original_lang,
996 sub_lang_list[sub_lang] = caption_url + '&' + params
998 # An extractor error can be raise by the download process if there are
999 # no automatic captions but there are subtitles
1000 except (KeyError, ExtractorError):
1001 self._downloader.report_warning(err_msg)
1005 def extract_id(cls, url):
1006 mobj = re.match(cls._VALID_URL, url, re.VERBOSE)
1008 raise ExtractorError(u'Invalid URL: %s' % url)
1009 video_id = mobj.group(2)
1012 def _extract_from_m3u8(self, manifest_url, video_id):
1014 def _get_urls(_manifest):
1015 lines = _manifest.split('\n')
1016 urls = filter(lambda l: l and not l.startswith('#'),
1019 manifest = self._download_webpage(manifest_url, video_id, u'Downloading formats manifest')
1020 formats_urls = _get_urls(manifest)
1021 for format_url in formats_urls:
1022 itag = self._search_regex(r'itag/(\d+?)/', format_url, 'itag')
1023 url_map[itag] = format_url
1026 def _extract_annotations(self, video_id):
1027 url = 'https://www.youtube.com/annotations_invideo?features=1&legacy=1&video_id=%s' % video_id
1028 return self._download_webpage(url, video_id, note=u'Searching for annotations.', errnote=u'Unable to download video annotations.')
1030 def _real_extract(self, url):
1032 u'http' if self._downloader.params.get('prefer_insecure', False)
1035 # Extract original video URL from URL with redirection, like age verification, using next_url parameter
1036 mobj = re.search(self._NEXT_URL_RE, url)
1038 url = proto + '://www.youtube.com/' + compat_urllib_parse.unquote(mobj.group(1)).lstrip('/')
1039 video_id = self.extract_id(url)
1042 url = proto + '://www.youtube.com/watch?v=%s&gl=US&hl=en&has_verified=1' % video_id
1043 video_webpage = self._download_webpage(url, video_id)
1045 # Attempt to extract SWF player URL
1046 mobj = re.search(r'swfConfig.*?"(https?:\\/\\/.*?watch.*?-.*?\.swf)"', video_webpage)
1047 if mobj is not None:
1048 player_url = re.sub(r'\\(.)', r'\1', mobj.group(1))
1053 self.report_video_info_webpage_download(video_id)
1054 if re.search(r'player-age-gate-content">', video_webpage) is not None:
1055 self.report_age_confirmation()
1057 # We simulate the access to the video from www.youtube.com/v/{video_id}
1058 # this can be viewed without login into Youtube
1059 data = compat_urllib_parse.urlencode({'video_id': video_id,
1060 'el': 'player_embedded',
1063 'eurl': 'https://youtube.googleapis.com/v/' + video_id,
1067 video_info_url = proto + '://www.youtube.com/get_video_info?' + data
1068 video_info_webpage = self._download_webpage(video_info_url, video_id,
1070 errnote='unable to download video info webpage')
1071 video_info = compat_parse_qs(video_info_webpage)
1074 for el_type in ['&el=embedded', '&el=detailpage', '&el=vevo', '']:
1075 video_info_url = (proto + '://www.youtube.com/get_video_info?&video_id=%s%s&ps=default&eurl=&gl=US&hl=en'
1076 % (video_id, el_type))
1077 video_info_webpage = self._download_webpage(video_info_url, video_id,
1079 errnote='unable to download video info webpage')
1080 video_info = compat_parse_qs(video_info_webpage)
1081 if 'token' in video_info:
1083 if 'token' not in video_info:
1084 if 'reason' in video_info:
1085 raise ExtractorError(
1086 u'YouTube said: %s' % video_info['reason'][0],
1087 expected=True, video_id=video_id)
1089 raise ExtractorError(
1090 u'"token" parameter not in video info for unknown reason',
1093 if 'view_count' in video_info:
1094 view_count = int(video_info['view_count'][0])
1098 # Check for "rental" videos
1099 if 'ypc_video_rental_bar_text' in video_info and 'author' not in video_info:
