RE: What's implementable in XPath expressions?

 -----Original Message-----
> From: www-ql-request@w3.org [mailto:www-ql-request@w3.org]On Behalf Of
> Michael Dyck
> Sent: Wednesday, April 25, 2001 11:22 PM
> To: Howard Katz
> Cc: www-ql@w3.org
> Subject: Re: What's implementable in XPath expressions?
>
> Howard Katz wrote:
> >
> > What's actually implementable at the head of an XPath
> expression? I'm trying to figure out what's allowed by the
> following snippet from the JavaCup grammar:
> > ...
> > Certainly $VarName/somePath and DOT/somePath are
> > going to be  commonly seen, but what about the others?
> > What's the meaning of [ElementConstructor]/somePath,
> > for example? Or even [Literal]/somePath? How should I
> > interpret these constructs?
>
> <http://www.w3.org/TR/xpath#node-sets>, the corresponding point
> in the XPath spec, says
>
>     The / and // operators compose an expression and a relative
>     location path. It is an error if the expression does not
>     evaluate to a node-set.

That's the bigger question then, isn't it: Does [ElementConstructor]
evaluate to a node-set, or is it simply there to decorate serialized XML? If
it produces node-sets, how do I compose against it?

Howard

> If XQuery has similar semantics, I think
> [ElementConstructor]/somePath could
> be valid (if perhaps pointless), e.g.
>     <foo><bar/></foo> / bar
> but [Literal]/somePath would never be valid.
>
> -Michael Dyck
>
>