Re: Copying, hash, and isEqual:
Re: Copying, hash, and isEqual:
- Subject: Re: Copying, hash, and isEqual:
- From: Marcel Weiher <email@hidden>
- Date: Mon, 7 Jun 2004 09:23:58 +0100
OK. Fine... let me rephrase. "But the key is not the same!"
I hope you don't mean this seriously? Are you trying to tell me that
an object is not "the same" as *itself*???
No, he's saying that an object is not the same as a copy of itself
No. (a) A copy of an object is never the same as the object itself.
It is just equal. The same would be object identity. ( == ) (b) the
key provided by the user for storing is the *exact* same key as the one
used for retrieval. Yes, I am aware of the fact that NSDictionary
makes a copy of the key, duh, that is the point of the discussion,
because the copy it makes is not identical to the copy I've given it.
And so you get an implementation that calls itself a dictionary but
doesn't actually work like a dictionary.
unless there is some way (e.g. -hash and -isEqual:) to tell if they
are the same and that dictionary keys are copies (which is
documented).
Hello?! Yes, it *is* documented. For the umpteenth time: that
doesn't mean it is not a bug. It just means it is a design bug, not an
implementation bug.
The bottom line is this:
- NS*Dictionary copies keys. It is documented as doing so and this
is the intended behavior, by design.
YES. That is why I have said numerous times that this is a *design*
bug. Goodness.
No, it's not a design bug.
Yes it is.
Go back to the very start of this thread and you'll find that the
original poster complained about the incompleteness (shallowness) of
the copying of the keys if they are containers.
Precisely!
If the keys are mutable then it is important that they get copied
since the original might get changed after it's inserted as a key but
the hash of the key is only taken at the time of insertion.
Does ANYBODY actually read what I post before shooting from the hip?
At all?
- if object (A) and object (B) are to be used as a key in a
dictionary, then -hash and -isEqual: need to be overridden
accordingly.
Which isn't documented. In fact, the documentation for -hash
recommends the *opposite*. But that's a very minor issue...
The documentation for hash says that it must return the same value for
two objects if -isEqual: returns YES for that pair. The documentation
for -isEqual: says that it's up to your class as to what is "equal"
and you need to define it if you add instance variables.
Precisely. Equality is application-specific. NSDictionary requiring a
very specific notion of equality to work around the fact that it
needlessly creates a new and different object as its key. Clearer now?
NSDictionary states that it uses -hash and -isEqual: so it should be
fairly obvious that these need to be properly implemented
The implementation in NSObject *is* proper. The only reason I created
a subclass is that NSObject doesn't implement NSCopying, which
NSDictionary unreasonable expects to be present.
in a class if your going to use instances of that class as keys. If
copies of objects are to be deemed equivalent that you need to
implement these messages.
What if I deem them equivalent exactly if they are the same, just as
NSObject does?
Your implementation does not follow the second rule. Your key is
not the *very same identical* key as you claim.
Of course it is!
The copy of the object is not the same as the object.
Yes. But that is not *my* key. That is the *new* key that
NSDictionary makes.
The base implementation of NSObject uses the object address in the
-hash and -isEqual: implementations but it also does not implement
-copyWithZone: so there is no way to get a copy of an existing
NSObject (or subclass), so you can't assume that the base
implementations are useful for copies of your subclass. If you
implement copying then you need to make a discussion as to the
equivalence of the original and the result of copying.
Saying that my objects are equal only if they are identical is a
legitimate choice.
@interface MyObject : NSObject <NSCopying>
{}
@end
@implementation MyObject
-copyWithZone:(NSZone*)zone
{
id copy=[[[self class] alloc] init];
return copy;
}
@end
Incidentally, given your object was immutable (implicitly, since is
had no instance variables) it would have been legitimate to simply
return self in this -copyWithZone: (which is what happens with
immutable strings).
Yes, but that is an optimization...
Marcel
--
Marcel Weiher Metaobject Software Technologies
email@hidden www.metaobject.com
Metaprogramming for the Graphic Arts. HOM, IDEAs, MetaAd etc.
1d480c25f397c4786386135f8e8938e4
_______________________________________________
cocoa-dev mailing list | email@hidden
Help/Unsubscribe/Archives:
http://www.lists.apple.com/mailman/listinfo/cocoa-dev
Do not post admin requests to the list. They will be ignored.
| References: | |
| | >Ugly bug in Foundation, beware! (From: Ondra Cada <email@hidden>) |
| | >Is that really a bug at all? (was: Re: Ugly bug in Foundation, beware!) (From: Alastair Houghton <email@hidden>) |
| | >Re: Is that really a bug at all? (was: Re: Ugly bug in Foundation, beware!) (From: Ondra Cada <email@hidden>) |
| | >Re: Is that really a bug at all? (was: Re: Ugly bug in Foundation, beware!) (From: Alastair Houghton <email@hidden>) |
| | >Re: Is that really a bug at all? (was: Re: Ugly bug in Foundation, beware!) (From: Brent Gulanowski <email@hidden>) |
| | >Re: Is that really a bug at all? (was: Re: Ugly bug in Foundation, beware!) (From: Alastair Houghton <email@hidden>) |
| | >Re: Is that really a bug at all? (was: Re: Ugly bug in Foundation, beware!) (From: Ondra Cada <email@hidden>) |
| | >Re: Is that really a bug at all? (was: Re: Ugly bug in Foundation, beware!) (From: Marcel Weiher <email@hidden>) |
| | >Re: Is that really a bug at all? (was: Re: Ugly bug in Foundation, beware!) (From: Brent Gulanowski <email@hidden>) |
| | >Re: Is that really a bug at all? (was: Re: Ugly bug in Foundation, beware!) (From: "Louis C. Sacha" <email@hidden>) |
| | >Re: Is that really a bug at all? (was: Re: Ugly bug in Foundation, beware!) (From: Marcel Weiher <email@hidden>) |
| | >Re: Is that really a bug at all? (was: Re: Ugly bug in Foundation, beware!) (From: Nat! <email@hidden>) |
| | >Re: NSDictionary design bug (was: Re: Ugly bug in Foundation, beware!) (From: Marcel Weiher <email@hidden>) |
| | >Copying, hash, and isEqual: (From: "b.bum" <email@hidden>) |
| | >Re: Copying, hash, and isEqual: (From: Marcel Weiher <email@hidden>) |
| | >Re: Copying, hash, and isEqual: (From: "b.bum" <email@hidden>) |
| | >Re: Copying, hash, and isEqual: (From: Marcel Weiher <email@hidden>) |
| | >Re: Copying, hash, and isEqual: (From: Nicko van Someren <email@hidden>) |