14 from .common
import InfoExtractor
, SearchInfoExtractor
15 from .subtitles
import SubtitlesInfoExtractor
16 from ..jsinterp
import JSInterpreter
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
= 'https://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
,
115 compat_urllib_parse
.urlencode(age_form
).encode('ascii'))
117 self
._download
_webpage
(
119 note
=u
'Confirming age', errnote
=u
'Unable to confirm age')
122 def _real_initialize(self
):
123 if self
._downloader
is None:
125 if not self
._set
_language
():
127 if not self
._login
():
132 class YoutubeIE(YoutubeBaseInfoExtractor
, SubtitlesInfoExtractor
):
133 IE_DESC
= u
'YouTube.com'
134 _VALID_URL
= r
"""(?x)^
136 (?:https?://|//)? # http(s):// or protocol-independent URL (optional)
137 (?:(?:(?:(?:\w+\.)?[yY][oO][uU][tT][uU][bB][eE](?:-nocookie)?\.com/|
138 (?:www\.)?deturl\.com/www\.youtube\.com/|
139 (?:www\.)?pwnyoutube\.com/|
140 (?:www\.)?yourepeat\.com/|
141 tube\.majestyc\.net/|
142 youtube\.googleapis\.com/) # the various hostnames, with wildcard subdomains
143 (?:.*?\#/)? # handle anchor (#/) redirect urls
144 (?: # the various things that can precede the ID:
145 (?:(?:v|embed|e)/) # v/ or embed/ or e/
146 |(?: # or the v= param in all its forms
147 (?:(?:watch|movie)(?:_popup)?(?:\.php)?/?)? # preceding watch(_popup|.php) or nothing (like /?v=xxxx)
148 (?:\?|\#!?) # the params delimiter ? or # or #!
149 (?:.*?&)? # any other preceding param (like /?s=tuff&v=xxxx)
153 |youtu\.be/ # just youtu.be/xxxx
154 |https?://(?:www\.)?cleanvideosearch\.com/media/action/yt/watch\?videoId=
156 )? # all until now is optional -> you can pass the naked ID
157 ([0-9A-Za-z_-]{11}) # here is it! the YouTube video ID
158 (?(1).+)? # if we found the ID, everything can follow
160 _NEXT_URL_RE
= r
'[\?&]next_url=([^&]+)'
162 '5': {'ext': 'flv', 'width': 400, 'height': 240}
,
163 '6': {'ext': 'flv', 'width': 450, 'height': 270}
,
164 '13': {'ext': '3gp'}
,
165 '17': {'ext': '3gp', 'width': 176, 'height': 144}
,
166 '18': {'ext': 'mp4', 'width': 640, 'height': 360}
,
167 '22': {'ext': 'mp4', 'width': 1280, 'height': 720}
,
168 '34': {'ext': 'flv', 'width': 640, 'height': 360}
,
169 '35': {'ext': 'flv', 'width': 854, 'height': 480}
,
170 '36': {'ext': '3gp', 'width': 320, 'height': 240}
,
171 '37': {'ext': 'mp4', 'width': 1920, 'height': 1080}
,
172 '38': {'ext': 'mp4', 'width': 4096, 'height': 3072}
,
173 '43': {'ext': 'webm', 'width': 640, 'height': 360}
,
174 '44': {'ext': 'webm', 'width': 854, 'height': 480}
,
175 '45': {'ext': 'webm', 'width': 1280, 'height': 720}
,
176 '46': {'ext': 'webm', 'width': 1920, 'height': 1080}
,
180 '82': {'ext': 'mp4', 'height': 360, 'format_note': '3D', 'preference': -20}
,
181 '83': {'ext': 'mp4', 'height': 480, 'format_note': '3D', 'preference': -20}
,
182 '84': {'ext': 'mp4', 'height': 720, 'format_note': '3D', 'preference': -20}
,
183 '85': {'ext': 'mp4', 'height': 1080, 'format_note': '3D', 'preference': -20}
,
184 '100': {'ext': 'webm', 'height': 360, 'format_note': '3D', 'preference': -20}
,
185 '101': {'ext': 'webm', 'height': 480, 'format_note': '3D', 'preference': -20}
,
186 '102': {'ext': 'webm', 'height': 720, 'format_note': '3D', 'preference': -20}
,
188 # Apple HTTP Live Streaming
189 '92': {'ext': 'mp4', 'height': 240, 'format_note': 'HLS', 'preference': -10}
,
190 '93': {'ext': 'mp4', 'height': 360, 'format_note': 'HLS', 'preference': -10}
,
191 '94': {'ext': 'mp4', 'height': 480, 'format_note': 'HLS', 'preference': -10}
,
192 '95': {'ext': 'mp4', 'height': 720, 'format_note': 'HLS', 'preference': -10}
,
193 '96': {'ext': 'mp4', 'height': 1080, 'format_note': 'HLS', 'preference': -10}
,
194 '132': {'ext': 'mp4', 'height': 240, 'format_note': 'HLS', 'preference': -10}
,
195 '151': {'ext': 'mp4', 'height': 72, 'format_note': 'HLS', 'preference': -10}
,
198 '133': {'ext': 'mp4', 'height': 240, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40}
,
199 '134': {'ext': 'mp4', 'height': 360, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40}
,
200 '135': {'ext': 'mp4', 'height': 480, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40}
,
201 '136': {'ext': 'mp4', 'height': 720, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40}
,
202 '137': {'ext': 'mp4', 'height': 1080, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40}
,
203 '138': {'ext': 'mp4', 'height': 2160, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40}
,
204 '160': {'ext': 'mp4', 'height': 144, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40}
,
205 '264': {'ext': 'mp4', 'height': 1440, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40}
,
208 '139': {'ext': 'm4a', 'format_note': 'DASH audio', 'vcodec': 'none', 'abr': 48, 'preference': -50}
,
209 '140': {'ext': 'm4a', 'format_note': 'DASH audio', 'vcodec': 'none', 'abr': 128, 'preference': -50}
,
210 '141': {'ext': 'm4a', 'format_note': 'DASH audio', 'vcodec': 'none', 'abr': 256, 'preference': -50}
,
213 '167': {'ext': 'webm', 'height': 360, 'width': 640, 'format_note': 'DASH video', 'acodec': 'none', 'container': 'webm', 'vcodec': 'VP8', 'preference': -40}
,
214 '168': {'ext': 'webm', 'height': 480, 'width': 854, 'format_note': 'DASH video', 'acodec': 'none', 'container': 'webm', 'vcodec': 'VP8', 'preference': -40}
,
215 '169': {'ext': 'webm', 'height': 720, 'width': 1280, 'format_note': 'DASH video', 'acodec': 'none', 'container': 'webm', 'vcodec': 'VP8', 'preference': -40}
,
216 '170': {'ext': 'webm', 'height': 1080, 'width': 1920, 'format_note': 'DASH video', 'acodec': 'none', 'container': 'webm', 'vcodec': 'VP8', 'preference': -40}
,
217 '218': {'ext': 'webm', 'height': 480, 'width': 854, 'format_note': 'DASH video', 'acodec': 'none', 'container': 'webm', 'vcodec': 'VP8', 'preference': -40}
,
218 '219': {'ext': 'webm', 'height': 480, 'width': 854, 'format_note': 'DASH video', 'acodec': 'none', 'container': 'webm', 'vcodec': 'VP8', 'preference': -40}
,
219 '242': {'ext': 'webm', 'height': 240, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40}
,
220 '243': {'ext': 'webm', 'height': 360, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40}
,
221 '244': {'ext': 'webm', 'height': 480, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40}
,
222 '245': {'ext': 'webm', 'height': 480, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40}
,
223 '246': {'ext': 'webm', 'height': 480, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40}
,
224 '247': {'ext': 'webm', 'height': 720, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40}
,
225 '248': {'ext': 'webm', 'height': 1080, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40}
,
226 '271': {'ext': 'webm', 'height': 1440, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40}
,
227 '272': {'ext': 'webm', 'height': 2160, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40}
,
230 '171': {'ext': 'webm', 'vcodec': 'none', 'format_note': 'DASH audio', 'abr': 48, 'preference': -50}
,
231 '172': {'ext': 'webm', 'vcodec': 'none', 'format_note': 'DASH audio', 'abr': 256, 'preference': -50}
,
234 '_rtmp': {'protocol': 'rtmp'}
,
240 u
"url": u
"http://www.youtube.com/watch?v=BaW_jenozKc",
241 u
"file": u
"BaW_jenozKc.mp4",
243 u
"title": u
"youtube-dl test video \"'/\\ä↭𝕐",
244 u
"uploader": u
"Philipp Hagemeister",
245 u
"uploader_id": u
"phihag",
246 u
"upload_date": u
"20121002",
247 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 .",
248 u
"categories": [u
'Science & Technology'],
252 u
"url": u
"http://www.youtube.com/watch?v=UxxajLWwzqY",
253 u
"file": u
"UxxajLWwzqY.mp4",
254 u
"note": u
"Test generic use_cipher_signature video (#897)",
256 u
