CGAL users discussion list

["Sebastien Loriot (GeometryFactory)" <>]

Sorry, I was working thinking to HDS which has different level check.

Actually the level 1 in polyhedron checks the boundary so if you didn't
call normalize_border you will have an error.

You first input is a valid polyhedron, but it cannot be directly
converted into a nef polyhedron because each of the two larger facets
does not have its points exactly on a plane. Thus nef is not able to
consider them as a single face.

A solution is to triangulate the polyhedron before feeding nef with it.

About the dangling facet, it was a bug in my viewer.

Sebastien.

On 03/28/2013 06:47 PM, Cody Rose wrote:
> Thank you for your reply. I had not through to use higher levels of
> checking for is_valid. However, I guess I still don't understand what
> those levels do exactly, because when I load the attached simple
> rectangular prism into a Polyhedron_3 and call is_valid(true, 1) then
> the validation still fails, and I cannot see how it could possibly do that.
>
> I discovered this because I still cannot figure out what's actually
> wrong with my original polyhedron - I cannot identify the geometry of
> the dangling face you describe. One OFF viewer I've tried renders two
> dangling faces, but another does not, and when I manually inspect the
> coordinates in the .off they looks fine. I suspect the rendering problem
> has to do with the fact that two of my faces are concave, but I can't
> imagine that CGAL is having a similar issue.
>
> (Incidentally, loading from the .off directly into a Nef_polyhedron
> doesn't help me, because in my actual use case these polyhedrons are
> generated in code, not read from a file.)
>
> Thank you,
> Cody
>
> On 3/28/2013 12:41 AM, Sebastien Loriot (GeometryFactory) wrote:
> > Hello,
> >
> > the polygon is actually invalid
> > (increase the check level to 3
> > if (!poly.is_valid(false,3)) {
> > std::cout << "poly is invalid\n";
> > }
> > ).
> >
> > You have a dangling face with make the surface non-manifold.
> >
> > One solution is to use this:
> >
> > if (!poly.is_valid(false,3)) {
> > std::cout << "poly is invalid\n";
> > stream.open("poly.off");
> > nef_polyhedron nef;
> > CGAL::OFF_to_nef_3 (stream, nef);
> > std::cout << "created nef\n";
> > }
> >
> >
> > Sebastien.
> >
> >
> > On 03/27/2013 09:28 PM, Cody Rose wrote:
> > > Hello,
> > >
> > > I receive an assertion failure when trying to convert the attached
> > > polyhedron to a nef polyhedron. The attached sample code demonstrates my
> > > problem. I am running this with CGAL 4.1, Boost 1.47, and LEDA 6.3. I am
> > > not sure if there is something about this polyhedron that causes the
> > > problem, but I could not find anything in the documentation to help me
> > > resolve this issue, and a cursory inspection of the .off doesn't show
> > > anything wrong to me (I'm probably just missing something simple).
> > >
> > > The example program out is as follows:
> > >
> > > loaded good poly with 16 vertices
> > > CGAL error: assertion violation!
> > > Expression : pe_prev->is_border() ||
> > > internal::Plane_constructor<Plane>::get_plane(pe_prev->facet(),pe_prev->facet()->plane()).
> > >
> > > has_on(pe_prev->opposite()->vertex()->point())
> > > File : c:\program files
> > > (x86)\cgal\include\cgal\nef_3\polyhedron_3_to_nef_3.h
> > > Line : 254
> > > Explanation:
> > > Refer to the bug-reporting instructions at
> > > http://www.cgal.org/bug_report.html
> > >
> > > Thank you,
> > > Cody Rose