App that fetchet tweets from Twitter and shows their basic statistics.

Background task fetches k recent tweets by phrase. Exposed API allows to see those tweets, their top hashtags, top user who made max amount of tweets, total amount of recently fetched tweets.

• python 3.7
• postgres 11+
• redis

Background task performed with schedule library, which is lightweight and simple in comparison to celery.

API is created with starlette and hosted with uvicorn

All data for API responses is cached in redis.

• Clone repository

:$ git clone https://github.com/pgtuk/tweets.git

:$ cd tweets

• Create .env

:$ cp .env.example .env

• Put your twitter consumer key to TWITTER_CONSUMER_KEY and secter to TWITTER_CONSUMER_SECRET in .env. Don't hesitate to change querying phrase PHRASE, tweets count TWEETS_COUNT and fetching interval TIME_INTERVAL. Check limits at Twitter API docs

:$ docker-compose up

Note: docker will run linters before starting app, but not tests - requires to add test database.

App will be available at '0.0.0.0:5000' by default

• If you want to run it locally you'll need python3.7, postgres and redis. Set up databases for deployment and tests. Test DB name assumed to be same as deployment db's name prefixed with test_.
• Set up python virtual enviroment and install package dependencies

$: python3.7 -m venv venv

$: venv/bin/pip install -r requirements.txt

• Use make to run linters, tests and app itself. See make help for more information.