Jose-Theam-CRM

This project implements a CRM solution for a shop with Users, Admins and Customers.

Tech Stack

This project is being developed with:

• Gradle
• SpringBoot + Kotlin
• PostgreSQL
• Flyway
• Docker compose

How to run it locally

Using the generated local docker image

1. Run the script ./scripts/create-docker-image.sh which will create a local docker image of the app.
2. Run docker-compose up which will spin up a dockerized Posgres database and then the Spring boot application.
3. Run docker-compose down at any moment to stop both containers.

Using your own environment

1. Load the Environment variables defined in the file .env into your environment.
2. Run docker-compose up postgres_db to spin up the PSQL database docker container.
   • Changing the values in the .env file will affect both to docker compose and the Spring app, so feel free to modify and customize it to your machine.
3. Run ./gradlew bootRun.
4. Your application should be running and listening on the port 8080.

How to run all tests

• This project is using TestContainers, which will spin up a PSQLContainer at the start of the test cycle using a random port.

1. Run ./gradlew test

Documentation

You can find some examples to the different API endpoints under /docs/postman.

About Image uploading

Image uploading for rest APIs is a challenge itself since they are supposed to operate with JSON, therefore multipart files uploads are not so easy (as they require multipart/form-data header for requests). After some investigation I found out that a common way to handle file uploads through Base64 encoded strings, which can be part of a valid JSON body request. This encoded string tends to be quite consume around a 25% extra bandwidth, but in this case I weighed design over performance.

Also, to avoid sending a huge request at the moment of creating a customer, the image can be uploaded in a request beforehand and have it ready to be attached on customer creation.

Therefore, in this application the flow to upload a picture to a Customer is the following:

1. Upload a picture MIME Base64 encoded String through a POST request to /pictures.
   • Its up to the client to create the Base64 encoded string with MIME headers. It will be validated before saving it.
   • Answer from this request will include a pictureId.
2. At this point, when creating or updating a Customer the picture can be attached to it including the pictureId in the request body.
   • This field is optional, meaning that customers can be created without a Picture. That can be easily changed, but I decided to keep customer creation making pictures not mandatory.
   • This has the drawback of having uploaded pictures not attached to any Customer entity. An easy solution for this would be a scheduled job cleaning those from time to time.

Assumptions are that users have access to all customer's pictures.