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},
226 '271': {'ext': 'webm', 'height': 1440, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
227 '272': {'ext': 'webm', 'height': 2160, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
230 '171': {'ext': 'webm', 'vcodec': 'none', 'format_note': 'DASH audio', 'abr': 48, 'preference': -50},
231 '172': {'ext': 'webm', 'vcodec': 'none', 'format_note': 'DASH audio', 'abr': 256, 'preference': -50},
234 '_rtmp': {'protocol': 'rtmp'},
240 u"url": u"http://www.youtube.com/watch?v=BaW_jenozKc",
241 u"file": u"BaW_jenozKc.mp4",
243 u"title": u"youtube-dl test video \"'/\\ä↭𝕐",
244 u"uploader": u"Philipp Hagemeister",
245 u"uploader_id": u"phihag",
246 u"upload_date": u"20121002",
247 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 .",
248 u"categories": [u'Science & Technology'],
252 u"url": u"http://www.youtube.com/watch?v=UxxajLWwzqY",
253 u"file": u"UxxajLWwzqY.mp4",
254 u"note": u"Test generic use_cipher_signature video (#897)",
256 u"upload_date": u"20120506",
257 u"title": u"Icona Pop - I Love It (feat. Charli XCX) [OFFICIAL VIDEO]",
258 u"description": u"md5:fea86fda2d5a5784273df5c7cc994d9f",
259 u"uploader": u"Icona Pop",
260 u"uploader_id": u"IconaPop"
264 u"url": u"https://www.youtube.com/watch?v=07FYdnEawAQ",
265 u"file": u"07FYdnEawAQ.mp4",
266 u"note": u"Test VEVO video with age protection (#956)",
268 u"upload_date": u"20130703",
269 u"title": u"Justin Timberlake - Tunnel Vision (Explicit)",
270 u"description": u"md5:64249768eec3bc4276236606ea996373",
271 u"uploader": u"justintimberlakeVEVO",
272 u"uploader_id": u"justintimberlakeVEVO"
276 u"url": u"//www.YouTube.com/watch?v=yZIXLfi8CZQ",
277 u"file": u"yZIXLfi8CZQ.mp4",
278 u"note": u"Embed-only video (#1746)",
280 u"upload_date": u"20120608",
281 u"title": u"Principal Sexually Assaults A Teacher - Episode 117 - 8th June 2012",
282 u"description": u"md5:09b78bd971f1e3e289601dfba15ca4f7",
283 u"uploader": u"SET India",
284 u"uploader_id": u"setindia"
288 u"url": u"http://www.youtube.com/watch?v=a9LDPn-MO4I",
289 u"file": u"a9LDPn-MO4I.m4a",
290 u"note": u"256k DASH audio (format 141) via DASH manifest",
292 u"upload_date": "20121002",
293 u"uploader_id": "8KVIDEO",
294 u"description": "No description available.",
295 u"uploader": "8KVIDEO",
296 u"title": "UHDTV TEST 8K VIDEO.mp4"
299 u"youtube_include_dash_manifest": True,
303 # DASH manifest with encrypted signature
305 u'url': u'https://www.youtube.com/watch?v=IB3lcPjvWLA',
307 u'id': u'IB3lcPjvWLA',
309 u'title': u'Afrojack - The Spark ft. Spree Wilson',
310 u'description': u'md5:9717375db5a9a3992be4668bbf3bc0a8',
311 u'uploader': u'AfrojackVEVO',
312 u'uploader_id': u'AfrojackVEVO',
313 u'upload_date': u'20131011',
316 u'youtube_include_dash_manifest': True,
324 def suitable(cls, url):
325 """Receives a URL and returns True if suitable for this IE."""
326 if YoutubePlaylistIE.suitable(url): return False
327 return re.match(cls._VALID_URL, url) is not None
329 def __init__(self, *args, **kwargs):
330 super(YoutubeIE, self).__init__(*args, **kwargs)
331 self._player_cache = {}
333 def report_video_info_webpage_download(self, video_id):
334 """Report attempt to download video info webpage."""
335 self.to_screen(u'%s: Downloading video info webpage' % video_id)
337 def report_information_extraction(self, video_id):
338 """Report attempt to extract video information."""
339 self.to_screen(u'%s: Extracting video information' % video_id)
341 def report_unavailable_format(self, video_id, format):
342 """Report extracted video URL."""
343 self.to_screen(u'%s: Format %s not available' % (video_id, format))
345 def report_rtmp_download(self):
346 """Indicate the download will use the RTMP protocol."""
347 self.to_screen(u'RTMP download detected')
349 def _extract_signature_function(self, video_id, player_url, slen):
350 id_m = re.match(r'.*-(?P<id>[a-zA-Z0-9_-]+)\.(?P<ext>[a-z]+)$',
352 player_type = id_m.group('ext')
353 player_id = id_m.group('id')
355 # Read from filesystem cache
356 func_id = '%s_%s_%d' % (player_type, player_id, slen)
357 assert os.path.basename(func_id) == func_id
358 cache_dir = get_cachedir(self._downloader.params)
360 cache_enabled = cache_dir is not None
362 cache_fn = os.path.join(os.path.expanduser(cache_dir),
366 with io.open(cache_fn, 'r', encoding='utf-8') as cachef:
367 cache_spec = json.load(cachef)
368 return lambda s: u''.join(s[i] for i in cache_spec)
370 pass # No cache available
372 if player_type == 'js':
373 code = self._download_webpage(
374 player_url, video_id,
375 note=u'Downloading %s player %s' % (player_type, player_id),
376 errnote=u'Download of %s failed' % player_url)
377 res = self._parse_sig_js(code)
378 elif player_type == 'swf':
379 urlh = self._request_webpage(
380 player_url, video_id,
381 note=u'Downloading %s player %s' % (player_type, player_id),
382 errnote=u'Download of %s failed' % player_url)
384 res = self._parse_sig_swf(code)
386 assert False, 'Invalid player type %r' % player_type
390 test_string = u''.join(map(compat_chr, range(slen)))
391 cache_res = res(test_string)
