Sha256: e4e94bc6ffdd6a39733d4a99982d75bcb453003987c9aa78aa433616972893ac
Contents?: true
Size: 1.38 KB
Versions: 13
Compression:
Stored size: 1.38 KB
Contents
# frozen_string_literal: true
require 'benchmark/ips'
# This benchmark compares accessing an instance variable vs accessing a struct member (via a function). The actual method dispatch is about 25% slower.
puts "Ruby #{RUBY_VERSION} at #{Time.now}"
ItemsStruct = Struct.new(:items) do
def initialize
super []
end
def push_me_pull_you(value = :x)
items = self.items
items << value
items.pop
end
def empty?
self.items.empty?
end
end
class ItemsClass
def initialize
@items = []
end
def push_me_pull_you(value = :x)
items = @items
items << value
items.pop
end
def empty?
@items.empty?
end
end
# There IS a measuarble difference:
Benchmark.ips do |x|
x.report("Struct#empty?") do |times|
i = 0
instance = ItemsStruct.new
while i < times
break unless instance.empty?
i += 1
end
end
x.report("Class#empty?") do |times|
i = 0
instance = ItemsClass.new
while i < times
break unless instance.empty?
i += 1
end
end
x.compare!
end
# This shows that in the presence of additional work, the difference is neglegible.
Benchmark.ips do |x|
x.report("Struct#push_me_pull_you") do |times|
i = 0
a = A.new
while i < times
a.push_me_pull_you(i)
i += 1
end
end
x.report("Class#push_me_pull_you") do |times|
i = 0
b = B.new
while i < times
b.push_me_pull_you(i)
i += 1
end
end
x.compare!
end
Version data entries
13 entries across 13 versions & 1 rubygems