"upload_date": u
"20120506",
257 u
"title": u
"Icona Pop - I Love It (feat. Charli XCX) [OFFICIAL VIDEO]",
258 u
"description": u
"md5:fea86fda2d5a5784273df5c7cc994d9f",
259 u
"uploader": u
"Icona Pop",
260 u
"uploader_id": u
"IconaPop"
264 u
"url": u
"https://www.youtube.com/watch?v=07FYdnEawAQ",
265 u
"file": u
"07FYdnEawAQ.mp4",
266 u
"note": u
"Test VEVO video with age protection (#956)",
268 u
"upload_date": u
"20130703",
269 u
"title": u
"Justin Timberlake - Tunnel Vision (Explicit)",
270 u
"description": u
"md5:64249768eec3bc4276236606ea996373",
271 u
"uploader": u
"justintimberlakeVEVO",
272 u
"uploader_id": u
"justintimberlakeVEVO"
276 u
"url": u
"//www.YouTube.com/watch?v=yZIXLfi8CZQ",
277 u
"file": u
"yZIXLfi8CZQ.mp4",
278 u
"note": u
"Embed-only video (#1746)",
280 u
"upload_date": u
"20120608",
281 u
"title": u
"Principal Sexually Assaults A Teacher - Episode 117 - 8th June 2012",
282 u
"description": u
"md5:09b78bd971f1e3e289601dfba15ca4f7",
283 u
"uploader": u
"SET India",
284 u
"uploader_id": u
"setindia"
288 u
"url": u
"http://www.youtube.com/watch?v=a9LDPn-MO4I",
289 u
"file": u
"a9LDPn-MO4I.m4a",
290 u
"note": u
"256k DASH audio (format 141) via DASH manifest",
292 u
"upload_date": "20121002",
293 u
"uploader_id": "8KVIDEO",
294 u
"description": "No description available.",
295 u
"uploader": "8KVIDEO",
296 u
"title": "UHDTV TEST 8K VIDEO.mp4"
299 u
"youtube_include_dash_manifest": True,
303 # DASH manifest with encrypted signature
305 u
'url': u
'https://www.youtube.com/watch?v=IB3lcPjvWLA',
307 u
'id': u
'IB3lcPjvWLA',
309 u
'title': u
'Afrojack - The Spark ft. Spree Wilson',
310 u
'description': u
'md5:9717375db5a9a3992be4668bbf3bc0a8',
311 u
'uploader': u
'AfrojackVEVO',
312 u
'uploader_id': u
'AfrojackVEVO',
313 u
'upload_date': u
'20131011',
316 u
'youtube_include_dash_manifest': True,
324 def suitable(cls
, url
):
325 """Receives a URL and returns True if suitable for this IE."""
326 if YoutubePlaylistIE
.suitable(url
): return False
327 return re
.match(cls
._VALID
_URL
, url
) is not None
329 def __init__(self
, *args
, **kwargs
):
330 super(YoutubeIE
, self
).__init
__(*args
, **kwargs
)
331 self
._player
_cache
= {}
333 def report_video_info_webpage_download(self
, video_id
):
334 """Report attempt to download video info webpage."""
335 self
.to_screen(u
'%s: Downloading video info webpage' % video_id
)
337 def report_information_extraction(self
, video_id
):
338 """Report attempt to extract video information."""
339 self
.to_screen(u
'%s: Extracting video information' % video_id
)
341 def report_unavailable_format(self
, video_id
, format
):
342 """Report extracted video URL."""
343 self
.to_screen(u
'%s: Format %s not available' % (video_id
, format
))
345 def report_rtmp_download(self
):
346 """Indicate the download will use the RTMP protocol."""
347 self
.to_screen(u
'RTMP download detected')
349 def _extract_signature_function(self
, video_id
, player_url
, slen
):
350 id_m
= re
.match(r
'.*-(?P<id>[a-zA-Z0-9_-]+)\.(?P<ext>[a-z]+)$',
352 player_type
= id_m
.group('ext')
353 player_id
= id_m
.group('id')
355 # Read from filesystem cache
356 func_id
= '%s_%s_%d' % (player_type
, player_id
, slen
)
357 assert os
.path
.basename(func_id
) == func_id
358 cache_dir
= get_cachedir(self
._downloader
.params
)
360 cache_enabled
= cache_dir
is not None
362 cache_fn
= os
.path
.join(os
.path
.expanduser(cache_dir
),
366 with io
.open(cache_fn
, 'r', encoding
='utf-8') as cachef
:
367 cache_spec
= json
.load(cachef
)
368 return lambda s
: u
''.join(s
[i
] for i
in cache_spec
)
370 pass # No cache available
372 if player_type
== 'js':
373 code
= self
._download
_webpage
(
374 player_url
, video_id
,
375 note
=u
'Downloading %s player %s' % (player_type
, player_id
),
376 errnote
=u
'Download of %s failed' % player_url
)
377 res
= self
._parse
_sig
_js
(code
)
378 elif player_type
== 'swf':
379 urlh
= self
._request
_webpage
(
380 player_url
, video_id
,
381 note
=u
'Downloading %s player %s' % (player_type
, player_id
),
382 errnote
=u
'Download of %s failed' % player_url
)
384 res
= self
._parse
_sig
_swf
(code
)
386 assert False, 'Invalid player type %r' % player_type
390 test_string
= u
''.join(map(compat_chr
, range(slen
)))
391 cache_res
= res(test_string
)
392 cache_spec
= [ord(c
) for c
in cache_res
]
394 os
.makedirs(os
.path
.dirname(cache_fn
))
395 except OSError as ose
:
396 if ose
.errno
!= errno
.EEXIST
:
398 write_json_file(cache_spec
, cache_fn
)
400 tb
= traceback
.format_exc()
401 self
._downloader
.report_warning(
402 u
'Writing cache to %r failed: %s' % (cache_fn
, tb
))
406 def _print_sig_code(self
, func
, slen
):
407 def gen_sig_code(idxs
):
408 def _genslice(start
, end
, step
):
409 starts
= u
'' if start
== 0 else str(start
)
410 ends
= (u
':%d' % (end
+step
)) if end
+ step
>= 0 else u
':'
411 steps
= u
'' if step
== 1 else (u
':%d' % step
)
412 return u
's[%s%s%s]' % (starts
, ends
, steps
)
415 start
= '(Never used)' # Quelch pyflakes warnings - start will be
416 # set as soon as step is set
417 for i
, prev
in zip(idxs
[1:], idxs
[:-1]):
421 yield _genslice(start
, prev
, step
)
424 if i
- prev
in [-1, 1]:
429 yield u
's[%d]' % prev
433 yield _genslice(start
, i
, step
)
435 test_string
= u
''.join(map(compat_chr
, range(slen
)))
436 cache_res
= func(test_string
)
437 cache_spec
= [ord(c
) for c
in cache_res
]
438 expr_code
= u
' + '.join(gen_sig_code(cache_spec
))
439 code
= u
'if len(s) == %d:\n return %s\n' % (slen
, expr_code
)
440 self
.to_screen(u
'Extracted signature function:\n' + code
)
442 def _parse_sig_js(self
, jscode
):
443 funcname
= self
._search
_regex
(
444 r
'signature=([$a-zA-Z]+)', jscode
,
445 u
'Initial JS player signature function name')
447 jsi
= JSInterpreter(jscode
)
448 initial_function
= jsi
.extract_function(funcname
)
449 return lambda s
: initial_function([s
])
451 def _parse_sig_swf(self
, file_contents
):
452 if file_contents
[1:3] != b
'WS':
453 raise ExtractorError(
454 u
'Not an SWF file; header is %r' % file_contents
[:3])
455 if file_contents
[:1] == b
'C':
456 content
= zlib
.decompress(file_contents
[8:])
458 raise NotImplementedError(u
'Unsupported compression format %r' %
461 def extract_tags(content
):
463 while pos
< len(content
):
464 header16
= struct
.unpack('<H', content
[pos
:pos
+2])[0]
466 tag_code
= header16
>> 6
467 tag_len
= header16
& 0x3f
469 tag_len
= struct
.unpack('<I', content
[pos
:pos
+4])[0]
471 assert pos
+tag_len
<= len(content
)
472 yield (tag_code
, content
[pos
:pos
+tag_len
])
476 for tag_code
, tag
in extract_tags(content
)
478 p
= code_tag
.index(b
'\0', 4) + 1
479 code_reader
= io
.BytesIO(code_tag
[p
:])
481 # Parse ABC (AVM2 ByteCode)
482 def read_int(reader
=None):
490 b
= struct
.unpack('<B', buf
)[0]
491 res
= res |
((b
& 0x7f) << shift
)
497 def u30(reader
=None):
498 res
= read_int(reader
)
499 assert res
& 0xf0000000 == 0
503 def s32(reader
=None):
505 if v
& 0x80000000 != 0:
506 v
= - ((v ^
0xffffffff) + 1)
509 def read_string(reader
=None):
513 resb
= reader
.read(slen
)
514 assert len(resb
) == slen
515 return resb
.decode('utf-8')
517 def read_bytes(count
, reader
=None):
520 resb
= reader
.read(count
)
521 assert len(resb
) == count
524 def read_byte(reader
=None):
525 resb
= read_bytes(1, reader
=reader
)
526 res
= struct
.unpack('<B', resb
)[0]
529 # minor_version + major_version
534 for _c
in range(1, int_count
):
537 for _c
in range(1, uint_count
):
540 read_bytes((double_count
-1) * 8)
542 constant_strings
= [u
'']
543 for _c
in range(1, string_count
):
545 constant_strings
.append(s
)
546 namespace_count
= u30()
547 for _c
