mirror of https://github.com/ZhongFuCheng3y/austin
# Conflicts: # austin-handler/src/main/java/com/java3y/austin/handler/pending/TaskPendingHolder.javapull/69/head
commit
8bbc026d11
@ -0,0 +1,186 @@
|
||||
package com.java3y.austin.handler.action;
|
||||
|
||||
import com.java3y.austin.common.domain.TaskInfo;
|
||||
import com.java3y.austin.common.dto.model.*;
|
||||
import com.java3y.austin.common.pipeline.BusinessProcess;
|
||||
import com.java3y.austin.common.pipeline.ProcessContext;
|
||||
import com.java3y.austin.handler.config.SensitiveWordsConfig;
|
||||
import org.springframework.beans.factory.annotation.Autowired;
|
||||
import org.springframework.data.redis.core.RedisTemplate;
|
||||
import org.springframework.stereotype.Service;
|
||||
import org.springframework.util.ObjectUtils;
|
||||
|
||||
import java.util.*;
|
||||
|
||||
/**
|
||||
* 敏感词过滤
|
||||
*
|
||||
* @author xiaoxiamao
|
||||
* @date 2024/08/17
|
||||
*/
|
||||
@Service
|
||||
public class SensWordsAction implements BusinessProcess<TaskInfo> {
|
||||
|
||||
|
||||
@Autowired
|
||||
private RedisTemplate<String, String> redisTemplate;
|
||||
|
||||
/**
|
||||
* 过滤逻辑
|
||||
*
|
||||
* @param context
|
||||
*
|
||||
* @see com.java3y.austin.common.enums.ChannelType
|
||||
*/
|
||||
@Override
|
||||
public void process(ProcessContext<TaskInfo> context) {
|
||||
// 获取敏感词典
|
||||
Set<String> sensDict = Optional.ofNullable(redisTemplate.opsForSet().members(SensitiveWordsConfig.SENS_WORDS_DICT))
|
||||
.orElse(Collections.emptySet());
|
||||
// 如果敏感词典为空,不过滤
|
||||
if (ObjectUtils.isEmpty(sensDict)) {
|
||||
return;
|
||||
}
|
||||
switch (context.getProcessModel().getMsgType()) {
|
||||
// IM
|
||||
case 10:
|
||||
// 无文本内容,暂不做过滤处理
|
||||
break;
|
||||
// PUSH
|
||||
case 20:
|
||||
PushContentModel pushContentModel =
|
||||
(PushContentModel) context.getProcessModel().getContentModel();
|
||||
pushContentModel.setContent(filter(pushContentModel.getContent(), sensDict));
|
||||
break;
|
||||
// SMS
|
||||
case 30:
|
||||
SmsContentModel smsContentModel =
|
||||
(SmsContentModel) context.getProcessModel().getContentModel();
|
||||
smsContentModel.setContent(filter(smsContentModel.getContent(), sensDict));
|
||||
break;
|
||||
// EMAIL
|
||||
case 40:
|
||||
EmailContentModel emailContentModel =
|
||||
(EmailContentModel) context.getProcessModel().getContentModel();
|
||||
emailContentModel.setContent(filter(emailContentModel.getContent(), sensDict));
|
||||
break;
|
||||
// OFFICIAL_ACCOUNT
|
||||
case 50:
|
||||
// 无文本内容,暂不做过滤处理
|
||||
break;
|
||||
// MINI_PROGRAM
|
||||
case 60:
|
||||
// 无文本内容,暂不做过滤处理
|
||||
break;
|
||||
// ENTERPRISE_WE_CHAT
|
||||
case 70:
|
||||
EnterpriseWeChatContentModel enterpriseWeChatContentModel =
|
||||
(EnterpriseWeChatContentModel) context.getProcessModel().getContentModel();
|
||||
enterpriseWeChatContentModel.setContent(filter(enterpriseWeChatContentModel.getContent(), sensDict));
|
||||
break;
|
||||
// DING_DING_ROBOT
|
||||
case 80:
|
||||
DingDingRobotContentModel dingDingRobotContentModel =
|
||||
(DingDingRobotContentModel) context.getProcessModel().getContentModel();
