Windows下比较简单的获取网页源码的方法

第一个方法是使用MFC里面的 <afxinet.h>

 CString GetHttpFileData(CString strUrl)
{
     CInternetSession Session("Internet Explorer", );
     CHttpFile *pHttpFile = NULL;
     CString strData;
     CString strClip;
     pHttpFile = (CHttpFile*)Session.OpenURL(strUrl);
    while ( pHttpFile->ReadString(strClip) )
     {
      strData += strClip;
      }
     return strData;
}

要讲一下，pHttpFile->ReadString() 每次可能只读一个数据片断，读多少次取决于网络状况，所以要把每次读到的数据加到总数据的尾部，用了CString 省去了缓冲区处理:) 
别忘了包含头文件#include <afxinet.h> 在工程设置，里面要选择 using MFC 要不然编译不了

第二种是使用WinNet的纯API实现的

#define MAXBLOCKSIZE 1024
#include <windows.h>
#include <wininet.h>

#pragma comment(lib, "wininet.lib")

void GetWebSrcCode(const char *Url);

int _tmain(int argc, _TCHAR* argv[])
{
    GetWebSrcCode("http://www.cnblogs.com/");

    return ;
}

void GetWebSrcCode(const char *Url)
{
    HINTERNET hSession = InternetOpen("zwt", INTERNET_OPEN_TYPE_PRECONFIG, NULL, NULL, );
    if (hSession != NULL)
    {
        HINTERNET hURL = InternetOpenUrl(hSession, Url, NULL, , INTERNET_FLAG_DONT_CACHE, );
        if (hURL != NULL)
        {
            char Temp[MAXBLOCKSIZE] = {};
            ULONG Number = ;

            FILE *stream;
            if( (stream = fopen( "E:\\test.html", "wb" )) != NULL )
            {
                while (Number > )
                {
                    InternetReadFile(hURL, Temp, MAXBLOCKSIZE - , &Number);
                    fwrite(Temp, sizeof (char), Number , stream);
                }
                fclose( stream );
            }

            InternetCloseHandle(hURL);
            hURL = NULL;
        }

        InternetCloseHandle(hSession);
        hSession = NULL;
    }
}

第三种就是使用非封装过的Socket实现了

int main(int argc, char* argv[])
{
    SOCKET hsocket;
    SOCKADDR_IN saServer;
    WSADATA wsadata;
    LPHOSTENT lphostent;
    int nRet;
    char Dest[];
    char* host_name="blog.sina.com.cn";
    char* req="GET /s/blog_44acab2f01016gz3.html HTTP/1.1\r\n"
        "User-Agent: Mozilla/4.0 (compatible; MSIE 8.0; Windows NT 5.1; Trident/4.0; .NET CLR 1.1.4322; .NET CLR 2.0.50727; .NET4.0C; .NET4.0E; .NET CLR 3.0.4506.2152; .NET CLR 3.5.30729)\r\n"
        "Host:blog.sina.com.cn\r\n\r\n";

    // 初始化套接字
    if(WSAStartup(MAKEWORD(,),&wsadata))
        printf("初始化SOCKET出错！");
    lphostent=gethostbyname(host_name);
    if(lphostent==NULL)
        printf("lphostent为空！");
    hsocket = socket(AF_INET,SOCK_STREAM,IPPROTO_TCP);
    saServer.sin_family = AF_INET;
    saServer.sin_port = htons();
    saServer.sin_addr =*((LPIN_ADDR)*lphostent->h_addr_list);
    // 利用SOCKET连接
    nRet = connect(hsocket,(LPSOCKADDR)&saServer,sizeof(SOCKADDR_IN));
    if(nRet == SOCKET_ERROR)
    {
        printf("建立连接时出错！");
        closesocket(hsocket);
        return ;
    }
    // 利用SOCKET发送   

    nRet = send(hsocket,req,strlen(req),);
    if(nRet==SOCKET_ERROR)
    {
        printf("发送数据包时出错！");
        closesocket(hsocket);
    }
    nRet=;
    while(nRet>)
    {
        // 接收返回数据包
        nRet=recv(hsocket,(LPSTR)Dest,sizeof(Dest),);
        if(nRet>)
            Dest[nRet]=;
        else
            Dest[]=;
        char sDest[] = {};
        UTF8_2_GB2312(sDest,nRet,Dest,nRet);
        // 显示返回数据包的大小、内容
        //printf("\nReceived bytes:%d\n",nRet);
        printf("Result:\n%s",sDest);
    }
}

另外，以上我们获取网页的时候，获取到的可能是UTF8，似乎目前大多数网站都用的这种编码吧！下面是编码转换。

void UTF_8ToUnicode(wchar_t* pOut,char *pText)
{
    char* uchar = (char *)pOut;
    uchar[] = ((pText[] & 0x0F) << ) + ((pText[] >> ) & 0x0F);
    uchar[] = ((pText[] & 0x03) << ) + (pText[] & 0x3F);
}
void Gb2312ToUnicode(wchar_t* pOut,char *gbBuffer)
{
    ::MultiByteToWideChar(CP_ACP,MB_PRECOMPOSED,gbBuffer,,pOut,);
}
void UTF_8ToGB2312(char*pOut, char *pText, int pLen)
{
    char Ctemp[];
    memset(Ctemp,,);
    int i = ,j = ;
    while(i < pLen)
    {
        if(pText[i] >= )
        {
            pOut[j++] = pText[i++];
        }
        else
        {
            WCHAR Wtemp;
            UTF_8ToUnicode(&Wtemp,pText + i);
            UnicodeToGB2312(Ctemp,Wtemp);
            pOut[j] = Ctemp[];
            pOut[j + ] = Ctemp[];
            i += ;
            j += ;
        }
    }
    pOut[j] ='\n';
    return;
} 

这是是转换成GB2312的代码