15 from .common
import InfoExtractor
, SearchInfoExtractor
16 from .subtitles
import SubtitlesInfoExtractor
21 compat_urllib_request
,
28 get_element_by_attribute
,
37 class YoutubeBaseInfoExtractor(InfoExtractor
):
38 """Provide base functions for Youtube extractors"""
39 _LOGIN_URL
= 'https://accounts.google.com/ServiceLogin'
40 _LANG_URL
= r
'https://www.youtube.com/?hl=en&persist_hl=1&gl=US&persist_gl=1&opt_out_ackd=1'
41 _AGE_URL
= 'http://www.youtube.com/verify_age?next_url=/&gl=US&hl=en'
42 _NETRC_MACHINE
= 'youtube'
43 # If True it will raise an error if no login info is provided
44 _LOGIN_REQUIRED
= False
46 def _set_language(self
):
47 return bool(self
._download
_webpage
(
49 note
=u
'Setting language', errnote
='unable to set language',
53 (username
, password
) = self
._get
_login
_info
()
54 # No authentication to be performed
56 if self
._LOGIN
_REQUIRED
:
57 raise ExtractorError(u
'No login info available, needed for using %s.' % self
.IE_NAME
, expected
=True)
60 login_page
= self
._download
_webpage
(
61 self
._LOGIN
_URL
, None,
62 note
=u
'Downloading login page',
63 errnote
=u
'unable to fetch login page', fatal
=False)
64 if login_page
is False:
67 galx
= self
._search
_regex
(r
'(?s)<input.+?name="GALX".+?value="(.+?)"',
68 login_page
, u
'Login GALX parameter')
72 u
'continue': u
'https://www.youtube.com/signin?action_handle_signin=true&feature=sign_in_button&hl=en_US&nomobiletemp=1',
76 u
'PersistentCookie': u
'yes',
78 u
'bgresponse': u
'js_disabled',
79 u
'checkConnection': u
'',
80 u
'checkedDomains': u
'youtube',
85 u
'signIn': u
'Sign in',
87 u
'service': u
'youtube',
91 # Convert to UTF-8 *before* urlencode because Python 2.x's urlencode
93 login_form
= dict((k
.encode('utf-8'), v
.encode('utf-8')) for k
,v
in login_form_strs
.items())
94 login_data
= compat_urllib_parse
.urlencode(login_form
).encode('ascii')
96 req
= compat_urllib_request
.Request(self
._LOGIN
_URL
, login_data
)
97 login_results
= self
._download
_webpage
(
99 note
=u
'Logging in', errnote
=u
'unable to log in', fatal
=False)
100 if login_results
is False:
102 if re
.search(r
'(?i)<form[^>]* id="gaia_loginform"', login_results
) is not None:
103 self
._downloader
.report_warning(u
'unable to log in: bad username or password')
107 def _confirm_age(self
):
110 'action_confirm': 'Confirm',
112 req
= compat_urllib_request
.Request(self
._AGE
_URL
, compat_urllib_parse
.urlencode(age_form
))
114 self
._download
_webpage
(
116 note
=u
'Confirming age', errnote
=u
'Unable to confirm age')
119 def _real_initialize(self
):
120 if self
._downloader
is None:
122 if not self
._set
_language
():
124 if not self
._login
():
129 class YoutubeIE(YoutubeBaseInfoExtractor
, SubtitlesInfoExtractor
):
130 IE_DESC
= u
'YouTube.com'
131 _VALID_URL
= r
"""(?x)^
133 (?:https?://|//)? # http(s):// or protocol-independent URL (optional)
134 (?:(?:(?:(?:\w+\.)?[yY][oO][uU][tT][uU][bB][eE](?:-nocookie)?\.com/|
135 (?:www\.)?deturl\.com/www\.youtube\.com/|
136 (?:www\.)?pwnyoutube\.com|
137 tube\.majestyc\.net/|
138 youtube\.googleapis\.com/) # the various hostnames, with wildcard subdomains
139 (?:.*?\#/)? # handle anchor (#/) redirect urls
140 (?: # the various things that can precede the ID:
141 (?:(?:v|embed|e)/) # v/ or embed/ or e/
142 |(?: # or the v= param in all its forms
143 (?:(?:watch|movie)(?:_popup)?(?:\.php)?)? # preceding watch(_popup|.php) or nothing (like /?v=xxxx)
144 (?:\?|\#!?) # the params delimiter ? or # or #!
145 (?:.*?&)? # any other preceding param (like /?s=tuff&v=xxxx)
149 |youtu\.be/ # just youtu.be/xxxx
151 )? # all until now is optional -> you can pass the naked ID
152 ([0-9A-Za-z_-]{11}) # here is it! the YouTube video ID
153 (?(1).+)? # if we found the ID, everything can follow
155 _NEXT_URL_RE
= r
'[\?&]next_url=([^&]+)'
157 '5': {'ext': 'flv', 'width': 400, 'height': 240}
,
158 '6': {'ext': 'flv', 'width': 450, 'height': 270}
,
159 '13': {'ext': '3gp'}
,
160 '17': {'ext': '3gp', 'width': 176, 'height': 144}
,
161 '18': {'ext': 'mp4', 'width': 640, 'height': 360}
,
162 '22': {'ext': 'mp4', 'width': 1280, 'height': 720}
,
163 '34': {'ext': 'flv', 'width': 640, 'height': 360}
,
164 '35': {'ext': 'flv', 'width': 854, 'height': 480}
,
165 '36': {'ext': '3gp', 'width': 320, 'height': 240}
,
166 '37': {'ext': 'mp4', 'width': 1920, 'height': 1080}
,
167 '38': {'ext': 'mp4', 'width': 4096, 'height': 3072}
,
168 '43': {'ext': 'webm', 'width': 640, 'height': 360}
,
169 '44': {'ext': 'webm', 'width': 854, 'height': 480}
,
170 '45': {'ext': 'webm', 'width': 1280, 'height': 720}
,
171 '46': {'ext': 'webm', 'width': 1920, 'height': 1080}
,
175 '82': {'ext': 'mp4', 'height': 360, 'resolution': '360p', 'format_note': '3D', 'preference': -20}
,
176 '83': {'ext': 'mp4', 'height': 480, 'resolution': '480p', 'format_note': '3D', 'preference': -20}
,
177 '84': {'ext': 'mp4', 'height': 720, 'resolution': '720p', 'format_note': '3D', 'preference': -20}
,
178 '85': {'ext': 'mp4', 'height': 1080, 'resolution': '1080p', 'format_note': '3D', 'preference': -20}
,
179 '100': {'ext': 'webm', 'height': 360, 'resolution': '360p', 'format_note': '3D', 'preference': -20}
,
180 '101': {'ext': 'webm', 'height': 480, 'resolution': '480p', 'format_note': '3D', 'preference': -20}
,
181 '102': {'ext': 'webm', 'height': 720, 'resolution': '720p', 'format_note': '3D', 'preference': -20}
,
183 # Apple HTTP Live Streaming
184 '92': {'ext': 'mp4', 'height': 240, 'resolution': '240p', 'format_note': 'HLS', 'preference': -10}
,
185 '93': {'ext': 'mp4', 'height': 360, 'resolution': '360p', 'format_note': 'HLS', 'preference': -10}
,
186 '94': {'ext': 'mp4', 'height': 480, 'resolution': '480p', 'format_note': 'HLS', 'preference': -10}
,
187 '95': {'ext': 'mp4', 'height': 720, 'resolution': '720p', 'format_note': 'HLS', 'preference': -10}
,
188 '96': {'ext': 'mp4', 'height': 1080, 'resolution': '1080p', 'format_note': 'HLS', 'preference': -10}
,
189 '132': {'ext': 'mp4', 'height': 240, 'resolution': '240p', 'format_note': 'HLS', 'preference': -10}
,
190 '151': {'ext': 'mp4', 'height': 72, 'resolution': '72p', 'format_note': 'HLS', 'preference': -10}
,
193 '133': {'ext': 'mp4', 'height': 240, 'resolution': '240p', 'format_note': 'DASH video', 'preference': -40}
,
194 '134': {'ext': 'mp4', 'height': 360, 'resolution': '360p', 'format_note': 'DASH video', 'preference': -40}
,
195 '135': {'ext': 'mp4', 'height': 480, 'resolution': '480p', 'format_note': 'DASH video', 'preference': -40}
,
196 '136': {'ext': 'mp4', 'height': 720, 'resolution': '720p', 'format_note': 'DASH video', 'preference': -40}
,
197 '137': {'ext': 'mp4', 'height': 1080, 'resolution': '1080p', 'format_note': 'DASH video', 'preference': -40}
,
198 '138': {'ext': 'mp4', 'height': 1081, 'resolution': '>1080p', 'format_note': 'DASH video', 'preference': -40}
,
199 '160': {'ext': 'mp4', 'height': 192, 'resolution': '192p', 'format_note': 'DASH video', 'preference': -40}
,
200 '264': {'ext': 'mp4', 'height': 1080, 'resolution': '1080p', 'format_note': 'DASH video', 'preference': -40}
,
203 '139': {'ext': 'm4a', 'format_note': 'DASH audio', 'vcodec': 'none', 'abr': 48, 'preference': -50}
,
204 '140': {'ext': 'm4a', 'format_note': 'DASH audio', 'vcodec': 'none', 'abr': 128, 'preference': -50}
,
205 '141': {'ext': 'm4a', 'format_note': 'DASH audio', 'vcodec': 'none', 'abr': 256, 'preference': -50}
,
208 '242': {'ext': 'webm', 'height': 240, 'resolution': '240p', 'format_note': 'DASH webm', 'preference': -40}
,
209 '243': {'ext': 'webm', 'height': 360, 'resolution': '360p', 'format_note': 'DASH webm', 'preference': -40}
,
210 '244': {'ext': 'webm', 'height': 480, 'resolution': '480p', 'format_note': 'DASH webm', 'preference': -40}
,
211 '245': {'ext': 'webm', 'height': 480, 'resolution': '480p', 'format_note': 'DASH webm', 'preference': -40}
,
212 '246': {'ext': 'webm', 'height': 480, 'resolution': '480p', 'format_note': 'DASH webm', 'preference': -40}
,
213 '247': {'ext': 'webm', 'height': 720, 'resolution': '720p', 'format_note': 'DASH webm', 'preference': -40}
,
214 '248': {'ext': 'webm', 'height': 1080, 'resolution': '1080p', 'format_note': 'DASH webm', 'preference': -40}
,
217 '171': {'ext': 'webm', 'vcodec': 'none', 'format_note': 'DASH webm audio', 'abr': 48, 'preference': -50}
,
218 '172': {'ext': 'webm', 'vcodec': 'none', 'format_note': 'DASH webm audio', 'abr': 256, 'preference': -50}
,
221 '_rtmp': {'protocol': 'rtmp'}
,
227 u
"url": u
"http://www.youtube.com/watch?v=BaW_jenozKc",
228 u
"file": u
"BaW_jenozKc.mp4",
230 u
"title": u
"youtube-dl test video \"'/\\ä↭𝕐",
231 u
"uploader": u
"Philipp Hagemeister",
232 u
"uploader_id": u
"phihag",
233 u
"upload_date": u
"20121002",
234 u
"description": u
"test chars: \"'/\\ä↭𝕐\ntest URL: https://github.com/rg3/youtube-dl/issues/1892\n\nThis is a test video for youtube-dl.\n\nFor more information, contact phihag@phihag.de ."
