int getline(FILE *stream, char *buf, size_t buflen, const char **errormsg)
Read a line from the FILE stream into buf/buflen using fgets(), so up
to buflen-1 chars will be read and the result will be NUL terminated.
If the line has a trailing newline it will be removed.
If the line is too long, excess characters will be read until
newline/EOF/error.
Various -ve return values indicate different errors, and errormsg
will be changed to an error description if it's not NULL.
Convert to use getline() instead of fgets() whenever reading user input
to ensure that an overly long input line doesn't leave excess characters
for the next input operation to accidentally use as input.
Zero out the password & account after we've finished with it.
Consistently use getpass(3) (i.e, character echo suppressed) when
reading the account data. For some reason, historically the "login"
code suppressed echo for Account: yet the "user" command did not!
Display the hostname in the "getaddrinfo failed" warning.
Appease some -Wcast-qual warnings. Fixing all of these requires
significant code refactoring. (mmm, legacy code).
realpath(3) on non-NetBSD systems may fail if the target filename doesn't
exist, so instead use realpath(3) on the parent directory of `file'.
Per discussion with Todd Eigenschink.
connect(2) in xconnect() by temporarily setting O_NONBLOCK
on the socket and using xpoll() to wait for the operation
to succeed.
The timeout used is the '-q quittime' argument (defaults to
60s for accept(2), and the system default for connect(2)).
Idea inspired by discussion with Chuck Cranor.
This may (indirectly) fix various problems with timeouts
in active mode through broken firewalls.
Implement xpoll() as a wrapper around poll(2), to make it
easier to replace on systems without a functional poll(2).
Unconditionally use xpoll() instead of conditionally using
select(2) or poll(2).
making functions that should be static be static, and cleaning up
const usage. Added a guard against buffer overflow, but the domap function
is a bit too complicated for me to tackle right now. I will leave it
to the author; hi luke!
current directory.
The previous behaviour (of trusting the remote server's response when
retrieving the list of files to mget with prompting disabled) has been
in ftp ~forever, and has been a "known issue" for a long time.
Recently an advisory was published by D.J. Bernstein on behalf of
Yosef Klein warning of the problems with the previous behaviour, so
to alleviate concern I've fixed this with a sledgehammer.
Remember the local cwd after any operation which may change it.
Use "remotecwd" instead of "remotepwd".
* Only remember {WWW,Proxy}-Authenticate "Basic" challenges; no point
in tracking any others since ftp doesn't support them.
* Improve the parsing of HTTP responses.
and sigint_raised is non-zero, reset the handler for SIGINT to SIG_DFL
and raise(SIGINT) so that the appropriate wait(3) status is setup.
Based on solution proposed by Ognyan Kulev.
This should really fix PR [pkg/26351].
skip LWS (linear white space; CR, LF, space, tab) and the end of lines and
between the field name and the field value. This still isn't 100% compliant,
since we don't support "multi line" responses at this time.
This should fix PR [bin/22611] from TAMURA Kent (although I can't easily
find a http server to reproduce the problem against.)
Fix a minor memory leak when parsing HTTP response headers.
exit with 130 instead of 1 (or rarely, 0).
This allows an ftp auto-fetch in a shell loop to correctly terminate the loop.
Should fix PR [pkg/26351], and possibly others.
until it's writable or it fails with something other than EINTR.
This matches the behaviour in SUSv3, and prevents the problem when
pressing ^T (SIGINFO, which is marked as restartable) during connection
setup would cause ftp to fail with EADDRINUSE or EALREADY when the
second connect(2) was attempted on the same socket.
Problem found and solution provided by Maxime Henrion <mux@freebsd.org>.
enabled. Suggested by Todd Vierling.
Allow empty passwords in ftp://user:@host/file auto-fetch URLs,
per RFC 1738. Requested by Simon Poole.
Update version.
* Work around broken ftp servers (notably ProFTPd) that can't even follow
RFC 2389, and skip any amount of whitespace before a FEATure response.
The RFC says 'single space' yet ProFTPd puts two.
Noted by DervishD <raul@pleyades.net>.
* Improve formatting of features[] debug dump.
