CGAL users discussion list

["Sebastien Loriot (GeometryFactory)" <>]

Note that even if do_intersect is a predicates, this is not a predicate
from the kernel and the exact_predicates_inexact_constructions_kernel
is not sufficient.

The documentation is here:
http://www.cgal.org/Manual/latest/doc_html/cgal_manual/Boolean_set_operations_2_ref/Function_do_intersect.html#Cross_link_anchor_1015

There is nothing that allow you to convert a polygon of  kernel
to another.
Cartesian_converter only works for objects that are in the kernel
(that is points for example, see here:
http://www.cgal.org/Manual/latest/doc_html/cgal_manual/Kernel_23_ref/Class_Cartesian_converter.html
)

Anyway, it would not be efficient at all since you will have to convert
all the points each time you want to make an intersection test.
The best solution would be to make this predicate safe with an
an exact predicates only kernel.

I will have a look at it.

In the meantime, could you provide the example polygon that lead to the
assertion.

Sebastien.

On 07/20/2011 03:38 AM, Gillian Ng wrote:
> Oh, I didn't know it required an exact constructions kernel. It works
> perfectly now, thanks a lot!
>
> By the way, is there any way to convert kernel types easily? My code has
> been using exact_predicates_inexact_constructions, but when I switched it to
> using exact_predicates_exact_constructions, it gave a ridiculous amount of
> errors. Is this even normal? I ended up storing two Polygons, one with exact
> constructions and one with inexact constructions, pretty clunky...
>
> --
> View this message in context: 
> http://cgal-discuss.949826.n4.nabble.com/CGAL-polygon-intersection-predicate-precondition-exception-tp3677725p3679799.html
> Sent from the cgal-discuss mailing list archive at Nabble.com.