238 u
"url": u
"http://www.youtube.com/watch?v=UxxajLWwzqY",
239 u
"file": u
"UxxajLWwzqY.mp4",
240 u
"note": u
"Test generic use_cipher_signature video (#897)",
242 u
"upload_date": u
"20120506",
243 u
"title": u
"Icona Pop - I Love It (feat. Charli XCX) [OFFICIAL VIDEO]",
244 u
"description": u
"md5:5b292926389560516e384ac437c0ec07",
245 u
"uploader": u
"Icona Pop",
246 u
"uploader_id": u
"IconaPop"
250 u
"url": u
"https://www.youtube.com/watch?v=07FYdnEawAQ",
251 u
"file": u
"07FYdnEawAQ.mp4",
252 u
"note": u
"Test VEVO video with age protection (#956)",
254 u
"upload_date": u
"20130703",
255 u
"title": u
"Justin Timberlake - Tunnel Vision (Explicit)",
256 u
"description": u
"md5:64249768eec3bc4276236606ea996373",
257 u
"uploader": u
"justintimberlakeVEVO",
258 u
"uploader_id": u
"justintimberlakeVEVO"
262 u
"url": u
"//www.YouTube.com/watch?v=yZIXLfi8CZQ",
263 u
"file": u
"yZIXLfi8CZQ.mp4",
264 u
"note": u
"Embed-only video (#1746)",
266 u
"upload_date": u
"20120608",
267 u
"title": u
"Principal Sexually Assaults A Teacher - Episode 117 - 8th June 2012",
268 u
"description": u
"md5:09b78bd971f1e3e289601dfba15ca4f7",
269 u
"uploader": u
"SET India",
270 u
"uploader_id": u
"setindia"
274 u
"url": u
"http://www.youtube.com/watch?v=a9LDPn-MO4I",
275 u
"file": u
"a9LDPn-MO4I.m4a",
276 u
"note": u
"256k DASH audio (format 141) via DASH manifest",
281 u
"upload_date": "20121002",
282 u
"uploader_id": "8KVIDEO",
283 u
"description": "No description available.",
284 u
"uploader": "8KVIDEO",
285 u
"title": "UHDTV TEST 8K VIDEO.mp4"
292 def suitable(cls
, url
):
293 """Receives a URL and returns True if suitable for this IE."""
294 if YoutubePlaylistIE
.suitable(url
): return False
295 return re
.match(cls
._VALID
_URL
, url
) is not None
297 def __init__(self
, *args
, **kwargs
):
298 super(YoutubeIE
, self
).__init
__(*args
, **kwargs
)
299 self
._player
_cache
= {}
301 def report_video_info_webpage_download(self
, video_id
):
302 """Report attempt to download video info webpage."""
303 self
.to_screen(u
'%s: Downloading video info webpage' % video_id
)
305 def report_information_extraction(self
, video_id
):
306 """Report attempt to extract video information."""
307 self
.to_screen(u
'%s: Extracting video information' % video_id
)
309 def report_unavailable_format(self
, video_id
, format
):
310 """Report extracted video URL."""
311 self
.to_screen(u
'%s: Format %s not available' % (video_id
, format
))
313 def report_rtmp_download(self
):
314 """Indicate the download will use the RTMP protocol."""
315 self
.to_screen(u
'RTMP download detected')
317 def _extract_signature_function(self
, video_id
, player_url
, slen
):
318 id_m
= re
.match(r
'.*-(?P<id>[a-zA-Z0-9_-]+)\.(?P<ext>[a-z]+)$',
320 player_type
= id_m
.group('ext')
321 player_id
= id_m
.group('id')
323 # Read from filesystem cache
324 func_id
= '%s_%s_%d' % (player_type
, player_id
, slen
)
325 assert os
.path
.basename(func_id
) == func_id
326 cache_dir
= get_cachedir(self
._downloader
.params
)
328 cache_enabled
= cache_dir
is not None
330 cache_fn
= os
.path
.join(os
.path
.expanduser(cache_dir
),
334 with io
.open(cache_fn
, 'r', encoding
='utf-8') as cachef
:
335 cache_spec
= json
.load(cachef
)
336 return lambda s
: u
''.join(s
[i
] for i
in cache_spec
)
338 pass # No cache available
340 if player_type
== 'js':
341 code
= self
._download
_webpage
(
342 player_url
, video_id
,
343 note
=u
'Downloading %s player %s' % (player_type
, player_id
),
344 errnote
=u
'Download of %s failed' % player_url
)
345 res
= self
._parse
_sig
_js
(code
)
346 elif player_type
== 'swf':
347 urlh
= self
._request
_webpage
(
348 player_url
, video_id
,
349 note
=u
'Downloading %s player %s' % (player_type
, player_id
),
350 errnote
=u
'Download of %s failed' % player_url
)
352 res
= self
._parse
_sig
_swf
(code
)
354 assert False, 'Invalid player type %r' % player_type
358 test_string
= u
''.join(map(compat_chr
, range(slen
)))
359 cache_res
= res(test_string
)
360 cache_spec
= [ord(c
) for c
in cache_res
]
362 os
.makedirs(os
.path
.dirname(cache_fn
))
363 except OSError as ose
:
364 if ose
.errno
!= errno
.EEXIST
:
366 write_json_file(cache_spec
, cache_fn
)
368 tb
= traceback
.format_exc()
369 self
._downloader
.report_warning(
370 u
'Writing cache to %r failed: %s' % (cache_fn
, tb
))
374 def _print_sig_code(self
, func
, slen
):
375 def gen_sig_code(idxs
):
376 def _genslice(start
, end
, step
):
377 starts
= u
'' if start
== 0 else str(start
)
378 ends
= (u
':%d' % (end
+step
)) if end
+ step
>= 0 else u
':'
379 steps
= u
'' if step
== 1 else (u
':%d' % step
)
380 return u
's[%s%s%s]' % (starts
, ends
, steps
)
383 start
= '(Never used)' # Quelch pyflakes warnings - start will be
384 # set as soon as step is set
385 for i
, prev
in zip(idxs
[1:], idxs
[:-1]):
389 yield _genslice(start
, prev
, step
)
392 if i
- prev
in [-1, 1]:
397 yield u
's[%d]' % prev
401 yield _genslice(start
, i
, step
)
403 test_string
= u
''.join(map(compat_chr
, range(slen
)))
404 cache_res
= func(test_string
)
405 cache_spec
= [ord(c
) for c
in cache_res
]
406 expr_code
= u
' + '.join(gen_sig_code(cache_spec
))
407 code
= u
'if len(s) == %d:\n return %s\n' % (slen
, expr_code
)
408 self
.to_screen(u
'Extracted signature function:\n' + code
)
410 def _parse_sig_js(self
, jscode
):
411 funcname
= self
._search
_regex
(
412 r
'signature=([a-zA-Z]+)', jscode
,
413 u
'Initial JS player signature function name')
418 return string
.lowercase
.index(varname
)
420 def interpret_statement(stmt
, local_vars
, allow_recursion
=20):
421 if allow_recursion
< 0:
422 raise ExtractorError(u
'Recursion limit reached')
424 if stmt
.startswith(u
'var '):
425 stmt
= stmt
[len(u
'var '):]
426 ass_m
= re
.match(r
'^(?P<out>[a-z]+)(?:\[(?P<index>[^\]]+)\])?' +
427 r
'=(?P<expr>.*)$', stmt
)
429 if ass_m
.groupdict().get('index'):
431 lvar
= local_vars
