15 from .common import InfoExtractor, SearchInfoExtractor
16 from .subtitles import SubtitlesInfoExtractor
21 compat_urllib_request,
28 get_element_by_attribute,
37 class YoutubeBaseInfoExtractor(InfoExtractor):
38 """Provide base functions for Youtube extractors"""
39 _LOGIN_URL = 'https://accounts.google.com/ServiceLogin'
40 _LANG_URL = r'https://www.youtube.com/?hl=en&persist_hl=1&gl=US&persist_gl=1&opt_out_ackd=1'
41 _AGE_URL = 'http://www.youtube.com/verify_age?next_url=/&gl=US&hl=en'
42 _NETRC_MACHINE = 'youtube'
43 # If True it will raise an error if no login info is provided
44 _LOGIN_REQUIRED = False
46 def _set_language(self):
47 return bool(self._download_webpage(
49 note=u'Setting language', errnote='unable to set language',
53 (username, password) = self._get_login_info()
54 # No authentication to be performed
56 if self._LOGIN_REQUIRED:
57 raise ExtractorError(u'No login info available, needed for using %s.' % self.IE_NAME, expected=True)
60 login_page = self._download_webpage(
61 self._LOGIN_URL, None,
62 note=u'Downloading login page',
63 errnote=u'unable to fetch login page', fatal=False)
64 if login_page is False:
67 galx = self._search_regex(r'(?s)<input.+?name="GALX".+?value="(.+?)"',
68 login_page, u'Login GALX parameter')
72 u'continue': u'https://www.youtube.com/signin?action_handle_signin=true&feature=sign_in_button&hl=en_US&nomobiletemp=1',
76 u'PersistentCookie': u'yes',
78 u'bgresponse': u'js_disabled',
79 u'checkConnection': u'',
80 u'checkedDomains': u'youtube',
85 u'signIn': u'Sign in',
87 u'service': u'youtube',
91 # Convert to UTF-8 *before* urlencode because Python 2.x's urlencode
93 login_form = dict((k.encode('utf-8'), v.encode('utf-8')) for k,v in login_form_strs.items())
94 login_data = compat_urllib_parse.urlencode(login_form).encode('ascii')
96 req = compat_urllib_request.Request(self._LOGIN_URL, login_data)
97 login_results = self._download_webpage(
99 note=u'Logging in', errnote=u'unable to log in', fatal=False)
100 if login_results is False:
102 if re.search(r'(?i)<form[^>]* id="gaia_loginform"', login_results) is not None:
103 self._downloader.report_warning(u'unable to log in: bad username or password')
107 def _confirm_age(self):
110 'action_confirm': 'Confirm',
112 req = compat_urllib_request.Request(self._AGE_URL, compat_urllib_parse.urlencode(age_form))
114 self._download_webpage(
116 note=u'Confirming age', errnote=u'Unable to confirm age')
119 def _real_initialize(self):
120 if self._downloader is None:
122 if not self._set_language():
124 if not self._login():
129 class YoutubeIE(YoutubeBaseInfoExtractor, SubtitlesInfoExtractor):
130 IE_DESC = u'YouTube.com'
131 _VALID_URL = r"""(?x)^
133 (?:https?://|//)? # http(s):// or protocol-independent URL (optional)
134 (?:(?:(?:(?:\w+\.)?[yY][oO][uU][tT][uU][bB][eE](?:-nocookie)?\.com/|
135 (?:www\.)?deturl\.com/www\.youtube\.com/|
136 (?:www\.)?pwnyoutube\.com|
137 tube\.majestyc\.net/|
138 youtube\.googleapis\.com/) # the various hostnames, with wildcard subdomains
139 (?:.*?\#/)? # handle anchor (#/) redirect urls
140 (?: # the various things that can precede the ID:
141 (?:(?:v|embed|e)/) # v/ or embed/ or e/
142 |(?: # or the v= param in all its forms
143 (?:(?:watch|movie)(?:_popup)?(?:\.php)?)? # preceding watch(_popup|.php) or nothing (like /?v=xxxx)
144 (?:\?|\#!?) # the params delimiter ? or # or #!
145 (?:.*?&)? # any other preceding param (like /?s=tuff&v=xxxx)
149 |youtu\.be/ # just youtu.be/xxxx
151 )? # all until now is optional -> you can pass the naked ID
152 ([0-9A-Za-z_-]{11}) # here is it! the YouTube video ID
153 (?(1).+)? # if we found the ID, everything can follow
155 _NEXT_URL_RE = r'[\?&]next_url=([^&]+)'
157 '5': {'ext': 'flv', 'width': 400, 'height': 240},
158 '6': {'ext': 'flv', 'width': 450, 'height': 270},
159 '13': {'ext': '3gp'},
160 '17': {'ext': '3gp', 'width': 176, 'height': 144},
161 '18': {'ext': 'mp4', 'width': 640, 'height': 360},
162 '22': {'ext': 'mp4', 'width': 1280, 'height': 720},
163 '34': {'ext': 'flv', 'width': 640, 'height': 360},
164 '35': {'ext': 'flv', 'width': 854, 'height': 480},
165 '36': {'ext': '3gp', 'width': 320, 'height': 240},
166 '37': {'ext': 'mp4', 'width': 1920, 'height': 1080},
167 '38': {'ext': 'mp4', 'width': 4096, 'height': 3072},
168 '43': {'ext': 'webm', 'width': 640, 'height': 360},
169 '44': {'ext': 'webm', 'width': 854, 'height': 480},
170 '45': {'ext': 'webm', 'width': 1280, 'height': 720},
171 '46': {'ext': 'webm', 'width': 1920, 'height': 1080},
175 '82': {'ext': 'mp4', 'height': 360, 'resolution': '360p', 'format_note': '3D', 'preference': -20},
176 '83': {'ext': 'mp4', 'height': 480, 'resolution': '480p', 'format_note': '3D', 'preference': -20},
177 '84': {'ext': 'mp4', 'height': 720, 'resolution': '720p', 'format_note': '3D', 'preference': -20},
178 '85': {'ext': 'mp4', 'height': 1080, 'resolution': '1080p', 'format_note': '3D', 'preference': -20},
179 '100': {'ext': 'webm', 'height': 360, 'resolution': '360p', 'format_note': '3D', 'preference': -20},
180 '101': {'ext': 'webm', 'height': 480, 'resolution': '480p', 'format_note': '3D', 'preference': -20},
181 '102': {'ext': 'webm', 'height': 720, 'resolution': '720p', 'format_note': '3D', 'preference': -20},
183 # Apple HTTP Live Streaming
184 '92': {'ext': 'mp4', 'height': 240, 'resolution': '240p', 'format_note': 'HLS', 'preference': -10},
185 '93': {'ext': 'mp4', 'height': 360, 'resolution': '360p', 'format_note': 'HLS', 'preference': -10},
186 '94': {'ext': 'mp4', 'height': 480, 'resolution': '480p', 'format_note': 'HLS', 'preference': -10},
187 '95': {'ext': 'mp4', 'height': 720, 'resolution': '720p', 'format_note': 'HLS', 'preference': -10},
188 '96': {'ext': 'mp4', 'height': 1080, 'resolution': '1080p', 'format_note': 'HLS', 'preference': -10},
189 '132': {'ext': 'mp4', 'height': 240, 'resolution': '240p', 'format_note': 'HLS', 'preference': -10},
190 '151': {'ext': 'mp4', 'height': 72, 'resolution': '72p', 'format_note': 'HLS', 'preference': -10},
193 '133': {'ext': 'mp4', 'height': 240, 'resolution': '240p', 'format_note': 'DASH video', 'preference': -40},
194 '134': {'ext': 'mp4', 'height': 360, 'resolution': '360p', 'format_note': 'DASH video', 'preference': -40},
195 '135': {'ext': 'mp4', 'height': 480, 'resolution': '480p', 'format_note': 'DASH video', 'preference': -40},
196 '136': {'ext': 'mp4', 'height': 720, 'resolution': '720p', 'format_note': 'DASH video', 'preference': -40},
197 '137': {'ext': 'mp4', 'height': 1080, 'resolution': '1080p', 'format_note': 'DASH video', 'preference': -40},
