csvroll 0.1.0

// the whole file taken from pikkr
use x86intrin::{m256i, mm256_setr_epi8};

#[inline]
pub fn mm256i(i: i8) -> m256i {
    mm256_setr_epi8(
        i,
        i,
        i,
        i,
        i,
        i,
        i,
        i,
        i,
        i,
        i,
        i,
        i,
        i,
        i,
        i,
        i,
        i,
        i,
        i,
        i,
        i,
        i,
        i,
        i,
        i,
        i,
        i,
        i,
        i,
        i,
        i,
    )
}

/// Create a value of `m256i` from the slice of bytes.
///
/// # Safety
/// This function guarantees the safety only if the length of `s` is greater than or equal to `i + 32`.
#[inline(always)]
pub unsafe fn u8_to_m256i(s: &[u8], i: usize) -> m256i {
    mm256_setr_epi8(
        *s.as_ptr().offset(i as isize + 0) as i8,
        *s.as_ptr().offset(i as isize + 1) as i8,
        *s.as_ptr().offset(i as isize + 2) as i8,
        *s.as_ptr().offset(i as isize + 3) as i8,
        *s.as_ptr().offset(i as isize + 4) as i8,
        *s.as_ptr().offset(i as isize + 5) as i8,
        *s.as_ptr().offset(i as isize + 6) as i8,
        *s.as_ptr().offset(i as isize + 7) as i8,
        *s.as_ptr().offset(i as isize + 8) as i8,
        *s.as_ptr().offset(i as isize + 9) as i8,
        *s.as_ptr().offset(i as isize + 10) as i8,
        *s.as_ptr().offset(i as isize + 11) as i8,
        *s.as_ptr().offset(i as isize + 12) as i8,
        *s.as_ptr().offset(i as isize + 13) as i8,
        *s.as_ptr().offset(i as isize + 14) as i8,
        *s.as_ptr().offset(i as isize + 15) as i8,
        *s.as_ptr().offset(i as isize + 16) as i8,
        *s.as_ptr().offset(i as isize + 17) as i8,
        *s.as_ptr().offset(i as isize + 18) as i8,
        *s.as_ptr().offset(i as isize + 19) as i8,
        *s.as_ptr().offset(i as isize + 20) as i8,
        *s.as_ptr().offset(i as isize + 21) as i8,
        *s.as_ptr().offset(i as isize + 22) as i8,
        *s.as_ptr().offset(i as isize + 23) as i8,
        *s.as_ptr().offset(i as isize + 24) as i8,
        *s.as_ptr().offset(i as isize + 25) as i8,
        *s.as_ptr().offset(i as isize + 26) as i8,
        *s.as_ptr().offset(i as isize + 27) as i8,
        *s.as_ptr().offset(i as isize + 28) as i8,
        *s.as_ptr().offset(i as isize + 29) as i8,
        *s.as_ptr().offset(i as isize + 30) as i8,
        *s.as_ptr().offset(i as isize + 31) as i8,
    )
}

#[inline]
pub unsafe fn u8_to_m256i_rest(s: &[u8], i: usize) -> m256i {
    match s.len() - i {
        31 => mm256_setr_epi8(
            *s.as_ptr().offset(i as isize + 0) as i8,
            *s.as_ptr().offset(i as isize + 1) as i8,
            *s.as_ptr().offset(i as isize + 2) as i8,
            *s.as_ptr().offset(i as isize + 3) as i8,
            *s.as_ptr().offset(i as isize + 4) as i8,
            *s.as_ptr().offset(i as isize + 5) as i8,
            *s.as_ptr().offset(i as isize + 6) as i8,
            *s.as_ptr().offset(i as isize + 7) as i8,
            *s.as_ptr().offset(i as isize + 8) as i8,
            *s.as_ptr().offset(i as isize + 9) as i8,
            *s.as_ptr().offset(i as isize + 10) as i8,
            *s.as_ptr().offset(i as isize + 11) as i8,
            *s.as_ptr().offset(i as isize + 12) as i8,
            *s.as_ptr().offset(i as isize + 13) as i8,
            *s.as_ptr().offset(i as isize + 14) as i8,
            *s.as_ptr().offset(i as isize + 15) as i8,
            *s.as_ptr().offset(i as isize + 16) as i8,
            *s.as_ptr().offset(i as isize + 17) as i8,
            *s.as_ptr().offset(i as isize + 18) as i8,
            *s.as_ptr().offset(i as isize + 19) as i8,
            *s.as_ptr().offset(i as isize + 20) as i8,
            *s.as_ptr().offset(i as isize + 21) as i8,
            *s.as_ptr().offset(i as isize + 22) as i8,
            *s.as_ptr().offset(i as isize + 23) as i8,
            *s.as_ptr().offset(i as isize + 24) as i8,
            *s.as_ptr().offset(i as isize + 25) as i8,
            *s.as_ptr().offset(i as isize + 26) as i8,
            *s.as_ptr().offset(i as isize + 27) as i8,
            *s.as_ptr().offset(i as isize + 28) as i8,
            *s.as_ptr().offset(i as isize + 29) as i8,
