Documentation request: Example on how to save as buffer

[johann1301s]

Documentation request

I would very much like to see an example of how to properly save as buffer.

Say i have to 4 strings:

const fileNameA = "file-a.txt"
const fileContentA = "foo"

const fileNameA = "file-b.txt"
const fileContentA = "bar"

How can i use Archiver to save these files as a tgz-buffer in memory?

[simoco0781]

I am also interested in. I would like to generate a zip and download it like a "blob" for exemple.

[Arnesfield]

You can pipe archive to a Writable stream and manually handle the chunks. Example:

import archiver from 'archiver';
import fs from 'fs';
import { Writable } from 'stream';

/**
 * @param {archiver.Archiver} archive
 */
async function toBuffer(archive) {
  // create writable and save chunks
  /** @type {Uint8Array[]} */
  const chunks = [];
  const writable = new Writable();
  writable._write = (chunk, encoding, callback) => {
    // save to array to concatenate later
    chunks.push(chunk);
    callback();
  };

  // pipe to writable
  archive.pipe(writable);
  await archive.finalize();

  // once done, concatenate chunks
  return Buffer.concat(chunks);
}

const archive = archiver('tar');
archive.file('./file-a.txt', { name: 'file-a.txt' });
archive.file('./file-b.txt', { name: 'file-b.txt' });

toBuffer(archive).then(buffer => {
  console.log('buffer:', buffer.toString());

  // additional: save buffer to disk to verify contents
  fs.writeFileSync('./my-zip.tgz', buffer);
});

The idea is to implement Writable._write (either by extending Writable or assigning a function), save the chunks in an array, and finally concatenate them once the archive is done.

[newmanw]

Wondering if this is a viable addition to the library?