BezierEasing provides Cubic Bezier Curve easing which generalizes easing functions (ease-in, ease-out, ease-in-out, ...any other custom curve) exactly like in CSS Transitions.
Implementing efficient lookup is not easy because it implies projecting the X coordinate to a Bezier Curve. This micro library uses fast heuristics (involving dichotomic search, newton-raphson, sampling) to focus on performance and precision.
It is heavily based on implementations available in Firefox and Chrome (for the CSS transition-timing-function property).
var easing = BezierEasing(0, 0, 1, 0.5);
// easing allows to project x in [0.0,1.0] range onto the bezier-curve defined by the 4 points (see schema below).
console.log(easing.get(0.0)); // 0.0
console.log(easing.get(0.5)); // 0.3125
console.log(easing.get(1.0)); // 1.0
(this schema is from the CSS spec)
It is the equivalent to CSS Transitions' transition-timing-function
.
In the same way you can define in CSS cubic-bezier(0.42, 0, 0.58, 1)
,
with BezierEasing, you can define it using BezierEasing(0.42, 0, 0.58, 1)
which have the .get
function taking an X and computing the Y interpolated easing value (see schema).
bezier-easing also define a mapping from existing CSS transition-timing-function
:
BezierEasing.css = {
"ease": BezierEasing.ease = BezierEasing(0.25, 0.1, 0.25, 1.0),
"linear": BezierEasing.linear = BezierEasing(0.00, 0.0, 1.00, 1.0),
"ease-in": BezierEasing.easeIn = BezierEasing(0.42, 0.0, 1.00, 1.0),
"ease-out": BezierEasing.easeOut = BezierEasing(0.00, 0.0, 0.58, 1.0),
"ease-in-out": BezierEasing.easeInOut = BezierEasing(0.42, 0.0, 0.58, 1.0)
};
There is also a toCSS()
method that returns the transition-timing-function value string (so the library is agnostic).
MIT License.
npm test
- Apple® :)
- Velocity.js
- GLSL.io and Diaporama Maker
Implementation based on this article.
You need a node
installed.
Install the deps:
npm install
The library is in index.js
.
Ensure any modication will:
- keep validating the tests (run
npm test
) - not bring performance regression (compare with
npm run benchmark
– don't rely 100% on its precision but it still helps to notice big gaps) - Run the visual example:
npm run visual