1
0
mirror of https://github.com/ytdl-org/youtube-dl.git synced 2024-12-22 07:56:49 +00:00

Compare commits

...

22 Commits

Author SHA1 Message Date
Kaspar V.
a097a5a865
Merge a4759ae8ea into d55d1f423d 2024-12-21 10:50:30 +01:00
dirkf
d55d1f423d [YouTube] Always extract using MWEB API client
* temporary fix-up for 403 on download
* MWEB parameters from yt-dlp 2024-12-06
2024-12-16 12:38:51 +00:00
dirkf
eeafbbc3e5 [YouTube] Fix signature function extraction for 2f1832d2
* `_` was omitted from patterns
* thx yt-dlp/yt-dlp#11801

Co-authored-by: bashonly
2024-12-16 12:38:51 +00:00
dirkf
cd7c7b5edb [YouTube] Simplify pattern for nsig function name extraction 2024-12-16 12:38:51 +00:00
dirkf
eed784e15f [YouTube] Pass nsig value as return hook, fixes player 3bb1f723 2024-12-16 12:38:51 +00:00
dirkf
b4469a0f65 [YouTube] Handle player 3bb1f723
* fix signature code extraction
* raise if n function returns input value
* add new tests from yt-dlp

Co-authored-by: bashonly
2024-12-16 12:38:51 +00:00
dirkf
ce1e556b8f [jsinterp] Add return hook for player 3bb1f723
* set var `_ytdl_do_not_return` to a specific value in the scope of a function
* if an expression to be returned has that value, `return` becomes `void`
2024-12-16 12:38:51 +00:00
dirkf
f487b4a02a [jsinterp] Strip /* comments */ when parsing
* NB: _separate() is looking creaky
2024-12-16 12:38:51 +00:00
dirkf
60835ca16c [jsinterp] Fix and improve "methods"
* push, unshift return new length
* impove edge cases for push/pop, shift/unshift, forEach, indexOf, charCodeAt
* increase test coverage
2024-12-16 12:38:51 +00:00
dirkf
94fd774608 [jsinterp] Fix and improve split/join
* improve split/join edge cases
* correctly implement regex split (not like re.split)
2024-12-16 12:38:51 +00:00
dirkf
5dee6213ed [jsinterp] Fix and improve arithmetic operations
* addition becomes concat with a string operand
* improve handling of edgier cases
* arithmetic in float like JS (more places need cast to int?)
* increase test coverage
2024-12-16 12:38:51 +00:00
dirkf
81e64cacf2 [jsinterp] Support multiple indexing (eg a[1][2])
* extend single indexing with improved RE (should probably use/have used _separate_at_paren())
* fix some cases that should have given undefined, not throwing
* standardise RE group names
* support length of objects, like {1: 2, 3: 4, length: 42}
2024-12-16 12:38:51 +00:00
dirkf
c1a03b1ac3 [jsinterp] Fix and improve loose and strict equality operations
* reimplement loose equality according to MDN (eg, 1 == "1")
* improve strict equality (eg, "abc" === "abc" but 'abc' is not 'abc')
* add tests for above
2024-12-16 12:38:51 +00:00
dirkf
118c6d7a17 [jsinterp] Implement typeof operator 2024-12-16 12:38:51 +00:00
dirkf
f28d7178e4 [InfoExtractor] Use kwarg maxsplit for re.split
* May become kw-only in future Pythons
2024-12-16 12:38:51 +00:00
dirkf
a4759ae8ea
Fix ArteTVEmbedIE 2023-02-15 16:37:05 +00:00
dirkf
3bb4530d42
langauge_code 2023-02-15 00:29:39 +00:00
dirkf
b3b0520c82
Update
Based on yt-dlp/yt-dlp 051d6b4 merging [extractor/arte] yt-dlp/yt-dlp#3302: Move to v2 API

Authored by: fstirlitz, pukkandan
2023-02-15 00:24:38 +00:00
dirkf
cdcb4af1c3
Linted? 2023-02-14 18:29:26 +00:00
dirkf
82833b44d7
Merge branch 'master' into extractor/arte/improvement-switch-to-config-api-v2 2023-02-14 18:24:27 +00:00
dirkf
0c149f04d4
Update from yt-dlp 2023-02-14 18:18:47 +00:00
Kaspar Vollenweider
66eec4e640
WIP: arte config api v2 2021-07-26 07:56:29 +02:00
6 changed files with 882 additions and 388 deletions

View File

@ -1,4 +1,5 @@
#!/usr/bin/env python
# coding: utf-8
from __future__ import unicode_literals
@ -11,7 +12,7 @@ sys.path.insert(0, os.path.dirname(os.path.dirname(os.path.abspath(__file__))))
import math
import re
from youtube_dl.compat import compat_str
from youtube_dl.compat import compat_str as str
from youtube_dl.jsinterp import JS_Undefined, JSInterpreter
NaN = object()
@ -19,7 +20,7 @@ NaN = object()
class TestJSInterpreter(unittest.TestCase):
def _test(self, jsi_or_code, expected, func='f', args=()):
if isinstance(jsi_or_code, compat_str):
if isinstance(jsi_or_code, str):
jsi_or_code = JSInterpreter(jsi_or_code)
got = jsi_or_code.call_function(func, *args)
if expected is NaN:
@ -40,16 +41,27 @@ class TestJSInterpreter(unittest.TestCase):
self._test('function f(){return 42 + 7;}', 49)
self._test('function f(){return 42 + undefined;}', NaN)
self._test('function f(){return 42 + null;}', 42)
self._test('function f(){return 1 + "";}', '1')
self._test('function f(){return 42 + "7";}', '427')
self._test('function f(){return false + true;}', 1)
self._test('function f(){return "false" + true;}', 'falsetrue')
self._test('function f(){return '
'1 + "2" + [3,4] + {k: 56} + null + undefined + Infinity;}',
'123,4[object Object]nullundefinedInfinity')
def test_sub(self):
self._test('function f(){return 42 - 7;}', 35)
self._test('function f(){return 42 - undefined;}', NaN)
self._test('function f(){return 42 - null;}', 42)
self._test('function f(){return 42 - "7";}', 35)
self._test('function f(){return 42 - "spam";}', NaN)
def test_mul(self):
self._test('function f(){return 42 * 7;}', 294)
self._test('function f(){return 42 * undefined;}', NaN)
self._test('function f(){return 42 * null;}', 0)
self._test('function f(){return 42 * "7";}', 294)
self._test('function f(){return 42 * "eggs";}', NaN)
def test_div(self):
jsi = JSInterpreter('function f(a, b){return a / b;}')
@ -57,17 +69,26 @@ class TestJSInterpreter(unittest.TestCase):
self._test(jsi, NaN, args=(JS_Undefined, 1))
self._test(jsi, float('inf'), args=(2, 0))
self._test(jsi, 0, args=(0, 3))
self._test(jsi, 6, args=(42, 7))
self._test(jsi, 0, args=(42, float('inf')))
self._test(jsi, 6, args=("42", 7))
self._test(jsi, NaN, args=("spam", 7))
def test_mod(self):
self._test('function f(){return 42 % 7;}', 0)
self._test('function f(){return 42 % 0;}', NaN)
self._test('function f(){return 42 % undefined;}', NaN)
self._test('function f(){return 42 % "7";}', 0)
self._test('function f(){return 42 % "beans";}', NaN)
def test_exp(self):
self._test('function f(){return 42 ** 2;}', 1764)
self._test('function f(){return 42 ** undefined;}', NaN)
self._test('function f(){return 42 ** null;}', 1)
self._test('function f(){return undefined ** 0;}', 1)
self._test('function f(){return undefined ** 42;}', NaN)
self._test('function f(){return 42 ** "2";}', 1764)
self._test('function f(){return 42 ** "spam";}', NaN)
def test_calc(self):
self._test('function f(a){return 2*a+1;}', 7, args=[3])
@ -89,7 +110,35 @@ class TestJSInterpreter(unittest.TestCase):
self._test('function f(){return 19 & 21;}', 17)
self._test('function f(){return 11 >> 2;}', 2)
self._test('function f(){return []? 2+3: 4;}', 5)
# equality
self._test('function f(){return 1 == 1}', True)
self._test('function f(){return 1 == 1.0}', True)
self._test('function f(){return 1 == "1"}', True)
self._test('function f(){return 1 == 2}', False)
self._test('function f(){return 1 != "1"}', False)
self._test('function f(){return 1 != 2}', True)
self._test('function f(){var x = {a: 1}; var y = x; return x == y}', True)
self._test('function f(){var x = {a: 1}; return x == {a: 1}}', False)
self._test('function f(){return NaN == NaN}', False)
self._test('function f(){return null == undefined}', True)
self._test('function f(){return "spam, eggs" == "spam, eggs"}', True)
# strict equality
self._test('function f(){return 1 === 1}', True)
self._test('function f(){return 1 === 1.0}', True)
self._test('function f(){return 1 === "1"}', False)
self._test('function f(){return 1 === 2}', False)
self._test('function f(){var x = {a: 1}; var y = x; return x === y}', True)
self._test('function f(){var x = {a: 1}; return x === {a: 1}}', False)
self._test('function f(){return NaN === NaN}', False)
self._test('function f(){return null === undefined}', False)
self._test('function f(){return null === null}', True)
self._test('function f(){return undefined === undefined}', True)
self._test('function f(){return "uninterned" === "uninterned"}', True)
self._test('function f(){return 1 === 1}', True)
self._test('function f(){return 1 === "1"}', False)
self._test('function f(){return 1 !== 1}', False)
self._test('function f(){return 1 !== "1"}', True)
# expressions
self._test('function f(){return 0 && 1 || 2;}', 2)
self._test('function f(){return 0 ?? 42;}', 0)
self._test('function f(){return "life, the universe and everything" < 42;}', False)
@ -111,7 +160,6 @@ class TestJSInterpreter(unittest.TestCase):
self._test('function f(){var x = 20; x += 30 + 1; return x;}', 51)
self._test('function f(){var x = 20; x -= 30 + 1; return x;}', -11)
@unittest.skip('Not yet fully implemented')
def test_comments(self):
self._test('''
function f() {
@ -130,6 +178,15 @@ class TestJSInterpreter(unittest.TestCase):
}
''', 3)
self._test('''
function f() {
var x = ( /* 1 + */ 2 +
/* 30 * 40 */
50);
return x;
}
''', 52)
def test_precedence(self):
self._test('''
function f() {
@ -266,7 +323,20 @@ class TestJSInterpreter(unittest.TestCase):
self._test('function f() { return (l=[0,1,2,3], function(a, b){return a+b})((l[1], l[2]), l[3]) }', 5)
def test_void(self):
self._test('function f() { return void 42; }', None)
self._test('function f() { return void 42; }', JS_Undefined)
def test_typeof(self):
self._test('function f() { return typeof undefined; }', 'undefined')
self._test('function f() { return typeof NaN; }', 'number')
self._test('function f() { return typeof Infinity; }', 'number')
self._test('function f() { return typeof true; }', 'boolean')
self._test('function f() { return typeof null; }', 'object')
self._test('function f() { return typeof "a string"; }', 'string')
self._test('function f() { return typeof 42; }', 'number')
self._test('function f() { return typeof 42.42; }', 'number')
self._test('function f() { var g = function(){}; return typeof g; }', 'function')
self._test('function f() { return typeof {key: "value"}; }', 'object')
# not yet implemented: Symbol, BigInt
def test_return_function(self):
jsi = JSInterpreter('''
@ -324,6 +394,16 @@ class TestJSInterpreter(unittest.TestCase):
self._test('function f() { let a; return a?.qq; }', JS_Undefined)
self._test('function f() { let a = {m1: 42, m2: 0 }; return a?.qq; }', JS_Undefined)
def test_indexing(self):
self._test('function f() { return [1, 2, 3, 4][3]}', 4)
self._test('function f() { return [1, [2, [3, [4]]]][1][1][1][0]}', 4)
self._test('function f() { var o = {1: 2, 3: 4}; return o[3]}', 4)
self._test('function f() { var o = {1: 2, 3: 4}; return o["3"]}', 4)
self._test('function f() { return [1, [2, {3: [4]}]][1][1]["3"][0]}', 4)
self._test('function f() { return [1, 2, 3, 4].length}', 4)
self._test('function f() { var o = {1: 2, 3: 4}; return o.length}', JS_Undefined)
self._test('function f() { var o = {1: 2, 3: 4}; o["length"] = 42; return o.length}', 42)
def test_regex(self):
self._test('function f() { let a=/,,[/,913,/](,)}/; }', None)
@ -411,6 +491,13 @@ class TestJSInterpreter(unittest.TestCase):
self._test(jsi, 't-e-s-t', args=[test_input, '-'])
self._test(jsi, '', args=[[], '-'])
self._test('function f(){return '
'[1, 1.0, "abc", {a: 1}, null, undefined, Infinity, NaN].join()}',
'1,1,abc,[object Object],,,Infinity,NaN')
self._test('function f(){return '
'[1, 1.0, "abc", {a: 1}, null, undefined, Infinity, NaN].join("~")}',
'1~1~abc~[object Object]~~~Infinity~NaN')
def test_split(self):
test_result = list('test')
tests = [
@ -424,6 +511,18 @@ class TestJSInterpreter(unittest.TestCase):
self._test(jsi, test_result, args=['t-e-s-t', '-'])
self._test(jsi, [''], args=['', '-'])
self._test(jsi, [], args=['', ''])
# RegExp split
self._test('function f(){return "test".split(/(?:)/)}',
['t', 'e', 's', 't'])
self._test('function f(){return "t-e-s-t".split(/[es-]+/)}',
['t', 't'])
# from MDN: surrogate pairs aren't handled: case 1 fails
# self._test('function f(){return "😄😄".split(/(?:)/)}',
# ['\ud83d', '\ude04', '\ud83d', '\ude04'])
# case 2 beats Py3.2: it gets the case 1 result
if sys.version_info >= (2, 6) and not ((3, 0) <= sys.version_info < (3, 3)):
self._test('function f(){return "😄😄".split(/(?:)/u)}',
['😄', '😄'])
def test_slice(self):
self._test('function f(){return [0, 1, 2, 3, 4, 5, 6, 7, 8].slice()}', [0, 1, 2, 3, 4, 5, 6, 7, 8])
@ -453,6 +552,40 @@ class TestJSInterpreter(unittest.TestCase):
self._test('function f(){return "012345678".slice(-1, 1)}', '')
self._test('function f(){return "012345678".slice(-3, -1)}', '67')
def test_pop(self):
# pop
self._test('function f(){var a = [0, 1, 2, 3, 4, 5, 6, 7, 8]; return [a.pop(), a]}',
[8, [0, 1, 2, 3, 4, 5, 6, 7]])
self._test('function f(){return [].pop()}', JS_Undefined)
# push
self._test('function f(){var a = [0, 1, 2]; return [a.push(3, 4), a]}',
[5, [0, 1, 2, 3, 4]])
self._test('function f(){var a = [0, 1, 2]; return [a.push(), a]}',
[3, [0, 1, 2]])
def test_shift(self):
# shift
self._test('function f(){var a = [0, 1, 2, 3, 4, 5, 6, 7, 8]; return [a.shift(), a]}',
[0, [1, 2, 3, 4, 5, 6, 7, 8]])
self._test('function f(){return [].shift()}', JS_Undefined)
# unshift
self._test('function f(){var a = [0, 1, 2]; return [a.unshift(3, 4), a]}',
[5, [3, 4, 0, 1, 2]])
self._test('function f(){var a = [0, 1, 2]; return [a.unshift(), a]}',
[3, [0, 1, 2]])
def test_forEach(self):
self._test('function f(){var ret = []; var l = [4, 2]; '
'var log = function(e,i,a){ret.push([e,i,a]);}; '
'l.forEach(log); '
'return [ret.length, ret[0][0], ret[1][1], ret[0][2]]}',
[2, 4, 1, [4, 2]])
self._test('function f(){var ret = []; var l = [4, 2]; '
'var log = function(e,i,a){this.push([e,i,a]);}; '
'l.forEach(log, ret); '
'return [ret.length, ret[0][0], ret[1][1], ret[0][2]]}',
[2, 4, 1, [4, 2]])
if __name__ == '__main__':
unittest.main()

