15 from .common import InfoExtractor, SearchInfoExtractor
16 from .subtitles import SubtitlesInfoExtractor
21 compat_urllib_request,
28 get_element_by_attribute,
38 class YoutubeBaseInfoExtractor(InfoExtractor):
39 """Provide base functions for Youtube extractors"""
40 _LOGIN_URL = 'https://accounts.google.com/ServiceLogin'
41 _LANG_URL = r'https://www.youtube.com/?hl=en&persist_hl=1&gl=US&persist_gl=1&opt_out_ackd=1'
42 _AGE_URL = 'http://www.youtube.com/verify_age?next_url=/&gl=US&hl=en'
43 _NETRC_MACHINE = 'youtube'
44 # If True it will raise an error if no login info is provided
45 _LOGIN_REQUIRED = False
47 def _set_language(self):
48 return bool(self._download_webpage(
50 note=u'Setting language', errnote='unable to set language',
54 (username, password) = self._get_login_info()
55 # No authentication to be performed
57 if self._LOGIN_REQUIRED:
58 raise ExtractorError(u'No login info available, needed for using %s.' % self.IE_NAME, expected=True)
61 login_page = self._download_webpage(
62 self._LOGIN_URL, None,
63 note=u'Downloading login page',
64 errnote=u'unable to fetch login page', fatal=False)
65 if login_page is False:
68 galx = self._search_regex(r'(?s)<input.+?name="GALX".+?value="(.+?)"',
69 login_page, u'Login GALX parameter')
73 u'continue': u'https://www.youtube.com/signin?action_handle_signin=true&feature=sign_in_button&hl=en_US&nomobiletemp=1',
77 u'PersistentCookie': u'yes',
79 u'bgresponse': u'js_disabled',
80 u'checkConnection': u'',
81 u'checkedDomains': u'youtube',
86 u'signIn': u'Sign in',
88 u'service': u'youtube',
92 # Convert to UTF-8 *before* urlencode because Python 2.x's urlencode
94 login_form = dict((k.encode('utf-8'), v.encode('utf-8')) for k,v in login_form_strs.items())
95 login_data = compat_urllib_parse.urlencode(login_form).encode('ascii')
97 req = compat_urllib_request.Request(self._LOGIN_URL, login_data)
98 login_results = self._download_webpage(
100 note=u'Logging in', errnote=u'unable to log in', fatal=False)
101 if login_results is False:
103 if re.search(r'(?i)<form[^>]* id="gaia_loginform"', login_results) is not None:
104 self._downloader.report_warning(u'unable to log in: bad username or password')
108 def _confirm_age(self):
111 'action_confirm': 'Confirm',
113 req = compat_urllib_request.Request(self._AGE_URL, compat_urllib_parse.urlencode(age_form))
115 self._download_webpage(
117 note=u'Confirming age', errnote=u'Unable to confirm age')
120 def _real_initialize(self):
121 if self._downloader is None:
123 if not self._set_language():
125 if not self._login():
130 class YoutubeIE(YoutubeBaseInfoExtractor, SubtitlesInfoExtractor):
131 IE_DESC = u'YouTube.com'
132 _VALID_URL = r"""(?x)^
134 (?:https?://|//)? # http(s):// or protocol-independent URL (optional)
135 (?:(?:(?:(?:\w+\.)?[yY][oO][uU][tT][uU][bB][eE](?:-nocookie)?\.com/|
136 (?:www\.)?deturl\.com/www\.youtube\.com/|
137 (?:www\.)?pwnyoutube\.com|
138 tube\.majestyc\.net/|
139 youtube\.googleapis\.com/) # the various hostnames, with wildcard subdomains
140 (?:.*?\#/)? # handle anchor (#/) redirect urls
141 (?: # the various things that can precede the ID:
142 (?:(?:v|embed|e)/) # v/ or embed/ or e/
143 |(?: # or the v= param in all its forms
144 (?:(?:watch|movie)(?:_popup)?(?:\.php)?)? # preceding watch(_popup|.php) or nothing (like /?v=xxxx)
145 (?:\?|\#!?) # the params delimiter ? or # or #!
146 (?:.*?&)? # any other preceding param (like /?s=tuff&v=xxxx)
150 |youtu\.be/ # just youtu.be/xxxx
152 )? # all until now is optional -> you can pass the naked ID
153 ([0-9A-Za-z_-]{11}) # here is it! the YouTube video ID
154 (?(1).+)? # if we found the ID, everything can follow
156 _NEXT_URL_RE = r'[\?&]next_url=([^&]+)'
158 '5': {'ext': 'flv', 'width': 400, 'height': 240},
159 '6': {'ext': 'flv', 'width': 450, 'height': 270},
160 '13': {'ext': '3gp'},
161 '17': {'ext': '3gp', 'width': 176, 'height': 144},
162 '18': {'ext': 'mp4', 'width': 640, 'height': 360},
163 '22': {'ext': 'mp4', 'width': 1280, 'height': 720},
164 '34': {'ext': 'flv', 'width': 640, 'height': 360},
165 '35': {'ext': 'flv', 'width': 854, 'height': 480},
166 '36': {'ext': '3gp', 'width': 320, 'height': 240},
167 '37': {'ext': 'mp4', 'width': 1920, 'height': 1080},
168 '38': {'ext': 'mp4', 'width': 4096, 'height': 3072},
169 '43': {'ext': 'webm', 'width': 640, 'height': 360},
170 '44': {'ext': 'webm', 'width': 854, 'height': 480},
171 '45': {'ext': 'webm', 'width': 1280, 'height': 720},
172 '46': {'ext': 'webm', 'width': 1920, 'height': 1080},
176 '82': {'ext': 'mp4', 'height': 360, 'resolution': '360p', 'format_note': '3D', 'preference': -20},
177 '83': {'ext': 'mp4', 'height': 480, 'resolution': '480p', 'format_note': '3D', 'preference': -20},
178 '84': {'ext': 'mp4', 'height': 720, 'resolution': '720p', 'format_note': '3D', 'preference': -20},
179 '85': {'ext': 'mp4', 'height': 1080, 'resolution': '1080p', 'format_note': '3D', 'preference': -20},
180 '100': {'ext': 'webm', 'height': 360, 'resolution': '360p', 'format_note': '3D', 'preference': -20},
181 '101': {'ext': 'webm', 'height': 480, 'resolution': '480p', 'format_note': '3D', 'preference': -20},
182 '102': {'ext': 'webm', 'height': 720, 'resolution': '720p', 'format_note': '3D', 'preference': -20},
184 # Apple HTTP Live Streaming
185 '92': {'ext': 'mp4', 'height': 240, 'resolution': '240p', 'format_note': 'HLS', 'preference': -10},
186 '93': {'ext': 'mp4', 'height': 360, 'resolution': '360p', 'format_note': 'HLS', 'preference': -10},
187 '94': {'ext': 'mp4', 'height': 480, 'resolution': '480p', 'format_note': 'HLS', 'preference': -10},
188 '95': {'ext': 'mp4', 'height': 720, 'resolution': '720p', 'format_note': 'HLS', 'preference': -10},
189 '96': {'ext': 'mp4', 'height': 1080, 'resolution': '1080p', 'format_note': 'HLS', 'preference': -10},
190 '132': {'ext': 'mp4', 'height': 240, 'resolution': '240p', 'format_note': 'HLS', 'preference': -10},
191 '151': {'ext': 'mp4', 'height': 72, 'resolution': '72p', 'format_note': 'HLS', 'preference': -10},
194 '133': {'ext': 'mp4', 'height': 240, 'resolution': '240p', 'format_note': 'DASH video', 'preference': -40},
195 '134': {'ext': 'mp4', 'height': 360, 'resolution': '360p', 'format_note': 'DASH video', 'preference': -40},
196 '135': {'ext': 'mp4', 'height': 480, 'resolution': '480p', 'format_note': 'DASH video', 'preference': -40},
197 '136': {'ext': 'mp4', 'height': 720, 'resolution': '720p', 'format_note': 'DASH video', 'preference': -40},
198 '137': {'ext': 'mp4', 'height': 1080, 'resolution': '1080p', 'format_note': 'DASH video', 'preference': -40},
