Hi all,
OK after reviewing the parsing of <default-locale> / <supported-locales>
it appears that we allow an additional separator which BCP 47 does not
allow. So my proposal here is to mandate the BCP 47 is used, but if no
language is found in the parsed Locale then to default back to the old
algorithm and use it to get the Locale.
I'll go ahead with adding that to the Javadoc and specification in the
relevant sections if nobody objects before 8/9.
Thanks!
Kind regards,
Manfred Riem
On 8/27/15, 5:03 AM, Bauke Scholtz wrote:
> I'd say +1 for using Locale#forLanguageTag() anyway. Let the
> language/country interpreting job up to the standard API instead of
> manually doing it.
>
> I only wonder for which cases exactly the old code would work where
> Locale#forLanguageTag() fails. If there's none, we could reduce code.
>
> Cheers, B
>
>
> On Wed, Aug 26, 2015 at 5:47 PM, manfred riem <manfred.riem_at_oracle.com
> <mailto:manfred.riem_at_oracle.com>> wrote:
>
> Hi all,
>
> The implementation is pretty much already done in Mojarra, but we
> have not required BCP 47 support as it required JavaSE 7. Now that
> JSF 2.3 is requiring JavaSE 8 we could easily state it is a
> requirement.
>
> Howerver we do have specification adjustments that need to be done
> in how <supported-locales> and <default-locale> in
> faces-config.xml need to be parsed (as that is in the specification).
>
> As I said I don't feel strong either way.
>
> So far we have one EG member saying +1.
>
> Thanks!
>
> Kind regards,
> Manfred Riem
>
> On 8/25/15, 10:14 PM, Josh Juneau wrote:
>> No preference on this, but how difficult is the implementation?
>> If this is going to help make the Locale support in JSF more
>> current, then perhaps it is worth the fix.
>>
>> Josh Juneau
>> juneau001_at_gmail.com <mailto:juneau001_at_gmail.com>
>> http://jj-blogger.blogspot.com
>> https://www.apress.com/index.php/author/author/view/id/1866
>>
>>
>> On Tue, Aug 25, 2015 at 8:07 AM, manfred riem
>> <manfred.riem_at_oracle.com <mailto:manfred.riem_at_oracle.com>> wrote:
>>
>> Hi all,
>>
>> While I do not have a preference if we do not get a response
>> we'll go with option #1.
>>
>> Thanks!
>>
>> Kind regards,
>> Manfred Riem
>>
>> On 8/24/15, 11:39 AM, manfred riem wrote:
>>
>> Hi all,
>>
>> Looking at this issue I think we have 2 options:
>>
>> 1. Do nothing and let the JSF runtime do what is has done
>> in the past (status quo).
>> 2. Make it so that when dealing with Locales the
>> Locale.forLanguageTag() is consulted first,
>> and only if it fails let the old code do its work.
>>
>> I have no particular preference, so please voice your
>> support for either.
>>
>> Note I like to close the loop on this before 8/31.
>>
>> Thanks!
>>
>> Kind regards,
>> Manfred Riem
>>
>>
>>
>