IE9 NOT getting 'children' of XML node

Question

I have the following XML sitting in a var called RoomPriceInfo in javascript:

<?xml version="1.0" encoding="UTF-8"?>
<BkgItemHotelRoomPrices CurrCode="EUR">
  <RoomType Code="DB" Count="1" Desc="Double" Age="0">
    <PriceInfo EndDate="2011-12-17" AgentMarkup="0.0" MarkupPerc="0.1075" FitRdg="0.25"  MarkupPrice="48.73" AgentPrice="48.75" StartDate="2011-12-11" Nights="7" FitRdgPrice="48.75" CurrDec="2" CurrDecPrice="48.75" SuppPrice="44.0"/>
  </RoomType>
</BkgItemHotelRoomPrices>

and the following code:

DBRoomPrice = RoomPriceInfo.doXPath("//RoomType[@Code='DB']");
alert(DBRoomPrice[0].children.length);

Under FF7 on Ubuntu and FF8 on WinXP I get an alert of 1 which is correct. However under IE8 on WinXP and IE9 on Windows 7 nothing happens. It just dies silently.

Please can anyone shed any light on this? If I do a getElementById on the DOM object and then ask for children on that, then IE8 & IE9 behave correctly.

Answer 1

Internet Explorer (including version 11!) does not support the .children property om XML elements.

If you want to get the number of child elements, use element.childElementCount (IE9+):

element.children.length;   // Does not work in IE on XML elements
element.childElementCount; // Works in every browser

If you merely want to know whether an element has any children, you can also check whether element.firstElementChild (or element.lastElementChild) is not null. This property is supported in IE9+:

element.children.length === 0;      // All real browsers
element.firstElementChild !== null; // IE 9+

If you want to iterate through all child elements of the XML node, use childNodes and exclude the non-element nodes via their nodeType:

for (var i = 0, len = element.childNodes.length; i < l; ++i) {
    var child = element.childNodes[i];
    if (child.nodeType !== 1/*Node.ELEMENT_NODE*/) continue;
    // Now, do whatever you want with the child element.
}