public function testHTMLClean()
{
$cleaner = HTMLCleaner::inst();
if ($cleaner) {
$this->assertEquals($cleaner->cleanHTML('<p>wrong <b>nesting</i></p>'), '<p>wrong <b>nesting</b></p>', "HTML cleaned properly");
$this->assertEquals($cleaner->cleanHTML('<p>unclosed paragraph'), '<p>unclosed paragraph</p>', "HTML cleaned properly");
} else {
$this->markTestSkipped('No HTMLCleaner library available (tidy or HTMLBeautifier)');
}
}
/**
* Attempt to clean invalid HTML, which messes up diffs.
* This cleans code if possible, using an instance of HTMLCleaner
*
* NB: By default, only extremely simple tidying is performed,
* by passing through DomDocument::loadHTML and saveXML
*
* @param string $content HTML content
* @param HTMLCleaner $cleaner Optional instance of a HTMLCleaner class to
* use, overriding self::$html_cleaner_class
* @return mixed|string
*/
public static function cleanHTML($content, $cleaner = null)
{
if (!$cleaner) {
if (self::$html_cleaner_class && class_exists(self::$html_cleaner_class)) {
$cleaner = Injector::inst()->create(self::$html_cleaner_class);
} else {
//load cleaner if the dependent class is available
$cleaner = HTMLCleaner::inst();
}
}
if ($cleaner) {
$content = $cleaner->cleanHTML($content);
} else {
// At most basic level of cleaning, use DOMDocument to save valid XML.
$doc = Injector::inst()->create('HTMLValue', $content);
$content = $doc->getContent();
}
// Remove empty <ins /> and <del /> tags because browsers hate them
$content = preg_replace('/<(ins|del)[^>]*\\/>/', '', $content);
return $content;
}
