HAT - Human Assisted Translator project

This is a CAT to work with texts. Now in experimental stage. Uses Python and FastAPI as a backend library. For a frontend Vue with its stack is used.

Running the backend

To run the backed navigate to backend directory and create virtual environment:

python3 -m venv .venv
source .venv/bin/activate

Install all dependencies:

pip install -r requirements.txt
python3 -m nltk.downloader punkt_tab

Then run FastAPI server:

uvicorn asgi:app

You will need PostgreSQL server running on a local machine. Set up connection string to the environment variable DATABASE_URL or update settings.py file.

Running backend's tests

To run tests you need to install pytest and pytest-asyncio packages. They are already listed in requirements.txt file, so you if you installed all dependencies from a previous section you are good to go.

It order to run tests you need to run pytest command from the backend directory.

pytest

Backend linting

To run linting for backend code execute the following command:

ruff format . && ruff check --select I --fix . && ruff check  --fix .

Running the worker

To run the worker navigate to backend directory and create virtual environment if not yet:

python3 -m venv.venv
source.venv/bin/activate

Then install all dependencies:

pip install -r requirements.txt
python3 -m nltk.downloader punkt_tab

Then run the worker:

python3 worker.py

The worker is being tested by the same test suite as the backend.

Running the frontend

To run the frontend you need to install all dependencies first. Navigate to frontend directory and run:

pnpm install

Then you can run the development server:

pnpm run dev

This will start the development server on localhost:5173 address. Open your browser and check the website. When development server is run you need to run backend server as well.

Running frontend's tests

To run frontend's tests you need to run:

pnpm run test

Running the production version

Before setting up the production version you need to adjust environment configs. To do this copy .env.example to .env and fill in all needed variables.

Also you have to adjust alembic.ini file to match your database settings. It is located in backend directory.

To run production version of the tool you need to use docker compose:

docker compose up -d --build

You can have two separate .env files for local testing and for production deployment. To do this make two copies of .env.template with different names: .env-local for local testing and .env-prod for production and specify which file to use by setting --env-file parameter when running docker-compose.

docker compose --env-file .env.prod up -d --build

This will build all needed images and run services in detached mode. Please note that reverse proxy is listening to 6916 port on the localhost, so you need to set up something to listed on the domain name if you need.

The production version is located at https://hat.codecliffs.ru

Getting into database in Docker compose deployment

There is a PGAdmin instance as a container which is set up to automatically connect to the database in another container. You should run docker compose with additional profile named debug like this:

docker compose --env-file .env-prod --profile debug up -d --build

After everything is running you can access PGAdming instance on http://localhost:8082/ address. It will automatically set up and lets you use it without login and password. This set up might be used to debug database issues or do mundane maintenance tasks without having to use CLI.

Since it is set up to be accessed by anyone who has an access to the server, make sure you've removed a container after doing debugging or limit access to people with such access.