grosser-parallel 0.2.0 → 0.3.0

data/README.markdown

@@ -8,22 +8,50 @@ Install
 
 Usage
 =====
+### Processes
+ - Speedup through multiple cpus
+ - Speedup for blocking operations
+ - Protects global data
+ - Extra memory used
 
-    #i -> 0..number_of_your_cpus
+### Threads
+ - Speedup for blocking operations
+ - Global data can be modified
+ - No extra memory used
+
+Map-Reduce-Style
+    # 2 Cpus -> finished after 2 runs (a,b + c)
+    results = Parallel.map(['a','b','c']) do |one_letter|
+      expensive_calculation(letter)
+    end
+
+    # 3 Processes -> finished after 1 run
+    results = Parallel.map(['a','b','c'], :in_processes=>3){|one_letter| ... }
+
+    # 3 Threads -> finished after 1 run
+    results = Parallel.map(['a','b','c'], :in_threads=>3){|one_letter| ... }
+
+
+Normal
+    #i -> 0...number_of_your_cpus
     results = Parallel.in_processes do |i|
       expensive_computation(data[i])
     end
 
-    #i -> 0..4
+    #i -> 0...4
     results = Parallel.in_processes(4) do |i|
       expensive_computation(data[i])
     end
 
-    #same with threads (no speedup through multiple cpus, but speedup for blocking operations)
+    # Threads
     results = Parallel.in_threads(4) do |i|
       blocking_computation(data[i])
     end
 
+TODO
+====
+ - optimize Parallel.map by not waiting for a group to finish: start new when one process finishes
+
 Author
 ======
 [Michael Grosser](http://pragmatig.wordpress.com)  

data/VERSION

@@ -1 +1 @@
-0.2.0
+0.3.0

data/lib/parallel.rb

@@ -49,6 +49,24 @@ class Parallel
     out.map{|x| Marshal.load(x)} #deserialize
   end
 
+  def self.map(array, options={})
+    count = if options[:in_threads]
+      method = 'in_threads'
+      options[:in_threads]
+    else
+      method = 'in_processes'
+      options[:in_processes] || processor_count
+    end
+
+    results = []
+    in_groups_of(array, count).each do |group|
+      results += send(method, group.size) do |i|
+        yield group[i]
+      end
+    end
+    results
+  end
+
   def self.processor_count
     case RUBY_PLATFORM
     when /darwin/
@@ -60,6 +78,19 @@ class Parallel
 
   private
 
+  def self.in_groups_of(array, count)
+    results = []
+    loop do
+      slice = array[(results.size * count)...((results.size+1) * count)]
+      if slice.nil? or slice.empty?
+        break
+      else
+        results << slice
+      end
+    end
+    results
+  end
+
   #handle user interrup (Ctrl+c)
   def self.kill_on_ctrl_c(pids)
     Signal.trap 'SIGINT' do

data/parallel.gemspec

@@ -2,11 +2,11 @@
 
 Gem::Specification.new do |s|
   s.name = %q{parallel}
-  s.version = "0.2.0"
+  s.version = "0.3.0"
 
   s.required_rubygems_version = Gem::Requirement.new(">= 0") if s.respond_to? :required_rubygems_version=
   s.authors = ["Michael Grosser"]
-  s.date = %q{2009-08-13}
+  s.date = %q{2009-08-26}
   s.email = %q{grosser.michael@gmail.com}
   s.extra_rdoc_files = [
     "README.markdown"
@@ -18,6 +18,8 @@ Gem::Specification.new do |s|
      "lib/parallel.rb",
      "parallel.gemspec",
      "spec/cases/parallel_influence_outside_data.rb",
+     "spec/cases/parallel_map.rb",
+     "spec/cases/parallel_map_sleeping.rb",
      "spec/cases/parallel_sleeping_2.rb",
      "spec/cases/parallel_start_and_kill.rb",
      "spec/cases/parallel_with_detected_cpus.rb",
@@ -37,7 +39,9 @@ Gem::Specification.new do |s|
      "spec/cases/parallel_start_and_kill.rb",
      "spec/cases/parallel_with_set_processes.rb",
      "spec/cases/parallel_influence_outside_data.rb",
-     "spec/cases/parallel_with_detected_cpus.rb"
+     "spec/cases/parallel_map_sleeping.rb",
+     "spec/cases/parallel_with_detected_cpus.rb",
+     "spec/cases/parallel_map.rb"
   ]
 
