<i18n dev> RFR: 8258805: Japanese characters not entered by mouse click on Windows 10

Dmitry Markov dmarkov at openjdk.java.net
Thu Jan 21 13:51:59 UTC 2021


On Wed, 20 Jan 2021 12:28:36 GMT, Alexey Ivanov <aivanov at openjdk.org> wrote:

>>> > Fix:
>>> > It is necessary to take care of unconfirmed composition string once the IME is going to be disabled.
>>> 
>>> The fix commits the unconfirmed composition string. Committing is better than discarding. Is it possible to preserve the 
>> state and to leave the string uncommitted?
>> 
>> The fix reverts the previous (correct) behaviour back. It is unnecessary to store the state and keep the string. That activity may be quite complicated and requires additional resources especially if there are several components with active IME
>
>> > The fix commits the unconfirmed composition string. Committing is better than discarding. Is it possible to preserve the
>> > state and to leave the string uncommitted?
>> 
>> The fix reverts the previous (correct) behaviour back.
> 
> According to the bug description, it used to stay in uncommitted state.
> 
>> It is unnecessary to store the state and keep the string. That activity may be quite complicated and requires additional resources especially if there are several components with active IME
> 
> I definitely agree that committing the text is better behaviour than discarding uncommitted text.
> 
> I am for accepting these changes unless we can keep the text uncommitted and allow the user to work with text when the focus moves back to the text component.

An interesting thing is that the test (see AWTTextTest1.java) behaves like a native application if TextField is replaced with JTextField. It appears AWT and Swing components behave differently while processing input method requests. That’s a great topic to discuss but it is out of scope for this PR. We can open a separate bug for it, if any.

Going back to the current PR. I updated the fix. Now we will commit the composition string if there is an active client. That changes eliminates the issue described in JDK-8258805. Also the behaviour of AWTTextTest1 is the same as it used to be on the previous versions w/o the fix.

-------------

PR: https://git.openjdk.java.net/jdk/pull/2142


More information about the i18n-dev mailing list