Ruby Hash Iteration

So a couple weeks ago I realized a cool way you can iterate over hashes in ruby. This is probably nothing new for a lot of people, but awesome nonetheless. For some reason I needed the index, so I’ll show that in my example here.

Basic iteration

2.2.3 :005 > h
 => {:a=>1, :b=>2, :c=>3, :d=>4}
2.2.3 :006 > h.each_with_index do |p, i|
2.2.3 :007 >     puts "Object is: #{p}, with index of: #{i}"
2.2.3 :008?>   end
Object is: [:a, 1], with index of: 0
Object is: [:b, 2], with index of: 1
Object is: [:c, 3], with index of: 2
Object is: [:d, 4], with index of: 3
 => {:a=>1, :b=>2, :c=>3, :d=>4}

Okay, so what’s the problem there. Actually, nothing really if that’s the output you want.

I was hoping to be able to iterate over the hash and fetch keys and values direcly, without doing something like p[0] or p[1] to obtain the key and value. (referencing example above)


2.2.3 :009 > h.each_with_index do |(k, v), i|
2.2.3 :010 >     puts "Key is: #{k}, with value of #{v}, with index of: #{i}"
2.2.3 :011?>   end
Key is: a, with value of 1, with index of: 0
Key is: b, with value of 2, with index of: 1
Key is: c, with value of 3, with index of: 2
Key is: d, with value of 4, with index of: 3
 => {:a=>1, :b=>2, :c=>3, :d=>4}

Notice the parentheses passed in the block. Obiously, the output difference is clear.


I was curious as to if there were any performance difference between the two. So I wrote a quick benchmark script.

require 'benchmark'

n = 5_000_000
TestHash = {a: 1, b: 2, c: 3, d: 4}

def parentheses(n)
  n.times do
    TestHash.each_with_index { |(k, v), i| b = "1" }

def no_parentheses(n)
  n.times do
    TestHash.each_with_index { |k, v| b = "1" }
end do |x|"no_parentheses:") { no_parentheses(n) }"   parentheses:") { parentheses(n) }


Turns out, virtually no difference in performance. Technically, with parentheses was slightly slower (very slightly). This example is 5,000,000 iterations.

Just thought this was cool, hopefully it helps somone else out.

comments powered by Disqus