198 '138': {'ext': 'mp4', 'height': 1081, 'resolution': '>1080p', 'format_note': 'DASH video', 'preference': -40},
199 '160': {'ext': 'mp4', 'height': 192, 'resolution': '192p', 'format_note': 'DASH video', 'preference': -40},
200 '264': {'ext': 'mp4', 'height': 1080, 'resolution': '1080p', 'format_note': 'DASH video', 'preference': -40},
203 '139': {'ext': 'm4a', 'format_note': 'DASH audio', 'vcodec': 'none', 'abr': 48, 'preference': -50},
204 '140': {'ext': 'm4a', 'format_note': 'DASH audio', 'vcodec': 'none', 'abr': 128, 'preference': -50},
205 '141': {'ext': 'm4a', 'format_note': 'DASH audio', 'vcodec': 'none', 'abr': 256, 'preference': -50},
208 '242': {'ext': 'webm', 'height': 240, 'resolution': '240p', 'format_note': 'DASH webm', 'preference': -40},
209 '243': {'ext': 'webm', 'height': 360, 'resolution': '360p', 'format_note': 'DASH webm', 'preference': -40},
210 '244': {'ext': 'webm', 'height': 480, 'resolution': '480p', 'format_note': 'DASH webm', 'preference': -40},
211 '245': {'ext': 'webm', 'height': 480, 'resolution': '480p', 'format_note': 'DASH webm', 'preference': -40},
212 '246': {'ext': 'webm', 'height': 480, 'resolution': '480p', 'format_note': 'DASH webm', 'preference': -40},
213 '247': {'ext': 'webm', 'height': 720, 'resolution': '720p', 'format_note': 'DASH webm', 'preference': -40},
214 '248': {'ext': 'webm', 'height': 1080, 'resolution': '1080p', 'format_note': 'DASH webm', 'preference': -40},
217 '171': {'ext': 'webm', 'vcodec': 'none', 'format_note': 'DASH webm audio', 'abr': 48, 'preference': -50},
218 '172': {'ext': 'webm', 'vcodec': 'none', 'format_note': 'DASH webm audio', 'abr': 256, 'preference': -50},
221 '_rtmp': {'protocol': 'rtmp'},
227 u"url": u"http://www.youtube.com/watch?v=BaW_jenozKc",
228 u"file": u"BaW_jenozKc.mp4",
230 u"title": u"youtube-dl test video \"'/\\ä↭𝕐",
231 u"uploader": u"Philipp Hagemeister",
232 u"uploader_id": u"phihag",
233 u"upload_date": u"20121002",
234 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 ."
238 u"url": u"http://www.youtube.com/watch?v=UxxajLWwzqY",
239 u"file": u"UxxajLWwzqY.mp4",
240 u"note": u"Test generic use_cipher_signature video (#897)",
242 u"upload_date": u"20120506",
243 u"title": u"Icona Pop - I Love It (feat. Charli XCX) [OFFICIAL VIDEO]",
244 u"description": u"md5:5b292926389560516e384ac437c0ec07",
245 u"uploader": u"Icona Pop",
246 u"uploader_id": u"IconaPop"
250 u"url": u"https://www.youtube.com/watch?v=07FYdnEawAQ",
251 u"file": u"07FYdnEawAQ.mp4",
252 u"note": u"Test VEVO video with age protection (#956)",
254 u"upload_date": u"20130703",
255 u"title": u"Justin Timberlake - Tunnel Vision (Explicit)",
256 u"description": u"md5:64249768eec3bc4276236606ea996373",
257 u"uploader": u"justintimberlakeVEVO",
258 u"uploader_id": u"justintimberlakeVEVO"
262 u"url": u"//www.YouTube.com/watch?v=yZIXLfi8CZQ",
263 u"file": u"yZIXLfi8CZQ.mp4",
264 u"note": u"Embed-only video (#1746)",
266 u"upload_date": u"20120608",
267 u"title": u"Principal Sexually Assaults A Teacher - Episode 117 - 8th June 2012",
268 u"description": u"md5:09b78bd971f1e3e289601dfba15ca4f7",
269 u"uploader": u"SET India",
270 u"uploader_id": u"setindia"
277 def suitable(cls, url):
278 """Receives a URL and returns True if suitable for this IE."""
279 if YoutubePlaylistIE.suitable(url): return False
280 return re.match(cls._VALID_URL, url) is not None
282 def __init__(self, *args, **kwargs):
283 super(YoutubeIE, self).__init__(*args, **kwargs)
284 self._player_cache = {}
286 def report_video_info_webpage_download(self, video_id):
287 """Report attempt to download video info webpage."""
288 self.to_screen(u'%s: Downloading video info webpage' % video_id)
290 def report_information_extraction(self, video_id):
291 """Report attempt to extract video information."""
292 self.to_screen(u'%s: Extracting video information' % video_id)
294 def report_unavailable_format(self, video_id, format):
295 """Report extracted video URL."""
296 self.to_screen(u'%s: Format %s not available' % (video_id, format))
298 def report_rtmp_download(self):
299 """Indicate the download will use the RTMP protocol."""
300 self.to_screen(u'RTMP download detected')
302 def _extract_signature_function(self, video_id, player_url, slen):
303 id_m = re.match(r'.*-(?P<id>[a-zA-Z0-9_-]+)\.(?P<ext>[a-z]+)$',
305 player_type = id_m.group('ext')
306 player_id = id_m.group('id')
308 # Read from filesystem cache
309 func_id = '%s_%s_%d' % (player_type, player_id, slen)
310 assert os.path.basename(func_id) == func_id
311 cache_dir = get_cachedir(self._downloader.params)
313 cache_enabled = cache_dir is not None
315 cache_fn = os.path.join(os.path.expanduser(cache_dir),
319 with io.open(cache_fn, 'r', encoding='utf-8') as cachef:
320 cache_spec = json.load(cachef)
321 return lambda s: u''.join(s[i] for i in cache_spec)
323 pass # No cache available
325 if player_type == 'js':
326 code = self._download_webpage(
327 player_url, video_id,
328 note=u'Downloading %s player %s' % (player_type, player_id),
329 errnote=u'Download of %s failed' % player_url)
330 res = self._parse_sig_js(code)
331 elif player_type == 'swf':
332 urlh = self._request_webpage(
333 player_url, video_id,
334 note=u'Downloading %s player %s' % (player_type, player_id),
335 errnote=u'Download of %s failed' % player_url)
337 res = self._parse_sig_swf(code)
339 assert False, 'Invalid player type %r' % player_type
343 test_string = u''.join(map(compat_chr, range(slen)))
344 cache_res = res(test_string)
345 cache_spec = [ord(c) for c in cache_res]
347 os.makedirs(os.path.dirname(cache_fn))
348 except OSError as ose:
349 if ose.errno != errno.EEXIST:
351 write_json_file(cache_spec, cache_fn)
353 tb = traceback.format_exc()
354 self._downloader.report_warning(
355 u'Writing cache to %r failed: %s' % (cache_fn, tb))
359 def _print_sig_code(self, func, slen):
360 def gen_sig_code(idxs):
361 def _genslice(start, end, step):
362 starts = u'' if start == 0 else str(start)
363 ends = (u':%d' % (end+step)) if end + step >= 0 else u':'
364 steps = u'' if step == 1 else (u':%d' % step)
365 return u's[%s%s%s]' % (starts, ends, steps)
368 start = '(Never used)' # Quelch pyflakes warnings - start will be
369 # set as soon as step is set
370 for i, prev in zip(idxs[1:], idxs[:-1]):
374 yield _genslice(start, prev, step)
377 if i - prev in [-1, 1]:
382 yield u's[%d]' % prev
386 yield _genslice(start, i, step)
388 test_string = u''.join(map(compat_chr, range(slen)))
