Parolla Bot

Use cases:

1. 1&1 conversation
2. Group discussion intervention
3. Translation
4. Configuration profil (roadmap or payment or learning strategy)

Services

1. Handle new message
2. Handle user join event
3. Handle update profile request (from UI)
4. Handle payment

Project architecture

The app consists of several domains(like users, items, etc.) Each domain has 3 layers of abstraction:

1. Handlers layer. Responsible for creating API route function, validation input user data, and creating/closing database connection at the start/end of each request.
2. Service layer. It's a place for all business logic and additional validation.
3. Database layer. Responsible for creating actual SQL queries and executing them.

Tools

• FastAPI
• SQLAlchemy 2.0
• Poetry
• Linters and formatters: black, isort, ruff
• Commands: Typer
• pytest
• Docker
• Celery and Redis
• Sentry
• Kafka(WIP)

Before first commit

poetry install --only=dev
pre-commit install

Environment variables

Environment variables for development can be found in .env.example. You must create app/.env file and fill missing variables with value SET_VALUE from stage environment.

Run application

• Using docker compose or prepared commands from makefile

docker compose up app

make dc-up

• Using local shell

1. Create virtual environment

poetry shell

2. Install all dependencies

poetry install --no-root

3. Start application

docker/start.sh

Run worker

• Using docker compose

docker compose up worker

Tests

• Using docker compose with any pytest flag or prepared makefile

docker compose run --rm test -s

make dc-test

make dc-test tests/integration

make dc-test-parallel

make dc-test-unit tests/unit

# run test and restart if code base is changed 
make dc-test-watch 

• All test running using pytest
• Option --reuse-db allow to use the same database without deleting it
• Parallel tests allow to run tests in parallel procceses. For each proccess created it's own database based on worker_id
• Option --without-db allow to run withou creating database and only for unit tests

Database migration

• autogenerate migration

docker compose run --rm migrations alembic revision --autogenerate -m 'Migration message'

make dc-migrations 'Migration message'

• upgrade to latest version

docker compose run --rm migrations upgrade heads

make dc-migrate

• upgrade to previous version

docker compose run --rm alembic downgrade -1

make dc-migrate-rollback

Run linters

• Run only check

make lint-check

• Run linters and modify files

make lint

Acknowledge

This project has been inspered a lot by the 2 following project. To create a clean architecture with a bit less of dependencies and fewer layer of complexity.

• https://github.com/t1m4/fastapi_template
• https://github.dev/v-v-d/carts