java.util patch for Warning Cleanup Day

Naoto Sato naoto.sato at oracle.com
Tue Dec 6 21:58:34 UTC 2011


On 12/6/11 1:29 PM, Mike Duigou wrote:
> I have reviewed this patch and also run it against the java/util jtreg test. It looks good to me.
>
> Some comments:
>
> Currency.java:
> - The current mercurial source has one of your changes but the remaining two changes, diamond and @SW are still valid.

For the @SW, keep the one in the current TL repo, as it has narrower scope.

Naoto

>
> ResourceBundle:
> - This file also has changes in the current mercurial source. Your version, IMO, is an improvement over the mercurial version and should be applied.
>
> HashMap:
> -The changes to use wildcard instead of the type parameters are a bit disappointing but reasonable.
>
> TreeSet:
> - The change to readObject() (using a single constructor) is theoretically outside of the scope of "warnings only" but I am OK with it.
>
> Mike
>
> On Dec 3 2011, at 07:12 , Rémi Forax wrote:
>
>> On 12/03/2011 03:33 PM, Alan Bateman wrote:
>>> On 03/12/2011 12:37, Rémi Forax wrote:
>>>> I've started to remove warnings from java.util
>>>> and I'm not able to cleanly retrofit java.util.PropertyPermission
>>>> because java.util.Collections.enumeration is wrongly typed.
>>>>
>>> Rémi - you might want to sync up with Mike Duigou as I believe he has been working on the java.util area including collections. Also best to move any discussions on collections/util area to core-libs-dev too.
>>>
>>> -Alan
>>
>> too late :)
>>
>> Here is a webrev that remove most of the unchecked cast/raw generics warnings
>> http://cr.openjdk.java.net/~forax/WarningDay1/webrev/
>>
>> Rémi
>>
>




More information about the core-libs-dev mailing list