Regular expression, parsing bad html in a xml document (strange)

I'm having problem with regular expression, not a good eek this week it seen like I alway's get a wall of problem. I know that it surely been ask a 1000 times, I look around, didn't find anythings, if you find somethings please point me out. So here what I want to do, I need to parse a xml document , but before to parse it I need to get rid of bad html that I don't want, but the document that I want require some stuff that I need too, so I don't want to get ride of all they HTML. So what I want to do, I already did a little bite of code that get out my good element and check for bad stuff, the only bad thing is that "text<text-1" is a good stuff, but I need to change < to &lt; or it will do bad things with my xml parser. Here what I try $simple = <<<XMLDATA <?xml version='1.0'?> <!DOCTYPE chapter SYSTEM "/just/a/test.dtd" [ <!ENTITY plainEntity "FOO entity"> <!ENTITY systemEntity SYSTEM "xmltest2.xml"> ]> <item> text <bad stuff> text<text-1 text <image title="Ceci est mon titre2" description="Ceci est ma description" link="http://www.windplanet.com/" url="http://www.windplanet.com/images/news/988991159.gif" align="left" width="235" height="131" size="13310"/> text text <image title="Ceci est mon titre" description="Ceci est ma description" link="http://www.windplanet.com/" url="http://www.windplanet.com/images/news/988991159.gif" align="left" width="235" height="131" size="13310"/> </item> XMLDATA; //$simple = str_replace("\n\n"," &lt;br/> &lt;br/> ",$simple); /* trouve moi tous les < sauf suivant ceci ... */ $data = $simple; print $data; if(preg_match_all("/\<(?:(?:\!|\/|\?|)(?:<!xml|<!DOCTYPE|<!ENTITY|<!image|<!item|))/",$data,$cbadhtml)){ foreach( $cbadhtml as $key => $myarray){ foreach( $myarray as $key2 => $myarray2){ print "<p><font color='red'>You can't use HTML here so ". htmlentities($myarray2) ." is not allowed</font></p>\n"; } } // what html? we exit //exit; } It find all the < but doesnt' remove the one that I accept, so how can I find the bad < and transform them to &lt; ? Thank you and have a nice day. -- Francis Fillion, BAA SI Broadcasting live from his linux box. And the maintainer of http://www.windplanet.com