[postgis-devel] Typmod Again

Paragon Corporation lr at pcorp.us
Wed Jul 22 14:23:32 PDT 2009


Paul,
What about the fact that we allow

POINT(1 2 3 4) = POINTZM(1 2 3 4)  --> which technically I think is supposed
to be POINT ZM(1,2,3,4) (neither POINTZM or POINT ZM seem to be valid in our
current releases)

Or should we break that and not allow that anymore and make POINT ZM or
POINTZM the new standard.

Thanks,
Regina

-----Original Message-----
From: postgis-devel-bounces at postgis.refractions.net
[mailto:postgis-devel-bounces at postgis.refractions.net] On Behalf Of Paul
Ramsey
Sent: Wednesday, July 22, 2009 5:14 PM
To: PostGIS Development Discussion
Subject: [postgis-devel] Typmod Again

OK, working into the typmod implementation, I found this interesting quirk.
Before, I was talking about

geometry(srid, type, dims)

But, in fact, dims is implicit in type, and you probably want that.
IE, 'POINTZM' (as defined in Table 4 of SFSQL 1.2.0) is a valid type, and
implies a dimensionality of four. We want our types to include that
information so we can distinguish between POINTM and POINTZ, otherwise, we
could define geometry(4326, POINT, 3) and be able to accept both POINTM and
POINTZ.

So that brings us down to just

geometry(srid, type)

Anything I'm missing?

P
_______________________________________________
postgis-devel mailing list
postgis-devel at postgis.refractions.net
http://postgis.refractions.net/mailman/listinfo/postgis-devel






More information about the postgis-devel mailing list