[ass_m
.group('out')]
432 idx
= interpret_expression(ass_m
.group('index'),
433 local_vars
, allow_recursion
)
434 assert isinstance(idx
, int)
437 expr
= ass_m
.group('expr')
440 local_vars
[ass_m
.group('out')] = val
442 expr
= ass_m
.group('expr')
443 elif stmt
.startswith(u
'return '):
445 expr
= stmt
[len(u
'return '):]
447 raise ExtractorError(
448 u
'Cannot determine left side of statement in %r' % stmt
)
450 v
= interpret_expression(expr
, local_vars
, allow_recursion
)
453 def interpret_expression(expr
, local_vars
, allow_recursion
):
458 return local_vars
[expr
]
460 m
= re
.match(r
'^(?P<in>[a-z]+)\.(?P<member>.*)$', expr
)
462 member
= m
.group('member')
463 val
= local_vars
[m
.group('in')]
464 if member
== 'split("")':
466 if member
== 'join("")':
468 if member
== 'length':
470 if member
== 'reverse()':
472 slice_m
= re
.match(r
'slice\((?P<idx>.*)\)', member
)
474 idx
= interpret_expression(
475 slice_m
.group('idx'), local_vars
, allow_recursion
-1)
479 r
'^(?P<in>[a-z]+)\[(?P<idx>.+)\]$', expr
)
481 val
= local_vars
[m
.group('in')]
482 idx
= interpret_expression(m
.group('idx'), local_vars
,
486 m
= re
.match(r
'^(?P<a>.+?)(?P<op>[%])(?P<b>.+?)$', expr
)
488 a
= interpret_expression(m
.group('a'),
489 local_vars
, allow_recursion
)
490 b
= interpret_expression(m
.group('b'),
491 local_vars
, allow_recursion
)
495 r
'^(?P<func>[a-zA-Z]+)\((?P<args>[a-z0-9,]+)\)$', expr
)
497 fname
= m
.group('func')
498 if fname
not in functions
:
499 functions
[fname
] = extract_function(fname
)
500 argvals
= [int(v
) if v
.isdigit() else local_vars
[v
]
501 for v
in m
.group('args').split(',')]
502 return functions
[fname
](argvals
)
503 raise ExtractorError(u
'Unsupported JS expression %r' % expr
)
505 def extract_function(funcname
):
507 r
'function ' + re
.escape(funcname
) +
508 r
'\((?P<args>[a-z,]+)\){(?P<code>[^}]+)}',
510 argnames
= func_m
.group('args').split(',')
513 local_vars
= dict(zip(argnames
, args
))
514 for stmt
in func_m
.group('code').split(';'):
515 res
= interpret_statement(stmt
, local_vars
)
519 initial_function
= extract_function(funcname
)
520 return lambda s
: initial_function([s
])
522 def _parse_sig_swf(self
, file_contents
):
523 if file_contents
[1:3] != b
'WS':
524 raise ExtractorError(
525 u
'Not an SWF file; header is %r' % file_contents
[:3])
526 if file_contents
[:1] == b
'C':
527 content
= zlib
.decompress(file_contents
[8:])
529 raise NotImplementedError(u
'Unsupported compression format %r' %
532 def extract_tags(content
):
534 while pos
< len(content
):
535 header16
= struct
.unpack('<H', content
[pos
:pos
+2])[0]
537 tag_code
= header16
>> 6
538 tag_len
= header16
& 0x3f
540 tag_len
= struct
.unpack('<I', content
[pos
:pos
+4])[0]
542 assert pos
+tag_len
<= len(content
)
543 yield (tag_code
, content
[pos
:pos
+tag_len
])
547 for tag_code
, tag
in extract_tags(content
)
549 p
= code_tag
.index(b
'\0', 4) + 1
550 code_reader
= io
.BytesIO(code_tag
[p
:])
552 # Parse ABC (AVM2 ByteCode)
553 def read_int(reader
=None):
561 b
= struct
.unpack('<B', buf
)[0]
562 res
= res |
((b
& 0x7f) << shift
)
568 def u30(reader
=None):
569 res
= read_int(reader
)
570 assert res
& 0xf0000000 == 0
574 def s32(reader
=None):
576 if v
& 0x80000000 != 0:
577 v
= - ((v ^
0xffffffff) + 1)
580 def read_string(reader
=None):
584 resb
= reader
.read(slen
)
585 assert len(resb
) == slen
586 return resb
.decode('utf-8')
588 def read_bytes(count
, reader
=None):
591 resb
= reader
.read(count
)
592 assert len(resb
) == count
595 def read_byte(reader
=None):
596 resb
= read_bytes(1, reader
=reader
)
597 res
= struct
.unpack('<B', resb
)[0]
600 # minor_version + major_version
605 for _c
in range(1, int_count
):
608 for _c
in range(1, uint_count
):
611 read_bytes((double_count
-1) * 8)
613 constant_strings
= [u
'']
614 for _c
in range(1, string_count
):
616 constant_strings
.append(s
)
617 namespace_count
= u30()
618 for _c
in range(1, namespace_count
):
622 for _c
in range(1, ns_set_count
):
624 for _c2
in range(count
):
626 multiname_count
= u30()
635 0x0e: 2, # MultinameA
636 0x1b: 1, # MultinameL
637 0x1c: 1, # MultinameLA
640 for _c
in range(1, multiname_count
):
642 assert kind
in MULTINAME_SIZES
, u
'Invalid multiname kind %r' % kind
644 u30() # namespace_idx
646 multinames
.append(constant_strings
[name_idx
])
648 multinames
.append('[MULTINAME kind: %d]' % kind
)
649 for _c2
in range(MULTINAME_SIZES
[kind
]):
654 MethodInfo
= collections
.namedtuple(
656 ['NEED_ARGUMENTS', 'NEED_REST'])
658 for method_id
in range(method_count
):
661 for _
in range(param_count
):
663 u30() # name index (always 0 for youtube)
665 if flags
& 0x08 != 0:
668 for c
in range(option_count
):
671 if flags
& 0x80 != 0:
672 # Param names present
673 for _
in range(param_count
):
675 mi
= MethodInfo(flags
& 0x01 != 0, flags
& 0x04 != 0)
676 method_infos
.append(mi
)
679 metadata_count
= u30()
680 for _c
in range(metadata_count
):
683 for _c2
in range(item_count
):
687 def parse_traits_info():
688 trait_name_idx
= u30()
689 kind_full
= read_byte()
690 kind
= kind_full
& 0x0f
691 attrs
= kind_full
>> 4
693 if kind
in [0x00, 0x06]: # Slot or Const
695 u30() # type_name_idx
699 elif kind
in [0x01, 0x02, 0x03]: # Method / Getter / Setter
702 methods
[multinames
[trait_name_idx
]] = method_idx
703 elif kind
== 0x04: # Class
706 elif kind
== 0x05: # Function
709 methods
[function_idx
] = multinames
[trait_name_idx
]
711 raise ExtractorError(u
'Unsupported trait kind %d' % kind
)
713 if attrs
& 0x4 != 0: # Metadata present
714 metadata_count
= u30()
715 for _c3
in range(metadata_count
):
716 u30() # metadata index
721 TARGET_CLASSNAME
= u
'SignatureDecipher'
722 searched_idx
= multinames
.index(TARGET_CLASSNAME
)
723 searched_class_id
= None
725 for class_id
in range(class_count
):
727 if name_idx
== searched_idx
:
728 # We found the class we're looking for!
