pprepair hangs about pprepair HOT 3 CLOSED

Oh, I should have mentioned, version was the 'master' branch, downloaded and compiled yesterday. GDAL and CGAL are both whatever is current on Ubuntu Xenial, I can retrieve that information if it makes a difference.

from pprepair.

After reading the conversations on several other issues here, I tried building from the 'new' branch.

Initially, it seems much better. The program runs impressively fast, and the irregular boundaries between the polygons are dissolved quite as I expected.

The one thing that becomes a surprise is that holes in the polygons are now filled. In the data set that I uploaded earlier, for instance, the polygon with ogc_fid=338 name='Sand Hill Unit' has had the two large enclaves on its southwestern border filled in.

I think I can work around this by adding extra polygons in upstream processing so that the holes will be occupied, and only then using pprepair to dissolve borders.

Is the filling of holes expected and unavoidable, or am I missing something?

from pprepair.

yes, indeed branch "new" is better sometimes, we might merge it with master soon, if we get back to working on this.

Holes will be filled, because that's the whole point of pprepair: to create a PP from a set of input. The branch "new" doesn't have yet a way to tell it that certain polygons shouldn't be filled (the master branch let,s you define a point); we might add that in the future. In the meantime, I suggest you use your hack. But

from pprepair.