Problem with using AJAX in Mozilla

vinay.sindri · January 25th, 2008, 03:36 AM

Hi!

I am reading an XML response in Mozilla. The response XML is like below:

<?xml version="1.0" encoding="ISO-8859-1" ?>
<Categories>
<Category>
<CategoryId>4</CategoryId>
<CategoryName>Consultant</CategoryName>
<DepartmentId>11</DepartmentId>
</Category>
<Category>
<CategoryId>3</CategoryId>
<CategoryName>MST Program Development</CategoryName>
<DepartmentId>11</DepartmentId>
</Category>
<Category>
<CategoryId>2</CategoryId>
<CategoryName>MSTnet Testing</CategoryName>
<DepartmentId>11</DepartmentId>
</Category>
<Category>
<CategoryId>1</CategoryId>
<CategoryName>Videotape Training Project</CategoryName>
<DepartmentId>11</DepartmentId>
</Category>
</Categories>

I am trying follwoing code:

xmlHttp.responseXML.documentElement.childNodes[i-1].childNodes[1].text;

But in Mozilla, it is saying undefined. Whereas, it is coming in IE.

samjudson · January 25th, 2008, 04:32 AM

Thats because according to the W3C DOM standard there is no .text property on a node.

The following should work:

xmlHttp.responseXML.documentElement.childNodes[i-1].childNodes[1].firstChild.nodeValue

That gets the nodeValue of the text node contained within the CategoryName node.

/- Sam Judson : Wrox Technical Editor -/

Martin Honnen · January 25th, 2008, 10:30 AM

IE uses MSXML for XML parsing and the DOM implementation of MSXML has a property named text that is not standardized.
Mozilla follows the W3C DOM where the Level 3 Core DOM defines a property named textContent which you can use to access the text contents of an element node.
You should also be aware that accessing nodes by indexing the childNodes collection is error-prone in a cross-browser way as there are different ways white space text between elements is treated when building the DOM model. IE/MSXML ignore such white space while Mozilla includes it as text nodes so for instance with the following XML

Code:

<root>
  <foo>
    <bar>value</bar>
  </foo>
</root>

you can use xmlDocumentInstance.documentElement.firstChild to access the foo element with IE/MSXML while with Mozilla that expression accesses a text node with white space.
A better approach is to use getElementsByTagName or XPath to make sure you access only elements.