C#去除HTML标记

最新推荐文章于 2023-07-13 15:51:32 发布

放开那段代码让我来

最新推荐文章于 2023-07-13 15:51:32 发布

阅读量516

点赞数

分类专栏： net 文章标签： net

本文链接：https://blog.csdn.net/u014424282/article/details/78720043

版权

net 专栏收录该内容

19 篇文章 0 订阅

订阅专栏

1、

using System;
using System.Collections.Generic;
using System.Linq;
using System.Text.RegularExpressions;
using System.Web;
using System.Web.UI;
using System.Web.UI.WebControls;

namespace WebApplication2
{
    public partial class WebForm1 : System.Web.UI.Page
    {
        protected void Page_Load(object sender, EventArgs e)
        {

        }

        ZKWebForm1219_20171128Entities db = new ZKWebForm1219_20171128Entities();
        protected void Button1_Click(object sender, EventArgs e)
        {
            var htmlstring = "<html xmlns='http://www.w3.org/1999/xhtml'><head runat='server'><meta http-equiv='Content-Type' content='text/html; charset=utf-8'/><title></title>  </head><body><form id='form1' runat='server'><div><asp:Button ID='Button1' runat='server' Text='Button'  /></div><div>武汉</div> </form><script src='Scripts/jquery-1.10.2.min.js'></script></body></html>";

            htmlstring = Extenions.RemoveHTML(htmlstring);
            Response.Write(htmlstring);      
        }
    }

    public static class Extenions
    {
        ///   <summary>   
        ///   去除HTML标记   
        ///   </summary>   
        ///   <param   name="NoHTML">包括HTML的源码   </param>   
        ///   <returns>已经去除后的文字</returns>   
        public static string RemoveHTML(this string Htmlstring)
        {
                //删除脚本   
                Htmlstring = Regex.Replace(Htmlstring, @"<script[^>]*?>.*?</script>", "", RegexOptions.IgnoreCase);
                //删除HTML   
                Htmlstring = Regex.Replace(Htmlstring, @"<(.[^>]*)>", "", RegexOptions.IgnoreCase);
               Htmlstring = Regex.Replace(Htmlstring, @"([\r\n])[\s]+", "", RegexOptions.IgnoreCase);
                Htmlstring = Regex.Replace(Htmlstring, @"-->", "", RegexOptions.IgnoreCase);
               Htmlstring = Regex.Replace(Htmlstring, @"<!--.*", "", RegexOptions.IgnoreCase);

               Htmlstring = Regex.Replace(Htmlstring, @"&(quot|#34);", "\"", RegexOptions.IgnoreCase);
               Htmlstring = Regex.Replace(Htmlstring, @"&(amp|#38);", "&", RegexOptions.IgnoreCase);
                Htmlstring = Regex.Replace(Htmlstring, @"&(lt|#60);", "<", RegexOptions.IgnoreCase);
               Htmlstring = Regex.Replace(Htmlstring, @"&(gt|#62);", ">", RegexOptions.IgnoreCase);
                Htmlstring = Regex.Replace(Htmlstring, @"&(nbsp|#160);", "   ", RegexOptions.IgnoreCase);
                Htmlstring = Regex.Replace(Htmlstring, @"&(iexcl|#161);", "\xa1", RegexOptions.IgnoreCase);
               Htmlstring = Regex.Replace(Htmlstring, @"&(cent|#162);", "\xa2", RegexOptions.IgnoreCase);
                Htmlstring = Regex.Replace(Htmlstring, @"&(pound|#163);", "\xa3", RegexOptions.IgnoreCase);
               Htmlstring = Regex.Replace(Htmlstring, @"&(copy|#169);", "\xa9", RegexOptions.IgnoreCase);
                Htmlstring = Regex.Replace(Htmlstring, @"&#(\d+);", "", RegexOptions.IgnoreCase);

               Htmlstring.Replace("<", "");
                Htmlstring.Replace(">", "");
               Htmlstring.Replace("\r\n", "");
               Htmlstring = HttpContext.Current.Server.HtmlEncode(Htmlstring).Trim();

            return Htmlstring;
        }

    }
}