AbinavRavi/simple-prod-api

A simple example of a production level API and list out components that make it useful for production

simple-prod-api

A simple example of a production level API and list out components that make it useful for production

ENVIRONMENT SETUP

1. Please create a virtual environment either by using python virtualenv or conda.
2. Once the virtualenv is created then install poetry for dependency management

pip install poetry

3. Once poetry is installed then install dependencies by just running

poetry install

4. This project makes use of pre-commit hooks to maintain style and quality of codeso before doing any development please run

pre-commit install

This will not allow to push code that doesnt meet standards mentioned in the pre-commit-config.yaml file

HOW TO RUN THE CODE

1. Inside the environment you can startup the server now using uvicorn ASGI server on your localhost using the following command

uvicorn src.main:app --reload (if not in dev environment please dont use the reload flag)

2. Please go to the browser and go to 127.0.0.1:8000/docs from which we can perform a secure login and access the API
3. To perform secure login please choose

/token POST endpoint

and click on try it out after expanding on the endpoint. 4. In the username and password fields please enter the following

username: admin
password: admin 

for testing purposes. This will return an access token and token_type as bearer 5. Once you have logged in you can now access the functionality endpoint

/get_array POST endpoint

click on the lock symbol and re enter the username and password as mentioned in the previous step, now we have authenticated our access to the endpoint

6. Since the app takes a sentence and generates 500 random numbers the sentence input can be provided as a json using either curl or postman or a client of your choice

curl -X POST -H "Content-Type: application/json" -d '{"sentence": "Hello World"}' http://localhost:8000/get_array

or use the request body on the UI and modify the sentence. This should provide a correct response on the terminal 7. To run the tests locally you can run

poetry run pytest tests/

Features added to make the functionality production ready are

1. pre-commit hooks to maintain quality and style of code
2. tests that can be run on development environment and also ci using github actions.
3. The github actions will run for every commit pushed into the repository, since this project is very small no limitations has been placed on the master branch.
4. Typing annotations have been used in order to understand the input and output types, some of the custom types are created using pydantic which comes inbuilt with fastapi.
5. Instead of storing dependencies in a requirements.txt I have made use of poetry package management which comes with poetry.lock file for easy reproduction of the environment and the same has been used in the CI to replicate build and testing.
6. A health_check endpoint has been added in addition to the functionality endpoint so that there can be checks on whether the issue is with connection to the server or the problem is with a specific endpoint

Tests

The test suite mainly consists of 3 files

1. conftest.py - which issues a client setup across the test suite and changes the root folder into src/
2. pytest.ini - contains a list of modules grouped by names, currently there is only one which is for endpoint checks
3. test_endpoints.py - contains test that validate the response and status codes of the health_check endpoint and the functionality of get_array endpoint.