On Thu, Mar 15, 2012 at 05:59:00PM +0100, Sandro Santilli wrote: > On Thu, Mar 15, 2012 at 12:56:38PM -0400, Pierre Racine wrote: > > > 1. find out intersecting pairs into cache (id1,id2,geom1,geom2) 2. > > > update cache > > > set geom1 = st_difference(geom1, geom2) 3. update main set geom = c.geom1 > > > from cache where id1 = id > > > > Hum... Life is never that easy. > > > > That would work if one polygon would always overlap with only one polygon > > but if one polygon is overlapped by two polygons or more the same polygon > > as two differences associated with it in the second step and only the last > > one will remains in the last update... > > True, should be an iterative process until "cache" becomes empty.
I would add that every iteration (ST_Difference call) would introduce small changes in your boundaries possibly introducing gaps or new overlaps. You should really be constructing a topology out of that dataset, you know ? Will be probably slow but I'm sure you can survive :) --strk; ,------o-. | __/ | Delivering high quality PostGIS 2.0 ! | / 2.0 | http://strk.keybit.net - http://vizzuality.com `-o------' _______________________________________________ postgis-users mailing list postgis-users@postgis.refractions.net http://postgis.refractions.net/mailman/listinfo/postgis-users
