国产精品久久久久7777婷婷,久久福利毛片,欧美三级乱码

天下 — Thu, 13 Jun 2013 08:29:00 GMT

#include "stdafx.h"

int _tmain(int argc, _TCHAR* argv[])
{
    WORD wVersionRequested;
    WSADATA wsaData;
    int err;
    int so_broadcast=1;
    int ret;

    wVersionRequested = MAKEWORD( 2, 2 );
    err = WSAStartup( wVersionRequested, &wsaData );
    if ( err != 0 ) {
        return 1;
    }
    SOCKET fd = socket(AF_INET, SOCK_DGRAM, IPPROTO_UDP);
    if (fd == INVALID_SOCKET) {
        printf("Error at socket(): %ld\n", WSAGetLastError() );
        WSACleanup();
        return 1;
    }

    err = setsockopt(fd,SOL_SOCKET,SO_BROADCAST,(char*)&so_broadcast,sizeof(so_broadcast));
    if ( err != 0 ) {
        printf("Error at setsockopt(): %ld\n", WSAGetLastError() );
        WSACleanup();
        return 1;
    }


    char buf[256];
    sprintf(buf,"%d",12345678);
    sockaddr_in addr;
    addr.sin_family = AF_INET;

    //受限�q�播地址,需要设�|�调用setsockopt(),用SO_BROADCAST选项!
    //addr.sin_addr.s_addr = inet_addr( "255.255.255.255");

    //直接�q�播地址
    addr.sin_addr.s_addr = inet_addr( "192.168.1.255");
    addr.sin_port = htons( 5005);

    ret = sendto(fd,buf,strlen(buf)+1,0,(sockaddr*)&addr,sizeof(addr));
    printf("sended len=%d \n",ret);
    if (ret==-1){
        err = WSAGetLastError();
        printf("err=%d \n",err);
    }

    WSACleanup();
    return 0;
}

天下 2013-06-13 16:29 发表评论

天下 — Tue, 09 Oct 2012 09:22:00 GMT

摘要: Code highlighting produced by Actipro CodeHighlighter (freeware)http://www.CodeHighlighter.com/-->//获取�q�播地址//在udp接收�q�播�?如果有多块网�?会重复收�?//所以udp发送广播时,最好指定直接的�q�播地址,不要用受限的�q�播地址255.255.255.255unsigned int&... 阅读全文

天下 2012-10-09 17:22 发表评论

getaddrinfo函数

天下 — Tue, 15 May 2012 08:57:00 GMT

#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#include <vector>

#ifdef WIN32
    #include <WinSock2.h>
    #include <ws2tcpip.h>
    #pragma comment(lib,"Ws2_32.lib")
    //end WIN32
#else //Linux
    #include <unistd.h>
    #include <sys/socket.h>
    #include <netinet/in.h>
    #include <arpa/inet.h>
    #include <netdb.h>
#endif //end Linux

typedef std::vector<struct in_addr> array_in_addr;
int GetHostsByName( const char *host_name,array_in_addr* pArrayaddr)
{
    int n=0;
    struct addrinfo hints,*res;
    struct addrinfo* p_addrinfo = NULL;

    memset( &hints,0,sizeof(struct addrinfo) );
    hints.ai_flags = AI_CANONNAME;   //客户端一般就用这�?nbsp;AI_CANONNAME,服务端徏议用:AI_PASSIVE
    hints.ai_family = AF_INET;        //AF_INET:IPv4 ,AF_INET6:IPv6

    hints.ai_socktype = SOCK_STREAM; //IPPROTO_IP 可能会��生多个重复IP
    sockaddr_in* p_sockaddr_in;
    if( (n = getaddrinfo( host_name,NULL, &hints, &res) ) != 0)
       return 0;

    printf("domain:%s \r\n",host_name);
    p_addrinfo = res;
    while(p_addrinfo!=NULL)
    {
        pArrayaddr->push_back( ((sockaddr_in*)p_addrinfo->ai_addr)->sin_addr);
        printf("ip%d: %s \r\n",pArrayaddr->size(),inet_ntoa(((sockaddr_in*)p_addrinfo->ai_addr)->sin_addr));
        p_addrinfo = p_addrinfo->ai_next;
    }
    freeaddrinfo(res);
    return pArrayaddr->size();
}

int main(int argc, char *argv[])
{
    int nRet;
    const char* pszHost = NULL;
    array_in_addr inaddrs;
    if (argc != 2) {
        printf("Usage:%s  \r\n",argv[0]);
        return -1;
    }
    pszHost = argv[1];

