fix TestMassiveRuneDiffConversion by skip invalid unicode code point

Author: iambus

diffmatchpatch/diff.go

@@ -388,24 +388,29 @@ func (dmp *DiffMatchPatch) diffBisectSplit(runes1, runes2 []rune, x, y int,
 }
 
 // DiffLinesToChars splits two texts into a list of strings, and educes the texts to a string of hashes where each Unicode character represents one line.
-// It's slightly faster to call DiffLinesToRunes first, followed by DiffMainRunes.
 func (dmp *DiffMatchPatch) DiffLinesToChars(text1, text2 string) (string, string, []string) {
-	chars1, chars2, lineArray := dmp.diffLinesToStrings(text1, text2)
-	return chars1, chars2, lineArray
+	chars1, chars2, lineArray := dmp.diffLinesToIndexes(text1, text2)
+	return indexesToString(chars1), indexesToString(chars2), lineArray
 }
 
 // DiffLinesToRunes splits two texts into a list of runes.
 func (dmp *DiffMatchPatch) DiffLinesToRunes(text1, text2 string) ([]rune, []rune, []string) {
+	chars1, chars2, lineArray := dmp.diffLinesToIndexes(text1, text2)
+	return []rune(indexesToString(chars1)), []rune(indexesToString(chars2)), lineArray
+}
+
+// diffLinesToIndexes splits two texts into a list of indexes
+func (dmp *DiffMatchPatch) diffLinesToIndexes(text1, text2 string) ([]index, []index, []string) {
 	chars1, chars2, lineArray := dmp.diffLinesToStrings(text1, text2)
-	return []rune(chars1), []rune(chars2), lineArray
+	return chars1, chars2, lineArray
 }
 
 // DiffCharsToLines rehydrates the text in a diff from a string of line hashes to real lines of text.
 func (dmp *DiffMatchPatch) DiffCharsToLines(diffs []Diff, lineArray []string) []Diff {
 	hydrated := make([]Diff, 0, len(diffs))
 	for _, aDiff := range diffs {
 		var sb strings.Builder
-		for _, i := range []rune(aDiff.Text) {
+		for _, i := range stringToIndex(aDiff.Text) {
 			sb.WriteString(lineArray[i])
 		}
 		aDiff.Text = sb.String()
@@ -1301,24 +1306,24 @@ func (dmp *DiffMatchPatch) DiffFromDelta(text1 string, delta string) (diffs []Di
 }
 
 // diffLinesToStrings splits two texts into a list of strings. Each string represents one line.
-func (dmp *DiffMatchPatch) diffLinesToStrings(text1, text2 string) (string, string, []string) {
+func (dmp *DiffMatchPatch) diffLinesToStrings(text1, text2 string) ([]index, []index, []string) {
 	// '\x00' is a valid character, but various debuggers don't like it. So we'll insert a junk entry to avoid generating a null character.
 	lineArray := []string{""} // e.g. lineArray[4] == 'Hello\n'
 
 	//Each string has the index of lineArray which it points to
 	strIndexArray1 := dmp.diffLinesToStringsMunge(text1, &lineArray)
 	strIndexArray2 := dmp.diffLinesToStringsMunge(text2, &lineArray)
 
-	return intArrayToString(strIndexArray1), intArrayToString(strIndexArray2), lineArray
+	return strIndexArray1, strIndexArray2, lineArray
 }
 
 // diffLinesToStringsMunge splits a text into an array of strings, and reduces the texts to a []string.
-func (dmp *DiffMatchPatch) diffLinesToStringsMunge(text string, lineArray *[]string) []uint32 {
+func (dmp *DiffMatchPatch) diffLinesToStringsMunge(text string, lineArray *[]string) []index {
 	// Walk the text, pulling out a substring for each line. text.split('\n') would would temporarily double our memory footprint. Modifying text would create many large strings to garbage collect.
 	lineHash := map[string]int{} // e.g. lineHash['Hello\n'] == 4
 	lineStart := 0
 	lineEnd := -1
-	strs := []uint32{}
+	strs := []index{}
 
 	for lineEnd < len(text)-1 {
 		lineEnd = indexOf(text, "\n", lineStart)
@@ -1332,11 +1337,11 @@ func (dmp *DiffMatchPatch) diffLinesToStringsMunge(text string, lineArray *[]str
 		lineValue, ok := lineHash[line]
 
 		if ok {
-			strs = append(strs, uint32(lineValue))
+			strs = append(strs, index(lineValue))
 		} else {
 			*lineArray = append(*lineArray, line)
 			lineHash[line] = len(*lineArray) - 1
-			strs = append(strs, uint32(len(*lineArray)-1))
+			strs = append(strs, index(len(*lineArray)-1))
 		}
 	}
 

diffmatchpatch/index.go

@@ -0,0 +1,32 @@
+package diffmatchpatch
+
+type index uint32
+
+const runeSkipStart = 0xd800
+const runeSkipEnd = 0xdfff + 1
+const runeMax = 0x110000 // next invalid code point
+
+func stringToIndex(text string) []index {
+	runes := []rune(text)
+	indexes := make([]index, len(runes))
+	for i, r := range runes {
+		if r < runeSkipEnd {
+			indexes[i] = index(r)
+		} else {
+			indexes[i] = index(r) - (runeSkipEnd - runeSkipStart)
+		}
+	}
+	return indexes
+}
+
+func indexesToString(indexes []index) string {
+	runes := make([]rune, len(indexes))
+	for i, index := range indexes {
+		if index < runeSkipStart {
+			runes[i] = rune(index)
+		} else {
+			runes[i] = rune(index + (runeSkipEnd - runeSkipStart))
+		}
+	}
+	return string(runes)
+}

diffmatchpatch/index_test.go

@@ -0,0 +1,16 @@
+package diffmatchpatch
+
+import (
+	"github.com/stretchr/testify/assert"
+	"testing"
+)
+
+func TestIndexConversion(t *testing.T) {
+	n := runeMax - (runeSkipEnd - runeSkipStart)
+	indexes := make([]index, n)
+	for i := 0; i < n; i++ {
+		indexes[i] = index(i)
+	}
+	indexes2 := stringToIndex(indexesToString(indexes))
+	assert.EqualValues(t, indexes, indexes2)
+}

diffmatchpatch/stringutil.go

@@ -86,11 +86,3 @@ func runesIndex(r1, r2 []rune) int {
 	}
 	return -1
 }
-
-func intArrayToString(ns []uint32) string {
-	runes := make([]rune, len(ns))
-	for i := 0; i < len(ns); i++ {
-		runes[i] = rune(ns[i])
-	}
-	return string(runes)
-}