            *s.as_ptr().offset(i as isize + 30) as i8,
            0,
        ),
        30 => mm256_setr_epi8(
            *s.as_ptr().offset(i as isize + 0) as i8,
            *s.as_ptr().offset(i as isize + 1) as i8,
            *s.as_ptr().offset(i as isize + 2) as i8,
            *s.as_ptr().offset(i as isize + 3) as i8,
            *s.as_ptr().offset(i as isize + 4) as i8,
            *s.as_ptr().offset(i as isize + 5) as i8,
            *s.as_ptr().offset(i as isize + 6) as i8,
            *s.as_ptr().offset(i as isize + 7) as i8,
            *s.as_ptr().offset(i as isize + 8) as i8,
            *s.as_ptr().offset(i as isize + 9) as i8,
            *s.as_ptr().offset(i as isize + 10) as i8,
            *s.as_ptr().offset(i as isize + 11) as i8,
            *s.as_ptr().offset(i as isize + 12) as i8,
            *s.as_ptr().offset(i as isize + 13) as i8,
            *s.as_ptr().offset(i as isize + 14) as i8,
            *s.as_ptr().offset(i as isize + 15) as i8,
            *s.as_ptr().offset(i as isize + 16) as i8,
            *s.as_ptr().offset(i as isize + 17) as i8,
            *s.as_ptr().offset(i as isize + 18) as i8,
            *s.as_ptr().offset(i as isize + 19) as i8,
            *s.as_ptr().offset(i as isize + 20) as i8,
            *s.as_ptr().offset(i as isize + 21) as i8,
            *s.as_ptr().offset(i as isize + 22) as i8,
            *s.as_ptr().offset(i as isize + 23) as i8,
            *s.as_ptr().offset(i as isize + 24) as i8,
            *s.as_ptr().offset(i as isize + 25) as i8,
            *s.as_ptr().offset(i as isize + 26) as i8,
            *s.as_ptr().offset(i as isize + 27) as i8,
            *s.as_ptr().offset(i as isize + 28) as i8,
            *s.as_ptr().offset(i as isize + 29) as i8,
            0,
            0,
        ),
        29 => mm256_setr_epi8(
            *s.as_ptr().offset(i as isize + 0) as i8,
            *s.as_ptr().offset(i as isize + 1) as i8,
            *s.as_ptr().offset(i as isize + 2) as i8,
            *s.as_ptr().offset(i as isize + 3) as i8,
            *s.as_ptr().offset(i as isize + 4) as i8,
            *s.as_ptr().offset(i as isize + 5) as i8,
            *s.as_ptr().offset(i as isize + 6) as i8,
            *s.as_ptr().offset(i as isize + 7) as i8,
            *s.as_ptr().offset(i as isize + 8) as i8,
            *s.as_ptr().offset(i as isize + 9) as i8,
            *s.as_ptr().offset(i as isize + 10) as i8,
            *s.as_ptr().offset(i as isize + 11) as i8,
            *s.as_ptr().offset(i as isize + 12) as i8,
            *s.as_ptr().offset(i as isize + 13) as i8,
            *s.as_ptr().offset(i as isize + 14) as i8,
            *s.as_ptr().offset(i as isize + 15) as i8,
            *s.as_ptr().offset(i as isize + 16) as i8,
            *s.as_ptr().offset(i as isize + 17) as i8,
            *s.as_ptr().offset(i as isize + 18) as i8,
            *s.as_ptr().offset(i as isize + 19) as i8,
            *s.as_ptr().offset(i as isize + 20) as i8,
            *s.as_ptr().offset(i as isize + 21) as i8,
            *s.as_ptr().offset(i as isize + 22) as i8,
            *s.as_ptr().offset(i as isize + 23) as i8,
            *s.as_ptr().offset(i as isize + 24) as i8,
            *s.as_ptr().offset(i as isize + 25) as i8,
            *s.as_ptr().offset(i as isize + 26) as i8,
            *s.as_ptr().offset(i as isize + 27) as i8,
            *s.as_ptr().offset(i as isize + 28) as i8,
            0,
            0,
            0,
        ),
        28 => mm256_setr_epi8(
            *s.as_ptr().offset(i as isize + 0) as i8,
            *s.as_ptr().offset(i as isize + 1) as i8,
            *s.as_ptr().offset(i as isize + 2) as i8,
            *s.as_ptr().offset(i as isize + 3) as i8,
            *s.as_ptr().offset(i as isize + 4) as i8,
            *s.as_ptr().offset(i as isize + 5) as i8,
            *s.as_ptr().offset(i as isize + 6) as i8,
            *s.as_ptr().offset(i as isize + 7) as i8,
            *s.as_ptr().offset(i as isize + 8) as i8,
            *s.as_ptr().offset(i as isize + 9) as i8,
            *s.as_ptr().offset(i as isize + 10) as i8,
            *s.as_ptr().offset(i as isize + 11) as i8,
            *s.as_ptr().offset(i as isize + 12) as i8,
