Performance penalty of Ruby's block shortcut

[sudo-nice]

Hello!

In Proc & Block section you've mentioned:

There is a small performance penalty in Ruby for using &:shortcut

But does that really happen to you? I don't remember the details, but I believe the shortcut's performance has been optimized already. Here are some results of mine:

require "benchmark/ips"

Item = Struct.new(:value)
ITEMS = %w[foo bar baz].map(&Item.method(:new))

def method_1
  ITEMS.map(&:value)
end

def method_2
  ITEMS.map { |i| i.value }
end

Benchmark.ips do |bm|
  bm.report("method_1") { method_1 }
  bm.report("method_2") { method_2 }
  bm.compare!
end

# >> Warming up --------------------------------------
# >>             method_1   210.596k i/100ms
# >>             method_2   213.856k i/100ms
# >> Calculating -------------------------------------
# >>             method_1      3.472M (± 0.1%) i/s -     17.479M in   5.034397s
# >>             method_2      3.434M (± 0.1%) i/s -     17.322M in   5.044515s
# >>
# >> Comparison:
# >>             method_1:  3472014.9 i/s
# >>             method_2:  3433897.9 i/s - 1.01x  slower
# >>

---

ruby 2.5.3p105 (2018-10-18 revision 65156) [x86_64-linux-gnu]

[konung]

Yep, I think you right. I was just going directly off of the fast-ruby repo's Readme.md. In my case, the difference is within 2 %. so not worth mentioning anymore, probably.

[sudo-nice]

Thanks for the answer, good to know that.

[konung]

I;ll keep this open, so I don't forget to update the wording.