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
|
|
|
|
45
|
|
|
class RawDataParser |
46
|
|
|
{ |
47
|
|
|
/** |
48
|
|
|
* Configuration array. |
49
|
|
|
*/ |
50
|
|
|
protected $cfg = [ |
51
|
|
|
// if `true` ignore filter decoding errors |
52
|
|
|
'ignore_filter_decoding_errors' => true, |
53
|
|
|
// if `true` ignore missing filter decoding errors |
54
|
|
|
'ignore_missing_filter_decoders' => true, |
55
|
|
|
]; |
56
|
|
|
|
57
|
|
|
protected $filterHelper; |
58
|
|
|
protected $objects; |
59
|
|
|
|
60
|
|
|
/** |
61
|
|
|
* @param array $cfg Configuration array, default is [] |
62
|
|
|
*/ |
63
|
18 |
|
public function __construct($cfg = []) |
64
|
|
|
{ |
65
|
|
|
// merge given array with default values |
66
|
18 |
|
$this->cfg = array_merge($this->cfg, $cfg); |
67
|
|
|
|
68
|
18 |
|
$this->filterHelper = new FilterHelper(); |
69
|
18 |
|
} |
70
|
|
|
|
71
|
|
|
/** |
72
|
|
|
* Decode the specified stream. |
73
|
|
|
* |
74
|
|
|
* @param string $pdfData PDF data |
75
|
|
|
* @param array $xref |
76
|
|
|
* @param array $sdic Stream's dictionary array |
77
|
|
|
* @param string $stream Stream to decode |
78
|
|
|
* |
79
|
|
|
* @return array containing decoded stream data and remaining filters |
80
|
|
|
*/ |
81
|
17 |
|
protected function decodeStream($pdfData, $xref, $sdic, $stream) |
82
|
|
|
{ |
83
|
|
|
// get stream length and filters |
84
|
17 |
|
$slength = \strlen($stream); |
85
|
17 |
|
if ($slength <= 0) { |
86
|
|
|
return ['', []]; |
87
|
|
|
} |
88
|
17 |
|
$filters = []; |
89
|
17 |
|
foreach ($sdic as $k => $v) { |
90
|
17 |
|
if ('/' == $v[0]) { |
91
|
17 |
|
if (('Length' == $v[1]) and (isset($sdic[($k + 1)])) and ('numeric' == $sdic[($k + 1)][0])) { |
92
|
|
|
// get declared stream length |
93
|
17 |
|
$declength = (int) ($sdic[($k + 1)][1]); |
94
|
17 |
|
if ($declength < $slength) { |
95
|
17 |
|
$stream = substr($stream, 0, $declength); |
96
|
17 |
|
$slength = $declength; |
97
|
|
|
} |
98
|
17 |
|
} elseif (('Filter' == $v[1]) and (isset($sdic[($k + 1)]))) { |
99
|
|
|
// resolve indirect object |
100
|
17 |
|
$objval = $this->getObjectVal($pdfData, $xref, $sdic[($k + 1)]); |
101
|
17 |
|
if ('/' == $objval[0]) { |
102
|
|
|
// single filter |
103
|
17 |
|
$filters[] = $objval[1]; |
104
|
1 |
|
} elseif ('[' == $objval[0]) { |
105
|
|
|
// array of filters |
106
|
1 |
|
foreach ($objval[1] as $flt) { |
107
|
1 |
|
if ('/' == $flt[0]) { |
108
|
1 |
|
$filters[] = $flt[1]; |
109
|
|
|
} |
110
|
|
|
} |
111
|
|
|
} |
112
|
|
|
} |
113
|
|
|
} |
114
|
|
|
} |
115
|
|
|
|
116
|
|
|
// decode the stream |
117
|
17 |
|
$remaining_filters = []; |
118
|
17 |
|
foreach ($filters as $filter) { |
119
|
17 |
|
if (\in_array($filter, $this->filterHelper->getAvailableFilters())) { |
120
|
|
|
try { |
121
|
17 |
|
$stream = $this->filterHelper->decodeFilter($filter, $stream); |
122
|
1 |
|
} catch (Exception $e) { |
123
|
1 |
|
$emsg = $e->getMessage(); |
124
|
1 |
|
if ((('~' == $emsg[0]) && !$this->cfg['ignore_missing_filter_decoders']) |
125
|
1 |
|
|| (('~' != $emsg[0]) && !$this->cfg['ignore_filter_decoding_errors']) |
126
|
|
|
) { |
127
|
17 |
|
throw new Exception($e->getMessage()); |
128
|
|
|
} |
129
|
|
|
} |
130
|
|
|
} else { |
131
|
|
|
// add missing filter to array |
132
|
3 |
|
$remaining_filters[] = $filter; |
133
|
|
|
} |
134
|
|
|
} |
135
|
|
|
|
136
|
17 |
|
return [$stream, $remaining_filters]; |
137
|
|
|
} |
138
|
|
|
|
139
|
|
|
/** |
140
|
|
|
* Decode the Cross-Reference section |
141
|
|
|
* |
142
|
|
|
* @param string $pdfData PDF data |
143
|
|
|
* @param int $startxref Offset at which the xref section starts (position of the 'xref' keyword) |
144
|
|
|
* @param array $xref Previous xref array (if any) |
145
|
|
|
* |
146
|
|
|
* @return array containing xref and trailer data |
147
|
|
|
*/ |
148
|
16 |
|
protected function decodeXref($pdfData, $startxref, $xref = []) |
149
|
|
|
{ |
150
|
16 |
|
$startxref += 4; // 4 is the length of the word 'xref' |
151
|
|
|