392 cache_spec = [ord(c) for c in cache_res]
394 os.makedirs(os.path.dirname(cache_fn))
395 except OSError as ose:
396 if ose.errno != errno.EEXIST:
398 write_json_file(cache_spec, cache_fn)
400 tb = traceback.format_exc()
401 self._downloader.report_warning(
402 u'Writing cache to %r failed: %s' % (cache_fn, tb))
406 def _print_sig_code(self, func, slen):
407 def gen_sig_code(idxs):
408 def _genslice(start, end, step):
409 starts = u'' if start == 0 else str(start)
410 ends = (u':%d' % (end+step)) if end + step >= 0 else u':'
411 steps = u'' if step == 1 else (u':%d' % step)
412 return u's[%s%s%s]' % (starts, ends, steps)
415 start = '(Never used)' # Quelch pyflakes warnings - start will be
416 # set as soon as step is set
417 for i, prev in zip(idxs[1:], idxs[:-1]):
421 yield _genslice(start, prev, step)
424 if i - prev in [-1, 1]:
429 yield u's[%d]' % prev
433 yield _genslice(start, i, step)
435 test_string = u''.join(map(compat_chr, range(slen)))
436 cache_res = func(test_string)
437 cache_spec = [ord(c) for c in cache_res]
438 expr_code = u' + '.join(gen_sig_code(cache_spec))
439 code = u'if len(s) == %d:\n return %s\n' % (slen, expr_code)
440 self.to_screen(u'Extracted signature function:\n' + code)
442 def _parse_sig_js(self, jscode):
443 funcname = self._search_regex(
444 r'signature=([$a-zA-Z]+)', jscode,
445 u'Initial JS player signature function name')
447 jsi = JSInterpreter(jscode)
448 initial_function = jsi.extract_function(funcname)
449 return lambda s: initial_function([s])
451 def _parse_sig_swf(self, file_contents):
452 if file_contents[1:3] != b'WS':
453 raise ExtractorError(
454 u'Not an SWF file; header is %r' % file_contents[:3])
455 if file_contents[:1] == b'C':
456 content = zlib.decompress(file_contents[8:])
458 raise NotImplementedError(u'Unsupported compression format %r' %
461 def extract_tags(content):
463 while pos < len(content):
464 header16 = struct.unpack('<H', content[pos:pos+2])[0]
466 tag_code = header16 >> 6
467 tag_len = header16 & 0x3f
469 tag_len = struct.unpack('<I', content[pos:pos+4])[0]
471 assert pos+tag_len <= len(content)
472 yield (tag_code, content[pos:pos+tag_len])
476 for tag_code, tag in extract_tags(content)
478 p = code_tag.index(b'\0', 4) + 1
479 code_reader = io.BytesIO(code_tag[p:])
481 # Parse ABC (AVM2 ByteCode)
482 def read_int(reader=None):
490 b = struct.unpack('<B', buf)[0]
491 res = res | ((b & 0x7f) << shift)
497 def u30(reader=None):
498 res = read_int(reader)
499 assert res & 0xf0000000 == 0
503 def s32(reader=None):
505 if v & 0x80000000 != 0:
506 v = - ((v ^ 0xffffffff) + 1)
509 def read_string(reader=None):
513 resb = reader.read(slen)
514 assert len(resb) == slen
515 return resb.decode('utf-8')
517 def read_bytes(count, reader=None):
520 resb = reader.read(count)
521 assert len(resb) == count
524 def read_byte(reader=None):
525 resb = read_bytes(1, reader=reader)
526 res = struct.unpack('<B', resb)[0]
529 # minor_version + major_version
534 for _c in range(1, int_count):
537 for _c in range(1, uint_count):
540 read_bytes((double_count-1) * 8)
542 constant_strings = [u'']
543 for _c in range(1, string_count):
545 constant_strings.append(s)
546 namespace_count = u30()
547 for _c in range(1, namespace_count):
551 for _c in range(1, ns_set_count):
553 for _c2 in range(count):
555 multiname_count = u30()
564 0x0e: 2, # MultinameA
565 0x1b: 1, # MultinameL
566 0x1c: 1, # MultinameLA
569 for _c in range(1, multiname_count):
571 assert kind in MULTINAME_SIZES, u'Invalid multiname kind %r' % kind
573 u30() # namespace_idx
575 multinames.append(constant_strings[name_idx])
577 multinames.append('[MULTINAME kind: %d]' % kind)
578 for _c2 in range(MULTINAME_SIZES[kind]):
583 MethodInfo = collections.namedtuple(
585 ['NEED_ARGUMENTS', 'NEED_REST'])
587 for method_id in range(method_count):
590 for _ in range(param_count):
592 u30() # name index (always 0 for youtube)
594 if flags & 0x08 != 0:
597 for c in range(option_count):
600 if flags & 0x80 != 0:
601 # Param names present
602 for _ in range(param_count):
604 mi = MethodInfo(flags & 0x01 != 0, flags & 0x04 != 0)
605 method_infos.append(mi)
608 metadata_count = u30()
609 for _c in range(metadata_count):
612 for _c2 in range(item_count):
616 def parse_traits_info():
617 trait_name_idx = u30()
618 kind_full = read_byte()
619 kind = kind_full & 0x0f
620 attrs = kind_full >> 4
622 if kind in [0x00, 0x06]: # Slot or Const
624 u30() # type_name_idx
628 elif kind in [0x01, 0x02, 0x03]: # Method / Getter / Setter
631 methods[multinames[trait_name_idx]] = method_idx
632 elif kind == 0x04: # Class
635 elif kind == 0x05: # Function
638 methods[function_idx] = multinames[trait_name_idx]
640 raise ExtractorError(u'Unsupported trait kind %d' % kind)
642 if attrs & 0x4 != 0: # Metadata present
643 metadata_count = u30()
644 for _c3 in range(metadata_count):
645 u30() # metadata index
650 TARGET_CLASSNAME = u'SignatureDecipher'
651 searched_idx = multinames.index(TARGET_CLASSNAME)
652 searched_class_id = None
654 for class_id in range(class_count):
656 if name_idx == searched_idx:
657 # We found the class we're looking for!
