java基于Map实现DFA算法

最新推荐文章于 2022-03-02 16:50:42 发布

毛豆有毛没豆

最新推荐文章于 2022-03-02 16:50:42 发布

阅读量351

点赞数

分类专栏：算法+数据结构文章标签：算法 dfa

本文链接：https://blog.youkuaiyun.com/qq_22049773/article/details/119111909

版权

算法+数据结构专栏收录该内容

1 篇文章

订阅专栏

package com.yangkaile.generator;

import lombok.extern.slf4j.Slf4j;
import org.junit.jupiter.api.Test;

import java.util.*;

/**
 * @description: DFA算法案例
 * @class Name: ApplicationTest
 * @author: wangdong
 * @Date: 2021/7/26 15:56
 */
@Slf4j
public class ApplicationTest {

    @Test
    public void test1(){
        Set<String> keyWordSet=new HashSet<>();
        keyWordSet.add("五连鞭");
        keyWordSet.add("接化发");
        keyWordSet.add("一鞭");
        keyWordSet.add("二鞭");
        keyWordSet.add("三鞭");
        keyWordSet.add("四鞭");
        keyWordSet.add("五鞭");
        keyWordSet.add("混元形意太极掌门人");
        Map dfa_map=addSensitiveWordToHashMap(keyWordSet);
        Set<String> result=getTriggerOverWord("一鞭后直接五鞭,",dfa_map);
        System.out.println(result);
    }
    /**
     * 构建成DFA算法模型
     * @param keyWordSet
     */
    public Map addSensitiveWordToHashMap (Set<String> keyWordSet) {
        Map sensitiveWordMap = new HashMap(keyWordSet.size());     //初始化关键词容器，减少扩容操作
        String key = null;
        Map nowMap = null;
        Map<String, String> newWorMap = null;
        //迭代keyWordSet
        Iterator<String> iterator = keyWordSet.iterator();
        while (iterator.hasNext()) {
            key = iterator.next();    //关键字
            nowMap = sensitiveWordMap;
            for (int i = 0; i < key.length(); i++) {
                char keyChar = key.charAt(i);       //转换成char型
                Object wordMap = nowMap.get(keyChar);       //获取

                if (wordMap != null) {        //如果存在该key，直接赋值
                    nowMap = (Map) wordMap;
                } else {     //不存在则，则构建一个map，同时将isEnd设置为0，因为他不是最后一个
                    newWorMap = new HashMap<String, String>();
                    newWorMap.put("isEnd", "0");     //不是最后一个
                    nowMap.put(keyChar, newWorMap);
                    nowMap = newWorMap;
                }

                if (i == key.length() - 1) {
                    nowMap.put("isEnd", "1");    //最后一个
                }
            }
        }
        return sensitiveWordMap;
    }

    /**
     * 判断从start后的词是否是关键词 -->基于DFA模型判断
     * @param content
     * @param start
     * @return
     */
    public int getOverWordLength(Map keyWordSet,String content,int start,int matchType){
        boolean flag = false;
        int length = 0;
        char word = 0;
        Map nowMap = keyWordSet;
        for (int i = start ; i < content.length() ; i++ ){
            word = content.charAt(i);
            nowMap = (Map) nowMap.get(word);
            if(nowMap == null){
                break;
            }else {
                length++;
                if("1".equals(nowMap.get("isEnd"))){
                    flag = true;
                    if(1 == matchType){
                        break;
                    }
                }
            }
        }
        if (length < 1 || !flag){
            length = 0;
        }
        return length;
    }

    /**
     * 基于DFA模型匹配关键词
     * @param content 待匹配文本
     * @param keyWordSet  关键词
     * @return 返回匹配到的关键词
     */
    public Set<String> getTriggerOverWord(String content,Map keyWordSet){

        Set<String> words = new HashSet<>();

        for (int i = 0 ; i < content.length() ; i++ ){
            int length = getOverWordLength(keyWordSet, content, i,2);
            if(length > 0){
                words.add(content.substring(i,i+length));
                i = i + length - 1;
            }
        }

        return words;
    }
}