Add Sutherland–Hodgman VPolygon intersection algorithm

[mforets]

This is a standard algorithm, see https://en.wikipedia.org/wiki/Sutherland%E2%80%93Hodgman_algorithm. The code follows that of http://www.rosettacode.org/wiki/Sutherland-Hodgman_polygon_clipping#Julia.

Related to #698.

[mforets]

if the intersection is empty the new function should rather return EmptySet -- for more advanced users, type stability is ensured using _intersection_vrep directly.

[schillic]

IMO VPolygons should always be created using a convex hull unless the algorithm guarantees that result anyway.

[mforets]

yes , i agree. i left the option as true by default.

[schillic]

I still think the option apply_convex_hull is misleading because you may want to set it to false to save time, but that will violate our type assumption of VPolygon.

[mforets]

I still think the option apply_convex_hull is misleading because you may want to set it to false to save time, but that will violate our type assumption of VPolygon.

yes, but if an application just requires the vertices of the intersection (without any particular sorting assumption) it's reasonable to pass apply_convex_hull=false, instead of digging & using internal methods to get the vertices directly.