199 '138': {'ext': 'mp4', 'height': 1081, 'resolution': '>1080p', 'format_note': 'DASH video', 'preference': -40},
200 '160': {'ext': 'mp4', 'height': 192, 'resolution': '192p', 'format_note': 'DASH video', 'preference': -40},
201 '264': {'ext': 'mp4', 'height': 1080, 'resolution': '1080p', 'format_note': 'DASH video', 'preference': -40},
204 '139': {'ext': 'm4a', 'format_note': 'DASH audio', 'vcodec': 'none', 'abr': 48, 'preference': -50},
205 '140': {'ext': 'm4a', 'format_note': 'DASH audio', 'vcodec': 'none', 'abr': 128, 'preference': -50},
206 '141': {'ext': 'm4a', 'format_note': 'DASH audio', 'vcodec': 'none', 'abr': 256, 'preference': -50},
209 '242': {'ext': 'webm', 'height': 240, 'resolution': '240p', 'format_note': 'DASH webm', 'preference': -40},
210 '243': {'ext': 'webm', 'height': 360, 'resolution': '360p', 'format_note': 'DASH webm', 'preference': -40},
211 '244': {'ext': 'webm', 'height': 480, 'resolution': '480p', 'format_note': 'DASH webm', 'preference': -40},
212 '245': {'ext': 'webm', 'height': 480, 'resolution': '480p', 'format_note': 'DASH webm', 'preference': -40},
213 '246': {'ext': 'webm', 'height': 480, 'resolution': '480p', 'format_note': 'DASH webm', 'preference': -40},
214 '247': {'ext': 'webm', 'height': 720, 'resolution': '720p', 'format_note': 'DASH webm', 'preference': -40},
215 '248': {'ext': 'webm', 'height': 1080, 'resolution': '1080p', 'format_note': 'DASH webm', 'preference': -40},
218 '171': {'ext': 'webm', 'vcodec': 'none', 'format_note': 'DASH webm audio', 'abr': 48, 'preference': -50},
219 '172': {'ext': 'webm', 'vcodec': 'none', 'format_note': 'DASH webm audio', 'abr': 256, 'preference': -50},
222 '_rtmp': {'protocol': 'rtmp'},
228 u"url": u"http://www.youtube.com/watch?v=BaW_jenozKc",
229 u"file": u"BaW_jenozKc.mp4",
231 u"title": u"youtube-dl test video \"'/\\ä↭𝕐",
232 u"uploader": u"Philipp Hagemeister",
233 u"uploader_id": u"phihag",
234 u"upload_date": u"20121002",
235 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 ."
239 u"url": u"http://www.youtube.com/watch?v=UxxajLWwzqY",
240 u"file": u"UxxajLWwzqY.mp4",
241 u"note": u"Test generic use_cipher_signature video (#897)",
243 u"upload_date": u"20120506",
244 u"title": u"Icona Pop - I Love It (feat. Charli XCX) [OFFICIAL VIDEO]",
245 u"description": u"md5:5b292926389560516e384ac437c0ec07",
246 u"uploader": u"Icona Pop",
247 u"uploader_id": u"IconaPop"
251 u"url": u"https://www.youtube.com/watch?v=07FYdnEawAQ",
252 u"file": u"07FYdnEawAQ.mp4",
253 u"note": u"Test VEVO video with age protection (#956)",
255 u"upload_date": u"20130703",
256 u"title": u"Justin Timberlake - Tunnel Vision (Explicit)",
257 u"description": u"md5:64249768eec3bc4276236606ea996373",
258 u"uploader": u"justintimberlakeVEVO",
259 u"uploader_id": u"justintimberlakeVEVO"
263 u"url": u"//www.YouTube.com/watch?v=yZIXLfi8CZQ",
264 u"file": u"yZIXLfi8CZQ.mp4",
265 u"note": u"Embed-only video (#1746)",
267 u"upload_date": u"20120608",
268 u"title": u"Principal Sexually Assaults A Teacher - Episode 117 - 8th June 2012",
269 u"description": u"md5:09b78bd971f1e3e289601dfba15ca4f7",
270 u"uploader": u"SET India",
271 u"uploader_id": u"setindia"
278 def suitable(cls, url):
279 """Receives a URL and returns True if suitable for this IE."""
280 if YoutubePlaylistIE.suitable(url): return False
281 return re.match(cls._VALID_URL, url) is not None
283 def __init__(self, *args, **kwargs):
284 super(YoutubeIE, self).__init__(*args, **kwargs)
285 self._player_cache = {}
287 def report_video_info_webpage_download(self, video_id):
288 """Report attempt to download video info webpage."""
289 self.to_screen(u'%s: Downloading video info webpage' % video_id)
291 def report_information_extraction(self, video_id):
292 """Report attempt to extract video information."""
293 self.to_screen(u'%s: Extracting video information' % video_id)
295 def report_unavailable_format(self, video_id, format):
296 """Report extracted video URL."""
297 self.to_screen(u'%s: Format %s not available' % (video_id, format))
299 def report_rtmp_download(self):
300 """Indicate the download will use the RTMP protocol."""
301 self.to_screen(u'RTMP download detected')
303 def _extract_signature_function(self, video_id, player_url, slen):
304 id_m = re.match(r'.*-(?P<id>[a-zA-Z0-9_-]+)\.(?P<ext>[a-z]+)$',
306 player_type = id_m.group('ext')
307 player_id = id_m.group('id')
309 # Read from filesystem cache
310 func_id = '%s_%s_%d' % (player_type, player_id, slen)
311 assert os.path.basename(func_id) == func_id
312 cache_dir = get_cachedir(self._downloader.params)
314 cache_enabled = cache_dir is not None
316 cache_fn = os.path.join(os.path.expanduser(cache_dir),
320 with io.open(cache_fn, 'r', encoding='utf-8') as cachef:
321 cache_spec = json.load(cachef)
322 return lambda s: u''.join(s[i] for i in cache_spec)
324 pass # No cache available
326 if player_type == 'js':
327 code = self._download_webpage(
328 player_url, video_id,
329 note=u'Downloading %s player %s' % (player_type, player_id),
330 errnote=u'Download of %s failed' % player_url)
331 res = self._parse_sig_js(code)
332 elif player_type == 'swf':
333 urlh = self._request_webpage(
334 player_url, video_id,
335 note=u'Downloading %s player %s' % (player_type, player_id),
336 errnote=u'Download of %s failed' % player_url)
338 res = self._parse_sig_swf(code)
340 assert False, 'Invalid player type %r' % player_type
344 test_string = u''.join(map(compat_chr, range(slen)))
345 cache_res = res(test_string)
346 cache_spec = [ord(c) for c in cache_res]
348 os.makedirs(os.path.dirname(cache_fn))
349 except OSError as ose:
350 if ose.errno != errno.EEXIST:
352 write_json_file(cache_spec, cache_fn)
354 tb = traceback.format_exc()
355 self._downloader.report_warning(
356 u'Writing cache to %r failed: %s' % (cache_fn, tb))
360 def _print_sig_code(self, func, slen):
361 def gen_sig_code(idxs):
362 def _genslice(start, end, step):
363 starts = u'' if start == 0 else str(start)
364 ends = (u':%d' % (end+step)) if end + step >= 0 else u':'
365 steps = u'' if step == 1 else (u':%d' % step)
366 return u's[%s%s%s]' % (starts, ends, steps)
369 start = '(Never used)' # Quelch pyflakes warnings - start will be
370 # set as soon as step is set
371 for i, prev in zip(idxs[1:], idxs[:-1]):
375 yield _genslice(start, prev, step)
378 if i - prev in [-1, 1]:
383 yield u's[%d]' % prev
387 yield _genslice(start, i, step)
389 test_string = u''.join(map(compat_chr, range(slen)))
