b1f53045d7
This is a bug fix release of RemexHtml, required by the latest version of Parsoid. RemexHtml migrated to a new namespace in 2.3.2. Since we don't support aliases in our phan configuration in core, update all uses to the new namespace to satisfy phan. Depends-On: I30f01f4a2a5479bb82c9b952ffa68a478215828a Depends-On: Iedf446635ee2112cfe637d8ebcf8092f0976bd17 Change-Id: I74fc929e4a66b28bfb1800ff0cd751c86e4a9f50
96 lines
2.4 KiB
PHP
96 lines
2.4 KiB
PHP
<?php
|
|
|
|
use Wikimedia\RemexHtml\Tokenizer\Attributes;
|
|
use Wikimedia\RemexHtml\Tokenizer\NullTokenHandler;
|
|
|
|
/**
|
|
* @internal
|
|
*/
|
|
class RemexStripTagHandler extends NullTokenHandler {
|
|
private $text = '';
|
|
|
|
public function getResult() {
|
|
return $this->text;
|
|
}
|
|
|
|
public function characters( $text, $start, $length, $sourceStart, $sourceLength ) {
|
|
$this->text .= substr( $text, $start, $length );
|
|
}
|
|
|
|
public function startTag( $name, Attributes $attrs, $selfClose, $sourceStart, $sourceLength ) {
|
|
// Inject whitespace for typical block-level tags to
|
|
// prevent merging unrelated<br>words.
|
|
if ( $this->isBlockLevelTag( $name ) ) {
|
|
$this->text .= ' ';
|
|
}
|
|
}
|
|
|
|
public function endTag( $name, $sourceStart, $sourceLength ) {
|
|
// Inject whitespace for typical block-level tags to
|
|
// prevent merging unrelated<br>words.
|
|
if ( $this->isBlockLevelTag( $name ) ) {
|
|
$this->text .= ' ';
|
|
}
|
|
}
|
|
|
|
// Per https://developer.mozilla.org/en-US/docs/Web/HTML/Block-level_elements
|
|
// retrieved on sept 12, 2018. <br> is not block level but was added anyways.
|
|
// The following is a complete list of all HTML block level elements
|
|
// (although "block-level" is not technically defined for elements that are
|
|
// new in HTML5).
|
|
// Structured as tag => true to allow O(1) membership test.
|
|
private const BLOCK_LEVEL_TAGS = [
|
|
'address' => true,
|
|
'article' => true,
|
|
'aside' => true,
|
|
'blockquote' => true,
|
|
'br' => true,
|
|
'canvas' => true,
|
|
'dd' => true,
|
|
'div' => true,
|
|
'dl' => true,
|
|
'dt' => true,
|
|
'fieldset' => true,
|
|
'figcaption' => true,
|
|
'figure' => true,
|
|
'footer' => true,
|
|
'form' => true,
|
|
'h1' => true,
|
|
'h2' => true,
|
|
'h3' => true,
|
|
'h4' => true,
|
|
'h5' => true,
|
|
'h6' => true,
|
|
'header' => true,
|
|
'hgroup' => true,
|
|
'hr' => true,
|
|
'li' => true,
|
|
'main' => true,
|
|
'nav' => true,
|
|
'noscript' => true,
|
|
'ol' => true,
|
|
'output' => true,
|
|
'p' => true,
|
|
'pre' => true,
|
|
'section' => true,
|
|
'table' => true,
|
|
'td' => true,
|
|
'tfoot' => true,
|
|
'th' => true,
|
|
'tr' => true,
|
|
'ul' => true,
|
|
'video' => true,
|
|
];
|
|
|
|
/**
|
|
* Detect block level tags. Of course css can make anything a block
|
|
* level tag, but this is still better than nothing.
|
|
*
|
|
* @param string $tagName HTML tag name
|
|
* @return bool True when tag is an html block level element
|
|
*/
|
|
private function isBlockLevelTag( $tagName ) {
|
|
$key = strtolower( trim( $tagName ) );
|
|
return isset( self::BLOCK_LEVEL_TAGS[$key] );
|
|
}
|
|
}
|