15 from .common import InfoExtractor, SearchInfoExtractor
16 from .subtitles import SubtitlesInfoExtractor
21 compat_urllib_request,
28 get_element_by_attribute,
39 class YoutubeBaseInfoExtractor(InfoExtractor):
40 """Provide base functions for Youtube extractors"""
41 _LOGIN_URL = 'https://accounts.google.com/ServiceLogin'
42 _LANG_URL = r'https://www.youtube.com/?hl=en&persist_hl=1&gl=US&persist_gl=1&opt_out_ackd=1'
43 _AGE_URL = 'http://www.youtube.com/verify_age?next_url=/&gl=US&hl=en'
44 _NETRC_MACHINE = 'youtube'
45 # If True it will raise an error if no login info is provided
46 _LOGIN_REQUIRED = False
48 def _set_language(self):
49 return bool(self._download_webpage(
51 note=u'Setting language', errnote='unable to set language',
55 (username, password) = self._get_login_info()
56 # No authentication to be performed
58 if self._LOGIN_REQUIRED:
59 raise ExtractorError(u'No login info available, needed for using %s.' % self.IE_NAME, expected=True)
62 login_page = self._download_webpage(
63 self._LOGIN_URL, None,
64 note=u'Downloading login page',
65 errnote=u'unable to fetch login page', fatal=False)
66 if login_page is False:
69 galx = self._search_regex(r'(?s)<input.+?name="GALX".+?value="(.+?)"',
70 login_page, u'Login GALX parameter')
74 u'continue': u'https://www.youtube.com/signin?action_handle_signin=true&feature=sign_in_button&hl=en_US&nomobiletemp=1',
78 u'PersistentCookie': u'yes',
80 u'bgresponse': u'js_disabled',
81 u'checkConnection': u'',
82 u'checkedDomains': u'youtube',
87 u'signIn': u'Sign in',
89 u'service': u'youtube',
93 # Convert to UTF-8 *before* urlencode because Python 2.x's urlencode
95 login_form = dict((k.encode('utf-8'), v.encode('utf-8')) for k,v in login_form_strs.items())
96 login_data = compat_urllib_parse.urlencode(login_form).encode('ascii')
98 req = compat_urllib_request.Request(self._LOGIN_URL, login_data)
99 login_results = self._download_webpage(
101 note=u'Logging in', errnote=u'unable to log in', fatal=False)
102 if login_results is False:
104 if re.search(r'(?i)<form[^>]* id="gaia_loginform"', login_results) is not None:
105 self._downloader.report_warning(u'unable to log in: bad username or password')
109 def _confirm_age(self):
112 'action_confirm': 'Confirm',
114 req = compat_urllib_request.Request(self._AGE_URL, compat_urllib_parse.urlencode(age_form))
116 self._download_webpage(
118 note=u'Confirming age', errnote=u'Unable to confirm age')
121 def _real_initialize(self):
122 if self._downloader is None:
124 if not self._set_language():
126 if not self._login():
131 class YoutubeIE(YoutubeBaseInfoExtractor, SubtitlesInfoExtractor):
132 IE_DESC = u'YouTube.com'
133 _VALID_URL = r"""(?x)^
135 (?:https?://|//)? # http(s):// or protocol-independent URL (optional)
136 (?:(?:(?:(?:\w+\.)?[yY][oO][uU][tT][uU][bB][eE](?:-nocookie)?\.com/|
137 (?:www\.)?deturl\.com/www\.youtube\.com/|
138 (?:www\.)?pwnyoutube\.com|
139 tube\.majestyc\.net/|
140 youtube\.googleapis\.com/) # the various hostnames, with wildcard subdomains
141 (?:.*?\#/)? # handle anchor (#/) redirect urls
142 (?: # the various things that can precede the ID:
143 (?:(?:v|embed|e)/) # v/ or embed/ or e/
144 |(?: # or the v= param in all its forms
145 (?:(?:watch|movie)(?:_popup)?(?:\.php)?)? # preceding watch(_popup|.php) or nothing (like /?v=xxxx)
146 (?:\?|\#!?) # the params delimiter ? or # or #!
147 (?:.*?&)? # any other preceding param (like /?s=tuff&v=xxxx)
151 |youtu\.be/ # just youtu.be/xxxx
153 )? # all until now is optional -> you can pass the naked ID
154 ([0-9A-Za-z_-]{11}) # here is it! the YouTube video ID
155 (?(1).+)? # if we found the ID, everything can follow
157 _NEXT_URL_RE = r'[\?&]next_url=([^&]+)'
159 '5': {'ext': 'flv', 'width': 400, 'height': 240},
160 '6': {'ext': 'flv', 'width': 450, 'height': 270},
161 '13': {'ext': '3gp'},
162 '17': {'ext': '3gp', 'width': 176, 'height': 144},
163 '18': {'ext': 'mp4', 'width': 640, 'height': 360},
164 '22': {'ext': 'mp4', 'width': 1280, 'height': 720},
165 '34': {'ext': 'flv', 'width': 640, 'height': 360},
166 '35': {'ext': 'flv', 'width': 854, 'height': 480},
167 '36': {'ext': '3gp', 'width': 320, 'height': 240},
168 '37': {'ext': 'mp4', 'width': 1920, 'height': 1080},
169 '38': {'ext': 'mp4', 'width': 4096, 'height': 3072},
170 '43': {'ext': 'webm', 'width': 640, 'height': 360},
171 '44': {'ext': 'webm', 'width': 854, 'height': 480},
172 '45': {'ext': 'webm', 'width': 1280, 'height': 720},
173 '46': {'ext': 'webm', 'width': 1920, 'height': 1080},
177 '82': {'ext': 'mp4', 'height': 360, 'resolution': '360p', 'format_note': '3D', 'preference': -20},
178 '83': {'ext': 'mp4', 'height': 480, 'resolution': '480p', 'format_note': '3D', 'preference': -20},
179 '84': {'ext': 'mp4', 'height': 720, 'resolution': '720p', 'format_note': '3D', 'preference': -20},
180 '85': {'ext': 'mp4', 'height': 1080, 'resolution': '1080p', 'format_note': '3D', 'preference': -20},
181 '100': {'ext': 'webm', 'height': 360, 'resolution': '360p', 'format_note': '3D', 'preference': -20},
182 '101': {'ext': 'webm', 'height': 480, 'resolution': '480p', 'format_note': '3D', 'preference': -20},
183 '102': {'ext': 'webm', 'height': 720, 'resolution': '720p', 'format_note': '3D', 'preference': -20},
185 # Apple HTTP Live Streaming
186 '92': {'ext': 'mp4', 'height': 240, 'resolution': '240p', 'format_note': 'HLS', 'preference': -10},
187 '93': {'ext': 'mp4', 'height': 360, 'resolution': '360p', 'format_note': 'HLS', 'preference': -10},
188 '94': {'ext': 'mp4', 'height': 480, 'resolution': '480p', 'format_note': 'HLS', 'preference': -10},
189 '95': {'ext': 'mp4', 'height': 720, 'resolution': '720p', 'format_note': 'HLS', 'preference': -10},
190 '96': {'ext': 'mp4', 'height': 1080, 'resolution': '1080p', 'format_note': 'HLS', 'preference': -10},
191 '132': {'ext': 'mp4', 'height': 240, 'resolution': '240p', 'format_note': 'HLS', 'preference': -10},
