Python CRUD Rest API using Flask, SQLAlchemy, Postgres, Docker, Docker Compose

Let's create a CRUD Rest API in Python, using the following:

Flask (Python web framework)
SQLAlchemy (ORM)
Postgres (database)
Docker (containerization)
Docker Compose (to run the application and the database in containers)

If you prefer a video version:

All the code is available in the GitHub repository (link in the video description): https://youtube.com/live/fHQWTsWqBdE

🏁 Intro

Here is a schema of the architecture of the application we are going to create:

We will create 5 endpoints for basic CRUD operations:

Create
Read all
Read one
Update
Delete

Here are the steps we are going through:

Create a Flask application using SQLALchemy as an ORM.
Dockerize the Flask application writing a Dockerfile and a docker-compose.yml file to run the application and the database
Run the Postgres database in a container using Docker Compose, and test it with TablePlus
Run the Flask application in a container using Docker Compose, and test it with Postman

We will go with a step-by-step guide, so you can follow along.

🏁 Create a Flask application using SQLALchemy as an ORM

Create a new folder:

mkdir flask-crud-api

step into the folder:

cd flask-crud-api

Open the folder with your favorite IDE. I am using VSCode, so I will use the command:

code .

We need just 4 files for the Flask application, including containerization.

You can create these files in different ways. One of them is to create them manually, the other one is to create them with the command line:

touch requirements.txt app.py Dockerfile docker-compose.yml

Your folder structure should look like this:

🗒️requirements.txt file

The requirements.txt file contains all the dependencies of the project. In our case we will need just 3.

Let's add them to the requirements.txt file:

flask
psycopg2-binary
Flask-SQLAlchemy

Short explanation of the dependencies:

flask is the Python web framework we are gonna use.

psycopg2-binary is the driver to make the connection with the Postgres database.

Flask-SQLAlchemy is the ORM to make the queries to the database.

🐍 app.py file

The app.py file is the main file of the application: it contains all the endpoints and the logic of the application.

Populate the app.py file as follows:

from flask import Flask, request, jsonify, make_response
from flask_sqlalchemy import SQLAlchemy
from os import environ

app = Flask(__name__)
app.config['SQLALCHEMY_DATABASE_URI'] = environ.get('DB_URL')
db = SQLAlchemy(app)

class User(db.Model):
    __tablename__ = 'users'

    id = db.Column(db.Integer, primary_key=True)
    username = db.Column(db.String(80), unique=True, nullable=False)
    email = db.Column(db.String(120), unique=True, nullable=False)

    def json(self):
        return {'id': self.id,'username': self.username, 'email': self.email}

db.create_all()

#create a test route
@app.route('/test', methods=['GET'])
def test():
  return make_response(jsonify({'message': 'test route'}), 200)


# create a user
@app.route('/users', methods=['POST'])
def create_user():
  try:
    data = request.get_json()
    new_user = User(username=data['username'], email=data['email'])
    db.session.add(new_user)
    db.session.commit()
    return make_response(jsonify({'message': 'user created'}), 201)
  except e:
    return make_response(jsonify({'message': 'error creating user'}), 500)

# get all users
@app.route('/users', methods=['GET'])
def get_users():
  try:
    users = User.query.all()
    return make_response(jsonify([user.json() for user in users]), 200)
  except e:
    return make_response(jsonify({'message': 'error getting users'}), 500)

# get a user by id
@app.route('/users/<int:id>', methods=['GET'])
def get_user(id):
  try:
    user = User.query.filter_by(id=id).first()
    if user:
      return make_response(jsonify({'user': user.json()}), 200)
    return make_response(jsonify({'message': 'user not found'}), 404)
  except e:
    return make_response(jsonify({'message': 'error getting user'}), 500)

# update a user
@app.route('/users/<int:id>', methods=['PUT'])
def update_user(id):
  try:
    user = User.query.filter_by(id=id).first()
    if user:
      data = request.get_json()
      user.username = data['username']
      user.email = data['email']
      db.session.commit()
      return make_response(jsonify({'message': 'user updated'}), 200)
    return make_response(jsonify({'message': 'user not found'}), 404)
  except e:
    return make_response(jsonify({'message': 'error updating user'}), 500)