389 cache_res = func(test_string)
390 cache_spec = [ord(c) for c in cache_res]
391 expr_code = u' + '.join(gen_sig_code(cache_spec))
392 code = u'if len(s) == %d:\n return %s\n' % (slen, expr_code)
393 self.to_screen(u'Extracted signature function:\n' + code)
395 def _parse_sig_js(self, jscode):
396 funcname = self._search_regex(
397 r'signature=([a-zA-Z]+)', jscode,
398 u'Initial JS player signature function name')
403 return string.lowercase.index(varname)
405 def interpret_statement(stmt, local_vars, allow_recursion=20):
406 if allow_recursion < 0:
407 raise ExtractorError(u'Recursion limit reached')
409 if stmt.startswith(u'var '):
410 stmt = stmt[len(u'var '):]
411 ass_m = re.match(r'^(?P<out>[a-z]+)(?:\[(?P<index>[^\]]+)\])?' +
412 r'=(?P<expr>.*)$', stmt)
414 if ass_m.groupdict().get('index'):
416 lvar = local_vars[ass_m.group('out')]
417 idx = interpret_expression(ass_m.group('index'),
418 local_vars, allow_recursion)
419 assert isinstance(idx, int)
422 expr = ass_m.group('expr')
425 local_vars[ass_m.group('out')] = val
427 expr = ass_m.group('expr')
428 elif stmt.startswith(u'return '):
430 expr = stmt[len(u'return '):]
432 raise ExtractorError(
433 u'Cannot determine left side of statement in %r' % stmt)
435 v = interpret_expression(expr, local_vars, allow_recursion)
438 def interpret_expression(expr, local_vars, allow_recursion):
443 return local_vars[expr]
445 m = re.match(r'^(?P<in>[a-z]+)\.(?P<member>.*)$', expr)
447 member = m.group('member')
448 val = local_vars[m.group('in')]
449 if member == 'split("")':
451 if member == 'join("")':
453 if member == 'length':
455 if member == 'reverse()':
457 slice_m = re.match(r'slice\((?P<idx>.*)\)', member)
459 idx = interpret_expression(
460 slice_m.group('idx'), local_vars, allow_recursion-1)
464 r'^(?P<in>[a-z]+)\[(?P<idx>.+)\]$', expr)
466 val = local_vars[m.group('in')]
467 idx = interpret_expression(m.group('idx'), local_vars,
471 m = re.match(r'^(?P<a>.+?)(?P<op>[%])(?P<b>.+?)$', expr)
473 a = interpret_expression(m.group('a'),
474 local_vars, allow_recursion)
475 b = interpret_expression(m.group('b'),
476 local_vars, allow_recursion)
480 r'^(?P<func>[a-zA-Z]+)\((?P<args>[a-z0-9,]+)\)$', expr)
482 fname = m.group('func')
483 if fname not in functions:
484 functions[fname] = extract_function(fname)
485 argvals = [int(v) if v.isdigit() else local_vars[v]
486 for v in m.group('args').split(',')]
487 return functions[fname](argvals)
488 raise ExtractorError(u'Unsupported JS expression %r' % expr)
490 def extract_function(funcname):
492 r'function ' + re.escape(funcname) +
493 r'\((?P<args>[a-z,]+)\){(?P<code>[^}]+)}',
495 argnames = func_m.group('args').split(',')
498 local_vars = dict(zip(argnames, args))
499 for stmt in func_m.group('code').split(';'):
500 res = interpret_statement(stmt, local_vars)
504 initial_function = extract_function(funcname)
505 return lambda s: initial_function([s])
507 def _parse_sig_swf(self, file_contents):
508 if file_contents[1:3] != b'WS':
509 raise ExtractorError(
510 u'Not an SWF file; header is %r' % file_contents[:3])
511 if file_contents[:1] == b'C':
512 content = zlib.decompress(file_contents[8:])
514 raise NotImplementedError(u'Unsupported compression format %r' %
517 def extract_tags(content):
519 while pos < len(content):
520 header16 = struct.unpack('<H', content[pos:pos+2])[0]
522 tag_code = header16 >> 6
523 tag_len = header16 & 0x3f
525 tag_len = struct.unpack('<I', content[pos:pos+4])[0]
527 assert pos+tag_len <= len(content)
528 yield (tag_code, content[pos:pos+tag_len])
532 for tag_code, tag in extract_tags(content)
534 p = code_tag.index(b'\0', 4) + 1
535 code_reader = io.BytesIO(code_tag[p:])
537 # Parse ABC (AVM2 ByteCode)
538 def read_int(reader=None):
546 b = struct.unpack('<B', buf)[0]
547 res = res | ((b & 0x7f) << shift)
553 def u30(reader=None):
554 res = read_int(reader)
555 assert res & 0xf0000000 == 0
559 def s32(reader=None):
561 if v & 0x80000000 != 0:
562 v = - ((v ^ 0xffffffff) + 1)
565 def read_string(reader=None):
569 resb = reader.read(slen)
570 assert len(resb) == slen
571 return resb.decode('utf-8')
573 def read_bytes(count, reader=None):
576 resb = reader.read(count)
577 assert len(resb) == count
580 def read_byte(reader=None):
581 resb = read_bytes(1, reader=reader)
582 res = struct.unpack('<B', resb)[0]
585 # minor_version + major_version
590 for _c in range(1, int_count):
593 for _c in range(1, uint_count):
596 read_bytes((double_count-1) * 8)
598 constant_strings = [u'']
599 for _c in range(1, string_count):
601 constant_strings.append(s)
602 namespace_count = u30()
603 for _c in range(1, namespace_count):
607 for _c in range(1, ns_set_count):
609 for _c2 in range(count):
611 multiname_count = u30()
620 0x0e: 2, # MultinameA
621 0x1b: 1, # MultinameL
622 0x1c: 1, # MultinameLA
625 for _c in range(1, multiname_count):
627 assert kind in MULTINAME_SIZES, u'Invalid multiname kind %r' % kind
629 u30() # namespace_idx
631 multinames.append(constant_strings[name_idx])
633 multinames.append('[MULTINAME kind: %d]' % kind)
634 for _c2 in range(MULTINAME_SIZES[kind]):
639 MethodInfo = collections.namedtuple(
641 ['NEED_ARGUMENTS', 'NEED_REST'])
643 for method_id in range(method_count):
646 for _ in range(param_count):
648 u30() # name index (always 0 for youtube)
650 if flags & 0x08 != 0:
653 for c in range(option_count):
656 if flags & 0x80 != 0:
657 # Param names present
658 for _ in range(param_count):
660 mi = MethodInfo(flags & 0x01 != 0, flags & 0x04 != 0)
661 method_infos.append(mi)
664 metadata_count = u30()
665 for _c in range(metadata_count):
668 for _c2 in range(item_count):
672 def parse_traits_info():
673 trait_name_idx = u30()
674 kind_full = read_byte()
675 kind = kind_full & 0x0f
676 attrs = kind_full >> 4
678 if kind in [0x00, 0x06]: # Slot or Const
680 u30() # type_name_idx
684 elif kind in [0x01, 0x02, 0x03]: # Method / Getter / Setter
687 methods[multinames[trait_name_idx]] = method_idx
688 elif kind == 0x04: # Class
691 elif kind == 0x05: # Function
694 methods[function_idx] = multinames[trait_name_idx]
696 raise ExtractorError(u'Unsupported trait kind %d' % kind)
698 if attrs & 0x4 != 0: # Metadata present
699 metadata_count = u30()
700 for _c3 in range(metadata_count):
701 u30() # metadata index
706 TARGET_CLASSNAME = u'SignatureDecipher'
707 searched_idx = multinames.index(TARGET_CLASSNAME)
708 searched_class_id = None
710 for class_id in range(class_count):
712 if name_idx == searched_idx:
713 # We found the class we're looking for!
