I would completely agree that it is best to avoid heterogeneous collections if at all possible Unfortunately we have to parse and store a data feed. While we can currently do this using by parsing the WKT and then storing in our database, writing tests requires the ability to create the collections in Python. I will take a look at adding support in 1.2.
Thanks, Aron On Wed, Sep 16, 2009 at 3:38 AM, Sean Gillies <[email protected]> wrote: > On Sep 15, 2009, at 4:39 PM, Aron Bierbaum wrote: > >> We have recently come across a need to support heterogeneous geometry >> collections. It appears that Shapely supports handling these >> geometries, but does not allow the user to create them from Python. >> There is actually a note in the GeometryCollection class that says "# >> Not to be created, but to be produced as a product of operations." Is >> there a specific reason that this was left out, or do we just need to >> find time to implement the correct factory methods in Python? >> >> Thanks, >> Aron > > It's been my opinion that heterogeneous collections are something to > avoid, but I'm not opposed to supporting them in 1.2. > > -- > Sean > > _______________________________________________ > Community mailing list > [email protected] > http://lists.gispython.org/mailman/listinfo/community > _______________________________________________ Community mailing list [email protected] http://lists.gispython.org/mailman/listinfo/community