658 searched_class_id = class_id
659 u30() # super_name idx
661 if flags & 0x08 != 0: # Protected namespace is present
662 u30() # protected_ns_idx
664 for _c2 in range(intrf_count):
668 for _c2 in range(trait_count):
671 if searched_class_id is None:
672 raise ExtractorError(u'Target class %r not found' %
677 for class_id in range(class_count):
680 for _c2 in range(trait_count):
681 trait_methods = parse_traits_info()
682 if class_id == searched_class_id:
683 method_names.update(trait_methods.items())
684 method_idxs.update(dict(
686 for name, idx in trait_methods.items()))
690 for _c in range(script_count):
693 for _c2 in range(trait_count):
697 method_body_count = u30()
698 Method = collections.namedtuple('Method', ['code', 'local_count'])
700 for _c in range(method_body_count):
704 u30() # init_scope_depth
705 u30() # max_scope_depth
707 code = read_bytes(code_length)
708 if method_idx in method_idxs:
709 m = Method(code, local_count)
710 methods[method_idxs[method_idx]] = m
711 exception_count = u30()
712 for _c2 in range(exception_count):
719 for _c2 in range(trait_count):
722 assert p + code_reader.tell() == len(code_tag)
723 assert len(methods) == len(method_idxs)
725 method_pyfunctions = {}
727 def extract_function(func_name):
728 if func_name in method_pyfunctions:
729 return method_pyfunctions[func_name]
730 if func_name not in methods:
731 raise ExtractorError(u'Cannot find function %r' % func_name)
732 m = methods[func_name]
735 registers = ['(this)'] + list(args) + [None] * m.local_count
737 coder = io.BytesIO(m.code)
739 opcode = struct.unpack('!B', coder.read(1))[0]
740 if opcode == 36: # pushbyte
741 v = struct.unpack('!B', coder.read(1))[0]
743 elif opcode == 44: # pushstring
745 stack.append(constant_strings[idx])
746 elif opcode == 48: # pushscope
747 # We don't implement the scope register, so we'll just
748 # ignore the popped value
750 elif opcode == 70: # callproperty
752 mname = multinames[index]
753 arg_count = u30(coder)
754 args = list(reversed(
755 [stack.pop() for _ in range(arg_count)]))
757 if mname == u'split':
758 assert len(args) == 1
759 assert isinstance(args[0], compat_str)
760 assert isinstance(obj, compat_str)
764 res = obj.split(args[0])
766 elif mname == u'slice':
767 assert len(args) == 1
768 assert isinstance(args[0], int)
769 assert isinstance(obj, list)
772 elif mname == u'join':
773 assert len(args) == 1
774 assert isinstance(args[0], compat_str)
775 assert isinstance(obj, list)
776 res = args[0].join(obj)
778 elif mname in method_pyfunctions:
779 stack.append(method_pyfunctions[mname](args))
781 raise NotImplementedError(
782 u'Unsupported property %r on %r'
784 elif opcode == 72: # returnvalue
787 elif opcode == 79: # callpropvoid
789 mname = multinames[index]
790 arg_count = u30(coder)
791 args = list(reversed(
792 [stack.pop() for _ in range(arg_count)]))
794 if mname == u'reverse':
795 assert isinstance(obj, list)
798 raise NotImplementedError(
799 u'Unsupported (void) property %r on %r'
801 elif opcode == 93: # findpropstrict
803 mname = multinames[index]
804 res = extract_function(mname)
806 elif opcode == 97: # setproperty
811 assert isinstance(obj, list)
812 assert isinstance(idx, int)
814 elif opcode == 98: # getlocal
816 stack.append(registers[index])
817 elif opcode == 99: # setlocal
820 registers[index] = value
821 elif opcode == 102: # getproperty
823 pname = multinames[index]
824 if pname == u'length':
826 assert isinstance(obj, list)
827 stack.append(len(obj))
828 else: # Assume attribute access
830 assert isinstance(idx, int)
832 assert isinstance(obj, list)
833 stack.append(obj[idx])
834 elif opcode == 128: # coerce
836 elif opcode == 133: # coerce_s
837 assert isinstance(stack[-1], (type(None), compat_str))
838 elif opcode == 164: # modulo
841 res = value1 % value2
843 elif opcode == 208: # getlocal_0
844 stack.append(registers[0])
845 elif opcode == 209: # getlocal_1
846 stack.append(registers[1])
847 elif opcode == 210: # getlocal_2
848 stack.append(registers[2])
849 elif opcode == 211: # getlocal_3
850 stack.append(registers[3])
851 elif opcode == 214: # setlocal_2
852 registers[2] = stack.pop()
853 elif opcode == 215: # setlocal_3
854 registers[3] = stack.pop()
856 raise NotImplementedError(
857 u'Unsupported opcode %d' % opcode)
859 method_pyfunctions[func_name] = resfunc
862 initial_function = extract_function(u'decipher')
863 return lambda s: initial_function([s])
865 def _decrypt_signature(self, s, video_id, player_url, age_gate=False):
866 """Turn the encrypted s field into a working signature"""
868 if player_url is not None:
869 if player_url.startswith(u'//'):
870 player_url = u'https:' + player_url
872 player_id = (player_url, len(s))
873 if player_id not in self._player_cache:
874 func = self._extract_signature_function(
875 video_id, player_url, len(s)
877 self._player_cache[player_id] = func
878 func = self._player_cache[player_id]
879 if self._downloader.params.get('youtube_print_sig_code'):
880 self._print_sig_code(func, len(s))
882 except Exception as e:
883 tb = traceback.format_exc()
884 raise ExtractorError(
885 u'Automatic signature extraction failed: ' + tb, cause=e)
887 return self._static_decrypt_signature(
888 s, video_id, player_url, age_gate)
890 def _get_available_subtitles(self, video_id, webpage):
892 sub_list = self._download_webpage(
893 'https://video.google.com/timedtext?hl=en&type=list&v=%s' % video_id,
894 video_id, note=False)
895 except ExtractorError as err:
896 self._downloader.report_warning(u'unable to download video subtitles: %s' % compat_str(err))
898 lang_list = re.findall(r'name="([^"]*)"[^>]+lang_code="([\w\-]+)"', sub_list)
903 params = compat_urllib_parse.urlencode({
906 'fmt': self._downloader.params.get('subtitlesformat', 'srt'),
907 'name': unescapeHTML(l[0]).encode('utf-8'),
909 url = u'https://www.youtube.com/api/timedtext?' + params
910 sub_lang_list[lang] = url
911 if not sub_lang_list:
912 self._downloader.report_warning(u'video doesn\'t have subtitles')
916 def _get_available_automatic_caption(self, video_id, webpage):
917 """We need the webpage for getting the captions url, pass it as an
918 argument to speed up the process."""
919 sub_format = self._downloader.params.get('subtitlesformat', 'srt')
920 self.to_screen(u'%s: Looking for automatic captions' % video_id)
921 mobj = re.search(r';ytplayer.config = ({.*?});', webpage)
922 err_msg = u'Couldn\'t find automatic captions for %s' % video_id
924 self._downloader.report_warning(err_msg)
926 player_config = json.loads(mobj.group(1))
928 args = player_config[u'args']
929 caption_url = args[u'ttsurl']
930 timestamp = args[u'timestamp']
931 # We get the available subtitles
932 list_params = compat_urllib_parse.urlencode({
937 list_url = caption_url + '&' + list_params
938 caption_list = self._download_xml(list_url, video_id)
939 original_lang_node = caption_list.find('track')
940 if original_lang_node is None or original_lang_node.attrib.get('kind') != 'asr' :
941 self._downloader.report_warning(u'Video doesn\'t have automatic captions')
943 original_lang = original_lang_node.attrib['lang_code']
946 for lang_node in caption_list.findall('target'):
947 sub_lang = lang_node.attrib['lang_code']
948 params = compat_urllib_parse.urlencode({
949 'lang': original_lang,
955 sub_lang_list[sub_lang] = caption_url + '&' + params
957 # An extractor error can be raise by the download process if there are
958 # no automatic captions but there are subtitles
959 except (KeyError, ExtractorError):
960 self._downloader.report_warning(err_msg)
964 def extract_id(cls, url):
965 mobj = re.match(cls._VALID_URL, url, re.VERBOSE)
967 raise ExtractorError(u'Invalid URL: %s' % url)
968 video_id = mobj.group(2)
971 def _extract_from_m3u8(self, manifest_url, video_id):
973 def _get_urls(_manifest):
974 lines = _manifest.split('\n')
975 urls = filter(lambda l: l and not l.startswith('#'),
978 manifest = self._download_webpage(manifest_url, video_id, u'Downloading formats manifest')
979 formats_urls = _get_urls(manifest)
980 for format_url in formats_urls:
981 itag = self._search_regex(r'itag/(\d+?)/', format_url, 'itag')
982 url_map[itag] = format_url
985 def _extract_annotations(self, video_id):
986 url = 'https://www.youtube.com/annotations_invideo?features=1&legacy=1&video_id=%s' % video_id
987 return self._download_webpage(url, video_id, note=u'Searching for annotations.', errnote=u'Unable to download video annotations.')
