cluFullTransmute

A more complete and advanced version of data transmutation.

Opportunities

1. Reduction of any A to any B, without checking the dimensionality of the data.
2. The ability to use transmute with const functions.
3. Possibility of delayed transmutation.
4. The library uses #![no_std]

A warning!

1. This library only works in a nightly compiler, we expect stabilization features.
2. You really need to understand what you are doing.

Use

1. Easy

use cluFullTransmute::mem::full_transmute;

fn main() {
	let a: bool = unsafe{ full_transmute(1u8) };
	assert_eq!(a, true);
	
	let b: bool = unsafe{ full_transmute(0u8) };
	assert_eq!(b, false);
	
	// Why does this work?
	//
	// Is bool one bit?
	// No, bool is not one bit, but u8.
	//
	assert_eq!(std::mem::size_of::<bool>(), 1);
}

2. GenericType

use cluFullTransmute::mem::full_transmute;

#[allow(dead_code)]
struct A<T>(T);

impl<T> Drop for A<T> {
	fn drop(&mut self) {
		panic!("Strange behavior of the internal library.");
	}
}

#[allow(dead_code)]
struct B<T>(T);

impl<T> B<T> {
	pub fn my_fn(&self) {}
}

fn main() {
	let data = A(9999usize); //ignore drop!
	
	let b: B<usize> = unsafe{ full_transmute(data) };
	assert_eq!(b.0, 9999usize);
	
	b.my_fn();
}

License

Copyright 2019 #UlinProject Denis Kotlyarov (Денис Котляров)

Licensed under the Apache License, Version 2.0