How do we avoid creating new columns?

I’m working on integrating with Stripe to allow our customers to process payments and manage subscriptions. Our customers have to input their card details every month at our checkout.

What problems did I encounter?

When creating a Subscription, Stripe requires certain information, such as customer, price items, and recurrence interval. However, to avoid making requests to create items and customers, I decided to save the Stripe IDs in our database. I didn’t want to save this data in our customer or plan tables because not all customers or plans would have this data.

How did I solve this problem?

To avoid creating columns with names like customer_stripe_id or price_item_id in our tables, I created a table called entity_field_values with four columns (entity_name, entity_field_id, field_name, and field_value), and the columns (entity_name, entity_field_id, and field_name) as a unique key.

How do I use this new table?

The EntityFieldValue class has only two methods: save_field and find_by_entity_and_field. In save_field, the entity, field_name, and field_value are passed. If the data is already saved in the database, we update it; otherwise, we create a new entry. The find_by_entity_and_field method takes two parameters: entity and field_name. If the data is found, it returns the object. Otherwise, it returns null (nil since I’m using Ruby).

Results obtained from this change?

Now, we avoid creating columns in our tables for specific things and can easily add new information without having to create migrations constantly. Our tables perform better because they now have less data stored, and when we select, we only use the data necessary for that functionality.