15 from .common import InfoExtractor, SearchInfoExtractor
16 from .subtitles import SubtitlesInfoExtractor
21 compat_urllib_request,
28 get_element_by_attribute,
36 class YoutubeBaseInfoExtractor(InfoExtractor):
37 """Provide base functions for Youtube extractors"""
38 _LOGIN_URL = 'https://accounts.google.com/ServiceLogin'
39 _LANG_URL = r'https://www.youtube.com/?hl=en&persist_hl=1&gl=US&persist_gl=1&opt_out_ackd=1'
40 _AGE_URL = 'http://www.youtube.com/verify_age?next_url=/&gl=US&hl=en'
41 _NETRC_MACHINE = 'youtube'
42 # If True it will raise an error if no login info is provided
43 _LOGIN_REQUIRED = False
45 def _set_language(self):
46 return bool(self._download_webpage(
48 note=u'Setting language', errnote='unable to set language',
52 (username, password) = self._get_login_info()
53 # No authentication to be performed
55 if self._LOGIN_REQUIRED:
56 raise ExtractorError(u'No login info available, needed for using %s.' % self.IE_NAME, expected=True)
59 login_page = self._download_webpage(
60 self._LOGIN_URL, None,
61 note=u'Downloading login page',
62 errnote=u'unable to fetch login page', fatal=False)
63 if login_page is False:
66 galx = self._search_regex(r'(?s)<input.+?name="GALX".+?value="(.+?)"',
67 login_page, u'Login GALX parameter')
71 u'continue': u'https://www.youtube.com/signin?action_handle_signin=true&feature=sign_in_button&hl=en_US&nomobiletemp=1',
75 u'PersistentCookie': u'yes',
77 u'bgresponse': u'js_disabled',
78 u'checkConnection': u'',
79 u'checkedDomains': u'youtube',
84 u'signIn': u'Sign in',
86 u'service': u'youtube',
90 # Convert to UTF-8 *before* urlencode because Python 2.x's urlencode
92 login_form = dict((k.encode('utf-8'), v.encode('utf-8')) for k,v in login_form_strs.items())
93 login_data = compat_urllib_parse.urlencode(login_form).encode('ascii')
95 req = compat_urllib_request.Request(self._LOGIN_URL, login_data)
96 login_results = self._download_webpage(
98 note=u'Logging in', errnote=u'unable to log in', fatal=False)
99 if login_results is False:
101 if re.search(r'(?i)<form[^>]* id="gaia_loginform"', login_results) is not None:
102 self._downloader.report_warning(u'unable to log in: bad username or password')
106 def _confirm_age(self):
109 'action_confirm': 'Confirm',
111 req = compat_urllib_request.Request(self._AGE_URL, compat_urllib_parse.urlencode(age_form))
113 self._download_webpage(
115 note=u'Confirming age', errnote=u'Unable to confirm age')
118 def _real_initialize(self):
119 if self._downloader is None:
121 if not self._set_language():
123 if not self._login():
128 class YoutubeIE(YoutubeBaseInfoExtractor, SubtitlesInfoExtractor):
129 IE_DESC = u'YouTube.com'
130 _VALID_URL = r"""(?x)^
132 (?:https?://|//)? # http(s):// or protocol-independent URL (optional)
133 (?:(?:(?:(?:\w+\.)?[yY][oO][uU][tT][uU][bB][eE](?:-nocookie)?\.com/|
134 tube\.majestyc\.net/|
135 youtube\.googleapis\.com/) # the various hostnames, with wildcard subdomains
136 (?:.*?\#/)? # handle anchor (#/) redirect urls
137 (?: # the various things that can precede the ID:
138 (?:(?:v|embed|e)/) # v/ or embed/ or e/
139 |(?: # or the v= param in all its forms
140 (?:(?:watch|movie)(?:_popup)?(?:\.php)?)? # preceding watch(_popup|.php) or nothing (like /?v=xxxx)
141 (?:\?|\#!?) # the params delimiter ? or # or #!
142 (?:.*?&)? # any other preceding param (like /?s=tuff&v=xxxx)
146 |youtu\.be/ # just youtu.be/xxxx
148 )? # all until now is optional -> you can pass the naked ID
149 ([0-9A-Za-z_-]{11}) # here is it! the YouTube video ID
150 (?(1).+)? # if we found the ID, everything can follow
152 _NEXT_URL_RE = r'[\?&]next_url=([^&]+)'
154 '5': {'ext': 'flv', 'width': 400, 'height': 240},
155 '6': {'ext': 'flv', 'width': 450, 'height': 270},
156 '13': {'ext': '3gp'},
157 '17': {'ext': '3gp', 'width': 176, 'height': 144},
158 '18': {'ext': 'mp4', 'width': 640, 'height': 360},
159 '22': {'ext': 'mp4', 'width': 1280, 'height': 720},
160 '34': {'ext': 'flv', 'width': 640, 'height': 360},
161 '35': {'ext': 'flv', 'width': 854, 'height': 480},
162 '36': {'ext': '3gp', 'width': 320, 'height': 240},
163 '37': {'ext': 'mp4', 'width': 1920, 'height': 1080},
164 '38': {'ext': 'mp4', 'width': 4096, 'height': 3072},
165 '43': {'ext': 'webm', 'width': 640, 'height': 360},
166 '44': {'ext': 'webm', 'width': 854, 'height': 480},
167 '45': {'ext': 'webm', 'width': 1280, 'height': 720},
168 '46': {'ext': 'webm', 'width': 1920, 'height': 1080},
172 '82': {'ext': 'mp4', 'height': 360, 'resolution': '360p', 'format_note': '3D', 'preference': -20},
173 '83': {'ext': 'mp4', 'height': 480, 'resolution': '480p', 'format_note': '3D', 'preference': -20},
174 '84': {'ext': 'mp4', 'height': 720, 'resolution': '720p', 'format_note': '3D', 'preference': -20},
175 '85': {'ext': 'mp4', 'height': 1080, 'resolution': '1080p', 'format_note': '3D', 'preference': -20},
176 '100': {'ext': 'webm', 'height': 360, 'resolution': '360p', 'format_note': '3D', 'preference': -20},
177 '101': {'ext': 'webm', 'height': 480, 'resolution': '480p', 'format_note': '3D', 'preference': -20},
178 '102': {'ext': 'webm', 'height': 720, 'resolution': '720p', 'format_note': '3D', 'preference': -20},
180 # Apple HTTP Live Streaming
181 '92': {'ext': 'mp4', 'height': 240, 'resolution': '240p', 'format_note': 'HLS', 'preference': -10},
182 '93': {'ext': 'mp4', 'height': 360, 'resolution': '360p', 'format_note': 'HLS', 'preference': -10},
183 '94': {'ext': 'mp4', 'height': 480, 'resolution': '480p', 'format_note': 'HLS', 'preference': -10},
184 '95': {'ext': 'mp4', 'height': 720, 'resolution': '720p', 'format_note': 'HLS', 'preference': -10},
185 '96': {'ext': 'mp4', 'height': 1080, 'resolution': '1080p', 'format_note': 'HLS', 'preference': -10},
186 '132': {'ext': 'mp4', 'height': 240, 'resolution': '240p', 'format_note': 'HLS', 'preference': -10},
187 '151': {'ext': 'mp4', 'height': 72, 'resolution': '72p', 'format_note': 'HLS', 'preference': -10},
190 '133': {'ext': 'mp4', 'height': 240, 'resolution': '240p', 'format_note': 'DASH video', 'preference': -40},
191 '134': {'ext': 'mp4', 'height': 360, 'resolution': '360p', 'format_note': 'DASH video', 'preference': -40},
192 '135': {'ext': 'mp4', 'height': 480, 'resolution': '480p', 'format_note': 'DASH video', 'preference': -40},
193 '136': {'ext': 'mp4', 'height': 720, 'resolution': '720p', 'format_note': 'DASH video', 'preference': -40},
194 '137': {'ext': 'mp4', 'height': 1080, 'resolution': '1080p', 'format_note': 'DASH video', 'preference': -40},
