Re: Q) NSNumber class factory methods, such as numberWithXXX

[Richard Frith-Macdonald]

On 16 Mar 2006, at 04:59, Jaeyang Park wrote:

<snip>

> I assume the if-statement, "if (self == abstractClass)" checks whether
> the message is sent to NSNumber or any of concrete sub classes of
> NSNumber.
> As in
>    [NSNumber numberWithFloat:13.0]      ---- A)
>
> or [NSFloatNumber numberWithFloat:13.0]  ---- B)

<snip>
Well, it would have that effect, but we never call +numberWithFloat:  
on NSFloatNumber. It's actually to differentiate between messages  
sent to NSNumber and to (say) MyCustomClass (a subclass of NSNumber  
written by someone else).

> Now, the question is that what is the difference(s) between two body
> of statements.
> If I follow the second (else-part) part,
> I get an instance from one of concrete sub class, in this case of B,
> NSFloatNumber

Call it MyCustomClass rather than NSFloatNumber

> using Zone. Then, the message initWithFloat: is sent to
> that instance. Since, NSFloatNumber does not override initWithFloat:
> method, the one in NSNumber will be used.

<snip>
Ah ... but this never happens for NSFloatNumber ... it only ever  
happens for some subclass you have written ... in which you *should*  
normally have overridden initWithFloat:
If you haven't overridden it ... your class lacks an initialiser, so  
it will fall back on the implementation in NSNumber ... which will  
release your custom class instance and provide a working  
NSFloatNumber  instance instead.

<snip>

> Am I missing something? (The odds are pretty high in this case, since
> I'm a virtually novice in Objective-C and GNUstep/Cocoa/OpenStep)

What this code does is ....

If you call +numberWithFloat: on NSNumber, it returns an instance of  
a private subclass of NSNumber encapsulating a float.
If you call +numberWithFloat: on your own subclass of NSNumber then  
an instance of your subclass is created
  if your subclass has the initWithFloat: initialiser, it is called  
to initialise it.
  if your subclass does not have the initialiser, the superclass  
implementation destroys it
  (because it can't know how to initialise your subclass) and  
creates an object which will work.