@scure/base
Version:
Secure, audited & 0-dep implementation of base64, bech32, base58, base32 & base16
1,352 lines (1,291 loc) • 51.5 kB
text/typescript
/*! scure-base - MIT License (c) 2022 Paul Miller (paulmillr.com) */
/** Transforms values between two representations. */
export interface Coder<F, T> {
/**
* Converts a value from the input representation to the output representation.
* @param from - Value in the source representation.
* @returns Converted value.
*/
encode(from: F): T;
/**
* Converts a value from the output representation back to the input representation.
* @param to - Value in the target representation.
* @returns Converted value.
*/
decode(to: T): F;
}
/** Coder that works with byte arrays and strings. */
export interface BytesCoder extends Coder<Uint8Array, string> {
/**
* Encodes bytes into a string representation.
* @param data - Bytes to encode.
* @returns Encoded string.
*/
encode: (data: Uint8Array) => string;
/**
* Decodes a string representation into raw bytes.
* @param str - Encoded string.
* @returns Decoded bytes.
*/
decode: (str: string) => Uint8Array;
}
/**
* Bytes API type helpers for old + new TypeScript.
*
* TS 5.6 has `Uint8Array`, while TS 5.9+ made it generic `Uint8Array<ArrayBuffer>`.
* We can't use specific return type, because TS 5.6 will error.
* We can't use generic return type, because most TS 5.9 software will expect specific type.
*
* Maps typed-array input leaves to broad forms.
* These are compatibility adapters, not ownership guarantees.
*
* - `TArg` keeps byte inputs broad.
* - `TRet` marks byte outputs for TS 5.6 and TS 5.9+ compatibility.
*/
export type TypedArg<T> = T extends BigInt64Array
? BigInt64Array
: T extends BigUint64Array
? BigUint64Array
: T extends Float32Array
? Float32Array
: T extends Float64Array
? Float64Array
: T extends Int16Array
? Int16Array
: T extends Int32Array
? Int32Array
: T extends Int8Array
? Int8Array
: T extends Uint16Array
? Uint16Array
: T extends Uint32Array
? Uint32Array
: T extends Uint8ClampedArray
? Uint8ClampedArray
: T extends Uint8Array
? Uint8Array
: never;
/** Maps typed-array output leaves to narrow TS-compatible forms. */
export type TypedRet<T> = T extends BigInt64Array
? ReturnType<typeof BigInt64Array.of>
: T extends BigUint64Array
? ReturnType<typeof BigUint64Array.of>
: T extends Float32Array
? ReturnType<typeof Float32Array.of>
: T extends Float64Array
? ReturnType<typeof Float64Array.of>
: T extends Int16Array
? ReturnType<typeof Int16Array.of>
: T extends Int32Array
? ReturnType<typeof Int32Array.of>
: T extends Int8Array
? ReturnType<typeof Int8Array.of>
: T extends Uint16Array
? ReturnType<typeof Uint16Array.of>
: T extends Uint32Array
? ReturnType<typeof Uint32Array.of>
: T extends Uint8ClampedArray
? ReturnType<typeof Uint8ClampedArray.of>
: T extends Uint8Array
? ReturnType<typeof Uint8Array.of>
: never;
/** Recursively adapts byte-carrying API input types. See {@link TypedArg}. */
export type TArg<T> =
| T
| ([TypedArg<T>] extends [never]
? T extends (...args: infer A) => infer R
? ((...args: { [K in keyof A]: TRet<A[K]> }) => TArg<R>) & {
[K in keyof T]: T[K] extends (...args: any) => any ? T[K] : TArg<T[K]>;
}
: T extends [infer A, ...infer R]
? [TArg<A>, ...{ [K in keyof R]: TArg<R[K]> }]
: T extends readonly [infer A, ...infer R]
? readonly [TArg<A>, ...{ [K in keyof R]: TArg<R[K]> }]
: T extends (infer A)[]
? TArg<A>[]
: T extends readonly (infer A)[]
? readonly TArg<A>[]
: T extends Promise<infer A>
? Promise<TArg<A>>
: T extends object
? { [K in keyof T]: TArg<T[K]> }
: T
: TypedArg<T>);
/** Recursively adapts byte-carrying API output types. See {@link TypedArg}. */
export type TRet<T> = T extends unknown
? T &
([TypedRet<T>] extends [never]
? T extends (...args: infer A) => infer R
? ((...args: { [K in keyof A]: TArg<A[K]> }) => TRet<R>) & {
[K in keyof T]: T[K] extends (...args: any) => any ? T[K] : TRet<T[K]>;
}
: T extends [infer A, ...infer R]
? [TRet<A>, ...{ [K in keyof R]: TRet<R[K]> }]
: T extends readonly [infer A, ...infer R]
? readonly [TRet<A>, ...{ [K in keyof R]: TRet<R[K]> }]
: T extends (infer A)[]
? TRet<A>[]
: T extends readonly (infer A)[]
? readonly TRet<A>[]
: T extends Promise<infer A>
? Promise<TRet<A>>
: T extends object
? { [K in keyof T]: TRet<T[K]> }
: T
: TypedRet<T>)
: never;
function isBytes(a: unknown): a is Uint8Array {
// Plain `instanceof Uint8Array` is too strict for some Buffer / proxy / cross-realm cases. The
// fallback still requires a real ArrayBuffer view, so plain JSON-deserialized
// `{ constructor: ... }` spoofing is rejected. `BYTES_PER_ELEMENT === 1` keeps the
// fallback on byte-oriented views.
return (
a instanceof Uint8Array ||
(ArrayBuffer.isView(a) &&
a.constructor.name === 'Uint8Array' &&
'BYTES_PER_ELEMENT' in a &&
a.BYTES_PER_ELEMENT === 1)
);
}
/** Asserts something is Uint8Array. */
function abytes(b: TArg<Uint8Array | undefined>): void {
if (!isBytes(b)) throw new TypeError('Uint8Array expected');
}
function isArrayOf(isString: boolean, arr: any[]) {
if (!Array.isArray(arr)) return false;
if (arr.length === 0) return true;
if (isString) {
return arr.every((item) => typeof item === 'string');
} else {
return arr.every((item) => Number.isSafeInteger(item));
}
}
function afn(input: Function): input is Function {
if (typeof input !== 'function') throw new TypeError('function expected');
return true;
}
function astr(label: string, input: unknown): input is string {
if (typeof input !== 'string') throw new TypeError(`${label}: string expected`);
return true;
}
function anumber(n: number): void {
if (typeof n !== 'number') throw new TypeError(`number expected, got ${typeof n}`);
if (!Number.isSafeInteger(n)) throw new RangeError(`invalid integer: ${n}`);
}
function aArr(input: any[]) {
if (!Array.isArray(input)) throw new TypeError('array expected');
}
function astrArr(label: string, input: string[]) {
if (!isArrayOf(true, input)) throw new TypeError(`${label}: array of strings expected`);
}
function anumArr(label: string, input: number[]) {
if (!isArrayOf(false, input)) throw new TypeError(`${label}: array of numbers expected`);
}
// TODO: some recusive type inference so it would check correct order of input/output inside rest?
