2000-09-04 20:45:55 +05:30
|
|
|
/* vi: set sw=4 ts=4: */
|
|
|
|
/*
|
2001-04-06 03:15:54 +05:30
|
|
|
* wget - retrieve a file using HTTP or FTP
|
2000-09-04 20:45:55 +05:30
|
|
|
*
|
2000-11-15 04:59:24 +05:30
|
|
|
* Chip Rosenthal Covad Communications <chip@laserlink.net>
|
2010-08-16 23:44:46 +05:30
|
|
|
* Licensed under GPLv2, see file LICENSE in this source tree.
|
2010-08-08 06:21:20 +05:30
|
|
|
*
|
|
|
|
* Copyright (C) 2010 Bradley M. Kuhn <bkuhn@ebb.org>
|
2010-10-29 15:16:52 +05:30
|
|
|
* Kuhn's copyrights are licensed GPLv2-or-later. File as a whole remains GPLv2.
|
2000-09-04 20:45:55 +05:30
|
|
|
*/
|
2015-10-07 05:09:40 +05:30
|
|
|
//config:config WGET
|
2017-07-19 01:31:24 +05:30
|
|
|
//config: bool "wget (35 kb)"
|
2015-10-07 05:09:40 +05:30
|
|
|
//config: default y
|
|
|
|
//config: help
|
2017-07-21 13:20:55 +05:30
|
|
|
//config: wget is a utility for non-interactive download of files from HTTP
|
|
|
|
//config: and FTP servers.
|
2015-10-07 05:09:40 +05:30
|
|
|
//config:
|
2017-01-10 19:28:54 +05:30
|
|
|
//config:config FEATURE_WGET_LONG_OPTIONS
|
|
|
|
//config: bool "Enable long options"
|
|
|
|
//config: default y
|
|
|
|
//config: depends on WGET && LONG_OPTS
|
|
|
|
//config:
|
2015-10-07 05:09:40 +05:30
|
|
|
//config:config FEATURE_WGET_STATUSBAR
|
2017-01-10 19:28:54 +05:30
|
|
|
//config: bool "Enable progress bar (+2k)"
|
2015-10-07 05:09:40 +05:30
|
|
|
//config: default y
|
|
|
|
//config: depends on WGET
|
|
|
|
//config:
|
|
|
|
//config:config FEATURE_WGET_AUTHENTICATION
|
|
|
|
//config: bool "Enable HTTP authentication"
|
|
|
|
//config: default y
|
|
|
|
//config: depends on WGET
|
|
|
|
//config: help
|
2017-07-21 13:20:55 +05:30
|
|
|
//config: Support authenticated HTTP transfers.
|
2015-10-07 05:09:40 +05:30
|
|
|
//config:
|
|
|
|
//config:config FEATURE_WGET_TIMEOUT
|
|
|
|
//config: bool "Enable timeout option -T SEC"
|
|
|
|
//config: default y
|
|
|
|
//config: depends on WGET
|
|
|
|
//config: help
|
2017-07-21 13:20:55 +05:30
|
|
|
//config: Supports network read and connect timeouts for wget,
|
|
|
|
//config: so that wget will give up and timeout, through the -T
|
|
|
|
//config: command line option.
|
2015-10-07 05:09:40 +05:30
|
|
|
//config:
|
2017-07-21 13:20:55 +05:30
|
|
|
//config: Currently only connect and network data read timeout are
|
|
|
|
//config: supported (i.e., timeout is not applied to the DNS query). When
|
|
|
|
//config: FEATURE_WGET_LONG_OPTIONS is also enabled, the --timeout option
|
|
|
|
//config: will work in addition to -T.
|
2015-10-07 05:09:40 +05:30
|
|
|
//config:
|
2017-01-23 05:38:16 +05:30
|
|
|
//config:config FEATURE_WGET_HTTPS
|
|
|
|
//config: bool "Support HTTPS using internal TLS code"
|
2018-02-06 19:45:08 +05:30
|
|
|
//it also enables FTPS support, but it's not well tested yet
|
2017-01-23 05:38:16 +05:30
|
|
|
//config: default y
|
|
|
|
//config: depends on WGET
|
|
|
|
//config: select TLS
|
|
|
|
//config: help
|
2017-07-21 13:20:55 +05:30
|
|
|
//config: wget will use internal TLS code to connect to https:// URLs.
|
|
|
|
//config: Note:
|
|
|
|
//config: On NOMMU machines, ssl_helper applet should be available
|
|
|
|
//config: in the $PATH for this to work. Make sure to select that applet.
|
2017-01-23 05:38:16 +05:30
|
|
|
//config:
|
2017-07-21 13:20:55 +05:30
|
|
|
//config: Note: currently, TLS code only makes TLS I/O work, it
|
|
|
|
//config: does *not* check that the peer is who it claims to be, etc.
|
|
|
|
//config: IOW: it uses peer-supplied public keys to establish encryption
|
|
|
|
//config: and signing keys, then encrypts and signs outgoing data and
|
|
|
|
//config: decrypts incoming data.
|
|
|
|
//config: It does not check signature hashes on the incoming data:
|
|
|
|
//config: this means that attackers manipulating TCP packets can
|
|
|
|
//config: send altered data and we unknowingly receive garbage.
|
|
|
|
//config: (This check might be relatively easy to add).
|
|
|
|
//config: It does not check public key's certificate:
|
|
|
|
//config: this means that the peer may be an attacker impersonating
|
|
|
|
//config: the server we think we are talking to.
|
2017-01-30 20:57:37 +05:30
|
|
|
//config:
|
2017-07-21 13:20:55 +05:30
|
|
|
//config: If you think this is unacceptable, consider this. As more and more
|
|
|
|
//config: servers switch to HTTPS-only operation, without such "crippled"
|
|
|
|
//config: TLS code it is *impossible* to simply download a kernel source
|
|
|
|
//config: from kernel.org. Which can in real world translate into
|
|
|
|
//config: "my small automatic tooling to build cross-compilers from sources
|
|
|
|
//config: no longer works, I need to additionally keep a local copy
|
|
|
|
//config: of ~4 megabyte source tarball of a SSL library and ~2 megabyte
|
|
|
|
//config: source of wget, need to compile and built both before I can
|
|
|
|
//config: download anything. All this despite the fact that the build
|
|
|
|
//config: is done in a QEMU sandbox on a machine with absolutely nothing
|
|
|
|
//config: worth stealing, so I don't care if someone would go to a lot
|
|
|
|
//config: of trouble to intercept my HTTPS download to send me an altered
|
|
|
|
//config: kernel tarball".
|
2017-01-30 20:57:37 +05:30
|
|
|
//config:
|
2017-07-21 13:20:55 +05:30
|
|
|
//config: If you still think this is unacceptable, send patches.
|
2017-01-30 20:57:37 +05:30
|
|
|
//config:
|
2017-07-21 13:20:55 +05:30
|
|
|
//config: If you still think this is unacceptable, do not want to send
|
|
|
|
//config: patches, but do want to waste bandwidth expaining how wrong
|
|
|
|
//config: it is, you will be ignored.
|
2017-01-30 20:57:37 +05:30
|
|
|
//config:
|
2015-10-07 06:10:53 +05:30
|
|
|
//config:config FEATURE_WGET_OPENSSL
|
|
|
|
//config: bool "Try to connect to HTTPS using openssl"
|
|
|
|
//config: default y
|
2015-10-07 05:09:40 +05:30
|
|
|
//config: depends on WGET
|
|
|
|
//config: help
|
2017-07-21 13:20:55 +05:30
|
|
|
//config: Try to use openssl to handle HTTPS.
|
2015-10-07 05:09:40 +05:30
|
|
|
//config:
|
2017-07-21 13:20:55 +05:30
|
|
|
//config: OpenSSL has a simple SSL client for debug purposes.
|
|
|
|
//config: If you select this option, wget will effectively run:
|
|
|
|
//config: "openssl s_client -quiet -connect hostname:443
|
|
|
|
//config: -servername hostname 2>/dev/null" and pipe its data
|
|
|
|
//config: through it. -servername is not used if hostname is numeric.
|
|
|
|
//config: Note inconvenient API: host resolution is done twice,
|
|
|
|
//config: and there is no guarantee openssl's idea of IPv6 address
|
|
|
|
//config: format is the same as ours.
|
|
|
|
//config: Another problem is that s_client prints debug information
|
|
|
|
//config: to stderr, and it needs to be suppressed. This means
|
|
|
|
//config: all error messages get suppressed too.
|
|
|
|
//config: openssl is also a big binary, often dynamically linked
|
|
|
|
//config: against ~15 libraries.
|
2015-10-07 05:09:40 +05:30
|
|
|
//config:
|
2017-07-21 13:20:55 +05:30
|
|
|
//config: If openssl can't be executed, internal TLS code will be used
|
|
|
|
//config: (if you enabled it); if openssl can be executed but fails later,
|
|
|
|
//config: wget can't detect this, and download will fail.
|
2015-10-07 05:09:40 +05:30
|
|
|
|
|
|
|
//applet:IF_WGET(APPLET(wget, BB_DIR_USR_BIN, BB_SUID_DROP))
|
|
|
|
|
|
|
|
//kbuild:lib-$(CONFIG_WGET) += wget.o
|
|
|
|
|
2011-03-21 05:07:05 +05:30
|
|
|
//usage:#define wget_trivial_usage
|
|
|
|
//usage: IF_FEATURE_WGET_LONG_OPTIONS(
|
2016-10-04 07:53:09 +05:30
|
|
|
//usage: "[-c|--continue] [--spider] [-q|--quiet] [-O|--output-document FILE]\n"
|
2011-03-21 05:07:05 +05:30
|
|
|
//usage: " [--header 'header: value'] [-Y|--proxy on/off] [-P DIR]\n"
|
2012-06-13 20:59:41 +05:30
|
|
|
/* Since we ignore these opts, we don't show them in --help */
|
2015-10-21 01:21:52 +05:30
|
|
|
/* //usage: " [--no-check-certificate] [--no-cache] [--passive-ftp] [-t TRIES]" */
|
|
|
|
/* //usage: " [-nv] [-nc] [-nH] [-np]" */
|
2017-01-25 02:11:43 +05:30
|
|
|
//usage: " [-S|--server-response] [-U|--user-agent AGENT]" IF_FEATURE_WGET_TIMEOUT(" [-T SEC]") " URL..."
|
2011-03-21 05:07:05 +05:30
|
|
|
//usage: )
|
|
|
|
//usage: IF_NOT_FEATURE_WGET_LONG_OPTIONS(
|
2017-01-25 02:11:43 +05:30
|
|
|
//usage: "[-cq] [-O FILE] [-Y on/off] [-P DIR] [-S] [-U AGENT]"
|
2011-03-21 05:07:05 +05:30
|
|
|
//usage: IF_FEATURE_WGET_TIMEOUT(" [-T SEC]") " URL..."
|
|
|
|
//usage: )
|
|
|
|
//usage:#define wget_full_usage "\n\n"
|
|
|
|
//usage: "Retrieve files via HTTP or FTP\n"
|
2016-10-04 07:53:09 +05:30
|
|
|
//usage: IF_FEATURE_WGET_LONG_OPTIONS(
|
2017-01-25 02:11:43 +05:30
|
|
|
//usage: "\n --spider Only check URL existence: $? is 0 if exists"
|
2018-05-28 18:06:26 +05:30
|
|
|
///////: "\n --no-check-certificate Don't validate the server's certificate"
|
2016-10-04 07:53:09 +05:30
|
|
|
//usage: )
|
|
|
|
//usage: "\n -c Continue retrieval of aborted transfer"
|
|
|
|
//usage: "\n -q Quiet"
|
|
|
|
//usage: "\n -P DIR Save to DIR (default .)"
|
2017-01-25 02:11:43 +05:30
|
|
|
//usage: "\n -S Show server response"
|
2011-03-21 05:07:05 +05:30
|
|
|
//usage: IF_FEATURE_WGET_TIMEOUT(
|
2016-10-04 07:53:09 +05:30
|
|
|
//usage: "\n -T SEC Network read timeout is SEC seconds"
|
2011-03-21 05:07:05 +05:30
|
|
|
//usage: )
|
2016-10-04 07:53:09 +05:30
|
|
|
//usage: "\n -O FILE Save to FILE ('-' for stdout)"
|
|
|
|
//usage: "\n -U STR Use STR for User-Agent header"
|
|
|
|
//usage: "\n -Y on/off Use proxy"
|
2011-03-21 05:07:05 +05:30
|
|
|
|
2007-05-27 00:30:18 +05:30
|
|
|
#include "libbb.h"
|
2006-09-26 14:52:12 +05:30
|
|
|
|
2012-06-13 20:59:41 +05:30
|
|
|
#if 0
|
|
|
|
# define log_io(...) bb_error_msg(__VA_ARGS__)
|
2015-02-19 01:11:02 +05:30
|
|
|
# define SENDFMT(fp, fmt, ...) \
|
|
|
|
do { \
|
|
|
|
log_io("> " fmt, ##__VA_ARGS__); \
|
|
|
|
fprintf(fp, fmt, ##__VA_ARGS__); \
|
|
|
|
} while (0);
|
2012-06-13 20:59:41 +05:30
|
|
|
#else
|
|
|
|
# define log_io(...) ((void)0)
|
2015-02-19 01:11:02 +05:30
|
|
|
# define SENDFMT(fp, fmt, ...) fprintf(fp, fmt, ##__VA_ARGS__)
|
2012-06-13 20:59:41 +05:30
|
|
|
#endif
|
2011-02-11 03:32:28 +05:30
|
|
|
|
|
|
|
|
2017-01-23 05:38:16 +05:30
|
|
|
#define SSL_SUPPORTED (ENABLE_FEATURE_WGET_OPENSSL || ENABLE_FEATURE_WGET_HTTPS)
|
wget: fix for brain-damaged HTTP servers. Closes 9471
write(3, "GET / HTTP/1.1\r\nUser-Agent: Wget\r\nConnection: close\r\n\r\n", 74) = 74
shutdown(3, SHUT_WR) = 0
alarm(900) = 900
read(3, "", 1024) = 0
write(2, "wget: error getting response\n", 29) = 29
exit(1)
The peer simply does not return anything. It closes its connection.
Probably it detects wget closing its writing end: shutdown(3, SHUT_WR).
The point it, closing write side of the socket is _valid_ for HTTP.
wget sent the full request, it won't be sending anything more:
it will only receive the response, and that's it.
Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>
2017-01-12 00:46:45 +05:30
|
|
|
|
2001-04-06 03:15:54 +05:30
|
|
|
struct host_info {
|
2011-02-13 07:03:11 +05:30
|
|
|
char *allocated;
|
2007-09-24 23:57:04 +05:30
|
|
|
const char *path;
|
2014-02-03 18:39:42 +05:30
|
|
|
char *user;
|
2014-02-22 18:42:29 +05:30
|
|
|
const char *protocol;
|
2007-09-24 23:57:04 +05:30
|
|
|
char *host;
|
|
|
|
int port;
|
2001-04-06 03:15:54 +05:30
|
|
|
};
|
2016-04-22 21:39:21 +05:30
|
|
|
static const char P_FTP[] ALIGN1 = "ftp";
|
|
|
|
static const char P_HTTP[] ALIGN1 = "http";
|
wget: fix for brain-damaged HTTP servers. Closes 9471
write(3, "GET / HTTP/1.1\r\nUser-Agent: Wget\r\nConnection: close\r\n\r\n", 74) = 74
shutdown(3, SHUT_WR) = 0
alarm(900) = 900
read(3, "", 1024) = 0
write(2, "wget: error getting response\n", 29) = 29
exit(1)
The peer simply does not return anything. It closes its connection.
