Bruce,



If I can add to this conversation a bit...



As far as C and C++ is concerned, no, there are no complete solutions that give 
you what XSLT 2.0 gives you beyond what Michael has suggested via a Web Service 
interface.  However, if I can put in a quick plug for the Saxon.NET project 
(which is currently in beta for Saxon 8.0-B and I am about to release an 8.1-B 
release to the SF.net site and Saxon.NET project site) and suggest to you that 
if it is C-styled syntax you are interested in -- or if in particular your are 
interested in using XSLT 2.0 on a Microsoft platform -- then maybe Saxon.NET 
will give you the capability you are looking for.



But let me first preface this with the following... The goal of the Saxon.NET 
project is to provide to the .NET community an API that is 100% compliant with 
that of Dr. Michael Kay's latest Saxon API and that can be used to implement a 
XSLT 2.0, XPath 2.0, and XQuery 1.0 transformation via any language and 
subsequent compiler that implements a solution supported by version 1.1+ of the 
.NET framework.  If the solution you are using (via C++.NET or C#) falls into 
this category, then great!  If not, then this is not the solution for your 
project.  But don't fret!  If not now, there will be XSLT 2.0 solutions, via 
your desired mechanism, in the months, not years, to come.



With all of this said, Dr. Kay's suggested solution is by far and beyond the 
best solution for dealing with projects that have a need for cross-platform 
transformation of XML data via a XSLT 2.0 processor.  Saxon.NET is currently 
built upon the Basic version of the specification set forth in the latest 
working draft from the W3C and as such does not provide the expanded 
Schema-Aware capabilities that come from the Saxon-SA product from Dr. Kay and 
Saxonica.



However, I should mention that the Saxon.NET project is currently in discussion 
with Saxonica about porting Saxon-SA to the .NET platform once the Saxon-B port 
is complete and proven.  But the focus on the Saxon-B port being complete and 
proven, if not already obvious, should be brought to the forefront of emphasis. 
 At the present time the Saxon.NET-B port is in beta form and is not complete 
nor is it proven and as such should not be a considered as a complete XSLT 2.0. 
XPath 2.0, and XQuery 1.0 solution for the .NET platform.  But this is something 
that we are aggressively working on obtaining and we look forward to the time 
that we can label the release as "complete and proven" and ready to be used 
within production level code on the .NET platform.  As such we also look forward 
to the time that Dr. Kay and Saxonica announce that Saxon-SA is not only 
available for the Java platform but for the .NET platform as well.



Best regards and best of luck to you as you decide on which solution will work 
best for you and your project.



<M:D/>







Bruce D'Arcus wrote:

On Oct 1, 2004, at 5:03 PM, Michael Kay wrote:



Consider implementing the transformation as a web service and invoking it

from the client application via HTTP calls. The client need never know 
that

the transformation is done using XSLT, let alone that it's done using an

XSLT processor written in Java.




Interesting idea.  Any examples I can point people to?



As I see it, it'd just involve passing the document and a single 
parameter to the web service, and getting back the formatted document.



Contexts in which it would likely be used?  Web applications (often 
written in PHP), as well as desktop applications like OpenOffice.



As a possibly relevant aside, the OpenOffice bibliographic project which 
I hope to see this become part of is being designed around a web service 
record query model (for sucking in records from remote databases).



And does such a service get around the JVM startup problem?



Bruce