729 searched_class_id
= class_id
730 u30() # super_name idx
732 if flags
& 0x08 != 0: # Protected namespace is present
733 u30() # protected_ns_idx
735 for _c2
in range(intrf_count
):
739 for _c2
in range(trait_count
):
742 if searched_class_id
is None:
743 raise ExtractorError(u
'Target class %r not found' %
748 for class_id
in range(class_count
):
751 for _c2
in range(trait_count
):
752 trait_methods
= parse_traits_info()
753 if class_id
== searched_class_id
:
754 method_names
.update(trait_methods
.items())
755 method_idxs
.update(dict(
757 for name
, idx
in trait_methods
.items()))
761 for _c
in range(script_count
):
764 for _c2
in range(trait_count
):
768 method_body_count
= u30()
769 Method
= collections
.namedtuple('Method', ['code', 'local_count'])
771 for _c
in range(method_body_count
):
775 u30() # init_scope_depth
776 u30() # max_scope_depth
778 code
= read_bytes(code_length
)
779 if method_idx
in method_idxs
:
780 m
= Method(code
, local_count
)
781 methods
[method_idxs
[method_idx
]] = m
782 exception_count
= u30()
783 for _c2
in range(exception_count
):
790 for _c2
in range(trait_count
):
793 assert p
+ code_reader
.tell() == len(code_tag
)
794 assert len(methods
) == len(method_idxs
)
796 method_pyfunctions
= {}
798 def extract_function(func_name
):
799 if func_name
in method_pyfunctions
:
800 return method_pyfunctions
[func_name
]
801 if func_name
not in methods
:
802 raise ExtractorError(u
'Cannot find function %r' % func_name
)
803 m
= methods
[func_name
]
806 registers
= ['(this)'] + list(args
) + [None] * m
.local_count
808 coder
= io
.BytesIO(m
.code
)
810 opcode
= struct
.unpack('!B', coder
.read(1))[0]
811 if opcode
== 36: # pushbyte
812 v
= struct
.unpack('!B', coder
.read(1))[0]
814 elif opcode
== 44: # pushstring
816 stack
.append(constant_strings
[idx
])
817 elif opcode
== 48: # pushscope
818 # We don't implement the scope register, so we'll just
819 # ignore the popped value
821 elif opcode
== 70: # callproperty
823 mname
= multinames
[index
]
824 arg_count
= u30(coder
)
825 args
= list(reversed(
826 [stack
.pop() for _
in range(arg_count
)]))
828 if mname
== u
'split':
829 assert len(args
) == 1
830 assert isinstance(args
[0], compat_str
)
831 assert isinstance(obj
, compat_str
)
835 res
= obj
.split(args
[0])
837 elif mname
== u
'slice':
838 assert len(args
) == 1
839 assert isinstance(args
[0], int)
840 assert isinstance(obj
, list)
843 elif mname
== u
'join':
844 assert len(args
) == 1
845 assert isinstance(args
[0], compat_str
)
846 assert isinstance(obj
, list)
847 res
= args
[0].join(obj
)
849 elif mname
in method_pyfunctions
:
850 stack
.append(method_pyfunctions
[mname
](args
))
852 raise NotImplementedError(
853 u
'Unsupported property %r on %r'
855 elif opcode
== 72: # returnvalue
858 elif opcode
== 79: # callpropvoid
860 mname
= multinames
[index
]
861 arg_count
= u30(coder
)
862 args
= list(reversed(
863 [stack
.pop() for _
in range(arg_count
)]))
865 if mname
== u
'reverse':
866 assert isinstance(obj
, list)
869 raise NotImplementedError(
870 u
'Unsupported (void) property %r on %r'
872 elif opcode
== 93: # findpropstrict
874 mname
= multinames
[index
]
875 res
= extract_function(mname
)
877 elif opcode
== 97: # setproperty
882 assert isinstance(obj
, list)
883 assert isinstance(idx
, int)
885 elif opcode
== 98: # getlocal
887 stack
.append(registers
[index
])
888 elif opcode
== 99: # setlocal
891 registers
[index
] = value
892 elif opcode
== 102: # getproperty
894 pname
= multinames
[index
]
895 if pname
== u
'length':
897 assert isinstance(obj
, list)
898 stack
.append(len(obj
))
899 else: # Assume attribute access
901 assert isinstance(idx
, int)
903 assert isinstance(obj
, list)
904 stack
.append(obj
[idx
])
905 elif opcode
== 128: # coerce
907 elif opcode
== 133: # coerce_s
908 assert isinstance(stack
[-1], (type(None), compat_str
))
909 elif opcode
== 164: # modulo
912 res
= value1
% value2
914 elif opcode
== 208: # getlocal_0
915 stack
.append(registers
[0])
916 elif opcode
== 209: # getlocal_1
917 stack
.append(registers
[1])
918 elif opcode
== 210: # getlocal_2
919 stack
.append(registers
[2])
920 elif opcode
== 211: # getlocal_3
921 stack
.append(registers
[3])
922 elif opcode
== 214: # setlocal_2
923 registers
[2] = stack
.pop()
924 elif opcode
== 215: # setlocal_3
925 registers
[3] = stack
.pop()
927 raise NotImplementedError(
928 u
'Unsupported opcode %d' % opcode
)
930 method_pyfunctions
[func_name
] = resfunc
933 initial_function
= extract_function(u
'decipher')
934 return lambda s
: initial_function([s
])
936 def _decrypt_signature(self
, s
, video_id
, player_url
, age_gate
=False):
937 """Turn the encrypted s field into a working signature"""
939 if player_url
is not None:
940 if player_url
.startswith(u
'//'):
941 player_url
= u
'https:' + player_url
943 player_id
= (player_url
, len(s
))
944 if player_id
not in self
._player
_cache
:
945 func
= self
._extract
_signature
_function
(
946 video_id
, player_url
, len(s
)
948 self
._player
_cache
[player_id
] = func
949 func
= self
._player
_cache
[player_id
]
950 if self
._downloader
.params
.get('youtube_print_sig_code'):
951 self
._print
_sig
_code
(func
, len(s
))
954 tb
= traceback
.format_exc()
955 self
._downloader
.report_warning(
956 u
'Automatic signature extraction failed: ' + tb
)
958 self
._downloader
.report_warning(
959 u
'Warning: Falling back to static signature algorithm')
961 return self
._static
_decrypt
_signature
(
962 s
, video_id
, player_url
, age_gate
)
964 def _static_decrypt_signature(self
, s
, video_id
, player_url
, age_gate
):
966 # The videos with age protection use another player, so the
967 # algorithms can be different.
969 return s
[2:63] + s
[82] + s
[64:82] + s
[63]
972 return s
[86:29:-1] + s
[88] + s
[28:5:-1]
974 return s
[25] + s
[3:25] + s
[0] + s
[26:42] + s
[79] + s
[43:79] + s
[91] + s
[80:83]
976 return s
[84:27:-1] + s
[86] + s
[26:5:-1]
978 return s
[25] + s
[3:25] + s
[2] + s
[26:40] + s
[77] + s
[41:77] + s
[89] + s
[78:81]
980 return s
[84:78:-1] + s
[87] + s
[77:60:-1] + s
[0] + s
[59:3:-1]
982 return s
[7:28] + s
[87] + s
[29:45] + s
[55] + s
[46:55] + s
[2] + s
[56:87] + s
[28]
984 return s
[6:27] + s
[4] + s
[28:39] + s
[27] + s
[40:59] + s
[2] + s
[60:]
986 return s
[80:72:-1] + s
[16] + s
[71:39:-1] + s
[72] + s
[38:16:-1] + s
[82] + s
[15::-1]
988 return s
[3:11] + s
[0] + s
[12:55] + s
[84] + s
[56:84]
990 return s
[78:70:-1] + s
[14] + s
[69:37:-1] + s
[70] + s
[36:14:-1] + s
[80] + s
[:14][::-1]
992 return s
[80:63:-1] + s
[0] + s
[62:0:-1] + s
[63]
994 return s
[80:37:-1] + s
[7] + s
[36:7:-1] + s
[0] + s
[6:0:-1] + s
[37]
996 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]
998 return s
[1:19] + s
[0] + s
[20:68] + s
[19] + s
[69:80]
1000 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]
1003 raise ExtractorError(u
'Unable to decrypt signature, key length %d not supported; retrying might work' % (len(s
)))
1005 def _get_available_subtitles(self
, video_id
, webpage
):
1007 sub_list
= self
._download
_webpage
(
1008 'http://video.google.com/timedtext?hl=en&type=list&v=%s' % video_id
,
1009 video_id
, note
=False)
1010 except ExtractorError
as err
:
1011 self
._downloader
.report_warning(u
'unable to download video subtitles: %s' % compat_str(err
))
1013 lang_list
= re
.findall(r
'name="([^"]*)"[^>]+lang_code="([\w\-]+)"', sub_list
)
1018 params
= compat_urllib_parse
.urlencode({
1021 'fmt': self
._downloader
.params
.get('subtitlesformat', 'srt'),
1022 'name': unescapeHTML(l
[0]).encode('utf-8'),
1024 url
= u
'http://www.youtube.com/api/timedtext?' + params
1025 sub_lang_list
[lang
] = url
1026 if not sub_lang_list
:
1027 self
._downloader
.report_warning(u
'video doesn\'t have subtitles')
1029 return sub_lang_list
1031 def _get_available_automatic_caption(self
, video_id
, webpage
):
1032 """We need the webpage for getting the captions url, pass it as an
1033 argument to speed up the process."""