View File

@ -1,4 +1,5 @@
#!/usr/bin/env python
# coding: utf-8
from __future__ import unicode_literals
@ -12,6 +13,7 @@ import re
import string
from youtube_dl.compat import (
compat_contextlib_suppress,
compat_open as open,
compat_str,
compat_urlretrieve,
@ -50,23 +52,38 @@ _SIG_TESTS = [
(
'https://s.ytimg.com/yts/jsbin/html5player-en_US-vflBb0OQx.js',
84,
'123456789abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQ0STUVWXYZ!"#$%&\'()*+,@./:;<=>'
'123456789abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQ0STUVWXYZ!"#$%&\'()*+,@./:;<=>',
),
(
'https://s.ytimg.com/yts/jsbin/html5player-en_US-vfl9FYC6l.js',
83,
'123456789abcdefghijklmnopqr0tuvwxyzABCDETGHIJKLMNOPQRS>UVWXYZ!"#$%&\'()*+,-./:;<=F'
'123456789abcdefghijklmnopqr0tuvwxyzABCDETGHIJKLMNOPQRS>UVWXYZ!"#$%&\'()*+,-./:;<=F',
),
(
'https://s.ytimg.com/yts/jsbin/html5player-en_US-vflCGk6yw/html5player.js',
'4646B5181C6C3020DF1D9C7FCFEA.AD80ABF70C39BD369CCCAE780AFBB98FA6B6CB42766249D9488C288',
'82C8849D94266724DC6B6AF89BBFA087EACCD963.B93C07FBA084ACAEFCF7C9D1FD0203C6C1815B6B'
'82C8849D94266724DC6B6AF89BBFA087EACCD963.B93C07FBA084ACAEFCF7C9D1FD0203C6C1815B6B',
),
(
'https://s.ytimg.com/yts/jsbin/html5player-en_US-vflKjOTVq/html5player.js',
'312AA52209E3623129A412D56A40F11CB0AF14AE.3EE09501CB14E3BCDC3B2AE808BF3F1D14E7FBF12',
'112AA5220913623229A412D56A40F11CB0AF14AE.3EE0950FCB14EEBCDC3B2AE808BF331D14E7FBF3',
)
),
(
'https://www.youtube.com/s/player/6ed0d907/player_ias.vflset/en_US/base.js',
'2aq0aqSyOoJXtK73m-uME_jv7-pT15gOFC02RFkGMqWpzEICs69VdbwQ0LDp1v7j8xx92efCJlYFYb1sUkkBSPOlPmXgIARw8JQ0qOAOAA',
'AOq0QJ8wRAIgXmPlOPSBkkUs1bYFYlJCfe29xx8j7v1pDL2QwbdV96sCIEzpWqMGkFR20CFOg51Tp-7vj_EMu-m37KtXJoOySqa0',
),
(
'https://www.youtube.com/s/player/3bb1f723/player_ias.vflset/en_US/base.js',
'2aq0aqSyOoJXtK73m-uME_jv7-pT15gOFC02RFkGMqWpzEICs69VdbwQ0LDp1v7j8xx92efCJlYFYb1sUkkBSPOlPmXgIARw8JQ0qOAOAA',
'MyOSJXtKI3m-uME_jv7-pT12gOFC02RFkGoqWpzE0Cs69VdbwQ0LDp1v7j8xx92efCJlYFYb1sUkkBSPOlPmXgIARw8JQ0qOAOAA',
),
(
'https://www.youtube.com/s/player/2f1832d2/player_ias.vflset/en_US/base.js',
'2aq0aqSyOoJXtK73m-uME_jv7-pT15gOFC02RFkGMqWpzEICs69VdbwQ0LDp1v7j8xx92efCJlYFYb1sUkkBSPOlPmXgIARw8JQ0qOAOAA',
'0QJ8wRAIgXmPlOPSBkkUs1bYFYlJCfe29xxAj7v1pDL0QwbdV96sCIEzpWqMGkFR20CFOg51Tp-7vj_EMu-m37KtXJ2OySqa0q',
),
]
_NSIG_TESTS = [
@ -142,6 +159,10 @@ _NSIG_TESTS = [
'https://www.youtube.com/s/player/5a3b6271/player_ias.vflset/en_US/base.js',
'B2j7f_UPT4rfje85Lu_e', 'm5DmNymaGQ5RdQ',
),
(
'https://www.youtube.com/s/player/7a062b77/player_ias.vflset/en_US/base.js',
'NRcE3y3mVtm_cV-W', 'VbsCYUATvqlt5w',
),
(
'https://www.youtube.com/s/player/dac945fd/player_ias.vflset/en_US/base.js',
'o8BkRxXhuYsBCWi6RplPdP', '3Lx32v_hmzTm6A',
@ -154,6 +175,10 @@ _NSIG_TESTS = [
'https://www.youtube.com/s/player/cfa9e7cb/player_ias.vflset/en_US/base.js',
'qO0NiMtYQ7TeJnfFG2', 'k9cuJDHNS5O7kQ',
),
(
'https://www.youtube.com/s/player/8c7583ff/player_ias.vflset/en_US/base.js',
'1wWCVpRR96eAmMI87L', 'KSkWAVv1ZQxC3A',
),
(
'https://www.youtube.com/s/player/b7910ca8/player_ias.vflset/en_US/base.js',
'_hXMCwMt9qE310D', 'LoZMgkkofRMCZQ',
@ -182,6 +207,18 @@ _NSIG_TESTS = [
'https://www.youtube.com/s/player/b12cc44b/player_ias.vflset/en_US/base.js',
'keLa5R2U00sR9SQK', 'N1OGyujjEwMnLw',
),
(
'https://www.youtube.com/s/player/3bb1f723/player_ias.vflset/en_US/base.js',
'gK15nzVyaXE9RsMP3z', 'ZFFWFLPWx9DEgQ',
),
(
'https://www.youtube.com/s/player/f8f53e1a/player_ias.vflset/en_US/base.js',
'VTQOUOv0mCIeJ7i8kZB', 'kcfD8wy0sNLyNQ',
),
(
'https://www.youtube.com/s/player/2f1832d2/player_ias.vflset/en_US/base.js',
'YWt1qdbe8SAfkoPHW5d', 'RrRjWQOJmBiP',
),
]
@ -216,11 +253,9 @@ class TestSignature(unittest.TestCase):
os.mkdir(self.TESTDATA_DIR)
def tearDown(self):
try:
with compat_contextlib_suppress(OSError):
for f in os.listdir(self.TESTDATA_DIR):
os.remove(f)
except OSError:
pass
def t_factory(name, sig_func, url_pattern):
@ -254,11 +289,12 @@ def signature(jscode, sig_input):
def n_sig(jscode, sig_input):
funcname = YoutubeIE(FakeYDL())._extract_n_function_name(jscode)
return JSInterpreter(jscode).call_function(funcname, sig_input)
return JSInterpreter(jscode).call_function(
funcname, sig_input, _ytdl_do_not_return=sig_input)
make_sig_test = t_factory(
'signature', signature, re.compile(r'.*-(?P<id>[a-zA-Z0-9_-]+)(?:/watch_as3|/html5player)?\.[a-z]+$'))
'signature', signature, re.compile(r'.*(?:-|/player/)(?P<id>[a-zA-Z0-9_-]+)(?:/.+\.js|(?:/watch_as3|/html5player)?\.[a-z]+)$'))
for test_spec in _SIG_TESTS:
make_sig_test(*test_spec)