390 cache_res = func(test_string)
391 cache_spec = [ord(c) for c in cache_res]
392 expr_code = u' + '.join(gen_sig_code(cache_spec))
393 code = u'if len(s) == %d:\n return %s\n' % (slen, expr_code)
394 self.to_screen(u'Extracted signature function:\n' + code)
396 def _parse_sig_js(self, jscode):
397 funcname = self._search_regex(
398 r'signature=([a-zA-Z]+)', jscode,
399 u'Initial JS player signature function name')
404 return string.lowercase.index(varname)
406 def interpret_statement(stmt, local_vars, allow_recursion=20):
407 if allow_recursion < 0:
408 raise ExtractorError(u'Recursion limit reached')
410 if stmt.startswith(u'var '):
411 stmt = stmt[len(u'var '):]
412 ass_m = re.match(r'^(?P<out>[a-z]+)(?:\[(?P<index>[^\]]+)\])?' +
413 r'=(?P<expr>.*)$', stmt)
415 if ass_m.groupdict().get('index'):
417 lvar = local_vars[ass_m.group('out')]
418 idx = interpret_expression(ass_m.group('index'),
419 local_vars, allow_recursion)
420 assert isinstance(idx, int)
423 expr = ass_m.group('expr')
426 local_vars[ass_m.group('out')] = val
428 expr = ass_m.group('expr')
429 elif stmt.startswith(u'return '):
431 expr = stmt[len(u'return '):]
433 raise ExtractorError(
434 u'Cannot determine left side of statement in %r' % stmt)
436 v = interpret_expression(expr, local_vars, allow_recursion)
439 def interpret_expression(expr, local_vars, allow_recursion):
444 return local_vars[expr]
446 m = re.match(r'^(?P<in>[a-z]+)\.(?P<member>.*)$', expr)
448 member = m.group('member')
449 val = local_vars[m.group('in')]
450 if member == 'split("")':
452 if member == 'join("")':
454 if member == 'length':
456 if member == 'reverse()':
458 slice_m = re.match(r'slice\((?P<idx>.*)\)', member)
460 idx = interpret_expression(
461 slice_m.group('idx'), local_vars, allow_recursion-1)
465 r'^(?P<in>[a-z]+)\[(?P<idx>.+)\]$', expr)
467 val = local_vars[m.group('in')]
468 idx = interpret_expression(m.group('idx'), local_vars,
472 m = re.match(r'^(?P<a>.+?)(?P<op>[%])(?P<b>.+?)$', expr)
474 a = interpret_expression(m.group('a'),
475 local_vars, allow_recursion)
476 b = interpret_expression(m.group('b'),
477 local_vars, allow_recursion)
481 r'^(?P<func>[a-zA-Z]+)\((?P<args>[a-z0-9,]+)\)$', expr)
483 fname = m.group('func')
484 if fname not in functions:
485 functions[fname] = extract_function(fname)
486 argvals = [int(v) if v.isdigit() else local_vars[v]
487 for v in m.group('args').split(',')]
488 return functions[fname](argvals)
489 raise ExtractorError(u'Unsupported JS expression %r' % expr)
491 def extract_function(funcname):
493 r'function ' + re.escape(funcname) +
494 r'\((?P<args>[a-z,]+)\){(?P<code>[^}]+)}',
496 argnames = func_m.group('args').split(',')
499 local_vars = dict(zip(argnames, args))
500 for stmt in func_m.group('code').split(';'):
501 res = interpret_statement(stmt, local_vars)
505 initial_function = extract_function(funcname)
506 return lambda s: initial_function([s])
508 def _parse_sig_swf(self, file_contents):
509 if file_contents[1:3] != b'WS':
510 raise ExtractorError(
511 u'Not an SWF file; header is %r' % file_contents[:3])
512 if file_contents[:1] == b'C':
513 content = zlib.decompress(file_contents[8:])
515 raise NotImplementedError(u'Unsupported compression format %r' %
518 def extract_tags(content):
520 while pos < len(content):
521 header16 = struct.unpack('<H', content[pos:pos+2])[0]
523 tag_code = header16 >> 6
524 tag_len = header16 & 0x3f
526 tag_len = struct.unpack('<I', content[pos:pos+4])[0]
528 assert pos+tag_len <= len(content)
529 yield (tag_code, content[pos:pos+tag_len])
533 for tag_code, tag in extract_tags(content)
535 p = code_tag.index(b'\0', 4) + 1
536 code_reader = io.BytesIO(code_tag[p:])
538 # Parse ABC (AVM2 ByteCode)
539 def read_int(reader=None):
547 b = struct.unpack('<B', buf)[0]
548 res = res | ((b & 0x7f) << shift)
554 def u30(reader=None):
555 res = read_int(reader)
556 assert res & 0xf0000000 == 0
560 def s32(reader=None):
562 if v & 0x80000000 != 0:
563 v = - ((v ^ 0xffffffff) + 1)
566 def read_string(reader=None):
570 resb = reader.read(slen)
571 assert len(resb) == slen
572 return resb.decode('utf-8')
574 def read_bytes(count, reader=None):
577 resb = reader.read(count)
578 assert len(resb) == count
581 def read_byte(reader=None):
582 resb = read_bytes(1, reader=reader)
583 res = struct.unpack('<B', resb)[0]
586 # minor_version + major_version
591 for _c in range(1, int_count):
594 for _c in range(1, uint_count):
597 read_bytes((double_count-1) * 8)
599 constant_strings = [u'']
600 for _c in range(1, string_count):
602 constant_strings.append(s)
603 namespace_count = u30()
604 for _c in range(1, namespace_count):
608 for _c in range(1, ns_set_count):
610 for _c2 in range(count):
612 multiname_count = u30()
621 0x0e: 2, # MultinameA
622 0x1b: 1, # MultinameL
623 0x1c: 1, # MultinameLA
626 for _c in range(1, multiname_count):
628 assert kind in MULTINAME_SIZES, u'Invalid multiname kind %r' % kind
630 u30() # namespace_idx
632 multinames.append(constant_strings[name_idx])
634 multinames.append('[MULTINAME kind: %d]' % kind)
635 for _c2 in range(MULTINAME_SIZES[kind]):
640 MethodInfo = collections.namedtuple(
642 ['NEED_ARGUMENTS', 'NEED_REST'])
644 for method_id in range(method_count):
647 for _ in range(param_count):
649 u30() # name index (always 0 for youtube)
651 if flags & 0x08 != 0:
654 for c in range(option_count):
657 if flags & 0x80 != 0:
658 # Param names present
659 for _ in range(param_count):
661 mi = MethodInfo(flags & 0x01 != 0, flags & 0x04 != 0)
662 method_infos.append(mi)
665 metadata_count = u30()
666 for _c in range(metadata_count):
669 for _c2 in range(item_count):
673 def parse_traits_info():
674 trait_name_idx = u30()
675 kind_full = read_byte()
676 kind = kind_full & 0x0f
677 attrs = kind_full >> 4
679 if kind in [0x00, 0x06]: # Slot or Const
681 u30() # type_name_idx
685 elif kind in [0x01, 0x02, 0x03]: # Method / Getter / Setter
688 methods[multinames[trait_name_idx]] = method_idx
689 elif kind == 0x04: # Class
692 elif kind == 0x05: # Function
695 methods[function_idx] = multinames[trait_name_idx]
697 raise ExtractorError(u'Unsupported trait kind %d' % kind)
699 if attrs & 0x4 != 0: # Metadata present
700 metadata_count = u30()
701 for _c3 in range(metadata_count):
702 u30() # metadata index
707 TARGET_CLASSNAME = u'SignatureDecipher'
708 searched_idx = multinames.index(TARGET_CLASSNAME)
709 searched_class_id = None
711 for class_id in range(class_count):
713 if name_idx == searched_idx:
714 # We found the class we're looking for!