#ifdef WIN32
    WSADATA wsaData;
    int err = WSAStartup( 0x0202, &wsaData );
    if ( err != 0 )  {
        printf("socket init(WSAStartup) failure \r\n");
        return 1;
    }
#endif
    GetHostsByName(pszHost,&inaddrs);

#ifdef WIN32
    WSACleanup();
#endif
    return 0;
}

参见�|�址:

http://baike.baidu.com/view/6757218.htm

天下 2012-05-15 16:57 发表评论

socket函数send和recv函数

天下 — Fri, 27 Apr 2012 01:14:00 GMT

摘要: Code highlighting produced by Actipro CodeHighlighter (freeware)http://www.CodeHighlighter.com/-->在发送端�Q�一�ơ发�?092个字节，在接收端�Q�一�ơ接�?092个字节，但是在接收端�Q�偶��会出现 socket.receive 接收不全的情�?nbsp;�Q�ret = ... 阅读全文

天下 2012-04-27 09:14 发表评论

天下 — Fri, 06 Jan 2012 09:03:00 GMT

TCP与UDP的不同接包处理方�?br />

1.UDP发包的问�?br />问：udp 发送两�ơ数据，�W�一��?nbsp;100字节 �Q�第二次200字节�Q?nbsp;接包方一�ơrecvfrom( 1000 ), 收到�?nbsp;100�Q�还�?00�Q�还�?00�Q?br />�{�：UDP是数据报文协议，是以数据包方式，所以每�ơ可以接�?00�Q?/span>200�Q�在理想情况下，�W�一�ơ是无论recvfrom多少都是接收�?00。当�Ӟ��可能�׃��|�络原因�Q�第二个包先到的话，有可能是200了。对可能会由于网�l�原因�ؕ序，所以可能先收到200�Q�所以自定义的udp协议包头里都要加上一个序列号�Q�标识发送与收包对应

2.TCP的发包问�?br />问：同样如果换成tcp, �W�一�ơ发�?nbsp;100字节 �Q�第二次发�?00字节�Q�recv( 1000 )会接收到多少�Q?br />�{�：tcp是流协议�Q�所以recv( 1000 )�Q�会收到300 tcp自己处理好了重传�Q�保证数据包的完整�?br />
3.有分片的情况下如下处�?br />问：如果MTU�?500�Q��用UDP发�?nbsp;2000�Q�那么recvfrom(2000)是收�?500�Q�还�?000?
�{�： �q�是接收2000�Q�数据分片由ip层处理了�Q�放到udp�q�是一个完整的包。接收到的包是由路由路径上最��的MTU来分片，注意转到UDP已经在是�l�装好的(�l�装出错的包会经crc校验出错而丢�?�Q�是一个完整的数据�?br />
4.分片后的处理
问：如果500那个片丢了怎么办？udp又没有重�?br />�{�：udp里有个crc��验，如果包不完整��׃��丢弃�Q�也不会通知是否接收成功�Q�所以UDP是不可靠的传输协议，而且TCP不存在这个问题，有自��q��重传机制。在内网来说�Q�UDP基本不会有丢包，可靠性还是有保障。当然如果是要求有时序性和高可靠性，�q�是走TCP�Q�不然就要自己提供重传和乱序处理( UDP内网发包处理量可以达 7w~10w/s )

5.不同�q�接到同一个端口的包处�?br />问：TCP
A -> C �?00
B -> C �?00
AB同时同一端口
C recv(1000) ,会收到多��？
�{�：A与C是一个tcp�q�接�Q�B与C又是另一个tcp�q�接�Q?nbsp;所以不同socket�Q�所以分开处理。每个socket有自��q��接收�~�冲和发送缓�?br />
6.什么是TCP�_�包

