Each table is referred to as a model, each model is simply just a python class and each attribute of that class becomes a column in an SQL table. The database is made up of multiple models. Just like with normal Python models can inherit from other models and share attributes with the parent model. This is very useful if you going to have models that will store similar types of data.

import flask_sqlalchemy

db = flask_sqlalchemy

class Pets(db.Model):
    __abstract__ = True
    name = db.Column(db.String(100))
    price = db.Column(db.Integer)
    breed = db.Column(db.String(100))

class Cats(Pets):
    __tablename__ = 'cats'
    id = db.Column(db.Integer, primary_key=True)

class Dogs(Pets):
    __tablename__ = 'dogs'
    id = db.Column(db.Integer, primary_key=True)

Taking a look at the models.py module, we define an abstract class called Pets. Which means SQLAlchemy will not create a table for that model. Our next two models Cats and Dogs inherit all the attributes form Pets. So Cats and Dog tables will each have a column called name, price and breed. The main advantage of this is if you ever need to change the models you just have to change it in once place. The more models that inherit from the base model.

from flask import Flask

from .models import db

uri = "mysql+pymysql://user:password@localhost:3306/test"
app.config['SQLALCHEMY_DATABASE_URI'] = uri
app.config['SQLALCHEMY_TRACK_MODIFICATIONS'] = False
app.app_context().push()

db.init_app(app)
db.create_all()

Above is an example __init__.py file to initialise the database and create all the database tables from the models. That’s it folks, thanks for reading.

Please note there are other ways to implement inheritance with SQLAlchemy, I personally found this way to be the cleanest in terms of code readability.

Appendix

• Example source code
• SQLAlchemy
• flask-sqlalchemy