715 searched_class_id = class_id
716 u30() # super_name idx
718 if flags & 0x08 != 0: # Protected namespace is present
719 u30() # protected_ns_idx
721 for _c2 in range(intrf_count):
725 for _c2 in range(trait_count):
728 if searched_class_id is None:
729 raise ExtractorError(u'Target class %r not found' %
734 for class_id in range(class_count):
737 for _c2 in range(trait_count):
738 trait_methods = parse_traits_info()
739 if class_id == searched_class_id:
740 method_names.update(trait_methods.items())
741 method_idxs.update(dict(
743 for name, idx in trait_methods.items()))
747 for _c in range(script_count):
750 for _c2 in range(trait_count):
754 method_body_count = u30()
755 Method = collections.namedtuple('Method', ['code', 'local_count'])
757 for _c in range(method_body_count):
761 u30() # init_scope_depth
762 u30() # max_scope_depth
764 code = read_bytes(code_length)
765 if method_idx in method_idxs:
766 m = Method(code, local_count)
767 methods[method_idxs[method_idx]] = m
768 exception_count = u30()
769 for _c2 in range(exception_count):
776 for _c2 in range(trait_count):
779 assert p + code_reader.tell() == len(code_tag)
780 assert len(methods) == len(method_idxs)
782 method_pyfunctions = {}
784 def extract_function(func_name):
785 if func_name in method_pyfunctions:
786 return method_pyfunctions[func_name]
787 if func_name not in methods:
788 raise ExtractorError(u'Cannot find function %r' % func_name)
789 m = methods[func_name]
792 registers = ['(this)'] + list(args) + [None] * m.local_count
794 coder = io.BytesIO(m.code)
796 opcode = struct.unpack('!B', coder.read(1))[0]
797 if opcode == 36: # pushbyte
798 v = struct.unpack('!B', coder.read(1))[0]
800 elif opcode == 44: # pushstring
802 stack.append(constant_strings[idx])
803 elif opcode == 48: # pushscope
804 # We don't implement the scope register, so we'll just
805 # ignore the popped value
807 elif opcode == 70: # callproperty
809 mname = multinames[index]
810 arg_count = u30(coder)
811 args = list(reversed(
812 [stack.pop() for _ in range(arg_count)]))
814 if mname == u'split':
815 assert len(args) == 1
816 assert isinstance(args[0], compat_str)
817 assert isinstance(obj, compat_str)
821 res = obj.split(args[0])
823 elif mname == u'slice':
824 assert len(args) == 1
825 assert isinstance(args[0], int)
826 assert isinstance(obj, list)
829 elif mname == u'join':
830 assert len(args) == 1
831 assert isinstance(args[0], compat_str)
832 assert isinstance(obj, list)
833 res = args[0].join(obj)
835 elif mname in method_pyfunctions:
836 stack.append(method_pyfunctions[mname](args))
838 raise NotImplementedError(
839 u'Unsupported property %r on %r'
841 elif opcode == 72: # returnvalue
844 elif opcode == 79: # callpropvoid
846 mname = multinames[index]
847 arg_count = u30(coder)
848 args = list(reversed(
849 [stack.pop() for _ in range(arg_count)]))
851 if mname == u'reverse':
852 assert isinstance(obj, list)
855 raise NotImplementedError(
856 u'Unsupported (void) property %r on %r'
858 elif opcode == 93: # findpropstrict
860 mname = multinames[index]
861 res = extract_function(mname)
863 elif opcode == 97: # setproperty
868 assert isinstance(obj, list)
869 assert isinstance(idx, int)
871 elif opcode == 98: # getlocal
873 stack.append(registers[index])
874 elif opcode == 99: # setlocal
877 registers[index] = value
878 elif opcode == 102: # getproperty
880 pname = multinames[index]
881 if pname == u'length':
883 assert isinstance(obj, list)
884 stack.append(len(obj))
885 else: # Assume attribute access
887 assert isinstance(idx, int)
889 assert isinstance(obj, list)
890 stack.append(obj[idx])
891 elif opcode == 128: # coerce
893 elif opcode == 133: # coerce_s
894 assert isinstance(stack[-1], (type(None), compat_str))
895 elif opcode == 164: # modulo
898 res = value1 % value2
900 elif opcode == 208: # getlocal_0
901 stack.append(registers[0])
902 elif opcode == 209: # getlocal_1
903 stack.append(registers[1])
904 elif opcode == 210: # getlocal_2
905 stack.append(registers[2])
906 elif opcode == 211: # getlocal_3
907 stack.append(registers[3])
908 elif opcode == 214: # setlocal_2
909 registers[2] = stack.pop()
910 elif opcode == 215: # setlocal_3
911 registers[3] = stack.pop()
913 raise NotImplementedError(
914 u'Unsupported opcode %d' % opcode)
916 method_pyfunctions[func_name] = resfunc
919 initial_function = extract_function(u'decipher')
920 return lambda s: initial_function([s])
922 def _decrypt_signature(self, s, video_id, player_url, age_gate=False):
923 """Turn the encrypted s field into a working signature"""
925 if player_url is not None:
926 if player_url.startswith(u'//'):
927 player_url = u'https:' + player_url
929 player_id = (player_url, len(s))
930 if player_id not in self._player_cache:
931 func = self._extract_signature_function(
932 video_id, player_url, len(s)
934 self._player_cache[player_id] = func
935 func = self._player_cache[player_id]
936 if self._downloader.params.get('youtube_print_sig_code'):
937 self._print_sig_code(func, len(s))
940 tb = traceback.format_exc()
941 self._downloader.report_warning(
942 u'Automatic signature extraction failed: ' + tb)
944 self._downloader.report_warning(
945 u'Warning: Falling back to static signature algorithm')
947 return self._static_decrypt_signature(
948 s, video_id, player_url, age_gate)
950 def _static_decrypt_signature(self, s, video_id, player_url, age_gate):
952 # The videos with age protection use another player, so the
953 # algorithms can be different.
955 return s[2:63] + s[82] + s[64:82] + s[63]
958 return s[86:29:-1] + s[88] + s[28:5:-1]
960 return s[25] + s[3:25] + s[0] + s[26:42] + s[79] + s[43:79] + s[91] + s[80:83]
962 return s[84:27:-1] + s[86] + s[26:5:-1]
964 return s[25] + s[3:25] + s[2] + s[26:40] + s[77] + s[41:77] + s[89] + s[78:81]
966 return s[84:78:-1] + s[87] + s[77:60:-1] + s[0] + s[59:3:-1]
968 return s[7:28] + s[87] + s[29:45] + s[55] + s[46:55] + s[2] + s[56:87] + s[28]
970 return s[6:27] + s[4] + s[28:39] + s[27] + s[40:59] + s[2] + s[60:]
972 return s[80:72:-1] + s[16] + s[71:39:-1] + s[72] + s[38:16:-1] + s[82] + s[15::-1]
974 return s[3:11] + s[0] + s[12:55] + s[84] + s[56:84]
976 return s[78:70:-1] + s[14] + s[69:37:-1] + s[70] + s[36:14:-1] + s[80] + s[:14][::-1]
978 return s[80:63:-1] + s[0] + s[62:0:-1] + s[63]
980 return s[80:37:-1] + s[7] + s[36:7:-1] + s[0] + s[6:0:-1] + s[37]
982 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]
984 return s[1:19] + s[0] + s[20:68] + s[19] + s[69:80]
986 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]
989 raise ExtractorError(u'Unable to decrypt signature, key length %d not supported; retrying might work' % (len(s)))
991 def _get_available_subtitles(self, video_id, webpage):
993 sub_list = self._download_webpage(
994 'http://video.google.com/timedtext?hl=en&type=list&v=%s' % video_id,
995 video_id, note=False)
996 except ExtractorError as err:
997 self._downloader.report_warning(u'unable to download video subtitles: %s' % compat_str(err))
999 lang_list = re.findall(r'name="([^"]*)"[^>]+lang_code="([\w\-]+)"', sub_list)
1004 params = compat_urllib_parse.urlencode({
1007 'fmt': self._downloader.params.get('subtitlesformat', 'srt'),
1008 'name': unescapeHTML(l[0]).encode('utf-8'),
1010 url = u'http://www.youtube.com/api/timedtext?' + params
1011 sub_lang_list[lang] = url
1012 if not sub_lang_list:
1013 self._downloader.report_warning(u'video doesn\'t have subtitles')
1015 return sub_lang_list
1017 def _get_available_automatic_caption(self, video_id, webpage):
1018 """We need the webpage for getting the captions url, pass it as an
1019 argument to speed up the process."""