�׃��TCP是流协议�Q�对于一个socket的包�Q�如发�?nbsp;10AAAAABBBBB两次�Q�由于网�l�原因第一�ơ又分成两次发送， 10AAAAAB和BBBB�Q�如果接包的时候先��d��10(包长�?再读入后�l�数据，当接收得快，发送的慢时�Q�就会出现先接收�?nbsp;10AAAAAB,会解释错�?nbsp;,再接到到BBBB10AAAAABBBBB�Q�也解释错误的情��c��这��是TCP的粘包�?br /> 解决的办法TLV方式�Q�先接收包头�Q�在包头里指定包体长度来接收。设�|�包头包��查位�Q�如��空�?x2开��_��0x3�l�束来检查一个包是否完整�Q�。对于TCP来说�Q?/span>1�Q�不存在丢包�Q�错包，所以不会出现数据出�?nbsp;2�Q�如果包头检��错误，即�ؓ非法或者请求，直接重置卛_��
7.

For message-oriented sockets, data is extracted from the first enqueued message, up to the size of the buffer specified. If the datagram or message is larger than the buffer specified, the buffer is filled with the first part of the datagram, and recvfrom generates the error WSAEMSGSIZE. For unreliable protocols (for example, UDP) the excess data is lost.

天下 2012-01-06 17:03 发表评论

天下 — Wed, 12 Oct 2011 08:01:00 GMT

#pragma pack(1)
typedef struct _tag_COMM_PACK_REQUEST
{
    BYTE    header[2];    //2字节,协议�?/span>
    UINT64    requestID;        //8字节,��水�?/span>
    BYTE    entryType;    //1字节,加密�c�d��
    DWORD    cmd;        //4字节,命��o
    DWORD    dataLen;    //4字节,数据长度
    BYTE    checksum;    //1字节,校验�?/span>
    char    data[0];    //0字节,
}COMM_PACK_REQUEST;        //�?20字节

typedef struct _tag_COMM_PACK_RESPONSE
{
    BYTE    header[2];    //2字节,协议�?/span>
    UINT64    responseID;        //8字节,��水�?/span>
    BYTE    entryType;    //1字节,加密�c�d��
    DWORD    err;        //4字节,命��o
    DWORD    dataLen;    //4字节,数据长度
    BYTE    checksum;    //1字节,校验�?/span>
    char    data[0];    //0字节,
}COMM_PACK_RESPONSE;    //�?20字节
#pragma pack()

#define COMM_PACK_REQUEST_SIZE    sizeof(COMM_PACK_REQUEST)
#define COMM_PACK_RESPONSE_SIZE    sizeof(COMM_PACK_RESPONSE)

天下 2011-10-12 16:01 发表评论

Linux Socket��d��模式服务端Demo 61.cpp

天下 — Thu, 08 Sep 2011 08:06:00 GMT

// File: 61.cpp
//客户端代�?nbsp;62.cpp
#include <stdio.h>
#include <unistd.h>
#include <stdlib.h>
#include <string.h>
#include <strings.h>
#include <sys/types.h>
#include <sys/socket.h>
#include <netinet/in.h>
#include <arpa/inet.h>

#define PORT 1234   /* Port that will be opened */
#define BACKLOG 5   /* Number of allowed connections */
#define MAXDATASIZE 4096

void process_cli(int connectfd,struct sockaddr_in client);

int main()  {
int listenfd, connectfd; /* socket descriptors */
struct sockaddr_in server; /* server's address information */
struct sockaddr_in client; /* client's address information */
socklen_t sin_size;

/* Create TCP socket  */
if ((listenfd = socket(AF_INET, SOCK_STREAM, 0)) == -1)  {
  /* handle exception */
  perror("Creating socket failed.\n");
  exit(1);
}

int opt = SO_REUSEADDR;
setsockopt(listenfd, SOL_SOCKET, SO_REUSEADDR, &opt, sizeof(opt));

bzero(&server,sizeof(server));
server.sin_family=AF_INET;
server.sin_port=htons(PORT);
server.sin_addr.s_addr = htonl (INADDR_ANY);
if (bind(listenfd, (struct sockaddr *)&server, sizeof(struct sockaddr)) == -1) {
  /* handle exception */
  perror("Bind error.\n");
  exit(1);
}

if(listen(listenfd,BACKLOG) == -1){  /* calls listen() */
  perror("listen() error\n");
  exit(1);
}

sin_size=sizeof(struct sockaddr_in);
while (1) {
  if ((connectfd = accept(listenfd,(struct sockaddr *)&client,&sin_size))==-1) {
   perror("accept() error\n");
   exit(1);
  }
  process_cli(connectfd, client);
}

close(listenfd);
}

void process_cli(int connectfd, struct sockaddr_in client)
{
int num;
char recvbuf[MAXDATASIZE], sendbuf[MAXDATASIZE];