1100 raise ExtractorError(u'"rental" videos not supported')
1102 # Start extracting information
1103 self.report_information_extraction(video_id)
1106 if 'author' not in video_info:
1107 raise ExtractorError(u'Unable to extract uploader name')
1108 video_uploader = compat_urllib_parse.unquote_plus(video_info['author'][0])
1111 video_uploader_id = None
1112 mobj = re.search(r'<link itemprop="url" href="http://www.youtube.com/(?:user|channel)/([^"]+)">', video_webpage)
1113 if mobj is not None:
1114 video_uploader_id = mobj.group(1)
1116 self._downloader.report_warning(u'unable to extract uploader nickname')
1119 if 'title' in video_info:
1120 video_title = video_info['title'][0]
1122 self._downloader.report_warning(u'Unable to extract video title')
1126 # We try first to get a high quality image:
1127 m_thumb = re.search(r'<span itemprop="thumbnail".*?href="(.*?)">',
1128 video_webpage, re.DOTALL)
1129 if m_thumb is not None:
1130 video_thumbnail = m_thumb.group(1)
1131 elif 'thumbnail_url' not in video_info:
1132 self._downloader.report_warning(u'unable to extract video thumbnail')
1133 video_thumbnail = None
1134 else: # don't panic if we can't find it
1135 video_thumbnail = compat_urllib_parse.unquote_plus(video_info['thumbnail_url'][0])
1139 mobj = re.search(r'id="eow-date.*?>(.*?)</span>', video_webpage, re.DOTALL)
1140 if mobj is not None:
1141 upload_date = ' '.join(re.sub(r'[/,-]', r' ', mobj.group(1)).split())
1142 upload_date = unified_strdate(upload_date)
1145 video_description = get_element_by_id("eow-description", video_webpage)
1146 if video_description:
1147 video_description = re.sub(r'''(?x)
1149 (?:[a-zA-Z-]+="[^"]+"\s+)*?
1151 (?:[a-zA-Z-]+="[^"]+"\s+)*?
1152 class="yt-uix-redirect-link"\s*>
1155 ''', r'\1', video_description)
1156 video_description = clean_html(video_description)
1158 fd_mobj = re.search(r'<meta name="description" content="([^"]+)"', video_webpage)
1160 video_description = unescapeHTML(fd_mobj.group(1))
1162 video_description = u''
1164 def _extract_count(klass):
1165 count = self._search_regex(
1166 r'class="%s">([\d,]+)</span>' % re.escape(klass),
1167 video_webpage, klass, default=None)
1168 if count is not None:
1169 return int(count.replace(',', ''))
1171 like_count = _extract_count(u'likes-count')
1172 dislike_count = _extract_count(u'dislikes-count')
1175 video_subtitles = self.extract_subtitles(video_id, video_webpage)
1177 if self._downloader.params.get('listsubtitles', False):
1178 self._list_available_subtitles(video_id, video_webpage)
1181 if 'length_seconds' not in video_info:
1182 self._downloader.report_warning(u'unable to extract video duration')
1183 video_duration = None
1185 video_duration = int(compat_urllib_parse.unquote_plus(video_info['length_seconds'][0]))
1188 video_annotations = None
1189 if self._downloader.params.get('writeannotations', False):
1190 video_annotations = self._extract_annotations(video_id)
1192 # Decide which formats to download
1194 mobj = re.search(r';ytplayer\.config\s*=\s*({.*?});', video_webpage)
1196 raise ValueError('Could not find vevo ID')
1197 json_code = uppercase_escape(mobj.group(1))
1198 ytplayer_config = json.loads(json_code)
1199 args = ytplayer_config['args']
1200 # Easy way to know if the 's' value is in url_encoded_fmt_stream_map
1201 # this signatures are encrypted
1202 if 'url_encoded_fmt_stream_map' not in args:
1203 raise ValueError(u'No stream_map present') # caught below
1204 re_signature = re.compile(r'[&,]s=')
1205 m_s = re_signature.search(args['url_encoded_fmt_stream_map'])