989 def _real_extract(self, url):
991 u'http' if self._downloader.params.get('prefer_insecure', False)
994 # Extract original video URL from URL with redirection, like age verification, using next_url parameter
995 mobj = re.search(self._NEXT_URL_RE, url)
997 url = proto + '://www.youtube.com/' + compat_urllib_parse.unquote(mobj.group(1)).lstrip('/')
998 video_id = self.extract_id(url)
1001 url = proto + '://www.youtube.com/watch?v=%s&gl=US&hl=en&has_verified=1' % video_id
1002 video_webpage = self._download_webpage(url, video_id)
1004 # Attempt to extract SWF player URL
1005 mobj = re.search(r'swfConfig.*?"(https?:\\/\\/.*?watch.*?-.*?\.swf)"', video_webpage)
1006 if mobj is not None:
1007 player_url = re.sub(r'\\(.)', r'\1', mobj.group(1))
1012 self.report_video_info_webpage_download(video_id)
1013 if re.search(r'player-age-gate-content">', video_webpage) is not None:
1014 self.report_age_confirmation()
1016 # We simulate the access to the video from www.youtube.com/v/{video_id}
1017 # this can be viewed without login into Youtube
1018 data = compat_urllib_parse.urlencode({'video_id': video_id,
1019 'el': 'player_embedded',
1022 'eurl': 'https://youtube.googleapis.com/v/' + video_id,
1026 video_info_url = proto + '://www.youtube.com/get_video_info?' + data
1027 video_info_webpage = self._download_webpage(video_info_url, video_id,
1029 errnote='unable to download video info webpage')
1030 video_info = compat_parse_qs(video_info_webpage)
1033 for el_type in ['&el=embedded', '&el=detailpage', '&el=vevo', '']:
1034 video_info_url = (proto + '://www.youtube.com/get_video_info?&video_id=%s%s&ps=default&eurl=&gl=US&hl=en'
1035 % (video_id, el_type))
1036 video_info_webpage = self._download_webpage(video_info_url, video_id,
1038 errnote='unable to download video info webpage')
1039 video_info = compat_parse_qs(video_info_webpage)
1040 if 'token' in video_info:
1042 if 'token' not in video_info:
1043 if 'reason' in video_info:
1044 raise ExtractorError(
1045 u'YouTube said: %s' % video_info['reason'][0],
1046 expected=True, video_id=video_id)
1048 raise ExtractorError(
1049 u'"token" parameter not in video info for unknown reason',
1052 if 'view_count' in video_info:
1053 view_count = int(video_info['view_count'][0])
1057 # Check for "rental" videos
1058 if 'ypc_video_rental_bar_text' in video_info and 'author' not in video_info:
1059 raise ExtractorError(u'"rental" videos not supported')
1061 # Start extracting information
1062 self.report_information_extraction(video_id)
1065 if 'author' not in video_info:
1066 raise ExtractorError(u'Unable to extract uploader name')
1067 video_uploader = compat_urllib_parse.unquote_plus(video_info['author'][0])
1070 video_uploader_id = None
1071 mobj = re.search(r'<link itemprop="url" href="http://www.youtube.com/(?:user|channel)/([^"]+)">', video_webpage)
1072 if mobj is not None:
1073 video_uploader_id = mobj.group(1)
1075 self._downloader.report_warning(u'unable to extract uploader nickname')
1078 if 'title' in video_info:
1079 video_title = video_info['title'][0]
1081 self._downloader.report_warning(u'Unable to extract video title')
1085 # We try first to get a high quality image:
1086 m_thumb = re.search(r'<span itemprop="thumbnail".*?href="(.*?)">',
1087 video_webpage, re.DOTALL)
1088 if m_thumb is not None:
1089 video_thumbnail = m_thumb.group(1)
1090 elif 'thumbnail_url' not in video_info:
1091 self._downloader.report_warning(u'unable to extract video thumbnail')
1092 video_thumbnail = None
1093 else: # don't panic if we can't find it
1094 video_thumbnail = compat_urllib_parse.unquote_plus(video_info['thumbnail_url'][0])
1098 mobj = re.search(r'(?s)id="eow-date.*?>(.*?)</span>', video_webpage)
1101 r'(?s)id="watch-uploader-info".*?>.*?(?:Published|Uploaded|Streamed live) on (.*?)</strong>',
1103 if mobj is not None:
1104 upload_date = ' '.join(re.sub(r'[/,-]', r' ', mobj.group(1)).split())
1105 upload_date = unified_strdate(upload_date)
1107 m_cat_container = get_element_by_id("eow-category", video_webpage)
1109 category = self._html_search_regex(
1110 r'(?s)<a[^<]+>(.*?)</a>', m_cat_container, 'category',
1112 video_categories = None if category is None else [category]
1114 video_categories = None
1117 video_description = get_element_by_id("eow-description", video_webpage)
1118 if video_description:
1119 video_description = re.sub(r'''(?x)
1121 (?:[a-zA-Z-]+="[^"]+"\s+)*?
1123 (?:[a-zA-Z-]+="[^"]+"\s+)*?