View File

@ -4,208 +4,272 @@ from __future__ import unicode_literals
import re
from .common import InfoExtractor
from ..compat import (
compat_str,
compat_urlparse,
)
from ..compat import compat_str
from ..utils import (
ExtractorError,
GeoRestrictedError,
int_or_none,
qualities,
merge_dicts,
parse_iso8601,
parse_qs,
strip_or_none,
try_get,
unified_strdate,
traverse_obj,
url_or_none,
urljoin,
)
class ArteTVBaseIE(InfoExtractor):
_ARTE_LANGUAGES = 'fr|de|en|es|it|pl'
_API_BASE = 'https://api.arte.tv/api/player/v1'
_API_BASE = 'https://api.arte.tv/api/player/v2'
# yt-dlp shims
@classmethod
def _match_valid_url(cls, url):
return re.match(cls._VALID_URL, url)
def _extract_m3u8_formats_and_subtitles(self, *args, **kwargs):
return self._extract_m3u8_formats(*args, **kwargs), {}
class ArteTVIE(ArteTVBaseIE):
_VALID_URL = r'''(?x)
https?://
(?:https?://
(?:
(?:www\.)?arte\.tv/(?P<lang>%(langs)s)/videos|
api\.arte\.tv/api/player/v\d+/config/(?P<lang_2>%(langs)s)
)
/(?P<id>\d{6}-\d{3}-[AF])
|arte://program)
/(?P<id>\d{6}-\d{3}-[AF]|LIVE)
''' % {'langs': ArteTVBaseIE._ARTE_LANGUAGES}
_TESTS = [{
'url': 'https://www.arte.tv/en/videos/088501-000-A/mexico-stealing-petrol-to-survive/',
'info_dict': {
'id': '088501-000-A',
'ext': 'mp4',
'title': 'Mexico: Stealing Petrol to Survive',
'upload_date': '20190628',
},
'only_matching': True,
}, {
'url': 'https://www.arte.tv/pl/videos/100103-000-A/usa-dyskryminacja-na-porodowce/',
'only_matching': True,
'info_dict': {
'id': '100103-000-A',
'title': 'USA: Dyskryminacja na porodówce',
'description': 'md5:242017b7cce59ffae340a54baefcafb1',
'alt_title': 'ARTE Reportage',
'timestamp': 1604417980,
'upload_date': '20201103',
'duration': 554,
# test format sort
'height': 720,
'thumbnail': r're:https://api-cdn\.arte\.tv/.+940x530',
'ext': 'mp4',
},
'params': {
'format': 'bestvideo',
'skip_download': 'm3u8',
},
}, {
'note': 'No alt_title',
'url': 'https://www.arte.tv/fr/videos/110371-000-A/la-chaleur-supplice-des-arbres-de-rue/',
'info_dict': {
'id': '110371-000-A',
'ext': 'mp4',
'upload_date': '20220718',
'duration': 154,
'timestamp': 1658162460,
'description': 'md5:5890f36fe7dccfadb8b7c0891de54786',
'title': 'La chaleur, supplice des arbres de rue',
'thumbnail': 'https://api-cdn.arte.tv/img/v2/image/CPE2sQDtD8GLQgt8DuYHLf/940x530',
},
'params': {
'format': 'bestvideo',
'skip_download': 'm3u8',
},
}, {
'url': 'https://api.arte.tv/api/player/v2/config/de/100605-013-A',
'only_matching': True,
}, {
'url': 'https://api.arte.tv/api/player/v2/config/de/LIVE',
'only_matching': True,
}]
def _real_extract(self, url):
mobj = re.match(self._VALID_URL, url)
video_id = mobj.group('id')
lang = mobj.group('lang') or mobj.group('lang_2')
_GEO_BYPASS = True
info = self._download_json(
'%s/config/%s/%s' % (self._API_BASE, lang, video_id), video_id)
player_info = info['videoJsonPlayer']
vsr = try_get(player_info, lambda x: x['VSR'], dict)
if not vsr:
error = None
if try_get(player_info, lambda x: x['custom_msg']['type']) == 'error':
error = try_get(
player_info, lambda x: x['custom_msg']['msg'], compat_str)
if not error:
error = 'Video %s is not available' % player_info.get('VID') or video_id
raise ExtractorError(error, expected=True)
upload_date_str = player_info.get('shootingDate')
if not upload_date_str:
upload_date_str = (player_info.get('VRA') or player_info.get('VDA') or '').split(' ')[0]
title = (player_info.get('VTI') or player_info['VID']).strip()
subtitle = player_info.get('VSU', '').strip()
if subtitle:
title += ' - %s' % subtitle
qfunc = qualities(['MQ', 'HQ', 'EQ', 'SQ'])
LANGS = {
'fr': 'F',
_LANG_MAP = { # ISO639 -> French abbreviations
'r': 'F',
'de': 'A',
'en': 'E[ANG]',
'es': 'E[ESP]',
'it': 'E[ITA]',
'pl': 'E[POL]',
# XXX: probably means mixed; <https://www.arte.tv/en/videos/107710-029-A/dispatches-from-ukraine-local-journalists-report/>
# uses this code for audio that happens to be in Ukrainian, but the manifest uses the ISO code 'mul' (mixed)
'mul': 'EU',
}
langcode = LANGS.get(lang, lang)
_VERSION_CODE_RE = re.compile(r'''(?x)
V
(?P<original_voice>O?)
(?P<vlang>[FA]|E\[[A-Z]+\]|EU)?
(?P<audio_desc>AUD|)
(?:
(?P<has_sub>-ST)
(?P<sdh_sub>M?)
(?P<sub_lang>[FA]|E\[[A-Z]+\]|EU)
)?
''')
formats = []
for format_id, format_dict in vsr.items():
f = dict(format_dict)
format_url = url_or_none(f.get('url'))
streamer = f.get('streamer')
if not format_url and not streamer:
continue
versionCode = f.get('versionCode')
l = re.escape(langcode)
# all obtained by exhaustive testing
_COUNTRIES_MAP = {
'DE_FR': (
'BL', 'DE', 'FR', 'GF', 'GP', 'MF', 'MQ', 'NC',
'PF', 'PM', 'RE', 'WF', 'YT',
),
# with both of the below 'BE' sometimes works, sometimes doesn't
'EUR_DE_FR': (
'AT', 'BL', 'CH', 'DE', 'FR', 'GF', 'GP', 'LI',
'MC', 'MF', 'MQ', 'NC', 'PF', 'PM', 'RE', 'WF',
'YT',
),
'SAT': (
'AD', 'AT', 'AX', 'BG', 'BL', 'CH', 'CY', 'CZ',
'DE', 'DK', 'EE', 'ES', 'FI', 'FR', 'GB', 'GF',
'GR', 'HR', 'HU', 'IE', 'IS', 'IT', 'KN', 'LI',
'LT', 'LU', 'LV', 'MC', 'MF', 'MQ', 'MT', 'NC',
'NL', 'NO', 'PF', 'PL', 'PM', 'PT', 'RE', 'RO',
'SE', 'SI', 'SK', 'SM', 'VA', 'WF', 'YT',
),
}
# Language preference from most to least priority
# Reference: section 6.8 of
# https://www.arte.tv/sites/en/corporate/files/complete-technical-guidelines-arte-geie-v1-07-1.pdf
PREFERENCES = (
# original version in requested language, without subtitles
r'VO{0}$'.format(l),
# original version in requested language, with partial subtitles in requested language
r'VO{0}-ST{0}$'.format(l),
# original version in requested language, with subtitles for the deaf and hard-of-hearing in requested language
r'VO{0}-STM{0}$'.format(l),
# non-original (dubbed) version in requested language, without subtitles
r'V{0}$'.format(l),
# non-original (dubbed) version in requested language, with subtitles partial subtitles in requested language
r'V{0}-ST{0}$'.format(l),
# non-original (dubbed) version in requested language, with subtitles for the deaf and hard-of-hearing in requested language
r'V{0}-STM{0}$'.format(l),
# original version in requested language, with partial subtitles in different language
r'VO{0}-ST(?!{0}).+?$'.format(l),
# original version in requested language, with subtitles for the deaf and hard-of-hearing in different language
r'VO{0}-STM(?!{0}).+?$'.format(l),
# original version in different language, with partial subtitles in requested language
r'VO(?:(?!{0}).+?)?-ST{0}$'.format(l),
# original version in different language, with subtitles for the deaf and hard-of-hearing in requested language
r'VO(?:(?!{0}).+?)?-STM{0}$'.format(l),
# original version in different language, without subtitles
r'VO(?:(?!{0}))?$'.format(l),
# original version in different language, with partial subtitles in different language
r'VO(?:(?!{0}).+?)?-ST(?!{0}).+?$'.format(l),
# original version in different language, with subtitles for the deaf and hard-of-hearing in different language
r'VO(?:(?!{0}).+?)?-STM(?!{0}).+?$'.format(l),
)
def _real_extract(self, url):
mobj = self._match_valid_url(url)
video_id = mobj.group('id')
lang = mobj.group('lang') or mobj.group('lang_2')
language_code = self._LANG_MAP.get(lang)
config = self._download_json('{0}/config/{1}/{2}'.format(self._API_BASE, lang, video_id), video_id)
geoblocking = traverse_obj(config, ('data', 'attributes', 'restriction', 'geoblocking')) or {}
if geoblocking.get('restrictedArea'):
raise GeoRestrictedError('Video restricted to {0!r}'.format(geoblocking['code']),
countries=self._COUNTRIES_MAP.get(geoblocking['code'], ('DE', 'FR')))
if not traverse_obj(config, ('data', 'attributes', 'rights')):
# Eg: https://www.arte.tv/de/videos/097407-215-A/28-minuten
# Eg: https://www.arte.tv/es/videos/104351-002-A/serviteur-du-peuple-1-23
raise ExtractorError(
'Video is not available in this language edition of Arte or broadcast rights expired', expected=True)
formats, subtitles = [], {}
secondary_formats = []
for stream in config['data']['attributes']['streams']:
# official player contains code like `e.get("versions")[0].eStat.ml5`
stream_version = stream['versions'][0]
stream_version_code = stream_version['eStat']['ml5']
for pref, p in enumerate(PREFERENCES):
if re.match(p, versionCode):
lang_pref = len(PREFERENCES) - pref
break
else:
lang_pref = -1
m = self._VERSION_CODE_RE.match(stream_version_code)
if m:
lang_pref = int(''.join('01'[x] for x in (
m.group('vlang') == language_code, # we prefer voice in the requested language
not m.group('audio_desc'), # and not the audio description version
bool(m.group('original_voice')), # but if voice is not in the requested language, at least choose the original voice
m.group('sub_lang') == language_code, # if subtitles are present, we prefer them in the requested language
not m.group('has_sub'), # but we prefer no subtitles otherwise
not m.group('sdh_sub'), # and we prefer not the hard-of-hearing subtitles if there are subtitles
)))
media_type = f.get('mediaType')
if media_type == 'hls':
m3u8_formats = self._extract_m3u8_formats(
format_url, video_id, 'mp4', entry_protocol='m3u8_native',
m3u8_id=format_id, fatal=False)
for m3u8_format in m3u8_formats:
m3u8_format['language_preference'] = lang_pref
formats.extend(m3u8_formats)
continue
format = {
'format_id': format_id,
'preference': -10 if f.get('videoFormat') == 'M3U8' else None,