/* prints client's IP */
printf("%s:%d Client was connected\n",inet_ntoa(client.sin_addr), ntohs( client.sin_port) );

while (1) {
  num = recv(connectfd, recvbuf, MAXDATASIZE,0);
  if (num == 0) {
   close(connectfd);
   printf("%s:%d Client was disconnected \n",inet_ntoa(client.sin_addr), ntohs( client.sin_port) );
   break;
  }
  if (num == -1) {
   perror("recv() error\n");
   break;
  }
  printf("recvbuf[0]:%d,recvbuf[num-1]:%d,num:%d\n",recvbuf[0],recvbuf[num-1],num);
  recvbuf[num] = '\0';
  printf("Received message From Client : %s\n",recvbuf);
  for (int i = 0; i < num; i++) {
   sendbuf[i] = recvbuf[num-i-1];
   //printf("sendbuf[%d]:%d\n",i,sendbuf[i]);
  }
  //printf("sendbuf[0]:%d,sendbuf[num-1]:%d,num:%d\n",sendbuf[0],sendbuf[num-1],num);
  send(connectfd,sendbuf,num,0); /* send to the client welcome message */
}
close(connectfd); /*  close connectfd */
}

天下 2011-09-08 16:06 发表评论

Linux Socket子进�E�模式服务端Demo 63.cpp

天下 — Thu, 08 Sep 2011 08:06:00 GMT

// File: 63.cpp
//客户端代�?nbsp;62.cpp
#include <stdio.h>
#include <unistd.h>
#include <stdlib.h>
#include <string.h>
#include <strings.h>
#include <sys/types.h>
#include <sys/socket.h>
#include <netinet/in.h>
#include <arpa/inet.h>

#define PORT 1234   /* Port that will be opened */
#define BACKLOG 5   /* Number of allowed connections */
#define MAXDATASIZE 4096

void process_cli(int connectfd,struct sockaddr_in client);

int main()
{
int listenfd, connectfd; /* socket descriptors */
pid_t pid;
struct sockaddr_in server; /* server's address information */
struct sockaddr_in client; /* client's address information */
socklen_t sin_size;

/* Create TCP socket  */
if ((listenfd = socket(AF_INET, SOCK_STREAM, 0)) == -1) {
  /* handle exception */
  perror("Creating socket failed.\n");
  exit(1);
}

int opt = SO_REUSEADDR;
setsockopt(listenfd, SOL_SOCKET, SO_REUSEADDR, &opt, sizeof(opt));

bzero(&server,sizeof(server));
server.sin_family=AF_INET;
server.sin_port=htons(PORT);
server.sin_addr.s_addr = htonl(INADDR_ANY);
if (bind(listenfd, (struct sockaddr *)&server, sizeof(struct sockaddr)) == -1) {
  /* handle exception */
  perror("Bind error.\n");
  exit(1);
}

if(listen(listenfd,BACKLOG) == -1){  /* calls listen() */
  perror("listen() error\n");
  exit(1);
}

sin_size=sizeof(struct sockaddr_in);

while(1)
{
  /*accept connection */
  if ((connectfd = accept(listenfd,(struct sockaddr *)&client,&sin_size))==-1) {
   perror("accept() error\n");
   exit(1);
  }
  /*  Create child process to service client */
  if ((pid=fork())>0) {
   /* parent process */
   close(connectfd);
   continue;
  }
  else if (pid==0) {
   /*child process*/
   close(listenfd);
   process_cli(connectfd, client);
   exit(0);
  }
  else {
   perror("fork() error\n");
   exit(0);
  }
}
close(listenfd);   /* close listenfd */
}

void process_cli(int connectfd, struct sockaddr_in client)
{
int num;
char recvbuf[MAXDATASIZE], sendbuf[MAXDATASIZE];

