Awesome

More info on active projects and modules at dat-ecosystem.org <img src="https://i.imgur.com/qZWlO1y.jpg" width="30" height="30" />

multidat

Manage multiple dat instances in multiple locations.

Usage

var Multidat = require('multidat')
var toilet = require('toiletdb')

var db = toilet('/tmp/dat')
Multidat(db, function (err, multidat) {
  if (err) throw err

  multidat.create(opts, function (err, dat) {
    if (err) throw err

    var dats = multidat.list()
    console.log(dats)

    multidat.close(dat.archive.key, function (err) {
      if (err) throw err
      console.log()
    })
  })
})

Error handling

If there is an error initializing a dat, instead of the whole process failing, an error object with attached .dir property will be pushed into the list of dats instead. That means when consuming multidat.list(), you should check for errors:

var dats = multidat.list()
dats.forEach(function (dat) {
  if (dat instanceof Error) {
    var err = dat
    console.log('failed to initialize dat in %s: %s', err.dir, err.message)
  }
})

This way you can decide for yourself whether an individual initialization failure should cause the whole process to fail or not.

API

Multidat(db, opts, callback(err, multidat))

Creat a new Multidat instance. Takes a toiletdb instance and a callback.

Options:

Dat: Use provided dat factory instead of dat-node

multidat.create(dir, opts, callback(err, dat))

Create a new dat archive.

dats = multidat.list()

List all available dat archives.

multidat.close(key, callback(err))

Close and remove a dat archive.

multidat.readManifest(dat, callback(err, manifest))

Read the dat.json file from the dat archive. This method is expected to be deprecated once dat archives provide a built-in method to return archives.

Why?

This package exists to manage multiple dat archives in different directories. The dat-node package is mostly stateless; all state is persisted into the archives themselves. This package acts as a layer on top to keep track of where archives are located and manage them between sessions.

When not to use this

If you're running a server, it's usually enough to run mafintosh/hypercore-archiver which is more consistent and simpler. If you're building a tool that only needs to manage a single dat archive at the time it's recommended to use datproject/dat-node instead.

License

MIT