Question

How can I refactor the following code to avoid repetition?

  r = @lots.fetch @lots.keys.sample      
  until (neighbours r)
    r = @lots.fetch @lots.keys.sample
  end

I basically have a new r object that is picked randomly and I need to pick r till the selected one doesn’t respond to certain criteria (neighbours r). How can I refactor it to avoid the repetition of getting r and keep getting it till a condition is reached? Thanks

Answer 1

Personally I would shuffle @lots (turn it into a two element array if needed) and then use take_while like this:

>> h = {:a => 1, :b => 2, :c => 3} #=> {:a=>1, :b=>2, :c=>3}
>> h.to_a.shuffle.take_while { |k, v| v < 3 } #=> []
>> h.to_a.shuffle.take_while { |k, v| v < 3 } #=> [[:b, 2]]
>> h.to_a.shuffle.take_while { |k, v| v < 3 } #=> [[:a, 1]]
>> h.to_a.shuffle.take_while { |k, v| v < 3 } #=> [[:b, 2], [:a, 1]]

If you need the result as a hash again, just call Hash[] on it:

 >> Hash[h.to_a.shuffle.take_while { |k, v| v < 3 }] #=> {:a=>1}

In your case the condition for take while would be something like { |k, v| !(neighbours v)}. Also if you want the elements from the starting hash to be repeatable, you’ll have to do use something else instead of shuffle.