Question

I have an array of hashes:

[{ item_id: 1, relationship_1: 1, relationship_2: 1, value: 'go' },
 { item_id: 2, relationship_1: 2, relationship_2: 2, value: 'stop' },
 { item_id: 3, relationship_1: 2, relationship_2: 1, value: 'stop' }, #remove
 { item_id: 4, relationship_1: 3, relationship_2: 1, value: 'go' },
 { item_id: 5, relationship_1: 1, relationship_2: 2, value: 'go' }] #remove

I want the lines commented to be removed. The need is to remove all lines that have relationship_1 and value in common. The only way I can think of is:

items.each do |i|
  items.each do |k|
    if i.item_id != k.item_id and i.relationship_1 == k.relationship_1 and i.value == k.value
      items.remove(k)
    end
  end
end

This is not working as intended. What is the most “Ruby” way to remove those offending items?

Answer 1

Unfortunately, uniq_by isn’t available in Ruby core. Pull it in with require 'activesupport'.

items.uniq_by {|h| [h[:replationship_1], h[:value]] }

Edit: As noted by @mu below, Ruby 1.9’s uniq also works:

items.uniq{|h| [h[:replationship_1], h[:value]] }