1 /* vi: set sw=4 ts=4: */
3 * wget - retrieve a file using HTTP
5 * Chip Rosenthal Covad Communications <chip@laserlink.net>
7 * Note: According to RFC2616 section 3.6.1, "All HTTP/1.1 applications MUST be
8 * able to receive and decode the "chunked" transfer-coding, and MUST ignore
9 * chunk-extension extensions they do not understand."
11 * This prevents this particular wget app from completely RFC compliant, and as
12 * such, prevents it from being used as a general purpose web browser... This
13 * is a design decision, since it makes the code smaller.
25 #include <sys/ioctl.h>
28 #include <sys/types.h>
30 #include <sys/socket.h>
31 #include <netinet/in.h>
32 #include <arpa/inet.h>
36 void parse_url(char *url, char **uri_host, int *uri_port, char **uri_path);
37 FILE *open_socket(char *host, int port);
38 char *gethdr(char *buf, size_t bufsiz, FILE *fp, int *istrunc);
39 void progressmeter(int flag);
41 /* Globals (can be accessed from signal handlers */
42 static off_t filesize = 0; /* content-length of the file */
43 #ifdef BB_FEATURE_STATUSBAR
44 static char *curfile; /* Name of current file being transferred. */
45 static struct timeval start; /* Time a transfer started. */
46 volatile unsigned long statbytes; /* Number of bytes transferred so far. */
47 /* For progressmeter() -- number of seconds before xfer considered "stalled" */
51 int wget_main(int argc, char **argv)
54 char *proxy, *proxy_host;
55 int uri_port, proxy_port;
59 FILE *sfp; /* socket to web server */
60 char *uri_host, *uri_path; /* parsed from command line url */
61 char *fname_out = NULL; /* where to direct output (-O) */
62 int do_continue = 0; /* continue a prev transfer (-c) */
63 long beg_range = 0L; /* range at which continue begins */
64 int got_clen = 0; /* got content-length: from server */
65 FILE *output; /* socket to web server */
66 int quiet_flag = FALSE; /* Be verry, verry quiet... */
71 while ((n = getopt(argc, argv, "cqO:")) != EOF) {
80 /* can't set fname_out to NULL if outputting to stdout, because
81 * this gets interpreted as the auto-gen output filename
82 * case below - tausq@debian.org
84 fname_out = (strcmp(optarg, "-") == 0 ? (char *)1 : optarg);
91 if (argc - optind != 1)
94 if (do_continue && !fname_out)
95 error_msg_and_die("cannot specify continue (-c) without a filename (-O)\n");
98 * Use the proxy if necessary.
100 if ((proxy = getenv("http_proxy")) != NULL) {
101 proxy = xstrdup(proxy);
102 parse_url(proxy, &proxy_host, &proxy_port, &uri_path);
103 parse_url(argv[optind], &uri_host, &uri_port, &uri_path);
106 * Parse url into components.
108 parse_url(argv[optind], &uri_host, &uri_port, &uri_path);
113 /* Guess an output filename */
116 #ifdef BB_FEATURE_STATUSBAR
119 get_last_path_component(uri_path);
120 if (fname_out==NULL || strlen(fname_out)<1) {
122 #ifdef BB_FEATURE_STATUSBAR
127 #ifdef BB_FEATURE_STATUSBAR
129 curfile=argv[optind];
135 * Open socket to server.
137 sfp = open_socket(proxy_host, proxy_port);
139 /* Make the assumption that if the file already exists
140 * on disk that the intention is to continue downloading
141 * a previously aborted download -Erik */
142 if (stat(fname_out, &sbuf) == 0) {
147 * Open the output file stream.
149 if (fname_out != (char *)1) {
150 if ( (output=fopen(fname_out, (do_continue ? "a" : "w")))
152 perror_msg_and_die("fopen(%s)", fname_out);
158 * Determine where to start transfer.
161 if (fstat(fileno(output), &sbuf) < 0)
162 error_msg_and_die("fstat()");
163 if (sbuf.st_size > 0)
164 beg_range = sbuf.st_size;
172 fprintf(sfp, "GET http://%s:%d/%s HTTP/1.1\r\n",
173 uri_host, uri_port, uri_path);
174 fprintf(sfp, "Host: %s\r\nUser-Agent: Wget\r\n", uri_host);
176 fprintf(sfp, "Range: bytes=%ld-\r\n", beg_range);
177 fprintf(sfp,"Connection: close\r\n\r\n");
180 * Retrieve HTTP response line and check for "200" status code.
