<Swing Dev> [8] Review request for 8023392: Swing text components printed with spaces between chars
anton nashatyrev
anton.nashatyrev at oracle.com
Mon Aug 26 12:41:52 UTC 2013
Hello Alexander,
On 26.08.2013 16:05, Alexander Scherbatiy wrote:
> I see that SwingUtilities2.drawString(..) and
> SwingUtilities2.drawChars(...) are used a lot in the Swing but the
> issue is only about printing.
> Could removing trailing spaces in these methods affect other
> non-printed components?
All the changes are made under isPrinting() condition, so the behavior
when rendering on screen/buffer should remain the same.
> Is it possible to add a test for the fix?
Sure, but I believe only manual is suitable for this case.
I will add one...
Thanks!
Anton.
>
> Thanks,
> Alexandr.
>
>
> On 8/22/2013 3:57 PM, anton nashatyrev wrote:
>> Hello,
>> could you please review the following fix:
>>
>> fix: http://cr.openjdk.java.net/~vkarnauk/8023392/jdk8/webrev.00/
>> <http://cr.openjdk.java.net/%7Evkarnauk/8023392/jdk8/webrev.00/>
>> bug: http://bugs.sun.com/view_bug.do?bug_id=80223392 (the bug is
>> still not replicated here, so internal link might be used if available)
>>
>> The problem description: While printing the text output is
>> handled a bit differently: the text layout is justified for the text
>> width (in theory on the screen this text should look identically as
>> without justification), which helps printer to output text more
>> evenly (https://bugs.openjdk.java.net/browse/JDK-6488219).
>> But the problem is the TextLayout.getJustifiedLayout() trim the
>> trailing spaces from the string before laying it out. In the case
>> when the string contains trailing spaces it is justified for the
>> width greater than required.
>>
>> The fix description: as far as we likely don't want to change
>> TextLayout.getJustifiedLayout() behavior, we should respect that
>> implementation feature and justify the text for the width of the
>> string without trailing spaces. I.e. trim trailing spaces before
>> TextLayout.getAdvance().
>>
>> Thanks!
>> Anton.
>
More information about the swing-dev
mailing list