I wonder if we need a "complex types" context set.

I still need a bag structure modifier to music - but I'm not convinced
that a music context set is the appropriate place.

Eliot needs a box structure (and potential a polygon structure) for his
searches. These are applicable to any area searches, not just GILS
(perhaps mathematics).

Similarly by analogy bounding cube, and/or polyhedra might be required
for 3d spatial searching.

Matthew

> -----Original Message-----
> From: Z39.50 Next-Generation Initiative [mailto:[log in to unmask]] 
> On Behalf Of Rob Sanderson
> Sent: Thursday, June 23, 2005 1:08 PM
> To: [log in to unmask]
> Subject: Re: Geospatial "bounding box" search in CQL
> 
> On Thu, 2005-06-23 at 05:43 -0400, Eliot Christian wrote:
> > There is a need to express geospatial "bounding box"
> > searches using the Common Query Language (CQL) syntax.
> 
> > Also, "nwse" could be used for the "Coordinate String"
> > Structure Attribute. (This has the great advantage of
> > making obvious the order of lat/lon coordinates.)
> >
> > Using these recommendations and my example search,
> > I guess the CQL expression for a "bounding box"
> > geospatial search would be like this:
> >
> >   bounds within/partial/nwse "36.5 -106.7 25.8 -93.5"
> >
> > Is that correct?
> 
> You would need to define a context set to look after the nwse relation
> modifier and the bounds index, but otherwise that looks fine :)
> 
> Eg:
> 
> gils.bounds within/partial/gils.nwse "..."
> 
> 
> Rob
>