Upgraded to latest rivo/uniseg. Also implemented basic emoji handling in StringWidth. See #59

Author: rivo

.github/workflows/test.yaml

@@ -11,7 +11,7 @@ jobs:
     strategy:
       matrix:
         os: [windows-latest, macos-latest, ubuntu-latest]
-        go: ["1.15", "1.16"]
+        go: ["1.18.x", "1.19.x"]
     runs-on: ${{ matrix.os }}
     steps:
       - uses: actions/checkout@v2

benchmark_test.go

@@ -113,21 +113,19 @@ func benchString1Width(b *testing.B, eastAsianWidth bool, start, stop rune, want
 	return n
 }
 func BenchmarkString1WidthAll(b *testing.B) {
-	benchSink = benchString1Width(b, false, 0, utf8.MaxRune+1, 1295990)
+	benchSink = benchString1Width(b, false, 0, utf8.MaxRune+1, 1298422)
 }
 func BenchmarkString1Width768(b *testing.B) {
 	benchSink = benchString1Width(b, false, 0, 0x300, 702)
 }
 func BenchmarkString1WidthAllEastAsian(b *testing.B) {
-	benchSink = benchString1Width(b, true, 0, utf8.MaxRune+1, 1436664)
+	benchSink = benchString1Width(b, true, 0, utf8.MaxRune+1, 1439014)
 }
 func BenchmarkString1Width768EastAsian(b *testing.B) {
 	benchSink = benchString1Width(b, true, 0, 0x300, 794)
 }
 
-//
 // tables
-//
 func benchTable(b *testing.B, tbl table) int {
 	n := 0
 	for i := 0; i < b.N; i++ {

go.mod

@@ -1,5 +1,5 @@
 module github.com/mattn/go-runewidth
 
-go 1.9
+go 1.18
 
-require github.com/rivo/uniseg v0.3.1
+require github.com/rivo/uniseg v0.4.2

go.sum

@@ -1,2 +1,4 @@
 github.com/rivo/uniseg v0.3.1 h1:SDPP7SHNl1L7KrEFCSJslJ/DM9DT02Nq2C61XrfHMmk=
 github.com/rivo/uniseg v0.3.1/go.mod h1:FN3SvrM+Zdj16jyLfmOkMNblXMcoc8DfTHruCPUcx88=
+github.com/rivo/uniseg v0.4.2 h1:YwD0ulJSJytLpiaWua0sBDusfsCZohxjxzVTYjwxfV8=
+github.com/rivo/uniseg v0.4.2/go.mod h1:FN3SvrM+Zdj16jyLfmOkMNblXMcoc8DfTHruCPUcx88=

runewidth.go

@@ -179,11 +179,12 @@ func (c *Condition) StringWidth(s string) (width int) {
 	for len(s) > 0 {
 		var chWidth int
 		cl, s, _, state = uniseg.FirstGraphemeClusterInString(s, state)
-		for _, r := range cl {
-			chWidth = c.RuneWidth(r)
-			if chWidth > 0 {
-				break // Our best guess at this point is to use the width of the first non-zero-width rune.
+		for index, r := range cl {
+			if index == 0 && inTable(r, emoji) {
+				chWidth = 2 // Not the optimal solution but it will work in most cases.
+				break
 			}
+			chWidth += c.RuneWidth(r)
 		}
 		width += chWidth
 	}
@@ -204,11 +205,12 @@ func (c *Condition) Truncate(s string, w int, tail string) string {
 			chWidth int
 		)
 		ch, substr, _, state = uniseg.FirstGraphemeClusterInString(substr, state)
-		for _, r := range ch {
-			chWidth = c.RuneWidth(r)
-			if chWidth > 0 {
-				break // See StringWidth() for details.
+		for index, r := range ch {
+			if index == 0 && inTable(r, emoji) {
+				chWidth = 2 // Not the optimal solution but it will work in most cases.
+				break
 			}
+			chWidth += c.RuneWidth(r)
 		}
 		if width+chWidth > w {
 			break

runewidth_table.go

@@ -44,7 +44,7 @@ var tables = []tableInfo{
 	{combining, "combining", 465, "3cce13deb5e23f9f7327f2b1ef162328285a7dcf277a98302a8f7cdd43971268"},
 	{doublewidth, "doublewidth", 182440, "3d16eda8650dc2c92d6318d32f0b4a74fda5a278db2d4544b1dd65863394823c"},
 	{ambiguous, "ambiguous", 138739, "d05e339a10f296de6547ff3d6c5aee32f627f6555477afebd4a3b7e3cf74c9e3"},
-	{emoji, "emoji", 3535, "9ec17351601d49c535658de8d129c1d0ccda2e620669fc39a2faaee7dedcef6d"},
+	{emoji, "emoji", 3561, "3f1ae3b21ead544a59ab68fe541e8b098adc4fe1ffe3389a523a01dbfd8c1d75"},
 	{narrow, "narrow", 111, "fa897699c5e3cd9141c638d539331b0bdd508b874e22996c5e929767d455fc5a"},
 	{neutral, "neutral", 27333, "5455f5e75c307f70b4e9b2384dc5a8bcd91a4c5e2b24b2b185dfad4d860ee5c2"},
 }
@@ -462,7 +462,7 @@ func TestZeroWidthJoiner(t *testing.T) {
 		{"\u200d🍳", 2},
 		{"👨\u200d👨", 2},
 		{"👨\u200d👨\u200d👧", 2},
-		{"🏳️\u200d🌈", 1},
+		{"🏳️\u200d🌈", 2},
 		{"あ👩\u200d🍳い", 6},
 		{"あ\u200d🍳い", 6},
 		{"あ\u200dい", 4},

runewidth_test.go

@@ -1,3 +1,4 @@
+//go:build ignore
 // +build ignore
 
 // Generate runewidth_table.go from data at https://unicode.org/
@@ -166,6 +167,16 @@ func emoji(out io.Writer, in io.Reader) error {
 		})
 	}
 
+	// We also want regional indicator symbols (flags) to be part of the Emoji
+	// table. They are U+1F1E6..U+1F1FF.
+	for index, r := range arr {
+		if r.lo >= 0x1f1ff {
+			arr = append(arr[:index+1], arr[index:]...)
+			arr[index] = rrange{lo: 0x1f1e6, hi: 0x1f1ff}
+			break
+		}
+	}
+
 	shapeup(&arr)
 	generate(out, "emoji", arr)