Probably it detects wget closing its writing end: shutdown(3, SHUT_WR).
The point it, closing write side of the socket is _valid_ for HTTP.
wget sent the full request, it won't be sending anything more:
it will only receive the response, and that's it.
Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>
2017-01-12 00:46:45 +05:30
|
|
|
#if SSL_SUPPORTED
|
2018-02-06 19:45:08 +05:30
|
|
|
# if ENABLE_FEATURE_WGET_HTTPS
|
|
|
|
static const char P_FTPS[] ALIGN1 = "ftps";
|
|
|
|
# endif
|
2016-04-22 21:39:21 +05:30
|
|
|
static const char P_HTTPS[] ALIGN1 = "https";
|
2015-10-12 15:21:25 +05:30
|
|
|
#endif
|
2001-04-06 03:15:54 +05:30
|
|
|
|
2015-02-19 01:11:02 +05:30
|
|
|
#if ENABLE_FEATURE_WGET_LONG_OPTIONS
|
|
|
|
/* User-specified headers prevent using our corresponding built-in headers. */
|
|
|
|
enum {
|
|
|
|
HDR_HOST = (1<<0),
|
|
|
|
HDR_USER_AGENT = (1<<1),
|
|
|
|
HDR_RANGE = (1<<2),
|
|
|
|
HDR_AUTH = (1<<3) * ENABLE_FEATURE_WGET_AUTHENTICATION,
|
|
|
|
HDR_PROXY_AUTH = (1<<4) * ENABLE_FEATURE_WGET_AUTHENTICATION,
|
|
|
|
};
|
|
|
|
static const char wget_user_headers[] ALIGN1 =
|
|
|
|
"Host:\0"
|
|
|
|
"User-Agent:\0"
|
|
|
|
"Range:\0"
|
|
|
|
# if ENABLE_FEATURE_WGET_AUTHENTICATION
|
|
|
|
"Authorization:\0"
|
|
|
|
"Proxy-Authorization:\0"
|
|
|
|
# endif
|
|
|
|
;
|
|
|
|
# define USR_HEADER_HOST (G.user_headers & HDR_HOST)
|
|
|
|
# define USR_HEADER_USER_AGENT (G.user_headers & HDR_USER_AGENT)
|
|
|
|
# define USR_HEADER_RANGE (G.user_headers & HDR_RANGE)
|
|
|
|
# define USR_HEADER_AUTH (G.user_headers & HDR_AUTH)
|
|
|
|
# define USR_HEADER_PROXY_AUTH (G.user_headers & HDR_PROXY_AUTH)
|
|
|
|
#else /* No long options, no user-headers :( */
|
|
|
|
# define USR_HEADER_HOST 0
|
|
|
|
# define USR_HEADER_USER_AGENT 0
|
|
|
|
# define USR_HEADER_RANGE 0
|
|
|
|
# define USR_HEADER_AUTH 0
|
|
|
|
# define USR_HEADER_PROXY_AUTH 0
|
|
|
|
#endif
|
2007-09-24 20:34:00 +05:30
|
|
|
|
2010-08-08 06:21:20 +05:30
|
|
|
/* Globals */
|
2007-09-24 20:34:00 +05:30
|
|
|
struct globals {
|
|
|
|
off_t content_len; /* Content-length of the file */
|
|
|
|
off_t beg_range; /* Range at which continue begins */
|
|
|
|
#if ENABLE_FEATURE_WGET_STATUSBAR
|
|
|
|
off_t transferred; /* Number of bytes transferred so far */
|
|
|
|
const char *curfile; /* Name of current file being transferred */
|
2009-11-08 21:04:43 +05:30
|
|
|
bb_progress_t pmt;
|
2010-08-08 06:21:20 +05:30
|
|
|
#endif
|
2013-07-30 15:22:58 +05:30
|
|
|
char *dir_prefix;
|
2011-02-13 07:03:11 +05:30
|
|
|
#if ENABLE_FEATURE_WGET_LONG_OPTIONS
|
2013-07-30 15:22:58 +05:30
|
|
|
char *post_data;
|
|
|
|
char *extra_headers;
|
2015-02-19 01:11:02 +05:30
|
|
|
unsigned char user_headers; /* Headers mentioned by the user */
|
2011-02-13 07:03:11 +05:30
|
|
|
#endif
|
2013-07-30 15:22:58 +05:30
|
|
|
char *fname_out; /* where to direct output (-O) */
|
|
|
|
const char *proxy_flag; /* Use proxies if env vars are set */
|
|
|
|
const char *user_agent; /* "User-Agent" header field */
|
2018-11-25 02:26:21 +05:30
|
|
|
int output_fd;
|
|
|
|
int o_flags;
|
2010-08-08 06:21:20 +05:30
|
|
|
#if ENABLE_FEATURE_WGET_TIMEOUT
|
|
|
|
unsigned timeout_seconds;
|
2018-11-25 02:26:21 +05:30
|
|
|
smallint die_if_timed_out;
|
2007-09-24 20:34:00 +05:30
|
|
|
#endif
|
2009-06-28 04:32:24 +05:30
|
|
|
smallint chunked; /* chunked transfer encoding */
|
|
|
|
smallint got_clen; /* got content-length: from server */
|
2011-02-10 10:59:06 +05:30
|
|
|
/* Local downloads do benefit from big buffer.
|
|
|
|
* With 512 byte buffer, it was measured to be
|
|
|
|
* an order of magnitude slower than with big one.
|
|
|
|
*/
|
2018-11-25 02:26:21 +05:30
|
|
|
char wget_buf[CONFIG_FEATURE_COPYBUF_KB*1024] ALIGNED(sizeof(long));
|
2010-02-04 19:30:15 +05:30
|
|
|
} FIX_ALIASING;
|
2011-02-10 10:59:06 +05:30
|
|
|
#define G (*ptr_to_globals)
|
2010-08-08 06:21:20 +05:30
|
|
|
#define INIT_G() do { \
|
2013-07-30 15:22:58 +05:30
|
|
|
SET_PTR_TO_GLOBALS(xzalloc(sizeof(G))); \
|
2010-08-08 06:21:20 +05:30
|
|
|
} while (0)
|
2013-10-16 18:13:30 +05:30
|
|
|
#define FINI_G() do { \
|
|
|
|
FREE_PTR_TO_GLOBALS(); \
|
|
|
|
} while (0)
|
2007-09-24 20:34:00 +05:30
|
|
|
|
|
|
|
|
2010-08-08 06:21:20 +05:30
|
|
|
/* Must match option string! */
|
|
|
|
enum {
|
|
|
|
WGET_OPT_CONTINUE = (1 << 0),
|
2016-10-04 07:53:09 +05:30
|
|
|
WGET_OPT_QUIET = (1 << 1),
|
2017-01-25 02:11:43 +05:30
|
|
|
WGET_OPT_SERVER_RESPONSE = (1 << 2),
|
|
|
|
WGET_OPT_OUTNAME = (1 << 3),
|
|
|
|
WGET_OPT_PREFIX = (1 << 4),
|
|
|
|
WGET_OPT_PROXY = (1 << 5),
|
|
|
|
WGET_OPT_USER_AGENT = (1 << 6),
|
|
|
|
WGET_OPT_NETWORK_READ_TIMEOUT = (1 << 7),
|
|
|
|
WGET_OPT_RETRIES = (1 << 8),
|
|
|
|
WGET_OPT_nsomething = (1 << 9),
|
|
|
|
WGET_OPT_HEADER = (1 << 10) * ENABLE_FEATURE_WGET_LONG_OPTIONS,
|
|
|
|
WGET_OPT_POST_DATA = (1 << 11) * ENABLE_FEATURE_WGET_LONG_OPTIONS,
|
|
|
|
WGET_OPT_SPIDER = (1 << 12) * ENABLE_FEATURE_WGET_LONG_OPTIONS,
|
2018-05-28 18:06:26 +05:30
|
|
|
WGET_OPT_NO_CHECK_CERT = (1 << 13) * ENABLE_FEATURE_WGET_LONG_OPTIONS,
|
2010-08-08 06:21:20 +05:30
|
|
|
};
|
2007-09-24 23:54:17 +05:30
|
|
|
|
2010-08-08 06:21:20 +05:30
|
|
|
enum {
|
|
|
|
PROGRESS_START = -1,
|
|
|
|
PROGRESS_END = 0,
|
|
|
|
PROGRESS_BUMP = 1,
|
|
|
|
};
|
|
|
|
#if ENABLE_FEATURE_WGET_STATUSBAR
|
2008-11-24 13:04:42 +05:30
|
|
|
static void progress_meter(int flag)
|
2007-09-24 23:54:17 +05:30
|
|
|
{
|
2018-11-23 23:44:52 +05:30
|
|
|
int notty;
|
|
|
|
|
2010-08-08 06:21:20 +05:30
|
|
|
if (option_mask32 & WGET_OPT_QUIET)
|
|
|
|
return;
|
2007-09-24 23:54:17 +05:30
|
|
|
|
2010-08-08 06:21:20 +05:30
|
|
|
if (flag == PROGRESS_START)
|
2011-02-11 23:26:13 +05:30
|
|
|
bb_progress_init(&G.pmt, G.curfile);
|
2007-09-24 23:54:17 +05:30
|
|
|
|
2018-11-23 23:44:52 +05:30
|
|
|
notty = bb_progress_update(&G.pmt,
|
2011-02-15 05:28:36 +05:30
|
|
|
G.beg_range,
|
|
|
|
G.transferred,
|
|
|
|
(G.chunked || !G.got_clen) ? 0 : G.beg_range + G.transferred + G.content_len
|
|
|
|
);
|
2007-09-24 23:54:17 +05:30
|
|
|
|
2010-08-08 06:21:20 +05:30
|
|
|
if (flag == PROGRESS_END) {
|
2011-02-13 07:03:11 +05:30
|
|
|
bb_progress_free(&G.pmt);
|
2018-11-23 23:44:52 +05:30
|
|
|
if (notty == 0)
|
|
|
|
bb_putchar_stderr('\n'); /* it's tty */
|
2009-12-11 17:06:10 +05:30
|
|
|
G.transferred = 0;
|
2007-09-24 23:54:17 +05:30
|
|
|
}
|
|
|
|
}
|
2010-08-08 06:21:20 +05:30
|
|
|
#else
|
2008-11-24 13:04:42 +05:30
|
|
|
static ALWAYS_INLINE void progress_meter(int flag UNUSED_PARAM) { }
|
2000-10-03 05:51:45 +05:30
|
|
|
#endif
|
2004-03-15 13:59:22 +05:30
|
|
|
|
2007-09-24 23:54:17 +05:30
|
|
|
|
2009-07-01 00:06:27 +05:30
|
|
|
/* IPv6 knows scoped address types i.e. link and site local addresses. Link
|
|
|
|
* local addresses can have a scope identifier to specify the
|
|
|
|
* interface/link an address is valid on (e.g. fe80::1%eth0). This scope
|
|
|
|
* identifier is only valid on a single node.
|
|
|
|
*
|
|
|
|
* RFC 4007 says that the scope identifier MUST NOT be sent across the wire,
|
|
|
|
* unless all nodes agree on the semantic. Apache e.g. regards zone identifiers
|
|
|
|
* in the Host header as invalid requests, see
|
|
|
|
* https://issues.apache.org/bugzilla/show_bug.cgi?id=35122
|
|
|
|
*/
|
|
|
|
static void strip_ipv6_scope_id(char *host)
|
|
|
|
{
|
|
|
|
char *scope, *cp;
|
|
|
|
|
|
|
|
/* bbox wget actually handles IPv6 addresses without [], like
|
|
|
|
* wget "http://::1/xxx", but this is not standard.
|
|
|
|
* To save code, _here_ we do not support it. */
|
|
|
|
|
|
|
|
if (host[0] != '[')
|
|
|
|
return; /* not IPv6 */
|
|
|
|
|
|
|
|
scope = strchr(host, '%');
|
|
|
|
if (!scope)
|
|
|
|
return;
|
|
|
|
|
|
|
|
/* Remove the IPv6 zone identifier from the host address */
|
|
|
|
cp = strchr(host, ']');
|
|
|
|
if (!cp || (cp[1] != ':' && cp[1] != '\0')) {
|
|
|
|
/* malformed address (not "[xx]:nn" or "[xx]") */
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* cp points to "]...", scope points to "%eth0]..." */
|
|
|
|
overlapping_strcpy(scope, cp);
|
|
|
|
}
|
|
|
|
|
2006-11-21 16:13:02 +05:30
|
|
|
#if ENABLE_FEATURE_WGET_AUTHENTICATION
|
2011-02-10 10:59:06 +05:30
|
|
|
/* Base64-encode character string. */
|
|
|
|
static char *base64enc(const char *str)
|
2006-09-09 17:50:57 +05:30
|
|
|
{
|
2018-11-25 02:26:21 +05:30
|
|
|
/* paranoia */
|
|
|
|
unsigned len = strnlen(str, sizeof(G.wget_buf)/4*3 - 10);
|
2011-02-10 10:59:06 +05:30
|
|
|
bb_uuencode(G.wget_buf, str, len, bb_uuenc_tbl_base64);
|
|
|
|
return G.wget_buf;
|
2001-04-06 03:15:54 +05:30
|
|
|
}
|
|
|
|
#endif
|
|
|
|
|
2013-10-13 01:17:07 +05:30
|
|
|
#if ENABLE_FEATURE_WGET_TIMEOUT
|
|
|
|
static void alarm_handler(int sig UNUSED_PARAM)
|
|
|
|
{
|
|
|
|
/* This is theoretically unsafe (uses stdio and malloc in signal handler) */
|
2016-03-17 20:28:16 +05:30
|
|
|
if (G.die_if_timed_out)
|
2013-10-13 01:17:07 +05:30
|
|
|
bb_error_msg_and_die("download timed out");
|
|
|
|
}
|
2016-03-17 20:28:16 +05:30
|
|
|
static void set_alarm(void)
|
|
|
|
{
|
|
|
|
if (G.timeout_seconds) {
|
|
|
|
alarm(G.timeout_seconds);
|
|
|
|
G.die_if_timed_out = 1;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
# define clear_alarm() ((void)(G.die_if_timed_out = 0))
|
|
|
|
#else
|
|
|
|
# define set_alarm() ((void)0)
|
|
|
|
# define clear_alarm() ((void)0)
|
2013-10-13 01:17:07 +05:30
|
|
|
#endif
|
|
|
|
|
2016-07-26 01:04:57 +05:30
|
|
|
#if ENABLE_FEATURE_WGET_OPENSSL
|
|
|
|
/*
|
|
|
|
* is_ip_address() attempts to verify whether or not a string
|
|
|
|
* contains an IPv4 or IPv6 address (vs. an FQDN). The result
|
|
|
|
* of inet_pton() can be used to determine this.
