Answering my own question, g can be LineString and GeometryCollection, I haven't tested MultiLineString. The returned objects are MultiLineString.
Testing with the original problem (a twisted bowtie polygon coerced to LineString), with a 2D rectangular random walk with 999 segments, and 100 random segments shows no issues, with the noded output feeding through to polygonizing without obvious failures. Roger Roger Bivand wrote > I'd like to ask for pointers to example code using GEOSNode/GEOSNode_r in > CAPI. What should the geometry g be - GEOS_MULTILINESTRING, > GEOS_LINESTRING, or what? Looking at the C++ code, it appears to be a > SegmentString. If so, can a GEOS_MULTILINESTRING be converted to such an > object in the CAPI? > > I'd like to add noding to rgeos, a user has occasional incoming polygon > geometries that are not valid, and whose creation he does not control, and > needs to node them after coercion to line strings, before going back out > through polygonize. > > Grateful for any ideas, > > Roger > > -- > Roger Bivand > Department of Economics, Norwegian School of Economics, > Helleveien 30, N-5045 Bergen, Norway. > voice: +47 55 95 93 55; fax +47 55 95 91 00 > e-mail: > Roger.Bivand@ > > _______________________________________________ > geos-devel mailing list > geos-devel@.osgeo > http://lists.osgeo.org/mailman/listinfo/geos-devel ----- Roger Bivand NHH Norwegian School of Economics, Bergen, Norway -- View this message in context: http://osgeo-org.1560.x6.nabble.com/GEOSNode-usage-what-should-g-be-tp5130497p5130512.html Sent from the GEOS Developers mailing list archive at Nabble.com. _______________________________________________ geos-devel mailing list geos-devel@lists.osgeo.org http://lists.osgeo.org/mailman/listinfo/geos-devel