/* prints client's IP */
printf("%s:%d Client was connected\n",inet_ntoa(client.sin_addr), ntohs( client.sin_port) );

while (1) {
  num = recv(connectfd, recvbuf, MAXDATASIZE,0);
  if (num == 0) {
   close(connectfd);
   printf("%s:%d Client was disconnected \n",inet_ntoa(client.sin_addr), ntohs( client.sin_port) );
   break;
  }
  if (num == -1) {
   perror("recv() error\n");
   break;
  }
  printf("recvbuf[0]:%d,recvbuf[num-1]:%d,num:%d\n",recvbuf[0],recvbuf[num-1],num);
  recvbuf[num] = '\0';
  printf("Received message From Client : %s\n",recvbuf);
  for (int i = 0; i < num; i++) {
   sendbuf[i] = recvbuf[num-i-1];
   //printf("sendbuf[%d]:%d\n",i,sendbuf[i]);
  }
  //printf("sendbuf[0]:%d,sendbuf[num-1]:%d,num:%d\n",sendbuf[0],sendbuf[num-1],num);
  send(connectfd,sendbuf,num,0); /* send to the client welcome message */
}
close(connectfd); /*  close connectfd */
}

天下 2011-09-08 16:06 发表评论

Linux Socket多线�E�模式服务端Demo 64.cpp

天下 — Thu, 08 Sep 2011 08:04:00 GMT

// File: 64.cpp
//g++ -o 64 64.cpp -lpthread
//客户端代�?nbsp;62.cpp
#include <stdio.h>
#include <unistd.h>
#include <stdlib.h>
#include <string.h>
#include <strings.h>
#include <sys/types.h>
#include <sys/socket.h>
#include <netinet/in.h>
#include <arpa/inet.h>
#include <pthread.h>

#define PORT 1234   /* Port that will be opened */
#define BACKLOG 5   /* Number of allowed connections */
#define MAXDATASIZE 4096

void process_cli(int connectfd,struct sockaddr_in client);

/* function to be executed by the new thread */
void* start_routine(void* arg);
struct  ARG  {
int connfd;
sockaddr_in client;
};

int main()
{
int listenfd, connectfd; /* socket descriptors */
pthread_t  thread;
ARG *arg;
struct sockaddr_in server; /* server's address information */
struct sockaddr_in client; /* client's address information */
socklen_t sin_size;

/* Create TCP socket */
if ((listenfd = socket(AF_INET, SOCK_STREAM, 0)) == -1) {
  /* handle exception */
  perror("Creating socket failed.\n");
  exit(1);
}

int opt = SO_REUSEADDR;
setsockopt(listenfd, SOL_SOCKET, SO_REUSEADDR, &opt, sizeof(opt));

bzero(&server,sizeof(server));
server.sin_family=AF_INET;
server.sin_port=htons(PORT);
server.sin_addr.s_addr = htonl(INADDR_ANY);
if (bind(listenfd, (struct sockaddr *)&server, sizeof(struct sockaddr)) == -1) {
  /* handle exception */
  perror("Bind error.");
  exit(1);
}

if(listen(listenfd,BACKLOG) == -1){  /* calls listen() */
  perror("listen() error\n");
  exit(1);
}

sin_size=sizeof(struct sockaddr_in);
while(1)
{
  /* Accept connection */
  if ((connectfd = accept(listenfd,(struct sockaddr *)&client,&sin_size))==-1) {
   perror("accept() error\n");
   exit(1);
  }
  /*  Create thread*/

  arg = new ARG;
  arg->connfd = connectfd;
  memcpy((void *)&arg->client, &client, sizeof(client));

  if (pthread_create(&thread, NULL, start_routine, (void*)arg)) {
   /* handle exception */
   perror("Pthread_create() error");
   exit(1);
  }
}
close(listenfd);   /* close listenfd */
}

void process_cli(int connectfd, struct sockaddr_in client)
{
int num;
char recvbuf[MAXDATASIZE], sendbuf[MAXDATASIZE];

