DOMParser, XML and nested nodes - help needed.

[willywongi]

Hi, I'm trying to build a filemanager for a cms. I'd like to have a page with JS that asks the server for the filelist, which in turn answers with a XML document.

This is a sample of the XML answer:

<?xml version="1.0" encoding="ISO-8859-15"?>
<items tot="1">
<item iid="60dddfb31b978dc4c36d4475211e6672" key="fid" title=" C-nar.jpg">
<node nType="icon" editable="false" key="thumb" width="720" height="576">[[url]]</node>
<node nType="parent" editable="false">
    <subnode nType="value" editable="true" key="title" title="Titolo"> C-nar.jpg</subnode>
    <subnode nType="value" editable="true" key="caption" title="Descrizione">Amaro al Sapor di carciofo</subnode>
    <subnode nType="value" editable="false" key="link" title="Link">[[url]]</subnode>
</node>
 
<node nType="select" editable="true" key="uid" title="Proprietario">
    <option value="1" selected="true">Amministratore</option>
    <option value="10000001" selected="false">SuperAdmin</option>
    <option value="2" selected="false">Pongi</option>
</node>
<node nType="select-multi" editable="true" key="groups:::gid" title="Visibile ai gruppi">
<option value="1" selected="true">Amministratori</option>
<option value="2" selected="false">Reggio Emilia</option>
<option value="3" selected="true">Pippo</option>
<option value="4" selected="false">Newsletter</option>
</node>
<node nType="value" editable="false" key="fsize">61.826KB</node>
<node nType="remove" editable="false"></node>
</item>
</items>

I did so with a number of others scripts, but this one is the only that seems to give me problems.
Firefox doesn't alert me of XML syntax errors, but when I'm going to traverse it through Javascript I can get the first childNodes (the <item> nodes) but everything below this level is parsed as a single #text node. How could it be?
I hope you can spot a big mistake in my xml.
Thanks everyone.

[acoder]

Welcome to TSDN!

Can you post the code you are using?

[willywongi]

Ok, that's not the actual code but I just skimmed some unimportant things

var tb = document.getElementById("myTable").tBody;
var XmlObj = parsexml(http.responseText);
for(var j=0;j<XmlObj.childNodes.length;j++) {
    if(XmlObj.childNodes[j].nodeType != 1) continue;
    var aItem = XmlObj.childNodes[j];
    aRow = document.createElement("TR");
    for(var i=0;i<XmlObj.childNodes[j].childNodes[i].length;i++) {
        var aItemNode = XmlObj.childNodes[j].childNodes[i];
        if(aItemNode.nodeType != 1) continue;
        // aValue is a Cell within the Row.
        var aValue = XmlObj.childNodes[j].childNodes[i];
        // append an item (aValue) to another element (aRow).
        // myAppendItem(aValue, aRow);
    }
    tb.appendChild(aRow);
}
 

The inner "for" iteration runs just once to find out that the only childNode is a #text one (nodeType != 1).

[acoder]

That is probably your problem. There are 12 different node types - see link. If you want to check for a text node, check for node type 3.

[willywongi]

Ok, that's true, but even if I skip the check for the nodeType, it turns out to be a node with no childNodes, as if the DOMParser won't traverse completely my xml, but without warnings or errors. That's why I'm lost.