Re: Ncat proposal: new environment variables

[Jacek Wielemborek <wielemborekj1 () gmail com>]

I had a meeting with David this Tuesday and was assigned to implement
the proposal from my previous post, without ncatin/ncatout FD numbers
yet (I will probably be adding them next week). So, here's a list of
environment variables that I introduced to the
--exec/--sh-exec/--lua-exec child processes (along with a quote from
the docs written by me):

 * NCAT_REMOTE_ADDR - the remote address - in connect mode, it's the
one we're connecting to, in listen mode - it's the one that connected
to us,
 * NCAT_REMOTE_PORT - the remote port number, where "remote" means the
same as above,
 * NCAT_LOCAL_ADDR - our local address used for
establishing/receiving the connection,
 * NCAT_LOCAL_PORT - our local port number,
 * NCAT_PROTO - the protocol name - TCP, UDP or SCTP,
 * NCAT_SSL - set to "1" if we're using SSL. Not set otherwise.

In addition to the code (which works under Windows as well) and its
tests and documentation, I also added some examples. This includes a
p0fme.py script, which tells the connecting user everything it could
gather on her, using TCP fingerprinting (requires an external tool -
p0f, see: http://lcamtuf.coredump.cx/p0f3/ ). There's also a simple
shell script demonstrating logging and a Lua "whatismyip" script.

Feel free to pull the nmap-exp/d33tah/ncat-env-conninfo branch, try it
out and - if you have some time - read the diff (against r31368).
Feedback is more than welcome! :)

I'm also really interested about the Ncat tricks you invented that
make use of these variables.

--
Jacek Wielemborek
_______________________________________________
Sent through the dev mailing list
http://nmap.org/mailman/listinfo/dev
Archived at http://seclists.org/nmap-dev/