1
|
|
|
<?php |
2
|
|
|
|
3
|
|
|
namespace PhpOffice\PhpSpreadsheet\Reader\Security; |
4
|
|
|
|
5
|
|
|
use PhpOffice\PhpSpreadsheet\Reader; |
6
|
|
|
use PhpOffice\PhpSpreadsheet\Settings; |
7
|
|
|
|
8
|
|
|
class XmlScanner |
9
|
|
|
{ |
10
|
|
|
/** |
11
|
|
|
* Identifies whether the thread-safe libxmlDisableEntityLoader() function is available. |
12
|
|
|
* |
13
|
|
|
* @var bool |
14
|
|
|
*/ |
15
|
|
|
private $libxmlDisableEntityLoader = false; |
|
|
|
|
16
|
|
|
|
17
|
|
|
/** |
18
|
|
|
* String used to identify risky xml elements. |
19
|
|
|
* |
20
|
|
|
* @var string |
21
|
|
|
*/ |
22
|
|
|
private $pattern; |
23
|
|
|
|
24
|
|
|
private $callback; |
25
|
|
|
|
26
|
|
|
private static $libxmlDisableEntityLoaderValue; |
27
|
|
|
|
28
|
128 |
|
public function __construct($pattern = '<!DOCTYPE') |
29
|
|
|
{ |
30
|
128 |
|
$this->pattern = $pattern; |
31
|
|
|
|
32
|
128 |
|
$this->disableEntityLoaderCheck(); |
33
|
|
|
|
34
|
|
|
// A fatal error will bypass the destructor, so we register a shutdown here |
35
|
128 |
|
register_shutdown_function([$this, '__destruct']); |
36
|
128 |
|
} |
37
|
|
|
|
38
|
128 |
|
public static function getInstance(Reader\IReader $reader) |
39
|
|
|
{ |
40
|
|
|
switch (true) { |
41
|
128 |
|
case $reader instanceof Reader\Html: |
42
|
26 |
|
return new self('<!ENTITY'); |
43
|
102 |
|
case $reader instanceof Reader\Xlsx: |
44
|
48 |
|
case $reader instanceof Reader\Xml: |
45
|
21 |
|
case $reader instanceof Reader\Ods: |
46
|
4 |
|
case $reader instanceof Reader\Gnumeric: |
47
|
102 |
|
return new self('<!DOCTYPE'); |
48
|
|
|
default: |
49
|
|
|
return new self('<!DOCTYPE'); |
50
|
|
|
} |
51
|
|
|
} |
52
|
|
|
|
53
|
|
|
public static function threadSafeLibxmlDisableEntityLoaderAvailability() |
54
|
|
|
{ |
55
|
|
|
if (PHP_MAJOR_VERSION == 7) { |
56
|
|
|
switch (PHP_MINOR_VERSION) { |
57
|
|
|
case 2: |
58
|
|
|
return PHP_RELEASE_VERSION >= 1; |
59
|
|
|
case 1: |
60
|
|
|
return PHP_RELEASE_VERSION >= 13; |
61
|
|
|
case 0: |
62
|
|
|
return PHP_RELEASE_VERSION >= 27; |
63
|
|
|
} |
64
|
|
|
|
65
|
|
|
return true; |
66
|
|
|
} |
67
|
|
|
|
68
|
|
|
return false; |
69
|
|
|
} |
70
|
|
|
|
71
|
128 |
|
private function disableEntityLoaderCheck() |
72
|
|
|
{ |
73
|
128 |
|
if (Settings::getLibXmlDisableEntityLoader()) { |
74
|
128 |
|
$libxmlDisableEntityLoaderValue = libxml_disable_entity_loader(true); |
75
|
|
|
|
76
|
128 |
|
if (self::$libxmlDisableEntityLoaderValue === null) { |
77
|
18 |
|
self::$libxmlDisableEntityLoaderValue = $libxmlDisableEntityLoaderValue; |
78
|
|
|
} |
79
|
|
|
} |
80
|
128 |
|
} |
81
|
|
|
|
82
|
|
|
private function shutdown() |
83
