Benchmarking Ruby's fetch against ActiveSupport's try.

ruby_try_vs_fetch_benchmark.rb

#!/usr/bin/env ruby
# frozen_string_literal: true

require 'benchmark'
require 'active_support/core_ext/object/try'

n = 5000000

puts "Key not found:"
sad_hash = {}
Benchmark.bm(7) do |x|
  x.report("try")   { n.times { sad_hash.try(:[], :test) } }
  x.report("fetch") { n.times { sad_hash.fetch(:test, nil) } }
end

puts "\n---\n\n"

puts "Key found:"
happy_hash = { test: true }
Benchmark.bm(7) do |x|
  x.report("try")   { n.times { happy_hash.try(:[], :test) } }
  x.report("fetch") { n.times { happy_hash.fetch(:test) } }
end

$ ./ruby_try_vs_fetch_benchmark.rb
Key not found:
              user     system      total        real
try       1.083426   0.021701   1.105127 (  1.105285)
fetch     0.195870   0.003290   0.199160 (  0.199291)

---

Key found:
              user     system      total        real
try       1.077929   0.014164   1.092093 (  1.092481)
fetch     0.194351   0.002286   0.196637 (  0.196766)