short_label = traverse_obj(stream_version, 'shortLabel', expected_type=compat_str, default='?')
if stream['protocol'].startswith('HLS'):
fmts, subs = self._extract_m3u8_formats_and_subtitles(
stream['url'], video_id=video_id, ext='mp4', m3u8_id=stream_version_code, fatal=False)
for fmt in fmts:
fmt.update({
'format_note': '{0} [{1}]'.format(stream_version.get("label", "unknown"), short_label),
'language_preference': lang_pref,
'format_note': '%s, %s' % (f.get('versionCode'), f.get('versionLibelle')),
'width': int_or_none(f.get('width')),
'height': int_or_none(f.get('height')),
'tbr': int_or_none(f.get('bitrate')),
'quality': qfunc(f.get('quality')),
}
if media_type == 'rtmp':
format['url'] = f['streamer']
format['play_path'] = 'mp4:' + f['url']
format['ext'] = 'flv'
})
if any(map(short_label.startswith, ('cc', 'OGsub'))):
secondary_formats.extend(fmts)
else:
format['url'] = f['url']
formats.extend(fmts)
for sub in subs:
subtitles = self._merge_subtitles(subtitles, sub)
formats.append(format)
elif stream['protocol'] in ('HTTPS', 'RTMP'):
formats.append({
'format_id': '{0}-{1}'.format(stream["protocol"], stream_version_code),
'url': stream['url'],
'format_note': '{0} [{1}]'.format(stream_version.get("label", "unknown"), short_label),
'language_preference': lang_pref,
# 'ext': 'mp4', # XXX: may or may not be necessary, at least for HTTPS
})
else:
self.report_warning('Skipping stream with unknown protocol {0}'.format(stream["protocol"]))
# TODO: chapters from stream['segments']?
# The JS also looks for chapters in config['data']['attributes']['chapters'],
# but I am yet to find a video having those
formats.extend(secondary_formats)
self._remove_duplicate_formats(formats)
self._sort_formats(formats)
metadata = config['data']['attributes']['metadata']
return {
'id': player_info.get('VID') or video_id,
'title': title,
'description': player_info.get('VDE'),
'upload_date': unified_strdate(upload_date_str),
'thumbnail': player_info.get('programImage') or player_info.get('VTU', {}).get('IUR'),
'id': metadata['providerId'],
'webpage_url': traverse_obj(metadata, ('link', 'url')),
'title': traverse_obj(metadata, 'subtitle', 'title'),
'alt_title': metadata.get('subtitle') and metadata.get('title'),
'description': metadata.get('description'),
'duration': traverse_obj(metadata, ('duration', 'seconds')),
'language': metadata.get('language'),
'timestamp': traverse_obj(config, ('data', 'attributes', 'rights', 'begin'), expected_type=parse_iso8601),
'is_live': config['data']['attributes'].get('live', False),
'formats': formats,
'subtitles': subtitles,
'thumbnails': [
{'url': image['url'], 'id': image.get('caption')}
for image in metadata.get('images') or [] if url_or_none(image.get('url'))
],
}
class ArteTVEmbedIE(InfoExtractor):
_VALID_URL = r'https?://(?:www\.)?arte\.tv/player/v\d+/index\.php\?.*?\bjson_url=.+'
_EMBED_REGEX = [r'<(?:iframe|script)[^>]+src=(["\'])(?P<url>(?:https?:)?//(?:www\.)?arte\.tv/player/v\d+/index\.php\?.*?\bjson_url=.+?)\1']
_TESTS = [{
'url': 'https://www.arte.tv/player/v5/index.php?json_url=https%3A%2F%2Fapi.arte.tv%2Fapi%2Fplayer%2Fv2%2Fconfig%2Fde%2F100605-013-A&lang=de&autoplay=true&mute=0100605-013-A',
'only_matching': True,
'skip': 'Video is not available in this language edition of Arte or broadcast rights expired'
}, {
'url': 'https://www.arte.tv/player/v5/index.php?json_url=https%3A%2F%2Fapi.arte.tv%2Fapi%2Fplayer%2Fv2%2Fconfig%2Fpl%2F100103-000-A&lang=pl&autoplay=true&mute=100103-000-A',
'info_dict': {
'id': '100605-013-A',
'id': '100103-000-A',
'ext': 'mp4',
'title': 'United we Stream November Lockdown Edition #13',
'description': 'md5:be40b667f45189632b78c1425c7c2ce1',
'upload_date': '20201116',
'title': 'USA: Dyskryminacja na porodówce',
'timestamp': 1604417980,
'upload_date': '20201103',
'description': 'md5:242017b7cce59ffae340a54baefcafb1',
'duration': 554,
},
'params': {
'format': 'bestvideo',
'skip_download': 'm3u8',
},
}, {
'url': 'https://www.arte.tv/player/v3/index.php?json_url=https://api.arte.tv/api/player/v2/config/de/100605-013-A',
'only_matching': True,
}]
@staticmethod
def _extract_urls(webpage):
return [url for _, url in re.findall(
r'<(?:iframe|script)[^>]+src=(["\'])(?P<url>(?:https?:)?//(?:www\.)?arte\.tv/player/v\d+/index\.php\?.*?\bjson_url=.+?)\1',
webpage)]
@classmethod
def _extract_urls(cls, webpage):
import itertools # just until this is lifted into IE
return list(itertools.chain(*(
(url for _, url in re.findall(erx, webpage)) for erx in cls._EMBED_REGEX)
))
def _real_extract(self, url):
qs = compat_urlparse.parse_qs(compat_urlparse.urlparse(url).query)
json_url = qs['json_url'][0]
qs = parse_qs(url)
json_url = qs['json_url'][-1]
video_id = ArteTVIE._match_id(json_url)
return self.url_result(
json_url, ie=ArteTVIE.ie_key(), video_id=video_id)
@ -215,44 +279,45 @@ class ArteTVPlaylistIE(ArteTVBaseIE):
_VALID_URL = r'https?://(?:www\.)?arte\.tv/(?P<lang>%s)/videos/(?P<id>RC-\d{6})' % ArteTVBaseIE._ARTE_LANGUAGES
_TESTS = [{
'url': 'https://www.arte.tv/en/videos/RC-016954/earn-a-living/',
'info_dict': {
'id': 'RC-016954',
'title': 'Earn a Living',
'description': 'md5:d322c55011514b3a7241f7fb80d494c2',
},
'playlist_mincount': 6,
'only_matching': True,
}, {
'url': 'https://www.arte.tv/pl/videos/RC-014123/arte-reportage/',
'only_matching': True,
'playlist_mincount': 100,
'info_dict': {
'description': 'md5:84e7bf1feda248bc325ebfac818c476e',
'id': 'RC-014123',
'title': 'ARTE Reportage - najlepsze reportaże',
},
'skip': '404 Not Found',
}, {
'url': 'https://www.arte.tv/en/videos/RC-016979/war-in-ukraine/',
'playlist_mincount': 79,
'info_dict': {
'id': 'RC-016979',
'title': 'War in Ukraine',
'description': 'On 24 February, Russian armed forces invaded Ukraine. We follow the war day by day and provide background information with special insights, reports and documentaries.',
},
}]
def _real_extract(self, url):
lang, playlist_id = re.match(self._VALID_URL, url).groups()
collection = self._download_json(
'%s/collectionData/%s/%s?source=videos'
% (self._API_BASE, lang, playlist_id), playlist_id)
entries = []
for video in collection['videos']:
if not isinstance(video, dict):
continue
video_url = url_or_none(video.get('url')) or url_or_none(video.get('jsonUrl'))
if not video_url:
continue
video_id = video.get('programId')
entries.append({
lang, playlist_id = self._match_valid_url(url).group('lang', 'id')
playlist = self._download_json(
'{0}/playlist/{1}/{2}'.format(self._API_BASE, lang, playlist_id), playlist_id)['data']['attributes']
entries = [{
'_type': 'url_transparent',
'url': video_url,
'id': video_id,
'url': video['config']['url'],
'ie_key': ArteTVIE.ie_key(),
'id': video.get('providerId'),
'title': video.get('title'),
'alt_title': video.get('subtitle'),
'thumbnail': url_or_none(try_get(video, lambda x: x['mainImage']['url'], compat_str)),
'duration': int_or_none(video.get('durationSeconds')),
'view_count': int_or_none(video.get('views')),
'ie_key': ArteTVIE.ie_key(),
})
title = collection.get('title')
description = collection.get('shortDescription') or collection.get('teaserText')
return self.playlist_result(entries, playlist_id, title, description)
'thumbnail': url_or_none(traverse_obj(video, ('mainImage', 'url'))),
'duration': int_or_none(traverse_obj(video, ('duration', 'seconds'))),
} for video in traverse_obj(playlist, ('items', lambda _, v: v['config']['url']))]
return self.playlist_result(entries, playlist_id,
traverse_obj(playlist, ('metadata', 'title')),
traverse_obj(playlist, ('metadata', 'description')))
class ArteTVCategoryIE(ArteTVBaseIE):
@ -262,11 +327,10 @@ class ArteTVCategoryIE(ArteTVBaseIE):
'info_dict': {
'id': 'politics-and-society',
'title': 'Politics and society',
'description': 'Investigative documentary series, geopolitical analysis, and international commentary',
'description': 'Watch documentaries and reportage about politics, society and current affairs.',
},
'playlist_mincount': 13,
},
]
}]
@classmethod
def suitable(cls, url):
@ -275,27 +339,23 @@ class ArteTVCategoryIE(ArteTVBaseIE):
and super(ArteTVCategoryIE, cls).suitable(url))
def _real_extract(self, url):
lang, playlist_id = re.match(self._VALID_URL, url).groups()
lang, playlist_id = self._match_valid_url(url).groups()
webpage = self._download_webpage(url, playlist_id)
items = []
for video in re.finditer(
r'<a\b[^>]*?href\s*=\s*(?P<q>"|\'|\b)(?P<url>https?://www\.arte\.tv/%s/videos/[\w/-]+)(?P=q)' % lang,
r'<a\b[^>]+\bhref\s*=\s*(?P<q>"|\'|\b)(?P<url>(?:https?://www\.arte\.tv)?/%s/videos/[\w/-]+)(?P=q)' % lang,
webpage):
video = video.group('url')
video = urljoin(url, video.group('url'))
if video == url:
continue
if any(ie.suitable(video) for ie in (ArteTVIE, ArteTVPlaylistIE, )):
items.append(video)
if items:
title = (self._og_search_title(webpage, default=None)
or self._html_search_regex(r'<title\b[^>]*>([^<]+)</title>', default=None))
title = strip_or_none(title.rsplit('|', 1)[0]) or self._generic_title(url)
result = self.playlist_from_matches(items, playlist_id=playlist_id, playlist_title=title)
if result:
description = self._og_search_description(webpage, default=None)
if description:
result['description'] = description
return result
return merge_dicts(
self.playlist_from_matches(items, playlist_id=playlist_id, playlist_title=title),
{'description': self._og_search_description(webpage, default=None)})

