JarCastingSHA-3 for JavaScript
A pure JavaScript implementation of the Keccak family of cryptographic hashing algorithms, most notably including Keccak and SHA3.
💡 Legacy Note: In previous versions of this library, theSHA3Hashobject provided a Keccak hash, not what we currently know as a SHA-3 hash. For backwards-compatibility, this object is still exported. However, users are encouraged to switch to using theSHA3orKeccakobjects instead, which provide the SHA-3 and Keccak hashing algorithms, respectively.
Installation
Via npm:
$ npm install sha3
Via yarn:
$ yarn add sha3
Usage
You can use this library from Node.js, from web browsers, and/or using ES6 imports.
Node.js (CommonJS style)
// Standard FIPS 202 SHA-3 implementation
const { SHA3 } = require('sha3');
// The Keccak hash function is also available
const { Keccak } = require('sha3');
ES6
// Standard FIPS 202 SHA-3 implementation
import { SHA3 } from 'sha3';
// The Keccak hash function is also available
import { Keccak } from 'sha3';
What's in the box
FIPS-compatible interfaces for the following algorithms:
SHA3: The SHA3 algorithm.Keccak: The Keccak algorithm.SHAKE: The SHAKE XOF algorithm.
💡 Legacy Note: Savvy inspectors may notice thatSHA3Hashis also provided. Prior to v2.0.0, this library only implemented an early version of the SHA3 algorithm. Since then, SHA3 has diverged from Keccak and is using a different padding scheme, but for compatibility, this alias is sticking around for a bit longer.
Examples
Generating a SHA3-512 hash
import { SHA3 } from 'sha3';
const hash = new SHA3(512);
hash.update('foo');
hash.digest('hex');
Generating a Keccak-256 hash
import { Keccak } from 'sha3';
const hash = new Keccak(256);
hash.update('foo');
hash.digest('hex');
Generating a SHAKE128 hash with 2048 bytes
import { SHAKE } from 'sha3';
const hash = new SHAKE(128);
hash.update('foo');
hash.digest({ buffer: Buffer.alloc(2048), format: 'hex' });
API Reference
All hash implementations provided by this library conform to the following API specification.
#constructor([size=512])
The constructor for each hash (e.g: Keccak, SHA3), expects the following parameters:
size(Number): Optional. The size of the hash to create, in bits. If provided, this must be one of224,256,384, or512. Defaults to512.
Example
// Construct a new Keccak hash of size 256
const hash = new Keccak(256);
#update(data, [encoding='utf8'])
Updates the hash content with the given data. Returns the hash object itself.
data(Buffer|string): Required. The data to read into the hash.encoding(string): Optional. The encoding of the givendata, if of typestring. Defaults to'utf8'.
💡 See Buffers and Character Encodings for a list of allowed encodings.
Example
const hash = new Keccak(256);
hash.update('hello');
hash.update('we can also chain these').update('together');
#digest([encoding='binary'])
Digests the hash and returns the result. After calling this function, the hash may continue to receive input.
encoding(string): Optional. The encoding to use for the returned digest. Defaults to'binary'.
If an encoding is provided and is a value other than 'binary', then this function returns a string. Otherwise, it returns a Buffer.
💡 See Buffers and Character Encodings for a list of allowed encodings.
Example
const hash = new Keccak(256);
hash.update('hello');
hash.digest('hex');
// => hash of 'hello' as a hex-encoded string
#digest([options={}])
Digests the hash and returns the result. After calling this function, the hash may continue to receive input.
Options include:
buffer(Buffer): Optional. A pre-allocated buffer to fill with output bytes. This is how XOF algorithms like SHAKE can be used to obtain an arbitrary number of hash bytes.format(string): Optional. The encoding to use for the returned digest. Defaults to'binary'. Ifbufferis also provided, this value will passed directly intoBuffer#toString()on the given buffer.padding(byte): Optional. Override the padding used to pad the input bytes to the algorithm's block size. Typically this should be omitted, but may be required if building additional cryptographic algorithms on top of this library.
If a format is provided and is a value other than 'binary', then this function returns a string. Otherwise, it returns a Buffer.
Example
const hash = new Keccak(256);
hash.update('hello');
hash.digest({ buffer: Buffer.alloc(32), format: 'hex' });
// => hash of 'hello' as a hex-encoded string
#reset()
Resets a hash to its initial state.
- All input buffers are cleared from memory.
- The hash object can safely be reused to compute another hash.
Example
const hash = new Keccak(256);
hash.update('hello');
hash.digest();
// => hash of 'hello'
hash.reset();
hash.update('world');
hash.digest();
// => hash of 'world'
Testing
Run yarn test for the full test suite.
Disclaimer
Cryptographic hashes provide integrity, but do not provide authenticity or confidentiality. Hash functions are one part of the cryptographic ecosystem, alongside other primitives like ciphers and MACs. If considering this library for the purpose of protecting passwords, you may actually be looking for a key derivation function, which can provide much better security guarantees for this use case.
Special Thanks
The following resources were invaluable to this implementation and deserve special thanks for work well done:
-
Keccak pseudocode: The Keccak team's excellent pseudo-code and technical descriptions.
-
mjosaarinen/tiny_sha3: Markku-Juhani O. Saarinen's compact, legible, and hackable implementation.
-
Phusion: For the initial release and maintenance of this project, and gracious hand-off to Twuni for continued development and maintenance.
