bloom_filter 0.6.1 → 0.6.2

data/README.md

@@ -48,6 +48,7 @@ Remote service:
 * CLI errors
 * CLI help command
 * Improve load/dump workflow
+  * periodic dumps
   
   
   

data/VERSION

@@ -1 +1 @@
-0.6.1
+0.6.2

data/bloom_filter.gemspec

@@ -5,11 +5,11 @@
 
 Gem::Specification.new do |s|
   s.name = %q{bloom_filter}
-  s.version = "0.6.1"
+  s.version = "0.6.2"
 
   s.required_rubygems_version = Gem::Requirement.new(">= 0") if s.respond_to? :required_rubygems_version=
   s.authors = ["Arya Asemanfar"]
-  s.date = %q{2010-01-26}
+  s.date = %q{2010-01-27}
   s.default_executable = %q{bloom_filter_server}
   s.description = %q{}
   s.email = %q{misterfunnyarsal@gmail.com}
@@ -28,9 +28,11 @@ Gem::Specification.new do |s|
      "bin/bloom_filter_server",
      "bloom_filter.gemspec",
      "lib/bloom_filter.rb",
+     "lib/bloom_filter/bit_vector.rb",
      "lib/bloom_filter/client.rb",
      "lib/bloom_filter/protocol.rb",
      "lib/bloom_filter/server.rb",
+     "test/bloom_filter_bit_vector_test.rb",
      "test/bloom_filter_client_test.rb",
      "test/bloom_filter_server_test.rb",
      "test/bloom_filter_test.rb",
@@ -42,7 +44,8 @@ Gem::Specification.new do |s|
   s.rubygems_version = %q{1.3.5}
   s.summary = %q{A simple BloomFilter implementation, usable in-process or as an EventMachine daemon.}
   s.test_files = [
-    "test/bloom_filter_client_test.rb",
+    "test/bloom_filter_bit_vector_test.rb",
+     "test/bloom_filter_client_test.rb",
      "test/bloom_filter_server_test.rb",
      "test/bloom_filter_test.rb",
      "test/test_helper.rb"

data/lib/bloom_filter/bit_vector.rb

@@ -0,0 +1,48 @@
+class BloomFilter
+  class BitVector
+    attr_reader :size
+    
+    def initialize(bits, from_str = nil)
+      @bytes = Array.new((bits.to_f / 8).ceil, 0)
+      @size = bits
+      
+      if from_str
+        from_str.size.times do |i|
+          @bytes[i] = from_str[i].ord
+        end
+      end
+    end
+    
+    def [](index)
+      (@bytes[index.to_i / 8] >> (index % 8)) & 0b1
+    end
+    
+    def []=(index, value)
+      new_value = value && value != 0 ? 1 : 0
+      current_value = self[index]
+      if current_value != new_value
+        if new_value == 0
+          @bytes[index.to_i / 8] -= 1 << (index % 8)
+        else
+          @bytes[index.to_i / 8] += 1 << (index % 8)
+        end
+      end
+    end
+    
+    def to_s
+      str = ""
+      @bytes.each { |byte| str << byte.chr }
+      str
+    end
+    
+    def eql?(o)
+      return false unless o.size == self.size
+      self.size.times do |i|
+        return false unless self[i] == o[i]
+      end
+      true
+    end
+    alias_method :==, :eql?
+    
+  end
+end

data/lib/bloom_filter/client.rb

@@ -1,12 +1,13 @@
 require 'socket'
 require 'bloom_filter/protocol'
+require 'bloom_filter/bit_vector'
 
 class BloomFilter
   class Timeout < StandardError; end
     
   class Client    
-    
     PACK_N = "N"
+    
     def initialize(host, port, options = {})
       @host, @port = host, port
       @timeout = options[:timeout]
@@ -40,10 +41,10 @@ class BloomFilter
         @socket.write("#{[elements.size + 1].pack(PACK_N)}#{Protocol::INCLUDE_MANY}#{elements}")
         
         timeout_or_default(els) do
-          response = @socket.read(@socket.read(4).unpack(PACK_N).first)
+          response = BitVector.new(els.size, @socket.read(@socket.read(4).unpack(PACK_N).first))
           result = []
           els.size.times do |i|
-            result << els[i] if response[i,1] == Protocol::TRUE
+            result << els[i] if response[i] == 1
           end
           result
         end

data/lib/bloom_filter/protocol.rb

@@ -1,7 +1,6 @@
 class BloomFilter
   module Protocol
-    FALSE = "0"
-    TRUE = "1"
+    TRUE = 1.chr
     DEFAULT_SEPARATOR = ","    
     ADD = 0
     INCLUDE = 1

data/lib/bloom_filter/server.rb

@@ -1,5 +1,6 @@
 require 'bloom_filter'
 require 'bloom_filter/protocol'
+require 'bloom_filter/bit_vector'
 require 'eventmachine'
 
