linux 运行socket文件下载,Linux下使用Socket实现http文件下载

// 转载自 http://blog.csdn.net/xiongyangg/article/details/50767482

// g++ download_demo.cpp -lpthread

#include

#include

#include

#include

#include

#include

#include

#include

#include

#include

#include

#include

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;

/*通过url解析出域名, 端口, 以及文件名*/

void parse_url(const char *url, char *domain, int *port, char *file_name){

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]);

}

}

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

domain[j] = url[i];

}

domain[j] = '\0';

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

if (pos){

sscanf(pos, ":%d", port);

}

for (int i = 0; i < (int)strlen(domain); i++){//删除域名端口号

if (domain[i] == ':'){

domain[i] = '\0';

break;

}

}

j = 0;

for (int i = start; url[i] != '\0'; i++){ //获取下载文件名

if (url[i] == '/'){

if (i != strlen(url) - 1){

j = 0;

}

continue;

}else{

file_name[j++] = url[i];

}

}

file_name[j] = '\0';

}

/*获取响应头的信息*/

struct resp_header get_resp_header(const char *response){

struct resp_header resp;

char *pos = (char*)strstr(response, "HTTP/");

if (pos)

sscanf(pos, "%*s %d", &resp.status_code);//返回状态码

pos = (char*)strstr(response, "Content-Type:");//返回内容类型

if (pos)

sscanf(pos, "%*s %s", resp.content_type);

pos = (char*)strstr(response, "Content-Length:");//内容的长度(字节)

if (pos)

sscanf(pos, "%*s %ld", &resp.content_length);

return resp;

}

// 向DNS服务器请求IP地址

void get_ip_addr(char *domain, char *ip_addr){

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;

}

}

/*用于显示下载进度条*/

void progressBar(long cur_size, long total_size){

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);

}

// client_socket 指向服务器下载

// fd 指向本地文件保存

void * download(void * socket_d){

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;

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));

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){

break;

}

}

if (length == resp.content_length){

printf("\nDownload successful ^_^\n\n");

}

}

int main(int argc, char const *argv[]){

char url[2048] = "http://seopic.699pic.com/photo/50010/8515.jpg_wh1200.jpg"; // 0.53M

char domain[64] = {0};

char ip_addr[16] = {0};

int port = 80;

char file_name[256] = {0};

char header[2048] = {0};

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);

sprintf(header, \

"GET %s HTTP/1.1\r\n"\

"User-Agent: Mozilla/4.0 (compatible; MSIE 5.00; Windows 98)\r\n"\

"Accept: */*\r\n"\

"Host:%s\r\n"\

"\r\n"\

,url, domain);

printf("---------------------------\n%s", header);

printf("---------------------------\n\n\n");

//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++){

// nothing todo

}

if (flag == 4){

break;

}

length += len;

}

printf("---------------------------\n%s", response);

printf("---------------------------\n\n\n");

resp = get_resp_header(response);

printf("(%d)status_code=%d\n", __LINE__, resp.status_code);

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;

}

运行结果

$ ./a.out

1: Parsing url...

2: Get ip address...

>>>>Detail<<<<

URL: http://seopic.699pic.com/photo/50010/8515.jpg_wh1200.jpg

DOMAIN: seopic.699pic.com

IP: 119.38.141.130

PORT: 80

FILENAME: 8515.jpg_wh1200.jpg

---------------------------

GET http://seopic.699pic.com/photo/50010/8515.jpg_wh1200.jpg HTTP/1.1

User-Agent: Mozilla/4.0 (compatible; MSIE 5.00; Windows 98)

Accept: */*

Host:seopic.699pic.com

---------------------------

3: Connect server...

4: Send request...

---------------------------

HTTP/1.1 200 OK

Server: marco/1.6

Date: Thu, 17 Aug 2017 08:51:00 GMT

Content-Type: image/jpeg

Content-Length: 560437

Connection: keep-alive

X-Request-Id: 96aaeffc8c0ece839ba5495988d22dc5; 1f6cb296b097a0c7ec2195a20b484587

X-Source: U/304

ETag: "82e871eb8d245fef907c9e5ef8cd8809"

X-Slice-Complete-Length: 560437

Last-Modified: Thu, 06 Apr 2017 12:59:05 GMT

X-Slice-Size: 65536

Expires: Wed, 23 Aug 2017 17:02:27 GMT

Cache-Control: max-age=691200

Accept-Ranges: bytes

Age: 393184

Via: T.2424.H.1, V.mix-gd-can-008, T.141134.R.1, M.cun-gd-zhs-134

---------------------------

(242)status_code=200

5: Start thread to download...

100.00% [==================================================] 0.53/0.53MB

Download successful ^_^

主要注意两个点。

1 组织HTTP协议的应用层数据包发起请求。

2 利用服务器返回的数据格式中连续两次\r\n解析出头部信息(包含文件大小)和文件原始数据(字节流)；