714 searched_class_id = class_id
715 u30() # super_name idx
717 if flags & 0x08 != 0: # Protected namespace is present
718 u30() # protected_ns_idx
720 for _c2 in range(intrf_count):
724 for _c2 in range(trait_count):
727 if searched_class_id is None:
728 raise ExtractorError(u'Target class %r not found' %
733 for class_id in range(class_count):
736 for _c2 in range(trait_count):
737 trait_methods = parse_traits_info()
738 if class_id == searched_class_id:
739 method_names.update(trait_methods.items())
740 method_idxs.update(dict(
742 for name, idx in trait_methods.items()))
746 for _c in range(script_count):
749 for _c2 in range(trait_count):
753 method_body_count = u30()
754 Method = collections.namedtuple('Method', ['code', 'local_count'])
756 for _c in range(method_body_count):
760 u30() # init_scope_depth
761 u30() # max_scope_depth
763 code = read_bytes(code_length)
764 if method_idx in method_idxs:
765 m = Method(code, local_count)
766 methods[method_idxs[method_idx]] = m
767 exception_count = u30()
768 for _c2 in range(exception_count):
775 for _c2 in range(trait_count):
778 assert p + code_reader.tell() == len(code_tag)
779 assert len(methods) == len(method_idxs)
781 method_pyfunctions = {}
783 def extract_function(func_name):
784 if func_name in method_pyfunctions:
785 return method_pyfunctions[func_name]
786 if func_name not in methods:
787 raise ExtractorError(u'Cannot find function %r' % func_name)
788 m = methods[func_name]
791 registers = ['(this)'] + list(args) + [None] * m.local_count
793 coder = io.BytesIO(m.code)
795 opcode = struct.unpack('!B', coder.read(1))[0]
796 if opcode == 36: # pushbyte
797 v = struct.unpack('!B', coder.read(1))[0]
799 elif opcode == 44: # pushstring
801 stack.append(constant_strings[idx])
802 elif opcode == 48: # pushscope
803 # We don't implement the scope register, so we'll just
804 # ignore the popped value
806 elif opcode == 70: # callproperty
808 mname = multinames[index]
809 arg_count = u30(coder)
810 args = list(reversed(
811 [stack.pop() for _ in range(arg_count)]))
813 if mname == u'split':
814 assert len(args) == 1
815 assert isinstance(args[0], compat_str)
816 assert isinstance(obj, compat_str)
820 res = obj.split(args[0])
822 elif mname == u'slice':
823 assert len(args) == 1
824 assert isinstance(args[0], int)
825 assert isinstance(obj, list)
828 elif mname == u'join':
829 assert len(args) == 1
830 assert isinstance(args[0], compat_str)
831 assert isinstance(obj, list)
832 res = args[0].join(obj)
834 elif mname in method_pyfunctions:
835 stack.append(method_pyfunctions[mname](args))
837 raise NotImplementedError(
838 u'Unsupported property %r on %r'
840 elif opcode == 72: # returnvalue
843 elif opcode == 79: # callpropvoid
845 mname = multinames[index]
846 arg_count = u30(coder)
847 args = list(reversed(
848 [stack.pop() for _ in range(arg_count)]))
850 if mname == u'reverse':
851 assert isinstance(obj, list)
854 raise NotImplementedError(
855 u'Unsupported (void) property %r on %r'
857 elif opcode == 93: # findpropstrict
859 mname = multinames[index]
860 res = extract_function(mname)
862 elif opcode == 97: # setproperty
867 assert isinstance(obj, list)
868 assert isinstance(idx, int)
870 elif opcode == 98: # getlocal
872 stack.append(registers[index])
873 elif opcode == 99: # setlocal
876 registers[index] = value
877 elif opcode == 102: # getproperty
879 pname = multinames[index]
880 if pname == u'length':
882 assert isinstance(obj, list)
883 stack.append(len(obj))
884 else: # Assume attribute access
886 assert isinstance(idx, int)
888 assert isinstance(obj, list)
889 stack.append(obj[idx])
890 elif opcode == 128: # coerce
892 elif opcode == 133: # coerce_s
893 assert isinstance(stack[-1], (type(None), compat_str))
894 elif opcode == 164: # modulo
897 res = value1 % value2
899 elif opcode == 208: # getlocal_0
900 stack.append(registers[0])
901 elif opcode == 209: # getlocal_1
902 stack.append(registers[1])
903 elif opcode == 210: # getlocal_2
904 stack.append(registers[2])
905 elif opcode == 211: # getlocal_3
906 stack.append(registers[3])
907 elif opcode == 214: # setlocal_2
908 registers[2] = stack.pop()
909 elif opcode == 215: # setlocal_3
910 registers[3] = stack.pop()
912 raise NotImplementedError(
913 u'Unsupported opcode %d' % opcode)
915 method_pyfunctions[func_name] = resfunc
918 initial_function = extract_function(u'decipher')
919 return lambda s: initial_function([s])
921 def _decrypt_signature(self, s, video_id, player_url, age_gate=False):
922 """Turn the encrypted s field into a working signature"""
924 if player_url is not None:
925 if player_url.startswith(u'//'):
926 player_url = u'https:' + player_url
928 player_id = (player_url, len(s))
929 if player_id not in self._player_cache:
930 func = self._extract_signature_function(
931 video_id, player_url, len(s)
933 self._player_cache[player_id] = func
934 func = self._player_cache[player_id]
935 if self._downloader.params.get('youtube_print_sig_code'):
936 self._print_sig_code(func, len(s))
939 tb = traceback.format_exc()
940 self._downloader.report_warning(
941 u'Automatic signature extraction failed: ' + tb)
943 self._downloader.report_warning(
944 u'Warning: Falling back to static signature algorithm')
946 return self._static_decrypt_signature(
947 s, video_id, player_url, age_gate)
949 def _static_decrypt_signature(self, s, video_id, player_url, age_gate):
951 # The videos with age protection use another player, so the
952 # algorithms can be different.
954 return s[2:63] + s[82] + s[64:82] + s[63]
957 return s[86:29:-1] + s[88] + s[28:5:-1]
959 return s[25] + s[3:25] + s[0] + s[26:42] + s[79] + s[43:79] + s[91] + s[80:83]
961 return s[84:27:-1] + s[86] + s[26:5:-1]
963 return s[25] + s[3:25] + s[2] + s[26:40] + s[77] + s[41:77] + s[89] + s[78:81]
965 return s[84:78:-1] + s[87] + s[77:60:-1] + s[0] + s[59:3:-1]
967 return s[7:28] + s[87] + s[29:45] + s[55] + s[46:55] + s[2] + s[56:87] + s[28]
969 return s[6:27] + s[4] + s[28:39] + s[27] + s[40:59] + s[2] + s[60:]
971 return s[80:72:-1] + s[16] + s[71:39:-1] + s[72] + s[38:16:-1] + s[82] + s[15::-1]
973 return s[3:11] + s[0] + s[12:55] + s[84] + s[56:84]
975 return s[78:70:-1] + s[14] + s[69:37:-1] + s[70] + s[36:14:-1] + s[80] + s[:14][::-1]
977 return s[80:63:-1] + s[0] + s[62:0:-1] + s[63]
979 return s[80:37:-1] + s[7] + s[36:7:-1] + s[0] + s[6:0:-1] + s[37]
981 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]
983 return s[1:19] + s[0] + s[20:68] + s[19] + s[69:80]
985 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]
988 raise ExtractorError(u'Unable to decrypt signature, key length %d not supported; retrying might work' % (len(s)))
990 def _get_available_subtitles(self, video_id, webpage):
992 sub_list = self._download_webpage(
993 'http://video.google.com/timedtext?hl=en&type=list&v=%s' % video_id,
994 video_id, note=False)
995 except ExtractorError as err:
996 self._downloader.report_warning(u'unable to download video subtitles: %s' % compat_str(err))
998 lang_list = re.findall(r'name="([^"]*)"[^>]+lang_code="([\w\-]+)"', sub_list)
1003 params = compat_urllib_parse.urlencode({
1006 'fmt': self._downloader.params.get('subtitlesformat', 'srt'),
1007 'name': unescapeHTML(l[0]).encode('utf-8'),
1009 url = u'http://www.youtube.com/api/timedtext?' + params
1010 sub_lang_list[lang] = url
1011 if not sub_lang_list:
1012 self._downloader.report_warning(u'video doesn\'t have subtitles')
1014 return sub_lang_list
1016 def _get_available_automatic_caption(self, video_id, webpage):
1017 """We need the webpage for getting the captions url, pass it as an
1018 argument to speed up the process."""
