1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
|
<?php
use MediaWiki\MediaWikiServices;
/**
* Tokenizer for AbuseFilter rules.
*/
class AbuseFilterTokenizer {
/** @var int Tokenizer cache version. Increment this when changing the syntax. **/
const CACHE_VERSION = 1;
const COMMENT_START_RE = '/\s*\/\*/A';
const ID_SYMBOL_RE = '/[0-9A-Za-z_]+/A';
const OPERATOR_RE =
'/(\!\=\=|\!\=|\!|\*\*|\*|\/|\+|\-|%|&|\||\^|\:\=|\?|\:|\<\=|\<|\>\=|\>|\=\=\=|\=\=|\=)/A';
const RADIX_RE = '/([0-9A-Fa-f]+(?:\.\d*)?|\.\d+)([bxo])?/Au';
const WHITESPACE = "\011\012\013\014\015\040";
// Order is important. The punctuation-matching regex requires that
// ** comes before *, etc. They are sorted to make it easy to spot
// such errors.
public static $operators = [
'!==', '!=', '!', // Inequality
'**', '*', // Multiplication/exponentiation
'/', '+', '-', '%', // Other arithmetic
'&', '|', '^', // Logic
':=', // Setting
'?', ':', // Ternery
'<=', '<', // Less than
'>=', '>', // Greater than
'===', '==', '=', // Equality
];
public static $punctuation = [
',' => AFPToken::TCOMMA,
'(' => AFPToken::TBRACE,
')' => AFPToken::TBRACE,
'[' => AFPToken::TSQUAREBRACKET,
']' => AFPToken::TSQUAREBRACKET,
';' => AFPToken::TSTATEMENTSEPARATOR,
];
public static $bases = [
'b' => 2,
'x' => 16,
'o' => 8
];
public static $baseCharsRe = [
2 => '/^[01]+$/',
8 => '/^[0-8]+$/',
16 => '/^[0-9A-Fa-f]+$/',
10 => '/^[0-9.]+$/',
];
public static $keywords = [
'in', 'like', 'true', 'false', 'null', 'contains', 'matches',
'rlike', 'irlike', 'regex', 'if', 'then', 'else', 'end',
];
/**
* @param string $code
* @return array
* @throws AFPException
* @throws AFPUserVisibleException
*/
static function tokenize( $code ) {
static $tokenizerCache = null;
if ( !$tokenizerCache ) {
$tokenizerCache = ObjectCache::getLocalServerInstance( 'hash' );
}
static $stats = null;
if ( !$stats ) {
$stats = MediaWikiServices::getInstance()->getStatsdDataFactory();
}
$cacheKey = wfGlobalCacheKey( __CLASS__, self::CACHE_VERSION, crc32( $code ) );
$tokens = $tokenizerCache->get( $cacheKey );
if ( $tokens ) {
$stats->increment( 'AbuseFilter.tokenizerCache.hit' );
return $tokens;
}
$stats->increment( 'AbuseFilter.tokenizerCache.miss' );
$tokens = [];
$curPos = 0;
do {
$prevPos = $curPos;
$token = self::nextToken( $code, $curPos );
$tokens[ $token->pos ] = [ $token, $curPos ];
} while ( $curPos !== $prevPos );
$tokenizerCache->set( $cacheKey, $tokens, 60 * 60 * 24 );
return $tokens;
}
/**
* @param string $code
* @param int &$offset
* @return AFPToken
* @throws AFPException
* @throws AFPUserVisibleException
*/
protected static function nextToken( $code, &$offset ) {
$matches = [];
$start = $offset;
// Read past comments
while ( preg_match( self::COMMENT_START_RE, $code, $matches, 0, $offset ) ) {
if ( strpos( $code, '*/', $offset ) === false ) {
throw new AFPUserVisibleException(
'unclosedcomment', $offset, [] );
}
$offset = strpos( $code, '*/', $offset ) + 2;
}
// Spaces
$offset += strspn( $code, self::WHITESPACE, $offset );
if ( $offset >= strlen( $code ) ) {
return new AFPToken( AFPToken::TNONE, '', $start );
}
$chr = $code[$offset];
// Punctuation
if ( isset( self::$punctuation[$chr] ) ) {
$offset++;
return new AFPToken( self::$punctuation[$chr], $chr, $start );
}
// String literal
if ( $chr === '"' || $chr === "'" ) {
return self::readStringLiteral( $code, $offset, $start );
}
$matches = [];
// Operators
if ( preg_match( self::OPERATOR_RE, $code, $matches, 0, $offset ) ) {
$token = $matches[0];
$offset += strlen( $token );
return new AFPToken( AFPToken::TOP, $token, $start );
}
// Numbers
if ( preg_match( self::RADIX_RE, $code, $matches, 0, $offset ) ) {
$token = $matches[0];
$input = $matches[1];
$baseChar = isset( $matches[2] ) ? $matches[2] : null;
// Sometimes the base char gets mixed in with the rest of it because
// the regex targets hex, too.