/* prints client's IP */
printf("%s:%d Client was connected\n",inet_ntoa(client.sin_addr), ntohs( client.sin_port) );

while (1) {
  num = recv(connectfd, recvbuf, MAXDATASIZE,0);
  if (num == 0) {
   close(connectfd);
   printf("%s:%d Client was disconnected \n",inet_ntoa(client.sin_addr), ntohs( client.sin_port) );
   break;
  }
  if (num == -1) {
   perror("recv() error\n");
   break;
  }
  printf("recvbuf[0]:%d,recvbuf[num-1]:%d,num:%d\n",recvbuf[0],recvbuf[num-1],num);
  recvbuf[num] = '\0';
  printf("Received message From Client : %s\n",recvbuf);
  for (int i = 0; i < num; i++) {
   sendbuf[i] = recvbuf[num-i-1];
   //printf("sendbuf[%d]:%d\n",i,sendbuf[i]);
  }
  //printf("sendbuf[0]:%d,sendbuf[num-1]:%d,num:%d\n",sendbuf[0],sendbuf[num-1],num);
  send(connectfd,sendbuf,num,0); /* send to the client welcome message */
}
close(connectfd); /*  close connectfd */
}

void* start_routine(void* arg)
{
ARG *info;
info = (ARG *)arg;

/* handle client’s requirement */
process_cli(info->connfd, info->client);

delete (ARG*)arg;
pthread_exit(NULL);
}

天下 2011-09-08 16:04 发表评论

linux Socket多�\复用模式服务�?69.cpp

天下 — Thu, 08 Sep 2011 08:03:00 GMT

// File: 69.cpp
#include <iostream>
#include <map>
#include <stdio.h>
#include <unistd.h>
#include <stdlib.h>
#include <math.h>
#include <string.h>
#include <strings.h>
#include <sys/types.h>
#include <sys/socket.h>
#include <netinet/in.h>
#include <arpa/inet.h>
#include <bitset>
using namespace std;
#define PORT 1234   /* Port that will be opened */
#define BACKLOG 5   /* Number of allowed connections */
#define MAXDATASIZE 4096

typedef map<unsigned int,sockaddr_in> connects;
void process_cli(int sockfd,sockaddr_in addr,char* recvbuf, int len);

void reverser(char* szIn,char* szOut,unsigned int nLen)
{
for (int i=0;i<nLen;i++) {
  szOut[i] = szIn[nLen-i-1];
}
}

int main()
{
connects connect_fds;
int sock_fd;
ssize_t recv_size;
fd_set read_set;
fd_set all_set;
int listen_fd;
int connect_fd;
int ready_size;
socklen_t addr_size;
sockaddr_in srv_addr;
sockaddr_in addr;
unsigned int max_fd = 0;
char recv_buf[MAXDATASIZE];

if ((listen_fd = socket(AF_INET, SOCK_STREAM, 0)) == -1) {
  perror("Creating socket failed.");
  exit(1);
}

int opt = SO_REUSEADDR;
setsockopt(listen_fd, SOL_SOCKET, SO_REUSEADDR, &opt, sizeof(opt));

bzero(&srv_addr,sizeof(srv_addr));
srv_addr.sin_family=AF_INET;
srv_addr.sin_port=htons(PORT);
srv_addr.sin_addr.s_addr = htonl (INADDR_ANY);
if (bind(listen_fd, (struct sockaddr *)&srv_addr, sizeof(struct sockaddr)) == -1) {
  perror("Bind error.");
  exit(1);
}

if(listen(listen_fd,BACKLOG) == -1) {
  perror("listen() error\n");
  exit(1);
}
max_fd = listen_fd;
addr_size=sizeof(sockaddr_in);

