Better handling for zero length arrays, remove need for return keywords, and more consistent application of equality checks on string compare methods.

4 files changed, 43 insertions, 47 deletions

diff --git a/core/source/core/scala/org/hashtree/stringmetric/distance/HammingMetric.scala b/core/source/core/scala/org/hashtree/stringmetric/distance/HammingMetric.scala
index 657c818..f40062e 100755
--- a/core/source/core/scala/org/hashtree/stringmetric/distance/HammingMetric.scala
+++ b/core/source/core/scala/org/hashtree/stringmetric/distance/HammingMetric.scala
@@ -8,24 +8,24 @@ object HammingMetric extends StringMetric {
 		val ca1 = stringCleaner.clean(charArray1)
 		val ca2 = stringCleaner.clean(charArray2)
 
-		if (ca1.length == 0 || ca2.length == 0 || ca1.length != ca2.length)
-			None
-		else
-			Some(hamming(ca1, ca2))
+		if (ca1.length == 0 || ca2.length == 0 || ca1.length != ca2.length) None
+		else Some(hamming(ca1, ca2))
 	}
 
 	override def compare(string1: String, string2: String)(implicit stringCleaner: StringCleaner): Option[Int] = {
-		compare(
-			stringCleaner.clean(string1.toCharArray),
-			stringCleaner.clean(string2.toCharArray)
-		)(new StringCleanerDelegate)
+		if (string1.length > 0 && string1.length == string2.length && string1 == string2) Some(0)
+		else
+			compare(
+				stringCleaner.clean(string1.toCharArray),
+				stringCleaner.clean(string2.toCharArray)
+			)(new StringCleanerDelegate)
 	}
 
 	private[this] def hamming(ct: CompareTuple[Char]) = {
-		require(ct._1.length > 0)
-		require(ct._2.length > 0)
 		require(ct._1.length == ct._2.length)
 
-		ct._1.zip(ct._2).count(t => t._1 != t._2)
+		if (ct._1.length == 0) 0
+		else
+			ct._1.zip(ct._2).count(t => t._1 != t._2)
 	}
 }
\ No newline at end of file
diff --git a/core/source/core/scala/org/hashtree/stringmetric/distance/JaroMetric.scala b/core/source/core/scala/org/hashtree/stringmetric/distance/JaroMetric.scala
index a77811b..5da024e 100755
--- a/core/source/core/scala/org/hashtree/stringmetric/distance/JaroMetric.scala
+++ b/core/source/core/scala/org/hashtree/stringmetric/distance/JaroMetric.scala
@@ -13,30 +13,28 @@ object JaroMetric extends StringMetric {
 		val ca1 = stringCleaner.clean(charArray1)
 		val ca2 = stringCleaner.clean(charArray2)
 
-		// Return None if either character array lacks length.
-		if (ca1.length == 0 || ca2.length == 0) return None
-
-		val mt = `match`((ca1, ca2))
-		val ms = scoreMatches((mt._1, mt._2))
-		val ts = scoreTranspositions((mt._1, mt._2))
-
-		// Return 0 if matches score is 0.
-		if (ms == 0) return Some(0f)
-
-		Some(((ms.toFloat / ca1.length) + (ms.toFloat / ca2.length) + ((ms.toFloat - ts) / ms)) / 3)
+		if (ca1.length == 0 || ca2.length == 0) None
+		else {
+			val mt = `match`((ca1, ca2))
+			val ms = scoreMatches((mt._1, mt._2))
+			val ts = scoreTranspositions((mt._1, mt._2))
+
+			if (ms == 0) Some(0f)
+			else
+				Some(((ms.toFloat / ca1.length) + (ms.toFloat / ca2.length) + ((ms.toFloat - ts) / ms)) / 3)
+		}
 	}
 
 	override def compare(string1: String, string2: String)(implicit stringCleaner: StringCleaner): Option[Float] = {
-		// Return 1 if strings are an exact match.
-		if (string1.length > 0 && string1 == string2) return Some(1f)
-
-		compare(
-			stringCleaner.clean(string1.toCharArray),
-			stringCleaner.clean(string2.toCharArray)
-		)(new StringCleanerDelegate)
+		if (string1.length > 0 && string1.length == string2.length && string1 == string2) Some(1f)
+		else
+			compare(
+				stringCleaner.clean(string1.toCharArray),
+				stringCleaner.clean(string2.toCharArray)
+			)(new StringCleanerDelegate)
 	}
 
