"lallous" <la*****@lgwm.org> wrote in message
news:3b**************************@posting.google.c om...
Sure, but it's a little misleading. It's not really a "self" delete.
It's merely a delete function that does a delete on the pointer you send it.
You could just as well write
CTest* x = new CTest;
int* p = new int;
x->selfDelete(p);
will that work? I mean even w/o casting the 'int *' to 'CTest *' ?
It's true I forgot about the cast to int*, but yes, it would work. Since
you're in a static function of CTest, you're not using anything specific to
any object (instance) of CTest. It's basically just working like a regular
function at that point (other than the fact that it's "inside" the class -
in this sense it works like a namespace.) You don't really even need the x
object or the x-> pointer notation to. You could just do this.
int* p = new int;
CTest::selfDelete( (CTest*)p );
The point I was trying to make was that "self" is really misleading. When a
function is static, there is no such thing as "self" or "this". (BTW, in
the Smalltalk language, "self" is actually a term analogous to "this" in
C++).
And you could just as easily take that selfDelete function out of the
class and just call it Delete. See why it's misleading? This is much more
interesting
void selfDelete()
{
delete this;
}
how would 'this' exist if I have the selfDelete() outside of the class?
You're right, it would not. I wrote that as a quick snippet out of context.
The context I should have written was
class CTest
{
public:
void selfDelete() { delete this; }
};
or alternatively
void CTest::selfDelete { delete this; }