1019 sub_format = self._downloader.params.get('subtitlesformat', 'srt')
1020 self.to_screen(u'%s: Looking for automatic captions' % video_id)
1021 mobj = re.search(r';ytplayer.config = ({.*?});', webpage)
1022 err_msg = u'Couldn\'t find automatic captions for %s' % video_id
1024 self._downloader.report_warning(err_msg)
1026 player_config = json.loads(mobj.group(1))
1028 args = player_config[u'args']
1029 caption_url = args[u'ttsurl']
1030 timestamp = args[u'timestamp']
1031 # We get the available subtitles
1032 list_params = compat_urllib_parse.urlencode({
1037 list_url = caption_url + '&' + list_params
1038 caption_list = self._download_xml(list_url, video_id)
1039 original_lang_node = caption_list.find('track')
1040 if original_lang_node is None or original_lang_node.attrib.get('kind') != 'asr' :
1041 self._downloader.report_warning(u'Video doesn\'t have automatic captions')
1043 original_lang = original_lang_node.attrib['lang_code']
1046 for lang_node in caption_list.findall('target'):
1047 sub_lang = lang_node.attrib['lang_code']
1048 params = compat_urllib_parse.urlencode({
1049 'lang': original_lang,
1055 sub_lang_list[sub_lang] = caption_url + '&' + params
1056 return sub_lang_list
1057 # An extractor error can be raise by the download process if there are
1058 # no automatic captions but there are subtitles
1059 except (KeyError, ExtractorError):
1060 self._downloader.report_warning(err_msg)
1063 def _extract_id(self, url):
1064 mobj = re.match(self._VALID_URL, url, re.VERBOSE)
1066 raise ExtractorError(u'Invalid URL: %s' % url)
1067 video_id = mobj.group(2)
1070 def _get_video_url_list(self, url_map):
1072 Transform a dictionary in the format {itag:url} to a list of (itag, url)
1073 with the requested formats.
1075 existing_formats = [x for x in self._formats if x in url_map]
1076 if len(existing_formats) == 0:
1077 raise ExtractorError(u'no known formats available for video')
1078 video_url_list = [(f, url_map[f]) for f in existing_formats] # All formats
1079 video_url_list.reverse() # order worst to best
1080 return video_url_list
1082 def _extract_from_m3u8(self, manifest_url, video_id):
1084 def _get_urls(_manifest):
1085 lines = _manifest.split('\n')
1086 urls = filter(lambda l: l and not l.startswith('#'),
1089 manifest = self._download_webpage(manifest_url, video_id, u'Downloading formats manifest')
1090 formats_urls = _get_urls(manifest)
1091 for format_url in formats_urls:
1092 itag = self._search_regex(r'itag/(\d+?)/', format_url, 'itag')
1093 url_map[itag] = format_url
1096 def _extract_annotations(self, video_id):
1097 url = 'https://www.youtube.com/annotations_invideo?features=1&legacy=1&video_id=%s' % video_id
1098 return self._download_webpage(url, video_id, note=u'Searching for annotations.', errnote=u'Unable to download video annotations.')
1100 def _real_extract(self, url):
1101 # Extract original video URL from URL with redirection, like age verification, using next_url parameter
1102 mobj = re.search(self._NEXT_URL_RE, url)
1104 url = 'https://www.youtube.com/' + compat_urllib_parse.unquote(mobj.group(1)).lstrip('/')
1105 video_id = self._extract_id(url)
1108 url = 'https://www.youtube.com/watch?v=%s&gl=US&hl=en&has_verified=1' % video_id
1109 video_webpage = self._download_webpage(url, video_id)
1111 # Attempt to extract SWF player URL
1112 mobj = re.search(r'swfConfig.*?"(https?:\\/\\/.*?watch.*?-.*?\.swf)"', video_webpage)
1113 if mobj is not None:
1114 player_url = re.sub(r'\\(.)', r'\1', mobj.group(1))
1119 self.report_video_info_webpage_download(video_id)
1120 if re.search(r'player-age-gate-content">', video_webpage) is not None:
1121 self.report_age_confirmation()
1123 # We simulate the access to the video from www.youtube.com/v/{video_id}
1124 # this can be viewed without login into Youtube
1125 data = compat_urllib_parse.urlencode({'video_id': video_id,
1126 'el': 'player_embedded',
1129 'eurl': 'https://youtube.googleapis.com/v/' + video_id,
1133 video_info_url = 'https://www.youtube.com/get_video_info?' + data
1134 video_info_webpage = self._download_webpage(video_info_url, video_id,
1136 errnote='unable to download video info webpage')
1137 video_info = compat_parse_qs(video_info_webpage)
1140 for el_type in ['&el=embedded', '&el=detailpage', '&el=vevo', '']:
1141 video_info_url = ('https://www.youtube.com/get_video_info?&video_id=%s%s&ps=default&eurl=&gl=US&hl=en'
1142 % (video_id, el_type))
1143 video_info_webpage = self._download_webpage(video_info_url, video_id,
1145 errnote='unable to download video info webpage')
1146 video_info = compat_parse_qs(video_info_webpage)
1147 if 'token' in video_info:
1149 if 'token' not in video_info:
1150 if 'reason' in video_info:
1151 raise ExtractorError(u'YouTube said: %s' % video_info['reason'][0], expected=True)
1153 raise ExtractorError(u'"token" parameter not in video info for unknown reason')
1155 if 'view_count' in video_info:
1156 view_count = int(video_info['view_count'][0])
1160 # Check for "rental" videos
1161 if 'ypc_video_rental_bar_text' in video_info and 'author' not in video_info:
1162 raise ExtractorError(u'"rental" videos not supported')
1164 # Start extracting information
1165 self.report_information_extraction(video_id)
1168 if 'author' not in video_info:
1169 raise ExtractorError(u'Unable to extract uploader name')
1170 video_uploader = compat_urllib_parse.unquote_plus(video_info['author'][0])
1173 video_uploader_id = None
1174 mobj = re.search(r'<link itemprop="url" href="http://www.youtube.com/(?:user|channel)/([^"]+)">', video_webpage)
1175 if mobj is not None:
1176 video_uploader_id = mobj.group(1)
1178 self._downloader.report_warning(u'unable to extract uploader nickname')
1181 if 'title' in video_info:
1182 video_title = compat_urllib_parse.unquote_plus(video_info['title'][0])
1184 self._downloader.report_warning(u'Unable to extract video title')
1188 # We try first to get a high quality image:
1189 m_thumb = re.search(r'<span itemprop="thumbnail".*?href="(.*?)">',
1190 video_webpage, re.DOTALL)
1191 if m_thumb is not None:
1192 video_thumbnail = m_thumb.group(1)
1193 elif 'thumbnail_url' not in video_info:
1194 self._downloader.report_warning(u'unable to extract video thumbnail')
1195 video_thumbnail = None
1196 else: # don't panic if we can't find it
1197 video_thumbnail = compat_urllib_parse.unquote_plus(video_info['thumbnail_url'][0])
1201 mobj = re.search(r'id="eow-date.*?>(.*?)</span>', video_webpage, re.DOTALL)
1202 if mobj is not None:
1203 upload_date = ' '.join(re.sub(r'[/,-]', r' ', mobj.group(1)).split())
1204 upload_date = unified_strdate(upload_date)
1207 video_description = get_element_by_id("eow-description", video_webpage)
1208 if video_description:
1209 video_description = re.sub(r'''(?x)
1211 (?:[a-zA-Z-]+="[^"]+"\s+)*?
1213 (?:[a-zA-Z-]+="[^"]+"\s+)*?
