Re: Script Arguments and IDs

[jah <jah () zadkiel plus com>]

On 02/06/2008 01:41, Kris Katterjohn wrote:
> So instead of using the script id, the subtable name checked for may
> need to
> be hardcoded in (presumably using the filename).  But even then it can be
> messed up (e.g. "SSLv2-support").
>
> I think "nmap.registry.args[id]" is cleaner than some hardcoded name.
>
> So how should this work?  Should we just specify what to name the
> subtable in
> the individual scripts rather than assuming the ID, or should IDs just be
> valid if args are used?
>
> Also: should IDs and filenames be closely related, or should it not
> matter?
I think that the ID that is displayed in the host script results should
closely match the filename to make it apparent which script was output. 
There'll come a time when there's too many scripts to guess which
displayed ID matches which script.
How about if ID was a table?
id= {"long"="SSLv2-support", "short" = "sslv2"}
id.long could be enforced programmatically, i.e. not specified in the
script; and this would be the filename and displayed in host script results.
id.short would be specified in the script and could be a less restricted
choice.
A script might then be able to hunt for arguments in either
nmap.registry.args.id.short or nmap.registry.args.id.long.
There could be some jiggery-pokery with id.long to replace hyphens and
underscores with spaces for presentation purposes.
Might also be able to "call" a script with either ID value...

just a thought, like.

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