|
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\Parser; |
|
16
|
|
|
|
|
17
|
|
|
use League\CommonMark\Delimiter\Delimiter; |
|
18
|
|
|
use League\CommonMark\Delimiter\Processor\DelimiterProcessorInterface; |
|
19
|
|
|
use League\CommonMark\Environment\EnvironmentInterface; |
|
20
|
|
|
use League\CommonMark\Node\Block\AbstractStringContainerBlock; |
|
21
|
|
|
use League\CommonMark\Node\Inline\AdjacentTextMerger; |
|
22
|
|
|
use League\CommonMark\Node\Inline\Text; |
|
23
|
|
|
use League\CommonMark\Node\Node; |
|
24
|
|
|
use League\CommonMark\Reference\ReferenceMapInterface; |
|
25
|
|
|
use League\CommonMark\Util\RegexHelper; |
|
26
|
|
|
|
|
27
|
|
|
/** |
|
28
|
|
|
* @internal |
|
29
|
|
|
*/ |
|
30
|
|
|
final class InlineParserEngine |
|
31
|
|
|
{ |
|
32
|
|
|
/** @var EnvironmentInterface */ |
|
33
|
|
|
protected $environment; |
|
34
|
|
|
|
|
35
|
2505 |
|
public function __construct(EnvironmentInterface $environment) |
|
36
|
|
|
{ |
|
37
|
2505 |
|
$this->environment = $environment; |
|
38
|
2505 |
|
} |
|
39
|
|
|
|
|
40
|
2448 |
|
public function parse(AbstractStringContainerBlock $container, ReferenceMapInterface $referenceMap): void |
|
41
|
|
|
{ |
|
42
|
2448 |
|
$inlineParserContext = new InlineParserContext($container, $referenceMap); |
|
43
|
2448 |
|
$cursor = $inlineParserContext->getCursor(); |
|
44
|
2448 |
|
while (($character = $cursor->getCharacter()) !== null) { |
|
45
|
2433 |
|
if (!$this->parseCharacter($character, $inlineParserContext)) { |
|
46
|
2256 |
|
$this->addPlainText($character, $container, $inlineParserContext); |
|
47
|
|
|
} |
|
48
|
|
|
} |
|
49
|
|
|
|
|
50
|
2448 |
|
$this->processInlines($inlineParserContext); |
|
51
|
|
|
|
|
52
|
2448 |
|
AdjacentTextMerger::mergeChildNodes($container); |
|
53
|
2448 |
|
} |
|
54
|
|
|
|
|
55
|
|
|
/** |
|
56
|
|
|
* @param string $character |
|
57
|
|
|
* @param InlineParserContext $inlineParserContext |
|
58
|
|
|
* |
|
59
|
|
|
* @return bool Whether we successfully parsed a character at that position |
|
60
|
|
|
*/ |
|
61
|
2433 |
|
private function parseCharacter(string $character, InlineParserContext $inlineParserContext): bool |
|
62
|
|
|
{ |
|
63
|
2433 |
|
foreach ($this->environment->getInlineParsersForCharacter($character) as $parser) { |
|
64
|
1365 |
|
if ($parser->parse($inlineParserContext)) { |
|
65
|
1293 |
|
return true; |
|
66
|
|
|
} |
|
67
|
|
|
} |
|
68
|
|
|
|
|
69
|
2268 |
|
if ($delimiterProcessor = $this->environment->getDelimiterProcessors()->getDelimiterProcessor($character)) { |
|
70
|
726 |
|
return $this->parseDelimiters($delimiterProcessor, $inlineParserContext); |
|
71
|
|
|
} |
|
72
|
|
|
|
|
73
|
2256 |
|
return false; |
|
74
|
|
|
} |
|
75
|
|
|
|
|
76
|
726 |
|
private function parseDelimiters(DelimiterProcessorInterface $delimiterProcessor, InlineParserContext $inlineContext): bool |
|
77
|
|
|
{ |
|
78
|
726 |
|
$cursor = $inlineContext->getCursor(); |
|
79
|
726 |
|
$character = $cursor->getCharacter(); |
|
80
|
726 |
|
$numDelims = 0; |
|
81
|
|
|
|
|
82
|
726 |
|
$charBefore = $cursor->peek(-1); |
|
83
|
726 |
|
if ($charBefore === null) { |
|
84
|
411 |
|
$charBefore = "\n"; |
|
85
|
|
|
} |
|
86
|
|
|
|
|
87
|
726 |
|
while ($cursor->peek($numDelims) === $character) { |
|
88
|
726 |
|
++$numDelims; |
|
89
|
|
|
} |
|
90
|
|
|
|
|
91
|
726 |
|
if ($numDelims < $delimiterProcessor->getMinLength()) { |
|
92
|
6 |
|
return false; |
|
93
|
|
|
} |
|
94
|
|
|
|
|
95
|
720 |
|
$cursor->advanceBy($numDelims); |
|
96
|
|
|
|
|
97
|
720 |
|
$charAfter = $cursor->getCharacter(); |
|
98
|
720 |
|
if ($charAfter === null) { |
|
99
|
435 |
|
$charAfter = "\n"; |
|
100
|
|
|
} |
|
101
|
|
|
|
|
102
|
720 |
|
[$canOpen, $canClose] = self::determineCanOpenOrClose($charBefore, $charAfter, $character, $delimiterProcessor); |
|
|
|
|
|
|
103
|
|
|
|
|
104
|
720 |
|
$node = new Text(\str_repeat($character, $numDelims), [ |
|
105
|
720 |
|
'delim' => true, |
|
