Slambda

Install

npm i --save slambda

Usage

const Slambda = require('slambda');
 
// Default values
const slambda = new Slambda({
  autoDeploy: true,
  container: {
    language: 'nodejs4.3',
    lifecycle: {
      init: () => {},
      pre: () => {},
      post: () => {},
    },
    memory: 1024,
    timeout: 10,
  },
  storage: new MemoryStore(),
  execution: new MemoryExec(),
});
 
// Create a container
const container = slambda.container('myapi');
 
// Create a method
container.method('get:user', (user) => {
  delete user.ssn;
  return user;
});
 
// Deploy
container.deploy();
 
// Call method
container
  .run('get:user', [{ id: 123, name: 'seth', ssn: '000-00-0000' }])
  .tap(user => console.log('user', user));

Chainable!? The above code can be shortened, as all methods except .run() are chainable:

// Create a container
slambda
  .container('myapi');
  .method('get:user', (user) => {
    delete user.ssn;
    return user;
  })
  .deploy()
  .run('get:user', [{ id: 123, name: 'seth', ssn: '000-00-0000' }])
  .tap(user => console.log('user', user));

Adapters

Slambda has two types of adapters, Storage and Execution.

All required adapter methods must return a Bluebird promise.

Storage

Storage adapters tell Slambda how to persist methods and containers. The default is Memory. Storage adapters can take whatever configuration they need.

Required methods:

get(String table, String id)

Returns an entity by ID

findById(String table, String index, String id)

Query table by index. Resolves to an array.

list(String table)

Resolves a list of entities from the specified table

put(String table, Object item)

Upserts an entity

delete(String table, String id)

Deletes specified entity by ID

Note: All storage adapter methods must return a Bluebird promise

Officially supported storage adapters:

• Memory (default)
• DynamoDB
• FileSystem (Coming soon)
• AWS S3 (Coming soon)
• MongoDB (Coming soon)
• Github (Coming soon)

Execution

Execution adapters tell Slambda how to run your code snippets. The default is Memory.

Required methods:

run(String containerId, String methodId, Array methodArguments)

Run a single method with arguments. Returns a promise with the results

deploy(Container container, Array<Method> methods)

Deploy a container and it's functions to the execution layer

execute(String containerId, Array<String id, Array arguments> calls)

Run a batch of method calls. Must return an equal length array as calls. Array order must not change.

Note: All storage adapter methods must return a Bluebird promise

Where possible, the .run() method should try to batch requests.

Officially supported execution adapters:

• Memory (default)
• Local (File system)
• AWS Lambda
• Google Cloud Functions (Coming soon)
• Docker (Coming soon)
• AWS EC2 (Coming soon)

API

Slambda

constructor(Object options)

Parameters

Object options

Defaults

{
  autoDeploy: true,
  container: {
    language: 'nodejs4.3',
    lifecycle: {
      init: () => {},
      pre: () => {},
      post: () => {},
    },
    memory: 1024,
    timeout: 10,
  },
  storage: new MemoryStore(),
  execution: new MemoryExec(),
}

container(String id, Object options)

Returns

Container

Parameters

String id Container ID Required

Object options Defaults to this.options.container

Container

constructor(Object options)

Parameters

Object options

Defaults

{
  language: 'nodejs4.3',
  lifecycle: {
    init: () => {},
    pre: () => {},
    post: () => {},
  },
  memory: 1024,
  timeout: 10,
}

method(String id, String|Function code)

Returns

Container

Parameters

String id Container-unique identifier for the method

String|Function code Method code.

run(String id, Array args)

Returns

Promise

Parameters

String id Method ID to run

Array args Arguments passed to method

deploy()

Deploys function to execution layer. Queues .run() commands behind deployment

Returns

Container