1214 class="yt-uix-redirect-link"\s*>
1217 ''', r'\1', video_description)
1218 video_description = clean_html(video_description)
1220 fd_mobj = re.search(r'<meta name="description" content="([^"]+)"', video_webpage)
1222 video_description = unescapeHTML(fd_mobj.group(1))
1224 video_description = u''
1226 def _extract_count(klass):
1227 count = self._search_regex(
1228 r'class="%s">([\d,]+)</span>' % re.escape(klass),
1229 video_webpage, klass, default=None)
1230 if count is not None:
1231 return int(count.replace(',', ''))
1233 like_count = _extract_count(u'likes-count')
1234 dislike_count = _extract_count(u'dislikes-count')
1237 video_subtitles = self.extract_subtitles(video_id, video_webpage)
1239 if self._downloader.params.get('listsubtitles', False):
1240 self._list_available_subtitles(video_id, video_webpage)
1243 if 'length_seconds' not in video_info:
1244 self._downloader.report_warning(u'unable to extract video duration')
1245 video_duration = None
1247 video_duration = int(compat_urllib_parse.unquote_plus(video_info['length_seconds'][0]))
1250 video_annotations = None
1251 if self._downloader.params.get('writeannotations', False):
1252 video_annotations = self._extract_annotations(video_id)
1254 # Decide which formats to download
1257 mobj = re.search(r';ytplayer.config = ({.*?});', video_webpage)
1259 raise ValueError('Could not find vevo ID')
1260 info = json.loads(mobj.group(1))
1262 # Easy way to know if the 's' value is in url_encoded_fmt_stream_map
1263 # this signatures are encrypted
1264 if 'url_encoded_fmt_stream_map' not in args:
1265 raise ValueError(u'No stream_map present') # caught below
1266 re_signature = re.compile(r'[&,]s=')
1267 m_s = re_signature.search(args['url_encoded_fmt_stream_map'])
1269 self.to_screen(u'%s: Encrypted signatures detected.' % video_id)
1270 video_info['url_encoded_fmt_stream_map'] = [args['url_encoded_fmt_stream_map']]
1271 m_s = re_signature.search(args.get('adaptive_fmts', u''))
1273 if 'adaptive_fmts' in video_info:
1274 video_info['adaptive_fmts'][0] += ',' + args['adaptive_fmts']
1276 video_info['adaptive_fmts'] = [args['adaptive_fmts']]
1280 if 'conn' in video_info and video_info['conn'][0].startswith('rtmp'):
1281 self.report_rtmp_download()
1282 video_url_list = [('_rtmp', video_info['conn'][0])]
1283 elif len(video_info.get('url_encoded_fmt_stream_map', [])) >= 1 or len(video_info.get('adaptive_fmts', [])) >= 1:
1284 encoded_url_map = video_info.get('url_encoded_fmt_stream_map', [''])[0] + ',' + video_info.get('adaptive_fmts',[''])[0]
1285 if 'rtmpe%3Dyes' in encoded_url_map:
1286 raise ExtractorError('rtmpe downloads are not supported, see https://github.com/rg3/youtube-dl/issues/343 for more information.', expected=True)
1288 for url_data_str in encoded_url_map.split(','):
1289 url_data = compat_parse_qs(url_data_str)
1290 if 'itag' in url_data and 'url' in url_data:
1291 url = url_data['url'][0]
1292 if 'sig' in url_data:
1293 url += '&signature=' + url_data['sig'][0]
1294 elif 's' in url_data:
1295 encrypted_sig = url_data['s'][0]
1296 if self._downloader.params.get('verbose'):
1298 if player_url is None:
1299 player_version = 'unknown'
1301 player_version = self._search_regex(
1302 r'-(.+)\.swf$', player_url,
1303 u'flash player', fatal=False)
1304 player_desc = 'flash player %s' % player_version
1306 player_version = self._search_regex(
1307 r'html5player-(.+?)\.js', video_webpage,
1308 'html5 player', fatal=False)
1309 player_desc = u'html5 player %s' % player_version
1311 parts_sizes = u'.'.join(compat_str(len(part)) for part in encrypted_sig.split('.'))
1312 self.to_screen(u'encrypted signature length %d (%s), itag %s, %s' %
1313 (len(encrypted_sig), parts_sizes, url_data['itag'][0], player_desc))
1316 jsplayer_url_json = self._search_regex(
1317 r'"assets":.+?"js":\s*("[^"]+")',
1318 video_webpage, u'JS player URL')
1319 player_url = json.loads(jsplayer_url_json)
1321 signature = self._decrypt_signature(
1322 encrypted_sig, video_id, player_url, age_gate)
1323 url += '&signature=' + signature
1324 if 'ratebypass' not in url:
1325 url += '&ratebypass=yes'
1326 url_map[url_data['itag'][0]] = url
1327 video_url_list = self._get_video_url_list(url_map)
1328 elif video_info.get('hlsvp'):
1329 manifest_url = video_info['hlsvp'][0]
1330 url_map = self._extract_from_m3u8(manifest_url, video_id)
1331 video_url_list = self._get_video_url_list(url_map)
1333 raise ExtractorError(u'no conn, hlsvp or url_encoded_fmt_stream_map information found in video info')
1336 for itag, video_real_url in video_url_list:
1339 'url': video_real_url,
1340 'player_url': player_url,
1342 dct.update(self._formats[itag])
1345 self._sort_formats(formats)
1349 'uploader': video_uploader,
1350 'uploader_id': video_uploader_id,
1351 'upload_date': upload_date,
1352 'title': video_title,
1353 'thumbnail': video_thumbnail,
1354 'description': video_description,
1355 'subtitles': video_subtitles,
1356 'duration': video_duration,
1357 'age_limit': 18 if age_gate else 0,
1358 'annotations': video_annotations,
1359 'webpage_url': 'https://www.youtube.com/watch?v=%s' % video_id,
1360 'view_count': view_count,
1361 'like_count': like_count,
1362 'dislike_count': dislike_count,
1366 class YoutubePlaylistIE(YoutubeBaseInfoExtractor):
1367 IE_DESC = u'YouTube.com playlists'
1368 _VALID_URL = r"""(?:
1373 (?:course|view_play_list|my_playlists|artist|playlist|watch)
1374 \? (?:.*?&)*? (?:p|a|list)=
1377 ((?:PL|EC|UU|FL|RD)?[0-9A-Za-z-_]{10,})
1380 ((?:PL|EC|UU|FL|RD)[0-9A-Za-z-_]{10,})
1382 _TEMPLATE_URL = 'https://www.youtube.com/playlist?list=%s&page=%s'
1383 _MORE_PAGES_INDICATOR = r'data-link-type="next"'
1384 _VIDEO_RE = r'href="/watch\?v=(?P<id>[0-9A-Za-z_-]{11})&[^"]*?index=(?P<index>\d+)'
1385 IE_NAME = u'youtube:playlist'
1388 def suitable(cls, url):
1389 """Receives a URL and returns True if suitable for this IE."""