View File

@ -3170,7 +3170,7 @@ class InfoExtractor(object):
# See com/longtailvideo/jwplayer/media/RTMPMediaProvider.as
# of jwplayer.flash.swf
rtmp_url_parts = re.split(
r'((?:mp4|mp3|flv):)', source_url, 1)
r'((?:mp4|mp3|flv):)', source_url, maxsplit=1)
if len(rtmp_url_parts) == 3:
rtmp_url, prefix, play_path = rtmp_url_parts
a_format.update({

View File

@ -3,11 +3,13 @@
from __future__ import unicode_literals
import collections
import hashlib
import itertools
import json
import os.path
import random
import re
import time
import traceback
from .common import InfoExtractor, SearchInfoExtractor
@ -290,6 +292,33 @@ class YoutubeBaseInfoExtractor(InfoExtractor):
_YT_INITIAL_PLAYER_RESPONSE_RE = r'ytInitialPlayerResponse\s*=\s*({.+?})\s*;'
_YT_INITIAL_BOUNDARY_RE = r'(?:var\s+meta|</script|\n)'
_SAPISID = None
def _generate_sapisidhash_header(self, origin='https://www.youtube.com'):
time_now = round(time.time())
if self._SAPISID is None:
yt_cookies = self._get_cookies('https://www.youtube.com')
# Sometimes SAPISID cookie isn't present but __Secure-3PAPISID is.
# See: https://github.com/yt-dlp/yt-dlp/issues/393
sapisid_cookie = dict_get(
yt_cookies, ('__Secure-3PAPISID', 'SAPISID'))
if sapisid_cookie and sapisid_cookie.value:
self._SAPISID = sapisid_cookie.value
self.write_debug('Extracted SAPISID cookie')
# SAPISID cookie is required if not already present
if not yt_cookies.get('SAPISID'):
self.write_debug('Copying __Secure-3PAPISID cookie to SAPISID cookie')
self._set_cookie(
'.youtube.com', 'SAPISID', self._SAPISID, secure=True, expire_time=time_now + 3600)
else:
self._SAPISID = False
if not self._SAPISID:
return None
# SAPISIDHASH algorithm from https://stackoverflow.com/a/32065323
sapisidhash = hashlib.sha1(
'{0} {1} {2}'.format(time_now, self._SAPISID, origin).encode('utf-8')).hexdigest()
return 'SAPISIDHASH {0}_{1}'.format(time_now, sapisidhash)
def _call_api(self, ep, query, video_id, fatal=True, headers=None):
data = self._DEFAULT_API_DATA.copy()
data.update(query)
@ -1579,20 +1608,27 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
self.to_screen('Extracted signature function:\n' + code)
def _parse_sig_js(self, jscode):
# Examples where `sig` is funcname:
# sig=function(a){a=a.split(""); ... ;return a.join("")};
# ;c&&(c=sig(decodeURIComponent(c)),a.set(b,encodeURIComponent(c)));return a};
# {var l=f,m=h.sp,n=sig(decodeURIComponent(h.s));l.set(m,encodeURIComponent(n))}
# sig=function(J){J=J.split(""); ... ;return J.join("")};
# ;N&&(N=sig(decodeURIComponent(N)),J.set(R,encodeURIComponent(N)));return J};
# {var H=u,k=f.sp,v=sig(decodeURIComponent(f.s));H.set(k,encodeURIComponent(v))}
funcname = self._search_regex(
(r'\b[cs]\s*&&\s*[adf]\.set\([^,]+\s*,\s*encodeURIComponent\s*\(\s*(?P<sig>[a-zA-Z0-9$]+)\(',
r'\b[a-zA-Z0-9]+\s*&&\s*[a-zA-Z0-9]+\.set\([^,]+\s*,\s*encodeURIComponent\s*\(\s*(?P<sig>[a-zA-Z0-9$]+)\(',
r'\bm=(?P<sig>[a-zA-Z0-9$]{2,})\(decodeURIComponent\(h\.s\)\)',
r'\bc&&\(c=(?P<sig>[a-zA-Z0-9$]{2,})\(decodeURIComponent\(c\)\)',
r'(?:\b|[^a-zA-Z0-9$])(?P<sig>[a-zA-Z0-9$]{2,})\s*=\s*function\(\s*a\s*\)\s*{\s*a\s*=\s*a\.split\(\s*""\s*\)(?:;[a-zA-Z0-9$]{2}\.[a-zA-Z0-9$]{2}\(a,\d+\))?',
r'(?P<sig>[a-zA-Z0-9$]+)\s*=\s*function\(\s*a\s*\)\s*{\s*a\s*=\s*a\.split\(\s*""\s*\)',
(r'\b(?P<var>[\w$]+)&&\((?P=var)=(?P<sig>[\w$]{2,})\(decodeURIComponent\((?P=var)\)\)',
r'(?P<sig>[\w$]+)\s*=\s*function\(\s*(?P<arg>[\w$]+)\s*\)\s*{\s*(?P=arg)\s*=\s*(?P=arg)\.split\(\s*""\s*\)\s*;\s*[^}]+;\s*return\s+(?P=arg)\.join\(\s*""\s*\)',
r'(?:\b|[^\w$])(?P<sig>[\w$]{2,})\s*=\s*function\(\s*a\s*\)\s*{\s*a\s*=\s*a\.split\(\s*""\s*\)(?:;[\w$]{2}\.[\w$]{2}\(a,\d+\))?',
# Old patterns
r'\b[cs]\s*&&\s*[adf]\.set\([^,]+\s*,\s*encodeURIComponent\s*\(\s*(?P<sig>[\w$]+)\(',
r'\b[\w]+\s*&&\s*[\w]+\.set\([^,]+\s*,\s*encodeURIComponent\s*\(\s*(?P<sig>[\w$]+)\(',
r'\bm=(?P<sig>[\w$]{2,})\(decodeURIComponent\(h\.s\)\)',
# Obsolete patterns
r'("|\')signature\1\s*,\s*(?P<sig>[a-zA-Z0-9$]+)\(',
r'\.sig\|\|(?P<sig>[a-zA-Z0-9$]+)\(',
r'yt\.akamaized\.net/\)\s*\|\|\s*.*?\s*[cs]\s*&&\s*[adf]\.set\([^,]+\s*,\s*(?:encodeURIComponent\s*\()?\s*(?P<sig>[a-zA-Z0-9$]+)\(',
r'\b[cs]\s*&&\s*[adf]\.set\([^,]+\s*,\s*(?P<sig>[a-zA-Z0-9$]+)\(',
r'\b[a-zA-Z0-9]+\s*&&\s*[a-zA-Z0-9]+\.set\([^,]+\s*,\s*(?P<sig>[a-zA-Z0-9$]+)\(',
r'\bc\s*&&\s*[a-zA-Z0-9]+\.set\([^,]+\s*,\s*\([^)]*\)\s*\(\s*(?P<sig>[a-zA-Z0-9$]+)\('),
r'("|\')signature\1\s*,\s*(?P<sig>[\w$]+)\(',
r'\.sig\|\|(?P<sig>[\w$]+)\(',
r'yt\.akamaized\.net/\)\s*\|\|\s*.*?\s*[cs]\s*&&\s*[adf]\.set\([^,]+\s*,\s*(?:encodeURIComponent\s*\()?\s*(?P<sig>[\w$]+)\(',
r'\b[cs]\s*&&\s*[adf]\.set\([^,]+\s*,\s*(?P<sig>[\w$]+)\(',
r'\bc\s*&&\s*[\w]+\.set\([^,]+\s*,\s*\([^)]*\)\s*\(\s*(?P<sig>[\w$]+)\('),
jscode, 'Initial JS player signature function name', group='sig')
jsi = JSInterpreter(jscode)
@ -1658,36 +1694,29 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
def _extract_n_function_name(self, jscode):
func_name, idx = self._search_regex(
# new: (b=String.fromCharCode(110),c=a.get(b))&&c=nfunc[idx](c)
# or: (b="nn"[+a.D],c=a.get(b))&&(c=nfunc[idx](c)
# or: (PL(a),b=a.j.n||null)&&(b=nfunc[idx](b)
# (y=NuD(),Mw(k),q=k.Z[y]||null)&&(q=narray[idx](q),k.set(y,q),k.V||NuD(''))}};
# (R="nn"[+J.Z],mW(J),N=J.K[R]||null)&&(N=narray[idx](N),J.set(R,N))}};
# or: (b=String.fromCharCode(110),c=a.get(b))&&c=narray[idx](c)
# or: (b="nn"[+a.D],c=a.get(b))&&(c=narray[idx](c)
# or: (PL(a),b=a.j.n||null)&&(b=narray[idx](b)
# or: (b="nn"[+a.D],vL(a),c=a.j[b]||null)&&(c=narray[idx](c),a.set(b,c),narray.length||nfunc("")
# old: (b=a.get("n"))&&(b=nfunc[idx](b)(?P<c>[a-z])\s*=\s*[a-z]\s*
# old: (b=a.get("n"))&&(b=narray[idx](b)(?P<c>[a-z])\s*=\s*[a-z]\s*
# older: (b=a.get("n"))&&(b=nfunc(b)
r'''(?x)
\((?:[\w$()\s]+,)*?\s* # (
(?P<b>[a-z])\s*=\s* # b=
(?:
(?: # expect ,c=a.get(b) (etc)
String\s*\.\s*fromCharCode\s*\(\s*110\s*\)|
"n+"\[\s*\+?s*[\w$.]+\s*]
)\s*(?:,[\w$()\s]+(?=,))*|
(?P<old>[\w$]+) # a (old[er])
)\s*
(?(old)
# b.get("n")
(?:\.\s*[\w$]+\s*|\[\s*[\w$]+\s*]\s*)*?
(?:\.\s*n|\[\s*"n"\s*]|\.\s*get\s*\(\s*"n"\s*\))
| # ,c=a.get(b)
,\s*(?P<c>[a-z])\s*=\s*[a-z]\s*
(?:\.\s*[\w$]+\s*|\[\s*[\w$]+\s*]\s*)*?
(?:\[\s*(?P=b)\s*]|\.\s*get\s*\(\s*(?P=b)\s*\))
)
# interstitial junk
\s*(?:\|\|\s*null\s*)?(?:\)\s*)?&&\s*(?:\(\s*)?
(?(c)(?P=c)|(?P=b))\s*=\s* # [c|b]=
# nfunc|nfunc[idx]
(?P<nfunc>[a-zA-Z_$][\w$]*)(?:\s*\[(?P<idx>\d+)\])?\s*\(\s*[\w$]+\s*\)
# (expr, ...,
\((?:(?:\s*[\w$]+\s*=)?(?:[\w$"+\.\s(\[]+(?:[)\]]\s*)?),)*
# b=...
(?P<b>[\w$]+)\s*=\s*(?!(?P=b)[^\w$])[\w$]+\s*(?:(?:
\.\s*[\w$]+ |
\[\s*[\w$]+\s*\] |
\.\s*get\s*\(\s*[\w$"]+\s*\)
)\s*){,2}(?:\s*\|\|\s*null(?=\s*\)))?\s*
\)\s*&&\s*\( # ...)&&(
# b = nfunc, b = narray[idx]
(?P=b)\s*=\s*(?P<nfunc>[\w$]+)\s*
(?:\[\s*(?P<idx>[\w$]+)\s*\]\s*)?
# (...)
\(\s*[\w$]+\s*\)
''', jscode, 'Initial JS player n function name', group=('nfunc', 'idx'),
default=(None, None))
# thx bashonly: yt-dlp/yt-dlp/pull/10611
@ -1697,15 +1726,19 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
r'''(?xs)
(?:(?<=[^\w$])|^) # instead of \b, which ignores $
(?P<name>(?!\d)[a-zA-Z\d_$]+)\s*=\s*function\((?!\d)[a-zA-Z\d_$]+\)
\s*\{(?:(?!};).)+?["']enhanced_except_
\s*\{(?:(?!};).)+?(?:
["']enhanced_except_ |
return\s*(?P<q>"|')[a-zA-Z\d-]+_w8_(?P=q)\s*\+\s*[\w$]+
)
''', jscode, 'Initial JS player n function name', group='name')
if not idx:
return func_name
return self._parse_json(self._search_regex(
r'var\s+{0}\s*=\s*(\[.+?\])\s*[,;]'.format(re.escape(func_name)), jscode,
'Initial JS player n function list ({0}.{1})'.format(func_name, idx)),
func_name, transform_source=js_to_json)[int(idx)]
return self._search_json(
r'var\s+{0}\s*='.format(re.escape(func_name)), jscode,
'Initial JS player n function list ({0}.{1})'.format(func_name, idx),
func_name, contains_pattern=r'\[[\s\S]+\]', end_pattern='[,;]',
transform_source=js_to_json)[int(idx)]
def _extract_n_function_code(self, video_id, player_url):
player_id = self._extract_player_info(player_url)
@ -1728,13 +1761,13 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
def extract_nsig(s):
try:
ret = func([s])
ret = func([s], kwargs={'_ytdl_do_not_return': s})
except JSInterpreter.Exception:
raise
except Exception as e:
raise JSInterpreter.Exception(traceback.format_exc(), cause=e)
if ret.startswith('enhanced_except_'):
if ret.startswith('enhanced_except_') or ret.endswith(s):
raise JSInterpreter.Exception('Signature function returned an exception')
return ret
@ -1910,9 +1943,50 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
player_response = self._extract_yt_initial_variable(
webpage, self._YT_INITIAL_PLAYER_RESPONSE_RE,
video_id, 'initial player response')
if not player_response:
if False and not player_response:
player_response = self._call_api(
'player', {'videoId': video_id}, video_id)
if True or not player_response:
origin = 'https://www.youtube.com'
pb_context = {'html5Preference': 'HTML5_PREF_WANTS'}
player_url = self._extract_player_url(webpage)
ytcfg = self._extract_ytcfg(video_id, webpage)
sts = self._extract_signature_timestamp(video_id, player_url, ytcfg)
if sts:
pb_context['signatureTimestamp'] = sts
query = {
'playbackContext': {
'contentPlaybackContext': pb_context,
'contentCheckOk': True,
'racyCheckOk': True,
},
'context': {
'client': {
'clientName': 'MWEB',
'clientVersion': '2.20241202.07.00',
'hl': 'en',
'userAgent': 'Mozilla/5.0 (iPad; CPU OS 16_7_10 like Mac OS X) AppleWebKit/605.1.15 (KHTML, like Gecko) Version/16.6 Mobile/15E148 Safari/604.1,gzip(gfe)',
'timeZone': 'UTC',
'utcOffsetMinutes': 0,
},
},
'videoId': video_id,
}
headers = {
'X-YouTube-Client-Name': '2',
'X-YouTube-Client-Version': '2.20241202.07.00',
'Origin': origin,
'Sec-Fetch-Mode': 'navigate',
'User-Agent': query['context']['client']['userAgent'],
}
auth = self._generate_sapisidhash_header(origin)
if auth is not None:
headers['Authorization'] = auth
headers['X-Origin'] = origin
player_response = self._call_api('player', query, video_id, fatal=False, headers=headers)
def is_agegated(playability):
if not isinstance(playability, dict):
@ -2219,12 +2293,12 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
formats.append(f)
playable_formats = [f for f in formats if not f.get('has_drm')]
if formats and not playable_formats:
if formats:
if not playable_formats:
# If there are no formats that definitely don't have DRM, all have DRM
self.report_drm(video_id)
formats[:] = playable_formats
if not formats:
else:
if streaming_data.get('licenseInfos'):
raise ExtractorError(
'This video is DRM protected.', expected=True)

