2.1.0 • Published 2 years ago

rc5 v2.1.0

Weekly downloads
47
License
MIT
Repository
github
Last release
2 years ago

RC5 · [Travis Status]

What is RC5: https://en.wikipedia.org/wiki/RC5

This implementation keep to paper: http://people.csail.mit.edu/rivest/Rivest-rc5rev.pdf

Usage

installation

npm install rc5

example

import RC5 from "rc5";

let key = "whisper";

let rc5 = new RC5(key);

// Encryption
let enBuf = rc5.encrypt("plain");
// => <Buffer 47 f1 b4 5c 8d e5 6d e3>

// Decryption
let deBuf = rc5.decrypt(enBuf);
// => <Buffer 70 6c 61 69 6e>
deBuf.toString();
// => plain

Documentation

The rc5 module provides RC-5 protocol encryption algorithm. use require('rc5') to access this module. to use rc5 we must create a instance of RC5 first, and provide 3 initial parameters:

  • key // secret key which user selected
  • w // bits size of each block
  • r // round of encryption

after that, we can use instance of RC5 to encrypt or decrypt data. NOTE that, encryption inputs and outputs are both should be Buffer, for convinent, encrypt method accept string type parameter, and will automatically convert to Buffer, by Buffer.from method.

new RC5(key, w, r)

  • key: Buffer | String, secret key that used in encryption/decryption. length of key should less than 255 bytes. Default: ""
  • w: Integer, which block size selected, w ∈ 16, 32, 64. Default: 32
  • r: Integer, how many rounds during encryption/decryption, r ∈ 0, 255 . Default: 12

Example: create RC5 instance

const RC5 = require("rc5");
let rc5; // instance of RC5

// create a normal instance
rc5 = new RC5("secret key", 32, 12);

// use default block size , rounds without key
rc5 = new RC5();

// use most complex secure level
rct = new RC5(Buffer.from("... a 255 bytes length key..."), 64, 255);

rc5.encrypt(plaintext)

  • plaintext: Buffer | String, plain resource to be encrypted.

Example: encryption

const RC5 = require("rc5");
const rc5 = new RC5("key");
rc5.encrypt(Buffer.from([1, 2, 3, 4, 5, 6]));
// => <Buffer fe 14 e1 42 64 2b db de>
rc5.encrypt("桜の花");
// => <Buffer d8 66 d5 3b b5 d4 91 7e c5 06 98 10 8a 63 d7 d3>

rc5.decrypt(ciphertext)

  • ciphertext: Buffer| String, cipher text to be decrypted.
  • options
    • trim: weather trim the tailing zero or not, default is true, see issue.

Example: decryption

const RC5 = require("rc5");
const rc5 = new RC5("key");
rc5.decrypt(Buffer.from([0xfe, 0x14, 0xe1, 0x42, 0x64, 0x2b, 0xdb, 0xde]));
// => <Buffer 01 02 03 04 05 06>
let buf = rc5.decrypt(
  Buffer.from([
    0xd8, 0x66, 0xd5, 0x3b, 0xb5, 0xd4, 0x91, 0x7e, 0xc5, 0x06, 0x98, 0x10,
    0x8a, 0x63, 0xd7, 0xd3,
  ])
);
// => <Buffer e6 a1 9c e3 81 ae e8 8a b1>
buf.toString();
// => '桜の花'

note rc5 is block cipher algorithm, every block has the same length, if the last block is not long enough, it will completing with 0x00, and rc5.js will trim the tailing 0x00 by default, but maybe it's wrong because the source data is the 0x00. In such case, you should pass set option trim to false, not finished yet! then you should calculate the correct bytes by yourself, see the example:

// https://github.com/DeronW/rc5.js/issues/12
import RC5 from "rc5";

const rc5 = new RC5("a key", 64, 255);
const ptBinary = Uint8Array.from([1, 2, 3, 4, 5, 0, 0]);
const encrypted = rc5.encrypt(Buffer.from(ptBinary));

console.log(rc5.decrypt(encrypted));
// => <Buffer 01 02 03 04 05>

console.log(rc5.decrypt(encrypted, { trim: false }));
// => <Buffer 01 02 03 04 05 00 00 00>

// none of them is right, pick up the bytes properly

Specification

of disadvantages in RC5 scenario with JavaScript.

RC5 need even size of blocks in encryption algorithm, and For simplicity it is proposed here that only the values 16, 32 and 64 be "allowable", which means RC5 need a 64 bits data type, that natively support add, subtract and shift(left or right) operations.

  • Number, BigInt

JavaScript has two foundamental numeral type Number and BigInt. But none of them could be used in RC5 cryption. Number has the biggest safe number: 2^53-1. why this number ? Because JavaScript followed IEEE754 where specified double precision float number, in easy way, Number can't represent Int64 type number accurate. BigInt is designed for represent big nubmer, really big nubmer without any inaccuracy, but, it does't support shift operation natively.

  • UInt8Array, UInt16Array, UInt32Array, Float64Array

None of such type could be used in RC5 cryption either. none of these type could represent UInt64 data.

all of them has same problem

TODO

  • Unit testing & 100% test coverage
  • Travis integrate
  • Browser support
  • TypeScript support
2.1.0

2 years ago

2.0.0

4 years ago

2.0.0-alpha.1

4 years ago

2.0.0-alpha

4 years ago

1.0.0

5 years ago

0.6.0

5 years ago

0.5.0

5 years ago

0.4.0

5 years ago

0.2.1

5 years ago