1020 sub_format = self._downloader.params.get('subtitlesformat', 'srt')
1021 self.to_screen(u'%s: Looking for automatic captions' % video_id)
1022 mobj = re.search(r';ytplayer.config = ({.*?});', webpage)
1023 err_msg = u'Couldn\'t find automatic captions for %s' % video_id
1025 self._downloader.report_warning(err_msg)
1027 player_config = json.loads(mobj.group(1))
1029 args = player_config[u'args']
1030 caption_url = args[u'ttsurl']
1031 timestamp = args[u'timestamp']
1032 # We get the available subtitles
1033 list_params = compat_urllib_parse.urlencode({
1038 list_url = caption_url + '&' + list_params
1039 caption_list = self._download_xml(list_url, video_id)
1040 original_lang_node = caption_list.find('track')
1041 if original_lang_node is None or original_lang_node.attrib.get('kind') != 'asr' :
1042 self._downloader.report_warning(u'Video doesn\'t have automatic captions')
1044 original_lang = original_lang_node.attrib['lang_code']
1047 for lang_node in caption_list.findall('target'):
1048 sub_lang = lang_node.attrib['lang_code']
1049 params = compat_urllib_parse.urlencode({
1050 'lang': original_lang,
1056 sub_lang_list[sub_lang] = caption_url + '&' + params
1057 return sub_lang_list
1058 # An extractor error can be raise by the download process if there are
1059 # no automatic captions but there are subtitles
1060 except (KeyError, ExtractorError):
1061 self._downloader.report_warning(err_msg)
1064 def _extract_id(self, url):
1065 mobj = re.match(self._VALID_URL, url, re.VERBOSE)
1067 raise ExtractorError(u'Invalid URL: %s' % url)
1068 video_id = mobj.group(2)
1071 def _get_video_url_list(self, url_map):
1073 Transform a dictionary in the format {itag:url} to a list of (itag, url)
1074 with the requested formats.
1076 existing_formats = [x for x in self._formats if x in url_map]
1077 if len(existing_formats) == 0:
1078 raise ExtractorError(u'no known formats available for video')
1079 video_url_list = [(f, url_map[f]) for f in existing_formats] # All formats
1080 video_url_list.reverse() # order worst to best
1081 return video_url_list
1083 def _extract_from_m3u8(self, manifest_url, video_id):
1085 def _get_urls(_manifest):
1086 lines = _manifest.split('\n')
1087 urls = filter(lambda l: l and not l.startswith('#'),
1090 manifest = self._download_webpage(manifest_url, video_id, u'Downloading formats manifest')
1091 formats_urls = _get_urls(manifest)
1092 for format_url in formats_urls:
1093 itag = self._search_regex(r'itag/(\d+?)/', format_url, 'itag')
1094 url_map[itag] = format_url
1097 def _extract_annotations(self, video_id):
1098 url = 'https://www.youtube.com/annotations_invideo?features=1&legacy=1&video_id=%s' % video_id
1099 return self._download_webpage(url, video_id, note=u'Searching for annotations.', errnote=u'Unable to download video annotations.')
1101 def _real_extract(self, url):
1102 # Extract original video URL from URL with redirection, like age verification, using next_url parameter
1103 mobj = re.search(self._NEXT_URL_RE, url)
1105 url = 'https://www.youtube.com/' + compat_urllib_parse.unquote(mobj.group(1)).lstrip('/')
1106 video_id = self._extract_id(url)
1109 url = 'https://www.youtube.com/watch?v=%s&gl=US&hl=en&has_verified=1' % video_id
1110 video_webpage = self._download_webpage(url, video_id)
1112 # Attempt to extract SWF player URL
1113 mobj = re.search(r'swfConfig.*?"(https?:\\/\\/.*?watch.*?-.*?\.swf)"', video_webpage)
1114 if mobj is not None:
1115 player_url = re.sub(r'\\(.)', r'\1', mobj.group(1))
1120 self.report_video_info_webpage_download(video_id)
1121 if re.search(r'player-age-gate-content">', video_webpage) is not None:
1122 self.report_age_confirmation()
1124 # We simulate the access to the video from www.youtube.com/v/{video_id}
1125 # this can be viewed without login into Youtube
1126 data = compat_urllib_parse.urlencode({'video_id': video_id,
1127 'el': 'player_embedded',
1130 'eurl': 'https://youtube.googleapis.com/v/' + video_id,
1134 video_info_url = 'https://www.youtube.com/get_video_info?' + data
1135 video_info_webpage = self._download_webpage(video_info_url, video_id,
1137 errnote='unable to download video info webpage')
1138 video_info = compat_parse_qs(video_info_webpage)
1141 for el_type in ['&el=embedded', '&el=detailpage', '&el=vevo', '']:
1142 video_info_url = ('https://www.youtube.com/get_video_info?&video_id=%s%s&ps=default&eurl=&gl=US&hl=en'
1143 % (video_id, el_type))
1144 video_info_webpage = self._download_webpage(video_info_url, video_id,
1146 errnote='unable to download video info webpage')
1147 video_info = compat_parse_qs(video_info_webpage)
1148 if 'token' in video_info:
1150 if 'token' not in video_info:
1151 if 'reason' in video_info:
1152 raise ExtractorError(u'YouTube said: %s' % video_info['reason'][0], expected=True)
1154 raise ExtractorError(u'"token" parameter not in video info for unknown reason')
1156 if 'view_count' in video_info:
1157 view_count = int(video_info['view_count'][0])
1161 # Check for "rental" videos
1162 if 'ypc_video_rental_bar_text' in video_info and 'author' not in video_info:
1163 raise ExtractorError(u'"rental" videos not supported')
1165 # Start extracting information
1166 self.report_information_extraction(video_id)
1169 if 'author' not in video_info:
1170 raise ExtractorError(u'Unable to extract uploader name')
1171 video_uploader = compat_urllib_parse.unquote_plus(video_info['author'][0])
1174 video_uploader_id = None
1175 mobj = re.search(r'<link itemprop="url" href="http://www.youtube.com/(?:user|channel)/([^"]+)">', video_webpage)
1176 if mobj is not None:
1177 video_uploader_id = mobj.group(1)
1179 self._downloader.report_warning(u'unable to extract uploader nickname')
1182 if 'title' in video_info:
1183 video_title = compat_urllib_parse.unquote_plus(video_info['title'][0])
1185 self._downloader.report_warning(u'Unable to extract video title')
1189 # We try first to get a high quality image:
1190 m_thumb = re.search(r'<span itemprop="thumbnail".*?href="(.*?)">',
1191 video_webpage, re.DOTALL)
1192 if m_thumb is not None:
1193 video_thumbnail = m_thumb.group(1)
1194 elif 'thumbnail_url' not in video_info:
1195 self._downloader.report_warning(u'unable to extract video thumbnail')
1196 video_thumbnail = None
1197 else: # don't panic if we can't find it
1198 video_thumbnail = compat_urllib_parse.unquote_plus(video_info['thumbnail_url'][0])
1202 mobj = re.search(r'id="eow-date.*?>(.*?)</span>', video_webpage, re.DOTALL)
1203 if mobj is not None:
1204 upload_date = ' '.join(re.sub(r'[/,-]', r' ', mobj.group(1)).split())
1205 upload_date = unified_strdate(upload_date)
1208 video_description = get_element_by_id("eow-description", video_webpage)
1209 if video_description:
1210 video_description = re.sub(r'''(?x)
1212 (?:[a-zA-Z-]+="[^"]+"\s+)*?
1214 (?:[a-zA-Z-]+="[^"]+"\s+)*?