195 '138': {'ext': 'mp4', 'height': 1081, 'resolution': '>1080p', 'format_note': 'DASH video', 'preference': -40},
196 '160': {'ext': 'mp4', 'height': 192, 'resolution': '192p', 'format_note': 'DASH video', 'preference': -40},
199 '139': {'ext': 'm4a', 'format_note': 'DASH audio', 'vcodec': 'none', 'abr': 48, 'preference': -50},
200 '140': {'ext': 'm4a', 'format_note': 'DASH audio', 'vcodec': 'none', 'abr': 128, 'preference': -50},
201 '141': {'ext': 'm4a', 'format_note': 'DASH audio', 'vcodec': 'none', 'abr': 256, 'preference': -50},
204 '242': {'ext': 'webm', 'height': 240, 'resolution': '240p', 'format_note': 'DASH webm', 'preference': -40},
205 '243': {'ext': 'webm', 'height': 360, 'resolution': '360p', 'format_note': 'DASH webm', 'preference': -40},
206 '244': {'ext': 'webm', 'height': 480, 'resolution': '480p', 'format_note': 'DASH webm', 'preference': -40},
207 '245': {'ext': 'webm', 'height': 480, 'resolution': '480p', 'format_note': 'DASH webm', 'preference': -40},
208 '246': {'ext': 'webm', 'height': 480, 'resolution': '480p', 'format_note': 'DASH webm', 'preference': -40},
209 '247': {'ext': 'webm', 'height': 720, 'resolution': '720p', 'format_note': 'DASH webm', 'preference': -40},
210 '248': {'ext': 'webm', 'height': 1080, 'resolution': '1080p', 'format_note': 'DASH webm', 'preference': -40},
213 '171': {'ext': 'webm', 'vcodec': 'none', 'format_note': 'DASH webm audio', 'abr': 48, 'preference': -50},
214 '172': {'ext': 'webm', 'vcodec': 'none', 'format_note': 'DASH webm audio', 'abr': 256, 'preference': -50},
220 u"url": u"http://www.youtube.com/watch?v=BaW_jenozKc",
221 u"file": u"BaW_jenozKc.mp4",
223 u"title": u"youtube-dl test video \"'/\\ä↭𝕐",
224 u"uploader": u"Philipp Hagemeister",
225 u"uploader_id": u"phihag",
226 u"upload_date": u"20121002",
227 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 ."
231 u"url": u"http://www.youtube.com/watch?v=UxxajLWwzqY",
232 u"file": u"UxxajLWwzqY.mp4",
233 u"note": u"Test generic use_cipher_signature video (#897)",
235 u"upload_date": u"20120506",
236 u"title": u"Icona Pop - I Love It (feat. Charli XCX) [OFFICIAL VIDEO]",
237 u"description": u"md5:5b292926389560516e384ac437c0ec07",
238 u"uploader": u"Icona Pop",
239 u"uploader_id": u"IconaPop"
243 u"url": u"https://www.youtube.com/watch?v=07FYdnEawAQ",
244 u"file": u"07FYdnEawAQ.mp4",
245 u"note": u"Test VEVO video with age protection (#956)",
247 u"upload_date": u"20130703",
248 u"title": u"Justin Timberlake - Tunnel Vision (Explicit)",
249 u"description": u"md5:64249768eec3bc4276236606ea996373",
250 u"uploader": u"justintimberlakeVEVO",
251 u"uploader_id": u"justintimberlakeVEVO"
255 u"url": u"//www.YouTube.com/watch?v=yZIXLfi8CZQ",
256 u"file": u"yZIXLfi8CZQ.mp4",
257 u"note": u"Embed-only video (#1746)",
259 u"upload_date": u"20120608",
260 u"title": u"Principal Sexually Assaults A Teacher - Episode 117 - 8th June 2012",
261 u"description": u"md5:09b78bd971f1e3e289601dfba15ca4f7",
262 u"uploader": u"SET India",
263 u"uploader_id": u"setindia"
270 def suitable(cls, url):
271 """Receives a URL and returns True if suitable for this IE."""
272 if YoutubePlaylistIE.suitable(url): return False
273 return re.match(cls._VALID_URL, url) is not None
275 def __init__(self, *args, **kwargs):
276 super(YoutubeIE, self).__init__(*args, **kwargs)
277 self._player_cache = {}
279 def report_video_info_webpage_download(self, video_id):
280 """Report attempt to download video info webpage."""
281 self.to_screen(u'%s: Downloading video info webpage' % video_id)
283 def report_information_extraction(self, video_id):
284 """Report attempt to extract video information."""
285 self.to_screen(u'%s: Extracting video information' % video_id)
287 def report_unavailable_format(self, video_id, format):
288 """Report extracted video URL."""
289 self.to_screen(u'%s: Format %s not available' % (video_id, format))
291 def report_rtmp_download(self):
292 """Indicate the download will use the RTMP protocol."""
293 self.to_screen(u'RTMP download detected')
295 def _extract_signature_function(self, video_id, player_url, slen):
296 id_m = re.match(r'.*-(?P<id>[a-zA-Z0-9_-]+)\.(?P<ext>[a-z]+)$',
298 player_type = id_m.group('ext')
299 player_id = id_m.group('id')
301 # Read from filesystem cache
302 func_id = '%s_%s_%d' % (player_type, player_id, slen)
303 assert os.path.basename(func_id) == func_id
304 cache_dir = get_cachedir(self._downloader.params)
306 cache_enabled = cache_dir is not None
308 cache_fn = os.path.join(os.path.expanduser(cache_dir),
312 with io.open(cache_fn, 'r', encoding='utf-8') as cachef:
313 cache_spec = json.load(cachef)
314 return lambda s: u''.join(s[i] for i in cache_spec)
316 pass # No cache available
318 if player_type == 'js':
319 code = self._download_webpage(
320 player_url, video_id,
321 note=u'Downloading %s player %s' % (player_type, player_id),
322 errnote=u'Download of %s failed' % player_url)
323 res = self._parse_sig_js(code)
324 elif player_type == 'swf':
325 urlh = self._request_webpage(
326 player_url, video_id,
327 note=u'Downloading %s player %s' % (player_type, player_id),
328 errnote=u'Download of %s failed' % player_url)
330 res = self._parse_sig_swf(code)
332 assert False, 'Invalid player type %r' % player_type
336 test_string = u''.join(map(compat_chr, range(slen)))
337 cache_res = res(test_string)
338 cache_spec = [ord(c) for c in cache_res]
340 os.makedirs(os.path.dirname(cache_fn))
341 except OSError as ose:
342 if ose.errno != errno.EEXIST:
344 write_json_file(cache_spec, cache_fn)
346 tb = traceback.format_exc()
347 self._downloader.report_warning(
348 u'Writing cache to %r failed: %s' % (cache_fn, tb))
352 def _print_sig_code(self, func, slen):
353 def gen_sig_code(idxs):
354 def _genslice(start, end, step):
355 starts = u'' if start == 0 else str(start)
356 ends = (u':%d' % (end+step)) if end + step >= 0 else u':'
357 steps = u'' if step == 1 else (u':%d' % step)
358 return u's[%s%s%s]' % (starts, ends, steps)
361 start = '(Never used)' # Quelch pyflakes warnings - start will be
362 # set as soon as step is set
363 for i, prev in zip(idxs[1:], idxs[:-1]):
367 yield _genslice(start, prev, step)
370 if i - prev in [-1, 1]:
375 yield u's[%d]' % prev
379 yield _genslice(start, i, step)
381 test_string = u''.join(map(compat_chr, range(slen)))
382 cache_res = func(test_string)
383 cache_spec = [ord(c) for c in cache_res]
