|
1
|
|
|
<?php |
|
2
|
|
|
|
|
3
|
|
|
/** |
|
4
|
|
|
* @file |
|
5
|
|
|
* This file is part of the PdfParser library. |
|
6
|
|
|
* |
|
7
|
|
|
* @author Sébastien MALOT <[email protected]> |
|
8
|
|
|
* @date 2017-01-03 |
|
9
|
|
|
* |
|
10
|
|
|
* @license LGPLv3 |
|
11
|
|
|
* @url <https://github.com/smalot/pdfparser> |
|
12
|
|
|
* |
|
13
|
|
|
* PdfParser is a pdf library written in PHP, extraction oriented. |
|
14
|
|
|
* Copyright (C) 2017 - Sébastien MALOT <[email protected]> |
|
15
|
|
|
* |
|
16
|
|
|
* This program is free software: you can redistribute it and/or modify |
|
17
|
|
|
* it under the terms of the GNU Lesser General Public License as published by |
|
18
|
|
|
* the Free Software Foundation, either version 3 of the License, or |
|
19
|
|
|
* (at your option) any later version. |
|
20
|
|
|
* |
|
21
|
|
|
* This program is distributed in the hope that it will be useful, |
|
22
|
|
|
* but WITHOUT ANY WARRANTY; without even the implied warranty of |
|
23
|
|
|
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the |
|
24
|
|
|
* GNU Lesser General Public License for more details. |
|
25
|
|
|
* |
|
26
|
|
|
* You should have received a copy of the GNU Lesser General Public License |
|
27
|
|
|
* along with this program. |
|
28
|
|
|
* If not, see <http://www.pdfparser.org/sites/default/LICENSE.txt>. |
|
29
|
|
|
*/ |
|
30
|
|
|
|
|
31
|
|
|
namespace Smalot\PdfParser; |
|
32
|
|
|
|
|
33
|
|
|
use Smalot\PdfParser\Element\ElementArray; |
|
34
|
|
|
use Smalot\PdfParser\Element\ElementBoolean; |
|
35
|
|
|
use Smalot\PdfParser\Element\ElementDate; |
|
36
|
|
|
use Smalot\PdfParser\Element\ElementHexa; |
|
37
|
|
|
use Smalot\PdfParser\Element\ElementName; |
|
38
|
|
|
use Smalot\PdfParser\Element\ElementNull; |
|
39
|
|
|
use Smalot\PdfParser\Element\ElementNumeric; |
|
40
|
|
|
use Smalot\PdfParser\Element\ElementString; |
|
41
|
|
|
use Smalot\PdfParser\Element\ElementStruct; |
|
42
|
|
|
use Smalot\PdfParser\Element\ElementXRef; |
|
43
|
|
|
|
|
44
|
|
|
/** |
|
45
|
|
|
* Class Element |
|
46
|
|
|
*/ |
|
47
|
|
|
class Element |
|
48
|
|
|
{ |
|
49
|
|
|
/** |
|
50
|
|
|
* @var Document |
|
51
|
|
|
*/ |
|
52
|
|
|
protected $document = null; |
|
53
|
|
|
|
|
54
|
|
|
protected $value = null; |
|
55
|
|
|
|
|
56
|
|
|
/** |
|
57
|
|
|
* @param Document $document |
|
58
|
|
|
*/ |
|
59
|
86 |
|
public function __construct($value, Document $document = null) |
|
60
|
|
|
{ |
|
61
|
86 |
|
$this->value = $value; |
|
62
|
86 |
|
$this->document = $document; |
|
63
|
86 |
|
} |
|
64
|
|
|
|
|
65
|
30 |
|
public function init() |
|
66
|
|
|
{ |
|
67
|
30 |
|
} |
|
68
|
|
|
|
|
69
|
|
|
/** |
|
70
|
|
|
* @return bool |
|
71
|
|
|
*/ |
|
72
|
7 |
|
public function equals($value) |
|
73
|
|
|
{ |
|
74
|
7 |
|
return $value == $this->value; |
|
75
|
|
|
} |
|
76
|
|
|
|
|
77
|
|
|
/** |
|
78
|
|
|
* @return bool |
|
79
|
|
|
*/ |
|
80
|
11 |
|
public function contains($value) |
|
81
|
|
|
{ |
|
82
|
11 |
|
if (\is_array($this->value)) { |
|
83
|
|
|
/** @var Element $val */ |
|
84
|
4 |
|
foreach ($this->value as $val) { |
|
85
|
4 |
|
if ($val->equals($value)) { |
|
86
|
4 |
|
return true; |
|
87
|
|
|
} |
|
88
|
|
|
} |
|
89
|
|
|
|
|
90
|
3 |
|
return false; |
