JavaScript String compression

string.compress.js

/*
	@fliptopbox
	
	LZW Compression/Decompression for Strings
	Implementation of LZW algorithms from:
	http://rosettacode.org/wiki/LZW_compression#JavaScript

	Usage:
	var a = 'a very very long string to be squashed';
	var b = a.compress(); // 'a veryāăąlong striċ to bečquashed'
	var c = b.uncompress(); // 'a very very long string to be squashed'
	console.log(a === c); // True

	var d = a.compress(true); // return as Array
	console.log(d); // [97, 32, 118 .... 101, 100] an Array of ASCII codes

*/

String.prototype.compress = function (asArray) {
	"use strict";
	// Build the dictionary.
	asArray = (asArray === true);
	var i,
		dictionary = {},
		uncompressed = this,
		c,
		wc,
		w = "",
		result = [],
		ASCII = '',
		dictSize = 256;
	for (i = 0; i < 256; i += 1) {
		dictionary[String.fromCharCode(i)] = i;
	}

	for (i = 0; i < uncompressed.length; i += 1) {
		c = uncompressed.charAt(i);
		wc = w + c;
		//Do not use dictionary[wc] because javascript arrays
		//will return values for array['pop'], array['push'] etc
	   // if (dictionary[wc]) {
		if (dictionary.hasOwnProperty(wc)) {
			w = wc;
		} else {
			result.push(dictionary[w]);
			ASCII += String.fromCharCode(dictionary[w]);
			// Add wc to the dictionary.
			dictionary[wc] = dictSize++;
			w = String(c);
		}
	}

	// Output the code for w.
	if (w !== "") {
		result.push(dictionary[w]);
		ASCII += String.fromCharCode(dictionary[w]);
	}
	return asArray ? result : ASCII;
};

String.prototype.decompress = function () {
	"use strict";
	// Build the dictionary.
	var i, tmp = [],
		dictionary = [],
		compressed = this,
		w,
		result,
		k,
		entry = "",
		dictSize = 256;
	for (i = 0; i < 256; i += 1) {
		dictionary[i] = String.fromCharCode(i);
	}

	if(compressed && typeof compressed === 'string') {
		// convert string into Array.
		for(i = 0; i < compressed.length; i += 1) {
			tmp.push(compressed[i].charCodeAt(0));
		}
		compressed = tmp;
		tmp = null;
	}

	w = String.fromCharCode(compressed[0]);
	result = w;
	for (i = 1; i < compressed.length; i += 1) {
		k = compressed[i];
		if (dictionary[k]) {
			entry = dictionary[k];
		} else {
			if (k === dictSize) {
				entry = w + w.charAt(0);
			} else {
				return null;
			}
		}

		result += entry;

		// Add w+entry[0] to the dictionary.
		dictionary[dictSize++] = w + entry.charAt(0);

		w = entry;
	}
	return result;
};

Fantastic piece of code. Short, sweet, fast and very easy to use.

It looks to me that this implementation mixed up ASCII and Unicode characters. The line

ASCII += String.fromCharCode(dictionary[w]);

will actually produce no ASCII, but Unicode characters. This may be a way to easily represent the codes 256+ in JavaScript, but I argue that this is not the LZW compression as other implementations will interpret it?!

For testing: https://jsfiddle.net/aers9b6g/

@BurninLeo hey, yes you are absolutely right. it is a unicode string. you argument is 100% correct.
this gist was written a very long time ago. i had actually forgotten about it.

you can return the ASCII value array like this String.compress(true) the true switch will bypasses the Unicode conversion.

thanks for the feedback :D i hope people read this comment