384 expr_code = u' + '.join(gen_sig_code(cache_spec))
385 code = u'if len(s) == %d:\n return %s\n' % (slen, expr_code)
386 self.to_screen(u'Extracted signature function:\n' + code)
388 def _parse_sig_js(self, jscode):
389 funcname = self._search_regex(
390 r'signature=([a-zA-Z]+)', jscode,
391 u'Initial JS player signature function name')
396 return string.lowercase.index(varname)
398 def interpret_statement(stmt, local_vars, allow_recursion=20):
399 if allow_recursion < 0:
400 raise ExtractorError(u'Recursion limit reached')
402 if stmt.startswith(u'var '):
403 stmt = stmt[len(u'var '):]
404 ass_m = re.match(r'^(?P<out>[a-z]+)(?:\[(?P<index>[^\]]+)\])?' +
405 r'=(?P<expr>.*)$', stmt)
407 if ass_m.groupdict().get('index'):
409 lvar = local_vars[ass_m.group('out')]
410 idx = interpret_expression(ass_m.group('index'),
411 local_vars, allow_recursion)
412 assert isinstance(idx, int)
415 expr = ass_m.group('expr')
418 local_vars[ass_m.group('out')] = val
420 expr = ass_m.group('expr')
421 elif stmt.startswith(u'return '):
423 expr = stmt[len(u'return '):]
425 raise ExtractorError(
426 u'Cannot determine left side of statement in %r' % stmt)
428 v = interpret_expression(expr, local_vars, allow_recursion)
431 def interpret_expression(expr, local_vars, allow_recursion):
436 return local_vars[expr]
438 m = re.match(r'^(?P<in>[a-z]+)\.(?P<member>.*)$', expr)
440 member = m.group('member')
441 val = local_vars[m.group('in')]
442 if member == 'split("")':
444 if member == 'join("")':
446 if member == 'length':
448 if member == 'reverse()':
450 slice_m = re.match(r'slice\((?P<idx>.*)\)', member)
452 idx = interpret_expression(
453 slice_m.group('idx'), local_vars, allow_recursion-1)
457 r'^(?P<in>[a-z]+)\[(?P<idx>.+)\]$', expr)
459 val = local_vars[m.group('in')]
460 idx = interpret_expression(m.group('idx'), local_vars,
464 m = re.match(r'^(?P<a>.+?)(?P<op>[%])(?P<b>.+?)$', expr)
466 a = interpret_expression(m.group('a'),
467 local_vars, allow_recursion)
468 b = interpret_expression(m.group('b'),
469 local_vars, allow_recursion)
473 r'^(?P<func>[a-zA-Z]+)\((?P<args>[a-z0-9,]+)\)$', expr)
475 fname = m.group('func')
476 if fname not in functions:
477 functions[fname] = extract_function(fname)
478 argvals = [int(v) if v.isdigit() else local_vars[v]
479 for v in m.group('args').split(',')]
480 return functions[fname](argvals)
481 raise ExtractorError(u'Unsupported JS expression %r' % expr)
483 def extract_function(funcname):
485 r'function ' + re.escape(funcname) +
486 r'\((?P<args>[a-z,]+)\){(?P<code>[^}]+)}',
488 argnames = func_m.group('args').split(',')
491 local_vars = dict(zip(argnames, args))
492 for stmt in func_m.group('code').split(';'):
493 res = interpret_statement(stmt, local_vars)
497 initial_function = extract_function(funcname)
498 return lambda s: initial_function([s])
500 def _parse_sig_swf(self, file_contents):
501 if file_contents[1:3] != b'WS':
502 raise ExtractorError(
503 u'Not an SWF file; header is %r' % file_contents[:3])
504 if file_contents[:1] == b'C':
505 content = zlib.decompress(file_contents[8:])
507 raise NotImplementedError(u'Unsupported compression format %r' %
510 def extract_tags(content):
512 while pos < len(content):
513 header16 = struct.unpack('<H', content[pos:pos+2])[0]
515 tag_code = header16 >> 6
516 tag_len = header16 & 0x3f
518 tag_len = struct.unpack('<I', content[pos:pos+4])[0]
520 assert pos+tag_len <= len(content)
521 yield (tag_code, content[pos:pos+tag_len])
525 for tag_code, tag in extract_tags(content)
527 p = code_tag.index(b'\0', 4) + 1
528 code_reader = io.BytesIO(code_tag[p:])
530 # Parse ABC (AVM2 ByteCode)
531 def read_int(reader=None):
539 b = struct.unpack('<B', buf)[0]
540 res = res | ((b & 0x7f) << shift)
546 def u30(reader=None):
547 res = read_int(reader)
548 assert res & 0xf0000000 == 0
552 def s32(reader=None):
554 if v & 0x80000000 != 0:
555 v = - ((v ^ 0xffffffff) + 1)
558 def read_string(reader=None):
562 resb = reader.read(slen)
563 assert len(resb) == slen
564 return resb.decode('utf-8')
566 def read_bytes(count, reader=None):
569 resb = reader.read(count)
570 assert len(resb) == count
573 def read_byte(reader=None):
574 resb = read_bytes(1, reader=reader)
575 res = struct.unpack('<B', resb)[0]
578 # minor_version + major_version
583 for _c in range(1, int_count):
586 for _c in range(1, uint_count):
589 read_bytes((double_count-1) * 8)
591 constant_strings = [u'']
592 for _c in range(1, string_count):
594 constant_strings.append(s)
595 namespace_count = u30()
596 for _c in range(1, namespace_count):
600 for _c in range(1, ns_set_count):
602 for _c2 in range(count):
604 multiname_count = u30()
613 0x0e: 2, # MultinameA
614 0x1b: 1, # MultinameL
615 0x1c: 1, # MultinameLA
618 for _c in range(1, multiname_count):
620 assert kind in MULTINAME_SIZES, u'Invalid multiname kind %r' % kind
622 u30() # namespace_idx
624 multinames.append(constant_strings[name_idx])
626 multinames.append('[MULTINAME kind: %d]' % kind)
627 for _c2 in range(MULTINAME_SIZES[kind]):
632 MethodInfo = collections.namedtuple(
634 ['NEED_ARGUMENTS', 'NEED_REST'])
636 for method_id in range(method_count):
639 for _ in range(param_count):
641 u30() # name index (always 0 for youtube)
643 if flags & 0x08 != 0:
646 for c in range(option_count):
649 if flags & 0x80 != 0:
650 # Param names present
651 for _ in range(param_count):
653 mi = MethodInfo(flags & 0x01 != 0, flags & 0x04 != 0)
654 method_infos.append(mi)
657 metadata_count = u30()
658 for _c in range(metadata_count):
661 for _c2 in range(item_count):
665 def parse_traits_info():
666 trait_name_idx = u30()
667 kind_full = read_byte()
668 kind = kind_full & 0x0f
669 attrs = kind_full >> 4
671 if kind in [0x00, 0x06]: # Slot or Const
673 u30() # type_name_idx
677 elif kind in [0x01, 0x02, 0x03]: # Method / Getter / Setter
680 methods[multinames[trait_name_idx]] = method_idx
681 elif kind == 0x04: # Class
684 elif kind == 0x05: # Function
687 methods[function_idx] = multinames[trait_name_idx]
689 raise ExtractorError(u'Unsupported trait kind %d' % kind)
691 if attrs & 0x4 != 0: # Metadata present
692 metadata_count = u30()
693 for _c3 in range(metadata_count):
694 u30() # metadata index
699 TARGET_CLASSNAME = u'SignatureDecipher'
700 searched_idx = multinames.index(TARGET_CLASSNAME)
701 searched_class_id = None
703 for class_id in range(class_count):
705 if name_idx == searched_idx:
706 # We found the class we're looking for!
