sha256: Slightly faster implementation

This commit is contained in:
quadbyte
2025-08-19 13:36:16 -04:00
parent b096f146de
commit 969aeeb5c9

View File

@@ -1,173 +1,250 @@
function sha256(message) {
// SHA-256 constants
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,
];
// Pre-computed SHA-256 constants (avoid array lookup overhead)
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 values
let h0 = 0x6a09e667;
let h1 = 0xbb67ae85;
let h2 = 0x3c6ef372;
let h3 = 0xa54ff53a;
let h4 = 0x510e527f;
let h5 = 0x9b05688c;
let h6 = 0x1f83d9ab;
let h7 = 0x5be0cd19;
// Initial hash values
let h0 = 0x6a09e667, h1 = 0xbb67ae85, h2 = 0x3c6ef372, h3 = 0xa54ff53a;
let h4 = 0x510e527f, h5 = 0x9b05688c, h6 = 0x1f83d9ab, h7 = 0x5be0cd19;
// Convert string to UTF-8 bytes manually
const msgBytes = stringToUtf8Bytes(message);
const msgLength = msgBytes.length;
const bitLength = msgLength * 8;
// Fast UTF-8 encoding for ASCII-only strings (common case)
let msgBytes;
let isAscii = true;
// Quick ASCII check
for (let i = 0; i < message.length; i++) {
if (message.charCodeAt(i) > 127) {
isAscii = false;
break;
}
}
if (isAscii) {
// Fast path for ASCII strings
msgBytes = new Array(message.length);
for (let i = 0; i < message.length; i++) {
msgBytes[i] = message.charCodeAt(i);
}
} else {
// Full UTF-8 encoding
msgBytes = stringToUtf8Bytes(message);
}
// Calculate padding
// Message + 1 bit (0x80) + padding zeros + 8 bytes for length = multiple of 64 bytes
const totalBitsNeeded = bitLength + 1 + 64; // message bits + padding bit + 64-bit length
const totalBytesNeeded = Math.ceil(totalBitsNeeded / 8);
const paddedLength = Math.ceil(totalBytesNeeded / 64) * 64; // Round up to multiple of 64
const msgLength = msgBytes.length;
const bitLength = msgLength * 8;
const paddedMsg = new Array(paddedLength).fill(0);
// Optimized padding calculation
const totalBitsNeeded = bitLength + 1 + 64;
const paddedLength = ((totalBitsNeeded + 511) >>> 9) << 6; // Faster than Math.ceil
// Pre-allocate padded message with exact size
const paddedMsg = new Array(paddedLength);
// Fast copy using simple loop (faster than copying one by one in some engines)
let i = 0;
while (i < msgLength) {
paddedMsg[i] = msgBytes[i];
i++;
}
// Fill remaining with zeros (explicit is often faster than fill)
while (i < paddedLength) {
paddedMsg[i] = 0;
i++;
}
// Add padding bit
paddedMsg[msgLength] = 0x80;
// Add length as 64-bit big-endian (optimized bit operations)
const highBits = (bitLength / 0x100000000) >>> 0;
const lowBits = bitLength >>> 0;
const lengthOffset = paddedLength - 8;
paddedMsg[lengthOffset] = highBits >>> 24;
paddedMsg[lengthOffset + 1] = (highBits >>> 16) & 0xFF;
paddedMsg[lengthOffset + 2] = (highBits >>> 8) & 0xFF;
paddedMsg[lengthOffset + 3] = highBits & 0xFF;
paddedMsg[lengthOffset + 4] = lowBits >>> 24;
paddedMsg[lengthOffset + 5] = (lowBits >>> 16) & 0xFF;
