[geos-devel] MULTI* and Invalid

[Sandro Santilli]

On Fri, Jun 14, 2013 at 01:16:02PM +0200, Oliver Courtin wrote:
> Hi,
> 
> 
> Below an example, throught PostGIS 
> with GEOS 3.3.5-CAPI-1.7.5:
> 
> 
> SELECT ST_IsValidReason(
>   'MULTIPOLYGON(((0 0,1 0,1 1,0 1,0 0)),((1 0,2 0,2 1,1 1,1 0)))'::geometry
>   );
> 
> 
> Output:   Self-intersection[1 0]
> 
> 
> After looking in SFS and ISO 19107, 
> we didn't find at first sight 
> why such a geometry should be invalid.
> 
> 
> Could you give us some light on it ?

The two components have a lineal intersection.
Puntal intersection would be considered valid
(ie: moving vertex  1,1  to 1,1.5).

IIRC there was a "single point" wording in the spec, but
Martin will know more about this.

--strk;