707 searched_class_id = class_id
708 u30() # super_name idx
710 if flags & 0x08 != 0: # Protected namespace is present
711 u30() # protected_ns_idx
713 for _c2 in range(intrf_count):
717 for _c2 in range(trait_count):
720 if searched_class_id is None:
721 raise ExtractorError(u'Target class %r not found' %
726 for class_id in range(class_count):
729 for _c2 in range(trait_count):
730 trait_methods = parse_traits_info()
731 if class_id == searched_class_id:
732 method_names.update(trait_methods.items())
733 method_idxs.update(dict(
735 for name, idx in trait_methods.items()))
739 for _c in range(script_count):
742 for _c2 in range(trait_count):
746 method_body_count = u30()
747 Method = collections.namedtuple('Method', ['code', 'local_count'])
749 for _c in range(method_body_count):
753 u30() # init_scope_depth
754 u30() # max_scope_depth
756 code = read_bytes(code_length)
757 if method_idx in method_idxs:
758 m = Method(code, local_count)
759 methods[method_idxs[method_idx]] = m
760 exception_count = u30()
761 for _c2 in range(exception_count):
768 for _c2 in range(trait_count):
771 assert p + code_reader.tell() == len(code_tag)
772 assert len(methods) == len(method_idxs)
774 method_pyfunctions = {}
776 def extract_function(func_name):
777 if func_name in method_pyfunctions:
778 return method_pyfunctions[func_name]
779 if func_name not in methods:
780 raise ExtractorError(u'Cannot find function %r' % func_name)
781 m = methods[func_name]
784 registers = ['(this)'] + list(args) + [None] * m.local_count
786 coder = io.BytesIO(m.code)
788 opcode = struct.unpack('!B', coder.read(1))[0]
789 if opcode == 36: # pushbyte
790 v = struct.unpack('!B', coder.read(1))[0]
792 elif opcode == 44: # pushstring
794 stack.append(constant_strings[idx])
795 elif opcode == 48: # pushscope
796 # We don't implement the scope register, so we'll just
797 # ignore the popped value
799 elif opcode == 70: # callproperty
801 mname = multinames[index]
802 arg_count = u30(coder)
803 args = list(reversed(
804 [stack.pop() for _ in range(arg_count)]))
806 if mname == u'split':
807 assert len(args) == 1
808 assert isinstance(args[0], compat_str)
809 assert isinstance(obj, compat_str)
813 res = obj.split(args[0])
815 elif mname == u'slice':
816 assert len(args) == 1
817 assert isinstance(args[0], int)
818 assert isinstance(obj, list)
821 elif mname == u'join':
822 assert len(args) == 1
823 assert isinstance(args[0], compat_str)
824 assert isinstance(obj, list)
825 res = args[0].join(obj)
827 elif mname in method_pyfunctions:
828 stack.append(method_pyfunctions[mname](args))
830 raise NotImplementedError(
831 u'Unsupported property %r on %r'
833 elif opcode == 72: # returnvalue
836 elif opcode == 79: # callpropvoid
838 mname = multinames[index]
839 arg_count = u30(coder)
840 args = list(reversed(
841 [stack.pop() for _ in range(arg_count)]))
843 if mname == u'reverse':
844 assert isinstance(obj, list)
847 raise NotImplementedError(
848 u'Unsupported (void) property %r on %r'
850 elif opcode == 93: # findpropstrict
852 mname = multinames[index]
853 res = extract_function(mname)
855 elif opcode == 97: # setproperty
860 assert isinstance(obj, list)
861 assert isinstance(idx, int)
863 elif opcode == 98: # getlocal
865 stack.append(registers[index])
866 elif opcode == 99: # setlocal
869 registers[index] = value
870 elif opcode == 102: # getproperty
872 pname = multinames[index]
873 if pname == u'length':
875 assert isinstance(obj, list)
876 stack.append(len(obj))
877 else: # Assume attribute access
879 assert isinstance(idx, int)
881 assert isinstance(obj, list)
882 stack.append(obj[idx])
883 elif opcode == 128: # coerce
885 elif opcode == 133: # coerce_s
886 assert isinstance(stack[-1], (type(None), compat_str))
887 elif opcode == 164: # modulo
890 res = value1 % value2
892 elif opcode == 208: # getlocal_0
893 stack.append(registers[0])
894 elif opcode == 209: # getlocal_1
895 stack.append(registers[1])
896 elif opcode == 210: # getlocal_2
897 stack.append(registers[2])
898 elif opcode == 211: # getlocal_3
899 stack.append(registers[3])
900 elif opcode == 214: # setlocal_2
901 registers[2] = stack.pop()
902 elif opcode == 215: # setlocal_3
903 registers[3] = stack.pop()
905 raise NotImplementedError(
906 u'Unsupported opcode %d' % opcode)
908 method_pyfunctions[func_name] = resfunc
911 initial_function = extract_function(u'decipher')
912 return lambda s: initial_function([s])
914 def _decrypt_signature(self, s, video_id, player_url, age_gate=False):
915 """Turn the encrypted s field into a working signature"""
917 if player_url is not None:
918 if player_url.startswith(u'//'):
919 player_url = u'https:' + player_url
921 player_id = (player_url, len(s))
922 if player_id not in self._player_cache:
923 func = self._extract_signature_function(
924 video_id, player_url, len(s)
926 self._player_cache[player_id] = func
927 func = self._player_cache[player_id]
928 if self._downloader.params.get('youtube_print_sig_code'):
929 self._print_sig_code(func, len(s))
932 tb = traceback.format_exc()
933 self._downloader.report_warning(
934 u'Automatic signature extraction failed: ' + tb)
936 self._downloader.report_warning(
937 u'Warning: Falling back to static signature algorithm')
939 return self._static_decrypt_signature(
940 s, video_id, player_url, age_gate)
942 def _static_decrypt_signature(self, s, video_id, player_url, age_gate):
944 # The videos with age protection use another player, so the
945 # algorithms can be different.
947 return s[2:63] + s[82] + s[64:82] + s[63]
950 return s[86:29:-1] + s[88] + s[28:5:-1]
952 return s[25] + s[3:25] + s[0] + s[26:42] + s[79] + s[43:79] + s[91] + s[80:83]
954 return s[84:27:-1] + s[86] + s[26:5:-1]
956 return s[25] + s[3:25] + s[2] + s[26:40] + s[77] + s[41:77] + s[89] + s[78:81]
958 return s[84:78:-1] + s[87] + s[77:60:-1] + s[0] + s[59:3:-1]
960 return s[7:28] + s[87] + s[29:45] + s[55] + s[46:55] + s[2] + s[56:87] + s[28]
962 return s[6:27] + s[4] + s[28:39] + s[27] + s[40:59] + s[2] + s[60:]
964 return s[80:72:-1] + s[16] + s[71:39:-1] + s[72] + s[38:16:-1] + s[82] + s[15::-1]
966 return s[3:11] + s[0] + s[12:55] + s[84] + s[56:84]
968 return s[78:70:-1] + s[14] + s[69:37:-1] + s[70] + s[36:14:-1] + s[80] + s[:14][::-1]
970 return s[80:63:-1] + s[0] + s[62:0:-1] + s[63]
972 return s[80:37:-1] + s[7] + s[36:7:-1] + s[0] + s[6:0:-1] + s[37]
974 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]
976 return s[1:19] + s[0] + s[20:68] + s[19] + s[69:80]
978 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]
981 raise ExtractorError(u'Unable to decrypt signature, key length %d not supported; retrying might work' % (len(s)))
983 def _get_available_subtitles(self, video_id, webpage):
985 sub_list = self._download_webpage(
986 'http://video.google.com/timedtext?hl=en&type=list&v=%s' % video_id,
987 video_id, note=False)
988 except ExtractorError as err:
989 self._downloader.report_warning(u'unable to download video subtitles: %s' % compat_str(err))
991 lang_list = re.findall(r'name="([^"]*)"[^>]+lang_code="([\w\-]+)"', sub_list)
996 params = compat_urllib_parse.urlencode({
999 'fmt': self._downloader.params.get('subtitlesformat', 'srt'),
1000 'name': l[0].encode('utf-8'),
1002 url = u'http://www.youtube.com/api/timedtext?' + params
1003 sub_lang_list[lang] = url
1004 if not sub_lang_list:
1005 self._downloader.report_warning(u'video doesn\'t have subtitles')
1007 return sub_lang_list
1009 def _get_available_automatic_caption(self, video_id, webpage):
1010 """We need the webpage for getting the captions url, pass it as an
1011 argument to speed up the process."""