1124 class="yt-uix-redirect-link"\s*>
1127 ''', r'\1', video_description)
1128 video_description = clean_html(video_description)
1130 fd_mobj = re.search(r'<meta name="description" content="([^"]+)"', video_webpage)
1132 video_description = unescapeHTML(fd_mobj.group(1))
1134 video_description = u''
1136 def _extract_count(klass):
1137 count = self._search_regex(
1138 r'class="%s">([\d,]+)</span>' % re.escape(klass),
1139 video_webpage, klass, default=None)
1140 if count is not None:
1141 return int(count.replace(',', ''))
1143 like_count = _extract_count(u'likes-count')
1144 dislike_count = _extract_count(u'dislikes-count')
1147 video_subtitles = self.extract_subtitles(video_id, video_webpage)
1149 if self._downloader.params.get('listsubtitles', False):
1150 self._list_available_subtitles(video_id, video_webpage)
1153 if 'length_seconds' not in video_info:
1154 self._downloader.report_warning(u'unable to extract video duration')
1155 video_duration = None
1157 video_duration = int(compat_urllib_parse.unquote_plus(video_info['length_seconds'][0]))
1160 video_annotations = None
1161 if self._downloader.params.get('writeannotations', False):
1162 video_annotations = self._extract_annotations(video_id)
1164 # Decide which formats to download
1166 mobj = re.search(r';ytplayer\.config\s*=\s*({.*?});', video_webpage)
1168 raise ValueError('Could not find vevo ID')
1169 json_code = uppercase_escape(mobj.group(1))
1170 ytplayer_config = json.loads(json_code)
1171 args = ytplayer_config['args']
1172 # Easy way to know if the 's' value is in url_encoded_fmt_stream_map
1173 # this signatures are encrypted
1174 if 'url_encoded_fmt_stream_map' not in args:
1175 raise ValueError(u'No stream_map present') # caught below
1176 re_signature = re.compile(r'[&,]s=')
1177 m_s = re_signature.search(args['url_encoded_fmt_stream_map'])
1179 self.to_screen(u'%s: Encrypted signatures detected.' % video_id)
1180 video_info['url_encoded_fmt_stream_map'] = [args['url_encoded_fmt_stream_map']]
1181 m_s = re_signature.search(args.get('adaptive_fmts', u''))
1183 if 'adaptive_fmts' in video_info:
1184 video_info['adaptive_fmts'][0] += ',' + args['adaptive_fmts']
1186 video_info['adaptive_fmts'] = [args['adaptive_fmts']]
1190 def _map_to_format_list(urlmap):
1192 for itag, video_real_url in urlmap.items():
1195 'url': video_real_url,
1196 'player_url': player_url,
1198 if itag in self._formats:
1199 dct.update(self._formats[itag])
1203 if 'conn' in video_info and video_info['conn'][0].startswith('rtmp'):
1204 self.report_rtmp_download()
1206 'format_id': '_rtmp',
1208 'url': video_info['conn'][0],
1209 'player_url': player_url,
1211 elif len(video_info.get('url_encoded_fmt_stream_map', [])) >= 1 or len(video_info.get('adaptive_fmts', [])) >= 1:
1212 encoded_url_map = video_info.get('url_encoded_fmt_stream_map', [''])[0] + ',' + video_info.get('adaptive_fmts',[''])[0]
1213 if 'rtmpe%3Dyes' in encoded_url_map:
1214 raise ExtractorError('rtmpe downloads are not supported, see https://github.com/rg3/youtube-dl/issues/343 for more information.', expected=True)
1216 for url_data_str in encoded_url_map.split(','):
1217 url_data = compat_parse_qs(url_data_str)
1218 if 'itag' in url_data and 'url' in url_data:
1219 url = url_data['url'][0]
1220 if 'sig' in url_data:
1221 url += '&signature=' + url_data['sig'][0]
1222 elif 's' in url_data:
1223 encrypted_sig = url_data['s'][0]
1224 if self._downloader.params.get('verbose'):
1226 if player_url is None:
1227 player_version = 'unknown'
1229 player_version = self._search_regex(
1230 r'-(.+)\.swf$', player_url,
1231 u'flash player', fatal=False)
1232 player_desc = 'flash player %s' % player_version
1234 player_version = self._search_regex(
1235 r'html5player-(.+?)\.js', video_webpage,
1236 'html5 player', fatal=False)
1237 player_desc = u'html5 player %s' % player_version
1239 parts_sizes = u'.'.join(compat_str(len(part)) for part in encrypted_sig.split('.'))
1240 self.to_screen(u'encrypted signature length %d (%s), itag %s, %s' %
1241 (len(encrypted_sig), parts_sizes, url_data['itag'][0], player_desc))
1244 jsplayer_url_json = self._search_regex(
1245 r'"assets":.+?"js":\s*("[^"]+")',
1246 video_webpage, u'JS player URL')
1247 player_url = json.loads(jsplayer_url_json)
1249 signature = self._decrypt_signature(
1250 encrypted_sig, video_id, player_url, age_gate)
1251 url += '&signature=' + signature
1252 if 'ratebypass' not in url:
1253 url += '&ratebypass=yes'
1254 url_map[url_data['itag'][0]] = url
1255 formats = _map_to_format_list(url_map)
1256 elif video_info.get('hlsvp'):
1257 manifest_url = video_info['hlsvp'][0]
1258 url_map = self._extract_from_m3u8(manifest_url, video_id)
1259 formats = _map_to_format_list(url_map)
1261 raise ExtractorError(u'no conn, hlsvp or url_encoded_fmt_stream_map information found in video info')
1263 # Look for the DASH manifest
1264 if (self._downloader.params.get('youtube_include_dash_manifest', False)):
1266 # The DASH manifest used needs to be the one from the original video_webpage.
1267 # The one found in get_video_info seems to be using different signatures.
1268 # However, in the case of an age restriction there won't be any embedded dashmpd in the video_webpage.
1269 # Luckily, it seems, this case uses some kind of default signature (len == 86), so the
1270 # combination of get_video_info and the _static_decrypt_signature() decryption fallback will work here.
1272 dash_manifest_url = video_info.get('dashmpd')[0]
1274 dash_manifest_url = ytplayer_config['args']['dashmpd']
1275 def decrypt_sig(mobj):
1277 dec_s = self._decrypt_signature(s, video_id, player_url, age_gate)
1278 return '/signature/%s' % dec_s
1279 dash_manifest_url = re.sub(r'/s/([\w\.]+)', decrypt_sig, dash_manifest_url)
1280 dash_doc = self._download_xml(
1281 dash_manifest_url, video_id,
1282 note=u'Downloading DASH manifest',
1283 errnote=u'Could not download DASH manifest')
1284 for r in dash_doc.findall(u'.//{urn:mpeg:DASH:schema:MPD:2011}Representation'):
1285 url_el = r.find('{urn:mpeg:DASH:schema:MPD:2011}BaseURL')
1288 format_id = r.attrib['id']
1289 video_url = url_el.text
1290 filesize = int_or_none(url_el.attrib.get('{http://youtube.com/yt/2012/10/10}contentLength'))
1292 'format_id': format_id,
1294 'width': int_or_none(r.attrib.get('width')),
1295 'tbr': int_or_none(r.attrib.get('bandwidth'), 1000),
1296 'asr': int_or_none(r.attrib.get('audioSamplingRate')),
1297 'filesize': filesize,
1300 existing_format = next(
1301 fo for fo in formats
1302 if fo['format_id'] == format_id)
1303 except StopIteration:
1304 f.update(self._formats.get(format_id, {}))
1307 existing_format.update(f)
1309 except (ExtractorError, KeyError) as e:
1310 self.report_warning(u'Skipping DASH manifest: %s' % e, video_id)
1312 self._sort_formats(formats)
1316 'uploader': video_uploader,
1317 'uploader_id': video_uploader_id,
1318 'upload_date': upload_date,
1319 'title': video_title,
1320 'thumbnail': video_thumbnail,
1321 'description': video_description,
1322 'categories': video_categories,
1323 'subtitles': video_subtitles,
1324 'duration': video_duration,
1325 'age_limit': 18 if age_gate else 0,
1326 'annotations': video_annotations,
1327 'webpage_url': proto + '://www.youtube.com/watch?v=%s' % video_id,
1328 'view_count': view_count,
1329 'like_count': like_count,
1330 'dislike_count': dislike_count,
1334 class YoutubePlaylistIE(YoutubeBaseInfoExtractor):
1335 IE_DESC = u'YouTube.com playlists'
1336 _VALID_URL = r"""(?x)(?:
1341 (?:course|view_play_list|my_playlists|artist|playlist|watch)
1342 \? (?:.*?&)*? (?:p|a|list)=
1346 (?:PL|LL|EC|UU|FL|RD)?[0-9A-Za-z-_]{10,}
1347 # Top tracks, they can also include dots
1352 ((?:PL|LL|EC|UU|FL|RD)[0-9A-Za-z-_]{10,})
1354 _TEMPLATE_URL = 'https://www.youtube.com/playlist?list=%s'
1355 _MORE_PAGES_INDICATOR = r'data-link-type="next"'
1356 _VIDEO_RE = r'href="\s*/watch\?v=(?P<id>[0-9A-Za-z_-]{11})&[^"]*?index=(?P<index>\d+)'
1357 IE_NAME = u'youtube:playlist'
1359 def _real_initialize(self):
1362 def _ids_to_results(self, ids):
1363 return [self.url_result(vid_id, 'Youtube', video_id=vid_id)
1366 def _extract_mix(self, playlist_id):
1367 # The mixes are generated from a a single video
1368 # the id of the playlist is just 'RD' + video_id
1369 url = 'https://youtube.com/watch?v=%s&list=%s' % (playlist_id[-11:], playlist_id)
1370 webpage = self._download_webpage(url, playlist_id, u'Downloading Youtube mix')
1371 search_title = lambda class_name: get_element_by_attribute('class', class_name, webpage)
1372 title_span = (search_title('playlist-title') or
1373 search_title('title long-title') or search_title('title'))
1374 title = clean_html(title_span)
1375 video_re = r'''(?x)data-video-username=".*?".*?
