去除网站中HTML标签的正则表达式

string str = a.ToString();

str = Regex.Replace(str, @"</?span[^>]*>", "", RegexOptions.IgnoreCase);

str = Regex.Replace(str, @"&#[^>]*;", "", RegexOptions.IgnoreCase);
str = Regex.Replace(str, @"</?marquee[^>]*>", "", RegexOptions.IgnoreCase);
str = Regex.Replace(str, @"</?object[^>]*>", "", RegexOptions.IgnoreCase);
str = Regex.Replace(str, @"</?param[^>]*>", "", RegexOptions.IgnoreCase);
str = Regex.Replace(str, @"</?embed[^>]*>","",RegexOptions.IgnoreCase);
str = Regex.Replace(str, @"</?table[^>]*>","",RegexOptions.IgnoreCase);
str = Regex.Replace(str, @" ","",RegexOptions.IgnoreCase);
str = Regex.Replace(str, @"</?tr[^>]*>", "", RegexOptions.IgnoreCase);
str = Regex.Replace(str, @"</?th[^>]*>","",RegexOptions.IgnoreCase);
str = Regex.Replace(str, @"</?p[^>]*>","",RegexOptions.IgnoreCase);
str = Regex.Replace(str, @"</?a[^>]*>","",RegexOptions.IgnoreCase);
str = Regex.Replace(str, @"</?img[^>]*>","",RegexOptions.IgnoreCase);

str = Regex.Replace(str, @"</?tbody[^>]*>","",RegexOptions.IgnoreCase);
str = Regex.Replace(str, @"</?li[^>]*>","",RegexOptions.IgnoreCase);
str = Regex.Replace(str, @"</?span[^>]*>","",RegexOptions.IgnoreCase);
str = Regex.Replace(str, @"</?div[^>]*>","",RegexOptions.IgnoreCase);
str = Regex.Replace(str, @"</?th[^>]*>", "", RegexOptions.IgnoreCase);
str = Regex.Replace(str, @"</?td[^>]*>", "", RegexOptions.IgnoreCase);
str = Regex.Replace(str, @"</?script[^>]*>", "", RegexOptions.IgnoreCase);
str = Regex.Replace(str, @"(javascript|jscript|vbscript|vbs):", "", RegexOptions.IgnoreCase);
str = Regex.Replace(str, @"on(mouse|exit|error|click|key)", "", RegexOptions.IgnoreCase);
str = Regex.Replace(str, @"<//?xml[^>]*>", "", RegexOptions.IgnoreCase);
str = Regex.Replace(str, @"<//?[a-z]+:[^>]*>", "", RegexOptions.IgnoreCase);
str = Regex.Replace(str, @"</?font[^>]*>", "", RegexOptions.IgnoreCase);
str = Regex.Replace(str, @"</?b[^>]*>","",RegexOptions.IgnoreCase);
str = Regex.Replace(str, @"</?u[^>]*>","",RegexOptions.IgnoreCase);
str = Regex.Replace(str, @"</?i[^>]*>","",RegexOptions.IgnoreCase);
str = Regex.Replace(str, @"</?strong[^>]*>","",RegexOptions.IgnoreCase);

 

  • 0
    点赞
  • 1
    收藏
    觉得还不错? 一键收藏
  • 0
    评论
评论
添加红包

请填写红包祝福语或标题

红包个数最小为10个

红包金额最低5元

当前余额3.43前往充值 >
需支付:10.00
成就一亿技术人!
领取后你会自动成为博主和红包主的粉丝 规则
hope_wisdom
发出的红包
实付
使用余额支付
点击重新获取
扫码支付
钱包余额 0

抵扣说明:

1.余额是钱包充值的虚拟货币,按照1:1的比例进行支付金额的抵扣。
2.余额无法直接购买下载,可以购买VIP、付费专栏及课程。

余额充值