|
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
|
|
|
* @license LGPLv3 |
|
10
|
|
|
* @url <https://github.com/smalot/pdfparser> |
|
11
|
|
|
* |
|
12
|
|
|
* PdfParser is a pdf library written in PHP, extraction oriented. |
|
13
|
|
|
* Copyright (C) 2017 - Sébastien MALOT <[email protected]> |
|
14
|
|
|
* |
|
15
|
|
|
* This program is free software: you can redistribute it and/or modify |
|
16
|
|
|
* it under the terms of the GNU Lesser General Public License as published by |
|
17
|
|
|
* the Free Software Foundation, either version 3 of the License, or |
|
18
|
|
|
* (at your option) any later version. |
|
19
|
|
|
* |
|
20
|
|
|
* This program is distributed in the hope that it will be useful, |
|
21
|
|
|
* but WITHOUT ANY WARRANTY; without even the implied warranty of |
|
22
|
|
|
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the |
|
23
|
|
|
* GNU Lesser General Public License for more details. |
|
24
|
|
|
* |
|
25
|
|
|
* You should have received a copy of the GNU Lesser General Public License |
|
26
|
|
|
* along with this program. |
|
27
|
|
|
* If not, see <http://www.pdfparser.org/sites/default/LICENSE.txt>. |
|
28
|
|
|
* |
|
29
|
|
|
*/ |
|
30
|
|
|
|
|
31
|
|
|
namespace Smalot\PdfParser; |
|
32
|
|
|
|
|
33
|
|
|
use Smalot\PdfParser\Element\ElementArray; |
|
34
|
|
|
use Smalot\PdfParser\Element\ElementMissing; |
|
35
|
|
|
use Smalot\PdfParser\Element\ElementXRef; |
|
36
|
|
|
use Smalot\PdfParser\Element\ElementNull; |
|
37
|
|
|
|
|
38
|
|
|
/** |
|
39
|
|
|
* Class Page |
|
40
|
|
|
* |
|
41
|
|
|
* @package Smalot\PdfParser |
|
42
|
|
|
*/ |
|
43
|
|
|
class Page extends PDFObject |
|
44
|
|
|
{ |
|
45
|
|
|
/** |
|
46
|
|
|
* @var Font[] |
|
47
|
|
|
*/ |
|
48
|
|
|
protected $fonts = null; |
|
49
|
|
|
|
|
50
|
|
|
/** |
|
51
|
|
|
* @var PDFObject[] |
|
52
|
|
|
*/ |
|
53
|
|
|
protected $xobjects = null; |
|
54
|
|
|
|
|
55
|
|
|
/** |
|
56
|
|
|
* @return Font[] |
|
57
|
|
|
*/ |
|
58
|
|
|
public function getFonts() |
|
59
|
|
|
{ |
|
60
|
|
|
if (!is_null($this->fonts)) { |
|
|
|
|
|
|
61
|
|
|
return $this->fonts; |
|
62
|
|
|
} |
|
63
|
|
|
|
|
64
|
|
|
$resources = $this->get('Resources'); |
|
65
|
|
|
|
|
66
|
|
|
if (method_exists($resources, 'has') && $resources->has('Font')) { |
|
67
|
|
|
|
|
68
|
|
|
if ($resources->get('Font') instanceof Header) { |
|
69
|
|
|
$fonts = $resources->get('Font')->getElements(); |
|
70
|
|
|
} else { |
|
71
|
|
|
$fonts = $resources->get('Font')->getHeader()->getElements(); |
|
72
|
|
|
} |
|
73
|
|
|
|
|
74
|
|
|
$table = array(); |
|
75
|
|
|
|
|
76
|
|
|
foreach ($fonts as $id => $font) { |
|
77
|
|
|
if ($font instanceof Font) { |
|
78
|
|
|
$table[$id] = $font; |
|
79
|
|
|
|
|
80
|
|
|
// Store too on cleaned id value (only numeric) |
|
81
|
|
|
$id = preg_replace('/[^0-9\.\-_]/', '', $id); |
|
82
|
|
|
if ($id != '') { |
|
83
|
|
|
$table[$id] = $font; |
|
84
|
|
|
} |
|
85
|
|
|
} |
|
86
|
|
|
} |
|
87
|
|
|
|
|
88
|
|
|
return ($this->fonts = $table); |
|
89
|
|
|
} else { |
|
90
|
|
|
return array(); |
|
91
|
|
|
} |
|
92
|
|
|