in range(1, namespace_count
):
551 for _c
in range(1, ns_set_count
):
553 for _c2
in range(count
):
555 multiname_count
= u30()
564 0x0e: 2, # MultinameA
565 0x1b: 1, # MultinameL
566 0x1c: 1, # MultinameLA
569 for _c
in range(1, multiname_count
):
571 assert kind
in MULTINAME_SIZES
, u
'Invalid multiname kind %r' % kind
573 u30() # namespace_idx
575 multinames
.append(constant_strings
[name_idx
])
577 multinames
.append('[MULTINAME kind: %d]' % kind
)
578 for _c2
in range(MULTINAME_SIZES
[kind
]):
583 MethodInfo
= collections
.namedtuple(
585 ['NEED_ARGUMENTS', 'NEED_REST'])
587 for method_id
in range(method_count
):
590 for _
in range(param_count
):
592 u30() # name index (always 0 for youtube)
594 if flags
& 0x08 != 0:
597 for c
in range(option_count
):
600 if flags
& 0x80 != 0:
601 # Param names present
602 for _
in range(param_count
):
604 mi
= MethodInfo(flags
& 0x01 != 0, flags
& 0x04 != 0)
605 method_infos
.append(mi
)
608 metadata_count
= u30()
609 for _c
in range(metadata_count
):
612 for _c2
in range(item_count
):
616 def parse_traits_info():
617 trait_name_idx
= u30()
618 kind_full
= read_byte()
619 kind
= kind_full
& 0x0f
620 attrs
= kind_full
>> 4
622 if kind
in [0x00, 0x06]: # Slot or Const
624 u30() # type_name_idx
628 elif kind
in [0x01, 0x02, 0x03]: # Method / Getter / Setter
631 methods
[multinames
[trait_name_idx
]] = method_idx
632 elif kind
== 0x04: # Class
635 elif kind
== 0x05: # Function
638 methods
[function_idx
] = multinames
[trait_name_idx
]
640 raise ExtractorError(u
'Unsupported trait kind %d' % kind
)
642 if attrs
& 0x4 != 0: # Metadata present
643 metadata_count
= u30()
644 for _c3
in range(metadata_count
):
645 u30() # metadata index
650 TARGET_CLASSNAME
= u
'SignatureDecipher'
651 searched_idx
= multinames
.index(TARGET_CLASSNAME
)
652 searched_class_id
= None
654 for class_id
in range(class_count
):
656 if name_idx
== searched_idx
:
657 # We found the class we're looking for!
658 searched_class_id
= class_id
659 u30() # super_name idx
661 if flags
& 0x08 != 0: # Protected namespace is present
662 u30() # protected_ns_idx
664 for _c2
in range(intrf_count
):
668 for _c2
in range(trait_count
):
671 if searched_class_id
is None:
672 raise ExtractorError(u
'Target class %r not found' %
677 for class_id
in range(class_count
):
680 for _c2
in range(trait_count
):
681 trait_methods
= parse_traits_info()
682 if class_id
== searched_class_id
:
683 method_names
.update(trait_methods
.items())
684 method_idxs
.update(dict(
686 for name
, idx
in trait_methods
.items()))
690 for _c
in range(script_count
):
693 for _c2
in range(trait_count
):
697 method_body_count
= u30()
698 Method
= collections
.namedtuple('Method', ['code', 'local_count'])
700 for _c
in range(method_body_count
):
704 u30() # init_scope_depth
705 u30() # max_scope_depth
707 code
= read_bytes(code_length
)
708 if method_idx
in method_idxs
:
709 m
= Method(code
, local_count
)
710 methods
[method_idxs
[method_idx
]] = m
711 exception_count
= u30()
712 for _c2
in range(exception_count
):
719 for _c2
in range(trait_count
):
722 assert p
+ code_reader
.tell() == len(code_tag
)
723 assert len(methods
) == len(method_idxs
)
725 method_pyfunctions
= {}
727 def extract_function(func_name
):
728 if func_name
in method_pyfunctions
:
729 return method_pyfunctions
[func_name
]
730 if func_name
not in methods
:
731 raise ExtractorError(u
'Cannot find function %r' % func_name
)
732 m
= methods
[func_name
]
735 registers
= ['(this)'] + list(args
) + [None] * m
.local_count
737 coder
= io
.BytesIO(m
.code
)
739 opcode
= struct
.unpack('!B', coder
.read(1))[0]
740 if opcode
== 36: # pushbyte
741 v
= struct
.unpack('!B', coder
.read(1))[0]
743 elif opcode
== 44: # pushstring
745 stack
.append(constant_strings
[idx
])
746 elif opcode
== 48: # pushscope
747 # We don't implement the scope register, so we'll just
748 # ignore the popped value
750 elif opcode
== 70: # callproperty
752 mname
= multinames
[index
]
753 arg_count
= u30(coder
)
754 args
= list(reversed(
755 [stack
.pop() for _
in range(arg_count
)]))
757 if mname
== u
'split':
758 assert len(args
) == 1
759 assert isinstance(args
[0], compat_str
)
760 assert isinstance(obj
, compat_str
)
764 res
= obj
.split(args
[0])
766 elif mname
== u
'slice':
767 assert len(args
) == 1
768 assert isinstance(args
[0], int)
769 assert isinstance(obj
, list)
772 elif mname
== u
'join':
773 assert len(args
) == 1
774 assert isinstance(args
[0], compat_str
)
775 assert isinstance(obj
, list)
776 res
= args
[0].join(obj
)
778 elif mname
in method_pyfunctions
:
779 stack
.append(method_pyfunctions
[mname
](args
))
781 raise NotImplementedError(
782 u
'Unsupported property %r on %r'
784 elif opcode
== 72: # returnvalue
787 elif opcode
== 79: # callpropvoid
789 mname
= multinames
[index
]
790 arg_count
= u30(coder
)
791 args
= list(reversed(
792 [stack
.pop() for _
in range(arg_count
)]))
794 if mname
== u
'reverse':
795 assert isinstance(obj
, list)
798 raise NotImplementedError(
799 u
'Unsupported (void) property %r on %r'
801 elif opcode
== 93: # findpropstrict
803 mname
= multinames
[index
]
804 res
= extract_function(mname
)
806 elif opcode
== 97: # setproperty
811 assert isinstance(obj
, list)
812 assert isinstance(idx
, int)
814 elif opcode
== 98: # getlocal
816 stack
.append(registers
[index
])
817 elif opcode
== 99: # setlocal
820 registers
[index
] = value
821 elif opcode
== 102: # getproperty
823 pname
= multinames
[index
]
824 if pname
== u
'length':
826 assert isinstance(obj
, list)
827 stack
.append(len(obj
))
828 else: # Assume attribute access
830 assert isinstance(idx
, int)
832 assert isinstance(obj
, list)
833 stack
.append(obj
[idx
])
834 elif opcode
== 128: # coerce
836 elif opcode
== 133: # coerce_s
837 assert isinstance(stack
[-1], (type(None), compat_str
))
838 elif opcode
== 164: # modulo
841 res
= value1
% value2
843 elif opcode
== 208: # getlocal_0
844 stack
.append(registers
[0])
845 elif opcode
== 209: # getlocal_1
846 stack
.append(registers
[1])
847 elif opcode
== 210: # getlocal_2
848 stack
.append(registers
[2])
849 elif opcode
== 211: # getlocal_3
850 stack
.append(registers
[3])
851 elif opcode
== 214: # setlocal_2
852 registers
[2] = stack
.pop()
853 elif opcode
== 215: # setlocal_3
854 registers
[3] = stack
.pop()
856 raise NotImplementedError(
857 u
'Unsupported opcode %d' % opcode
)
859 method_pyfunctions
[func_name
] = resfunc
862 initial_function
= extract_function(u
'decipher')
863 return lambda s
: initial_function([s
])
865 def _decrypt_signature(self
, s
, video_id
, player_url
, age_gate
=False):
866 """Turn the encrypted s field into a working signature"""
868 if player_url
is not None:
869 if player_url
.startswith(u
'//'):
870 player_url
= u
'https:' + player_url
872 player_id
= (player_url
, len(s
))
873 if player_id
not in self
._player
_cache
:
874 func
= self
._extract
_signature
_function
(
875 video_id
, player_url
, len(s
)
877 self
._player
_cache
[player_id
] = func
878 func
= self
._player
_cache
[player_id
]
879 if self
._downloader
.params
.get('youtube_print_sig_code'):
880 self
._print
_sig
_code
(func
, len(s
))
883 tb
= traceback
.format_exc()
884 self
._downloader
.report_warning(
885 u
'Automatic signature extraction failed: ' + tb
)
887 self
._downloader
.report_warning(
888 u
'Warning: Falling back to static signature algorithm')
890 return self
._static
_decrypt
_signature
(
891 s
, video_id
, player_url
, age_gate
)
893 def _static_decrypt_signature(self
, s
, video_id
, player_url
, age_gate
):
895 # The videos with age protection use another player, so the
896 # algorithms can be different.
