Ray Casting

Retrieve information about where a ray, defined by an origin and a direction, hits a target. Generally, the actual hit point can be calculated using the fraction value returned by the appropriate functions using the formula:

$$\textbf{hit} = \textbf{origin} + \textrm{fraction} \times \textbf{direction}, \quad \textrm{with fraction} \in [0,\infty).$$

The direction given to Jitter does not need to be normalized. The normal returned by Jitter is the normalized surface normal of the collision target at the hit point.

Generic Ray Cast for ISupportMap

To cast a ray against any object implementing ISupportMap, use NarrowPhase.RayCast:

public static bool RayCast(ISupportMap support, 
                           ref JMatrix orientation, ref JVector position,
                           ref JVector origin, ref JVector direction, 
                           out float fraction, out JVector normal)

Ray Cast Against Shapes

The world class offers two ray cast methods:

public bool RayCast(Shape shape, JVector origin, JVector direction,
                    out JVector normal, out float fraction)

public bool RayCast(JVector origin, JVector direction, 
                    RayCastFilterPre? pre, RayCastFilterPost? post,
                    out Shape? shape, out JVector normal, out float fraction)

The first method can be used to cast a ray against a single shape. The second method tests a ray against the world. Use the ray cast filter to get information about possible candidates or to avoid collision with specific shapes entirely. For example, a bullet may always be able to penetrate a wall; the pre-filter can be used to discard any collision with the wall. For a wall with non-uniform thickness, the bullet may only be able to penetrate the thinner part. In this case, the post-filter can be used to get information on where the ray hit the wall and, based on that information, decide whether to drop the collision.