socket http文件下载器c语言实现

socket是网络编程的基石, 本文介绍如何使用c语言使用socket实现一个http文件下载器.

下载分为以下几个过程

解析出下载地址中的域名和文件名

通过域名获取服务器的IP地址

与目标服务器建立连接

构建http请求头并将其发送到服务器

等待服务器响应然后接收响应头

解析响应头, 判断返回码, 分离开响应头, 并且响应的正文内容以字节形式写入文件, 正文内容与头部用两个\n\r分开

具体实现完全可以通过代码看明白, 只需要看main函数和download函数即可, 其他函数都不是核心

完整的实现(Linux平台)

一下代码是单线程执行, 所以我把线程加进去感觉有点多此一举, 算是描述多线程下载吧, 这里补上没有使用线程的代码并能显示下载速度的代码 https://github.com/xanarry/http-downloader/blob/master/http_downloader.c

#include <stdio.h>
#include <string.h>
#include <sys/socket.h>
#include <arpa/inet.h>
#include <fcntl.h>
#include <unistd.h>
#include <netdb.h>
#include <stdlib.h>
#include <sys/types.h>
#include <sys/stat.h>
#include <pthread.h>
#include <sys/time.h>

struct resp_header//保持相应头信息
{
int status_code;//HTTP/1.1 '200' OK
char content_type[128];//Content-Type: application/gzip
long content_length;//Content-Length: 11683079
char file_name[256];
};

struct resp_header resp;//全剧变量以便在多个进程中使用

void parse_url(const char *url, char *domain, int *port, char *file_name)
{
/*通过url解析出域名, 端口, 以及文件名*/
int j = 0;
int start = 0;
*port = 80;
char *patterns[] = {"http://", "https://", NULL};

for (int i = 0; patterns[i]; i++)
if (strncmp(url, patterns[i], strlen(patterns[i])) == 0)
start = strlen(patterns[i]);

//解析域名, 这里处理时域名后面的端口号会保留
for (int i = start; url[i] != '/' && url[i] != '\0'; i++, j++)
domain[j] = url[i];
domain[j] = '\0';

//解析端口号, 如果没有, 那么设置端口为80
char *pos = strstr(domain, ":");
if (pos)
sscanf(pos, ":%d", port);

//删除域名端口号

for (int i = 0; i < (int)strlen(domain); i++)
{
if (domain[i] == ':')
{
domain[i] = '\0';
break;
}
}

//获取下载文件名
j = 0;
for (int i = start; url[i] != '\0'; i++)
{
if (url[i] == '/')
{
if (i !=  strlen(url) - 1)
j = 0;
continue;
}
else
file_name[j++] = url[i];
}
file_name[j] = '\0';
}

struct resp_header get_resp_header(const char *response)
{
/*获取响应头的信息*/
struct resp_header resp;

char *pos = strstr(response, "HTTP/");
if (pos)
sscanf(pos, "%*s %d", &resp.status_code);//返回状态码

pos = strstr(response, "Content-Type:");//返回内容类型
if (pos)
sscanf(pos, "%*s %s", resp.content_type);

pos = strstr(response, "Content-Length:");//内容的长度(字节)
if (pos)
sscanf(pos, "%*s %ld", &resp.content_length);

return resp;
}

void get_ip_addr(char *domain, char *ip_addr)
{
/*通过域名得到相应的ip地址*/
struct hostent *host = gethostbyname(domain);
if (!host)
{
ip_addr = NULL;
return;
}

for (int i = 0; host->h_addr_list[i]; i++)
{
strcpy(ip_addr, inet_ntoa( * (struct in_addr*) host->h_addr_list[i]));
break;
}
}

void progressBar(long cur_size, long total_size)
{
/*用于显示下载进度条*/
float percent = (float) cur_size / total_size;
const int numTotal = 50;
int numShow = (int)(numTotal * percent);

if (numShow == 0)
numShow = 1;

if (numShow > numTotal)
numShow = numTotal;

char sign[51] = {0};
memset(sign, '=', numTotal);

printf("\r%.2f%%\t[%-*.*s] %.2f/%.2fMB", percent * 100, numTotal, numShow, sign, cur_size / 1024.0 / 1024.0, total_size / 1024.0 / 1024.0);
fflush(stdout);

if (numShow == numTotal)
printf("\n");
}

void * download(void * socket_d)
{
/*下载文件函数, 放在线程中执行*/
int client_socket = *(int *) socket_d;
int length = 0;
int mem_size = 4096;//mem_size might be enlarge, so reset it
int buf_len = mem_size;//read 4k each time
int len;

//创建文件描述符
int fd = open(resp.file_name, O_CREAT | O_WRONLY, S_IRWXG | S_IRWXO | S_IRWXU);
if (fd < 0)
{
printf("Create file failed\n");
exit(0);
}

char *buf = (char *) malloc(mem_size * sizeof(char));