 
@@ -86,7 +87,10 @@ class BloomFilter
     end
 
     def write_response(response)
-      str = Array(response).inject("") { |s, t| s << (t ? Protocol::TRUE : Protocol::FALSE) }
+      response = Array(response)
+      vector = BitVector.new(response.size)
+      response.each_with_index { |t, index| vector[index] = t }
+      str = vector.to_s
       send_data([str.size].pack(PACK_N) + str)
     end
   end

data/test/bloom_filter_bit_vector_test.rb

@@ -0,0 +1,44 @@
+require 'test_helper'
+require 'bloom_filter/bit_vector'
+
+class BloomFilterBitVectorTest < Test::Unit::TestCase
+  context "with a bit vector of size 16" do
+    setup do
+      @bit_vector = BloomFilter::BitVector.new(16)
+    end
+    
+    should "return all 0s at first" do
+      @bit_vector.size.times do |i|
+        fail "returned #{@bit_vector[i]} for bit #{i}" if @bit_vector[i] != 0
+      end
+    end
+    
+    should "sets bits to 1 and returns 1 when retrieved" do
+      @bit_vector.size.times do |i|
+        @bit_vector[i] = 1
+        assert_equal 1, @bit_vector[i]
+      end
+    end
+    
+    should "sets bits back to 0" do
+      @bit_vector.size.times do |i|
+        @bit_vector[i] = 1
+        @bit_vector[i] = 0
+        assert_equal 0, @bit_vector[i]
+      end
+    end
+    
+    should "convert the bit vector into a string" do
+      assert @bit_vector.to_s.is_a?(String)
+    end
+    
+    should "convert back to the same bit vector" do
+      [0,3,6,7].each do |bit|
+        @bit_vector[bit] = 1
+      end
+      
+      assert_equal @bit_vector, BloomFilter::BitVector.new(16, @bit_vector.to_s)
+    end
+  end
+  
+end

data/test/bloom_filter_client_test.rb

@@ -13,42 +13,48 @@ class BloomFilterClientTest < Test::Unit::TestCase
     should "add elements" do
       @socket.expects(:write).with("#{[6].pack("N")}#{BloomFilter::Protocol::ADD}hello")
       @socket.expects(:read).with(4).returns([1].pack("N"))
-      @socket.expects(:read).with(1).returns("1")
+      @socket.expects(:read).with(1).returns(BloomFilter::Protocol::TRUE)
       assert @client.add("hello")
     end
     
     should "check for elements" do
       @socket.expects(:write).with("#{[6].pack("N")}#{BloomFilter::Protocol::INCLUDE}hello")
       @socket.expects(:read).with(4).returns([1].pack("N"))
-      @socket.expects(:read).with(1).returns("1")
+      @socket.expects(:read).with(1).returns(BloomFilter::Protocol::TRUE)
       assert @client.include?("hello")
     end
     
     should "return false for elements that arent there" do
       @socket.expects(:write).with("#{[4].pack("N")}#{BloomFilter::Protocol::INCLUDE}bye")
       @socket.expects(:read).with(4).returns([1].pack("N"))
-      @socket.expects(:read).with(1).returns("0")
+      @socket.expects(:read).with(1).returns(0.chr)
       assert !@client.include?("bye")
     end
     
     should "check for multiple elements" do
       @socket.expects(:write).with("#{[10].pack("N")}#{BloomFilter::Protocol::INCLUDE_MANY}hello,bye")
-      @socket.expects(:read).with(4).returns([2].pack("N"))
-      @socket.expects(:read).with(2).returns("10")
+      @socket.expects(:read).with(4).returns([1].pack("N"))
+      @socket.expects(:read).with(1).returns(1.chr)
       assert_equal ["hello"], @client & ["hello", "bye"]
+      
+      @socket.expects(:write).with("#{[10].pack("N")}#{BloomFilter::Protocol::INCLUDE_MANY}bye,hello")
+      @socket.expects(:read).with(4).returns([1].pack("N"))
+      @socket.expects(:read).with(1).returns(2.chr)
+      assert_equal ["hello"], @client & ["bye", "hello"]
+      
     end
     
     should "send dump" do
       @socket.expects(:write).with("#{[7].pack("N")}#{BloomFilter::Protocol::DUMP}/tmp/f")
       @socket.expects(:read).with(4).returns([1].pack("N"))
-      @socket.expects(:read).with(1).returns("1")
+      @socket.expects(:read).with(1).returns(BloomFilter::Protocol::TRUE)
       assert @client.dump("/tmp/f")
     end
     
     should "send load" do
       @socket.expects(:write).with("#{[7].pack("N")}#{BloomFilter::Protocol::LOAD}/tmp/f")
       @socket.expects(:read).with(4).returns([1].pack("N"))
-      @socket.expects(:read).with(1).returns("1")
+      @socket.expects(:read).with(1).returns(BloomFilter::Protocol::TRUE)
       assert @client.load("/tmp/f")
     end
     

data/test/bloom_filter_server_test.rb

@@ -15,22 +15,24 @@ class BloomFilterServerTest < Test::Unit::TestCase
     
     should "add elements" do
       @filter.expects(:add).with("hello")
-      expects_send_data("#{[1].pack('N')}1")
+      expects_send_data("#{[1].pack('N')}#{BloomFilter::Protocol::TRUE}")
       