            *s.as_ptr().offset(i as isize + 13) as i8,
            *s.as_ptr().offset(i as isize + 14) as i8,
            *s.as_ptr().offset(i as isize + 15) as i8,
            *s.as_ptr().offset(i as isize + 16) as i8,
            *s.as_ptr().offset(i as isize + 17) as i8,
            *s.as_ptr().offset(i as isize + 18) as i8,
            *s.as_ptr().offset(i as isize + 19) as i8,
            *s.as_ptr().offset(i as isize + 20) as i8,
            *s.as_ptr().offset(i as isize + 21) as i8,
            *s.as_ptr().offset(i as isize + 22) as i8,
            *s.as_ptr().offset(i as isize + 23) as i8,
            *s.as_ptr().offset(i as isize + 24) as i8,
            *s.as_ptr().offset(i as isize + 25) as i8,
            *s.as_ptr().offset(i as isize + 26) as i8,
            *s.as_ptr().offset(i as isize + 27) as i8,
            0,
            0,
            0,
            0,
        ),
        27 => mm256_setr_epi8(
            *s.as_ptr().offset(i as isize + 0) as i8,
            *s.as_ptr().offset(i as isize + 1) as i8,
            *s.as_ptr().offset(i as isize + 2) as i8,
            *s.as_ptr().offset(i as isize + 3) as i8,
            *s.as_ptr().offset(i as isize + 4) as i8,
            *s.as_ptr().offset(i as isize + 5) as i8,
            *s.as_ptr().offset(i as isize + 6) as i8,
            *s.as_ptr().offset(i as isize + 7) as i8,
            *s.as_ptr().offset(i as isize + 8) as i8,
            *s.as_ptr().offset(i as isize + 9) as i8,
            *s.as_ptr().offset(i as isize + 10) as i8,
            *s.as_ptr().offset(i as isize + 11) as i8,
            *s.as_ptr().offset(i as isize + 12) as i8,
            *s.as_ptr().offset(i as isize + 13) as i8,
            *s.as_ptr().offset(i as isize + 14) as i8,
            *s.as_ptr().offset(i as isize + 15) as i8,
            *s.as_ptr().offset(i as isize + 16) as i8,
            *s.as_ptr().offset(i as isize + 17) as i8,
            *s.as_ptr().offset(i as isize + 18) as i8,
            *s.as_ptr().offset(i as isize + 19) as i8,
            *s.as_ptr().offset(i as isize + 20) as i8,
            *s.as_ptr().offset(i as isize + 21) as i8,
            *s.as_ptr().offset(i as isize + 22) as i8,
            *s.as_ptr().offset(i as isize + 23) as i8,
            *s.as_ptr().offset(i as isize + 24) as i8,
            *s.as_ptr().offset(i as isize + 25) as i8,
            *s.as_ptr().offset(i as isize + 26) as i8,
            0,
            0,
            0,
            0,
            0,
        ),
        26 => mm256_setr_epi8(
            *s.as_ptr().offset(i as isize + 0) as i8,
            *s.as_ptr().offset(i as isize + 1) as i8,
            *s.as_ptr().offset(i as isize + 2) as i8,
            *s.as_ptr().offset(i as isize + 3) as i8,
            *s.as_ptr().offset(i as isize + 4) as i8,
            *s.as_ptr().offset(i as isize + 5) as i8,
            *s.as_ptr().offset(i as isize + 6) as i8,
            *s.as_ptr().offset(i as isize + 7) as i8,
            *s.as_ptr().offset(i as isize + 8) as i8,
            *s.as_ptr().offset(i as isize + 9) as i8,
            *s.as_ptr().offset(i as isize + 10) as i8,
            *s.as_ptr().offset(i as isize + 11) as i8,
            *s.as_ptr().offset(i as isize + 12) as i8,
            *s.as_ptr().offset(i as isize + 13) as i8,
            *s.as_ptr().offset(i as isize + 14) as i8,
            *s.as_ptr().offset(i as isize + 15) as i8,
            *s.as_ptr().offset(i as isize + 16) as i8,
            *s.as_ptr().offset(i as isize + 17) as i8,
            *s.as_ptr().offset(i as isize + 18) as i8,
            *s.as_ptr().offset(i as isize + 19) as i8,
            *s.as_ptr().offset(i as isize + 20) as i8,
            *s.as_ptr().offset(i as isize + 21) as i8,
            *s.as_ptr().offset(i as isize + 22) as i8,
            *s.as_ptr().offset(i as isize + 23) as i8,
            *s.as_ptr().offset(i as isize + 24) as i8,
            *s.as_ptr().offset(i as isize + 25) as i8,
            0,
            0,
            0,
            0,
            0,
            0,
        ),
        25 => mm256_setr_epi8(
            *s.as_ptr().offset(i as isize + 0) as i8,
            *s.as_ptr().offset(i as isize + 1) as i8,
