Hi Bruce,

Not sure if "()" always indicates function call.
XPath 1.0 grammar says

PathExpr = FilterExpr | . . . [ignoring others like LocationPath for ]
FilterExpr = PrimaryExpr | FilterExpr Predicate
PrimaryExpr = '(' Expr ')' | . . .  [others too in the list, ignoring
them for convenience]
Where Expr is an 'OrExpr', which can be another PathExpr.

See [19] in section 3.1 of Xpath 1.0
specification.[http://www.w3.org/TR/xpath]

-Menon

On 5/2/05, Bruce D'Arcus <bdarcus@xxxxxxxxx> wrote:
> On 5/2/05, Ramkumar Menon <ramkumar.menon@xxxxxxxxx> wrote:
>
> > Sorry for not being clear enough, but I had meant to ask "why are
> > patterns like (foo) or (/child::foo) considered invalid patterns as
> > per XSLT specification ?"
> > Curious  as to whether this kind of a pattern breaks any semantics of XSLT
.
> > rgds,
>
> I'll let someone else step in to correct me if I get the language wrong,
but:
>
> / or foo are nodes, while () indicates a function call.  So
> "count(current())" or "count(foo)" would both be valid, because the
> current() in the first is a function, while the foo in the second is
> just a node..
>
> Bruce
>
>


--
Shift to the left, shift to the right!
Pop up, push down, byte, byte, byte!

-Ramkumar Menon
 A typical Macroprocessor