http协议多线程下载

本文介绍了一个基于C++的多线程HTTP文件下载程序,支持断点续传功能。通过发送带有Range头的HTTP请求来实现文件的分块下载,并在下载完成后自动合并文件。

#include <iostream>
#include <iosfwd>
#include <fstream>
#include <regex>
#include <string>
#include <winsock2.h>
#include <pthread.h>
#include <map>
#include <queue>

/**
*要实现多线程http协议下载文件的功能,支持断点续传和多线程同时下载
*http请求使用Range: bytes完成
*同时写n个临时文件,最后执行完成拼接在一起
*目前支持多线程下载,自动拼接文件,
*后续增加一个守护进程,判断是否都下载完成,记录下载进度,若是未下载完成,继续下载,若是下载完成,就完成文件拼接
**/

using namespace std;

SOCKET sock[100];
//测试主机和端口
const char *testHostName="www.qxmuye.com";
const short testPort=80;
const string testPortChar="80";
//文件地址
const char *url="/article/UploadPic/2015-9/2015982144994925.gif";
const string dataFile="D://testmap//test.gif";
const string tmpFile="D://testmap//tmpfile";
//分块下载的块大小
const int getSize=100000;

int fileNameInt;

//发送http请求包
bool sendHttpQuery(string sendQueryStr,int sockNum){
    int n=0;
   //初始化socket
    sock[sockNum] = socket(AF_INET, SOCK_STREAM, 0);

    if (sock[sockNum] == INVALID_SOCKET)
    {
        cout << "建立socket失败! 错误码: " << WSAGetLastError() << endl;
        return false;
    }
    sockaddr_in sa = { AF_INET };
    u_short port=0;
    sa.sin_port = htons(port);
    sa.sin_addr.s_addr = htonl(INADDR_ANY);
    n = bind(sock[sockNum], (sockaddr*)&sa, sizeof(sa));
    if (n == SOCKET_ERROR)
    {
        cout << "bind函数失败! 错误码: " << WSAGetLastError() << endl;
        return false;
    }
    struct hostent *p = gethostbyname(testHostName);
    if (p == NULL)
    {
        cout << "主机无法解析出ip! 错误吗: " << WSAGetLastError() << endl;
        return false;
    }

    sa.sin_port = htons(testPort);//随机分配
    memcpy(&sa.sin_addr, p->h_addr, 4);
    //sa.sin_addr.S_un.S_addr = inet_addr(*(p->h_addr_list));
    //cout << *(p->h_addr_list) << endl;
    //连接
    n = connect(sock[sockNum], (sockaddr*)&sa, sizeof(sa));
    if (n == SOCKET_ERROR)
    {
        cout << "connect函数失败! 错误码: " << WSAGetLastError() << endl;
        return false;
    }

    //按照http送GET请求
    cout<<"发送http请求:"+sendQueryStr<<endl;
    if (SOCKET_ERROR == send(sock[sockNum], sendQueryStr.c_str(), sendQueryStr.size(), 0))
    {
        cout << "send error! 错误码: " << WSAGetLastError() << endl;
        closesocket(sock[sockNum]);
        return false;
    }
    return true;
}

//文件拼接,将第二个参数文件中的内容复制给第一个参数中的文件,两个文件要求已打开
void getGolobalFile(ofstream &fGolobal,ifstream &fPart){
    //fGolobal.open(dataFile,ios::app|ios::binary);
    char tmpFileChar[1001];
    if(!fPart.is_open()){
        cout<<"临时文件未打开!"<<endl;
    }
    if(!fGolobal.is_open()){
        cout<<"总文件未打开!"<<endl;
    }
    while(!fPart.eof()){
        fPart.read(tmpFileChar,1000);
        fGolobal.write(tmpFileChar,fPart.gcount());
        //char readNumChar[10];
        //sprintf(readNumChar,"%d",fPart.gcount());
        //cout<<"复制字符长度:"+(string)readNumChar<<endl;
        //memset(tmpFileChar,0,1024);
    }
    cout<<"文件复制完成!"<<endl;
}

//从http的返回头信息中解析出文件长度
char * getResponseLeng(const char receiveBuf[1024]){

    //找到Content-Range: bytes 0-1/之后的字符串lengStartChar
    char *lengStartChar = strstr(receiveBuf, "Content-Range: bytes 0-1/");
    int tmpSize=sizeof("Content-Range: bytes 0-1/");//注意sizeof长度将结束符计算进去,后面需要减掉1
    //char *tmpSizeChar;
    //sprintf(tmpSizeChar,"%d",tmpSize);
    //cout<<(string)tmpSizeChar<<endl;
    //char *lengEndchar=strstr(buf, "\r\n");
    int i=0;
    bool isEnd=false;
    //在lengStartChar找到\r\n之前的位置
    while(!isEnd){
        if(lengStartChar[i+1]!='\r')i++;
        else{
            if(lengStartChar[i+2]=='\n')isEnd=true;
            else i++;
        }
    }