|
|
|
|
*
|
|
|
|
* TODO add proper error checking when inet_pton() returns -1
|
|
|
|
* (some form of system error has occurred, and errno is set)
|
|
|
|
*/
|
|
|
|
static int is_ip_address(const char *string)
|
|
|
|
{
|
|
|
|
struct sockaddr_in sa;
|
|
|
|
|
|
|
|
int result = inet_pton(AF_INET, string, &(sa.sin_addr));
|
|
|
|
# if ENABLE_FEATURE_IPV6
|
|
|
|
if (result == 0) {
|
|
|
|
struct sockaddr_in6 sa6;
|
|
|
|
result = inet_pton(AF_INET6, string, &(sa6.sin6_addr));
|
|
|
|
}
|
|
|
|
# endif
|
|
|
|
return (result == 1);
|
|
|
|
}
|
|
|
|
#endif
|
|
|
|
|
2007-09-24 23:54:17 +05:30
|
|
|
static FILE *open_socket(len_and_sockaddr *lsa)
|
2000-09-04 20:45:55 +05:30
|
|
|
{
|
2013-10-13 01:17:07 +05:30
|
|
|
int fd;
|
2007-09-24 23:54:17 +05:30
|
|
|
FILE *fp;
|
2007-09-24 20:34:00 +05:30
|
|
|
|
2016-03-17 20:28:16 +05:30
|
|
|
set_alarm();
|
2013-10-13 01:17:07 +05:30
|
|
|
fd = xconnect_stream(lsa);
|
2016-03-17 20:28:16 +05:30
|
|
|
clear_alarm();
|
2013-10-13 01:17:07 +05:30
|
|
|
|
2007-09-24 23:54:17 +05:30
|
|
|
/* glibc 2.4 seems to try seeking on it - ??! */
|
|
|
|
/* hopefully it understands what ESPIPE means... */
|
2013-10-13 01:17:07 +05:30
|
|
|
fp = fdopen(fd, "r+");
|
2014-02-22 18:42:29 +05:30
|
|
|
if (!fp)
|
2018-04-01 23:29:37 +05:30
|
|
|
bb_die_memory_exhausted();
|
2007-09-24 20:34:00 +05:30
|
|
|
|
2007-09-24 23:54:17 +05:30
|
|
|
return fp;
|
|
|
|
}
|
2007-09-24 20:34:00 +05:30
|
|
|
|
2018-02-12 21:16:13 +05:30
|
|
|
/* We balk at any control chars in other side's messages.
|
|
|
|
* This prevents nasty surprises (e.g. ESC sequences) in "Location:" URLs
|
|
|
|
* and error messages.
|
|
|
|
*
|
|
|
|
* The only exception is tabs, which are converted to (one) space:
|
|
|
|
* HTTP's "headers: <whitespace> values" may have those.
|
|
|
|
*/
|
|
|
|
static char* sanitize_string(char *s)
|
|
|
|
{
|
|
|
|
unsigned char *p = (void *) s;
|
|
|
|
while (*p) {
|
|
|
|
if (*p < ' ') {
|
|
|
|
if (*p != '\t')
|
|
|
|
break;
|
|
|
|
*p = ' ';
|
|
|
|
}
|
|
|
|
p++;
|
|
|
|
}
|
|
|
|
*p = '\0';
|
|
|
|
return s;
|
|
|
|
}
|
|
|
|
|
2011-02-11 03:32:28 +05:30
|
|
|
/* Returns '\n' if it was seen, else '\0'. Trims at first '\r' or '\n' */
|
2018-02-12 21:16:13 +05:30
|
|
|
static char fgets_trim_sanitize(FILE *fp, const char *fmt)
|
2011-02-11 03:32:28 +05:30
|
|
|
{
|
|
|
|
char c;
|
|
|
|
char *buf_ptr;
|
|
|
|
|
2016-03-17 20:28:16 +05:30
|
|
|
set_alarm();
|
2018-02-12 21:16:13 +05:30
|
|
|
if (fgets(G.wget_buf, sizeof(G.wget_buf), fp) == NULL)
|
2011-02-11 03:32:28 +05:30
|
|
|
bb_perror_msg_and_die("error getting response");
|
2016-03-17 20:28:16 +05:30
|
|
|
clear_alarm();
|
2011-02-11 03:32:28 +05:30
|
|
|
|
|
|
|
buf_ptr = strchrnul(G.wget_buf, '\n');
|
|
|
|
c = *buf_ptr;
|
2018-02-12 21:16:13 +05:30
|
|
|
#if 1
|
|
|
|
/* Disallow any control chars: trim at first char < 0x20 */
|
|
|
|
sanitize_string(G.wget_buf);
|
|
|
|
#else
|
2011-02-11 03:32:28 +05:30
|
|
|
*buf_ptr = '\0';
|
|
|
|
buf_ptr = strchrnul(G.wget_buf, '\r');
|
|
|
|
*buf_ptr = '\0';
|
2018-02-12 21:16:13 +05:30
|
|
|
#endif
|
2011-02-11 03:32:28 +05:30
|
|
|
|
|
|
|
log_io("< %s", G.wget_buf);
|
|
|
|
|
2017-01-25 02:11:43 +05:30
|
|
|
if (fmt && (option_mask32 & WGET_OPT_SERVER_RESPONSE))
|
|
|
|
fprintf(stderr, fmt, G.wget_buf);
|
|
|
|
|
2011-02-11 03:32:28 +05:30
|
|
|
return c;
|
|
|
|
}
|
|
|
|
|
2011-02-10 10:59:06 +05:30
|
|
|
static int ftpcmd(const char *s1, const char *s2, FILE *fp)
|
2007-09-24 23:54:17 +05:30
|
|
|
{
|
|
|
|
int result;
|
|
|
|
if (s1) {
|
2011-02-11 03:32:28 +05:30
|
|
|
if (!s2)
|
|
|
|
s2 = "";
|
2007-09-24 23:54:17 +05:30
|
|
|
fprintf(fp, "%s%s\r\n", s1, s2);
|
2017-01-25 02:11:43 +05:30
|
|
|
/* With --server-response, wget also shows its ftp commands */
|
|
|
|
if (option_mask32 & WGET_OPT_SERVER_RESPONSE)
|
|
|
|
fprintf(stderr, "--> %s%s\n\n", s1, s2);
|
2007-09-24 23:54:17 +05:30
|
|
|
fflush(fp);
|
2011-02-11 03:32:28 +05:30
|
|
|
log_io("> %s%s", s1, s2);
|
2007-09-24 23:54:17 +05:30
|
|
|
}
|
|
|
|
|
2018-02-12 21:16:13 +05:30
|
|
|
/* Read until "Nxx something" is received */
|
|
|
|
G.wget_buf[3] = 0;
|
2007-09-24 23:54:17 +05:30
|
|
|
do {
|
2018-02-12 21:16:13 +05:30
|
|
|
fgets_trim_sanitize(fp, "%s\n");
|
2011-02-10 10:59:06 +05:30
|
|
|
} while (!isdigit(G.wget_buf[0]) || G.wget_buf[3] != ' ');
|
2007-09-24 23:54:17 +05:30
|
|
|
|
2011-02-10 10:59:06 +05:30
|
|
|
G.wget_buf[3] = '\0';
|
|
|
|
result = xatoi_positive(G.wget_buf);
|
|
|
|
G.wget_buf[3] = ' ';
|
2007-09-24 23:54:17 +05:30
|
|
|
return result;
|
|
|
|
}
|
|
|
|
|
2011-02-13 07:03:11 +05:30
|
|
|
static void parse_url(const char *src_url, struct host_info *h)
|
2007-09-24 23:54:17 +05:30
|
|
|
{
|
|
|
|
char *url, *p, *sp;
|
|
|
|
|
2011-02-13 07:03:11 +05:30
|
|
|
free(h->allocated);
|
|
|
|
h->allocated = url = xstrdup(src_url);
|
2007-09-24 23:54:17 +05:30
|
|
|
|
2014-02-22 18:42:29 +05:30
|
|
|
h->protocol = P_FTP;
|
|
|
|
p = strstr(url, "://");
|
|
|
|
if (p) {
|
|
|
|
*p = '\0';
|
|
|
|
h->host = p + 3;
|
|
|
|
if (strcmp(url, P_FTP) == 0) {
|
2018-04-17 16:13:54 +05:30
|
|
|
h->port = bb_lookup_std_port(P_FTP, "tcp", 21);
|
2014-02-22 18:42:29 +05:30
|
|
|
} else
|
wget: fix for brain-damaged HTTP servers. Closes 9471
write(3, "GET / HTTP/1.1\r\nUser-Agent: Wget\r\nConnection: close\r\n\r\n", 74) = 74
shutdown(3, SHUT_WR) = 0
alarm(900) = 900
read(3, "", 1024) = 0
write(2, "wget: error getting response\n", 29) = 29
exit(1)
The peer simply does not return anything. It closes its connection.
Probably it detects wget closing its writing end: shutdown(3, SHUT_WR).
The point it, closing write side of the socket is _valid_ for HTTP.
wget sent the full request, it won't be sending anything more:
it will only receive the response, and that's it.
Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>
2017-01-12 00:46:45 +05:30
|
|
|
#if SSL_SUPPORTED
|
2018-02-06 19:45:08 +05:30
|
|
|
# if ENABLE_FEATURE_WGET_HTTPS
|
|
|
|
if (strcmp(url, P_FTPS) == 0) {
|
2018-04-17 16:13:54 +05:30
|
|
|
h->port = bb_lookup_std_port(P_FTPS, "tcp", 990);
|
2018-02-06 19:45:08 +05:30
|
|
|
h->protocol = P_FTPS;
|
|
|
|
} else
|
|
|
|
# endif
|
2014-02-22 18:42:29 +05:30
|
|
|
if (strcmp(url, P_HTTPS) == 0) {
|
2018-04-17 16:13:54 +05:30
|
|
|
h->port = bb_lookup_std_port(P_HTTPS, "tcp", 443);
|
2014-02-22 18:42:29 +05:30
|
|
|
h->protocol = P_HTTPS;
|
|
|
|
} else
|
2015-10-12 15:21:25 +05:30
|
|
|
#endif
|
2014-02-22 18:42:29 +05:30
|
|
|
if (strcmp(url, P_HTTP) == 0) {
|
2013-12-17 23:33:41 +05:30
|
|
|
http:
|
2018-04-17 16:13:54 +05:30
|
|
|
h->port = bb_lookup_std_port(P_HTTP, "tcp", 80);
|
2014-02-22 18:42:29 +05:30
|
|
|
h->protocol = P_HTTP;
|
|
|
|
} else {
|
|
|
|
*p = ':';
|
2018-02-12 21:16:13 +05:30
|
|
|
bb_error_msg_and_die("not an http or ftp url: %s", url);
|
2014-02-22 18:42:29 +05:30
|
|
|
}
|
|
|
|
} else {
|
2013-12-17 23:33:41 +05:30
|
|
|
// GNU wget is user-friendly and falls back to http://
|
|
|
|
h->host = url;
|
|
|
|
goto http;
|
2014-02-22 18:42:29 +05:30
|
|
|
}
|
2007-09-24 23:54:17 +05:30
|
|
|
|
|
|
|
// FYI:
|
|
|
|
// "Real" wget 'http://busybox.net?var=a/b' sends this request:
|
2017-01-20 18:42:10 +05:30
|
|
|
// 'GET /?var=a/b HTTP/1.0'
|
2007-09-24 23:54:17 +05:30
|
|
|
// and saves 'index.html?var=a%2Fb' (we save 'b')
|
|
|
|
// wget 'http://busybox.net?login=john@doe':
|
|
|
|
// request: 'GET /?login=john@doe HTTP/1.0'
|
2018-04-24 17:05:32 +05:30
|
|
|
// saves: 'index.html?login=john@doe' (we save 'login=john@doe')
|
2007-09-24 23:54:17 +05:30
|
|
|
// wget 'http://busybox.net#test/test':
|
|
|
|
// request: 'GET / HTTP/1.0'
|
|
|
|
// saves: 'index.html' (we save 'test')
|
|
|
|
//
|
|
|
|
// We also don't add unique .N suffix if file exists...
|
|
|
|
sp = strchr(h->host, '/');
|
|
|
|
p = strchr(h->host, '?'); if (!sp || (p && sp > p)) sp = p;
|
|
|
|
p = strchr(h->host, '#'); if (!sp || (p && sp > p)) sp = p;
|
|
|
|
if (!sp) {
|
2007-09-24 23:57:04 +05:30
|
|
|
h->path = "";
|
2007-09-24 23:54:17 +05:30
|
|
|
} else if (*sp == '/') {
|
|
|
|
*sp = '\0';
|
|
|
|
h->path = sp + 1;
|
2018-04-24 17:05:32 +05:30
|
|
|
} else {
|
|
|
|
// sp points to '#' or '?'
|
|
|
|
// Note:
|
2007-09-24 23:54:17 +05:30
|
|
|
// http://busybox.net?login=john@doe is a valid URL
|
2018-04-24 17:05:32 +05:30
|
|
|
// (without '/' between ".net" and "?"),
|
|
|
|
// can't store NUL at sp[-1] - this destroys hostname.
|
|
|
|
*sp++ = '\0';
|
2007-09-24 23:54:17 +05:30
|
|
|
h->path = sp;
|
|
|
|
}
|
|
|
|
|
|
|
|
sp = strrchr(h->host, '@');
|
|
|
|
if (sp != NULL) {
|
2011-09-12 00:34:02 +05:30
|
|
|
// URL-decode "user:password" string before base64-encoding:
|
|
|
|
// wget http://test:my%20pass@example.com should send
|
|
|
|
// Authorization: Basic dGVzdDpteSBwYXNz
|
|
|
|
// which decodes to "test:my pass".
|
|
|
|
// Standard wget and curl do this too.
|
2007-09-24 23:54:17 +05:30
|
|
|
*sp = '\0';
|
2014-02-03 18:39:42 +05:30
|
|
|
free(h->user);
|
|
|
|
h->user = xstrdup(percent_decode_in_place(h->host, /*strict:*/ 0));
|
2007-09-24 23:54:17 +05:30
|
|
|
h->host = sp + 1;
|
|
|
|
}
|
2014-02-03 18:39:42 +05:30
|
|
|
/* else: h->user remains NULL, or as set by original request
|
|
|
|
* before redirect (if we are here after a redirect).
|
|
|
|
*/
|
2007-09-24 23:54:17 +05:30
|
|
|
}
|
|
|
|
|
2018-02-12 21:16:13 +05:30
|
|
|
static char *get_sanitized_hdr(FILE *fp)
|
2007-09-24 23:54:17 +05:30
|
|
|
{
|
|
|
|
char *s, *hdrval;
|
|
|
|
int c;
|
|
|
|
|
|
|
|
/* retrieve header line */
|
2018-02-12 21:16:13 +05:30
|
|
|
c = fgets_trim_sanitize(fp, " %s\n");
|
2007-09-24 23:54:17 +05:30
|
|
|
|
2011-02-11 03:32:28 +05:30
|
|
|
/* end of the headers? */
|
|
|
|
if (G.wget_buf[0] == '\0')
|
2007-09-24 23:54:17 +05:30
|
|
|
return NULL;
|
|
|
|
|
|
|
|
/* convert the header name to lower case */
|
2013-07-01 18:31:50 +05:30
|
|
|
for (s = G.wget_buf; isalnum(*s) || *s == '-' || *s == '.' || *s == '_'; ++s) {
|
|
|
|
/*
|
|
|
|
* No-op for 20-3f and 60-7f. "0-9a-z-." are in these ranges.
