|
1
|
|
|
<?php |
|
2
|
|
|
/** |
|
3
|
|
|
* @author: Viskov Sergey |
|
4
|
|
|
* @date : 8/4/15 |
|
5
|
|
|
* @time : 9:31 PM |
|
6
|
|
|
*/ |
|
7
|
|
|
|
|
8
|
|
|
namespace LTDBeget\dns; |
|
9
|
|
|
|
|
10
|
|
|
use LTDBeget\ascii\AsciiChar; |
|
11
|
|
|
use LTDBeget\dns\record\Record; |
|
12
|
|
|
use LTDBeget\stringstream\StringStream; |
|
13
|
|
|
|
|
14
|
|
|
/** |
|
15
|
|
|
* Class Tokenizer |
|
16
|
|
|
* |
|
17
|
|
|
* @package LTDBeget\dns |
|
18
|
|
|
*/ |
|
19
|
|
|
final class Tokenizer |
|
20
|
|
|
{ |
|
21
|
|
|
/** |
|
22
|
|
|
* @var StringStream |
|
23
|
|
|
*/ |
|
24
|
|
|
private $stream; |
|
25
|
|
|
/** |
|
26
|
|
|
* Result of tokenize input string |
|
27
|
|
|
* |
|
28
|
|
|
* @var array |
|
29
|
|
|
*/ |
|
30
|
|
|
private $tokens = []; |
|
31
|
|
|
/** |
|
32
|
|
|
* @var string |
|
33
|
|
|
*/ |
|
34
|
|
|
private $ttl; |
|
35
|
|
|
/** |
|
36
|
|
|
* @var null |
|
37
|
|
|
*/ |
|
38
|
|
|
private $origin; |
|
39
|
|
|
/** |
|
40
|
|
|
* @var array |
|
41
|
|
|
*/ |
|
42
|
|
|
private $allowedGlobalVariables = [ |
|
43
|
|
|
'origin' => true, |
|
44
|
|
|
'ttl' => true |
|
45
|
|
|
]; |
|
46
|
|
|
|
|
47
|
|
|
/** |
|
48
|
|
|
* @var int |
|
49
|
|
|
*/ |
|
50
|
|
|
private $recordsAmmount = 0; |
|
51
|
|
|
|
|
52
|
|
|
/** |
|
53
|
|
|
* Tokenizer constructor. |
|
54
|
|
|
* |
|
55
|
|
|
* @param string $string |
|
56
|
|
|
*/ |
|
57
|
7 |
|
private function __construct(string $string) |
|
58
|
|
|
{ |
|
59
|
7 |
|
$this->stream = new StringStream($string); |
|
60
|
7 |
|
} |
|
61
|
|
|
|
|
62
|
|
|
/** |
|
63
|
|
|
* @param string $plainData |
|
64
|
|
|
* @return array |
|
65
|
|
|
*/ |
|
66
|
7 |
|
public static function tokenize(string $plainData) : array |
|
67
|
|
|
{ |
|
68
|
7 |
|
return (new self($plainData))->tokenizeInternal()->tokens; |
|
69
|
|
|
} |
|
70
|
|
|
|
|
71
|
|
|
/** |
|
72
|
|
|
* @return Tokenizer |
|
73
|
|
|
*/ |
|
74
|
7 |
|
private function tokenizeInternal() : Tokenizer |
|
75
|
|
|
{ |
|
76
|
|
|
do { |
|
77
|
7 |
|
$this->stream->ignoreWhitespace(); |
|
78
|
7 |
|
if ($this->stream->currentAscii()->is(AsciiChar::DOLLAR)) { |
|
79
|
3 |
|
$this->stream->next(); |
|
80
|
3 |
|
$this->extractGlobalVariable(); |
|
81
|
7 |
|
} elseif ($this->stream->currentAscii()->is(AsciiChar::SEMICOLON)) { |
|
82
|
5 |
|
$this->ignoreComment(); |
|
83
|
|
|
} else { |
|
84
|
7 |
|
$this->extractRecord(); |
|
85
|
6 |
|
$this->stream->ignoreWhitespace(); |
|
86
|
|
|
} |
|
87
|
6 |
|
} while (!$this->stream->isEnd()); |
|
88
|
|
|
|
|
89
|
6 |
|
return $this; |
|
90
|
|
|
} |
|
91
|
|
|
|
|
92
|
3 |
|
private function extractGlobalVariable() |
|
93
|
|
|
{ |
|
94
|
3 |
|
$variableName = ''; |
|
95
|
|
|
start: |
|
96
|
3 |
|
if ($this->stream->currentAscii()->isLetter()) { |
|
97
|
3 |
|
$variableName .= $this->stream->current(); |
|
98
|
3 |
|
