Re: EXSLT for XSLT 2.0: a new life for EXSLT?

Hi everyone,

Let me introduce myself - my name is John Snelson, and I work for Oracle 
implementing Berkeley DB XML (an XML database), and XQilla (a 
stand-alone XQuery implementation, working towards also being an XSLT 
2.0 implementation). I'd like to offer any help I can give towards the 
EXSLT 2.0 initiative.

Micah Dubinko wrote:
> 1) We should be aware the context of the extension, namely XPath 2.0, 
> XSLT 2.0, or XQuery. A large part of EXSLT 1.0 was actually XPath 
> extensions. Now that multiple langauges are based on XPath, the 
> distinction is more relevant.

I'm a member of the XQuery working group, where I've been pushing to see 
if we can start some community initiatives (EXQuery?) for a while. I 
think both XQuery and XSLT users stand to benefit from a community 
focused on defining useful extensions to this family of languages.

> 2) In particular we talked about a few categories of functions:
> * Eval
> * Functions interesting to implementations that support both XSLT and 
> XQuery
> * Request/Response stuff

I agree completely - I've had a number of detailed discussions about all 
these areas recently. There are people who've already thought about 
these a lot, and I'm sure they could be persuaded to take part if 
appropriate.

Other ideas are:

* Inheriting the functions that are still useful from EXSLT (math, 
random...)
* Full support for REST calls out and in
* SQL access
* JSON parsing / serialization

A few more ambitious things could probably be achieved with functions, 
but it might make more sense to have new syntax for:

* Higher order functions (wrt Dimitre's work in XSLT)
* Tuples (named members, type known at compile time)
* Maps (more loosely typed, members only known at runtime)

John

-- 
John Snelson, Oracle Corporation            http://snelson.org.uk/john
Berkeley DB XML:            http://oracle.com/database/berkeley-db/xml
XQilla:                                  http://xqilla.sourceforge.net