|
|
|
|
* 40-5f range ("@A-Z[\]^_") maps to 60-7f.
|
|
|
|
* "A-Z" maps to "a-z".
|
|
|
|
* "@[\]" can't occur in header names.
|
|
|
|
* "^_" maps to "~,DEL" (which is wrong).
|
|
|
|
* "^" was never seen yet, "_" was seen from web.archive.org
|
|
|
|
* (x-archive-orig-x_commoncrawl_Signature: HEXSTRING).
|
|
|
|
*/
|
2011-02-11 03:32:28 +05:30
|
|
|
*s |= 0x20;
|
2010-04-04 18:59:32 +05:30
|
|
|
}
|
2007-09-24 23:54:17 +05:30
|
|
|
|
|
|
|
/* verify we are at the end of the header name */
|
|
|
|
if (*s != ':')
|
2018-02-12 21:16:13 +05:30
|
|
|
bb_error_msg_and_die("bad header line: %s", G.wget_buf);
|
2007-09-24 23:54:17 +05:30
|
|
|
|
|
|
|
/* locate the start of the header value */
|
|
|
|
*s++ = '\0';
|
|
|
|
hdrval = skip_whitespace(s);
|
|
|
|
|
2011-02-11 03:32:28 +05:30
|
|
|
if (c != '\n') {
|
|
|
|
/* Rats! The buffer isn't big enough to hold the entire header value */
|
|
|
|
while (c = getc(fp), c != EOF && c != '\n')
|
|
|
|
continue;
|
2007-09-24 23:54:17 +05:30
|
|
|
}
|
|
|
|
|
|
|
|
return hdrval;
|
|
|
|
}
|
|
|
|
|
2012-06-13 21:01:07 +05:30
|
|
|
static void reset_beg_range_to_zero(void)
|
|
|
|
{
|
2012-06-17 23:22:25 +05:30
|
|
|
bb_error_msg("restart failed");
|
2012-06-13 21:01:07 +05:30
|
|
|
G.beg_range = 0;
|
|
|
|
xlseek(G.output_fd, 0, SEEK_SET);
|
2012-06-17 23:22:25 +05:30
|
|
|
/* Done at the end instead: */
|
|
|
|
/* ftruncate(G.output_fd, 0); */
|
2012-06-13 21:01:07 +05:30
|
|
|
}
|
|
|
|
|
2015-10-07 05:09:40 +05:30
|
|
|
#if ENABLE_FEATURE_WGET_OPENSSL
|
2015-10-07 06:10:53 +05:30
|
|
|
static int spawn_https_helper_openssl(const char *host, unsigned port)
|
2014-02-22 18:42:29 +05:30
|
|
|
{
|
|
|
|
char *allocated = NULL;
|
2016-07-26 01:04:57 +05:30
|
|
|
char *servername;
|
2014-02-22 18:42:29 +05:30
|
|
|
int sp[2];
|
|
|
|
int pid;
|
2017-01-23 05:38:16 +05:30
|
|
|
IF_FEATURE_WGET_HTTPS(volatile int child_failed = 0;)
|
2014-02-22 18:42:29 +05:30
|
|
|
|
|
|
|
if (socketpair(AF_UNIX, SOCK_STREAM, 0, sp) != 0)
|
|
|
|
/* Kernel can have AF_UNIX support disabled */
|
|
|
|
bb_perror_msg_and_die("socketpair");
|
|
|
|
|
|
|
|
if (!strchr(host, ':'))
|
|
|
|
host = allocated = xasprintf("%s:%u", host, port);
|
2016-07-26 01:04:57 +05:30
|
|
|
servername = xstrdup(host);
|
|
|
|
strrchr(servername, ':')[0] = '\0';
|
2014-02-22 18:42:29 +05:30
|
|
|
|
2015-10-07 06:10:53 +05:30
|
|
|
fflush_all();
|
|
|
|
pid = xvfork();
|
2014-02-22 18:42:29 +05:30
|
|
|
if (pid == 0) {
|
|
|
|
/* Child */
|
2016-07-26 01:04:57 +05:30
|
|
|
char *argv[8];
|
2014-02-22 18:42:29 +05:30
|
|
|
|
|
|
|
close(sp[0]);
|
|
|
|
xmove_fd(sp[1], 0);
|
|
|
|
xdup2(0, 1);
|
|
|
|
/*
|
|
|
|
* openssl s_client -quiet -connect www.kernel.org:443 2>/dev/null
|
|
|
|
* It prints some debug stuff on stderr, don't know how to suppress it.
|
|
|
|
* Work around by dev-nulling stderr. We lose all error messages :(
|
|
|
|
*/
|
|
|
|
xmove_fd(2, 3);
|
|
|
|
xopen("/dev/null", O_RDWR);
|
2016-07-26 01:04:57 +05:30
|
|
|
memset(&argv, 0, sizeof(argv));
|
2014-02-22 18:42:29 +05:30
|
|
|
argv[0] = (char*)"openssl";
|
|
|
|
argv[1] = (char*)"s_client";
|
|
|
|
argv[2] = (char*)"-quiet";
|
|
|
|
argv[3] = (char*)"-connect";
|
|
|
|
argv[4] = (char*)host;
|
2016-07-26 01:04:57 +05:30
|
|
|
/*
|
|
|
|
* Per RFC 6066 Section 3, the only permitted values in the
|
|
|
|
* TLS server_name (SNI) field are FQDNs (DNS hostnames).
|
|
|
|
* IPv4 and IPv6 addresses, port numbers are not allowed.
|
|
|
|
*/
|
|
|
|
if (!is_ip_address(servername)) {
|
|
|
|
argv[5] = (char*)"-servername";
|
|
|
|
argv[6] = (char*)servername;
|
|
|
|
}
|
|
|
|
|
2014-02-22 18:42:29 +05:30
|
|
|
BB_EXECVP(argv[0], argv);
|
|
|
|
xmove_fd(3, 2);
|
2017-01-23 05:38:16 +05:30
|
|
|
# if ENABLE_FEATURE_WGET_HTTPS
|
2015-10-07 06:10:53 +05:30
|
|
|
child_failed = 1;
|
|
|
|
xfunc_die();
|
|
|
|
# else
|
2014-02-22 18:42:29 +05:30
|
|
|
bb_perror_msg_and_die("can't execute '%s'", argv[0]);
|
2015-10-07 06:10:53 +05:30
|
|
|
# endif
|
2014-02-22 18:42:29 +05:30
|
|
|
/* notreached */
|
|
|
|
}
|
|
|
|
|
2014-02-24 04:09:47 +05:30
|
|
|
/* Parent */
|
2016-07-26 01:04:57 +05:30
|
|
|
free(servername);
|
2014-02-22 18:42:29 +05:30
|
|
|
free(allocated);
|
|
|
|
close(sp[1]);
|
2017-01-23 05:38:16 +05:30
|
|
|
# if ENABLE_FEATURE_WGET_HTTPS
|
2015-10-07 06:10:53 +05:30
|
|
|
if (child_failed) {
|
|
|
|
close(sp[0]);
|
|
|
|
return -1;
|
|
|
|
}
|
|
|
|
# endif
|
2014-02-22 18:42:29 +05:30
|
|
|
return sp[0];
|
|
|
|
}
|
2015-10-07 05:09:40 +05:30
|
|
|
#endif
|
2014-02-22 18:42:29 +05:30
|
|
|
|
2017-01-23 05:38:16 +05:30
|
|
|
#if ENABLE_FEATURE_WGET_HTTPS
|
2018-02-06 19:45:08 +05:30
|
|
|
static void spawn_ssl_client(const char *host, int network_fd, int flags)
|
2014-02-24 04:09:47 +05:30
|
|
|
{
|
|
|
|
int sp[2];
|
|
|
|
int pid;
|
2017-01-23 05:38:16 +05:30
|
|
|
char *servername, *p;
|
|
|
|
|
2018-11-13 16:30:19 +05:30
|
|
|
if (!(option_mask32 & WGET_OPT_NO_CHECK_CERT)) {
|
|
|
|
option_mask32 |= WGET_OPT_NO_CHECK_CERT;
|
2018-11-14 16:05:36 +05:30
|
|
|
bb_error_msg("note: TLS certificate validation not implemented");
|
2018-11-13 16:30:19 +05:30
|
|
|
}
|
2018-05-28 18:06:26 +05:30
|
|
|
|
2017-01-23 05:38:16 +05:30
|
|
|
servername = xstrdup(host);
|
|
|
|
p = strrchr(servername, ':');
|
|
|
|
if (p) *p = '\0';
|
2014-02-24 04:09:47 +05:30
|
|
|
|
|
|
|
if (socketpair(AF_UNIX, SOCK_STREAM, 0, sp) != 0)
|
|
|
|
/* Kernel can have AF_UNIX support disabled */
|
|
|
|
bb_perror_msg_and_die("socketpair");
|
|
|
|
|
2017-01-23 05:38:16 +05:30
|
|
|
fflush_all();
|
2014-02-24 04:09:47 +05:30
|
|
|
pid = BB_MMU ? xfork() : xvfork();
|
|
|
|
if (pid == 0) {
|
|
|
|
/* Child */
|
|
|
|
close(sp[0]);
|
|
|
|
xmove_fd(sp[1], 0);
|
|
|
|
xdup2(0, 1);
|
2017-01-23 05:38:16 +05:30
|
|
|
if (BB_MMU) {
|
|
|
|
tls_state_t *tls = new_tls_state();
|
|
|
|
tls->ifd = tls->ofd = network_fd;
|
|
|
|
tls_handshake(tls, servername);
|
2018-02-06 19:45:08 +05:30
|
|
|
tls_run_copy_loop(tls, flags);
|
2017-01-23 05:38:16 +05:30
|
|
|
exit(0);
|
|
|
|
} else {
|
2018-02-06 19:45:08 +05:30
|
|
|
char *argv[6];
|
|
|
|
|
2017-01-23 05:38:16 +05:30
|
|
|
xmove_fd(network_fd, 3);
|
|
|
|
argv[0] = (char*)"ssl_client";
|
|
|
|
argv[1] = (char*)"-s3";
|
|
|
|
//TODO: if (!is_ip_address(servername))...
|
|
|
|
argv[2] = (char*)"-n";
|
|
|
|
argv[3] = servername;
|
2018-02-06 19:45:08 +05:30
|
|
|
argv[4] = (flags & TLSLOOP_EXIT_ON_LOCAL_EOF ? (char*)"-e" : NULL);
|
|
|
|
argv[5] = NULL;
|
2017-01-23 05:38:16 +05:30
|
|
|
BB_EXECVP(argv[0], argv);
|
|
|
|
bb_perror_msg_and_die("can't execute '%s'", argv[0]);
|
|
|
|
}
|
2014-02-24 04:09:47 +05:30
|
|
|
/* notreached */
|
|
|
|
}
|
|
|
|
|
|
|
|
/* Parent */
|
2017-01-23 05:38:16 +05:30
|
|
|
free(servername);
|
2014-02-24 04:09:47 +05:30
|
|
|
close(sp[1]);
|
|
|
|
xmove_fd(sp[0], network_fd);
|
|
|
|
}
|
|
|
|
#endif
|
|
|
|
|
2018-02-06 19:32:16 +05:30
|
|
|
static FILE* prepare_ftp_session(FILE **dfpp, struct host_info *target, len_and_sockaddr *lsa)
|
|
|
|
{
|
|
|
|
FILE *sfp;
|
2018-04-07 16:52:52 +05:30
|
|
|
char *pass;
|
2018-02-06 19:32:16 +05:30
|
|
|
int port;
|
|
|
|
|
|
|
|
sfp = open_socket(lsa);
|
2018-02-06 19:45:08 +05:30
|
|
|
#if ENABLE_FEATURE_WGET_HTTPS
|
|
|
|
if (target->protocol == P_FTPS)
|
|
|
|
spawn_ssl_client(target->host, fileno(sfp), TLSLOOP_EXIT_ON_LOCAL_EOF);
|
|
|
|
#endif
|
|
|
|
|
2018-02-06 19:32:16 +05:30
|
|
|
if (ftpcmd(NULL, NULL, sfp) != 220)
|
2018-02-12 21:16:13 +05:30
|
|
|
bb_error_msg_and_die("%s", G.wget_buf);
|
|
|
|
/* note: ftpcmd() sanitizes G.wget_buf, ok to print */
|
2018-02-06 19:32:16 +05:30
|
|
|
|
2018-04-07 16:52:52 +05:30
|
|
|
/* Split username:password pair */
|
|
|
|
pass = (char*)"busybox"; /* password for "anonymous" */
|
|
|
|
if (target->user) {
|
|
|
|
pass = strchr(target->user, ':');
|
|
|
|
if (pass)
|
|
|
|
*pass++ = '\0';
|
|
|
|
}
|
|
|
|
|
|
|
|
/* Log in */
|
|
|
|
switch (ftpcmd("USER ", target->user ?: "anonymous", sfp)) {
|
2018-02-06 19:32:16 +05:30
|
|
|
case 230:
|
|
|
|
break;
|
|
|
|
case 331:
|
2018-04-07 16:52:52 +05:30
|
|
|
if (ftpcmd("PASS ", pass, sfp) == 230)
|
2018-02-06 19:32:16 +05:30
|
|
|
break;
|
|
|
|
/* fall through (failed login) */
|
|
|
|
default:
|
2018-02-12 21:16:13 +05:30
|
|
|
bb_error_msg_and_die("ftp login: %s", G.wget_buf);
|
2018-02-06 19:32:16 +05:30
|
|
|
}
|
|
|
|
|
|
|
|
ftpcmd("TYPE I", NULL, sfp);
|
|
|
|
|
2018-04-07 16:52:52 +05:30
|
|
|
/* Query file size */
|
2018-02-06 19:32:16 +05:30
|
|
|
if (ftpcmd("SIZE ", target->path, sfp) == 213) {
|
|
|
|
G.content_len = BB_STRTOOFF(G.wget_buf + 4, NULL, 10);
|
|
|
|
if (G.content_len < 0 || errno) {
|
2018-04-08 21:36:24 +05:30
|
|
|
bb_error_msg_and_die("bad SIZE value '%s'", G.wget_buf + 4);
|
2018-02-06 19:32:16 +05:30
|
|
|
}
|
|
|
|
G.got_clen = 1;
|
|
|
|
}
|
|
|
|
|
2018-04-07 16:52:52 +05:30
|
|
|
/* Enter passive mode */
|
2018-02-06 20:18:12 +05:30
|
|
|
if (ENABLE_FEATURE_IPV6 && ftpcmd("EPSV", NULL, sfp) == 229) {
|
|
|
|
/* good */
|
|
|
|
} else
|
2018-02-06 19:32:16 +05:30
|
|
|
if (ftpcmd("PASV", NULL, sfp) != 227) {
|
|
|
|
pasv_error:
|
2018-02-12 21:16:13 +05:30
|
|
|
bb_error_msg_and_die("bad response to %s: %s", "PASV", G.wget_buf);
|
2018-02-06 19:32:16 +05:30
|
|
|
}
|
2018-02-06 20:18:12 +05:30
|
|
|
port = parse_pasv_epsv(G.wget_buf);
|
|
|
|
if (port < 0)
|
|
|
|
goto pasv_error;
|
|
|
|
|
2018-02-06 19:32:16 +05:30
|
|
|
set_nport(&lsa->u.sa, htons(port));
|
|
|
|
|
|
|
|
*dfpp = open_socket(lsa);
|
|
|
|
|
2018-02-07 01:19:27 +05:30
|
|
|
#if ENABLE_FEATURE_WGET_HTTPS
|
2018-02-08 04:58:30 +05:30
|
|
|
if (target->protocol == P_FTPS) {
|
|
|
|
/* "PROT P" enables encryption of data stream.
