1
|
|
|
<?php |
2
|
|
|
|
3
|
|
|
namespace Stichoza\GoogleTranslate\Tokens; |
4
|
|
|
|
5
|
|
|
/** |
6
|
|
|
* Google Token Generator. |
7
|
|
|
* |
8
|
|
|
* Thanks to @helen5106 and @tehmaestro and few other cool guys |
9
|
|
|
* at https://github.com/Stichoza/google-translate-php/issues/32 |
10
|
|
|
*/ |
11
|
|
|
class GoogleTokenGenerator implements TokenProviderInterface |
12
|
|
|
{ |
13
|
|
|
/** |
14
|
|
|
* Generate and return a token. |
15
|
|
|
* |
16
|
|
|
* @param string $source Source language |
17
|
|
|
* @param string $target Target language |
18
|
|
|
* @param string $text Text to translate |
19
|
|
|
* |
20
|
|
|
* @return mixed A token |
21
|
|
|
*/ |
22
|
|
|
public function generateToken($source, $target, $text) |
23
|
|
|
{ |
24
|
|
|
return $this->TL($text); |
25
|
|
|
} |
26
|
|
|
|
27
|
|
|
/** |
28
|
|
|
* Generate a valid Google Translate request token. |
29
|
|
|
* |
30
|
|
|
* @param string $a text to translate |
31
|
|
|
* |
32
|
|
|
* @return string |
33
|
|
|
*/ |
34
|
|
|
private function TL($a) |
35
|
|
|
{ |
36
|
|
|
$tkk = $this->TKK(); |
37
|
|
|
$b = $tkk[0]; |
38
|
|
|
|
39
|
|
|
for ($d = [], $e = 0, $f = 0; $f < mb_strlen($a, 'UTF-8'); $f++) { |
40
|
|
|
$g = $this->charCodeAt($a, $f); |
41
|
|
|
if (128 > $g) { |
42
|
|
|
$d[$e++] = $g; |
43
|
|
|
} else { |
44
|
|
|
if (2048 > $g) { |
45
|
|
|
$d[$e++] = $g >> 6 | 192; |
46
|
|
|
} else { |
47
|
|
|
if (55296 == ($g & 64512) && $f + 1 < mb_strlen($a, 'UTF-8') && 56320 == ($this->charCodeAt($a, $f + 1) & 64512)) { |
48
|
|
|
$g = 65536 + (($g & 1023) << 10) + ($this->charCodeAt($a, ++$f) & 1023); |
49
|
|
|
$d[$e++] = $g >> 18 | 240; |
50
|
|
|
$d[$e++] = $g >> 12 & 63 | 128; |
51
|
|
|
} else { |
52
|
|
|
$d[$e++] = $g >> 12 | 224; |
53
|
|
|
} |
54
|
|
|
$d[$e++] = $g >> 6 & 63 | 128; |
55
|
|
|
} |
56
|
|
|
$d[$e++] = $g & 63 | 128; |
57
|
|
|
} |
58
|
|
|
} |
59
|
|
|
$a = $b; |
60
|
|
|
for ($e = 0; $e < count($d); $e++) { |
|
|
|
|
61
|
|
|
$a += $d[$e]; |
62
|
|
|
$a = $this->RL($a, '+-a^+6'); |
63
|
|
|
} |
64
|
|
|
$a = $this->RL($a, '+-3^+b+-f'); |
65
|
|
|
$a ^= $tkk[1] ? $tkk[1] + 0 : 0; |
66
|
|
|
if (0 > $a) { |
67
|
|
|
$a = ($a & 2147483647) + 2147483648; |
68
|
|
|
} |
69
|
|
|
$a = fmod($a, pow(10, 6)); |
70
|
|
|
|
71
|
|
|
return $a.'.'.($a ^ $b); |
72
|
|
|
} |
73
|
|
|
|
74
|
|
|
/** |
75
|
|
|
* @return array |
76
|
|
|
*/ |
77
|
|
|
private function TKK() |
78
|
|
|
{ |
79
|
|
|
return ['406398', (561666268 + 1526272306)]; |
80
|
|
|
} |