// skip initial white space chars: \x00 null (NUL), \x09 horizontal tab (HT), \x0A line feed (LF), \x0C form feed (FF), \x0D carriage return (CR), \x20 space (SP) |
152
|
16 |
|
$offset = $startxref + strspn($pdfData, "\x00\x09\x0a\x0c\x0d\x20", $startxref); |
153
|
|
|
// initialize object number |
154
|
16 |
|
$obj_num = 0; |
155
|
|
|
// search for cross-reference entries or subsection |
156
|
16 |
|
while (preg_match('/([0-9]+)[\x20]([0-9]+)[\x20]?([nf]?)(\r\n|[\x20]?[\r\n])/', $pdfData, $matches, PREG_OFFSET_CAPTURE, $offset) > 0) { |
157
|
16 |
|
if ($matches[0][1] != $offset) { |
158
|
|
|
// we are on another section |
159
|
3 |
|
break; |
160
|
|
|
} |
161
|
16 |
|
$offset += \strlen($matches[0][0]); |
162
|
16 |
|
if ('n' == $matches[3][0]) { |
163
|
|
|
// create unique object index: [object number]_[generation number] |
164
|
16 |
|
$index = $obj_num.'_'.(int) ($matches[2][0]); |
165
|
|
|
// check if object already exist |
166
|
16 |
|
if (!isset($xref['xref'][$index])) { |
167
|
|
|
// store object offset position |
168
|
16 |
|
$xref['xref'][$index] = (int) ($matches[1][0]); |
169
|
|
|
} |
170
|
16 |
|
++$obj_num; |
171
|
16 |
|
} elseif ('f' == $matches[3][0]) { |
172
|
16 |
|
++$obj_num; |
173
|
|
|
} else { |
174
|
|
|
// object number (index) |
175
|
16 |
|
$obj_num = (int) ($matches[1][0]); |
176
|
|
|
} |
177
|
|
|
} |
178
|
|
|
// get trailer data |
179
|
16 |
|
if (preg_match('/trailer[\s]*<<(.*)>>/isU', $pdfData, $matches, PREG_OFFSET_CAPTURE, $offset) > 0) { |
180
|
16 |
|
$trailer_data = $matches[1][0]; |
181
|
16 |
|
if (!isset($xref['trailer']) or empty($xref['trailer'])) { |
182
|
|
|
// get only the last updated version |
183
|
16 |
|
$xref['trailer'] = []; |
184
|
|
|
// parse trailer_data |
185
|
16 |
|
if (preg_match('/Size[\s]+([0-9]+)/i', $trailer_data, $matches) > 0) { |
186
|
16 |
|
$xref['trailer']['size'] = (int) ($matches[1]); |
187
|
|
|
} |
188
|
16 |
|
if (preg_match('/Root[\s]+([0-9]+)[\s]+([0-9]+)[\s]+R/i', $trailer_data, $matches) > 0) { |
189
|
16 |
|
$xref['trailer']['root'] = (int) ($matches[1]).'_'.(int) ($matches[2]); |
190
|
|
|
} |
191
|
16 |
|
if (preg_match('/Encrypt[\s]+([0-9]+)[\s]+([0-9]+)[\s]+R/i', $trailer_data, $matches) > 0) { |
192
|
|
|
$xref['trailer']['encrypt'] = (int) ($matches[1]).'_'.(int) ($matches[2]); |
193
|
|
|
} |
194
|
16 |
|
if (preg_match('/Info[\s]+([0-9]+)[\s]+([0-9]+)[\s]+R/i', $trailer_data, $matches) > 0) { |
195
|
16 |
|
$xref['trailer']['info'] = (int) ($matches[1]).'_'.(int) ($matches[2]); |
196
|
|
|
} |
197
|
16 |
|
if (preg_match('/ID[\s]*[\[][\s]*[<]([^>]*)[>][\s]*[<]([^>]*)[>]/i', $trailer_data, $matches) > 0) { |
198
|
15 |
|
$xref['trailer']['id'] = []; |
199
|
15 |
|
$xref['trailer']['id'][0] = $matches[1]; |
200
|
15 |
|
$xref['trailer']['id'][1] = $matches[2]; |
201
|
|
|
} |
202
|
|
|
} |
203
|
16 |
|
if (preg_match('/Prev[\s]+([0-9]+)/i', $trailer_data, $matches) > 0) { |
204
|
|
|
// get previous xref |
205
|
16 |
|
$xref = $this->getXrefData($pdfData, (int) ($matches[1]), $xref); |
206
|
|
|
} |
207
|
|
|
} else { |
208
|
|
|
throw new Exception('Unable to find trailer'); |
209
|
|
|
} |
210
|
|
|
|
211
|
16 |
|
return $xref; |
212
|
|
|
} |
213
|
|
|
|
214
|
|
|
/** |
215
|
|
|
* Decode the Cross-Reference Stream section |
216
|
|
|
* |
217
|
|
|
* @param string $pdfData PDF data |
218
|
|
|
* @param int $startxref Offset at which the xref section starts |
219
|
|
|
* @param array $xref Previous xref array (if any) |
220
|
|
|
* |
221
|
|
|
* @return array containing xref and trailer data |
222
|
|
|
* |
223
|
|
|
* @throws Exception if unknown PNG predictor detected |
224
|
|
|
*/ |
225
|
2 |
|
protected function decodeXrefStream($pdfData, $startxref, $xref = []) |
226
|
|
|
{ |
227
|
|
|
// try to read Cross-Reference Stream |
228
|
2 |
|
$xrefobj = $this->getRawObject($pdfData, $startxref); |
229
|
2 |
|
$xrefcrs = $this->getIndirectObject($pdfData, $xref, $xrefobj[1], $startxref, true); |
230
|
2 |
|
if (!isset($xref['trailer']) or empty($xref['trailer'])) { |