            *s.as_ptr().offset(i as isize + 2) as i8,
            *s.as_ptr().offset(i as isize + 3) as i8,
            *s.as_ptr().offset(i as isize + 4) as i8,
            *s.as_ptr().offset(i as isize + 5) as i8,
            *s.as_ptr().offset(i as isize + 6) as i8,
            *s.as_ptr().offset(i as isize + 7) as i8,
            *s.as_ptr().offset(i as isize + 8) as i8,
            *s.as_ptr().offset(i as isize + 9) as i8,
            *s.as_ptr().offset(i as isize + 10) as i8,
            *s.as_ptr().offset(i as isize + 11) as i8,
            *s.as_ptr().offset(i as isize + 12) as i8,
            *s.as_ptr().offset(i as isize + 13) as i8,
            *s.as_ptr().offset(i as isize + 14) as i8,
            *s.as_ptr().offset(i as isize + 15) as i8,
            *s.as_ptr().offset(i as isize + 16) as i8,
            *s.as_ptr().offset(i as isize + 17) as i8,
            *s.as_ptr().offset(i as isize + 18) as i8,
            *s.as_ptr().offset(i as isize + 19) as i8,
            *s.as_ptr().offset(i as isize + 20) as i8,
            *s.as_ptr().offset(i as isize + 21) as i8,
            *s.as_ptr().offset(i as isize + 22) as i8,
            *s.as_ptr().offset(i as isize + 23) as i8,
            *s.as_ptr().offset(i as isize + 24) as i8,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
        ),
        24 => mm256_setr_epi8(
            *s.as_ptr().offset(i as isize + 0) as i8,
            *s.as_ptr().offset(i as isize + 1) as i8,
            *s.as_ptr().offset(i as isize + 2) as i8,
            *s.as_ptr().offset(i as isize + 3) as i8,
            *s.as_ptr().offset(i as isize + 4) as i8,
            *s.as_ptr().offset(i as isize + 5) as i8,
            *s.as_ptr().offset(i as isize + 6) as i8,
            *s.as_ptr().offset(i as isize + 7) as i8,
            *s.as_ptr().offset(i as isize + 8) as i8,
            *s.as_ptr().offset(i as isize + 9) as i8,
            *s.as_ptr().offset(i as isize + 10) as i8,
            *s.as_ptr().offset(i as isize + 11) as i8,
            *s.as_ptr().offset(i as isize + 12) as i8,
            *s.as_ptr().offset(i as isize + 13) as i8,
            *s.as_ptr().offset(i as isize + 14) as i8,
            *s.as_ptr().offset(i as isize + 15) as i8,
            *s.as_ptr().offset(i as isize + 16) as i8,
            *s.as_ptr().offset(i as isize + 17) as i8,
            *s.as_ptr().offset(i as isize + 18) as i8,
            *s.as_ptr().offset(i as isize + 19) as i8,
            *s.as_ptr().offset(i as isize + 20) as i8,
            *s.as_ptr().offset(i as isize + 21) as i8,
            *s.as_ptr().offset(i as isize + 22) as i8,
            *s.as_ptr().offset(i as isize + 23) as i8,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
        ),
        23 => mm256_setr_epi8(
            *s.as_ptr().offset(i as isize + 0) as i8,
            *s.as_ptr().offset(i as isize + 1) as i8,
            *s.as_ptr().offset(i as isize + 2) as i8,
            *s.as_ptr().offset(i as isize + 3) as i8,
            *s.as_ptr().offset(i as isize + 4) as i8,
            *s.as_ptr().offset(i as isize + 5) as i8,
            *s.as_ptr().offset(i as isize + 6) as i8,
            *s.as_ptr().offset(i as isize + 7) as i8,
            *s.as_ptr().offset(i as isize + 8) as i8,
            *s.as_ptr().offset(i as isize + 9) as i8,
            *s.as_ptr().offset(i as isize + 10) as i8,
            *s.as_ptr().offset(i as isize + 11) as i8,
            *s.as_ptr().offset(i as isize + 12) as i8,
            *s.as_ptr().offset(i as isize + 13) as i8,
            *s.as_ptr().offset(i as isize + 14) as i8,
            *s.as_ptr().offset(i as isize + 15) as i8,
            *s.as_ptr().offset(i as isize + 16) as i8,
            *s.as_ptr().offset(i as isize + 17) as i8,
            *s.as_ptr().offset(i as isize + 18) as i8,
            *s.as_ptr().offset(i as isize + 19) as i8,
            *s.as_ptr().offset(i as isize + 20) as i8,
            *s.as_ptr().offset(i as isize + 21) as i8,
            *s.as_ptr().offset(i as isize + 22) as i8,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
        ),
        22 => mm256_setr_epi8(
            *s.as_ptr().offset(i as isize + 0) as i8,
            *s.as_ptr().offset(i as isize + 1) as i8,