// like <string, number>, <number, bytes>, <bytes, float>
type Chain = [Coder<any, any>, ...Coder<any, any>[]];
// Extract info from Coder type
type Input<F> = F extends Coder<infer T, any> ? T : never;
type Output<F> = F extends Coder<any, infer T> ? T : never;
// Generic function for arrays
type First<T> = T extends [infer U, ...any[]] ? U : never;
type Last<T> = T extends [...any[], infer U] ? U : never;
type Tail<T> = T extends [any, ...infer U] ? U : never;
type AsChain<C extends Chain, Rest = Tail<C>> = {
// C[K] = Coder<Input<C[K]>, Input<Rest[k]>>
[K in keyof C]: Coder<Input<C[K]>, Input<K extends keyof Rest ? Rest[K] : any>>;
};
/**
* @__NO_SIDE_EFFECTS__
*/
function chain<T extends Chain & AsChain<T>>(...args: T): Coder<Input<First<T>>, Output<Last<T>>> {
const id = (a: any) => a;
// Wrap call in closure so JIT can inline calls
const wrap = (a: any, b: any) => (c: any) => a(b(c));
// Construct chain of args[-1].encode(args[-2].encode([...]))
const encode = args.map((x) => x.encode).reduceRight(wrap, id);
// Construct chain of args[0].decode(args[1].decode(...))
const decode = args.map((x) => x.decode).reduce(wrap, id);
return { encode, decode };
}
/**
* Encodes integer radix representation to array of strings using alphabet and back.
* Could also be array of strings.
* @__NO_SIDE_EFFECTS__
*/
function alphabet(letters: string | string[]): Coder<number[], string[]> {
// mapping 1 to "b"
const lettersA = typeof letters === 'string' ? letters.split('') : letters;
const len = lettersA.length;
astrArr('alphabet', lettersA);
// mapping "b" to 1
const indexes = new Map(lettersA.map((l, i) => [l, i]));
return {
encode: (digits: number[]) => {
aArr(digits);
return digits.map((i) => {
if (!Number.isSafeInteger(i) || i < 0 || i >= len)
throw new Error(
`alphabet.encode: digit index outside alphabet "${i}". Allowed: ${letters}`
);
return lettersA[i]!;
});
},
decode: (input: string[]): number[] => {
aArr(input);
return input.map((letter) => {
astr('alphabet.decode', letter);
const i = indexes.get(letter);
if (i === undefined) throw new Error(`Unknown letter: "${letter}". Allowed: ${letters}`);
return i;
});
},
};
}
/**
* @__NO_SIDE_EFFECTS__
*/
function join(separator = ''): Coder<string[], string> {
astr('join', separator);
// join('') is only lossless when each chunk is already unambiguous, such as single-symbol alphabets.
// Multi-character tokens need a separator that cannot appear inside the chunks.
return {
encode: (from) => {
astrArr('join.decode', from);
return from.join(separator);
},
decode: (to) => {
astr('join.decode', to);
return to.split(separator);
},
};
}
/**
* Pad strings array so it has integer number of bits
* @__NO_SIDE_EFFECTS__
*/
function padding(bits: number, chr = '='): Coder<string[], string[]> {
anumber(bits);
astr('padding', chr);
return {
encode(data: string[]): string[] {
astrArr('padding.encode', data);
// Mutates the intermediate token array in place while appending pad chars.
// utils.padding callers that need to preserve their input should pass a copy.
while ((data.length * bits) % 8) data.push(chr);
return data;
},
decode(input: string[]): string[] {
astrArr('padding.decode', input);
let end = input.length;
if ((end * bits) % 8)
throw new Error('padding: invalid, string should have whole number of bytes');
for (; end > 0 && input[end - 1] === chr; end--) {
const last = end - 1;
const byte = last * bits;
if (byte % 8 === 0) throw new Error('padding: invalid, string has too much padding');
}
return input.slice(0, end);
},
};
}
/**
* @__NO_SIDE_EFFECTS__
*/
function normalize<T>(fn: (val: T) => T): Coder<T, T> {
afn(fn);
return { encode: (from: T) => from, decode: (to: T) => fn(to) };
}
/**
* Slow: O(n^2) time complexity
*/
function convertRadix(data: number[], from: number, to: number): number[] {
// base 1 is impossible
if (from < 2)
throw new RangeError(`convertRadix: invalid from=${from}, base cannot be less than 2`);
if (to < 2) throw new RangeError(`convertRadix: invalid to=${to}, base cannot be less than 2`);
aArr(data);
if (!data.length) return [];
let pos = 0;
const res = [];
const digits = Array.from(data, (d) => {
anumber(d);
if (d < 0 || d >= from) throw new Error(`invalid integer: ${d}`);
return d;
});
const dlen = digits.length;
while (true) {
let carry = 0;
let done = true;
for (let i = pos; i < dlen; i++) {
const digit = digits[i]!;
const fromCarry = from * carry;
const digitBase = fromCarry + digit;
if (
!Number.isSafeInteger(digitBase) ||
fromCarry / from !== carry ||
digitBase - digit !== fromCarry
) {
throw new Error('convertRadix: carry overflow');
}
const div = digitBase / to;
carry = digitBase % to;
const rounded = Math.floor(div);
digits[i] = rounded;
if (!Number.isSafeInteger(rounded) || rounded * to + carry !== digitBase)
throw new Error('convertRadix: carry overflow');
if (!done) continue;
else if (!rounded) pos = i;
else done = false;
}
res.push(carry);
if (done) break;
}
// Preserve explicit leading zero digits so callers like base58 keep zero-prefix semantics.
for (let i = 0; i < data.length - 1 && data[i] === 0; i++) res.push(0);
return res.reverse();
}
const gcd = (a: number, b: number): number => (b === 0 ? a : gcd(b, a % b));
// Maximum carry width before the `pos` cycle repeats.
