On Wed, Mar 5, 2008 at 4:14 PM, mmalc crawford <[EMAIL PROTECTED]> wrote:
>
> On Mar 5, 2008, at 1:40 PM, Adam P Jenkins wrote:
>
> > If you define a @property in the interface, then in the
> > implementation you either need use @synthesize to have the compiler
> > automatically generate a getter and setter, or use @dynamic to
> > inform the compiler that you will provide the appropriately named
> > getter and setter methods.
> >
> There is no *need* to specify @dynamic. In fact, doing so will
> conceal a bug (at least until runtime...) if you don't intend to
> create the appropriate methods at runtime (as do, for example, managed
> objects) but forget to provide an implementation. If you don't
> specify @dynamic, then the compiler will warn you if you don't supply
> appropriate methods at compile time.
>
>
> > If you need to define the accessors yourself because the auto-
> > generated ones don't do what you want, then use @dynamic instead
> > like this:
> > @implementation Person
> > @dynamic firstName;
> >
> > - (NSString*)firstName { ... }
> > - (void)setFirstName:(NSString *)newName { ... }
> > @end
>
> It's perfectly reasonable also to do either:
>
>
> @implementation Person
>
>
> - (NSString*)firstName { ... }
> - (void)setFirstName:(NSString *)newName { ... }
> @end
>
>
> or, for example:
>
>
> @implementation Person
> @synthesize firstName;
>
>
> - (void)setFirstName:(NSString *)newName { ... }
> @end
>
>
> In the latter case, the compiler will just synthesise -firstName.
>
>
> One further issue for the sake of raising it:
>
> > @property (setter=mySetMethod:,getter=myMethod) id valueTest;
> >
>
> Note that this implies that the accessor methods are atomic. It's
> comparatively rare (unless you're using GC) that Cocoa developers
> implement atomic accessor methods, so you'd typically specify:
>
> @property (nonatomic, setter=mySetMethod:,getter=myMethod) id valueTest;
>
>
> mmalc
>
Ok first, mmalc, thank you for taking the time to point out exactly
what I needed to see to understand where I was going wrong. My
problem stemmed from the misguided idea that properties were required
for KVC/KVO. Chris' comment about "using non-default names" makes
much more sense now.
Second, for my situation, I ended up with an interface that looks like this:
@property (getter=getValueForBindings, setter=setValueForBindings:,
assign, nonatomic) id value;
And an implementation:
-(void) setValueForBindings:(id)_value
{
[self willChangeValueForKey:@"value"];
{
BOOL didSetOK = [self setStringValue:(NSString *)_value];
// Do something with didSetOK
}
[self didChangeValueForKey:@"value"];
}
-(id) getValueForBindings
{
return( value );
}
Then binding my interface to the key 'valueForBindings'. Not the
prettiest, but it's required for my situation.
Thanks again to everyone for all the help.
--
Jim
http://nukethemfromorbit.com
_______________________________________________
Cocoa-dev mailing list (Cocoa-dev@lists.apple.com)
Please do not post admin requests or moderator comments to the list.
Contact the moderators at cocoa-dev-admins(at)lists.apple.com
Help/Unsubscribe/Update your Subscription:
http://lists.apple.com/mailman/options/cocoa-dev/archive%40mail-archive.com
This email sent to [EMAIL PROTECTED]
