CGAL users discussion list

["Miguel Armando Riem de Oliveira" <>]

Hi Winnie

Bearing in mind what you said I will try to re implement my code always
checking for the TDS validity every time I make a change.

Thanks very much for the help

Miguel

-----Original Message-----
From: Winnie Hellmann 
[mailto:]
 On Behalf Of
Winnie Hellmann
Sent: Tuesday, March 20, 2012 3:34 PM
To: 

Subject: Re: [cgal-discuss] Re: Constraned Triangulation 2 problem with
remove vertice

> Do you think I should test the validity of the face
> fh->is_valid();
Yes, this is what I meant. We can see in the second image that the
triangulation is not valid, but this might be caused by operating on an
invalid face handle beforehand.

> Because I used the previous and at some point it crashes in an assertion.
Then this should be the point where you use an invalid(ated) face handle.

> Are they copy constructable? Can I put them in an STL container like 
> std::vector<Face_handle>
Yes, they are. But keep in mind that they are not updated in any way. So
if you store a face handle and the face changes, the vector contains an
invalid face handle afterwards.

> But does TDS change when I add/remove constraints?
Yes, it does if those constrained edges were not part of the
triangulation before. As soon as faces change (which is the case if you
add new edges to the triangulation) their handles are invalidated and
sometimes also the handles of their neighbors (that depends if the new
edge is totally inside one face).