* Invalidate remote directory completion cache if any command which
may change the remote contents completes successfully, including:
del, mdel, ren, mkdir, rmdir, quote, and all upload commands
Patch from Yar Tikhiy <yar@freebsd.org>.
may change the remote contents completes successfully, including:
del, mdel, ren, mkdir, rmdir, quote, and all upload commands
Patch from Yar Tikhiy <yar@comp.chem.msu.su>.
RFC 2389, and skip any amount of whitespace before a FEATure response.
The RFC says 'single space' yet ProFTPd puts two.
* Improve formatting of features[] debug dump.
in external programs (conditionalized on -DSTANDALONE_PROGRESS).
The following moved from util.c to progressbar.c:
alarmtimer(), progressmeter(), psummary(), ptransfer(),
xsignal(), xsignal_restart()
The following moved from extern.h and ftp_var.h to progressbar.h:
STALLTIME, verbose, fromatty, progress, quit_time, ttywidth
use it to disable extended passive mode when your firewall is obnoxious.
This did in fact solve a problem I was having with a firewall at work, and
the status message was taken directly from a problem session fixed by this.
elapsed time. (suggested by simonb)
- actually display transfer stats after a URL fetch. (bug introduced a
*long* time ago)
- update copyright & version
Fixes [bin/15415] from Takahiro Kambe <taca@sky.yamashina.kyoto.jp>
- Fix bad mode passed by mls() to recvrequest().
Fixes [bin/16642] from <steve.mcclellan@radisys.com>
From Hajimu UMEMOTO, via Mike Heffner of FreeBSD.
(FreeBSD has imported NetBSD's ftp as their ftp client;
Mike is sending back some of their local changes).
- add -N netrc and $NETRC, as methods to select an alternative .netrc file
- cache local user name and home directory for further use
- in mget(), use docase() instead of a local version to do the case
conversion.
after the chunksize (before the \r\n), Apache 1.3.11 puts *multiple*
trailing spaces after the chunksize. I 'm fairly certain that this is
contrary to RFC 2068 section 3.6, but whatever...
Found by David Brownlee <abs@mono.org>
it to a numeric port number
* use parseport() in parse_url() and hookup()
* don't try and lookup the port number using getaddrinfo(), as it's too hard
to separate a failed host name lookup from a failed service name lookup.
this was causing lossage on systems that don't have `http' in services(5)
(such as solaris), but only crept in when we started using getaddrinfo()
unconditionally.
equivalents. name change suggested by Klaus Klein <kjk@netbsd.org>
- change defined(BSD4_4) || HAVE_SIN_LEN tests into HAVE_SOCKADDR_SA_LEN,
and set the latter if BSD4_4 exists
paths. (lukemftp will provide replacements for these on older systems)
* rename __USE_SELECT to USE_SELECT
* rename BSD4_4 to HAVE_SIN_LEN
* replace union sockunion {} with struct sockinet {}, and modify the code
accordingly. this is possibly more portable, as it doesn't rely upon the
structure alignment within the union for our own stuff.
(XXX: haven't tested the ipv6 stuff)
#define NOQUAD ! NOQUAD
------- ------ - ------
QUADF "%ld" "%lld"
QUADFP(x) "%" x "ld" "%" x "lld"
QUADT long long long
STRTOL(x,y,z) strtol(x,y,z) strtoll(x,y,z)
(from RFC 2389).
add support for MLST & MLSD (machine parseble listings) with 'mlst', 'mlsd'
and 'pmlsd' (mlsd |$PAGER) commands. (from draft-ietf-ftpext-mlst-11)
rename remotesyst() to getremoteinfo(), and modify to parse the result from
FEAT (if supported), and take into account the support for the various
extensions such as MDTM, SIZE, REST (STREAM), MLSD, and FEAT/OPTS.
put each feature into one of the following categories:
- known to work (explicit FEAT)
- unknown but assume works until explicit failure, when it's
then tagged as `known not to work'.
