CGAL users discussion list

["Laurent Rineau (GeometryFactory)" <>]

On Friday 26 March 2010 11:55:05 N2a wrote:
> Hello,
> I was making tests on the mesh 2D generation and I have 2 things I do not
> understand.
> I run a constrained triangulation on a non convex hull. The boundary is
> constrained.
> 
> Problem 1: I iterate over FINITE faces and when I print the vertex index
> for each one I get three faces (over more than 1000) with the infinite
> vertex 0.
[...]
> 
> Thank You in advance,
> Benjamin
> 
> Here is the a piece of code:
> Declaration are the ones of the mesh_global.cpp example.
> 
>   std::map<Vertex_handle,int> vertex_index;
>   int idx = 0;
>   int nn=cdt.number_of_vertices();
>   std::vector<Vertex_handle> TV(nn);
>   int ii=0;
>   for(CDT::Vertex_iterator vit=cdt.vertices_begin(); vit !=
> cdt.vertices_end(); ++vit)
>    {
>          vertex_index[vit]=idx++;
>               TV[ii++]=vit;
>    }

Here you give an index to all vertices of the triangulation (including the 
infinite one).

TV[0] is cdt.vertices_begin(). There is no reason that vertex is the infinite 
one (you are iterating of finite vertices: vertex_begin() is a deprecated 
name 
for finite_vertices_begin()). Then, it is normal that a finite face can be 
incident to that vertex numbered 0.

> Problem 2: Each triangle have a positive signed area. I want to
> isolate the mesh from the triangulated part. Do I miss something in the
> boundary orientation?

I do not really understand your question about the boundary orientation.

If "fit" is a handle (or iterator) to a facet, then "fit->is_in_domain()" is 
true if and only if the facet is in the mesh domain (I assume that what you 
call "the mesh" is what I call "the mesh domain").

-- 
Laurent Rineau, PhD
R&D Engineer at GeometryFactory           http://www.geometryfactory.com/
Release Manager of the CGAL Project       http://www.cgal.org/