231
|
|
|
// get only the last updated version |
232
|
2 |
|
$xref['trailer'] = []; |
233
|
2 |
|
$filltrailer = true; |
234
|
|
|
} else { |
235
|
2 |
|
$filltrailer = false; |
236
|
|
|
} |
237
|
2 |
|
if (!isset($xref['xref'])) { |
238
|
2 |
|
$xref['xref'] = []; |
239
|
|
|
} |
240
|
2 |
|
$valid_crs = false; |
241
|
2 |
|
$columns = 0; |
242
|
2 |
|
$sarr = $xrefcrs[0][1]; |
243
|
2 |
|
if (!\is_array($sarr)) { |
244
|
|
|
$sarr = []; |
245
|
|
|
} |
246
|
|
|
|
247
|
2 |
|
$wb = []; |
248
|
|
|
|
249
|
2 |
|
foreach ($sarr as $k => $v) { |
250
|
|
|
if ( |
251
|
2 |
|
('/' == $v[0]) |
252
|
2 |
|
&& ('Type' == $v[1]) |
253
|
|
|
&& ( |
254
|
2 |
|
isset($sarr[($k + 1)]) |
255
|
2 |
|
&& '/' == $sarr[($k + 1)][0] |
256
|
2 |
|
&& 'XRef' == $sarr[($k + 1)][1] |
257
|
|
|
) |
258
|
|
|
) { |
259
|
2 |
|
$valid_crs = true; |
260
|
2 |
|
} elseif (('/' == $v[0]) and ('Index' == $v[1]) and (isset($sarr[($k + 1)]))) { |
261
|
|
|
// first object number in the subsection |
262
|
2 |
|
$index_first = (int) ($sarr[($k + 1)][1][0][1]); |
263
|
2 |
|
} elseif (('/' == $v[0]) and ('Prev' == $v[1]) and (isset($sarr[($k + 1)]) and ('numeric' == $sarr[($k + 1)][0]))) { |
264
|
|
|
// get previous xref offset |
265
|
2 |
|
$prevxref = (int) ($sarr[($k + 1)][1]); |
266
|
2 |
|
} elseif (('/' == $v[0]) and ('W' == $v[1]) and (isset($sarr[($k + 1)]))) { |
267
|
|
|
// number of bytes (in the decoded stream) of the corresponding field |
268
|
2 |
|
$wb[0] = (int) ($sarr[($k + 1)][1][0][1]); |
269
|
2 |
|
$wb[1] = (int) ($sarr[($k + 1)][1][1][1]); |
270
|
2 |
|
$wb[2] = (int) ($sarr[($k + 1)][1][2][1]); |
271
|
2 |
|
} elseif (('/' == $v[0]) and ('DecodeParms' == $v[1]) and (isset($sarr[($k + 1)][1]))) { |
272
|
2 |
|
$decpar = $sarr[($k + 1)][1]; |
273
|
2 |
|
foreach ($decpar as $kdc => $vdc) { |
274
|
|
|
if ( |
275
|
2 |
|
'/' == $vdc[0] |
276
|
2 |
|
&& 'Columns' == $vdc[1] |
277
|
|
|
&& ( |
278
|
2 |
|
isset($decpar[($kdc + 1)]) |
279
|
2 |
|
&& 'numeric' == $decpar[($kdc + 1)][0] |
280
|
|
|
) |
281
|
|
|
) { |
282
|
2 |
|
$columns = (int) ($decpar[($kdc + 1)][1]); |
283
|
|
|
} elseif ( |
284
|
2 |
|
'/' == $vdc[0] |
285
|
2 |
|
&& 'Predictor' == $vdc[1] |
286
|
|
|
&& ( |
287
|
2 |
|
isset($decpar[($kdc + 1)]) |
288
|
2 |
|
&& 'numeric' == $decpar[($kdc + 1)][0] |
289
|
|
|
) |
290
|
|
|
) { |
291
|
2 |
|
$predictor = (int) ($decpar[($kdc + 1)][1]); |
|
|
|
|
292
|
|
|
} |
293
|
|
|
} |
294
|
2 |
|
} elseif ($filltrailer) { |
295
|
2 |
|
if (('/' == $v[0]) and ('Size' == $v[1]) and (isset($sarr[($k + 1)]) and ('numeric' == $sarr[($k + 1)][0]))) { |
296
|
2 |
|
$xref['trailer']['size'] = $sarr[($k + 1)][1]; |
297
|
2 |
|
} elseif (('/' == $v[0]) and ('Root' == $v[1]) and (isset($sarr[($k + 1)]) and ('objref' == $sarr[($k + 1)][0]))) { |
298
|
2 |
|
$xref['trailer']['root'] = $sarr[($k + 1)][1]; |
299
|
2 |
|
} elseif (('/' == $v[0]) and ('Info' == $v[1]) and (isset($sarr[($k + 1)]) and ('objref' == $sarr[($k + 1)][0]))) { |
300
|
2 |
|
$xref['trailer']['info'] = $sarr[($k + 1)][1]; |
301
|
2 |
|
} elseif (('/' == $v[0]) and ('Encrypt' == $v[1]) and (isset($sarr[($k + 1)]) and ('objref' == $sarr[($k + 1)][0]))) { |
302
|
|
|
$xref['trailer']['encrypt'] = $sarr[($k + 1)][1]; |
303
|
2 |
|
} elseif (('/' == $v[0]) and ('ID' == $v[1]) and (isset($sarr[($k + 1)]))) { |
304
|
2 |
|
$xref['trailer']['id'] = []; |
305
|
2 |
|
$xref['trailer']['id'][0] = $sarr[($k + 1)][1][0][1]; |
306
|
2 |
|
$xref['trailer']['id'][1] = $sarr[($k + 1)][1][1][1]; |
307
|
|
|
} |
308
|
|
|
} |
309
|
|
|
} |
310
|
|
|
|
311
|
|
|
// decode data |
312
|
2 |
|
if ($valid_crs and isset($xrefcrs[1][3][0])) { |
313
|
|
|
// number of bytes in a row |
314
|
2 |
|
$rowlen = ($columns + 1); |
315
|
|
|
// convert the stream into an array of integers |
316
|
2 |
|
$sdata = unpack('C*', $xrefcrs[1][3][0]); |
317
|
|
|
// split the rows |
318
|
2 |
|
$sdata = array_chunk($sdata, $rowlen); |
|
|
|
|
319
|
|
|
// initialize decoded array |
320
|
2 |
|
$ddata = []; |
321
|
|
|
// initialize first row with zeros |
322
|
2 |
|
$prev_row = array_fill(0, $rowlen, 0); |
323
|
|
|
// for each row apply PNG unpredictor |
324
|
2 |
|
foreach ($sdata as $k => $row) { |
325
|
|
|
// initialize new row |
326
|
2 |
|
$ddata[$k] = []; |
327
|
|
|
// get PNG predictor value |
328
|
2 |
|
$predictor = (10 + $row[0]); |
329
|
|
|
// for each byte on the row |
330
|
2 |
|
for ($i = 1; $i <= $columns; ++$i) { |
331
|
|
|
// new index |
332
|
2 |
|
$j = ($i - 1); |
333
|
2 |
|
$row_up = $prev_row[$j]; |
334
|
2 |
|
if (1 == $i) { |
335
|
2 |
|
$row_left = 0; |
336
|
2 |
|
$row_upleft = 0; |
337
|
|
|
} else { |
338
|
2 |
|
$row_left = $row[($i - 1)]; |
339
|
2 |
|
$row_upleft = $prev_row[($j - 1)]; |
340
|
|
|
} |
341
|
2 |
|
switch ($predictor) { |
342
|
2 |
|
case 10: // PNG prediction (on encoding, PNG None on all rows) |
343
|
|
|
$ddata[$k][$j] = $row[$i]; |
344
|
|
|
break; |
345
|
|
|
|
346
|
2 |
|
case 11: // PNG prediction (on encoding, PNG Sub on all rows) |
347
|
|
|
$ddata[$k][$j] = (($row[$i] + $row_left) & 0xff); |
348
|
|
|
break; |
349
|
|
|
|
350
|
2 |
|
case 12: // PNG prediction (on encoding, PNG Up on all rows) |
351
|
2 |
|
$ddata[$k][$j] = (($row[$i] + $row_up) & 0xff); |
352
|
2 |
|
break; |
353
|
|
|
|
354
|
|
|
case 13: // PNG prediction (on encoding, PNG Average on all rows) |
355
|
|
|
$ddata[$k][$j] = (($row[$i] + (($row_left + $row_up) / 2)) & 0xff); |
356
|
|
|
break; |
357
|
|
|
|
358
|
|
|
case 14: // PNG prediction (on encoding, PNG Paeth on all rows) |
359
|
|
|
// initial estimate |
360
|
|
|
$p = ($row_left + $row_up - $row_upleft); |
361
|
|
|
// distances |
362
|
|
|
$pa = abs($p - $row_left); |
363
|
|
|
$pb = abs($p - $row_up); |
364
|
|
|
$pc = abs($p - $row_upleft); |
365
|
|
|
$pmin = min($pa, $pb, $pc); |
366
|
|
|
// return minimum distance |
367
|
|
|
switch ($pmin) { |
368
|
|
|
case $pa: |
369
|
|
|
$ddata[$k][$j] = (($row[$i] + $row_left) & 0xff); |
370
|
|
|
break; |
371
|
|
|
|
372
|
|
|
case $pb: |
373
|
|
|
$ddata[$k][$j] = (($row[$i] + $row_up) & 0xff); |
374
|
|
|
break; |
375
|
|
|
|
376
|
|
|
case $pc: |
377
|
|
|
$ddata[$k][$j] = (($row[$i] + $row_upleft) & 0xff); |
378
|
|
|
break; |
379
|
|
|
} |
380
|
|
|
break; |
381
|
|
|
|
382
|
|
|
default: // PNG prediction (on encoding, PNG optimum) |
383
|
|
|
throw new Exception('Unknown PNG predictor'); |
384
|
|
|
} |
385
|
|
|
} |
386
|
2 |
|
$prev_row = $ddata[$k]; |
387
|
|
|
} // end for each row |
388
|
|
|
// complete decoding |
389
|
2 |
|
$sdata = []; |
390
|
|
|
// for every row |
391
|
2 |
|
foreach ($ddata as $k => $row) { |
392
|
|
|
// initialize new row |
393
|
2 |
|
$sdata[$k] = [0, 0, 0]; |
394
|
2 |
|
if (0 == $wb[0]) { |
395
|
|
|
// default type field |
396
|
|
|
$sdata[$k][0] = 1; |
397
|
|
|
} |
398
|
2 |
|
$i = 0; // count bytes in the row |
399
|
|
|
// for every column |
400
|
2 |
|
for ($c = 0; $c < 3; ++$c) { |
401
|
|
|
// for every byte on the column |
402
|
2 |
|
for ($b = 0; $b < $wb[$c]; ++$b) { |
403
|
2 |
|
if (isset($row[$i])) { |
404
|
2 |
|
$sdata[$k][$c] += ($row[$i] << (($wb[$c] - 1 - $b) * 8)); |
405
|
|
|
} |
406
|
2 |
|
++$i; |
407
|
|
|
} |
408
|
|
|
} |
409
|
|
|
} |
410
|
2 |
|
$ddata = []; |
|
|
|
|
411
|
|
|
// fill xref |
412
|
2 |
|
if (isset($index_first)) { |
413
|
2 |
|
$obj_num = $index_first; |
414
|
|
|
} else { |
415
|
2 |
|
$obj_num = 0; |
416
|
|
|
} |
417
|
2 |
|
foreach ($sdata as $k => $row) { |
418
|
2 |
|
switch ($row[0]) { |
419
|
2 |
|
case 0: // (f) linked list of free objects |
420
|
2 |
|
break; |
421
|
|
|
|
422
|
2 |
|
case 1: // (n) objects that are in use but are not compressed |
423
|
|
|
// create unique object index: [object number]_[generation number] |
424
|
2 |
|
$index = $obj_num.'_'.$row[2]; |
425
|
|
|
// check if object already exist |
426
|
2 |
|
if (!isset($xref['xref'][$index])) { |
427
|
|
|
// store object offset position |
428
|
2 |
|
$xref['xref'][$index] = $row[1]; |
429
|
|
|
} |
430
|
2 |
|
break; |
431
|
|
|
|
432
|
2 |
|
case 2: // compressed objects |
433
|
|
|
// $row[1] = object number of the object stream in which this object is stored |
434
|
|
|
// $row[2] = index of this object within the object stream |
435
|
2 |
|
$index = $row[1].'_0_'.$row[2]; |
436
|
2 |
|
$xref['xref'][$index] = -1; |
437
|
2 |
|
break; |
438
|
|
|
|
439
|
|
|
default: // null objects |
440
|
|
|
break; |
441
|
|
|
} |
442
|
2 |
|
++$obj_num; |
443
|
|
|
} |
444
|
|
|
} // end decoding data |
445
|
2 |
|
if (isset($prevxref)) { |
446
|
|
|
// get previous xref |
447
|
2 |
|
$xref = $this->getXrefData($pdfData, $prevxref, $xref); |
448
|
|
|
} |
449
|
|
|
|
450
|
2 |
|
return $xref; |
451
|
|
|
} |
452
|
|
|
|
453
|
|
|
/** |
454
|
|
|
* Get content of indirect object. |
455
|
|
|
* |
456
|
|
|
* @param string $pdfData PDF data |
457
|
|
|
* @param array $xref |
458
|
|
|
* @param string $objRef Object number and generation number separated by underscore character |
459
|
|
|
* @param int $offset Object offset |
460
|
|
|
* @param bool $decoding If true decode streams |
461
|
|
|
* |
462
|
|
|
* @return array containing object data |
463
|
|
|
* |
464
|
|
|
* @throws Exception if invalid object reference found |
465
|
|
|
*/ |
466
|
17 |
|
protected function getIndirectObject($pdfData, $xref, $objRef, $offset = 0, $decoding = true) |
467
|
|
|
{ |
468
|
|
|
/* |
469
|
|
|
* build indirect object header |
470
|
|
|
*/ |
471
|
|
|
// $objHeader = "[object number] [generation number] obj" |
472
|
17 |
|
$objRefArr = explode('_', $objRef); |
473
|
17 |
|
if (2 !== \count($objRefArr)) { |
474
|
|
|
throw new Exception('Invalid object reference for $obj.'); |
475
|
|
|
} |
476
|
17 |
|
$objHeader = $objRefArr[0].' '.$objRefArr[1].' obj'; |
477
|
|
|
|
478
|
|
|
/* |
479
|
|
|
* check if we are in position |
480
|
|
|
*/ |
481
|
|
|
// ignore whitespace characters at offset (NUL, HT, LF, FF, CR, SP) |
482
|
17 |
|
$offset += strspn($pdfData, "\0\t\n\f\r ", $offset); |
483
|
|
|
// ignore leading zeros for object number |
484
|
17 |
|
$offset += strspn($pdfData, '0', $offset); |
485
|
17 |
|
if (substr($pdfData, $offset, \strlen($objHeader)) !== $objHeader) { |
486
|
|
|
// an indirect reference to an undefined object shall be considered a reference to the null object |
487
|
|
|
return ['null', 'null', $offset]; |
488
|
|
|
} |
489
|
|
|
|
490
|
|
|
/* |
491
|
|
|
* get content |
492
|
|
|
*/ |
493
|
|
|
// starting position of object content |
494
|
17 |
|
$offset += \strlen($objHeader); |
495
|
17 |
|
$objContentArr = []; |
496
|
17 |
|
$i = 0; // object main index |
497
|
|
|
do { |
498
|
17 |
|
$oldOffset = $offset; |
499
|
|
|
// get element |
500
|
17 |
|
$element = $this->getRawObject($pdfData, $offset); |
501
|
17 |
|
$offset = $element[2]; |
502
|
|
|
// decode stream using stream's dictionary information |
503
|
17 |
|
if ($decoding && ('stream' === $element[0]) && (isset($objContentArr[($i - 1)][0])) && ('<<' === $objContentArr[($i - 1)][0])) { |
504
|
17 |
|
$element[3] = $this->decodeStream($pdfData, $xref, $objContentArr[($i - 1)][1], $element[1]); |
505
|
|
|
} |
506
|
17 |
|
$objContentArr[$i] = $element; |
507
|
17 |
|
++$i; |
508
|
17 |
|
} while (('endobj' !== $element[0]) && ($offset !== $oldOffset)); |
509
|
|
|
// remove closing delimiter |
510
|
17 |
|
array_pop($objContentArr); |
511
|
|
|
|
512
|
|
|
/* |
513
|
|
|
* return raw object content |
514
|
|
|
*/ |
515
|
17 |
|
return $objContentArr; |
516
|
|
|
} |
517
|
|
|
|
518
|
|
|
/** |
519
|
|
|
* Get the content of object, resolving indirect object reference if necessary. |
520
|
|
|
* |
521
|
|
|
* @param string $pdfData PDF data |
522
|
|
|
* @param array $obj Object value |
523
|
|
|
* |
524
|
|
|
* @return array containing object data |
525
|
|
|
* |
526
|
|
|
* @throws Exception |
527
|
|
|
*/ |
528
|
17 |
|
protected function getObjectVal($pdfData, $xref, $obj) |
529
|
|
|
{ |
530
|
17 |
|
if ('objref' == $obj[0]) { |
531
|
|
|
// reference to indirect object |
532
|
|
|
if (isset($this->objects[$obj[1]])) { |
533
|
|
|
// this object has been already parsed |
534
|
|
|
return $this->objects[$obj[1]]; |
535
|
|
|
} elseif (isset($xref[$obj[1]])) { |
536
|
|
|
// parse new object |
537
|
|
|
$this->objects[$obj[1]] = $this->getIndirectObject($pdfData, $xref, $obj[1], $xref[$obj[1]], false); |
538
|
|
|
|
539
|
|
|
return $this->objects[$obj[1]]; |
540
|
|
|
} |
541
|
|
|
} |
542
|
|
|
|
543
|
17 |
|
return $obj; |
544
|
|
|
} |
545
|
|
|
|
546
|
|
|
/** |
547
|
|
|
* Get object type, raw value and offset to next object |
548
|
|
|
* |
549
|
|
|
* @param int $offset Object offset |
550
|
|
|
* |
551
|
|
|
* @return array containing object type, raw value and offset to next object |
552
|
|
|
*/ |
553
|
17 |
|
protected function getRawObject($pdfData, $offset = 0) |
554
|
|
|
{ |
555
|
17 |
|
$objtype = ''; // object type to be returned |
556
|
17 |
|
$objval = ''; // object value to be returned |
557
|
|
|
|
558
|
|
|
/* |
559
|
|
|
* skip initial white space chars: |
560
|
|
|
* \x00 null (NUL) |
561
|
|
|
* \x09 horizontal tab (HT) |
562
|
|
|
* \x0A line feed (LF) |
563
|
|
|
* \x0C form feed (FF) |
564
|
|
|
* \x0D carriage return (CR) |
565
|
|
|
* \x20 space (SP) |
566
|
|
|
*/ |
567
|
17 |
|
$offset += strspn($pdfData, "\x00\x09\x0a\x0c\x0d\x20", $offset); |
568
|
|
|
|
569
|
|
|
// get first char |
570
|
17 |
|
$char = $pdfData[$offset]; |
571
|
|
|
// get object type |
572
|
17 |
|
switch ($char) { |
573
|
17 |
|
case '%': // \x25 PERCENT SIGN |
574
|
|
|
// skip comment and search for next token |
575
|
|
|
$next = strcspn($pdfData, "\r\n", $offset); |
576
|
|
|
if ($next > 0) { |
577
|
|
|
$offset += $next; |
578
|
|
|
|
579
|
|
|
return $this->getRawObject($pdfData, $offset); |
580
|
|
|
} |
581
|
|
|
break; |
582
|
|
|
|
583
|
17 |
|
case '/': // \x2F SOLIDUS |
584
|
|
|
// name object |
585
|
17 |
|
$objtype = $char; |
586
|
17 |
|
++$offset; |
587
|
17 |
|
$pregResult = preg_match( |
588
|
17 |
|
'/^([^\x00\x09\x0a\x0c\x0d\x20\s\x28\x29\x3c\x3e\x5b\x5d\x7b\x7d\x2f\x25]+)/', |
589
|
17 |
|
substr($pdfData, $offset, 256), |
590
|
|
|
$matches |
591
|
|
|
); |
592
|
17 |
|
if (1 == $pregResult) { |
593
|
17 |
|
$objval = $matches[1]; // unescaped value |
594
|
17 |
|
$offset += \strlen($objval); |
595
|
|
|
} |
596
|
17 |
|
break; |
597
|
|
|
|
598
|
17 |
|
case '(': // \x28 LEFT PARENTHESIS |
599
|
17 |
|
case ')': // \x29 RIGHT PARENTHESIS |
600
|
|
|
// literal string object |
601
|
17 |
|
$objtype = $char; |
602
|
17 |
|
++$offset; |
603
|
17 |
|
$strpos = $offset; |
604
|
17 |
|
if ('(' == $char) { |
605
|
17 |
|
$open_bracket = 1; |
606
|
17 |
|
while ($open_bracket > 0) { |
607
|
17 |
|
if (!isset($pdfData[$strpos])) { |
608
|
|
|
break; |
609
|
|
|
} |
610
|
17 |
|
$ch = $pdfData[$strpos]; |
611
|
17 |
|
switch ($ch) { |
612
|
17 |
|
case '\\': // REVERSE SOLIDUS (5Ch) (Backslash) |
613
|
|
|
// skip next character |
614
|
14 |
|
++$strpos; |
615
|
14 |
|
break; |
616
|
|
|
|
617
|
17 |
|
case '(': // LEFT PARENHESIS (28h) |
618
|
|
|
++$open_bracket; |
619
|
|
|
break; |
620
|
|
|
|
621
|
17 |
|
case ')': // RIGHT PARENTHESIS (29h) |
622
|
17 |
|
--$open_bracket; |
623
|
17 |
|
break; |
624
|
|
|
} |
625
|
17 |
|
++$strpos; |
626
|
|
|
} |
627
|
17 |
|
$objval = substr($pdfData, $offset, ($strpos - $offset - 1)); |
628
|
17 |
|
$offset = $strpos; |
629
|
|
|
} |
630
|
17 |
|
break; |
631
|
|
|
|
632
|
17 |
|
case '[': // \x5B LEFT SQUARE BRACKET |
633
|
17 |
|
case ']': // \x5D RIGHT SQUARE BRACKET |
634
|
|
|
// array object |
635
|
17 |
|
$objtype = $char; |
636
|
17 |
|
++$offset; |
637
|
17 |
|
if ('[' == $char) { |
638
|
|
|
// get array content |
639
|
17 |
|
$objval = []; |
640
|
|
|
do { |
641
|
|
|
// get element |
642
|
17 |
|
$element = $this->getRawObject($pdfData, $offset); |
643
|
17 |
|
$offset = $element[2]; |
644
|
17 |
|
$objval[] = $element; |
645
|
17 |
|
} while (']' != $element[0]); |
646
|
|
|
// remove closing delimiter |
647
|
17 |
|
array_pop($objval); |
648
|
|
|
} |
649
|
17 |
|
break; |
650
|
|
|
|
651
|
17 |
|
case '<': // \x3C LESS-THAN SIGN |
652
|
17 |
|
case '>': // \x3E GREATER-THAN SIGN |
653
|
17 |
|
if (isset($pdfData[($offset + 1)]) and ($pdfData[($offset + 1)] == $char)) { |
654
|
|
|
// dictionary object |
655
|
17 |
|
$objtype = $char.$char; |
656
|
17 |
|
$offset += 2; |
657
|
17 |
|
if ('<' == $char) { |
658
|
|
|
// get array content |
659
|
17 |
|
$objval = []; |
660
|
|
|
do { |
661
|
|
|
// get element |
662
|
17 |
|
$element = $this->getRawObject($pdfData, $offset); |
663
|
17 |
|
$offset = $element[2]; |
664
|
17 |
|
$objval[] = $element; |
665
|
17 |
|
} while ('>>' != $element[0]); |
666
|
|
|
// remove closing delimiter |
667
|
17 |
|
array_pop($objval); |
668
|
|
|
} |
669
|
|
|
} else { |
670
|
|
|
// hexadecimal string object |
671
|
4 |
|
$objtype = $char; |
672
|
4 |
|
++$offset; |
673
|
4 |
|
$pregResult = preg_match( |
674
|
4 |
|
'/^([0-9A-Fa-f\x09\x0a\x0c\x0d\x20]+)>/iU', |
675
|
4 |
|
substr($pdfData, $offset), |
676
|
|
|
$matches |
677
|
|
|
); |
678
|
4 |
|
if (('<' == $char) && 1 == $pregResult) { |
679
|
|
|
// remove white space characters |
680
|
4 |
|
$objval = strtr($matches[1], "\x09\x0a\x0c\x0d\x20", ''); |
681
|
4 |
|
$offset += \strlen($matches[0]); |
682
|
|
|
} elseif (false !== ($endpos = strpos($pdfData, '>', $offset))) { |
683
|
|
|
$offset = $endpos + 1; |
684
|
|
|
} |
685
|
|
|
} |
686
|
17 |
|
break; |
687
|
|
|
|
688
|
|
|
default: |
689
|
17 |
|
if ('endobj' == substr($pdfData, $offset, 6)) { |
690
|
|
|
// indirect object |
691
|
17 |
|
$objtype = 'endobj'; |
692
|
17 |
|
$offset += 6; |
693
|
17 |
|
} elseif ('null' == substr($pdfData, $offset, 4)) { |
694
|
|
|
// null object |
695
|
2 |
|
$objtype = 'null'; |
696
|
2 |
|
$offset += 4; |
697
|
2 |
|
$objval = 'null'; |
698
|
17 |
|
} elseif ('true' == substr($pdfData, $offset, 4)) { |
699
|
|
|
// boolean true object |
700
|
4 |
|
$objtype = 'boolean'; |
701
|
4 |
|
$offset += 4; |
702
|
4 |
|
$objval = 'true'; |
703
|
17 |
|
} elseif ('false' == substr($pdfData, $offset, 5)) { |
704
|
|
|
// boolean false object |
705
|
1 |
|
$objtype = 'boolean'; |
706
|
1 |
|
$offset += 5; |
707
|
1 |
|
$objval = 'false'; |
708
|
17 |
|
} elseif ('stream' == substr($pdfData, $offset, 6)) { |
709
|
|
|
// start stream object |
710
|
17 |
|
$objtype = 'stream'; |
711
|
17 |
|
$offset += 6; |
712
|
17 |
|
if (1 == preg_match('/^([\r]?[\n])/isU', substr($pdfData, $offset), $matches)) { |
713
|
17 |
|
$offset += \strlen($matches[0]); |
714
|
17 |
|
$pregResult = preg_match( |
715
|
17 |
|
'/(endstream)[\x09\x0a\x0c\x0d\x20]/isU', |
716
|
17 |
|
substr($pdfData, $offset), |
717
|
|
|
$matches, |
718
|
17 |
|
PREG_OFFSET_CAPTURE |
719
|
|
|
); |
720
|
17 |
|
if (1 == $pregResult) { |
721
|
17 |
|
$objval = substr($pdfData, $offset, $matches[0][1]); |
722
|
17 |
|
$offset += $matches[1][1]; |
723
|
|
|
} |
724
|
|
|
} |
725
|
17 |
|
} elseif ('endstream' == substr($pdfData, $offset, 9)) { |
726
|
|
|
// end stream object |
727
|
17 |
|
$objtype = 'endstream'; |
728
|
17 |
|
$offset += 9; |
729
|
17 |
|
} elseif (1 == preg_match('/^([0-9]+)[\s]+([0-9]+)[\s]+R/iU', substr($pdfData, $offset, 33), $matches)) { |
730
|
|
|
// indirect object reference |
731
|
17 |
|
$objtype = 'objref'; |
732
|
17 |
|
$offset += \strlen($matches[0]); |
733
|
17 |
|
$objval = (int) ($matches[1]).'_'.(int) ($matches[2]); |
734
|
17 |
|
} elseif (1 == preg_match('/^([0-9]+)[\s]+([0-9]+)[\s]+obj/iU', substr($pdfData, $offset, 33), $matches)) { |
735
|
|
|
// object start |
736
|
2 |
|
$objtype = 'obj'; |
737
|
2 |
|
$objval = (int) ($matches[1]).'_'.(int) ($matches[2]); |
738
|
2 |
|
$offset += \strlen($matches[0]); |
739
|
17 |
|
} elseif (($numlen = strspn($pdfData, '+-.0123456789', $offset)) > 0) { |
740
|
|
|
// numeric object |
741
|
17 |
|
$objtype = 'numeric'; |
742
|
17 |
|
$objval = substr($pdfData, $offset, $numlen); |
743
|
17 |
|
$offset += $numlen; |
744
|
|
|
} |
745
|
17 |
|
break; |
746
|
|
|
} |
747
|
|
|
|
748
|
17 |
|
return [$objtype, $objval, $offset]; |
749
|
|
|
} |
750
|
|
|
|
751
|
|
|
/** |
