I was assuming that Rob wasn't trying to generate a result set of "all
records that don't have any of these words" i.e. almost the whole database
(for the same reason that Z39.50 doesn't let you do it but instead defines
and_not, a binary operator, and doesn't allow a unary NOT), but rather that
he was talking about doing so as part of a boolean query as in "any of these
but none of those". Possible I misread him (but I doubt it).
----- Original Message -----
From: "Mike Taylor" <[log in to unmask]>
To: <[log in to unmask]>
Sent: Wednesday, May 28, 2003 7:17 PM
Subject: Re: CQL Range Search Syntax
> > Date: Wed, 28 May 2003 17:30:57 -0400
> > From: Ray Denenberg <[log in to unmask]>
> > > > dc.creator none "sanderson dovey levan"
> > >
> > > But we can't express that in Type-1.
> > And we can express 'all' and 'any' in type-1?
> dc.title all "elements programming"
> is equivalent to
> @and @attr 1=4 elements @attr 1=4 programming
> > In any case, if you can say "any of these words" it follows that you
> > can express "none of these words" using boolean and-not.
> Not so. For
> dc.title none "elements programming"
> You'd want to use
> @andnot @andnot SOMETHING @attr 1=4 elements @attr 1=4 programming
> but we have no SOMETHING in the Type-1 query that allows us to
> generate a list of all the records in the database. Put another way,
> Type-1 has binary ANDNOT operator, but no unary NOT operator -- which
> is what you'd need to map Rob's proposed "none" relation to.
> _/|_ _______________________________________________________________
> /o ) \/ Mike Taylor <[log in to unmask]> http://www.miketaylor.org.uk
> )_v__/\ "Historically, Taunton is part of Minehead already" --
> Monty Python.
> Listen to my wife's new CD of kids' music, _Child's Play_, at