            *s.as_ptr().offset(i as isize + 2) as i8,
            *s.as_ptr().offset(i as isize + 3) as i8,
            *s.as_ptr().offset(i as isize + 4) as i8,
            *s.as_ptr().offset(i as isize + 5) as i8,
            *s.as_ptr().offset(i as isize + 6) as i8,
            *s.as_ptr().offset(i as isize + 7) as i8,
            *s.as_ptr().offset(i as isize + 8) as i8,
            *s.as_ptr().offset(i as isize + 9) as i8,
            *s.as_ptr().offset(i as isize + 10) as i8,
            *s.as_ptr().offset(i as isize + 11) as i8,
            *s.as_ptr().offset(i as isize + 12) as i8,
            *s.as_ptr().offset(i as isize + 13) as i8,
            *s.as_ptr().offset(i as isize + 14) as i8,
            *s.as_ptr().offset(i as isize + 15) as i8,
            *s.as_ptr().offset(i as isize + 16) as i8,
            *s.as_ptr().offset(i as isize + 17) as i8,
            *s.as_ptr().offset(i as isize + 18) as i8,
            *s.as_ptr().offset(i as isize + 19) as i8,
            *s.as_ptr().offset(i as isize + 20) as i8,
            *s.as_ptr().offset(i as isize + 21) as i8,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
        ),
        21 => mm256_setr_epi8(
            *s.as_ptr().offset(i as isize + 0) as i8,
            *s.as_ptr().offset(i as isize + 1) as i8,
            *s.as_ptr().offset(i as isize + 2) as i8,
            *s.as_ptr().offset(i as isize + 3) as i8,
            *s.as_ptr().offset(i as isize + 4) as i8,
            *s.as_ptr().offset(i as isize + 5) as i8,
            *s.as_ptr().offset(i as isize + 6) as i8,
            *s.as_ptr().offset(i as isize + 7) as i8,
            *s.as_ptr().offset(i as isize + 8) as i8,
            *s.as_ptr().offset(i as isize + 9) as i8,
            *s.as_ptr().offset(i as isize + 10) as i8,
            *s.as_ptr().offset(i as isize + 11) as i8,
            *s.as_ptr().offset(i as isize + 12) as i8,
            *s.as_ptr().offset(i as isize + 13) as i8,
            *s.as_ptr().offset(i as isize + 14) as i8,
            *s.as_ptr().offset(i as isize + 15) as i8,
            *s.as_ptr().offset(i as isize + 16) as i8,
            *s.as_ptr().offset(i as isize + 17) as i8,
            *s.as_ptr().offset(i as isize + 18) as i8,
            *s.as_ptr().offset(i as isize + 19) as i8,
            *s.as_ptr().offset(i as isize + 20) as i8,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
        ),
        20 => mm256_setr_epi8(
            *s.as_ptr().offset(i as isize + 0) as i8,
            *s.as_ptr().offset(i as isize + 1) as i8,
            *s.as_ptr().offset(i as isize + 2) as i8,
            *s.as_ptr().offset(i as isize + 3) as i8,
            *s.as_ptr().offset(i as isize + 4) as i8,
            *s.as_ptr().offset(i as isize + 5) as i8,
            *s.as_ptr().offset(i as isize + 6) as i8,
            *s.as_ptr().offset(i as isize + 7) as i8,
            *s.as_ptr().offset(i as isize + 8) as i8,
            *s.as_ptr().offset(i as isize + 9) as i8,
            *s.as_ptr().offset(i as isize + 10) as i8,
            *s.as_ptr().offset(i as isize + 11) as i8,
            *s.as_ptr().offset(i as isize + 12) as i8,
            *s.as_ptr().offset(i as isize + 13) as i8,
            *s.as_ptr().offset(i as isize + 14) as i8,
            *s.as_ptr().offset(i as isize + 15) as i8,
            *s.as_ptr().offset(i as isize + 16) as i8,
            *s.as_ptr().offset(i as isize + 17) as i8,
            *s.as_ptr().offset(i as isize + 18) as i8,
            *s.as_ptr().offset(i as isize + 19) as i8,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
        ),
        19 => mm256_setr_epi8(
            *s.as_ptr().offset(i as isize + 0) as i8,
            *s.as_ptr().offset(i as isize + 1) as i8,
            *s.as_ptr().offset(i as isize + 2) as i8,
            *s.as_ptr().offset(i as isize + 3) as i8,
            *s.as_ptr().offset(i as isize + 4) as i8,
            *s.as_ptr().offset(i as isize + 5) as i8,
            *s.as_ptr().offset(i as isize + 6) as i8,
            *s.as_ptr().offset(i as isize + 7) as i8,
            *s.as_ptr().offset(i as isize + 8) as i8,
            *s.as_ptr().offset(i as isize + 9) as i8,