    //将lengStartChar中关于长度的一段赋值给长度字符串,注意减去结束符
    char *leng=new char(10);
    memset(leng,0x00,10);
    memcpy(leng,lengStartChar+tmpSize-1,i-tmpSize+2);
    //cout<<"111  "+(string)leng+"  111"<<endl;

    return leng;
}

//下载线程,将特定的文件块下载到特定的临时文件中,根据参数设置特定的临时文件
void *downloadTmpFile(void *args){
    int n;
    //修改子线程为分离状态,不阻塞主线程
    pthread_detach(pthread_self());
    int tmpNum=*(int *)args;//先将指针转为int型指针,再取值
    char tmpBuf[1024];
    memset(tmpBuf, 0, sizeof(tmpBuf));

    char tmpNumChar[10];
    sprintf(tmpNumChar,"%d",tmpNum);
    ofstream fileTmp;
    //临时文件新建
    fileTmp.open(tmpFile+(string)tmpNumChar ,ios::out|ios::binary);

    cout << " 文档打开 "+tmpFile+(string)tmpNumChar  << endl;
    if(!fileTmp.is_open()){
       cout << " 文档打开失败! "+tmpFile+(string)tmpNumChar << endl;
    }

    char *startMark=new char(10);
    char *endMark=new char(10);//设置开始地址,结束地址,注意
    memset(startMark,0,10);
    memset(endMark,0,10);
    sprintf(startMark,"%d",(tmpNum*getSize+0));
    sprintf(endMark,"%d",(tmpNum*getSize+(getSize-1)));
    string strFirst="GET "+(string)url+" HTTP/1.1\r\nHost:"+(string)testHostName+":"+testPortChar+"\r\nRange:bytes="+(string)startMark+"-"+(string)endMark+"\r\nConnection:Close\r\n\r\n";
    //发送http请求
    if(!sendHttpQuery(strFirst,tmpNum)){
        cout<<" http 请求发送失败! "<<endl;
    }
    n = recv(sock[tmpNum], tmpBuf, sizeof(tmpBuf)-1, 0);
    char *cpos = strstr(tmpBuf, "\r\n\r\n");
    fileTmp.write(cpos + strlen("\r\n\r\n"), n - (cpos - tmpBuf) - strlen("\r\n\r\n"));
    while ((n = recv(sock[tmpNum], tmpBuf, sizeof(tmpBuf)-1, 0)) > 0)
    {
        //写入临时文件
        fileTmp.write(tmpBuf, n);
    }
    fileTmp.close();
}

int main()
{
    char buf[1024];
    memset(buf, 0, sizeof(buf));
    fstream file;

    WORD version(0);
    WSADATA wsadata;
    int socket_return(0);
    version = MAKEWORD(2,0);
    socket_return = WSAStartup(version,&wsadata);
    if (socket_return != 0)
    {
        return 0;
    }

    string strFirst="GET "+(string)url+" HTTP/1.1\r\nHost:"+(string)testHostName+":"+testPortChar+"\r\nRange:bytes=0-1\r\nConnection:Close\r\n\r\n";
    //发送http请求
    if(!sendHttpQuery(strFirst,0)){
        cout<<" http 请求发送失败! "<<endl;
        return 0;
    }

    //思路:将Content-Range: bytes 0-1/后面的值作为长度解析出来
    //再根据文件长度决定循环几次请求(后续实现分配几个线程下载)
    recv(sock[0], buf, sizeof(buf)-1, 0);

    cout<<buf<<endl;
    char *lengChar=getResponseLeng(buf);
    cout<<"长度为 :"+(string)lengChar<<endl;
    int lengInt=atoi(lengChar);
    //得到每次取1000长度总共需要循环的次数
    int getTimeInt=lengInt/getSize;

    int i=0;
    int ret[getTimeInt+1];
    pthread_t testthreadDownload[getTimeInt+1];

    //ofstream fBig;
    //fBig.open()

    while(i<=getTimeInt){
        //注意,使用sleep让线程得以创建,不然可能还没创建线程就已经销毁,考虑使用信号量

        ret[i]= pthread_create(&testthreadDownload[i],NULL,&downloadTmpFile,&i);
        Sleep(1000);
        if(ret[i]){
            cout<<"创建临时文件线程失败!"<<endl;
            return 1;
        }
        i++;
    }
    i=0;

    while(i<=getTimeInt){
       pthread_join(testthreadDownload[i],NULL);
       i++;
    }
   //
    Sleep(40000);
    i=0;
    //pthread_t testthreadFileAdd[getTimeInt+1];
    ofstream fileGolobal;
    ifstream fileTmp[getTimeInt+1];
    while(i<=getTimeInt){

        fileGolobal.open(dataFile,ios::app|ios::out|ios::binary);
        char tmpNumChar[5];
        sprintf(tmpNumChar,"%d",i);
        cout<<"打开文件"+tmpFile+(string)tmpNumChar<<endl;
        //if(fileTmp[i].is_open()){
        //    fileTmp[i].close();
        //}
        fileTmp[i].open(tmpFile+(string)tmpNumChar,ios::in|ios::binary);
        cout<<"文件打开成功!"+tmpFile+(string)tmpNumChar<<endl;
        getGolobalFile(fileGolobal,fileTmp[i]);
        fileGolobal.close();
        fileTmp[i].close();
        i++;
    }
    //file.close();
    return 0;
}

评论
成就一亿技术人!
拼手气红包6.0元
还能输入1000个字符
 
红包 添加红包
表情包 插入表情
 条评论被折叠 查看
添加红包

请填写红包祝福语或标题

红包个数最小为10个

红包金额最低5元

当前余额3.43前往充值 >
需支付:10.00
成就一亿技术人!
领取后你会自动成为博主和红包主的粉丝 规则
hope_wisdom
发出的红包
实付
使用余额支付
点击重新获取
扫码支付
钱包余额 0

抵扣说明:

1.余额是钱包充值的虚拟货币,按照1:1的比例进行支付金额的抵扣。
2.余额无法直接购买下载,可以购买VIP、付费专栏及课程。

余额充值