   if s.respond_to? :specification_version then

data/spec/cases/parallel_map.rb

@@ -0,0 +1,6 @@
+require 'spec/spec_helper.rb'
+
+result = Parallel.map(['a','b','c','d']) do |x|
+  "-#{x}-"
+end
+print result * ' '

data/spec/cases/parallel_map_sleeping.rb

@@ -0,0 +1,5 @@
+require 'spec/spec_helper.rb'
+
+Parallel.map(['a','b','c','d']) do |x|
+  sleep 1
+end

data/spec/parallel_spec.rb

@@ -54,4 +54,42 @@ describe Parallel do
       Parallel.in_threads(4){|i| "XXX#{i}"}.should == ["XXX0",'XXX1','XXX2','XXX3']
     end
   end
+
+  describe :map do
+    it "saves time" do
+      t = Time.now
+      `ruby spec/cases/parallel_map_sleeping.rb`
+      Time.now.should be_close(t, 3)
+    end
+
+    it "executes with given parameters" do
+      `ruby spec/cases/parallel_map.rb`.should == "-a- -b- -c- -d-"
+    end
+
+    it "does not flatten results" do
+      Parallel.map([1,2,3], :in_threads=>2){|x| [x,x]}.should == [[1,1],[2,2],[3,3]]
+    end
+
+    it "can run in threads" do
+      Parallel.map([1,2,3,4,5,6,7,8,9], :in_threads=>4){|x| x+2 }.should == [3,4,5,6,7,8,9,10,11]
+    end
+  end
+
+  describe :in_groups_of do
+    it "works for empty" do
+      Parallel.send(:in_groups_of, [], 3).should == []
+    end
+
+    it "works for smaller then count" do
+      Parallel.send(:in_groups_of, [1,2], 3).should == [[1,2]]
+    end
+
+    it "works for count" do
+      Parallel.send(:in_groups_of, [1,2,3], 3).should == [[1,2,3]]
+    end
+
+    it "works for larger than count" do
+      Parallel.send(:in_groups_of, [1,2,3,4], 3).should == [[1,2,3],[4]]
+    end
+  end
 end

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification 
 name: grosser-parallel
 version: !ruby/object:Gem::Version 
-  version: 0.2.0
+  version: 0.3.0
 platform: ruby
 authors: 
 - Michael Grosser
@@ -9,7 +9,7 @@ autorequire:
 bindir: bin
 cert_chain: []
 
-date: 2009-08-13 00:00:00 -07:00
+date: 2009-08-26 00:00:00 -07:00
 default_executable: 
 dependencies: []
 
@@ -28,6 +28,8 @@ files:
 - lib/parallel.rb
 - parallel.gemspec
 - spec/cases/parallel_influence_outside_data.rb
+- spec/cases/parallel_map.rb
+- spec/cases/parallel_map_sleeping.rb
 - spec/cases/parallel_sleeping_2.rb
 - spec/cases/parallel_start_and_kill.rb
 - spec/cases/parallel_with_detected_cpus.rb
@@ -36,7 +38,6 @@ files:
 - spec/spec_helper.rb
 has_rdoc: false
 homepage: http://github.com/grosser/parallel
-licenses: 
 post_install_message: 
 rdoc_options: 
 - --charset=UTF-8
@@ -57,7 +58,7 @@ required_rubygems_version: !ruby/object:Gem::Requirement
 requirements: []
 
 rubyforge_project: 
-rubygems_version: 1.3.5
+rubygems_version: 1.2.0
 signing_key: 
 specification_version: 3
 summary: Run any kind of code in parallel processes
@@ -68,4 +69,6 @@ test_files:
 - spec/cases/parallel_start_and_kill.rb
 - spec/cases/parallel_with_set_processes.rb
 - spec/cases/parallel_influence_outside_data.rb
+- spec/cases/parallel_map_sleeping.rb
 - spec/cases/parallel_with_detected_cpus.rb
+- spec/cases/parallel_map.rb