@codezilluh/bitarray.js

BitArray

An easy-to-use ArrayBuffer-like concept for storing indivdual bits.

Purpose

The purpose of this package is to make data optimization easier. Why store a value as 8 bits when you only need 5? Although this package still stores data into bytes, you can easily pack those bytes however you want.

Installation

You can install the package with npm

npm i @codezilluh/bitarray.js

Usage

You can store values as integers, unsigned integers, and bits. You can set a scale for the value you wish to store. The scale determines the range of possible values. A scale of 8 for a uint (uint-8) can store any number from 0 to 255. More info on ranges later.

To encode (store) data:

// Create a new BitArray
const array = new BitArray();

// Add a uInt to the BitArray
// 223 is the uInt to store
// 8 is the amount of bits (this example is of a uint-8)
array.addUint(223, 8);

// You could also specify where you want to store the uInt
// 0 is the position in bits
array.setUint(223, 8, 0);

// Regular integers can also be added
// -45 is the integer to store
// 7 is the amount of bits (this could be called an int-7)
array.addInt(-45, 7);

// Bits (booleans) may be stored
array.addBit(true);

// Encode the BitArray into an ArrayBuffer
array.encode();

To decode (read) data:

// There are three options (differing only in style)
const array = decodeBitArray(buffer);
// OR
const array = new BitArray(buffer);
// OR
const array = existingBitArray.decode(buffer);

// You can read a uInt with an optional position
// 8 is the scale (this is a uint-8)
// 0 is the position (in bits)
array.getUint(8, 0);

// You read ints the same way
// 7 is the scale
// Notice the position is optional for reading in order
array.getInt(7);

// Bits are the same
// Think of this example as "get next bit"
array.getBit();

Ranges

As mentioned, you can store data with a custom scale. There is no official limit to the scale, however doing anything larger than 64-bit seems impractical and is not what this package is meant for.

These are the general formulas for finding the range of a particular scale:

Formula for uInts with scale of n:
    min: 0,
    max: (2^n) - 1

Formula for ints with scale of n:
    min: -(2^(n-1))
    max: (2^(n-1)) - 1

I have built functions into this package to output ranges:

// This is what would be outputted for the 8-bit number scale
// [MIN, MAX]
scaleRangeInt(8); // [-128, 127]
scaleRangeUint(8); // [0, 255]

Conclusion

I hope you find this easy-to-use. I will be using this mostly for optimizing real-time communication, but it could also be used to optimize database or data storage in general.