81
|
|
|
|
82
|
|
|
/** |
83
|
|
|
* Process token data by applying multiple operations. |
84
|
|
|
* (Params are safe, no need for multibyte functions) |
85
|
|
|
* |
86
|
|
|
* @param int $a |
87
|
|
|
* @param string $b |
88
|
|
|
* |
89
|
|
|
* @return int |
90
|
|
|
*/ |
91
|
|
|
private function RL($a, $b) |
92
|
|
|
{ |
93
|
|
|
for ($c = 0; $c < strlen($b) - 2; $c += 3) { |
94
|
|
|
$d = $b[$c + 2]; |
95
|
|
|
$d = 'a' <= $d ? ord($d[0]) - 87 : intval($d); |
96
|
|
|
$d = '+' == $b[$c + 1] ? $this->unsignedRightShift($a, $d) : $a << $d; |
97
|
|
|
$a = '+' == $b[$c] ? ($a + $d & 4294967295) : $a ^ $d; |
98
|
|
|
} |
99
|
|
|
|
100
|
|
|
return $a; |
101
|
|
|
} |
102
|
|
|
|
103
|
|
|
/** |
104
|
|
|
* Unsigned right shift implementation |
105
|
|
|
* https://msdn.microsoft.com/en-us/library/342xfs5s(v=vs.94).aspx |
106
|
|
|
* http://stackoverflow.com/a/43359819/2953830 |
107
|
|
|
* |
108
|
|
|
* @param $a |
109
|
|
|
* @param $b |
110
|
|
|
* |
111
|
|
|
* @return number |
112
|
|
|
*/ |
113
|
|
|
private function unsignedRightShift($a, $b) |
114
|
|
|
{ |
115
|
|
|
if ($b >= 32 || $b < -32) { |
116
|
|
|
$m = (int)($b / 32); |
117
|
|
|
$b = $b - ($m * 32); |
118
|
|
|
} |
119
|
|
|
|
120
|
|
|
if ($b < 0) { |
121
|
|
|
$b = 32 + $b; |
122
|
|
|
} |
123
|
|
|
|
124
|
|
|
if ($b == 0) { |
125
|
|
|
return (($a >> 1) & 0x7fffffff) * 2 + (($a >> $b) & 1); |
126
|
|
|
} |
127
|
|
|
|
128
|
|
|
if ($a < 0) { |
129
|
|
|
$a = ($a >> 1); |
130
|
|
|
$a &= 2147483647; |
131
|
|
|
$a |= 0x40000000; |
132
|
|
|
$a = ($a >> ($b - 1)); |
133
|
|
|
} else { |
134
|
|
|
$a = ($a >> $b); |
135
|
|
|
} |
136
|
|
|
|
137
|
|
|
return $a; |
138
|
|
|
} |
139
|
|
|
|
140
|
|
|
/** |
141
|
|
|
* Get the Unicode of the character at the specified index in a string. |
142
|
|
|
* |
143
|
|
|
* @param string $str |
144
|
|
|
* @param int $index |
145
|
|
|
* |
146
|
|
|
* @return null|number |
147
|
|
|
*/ |
148
|
|
|
private function charCodeAt($str, $index) |
149
|
|
|
{ |
150
|
|
|
$char = mb_substr($str, $index, 1, 'UTF-8'); |
151
|
|
|
if (mb_check_encoding($char, 'UTF-8')) { |
152
|
|
|
$ret = mb_convert_encoding($char, 'UTF-32BE', 'UTF-8'); |
153
|
|
|
$result = hexdec(bin2hex($ret)); |
154
|
|
|
|
155
|
|
|
return $result; |
156
|
|
|
} |
157
|
|
|
|
158
|
|
|
return; |
159
|
|
|
} |
160
|
|
|
} |
161
|
|
|
|
If the size of the collection does not change during the iteration, it is generally a good practice to compute it beforehand, and not on each iteration: