| 148 |
lars |
1 |
<?php
|
|
|
2 |
|
|
|
3 |
/*
|
|
|
4 |
* This file is part of the Symfony package.
|
|
|
5 |
*
|
|
|
6 |
* (c) Fabien Potencier <fabien@symfony.com>
|
|
|
7 |
*
|
|
|
8 |
* For the full copyright and license information, please view the LICENSE
|
|
|
9 |
* file that was distributed with this source code.
|
|
|
10 |
*/
|
|
|
11 |
|
|
|
12 |
namespace Symfony\Component\CssSelector\Parser\Tokenizer;
|
|
|
13 |
|
|
|
14 |
/**
|
|
|
15 |
* CSS selector tokenizer escaping applier.
|
|
|
16 |
*
|
|
|
17 |
* This component is a port of the Python cssselect library,
|
|
|
18 |
* which is copyright Ian Bicking, @see https://github.com/SimonSapin/cssselect.
|
|
|
19 |
*
|
|
|
20 |
* @author Jean-François Simon <jeanfrancois.simon@sensiolabs.com>
|
|
|
21 |
*
|
|
|
22 |
* @internal
|
|
|
23 |
*/
|
|
|
24 |
class TokenizerEscaping
|
|
|
25 |
{
|
|
|
26 |
private TokenizerPatterns $patterns;
|
|
|
27 |
|
|
|
28 |
public function __construct(TokenizerPatterns $patterns)
|
|
|
29 |
{
|
|
|
30 |
$this->patterns = $patterns;
|
|
|
31 |
}
|
|
|
32 |
|
|
|
33 |
public function escapeUnicode(string $value): string
|
|
|
34 |
{
|
|
|
35 |
$value = $this->replaceUnicodeSequences($value);
|
|
|
36 |
|
|
|
37 |
return preg_replace($this->patterns->getSimpleEscapePattern(), '$1', $value);
|
|
|
38 |
}
|
|
|
39 |
|
|
|
40 |
public function escapeUnicodeAndNewLine(string $value): string
|
|
|
41 |
{
|
|
|
42 |
$value = preg_replace($this->patterns->getNewLineEscapePattern(), '', $value);
|
|
|
43 |
|
|
|
44 |
return $this->escapeUnicode($value);
|
|
|
45 |
}
|
|
|
46 |
|
|
|
47 |
private function replaceUnicodeSequences(string $value): string
|
|
|
48 |
{
|
|
|
49 |
return preg_replace_callback($this->patterns->getUnicodeEscapePattern(), function ($match) {
|
|
|
50 |
$c = hexdec($match[1]);
|
|
|
51 |
|
|
|
52 |
if (0x80 > $c %= 0x200000) {
|
|
|
53 |
return \chr($c);
|
|
|
54 |
}
|
|
|
55 |
if (0x800 > $c) {
|
|
|
56 |
return \chr(0xC0 | $c >> 6).\chr(0x80 | $c & 0x3F);
|
|
|
57 |
}
|
|
|
58 |
if (0x10000 > $c) {
|
|
|
59 |
return \chr(0xE0 | $c >> 12).\chr(0x80 | $c >> 6 & 0x3F).\chr(0x80 | $c & 0x3F);
|
|
|
60 |
}
|
|
|
61 |
|
|
|
62 |
return '';
|
|
|
63 |
}, $value);
|
|
|
64 |
}
|
|
|
65 |
}
|