CGAL users discussion list

["Sebastien Loriot (GeometryFactory)" <>]

I confirm this is a bug. I've added it to our tracker and let you know
as soon as we have fixed it.

Thanks,

Sebastien.

On 02/18/2015 10:41 AM, Marcel wrote:
> Dear CGAL-Team,
>
> I am working with your Straight Skeleton implementation. I encountered
> a problem with the following polygon:
>
> polygon[(320, 1120),(-141.55499021951055738,
> 1658.4808219227620611),(-1115.9820207544369168,
> 1835.6551586833445526),(760, 520),(800, 560),(840, 600),(800,
> 640),(400, 1040)]
>
> I try to generate the straight skeleton with the following call:
>
> typedef CGAL::Cartesian<CGAL::Gmpq> Kernel;
> Kernel c;
> CGAL::create_interior_straight_skeleton_2<CGAL::Polygon_2<Kernel>>(polygon,
> c);
>
> The algorithm aborts after a few seconds with the following assertion
> violated.
>
>   /usr/include/boost/smart_ptr/shared_ptr.hpp:646: typename
> boost::detail::sp_dereference<T>::type
> boost::shared_ptr<T>::operator*() const [with T =
> CGAL::Straight_skeleton_2<CGAL::Cartesian<CGAL::Gmpq>,
> CGAL::Straight_skeleton_items_2, std::allocator<int> >; typename
> boost::detail::sp_dereference<T>::type =
> CGAL::Straight_skeleton_2<CGAL::Cartesian<CGAL::Gmpq>,
> CGAL::Straight_skeleton_items_2, std::allocator<int> >&]: Assertion
> `px != 0' failed.
>
> The subsequent collinear points seems to be troubling. Removing the
> collinear points yields to a successful computation. As far as I see,
> collinear points are not considered a degenerated case
> (http://doc.cgal.org/latest/Straight_skeleton_2/index.html, 1.1 2D
> Contours). So, I suppose it is bug?
>
> Kind regards
> Marcel
>
> System Details:
>
> OpenSuse 13.2 (Harlequin) x86_64
> Intel (R) Xeon (R) CPU E5430 @ 2.66Ghz
> GCC 4.8.3, flags: -frounding-math -std=c++11 -O0 -g -gdwarf-3 -W -Wall
> -Wextra -pedantic -fopenmp
> CGAL 4.5.2 - compiled without any additional flags