# delete a user
@app.route('/users/<int:id>', methods=['DELETE'])
def delete_user(id):
  try:
    user = User.query.filter_by(id=id).first()
    if user:
      db.session.delete(user)
      db.session.commit()
      return make_response(jsonify({'message': 'user deleted'}), 200)
    return make_response(jsonify({'message': 'user not found'}), 404)
  except e:
    return make_response(jsonify({'message': 'error deleting user'}), 500)

Explanation:

We are importing:

Flask as a framework
request to handle the HTTP
jsonify to handle the json format, not native in Python
make_response to handle the HTTP responses
flask_sqlalchemy to handle the db queries
environ to handle the environment variables

We are creating Flask app, configuring the database bu setting an environment variable called 'DB_URL'. We will set it later in the docker-compose.yml file.

Then we are creating a User class with an id, a username and an email. the id will be autoincremented automatically by SQLAlchemy when we will create the users. the __tablename__ = 'users' line is to define the name of the table in the database

An important line is db.create_all(). This will synchronize the database with the model defined, for example creating an "users" table.

Then we have 6 endpoints

test: just a test route
create a user: create a user with a username and an email
get all users: get all the users in the database
get one user: get one user by id
update one user: update one user by id
delete one user: delete one user by id

All the routes have error handling, for example if the user is not found, we will return a 404 HTTP response.

You can check a video-explanation here

🐳 Dockerize the Flask application

Let's populate the Dockerfile :

Dockerfile:

FROM python:3.6-slim-buster

WORKDIR /app

COPY requirements.txt ./

RUN pip install -r requirements.txt

COPY . .

EXPOSE 4000

CMD [ "flask", "run", "--host=0.0.0.0", "--port=4000"]

FROM sets the base image to use. In this case we are using the python 3.6 slim buster image

WORKDIR sets the working directory inside the image

COPY requirements.txt ./ copies the requirements.txt file to the working directory

RUN pip install -r requirements.txt installs the requirements

COPY . . copies all the files in the current directory to the working directory

EXPOSE 4000 exposes the port 4000

CMD [ "flask", "run", "--host=0.0.0.0", "--port=4000"] sets the command to run when the container starts

🐳🐳Docker compose

The term "Docker compose" might be a bit confusing because it's referred both to a file and to a set of CLI commands. Here we will use the term to refer to the file.

Populate the docker-compose.yml file:

version: "3.9"

services:
  flask_app:
    container_name: flask_app
    image: dockerhub-flask_live_app:1.0.0
    build: .
    ports:
      - "4000:4000"
    environment:
      - DB_URL=postgresql://postgres:postgres@flask_db:5432/postgres
    depends_on:
      - flask_db
  flask_db:
    container_name: flask_db
    image: postgres:12
    ports:
      - "5432:5432"
    environment:
      - POSTGRES_PASSWORD=postgres
      - POSTGRES_USER=postgres
      - POSTGRES_DB=postgres
    volumes:
      - pgdata:/var/lib/postgresql/data

volumes:
  pgdata: {}

We just defined 2 services: flask_app and flask_db

flask_app is the Flask application we just dockerized

flask_db is a Postgres container, to store the data. We will use the official Postgres image

Explanation:

version is the version of the docker-compose file. We are using the verwion 3.9

services is the list of services (containers) we want to run. In this case, we have 2 services: flask_app and flask_db

container_name is the name of the container. It's not mandatory, but it's a good practice to have a name for the container. Containers find each other by their name, so it's important to have a name for the containers we want to communicate with.

image is the name of the image we want to use. I recommend replacing "dockerhub-" with YOUR Dockerhub account (it's free).

build is the path to the Dockerfile. In this case, it's the current directory, so we are using .