// Residues advance in gcd(from, to) steps, so the largest pre-drain width is from + (to - gcd).
const radix2carry = /* @__NO_SIDE_EFFECTS__ */ (from: number, to: number) =>
from + (to - gcd(from, to));
const powers: number[] = /* @__PURE__ */ (() => {
let res = [];
for (let i = 0; i < 40; i++) res.push(2 ** i);
return res;
})();
/**
* Implemented with numbers, because BigInt is 5x slower
*/
function convertRadix2(data: number[], from: number, to: number, padding: boolean): number[] {
aArr(data);
if (from <= 0 || from > 32) throw new RangeError(`convertRadix2: wrong from=${from}`);
if (to <= 0 || to > 32) throw new RangeError(`convertRadix2: wrong to=${to}`);
if (radix2carry(from, to) > 32) {
throw new Error(
`convertRadix2: carry overflow from=${from} to=${to} carryBits=${radix2carry(from, to)}`
);
}
let carry = 0;
let pos = 0; // bitwise position in current element
const max = powers[from]!;
const mask = powers[to]! - 1;
const res: number[] = [];
for (const n of data) {
anumber(n);
if (n >= max) throw new Error(`convertRadix2: invalid data word=${n} from=${from}`);
carry = (carry << from) | n;
if (pos + from > 32) throw new Error(`convertRadix2: carry overflow pos=${pos} from=${from}`);
pos += from;
for (; pos >= to; pos -= to) res.push(((carry >> (pos - to)) & mask) >>> 0);
const pow = powers[pos];
if (pow === undefined) throw new Error('invalid carry');
carry &= pow - 1; // clean carry, otherwise it will cause overflow
}
carry = (carry << (to - pos)) & mask;
// Canonical decode paths reject leftover whole input words and non-zero pad bits.
// For Bech32 5->8 regrouping, this is the "4 bits or less, all zeroes" tail rule.
if (!padding && pos >= from) throw new Error('Excess padding');
if (!padding && carry > 0) throw new Error(`Non-zero padding: ${carry}`);
if (padding && pos > 0) res.push(carry >>> 0);
return res;
}
/**
* @__NO_SIDE_EFFECTS__
*/
function radix(num: number): TRet<Coder<Uint8Array, number[]>> {
anumber(num);
const _256 = 2 ** 8;
// Base-range and carry-overflow checks live in convertRadix so encode/decode reject unsupported bases symmetrically.
return {
encode: (bytes: TArg<Uint8Array>) => {
if (!isBytes(bytes)) throw new TypeError('radix.encode input should be Uint8Array');
return convertRadix(Array.from(bytes), _256, num);
},
decode: (digits: number[]) => {
anumArr('radix.decode', digits);
return Uint8Array.from(convertRadix(digits, num, _256));
},
};
}
/**
* If both bases are power of same number (like `2**8 <-> 2**64`),
* there is a linear algorithm. For now we have implementation for power-of-two bases only.
* @__NO_SIDE_EFFECTS__
*/
function radix2(bits: number, revPadding = false): TRet<Coder<Uint8Array, number[]>> {
anumber(bits);
if (bits <= 0 || bits > 32) throw new RangeError('radix2: bits should be in (0..32]');
if (radix2carry(8, bits) > 32 || radix2carry(bits, 8) > 32)
throw new RangeError('radix2: carry overflow');
// revPadding flips which direction allows a partial zero tail.
// Default pads 8->bits and rejects extra bits on bits->8; `true` does the opposite.
return {
encode: (bytes: TArg<Uint8Array>) => {
if (!isBytes(bytes)) throw new TypeError('radix2.encode input should be Uint8Array');
return convertRadix2(Array.from(bytes), 8, bits, !revPadding);
},
decode: (digits: number[]) => {
anumArr('radix2.decode', digits);
return Uint8Array.from(convertRadix2(digits, bits, 8, revPadding));
},
};
}
type ArgumentTypes<F extends Function> = F extends (...args: infer A) => any ? A : never;
type BytesFn = (data: TArg<Uint8Array>) => TRet<Uint8Array>;
function unsafeWrapper<T extends (...args: any) => any>(fn: T) {
afn(fn);
return function (...args: ArgumentTypes<T>): ReturnType<T> | void {
// Only for *Unsafe APIs that intentionally collapse validation failures to `undefined`.
// Do not wrap code that needs to preserve exception details.
try {
return fn.apply(null, args);
} catch (e) {}
};
}
function checksum(len: number, fn: TArg<BytesFn>): TRet<Coder<Uint8Array, Uint8Array>> {
anumber(len);
// Reject degenerate zero-byte checksums up front so callers don't accidentally
// build a no-op checksum stage.
if (len <= 0) throw new RangeError(`checksum length must be positive: ${len}`);
afn(fn);
const _fn = fn as BytesFn;
// Uses the first `len` bytes of fn(data) in both directions.
// Current call sites rely on `len > 0` and checksum functions that return at least that many bytes.
return {
encode(data: TArg<Uint8Array>) {
if (!isBytes(data)) throw new TypeError('checksum.encode: input should be Uint8Array');
const sum = _fn(data).slice(0, len);
const res = new Uint8Array(data.length + len);
res.set(data);
res.set(sum, data.length);
return res;
},
decode(data: TArg<Uint8Array>) {
if (!isBytes(data)) throw new TypeError('checksum.decode: input should be Uint8Array');
const payload = data.slice(0, -len);
const oldChecksum = data.slice(-len);
const newChecksum = _fn(payload).slice(0, len);
for (let i = 0; i < len; i++)
if (newChecksum[i] !== oldChecksum[i]) throw new Error('Invalid checksum');
return payload;
},
};
}
// prettier-ignore
/**
* Low-level building blocks used by the exported codecs.
* @example
* Build a radix-32 coder from the low-level helpers.
* ```ts
* import { utils } from '@scure/base';
* utils.radix2(5).encode(Uint8Array.from([1, 2, 3]));
* ```
*/
export const utils: { alphabet: typeof alphabet; chain: typeof chain; checksum: typeof checksum; convertRadix: typeof convertRadix; convertRadix2: typeof convertRadix2; radix: typeof radix; radix2: typeof radix2; join: typeof join; padding: typeof padding; } = /* @__PURE__ */ Object.freeze({
alphabet, chain, checksum, convertRadix, convertRadix2, radix, radix2, join, padding,
});
// RFC 4648 aka RFC 3548
// ---------------------
/**
* base16 encoding from RFC 4648.
* This codec uses RFC 4648 Table 5's uppercase alphabet directly.
* RFC 4648 §8 calls base16 "case-insensitive hex encoding", but we intentionally do not case-fold decode input here.
* Use `hex` for case-insensitive hex decoding.
* @example
* ```js
* base16.encode(Uint8Array.from([0x12, 0xab]));
* // => '12AB'
* ```
*/
export const base16: BytesCoder = /* @__PURE__ */ Object.freeze(
chain(radix2(4), alphabet('0123456789ABCDEF'), join(''))
);
/**
* base32 encoding from RFC 4648. Has padding.
