| 1 |
<?php declare(strict_types=1); |
| 2 |
|
| 3 |
namespace PhpParser; |
| 4 |
|
| 5 |
class Comment implements \JsonSerializable { |
| 6 |
protected string $text; |
| 7 |
protected int $startLine; |
| 8 |
protected int $startFilePos; |
| 9 |
protected int $startTokenPos; |
| 10 |
protected int $endLine; |
| 11 |
protected int $endFilePos; |
| 12 |
protected int $endTokenPos; |
| 13 |
|
| 14 |
/** |
| 15 |
* Constructs a comment node. |
| 16 |
* |
| 17 |
* @param string $text Comment text (including comment delimiters like /*) |
| 18 |
* @param int $startLine Line number the comment started on |
| 19 |
* @param int $startFilePos File offset the comment started on |
| 20 |
* @param int $startTokenPos Token offset the comment started on |
| 21 |
*/ |
| 22 |
public function __construct( |
| 23 |
string $text, |
| 24 |
int $startLine = -1, int $startFilePos = -1, int $startTokenPos = -1, |
| 25 |
int $endLine = -1, int $endFilePos = -1, int $endTokenPos = -1 |
| 26 |
) { |
| 27 |
$this->text = $text; |
| 28 |
$this->startLine = $startLine; |
| 29 |
$this->startFilePos = $startFilePos; |
| 30 |
$this->startTokenPos = $startTokenPos; |
| 31 |
$this->endLine = $endLine; |
| 32 |
$this->endFilePos = $endFilePos; |
| 33 |
$this->endTokenPos = $endTokenPos; |
| 34 |
} |
| 35 |
|
| 36 |
/** |
| 37 |
* Gets the comment text. |
| 38 |
* |
| 39 |
* @return string The comment text (including comment delimiters like /*) |
| 40 |
*/ |
| 41 |
public function getText(): string { |
| 42 |
return $this->text; |
| 43 |
} |
| 44 |
|
| 45 |
/** |
| 46 |
* Gets the line number the comment started on. |
| 47 |
* |
| 48 |
* @return int Line number (or -1 if not available) |
| 49 |
* @phpstan-return -1|positive-int |
| 50 |
*/ |
| 51 |
public function getStartLine(): int { |
| 52 |
return $this->startLine; |
| 53 |
} |
| 54 |
|
| 55 |
/** |
| 56 |
* Gets the file offset the comment started on. |
| 57 |
* |
| 58 |
* @return int File offset (or -1 if not available) |
| 59 |
*/ |
| 60 |
public function getStartFilePos(): int { |
| 61 |
return $this->startFilePos; |
| 62 |
} |
| 63 |
|
| 64 |
/** |
| 65 |
* Gets the token offset the comment started on. |
| 66 |
* |
| 67 |
* @return int Token offset (or -1 if not available) |
| 68 |
*/ |
| 69 |
public function getStartTokenPos(): int { |
| 70 |
return $this->startTokenPos; |
| 71 |
} |
| 72 |
|
| 73 |
/** |
| 74 |
* Gets the line number the comment ends on. |
| 75 |
* |
| 76 |
* @return int Line number (or -1 if not available) |
| 77 |
* @phpstan-return -1|positive-int |
| 78 |
*/ |
| 79 |
public function getEndLine(): int { |
| 80 |
return $this->endLine; |
| 81 |
} |
| 82 |
|
| 83 |
/** |
| 84 |
* Gets the file offset the comment ends on. |
| 85 |
* |
| 86 |
* @return int File offset (or -1 if not available) |
| 87 |
*/ |
| 88 |
public function getEndFilePos(): int { |
| 89 |
return $this->endFilePos; |
| 90 |
} |
| 91 |
|
| 92 |
/** |
| 93 |
* Gets the token offset the comment ends on. |
| 94 |
* |
| 95 |
* @return int Token offset (or -1 if not available) |
| 96 |
*/ |
| 97 |
public function getEndTokenPos(): int { |
| 98 |
return $this->endTokenPos; |
| 99 |
} |
| 100 |
|
| 101 |
/** |
| 102 |
* Gets the comment text. |
| 103 |
* |
| 104 |
* @return string The comment text (including comment delimiters like /*) |
| 105 |
*/ |
| 106 |
public function __toString(): string { |
| 107 |
return $this->text; |
| 108 |
} |
| 109 |
|
| 110 |
/** |
| 111 |
* Gets the reformatted comment text. |
| 112 |
* |
| 113 |
* "Reformatted" here means that we try to clean up the whitespace at the |
| 114 |
* starts of the lines. This is necessary because we receive the comments |
| 115 |
* without leading whitespace on the first line, but with leading whitespace |