1012 sub_format = self._downloader.params.get('subtitlesformat', 'srt')
1013 self.to_screen(u'%s: Looking for automatic captions' % video_id)
1014 mobj = re.search(r';ytplayer.config = ({.*?});', webpage)
1015 err_msg = u'Couldn\'t find automatic captions for %s' % video_id
1017 self._downloader.report_warning(err_msg)
1019 player_config = json.loads(mobj.group(1))
1021 args = player_config[u'args']
1022 caption_url = args[u'ttsurl']
1023 timestamp = args[u'timestamp']
1024 # We get the available subtitles
1025 list_params = compat_urllib_parse.urlencode({
1030 list_url = caption_url + '&' + list_params
1031 caption_list = self._download_xml(list_url, video_id)
1032 original_lang_node = caption_list.find('track')
1033 if original_lang_node is None or original_lang_node.attrib.get('kind') != 'asr' :
1034 self._downloader.report_warning(u'Video doesn\'t have automatic captions')
1036 original_lang = original_lang_node.attrib['lang_code']
1039 for lang_node in caption_list.findall('target'):
1040 sub_lang = lang_node.attrib['lang_code']
1041 params = compat_urllib_parse.urlencode({
1042 'lang': original_lang,
1048 sub_lang_list[sub_lang] = caption_url + '&' + params
1049 return sub_lang_list
1050 # An extractor error can be raise by the download process if there are
1051 # no automatic captions but there are subtitles
1052 except (KeyError, ExtractorError):
1053 self._downloader.report_warning(err_msg)
1056 def _extract_id(self, url):
1057 mobj = re.match(self._VALID_URL, url, re.VERBOSE)
1059 raise ExtractorError(u'Invalid URL: %s' % url)
1060 video_id = mobj.group(2)
1063 def _get_video_url_list(self, url_map):
1065 Transform a dictionary in the format {itag:url} to a list of (itag, url)
1066 with the requested formats.
1068 existing_formats = [x for x in self._formats if x in url_map]
1069 if len(existing_formats) == 0:
1070 raise ExtractorError(u'no known formats available for video')
1071 video_url_list = [(f, url_map[f]) for f in existing_formats] # All formats
1072 video_url_list.reverse() # order worst to best
1073 return video_url_list
1075 def _extract_from_m3u8(self, manifest_url, video_id):
1077 def _get_urls(_manifest):
1078 lines = _manifest.split('\n')
1079 urls = filter(lambda l: l and not l.startswith('#'),
1082 manifest = self._download_webpage(manifest_url, video_id, u'Downloading formats manifest')
1083 formats_urls = _get_urls(manifest)
1084 for format_url in formats_urls:
1085 itag = self._search_regex(r'itag/(\d+?)/', format_url, 'itag')
1086 url_map[itag] = format_url
1089 def _extract_annotations(self, video_id):
1090 url = 'https://www.youtube.com/annotations_invideo?features=1&legacy=1&video_id=%s' % video_id
1091 return self._download_webpage(url, video_id, note=u'Searching for annotations.', errnote=u'Unable to download video annotations.')
1093 def _real_extract(self, url):
1094 # Extract original video URL from URL with redirection, like age verification, using next_url parameter
1095 mobj = re.search(self._NEXT_URL_RE, url)
1097 url = 'https://www.youtube.com/' + compat_urllib_parse.unquote(mobj.group(1)).lstrip('/')
1098 video_id = self._extract_id(url)
1101 url = 'https://www.youtube.com/watch?v=%s&gl=US&hl=en&has_verified=1' % video_id
1102 video_webpage = self._download_webpage(url, video_id)
1104 # Attempt to extract SWF player URL
1105 mobj = re.search(r'swfConfig.*?"(https?:\\/\\/.*?watch.*?-.*?\.swf)"', video_webpage)
1106 if mobj is not None:
1107 player_url = re.sub(r'\\(.)', r'\1', mobj.group(1))
1112 self.report_video_info_webpage_download(video_id)
1113 if re.search(r'player-age-gate-content">', video_webpage) is not None:
1114 self.report_age_confirmation()
1116 # We simulate the access to the video from www.youtube.com/v/{video_id}
1117 # this can be viewed without login into Youtube
1118 data = compat_urllib_parse.urlencode({'video_id': video_id,
1119 'el': 'player_embedded',
1122 'eurl': 'https://youtube.googleapis.com/v/' + video_id,
1126 video_info_url = 'https://www.youtube.com/get_video_info?' + data
1127 video_info_webpage = self._download_webpage(video_info_url, video_id,
1129 errnote='unable to download video info webpage')
1130 video_info = compat_parse_qs(video_info_webpage)
1133 for el_type in ['&el=embedded', '&el=detailpage', '&el=vevo', '']:
1134 video_info_url = ('https://www.youtube.com/get_video_info?&video_id=%s%s&ps=default&eurl=&gl=US&hl=en'
1135 % (video_id, el_type))
1136 video_info_webpage = self._download_webpage(video_info_url, video_id,
1138 errnote='unable to download video info webpage')
1139 video_info = compat_parse_qs(video_info_webpage)
1140 if 'token' in video_info:
1142 if 'token' not in video_info:
1143 if 'reason' in video_info:
1144 raise ExtractorError(u'YouTube said: %s' % video_info['reason'][0], expected=True)
1146 raise ExtractorError(u'"token" parameter not in video info for unknown reason')
1148 if 'view_count' in video_info:
1149 view_count = int(video_info['view_count'][0])
1153 # Check for "rental" videos
1154 if 'ypc_video_rental_bar_text' in video_info and 'author' not in video_info:
1155 raise ExtractorError(u'"rental" videos not supported')
1157 # Start extracting information
1158 self.report_information_extraction(video_id)
1161 if 'author' not in video_info:
1162 raise ExtractorError(u'Unable to extract uploader name')
1163 video_uploader = compat_urllib_parse.unquote_plus(video_info['author'][0])
1166 video_uploader_id = None
1167 mobj = re.search(r'<link itemprop="url" href="http://www.youtube.com/(?:user|channel)/([^"]+)">', video_webpage)
1168 if mobj is not None:
1169 video_uploader_id = mobj.group(1)
1171 self._downloader.report_warning(u'unable to extract uploader nickname')
1174 if 'title' in video_info:
1175 video_title = compat_urllib_parse.unquote_plus(video_info['title'][0])
1177 self._downloader.report_warning(u'Unable to extract video title')
1181 # We try first to get a high quality image:
1182 m_thumb = re.search(r'<span itemprop="thumbnail".*?href="(.*?)">',
1183 video_webpage, re.DOTALL)
1184 if m_thumb is not None:
1185 video_thumbnail = m_thumb.group(1)
1186 elif 'thumbnail_url' not in video_info:
1187 self._downloader.report_warning(u'unable to extract video thumbnail')
1188 video_thumbnail = None
1189 else: # don't panic if we can't find it
1190 video_thumbnail = compat_urllib_parse.unquote_plus(video_info['thumbnail_url'][0])
1194 mobj = re.search(r'id="eow-date.*?>(.*?)</span>', video_webpage, re.DOTALL)
1195 if mobj is not None:
1196 upload_date = ' '.join(re.sub(r'[/,-]', r' ', mobj.group(1)).split())
1197 upload_date = unified_strdate(upload_date)
1200 video_description = get_element_by_id("eow-description", video_webpage)
1201 if video_description:
1202 video_description = re.sub(r'''(?x)
1204 (?:[a-zA-Z-]+="[^"]+"\s+)*?
1206 (?:[a-zA-Z-]+="[^"]+"\s+)*?
