1
|
|
|
<?php |
2
|
|
|
|
3
|
|
|
/* |
4
|
|
|
* This file is part of the league/commonmark package. |
5
|
|
|
* |
6
|
|
|
* (c) Colin O'Dell <[email protected]> |
7
|
|
|
* |
8
|
|
|
* Original code based on the CommonMark JS reference parser (https://bitly.com/commonmark-js) |
9
|
|
|
* - (c) John MacFarlane |
10
|
|
|
* |
11
|
|
|
* For the full copyright and license information, please view the LICENSE |
12
|
|
|
* file that was distributed with this source code. |
13
|
|
|
*/ |
14
|
|
|
|
15
|
|
|
namespace League\CommonMark\Util; |
16
|
|
|
|
17
|
|
|
use League\CommonMark\Cursor; |
18
|
|
|
|
19
|
|
|
final class LinkParserHelper |
20
|
|
|
{ |
21
|
|
|
/** |
22
|
|
|
* Attempt to parse link destination |
23
|
|
|
* |
24
|
|
|
* @param Cursor $cursor |
25
|
|
|
* |
26
|
|
|
* @return null|string The string, or null if no match |
27
|
|
|
*/ |
28
|
420 |
|
public static function parseLinkDestination(Cursor $cursor): ?string |
29
|
|
|
{ |
30
|
420 |
|
if ($res = $cursor->match(RegexHelper::REGEX_LINK_DESTINATION_BRACES)) { |
31
|
|
|
// Chop off surrounding <..>: |
32
|
30 |
|
return UrlEncoder::unescapeAndEncode( |
33
|
30 |
|
RegexHelper::unescape(\substr($res, 1, -1)) |
34
|
|
|
); |
35
|
|
|
} |
36
|
|
|
|
37
|
393 |
|
if ($cursor->getCharacter() === '<') { |
38
|
9 |
|
return null; |
39
|
|
|
} |
40
|
|
|
|
41
|
384 |
|
$destination = self::manuallyParseLinkDestination($cursor); |
42
|
384 |
|
if ($destination === null) { |
43
|
3 |
|
return null; |
44
|
|
|
} |
45
|
|
|
|
46
|
381 |
|
return UrlEncoder::unescapeAndEncode( |
47
|
381 |
|
RegexHelper::unescape($destination) |
48
|
|
|
); |
49
|
|
|
} |
50
|
|
|
|
51
|
|
|
/** |
52
|
|
|
* @param Cursor $cursor |
53
|
|
|
* |
54
|
|
|
* @return int |
55
|
|
|
*/ |
56
|
426 |
|
public static function parseLinkLabel(Cursor $cursor): int |
57
|
|
|
{ |
58
|
426 |
|
$match = $cursor->match('/^\[(?:[^\\\\\[\]]|\\\\.){0,1000}\]/'); |
59
|
426 |
|
if ($match === null) { |
60
|
228 |
|
return 0; |
61
|
|
|
} |
62
|
|
|
|
63
|
399 |
|
$length = \mb_strlen($match, 'utf-8'); |
64
|
|
|
|
65
|
399 |
|
if ($length > 1001) { |
66
|
|
|
return 0; |
67
|
|
|
} |
68
|
|
|
|
69
|
399 |
|
return $length; |
70
|
|
|
} |
71
|
|
|
|
72
|
|
|
/** |
73
|
|
|
* Attempt to parse link title (sans quotes) |
74
|
|
|
* |
75
|
|
|
* @param Cursor $cursor |
76
|
|
|
* |
77
|
|
|
* @return null|string The string, or null if no match |
78
|
|
|
*/ |
79
|
183 |
|
public static function parseLinkTitle(Cursor $cursor): ?string |
80
|
|
|
{ |
81
|
183 |
|
if ($title = $cursor->match('/' . RegexHelper::PARTIAL_LINK_TITLE . '/')) { |
82
|
|
|
// Chop off quotes from title and unescape |
83
|
138 |
|
return RegexHelper::unescape(\substr($title, 1, -1)); |
84
|
|
|
} |
85
|
|
|
|
86
|
45 |
|
return null; |
87
|
|
|
} |
88
|
|
|
|
89
|
|
|
/** |
90
|
|
|
* @param Cursor $cursor |
91
|
|
|
* |
92
|
|
|
* @return string|null |
93
|
|
|
*/ |
94
|
384 |
|
private static function manuallyParseLinkDestination(Cursor $cursor): ?string |
95
|
|
|
{ |
96
|
384 |
|
$oldPosition = $cursor->getPosition(); |
97
|
384 |
|
$oldState = $cursor->saveState(); |
98
|
|
|
|
99
|
384 |
|
$openParens = 0; |
100
|
384 |
|
while (($c = $cursor->getCharacter()) !== null) { |
101
|
381 |
|
if ($c === '\\' && $cursor->peek() !== null && RegexHelper::isEscapable($cursor->peek())) { |
102
|
18 |
|
$cursor->advanceBy(2); |
103
|
381 |
|
} elseif ($c === '(') { |
104
|
12 |
|
$cursor->advanceBy(1); |
105
|
12 |
|
$openParens++; |
106
|
381 |
|
} elseif ($c === ')') { |
107
|
117 |
|
if ($openParens < 1) { |
108
|
114 |
|
break; |
109
|
|
|
} |
110
|
|
|
|
111
|
12 |
|
$cursor->advanceBy(1); |
112
|
12 |
|
$openParens--; |
113
|
378 |
|
} elseif (\preg_match(RegexHelper::REGEX_WHITESPACE_CHAR, $c)) { |
114
|
177 |
|
break; |
115
|
|
|
} else { |
116
|
378 |
|
$cursor->advanceBy(1); |
117
|
|
|
} |
118
|
|
|
} |
119
|
|
|
|
120
|
384 |
|
if ($openParens !== 0) { |
121
|
|
|
return null; |
122
|
|
|
} |
123
|
|
|
|
124
|
384 |
|
if ($cursor->getPosition() === $oldPosition && $c !== ')') { |
125
|
3 |
|
return null; |
126
|
|
|
} |
127
|
|
|
|
128
|
381 |
|
$newPos = $cursor->getPosition(); |
129
|
381 |
|
$cursor->restoreState($oldState); |
130
|
|
|
|
131
|
381 |
|
$cursor->advanceBy($newPos - $cursor->getPosition()); |
132
|
|
|
|
133
|
381 |
|
return $cursor->getPreviousText(); |
134
|
|
|
} |
135
|
|
|
} |
136
|
|
|
|