Re: Numeric top-level predicates in patterns

On 2-4-2014 17:05, David Rudel wrote:
> On Wed, Apr 2, 2014 at 4:51 PM, Abel Braaksma (Exselt) <abel@xxxxxxxxxx> wrote:
>
>> Then, to make it tricky, what happens here, if the applied set is //foo?
>> (foo[2])[3]
>>
>> I did not find that well-defined in the spec or in Michael Kay's XSLT
>> 2.0 Programmer's Reference book.
> I agree (though I could be wrong) with everything you had up until the
> above. Kay does deal with this topic. See the section on "Filter
> Expressions."
>

That section (in 4th ed, pp 638) does not explain sibling predicates of
which more than one is a numeric predicate, and in the section on
patterns, that is neither mentioned. I agree it is explained in more
general terms, but I tried to find a source for those cases I laid out
in the original post, and couldn't readily find one (that doesn't mean
there is none, of course).

Cheers,
Abel