1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101
use glam::f32::Vec3A;
///
/// Implements spherical interpolation along the great arc created by
/// the initial points. This returns a new point `p` percent of the way
/// along that arc.
///
/// Note: `a` and `b` should both be normalized for normalized results.
///
pub fn geometric_slerp(a: Vec3A, b: Vec3A, p: f32) -> Vec3A {
let angle = a.dot(b).acos();
let sin = angle.sin().recip();
a * (((1.0 - p) * angle).sin() * sin) + b * ((p * angle).sin() * sin)
}
///
/// This is an optimization for the `geometric_slerp` in the case where `p`
/// is `0.5` or 50%.
///
/// Note: `a` and `b` should both be normalized for normalized results.
///
pub fn geometric_slerp_half(a: Vec3A, b: Vec3A) -> Vec3A {
(a + b) * (2.0 * (1.0 + a.dot(b))).sqrt().recip()
}
///
/// This is an optimization for the case where multiple points require the
/// calculation of varying values of `p` for the same start and end points.
///
/// See the intended use in [`BaseShape::interpolate_multiple`](crate::BaseShape::interpolate_multiple).
///
/// Note: `a` and `b` should both be normalized for normalized results.
///
pub fn geometric_slerp_multiple(a: Vec3A, b: Vec3A, indices: &[u32], points: &mut [Vec3A]) {
let angle = a.dot(b).acos();
let sin = angle.sin().recip();
for (percent, index) in indices.iter().enumerate() {
let percent = (percent + 1) as f32 / (indices.len() + 1) as f32;
points[*index as usize] =
a * (((1.0 - percent) * angle).sin() * sin) + b * ((percent * angle).sin() * sin);
}
}
///
/// Performs normalized linear interpolation. This creates distortion when
/// compared with spherical interpolation along an arc, however this is most
/// likely faster, as though this avoids expensive sin and acos calculations.
///
pub fn normalized_lerp(a: Vec3A, b: Vec3A, p: f32) -> Vec3A {
((1.0 - p) * a + p * b).normalize()
}
///
/// This is an optimization of `normalized_lerp` which avoids a multiplication.
///
pub fn normalized_lerp_half(a: Vec3A, b: Vec3A) -> Vec3A {
(a + b).normalize()
}
///
/// This is provided as a plug in for people who need it, but this implements
/// essentially the same algorithm as `BaseShape` would without ever being
/// reimplemented.
///
pub fn normalized_lerp_multiple(a: Vec3A, b: Vec3A, indices: &[u32], points: &mut [Vec3A]) {
for (percent, index) in indices.iter().enumerate() {
let percent = (percent + 1) as f32 / (indices.len() + 1) as f32;
points[*index as usize] = ((1.0 - percent) * a + percent * b).normalize();
}
}
///
/// Simple linear interpolation. No weirdness here.
///
pub fn lerp(a: Vec3A, b: Vec3A, p: f32) -> Vec3A {
(1.0 - p) * a + p * b
}
///
/// Gives the average of the two points.
///
pub fn lerp_half(a: Vec3A, b: Vec3A) -> Vec3A {
(a + b) * 0.5
}
///
/// This is provided as a plug in for people who need it, but this implements
/// essentially the same algorithm as `BaseShape` would without ever being
/// reimplemented.
///
pub fn lerp_multiple(a: Vec3A, b: Vec3A, indices: &[u32], points: &mut [Vec3A]) {
for (percent, index) in indices.iter().enumerate() {
let percent = (percent + 1) as f32 / (indices.len() + 1) as f32;
points[*index as usize] = (1.0 - percent) * a + percent * b;
}
}