// Copyright 2018 Developers of the Rand project.
//
// Licensed under the Apache License, Version 2.0 <LICENSE-APACHE or
// https://www.apache.org/licenses/LICENSE-2.0> or the MIT license
// <LICENSE-MIT or https://opensource.org/licenses/MIT>, at your
// option. This file may not be copied, modified, or distributed
// except according to those terms.

//! Helper functions for implementing `RngCore` functions.
//!
//! For cross-platform reproducibility, these functions all use Little Endian:
//! least-significant part first. For example, `next_u64_via_u32` takes `u32`
//! values `x, y`, then outputs `(y << 32) | x`. To implement `next_u32`
//! from `next_u64` in little-endian order, one should use `next_u64() as u32`.
//!
//! Byte-swapping (like the std `to_le` functions) is only needed to convert
//! to/from byte sequences, and since its purpose is reproducibility,
//! non-reproducible sources (e.g. `OsRng`) need not bother with it.

use core::ptr::copy_nonoverlapping;
use core::slice;
use core::cmp::min;
use core::mem::size_of;
use crate::RngCore;


/// Implement `next_u64` via `next_u32`, little-endian order.
pub fn next_u64_via_u32<R: RngCore + ?Sized>(rng: &mut R) -> u64 {
    // Use LE; we explicitly generate one value before the next.
    let x = u64::from(rng.next_u32());
    let y = u64::from(rng.next_u32());
    (y << 32) | x
}

/// Implement `fill_bytes` via `next_u64` and `next_u32`, little-endian order.
///
/// The fastest way to fill a slice is usually to work as long as possible with
/// integers. That is why this method mostly uses `next_u64`, and only when
/// there are 4 or less bytes remaining at the end of the slice it uses
/// `next_u32` once.
pub fn fill_bytes_via_next<R: RngCore + ?Sized>(rng: &mut R, dest: &mut [u8]) {
    let mut left = dest;
    while left.len() >= 8 {
        let (l, r) = {left}.split_at_mut(8);
        left = r;
        let chunk: [u8; 8] = rng.next_u64().to_le_bytes();
        l.copy_from_slice(&chunk);
    }
    let n = left.len();
    if n > 4 {
        let chunk: [u8; 8] = rng.next_u64().to_le_bytes();
        left.copy_from_slice(&chunk[..n]);
    } else if n > 0 {
        let chunk: [u8; 4] = rng.next_u32().to_le_bytes();
        left.copy_from_slice(&chunk[..n]);
    }
}

macro_rules! impl_uint_from_fill {
    ($rng:expr, $ty:ty, $N:expr) => ({
        debug_assert!($N == size_of::<$ty>());

        let mut int: $ty = 0;
        unsafe {
            let ptr = &mut int as *mut $ty as *mut u8;
            let slice = slice::from_raw_parts_mut(ptr, $N);
            $rng.fill_bytes(slice);
        }
        int
    });
}

macro_rules! fill_via_chunks {
    ($src:expr, $dst:expr, $ty:ty, $size:expr) => ({
        let chunk_size_u8 = min($src.len() * $size, $dst.len());
        let chunk_size = (chunk_size_u8 + $size - 1) / $size;
        if cfg!(target_endian="little") {
            unsafe {
                copy_nonoverlapping(
                    $src.as_ptr() as *const u8,
                    $dst.as_mut_ptr(),
                    chunk_size_u8);
            }
        } else {
            for (&n, chunk) in $src.iter().zip($dst.chunks_mut($size)) {
                let tmp = n.to_le();
                let src_ptr = &tmp as *const $ty as *const u8;
                unsafe {
                    copy_nonoverlapping(src_ptr,
                                        chunk.as_mut_ptr(),
                                        chunk.len());
                }
            }
        }

        (chunk_size, chunk_size_u8)
    });
}

/// Implement `fill_bytes` by reading chunks from the output buffer of a block
/// based RNG.
///
/// The return values are `(consumed_u32, filled_u8)`.
///
/// `filled_u8` is the number of filled bytes in `dest`, which may be less than
/// the length of `dest`.
/// `consumed_u32` is the number of words consumed from `src`, which is the same
/// as `filled_u8 / 4` rounded up.
///
/// # Example
/// (from `IsaacRng`)
///
/// ```ignore
/// fn fill_bytes(&mut self, dest: &mut [u8]) {
///     let mut read_len = 0;
///     while read_len < dest.len() {
///         if self.index >= self.rsl.len() {
///             self.isaac();
///         }
///
///         let (consumed_u32, filled_u8) =
///             impls::fill_via_u32_chunks(&mut self.rsl[self.index..],
///                                        &mut dest[read_len..]);
///
///         self.index += consumed_u32;
///         read_len += filled_u8;
///     }
/// }
/// ```
pub fn fill_via_u32_chunks(src: &[u32], dest: &mut [u8]) -> (usize, usize) {
    fill_via_chunks!(src, dest, u32, 4)
}

/// Implement `fill_bytes` by reading chunks from the output buffer of a block
/// based RNG.
///
/// The return values are `(consumed_u64, filled_u8)`.
/// `filled_u8` is the number of filled bytes in `dest`, which may be less than
/// the length of `dest`.
/// `consumed_u64` is the number of words consumed from `src`, which is the same
/// as `filled_u8 / 8` rounded up.
///
/// See `fill_via_u32_chunks` for an example.
pub fn fill_via_u64_chunks(src: &[u64], dest: &mut [u8]) -> (usize, usize) {
    fill_via_chunks!(src, dest, u64, 8)
}

/// Implement `next_u32` via `fill_bytes`, little-endian order.
pub fn next_u32_via_fill<R: RngCore + ?Sized>(rng: &mut R) -> u32 {
    impl_uint_from_fill!(rng, u32, 4)
}

/// Implement `next_u64` via `fill_bytes`, little-endian order.
pub fn next_u64_via_fill<R: RngCore + ?Sized>(rng: &mut R) -> u64 {
    impl_uint_from_fill!(rng, u64, 8)
}

// TODO: implement tests for the above