1
|
|
|
<?php |
2
|
|
|
|
3
|
|
|
declare(strict_types=1); |
4
|
|
|
|
5
|
|
|
/* |
6
|
|
|
* This file is part of the league/commonmark package. |
7
|
|
|
* |
8
|
|
|
* (c) Colin O'Dell <[email protected]> |
9
|
|
|
* |
10
|
|
|
* Original code based on the CommonMark JS reference parser (https://bitly.com/commonmark-js) |
11
|
|
|
* - (c) John MacFarlane |
12
|
|
|
* |
13
|
|
|
* For the full copyright and license information, please view the LICENSE |
14
|
|
|
* file that was distributed with this source code. |
15
|
|
|
*/ |
16
|
|
|
|
17
|
|
|
namespace League\CommonMark\Parser; |
18
|
|
|
|
19
|
|
|
use League\CommonMark\Environment\EnvironmentInterface; |
20
|
|
|
use League\CommonMark\Node\Block\AbstractBlock; |
21
|
|
|
use League\CommonMark\Node\Inline\AdjacentTextMerger; |
22
|
|
|
use League\CommonMark\Node\Inline\Text; |
23
|
|
|
use League\CommonMark\Parser\Inline\InlineParserInterface; |
24
|
|
|
use League\CommonMark\Reference\ReferenceMapInterface; |
25
|
|
|
|
26
|
|
|
/** |
27
|
|
|
* @internal |
28
|
|
|
*/ |
29
|
|
|
final class InlineParserEngine implements InlineParserEngineInterface |
30
|
|
|
{ |
31
|
|
|
/** |
32
|
|
|
* @var EnvironmentInterface |
33
|
|
|
* |
34
|
|
|
* @psalm-readonly |
35
|
|
|
*/ |
36
|
|
|
private $environment; |
37
|
|
|
|
38
|
|
|
/** |
39
|
|
|
* @var ReferenceMapInterface |
40
|
|
|
* |
41
|
|
|
* @psalm-readonly |
42
|
|
|
*/ |
43
|
|
|
private $referenceMap; |
44
|
|
|
|
45
|
|
|
/** |
46
|
|
|
* @var array<int, InlineParserInterface|string|bool> |
47
|
|
|
* @psalm-var list<array{0: InlineParserInterface, 1: string, 2: bool}> |
48
|
|
|
* @phpstan-var array<int, array{0: InlineParserInterface, 1: string, 2: bool}> |
49
|
|
|
*/ |
50
|
|
|
private $parsers; |
51
|
|
|
|
52
|
2988 |
|
public function __construct(EnvironmentInterface $environment, ReferenceMapInterface $referenceMap) |
53
|
|
|
{ |
54
|
2988 |
|
$this->environment = $environment; |
55
|
2988 |
|
$this->referenceMap = $referenceMap; |
56
|
|
|
|
57
|
2988 |
|
foreach ($environment->getInlineParsers() as $parser) { |
58
|
|
|
\assert($parser instanceof InlineParserInterface); |
59
|
2988 |
|
$regex = $parser->getMatchDefinition()->getRegex(); |
60
|
|
|
|
61
|
2988 |
|
$this->parsers[] = [$parser, $regex, \strlen($regex) !== \mb_strlen($regex)]; |
62
|
|
|
} |
63
|
2988 |
|
} |
64
|
|
|
|
65
|
2694 |
|
public function parse(string $contents, AbstractBlock $block): void |
66
|
|
|
{ |
67
|
2694 |
|
$contents = \trim($contents); |
68
|
2694 |
|
$cursor = new Cursor($contents); |
69
|
|
|
|
70
|
2694 |
|
$inlineParserContext = new InlineParserContext($cursor, $block, $this->referenceMap); |
71
|
|
|
|
72
|
|
|
// Have all parsers look at the line to determine what they might want to parse and what positions they exist at |
73
|
2694 |
|
foreach ($this->matchParsers($contents) as $matchPosition => $parsers) { |
74
|
2070 |
|
$currentPosition = $cursor->getPosition(); |
75
|
|
|
// We've already gone past this point |
76
|
2070 |
|
if ($currentPosition > $matchPosition) { |
77
|
927 |
|
continue; |
78
|
|
|
} |
79
|
|
|
|
80
|
|
|
// We've skipped over some uninteresting text that should be added as a plain text node |
81
|
2070 |
|
if ($currentPosition < $matchPosition) { |
82
|
1797 |
|
$cursor->advanceBy($matchPosition - $currentPosition); |
83
|
1797 |
|
$this->addPlainText($cursor->getPreviousText(), $block); |
84
|
|
|
} |
85
|
|
|
|
86
|
|
|
// We're now at a potential start - see which of the current parsers can handle it |
87
|
2070 |
|
$parsed = false; |
88
|
2070 |
|
foreach ($parsers as [$parser, $matches]) { |
89
|
|
|
\assert($parser instanceof InlineParserInterface); |
90
|
2070 |
|
if ($parser->parse($inlineParserContext->withMatches($matches))) { |
91
|
|
|
// A parser has successfully handled the text at the given position; don't consider any others at this position |
92
|
2058 |
|
$parsed = true; |
93
|
2058 |
|
break; |
94
|
|
|
} |
95
|
|
|
} |
96
|
|
|
|
97
|
