1
|
|
|
<?php |
2
|
|
|
|
3
|
|
|
/** |
4
|
|
|
* This file is based on code of tecnickcom/TCPDF PDF library. |
5
|
|
|
* |
6
|
|
|
* Original author Nicola Asuni ([email protected]) and |
7
|
|
|
* contributors (https://github.com/tecnickcom/TCPDF/graphs/contributors). |
8
|
|
|
* |
9
|
|
|
* @see https://github.com/tecnickcom/TCPDF |
10
|
|
|
* |
11
|
|
|
* Original code was licensed on the terms of the LGPL v3. |
12
|
|
|
* |
13
|
|
|
* ------------------------------------------------------------------------------ |
14
|
|
|
* |
15
|
|
|
* @file This file is part of the PdfParser library. |
16
|
|
|
* |
17
|
|
|
* @author Konrad Abicht <[email protected]> |
18
|
|
|
* @date 2020-01-06 |
19
|
|
|
* |
20
|
|
|
* @license LGPLv3 |
21
|
|
|
* @url <https://github.com/smalot/pdfparser> |
22
|
|
|
* |
23
|
|
|
* PdfParser is a pdf library written in PHP, extraction oriented. |
24
|
|
|
* Copyright (C) 2017 - Sébastien MALOT <[email protected]> |
25
|
|
|
* |
26
|
|
|
* This program is free software: you can redistribute it and/or modify |
27
|
|
|
* it under the terms of the GNU Lesser General Public License as published by |
28
|
|
|
* the Free Software Foundation, either version 3 of the License, or |
29
|
|
|
* (at your option) any later version. |
30
|
|
|
* |
31
|
|
|
* This program is distributed in the hope that it will be useful, |
32
|
|
|
* but WITHOUT ANY WARRANTY; without even the implied warranty of |
33
|
|
|
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the |
34
|
|
|
* GNU Lesser General Public License for more details. |
35
|
|
|
* |
36
|
|
|
* You should have received a copy of the GNU Lesser General Public License |
37
|
|
|
* along with this program. |
38
|
|
|
* If not, see <http://www.pdfparser.org/sites/default/LICENSE.txt>. |
39
|
|
|
*/ |
40
|
|
|
|
41
|
|
|
namespace Smalot\PdfParser\RawData; |
42
|
|
|
|
43
|
|
|
use Exception; |
44
|
|
|
use Smalot\PdfParser\Config; |
45
|
|
|
|
46
|
|
|
class RawDataParser |
47
|
|
|
{ |
48
|
|
|
/** |
49
|
|
|
* @var \Smalot\PdfParser\Config |
50
|
|
|
*/ |
51
|
|
|
private $config; |
52
|
|
|
|
53
|
|
|
/** |
54
|
|
|
* Configuration array. |
55
|
|
|
*/ |
56
|
|
|
protected $cfg = [ |
57
|
|
|
// if `true` ignore filter decoding errors |
58
|
|
|
'ignore_filter_decoding_errors' => true, |
59
|
|
|
// if `true` ignore missing filter decoding errors |
60
|
|
|
'ignore_missing_filter_decoders' => true, |
61
|
|
|
]; |
62
|
|
|
|
63
|
|
|
protected $filterHelper; |
64
|
|
|
protected $objects; |
65
|
|
|
|
66
|
|
|
/** |
67
|
|
|
* @param array $cfg Configuration array, default is [] |
68
|
|
|
*/ |
69
|
33 |
|
public function __construct($cfg = [], Config $config = null) |
70
|
|
|
{ |
71
|
|
|
// merge given array with default values |
72
|
33 |
|
$this->cfg = array_merge($this->cfg, $cfg); |
73
|
|
|
|
74
|
33 |
|
$this->filterHelper = new FilterHelper(); |
75
|
33 |
|
$this->config = $config ?: new Config(); |
76
|
33 |
|
} |
77
|
|
|
|
78
|
|
|
/** |
79
|
|
|
* Decode the specified stream. |
80
|
|
|
* |
81
|
|
|
* @param string $pdfData PDF data |
82
|
|
|
* @param array $xref |
83
|
|
|
* @param array $sdic Stream's dictionary array |
84
|
|
|
* @param string $stream Stream to decode |
85
|
|
|
* |
86
|
|
|
* @return array containing decoded stream data and remaining filters |
87
|
|
|
*/ |
88
|
29 |
|
protected function decodeStream($pdfData, $xref, $sdic, $stream) |
89
|
|
|
{ |
90
|
|
|
// get stream length and filters |
91
|
29 |
|
$slength = \strlen($stream); |
92
|
29 |
|
if ($slength <= 0) { |
93
|
|
|
return ['', []]; |
94
|
|
|
} |
95
|
29 |
|
$filters = []; |
96
|
29 |
|
foreach ($sdic as $k => $v) { |
97
|
29 |
|
if ('/' == $v[0]) { |
98
|
29 |
|
if (('Length' == $v[1]) && (isset($sdic[($k + 1)])) && ('numeric' == $sdic[($k + 1)][0])) { |
99
|
|
|
// get declared stream length |
100
|
28 |
|
$declength = (int) ($sdic[($k + 1)][1]); |
101
|
28 |
|
if ($declength < $slength) { |
102
|
28 |
|
$stream = substr($stream, 0, $declength); |
103
|
28 |
|
$slength = $declength; |
104
|
|
|
} |
105
|
29 |
|
} elseif (('Filter' == $v[1]) && (isset($sdic[($k + 1)]))) { |
106
|
|
|
// resolve indirect object |
107
|
29 |
|
$objval = $this->getObjectVal($pdfData, $xref, $sdic[($k + 1)]); |
108
|
29 |
|
if ('/' == $objval[0]) { |
109
|
|
|
// single filter |
110
|
29 |
|
$filters[] = $objval[1]; |
111
|
1 |
|
} elseif ('[' == $objval[0]) { |
