On 4/4/06, Larry Wall <[EMAIL PROTECTED]> wrote:
> On the other hand, if junctions really are sets of sets, then maybe it's
> a mistake to autocoerce junctions to sets by swiping their internal set
> of values.  Arguably any(1,2,3) should coerce not to
>
>     (1,2,3)
>
> but to
>
>     (
>         (1),
>         (2),
>         (3),
>         (1,2),
>         (1,3),
>         (2,3),
>         (1,2,3),
>     )

I don't follow.  Why is that the representation of any(1,2,3)?  Is
this a disjunctive normal form; i.e. is 2 < any(1,2,3) equivalent to
the test:

     2 < 1
  || 2 < 2
  || 2 < 3
  || 2 < 1 && 2 < 2
  || ...

Or did you have something else in mind?

Luke

Reply via email to