| 116 |
* on all subsequent lines. |
| 117 |
* |
| 118 |
* Additionally, this normalizes CRLF newlines to LF newlines. |
| 119 |
*/ |
| 120 |
public function getReformattedText(): string { |
| 121 |
$text = str_replace("\r\n", "\n", $this->text); |
| 122 |
$newlinePos = strpos($text, "\n"); |
| 123 |
if (false === $newlinePos) { |
| 124 |
// Single line comments don't need further processing |
| 125 |
return $text; |
| 126 |
} |
| 127 |
if (preg_match('(^.*(?:\n\s+\*.*)+$)', $text)) { |
| 128 |
// Multi line comment of the type |
| 129 |
// |
| 130 |
// /* |
| 131 |
// * Some text. |
| 132 |
// * Some more text. |
| 133 |
// */ |
| 134 |
// |
| 135 |
// is handled by replacing the whitespace sequences before the * by a single space |
| 136 |
return preg_replace('(^\s+\*)m', ' *', $text); |
| 137 |
} |
| 138 |
if (preg_match('(^/\*\*?\s*\n)', $text) && preg_match('(\n(\s*)\*/$)', $text, $matches)) { |
| 139 |
// Multi line comment of the type |
| 140 |
// |
| 141 |
// /* |
| 142 |
// Some text. |
| 143 |
// Some more text. |
| 144 |
// */ |
| 145 |
// |
| 146 |
// is handled by removing the whitespace sequence on the line before the closing |
| 147 |
// */ on all lines. So if the last line is " */", then " " is removed at the |
| 148 |
// start of all lines. |
| 149 |
return preg_replace('(^' . preg_quote($matches[1]) . ')m', '', $text); |
| 150 |
} |
| 151 |
if (preg_match('(^/\*\*?\s*(?!\s))', $text, $matches)) { |
| 152 |
// Multi line comment of the type |
| 153 |
// |
| 154 |
// /* Some text. |
| 155 |
// Some more text. |
| 156 |
// Indented text. |
| 157 |
// Even more text. */ |
| 158 |
// |
| 159 |
// is handled by removing the difference between the shortest whitespace prefix on all |
| 160 |
// lines and the length of the "/* " opening sequence. |
| 161 |
$prefixLen = $this->getShortestWhitespacePrefixLen(substr($text, $newlinePos + 1)); |
| 162 |
$removeLen = $prefixLen - strlen($matches[0]); |
| 163 |
return preg_replace('(^\s{' . $removeLen . '})m', '', $text); |
| 164 |
} |
| 165 |
|
| 166 |
// No idea how to format this comment, so simply return as is |
| 167 |
return $text; |
| 168 |
} |
| 169 |
|
| 170 |
/** |
| 171 |
* Get length of shortest whitespace prefix (at the start of a line). |
| 172 |
* |
| 173 |
* If there is a line with no prefix whitespace, 0 is a valid return value. |
| 174 |
* |
| 175 |
* @param string $str String to check |
| 176 |
* @return int Length in characters. Tabs count as single characters. |
| 177 |
*/ |
| 178 |
private function getShortestWhitespacePrefixLen(string $str): int { |
| 179 |
$lines = explode("\n", $str); |
| 180 |
$shortestPrefixLen = \PHP_INT_MAX; |
| 181 |
foreach ($lines as $line) { |
| 182 |
preg_match('(^\s*)', $line, $matches); |
| 183 |
$prefixLen = strlen($matches[0]); |
| 184 |
if ($prefixLen < $shortestPrefixLen) { |
| 185 |
$shortestPrefixLen = $prefixLen; |
| 186 |
} |
| 187 |
} |
| 188 |
return $shortestPrefixLen; |
| 189 |
} |
| 190 |
|
| 191 |
/** |
| 192 |
* @return array{nodeType:string, text:mixed, line:mixed, filePos:mixed} |
| 193 |
*/ |
| 194 |
public function jsonSerialize(): array { |
| 195 |
// Technically not a node, but we make it look like one anyway |
| 196 |
$type = $this instanceof Comment\Doc ? 'Comment_Doc' : 'Comment'; |
| 197 |
return [ |
| 198 |
'nodeType' => $type, |
| 199 |
'text' => $this->text, |
| 200 |
// TODO: Rename these to include "start". |
| 201 |
'line' => $this->startLine, |
| 202 |
'filePos' => $this->startFilePos, |
| 203 |
'tokenPos' => $this->startTokenPos, |
| 204 |
'endLine' => $this->endLine, |
| 205 |
'endFilePos' => $this->endFilePos, |
| 206 |
'endTokenPos' => $this->endTokenPos, |
| 207 |
]; |
| 208 |
} |
| 209 |
} |
| 210 |
|