• Derek Parnell (12/12) Jun 07 2006 What is the rationale for not allowing the '!' operator to be overloaded...
• Oskar Linde (8/16) Jun 07 2006 I guess that for classes, the reason is that !c is defined as

Derek Parnell <derek psych.ward> writes:

What is the rationale for not allowing the '!' operator to be overloaded?

I ask because I had developed a Bool class but I can't use it in place of
'bool' in cases where the syntax is ...

  if (! mybool ) ...

I can overload the '~' operator but that is not as natural for boolean
thinking as that operation is more suited to arithmetical functionality.

-- 
Derek
(skype: derek.j.parnell)
Melbourne, Australia
"Down with mediocracy!"
7/06/2006 5:05:59 PM

Oskar Linde <oskar.lindeREM OVEgmail.com> writes:

Derek Parnell skrev:
 What is the rationale for not allowing the '!' operator to be overloaded?

I guess that for classes, the reason is that !c is defined as
"c is null", i.e. the ! refers to the reference, not the class instance. 
I.e the same reason as why = can't be overloaded for classes.

For structs, I fail to see any reason not to allow overloading of either 
! or =.

 I ask because I had developed a Bool class but I can't use it in place of
 'bool' in cases where the syntax is ...
 
   if (! mybool ) ...
 
 I can overload the '~' operator but that is not as natural for boolean
 thinking as that operation is more suited to arithmetical functionality.

Matlab uses ~ for boolean negation... But I agree with you.

/Oskar