112
|
|
|
// array of filters |
113
|
1 |
|
foreach ($objval[1] as $flt) { |
114
|
1 |
|
if ('/' == $flt[0]) { |
115
|
1 |
|
$filters[] = $flt[1]; |
116
|
|
|
} |
117
|
|
|
} |
118
|
|
|
} |
119
|
|
|
} |
120
|
|
|
} |
121
|
|
|
} |
122
|
|
|
|
123
|
|
|
// decode the stream |
124
|
29 |
|
$remaining_filters = []; |
125
|
29 |
|
foreach ($filters as $filter) { |
126
|
29 |
|
if (\in_array($filter, $this->filterHelper->getAvailableFilters())) { |
127
|
|
|
try { |
128
|
29 |
|
$stream = $this->filterHelper->decodeFilter($filter, $stream); |
129
|
1 |
|
} catch (Exception $e) { |
130
|
1 |
|
$emsg = $e->getMessage(); |
131
|
1 |
|
if ((('~' == $emsg[0]) && !$this->cfg['ignore_missing_filter_decoders']) |
132
|
1 |
|
|| (('~' != $emsg[0]) && !$this->cfg['ignore_filter_decoding_errors']) |
133
|
|
|
) { |
134
|
29 |
|
throw new Exception($e->getMessage()); |
135
|
|
|
} |
136
|
|
|
} |
137
|
|
|
} else { |
138
|
|
|
// add missing filter to array |
139
|
4 |
|
$remaining_filters[] = $filter; |
140
|
|
|
} |
141
|
|
|
} |
142
|
|
|
|
143
|
29 |
|
return [$stream, $remaining_filters]; |
144
|
|
|
} |
145
|
|
|
|
146
|
|
|
/** |
147
|
|
|
* Decode the Cross-Reference section |
148
|
|
|
* |
149
|
|
|
* @param string $pdfData PDF data |
150
|
|
|
* @param int $startxref Offset at which the xref section starts (position of the 'xref' keyword) |
151
|
|
|
* @param array $xref Previous xref array (if any) |
152
|
|
|
* |
153
|
|
|
* @return array containing xref and trailer data |
154
|
|
|
*/ |
155
|
24 |
|
protected function decodeXref($pdfData, $startxref, $xref = []) |
156
|
|
|
{ |
157
|
24 |
|
$startxref += 4; // 4 is the length of the word 'xref' |
158
|
|
|
// skip initial white space chars |
159
|
24 |
|
$offset = $startxref + strspn($pdfData, $this->config->getPdfWhitespaces(), $startxref); |
160
|
|
|
// initialize object number |
161
|
24 |
|
$obj_num = 0; |
162
|
|
|
// search for cross-reference entries or subsection |
163
|
24 |
|
while (preg_match('/([0-9]+)[\x20]([0-9]+)[\x20]?([nf]?)(\r\n|[\x20]?[\r\n])/', $pdfData, $matches, \PREG_OFFSET_CAPTURE, $offset) > 0) { |
164
|
24 |
|
if ($matches[0][1] != $offset) { |
165
|
|
|
// we are on another section |
166
|
5 |
|
break; |
167
|
|
|
} |
168
|
24 |
|
$offset += \strlen($matches[0][0]); |
169
|
24 |
|
if ('n' == $matches[3][0]) { |
170
|
|
|
// create unique object index: [object number]_[generation number] |
171
|
24 |
|
$index = $obj_num.'_'.(int) ($matches[2][0]); |
172
|
|
|
// check if object already exist |
173
|
24 |
|
if (!isset($xref['xref'][$index])) { |
174
|
|
|
// store object offset position |
175
|
24 |
|
$xref['xref'][$index] = (int) ($matches[1][0]); |
176
|
|
|
} |
177
|
24 |
|
++$obj_num; |
178
|
24 |
|
} elseif ('f' == $matches[3][0]) { |
179
|
24 |
|
++$obj_num; |
180
|
|
|
} else { |
181
|
|
|
// object number (index) |
182
|
24 |
|
$obj_num = (int) ($matches[1][0]); |
183
|
|
|
} |
184
|
|
|
} |
185
|
|
|
// get trailer data |
186
|
24 |
|
if (preg_match('/trailer[\s]*<<(.*)>>/isU', $pdfData, $matches, \PREG_OFFSET_CAPTURE, $offset) > 0) { |
187
|
24 |
|
$trailer_data = $matches[1][0]; |
188
|
24 |
|
if (!isset($xref['trailer']) || empty($xref['trailer'])) { |
189
|
|
|
// get only the last updated version |
190
|
24 |
|
$xref['trailer'] = []; |
191
|
|
|
// parse trailer_data |
192
|
24 |
|
if (preg_match('/Size[\s]+([0-9]+)/i', $trailer_data, $matches) > 0) { |
193
|
24 |
|
$xref['trailer']['size'] = (int) ($matches[1]); |
194
|
|
|
} |
195
|
24 |
|
if (preg_match('/Root[\s]+([0-9]+)[\s]+([0-9]+)[\s]+R/i', $trailer_data, $matches) > 0) { |
196
|
24 |
|
$xref['trailer']['root'] = (int) ($matches[1]).'_'.(int) ($matches[2]); |
197
|
|
|
} |
198
|
24 |
|
if (preg_match('/Encrypt[\s]+([0-9]+)[\s]+([0-9]+)[\s]+R/i', $trailer_data, $matches) > 0) { |
199
|
|
|
$xref['trailer']['encrypt'] = (int) ($matches[1]).'_'.(int) ($matches[2]); |
200
|
|
|
} |
201
|
24 |
|
if (preg_match('/Info[\s]+([0-9]+)[\s]+([0-9]+)[\s]+R/i', $trailer_data, $matches) > 0) { |
202
|
24 |
|
$xref['trailer']['info'] = (int) ($matches[1]).'_'.(int) ($matches[2]); |
203
|
|
|
} |
204
|
24 |
|
if (preg_match('/ID[\s]*[\[][\s]*[<]([^>]*)[>][\s]*[<]([^>]*)[>]/i', $trailer_data, $matches) > 0) { |
205
|
22 |
|
$xref['trailer']['id'] = []; |
206
|
22 |
|
$xref['trailer']['id'][0] = $matches[1]; |
207
|
22 |
|
$xref['trailer']['id'][1] = $matches[2]; |
208
|