FD_ZERO(&all_set);
FD_SET(listen_fd, &all_set);

connects::iterator iter;

printf("clients:(%u) ,current listen socket:%u\n",connect_fds.size(),listen_fd);
while(1)
{
  read_set = all_set;
  ready_size = select(max_fd+1, &read_set, NULL, NULL, NULL);
  if ( FD_ISSET(listen_fd,&read_set) ) {
   connect_fd = accept(listen_fd,(struct sockaddr *)&addr,&addr_size);
   if (connect_fd ==-1 ) {
    perror("accept() error\n");
    continue;
   }
   connect_fds[connect_fd]  = addr;
   max_fd = connect_fd;
   FD_SET(connect_fd, &all_set);
   printf("clients:(%u) ,%s:%d connected ,connect socket:%u\n",connect_fds.size(),inet_ntoa(addr.sin_addr),ntohs(addr.sin_port),connect_fd);
  }

  for ( iter = connect_fds.begin(); iter!=connect_fds.end(); iter++) {
   sock_fd = iter->first;
   addr = iter->second;
   if ( FD_ISSET(sock_fd,&read_set) ) {
    recv_size = recv(sock_fd, recv_buf, MAXDATASIZE,0);
    if ( recv_size == 0) {
     close(sock_fd);
     FD_CLR(sock_fd, &all_set);
     connect_fds.erase(sock_fd);
     printf("%s:%d disconnected\n",inet_ntoa(addr.sin_addr),ntohs(addr.sin_port) );
    }
    else {
     process_cli(sock_fd,addr,recv_buf,recv_size);
    }
   }
  }
}
close(listen_fd);
}

void process_cli(int sockfd,sockaddr_in addr,char* recvbuf, int len)
{
char sendbuf[MAXDATASIZE];
printf("Received %s:%d message: %s\n",inet_ntoa(addr.sin_addr),ntohs(addr.sin_port), recvbuf);
reverser(recvbuf,sendbuf,len);
printf("Send To  %s:%d message: %s\n",inet_ntoa(addr.sin_addr),ntohs(addr.sin_port), sendbuf);
send(sockfd,sendbuf,len,0);
}

天下 2011-09-08 16:03 发表评论

�?TCP socket ��用 Nagle ��法

天下 — Sun, 07 Aug 2011 08:30:00 GMT

      //�?nbsp;TCP socket ��用 Nagle ��法
　　/* Create new stream socket */
　　sock = socket ( AF_INET, SOCK_STREAM, 0 );

　　/* Disable the Nagle (TCP No Delay) algorithm */
　　flag = 1;
　　int ret = setsockopt ( sock, IPPROTO_TCP, TCP_NODELAY, (char *)&flag, sizeof(flag) );

　　if (ret == -1) {
　　   printf("Couldn't setsockopt(TCP_NODELAY)\n");
　　   exit(-1);
　　}

天下 2011-08-07 16:30 发表评论

天下 — Sun, 07 Aug 2011 03:20:00 GMT

#include "netinet/tcp.h"
int keepAlive    = 1;        //讑֮�KeepAlive
int keepIdle    = 30;        //开始首�ơKeepAlive探测前的TCP�I�闭旉��
int keepInterval    = 30;    //两次KeepAlive探测间的旉��间隔
int keepCount        = 3;    //判定断开前的KeepAlive探测�ơ数

if(setsockopt(sockfd,SOL_SOCKET,SO_KEEPALIVE,(void*)&keepAlive,sizeof(keepAlive)) == -1)
{
    printf("Socket Option setting(SO_KEEPALIVE) fail !\n");
    return FALSE;
}

if(setsockopt(sockfd,SOL_TCP,TCP_KEEPIDLE,(void *)&keepIdle,sizeof(keepIdle)) == -1)
{
    printf("Socket Option setting(SO_KEEPALIVE) fail !\n");
    return FALSE;
}

if(setsockopt(sockfd,SOL_TCP,TCP_KEEPINTVL,(void *)&keepInterval,sizeof(keepInterval)) == -1)
{
    printf("Socket Option setting(TCP_KEEPIDLE) fail !\n");
    return FALSE;
}

if(setsockopt(sockfd,SOL_TCP,TCP_KEEPCNT,(void *)&keepCount,sizeof(keepCount)) == -1)
{
    printf("Socket Option setting(TCP_KEEPCNT) fail !\n");
    return FALSE;
}

天下 2011-08-07 11:20 发表评论

字节��序�Q�高位优�?big-endian)和低位优�?little-endian)

