Some Eclipse Foundation services are deprecated, or will be soon. Please ensure you've read this important communication.

Bug 312708

Summary: TVT36:TCT202: CHS: Corrupted Charaters
Product: [Modeling] GMF-Runtime Reporter: CDE Administration <cdeadmin>
Component: GeneralAssignee: Inbox <gmf.runtime.diagram-inbox>
Status: CLOSED FIXED QA Contact:
Severity: normal    
Priority: P3 CC: ahunter.eclipse, camle, kitlo
Version: 2.3   
Target Milestone: ---   
Hardware: PC   
OS: Windows 7   
Whiteboard:
Attachments:
Description Flags
13.000.080_corruption.gif
none
13.000.070.gif
none
202_diagram.gif
none
202_preview.gif
none
202_ja.gif none

Description CDE Administration CLA 2010-05-12 16:21:40 EDT
<response_by> Ma Xiao Long at 2010.05.07.22.42.52 </response_by>
NLS Build Label: 20100506-1500-nl1
Language: S.Chinese

- Steps To Recreate Problem:
1)  Double click geoshapes diagram you just created to view it, if it is not already displayed
2)  Click on diagram surface
3)  Select Add Rectangle.

- Problem Description:
We found "Text" in the added Rectagngle is corrupted after we click any other places, pls help to check, thanks!

<response_by> rewagner at 2010.05.10.10.01.45 </response_by>
I was unable to recreate this problem.  I would suggest starting a new workspace and retrying this test.

Thanks
Ross

<response_by> Keiko Terada at 2010.05.11.20.04.58 </response_by>
same problem for Japanese
build label: 20100509-1500

Translated "Text" is shown as default when Rectangle is shown on the diagram.
When I add 1 after the word and press Return key, the translated string is not shown.

<response_by> Keiko Terada at 2010.05.11.20.29.08 </response_by>
Print preview shows the translation without corruption.  The dialog view brings corruptions.
Please compare the attached images.

<response_by> Ma Xiao Long at 2010.05.11.21.56.07 </response_by>
We tried with a new workspace but the problem is still there.  As Keiko-san said, when we first drag a Rectangle to Geoshape Diagram, Text is highlighted and it shows correctly, but if we clicked other place in Geoshape Diagram or drag another item, it will show corrupted.  Pls advise, thanks!

<response_by> rewagner at 2010.05.12.08.19.56 </response_by>
This is very strange.  I cannot recreate this on my test machine, but I have tried it on both of your test machines and I see the problem.  It may have something to do with the machines themselves.  I will look into this further and get back to you as soon as I can with an answer.

Thanks,
Ross

<response_by> rewagner at 2010.05.12.15.17.52 </response_by>
Upon further testing, it seems that this problem happens with only Double Byte (DBCS) languages such as Japanese, Chinese and Korean.  I have tried it on my machine with those languages, but the problem doesn't appear.  However, I have tried it with the respective lab machines and I have seen it occur.

I will create a bug and this should be fixed.

Thanks,
Ross

<response_by> rewagner at 2010.05.12.15.18.07 </response_by>
This article was reassigned from Category:''TVT/Testing,Inbox''.
Comment 1 CDE Administration CLA 2010-05-12 16:21:46 EDT
Created attachment 168251 [details]
13.000.080_corruption.gif
Comment 2 CDE Administration CLA 2010-05-12 16:21:48 EDT
Created attachment 168252 [details]
13.000.070.gif
Comment 3 CDE Administration CLA 2010-05-12 16:21:54 EDT
Created attachment 168253 [details]
202_diagram.gif
Comment 4 CDE Administration CLA 2010-05-12 16:21:56 EDT
Created attachment 168254 [details]
202_preview.gif
Comment 5 CDE Administration CLA 2010-05-12 16:22:04 EDT
<cde:tctdetail>
Testcase: 13.000.070 - GMF - CREATE A RECTANGLE
Project: WSW36
Component: Xfer - GMF/Models
Priority: 2
Subject: CHS: Corrupted Charaters
Article ID: 202
Originator: maxl@szchkj.com.cn
</cde:tctdetail>
Comment 6 CDE Administration CLA 2010-05-12 22:51:08 EDT
<response_by> Kathy Li at 2010.05.12.21.39.59 </response_by>
Must-Fix: Yes
Comment 7 CDE Administration CLA 2010-05-13 04:21:03 EDT
<response_by> MiJung Cho at 2010.05.13.03.11.21 </response_by>
Hello,

The same for korean

Thanks
MJ
Comment 8 Anthony Hunter CLA 2010-05-13 10:11:12 EDT
What happens when you try to enter DBCS characters into the text field?

