Building RESTful APIs using Flask

By now, you should already know what a REST API is, if you don't, I recommend you read about it here.

As a very brief summary, building an API in a RESTful way means that you have to build your URL endpoints grouped by "resources". A resource is something you want to manage, e.g: Student, User, Car, etc. A resource is something similar to a database table, but we call them "resources" because of a few exceptions.

Here is an example of RESTful API endpoints to manage Students:

Method	URL	Description
GET	/student	Should return all the students
GET	/student/1	Should return a single student with the id=1
GET	/cohort/1/students	Should return all the students from the cohort with id=1
POST	/student	Should create a new student
PUT	/student/1	Should update the information of the student with the id=1
DELETE	/student/1	Should delete the student with id=1

Take a look at the URLs; they follow a pattern. After a while, the endpoints will speak for themselves, becoming self-explanatory. You will be able to guess what they do or even guess some endpoints. That is the whole idea.

👉 You can read more about REST APIs on this 4Geeks Lesson.
Here is a cool 8 min video explaining REST: https://www.youtube.com/watch?v=7YcW25PHnAA

Now let's talk about Flask

Flask is amazing! It is very similar to Node.js Express Server, and that makes it even cooler because you will be able to work with both technologies without much of a learning curve.

Flask is a library for creating web servers and APIs, basically when you run a Python script that contains the following lines, the computer will start listening for HTTP requests:

1from flask import Flask
2app = Flask(__name__)
3
4@app.route("/")
5def hello():
6    return "Hello World!"
7
8app.run(host='0.0.0.0')

Click to test this code live

Flask Hello-World explained

1from flask import Flask  # Here we import the Flask library into our file
2app = Flask(__name__)  # Here we create a new instance of the Flask server
3
4@app.route("/")  # Here we define the first API path: GET /
5def hello():  # This method will be called when the request is called from any client
6    return "Hello World!"  # Flask will return "Hello World!", this could be an HTML string or a JSON string
7
8app.run(host='0.0.0.0')  # Lastly we start the server on localhost

In Flask we can add new endpoints by using the @app.route decorator, don't worry if this is the first time you see a decorator, the concept is very simple and here is a 5 min video explaining it.

Adding new endpoints

If you want to add another endpoint to your API that runs when a client calls GET /person you will have to add another block of code like this:

1@app.route("/person")  # Here we specify the path for the endpoint
2def handle_person():  # Here we declare a function that will be called when a request is made to that URL
3    return "Hello Person!"  # Here we specify the string we want to respond to the client

Specifying the method: GET, PUT, POST, DELETE

If you want your endpoint to answer to POST, PUT or DELETE you can specify that on the decorator like this:

1from flask import Flask, request
2
3@app.route("/person", methods=['POST', 'GET'])  # Here we specify that this endpoint accepts POST and GET requests
4def handle_person():
5  if request.method == 'POST':  # We can understand what type of request we are handling using a conditional
6    return "A POST has been received!"
7  else:
8    return "A GET has been received!"

Responding a JSON body

The response can basically be whatever you want as long as it is a string: HTML, JSON, CSS, Images, etc. Just make sure you convert into a string whatever you want to respond.

In the following example we are using the jsonify method to convert a dictionary called person1 into a JSON string before returning it to the client.

1from flask import Flask, jsonify
2
3@app.route("/person")
4def handle_person():
5    person1 = {
6      "name": "Bob"
7    }
8
9    return jsonify(person1)

The Response Code

The response code is 200 by default and 500 if there is an unknown error. If you want to respond to the client with a different code, you will have to specify it like this:

1from flask import Flask, jsonify
2
3@app.route("/person")
4def handle_person():
5    content = {
6      "details": "Hey, there has been an error on your request"
7    }
8    resp = jsonify(content)
9    resp.status_code = 400  # Here we change the status code to 400 (typical code for request errors)
10
11    return resp

Another way of changing the response code using a comma ,:

1@app.route("/person")
2def handle_person():
3    content = {
4      "details": "Hey, there has been an error on your request"
5    }
6
7    return jsonify(content), 400

Handling errors and validations

But what if the request comes with errors? For example: If we have an endpoint to create a person and must specify both the first_name AND last_name but only the first_name was found on the request, this is how we would validate it:

1@app.route('/person', methods=['POST'])
2def create_person():
3    # POST request
4    body = request.get_json()  # Get the request body content
5    if body is None:
6        return "The request body is null", 400
7    if 'first_name' not in body:
8        return 'You need to specify the first_name', 400
9    if 'last_name' not in body:
10        return 'You need to specify the last_name', 400
11
12    return "ok", 200

Defining a model

There are different ways to integrate Flask to a database server, but this time we will be explaining the integration with SQLAlchemy.

There is a great Python library that integrates Flask + SQLAlchemy in a seamless way: Flask-SQLAlchemy. We suggest you read this lesson about SQLAlchemy first and come back here.

To integrate with SQLAlchemy all you have to do is install the package and import it into your files like this:

1from flask_sqlalchemy import SQLAlchemy
2db = SQLAlchemy()

Once it is imported, you can start declaring your database models like this:

1class Person(db.Model):
2    id = db.Column(db.Integer, primary_key=True)
3    username = db.Column(db.String(80), unique=True, nullable=False)
4    email = db.Column(db.String(120), unique=True, nullable=False)
5
6    def __repr__(self):
7        return '<Person %r>' % self.username
8
9    def serialize(self):
10        return {
11            "username": self.username,
12            "email": self.email
13        }

You can add as many models as you like.

Ready to start coding?

We have prepared a boilerplate that you can run yourself in GitHub Codespaces or Gitpod and use it as a base for your project.

Flask Rest Hello: https://github.com/4GeeksAcademy/flask-rest-hello