When loading a XSLT document that contains a script you must enable the AllowXsltScript property. See this forum post for more info. This will enable the use of msxml:script element functionality. The default value is changed to false in version 6.0. We use XSLT in our report generator to easily generate reports in any format (HTML, XML, Excel, CSV and custom formats). Read more on XSLT Security.

Technorati: MSXML, XML, XSLT

Sample script which goes after the xslt namespace:

<msxml:script language="VBScript" implements-prefix="VBCustomScript">
<![CDATA[
function VBCustomScript(str)
VBCustomScript= "test" & str
 end function
]]>
</msxml:script>

add to xslt namespace:

<xsl:stylesheet version="1.0"
xmlns:xsl="http://www.w3.org/1999/XSL/Transform"
xmlns:msxml="urn:schemas-microsoft-com:xslt"
xmlns:VBCustomScript="#VBCustomScript"
>

in xslt:

<xsl:value-of select=""VBCustomScript:VBCustomScript(string(@Field))""/>

The changes from the current XML sniffer will be:

- Check if IE or Mozilla/Opera.

- Array with all Microsoft XML Parser versions. Loop through and find newest version installed. Better than the current try/catch of all versions which is currently used.

- Include W3C standard of XMLHTTPRequest and DOMDocument together with all the current included objects from Microsoft XML Parser.

Anyone got some thoughts regarding this?

This post will explain how to load a XML document and then loop through all elements and display the value. It will give you a brief introduction in how to use the DOMCDocument object in Microsoft XML Parser and how the DOM works.

XML looks like this:

<?xml version="1.0" encoding="UTF-8"?>
<Users>
	<User>Espen</User>
	<User>Bill</User>
	<User>Steve</User>
</Users>

I have used a simple XML with one root element and three child nodes, each with their own node text. The DOM represent the XML like this:

Users (documentElement)

User (childNode 2) – Steve (childNode 0 nodeText)

The is made up of node elements like shown above. These elements can have child and/or parent elements, attributes and a node text. We can use the DOM in JavaScript to retrieve and manipulate these elements. See the DOM specification for more information.

Below is the javascript code. It requires this XML Parser sniffer!

<script type="text/javascript" src="msxml_sniffer.js"></script>

<script type="text/javascript">
// requires XML Parser sniffer
var formValues_DOM = NewXMLParser('DOMDocument');
var formValues;

function getData() {
	var url;
	url = 'users.xml';
	if(typeof FormValues_DOM != 'object') {
		if(formValues_DOM > '') {
			formValues_DOM = new ActiveXObject(formValues_DOM)
		} else {
			alert('XML Parser not installed');
		}
	}
	formValues_DOM.async = false;
	formValues_DOM.load(url);
	if(formValues_DOM.parseError == 0) {
		formValues = formValues_DOM.documentElement;
	insertValues();
	}
	else {
		alert('XML not loaded');
	}
}

function insertValues() {
	var nodeValue;
	for(var i=0;i < formValues.childNodes.length;i++) {
		nodeValue= formValues.childNodes[i].text;
		// do something with the value here:
		alert(nodeValue);
	}
}
</script>

I posted a question on Expert Exchange and got an answer from someone who had used indent=”yes” in XSLT output when using method=”xml”. That worked fine when used with my Excelt XSLT file.

But when I was going to make a CSV export using XSLT, the same formula didn’t work. After some googling and testing I found this solution: Use method=”text” and indent=”yes” in xsl:output. Use xsl text and put a line break (#x0D;) and use xml:space=”preserve”.

Note the key here is xml:space=”preserve”. Without it won’t work. It only worked if I added some text before the line break. But obviously, in a CSV, I don’t wont test at the end of the line.

Arghh..How can i post html/xml code in my blog? I have used the CODE tag earlier when posting ASP and Javascript code. But doesn’t work with XML, it doesn’t show up in the post.