Hi,
i've been using (like you can see in the last topic) a XPath processor for Java, the problem is that it's very slow, because i want to parse a XML file directly from the Internet (from a URL)
I have this code

XPath xpath = XPathFactory.newInstance().newXPath();
xpath.setNamespaceContext(new RBTNameSpace());
InputSource inputSource = new InputSource(uaprof);

String expr = "/rdf:RDF/rdf:Description/prf:component/rdf:Description/prf:Vendor/text()";
String brand = xpath.evaluate(expr,inputSource);

XPath used is

import javax.xml.xpath.XPath;

Is there any different (faster) XPath Class, or should i use Xalan or Saxon ?

I think the problem is when i create a new InputSource with the file.
I know that XTags from Jakarta is very fast, the only problem is that it only work with JSP

Thank you

My first guess would be that the response time you are seeing is dominated by the time taken to retrieve the XML file from the remote site and parse it, and therefore your choice of XPath processor (which comes into play only when that process is complete) is going to make very little difference. How big is the file?

Michael Kay
http://www.saxonica.com/
Author, XSLT Programmer's Reference and XPath 2.0 Programmer's Reference

Incidentally, your import statement only declares the XPath API interface, this doesn't determine which XPath implementation you are using. You're probably using the default one in the JDK (Xalan), but your code wouldn't have to change to use Saxon, you just have to put Saxon on the classpath.

Michael Kay
http://www.saxonica.com/
Author, XSLT Programmer's Reference and XPath 2.0 Programmer's Reference

The file is very small, it's less then 20kb (17kb i think)
What i find strange, is that with the Jakarta XTags Lib, it's very fast

Thanks, i found the problem
i have several expression to evaluate, and i thought that the file was read in to memory when i wrote
InputSource inputSource = new InputSource(FILE);

But the file is read in each
String brand = xpath.evaluate(expr,inputSource);

Now i change it with a
Document document = builder.parse(new InputSource(uaprof));
On top of the code, then the document is read and each
String brand = xpath.evaluate(expr,document);

Now it's much faster :D
Thanks