1207 self.to_screen(u'%s: Encrypted signatures detected.' % video_id)
1208 video_info['url_encoded_fmt_stream_map'] = [args['url_encoded_fmt_stream_map']]
1209 m_s = re_signature.search(args.get('adaptive_fmts', u''))
1211 if 'adaptive_fmts' in video_info:
1212 video_info['adaptive_fmts'][0] += ',' + args['adaptive_fmts']
1214 video_info['adaptive_fmts'] = [args['adaptive_fmts']]
1218 def _map_to_format_list(urlmap):
1220 for itag, video_real_url in urlmap.items():
1223 'url': video_real_url,
1224 'player_url': player_url,
1226 if itag in self._formats:
1227 dct.update(self._formats[itag])
1231 if 'conn' in video_info and video_info['conn'][0].startswith('rtmp'):
1232 self.report_rtmp_download()
1234 'format_id': '_rtmp',
1236 'url': video_info['conn'][0],
1237 'player_url': player_url,
1239 elif len(video_info.get('url_encoded_fmt_stream_map', [])) >= 1 or len(video_info.get('adaptive_fmts', [])) >= 1:
1240 encoded_url_map = video_info.get('url_encoded_fmt_stream_map', [''])[0] + ',' + video_info.get('adaptive_fmts',[''])[0]
1241 if 'rtmpe%3Dyes' in encoded_url_map:
1242 raise ExtractorError('rtmpe downloads are not supported, see https://github.com/rg3/youtube-dl/issues/343 for more information.', expected=True)
1244 for url_data_str in encoded_url_map.split(','):
1245 url_data = compat_parse_qs(url_data_str)
1246 if 'itag' in url_data and 'url' in url_data:
1247 url = url_data['url'][0]
1248 if 'sig' in url_data:
1249 url += '&signature=' + url_data['sig'][0]
1250 elif 's' in url_data:
1251 encrypted_sig = url_data['s'][0]
1252 if self._downloader.params.get('verbose'):
1254 if player_url is None:
1255 player_version = 'unknown'
1257 player_version = self._search_regex(
1258 r'-(.+)\.swf$', player_url,
1259 u'flash player', fatal=False)
1260 player_desc = 'flash player %s' % player_version
1262 player_version = self._search_regex(
1263 r'html5player-(.+?)\.js', video_webpage,
1264 'html5 player', fatal=False)
1265 player_desc = u'html5 player %s' % player_version
1267 parts_sizes = u'.'.join(compat_str(len(part)) for part in encrypted_sig.split('.'))
1268 self.to_screen(u'encrypted signature length %d (%s), itag %s, %s' %
1269 (len(encrypted_sig), parts_sizes, url_data['itag'][0], player_desc))
1272 jsplayer_url_json = self._search_regex(
1273 r'"assets":.+?"js":\s*("[^"]+")',
1274 video_webpage, u'JS player URL')
1275 player_url = json.loads(jsplayer_url_json)
1277 signature = self._decrypt_signature(
1278 encrypted_sig, video_id, player_url, age_gate)
1279 url += '&signature=' + signature
1280 if 'ratebypass' not in url:
1281 url += '&ratebypass=yes'
1282 url_map[url_data['itag'][0]] = url
1283 formats = _map_to_format_list(url_map)
1284 elif video_info.get('hlsvp'):
1285 manifest_url = video_info['hlsvp'][0]
1286 url_map = self._extract_from_m3u8(manifest_url, video_id)
1287 formats = _map_to_format_list(url_map)
1289 raise ExtractorError(u'no conn, hlsvp or url_encoded_fmt_stream_map information found in video info')
1291 # Look for the DASH manifest
1292 if (self._downloader.params.get('youtube_include_dash_manifest', False)):
1294 # The DASH manifest used needs to be the one from the original video_webpage.
1295 # The one found in get_video_info seems to be using different signatures.
1296 # However, in the case of an age restriction there won't be any embedded dashmpd in the video_webpage.
1297 # Luckily, it seems, this case uses some kind of default signature (len == 86), so the
1298 # combination of get_video_info and the _static_decrypt_signature() decryption fallback will work here.