1207 class="yt-uix-redirect-link"\s*>
1210 ''', r'\1', video_description)
1211 video_description = clean_html(video_description)
1213 fd_mobj = re.search(r'<meta name="description" content="([^"]+)"', video_webpage)
1215 video_description = unescapeHTML(fd_mobj.group(1))
1217 video_description = u''
1219 def _extract_count(klass):
1220 count = self._search_regex(
1221 r'class="%s">([\d,]+)</span>' % re.escape(klass),
1222 video_webpage, klass, default=None)
1223 if count is not None:
1224 return int(count.replace(',', ''))
1226 like_count = _extract_count(u'likes-count')
1227 dislike_count = _extract_count(u'dislikes-count')
1230 video_subtitles = self.extract_subtitles(video_id, video_webpage)
1232 if self._downloader.params.get('listsubtitles', False):
1233 self._list_available_subtitles(video_id, video_webpage)
1236 if 'length_seconds' not in video_info:
1237 self._downloader.report_warning(u'unable to extract video duration')
1238 video_duration = None
1240 video_duration = int(compat_urllib_parse.unquote_plus(video_info['length_seconds'][0]))
1243 video_annotations = None
1244 if self._downloader.params.get('writeannotations', False):
1245 video_annotations = self._extract_annotations(video_id)
1247 # Decide which formats to download
1250 mobj = re.search(r';ytplayer.config = ({.*?});', video_webpage)
1252 raise ValueError('Could not find vevo ID')
1253 info = json.loads(mobj.group(1))
1255 # Easy way to know if the 's' value is in url_encoded_fmt_stream_map
1256 # this signatures are encrypted
1257 if 'url_encoded_fmt_stream_map' not in args:
1258 raise ValueError(u'No stream_map present') # caught below
1259 re_signature = re.compile(r'[&,]s=')
1260 m_s = re_signature.search(args['url_encoded_fmt_stream_map'])
1262 self.to_screen(u'%s: Encrypted signatures detected.' % video_id)
1263 video_info['url_encoded_fmt_stream_map'] = [args['url_encoded_fmt_stream_map']]
1264 m_s = re_signature.search(args.get('adaptive_fmts', u''))
1266 if 'adaptive_fmts' in video_info:
1267 video_info['adaptive_fmts'][0] += ',' + args['adaptive_fmts']
1269 video_info['adaptive_fmts'] = [args['adaptive_fmts']]
1273 if 'conn' in video_info and video_info['conn'][0].startswith('rtmp'):
1274 self.report_rtmp_download()
1275 video_url_list = [(None, video_info['conn'][0])]
1276 elif len(video_info.get('url_encoded_fmt_stream_map', [])) >= 1 or len(video_info.get('adaptive_fmts', [])) >= 1:
1277 encoded_url_map = video_info.get('url_encoded_fmt_stream_map', [''])[0] + ',' + video_info.get('adaptive_fmts',[''])[0]
1278 if 'rtmpe%3Dyes' in encoded_url_map:
1279 raise ExtractorError('rtmpe downloads are not supported, see https://github.com/rg3/youtube-dl/issues/343 for more information.', expected=True)
1281 for url_data_str in encoded_url_map.split(','):
1282 url_data = compat_parse_qs(url_data_str)
1283 if 'itag' in url_data and 'url' in url_data:
1284 url = url_data['url'][0]
1285 if 'sig' in url_data:
1286 url += '&signature=' + url_data['sig'][0]
1287 elif 's' in url_data:
1288 encrypted_sig = url_data['s'][0]
1289 if self._downloader.params.get('verbose'):
1291 if player_url is None:
1292 player_version = 'unknown'
1294 player_version = self._search_regex(
1295 r'-(.+)\.swf$', player_url,
1296 u'flash player', fatal=False)
1297 player_desc = 'flash player %s' % player_version
1299 player_version = self._search_regex(
1300 r'html5player-(.+?)\.js', video_webpage,
1301 'html5 player', fatal=False)
1302 player_desc = u'html5 player %s' % player_version
1304 parts_sizes = u'.'.join(compat_str(len(part)) for part in encrypted_sig.split('.'))
1305 self.to_screen(u'encrypted signature length %d (%s), itag %s, %s' %
1306 (len(encrypted_sig), parts_sizes, url_data['itag'][0], player_desc))
1309 jsplayer_url_json = self._search_regex(
1310 r'"assets":.+?"js":\s*("[^"]+")',
1311 video_webpage, u'JS player URL')
1312 player_url = json.loads(jsplayer_url_json)
1314 signature = self._decrypt_signature(
1315 encrypted_sig, video_id, player_url, age_gate)
1316 url += '&signature=' + signature
1317 if 'ratebypass' not in url:
1318 url += '&ratebypass=yes'
1319 url_map[url_data['itag'][0]] = url
1320 video_url_list = self._get_video_url_list(url_map)
1321 elif video_info.get('hlsvp'):
1322 manifest_url = video_info['hlsvp'][0]
1323 url_map = self._extract_from_m3u8(manifest_url, video_id)
1324 video_url_list = self._get_video_url_list(url_map)
1326 raise ExtractorError(u'no conn, hlsvp or url_encoded_fmt_stream_map information found in video info')
1329 for itag, video_real_url in video_url_list:
1332 'url': video_real_url,
1333 'player_url': player_url,
1335 dct.update(self._formats[itag])
1338 self._sort_formats(formats)
1342 'uploader': video_uploader,
1343 'uploader_id': video_uploader_id,
1344 'upload_date': upload_date,
1345 'title': video_title,
1346 'thumbnail': video_thumbnail,
1347 'description': video_description,
1348 'subtitles': video_subtitles,
1349 'duration': video_duration,
1350 'age_limit': 18 if age_gate else 0,
1351 'annotations': video_annotations,
1352 'webpage_url': 'https://www.youtube.com/watch?v=%s' % video_id,
1353 'view_count': view_count,
1354 'like_count': like_count,
1355 'dislike_count': dislike_count,
1359 class YoutubePlaylistIE(YoutubeBaseInfoExtractor):
1360 IE_DESC = u'YouTube.com playlists'
1361 _VALID_URL = r"""(?:
1366 (?:course|view_play_list|my_playlists|artist|playlist|watch)
1367 \? (?:.*?&)*? (?:p|a|list)=
1370 ((?:PL|EC|UU|FL|RD)?[0-9A-Za-z-_]{10,})
1373 ((?:PL|EC|UU|FL|RD)[0-9A-Za-z-_]{10,})
1375 _TEMPLATE_URL = 'https://www.youtube.com/playlist?list=%s&page=%s'
1376 _MORE_PAGES_INDICATOR = r'data-link-type="next"'
1377 _VIDEO_RE = r'href="/watch\?v=(?P<id>[0-9A-Za-z_-]{11})&[^"]*?index=(?P<index>\d+)'
1378 IE_NAME = u'youtube:playlist'
1381 def suitable(cls, url):
1382 """Receives a URL and returns True if suitable for this IE."""
