详解Java递归实现树形结构的两种方式

0、引言

在开发的过程中，很多业务场景需要一个树形结构的结果集进行前端展示，也可以理解为是一个无限父子结构，常见的有报表指标结构、菜单结构等。Java中递归实现树形结构的两种常见方式如下：

Java7及以下纯Java递归实现
Java8及以上借助lamda表达式实现

1、数据准备

Java实体类NodePO对应数据库表

package com.wbs.pojo;

import lombok.Data;
import lombok.NoArgsConstructor;

import java.util.List;

@Data
@NoArgsConstructor
public class NodePO {

    /**
     * 当前节点id
     */
    private String id;

    /**
     * 当前节点名称
     */
    private String name;

    /**
     * 父级节点id
     */
    private String parentId;

    /**
     * 当前节点序号
     */
    private String orderNo;

    /**
     * 子集节点
     */
    private List children;

    /**
     * 构造函数
     * @param id
     * @param name
     * @param parentId
     * @param orderNo
     */
    public NodePO(String id,String name,String parentId,String orderNo){
        this.id = id;
        this.name = name;
        this.parentId = parentId;
        this.orderNo = orderNo;
    }
}

自己造一些数据模拟从数据库中查询出来的数据：

static final List nodePOs = Arrays.asList(
            new NodePO("1","一级节点1",null,"_0001"),
            new NodePO("2","二级节点1.1","1","_0002"),
            new NodePO("3","二级节点1.2","1","_0003"),

            new NodePO("4","一级节点2",null,"_0004"),
            new NodePO("5","二级节点2.1","4","_0005"),
            new NodePO("6","二级节点2.2","4","_0006"),
            new NodePO("7","三级节点2.2.1","6","_0007"),

            new NodePO("8","一级节点3",null,"_0008"),
            new NodePO("9","二级节点3.1","8","_0009"),
            new NodePO("10","三级节点3.1.1","9","_0010"),
            new NodePO("11","四级节点3.1.1.1","10","_0011"),
            new NodePO("12","五级节点3.1.1.1.1","11","_0012")
    );

2、类型转化

从开发的过程中发现直接操作实体类集合，专门指定某一个实体类封装的方法是不具有普适性的，所以将实体类集合统一转化为Map集合，操作方便，具有一定的普适性：

List> mapList = BeanMapUtils.listBeanToListMap(jsonObject);

BeanMapUtils自己简单封装一个工具类（不惧普适性勿喷）：

package com.wbs.util;

import com.alibaba.fastjson.JSON;
import com.alibaba.fastjson.JSONObject;
import com.google.common.collect.Lists;
import com.google.common.collect.Maps;
import lombok.SneakyThrows;
import org.springframework.cglib.beans.BeanMap;

import java.util.*;
import java.util.function.Function;
import java.util.stream.Collectors;

/**
 * @author 一宿君
 * @version Id: BeanMapUtils.java, v 0.1 Administrator Exp $$
 * @date 2022-10-13 14:24:20
 * @desc java实体类和map相互转换工具类
 */
public class BeanMapUtils {

    /**
     * 将实体类对象属性转化为map对象
     * @param t
     * @param 
     * @return
     */
    public static  Map beanToMap(T t) {
        Map map = new HashMap<>();
        if (t != null) {
            if (t instanceof JSONObject){
                return (JSONObject)t;
            }
            BeanMap beanMap = BeanMap.create(t);
            for (Object key : beanMap.keySet()) {
                map.put(key.toString(), beanMap.get(key));
            }
        }
        return map;
    }


    /**
     * 将map对象中转化为实体类对象
     * @param map
     * @param clazz
     * @param 
     * @return
     * @throws Exception
     */
    public static  T mapToBean(Map map,Class clazz) throws Exception {
        T bean = clazz.newInstance();
        if (bean instanceof JSONObject){
            JSONObject jsonObject = (JSONObject)bean;
            Set> entries = map.entrySet();
            for (Map.Entry entry : entries) {
                jsonObject.put(entry.getKey(),entry.getValue());
            }
            return (T)jsonObject;
        }
        BeanMap beanMap = BeanMap.create(bean);
        beanMap.putAll(map);
        return bean;
    }

    /**
     * 通过lambda表达式将List转化为List>
     * @param objList
     * @param 
     * @return
     */
    public static  List> listBeanToListMap(List objList) {
        return objList.stream().map(new Function>() {
            @Override
            public Map apply(T t) {
                Map map = Maps.newHashMap();
                if (t instanceof JSONObject){
                    return (JSONObject)t;
                }
                BeanMap beanMap = BeanMap.create(t);
                for (Object key : beanMap.keySet()) {
                    map.put(key.toString(), beanMap.get(key));
                }
                return map;
            }
        }).collect(Collectors.toList());
    }

    /**
     * 通过lambda表达式将List>转化为List
     * @param mapList
     * @param 
     * @return
     */
    public static  List listMapToListBean(List> mapList,Class clazz) {
        return mapList.stream().map(new Function,T>() {
            @SneakyThrows
            @Override
            public T apply(Map map) {
                T t = clazz.newInstance();
                if (t instanceof JSONObject){
                    return (T)map;
                }
                BeanMap beanMap = BeanMap.create(t);
                beanMap.putAll(map);
                return t;
            }
        }).collect(Collectors.toList());
    }
}

其中org.springframework.cglib.beans.BeanMap;是org.springframework:spring-core依赖下的工具包，spring-core核心依赖只要导入spring-boot-starter依赖即可


    org.springframework.boot
    spring-boot-starter
    2.2.0.RELEASE

3、递归实现方法

3.1、Java7及以下纯Java递归实现

既然是Java7及以下实现方式，那排序也用最原始的冒泡排序：

/**
     * 冒泡排序,小的在前，大的在后
     * @param list
     * @return
     */
    public static List> sortJava7Map(List> list){
        if(CollectionUtils.isEmpty(list)){
            return Lists.newArrayList();
        }
        boolean flag;
        int size = list.size();
        for (int i = 0; i < size - 1; i++) {
            flag = false;
            for (int j = 1; j < size - i; j++) {
                Map frontMap = list.get(j - 1);
                Map afterMap = list.get(j);
                if (String.valueOf(frontMap.get("orderNo")).compareTo(String.valueOf(afterMap.get("orderNo"))) > 0){
                    list.set(j - 1,afterMap);
                    list.set(j,frontMap);
                    flag = true;
                }
            }
            //如果没有发生位置互换，则退出循环
            if (!flag){
                break;
            }
        }
        return list;
    }

给定一个节点，获取它的所有子节点：

/**
     * Java7及以下版本获取子节点的方式
     * @param parentNode
     * @param allList
     * @return
     */
    public static List> getJava7Children(Map parentNode,List> allList){

        //存放当前节点的直系子节点
        List> curNodeChildrenList = Lists.newArrayList();

        //存放直系子节点以外的节点
        List> otherNodeList = Lists.newArrayList();

        Object pId = parentNode.get("id");
        for (Map map : allList) {
            Object curPId = map.get("parentId");
            if (ObjectUtils.isNotEmpty(curPId) && Objects.equals(pId,curPId)){
                curNodeChildrenList.add(map);
            }else {
                otherNodeList.add(map);
            }
        }
        if (curNodeChildrenList.isEmpty()){
            return curNodeChildrenList;
        }
        //每一层级都进行排序
        curNodeChildrenList = sortJava7Map(curNodeChildrenList);

        //迭代直系子节点再获取子节点
        for (Map map : curNodeChildrenList) {
            map.put("children",getJava7Children(map,otherNodeList));
        }
        return curNodeChildrenList;
    }

给出一个结果集，构建树形结果集：

/**
     * 使用Java7的方式获取树形结构
     * @param allList
     * @return
     */
    public static List> getJava7ResultTree(List> allList){
        //存放所有的一级节点
        List> oneLevelNodeList = Lists.newArrayList();

        for (Map map : allList) {
            if (ObjectUtils.isEmpty(map.get("parentId"))){
                map.put("children",getJava7Children(map,allList));
                oneLevelNodeList.add(map);
            }
        }
        return sortJava8Map(oneLevelNodeList);
    }

获取树形结构：

//转化为Map集合
List> mapList = BeanMapUtils.listBeanToListMap(nodePOs);
//获取树形结构
List> java7ResultTree = getJava7ResultTree(mapList);
//打印输出
System.out.println(JSON.toJSONString(java7ResultTree));

打印结果：

[{"orderNo":"_0001","children":[{"orderNo":"_0002","children":[],"name":"二级节点1.1","id":"2","parentId":"1"},{"orderNo":"_0003","children":[],"name":"二级节点1.2","id":"3","parentId":"1"}],"name":"一级节点1","id":"1"},{"orderNo":"_0004","children":[{"orderNo":"_0005","children":[],"name":"二级节点2.1","id":"5","parentId":"4"},{"orderNo":"_0006","children":[{"orderNo":"_0007","children":[],"name":"三级节点2.2.1","id":"7","parentId":"6"}],"name":"二级节点2.2","id":"6","parentId":"4"}],"name":"一级节点2","id":"4"},{"orderNo":"_0008","children":[{"orderNo":"_0009","children":[{"orderNo":"_0010","children":[{"orderNo":"_0011","children":[{"orderNo":"_0012","children":[],"name":"五级节点3.1.1.1.1","id":"12","parentId":"11"}],"name":"四级节点3.1.1.1","id":"11","parentId":"10"}],"name":"三级节点3.1.1","id":"10","parentId":"9"}],"name":"二级节点3.1","id":"9","parentId":"8"}],"name":"一级节点3","id":"8"}]

树形结构搞定！

3.2、Java8及以上借助lamda表达式实现

Java7的方式虽然实现了树形结构，但是有一定的缺点，比如：代码量比较大，逻辑相对较复杂，那Java8是如何简化，如下所示：

既然Java8有lamda表达式，那代码我们能省就省，先看排序，一行代码搞定：

/**
     * 根据orderNo排序树形结构的每一个层级
     * @param list
     * @return
     */
    public static List> sortJava8Map(List> list){
        if(CollectionUtils.isEmpty(list)){
            return Lists.newArrayList();
        }
        //关键之处，一行代码搞定
        list.sort(Comparator.comparing(m -> String.valueOf(m.get("orderNo"))));
        return list;
    }

给定一个节点，获取它的所有子节点：

释义：
filter： 过滤，相当于for循环，再if条件判断。
peek： 给定一个节点，往它的children塞子节点。

/**
     * 根据父级节点获取所有的子集节点
     * @param parentNode
     * @param allList
     * @return
     */
    public static List> getJava8Children(Map parentNode, List> allList){
        return allList.stream()
                .filter(curNode -> ObjectUtils.isNotEmpty(curNode.get("parentId")) && Objects.equals(curNode.get("parentId"),parentNode.get("id")))
                .peek(m -> m.put("children", getJava8Children(m,allList))).collect(Collectors.toList());
    }