106
|
|
|
]); |
|
107
|
720 |
|
$inlineContext->getContainer()->appendChild($node); |
|
108
|
|
|
|
|
109
|
|
|
// Add entry to stack to this opener |
|
110
|
720 |
|
if ($canOpen || $canClose) { |
|
111
|
651 |
|
$delimiter = new Delimiter($character, $numDelims, $node, $canOpen, $canClose); |
|
112
|
651 |
|
$inlineContext->getDelimiterStack()->push($delimiter); |
|
113
|
|
|
} |
|
114
|
|
|
|
|
115
|
720 |
|
return true; |
|
116
|
|
|
} |
|
117
|
|
|
|
|
118
|
2448 |
|
private function processInlines(InlineParserContext $inlineParserContext): void |
|
119
|
|
|
{ |
|
120
|
2448 |
|
$delimiterStack = $inlineParserContext->getDelimiterStack(); |
|
121
|
2448 |
|
$delimiterStack->processDelimiters(null, $this->environment->getDelimiterProcessors()); |
|
122
|
|
|
|
|
123
|
|
|
// Remove all delimiters |
|
124
|
2448 |
|
$delimiterStack->removeAll(); |
|
125
|
2448 |
|
} |
|
126
|
|
|
|
|
127
|
2256 |
|
private function addPlainText(string $character, Node $container, InlineParserContext $inlineParserContext): void |
|
128
|
|
|
{ |
|
129
|
|
|
// We reach here if none of the parsers can handle the input |
|
130
|
|
|
// Attempt to match multiple non-special characters at once |
|
131
|
2256 |
|
$text = $inlineParserContext->getCursor()->match($this->environment->getInlineParserCharacterRegex()); |
|
132
|
|
|
// This might fail if we're currently at a special character which wasn't parsed; if so, just add that character |
|
133
|
2256 |
|
if ($text === null) { |
|
134
|
309 |
|
$inlineParserContext->getCursor()->advanceBy(1); |
|
135
|
309 |
|
$text = $character; |
|
136
|
|
|
} |
|
137
|
|
|
|
|
138
|
2256 |
|
$lastInline = $container->lastChild(); |
|
139
|
2256 |
|
if ($lastInline instanceof Text && !isset($lastInline->data['delim'])) { |
|
140
|
354 |
|
$lastInline->append($text); |
|
141
|
|
|
} else { |
|
142
|
2217 |
|
$container->appendChild(new Text($text)); |
|
143
|
|
|
} |
|
144
|
2256 |
|
} |
|
145
|
|
|
|
|
146
|
|
|
/** |
|
147
|
|
|
* @param string $charBefore |
|
148
|
|
|
* @param string $charAfter |
|
149
|
|
|
* @param string $character |
|
150
|
|
|
* @param DelimiterProcessorInterface $delimiterProcessor |
|
151
|
|
|
* |
|
152
|
|
|
* @return bool[] |
|
153
|
|
|
*/ |
|
154
|
720 |
|
private static function determineCanOpenOrClose(string $charBefore, string $charAfter, string $character, DelimiterProcessorInterface $delimiterProcessor) |
|
155
|
|
|
{ |
|
156
|
720 |
|
$afterIsWhitespace = \preg_match(RegexHelper::REGEX_UNICODE_WHITESPACE_CHAR, $charAfter); |
|
157
|
720 |
|
$afterIsPunctuation = \preg_match(RegexHelper::REGEX_PUNCTUATION, $charAfter); |
|
158
|
720 |
|
$beforeIsWhitespace = \preg_match(RegexHelper::REGEX_UNICODE_WHITESPACE_CHAR, $charBefore); |
|
159
|
720 |
|
$beforeIsPunctuation = \preg_match(RegexHelper::REGEX_PUNCTUATION, $charBefore); |
|
160
|
|
|
|
|
161
|
720 |
|
$leftFlanking = !$afterIsWhitespace && (!$afterIsPunctuation || $beforeIsWhitespace || $beforeIsPunctuation); |
|
162
|
720 |
|
$rightFlanking = !$beforeIsWhitespace && (!$beforeIsPunctuation || $afterIsWhitespace || $afterIsPunctuation); |
|
163
|
|
|
|
|
164
|
720 |
|
if ($character === '_') { |
|
165
|
237 |
|
$canOpen = $leftFlanking && (!$rightFlanking || $beforeIsPunctuation); |
|
166
|
237 |
|
$canClose = $rightFlanking && (!$leftFlanking || $afterIsPunctuation); |
|
167
|
|
|
} else { |
|
168
|
522 |
|
$canOpen = $leftFlanking && $character === $delimiterProcessor->getOpeningCharacter(); |
|
169
|
522 |
|
$canClose = $rightFlanking && $character === $delimiterProcessor->getClosingCharacter(); |
|
170
|
|
|
} |
|
171
|
|
|
|
|
172
|
720 |
|
return [$canOpen, $canClose]; |
|
173
|
|
|
} |
|
174
|
|
|
} |
|
175
|
|
|
|
thephpleague /
commonmark
Loading history...
This check marks access to variables or properties that have not been declared yet. While PHP has no explicit notion of declaring a variable, accessing it before a value is assigned to it is most likely a bug.