djb2a

DJB2a non-cryptographic hash function

DJB2a has good distribution and collisions are rare.

The implementation uses BigInt internally for precision, avoiding integer overflow issues that can occur with JavaScript's regular number type when processing long strings.

Install

npm install djb2a

Usage

import djb2a from 'djb2a';

// Hash a string (UTF-16 code units by default)
djb2a('🦄🌈');
//=> 1484783307

// Hash a string as UTF-8 bytes for cross-language compatibility
djb2a('🦄🌈', {bytes: true});
//=> 1642026147

// Hash bytes directly
djb2a(new Uint8Array([1, 2, 3]));
//=> 193375973

// Return a 64-bit BigInt instead of a 32-bit number
djb2a.bigint('🦄🌈');
//=> 5779750603n

// BigInt with UTF-8 bytes
djb2a.bigint('🦄🌈', {bytes: true});
//=> 7564289453806755n

API

djb2a(input, options?)

Returns the hash as a positive 32-bit integer.

input

Type: string | Uint8Array

The input to hash. Strings are hashed as UTF-16 code units by default (see options.bytes to change this).

options

Type: object

bytes

Type: boolean\ Default: false

Interpret the string as UTF-8 encoded bytes instead of UTF-16 code units.

This provides compatibility with DJB2a implementations in other languages, which typically operate on UTF-8 bytes.

djb2a.bigint(input, options?)

Returns the hash as a positive 64-bit BigInt.

Accepts the same arguments as the main function.

Implementation notes

By default, this implementation processes strings as UTF-16 code units using JavaScript's charCodeAt method. This makes it a JavaScript-specific variant of DJB2a that differs from the original C implementation (which operates on bytes) for non-ASCII characters.

Use the bytes option to encode strings as UTF-8 for compatibility with other language implementations.

Related

• fnv1a - FNV-1a non-cryptographic hash function
• sdbm - SDBM non-cryptographic hash function