I am using XSLT for some years now for doing all kinds of processing like importing and exporting databases, converting markup code and generating web pages. Because I was using almost all of XSLT‘s features and some important extensions of the standard EXSLT function set, I needed a conforming and feature-rich XSLT processor.
So my first choice always was Michael Kay‘s most excellent Saxon which not only covers the published standards completely, but is always a bit ahead of the pack. This comes as no surprise as Michael Kay is the maintainer of the upcoming XSLT 2.0 standard and has written a highly recommendable book on XSLT which unfortunately seems to be out of print as the publisher Wrox came into deep trouble recently and their fate is unclear to me.
However, Saxon is written in Java and therefore it is pretty slow. The good news is that the other XSLT processors have finally catched up with standards support and I have happily switched to the XSLT C Library of the GNOME project which is much faster and has good support for EXSLT, multiple output documents and some SAXON extensions (like explicit XHTML support). Well done, guys! This made my day.