This is the documentation for take-home test project for Ubersnap Middle Backend Position.
The requirement is based on this Gist
This program is writen in Go, with Fiber as the Web Framework, and uses FFMPEG for the image processing functionality.
I use Linux for making this project, so I can't write a guide on Windows/MacOS, but for MacOS I believe the steps is almost similar.
This project needs few dependencies to be installed, such as:
- Go (the version I use is 1.22.1)
- FFMPEG
- Docker or Podman (optional if you want to run it in a container)
Also before you compile it please add the following environment variables to your .env file:
- copy the .env.example to .env
- update the values
The .env variables needed is just PORT at the moment, but in the future is a good practice to have sensitive variables in the .env file instead of hard coding them in the code.
Please install Go and FFMPEG according to your OS, Then go to the root of the project and run
go build main.go.
It will download all the dependencies and compile the project.
Then once you see the main binary, run it with ./main
For containerizing the project, you need to use Docker or Podman installed.
In your terminal, run:
podman build . --tag "<IMAGE-NAME>"
It will download base image, and the dependencies and compile the project.
podman run -p 8080:8080 --env-file=.env <IMAGE-NAME>
That will port forward the port 8080, passed to the .env file and run the project in the container.
For testing purposes, I provide Hoppscotch json file to be used for testing, it's a Postman alternative that I personally used.
For importing the json to Hoppscotch, please go to this docs
I provided the necessarry Request Body and Request Header for testing purposes.
For the unit test, I only cover the converter directory because that's the main functionality lies, for the testing I just run go test ./converter
This project is not perfect, there are few things that I would like to improve.
- Improving the route handler, to make it neater.
- The output image is written in
/tmp/, which is not ideal, for the production code use S3 or any other cloud storage to serve it. - This project isn't tested in Cloud, so it's not production ready.
- This docs is written in rush, if I have more time I can write it in a more structured way.
- This project is not reviewed by anyone, so I can't tell what else I need to Improve.