1215 class="yt-uix-redirect-link"\s*>
1218 ''', r'\1', video_description)
1219 video_description = clean_html(video_description)
1221 fd_mobj = re.search(r'<meta name="description" content="([^"]+)"', video_webpage)
1223 video_description = unescapeHTML(fd_mobj.group(1))
1225 video_description = u''
1227 def _extract_count(klass):
1228 count = self._search_regex(
1229 r'class="%s">([\d,]+)</span>' % re.escape(klass),
1230 video_webpage, klass, default=None)
1231 if count is not None:
1232 return int(count.replace(',', ''))
1234 like_count = _extract_count(u'likes-count')
1235 dislike_count = _extract_count(u'dislikes-count')
1238 video_subtitles = self.extract_subtitles(video_id, video_webpage)
1240 if self._downloader.params.get('listsubtitles', False):
1241 self._list_available_subtitles(video_id, video_webpage)
1244 if 'length_seconds' not in video_info:
1245 self._downloader.report_warning(u'unable to extract video duration')
1246 video_duration = None
1248 video_duration = int(compat_urllib_parse.unquote_plus(video_info['length_seconds'][0]))
1251 video_annotations = None
1252 if self._downloader.params.get('writeannotations', False):
1253 video_annotations = self._extract_annotations(video_id)
1255 # Decide which formats to download
1258 mobj = re.search(r';ytplayer.config = ({.*?});', video_webpage)
1260 raise ValueError('Could not find vevo ID')
1261 info = json.loads(mobj.group(1))
1263 # Easy way to know if the 's' value is in url_encoded_fmt_stream_map
1264 # this signatures are encrypted
1265 if 'url_encoded_fmt_stream_map' not in args:
1266 raise ValueError(u'No stream_map present') # caught below
1267 re_signature = re.compile(r'[&,]s=')
1268 m_s = re_signature.search(args['url_encoded_fmt_stream_map'])
1270 self.to_screen(u'%s: Encrypted signatures detected.' % video_id)
1271 video_info['url_encoded_fmt_stream_map'] = [args['url_encoded_fmt_stream_map']]
1272 m_s = re_signature.search(args.get('adaptive_fmts', u''))
1274 if 'adaptive_fmts' in video_info:
1275 video_info['adaptive_fmts'][0] += ',' + args['adaptive_fmts']
1277 video_info['adaptive_fmts'] = [args['adaptive_fmts']]
1281 if 'conn' in video_info and video_info['conn'][0].startswith('rtmp'):
1282 self.report_rtmp_download()
1283 video_url_list = [('_rtmp', video_info['conn'][0])]
1284 elif len(video_info.get('url_encoded_fmt_stream_map', [])) >= 1 or len(video_info.get('adaptive_fmts', [])) >= 1:
1285 encoded_url_map = video_info.get('url_encoded_fmt_stream_map', [''])[0] + ',' + video_info.get('adaptive_fmts',[''])[0]
1286 if 'rtmpe%3Dyes' in encoded_url_map:
1287 raise ExtractorError('rtmpe downloads are not supported, see https://github.com/rg3/youtube-dl/issues/343 for more information.', expected=True)
1289 for url_data_str in encoded_url_map.split(','):
1290 url_data = compat_parse_qs(url_data_str)
1291 if 'itag' in url_data and 'url' in url_data:
1292 url = url_data['url'][0]
1293 if 'sig' in url_data:
1294 url += '&signature=' + url_data['sig'][0]
1295 elif 's' in url_data:
1296 encrypted_sig = url_data['s'][0]
1297 if self._downloader.params.get('verbose'):
1299 if player_url is None:
1300 player_version = 'unknown'
1302 player_version = self._search_regex(
1303 r'-(.+)\.swf$', player_url,
1304 u'flash player', fatal=False)
1305 player_desc = 'flash player %s' % player_version
1307 player_version = self._search_regex(
1308 r'html5player-(.+?)\.js', video_webpage,
1309 'html5 player', fatal=False)
1310 player_desc = u'html5 player %s' % player_version
1312 parts_sizes = u'.'.join(compat_str(len(part)) for part in encrypted_sig.split('.'))
1313 self.to_screen(u'encrypted signature length %d (%s), itag %s, %s' %
1314 (len(encrypted_sig), parts_sizes, url_data['itag'][0], player_desc))
1317 jsplayer_url_json = self._search_regex(
1318 r'"assets":.+?"js":\s*("[^"]+")',
1319 video_webpage, u'JS player URL')
1320 player_url = json.loads(jsplayer_url_json)
1322 signature = self._decrypt_signature(
1323 encrypted_sig, video_id, player_url, age_gate)
1324 url += '&signature=' + signature
1325 if 'ratebypass' not in url:
1326 url += '&ratebypass=yes'
1327 url_map[url_data['itag'][0]] = url
1328 video_url_list = self._get_video_url_list(url_map)
1329 elif video_info.get('hlsvp'):
1330 manifest_url = video_info['hlsvp'][0]
1331 url_map = self._extract_from_m3u8(manifest_url, video_id)
1332 video_url_list = self._get_video_url_list(url_map)
1334 raise ExtractorError(u'no conn, hlsvp or url_encoded_fmt_stream_map information found in video info')
1337 for itag, video_real_url in video_url_list:
1340 'url': video_real_url,
1341 'player_url': player_url,
1343 dct.update(self._formats[itag])
1346 self._sort_formats(formats)
1350 'uploader': video_uploader,
1351 'uploader_id': video_uploader_id,
1352 'upload_date': upload_date,
1353 'title': video_title,
1354 'thumbnail': video_thumbnail,
1355 'description': video_description,
1356 'subtitles': video_subtitles,
1357 'duration': video_duration,
1358 'age_limit': 18 if age_gate else 0,
1359 'annotations': video_annotations,
1360 'webpage_url': 'https://www.youtube.com/watch?v=%s' % video_id,
1361 'view_count': view_count,
1362 'like_count': like_count,
1363 'dislike_count': dislike_count,
1367 class YoutubePlaylistIE(YoutubeBaseInfoExtractor):
1368 IE_DESC = u'YouTube.com playlists'
1369 _VALID_URL = r"""(?:
1374 (?:course|view_play_list|my_playlists|artist|playlist|watch)
1375 \? (?:.*?&)*? (?:p|a|list)=
1378 ((?:PL|EC|UU|FL|RD)?[0-9A-Za-z-_]{10,})
1381 ((?:PL|EC|UU|FL|RD)[0-9A-Za-z-_]{10,})
1383 _TEMPLATE_URL = 'https://www.youtube.com/playlist?list=%s&page=%s'
1384 _MORE_PAGES_INDICATOR = r'data-link-type="next"'
1385 _VIDEO_RE = r'href="/watch\?v=(?P<id>[0-9A-Za-z_-]{11})&[^"]*?index=(?P<index>\d+)'
1386 IE_NAME = u'youtube:playlist'
1389 def suitable(cls, url):
1390 """Receives a URL and returns True if suitable for this IE."""
