tar

fast, browser-oriented tarball manipulation with zero dependencies

why?

this is a rewrite of tarballjs. enhancements:

• fixes imports by using esmodules
• works in node
• allows tree-shaking
• smaller bundle size
• better developer experience

usage

untar

takes an ArrayBuffer and returns an array of entries representing files and folders

import { untar } from '@immutabl3/tar';

const res = await fetch(url);
const buffer = await res.arrayBuffer();
const entries = await untar(buffer);

Entry

• .path: string - file's tar path
• .ext: string - file's extension (e.g. png, jpg)
• .type: string - 'file' or 'directory'
• .size: number - file's byte size
• .getText(): string - read the file as a text string
• .getBinary(): Uint8Array - read the file as bytes
• .getBlob(mimetype): Blob - read the file as a Blob with the provided mime

tar

creates a writer to add folders and files to. writing the tar returns a Uint8Array

import { tar } from '@immutabl3/tar';

const writer = tar();
writer.addFolder('myfolder/');
writer.addText('myfolder/first.txt', 'this is some text 🙂');
writer.addText('myfolder/second.txt', 'some more text with 🙃 emojis');
const result = await writer.write();

limitations

• File name (including path) has to be less than 100 characters.
• Maximum total file size seems to be limited to somewhere between 500MB to 1GB (exact limit is unknown).

tests

• npm install
• npm test

references

• https://en.wikipedia.org/wiki/Tar_(computing)
• https://www.gnu.org/software/tar/manual/html_node/Standard.html