752
|
|
|
* Get Cross-Reference (xref) table and trailer data from PDF document data. |
753
|
|
|
* |
754
|
|
|
* @param string $pdfData |
755
|
|
|
* @param int $offset xref offset (if know) |
756
|
|
|
* @param array $xref previous xref array (if any) |
757
|
|
|
* |
758
|
|
|
* @return array containing xref and trailer data |
759
|
|
|
* |
760
|
|
|
* @throws Exception if it was unable to find startxref |
761
|
|
|
* @throws Exception if it was unable to find xref |
762
|
|
|
*/ |
763
|
18 |
|
protected function getXrefData($pdfData, $offset = 0, $xref = []) |
764
|
|
|
{ |
765
|
18 |
|
$startxrefPreg = preg_match( |
766
|
18 |
|
'/[\r\n]startxref[\s]*[\r\n]+([0-9]+)[\s]*[\r\n]+%%EOF/i', |
767
|
|
|
$pdfData, |
768
|
|
|
$matches, |
769
|
18 |
|
PREG_OFFSET_CAPTURE, |
770
|
|
|
$offset |
771
|
|
|
); |
772
|
|
|
|
773
|
18 |
|
if (0 == $offset) { |
774
|
|
|
// find last startxref |
775
|
18 |
|
$pregResult = preg_match_all( |
776
|
18 |
|
'/[\r\n]startxref[\s]*[\r\n]+([0-9]+)[\s]*[\r\n]+%%EOF/i', |
777
|
|
|
$pdfData, $matches, |
778
|
18 |
|
PREG_SET_ORDER, |
779
|
|
|
$offset |
780
|
|
|
); |
781
|
18 |
|
if (0 == $pregResult) { |
782
|
|
|
throw new Exception('Unable to find startxref'); |
783
|
|
|
} |
784
|
18 |
|
$matches = array_pop($matches); |
785
|
18 |
|
$startxref = $matches[1]; |
786
|
4 |
|
} elseif (strpos($pdfData, 'xref', $offset) == $offset) { |
787
|
|
|
// Already pointing at the xref table |
788
|
3 |
|
$startxref = $offset; |
789
|
2 |
|
} elseif (preg_match('/([0-9]+[\s][0-9]+[\s]obj)/i', $pdfData, $matches, PREG_OFFSET_CAPTURE, $offset)) { |
790
|
|
|
// Cross-Reference Stream object |
791
|
2 |
|
$startxref = $offset; |
792
|
|
|
} elseif ($startxrefPreg) { |
793
|
|
|
// startxref found |
794
|
|
|
$startxref = $matches[1][0]; |
795
|
|
|
} else { |
796
|
|
|
throw new Exception('Unable to find startxref'); |
797
|
|
|
} |
798
|
|
|
|
799
|
18 |
|
if ($startxref > \strlen($pdfData)) { |
800
|
1 |
|
throw new Exception('Unable to find xref (PDF corrupted?)'); |
801
|
|
|
} |
802
|
|
|
|
803
|
|
|
// check xref position |
804
|
17 |
|
if (strpos($pdfData, 'xref', $startxref) == $startxref) { |
805
|
|
|
// Cross-Reference |
806
|
16 |
|
$xref = $this->decodeXref($pdfData, $startxref, $xref); |
807
|
|
|
} else { |
808
|
|
|
// Cross-Reference Stream |
809
|
2 |
|
$xref = $this->decodeXrefStream($pdfData, $startxref, $xref); |
810
|
|
|
} |
811
|
17 |
|
if (empty($xref)) { |
812
|
|
|
throw new Exception('Unable to find xref'); |
813
|
|
|
} |
814
|
|
|
|
815
|
17 |
|
return $xref; |
816
|
|
|
} |
817
|
|
|
|
818
|
|
|
/** |
819
|
|
|
* Parses PDF data and returns extracted data as array. |
820
|
|
|
* |
821
|
|
|
* @param string $data PDF data to parse |
822
|
|
|
* |
823
|
|
|
* @return array array of parsed PDF document objects |
824
|
|
|
* |
825
|
|
|
* @throws Exception if empty PDF data given |
826
|
|
|
* @throws Exception if PDF data missing %PDF header |
827
|
|
|
*/ |
828
|
18 |
|
public function parseData($data) |
829
|
|
|
{ |
830
|
18 |
|
if (empty($data)) { |
831
|
|
|
throw new Exception('Empty PDF data given.'); |
832
|
|
|
} |
833
|
|
|
// find the pdf header starting position |
834
|
18 |
|
if (false === ($trimpos = strpos($data, '%PDF-'))) { |
835
|
|
|
throw new Exception('Invalid PDF data: missing %PDF header.'); |
836
|
|
|
} |
837
|
|
|
|
838
|
|
|
// get PDF content string |
839
|
18 |
|
$pdfData = substr($data, $trimpos); |
840
|
|
|
|
841
|
|
|
// get xref and trailer data |
842
|
18 |
|
$xref = $this->getXrefData($pdfData); |
843
|
|
|
|
844
|
|
|
// parse all document objects |
845
|
17 |
|
$objects = []; |
846
|
17 |
|
foreach ($xref['xref'] as $obj => $offset) { |
847
|
17 |
|
if (!isset($objects[$obj]) and ($offset > 0)) { |
848
|
|
|
// decode objects with positive offset |
849
|
17 |
|
$objects[$obj] = $this->getIndirectObject($pdfData, $xref, $obj, $offset, true); |
850
|
|
|
} |
851
|
|
|
} |
852
|
|
|
|
853
|
17 |
|
return [$xref, $objects]; |
854
|
|
|
} |
855
|
|
|
} |
856
|
|
|
|