Updated smetrics package and changed the Jaro implementation to inclu…

…de rounding of transposition halves

finder/algorithm.go

@@ -56,7 +56,11 @@ func NewWagnerFischer(insert, delete, substitution int) Algorithm {
 }
 
 // NewJaro returns the default Jaro algorithm
-// @see https://rosettacode.org/wiki/Jaro_distance#Go
+// @see Original https://rosettacode.org/wiki/Jaro_distance#Go
+// Relevant discussions: https://github.com/xrash/smetrics/issues/7#issuecomment-664794681
+// Changes over original:
+// - Reduced allocations
+// - Added rounding on the unaligned matches as per: http://www.alias-i.com/lingpipe/docs/api/com/aliasi/spell/JaroWinklerDistance.html
 //nolint:gocyclo
 func NewJaro() Algorithm {
 	return func(a, b string) float64 {
@@ -127,6 +131,6 @@ func NewJaro() Algorithm {
 
 		return (matches/float64(len(a)) +
 			matches/float64(len(b)) +
-			(matches-(transpositions/2))/matches) / 3
+			(matches-math.Floor(transpositions/2))/matches) / 3
 	}
 }

finder/algorithm_test.go

@@ -501,7 +501,7 @@ func TestHomoPhoneJaroImplementations(t *testing.T) {
 		var scores = make([]float64, 4)
 		scores[smetricsJaro] = smetrics.Jaro(a, b)
 		scores[rosettaJaroV0] = RosettaJaroV0(a, b)
-		scores[rosettaJaroV1] = RosettaJaroV1(a, b)
+		scores[rosettaJaroV1] = NewJaro()(a, b)
 		scores[jaroDistanceMasatana] = func() float64 {
 			s, _ := JaroDistanceMasatana(a, b)
 			return s
@@ -525,6 +525,7 @@ func TestHomoPhoneJaroImplementations(t *testing.T) {
 }
 
 func TestJaroImplementations(t *testing.T) {
+	RosettaJaroV1 := NewJaro()
 	for _, tt := range jaroReferenceList {
 		score := NewJaro()(tt.a, tt.b)
 
@@ -670,6 +671,7 @@ func BenchmarkJaroImplementations(b *testing.B) {
 			})
 
 			b.Run("RosettaJaro V1", func(b *testing.B) {
+				RosettaJaroV1 := NewJaro()
 				b.ResetTimer()
 				b.ReportAllocs()
 				for i := 0; i < b.N; i++ {
@@ -711,6 +713,7 @@ func BenchmarkRosettaJaro(b *testing.B) {
 		}
 	})
 	b.Run("Single alloc", func(b *testing.B) {
+		RosettaJaroV1 := NewJaro()
 		b.ReportAllocs()
 		b.ResetTimer()
 
@@ -792,75 +795,3 @@ func RosettaJaroV0(a, b string) float64 {
 		matches/float64(len(b)) +
 		(matches-(transpositions/2))/matches) / 3
 }
-
-// @see https://rosettacode.org/wiki/Jaro_distance#Go
-// Changes:
-// - Allocation reduction
-func RosettaJaroV1(a, b string) float64 {
-	if len(a) == 0 && len(b) == 0 {
-		return 1
-	}
-	if len(a) == 0 || len(b) == 0 {
-		return 0
-	}
-	matchDistance := len(a)
-	if len(b) > matchDistance {
-		matchDistance = len(b)
-	}
-
-	matchDistance = matchDistance/2 - 1
-	matchesCollected := make([]bool, len(a)+len(b))
-
-	var matches float64
-	var transpositions float64
-	for i := range a {
-		start := i - matchDistance
-		if start < 0 {
-			start = 0
-		}
-
-		end := i + matchDistance + 1
-		if end > len(b) {
-			end = len(b)
-		}
-
-		for k := start; k < end; k++ {
-			if matchesCollected[k+len(a)] {
-				continue
-			}
-			if a[i] != b[k] {
-				continue
-			}
-
-			matchesCollected[i] = true
-			matchesCollected[k+len(a)] = true
-			matches++
-			break
-		}
-	}
-
-	if matches == 0 {
-		return 0
-	}
-
-	k := 0
-	for i := range a {
-		if !matchesCollected[i] {
-			continue
-		}
-
-		for !matchesCollected[k+len(a)] {
-			k++
-		}
-
-		if a[i] != b[k] {
-			transpositions++
-		}
-
-		k++
-	}
-
-	return (matches/float64(len(a)) +
-		matches/float64(len(b)) +
-		(matches-(transpositions/2))/matches) / 3
-}

go.mod

@@ -9,7 +9,7 @@ require (
 	github.com/mitchellh/gox v1.0.1 // indirect
 	github.com/rs/cors v1.7.0
 	github.com/sirupsen/logrus v1.5.0
-	github.com/xrash/smetrics v0.0.0-20170218160415-a3153f7040e9
+	github.com/xrash/smetrics v0.0.0-20200723181607-f06e43cca1ab
 	github.com/zikes/sift4 v0.0.0-20151103205100-8e89a8aebc1f
 	golang.org/x/crypto v0.0.0-20200323165209-0ec3e9974c59 // indirect
 	golang.org/x/net v0.0.0-20200324143707-d3edc9973b7e // indirect

go.sum

@@ -40,6 +40,8 @@ github.com/stretchr/testify v1.2.2/go.mod h1:a8OnRcib4nhh0OaRAV+Yts87kKdq0PP7pXf
 github.com/stretchr/testify v1.3.0/go.mod h1:M5WIy9Dh21IEIfnGCwXGc5bZfKNJtfHm1UVUgZn+9EI=
 github.com/xrash/smetrics v0.0.0-20170218160415-a3153f7040e9 h1:w8V9v0qVympSF6GjdjIyeqR7+EVhAF9CBQmkmW7Zw0w=
 github.com/xrash/smetrics v0.0.0-20170218160415-a3153f7040e9/go.mod h1:N3UwUGtsrSj3ccvlPHLoLsHnpR27oXr4ZE984MbSER8=
+github.com/xrash/smetrics v0.0.0-20200723181607-f06e43cca1ab h1:UJJSloV3F3NHvn3mWDJ6s2441UlIzjcyn2YUcEPgL0c=
+github.com/xrash/smetrics v0.0.0-20200723181607-f06e43cca1ab/go.mod h1:N3UwUGtsrSj3ccvlPHLoLsHnpR27oXr4ZE984MbSER8=
 github.com/zikes/sift4 v0.0.0-20151103205100-8e89a8aebc1f h1:uwsos4EwJT8IjkwPbPl78dIoouY41YQt4ocQAAKo5+o=
 github.com/zikes/sift4 v0.0.0-20151103205100-8e89a8aebc1f/go.mod h1:RkT6UOQ/xx22IhYMzJoJsW3eMFc96fptPf6k/0dgncE=
 golang.org/x/crypto v0.0.0-20180904163835-0709b304e793/go.mod h1:6SG95UA2DQfeDnfUPMdvaQW0Q7yPrPDi9nlGo2tz2b4=