|
|
} |
209
|
|
|
} |
210
|
24 |
|
if (preg_match('/Prev[\s]+([0-9]+)/i', $trailer_data, $matches) > 0) { |
211
|
|
|
// get previous xref |
212
|
24 |
|
$xref = $this->getXrefData($pdfData, (int) ($matches[1]), $xref); |
213
|
|
|
} |
214
|
|
|
} else { |
215
|
|
|
throw new Exception('Unable to find trailer'); |
216
|
|
|
} |
217
|
|
|
|
218
|
24 |
|
return $xref; |
219
|
|
|
} |
220
|
|
|
|
221
|
|
|
/** |
222
|
|
|
* Decode the Cross-Reference Stream section |
223
|
|
|
* |
224
|
|
|
* @param string $pdfData PDF data |
225
|
|
|
* @param int $startxref Offset at which the xref section starts |
226
|
|
|
* @param array $xref Previous xref array (if any) |
227
|
|
|
* |
228
|
|
|
* @return array containing xref and trailer data |
229
|
|
|
* |
230
|
|
|
* @throws Exception if unknown PNG predictor detected |
231
|
|
|
*/ |
232
|
6 |
|
protected function decodeXrefStream($pdfData, $startxref, $xref = []) |
233
|
|
|
{ |
234
|
|
|
// try to read Cross-Reference Stream |
235
|
6 |
|
$xrefobj = $this->getRawObject($pdfData, $startxref); |
236
|
6 |
|
$xrefcrs = $this->getIndirectObject($pdfData, $xref, $xrefobj[1], $startxref, true); |
237
|
6 |
|
if (!isset($xref['trailer']) || empty($xref['trailer'])) { |
238
|
|
|
// get only the last updated version |
239
|
6 |
|
$xref['trailer'] = []; |
240
|
6 |
|
$filltrailer = true; |
241
|
|
|
} else { |
242
|
4 |
|
$filltrailer = false; |
243
|
|
|
} |
244
|
6 |
|
if (!isset($xref['xref'])) { |
245
|
6 |
|
$xref['xref'] = []; |
246
|
|
|
} |
247
|
6 |
|
$valid_crs = false; |
248
|
6 |
|
$columns = 0; |
249
|
6 |
|
$predictor = null; |
250
|
6 |
|
$sarr = $xrefcrs[0][1]; |
251
|
6 |
|
if (!\is_array($sarr)) { |
252
|
|
|
$sarr = []; |
253
|
|
|
} |
254
|
|
|
|
255
|
6 |
|
$wb = []; |
256
|
|
|
|
257
|
6 |
|
foreach ($sarr as $k => $v) { |
258
|
|
|
if ( |
259
|
6 |
|
('/' == $v[0]) |
260
|
6 |
|
&& ('Type' == $v[1]) |
261
|
|
|
&& ( |
262
|
6 |
|
isset($sarr[($k + 1)]) |
263
|
6 |
|
&& '/' == $sarr[($k + 1)][0] |
264
|
6 |
|
&& 'XRef' == $sarr[($k + 1)][1] |
265
|
|
|
) |
266
|
|
|
) { |
267
|
6 |
|
$valid_crs = true; |
268
|
6 |
|
} elseif (('/' == $v[0]) && ('Index' == $v[1]) && (isset($sarr[($k + 1)]))) { |
269
|
|
|
// first object number in the subsection |
270
|
4 |
|
$index_first = (int) ($sarr[($k + 1)][1][0][1]); |
271
|
6 |
|
} elseif (('/' == $v[0]) && ('Prev' == $v[1]) && (isset($sarr[($k + 1)]) && ('numeric' == $sarr[($k + 1)][0]))) { |
272
|
|
|
// get previous xref offset |
273
|
4 |
|
$prevxref = (int) ($sarr[($k + 1)][1]); |
274
|
6 |
|
} elseif (('/' == $v[0]) && ('W' == $v[1]) && (isset($sarr[($k + 1)]))) { |
275
|
|
|
// number of bytes (in the decoded stream) of the corresponding field |
276
|
6 |
|
$wb[0] = (int) ($sarr[($k + 1)][1][0][1]); |
277
|
6 |
|
$wb[1] = (int) ($sarr[($k + 1)][1][1][1]); |
278
|
6 |
|
$wb[2] = (int) ($sarr[($k + 1)][1][2][1]); |
279
|
6 |
|
} elseif (('/' == $v[0]) && ('DecodeParms' == $v[1]) && (isset($sarr[($k + 1)][1]))) { |
280
|
5 |
|
$decpar = $sarr[($k + 1)][1]; |
281
|
5 |
|
foreach ($decpar as $kdc => $vdc) { |
282
|
|
|
if ( |
283
|
5 |
|
'/' == $vdc[0] |
284
|
5 |
|
&& 'Columns' == $vdc[1] |
285
|
|
|
&& ( |
286
|
5 |
|
isset($decpar[($kdc + 1)]) |
287
|
5 |
|
&& 'numeric' == $decpar[($kdc + 1)][0] |
288
|
|
|
) |
289
|
|
|
) { |
290
|
5 |
|
$columns = (int) ($decpar[($kdc + 1)][1]); |
291
|
|
|
} elseif ( |
292
|
5 |
|
'/' == $vdc[0] |
293
|
5 |
|
&& 'Predictor' == $vdc[1] |
294
|
|
|
&& ( |
295
|
5 |
|
isset($decpar[($kdc + 1)]) |
296
|
5 |
|
&& 'numeric' == $decpar[($kdc + 1)][0] |
297
|
|
|
) |
298
|
|
|
) { |
299
|
5 |
|
$predictor = (int) ($decpar[($kdc + 1)][1]); |
300
|
|
|
} |
301
|
|
|
} |
302
|
6 |
|
} elseif ($filltrailer) { |
303
|
6 |
|
if (('/' == $v[0]) && ('Size' == $v[1]) && (isset($sarr[($k + 1)]) && ('numeric' == $sarr[($k + 1)][0]))) { |
304
|
6 |
|
$xref['trailer']['size'] = $sarr[($k + 1)][1]; |
305
|
6 |
|
} elseif (('/' == $v[0]) && ('Root' == $v[1]) && (isset($sarr[($k + 1)]) && ('objref' == $sarr[($k + 1)][0]))) { |
306
|
6 |
|
$xref['trailer']['root'] = $sarr[($k + 1)][1]; |
307
|
6 |
|
} elseif (('/' == $v[0]) && ('Info' == $v[1]) && (isset($sarr[($k + 1)]) && ('objref' == $sarr[($k + 1)][0]))) { |
308
|
6 |
|
$xref['trailer']['info'] = $sarr[($k + 1)][1]; |
309
|
6 |
|
} elseif (('/' == $v[0]) && ('Encrypt' == $v[1]) && (isset($sarr[($k + 1)]) && ('objref' == $sarr[($k + 1)][0]))) { |
