On lundi 25 septembre 2017 22:24:21 CEST Markus Metz wrote: > On Mon, Sep 25, 2017 at 8:23 PM, Even Rouault <[email protected]> > > wrote: > > On lundi 25 septembre 2017 11:04:35 CEST Rich Shepard wrote: > > > On Mon, 25 Sep 2017, Markus Metz wrote: > > > > these multi-surfaces are in the layer with multi-polygons. OGR calls > > these > > > > > geometries multi-surfaces. Investigating a bit more, I found > > > > > > > > curve-polygons within these multi-surfaces. Curve-polygons are in > > turn not > > > > > ordinary polygons, but defined by curves, and v.in.ogr (most GIS > > software > > > > > in general) does not support curves. > > > > > > Markus, > > > > > > > > > > > > Thanks for exploring the data files. Please suggest how I could extract > > > > > > the useful polygons and attribute data. I know that agencies assume > > everyone > > > > buys ESRI's software and they're not concerned with accommodating those > > of > > > > us who do not. In the mid-1990s everyone used ESRI's export format > > (.e00) > > > > and it was easy to write input filters for grass back then. > > > > You could for example use ogr2ogr to convert the curve geometries to > > linear ones with > > > ogr2ogr -f sqlite out.db in.gdb -nlt CONVERT_TO_LINEAR > > Thanks a lot for this suggestion! How are these curves converted to linear > features, i.e. how does ogr2ogr decide on reasonable conversion parameters > such as distance between vertices?
It uses (actually OGR_G_ForceTo() which I mentionned and underlying functions) a 4 degree step for circular portions. This can be tuned by setting the OGR_ARC_STEPSIZE configuration option/env variable to a value in degree. -- Spatialys - Geospatial professional services http://www.spatialys.com
_______________________________________________ grass-user mailing list [email protected] https://lists.osgeo.org/mailman/listinfo/grass-user
