Merge pull request #27 from adrg/metrics-improvements

Fix support for multibyte characters

Author: adrg

.github/workflows/ci.yml

@@ -10,15 +10,15 @@ jobs:
     runs-on: ubuntu-latest
     strategy:
       matrix:
-        go: ['1.16']
+        go: ['1.17']
     steps:
     - name: Setup
       uses: actions/setup-go@v4
       with:
         go-version: ${{ matrix.go }}
 
     - name: Checkout
-      uses: actions/checkout@v3
+      uses: actions/checkout@v4
 
     - name: Dependencies
       run: |

.github/workflows/codeql-analysis.yml

@@ -24,7 +24,7 @@ jobs:
 
     steps:
     - name: Checkout repository
-      uses: actions/checkout@v3
+      uses: actions/checkout@v4
 
     - name: Initialize CodeQL
       uses: github/codeql-action/init@v2

go.mod

@@ -2,4 +2,4 @@ module github.com/adrg/strutil
 
 go 1.14
 
-require github.com/stretchr/testify v1.8.2
+require github.com/stretchr/testify v1.8.4

go.sum

@@ -8,8 +8,8 @@ github.com/stretchr/objx v0.4.0/go.mod h1:YvHI0jy2hoMjB+UWwv71VJQ9isScKT/TqJzVSS
 github.com/stretchr/objx v0.5.0/go.mod h1:Yh+to48EsGEfYuaHDzXPcE3xhTkx73EhmCGUpEOglKo=
 github.com/stretchr/testify v1.7.1/go.mod h1:6Fq8oRcR53rry900zMqJjRRixrwX3KX962/h/Wwjteg=
 github.com/stretchr/testify v1.8.0/go.mod h1:yNjHg4UonilssWZ8iaSj1OCr/vHnekPRkoO+kdMU+MU=
-github.com/stretchr/testify v1.8.2 h1:+h33VjcLVPDHtOdpUCuF+7gSuG3yGIftsP1YvFihtJ8=
-github.com/stretchr/testify v1.8.2/go.mod h1:w2LPCIKwWwSfY2zedu0+kehJoqGctiVI29o6fzry7u4=
+github.com/stretchr/testify v1.8.4 h1:CcVxjf3Q8PM0mHUKJCdn+eZZtm5yQwehR5yeSVQQcUk=
+github.com/stretchr/testify v1.8.4/go.mod h1:sz/lmYIOXD/1dqDmKjjqLyZ2RngseejIcXlSw2iwfAo=
 gopkg.in/check.v1 v0.0.0-20161208181325-20d25e280405 h1:yhCVgyC4o1eVCa2tZl7eS0r+SDo693bJlVdllGtEeKM=
 gopkg.in/check.v1 v0.0.0-20161208181325-20d25e280405/go.mod h1:Co6ibVJAznAaIkqp8huTwlJQCZ016jof/cbN4VW5Yz0=
 gopkg.in/yaml.v3 v3.0.0-20200313102051-9f266ea9e77c/go.mod h1:K4uyk7z7BCEPqu6E+C64Yfv1cQ7kz7rIZviUmN+EgEM=

internal/stringutil/stringutil.go

@@ -1,17 +1,15 @@
 package stringutil
 
-import "unicode/utf8"
-
 // CommonPrefix returns the common prefix of the specified strings. An empty
 // string is returned if the parameters have no prefix in common.
 func CommonPrefix(first, second string) string {
-	if utf8.RuneCountInString(first) > utf8.RuneCountInString(second) {
-		first, second = second, first
+	fRunes, sRunes := []rune(first), []rune(second)
+	if len(fRunes) > len(sRunes) {
+		fRunes, sRunes = sRunes, fRunes
 	}
 
 	var commonLen int
-	sRunes := []rune(second)
-	for i, r := range first {
+	for i, r := range fRunes {
 		if r != sRunes[i] {
 			break
 		}

internal/stringutil/stringutil_test.go

@@ -17,6 +17,14 @@ func TestCommonPrefix(t *testing.T) {
 		{"a", stringutil.CommonPrefix("aab", "ab")},
 		{"aa", stringutil.CommonPrefix("aab", "aaab")},
 		{"aa", stringutil.CommonPrefix("aaab", "aab")},
+		{"忧郁的乌龟", stringutil.CommonPrefix("忧郁的乌龟", "忧郁的乌龟")},
+		{"忧郁的", stringutil.CommonPrefix("忧郁的", "忧郁的乌龟")},
+		{"忧郁的", stringutil.CommonPrefix("忧郁的乌龟", "忧郁的")},
+		{"", stringutil.CommonPrefix("忧郁的乌龟", "郁的乌龟")},
+		{"", stringutil.CommonPrefix("郁的乌龟", "忧郁的乌龟")},
+		{"\u2019", stringutil.CommonPrefix("\u2019a", "\u2019b")},
+		{"a\u2019bc", stringutil.CommonPrefix("a\u2019bcd", "a\u2019bce")},
+		{"abc", stringutil.CommonPrefix("abc\u2019d", "abc\u2020d")},
 	})
 }
 