192 '151': {'ext': 'mp4', 'height': 72, 'resolution': '72p', 'format_note': 'HLS', 'preference': -10},
195 '133': {'ext': 'mp4', 'height': 240, 'resolution': '240p', 'format_note': 'DASH video', 'preference': -40},
196 '134': {'ext': 'mp4', 'height': 360, 'resolution': '360p', 'format_note': 'DASH video', 'preference': -40},
197 '135': {'ext': 'mp4', 'height': 480, 'resolution': '480p', 'format_note': 'DASH video', 'preference': -40},
198 '136': {'ext': 'mp4', 'height': 720, 'resolution': '720p', 'format_note': 'DASH video', 'preference': -40},
199 '137': {'ext': 'mp4', 'height': 1080, 'resolution': '1080p', 'format_note': 'DASH video', 'preference': -40},
200 '138': {'ext': 'mp4', 'height': 1081, 'resolution': '>1080p', 'format_note': 'DASH video', 'preference': -40},
201 '160': {'ext': 'mp4', 'height': 192, 'resolution': '192p', 'format_note': 'DASH video', 'preference': -40},
202 '264': {'ext': 'mp4', 'height': 1080, 'resolution': '1080p', 'format_note': 'DASH video', 'preference': -40},
205 '139': {'ext': 'm4a', 'format_note': 'DASH audio', 'vcodec': 'none', 'abr': 48, 'preference': -50},
206 '140': {'ext': 'm4a', 'format_note': 'DASH audio', 'vcodec': 'none', 'abr': 128, 'preference': -50},
207 '141': {'ext': 'm4a', 'format_note': 'DASH audio', 'vcodec': 'none', 'abr': 256, 'preference': -50},
210 '167': {'ext': 'webm', 'height': 360, 'width': 640, 'format_note': 'DASH video', 'container': 'webm', 'vcodec': 'VP8', 'acodec': 'none', 'preference': -40},
211 '168': {'ext': 'webm', 'height': 480, 'width': 854, 'format_note': 'DASH video', 'container': 'webm', 'vcodec': 'VP8', 'acodec': 'none', 'preference': -40},
212 '168': {'ext': 'webm', 'height': 1080, 'width': 1920, 'format_note': 'DASH video', 'container': 'webm', 'vcodec': 'VP8', 'acodec': 'none', 'preference': -40},
213 '218': {'ext': 'webm', 'height': 480, 'width': 854, 'format_note': 'DASH video', 'container': 'webm', 'vcodec': 'VP8', 'acodec': 'none', 'preference': -40},
214 '219': {'ext': 'webm', 'height': 480, 'width': 854, 'format_note': 'DASH video', 'container': 'webm', 'vcodec': 'VP8', 'acodec': 'none', 'preference': -40},
215 '242': {'ext': 'webm', 'height': 240, 'resolution': '240p', 'format_note': 'DASH webm', 'preference': -40},
216 '243': {'ext': 'webm', 'height': 360, 'resolution': '360p', 'format_note': 'DASH webm', 'preference': -40},
217 '244': {'ext': 'webm', 'height': 480, 'resolution': '480p', 'format_note': 'DASH webm', 'preference': -40},
218 '245': {'ext': 'webm', 'height': 480, 'resolution': '480p', 'format_note': 'DASH webm', 'preference': -40},
219 '246': {'ext': 'webm', 'height': 480, 'resolution': '480p', 'format_note': 'DASH webm', 'preference': -40},
220 '247': {'ext': 'webm', 'height': 720, 'resolution': '720p', 'format_note': 'DASH webm', 'preference': -40},
221 '248': {'ext': 'webm', 'height': 1080, 'resolution': '1080p', 'format_note': 'DASH webm', 'preference': -40},
224 '171': {'ext': 'webm', 'vcodec': 'none', 'format_note': 'DASH webm audio', 'abr': 48, 'preference': -50},
225 '172': {'ext': 'webm', 'vcodec': 'none', 'format_note': 'DASH webm audio', 'abr': 256, 'preference': -50},
228 '_rtmp': {'protocol': 'rtmp'},
234 u"url": u"http://www.youtube.com/watch?v=BaW_jenozKc",
235 u"file": u"BaW_jenozKc.mp4",
237 u"title": u"youtube-dl test video \"'/\\ä↭𝕐",
238 u"uploader": u"Philipp Hagemeister",
239 u"uploader_id": u"phihag",
240 u"upload_date": u"20121002",
241 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 ."
245 u"url": u"http://www.youtube.com/watch?v=UxxajLWwzqY",
246 u"file": u"UxxajLWwzqY.mp4",
247 u"note": u"Test generic use_cipher_signature video (#897)",
249 u"upload_date": u"20120506",
250 u"title": u"Icona Pop - I Love It (feat. Charli XCX) [OFFICIAL VIDEO]",
251 u"description": u"md5:5b292926389560516e384ac437c0ec07",
252 u"uploader": u"Icona Pop",
253 u"uploader_id": u"IconaPop"
257 u"url": u"https://www.youtube.com/watch?v=07FYdnEawAQ",
258 u"file": u"07FYdnEawAQ.mp4",
259 u"note": u"Test VEVO video with age protection (#956)",
261 u"upload_date": u"20130703",
262 u"title": u"Justin Timberlake - Tunnel Vision (Explicit)",
263 u"description": u"md5:64249768eec3bc4276236606ea996373",
264 u"uploader": u"justintimberlakeVEVO",
265 u"uploader_id": u"justintimberlakeVEVO"
269 u"url": u"//www.YouTube.com/watch?v=yZIXLfi8CZQ",
270 u"file": u"yZIXLfi8CZQ.mp4",
271 u"note": u"Embed-only video (#1746)",
273 u"upload_date": u"20120608",
274 u"title": u"Principal Sexually Assaults A Teacher - Episode 117 - 8th June 2012",
275 u"description": u"md5:09b78bd971f1e3e289601dfba15ca4f7",
276 u"uploader": u"SET India",
277 u"uploader_id": u"setindia"
281 u"url": u"http://www.youtube.com/watch?v=a9LDPn-MO4I",
282 u"file": u"a9LDPn-MO4I.m4a",
283 u"note": u"256k DASH audio (format 141) via DASH manifest",
285 u"upload_date": "20121002",
286 u"uploader_id": "8KVIDEO",
287 u"description": "No description available.",
288 u"uploader": "8KVIDEO",
289 u"title": "UHDTV TEST 8K VIDEO.mp4"
292 u"youtube_include_dash_manifest": True,
300 def suitable(cls, url):
301 """Receives a URL and returns True if suitable for this IE."""
302 if YoutubePlaylistIE.suitable(url): return False
303 return re.match(cls._VALID_URL, url) is not None
305 def __init__(self, *args, **kwargs):
306 super(YoutubeIE, self).__init__(*args, **kwargs)
307 self._player_cache = {}
309 def report_video_info_webpage_download(self, video_id):
310 """Report attempt to download video info webpage."""
311 self.to_screen(u'%s: Downloading video info webpage' % video_id)
313 def report_information_extraction(self, video_id):
314 """Report attempt to extract video information."""
315 self.to_screen(u'%s: Extracting video information' % video_id)
317 def report_unavailable_format(self, video_id, format):
318 """Report extracted video URL."""
319 self.to_screen(u'%s: Format %s not available' % (video_id, format))
321 def report_rtmp_download(self):
322 """Indicate the download will use the RTMP protocol."""