1300 dash_manifest_url = video_info.get('dashmpd')[0]
1302 dash_manifest_url = ytplayer_config['args']['dashmpd']
1303 def decrypt_sig(mobj):
1305 dec_s = self._decrypt_signature(s, video_id, player_url, age_gate)
1306 return '/signature/%s' % dec_s
1307 dash_manifest_url = re.sub(r'/s/([\w\.]+)', decrypt_sig, dash_manifest_url)
1308 dash_doc = self._download_xml(
1309 dash_manifest_url, video_id,
1310 note=u'Downloading DASH manifest',
1311 errnote=u'Could not download DASH manifest')
1312 for r in dash_doc.findall(u'.//{urn:mpeg:DASH:schema:MPD:2011}Representation'):
1313 url_el = r.find('{urn:mpeg:DASH:schema:MPD:2011}BaseURL')
1316 format_id = r.attrib['id']
1317 video_url = url_el.text
1318 filesize = int_or_none(url_el.attrib.get('{http://youtube.com/yt/2012/10/10}contentLength'))
1320 'format_id': format_id,
1322 'width': int_or_none(r.attrib.get('width')),
1323 'tbr': int_or_none(r.attrib.get('bandwidth'), 1000),
1324 'asr': int_or_none(r.attrib.get('audioSamplingRate')),
1325 'filesize': filesize,
1328 existing_format = next(
1329 fo for fo in formats
1330 if fo['format_id'] == format_id)
1331 except StopIteration:
1332 f.update(self._formats.get(format_id, {}))
1335 existing_format.update(f)
1337 except (ExtractorError, KeyError) as e:
1338 self.report_warning(u'Skipping DASH manifest: %s' % e, video_id)
1340 self._sort_formats(formats)
1344 'uploader': video_uploader,
1345 'uploader_id': video_uploader_id,
1346 'upload_date': upload_date,
1347 'title': video_title,
1348 'thumbnail': video_thumbnail,
1349 'description': video_description,
1350 'subtitles': video_subtitles,
1351 'duration': video_duration,
1352 'age_limit': 18 if age_gate else 0,
1353 'annotations': video_annotations,
1354 'webpage_url': proto + '://www.youtube.com/watch?v=%s' % video_id,
1355 'view_count': view_count,
1356 'like_count': like_count,
1357 'dislike_count': dislike_count,
1361 class YoutubePlaylistIE(YoutubeBaseInfoExtractor):
1362 IE_DESC = u'YouTube.com playlists'
1363 _VALID_URL = r"""(?x)(?:
1368 (?:course|view_play_list|my_playlists|artist|playlist|watch)
1369 \? (?:.*?&)*? (?:p|a|list)=
1373 (?:PL|EC|UU|FL|RD)?[0-9A-Za-z-_]{10,}
1374 # Top tracks, they can also include dots
1379 ((?:PL|EC|UU|FL|RD)[0-9A-Za-z-_]{10,})
1381 _TEMPLATE_URL = 'https://www.youtube.com/playlist?list=%s'
1382 _MORE_PAGES_INDICATOR = r'data-link-type="next"'
1383 _VIDEO_RE = r'href="\s*/watch\?v=(?P<id>[0-9A-Za-z_-]{11})&[^"]*?index=(?P<index>\d+)'
1384 IE_NAME = u'youtube:playlist'
1386 def _real_initialize(self):
1389 def _ids_to_results(self, ids):
1390 return [self.url_result(vid_id, 'Youtube', video_id=vid_id)
1393 def _extract_mix(self, playlist_id):
1394 # The mixes are generated from a a single video
1395 # the id of the playlist is just 'RD' + video_id
1396 url = 'https://youtube.com/watch?v=%s&list=%s' % (playlist_id[-11:], playlist_id)
1397 webpage = self._download_webpage(url, playlist_id, u'Downloading Youtube mix')
1398 search_title = lambda class_name: get_element_by_attribute('class', class_name, webpage)
1399 title_span = (search_title('playlist-title') or
1400 search_title('title long-title') or search_title('title'))
1401 title = clean_html(title_span)
1402 video_re = r'''(?x)data-video-username="(.*?)".*?
