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
|
|
|
* |
19
|
|
|
* @date 2020-01-06 |
20
|
|
|
* |
21
|
|
|
* @license LGPLv3 |
22
|
|
|
* |
23
|
|
|
* @url <https://github.com/smalot/pdfparser> |
24
|
|
|
* |
25
|
|
|
* PdfParser is a pdf library written in PHP, extraction oriented. |
26
|
|
|
* Copyright (C) 2017 - Sébastien MALOT <[email protected]> |
27
|
|
|
* |
28
|
|
|
* This program is free software: you can redistribute it and/or modify |
29
|
|
|
* it under the terms of the GNU Lesser General Public License as published by |
30
|
|
|
* the Free Software Foundation, either version 3 of the License, or |
31
|
|
|
* (at your option) any later version. |
32
|
|
|
* |
33
|
|
|
* This program is distributed in the hope that it will be useful, |
34
|
|
|
* but WITHOUT ANY WARRANTY; without even the implied warranty of |
35
|
|
|
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the |
36
|
|
|
* GNU Lesser General Public License for more details. |
37
|
|
|
* |
38
|
|
|
* You should have received a copy of the GNU Lesser General Public License |
39
|
|
|
* along with this program. |
40
|
|
|
* If not, see <http://www.pdfparser.org/sites/default/LICENSE.txt>. |
41
|
|
|
*/ |
42
|
|
|
|
43
|
|
|
namespace Smalot\PdfParser\RawData; |
44
|
|
|
|
45
|
|
|
use Smalot\PdfParser\Config; |
46
|
|
|
|
47
|
|
|
class RawDataParser |
48
|
|
|
{ |
49
|
|
|
/** |
50
|
|
|
* @var Config |
51
|
|
|
*/ |
52
|
|
|
private $config; |
53
|
|
|
|
54
|
|
|
/** |
55
|
|
|
* Configuration array. |
56
|
|
|
* |
57
|
|
|
* @var array<string,bool> |
58
|
|
|
*/ |
59
|
|
|
protected $cfg = [ |
60
|
|
|
// if `true` ignore filter decoding errors |
61
|
|
|
'ignore_filter_decoding_errors' => true, |
62
|
|
|
// if `true` ignore missing filter decoding errors |
63
|
|
|
'ignore_missing_filter_decoders' => true, |
64
|
|
|
]; |
65
|
|
|
|
66
|
|
|
protected $filterHelper; |
67
|
|
|
protected $objects; |
68
|
|
|
|
69
|
|
|
/** |
70
|
|
|
* @param array $cfg Configuration array, default is [] |
71
|
|
|
*/ |
72
|
72 |
|
public function __construct($cfg = [], ?Config $config = null) |
73
|
|
|
{ |
74
|
|
|
// merge given array with default values |
75
|
72 |
|
$this->cfg = array_merge($this->cfg, $cfg); |
76
|
|
|
|
77
|
72 |
|
$this->filterHelper = new FilterHelper(); |
78
|
72 |
|
$this->config = $config ?: new Config(); |
79
|
|
|
} |
80
|
|
|
|
81
|
|
|
/** |
82
|
|
|
* Decode the specified stream. |
83
|
|
|
* |
84
|
|
|
* @param string $pdfData PDF data |
85
|
|
|
* @param array $sdic Stream's dictionary array |
86
|
|
|
* @param string $stream Stream to decode |
87
|
|
|
* |
88
|
|
|
* @return array containing decoded stream data and remaining filters |
89
|
|
|
* |
90
|
|
|
* @throws \Exception |
91
|
|
|
*/ |
92
|
65 |
|
protected function decodeStream(string $pdfData, array $xref, array $sdic, string $stream): array |
93
|
|
|
{ |
94
|
|
|
// get stream length and filters |
95
|
65 |
|
$slength = \strlen($stream); |
96
|
65 |
|
if ($slength <= 0) { |
97
|
|
|
return ['', []]; |
98
|
|
|
} |
99
|
65 |
|
$filters = []; |
100
|
65 |
|
foreach ($sdic as $k => $v) { |
101
|
65 |
|
if ('/' == $v[0]) { |
102
|
65 |
|
if (('Length' == $v[1]) && (isset($sdic[$k + 1])) && ('numeric' == $sdic[$k + 1][0])) { |
103
|
|
|
// get declared stream length |
104
|
60 |
|
$declength = (int) $sdic[$k + 1][1]; |
105
|
60 |
|
if ($declength < $slength) { |
106
|
60 |
|
$stream = substr($stream, 0, $declength); |
107
|
60 |
|
$slength = $declength; |
108
|
|
|
} |
109
|
65 |
|
} elseif (('Filter' == $v[1]) && (isset($sdic[$k + 1]))) { |
110
|
|
|
// resolve indirect object |
111
|
65 |
|
$objval = $this->getObjectVal($pdfData, $xref, $sdic[$k + 1]); |
112
|
65 |
|
if ('/' == $objval[0]) { |
113
|
|
|
// single filter |
114
|
65 |
|
$filters[] = $objval[1]; |
115
|
4 |
|
} elseif ('[' == $objval[0]) { |
116
|
|
|
// array of filters |
117
|
4 |
|
foreach ($objval[1] as $flt) { |
118
|
4 |
|
if ('/' == $flt[0]) { |
119
|
4 |
|
$filters[] = $flt[1]; |
120
|
|
|
} |
121
|
|
|
} |
122
|
|
|
} |
123
|
|
|
} |
124
|
|
|
} |
125
|
|
|
} |
126
|
|
|
|
127
|
|
|
// decode the stream |
128
|
65 |
|
$remaining_filters = []; |
129
|
65 |
|
foreach ($filters as $filter) { |
130
|
65 |
|
if (\in_array($filter, $this->filterHelper->getAvailableFilters(), true)) { |
131
|
|
|
try { |
132
|
65 |
|
$stream = $this->filterHelper->decodeFilter($filter, $stream, $this->config->getDecodeMemoryLimit()); |
133
|
2 |
|
} catch (\Exception $e) { |
134
|
2 |
|
$emsg = $e->getMessage(); |
135
|
2 |
|
if ((('~' == $emsg[0]) && !$this->cfg['ignore_missing_filter_decoders']) |
136
|
2 |
|
|| (('~' != $emsg[0]) && !$this->cfg['ignore_filter_decoding_errors']) |
137
|
|
|
) { |
138
|
65 |
|
throw new \Exception($e->getMessage()); |
139
|
|
|
} |
140
|
|
|
} |
141
|
|
|
} else { |
142
|
|
|
// add missing filter to array |
143
|
9 |
|
$remaining_filters[] = $filter; |
144
|
|
|
} |
145
|
|
|
} |
146
|
|
|
|
147
|
65 |
|
