Using pluck can save a bunch of memory

If you only need to use a single field on an Active Record model, for example, the ids, then Rails has your back.

Instead of…

…iterating over objects fully loaded into memory.

Book.paperbacks.map { |book| book.title }
#=> ["Eloquent Ruby", "Sapiens", "Agile Web Development With Rails"]
Book.paperbacks.map(&:title)
#=> ["Eloquent Ruby", "Sapiens", "Agile Web Development With Rails"]

Use…

…the ActiveRelation#pluck method to pull the required fields directly from the database.

Book.paperbacks.pluck(:title)
#=> ["Eloquent Ruby", "Sapiens", "Agile Web Development With Rails"]

But why?

This is about speed and efficient memory usage.

The problem of Active Record models taking up a lot of memory becomes particularly painful if you’re operating over large collections. In the first example, each row from the database is transformed into an object and yet we’re only using a single field. All the extra functionality that a model has—and Rails prepares for you in memory—is unneeded.

Only loading the fields that you need means your application will run faster and use less memory.

Why not?

The #pluck method returns only the values you request in a plain Ruby array—no methods are loaded from your model.

If you do need the actual Active Record object or need to update the model at some point after using #pluck, you’re out of luck.

You might also consider using a select clause in your scope, which would provide you with a full Active Record model rather than a plain array, but only with the fields you request. This is still nowhere near as memory efficient as using #pluck.

Book.paperbacks.select(:title).map { |book| book.title }
#=> ["Eloquent Ruby", "Sapiens", "Agile Web Development With Rails"]