[geos-devel] Proposals for ensuring correct semantics?

[Martin Davis]

Sandro's (sp?) buffer question points up a potential issue with ensuring that both JTS and GEOS continue to have identical semantics.  Any ideas for how to ensure this?

We do have the JTS Test Suite, but I'm not sure how well it's implemented in GEOS.  Also, it's not complete for spatial functions, and it certainly doesn't address non-Geometry methods.

We do have a set of JUnit tests for JTS; it might be worthwhile porting them to C++ and putting a framework in place to validate new builds of GEOS.

One difficulty with validating output of constructive functions is that the output may be topologically correct but not quite identical, due to vagaries of implementation and floating point.  I don't have a great solution for this right now - it would be worth discussing.

Martin

Martin Davis, Senior Technical Architect
Vivid Solutions Inc.
Suite #1A-2328 Government Street   Victoria, B.C.   V8T 5G5
Phone: (250) 385 6040 - Local 308      Fax: (250) 385 6046