            *s.as_ptr().offset(i as isize + 10) as i8,
            *s.as_ptr().offset(i as isize + 11) as i8,
            *s.as_ptr().offset(i as isize + 12) as i8,
            *s.as_ptr().offset(i as isize + 13) as i8,
            *s.as_ptr().offset(i as isize + 14) as i8,
            *s.as_ptr().offset(i as isize + 15) as i8,
            *s.as_ptr().offset(i as isize + 16) as i8,
            *s.as_ptr().offset(i as isize + 17) as i8,
            *s.as_ptr().offset(i as isize + 18) as i8,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
        ),
        18 => mm256_setr_epi8(
            *s.as_ptr().offset(i as isize + 0) as i8,
            *s.as_ptr().offset(i as isize + 1) as i8,
            *s.as_ptr().offset(i as isize + 2) as i8,
            *s.as_ptr().offset(i as isize + 3) as i8,
            *s.as_ptr().offset(i as isize + 4) as i8,
            *s.as_ptr().offset(i as isize + 5) as i8,
            *s.as_ptr().offset(i as isize + 6) as i8,
            *s.as_ptr().offset(i as isize + 7) as i8,
            *s.as_ptr().offset(i as isize + 8) as i8,
            *s.as_ptr().offset(i as isize + 9) as i8,
            *s.as_ptr().offset(i as isize + 10) as i8,
            *s.as_ptr().offset(i as isize + 11) as i8,
            *s.as_ptr().offset(i as isize + 12) as i8,
            *s.as_ptr().offset(i as isize + 13) as i8,
            *s.as_ptr().offset(i as isize + 14) as i8,
            *s.as_ptr().offset(i as isize + 15) as i8,
            *s.as_ptr().offset(i as isize + 16) as i8,
            *s.as_ptr().offset(i as isize + 17) as i8,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
        ),
        17 => mm256_setr_epi8(
            *s.as_ptr().offset(i as isize + 0) as i8,
            *s.as_ptr().offset(i as isize + 1) as i8,
            *s.as_ptr().offset(i as isize + 2) as i8,
            *s.as_ptr().offset(i as isize + 3) as i8,
            *s.as_ptr().offset(i as isize + 4) as i8,
            *s.as_ptr().offset(i as isize + 5) as i8,
            *s.as_ptr().offset(i as isize + 6) as i8,
            *s.as_ptr().offset(i as isize + 7) as i8,
            *s.as_ptr().offset(i as isize + 8) as i8,
            *s.as_ptr().offset(i as isize + 9) as i8,
            *s.as_ptr().offset(i as isize + 10) as i8,
            *s.as_ptr().offset(i as isize + 11) as i8,
            *s.as_ptr().offset(i as isize + 12) as i8,
            *s.as_ptr().offset(i as isize + 13) as i8,
            *s.as_ptr().offset(i as isize + 14) as i8,
            *s.as_ptr().offset(i as isize + 15) as i8,
            *s.as_ptr().offset(i as isize + 16) as i8,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
        ),
        16 => mm256_setr_epi8(
            *s.as_ptr().offset(i as isize + 0) as i8,
            *s.as_ptr().offset(i as isize + 1) as i8,
            *s.as_ptr().offset(i as isize + 2) as i8,
            *s.as_ptr().offset(i as isize + 3) as i8,
            *s.as_ptr().offset(i as isize + 4) as i8,
            *s.as_ptr().offset(i as isize + 5) as i8,
            *s.as_ptr().offset(i as isize + 6) as i8,
            *s.as_ptr().offset(i as isize + 7) as i8,
            *s.as_ptr().offset(i as isize + 8) as i8,
            *s.as_ptr().offset(i as isize + 9) as i8,
            *s.as_ptr().offset(i as isize + 10) as i8,
            *s.as_ptr().offset(i as isize + 11) as i8,
            *s.as_ptr().offset(i as isize + 12) as i8,
            *s.as_ptr().offset(i as isize + 13) as i8,
            *s.as_ptr().offset(i as isize + 14) as i8,
            *s.as_ptr().offset(i as isize + 15) as i8,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
        ),
        15 => mm256_setr_epi8(
            *s.as_ptr().offset(i as isize + 0) as i8,
            *s.as_ptr().offset(i as isize + 1) as i8,
            *s.as_ptr().offset(i as isize + 2) as i8,
            *s.as_ptr().offset(i as isize + 3) as i8,
            *s.as_ptr().offset(i as isize + 4) as i8,
            *s.as_ptr().offset(i as isize + 5) as i8,
            *s.as_ptr().offset(i as isize + 6) as i8,
            *s.as_ptr().offset(i as isize + 7) as i8,
            *s.as_ptr().offset(i as isize + 8) as i8,
            *s.as_ptr().offset(i as isize + 9) as i8,
