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.
Summary: | I18N - Output window doesn't show i18n'ed characters. | ||
---|---|---|---|
Product: | projects | Reporter: | Jun Qian <jqian> |
Component: | Ant | Assignee: | Jesse Glick <jglick> |
Status: | RESOLVED WONTFIX | ||
Severity: | blocker | CC: | jglick, kaa, kfrank, mkleint |
Priority: | P3 | Keywords: | I18N |
Version: | 6.x | ||
Hardware: | All | ||
OS: | All | ||
Issue Type: | DEFECT | Exception Reporter: | |
Attachments: |
Output window I18N Problem
screenshot |
Description
Jun Qian
2007-09-11 02:07:35 UTC
Created attachment 48537 [details]
Output window I18N Problem
might not be output window related. Possibly ant (or some other source of output) doesn't handle i18n'ed chars in this particular case Probably dependent on both project type (~ build script) and operating system. If not reproducible on a Linux system using UTF-8 system encoding, probably need to close as WONTFIX - I have tried and failed to get XP to properly handle non-ASCII characters in stdio. what output is not being shown correctly ? what are the user actions and steps ? ken.frank@sun.com To reproduce: 1. Create a SOA SynchronousSample 2. Deploy SynchronousSampleApplication 3. Rename the default test case to something containing i18n'ed characters 4. Run the test case In the attached screenshot, there are three test cases (TestCase0, ...) in the project three. After running all three test cases, the output window shows "??" instead of the i18n'ed test case names. I just realized that there has never been a requirement to i18n-ize strings defined in NetBeans' ant tasks. I guess we don't support i18n in ant task output? but I think in general its true that what comes from ant is not translated. I think there are some msgs that are seen in ow as result of compiling that are but many are not. (Jesse can comment with more accuracy about this) But for this issue, I am wondering if still the multibyte characters themselves should be shown ok, especially if the same encoding for the characters are that of the project, which in this case it is, so its not a situation that would require encoding detection, at least within the ide. ken.frank@sun.com I tried to print chinese chars directly to output window to a module without a problem. when running a test with System.out in chinese on macosx I get ?? chars for output but correct chars in exception stacktrace. The project was setup to use UTF-8 BTW. This leads me to assume the problem is somewhere in the ant output processing. Created attachment 50162 [details]
screenshot
I've run the test on my macosx machine which has Macroman encoding as default btw. The project was setup to use UTF-8. System.out is a bytestream. There is necessarily a conversion from chars to bytes done when text is printed, which is beyond the control of NB or Ant AFAIK. If you want to use international characters in text I/O, you had better use an OS in which UTF-8 is the universally used encoding, e.g. modern Linux distributions. I can't speak for Mac OS X, but I have tried in the past to get this to work in Windows XP even without an IDE, to no avail. |