1034 sub_format
= self
._downloader
.params
.get('subtitlesformat', 'srt')
1035 self
.to_screen(u
'%s: Looking for automatic captions' % video_id
)
1036 mobj
= re
.search(r
';ytplayer.config = ({.*?});', webpage
)
1037 err_msg
= u
'Couldn\'t find automatic captions for %s' % video_id
1039 self
._downloader
.report_warning(err_msg
)
1041 player_config
= json
.loads(mobj
.group(1))
1043 args
= player_config
[u
'args']
1044 caption_url
= args
[u
'ttsurl']
1045 timestamp
= args
[u
'timestamp']
1046 # We get the available subtitles
1047 list_params
= compat_urllib_parse
.urlencode({
1052 list_url
= caption_url
+ '&' + list_params
1053 caption_list
= self
._download
_xml
(list_url
, video_id
)
1054 original_lang_node
= caption_list
.find('track')
1055 if original_lang_node
is None or original_lang_node
.attrib
.get('kind') != 'asr' :
1056 self
._downloader
.report_warning(u
'Video doesn\'t have automatic captions')
1058 original_lang
= original_lang_node
.attrib
['lang_code']
1061 for lang_node
in caption_list
.findall('target'):
1062 sub_lang
= lang_node
.attrib
['lang_code']
1063 params
= compat_urllib_parse
.urlencode({
1064 'lang': original_lang
,
1070 sub_lang_list
[sub_lang
] = caption_url
+ '&' + params
1071 return sub_lang_list
1072 # An extractor error can be raise by the download process if there are
1073 # no automatic captions but there are subtitles
1074 except (KeyError, ExtractorError
):
1075 self
._downloader
.report_warning(err_msg
)
1078 def _extract_id(self
, url
):
1079 mobj
= re
.match(self
._VALID
_URL
, url
, re
.VERBOSE
)
1081 raise ExtractorError(u
'Invalid URL: %s' % url
)
1082 video_id
= mobj
.group(2)
1085 def _extract_from_m3u8(self
, manifest_url
, video_id
):
1087 def _get_urls(_manifest
):
1088 lines
= _manifest
.split('\n')
1089 urls
= filter(lambda l
: l
and not l
.startswith('#'),
1092 manifest
= self
._download
_webpage
(manifest_url
, video_id
, u
'Downloading formats manifest')
1093 formats_urls
= _get_urls(manifest
)
1094 for format_url
in formats_urls
:
1095 itag
= self
._search
_regex
(r
'itag/(\d+?)/', format_url
, 'itag')
1096 url_map
[itag
] = format_url
1099 def _extract_annotations(self
, video_id
):
1100 url
= 'https://www.youtube.com/annotations_invideo?features=1&legacy=1&video_id=%s' % video_id
1101 return self
._download
_webpage
(url
, video_id
, note
=u
'Searching for annotations.', errnote
=u
'Unable to download video annotations.')
1103 def _real_extract(self
, url
):
1104 # Extract original video URL from URL with redirection, like age verification, using next_url parameter
1105 mobj
= re
.search(self
._NEXT
_URL
_RE
, url
)
1107 url
= 'https://www.youtube.com/' + compat_urllib_parse
.unquote(mobj
.group(1)).lstrip('/')
1108 video_id
= self
._extract
_id
(url
)
1111 url
= 'https://www.youtube.com/watch?v=%s&gl=US&hl=en&has_verified=1' % video_id
1112 video_webpage
= self
._download
_webpage
(url
, video_id
)
1114 # Attempt to extract SWF player URL
1115 mobj
= re
.search(r
'swfConfig.*?"(https?:\\/\\/.*?watch.*?-.*?\.swf)"', video_webpage
)
1116 if mobj
is not None:
1117 player_url
= re
.sub(r
'\\(.)', r
'\1', mobj
.group(1))
1122 self
.report_video_info_webpage_download(video_id
)
1123 if re
.search(r
'player-age-gate-content">', video_webpage
) is not None:
1124 self
.report_age_confirmation()
1126 # We simulate the access to the video from www.youtube.com/v/{video_id}
1127 # this can be viewed without login into Youtube
1128 data
= compat_urllib_parse
.urlencode({'video_id': video_id
,
1129 'el': 'player_embedded',
1132 'eurl': 'https://youtube.googleapis.com/v/' + video_id
,
1136 video_info_url
= 'https://www.youtube.com/get_video_info?' + data
1137 video_info_webpage
= self
._download
_webpage
(video_info_url
, video_id
,
1139 errnote
='unable to download video info webpage')
1140 video_info
= compat_parse_qs(video_info_webpage
)
1143 for el_type
in ['&el=embedded', '&el=detailpage', '&el=vevo', '']:
1144 video_info_url
= ('https://www.youtube.com/get_video_info?&video_id=%s%s&ps=default&eurl=&gl=US&hl=en'
1145 % (video_id
, el_type
))
1146 video_info_webpage
= self
._download
_webpage
(video_info_url
, video_id
,
1148 errnote
='unable to download video info webpage')
1149 video_info
= compat_parse_qs(video_info_webpage
)
1150 if 'token' in video_info
:
1152 if 'token' not in video_info
:
1153 if 'reason' in video_info
:
1154 raise ExtractorError(u
'YouTube said: %s' % video_info
['reason'][0], expected
=True)
1156 raise ExtractorError(u
'"token" parameter not in video info for unknown reason')
1158 if 'view_count' in video_info
:
1159 view_count
= int(video_info
['view_count'][0])
1163 # Check for "rental" videos
1164 if 'ypc_video_rental_bar_text' in video_info
and 'author' not in video_info
:
1165 raise ExtractorError(u
'"rental" videos not supported')
1167 # Start extracting information
1168 self
.report_information_extraction(video_id
)
1171 if 'author' not in video_info
:
1172 raise ExtractorError(u
'Unable to extract uploader name')
1173 video_uploader
= compat_urllib_parse
.unquote_plus(video_info
['author'][0])
1176 video_uploader_id
= None
1177 mobj
= re
.search(r
'<link itemprop="url" href="http://www.youtube.com/(?:user|channel)/([^"]+)">', video_webpage
)
1178 if mobj
is not None:
1179 video_uploader_id
= mobj
.group(1)
1181 self
._downloader
.report_warning(u
'unable to extract uploader nickname')
1184 if 'title' in video_info
:
1185 video_title
= compat_urllib_parse
.unquote_plus(video_info
['title'][0])
1187 self
._downloader
.report_warning(u
'Unable to extract video title')
1191 # We try first to get a high quality image:
1192 m_thumb
= re
.search(r
'<span itemprop="thumbnail".*?href="(.*?)">',
1193 video_webpage
, re
.DOTALL
)
1194 if m_thumb
is not None:
1195 video_thumbnail
= m_thumb
.group(1)
1196 elif 'thumbnail_url' not in video_info
:
1197 self
._downloader
.report_warning(u
'unable to extract video thumbnail')
1198 video_thumbnail
= None
1199 else: # don't panic if we can't find it
1200 video_thumbnail
= compat_urllib_parse
.unquote_plus(video_info
['thumbnail_url'][0])
1204 mobj
= re
.search(r
'id="eow-date.*?>(.*?)</span>', video_webpage
, re
.DOTALL
)
1205 if mobj
is not None:
1206 upload_date
= ' '.join(re
.sub(r
'[/,-]', r
' ', mobj
.group(1)).split())
1207 upload_date
= unified_strdate(upload_date
)
1210 video_description
= get_element_by_id("eow-description", video_webpage
)
1211 if video_description
:
1212 video_description
= re
.sub(r
'''(?x)
1214 (?:[a-zA-Z-]+="[^"]+"\s+)*?
1216 (?:[a-zA-Z-]+="[^"]+"\s+)*?
