Implemented in trunk now. --> Full rebuild (make clean; make) of GDAL necessary due to addition of virtual methods.
Best regards, Even Le samedi 14 août 2010 20:50:18, Even Rouault a écrit : > Le samedi 14 août 2010 19:57:52, Frank Warmerdam a écrit : > > On Sat, Aug 14, 2010 at 11:18 AM, Even Rouault > > > > Even, > > > > I'm generally ok with this change. I would like to see it made explicit > > that the name returned by OGRLayer::GetName() must always match the name > > returned by the OGRFeatureDefn::GetName(). > > Yes 100% agreed : here's what is currently stated in the doc of the method > : "This returns the same result as GetLayerDefn()->GetName()" > > Is that clear enough ? > > > I am also wondering if there are any other aspects that should be handled > > this way. Many applications want to show a list of laye names, and the > > geometry type of the layer (ie. ogrinfo, QGIS, etc). Should we also be > > making the geometry type a virtual method fetch on the layer? > > Yes, that makes sense although this is an information that is sometimes a > bit harder to get than the name, however I can add easily it : > > OGRGeometryType OGRLayer::GetGeomType() > > which should return the same as GetLayerDefn()->GetGeomType() > > > I can see the change you propose being of benefit to essentially all the > > RDBMS drivers. > > Yes, a quick look at the MYSQL driver shows that it could benefit from that > too. > > > Best regards, > > _______________________________________________ > gdal-dev mailing list > [email protected] > http://lists.osgeo.org/mailman/listinfo/gdal-dev _______________________________________________ gdal-dev mailing list [email protected] http://lists.osgeo.org/mailman/listinfo/gdal-dev