182 if (fgets(buf, sizeof(buf), sfp) == NULL) {
183 error_msg_and_die("no response from server\n");
185 for (s = buf ; *s != '\0' && !isspace(*s) ; ++s)
187 for ( ; isspace(*s) ; ++s)
193 error_msg_and_die("server does not support ranges\n");
199 error_msg_and_die("server returned error: %s", buf);
203 * Retrieve HTTP headers.
205 while ((s = gethdr(buf, sizeof(buf), sfp, &n)) != NULL) {
206 if (strcmp(buf, "content-length") == 0) {
211 if (strcmp(buf, "transfer-encoding") == 0) {
212 error_msg_and_die("server wants to do %s transfer encoding\n", s);
218 * Retrieve HTTP body.
220 #ifdef BB_FEATURE_STATUSBAR
222 if (quiet_flag==FALSE)
225 while (filesize > 0 && (n = fread(buf, 1, sizeof(buf), sfp)) > 0) {
226 fwrite(buf, 1, n, output);
227 #ifdef BB_FEATURE_STATUSBAR
229 if (quiet_flag==FALSE)
235 if (n == 0 && ferror(sfp))
236 perror_msg_and_die("network read error");
242 void parse_url(char *url, char **uri_host, int *uri_port, char **uri_path)
245 static char *defaultpath = "/";
249 if (strncmp(url, "http://", 7) != 0)
250 error_msg_and_die("not an http url: %s\n", url);
252 /* pull the host portion to the front of the buffer */
253 for (s = url, h = url+7 ; *h != '/' && *h != 0; ++h) {
255 *uri_port = atoi(h+1);
262 if (*h == 0) h = defaultpath;
267 if (!strcmp( *uri_host, *uri_path))
268 *uri_path = defaultpath;
272 FILE *open_socket(char *host, int port)
274 struct sockaddr_in sin;
279 memzero(&sin, sizeof(sin));
280 sin.sin_family = AF_INET;
281 if ((hp = (struct hostent *) gethostbyname(host)) == NULL)
282 error_msg_and_die("cannot resolve %s\n", host);
283 memcpy(&sin.sin_addr, hp->h_addr_list[0], hp->h_length);
284 sin.sin_port = htons(port);
287 * Get the server onto a stdio stream.
289 if ((fd = socket(AF_INET, SOCK_STREAM, 0)) < 0)
290 perror_msg_and_die("socket()");
291 if (connect(fd, (struct sockaddr *) &sin, sizeof(sin)) < 0)
292 perror_msg_and_die("connect(%s)", host);
293 if ((fp = fdopen(fd, "r+")) == NULL)
294 perror_msg_and_die("fdopen()");
300 char *gethdr(char *buf, size_t bufsiz, FILE *fp, int *istrunc)
307 /* retrieve header line */
308 if (fgets(buf, bufsiz, fp) == NULL)
311 /* see if we are at the end of the headers */
312 for (s = buf ; *s == '\r' ; ++s)
317 /* convert the header name to lower case */
318 for (s = buf ; isalnum(*s) || *s == '-' ; ++s)
321 /* verify we are at the end of the header name */
323 error_msg_and_die("bad header line: %s\n", buf);
325 /* locate the start of the header value */
326 for (*s++ = '\0' ; *s == ' ' || *s == '\t' ; ++s)
330 /* locate the end of header */
331 while (*s != '\0' && *s != '\r' && *s != '\n')
334 /* end of header found */
340 /* Rats! The buffer isn't big enough to hold the entire header value. */
341 while (c = getc(fp), c != EOF && c != '\n')
347 #ifdef BB_FEATURE_STATUSBAR
348 /* Stuff below is from BSD rcp util.c, as added to openshh.
349 * Original copyright notice is retained at the end of this file.
357 struct winsize winsize;
359 if (ioctl(fileno(stdout), TIOCGWINSZ, &winsize) != -1)
360 return (winsize.ws_col ? winsize.ws_col : 80);
366 updateprogressmeter(int ignore)
368 int save_errno = errno;
377 struct itimerval itv;
379 itv.it_value.tv_sec = wait;
380 itv.it_value.tv_usec = 0;
381 itv.it_interval = itv.it_value;
382 setitimer(ITIMER_REAL, &itv, NULL);
387 progressmeter(int flag)
389 static const char prefixes[] = " KMGTP";
390 static struct timeval lastupdate;
391 static off_t lastsize;
392 struct timeval now, td, wait;
393 off_t cursize, abbrevsize;
395 int ratio, barlength, i, remaining;
399 (void) gettimeofday(&start, (struct timezone *) 0);
404 (void) gettimeofday(&now, (struct timezone *) 0);
407 ratio = 100.0 * cursize / filesize;
408 ratio = MAX(ratio, 0);
409 ratio = MIN(ratio, 100);
413 snprintf(buf, sizeof(buf), "\r%-20.20s %3d%% ", curfile, ratio);
415 barlength = getttywidth() - 51;
417 i = barlength * ratio / 100;
418 snprintf(buf + strlen(buf), sizeof(buf) - strlen(buf),
420 "*****************************************************************************"
421 "*****************************************************************************",
425 abbrevsize = cursize;
426 while (abbrevsize >= 100000 && i < sizeof(prefixes)) {
430 snprintf(buf + strlen(buf), sizeof(buf) - strlen(buf), " %5d %c%c ",
431 (int) abbrevsize, prefixes[i], prefixes[i] == ' ' ? ' ' :
434 timersub(&now, &lastupdate, &wait);
435 if (cursize > lastsize) {
438 if (wait.tv_sec >= STALLTIME) {
439 start.tv_sec += wait.tv_sec;
440 start.tv_usec += wait.tv_usec;
444 timersub(&now, &start, &td);
445 elapsed = td.tv_sec + (td.tv_usec / 1000000.0);
447 if (statbytes <= 0 || elapsed <= 0.0 || cursize > filesize) {
448 snprintf(buf + strlen(buf), sizeof(buf) - strlen(buf),
450 } else if (wait.tv_sec >= STALLTIME) {
451 snprintf(buf + strlen(buf), sizeof(buf) - strlen(buf),
454 remaining = (int) (filesize / (statbytes / elapsed) - elapsed);
455 i = remaining / 3600;
457 snprintf(buf + strlen(buf), sizeof(buf) - strlen(buf),
460 snprintf(buf + strlen(buf), sizeof(buf) - strlen(buf),
462 i = remaining % 3600;
463 snprintf(buf + strlen(buf), sizeof(buf) - strlen(buf),
464 "%02d:%02d ETA", i / 60, i % 60);
466 write(fileno(stderr), buf, strlen(buf));
470 sa.sa_handler = updateprogressmeter;
471 sigemptyset(&sa.sa_mask);
472 sa.sa_flags = SA_RESTART;
473 sigaction(SIGALRM, &sa, NULL);
475 } else if (flag == 1) {
482 /* Original copyright notice which applies to the BB_FEATURE_STATUSBAR stuff,
483 * much of which was blatently stolen from openssh. */
486 * Copyright (c) 1992, 1993
487 * The Regents of the University of California. All rights reserved.
489 * Redistribution and use in source and binary forms, with or without
490 * modification, are permitted provided that the following conditions
492 * 1. Redistributions of source code must retain the above copyright
493 * notice, this list of conditions and the following disclaimer.
494 * 2. Redistributions in binary form must reproduce the above copyright
495 * notice, this list of conditions and the following disclaimer in the
496 * documentation and/or other materials provided with the distribution.
498 * 3. <BSD Advertising Clause omitted per the July 22, 1999 licensing change
499 * ftp://ftp.cs.berkeley.edu/pub/4bsd/README.Impt.License.Change>
501 * 4. Neither the name of the University nor the names of its contributors
502 * may be used to endorse or promote products derived from this software
503 * without specific prior written permission.
505 * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
506 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
507 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
508 * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
509 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
510 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
511 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
512 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
513 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
514 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
517 * $Id: wget.c,v 1.13 2000/12/09 16:55:35 andersen Exp $
524 c-file-style: "linux"