metrics/jaro.go

@@ -62,22 +62,25 @@ func (m *Jaro) Compare(a, b string) float64 {
 }
 
 func matchingRunes(a, b string, limit int) []rune {
-	common := []rune{}
-	runesB := []rune(b)
-	lenB := len(runesB)
-
-	for i, r := range a {
+	var (
+		runesA      = []rune(a)
+		runesB      = []rune(b)
+		runesCommon = []rune{}
+		lenB        = len(runesB)
+	)
+
+	for i, r := range runesA {
 		end := mathutil.Min(i+limit+1, lenB)
 		for j := mathutil.Max(0, i-limit); j < end; j++ {
 			if r == runesB[j] && runesB[j] != -1 {
-				common = append(common, runesB[j])
+				runesCommon = append(runesCommon, runesB[j])
 				runesB[j] = -1
 				break
 			}
 		}
 	}
 
-	return common
+	return runesCommon
 }
 
 func transpositions(a, b []rune) int {

metrics/levenshtein.go

@@ -2,7 +2,6 @@ package metrics
 
 import (
 	"strings"
-	"unicode/utf8"
 
 	"github.com/adrg/strutil/internal/mathutil"
 )
@@ -56,8 +55,15 @@ func (m *Levenshtein) Distance(a, b string) int {
 }
 
 func (m *Levenshtein) distance(a, b string) (int, int) {
+	// Lower terms if case insensitive comparison is specified.
+	if !m.CaseSensitive {
+		a = strings.ToLower(a)
+		b = strings.ToLower(b)
+	}
+	runesA, runesB := []rune(a), []rune(b)
+
 	// Check if both terms are empty.
-	lenA, lenB := utf8.RuneCountInString(a), utf8.RuneCountInString(b)
+	lenA, lenB := len(runesA), len(runesB)
 	if lenA == 0 && lenB == 0 {
 		return 0, 0
 	}
@@ -71,12 +77,6 @@ func (m *Levenshtein) distance(a, b string) (int, int) {
 		return m.DeleteCost * lenA, maxLen
 	}
 
-	// Lower terms if case insensitive comparison is specified.
-	if !m.CaseSensitive {
-		a = strings.ToLower(a)
-		b = strings.ToLower(b)
-	}
-
 	// Initialize cost slice.
 	prevCol := make([]int, lenB+1)
 	for i := 0; i <= lenB; i++ {
@@ -92,7 +92,7 @@ func (m *Levenshtein) distance(a, b string) (int, int) {
 			insCost := col[j] + m.InsertCost
 
 			subCost := prevCol[j]
-			if a[i] != b[j] {
+			if runesA[i] != runesB[j] {
 				subCost += m.ReplaceCost
 			}
 