323 self.to_screen(u'RTMP download detected')
325 def _extract_signature_function(self, video_id, player_url, slen):
326 id_m = re.match(r'.*-(?P<id>[a-zA-Z0-9_-]+)\.(?P<ext>[a-z]+)$',
328 player_type = id_m.group('ext')
329 player_id = id_m.group('id')
331 # Read from filesystem cache
332 func_id = '%s_%s_%d' % (player_type, player_id, slen)
333 assert os.path.basename(func_id) == func_id
334 cache_dir = get_cachedir(self._downloader.params)
336 cache_enabled = cache_dir is not None
338 cache_fn = os.path.join(os.path.expanduser(cache_dir),
342 with io.open(cache_fn, 'r', encoding='utf-8') as cachef:
343 cache_spec = json.load(cachef)
344 return lambda s: u''.join(s[i] for i in cache_spec)
346 pass # No cache available
348 if player_type == 'js':
349 code = self._download_webpage(
350 player_url, video_id,
351 note=u'Downloading %s player %s' % (player_type, player_id),
352 errnote=u'Download of %s failed' % player_url)
353 res = self._parse_sig_js(code)
354 elif player_type == 'swf':
355 urlh = self._request_webpage(
356 player_url, video_id,
357 note=u'Downloading %s player %s' % (player_type, player_id),
358 errnote=u'Download of %s failed' % player_url)
360 res = self._parse_sig_swf(code)
362 assert False, 'Invalid player type %r' % player_type
366 test_string = u''.join(map(compat_chr, range(slen)))
367 cache_res = res(test_string)
368 cache_spec = [ord(c) for c in cache_res]
370 os.makedirs(os.path.dirname(cache_fn))
371 except OSError as ose:
372 if ose.errno != errno.EEXIST:
374 write_json_file(cache_spec, cache_fn)
376 tb = traceback.format_exc()
377 self._downloader.report_warning(
378 u'Writing cache to %r failed: %s' % (cache_fn, tb))
382 def _print_sig_code(self, func, slen):
383 def gen_sig_code(idxs):
384 def _genslice(start, end, step):
385 starts = u'' if start == 0 else str(start)
386 ends = (u':%d' % (end+step)) if end + step >= 0 else u':'
387 steps = u'' if step == 1 else (u':%d' % step)
388 return u's[%s%s%s]' % (starts, ends, steps)
391 start = '(Never used)' # Quelch pyflakes warnings - start will be
392 # set as soon as step is set
393 for i, prev in zip(idxs[1:], idxs[:-1]):
397 yield _genslice(start, prev, step)
400 if i - prev in [-1, 1]:
405 yield u's[%d]' % prev
409 yield _genslice(start, i, step)
411 test_string = u''.join(map(compat_chr, range(slen)))
412 cache_res = func(test_string)
413 cache_spec = [ord(c) for c in cache_res]
414 expr_code = u' + '.join(gen_sig_code(cache_spec))
415 code = u'if len(s) == %d:\n return %s\n' % (slen, expr_code)
416 self.to_screen(u'Extracted signature function:\n' + code)
418 def _parse_sig_js(self, jscode):
419 funcname = self._search_regex(
420 r'signature=([a-zA-Z]+)', jscode,
421 u'Initial JS player signature function name')
426 return string.lowercase.index(varname)
428 def interpret_statement(stmt, local_vars, allow_recursion=20):
429 if allow_recursion < 0:
430 raise ExtractorError(u'Recursion limit reached')
432 if stmt.startswith(u'var '):
433 stmt = stmt[len(u'var '):]
434 ass_m = re.match(r'^(?P<out>[a-z]+)(?:\[(?P<index>[^\]]+)\])?' +
435 r'=(?P<expr>.*)$', stmt)
437 if ass_m.groupdict().get('index'):
439 lvar = local_vars[ass_m.group('out')]
440 idx = interpret_expression(ass_m.group('index'),
441 local_vars, allow_recursion)
442 assert isinstance(idx, int)
445 expr = ass_m.group('expr')
448 local_vars[ass_m.group('out')] = val
450 expr = ass_m.group('expr')
451 elif stmt.startswith(u'return '):
453 expr = stmt[len(u'return '):]
455 raise ExtractorError(
456 u'Cannot determine left side of statement in %r' % stmt)
458 v = interpret_expression(expr, local_vars, allow_recursion)
461 def interpret_expression(expr, local_vars, allow_recursion):
466 return local_vars[expr]
468 m = re.match(r'^(?P<in>[a-z]+)\.(?P<member>.*)$', expr)
470 member = m.group('member')
471 val = local_vars[m.group('in')]
472 if member == 'split("")':
474 if member == 'join("")':
476 if member == 'length':
478 if member == 'reverse()':
480 slice_m = re.match(r'slice\((?P<idx>.*)\)', member)
482 idx = interpret_expression(
483 slice_m.group('idx'), local_vars, allow_recursion-1)
487 r'^(?P<in>[a-z]+)\[(?P<idx>.+)\]$', expr)
489 val = local_vars[m.group('in')]
490 idx = interpret_expression(m.group('idx'), local_vars,
494 m = re.match(r'^(?P<a>.+?)(?P<op>[%])(?P<b>.+?)$', expr)
496 a = interpret_expression(m.group('a'),
497 local_vars, allow_recursion)
498 b = interpret_expression(m.group('b'),
499 local_vars, allow_recursion)
503 r'^(?P<func>[a-zA-Z]+)\((?P<args>[a-z0-9,]+)\)$', expr)
505 fname = m.group('func')
506 if fname not in functions:
507 functions[fname] = extract_function(fname)
508 argvals = [int(v) if v.isdigit() else local_vars[v]
509 for v in m.group('args').split(',')]
510 return functions[fname](argvals)
511 raise ExtractorError(u'Unsupported JS expression %r' % expr)
513 def extract_function(funcname):
515 r'function ' + re.escape(funcname) +
516 r'\((?P<args>[a-z,]+)\){(?P<code>[^}]+)}',
518 argnames = func_m.group('args').split(',')
521 local_vars = dict(zip(argnames, args))
522 for stmt in func_m.group('code').split(';'):
523 res = interpret_statement(stmt, local_vars)
527 initial_function = extract_function(funcname)
528 return lambda s: initial_function([s])
530 def _parse_sig_swf(self, file_contents):
531 if file_contents[1:3] != b'WS':
532 raise ExtractorError(
533 u'Not an SWF file; header is %r' % file_contents[:3])
534 if file_contents[:1] == b'C':
535 content = zlib.decompress(file_contents[8:])
537 raise NotImplementedError(u'Unsupported compression format %r' %
540 def extract_tags(content):
542 while pos < len(content):
543 header16 = struct.unpack('<H', content[pos:pos+2])[0]
545 tag_code = header16 >> 6
546 tag_len = header16 & 0x3f
548 tag_len = struct.unpack('<I', content[pos:pos+4])[0]
550 assert pos+tag_len <= len(content)
551 yield (tag_code, content[pos:pos+tag_len])
555 for tag_code, tag in extract_tags(content)
557 p = code_tag.index(b'\0', 4) + 1
558 code_reader = io.BytesIO(code_tag[p:])
560 # Parse ABC (AVM2 ByteCode)
561 def read_int(reader=None):
569 b = struct.unpack('<B', buf)[0]
570 res = res | ((b & 0x7f) << shift)
576 def u30(reader=None):
577 res = read_int(reader)
578 assert res & 0xf0000000 == 0
582 def s32(reader=None):
584 if v & 0x80000000 != 0:
585 v = - ((v ^ 0xffffffff) + 1)
588 def read_string(reader=None):
592 resb = reader.read(slen)
593 assert len(resb) == slen
594 return resb.decode('utf-8')
596 def read_bytes(count, reader=None):
599 resb = reader.read(count)
600 assert len(resb) == count
603 def read_byte(reader=None):
604 resb = read_bytes(1, reader=reader)
605 res = struct.unpack('<B', resb)[0]
608 # minor_version + major_version
613 for _c in range(1, int_count):
616 for _c in range(1, uint_count):
619 read_bytes((double_count-1) * 8)
621 constant_strings = [u'']
622 for _c in range(1, string_count):
624 constant_strings.append(s)
625 namespace_count = u30()
626 for _c in range(1, namespace_count):
630 for _c in range(1, ns_set_count):
632 for _c2 in range(count):
634 multiname_count = u30()
643 0x0e: 2, # MultinameA
644 0x1b: 1, # MultinameL
645 0x1c: 1, # MultinameLA
648 for _c in range(1, multiname_count):
650 assert kind in MULTINAME_SIZES, u'Invalid multiname kind %r' % kind
652 u30() # namespace_idx
654 multinames.append(constant_strings[name_idx])
656 multinames.append('[MULTINAME kind: %d]' % kind)
657 for _c2 in range(MULTINAME_SIZES[kind]):
662 MethodInfo = collections.namedtuple(
664 ['NEED_ARGUMENTS', 'NEED_REST'])
666 for method_id in range(method_count):
669 for _ in range(param_count):
671 u30() # name index (always 0 for youtube)
673 if flags & 0x08 != 0:
676 for c in range(option_count):
679 if flags & 0x80 != 0:
680 # Param names present
681 for _ in range(param_count):
683 mi = MethodInfo(flags & 0x01 != 0, flags & 0x04 != 0)
684 method_infos.append(mi)
687 metadata_count = u30()
688 for _c in range(metadata_count):
691 for _c2 in range(item_count):
695 def parse_traits_info():
696 trait_name_idx = u30()
697 kind_full = read_byte()
698 kind = kind_full & 0x0f
699 attrs = kind_full >> 4
701 if kind in [0x00, 0x06]: # Slot or Const
703 u30() # type_name_idx
707 elif kind in [0x01, 0x02, 0x03]: # Method / Getter / Setter
710 methods[multinames[trait_name_idx]] = method_idx
711 elif kind == 0x04: # Class
714 elif kind == 0x05: # Function
717 methods[function_idx] = multinames[trait_name_idx]
719 raise ExtractorError(u'Unsupported trait kind %d' % kind)
721 if attrs & 0x4 != 0: # Metadata present
722 metadata_count = u30()
723 for _c3 in range(metadata_count):
724 u30() # metadata index
729 TARGET_CLASSNAME = u'SignatureDecipher'
730 searched_idx = multinames.index(TARGET_CLASSNAME)
731 searched_class_id = None
733 for class_id in range(class_count):
735 if name_idx == searched_idx:
736 # We found the class we're looking for!
