TransformBlend

class TransformBlend

This defines a single entry in a TransformBlendTable. It represents a unique combination of VertexTransform pointers and blend amounts.

Inheritance diagram

TransformBlend(VertexTransform const *transform0, PN_stdfloat weight0, VertexTransform const *transform1, PN_stdfloat weight1)

TransformBlend(VertexTransform const *transform0, PN_stdfloat weight0, VertexTransform const *transform1, PN_stdfloat weight1, VertexTransform const *transform2, PN_stdfloat weight2)

TransformBlend(VertexTransform const *transform0, PN_stdfloat weight0, VertexTransform const *transform1, PN_stdfloat weight1, VertexTransform const *transform2, PN_stdfloat weight2, VertexTransform const *transform3, PN_stdfloat weight3)

void add_transform(VertexTransform const *transform, PN_stdfloat weight): Adds a new transform to the blend. If the transform already existed, increases its weight factor.

int compare_to(TransformBlend const &other) const: Defines an arbitrary ordering for TransformBlend objects.

void get_blend(LMatrix4 &result, Thread *current_thread) const

Returns the current value of the blend, based on the current value of all of the nested transform objects and their associated weights.

You should call update_blend() to ensure that the cache is up-to-date before calling this.

UpdateSeq get_modified(Thread *current_thread = Thread::get_current_thread()) const: Returns a counter which is guaranteed to increment at least as often as the result of get_blend() changes.

std::size_t get_num_transforms(void) const: Returns the number of transforms stored in the blend object.

VertexTransform const *get_transform(std::size_t n) const: Returns the nth transform stored in the blend object.

PN_stdfloat get_weight(std::size_t n) const

Returns the weight associated with the nth transform stored in the blend object.

Returns the weight associated with the indicated transform, or 0 if there is no entry for the transform.

bool has_transform(VertexTransform const *transform) const: Returns true if the blend has the indicated transform, false otherwise.

void limit_transforms(int max_transforms): If the total number of transforms in the blend exceeds max_transforms, removes the n least-important transforms as needed to reduce the number of transforms to max_transforms.

void normalize_weights(void): Rescales all of the weights on the various transforms so that they sum to 1.0. It is generally a good idea to call this after adding or removing transforms from the blend.

void remove_transform(std::size_t n)

Removes the nth transform stored in the blend object.

Removes the indicated transform from the blend.

void set_transform(std::size_t n, VertexTransform const *transform): Replaces the nth transform stored in the blend object.

void set_weight(std::size_t n, PN_stdfloat weight): Replaces the weight associated with the nth transform stored in the blend object.

void transform_point(LPoint3d &point, Thread *current_thread) const

Transforms the indicated point by the blend matrix.

You should call update_blend() to ensure that the cache is up-to-date before calling this.

void transform_vector(LVector3d &point, Thread *current_thread) const

Transforms the indicated vector by the blend matrix.

You should call update_blend() to ensure that the cache is up-to-date before calling this.

void update_blend(Thread *current_thread) const: Recomputes the internal representation of the blend value, if necessary. You should call this before calling get_blend() or transform_point().