2070 |
|
if ($parsed) { |
98
|
2058 |
|
continue; |
99
|
|
|
} |
100
|
|
|
|
101
|
|
|
// Despite potentially being interested, nothing actually parsed text here, so add the current character and continue onwards |
102
|
186 |
|
$this->addPlainText((string) $cursor->getCharacter(), $block); |
103
|
186 |
|
$cursor->advance(); |
104
|
|
|
} |
105
|
|
|
|
106
|
|
|
// Add any remaining text that wasn't parsed |
107
|
2694 |
|
if (! $cursor->isAtEnd()) { |
108
|
1371 |
|
$this->addPlainText($cursor->getRemainder(), $block); |
109
|
|
|
} |
110
|
|
|
|
111
|
|
|
// Process any delimiters that were found |
112
|
2694 |
|
$delimiterStack = $inlineParserContext->getDelimiterStack(); |
113
|
2694 |
|
$delimiterStack->processDelimiters(null, $this->environment->getDelimiterProcessors()); |
114
|
2694 |
|
$delimiterStack->removeAll(); |
115
|
|
|
|
116
|
|
|
// Combine adjacent text notes into one |
117
|
2694 |
|
AdjacentTextMerger::mergeChildNodes($block); |
118
|
2694 |
|
} |
119
|
|
|
|
120
|
2493 |
|
private function addPlainText(string $text, AbstractBlock $container): void |
121
|
|
|
{ |
122
|
2493 |
|
$lastInline = $container->lastChild(); |
123
|
2493 |
|
if ($lastInline instanceof Text && ! $lastInline->data->has('delim')) { |
124
|
270 |
|
$lastInline->append($text); |
125
|
|
|
} else { |
126
|
2457 |
|
$container->appendChild(new Text($text)); |
127
|
|
|
} |
128
|
2493 |
|
} |
129
|
|
|
|
130
|
|
|
/** |
131
|
|
|
* Given the current line, ask all the parsers which parts of the text they would be interested in parsing. |
132
|
|
|
* |
133
|
|
|
* The resulting array provides a list of character positions, which parsers are interested in trying to parse |
134
|
|
|
* the text at those points, and (for convenience/optimization) what the matching text happened to be. |
135
|
|
|
* |
136
|
|
|
* @return array<array<int, InlineParserInterface|string>> |
137
|
|
|
* |
138
|
|
|
* @psalm-return array<int, list<array{0: InlineParserInterface, 1: non-empty-array<string>}>> |
139
|
|
|
* |
140
|
|
|
* @phpstan-return array<int, array<int, array{0: InlineParserInterface, 1: non-empty-array<string>}>> |
141
|
|
|
*/ |
142
|
2694 |
|
private function matchParsers(string $contents): array |
143
|
|
|
{ |
144
|
2694 |
|
$contents = \trim($contents); |
145
|
2694 |
|
$isMultibyte = \mb_strlen($contents) !== \strlen($contents); |
146
|
|
|
|
147
|
2694 |
|
$ret = []; |
148
|
|
|
|
149
|
2694 |
|
foreach ($this->parsers as [$parser, $regex, $isRegexMultibyte]) { |
150
|
2694 |
|
if ($isMultibyte || $isRegexMultibyte) { |
151
|
111 |
|
$regex .= 'u'; |
152
|
|
|
} |
153
|
|
|
|
154
|
|
|
// See if the parser's InlineParserMatch regex matched against any part of the string |
155
|
2694 |
|
if (! \preg_match_all($regex, $contents, $matches, \PREG_OFFSET_CAPTURE | \PREG_SET_ORDER)) { |
156
|
2688 |
|
continue; |
157
|
|
|
} |
158
|
|
|
|
159
|
|
|
// For each part that matched... |
160
|
2070 |
|
foreach ($matches as $match) { |
161
|
2070 |
|
if ($isMultibyte) { |
162
|
|
|
// PREG_OFFSET_CAPTURE always returns the byte offset, not the char offset, which is annoying |
163
|
57 |
|
$offset = \mb_strlen(\substr($contents, 0, $match[0][1]), 'UTF-8'); |
164
|
|
|
} else { |
165
|
2013 |
|
$offset = $match[0][1]; |
166
|
|
|
} |
167
|
|
|
|
168
|
|
|
\assert(\is_int($offset)); |
169
|
|
|
|
170
|
|
|
// Remove the offsets, keeping only the matched text |
171
|
1380 |
|
$m = \array_map(static function (array $s): string { |
172
|
2070 |
|
return (string) $s[0]; |
173
|
2070 |
|
}, $match); |
174
|
|
|
|
175
|
2070 |
|
if ($m === []) { |
176
|
|
|
continue; |
177
|
|
|
} |
178
|
|
|
|
179
|
|
|
// Add this match to the list of character positions to stop at |
180
|
2070 |
|
$ret[$offset][] = [$parser, $m]; |
181
|
|
|
} |
182
|
|
|
} |
183
|
|
|
|
184
|
|
|
// Sort matches by position so we visit them in order |
185
|
2694 |
|
\ksort($ret); |
186
|
|
|
|
187
|
2694 |
|
return $ret; |
188
|
|
|
} |
189
|
|
|
} |
190
|
|
|
|