Thank you, Phil. I have updated the fix based on your suggestion. Please find
the new version here: http://cr.openjdk.java.net/~dmarkov/8163979/webrev.03/
Thanks,
Dmitry
> On 14 Feb 2017, at 01:54, Phil Race <philip.r...@oracle.com> wrote:
>
> 63 if (workList == null || (workList != null && workList.isEmpty()))
> {
>
>
> I think we can safely write just :
>
> if (workList == null || workList.isEmpty()) {
>
> Other than that it seems fine.
>
> -phil.
>
> On 02/13/2017 11:43 AM, Dmitry Markov wrote:
>> Hi Sergey,
>>
>> Thank you for the review. You suggested a really good solution for the issue.
>> I have updated the fix based on your recommendation. The new version is
>> located at http://cr.openjdk.java.net/~dmarkov/8163979/webrev.02/
>> <http://cr.openjdk.java.net/~dmarkov/8163979/webrev.02/>
>>
>> Thanks,
>> Dmitry
>>> On 13 Feb 2017, at 20:22, Sergey Bylokhov <sergey.bylok...@oracle.com>
>>> <mailto:sergey.bylok...@oracle.com> wrote:
>>>
>>> Hi, Dmitry.
>>> Probably we can reuse Java_sun_lwawt_macosx_CInputMethod_getNativeLocale?
>>>
>>> And change the java code in the CInputMethodDescriptor to something like:
>>>
>>> static Object[] getAvailableLocalesInternal() {
>>> List<?> workList = nativeGetAvailableLocales();
>>> Locale nativeLocale = CIM.getNativeLocale()
>>>
>>> if (workList == null && nativeLocale == null)) {
>>> return new Object[] {
>>> Locale.getDefault()
>>> };
>>> }
>>> if(workList != null && !workList.isEmpty() && nativeLocale == null) {
>>> return workList.toArray();
>>> }
>>> if((workList == null || workList.isEmpty()) && nativeLocale != null) {
>>> return new Object[] {
>>> nativeLocale;
>>> };
>>> }
>>> if (!workList.contains(nativeLocale)){
>>> workList.add(nativeLocale);
>>> }
>>> return workList.toArray();
>>> }
>>>
>>>
>>>> I am sorry, I sent the incorrect version of the fix, (i.e.
>>>> http://cr.openjdk.java.net/~dmarkov/8163979/webrev.00/
>>>> <http://cr.openjdk.java.net/~dmarkov/8163979/webrev.00/>). It releases
>>>> currenLocale two times.
>>>> The correct version is located at
>>>> http://cr.openjdk.java.net/~dmarkov/8163979/webrev.01/
>>>> <http://cr.openjdk.java.net/~dmarkov/8163979/webrev.01/>
>>>> Could you review the updated fix, please?
>>>>
>>>> Thanks,
>>>> Dmitry
>>>>> On 13 Feb 2017, at 16:58, Dmitry Markov <dmitry.mar...@oracle.com>
>>>>> <mailto:dmitry.mar...@oracle.com> wrote:
>>>>>
>>>>> Hello,
>>>>>
>>>>> Could you review a fix for jdk9, please?
>>>>>
>>>>> bug: https://bugs.openjdk.java.net/browse/JDK-8163979
>>>>> <https://bugs.openjdk.java.net/browse/JDK-8163979>
>>>>> webrev: http://cr.openjdk.java.net/~dmarkov/8163979/webrev.00/
>>>>> <http://cr.openjdk.java.net/~dmarkov/8163979/webrev.00/>
>>>>>
>>>>> Problem description:
>>>>> If the current keyboard layout is set to non-default value, (e.g. ABC,
>>>>> ABC Extended, etc) during Java launch, the input method functionality is
>>>>> not initialised correctly and may not work for some input methods, (e.g.
>>>>> OpenVanilla input method). The initialisation problems are caused by
>>>>> nativeGetAvailableLocales() function from CInputMethod.m. The function
>>>>> tries to obtain the list of available input method locales using platform
>>>>> API. If the retrieval is failed for some reason,
>>>>> nativeGetAvailableLocales() returns empty list or null.
>>>>>
>>>>> Fix:
>>>>> The function nativeGetAvailableLocales() should return the list contained
>>>>> the current input method locale, if it cannot retrieve the list of
>>>>> available locales from the platform.
>>>>>
>>>>>
>>>>> Thanks,
>>>>> Dmitry
>
