wget

#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#include <netdb.h>
#include <sys/types.h>
#include <netinet/in.h>
#include <sys/socket.h>

#include <fcntl.h>
#include <libgen.h>

char *gethostname(char *url)
{
    if(strstr(url, "http://"))
    {
        url += strlen("http://");
    }
    if(strstr(url, "https://"))
    {
        url += strlen("https://");
    }
    if(strchr(url, '/'))
    {
        strchr(url, '/')[0] = '\0';
    }

    return url;
}

char *getfilename(char *file)
{
    if(file[0] == '\0')
        strcpy(file, "test.txt");
    if(file[0] == '/' && file[1] == '\0')
        strcpy(file, "test.txt");

    return file;
}


#define PORT 80

int connect_to_ip(char *host)
{
    int sock;
    int sockfd;
    struct hostent *he;
    struct sockaddr_in their_addr;


    if((he = gethostbyname(host)) == NULL)
    {
        perror("gethostbyname(): fail: ");
        exit(1);
    }
    else
        printf("Client-The remote host is: %s\n", host);

    if((sockfd = socket(AF_INET, SOCK_STREAM, 0)) == -1)
    {
        perror("socket()");
        exit(1);
    }
    else
        printf("Client-The socket() sockfd is OK...\n");

    their_addr.sin_family = AF_INET;
    printf("Server-Using %s and port %d...\n", host, PORT);
    their_addr.sin_port = htons(PORT);
    their_addr.sin_addr = *((struct in_addr *)he->h_addr);
    memset(&(their_addr.sin_zero), '\0', 8);

    if(connect(sockfd, (struct sockaddr *)&their_addr, sizeof(struct sockaddr)) == -1)
    {
        perror("connect()");
        exit(1);
    }
    else
        printf("Client-The connect() is OK...\n");

    return sockfd;
}

char *gethdr(char *buff, int size, FILE *sfp)
{
    char *s;
    memset((char*)buff, 0, size);

    fgets(buff, size, sfp);

    for(s=buff; *s == '\r'; ++s) continue;
    if(*s == '\n') return NULL;

    /* locate the end of header */
    while (*s && *s != '\r' && *s != '\n')
        ++s;

    *s = '\0';

    return buff;
}


#define isspace(s) (*s == ' ')

char *skip_whitespace(char *str)
{
    while(isspace(str)) ++str;

    return str;
}

char *skip_non_whitespace(char *str)
{
    while(*str && !isspace(str)) ++str;

    return str;
}


void getfile(int sockfd, char *host, char *file)
{
    char cmd[1024] = {};
    char buff[512] = {};
    char *fname;
    int fd, len=0;
    int hack = 0;
    FILE *sfp;
    char *str;

    /* open outpt file */
    fname = basename(file);
    fd = open(fname, O_CREAT|O_RDWR, 0755);

    sprintf(cmd, "GET %s HTTP/1.1\r\n", file);
    sprintf(cmd, "%sHost: %s\r\n", cmd, host);
    sprintf(cmd, "%s\r\n", cmd);

    write(sockfd, cmd, strlen(cmd));
    printf("cmd: %s\n", cmd);

    /* skip the HEADER */
    /* START */
    sfp = fdopen(sockfd, "r+");
    gethdr(buff, 20, sfp);

    str = buff;
    str = skip_non_whitespace(str);
    str = skip_whitespace(str);
    printf("%s, status: %d\n", buff, atoi(str));

    while(gethdr(buff, sizeof(buff), sfp) != NULL)
        printf("%s\n", buff);

    fflush(sfp);
    /* END */

    while(1)
    {
        memset((char*)buff, 0, sizeof(buff));
        len=read(sockfd, buff, sizeof(buff));

        write(fd, buff, len);

        if(len <= 0) break;
    }

    close(fd);
    close(sockfd);

    return;
}

int main(int argc, char *argv[])
{
    char *host, *file;
    char url[256] = {};
    int sockfd;

    if(argc < 2)
    {
        printf("usage: %s \n", argv[0]);
        exit(0);
    }

    strcpy(url, argv[1]);
    host = gethostname(url);
    file = argv[1];
    file = getfilename(&file[host+strlen(host)-url]);
    printf("url: %s, file: %s\n", url, file);

    sockfd = connect_to_ip(host);
    getfile(sockfd, host, file);

    printf("\nsuccessfully written to \"%s\"\n\n",  basename(file));

    return 0;
}

gcc -o wget wget.c
./wget http://www.dhpc.adelaide.edu.au/education/dhpc/lab/MPI/hello.c
./wget www.physics.arizona.edu/~physreu/dox/cprogs/day1/hello.c