.NET 6 实现敏感词过滤
一、什么是敏感词过滤？
敏感词过滤是一种处理网络内容的技术，可以检测和过滤出网络中的敏感/违禁词汇。它通过给定的关键字或字符串，判断网络内容是否包含某些敏感信息，从而防止违反法律法规的信息流通。
通常，可以使用两种方法来过滤敏感词：
- 黑名单过滤：即定义一个黑名单，将所有敏感词择记录在其中，然后对输入的文本进行对比，如果发现有敏感词，就将其过滤掉。
- 白名单过滤：即定义一个白名单，将所有不敏感的词汇记录在其中，然后对输入的文本进行对比，如果发现有不在白名单中的词汇，就将其过滤掉。
二、ToolGood.Words是什么？
ToolGood.Words是一款高性能非法词(敏感词)检测组件，附带繁体简体互换，支持全角半角互换，获取拼音首字母，获取拼音字母，拼音模糊搜索等功能。
ToolGood.Words的源码网站：ToolGood.Words源码网站
三、在Visual Studio中安装ToolGood.Words
3.1、右键项目解决方案，选择“管理NuGet程序包”，如下图所示：
3.2、切换到“浏览”选项卡，搜索“ToolGood.Words”并安装：
安装完之后最好重新编译生成项目
四、创建“subContentCheck”类
敏感/违禁词汇因特殊内容不便上传，可自行在网站上查找
using Microsoft.AspNetCore.DataProtection.KeyManagement;
using Microsoft.AspNetCore.Http;
using Microsoft.CodeAnalysis.Text;
using Newtonsoft.Json;
using System.Collections;
using System.Text;
using ToolGood.Words;
using static System.Net.Mime.MediaTypeNames;
using IHostingEnvironment = Microsoft.AspNetCore.Hosting.IHostingEnvironment;
namespace WebApplication1 //放在自己项目中时，需要更换为自己的命名空间
{
public class keywords
{
public List<string> IllegalKeywords { get; set; }
}
public class urlwords
{
public List<string> IllegalUrls { get; set; }
}
/// <summary>
/// 提交的内容敏感违禁词检查类
/// </summary>
public class subContentCheck
{
/// <summary>
/// 本地静态文件地址路径
/// </summary>
private IHostingEnvironment _hostingEnv;
/// <summary>
/// 敏感词库
/// </summary>
private string dictionaryPath = "/sensitiveWords/sensitiveWords.txt";
/// <summary>
/// 敏感链接、网站、网址库
/// </summary>
private string urlsPath = "/sensitiveWords/IllegalUrls.txt";
/// <summary>
/// 保存敏感词组
/// </summary>
public string[] Words { get; set; }
/// <summary>
/// 一个参数的构造函数
/// </summary>
/// <param name="hostingEnv">本地静态文件地址路径</param>
public subContentCheck(IHostingEnvironment hostingEnv)
{
_hostingEnv = hostingEnv;
InitDictionary();
}
/// <summary>
/// 初始化内存敏感词库
/// </summary>
public void InitDictionary()
{
Words = new string[] { };
string wordsPath = _hostingEnv.WebRootPath + dictionaryPath;
string urlPath = _hostingEnv.WebRootPath + urlsPath;
//List<keywords> keys = new List<keywords>();
//List<urlwords> urls = new List<urlwords>();
string[] readAllWords = System.IO.File.ReadAllLines(wordsPath, System.Text.Encoding.UTF8);
string[] readAllurl = System.IO.File.ReadAllLines(urlPath, System.Text.Encoding.UTF8);
//由于数组是非动态的，不能进行动态的添加，所有先将它转成list，操作
ArrayList arrayList = new ArrayList(Words.ToList());
if (readAllWords.Length > 0 || readAllurl.Length > 0)
{
if (readAllWords.Length > 1)
{
//keywords key = new keywords();
//key.IllegalKeywords = new List<string>();
foreach (string itemWords in readAllWords)
{
string[] allSplitWords = itemWords.Split('|');
foreach (string itemSplitWords in allSplitWords)
{
if (!string.IsNullOrEmpty(itemSplitWords))
{
arrayList.Add(itemSplitWords);
//string aaa = itemSplitWords;
//key.IllegalKeywords.Add(aaa);
//IllegalKeywords.Add(itemSplitWords);
}
}
}
//keys.Add(key);
}
else
{
if (readAllWords.Length == 1)
{
string[] allSplitWords = readAllWords[0].Split('|');
//keywords key = new keywords();
//key.IllegalKeywords = new List<string>();
foreach (string itemSplitWords in allSplitWords)
{
if (!string.IsNullOrEmpty(itemSplitWords))
{
arrayList.Add(itemSplitWords);
//string aaa = itemSplitWords;
//key.IllegalKeywords.Add(aaa);
//IllegalKeywords.Add(itemSplitWords);
}
}
//keys.Add(key);
}
}
if (readAllurl.Length > 1)
{
//urlwords url = new urlwords();
//url.IllegalUrls = new List<string>();
foreach (string itemUrls in readAllurl)
{
string[] allSplitUrls = itemUrls.Split('|');
foreach (string itemSplitUrls in allSplitUrls)
{
if (!string.IsNullOrEmpty(itemSplitUrls))
{
arrayList.Add(itemSplitUrls);
//string Keyword = itemSplitUrls;
//url.IllegalUrls.Add(Keyword);
//IllegalUrls.Add(itemSplitUrls);