1391 return re.match(cls._VALID_URL, url, re.VERBOSE) is not None
1393 def _real_initialize(self):
1396 def _ids_to_results(self, ids):
1397 return [self.url_result(vid_id, 'Youtube', video_id=vid_id)
1400 def _extract_mix(self, playlist_id):
1401 # The mixes are generated from a a single video
1402 # the id of the playlist is just 'RD' + video_id
1403 url = 'https://youtube.com/watch?v=%s&list=%s' % (playlist_id[-11:], playlist_id)
1404 webpage = self._download_webpage(url, playlist_id, u'Downloading Youtube mix')
1405 title_span = (get_element_by_attribute('class', 'title long-title', webpage) or
1406 get_element_by_attribute('class', 'title ', webpage))
1407 title = clean_html(title_span)
1408 video_re = r'data-index="\d+".*?href="/watch\?v=([0-9A-Za-z_-]{11})&[^"]*?list=%s' % re.escape(playlist_id)
1409 ids = orderedSet(re.findall(video_re, webpage))
1410 url_results = self._ids_to_results(ids)
1412 return self.playlist_result(url_results, playlist_id, title)
1414 def _real_extract(self, url):
1415 # Extract playlist id
1416 mobj = re.match(self._VALID_URL, url, re.VERBOSE)
1418 raise ExtractorError(u'Invalid URL: %s' % url)
1419 playlist_id = mobj.group(1) or mobj.group(2)
1421 # Check if it's a video-specific URL
1422 query_dict = compat_urlparse.parse_qs(compat_urlparse.urlparse(url).query)
1423 if 'v' in query_dict:
1424 video_id = query_dict['v'][0]
1425 if self._downloader.params.get('noplaylist'):
1426 self.to_screen(u'Downloading just video %s because of --no-playlist' % video_id)
1427 return self.url_result(video_id, 'Youtube', video_id=video_id)
1429 self.to_screen(u'Downloading playlist PL%s - add --no-playlist to just download video %s' % (playlist_id, video_id))
1431 if playlist_id.startswith('RD'):
1432 # Mixes require a custom extraction process
1433 return self._extract_mix(playlist_id)
1434 if playlist_id.startswith('TL'):
1435 raise ExtractorError(u'For downloading YouTube.com top lists, use '
1436 u'the "yttoplist" keyword, for example "youtube-dl \'yttoplist:music:Top Tracks\'"', expected=True)
1438 # Extract the video ids from the playlist pages
1441 for page_num in itertools.count(1):
1442 url = self._TEMPLATE_URL % (playlist_id, page_num)
1443 page = self._download_webpage(url, playlist_id, u'Downloading page #%s' % page_num)
1444 matches = re.finditer(self._VIDEO_RE, page)
1445 # We remove the duplicates and the link with index 0
1446 # (it's not the first video of the playlist)
1447 new_ids = orderedSet(m.group('id') for m in matches if m.group('index') != '0')
1450 if re.search(self._MORE_PAGES_INDICATOR, page) is None:
1454 playlist_title = self._og_search_title(page)
1455 except RegexNotFoundError:
1456 self.report_warning(
1457 u'Playlist page is missing OpenGraph title, falling back ...',
1459 playlist_title = self._html_search_regex(
1460 r'<h1 class="pl-header-title">(.*?)</h1>', page, u'title')
1462 url_results = self._ids_to_results(ids)
1463 return self.playlist_result(url_results, playlist_id, playlist_title)
1466 class YoutubeTopListIE(YoutubePlaylistIE):
1467 IE_NAME = u'youtube:toplist'
1468 IE_DESC = (u'YouTube.com top lists, "yttoplist:{channel}:{list title}"'
1469 u' (Example: "yttoplist:music:Top Tracks")')
1470 _VALID_URL = r'yttoplist:(?P<chann>.*?):(?P<title>.*?)$'
1472 def _real_extract(self, url):
1473 mobj = re.match(self._VALID_URL, url)
1474 channel = mobj.group('chann')
1475 title = mobj.group('title')
1476 query = compat_urllib_parse.urlencode({'title': title})
1477 playlist_re = 'href="([^"]+?%s[^"]+?)"' % re.escape(query)
1478 channel_page = self._download_webpage('https://www.youtube.com/%s' % channel, title)
1479 link = self._html_search_regex(playlist_re, channel_page, u'list')
1480 url = compat_urlparse.urljoin('https://www.youtube.com/', link)
1482 video_re = r'data-index="\d+".*?data-video-id="([0-9A-Za-z_-]{11})"'
1484 # sometimes the webpage doesn't contain the videos
1485 # retry until we get them
1486 for i in itertools.count(0):
1487 msg = u'Downloading Youtube mix'
1489 msg += ', retry #%d' % i
1490 webpage = self._download_webpage(url, title, msg)
1491 ids = orderedSet(re.findall(video_re, webpage))
1494 url_results = self._ids_to_results(ids)
1495 return self.playlist_result(url_results, playlist_title=title)
1498 class YoutubeChannelIE(InfoExtractor):
1499 IE_DESC = u'YouTube.com channels'
1500 _VALID_URL = r"^(?:https?://)?(?:youtu\.be|(?:\w+\.)?youtube(?:-nocookie)?\.com)/channel/([0-9A-Za-z_-]+)"
1501 _MORE_PAGES_INDICATOR = 'yt-uix-load-more'
1502 _MORE_PAGES_URL = 'http://www.youtube.com/c4_browse_ajax?action_load_more_videos=1&flow=list&paging=%s&view=0&sort=da&channel_id=%s'
1503 IE_NAME = u'youtube:channel'
1505 def extract_videos_from_page(self, page):
1507 for mobj in re.finditer(r'href="/watch\?v=([0-9A-Za-z_-]+)&?', page):
1508 if mobj.group(1) not in ids_in_page:
1509 ids_in_page.append(mobj.group(1))
1512 def _real_extract(self, url):
1513 # Extract channel id
1514 mobj = re.match(self._VALID_URL, url)
1516 raise ExtractorError(u'Invalid URL: %s' % url)
1518 # Download channel page
1519 channel_id = mobj.group(1)
1521 url = 'https://www.youtube.com/channel/%s/videos' % channel_id
1522 channel_page = self._download_webpage(url, channel_id)
1523 autogenerated = re.search(r'''(?x)
1525 channel-header-autogenerated-label|
1526 yt-channel-title-autogenerated
1527 )[^"]*"''', channel_page) is not None
1530 # The videos are contained in a single page
1531 # the ajax pages can't be used, they are empty
1532 video_ids = self.extract_videos_from_page(channel_page)
1534 # Download all channel pages using the json-based channel_ajax query
1535 for pagenum in itertools.count(1):
1536 url = self._MORE_PAGES_URL % (pagenum, channel_id)
1537 page = self._download_webpage(url, channel_id,
1538 u'Downloading page #%s' % pagenum)
1540 page = json.loads(page)
1542 ids_in_page = self.extract_videos_from_page(page['content_html'])
1543 video_ids.extend(ids_in_page)
1545 if self._MORE_PAGES_INDICATOR not in page['load_more_widget_html']:
1548 self._downloader.to_screen(u'[youtube] Channel %s: Found %i videos' % (channel_id, len(video_ids)))
1550 url_entries = [self.url_result(video_id, 'Youtube', video_id=video_id)
1551 for video_id in video_ids]
1552 return self.playlist_result(url_entries, channel_id)
1555 class YoutubeUserIE(InfoExtractor):
1556 IE_DESC = u'YouTube.com user videos (URL or "ytuser" keyword)'
1557 _VALID_URL = r'(?:(?:(?:https?://)?(?:\w+\.)?youtube\.com/(?:user/)?(?!(?:attribution_link|watch)(?:$|[^a-z_A-Z0-9-])))|ytuser:)(?!feed/)([A-Za-z0-9_-]+)'
1558 _TEMPLATE_URL = 'http://gdata.youtube.com/feeds/api/users/%s'
1559 _GDATA_PAGE_SIZE = 50
1560 _GDATA_URL = 'http://gdata.youtube.com/feeds/api/users/%s/uploads?max-results=%d&start-index=%d&alt=json'
1561 IE_NAME = u'youtube:user'
1564 def suitable(cls, url):
1565 # Don't return True if the url can be extracted with other youtube
1566 # extractor, the regex would is too permissive and it would match.
