Re: SequenceTyping matching of 'element QName'

Kay, Michael wrote:

> I think that typeswitch is so central to the static typing approach 
> that the ideas you suggest for extending it wouldn't fly.

No, I expected that.  But is it possible to change the definition
of SequenceType to that 'element QName' can be used without a Schema?
Being able to specify "an element node with a given QName" is
*very* useful, and I suspect *more* useful than "has a specific name
and that has been validated as conforming to the schema definition for
that name", especially given the existence of Basic XQuery.

Expressing "an element node with a given QName" is very difficult with
the current definition, even though I think it is something a lot
of people will want to do, while expressing "has a specific name
and that has been validated as conforming to the schema definition for
that name" can be done using 'element QName1 of type QName2', as long as
the Scheme defines a name for the element's type.

Being able to succinctly express "element with a given QName" is a
very important use-case as I see it, and we should be able to express it
without requiring Schemas.  (Even after I implement Schemas Import, I
suspect many people will be put off if they are expected to write
schemas.)

(I didn't read carefully enough earlier, and I realize I should have
caught this sooner.)
-- 
	--Per Bothner
per@bothner.com   http://www.bothner.com/per/

Received on Friday, 18 October 2002 14:50:52 UTC