paddedMsg[lengthOffset + 6] = (lowBits >>> 8) & 0xFF;
paddedMsg[lengthOffset + 7] = lowBits & 0xFF;
// Copy original message
for (let i = 0; i < msgLength; i++) {
paddedMsg[i] = msgBytes[i];
}
// Add padding bit (0x80 = 10000000 in binary)
paddedMsg[msgLength] = 0x80;
// Add length as 64-bit big-endian integer at the end
// JavaScript numbers are not precise enough for 64-bit integers, so we handle high/low separately
const highBits = Math.floor(bitLength / 0x100000000);
const lowBits = bitLength % 0x100000000;
// Write 64-bit length in big-endian format
paddedMsg[paddedLength - 8] = (highBits >>> 24) & 0xff;
paddedMsg[paddedLength - 7] = (highBits >>> 16) & 0xff;
paddedMsg[paddedLength - 6] = (highBits >>> 8) & 0xff;
paddedMsg[paddedLength - 5] = highBits & 0xff;
paddedMsg[paddedLength - 4] = (lowBits >>> 24) & 0xff;
paddedMsg[paddedLength - 3] = (lowBits >>> 16) & 0xff;
paddedMsg[paddedLength - 2] = (lowBits >>> 8) & 0xff;
paddedMsg[paddedLength - 1] = lowBits & 0xff;
// Process message in 512-bit (64-byte) chunks
for (let chunk = 0; chunk < paddedLength; chunk += 64) {
// Pre-allocate working array (reused across chunks)
const w = new Array(64);
// Break chunk into sixteen 32-bit big-endian words
for (let i = 0; i < 16; i++) {
const offset = chunk + i * 4;
w[i] =
(paddedMsg[offset] << 24) |
(paddedMsg[offset + 1] << 16) |
(paddedMsg[offset + 2] << 8) |
paddedMsg[offset + 3];
// Ensure unsigned 32-bit
w[i] = w[i] >>> 0;
// Process message in 512-bit chunks
for (let chunk = 0; chunk < paddedLength; chunk += 64) {
// Unrolled word extraction for better performance
let offset = chunk;
w[0] = (paddedMsg[offset] << 24) | (paddedMsg[offset + 1] << 16) | (paddedMsg[offset + 2] << 8) | paddedMsg[offset + 3]; offset += 4;
w[1] = (paddedMsg[offset] << 24) | (paddedMsg[offset + 1] << 16) | (paddedMsg[offset + 2] << 8) | paddedMsg[offset + 3]; offset += 4;
w[2] = (paddedMsg[offset] << 24) | (paddedMsg[offset + 1] << 16) | (paddedMsg[offset + 2] << 8) | paddedMsg[offset + 3]; offset += 4;
w[3] = (paddedMsg[offset] << 24) | (paddedMsg[offset + 1] << 16) | (paddedMsg[offset + 2] << 8) | paddedMsg[offset + 3]; offset += 4;
w[4] = (paddedMsg[offset] << 24) | (paddedMsg[offset + 1] << 16) | (paddedMsg[offset + 2] << 8) | paddedMsg[offset + 3]; offset += 4;
w[5] = (paddedMsg[offset] << 24) | (paddedMsg[offset + 1] << 16) | (paddedMsg[offset + 2] << 8) | paddedMsg[offset + 3]; offset += 4;
w[6] = (paddedMsg[offset] << 24) | (paddedMsg[offset + 1] << 16) | (paddedMsg[offset + 2] << 8) | paddedMsg[offset + 3]; offset += 4;
w[7] = (paddedMsg[offset] << 24) | (paddedMsg[offset + 1] << 16) | (paddedMsg[offset + 2] << 8) | paddedMsg[offset + 3]; offset += 4;
w[8] = (paddedMsg[offset] << 24) | (paddedMsg[offset + 1] << 16) | (paddedMsg[offset + 2] << 8) | paddedMsg[offset + 3]; offset += 4;