|
91
|
|
|
} |
|
92
|
|
|
|
|
93
|
7 |
|
return $this->equals($value); |
|
94
|
|
|
} |
|
95
|
|
|
|
|
96
|
57 |
|
public function getContent() |
|
97
|
|
|
{ |
|
98
|
57 |
|
return $this->value; |
|
99
|
|
|
} |
|
100
|
|
|
|
|
101
|
|
|
/** |
|
102
|
|
|
* @return string |
|
103
|
|
|
*/ |
|
104
|
37 |
|
public function __toString() |
|
105
|
|
|
{ |
|
106
|
37 |
|
return (string) ($this->value); |
|
107
|
|
|
} |
|
108
|
|
|
|
|
109
|
|
|
/** |
|
110
|
|
|
* @param string $content |
|
111
|
|
|
* @param Document $document |
|
112
|
|
|
* @param int $position |
|
113
|
|
|
* |
|
114
|
|
|
* @return array |
|
115
|
|
|
* |
|
116
|
|
|
* @throws \Exception |
|
117
|
|
|
*/ |
|
118
|
21 |
|
public static function parse($content, Document $document = null, &$position = 0) |
|
119
|
|
|
{ |
|
120
|
21 |
|
$args = \func_get_args(); |
|
121
|
21 |
|
$only_values = isset($args[3]) ? $args[3] : false; |
|
122
|
21 |
|
$content = trim($content); |
|
123
|
21 |
|
$values = []; |
|
124
|
|
|
|
|
125
|
|
|
do { |
|
126
|
21 |
|
$old_position = $position; |
|
127
|
|
|
|
|
128
|
21 |
|
if (!$only_values) { |
|
129
|
19 |
|
if (!preg_match('/^\s*(?P<name>\/[A-Z0-9\._]+)(?P<value>.*)/si', substr($content, $position), $match)) { |
|
130
|
2 |
|
break; |
|
131
|
|
|
} else { |
|
132
|
19 |
|
$name = ltrim($match['name'], '/'); |
|
133
|
19 |
|
$value = $match['value']; |
|
134
|
19 |
|
$position = strpos($content, $value, $position + \strlen($match['name'])); |
|
135
|
|
|
} |
|
136
|
|
|
} else { |
|
137
|
11 |
|
$name = \count($values); |
|
138
|
11 |
|
$value = substr($content, $position); |
|
139
|
|
|
} |
|
140
|
|
|
|
|
141
|
21 |
|
if ($element = ElementName::parse($value, $document, $position)) { |
|
142
|
19 |
|
$values[$name] = $element; |
|
143
|
15 |
|
} elseif ($element = ElementXRef::parse($value, $document, $position)) { |
|
144
|
12 |
|
$values[$name] = $element; |
|
145
|
11 |
|
} elseif ($element = ElementNumeric::parse($value, $document, $position)) { |
|
146
|
9 |
|
$values[$name] = $element; |
|
147
|
9 |
|
} elseif ($element = ElementStruct::parse($value, $document, $position)) { |
|
148
|
7 |
|
$values[$name] = $element; |
|
149
|
9 |
|
} elseif ($element = ElementBoolean::parse($value, $document, $position)) { |
|
150
|
2 |
|
$values[$name] = $element; |
|
151
|
9 |
|
} elseif ($element = ElementNull::parse($value, $document, $position)) { |
|
152
|
1 |
|
$values[$name] = $element; |
|
153
|
9 |
|
} elseif ($element = ElementDate::parse($value, $document, $position)) { |
|
154
|
4 |
|
$values[$name] = $element; |
|
155
|
9 |
|
} elseif ($element = ElementString::parse($value, $document, $position)) { |
|
156
|
7 |
|
$values[$name] = $element; |
|
157
|
8 |
|
} elseif ($element = ElementHexa::parse($value, $document, $position)) { |
|
158
|
1 |
|
$values[$name] = $element; |
|
159
|
8 |
|
} elseif ($element = ElementArray::parse($value, $document, $position)) { |
|
160
|
8 |
|
$values[$name] = $element; |
|
161
|
|
|
} else { |
|
162
|
3 |
|
$position = $old_position; |
|
163
|
3 |
|
break; |
|
164
|
|
|
} |
|
165
|
21 |
|
} while ($position < \strlen($content)); |
|
166
|
|
|
|
|
167
|
21 |
|
return $values; |
|
168
|
|
|
} |
|
169
|
|
|
} |
|
170
|
|
|
|
smalot /
pdfparser