|
|
|
|
* Without it (or with "PROT C"), data is sent unencrypted.
|
|
|
|
*/
|
|
|
|
if (ftpcmd("PROT P", NULL, sfp) == 200)
|
|
|
|
spawn_ssl_client(target->host, fileno(*dfpp), /*flags*/ 0);
|
|
|
|
}
|
2018-02-07 01:19:27 +05:30
|
|
|
#endif
|
2018-02-06 19:45:08 +05:30
|
|
|
|
2018-02-06 19:32:16 +05:30
|
|
|
if (G.beg_range != 0) {
|
|
|
|
sprintf(G.wget_buf, "REST %"OFF_FMT"u", G.beg_range);
|
|
|
|
if (ftpcmd(G.wget_buf, NULL, sfp) == 350)
|
|
|
|
G.content_len -= G.beg_range;
|
|
|
|
else
|
|
|
|
reset_beg_range_to_zero();
|
|
|
|
}
|
|
|
|
|
2018-02-12 21:16:13 +05:30
|
|
|
//TODO: needs ftp-escaping 0xff and '\n' bytes here.
|
|
|
|
//Or disallow '\n' altogether via sanitize_string() in parse_url().
|
|
|
|
//But 0xff's are possible in valid utf8 filenames.
|
2018-02-06 19:32:16 +05:30
|
|
|
if (ftpcmd("RETR ", target->path, sfp) > 150)
|
2018-02-12 21:16:13 +05:30
|
|
|
bb_error_msg_and_die("bad response to %s: %s", "RETR", G.wget_buf);
|
2018-02-06 19:32:16 +05:30
|
|
|
|
|
|
|
return sfp;
|
|
|
|
}
|
|
|
|
|
2011-02-15 05:28:36 +05:30
|
|
|
static void NOINLINE retrieve_file_data(FILE *dfp)
|
2009-06-28 04:32:24 +05:30
|
|
|
{
|
2010-08-08 06:21:20 +05:30
|
|
|
#if ENABLE_FEATURE_WGET_STATUSBAR || ENABLE_FEATURE_WGET_TIMEOUT
|
|
|
|
# if ENABLE_FEATURE_WGET_TIMEOUT
|
2012-09-03 16:19:30 +05:30
|
|
|
unsigned second_cnt = G.timeout_seconds;
|
2010-08-08 06:21:20 +05:30
|
|
|
# endif
|
|
|
|
struct pollfd polldata;
|
|
|
|
|
|
|
|
polldata.fd = fileno(dfp);
|
|
|
|
polldata.events = POLLIN | POLLPRI;
|
|
|
|
#endif
|
|
|
|
progress_meter(PROGRESS_START);
|
2009-06-28 04:32:24 +05:30
|
|
|
|
|
|
|
if (G.chunked)
|
|
|
|
goto get_clen;
|
|
|
|
|
|
|
|
/* Loops only if chunked */
|
|
|
|
while (1) {
|
2011-02-12 02:53:23 +05:30
|
|
|
|
|
|
|
#if ENABLE_FEATURE_WGET_STATUSBAR || ENABLE_FEATURE_WGET_TIMEOUT
|
|
|
|
/* Must use nonblocking I/O, otherwise fread will loop
|
|
|
|
* and *block* until it reads full buffer,
|
|
|
|
* which messes up progress bar and/or timeout logic.
|
|
|
|
* Because of nonblocking I/O, we need to dance
|
|
|
|
* very carefully around EAGAIN. See explanation at
|
2012-09-03 16:19:30 +05:30
|
|
|
* clearerr() calls.
|
2011-02-12 02:53:23 +05:30
|
|
|
*/
|
|
|
|
ndelay_on(polldata.fd);
|
|
|
|
#endif
|
2009-12-11 17:06:10 +05:30
|
|
|
while (1) {
|
2009-06-28 04:32:24 +05:30
|
|
|
int n;
|
2009-12-11 17:06:10 +05:30
|
|
|
unsigned rdsz;
|
|
|
|
|
2010-08-08 06:21:20 +05:30
|
|
|
#if ENABLE_FEATURE_WGET_STATUSBAR || ENABLE_FEATURE_WGET_TIMEOUT
|
2011-02-12 02:12:00 +05:30
|
|
|
/* fread internally uses read loop, which in our case
|
|
|
|
* is usually exited when we get EAGAIN.
|
|
|
|
* In this case, libc sets error marker on the stream.
|
|
|
|
* Need to clear it before next fread to avoid possible
|
|
|
|
* rare false positive ferror below. Rare because usually
|
|
|
|
* fread gets more than zero bytes, and we don't fall
|
|
|
|
* into if (n <= 0) ...
|
|
|
|
*/
|
|
|
|
clearerr(dfp);
|
2011-02-12 02:31:33 +05:30
|
|
|
#endif
|
2012-09-03 16:19:30 +05:30
|
|
|
errno = 0;
|
|
|
|
rdsz = sizeof(G.wget_buf);
|
|
|
|
if (G.got_clen) {
|
|
|
|
if (G.content_len < (off_t)sizeof(G.wget_buf)) {
|
|
|
|
if ((int)G.content_len <= 0)
|
|
|
|
break;
|
|
|
|
rdsz = (unsigned)G.content_len;
|
|
|
|
}
|
|
|
|
}
|
2011-02-10 14:25:05 +05:30
|
|
|
n = fread(G.wget_buf, 1, rdsz, dfp);
|
2012-09-03 16:19:30 +05:30
|
|
|
|
|
|
|
if (n > 0) {
|
|
|
|
xwrite(G.output_fd, G.wget_buf, n);
|
|
|
|
#if ENABLE_FEATURE_WGET_STATUSBAR
|
|
|
|
G.transferred += n;
|
|
|
|
#endif
|
|
|
|
if (G.got_clen) {
|
|
|
|
G.content_len -= n;
|
|
|
|
if (G.content_len == 0)
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
#if ENABLE_FEATURE_WGET_TIMEOUT
|
|
|
|
second_cnt = G.timeout_seconds;
|
|
|
|
#endif
|
2014-03-27 21:20:29 +05:30
|
|
|
goto bump;
|
2012-09-03 16:19:30 +05:30
|
|
|
}
|
|
|
|
|
|
|
|
/* n <= 0.
|
|
|
|
* man fread:
|
2011-02-12 02:12:00 +05:30
|
|
|
* If error occurs, or EOF is reached, the return value
|
|
|
|
* is a short item count (or zero).
|
|
|
|
* fread does not distinguish between EOF and error.
|
|
|
|
*/
|
2012-09-03 16:19:30 +05:30
|
|
|
if (errno != EAGAIN) {
|
|
|
|
if (ferror(dfp)) {
|
|
|
|
progress_meter(PROGRESS_END);
|
2011-02-12 02:12:00 +05:30
|
|
|
bb_perror_msg_and_die(bb_msg_read_error);
|
2012-09-03 16:19:30 +05:30
|
|
|
}
|
2011-02-12 02:12:00 +05:30
|
|
|
break; /* EOF, not error */
|
2009-06-28 04:32:24 +05:30
|
|
|
}
|
2011-02-12 02:12:00 +05:30
|
|
|
|
2012-09-03 16:19:30 +05:30
|
|
|
#if ENABLE_FEATURE_WGET_STATUSBAR || ENABLE_FEATURE_WGET_TIMEOUT
|
|
|
|
/* It was EAGAIN. There is no data. Wait up to one second
|
|
|
|
* then abort if timed out, or update the bar and try reading again.
|
|
|
|
*/
|
|
|
|
if (safe_poll(&polldata, 1, 1000) == 0) {
|
|
|
|
# if ENABLE_FEATURE_WGET_TIMEOUT
|
|
|
|
if (second_cnt != 0 && --second_cnt == 0) {
|
|
|
|
progress_meter(PROGRESS_END);
|
|
|
|
bb_error_msg_and_die("download timed out");
|
|
|
|
}
|
|
|
|
# endif
|
|
|
|
/* We used to loop back to poll here,
|
|
|
|
* but there is no great harm in letting fread
|
|
|
|
* to try reading anyway.
|
|
|
|
*/
|
|
|
|
}
|
2014-03-27 21:20:29 +05:30
|
|
|
#endif
|
|
|
|
bump:
|
2012-09-03 16:19:30 +05:30
|
|
|
/* Need to do it _every_ second for "stalled" indicator
|
|
|
|
* to be shown properly.
|
|
|
|
*/
|
2010-08-08 06:21:20 +05:30
|
|
|
progress_meter(PROGRESS_BUMP);
|
2012-09-03 16:19:30 +05:30
|
|
|
} /* while (reading data) */
|
|
|
|
|
2011-02-12 02:53:23 +05:30
|
|
|
#if ENABLE_FEATURE_WGET_STATUSBAR || ENABLE_FEATURE_WGET_TIMEOUT
|
|
|
|
clearerr(dfp);
|
2011-02-12 03:36:21 +05:30
|
|
|
ndelay_off(polldata.fd); /* else fgets can get very unhappy */
|
2011-02-12 02:53:23 +05:30
|
|
|
#endif
|
2009-06-28 04:32:24 +05:30
|
|
|
if (!G.chunked)
|
|
|
|
break;
|
|
|
|
|
2018-04-08 21:36:24 +05:30
|
|
|
/* Each chunk ends with "\r\n" - eat it */
|
|
|
|
fgets_trim_sanitize(dfp, NULL);
|
2009-06-28 04:32:24 +05:30
|
|
|
get_clen:
|
2018-04-08 21:36:24 +05:30
|
|
|
/* chunk size format is "HEXNUM[;name[=val]]\r\n" */
|
2018-02-12 21:16:13 +05:30
|
|
|
fgets_trim_sanitize(dfp, NULL);
|
2018-04-08 21:36:24 +05:30
|
|
|
errno = 0;
|
2011-02-10 10:59:06 +05:30
|
|
|
G.content_len = STRTOOFF(G.wget_buf, NULL, 16);
|
2018-04-08 21:36:24 +05:30
|
|
|
/*
|
|
|
|
* Had a bug with inputs like "ffffffff0001f400"
|
|
|
|
* smashing the heap later. Ensure >= 0.
|
|
|
|
*/
|
|
|
|
if (G.content_len < 0 || errno)
|
|
|
|
bb_error_msg_and_die("bad chunk length '%s'", G.wget_buf);
|
2009-12-11 17:06:10 +05:30
|
|
|
if (G.content_len == 0)
|
2009-06-28 04:32:24 +05:30
|
|
|
break; /* all done! */
|
2009-12-11 17:06:10 +05:30
|
|
|
G.got_clen = 1;
|
2012-09-03 16:19:30 +05:30
|
|
|
/*
|
|
|
|
* Note that fgets may result in some data being buffered in dfp.
|
|
|
|
* We loop back to fread, which will retrieve this data.
|
|
|
|
* Also note that code has to be arranged so that fread
|
|
|
|
* is done _before_ one-second poll wait - poll doesn't know
|
|
|
|
* about stdio buffering and can result in spurious one second waits!
|
|
|
|
*/
|
2009-06-28 04:32:24 +05:30
|
|
|
}
|
|
|
|
|
2012-06-17 23:22:25 +05:30
|
|
|
/* If -c failed, we restart from the beginning,
|
|
|
|
* but we do not truncate file then, we do it only now, at the end.
|
|
|
|
* This lets user to ^C if his 99% complete 10 GB file download
|
|
|
|
* failed to restart *without* losing the almost complete file.
|
|
|
|
*/
|
|
|
|
{
|
|
|
|
off_t pos = lseek(G.output_fd, 0, SEEK_CUR);
|
|
|
|
if (pos != (off_t)-1)
|
|
|
|
ftruncate(G.output_fd, pos);
|
|
|
|
}
|
|
|
|
|
2011-02-13 07:03:11 +05:30
|
|
|
/* Draw full bar and free its resources */
|
2011-02-15 05:28:36 +05:30
|
|
|
G.chunked = 0; /* makes it show 100% even for chunked download */
|
|
|
|
G.got_clen = 1; /* makes it show 100% even for download of (formerly) unknown size */
|
2010-08-08 06:21:20 +05:30
|
|
|
progress_meter(PROGRESS_END);
|
2009-06-28 04:32:24 +05:30
|
|
|
}
|
|
|
|
|
2011-02-17 00:39:36 +05:30
|
|
|
static void download_one_url(const char *url)
|
2007-09-24 23:54:17 +05:30
|
|
|
{
|
2011-02-13 07:03:11 +05:30
|
|
|
bool use_proxy; /* Use proxies if env vars are set */
|
2009-06-28 07:03:57 +05:30
|
|
|
int redir_limit;
|
2011-02-13 07:03:11 +05:30
|
|
|
len_and_sockaddr *lsa;
|
2009-06-28 04:32:24 +05:30
|
|
|
FILE *sfp; /* socket to web/ftp server */
|
2007-09-27 20:37:23 +05:30
|
|
|
FILE *dfp; /* socket to ftp server (data) */
|
2011-02-13 07:03:11 +05:30
|
|
|
char *fname_out_alloc;
|
2011-12-18 09:41:56 +05:30
|
|
|
char *redirected_path = NULL;
|
2011-02-13 07:03:11 +05:30
|
|
|
struct host_info server;
|
|
|
|
struct host_info target;
|
2009-07-01 00:06:27 +05:30
|
|
|
|
2011-02-13 07:03:11 +05:30
|
|
|
server.allocated = NULL;
|
|
|
|
target.allocated = NULL;
|
|
|
|
server.user = NULL;
|
2009-10-05 05:48:01 +05:30
|
|
|
target.user = NULL;
|
2011-02-13 07:03:11 +05:30
|
|
|
|
|
|
|
parse_url(url, &target);
|
2001-04-06 03:15:54 +05:30
|
|
|
|
2007-04-10 15:07:29 +05:30
|
|
|
/* Use the proxy if necessary */
|
2011-02-13 07:03:11 +05:30
|
|
|
use_proxy = (strcmp(G.proxy_flag, "off") != 0);
|
2004-02-22 05:57:34 +05:30
|
|
|
if (use_proxy) {
|
2018-03-05 04:47:02 +05:30
|
|
|
char *proxy = getenv(target.protocol[0] == 'f' ? "ftp_proxy" : "http_proxy");
|
2014-02-22 18:42:29 +05:30
|
|
|
//FIXME: what if protocol is https? Ok to use http_proxy?