w[9] = (paddedMsg[offset] << 24) | (paddedMsg[offset + 1] << 16) | (paddedMsg[offset + 2] << 8) | paddedMsg[offset + 3]; offset += 4;
w[10] = (paddedMsg[offset] << 24) | (paddedMsg[offset + 1] << 16) | (paddedMsg[offset + 2] << 8) | paddedMsg[offset + 3]; offset += 4;
w[11] = (paddedMsg[offset] << 24) | (paddedMsg[offset + 1] << 16) | (paddedMsg[offset + 2] << 8) | paddedMsg[offset + 3]; offset += 4;
w[12] = (paddedMsg[offset] << 24) | (paddedMsg[offset + 1] << 16) | (paddedMsg[offset + 2] << 8) | paddedMsg[offset + 3]; offset += 4;
w[13] = (paddedMsg[offset] << 24) | (paddedMsg[offset + 1] << 16) | (paddedMsg[offset + 2] << 8) | paddedMsg[offset + 3]; offset += 4;
w[14] = (paddedMsg[offset] << 24) | (paddedMsg[offset + 1] << 16) | (paddedMsg[offset + 2] << 8) | paddedMsg[offset + 3]; offset += 4;
w[15] = (paddedMsg[offset] << 24) | (paddedMsg[offset + 1] << 16) | (paddedMsg[offset + 2] << 8) | paddedMsg[offset + 3];
// Extend words (partially unrolled for better performance)
for (let i = 16; i < 64; i += 4) {
// Process 4 words at once
let s0 = rightRotate(w[i - 15], 7) ^ rightRotate(w[i - 15], 18) ^ (w[i - 15] >>> 3);
let s1 = rightRotate(w[i - 2], 17) ^ rightRotate(w[i - 2], 19) ^ (w[i - 2] >>> 10);
w[i] = (w[i - 16] + s0 + w[i - 7] + s1) >>> 0;
s0 = rightRotate(w[i - 14], 7) ^ rightRotate(w[i - 14], 18) ^ (w[i - 14] >>> 3);
s1 = rightRotate(w[i - 1], 17) ^ rightRotate(w[i - 1], 19) ^ (w[i - 1] >>> 10);
w[i + 1] = (w[i - 15] + s0 + w[i - 6] + s1) >>> 0;
s0 = rightRotate(w[i - 13], 7) ^ rightRotate(w[i - 13], 18) ^ (w[i - 13] >>> 3);
s1 = rightRotate(w[i], 17) ^ rightRotate(w[i], 19) ^ (w[i] >>> 10);
w[i + 2] = (w[i - 14] + s0 + w[i - 5] + s1) >>> 0;
s0 = rightRotate(w[i - 12], 7) ^ rightRotate(w[i - 12], 18) ^ (w[i - 12] >>> 3);
s1 = rightRotate(w[i + 1], 17) ^ rightRotate(w[i + 1], 19) ^ (w[i + 1] >>> 10);
w[i + 3] = (w[i - 13] + s0 + w[i - 4] + s1) >>> 0;
}
// Initialize working variables
let a = h0, b = h1, c = h2, d = h3, e = h4, f = h5, g = h6, h = h7;
// Main loop (partially unrolled)
for (let i = 0; i < 64; i += 8) {
// Round 1
let S1 = rightRotate(e, 6) ^ rightRotate(e, 11) ^ rightRotate(e, 25);
let ch = (e & f) ^ (~e & g);
let temp1 = (h + S1 + ch + K[i] + w[i]) >>> 0;
let S0 = rightRotate(a, 2) ^ rightRotate(a, 13) ^ rightRotate(a, 22);
let maj = (a & b) ^ (a & c) ^ (b & c);
let temp2 = (S0 + maj) >>> 0;
h = g; g = f; f = e; e = (d + temp1) >>> 0; d = c; c = b; b = a; a = (temp1 + temp2) >>> 0;
// Round 2
S1 = rightRotate(e, 6) ^ rightRotate(e, 11) ^ rightRotate(e, 25);
ch = (e & f) ^ (~e & g);
temp1 = (h + S1 + ch + K[i + 1] + w[i + 1]) >>> 0;
S0 = rightRotate(a, 2) ^ rightRotate(a, 13) ^ rightRotate(a, 22);
maj = (a & b) ^ (a & c) ^ (b & c);
temp2 = (S0 + maj) >>> 0;
h = g; g = f; f = e; e = (d + temp1) >>> 0; d = c; c = b; b = a; a = (temp1 + temp2) >>> 0;
// Round 3
S1 = rightRotate(e, 6) ^ rightRotate(e, 11) ^ rightRotate(e, 25);