898 return s
[2:63] + s
[82] + s
[64:82] + s
[63]
901 return s
[86:29:-1] + s
[88] + s
[28:5:-1]
903 return s
[25] + s
[3:25] + s
[0] + s
[26:42] + s
[79] + s
[43:79] + s
[91] + s
[80:83]
905 return s
[84:27:-1] + s
[86] + s
[26:5:-1]
907 return s
[25] + s
[3:25] + s
[2] + s
[26:40] + s
[77] + s
[41:77] + s
[89] + s
[78:81]
909 return s
[84:78:-1] + s
[87] + s
[77:60:-1] + s
[0] + s
[59:3:-1]
911 return s
[7:28] + s
[87] + s
[29:45] + s
[55] + s
[46:55] + s
[2] + s
[56:87] + s
[28]
913 return s
[6:27] + s
[4] + s
[28:39] + s
[27] + s
[40:59] + s
[2] + s
[60:]
915 return s
[80:72:-1] + s
[16] + s
[71:39:-1] + s
[72] + s
[38:16:-1] + s
[82] + s
[15::-1]
917 return s
[3:11] + s
[0] + s
[12:55] + s
[84] + s
[56:84]
919 return s
[78:70:-1] + s
[14] + s
[69:37:-1] + s
[70] + s
[36:14:-1] + s
[80] + s
[:14][::-1]
921 return s
[80:63:-1] + s
[0] + s
[62:0:-1] + s
[63]
923 return s
[80:37:-1] + s
[7] + s
[36:7:-1] + s
[0] + s
[6:0:-1] + s
[37]
925 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]
927 return s
[1:19] + s
[0] + s
[20:68] + s
[19] + s
[69:80]
929 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]
932 raise ExtractorError(u
'Unable to decrypt signature, key length %d not supported; retrying might work' % (len(s
)))
934 def _get_available_subtitles(self
, video_id
, webpage
):
936 sub_list
= self
._download
_webpage
(
937 'https://video.google.com/timedtext?hl=en&type=list&v=%s' % video_id
,
938 video_id
, note
=False)
939 except ExtractorError
as err
:
940 self
._downloader
.report_warning(u
'unable to download video subtitles: %s' % compat_str(err
))
942 lang_list
= re
.findall(r
'name="([^"]*)"[^>]+lang_code="([\w\-]+)"', sub_list
)
947 params
= compat_urllib_parse
.urlencode({
950 'fmt': self
._downloader
.params
.get('subtitlesformat', 'srt'),
951 'name': unescapeHTML(l
[0]).encode('utf-8'),
953 url
= u
'https://www.youtube.com/api/timedtext?' + params
954 sub_lang_list
[lang
] = url
955 if not sub_lang_list
:
956 self
._downloader
.report_warning(u
'video doesn\'t have subtitles')
960 def _get_available_automatic_caption(self
, video_id
, webpage
):
961 """We need the webpage for getting the captions url, pass it as an
962 argument to speed up the process."""
963 sub_format
= self
._downloader
.params
.get('subtitlesformat', 'srt')
964 self
.to_screen(u
'%s: Looking for automatic captions' % video_id
)
965 mobj
= re
.search(r
';ytplayer.config = ({.*?});', webpage
)
966 err_msg
= u
'Couldn\'t find automatic captions for %s' % video_id
968 self
._downloader
.report_warning(err_msg
)
970 player_config
= json
.loads(mobj
.group(1))
972 args
= player_config
[u
'args']
973 caption_url
= args
[u
'ttsurl']
974 timestamp
= args
[u
'timestamp']
975 # We get the available subtitles
976 list_params
= compat_urllib_parse
.urlencode({
981 list_url
= caption_url
+ '&' + list_params
982 caption_list
= self
._download
_xml
(list_url
, video_id
)
983 original_lang_node
= caption_list
.find('track')
984 if original_lang_node
is None or original_lang_node
.attrib
.get('kind') != 'asr' :
985 self
._downloader
.report_warning(u
'Video doesn\'t have automatic captions')
987 original_lang
= original_lang_node
.attrib
['lang_code']
990 for lang_node
in caption_list
.findall('target'):
991 sub_lang
= lang_node
.attrib
['lang_code']
992 params
= compat_urllib_parse
.urlencode({
993 'lang': original_lang
,
999 sub_lang_list
[sub_lang
] = caption_url
+ '&' + params
1000 return sub_lang_list
1001 # An extractor error can be raise by the download process if there are
1002 # no automatic captions but there are subtitles
1003 except (KeyError, ExtractorError
):
1004 self
._downloader
.report_warning(err_msg
)
1008 def extract_id(cls
, url
):
1009 mobj
= re
.match(cls
._VALID
_URL
, url
, re
.VERBOSE
)
1011 raise ExtractorError(u
'Invalid URL: %s' % url
)
1012 video_id
= mobj
.group(2)
1015 def _extract_from_m3u8(self
, manifest_url
, video_id
):
1017 def _get_urls(_manifest
):
1018 lines
= _manifest
.split('\n')
1019 urls
= filter(lambda l
: l
and not l
.startswith('#'),
1022 manifest
= self
._download
_webpage
(manifest_url
, video_id
, u
'Downloading formats manifest')
1023 formats_urls
= _get_urls(manifest
)
1024 for format_url
in formats_urls
:
1025 itag
= self
._search
_regex
(r
'itag/(\d+?)/', format_url
, 'itag')
1026 url_map
[itag
] = format_url
1029 def _extract_annotations(self
, video_id
):
1030 url
= 'https://www.youtube.com/annotations_invideo?features=1&legacy=1&video_id=%s' % video_id
1031 return self
._download
_webpage
(url
, video_id
, note
=u
'Searching for annotations.', errnote
=u
'Unable to download video annotations.')