|
2011-02-15 05:28:36 +05:30
|
|
|
use_proxy = (proxy && proxy[0]);
|
|
|
|
if (use_proxy)
|
2006-10-07 19:58:55 +05:30
|
|
|
parse_url(proxy, &server);
|
2002-05-15 05:06:45 +05:30
|
|
|
}
|
2009-07-01 00:06:27 +05:30
|
|
|
if (!use_proxy) {
|
2018-03-05 04:47:02 +05:30
|
|
|
server.protocol = target.protocol;
|
2009-07-01 00:06:27 +05:30
|
|
|
server.port = target.port;
|
|
|
|
if (ENABLE_FEATURE_IPV6) {
|
2011-02-13 07:03:11 +05:30
|
|
|
//free(server.allocated); - can't be non-NULL
|
|
|
|
server.host = server.allocated = xstrdup(target.host);
|
2009-07-01 00:06:27 +05:30
|
|
|
} else {
|
|
|
|
server.host = target.host;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
if (ENABLE_FEATURE_IPV6)
|
|
|
|
strip_ipv6_scope_id(target.host);
|
2004-03-15 13:59:22 +05:30
|
|
|
|
2011-02-13 07:03:11 +05:30
|
|
|
/* If there was no -O FILE, guess output filename */
|
|
|
|
fname_out_alloc = NULL;
|
2011-02-13 07:19:43 +05:30
|
|
|
if (!(option_mask32 & WGET_OPT_OUTNAME)) {
|
2011-02-13 07:03:11 +05:30
|
|
|
G.fname_out = bb_get_last_path_component_nostrip(target.path);
|
2007-09-24 23:57:04 +05:30
|
|
|
/* handle "wget http://kernel.org//" */
|
2011-02-13 07:03:11 +05:30
|
|
|
if (G.fname_out[0] == '/' || !G.fname_out[0])
|
|
|
|
G.fname_out = (char*)"index.html";
|
2007-09-24 23:57:04 +05:30
|
|
|
/* -P DIR is considered only if there was no -O FILE */
|
2012-06-17 23:51:30 +05:30
|
|
|
if (G.dir_prefix)
|
|
|
|
G.fname_out = fname_out_alloc = concat_path_file(G.dir_prefix, G.fname_out);
|
2011-03-21 04:59:37 +05:30
|
|
|
else {
|
2012-06-17 23:51:30 +05:30
|
|
|
/* redirects may free target.path later, need to make a copy */
|
|
|
|
G.fname_out = fname_out_alloc = xstrdup(G.fname_out);
|
2011-03-21 04:59:37 +05:30
|
|
|
}
|
2007-09-24 23:57:04 +05:30
|
|
|
}
|
2006-11-21 16:13:02 +05:30
|
|
|
#if ENABLE_FEATURE_WGET_STATUSBAR
|
2011-02-13 07:03:11 +05:30
|
|
|
G.curfile = bb_get_last_path_component_nostrip(G.fname_out);
|
2000-12-09 22:25:35 +05:30
|
|
|
#endif
|
2007-09-24 23:57:04 +05:30
|
|
|
|
2007-04-10 15:07:29 +05:30
|
|
|
/* Determine where to start transfer */
|
2011-02-15 05:28:36 +05:30
|
|
|
G.beg_range = 0;
|
2011-02-13 07:03:11 +05:30
|
|
|
if (option_mask32 & WGET_OPT_CONTINUE) {
|
2011-02-15 05:28:36 +05:30
|
|
|
G.output_fd = open(G.fname_out, O_WRONLY);
|
|
|
|
if (G.output_fd >= 0) {
|
|
|
|
G.beg_range = xlseek(G.output_fd, 0, SEEK_END);
|
2006-09-23 23:19:09 +05:30
|
|
|
}
|
|
|
|
/* File doesn't exist. We do not create file here yet.
|
2011-02-13 04:10:31 +05:30
|
|
|
* We are not sure it exists on remote side */
|
2000-09-04 20:45:55 +05:30
|
|
|
}
|
|
|
|
|
2009-06-28 07:03:57 +05:30
|
|
|
redir_limit = 5;
|
|
|
|
resolve_lsa:
|
2007-02-04 08:09:08 +05:30
|
|
|
lsa = xhost2sockaddr(server.host, server.port);
|
2011-02-13 07:03:11 +05:30
|
|
|
if (!(option_mask32 & WGET_OPT_QUIET)) {
|
2009-06-28 07:03:57 +05:30
|
|
|
char *s = xmalloc_sockaddr2dotted(&lsa->u.sa);
|
|
|
|
fprintf(stderr, "Connecting to %s (%s)\n", server.host, s);
|
|
|
|
free(s);
|
2003-10-31 15:01:46 +05:30
|
|
|
}
|
2009-06-28 07:03:57 +05:30
|
|
|
establish_session:
|
2011-02-15 05:28:36 +05:30
|
|
|
/*G.content_len = 0; - redundant, got_clen = 0 is enough */
|
|
|
|
G.got_clen = 0;
|
|
|
|
G.chunked = 0;
|
2018-02-06 19:45:08 +05:30
|
|
|
if (use_proxy || target.protocol[0] != 'f' /*not ftp[s]*/) {
|
2001-04-06 03:15:54 +05:30
|
|
|
/*
|
|
|
|
* HTTP session
|
|
|
|
*/
|
2009-06-28 07:03:57 +05:30
|
|
|
char *str;
|
2009-06-28 04:32:24 +05:30
|
|
|
int status;
|
2001-04-06 03:15:54 +05:30
|
|
|
|
2014-02-22 18:42:29 +05:30
|
|
|
/* Open socket to http(s) server */
|
2015-10-07 05:09:40 +05:30
|
|
|
#if ENABLE_FEATURE_WGET_OPENSSL
|
2017-01-23 05:38:16 +05:30
|
|
|
/* openssl (and maybe internal TLS) support is configured */
|
2018-03-05 04:47:02 +05:30
|
|
|
if (server.protocol == P_HTTPS) {
|
2015-10-07 05:09:40 +05:30
|
|
|
/* openssl-based helper
|
|
|
|
* Inconvenient API since we can't give it an open fd
|
|
|
|
*/
|
2015-10-07 06:10:53 +05:30
|
|
|
int fd = spawn_https_helper_openssl(server.host, server.port);
|
2017-01-23 05:38:16 +05:30
|
|
|
# if ENABLE_FEATURE_WGET_HTTPS
|
|
|
|
if (fd < 0) { /* no openssl? try internal */
|
2015-10-07 06:10:53 +05:30
|
|
|
sfp = open_socket(lsa);
|
2018-02-06 19:45:08 +05:30
|
|
|
spawn_ssl_client(server.host, fileno(sfp), /*flags*/ 0);
|
2015-10-07 06:10:53 +05:30
|
|
|
goto socket_opened;
|
|
|
|
}
|
|
|
|
# else
|
|
|
|
/* We don't check for exec("openssl") failure in this case */
|
|
|
|
# endif
|
2014-02-22 18:42:29 +05:30
|
|
|
sfp = fdopen(fd, "r+");
|
|
|
|
if (!sfp)
|
2018-04-01 23:29:37 +05:30
|
|
|
bb_die_memory_exhausted();
|
2015-10-07 06:10:53 +05:30
|
|
|
goto socket_opened;
|
|
|
|
}
|
|
|
|
sfp = open_socket(lsa);
|
|
|
|
socket_opened:
|
2017-01-23 05:38:16 +05:30
|
|
|
#elif ENABLE_FEATURE_WGET_HTTPS
|
|
|
|
/* Only internal TLS support is configured */
|
2015-10-07 06:10:53 +05:30
|
|
|
sfp = open_socket(lsa);
|
2018-03-05 04:47:02 +05:30
|
|
|
if (server.protocol == P_HTTPS)
|
2018-02-06 19:45:08 +05:30
|
|
|
spawn_ssl_client(server.host, fileno(sfp), /*flags*/ 0);
|
2015-10-07 06:10:53 +05:30
|
|
|
#else
|
|
|
|
/* ssl (https) support is not configured */
|
|
|
|
sfp = open_socket(lsa);
|
2014-02-24 04:09:47 +05:30
|
|
|
#endif
|
2009-06-28 07:03:57 +05:30
|
|
|
/* Send HTTP request */
|
|
|
|
if (use_proxy) {
|
2015-02-19 01:11:02 +05:30
|
|
|
SENDFMT(sfp, "GET %s://%s/%s HTTP/1.1\r\n",
|
2014-02-22 18:42:29 +05:30
|
|
|
target.protocol, target.host,
|
2009-06-28 07:03:57 +05:30
|
|
|
target.path);
|
|
|
|
} else {
|
2015-02-19 01:11:02 +05:30
|
|
|
SENDFMT(sfp, "%s /%s HTTP/1.1\r\n",
|
2014-02-22 18:42:29 +05:30
|
|
|
(option_mask32 & WGET_OPT_POST_DATA) ? "POST" : "GET",
|
|
|
|
target.path);
|
2009-06-28 07:03:57 +05:30
|
|
|
}
|
2015-02-19 01:11:02 +05:30
|
|
|
if (!USR_HEADER_HOST)
|
|
|
|
SENDFMT(sfp, "Host: %s\r\n", target.host);
|
|
|
|
if (!USR_HEADER_USER_AGENT)
|
|
|
|
SENDFMT(sfp, "User-Agent: %s\r\n", G.user_agent);
|
2001-04-06 03:15:54 +05:30
|
|
|
|
2011-02-10 17:53:45 +05:30
|
|
|
/* Ask server to close the connection as soon as we are done
|
|
|
|
* (IOW: we do not intend to send more requests)
|
|
|
|
*/
|
2015-02-19 01:11:02 +05:30
|
|
|
SENDFMT(sfp, "Connection: close\r\n");
|
2011-02-10 17:53:45 +05:30
|
|
|
|
2006-11-21 16:13:02 +05:30
|
|
|
#if ENABLE_FEATURE_WGET_AUTHENTICATION
|
2015-02-19 01:11:02 +05:30
|
|
|
if (target.user && !USR_HEADER_AUTH) {
|
|
|
|
SENDFMT(sfp, "Proxy-Authorization: Basic %s\r\n"+6,
|
2011-02-10 10:59:06 +05:30
|
|
|
base64enc(target.user));
|
2009-06-28 07:03:57 +05:30
|
|
|
}
|
2015-02-19 01:11:02 +05:30
|
|
|
if (use_proxy && server.user && !USR_HEADER_PROXY_AUTH) {
|
|
|
|
SENDFMT(sfp, "Proxy-Authorization: Basic %s\r\n",
|
2011-02-10 10:59:06 +05:30
|
|
|
base64enc(server.user));
|
2009-06-28 07:03:57 +05:30
|
|
|
}
|
2001-04-06 03:15:54 +05:30
|
|
|
#endif
|
2000-09-04 20:45:55 +05:30
|
|
|
|
2015-02-19 01:11:02 +05:30
|
|
|
if (G.beg_range != 0 && !USR_HEADER_RANGE)
|
|
|
|
SENDFMT(sfp, "Range: bytes=%"OFF_FMT"u-\r\n", G.beg_range);
|
2011-02-10 17:53:45 +05:30
|
|
|
|
2006-10-25 06:03:44 +05:30
|
|
|
#if ENABLE_FEATURE_WGET_LONG_OPTIONS
|
2015-02-19 01:11:02 +05:30
|
|
|
if (G.extra_headers) {
|
|
|
|
log_io(G.extra_headers);
|
2011-02-13 07:03:11 +05:30
|
|
|
fputs(G.extra_headers, sfp);
|
2015-02-19 01:11:02 +05:30
|
|
|
}
|
2009-06-28 07:03:57 +05:30
|
|
|
|
2011-02-13 07:03:11 +05:30
|
|
|
if (option_mask32 & WGET_OPT_POST_DATA) {
|
2015-02-19 01:11:02 +05:30
|
|
|
SENDFMT(sfp,
|
2011-02-10 17:53:45 +05:30
|
|
|
"Content-Type: application/x-www-form-urlencoded\r\n"
|
|
|
|
"Content-Length: %u\r\n"
|
|
|
|
"\r\n"
|
|
|
|
"%s",
|
2011-03-28 02:03:13 +05:30
|
|
|
(int) strlen(G.post_data), G.post_data
|
2011-02-10 17:53:45 +05:30
|
|
|
);
|
2009-06-28 07:03:57 +05:30
|
|
|
} else
|
2006-10-25 06:03:44 +05:30
|
|
|
#endif
|
2011-02-10 17:53:45 +05:30
|
|
|
{
|
2015-02-19 01:11:02 +05:30
|
|
|
SENDFMT(sfp, "\r\n");
|
2009-06-28 07:03:57 +05:30
|
|
|
}
|
2001-04-06 03:15:54 +05:30
|
|
|
|
2010-09-23 18:48:41 +05:30
|
|
|
fflush(sfp);
|
wget: fix for brain-damaged HTTP servers. Closes 9471
write(3, "GET / HTTP/1.1\r\nUser-Agent: Wget\r\nConnection: close\r\n\r\n", 74) = 74
shutdown(3, SHUT_WR) = 0
alarm(900) = 900
read(3, "", 1024) = 0
write(2, "wget: error getting response\n", 29) = 29
exit(1)
The peer simply does not return anything. It closes its connection.
Probably it detects wget closing its writing end: shutdown(3, SHUT_WR).
The point it, closing write side of the socket is _valid_ for HTTP.
wget sent the full request, it won't be sending anything more:
it will only receive the response, and that's it.
Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>
2017-01-12 00:46:45 +05:30
|
|
|
|
2017-01-16 22:01:05 +05:30
|
|
|
/* Tried doing this unconditionally.
|
|
|
|
* Cloudflare and nginx/1.11.5 are shocked to see SHUT_WR on non-HTTPS.
|
|
|
|
*/
|
wget: fix for brain-damaged HTTP servers. Closes 9471
write(3, "GET / HTTP/1.1\r\nUser-Agent: Wget\r\nConnection: close\r\n\r\n", 74) = 74
shutdown(3, SHUT_WR) = 0
alarm(900) = 900
read(3, "", 1024) = 0
write(2, "wget: error getting response\n", 29) = 29
exit(1)
The peer simply does not return anything. It closes its connection.
Probably it detects wget closing its writing end: shutdown(3, SHUT_WR).
The point it, closing write side of the socket is _valid_ for HTTP.
wget sent the full request, it won't be sending anything more:
it will only receive the response, and that's it.
Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>
2017-01-12 00:46:45 +05:30
|
|
|
#if SSL_SUPPORTED
|
|
|
|
if (target.protocol == P_HTTPS) {
|
|
|
|
/* If we use SSL helper, keeping our end of the socket open for writing
|
|
|
|
* makes our end (i.e. the same fd!) readable (EAGAIN instead of EOF)
|
|
|
|
* even after child closes its copy of the fd.
|
|
|
|
* This helps:
|
|
|
|
*/
|
|
|
|
shutdown(fileno(sfp), SHUT_WR);
|
|
|
|
}
|
|
|
|
#endif
|
2010-09-23 18:48:41 +05:30
|
|
|
|
2009-06-28 07:03:57 +05:30
|
|
|
/*
|
|
|
|
* Retrieve HTTP response line and check for "200" status code.
|
|
|
|
*/
|
2006-10-15 22:35:55 +05:30
|
|
|
read_response:
|
2018-02-12 21:16:13 +05:30
|
|
|
fgets_trim_sanitize(sfp, " %s\n");
|
2009-06-28 07:03:57 +05:30
|
|
|
|
2011-02-10 10:59:06 +05:30
|
|
|
str = G.wget_buf;
|
2009-06-28 07:03:57 +05:30
|
|
|
str = skip_non_whitespace(str);
|
|
|
|
str = skip_whitespace(str);
|
|
|
|
// FIXME: no error check
|
|
|
|
// xatou wouldn't work: "200 OK"
|
|
|
|
status = atoi(str);
|
|
|
|
switch (status) {
|
|
|
|
case 0:
|
|
|
|
case 100:
|
2018-02-12 21:16:13 +05:30
|
|
|
while (get_sanitized_hdr(sfp) != NULL)
|
2009-06-28 07:03:57 +05:30
|
|
|
/* eat all remaining headers */;
|
|
|
|
goto read_response;
|
2016-08-31 16:58:53 +05:30
|
|
|
|
|
|
|
/* Success responses */
|
2009-06-28 07:03:57 +05:30
|
|
|
case 200:
|
2016-08-31 16:58:53 +05:30
|
|
|
/* fall through */
|
|
|
|
case 201: /* 201 Created */
|
|
|
|
/* "The request has been fulfilled and resulted in a new resource being created" */
|
2016-09-01 14:46:22 +05:30
|
|
|
/* Standard wget is reported to treat this as success */
|
2016-08-31 16:58:53 +05:30
|
|
|
/* fall through */
|
|
|
|
case 202: /* 202 Accepted */
|
|
|
|
/* "The request has been accepted for processing, but the processing has not been completed" */
|
|
|
|
/* Treat as success: fall through */
|
|
|
|
case 203: /* 203 Non-Authoritative Information */
|
|
|
|
/* "Use of this response code is not required and is only appropriate when the response would otherwise be 200 (OK)" */
|
|
|
|
/* fall through */
|
|
|
|
case 204: /* 204 No Content */
|
2008-06-22 21:58:02 +05:30
|
|
|
/*
|
|
|
|
Response 204 doesn't say "null file", it says "metadata
|
|
|
|
has changed but data didn't":
|
|
|
|
|
|
|
|
"10.2.5 204 No Content
|
|
|
|
The server has fulfilled the request but does not need to return
|
|
|
|
an entity-body, and might want to return updated metainformation.
|
|
|
|
The response MAY include new or updated metainformation in the form
|
|
|
|
of entity-headers, which if present SHOULD be associated with
|
|
|
|
the requested variant.
|
|
|
|
|
|
|
|
If the client is a user agent, it SHOULD NOT change its document
|
|
|
|
view from that which caused the request to be sent. This response
|
|
|
|
is primarily intended to allow input for actions to take place
|
|
|
|
without causing a change to the user agent's active document view,
|
|
|
|
although any new or updated metainformation SHOULD be applied
|
|
|
|
to the document currently in the user agent's active view.
|
|
|
|
|
|
|
|
The 204 response MUST NOT include a message-body, and thus
|
|
|
|
is always terminated by the first empty line after the header fields."
|
|
|
|
|
|
|
|
However, in real world it was observed that some web servers
|
|
|
|
(e.g. Boa/0.94.14rc21) simply use code 204 when file size is zero.
|
|
|
|
*/
|
2012-06-13 21:01:07 +05:30
|
|
|
if (G.beg_range != 0) {
|
|
|
|
/* "Range:..." was not honored by the server.
|
|
|
|
* Restart download from the beginning.
|
|
|
|
*/
|
|
|
|
reset_beg_range_to_zero();
|
|
|
|
}
|
2009-06-28 07:03:57 +05:30
|
|
|
break;
|
2016-08-31 16:58:53 +05:30
|
|
|
/* 205 Reset Content ?? what to do on this ?? */
|
|
|
|
|
2010-10-29 15:16:52 +05:30
|
|
|
case 300: /* redirection */
|
2009-06-28 07:03:57 +05:30
|
|
|
case 301:
|
|
|
|
case 302:
|
|
|
|
case 303:
|
|
|
|
break;
|
2016-08-31 16:58:53 +05:30
|
|
|
|
2012-06-13 20:59:41 +05:30
|
|
|
case 206: /* Partial Content */
|
|
|
|
if (G.beg_range != 0)
|
|
|
|
/* "Range:..." worked. Good. */
|
2006-10-15 22:35:55 +05:30
|
|
|
break;
|
2012-06-13 20:59:41 +05:30
|
|
|
/* Partial Content even though we did not ask for it??? */
|
2009-06-28 07:03:57 +05:30
|
|
|
/* fall through */
|
|
|
|
default:
|
2018-02-12 21:16:13 +05:30
|
|
|
bb_error_msg_and_die("server returned error: %s", G.wget_buf);
|
2009-06-28 07:03:57 +05:30
|
|
|
}
|
2004-03-15 13:59:22 +05:30
|
|
|
|
2009-06-28 07:03:57 +05:30
|
|
|
/*
|
|
|
|
* Retrieve HTTP headers.
|
|
|
|
*/
|
2018-02-12 21:16:13 +05:30
|
|
|
while ((str = get_sanitized_hdr(sfp)) != NULL) {
|
2011-02-13 07:03:11 +05:30
|
|
|
static const char keywords[] ALIGN1 =
|
|
|
|
"content-length\0""transfer-encoding\0""location\0";
|
|
|
|
enum {
|
|
|
|
KEY_content_length = 1, KEY_transfer_encoding, KEY_location
|
|
|
|
};
|
2009-08-22 23:49:48 +05:30
|
|
|
smalluint key;
|
2011-02-13 07:03:11 +05:30
|
|
|
|
2018-02-12 21:16:13 +05:30
|
|
|
/* get_sanitized_hdr converted "FOO:" string to lowercase */
|
2011-02-13 07:03:11 +05:30
|
|
|
|
2009-08-22 23:49:48 +05:30
|
|
|
/* strip trailing whitespace */
|
|
|
|
char *s = strchrnul(str, '\0') - 1;
|
|
|
|
while (s >= str && (*s == ' ' || *s == '\t')) {
|
|
|
|
*s = '\0';
|
|
|
|
s--;
|
|
|
|
}
|
2011-02-10 10:59:06 +05:30
|
|
|
key = index_in_strings(keywords, G.wget_buf) + 1;
|
2009-06-28 07:03:57 +05:30
|
|
|
if (key == KEY_content_length) {
|
2009-12-11 17:06:10 +05:30
|
|
|
G.content_len = BB_STRTOOFF(str, NULL, 10);
|
|
|
|
if (G.content_len < 0 || errno) {
|
2018-02-12 21:16:13 +05:30
|
|
|
bb_error_msg_and_die("content-length %s is garbage", str);
|
2001-04-10 23:47:05 +05:30
|
|
|
}
|
2009-06-28 07:03:57 +05:30
|
|
|
G.got_clen = 1;
|
|
|
|
continue;
|
|
|
|
}
|
|
|
|
if (key == KEY_transfer_encoding) {
|
2011-02-13 07:03:11 +05:30
|
|
|
if (strcmp(str_tolower(str), "chunked") != 0)
|
2018-02-12 21:16:13 +05:30
|
|
|
bb_error_msg_and_die("transfer encoding '%s' is not supported", str);
|
2011-02-13 07:03:11 +05:30
|
|
|
G.chunked = 1;
|
2009-06-28 07:03:57 +05:30
|
|
|
}
|
|
|
|
if (key == KEY_location && status >= 300) {
|
|
|
|
if (--redir_limit == 0)
|
|
|
|
bb_error_msg_and_die("too many redirections");
|
|
|
|
fclose(sfp);
|
2011-02-13 07:03:11 +05:30
|
|
|
if (str[0] == '/') {
|
2011-12-18 09:41:56 +05:30
|
|
|
free(redirected_path);
|
2018-02-12 21:16:13 +05:30
|
|
|
target.path = redirected_path = xstrdup(str + 1);
|
2009-06-28 07:03:57 +05:30
|
|
|
/* lsa stays the same: it's on the same server */
|
2011-02-13 07:03:11 +05:30
|
|
|
} else {
|
2009-06-28 07:03:57 +05:30
|
|
|
parse_url(str, &target);
|
|
|
|
if (!use_proxy) {
|
2014-02-03 18:39:42 +05:30
|
|
|
/* server.user remains untouched */
|
2011-02-13 07:03:11 +05:30
|
|
|
free(server.allocated);
|
2011-02-15 04:26:07 +05:30
|
|
|
server.allocated = NULL;
|
2018-07-02 22:01:02 +05:30
|
|
|
server.protocol = target.protocol;
|
2009-06-28 07:03:57 +05:30
|
|
|
server.host = target.host;
|
2009-07-01 00:06:27 +05:30
|
|
|
/* strip_ipv6_scope_id(target.host); - no! */
|
|
|
|
/* we assume remote never gives us IPv6 addr with scope id */
|
2009-06-28 07:03:57 +05:30
|
|
|
server.port = target.port;
|
2007-01-12 16:05:23 +05:30
|
|
|
free(lsa);
|
2009-06-28 07:03:57 +05:30
|
|
|
goto resolve_lsa;
|
|
|
|
} /* else: lsa stays the same: we use proxy */
|
2001-04-06 03:15:54 +05:30
|
|
|
}
|
2009-06-28 07:03:57 +05:30
|
|
|
goto establish_session;
|
2001-04-06 03:15:54 +05:30
|
|
|
}
|
2009-06-28 07:03:57 +05:30
|
|
|
}
|
|
|
|
// if (status >= 300)
|
|
|
|
// bb_error_msg_and_die("bad redirection (no Location: header from server)");
|
2004-03-15 13:59:22 +05:30
|
|
|
|
2009-06-28 07:03:57 +05:30
|
|
|
/* For HTTP, data is pumped over the same connection */
|
2001-04-06 03:15:54 +05:30
|
|
|
dfp = sfp;
|
2006-10-07 19:58:55 +05:30
|
|
|
} else {
|
2001-04-06 03:15:54 +05:30
|
|
|
/*
|
|
|
|
* FTP session
|
|
|
|
*/
|
2009-06-28 04:32:24 +05:30
|
|
|
sfp = prepare_ftp_session(&dfp, &target, lsa);
|
2000-09-04 20:45:55 +05:30
|
|
|
}
|
2007-09-24 20:34:00 +05:30
|
|
|
|
2011-02-13 07:03:11 +05:30
|
|
|
free(lsa);
|
2001-04-06 03:15:54 +05:30
|
|
|
|
2011-02-13 07:19:43 +05:30
|
|
|
if (!(option_mask32 & WGET_OPT_SPIDER)) {
|
2011-02-15 05:28:36 +05:30
|
|
|
if (G.output_fd < 0)
|
|
|
|
G.output_fd = xopen(G.fname_out, G.o_flags);
|
|
|
|
retrieve_file_data(dfp);
|
|
|
|
if (!(option_mask32 & WGET_OPT_OUTNAME)) {
|
|
|
|
xclose(G.output_fd);
|
|
|
|
G.output_fd = -1;
|
2011-02-13 07:19:43 +05:30
|
|
|
}
|
2007-09-27 20:37:23 +05:30
|
|
|
}
|
2006-10-01 16:28:54 +05:30
|
|
|
|
2009-06-28 07:03:57 +05:30
|
|
|
if (dfp != sfp) {
|
2011-02-13 07:03:11 +05:30
|
|
|
/* It's ftp. Close data connection properly */
|
2001-04-06 03:15:54 +05:30
|
|
|
fclose(dfp);
|
2011-02-10 10:59:06 +05:30
|
|
|
if (ftpcmd(NULL, NULL, sfp) != 226)
|
2018-02-12 21:16:13 +05:30
|
|
|
bb_error_msg_and_die("ftp error: %s", G.wget_buf);
|
2011-02-10 10:59:06 +05:30
|
|
|
/* ftpcmd("QUIT", NULL, sfp); - why bother? */
|
2001-04-06 03:15:54 +05:30
|
|
|
}
|
2011-02-13 07:03:11 +05:30
|
|
|
fclose(sfp);
|
2007-09-24 20:34:00 +05:30
|
|
|
|
2011-02-13 07:19:43 +05:30
|
|
|
free(server.allocated);
|
|
|
|
free(target.allocated);
|
2014-02-03 18:39:42 +05:30
|
|
|
free(server.user);
|
|
|
|
free(target.user);
|
2011-02-13 07:19:43 +05:30
|
|
|
free(fname_out_alloc);
|
2011-12-18 09:41:56 +05:30
|
|
|
free(redirected_path);
|
2001-04-06 03:15:54 +05:30
|
|
|
}
|
2011-02-13 07:03:11 +05:30
|
|
|
|
|
|
|
int wget_main(int argc, char **argv) MAIN_EXTERNALLY_VISIBLE;
|
|
|
|
int wget_main(int argc UNUSED_PARAM, char **argv)
|
|
|
|
{
|
|
|
|
#if ENABLE_FEATURE_WGET_LONG_OPTIONS
|
|
|
|
static const char wget_longopts[] ALIGN1 =
|
|
|
|
/* name, has_arg, val */
|
|
|
|
"continue\0" No_argument "c"
|
|
|
|
"quiet\0" No_argument "q"
|
2017-01-25 02:11:43 +05:30
|
|
|
"server-response\0" No_argument "S"
|
2011-02-13 07:03:11 +05:30
|
|
|
"output-document\0" Required_argument "O"
|
|
|
|
"directory-prefix\0" Required_argument "P"
|
|
|
|
"proxy\0" Required_argument "Y"
|
|
|
|
"user-agent\0" Required_argument "U"
|
2015-10-21 01:21:52 +05:30
|
|
|
IF_FEATURE_WGET_TIMEOUT(
|
|
|
|
"timeout\0" Required_argument "T")
|
2011-02-13 07:03:11 +05:30
|
|
|
/* Ignored: */
|
2015-10-21 01:21:52 +05:30
|
|
|
IF_DESKTOP( "tries\0" Required_argument "t")
|
|
|
|
"header\0" Required_argument "\xff"
|
|
|
|
"post-data\0" Required_argument "\xfe"
|
2016-10-04 07:53:09 +05:30
|
|
|
"spider\0" No_argument "\xfd"
|
2018-05-28 18:06:26 +05:30
|
|
|
"no-check-certificate\0" No_argument "\xfc"
|
2011-02-13 07:03:11 +05:30
|
|
|
/* Ignored (we always use PASV): */
|
2015-10-21 01:21:52 +05:30
|
|
|
IF_DESKTOP( "passive-ftp\0" No_argument "\xf0")
|
2012-06-13 20:59:41 +05:30
|
|
|
/* Ignored (we don't support caching) */
|
2015-10-21 01:21:52 +05:30
|
|
|
IF_DESKTOP( "no-cache\0" No_argument "\xf0")
|
|
|
|
IF_DESKTOP( "no-verbose\0" No_argument "\xf0")
|
|
|
|
IF_DESKTOP( "no-clobber\0" No_argument "\xf0")
|
|
|
|
IF_DESKTOP( "no-host-directories\0" No_argument "\xf0")
|
|
|
|
IF_DESKTOP( "no-parent\0" No_argument "\xf0")
|
2011-02-13 07:03:11 +05:30
|
|
|
;
|
getopt32: remove applet_long_options
FEATURE_GETOPT_LONG made dependent on LONG_OPTS.
The folloving options are removed, now LONG_OPTS enables long options
for affected applets:
FEATURE_ENV_LONG_OPTIONS FEATURE_EXPAND_LONG_OPTIONS
FEATURE_UNEXPAND_LONG_OPTIONS FEATURE_MKDIR_LONG_OPTIONS
FEATURE_MV_LONG_OPTIONS FEATURE_RMDIR_LONG_OPTIONS
FEATURE_ADDGROUP_LONG_OPTIONS FEATURE_ADDUSER_LONG_OPTIONS
FEATURE_HWCLOCK_LONG_OPTIONS FEATURE_NSENTER_LONG_OPTS
FEATURE_CHCON_LONG_OPTIONS FEATURE_RUNCON_LONG_OPTIONS
They either had a small number of long options, or their long options are
essential.