ports is the list of ports we want to expose. In this case, we are exposing the port 4000 of the flask_app container, and the port 5432 of the flask_db container. The format is host_port:container_port

depends_on is the list of services we want to start before this one. In this case, we want to start the flask_db container before the flask_app container

environment is to define the environment variables. for the flask_app, we will have a database url to configure the configuration. For the flask_db container, we will have the environment variables we have to define when we wan to use the Postgres container (we can't change the keys here, because we are using the Postgres image, defined by the Postgres team).

volumes in the flask_db defines a named volume we will use for persistency. Containers are ephimerals by definition, so we need this additional feature to make our data persist when the container will be removed (a container is just a process).

volumes at the end of the file is the list of volumes we want to create. In this case, we are creating a volume called pgdata. The format is volume_name: {}

👟 Run the Postgres container and test it with TablePlus

To run the Postgres container, type:

docker compose up -d flask_db

The -d flag is to run the container in detached mode, so it will run in the background.

You should see something like this:

Docker is pulling (downloading) the Postgres image on our local machine and it's running a container based on that Postgres image.

To check if the container is running, type:

docker compose logs

If everything is ok, you should see something like this:

If the last line is LOG: database system is ready to accept connections, it means that the container is running and the Postgres server is ready to accept connections.

But to be sure, let's make another test.

To show all the containers (running and stopped ones) type:

docker ps -a

The output should be similar to this:

Now, to test the db connection, we can use any tool we want. Personally, I use TablePlus.

Use the following configuration:

Host: localhost

Port: 5432

User: postgres

Password: postgres

Database: postgres

Then hit "Test" (at the bottom-right).

If you get the message "connection is OK" you are good to go.

You can also click "Connect" and you will see an empty database. This is correct.

🔨 Build and run the Flask application

Now, let's build and run the Flask application.

Let's go back to the folder where the docker-compose.yml is located and type:

docker compose build

This should BUILD the flask_app image, with the name defined in the "image" value. In my case it's francescoxx/flask_live_app:1.0.0 because that's my Dockerhub username. You should replace "francescoxx" with your Dockerhub username.

You can also see all the steps docker did to build the image, layer by layer. You might recognize some of them, because we defined them in the Dockerfile.

Now, to check if the image has been built successfully, type:

docker images

We should see a similar result, with the image we just built:

⚗️ Run the flask_app service

We are almost done, but one last step is to run a container based on the image we just built.

To do that, we can just type:

docker compose up flask_app

In this case we don't use the -d flag, because we want to see the logs in the terminal.

We should see something like this:

🔍 Test the application

Let's test our application. First of all, let's just go to any browser and visit localhost:4000/test

You shoulds see this result:

(note that if you visit localhost:4000 you get an error because there is no route associated to this endpoint, but by getting an error is a good thing, because it means that the server is running!)

Now it's time to test all the endpints using Postman. Feel free to use any tool you want.

If we make a GET request to localhost:4000/users we will get an empty array. This is correct

📝 Create a user

Now let's create a user, making a POST request to localhost:4000/users with the body below as a request body:

Let's create another one:

One more:

📝 Get all users

Now, let's make a GET request to localhost:4000/users to get all the users:

We just created 3 users.

📝 Get a specific user

If you want to get a specific user, you can make a GET request to localhost:4000/users/<user_id>.

For example, to get the user with id 2, you can make a GET request to localhost:4000/users/2

📝 Update a user

If you want to update a user, you can make a PUT request to localhost:4000/users/<user_id>.

For example, to update the user with id 2, you can make a PUT request to localhost:4000/users/2 with the body below as a request body:

To check if the user has been updated, you can make a GET request to localhost:4000/users/2

📝 Delete a user

To delete a user, you can make a DELETE request to localhost:4000/users/<user_id>.

For Example, to delete the user with id 2, you can make a DELETE request to localhost:4000/users/2

To check if the user has been deleted, you can make a GET request to localhost:4000/users

As you can see the user with id 2 is not there anymore.

🏁 Conclusion

We made it! We have built a CRUD rest API in Python, using Flask, SQLAlchemy, Postgres, Docker and Docker compose.

This is just an example, but you can use this as a starting point to build your own application.