| Register | FAQ | Calendar | Search | Today's Posts | Mark Forums Read |
|
#1
| |||
| |||
| [This is a supercede, just to correct a broken Subject] I've got an enum that contains Things that come in pairs. I'd like each Thing to have a *final* reference to it's mate, that in half of the cases appears later in the list of Thing-instances. e.g.: enum Thing { T1, T2; final Thing mate; } What could I do, to make T2 "final"ly the mate of T1 and vice versa ? I know some alternatives, myself, like just making the reference non-final (and initialize it in the enum's static {...} block) or defining an abstract method in the enum like "abstract Thing getMate();" and implement it for each instance to return its mate... Please only followup, if you know a trick for final fields, or if you know that it is strictly impossible that way. |
|
#2
| |||
| |||
| On Mon, 25 Aug 2008 12:51:30 +0000, Andreas Leitgeb wrote: > [This is a supercede, just to correct a broken Subject] > > I've got an enum that contains Things that come in pairs. > > I'd like each Thing to have a *final* reference to it's mate, that in > half of the cases appears later in the list of Thing-instances. > > e.g.: > enum Thing { > T1, T2; > final Thing mate; > } [Snip] This is a problem because you really want to create a pair at a time not a single thing at a time. You need to initialize the two mates at creation, but you can't since the other thing hasn't been created yet. I would suggest having a class called pair instead of a class called Thing. It would make these problems go away. Otherwise you're going to have to give up on the finality of mate. -- Kenneth P. Turvey <kt-usenet@squeakydolphin.com> http://www.electricsenator.net I took the initiative in creating the Internet. -- Al Gore |
|
#3
| |||
| |||
| Andreas Leitgeb wrote: > [This is a supercede, just to correct a broken Subject] > > I've got an enum that contains Things that come in pairs. > > I'd like each Thing to have a *final* reference to > it's mate, that in half of the cases appears later > in the list of Thing-instances. > > e.g.: > enum Thing { > T1, T2; > final Thing mate; > } > > What could I do, to make T2 "final"ly the mate of T1 > and vice versa ? > > I know some alternatives, myself, like just making the > reference non-final (and initialize it in the enum's > static {...} block) or defining an abstract method in > the enum like "abstract Thing getMate();" and implement > it for each instance to return its mate... > > Please only followup, if you know a trick for final fields, > or if you know that it is strictly impossible that way. Not having played with enums, I don't know what games you can play with their constructors, but this kind of thing will work for a "normal" class. (The trick is to construct the mate if and only if it doesn't already exist.) import java.util.Map; import java.util.HashMap; public class Mates { public final Mates myMate; private final int value; private static Map<Integer, Mates> instances = new HashMap<Integer, Mates>(); public Mates(int i) { value = i; instances.put(i, this); int mateValue = (i % 2 == 0) ? i + 1 : i - 1; if (instances.get(mateValue) == null) new Mates(mateValue); myMate = instances.get(mateValue); } public String toString() { return Integer.valueOf(value).toString(); } public static void main(String[] args) { Mates[] arr = new Mates[10]; for (int i = 0; i < 10; i++) arr[i] = new Mates(i); for (int i = 0; i < 10; i++) { System.out.println(arr[i] + " is the mate of " + arr[i].myMate); } } } |
|
#4
| |||
| |||
| Mike Schilling <mscottschilling@hotmail.com> wrote: > Not having played with enums, I don't know what games you can play with > their constructors, Unfortunately, not enough :-( The "next" enum-instance isn't even allocated, while the current instance's constructor is not completed. As it seems there really is no way. In my case at hand, I've now changed my code to not really use the enum-instances as such, but in the enum's static initializer I create a few arrays, that are filled with return values of method calls on the instances. Later I do no longer touch the instances, but only read the tables to get the infos I need. Since the pairs were subsequent and I now use int's instead of Thing-references, accessing the mate's results is now just a matter of indexing with "i^1" instead of "i". Thanks to all. Especially the trick for non-enums with pair-wise creation (posted by a few) was new to me. PS: the question about that "final" field was rather of educational purpose. Even if it had worked, I'd have switched to int-based processing at some time. PPS: doing a switch on the index turned out to be noticably faster than overriding an abstract method in the enum. I really tried it, and wrote 40 anonymous subclasses for my 40 instances. I still wouldn't claim that this always needs to be the case. (Yes, I did the profiling first, and it did indicate, that it would pay to optimize.) PPPS: since an enum's .values() always does a .clone on it's internally kept inventory-array ($VALUES), I also added a static field: private static final Thing things[]=Thing.values(); to help me get the Instance by index without cloning the array each time. To my surprise I also didn't find a method like valueOf(int). |
![]() |
| Thread Tools | |
| Display Modes | |
In an effort to better serve ads to our visitors, cookies are used on objectmix.com. For more information, check out our Privacy Policy.