-
Notifications
You must be signed in to change notification settings - Fork 22
Commit
This commit does not belong to any branch on this repository, and may belong to a fork outside of the repository.
Added inline configuration example, benchmarking a few FizzBuzz solut…
…ions for fun and profit.
- Loading branch information
Showing
1 changed file
with
223 additions
and
0 deletions.
There are no files selected for viewing
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Original file line number | Diff line number | Diff line change |
---|---|---|
@@ -0,0 +1,223 @@ | ||
#!/usr/bin/env jruby | ||
|
||
# This example demonstrates using Doubleshot inline, | ||
# declaring and resolving dependencies without having | ||
# to create a separate Doubleshot file. Useful for | ||
# little one-off scripts like this. For example: | ||
# As long as you have doubleshot installed, any | ||
# single-file Gist would be self-executable, defining | ||
# it's dependencies inline. | ||
require "doubleshot" | ||
|
||
Doubleshot.new do |config| | ||
config.gem "perfer", "= 0.2.0" | ||
end.setup! | ||
|
||
require "perfer" | ||
require "stringio" | ||
require "singleton" | ||
|
||
class Solutions | ||
include Singleton | ||
|
||
def initialize | ||
@solutions = [] | ||
@names = {} | ||
end | ||
|
||
def add(name, &b) | ||
@solutions << b | ||
@names[b] = name | ||
end | ||
|
||
def benchmark! | ||
raise StandardError.new("Solutions do not match!") unless valid? | ||
|
||
Perfer::session "FizzBuzz" do |x| | ||
@solutions.each do |solution| | ||
x.iterate(@names[solution]) do | ||
solution.call StringIO.new | ||
end | ||
end | ||
end.run | ||
end | ||
|
||
private | ||
def valid? | ||
raise NoSolutionsDefinedError.new if @solutions.empty? | ||
return true if @solutions.size == 1 | ||
|
||
reference = render(@solutions[0]).string | ||
@solutions[1..-1].all? { |solution| render(solution).string == reference } | ||
end | ||
|
||
def render(solution) | ||
buffer = StringIO.new | ||
solution.call(buffer) | ||
buffer | ||
end | ||
|
||
class NoSolutionsDefinedError < StandardError | ||
end | ||
end | ||
|
||
def solution(name, &b) | ||
Solutions.instance.add(name, &b) | ||
end | ||
|
||
solution "Reference" do |out| | ||
def self.multiple_of_3(number) | ||
number % 3 == 0 | ||
end | ||
|
||
def self.multiple_of_5(number) | ||
number % 5 == 0 | ||
end | ||
1.upto(100) do |i| | ||
output = "" | ||
|
||
if !multiple_of_3(i) && !multiple_of_5(i) | ||
output = i | ||
else | ||
output << "Fizz" if multiple_of_3(i) | ||
output << "Buzz" if multiple_of_5(i) | ||
end | ||
|
||
out.puts output | ||
end | ||
end | ||
|
||
# In SolutionTwo we inline the modulo operations, | ||
# and re-order the conditionals to avoid evaluating | ||
# the same expression more than once. | ||
solution "Evaluate Once" do |out| | ||
1.upto(100) do |i| | ||
if i % 3 == 0 | ||
if i % 5 == 0 | ||
out.puts "FizzBuzz" | ||
else | ||
out.puts "Fizz" | ||
end | ||
else | ||
if i % 5 == 0 | ||
out.puts "Buzz" | ||
else | ||
out.puts i | ||
end | ||
end | ||
end | ||
end | ||
|
||
# In SolutionThree we optimize SolutionTwo further | ||
# by avoiding the modulo operations, and switching | ||
# to a while loop to avoid the dispatch overhead of | ||
# the block iterator. | ||
# | ||
# So far our solutions are all (mostly) fundamentally | ||
# the same, each step simply inlining operations. | ||
# The only real algorithmic advancement is to re-order | ||
# the conditionals to avoid needing to evaluate the | ||
# "3" or "5" case more than once when only one of those | ||
# is true. | ||
# | ||
# NOTE: This solution is the best performing, by a | ||
# pretty good margin, around two-to-one over the short, | ||
# idiomatic solutions. Just goes to show, | ||
# method-dispatch (yes, including block-yielding) is | ||
# the enemy of fast Ruby code. | ||
solution "While Loop" do |out| | ||
i = 1 | ||
threes = 1 | ||
fives = 1 | ||
|
||
while i < 101 | ||
if threes == 3 | ||
threes = 0 | ||
if fives == 5 | ||
fives = 0 | ||
out.puts "FizzBuzz" | ||
else | ||
out.puts "Fizz" | ||
end | ||
else | ||
if fives == 5 | ||
fives = 0 | ||
out.puts "Buzz" | ||
else | ||
out.puts i | ||
end | ||
end | ||
|
||
threes += 1 | ||
fives += 1 | ||
i += 1 | ||
end | ||
end | ||
|
||
# This solution by Daniel Martin is cuourtesy of Ruby Quiz: | ||
# http://www.rubyquiz.com/quiz126.html | ||
solution "Martinization" do |out| | ||
(1..100).each do |i| | ||
x = "" | ||
x += "Fizz" if i % 3 == 0 | ||
x += "Buzz" if i % 5 == 0 | ||
out.puts( x.empty? ? i : x ) | ||
end | ||
end | ||
|
||
# This is the same as SolutionFour, but with an experiment | ||
# to optimize out the temporary string: | ||
solution "No Temporary String" do |out| | ||
alt = nil | ||
|
||
(1..100).each do |i| | ||
if i % 3 == 0 | ||
alt = true | ||
out.print "Fizz" | ||
end | ||
|
||
if i % 5 == 0 | ||
alt = true | ||
out.print "Buzz" | ||
end | ||
|
||
if alt | ||
alt = nil | ||
out << $/ | ||
else | ||
out.puts i | ||
end | ||
end | ||
end | ||
|
||
# Another alternative to solution four, seeing if | ||
# precalculating the range to an Array improves | ||
# performance at all (since it's a small range), | ||
# there's no real reason not to, but depending | ||
# on the implementation of Range#to_a, we may | ||
# not avoid calling Fixnum#succ anyway, so it | ||
# might be a useless exercise. | ||
solution "Precalculated Array" do |out| | ||
(1..100).to_a.each do |i| | ||
x = "" | ||
x += "Fizz" if i % 3 == 0 | ||
x += "Buzz" if i % 5 == 0 | ||
out.puts( x.empty? ? i : x ) | ||
end | ||
end | ||
|
||
Solutions.instance.benchmark! | ||
|
||
__END__ | ||
|
||
######################################################### | ||
Output of solutions match? true | ||
Session FizzBuzz with jruby 1.7.0 (1.9.3p203) 2012-10-22 ff1ebbe on Java HotSpot(TM) 64-Bit Server VM 1.7.0_04-b21 [darwin-x86_64] | ||
|
||
Taking 10 measurements of at least 1.0s | ||
Reference 87.38 µs/i ± 2.371 ( 2.7%) <=> 11443 ips | ||
Evaluate Once 49.78 µs/i ± 0.274 ( 0.6%) <=> 20085 ips | ||
While Loop 40.81 µs/i ± 0.833 ( 2.0%) <=> 24500 ips | ||
Martinization 58.31 µs/i ± 0.281 ( 0.5%) <=> 17148 ips | ||
No Temporary String 47.81 µs/i ± 0.220 ( 0.5%) <=> 20916 ips | ||
Precalculated Array 59.24 µs/i ± 1.466 ( 2.5%) <=> 16879 ips |