737 searched_class_id = class_id
738 u30() # super_name idx
740 if flags & 0x08 != 0: # Protected namespace is present
741 u30() # protected_ns_idx
743 for _c2 in range(intrf_count):
747 for _c2 in range(trait_count):
750 if searched_class_id is None:
751 raise ExtractorError(u'Target class %r not found' %
756 for class_id in range(class_count):
759 for _c2 in range(trait_count):
760 trait_methods = parse_traits_info()
761 if class_id == searched_class_id:
762 method_names.update(trait_methods.items())
763 method_idxs.update(dict(
765 for name, idx in trait_methods.items()))
769 for _c in range(script_count):
772 for _c2 in range(trait_count):
776 method_body_count = u30()
777 Method = collections.namedtuple('Method', ['code', 'local_count'])
779 for _c in range(method_body_count):
783 u30() # init_scope_depth
784 u30() # max_scope_depth
786 code = read_bytes(code_length)
787 if method_idx in method_idxs:
788 m = Method(code, local_count)
789 methods[method_idxs[method_idx]] = m
790 exception_count = u30()
791 for _c2 in range(exception_count):
798 for _c2 in range(trait_count):
801 assert p + code_reader.tell() == len(code_tag)
802 assert len(methods) == len(method_idxs)
804 method_pyfunctions = {}
806 def extract_function(func_name):
807 if func_name in method_pyfunctions:
808 return method_pyfunctions[func_name]
809 if func_name not in methods:
810 raise ExtractorError(u'Cannot find function %r' % func_name)
811 m = methods[func_name]
814 registers = ['(this)'] + list(args) + [None] * m.local_count
816 coder = io.BytesIO(m.code)
818 opcode = struct.unpack('!B', coder.read(1))[0]
819 if opcode == 36: # pushbyte
820 v = struct.unpack('!B', coder.read(1))[0]
822 elif opcode == 44: # pushstring
824 stack.append(constant_strings[idx])
825 elif opcode == 48: # pushscope
826 # We don't implement the scope register, so we'll just
827 # ignore the popped value
829 elif opcode == 70: # callproperty
831 mname = multinames[index]
832 arg_count = u30(coder)
833 args = list(reversed(
834 [stack.pop() for _ in range(arg_count)]))
836 if mname == u'split':
837 assert len(args) == 1
838 assert isinstance(args[0], compat_str)
839 assert isinstance(obj, compat_str)
843 res = obj.split(args[0])
845 elif mname == u'slice':
846 assert len(args) == 1
847 assert isinstance(args[0], int)
848 assert isinstance(obj, list)
851 elif mname == u'join':
852 assert len(args) == 1
853 assert isinstance(args[0], compat_str)
854 assert isinstance(obj, list)
855 res = args[0].join(obj)
857 elif mname in method_pyfunctions:
858 stack.append(method_pyfunctions[mname](args))
860 raise NotImplementedError(
861 u'Unsupported property %r on %r'
863 elif opcode == 72: # returnvalue
866 elif opcode == 79: # callpropvoid
868 mname = multinames[index]
869 arg_count = u30(coder)
870 args = list(reversed(
871 [stack.pop() for _ in range(arg_count)]))
873 if mname == u'reverse':
874 assert isinstance(obj, list)
877 raise NotImplementedError(
878 u'Unsupported (void) property %r on %r'
880 elif opcode == 93: # findpropstrict
882 mname = multinames[index]
883 res = extract_function(mname)
885 elif opcode == 97: # setproperty
890 assert isinstance(obj, list)
891 assert isinstance(idx, int)
893 elif opcode == 98: # getlocal
895 stack.append(registers[index])
896 elif opcode == 99: # setlocal
899 registers[index] = value
900 elif opcode == 102: # getproperty
902 pname = multinames[index]
903 if pname == u'length':
905 assert isinstance(obj, list)
906 stack.append(len(obj))
907 else: # Assume attribute access
909 assert isinstance(idx, int)
911 assert isinstance(obj, list)
912 stack.append(obj[idx])
913 elif opcode == 128: # coerce
915 elif opcode == 133: # coerce_s
916 assert isinstance(stack[-1], (type(None), compat_str))
917 elif opcode == 164: # modulo
920 res = value1 % value2
922 elif opcode == 208: # getlocal_0
923 stack.append(registers[0])
924 elif opcode == 209: # getlocal_1
925 stack.append(registers[1])
926 elif opcode == 210: # getlocal_2
927 stack.append(registers[2])
928 elif opcode == 211: # getlocal_3
929 stack.append(registers[3])
930 elif opcode == 214: # setlocal_2
931 registers[2] = stack.pop()
932 elif opcode == 215: # setlocal_3
933 registers[3] = stack.pop()
935 raise NotImplementedError(
936 u'Unsupported opcode %d' % opcode)
938 method_pyfunctions[func_name] = resfunc
941 initial_function = extract_function(u'decipher')
942 return lambda s: initial_function([s])
944 def _decrypt_signature(self, s, video_id, player_url, age_gate=False):
945 """Turn the encrypted s field into a working signature"""
947 if player_url is not None:
948 if player_url.startswith(u'//'):
949 player_url = u'https:' + player_url
951 player_id = (player_url, len(s))
952 if player_id not in self._player_cache:
953 func = self._extract_signature_function(
954 video_id, player_url, len(s)
956 self._player_cache[player_id] = func
957 func = self._player_cache[player_id]
958 if self._downloader.params.get('youtube_print_sig_code'):
959 self._print_sig_code(func, len(s))
962 tb = traceback.format_exc()
963 self._downloader.report_warning(
964 u'Automatic signature extraction failed: ' + tb)
966 self._downloader.report_warning(
967 u'Warning: Falling back to static signature algorithm')
969 return self._static_decrypt_signature(
970 s, video_id, player_url, age_gate)
972 def _static_decrypt_signature(self, s, video_id, player_url, age_gate):
974 # The videos with age protection use another player, so the
975 # algorithms can be different.
977 return s[2:63] + s[82] + s[64:82] + s[63]
980 return s[86:29:-1] + s[88] + s[28:5:-1]
982 return s[25] + s[3:25] + s[0] + s[26:42] + s[79] + s[43:79] + s[91] + s[80:83]
984 return s[84:27:-1] + s[86] + s[26:5:-1]
986 return s[25] + s[3:25] + s[2] + s[26:40] + s[77] + s[41:77] + s[89] + s[78:81]
988 return s[84:78:-1] + s[87] + s[77:60:-1] + s[0] + s[59:3:-1]
990 return s[7:28] + s[87] + s[29:45] + s[55] + s[46:55] + s[2] + s[56:87] + s[28]
992 return s[6:27] + s[4] + s[28:39] + s[27] + s[40:59] + s[2] + s[60:]
994 return s[80:72:-1] + s[16] + s[71:39:-1] + s[72] + s[38:16:-1] + s[82] + s[15::-1]
996 return s[3:11] + s[0] + s[12:55] + s[84] + s[56:84]
998 return s[78:70:-1] + s[14] + s[69:37:-1] + s[70] + s[36:14:-1] + s[80] + s[:14][::-1]
1000 return s[80:63:-1] + s[0] + s[62:0:-1] + s[63]
1002 return s[80:37:-1] + s[7] + s[36:7:-1] + s[0] + s[6:0:-1] + s[37]
1004 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]
1006 return s[1:19] + s[0] + s[20:68] + s[19] + s[69:80]
1008 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]
1011 raise ExtractorError(u'Unable to decrypt signature, key length %d not supported; retrying might work' % (len(s)))
1013 def _get_available_subtitles(self, video_id, webpage):
1015 sub_list = self._download_webpage(
1016 'http://video.google.com/timedtext?hl=en&type=list&v=%s' % video_id,
1017 video_id, note=False)
1018 except ExtractorError as err:
1019 self._downloader.report_warning(u'unable to download video subtitles: %s' % compat_str(err))
1021 lang_list = re.findall(r'name="([^"]*)"[^>]+lang_code="([\w\-]+)"', sub_list)
1026 params = compat_urllib_parse.urlencode({
1029 'fmt': self._downloader.params.get('subtitlesformat', 'srt'),
1030 'name': unescapeHTML(l[0]).encode('utf-8'),
1032 url = u'http://www.youtube.com/api/timedtext?' + params
1033 sub_lang_list[lang] = url
1034 if not sub_lang_list:
1035 self._downloader.report_warning(u'video doesn\'t have subtitles')
1037 return sub_lang_list
1039 def _get_available_automatic_caption(self, video_id, webpage):
1040 """We need the webpage for getting the captions url, pass it as an
1041 argument to speed up the process."""
