has_and_belongs_to_many, avoiding dupes in the join table

Solution 1:

Prevent duplicates in the view only (Lazy solution)

The following does not prevent writing duplicate relationships to the database, it only ensures find methods ignore duplicates.

In Rails 5:

has_and_belongs_to_many :tags, -> { distinct }

Note: Relation#uniq was depreciated in Rails 5 (commit)

In Rails 4

has_and_belongs_to_many :tags, -> { uniq }

Prevent duplicate data from being saved (best solution)

Option 1: Prevent duplicates from the controller:

post.tags << tag unless post.tags.include?(tag)

However, multiple users could attempt post.tags.include?(tag) at the same time, thus this is subject to race conditions. This is discussed here.

For robustness you can also add this to the Post model (post.rb)

def tag=(tag)
  tags << tag unless tags.include?(tag)
end

Option 2: Create a unique index

The most foolproof way of preventing duplicates is to have duplicate constraints at the database layer. This can be achieved by adding a unique index on the table itself.

rails g migration add_index_to_posts
# migration file
add_index :posts_tags, [:post_id, :tag_id], :unique => true
add_index :posts_tags, :tag_id

Once you have the unique index, attempting to add a duplicate record will raise an ActiveRecord::RecordNotUnique error. Handling this is out of the scope of this question. View this SO question.

rescue_from ActiveRecord::RecordNotUnique, :with => :some_method

Solution 2:

In addition the suggestions above:

add :uniq to the has_and_belongs_to_many association
adding unique index on the join table

I would do an explicit check to determine if the relationship already exists. For instance:

post = Post.find(1)
tag = Tag.find(2)
post.tags << tag unless post.tags.include?(tag)

Solution 3:

You can pass the :uniq option as described in the documentation. Also note that the :uniq options doesn't prevent the creation of duplicate relationships, it only ensures accessor/find methods will select them once.

If you want to prevent duplicates in the association table you should create an unique index and handle the exception. Also validates_uniqueness_of doesn't work as expected because you can fall into the case a second request is writing to the database between the time the first request checks for duplicates and writes into the database.

Solution 4:

In Rails4:

class Post < ActiveRecord::Base 
  has_and_belongs_to_many :tags, -> { uniq }

(beware, the -> { uniq } must be directly after the relation name, before other params)

Rails documentation