RE: Experiences comparing performance of XSLT vs. "raw

> I know this is a completely open question, but what kind of experience
> have people had comparing the runtime performance and 
> throughput of XSLT
> vs. "raw Java code"?  I would guess the only scope limitation I could
> give for this would be using Xalan and JDK 1.4.2.

Please don't measure with Xalan and assume that the results apply to XSLT
processors in general: other processors might be 10 times faster, or more.

I see many XSLT applications where the performance is dominated by parsing,
tree-building, and serialization. In such situations, Java isn't going to be
any faster.

Michael Kay
http://www.saxonica.com/