CGAL users discussion list

[Sebastien Loriot <>]

On this page, there are 3 examples of how to deal with output of 
intersection function:
https://doc.cgal.org/latest/Kernel_23/group__intersection__linear__grp.html

Maybe the second one (apply_visitor) will better fit your need
(using a template operator()).

Best,

Sebastien.

On 2/21/22 00:58, Bob Bill ( via cgal-discuss Mailing 
List) wrote:
> Thanks for your prompt reply, Marc!
>
> Okay, I got what the problem is. However, I don't see how to dispatch 
> "automatically". I mean, how can I get the actual type of ´first´ in 
> order to pass it as an argument of the proper type to CGAL::intersection() ?
>
> I cannot do something like:
>
> if (const Segment *s = boost::get<Segment>(&first)) {
> Segment first_s = *s;
> } else if (const Rectangle *p = boost::get<Rectangle>(&first)) {
> Rectangle first_r = *p;
> } else if (const Triangle *t = boost::get<Triangle>(&first)) {
> Triangle first_t = *t;
> }
>
> because in this way I will not be able to pass the original `first`. 
> This is the first time I'm using boost libraries, so I'm really sorry if 
> that's a trivial question.
>
> Best,
> Bob
> Il domenica 20 febbraio 2022, 21:14:43 CET, Marc Glisse 
> <> ha scritto:
>
>
> On Sun, 20 Feb 2022, Bob Bill  wrote:
>
>
>  > I am facing the following problem: I have two 2D objects. Each one 
> could be a triangle, a rectangle or a segment, and I need to compute 
> their intersection using CGAL::intersection().
>  > So I created a function `create_suitable_object` that takes the 
> vector with the vertices and create the corresponding CGAL object. As 
> the return type is not unique, I used `boost::variant`.
>  > Then in my main I created the two objects calling the above mentioned 
> function for each vector of vertices, and then I wanted to pass those 
> objects to CGAL::intersection().
>
>
> Where in the documentation did you see that it was ok to pass a variant to
> CGAL::intersection? As far as I can see, it only lists overloads like
> intersection(segment, triangle), etc, so you need to handle the
> dispatching yourself.
>
> --
> Marc Glisse
>
> --
> You are currently subscribed to cgal-discuss.
> To unsubscribe or access the archives, go to
> https://sympa.inria.fr/sympa/info/cgal-discuss 
> <https://sympa.inria.fr/sympa/info/cgal-discuss>
>
>
>
>
> --
> You are currently subscribed to cgal-discuss.
> To unsubscribe or access the archives, go to
> https://sympa.inria.fr/sympa/info/cgal-discuss