}
}
}
//urls.Add(url);
}
else
{
if (readAllurl.Length == 1)
{
string[] allSplitUrls = readAllurl[0].Split('|');
//urlwords url = new urlwords();
//url.IllegalUrls = new List<string>();
foreach (string itemSplitUrls in allSplitUrls)
{
if (!string.IsNullOrEmpty(itemSplitUrls))
{
arrayList.Add(itemSplitUrls);
//IllegalUrls.Add(itemSplitUrls);
//string Keyword = itemSplitUrls;
//url.IllegalUrls.Add(Keyword);
}
}
//urls.Add(url);
}
}
}
//我们在将list转换成String[]数组
Words = (string[])arrayList.ToArray(typeof(string));
}
/// <summary>
/// 过滤替换敏感词
/// </summary>
/// <param name="sourceText">需要过滤替换的原内容</param>
/// <param name="replaceChar">敏感词替换的字符；默认替换为‘*’</param>
/// <returns>返回状态码；为空则表示传入的内容为空；“0”：设置违禁词时发生错误；“1”：敏感内容替换时发生错误；“2”：需要替换的文本内容为空；其余则返回替换成功的字符串内容</returns>
public string FilterWithChar(string sourceText, char replaceChar = '*')
{
if (!string.IsNullOrEmpty(sourceText))
{
string result = "";
WordsSearch wordsSearch = new WordsSearch();
try
{
wordsSearch.SetKeywords(Words);
}
catch (Exception ex)
{
result = "0";
return result;
}
try
{
result = wordsSearch.Replace(sourceText, replaceChar);
return result;
}
catch (Exception ex)
{
return result = "1";
}
}
else
{
return "2";
}
}
/// <summary>
/// 查找原内容中知否包含敏感/违禁词
/// </summary>
/// <param name="sourceText">需要判断的原内容</param>
/// <returns>返回状态码；为空则表示传入的内容为空；“0”：设置违禁词时发生错误；“1”：敏感内容查询时发生错误；“2”：需要替换的文本内容为空；“3”：原内容中包含敏感/违禁词汇；“4”：原内容中不包含敏感/违禁词汇</returns>
public string FindSensitiveKey(string sourceText)
{
string result = "";
if (!string.IsNullOrEmpty(sourceText))
{
WordsSearch wordsSearch = new WordsSearch();
try
{
wordsSearch.SetKeywords(Words);
}
catch (Exception ex)
{
result = "0";
return result;
}
try
{
bool res = wordsSearch.ContainsAny(sourceText);
if (res)
{
result = "3";
return result;
}
else
{
result = "4";
return result;
}
}
catch (Exception ex)
{
return result = "1";
}
}
else
{
result = "2";
}
return result;
}
/// <summary>
/// 把对象写入到json文件中
/// </summary>
/// <param name="obj"></param>
/// <returns></returns>
public static void Write(List<keywords> jsonData, List<urlwords> urlJsonData, string filename)
{
var directorypath = Directory.GetCurrentDirectory();
string strFileName = directorypath + "\\" + filename + ".json";
string ListJson = "";
if (jsonData != null)
{
ListJson = JsonConvert.SerializeObject(jsonData);
}
else
{
ListJson = JsonConvert.SerializeObject(urlJsonData);
}
Console.WriteLine(ListJson);
writeJsonFile(strFileName, ListJson);
//将序列化的json字符串内容写入Json文件，并且保存
void writeJsonFile(string path, string jsonConents)
{
using (FileStream fs = new FileStream(path, FileMode.OpenOrCreate, System.IO.FileAccess.ReadWrite, FileShare.ReadWrite))
{
//如果json文件中有中文数据，可能会出现乱码的现象，那么需要加上如下代码
Encoding.RegisterProvider(CodePagesEncodingProvider.Instance);
using (StreamWriter sw = new StreamWriter(fs, Encoding.GetEncoding("GB2312")))
{
sw.WriteLine(jsonConents);
}
}
}
}
}
}
五、写API接口
/// <summary>
/// 进行敏感词脱敏
/// </summary>
/// <param name="sourctText">需要脱敏的文本内容</param>
/// <returns></returns>
[HttpPost]
public IActionResult sensitive_words_replace2(string sourctText)
{
string resultStr = "";
//实例化敏感词库
subContentCheck strCheck = new subContentCheck(_hostingEnv);
if (string.IsNullOrEmpty(sourctText))
{
return Json(new { code = 230, msg = "需要替换的文本内容为空！", resultStr = resultStr });
}
try
{
resultStr = strCheck.FilterWithChar(sourctText);
string resMsg = "";
int resCode = 200;
if (resultStr=="0")
{
resCode = 210;
resultStr = "";
resMsg = "设置违禁词时发生错误，请联系管理员！";
}else if (resultStr=="1")
{
resCode = 240;
resultStr = "";
resMsg = "敏感内容替换时发生错误！";
}
else if (resultStr == "2")
{
resCode = 260;
resultStr = "";
resMsg = "需要替换的文本内容为空！";
}
else
{
resCode = 200;
