[postgis-devel] POINT EMPTY in database dumps (EWKB)

[Even Rouault]

Selon Paul Ramsey <pramsey at cleverelephant.ca>:

> Thought of two other ways to represent POINT EMPTY when asked for the EWKB:
>
> POINT(NaN NaN)

--> that one at least is "standardized", as being the solution taken by
GeoPackage.

> NULL

How would you distinguish a "real" NULL from a POINT EMPTY represented as a
NULL?

>
> Neither is ideal, but each at least matches the type of the original,
> so wouldn't cause the constraint failure we have here.
>
> On Tue, Mar 17, 2015 at 11:44 AM, Paul Ramsey <pramsey at cleverelephant.ca>
> wrote:
> > http://trac.osgeo.org/postgis/ticket/3031
> >
> > I didn’t anticipate this *particular* problem when hacking the POINT
> EMPTY solution (which is to emit MULTIPOINT EMPTY (really, a WKB multipoint
> with zero members)).
> >
> > The only solution I can see would be to add an arbitrary additional WKB
> type number for the extended WKB format and use that exclusively to handling
> POINT EMPTY. That would allow dump/restore to work. For anyone who is
> requesting extended WKB and consuming it though, it would be a breaking
> change as this (surprise!) new WKB type number shows up in the stream.
> >
> > Thoughts?
> >
> > P
> >
> >
> > --
> > http://postgis.net
> > http://cleverelephant.ca
> >
> >
> _______________________________________________
> postgis-devel mailing list
> postgis-devel at lists.osgeo.org
> http://lists.osgeo.org/cgi-bin/mailman/listinfo/postgis-devel


-- 
Spatialys - Geospatial professional services
http://www.spatialys.com