Serverlessrest

Serverless implementation of "REST" using DynamoDB and GraphQL. Easy to setup - create one config file, deploy and ready to use.

Getting Started

Prerequisites

• Download node
• Install npm
• Setup Serverless:
• Setup yarn

Installing

yarn install
sls dynamodb install

Running the tests

  yarn test

or for e2e tests:

  yarn test:e2e

Configuration

Basic

To create simple CRUD (create, read, update and delete) for any resource you have to add config object similar to:

{
  tableName: 'objects',
  item: 'Object',
  attributes: [
    { name: 'text_attr_1', type: 'String', required: true },
    { name: 'num_attr_2', type: 'Int' },
    { name: 'text_arr_attr_3', type: '[String]' },
  ],
  queries: [
    { type: 'get' },
    { type: 'index' },
  ],
  mutations: [
    {
      type: 'create',
      attributes: [
        { name: 'text_attr_1', type: 'String!' },
        { name: 'text_arr_attr_3', type: '[String]' },
      ],
    },
    {
      type: 'update',
      attributes: [
        { name: 'text_attr_1', type: 'String!' },
      ],
    },
    {
      type: 'delete',
    },
  ],
}

It will generate GraphQL server for resource Object that has such attributes:

• Required text_attr_1 that has to be String
• num_attr_2 that has to be Integer
• text_arr_attr_3 that is an array of String values autogenerated:
• ID

also it will create GraphQL queries and mutations that should work like CRUD:

• get
• index (by ID )
• create (you have to pass text_attr_1 and text_arr_attr_3 - as specified in config)
• update (you have to pass text_attr_1 - as specified in config)
• delete (by ID)

GraphQL schema that is generated from this configuration:

schema {
  query: Query
  mutation: Mutation
}

type Query {
  objects: [Object]
  object(id: ID!): Object
}

type Mutation {
  createObject(text_attr_1: String!, text_arr_attr_3: String!): Object
  updateObject(id: ID!, text_attr_1: String!): Object
  deleteObject(id: ID!): Object
}

type Object {
  id: ID!
  title: text_attr_1!
  num_attr_2: Int
  text_arr_attr_3: [String]
}

All resolvers are also auto generated.

Enum

You can create ENUM type, add such object to config file:

{
  type: 'enum',
  name: 'ObjectType',
  values: ['SIMPLE', 'COMPLEX', 'EMPTY'],
}

then you can use it as attribute on object config:

{
  tableName: 'objects',
  item: 'Object',
  attributes: [
    ...
    { name: 'object_type', type: 'ObjectType' },
    ...
  ],
  ...

Custom resolver

Not only you can resource by ID but also by custom value, for example by name. You have to create custom query, to do so, add something similar to queries in config file:

queries: [
    ...
    {
      type: 'custom',
      name: 'objectByName',
      attributes: [
        {
          name: 'name',
          type: 'String',
          required: true,
        },
      ],
      returnType: 'collection',
    },
  ],

It should return all object with name passed as argument.

Auth

You can set authorization for every object in config, by default it is not set. If you want to add it you have to add auth: true, value to object config:

{
  tableName: 'objects',
  item: 'Object',
  auth: true,
  ...

it will add authorization to this resource, you can learn more about it in Examples section.

GraphQL Examples

All examples should work with Complex Config -> examples/complex/complex.config.js, to run it:

yarn run sls:offline --printOutput  -- --config='examples/complex/complex.config.js'

CREATE

mutation {
  createEpic (description: "Some epic", sprint: "10", due_date: "20.02.2020") {
    id, description, sprint, due_date
  }
}

READ

INDEX - get all records

query {
  epics {
    id, description, created_at, due_date
  }
}

GET - get single record by ID

query {
  epic (id: "__SPECIFY_ID_HERE__") {
    id, description, created_at, due_date
  }
}

CUSTOM - get single record by attribute

query {
  epicsBySprint (sprint: "10") {
    id, description, created_at, due_date
  }
}

UPDATE

mutation {
  updateEpic (id: "__SPECIFY_ID_HERE__", description: "Some epic v2") {
    id, description
  }
}

DELETE

mutation {
  deleteEpic (id: "22f2d3f0-b765-11e9-b389-e358651b1ef5") {
    id
  }
}

AUTH

If resource is marked as authorized, then you have to login before you can perform any action on it, Otherwise you will get error:

You are not authenticated!

To create account you have to perform signup mutation:

mutation {
  signup(username: "uname", password: "password", email: "some@mail.com")
}

as response you will get jsonwebtoken that you can use to authorize all your future requests.

You can also login:

mutation {
  login (password: "password", email: "some@mail.com")
}

To authorize request, you have to add HTTP Header to request:

Authorization	Bearer __JSONWEBTOKEN__

Deployment

Local

yarn run sls:offline

You can visit http://localhost:8000/shell for Dynamodb Local Shell. Serverless:Offline is listening on http://localhost:3000.

AWS

Setup an AWS Account for use with Serverless

You can create custom profile serverlessrest:

  serverless config credentials --provider aws --key ___AWS_KEY___ --secret ___AWS_SECRET_KEY___ --profile serverlessrest

To deploy you can run:

  serverless deploy --aws-profile serverlessrest

Built With

• Serverless - The easy, open way to build serverless applications
• GraphQL - A query language for your API
• AWS - Cloud Computing Services

License

This project is licensed under the MIT License - see the LICENSE.md file for details.