       @server.receive_data("#{[6].pack("N")}#{BloomFilter::Protocol::ADD}hello")
     end
   
     should "check for elements" do
       @filter.expects(:include?).with("hello").returns(true)
-      expects_send_data("#{[1].pack("N")}1")
+      expects_send_data("#{[1].pack("N")}#{BloomFilter::Protocol::TRUE}")
       @server.receive_data("#{[6].pack("N")}#{BloomFilter::Protocol::INCLUDE}hello")
     end
     
     should "check for multiple elements" do
-      @filter.expects(:include?).with("hello").returns(true)
-      @filter.expects(:include?).with("bye").returns(false)
-      expects_send_data("#{[2].pack("N")}10")
+      @filter.expects(:include?).with("hello").returns(true).twice
+      @filter.expects(:include?).with("bye").returns(false).twice
+      expects_send_data("#{[1].pack("N")}#{1.chr}") # 1.chr => false, true
+      expects_send_data("#{[1].pack("N")}#{2.chr}") # 2.chr => true, false
       @server.receive_data("#{[10].pack("N")}#{BloomFilter::Protocol::INCLUDE_MANY}hello,bye")
+      @server.receive_data("#{[10].pack("N")}#{BloomFilter::Protocol::INCLUDE_MANY}bye,hello")
     end
     
     
@@ -42,7 +44,7 @@ class BloomFilterServerTest < Test::Unit::TestCase
       File.expects(:open).with(path, 'w').yields(file)
       file.expects(:write).with("dumped_data")
       
-      expects_send_data("#{[1].pack("N")}1")
+      expects_send_data("#{[1].pack("N")}#{BloomFilter::Protocol::TRUE}")
       
       
       @server.receive_data("#{[path.size + 1].pack("N")}#{BloomFilter::Protocol::DUMP}#{path}")
@@ -54,7 +56,7 @@ class BloomFilterServerTest < Test::Unit::TestCase
       @filter.expects(:replace).with(file)
       File.expects(:open).with(path, 'r').yields(file)
       
-      expects_send_data("#{[1].pack("N")}1")
+      expects_send_data("#{[1].pack("N")}#{BloomFilter::Protocol::TRUE}")
       
       @server.receive_data("#{[path.size + 1].pack("N")}#{BloomFilter::Protocol::LOAD}#{path}")
     end
@@ -62,7 +64,7 @@ class BloomFilterServerTest < Test::Unit::TestCase
     
     should "properly buffer partial messages" do
       @filter.expects(:add).with("hello")
-      expects_send_data("#{[1].pack('N')}1")
+      expects_send_data("#{[1].pack('N')}#{BloomFilter::Protocol::TRUE}")
       
       @server.receive_data("#{[6].pack("N")}#{BloomFilter::Protocol::ADD}hel")
       @server.receive_data("lo")
@@ -71,7 +73,7 @@ class BloomFilterServerTest < Test::Unit::TestCase
     
     should "handle two messages sent at once" do
       @filter.expects(:add).with("hello")
-      expects_send_data("#{[1].pack('N')}1").twice()
+      expects_send_data("#{[1].pack('N')}#{BloomFilter::Protocol::TRUE}").twice()
       @filter.expects(:add).with("bye")
 
       @server.receive_data("#{[6].pack("N")}#{BloomFilter::Protocol::ADD}hello#{[4].pack("N")}#{BloomFilter::Protocol::ADD}bye")

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification 
 name: bloom_filter
 version: !ruby/object:Gem::Version 
-  version: 0.6.1
+  version: 0.6.2
 platform: ruby
 authors: 
 - Arya Asemanfar
@@ -9,7 +9,7 @@ autorequire:
 bindir: bin
 cert_chain: []
 
-date: 2010-01-26 00:00:00 -08:00
+date: 2010-01-27 00:00:00 -08:00
 default_executable: bloom_filter_server
 dependencies: 
 - !ruby/object:Gem::Dependency 
@@ -61,9 +61,11 @@ files:
 - bin/bloom_filter_server
 - bloom_filter.gemspec
 - lib/bloom_filter.rb
+- lib/bloom_filter/bit_vector.rb
 - lib/bloom_filter/client.rb
 - lib/bloom_filter/protocol.rb
 - lib/bloom_filter/server.rb
+- test/bloom_filter_bit_vector_test.rb
 - test/bloom_filter_client_test.rb
 - test/bloom_filter_server_test.rb
 - test/bloom_filter_test.rb
@@ -97,6 +99,7 @@ signing_key:
 specification_version: 3
 summary: A simple BloomFilter implementation, usable in-process or as an EventMachine daemon.
 test_files: 
+- test/bloom_filter_bit_vector_test.rb
 - test/bloom_filter_client_test.rb
 - test/bloom_filter_server_test.rb
 - test/bloom_filter_test.rb