neyuki 0.6.3

-- SHA-256 written in Neyuki itself, as an exercise in the language's
-- arbitrary-precision integers and bit operators. `@neyuki/crypto` hashes
-- natively; this file checks its own answer against that.
local table = require("@neyuki/table")
local string = require("@neyuki/string")
local crypto = require("@neyuki/crypto")

const MASK32 = 0xffffffff

-- SHA-256 [FIPS 180-4]
-- Neyuki integers are arbitrary-precision, so every operation that can grow
-- past 32 bits (addition, left shift) is masked back down with `& MASK32`.

-- round constants [§4.2.2]
const K = {
    0x428a2f98, 0x71374491, 0xb5c0fbcf, 0xe9b5dba5, 0x3956c25b, 0x59f111f1, 0x923f82a4, 0xab1c5ed5,
    0xd807aa98, 0x12835b01, 0x243185be, 0x550c7dc3, 0x72be5d74, 0x80deb1fe, 0x9bdc06a7, 0xc19bf174,
    0xe49b69c1, 0xefbe4786, 0x0fc19dc6, 0x240ca1cc, 0x2de92c6f, 0x4a7484aa, 0x5cb0a9dc, 0x76f988da,
    0x983e5152, 0xa831c66d, 0xb00327c8, 0xbf597fc7, 0xc6e00bf3, 0xd5a79147, 0x06ca6351, 0x14292967,
    0x27b70a85, 0x2e1b2138, 0x4d2c6dfc, 0x53380d13, 0x650a7354, 0x766a0abb, 0x81c2c92e, 0x92722c85,
    0xa2bfe8a1, 0xa81a664b, 0xc24b8b70, 0xc76c51a3, 0xd192e819, 0xd6990624, 0xf40e3585, 0x106aa070,
    0x19a4c116, 0x1e376c08, 0x2748774c, 0x34b0bcb5, 0x391c0cb3, 0x4ed8aa4a, 0x5b9cca4f, 0x682e6ff3,
    0x748f82ee, 0x78a5636f, 0x84c87814, 0x8cc70208, 0x90befffa, 0xa4506ceb, 0xbef9a3f7, 0xc67178f2
}

-- initial hash value [§5.3.3]
const H0 = {
    0x6a09e667, 0xbb67ae85, 0x3c6ef372, 0xa54ff53a, 0x510e527f, 0x9b05688c, 0x1f83d9ab, 0x5be0cd19
}

function rotr(x: int, n: int): int
    return ((x >> n) | (x << (32 - n))) & MASK32
end

-- logical functions [§4.1.2]
function ch(x: int, y: int, z: int): int
    return (x & y) ~ ((x ~ MASK32) & z)
end

function maj(x: int, y: int, z: int): int
    return (x & y) ~ (x & z) ~ (y & z)
end

function bigSigma0(x: int): int
    return rotr(x, 2) ~ rotr(x, 13) ~ rotr(x, 22)
end

function bigSigma1(x: int): int
    return rotr(x, 6) ~ rotr(x, 11) ~ rotr(x, 25)
end

function smallSigma0(x: int): int
    return rotr(x, 7) ~ rotr(x, 18) ~ (x >> 3)
end

function smallSigma1(x: int): int
    return rotr(x, 17) ~ rotr(x, 19) ~ (x >> 10)
end

function sha256(msg: string): string
    -- PREPROCESSING [§5.1.1]: work on the UTF-8 bytes of msg; append 0x80,
    -- zero-pad to 56 mod 64 bytes, then the original length in bits as a
    -- 64-bit big-endian integer
    const bitLength = #msg * 8
    -- (appending via `bytes[#bytes + 1]` rather than table.insert, which
    -- rescans the table to find its length and would make this quadratic)
    local bytes = {}
    for offset = 1, #msg, 4096 do
        for _, byte in {string.byte(msg, offset, offset + 4095)} do
            bytes[#bytes + 1] = byte
        end
    end
    bytes[#bytes + 1] = 0x80
    while (#bytes % 64 != 56) do
        bytes[#bytes + 1] = 0
    end
    for shift = 56, 0, -8 do
        bytes[#bytes + 1] = (bitLength >> shift) & 0xff
    end

    local H = table.clone(H0)
    local W = {}

    -- HASH COMPUTATION [§6.2.2], one 64-byte block at a time
    for offset = 0, #bytes - 1, 64 do
        -- 1 - prepare message schedule W (big-endian words)
        for t = 1, 16 do
            const i = offset + (t - 1) * 4
            W[t] = (bytes[i + 1] << 24) | (bytes[i + 2] << 16) | (bytes[i + 3] << 8) | bytes[i + 4]
        end
        for t = 17, 64 do
            W[t] = (smallSigma1(W[t - 2]) + W[t - 7] + smallSigma0(W[t - 15]) + W[t - 16]) & MASK32
        end

        -- 2 - initialise working variables with the previous hash value
        local a, b, c, d, e, f, g, h = H[1], H[2], H[3], H[4], H[5], H[6], H[7], H[8]

        -- 3 - main loop
        for t = 1, 64 do
            const T1 = (h + bigSigma1(e) + ch(e, f, g) + K[t] + W[t]) & MASK32
            const T2 = (bigSigma0(a) + maj(a, b, c)) & MASK32
            h = g
            g = f
            f = e
            e = (d + T1) & MASK32
            d = c
            c = b
            b = a
            a = (T1 + T2) & MASK32
        end

        -- 4 - compute the new intermediate hash value
        H[1] = (H[1] + a) & MASK32
        H[2] = (H[2] + b) & MASK32
        H[3] = (H[3] + c) & MASK32
        H[4] = (H[4] + d) & MASK32
        H[5] = (H[5] + e) & MASK32
        H[6] = (H[6] + f) & MASK32
        H[7] = (H[7] + g) & MASK32
        H[8] = (H[8] + h) & MASK32
    end

    -- hex digest, 8 zero-padded digits per word
    local parts = {}
    for i, word in H do
        parts[i] = string.format("%08x", word)
    end
    return table.concat(parts, "")
end

for _, message in {"", "abc", "The quick brown fox jumps over the lazy dog"} do
    local ours = sha256(message)
    local native = crypto.hash(message, "sha256")
        local status = "MISMATCH"
    if (ours == native) then status = "ok" end
    print(string.format("%-45s %s %s", "\"" .. message .. "\"", ours, status))
end