|
1
|
|
|
<?php |
|
2
|
|
|
|
|
3
|
|
|
/** |
|
4
|
|
|
* @file This file is part of the PdfParser library. |
|
5
|
|
|
* |
|
6
|
|
|
* @author Konrad Abicht <[email protected]> |
|
7
|
|
|
* |
|
8
|
|
|
* @date 2020-06-02 |
|
9
|
|
|
* |
|
10
|
|
|
* @author Sébastien MALOT <[email protected]> |
|
11
|
|
|
* |
|
12
|
|
|
* @date 2017-01-03 |
|
13
|
|
|
* |
|
14
|
|
|
* @license LGPLv3 |
|
15
|
|
|
* |
|
16
|
|
|
* @url <https://github.com/smalot/pdfparser> |
|
17
|
|
|
* |
|
18
|
|
|
* PdfParser is a pdf library written in PHP, extraction oriented. |
|
19
|
|
|
* Copyright (C) 2017 - Sébastien MALOT <[email protected]> |
|
20
|
|
|
* |
|
21
|
|
|
* This program is free software: you can redistribute it and/or modify |
|
22
|
|
|
* it under the terms of the GNU Lesser General Public License as published by |
|
23
|
|
|
* the Free Software Foundation, either version 3 of the License, or |
|
24
|
|
|
* (at your option) any later version. |
|
25
|
|
|
* |
|
26
|
|
|
* This program is distributed in the hope that it will be useful, |
|
27
|
|
|
* but WITHOUT ANY WARRANTY; without even the implied warranty of |
|
28
|
|
|
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the |
|
29
|
|
|
* GNU Lesser General Public License for more details. |
|
30
|
|
|
* |
|
31
|
|
|
* You should have received a copy of the GNU Lesser General Public License |
|
32
|
|
|
* along with this program. |
|
33
|
|
|
* If not, see <http://www.pdfparser.org/sites/default/LICENSE.txt>. |
|
34
|
|
|
*/ |
|
35
|
|
|
|
|
36
|
|
|
namespace PHPUnitTests\Integration\Element; |
|
37
|
|
|
|
|
38
|
|
|
use PHPUnitTests\TestCase; |
|
39
|
|
|
use Smalot\PdfParser\Element\ElementDate; |
|
40
|
|
|
use Smalot\PdfParser\Element\ElementHexa; |
|
41
|
|
|
use Smalot\PdfParser\Element\ElementString; |
|
42
|
|
|
|
|
43
|
|
|
class ElementHexaTest extends TestCase |
|
44
|
|
|
{ |
|
45
|
|
|
public function testParse(): void |
|
46
|
|
|
{ |
|
47
|
|
|
// Skipped. |
|
48
|
|
|
$offset = 0; |
|
49
|
|
|
$element = ElementHexa::parse('ABC', null, $offset); |
|
50
|
|
|
$this->assertFalse($element); |
|
51
|
|
|
$this->assertEquals(0, $offset); |
|
52
|
|
|
|
|
53
|
|
|
$offset = 0; |
|
54
|
|
|
$element = ElementHexa::parse(' [ <0020> 5 6 ]', null, $offset); |
|
55
|
|
|
$this->assertFalse($element); |
|
56
|
|
|
$this->assertEquals(0, $offset); |
|
57
|
|
|
|
|
58
|
|
|
$offset = 0; |
|
59
|
|
|
$element = ElementHexa::parse(' << <0020> >>', null, $offset); |
|
60
|
|
|
$this->assertFalse($element); |
|
61
|
|
|
$this->assertEquals(0, $offset); |
|
62
|
|
|
|
|
63
|
|
|
$offset = 0; |
|
64
|
|
|
$element = ElementHexa::parse(' / <0020> ', null, $offset); |
|
65
|
|
|
$this->assertFalse($element); |
|
66
|
|
|
$this->assertEquals(0, $offset); |
|
67
|
|
|
|
|
68
|
|
|
$offset = 0; |
|
69
|
|
|
$element = ElementHexa::parse(' 0 <0020> ', null, $offset); |
|
70
|
|
|
$this->assertFalse($element); |
|
71
|
|
|
$this->assertEquals(0, $offset); |
|
72
|
|
|
|
|
73
|
|
|
$offset = 0; |
|
74
|
|
|
$element = ElementHexa::parse(" 0 \n <0020> ", null, $offset); |
|
75
|
|
|
$this->assertFalse($element); |
|
76
|
|
|
$this->assertEquals(0, $offset); |
|
77
|
|
|
|
|
78
|
|
|
// Valid. |
|
79
|
|
|
$offset = 0; |
|
80
|
|
|
$element = ElementHexa::parse(' <0020> ', null, $offset); |
|
81
|
|
|
$this->assertEquals(' ', $element->getContent()); |
|
82
|
|
|
$this->assertEquals(7, $offset); |
|
83
|
|
|
|
|
84
|
|
|
$offset = 0; |
|
85
|
|
|
$element = ElementHexa::parse(' <0020> ', null, $offset); |