* RFC 4648 §6 Table 3 uses uppercase letters, and RFC 4648 §3.4 allows applications to choose
* upper- or lowercase alphabets. We keep the published uppercase table and do not case-fold decode input.
* Use `base32nopad` for unpadded version.
* Also check out `base32hex`, `base32hexnopad`, `base32crockford`.
* @example
* ```js
* base32.encode(Uint8Array.from([0x12, 0xab]));
* // => 'CKVQ===='
* base32.decode('CKVQ====');
* // => Uint8Array.from([0x12, 0xab])
* ```
*/
export const base32: BytesCoder = /* @__PURE__ */ Object.freeze(
chain(radix2(5), alphabet('ABCDEFGHIJKLMNOPQRSTUVWXYZ234567'), padding(5), join(''))
);
/**
* base32 encoding from RFC 4648. No padding.
* This variant inherits RFC 4648 base32's uppercase table and intentionally does not case-fold decode input.
* Use `base32` for padded version.
* Also check out `base32hex`, `base32hexnopad`, `base32crockford`.
* @example
* ```js
* base32nopad.encode(Uint8Array.from([0x12, 0xab]));
* // => 'CKVQ'
* base32nopad.decode('CKVQ');
* // => Uint8Array.from([0x12, 0xab])
* ```
*/
export const base32nopad: BytesCoder = /* @__PURE__ */ Object.freeze(
chain(radix2(5), alphabet('ABCDEFGHIJKLMNOPQRSTUVWXYZ234567'), join(''))
);
/**
* base32 encoding from RFC 4648. Padded. Compared to ordinary `base32`, slightly different alphabet.
* RFC 4648 §7 Table 4 uses uppercase letters, and we intentionally keep that table without case-folding decode input.
* Use `base32hexnopad` for unpadded version.
* @example
* ```js
* base32hex.encode(Uint8Array.from([0x12, 0xab]));
* // => '2ALG===='
* base32hex.decode('2ALG====');
* // => Uint8Array.from([0x12, 0xab])
* ```
*/
export const base32hex: BytesCoder = /* @__PURE__ */ Object.freeze(
chain(radix2(5), alphabet('0123456789ABCDEFGHIJKLMNOPQRSTUV'), padding(5), join(''))
);
/**
* base32 encoding from RFC 4648. No padding. Compared to ordinary `base32`, slightly different alphabet.
* This variant inherits RFC 4648 base32hex's uppercase table and intentionally does not case-fold decode input.
* Use `base32hex` for padded version.
* @example
* ```js
* base32hexnopad.encode(Uint8Array.from([0x12, 0xab]));
* // => '2ALG'
* base32hexnopad.decode('2ALG');
* // => Uint8Array.from([0x12, 0xab])
* ```
*/
export const base32hexnopad: BytesCoder = /* @__PURE__ */ Object.freeze(
chain(radix2(5), alphabet('0123456789ABCDEFGHIJKLMNOPQRSTUV'), join(''))
);
/**
* base32 encoding from RFC 4648. Doug Crockford's version.
* See {@link https://www.crockford.com/base32.html | Douglas Crockford's Base32}.
* @example
* ```js
* base32crockford.encode(Uint8Array.from([0x12, 0xab]));
* // => '2ANG'
* base32crockford.decode('2ANG');
* // => Uint8Array.from([0x12, 0xab])
* ```
*/
export const base32crockford: BytesCoder = /* @__PURE__ */ Object.freeze(
chain(
radix2(5),
alphabet('0123456789ABCDEFGHJKMNPQRSTVWXYZ'),
join(''),
normalize((s: string) => s.toUpperCase().replace(/O/g, '0').replace(/[IL]/g, '1'))
)
);
// Built-in base64 conversion https://caniuse.com/mdn-javascript_builtins_uint8array_frombase64
// Require both directions before taking the native fast path, so base64/base64url don't mix native and JS behavior.
// prettier-ignore
const hasBase64Builtin: boolean = /* @__PURE__ */ (() =>
typeof (Uint8Array as any).from([]).toBase64 === 'function' &&
typeof (Uint8Array as any).fromBase64 === 'function')();
// Native `Uint8Array.fromBase64()` accepts these ASCII whitespace chars.
// Reject them first so the native base64 path still follows RFC 4648 §3.3.
// ASCII whitespace is U+0009 TAB, U+000A LF, U+000C FF, U+000D CR, or U+0020 SPACE
const ASCII_WHITESPACE = /[\t\n\f\r ]/;
const decodeBase64Builtin = (s: string, isUrl: boolean) => {
astr('base64', s);
const alphabet = isUrl ? 'base64url' : 'base64';
// Per spec, .fromBase64 already throws on any other non-alphabet symbols except ASCII whitespace
// And checking just for whitespace makes decoding about 3x faster than a full range check.
// lastChunkHandling: 'strict' rejects loose tails and non-zero pad bits so native decoding stays canonical.
if (s.length > 0 && ASCII_WHITESPACE.test(s)) throw new Error('invalid base64');
return (Uint8Array as any).fromBase64(s, { alphabet, lastChunkHandling: 'strict' });
};
/**
* base64 from RFC 4648. Padded.
* Use `base64nopad` for unpadded version.
* Also check out `base64url`, `base64urlnopad`.
* Falls back to built-in function, when available.
* @example
* ```js
* base64.encode(Uint8Array.from([0x12, 0xab]));
* // => 'Eqs='
* base64.decode('Eqs=');
* // => Uint8Array.from([0x12, 0xab])
* ```
*/
// prettier-ignore
export const base64: BytesCoder = /* @__PURE__ */ Object.freeze(hasBase64Builtin ? {
encode(b) { abytes(b); return (b as any).toBase64(); },
decode(s) { return decodeBase64Builtin(s, false); },
} : chain(
radix2(6),
alphabet('ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789+/'),
padding(6),
join('')
));
/**
* base64 from RFC 4648. No padding.
* Use `base64` for padded version.
* @example
* ```js
* base64nopad.encode(Uint8Array.from([0x12, 0xab]));
* // => 'Eqs'
* base64nopad.decode('Eqs');
* // => Uint8Array.from([0x12, 0xab])
* ```
*/
export const base64nopad: BytesCoder = /* @__PURE__ */ Object.freeze(
chain(
radix2(6),
alphabet('ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789+/'),
join('')
)
);
/**
* base64 from RFC 4648, using URL-safe alphabet. Padded.
* Use `base64urlnopad` for unpadded version.
* Falls back to built-in function, when available.