310
|
|
|
$xref['trailer']['encrypt'] = $sarr[($k + 1)][1]; |
311
|
6 |
|
} elseif (('/' == $v[0]) && ('ID' == $v[1]) && (isset($sarr[($k + 1)]))) { |
312
|
6 |
|
$xref['trailer']['id'] = []; |
313
|
6 |
|
$xref['trailer']['id'][0] = $sarr[($k + 1)][1][0][1]; |
314
|
6 |
|
$xref['trailer']['id'][1] = $sarr[($k + 1)][1][1][1]; |
315
|
|
|
} |
316
|
|
|
} |
317
|
|
|
} |
318
|
|
|
|
319
|
|
|
// decode data |
320
|
6 |
|
if ($valid_crs && isset($xrefcrs[1][3][0])) { |
321
|
6 |
|
if (null !== $predictor) { |
322
|
|
|
// number of bytes in a row |
323
|
5 |
|
$rowlen = ($columns + 1); |
324
|
|
|
// convert the stream into an array of integers |
325
|
5 |
|
$sdata = unpack('C*', $xrefcrs[1][3][0]); |
326
|
|
|
// split the rows |
327
|
5 |
|
$sdata = array_chunk($sdata, $rowlen); |
328
|
|
|
|
329
|
|
|
// initialize decoded array |
330
|
5 |
|
$ddata = []; |
331
|
|
|
// initialize first row with zeros |
332
|
5 |
|
$prev_row = array_fill(0, $rowlen, 0); |
333
|
|
|
// for each row apply PNG unpredictor |
334
|
5 |
|
foreach ($sdata as $k => $row) { |
335
|
|
|
// initialize new row |
336
|
5 |
|
$ddata[$k] = []; |
337
|
|
|
// get PNG predictor value |
338
|
5 |
|
$predictor = (10 + $row[0]); |
339
|
|
|
// for each byte on the row |
340
|
5 |
|
for ($i = 1; $i <= $columns; ++$i) { |
341
|
|
|
// new index |
342
|
5 |
|
$j = ($i - 1); |
343
|
5 |
|
$row_up = $prev_row[$j]; |
344
|
5 |
|
if (1 == $i) { |
345
|
5 |
|
$row_left = 0; |
346
|
5 |
|
$row_upleft = 0; |
347
|
|
|
} else { |
348
|
5 |
|
$row_left = $row[($i - 1)]; |
349
|
5 |
|
$row_upleft = $prev_row[($j - 1)]; |
350
|
|
|
} |
351
|
5 |
|
switch ($predictor) { |
352
|
5 |
|
case 10: // PNG prediction (on encoding, PNG None on all rows) |
353
|
|
|
$ddata[$k][$j] = $row[$i]; |
354
|
|
|
break; |
355
|
|
|
|
356
|
5 |
|
case 11: // PNG prediction (on encoding, PNG Sub on all rows) |
357
|
|
|
$ddata[$k][$j] = (($row[$i] + $row_left) & 0xff); |
358
|
|
|
break; |
359
|
|
|
|
360
|
5 |
|
case 12: // PNG prediction (on encoding, PNG Up on all rows) |
361
|
5 |
|
$ddata[$k][$j] = (($row[$i] + $row_up) & 0xff); |
362
|
5 |
|
break; |
363
|
|
|
|
364
|
|
|
case 13: // PNG prediction (on encoding, PNG Average on all rows) |
365
|
|
|
$ddata[$k][$j] = (($row[$i] + (($row_left + $row_up) / 2)) & 0xff); |
366
|
|
|
break; |
367
|
|
|
|
368
|
|
|
case 14: // PNG prediction (on encoding, PNG Paeth on all rows) |
369
|
|
|
// initial estimate |
370
|
|
|
$p = ($row_left + $row_up - $row_upleft); |
371
|
|
|
// distances |
372
|
|
|
$pa = abs($p - $row_left); |
373
|
|
|
$pb = abs($p - $row_up); |
374
|
|
|
$pc = abs($p - $row_upleft); |
375
|
|
|
$pmin = min($pa, $pb, $pc); |
376
|
|
|
// return minimum distance |
377
|
|
|
switch ($pmin) { |
378
|
|
|
case $pa: |
379
|
|
|
$ddata[$k][$j] = (($row[$i] + $row_left) & 0xff); |
380
|
|
|
break; |
381
|
|
|
|
382
|
|
|
case $pb: |
383
|
|
|
$ddata[$k][$j] = (($row[$i] + $row_up) & 0xff); |
384
|
|
|
break; |
385
|
|
|
|
386
|
|
|
case $pc: |
387
|
|
|
$ddata[$k][$j] = (($row[$i] + $row_upleft) & 0xff); |
388
|
|
|
break; |
389
|
|
|
} |
390
|
|
|
break; |
391
|
|
|
|
392
|
|
|
default: // PNG prediction (on encoding, PNG optimum) |
393
|
|
|
throw new Exception('Unknown PNG predictor: '.$predictor); |
394
|
|
|
} |
395
|
|
|
} |
396
|
5 |
|
$prev_row = $ddata[$k]; |
397
|
|
|
} // end for each row |
398
|
|
|
// complete decoding |
399
|
|
|
} else { |
400
|
|
|
// number of bytes in a row |
401
|
2 |
|
$rowlen = array_sum($wb); |
402
|
|
|
// convert the stream into an array of integers |
403
|
2 |
|
$sdata = unpack('C*', $xrefcrs[1][3][0]); |
404
|
|
|
// split the rows |
405
|
2 |
|
$ddata = array_chunk($sdata, $rowlen); |
|
|
|
|
406
|
|
|
} |
407
|
|
|
|
408
|
6 |
|
$sdata = []; |
409
|
|
|
|
410
|
|
|
// for every row |
411
|
6 |
|
foreach ($ddata as $k => $row) { |
412
|
|
|
// initialize new row |
413
|
6 |
|
$sdata[$k] = [0, 0, 0]; |
414
|
6 |
|
if (0 == $wb[0]) { |
415
|
|
|
// default type field |
416
|
|
|
$sdata[$k][0] = 1; |
417
|
|
|
} |
418
|
6 |
|
$i = 0; // count bytes in the row |
419
|
|
|
// for every column |
420
|
6 |
|
for ($c = 0; $c < 3; ++$c) { |
421
|
|
|
// for every byte on the column |
422
|
6 |
|
for ($b = 0; $b < $wb[$c]; ++$b) { |
423
|
6 |
|
if (isset($row[$i])) { |
424
|
6 |
|
$sdata[$k][$c] += ($row[$i] << (($wb[$c] - 1 - $b) * 8)); |
425
|
|
|
} |
426
|
6 |
|
++$i; |
