hyperdrive-http

Hyperdrive Http

Serve a hyperdrive archive or hypercore feed over HTTP. For an example of use, see dat.haus.

Usage

Hyperdrive-http returns a function to call when you receive a http request:

var server = http.createServer()
server.on('request', hyperdriveHttp(archive))

Setup

To use hyperdrive-http you will need to:

• Create your own http server
• Setup your hyperdrive archive(s)
• Connect to the swarm before serving archive

API

Hyperdrive works with many archives/feeds or a single archive.

Multiple Archives

If you have multiple archives, you will need to look the archive to return using the key.

Initiate with an archive lookup function: var onrequest = hyperdriveHttp(getArchive)

The archive lookup function may look like this:

var getArchive = function (datInfo, cb) {
  // datInfo = {
  //   key: archive.key,
  //   filename: filename.txt // If file is requested in URL
  //   op: 'get' or 'changes'
  // }

  // Find the archive to return:
  var archive = cache.get(datInfo.key)
  if (!archive) {
    archive = drive.createArchive(datInfo.key)
    // Make sure you join the swarm before callback
    sw.join(archive.discoveryKey)
  }
  cb(null, archive) // callback with your found archive
}

Single Archive

Hyperdrive-http works great with a single archive too. It exposes the metadata at the root path and files are available without using the key.

Pass a single archive on initiation: var onrequest = hyperdriveHttp(archive)

Now your archive metadata will be available at http://example.com/

Hypercore Feed(s)

You can also use a hypercore feed: hyperdriveHttp(feed) (or using a similar getArchive function)

URL Format

Hyperdrive-http responds to any URL with a specific format. If the URL does cannot be parsed, it will return a 404.

Multiple archives on one site

• Get metadata for archive: http://dat.haus/c5dbfe5521d8dddba683544ee4b1c7f6ce1c7b23bd387bd850397e4aaf9afbd9/
• Get file from archive: http://dat.haus/c5dbfe5521d8dddba683544ee4b1c7f6ce1c7b23bd387bd850397e4aaf9afbd9/filename.pdf
• Upload file: POST http://archive-example.com/ or POST http://archive-example.com/c5dbfe5521d8dddba683544ee4b1c7f6ce1c7b23bd387bd850397e4aaf9afbd9

Single Archive Mode

• Get metadata for archive: http://archive-example.com/
• Get file from archive: http://archive-example.com/filename.pdf
• Upload file: POST http://archive-example.com/

Hypercore Mode

For hypercore feeds, the data is available with the same logic as above for a single or multiple feeds.

Example

var hyperdriveHttp = require('hyperdrive-http')

var getArchive = function (datInfo, cb) {
  // find the archive to serve
  var discoveryKey = crypto.createHmac('sha256', Buffer(datInfo.key, 'hex')).update('hypercore').digest('hex')
  var archive = cache.get(discoveryKey)
  if (!archive) {
    archive = drive.createArchive(datInfo.key)
    // connect to swarm, if necessary
    sw.join(archive.discoveryKey)
  }
  cb(null, archive) // callback with your found archive
}

var onrequest = hyperdriveHttp(getArchive)
var server = http.createServer()
server.listen(8000)
server.on('request', onrequest)

Pass an archive lookup function for the first argument of hyperdriveHttp. The function is called with datInfo and a callback.

datInfo = {
  key: archive.key,
  filename: someFile.txt,
  op: 'get' // or 'changes'
}