ch = (e & f) ^ (~e & g);
temp1 = (h + S1 + ch + K[i + 2] + w[i + 2]) >>> 0;
S0 = rightRotate(a, 2) ^ rightRotate(a, 13) ^ rightRotate(a, 22);
maj = (a & b) ^ (a & c) ^ (b & c);
temp2 = (S0 + maj) >>> 0;
h = g; g = f; f = e; e = (d + temp1) >>> 0; d = c; c = b; b = a; a = (temp1 + temp2) >>> 0;
// Round 4
S1 = rightRotate(e, 6) ^ rightRotate(e, 11) ^ rightRotate(e, 25);
ch = (e & f) ^ (~e & g);
temp1 = (h + S1 + ch + K[i + 3] + w[i + 3]) >>> 0;
S0 = rightRotate(a, 2) ^ rightRotate(a, 13) ^ rightRotate(a, 22);
maj = (a & b) ^ (a & c) ^ (b & c);
temp2 = (S0 + maj) >>> 0;
h = g; g = f; f = e; e = (d + temp1) >>> 0; d = c; c = b; b = a; a = (temp1 + temp2) >>> 0;
// Round 5
S1 = rightRotate(e, 6) ^ rightRotate(e, 11) ^ rightRotate(e, 25);
ch = (e & f) ^ (~e & g);
temp1 = (h + S1 + ch + K[i + 4] + w[i + 4]) >>> 0;
S0 = rightRotate(a, 2) ^ rightRotate(a, 13) ^ rightRotate(a, 22);
maj = (a & b) ^ (a & c) ^ (b & c);
temp2 = (S0 + maj) >>> 0;
h = g; g = f; f = e; e = (d + temp1) >>> 0; d = c; c = b; b = a; a = (temp1 + temp2) >>> 0;
// Round 6
S1 = rightRotate(e, 6) ^ rightRotate(e, 11) ^ rightRotate(e, 25);
ch = (e & f) ^ (~e & g);
temp1 = (h + S1 + ch + K[i + 5] + w[i + 5]) >>> 0;
S0 = rightRotate(a, 2) ^ rightRotate(a, 13) ^ rightRotate(a, 22);
maj = (a & b) ^ (a & c) ^ (b & c);
temp2 = (S0 + maj) >>> 0;
h = g; g = f; f = e; e = (d + temp1) >>> 0; d = c; c = b; b = a; a = (temp1 + temp2) >>> 0;
// Round 7
S1 = rightRotate(e, 6) ^ rightRotate(e, 11) ^ rightRotate(e, 25);
ch = (e & f) ^ (~e & g);
temp1 = (h + S1 + ch + K[i + 6] + w[i + 6]) >>> 0;
S0 = rightRotate(a, 2) ^ rightRotate(a, 13) ^ rightRotate(a, 22);
maj = (a & b) ^ (a & c) ^ (b & c);
temp2 = (S0 + maj) >>> 0;
h = g; g = f; f = e; e = (d + temp1) >>> 0; d = c; c = b; b = a; a = (temp1 + temp2) >>> 0;
// Round 8
S1 = rightRotate(e, 6) ^ rightRotate(e, 11) ^ rightRotate(e, 25);
ch = (e & f) ^ (~e & g);
temp1 = (h + S1 + ch + K[i + 7] + w[i + 7]) >>> 0;
S0 = rightRotate(a, 2) ^ rightRotate(a, 13) ^ rightRotate(a, 22);
maj = (a & b) ^ (a & c) ^ (b & c);
temp2 = (S0 + maj) >>> 0;
h = g; g = f; f = e; e = (d + temp1) >>> 0; d = c; c = b; b = a; a = (temp1 + temp2) >>> 0;
}
// Add chunk's hash to result
h0 = (h0 + a) >>> 0; h1 = (h1 + b) >>> 0; h2 = (h2 + c) >>> 0; h3 = (h3 + d) >>> 0;
h4 = (h4 + e) >>> 0; h5 = (h5 + f) >>> 0; h6 = (h6 + g) >>> 0; h7 = (h7 + h) >>> 0;
}
// Extend the sixteen 32-bit words into sixty-four 32-bit words
for (let i = 16; i < 64; i++) {
const s0 =
rightRotate(w[i - 15], 7) ^
rightRotate(w[i - 15], 18) ^
(w[i - 15] >>> 3);
const s1 =
rightRotate(w[i - 2], 17) ^
rightRotate(w[i - 2], 19) ^
(w[i - 2] >>> 10);
w[i] = (w[i - 16] + s0 + w[i - 7] + s1) >>> 0;
// Fast hex conversion (pre-computed lookup table would be even faster)
const hex = '0123456789abcdef';
let result = '';
const hashes = [h0, h1, h2, h3, h4, h5, h6, h7];
for (let i = 0; i < 8; i++) {