//从套接字中读取文件流
while ((len = read(client_socket, buf, buf_len)) != 0 && length < resp.content_length)
{
write(fd, buf, len);
length += len;
progressBar(length, resp.content_length);
}

if (length == resp.content_length)
printf("Download successful ^_^\n\n");
}

int main(int argc, char const *argv[])
{
/*
test url:
1. https://nodejs.org/dist/v4.2.3/node-v4.2.3-linux-x64.tar.gz 2. http://img.ivsky.com/img/tupian/pre/201312/04/nelumbo_nucifera-009.jpg */
char url[2048] = "127.0.0.1";
char domain[64] = {0};
char ip_addr[16] = {0};
int port = 80;
char file_name[256] = {0};

if (argc == 1)
{
printf("Input a valid URL please\n");
exit(0);
}
else
strcpy(url, argv[1]);

puts("1: Parsing url...");
parse_url(url, domain, &port, file_name);

if (argc == 3)
strcpy(file_name, argv[2]);

puts("2: Get ip address...");
get_ip_addr(domain, ip_addr);
if (strlen(ip_addr) == 0)
{
printf("can not get ip address\n");
return 0;
}

puts("\n>>>>Detail<<<<");
printf("URL: %s\n", url);
printf("DOMAIN: %s\n", domain);
printf("IP: %s\n", ip_addr);
printf("PORT: %d\n", port);
printf("FILENAME: %s\n\n", file_name);

//设置http请求头信息
char header[2048] = {0};
sprintf(header, \
"GET %s HTTP/1.1\r\n"\
"Accept:text/html,application/xhtml+xml,application/xml;q=0.9,image/webp,*/*;q=0.8\r\n"\
"User-Agent:Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537(KHTML, like Gecko) Chrome/47.0.2526Safari/537.36\r\n"\
"Host:%s\r\n"\
"Connection:close\r\n"\
"\r\n"\
,url, domain);

//printf("%s\n%d", header, (int) strlen(header));

//创建套接字
int client_socket = socket(AF_INET, SOCK_STREAM, IPPROTO_TCP);
if (client_socket < 0)
{
printf("invalid socket descriptor: %d\n", client_socket);
exit(-1);
}

//创建地址结构体
struct sockaddr_in addr;
memset(&addr, 0, sizeof(addr));
addr.sin_family = AF_INET;
addr.sin_addr.s_addr = inet_addr(ip_addr);
addr.sin_port = htons(port);

//连接服务器
puts("3: Connect server...");
int res = connect(client_socket, (struct sockaddr *) &addr, sizeof(addr));
if (res == -1)
{
printf("connect failed, return: %d\n", res);
exit(-1);
}

puts("4: Send request...");//向服务器发送下载请求
write(client_socket, header, strlen(header));

int mem_size = 4096;
int length = 0;
int len;
char *buf = (char *) malloc(mem_size * sizeof(char));
char *response = (char *) malloc(mem_size * sizeof(char));

//每次单个字符读取响应头信息, 仅仅读取的是响应部分的头部, 后面单独开线程下载
while ((len = read(client_socket, buf, 1)) != 0)
{
if (length + len > mem_size)
{
//动态内存申请, 因为无法确定响应头内容长度
mem_size *= 2;
char * temp = (char *) realloc(response, sizeof(char) * mem_size);
if (temp == NULL)
{
printf("realloc failed\n");
exit(-1);
}
response = temp;
}

buf[len] = '\0';
strcat(response, buf);

//找到响应头的头部信息, 两个"\n\r"为分割点
int flag = 0;
for (int i = strlen(response) - 1; response[i] == '\n' || response[i] == '\r'; i--, flag++);
if (flag == 4)
break;

length += len;
}

//printf("\n>>>>Response header:<<<<\n%s", response);

resp = get_resp_header(response);
strcpy(resp.file_name, file_name);

printf("5: Start thread to download...\n");
/*开新的线程下载文件*/
pthread_t download_thread;
pthread_create(&download_thread, NULL, download, (void *) &client_socket);
pthread_join(download_thread, NULL);
return 0;
}

执行过程, 注意使用了线程编译时要加 -lpthread

xanarry@ThinkPad:~/t$ gcc socket\ http下载器.c -std=c99 -lpthread
xanarry@ThinkPad:~/t$ ./a.out https://nodejs.org/dist/v4.2.3/node-v4.2.3-linux-x64.tar.gz 1: Parsing url...
2: Get ip address...

>>>>Detail<<<<
URL: https://nodejs.org/dist/v4.2.3/node-v4.2.3-linux-x64.tar.gz DOMAIN: nodejs.org
IP: 104.20.22.46
PORT: 80
FILENAME: node-v4.2.3-linux-x64.tar.gz

3: Connect server...
4: Send request...
5: Start thread to download...
3.11%   [==                                         ] 0.35/11.14MB

这仅仅实现的文件的单线程下载, http的请求头中有个Content-Length属性, 可以通过这个属性轻松实现断点下载和多线程下载. 如下是实现思路:

断点下载

在新的下载之前, 发送两次请求, 第一次请求中我们子读取响应头中的头部, 得到文件长度, 用该长度与本地的文件长度做比较, 如果长度不一致, 那么再发送第二次请求, 在第二次请求中我们给定下载内容的起点和终点, 返回的字节依次写入本地文件末尾就OK了

多线程下载

假如分3个线程下载, 文件的大小是100kb, 第一个线程下线0-33kb, 保存到临时文件part1, 第二线程下载34-66kb, 保存到临时文件part2, 第三个线程下载67-100kb, 保存到文件part3, 最后再把三个临时文件按顺序合并到一起就完成下载了