Re: *which* alternative that matches nothing? (was Re: repetition)

I’m suggesting my preference for a single representation of an empty set,
from the selection that Steven offered- on grounds of simplicity.
 However you name it, null, empty set or empty sequence ( end user view cf
implementers view)

Regards


On Fri, 17 Dec 2021 at 14:36, C. M. Sperberg-McQueen <
cmsmcq@blackmesatech.com> wrote:

> Dave, you keep switching between () and [] — that way lies
> confusion, since they mean different things. [] is a character
> inclusion with no members and matches nothing, () is a set of
> alternatives containing a single alternative matching the empty
> sequence.
>
> Michael
>
>
> > On 17,Dec2021, at 2:45 AM, Dave Pawson <dave.pawson@gmail.com> wrote:
> >
> > On Fri, 17 Dec 2021 at 09:43, Steven Pemberton <steven.pemberton@cwi.nl>
> wrote:
> >>
> >>> Tks for the clarification Steven.
> >>> <myView> KISS principle, have one option - from the given list ()
> >>> seems clearest</myView>
> >>
> >> I completely agree with the KISS principle, but the use of () doesn't
> come
> >> from a design for representing empty, but from generality and
> consistency.
> >> There are a number of ways you could explicitly mark empty alternatives,
> >> but they emerge from generality, not from a use case.
> >>
> >> I did once consider allowing empty strings
> >>
> >>   empty: "".
> >>
> >> but it didn't add any functionality.
> >
> > I'll reiterate, [] says empty most clearly for me?
> >
> > regards
> >
> >
> >
> > --
> > Dave Pawson
> > XSLT XSL-FO FAQ.
> > Docbook FAQ.
> >
>
> --
Dave Pawson
XSLT XSL-FO FAQ.
Docbook FAQ.