} |
|
93
|
|
|
|
|
94
|
|
|
/** |
|
95
|
|
|
* @param string $id |
|
96
|
|
|
* |
|
97
|
|
|
* @return Font |
|
98
|
|
|
*/ |
|
99
|
|
|
public function getFont($id) |
|
100
|
|
|
{ |
|
101
|
|
|
$fonts = $this->getFonts(); |
|
102
|
|
|
|
|
103
|
|
|
if (isset($fonts[$id])) { |
|
104
|
|
|
return $fonts[$id]; |
|
105
|
|
|
} else { |
|
106
|
|
|
$id = preg_replace('/[^0-9\.\-_]/', '', $id); |
|
107
|
|
|
|
|
108
|
|
|
if (isset($fonts[$id])) { |
|
109
|
|
|
return $fonts[$id]; |
|
110
|
|
|
} else { |
|
111
|
|
|
return null; |
|
112
|
|
|
} |
|
113
|
|
|
} |
|
114
|
|
|
} |
|
115
|
|
|
|
|
116
|
|
|
/** |
|
117
|
|
|
* Support for XObject |
|
118
|
|
|
* |
|
119
|
|
|
* @return PDFObject[] |
|
120
|
|
|
*/ |
|
121
|
|
|
public function getXObjects() |
|
122
|
|
|
{ |
|
123
|
|
|
if (!is_null($this->xobjects)) { |
|
|
|
|
|
|
124
|
|
|
return $this->xobjects; |
|
125
|
|
|
} |
|
126
|
|
|
|
|
127
|
|
|
$resources = $this->get('Resources'); |
|
128
|
|
|
|
|
129
|
|
|
if (method_exists($resources, 'has') && $resources->has('XObject')) { |
|
130
|
|
|
|
|
131
|
|
|
if ($resources->get('XObject') instanceof Header) { |
|
132
|
|
|
$xobjects = $resources->get('XObject')->getElements(); |
|
133
|
|
|
} else { |
|
134
|
|
|
$xobjects = $resources->get('XObject')->getHeader()->getElements(); |
|
135
|
|
|
} |
|
136
|
|
|
|
|
137
|
|
|
$table = array(); |
|
138
|
|
|
|
|
139
|
|
|
foreach ($xobjects as $id => $xobject) { |
|
140
|
|
|
$table[$id] = $xobject; |
|
141
|
|
|
|
|
142
|
|
|
// Store too on cleaned id value (only numeric) |
|
143
|
|
|
$id = preg_replace('/[^0-9\.\-_]/', '', $id); |
|
144
|
|
|
if ($id != '') { |
|
145
|
|
|
$table[$id] = $xobject; |
|
146
|
|
|
} |
|
147
|
|
|
} |
|
148
|
|
|
|
|
149
|
|
|
return ($this->xobjects = $table); |
|
150
|
|
|
} else { |
|
151
|
|
|
return array(); |
|
152
|
|
|
} |
|
153
|
|
|
} |
|
154
|
|
|
|
|
155
|
|
|
/** |
|
156
|
|
|
* @param string $id |
|
157
|
|
|
* |
|
158
|
|
|
* @return PDFObject |
|
159
|
|
|
*/ |
|
160
|
|
|
public function getXObject($id) |
|
161
|
|
|
{ |
|
162
|
|
|
$xobjects = $this->getXObjects(); |
|
163
|
|
|
|
|
164
|
|
|
if (isset($xobjects[$id])) { |
|
165
|
|
|
return $xobjects[$id]; |
|
166
|
|
|
} else { |
|
167
|
|
|
return null; |
|
168
|
|
|
/*$id = preg_replace('/[^0-9\.\-_]/', '', $id); |
|
169
|
|
|
|
|
170
|
|
|
if (isset($xobjects[$id])) { |
|
171
|
|
|
return $xobjects[$id]; |
|
172
|
|
|
} else { |
|
173
|
|
|
return null; |
|
174
|
|
|
}*/ |
|
175
|
|
|
} |
|
176
|
|
|
} |
|
177
|
|
|
|
|
178
|
|
|
/** |
|
179
|
|
|
* @param Page |
|
180
|
|
|
* |
|
181
|
|
|
* @return string |
|
182
|
|
|
*/ |
|
183
|
|
|
public function getText(Page $page = null) |
|
184
|
|
|
{ |
|
185
|
|
|
if ($contents = $this->get('Contents')) { |
|
186
|
|
|
|
|
187
|
|
|
if ($contents instanceof ElementMissing) { |
|
188
|
|
|
return ''; |
|
189
|
|
|
} elseif ($contents instanceof ElementNull) { |
|
190
|
|
|
return ''; |
|
191
|
|
|
} elseif ($contents instanceof PDFObject) { |
|
|
|
|
|
|
192
|
|
|
$elements = $contents->getHeader()->getElements(); |
|
193
|
|