* @example
* ```js
* base64url.encode(Uint8Array.from([0x12, 0xab]));
* // => 'Eqs='
* base64url.decode('Eqs=');
* // => Uint8Array.from([0x12, 0xab])
* ```
*/
// prettier-ignore
export const base64url: BytesCoder = /* @__PURE__ */ Object.freeze(hasBase64Builtin ? {
encode(b) { abytes(b); return (b as any).toBase64({ alphabet: 'base64url' }); },
decode(s) { return decodeBase64Builtin(s, true); },
} : chain(
radix2(6),
alphabet('ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789-_'),
padding(6),
join('')
));
/**
* base64 from RFC 4648, using URL-safe alphabet. No padding.
* Use `base64url` for padded version.
* @example
* ```js
* base64urlnopad.encode(Uint8Array.from([0x12, 0xab]));
* // => 'Eqs'
* base64urlnopad.decode('Eqs');
* // => Uint8Array.from([0x12, 0xab])
* ```
*/
export const base64urlnopad: BytesCoder = /* @__PURE__ */ Object.freeze(
chain(
radix2(6),
alphabet('ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789-_'),
join('')
)
);
// base58 code
// -----------
const genBase58 = /* @__NO_SIDE_EFFECTS__ */ (abc: string) =>
chain(radix(58), alphabet(abc), join(''));
/**
* base58: base64 without ambigous characters +, /, 0, O, I, l.
* Quadratic (O(n^2)) - so, can't be used on large inputs.
* @example
* ```js
* const text = base58.encode(Uint8Array.from([0, 1, 2]));
* base58.decode(text);
* // => Uint8Array.from([0, 1, 2])
* ```
*/
export const base58: BytesCoder = /* @__PURE__ */ Object.freeze(
genBase58('123456789ABCDEFGHJKLMNPQRSTUVWXYZabcdefghijkmnopqrstuvwxyz')
);
/**
* base58: flickr version. Check out `base58`.
* @example
* Round-trip bytes with the Flickr alphabet.
* ```ts
* const text = base58flickr.encode(Uint8Array.from([0, 1, 2]));
* base58flickr.decode(text);
* ```
*/
export const base58flickr: BytesCoder = /* @__PURE__ */ Object.freeze(
genBase58('123456789abcdefghijkmnopqrstuvwxyzABCDEFGHJKLMNPQRSTUVWXYZ')
);
/**
* base58: XRP version. Check out `base58`.
* @example
* Round-trip bytes with the XRP alphabet.
* ```ts
* const text = base58xrp.encode(Uint8Array.from([0, 1, 2]));
* base58xrp.decode(text);
* ```
*/
export const base58xrp: BytesCoder = /* @__PURE__ */ Object.freeze(
genBase58('rpshnaf39wBUDNEGHJKLM4PQRST7VWXYZ2bcdeCg65jkm8oFqi1tuvAxyz')
);
// Data len (index) -> encoded block len.
// Monero pads each 1..8-byte block to this fixed base58 width so decode can recover the tail length.
const XMR_BLOCK_LEN = [0, 2, 3, 5, 6, 7, 9, 10, 11];
/**
* base58: XMR version. Check out `base58`.
* Done in 8-byte blocks (which equals 11 chars in decoding). Last (non-full) block padded with '1' to size in XMR_BLOCK_LEN.
* Block encoding significantly reduces quadratic complexity of base58.
* @example
* Round-trip bytes with the Monero block codec.
* ```ts
* const text = base58xmr.encode(Uint8Array.from([0, 1, 2]));
* base58xmr.decode(text);
* ```
*/
export const base58xmr: BytesCoder = /* @__PURE__ */ Object.freeze({
encode(data: TArg<Uint8Array>) {
abytes(data);
let res = '';
for (let i = 0; i < data.length; i += 8) {
const block = data.subarray(i, i + 8);
res += base58.encode(block).padStart(XMR_BLOCK_LEN[block.length]!, '1');
}
return res;
},
decode(str: string) {
astr('base58xmr.decode', str);
let res: number[] = [];
for (let i = 0; i < str.length; i += 11) {
const slice = str.slice(i, i + 11);
const blockLen = XMR_BLOCK_LEN.indexOf(slice.length);
const block = base58.decode(slice);
for (let j = 0; j < block.length - blockLen; j++) {
if (block[j] !== 0) throw new Error('base58xmr: wrong padding');
}
res = res.concat(Array.from(block.slice(block.length - blockLen)));
}
return Uint8Array.from(res);
},
});
/**
* Method, which creates base58check encoder.
* Requires function, calculating sha256.
* Callers must include any version bytes in `data`; this helper only applies the
* 4-byte double-SHA256 checksum used by Bitcoin Base58Check.
* @param sha256 - Function used to calculate the checksum hash.
* @returns base58check codec using 4 checksum bytes.
* @throws On wrong argument types. {@link TypeError}
* @example
* Create a base58check codec from a SHA-256 implementation.
* ```ts
* import { createBase58check } from '@scure/base';
* import { sha256 } from '@noble/hashes/sha2.js';
* const coder = createBase58check(sha256);
* coder.encode(Uint8Array.from([1, 2, 3]));
* ```
*/
export const createBase58check = (sha256: TArg<BytesFn>): BytesCoder => {
// Validate the hash function at construction time so wrong inputs fail before returning a coder.
afn(sha256);
const _sha256 = sha256 as BytesFn;
return chain(
checksum(4, (data: TArg<Uint8Array>) => _sha256(_sha256(data))),
base58
);
};
/**
* Use `createBase58check` instead.
* @deprecated Use {@link createBase58check} instead.
* Callers must include any version bytes in `data`; this alias keeps the same
* 4-byte double-SHA256 checksum behavior as `createBase58check`.
* @param sha256 - Function used to calculate the checksum hash.
* @returns base58check codec using 4 checksum bytes.
* @example
* Create a base58check codec with the deprecated alias.
* ```ts
* import { base58check } from '@scure/base';
* import { sha256 } from '@noble/hashes/sha2.js';
* const coder = base58check(sha256);
* coder.encode(Uint8Array.from([1, 2, 3]));
* ```
*/
export const base58check: (sha256: TArg<BytesFn>) => BytesCoder = createBase58check;
// Bech32 code
// -----------
/** Result of bech32 decoding. */
export interface Bech32Decoded<Prefix extends string = string> {
/** Human-readable bech32 prefix. */
prefix: Prefix;
/** Decoded 5-bit word payload. */
words: number[];
}
/** Result of bech32 decoding with original bytes attached. */
export interface Bech32DecodedWithArray<Prefix extends string = string> {
/** Human-readable bech32 prefix. */
prefix: Prefix;
/** Decoded 5-bit word payload. */
words: number[];
/** Decoded payload converted back into raw bytes. */
bytes: Uint8Array;
}
// BIP 173 character table: data values 0..31 map to `qpzry9x8gf2tvdw0s3jn54khce6mua7l`.
