Nettoyer une page HTML de ces balises

1. function html2txt($document){
2.         $search = array(‘@<script[^>]*?>.*?</script>@si’, // Strip out javascript
3.         ‘@<style[^>]*?>.*?</style>@siU’, // Strip style tags properly
4.         ‘@<[?]php[^>].*?[?]>@si’, //scripts php
5.         ‘@<[?][^>].*?[?]>@si’, //scripts php
6.         ‘@<[\/\!]*?[^<>]*?>@si’, // Strip out HTML tags
7.         ‘@<![\s\S]*?–[ \t\n\r]*>@’, // Strip multi-line comments including CDATA
8.         );
9.         $text = preg_replace($search, ”, $document);
10.         return $text;
11. }

Note