谈谈如果移除HTML中的HTML Tag

最近在做把包含了HTML代码的文本显示在Grid里面, 所以就需要把HTML tag给移除调. 用了一些方法都不是太理想,分享一下经验.

1. 用正则表达式.

 1public static string RemoveHtmlTags(string strHtml)
 2        {
 3            string[] aryReg = {
 4                                  @"<script[^>]*?>.*?</script>",
 5                                  @"<(\/\s*)?!?((\w+:)?\w+)(\w+-?\w+(\s*=?\s*(([""'])(\\[""'tbnr]|[^\7])*?\7|\w+)|.{0})|\s)*?(\/\s*)?>"
 6                                  ,
 7                                  "<\\s*/{0,1}\\s*(font)[^>]*>",
 8                                  @"([\r\n])[\s]+",
 9                                  @"&(quot|#34);",
10                                  @"&(amp|#38);",
11                                  @"&(lt|#60);",
12                                  @"&(gt|#62);",
13                                  @"&(nbsp|#160);",
14                                  @"&(iexcl|#161);",
15                                  @"&(cent|#162);",
16                                  @"&(pound|#163);",
17                                  @"&(copy|#169);",
18                                  @"&#(\d+);",
19                                  @"-->",
20                                  @"<!--.*\n"
21                              };
22
23            string[] aryRep = {
24                                  "",
25                                  "",
26                                  "",
27                                  "",
28                                  "\"",
29                                  "&",
30                                  "<",
31                                  ">",
32                                  " ",
33                                  "\xa1", //chr(161),
34                                  "\xa2", //chr(162),
35                                  "\xa3", //chr(163),
36                                  "\xa9", //chr(169),
37                                  "",
38                                  "\r\n",
39                                  ""
40                              };
41
42            string strOutput = strHtml;
43            for (int i = 0; i < aryReg.Length; i++)
44            {
45                Regex regex = new Regex(aryReg[i], RegexOptions.IgnoreCase);
46                strOutput = regex.Replace(strOutput, aryRep[i]);
47            }
48            strOutput.Replace("<", "");
49            strOutput.Replace(">", "");
50            strOutput.Replace("\r\n", "");
51            return strOutput;
52        }

这种方法的缺点是正则表达式执行比较慢.在数据比较多的时候会占用大量系统资源,引起程序失去响应

2.用WebBrowser的InnerText功能

 1        public static string RemoveHtmlTags(string strHtml)
 2        {
 3            WebBrowser wb = new WebBrowser();
 4            wb.ScriptErrorsSuppressed = false;
 5            wb.Navigate("about:blank");
 6            while (wb.ReadyState != WebBrowserReadyState.Complete)
 7            {
 8                Application.DoEvents();
 9            }
10            try
11            {
12                wb.Document.Write(strHtml);
13                return wb.Document.Body.InnerText;
14            }
15            finally
16            {
17                wb.Dispose();
18            }
19        }

这种方法,我觉得不好,不易多用,因为WebBrowser这个东东毕竟是包装COM的东西.

3.写个HTML Parser自己来解析HTML,想怎么去Tag都可以了.
我没有自己去写,在CodePlex里面去弄了一个来用了,感觉比上面两种的效率都要好.给个链接有需要的去看看吧
http://www.codeplex.com/htmlagilitypack

也贴点这个项目里面的代码吧.

HTMLaglityPack
 1public static string RemoveHtmlTags(string strHtml)
 2        {
 3            HtmlDocument doc = new HtmlDocument();
 4            doc.LoadHtml(strHtml);
 5
 6            StringWriter sw = new StringWriter();
 7            ConvertTo(doc.DocumentNode, sw);
 8            sw.Flush();
 9            return sw.ToString();
10        }
11
12        private static void ConvertContentTo(HtmlNode node, TextWriter outText)
13        {
14            foreach (HtmlNode subnode in node.ChildNodes)
15            {
16                ConvertTo(subnode, outText);
17            }
18        }
19
20        private static void ConvertTo(HtmlNode node, TextWriter outText)
21        {
22            switch (node.NodeType)
23            {
24                case HtmlNodeType.Comment:
25                    // don't output comments
26                    break;
27
28                case HtmlNodeType.Document:
29                    ConvertContentTo(node, outText);
30                    break;
31
32                case HtmlNodeType.Text:
33                    // script and style must not be output
34                    string parentName = node.ParentNode.Name;
35                    if ((parentName == "script") || (parentName == "style"))
36                        break;
37
38                    // get text
39                    string html = ((HtmlTextNode)node).Text;
40
41                    // is it in fact a special closing node output as text?
42                    if (HtmlNode.IsOverlappedClosingElement(html))
43                        break;
44
45                    // check the text is meaningful and not a bunch of whitespaces
46                    if (html.Trim().Length > 0)
47                    {
48                        outText.Write(HtmlEntity.DeEntitize(html));
49                    }
50                    break;
51
52                case HtmlNodeType.Element:
53                    switch (node.Name)
54                    {
55                        case "p":
56                            // treat paragraphs as crlf
57                            outText.Write("\r\n");
58                            break;
59                    }
60
61                    if (node.HasChildNodes)
62                    {
63                        ConvertContentTo(node, outText);
64                    }
65                    break;
66            }
67        }