1042 sub_format = self._downloader.params.get('subtitlesformat', 'srt')
1043 self.to_screen(u'%s: Looking for automatic captions' % video_id)
1044 mobj = re.search(r';ytplayer.config = ({.*?});', webpage)
1045 err_msg = u'Couldn\'t find automatic captions for %s' % video_id
1047 self._downloader.report_warning(err_msg)
1049 player_config = json.loads(mobj.group(1))
1051 args = player_config[u'args']
1052 caption_url = args[u'ttsurl']
1053 timestamp = args[u'timestamp']
1054 # We get the available subtitles
1055 list_params = compat_urllib_parse.urlencode({
1060 list_url = caption_url + '&' + list_params
1061 caption_list = self._download_xml(list_url, video_id)
1062 original_lang_node = caption_list.find('track')
1063 if original_lang_node is None or original_lang_node.attrib.get('kind') != 'asr' :
1064 self._downloader.report_warning(u'Video doesn\'t have automatic captions')
1066 original_lang = original_lang_node.attrib['lang_code']
1069 for lang_node in caption_list.findall('target'):
1070 sub_lang = lang_node.attrib['lang_code']
1071 params = compat_urllib_parse.urlencode({
1072 'lang': original_lang,
1078 sub_lang_list[sub_lang] = caption_url + '&' + params
1079 return sub_lang_list
1080 # An extractor error can be raise by the download process if there are
1081 # no automatic captions but there are subtitles
1082 except (KeyError, ExtractorError):
1083 self._downloader.report_warning(err_msg)
1086 def _extract_id(self, url):
1087 mobj = re.match(self._VALID_URL, url, re.VERBOSE)
1089 raise ExtractorError(u'Invalid URL: %s' % url)
1090 video_id = mobj.group(2)
1093 def _extract_from_m3u8(self, manifest_url, video_id):
1095 def _get_urls(_manifest):
1096 lines = _manifest.split('\n')
1097 urls = filter(lambda l: l and not l.startswith('#'),
1100 manifest = self._download_webpage(manifest_url, video_id, u'Downloading formats manifest')
1101 formats_urls = _get_urls(manifest)
1102 for format_url in formats_urls:
1103 itag = self._search_regex(r'itag/(\d+?)/', format_url, 'itag')
1104 url_map[itag] = format_url
1107 def _extract_annotations(self, video_id):
1108 url = 'https://www.youtube.com/annotations_invideo?features=1&legacy=1&video_id=%s' % video_id
1109 return self._download_webpage(url, video_id, note=u'Searching for annotations.', errnote=u'Unable to download video annotations.')
1111 def _real_extract(self, url):
1112 # Extract original video URL from URL with redirection, like age verification, using next_url parameter
1113 mobj = re.search(self._NEXT_URL_RE, url)
1115 url = 'https://www.youtube.com/' + compat_urllib_parse.unquote(mobj.group(1)).lstrip('/')
1116 video_id = self._extract_id(url)
1119 url = 'https://www.youtube.com/watch?v=%s&gl=US&hl=en&has_verified=1' % video_id
1120 video_webpage = self._download_webpage(url, video_id)
1122 # Attempt to extract SWF player URL
1123 mobj = re.search(r'swfConfig.*?"(https?:\\/\\/.*?watch.*?-.*?\.swf)"', video_webpage)
1124 if mobj is not None:
1125 player_url = re.sub(r'\\(.)', r'\1', mobj.group(1))
1130 self.report_video_info_webpage_download(video_id)
1131 if re.search(r'player-age-gate-content">', video_webpage) is not None:
1132 self.report_age_confirmation()
1134 # We simulate the access to the video from www.youtube.com/v/{video_id}
1135 # this can be viewed without login into Youtube
1136 data = compat_urllib_parse.urlencode({'video_id': video_id,
1137 'el': 'player_embedded',
1140 'eurl': 'https://youtube.googleapis.com/v/' + video_id,
1144 video_info_url = 'https://www.youtube.com/get_video_info?' + data
1145 video_info_webpage = self._download_webpage(video_info_url, video_id,
1147 errnote='unable to download video info webpage')
1148 video_info = compat_parse_qs(video_info_webpage)
1151 for el_type in ['&el=embedded', '&el=detailpage', '&el=vevo', '']:
1152 video_info_url = ('https://www.youtube.com/get_video_info?&video_id=%s%s&ps=default&eurl=&gl=US&hl=en'
1153 % (video_id, el_type))
1154 video_info_webpage = self._download_webpage(video_info_url, video_id,
1156 errnote='unable to download video info webpage')
1157 video_info = compat_parse_qs(video_info_webpage)
1158 if 'token' in video_info:
1160 if 'token' not in video_info:
1161 if 'reason' in video_info:
1162 raise ExtractorError(u'YouTube said: %s' % video_info['reason'][0], expected=True)
1164 raise ExtractorError(u'"token" parameter not in video info for unknown reason')
1166 if 'view_count' in video_info:
1167 view_count = int(video_info['view_count'][0])
1171 # Check for "rental" videos
1172 if 'ypc_video_rental_bar_text' in video_info and 'author' not in video_info:
1173 raise ExtractorError(u'"rental" videos not supported')
1175 # Start extracting information
1176 self.report_information_extraction(video_id)
1179 if 'author' not in video_info:
1180 raise ExtractorError(u'Unable to extract uploader name')
1181 video_uploader = compat_urllib_parse.unquote_plus(video_info['author'][0])
1184 video_uploader_id = None
1185 mobj = re.search(r'<link itemprop="url" href="http://www.youtube.com/(?:user|channel)/([^"]+)">', video_webpage)
1186 if mobj is not None:
1187 video_uploader_id = mobj.group(1)
1189 self._downloader.report_warning(u'unable to extract uploader nickname')
1192 if 'title' in video_info:
1193 video_title = compat_urllib_parse.unquote_plus(video_info['title'][0])
1195 self._downloader.report_warning(u'Unable to extract video title')
1199 # We try first to get a high quality image:
1200 m_thumb = re.search(r'<span itemprop="thumbnail".*?href="(.*?)">',
1201 video_webpage, re.DOTALL)
1202 if m_thumb is not None:
1203 video_thumbnail = m_thumb.group(1)
1204 elif 'thumbnail_url' not in video_info:
1205 self._downloader.report_warning(u'unable to extract video thumbnail')
1206 video_thumbnail = None
1207 else: # don't panic if we can't find it
1208 video_thumbnail = compat_urllib_parse.unquote_plus(video_info['thumbnail_url'][0])
1212 mobj = re.search(r'id="eow-date.*?>(.*?)</span>', video_webpage, re.DOTALL)
1213 if mobj is not None:
1214 upload_date = ' '.join(re.sub(r'[/,-]', r' ', mobj.group(1)).split())
1215 upload_date = unified_strdate(upload_date)
1218 video_description = get_element_by_id("eow-description", video_webpage)
1219 if video_description:
1220 video_description = re.sub(r'''(?x)
1222 (?:[a-zA-Z-]+="[^"]+"\s+)*?
1224 (?:[a-zA-Z-]+="[^"]+"\s+)*?
