CGAL users discussion list

Le Mar 17 juillet 2007 20:54, Thomas Zangl - Home a écrit :
> In my case, this must not be true. I iterate over a triangulation
> using a the facet iterator. If a facet fullfills a criterion I add it
> to my list of points and facets and later they get inserted into a
> Polyhedron.

Can you guarantee that you obtain a surface in the end?

If you can, you could program a flooding algorithm: start with a triangle
T of the surface you are looking for, and for each edge e of T, find among
the other triangles incident to e the one that also belongs to the
surface. Then iterate on the new edges of your patch.
[it may also be much less expensive than checking all the faces of your 3D
triangulation]

If you cannot, Polyhedron_3 is not an adequate data structure.
-- 
Camille