1383 return re.match(cls._VALID_URL, url, re.VERBOSE) is not None
1385 def _real_initialize(self):
1388 def _ids_to_results(self, ids):
1389 return [self.url_result(vid_id, 'Youtube', video_id=vid_id)
1392 def _extract_mix(self, playlist_id):
1393 # The mixes are generated from a a single video
1394 # the id of the playlist is just 'RD' + video_id
1395 url = 'https://youtube.com/watch?v=%s&list=%s' % (playlist_id[-11:], playlist_id)
1396 webpage = self._download_webpage(url, playlist_id, u'Downloading Youtube mix')
1397 title_span = (get_element_by_attribute('class', 'title long-title', webpage) or
1398 get_element_by_attribute('class', 'title ', webpage))
1399 title = clean_html(title_span)
1400 video_re = r'data-index="\d+".*?href="/watch\?v=([0-9A-Za-z_-]{11})&[^"]*?list=%s' % re.escape(playlist_id)
1401 ids = orderedSet(re.findall(video_re, webpage))
1402 url_results = self._ids_to_results(ids)
1404 return self.playlist_result(url_results, playlist_id, title)
1406 def _real_extract(self, url):
1407 # Extract playlist id
1408 mobj = re.match(self._VALID_URL, url, re.VERBOSE)
1410 raise ExtractorError(u'Invalid URL: %s' % url)
1411 playlist_id = mobj.group(1) or mobj.group(2)
1413 # Check if it's a video-specific URL
1414 query_dict = compat_urlparse.parse_qs(compat_urlparse.urlparse(url).query)
1415 if 'v' in query_dict:
1416 video_id = query_dict['v'][0]
1417 if self._downloader.params.get('noplaylist'):
1418 self.to_screen(u'Downloading just video %s because of --no-playlist' % video_id)
1419 return self.url_result(video_id, 'Youtube', video_id=video_id)
1421 self.to_screen(u'Downloading playlist PL%s - add --no-playlist to just download video %s' % (playlist_id, video_id))
1423 if playlist_id.startswith('RD'):
1424 # Mixes require a custom extraction process
1425 return self._extract_mix(playlist_id)
1426 if playlist_id.startswith('TL'):
1427 raise ExtractorError(u'For downloading YouTube.com top lists, use '
1428 u'the "yttoplist" keyword, for example "youtube-dl \'yttoplist:music:Top Tracks\'"', expected=True)
1430 # Extract the video ids from the playlist pages
1433 for page_num in itertools.count(1):
1434 url = self._TEMPLATE_URL % (playlist_id, page_num)
1435 page = self._download_webpage(url, playlist_id, u'Downloading page #%s' % page_num)
1436 matches = re.finditer(self._VIDEO_RE, page)
1437 # We remove the duplicates and the link with index 0
1438 # (it's not the first video of the playlist)
1439 new_ids = orderedSet(m.group('id') for m in matches if m.group('index') != '0')
1442 if re.search(self._MORE_PAGES_INDICATOR, page) is None:
1445 playlist_title = self._og_search_title(page)
1447 url_results = self._ids_to_results(ids)
1448 return self.playlist_result(url_results, playlist_id, playlist_title)
1451 class YoutubeTopListIE(YoutubePlaylistIE):
1452 IE_NAME = u'youtube:toplist'
1453 IE_DESC = (u'YouTube.com top lists, "yttoplist:{channel}:{list title}"'
1454 u' (Example: "yttoplist:music:Top Tracks")')
1455 _VALID_URL = r'yttoplist:(?P<chann>.*?):(?P<title>.*?)$'
1457 def _real_extract(self, url):
1458 mobj = re.match(self._VALID_URL, url)
1459 channel = mobj.group('chann')
1460 title = mobj.group('title')
1461 query = compat_urllib_parse.urlencode({'title': title})
1462 playlist_re = 'href="([^"]+?%s[^"]+?)"' % re.escape(query)
1463 channel_page = self._download_webpage('https://www.youtube.com/%s' % channel, title)
1464 link = self._html_search_regex(playlist_re, channel_page, u'list')
1465 url = compat_urlparse.urljoin('https://www.youtube.com/', link)
1467 video_re = r'data-index="\d+".*?data-video-id="([0-9A-Za-z_-]{11})"'
1469 # sometimes the webpage doesn't contain the videos
1470 # retry until we get them
1471 for i in itertools.count(0):
1472 msg = u'Downloading Youtube mix'
1474 msg += ', retry #%d' % i
1475 webpage = self._download_webpage(url, title, msg)
1476 ids = orderedSet(re.findall(video_re, webpage))
1479 url_results = self._ids_to_results(ids)
1480 return self.playlist_result(url_results, playlist_title=title)
1483 class YoutubeChannelIE(InfoExtractor):
1484 IE_DESC = u'YouTube.com channels'
1485 _VALID_URL = r"^(?:https?://)?(?:youtu\.be|(?:\w+\.)?youtube(?:-nocookie)?\.com)/channel/([0-9A-Za-z_-]+)"
1486 _MORE_PAGES_INDICATOR = 'yt-uix-load-more'
1487 _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'
1488 IE_NAME = u'youtube:channel'
1490 def extract_videos_from_page(self, page):
1492 for mobj in re.finditer(r'href="/watch\?v=([0-9A-Za-z_-]+)&?', page):
1493 if mobj.group(1) not in ids_in_page:
1494 ids_in_page.append(mobj.group(1))
1497 def _real_extract(self, url):
1498 # Extract channel id
1499 mobj = re.match(self._VALID_URL, url)
1501 raise ExtractorError(u'Invalid URL: %s' % url)
1503 # Download channel page
1504 channel_id = mobj.group(1)
1506 url = 'https://www.youtube.com/channel/%s/videos' % channel_id
1507 channel_page = self._download_webpage(url, channel_id)
1508 autogenerated = re.search(r'''(?x)
1510 channel-header-autogenerated-label|
1511 yt-channel-title-autogenerated
1512 )[^"]*"''', channel_page) is not None
1515 # The videos are contained in a single page
1516 # the ajax pages can't be used, they are empty
1517 video_ids = self.extract_videos_from_page(channel_page)
1519 # Download all channel pages using the json-based channel_ajax query
1520 for pagenum in itertools.count(1):
1521 url = self._MORE_PAGES_URL % (pagenum, channel_id)
1522 page = self._download_webpage(url, channel_id,
1523 u'Downloading page #%s' % pagenum)
1525 page = json.loads(page)
1527 ids_in_page = self.extract_videos_from_page(page['content_html'])
1528 video_ids.extend(ids_in_page)
1530 if self._MORE_PAGES_INDICATOR not in page['load_more_widget_html']:
1533 self._downloader.to_screen(u'[youtube] Channel %s: Found %i videos' % (channel_id, len(video_ids)))
1535 url_entries = [self.url_result(video_id, 'Youtube', video_id=video_id)
1536 for video_id in video_ids]
1537 return self.playlist_result(url_entries, channel_id)
1540 class YoutubeUserIE(InfoExtractor):
1541 IE_DESC = u'YouTube.com user videos (URL or "ytuser" keyword)'
1542 _VALID_URL = r'(?:(?:(?:https?://)?(?:\w+\.)?youtube\.com/(?:user/)?(?!(?:attribution_link|watch)(?:$|[^a-z_A-Z0-9-])))|ytuser:)(?!feed/)([A-Za-z0-9_-]+)'
1543 _TEMPLATE_URL = 'http://gdata.youtube.com/feeds/api/users/%s'
1544 _GDATA_PAGE_SIZE = 50
1545 _GDATA_URL = 'http://gdata.youtube.com/feeds/api/users/%s/uploads?max-results=%d&start-index=%d&alt=json'
1546 IE_NAME = u'youtube:user'
1549 def suitable(cls, url):
1550 # Don't return True if the url can be extracted with other youtube
1551 # extractor, the regex would is too permissive and it would match.
