iulianav / chatbot-api

The project represents an API used by data scientists to further improve an existing chatbot. Requests are sent by a background job which pushes the customer data and consent via HTTP.

chatbot-api

The project represents a Data API used by data scientists to further improve an existing chatbot. Requests are sent by a background job which pushes the customer data and consent via HTTP.

How to run

To run this project, you need to have Docker and Compose installed.

The docker-compose commands are already abstracted through Makefile rules for simplicity. Run the following commands to:

1. Build the app: make build-app
2. Run the app: make run-app
3. Shut down the app: make kill-app

Now the Data API app should be running by default at http://0.0.0.0:80, together with a PostgreSQL database that it is connected to. The app and the database each run in their own container.

The default database's settings can be found in .env. Feel free to change those according to your needs by having your own .env file locally.

To remove the persistent volume created by the database container, and hence delete the database contents, run:

1. Find db volume name: docker volume ls
2. Delete db volume: docker volume rm <volume_name>

Development

To further develop the project, you should create a virtual environment with virtualenv or anaconda, and then install both the project requirements and the development requirements inside of it:

1. Create environment: conda create <env_name> python=3.9
2. Activate environment: conda activate <env_name>
3. Install requirements : pip install -r requirements.txt
4. Install development requirements: pip install -r requirements-dev.txt

You can find additional Makefile rules to help with checking the code quality in makefiles/Makefile.python:

1. Sort imports: make isort
2. Format code: make black
3. Lint code: make lint
4. Type checking: make analyze

Manual testing

To manually test the API, you can go to http://0.0.0.0:80/docs while the app is running. It will display a default UI to manually test each endpoint by inputting valid data. The UI also displays the status codes and responses.

To check the content of the database after each request, you should:

1. Enter the db container: docker exec -it postgres-db bash.
2. Connect to PostgreSQL as user postgres: psql -U postgres
3. Get access to the chatbotapi db: \c chatbotapi
4. Retrieve the records of interest after each request: e.g. SELECT * from user_inputs;

Sometimes, especially after deleting the database persistent volume, there might be an error stating that the relation user_inputs does not exist. In this case: kill, re-build and re-start the app.

Unit testing

To run the unit tests in the development virtual environment, run the command: python -m pytest tests

TODO

• Add docstrings & comments