return [$stream, $remaining_filters]; |
148
|
|
|
} |
149
|
|
|
|
150
|
|
|
/** |
151
|
|
|
* Decode the Cross-Reference section |
152
|
|
|
* |
153
|
|
|
* @param string $pdfData PDF data |
154
|
|
|
* @param int $startxref Offset at which the xref section starts (position of the 'xref' keyword) |
155
|
|
|
* @param array $xref Previous xref array (if any) |
156
|
|
|
* |
157
|
|
|
* @return array containing xref and trailer data |
158
|
|
|
* |
159
|
|
|
* @throws \Exception |
160
|
|
|
*/ |
161
|
52 |
|
protected function decodeXref(string $pdfData, int $startxref, array $xref = []): array |
162
|
|
|
{ |
163
|
52 |
|
$startxref += 4; // 4 is the length of the word 'xref' |
164
|
|
|
// skip initial white space chars |
165
|
52 |
|
$offset = $startxref + strspn($pdfData, $this->config->getPdfWhitespaces(), $startxref); |
166
|
|
|
// initialize object number |
167
|
52 |
|
$obj_num = 0; |
168
|
|
|
// search for cross-reference entries or subsection |
169
|
52 |
|
while (preg_match('/([0-9]+)[\x20]([0-9]+)[\x20]?([nf]?)(\r\n|[\x20]?[\r\n])/', $pdfData, $matches, \PREG_OFFSET_CAPTURE, $offset) > 0) { |
170
|
52 |
|
if ($matches[0][1] != $offset) { |
171
|
|
|
// we are on another section |
172
|
12 |
|
break; |
173
|
|
|
} |
174
|
52 |
|
$offset += \strlen($matches[0][0]); |
175
|
52 |
|
if ('n' == $matches[3][0]) { |
176
|
|
|
// create unique object index: [object number]_[generation number] |
177
|
52 |
|
$index = $obj_num.'_'.(int) $matches[2][0]; |
178
|
|
|
// check if object already exist |
179
|
52 |
|
if (!isset($xref['xref'][$index])) { |
180
|
|
|
// store object offset position |
181
|
52 |
|
$xref['xref'][$index] = (int) $matches[1][0]; |
182
|
|
|
} |
183
|
52 |
|
++$obj_num; |
184
|
52 |
|
} elseif ('f' == $matches[3][0]) { |
185
|
52 |
|
++$obj_num; |
186
|
|
|
} else { |
187
|
|
|
// object number (index) |
188
|
52 |
|
$obj_num = (int) $matches[1][0]; |
189
|
|
|
} |
190
|
|
|
} |
191
|
|
|
// get trailer data |
192
|
52 |
|
if (preg_match('/trailer[\s]*<<(.*)>>/isU', $pdfData, $matches, \PREG_OFFSET_CAPTURE, $offset) > 0) { |
193
|
52 |
|
$trailer_data = $matches[1][0]; |
194
|
52 |
|
if (!isset($xref['trailer']) || empty($xref['trailer'])) { |
195
|
|
|
// get only the last updated version |
196
|
52 |
|
$xref['trailer'] = []; |
197
|
|
|
// parse trailer_data |
198
|
52 |
|
if (preg_match('/Size[\s]+([0-9]+)/i', $trailer_data, $matches) > 0) { |
199
|
52 |
|
$xref['trailer']['size'] = (int) $matches[1]; |
200
|
|
|
} |
201
|
52 |
|
if (preg_match('/Root[\s]+([0-9]+)[\s]+([0-9]+)[\s]+R/i', $trailer_data, $matches) > 0) { |
202
|
52 |
|
$xref['trailer']['root'] = (int) $matches[1].'_'.(int) $matches[2]; |
203
|
|
|
} |
204
|
52 |
|
if (preg_match('/Encrypt[\s]+([0-9]+)[\s]+([0-9]+)[\s]+R/i', $trailer_data, $matches) > 0) { |
205
|
2 |
|
$xref['trailer']['encrypt'] = (int) $matches[1].'_'.(int) $matches[2]; |
206
|
|
|
} |
207
|
52 |
|
if (preg_match('/Info[\s]+([0-9]+)[\s]+([0-9]+)[\s]+R/i', $trailer_data, $matches) > 0) { |
208
|
49 |
|
$xref['trailer']['info'] = (int) $matches[1].'_'.(int) $matches[2]; |
209
|
|
|
} |
210
|
52 |
|
if (preg_match('/ID[\s]*[\[][\s]*[<]([^>]*)[>][\s]*[<]([^>]*)[>]/i', $trailer_data, $matches) > 0) { |
211
|
41 |
|
$xref['trailer']['id'] = []; |
212
|
41 |
|
$xref['trailer']['id'][0] = $matches[1]; |
213
|
41 |
|
$xref['trailer']['id'][1] = $matches[2]; |
214
|
|
|
} |
215
|
|
|
} |
216
|
52 |
|
if (preg_match('/Prev[\s]+([0-9]+)/i', $trailer_data, $matches) > 0) { |
217
|
|
|
// get previous xref |
218
|
52 |
|
$xref = $this->getXrefData($pdfData, (int) $matches[1], $xref); |
219
|
|
|
} |
220
|
|
|
} else { |
221
|
|
|
throw new \Exception('Unable to find trailer'); |
222
|
|
|
} |
223
|
|
|
|
224
|
52 |
|
return $xref; |
225
|
|
|
} |
226
|
|
|
|
227
|
|
|
/** |
228
|
|
|
* Decode the Cross-Reference Stream section |
229
|
|
|
* |
230
|
|
|
* @param string $pdfData PDF data |
231
|
|
|
* @param int $startxref Offset at which the xref section starts |
232
|
|
|
* @param array $xref Previous xref array (if any) |
233
|
|
|
* |
234
|
|
|
* @return array containing xref and trailer data |
235
|
|
|
* |
236
|
|
|
* @throws \Exception if unknown PNG predictor detected |
237
|
|
|
*/ |
238
|
13 |
|
protected function decodeXrefStream(string $pdfData, int $startxref, array $xref = []): array |
239
|
|
|
{ |
240
|
|
|
// try to read Cross-Reference Stream |
241
|
13 |
|
$xrefobj = $this->getRawObject($pdfData, $startxref); |
242
|
13 |
|
$xrefcrs = $this->getIndirectObject($pdfData, $xref, $xrefobj[1], $startxref, true); |
243
|
13 |
|
if (!isset($xref['trailer']) || empty($xref['trailer'])) { |
244
|
|
|
// get only the last updated version |
245
|
13 |
|
$xref['trailer'] = []; |
246
|
13 |
|
$filltrailer = true; |
247
|
|
|
} else { |
248
|
11 |
|
$filltrailer = false; |
249
|
|
|
} |
250
|
13 |
|
if (!isset($xref['xref'])) { |
251
|
13 |
|
$xref['xref'] = []; |
252
|
|
|
} |
253
|
13 |
|
$valid_crs = false; |
