octonato / shopping-cart-scala

Lagom Scala shopping cart sample application

Shopping Cart

This sample application demonstrates a simple shopping cart built with Lagom. It contains two services, a shopping cart service, for managing shopping carts, and an inventory service, for tracking inventory.

The shopping cart service persists its data to a relational database using Lagom's persistence API, and is intended to demonstrate how to persist state using Lagom.

The inventory service consumes a stream of events published to Kafka by the shopping cart service, and is intended to demonstrate how Kafka event streams can be consumed in Lagom. However, it doesn't persist its state to a database, it just stores it in memory, and this memory is not shared across nodes. Hence, it should not be used as an example of how to store state in Lagom.

Setup

You will need to install PostgreSQL. In development, the application expects it to be running on localhost on the default port (9042), and it expects there to be a database called shopping_cart, with a user called shopping_cart with password shopping_cart that has full access to it. This can be created using the following SQL:

CREATE DATABASE shopping_cart;
CREATE USER shopping_cart WITH PASSWORD 'shopping_cart';
GRANT ALL PRIVILEGES ON DATABASE shopping_cart TO shopping_cart;

Once PostgreSQL is setup, you can start the system by running:

sbt runAll

Shopping cart service

The shopping cart service offers three REST endpoints:

• Get the current contents of the shopping cart:

  curl http://localhost:9000/shoppingcart/123
  

• Update the quantity of an item in the shopping cart:

  curl -H "Content-Type: application/json" -d '{"productId": "456", "quantity": 2}' -X POST http://localhost:9000/shoppingcart/123
  

• Checkout the shopping cart (ie, complete the transaction)

  curl -X POST http://localhost:9000/shoppingcart/123/checkout
  

For simplicity, no authentication is implemented, shopping cart IDs are arbitrary and whoever makes the request can use whatever ID they want, and product IDs are also arbitrary and trusted. An a real world application, the shopping cart IDs would likely be random UUIDs to ensure uniqueness, and product IDs would be validated against a product database.

When the shopping cart is checked out, an event is published to the Kafka called shopping-cart by the shopping cart service, such events look like this:

{
  "id": "123",
  "items": [
    {"productId": "456", "quantity": 2},
    {"productId": "789", "quantity": 1}
  ],
  "checkedOut": true
}

Inventory service

The inventory service offers two REST endpoints:

• Get the inventory of an item:

  curl http://localhost:9000/inventory/456
  

• Add to the inventory of an item:

  curl -H "Content-Type: application/json" -d 4 -X POST http://localhost:9000/inventory/456
  

The inventory service consumes the shopping-cart topic from Kafka, and decrements the inventory according to the events.