pic-scale 0.7.5

/*
 * Copyright (c) Radzivon Bartoshyk. All rights reserved.
 *
 * Redistribution and use in source and binary forms, with or without modification,
 * are permitted provided that the following conditions are met:
 *
 * 1.  Redistributions of source code must retain the above copyright notice, this
 * list of conditions and the following disclaimer.
 *
 * 2.  Redistributions in binary form must reproduce the above copyright notice,
 * this list of conditions and the following disclaimer in the documentation
 * and/or other materials provided with the distribution.
 *
 * 3.  Neither the name of the copyright holder nor the names of its
 * contributors may be used to endorse or promote products derived from
 * this software without specific prior written permission.
 *
 * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
 * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
 * DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT HOLDER OR CONTRIBUTORS BE LIABLE
 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR
 * SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER
 * CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY,
 * OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
 * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
 */
#![forbid(unsafe_code)]

use crate::WorkloadStrategy;
#[cfg(all(target_arch = "x86_64", feature = "avx"))]
use crate::avx2::{avx_premultiply_alpha_rgba, avx_unpremultiply_alpha_rgba};
#[cfg(all(target_arch = "aarch64", feature = "neon"))]
use crate::neon::{neon_premultiply_alpha_rgba, neon_unpremultiply_alpha_rgba};
#[cfg(all(any(target_arch = "x86_64", target_arch = "x86"), feature = "sse"))]
use crate::sse::*;
#[cfg(all(target_arch = "wasm32", target_feature = "simd128",))]
use crate::wasm32::{wasm_premultiply_alpha_rgba, wasm_unpremultiply_alpha_rgba};
use novtb::{ParallelZonedIterator, TbSliceMut};
use std::sync::OnceLock;

#[inline]
/// Divides value by 255 with rounding to nearest
pub(crate) fn div_by_255(v: u16) -> u8 {
    ((((v + 0x80) >> 8) + v + 0x80) >> 8) as u8
}

pub(crate) fn premultiply_alpha_rgba_row_impl(dst: &mut [u8], src: &[u8]) {
    for (dst, src) in dst
        .as_chunks_mut::<4>()
        .0
        .iter_mut()
        .zip(src.as_chunks::<4>().0.iter())
    {
        let a = src[3] as u16;
        dst[0] = div_by_255(src[0] as u16 * a);
        dst[1] = div_by_255(src[1] as u16 * a);
        dst[2] = div_by_255(src[2] as u16 * a);
        dst[3] = div_by_255(255 * a);
    }
}

pub(crate) fn premultiply_alpha_gray_alpha_row_impl(dst: &mut [u8], src: &[u8]) {
    for (dst, src) in dst
        .as_chunks_mut::<2>()
        .0
        .iter_mut()
        .zip(src.as_chunks::<2>().0.iter())
    {
        let a = src[1] as u16;
        dst[0] = div_by_255(src[0] as u16 * a);
        dst[1] = div_by_255(255 * a);
    }
}

fn premultiply_alpha_rgba_impl(dst: &mut [u8], src: &[u8]) {
    premultiply_alpha_rgba_row_impl(dst, src);
}

fn premultiply_alpha_gray_alpha_impl(
    dst: &mut [u8],
    dst_stride: usize,
    src: &[u8],
    _: usize,
    _: usize,
    stride: usize,
    pool: &novtb::ThreadPool,
) {
    dst.tb_par_chunks_mut(dst_stride)
        .zip(src.chunks(stride))
        .for_each(pool, |(dst, src)| {
            premultiply_alpha_gray_alpha_row_impl(dst, src);
        });
}

static UNPREMULTIPLICATION_TABLE: OnceLock<Box<[u8; 65536]>> = OnceLock::new();

pub(crate) fn unpremultiplication_table() -> &'static [u8; 65536] {
    UNPREMULTIPLICATION_TABLE.get_or_init(|| {
        let mut buf = Box::new([0u8; 65536]);
        for alpha in 0..256 {
            for pixel in 0..256 {
                #[allow(clippy::manual_checked_ops)]
                if alpha == 0 {
                    buf[alpha * 255 + pixel] = 0;
                } else {
                    let value = (pixel * 255 + alpha / 2) / alpha;
                    buf[alpha * 255 + pixel] = if value > 255 { 255 } else { value as u8 };
                }
            }
        }
        buf
    })
}

#[inline]
pub(crate) fn unpremultiply_alpha_rgba_row_impl(in_place: &mut [u8]) {
    let table = unpremultiplication_table();
    for dst in in_place.as_chunks_mut::<4>().0.iter_mut() {
        let a = dst[3];
        let z = a as u16 * 255;
        dst[0] = table[(z + dst[0] as u16) as usize];
        dst[1] = table[(z + dst[1] as u16) as usize];
        dst[2] = table[(z + dst[2] as u16) as usize];
    }
}