1225 class="yt-uix-redirect-link"\s*>
1228 ''', r'\1', video_description)
1229 video_description = clean_html(video_description)
1231 fd_mobj = re.search(r'<meta name="description" content="([^"]+)"', video_webpage)
1233 video_description = unescapeHTML(fd_mobj.group(1))
1235 video_description = u''
1237 def _extract_count(klass):
1238 count = self._search_regex(
1239 r'class="%s">([\d,]+)</span>' % re.escape(klass),
1240 video_webpage, klass, default=None)
1241 if count is not None:
1242 return int(count.replace(',', ''))
1244 like_count = _extract_count(u'likes-count')
1245 dislike_count = _extract_count(u'dislikes-count')
1248 video_subtitles = self.extract_subtitles(video_id, video_webpage)
1250 if self._downloader.params.get('listsubtitles', False):
1251 self._list_available_subtitles(video_id, video_webpage)
1254 if 'length_seconds' not in video_info:
1255 self._downloader.report_warning(u'unable to extract video duration')
1256 video_duration = None
1258 video_duration = int(compat_urllib_parse.unquote_plus(video_info['length_seconds'][0]))
1261 video_annotations = None
1262 if self._downloader.params.get('writeannotations', False):
1263 video_annotations = self._extract_annotations(video_id)
1265 # Decide which formats to download
1267 mobj = re.search(r';ytplayer.config = ({.*?});', video_webpage)
1269 raise ValueError('Could not find vevo ID')
1270 info = json.loads(mobj.group(1))
1272 # Easy way to know if the 's' value is in url_encoded_fmt_stream_map
1273 # this signatures are encrypted
1274 if 'url_encoded_fmt_stream_map' not in args:
1275 raise ValueError(u'No stream_map present') # caught below
1276 re_signature = re.compile(r'[&,]s=')
1277 m_s = re_signature.search(args['url_encoded_fmt_stream_map'])
1279 self.to_screen(u'%s: Encrypted signatures detected.' % video_id)
1280 video_info['url_encoded_fmt_stream_map'] = [args['url_encoded_fmt_stream_map']]
1281 m_s = re_signature.search(args.get('adaptive_fmts', u''))
1283 if 'adaptive_fmts' in video_info:
1284 video_info['adaptive_fmts'][0] += ',' + args['adaptive_fmts']
1286 video_info['adaptive_fmts'] = [args['adaptive_fmts']]
1290 def _map_to_format_list(urlmap):
1292 for itag, video_real_url in urlmap.items():
1295 'url': video_real_url,
1296 'player_url': player_url,
1298 if itag in self._formats:
1299 dct.update(self._formats[itag])
1303 if 'conn' in video_info and video_info['conn'][0].startswith('rtmp'):
1304 self.report_rtmp_download()
1306 'format_id': '_rtmp',
1308 'url': video_info['conn'][0],
1309 'player_url': player_url,
1311 elif len(video_info.get('url_encoded_fmt_stream_map', [])) >= 1 or len(video_info.get('adaptive_fmts', [])) >= 1:
1312 encoded_url_map = video_info.get('url_encoded_fmt_stream_map', [''])[0] + ',' + video_info.get('adaptive_fmts',[''])[0]
1313 if 'rtmpe%3Dyes' in encoded_url_map:
1314 raise ExtractorError('rtmpe downloads are not supported, see https://github.com/rg3/youtube-dl/issues/343 for more information.', expected=True)
1316 for url_data_str in encoded_url_map.split(','):
1317 url_data = compat_parse_qs(url_data_str)
1318 if 'itag' in url_data and 'url' in url_data:
1319 url = url_data['url'][0]
1320 if 'sig' in url_data:
1321 url += '&signature=' + url_data['sig'][0]
1322 elif 's' in url_data:
1323 encrypted_sig = url_data['s'][0]
1324 if self._downloader.params.get('verbose'):
1326 if player_url is None:
1327 player_version = 'unknown'
1329 player_version = self._search_regex(
1330 r'-(.+)\.swf$', player_url,
1331 u'flash player', fatal=False)
1332 player_desc = 'flash player %s' % player_version
1334 player_version = self._search_regex(
1335 r'html5player-(.+?)\.js', video_webpage,
1336 'html5 player', fatal=False)
1337 player_desc = u'html5 player %s' % player_version
1339 parts_sizes = u'.'.join(compat_str(len(part)) for part in encrypted_sig.split('.'))
1340 self.to_screen(u'encrypted signature length %d (%s), itag %s, %s' %
1341 (len(encrypted_sig), parts_sizes, url_data['itag'][0], player_desc))
1344 jsplayer_url_json = self._search_regex(
1345 r'"assets":.+?"js":\s*("[^"]+")',
1346 video_webpage, u'JS player URL')
1347 player_url = json.loads(jsplayer_url_json)
1349 signature = self._decrypt_signature(
1350 encrypted_sig, video_id, player_url, age_gate)
1351 url += '&signature=' + signature
1352 if 'ratebypass' not in url:
1353 url += '&ratebypass=yes'
1354 url_map[url_data['itag'][0]] = url
1355 formats = _map_to_format_list(url_map)
1356 elif video_info.get('hlsvp'):
1357 manifest_url = video_info['hlsvp'][0]
1358 url_map = self._extract_from_m3u8(manifest_url, video_id)
1359 formats = _map_to_format_list(url_map)
1361 raise ExtractorError(u'no conn, hlsvp or url_encoded_fmt_stream_map information found in video info')
1363 # Look for the DASH manifest
1364 dash_manifest_url_lst = video_info.get('dashmpd')
1365 if (dash_manifest_url_lst and dash_manifest_url_lst[0] and
1366 self._downloader.params.get('youtube_include_dash_manifest', False)):
1368 dash_doc = self._download_xml(
1369 dash_manifest_url_lst[0], video_id,
1370 note=u'Downloading DASH manifest',
1371 errnote=u'Could not download DASH manifest')
1372 for r in dash_doc.findall(u'.//{urn:mpeg:DASH:schema:MPD:2011}Representation'):
1373 url_el = r.find('{urn:mpeg:DASH:schema:MPD:2011}BaseURL')
1376 format_id = r.attrib['id']
1377 video_url = url_el.text
1378 filesize = int_or_none(url_el.attrib.get('{http://youtube.com/yt/2012/10/10}contentLength'))
1380 'format_id': format_id,
1382 'width': int_or_none(r.attrib.get('width')),
1383 'tbr': int_or_none(r.attrib.get('bandwidth'), 1000),
1384 'asr': int_or_none(r.attrib.get('audioSamplingRate')),
1385 'filesize': filesize,
1388 existing_format = next(
1389 fo for fo in formats
1390 if fo['format_id'] == format_id)
1391 except StopIteration:
1392 f.update(self._formats.get(format_id, {}))
1395 existing_format.update(f)
1397 except (ExtractorError, KeyError) as e:
1398 self.report_warning(u'Skipping DASH manifest: %s' % e, video_id)
1400 self._sort_formats(formats)
1404 'uploader': video_uploader,
1405 'uploader_id': video_uploader_id,
1406 'upload_date': upload_date,
1407 'title': video_title,
1408 'thumbnail': video_thumbnail,
1409 'description': video_description,
1410 'subtitles': video_subtitles,
1411 'duration': video_duration,
1412 'age_limit': 18 if age_gate else 0,
1413 'annotations': video_annotations,
1414 'webpage_url': 'https://www.youtube.com/watch?v=%s' % video_id,
1415 'view_count': view_count,
1416 'like_count': like_count,
1417 'dislike_count': dislike_count,
1421 class YoutubePlaylistIE(YoutubeBaseInfoExtractor):
1422 IE_DESC = u'YouTube.com playlists'
1423 _VALID_URL = r"""(?:
1428 (?:course|view_play_list|my_playlists|artist|playlist|watch)
1429 \? (?:.*?&)*? (?:p|a|list)=
1432 ((?:PL|EC|UU|FL|RD)?[0-9A-Za-z-_]{10,})
1435 ((?:PL|EC|UU|FL|RD)[0-9A-Za-z-_]{10,})
1437 _TEMPLATE_URL = 'https://www.youtube.com/playlist?list=%s&page=%s'
1438 _MORE_PAGES_INDICATOR = r'data-link-type="next"'
1439 _VIDEO_RE = r'href="/watch\?v=(?P<id>[0-9A-Za-z_-]{11})&[^"]*?index=(?P<index>\d+)'
1440 IE_NAME = u'youtube:playlist'
1443 def suitable(cls, url):
1444 """Receives a URL and returns True if suitable for this IE."""