1376 href="/watch\?v=([0-9A-Za-z_-]{11})&[^"]*?list=%s''' % re.escape(playlist_id)
1377 ids = orderedSet(re.findall(video_re, webpage, flags=re.DOTALL))
1378 url_results = self._ids_to_results(ids)
1380 return self.playlist_result(url_results, playlist_id, title)
1382 def _real_extract(self, url):
1383 # Extract playlist id
1384 mobj = re.match(self._VALID_URL, url)
1386 raise ExtractorError(u'Invalid URL: %s' % url)
1387 playlist_id = mobj.group(1) or mobj.group(2)
1389 # Check if it's a video-specific URL
1390 query_dict = compat_urlparse.parse_qs(compat_urlparse.urlparse(url).query)
1391 if 'v' in query_dict:
1392 video_id = query_dict['v'][0]
1393 if self._downloader.params.get('noplaylist'):
1394 self.to_screen(u'Downloading just video %s because of --no-playlist' % video_id)
1395 return self.url_result(video_id, 'Youtube', video_id=video_id)
1397 self.to_screen(u'Downloading playlist %s - add --no-playlist to just download video %s' % (playlist_id, video_id))
1399 if playlist_id.startswith('RD'):
1400 # Mixes require a custom extraction process
1401 return self._extract_mix(playlist_id)
1402 if playlist_id.startswith('TL'):
1403 raise ExtractorError(u'For downloading YouTube.com top lists, use '
1404 u'the "yttoplist" keyword, for example "youtube-dl \'yttoplist:music:Top Tracks\'"', expected=True)
1406 url = self._TEMPLATE_URL % playlist_id
1407 page = self._download_webpage(url, playlist_id)
1408 more_widget_html = content_html = page
1410 # Check if the playlist exists or is private
1411 if re.search(r'<div class="yt-alert-message">[^<]*?(The|This) playlist (does not exist|is private)[^<]*?</div>', page) is not None:
1412 raise ExtractorError(
1413 u'The playlist doesn\'t exist or is private, use --username or '
1414 '--netrc to access it.',
1417 # Extract the video ids from the playlist pages
1420 for page_num in itertools.count(1):
1421 matches = re.finditer(self._VIDEO_RE, content_html)
1422 # We remove the duplicates and the link with index 0
1423 # (it's not the first video of the playlist)
1424 new_ids = orderedSet(m.group('id') for m in matches if m.group('index') != '0')
1427 mobj = re.search(r'data-uix-load-more-href="/?(?P<more>[^"]+)"', more_widget_html)
1431 more = self._download_json(
1432 'https://youtube.com/%s' % mobj.group('more'), playlist_id,
1433 'Downloading page #%s' % page_num,
1434 transform_source=uppercase_escape)
1435 content_html = more['content_html']
1436 more_widget_html = more['load_more_widget_html']
1438 playlist_title = self._html_search_regex(
1439 r'(?s)<h1 class="pl-header-title[^"]*">\s*(.*?)\s*</h1>',
1442 url_results = self._ids_to_results(ids)
1443 return self.playlist_result(url_results, playlist_id, playlist_title)
1446 class YoutubeTopListIE(YoutubePlaylistIE):
1447 IE_NAME = u'youtube:toplist'
1448 IE_DESC = (u'YouTube.com top lists, "yttoplist:{channel}:{list title}"'
1449 u' (Example: "yttoplist:music:Top Tracks")')
1450 _VALID_URL = r'yttoplist:(?P<chann>.*?):(?P<title>.*?)$'
1452 def _real_extract(self, url):
1453 mobj = re.match(self._VALID_URL, url)
1454 channel = mobj.group('chann')
1455 title = mobj.group('title')
1456 query = compat_urllib_parse.urlencode({'title': title})
1457 playlist_re = 'href="([^"]+?%s.*?)"' % re.escape(query)
1458 channel_page = self._download_webpage('https://www.youtube.com/%s' % channel, title)
1459 link = self._html_search_regex(playlist_re, channel_page, u'list')
1460 url = compat_urlparse.urljoin('https://www.youtube.com/', link)
1462 video_re = r'data-index="\d+".*?data-video-id="([0-9A-Za-z_-]{11})"'
1464 # sometimes the webpage doesn't contain the videos
1465 # retry until we get them
1466 for i in itertools.count(0):
1467 msg = u'Downloading Youtube mix'
1469 msg += ', retry #%d' % i
1470 webpage = self._download_webpage(url, title, msg)
1471 ids = orderedSet(re.findall(video_re, webpage))
1474 url_results = self._ids_to_results(ids)
1475 return self.playlist_result(url_results, playlist_title=title)
1478 class YoutubeChannelIE(InfoExtractor):
1479 IE_DESC = u'YouTube.com channels'
1480 _VALID_URL = r"^(?:https?://)?(?:youtu\.be|(?:\w+\.)?youtube(?:-nocookie)?\.com)/channel/([0-9A-Za-z_-]+)"
1481 _MORE_PAGES_INDICATOR = 'yt-uix-load-more'
1482 _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'
1483 IE_NAME = u'youtube:channel'
1485 def extract_videos_from_page(self, page):
1487 for mobj in re.finditer(r'href="/watch\?v=([0-9A-Za-z_-]+)&?', page):
1488 if mobj.group(1) not in ids_in_page:
1489 ids_in_page.append(mobj.group(1))
1492 def _real_extract(self, url):
1493 # Extract channel id
1494 mobj = re.match(self._VALID_URL, url)
1496 raise ExtractorError(u'Invalid URL: %s' % url)
1498 # Download channel page
1499 channel_id = mobj.group(1)
1501 url = 'https://www.youtube.com/channel/%s/videos' % channel_id
1502 channel_page = self._download_webpage(url, channel_id)
1503 autogenerated = re.search(r'''(?x)
1505 channel-header-autogenerated-label|
1506 yt-channel-title-autogenerated
1507 )[^"]*"''', channel_page) is not None
1510 # The videos are contained in a single page
1511 # the ajax pages can't be used, they are empty
1512 video_ids = self.extract_videos_from_page(channel_page)
1514 # Download all channel pages using the json-based channel_ajax query
1515 for pagenum in itertools.count(1):
1516 url = self._MORE_PAGES_URL % (pagenum, channel_id)
1517 page = self._download_json(
1518 url, channel_id, note=u'Downloading page #%s' % pagenum,
1519 transform_source=uppercase_escape)
1521 ids_in_page = self.extract_videos_from_page(page['content_html'])
1522 video_ids.extend(ids_in_page)
1524 if self._MORE_PAGES_INDICATOR not in page['load_more_widget_html']:
1527 self._downloader.to_screen(u'[youtube] Channel %s: Found %i videos' % (channel_id, len(video_ids)))
