php DOMDocument: html to node tree

This function converts html into a node tree (<ul> structure). However, every node is returned as a child node of the preceding node even if that node was a sibling of the current node.

$xml = '
<div>

   <div>
       <b></b>
   </div>

   <p></p>

</div>
';


function xml2array($xml,&$result = '') {

    foreach($xml->children() as $name => $xmlchild) {
         xml2array($xmlchild, $result);

    }
    $result = "<ul><li>".$xml->getName().$result."</li></ul>";

}


$result='';
$dd = xml2array(simplexml_load_string($xml), $result);


echo "<pre>";
print_r($result);

the above code returns this:

   <ul>
      <li>div
         <ul>
            <li>p
            <ul>
                <li>div
                       <ul>
                           <li>b</li>
                       </ul>
                 </li>
           </ul>
          </li>
      开发者_Go百科</ul>
   </li>
  </ul>

you notice 'div' is now a child of 'p' even though they are siblings, and 'div' comes after 'p' unlike the orignal input.

this is how it should look:

<ul>
    <li>div
    <ul>
     <li>div
       <ul>b</ul>
    </li>

    <li>p</li>

    </ul>
    </li>

</ul>

Here you go, a non-broken function, with the call by reference removed because if you don't know exactly what they are, possibilities of obscure bugs occur (and you might know how/when to use them, the next coder might struggle, references are best used with a very clear reason or not at all):

<?php
$xml = '
<div>
   <div>
       <b></b>
   </div>
   <p></p>
</div>
';


function xml2ul($xml) {
    $children = $xml->children();
    if(empty($children)) return '';
    $result = '<ul>';
    foreach($xml->children() as $name => $xmlchild) {
        $result .= '<li>'.$name.xml2ul($xmlchild).'</li>';
    }
    $result .= '</ul>';
    return $result;
}

echo xml2ul(simplexml_load_string($xml), $result);