View File

@ -1,3 +1,4 @@
# coding: utf-8
from __future__ import unicode_literals
import itertools
@ -5,11 +6,12 @@ import json
import operator
import re
from functools import update_wrapper
from functools import update_wrapper, wraps
from .utils import (
error_to_compat_str,
ExtractorError,
float_or_none,
js_to_json,
remove_quotes,
unified_timestamp,
@ -20,9 +22,11 @@ from .compat import (
compat_basestring,
compat_chr,
compat_collections_chain_map as ChainMap,
compat_contextlib_suppress,
compat_filter as filter,
compat_itertools_zip_longest as zip_longest,
compat_map as map,
compat_numeric_types,
compat_str,
)
@ -62,6 +66,10 @@ _NaN = float('nan')
_Infinity = float('inf')
class JS_Undefined(object):
pass
def _js_bit_op(op):
def zeroise(x):
@ -74,43 +82,114 @@ def _js_bit_op(op):
return wrapped
def _js_arith_op(op):
def _js_arith_op(op, div=False):
@wraps_op(op)
def wrapped(a, b):
if JS_Undefined in (a, b):
return _NaN
return op(a or 0, b or 0)
# null, "" --> 0
a, b = (float_or_none(
(x.strip() if isinstance(x, compat_basestring) else x) or 0,
default=_NaN) for x in (a, b))
if _NaN in (a, b):
return _NaN
try:
return op(a, b)
except ZeroDivisionError:
return _NaN if not (div and (a or b)) else _Infinity
return wrapped
def _js_div(a, b):
if JS_Undefined in (a, b) or not (a or b):
return _NaN
return operator.truediv(a or 0, b) if b else _Infinity
_js_arith_add = _js_arith_op(operator.add)
def _js_mod(a, b):
if JS_Undefined in (a, b) or not b:
return _NaN
return (a or 0) % b
def _js_add(a, b):
if not (isinstance(a, compat_basestring) or isinstance(b, compat_basestring)):
return _js_arith_add(a, b)
if not isinstance(a, compat_basestring):
a = _js_toString(a)
elif not isinstance(b, compat_basestring):
b = _js_toString(b)
return operator.concat(a, b)
_js_mod = _js_arith_op(operator.mod)
__js_exp = _js_arith_op(operator.pow)
def _js_exp(a, b):
if not b:
return 1 # even 0 ** 0 !!
elif JS_Undefined in (a, b):
return _NaN
return (a or 0) ** b
return __js_exp(a, b)
def _js_eq_op(op):
def _js_to_primitive(v):
return (
','.join(map(_js_toString, v)) if isinstance(v, list)
else '[object Object]' if isinstance(v, dict)
else compat_str(v) if not isinstance(v, (
compat_numeric_types, compat_basestring))
else v
)
def _js_toString(v):
return (
'undefined' if v is JS_Undefined
else 'Infinity' if v == _Infinity
else 'NaN' if v is _NaN
else 'null' if v is None
# bool <= int: do this first
else ('false', 'true')[v] if isinstance(v, bool)
else '{0:.7f}'.format(v).rstrip('.0') if isinstance(v, compat_numeric_types)
else _js_to_primitive(v))
_nullish = frozenset((None, JS_Undefined))
def _js_eq(a, b):
# NaN != any
if _NaN in (a, b):
return False
# Object is Object
if isinstance(a, type(b)) and isinstance(b, (dict, list)):
return operator.is_(a, b)
# general case
if a == b:
return True
# null == undefined
a_b = set((a, b))
if a_b & _nullish:
return a_b <= _nullish
a, b = _js_to_primitive(a), _js_to_primitive(b)
if not isinstance(a, compat_basestring):
a, b = b, a
# Number to String: convert the string to a number
# Conversion failure results in ... false
if isinstance(a, compat_basestring):
return float_or_none(a) == b
return a == b
def _js_neq(a, b):
return not _js_eq(a, b)
def _js_id_op(op):
@wraps_op(op)
def wrapped(a, b):
if set((a, b)) <= set((None, JS_Undefined)):
return op(a, a)
if _NaN in (a, b):
return op(_NaN, None)
if not isinstance(a, (compat_basestring, compat_numeric_types)):
a, b = b, a
# strings are === if ==
# why 'a' is not 'a': https://stackoverflow.com/a/1504848
if isinstance(a, (compat_basestring, compat_numeric_types)):
return a == b if op(0, 0) else a != b
return op(a, b)
return wrapped
@ -138,25 +217,57 @@ def _js_ternary(cndn, if_true=True, if_false=False):
return if_true
def _js_unary_op(op):
@wraps_op(op)
def wrapped(_, a):
return op(a)
return wrapped
# https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Operators/typeof
def _js_typeof(expr):
with compat_contextlib_suppress(TypeError, KeyError):
return {
JS_Undefined: 'undefined',
_NaN: 'number',
_Infinity: 'number',
True: 'boolean',
False: 'boolean',
None: 'object',
}[expr]
for t, n in (
(compat_basestring, 'string'),
(compat_numeric_types, 'number'),
):
if isinstance(expr, t):
return n
if callable(expr):
return 'function'
# TODO: Symbol, BigInt
return 'object'
# (op, definition) in order of binding priority, tightest first
# avoid dict to maintain order
# definition None => Defined in JSInterpreter._operator
_OPERATORS = (
('>>', _js_bit_op(operator.rshift)),
('<<', _js_bit_op(operator.lshift)),
('+', _js_arith_op(operator.add)),
('+', _js_add),
('-', _js_arith_op(operator.sub)),
('*', _js_arith_op(operator.mul)),
('%', _js_mod),
('/', _js_div),
('/', _js_arith_op(operator.truediv, div=True)),
('**', _js_exp),
)
_COMP_OPERATORS = (
('===', operator.is_),
('!==', operator.is_not),
('==', _js_eq_op(operator.eq)),
('!=', _js_eq_op(operator.ne)),
('===', _js_id_op(operator.is_)),
('!==', _js_id_op(operator.is_not)),
('==', _js_eq),
('!=', _js_neq),
('<=', _js_comp_op(operator.le)),
('>=', _js_comp_op(operator.ge)),
('<', _js_comp_op(operator.lt)),
@ -176,6 +287,11 @@ _SC_OPERATORS = (
('&&', None),
)
_UNARY_OPERATORS_X = (
('void', _js_unary_op(lambda _: JS_Undefined)),
('typeof', _js_unary_op(_js_typeof)),
)
_OPERATOR_RE = '|'.join(map(lambda x: re.escape(x[0]), _OPERATORS + _LOG_OPERATORS))
_NAME_RE = r'[a-zA-Z_$][\w$]*'
@ -183,10 +299,6 @@ _MATCHING_PARENS = dict(zip(*zip('()', '{}', '[]')))
_QUOTES = '\'"/'
class JS_Undefined(object):
pass
class JS_Break(ExtractorError):
def __init__(self):
ExtractorError.__init__(self, 'Invalid break')
@ -242,6 +354,7 @@ class Debugger(object):
@classmethod
def wrap_interpreter(cls, f):
@wraps(f)
def interpret_statement(self, stmt, local_vars, allow_recursion, *args, **kwargs):
if cls.ENABLED and stmt.strip():
cls.write(stmt, level=allow_recursion)
@ -255,7 +368,7 @@ class Debugger(object):
raise
if cls.ENABLED and stmt.strip():
if should_ret or repr(ret) != stmt:
cls.write(['->', '=>'][should_ret], repr(ret), '<-|', stmt, level=allow_recursion)
cls.write(['->', '=>'][bool(should_ret)], repr(ret), '<-|', stmt, level=allow_recursion)
return ret, should_ret
return interpret_statement
@ -284,6 +397,9 @@ class JSInterpreter(object):
RE_FLAGS = {
# special knowledge: Python's re flags are bitmask values, current max 128
# invent new bitmask values well above that for literal parsing
# JS 'u' flag is effectively always set (surrogate pairs aren't seen),
# but \u{...} and \p{...} escapes aren't handled); no additional JS 'v'
# features are supported
# TODO: execute matches with these flags (remaining: d, y)
'd': 1024, # Generate indices for substring matches
'g': 2048, # Global search
@ -291,6 +407,7 @@ class JSInterpreter(object):
'm': re.M, # Multi-line search
's': re.S, # Allows . to match newline characters
'u': re.U, # Treat a pattern as a sequence of unicode code points
'v': re.U, # Like 'u' with extended character class and \p{} syntax
'y': 4096, # Perform a "sticky" search that matches starting at the current position in the target string
}
@ -347,6 +464,8 @@ class JSInterpreter(object):
def __op_chars(cls):
op_chars = set(';,[')
for op in cls._all_operators():
if op[0].isalpha():
continue
op_chars.update(op[0])
return op_chars
@ -369,9 +488,18 @@ class JSInterpreter(object):
skipping = 0
if skip_delims:
skip_delims = variadic(skip_delims)
skip_txt = None
for idx, char in enumerate(expr):
if skip_txt and idx <= skip_txt[1]:
continue
paren_delta = 0
if not in_quote:
if char == '/' and expr[idx:idx + 2] == '/*':
# skip a comment
skip_txt = expr[idx:].find('*/', 2)
skip_txt = [idx, idx + skip_txt + 1] if skip_txt >= 2 else None
if skip_txt:
continue
if char in _MATCHING_PARENS:
counters[_MATCHING_PARENS[char]] += 1
paren_delta = 1
@ -404,11 +532,18 @@ class JSInterpreter(object):
if pos < delim_len:
pos += 1
continue
if skip_txt and skip_txt[0] >= start and skip_txt[1] <= idx - delim_len:
yield expr[start:skip_txt[0]] + expr[skip_txt[1] + 1: idx - delim_len]
else:
yield expr[start: idx - delim_len]
skip_txt = None
start, pos = idx + 1, 0
splits += 1
if max_split and splits >= max_split:
break
if skip_txt and skip_txt[0] >= start:
yield expr[start:skip_txt[0]] + expr[skip_txt[1] + 1:]
else:
yield expr[start:]
@classmethod
@ -425,7 +560,7 @@ class JSInterpreter(object):
if not _cached:
_cached.extend(itertools.chain(
# Ref: https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Operators/Operator_Precedence
_SC_OPERATORS, _LOG_OPERATORS, _COMP_OPERATORS, _OPERATORS))
_SC_OPERATORS, _LOG_OPERATORS, _COMP_OPERATORS, _OPERATORS, _UNARY_OPERATORS_X))
return _cached
def _operator(self, op, left_val, right_expr, expr, local_vars, allow_recursion):
@ -449,13 +584,14 @@ class JSInterpreter(object):
except Exception as e:
raise self.Exception('Failed to evaluate {left_val!r:.50} {op} {right_val!r:.50}'.format(**locals()), expr, cause=e)
def _index(self, obj, idx, allow_undefined=False):
if idx == 'length':
def _index(self, obj, idx, allow_undefined=True):
if idx == 'length' and isinstance(obj, list):
return len(obj)
try:
return obj[int(idx)] if isinstance(obj, list) else obj[idx]
except Exception as e:
return obj[int(idx)] if isinstance(obj, list) else obj[compat_str(idx)]
except (TypeError, KeyError, IndexError) as e:
if allow_undefined:
# when is not allowed?
return JS_Undefined
raise self.Exception('Cannot get index {idx!r:.100}'.format(**locals()), expr=repr(obj), cause=e)
@ -467,7 +603,7 @@ class JSInterpreter(object):
# used below
_VAR_RET_THROW_RE = re.compile(r'''(?x)
(?P<var>(?:var|const|let)\s)|return(?:\s+|(?=["'])|$)|(?P<throw>throw\s+)
(?:(?P<var>var|const|let)\s+|(?P<ret>return)(?:\s+|(?=["'])|$)|(?P<throw>throw)\s+)
''')
_COMPOUND_RE = re.compile(r'''(?x)
(?P<try>try)\s*\{|
@ -479,6 +615,52 @@ class JSInterpreter(object):
_FINALLY_RE = re.compile(r'finally\s*\{')
_SWITCH_RE = re.compile(r'switch\s*\(')
def handle_operators(self, expr, local_vars, allow_recursion):
for op, _ in self._all_operators():
# hackety: </> have higher priority than <</>>, but don't confuse them
skip_delim = (op + op) if op in '<>*?' else None
if op == '?':
skip_delim = (skip_delim, '?.')
separated = list(self._separate(expr, op, skip_delims=skip_delim))
if len(separated) < 2:
continue
right_expr = separated.pop()
# handle operators that are both unary and binary, minimal BODMAS
if op in ('+', '-'):
# simplify/adjust consecutive instances of these operators
undone = 0
separated = [s.strip() for s in separated]
while len(separated) > 1 and not separated[-1]:
undone += 1
separated.pop()
if op == '-' and undone % 2 != 0:
right_expr = op + right_expr
elif op == '+':
while len(separated) > 1 and set(separated[-1]) <= self.OP_CHARS:
right_expr = separated.pop() + right_expr
if separated[-1][-1:] in self.OP_CHARS:
right_expr = separated.pop() + right_expr
# hanging op at end of left => unary + (strip) or - (push right)
left_val = separated[-1] if separated else ''
for dm_op in ('*', '%', '/', '**'):
bodmas = tuple(self._separate(left_val, dm_op, skip_delims=skip_delim))
if len(bodmas) > 1 and not bodmas[-1].strip():
expr = op.join(separated) + op + right_expr
if len(separated) > 1:
separated.pop()
right_expr = op.join((left_val, right_expr))
else:
separated = [op.join((left_val, right_expr))]
right_expr = None
break
if right_expr is None:
continue
left_val = self.interpret_expression(op.join(separated), local_vars, allow_recursion)
return self._operator(op, left_val, right_expr, expr, local_vars, allow_recursion), True
@Debugger.wrap_interpreter
def interpret_statement(self, stmt, local_vars, allow_recursion=100):
if allow_recursion < 0:
@ -501,7 +683,7 @@ class JSInterpreter(object):
expr = stmt[len(m.group(0)):].strip()
if m.group('throw'):
raise JS_Throw(self.interpret_expression(expr, local_vars, allow_recursion))
should_return = not m.group('var')
should_return = 'return' if m.group('ret') else False
if not expr:
return None, should_return
@ -533,9 +715,15 @@ class JSInterpreter(object):
else:
raise self.Exception('Unsupported object {obj:.100}'.format(**locals()), expr=expr)
if expr.startswith('void '):
left = self.interpret_expression(expr[5:], local_vars, allow_recursion)
return None, should_return
for op, _ in _UNARY_OPERATORS_X:
if not expr.startswith(op):
continue
operand = expr[len(op):]
if not operand or operand[0] != ' ':
continue
op_result = self.handle_operators(expr, local_vars, allow_recursion)
if op_result:
return op_result[0], should_return
if expr.startswith('{'):
inner, outer = self._separate_at_paren(expr)
@ -582,7 +770,7 @@ class JSInterpreter(object):
if_expr, expr = self._separate_at_paren(expr)
else:
# may lose ... else ... because of ll.368-374
if_expr, expr = self._separate_at_paren(expr, delim=';')
if_expr, expr = self._separate_at_paren(' %s;' % (expr,), delim=';')
else_expr = None
m = re.match(r'else\s*(?P<block>\{)?', expr)
if m:
@ -720,7 +908,7 @@ class JSInterpreter(object):
start, end = m.span()
sign = m.group('pre_sign') or m.group('post_sign')
ret = local_vars[var]
local_vars[var] += 1 if sign[0] == '+' else -1
local_vars[var] = _js_add(ret, 1 if sign[0] == '+' else -1)
if m.group('pre_sign'):
ret = local_vars[var]
expr = expr[:start] + self._dump(ret, local_vars) + expr[end:]
@ -730,13 +918,13 @@ class JSInterpreter(object):
m = re.match(r'''(?x)
(?P<assign>
(?P<out>{_NAME_RE})(?:\[(?P<index>[^\]]+?)\])?\s*
(?P<out>{_NAME_RE})(?:\[(?P<out_idx>(?:.+?\]\s*\[)*.+?)\])?\s*
(?P<op>{_OPERATOR_RE})?
=(?!=)(?P<expr>.*)$
)|(?P<return>
(?!if|return|true|false|null|undefined|NaN|Infinity)(?P<name>{_NAME_RE})$
)|(?P<indexing>
(?P<in>{_NAME_RE})\[(?P<idx>.+)\]$
(?P<in>{_NAME_RE})\[(?P<in_idx>(?:.+?\]\s*\[)*.+?)\]$
)|(?P<attribute>
(?P<var>{_NAME_RE})(?:(?P<nullish>\?)?\.(?P<member>[^(]+)|\[(?P<member2>[^\]]+)\])\s*
)|(?P<function>
@ -746,19 +934,23 @@ class JSInterpreter(object):
if md.get('assign'):
left_val = local_vars.get(m.group('out'))
if not m.group('index'):
if not m.group('out_idx'):
local_vars[m.group('out')] = self._operator(
m.group('op'), left_val, m.group('expr'), expr, local_vars, allow_recursion)
return local_vars[m.group('out')], should_return
elif left_val in (None, JS_Undefined):
raise self.Exception('Cannot index undefined variable ' + m.group('out'), expr=expr)
idx = self.interpret_expression(m.group('index'), local_vars, allow_recursion)
if not isinstance(idx, (int, float)):
raise self.Exception('List index %s must be integer' % (idx, ), expr=expr)
indexes = re.split(r'\]\s*\[', m.group('out_idx'))
for i, idx in enumerate(indexes, 1):
idx = self.interpret_expression(idx, local_vars, allow_recursion)
if i < len(indexes):
left_val = self._index(left_val, idx)
if isinstance(idx, float):
idx = int(idx)
left_val[idx] = self._operator(
m.group('op'), self._index(left_val, idx), m.group('expr'), expr, local_vars, allow_recursion)
m.group('op'), self._index(left_val, idx) if m.group('op') else None,
m.group('expr'), expr, local_vars, allow_recursion)
return left_val[idx], should_return
elif expr.isdigit():
@ -776,63 +968,31 @@ class JSInterpreter(object):
return _Infinity, should_return
elif md.get('return'):
return local_vars[m.group('name')], should_return
ret = local_vars[m.group('name')]
# challenge may try to force returning the original value
# use an optional internal var to block this
if should_return == 'return':
if '_ytdl_do_not_return' not in local_vars:
return ret, True
return (ret, True) if ret != local_vars['_ytdl_do_not_return'] else (ret, False)
else:
return ret, should_return
try:
with compat_contextlib_suppress(ValueError):
ret = json.loads(js_to_json(expr)) # strict=True)
if not md.get('attribute'):
return ret, should_return
except ValueError:
pass
if md.get('indexing'):
val = local_vars[m.group('in')]
idx = self.interpret_expression(m.group('idx'), local_vars, allow_recursion)
return self._index(val, idx), should_return
