Defining has_many :through Requires Explicit Association with the Intermediate Table

Target audience: Rails beginners

When using has_many :through, an error occurs if you forget to define a direct association with the intermediate table you are going through.
This is because the information required to construct the SQL that Rails issues is missing.

---

1. Common pattern of missing definitions

Example of problematic model definition:

# app/models/user.rb
class User < ApplicationRecord
  # NG: No direct association defined with the intermediate table (registrations)
  has_many :events, through: :registrations
end

# app/models/registration.rb
class Registration < ApplicationRecord
  belongs_to :user
  belongs_to :event
end

# app/models/event.rb
class Event < ApplicationRecord
  has_many :registrations
  has_many :participants, through: :registrations, source: :user
end

In the User model above, has_many :events, through: :registrations is written, but there is no direct association definition has_many :registrations between the User model and the intermediate registrations table prior to it.

---

2. Impact on SQL generation

If you try to retrieve data like User.first.events, Rails will attempt to generate SQL that JOINS the users table, registrations table, and events table.

Specifically, it needs to first JOIN the users table and registrations table, and then JOIN that result with the events table.

However, if the User model does not have the has_many :registrations definition, Rails cannot resolve which foreign key (e.g., registrations.user_id) to use to JOIN the users table and registrations table. As a result, it cannot assemble the appropriate SQL, leading to an error.

---

3. Correct Model Definition

You need to explicitly define the association with the intermediate table.

# app/models/user.rb
class User < ApplicationRecord
  has_many :registrations # Define association with the intermediate table
  has_many :events, through: :registrations
end

With this has_many :registrations definition, Rails can understand that it should JOIN the id column of the users table with the user_id column of the registrations table.

---

4. SQL Issued with Correct Definition (Image)

When User.find(1).events is executed, the following SQL is issued:

SELECT "events".*
FROM "events"
INNER JOIN "registrations" ON "events"."id" = "registrations"."event_id"
WHERE "registrations"."user_id" = 1;

For this SQL to function correctly, it is a prerequisite that Rails recognizes the association between the users table and the registrations table (in this case, registrations.user_id = users.id). That is why has_many :registrations is necessary.

---

Conclusion

When using has_many :through, always check the following points:

• The has_many definition (e.g., has_many :registrations) for the association name specified by through: (e.g., :registrations) must exist in the originating model.