|
||||
dingDingRobotContentModel.setContent(filter(dingDingRobotContentModel.getContent(), sensDict));
|
||||
break;
|
||||
// DING_DING_WORK_NOTICE
|
||||
case 90:
|
||||
DingDingWorkContentModel dingDingWorkContentModel =
|
||||
(DingDingWorkContentModel) context.getProcessModel().getContentModel();
|
||||
dingDingWorkContentModel.setContent(filter(dingDingWorkContentModel.getContent(), sensDict));
|
||||
break;
|
||||
// ENTERPRISE_WE_CHAT_ROBOT
|
||||
case 100:
|
||||
EnterpriseWeChatRobotContentModel enterpriseWeChatRobotContentModel =
|
||||
(EnterpriseWeChatRobotContentModel) context.getProcessModel().getContentModel();
|
||||
enterpriseWeChatRobotContentModel.setContent(filter(enterpriseWeChatRobotContentModel.getContent(), sensDict));
|
||||
break;
|
||||
// FEI_SHU_ROBOT
|
||||
case 110:
|
||||
FeiShuRobotContentModel feiShuRobotContentModel =
|
||||
(FeiShuRobotContentModel) context.getProcessModel().getContentModel();
|
||||
feiShuRobotContentModel.setContent(filter(feiShuRobotContentModel.getContent(), sensDict));
|
||||
break;
|
||||
// ALIPAY_MINI_PROGRAM
|
||||
case 120:
|
||||
// 无文本内容,暂不做过滤处理
|
||||
break;
|
||||
default:
|
||||
break;
|
||||
}
|
||||
}
|
||||
|
||||
/**
|
||||
* 敏感词替换成对应长度'*'
|
||||
*
|
||||
* @param content
|
||||
* @param sensDict
|
||||
* @return
|
||||
*/
|
||||
private String filter(String content, Set<String> sensDict) {
|
||||
if (ObjectUtils.isEmpty(content) || ObjectUtils.isEmpty(sensDict)) {
|
||||
return content;
|
||||
}
|
||||
// 构建字典树
|
||||
TrieNode root = buildTrie(sensDict);
|
||||
StringBuilder result = new StringBuilder();
|
||||
int n = content.length();
|
||||
int i = 0;
|
||||
|
||||
while (i < n) {
|
||||
TrieNode node = root;
|
||||
int j = i;
|
||||
int lastMatchEnd = -1;
|
||||
|
||||
while (j < n && node != null) {
|
||||
node = node.children.get(content.charAt(j));
|
||||
if (node != null && node.isEnd) {
|
||||
lastMatchEnd = j;
|
||||
}
|
||||
j++;
|
||||
}
|
||||
|
||||
if (lastMatchEnd != -1) {
|
||||
// 找到敏感词,用'*'替换
|
||||
for (int k = i; k <= lastMatchEnd; k++) {
|
||||
result.append('*');
|
||||
}
|
||||
i = lastMatchEnd + 1;
|
||||
} else {
|
||||
result.append(content.charAt(i));
|
||||
i++;
|
||||
}
|
||||
}
|
||||
|
||||
return result.toString();
|
||||
}
|
||||
|
||||
/**
|
||||
* 构建字典树
|
||||
*
|
||||
* @param sensDict
|
||||
* @return
|
||||
*/
|
||||
private TrieNode buildTrie(Set<String> sensDict) {
|
||||
TrieNode root = new TrieNode();
|
||||
for (String word : sensDict) {
|
||||
TrieNode node = root;
|
||||
for (char c : word.toCharArray()) {
|
||||
node = node.children.computeIfAbsent(c, k -> new TrieNode());
|
||||
}
|
||||
node.isEnd = true;
|
||||
}
|
||||
return root;
|
||||
}
|
||||
|
||||
/**
|
||||
* 树节点
|
||||
*/
|
||||
private static class TrieNode {
|
||||
Map<Character, TrieNode> children = new HashMap<>();
|
||||
// 是否为叶子节点
|
||||
boolean isEnd = false;
|
||||
}
|
||||
|
||||
}
|
@ -0,0 +1,7 @@
|
||||
机密信息
|
||||
政治敏感
|
||||
违法犯罪
|
||||
黑客攻击
|
||||
网络谩骂
|
||||
admin
|
||||
password
|
Loading…
Reference in new issue