1403 href="/watch\?v=([0-9A-Za-z_-]{11})&[^"]*?list=%s''' % re.escape(playlist_id)
1404 matches = orderedSet(re.findall(video_re, webpage, flags=re.DOTALL))
1405 # Some of the videos may have been deleted, their username field is empty
1406 ids = [video_id for (username, video_id) in matches if username]
1407 url_results = self._ids_to_results(ids)
1409 return self.playlist_result(url_results, playlist_id, title)
1411 def _real_extract(self, url):
1412 # Extract playlist id
1413 mobj = re.match(self._VALID_URL, url)
1415 raise ExtractorError(u'Invalid URL: %s' % url)
1416 playlist_id = mobj.group(1) or mobj.group(2)
1418 # Check if it's a video-specific URL
1419 query_dict = compat_urlparse.parse_qs(compat_urlparse.urlparse(url).query)
1420 if 'v' in query_dict:
1421 video_id = query_dict['v'][0]
1422 if self._downloader.params.get('noplaylist'):
1423 self.to_screen(u'Downloading just video %s because of --no-playlist' % video_id)
1424 return self.url_result(video_id, 'Youtube', video_id=video_id)
1426 self.to_screen(u'Downloading playlist %s - add --no-playlist to just download video %s' % (playlist_id, video_id))
1428 if playlist_id.startswith('RD'):
1429 # Mixes require a custom extraction process
1430 return self._extract_mix(playlist_id)
1431 if playlist_id.startswith('TL'):
1432 raise ExtractorError(u'For downloading YouTube.com top lists, use '
1433 u'the "yttoplist" keyword, for example "youtube-dl \'yttoplist:music:Top Tracks\'"', expected=True)
1435 url = self._TEMPLATE_URL % playlist_id
1436 page = self._download_webpage(url, playlist_id)
1437 more_widget_html = content_html = page
1439 # Check if the playlist exists or is private
1440 if re.search(r'<div class="yt-alert-message">[^<]*?(The|This) playlist (does not exist|is private)[^<]*?</div>', page) is not None:
1441 raise ExtractorError(
1442 u'The playlist doesn\'t exist or is private, use --username or '
1443 '--netrc to access it.',
1446 # Extract the video ids from the playlist pages
1449 for page_num in itertools.count(1):
1450 matches = re.finditer(self._VIDEO_RE, content_html)
1451 # We remove the duplicates and the link with index 0
1452 # (it's not the first video of the playlist)
1453 new_ids = orderedSet(m.group('id') for m in matches if m.group('index') != '0')
1456 mobj = re.search(r'data-uix-load-more-href="/?(?P<more>[^"]+)"', more_widget_html)
1460 more = self._download_json(
1461 'https://youtube.com/%s' % mobj.group('more'), playlist_id,
1462 'Downloading page #%s' % page_num,
1463 transform_source=uppercase_escape)
1464 content_html = more['content_html']
1465 more_widget_html = more['load_more_widget_html']
1467 playlist_title = self._html_search_regex(
1468 r'(?s)<h1 class="pl-header-title[^"]*">\s*(.*?)\s*</h1>',
1471 url_results = self._ids_to_results(ids)
1472 return self.playlist_result(url_results, playlist_id, playlist_title)
1475 class YoutubeTopListIE(YoutubePlaylistIE):
1476 IE_NAME = u'youtube:toplist'
1477 IE_DESC = (u'YouTube.com top lists, "yttoplist:{channel}:{list title}"'
1478 u' (Example: "yttoplist:music:Top Tracks")')
1479 _VALID_URL = r'yttoplist:(?P<chann>.*?):(?P<title>.*?)$'
1481 def _real_extract(self, url):
1482 mobj = re.match(self._VALID_URL, url)
1483 channel = mobj.group('chann')
1484 title = mobj.group('title')
1485 query = compat_urllib_parse.urlencode({'title': title})
1486 playlist_re = 'href="([^"]+?%s.*?)"' % re.escape(query)
1487 channel_page = self._download_webpage('https://www.youtube.com/%s' % channel, title)
1488 link = self._html_search_regex(playlist_re, channel_page, u'list')
1489 url = compat_urlparse.urljoin('https://www.youtube.com/', link)
1491 video_re = r'data-index="\d+".*?data-video-id="([0-9A-Za-z_-]{11})"'
1493 # sometimes the webpage doesn't contain the videos
1494 # retry until we get them
1495 for i in itertools.count(0):
1496 msg = u'Downloading Youtube mix'
