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; |
16
|
|
|
|
17
|
|
|
use League\CommonMark\Block\Element\AbstractStringContainerBlock; |
18
|
|
|
use League\CommonMark\Delimiter\Delimiter; |
19
|
|
|
use League\CommonMark\Delimiter\Processor\DelimiterProcessorInterface; |
20
|
|
|
use League\CommonMark\Inline\AdjacentTextMerger; |
21
|
|
|
use League\CommonMark\Inline\Element\Text; |
22
|
|
|
use League\CommonMark\Node\Node; |
23
|
|
|
use League\CommonMark\Reference\ReferenceMap; |
24
|
|
|
use League\CommonMark\Util\RegexHelper; |
25
|
|
|
|
26
|
|
|
/** |
27
|
|
|
* @internal |
28
|
|
|
*/ |
29
|
|
|
final class InlineParserEngine |
30
|
|
|
{ |
31
|
|
|
protected $environment; |
32
|
|
|
|
33
|
2079 |
|
public function __construct(EnvironmentInterface $environment) |
34
|
|
|
{ |
35
|
2079 |
|
$this->environment = $environment; |
36
|
2079 |
|
} |
37
|
|
|
|
38
|
|
|
/** |
39
|
|
|
* @param AbstractStringContainerBlock $container |
40
|
|
|
* @param ReferenceMap $referenceMap |
41
|
|
|
*/ |
42
|
2025 |
|
public function parse(AbstractStringContainerBlock $container, ReferenceMap $referenceMap) |
43
|
|
|
{ |
44
|
2025 |
|
$inlineParserContext = new InlineParserContext($container, $referenceMap); |
45
|
2025 |
|
while (($character = $inlineParserContext->getCursor()->getCharacter()) !== null) { |
46
|
2010 |
|
if (!$this->parseCharacter($character, $inlineParserContext)) { |
47
|
1839 |
|
$this->addPlainText($character, $container, $inlineParserContext); |
48
|
|
|
} |
49
|
|
|
} |
50
|
|
|
|
51
|
2025 |
|
$this->processInlines($inlineParserContext); |
52
|
|
|
|
53
|
2025 |
|
AdjacentTextMerger::mergeChildNodes($container); |
54
|
2025 |
|
} |
55
|
|
|
|
56
|
|
|
/** |
57
|
|
|
* @param string $character |
58
|
|
|
* @param InlineParserContext $inlineParserContext |
59
|
|
|
* |
60
|
|
|
* @return bool Whether we successfully parsed a character at that position |
61
|
|
|
*/ |
62
|
2010 |
|
private function parseCharacter(string $character, InlineParserContext $inlineParserContext): bool |
63
|
|
|
{ |
64
|
2010 |
|
foreach ($this->environment->getInlineParsersForCharacter($character) as $parser) { |
65
|
1200 |
|
if ($parser->parse($inlineParserContext)) { |
66
|
1174 |
|
return true; |
67
|
|
|
} |
68
|
|
|
} |
69
|
|
|
|
70
|
1851 |
|
if ($delimiterProcessor = $this->environment->getDelimiterProcessors()->getDelimiterProcessor($character)) { |
71
|
618 |
|
return $this->parseDelimiters($delimiterProcessor, $inlineParserContext); |
72
|
|
|
} |
73
|
|
|
|
74
|
1839 |
|
return false; |
75
|
|
|
} |
76
|
|
|
|
77
|
618 |
|
private function parseDelimiters(DelimiterProcessorInterface $delimiterProcessor, InlineParserContext $inlineContext): bool |
78
|
|
|
{ |
79
|
618 |
|
$character = $inlineContext->getCursor()->getCharacter(); |
80
|
618 |
|
$numDelims = 0; |
81
|
|
|
|
82
|
618 |
|
$cursor = $inlineContext->getCursor(); |
83
|
618 |
|
$charBefore = $cursor->peek(-1); |
84
|
618 |
|
if ($charBefore === null) { |
85
|
387 |
|
$charBefore = "\n"; |
86
|
|
|
} |
87
|
|
|
|
88
|
618 |
|
while ($cursor->peek($numDelims) === $character) { |
89
|
618 |
|
++$numDelims; |
90
|
|
|
} |
91
|
|
|
|
92
|
618 |
|
if ($numDelims < $delimiterProcessor->getMinLength()) { |
93
|
|
|
return false; |
94
|
|
|
} |
95
|
|
|
|
96
|
618 |
|
$cursor->advanceBy($numDelims); |
97
|
|
|
|
98
|
618 |
|
$charAfter = $cursor->getCharacter(); |
99
|
618 |
|
if ($charAfter === null) { |
100
|
393 |
|
$charAfter = "\n"; |
101
|
|
|
} |
102
|
|
|
|
103
|
618 |
|
list($canOpen, $canClose) = self::determineCanOpenOrClose($charBefore, $charAfter, $character, $delimiterProcessor); |
104
|
|
|
|
105
|
618 |
|
$node = new Text($cursor->getPreviousText(), [ |
106
|
618 |
|
'delim' => true, |
107
|
