Two things: speed & size.
LZMA-JS 2.x now minifies to smaller than one fourth of 1.x and in some cases is 1,000x faster (particularly with high compression).
It is also more modular. The compression and decompression algorithms can be optionally separated to shrink the file size even more.
Here are some file size stats:
|lzma_worker.js||both||23.4 KB||9.2 KB|
|lzma-c.js||compression||17.9 KB||7.3 KB|
|lzma-d.js||decompression||6.8 KB||3.0 KB|
Also, older versions returned compressed data as unsigned bytes. Now, it returns signed bytes.
Live demos can be found here.
LZMA-JS is available in the npm repository.
npm install lzma
If you are using bower, you can download the source like this:
bower install lzma
First, load the bootstrapping code.
<!-- In a browser -->
Create the LZMA object.
/// LZMA([optional path])/// If lzma_worker.js is in the same directory, you don't need to set the path.var my_lzma = "../src/lzma_worker.js";
(De)Compress stuff asynchronously:
/// To compress:///NOTE: mode can be 1-9 (1 is fast and pretty good; 9 is slower and probably much better).///NOTE: compress() can take a string or an array of bytes./// (A Node.js Buffer or a Uint8Array instance counts as an array of bytes.)my_lzma;/// To decompress:///NOTE: By default, the result will be returned as a string if it decodes as valid UTF-8 text;/// otherwise, it will return a Uint8Array instance.my_lzma;
(De)Compress stuff synchronously (not recommended; may cause the client to freeze):
/// To compress:///NOTE: You'll need to do your own error catching.result = my_lzma;/// To decompress:result = my_lzma;
After installing with npm, it can be loaded with the following code:
var my_lzma = ;
decompress() function needs an array of bytes or a Node.js
If the decompression progress is unable to be calculated, the
on_progress() function will be triggered once with the value
LZMA-JS will try to use Web Workers if they are available. If the environment does not support Web Workers,
it will just do something else, and it won't pollute the global scope.
Each call to
LZMA() will create a new Web Worker, which can be accessed via
If you'd prefer not to bother with Web Workers, you can just include
lzma_worker.js directly. For example:
That will create a global
LZMA object that you can use directly. Like this:
Note that this
LZMA variable is an
Object, not a
In Node.js, the Web Worker code is already skipped, so there's no need to do this.
And if you only need to compress or decompress and you're looking to save some bytes, instead of loading lzma_worker.js, you can simply load lzma-c.js (for compression) or lzma-d.js (for decompression).
Of course, you'll want to load the minified versions if you're sending data over the wire.
LZMA-JS is compatible with anything that is compatible with the reference implementation of LZMA, for example, the