free & open source (AGPL-3.0 licensed) webhook and socket events for lifecylce methods REST API out of the box can be started from the command line docker container for out of the box deploys connects to any existing mongo, and can utilize existing data easily extensible / reusable
Don't reinvent the wheel. When you need a simple backend persistence layer, you don't need to write an endpoint and controller for everything. 99% of CRUD tasks can be accomplished with a REST interface. Instead, focus on the other difficult parts of your application.
Since each method is exposed on the root, you can pick and choose what methods you want to use, or you can just mount everything to a custom endpoint and go from there.
npm install --save ubermind
const express = require('express')
const ubermind = require('ubermind')
const app = express()
// mount it to an endpoint
app.use('/ubermind', ubermind({
timestamps: true, // adds timestamps to every object
webhook: 'localhost:3000', // where it will psot events to
socket: 'localhost:4000', // socket.io url
db: 'localhost:27017/ubermind' // specify the database connection string
}));
app.listen(3000, () => {
console.log('listening on port 3000')
});
This will expose the ubermind middleware on to /ubermind
of your server.
If you do a GET
request to the root /ubermind
URL, you'll see a specification of your ubermind endpoint, version, etc...
To create a document, you can simply send a POST
request to the /ubermind
endpoint.
POST
request to /ubermind
takes a data paylaod of
{
model: 'todos', // the name of the collection
data: {
// any fucking data you want
}
}
You can also send the model as a url parameter, like this
/ubermind/todos
and then in the body of the request, you'd put the data
data: {
// any other data
}
You don't need to define the models anywhere. Mongo doesn't enforce schemas, so why would we?
If you want to query collections, you can query them with query params on the model.
This is the endpoin that will be doing most of the heavy lifting for querying.
You can query by any key on your documents using this endpoint, as well as add pagination with skip
, limit
, and offset
query params.
GET
/ubermind/todos?limit=50?skip=5
will return all the todos
collection.
GET
/ubermind?model=todos?limit=50
will return the first 50 documents in the todos
collection.
Both endpoints allow you to add query params.
PUT
/ubermind
{
model: 'todos',
id: '1234',
data: {
// data to update document with
}
}
or you can do a similar URL parameter request
PUT
/ubermind/todos
{
id: '1234',
data: {
// data to update document with
}
}
or
PUT
/ubermind/todos/1234
and this will accept an updated document
{
// any data goes here
}
You can also delete documents with a DELETE
request.
DELETE
/ubermind
{
model: 'todos',
id: '1234'
}
Delete all documents in a collection.
Since this is a dangerous operation, it requires the deleteAll
property to be set.
You can turn this action off in the configs object (see configs section)
DELETE
request
{
model: 'todos',
}
deleteAll: true
must be set in the config
You can delete by parameter as well
{
model: 'todos',
query: {
// add any query params. anything that matches will be deleted
}
}
The server sends back a response object for every request that has the following format:
{
status: xxx,
message: <success or error message>,
data: {
// your response data
}
}
You can add any authentication middleware in front of the ubermind middleware to lock down the endpoint.
The middleware takes an optional configuration option
deleteAll
// not implemented
timestamps
logger
// not implemented
Must be a valid connection string. Follows mongojs connection string formatting. You can read more about that here.
If timestamps
is set to true in the configuration, then each object will be created with a created_at
key.
On update, each object will have a updated_at
key either added or updated, depending on if the document has been updated before.
Out of the box, ubermind can start a server from the command line that will listen on port 1337.
If you want to start it up by the command line`
npm install -g ubermind
Then you can run
ubermind --help
for a print out of all the available commands.
You can also start a docker container that will host the hivmeind server.
docker run -p 1337:1337 --link mongo:mongo -e MONGO_URI=<your_mongo_db_uri> -d --name ubermind ubermindapps/ubermind
Or there's a docker-compose file that you can curl down and docker-compose up
db_connected
- Emitted on database successful connection
db_error
- Emitted on database error
These are fired when actions on a collection happen. Events are fired in the EventEmitter (Node lifecycle), to the Webhook URL, and to the Socket.io URL. They all have the same naming convention and payloads.
<collection>:created
- Emitted when a model is created. They payload of this event is the object that was created.
<collection>:updated
- Emitted when a model is updated. The payload this event is the <key>:<value>
pairs that were changed.
<collection>:deleted
- Emitted when a model is deleted. The payload of this event is the id
(s) of the documents that were deleted.
2.0 Goals
- Configurable but simple authentication setup
- Socket.io introduction (currently only supports webhook)
- Sharding / replication setup
- Docker container
- Command line startup via
npm start
I use nodemon to detect changes in local development.
You can run example.js
with this and it loads in the library.
nodemon example.js
Got an improvement? Fixed a bug? Make a PR, I'll check it out.
AGPL-3.0 licensed