Finished a basic implementation in Ruby, adding it to the build.

README

@@ -39,6 +39,7 @@
  	- PHP
  	- Python
  	- Java
+ 	- Ruby
 
  	[A language missing? Please add it to the project.] 
 

build.xml

@@ -202,21 +202,17 @@
 			<fail message="Ruby unit-test failed: ${rubyunittestresults}">
 				<condition>
 					<not>
-						<matches pattern="OK"
+						<matches pattern="0 failures, 0 errors"
 						         string="${rubyunittestresults}" />
 					</not>
 				</condition>
 			</fail>
-			<echo>Ruby unit-tests: OK</echo>
+			<echo>Ruby Unit Test: OK</echo>
 		</sequential>
 	</macrodef>
 
 	<!-- TARGETS -->
 
-	<target name="rbtest">
-		<rubyunittest />
-	</target>
-
 	<target name="test" description="Runs the project unit tests">
 		<phpunit testcase="LZ77Test" file="${src.dir}/test/php/lz77Test.php" />
 		<jslint files="${src.dir}/main/js/lz77.js" />

src/main/ruby/lz77.rb

@@ -20,12 +20,138 @@
 # OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
 # THE SOFTWARE.
 #
+# @author Olle Törnström olle[at]studiomediatech[dot]com
+# @created 2010-02-07
+#
 class Compressor
-	def compress(data)
-		return "TODO: Implement!"
+
+	def initialize
+		@referencePrefix = "`"
+	    @referencePrefixCode = @referencePrefix[0]
+		@referenceIntBase = 96
+		@referenceIntFloorCode = " "[0]
+		@referenceIntCeilCode = @referenceIntFloorCode + @referenceIntBase - 1
+		@maxStringDistance = @referenceIntBase ** 2 - 1
+		@minStringLength = 5
+		@maxStringLength = @referenceIntBase ** 1 - 1 + @minStringLength
+		@maxWindowLength = @maxStringDistance + @minStringLength;
+		@defaultWindowLength = 144
+	end
+
+	def compress(data, windowLength = @defaultWindowLength)
+		compressed = ""
+		pos = 0
+		lastPos = data.length - @minStringLength
+		while pos < lastPos do
+			searchStart = [pos - windowLength, 0].max
+			matchLength = @minStringLength
+			foundMatch = false
+			bestMatchDistance = @maxStringDistance
+			bestMatchLength = 0
+			newCompressed = nil
+			while (searchStart + matchLength) < pos do
+				m1 = data[searchStart, matchLength]
+				m2 = data[pos, matchLength]
+				isValidMatch = (m1 == m2 and matchLength < @maxStringLength)
+				if isValidMatch then
+					matchLength += 1
+					foundMatch = true
+				else
+					realMatchLength = matchLength - 1
+					if foundMatch and realMatchLength > bestMatchLength then
+						bestMatchDistance = pos - searchStart - realMatchLength
+						bestMatchLength = realMatchLength
+					end
+					matchLength = @minStringLength
+					searchStart += 1
+					foundMatch = false
+				end
+			end
+			if bestMatchLength > 0 then
+				head = "" << @referencePrefix
+				head << encodeReferenceInt(bestMatchDistance, 2)
+				tail = encodeReferenceLength(bestMatchLength)
+				newCompressed = ""
+				newCompressed = head << tail
+				pos += bestMatchLength
+			else
+				if data[pos, 1] != @referencePrefix then
+					newCompressed = data[pos, 1]
+				else
+					newCompressed = @referencePrefix << @referencePrefix
+				end
+				pos += 1
+			end
+			compressed << newCompressed
+		end
+		return compressed + data[pos..-1].gsub(/`/, "``")
 	end
 
 	def decompress(data)
-		return "TODO: Implement!"
+		decompressed = ""
+		pos = 0
+		while pos < data.length do
+			currentChar = data[pos, 1]
+			if currentChar != @referencePrefix then
+				decompressed << currentChar				
+				pos += 1
+			else
+				nextChar = data[pos + 1, 1]
+				if nextChar != @referencePrefix then
+					distance = decodeReferenceInt(data[pos + 1, 2], 2)
+					length = decodeReferenceLength(data[pos + 3, 1])
+					start = decompressed.length - distance - length
+					stop = start + length
+					decompressed << decompressed[start..stop-1]
+					pos += @minStringLength - 1
+				else
+					decompressed << @referencePrefix
+					pos += 2
+				end
+			end
+		end
+		return decompressed
+	end
+
+	private # PRIVATE METHODS **************************************************
+
+	def encodeReferenceInt(value, width)
+		encoded = ""
+		if value >= 0 and value < ((@referenceIntBase ** width) - 1) then
+			while value > 0 do
+				encoded = ((value % @referenceIntBase) + @referenceIntFloorCode).chr << encoded
+				value = value / @referenceIntBase
+			end
+			missingLength = width - encoded.length
+			(0..missingLength-1).each do
+				encoded = (@referenceIntFloorCode).chr << encoded
+			end
+			return encoded
+		else
+			raise "Reference value out of range"
+		end
+	end
+
+	def encodeReferenceLength(length)
+		return encodeReferenceInt(length - @minStringLength, 1)
 	end
+
+	def decodeReferenceInt(data, width)
+		value = 0
+		(0..width-1).each do |i|
+			value *= @referenceIntBase			
+			charCode = data[i]
+			if charCode >= @referenceIntFloorCode and charCode <= @referenceIntCeilCode then
+				value += (charCode - @referenceIntFloorCode)
+			else
+				raise "Invalid char code"
+			end
+		end
+		return value
+	end
+
+	def decodeReferenceLength(data)
+		return decodeReferenceInt(data, 1) + @minStringLength
+	end
+
 end

src/test/php/lz77Test.php

@@ -9,7 +9,6 @@ class LZ77Test extends PHPUnit_Framework_TestCase {
 	protected $encodedTextData = "ababassbabasbabbba` '&bs` 1 sb` 4!` . abaaaa` *!";
 
 	protected function setUp() {
-
 		$this->compressor = new LZ77();
 	}
 
@@ -27,4 +26,5 @@ public function testDecompress() {
 		$this->assertEquals($this->textData, $result);
 	}
 }
+
 ?>

src/test/ruby/lz77test.rb

@@ -9,10 +9,12 @@ def setup
 	end
 
 	def testCompress
-		assert_equal(@encodedTextData, @compressor.compress(@textData))
+		result = @compressor.compress(@textData)
+		assert_equal(@encodedTextData, result)
 	end
 
 	def testDecompress
-		assert_equal(@textData, @compressor.decompress(@encodedTextData))
+		result = @compressor.decompress(@encodedTextData)
+		assert_equal(@textData, result)
 	end
 end