|
|
|
194
|
|
|
if (is_numeric(key($elements))) { |
|
195
|
|
|
$new_content = ''; |
|
196
|
|
|
|
|
197
|
|
|
foreach ($elements as $element) { |
|
198
|
|
|
if ($element instanceof ElementXRef) { |
|
199
|
|
|
$new_content .= $element->getObject()->getContent(); |
|
200
|
|
|
} else { |
|
201
|
|
|
$new_content .= $element->getContent(); |
|
202
|
|
|
} |
|
203
|
|
|
} |
|
204
|
|
|
|
|
205
|
|
|
$header = new Header(array(), $this->document); |
|
206
|
|
|
$contents = new PDFObject($this->document, $header, $new_content); |
|
207
|
|
|
} |
|
208
|
|
|
} elseif ($contents instanceof ElementArray) { |
|
209
|
|
|
// Create a virtual global content. |
|
210
|
|
|
$new_content = ''; |
|
211
|
|
|
|
|
212
|
|
|
foreach ($contents->getContent() as $content) { |
|
213
|
|
|
$new_content .= $content->getContent() . "\n"; |
|
214
|
|
|
} |
|
215
|
|
|
|
|
216
|
|
|
$header = new Header(array(), $this->document); |
|
217
|
|
|
$contents = new PDFObject($this->document, $header, $new_content); |
|
218
|
|
|
} |
|
219
|
|
|
|
|
220
|
|
|
return $contents->getText($this); |
|
|
|
|
|
|
221
|
|
|
} |
|
222
|
|
|
|
|
223
|
|
|
return ''; |
|
224
|
|
|
} |
|
225
|
|
|
|
|
226
|
|
|
/** |
|
227
|
|
|
* @param Page |
|
228
|
|
|
* |
|
229
|
|
|
* @return array |
|
230
|
|
|
*/ |
|
231
|
|
|
public function getTextArray(Page $page = null) |
|
232
|
|
|
{ |
|
233
|
|
|
if ($contents = $this->get('Contents')) { |
|
234
|
|
|
|
|
235
|
|
|
if ($contents instanceof ElementMissing) { |
|
236
|
|
|
return array(); |
|
237
|
|
|
} elseif ($contents instanceof ElementNull) { |
|
238
|
|
|
return array(); |
|
239
|
|
|
} elseif ($contents instanceof PDFObject) { |
|
|
|
|
|
|
240
|
|
|
$elements = $contents->getHeader()->getElements(); |
|
241
|
|
|
|
|
242
|
|
|
if (is_numeric(key($elements))) { |
|
243
|
|
|
$new_content = ''; |
|
244
|
|
|
|
|
245
|
|
|
/** @var PDFObject $element */ |
|
246
|
|
|
foreach ($elements as $element) { |
|
247
|
|
|
if ($element instanceof ElementXRef) { |
|
248
|
|
|
$new_content .= $element->getObject()->getContent(); |
|
249
|
|
|
} else { |
|
250
|
|
|
$new_content .= $element->getContent(); |
|
251
|
|
|
} |
|
252
|
|
|
} |
|
253
|
|
|
|
|
254
|
|
|
$header = new Header(array(), $this->document); |
|
255
|
|
|
$contents = new PDFObject($this->document, $header, $new_content); |
|
256
|
|
|
} |
|
257
|
|
|
} elseif ($contents instanceof ElementArray) { |
|
258
|
|
|
// Create a virtual global content. |
|
259
|
|
|
$new_content = ''; |
|
260
|
|
|
|
|
261
|
|
|
/** @var PDFObject $content */ |
|
262
|
|
|
foreach ($contents->getContent() as $content) { |
|
263
|
|
|
$new_content .= $content->getContent() . "\n"; |
|
264
|
|
|
} |
|
265
|
|
|
|
|
266
|
|
|
$header = new Header(array(), $this->document); |
|
267
|
|
|
$contents = new PDFObject($this->document, $header, $new_content); |
|
268
|
|
|
} |
|
269
|
|
|
|
|
270
|
|
|
return $contents->getTextArray($this); |
|
|
|
|
|
|
271
|
|
|
} |
|
272
|
|
|
|
|
273
|
|
|
return array(); |
|
274
|
|
|
} |
|
275
|
|
|
} |
|
276
|
|
|
|
smalot /
pdfparser
Loading history...