metrics/metrics_test.go

@@ -17,6 +17,9 @@ func TestHamming(t *testing.T) {
 	require.Equal(t, 0, h.Distance("", ""))
 	require.Equal(t, "0.75", sf(h.Compare("text", "test")))
 	require.Equal(t, "0.50", sf(h.Compare("once", "one")))
+	require.Equal(t, "1.00", sf(h.Compare("ab\u2019c", "ab\u2019c")))
+	require.Equal(t, "0.75", sf(h.Compare("ab\u2019d", "ab\u2019c")))
+	require.Equal(t, "0.75", sf(h.Compare("ab\u2018c", "ab\u2019c")))
 	h.CaseSensitive = false
 	require.Equal(t, "0.50", sf(h.Compare("one", "ONCE")))
 }
@@ -25,6 +28,9 @@ func TestJaccard(t *testing.T) {
 	j := metrics.NewJaccard()
 	require.Equal(t, "1.00", sf(j.Compare("", "")))
 	require.Equal(t, "0.00", sf(j.Compare("a", "b")))
+	require.Equal(t, "1.00", sf(j.Compare("ab\u2019c", "ab\u2019c")))
+	require.Equal(t, "0.50", sf(j.Compare("ab\u2019d", "ab\u2019c")))
+	require.Equal(t, "0.20", sf(j.Compare("ab\u2018c", "ab\u2019c")))
 	require.Equal(t, "0.43", sf(j.Compare("night", "alright")))
 	j.NgramSize = 0
 	require.Equal(t, "0.43", sf(j.Compare("night", "alright")))
@@ -37,6 +43,9 @@ func TestJaro(t *testing.T) {
 	j := metrics.NewJaro()
 	require.Equal(t, "1.00", sf(j.Compare("", "")))
 	require.Equal(t, "0.00", sf(j.Compare("test", "")))
+	require.Equal(t, "1.00", sf(j.Compare("ab\u2019c", "ab\u2019c")))
+	require.Equal(t, "0.83", sf(j.Compare("ab\u2019d", "ab\u2019c")))
+	require.Equal(t, "0.83", sf(j.Compare("ab\u2018c", "ab\u2019c")))
 	require.Equal(t, "0.00", sf(j.Compare("a", "b")))
 	require.Equal(t, "0.78", sf(j.Compare("sort", "shirt")))
 	require.Equal(t, "0.64", sf(j.Compare("sort", "report")))
@@ -48,6 +57,9 @@ func TestJaroWinkler(t *testing.T) {
 	j := metrics.NewJaroWinkler()
 	require.Equal(t, "1.00", sf(j.Compare("", "")))
 	require.Equal(t, "0.00", sf(j.Compare("test", "")))
+	require.Equal(t, "1.00", sf(j.Compare("ab\u2019c", "ab\u2019c")))
+	require.Equal(t, "0.88", sf(j.Compare("ab\u2019d", "ab\u2019c")))
+	require.Equal(t, "0.87", sf(j.Compare("ab\u2018c", "ab\u2019c")))
 	require.Equal(t, "0.80", sf(j.Compare("sort", "shirt")))
 	require.Equal(t, "0.94", sf(j.Compare("charm", "charmed")))
 	j.CaseSensitive = false
@@ -59,11 +71,19 @@ func TestLevenshtein(t *testing.T) {
 	require.Equal(t, 0, l.Distance("", ""))
 	require.Equal(t, 4, l.Distance("test", ""))
 	require.Equal(t, 4, l.Distance("", "test"))
+	require.Equal(t, 0, l.Distance("ab\u2019c", "ab\u2019c"))
+	require.Equal(t, 1, l.Distance("ab\u2019d", "ab\u2019c"))
+	require.Equal(t, 1, l.Distance("ab\u2018c", "ab\u2019c"))
 	require.Equal(t, "0.40", sf(l.Compare("book", "brick")))
+	require.Equal(t, "0.75", sf(l.Compare("ab\u2019d", "ab\u2019c")))
+	require.Equal(t, "0.75", sf(l.Compare("ab\u2018c", "ab\u2019c")))
 	l.CaseSensitive = false
 	require.Equal(t, "0.80", sf(l.Compare("hello", "jello")))
 	l.ReplaceCost = 2
 	require.Equal(t, "0.60", sf(l.Compare("hello", "JELLO")))
+	require.Equal(t, "1.00", sf(l.Compare("ab\u2019c", "ab\u2019c")))
+	require.Equal(t, "0.50", sf(l.Compare("ab\u2019d", "ab\u2019c")))
+	require.Equal(t, "0.50", sf(l.Compare("ab\u2018c", "ab\u2019c")))
 }
 
 func TestOperlapCoefficient(t *testing.T) {
@@ -72,6 +92,9 @@ func TestOperlapCoefficient(t *testing.T) {
 	require.Equal(t, "0.75", sf(o.Compare("night", "alright")))
 	require.Equal(t, "0.00", sf(o.Compare("aa", "")))
 	require.Equal(t, "0.00", sf(o.Compare("bb", "")))
+	require.Equal(t, "1.00", sf(o.Compare("ab\u2019c", "ab\u2019c")))
+	require.Equal(t, "0.67", sf(o.Compare("ab\u2019d", "ab\u2019c")))
+	require.Equal(t, "0.33", sf(o.Compare("ab\u2018c", "ab\u2019c")))
 	o.NgramSize = 0
 	require.Equal(t, "0.75", sf(o.Compare("night", "alright")))
 	require.Equal(t, "1.00", sf(o.Compare("aa", "aaaa")))
@@ -87,6 +110,9 @@ func TestSmithWatermanGotoh(t *testing.T) {
 	require.Equal(t, "0.00", sf(s.Compare("test", "")))
 	require.Equal(t, "0.00", sf(s.Compare("", "test")))
 	require.Equal(t, "0.88", sf(s.Compare("a pink kitten", "a kitten")))
+	require.Equal(t, "1.00", sf(s.Compare("ab\u2019c", "ab\u2019c")))
+	require.Equal(t, "0.75", sf(s.Compare("ab\u2019d", "ab\u2019c")))
+	require.Equal(t, "0.50", sf(s.Compare("ab\u2018c", "ab\u2019c")))
 	s.Substitution = nil
 	require.Equal(t, "0.88", sf(s.Compare("a pink kitten", "a kitten")))
 	s.CaseSensitive = false
@@ -103,6 +129,9 @@ func TestSorensenDice(t *testing.T) {
 	require.Equal(t, "1.00", sf(s.Compare("", "")))
 	require.Equal(t, "0.00", sf(s.Compare("a", "b")))
 	require.Equal(t, "0.60", sf(s.Compare("night", "alright")))
+	require.Equal(t, "1.00", sf(s.Compare("ab\u2019c", "ab\u2019c")))
+	require.Equal(t, "0.67", sf(s.Compare("ab\u2019d", "ab\u2019c")))
+	require.Equal(t, "0.33", sf(s.Compare("ab\u2018c", "ab\u2019c")))
 	s.NgramSize = 0
 	require.Equal(t, "0.60", sf(s.Compare("night", "alright")))
 	s.CaseSensitive = false