| Re: Cleaning up inherited object classes [message #37171] |
Wed, 03 December 2003 20:15 |
David Fanning
Messages: 11724
Registered: August 2001
|
Senior Member |
|
|
Marc Schellens writes:
> Well another way would be to consider it already
> when writing your program.
> I always have the cleanup method just before the
> XX__define procedure. So for every inheritance I immediately add
> the superclass cleanup call and for every pointer the ptr_free (object,
> obj_destroy) function.
Well, yeah. Now I do too. :-)
Cheers,
David
--
David W. Fanning, Ph.D.
Fanning Software Consulting, Inc.
Phone: 970-221-0438, E-mail: david@dfanning.com
Coyote's Guide to IDL Programming: http://www.dfanning.com/
Toll-Free IDL Book Orders: 1-888-461-0155
|
|
|
|
| Re: Cleaning up inherited object classes [message #37172 is a reply to message #37171] |
Wed, 03 December 2003 18:38  |
marc schellens[1]
Messages: 183
Registered: January 2000
|
Senior Member |
|
|
Well another way would be to consider it already
when writing your program.
I always have the cleanup method just before the
XX__define procedure. So for every inheritance I immediately add
the superclass cleanup call and for every pointer the ptr_free (object,
obj_destroy) function.
marc
David Fanning wrote:
> Ben Tupper writes:
>
>
>> I would defer to other's on the question if a subclass MUST have its own
>> CLEANUP method. I haven't tried it, but I'm not sure that it does
>> (unless it has its own pointers and objects to cleanup.)
>
>
> I'll tell you this, I have spent an *untold* number
> of hours chasing down memory leaks only to find that
> I forget to call a superclass CLEANUP method from my
> object's CLEANUP method! I've done it so often this
> is the *FIRST* thing I look for when I'm chasing
> memory leaks now.
>
> Who says you can't teach an old dog new tricks?
>
> Cheers,
>
> David
>
|
|
|
|
| Re: Cleaning up inherited object classes [message #37177 is a reply to message #37172] |
Wed, 03 December 2003 12:28 |
David Fanning
Messages: 11724
Registered: August 2001
|
Senior Member |
|
|
Ben Tupper writes:
> I would defer to other's on the question if a subclass MUST have its own
> CLEANUP method. I haven't tried it, but I'm not sure that it does
> (unless it has its own pointers and objects to cleanup.)
I'll tell you this, I have spent an *untold* number
of hours chasing down memory leaks only to find that
I forget to call a superclass CLEANUP method from my
object's CLEANUP method! I've done it so often this
is the *FIRST* thing I look for when I'm chasing
memory leaks now.
Who says you can't teach an old dog new tricks?
Cheers,
David
--
David W. Fanning, Ph.D.
Fanning Software Consulting, Inc.
Phone: 970-221-0438, E-mail: david@dfanning.com
Coyote's Guide to IDL Programming: http://www.dfanning.com/
Toll-Free IDL Book Orders: 1-888-461-0155
|
|
|
|
| Re: Cleaning up inherited object classes [message #37179 is a reply to message #37177] |
Wed, 03 December 2003 12:17 |
btt
Messages: 345
Registered: December 2000
|
Senior Member |
|
|
M. Katz wrote:
> Cleaning up is my least favorite activity. Were my living room an IDL
> object I'm sure it'd be full of dangling pointer references. Here's a
> question regarding objects' Cleanup methods and inheritance.
>
> When an object inherits another object, methods can be overridden. So
> what happens to the CleanUp method? It is special.
>
> If my House object inherits the Living_Room and Bathroom object
> classes, will a call to HOUSE::CleanUp also call Living_Room::CleanUp
> and Bathroom::Cleanup when obj_destroy, self is called?
>
> Let me put that another way. Suppose an object class, A, has pointer
> fields. Unless someone tells me otherwise, I assume it's a good idea
> two specifically free the pointers in that object's Cleanup routine.
> Now, suppose another object class, B, inherits A. B has its own
> pointers to clean up as well, so I write that into its cleanup
> routine.
>
> It is sufficient to write the Cleanup methods like this?
>
> pro Bobj::CleanUp
> ptr_free, self.Bpointer
> obj_destroy, self
> end
>
> pro Aobj::CleanUp
> ptr_free, self.Apointer
> obj_destroy, self
> end
>
> Will Bobj::CleanUp's call to "obj_destroy, self" also call
> Aobj::Cleanup so that self.Apointer can be freed as the object is
> destroyed?
>
> Also, does the destruction of an object that contains a pointer field
> also inherently free the pointer? or is it necessary to specifically
> ask for that in the Cleanup?
>
> Now if I could only get the House::TakeOutTheTrash method to work
> reliably my wife would be thrilled.
>
> M. Katz
Hello,
I think you simply call the cleanup method for each superclass.
The following is the way I do it.
PRO House::Cleanup
DoMyOwnCleanUpofLocalPointersAndObjects
Self->Living_Room::CleanUp
Self->Bath_Room::Cleanup
END
Provided that HOUSE was defined this way.
PRO House__Define
struct = {House, $
INHERITS Bath_Room, $
INHERITS Living_Room}
END
If Bath_Room inherits from some other object, such as READING_ROOM, then
it will call that superclass' cleanup method if you have Bath_Room's
cleanup as ...
PRO Bath_Room::Cleanup
self->Reading_Room::Cleanup
END
I would defer to other's on the question if a subclass MUST have its own
CLEANUP method. I haven't tried it, but I'm not sure that it does
(unless it has its own pointers and objects to cleanup.)
Cheers,
Ben
|
|
|
|
comp.lang.idl-pvwave archive
Members
Search
Help
Login
Home