254
|
13 |
|
$columns = 0; |
255
|
13 |
|
$predictor = null; |
256
|
13 |
|
$sarr = $xrefcrs[0][1]; |
257
|
13 |
|
if (!\is_array($sarr)) { |
258
|
|
|
$sarr = []; |
259
|
|
|
} |
260
|
|
|
|
261
|
13 |
|
$wb = []; |
262
|
|
|
|
263
|
13 |
|
foreach ($sarr as $k => $v) { |
264
|
|
|
if ( |
265
|
13 |
|
('/' == $v[0]) |
266
|
13 |
|
&& ('Type' == $v[1]) |
267
|
13 |
|
&& (isset($sarr[$k + 1]) |
268
|
13 |
|
&& '/' == $sarr[$k + 1][0] |
269
|
13 |
|
&& 'XRef' == $sarr[$k + 1][1] |
270
|
|
|
) |
271
|
|
|
) { |
272
|
13 |
|
$valid_crs = true; |
273
|
13 |
|
} elseif (('/' == $v[0]) && ('Index' == $v[1]) && (isset($sarr[$k + 1]))) { |
274
|
|
|
// initialize list for: first object number in the subsection / number of objects |
275
|
11 |
|
$index_blocks = []; |
276
|
11 |
|
for ($m = 0; $m < \count($sarr[$k + 1][1]); $m += 2) { |
|
|
|
|
277
|
11 |
|
$index_blocks[] = [$sarr[$k + 1][1][$m][1], $sarr[$k + 1][1][$m + 1][1]]; |
278
|
|
|
} |
279
|
13 |
|
} elseif (('/' == $v[0]) && ('Prev' == $v[1]) && (isset($sarr[$k + 1]) && ('numeric' == $sarr[$k + 1][0]))) { |
280
|
|
|
// get previous xref offset |
281
|
11 |
|
$prevxref = (int) $sarr[$k + 1][1]; |
282
|
13 |
|
} elseif (('/' == $v[0]) && ('W' == $v[1]) && (isset($sarr[$k + 1]))) { |
283
|
|
|
// number of bytes (in the decoded stream) of the corresponding field |
284
|
13 |
|
$wb[0] = (int) $sarr[$k + 1][1][0][1]; |
285
|
13 |
|
$wb[1] = (int) $sarr[$k + 1][1][1][1]; |
286
|
13 |
|
$wb[2] = (int) $sarr[$k + 1][1][2][1]; |
287
|
13 |
|
} elseif (('/' == $v[0]) && ('DecodeParms' == $v[1]) && (isset($sarr[$k + 1][1]))) { |
288
|
11 |
|
$decpar = $sarr[$k + 1][1]; |
289
|
11 |
|
foreach ($decpar as $kdc => $vdc) { |
290
|
|
|
if ( |
291
|
11 |
|
'/' == $vdc[0] |
292
|
11 |
|
&& 'Columns' == $vdc[1] |
293
|
11 |
|
&& (isset($decpar[$kdc + 1]) |
294
|
11 |
|
&& 'numeric' == $decpar[$kdc + 1][0] |
295
|
|
|
) |
296
|
|
|
) { |
297
|
11 |
|
$columns = (int) $decpar[$kdc + 1][1]; |
298
|
|
|
} elseif ( |
299
|
11 |
|
'/' == $vdc[0] |
300
|
11 |
|
&& 'Predictor' == $vdc[1] |
301
|
11 |
|
&& (isset($decpar[$kdc + 1]) |
302
|
11 |
|
&& 'numeric' == $decpar[$kdc + 1][0] |
303
|
|
|
) |
304
|
|
|
) { |
305
|
11 |
|
$predictor = (int) $decpar[$kdc + 1][1]; |
306
|
|
|
} |
307
|
|
|
} |
308
|
13 |
|
} elseif ($filltrailer) { |
309
|
13 |
|
if (('/' == $v[0]) && ('Size' == $v[1]) && (isset($sarr[$k + 1]) && ('numeric' == $sarr[$k + 1][0]))) { |
310
|
13 |
|
$xref['trailer']['size'] = $sarr[$k + 1][1]; |
311
|
13 |
|
} elseif (('/' == $v[0]) && ('Root' == $v[1]) && (isset($sarr[$k + 1]) && ('objref' == $sarr[$k + 1][0]))) { |
312
|
13 |
|
$xref['trailer']['root'] = $sarr[$k + 1][1]; |
313
|
13 |
|
} elseif (('/' == $v[0]) && ('Info' == $v[1]) && (isset($sarr[$k + 1]) && ('objref' == $sarr[$k + 1][0]))) { |
314
|
13 |
|
$xref['trailer']['info'] = $sarr[$k + 1][1]; |
315
|
13 |
|
} elseif (('/' == $v[0]) && ('Encrypt' == $v[1]) && (isset($sarr[$k + 1]) && ('objref' == $sarr[$k + 1][0]))) { |
316
|
|
|
$xref['trailer']['encrypt'] = $sarr[$k + 1][1]; |
317
|
13 |
|
} elseif (('/' == $v[0]) && ('ID' == $v[1]) && (isset($sarr[$k + 1]))) { |
318
|
13 |
|
$xref['trailer']['id'] = []; |
319
|
13 |
|
$xref['trailer']['id'][0] = $sarr[$k + 1][1][0][1]; |
320
|
13 |
|
$xref['trailer']['id'][1] = $sarr[$k + 1][1][1][1]; |
321
|
|
|
} |
322
|
|
|
} |
323
|
|
|
} |
324
|
|
|
|
325
|
|
|
// decode data |
326
|
13 |
|
if ($valid_crs && isset($xrefcrs[1][3][0])) { |
327
|
13 |
|
if (null !== $predictor) { |
328
|
|
|
// number of bytes in a row |
329
|
11 |
|
$rowlen = ($columns + 1); |
330
|
|
|
// convert the stream into an array of integers |
331
|
|
|
/** @var array<int> */ |
332
|
11 |
|
$sdata = unpack('C*', $xrefcrs[1][3][0]); |
333
|
|
|
// TODO: Handle the case when unpack returns false |
334
|
|
|
|
335
|
|
|
// split the rows |
336
|
11 |
|
$sdata = array_chunk($sdata, $rowlen); |
337
|
|
|
|
338
|
|
|
// initialize decoded array |
339
|
11 |
|
$ddata = []; |
340
|
|
|
// initialize first row with zeros |
341
|
11 |
|
$prev_row = array_fill(0, $rowlen, 0); |
342
|
|
|
// for each row apply PNG unpredictor |
343
|
11 |
|
foreach ($sdata as $k => $row) { |
344
|
|
|
// initialize new row |
345
|
11 |
|
$ddata[$k] = []; |
346
|
|
|
// get PNG predictor value |
347
|
11 |
|
$predictor = (10 + $row[0]); |
348
|
|
|
// for each byte on the row |
349
|
11 |
|
for ($i = 1; $i <= $columns; ++$i) { |
350
|
|
|
// new index |
351
|
11 |
|
$j = ($i - 1); |
352
|
11 |
|
$row_up = $prev_row[$j]; |
353
|
11 |
|
if (1 == $i) { |
354
|
11 |
|
$row_left = 0; |
355
|
11 |
|
$row_upleft = 0; |
356
|
|
|
} else { |
357
|
11 |
|
$row_left = $row[$i - 1]; |
358
|
11 |
|
$row_upleft = $prev_row[$j - 1]; |
359
|
|
|
} |
360
|
|
|
switch ($predictor) { |
361
|
11 |
|
case 10: // PNG prediction (on encoding, PNG None on all rows) |
362
|
|
|
$ddata[$k][$j] = $row[$i]; |
363
|
|
|
break; |
364
|
|
|
|
365
|
11 |
|
case 11: // PNG prediction (on encoding, PNG Sub on all rows) |
