'commit'

pom.xml

@@ -379,6 +379,11 @@
             <artifactId>jakarta.json</artifactId>
             <version>2.0.1</version>
         </dependency>
+        <dependency>
+            <groupId>com.hankcs</groupId>
+            <artifactId>hanlp</artifactId>
+            <version>portable-1.8.4</version>
+        </dependency>
     </dependencies>
     <build>
         <plugins>

src/main/java/com/YunXiao/Study.java

@@ -1,7 +1,5 @@
 package com.YunXiao;
 
-import cn.hutool.core.date.DateTime;
-import cn.hutool.core.io.file.FileWriter;
 import com.YunXiao.Model.BaseModel;
 import com.YunXiao.Util.SyncUtil;
 import com.alibaba.fastjson.JSONArray;
@@ -9,57 +7,69 @@ import com.alibaba.fastjson.JSONObject;
 
 import com.dsideal.QingLong.Util.RedisKit;
 import com.dsideal.QingLong.Util.SimilarayUtil;
+import com.hankcs.hanlp.seg.common.Term;
+import com.hankcs.hanlp.tokenizer.StandardTokenizer;
 import com.jfinal.plugin.activerecord.Db;
 import com.jfinal.plugin.activerecord.Record;
 
-import java.io.File;
+import java.util.*;
-import java.nio.charset.StandardCharsets;
-import java.util.ArrayList;
-import java.util.Collections;
-import java.util.List;
 
 public class Study {
     //Model层的实例，在此层，不允许调用Dao层，只能对接Model层
     public static BaseModel bm = new BaseModel();
 
-
+    public static String KEY = "TY_BUREAU";
-    static long start;
-    static boolean logFirst = false;
-    static String LOG_FILE = null;
-    static String directoryPath = null;
 
     /**
-     * 功能：记录日志，每个任务一个单独文件
+     * 功能:利用中文分词技术，将两个字符串都进行分词，通过对比完全命中的数量来评估是不是两个单位名称是同一个单位
      *
-     * @param msg
+     * @param s1
+     * @param s2
+     * @return
      */
-    public static void log(String msg) {
+    public static boolean check(String s1, String s2) {
-        if (!logFirst) {
+        //静态变量
-            // 记录到日志文件
+        String[] area = {
-            LOG_FILE = SyncUtil.WEBPATH + "/Logs/{start_time}.log";
+                "南关区", "宽城区", "朝阳区", "二道区", "绿园区", "双阳区", "九台市",
-            // 创建 File 对象
+                "农安县", "经开区", "德惠市", "高新区", "净月区", "汽开区", "榆树市", "公主岭市", "长春市"
-            File file = new File(LOG_FILE);
+        };
-            // 获取目录部分
+        List<String> areaList = Arrays.stream(area).toList();
-            directoryPath = file.getParent();
+
-            // 判断目录是否存在，如果不存在则创建
+        List<Term> termList1 = StandardTokenizer.segment(s1);
-            File directory = new File(directoryPath);
+        Map<String, Integer> map = new HashMap<>();
-            if (!directory.exists()) {
+        for (Term term : termList1) {
-                if (directory.mkdirs()) {
+            String key = SimilarayUtil.Convert(term.word);
-                    System.out.println(DateTime.now() + "       目录已创建：" + directoryPath);
+
-                } else {
+            for (int i = 0; i < areaList.size(); i++) {
-                    System.out.println("创建日志目录失败，请检查！");
+                if (areaList.get(i).startsWith(key)) {
-                    System.exit(0);
+                    key = areaList.get(i);
+                    break;
                 }
             }
-            LOG_FILE = LOG_FILE.replace("{start_time}", String.valueOf(start));
+            int x = 0;
-            logFirst = true;
+            if (map.containsKey(key)) x = map.get(key);
+            map.put(key, x + 1);
         }
-        System.out.println(DateTime.now() + "     " + msg);
-        FileWriter writer = new FileWriter(LOG_FILE, StandardCharsets.UTF_8);
-        writer.append(DateTime.now() + "  " + msg + "\n");
-    }
 
-    public static String KEY = "TY_BUREAU";
+        List<Term> termList2 = StandardTokenizer.segment(s2);
+        for (Term term : termList2) {
+            String key = SimilarayUtil.Convert(term.word);
+            for (int i = 0; i < areaList.size(); i++) {
+                if (areaList.get(i).startsWith(key)) {
+                    key = areaList.get(i);
+                    break;
+                }
+            }
+            int x = 0;
+            if (map.containsKey(key)) x = map.get(key);
+            map.put(key, x + 1);
+        }
+        // 使用 for-each 循环遍历 Map
+        for (Map.Entry<String, Integer> entry : map.entrySet()) {
+            if (entry.getValue() == 1) return false;
+        }
+        return true;
+    }
 
     public static void main(String[] args) throws Exception {
         //初始化
@@ -108,9 +118,12 @@ public class Study {
                 sql = "update t_base_organization set third_party_id=? where org_id=?";
                 Db.update(sql, ty_org_id, org_id);
             } else {
-                System.out.println("QingLong系统 org_id:" + org_id + ",org_name：" + org_name + "\n天喻平台  ty_org_id:" + ty_org_id + ",ty_org_name：" + ty_org_name );
+                boolean success = check(org_name, ty_org_name);
-                System.out.println("相似度：" + similary);
+                if (success) {
-                System.out.println();
+                    System.out.println("QingLong系统 org_id:" + org_id + ",org_name：" + org_name + "\n天喻平台  ty_org_id:" + ty_org_id + ",ty_org_name：" + ty_org_name);
+                } else {
+                    System.out.println("QingLong系统 org_id:" + org_id + ",org_name：" + org_name + " 无法推荐出合适的单位名，请人工检索！");
+                }
             }
         }
         // 关闭连接

src/main/java/com/dsideal/QingLong/Util/SimilarayUtil.java

@@ -35,7 +35,7 @@ public class SimilarayUtil {
      * @param chineseNumber
      * @return
      */
-    public static String ChineseToArabic(String chineseNumber) {
+    private static String ChineseToArabic(String chineseNumber) {
         int arabicNumber = 0;
         for (int i = 0; i < chineseNumber.length(); i++) {
             char chineseChar = chineseNumber.charAt(i);
@@ -51,7 +51,7 @@ public class SimilarayUtil {
      * @param arabicNumberStr
      * @return
      */
-    public static String ArabicToChinese(String arabicNumberStr) {
+    private static String ArabicToChinese(String arabicNumberStr) {
         String chineseNumberStr = "";
         for (int i = 0; i < arabicNumberStr.length(); i++) {
             char arabicChar = arabicNumberStr.charAt(i);