const BECH_ALPHABET: Coder<number[], string> = chain(
alphabet('qpzry9x8gf2tvdw0s3jn54khce6mua7l'),
join('')
);
// BIP 173 `bech32_polymod` GEN coefficients.
const POLYMOD_GENERATORS = [0x3b6a57b2, 0x26508e6d, 0x1ea119fa, 0x3d4233dd, 0x2a1462b3];
// BIP 173 step split: this applies the polymod state transition before callers xor in the next 5-bit value.
function bech32Polymod(pre: number): number {
const b = pre >> 25;
let chk = (pre & 0x1ffffff) << 5;
for (let i = 0; i < POLYMOD_GENERATORS.length; i++) {
if (((b >> i) & 1) === 1) chk ^= POLYMOD_GENERATORS[i]!;
}
return chk;
}
function bechChecksum(prefix: string, words: number[], encodingConst = 1): string {
const len = prefix.length;
let chk = 1;
for (let i = 0; i < len; i++) {
const c = prefix.charCodeAt(i);
if (c < 33 || c > 126) throw new Error(`Invalid prefix (${prefix})`);
chk = bech32Polymod(chk) ^ (c >> 5);
}
chk = bech32Polymod(chk);
for (let i = 0; i < len; i++) chk = bech32Polymod(chk) ^ (prefix.charCodeAt(i) & 0x1f);
for (let v of words) chk = bech32Polymod(chk) ^ v;
for (let i = 0; i < 6; i++) chk = bech32Polymod(chk);
// BIP 173/BIP 350: xor the final checksum constant, then emit the 30-bit state as six 5-bit symbols.
chk ^= encodingConst;
return BECH_ALPHABET.encode(convertRadix2([chk % powers[30]!], 30, 5, false));
}
/** bech32 codec surface. */
export interface Bech32 {
/**
* Encodes a human-readable prefix and 5-bit words into a bech32 string.
* @param prefix - Human-readable prefix.
* @param words - 5-bit words or raw bytes.
* @param limit - Maximum accepted output length, or `false` to disable the limit.
* @returns Encoded bech32 string.
*/
encode<Prefix extends string>(
prefix: Prefix,
words: number[] | Uint8Array,
limit?: number | false
): `${Lowercase<Prefix>}1${string}`;
/**
* Decodes a bech32 string into prefix and words.
* @param str - Encoded bech32 string.
* @param limit - Maximum accepted input length, or `false` to disable the limit.
* @returns Decoded prefix and 5-bit words.
*/
decode<Prefix extends string>(
str: `${Prefix}1${string}`,
limit?: number | false
): Bech32Decoded<Prefix>;
decode(str: string, limit?: number | false): Bech32Decoded;
/**
* Encodes raw bytes by first converting them to 5-bit words.
* @param prefix - Human-readable prefix.
* @param bytes - Raw bytes to encode.
* @returns Encoded bech32 string.
*/
encodeFromBytes(prefix: string, bytes: Uint8Array): string;
/**
* Decodes a bech32 string and converts the payload back into bytes.
* @param str - Encoded bech32 string.
* @returns Decoded prefix, words, and bytes.
*/
decodeToBytes(str: string): Bech32DecodedWithArray;
/**
* Decodes a bech32 string, returning `undefined` instead of throwing on invalid input.
* @param str - Encoded bech32 string.
* @param limit - Maximum accepted input length, or `false` to disable the limit.
* @returns Decoded prefix and words, or `undefined` for invalid input.
*/
decodeUnsafe(str: string, limit?: number | false): void | Bech32Decoded<string>;
/**
* Converts 5-bit words back into raw bytes.
* @param to - 5-bit words to decode.
* @returns Decoded bytes.
*/
fromWords(to: number[]): Uint8Array;
/**
* Converts 5-bit words back into raw bytes, returning `undefined` instead of throwing.
* @param to - 5-bit words to decode.
* @returns Decoded bytes, or `undefined` for invalid input.
*/
fromWordsUnsafe(to: number[]): void | Uint8Array;
/**
* Converts raw bytes into 5-bit words for bech32 encoding.
* @param from - Raw bytes to convert.
* @returns 5-bit words.
*/
toWords(from: Uint8Array): number[];
}
/**
* @__NO_SIDE_EFFECTS__
*/
function genBech32(encoding: 'bech32' | 'bech32m'): TRet<Bech32> {
// BIP 173 uses final xor constant 1; BIP 350 swaps in 0x2bc830a3 for Bech32m.
const ENCODING_CONST = encoding === 'bech32' ? 1 : 0x2bc830a3;
const _words = radix2(5);
const fromWords = _words.decode;
const toWords = _words.encode;
const fromWordsUnsafe = unsafeWrapper(fromWords);
function encode<Prefix extends string>(
prefix: Prefix,
words: TArg<number[] | Uint8Array>,
limit: number | false = 90
): `${Lowercase<Prefix>}1${string}` {
astr('bech32.encode prefix', prefix);
if (isBytes(words)) words = Array.from(words);
anumArr('bech32.encode', words);
const plen = prefix.length;
if (plen === 0) throw new TypeError(`Invalid prefix length ${plen}`);
// Total output is hrp + `1` separator + payload words + 6 checksum chars.
const actualLength = plen + 7 + words.length;
if (limit !== false && actualLength > limit)
throw new TypeError(`Length ${actualLength} exceeds limit ${limit}`);
const lowered = prefix.toLowerCase();
const sum = bechChecksum(lowered, words, ENCODING_CONST);
return `${lowered}1${BECH_ALPHABET.encode(words)}${sum}` as `${Lowercase<Prefix>}1${string}`;
}
function decode<Prefix extends string>(
str: `${Prefix}1${string}`,
limit?: number | false
): Bech32Decoded<Prefix>;
function decode(str: string, limit?: number | false): Bech32Decoded;
function decode(str: string, limit: number | false = 90): Bech32Decoded {
astr('bech32.decode input', str);
const slen = str.length;
// Minimum length is 1-char hrp + `1` separator + 6-char checksum.
if (slen < 8 || (limit !== false && slen > limit))
throw new TypeError(`invalid string length: ${slen} (${str}). Expected (8..${limit})`);
// don't allow mixed case
const lowered = str.toLowerCase();
if (str !== lowered && str !== str.toUpperCase())
throw new Error(`String must be lowercase or uppercase`);
const sepIndex = lowered.lastIndexOf('1');
if (sepIndex === 0 || sepIndex === -1)
throw new Error(`Letter "1" must be present between prefix and data only`);
const prefix = lowered.slice(0, sepIndex);
const data = lowered.slice(sepIndex + 1);
if (data.length < 6) throw new Error('Data must be at least 6 characters long');
const words = BECH_ALPHABET.decode(data).slice(0, -6);
const sum = bechChecksum(prefix, words, ENCODING_CONST);
if (!data.endsWith(sum)) throw new Error(`Invalid checksum in ${str}: expected "${sum}"`);
return { prefix, words };
}
const decodeUnsafe = unsafeWrapper(decode);
function decodeToBytes(str: string): TRet<Bech32DecodedWithArray> {
// Keep the byte helper unbounded; callers that need the default BIP 173 length cap should use decode(str).
