CGAL users discussion list

[GiorgioCGAL <>]

Hi,
I am trying to use CGAL in order to perform the difference of two
polyhedrons. I work with Polyhedron_3 with the kernel
CGAL::Simple_cartesian<CGAL::Gmpq> and I convert the polyhedrons to
Nef_polyhedron_3 for performing the difference operation.

#include <CGAL/Simple_cartesian.h>
#include <CGAL/Polyhedron_3.h>
#include <CGAL/Nef_polyhedron_3.h>
#include <CGAL/Gmpq.h>
#include <CGAL/Polyhedron_incremental_builder_3.h>
#include <CGAL/IO/Polyhedron_iostream.h>

typedef CGAL::Simple_cartesian<CGAL::Gmpq> Kernel;
typedef CGAL::Polyhedron_3<Kernel> Polyhedron;
typedef CGAL::Nef_polyhedron_3<Kernel> Nef_Polyhedron;
typedef Polyhedron::HalfedgeDS HalfedgeDS;

Polyhedron diff(Polyhedron & i_poly1, Polyhedron & i_poly2)
{
        try
        {
                if (i_poly1.is_closed())
                {
                        Nef_Polyhedron NP1(i_poly1);
                        Nef_Polyhedron NP2(i_poly2);
                        Nef_Polyhedron o_NP = NP1 - NP2;
                        Polyhedron o_poly;
                        o_NP.convert_to_Polyhedron(o_poly);
                        return o_poly;
                }
                else
                {
                        throw Utils::ProcessControl::AREException(
                                        "At least one geometry is not 
closed.", __FUNCTION__);
                }
        }
        catch (Utils::ProcessControl::AREException & exception)
        {
                throw exception;
        }
}

It works fine but when I try to perform the difference between A and B and B
is strictly inside A (generating an holed polyhedron) the orientation of the
facets of the contour of the hole is not as I would expect. 

According to the documentation 
“The convention is that the halfedges are oriented counterclockwise around
facets as seen from the outside of the polyhedron. An implication is that
the halfedges are oriented clockwise around the vertices. The notion of the
solid side of a facet as defined by the halfedge orientation extends to
polyhedral surfaces with border edges although they do not define a closed
object. If normal vectors are considered for the facets, normals point
outwards (following the right-hand rule).”      

I would expect that the normal of the facets on the hole of the resulting
polyhedron (following the right-hand rule) will point to the hole itself but
they point to the interior of the solid.
What I am missing? There is a more suitable kernel for working with holed
polyhedron and that is consistent with the right-hand rule?

Thank you in advance,
Giorgio




--
View this message in context: 
http://cgal-discuss.949826.n4.nabble.com/Polyhedron-3-and-Nef-polyhedron-3-difference-for-holed-polyhedrons-tp4659409.html
Sent from the cgal-discuss mailing list archive at Nabble.com.