Module version of UZIP.js
npm install uzip-modulezlib inflate, deflate, inflateRaw, deflateRaw, as well
as simple in memory zip creation and parsing.
This is a ES6 module version of UZIP.js
It's faster than pako in my tests.
``
import {
deflate,
deflateRaw,
inflate,
inflateRaw,
encode,
parse,
} from 'uzip-module';
Includes header and footer
`
const compressed = deflate(uint8Array);
or
`
const compressed = deflate(uint8Array, {level:9});
`
const uncompressedUint8Array = inflate(compressedUint8Array);
or
`
const uncompressedUint8Array = inflate(
compressedUint8Array,
destinationUint8Array);
These take the exact same arguments as inflate and deflate
but don't store or expect the header or the footer
Creates a zip file. You pass a JavaScript object of filenames
to Uint8Arrays it returns a Uint8Array zip file
`
const utf8Encoder = new TextEncoder();
const files = {
'stuff/': utf8Encoder.encode(''),
'stuff/dog.txt': utf8Encoder.encode('german shepard\n'),
'stuff/birds/': utf8Encoder.encode(''),
'stuff/birds/bird.txt': utf8Encoder.encode('parrot\n'),
'stuff/cat.txt': utf8Encoder.encode('siamese\n'),
'stuff/long.txt': utf8Encoder.encode(),`
}
const zipUint8Array = encode(files);
Does the opposite of encode. Takes a zip Uint8Array
and returns a JavaScript object of filenames to Uint8Arrays
Calling parse on the zipUint8Array from the previousfiles
example will return the same data seen in above
`
const unzippedFiles = parse(files);
You can also call it with an extra true in which case it willsize
just return the filenames and a and csize for each one.
The uncompressed and compressed sizes;
`
const unzippedFileSizes = parse(files, true);
unzippedFileSizes would have a structure like
`
{
"stuff/": {
"size": 0,
"csize": 0
},
"stuff/dog.txt": {
"size": 15,
"csize": 15
},
"stuff/long.txt": {
"size": 1601,
"csize": 24
}
}
All credit goes to the original author Photopea.
I ported this to ES6 modules to use in another library.
Originally I thought about putting the various parts in more separate files
like the inflate in one file, deflate in another, parse, and encode in others
but tree shaking is supposed to handle this stuff more or less
so I think just moving it to ES6 modules is enough to get any
unused code stripped. The only other think maybe left to do
is change the bin and f imports in UZIP.js` to import each
individual identifier instead of all of them.
Also I spent about an hour trying to get ES6 modules to work with mocha in node
but failed so got sick of wasting time and used puppeteer. Patches welcome
to remove that dependency.
I didn't make a non-ES6 version. If it's important submit a PR (babel) and I'll
take a look.