1390 return re.match(cls._VALID_URL, url, re.VERBOSE) is not None
1392 def _real_initialize(self):
1395 def _ids_to_results(self, ids):
1396 return [self.url_result(vid_id, 'Youtube', video_id=vid_id)
1399 def _extract_mix(self, playlist_id):
1400 # The mixes are generated from a a single video
1401 # the id of the playlist is just 'RD' + video_id
1402 url = 'https://youtube.com/watch?v=%s&list=%s' % (playlist_id[-11:], playlist_id)
1403 webpage = self._download_webpage(url, playlist_id, u'Downloading Youtube mix')
1404 title_span = (get_element_by_attribute('class', 'title long-title', webpage) or
1405 get_element_by_attribute('class', 'title ', webpage))
1406 title = clean_html(title_span)
1407 video_re = r'data-index="\d+".*?href="/watch\?v=([0-9A-Za-z_-]{11})&[^"]*?list=%s' % re.escape(playlist_id)
1408 ids = orderedSet(re.findall(video_re, webpage))
1409 url_results = self._ids_to_results(ids)
1411 return self.playlist_result(url_results, playlist_id, title)
1413 def _real_extract(self, url):
1414 # Extract playlist id
1415 mobj = re.match(self._VALID_URL, url, re.VERBOSE)
1417 raise ExtractorError(u'Invalid URL: %s' % url)
1418 playlist_id = mobj.group(1) or mobj.group(2)
1420 # Check if it's a video-specific URL
1421 query_dict = compat_urlparse.parse_qs(compat_urlparse.urlparse(url).query)
1422 if 'v' in query_dict:
1423 video_id = query_dict['v'][0]
1424 if self._downloader.params.get('noplaylist'):
1425 self.to_screen(u'Downloading just video %s because of --no-playlist' % video_id)
1426 return self.url_result(video_id, 'Youtube', video_id=video_id)
1428 self.to_screen(u'Downloading playlist PL%s - add --no-playlist to just download video %s' % (playlist_id, video_id))
1430 if playlist_id.startswith('RD'):
1431 # Mixes require a custom extraction process
1432 return self._extract_mix(playlist_id)
1433 if playlist_id.startswith('TL'):
1434 raise ExtractorError(u'For downloading YouTube.com top lists, use '
1435 u'the "yttoplist" keyword, for example "youtube-dl \'yttoplist:music:Top Tracks\'"', expected=True)
1437 # Extract the video ids from the playlist pages
1440 for page_num in itertools.count(1):
1441 url = self._TEMPLATE_URL % (playlist_id, page_num)
1442 page = self._download_webpage(url, playlist_id, u'Downloading page #%s' % page_num)
1443 matches = re.finditer(self._VIDEO_RE, page)
1444 # We remove the duplicates and the link with index 0
1445 # (it's not the first video of the playlist)
1446 new_ids = orderedSet(m.group('id') for m in matches if m.group('index') != '0')
1449 if re.search(self._MORE_PAGES_INDICATOR, page) is None:
1453 playlist_title = self._og_search_title(page)
1454 except RegexNotFoundError:
1455 self.report_warning(
1456 u'Playlist page is missing OpenGraph title, falling back ...',
1458 playlist_title = self._html_search_regex(
1459 r'<h1 class="pl-header-title">(.*?)</h1>', page, u'title')
1461 url_results = self._ids_to_results(ids)
1462 return self.playlist_result(url_results, playlist_id, playlist_title)
1465 class YoutubeTopListIE(YoutubePlaylistIE):
1466 IE_NAME = u'youtube:toplist'
1467 IE_DESC = (u'YouTube.com top lists, "yttoplist:{channel}:{list title}"'
1468 u' (Example: "yttoplist:music:Top Tracks")')
1469 _VALID_URL = r'yttoplist:(?P<chann>.*?):(?P<title>.*?)$'
1471 def _real_extract(self, url):
1472 mobj = re.match(self._VALID_URL, url)
1473 channel = mobj.group('chann')
1474 title = mobj.group('title')
1475 query = compat_urllib_parse.urlencode({'title': title})
1476 playlist_re = 'href="([^"]+?%s[^"]+?)"' % re.escape(query)
1477 channel_page = self._download_webpage('https://www.youtube.com/%s' % channel, title)
1478 link = self._html_search_regex(playlist_re, channel_page, u'list')
1479 url = compat_urlparse.urljoin('https://www.youtube.com/', link)
1481 video_re = r'data-index="\d+".*?data-video-id="([0-9A-Za-z_-]{11})"'
1483 # sometimes the webpage doesn't contain the videos
1484 # retry until we get them
1485 for i in itertools.count(0):
1486 msg = u'Downloading Youtube mix'
1488 msg += ', retry #%d' % i
1489 webpage = self._download_webpage(url, title, msg)
1490 ids = orderedSet(re.findall(video_re, webpage))
1493 url_results = self._ids_to_results(ids)
1494 return self.playlist_result(url_results, playlist_title=title)
1497 class YoutubeChannelIE(InfoExtractor):
1498 IE_DESC = u'YouTube.com channels'
1499 _VALID_URL = r"^(?:https?://)?(?:youtu\.be|(?:\w+\.)?youtube(?:-nocookie)?\.com)/channel/([0-9A-Za-z_-]+)"
1500 _MORE_PAGES_INDICATOR = 'yt-uix-load-more'
1501 _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'
1502 IE_NAME = u'youtube:channel'
1504 def extract_videos_from_page(self, page):
1506 for mobj in re.finditer(r'href="/watch\?v=([0-9A-Za-z_-]+)&?', page):
1507 if mobj.group(1) not in ids_in_page:
1508 ids_in_page.append(mobj.group(1))
1511 def _real_extract(self, url):
1512 # Extract channel id
1513 mobj = re.match(self._VALID_URL, url)
1515 raise ExtractorError(u'Invalid URL: %s' % url)
1517 # Download channel page
1518 channel_id = mobj.group(1)
1520 url = 'https://www.youtube.com/channel/%s/videos' % channel_id
1521 channel_page = self._download_webpage(url, channel_id)
1522 autogenerated = re.search(r'''(?x)
1524 channel-header-autogenerated-label|
1525 yt-channel-title-autogenerated
1526 )[^"]*"''', channel_page) is not None
1529 # The videos are contained in a single page
1530 # the ajax pages can't be used, they are empty
1531 video_ids = self.extract_videos_from_page(channel_page)
1533 # Download all channel pages using the json-based channel_ajax query
1534 for pagenum in itertools.count(1):
1535 url = self._MORE_PAGES_URL % (pagenum, channel_id)
1536 page = self._download_webpage(url, channel_id,
1537 u'Downloading page #%s' % pagenum)
1539 page = json.loads(page)
1541 ids_in_page = self.extract_videos_from_page(page['content_html'])
1542 video_ids.extend(ids_in_page)
1544 if self._MORE_PAGES_INDICATOR not in page['load_more_widget_html']:
1547 self._downloader.to_screen(u'[youtube] Channel %s: Found %i videos' % (channel_id, len(video_ids)))
1549 url_entries = [self.url_result(video_id, 'Youtube', video_id=video_id)
1550 for video_id in video_ids]
1551 return self.playlist_result(url_entries, channel_id)
1554 class YoutubeUserIE(InfoExtractor):
1555 IE_DESC = u'YouTube.com user videos (URL or "ytuser" keyword)'
1556 _VALID_URL = r'(?:(?:(?:https?://)?(?:\w+\.)?youtube\.com/(?:user/)?(?!(?:attribution_link|watch)(?:$|[^a-z_A-Z0-9-])))|ytuser:)(?!feed/)([A-Za-z0-9_-]+)'
1557 _TEMPLATE_URL = 'http://gdata.youtube.com/feeds/api/users/%s'
1558 _GDATA_PAGE_SIZE = 50
1559 _GDATA_URL = 'http://gdata.youtube.com/feeds/api/users/%s/uploads?max-results=%d&start-index=%d&alt=json'
1560 IE_NAME = u'youtube:user'
1563 def suitable(cls, url):
1564 # Don't return True if the url can be extracted with other youtube
1565 # extractor, the regex would is too permissive and it would match.