            *s.as_ptr().offset(i as isize + 10) as i8,
            *s.as_ptr().offset(i as isize + 11) as i8,
            *s.as_ptr().offset(i as isize + 12) as i8,
            *s.as_ptr().offset(i as isize + 13) as i8,
            *s.as_ptr().offset(i as isize + 14) as i8,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
        ),
        14 => mm256_setr_epi8(
            *s.as_ptr().offset(i as isize + 0) as i8,
            *s.as_ptr().offset(i as isize + 1) as i8,
            *s.as_ptr().offset(i as isize + 2) as i8,
            *s.as_ptr().offset(i as isize + 3) as i8,
            *s.as_ptr().offset(i as isize + 4) as i8,
            *s.as_ptr().offset(i as isize + 5) as i8,
            *s.as_ptr().offset(i as isize + 6) as i8,
            *s.as_ptr().offset(i as isize + 7) as i8,
            *s.as_ptr().offset(i as isize + 8) as i8,
            *s.as_ptr().offset(i as isize + 9) as i8,
            *s.as_ptr().offset(i as isize + 10) as i8,
            *s.as_ptr().offset(i as isize + 11) as i8,
            *s.as_ptr().offset(i as isize + 12) as i8,
            *s.as_ptr().offset(i as isize + 13) as i8,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
        ),
        13 => mm256_setr_epi8(
            *s.as_ptr().offset(i as isize + 0) as i8,
            *s.as_ptr().offset(i as isize + 1) as i8,
            *s.as_ptr().offset(i as isize + 2) as i8,
            *s.as_ptr().offset(i as isize + 3) as i8,
            *s.as_ptr().offset(i as isize + 4) as i8,
            *s.as_ptr().offset(i as isize + 5) as i8,
            *s.as_ptr().offset(i as isize + 6) as i8,
            *s.as_ptr().offset(i as isize + 7) as i8,
            *s.as_ptr().offset(i as isize + 8) as i8,
            *s.as_ptr().offset(i as isize + 9) as i8,
            *s.as_ptr().offset(i as isize + 10) as i8,
            *s.as_ptr().offset(i as isize + 11) as i8,
            *s.as_ptr().offset(i as isize + 12) as i8,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
        ),
        12 => mm256_setr_epi8(
            *s.as_ptr().offset(i as isize + 0) as i8,
            *s.as_ptr().offset(i as isize + 1) as i8,
            *s.as_ptr().offset(i as isize + 2) as i8,
            *s.as_ptr().offset(i as isize + 3) as i8,
            *s.as_ptr().offset(i as isize + 4) as i8,
            *s.as_ptr().offset(i as isize + 5) as i8,
            *s.as_ptr().offset(i as isize + 6) as i8,
            *s.as_ptr().offset(i as isize + 7) as i8,
            *s.as_ptr().offset(i as isize + 8) as i8,
            *s.as_ptr().offset(i as isize + 9) as i8,
            *s.as_ptr().offset(i as isize + 10) as i8,
            *s.as_ptr().offset(i as isize + 11) as i8,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
        ),
        11 => mm256_setr_epi8(
            *s.as_ptr().offset(i as isize + 0) as i8,
            *s.as_ptr().offset(i as isize + 1) as i8,
            *s.as_ptr().offset(i as isize + 2) as i8,
            *s.as_ptr().offset(i as isize + 3) as i8,
            *s.as_ptr().offset(i as isize + 4) as i8,
            *s.as_ptr().offset(i as isize + 5) as i8,
            *s.as_ptr().offset(i as isize + 6) as i8,
            *s.as_ptr().offset(i as isize + 7) as i8,
            *s.as_ptr().offset(i as isize + 8) as i8,
            *s.as_ptr().offset(i as isize + 9) as i8,
            *s.as_ptr().offset(i as isize + 10) as i8,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
        ),
        10 => mm256_setr_epi8(
            *s.as_ptr().offset(i as isize + 0) as i8,
            *s.as_ptr().offset(i as isize + 1) as i8,
            *s.as_ptr().offset(i as isize + 2) as i8,
            *s.as_ptr().offset(i as isize + 3) as i8,
            *s.as_ptr().offset(i as isize + 4) as i8,
            *s.as_ptr().offset(i as isize + 5) as i8,
            *s.as_ptr().offset(i as isize + 6) as i8,
            *s.as_ptr().offset(i as isize + 7) as i8,
            *s.as_ptr().offset(i as isize + 8) as i8,
            *s.as_ptr().offset(i as isize + 9) as i8,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
        ),
        9 => mm256_setr_epi8(
            *s.as_ptr().offset(i as isize + 0) as i8,
            *s.as_ptr().offset(i as isize + 1) as i8,
