This is a generator for metadata according to the MOOChub format.
The application is made to be run un Ubuntu 22.04. Make sure the system is up-to-date.
sudo apt update && sudo apt upgrade
The source code can be cloned using Git.
sudo apt install git
The application is a Docker image. Docker can be installed via snap.
sudo snap install docker
Get the data from the GitHub repository:
git clone https://github.com/MaxThomasHPI/MetadataGen.git
Make sure to set all environment variables in the .env file before build/start the docker containers.
Since the program uses the Gemini-API from Google, a valid key for this application needs to be provided. It needs to be stored in an environment variable called "GEMINI_KEY". The key has to work with the gemini-2.5-flash model. Add the key in the .env variable at the root level.
# .env
GEMINI_KEY=<Your Gemini-API key>
There is a database running in the background to log input data and the suggestions made. The database is set up in a local bind and will take environment variables as provided in the .env file. This includes a username for the database as well as a password. A default database called "suggestions" will be set up as well. Please do not change the name of the database.
# .env
POSTGRES_USER=<your-psql-username>
POSTGRES_PASSWORD=<your-psql-password>
POSTGRES_DB=suggestions # the default database to be created at startup, do not change!
The repository contains a Dockerfile and a docker-compose.yml for setting up an image. Go to the directory containing the Dockerfile/docker-compose.yml and run:
sudo docker compose up --build
The application should start but can also be started with:
sudo docker compose up
after successfully building the image.
The application will run by default on Port 5000 (default Flask). Port 80 will be exposed as stated in the Dockerfile. In the docker-compose.yml the mapping (80:5000) is configured. Also, the nginx.conf contains the mapping. If you prefer another configuration make sure to update these files accordingly. With this setup the application frontend can be reached via accessing http://localhost in the browser.