1498 msg += ', retry #%d' % i
1499 webpage = self._download_webpage(url, title, msg)
1500 ids = orderedSet(re.findall(video_re, webpage))
1503 url_results = self._ids_to_results(ids)
1504 return self.playlist_result(url_results, playlist_title=title)
1507 class YoutubeChannelIE(InfoExtractor):
1508 IE_DESC = u'YouTube.com channels'
1509 _VALID_URL = r"^(?:https?://)?(?:youtu\.be|(?:\w+\.)?youtube(?:-nocookie)?\.com)/channel/([0-9A-Za-z_-]+)"
1510 _MORE_PAGES_INDICATOR = 'yt-uix-load-more'
1511 _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'
1512 IE_NAME = u'youtube:channel'
1514 def extract_videos_from_page(self, page):
1516 for mobj in re.finditer(r'href="/watch\?v=([0-9A-Za-z_-]+)&?', page):
1517 if mobj.group(1) not in ids_in_page:
1518 ids_in_page.append(mobj.group(1))
1521 def _real_extract(self, url):
1522 # Extract channel id
1523 mobj = re.match(self._VALID_URL, url)
1525 raise ExtractorError(u'Invalid URL: %s' % url)
1527 # Download channel page
1528 channel_id = mobj.group(1)
1530 url = 'https://www.youtube.com/channel/%s/videos' % channel_id
1531 channel_page = self._download_webpage(url, channel_id)
1532 autogenerated = re.search(r'''(?x)
1534 channel-header-autogenerated-label|
1535 yt-channel-title-autogenerated
1536 )[^"]*"''', channel_page) is not None
1539 # The videos are contained in a single page
1540 # the ajax pages can't be used, they are empty
1541 video_ids = self.extract_videos_from_page(channel_page)
1543 # Download all channel pages using the json-based channel_ajax query
1544 for pagenum in itertools.count(1):
1545 url = self._MORE_PAGES_URL % (pagenum, channel_id)
1546 page = self._download_json(
1547 url, channel_id, note=u'Downloading page #%s' % pagenum,
1548 transform_source=uppercase_escape)
1550 ids_in_page = self.extract_videos_from_page(page['content_html'])
1551 video_ids.extend(ids_in_page)
1553 if self._MORE_PAGES_INDICATOR not in page['load_more_widget_html']:
1556 self._downloader.to_screen(u'[youtube] Channel %s: Found %i videos' % (channel_id, len(video_ids)))
1558 url_entries = [self.url_result(video_id, 'Youtube', video_id=video_id)
1559 for video_id in video_ids]
1560 return self.playlist_result(url_entries, channel_id)
1563 class YoutubeUserIE(InfoExtractor):
1564 IE_DESC = u'YouTube.com user videos (URL or "ytuser" keyword)'
1565 _VALID_URL = r'(?:(?:(?:https?://)?(?:\w+\.)?youtube\.com/(?:user/)?(?!(?:attribution_link|watch|results)(?:$|[^a-z_A-Z0-9-])))|ytuser:)(?!feed/)([A-Za-z0-9_-]+)'
1566 _TEMPLATE_URL = 'https://gdata.youtube.com/feeds/api/users/%s'
1567 _GDATA_PAGE_SIZE = 50
1568 _GDATA_URL = 'https://gdata.youtube.com/feeds/api/users/%s/uploads?max-results=%d&start-index=%d&alt=json'
1569 IE_NAME = u'youtube:user'
1572 def suitable(cls, url):
1573 # Don't return True if the url can be extracted with other youtube
1574 # extractor, the regex would is too permissive and it would match.
1575 other_ies = iter(klass for (name, klass) in globals().items() if name.endswith('IE') and klass is not cls)
1576 if any(ie.suitable(url) for ie in other_ies): return False
1577 else: return super(YoutubeUserIE, cls).suitable(url)
1579 def _real_extract(self, url):
1581 mobj = re.match(self._VALID_URL, url)
1583 raise ExtractorError(u'Invalid URL: %s' % url)
1585 username = mobj.group(1)
1587 # Download video ids using YouTube Data API. Result size per
1588 # query is limited (currently to 50 videos) so we need to query
1589 # page by page until there are no video ids - it means we got
1592 def download_page(pagenum):
1593 start_index = pagenum * self._GDATA_PAGE_SIZE + 1
1595 gdata_url = self._GDATA_URL % (username, self._GDATA_PAGE_SIZE, start_index)
1596 page = self._download_webpage(
1597 gdata_url, username,
1598 u'Downloading video ids from %d to %d' % (
1599 start_index, start_index + self._GDATA_PAGE_SIZE))
1602 response = json.loads(page)