366
|
|
|
$ddata[$k][$j] = (($row[$i] + $row_left) & 0xFF); |
367
|
|
|
break; |
368
|
|
|
|
369
|
11 |
|
case 12: // PNG prediction (on encoding, PNG Up on all rows) |
370
|
11 |
|
$ddata[$k][$j] = (($row[$i] + $row_up) & 0xFF); |
371
|
11 |
|
break; |
372
|
|
|
|
373
|
|
|
case 13: // PNG prediction (on encoding, PNG Average on all rows) |
374
|
|
|
$ddata[$k][$j] = (($row[$i] + (($row_left + $row_up) / 2)) & 0xFF); |
375
|
|
|
break; |
376
|
|
|
|
377
|
|
|
case 14: // PNG prediction (on encoding, PNG Paeth on all rows) |
378
|
|
|
// initial estimate |
379
|
|
|
$p = ($row_left + $row_up - $row_upleft); |
380
|
|
|
// distances |
381
|
|
|
$pa = abs($p - $row_left); |
382
|
|
|
$pb = abs($p - $row_up); |
383
|
|
|
$pc = abs($p - $row_upleft); |
384
|
|
|
$pmin = min($pa, $pb, $pc); |
385
|
|
|
// return minimum distance |
386
|
|
|
switch ($pmin) { |
387
|
|
|
case $pa: |
388
|
|
|
$ddata[$k][$j] = (($row[$i] + $row_left) & 0xFF); |
389
|
|
|
break; |
390
|
|
|
|
391
|
|
|
case $pb: |
392
|
|
|
$ddata[$k][$j] = (($row[$i] + $row_up) & 0xFF); |
393
|
|
|
break; |
394
|
|
|
|
395
|
|
|
case $pc: |
396
|
|
|
$ddata[$k][$j] = (($row[$i] + $row_upleft) & 0xFF); |
397
|
|
|
break; |
398
|
|
|
} |
399
|
|
|
break; |
400
|
|
|
|
401
|
|
|
default: // PNG prediction (on encoding, PNG optimum) |
402
|
|
|
throw new \Exception('Unknown PNG predictor: '.$predictor); |
403
|
|
|
} |
404
|
|
|
} |
405
|
11 |
|
$prev_row = $ddata[$k]; |
406
|
|
|
} // end for each row |
407
|
|
|
// complete decoding |
408
|
|
|
} else { |
409
|
|
|
// number of bytes in a row |
410
|
2 |
|
$rowlen = array_sum($wb); |
411
|
2 |
|
if (0 < $rowlen) { |
412
|
|
|
// convert the stream into an array of integers |
413
|
2 |
|
$sdata = unpack('C*', $xrefcrs[1][3][0]); |
414
|
|
|
// split the rows |
415
|
2 |
|
$ddata = array_chunk($sdata, $rowlen); |
|
|
|
|
416
|
|
|
} else { |
417
|
|
|
// if the row length is zero, $ddata should be an empty array as well |
418
|
|
|
$ddata = []; |
419
|
|
|
} |
420
|
|
|
} |
421
|
|
|
|
422
|
13 |
|
$sdata = []; |
423
|
|
|
|
424
|
|
|
// for every row |
425
|
13 |
|
foreach ($ddata as $k => $row) { |
426
|
|
|
// initialize new row |
427
|
13 |
|
$sdata[$k] = [0, 0, 0]; |
428
|
13 |
|
if (0 == $wb[0]) { |
429
|
|
|
// default type field |
430
|
|
|
$sdata[$k][0] = 1; |
431
|
|
|
} |
432
|
13 |
|
$i = 0; // count bytes in the row |
433
|
|
|
// for every column |
434
|
13 |
|
for ($c = 0; $c < 3; ++$c) { |
435
|
|
|
// for every byte on the column |
436
|
13 |
|
for ($b = 0; $b < $wb[$c]; ++$b) { |
437
|
13 |
|
if (isset($row[$i])) { |
438
|
13 |
|
$sdata[$k][$c] += ($row[$i] << (($wb[$c] - 1 - $b) * 8)); |
439
|
|
|
} |
440
|
13 |
|
++$i; |
441
|
|
|
} |
442
|
|
|
} |
443
|
|
|
} |
444
|
|
|
|
445
|
|
|
// fill xref |
446
|
13 |
|
if (isset($index_blocks)) { |
447
|
|
|
// load the first object number of the first /Index entry |
448
|
11 |
|
$obj_num = $index_blocks[0][0]; |
449
|
|
|
} else { |
450
|
12 |
|
$obj_num = 0; |
451
|
|
|
} |
452
|
13 |
|
foreach ($sdata as $k => $row) { |
453
|
13 |
|
switch ($row[0]) { |
454
|
13 |
|
case 0: // (f) linked list of free objects |
455
|
13 |
|
break; |
456
|
|
|
|
457
|
13 |
|
case 1: // (n) objects that are in use but are not compressed |
458
|
|
|
// create unique object index: [object number]_[generation number] |
459
|
13 |
|
$index = $obj_num.'_'.$row[2]; |
460
|
|
|
// check if object already exist |
461
|
13 |
|
if (!isset($xref['xref'][$index])) { |
462
|
|
|
// store object offset position |
463
|
13 |
|
$xref['xref'][$index] = $row[1]; |
464
|
|
|
} |
465
|
13 |
|
break; |
466
|
|
|
|
467
|
13 |
|
case 2: // compressed objects |
468
|
|
|
// $row[1] = object number of the object stream in which this object is stored |
469
|
|
|
// $row[2] = index of this object within the object stream |
470
|
13 |
|
$index = $row[1].'_0_'.$row[2]; |
471
|
13 |
|
$xref['xref'][$index] = -1; |
472
|
13 |
|
break; |
473
|
|
|
|
474
|
|
|
default: // null objects |
475
|
|
|
break; |
476
|
|
|
} |
477
|
13 |
|
++$obj_num; |
478
|
13 |
|
if (isset($index_blocks)) { |
479
|
|
|
// reduce the number of remaining objects |
480
|
11 |
|
--$index_blocks[0][1]; |
481
|
11 |
|
if (0 == $index_blocks[0][1]) { |
|
|
|
|
482
|
|
|
// remove the actual used /Index entry |
483
|
11 |
|
array_shift($index_blocks); |
484
|
11 |
|
if (0 < \count($index_blocks)) { |
485
|
|
|
// load the first object number of the following /Index entry |
486
|
8 |
|
$obj_num = $index_blocks[0][0]; |
487
|
|
|
} else { |
488
|
|
|
// if there are no more entries, remove $index_blocks to avoid actions on an empty array |
489
|
11 |
|
unset($index_blocks); |
490
|
|
|
} |
491
|
|
|
} |
492
|
|
|
} |
493
|
|
|
} |
494
|
|
|
} // end decoding data |
495
|
13 |
|
if (isset($prevxref)) { |
496
|
|
|
// get previous xref |
497
|
11 |
|
$xref = $this->getXrefData($pdfData, $prevxref, $xref); |
498
|
|
|
} |
499
|
|
|
|
500
|
13 |
|