1445 return re.match(cls._VALID_URL, url, re.VERBOSE) is not None
1447 def _real_initialize(self):
1450 def _ids_to_results(self, ids):
1451 return [self.url_result(vid_id, 'Youtube', video_id=vid_id)
1454 def _extract_mix(self, playlist_id):
1455 # The mixes are generated from a a single video
1456 # the id of the playlist is just 'RD' + video_id
1457 url = 'https://youtube.com/watch?v=%s&list=%s' % (playlist_id[-11:], playlist_id)
1458 webpage = self._download_webpage(url, playlist_id, u'Downloading Youtube mix')
1459 title_span = (get_element_by_attribute('class', 'title long-title', webpage) or
1460 get_element_by_attribute('class', 'title ', webpage))
1461 title = clean_html(title_span)
1462 video_re = r'data-index="\d+".*?href="/watch\?v=([0-9A-Za-z_-]{11})&[^"]*?list=%s' % re.escape(playlist_id)
1463 ids = orderedSet(re.findall(video_re, webpage))
1464 url_results = self._ids_to_results(ids)
1466 return self.playlist_result(url_results, playlist_id, title)
1468 def _real_extract(self, url):
1469 # Extract playlist id
1470 mobj = re.match(self._VALID_URL, url, re.VERBOSE)
1472 raise ExtractorError(u'Invalid URL: %s' % url)
1473 playlist_id = mobj.group(1) or mobj.group(2)
1475 # Check if it's a video-specific URL
1476 query_dict = compat_urlparse.parse_qs(compat_urlparse.urlparse(url).query)
1477 if 'v' in query_dict:
1478 video_id = query_dict['v'][0]
1479 if self._downloader.params.get('noplaylist'):
1480 self.to_screen(u'Downloading just video %s because of --no-playlist' % video_id)
1481 return self.url_result(video_id, 'Youtube', video_id=video_id)
1483 self.to_screen(u'Downloading playlist PL%s - add --no-playlist to just download video %s' % (playlist_id, video_id))
1485 if playlist_id.startswith('RD'):
1486 # Mixes require a custom extraction process
1487 return self._extract_mix(playlist_id)
1488 if playlist_id.startswith('TL'):
1489 raise ExtractorError(u'For downloading YouTube.com top lists, use '
1490 u'the "yttoplist" keyword, for example "youtube-dl \'yttoplist:music:Top Tracks\'"', expected=True)
1492 # Extract the video ids from the playlist pages
1495 for page_num in itertools.count(1):
1496 url = self._TEMPLATE_URL % (playlist_id, page_num)
1497 page = self._download_webpage(url, playlist_id, u'Downloading page #%s' % page_num)
1498 matches = re.finditer(self._VIDEO_RE, page)
1499 # We remove the duplicates and the link with index 0
1500 # (it's not the first video of the playlist)
1501 new_ids = orderedSet(m.group('id') for m in matches if m.group('index') != '0')
1504 if re.search(self._MORE_PAGES_INDICATOR, page) is None:
1508 playlist_title = self._og_search_title(page)
1509 except RegexNotFoundError:
1510 self.report_warning(
1511 u'Playlist page is missing OpenGraph title, falling back ...',
1513 playlist_title = self._html_search_regex(
1514 r'<h1 class="pl-header-title">(.*?)</h1>', page, u'title')
1516 url_results = self._ids_to_results(ids)
1517 return self.playlist_result(url_results, playlist_id, playlist_title)
1520 class YoutubeTopListIE(YoutubePlaylistIE):
1521 IE_NAME = u'youtube:toplist'
1522 IE_DESC = (u'YouTube.com top lists, "yttoplist:{channel}:{list title}"'
1523 u' (Example: "yttoplist:music:Top Tracks")')
1524 _VALID_URL = r'yttoplist:(?P<chann>.*?):(?P<title>.*?)$'
1526 def _real_extract(self, url):
1527 mobj = re.match(self._VALID_URL, url)
1528 channel = mobj.group('chann')
1529 title = mobj.group('title')
1530 query = compat_urllib_parse.urlencode({'title': title})
1531 playlist_re = 'href="([^"]+?%s[^"]+?)"' % re.escape(query)
1532 channel_page = self._download_webpage('https://www.youtube.com/%s' % channel, title)
1533 link = self._html_search_regex(playlist_re, channel_page, u'list')
1534 url = compat_urlparse.urljoin('https://www.youtube.com/', link)
1536 video_re = r'data-index="\d+".*?data-video-id="([0-9A-Za-z_-]{11})"'
1538 # sometimes the webpage doesn't contain the videos
1539 # retry until we get them
1540 for i in itertools.count(0):
1541 msg = u'Downloading Youtube mix'
1543 msg += ', retry #%d' % i
1544 webpage = self._download_webpage(url, title, msg)
1545 ids = orderedSet(re.findall(video_re, webpage))
1548 url_results = self._ids_to_results(ids)
1549 return self.playlist_result(url_results, playlist_title=title)
1552 class YoutubeChannelIE(InfoExtractor):
1553 IE_DESC = u'YouTube.com channels'
1554 _VALID_URL = r"^(?:https?://)?(?:youtu\.be|(?:\w+\.)?youtube(?:-nocookie)?\.com)/channel/([0-9A-Za-z_-]+)"
1555 _MORE_PAGES_INDICATOR = 'yt-uix-load-more'
1556 _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'
1557 IE_NAME = u'youtube:channel'
1559 def extract_videos_from_page(self, page):
1561 for mobj in re.finditer(r'href="/watch\?v=([0-9A-Za-z_-]+)&?', page):
1562 if mobj.group(1) not in ids_in_page:
1563 ids_in_page.append(mobj.group(1))
1566 def _real_extract(self, url):
1567 # Extract channel id
1568 mobj = re.match(self._VALID_URL, url)
1570 raise ExtractorError(u'Invalid URL: %s' % url)
1572 # Download channel page
1573 channel_id = mobj.group(1)
1575 url = 'https://www.youtube.com/channel/%s/videos' % channel_id
1576 channel_page = self._download_webpage(url, channel_id)
1577 autogenerated = re.search(r'''(?x)
1579 channel-header-autogenerated-label|
1580 yt-channel-title-autogenerated
1581 )[^"]*"''', channel_page) is not None
1584 # The videos are contained in a single page
1585 # the ajax pages can't be used, they are empty
1586 video_ids = self.extract_videos_from_page(channel_page)
1588 # Download all channel pages using the json-based channel_ajax query
1589 for pagenum in itertools.count(1):
1590 url = self._MORE_PAGES_URL % (pagenum, channel_id)
1591 page = self._download_webpage(url, channel_id,
1592 u'Downloading page #%s' % pagenum)
1594 page = json.loads(page)
1596 ids_in_page = self.extract_videos_from_page(page['content_html'])
1597 video_ids.extend(ids_in_page)
1599 if self._MORE_PAGES_INDICATOR not in page['load_more_widget_html']:
1602 self._downloader.to_screen(u'[youtube] Channel %s: Found %i videos' % (channel_id, len(video_ids)))
1604 url_entries = [self.url_result(video_id, 'Youtube', video_id=video_id)
1605 for video_id in video_ids]
1606 return self.playlist_result(url_entries, channel_id)
1609 class YoutubeUserIE(InfoExtractor):
1610 IE_DESC = u'YouTube.com user videos (URL or "ytuser" keyword)'
1611 _VALID_URL = r'(?:(?:(?:https?://)?(?:\w+\.)?youtube\.com/(?:user/)?(?!(?:attribution_link|watch)(?:$|[^a-z_A-Z0-9-])))|ytuser:)(?!feed/)([A-Za-z0-9_-]+)'
1612 _TEMPLATE_URL = 'http://gdata.youtube.com/feeds/api/users/%s'
1613 _GDATA_PAGE_SIZE = 50
1614 _GDATA_URL = 'http://gdata.youtube.com/feeds/api/users/%s/uploads?max-results=%d&start-index=%d&alt=json'
1615 IE_NAME = u'youtube:user'
1618 def suitable(cls, url):
1619 # Don't return True if the url can be extracted with other youtube
1620 # extractor, the regex would is too permissive and it would match.