1603 except ValueError as err:
1604 raise ExtractorError(u'Invalid JSON in API response: ' + compat_str(err))
1605 if 'entry' not in response['feed']:
1608 # Extract video identifiers
1609 entries = response['feed']['entry']
1610 for entry in entries:
1611 title = entry['title']['$t']
1612 video_id = entry['id']['$t'].split('/')[-1]
1616 'ie_key': 'Youtube',
1620 url_results = PagedList(download_page, self._GDATA_PAGE_SIZE)
1622 return self.playlist_result(url_results, playlist_title=username)
1625 class YoutubeSearchIE(SearchInfoExtractor):
1626 IE_DESC = u'YouTube.com searches'
1627 _API_URL = u'https://gdata.youtube.com/feeds/api/videos?q=%s&start-index=%i&max-results=50&v=2&alt=jsonc'
1629 IE_NAME = u'youtube:search'
1630 _SEARCH_KEY = 'ytsearch'
1632 def _get_n_results(self, query, n):
1633 """Get a specified number of results for a query"""
1640 while (PAGE_SIZE * pagenum) < limit:
1641 result_url = self._API_URL % (
1642 compat_urllib_parse.quote_plus(query.encode('utf-8')),
1643 (PAGE_SIZE * pagenum) + 1)
1644 data_json = self._download_webpage(
1645 result_url, video_id=u'query "%s"' % query,
1646 note=u'Downloading page %s' % (pagenum + 1),
1647 errnote=u'Unable to download API page')
1648 data = json.loads(data_json)
1649 api_response = data['data']
1651 if 'items' not in api_response:
1652 raise ExtractorError(
1653 u'[youtube] No video results', expected=True)
1655 new_ids = list(video['id'] for video in api_response['items'])
1656 video_ids += new_ids
1658 limit = min(n, api_response['totalItems'])
1661 if len(video_ids) > n:
1662 video_ids = video_ids[:n]
1663 videos = [self.url_result(video_id, 'Youtube', video_id=video_id)
1664 for video_id in video_ids]
1665 return self.playlist_result(videos, query)
1668 class YoutubeSearchDateIE(YoutubeSearchIE):
1669 IE_NAME = YoutubeSearchIE.IE_NAME + ':date'
1670 _API_URL = 'https://gdata.youtube.com/feeds/api/videos?q=%s&start-index=%i&max-results=50&v=2&alt=jsonc&orderby=published'
1671 _SEARCH_KEY = 'ytsearchdate'
1672 IE_DESC = u'YouTube.com searches, newest videos first'
1675 class YoutubeSearchURLIE(InfoExtractor):
1676 IE_DESC = u'YouTube.com search URLs'
1677 IE_NAME = u'youtube:search_url'
1678 _VALID_URL = r'https?://(?:www\.)?youtube\.com/results\?(.*?&)?search_query=(?P<query>[^&]+)(?:[&]|$)'
1680 def _real_extract(self, url):
1681 mobj = re.match(self._VALID_URL, url)
1682 query = compat_urllib_parse.unquote_plus(mobj.group('query'))
1684 webpage = self._download_webpage(url, query)
1685 result_code = self._search_regex(
1686 r'(?s)<ol id="search-results"(.*?)</ol>', webpage, u'result HTML')
1688 part_codes = re.findall(
1689 r'(?s)<h3 class="yt-lockup-title">(.*?)</h3>', result_code)
1691 for part_code in part_codes:
1692 part_title = self._html_search_regex(
1693 r'(?s)title="([^"]+)"', part_code, 'item title', fatal=False)
1694 part_url_snippet = self._html_search_regex(
1695 r'(?s)href="([^"]+)"', part_code, 'item URL')
1696 part_url = compat_urlparse.urljoin(
1697 'https://www.youtube.com/', part_url_snippet)
1701 'title': part_title,
1705 '_type': 'playlist',
1711 class YoutubeShowIE(InfoExtractor):
1712 IE_DESC = u'YouTube.com (multi-season) shows'
1713 _VALID_URL = r'https?://www\.youtube\.com/show/(.*)'
1714 IE_NAME = u'youtube:show'
1716 def _real_extract(self, url):
1717 mobj = re.match(self._VALID_URL, url)
1718 show_name = mobj.group(1)
1719 webpage = self._download_webpage(url, show_name, u'Downloading show webpage')
1720 # There's one playlist for each season of the show
1721 m_seasons = list(re.finditer(r'href="(/playlist\?list=.*?)"', webpage))
1722 self.to_screen(u'%s: Found %s seasons' % (show_name, len(m_seasons)))
1723 return [self.url_result('https://www.youtube.com' + season.group(1), 'YoutubePlaylist') for season in m_seasons]
1726 class YoutubeFeedsInfoExtractor(YoutubeBaseInfoExtractor):