return $xref; |
501
|
|
|
} |
502
|
|
|
|
503
|
65 |
|
protected function getObjectHeaderPattern(array $objRefs): string |
504
|
|
|
{ |
505
|
|
|
// consider all whitespace character (PDF specifications) |
506
|
65 |
|
return '/'.$objRefs[0].$this->config->getPdfWhitespacesRegex().$objRefs[1].$this->config->getPdfWhitespacesRegex().'obj/'; |
507
|
|
|
} |
508
|
|
|
|
509
|
65 |
|
protected function getObjectHeaderLen(array $objRefs): int |
510
|
|
|
{ |
511
|
|
|
// "4 0 obj" |
512
|
|
|
// 2 whitespaces + strlen("obj") = 5 |
513
|
65 |
|
return 5 + \strlen($objRefs[0]) + \strlen($objRefs[1]); |
514
|
|
|
} |
515
|
|
|
|
516
|
|
|
/** |
517
|
|
|
* Get content of indirect object. |
518
|
|
|
* |
519
|
|
|
* @param string $pdfData PDF data |
520
|
|
|
* @param string $objRef Object number and generation number separated by underscore character |
521
|
|
|
* @param int $offset Object offset |
522
|
|
|
* @param bool $decoding If true decode streams |
523
|
|
|
* |
524
|
|
|
* @return array containing object data |
525
|
|
|
* |
526
|
|
|
* @throws \Exception if invalid object reference found |
527
|
|
|
*/ |
528
|
65 |
|
protected function getIndirectObject(string $pdfData, array $xref, string $objRef, int $offset = 0, bool $decoding = true): array |
529
|
|
|
{ |
530
|
|
|
/* |
531
|
|
|
* build indirect object header |
532
|
|
|
*/ |
533
|
|
|
// $objHeader = "[object number] [generation number] obj" |
534
|
65 |
|
$objRefArr = explode('_', $objRef); |
535
|
65 |
|
if (2 !== \count($objRefArr)) { |
536
|
|
|
throw new \Exception('Invalid object reference for $obj.'); |
537
|
|
|
} |
538
|
|
|
|
539
|
65 |
|
$objHeaderLen = $this->getObjectHeaderLen($objRefArr); |
540
|
|
|
|
541
|
|
|
/* |
542
|
|
|
* check if we are in position |
543
|
|
|
*/ |
544
|
|
|
// ignore whitespace characters at offset |
545
|
65 |
|
$offset += strspn($pdfData, $this->config->getPdfWhitespaces(), $offset); |
546
|
|
|
// ignore leading zeros for object number |
547
|
65 |
|
$offset += strspn($pdfData, '0', $offset); |
548
|
65 |
|
if (0 == preg_match($this->getObjectHeaderPattern($objRefArr), substr($pdfData, $offset, $objHeaderLen))) { |
549
|
|
|
// an indirect reference to an undefined object shall be considered a reference to the null object |
550
|
|
|
return ['null', 'null', $offset]; |
551
|
|
|
} |
552
|
|
|
|
553
|
|
|
/* |
554
|
|
|
* get content |
555
|
|
|
*/ |
556
|
|
|
// starting position of object content |
557
|
65 |
|
$offset += $objHeaderLen; |
558
|
65 |
|
$objContentArr = []; |
559
|
65 |
|
$i = 0; // object main index |
560
|
65 |
|
$header = null; |
561
|
|
|
do { |
562
|
65 |
|
$oldOffset = $offset; |
563
|
|
|
// get element |
564
|
65 |
|
$element = $this->getRawObject($pdfData, $offset, null != $header ? $header[1] : null); |
565
|
65 |
|
$offset = $element[2]; |
566
|
|
|
// decode stream using stream's dictionary information |
567
|
65 |
|
if ($decoding && ('stream' === $element[0]) && null != $header) { |
568
|
65 |
|
$element[3] = $this->decodeStream($pdfData, $xref, $header[1], $element[1]); |
569
|
|
|
} |
570
|
65 |
|
$objContentArr[$i] = $element; |
571
|
65 |
|
$header = isset($element[0]) && '<<' === $element[0] ? $element : null; |
572
|
65 |
|
++$i; |
573
|
65 |
|
} while (('endobj' !== $element[0]) && ($offset !== $oldOffset)); |
574
|
|
|
// remove closing delimiter |
575
|
65 |
|
array_pop($objContentArr); |
576
|
|
|
|
577
|
|
|
/* |
578
|
|
|
* return raw object content |
579
|
|
|
*/ |
580
|
65 |
|
return $objContentArr; |
581
|
|
|
} |
582
|
|
|
|
583
|
|
|
/** |
584
|
|
|
* Get the content of object, resolving indirect object reference if necessary. |
585
|
|
|
* |
586
|
|
|
* @param string $pdfData PDF data |
587
|
|
|
* @param array $obj Object value |
588
|
|
|
* |
589
|
|
|
* @return array containing object data |
590
|
|
|
* |
591
|
|
|
* @throws \Exception |
592
|
|
|
*/ |
593
|
65 |
|
protected function getObjectVal(string $pdfData, $xref, array $obj): array |
594
|
|
|
{ |
595
|
65 |
|
if ('objref' == $obj[0]) { |
596
|
|
|
// reference to indirect object |
597
|
|
|
if (isset($this->objects[$obj[1]])) { |
598
|
|
|
// this object has been already parsed |
599
|
|
|
return $this->objects[$obj[1]]; |
600
|
|
|
} elseif (isset($xref[$obj[1]])) { |
601
|
|
|
// parse new object |
602
|
|
|
$this->objects[$obj[1]] = $this->getIndirectObject($pdfData, $xref, $obj[1], $xref[$obj[1]], false); |
603
|
|
|
|
604
|
|
|
return $this->objects[$obj[1]]; |
605
|
|
|
} |
606
|
|
|
} |
607
|
|
|
|
608
|
65 |
|
return $obj; |
609
|
|
|
} |
610
|
|
|
|
611
|
|
|
/** |
612
|
|
|
* Get object type, raw value and offset to next object |
613
|
|
|
* |
614
|
|
|
* @param int $offset Object offset |
615
|
|
|
* @param array|null $headerDic obj header's dictionary, parsed by getRawObject. Used for stream parsing optimization |
616
|
|
|
* |
617
|
|
|
* @return array containing object type, raw value and offset to next object |
618
|
|
|
*/ |
619
|
66 |