1033 def _real_extract(self
, url
):
1035 u
'http' if self
._downloader
.params
.get('prefer_insecure', False)
1038 # Extract original video URL from URL with redirection, like age verification, using next_url parameter
1039 mobj
= re
.search(self
._NEXT
_URL
_RE
, url
)
1041 url
= proto
+ '://www.youtube.com/' + compat_urllib_parse
.unquote(mobj
.group(1)).lstrip('/')
1042 video_id
= self
.extract_id(url
)
1045 url
= proto
+ '://www.youtube.com/watch?v=%s&gl=US&hl=en&has_verified=1' % video_id
1046 video_webpage
= self
._download
_webpage
(url
, video_id
)
1048 # Attempt to extract SWF player URL
1049 mobj
= re
.search(r
'swfConfig.*?"(https?:\\/\\/.*?watch.*?-.*?\.swf)"', video_webpage
)
1050 if mobj
is not None:
1051 player_url
= re
.sub(r
'\\(.)', r
'\1', mobj
.group(1))
1056 self
.report_video_info_webpage_download(video_id
)
1057 if re
.search(r
'player-age-gate-content">', video_webpage
) is not None:
1058 self
.report_age_confirmation()
1060 # We simulate the access to the video from www.youtube.com/v/{video_id}
1061 # this can be viewed without login into Youtube
1062 data
= compat_urllib_parse
.urlencode({'video_id': video_id
,
1063 'el': 'player_embedded',
1066 'eurl': 'https://youtube.googleapis.com/v/' + video_id
,
1070 video_info_url
= proto
+ '://www.youtube.com/get_video_info?' + data
1071 video_info_webpage
= self
._download
_webpage
(video_info_url
, video_id
,
1073 errnote
='unable to download video info webpage')
1074 video_info
= compat_parse_qs(video_info_webpage
)
1077 for el_type
in ['&el=embedded', '&el=detailpage', '&el=vevo', '']:
1078 video_info_url
= (proto
+ '://www.youtube.com/get_video_info?&video_id=%s%s&ps=default&eurl=&gl=US&hl=en'
1079 % (video_id
, el_type
))
1080 video_info_webpage
= self
._download
_webpage
(video_info_url
, video_id
,
1082 errnote
='unable to download video info webpage')
1083 video_info
= compat_parse_qs(video_info_webpage
)
1084 if 'token' in video_info
:
1086 if 'token' not in video_info
:
1087 if 'reason' in video_info
:
1088 raise ExtractorError(
1089 u
'YouTube said: %s' % video_info
['reason'][0],
1090 expected
=True, video_id
=video_id
)
1092 raise ExtractorError(
1093 u
'"token" parameter not in video info for unknown reason',
1096 if 'view_count' in video_info
:
1097 view_count
= int(video_info
['view_count'][0])
1101 # Check for "rental" videos
1102 if 'ypc_video_rental_bar_text' in video_info
and 'author' not in video_info
:
1103 raise ExtractorError(u
'"rental" videos not supported')
1105 # Start extracting information
1106 self
.report_information_extraction(video_id
)
1109 if 'author' not in video_info
:
1110 raise ExtractorError(u
'Unable to extract uploader name')
1111 video_uploader
= compat_urllib_parse
.unquote_plus(video_info
['author'][0])
1114 video_uploader_id
= None
1115 mobj
= re
.search(r
'<link itemprop="url" href="http://www.youtube.com/(?:user|channel)/([^"]+)">', video_webpage
)
1116 if mobj
is not None:
1117 video_uploader_id
= mobj
.group(1)
1119 self
._downloader
.report_warning(u
'unable to extract uploader nickname')
1122 if 'title' in video_info
:
1123 video_title
= video_info
['title'][0]
1125 self
._downloader
.report_warning(u
'Unable to extract video title')
1129 # We try first to get a high quality image:
1130 m_thumb
= re
.search(r
'<span itemprop="thumbnail".*?href="(.*?)">',
1131 video_webpage
, re
.DOTALL
)
1132 if m_thumb
is not None:
1133 video_thumbnail
= m_thumb
.group(1)
1134 elif 'thumbnail_url' not in video_info
:
1135 self
._downloader
.report_warning(u
'unable to extract video thumbnail')
1136 video_thumbnail
= None
1137 else: # don't panic if we can't find it
1138 video_thumbnail
= compat_urllib_parse
.unquote_plus(video_info
['thumbnail_url'][0])
1142 mobj
= re
.search(r
'(?s)id="eow-date.*?>(.*?)</span>', video_webpage
)
1145 r
'(?s)id="watch-uploader-info".*?>.*?(?:Published|Uploaded|Streamed live) on (.*?)</strong>',
1147 if mobj
is not None:
1148 upload_date
= ' '.join(re
.sub(r
'[/,-]', r
' ', mobj
.group(1)).split())
1149 upload_date
= unified_strdate(upload_date
)
1151 m_cat_container
= get_element_by_id("eow-category", video_webpage
)
1153 category
= self
._html
_search
_regex
(
1154 r
'(?s)<a[^<]+>(.*?)</a>', m_cat_container
, 'category',
1156 video_categories
= None if category
is None else [category
]
1158 video_categories
= None
1161 video_description
= get_element_by_id("eow-description", video_webpage
)
1162 if video_description
:
1163 video_description
= re
.sub(r
'''(?x)
1165 (?:[a-zA-Z-]+="[^"]+"\s+)*?
1167 (?:[a-zA-Z-]+="[^"]+"\s+)*?
1168 class="yt-uix-redirect-link"\s*>
1171 ''', r
'\1', video_description
)
1172 video_description
= clean_html(video_description
)
1174 fd_mobj
= re
.search(r
'<meta name="description" content="([^"]+)"', video_webpage
)
1176 video_description
= unescapeHTML(fd_mobj
.group(1))
1178 video_description
= u
''
1180 def _extract_count(klass
):
1181 count
= self
._search
_regex
(
1182 r
'class="%s">([\d,]+)</span>' % re
.escape(klass
),
1183 video_webpage
, klass
, default
=None)
1184 if count
is not None:
1185 return int(count
.replace(',', ''))
1187 like_count
= _extract_count(u
'likes-count')
1188 dislike_count
= _extract_count(u
'dislikes-count')
1191 video_subtitles
= self
.extract_subtitles(video_id
, video_webpage
)
1193 if self
._downloader
.params
.get('listsubtitles', False):
1194 self
._list
_available
_subtitles
(video_id
, video_webpage
)
1197 if 'length_seconds' not in video_info
:
1198 self
._downloader
.report_warning(u
'unable to extract video duration')
1199 video_duration
= None
1201 video_duration
= int(compat_urllib_parse
.unquote_plus(video_info
['length_seconds'][0]))
1204 video_annotations
= None
1205 if self
._downloader
.params
.get('writeannotations', False):
1206 video_annotations
= self
._extract
_annotations
(video_id
)
1208 # Decide which formats to download
1210 mobj
= re
.search(r
';ytplayer\.config\s*=\s*({.*?});', video_webpage
)
1212 raise ValueError('Could not find vevo ID')
1213 json_code
= uppercase_escape(mobj
.group(1))
1214 ytplayer_config
= json
.loads(json_code
)
1215 args
= ytplayer_config
['args']
1216 # Easy way to know if the 's' value is in url_encoded_fmt_stream_map
1217 # this signatures are encrypted
1218 if 'url_encoded_fmt_stream_map' not in args
:
1219 raise ValueError(u
'No stream_map present') # caught below
1220 re_signature
= re
.compile(r
'[&,]s=')
1221 m_s
= re_signature
.search(args
['url_encoded_fmt_stream_map'])
1223 self
.to_screen(u
'%s: Encrypted signatures detected.' % video_id
)
1224 video_info
['url_encoded_fmt_stream_map'] = [args
['url_encoded_fmt_stream_map']]
1225 m_s
= re_signature
.search(args
.get('adaptive_fmts', u
''))
1227 if 'adaptive_fmts' in video_info
:
1228 video_info
['adaptive_fmts'][0] += ',' + args
['adaptive_fmts']
1230 video_info
['adaptive_fmts'] = [args
['adaptive_fmts']]
1234 def _map_to_format_list(urlmap
):
1236 for itag
, video_real_url
in urlmap
.items():
1239 'url': video_real_url
,
1240 'player_url': player_url
,
1242 if itag
in self
._formats
:
1243 dct
.update(self
._formats
[itag
])
1247 if 'conn' in video_info
and video_info
['conn'][0].startswith('rtmp'):
1248 self
.report_rtmp_download()
1250 'format_id': '_rtmp',
1252 'url': video_info
['conn'][0],
1253 'player_url': player_url
,
1255 elif len(video_info
.get('url_encoded_fmt_stream_map', [])) >= 1 or len(video_info
.get('adaptive_fmts', [])) >= 1:
1256 encoded_url_map
= video_info
.get('url_encoded_fmt_stream_map', [''])[0] + ',' + video_info
.get('adaptive_fmts',[''])[0]
1257 if 'rtmpe%3Dyes' in encoded_url_map
:
1258 raise ExtractorError('rtmpe downloads are not supported, see https://github.com/rg3/youtube-dl/issues/343 for more information.', expected
=True)
1260 for url_data_str
in encoded_url_map
.split(','):
1261 url_data
= compat_parse_qs(url_data_str
)
1262 if 'itag' in url_data
and 'url' in url_data
:
1263 url
= url_data
['url'][0]
1264 if 'sig' in url_data
:
1265 url
+= '&signature=' + url_data
['sig'][0]
1266 elif 's' in url_data
:
1267 encrypted_sig
= url_data
['s'][0]
1268 if self
._downloader
.params
.get('verbose'):
1270 if player_url
is None:
1271 player_version
= 'unknown'
1273 player_version
= self
._search
_regex
(
1274 r
'-(.+)\.swf$', player_url
,
1275 u
'flash player', fatal
=False)
1276 player_desc
= 'flash player %s' % player_version
1278 player_version
= self
._search
_regex
(
1279 r
'html5player-(.+?)\.js', video_webpage
,
1280 'html5 player', fatal
=False)
1281 player_desc
= u
'html5 player %s' % player_version
1283 parts_sizes
= u
'.'.join(compat_str(len(part
)) for part
in encrypted_sig
.split('.'))