1728 Base class for extractors that fetch info from
1729 http://www.youtube.com/feed_ajax
1730 Subclasses must define the _FEED_NAME and _PLAYLIST_TITLE properties.
1732 _LOGIN_REQUIRED = True
1733 # use action_load_personal_feed instead of action_load_system_feed
1734 _PERSONAL_FEED = False
1737 def _FEED_TEMPLATE(self):
1738 action = 'action_load_system_feed'
1739 if self._PERSONAL_FEED:
1740 action = 'action_load_personal_feed'
1741 return 'https://www.youtube.com/feed_ajax?%s=1&feed_name=%s&paging=%%s' % (action, self._FEED_NAME)
1745 return u'youtube:%s' % self._FEED_NAME
1747 def _real_initialize(self):
1750 def _real_extract(self, url):
1753 for i in itertools.count(1):
1754 info = self._download_json(self._FEED_TEMPLATE % paging,
1755 u'%s feed' % self._FEED_NAME,
1756 u'Downloading page %s' % i)
1757 feed_html = info.get('feed_html') or info.get('content_html')
1758 m_ids = re.finditer(r'"/watch\?v=(.*?)["&]', feed_html)
1759 ids = orderedSet(m.group(1) for m in m_ids)
1760 feed_entries.extend(
1761 self.url_result(video_id, 'Youtube', video_id=video_id)
1762 for video_id in ids)
1763 if info['paging'] is None:
1765 paging = info['paging']
1766 return self.playlist_result(feed_entries, playlist_title=self._PLAYLIST_TITLE)
1768 class YoutubeSubscriptionsIE(YoutubeFeedsInfoExtractor):
1769 IE_DESC = u'YouTube.com subscriptions feed, "ytsubs" keyword (requires authentication)'
1770 _VALID_URL = r'https?://www\.youtube\.com/feed/subscriptions|:ytsubs(?:criptions)?'
1771 _FEED_NAME = 'subscriptions'
1772 _PLAYLIST_TITLE = u'Youtube Subscriptions'
1774 class YoutubeRecommendedIE(YoutubeFeedsInfoExtractor):
1775 IE_DESC = u'YouTube.com recommended videos, "ytrec" keyword (requires authentication)'
1776 _VALID_URL = r'https?://www\.youtube\.com/feed/recommended|:ytrec(?:ommended)?'
1777 _FEED_NAME = 'recommended'
1778 _PLAYLIST_TITLE = u'Youtube Recommended videos'
1780 class YoutubeWatchLaterIE(YoutubeFeedsInfoExtractor):
1781 IE_DESC = u'Youtube watch later list, "ytwatchlater" keyword (requires authentication)'
1782 _VALID_URL = r'https?://www\.youtube\.com/feed/watch_later|:ytwatchlater'
1783 _FEED_NAME = 'watch_later'
1784 _PLAYLIST_TITLE = u'Youtube Watch Later'
1785 _PERSONAL_FEED = True
1787 class YoutubeHistoryIE(YoutubeFeedsInfoExtractor):
1788 IE_DESC = u'Youtube watch history, "ythistory" keyword (requires authentication)'
1789 _VALID_URL = u'https?://www\.youtube\.com/feed/history|:ythistory'
1790 _FEED_NAME = 'history'
1791 _PERSONAL_FEED = True
1792 _PLAYLIST_TITLE = u'Youtube Watch History'
1794 class YoutubeFavouritesIE(YoutubeBaseInfoExtractor):
1795 IE_NAME = u'youtube:favorites'
1796 IE_DESC = u'YouTube.com favourite videos, "ytfav" keyword (requires authentication)'
1797 _VALID_URL = r'https?://www\.youtube\.com/my_favorites|:ytfav(?:ou?rites)?'
1798 _LOGIN_REQUIRED = True
1800 def _real_extract(self, url):
1801 webpage = self._download_webpage('https://www.youtube.com/my_favorites', 'Youtube Favourites videos')
1802 playlist_id = self._search_regex(r'list=(.+?)["&]', webpage, u'favourites playlist id')
1803 return self.url_result(playlist_id, 'YoutubePlaylist')
1806 class YoutubeTruncatedURLIE(InfoExtractor):
1807 IE_NAME = 'youtube:truncated_url'
1808 IE_DESC = False # Do not list
1809 _VALID_URL = r'''(?x)
1810 (?:https?://)?[^/]+/watch\?(?:feature=[a-z_]+)?$|
1811 (?:https?://)?(?:www\.)?youtube\.com/attribution_link\?a=[^&]+$
1814 def _real_extract(self, url):
1815 raise ExtractorError(
1816 u'Did you forget to quote the URL? Remember that & is a meta '
1817 u'character in most shells, so you want to put the URL in quotes, '
1819 u'"http://www.youtube.com/watch?feature=foo&v=BaW_jenozKc" '
1820 u' or simply youtube-dl BaW_jenozKc .',