Uploaded image for project: 'JDK'
  1. JDK
  2. JDK-4473954

Class.newInscance can't handle some specific Japanese named classes or pack

XMLWordPrintable


      ******Testing enviroment***************************************************
      Merlin build: Merlin beta-refresh b69
      OS : Solaris8 Sparc/X86 ja locale,
                    RedHatLinux6.2Jupdate
                    WindowsME, Win2K
      ***************************************************************************


      Class.newInscance can't handle some specific classes or packages for which name some specific characters are used, therefore NoClassDefFoundError occurs for the classes.


      To reproduce,

      1. Extract an attached file newInstance_Test.tar.
         The included files are listed in "newInstance_Test_list.txt" (attached).

         *Some java files and a directory named in Japanese are included in it.
         The directory has some Japanese classes and java files(SJIS encoding) for the Japanese classes.

      2. Compile the java files below;
         PackageTest.java (EUC encoding)
         URLClassLoaderTest.java (EUC encoding)
         MyClassLoader.java
         Status.java
         Test.java
       
      3. Excute PackageTest ---- java PackageTest
      You'll see NoClassDefFoundError occurs.

      I attached the Exception log for "PackageTest_log.txt" which is included in newInstance_Test.tar because Japanese character can not be displayed properly here.
      Please refer the file to see the Exception.

      4. Excute URLClassLoaderTest ----
      java URLClassLoaderTest http://javasoft.japan.sun.com/i18n-test/URLClassLoaderTest_data/

      You'll see NoClassDefFoundError for the same character as above occurs here also.
      I attached the Exception log for "URLClassLoaderTest_log.txt" which is included in newInstance_Test.tar
      Please refer the file to see the problem.

      It seems that newInstance can't recognaize the character properly and recognize as other Japanese character.

      ***I attached "memo.txt" in which I described how the characrter are recognized incorrectly (EUCJP encoding).***


      This is not reproducible with the previous build, b68.






            iris Iris Clark
            mmtokunasunw Mtokunaga Mtokunaga (Inactive)
            Votes:
            0 Vote for this issue
            Watchers:
            0 Start watching this issue

              Created:
              Updated:
              Resolved:
              Imported:
              Indexed: