Spring Boot 敏感词过滤组件实现：基于DFA算法的高效敏感词检测与替换

Spring Boot 敏感词过滤组件实现

本文将介绍如何在Spring Boot项目中实现一个高效的敏感词过滤组件，包含两种实现方式：基于DFA算法的优化版本和简单直接的实现版本。

组件功能：

1. 从文件加载敏感词库

2. 敏感词检测功能

3. 敏感词替换功能

4. 支持Spring Boot自动装配

实现方案一：DFA算法优化版

这里放入第一个SensitiveWordFilter类的完整代码

package com.example.utils;

import jakarta.annotation.PostConstruct;
import org.springframework.core.io.ClassPathResource;
import org.springframework.stereotype.Component;

import java.io.BufferedReader;
import java.io.IOException;
import java.io.InputStream;
import java.io.InputStreamReader;
import java.util.*;

@Component
public class SensitiveWordFilter {

    private static final String REPLACEMENT = "***";
    private Map<Object, Object> sensitiveWordMap;

    @PostConstruct
    public void init() {
        try {
            // 读取敏感词文件
            Set<String> keyWordSet = new HashSet<>();
            ClassPathResource resource = new ClassPathResource("sensitive-words.txt");
            InputStream inputStream = resource.getInputStream();
            BufferedReader reader = new BufferedReader(new InputStreamReader(inputStream));
            String line;
            while ((line = reader.readLine()) != null) {
                keyWordSet.add(line.trim());
            }
            reader.close();

            // 构建DFA算法模型
            addSensitiveWordToHashMap(keyWordSet);
        } catch (IOException e) {
            e.printStackTrace();
        }
    }

    private void addSensitiveWordToHashMap(Set<String> keyWordSet) {
        sensitiveWordMap = new HashMap(keyWordSet.size());
        String key;
        Map nowMap;
        Map<String, String> newWorMap;

        for (String aKeyWordSet : keyWordSet) {
            key = aKeyWordSet;
            nowMap = sensitiveWordMap;
            for (int i = 0; i < key.length(); i++) {
                char keyChar = key.charAt(i);
                Object wordMap = nowMap.get(keyChar);

                if (wordMap != null) {
                    nowMap = (Map) wordMap;
                } else {
                    newWorMap = new HashMap<>();
                    newWorMap.put("isEnd", "0");
                    nowMap.put(keyChar, newWorMap);
                    nowMap = newWorMap;
                }

                if (i == key.length() - 1) {
                    nowMap.put("isEnd", "1");
                }
            }
        }
    }

    public String filter(String text) {
        if (text == null || text.trim().isEmpty()) {
            return text;
        }

        StringBuilder result = new StringBuilder();
        Map nowMap = sensitiveWordMap;
        int start = 0;
        int point = 0;

        while (point < text.length()) {
            char key = text.charAt(point);
            nowMap = (Map) nowMap.get(key);

            if (nowMap != null) {
                point++;
                if ("1".equals(nowMap.get("isEnd"))) {
                    result.append(REPLACEMENT);
                    start = point;
                    nowMap = sensitiveWordMap;
                }
            } else {
                result.append(text.charAt(start));
                point = ++start;
                nowMap = sensitiveWordMap;
            }
        }

        result.append(text.substring(start));
        return result.toString();
    }

    public boolean containsSensitiveWord(String text) {
        if (text == null || text.trim().isEmpty()) {
            return false;
        }

        Map nowMap = sensitiveWordMap;
        int point = 0;

        while (point < text.length()) {
            char key = text.charAt(point);
            nowMap = (Map) nowMap.get(key);

            if (nowMap != null) {
                point++;
                if ("1".equals(nowMap.get("isEnd"))) {
                    return true;
                }
            } else {
                point = point - (point > 0 ? 1 : 0);
                nowMap = sensitiveWordMap;
                point++;
            }
        }

        return false;
    }
}

DFA算法特点：

• 使用字典树(Trie)结构存储敏感词

• 时间复杂度接近O(n)，n为文本长度

• 适合大规模敏感词库

• 支持最长匹配原则

实现方案二：简单直接版

这里放入第二个SensitiveWordFilter类的完整代码

package com.example.utils;

import org.springframework.core.io.ClassPathResource;
import org.springframework.stereotype.Component;

import jakarta.annotation.PostConstruct;
import java.io.BufferedReader;
import java.io.IOException;
import java.io.InputStream;
import java.io.InputStreamReader;
import java.util.*;

@Component
public class SensitiveWordFilter {

    private static final String REPLACEMENT = "***";
    private static final Set<String> sensitiveWords = new HashSet<>();

    // 初始化敏感词库
    @PostConstruct
    public void init() {
        try {
            ClassPathResource resource = new ClassPathResource("sensitive-words.txt");
            InputStream inputStream = resource.getInputStream();
            BufferedReader reader = new BufferedReader(new InputStreamReader(inputStream));
            String line;
            while ((line = reader.readLine()) != null) {
                sensitiveWords.add(line.trim());
            }
            reader.close();
        } catch (IOException e) {
            e.printStackTrace();
        }
    }

    // 过滤敏感词
    public String filter(String text) {
        if (text == null || text.trim().isEmpty()) {
            return text;
        }
        
        for (String word : sensitiveWords) {
            if (text.contains(word)) {
                text = text.replaceAll(word, REPLACEMENT);
            }
        }
        return text;
    }

    // 检查是否包含敏感词
    public boolean containsSensitiveWord(String text) {
        if (text == null || text.trim().isEmpty()) {
            return false;
        }
        
        for (String word : sensitiveWords) {
            if (text.contains(word)) {
                return true;
            }
        }
        return false;
    }
}

简单版特点：

• 实现简单直观

• 适合小规模敏感词库

• 使用String.contains()方法检测

• 性能在词库较大时较差

使用说明

1. 在resources目录下创建sensitive-words.txt文件

   

2. 每行写入一个敏感词

3. 在Spring Boot项目中注入组件：

//以下引入二选一

// 引入
@Resource
private SensitiveWordFilter sensitiveWordFilter;

// 引入
@Autowired
private SensitiveWordFilter sensitiveWordFilter;

4. 调用方法

// 引入
@Resource
private SensitiveWordFilter sensitiveWordFilter;

/**
     * 新增
     */
    public void add(Comment comment) {

        // 过滤敏感词，哪个字段需要过滤就设置哪个
        // comment.setTitle(sensitiveWordFilter.filter(comment.getTitle()));
        comment.setContent(sensitiveWordFilter.filter(comment.getContent()));

        commentMapper.insert(comment);
    }

依赖配置

<!-- 敏感词处理依赖 -->
<dependency>
    <groupId>jakarta.annotation</groupId>
    <artifactId>jakarta.annotation-api</artifactId>
    <version>2.1.1</version>
</dependency>

前端效果

性能对比

方案	初始化时间	检测时间	内存占用
DFA算法	较高	O(n)	较高
简单版	低	O(n*m)	低

适用场景

• DFA算法版：适合敏感词数量多(1000+)、性能要求高的场景

• 简单版：适合敏感词数量少、快速开发的场景

扩展建议

1. 可以添加动态更新敏感词库功能

2. 支持多种替换策略(如随机替换、首字母保留等)

3. 添加日志记录功能

4. 支持多级敏感词(如政治、色情等分类)

总结

本文提供了两种不同复杂度的敏感词过滤实现，开发者可以根据项目需求选择合适的方案。DFA算法虽然实现复杂但性能优异，简单版则适合快速实现基本功能。