Purify.php

<?php

require_once('/home/alix/_.php');

ph();

function Purify($html, $tags = null)
{
	$html = preg_replace('~<(script|style)[^>]*>.*?(?:</\1>|\z)~is', '', $html);
	$html = strip_tags($html, '<' . implode('><', array_keys((array) $tags)) . '>');

	if (is_array($tags) === true)
	{
		$html = strip_tags($html, '<' . implode('><', array_keys($tags)) . '>');

		if (function_exists('dom_import_simplexml') === true)
		{
			if (is_object($dom = @dom_import_simplexml(ph()->Net->XML($html))) === true)
			{
				$dom = $dom->ownerDocument;

				foreach ($tags as $tag => $attributes)
				{
					foreach ($dom->getElementsByTagName($tag) as $node)
					{
						if (($attributes !== true) && ($node->hasAttributes() === true))
						{
							$attributes = explode('|', $attributes);

							foreach (range(max(0, $node->attributes->length - 1), 0) as $i)
							{
								$attribute = $node->attributes->item($i);

								if ((empty($attributes) === true) || (in_array($attribute->name, $attributes) !== true))
								{
									$node->removeAttributeNode($attribute);
								}
							}
						}
					}
				}

				$html = preg_replace(array('~<(?:!DOCTYPE|/?(?:html|body))[^>]*>[[:space:]]*~i', '~^<p>|</p>$~i'), '', $dom->saveHTML());
			}
		}
	}

	else
	{
		$html = strip_tags($html);
	}

	return $html;
}

$html = 'hel<a>lo wo<b ID="one" foo="bar" class="strong">r</b>ld</p>';

ph()->Dump(Purify($html, array('b' => 'id|class')));
ph()->Dump(Purify($html, array('b' => null)));
ph()->Dump(Purify($html, array('b' => false)));
ph()->Dump(Purify($html, array('b' => array())));

########################################################################

$tags = array(); // strip all tags

$tags = array
(
    'a' => '', // preserve <a> tags, but strip all attributes
    'b' => null, // preserve <b> tags, but strip all attributes
    'c' => false, // preserve <c> tags, but strip all attributes
    'd' => array(), // preserve <d> tags, but strip all attributes
);

$tags = array
(
    'a' => 'id|class', // preserve <a> tags, but strip all attributes except "id" or "class"
    'b' => array('id', 'class'), // preserve <b> tags, but strip all attributes except "id" or "class"
);

// TODO:
$tags = array
(
	'b' => array // preserve <b> tags
	(
		'id' => 'post[0-9]+|comment[0-9]+' // but strip all attributes except "id" if value matches regex
	),
);

?>