[Feature] add for new

src/main/java/com/github/houbb/sensitive/word/support/map/SensitiveWordMap.java

@@ -270,7 +270,7 @@ public class SensitiveWordMap implements IWordMap {
             return target;
         }
         // 用于结果构建
-        StringBuilder resultBuilder = new StringBuilder();
+        StringBuilder resultBuilder = new StringBuilder(target.length());
 
         for (int i = 0; i < target.length(); i++) {
             char currentChar = target.charAt(i);

src/test/java/com/github/houbb/sensitive/word/data/DataUtil.java

@@ -0,0 +1,27 @@
+package com.github.houbb.sensitive.word.data;
+
+import com.github.houbb.heaven.util.io.FileUtil;
+import com.github.houbb.heaven.util.util.CollectionUtil;
+
+import java.util.List;
+
+/**
+ * @author binbin.hou
+ * @since 0.0.3
+ */
+public class DataUtil {
+
+    /**
+     * 获取对应文件的独一无二内容
+     * @param name 名称
+     * @return 结果
+     * @since 0.0.1
+     */
+    public static List<String> distinctLines(final String name) {
+        final String dir = "D:\\github\\sensitive-word\\src\\main\\resources\\";
+        final String path = dir + name;
+        List<String> lines = FileUtil.readAllLines(path);
+        return CollectionUtil.distinct(lines);
+    }
+
+}

src/test/java/com/github/houbb/sensitive/word/data/StopWordTest.java

@@ -0,0 +1,44 @@
+package com.github.houbb.sensitive.word.data;
+
+import com.github.houbb.heaven.support.condition.ICondition;
+import com.github.houbb.heaven.support.filter.IFilter;
+import com.github.houbb.heaven.util.io.FileUtil;
+import com.github.houbb.heaven.util.lang.StringUtil;
+import com.github.houbb.heaven.util.util.CharsetUtil;
+import com.github.houbb.heaven.util.util.CollectionUtil;
+import org.junit.Ignore;
+import org.junit.Test;
+
+import java.util.Collections;
+import java.util.List;
+
+/**
+ * 停止词数据初始化
+ * @author binbin.hou
+ * @since 0.0.3
+ */
+public class StopWordTest {
+
+    /**
+     * 中文测试
+     * @since 0.0.3
+     */
+    @Test
+    @Ignore
+    public void zhTest() {
+        final String sourceFile = "stopword.txt";
+        final String targetFile = "D:\\github\\sensitive-word\\src\\main\\resources\\stopword_zh.txt";
+
+        List<String> allLines = DataUtil.distinctLines(sourceFile);
+
+        List<String> zhLines = CollectionUtil.conditionList(allLines, new ICondition<String>() {
+            @Override
+            public boolean condition(String s) {
+                return CharsetUtil.isAllChinese(s);
+            }
+        });
+
+        FileUtil.write(targetFile, zhLines);
+    }
+
+}