|
86
|
|
|
$this->assertEquals(' ', $element->getContent()); |
|
87
|
|
|
$this->assertEquals(7, $offset); |
|
88
|
|
|
|
|
89
|
|
|
$offset = 0; |
|
90
|
|
|
$element = ElementHexa::parse(' <0020>', null, $offset); |
|
91
|
|
|
$this->assertEquals(' ', $element->getContent()); |
|
92
|
|
|
$this->assertEquals(7, $offset); |
|
93
|
|
|
|
|
94
|
|
|
$offset = 0; |
|
95
|
|
|
$element = ElementHexa::parse('<0020>', null, $offset); |
|
96
|
|
|
$this->assertEquals(' ', $element->getContent()); |
|
97
|
|
|
$this->assertEquals(6, $offset); |
|
98
|
|
|
|
|
99
|
|
|
$offset = 0; |
|
100
|
|
|
$element = ElementHexa::parse(" \n <0020> ", null, $offset); |
|
101
|
|
|
$this->assertEquals(' ', $element->getContent()); |
|
102
|
|
|
$this->assertEquals(9, $offset); |
|
103
|
|
|
|
|
104
|
|
|
$offset = 0; |
|
105
|
|
|
$element = ElementHexa::parse(" \n <5465616d204d616e6167656d656e742053797374656d73> ", null, $offset); |
|
106
|
|
|
$this->assertEquals('Team Management Systems', $element->getContent()); |
|
107
|
|
|
$this->assertEquals(51, $offset); |
|
108
|
|
|
|
|
109
|
|
|
$offset = 0; |
|
110
|
|
|
$element = ElementHexa::parse(" \n <5265706f72744275696c646572> ", null, $offset); |
|
111
|
|
|
$this->assertTrue($element instanceof ElementString); |
|
112
|
|
|
$this->assertEquals('ReportBuilder', $element->getContent()); |
|
113
|
|
|
$this->assertEquals(31, $offset); |
|
114
|
|
|
|
|
115
|
|
|
$offset = 0; |
|
116
|
|
|
$element = ElementHexa::parse(" \n <443a3230313331323137313334303435303027303027> ", null, $offset); |
|
117
|
|
|
$this->assertTrue($element instanceof ElementDate); |
|
118
|
|
|
$this->assertEquals('2013-12-17T13:40:45+00:00', (string) $element); |
|
119
|
|
|
$this->assertEquals(49, $offset); |
|
120
|
|
|
|
|
121
|
|
|
// Test that a hexadecimal string 'dirty' with extra characters |
|
122
|
|
|
// such as newlines or spaces is properly decoded |
|
123
|
|
|
$element = ElementHexa::decode(' <feff007000610073007100750061002c0020007000720069006d00610076006500720061002c0020 |
|
124
|
|
|
00720065007 30075007200720065007a0069006f006e0065002c0020006600650073007400610020 |
|
125
|
|
|
0063007200690073007400690061006e0061002c002000670065007300f9002c00200075006f0076 |
|
126
|
|
|
0061002000640069 & 002000630069006f00630063006f006c00610074006100 Y 2c00200063006f006e |
|
127
|
|
|
00690067006c00690065007400740069002c0020007000750 / 06c00630069006e0069002c00200070 |
|
128
|
|
|
00610073007100750061006c0065 002c002000630061006d00700061006e0065002c002000640069 |
|
129
|
|
|
006e006100200072006500620075006300630069002c00200075006f007600610020006400690020 |
|
130
|
|
|
007000610 073007100750061002c0020> '); |
|
131
|
|
|
|
|
132
|
|
|
$this->assertEquals('pasqua, primavera, resurrezione, festa cristiana, gesù, uova di cioccolata, coniglietti, pulcini, pasquale, campane, dina rebucci, uova di pasqua, ', $element); |
|
133
|
|
|
} |
|
134
|
|
|
|
|
135
|
|
|
/** |
|
136
|
|
|
* Closing round bracket encoded in hexadecimal format breaks parsing - string is truncated. |
|
137
|
|
|
* |
|
138
|
|
|
* @see https://github.com/smalot/pdfparser/issues/715 |
|
139
|
|
|
*/ |
|
140
|
|
|
public function testIssue715(): void |
|
141
|
|
|
{ |
|
142
|
|
|
$offset = 0; |
|
143
|
|
|
$testString = '()\\'; |
|
144
|
|
|
$element = ElementHexa::parse('<'.bin2hex($testString).'>', null, $offset); |
|
145
|
|
|
$this->assertEquals($testString, (string) $element); |
|
146
|
|
|
} |
|
147
|
|
|
} |
|
148
|
|
|
|
smalot /
pdfparser
