Re: [slang-users] wrong predeclaration makes function unaccessable

[Jörg Sommer <joerg@xxxxxxxxxxxx>]

Hallo John,

"John E. Davis" <davis@xxxxxxxxxxxxx> wrote:
> =?UTF-8?Q?J=C3=B6rg?= Sommer <joerg@xxxxxxxxxxxx> wrote:
>>
>>#v+
>>implements("gaga");
>>
>>static define foo();			% wrong behaviour
>>% private define foo();			% correct behaviour
>>
>>static define bla()
>>{
>>    foo();
>>}
>
> When "bla" gets compiled, the only version of "foo" that is available
> is the one that exists in the static namespace.  Hence it is the one
> that gets linked to "bla".  And because the body of the function does
> not exist, the function will do nothing when called.

When you say the static function has an empty body this should be
equivalent to the definition above:

static define foo() {}

But this gives an error. I think, SLang should complain about the
previous definition of foo() with a different modifier as it does if the
functions have bodies.

Bye, Jörg.
-- 
Der Mensch hat die Atombombe erfunden.
Keine Maus der Welt käme auf die Idee, eine Mausefalle zu konstruieren.

_______________________________________________
To unsubscribe, visit http://jedsoft.org/slang/mailinglists.html