|
protected function getRawObject(string $pdfData, int $offset = 0, ?array $headerDic = null): array |
620
|
|
|
{ |
621
|
66 |
|
$objtype = ''; // object type to be returned |
622
|
66 |
|
$objval = ''; // object value to be returned |
623
|
|
|
|
624
|
|
|
// skip initial white space chars |
625
|
66 |
|
$offset += strspn($pdfData, $this->config->getPdfWhitespaces(), $offset); |
626
|
|
|
|
627
|
|
|
// get first char |
628
|
66 |
|
$char = $pdfData[$offset]; |
629
|
|
|
// get object type |
630
|
|
|
switch ($char) { |
631
|
66 |
|
case '%': // \x25 PERCENT SIGN |
632
|
|
|
// skip comment and search for next token |
633
|
3 |
|
$next = strcspn($pdfData, "\r\n", $offset); |
634
|
3 |
|
if ($next > 0) { |
635
|
3 |
|
$offset += $next; |
636
|
|
|
|
637
|
3 |
|
return $this->getRawObject($pdfData, $offset); |
638
|
|
|
} |
639
|
|
|
break; |
640
|
|
|
|
641
|
66 |
|
case '/': // \x2F SOLIDUS |
642
|
|
|
// name object |
643
|
66 |
|
$objtype = $char; |
644
|
66 |
|
++$offset; |
645
|
66 |
|
$span = strcspn($pdfData, "\x00\x09\x0a\x0c\x0d\x20\n\t\r\v\f\x28\x29\x3c\x3e\x5b\x5d\x7b\x7d\x2f\x25", $offset, 256); |
646
|
66 |
|
if ($span > 0) { |
647
|
66 |
|
$objval = substr($pdfData, $offset, $span); // unescaped value |
648
|
66 |
|
$offset += $span; |
649
|
|
|
} |
650
|
66 |
|
break; |
651
|
|
|
|
652
|
66 |
|
case '(': // \x28 LEFT PARENTHESIS |
653
|
66 |
|
case ')': // \x29 RIGHT PARENTHESIS |
654
|
|
|
// literal string object |
655
|
61 |
|
$objtype = $char; |
656
|
61 |
|
++$offset; |
657
|
61 |
|
$strpos = $offset; |
658
|
61 |
|
if ('(' == $char) { |
659
|
61 |
|
$open_bracket = 1; |
660
|
61 |
|
while ($open_bracket > 0) { |
661
|
61 |
|
if (!isset($pdfData[$strpos])) { |
662
|
|
|
break; |
663
|
|
|
} |
664
|
61 |
|
$ch = $pdfData[$strpos]; |
665
|
|
|
switch ($ch) { |
666
|
61 |
|
case '\\': // REVERSE SOLIDUS (5Ch) (Backslash) |
667
|
|
|
// skip next character |
668
|
29 |
|
++$strpos; |
669
|
29 |
|
break; |
670
|
|
|
|
671
|
61 |
|
case '(': // LEFT PARENHESIS (28h) |
672
|
2 |
|
++$open_bracket; |
673
|
2 |
|
break; |
674
|
|
|
|
675
|
61 |
|
case ')': // RIGHT PARENTHESIS (29h) |
676
|
61 |
|
--$open_bracket; |
677
|
61 |
|
break; |
678
|
|
|
} |
679
|
61 |
|
++$strpos; |
680
|
|
|
} |
681
|
61 |
|
$objval = substr($pdfData, $offset, $strpos - $offset - 1); |
682
|
61 |
|
$offset = $strpos; |
683
|
|
|
} |
684
|
61 |
|
break; |
685
|
|
|
|
686
|
66 |
|
case '[': // \x5B LEFT SQUARE BRACKET |
687
|
66 |
|
case ']': // \x5D RIGHT SQUARE BRACKET |
688
|
|
|
// array object |
689
|
65 |
|
$objtype = $char; |
690
|
65 |
|
++$offset; |
691
|
65 |
|
if ('[' == $char) { |
692
|
|
|
// get array content |
693
|
65 |
|
$objval = []; |
694
|
|
|
do { |
695
|
65 |
|
$oldOffset = $offset; |
696
|
|
|
// get element |
697
|
65 |
|
$element = $this->getRawObject($pdfData, $offset); |
698
|
65 |
|
$offset = $element[2]; |
699
|
65 |
|
$objval[] = $element; |
700
|
65 |
|
} while ((']' != $element[0]) && ($offset != $oldOffset)); |
701
|
|
|
// remove closing delimiter |
702
|
65 |
|
array_pop($objval); |
703
|
|
|
} |
704
|
65 |
|
break; |
705
|
|
|
|
706
|
66 |
|
case '<': // \x3C LESS-THAN SIGN |
707
|
66 |
|
case '>': // \x3E GREATER-THAN SIGN |
708
|
66 |
|
if (isset($pdfData[$offset + 1]) && ($pdfData[$offset + 1] == $char)) { |
709
|
|
|
// dictionary object |
710
|
66 |
|
$objtype = $char.$char; |
711
|
66 |
|
$offset += 2; |
712
|
66 |
|
if ('<' == $char) { |
713
|
|
|
// get array content |
714
|
66 |
|
$objval = []; |
715
|
|
|
do { |
716
|
66 |
|
$oldOffset = $offset; |
717
|
|
|
// get element |
718
|
66 |
|
$element = $this->getRawObject($pdfData, $offset); |
719
|
66 |
|
$offset = $element[2]; |
720
|
66 |
|
$objval[] = $element; |
721
|
66 |
|
} while (('>>' != $element[0]) && ($offset != $oldOffset)); |
722
|
|
|
// remove closing delimiter |
723
|
66 |
|
array_pop($objval); |
724
|
|
|
} |
725
|
|
|
} else { |
726
|
|
|
// hexadecimal string object |
727
|
32 |
|
$objtype = $char; |
728
|
32 |
|
++$offset; |
729
|
|
|
|
730
|
32 |
|
$span = strspn($pdfData, "0123456789abcdefABCDEF\x09\x0a\x0c\x0d\x20", $offset); |
731
|
32 |
|
$dataToCheck = $pdfData[$offset + $span] ?? null; |
732
|
32 |
|
if ('<' == $char && $span > 0 && '>' == $dataToCheck) { |
733
|
|
|
// remove white space characters |
734
|
32 |
|
$objval = strtr(substr($pdfData, $offset, $span), $this->config->getPdfWhitespaces(), ''); |
735
|
32 |
|
$offset += $span + 1; |
736
|
2 |
|
} elseif (false !== ($endpos = strpos($pdfData, '>', $offset))) { |
737
|
2 |
|
$offset = $endpos + 1; |
738
|
|
|
} |
739
|
|
|
} |
740
|
66 |
|
break; |
741
|
|
|
|
742
|
|
|
default: |
743
|
66 |
|
if ('endobj' == substr($pdfData, $offset, 6)) { |
744
|
|
|
// indirect object |
745
|
65 |
|
$objtype = 'endobj'; |
746
|
65 |
|
$offset += 6; |
747
|
66 |
|
} elseif ('null' == substr($pdfData, $offset, 4)) { |
748
|
|
|
// null object |
749
|
11 |
|
$objtype = 'null'; |
750
|
11 |
|
$offset += 4; |
