2019-04-21 17:54:24 +02:00
|
|
|
|
#[cfg(feature = "std")] use std::f32;
|
|
|
|
|
#[cfg(not(feature = "std"))] use core::f32;
|
|
|
|
|
#[cfg(not(feature = "std"))] use core::intrinsics::cosf32;
|
|
|
|
|
#[cfg(feature = "std")] use std::f64;
|
|
|
|
|
#[cfg(not(feature = "std"))] use core::f64;
|
|
|
|
|
#[cfg(not(feature = "std"))] use core::intrinsics::cosf64;
|
|
|
|
|
#[cfg(feature = "std")] use std::ops::{Add, Mul, Sub};
|
|
|
|
|
#[cfg(not(feature = "std"))] use core::ops::{Add, Mul, Sub};
|
2019-04-19 13:39:37 +02:00
|
|
|
|
|
|
|
|
|
/// Keys that can be interpolated in between. Implementing this trait is required to perform
|
|
|
|
|
/// sampling on splines.
|
|
|
|
|
///
|
|
|
|
|
/// `T` is the variable used to sample with. Typical implementations use `f32` or `f64`, but you’re
|
|
|
|
|
/// free to use the ones you like.
|
|
|
|
|
pub trait Interpolate<T>: Sized + Copy {
|
|
|
|
|
/// Linear interpolation.
|
|
|
|
|
fn lerp(a: Self, b: Self, t: T) -> Self;
|
|
|
|
|
|
|
|
|
|
/// Cubic hermite interpolation.
|
|
|
|
|
///
|
|
|
|
|
/// Default to `Self::lerp`.
|
|
|
|
|
fn cubic_hermite(_: (Self, T), a: (Self, T), b: (Self, T), _: (Self, T), t: T) -> Self {
|
|
|
|
|
Self::lerp(a.0, b.0, t)
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
|
2019-04-21 17:54:24 +02:00
|
|
|
|
/// A trait for anything that supports additions, subtraction, multiplication and division.
|
|
|
|
|
pub trait Additive:
|
|
|
|
|
Copy +
|
|
|
|
|
Add<Self, Output = Self> +
|
|
|
|
|
Sub<Self, Output = Self> {
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
impl<T> Additive for T
|
|
|
|
|
where T: Copy +
|
|
|
|
|
Add<Self, Output = Self> +
|
|
|
|
|
Sub<Self, Output = Self> {
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/// Linear combination.
|
|
|
|
|
pub trait Linear<T> {
|
|
|
|
|
/// Apply an outer multiplication law.
|
|
|
|
|
fn outer_mul(self, t: T) -> Self;
|
|
|
|
|
|
|
|
|
|
/// Apply an outer division law.
|
|
|
|
|
fn outer_div(self, t: T) -> Self;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
macro_rules! impl_linear_simple {
|
|
|
|
|
($t:ty) => {
|
|
|
|
|
impl Linear<$t> for $t {
|
|
|
|
|
fn outer_mul(self, t: $t) -> Self {
|
|
|
|
|
self * t
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/// Apply an outer division law.
|
|
|
|
|
fn outer_div(self, t: $t) -> Self {
|
|
|
|
|
self / t
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
impl_linear_simple!(f32);
|
|
|
|
|
impl_linear_simple!(f64);
|
|
|
|
|
|
|
|
|
|
macro_rules! impl_linear_cast {
|
|
|
|
|
($t:ty, $q:ty) => {
|
|
|
|
|
impl Linear<$t> for $q {
|
|
|
|
|
fn outer_mul(self, t: $t) -> Self {
|
|
|
|
|
self * t as $q
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/// Apply an outer division law.
|
|
|
|
|
fn outer_div(self, t: $t) -> Self {
|
|
|
|
|
self / t as $q
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
impl_linear_cast!(f32, f64);
|
|
|
|
|
impl_linear_cast!(f64, f32);
|
|
|
|
|
|
|
|
|
|
/// Types with a neutral element for multiplication.
|
|
|
|
|
pub trait One {
|
|
|
|
|
/// Return the neutral element for the multiplicative monoid.
|
|
|
|
|
fn one() -> Self;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
macro_rules! impl_one_float {
|
|
|
|
|
($t:ty) => {
|
|
|
|
|
impl One for $t {
|
|
|
|
|
#[inline(always)]
|
|
|
|
|
fn one() -> Self {
|
|
|
|
|
1.
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
impl_one_float!(f32);
|
|
|
|
|
impl_one_float!(f64);
|
|
|
|
|
|
|
|
|
|
/// Types with a sane definition of π and cosine.
|
|
|
|
|
pub trait Trigo {
|
|
|
|
|
/// π.
|
|
|
|
|
fn pi() -> Self;
|
|
|
|
|
|
|
|
|
|
/// Cosine of the argument.
|
|
|
|
|
fn cos(self) -> Self;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
impl Trigo for f32 {
|
|
|
|
|
#[inline(always)]
|
|
|
|
|
fn pi() -> Self {
|
|
|
|
|
f32::consts::PI
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
#[inline(always)]
|
|
|
|
|
fn cos(self) -> Self {
|
|
|
|
|
#[cfg(feature = "std")]
|
|
|
|
|
{
|
|
|
|
|
self.cos()
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
#[cfg(not(feature = "std"))]
|
|
|
|
|
{
|
|
|
|
|
unsafe { cosf32(self) }
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
impl Trigo for f64 {
|
|
|
|
|
#[inline(always)]
|
|
|
|
|
fn pi() -> Self {
|
|
|
|
|
f64::consts::PI
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
#[inline(always)]
|
|
|
|
|
fn cos(self) -> Self {
|
|
|
|
|
#[cfg(feature = "std")]
|
|
|
|
|
{
|
|
|
|
|
self.cos()
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
#[cfg(not(feature = "std"))]
|
|
|
|
|
{
|
|
|
|
|
unsafe { cosf64(self) }
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
|
2019-04-19 13:39:37 +02:00
|
|
|
|
// Default implementation of Interpolate::cubic_hermite.
|
|
|
|
|
//
|
|
|
|
|
// `V` is the value being interpolated. `T` is the sampling value (also sometimes called time).
|
|
|
|
|
pub(crate) fn cubic_hermite_def<V, T>(x: (V, T), a: (V, T), b: (V, T), y: (V, T), t: T) -> V
|
2019-04-21 17:54:24 +02:00
|
|
|
|
where V: Additive + Linear<T>,
|
|
|
|
|
T: Additive + Mul<T, Output = T> + One {
|
2019-04-19 13:39:37 +02:00
|
|
|
|
// some stupid generic constants, because Rust doesn’t have polymorphic literals…
|
2019-04-21 17:54:24 +02:00
|
|
|
|
let one_t = T::one();
|
|
|
|
|
let two_t = one_t + one_t; // lolololol
|
|
|
|
|
let three_t = two_t + one_t; // megalol
|
2019-04-19 13:39:37 +02:00
|
|
|
|
|
|
|
|
|
// sampler stuff
|
|
|
|
|
let t2 = t * t;
|
|
|
|
|
let t3 = t2 * t;
|
|
|
|
|
let two_t3 = t3 * two_t;
|
|
|
|
|
let three_t2 = t2 * three_t;
|
|
|
|
|
|
|
|
|
|
// tangents
|
2019-04-21 17:54:24 +02:00
|
|
|
|
let m0 = (b.0 - x.0).outer_div(b.1 - x.1);
|
|
|
|
|
let m1 = (y.0 - a.0).outer_div(y.1 - a.1);
|
2019-04-19 13:39:37 +02:00
|
|
|
|
|
2019-04-21 17:54:24 +02:00
|
|
|
|
a.0.outer_mul(two_t3 - three_t2 + one_t) + m0.outer_mul(t3 - t2 * two_t + t) + b.0.outer_mul(three_t2 - two_t3) + m1.outer_mul(t3 - t2)
|
2019-04-19 13:39:37 +02:00
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
macro_rules! impl_interpolate_simple {
|
|
|
|
|
($t:ty) => {
|
|
|
|
|
impl Interpolate<$t> for $t {
|
|
|
|
|
fn lerp(a: Self, b: Self, t: $t) -> Self {
|
|
|
|
|
a * (1. - t) + b * t
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
fn cubic_hermite(x: (Self, $t), a: (Self, $t), b: (Self, $t), y: (Self, $t), t: $t) -> Self {
|
|
|
|
|
cubic_hermite_def(x, a, b, y, t)
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
impl_interpolate_simple!(f32);
|
|
|
|
|
impl_interpolate_simple!(f64);
|
|
|
|
|
|
|
|
|
|
macro_rules! impl_interpolate_via {
|
|
|
|
|
($t:ty, $v:ty) => {
|
|
|
|
|
impl Interpolate<$t> for $v {
|
|
|
|
|
fn lerp(a: Self, b: Self, t: $t) -> Self {
|
|
|
|
|
a * (1. - t as $v) + b * t as $v
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
fn cubic_hermite((x, xt): (Self, $t), (a, at): (Self, $t), (b, bt): (Self, $t), (y, yt): (Self, $t), t: $t) -> Self {
|
|
|
|
|
cubic_hermite_def((x, xt as $v), (a, at as $v), (b, bt as $v), (y, yt as $v), t as $v)
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
impl_interpolate_via!(f32, f64);
|
|
|
|
|
impl_interpolate_via!(f64, f32);
|