Library for validating input data and parameters


9170.5.04 years ago5 years agoMinified + gzip package size for @cashfarm/plow in KB



Get the soil read to grow awesomeness :)

Documentation (Can't call it a proper doc, but here it goes)

The building blocks

Domain Events

Events represent change. Specifically a change that has already happened. They are immutable. Events are used inside Entities to apply the change to it's internal state and are propagated throughout the system to notify interested parties that something happened.

Here's is a simple event implementation that represents the fact that a task was completed:

import { Guid } from '@cashfarm/lang';
import { IDomainEvent } from '@cashfarm/plow';

class TaskCompleted {
        public readonly taskId: Guid,
        public readonly done: boolean
    ) {}

Entity and AggregateRoot<TId> classes

You model your domain using Entities and Aggregate Roots (AR). Only AR classes should be used to modify the system state. Aggregate Roots should protect their internal state, expose semantic methods and modify their internal state by applying events. Here's a simple example of what an AR should look like:

import { Guid, isEmpty } from '@cashfarm/lang';
import { AggregateRoot, Apply, IDomainEvent, Event } from '@cashfarm/plow';

import { TaskCompleted } from '../events';

class Task extends AggregateRoot<Guid> {
    private _description: string;
    public get description(): string {
        return this._description;

    private _done: boolean;
    public get done(): boolean {
        return this._done;

    constructor(desc: string) {
        if (isEmpty(desc)) {
            throw new DomainException('Task description cannot be empty');

        this._desc = desc;
        this._done = false;

    public complete() {
        this.applyChange(new TaskCompleted(true));

    public [Apply(TaskCompleted)](evt: TaskCompleted) {
        this._done = evt.done;


Repositories are how we call the classes that permanently store aggregate roots. If you are using GetEventStore to store your aggregates, all you need to do is extend the class EventSourcedRepositoryOf<TAggregate, TAggtId>;

If you are using Plow's container, you can simply inject IEventStore and IEventBus as they are already configured in the inversify container.

export class TaskRepository extends EventSourcedRepositoryOf<Task, Guid> {
    @inject(IEventStore) protected storage: IEventStore,
    @inject(IEventBus) eventBus: IEventBus
  ) {
    super(storage, Campaign, eventBus);


A projection class implements methods that handle events with the intent of projecting the data to a read model. The read model is a denormalized databsed optimized for reading.

Here's an example of a projection that sums up the number of completed tasks

import { Handle } from '@cashfarm/plow';

import { TaskCompleted } from '../events';

@Projection(TaskCompleted) // This decorator will register the projection in the event bus
export class TaskProjections {
      @inject(TaskStore) private tasks: TaskStore,
      @inject(UserStore) private users: UserStore) {}

  public [Handle(TaskCompleted)](event: TaskCompleted): void {
    debug('Running projection for TaskCompleted');
    const task = this.tasks.findById(event.taskId);
    const user = this.users.findById(task.userId);

    user.completedTasks += 1;




  • Plow now has an internal inversify container for dependency injection
  • Use @Projection, @Controller and @DomainEvent to register the respective classes
  • No need for [Symbol.EventLoader]() and [Symbol.MAPPER]() methods



  • create a generator (yeoman or vscode) to generate boilerplates for:
    • Stores and Repositories
    • events
    • AggregateRoots
    • Endpoints
    • Projections

If you find any bugs or have a feature request, please open an issue on github!

The npm package download data comes from npm's download counts api and package details come from npms.io.