In short, I have to generate a constrained delaunay triangulation from a set of 2D points that carry additional information.
Since there could be intersecting constraints, I need that the triangulation properly interpolates the satellite data whenever it generates the new point of intersection.
I tried with CGAL (C++) to define a custom kernel following the example here, but I've failed to compile it succesfully, basically because there is something fundamental of the underlying generic programming machinery of CGAL that I am missing (lot of type mismatches etc.).
Can you point me to a working example that extends points with satellite data that is properly interpolated at constrain intersections (or at least something similar) ?
In theory, I could use a vertex with info, but then I'd need some callback mechanism that let me handle the intersection (I actually did this in a dirty way by modifying cgal source code). Anyway, what I am looking for is a "cgal-way" for achieving what I just discussed.