for idx in re.split(r'\]\s*\[', m.group('in_idx')):
idx = self.interpret_expression(idx, local_vars, allow_recursion)
val = self._index(val, idx)
return val, should_return
for op, _ in self._all_operators():
# hackety: </> have higher priority than <</>>, but don't confuse them
skip_delim = (op + op) if op in '<>*?' else None
if op == '?':
skip_delim = (skip_delim, '?.')
separated = list(self._separate(expr, op, skip_delims=skip_delim))
if len(separated) < 2:
continue
right_expr = separated.pop()
# handle operators that are both unary and binary, minimal BODMAS
if op in ('+', '-'):
# simplify/adjust consecutive instances of these operators
undone = 0
separated = [s.strip() for s in separated]
while len(separated) > 1 and not separated[-1]:
undone += 1
separated.pop()
if op == '-' and undone % 2 != 0:
right_expr = op + right_expr
elif op == '+':
while len(separated) > 1 and set(separated[-1]) <= self.OP_CHARS:
right_expr = separated.pop() + right_expr
if separated[-1][-1:] in self.OP_CHARS:
right_expr = separated.pop() + right_expr
# hanging op at end of left => unary + (strip) or - (push right)
left_val = separated[-1] if separated else ''
for dm_op in ('*', '%', '/', '**'):
bodmas = tuple(self._separate(left_val, dm_op, skip_delims=skip_delim))
if len(bodmas) > 1 and not bodmas[-1].strip():
expr = op.join(separated) + op + right_expr
if len(separated) > 1:
separated.pop()
right_expr = op.join((left_val, right_expr))
else:
separated = [op.join((left_val, right_expr))]
right_expr = None
break
if right_expr is None:
continue
left_val = self.interpret_expression(op.join(separated), local_vars, allow_recursion)
return self._operator(op, left_val, right_expr, expr, local_vars, allow_recursion), should_return
op_result = self.handle_operators(expr, local_vars, allow_recursion)
if op_result:
return op_result[0], should_return
if md.get('attribute'):
variable, member, nullish = m.group('var', 'member', 'nullish')
@ -877,7 +1037,7 @@ class JSInterpreter(object):
# Member access
if arg_str is None:
return self._index(obj, member, nullish)
return self._index(obj, member)
# Function call
argvals = [
@ -904,7 +1064,7 @@ class JSInterpreter(object):
if obj is compat_str:
if member == 'fromCharCode':
assertion(argvals, 'takes one or more arguments')
return ''.join(map(compat_chr, argvals))
return ''.join(compat_chr(int(n)) for n in argvals)
raise self.Exception('Unsupported string method ' + member, expr=expr)
elif obj is float:
if member == 'pow':
@ -913,13 +1073,47 @@ class JSInterpreter(object):
raise self.Exception('Unsupported Math method ' + member, expr=expr)
if member == 'split':
assertion(argvals, 'takes one or more arguments')
assertion(len(argvals) == 1, 'with limit argument is not implemented')
return obj.split(argvals[0]) if argvals[0] else list(obj)
assertion(len(argvals) <= 2, 'takes at most two arguments')
if len(argvals) > 1:
limit = argvals[1]
assertion(isinstance(limit, int) and limit >= 0, 'integer limit >= 0')
if limit == 0:
return []
else:
limit = 0
if len(argvals) == 0:
argvals = [JS_Undefined]
elif isinstance(argvals[0], self.JS_RegExp):
# avoid re.split(), similar but not enough
def where():
for m in argvals[0].finditer(obj):
yield m.span(0)
yield (None, None)
def splits(limit=limit):
i = 0
for j, jj in where():
if j == jj == 0:
continue
if j is None and i >= len(obj):
break
yield obj[i:j]
if jj is None or limit == 1:
break
limit -= 1
i = jj
return list(splits())
return (
obj.split(argvals[0], limit - 1) if argvals[0] and argvals[0] != JS_Undefined
else list(obj)[:limit or None])
elif member == 'join':
assertion(isinstance(obj, list), 'must be applied on a list')
assertion(len(argvals) == 1, 'takes exactly one argument')
return argvals[0].join(obj)
assertion(len(argvals) <= 1, 'takes at most one argument')
return (',' if len(argvals) == 0 else argvals[0]).join(
('' if x in (None, JS_Undefined) else _js_toString(x))
for x in obj)
elif member == 'reverse':
assertion(not argvals, 'does not take any arguments')
obj.reverse()
@ -941,37 +1135,31 @@ class JSInterpreter(object):
index, how_many = map(int, (argvals + [len(obj)])[:2])
if index < 0:
index += len(obj)
add_items = argvals[2:]
res = []
for _ in range(index, min(index + how_many, len(obj))):
res.append(obj.pop(index))
for i, item in enumerate(add_items):
obj.insert(index + i, item)
res = [obj.pop(index)
for _ in range(index, min(index + how_many, len(obj)))]
obj[index:index] = argvals[2:]
return res
elif member == 'unshift':
assertion(isinstance(obj, list), 'must be applied on a list')
assertion(argvals, 'takes one or more arguments')
for item in reversed(argvals):
obj.insert(0, item)
return obj
elif member == 'pop':
elif member in ('shift', 'pop'):
assertion(isinstance(obj, list), 'must be applied on a list')
assertion(not argvals, 'does not take any arguments')
if not obj:
return
return obj.pop()
return obj.pop(0 if member == 'shift' else -1) if len(obj) > 0 else JS_Undefined
elif member == 'unshift':
assertion(isinstance(obj, list), 'must be applied on a list')
# not enforced: assertion(argvals, 'takes one or more arguments')
obj[0:0] = argvals
return len(obj)
elif member == 'push':
assertion(argvals, 'takes one or more arguments')
# not enforced: assertion(argvals, 'takes one or more arguments')
obj.extend(argvals)
return obj
return len(obj)
elif member == 'forEach':
assertion(argvals, 'takes one or more arguments')
assertion(len(argvals) <= 2, 'takes at-most 2 arguments')
assertion(len(argvals) <= 2, 'takes at most 2 arguments')
f, this = (argvals + [''])[:2]
return [f((item, idx, obj), {'this': this}, allow_recursion) for idx, item in enumerate(obj)]
elif member == 'indexOf':
assertion(argvals, 'takes one or more arguments')
assertion(len(argvals) <= 2, 'takes at-most 2 arguments')
assertion(len(argvals) <= 2, 'takes at most 2 arguments')
idx, start = (argvals + [0])[:2]
try:
return obj.index(idx, start)
@ -980,7 +1168,7 @@ class JSInterpreter(object):
elif member == 'charCodeAt':
assertion(isinstance(obj, compat_str), 'must be applied on a string')
# assertion(len(argvals) == 1, 'takes exactly one argument') # but not enforced
idx = argvals[0] if isinstance(argvals[0], int) else 0
idx = argvals[0] if len(argvals) > 0 and isinstance(argvals[0], int) else 0
if idx >= len(obj):
return None
return ord(obj[idx])
@ -1031,7 +1219,7 @@ class JSInterpreter(object):
yield self.interpret_expression(v, local_vars, allow_recursion)
def extract_object(self, objname):
_FUNC_NAME_RE = r'''(?:[a-zA-Z$0-9]+|"[a-zA-Z$0-9]+"|'[a-zA-Z$0-9]+')'''
_FUNC_NAME_RE = r'''(?:{n}|"{n}"|'{n}')'''.format(n=_NAME_RE)
obj = {}
fields = next(filter(None, (
obj_m.group('fields') for obj_m in re.finditer(
@ -1090,6 +1278,7 @@ class JSInterpreter(object):
def extract_function_from_code(self, argnames, code, *global_stack):
local_vars = {}
while True:
mobj = re.search(r'function\((?P<args>[^)]*)\)\s*{', code)
if mobj is None:
@ -1100,10 +1289,11 @@ class JSInterpreter(object):
[x.strip() for x in mobj.group('args').split(',')],
body, local_vars, *global_stack))
code = code[:start] + name + remaining
return self.build_function(argnames, code, local_vars, *global_stack)
def call_function(self, funcname, *args):
return self.extract_function(funcname)(args)
def call_function(self, funcname, *args, **kw_global_vars):
return self.extract_function(funcname)(args, kw_global_vars)
@classmethod
def build_arglist(cls, arg_text):
@ -1122,8 +1312,9 @@ class JSInterpreter(object):
global_stack = list(global_stack) or [{}]
argnames = tuple(argnames)
def resf(args, kwargs={}, allow_recursion=100):
global_stack[0].update(zip_longest(argnames, args, fillvalue=None))
def resf(args, kwargs=None, allow_recursion=100):
kwargs = kwargs or {}
global_stack[0].update(zip_longest(argnames, args, fillvalue=JS_Undefined))
global_stack[0].update(kwargs)
var_stack = LocalNameSpace(*global_stack)
ret, should_abort = self.interpret_statement(code.replace('\n', ' '), var_stack, allow_recursion - 1)