const { prefix, words } = decode(str, false);
return {
prefix,
words,
bytes: fromWords(words) as TRet<Uint8Array>,
} as TRet<Bech32DecodedWithArray>;
}
function encodeFromBytes(prefix: string, bytes: TArg<Uint8Array>) {
// Keep the convenience wrapper on encode()'s default 90-char cap; custom limits should call encode(prefix, toWords(bytes), limit).
return encode(prefix, toWords(bytes));
}
return {
encode,
decode,
encodeFromBytes,
decodeToBytes,
decodeUnsafe,
fromWords,
fromWordsUnsafe,
toWords,
};
}
/**
* bech32 from BIP 173. Operates on words.
* For high-level helpers, check out {@link https://github.com/paulmillr/scure-btc-signer | scure-btc-signer}.
* @example
* Convert bytes to words, encode them, then decode back.
* ```ts
* const words = bech32.toWords(Uint8Array.from([1, 2, 3]));
* const text = bech32.encode('bc', words);
* bech32.decode(text);
* ```
*/
export const bech32: TRet<Bech32> = /* @__PURE__ */ Object.freeze(genBech32('bech32'));
/**
* bech32m from BIP 350. Operates on words.
* It was to mitigate `bech32` weaknesses.
* For high-level helpers, check out {@link https://github.com/paulmillr/scure-btc-signer | scure-btc-signer}.
* @example
* Convert bytes to words, encode them with bech32m, then decode back.
* ```ts
* const words = bech32m.toWords(Uint8Array.from([1, 2, 3]));
* const text = bech32m.encode('bc', words);
* bech32m.decode(text);
* ```
*/
export const bech32m: TRet<Bech32> = /* @__PURE__ */ Object.freeze(genBech32('bech32m'));
declare const TextEncoder: any;
declare const TextDecoder: any;
/**
* ASCII-to-byte decoder. Rejects non-ASCII text and bytes instead of doing UTF-8 replacement.
* Method names follow `BytesCoder`, so `encode(bytes)` returns a string and `decode(string)` returns bytes.
* @example
* ```js
* const b = ascii.decode("ABC"); // => new Uint8Array([ 65, 66, 67 ])
* const str = ascii.encode(b); // "ABC"
* ```
*/
export const ascii: TRet<BytesCoder> = /* @__PURE__ */ Object.freeze({
encode(data: TArg<Uint8Array>) {
abytes(data);
let res = '';
for (let i = 0; i < data.length; i++) {
const byte = data[i]!;
// ASCII is 7-bit; reject bytes outside 0x00..0x7f instead of silently widening to
// Latin-1/UTF-8.
if (byte > 127) throw new RangeError(`bytes contain non-ASCII byte ${byte} at position ${i}`);
res += String.fromCharCode(byte);
}
return res;
},
decode(str: string) {
if (typeof str !== 'string') throw new TypeError('ascii string expected, got ' + typeof str);
const res = new Uint8Array(str.length);
for (let i = 0; i < str.length; i++) {
// Indexed access is much faster than Uint8Array.from(str, mapFn) here and keeps
// exact error positions.
const charCode = str.charCodeAt(i);
if (charCode > 127) {
throw new RangeError(
`string contains non-ASCII character "${str[i]}" with code ${charCode} at position ${i}`
);
}
res[i] = charCode;
}
return res;
},
});
const _isWellFormedShim = (str: string): boolean => {
// encodeURI rejects malformed UTF-16, giving a compact fallback that matches native
// isWellFormed on our tests/fuzz corpus.
try {
return encodeURI(str) !== null;
} catch {
return false;
}
};
const _isWellFormed: (str: string) => boolean = /* @__PURE__ */ (() =>
// Pick the native check once so utf8.decode doesn't re-probe String.prototype on every call.
typeof ('' as any).isWellFormed === 'function'
? (str) => (str as any).isWellFormed()
: _isWellFormedShim)();
// This fallback stays small because strict UTF-8 only needs fatal decoding plus well-formed
// UTF-16 checks, not the replacement, streaming, or legacy-encoding behavior of full platform
// text codecs.
const utf8Fallback: BytesCoder = /* @__PURE__ */ Object.freeze({
encode(data: TArg<Uint8Array>) {
abytes(data);
let res = '';
for (let i = 0; i < data.length; ) {
const a = data[i++]!;
if (a < 0b1000_0000) {
res += String.fromCharCode(a);
continue;
}
if (a < 0b1100_0010 || i >= data.length) throw new TypeError(`invalid utf8 at byte ${i - 1}`);
const b = data[i++]!;
if ((b & 0b1100_0000) !== 0b1000_0000) throw new TypeError(`invalid utf8 at byte ${i - 1}`);
let cp = ((a & 0b0001_1111) << 6) | (b & 0b0011_1111);
if (a >= 0b1110_0000) {
if (i >= data.length) throw new TypeError(`invalid utf8 at byte ${i - 1}`);
const c = data[i++]!;
if (
(c & 0b1100_0000) !== 0b1000_0000 ||
(a === 0b1110_0000 && b < 0b1010_0000) ||
(a === 0xed && b >= 0b1010_0000)
)
throw new TypeError(`invalid utf8 at byte ${i - 1}`);
cp = ((a & 0b0000_1111) << 12) | ((b & 0b0011_1111) << 6) | (c & 0b0011_1111);
if (a >= 0b1111_0000) {
if (i >= data.length) throw new TypeError(`invalid utf8 at byte ${i - 1}`);
const d = data[i++]!;
if (
a > 0b1111_0100 ||
(d & 0b1100_0000) !== 0b1000_0000 ||
(a === 0b1111_0000 && b < 0b1001_0000) ||
(a === 0b1111_0100 && b >= 0b1001_0000)
)
throw new TypeError(`invalid utf8 at byte ${i - 1}`);
cp =
((a & 7) << 18) |
((b & 0b0011_1111) << 12) |
((c & 0b0011_1111) << 6) |
(d & 0b0011_1111);
}
}
if (cp < 0x10000) res += String.fromCharCode(cp);
else {
cp -= 0x10000;
res += String.fromCharCode((cp >> 10) + 0xd800, (cp & 0x3ff) + 0xdc00);
}
}
return res;
},
decode(str: string) {
astr('utf8', str);
if (!_isWellFormed(str)) throw new TypeError('utf8 expected well-formed string');
// Direct Uint8Array writes are much faster than number[] + Uint8Array.from on Hermes and
// large Node inputs.
