Hello Gianugo,

I must admit that I am thinking about the Service more from the perspective
of someone looking to add support for XQuery to an implementation that
already has the XPath Service, and has people already using that XPath
Service.

To handle this I would like to introduce a new Service so that code built on
the XPath service will work in the same way without any changes.  New code
can take advantage of XQuery via the new Service, and old code can be
migrated to XQuery at a  time convenient for its developer (and clearly
"never" must be an option!).

Your point about how decoupled XQuery is from document instances is a good
one.  I seem to remember that in the past there was some general discussion
about the advantages and disadvantages of binding a Service to a
Collection....

Do you have an idea for an alternative approach?  I was thinking that adding
a new Service would fit nicely with the character of the API, but perhaps
there are sufficient arguments for introducing a new class (XQueryEngine?)
to the API - one which is not tied to a Collection?

I could see the interface being quite simple - perhaps just a method
accepting the query as a parameter and returning a ResourceSet, as with the
XPathQueryService.  Depending on how individual datastores implement XQuery,
maybe a generic method to configure the underlying XQuery engine would also
be useful.

Greetings,
Trevor.



> I for one would welcome this Service. The only problem I envision is
> that XQuery is *very* decoupled from a document instance or even a
> Collection (given its (ab?)use of document() functionalities) so it
> doesn't really make sense to have it bound to a particular collection.
> 
> How do you see this Service implemented?
> 
> Ciao,
> 
> -- 
> Gianugo Rabellino
> CTO
> Pro-netics s.r.l.
> http://www.pro-netics.com
----------------------------------------------------------------------
Post a message:         mailto:[EMAIL PROTECTED]
Unsubscribe:            mailto:[EMAIL PROTECTED]
Contact administrator:  mailto:[EMAIL PROTECTED]
Read archived messages: http://archive.xmldb.org/
----------------------------------------------------------------------

Reply via email to