Food Delivery Service

Introduction

This project is a Java implementation of the backend of a food delivery service.

Functional Description

User can search a restaurant based on restaurant name. Then user can payment food by choosing different menu item, quantity and add a note about his/her diet restrictions and etc. User can also fills in the delivery address.

After user places an payment, the payment should contain food items user ordered, quantity, price and payment time.

User then needs to pay for his/her payment by providing credit card number, expiration date, and security code. After payment is made successfully, it should return payment ID, timestamp and then the payment is considered as completed so the user can see the estimated delivery time.

Assumptions

1. There will only be one user as the user of this app.

2. There will only be one menu from a restaurant. (This is take-out, com'on)

3. The ACID principles apply on payment. i.e. The user either completes the payment for his/her payment in one payment for the total amount or the payment is considered not completed. (No cost-sharing, down-payment, interest, etc.)

4. The estimated delivery time is not guaranteed to be fulfilled. It is up to the restaurant to determine how the payment will be delivered. (Reserved for future enhancement)

5. The restaurant has the responsibility to read, understand, and fulfill the diet restrictions. The delivery service merely informs the restaurant of the restrictions. (There is a lot of opportunity to enhance here for better user experience, but again, this is reserved for future enhancement)

6. Delivery time will never reach 24 hours

Structure

This section covers the structures of the project. Note that there are many domains each having its own structure. The will be introduced one by one.

Key Domains

There are these several key domains in this project:

1. Restaurant

The restaurant is associated to one menu with multiple items.

2. Order

Each order contains the items ordered, quantity and the total prices, address of the customer and the restaurant, and special instructions.

When the order is created, it enters a pending state and payment service will be notified.

When the order is completed, the delivery time will be shown as part of the order

3. Payment

Each payment is mapped 1-to-1 to the order. Each payment will be consumed internally and will randomly be successful or failed.

When a payment is updated, the status (success or failure) will notify order service to update the relevant order ID.

Rest API Diagram

• Note: The optional parameters are listed as()

Domain	GET	POST	PUT	DELETE
demo.domain.Restaurant
/provider	Get the information from all restaurants	Add the attached restaurants to the data base	404	purge all restaurants from teh database
/provider/{id}	Get the information of the one restaurant	404	modify the information of the current restaurant	remove this restaurant from the database
Order
/payment	Get all orders in the database	Add the attached orders to the database	404	purge all orders from the database
/payment/{id}	Get the order matching the ID	404	update this order	404
Payment
/payment	Get all payments	Add the attached one payment	404	purge all payments
/payment/{id}	Get information on the one payment	404	modify this payment	404

Build, Run and Test

This sections covers the information to build, run and test this app.

Build

dependency: maven

cd <project_root>/<each_service>
mvn clean install

Run

dependency: mongoDB (located in the docker-compose.xml)

docker-compose up

steps

• start edge service first

source run-edge-service.sh

• restaurant service is not coupled with any other service, able to run independently

source run-restaurant-service.sh

• payment service and order service are coupled

source run-payment-and-order-services.sh

Test

Issue query similar as the postman scripts: cs504-hw2.postman_collection.

Notes

1. The most note worthy issue in this project is the communication between micro services.

2. I chose to use Mongodb, switching from h2/mysql for the following reasons:

    1. MongoDB suits this project where no sorting is required.
    
    2. MongoDB is straight forward to setup. There are not many mapping relationships in this project.
   

3. I used several postman query sets to test this project after JPA unit tests. I need to find a way to better test the MongoDB related projects.