Sequelize & Node JS ORMs in 2021
Since 2020 took a serious hit on the excitement in my life, I thought I’d start out 2021 on a high-note: taking Sequelize, a Node JS ORM, for a spin 🚗.
But first, the basics:
What is an ORM?
An ORM, short for Object Relational Mapper, is a tool that allows us to (in this case) interact with the database of our program in our preferred programming language.
Think of it as Google Translate between you and your database: it takes in commands in one language and maps them to another - allowing you to create, read, update and delete data without having to “speak” the databases’ language¹.
Why use a Node JS ORM?
If you’ve dealt with projects that required storing lots of organized information, you’ve likely encountered code written in Structured Query Language (SQL). SQL is a domain-specific language that allows you to manage data held in relational database management systems², such as PostgreSQL. You may have even written something like this:
// Select all names and phone numbers of hotels located in Sydney
SELECT name, phone_number
FROM hotels
WHERE city = 'Sydney';While SQL is a powerful language, some find the syntax to be awkward and error-prone. ORMs in their ideal form abstract some of the logic from database operations, resulting in fewer lines of code and a more efficient work-flow³.
Sounds great…is there a catch?
More complex database queries using ORMs can be difficult to achieve, leading to the argument that time may be better spent learning SQL rather than learning a tool to avoid having to learn SQL⁴.
Going for a Ride with Sequelize
While I’ve had some experience working with Active Record in Ruby on Rails, Sequelize is my first foray into using an ORM in Node JS and JavaScript. As one of the most downloaded⁵ Node JS ORMs out there, it felt like a good starting point.
To get acquainted with Sequelize, which is promise-based, I played around with a sample project using Express while perusing the offical Sequelize documentation.
The Set-Up
Setting up the sample project was straight-forward and allowed me to get right into making database inserts and queries.
Defining Associations
Similar to Active Record, Sequelize uses models to define database tables. It is convention to put each data model in its own file (with a pluralized name), which is then put in a folder called “models”.
As a quick overview, in databases, relationships between data are typically visualized using entity-relationship diagrams. Associations between tables (or models) of related data are described in terms such as one-to-one, one-to-many and many-to-many relationships.
In Sequelize, these relationships are defined by combining the following types of associations (which you can learn more about here):
HasOneBelongsToHasManyBelongsToMany
For example, for the one-to-many relationship between Travellers and Luggage, we would define it as:
const { traveller, luggage } = sequelize.models;traveller.hasMany(luggage);luggage.belongsTo(traveller);
Once relationships are defined, foreign keys are generated automatically by Sequelize when an insert is made into the database.
Database Queries
For queries into the database, Sequelize has a wide selection of methods available. A simple example is if we wanted to find all the information about the Luggage for the Traveller with the ID of 2:
// Sequelize
Luggage.findAll({ where: { traveller_id: 2 } }); // SQL equivalent
SELECT * FROM luggage WHERE traveller_id = 2;/* If you wanted the information about the traveller and their luggage: */// Sequelize
Traveller.findByPk(2, { include: models.luggage });// SQL equivalent
SELECT * FROM travellers, luggage
JOIN luggage ON travellers.id = luggage.traveller_id
WHERE travellers.id = 2;
Final Thoughts
Overall, I enjoyed learning a little about Sequelize. I found the syntax of defining relationships to be intuitive and really appreciated the automatic generation of foreign keys as that has been a pain-point for me in previous projects. But, while simple queries were a delight, I found queries involving more than one model with Sequelize’s syntax to be a challenge.
I’m now inspired to look into other Node JS ORM and query-building libraries like Objection.js and Knex for my next project.
Thanks for reading and let me know your thoughts on ORMs!
References
1 Object–relational mapping,Wikipedia https://en.wikipedia.org/wiki/Object%E2%80%93relational_mapping Retrieved Jan 8, 2021
2 SQL, Wikipedia https://en.wikipedia.org/wiki/SQL Retrieved Jan 8, 2021
3 Object-relational Mappers (ORMs), Full Stack Python https://www.fullstackpython.com/object-relational-mappers-orms.html Retrieved Jan 8, 2021
4 Node.js ORMs and why you shouldn’t use them, Thomas Hunter https://blog.logrocket.com/why-you-should-avoid-orms-with-examples-in-node-js-e0baab73fa5/ Retrieved Jan 8, 2021
5 Comparing bookshelf vs. knex vs. objection vs. orm vs. sequelize, NPM Compare https://npmcompare.com/compare/bookshelf,knex,objection,orm,sequelize Retrieved Jan 8, 2021
6 API Reference, Sequelize https://sequelize.org/master/manual/assocs.html Retrieved Jan 8, 2021
