<Swing Dev> RFR(S): 8234913 Improve parsing of Length Units in javax/swing/text/html/CSS

Volodin, Vladislav vladislav.volodin at sap.com
Sat Dec 21 12:49:22 UTC 2019


Hello Sergey,

thank you for the comment. I was able to write the automated test. I had to use some copy-paste technique from another test case to make this option "-verbose:class" work.
In addition to my previous idea, I decided to slightly improve the code by removing unnecessary boxing & unboxing, plus also corrected a type.

The patch is here: http://cr.openjdk.java.net/~clanger/webrevs/8234913.1/

If you think that my work might be helpful, I have identified few other places, where exceptions were either improperly handled, or thrown when it can be avoided. I can improve that part as well.
Regarding boxing & unboxing optimizations, I followed suggestions from IntelliJ.

Kind regards,
Vlad

-----Original Message-----
From: Sergey Bylokhov <Sergey.Bylokhov at oracle.com> 
Sent: Dienstag, 3. Dezember 2019 01:13
To: Volodin, Vladislav <vladislav.volodin at sap.com>
Cc: swing-dev at openjdk.java.net
Subject: Re: <Swing Dev> RFR(S): 8234913 Improve parsing of Length Units in javax/swing/text/html/CSS

Probably "java -verbose:class" might help? I guess after the fix
this exception should not be loaded on successful parsing of "%".

On 12/2/19 2:38 pm, Volodin, Vladislav wrote:
> Hello Sergey,
> 
> indeed I wish I can create it. The main difficulty is that the exception (when it occurs) is handled by try-catch and swallowed. So only if I turn on the debugger with “any exception” breakpoint, I will be able to find that the execution path was wrong.
> 
> I am new here, and I will appreciate if you can give me an idea how to test the exception presence? Something with reflection? E.g. count a number of NumberFormatException constructor calls?
> 
> Kind regards,
> Vlad
> 
> Sent from myFone
> 
>> On 2. Dec 2019, at 23:24, Sergey Bylokhov <Sergey.Bylokhov at oracle.com> wrote:
>>
>> Hi, Vladislav.
>>
>> Is it possible to provide an automated test for this change?
>>
>> On 11/28/19 2:08 am, Volodin, Vladislav wrote:
>>> Hello everyone,
>>> I'd like to contribute a little improvement to javax/swing/text/html/CSS. The issue is that "font-size: 100%" throws NumberFormatException for 100%, because of a wrong execution path. It is possible to reproduce the issue with the code below (but you should create Java exception breakpoints to see the place):
>>> package com.test;
>>> import javax.swing.text.MutableAttributeSet;
>>> import javax.swing.text.SimpleAttributeSet;
>>> import javax.swing.text.html.CSS;
>>> import javax.swing.text.html.StyleSheet;
>>> public class Main {
>>>     public static void main(String[] args) {
>>>         StyleSheet ss = new StyleSheet();
>>>         MutableAttributeSet attr = new SimpleAttributeSet();
>>>         ss.addCSSAttribute(attr, CSS.Attribute.FONT_SIZE, "100%");
>>>     }
>>> }
>>> Bug: https://bugs.openjdk.java.net/browse/JDK-8234913
>>> Webrev: http://cr.openjdk.java.net/~clanger/webrevs/8234913.0/
>>> Kind regards,
>>> Vlad
>>
>>
>> -- 
>> Best regards, Sergey.


-- 
Best regards, Sergey.


More information about the swing-dev mailing list