This Bugzilla instance is a read-only archive of historic NetBeans bug reports. To report a bug in NetBeans please follow the project's instructions for reporting issues.
Now bug 119307 has been fixed so we can type Japanese on Output window. But it does not work properly. It seems that the caret position is not changing during the composition. 1. Run the small program of bug 84375 2. Turn conversion on Output window 3. Type something I'll put screencast later. The scenario is to type something after "katakai". But after the composition of Japanese, "katakai" is broken and "kai" disappears and replaced with japanese. I don't see this issue on Solaris but it happens on Windows.
Created attachment 51643 [details] demo
confirmed. I'm trying the same with chinese pinyin input method on my mac. (i'm not familiar with japanese input methods) I'm getting somewhat different results but probably caused by the same bug. 1. when entering chinese characters at the end of the line, everything seems fine. 2. I can delete any arbitrary character 3. if I go back to the beginning of the line and type something, the complete line gets broken, deleting existing characters and leaving the english alphabet chars around.
possibly related to #114290
I have exactly the same problem as I reported at http://statistics.netbeans.org/exceptions/detail.do?id=9089 When I tried to input a single Japanese character on Output field by using SCIM running on Ubuntu Linux, I got ClassCastException. I have a java.util.Scanner type object in my program to get some characters from standard input. This problem happened only when I typed a Japanese character, so I could input alphabetical characters and get my program run successfully.
Created attachment 52362 [details] patched output2 module
attached is a patched output2 module that should fix the issue. to test it: 1. download the jar file 2. place it into your netbeans installation in the netbeans/platform8/modules folder (on mac it's Netbeans.app/Resources/Netbeans/platform8/modules folder) 3. start the ide and try typing english and japanese/chinese characters into output window.
I tried patch on my Mac and it works fine for me.
for the record: I asked martin_adamek to try out the chinese pinyin input method on mac. That one replaces characters typed in latin pinyin into chinese chars.
I confirm that the attached module fixes the problem. I tried on Mac OS X witch Chinese keyboard following Milos's instructions: without the patch (build 4256), and was able to reproduce the problem, with the patch, everything works correctly.
It's still not working properly for Japanese IME on Windows.
This patch works on Ubuntu and Leopard. I could input Japanese characters in Output field and get my program run successfully. I put the jar file under "platform7" directory instead of "platform8" since my NetBeans 6.0beta2 doesn't has a direcotry, platform8. I guess NetBeans 6.0 final might have this directory.
yokolet: my mistake, platform7 is correct, platform8 was a typo.
Created attachment 52389 [details] screencast: with patch and console output
commiting the partial fix. Checking in ui/AbstractOutputPane.java; /cvs/core/output2/src/org/netbeans/core/output2/ui/AbstractOutputPane.java,v <-- AbstractOutputPane.java new revision: 1.49; previous revision: 1.48 done Checking in OutputDocument.java; /cvs/core/output2/src/org/netbeans/core/output2/OutputDocument.java,v <-- OutputDocument.java new revision: 1.23; previous revision: 1.22 done
Created attachment 52444 [details] patched output module - second shot
the second patch doesn't seem to work either, I'd like to waive the issue. It has been partially fixed for some input methods. The once that don't work can be workarounded by copy&pasting the text into the OW.
I suppose the waiver is approved, no objections were raised.
can team see if this is one that might be fixed in upcoming patch or nb6.0.1 release (or at least for 6.1) ? ken.frank@sun.com
since many parts of this are now fixed, could they be part of 6.0 patch 2 - being organized now by sustaining ? I've added the required status whiteboard entry here in case it can be. I also have information on the internal process and alias notifications needed to sustaining.
I've put in status whiteboard request for this to be fixed for patch 3 - which will happen in next few weeks - can team look and see if that can happen ? ken.frank@sun.com
Status whiteboard should be release601_fixes_candidate1 (since 6.0.1 patch1 is the next planned patch).
fixing the status whiteboard. The issues planned for 6.0.1 patch1 should be fixed till 1/31. Milosi, is it fixable in this timeframe(tommorow)?
no. i still don't know how to fix.
no objections against the waiver for 6.1 -> waiving it
could team summarize what parts are fixed and what not and then 1. could fixed parts be put into trunk, so can go into upcoming 6.1 patch (it needs to be in trunk first before can be nominated) 2. could details of unfixed yet parts be provided and I'll open a separate issue ? ken.frank@sun.com
1. fixed part is in trunk and was part of 6.0 and 6.1 2. the unfixed part is entering japanese text. I've fixed entering chinese text (at least for the input method I was using on mac, there might be other platforms or input methods that fail still) what I would like to know is the exact setup that doens't work 1. OS platform 2. locale used 3. input method used 4. jdk version used (I think that's irrelevant but who knows)
thanks for the clarification on it. I'll try to find out details of what is not yet working. yokolet, does fix from after Nov 3, which is after your last comment, does it work for you using windows, since that is what originally reported on ? ken.frank@sun.com
Ken. Excuse me. I don't have Windows on which I can test this. At least, I've been able to input Japanese characters on Mac OSX and Ubuntu since Nov. 2.
Yokolet, thanks for confirming about mac and ubuntu that its ok now. ken.frank@sun.com
->t_h
Check please if the issue still exists in trunk. I tried Chinese（Pinyin)/ Japanese input on Win Vista and it seems it works for me. If the issue is still valid please specify steps how to reproduce (input method, letters).
Very sorry for late. It's still reproduce for me on Windpws XP and Vista on trunk. See the flash demos in attachment. 1. Start the program in Japanese locale, using the default IME of Windows 2. Input "katakai" in ASCII into output window 3. Turn conversion ON 4. Type k,a,t,a,k,a,i These strings will be converted into Hiragana, simple Japanese characters katakai[hiragana characters of katakai] 5. Hit space key multiple times it will convert Hiragana to Japanese Kanji characters, it will display several string for displaying a candidate. At this point, the candidate string should be displayed *after* 'katakai' like below, katakai[candidate of Kanji] However, the current result is breaking the existing katakai of ASCII like kata[candidate of Kanji]kai It seems that the current caret position is not staying at the end of 'katakai'.
masaki, thanks a lot for the detailed steps. I was able to reproduce and fix. http://hg.netbeans.org/main/rev/8e80a26e097c
Integrated into 'main-golden', will be available in build *200809301401* on http://bits.netbeans.org/dev/nightly/ (upload may still be in progress) Changeset: http://hg.netbeans.org/main/rev/8e80a26e097c User: Milos Kleint <mkleint@netbeans.org> Log: #119985 fix japanese hiragana input method.
Super! Thanks Milos! I verified your fix in 200809301401 on Windows, Solaris and Mac OS X.
Nice. Thanks, Masaki and Milos!