#[inline]
pub(crate) fn unpremultiply_alpha_gray_alpha_row_impl(in_place: &mut [u8]) {
    let table = unpremultiplication_table();
    for dst in in_place.as_chunks_mut::<2>().0.iter_mut() {
        let a = dst[1];
        let z = a as u16 * 255;
        dst[0] = table[(z + dst[0] as u16) as usize];
    }
}

fn unpremultiply_alpha_rgba_impl(in_place: &mut [u8], _: WorkloadStrategy) {
    unpremultiply_alpha_rgba_row_impl(in_place);
}

pub(crate) fn premultiply_alpha_rgba(
    dst: &mut [u8],
    dst_stride: usize,
    src: &[u8],
    width: usize,
    _: usize,
    src_stride: usize,
    pool: &novtb::ThreadPool,
) {
    let mut _dispatcher: fn(&mut [u8], &[u8]) = premultiply_alpha_rgba_impl;
    #[cfg(all(target_arch = "aarch64", feature = "neon"))]
    {
        _dispatcher = neon_premultiply_alpha_rgba;
    }
    #[cfg(all(any(target_arch = "x86_64", target_arch = "x86"), feature = "sse"))]
    {
        if std::arch::is_x86_feature_detected!("sse4.1") {
            _dispatcher = sse_premultiply_alpha_rgba;
        }
    }
    #[cfg(all(target_arch = "x86_64", feature = "avx"))]
    {
        if std::arch::is_x86_feature_detected!("avx2") {
            _dispatcher = avx_premultiply_alpha_rgba;
        }
    }
    #[cfg(all(target_arch = "wasm32", target_feature = "simd128"))]
    {
        _dispatcher = wasm_premultiply_alpha_rgba;
    }
    dst.tb_par_chunks_mut(dst_stride)
        .zip(src.chunks(src_stride))
        .for_each(pool, |(dst, src)| {
            _dispatcher(&mut dst[..width * 4], &src[..width * 4]);
        });
}

pub(crate) fn premultiply_alpha_gray_alpha(
    dst: &mut [u8],
    dst_stride: usize,
    src: &[u8],
    width: usize,
    height: usize,
    src_stride: usize,
    pool: &novtb::ThreadPool,
) {
    #[allow(clippy::type_complexity)]
    let mut _dispatcher: fn(&mut [u8], usize, &[u8], usize, usize, usize, &novtb::ThreadPool) =
        premultiply_alpha_gray_alpha_impl;
    _dispatcher(dst, dst_stride, src, width, height, src_stride, pool);
}

pub(crate) fn unpremultiply_alpha_rgba(
    in_place: &mut [u8],
    width: usize,
    _: usize,
    stride: usize,
    pool: &novtb::ThreadPool,
    workload_strategy: WorkloadStrategy,
) {
    let mut _dispatcher: fn(&mut [u8], WorkloadStrategy) = unpremultiply_alpha_rgba_impl;
    #[cfg(all(target_arch = "aarch64", feature = "neon"))]
    {
        _dispatcher = neon_unpremultiply_alpha_rgba;
    }
    #[cfg(all(any(target_arch = "x86_64", target_arch = "x86"), feature = "sse"))]
    {
        if std::arch::is_x86_feature_detected!("sse4.1") {
            _dispatcher = sse_unpremultiply_alpha_rgba;
        }
    }
    #[cfg(all(target_arch = "x86_64", feature = "avx"))]
    {
        if std::arch::is_x86_feature_detected!("avx2") {
            _dispatcher = avx_unpremultiply_alpha_rgba;
        }
    }
    #[cfg(all(target_arch = "wasm32", target_feature = "simd128"))]
    {
        _dispatcher = wasm_unpremultiply_alpha_rgba;
    }
    in_place.tb_par_chunks_mut(stride).for_each(pool, |row| {
        _dispatcher(&mut row[..width * 4], workload_strategy);
    });
}

fn unpremultiply_alpha_gray_alpha_impl(
    in_place: &mut [u8],
    width: usize,
    _: usize,
    stride: usize,
    pool: &novtb::ThreadPool,
    _: WorkloadStrategy,
) {
    in_place.tb_par_chunks_mut(stride).for_each(pool, |row| {
        unpremultiply_alpha_gray_alpha_row_impl(&mut row[..width * 2]);
    });
}

pub(crate) fn unpremultiply_alpha_gray_alpha(
    in_place: &mut [u8],
    width: usize,
    height: usize,
    stride: usize,
    pool: &novtb::ThreadPool,
    workload_strategy: WorkloadStrategy,
) {
    let mut _dispatcher: fn(&mut [u8], usize, usize, usize, &novtb::ThreadPool, WorkloadStrategy) =
        unpremultiply_alpha_gray_alpha_impl;
    _dispatcher(in_place, width, height, stride, pool, workload_strategy);
}