Ignore Compression when javascript

[jerry_louise]

I use a function to compress html output now i would like to know if anyone knows how to make it ignore everything in the <script> tag so it does not mess up my javascript.

function

ob_start("compress");
function compress($htmlOutput) 
{
	$htmlOutput = preg_replace('/<!--(.|\s)*?-->/', '', $htmlOutput);
    $htmlOutput = str_replace("\n\n","\n",str_replace("\r\n","\n",str_replace("  "," ",str_replace("\t","",$htmlOutput))));
    $htmlOutput = ereg_replace("\n", "", $htmlOutput);
    return $htmlOutput;
}

/// HTML stuff here

ob_end_flush();

[johanafm]

First off, ereg_* have been deprecated and will be dropped as of PHP 6, so you should stick exclusively to PCRE.

First replacing \r\n with \n, then \n\n with \n and finally \n with '' is the same as replacing \r\n with '', then \n with ''.

The easiest approach would proably be to use DOM.

$dom = new DOMDocument();
$dom->loadHTML($str);

foreach($dom->childNodes as $c)
	descendAndCleanText($c);

function descendAndCleanText($n) {
	if ($n->nodeName == '#text') {
		$from = array("\r\n", "\n", '  ', "\t");
		$to = array('', '', ' ', '');
		$n->nodeValue = str_replace($from, $to, $n->nodeValue);
	}
	else if ($n->hasChildNodes()) {
		$comments = array();
		foreach($n->childNodes as $c) {
			if ($c->nodeName == '#comment') {
				// can't delete child nodes while foreaching through them in this way
				$comments[] = $c;
			}
			// not really necessary, since script doesn't contain #textNode, but CDATA...
			// still, there is no use descending into script nodes only to find this out anyway.
			else if ($c->nodeName != 'script') {
				descendAndCleanText($c);
			}
		}
		foreach($comments as $c)
			$c->parentNode->removeChild($c);
		$n->normalize();
	}
}