CGAL users discussion list

[Andreas Fabri <>]

Thomas Zangl - Home wrote:
>  
> Hi!
>
> Given a RegularTriangulation3 and a number of edges of this
> triangulation. For each of these given edges I try to compute its dual
> plane (facet). The plane is formed by the set of vertices which are the
> dual of each regular triangulation cell, the edge is part of.
>
> I use some code like this:
>  
> (vh1, vh2 are vertex handles of an edge, t = RegularTriangultion3)
>  
>   Rt_Cell_handle c;
>   int i, j;
>   assert(t.is_edge(vh1, vh2, c,i,j));

Note that c, i, and j only get assigned values when you are compiling
in debug mode. Otherwise the assert statement is not executed, and
the next call is just random, and so is what follows.
>   Rt_Cell_circulator cCirc = t.incident_cells(c, i, j);
>   Rt_Cell_circulator cCircEnd = cCirc;
>   PointContainer pDual;
>   do {
>     if (!(t.is_infinite(cCirc))) {
>       Point p = t.dual(cCirc);
>       pDual.push_back(p);
>     }
>   } while(cCirc++!=cCircEnd);
>
>   if (pDual.size() == 1 ) {
>     cout << "hasIntersectionWithPD: got only one point." << endl;
>     return false;
>   } else if (pDual.size() == 2) {
>     cout << "hasIntersectionWithPD: got only one segment." << endl;
>     return false;
>   } else if (pDual.size() == 0) {
>     cout << "hasIntersectionWithPD: Got no dual point? Argl....!" << endl;
>     return false;
>   }
>
> I then happens that an edge has 0, 1 or 2 cells around it which
> produce the same number of dual points.
>
> Can that be true? I wonder because if an edge of a triangulation has 0
> dual points, this means that the edge is only part of infinite
> cells?
>
> Best regards,