1217 class="yt-uix-redirect-link"\s*>
1220 ''', r
'\1', video_description
)
1221 video_description
= clean_html(video_description
)
1223 fd_mobj
= re
.search(r
'<meta name="description" content="([^"]+)"', video_webpage
)
1225 video_description
= unescapeHTML(fd_mobj
.group(1))
1227 video_description
= u
''
1229 def _extract_count(klass
):
1230 count
= self
._search
_regex
(
1231 r
'class="%s">([\d,]+)</span>' % re
.escape(klass
),
1232 video_webpage
, klass
, default
=None)
1233 if count
is not None:
1234 return int(count
.replace(',', ''))
1236 like_count
= _extract_count(u
'likes-count')
1237 dislike_count
= _extract_count(u
'dislikes-count')
1240 video_subtitles
= self
.extract_subtitles(video_id
, video_webpage
)
1242 if self
._downloader
.params
.get('listsubtitles', False):
1243 self
._list
_available
_subtitles
(video_id
, video_webpage
)
1246 if 'length_seconds' not in video_info
:
1247 self
._downloader
.report_warning(u
'unable to extract video duration')
1248 video_duration
= None
1250 video_duration
= int(compat_urllib_parse
.unquote_plus(video_info
['length_seconds'][0]))
1253 video_annotations
= None
1254 if self
._downloader
.params
.get('writeannotations', False):
1255 video_annotations
= self
._extract
_annotations
(video_id
)
1257 # Decide which formats to download
1259 mobj
= re
.search(r
';ytplayer.config = ({.*?});', video_webpage
)
1261 raise ValueError('Could not find vevo ID')
1262 info
= json
.loads(mobj
.group(1))
1264 # Easy way to know if the 's' value is in url_encoded_fmt_stream_map
1265 # this signatures are encrypted
1266 if 'url_encoded_fmt_stream_map' not in args
:
1267 raise ValueError(u
'No stream_map present') # caught below
1268 re_signature
= re
.compile(r
'[&,]s=')
1269 m_s
= re_signature
.search(args
['url_encoded_fmt_stream_map'])
1271 self
.to_screen(u
'%s: Encrypted signatures detected.' % video_id
)
1272 video_info
['url_encoded_fmt_stream_map'] = [args
['url_encoded_fmt_stream_map']]
1273 m_s
= re_signature
.search(args
.get('adaptive_fmts', u
''))
1275 if 'adaptive_fmts' in video_info
:
1276 video_info
['adaptive_fmts'][0] += ',' + args
['adaptive_fmts']
1278 video_info
['adaptive_fmts'] = [args
['adaptive_fmts']]
1282 def _map_to_format_list(urlmap
):
1284 for itag
, video_real_url
in urlmap
.items():
1287 'url': video_real_url
,
1288 'player_url': player_url
,
1290 dct
.update(self
._formats
[itag
])
1294 if 'conn' in video_info
and video_info
['conn'][0].startswith('rtmp'):
1295 self
.report_rtmp_download()
1297 'format_id': '_rtmp',
1299 'url': video_info
['conn'][0],
1300 'player_url': player_url
,
1302 elif len(video_info
.get('url_encoded_fmt_stream_map', [])) >= 1 or len(video_info
.get('adaptive_fmts', [])) >= 1:
1303 encoded_url_map
= video_info
.get('url_encoded_fmt_stream_map', [''])[0] + ',' + video_info
.get('adaptive_fmts',[''])[0]
1304 if 'rtmpe%3Dyes' in encoded_url_map
:
1305 raise ExtractorError('rtmpe downloads are not supported, see https://github.com/rg3/youtube-dl/issues/343 for more information.', expected
=True)
1307 for url_data_str
in encoded_url_map
.split(','):
1308 url_data
= compat_parse_qs(url_data_str
)
1309 if 'itag' in url_data
and 'url' in url_data
:
1310 url
= url_data
['url'][0]
1311 if 'sig' in url_data
:
1312 url
+= '&signature=' + url_data
['sig'][0]
1313 elif 's' in url_data
:
1314 encrypted_sig
= url_data
['s'][0]
1315 if self
._downloader
.params
.get('verbose'):
1317 if player_url
is None:
1318 player_version
= 'unknown'
1320 player_version
= self
._search
_regex
(
1321 r
'-(.+)\.swf$', player_url
,
1322 u
'flash player', fatal
=False)
1323 player_desc
= 'flash player %s' % player_version
1325 player_version
= self
._search
_regex
(
1326 r
'html5player-(.+?)\.js', video_webpage
,
1327 'html5 player', fatal
=False)
1328 player_desc
= u
'html5 player %s' % player_version
1330 parts_sizes
= u
'.'.join(compat_str(len(part
)) for part
in encrypted_sig
.split('.'))
1331 self
.to_screen(u
'encrypted signature length %d (%s), itag %s, %s' %
1332 (len(encrypted_sig
), parts_sizes
, url_data
['itag'][0], player_desc
))
1335 jsplayer_url_json
= self
._search
_regex
(
1336 r
'"assets":.+?"js":\s*("[^"]+")',
1337 video_webpage
, u
'JS player URL')
1338 player_url
= json
.loads(jsplayer_url_json
)
1340 signature
= self
._decrypt
_signature
(
1341 encrypted_sig
, video_id
, player_url
, age_gate
)
1342 url
+= '&signature=' + signature
1343 if 'ratebypass' not in url
:
1344 url
+= '&ratebypass=yes'
1345 url_map
[url_data
['itag'][0]] = url
1346 formats
= _map_to_format_list(url_map
)
1347 elif video_info
.get('hlsvp'):
1348 manifest_url
= video_info
['hlsvp'][0]
1349 url_map
= self
._extract
_from
_m
3u8(manifest_url
, video_id
)
1350 formats
= _map_to_format_list(url_map
)
1352 raise ExtractorError(u
'no conn, hlsvp or url_encoded_fmt_stream_map information found in video info')
1354 # Look for the DASH manifest
1355 dash_manifest_url_lst
= video_info
.get('dashmpd')
1356 if dash_manifest_url_lst
and dash_manifest_url_lst
[0]:
1358 dash_doc
= self
._download
_xml
(
1359 dash_manifest_url_lst
[0], video_id
,
1360 note
=u
'Downloading DASH manifest',
1361 errnote
=u
'Could not download DASH manifest')
1362 for r
in dash_doc
.findall(u
'.//{urn:mpeg:DASH:schema:MPD:2011}Representation'):
1363 url_el
= r
.find('{urn:mpeg:DASH:schema:MPD:2011}BaseURL')
1366 format_id
= r
.attrib
['id']
1367 video_url
= url_el
.text
1368 filesize
= int_or_none(url_el
.attrib
.get('{http://youtube.com/yt/2012/10/10}contentLength'))
1370 'format_id': format_id
,
1372 'width': int_or_none(r
.attrib
.get('width')),
1373 'tbr': int_or_none(r
.attrib
.get('bandwidth'), 1000),
1374 'asr': int_or_none(r
.attrib
.get('audioSamplingRate')),
1375 'filesize': filesize
,
1378 existing_format
= next(
1379 fo
for fo
in formats
1380 if fo
['format_id'] == format_id
)
1381 except StopIteration:
1382 f
.update(self
._formats
.get(format_id
, {}))
1385 existing_format
.update(f
)
1387 except (ExtractorError
, KeyError) as e
:
1388 self
.report_warning(u
'Skipping DASH manifest: %s' % e
, video_id
)
1390 self
._sort
_formats
(formats
)
1394 'uploader': video_uploader
,
1395 'uploader_id': video_uploader_id
,
1396 'upload_date': upload_date
,
1397 'title': video_title
,
1398 'thumbnail': video_thumbnail
,
1399 'description': video_description
,
1400 'subtitles': video_subtitles
,
1401 'duration': video_duration
,
1402 'age_limit': 18 if age_gate
else 0,
1403 'annotations': video_annotations
,
1404 'webpage_url': 'https://www.youtube.com/watch?v=%s' % video_id
,
1405 'view_count': view_count
,
1406 'like_count': like_count
,
1407 'dislike_count': dislike_count
,
1411 class YoutubePlaylistIE(YoutubeBaseInfoExtractor
):
1412 IE_DESC
= u
'YouTube.com playlists'
1413 _VALID_URL
= r
"""(?:
1418 (?:course|view_play_list|my_playlists|artist|playlist|watch)
1419 \? (?:.*?&)*? (?:p|a|list)=
1422 ((?:PL|EC|UU|FL|RD)?[0-9A-Za-z-_]{10,})
1425 ((?:PL|EC|UU|FL|RD)[0-9A-Za-z-_]{10,})
1427 _TEMPLATE_URL
= 'https://www.youtube.com/playlist?list=%s&page=%s'
1428 _MORE_PAGES_INDICATOR
= r
'data-link-type="next"'
1429 _VIDEO_RE
= r
'href="/watch\?v=(?P<id>[0-9A-Za-z_-]{11})&[^"]*?index=(?P<index>\d+)'
1430 IE_NAME
= u
'youtube:playlist'
1433 def suitable(cls
, url
):
1434 """Receives a URL and returns True if suitable for this IE."""
