pascal language lesson 2

[cd]

in objpas.inc there is the following code for TObject

Code: [Select]

      procedure TObject.Free;
        begin
           // the call via self avoids a warning
           if self<>nil then
             self.destroy;
        end;
what i want to ask is - does the line

Code: [Select]

           if self<>nil then
has any effect? i mean, if self is nil then a call to Free looks like nil.Free and it will fail because Free is a normal method of an uninitialized object, not a class method.
Finally, could this code be reduced to

Code: [Select]

      procedure TObject.Free;
        begin
           // the call via self avoids a warning
             self.destroy;
        end;
without loss of functionality?

[eny]

Hm, maybe the comment explains it?

[Martin_fr]

Free is not a virtual method, so self can be nil.

you can not do
nil.free;

because nil isn't typed.
but

TObject(nil).Free;
will work.
 (as will any variable.free, if variable is of type TObject or descendand, even if variable is nil.

[Ñuño_Martínez]

in objpas.inc there is the following code for TObject

Code: [Select]

      procedure TObject.Free;
        begin
           // the call via self avoids a warning
           if self<>nil then
             self.destroy;
        end;
what i want to ask is - does the line

Code: [Select]

           if self<>nil then
has any effect? i mean, if self is nil then a call to Free looks like nil.Free and it will fail because Free is a normal method of an uninitialized object, not a class method.
Finally, could this code be reduced to

Code: [Select]

      procedure TObject.Free;
        begin
           // the call via self avoids a warning
             self.destroy;
        end;
without loss of functionality?

AFAIK, the compiler gets this code:

Code: [Select]

   PROCEDURE TObject.Free;
   BEGIN
   // the call via self avoids a warning
      IF SELF <> NIL THEN
         SELF.Destroy;
   END;
...
  Object.Free;
...
and translates it to:

Code: [Select]

   PROCEDURE _TObject_Free_ (SELF: TObjecPtr);
   BEGIN
   // the call via self avoids a warning
      IF SELF <> NIL THEN
         SELF^.TObject_VirtualTable.Destroy^ (SELF);
   END;
...
   _TObject_Free_ (Object);
....

That's why it checks if SELF is NIL.

Note: I know, I know, it doesn't do it in this exact way, but it's a simplification.

[edit] After read Martin_fr's explanation again:

Code: [Select]

TObject (NIL).Free; is translated to

Code: [Select]

_TObject_Free_ (TObjecPtr (NIL));
but

Code: [Select]

TObject (NIL).Destroy; is translated to

Code: [Select]

TObjecPtr (NIL)^.TObject_VirtualTable^.Destroy (TObjectPtr (NIL));
That's why the first doesn't generates a runtime error but the last does it, and also explains why "Free" is recommended.