427
|
|
|
} |
428
|
|
|
} |
429
|
|
|
} |
430
|
6 |
|
$ddata = []; |
|
|
|
|
431
|
|
|
// fill xref |
432
|
6 |
|
if (isset($index_first)) { |
433
|
4 |
|
$obj_num = $index_first; |
434
|
|
|
} else { |
435
|
6 |
|
$obj_num = 0; |
436
|
|
|
} |
437
|
6 |
|
foreach ($sdata as $k => $row) { |
438
|
6 |
|
switch ($row[0]) { |
439
|
6 |
|
case 0: // (f) linked list of free objects |
440
|
6 |
|
break; |
441
|
|
|
|
442
|
6 |
|
case 1: // (n) objects that are in use but are not compressed |
443
|
|
|
// create unique object index: [object number]_[generation number] |
444
|
6 |
|
$index = $obj_num.'_'.$row[2]; |
445
|
|
|
// check if object already exist |
446
|
6 |
|
if (!isset($xref['xref'][$index])) { |
447
|
|
|
// store object offset position |
448
|
6 |
|
$xref['xref'][$index] = $row[1]; |
449
|
|
|
} |
450
|
6 |
|
break; |
451
|
|
|
|
452
|
6 |
|
case 2: // compressed objects |
453
|
|
|
// $row[1] = object number of the object stream in which this object is stored |
454
|
|
|
// $row[2] = index of this object within the object stream |
455
|
6 |
|
$index = $row[1].'_0_'.$row[2]; |
456
|
6 |
|
$xref['xref'][$index] = -1; |
457
|
6 |
|
break; |
458
|
|
|
|
459
|
|
|
default: // null objects |
460
|
|
|
break; |
461
|
|
|
} |
462
|
6 |
|
++$obj_num; |
463
|
|
|
} |
464
|
|
|
} // end decoding data |
465
|
6 |
|
if (isset($prevxref)) { |
466
|
|
|
// get previous xref |
467
|
4 |
|
$xref = $this->getXrefData($pdfData, $prevxref, $xref); |
468
|
|
|
} |
469
|
|
|
|
470
|
6 |
|
return $xref; |
471
|
|
|
} |
472
|
|
|
|
473
|
29 |
|
protected function getObjectHeaderPattern($objRefArr): string |
474
|
|
|
{ |
475
|
|
|
// consider all whitespace character (PDF specifications) |
476
|
29 |
|
return '/'.$objRefArr[0].$this->config->getPdfWhitespacesRegex().$objRefArr[1].$this->config->getPdfWhitespacesRegex().'obj'.'/'; |
477
|
|
|
} |
478
|
|
|
|
479
|
29 |
|
protected function getObjectHeaderLen($objRefArr): int |
480
|
|
|
{ |
481
|
|
|
// "4 0 obj" |
482
|
|
|
// 2 whitespaces + strlen("obj") = 5 |
483
|
29 |
|
return 5 + \strlen($objRefArr[0]) + \strlen($objRefArr[1]); |
484
|
|
|
} |
485
|
|
|
|
486
|
|
|
/** |
487
|
|
|
* Get content of indirect object. |
488
|
|
|
* |
489
|
|
|
* @param string $pdfData PDF data |
490
|
|
|
* @param array $xref |
491
|
|
|
* @param string $objRef Object number and generation number separated by underscore character |
492
|
|
|
* @param int $offset Object offset |
493
|
|
|
* @param bool $decoding If true decode streams |
494
|
|
|
* |
495
|
|
|
* @return array containing object data |
496
|
|
|
* |
497
|
|
|
* @throws Exception if invalid object reference found |
498
|
|
|
*/ |
499
|
29 |
|
protected function getIndirectObject($pdfData, $xref, $objRef, $offset = 0, $decoding = true) |
500
|
|
|
{ |
501
|
|
|
/* |
502
|
|
|
* build indirect object header |
503
|
|
|
*/ |
504
|
|
|
// $objHeader = "[object number] [generation number] obj" |
505
|
29 |
|
$objRefArr = explode('_', $objRef); |
506
|
29 |
|
if (2 !== \count($objRefArr)) { |
507
|
|
|
throw new Exception('Invalid object reference for $obj.'); |
508
|
|
|
} |
509
|
|
|
|
510
|
29 |
|
$objHeaderLen = $this->getObjectHeaderLen($objRefArr); |
511
|
|
|
|
512
|
|
|
/* |
513
|
|
|
* check if we are in position |
514
|
|
|
*/ |
515
|
|
|
// ignore whitespace characters at offset |
516
|
29 |
|
$offset += strspn($pdfData, $this->config->getPdfWhitespaces(), $offset); |
517
|
|
|
// ignore leading zeros for object number |
518
|
29 |
|
$offset += strspn($pdfData, '0', $offset); |
519
|
29 |
|
if (0 == preg_match($this->getObjectHeaderPattern($objRefArr), substr($pdfData, $offset, $objHeaderLen))) { |
520
|
|
|
// an indirect reference to an undefined object shall be considered a reference to the null object |
521
|
2 |
|
return ['null', 'null', $offset]; |
522
|
|
|
} |
523
|
|
|
|
524
|
|
|
/* |
525
|
|
|
* get content |
526
|
|
|
*/ |
527
|
|
|
// starting position of object content |
528
|
29 |
|
$offset += $objHeaderLen; |
529
|
29 |
|
$objContentArr = []; |
530
|
29 |
|
$i = 0; // object main index |
531
|
|
|
do { |
532
|
29 |
|
$oldOffset = $offset; |
533
|
|
|
// get element |
534
|
29 |
|
$element = $this->getRawObject($pdfData, $offset); |
535
|
29 |
|
$offset = $element[2]; |
536
|
|
|
// decode stream using stream's dictionary information |
537
|
29 |
|