1435 return re
.match(cls
._VALID
_URL
, url
, re
.VERBOSE
) is not None
1437 def _real_initialize(self
):
1440 def _ids_to_results(self
, ids
):
1441 return [self
.url_result(vid_id
, 'Youtube', video_id
=vid_id
)
1444 def _extract_mix(self
, playlist_id
):
1445 # The mixes are generated from a a single video
1446 # the id of the playlist is just 'RD' + video_id
1447 url
= 'https://youtube.com/watch?v=%s&list=%s' % (playlist_id
[-11:], playlist_id
)
1448 webpage
= self
._download
_webpage
(url
, playlist_id
, u
'Downloading Youtube mix')
1449 title_span
= (get_element_by_attribute('class', 'title long-title', webpage
) or
1450 get_element_by_attribute('class', 'title ', webpage
))
1451 title
= clean_html(title_span
)
1452 video_re
= r
'data-index="\d+".*?href="/watch\?v=([0-9A-Za-z_-]{11})&[^"]*?list=%s' % re
.escape(playlist_id
)
1453 ids
= orderedSet(re
.findall(video_re
, webpage
))
1454 url_results
= self
._ids
_to
_results
(ids
)
1456 return self
.playlist_result(url_results
, playlist_id
, title
)
1458 def _real_extract(self
, url
):
1459 # Extract playlist id
1460 mobj
= re
.match(self
._VALID
_URL
, url
, re
.VERBOSE
)
1462 raise ExtractorError(u
'Invalid URL: %s' % url
)
1463 playlist_id
= mobj
.group(1) or mobj
.group(2)
1465 # Check if it's a video-specific URL
1466 query_dict
= compat_urlparse
.parse_qs(compat_urlparse
.urlparse(url
).query
)
1467 if 'v' in query_dict
:
1468 video_id
= query_dict
['v'][0]
1469 if self
._downloader
.params
.get('noplaylist'):
1470 self
.to_screen(u
'Downloading just video %s because of --no-playlist' % video_id
)
1471 return self
.url_result(video_id
, 'Youtube', video_id
=video_id
)
1473 self
.to_screen(u
'Downloading playlist PL%s - add --no-playlist to just download video %s' % (playlist_id
, video_id
))
1475 if playlist_id
.startswith('RD'):
1476 # Mixes require a custom extraction process
1477 return self
._extract
_mix
(playlist_id
)
1478 if playlist_id
.startswith('TL'):
1479 raise ExtractorError(u
'For downloading YouTube.com top lists, use '
1480 u
'the "yttoplist" keyword, for example "youtube-dl \'yttoplist:music:Top Tracks\'"', expected
=True)
1482 # Extract the video ids from the playlist pages
1485 for page_num
in itertools
.count(1):
1486 url
= self
._TEMPLATE
_URL
% (playlist_id
, page_num
)
1487 page
= self
._download
_webpage
(url
, playlist_id
, u
'Downloading page #%s' % page_num
)
1488 matches
= re
.finditer(self
._VIDEO
_RE
, page
)
1489 # We remove the duplicates and the link with index 0
1490 # (it's not the first video of the playlist)
1491 new_ids
= orderedSet(m
.group('id') for m
in matches
if m
.group('index') != '0')
1494 if re
.search(self
._MORE
_PAGES
_INDICATOR
, page
) is None:
1497 playlist_title
= self
._og
_search
_title
(page
)
1499 url_results
= self
._ids
_to
_results
(ids
)
1500 return self
.playlist_result(url_results
, playlist_id
, playlist_title
)
1503 class YoutubeTopListIE(YoutubePlaylistIE
):
1504 IE_NAME
= u
'youtube:toplist'
1505 IE_DESC
= (u
'YouTube.com top lists, "yttoplist:{channel}:{list title}"'
1506 u
' (Example: "yttoplist:music:Top Tracks")')
1507 _VALID_URL
= r
'yttoplist:(?P<chann>.*?):(?P<title>.*?)$'
1509 def _real_extract(self
, url
):
1510 mobj
= re
.match(self
._VALID
_URL
, url
)
1511 channel
= mobj
.group('chann')
1512 title
= mobj
.group('title')
1513 query
= compat_urllib_parse
.urlencode({'title': title}
)
1514 playlist_re
= 'href="([^"]+?%s[^"]+?)"' % re
.escape(query
)
1515 channel_page
= self
._download
_webpage
('https://www.youtube.com/%s' % channel
, title
)
1516 link
= self
._html
_search
_regex
(playlist_re
, channel_page
, u
'list')
1517 url
= compat_urlparse
.urljoin('https://www.youtube.com/', link
)
1519 video_re
= r
'data-index="\d+".*?data-video-id="([0-9A-Za-z_-]{11})"'
1521 # sometimes the webpage doesn't contain the videos
1522 # retry until we get them
1523 for i
in itertools
.count(0):
1524 msg
= u
'Downloading Youtube mix'
1526 msg
+= ', retry #%d' % i
1527 webpage
= self
._download
_webpage
(url
, title
, msg
)
1528 ids
= orderedSet(re
.findall(video_re
, webpage
))
1531 url_results
= self
._ids
_to
_results
(ids
)
1532 return self
.playlist_result(url_results
, playlist_title
=title
)
1535 class YoutubeChannelIE(InfoExtractor
):
1536 IE_DESC
= u
'YouTube.com channels'
1537 _VALID_URL
= r
"^(?:https?://)?(?:youtu\.be|(?:\w+\.)?youtube(?:-nocookie)?\.com)/channel/([0-9A-Za-z_-]+)"
1538 _MORE_PAGES_INDICATOR
= 'yt-uix-load-more'
1539 _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'
1540 IE_NAME
= u
'youtube:channel'
1542 def extract_videos_from_page(self
, page
):
1544 for mobj
in re
.finditer(r
'href="/watch\?v=([0-9A-Za-z_-]+)&?', page
):
1545 if mobj
.group(1) not in ids_in_page
:
1546 ids_in_page
.append(mobj
.group(1))
1549 def _real_extract(self
, url
):
1550 # Extract channel id
1551 mobj
= re
.match(self
._VALID
_URL
, url
)
1553 raise ExtractorError(u
'Invalid URL: %s' % url
)
1555 # Download channel page
1556 channel_id
= mobj
.group(1)
1558 url
= 'https://www.youtube.com/channel/%s/videos' % channel_id
1559 channel_page
= self
._download
_webpage
(url
, channel_id
)
1560 autogenerated
= re
.search(r
'''(?x)
1562 channel-header-autogenerated-label|
1563 yt-channel-title-autogenerated
1564 )[^"]*"''', channel_page
) is not None
1567 # The videos are contained in a single page
1568 # the ajax pages can't be used, they are empty
1569 video_ids
= self
.extract_videos_from_page(channel_page
)
1571 # Download all channel pages using the json-based channel_ajax query
1572 for pagenum
in itertools
.count(1):
1573 url
= self
._MORE
_PAGES
_URL
% (pagenum
, channel_id
)
1574 page
= self
._download
_webpage
(url
, channel_id
,
1575 u
'Downloading page #%s' % pagenum
)
1577 page
= json
.loads(page
)
1579 ids_in_page
= self
.extract_videos_from_page(page
['content_html'])
1580 video_ids
.extend(ids_in_page
)
1582 if self
._MORE
_PAGES
_INDICATOR
not in page
['load_more_widget_html']:
1585 self
._downloader
.to_screen(u
'[youtube] Channel %s: Found %i videos' % (channel_id
, len(video_ids
)))
1587 url_entries
= [self
.url_result(video_id
, 'Youtube', video_id
=video_id
)
1588 for video_id
in video_ids
]
1589 return self
.playlist_result(url_entries
, channel_id
)
1592 class YoutubeUserIE(InfoExtractor
):
1593 IE_DESC
= u
'YouTube.com user videos (URL or "ytuser" keyword)'
1594 _VALID_URL
= r
'(?:(?:(?:https?://)?(?:\w+\.)?youtube\.com/(?:user/)?(?!(?:attribution_link|watch)(?:$|[^a-z_A-Z0-9-])))|ytuser:)(?!feed/)([A-Za-z0-9_-]+)'
1595 _TEMPLATE_URL
= 'http://gdata.youtube.com/feeds/api/users/%s'
1596 _GDATA_PAGE_SIZE
= 50
1597 _GDATA_URL
= 'http://gdata.youtube.com/feeds/api/users/%s/uploads?max-results=%d&start-index=%d&alt=json'
1598 IE_NAME
= u
'youtube:user'
1601 def suitable(cls
, url
):
1602 # Don't return True if the url can be extracted with other youtube
1603 # extractor, the regex would is too permissive and it would match.
