网页解析：如何获得网页源码中嵌套的标签。

一：前言：网页源码中有很多嵌套的标签

例如div标签嵌套如：bUTP<DIV>finally<div>aurora</div>@126.com</div><div class=\"Cited1\">ggff</div>

我们的网页解析工作中有时候需要解嵌套。通俗的讲就是把嵌套的标签以线性表的形式表示出来。还拿上面的例子来说明。即解嵌套为

<div>aurora</div>

 <DIV>finally<div>aurora</div>@126.com</div>

div class=\"Cited1\">ggff</div>

核心代码如下：

 

 class ThemeIRAssist
    {
        public static void  GetNodesByTags( ref string rawtext,string tags,ref  List<string>result )
       {   //储存开始标签的位置
           List<Position > beginTagPos = new List<Position >();
            //储存结束标签的位置
           List<Position> endTagPos = new List<Position>();
            //匹配开始标签的正则表达式
           string sBeginTagPattern = "<" + tags;
           Regex regexBeginTag = new Regex(sBeginTagPattern, RegexOptions.IgnoreCase | RegexOptions.Singleline);
           //匹配结束标签的正则表达式。
           string sEndTagPattern = "</" + tags + ">";
           Regex regexEndTag = new Regex(sEndTagPattern, RegexOptions.Singleline | RegexOptions.IgnoreCase);
            //获得开始标签的集合
           MatchCollection beginTagCollection = regexBeginTag.Matches(rawtext);
            //获得结束标签的集合
           MatchCollection endTagCollection = regexEndTag.Matches(rawtext);
           foreach (Match mymatch in beginTagCollection)
           {
              Position pos=new Position();
               pos.nPos=mymatch.Index;
               pos.VistStatus=false;
               beginTagPos.Add(pos);
               
           }

           foreach (Match mymatch in endTagCollection)
           {
               Position pos = new Position();
               pos.nPos = mymatch.Index;
               pos.VistStatus = false;
               endTagPos.Add(pos);


           }
           for (int i = 0; i < endTagPos.Count; i++)
           {
               for (int j = beginTagPos.Count - 1; j >= 0; j--)
               {
                   if(endTagPos[i].nPos<beginTagPos[j].nPos)
                       continue;
                   else
                   {
                       if (beginTagPos[j].VistStatus)
                           continue;
                       else
                       {
                           result.Add(rawtext.Substring(beginTagPos[j].nPos,endTagPos[i].nPos-beginTagPos[j].nPos+6));
                           beginTagPos[j].VistStatus=true;
                           break;
                       }

                   }
               }
           }
            






 
       }
    }
    class Position
    {
       private int pos;
        private bool visited;
        public int nPos
        {
            get { return pos; }
            set { pos = value; }
       
        }
        public bool VistStatus
        {
            get { return visited; }
            set { visited = value; }
        }

    }

Main函数测试如下：

 

 static void Main(string[] args)
        {
            
            StreamReader sr = new StreamReader("D:\\finally.txt", Encoding.GetEncoding("utf-8"));
            string rawtext = sr.ReadToEnd();
           // string rawtext = "bUTP<DIV>finally<div>aurora</div>@126.com</div><div class=\"Cited1\">ggff</div>";
            List<string> result = new List<string>();
            ThemeIRAssist.GetNodesByTags(ref rawtext, "div", ref result);
            Console.WriteLine(result.Count);
            Regex regexStadard = new Regex(@"^<div\s+class=""Cited1""", RegexOptions.IgnoreCase | RegexOptions.Singleline);
            for (int i = result.Count - 1; i >= 0; i--)
            {
                if(!regexStadard.IsMatch(result[i]))
                    result.RemoveAt(i);
            }
            Console.WriteLine(result.Count);
            foreach(string s in result)
            {
                Console.WriteLine("***************************************************");
                Console.Write(s);
                Console.WriteLine("***************************************************");
            }
            Console.Read();
           
        }