CGAL users discussion list

["Wayne C. Gramlich" <>]

Sebastien:

Thanks.  I never saw CGAL Nef3 convert an .off file into a skin before.
It looks like I am having a tessellation problem of some sort.  Maybe
I'll add a "closed" primitive to demo/Nef_3 as well to help me catch
these problems in the future.

Thanks,

-Wayne

On 07/04/2011 10:58 PM, Sebastien Loriot (GeometryFactory) wrote:
> Hello,
>
> The problem comes from the fact that notched is not closed,
> thus it is interpreted as a skin.
>
> Sebastien.
>
> #include <CGAL/Simple_cartesian.h>
> #include <CGAL/Polyhedron_3.h>
> #include <CGAL/IO/Polyhedron_iostream.h>
> #include <iostream>
> #include <fstream>
>
> typedef CGAL::Simple_cartesian<double> K;
> typedef CGAL::Polyhedron_3<K> Polyhedron;
>
> int main(int,char** argv)
> {
> std::ifstream in(argv[1]);
> Polyhedron p;
> in >> p;
>
> if (p.is_closed()) std::cout << "closed" << std::endl;
>
> }
>
>
>
> Wayne C. Gramlich wrote:
> > Greetings:
> >
> > I have been using the demo/nef_3_homogeneous program quite extensively
> > in my code and it has been *extremely* robust.
> >
> > Recently I have come up with a relatively simple case where it fails
> > for me. The two attached files -- "cube.off" and "notched.off"
> > trigger tha failure as follows:
> >
> > nef_3_homogeneous loadoff notched.off loadoff cube.off union vis
> >
> > and the result looks correct. However, when I change the operation
> > from "union" to either "inters" or "diff", the result behaves as if
> > "notched.off" is just a skin, rather than a filled volume. I have
> > used the excellent "meshlab" program to verify that all the triangle
> > normals are pointed in the right direction for "notched.off".
> >
> > I am running CGAL 3.8 on Ubuntu 10.4 with gcc 4.4.3 using the 32-bit
> > x86 hardware.
> >
> > Does anybody have insight into what the issue might be?
> >
> > Thank you,
> >
> > -Wayne