// This mostly happens with binary
if ( !$baseChar && !empty( self::$bases[ substr( $input, - 1 ) ] ) ) {
$baseChar = substr( $input, - 1, 1 );
$input = substr( $input, 0, - 1 );
}
$base = $baseChar ? self::$bases[$baseChar] : 10;
// Check against the appropriate character class for input validation
if ( preg_match( self::$baseCharsRe[$base], $input ) ) {
$num = $base !== 10 ? base_convert( $input, $base, 10 ) : $input;
$offset += strlen( $token );
return ( strpos( $input, '.' ) !== false )
? new AFPToken( AFPToken::TFLOAT, floatval( $num ), $start )
: new AFPToken( AFPToken::TINT, intval( $num ), $start );
}
}
// IDs / Keywords
if ( preg_match( self::ID_SYMBOL_RE, $code, $matches, 0, $offset ) ) {
$token = $matches[0];
$offset += strlen( $token );
$type = in_array( $token, self::$keywords )
? AFPToken::TKEYWORD
: AFPToken::TID;
return new AFPToken( $type, $token, $start );
}
throw new AFPUserVisibleException(
'unrecognisedtoken', $start, [ substr( $code, $start ) ] );
}
/**
* @param string $code
* @param int &$offset
* @param int $start
* @return AFPToken
* @throws AFPException
* @throws AFPUserVisibleException
*/
protected static function readStringLiteral( $code, &$offset, $start ) {
$type = $code[$offset];
$offset++;
$length = strlen( $code );
$token = '';
while ( $offset < $length ) {
if ( $code[$offset] === $type ) {
$offset++;
return new AFPToken( AFPToken::TSTRING, $token, $start );
}
// Performance: Use a PHP function (implemented in C)
// to scan ahead.
$addLength = strcspn( $code, $type . "\\", $offset );
if ( $addLength ) {
$token .= substr( $code, $offset, $addLength );
$offset += $addLength;
} elseif ( $code[$offset] == '\\' ) {
switch ( $code[$offset + 1] ) {
case '\\':
$token .= '\\';
break;
case $type:
$token .= $type;
break;
case 'n';
$token .= "\n";
break;
case 'r':
$token .= "\r";
break;
case 't':
$token .= "\t";
break;
case 'x':
$chr = substr( $code, $offset + 2, 2 );
if ( preg_match( '/^[0-9A-Fa-f]{2}$/', $chr ) ) {
$chr = base_convert( $chr, 16, 10 );
$token .= chr( $chr );
$offset += 2; # \xXX -- 2 done later
} else {
$token .= 'x';
}
break;
default:
$token .= "\\" . $code[$offset + 1];
}
$offset += 2;
} else {
$token .= $code[$offset];
$offset++;
}
}
throw new AFPUserVisibleException( 'unclosedstring', $offset, [] );
}
}
|
