优化

finalseg/finalseg.go

@@ -10,88 +10,86 @@ var (
 	reSkip = regexp.MustCompile(`(\d+\.\d+|[a-zA-Z0-9]+)`)
 )
 
-func cutHan(sentence string) chan string {
-	result := make(chan string)
-	go func() {
-		runes := []rune(sentence)
-		_, posList := viterbi(runes, []byte{'B', 'M', 'E', 'S'})
-		begin, next := 0, 0
-		for i, char := range runes {
-			pos := posList[i]
-			switch pos {
-			case 'B':
-				begin = i
-			case 'E':
-				result <- string(runes[begin : i+1])
-				next = i + 1
-			case 'S':
-				result <- string(char)
-				next = i + 1
-			}
+func cutHan(sentence string) []string {
+	result := make([]string, 0, 10)
+
+	runes := []rune(sentence)
+	_, posList := viterbi(runes, []byte{'B', 'M', 'E', 'S'})
+	begin, next := 0, 0
+	for i, char := range runes {
+		pos := posList[i]
+		switch pos {
+		case 'B':
+			begin = i
+		case 'E':
+			result = append(result, string(runes[begin:i+1]))
+			next = i + 1
+		case 'S':
+			result = append(result, string(char))
+			next = i + 1
 		}
-		if next < len(runes) {
-			result <- string(runes[next:])
-		}
-		close(result)
-	}()
+	}
+	if next < len(runes) {
+		result = append(result, string(runes[next:]))
+	}
+
 	return result
 }
 
 // Cut cuts sentence into words using Hidden Markov Model with Viterbi
 // algorithm. It is used by Jiebago for unknonw words.
-func Cut(sentence string) chan string {
-	result := make(chan string)
+func Cut(sentence string) []string {
+	result := make([]string, 0, 10)
 	s := sentence
 	var hans string
 	var hanLoc []int
 	var nonhanLoc []int
-	go func() {
-		for {
-			hanLoc = reHan.FindStringIndex(s)
-			if hanLoc == nil {
-				if len(s) == 0 {
-					break
-				}
-			} else if hanLoc[0] == 0 {
-				hans = s[hanLoc[0]:hanLoc[1]]
-				s = s[hanLoc[1]:]
-				for han := range cutHan(hans) {
-					result <- han
-				}
+
+	for {
+		hanLoc = reHan.FindStringIndex(s)
+		if hanLoc == nil {
+			if len(s) == 0 {
+				break
+			}
+		} else if hanLoc[0] == 0 {
+			hans = s[hanLoc[0]:hanLoc[1]]
+			s = s[hanLoc[1]:]
+			for _, han := range cutHan(hans) {
+				result = append(result, han)
+			}
+			continue
+		}
+		nonhanLoc = reSkip.FindStringIndex(s)
+		if nonhanLoc == nil {
+			if len(s) == 0 {
+				break
+			}
+		} else if nonhanLoc[0] == 0 {
+			nonhans := s[nonhanLoc[0]:nonhanLoc[1]]
+			s = s[nonhanLoc[1]:]
+			if nonhans != "" {
+				result = append(result, nonhans)
 				continue
 			}
-			nonhanLoc = reSkip.FindStringIndex(s)
-			if nonhanLoc == nil {
-				if len(s) == 0 {
-					break
-				}
-			} else if nonhanLoc[0] == 0 {
-				nonhans := s[nonhanLoc[0]:nonhanLoc[1]]
-				s = s[nonhanLoc[1]:]
-				if nonhans != "" {
-					result <- nonhans
-					continue
-				}
-			}
-			var loc []int
-			if hanLoc == nil && nonhanLoc == nil {
-				if len(s) > 0 {
-					result <- s
-					break
-				}
-			} else if hanLoc == nil {
-				loc = nonhanLoc
-			} else if nonhanLoc == nil {
-				loc = hanLoc
-			} else if hanLoc[0] < nonhanLoc[0] {
-				loc = hanLoc
-			} else {
-				loc = nonhanLoc
-			}
-			result <- s[:loc[0]]
-			s = s[loc[0]:]
 		}
-		close(result)
-	}()
+		var loc []int
+		if hanLoc == nil && nonhanLoc == nil {
+			if len(s) > 0 {
+				result = append(result, s)
+				break
+			}
+		} else if hanLoc == nil {
+			loc = nonhanLoc
+		} else if nonhanLoc == nil {
+			loc = hanLoc
+		} else if hanLoc[0] < nonhanLoc[0] {
+			loc = hanLoc
+		} else {
+			loc = nonhanLoc
+		}
+		result = append(result, s[:loc[0]])
+		s = s[loc[0]:]
+	}
+
 	return result
 }

finalseg/finalseg_test.go

@@ -5,14 +5,6 @@ import (
 	"testing"
 )
 
-func chanToArray(ch chan string) []string {
-	var result []string
-	for word := range ch {
-		result = append(result, word)
-	}
-	return result
-}
-
 func TestViterbi(t *testing.T) {
 	obs := "我们是程序员"
 	states := []byte{'B', 'M', 'E', 'S'}
@@ -29,7 +21,7 @@ func TestViterbi(t *testing.T) {
 
 func TestCutHan(t *testing.T) {
 	obs := "我们是程序员"
-	result := chanToArray(cutHan(obs))
+	result := cutHan(obs)
 	if len(result) != 3 {
 		t.Fatal(result)
 	}
@@ -46,7 +38,7 @@ func TestCutHan(t *testing.T) {
 
 func TestCut(t *testing.T) {
 	sentence := "我们是程序员"
-	result := chanToArray(Cut(sentence))
+	result := Cut(sentence)
 	if len(result) != 3 {
 		t.Fatal(len(result))
 	}
@@ -59,11 +51,11 @@ func TestCut(t *testing.T) {
 	if result[2] != "程序员" {
 		t.Fatal(result[2])
 	}
-	result2 := chanToArray(Cut("I'm a programmer!"))
+	result2 := Cut("I'm a programmer!")
 	if len(result2) != 8 {
 		t.Fatal(result2)
 	}
-	result3 := chanToArray(Cut("程序员average年龄28.6岁。"))
+	result3 := Cut("程序员average年龄28.6岁。")
 	if len(result3) != 6 {
 		t.Fatal(result3)
 	}