1552 other_ies = iter(klass for (name, klass) in globals().items() if name.endswith('IE') and klass is not cls)
1553 if any(ie.suitable(url) for ie in other_ies): return False
1554 else: return super(YoutubeUserIE, cls).suitable(url)
1556 def _real_extract(self, url):
1558 mobj = re.match(self._VALID_URL, url)
1560 raise ExtractorError(u'Invalid URL: %s' % url)
1562 username = mobj.group(1)
1564 # Download video ids using YouTube Data API. Result size per
1565 # query is limited (currently to 50 videos) so we need to query
1566 # page by page until there are no video ids - it means we got
1571 for pagenum in itertools.count(0):
1572 start_index = pagenum * self._GDATA_PAGE_SIZE + 1
1574 gdata_url = self._GDATA_URL % (username, self._GDATA_PAGE_SIZE, start_index)
1575 page = self._download_webpage(gdata_url, username,
1576 u'Downloading video ids from %d to %d' % (start_index, start_index + self._GDATA_PAGE_SIZE))
1579 response = json.loads(page)
1580 except ValueError as err:
1581 raise ExtractorError(u'Invalid JSON in API response: ' + compat_str(err))
1582 if 'entry' not in response['feed']:
1583 # Number of videos is a multiple of self._MAX_RESULTS
1586 # Extract video identifiers
1587 entries = response['feed']['entry']
1588 for entry in entries:
1589 title = entry['title']['$t']
1590 video_id = entry['id']['$t'].split('/')[-1]
1591 url_results.append({
1594 'ie_key': 'Youtube',
1599 # A little optimization - if current page is not
1600 # "full", ie. does not contain PAGE_SIZE video ids then
1601 # we can assume that this page is the last one - there
1602 # are no more ids on further pages - no need to query
1605 if len(entries) < self._GDATA_PAGE_SIZE:
1608 return self.playlist_result(url_results, playlist_title=username)
1611 class YoutubeSearchIE(SearchInfoExtractor):
1612 IE_DESC = u'YouTube.com searches'
1613 _API_URL = 'https://gdata.youtube.com/feeds/api/videos?q=%s&start-index=%i&max-results=50&v=2&alt=jsonc'
1615 IE_NAME = u'youtube:search'
1616 _SEARCH_KEY = 'ytsearch'
1618 def _get_n_results(self, query, n):
1619 """Get a specified number of results for a query"""
1625 while (50 * pagenum) < limit:
1626 result_url = self._API_URL % (compat_urllib_parse.quote_plus(query), (50*pagenum)+1)
1627 data_json = self._download_webpage(
1628 result_url, video_id=u'query "%s"' % query,
1629 note=u'Downloading page %s' % (pagenum + 1),
1630 errnote=u'Unable to download API page')
1631 data = json.loads(data_json)
1632 api_response = data['data']
1634 if 'items' not in api_response:
1635 raise ExtractorError(u'[youtube] No video results')
1637 new_ids = list(video['id'] for video in api_response['items'])
1638 video_ids += new_ids
1640 limit = min(n, api_response['totalItems'])
1643 if len(video_ids) > n:
1644 video_ids = video_ids[:n]
1645 videos = [self.url_result(video_id, 'Youtube', video_id=video_id)
1646 for video_id in video_ids]
1647 return self.playlist_result(videos, query)
1649 class YoutubeSearchDateIE(YoutubeSearchIE):
1650 IE_NAME = YoutubeSearchIE.IE_NAME + ':date'
1651 _API_URL = 'https://gdata.youtube.com/feeds/api/videos?q=%s&start-index=%i&max-results=50&v=2&alt=jsonc&orderby=published'
1652 _SEARCH_KEY = 'ytsearchdate'
1653 IE_DESC = u'YouTube.com searches, newest videos first'
1655 class YoutubeShowIE(InfoExtractor):
1656 IE_DESC = u'YouTube.com (multi-season) shows'
1657 _VALID_URL = r'https?://www\.youtube\.com/show/(.*)'
1658 IE_NAME = u'youtube:show'
1660 def _real_extract(self, url):
1661 mobj = re.match(self._VALID_URL, url)
1662 show_name = mobj.group(1)
1663 webpage = self._download_webpage(url, show_name, u'Downloading show webpage')
1664 # There's one playlist for each season of the show
1665 m_seasons = list(re.finditer(r'href="(/playlist\?list=.*?)"', webpage))
1666 self.to_screen(u'%s: Found %s seasons' % (show_name, len(m_seasons)))
1667 return [self.url_result('https://www.youtube.com' + season.group(1), 'YoutubePlaylist') for season in m_seasons]
1670 class YoutubeFeedsInfoExtractor(YoutubeBaseInfoExtractor):
1672 Base class for extractors that fetch info from
1673 http://www.youtube.com/feed_ajax
1674 Subclasses must define the _FEED_NAME and _PLAYLIST_TITLE properties.
1676 _LOGIN_REQUIRED = True
1677 # use action_load_personal_feed instead of action_load_system_feed
1678 _PERSONAL_FEED = False
1681 def _FEED_TEMPLATE(self):
1682 action = 'action_load_system_feed'
1683 if self._PERSONAL_FEED:
1684 action = 'action_load_personal_feed'
1685 return 'http://www.youtube.com/feed_ajax?%s=1&feed_name=%s&paging=%%s' % (action, self._FEED_NAME)
1689 return u'youtube:%s' % self._FEED_NAME
1691 def _real_initialize(self):
1694 def _real_extract(self, url):
1697 for i in itertools.count(1):
1698 info = self._download_webpage(self._FEED_TEMPLATE % paging,
1699 u'%s feed' % self._FEED_NAME,
1700 u'Downloading page %s' % i)
1701 info = json.loads(info)
1702 feed_html = info['feed_html']
1703 m_ids = re.finditer(r'"/watch\?v=(.*?)["&]', feed_html)
1704 ids = orderedSet(m.group(1) for m in m_ids)
1705 feed_entries.extend(
1706 self.url_result(video_id, 'Youtube', video_id=video_id)
1707 for video_id in ids)
1708 if info['paging'] is None:
1710 paging = info['paging']
1711 return self.playlist_result(feed_entries, playlist_title=self._PLAYLIST_TITLE)
1713 class YoutubeSubscriptionsIE(YoutubeFeedsInfoExtractor):
1714 IE_DESC = u'YouTube.com subscriptions feed, "ytsubs" keyword(requires authentication)'
1715 _VALID_URL = r'https?://www\.youtube\.com/feed/subscriptions|:ytsubs(?:criptions)?'
1716 _FEED_NAME = 'subscriptions'
1717 _PLAYLIST_TITLE = u'Youtube Subscriptions'
1719 class YoutubeRecommendedIE(YoutubeFeedsInfoExtractor):
1720 IE_DESC = u'YouTube.com recommended videos, "ytrec" keyword (requires authentication)'
1721 _VALID_URL = r'https?://www\.youtube\.com/feed/recommended|:ytrec(?:ommended)?'
1722 _FEED_NAME = 'recommended'
1723 _PLAYLIST_TITLE = u'Youtube Recommended videos'
1725 class YoutubeWatchLaterIE(YoutubeFeedsInfoExtractor):
1726 IE_DESC = u'Youtube watch later list, "ytwatchlater" keyword (requires authentication)'
1727 _VALID_URL = r'https?://www\.youtube\.com/feed/watch_later|:ytwatchlater'
1728 _FEED_NAME = 'watch_later'
1729 _PLAYLIST_TITLE = u'Youtube Watch Later'
1730 _PERSONAL_FEED = True
1732 class YoutubeHistoryIE(YoutubeFeedsInfoExtractor):
1733 IE_DESC = u'Youtube watch history, "ythistory" keyword (requires authentication)'
1734 _VALID_URL = u'https?://www\.youtube\.com/feed/history|:ythistory'
1735 _FEED_NAME = 'history'
1736 _PERSONAL_FEED = True
1737 _PLAYLIST_TITLE = u'Youtube Watch History'
1739 class YoutubeFavouritesIE(YoutubeBaseInfoExtractor):
1740 IE_NAME = u'youtube:favorites'
1741 IE_DESC = u'YouTube.com favourite videos, "ytfav" keyword (requires authentication)'
1742 _VALID_URL = r'https?://www\.youtube\.com/my_favorites|:ytfav(?:ou?rites)?'
1743 _LOGIN_REQUIRED = True
1745 def _real_extract(self, url):
1746 webpage = self._download_webpage('https://www.youtube.com/my_favorites', 'Youtube Favourites videos')
1747 playlist_id = self._search_regex(r'list=(.+?)["&]', webpage, u'favourites playlist id')
1748 return self.url_result(playlist_id, 'YoutubePlaylist')
1751 class YoutubeTruncatedURLIE(InfoExtractor):
1752 IE_NAME = 'youtube:truncated_url'
1753 IE_DESC = False # Do not list
1754 _VALID_URL = r'(?:https?://)?[^/]+/watch\?feature=[a-z_]+$'
1756 def _real_extract(self, url):
1757 raise ExtractorError(
1758 u'Did you forget to quote the URL? Remember that & is a meta '
1759 u'character in most shells, so you want to put the URL in quotes, '
1761 u'\'http://www.youtube.com/watch?feature=foo&v=BaW_jenozKc\''
1762 u' (or simply youtube-dl BaW_jenozKc ).',