@cley_faye/lzw

Pure JavaScript implementation of LZW compression

Provide a low-level implementation of the LZW algorithm as well as minimal support for compression/decompression of byte buffers.

Installation

npm install @cley_faye/lzw

Usage

The package uses the JavaScript ES module syntax.

Low-level usage

At low-level, LZW only handle numeric "codes" for both input and output. It is possible to create an instance able to compress any input dictionary, but for convenience functions to create a "byte" dictionary (handling all 256 values) is provided.

How to compress and decompress a sequence of input codes:

import {CodeCompress, CodeDecompress} from "@cley_faye/lzw";

const input = [1, 2, 3, 4];
// Compression
const compress = CodeCompress.createSingleBytes();
const compressed = [];
input.forEach(inputCode => {
  const outputCodes = compress.addInput(inputCode);
  if (outputCodes) {
    compressed.push(...outputCodes);
  }
});
const finalCodes = compress.endInput();
if (finalCodes) {
  compressed.push(...finalCodes);
}
console.log(compressed);
// Decompression
const decompress = CodeDecompress.createSingleBytes();
const decompressed = [];
compressed.forEach(inputCode => {
  const outputCodes = decompress.addInput(inputCode);
  decompressed.push(...outputCodes);
});
console.log(decompressed);

Usage with Uint8Array

Encoding of output codes into sequence of bits/bytes is also included.

Here's a basic example:

import {CompressBuf8, DecompressBuf8} from "@cley_faye/lzw";

const input = new Uint8Array([1, 2, 3, 4]);
// Compression
const compressed = CompressBuf8.compressBytes(input);
// Decompression
const decompressed = DecompressBuf8.decompressBytes(compressed);

You can also process chunks by creating an instance of CompressBuf8 or DecompressBuf8.

Encoding

This is incompatible with streams made with the version 1 of this library. Compressed Uint8Array are prefixed with a single byte indicating the settings from the compressor. Beyond that first byte, dictionary codes are output using the least amount of bit necessary to represent the largest entry of the dictionary. If configured, a reset code (256) and a stop code (257, or 256 if no reset code) are added to the dictionary on start/reset. The initial 256 first values of the dictionary are just their corresponding byte.

Settings

The library exposes most settings expected when handling LZW. Most notably:

• Setting a limit to the dictionary size can be done using the maxCodeValue argument, either when creating a custom dictionary or with helper functions.
• Outputing a "stop" code can be disabled with the setStop argument
• Storage order (most significant bit first or not) is handled with the msb argument.

It is advised to leave the default values for setStop (true) and msb (true). Specifying a value for maxCodeValue will lower compression efficiency, but also limit the memory needed to compress/decompress, as the dictionary can grow quite large quite fast.