if ($decoding && ('stream' === $element[0]) && (isset($objContentArr[($i - 1)][0])) && ('<<' === $objContentArr[($i - 1)][0])) { |
538
|
29 |
|
$element[3] = $this->decodeStream($pdfData, $xref, $objContentArr[($i - 1)][1], $element[1]); |
539
|
|
|
} |
540
|
29 |
|
$objContentArr[$i] = $element; |
541
|
29 |
|
++$i; |
542
|
29 |
|
} while (('endobj' !== $element[0]) && ($offset !== $oldOffset)); |
543
|
|
|
// remove closing delimiter |
544
|
29 |
|
array_pop($objContentArr); |
545
|
|
|
|
546
|
|
|
/* |
547
|
|
|
* return raw object content |
548
|
|
|
*/ |
549
|
29 |
|
return $objContentArr; |
550
|
|
|
} |
551
|
|
|
|
552
|
|
|
/** |
553
|
|
|
* Get the content of object, resolving indirect object reference if necessary. |
554
|
|
|
* |
555
|
|
|
* @param string $pdfData PDF data |
556
|
|
|
* @param array $obj Object value |
557
|
|
|
* |
558
|
|
|
* @return array containing object data |
559
|
|
|
* |
560
|
|
|
* @throws Exception |
561
|
|
|
*/ |
562
|
29 |
|
protected function getObjectVal($pdfData, $xref, $obj) |
563
|
|
|
{ |
564
|
29 |
|
if ('objref' == $obj[0]) { |
565
|
|
|
// reference to indirect object |
566
|
|
|
if (isset($this->objects[$obj[1]])) { |
567
|
|
|
// this object has been already parsed |
568
|
|
|
return $this->objects[$obj[1]]; |
569
|
|
|
} elseif (isset($xref[$obj[1]])) { |
570
|
|
|
// parse new object |
571
|
|
|
$this->objects[$obj[1]] = $this->getIndirectObject($pdfData, $xref, $obj[1], $xref[$obj[1]], false); |
572
|
|
|
|
573
|
|
|
return $this->objects[$obj[1]]; |
574
|
|
|
} |
575
|
|
|
} |
576
|
|
|
|
577
|
29 |
|
return $obj; |
578
|
|
|
} |
579
|
|
|
|
580
|
|
|
/** |
581
|
|
|
* Get object type, raw value and offset to next object |
582
|
|
|
* |
583
|
|
|
* @param int $offset Object offset |
584
|
|
|
* |
585
|
|
|
* @return array containing object type, raw value and offset to next object |
586
|
|
|
*/ |
587
|
30 |
|
protected function getRawObject($pdfData, $offset = 0) |
588
|
|
|
{ |
589
|
30 |
|
$objtype = ''; // object type to be returned |
590
|
30 |
|
$objval = ''; // object value to be returned |
591
|
|
|
|
592
|
|
|
// skip initial white space chars |
593
|
30 |
|
$offset += strspn($pdfData, $this->config->getPdfWhitespaces(), $offset); |
594
|
|
|
|
595
|
|
|
// get first char |
596
|
30 |
|
$char = $pdfData[$offset]; |
597
|
|
|
// get object type |
598
|
30 |
|
switch ($char) { |
599
|
30 |
|
case '%': // \x25 PERCENT SIGN |
600
|
|
|
// skip comment and search for next token |
601
|
|
|
$next = strcspn($pdfData, "\r\n", $offset); |
602
|
|
|
if ($next > 0) { |
603
|
|
|
$offset += $next; |
604
|
|
|
|
605
|
|
|
return $this->getRawObject($pdfData, $offset); |
606
|
|
|
} |
607
|
|
|
break; |
608
|
|
|
|
609
|
30 |
|
case '/': // \x2F SOLIDUS |
610
|
|
|
// name object |
611
|
30 |
|
$objtype = $char; |
612
|
30 |
|
++$offset; |
613
|
30 |
|
$pregResult = preg_match( |
614
|
30 |
|
'/^([^\x00\x09\x0a\x0c\x0d\x20\s\x28\x29\x3c\x3e\x5b\x5d\x7b\x7d\x2f\x25]+)/', |
615
|
30 |
|
substr($pdfData, $offset, 256), |
616
|
|
|
$matches |
617
|
|
|
); |
618
|
30 |
|
if (1 == $pregResult) { |
619
|
30 |
|
$objval = $matches[1]; // unescaped value |
620
|
30 |
|
$offset += \strlen($objval); |
621
|
|
|
} |
622
|
30 |
|
break; |
623
|
|
|
|
624
|
30 |
|
case '(': // \x28 LEFT PARENTHESIS |
625
|
30 |
|
case ')': // \x29 RIGHT PARENTHESIS |
626
|
|
|
// literal string object |
627
|
27 |
|
$objtype = $char; |
628
|
27 |
|
++$offset; |
629
|
27 |
|
$strpos = $offset; |
630
|
27 |
|
if ('(' == $char) { |
631
|
27 |
|
$open_bracket = 1; |
632
|
27 |
|
while ($open_bracket > 0) { |
633
|
27 |
|
if (!isset($pdfData[$strpos])) { |
634
|
|
|
break; |
635
|
|
|
} |
636
|
27 |
|
$ch = $pdfData[$strpos]; |
637
|
27 |
|
switch ($ch) { |
638
|
27 |
|
case '\\': // REVERSE SOLIDUS (5Ch) (Backslash) |
639
|
|
|
// skip next character |
640
|
15 |
|
++$strpos; |
641
|
15 |
|
break; |
642
|
|
|
|
643
|
27 |
|
case '(': // LEFT PARENHESIS (28h) |
644
|
|
|
++$open_bracket; |
645
|
|
|
break; |
646
|
|
|
|
647
|
27 |
|
case ')': // RIGHT PARENTHESIS (29h) |
648
|
27 |
|
--$open_bracket; |
649
|
27 |
|
break; |
650
|
|
|
} |
651
|
27 |
|
++$strpos; |
652
|
|
|
} |
653
|
27 |
|
$objval = substr($pdfData, $offset, ($strpos - $offset - 1)); |
654
|
27 |
|
$offset = $strpos; |
655
|
|
|
} |
656
|
27 |
|
break; |
657
|
|
|
|
658
|
30 |
|
case '[': // \x5B LEFT SQUARE BRACKET |
659
|
30 |
|
case ']': // \x5D RIGHT SQUARE BRACKET |
