Hi I am going to merge all the multipolygon fixes. Including, if appropriate, the sea polygon changes. In addition to the patches sent to the list I also have permission to use the attached patch from Olaf Kähler which he describes as follows:
I guess, mkgmap has a bug with "vanishing islands" in case of multipolygons. The major problem seems to be, that in the polygon splitting process (PolygonSplitterBase.java), the Java awt methods mess around with the inner polygons, that were previously merged into the outer polygons. I included a patch, that somewhat works, but still seems to have a few issues. I've also included sample .osm files to explain these issues:
I'd appreciate it if anyone who understands the polygon issues takes a look at the patch. Also if you have an test cases which don't work with the current trunk but might be fixed by the outstanding patches let me know. ..Steve
Of course I forgot to attach the patch and examples... trying again. ..Steve
Steve Ratcliffe wrote:
Hi
I am going to merge all the multipolygon fixes. Including, if appropriate, the sea polygon changes.
In addition to the patches sent to the list I also have permission to use the attached patch from Olaf Kähler which he describes as follows:
I guess, mkgmap has a bug with "vanishing islands" in case of multipolygons. The major problem seems to be, that in the polygon splitting process (PolygonSplitterBase.java), the Java awt methods mess around with the inner polygons, that were previously merged into the outer polygons. I included a patch, that somewhat works, but still seems to have a few issues. I've also included sample .osm files to explain these issues:
I'd appreciate it if anyone who understands the polygon issues takes a look at the patch.
Also if you have an test cases which don't work with the current trunk but might be fixed by the outstanding patches let me know.
..Steve _______________________________________________ mkgmap-dev mailing list mkgmap-dev@lists.mkgmap.org.uk http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev
I tested the three samples with build 1129 before and after applying the patch. Before patch -------------------------------- test1=lake with island level0=fail (Island missing) level1=okay level2=okay test2=lake with 2 islands level0=fail (Islands missing) level1=okay level2=okay test3=lake with island + small lake on island level0=fail (Island & small lake missing) level1=okay level2=okay After patch -------------------------------- test1 all okay test2 all okay test3 level0=fail (Island & small lake missing) level1=okay level2=okay So test 3 still fails, even after applying the patch, or did I do something wrong applying the patch? The patch failed to apply properly for me, and I had to fix some errors by hand, so I could have made a mistake. I will test my own samples next. Thanks for the updates, Garvan
Steve Ratcliffe wrote:
Hi
So test 3 still fails, even after applying the patch, or did I do something wrong applying the patch?
No test 3 still failing was a known problem, there are still cases where it was known not to work.
..Steve
I had disappointing results when I tried this patch on a larger sample with real data. It works sometimes. I uploaded a sample file here: http://freefile.kristopherw.us/uploads/sisophon2001/sea.osm.gz Garvan
Dear Steve, I just tried the patch. Apart from test 3 not working, the modified PolygonSplitterBase seems not to improve the "flooded island problem". Best wishes Christian Steve Ratcliffe schrieb:
Hi
So test 3 still fails, even after applying the patch, or did I do something wrong applying the patch?
No test 3 still failing was a known problem, there are still cases where it was known not to work.
..Steve _______________________________________________ mkgmap-dev mailing list mkgmap-dev@lists.mkgmap.org.uk http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev
participants (3)
-
Christian Gawron -
Garvan & maew -
Steve Ratcliffe