1529 url_entries = [self.url_result(video_id, 'Youtube', video_id=video_id)
1530 for video_id in video_ids]
1531 return self.playlist_result(url_entries, channel_id)
1534 class YoutubeUserIE(InfoExtractor):
1535 IE_DESC = u'YouTube.com user videos (URL or "ytuser" keyword)'
1536 _VALID_URL = r'(?:(?:(?:https?://)?(?:\w+\.)?youtube\.com/(?:user/)?(?!(?:attribution_link|watch|results)(?:$|[^a-z_A-Z0-9-])))|ytuser:)(?!feed/)([A-Za-z0-9_-]+)'
1537 _TEMPLATE_URL = 'https://gdata.youtube.com/feeds/api/users/%s'
1538 _GDATA_PAGE_SIZE = 50
1539 _GDATA_URL = 'https://gdata.youtube.com/feeds/api/users/%s/uploads?max-results=%d&start-index=%d&alt=json'
1540 IE_NAME = u'youtube:user'
1543 def suitable(cls, url):
1544 # Don't return True if the url can be extracted with other youtube
1545 # extractor, the regex would is too permissive and it would match.
1546 other_ies = iter(klass for (name, klass) in globals().items() if name.endswith('IE') and klass is not cls)
1547 if any(ie.suitable(url) for ie in other_ies): return False
1548 else: return super(YoutubeUserIE, cls).suitable(url)
1550 def _real_extract(self, url):
1552 mobj = re.match(self._VALID_URL, url)
1554 raise ExtractorError(u'Invalid URL: %s' % url)
1556 username = mobj.group(1)
1558 # Download video ids using YouTube Data API. Result size per
1559 # query is limited (currently to 50 videos) so we need to query
1560 # page by page until there are no video ids - it means we got
1563 def download_page(pagenum):
1564 start_index = pagenum * self._GDATA_PAGE_SIZE + 1
1566 gdata_url = self._GDATA_URL % (username, self._GDATA_PAGE_SIZE, start_index)
1567 page = self._download_webpage(
1568 gdata_url, username,
1569 u'Downloading video ids from %d to %d' % (
1570 start_index, start_index + self._GDATA_PAGE_SIZE))
1573 response = json.loads(page)
1574 except ValueError as err:
1575 raise ExtractorError(u'Invalid JSON in API response: ' + compat_str(err))
1576 if 'entry' not in response['feed']:
1579 # Extract video identifiers
1580 entries = response['feed']['entry']
1581 for entry in entries:
1582 title = entry['title']['$t']
1583 video_id = entry['id']['$t'].split('/')[-1]
1587 'ie_key': 'Youtube',
1591 url_results = PagedList(download_page, self._GDATA_PAGE_SIZE)
1593 return self.playlist_result(url_results, playlist_title=username)
1596 class YoutubeSearchIE(SearchInfoExtractor):
1597 IE_DESC = u'YouTube.com searches'
1598 _API_URL = u'https://gdata.youtube.com/feeds/api/videos?q=%s&start-index=%i&max-results=50&v=2&alt=jsonc'
1600 IE_NAME = u'youtube:search'
1601 _SEARCH_KEY = 'ytsearch'
1603 def _get_n_results(self, query, n):
1604 """Get a specified number of results for a query"""
1611 while (PAGE_SIZE * pagenum) < limit:
1612 result_url = self._API_URL % (
1613 compat_urllib_parse.quote_plus(query.encode('utf-8')),
1614 (PAGE_SIZE * pagenum) + 1)
1615 data_json = self._download_webpage(
1616 result_url, video_id=u'query "%s"' % query,
1617 note=u'Downloading page %s' % (pagenum + 1),
1618 errnote=u'Unable to download API page')
1619 data = json.loads(data_json)
1620 api_response = data['data']
1622 if 'items' not in api_response:
1623 raise ExtractorError(
1624 u'[youtube] No video results', expected=True)
1626 new_ids = list(video['id'] for video in api_response['items'])
1627 video_ids += new_ids
1629 limit = min(n, api_response['totalItems'])
1632 if len(video_ids) > n:
1633 video_ids = video_ids[:n]
1634 videos = [self.url_result(video_id, 'Youtube', video_id=video_id)
1635 for video_id in video_ids]
1636 return self.playlist_result(videos, query)
1639 class YoutubeSearchDateIE(YoutubeSearchIE):
1640 IE_NAME = YoutubeSearchIE.IE_NAME + ':date'
1641 _API_URL = 'https://gdata.youtube.com/feeds/api/videos?q=%s&start-index=%i&max-results=50&v=2&alt=jsonc&orderby=published'
1642 _SEARCH_KEY = 'ytsearchdate'
1643 IE_DESC = u'YouTube.com searches, newest videos first'
1646 class YoutubeSearchURLIE(InfoExtractor):
1647 IE_DESC = u'YouTube.com search URLs'
1648 IE_NAME = u'youtube:search_url'
1649 _VALID_URL = r'https?://(?:www\.)?youtube\.com/results\?(.*?&)?search_query=(?P<query>[^&]+)(?:[&]|$)'
1651 def _real_extract(self, url):
1652 mobj = re.match(self._VALID_URL, url)
1653 query = compat_urllib_parse.unquote_plus(mobj.group('query'))
1655 webpage = self._download_webpage(url, query)
1656 result_code = self._search_regex(
1657 r'(?s)<ol class="item-section"(.*?)</ol>', webpage, u'result HTML')
1659 part_codes = re.findall(
1660 r'(?s)<h3 class="yt-lockup-title">(.*?)</h3>', result_code)
1662 for part_code in part_codes:
1663 part_title = self._html_search_regex(
1664 [r'(?s)title="([^"]+)"', r'>([^<]+)</a>'], part_code, 'item title', fatal=False)
1665 part_url_snippet = self._html_search_regex(
1666 r'(?s)href="([^"]+)"', part_code, 'item URL')
1667 part_url = compat_urlparse.urljoin(
1668 'https://www.youtube.com/', part_url_snippet)
1672 'title': part_title,
1676 '_type': 'playlist',
1682 class YoutubeShowIE(InfoExtractor):
1683 IE_DESC = u'YouTube.com (multi-season) shows'
1684 _VALID_URL = r'https?://www\.youtube\.com/show/(.*)'
1685 IE_NAME = u'youtube:show'
1687 def _real_extract(self, url):
1688 mobj = re.match(self._VALID_URL, url)
1689 show_name = mobj.group(1)
1690 webpage = self._download_webpage(url, show_name, u'Downloading show webpage')
1691 # There's one playlist for each season of the show
1692 m_seasons = list(re.finditer(r'href="(/playlist\?list=.*?)"', webpage))
1693 self.to_screen(u'%s: Found %s seasons' % (show_name, len(m_seasons)))
1694 return [self.url_result('https://www.youtube.com' + season.group(1), 'YoutubePlaylist') for season in m_seasons]
1697 class YoutubeFeedsInfoExtractor(YoutubeBaseInfoExtractor):
