You should be able to obtain passwords or usernames with this
function.
Signed-off-by: Andreas Schneider <mail@cynapses.org>
git-svn-id: svn+ssh://svn.berlios.de/svnroot/repos/libssh/trunk@198 7dcaeef0-15fb-0310-b436-a5af3365683c
0002-Use-const-whereever-it-makes-sense.patch,
0003-Implement-function-to-retrieve-userauth-possabilitie.patch
from Andreas Schneider
git-svn-id: svn+ssh://svn.berlios.de/svnroot/repos/libssh/trunk@191 7dcaeef0-15fb-0310-b436-a5af3365683c
refactoring of the socket class. Now the buffering happens in the socket class.
enhanced the logging system. Cleaned up some debugging messages.
Verified the working with ssh-1.
If this cleanup introduced bugs (it did but corrected the found ones) at least, they will be easier to find
also added the (c) and fixed dates for updated files
git-svn-id: svn+ssh://svn.berlios.de/svnroot/repos/libssh/trunk@169 7dcaeef0-15fb-0310-b436-a5af3365683c
Some documentation, and a new logging system.
some work must be done to get rid of the infamous ssh_say()
git-svn-id: svn+ssh://svn.berlios.de/svnroot/repos/libssh/trunk@166 7dcaeef0-15fb-0310-b436-a5af3365683c
include/libssh/priv.h: Add struct definitions that previously were in include/libssh/server.h
libssh/server.c: Include libssh/priv.h
git-svn-id: svn+ssh://svn.berlios.de/svnroot/repos/libssh/trunk@109 7dcaeef0-15fb-0310-b436-a5af3365683c
Ipv6 patch from Laurent Bigonville. When the ipv6 targets are
unreachable, it tries all possible targets from the getaddrinfo list.
Norbert: does it break anything ? it works for me but I didn't test it
extensively
git-svn-id: svn+ssh://svn.berlios.de/svnroot/repos/libssh/trunk@99 7dcaeef0-15fb-0310-b436-a5af3365683c
I am going to make the whole documentation as doxygen files.
git-svn-id: svn+ssh://svn.berlios.de/svnroot/repos/libssh/trunk@77 7dcaeef0-15fb-0310-b436-a5af3365683c
some packet nonblocking fixes.
reenable sftp from the sample client.
git-svn-id: svn+ssh://svn.berlios.de/svnroot/repos/libssh/trunk@52 7dcaeef0-15fb-0310-b436-a5af3365683c
and packet_read may return SSH_AGAIN if the session is nonblocking and
there is not enough data to be read.
I also added a socket buffering through session->in_socket_buffer. It is
more low-level than the packet buffer.
I should rename in_buffer with in_packet_buffer.
There is still work to do with the socket status, (opened, error,
closed, ...) and much more work to extend the nonblocking to session
opening, messages sending and such.
I find the switch(session->packet_state) solution very nice (especially
when the nonblocking function may have 10 differents states, like in a
connection.)
git-svn-id: svn+ssh://svn.berlios.de/svnroot/repos/libssh/trunk@49 7dcaeef0-15fb-0310-b436-a5af3365683c
I resolved a memory alloc problem into hmac_init (same kind that the one of md5_init).
It's still saying there is a memory corruption. Since the memory corruption happens before it is found (in malloc()), I'll have to run valgrind to locate it.
git-svn-id: svn+ssh://svn.berlios.de/svnroot/repos/libssh/trunk@46 7dcaeef0-15fb-0310-b436-a5af3365683c
It still doesn't compile because of the O1 thing passed to ld.
git-svn-id: svn+ssh://svn.berlios.de/svnroot/repos/libssh/trunk@41 7dcaeef0-15fb-0310-b436-a5af3365683c
I adapted the sample.c file. the select_loop function is bloated and fails to demonstrate how libssh is simple to handle...
it looks to run at first try.
git-svn-id: svn+ssh://svn.berlios.de/svnroot/repos/libssh/trunk@39 7dcaeef0-15fb-0310-b436-a5af3365683c
