动态网站推广与搜索引擎优化初探

最近接触动态网站推广问题，查阅资料后发现：难！搜索引擎的蜘蛛爬虫（robot）MS主要通过目录浏览方式进行搜集，而动态网站页面很少，绝大部分都是动态生成的，要被搜索引擎收录比较困难。目前大致以下几种方式进行推广：一是在各大搜索引擎提交自己的网站；二是通过各种大大小小的分类目录；三是做链接交换和邮件推广；四是自己生成网站地图和robot文件（还有其它很多五花八门的手段），本文要描述的是第四种方式。

       第一个是动态网站的各个链接生成网站地图（据报道google,microsoft和yahoo联合声明一个统一的标准sitemap 0.9，事实上目前只有google一家可以提交网站地图， 参见：http://www.google.com/support/webmasters/bin/answer.py?answer=40318&hl=zh_CN），站点地图范本如下：

<?xml version="1.0" encoding="UTF-8"?>
    < urlset xmlns="http://www.google.com/schemas/sitemap/0.84">
     < url>
        < loc>http://www.example.com/</loc>
        < lastmod>2005-01-01</lastmod>
        < changefreq>monthly</changefreq>
        < priority>0.8</priority>
     </url>    
    </urlset>

       我的做法是用一张表记录点击的链接，再写一个页面来生成网站地图（存储地图文件的目录需要目录写授权）文件Sitemap.xml,代码如下： 

<!---->        /**//*
         * 生成网站地图Sitemap.xml
         * sid:网站代码
         */
        private void CreateXMLFile(string sid)
        ...{
                SqlParameter param1 = new SqlParameter("@SID", SqlDbType.VarChar, 20);
                param1.Value = sid;
                IDataParameter[] parameters = new IDataParameter[] ...{ param1 };
                DbHelperSQL dbHelper = new DbHelperSQL(connStr);
                string outParams = "";
                DataSet ds = dbHelper.RunProcedure("spGetSiteMap", parameters, "TmpSiteMapInfo", ref outParams);
                if (ds.Tables[0].Rows.Count > 0)
                ...{
                        string XMLSpace = "http://www.google.com/schemas/sitemap/0.9";
                        DateTime dt = System.DateTime.Now;

                        XmlText xmltext;
                        XmlElement xmlelem;
                        // Create a new, empty document.
                        XmlDocument doc = new XmlDocument();
                        XmlDeclaration docNode = doc.CreateXmlDeclaration("1.0", "UTF-8", null);
                        doc.AppendChild(docNode);

                        // Create and insert a new element.
                        XmlNode urlset = doc.CreateNode(XmlNodeType.Element, "urlset", XMLSpace);
                        doc.AppendChild(urlset);

                        foreach (DataRow dr in ds.Tables[0].Rows)
                        ...{
                                // Create a nested element (with an attribute).
                                XmlElement url = doc.CreateElement("", "url", XMLSpace);
                                urlset.AppendChild(url);

                                xmlelem = doc.CreateElement("", "loc", XMLSpace);
                                xmltext = doc.CreateTextNode(dr["URL"].ToString());
                                xmlelem.AppendChild(xmltext);
                                url.AppendChild(xmlelem);

                                xmlelem = doc.CreateElement("", "lastmod", XMLSpace);
                                xmltext = doc.CreateTextNode(string.Format("{0:u}", dt).Substring(0,10));
                                xmlelem.AppendChild(xmltext);
                                url.AppendChild(xmlelem);

                                xmlelem = doc.CreateElement("", "changefreq", XMLSpace);
                                if (dr["Type"].ToString() == "1")
                                        xmltext = doc.CreateTextNode("daily");
                                else
                                        xmltext = doc.CreateTextNode("monthly");
                                xmlelem.AppendChild(xmltext);
                                url.AppendChild(xmlelem);

                                xmlelem = doc.CreateElement("", "priority", XMLSpace);
                                xmltext = doc.CreateTextNode(dr["OrderNo"].ToString());
                                xmlelem.AppendChild(xmltext);
                                url.AppendChild(xmlelem);
                        }

                        doc.Save(Server.MapPath("Sitemap.xml"));
                }
                return;
        }

       上面的代码比较简单，一个要点是如果上面结点（比如：urlset）带有命名空间，则下层结点也一定要带，否则下层结点会自动带一个空的命名空间（好像与习惯思维相反，这点花了偶不少时间）。

       第二个是robots（爬虫用的配置文件），也有相关标准，网上资料很多，下面是我写的生成robots文件的代码： 

 

<!---->        private void CreateRobotFile(string sid)
        ...{
                SqlParameter param1 = new SqlParameter("@SID", SqlDbType.VarChar, 20);
                param1.Value = sid;
                IDataParameter[] parameters = new IDataParameter[] ...{ param1 };
                DbHelperSQL dbHelper = new DbHelperSQL(connStr);
                string outParams = "";
                DataSet ds = dbHelper.RunProcedure("spGetSiteMap", parameters, "TmpSiteMapInfo", ref outParams);
                if (ds.Tables[0].Rows.Count > 0)
                ...{
                        FileStream fs = new FileStream(Server.MapPath("robots.txt"), FileMode.OpenOrCreate, FileAccess.Write);
                        StreamWriter m_streamWriter = new StreamWriter(fs);
                        m_streamWriter.Flush();
                        //    使用StreamWriter来往文件中写入内容
                        m_streamWriter.BaseStream.Seek(0, SeekOrigin.Begin);
                        //    把richTextBox1中的内容写入文件
                        m_streamWriter.WriteLine("# Robots.txt file from http://www.hugesoft.net");
                        m_streamWriter.WriteLine("# All robots will spider the domain");
                        m_streamWriter.WriteLine("");
                        m_streamWriter.WriteLine("Sitemap: http://www.hugesoft.net/Sitemap.xml");
                        m_streamWriter.WriteLine("User-agent: *");
                        m_streamWriter.WriteLine("Disallow: ");
                        foreach (DataRow dr in ds.Tables[0].Rows)
                        ...{
                                string str = dr["URL"].ToString().ToLower();
                                int index = str.IndexOf("http://");
                                if (index < 0)
                                        continue;
                                index = str.IndexOf("/",index + 7);
                                if (index < 0)
                                        continue;
                                str = str.Substring(index);
                                m_streamWriter.WriteLine("Allow: " + str);
                        }
                        //关闭此文件
                        m_streamWriter.Flush();
                        m_streamWriter.Close();
                }
        }

 

       由于我记录的是绝对URL，因此生成robots.txt时需要解析URL，去掉域名部分。别某些标记（比如：Sitemap,Allow）不一定能被所有类型的robot支持。

       robots.txt文件一定要放在网站根目录下，Sitemap.xml可以提交给google（目前只此一家）。