Re: NSE exceptions: false or nil?

[Stoiko Ivanov <stoiko () xover htu tuwien ac at>]

Hi,

On Thu, Nov 01, 2007 at 10:40:22PM -0700, David Fifield wrote:
> Are NSE exceptions supposed to return false or nil? At
> http://insecure.org/nmap/nse/nse-api.html#nse-exceptions it says that
> the first thing returned is a Boolean value. But trying to catch an
> exception from connect fails, as the attached example script (test.nse)
> shows. It produces this message:
>
>       Initiating SCRIPT ENGINE at 22:26
>       false   Name or service not known
>       SCRIPT ENGINE: In: nse_nmaplib.cc:474 Trying to finalize a non conforming function. Are you sure you return 
> true on success followed by the remaining return values and nil on failure followed by an error string?
>       QUITTING!
>
> We see that connect returns (false, "Name or service not known"), but
> this is "non conforming." Should NSE's internal functions be modified to
> return nil instead of false, or should false be considered an exception
> marker? Does the patch I attached do the right thing?
I think the latter would be more convenient. 
Since the network I/O functions return true on success I would expect a false 
value to indicate an exception. 
AFAIK this patch shouldn't cause any problems with the remaining NSE.

> David Fifield
>
> ..snip..
> Index: nse_nmaplib.cc
> ===================================================================
> --- nse_nmaplib.cc    (revision 6111)
> +++ nse_nmaplib.cc    (working copy)
> @@ -460,7 +460,7 @@
>  }
>  
>  static int l_exc_finalize(lua_State *l) {
> -     if (lua_isnil(l, 1)) {
> +     if (lua_isnil(l, 1) || !lua_toboolean(l, 1)) {
This looks good to me, however if I recall correctly the lua_isnil() is not
really needed, because lua_toboolean() should return false if there is nil
at the specified position on the stack.

>               lua_pushvalue(l, lua_upvalueindex(1));
>               lua_call(l, 0, 0);
>               lua_settop(l, 2);

cheers
stoiko


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