|
|
]); |
108
|
618 |
|
$inlineContext->getContainer()->appendChild($node); |
109
|
|
|
|
110
|
|
|
// Add entry to stack to this opener |
111
|
618 |
|
$delimiter = new Delimiter($character, $numDelims, $node, $canOpen, $canClose); |
112
|
618 |
|
$inlineContext->getDelimiterStack()->push($delimiter); |
113
|
|
|
|
114
|
618 |
|
return true; |
115
|
|
|
} |
116
|
|
|
|
117
|
|
|
/** |
118
|
|
|
* @param InlineParserContext $inlineParserContext |
119
|
|
|
*/ |
120
|
2025 |
|
private function processInlines(InlineParserContext $inlineParserContext) |
121
|
|
|
{ |
122
|
2025 |
|
$delimiterStack = $inlineParserContext->getDelimiterStack(); |
123
|
2025 |
|
$delimiterStack->processDelimiters(null, $this->environment->getDelimiterProcessors()); |
124
|
|
|
|
125
|
|
|
// Remove all delimiters |
126
|
2025 |
|
$delimiterStack->removeAll(); |
127
|
2025 |
|
} |
128
|
|
|
|
129
|
|
|
/** |
130
|
|
|
* @param string $character |
131
|
|
|
* @param Node $container |
132
|
|
|
* @param InlineParserContext $inlineParserContext |
133
|
|
|
*/ |
134
|
1839 |
|
private function addPlainText(string $character, Node $container, InlineParserContext $inlineParserContext) |
135
|
|
|
{ |
136
|
|
|
// We reach here if none of the parsers can handle the input |
137
|
|
|
// Attempt to match multiple non-special characters at once |
138
|
1839 |
|
$text = $inlineParserContext->getCursor()->match($this->environment->getInlineParserCharacterRegex()); |
139
|
|
|
// This might fail if we're currently at a special character which wasn't parsed; if so, just add that character |
140
|
1839 |
|
if ($text === null) { |
141
|
213 |
|
$inlineParserContext->getCursor()->advance(); |
142
|
213 |
|
$text = $character; |
143
|
|
|
} |
144
|
|
|
|
145
|
1839 |
|
$lastInline = $container->lastChild(); |
146
|
1839 |
|
if ($lastInline instanceof Text && !isset($lastInline->data['delim'])) { |
147
|
258 |
|
$lastInline->append($text); |
148
|
|
|
} else { |
149
|
1800 |
|
$container->appendChild(new Text($text)); |
150
|
|
|
} |
151
|
1839 |
|
} |
152
|
|
|
|
153
|
|
|
/** |
154
|
|
|
* @param string $charBefore |
155
|
|
|
* @param string $charAfter |
156
|
|
|
* @param string $character |
157
|
|
|
* @param DelimiterProcessorInterface $delimiterProcessor |
158
|
|
|
* |
159
|
|
|
* @return bool[] |
160
|
|
|
*/ |
161
|
618 |
|
private static function determineCanOpenOrClose(string $charBefore, string $charAfter, string $character, DelimiterProcessorInterface $delimiterProcessor) |
162
|
|
|
{ |
163
|
618 |
|
$afterIsWhitespace = \preg_match(RegexHelper::REGEX_UNICODE_WHITESPACE_CHAR, $charAfter); |
164
|
618 |
|
$afterIsPunctuation = \preg_match(RegexHelper::REGEX_PUNCTUATION, $charAfter); |
165
|
618 |
|
$beforeIsWhitespace = \preg_match(RegexHelper::REGEX_UNICODE_WHITESPACE_CHAR, $charBefore); |
166
|
618 |
|
$beforeIsPunctuation = \preg_match(RegexHelper::REGEX_PUNCTUATION, $charBefore); |
167
|
|
|
|
168
|
618 |
|
$leftFlanking = !$afterIsWhitespace && (!$afterIsPunctuation || $beforeIsWhitespace || $beforeIsPunctuation); |
169
|
618 |
|
$rightFlanking = !$beforeIsWhitespace && (!$beforeIsPunctuation || $afterIsWhitespace || $afterIsPunctuation); |
170
|
|
|
|
171
|
618 |
|
if ($character === '_') { |
172
|
210 |
|
$canOpen = $leftFlanking && (!$rightFlanking || $beforeIsPunctuation); |
173
|
210 |
|
$canClose = $rightFlanking && (!$leftFlanking || $afterIsPunctuation); |
174
|
|
|
} else { |
175
|
441 |
|
$canOpen = $leftFlanking && $character === $delimiterProcessor->getOpeningCharacter(); |
176
|
441 |
|
$canClose = $rightFlanking && $character === $delimiterProcessor->getClosingCharacter(); |
177
|
|
|
} |
178
|
|
|
|
179
|
618 |
|
return [$canOpen, $canClose]; |
180
|
|
|
} |
181
|
|
|
} |
182
|
|
|
|