fast_group_by 0.1.0

data/Rakefile

@@ -0,0 +1,49 @@
+require "rake"
+require "rake/clean"
+require "rake/gempackagetask"
+require 'rubygems'
+
+################################################################################
+### Gem
+################################################################################
+
+begin
+  # Parse gemspec using the github safety level.
+  file = Dir['*.gemspec'].first
+  data = File.read(file)
+  spec = nil
+  # FIXME: Lowered SAFE from 3 to 2 to work with Ruby 1.9 due to rubygems
+  # performing a require internally
+  Thread.new { spec = eval("$SAFE = 2\n%s" % data)}.join
+
+  # Create the gem tasks
+  Rake::GemPackageTask.new(spec) do |package|
+    package.gem_spec = spec
+  end
+rescue Exception => e
+  printf "WARNING: Error caught (%s): %s\n%s", e.class.name, e.message, e.backtrace[0...5].map {|l| '  %s' % l}.join("\n")
+end
+
+desc 'Package and install the gem for the current version'
+task :install => :gem do
+  system "sudo gem install -l pkg/%s-%s.gem" % [spec.name, spec.version]
+end
+
+desc 'Show files missing from gemspec'
+task :diff do
+  files = %w[
+    Rakefile
+    *README*
+    *LICENSE*
+    *.gemspec
+    bin/*
+    lib/**/*
+    spec/**/*
+  ].map {|pattern| Dir.glob(pattern)}.flatten.select{|f| File.file?(f)}
+  missing_files = files - spec.files
+  extra_files = spec.files - files
+  puts "Missing files:"
+  puts missing_files.join(" ")
+  puts "Extra files:"
+  puts extra_files.join(" ")
+end

data/fast_group_by.gemspec

@@ -0,0 +1,33 @@
+Gem::Specification.new do |s|
+  # Project
+  s.name         = 'fast_group_by'
+  s.summary      = "Fast group_by is an Enumerable#group_by implementation that uses hash instead of OrderedHash and is thus faster, but not ordered."
+  s.description  = s.summary
+  s.version      = '0.1.0'
+  s.date         = '2009-07-16'
+  s.platform     = Gem::Platform::RUBY
+  s.authors      = ["Wes Oldenbeuving"]
+  s.email        = "narnach@gmail.com"
+  s.homepage     = "http://www.github.com/Narnach/fast_group_by"
+
+  # Files
+  root_files     = %w[readme.rdoc Rakefile fast_group_by.gemspec]
+  bin_files      = []
+  lib_files      = %w[fast_group_by]
+  test_files     = %w[]
+  spec_files     = %w[fast_group_by]
+  other_files    = %w[]
+  s.bindir       = "bin"
+  s.require_path = "lib"
+  s.executables  = bin_files
+  s.test_files   = test_files.map {|f| 'test/%s_test.rb' % f} + spec_files.map {|f| 'spec/%s_spec.rb' % f}
+  s.files        = root_files + s.test_files + other_files + bin_files.map {|f| 'bin/%s' % f} + lib_files.map {|f| 'lib/%s.rb' % f}
+
+  # rdoc
+  s.has_rdoc         = true
+  s.extra_rdoc_files = %w[readme.rdoc]
+  s.rdoc_options << '--inline-source' << '--line-numbers' << '--main' << 'readme.rdoc'
+
+  # Requirements
+  s.required_ruby_version = ">= 1.8.0"
+end

data/lib/fast_group_by.rb

@@ -0,0 +1,13 @@
+module Enumerable
+  # Return a Hash with key-Array pairs. The keys are return values of the
+  # block that is called with each element in the collection.
+  # All elements with the same return key are added to the Array associated
+  # with that key.
+  def fast_group_by(&block)
+    res = Hash.new { |hash, key| hash[key] = [] }
+    each do |e|
+      res[block.call(e)] << e
+    end
+    res
+  end
+end

