more TreeMap questions (descendingMap, headMap and tailMap)

charlie hunt charlie.hunt at sun.com
Sat May 24 01:36:14 UTC 2008


If that's the definition given to "restricted range", I think there is 
some inconsistencies in the behavior of TreeMap.

If you play around with some of descendingMap(), subMap(), headMap() and 
tailMap() variations such as those in the example program in my original 
e-mail, I think you'll see there are some cases where 
IllegalArgumentException is thrown.

Are you suggesting that "could ever contain" to mean, for example, if I 
have Integer keys, that all fromKey/toKey must be between 
Integer.MIN_VALUE and Integer.MAX_VALUE?   Or, is it "restricted" to the 
bounds of the "backing" map?

Perhaps I'm just being naive, but it seems "restricted range" is rather 
ambigious.

charlie ...

Martin Buchholz wrote:
> The "restricted range" refers to the elements that the submap could
> possibly ever contain,
> not the elements that it currently contains.
>
> This may not be described in the clearest way in the doc,
> but I'm not motivated to work on clarifying it,
> partly because I've generally felt it was a design mistake
> to try to enforce these kinds of range restrictions,
> although I helped to implement the enforcing code.
>
> Martin
>
> On Fri, May 23, 2008 at 6:11 AM, charlie hunt <charlie.hunt at sun.com> wrote:
>   
>> I have run across a couple cases which may or may not be defects in TreeMap.
>>  Or, it's simply a case where I am not understanding the Java docs for
>> TreeMap & NavigableMap correctly.
>>
>> When looking at the Java docs for TreeMap / NavigableMap tailMap() I read,
>> "Returns a view of the portion of this map whose keys are greater than (or
>> equal to, if |inclusive| is true) |fromKey|. The returned map is backed by
>> this map, so changes in the returned map are reflected in this map, and
>> vice-versa. The returned map supports all optional map operations that this
>> map supports."  More importantly, I read this method will throw an
>> "||IllegalArgumentException if this map itself has a restricted range, and
>> |fromKey| lies outside the bounds of the range".
>>
>> I interpret the latter here to mean if I create a subMap(), headMap() or
>> tailMap() I have a map which has "a restricted range"?
>>
>> Assuming my interpretation is correct.  I wrote a test program on a TreeMap
>> containing Integer keys and String values.  I populated the TreeMap with
>> keys 5 - 5000 at every 5th integer, (i.e. 5,10,15,20,25, ....
>> 4985,4990,4995,5000).  And, for each key, I just created a string such as,
>> "Value stored is -> <key>".
>>
>> I found if I do a subMap(Integer.valueOf(5),true,Integer.valueOf(5000),true)
>> I get a NavigableMap back containing the same keys/values as the TreeMap.
>>  If I then do a tailMap(Integer.valueOf(4850),false), I get a NavigableMap
>> back containing keys 4855 - 5000 and their corresponding values.  All as
>> expected to this point.  If I now do a tailMap(Integer.valueOf(4850,false)
>> on that last NavigableMap I got back which contains keys 4855 - 5000,
>> according to the Java doc, I would expect an IllegalArgumentException to be
>> thrown since 4850 is outside the bounds of 4855 - 5000.   But, this is not
>> what happens. I get the same NavigableMap back containing keys 4855 - 5000.
>>
>> I can produce a similar result by taking the same initial TreeMap, doing a
>> descendingMap() on it.  Then, doing a headMap(Integer.valueOf(4850))
>> followed by a headMap(Integer.valueOf(4850)).  I am interpreting that a
>> headMap(K key) on a descendingMap is essentially the same as a tailMap(K
>> key, false) on an ascending map of the same underlying TreeMap.
>>
>> Am I interpreting some completely wrong?  Or, is the not throwing of an
>> IllegalArgumentException expected behavior?  Or, is this indeed a defect?
>>
>> Attached is a simple program which illustrates what I have described.
>>
>> thanks,
>>
>> charlie ...
>>
>>
>> --
>>
>> Charlie Hunt
>> Java Performance Engineer
>>
>> <http://java.sun.com/docs/performance/>
>>
>>
>>     




More information about the core-libs-dev mailing list