天下 — Mon, 28 Feb 2011 05:07:00 GMT

字节��序是指占内存多于一个字节类型的数据在内存中的存��N��序，通常有小端、大端两�U�字节顺序�?/span>

��端字节序指低字节数据存攑֜�内存低地址处，高字节数据存攑֜�内存高地址处；

大端字节序是高字节数据存攑֜�低地址处，低字节数据存攑֜�高地址处�?/span>

��Z��X86�q�_��?/font>PC机是��端字节序的

�|�络字节�? MSB 高字节前存法 Most Significant Bit (Big Edian)

��L��字节�? LSB 低字节前存法 Lest Significant Bit (Little Edian)

    因�ؓ现行的计��机都是以八位一个字节�ؓ存储单位�Q�那么一�?6位的整数�Q�也��是C语言中的short�Q�在内存中可能有两种存储��序big-endian和litte-endian。考虑一个short整数0x3132(0x32是低位，0x31是高�?�Q�把它赋值给一个short变量�Q�那么它在内存中的存储可能有如下两种情况�Q?br />

可以做个实验
在Windows上下如下�E�序
#include <stdio.h>
#include <assert.h>

int main(void)
{
        short test;
        FILE* fp;

        test = 0x3132; /**//* (31ASIIC码的'1', 32ASIIC码的'2') */
        if ((fp = fopen("c:\\test.txt", "wb")) == NULL)
              assert(0);
        fwrite(&test, sizeof(short), 1, fp);
        fclose(fp);
        return 0;
}
    然后在C盘下打开test.txt文�g�Q�可以看见内�Ҏ��21�Q�而test�{�于0x3132�Q�可以明昄��看出来x86的字节顺序是低位在前。如果我们把�q�段同样的代码放�?big-endian)的机器上执行�Q�那么打出来的文件就�?2。这在本��Z��使用是没有问题的。但当你把这个文件从一个big-endian机器复制��C��个little-endian机器上时��出现问题了�?br />    如上�q�C��子，我们在big-endian的机器上创徏了这个test文�g�Q�把其复制到little-endian的机器上再用fread��d��一个short里面�Q�我们得到的��׃��再是0x3132而是0x3231了，�q�样��d��的数据就是错误的�Q�所以在两个字节��序不一��L��机器上传输数据时需要特别小心字节顺序，理解了字节顺序在可以帮助我们写出�U�L��行更高的代码�?br />

正因为有字节��序的差别，所以在�|�络传输的时候定义了所有字节顺序相关的数据都��用big-endian�Q�BSD的代码中定义了四个宏来理�Q?br />#define ntohs(n)      //�|�络字节��序��C��机字节顺�?nbsp;n代表net, h代表host, s代表short
#define htons(n)      //��L��字节��序到网�l�字节顺�?nbsp;n代表net, h代表host, s代表short
#define ntohl(n)      //�|�络字节��序��C��机字节顺�?nbsp;n代表net, h代表host, s代表long
#define htonl(n)      //��L��字节��序到网�l�字节顺�?nbsp;n代表net, h代表host, s代表long

天下 2011-02-28 13:07 发表评论

sockaddr �?sockaddr_in �c�d��指针转换分析

天下 — Tue, 11 Jan 2011 13:19:00 GMT

#include <iostream>
#include <afxwin.h>
#include <Winsock2.h>
#pragma comment(lib,"Ws2_32.lib")
int main(){
    sockaddr_in addrSock1;
    addrSock1.sin_family = AF_INET;
    addrSock1.sin_port   = htons(5555);
    addrSock1.sin_addr.s_addr = inet_addr("127.0.0.1");

    sockaddr* addrSock2;

    addrSock2 = (sockaddr*)&addrSock1;

    u_long uPort1 = htons(5555);
    u_long uPort2 = MAKEWORD(addrSock2->sa_data[0],addrSock2->sa_data[1]);
    return 0;
}

//注意 addSock2->sa_data[0],addSock2->sa_data[1]及MAKEWORD

天下 2011-01-11 21:19 发表评论

[原创]��装了个��Z��Libcurl的类,大家评评,��情拍砖.

天下 — Thu, 18 Nov 2010 06:15:00 GMT

摘要: [原创]��装了个��Z��Libcurl的类,大家评评,��情拍砖. Code highlighting produced by Actipro CodeHighlighter (freeware)http://www.CodeHighlighter.com/--> 1 #pragma once 2 #include &... 阅读全文

天下 2010-11-18 14:15 发表评论