1621 other_ies = iter(klass for (name, klass) in globals().items() if name.endswith('IE') and klass is not cls)
1622 if any(ie.suitable(url) for ie in other_ies): return False
1623 else: return super(YoutubeUserIE, cls).suitable(url)
1625 def _real_extract(self, url):
1627 mobj = re.match(self._VALID_URL, url)
1629 raise ExtractorError(u'Invalid URL: %s' % url)
1631 username = mobj.group(1)
1633 # Download video ids using YouTube Data API. Result size per
1634 # query is limited (currently to 50 videos) so we need to query
1635 # page by page until there are no video ids - it means we got
1638 def download_page(pagenum):
1639 start_index = pagenum * self._GDATA_PAGE_SIZE + 1
1641 gdata_url = self._GDATA_URL % (username, self._GDATA_PAGE_SIZE, start_index)
1642 page = self._download_webpage(
1643 gdata_url, username,
1644 u'Downloading video ids from %d to %d' % (
1645 start_index, start_index + self._GDATA_PAGE_SIZE))
1648 response = json.loads(page)
1649 except ValueError as err:
1650 raise ExtractorError(u'Invalid JSON in API response: ' + compat_str(err))
1651 if 'entry' not in response['feed']:
1654 # Extract video identifiers
1655 entries = response['feed']['entry']
1656 for entry in entries:
1657 title = entry['title']['$t']
1658 video_id = entry['id']['$t'].split('/')[-1]
1662 'ie_key': 'Youtube',
1666 url_results = PagedList(download_page, self._GDATA_PAGE_SIZE)
1668 return self.playlist_result(url_results, playlist_title=username)
1671 class YoutubeSearchIE(SearchInfoExtractor):
1672 IE_DESC = u'YouTube.com searches'
1673 _API_URL = 'https://gdata.youtube.com/feeds/api/videos?q=%s&start-index=%i&max-results=50&v=2&alt=jsonc'
1675 IE_NAME = u'youtube:search'
1676 _SEARCH_KEY = 'ytsearch'
1678 def _get_n_results(self, query, n):
1679 """Get a specified number of results for a query"""
1685 while (50 * pagenum) < limit:
1686 result_url = self._API_URL % (compat_urllib_parse.quote_plus(query), (50*pagenum)+1)
1687 data_json = self._download_webpage(
1688 result_url, video_id=u'query "%s"' % query,
1689 note=u'Downloading page %s' % (pagenum + 1),
1690 errnote=u'Unable to download API page')
1691 data = json.loads(data_json)
1692 api_response = data['data']
1694 if 'items' not in api_response:
1695 raise ExtractorError(u'[youtube] No video results')
1697 new_ids = list(video['id'] for video in api_response['items'])
1698 video_ids += new_ids
1700 limit = min(n, api_response['totalItems'])
1703 if len(video_ids) > n:
1704 video_ids = video_ids[:n]
1705 videos = [self.url_result(video_id, 'Youtube', video_id=video_id)
1706 for video_id in video_ids]
1707 return self.playlist_result(videos, query)
1709 class YoutubeSearchDateIE(YoutubeSearchIE):
1710 IE_NAME = YoutubeSearchIE.IE_NAME + ':date'
1711 _API_URL = 'https://gdata.youtube.com/feeds/api/videos?q=%s&start-index=%i&max-results=50&v=2&alt=jsonc&orderby=published'
1712 _SEARCH_KEY = 'ytsearchdate'
1713 IE_DESC = u'YouTube.com searches, newest videos first'
1715 class YoutubeShowIE(InfoExtractor):
1716 IE_DESC = u'YouTube.com (multi-season) shows'
1717 _VALID_URL = r'https?://www\.youtube\.com/show/(.*)'
1718 IE_NAME = u'youtube:show'
1720 def _real_extract(self, url):
1721 mobj = re.match(self._VALID_URL, url)
1722 show_name = mobj.group(1)
1723 webpage = self._download_webpage(url, show_name, u'Downloading show webpage')
1724 # There's one playlist for each season of the show
1725 m_seasons = list(re.finditer(r'href="(/playlist\?list=.*?)"', webpage))
1726 self.to_screen(u'%s: Found %s seasons' % (show_name, len(m_seasons)))
1727 return [self.url_result('https://www.youtube.com' + season.group(1), 'YoutubePlaylist') for season in m_seasons]
1730 class YoutubeFeedsInfoExtractor(YoutubeBaseInfoExtractor):
1732 Base class for extractors that fetch info from
1733 http://www.youtube.com/feed_ajax
1734 Subclasses must define the _FEED_NAME and _PLAYLIST_TITLE properties.
1736 _LOGIN_REQUIRED = True
1737 # use action_load_personal_feed instead of action_load_system_feed
1738 _PERSONAL_FEED = False
1741 def _FEED_TEMPLATE(self):
1742 action = 'action_load_system_feed'
1743 if self._PERSONAL_FEED:
1744 action = 'action_load_personal_feed'
1745 return 'http://www.youtube.com/feed_ajax?%s=1&feed_name=%s&paging=%%s' % (action, self._FEED_NAME)
1749 return u'youtube:%s' % self._FEED_NAME
1751 def _real_initialize(self):
1754 def _real_extract(self, url):
1757 for i in itertools.count(1):
1758 info = self._download_webpage(self._FEED_TEMPLATE % paging,
1759 u'%s feed' % self._FEED_NAME,
1760 u'Downloading page %s' % i)
1761 info = json.loads(info)
1762 feed_html = info['feed_html']
1763 m_ids = re.finditer(r'"/watch\?v=(.*?)["&]', feed_html)
1764 ids = orderedSet(m.group(1) for m in m_ids)
1765 feed_entries.extend(
1766 self.url_result(video_id, 'Youtube', video_id=video_id)
1767 for video_id in ids)
1768 if info['paging'] is None:
1770 paging = info['paging']
1771 return self.playlist_result(feed_entries, playlist_title=self._PLAYLIST_TITLE)
1773 class YoutubeSubscriptionsIE(YoutubeFeedsInfoExtractor):
1774 IE_DESC = u'YouTube.com subscriptions feed, "ytsubs" keyword(requires authentication)'
1775 _VALID_URL = r'https?://www\.youtube\.com/feed/subscriptions|:ytsubs(?:criptions)?'
1776 _FEED_NAME = 'subscriptions'
1777 _PLAYLIST_TITLE = u'Youtube Subscriptions'
1779 class YoutubeRecommendedIE(YoutubeFeedsInfoExtractor):
1780 IE_DESC = u'YouTube.com recommended videos, "ytrec" keyword (requires authentication)'
1781 _VALID_URL = r'https?://www\.youtube\.com/feed/recommended|:ytrec(?:ommended)?'
1782 _FEED_NAME = 'recommended'
1783 _PLAYLIST_TITLE = u'Youtube Recommended videos'
1785 class YoutubeWatchLaterIE(YoutubeFeedsInfoExtractor):
1786 IE_DESC = u'Youtube watch later list, "ytwatchlater" keyword (requires authentication)'
1787 _VALID_URL = r'https?://www\.youtube\.com/feed/watch_later|:ytwatchlater'
1788 _FEED_NAME = 'watch_later'
1789 _PLAYLIST_TITLE = u'Youtube Watch Later'
1790 _PERSONAL_FEED = True
1792 class YoutubeHistoryIE(YoutubeFeedsInfoExtractor):
1793 IE_DESC = u'Youtube watch history, "ythistory" keyword (requires authentication)'
1794 _VALID_URL = u'https?://www\.youtube\.com/feed/history|:ythistory'
1795 _FEED_NAME = 'history'
1796 _PERSONAL_FEED = True
1797 _PLAYLIST_TITLE = u'Youtube Watch History'
1799 class YoutubeFavouritesIE(YoutubeBaseInfoExtractor):
1800 IE_NAME = u'youtube:favorites'
1801 IE_DESC = u'YouTube.com favourite videos, "ytfav" keyword (requires authentication)'
1802 _VALID_URL = r'https?://www\.youtube\.com/my_favorites|:ytfav(?:ou?rites)?'
1803 _LOGIN_REQUIRED = True
1805 def _real_extract(self, url):
1806 webpage = self._download_webpage('https://www.youtube.com/my_favorites', 'Youtube Favourites videos')
1807 playlist_id = self._search_regex(r'list=(.+?)["&]', webpage, u'favourites playlist id')
1808 return self.url_result(playlist_id, 'YoutubePlaylist')
1811 class YoutubeTruncatedURLIE(InfoExtractor):
1812 IE_NAME = 'youtube:truncated_url'
1813 IE_DESC = False # Do not list
1814 _VALID_URL = r'''(?x)
1815 (?:https?://)?[^/]+/watch\?feature=[a-z_]+$|
1816 (?:https?://)?(?:www\.)?youtube\.com/attribution_link\?a=[^&]+$
1819 def _real_extract(self, url):
1820 raise ExtractorError(
1821 u'Did you forget to quote the URL? Remember that & is a meta '
1822 u'character in most shells, so you want to put the URL in quotes, '
1824 u'"http://www.youtube.com/watch?feature=foo&v=BaW_jenozKc" '
1825 u' or simply youtube-dl BaW_jenozKc .',