1284 self
.to_screen(u
'encrypted signature length %d (%s), itag %s, %s' %
1285 (len(encrypted_sig
), parts_sizes
, url_data
['itag'][0], player_desc
))
1288 jsplayer_url_json
= self
._search
_regex
(
1289 r
'"assets":.+?"js":\s*("[^"]+")',
1290 video_webpage
, u
'JS player URL')
1291 player_url
= json
.loads(jsplayer_url_json
)
1293 signature
= self
._decrypt
_signature
(
1294 encrypted_sig
, video_id
, player_url
, age_gate
)
1295 url
+= '&signature=' + signature
1296 if 'ratebypass' not in url
:
1297 url
+= '&ratebypass=yes'
1298 url_map
[url_data
['itag'][0]] = url
1299 formats
= _map_to_format_list(url_map
)
1300 elif video_info
.get('hlsvp'):
1301 manifest_url
= video_info
['hlsvp'][0]
1302 url_map
= self
._extract
_from
_m
3u8(manifest_url
, video_id
)
1303 formats
= _map_to_format_list(url_map
)
1305 raise ExtractorError(u
'no conn, hlsvp or url_encoded_fmt_stream_map information found in video info')
1307 # Look for the DASH manifest
1308 if (self
._downloader
.params
.get('youtube_include_dash_manifest', False)):
1310 # The DASH manifest used needs to be the one from the original video_webpage.
1311 # The one found in get_video_info seems to be using different signatures.
1312 # However, in the case of an age restriction there won't be any embedded dashmpd in the video_webpage.
1313 # Luckily, it seems, this case uses some kind of default signature (len == 86), so the
1314 # combination of get_video_info and the _static_decrypt_signature() decryption fallback will work here.
1316 dash_manifest_url
= video_info
.get('dashmpd')[0]
1318 dash_manifest_url
= ytplayer_config
['args']['dashmpd']
1319 def decrypt_sig(mobj
):
1321 dec_s
= self
._decrypt
_signature
(s
, video_id
, player_url
, age_gate
)
1322 return '/signature/%s' % dec_s
1323 dash_manifest_url
= re
.sub(r
'/s/([\w\.]+)', decrypt_sig
, dash_manifest_url
)
1324 dash_doc
= self
._download
_xml
(
1325 dash_manifest_url
, video_id
,
1326 note
=u
'Downloading DASH manifest',
1327 errnote
=u
'Could not download DASH manifest')
1328 for r
in dash_doc
.findall(u
'.//{urn:mpeg:DASH:schema:MPD:2011}Representation'):
1329 url_el
= r
.find('{urn:mpeg:DASH:schema:MPD:2011}BaseURL')
1332 format_id
= r
.attrib
['id']
1333 video_url
= url_el
.text
1334 filesize
= int_or_none(url_el
.attrib
.get('{http://youtube.com/yt/2012/10/10}contentLength'))
1336 'format_id': format_id
,
1338 'width': int_or_none(r
.attrib
.get('width')),
1339 'tbr': int_or_none(r
.attrib
.get('bandwidth'), 1000),
1340 'asr': int_or_none(r
.attrib
.get('audioSamplingRate')),
1341 'filesize': filesize
,
1344 existing_format
= next(
1345 fo
for fo
in formats
1346 if fo
['format_id'] == format_id
)
1347 except StopIteration:
1348 f
.update(self
._formats
.get(format_id
, {}))
1351 existing_format
.update(f
)
1353 except (ExtractorError
, KeyError) as e
:
1354 self
.report_warning(u
'Skipping DASH manifest: %s' % e
, video_id
)
1356 self
._sort
_formats
(formats
)
1360 'uploader': video_uploader
,
1361 'uploader_id': video_uploader_id
,
1362 'upload_date': upload_date
,
1363 'title': video_title
,
1364 'thumbnail': video_thumbnail
,
1365 'description': video_description
,
1366 'categories': video_categories
,
1367 'subtitles': video_subtitles
,
1368 'duration': video_duration
,
1369 'age_limit': 18 if age_gate
else 0,
1370 'annotations': video_annotations
,
1371 'webpage_url': proto
+ '://www.youtube.com/watch?v=%s' % video_id
,
1372 'view_count': view_count
,
1373 'like_count': like_count
,
1374 'dislike_count': dislike_count
,
1378 class YoutubePlaylistIE(YoutubeBaseInfoExtractor
):
1379 IE_DESC
= u
'YouTube.com playlists'
1380 _VALID_URL
= r
"""(?x)(?:
1385 (?:course|view_play_list|my_playlists|artist|playlist|watch)
1386 \? (?:.*?&)*? (?:p|a|list)=
1390 (?:PL|LL|EC|UU|FL|RD)?[0-9A-Za-z-_]{10,}
1391 # Top tracks, they can also include dots
1396 ((?:PL|LL|EC|UU|FL|RD)[0-9A-Za-z-_]{10,})
1398 _TEMPLATE_URL
= 'https://www.youtube.com/playlist?list=%s'
1399 _MORE_PAGES_INDICATOR
= r
'data-link-type="next"'
1400 _VIDEO_RE
= r
'href="\s*/watch\?v=(?P<id>[0-9A-Za-z_-]{11})&[^"]*?index=(?P<index>\d+)'
1401 IE_NAME
= u
'youtube:playlist'
1403 def _real_initialize(self
):
1406 def _ids_to_results(self
, ids
):
1407 return [self
.url_result(vid_id
, 'Youtube', video_id
=vid_id
)
1410 def _extract_mix(self
, playlist_id
):
1411 # The mixes are generated from a a single video
1412 # the id of the playlist is just 'RD' + video_id
1413 url
= 'https://youtube.com/watch?v=%s&list=%s' % (playlist_id
[-11:], playlist_id
)
1414 webpage
= self
._download
_webpage
(url
, playlist_id
, u
'Downloading Youtube mix')
1415 search_title
= lambda class_name
: get_element_by_attribute('class', class_name
, webpage
)
1416 title_span
= (search_title('playlist-title') or
1417 search_title('title long-title') or search_title('title'))
1418 title
= clean_html(title_span
)
1419 video_re
= r
'''(?x)data-video-username=".*?".*?
1420 href="/watch\?v=([0-9A-Za-z_-]{11})&[^"]*?list=%s''' % re
.escape(playlist_id
)
1421 ids
= orderedSet(re
.findall(video_re
, webpage
, flags
=re
.DOTALL
))
1422 url_results
= self
._ids
_to
_results
(ids
)
1424 return self
.playlist_result(url_results
, playlist_id
, title
)
1426 def _real_extract(self
, url
):
1427 # Extract playlist id
1428 mobj
= re
.match(self
._VALID
_URL
, url
)
1430 raise ExtractorError(u
'Invalid URL: %s' % url
)
1431 playlist_id
= mobj
.group(1) or mobj
.group(2)
1433 # Check if it's a video-specific URL
1434 query_dict
= compat_urlparse
.parse_qs(compat_urlparse
.urlparse(url
).query
)
1435 if 'v' in query_dict
:
1436 video_id
= query_dict
['v'][0]
1437 if self
._downloader
.params
.get('noplaylist'):
1438 self
.to_screen(u
'Downloading just video %s because of --no-playlist' % video_id
)
1439 return self
.url_result(video_id
, 'Youtube', video_id
=video_id
)
1441 self
.to_screen(u
'Downloading playlist %s - add --no-playlist to just download video %s' % (playlist_id
, video_id
))
1443 if playlist_id
.startswith('RD'):
1444 # Mixes require a custom extraction process
1445 return self
._extract
_mix
(playlist_id
)
1446 if playlist_id
.startswith('TL'):
1447 raise ExtractorError(u
'For downloading YouTube.com top lists, use '
1448 u
'the "yttoplist" keyword, for example "youtube-dl \'yttoplist:music:Top Tracks\'"', expected
=True)
1450 url
= self
._TEMPLATE
_URL
% playlist_id
1451 page
= self
._download
_webpage
(url
, playlist_id
)
1452 more_widget_html
= content_html
= page
1454 # Check if the playlist exists or is private
1455 if re
.search(r
'<div class="yt-alert-message">[^<]*?(The|This) playlist (does not exist|is private)[^<]*?</div>', page
) is not None:
1456 raise ExtractorError(
1457 u
'The playlist doesn\'t exist or is private, use --username or '
1458 '--netrc to access it.',
1461 # Extract the video ids from the playlist pages
1464 for page_num
in itertools
.count(1):
1465 matches
= re
.finditer(self
._VIDEO
_RE
, content_html
)
1466 # We remove the duplicates and the link with index 0