data/readme.rdoc

@@ -0,0 +1,45 @@
+= Fast group_by
+
+An unordered group_by implementation that is a lot faster than
+the ordered version used in ActiveSupport version 2.2.x. This becomes apparent when used with large data sets.
+
+== How to install
+
+Once the gem appears on gems.github.com, you can install it with:
+  sudo gem install Narnach-fast_group_by
+
+The alternative is to clone the git repo and build the gem yourself:
+  git clone git://github.com/Narnach/fast_group_by.git
+  cd fast_group_by
+  rake install
+
+== Rationale
+
+ActiveSupport uses OrderedHash in its Enumerable#group_by. In version 2.2.x, it uses nested Arrays to store data. Array#assoc is used to find if keys are already defined.
+For small Arrays, Array#assoc's linear search is about as fast as Hash#[]' tree search, but for larger data sets, the difference in lookup speed becomes noticeable.
+In version 2.3.x, there is a new implementation of OrderedHash that uses Hash instead of Array and is thus a lot faster. Using fast_group_by with ActiveSupport 2.3.x does not make much sense.
+
+The reason why I wrote this is that the OrderedHash version took 66 seconds to group a collection of 50k items. The Hash version took less than 1 second.
+
+== License
+
+Copyright (c) 2009 Wes Oldenbeuving
+
+Permission is hereby granted, free of charge, to any person obtaining
+a copy of this software and associated documentation files (the
+"Software"), to deal in the Software without restriction, including
+without limitation the rights to use, copy, modify, merge, publish,
+distribute, sublicense, and/or sell copies of the Software, and to
+permit persons to whom the Software is furnished to do so, subject to
+the following conditions:
+
+The above copyright notice and this permission notice shall be
+included in all copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
+EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
+NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE
+LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION
+OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION
+WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.

data/spec/fast_group_by_spec.rb

@@ -0,0 +1,12 @@
+require File.join(File.dirname(__FILE__),%w[.. lib fast_group_by])
+
+describe Array, '#fast_group_by' do
+  it 'should return a Hash with key-Array pairs' do
+    ary = (1..10).to_a
+    groups = ary.fast_group_by {|e| e%2}
+    groups.should == {
+      0 => [2, 4, 6, 8, 10],
+      1 => [1, 3, 5, 7, 9]
+    }
+  end
+end

metadata

@@ -0,0 +1,62 @@
+--- !ruby/object:Gem::Specification 
+name: fast_group_by
+version: !ruby/object:Gem::Version 
+  version: 0.1.0
+platform: ruby
+authors: 
+- Wes Oldenbeuving
+autorequire: 
+bindir: bin
+cert_chain: []
+
+date: 2009-07-16 00:00:00 +02:00
+default_executable: 
+dependencies: []
+
+description: Fast group_by is an Enumerable#group_by implementation that uses hash instead of OrderedHash and is thus faster, but not ordered.
+email: narnach@gmail.com
+executables: []
+
+extensions: []
+
+extra_rdoc_files: 
+- readme.rdoc
+files: 
+- readme.rdoc
+- Rakefile
+- fast_group_by.gemspec
+- spec/fast_group_by_spec.rb
+- lib/fast_group_by.rb
+has_rdoc: true
+homepage: http://www.github.com/Narnach/fast_group_by
+licenses: []
+
+post_install_message: 
+rdoc_options: 
+- --inline-source
+- --line-numbers
+- --main
+- readme.rdoc
+require_paths: 
+- lib
+required_ruby_version: !ruby/object:Gem::Requirement 
+  requirements: 
+  - - ">="
+    - !ruby/object:Gem::Version 
+      version: 1.8.0
+  version: 
+required_rubygems_version: !ruby/object:Gem::Requirement 
+  requirements: 
+  - - ">="
+    - !ruby/object:Gem::Version 
+      version: "0"
+  version: 
+requirements: []
+
+rubyforge_project: 
+rubygems_version: 1.3.4
+signing_key: 
+specification_version: 3
+summary: Fast group_by is an Enumerable#group_by implementation that uses hash instead of OrderedHash and is thus faster, but not ordered.
+test_files: 
+- spec/fast_group_by_spec.rb