Question

I think the best way to explain this is with an example:

class A
    attr_accessor :test
    def initialize(x = nil)
        @test = x
    end

    def ==(other)
        return @test == other.test
    end
end

a1 = A.new(1) # => #<A:0x11b7118 @test=1>
a1.test # => 1

a2 = A.new(1) # => #<A:0x11fb0f8 @test=1>
a2.test # => 1

a1 == a2 # => true
[a1].include?(a2) # => true
[a1] - [a2] # => [#<A:0x11b7118 @test=1>]

In this example, how would I get [a1] – [a2] to return an empty array, as one would expect it to for any other Ruby class? Is there some method that I have to define for A that I’m missing?

Answer 1

You need to override eql? and hash. These are the ones that are used for those kinds of set operations.