resMsg = "敏感词替换请求成功！";
}
return Json(new { code = resCode, msg = resMsg, resultStr = resultStr });
}
catch (Exception ex)
{
return Json(new { code = 220, msg = "敏感内容替换时发生错误！", resultStr = "" });
}
}
/// <summary>
/// 进行敏感词判断
/// </summary>
/// <param name="sourctText">需要脱敏的文本内容</param>
/// <returns></returns>
[HttpPost]
public IActionResult whether_sensitive_words(string sourctText)
{
string resultStr = "";
//实例化敏感词库
subContentCheck strCheck = new subContentCheck(_hostingEnv);
if (string.IsNullOrEmpty(sourctText))
{
return Json(new { code = 230, msg = "需要替换的文本内容为空！", resultStr = resultStr });
}
try
{
resultStr = strCheck.FindSensitiveKey(sourctText);
string resMsg = "";
int resCode = 200;
if (resultStr == "0")
{
resCode = 210;
resultStr = "";
resMsg = "设置违禁词时发生错误，请联系管理员！";
}
else if (resultStr == "1")
{
resCode = 240;
resultStr = "";
resMsg = "敏感内容匹配时发生错误！";
}
else if (resultStr == "2")
{
resCode = 260;
resultStr = "";
resMsg = "需要判断的文本内容为空！";
}
else if (resultStr == "3")
{
resCode = 270;
resultStr = "";
resMsg = "内容中含有敏感/违禁词！";
}
else
{
resCode = 200;
resMsg = "内容中不含敏感/违禁词！";
}
return Json(new { code = resCode, msg = resMsg, resultStr = resultStr });
}
catch (Exception ex)
{
return Json(new { code = 220, msg = "敏感内容匹配时发生错误！", resultStr = "" });
}
}
六、前端封装JS方法
/**
* 敏感词/违禁词替换
* @param {string} sourctText 需要进行替换的内容
* @param {string} boxid 将替换成功之后的内容赋值的元素容器id属性名
* @param {object} layui Layui实例
* @returns 替换之后的文本内容
*/
function sensitive_words_replace(sourctText, boxid, layui) {
let resultStr = "";
//let url = ["/Home/sensitive_words_replace", "/Home/sensitive_words_replace1", "/Home/sensitive_words_replace2"];
$.ajax({
url: "/Home/sensitive_words_replace2",//请求后端接口的路径
dataType: "JSON",
type: "POST",
data: {
"sourctText": sourctText
},
success: function (res) {
let resCode = res.code;
let resMsg = res.msg;
if ((resCode == "210" || resCode == 210) || (resCode == 220 || resCode == "220") || (resCode == 230 || resCode == "230") || (resCode == 240 || resCode == "240") || (resCode == 260 || resCode == "260")) {
//返回数据后关闭loading
layer.closeAll();
resultStr = res.resultStr;
layui.layer.alert(resMsg, { icon: 5, title: "温馨提示", closeBtn: 0 });
} else if (resCode == 200 || resCode == "200") {
resultStr = res.resultStr;
$("#" + boxid).val(resultStr);
//返回数据后关闭loading
layer.closeAll();
}
},
error: function (error) {
//返回数据后关闭loading
layer.closeAll();
layui.layer.alert(error, { icon: 5, title: "温馨提示", closeBtn: 0 });
}
});
return resultStr;
}
/**
* 查询是否包含敏感/违禁词
* @param {string} sourctText 需要进行替换的内容
* @param {string} boxid 将替换成功之后的内容赋值的元素容器id属性名
* @param {object} layui Layui实例
* @returns 返回Bool；包含：“true”；不包含：“false”
*/
function whether_sensitive_words(sourctText, boxid, layui) {
let resultBool = false;
$.ajax({
url: "/Home/whether_sensitive_words",//请求后端接口的路径
dataType: "JSON",
type: "POST",
async: false,//此处需要注意的是要想获取ajax返回的值这个async属性必须设置成同步的，否则获取不到返回值
data: {
"sourctText": sourctText
},
success: function (res) {
let resCode = res.code;
let resMsg = res.msg;
if ((resCode == "210" || resCode == 210) || (resCode == 220 || resCode == "220") || (resCode == 230 || resCode == "230") || (resCode == 240 || resCode == "240") || (resCode == 260 || resCode == "260")) {
resultBool = false;
layui.layer.alert(resMsg, { icon: 5, title: "温馨提示", closeBtn: 0 });
} else if (resCode == 270 || resCode == "270") {
resultBool = true;
} else if (resCode == 200 || resCode == "200") {
resultBool = false;
//返回数据后关闭loading
layer.closeAll();
}
},
error: function (error) {
layui.layer.alert(error, { icon: 5, title: "温馨提示", closeBtn: 0 });
}
});
return resultBool;
}
本文来自博客园，作者：TomLucas，转载请注明原文链接：https://www.cnblogs.com/lucasDC/p/17255906.html