- known not to work (FEAT succeeded but didn't return anything,
or was unknown and then explicit failure)
assign results into features[] matrix.
add support to getreply() so that an optional callback will be called
for each line received from the server except for the first and last.
this is used in FEAT (and MLST) parsing.
modify various commands to check if REST (STREAM), MDTM and SIZE are
explicitly or implicitly supported before using.
fix `syst' when verbose is off.
minor knf (indent goto labels by one space, etc).
simply various command usage handlers by assuming that argv != NULL except
for quit() and disconnect().
call remotesyst() only when login has been successful
some servers don't let you run SYST until you've successfully logged in.
* in fetch_ftp(), always call setpeer() with autologin disabled, and use
the following ftp_login() to DTRT.
this prevents ftp from trying to login a second time if the first autologin
fails when connecting to a remote site anonymously using autofetch.
* reset unix_proxy and unix_server in cleanuppeer()
* missed a function conversion in the KNF sweep...
* initconn(): use in_addr_t instead of u_int32_t when manipulating IPv6
addresses (and assume anything with ipv6 has in_addr_t; if not, i'll
add an autoconf test for it)
* ai_unmapped(): not all systems have sin_len; so only set #ifdef BSD4_4
* fix some lint
- ftp(1): treats IPv4 mapped destination as IPv4 peer, not native IPv6 peer.
this does not support network with SIIT translator.
- rshd(8)/rlogind(8): rejects accesses from IPv4 mapped peer, to avoid
possible abuse of IPv4 mapped addr (rshd/rlogind use source address-based
auth so it is important to check the condition).
Now that after many years on not caring we find certain popular
ftp servers are starting to obey RFC959 to the letter of the law
and will only return a list of filenames (not directories or
other filetypes) in the output of `NLST', then `LIST' is more useful
in this case. (Note that the aforementioned pedanticness means that
filename completion isn't as useful as it could be...)
Fixes [bin/8937] by David A. Gatwood <dgatwood@deepspace.mklinux.org>
* sanity check a length (otherwise certain bogus responses can crash ftp)
* allow a transfer encoding type of `binary'; certain firewall vendors
return this bogus type...
send Host: directive with RFC2732 bracket notation for IPv6 numeric,
otherwise "host:port" is ambiguous to servers (clarification will be submitted
as update to RFC2732).
typedef void (*sigfunc) __P((int));
and replace use of sig_t and void (*)(int).
certain other OSes define sig_t differently to that (they add extra arguments),
and it causes problems due to function mismatches, etc...
the output of `MDTM'.
obviously the programmer of aforementioned lame ftpd's did something like
"19%02d", tm->tm_year
instead of
"%04d", tm->tm_year + TM_YEAR_BASE
fixes [bin/9289] by jbernard@mines.edu
the word is already complete then return CC_REFRESH so that the higher layer
may append a suffix if necessary. Fix from Launey Thomas <ljt@alum.mit.edu>.
Problem noticed/debugging assisted by giles lean <giles@nemeton.com.au>.
XXX: rate limiting with chunked xfers might not limit correctly (i.e,
the limit may be too high or too low); fixing this is non trivial,
and will probably occur if i ever rototill fetch_url()
* fix bug where the second press of <TAB> on an empty word (i.e, list
all options) may have resulted in an strncmp() against NULL. (detected
by _DIAGASSERT())
for listings of the current working directory; some ftp servers don't
like `NLST .'.
[noted by Giles Lean <giles@nemeton.com.au>]
* recvrequest(): treat remote=="" as remote==NULL when calling command().
(to support the above change)
* support `[user@]' in `[user@]host' and `[user@]host[:][path]'.
[based on idea (and initial code) from David Maxwell <david@fundy.ca>]
* `idle' may be invoked without any args
* reformat some comments
* reformat usage string in program and man page
* call updateremotepwd() after successful login, not after successful connect
* always call setsockopt(, IPPROTO_IP, IP_TOS, ) (et al); using #if
defined(IPPROTO_IP) doesn't work on certain foreign systems where
enums instead of #defines are used...
[noted by Matthias Pfaller <leo@dachau.marco.de>]
EL_RPROMPT support i added to editline(3).
- allow $FTPPROMPT and $FTPRPROMPT to override defaults for the relevant
prompts
- move `%' formatting code from prompt() to expandbuf().