$this->stream->next(); |
|
99
|
3 |
|
goto start; |
|
100
|
3 |
|
} elseif ($this->stream->currentAscii()->isHorizontalSpace()) { |
|
101
|
3 |
|
$variableName = mb_strtolower($variableName); |
|
102
|
3 |
|
if (!array_key_exists($variableName, $this->allowedGlobalVariables)) { |
|
103
|
|
|
throw new SyntaxErrorException($this->stream); |
|
104
|
|
|
} |
|
105
|
3 |
|
$this->stream->ignoreHorizontalSpace(); |
|
106
|
3 |
|
$this->extractGlobalVariableValue($variableName); |
|
107
|
|
|
} else { |
|
108
|
|
|
throw new SyntaxErrorException($this->stream); |
|
109
|
|
|
} |
|
110
|
3 |
|
} |
|
111
|
|
|
|
|
112
|
|
|
/** |
|
113
|
|
|
* @param string $variableName |
|
114
|
|
|
*/ |
|
115
|
3 |
|
private function extractGlobalVariableValue(string $variableName) |
|
116
|
|
|
{ |
|
117
|
3 |
|
$this->{$variableName} = ''; |
|
118
|
|
|
start: |
|
119
|
3 |
|
$char = $this->stream->currentAscii(); |
|
120
|
3 |
|
$this->{$variableName} .= ''; |
|
121
|
3 |
|
if ($char->isLetter() || |
|
122
|
3 |
|
$char->isDigit() || |
|
123
|
3 |
|
$char->is(AsciiChar::UNDERSCORE) || |
|
124
|
3 |
|
$char->is(AsciiChar::DOT) || |
|
125
|
3 |
|
$char->is(AsciiChar::HYPHEN) || |
|
126
|
3 |
|
$char->is(AsciiChar::AT_SYMBOL) || |
|
127
|
3 |
|
$char->is(AsciiChar::ASTERISK) |
|
128
|
|
|
) { |
|
129
|
3 |
|
$this->{$variableName} .= $this->stream->current(); |
|
130
|
3 |
|
$this->stream->next(); |
|
131
|
3 |
|
goto start; |
|
132
|
3 |
|
} elseif ($char->isWhiteSpace()) { |
|
133
|
3 |
|
return; |
|
134
|
|
|
} else { |
|
135
|
|
|
throw new SyntaxErrorException($this->stream); |
|
136
|
|
|
} |
|
137
|
|
|
} |
|
138
|
|
|
|
|
139
|
5 |
View Code Duplication |
private function ignoreComment() |
|
|
|
|
|
|
140
|
|
|
{ |
|
141
|
|
|
start: |
|
142
|
5 |
|
if (!$this->stream->currentAscii()->isVerticalSpace() && !$this->stream->isEnd()) { |
|
143
|
5 |
|
$this->stream->next(); |
|
144
|
5 |
|
goto start; |
|
145
|
|
|
} |
|
146
|
5 |
|
} |
|
147
|
|
|
|
|
148
|
7 |
|
private function extractRecord() |
|
149
|
|
|
{ |
|
150
|
7 |
|
$isFirst = $this->recordsAmmount === 0; |
|
151
|
7 |
|
if(! $isFirst ) { |
|
152
|
5 |
|
$lastParsedRecord = end($this->tokens); |
|
153
|
5 |
|
$previousName = $lastParsedRecord['NAME']; |
|
154
|
|
|
} else { |
|
155
|
7 |
|
$previousName = NULL; |
|
156
|
|
|
} |
|
157
|
7 |
|
$this->tokens[] = (new Record($this->stream, $this->origin, $this->ttl, $isFirst, $previousName))->tokenize(); |
|
158
|
6 |
|
$this->recordsAmmount++; |
|
159
|
|
|
} |
|
160
|
|
|
} |
LTD-Beget /
dns-zone-tokenizer
Duplicated code is one of the most pungent code smells. If you need to duplicate the same code in three or more different places, we strongly encourage you to look into extracting the code into a single class or operation.
You can also find more detailed suggestions in the “Code” section of your repository.