660
|
|
|
// array object |
661
|
29 |
|
$objtype = $char; |
662
|
29 |
|
++$offset; |
663
|
29 |
|
if ('[' == $char) { |
664
|
|
|
// get array content |
665
|
29 |
|
$objval = []; |
666
|
|
|
do { |
667
|
29 |
|
$oldOffset = $offset; |
668
|
|
|
// get element |
669
|
29 |
|
$element = $this->getRawObject($pdfData, $offset); |
670
|
29 |
|
$offset = $element[2]; |
671
|
29 |
|
$objval[] = $element; |
672
|
29 |
|
} while ((']' != $element[0]) && ($offset != $oldOffset)); |
673
|
|
|
// remove closing delimiter |
674
|
29 |
|
array_pop($objval); |
675
|
|
|
} |
676
|
29 |
|
break; |
677
|
|
|
|
678
|
30 |
|
case '<': // \x3C LESS-THAN SIGN |
679
|
30 |
|
case '>': // \x3E GREATER-THAN SIGN |
680
|
30 |
|
if (isset($pdfData[($offset + 1)]) && ($pdfData[($offset + 1)] == $char)) { |
681
|
|
|
// dictionary object |
682
|
30 |
|
$objtype = $char.$char; |
683
|
30 |
|
$offset += 2; |
684
|
30 |
|
if ('<' == $char) { |
685
|
|
|
// get array content |
686
|
30 |
|
$objval = []; |
687
|
|
|
do { |
688
|
30 |
|
$oldOffset = $offset; |
689
|
|
|
// get element |
690
|
30 |
|
$element = $this->getRawObject($pdfData, $offset); |
691
|
30 |
|
$offset = $element[2]; |
692
|
30 |
|
$objval[] = $element; |
693
|
30 |
|
} while (('>>' != $element[0]) && ($offset != $oldOffset)); |
694
|
|
|
// remove closing delimiter |
695
|
30 |
|
array_pop($objval); |
696
|
|
|
} |
697
|
|
|
} else { |
698
|
|
|
// hexadecimal string object |
699
|
12 |
|
$objtype = $char; |
700
|
12 |
|
++$offset; |
701
|
12 |
|
$pregResult = preg_match( |
702
|
12 |
|
'/^([0-9A-Fa-f\x09\x0a\x0c\x0d\x20]+)>/iU', |
703
|
12 |
|
substr($pdfData, $offset), |
704
|
|
|
$matches |
705
|
|
|
); |
706
|
12 |
|
if (('<' == $char) && 1 == $pregResult) { |
707
|
|
|
// remove white space characters |
708
|
12 |
|
$objval = strtr($matches[1], $this->config->getPdfWhitespaces(), ''); |
709
|
12 |
|
$offset += \strlen($matches[0]); |
710
|
|
|
} elseif (false !== ($endpos = strpos($pdfData, '>', $offset))) { |
711
|
|
|
$offset = $endpos + 1; |
712
|
|
|
} |
713
|
|
|
} |
714
|
30 |
|
break; |
715
|
|
|
|
716
|
|
|
default: |
717
|
30 |
|
if ('endobj' == substr($pdfData, $offset, 6)) { |
718
|
|
|
// indirect object |
719
|
29 |
|
$objtype = 'endobj'; |
720
|
29 |
|
$offset += 6; |
721
|
30 |
|
} elseif ('null' == substr($pdfData, $offset, 4)) { |
722
|
|
|
// null object |
723
|
4 |
|
$objtype = 'null'; |
724
|
4 |
|
$offset += 4; |
725
|
4 |
|
$objval = 'null'; |
726
|
30 |
|
} elseif ('true' == substr($pdfData, $offset, 4)) { |
727
|
|
|
// boolean true object |
728
|
9 |
|
$objtype = 'boolean'; |
729
|
9 |
|
$offset += 4; |
730
|
9 |
|
$objval = 'true'; |
731
|
30 |
|
} elseif ('false' == substr($pdfData, $offset, 5)) { |
732
|
|
|
// boolean false object |
733
|
2 |
|
$objtype = 'boolean'; |
734
|
2 |
|
$offset += 5; |
735
|
2 |
|
$objval = 'false'; |
736
|
30 |
|
} elseif ('stream' == substr($pdfData, $offset, 6)) { |
737
|
|
|
// start stream object |
738
|
29 |
|
$objtype = 'stream'; |
739
|
29 |
|
$offset += 6; |
740
|
29 |
|
if (1 == preg_match('/^([\r]?[\n])/isU', substr($pdfData, $offset), $matches)) { |
741
|
29 |
|
$offset += \strlen($matches[0]); |
742
|
29 |
|
$pregResult = preg_match( |
743
|
29 |
|
'/(endstream)[\x09\x0a\x0c\x0d\x20]/isU', |
744
|
29 |
|
substr($pdfData, $offset), |
745
|
|
|
$matches, |
746
|
29 |
|
\PREG_OFFSET_CAPTURE |
747
|
|
|
); |
748
|
29 |
|
if (1 == $pregResult) { |
749
|
29 |
|
$objval = substr($pdfData, $offset, $matches[0][1]); |
750
|
29 |
|
$offset += $matches[1][1]; |
751
|
|
|
} |
752
|
|
|
} |
753
|
30 |
|
} elseif ('endstream' == substr($pdfData, $offset, 9)) { |
754
|
|
|
// end stream object |
755
|
29 |
|
$objtype = 'endstream'; |
756
|
29 |
|
$offset += 9; |
757
|
30 |
|
} elseif (1 == preg_match('/^([0-9]+)[\s]+([0-9]+)[\s]+R/iU', substr($pdfData, $offset, 33), $matches)) { |
758
|
|
|
// indirect object reference |
759
|
29 |
|
$objtype = 'objref'; |
760
|
29 |
|
$offset += \strlen($matches[0]); |
761
|
29 |
|
$objval = (int) ($matches[1]).'_'.(int) ($matches[2]); |
762
|
30 |
|
} elseif (1 == preg_match('/^([0-9]+)[\s]+([0-9]+)[\s]+obj/iU', substr($pdfData, $offset, 33), $matches)) { |
763
|
|
|
// object start |
764
|
6 |
|
$objtype = 'obj'; |
765
|
6 |
|
$objval = (int) ($matches[1]).'_'.(int) ($matches[2]); |
766
|
6 |
|
$offset += \strlen($matches[0]); |
767
|
30 |
|
