Indeed, thanks Simon for the investigation, Mapinfo products call those multipolygon but you seems totaly right, There are actually polygon with multiple inner shells. And after investigation, it seems that jts-io actually load it well. I've been testing the geometry in/out oracle and in/out jts, displaying the wkt result at each step and the geometry is correct at each step.
So thanks a lot guys, and Martin sorry for the worries, the problem must be somewhere else. Jonathan -----Original Message----- From: [email protected] [mailto:[email protected]] On Behalf Of Martin Davis Sent: February 11, 2009 4:35 PM To: JTS Topology Suite Development Subject: Re: [jts-devel] Does Oracle support a "MultiPolygon which containsa MultiPolygon". Good stuff, Simon. So Jonathan, what behaviour are you actually seeing? Is an exception being thrown, or does the output not match what you'd expect to see? Simon Greener wrote: > Martin et al, > > My first reaction to Jonathan's post and your question was to say that, no, > Oracle (as an OGC SFS and SQL/MM compliant product) does not support a > "MultiPolygon which > contains a MultiPolygon". I thought, though, that I should examine the actual > geometry in detail so I converted Jonathan's WKT geometry below back into an > Oracle SDO_GEOMETRY and examined its SDO_ELEM_INFO array. This is the result: > > SDO_ELEM_INFO(1,1003,1,1743,2003,1,2041,2003,1,2071,1003,1,2779,2003,1) > > This is: a simple polygon outer shell (1,1003,1) followed by 4 inner shells > (2003): > > 1743,2003,1, > 2041,2003,1, > 2071,1003,1, > 2779,2003,1 > > Note that there is only 1 outer shell. Even if there were many outer shells > with each having many inner shells, the result would still be a single > multipolygon. > > My test harness is after my salutation at the very end of this email. > > regards > Simon > >> Date: Wed, 11 Feb 2009 11:24:23 -0800 >> From: Martin Davis <[email protected]> >> Subject: Re: [jts-devel] jts-io mutipolygon >> To: JTS Topology Suite Development <[email protected]> >> Message-ID: <[email protected]> >> Content-Type: text/plain; charset=windows-1252; format=flowed >> >> Under the OGC SFS spec there is no such thing as a "MultiPolygon which >> contains a MultiPolygon". Does Oracle actually support this? >> >> If this is truly the case, I think you're seeing the workaround - JTS >> maps the geometry to something which fits in its model. (Although this >> situation was probably never designed for - I'm actually surprised that >> it works). >> >> Houde, Jonathan wrote: >> >>> Hi all, >>> >>> I’ve been using jts-io to read geometries from oracle database. >>> >>> I have a certain case where the geometry is a MultiPolygon which >>> contains MultiPolygon, >>> >>> It seems that jts-io is not excepting this and doesn’t load it correctly. >>> >>> It read the inner Multipolygon as single Polygon >>> >>> You can see the geometry at the corresponding wkt geometry at the >>> bottom of my mail. >>> >>> Any fix planned for that or possible workaround? >>> >>> Thanks, >>> >>> Jonathan >>> >>> SDO_UTIL.TO_WKTGEOMETRY(GEOLOC) MULTIPOLYGON (((-71.33959188 >>> 46.81182897, -71.33935608 46.81166004, -71.33922 46.81155897, >>> -71.33990904 46.811142, -71.34054516 46.81077003, -71.34058512 >>> > > -- Martin Davis Senior Technical Architect Refractions Research, Inc. (250) 383-3022 _______________________________________________ jts-devel mailing list [email protected] http://lists.refractions.net/mailman/listinfo/jts-devel
_______________________________________________ jts-devel mailing list [email protected] http://lists.refractions.net/mailman/listinfo/jts-devel