- implement `%.' and `%c', similar to the same % codes in tcsh(1)
(functionality I added to tcsh nearly 6 years ago), except that `%.'
always does `...trailing' and `%c' always does `/<x>trailing'.
- unknown `%foo' codes get printed as `%foo'
to contain the remote working directory.
- add `set prompt', a user configurable prompt. (defaults to `ftp> ').
the following escape characters a la tcsh(1) are supported: %/, %m,
%M, and %n.
- add global var `username'; used by prompt code
- fix a couple of minor memory leaks
- bump version
the provided port is a valid number use that rather than trying to do
getservbyname() against it.
fixes a problem on foreign systems noted by Chuck Silvers <chuq@chuq.com>
- add `usage'; displays the usage of a command.
implemented by calling the c_handler() with argc = 0, argv = "funcname".
- add `passive auto'; does the same as $FTPMODE=auto.
- add `set [option value]'; display all options, or set an option to a value.
- add `unset option'; unset an option.
- add getoptionvalue() to retrieve an option's value, and replace a few
global variables with calls to this.
- implement cleanuppeer(), which resets various bits of state back to
`disconnected'. call in disconnect() and lostpeer().
- support completing on `options'.
- improve recovery after a SIGINT may have closed the connection.
XXX: there's still a couple to fix
other stuff:
- various consistency fixes in the man page.
- ensure that the command usage strings in the code and man page match reality.
- mput/mget: check that the connection still exists before each xfer.
- minor cosmetic changes in confirm().
- set code correctly in sizecmd() and modtime()
- don't need \n in err() strings.
- change lostpeer to take an argument (rather than casting (sig_t)lostpeer
in signal handlers)
- knf and whitespace police.
* simplify main loop (don't need `top' variable any more)
* use a struct sockaddr_in6.sin6_addr for the result from inet_pton(),
rather than u_char buf[16]
* add a few more comments
* confirm(): move from util.c to cmds.c. display mnemonic string in its prompt.
add support for `q' (terminate current xfer), `?' (show help list)
* in various signal handlers, output a linefeed only if fromatty.
* if fgets(stdin) returned NULL (i.e, EOF), clearerr(stdin) because you don't
want future fgets to fail. this is not done for the fgets() in the main
command loop, since ftp will quit at that point.
* unless ftp is invoked with -a, don't retain the anonftp setting between
hosts (`ftp somehost:' sets anonftp, but you don't want that to `stick'
if you close that connection and open a new one).
stage. if this occurs, just call lostpeer() to close the connection. whilst this
might be considered brutal, it's also extremely handy if you're impatient or there's
lossage at the remote end.
don't save the signal mask on some foreign systems.
* ensure signal handlers don't use stdio and do reset errno if they
don't exit with siglongjmp()
* use a common SIGINT handler for {send,recv}request()
(work around editline's override)
* minor cleanup of signal handler (along the lines of similar work in
recvrequest()). the handlers should now be reset everytime the cleanup
handler was callled.
prints out info if bytes > 0). only set the handler if SIGINFO is defined
* hijack SIGQUIT to be the same as SIGINFO (foreign ports have this, and it's
annoying to have SIGQUIT dump core on netbsd when it prints info on other
systems)
* in {recv,send}request(), factor a lot of duplicated code out into a
`cleanup' section at the end
* rework shell() a bit
* implement xsignal_restart(), which only sets the SA_RESTART flag if
specifically requested
* xsignal() is now a wrapper to xsignal_restart(). INFO, USR1, USR2 and WINCH
are restartable, ALRM, INT, PIPE and QUIT are not.
* improve getreply()'s timeout code to take advantage of the above.
other changes:
* improve wording of how globbing works for `classic' URLs (host:path).
suggested by John Refling <johnr@imageworks.com> in relation to PRs
[bin/8519] and [bin/8520]
* always compile in the `edit' command even if NO_EDITCOMPLETE defined.
it's just a no-op in the latter case, which is more consistent to
the users.
* always compile in about: support (i.e, remove NO_ABOUT).
i'm entitled to some vanity in this program...
* clean up some whitespace
the disabled state can be overridden by toggling epsv4.