const h = hashes[i];
result += hex[(h >>> 28) & 15] + hex[(h >>> 24) & 15] + hex[(h >>> 20) & 15] + hex[(h >>> 16) & 15] +
hex[(h >>> 12) & 15] + hex[(h >>> 8) & 15] + hex[(h >>> 4) & 15] + hex[h & 15];
}
// Initialize working variables for this chunk
let a = h0,
b = h1,
c = h2,
d = h3,
e = h4,
f = h5,
g = h6,
h = h7;
// Main loop
for (let i = 0; i < 64; i++) {
const S1 = rightRotate(e, 6) ^ rightRotate(e, 11) ^ rightRotate(e, 25);
const ch = (e & f) ^ (~e & g);
const temp1 = (h + S1 + ch + K[i] + w[i]) >>> 0;
const S0 = rightRotate(a, 2) ^ rightRotate(a, 13) ^ rightRotate(a, 22);
const maj = (a & b) ^ (a & c) ^ (b & c);
const temp2 = (S0 + maj) >>> 0;
h = g;
g = f;
f = e;
e = (d + temp1) >>> 0;
d = c;
c = b;
b = a;
a = (temp1 + temp2) >>> 0;
}
// Add this chunk's hash to result so far
h0 = (h0 + a) >>> 0;
h1 = (h1 + b) >>> 0;
h2 = (h2 + c) >>> 0;
h3 = (h3 + d) >>> 0;
h4 = (h4 + e) >>> 0;
h5 = (h5 + f) >>> 0;
h6 = (h6 + g) >>> 0;
h7 = (h7 + h) >>> 0;
}
// Produce the final hash value as a hex string
return [h0, h1, h2, h3, h4, h5, h6, h7]
.map((h) => h.toString(16).padStart(8, "0"))
.join("");
return result;
}
function stringToUtf8Bytes(str) {
const bytes = [];
for (let i = 0; i < str.length; i++) {
let code = str.charCodeAt(i);
if (code < 0x80) {
// 1-byte character (ASCII)
bytes.push(code);
} else if (code < 0x800) {
// 2-byte character
bytes.push(0xc0 | (code >> 6));
bytes.push(0x80 | (code & 0x3f));
} else if (code < 0xd800 || code > 0xdfff) {
// 3-byte character (not surrogate)
bytes.push(0xe0 | (code >> 12));
bytes.push(0x80 | ((code >> 6) & 0x3f));
bytes.push(0x80 | (code & 0x3f));
} else {
// 4-byte character (surrogate pair)
i++; // Move to next character
const code2 = str.charCodeAt(i);
const codePoint = 0x10000 + (((code & 0x3ff) << 10) | (code2 & 0x3ff));
bytes.push(0xf0 | (codePoint >> 18));
bytes.push(0x80 | ((codePoint >> 12) & 0x3f));
bytes.push(0x80 | ((codePoint >> 6) & 0x3f));
bytes.push(0x80 | (codePoint & 0x3f));
const bytes = [];
for (let i = 0; i < str.length; i++) {
let code = str.charCodeAt(i);
if (code < 0x80) {
bytes.push(code);
} else if (code < 0x800) {
bytes.push(0xC0 | (code >> 6), 0x80 | (code & 0x3F));
} else if (code < 0xD800 || code > 0xDFFF) {
bytes.push(0xE0 | (code >> 12), 0x80 | ((code >> 6) & 0x3F), 0x80 | (code & 0x3F));
} else {
i++;
const code2 = str.charCodeAt(i);
const codePoint = 0x10000 + (((code & 0x3FF) << 10) | (code2 & 0x3FF));
bytes.push(0xF0 | (codePoint >> 18), 0x80 | ((codePoint >> 12) & 0x3F),
0x80 | ((codePoint >> 6) & 0x3F), 0x80 | (codePoint & 0x3F));
}
}
}
return bytes;
return bytes;
}
// Inline right rotate function for better performance
function rightRotate(value, amount) {
return ((value >>> amount) | (value << (32 - amount))) >>> 0;
return ((value >>> amount) | (value << (32 - amount))) >>> 0;
}
// Test function to verify implementation