1566 other_ies = iter(klass for (name, klass) in globals().items() if name.endswith('IE') and klass is not cls)
1567 if any(ie.suitable(url) for ie in other_ies): return False
1568 else: return super(YoutubeUserIE, cls).suitable(url)
1570 def _real_extract(self, url):
1572 mobj = re.match(self._VALID_URL, url)
1574 raise ExtractorError(u'Invalid URL: %s' % url)
1576 username = mobj.group(1)
1578 # Download video ids using YouTube Data API. Result size per
1579 # query is limited (currently to 50 videos) so we need to query
1580 # page by page until there are no video ids - it means we got
1585 for pagenum in itertools.count(0):
1586 start_index = pagenum * self._GDATA_PAGE_SIZE + 1
1588 gdata_url = self._GDATA_URL % (username, self._GDATA_PAGE_SIZE, start_index)
1589 page = self._download_webpage(gdata_url, username,
1590 u'Downloading video ids from %d to %d' % (start_index, start_index + self._GDATA_PAGE_SIZE))
1593 response = json.loads(page)
1594 except ValueError as err:
1595 raise ExtractorError(u'Invalid JSON in API response: ' + compat_str(err))
1596 if 'entry' not in response['feed']:
1597 # Number of videos is a multiple of self._MAX_RESULTS
1600 # Extract video identifiers
1601 entries = response['feed']['entry']
1602 for entry in entries:
1603 title = entry['title']['$t']
1604 video_id = entry['id']['$t'].split('/')[-1]
1605 url_results.append({
1608 'ie_key': 'Youtube',
1613 # A little optimization - if current page is not
1614 # "full", ie. does not contain PAGE_SIZE video ids then
1615 # we can assume that this page is the last one - there
1616 # are no more ids on further pages - no need to query
1619 if len(entries) < self._GDATA_PAGE_SIZE:
1622 return self.playlist_result(url_results, playlist_title=username)
1625 class YoutubeSearchIE(SearchInfoExtractor):
1626 IE_DESC = u'YouTube.com searches'
1627 _API_URL = 'https://gdata.youtube.com/feeds/api/videos?q=%s&start-index=%i&max-results=50&v=2&alt=jsonc'
1629 IE_NAME = u'youtube:search'
1630 _SEARCH_KEY = 'ytsearch'
1632 def _get_n_results(self, query, n):
1633 """Get a specified number of results for a query"""
1639 while (50 * pagenum) < limit:
1640 result_url = self._API_URL % (compat_urllib_parse.quote_plus(query), (50*pagenum)+1)
1641 data_json = self._download_webpage(
1642 result_url, video_id=u'query "%s"' % query,
1643 note=u'Downloading page %s' % (pagenum + 1),
1644 errnote=u'Unable to download API page')
1645 data = json.loads(data_json)
1646 api_response = data['data']
1648 if 'items' not in api_response:
1649 raise ExtractorError(u'[youtube] No video results')
1651 new_ids = list(video['id'] for video in api_response['items'])
1652 video_ids += new_ids
1654 limit = min(n, api_response['totalItems'])
1657 if len(video_ids) > n:
1658 video_ids = video_ids[:n]
1659 videos = [self.url_result(video_id, 'Youtube', video_id=video_id)
1660 for video_id in video_ids]
1661 return self.playlist_result(videos, query)
1663 class YoutubeSearchDateIE(YoutubeSearchIE):
1664 IE_NAME = YoutubeSearchIE.IE_NAME + ':date'
1665 _API_URL = 'https://gdata.youtube.com/feeds/api/videos?q=%s&start-index=%i&max-results=50&v=2&alt=jsonc&orderby=published'
1666 _SEARCH_KEY = 'ytsearchdate'
1667 IE_DESC = u'YouTube.com searches, newest videos first'
1669 class YoutubeShowIE(InfoExtractor):
1670 IE_DESC = u'YouTube.com (multi-season) shows'
1671 _VALID_URL = r'https?://www\.youtube\.com/show/(.*)'
1672 IE_NAME = u'youtube:show'
1674 def _real_extract(self, url):
1675 mobj = re.match(self._VALID_URL, url)
1676 show_name = mobj.group(1)
1677 webpage = self._download_webpage(url, show_name, u'Downloading show webpage')
1678 # There's one playlist for each season of the show
1679 m_seasons = list(re.finditer(r'href="(/playlist\?list=.*?)"', webpage))
1680 self.to_screen(u'%s: Found %s seasons' % (show_name, len(m_seasons)))
1681 return [self.url_result('https://www.youtube.com' + season.group(1), 'YoutubePlaylist') for season in m_seasons]
1684 class YoutubeFeedsInfoExtractor(YoutubeBaseInfoExtractor):
1686 Base class for extractors that fetch info from
1687 http://www.youtube.com/feed_ajax
1688 Subclasses must define the _FEED_NAME and _PLAYLIST_TITLE properties.
1690 _LOGIN_REQUIRED = True
1691 # use action_load_personal_feed instead of action_load_system_feed
1692 _PERSONAL_FEED = False
1695 def _FEED_TEMPLATE(self):
1696 action = 'action_load_system_feed'
1697 if self._PERSONAL_FEED:
1698 action = 'action_load_personal_feed'
1699 return 'http://www.youtube.com/feed_ajax?%s=1&feed_name=%s&paging=%%s' % (action, self._FEED_NAME)
1703 return u'youtube:%s' % self._FEED_NAME
1705 def _real_initialize(self):
1708 def _real_extract(self, url):
1711 for i in itertools.count(1):
1712 info = self._download_webpage(self._FEED_TEMPLATE % paging,
1713 u'%s feed' % self._FEED_NAME,
1714 u'Downloading page %s' % i)
1715 info = json.loads(info)
1716 feed_html = info['feed_html']
1717 m_ids = re.finditer(r'"/watch\?v=(.*?)["&]', feed_html)
1718 ids = orderedSet(m.group(1) for m in m_ids)
1719 feed_entries.extend(
1720 self.url_result(video_id, 'Youtube', video_id=video_id)
1721 for video_id in ids)
1722 if info['paging'] is None:
1724 paging = info['paging']
1725 return self.playlist_result(feed_entries, playlist_title=self._PLAYLIST_TITLE)
1727 class YoutubeSubscriptionsIE(YoutubeFeedsInfoExtractor):
1728 IE_DESC = u'YouTube.com subscriptions feed, "ytsubs" keyword(requires authentication)'
1729 _VALID_URL = r'https?://www\.youtube\.com/feed/subscriptions|:ytsubs(?:criptions)?'
1730 _FEED_NAME = 'subscriptions'
1731 _PLAYLIST_TITLE = u'Youtube Subscriptions'
1733 class YoutubeRecommendedIE(YoutubeFeedsInfoExtractor):
1734 IE_DESC = u'YouTube.com recommended videos, "ytrec" keyword (requires authentication)'
1735 _VALID_URL = r'https?://www\.youtube\.com/feed/recommended|:ytrec(?:ommended)?'
1736 _FEED_NAME = 'recommended'
1737 _PLAYLIST_TITLE = u'Youtube Recommended videos'
1739 class YoutubeWatchLaterIE(YoutubeFeedsInfoExtractor):
1740 IE_DESC = u'Youtube watch later list, "ytwatchlater" keyword (requires authentication)'
1741 _VALID_URL = r'https?://www\.youtube\.com/feed/watch_later|:ytwatchlater'
1742 _FEED_NAME = 'watch_later'
1743 _PLAYLIST_TITLE = u'Youtube Watch Later'
1744 _PERSONAL_FEED = True
1746 class YoutubeHistoryIE(YoutubeFeedsInfoExtractor):
1747 IE_DESC = u'Youtube watch history, "ythistory" keyword (requires authentication)'
1748 _VALID_URL = u'https?://www\.youtube\.com/feed/history|:ythistory'
1749 _FEED_NAME = 'history'
1750 _PERSONAL_FEED = True
1751 _PLAYLIST_TITLE = u'Youtube Watch History'
1753 class YoutubeFavouritesIE(YoutubeBaseInfoExtractor):
1754 IE_NAME = u'youtube:favorites'
1755 IE_DESC = u'YouTube.com favourite videos, "ytfav" keyword (requires authentication)'
1756 _VALID_URL = r'https?://www\.youtube\.com/my_favorites|:ytfav(?:ou?rites)?'
1757 _LOGIN_REQUIRED = True
1759 def _real_extract(self, url):
1760 webpage = self._download_webpage('https://www.youtube.com/my_favorites', 'Youtube Favourites videos')
1761 playlist_id = self._search_regex(r'list=(.+?)["&]', webpage, u'favourites playlist id')
1762 return self.url_result(playlist_id, 'YoutubePlaylist')
1765 class YoutubeTruncatedURLIE(InfoExtractor):
1766 IE_NAME = 'youtube:truncated_url'
1767 IE_DESC = False # Do not list
1768 _VALID_URL = r'(?:https?://)?[^/]+/watch\?feature=[a-z_]+$'
1770 def _real_extract(self, url):
1771 raise ExtractorError(
1772 u'Did you forget to quote the URL? Remember that & is a meta '
1773 u'character in most shells, so you want to put the URL in quotes, '
1775 u'"http://www.youtube.com/watch?feature=foo&v=BaW_jenozKc" '
1776 u' or simply youtube-dl BaW_jenozKc .',