added test for load cutomized stop words file

analyse/analyse_test.go

@@ -240,6 +240,19 @@ var (
 		TfIdf{Word: "怯懦", Freq: 0.271099},
 		TfIdf{Word: "隨便", Freq: 0.168377},
 	}
+
+	LyciWeight2 = []TfIdf{
+		TfIdf{Word: "所謂", Freq: 1.215739},
+		TfIdf{Word: "一般", Freq: 0.731179},
+		TfIdf{Word: "雖然", Freq: 0.405246},
+		TfIdf{Word: "退縮", Freq: 0.405246},
+		TfIdf{Word: "肌迫", Freq: 0.405246},
+		TfIdf{Word: "矯作", Freq: 0.405246},
+		TfIdf{Word: "怯懦", Freq: 0.326238},
+		TfIdf{Word: "逼不得已", Freq: 0.202623},
+		TfIdf{Word: "右銘", Freq: 0.202623},
+		TfIdf{Word: "寬闊", Freq: 0.202623},
+	}
 )
 
 func TestExtractTags(t *testing.T) {
@@ -270,3 +283,16 @@ func TestExtratTagsWithWeight(t *testing.T) {
 		}
 	}
 }
+
+func TestExtractTagsWithStopWordsFile(t *testing.T) {
+	jiebago.SetDictionary("../dict.txt")
+	SetIdf("idf.txt")
+	SetStopWords("stop_words.txt")
+	result := ExtractTags(Lyric, 7)
+	for index, tag := range result {
+		if LyciWeight2[index].Word != tag.Word ||
+			math.Abs(LyciWeight2[index].Freq-tag.Freq) > 1e-6 {
+			t.Errorf("%v != %v", tag, LyciWeight2[index])
+		}
+	}
+}

analyse/idf.go

@@ -29,7 +29,7 @@ func NewIDFLoader() *IDFLoader {
 	return loader
 }
 
-func (loader *IDFLoader) NewPath(idfFilePath string) error {
+func (loader *IDFLoader) newPath(idfFilePath string) error {
 	if loader.Path == idfFilePath {
 		return nil
 	}
@@ -67,5 +67,5 @@ func SetIdf(idfFilePath string) error {
 		}
 		idfFilePath = filepath.Clean(filepath.Join(pwd, idfFilePath))
 	}
-	return idfLoader.NewPath(idfFilePath)
+	return idfLoader.newPath(idfFilePath)
 }

analyse/stop_words.txt

@@ -0,0 +1,51 @@
+the
+of
+is
+and
+to
+in
+that
+we
+for
+an
+are
+by
+be
+as
+on
+with
+can
+if
+from
+which
+you
+it
+this
+then
+at
+have
+all
+not
+one
+has
+or
+that
+的
+了
+和
+是
+就
+都
+而
+及
+與
+著
+或
+一個
+沒有
+我們
+你們
+妳們
+他們
+她們
+是否