} elseif (($numlen = strspn($pdfData, '+-.0123456789', $offset)) > 0) { |
768
|
|
|
// numeric object |
769
|
29 |
|
$objtype = 'numeric'; |
770
|
29 |
|
$objval = substr($pdfData, $offset, $numlen); |
771
|
29 |
|
$offset += $numlen; |
772
|
|
|
} |
773
|
30 |
|
break; |
774
|
|
|
} |
775
|
|
|
|
776
|
30 |
|
return [$objtype, $objval, $offset]; |
777
|
|
|
} |
778
|
|
|
|
779
|
|
|
/** |
780
|
|
|
* Get Cross-Reference (xref) table and trailer data from PDF document data. |
781
|
|
|
* |
782
|
|
|
* @param string $pdfData |
783
|
|
|
* @param int $offset xref offset (if know) |
784
|
|
|
* @param array $xref previous xref array (if any) |
785
|
|
|
* |
786
|
|
|
* @return array containing xref and trailer data |
787
|
|
|
* |
788
|
|
|
* @throws Exception if it was unable to find startxref |
789
|
|
|
* @throws Exception if it was unable to find xref |
790
|
|
|
*/ |
791
|
30 |
|
protected function getXrefData($pdfData, $offset = 0, $xref = []) |
792
|
|
|
{ |
793
|
30 |
|
$startxrefPreg = preg_match( |
794
|
30 |
|
'/[\r\n]startxref[\s]*[\r\n]+([0-9]+)[\s]*[\r\n]+%%EOF/i', |
795
|
|
|
$pdfData, |
796
|
|
|
$matches, |
797
|
30 |
|
\PREG_OFFSET_CAPTURE, |
798
|
|
|
$offset |
799
|
|
|
); |
800
|
|
|
|
801
|
30 |
|
if (0 == $offset) { |
802
|
|
|
// find last startxref |
803
|
30 |
|
$pregResult = preg_match_all( |
804
|
30 |
|
'/[\r\n]startxref[\s]*[\r\n]+([0-9]+)[\s]*[\r\n]+%%EOF/i', |
805
|
|
|
$pdfData, $matches, |
806
|
30 |
|
\PREG_SET_ORDER, |
807
|
|
|
$offset |
808
|
|
|
); |
809
|
30 |
|
if (0 == $pregResult) { |
810
|
|
|
throw new Exception('Unable to find startxref'); |
811
|
|
|
} |
812
|
30 |
|
$matches = array_pop($matches); |
813
|
30 |
|
$startxref = $matches[1]; |
814
|
8 |
|
} elseif (strpos($pdfData, 'xref', $offset) == $offset) { |
815
|
|
|
// Already pointing at the xref table |
816
|
5 |
|
$startxref = $offset; |
817
|
4 |
|
} elseif (preg_match('/([0-9]+[\s][0-9]+[\s]obj)/i', $pdfData, $matches, \PREG_OFFSET_CAPTURE, $offset)) { |
818
|
|
|
// Cross-Reference Stream object |
819
|
4 |
|
$startxref = $offset; |
820
|
|
|
} elseif ($startxrefPreg) { |
821
|
|
|
// startxref found |
822
|
|
|
$startxref = $matches[1][0]; |
823
|
|
|
} else { |
824
|
|
|
throw new Exception('Unable to find startxref'); |
825
|
|
|
} |
826
|
|
|
|
827
|
30 |
|
if ($startxref > \strlen($pdfData)) { |
828
|
1 |
|
throw new Exception('Unable to find xref (PDF corrupted?)'); |
829
|
|
|
} |
830
|
|
|
|
831
|
|
|
// check xref position |
832
|
29 |
|
if (strpos($pdfData, 'xref', $startxref) == $startxref) { |
833
|
|
|
// Cross-Reference |
834
|
24 |
|
$xref = $this->decodeXref($pdfData, $startxref, $xref); |
835
|
|
|
} else { |
836
|
|
|
// Cross-Reference Stream |
837
|
6 |
|
$xref = $this->decodeXrefStream($pdfData, $startxref, $xref); |
838
|
|
|
} |
839
|
29 |
|
if (empty($xref)) { |
840
|
|
|
throw new Exception('Unable to find xref'); |
841
|
|
|
} |
842
|
|
|
|
843
|
29 |
|
return $xref; |
844
|
|
|
} |
845
|
|
|
|
846
|
|
|
/** |
847
|
|
|
* Parses PDF data and returns extracted data as array. |
848
|
|
|
* |
849
|
|
|
* @param string $data PDF data to parse |
850
|
|
|
* |
851
|
|
|
* @return array array of parsed PDF document objects |
852
|
|
|
* |
853
|
|
|
* @throws Exception if empty PDF data given |
854
|
|
|
* @throws Exception if PDF data missing %PDF header |
855
|
|
|
*/ |
856
|
30 |
|
public function parseData($data) |
857
|
|
|
{ |
858
|
30 |
|
if (empty($data)) { |
859
|
|
|
throw new Exception('Empty PDF data given.'); |
860
|
|
|
} |
861
|
|
|
// find the pdf header starting position |
862
|
30 |
|
if (false === ($trimpos = strpos($data, '%PDF-'))) { |
863
|
|
|
throw new Exception('Invalid PDF data: missing %PDF header.'); |
864
|
|
|
} |
865
|
|
|
|
866
|
|
|
// get PDF content string |
867
|
30 |
|
$pdfData = substr($data, $trimpos); |
868
|
|
|
|
869
|
|
|
// get xref and trailer data |
870
|
30 |
|
$xref = $this->getXrefData($pdfData); |
871
|
|
|
|
872
|
|
|
// parse all document objects |
873
|
29 |
|
$objects = []; |
874
|
29 |
|
foreach ($xref['xref'] as $obj => $offset) { |
875
|
29 |
|
if (!isset($objects[$obj]) && ($offset > 0)) { |
876
|
|
|
// decode objects with positive offset |
877
|
29 |
|
$objects[$obj] = $this->getIndirectObject($pdfData, $xref, $obj, $offset, true); |
878
|
|
|
} |
879
|
|
|
} |
880
|
|
|
|
881
|
29 |
|
return [$xref, $objects]; |
882
|
|
|
} |
883
|
|
|
} |
884
|
|
|
|