CMSocial

A web application that builds a social coding platform used by https://training.olinfo.it/

Currently, it heavily depends on CMS.

The backend (in the backend/ folder of this repository) is written in Python + SQLAlchemy, structured similarly to CMS and using CMS under the hood. The frontend is in TypeScript and it's based on NextJS, TailwindCSS and DaisyUI.

Run a local development instance

Make sure you have docker installed. Clone this repository with --recursive and then run:

$ docker compose -f docker-compose.dev.yml up --build dev

The command will create a few containers:

• cmsocial-dev_db-1 - Postgres database instance.
• cmsocial-dev_cms-1 - CMS instance (exposed as /admin via the proxy).
• cmsocial-dev_backend-1 - Python backend (exposed as /api via the proxy).
• cmsocial-dev_frontend-1 - NextJS frontend (exposed as / via the proxy).
• cmsocial-dev_old_frontend-1 - (Deprecated) AngularJS frontend (exposed as /old via the proxy).
• cmsocial-dev-1 - Nginx proxy instance, running on localhost:8000.

When the backend runs for the first time, it will call cmsocialInitDevDB to initialize the database with some sample data, among which: an admin account with username admin and password admin, and a user account with username test and password password.

You can access your local instance by visiting http://dev.olinfo.it:8000/. Don't use localhost:8000 directly, or your browser will complain about CORS.

Deleting the database

If you want to start fresh with a clean database, just delete the container with:

docker rm -f cmsocial-dev_db-1

And then restart the other containers (mostly the backend one, to re-run cmsocialInitDevDB).

Avoid running out of disk space

When you make changes and run multiple times the docker compose command to spin up the containers, it's possible that docker will accumulate many unused containers / images. You can prune old containers / images by running the following command periodically:

docker system prune

Run the frontend only

If you just want to work on the frontend, connecting to the production backend which runs on https://training.olinfo.it, you can avoid using docker.

Make sure you can run pnpm on your system, then you can simply run pnpm dev to start the development server on http://dev.olinfo.it:3000/. Note that the port (3000) is different from the one used by the docker instructions above (8000).

Contributing

Before committing, make sure to run pnpm lint to check for linting errors.

If you've added or changed text, you also need to update the translations:

1. Run pnpm translate to extract the new messages.
2. Update src/locales/en/messages.po with the new translation.

Deployment

Read instructions at: https://github.com/olimpiadi-informatica/infrastructure?tab=readme-ov-file