751
|
11 |
|
$objval = 'null'; |
752
|
66 |
|
} elseif ('true' == substr($pdfData, $offset, 4)) { |
753
|
|
|
// boolean true object |
754
|
31 |
|
$objtype = 'boolean'; |
755
|
31 |
|
$offset += 4; |
756
|
31 |
|
$objval = 'true'; |
757
|
66 |
|
} elseif ('false' == substr($pdfData, $offset, 5)) { |
758
|
|
|
// boolean false object |
759
|
5 |
|
$objtype = 'boolean'; |
760
|
5 |
|
$offset += 5; |
761
|
5 |
|
$objval = 'false'; |
762
|
66 |
|
} elseif ('stream' == substr($pdfData, $offset, 6)) { |
763
|
|
|
// start stream object |
764
|
66 |
|
$objtype = 'stream'; |
765
|
66 |
|
$offset += 6; |
766
|
66 |
|
if (1 == preg_match('/^( *[\r]?[\n])/isU', substr($pdfData, $offset, 4), $matches)) { |
767
|
66 |
|
$offset += \strlen($matches[0]); |
768
|
|
|
|
769
|
|
|
// we get stream length here to later help preg_match test less data |
770
|
66 |
|
$streamLen = (int) $this->getHeaderValue($headerDic, 'Length', 'numeric', 0); |
771
|
66 |
|
$skip = false === $this->config->getRetainImageContent() && 'XObject' == $this->getHeaderValue($headerDic, 'Type', '/') && 'Image' == $this->getHeaderValue($headerDic, 'Subtype', '/'); |
772
|
|
|
|
773
|
66 |
|
$pregResult = preg_match( |
774
|
66 |
|
'/(endstream)[\x09\x0a\x0c\x0d\x20]/isU', |
775
|
66 |
|
$pdfData, |
776
|
66 |
|
$matches, |
777
|
66 |
|
\PREG_OFFSET_CAPTURE, |
778
|
66 |
|
$offset + $streamLen |
779
|
66 |
|
); |
780
|
|
|
|
781
|
66 |
|
if (1 == $pregResult) { |
782
|
66 |
|
$objval = $skip ? '' : substr($pdfData, $offset, $matches[0][1] - $offset); |
783
|
66 |
|
$offset = $matches[1][1]; |
784
|
|
|
} |
785
|
|
|
} |
786
|
66 |
|
} elseif ('endstream' == substr($pdfData, $offset, 9)) { |
787
|
|
|
// end stream object |
788
|
65 |
|
$objtype = 'endstream'; |
789
|
65 |
|
$offset += 9; |
790
|
66 |
|
} elseif (1 == preg_match('/^([0-9]+)[\s]+([0-9]+)[\s]+R/iU', substr($pdfData, $offset, 33), $matches)) { |
791
|
|
|
// indirect object reference |
792
|
65 |
|
$objtype = 'objref'; |
793
|
65 |
|
$offset += \strlen($matches[0]); |
794
|
65 |
|
$objval = (int) $matches[1].'_'.(int) $matches[2]; |
795
|
66 |
|
} elseif (1 == preg_match('/^([0-9]+)[\s]+([0-9]+)[\s]+obj/iU', substr($pdfData, $offset, 33), $matches)) { |
796
|
|
|
// object start |
797
|
14 |
|
$objtype = 'obj'; |
798
|
14 |
|
$objval = (int) $matches[1].'_'.(int) $matches[2]; |
799
|
14 |
|
$offset += \strlen($matches[0]); |
800
|
66 |
|
} elseif (($numlen = strspn($pdfData, '+-.0123456789', $offset)) > 0) { |
801
|
|
|
// numeric object |
802
|
65 |
|
$objtype = 'numeric'; |
803
|
65 |
|
$objval = substr($pdfData, $offset, $numlen); |
804
|
65 |
|
$offset += $numlen; |
805
|
|
|
} |
806
|
66 |
|
break; |
807
|
|
|
} |
808
|
|
|
|
809
|
66 |
|
return [$objtype, $objval, $offset]; |
810
|
|
|
} |
811
|
|
|
|
812
|
|
|
/** |
813
|
|
|
* Get value of an object header's section (obj << YYY >> part ). |
814
|
|
|
* |
815
|
|
|
* It is similar to Header::get('...')->getContent(), the only difference is it can be used during the parsing process, |
816
|
|
|
* when no Smalot\PdfParser\Header objects are created yet. |
817
|
|
|
* |
818
|
|
|
* @param string $key header's section name |
819
|
|
|
* @param string $type type of the section (i.e. 'numeric', '/', '<<', etc.) |
820
|
|
|
* @param string|array|null $default default value for header's section |
821
|
|
|
* |
822
|
|
|
* @return string|array|null value of obj header's section, or default value if none found, or its type doesn't match $type param |
823
|
|
|
*/ |
824
|
66 |
|
private function getHeaderValue(?array $headerDic, string $key, string $type, $default = '') |
825
|
|
|
{ |
826
|
66 |
|
if (false === \is_array($headerDic)) { |
|
|
|
|
827
|
1 |
|
return $default; |
828
|
|
|
} |
829
|
|
|
|
830
|
|
|
/* |
831
|
|
|
* It recieves dictionary of header fields, as it is returned by RawDataParser::getRawObject, |
832
|
|
|
* iterates over it, searching for section of type '/' whith requested key. |
833
|
|
|
* If such a section is found, it tries to receive it's value (next object in dictionary), |
834
|
|
|
* returning it, if it matches requested type, or default value otherwise. |
835
|
|
|
*/ |
836
|
65 |
|
foreach ($headerDic as $i => $val) { |
837
|
65 |
|
$isSectionName = \is_array($val) && 3 == \count($val) && '/' == $val[0]; |
838
|
|
|
if ( |
839
|
65 |
|
$isSectionName |
840
|
65 |
|
&& $val[1] == $key |
841
|
65 |
|
&& isset($headerDic[$i + 1]) |
842
|
|
|
) { |
843
|
65 |
|
$isSectionValue = \is_array($headerDic[$i + 1]) && 1 < \count($headerDic[$i + 1]); |
844
|
|
|
|
845
|
65 |
|
return $isSectionValue && $type == $headerDic[$i + 1][0] |
846
|
60 |
|
? $headerDic[$i + 1][1] |
847
|
65 |
|
: $default; |
848
|
|
|
} |
849
|
|
|
} |
850
|
|
|
|
851
|
|
|
return $default; |
852
|
|
|
} |
853
|
|
|
|
854
|
|
|
/** |
855
|
|
|
* Get Cross-Reference (xref) table and trailer data from PDF document data. |
856
|
|
|
* |
857
|
|
|
* @param int $offset xref offset (if known) |
858
|