1467 # (it's not the first video of the playlist)
1468 new_ids
= orderedSet(m
.group('id') for m
in matches
if m
.group('index') != '0')
1471 mobj
= re
.search(r
'data-uix-load-more-href="/?(?P<more>[^"]+)"', more_widget_html
)
1475 more
= self
._download
_json
(
1476 'https://youtube.com/%s' % mobj
.group('more'), playlist_id
,
1477 'Downloading page #%s' % page_num
,
1478 transform_source
=uppercase_escape
)
1479 content_html
= more
['content_html']
1480 more_widget_html
= more
['load_more_widget_html']
1482 playlist_title
= self
._html
_search
_regex
(
1483 r
'(?s)<h1 class="pl-header-title[^"]*">\s*(.*?)\s*</h1>',
1486 url_results
= self
._ids
_to
_results
(ids
)
1487 return self
.playlist_result(url_results
, playlist_id
, playlist_title
)
1490 class YoutubeTopListIE(YoutubePlaylistIE
):
1491 IE_NAME
= u
'youtube:toplist'
1492 IE_DESC
= (u
'YouTube.com top lists, "yttoplist:{channel}:{list title}"'
1493 u
' (Example: "yttoplist:music:Top Tracks")')
1494 _VALID_URL
= r
'yttoplist:(?P<chann>.*?):(?P<title>.*?)$'
1496 def _real_extract(self
, url
):
1497 mobj
= re
.match(self
._VALID
_URL
, url
)
1498 channel
= mobj
.group('chann')
1499 title
= mobj
.group('title')
1500 query
= compat_urllib_parse
.urlencode({'title': title}
)
1501 playlist_re
= 'href="([^"]+?%s.*?)"' % re
.escape(query
)
1502 channel_page
= self
._download
_webpage
('https://www.youtube.com/%s' % channel
, title
)
1503 link
= self
._html
_search
_regex
(playlist_re
, channel_page
, u
'list')
1504 url
= compat_urlparse
.urljoin('https://www.youtube.com/', link
)
1506 video_re
= r
'data-index="\d+".*?data-video-id="([0-9A-Za-z_-]{11})"'
1508 # sometimes the webpage doesn't contain the videos
1509 # retry until we get them
1510 for i
in itertools
.count(0):
1511 msg
= u
'Downloading Youtube mix'
1513 msg
+= ', retry #%d' % i
1514 webpage
= self
._download
_webpage
(url
, title
, msg
)
1515 ids
= orderedSet(re
.findall(video_re
, webpage
))
1518 url_results
= self
._ids
_to
_results
(ids
)
1519 return self
.playlist_result(url_results
, playlist_title
=title
)
1522 class YoutubeChannelIE(InfoExtractor
):
1523 IE_DESC
= u
'YouTube.com channels'
1524 _VALID_URL
= r
"^(?:https?://)?(?:youtu\.be|(?:\w+\.)?youtube(?:-nocookie)?\.com)/channel/([0-9A-Za-z_-]+)"
1525 _MORE_PAGES_INDICATOR
= 'yt-uix-load-more'
1526 _MORE_PAGES_URL
= 'https://www.youtube.com/c4_browse_ajax?action_load_more_videos=1&flow=list&paging=%s&view=0&sort=da&channel_id=%s'
1527 IE_NAME
= u
'youtube:channel'
1529 def extract_videos_from_page(self
, page
):
1531 for mobj
in re
.finditer(r
'href="/watch\?v=([0-9A-Za-z_-]+)&?', page
):
1532 if mobj
.group(1) not in ids_in_page
:
1533 ids_in_page
.append(mobj
.group(1))
1536 def _real_extract(self
, url
):
1537 # Extract channel id
1538 mobj
= re
.match(self
._VALID
_URL
, url
)
1540 raise ExtractorError(u
'Invalid URL: %s' % url
)
1542 # Download channel page
1543 channel_id
= mobj
.group(1)
1545 url
= 'https://www.youtube.com/channel/%s/videos' % channel_id
1546 channel_page
= self
._download
_webpage
(url
, channel_id
)
1547 autogenerated
= re
.search(r
'''(?x)
1549 channel-header-autogenerated-label|
1550 yt-channel-title-autogenerated
1551 )[^"]*"''', channel_page
) is not None
1554 # The videos are contained in a single page
1555 # the ajax pages can't be used, they are empty
1556 video_ids
= self
.extract_videos_from_page(channel_page
)
1558 # Download all channel pages using the json-based channel_ajax query
1559 for pagenum
in itertools
.count(1):
1560 url
= self
._MORE
_PAGES
_URL
% (pagenum
, channel_id
)
1561 page
= self
._download
_json
(
1562 url
, channel_id
, note
=u
'Downloading page #%s' % pagenum
,
1563 transform_source
=uppercase_escape
)
1565 ids_in_page
= self
.extract_videos_from_page(page
['content_html'])
1566 video_ids
.extend(ids_in_page
)
1568 if self
._MORE
_PAGES
_INDICATOR
not in page
['load_more_widget_html']:
1571 self
._downloader
.to_screen(u
'[youtube] Channel %s: Found %i videos' % (channel_id
, len(video_ids
)))
1573 url_entries
= [self
.url_result(video_id
, 'Youtube', video_id
=video_id
)
1574 for video_id
in video_ids
]
1575 return self
.playlist_result(url_entries
, channel_id
)
1578 class YoutubeUserIE(InfoExtractor
):
1579 IE_DESC
= u
'YouTube.com user videos (URL or "ytuser" keyword)'
1580 _VALID_URL
= r
'(?:(?:(?:https?://)?(?:\w+\.)?youtube\.com/(?:user/)?(?!(?:attribution_link|watch|results)(?:$|[^a-z_A-Z0-9-])))|ytuser:)(?!feed/)([A-Za-z0-9_-]+)'
1581 _TEMPLATE_URL
= 'https://gdata.youtube.com/feeds/api/users/%s'
1582 _GDATA_PAGE_SIZE
= 50
1583 _GDATA_URL
= 'https://gdata.youtube.com/feeds/api/users/%s/uploads?max-results=%d&start-index=%d&alt=json'
1584 IE_NAME
= u
'youtube:user'
1587 def suitable(cls
, url
):
1588 # Don't return True if the url can be extracted with other youtube
1589 # extractor, the regex would is too permissive and it would match.
1590 other_ies
= iter(klass
for (name
, klass
) in globals().items() if name
.endswith('IE') and klass
is not cls
)
1591 if any(ie
.suitable(url
) for ie
in other_ies
): return False
1592 else: return super(YoutubeUserIE
, cls
).suitable(url
)
1594 def _real_extract(self
, url
):
1596 mobj
= re
.match(self
._VALID
_URL
, url
)
1598 raise ExtractorError(u
'Invalid URL: %s' % url
)
1600 username
= mobj
.group(1)
1602 # Download video ids using YouTube Data API. Result size per
1603 # query is limited (currently to 50 videos) so we need to query
1604 # page by page until there are no video ids - it means we got
1607 def download_page(pagenum
):
1608 start_index
= pagenum
* self
._GDATA
_PAGE
_SIZE
+ 1
1610 gdata_url
= self
._GDATA
_URL
% (username
, self
._GDATA
_PAGE
_SIZE
, start_index
)
1611 page
= self
._download
_webpage
(
1612 gdata_url
, username
,
1613 u
'Downloading video ids from %d to %d' % (
1614 start_index
, start_index
+ self
._GDATA
_PAGE
_SIZE
))
1617 response
= json
.loads(page
)
1618 except ValueError as err
:
1619 raise ExtractorError(u
'Invalid JSON in API response: ' + compat_str(err
))
1620 if 'entry' not in response
['feed']:
1623 # Extract video identifiers
1624 entries
= response
['feed']['entry']
1625 for entry
in entries
:
1626 title
= entry
['title']['$t']
1627 video_id
= entry
['id']['$t'].split('/')[-1]
1631 'ie_key': 'Youtube',
1635 url_results
= PagedList(download_page
, self
._GDATA
_PAGE
_SIZE
)
1637 return self
.playlist_result(url_results
, playlist_title
=username
)
1640 class YoutubeSearchIE(SearchInfoExtractor
):
1641 IE_DESC
= u
'YouTube.com searches'
1642 _API_URL
= u
'https://gdata.youtube.com/feeds/api/videos?q=%s&start-index=%i&max-results=50&v=2&alt=jsonc'
1644 IE_NAME
= u
'youtube:search'
1645 _SEARCH_KEY
= 'ytsearch'
1647 def _get_n_results(self
, query
, n
):
1648 """Get a specified number of results for a query"""
1655 while (PAGE_SIZE
* pagenum
) < limit
:
1656 result_url
= self
._API
_URL
% (
1657 compat_urllib_parse
.quote_plus(query
.encode('utf-8')),