1604 other_ies
= iter(klass
for (name
, klass
) in globals().items() if name
.endswith('IE') and klass
is not cls
)
1605 if any(ie
.suitable(url
) for ie
in other_ies
): return False
1606 else: return super(YoutubeUserIE
, cls
).suitable(url
)
1608 def _real_extract(self
, url
):
1610 mobj
= re
.match(self
._VALID
_URL
, url
)
1612 raise ExtractorError(u
'Invalid URL: %s' % url
)
1614 username
= mobj
.group(1)
1616 # Download video ids using YouTube Data API. Result size per
1617 # query is limited (currently to 50 videos) so we need to query
1618 # page by page until there are no video ids - it means we got
1623 for pagenum
in itertools
.count(0):
1624 start_index
= pagenum
* self
._GDATA
_PAGE
_SIZE
+ 1
1626 gdata_url
= self
._GDATA
_URL
% (username
, self
._GDATA
_PAGE
_SIZE
, start_index
)
1627 page
= self
._download
_webpage
(gdata_url
, username
,
1628 u
'Downloading video ids from %d to %d' % (start_index
, start_index
+ self
._GDATA
_PAGE
_SIZE
))
1631 response
= json
.loads(page
)
1632 except ValueError as err
:
1633 raise ExtractorError(u
'Invalid JSON in API response: ' + compat_str(err
))
1634 if 'entry' not in response
['feed']:
1635 # Number of videos is a multiple of self._MAX_RESULTS
1638 # Extract video identifiers
1639 entries
= response
['feed']['entry']
1640 for entry
in entries
:
1641 title
= entry
['title']['$t']
1642 video_id
= entry
['id']['$t'].split('/')[-1]
1643 url_results
.append({
1646 'ie_key': 'Youtube',
1651 # A little optimization - if current page is not
1652 # "full", ie. does not contain PAGE_SIZE video ids then
1653 # we can assume that this page is the last one - there
1654 # are no more ids on further pages - no need to query
1657 if len(entries
) < self
._GDATA
_PAGE
_SIZE
:
1660 return self
.playlist_result(url_results
, playlist_title
=username
)
1663 class YoutubeSearchIE(SearchInfoExtractor
):
1664 IE_DESC
= u
'YouTube.com searches'
1665 _API_URL
= 'https://gdata.youtube.com/feeds/api/videos?q=%s&start-index=%i&max-results=50&v=2&alt=jsonc'
1667 IE_NAME
= u
'youtube:search'
1668 _SEARCH_KEY
= 'ytsearch'
1670 def _get_n_results(self
, query
, n
):
1671 """Get a specified number of results for a query"""
1677 while (50 * pagenum
) < limit
:
1678 result_url
= self
._API
_URL
% (compat_urllib_parse
.quote_plus(query
), (50*pagenum
)+1)
1679 data_json
= self
._download
_webpage
(
1680 result_url
, video_id
=u
'query "%s"' % query
,
1681 note
=u
'Downloading page %s' % (pagenum
+ 1),
1682 errnote
=u
'Unable to download API page')
1683 data
= json
.loads(data_json
)
1684 api_response
= data
['data']
1686 if 'items' not in api_response
:
1687 raise ExtractorError(u
'[youtube] No video results')
1689 new_ids
= list(video
['id'] for video
in api_response
['items'])
1690 video_ids
+= new_ids
1692 limit
= min(n
, api_response
['totalItems'])
1695 if len(video_ids
) > n
:
1696 video_ids
= video_ids
[:n
]
1697 videos
= [self
.url_result(video_id
, 'Youtube', video_id
=video_id
)
1698 for video_id
in video_ids
]
1699 return self
.playlist_result(videos
, query
)
1701 class YoutubeSearchDateIE(YoutubeSearchIE
):
1702 IE_NAME
= YoutubeSearchIE
.IE_NAME
+ ':date'
1703 _API_URL
= 'https://gdata.youtube.com/feeds/api/videos?q=%s&start-index=%i&max-results=50&v=2&alt=jsonc&orderby=published'
1704 _SEARCH_KEY
= 'ytsearchdate'
1705 IE_DESC
= u
'YouTube.com searches, newest videos first'
1707 class YoutubeShowIE(InfoExtractor
):
1708 IE_DESC
= u
'YouTube.com (multi-season) shows'
1709 _VALID_URL
= r
'https?://www\.youtube\.com/show/(.*)'
1710 IE_NAME
= u
'youtube:show'
1712 def _real_extract(self
, url
):
1713 mobj
= re
.match(self
._VALID
_URL
, url
)
1714 show_name
= mobj
.group(1)
1715 webpage
= self
._download
_webpage
(url
, show_name
, u
'Downloading show webpage')
1716 # There's one playlist for each season of the show
1717 m_seasons
= list(re
.finditer(r
'href="(/playlist\?list=.*?)"', webpage
))
1718 self
.to_screen(u
'%s: Found %s seasons' % (show_name
, len(m_seasons
)))
1719 return [self
.url_result('https://www.youtube.com' + season
.group(1), 'YoutubePlaylist') for season
in m_seasons
]
1722 class YoutubeFeedsInfoExtractor(YoutubeBaseInfoExtractor
):
1724 Base class for extractors that fetch info from
1725 http://www.youtube.com/feed_ajax
1726 Subclasses must define the _FEED_NAME and _PLAYLIST_TITLE properties.
1728 _LOGIN_REQUIRED
= True
1729 # use action_load_personal_feed instead of action_load_system_feed
1730 _PERSONAL_FEED
= False
1733 def _FEED_TEMPLATE(self
):
1734 action
= 'action_load_system_feed'
1735 if self
._PERSONAL
_FEED
:
1736 action
= 'action_load_personal_feed'
1737 return 'http://www.youtube.com/feed_ajax?%s=1&feed_name=%s&paging=%%s' % (action
, self
._FEED
_NAME
)
1741 return u
'youtube:%s' % self
._FEED
_NAME
1743 def _real_initialize(self
):
1746 def _real_extract(self
, url
):
1749 for i
in itertools
.count(1):
1750 info
= self
._download
_webpage
(self
._FEED
_TEMPLATE
% paging
,
1751 u
'%s feed' % self
._FEED
_NAME
,
1752 u
'Downloading page %s' % i
)
1753 info
= json
.loads(info
)
1754 feed_html
= info
['feed_html']
1755 m_ids
= re
.finditer(r
'"/watch\?v=(.*?)["&]', feed_html
)
1756 ids
= orderedSet(m
.group(1) for m
in m_ids
)
1757 feed_entries
.extend(
1758 self
.url_result(video_id
, 'Youtube', video_id
=video_id
)
1759 for video_id
in ids
)
1760 if info
['paging'] is None:
1762 paging
= info
['paging']
1763 return self
.playlist_result(feed_entries
, playlist_title
=self
._PLAYLIST
_TITLE
)
1765 class YoutubeSubscriptionsIE(YoutubeFeedsInfoExtractor
):
1766 IE_DESC
= u
'YouTube.com subscriptions feed, "ytsubs" keyword(requires authentication)'
1767 _VALID_URL
= r
'https?://www\.youtube\.com/feed/subscriptions|:ytsubs(?:criptions)?'
1768 _FEED_NAME
= 'subscriptions'
1769 _PLAYLIST_TITLE
= u
'Youtube Subscriptions'
1771 class YoutubeRecommendedIE(YoutubeFeedsInfoExtractor
):
1772 IE_DESC
= u
'YouTube.com recommended videos, "ytrec" keyword (requires authentication)'
1773 _VALID_URL
= r
'https?://www\.youtube\.com/feed/recommended|:ytrec(?:ommended)?'
1774 _FEED_NAME
= 'recommended'
1775 _PLAYLIST_TITLE
= u
'Youtube Recommended videos'
1777 class YoutubeWatchLaterIE(YoutubeFeedsInfoExtractor
):
1778 IE_DESC
= u
'Youtube watch later list, "ytwatchlater" keyword (requires authentication)'
1779 _VALID_URL
= r
'https?://www\.youtube\.com/feed/watch_later|:ytwatchlater'
1780 _FEED_NAME
= 'watch_later'
1781 _PLAYLIST_TITLE
= u
'Youtube Watch Later'
1782 _PERSONAL_FEED
= True
1784 class YoutubeHistoryIE(YoutubeFeedsInfoExtractor
):
1785 IE_DESC
= u
'Youtube watch history, "ythistory" keyword (requires authentication)'
1786 _VALID_URL
= u
'https?://www\.youtube\.com/feed/history|:ythistory'
1787 _FEED_NAME
= 'history'
1788 _PERSONAL_FEED
= True
1789 _PLAYLIST_TITLE
= u
'Youtube Watch History'
1791 class YoutubeFavouritesIE(YoutubeBaseInfoExtractor
):
1792 IE_NAME
= u
'youtube:favorites'
1793 IE_DESC
= u
'YouTube.com favourite videos, "ytfav" keyword (requires authentication)'
1794 _VALID_URL
= r
'https?://www\.youtube\.com/my_favorites|:ytfav(?:ou?rites)?'
1795 _LOGIN_REQUIRED
= True
1797 def _real_extract(self
, url
):
1798 webpage
= self
._download
_webpage
('https://www.youtube.com/my_favorites', 'Youtube Favourites videos')
1799 playlist_id
= self
._search
_regex
(r
'list=(.+?)["&]', webpage
, u
'favourites playlist id')
1800 return self
.url_result(playlist_id
, 'YoutubePlaylist')
1803 class YoutubeTruncatedURLIE(InfoExtractor
):
1804 IE_NAME
= 'youtube:truncated_url'
1805 IE_DESC
= False # Do not list
1806 _VALID_URL
= r
'(?:https?://)?[^/]+/watch\?feature=[a-z_]+$'
1808 def _real_extract(self
, url
):
1809 raise ExtractorError(
1810 u
'Did you forget to quote the URL? Remember that & is a meta '
1811 u
'character in most shells, so you want to put the URL in quotes, '
1813 u
'"http://www.youtube.com/watch?feature=foo&v=BaW_jenozKc" '
1814 u
' or simply youtube-dl BaW_jenozKc .',