↑
Main Page

Parsing errors

Because the

xml

property needs to be available on every type of node in a document, it’s best to add it to

the

Node

class itself (remember, all other node types inherit from

Node

Node.prototype.__defineGetter__(“xml”, function () {

var oSerializer = new XMLSerializer();

return oSerializer.serializeToString(this, “text/xml”);

});

The function assigned to the

xml

property is very simple, and the only change from the earlier example

is that

this

is the first argument for the

serializeToString()

method (remember, in this context

this

refers to the node). If you include this code in a page, it’s possible to use this custom

xml

property

in the same manner as Microsoft’s

xml

property:

oXmlDom.load(“test.xml”);

alert(oXmlDom.xml);

var oNode = oXmlDom.documentElements.childNodes[1];

alert(oNode.xml);

Parsing errors

When an error occurs in the parsing of an XML file, the XML DOM creates a document explaining the

error. Suppose you ran the following code:

var oParser = new DOMParser()

var oXmlDom = oParser.parseFromString(“<root><child></root>”);

Although no error is thrown,

oXmlDom

shows you the error. In this case, it presents this code:

XML Parsing Error: not well-formed

Location: file://c:/Chapter 15/examples/MozillaXmlDomExample.htm

Line Number 5, Column 1:<sourcetext><root><child></root>

--------------^</sourcetext>

</parsererror>

So to determine if there’s an error in the parsing of XML code, you must test the tag name of the docu-

ment element:

if (oXmlDom.documentElement.tagName != “parsererror”) {

//continue on, no errors

} else {

//do something else, there was an error

}

This code must also be surrounded by a browser detect because it only works in

Mozilla.

454

Chapter 15

18_579088 ch15.qxd 3/28/05 11:42 AM Page 454