1567 other_ies = iter(klass for (name, klass) in globals().items() if name.endswith('IE') and klass is not cls)
1568 if any(ie.suitable(url) for ie in other_ies): return False
1569 else: return super(YoutubeUserIE, cls).suitable(url)
1571 def _real_extract(self, url):
1573 mobj = re.match(self._VALID_URL, url)
1575 raise ExtractorError(u'Invalid URL: %s' % url)
1577 username = mobj.group(1)
1579 # Download video ids using YouTube Data API. Result size per
1580 # query is limited (currently to 50 videos) so we need to query
1581 # page by page until there are no video ids - it means we got
1584 def download_page(pagenum):
1585 start_index = pagenum * self._GDATA_PAGE_SIZE + 1
1587 gdata_url = self._GDATA_URL % (username, self._GDATA_PAGE_SIZE, start_index)
1588 page = self._download_webpage(
1589 gdata_url, username,
1590 u'Downloading video ids from %d to %d' % (
1591 start_index, start_index + self._GDATA_PAGE_SIZE))
1594 response = json.loads(page)
1595 except ValueError as err:
1596 raise ExtractorError(u'Invalid JSON in API response: ' + compat_str(err))
1597 if 'entry' not in response['feed']:
1600 # Extract video identifiers
1601 entries = response['feed']['entry']
1602 for entry in entries:
1603 title = entry['title']['$t']
1604 video_id = entry['id']['$t'].split('/')[-1]
1608 'ie_key': 'Youtube',
1612 url_results = PagedList(download_page, self._GDATA_PAGE_SIZE)
1614 return self.playlist_result(url_results, playlist_title=username)
1617 class YoutubeSearchIE(SearchInfoExtractor):
1618 IE_DESC = u'YouTube.com searches'
1619 _API_URL = 'https://gdata.youtube.com/feeds/api/videos?q=%s&start-index=%i&max-results=50&v=2&alt=jsonc'
1621 IE_NAME = u'youtube:search'
1622 _SEARCH_KEY = 'ytsearch'
1624 def _get_n_results(self, query, n):
1625 """Get a specified number of results for a query"""
1631 while (50 * pagenum) < limit:
1632 result_url = self._API_URL % (compat_urllib_parse.quote_plus(query), (50*pagenum)+1)
1633 data_json = self._download_webpage(
1634 result_url, video_id=u'query "%s"' % query,
1635 note=u'Downloading page %s' % (pagenum + 1),
1636 errnote=u'Unable to download API page')
1637 data = json.loads(data_json)
1638 api_response = data['data']
1640 if 'items' not in api_response:
1641 raise ExtractorError(u'[youtube] No video results')
1643 new_ids = list(video['id'] for video in api_response['items'])
1644 video_ids += new_ids
1646 limit = min(n, api_response['totalItems'])
1649 if len(video_ids) > n:
1650 video_ids = video_ids[:n]
1651 videos = [self.url_result(video_id, 'Youtube', video_id=video_id)
1652 for video_id in video_ids]
1653 return self.playlist_result(videos, query)
1655 class YoutubeSearchDateIE(YoutubeSearchIE):
1656 IE_NAME = YoutubeSearchIE.IE_NAME + ':date'
1657 _API_URL = 'https://gdata.youtube.com/feeds/api/videos?q=%s&start-index=%i&max-results=50&v=2&alt=jsonc&orderby=published'
1658 _SEARCH_KEY = 'ytsearchdate'
1659 IE_DESC = u'YouTube.com searches, newest videos first'
1661 class YoutubeShowIE(InfoExtractor):
1662 IE_DESC = u'YouTube.com (multi-season) shows'
1663 _VALID_URL = r'https?://www\.youtube\.com/show/(.*)'
1664 IE_NAME = u'youtube:show'
1666 def _real_extract(self, url):
1667 mobj = re.match(self._VALID_URL, url)
1668 show_name = mobj.group(1)
1669 webpage = self._download_webpage(url, show_name, u'Downloading show webpage')
1670 # There's one playlist for each season of the show
1671 m_seasons = list(re.finditer(r'href="(/playlist\?list=.*?)"', webpage))
1672 self.to_screen(u'%s: Found %s seasons' % (show_name, len(m_seasons)))
1673 return [self.url_result('https://www.youtube.com' + season.group(1), 'YoutubePlaylist') for season in m_seasons]
1676 class YoutubeFeedsInfoExtractor(YoutubeBaseInfoExtractor):
1678 Base class for extractors that fetch info from
1679 http://www.youtube.com/feed_ajax
1680 Subclasses must define the _FEED_NAME and _PLAYLIST_TITLE properties.
1682 _LOGIN_REQUIRED = True
1683 # use action_load_personal_feed instead of action_load_system_feed
1684 _PERSONAL_FEED = False
1687 def _FEED_TEMPLATE(self):
1688 action = 'action_load_system_feed'
1689 if self._PERSONAL_FEED:
1690 action = 'action_load_personal_feed'
1691 return 'http://www.youtube.com/feed_ajax?%s=1&feed_name=%s&paging=%%s' % (action, self._FEED_NAME)
1695 return u'youtube:%s' % self._FEED_NAME
1697 def _real_initialize(self):
1700 def _real_extract(self, url):
1703 for i in itertools.count(1):
1704 info = self._download_webpage(self._FEED_TEMPLATE % paging,
1705 u'%s feed' % self._FEED_NAME,
1706 u'Downloading page %s' % i)
1707 info = json.loads(info)
1708 feed_html = info['feed_html']
1709 m_ids = re.finditer(r'"/watch\?v=(.*?)["&]', feed_html)
1710 ids = orderedSet(m.group(1) for m in m_ids)
1711 feed_entries.extend(
1712 self.url_result(video_id, 'Youtube', video_id=video_id)
1713 for video_id in ids)
1714 if info['paging'] is None:
1716 paging = info['paging']
1717 return self.playlist_result(feed_entries, playlist_title=self._PLAYLIST_TITLE)
1719 class YoutubeSubscriptionsIE(YoutubeFeedsInfoExtractor):
1720 IE_DESC = u'YouTube.com subscriptions feed, "ytsubs" keyword(requires authentication)'
1721 _VALID_URL = r'https?://www\.youtube\.com/feed/subscriptions|:ytsubs(?:criptions)?'
1722 _FEED_NAME = 'subscriptions'
1723 _PLAYLIST_TITLE = u'Youtube Subscriptions'
1725 class YoutubeRecommendedIE(YoutubeFeedsInfoExtractor):
1726 IE_DESC = u'YouTube.com recommended videos, "ytrec" keyword (requires authentication)'
1727 _VALID_URL = r'https?://www\.youtube\.com/feed/recommended|:ytrec(?:ommended)?'
1728 _FEED_NAME = 'recommended'
1729 _PLAYLIST_TITLE = u'Youtube Recommended videos'
1731 class YoutubeWatchLaterIE(YoutubeFeedsInfoExtractor):
1732 IE_DESC = u'Youtube watch later list, "ytwatchlater" keyword (requires authentication)'
1733 _VALID_URL = r'https?://www\.youtube\.com/feed/watch_later|:ytwatchlater'
1734 _FEED_NAME = 'watch_later'
1735 _PLAYLIST_TITLE = u'Youtube Watch Later'
1736 _PERSONAL_FEED = True
1738 class YoutubeHistoryIE(YoutubeFeedsInfoExtractor):
1739 IE_DESC = u'Youtube watch history, "ythistory" keyword (requires authentication)'
1740 _VALID_URL = u'https?://www\.youtube\.com/feed/history|:ythistory'
1741 _FEED_NAME = 'history'
1742 _PERSONAL_FEED = True
1743 _PLAYLIST_TITLE = u'Youtube Watch History'
1745 class YoutubeFavouritesIE(YoutubeBaseInfoExtractor):
1746 IE_NAME = u'youtube:favorites'
1747 IE_DESC = u'YouTube.com favourite videos, "ytfav" keyword (requires authentication)'
1748 _VALID_URL = r'https?://www\.youtube\.com/my_favorites|:ytfav(?:ou?rites)?'
1749 _LOGIN_REQUIRED = True
1751 def _real_extract(self, url):
1752 webpage = self._download_webpage('https://www.youtube.com/my_favorites', 'Youtube Favourites videos')
1753 playlist_id = self._search_regex(r'list=(.+?)["&]', webpage, u'favourites playlist id')
1754 return self.url_result(playlist_id, 'YoutubePlaylist')
1757 class YoutubeTruncatedURLIE(InfoExtractor):
1758 IE_NAME = 'youtube:truncated_url'
1759 IE_DESC = False # Do not list
1760 _VALID_URL = r'(?:https?://)?[^/]+/watch\?feature=[a-z_]+$'
1762 def _real_extract(self, url):
1763 raise ExtractorError(
1764 u'Did you forget to quote the URL? Remember that & is a meta '
1765 u'character in most shells, so you want to put the URL in quotes, '
1767 u'"http://www.youtube.com/watch?feature=foo&v=BaW_jenozKc" '
1768 u' or simply youtube-dl BaW_jenozKc .',