请高手指教高效的Split函数

     /* *C++实现 */    

// String继承于std::string  
      
// 我的Split代码,vector操作很慢啊。 
                
typedef vector < String >  Strings;  
 
Strings Split( char *  by, bool  havenull = true )  
{   
    Strings strs;    
    typedef unsigned  int  UINT;   
    UINT bylen = strlen(by);   
    UINT start = 0 ;   
    UINT pos = 0 ;   
     char *  head = ( char * ) this -> c_str();   
     if (havenull)   
    {   
         while ((pos = this -> find(by,start)) !=- 1 )   
        {   
             * (head + pos) = 0 ;   
            strs.push_back(head + start);   
             * (head + pos) =* by;   
            start = pos + bylen;   
        };   
    }   
     else    
    {   
         while ((pos = this -> find(by,start)) !=- 1 )   
        {   
             if (pos - start > 0 )                                       
            {   
                 * (head + pos) = 0 ;   
                strs.push_back(head + start);   
                 * (head + pos) =* by;   
            };   
            start = pos + bylen;   
        };   
    };   
     if (start < this -> size())   
    {   
        strs.push_back( this -> substr(start, this -> size() - start));   
    };   
     return  strs;   
};   
// 测试代码：   
void  main()   
{   
    String s;   
    Strings ok;   
    s = " /c:;lihdaofhdo/al;jdf;lkajdf/adlfkadfj;ajdqwer/asdfja;ldsjqwerq/difjtt.cpp// " ;   
     for ( int  j = 0 ;j < 1000000 ;j ++ )   
        ok = s.Split( " / " , false );   
     for ( int  i = 0 ;i < ok.size();i ++ )   
    {   
        cout << ok[i] << endl;   
    };   
};   
           
     /* *C#实现的代码 */    
     using  System;   
     using  System.Collections.Generic;   
     using  System.Text;   
       
     namespace  ConsoleApplication1   
    {   
         class  Program   
        {   
             static   void  Main( string [] args)   
            {   
                String a  = " /c:;lihdaofhdo/al;jdf;lkajdf/adlfkadfj;ajdqwer/asdfja;ldsjqwerq/difjtt.cpp// " ;   
                 string  s  =   " / " ;   
                 char [] by  =  s.ToCharArray();   
                String [] strs;   
                Console.ReadKey();   
               strs  =  a.Split(by);   
                for  ( int  j  =   0 ; j  <   1000000 ; j ++ )   
                {   
                   strs  =  a.Split(by);   
               };   
                for  ( int  i  =   0 ; i  <  strs.Length; i ++ )   
               {   
                   Console.WriteLine(strs[i]);   
               };   
                Console.ReadKey();   
           }   
       }   
    }   

// 测试机器CPU：AMD64 3000+ ，操作系统Windows XPSP2。   
// 都执行字符串切分一百万次，这时C#的不到一秒完成，C++打开优化要六秒的样子。
// 哪位大侠知道高效的C++下Split代码？嵌入汇编完成应该会快点。