RFR JDK-8159039: sun/net/httpclient/hpack/HeaderTableTest.java fails on some locales

[Pavel Rappo]

Hi Naoto, 

Yes, you are right. This method execution affects all that go after. I will fix it. Thanks!

> On 13 Jun 2016, at 16:58, Naoto Sato <naoto.sato at oracle.com> wrote:
> 
> Does that mean each method test invocation on an separate vm, or the scope is .java? If it's the latter, I would still suggest doing try/finally, otherwise the later test invocations could be affected with the FRENCH locale.
> 
> Naoto
> 
> On 6/13/16 8:52 AM, Chris Hegarty wrote:
>> On 13/06/16 16:39, Naoto Sato wrote:
>>> Hi Pavel,
>>> 
>>> Locale.setDefault() sets the JVM wide default locale, so I'd suggest add
>>> try/finally block to resume the locale to the previous one.
>> 
>> I did suggest something like this to Pavel off-list, but then
>> noticed that the test runs in othervm mode, so won't affect
>> subsequent tests. But of course it could have other negative
>> affects on itself, but somewhat limited at least.
>> 
>> -Chris.
>> 
>>> Naoto
>>> 
>>> On 6/13/16 3:29 AM, Pavel Rappo wrote:
>>>> Hi,
>>>> 
>>>> Could you please review the following change for JDK-8159039?
>>>> 
>>>> http://cr.openjdk.java.net/~prappo/8159039/webrev.00/
>>>> 
>>>> This change contains a fix for 8159039 as well as a number of tiny
>>>> editorial
>>>> changes.
>>>> 
>>>> Thanks,
>>>> -Pavel
>>>>