bcbecker / Flask-GraphQL-API

Flask API using Graphene to query a MongoDB database. Containerized with Docker

Flask-GraphQL-API

Flask API using Graphene to query a MongoDB database. Containerized with Docker.

A simple learning project for Mongo and GraphQL (Graphene).

Table of Contents

• General Info
• Technologies Used
• Project Structure
• Features
• Setup with Docker
• Querying the API
• Improvements/To Do

General Information

In the past, I've primarily dealt with REST APIs and SQL databases. So, I felt it was time to dive into GraphQL and a NoSQL database. In this case, Graphene and MongoDB. This project started with the basic graphene-mongo tutorial, and has grown to include more queries and create/update/delete mutations for each Mongo document.

The testing is still in progress, as I've found testing GraphQL to be a tad tricky. As I learn, I'll continue to update the API, as I feel this is the best way to learn.

Technologies Used

• Flask, with the Flask-GraphQL extension

  • Flask-Mongoengine as the ORM
  • Graphene-mongo for GraphQL schema/querying
  • pytest, coverage, and mongomock for testing
  • gunicorn as the production WSGI server

• MongoDB, with prepopulated data

• Docker for containerization

Project Structure

~/flask-graphql-api
    |-- docker-compose.yml
    |__ /backend
        |-- Dockerfile.api
        |-- run.py
        |-- populate_db.py
        |-- config.py
        |-- requirements.txt
        |-- .env
        |__ /api
            |__ /department
                |-- __init__.py
                |-- mutations.py
                |-- types.py
            |__ /employee
                |-- __init__.py
                |-- mutations.py
                |-- types.py
            |__ /role
                |-- __init__.py
                |-- mutations.py
                |-- types.py
            |-- __init__.py
            |-- models.py
            |-- routes.py
            |-- schema.py
        |__ /tests
            |-- __init__.py
            |-- test_api.py
    |__ /mongodb
        |-- Dockerfile.mongo
        |__ /data
            |-- ...

Features

• GraphQL API querying a NoSQL dataset
  • Query all employees, departments, or roles
  • Query individual employees, departments, or roles
  • Mutate (create, update, delete) employees, departments, or roles

Setup with Docker

Ensure you have docker installed.

docker --version

Creating .env file

For this server to run, you must create a .env file within the package, setting the following parameters:

cd backend
touch .env

These are the .env variables, if your host/port/url etc vary, change it accordingly. Note that if you use Docker, your HOST will be the name of your service as defined in the docker-compose file, not localhost.

SECRET_KEY=
FLASK_APP=run.py

MONGODB_DB=<DB_NAME>
MONGODB_HOST=<DB_HOST>
MONGODB_PORT=<DB_PORT>

Set the secret key(s) to whatever you'd like (though they should be secure for production).

Running the Server

Start the docker daemon and build the containers (sudo if Linux). You may not need to use sudo, or start the daemon, depending on your configuration.

sudo dockerd
sudo docker-compose up -d --build

To ensure everything is up and running, run docker ps

sudo docker ps

Start/stop containers using the following:

sudo docker container <start/stop> <container-name>

Or, stop and remove running containers with:

sudo docker-compose down

Viewing the website

Site can be found here: http://127.0.0.1:5000 (for localhost) GraphiQL interface at: http://127.0.0.1:5000/graphql

Querying the API

Queries and mutations can be executed via the GraphiQL interface, or testing tools like Postman. To query all employees, returning all information including department, roles, manager, and tasks:

{
  allEmployees {
    edges {
      node {
        id,
        name,
        hiredOn,
        department {
          id,
          name
        },
        roles {
          edges {
            node {
              id,
              name
            }
          }
        },
        manager {
          id,
          name
        }
        tasks {
          edges {
            node {
              name,
              deadline
            }
          }
        }
      }
    }
  }
}

TODO

TODO:

• Fix task mutations
• Update testing