Re: [NSE] Reflection facilities for better verbose/debug output

[Fyodor <fyodor () insecure org>]

On Mon, Sep 24, 2012 at 01:48:08AM -0400, Patrick Donnelly wrote:
> a shorter version might also be:
>
> NSE: [http-default-accounts W:0xdeadbeef] ...
>
> compared to
>
> NSE: [http-default-accounts M:0xdeadbee] ...
>
> where M and W are respectively Master and Worker, naturally.
>
> A version that would include the parent might look like:
>
> NSE: [http-default-accounts M:0xdeadbee:W:0xdeadbeef] ...
>
> Thoughts?
>
> > Even that is quite long; maybe the thread id should only be there for
> > -d2 or higher.
>
> That's one possibility.

Thanks Patrick, this looks like great functionality!  As far as
whether to print the thread ID, let's look at how scripts are
reference (in debug mode) now.  When scripts start or stop, we print
out lines like:

NSE: Starting ssh-hostkey against 74.207.244.221:22
NSE: Finished ssh-hostkey against 74.207.244.221:22.

So if we were to use the thread ID in debug messages, we'd need to
include it in the lines above too so users can follow what is going
on.  But given that a thread ID like 0x38472a17 isn't a whole lot
shorter than "74.207.244.221:22", and given that the latter is a lot
more meaningful for end users, maybe we should stick to printing the
target IP (and port number if it is a portrule).  So the debug lines
could look like:

NSE [http-headers, 74.207.244.221:80] HTTP: Host supports HEAD.
NSE [http-default-accounts, 64.211.144.177:80] 5 fingerprints were loaded

And verbose lines could be like:

NSE [firewalk] not running due to lack of privileges.

(If a verbose line needs the target IP and/or port, it can include it
in the text itself).

For higher debug levels (e.g. at least 2), also including the thread
ID with debug messages seems reasonable.

Cheers,
Fyodor
_______________________________________________
Sent through the nmap-dev mailing list
http://cgi.insecure.org/mailman/listinfo/nmap-dev
Archived at http://seclists.org/nmap-dev/