Question

I’ve had a section of my JavaScript code that’s been working fine for a while now to grab values from incoming xml comms, but it has suddenly stopped reporting any found elements by certain tags. I did some looking and it seems that only tags that are numbers get affected by this, once I switch back to letters everything works great. Is there some reason why I can’t use just numbers?

Javascript Code:

xmlhttp.onreadystatechange=function()
{
    xmlDoc=xmlhttp.responseXML;
    x=xmlDoc.getElementsByTaName("VAR");
    alert(x.length); // <- reports 0 when numbers are used as tags
    var dataBack = [];
    for (j=0;j<x.length;j++) {
        dataBack[j] = x[0].getElementsByTagName(x[j])[0].childNodes[0].nodeValue);
    }
    useXMLdata(dataBack);
 }

I have tried adding a single letter to my number tags and it magically starts to work. I’m using this to actually catch the tags, but I’m curious why I can’t use only numbers.

Answer 1

Elements which are only numbers (e.g. <8>Some content</8>) do not appear to be valid XML.

The XML specification is pretty turgid, but section 3.1 defines start tag naming. A start tag name must begin with a NameStartChar character and then continue with any number of NameChar characters.

NameStartChar appears to be a subset of NameChar which does not include the digits 0-9 among other things. Therefore a valid tag name cannot begin with a numeric digit.

I doubt your XML will validate if it contains XML elements such as <8>Something</8>. Whether that is the reason that your JavaScript fails I have no idea, but it’s reason enough not to structure your XML in that way.

Edit

Try plugging the following XML into a validator:

<?xml version="1.0" ?> 
<Root>
  <Child>
    <1xyz>Content</1xyz>
  </Child>
</Root>

This gives the error Invalid element name for the <1xyz> tag. Even SO doesn’t like it hence the lack of syntax highlighting!