1658 (PAGE_SIZE
* pagenum
) + 1)
1659 data_json
= self
._download
_webpage
(
1660 result_url
, video_id
=u
'query "%s"' % query
,
1661 note
=u
'Downloading page %s' % (pagenum
+ 1),
1662 errnote
=u
'Unable to download API page')
1663 data
= json
.loads(data_json
)
1664 api_response
= data
['data']
1666 if 'items' not in api_response
:
1667 raise ExtractorError(
1668 u
'[youtube] No video results', expected
=True)
1670 new_ids
= list(video
['id'] for video
in api_response
['items'])
1671 video_ids
+= new_ids
1673 limit
= min(n
, api_response
['totalItems'])
1676 if len(video_ids
) > n
:
1677 video_ids
= video_ids
[:n
]
1678 videos
= [self
.url_result(video_id
, 'Youtube', video_id
=video_id
)
1679 for video_id
in video_ids
]
1680 return self
.playlist_result(videos
, query
)
1683 class YoutubeSearchDateIE(YoutubeSearchIE
):
1684 IE_NAME
= YoutubeSearchIE
.IE_NAME
+ ':date'
1685 _API_URL
= 'https://gdata.youtube.com/feeds/api/videos?q=%s&start-index=%i&max-results=50&v=2&alt=jsonc&orderby=published'
1686 _SEARCH_KEY
= 'ytsearchdate'
1687 IE_DESC
= u
'YouTube.com searches, newest videos first'
1690 class YoutubeSearchURLIE(InfoExtractor
):
1691 IE_DESC
= u
'YouTube.com search URLs'
1692 IE_NAME
= u
'youtube:search_url'
1693 _VALID_URL
= r
'https?://(?:www\.)?youtube\.com/results\?(.*?&)?search_query=(?P<query>[^&]+)(?:[&]|$)'
1695 def _real_extract(self
, url
):
1696 mobj
= re
.match(self
._VALID
_URL
, url
)
1697 query
= compat_urllib_parse
.unquote_plus(mobj
.group('query'))
1699 webpage
= self
._download
_webpage
(url
, query
)
1700 result_code
= self
._search
_regex
(
1701 r
'(?s)<ol id="search-results"(.*?)</ol>', webpage
, u
'result HTML')
1703 part_codes
= re
.findall(
1704 r
'(?s)<h3 class="yt-lockup-title">(.*?)</h3>', result_code
)
1706 for part_code
in part_codes
:
1707 part_title
= self
._html
_search
_regex
(
1708 r
'(?s)title="([^"]+)"', part_code
, 'item title', fatal
=False)
1709 part_url_snippet
= self
._html
_search
_regex
(
1710 r
'(?s)href="([^"]+)"', part_code
, 'item URL')
1711 part_url
= compat_urlparse
.urljoin(
1712 'https://www.youtube.com/', part_url_snippet
)
1716 'title': part_title
,
1720 '_type': 'playlist',
1726 class YoutubeShowIE(InfoExtractor
):
1727 IE_DESC
= u
'YouTube.com (multi-season) shows'
1728 _VALID_URL
= r
'https?://www\.youtube\.com/show/(.*)'
1729 IE_NAME
= u
'youtube:show'
1731 def _real_extract(self
, url
):
1732 mobj
= re
.match(self
._VALID
_URL
, url
)
1733 show_name
= mobj
.group(1)
1734 webpage
= self
._download
_webpage
(url
, show_name
, u
'Downloading show webpage')
1735 # There's one playlist for each season of the show
1736 m_seasons
= list(re
.finditer(r
'href="(/playlist\?list=.*?)"', webpage
))
1737 self
.to_screen(u
'%s: Found %s seasons' % (show_name
, len(m_seasons
)))
1738 return [self
.url_result('https://www.youtube.com' + season
.group(1), 'YoutubePlaylist') for season
in m_seasons
]
1741 class YoutubeFeedsInfoExtractor(YoutubeBaseInfoExtractor
):
1743 Base class for extractors that fetch info from
1744 http://www.youtube.com/feed_ajax
1745 Subclasses must define the _FEED_NAME and _PLAYLIST_TITLE properties.
1747 _LOGIN_REQUIRED
= True
1748 # use action_load_personal_feed instead of action_load_system_feed
1749 _PERSONAL_FEED
= False
1752 def _FEED_TEMPLATE(self
):
1753 action
= 'action_load_system_feed'
1754 if self
._PERSONAL
_FEED
:
1755 action
= 'action_load_personal_feed'
1756 return 'https://www.youtube.com/feed_ajax?%s=1&feed_name=%s&paging=%%s' % (action
, self
._FEED
_NAME
)
1760 return u
'youtube:%s' % self
._FEED
_NAME
1762 def _real_initialize(self
):
1765 def _real_extract(self
, url
):
1768 for i
in itertools
.count(1):
1769 info
= self
._download
_json
(self
._FEED
_TEMPLATE
% paging
,
1770 u
'%s feed' % self
._FEED
_NAME
,
1771 u
'Downloading page %s' % i
)
1772 feed_html
= info
.get('feed_html') or info
.get('content_html')
1773 m_ids
= re
.finditer(r
'"/watch\?v=(.*?)["&]', feed_html
)
1774 ids
= orderedSet(m
.group(1) for m
in m_ids
)
1775 feed_entries
.extend(
1776 self
.url_result(video_id
, 'Youtube', video_id
=video_id
)
1777 for video_id
in ids
)
1779 r
'data-uix-load-more-href="/?[^"]+paging=(?P<paging>\d+)',
1783 paging
= mobj
.group('paging')
1784 return self
.playlist_result(feed_entries
, playlist_title
=self
._PLAYLIST
_TITLE
)
1786 class YoutubeSubscriptionsIE(YoutubeFeedsInfoExtractor
):
1787 IE_DESC
= u
'YouTube.com subscriptions feed, "ytsubs" keyword (requires authentication)'
1788 _VALID_URL
= r
'https?://www\.youtube\.com/feed/subscriptions|:ytsubs(?:criptions)?'
1789 _FEED_NAME
= 'subscriptions'
1790 _PLAYLIST_TITLE
= u
'Youtube Subscriptions'
1792 class YoutubeRecommendedIE(YoutubeFeedsInfoExtractor
):
1793 IE_DESC
= u
'YouTube.com recommended videos, "ytrec" keyword (requires authentication)'
1794 _VALID_URL
= r
'https?://www\.youtube\.com/feed/recommended|:ytrec(?:ommended)?'
1795 _FEED_NAME
= 'recommended'
1796 _PLAYLIST_TITLE
= u
'Youtube Recommended videos'
1798 class YoutubeWatchLaterIE(YoutubeFeedsInfoExtractor
):
1799 IE_DESC
= u
'Youtube watch later list, "ytwatchlater" keyword (requires authentication)'
1800 _VALID_URL
= r
'https?://www\.youtube\.com/feed/watch_later|:ytwatchlater'
1801 _FEED_NAME
= 'watch_later'
1802 _PLAYLIST_TITLE
= u
'Youtube Watch Later'
1803 _PERSONAL_FEED
= True
1805 class YoutubeHistoryIE(YoutubeFeedsInfoExtractor
):
1806 IE_DESC
= u
'Youtube watch history, "ythistory" keyword (requires authentication)'
1807 _VALID_URL
= u
'https?://www\.youtube\.com/feed/history|:ythistory'
1808 _FEED_NAME
= 'history'
1809 _PERSONAL_FEED
= True
1810 _PLAYLIST_TITLE
= u
'Youtube Watch History'
1812 class YoutubeFavouritesIE(YoutubeBaseInfoExtractor
):
1813 IE_NAME
= u
'youtube:favorites'
1814 IE_DESC
= u
'YouTube.com favourite videos, "ytfav" keyword (requires authentication)'
1815 _VALID_URL
= r
'https?://www\.youtube\.com/my_favorites|:ytfav(?:ou?rites)?'
1816 _LOGIN_REQUIRED
= True
1818 def _real_extract(self
, url
):
1819 webpage
= self
._download
_webpage
('https://www.youtube.com/my_favorites', 'Youtube Favourites videos')
1820 playlist_id
= self
._search
_regex
(r
'list=(.+?)["&]', webpage
, u
'favourites playlist id')
1821 return self
.url_result(playlist_id
, 'YoutubePlaylist')
1824 class YoutubeTruncatedURLIE(InfoExtractor
):
1825 IE_NAME
= 'youtube:truncated_url'
1826 IE_DESC
= False # Do not list
1827 _VALID_URL
= r
'''(?x)
1828 (?:https?://)?[^/]+/watch\?(?:
1830 annotation_id=annotation_[^&]+
1832 (?:https?://)?(?:www\.)?youtube\.com/attribution_link\?a=[^&]+$
1836 'url': 'http://www.youtube.com/watch?annotation_id=annotation_3951667041',
1837 'only_matching': True,
1840 def _real_extract(self
, url
):
1841 raise ExtractorError(
1842 u
'Did you forget to quote the URL? Remember that & is a meta '
1843 u
'character in most shells, so you want to put the URL in quotes, '
1845 u
'"http://www.youtube.com/watch?feature=foo&v=BaW_jenozKc" '
1846 u
' or simply youtube-dl BaW_jenozKc .',