Re: [NSE] Better Handling for Require Errors

[David Fifield <david () bamsoftware com>]

On Wed, May 04, 2011 at 05:07:57PM -0400, Patrick Donnelly wrote:
> On Mon, Mar 28, 2011 at 1:02 PM, Patrick Donnelly <batrick () batbytes com> wrote:
> > So I wrote up a quick patch to make require errors easier for scripts.
> > Basically, you can now require libraries that may not exist without
> > having to use pcall. See also [1].
>
> This is committed as of revision 23085. The "root" module has been
> removed in favor of keeping the check for root access in the rule
> function. The require function has been modified so that NSE can
> silently catch require errors. If you want/need to use the real
> require function, it is available as stdnse.require.

This can cause some confusion during --script-updatedb. There was an
error in ncp.lua where "ipOps" was written "ipops" and failed to import
on a case-sensistive filesystem. The quiet require errors led to a
confusing error message:

$ ./nmap --script-updatedb

Starting Nmap 5.51SVN ( http://nmap.org ) at 2011-06-07 23:08 PDT
NSE: Updating rule database.
NSE: failed to initialize the script engine:
/home/david/nmap-git/nse_main.lua:1017: attempt to index local 'script' (a nil value)
stack traceback:
        /home/david/nmap-git/nse_main.lua:1017: in main chunk
        [C]: ?

QUITTING!

Running with -d shows a stack trace and makes it more obvious what the
problem is ("module 'ipops' not found"), but can we do anything about
this?

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