RSS reader

API to manage RSS feeds.

• subscribe users
• follow/unfollow feeds
• mark feed items as read/unread

Background (scheduled) worker to fetch feeds items.

Stack

• Python 3.9 + FastAPI + pipenv
• OAuth2 authentication with user/password
• Docker container to run app and its deps
• MongoDB as the domain storage and tasks control
• Dramatiq to run the background workers
• Redis as dramatiq message broker
• Apscheduler for the task scheduling
• Pytest for integration tests

Environment

Application loads env vars from a .env file in src/ folder.

Variable	Description	Example
FEEDS_UPDATE_INTERVAL	Interval in minutes to update feeds.	15
JWT_SECRET	Secret to encrypt user passwords.	26faa9...58d95d
MONGO_URL	Database url.	mongodb://127.0.0.1:21017
MONGO_DB_NAME	Database name.	rss_reader
TOKEN_EXPIRATION_MIN	Token expiration time in minutes.	15

JWT_SECRET can be randomly generated with openssl.

$ openssl rand -hex 32

Install and run

Use docker-compose to execute application and its dependencies altogether.

First go to src/ folder and create an updated requirements.txt file.

$ pipenv lock -r > requirements.txt

Build and run.

$ docker-compose build
$ docker-compose up -d

Server URL:

• http://127.0.0.1:8000/

Docs URL (OpenAPI):

• http://127.0.0.1:8000/docs

Development mode

Run mongo and redis in background.

$ docker-compose up -d mongo redis

Go to src/ folder, install all deps and run.

# on terminal 1

$ pipenv install --dev
$ pipenv run workers

# on terminal 2

$ pipenv run app

Test

Go to src/ folder, install dev deps and run.

$ pipenv install --dev
$ pipenv run test

Pylint

Static code analysis is made by pylint.

$ pipenv run lint

Improvements

List of things that can be improved:

• Cache. To improve the performance of the API requests.
• Service layer. That layer would help to abstract a little bit more the logic present in routers.
• Pagination. To go through items listed from the database.
• DB transactions. Started to implement it, but left the initial code in a separate branch.
• Events-driven architecture. Might be useful, but will add substantial complexity.
• More tests. Especially for the background workers.
• Documentation. Code and system documentation.