-	private[this] def `match`(ct: CompareTuple[Char]) = {
+	private[this] def `match`(ct: CompareTuple[Char]): MatchTuple[Char] = {
 		val window = math.abs((math.max(ct._1.length, ct._2.length) / 2f).floor.toInt - 1)
 		val one = ArrayBuffer.empty[Int]
 		val two = HashSet.empty[Int]
diff --git a/core/source/core/scala/org/hashtree/stringmetric/distance/JaroWinklerMetric.scala b/core/source/core/scala/org/hashtree/stringmetric/distance/JaroWinklerMetric.scala
index 568a69e..a8ae494 100755
--- a/core/source/core/scala/org/hashtree/stringmetric/distance/JaroWinklerMetric.scala
+++ b/core/source/core/scala/org/hashtree/stringmetric/distance/JaroWinklerMetric.scala
@@ -23,12 +23,11 @@ object JaroWinklerMetric extends StringMetric {
 	}
 
 	override def compare(string1: String, string2: String)(implicit stringCleaner: StringCleaner): Option[Float] = {
-		// Return 1 if strings are an exact match.
-		if (string1.length > 0 && string1 == string2) return Some(1f)
-
-		compare(
-			stringCleaner.clean(string1.toCharArray),
-			stringCleaner.clean(string2.toCharArray)
-		)(new StringCleanerDelegate)
+		if (string1.length > 0 && string1.length == string2.length && string1 == string2) Some(1f)
+		else
+			compare(
+				stringCleaner.clean(string1.toCharArray),
+				stringCleaner.clean(string2.toCharArray)
+			)(new StringCleanerDelegate)
 	}
 }
\ No newline at end of file
diff --git a/core/source/core/scala/org/hashtree/stringmetric/distance/LevenshteinMetric.scala b/core/source/core/scala/org/hashtree/stringmetric/distance/LevenshteinMetric.scala
index ade1178..4c58ff1 100755
--- a/core/source/core/scala/org/hashtree/stringmetric/distance/LevenshteinMetric.scala
+++ b/core/source/core/scala/org/hashtree/stringmetric/distance/LevenshteinMetric.scala
@@ -8,8 +8,7 @@ object LevenshteinMetric extends StringMetric {
 		val ca1 = stringCleaner.clean(charArray1)
 		val ca2 = stringCleaner.clean(charArray2)
 
-		if (ca1.length == 0 && ca2.length == 0)
-			None
+		if (ca1.length == 0 && ca2.length == 0) None
 		else {
 			val levenshteinMemoize = Memoize.Y(levenshtein)
 
@@ -18,17 +17,17 @@ object LevenshteinMetric extends StringMetric {
 	}
 
 	override def compare(string1: String, string2: String)(implicit stringCleaner: StringCleaner): Option[Int] = {
-		compare(
-			stringCleaner.clean(string1.toCharArray),
-			stringCleaner.clean(string2.toCharArray)
-		)(new StringCleanerDelegate)
+		if (string1.length > 0 && string1.length == string2.length && string1 == string2) Some(0)
+		else
+			compare(
+				stringCleaner.clean(string1.toCharArray),
+				stringCleaner.clean(string2.toCharArray)
+			)(new StringCleanerDelegate)
 	}
 
 	private[this] def levenshtein(f: CompareTuple[Char] => Int)(ct: CompareTuple[Char]): Int = {
-		if (ct._1.length == 0)
-			ct._2.length
-		else if (ct._2.length == 0)
-			ct._1.length
+		if (ct._1.length == 0) ct._2.length
+		else if (ct._2.length == 0) ct._1.length
 		else {
 			math.min(
 				math.min(