(I got sick of the errors about EPSV not being supported on almost
every server I connect to. This way we retain support for epsv4, but
it's not so whiny after the first failure...)
to "extern" if it's not set. define GLOBAL to (empty) in main.c.
this effectively moves all the globals into main.c whilst retaining
namespace access to them in other source files.
(global vars in header files confuse foreign linkers)
should be in host order. found/fixed by Matthias Pfaller <leo@dachau.marco.de>
* parse_url(): improve checking of portnum, and add an extra argument to pass
back the parsed portnum to the caller (reduces a bit of code duplication)
* Move the KAME/WIDE copyrights after the BSD/TNFi ones. Since there was
significantly less code added under the former, it's only fair on the latter.
ETA might be over and will appear to drop rapidly, rather than being under
and appear not to change. the original code makes sense when you're testing
by suspending & resuming the client. however, the unfudged number is probably
better in reality, especially for slow spurty networks.
requested by ITOH Yasufumi <itohy@netbsd.org> in [bin/7977]
lpage page local files
pdir as dir, but through your $PAGER
pls as ls, but through your $PAGER
* implement docase() (a la dotrans() et al) and use appropriately, rototilling
some duplicated code
* globulize(): modify to return a pointer to the strdup()ed result in all cases,
and hack the code that calls it to take this into account
* replace strcpy() and strncpy() with strlcpy()
* put(), getit(): use some aptly named local vars instead of argv[...]
* delint
prototype for idle() in <unistd.h> (which i thought was against namespace
and sensibility guidelines, but...)
* consistently use xsignal() instead of signal(). we get known behaviour
in all cases (SA_RESTART), which is good for some borken foreign systems.
* remove signal.h from most files; it's unnecessary now
* fetch_url(): use `long chunksize' instead of ssize_t; it's more portable, and
we're setting chunksize with strtol() anyway
* xsignal(): only use SA_RESTART if it exists. SunOS 4.x doesn't have it
but has the inverse (SA_INTERRUPT). the original function i was inspired
from had this support (lib/signal.c, W. Richard Stevens' `UNP 2nd ed Vol 1').
* remove <termios.h> from util.c; it should be unnecessary now
prototype for idle() in <unistd.h> (which i thought was against namespace
and sensibility guidelines, but...)
* consistently use xsignal() instead of signal(). we get known behaviour
in all cases (SA_RESTART), which is good for some borken foreign systems.
* remove signal.h from most files; it's unnecessary now
* fetch_url(): use `long chunksize' instead of ssize_t; it's more portable, and
we're setting chunksize with strtol() anyway
* xsignal(): only use SA_RESTART if it exists. SunOS 4.x doesn't have it
but has the inverse (SA_INTERRUPT). the original function i was inspired
from had this support (lib/signal.c, W. Richard Stevens' `UNP 2nd ed Vol 1').
* remove <termios.h> from util.c; it should be unnecessary now
it's more portable and more obvious
* remove the mkgmtime() && HAVE_TIMEGM stuff:
a) why should netbsd have to define HAVE_TIMEGM to compile cleanly?
b) foreign compiles of ftp should just be linked with working
timegm function
a more portable version of this ftp client will be released as a 3rdparty
product; no use polluting our code with half-baked attempts...
* abort_remote(): replace borken MIN(4,BUFSIZ) with just BUFSIZ; it
should have been MAX(4,BUFSIZ), but it's probably safe to assume that
BUFSIZ is at least 3... (fix from simonb)
* auth_url(): use the correct variable when calculating a buffer size.
add lots of comments about how to size up the buffers, and add extra
checks to hopefully ensure that there won't be an overflow (unless
someone modifies the length of the sprintf()s).
* as part of the above, slightly rework the way the `*' bar is calculated.
also fixes a display bug when > 160 stars were needed to be printed.
the maximum progress bar width at this time is 256.
* remove some code that checks the port that was #if 0-ed out as part of the
ipv6 migration; it's not going to be used again.
* document the above three commands
* rototill the way the sndbuf and rcvbuf work. remove resetsockbufsize()
* use the appropriate socket buffer size as the size of the buffer that
the read()/write() loops use. speeds up things in some cases.
wiz