            *s.as_ptr().offset(i as isize + 2) as i8,
            *s.as_ptr().offset(i as isize + 3) as i8,
            *s.as_ptr().offset(i as isize + 4) as i8,
            *s.as_ptr().offset(i as isize + 5) as i8,
            *s.as_ptr().offset(i as isize + 6) as i8,
            *s.as_ptr().offset(i as isize + 7) as i8,
            *s.as_ptr().offset(i as isize + 8) as i8,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
        ),
        8 => mm256_setr_epi8(
            *s.as_ptr().offset(i as isize + 0) as i8,
            *s.as_ptr().offset(i as isize + 1) as i8,
            *s.as_ptr().offset(i as isize + 2) as i8,
            *s.as_ptr().offset(i as isize + 3) as i8,
            *s.as_ptr().offset(i as isize + 4) as i8,
            *s.as_ptr().offset(i as isize + 5) as i8,
            *s.as_ptr().offset(i as isize + 6) as i8,
            *s.as_ptr().offset(i as isize + 7) as i8,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
        ),
        7 => mm256_setr_epi8(
            *s.as_ptr().offset(i as isize + 0) as i8,
            *s.as_ptr().offset(i as isize + 1) as i8,
            *s.as_ptr().offset(i as isize + 2) as i8,
            *s.as_ptr().offset(i as isize + 3) as i8,
            *s.as_ptr().offset(i as isize + 4) as i8,
            *s.as_ptr().offset(i as isize + 5) as i8,
            *s.as_ptr().offset(i as isize + 6) as i8,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
        ),
        6 => mm256_setr_epi8(
            *s.as_ptr().offset(i as isize + 0) as i8,
            *s.as_ptr().offset(i as isize + 1) as i8,
            *s.as_ptr().offset(i as isize + 2) as i8,
            *s.as_ptr().offset(i as isize + 3) as i8,
            *s.as_ptr().offset(i as isize + 4) as i8,
            *s.as_ptr().offset(i as isize + 5) as i8,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
        ),
        5 => mm256_setr_epi8(
            *s.as_ptr().offset(i as isize + 0) as i8,
            *s.as_ptr().offset(i as isize + 1) as i8,
            *s.as_ptr().offset(i as isize + 2) as i8,
            *s.as_ptr().offset(i as isize + 3) as i8,
            *s.as_ptr().offset(i as isize + 4) as i8,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
        ),
        4 => mm256_setr_epi8(
            *s.as_ptr().offset(i as isize + 0) as i8,
            *s.as_ptr().offset(i as isize + 1) as i8,
            *s.as_ptr().offset(i as isize + 2) as i8,
            *s.as_ptr().offset(i as isize + 3) as i8,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
        ),
        3 => mm256_setr_epi8(
            *s.as_ptr().offset(i as isize + 0) as i8,
            *s.as_ptr().offset(i as isize + 1) as i8,
            *s.as_ptr().offset(i as isize + 2) as i8,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
        ),
        2 => mm256_setr_epi8(
            *s.as_ptr().offset(i as isize + 0) as i8,
            *s.as_ptr().offset(i as isize + 1) as i8,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
        ),
        1 => mm256_setr_epi8(
            *s.as_ptr().offset(i as isize + 0) as i8,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
        ),
        _ => mm256_setr_epi8(
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
            0,
        ),
    }
}

#[cfg(test)]
mod tests {
    use super::*;

    #[test]
    fn test_mm256i() {
        let test_cases = vec![0, 1, 2, 3];
        for i in test_cases {
            let want = mm256_setr_epi8(
                i,
                i,
                i,
                i,
                i,
                i,
                i,
                i,
                i,
                i,
                i,
                i,
                i,
                i,
                i,
                i,
                i,
                i,
                i,
                i,
                i,
                i,
                i,
                i,
                i,
                i,
                i,
                i,
                i,
                i,
                i,
                i,
            );
            let got = mm256i(i);
            assert_eq!(want.as_u8x32().as_array(), got.as_u8x32().as_array());
        }
    }
}