digger-server

Warehouse routing for digger-client requests.

install

$ npm install digger-server

usage

Make a digger-level supplier and mount it.

var level = require('level');
var diggerserver = require('digger-server')
var diggerlevel = require('digger-level')

// create a new leveldb - this can also be a sub-level
var leveldb = level('/tmp/digger')

// create a level digger supplier
var supplier = diggerlevel(leveldb)

// create a digger server to mount our level supplier
var digger = diggerserver()

// mount the level supplier onto the server
digger.warehouse(supplier)

// access control
digger.access(function(path, user, mode, next){
	next()	
})

// create a HTTP server to host it
var server = http.createServer(digger.handler())

server.listen(80)

notes

The digger server accepts HTTP requests that have been converted to JSON objects.

A digger req is a read-stream in object mode with the following properties:

{
	method:'get',
	headers:{
		'x-digger-selector':'folder.red'
	},
	url:'/apples'
}

raw requests

You can convert raw HTTP requests using digger-http.

endpoints

Here are the notable endpoints:

• POST /digger/ship - callback based contracts
• POST /digger/stream - stream based contracts
• POST /digger/select - selector resolving
• • /* - warehouse

The first 3 are used for contract resolving - a contract is a complicated multi-step selector / append / save / delete request.

ship

The ship endpoint accepts a contract as its POST data and expects the body of each contract to be present.

stream

The stream endpoint accepts a contract in its x-digger-contract header and will stream the requests input to that contract.

select

This creates selector streams for each of the selector stages and pipes them accordingly.

The input is not JSON packets but context paths.

warehouse

This is the REST controller back to the storage engine - you can call standard HTTP on this (for example to load a single container just GET its url)

licence

MIT