Example: upstream addgroup and adduser have ONLY longopts,
we should probably go further and get rid
of non-standard short options.
To this end, make addgroup and adduser "select LONG_OPTS".
We had this breakage caused by us even in our own package!
#if ENABLE_LONG_OPTS || !ENABLE_ADDGROUP
/* We try to use --gid, not -g, because "standard" addgroup
* has no short option -g, it has only long --gid.
*/
argv[1] = (char*)"--gid";
#else
/* Breaks if system in fact does NOT use busybox addgroup */
argv[1] = (char*)"-g";
#endif
xargs: its lone longopt no longer depends on DESKTOP, only on LONG_OPTS.
hwclock TODO: get rid of incompatible -t, -l aliases to --systz, --localtime
Shorten help texts by omitting long option when short opt alternative exists.
Reduction of size comes from the fact that store of an immediate
(an address of longopts) to a fixed address (global variable)
is a longer insn than pushing that immediate or passing it in a register.
This effect is CPU-agnostic.
function old new delta
getopt32 1350 22 -1328
vgetopt32 - 1318 +1318
getopt32long - 24 +24
tftpd_main 562 567 +5
scan_recursive 376 380 +4
collect_cpu 545 546 +1
date_main 1096 1095 -1
hostname_main 262 259 -3
uname_main 259 255 -4
setpriv_main 362 358 -4
rmdir_main 191 187 -4
mv_main 562 558 -4
ipcalc_main 548 544 -4
ifenslave_main 641 637 -4
gzip_main 192 188 -4
gunzip_main 77 73 -4
fsfreeze_main 81 77 -4
flock_main 318 314 -4
deluser_main 337 333 -4
cp_main 374 370 -4
chown_main 175 171 -4
applet_long_options 4 - -4
xargs_main 894 889 -5
wget_main 2540 2535 -5
udhcpc_main 2767 2762 -5
touch_main 436 431 -5
tar_main 1014 1009 -5
start_stop_daemon_main 1033 1028 -5
sed_main 682 677 -5
script_main 1082 1077 -5
run_parts_main 330 325 -5
rtcwake_main 459 454 -5
od_main 2169 2164 -5
nl_main 201 196 -5
modprobe_main 773 768 -5
mkdir_main 160 155 -5
ls_main 568 563 -5
install_main 773 768 -5
hwclock_main 411 406 -5
getopt_main 622 617 -5
fstrim_main 256 251 -5
env_main 198 193 -5
dumpleases_main 635 630 -5
dpkg_main 3991 3986 -5
diff_main 1355 1350 -5
cryptpw_main 233 228 -5
cpio_main 593 588 -5
conspy_main 1135 1130 -5
chpasswd_main 313 308 -5
adduser_main 887 882 -5
addgroup_main 416 411 -5
ftpgetput_main 351 345 -6
get_terminal_width_height 242 234 -8
expand_main 690 680 -10
static.expand_longopts 18 - -18
static.unexpand_longopts 27 - -27
mkdir_longopts 28 - -28
env_longopts 30 - -30
static.ifenslave_longopts 34 - -34
mv_longopts 46 - -46
static.rmdir_longopts 48 - -48
packed_usage 31739 31687 -52
------------------------------------------------------------------------------
(add/remove: 2/8 grow/shrink: 3/49 up/down: 1352/-1840) Total: -488 bytes
text data bss dec hex filename
915681 485 6880 923046 e15a6 busybox_old
915428 485 6876 922789 e14a5 busybox_unstripped
Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>
2017-08-08 20:08:18 +05:30
|
|
|
# define GETOPT32 getopt32long
|
|
|
|
# define LONGOPTS ,wget_longopts
|
|
|
|
#else
|
|
|
|
# define GETOPT32 getopt32
|
|
|
|
# define LONGOPTS
|
2011-02-13 07:03:11 +05:30
|
|
|
#endif
|
|
|
|
|
|
|
|
#if ENABLE_FEATURE_WGET_LONG_OPTIONS
|
|
|
|
llist_t *headers_llist = NULL;
|
|
|
|
#endif
|
|
|
|
|
|
|
|
INIT_G();
|
|
|
|
|
2013-10-13 01:17:07 +05:30
|
|
|
#if ENABLE_FEATURE_WGET_TIMEOUT
|
|
|
|
G.timeout_seconds = 900;
|
|
|
|
signal(SIGALRM, alarm_handler);
|
|
|
|
#endif
|
2011-02-13 07:03:11 +05:30
|
|
|
G.proxy_flag = "on"; /* use proxies if env vars are set */
|
|
|
|
G.user_agent = "Wget"; /* "User-Agent" header field */
|
|
|
|
|
|
|
|
#if ENABLE_FEATURE_WGET_LONG_OPTIONS
|
|
|
|
#endif
|
2017-08-09 01:25:02 +05:30
|
|
|
GETOPT32(argv, "^"
|
|
|
|
"cqSO:P:Y:U:T:+"
|
2015-10-21 01:21:52 +05:30
|
|
|
/*ignored:*/ "t:"
|
|
|
|
/*ignored:*/ "n::"
|
|
|
|
/* wget has exactly four -n<letter> opts, all of which we can ignore:
|
|
|
|
* -nv --no-verbose: be moderately quiet (-q is full quiet)
|
|
|
|
* -nc --no-clobber: abort if exists, neither download to FILE.n nor overwrite FILE
|
|
|
|
* -nH --no-host-directories: wget -r http://host/ won't create host/
|
|
|
|
* -np --no-parent
|
|
|
|
* "n::" above says that we accept -n[ARG].
|
|
|
|
* Specifying "n:" would be a bug: "-n ARG" would eat ARG!
|
|
|
|
*/
|
2017-08-09 01:25:02 +05:30
|
|
|
"\0"
|
|
|
|
"-1" /* at least one URL */
|
|
|
|
IF_FEATURE_WGET_LONG_OPTIONS(":\xff::") /* --header is a list */
|
getopt32: remove applet_long_options
FEATURE_GETOPT_LONG made dependent on LONG_OPTS.
The folloving options are removed, now LONG_OPTS enables long options
for affected applets:
FEATURE_ENV_LONG_OPTIONS FEATURE_EXPAND_LONG_OPTIONS
FEATURE_UNEXPAND_LONG_OPTIONS FEATURE_MKDIR_LONG_OPTIONS
FEATURE_MV_LONG_OPTIONS FEATURE_RMDIR_LONG_OPTIONS
FEATURE_ADDGROUP_LONG_OPTIONS FEATURE_ADDUSER_LONG_OPTIONS
FEATURE_HWCLOCK_LONG_OPTIONS FEATURE_NSENTER_LONG_OPTS
FEATURE_CHCON_LONG_OPTIONS FEATURE_RUNCON_LONG_OPTIONS
They either had a small number of long options, or their long options are
essential.
Example: upstream addgroup and adduser have ONLY longopts,
we should probably go further and get rid
of non-standard short options.
To this end, make addgroup and adduser "select LONG_OPTS".
We had this breakage caused by us even in our own package!
#if ENABLE_LONG_OPTS || !ENABLE_ADDGROUP
/* We try to use --gid, not -g, because "standard" addgroup
* has no short option -g, it has only long --gid.
*/
argv[1] = (char*)"--gid";
#else
/* Breaks if system in fact does NOT use busybox addgroup */
argv[1] = (char*)"-g";
#endif
xargs: its lone longopt no longer depends on DESKTOP, only on LONG_OPTS.
hwclock TODO: get rid of incompatible -t, -l aliases to --systz, --localtime
Shorten help texts by omitting long option when short opt alternative exists.
Reduction of size comes from the fact that store of an immediate
(an address of longopts) to a fixed address (global variable)
is a longer insn than pushing that immediate or passing it in a register.
This effect is CPU-agnostic.
function old new delta
getopt32 1350 22 -1328
vgetopt32 - 1318 +1318
getopt32long - 24 +24
tftpd_main 562 567 +5
scan_recursive 376 380 +4
collect_cpu 545 546 +1
date_main 1096 1095 -1
hostname_main 262 259 -3
uname_main 259 255 -4
setpriv_main 362 358 -4
rmdir_main 191 187 -4
mv_main 562 558 -4
ipcalc_main 548 544 -4
ifenslave_main 641 637 -4
gzip_main 192 188 -4
gunzip_main 77 73 -4
fsfreeze_main 81 77 -4
flock_main 318 314 -4
deluser_main 337 333 -4
cp_main 374 370 -4
chown_main 175 171 -4
applet_long_options 4 - -4
xargs_main 894 889 -5
wget_main 2540 2535 -5
udhcpc_main 2767 2762 -5
touch_main 436 431 -5
tar_main 1014 1009 -5
start_stop_daemon_main 1033 1028 -5
sed_main 682 677 -5
script_main 1082 1077 -5
run_parts_main 330 325 -5
rtcwake_main 459 454 -5
od_main 2169 2164 -5
nl_main 201 196 -5
modprobe_main 773 768 -5
mkdir_main 160 155 -5
ls_main 568 563 -5
install_main 773 768 -5
hwclock_main 411 406 -5
getopt_main 622 617 -5
fstrim_main 256 251 -5
env_main 198 193 -5
dumpleases_main 635 630 -5
dpkg_main 3991 3986 -5
diff_main 1355 1350 -5
cryptpw_main 233 228 -5
cpio_main 593 588 -5
conspy_main 1135 1130 -5
chpasswd_main 313 308 -5
adduser_main 887 882 -5
addgroup_main 416 411 -5
ftpgetput_main 351 345 -6
get_terminal_width_height 242 234 -8
expand_main 690 680 -10
static.expand_longopts 18 - -18
static.unexpand_longopts 27 - -27
mkdir_longopts 28 - -28
env_longopts 30 - -30
static.ifenslave_longopts 34 - -34
mv_longopts 46 - -46
static.rmdir_longopts 48 - -48
packed_usage 31739 31687 -52
------------------------------------------------------------------------------
(add/remove: 2/8 grow/shrink: 3/49 up/down: 1352/-1840) Total: -488 bytes
text data bss dec hex filename
915681 485 6880 923046 e15a6 busybox_old
915428 485 6876 922789 e14a5 busybox_unstripped
Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>
2017-08-08 20:08:18 +05:30
|
|
|
LONGOPTS
|
2015-10-21 01:21:52 +05:30
|
|
|
, &G.fname_out, &G.dir_prefix,
|
2011-02-13 07:03:11 +05:30
|
|
|
&G.proxy_flag, &G.user_agent,
|
|
|
|
IF_FEATURE_WGET_TIMEOUT(&G.timeout_seconds) IF_NOT_FEATURE_WGET_TIMEOUT(NULL),
|
2015-10-21 01:21:52 +05:30
|
|
|
NULL, /* -t RETRIES */
|
|
|
|
NULL /* -n[ARG] */
|
2011-02-13 07:03:11 +05:30
|
|
|
IF_FEATURE_WGET_LONG_OPTIONS(, &headers_llist)
|
|
|
|
IF_FEATURE_WGET_LONG_OPTIONS(, &G.post_data)
|
|
|
|
);
|
2016-10-04 07:53:09 +05:30
|
|
|
#if 0 /* option bits debug */
|
|
|
|
if (option_mask32 & WGET_OPT_RETRIES) bb_error_msg("-t NUM");
|
|
|
|
if (option_mask32 & WGET_OPT_nsomething) bb_error_msg("-nsomething");
|
|
|
|
if (option_mask32 & WGET_OPT_HEADER) bb_error_msg("--header");
|
|
|
|
if (option_mask32 & WGET_OPT_POST_DATA) bb_error_msg("--post-data");
|
|
|
|
if (option_mask32 & WGET_OPT_SPIDER) bb_error_msg("--spider");
|
2018-05-28 18:06:26 +05:30
|
|
|
if (option_mask32 & WGET_OPT_NO_CHECK_CERT) bb_error_msg("--no-check-certificate");
|
2016-10-04 07:53:09 +05:30
|
|
|
exit(0);
|
|
|
|
#endif
|
2011-02-13 07:03:11 +05:30
|
|
|
argv += optind;
|
|
|
|
|
|
|
|
#if ENABLE_FEATURE_WGET_LONG_OPTIONS
|
|
|
|
if (headers_llist) {
|
2015-02-19 01:11:02 +05:30
|
|
|
int size = 0;
|
|
|
|
char *hdr;
|
2011-02-13 07:03:11 +05:30
|
|
|
llist_t *ll = headers_llist;
|
|
|
|
while (ll) {
|
|
|
|
size += strlen(ll->data) + 2;
|
|
|
|
ll = ll->link;
|
|
|
|
}
|
2015-02-19 01:11:02 +05:30
|
|
|
G.extra_headers = hdr = xmalloc(size + 1);
|
2011-02-13 07:03:11 +05:30
|
|
|
while (headers_llist) {
|
2015-02-19 01:11:02 +05:30
|
|
|
int bit;
|
|
|
|
const char *words;
|
|
|
|
|
|
|
|
size = sprintf(hdr, "%s\r\n",
|
|
|
|
(char*)llist_pop(&headers_llist));
|
|
|
|
/* a bit like index_in_substrings but don't match full key */
|
|
|
|
bit = 1;
|
|
|
|
words = wget_user_headers;
|
|
|
|
while (*words) {
|
|
|
|
if (strstr(hdr, words) == hdr) {
|
|
|
|
G.user_headers |= bit;
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
bit <<= 1;
|
|
|
|
words += strlen(words) + 1;
|
|
|
|
}
|
|
|
|
hdr += size;
|
2011-02-13 07:03:11 +05:30
|
|
|
}
|
|
|
|
}
|
|
|
|
#endif
|
|
|
|
|
2011-02-15 05:28:36 +05:30
|
|
|
G.output_fd = -1;
|
|
|
|
G.o_flags = O_WRONLY | O_CREAT | O_TRUNC | O_EXCL;
|
|
|
|
if (G.fname_out) { /* -O FILE ? */
|
|
|
|
if (LONE_DASH(G.fname_out)) { /* -O - ? */
|
|
|
|
G.output_fd = 1;
|
|
|
|
option_mask32 &= ~WGET_OPT_CONTINUE;
|
|
|
|
}
|
|
|
|
/* compat with wget: -O FILE can overwrite */
|
|
|
|
G.o_flags = O_WRONLY | O_CREAT | O_TRUNC;
|
|
|
|
}
|
|
|
|
|
2011-02-13 07:03:11 +05:30
|
|
|
while (*argv)
|
2011-02-17 00:39:36 +05:30
|
|
|
download_one_url(*argv++);
|
2011-02-13 07:03:11 +05:30
|
|
|
|
2011-02-15 15:33:53 +05:30
|
|
|
if (G.output_fd >= 0)
|
|
|
|
xclose(G.output_fd);
|
|
|
|
|
2013-10-16 18:13:30 +05:30
|
|
|
#if ENABLE_FEATURE_CLEAN_UP && ENABLE_FEATURE_WGET_LONG_OPTIONS
|
|
|
|
free(G.extra_headers);
|
|
|
|
#endif
|
|
|
|
FINI_G();
|
|
|
|
|
2011-02-17 00:39:36 +05:30
|
|
|
return EXIT_SUCCESS;
|
2011-02-13 07:03:11 +05:30
|
|
|
}
|