求一正则表达式

Question

求一正则表达式可以去除html文本中的,script, style, object 和基本的html标签

Answer 1

public static string NoHTML(string Htmlstring)
        {            Htmlstring = Regex.Replace(Htmlstring, @"<script[^>]*?>.*?</script>", "", RegexOptions.IgnoreCase);
            //Htmlstring = Regex.Replace(Htmlstring, @"<(.[^>]*)>", "", RegexOptions.IgnoreCase);            Htmlstring = Regex.Replace(Htmlstring, @"<([^pbul].[^>]*[^pbul])>", "", RegexOptions.IgnoreCase);            Htmlstring = Regex.Replace(Htmlstring, @"</[^pbul]*>", "", 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 = Regex.Replace(Htmlstring, @"<li>", "<br>", RegexOptions.IgnoreCase);            // Htmlstring = Htmlstring.Replace("<", "");            // Htmlstring = Htmlstring.Replace(">", "");            //Htmlstring = Htmlstring.Replace("\r\n", "");            //Htmlstring = Htmlstring.Replace("\n", "");            // Htmlstring = HttpContext.Current.Server.HtmlEncode(Htmlstring).Trim();            return Htmlstring;
        }