1699 Base class for extractors that fetch info from
1700 http://www.youtube.com/feed_ajax
1701 Subclasses must define the _FEED_NAME and _PLAYLIST_TITLE properties.
1703 _LOGIN_REQUIRED = True
1704 # use action_load_personal_feed instead of action_load_system_feed
1705 _PERSONAL_FEED = False
1708 def _FEED_TEMPLATE(self):
1709 action = 'action_load_system_feed'
1710 if self._PERSONAL_FEED:
1711 action = 'action_load_personal_feed'
1712 return 'https://www.youtube.com/feed_ajax?%s=1&feed_name=%s&paging=%%s' % (action, self._FEED_NAME)
1716 return u'youtube:%s' % self._FEED_NAME
1718 def _real_initialize(self):
1721 def _real_extract(self, url):
1724 for i in itertools.count(1):
1725 info = self._download_json(self._FEED_TEMPLATE % paging,
1726 u'%s feed' % self._FEED_NAME,
1727 u'Downloading page %s' % i)
1728 feed_html = info.get('feed_html') or info.get('content_html')
1729 m_ids = re.finditer(r'"/watch\?v=(.*?)["&]', feed_html)
1730 ids = orderedSet(m.group(1) for m in m_ids)
1731 feed_entries.extend(
1732 self.url_result(video_id, 'Youtube', video_id=video_id)
1733 for video_id in ids)
1735 r'data-uix-load-more-href="/?[^"]+paging=(?P<paging>\d+)',
1739 paging = mobj.group('paging')
1740 return self.playlist_result(feed_entries, playlist_title=self._PLAYLIST_TITLE)
1742 class YoutubeSubscriptionsIE(YoutubeFeedsInfoExtractor):
1743 IE_DESC = u'YouTube.com subscriptions feed, "ytsubs" keyword (requires authentication)'
1744 _VALID_URL = r'https?://www\.youtube\.com/feed/subscriptions|:ytsubs(?:criptions)?'
1745 _FEED_NAME = 'subscriptions'
1746 _PLAYLIST_TITLE = u'Youtube Subscriptions'
1748 class YoutubeRecommendedIE(YoutubeFeedsInfoExtractor):
1749 IE_DESC = u'YouTube.com recommended videos, "ytrec" keyword (requires authentication)'
1750 _VALID_URL = r'https?://www\.youtube\.com/feed/recommended|:ytrec(?:ommended)?'
1751 _FEED_NAME = 'recommended'
1752 _PLAYLIST_TITLE = u'Youtube Recommended videos'
1754 class YoutubeWatchLaterIE(YoutubeFeedsInfoExtractor):
1755 IE_DESC = u'Youtube watch later list, "ytwatchlater" keyword (requires authentication)'
1756 _VALID_URL = r'https?://www\.youtube\.com/feed/watch_later|:ytwatchlater'
1757 _FEED_NAME = 'watch_later'
1758 _PLAYLIST_TITLE = u'Youtube Watch Later'
1759 _PERSONAL_FEED = True
1761 class YoutubeHistoryIE(YoutubeFeedsInfoExtractor):
1762 IE_DESC = u'Youtube watch history, "ythistory" keyword (requires authentication)'
1763 _VALID_URL = u'https?://www\.youtube\.com/feed/history|:ythistory'
1764 _FEED_NAME = 'history'
1765 _PERSONAL_FEED = True
1766 _PLAYLIST_TITLE = u'Youtube Watch History'
1768 class YoutubeFavouritesIE(YoutubeBaseInfoExtractor):
1769 IE_NAME = u'youtube:favorites'
1770 IE_DESC = u'YouTube.com favourite videos, "ytfav" keyword (requires authentication)'
1771 _VALID_URL = r'https?://www\.youtube\.com/my_favorites|:ytfav(?:ou?rites)?'
1772 _LOGIN_REQUIRED = True
1774 def _real_extract(self, url):
1775 webpage = self._download_webpage('https://www.youtube.com/my_favorites', 'Youtube Favourites videos')
1776 playlist_id = self._search_regex(r'list=(.+?)["&]', webpage, u'favourites playlist id')
1777 return self.url_result(playlist_id, 'YoutubePlaylist')
1780 class YoutubeTruncatedURLIE(InfoExtractor):
1781 IE_NAME = 'youtube:truncated_url'
1782 IE_DESC = False # Do not list
1783 _VALID_URL = r'''(?x)
1784 (?:https?://)?[^/]+/watch\?(?:
1786 annotation_id=annotation_[^&]+
1788 (?:https?://)?(?:www\.)?youtube\.com/attribution_link\?a=[^&]+$
1792 'url': 'http://www.youtube.com/watch?annotation_id=annotation_3951667041',
1793 'only_matching': True,
1795 'url': 'http://www.youtube.com/watch?',
1796 'only_matching': True,
1799 def _real_extract(self, url):
1800 raise ExtractorError(
1801 u'Did you forget to quote the URL? Remember that & is a meta '
1802 u'character in most shells, so you want to put the URL in quotes, '
1804 u'"http://www.youtube.com/watch?feature=foo&v=BaW_jenozKc" '
1805 u' or simply youtube-dl BaW_jenozKc .',