const res = new Uint8Array(str.length * 3);
let pos = 0;
for (let i = 0; i < str.length; i++) {
let c = str.charCodeAt(i);
if (c < 0b1000_0000) {
res[pos++] = c;
continue;
}
if (c >= 0xd800 && c <= 0xdfff) {
const d = str.charCodeAt(++i);
c = 0x10000 + ((c - 0xd800) << 10) + d - 0xdc00;
}
if (c >= 0x10000) {
res[pos++] = (c >> 18) | 0b1111_0000;
res[pos++] = ((c >> 12) & 0b0011_1111) | 0b1000_0000;
} else if (c >= 0x800) res[pos++] = (c >> 12) | 0b1110_0000;
else res[pos++] = (c >> 6) | 0b1100_0000;
if (c >= 0x800) res[pos++] = ((c >> 6) & 0b0011_1111) | 0b1000_0000;
res[pos++] = (c & 0b0011_1111) | 0b1000_0000;
}
return res.subarray(0, pos);
},
});
/**
* Strict UTF-8-to-byte decoder. Uses built-in TextDecoder / TextEncoder when available.
* Method names follow `BytesCoder`, so `encode(bytes)` returns a string and
* `decode(string)` returns bytes.
* `encode(bytes)` requires Uint8Array input, preserves an explicit leading BOM, and
* throws on invalid UTF-8 bytes.
* `decode(string)` requires a primitive string and throws on malformed UTF-16 strings with
* lone surrogates.
* @example
* ```js
* const b = utf8.decode("hey"); // => new Uint8Array([ 104, 101, 121 ])
* const str = utf8.encode(b); // "hey"
* ```
*/
export const utf8: BytesCoder = /* @__PURE__ */ (() => {
let _utf8Encoder: any;
let _utf8Decoder: any;
const utf8Builtin: BytesCoder = {
// ignoreBOM preserves an explicit leading U+FEFF;
// fatal rejects invalid UTF-8 bytes instead of replacing them.
encode(data) {
abytes(data);
return (
_utf8Decoder || (_utf8Decoder = new TextDecoder('utf-8', { ignoreBOM: true, fatal: true }))
).decode(data);
},
decode(str) {
astr('utf8', str);
if (!_isWellFormed(str)) throw new TypeError('utf8 expected well-formed string');
return (_utf8Encoder || (_utf8Encoder = new TextEncoder())).encode(str);
},
};
return Object.freeze({
// Select each direction once at module init, since
// TextEncoder and TextDecoder can exist independently.
encode: typeof TextDecoder === 'function' ? utf8Builtin.encode : utf8Fallback.encode,
decode: typeof TextEncoder === 'function' ? utf8Builtin.decode : utf8Fallback.decode,
});
})();
// Keep fallback parity probes behind a test-only export until runtime fallback behavior is decided.
export const __TESTS: {
utf8Fallback: BytesCoder;
_isWellFormedShim: (str: string) => boolean;
} = /* @__PURE__ */ Object.freeze({
utf8Fallback: utf8Fallback,
_isWellFormedShim: _isWellFormedShim,
});
// Built-in hex conversion https://caniuse.com/mdn-javascript_builtins_uint8array_fromhex
// prettier-ignore
const hasHexBuiltin: boolean = /* @__PURE__ */ (() =>
// Require both directions before enabling the native hex path so encode/decode stay symmetric.
typeof (Uint8Array as any).from([]).toHex === 'function' &&
typeof (Uint8Array as any).fromHex === 'function')();
// prettier-ignore
const hexBuiltin: BytesCoder = {
// Keep local type guards so the native path preserves library-level input errors.
// Native toHex emits lowercase hex, matching the fallback alphabet and Node's hex strings.
encode(data) { abytes(data); return (data as any).toHex(); },
// Native fromHex accepts either hex case and rejects odd-length / non-hex syntax.
decode(s) { astr('hex', s); return (Uint8Array as any).fromHex(s); },
};
/**
* hex string decoder. Uses built-in function, when available.
* Lowercase codec; unlike `base16`, this variant accepts either hex case and emits lowercase.
* @example
* ```js
* const b = hex.decode("0102ff"); // => new Uint8Array([ 1, 2, 255 ])
* const str = hex.encode(b); // "0102ff"
* ```
*/
export const hex: BytesCoder = /* @__PURE__ */ Object.freeze(
hasHexBuiltin
? hexBuiltin
: chain(
radix2(4),
alphabet('0123456789abcdef'),
join(''),
normalize((s: string) => {
if (typeof s !== 'string' || s.length % 2 !== 0)
throw new TypeError(
`hex.decode: expected string, got ${typeof s} with length ${s.length}`
);
return s.toLowerCase();
})
)
);
/** Built-in codecs exposed through the deprecated string conversion helpers. */
export type SomeCoders = {
/** UTF-8 string codec. */
utf8: BytesCoder;
/** Hex codec. */
hex: BytesCoder;
/** Uppercase RFC 4648 base16 codec. */
base16: BytesCoder;
/** RFC 4648 base32 codec with padding. */
base32: BytesCoder;
/** RFC 4648 base64 codec with padding. */
base64: BytesCoder;
/** URL-safe base64 codec without `+` or `/`. */
base64url: BytesCoder;
/** Bitcoin-style base58 codec. */
base58: BytesCoder;
/** Monero-style base58 codec. */
base58xmr: BytesCoder;
};
// prettier-ignore
// Keep this registry aligned with CoderType/coderTypeError; only byte<->string codecs belong here.
const CODERS: SomeCoders = {
utf8, hex, base16, base32, base64, base64url, base58, base58xmr
};
type CoderType = keyof SomeCoders;
const coderTypeError =
'Invalid encoding type. Available types: utf8, hex, base16, base32, base64, base64url, base58, base58xmr';
/**
* Encodes bytes with one of the built-in codecs.
* @deprecated Use the codec directly, for example `hex.encode(bytes)`.
* @param type - Codec name.
* @param bytes - Bytes to encode.
* @returns Encoded string.
* @throws On wrong argument types. {@link TypeError}
* @example
* ```ts
* bytesToString('hex', Uint8Array.from([1, 2, 255]));
* ```
*/
export const bytesToString = (type: CoderType, bytes: TArg<Uint8Array>): string => {
if (typeof type !== 'string' || !CODERS.hasOwnProperty(type)) throw new TypeError(coderTypeError);
if (!isBytes(bytes)) throw new TypeError('bytesToString() expects Uint8Array');
return CODERS[