|
|
* @param array $xref previous xref array (if any) |
859
|
|
|
* |
860
|
|
|
* @return array containing xref and trailer data |
861
|
|
|
* |
862
|
|
|
* @throws \Exception if it was unable to find startxref |
863
|
|
|
* @throws \Exception if it was unable to find xref |
864
|
|
|
*/ |
865
|
66 |
|
protected function getXrefData(string $pdfData, int $offset = 0, array $xref = []): array |
866
|
|
|
{ |
867
|
|
|
// If the $offset is currently pointed at whitespace, bump it |
868
|
|
|
// forward until it isn't; affects loosely targetted offsets |
869
|
|
|
// for the 'xref' keyword |
870
|
|
|
// See: https://github.com/smalot/pdfparser/issues/673 |
871
|
66 |
|
$bumpOffset = $offset; |
872
|
66 |
|
while (preg_match('/\s/', substr($pdfData, $bumpOffset, 1))) { |
873
|
1 |
|
++$bumpOffset; |
874
|
|
|
} |
875
|
|
|
|
876
|
|
|
// Find all startxref tables from this $offset forward |
877
|
66 |
|
$startxrefPreg = preg_match_all( |
878
|
66 |
|
'/(?<=[\r\n])startxref[\s]*[\r\n]+([0-9]+)[\s]*[\r\n]+%%EOF/i', |
879
|
66 |
|
$pdfData, |
880
|
66 |
|
$startxrefMatches, |
881
|
66 |
|
\PREG_SET_ORDER, |
882
|
66 |
|
$offset |
883
|
66 |
|
); |
884
|
|
|
|
885
|
66 |
|
if (0 == $startxrefPreg) { |
|
|
|
|
886
|
|
|
// No startxref tables were found |
887
|
|
|
throw new \Exception('Unable to find startxref'); |
888
|
66 |
|
} elseif (0 == $offset) { |
889
|
|
|
// Use the last startxref in the document |
890
|
66 |
|
$startxref = (int) $startxrefMatches[\count($startxrefMatches) - 1][1]; |
891
|
23 |
|
} elseif (strpos($pdfData, 'xref', $bumpOffset) == $bumpOffset) { |
892
|
|
|
// Already pointing at the xref table |
893
|
12 |
|
$startxref = $bumpOffset; |
894
|
11 |
|
} elseif (preg_match('/([0-9]+[\s][0-9]+[\s]obj)/i', $pdfData, $matches, 0, $bumpOffset)) { |
895
|
|
|
// Cross-Reference Stream object |
896
|
11 |
|
$startxref = $bumpOffset; |
897
|
|
|
} else { |
898
|
|
|
// Use the next startxref from this $offset |
899
|
|
|
$startxref = (int) $startxrefMatches[0][1]; |
900
|
|
|
} |
901
|
|
|
|
902
|
66 |
|
if ($startxref > \strlen($pdfData)) { |
903
|
1 |
|
throw new \Exception('Unable to find xref (PDF corrupted?)'); |
904
|
|
|
} |
905
|
|
|
|
906
|
|
|
// check xref position |
907
|
65 |
|
if (strpos($pdfData, 'xref', $startxref) == $startxref) { |
908
|
|
|
// Cross-Reference |
909
|
52 |
|
$xref = $this->decodeXref($pdfData, $startxref, $xref); |
910
|
|
|
} else { |
911
|
|
|
// Check if the $pdfData might have the wrong line-endings |
912
|
13 |
|
$pdfDataUnix = str_replace("\r\n", "\n", $pdfData); |
913
|
13 |
|
if ($startxref < \strlen($pdfDataUnix) && strpos($pdfDataUnix, 'xref', $startxref) == $startxref) { |
914
|
|
|
// Return Unix-line-ending flag |
915
|
|
|
$xref = ['Unix' => true]; |
916
|
|
|
} else { |
917
|
|
|
// Cross-Reference Stream |
918
|
13 |
|
$xref = $this->decodeXrefStream($pdfData, $startxref, $xref); |
919
|
|
|
} |
920
|
|
|
} |
921
|
65 |
|
if (empty($xref)) { |
922
|
|
|
throw new \Exception('Unable to find xref'); |
923
|
|
|
} |
924
|
|
|
|
925
|
65 |
|
return $xref; |
926
|
|
|
} |
927
|
|
|
|
928
|
|
|
/** |
929
|
|
|
* Parses PDF data and returns extracted data as array. |
930
|
|
|
* |
931
|
|
|
* @param string $data PDF data to parse |
932
|
|
|
* |
933
|
|
|
* @return array array of parsed PDF document objects |
934
|
|
|
* |
935
|
|
|
* @throws \Exception if empty PDF data given |
936
|
|
|
* @throws \Exception if PDF data missing %PDF header |
937
|
|
|
*/ |
938
|
66 |
|
public function parseData(string $data): array |
939
|
|
|
{ |
940
|
66 |
|
if (empty($data)) { |
941
|
|
|
throw new \Exception('Empty PDF data given.'); |
942
|
|
|
} |
943
|
|
|
// find the pdf header starting position |
944
|
66 |
|
if (false === ($trimpos = strpos($data, '%PDF-'))) { |
945
|
|
|
throw new \Exception('Invalid PDF data: missing %PDF header.'); |
946
|
|
|
} |
947
|
|
|
|
948
|
|
|
// get PDF content string |
949
|
66 |
|
$pdfData = $trimpos > 0 ? substr($data, $trimpos) : $data; |
950
|
|
|
|
951
|
|
|
// get xref and trailer data |
952
|
66 |
|
$xref = $this->getXrefData($pdfData); |
953
|
|
|
|
954
|
|
|
// If we found Unix line-endings |
955
|
65 |
|
if (isset($xref['Unix'])) { |
956
|
|
|
$pdfData = str_replace("\r\n", "\n", $pdfData); |
957
|
|
|
$xref = $this->getXrefData($pdfData); |
958
|
|
|
} |
959
|
|
|
|
960
|
|
|
// parse all document objects |
961
|
65 |
|
$objects = []; |
962
|
65 |
|
foreach ($xref['xref'] as $obj => $offset) { |
963
|
65 |
|
if (!isset($objects[$obj]) && ($offset > 0)) { |
964
|
|
|
// decode objects with positive offset |
965
|
65 |
|
$objects[$obj] = $this->getIndirectObject($pdfData, $xref, $obj, $offset, true); |
966
|
|
|
} |
967
|
|
|
} |
968
|
|
|
|
969
|
65 |
|
return [$xref, $objects]; |
970
|
|
|
} |
971
|
|
|
} |
972
|
|
|
|
If the size of the collection does not change during the iteration, it is generally a good practice to compute it beforehand, and not on each iteration: