1 unstable release

0.1.0	Oct 29, 2022

#2008 in Rust patterns

4,855 downloads per month

MIT/Apache

6KB

`⚡ float-derive`

A crate that allows deriving Eq and Hash for types that contain floats 🦀

[dependencies]
float-derive = "0.1.0"

IEEE 754 floating point numbers can have the value NAN (not a number). It is unequal to any float, including itself! This is the reason `f32` and `f64` doesn’t implement the `Eq` trait.

This crate allows to derive Eq and Hash for traits that contain floating points by explicitly comparing floating points such that: NAN == NAN is true.

How to use

use float_derive::{FloatPartialEq, FloatEq, FloatHash};

#[derive(FloatPartialEq, FloatEq, FloatHash)]
struct MyStruct {
    a: i32,
    b: i32,
    my_float: f32,
    my_second_float: f64
}

#[derive(FloatPartialEq, FloatEq, FloatHash)]
enum MyEnum {
    A(i32, f32),
    B { a: i32, b: f64 },
    C
}

Credits

ordered-float for offering explicit float equality.

Dependencies

~1.5MB
~41K SLoC