|
|
|
{ |
84
|
|
|
if (self::$libxmlDisableEntityLoaderValue !== null) { |
85
|
|
|
libxml_disable_entity_loader(self::$libxmlDisableEntityLoaderValue); |
86
|
|
|
} |
87
|
|
|
} |
88
|
|
|
|
89
|
|
|
public function __destruct() |
90
|
|
|
{ |
91
|
|
|
$this->shutdown(); |
92
|
|
|
} |
93
|
|
|
|
94
|
1 |
|
public function setAdditionalCallback(callable $callback) |
95
|
|
|
{ |
96
|
1 |
|
$this->callback = $callback; |
97
|
1 |
|
} |
98
|
|
|
|
99
|
108 |
|
private function toUtf8($xml) |
100
|
|
|
{ |
101
|
108 |
|
$pattern = '/encoding="(.*?)"/'; |
102
|
108 |
|
$result = preg_match($pattern, $xml, $matches); |
103
|
108 |
|
$charset = $result ? $matches[1] : 'UTF-8'; |
104
|
|
|
|
105
|
108 |
|
if ($charset !== 'UTF-8') { |
106
|
4 |
|
$xml = mb_convert_encoding($xml, 'UTF-8', $charset); |
107
|
|
|
|
108
|
4 |
|
$result = preg_match($pattern, $xml, $matches); |
109
|
4 |
|
$charset = $result ? $matches[1] : 'UTF-8'; |
110
|
4 |
|
if ($charset !== 'UTF-8') { |
111
|
4 |
|
throw new Reader\Exception('Suspicious Double-encoded XML, spreadsheet file load() aborted to prevent XXE/XEE attacks'); |
112
|
|
|
} |
113
|
|
|
} |
114
|
|
|
|
115
|
104 |
|
return $xml; |
116
|
|
|
} |
117
|
|
|
|
118
|
|
|
/** |
119
|
|
|
* Scan the XML for use of <!ENTITY to prevent XXE/XEE attacks. |
120
|
|
|
* |
121
|
|
|
* @param mixed $xml |
122
|
|
|
* |
123
|
|
|
* @throws Reader\Exception |
124
|
|
|
* |
125
|
|
|
* @return string |
126
|
|
|
*/ |
127
|
108 |
|
public function scan($xml) |
128
|
|
|
{ |
129
|
108 |
|
$this->disableEntityLoaderCheck(); |
130
|
|
|
|
131
|
108 |
|
$xml = $this->toUtf8($xml); |
132
|
|
|
|
133
|
|
|
// Don't rely purely on libxml_disable_entity_loader() |
134
|
104 |
|
$pattern = '/\\0?' . implode('\\0?', str_split($this->pattern)) . '\\0?/'; |
135
|
|
|
|
136
|
104 |
|
if (preg_match($pattern, $xml)) { |
137
|
8 |
|
throw new Reader\Exception('Detected use of ENTITY in XML, spreadsheet file load() aborted to prevent XXE/XEE attacks'); |
138
|
|
|
} |
139
|
|
|
|
140
|
96 |
|
if ($this->callback !== null && is_callable($this->callback)) { |
141
|
1 |
|
$xml = call_user_func($this->callback, $xml); |
142
|
|
|
} |
143
|
|
|
|
144
|
96 |
|
return $xml; |
145
|
|
|
} |
146
|
|
|
|
147
|
|
|
/** |
148
|
|
|
* Scan theXML for use of <!ENTITY to prevent XXE/XEE attacks. |
149
|
|
|
* |
150
|
|
|
* @param string $filestream |
151
|
|
|
* |
152
|
|
|
* @throws Reader\Exception |
153
|
|
|
* |
154
|
|
|
* @return string |
155
|
|
|
*/ |
156
|
40 |
|
public function scanFile($filestream) |
157
|
|
|
{ |
158
|
40 |
|
return $this->scan(file_get_contents($filestream)); |
159
|
|
|
} |
160
|
|
|
} |
161
|
|
|
|