Assuming this is successful, I would suggest we leave the string blank

Text_InitialValue =

This would restore the behavior from Galileo where the default string was empty.
Comment 9 CDE Administration CLA 2010-05-13 11:21:11 EDT
<response_by> Ma Xiao Long at 2010.05.13.10.12.52 </response_by>
I can try your solution in the next build that contains our returned update translation, however, I wonder if there is any better solution to solve the problem?
Comment 10 CDE Administration CLA 2010-05-14 01:05:42 EDT
<response_by> Keiko Terada at 2010.05.14.00.02.28 </response_by>
build: 20100513-1500

Any Japanese character is not accepted in the diagram.
Please see the screenshot, thanks.
Comment 11 CDE Administration CLA 2010-05-14 01:05:44 EDT
Created attachment 168501 [details]
202_ja.gif
Comment 12 CDE Administration CLA 2010-05-14 03:35:45 EDT
<response_by> Ma Xiao Long at 2010.05.14.02.25.27 </response_by>
I tried to input other DBCS characters, but they all truncated.
Comment 13 Anthony Hunter CLA 2010-05-14 11:37:30 EDT
(In reply to comment #12)
> <response_by> Ma Xiao Long at 2010.05.14.02.25.27 </response_by>
> I tried to input other DBCS characters, but they all truncated.

So you cannot enter any DBCS characters into these text fields. Can you confirm this is a regression since IES 3.5 (Galileo)?
Comment 14 CDE Administration CLA 2010-05-14 16:06:15 EDT
<response_by> rewagner at 2010.05.14.15.01.52 </response_by>
I just tried this with a 3.5 release, and the characters still get corrupted if you input Kanji or hiragana inside the text box.  I have tried this one a 3.5 and 3.6 build on both Win XP and Win 7, and the only time I see this error is when I am on Win 7 set to Japanese or Chinese or Korean on their test machines.  I have not tried it on an Asian version of Win XP since I don't have access to one, but we should definitely try that since I believe it is a Win 7 quirk.

Thanks,
Ross
Comment 15 Anthony Hunter CLA 2010-05-18 17:43:05 EDT
So this is not a regression and this problem existed on IES 3.5 (Galileo).

What happens when you enter the same characters in the GEF Logic Example? This is in the GEF TVT tests.

(In reply to comment #14)
>  I have not tried it on an Asian version of Win XP since I don't have access to
> one, but we should definitely try that since I believe it is a Win 7 quirk.

Did we get anywhere with this?
Comment 16 CDE Administration CLA 2010-05-19 17:06:02 EDT
<response_by> rewagner at 2010.05.19.16.00.08 </response_by>
Kit and I tried to work with the configurations and encodings for the Japanese machine, and we found that when you go to the Control Panel, click on Regional and Languages and go to the Administrative tab and set the non-Unicode Programs language from English to the native language, the characters in the diagrams no longer become corrupted.  Kit believes that you can now probably look up the encodings and put in some simple conversions and this will fix the problem.

Thanks,
Ross
Comment 17 CDE Administration CLA 2010-05-22 16:06:43 EDT
<response_by> Ma Xiao Long at 2010.05.21.20.52.04 </response_by>
Hi Ross,

Thank you for your detailed instruction, I changed the Regional and Languages setting in Control Panel, the problem disappeared.  Thank you so much!
Comment 18 CDE Administration CLA 2010-05-22 16:07:04 EDT
[Added by CDE: Completed by the CDE Bridge]
Comment 19 CDE Administration CLA 2010-05-24 01:36:21 EDT
<response_by> Keiko Terada at 2010.05.24.00.32.51 </response_by>
I verified the JA characters shown properly on the test machine. Thanks.
Comment 20 CDE Administration CLA 2010-06-01 03:20:47 EDT
<response_by> MiJung Cho at 2010.06.01.02.19.40 </response_by>
Hello,

It is fixed for korean in 20100529 build as your instruction.

Thanks
MJ
Comment 21 Eclipse Webmaster CLA 2010-07-19 21:59:17 EDT
[GMF Restructure] Bug 319140 : product GMF and component
Runtime was the original product and component for this bug
Comment 22 CDE Administration CLA 2010-07-19 22:24:21 EDT
<response_by> Eric Ma at 2010.05.07.22.42.52 </response_by>








<response_by> Eric Ma at 2010.05.11.21.56.07 </response_by>














<response_by> Eric Ma at 2010.05.13.10.12.52 </response_by>



<response_by> Eric Ma at 2010.05.14.02.25.27 </response_by>










<response_by> Eric Ma at 2010.05.21.20.52.04 </response_by>