PHP去除文章的html格式

function cutstr_html($string,$length=0,$ellipsis='…'){
     $string=strip_tags($string);
       $string=preg_replace('/\n/is','',$string);
    $string=preg_replace('/ | /is','',$string);
       $string=preg_replace('/ /is','',$string);
       preg_match_all("/[\x01-\x7f]|[\xc2-\xdf][\x80-\xbf]|\xe0[\xa0-\xbf][\x80-\xbf]|[\xe1-\xef][\x80-\xbf][\x80-\xbf]|\xf0[\x90-\xbf][\x80-\xbf][\x80-\xbf]|[\xf1-\xf7][\x80-\xbf][\x80-\xbf][\x80-\xbf]/",$string,$string);        if(is_array($string)&&!empty($string[0])){
        if(is_numeric($length)&&$length){
            $string=join('',array_slice($string[0],0,$length)).$ellipsis;
        }else{
            $string=implode('',$string[0]);
       }
    }else{
        $string='';        }
   return $string;
}
function clearHtml($content)
{
    $content = preg_replace("/]*>/i", "", $content);
    @$content = preg_replace("/<\/a>/i", "", $content);
    $content = preg_replace("/]*>/i", "", $content);
    @$content = preg_replace("/<\/div>/i", "", $content);
    $content = preg_replace("//i", "", $content);//注释内容
    $content = preg_replace("/style=.+?['|\"]/i", '', $content);//去除样式
    $content = preg_replace("/class=.+?['|\"]/i", '', $content);//去除样式
    $content = preg_replace("/id=.+?['|\"]/i", '', $content);//去除样式
    $content = preg_replace("/lang=.+?['|\"]/i", '', $content);//去除样式
    $content = preg_replace("/width=.+?['|\"]/i", '', $content);//去除样式
    $content = preg_replace("/height=.+?['|\"]/i", '', $content);//去除样式
    $content = preg_replace("/border=.+?['|\"]/i", '', $content);//去除样式
    $content = preg_replace("/face=.+?['|\"]/i", '', $content);//去除样式
    $content = preg_replace("/face=.+?['|\"]/", '', $content);//去除样式 只允许小写 正则匹配没有带 i 参数
    return $content;
}

你可能感兴趣的:(PHP去除文章的html格式)