|
1
|
|
|
<?php |
|
2
|
|
|
|
|
3
|
|
|
/* |
|
4
|
|
|
* This file is part of the league/commonmark package. |
|
5
|
|
|
* |
|
6
|
|
|
* (c) Colin O'Dell <[email protected]> |
|
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
|
|
|
declare(strict_types=1); |
|
13
|
|
|
|
|
14
|
|
|
namespace League\CommonMark\Normalizer; |
|
15
|
|
|
|
|
16
|
|
|
/*** |
|
17
|
|
|
* Normalize text input using the steps given by the CommonMark spec to normalize labels |
|
18
|
|
|
* |
|
19
|
|
|
* @see https://spec.commonmark.org/0.29/#matches |
|
20
|
|
|
* |
|
21
|
|
|
* @psalm-immutable |
|
22
|
|
|
*/ |
|
23
|
|
|
final class TextNormalizer implements TextNormalizerInterface |
|
24
|
|
|
{ |
|
25
|
|
|
/** |
|
26
|
|
|
* Source: https://github.com/symfony/polyfill-mbstring/blob/master/Mbstring.php |
|
27
|
|
|
*/ |
|
28
|
|
|
private const CASE_FOLD = [ |
|
29
|
|
|
['µ', 'ſ', "\xCD\x85", 'ς', "\xCF\x90", "\xCF\x91", "\xCF\x95", "\xCF\x96", "\xCF\xB0", "\xCF\xB1", "\xCF\xB5", "\xE1\xBA\x9B", "\xE1\xBE\xBE", "\xC3\x9F", "\xE1\xBA\x9E"], |
|
30
|
|
|
['μ', 's', 'ι', 'σ', 'β', 'θ', 'φ', 'π', 'κ', 'ρ', 'ε', "\xE1\xB9\xA1", 'ι', 'ss', 'ss'], |
|
31
|
|
|
]; |
|
32
|
|
|
|
|
33
|
|
|
/** |
|
34
|
|
|
* {@inheritdoc} |
|
35
|
|
|
* |
|
36
|
|
|
* @psalm-pure |
|
37
|
|
|
*/ |
|
38
|
543 |
|
public function normalize(string $text, $context = null): string |
|
39
|
|
|
{ |
|
40
|
|
|
// Collapse internal whitespace to single space and remove |
|
41
|
|
|
// leading/trailing whitespace |
|
42
|
543 |
|
$text = \preg_replace('/\s+/', ' ', \trim($text)); |
|
43
|
|
|
\assert(\is_string($text)); |
|
44
|
|
|
|
|
45
|
543 |
|
if (! \defined('MB_CASE_FOLD')) { |
|
46
|
|
|
// We're not on a version of PHP (7.3+) which has this feature |
|
47
|
181 |
|
$text = \str_replace(self::CASE_FOLD[0], self::CASE_FOLD[1], $text); |
|
48
|
|
|
|
|
49
|
181 |
|
return \mb_strtolower($text, 'UTF-8'); |
|
50
|
|
|
} |
|
51
|
|
|
|
|
52
|
362 |
|
return \mb_convert_case($text, \MB_CASE_FOLD, 'UTF-8'); |
|
53
|
|
|
} |
|
54
|
|
|
} |
|
55
|
|
|
|
thephpleague /
commonmark
