Base 32 encoding/decoding for JavaScript

Base 32 is between hexadecimal notation and Base 64 encoding. It's intended to be a human-friendly -- you don't have to worry about punctuation, capitalization, or letters/numbers that are easy to confuse, making it easier to transmit in handwriting or over the phone.

One of the primary purposes is to have aesthetically pleasing SHA1 hashes. Compare:

• Hex: 17057684bea1f9331418b633a8f373119d765fd4
• B64: xE_ptB5SeclHm8JEsD0-ST1mTBM
• B32: 2w2qd15ym7wk650rprtuh4vk26eqcqym

Try giving out the Base 64 hash over the phone! "lowercase 'x', capital 'E', underscore, lowercase 'p', ..." Base 32 will work the same with upper- or lowercase, you can mistake a number for a similar-looking letter, and it will still decode to the same data.

Getting started

In your shell, install with npm:

npm install base32

In your code:

var base32 = require('base32')

// simple api

var encoded = base32.encode('some data to encode')
var decoded = base32.decode(encoded)

// streaming api
this.encoder = new Base32.encoder()
this.dataCallback = function(chunk) {
    this.emit(this.encoder(chunk))
}
this.closeCallback = function(chunk) {
    this.emit(this.finish()) // flush any remaining bits
}

// easy sha1 hash
var hash = base32.sha1(some_data_to_hash) // DONE.

On the command-line (to install system-wide, use npm install -g base32):

base32 -h
#> Usage: base32 [input_file] [-o output_file] [-d|--decode] [-s|--sha]
echo "Hello World" | base32
#> 91jprv3f41bpywkccg50
echo 'axqqeb10d5u20wk5c5p6ry90exqq4uvk44' | base32 -d
#> Wow, it really works!
base32 -s test/*
#> ky2t1raumjn9cghne773petngx3zz3q7  test/base32-test.coffee
#> 6b4bkjaveddmg5jh7hnyw132yht20g6e  test/compare.coffee

Warning: this is a Base 32 implementation, not the Base 32 implementation

There are about (128 choose 32) different specifications of something called "Base 32" - see Wikipedia for some of them.

This is a modified Crockford encoding that should be simple, less error-prone, and streamable (for Node).

Minispec

• The encoding alphabet consists of the numerals 0-9 and the letters a-z, excluding a few letters that might look like numbers, which we simply interpret as follows:

  • I -> 1
  • L -> 1
  • O -> 0
  • S -> 5

• When decoding, capital letters are converted to lowercase and the "ambiguous" letters mentioned above converted to their numeric counterparts.

• Each character corresponds to 5 bits of input, tail-padded with 0s where necessary.

• Lexicographic order of strings is preserved through Base 32 encoding.

decimal -> base32 Table

00xxx	01xxx	10xxx	11xxx
0 -> 0	8 -> 8	16 -> g	24 -> r
1 -> 1	9 -> 9	17 -> h	25 -> t
2 -> 2	10 -> a	18 -> j	26 -> u
3 -> 3	11 -> b	19 -> k	27 -> v
4 -> 4	12 -> c	20 -> m	28 -> w
5 -> 5	13 -> d	21 -> n	29 -> x
6 -> 6	14 -> e	22 -> p	30 -> y
7 -> 7	15 -> f	23 -> q	31 -> z

Formalia

Under MIT License.

Fork as much as you like, I'm more than amenable to pull requests. I'm trying to keep it reasonably node-ish, so bear that in mind.