Derby
  1. Derby
  2. DERBY-1914

test lang/wisconsin gives garbage output on zOS.

    Details

    • Type: Bug Bug
    • Status: Closed
    • Priority: Major Major
    • Resolution: Fixed
    • Affects Version/s: 10.2.1.6, 10.2.2.0, 10.3.1.4
    • Fix Version/s: 10.9.2.2, 10.10.1.1
    • Component/s: Test
    • Environment:
    • Urgency:
      Normal

      Description

      When running the test lang/wisconsin.java on zOS the output in the .tmp file shows non-native characters.
      However, this is most likely a test/test harness specific problem, for if one executes the simple statistics gathering commands through ij, there is no problem. Note that other tests do produce readable (native) output.

      It's impossible to tell if there is any actual problem this way.

      1. wisconsinOutputOnZos.zip
        204 kB
        Kathey Marsden
      2. wisconsinOutputAfterAsciiFtp.zip
        204 kB
        Kathey Marsden
      3. wisconsin_diff.txt
        3 kB
        Kathey Marsden
      4. derby-1914_diff.txt
        3 kB
        Kathey Marsden

        Activity

        Hide
        Kathey Marsden added a comment -

        I still see this in release testing for 10.5.2.0 on z/OS with
        java version "1.6.0"
        Java(TM) SE Runtime Environment (build pmz3160sr4-20090219_01(SR4))
        IBM J9 VM (build 2.4, J2RE 1.6.0 IBM J9 2.4 z/OS s390-31 jvmmz3160-2009
        J9VM - 20090215_029883_bHdSMr
        JIT - r9_20090213_2028
        GC - 20090213_AA)
        JCL - 20090218_01

        Show
        Kathey Marsden added a comment - I still see this in release testing for 10.5.2.0 on z/OS with java version "1.6.0" Java(TM) SE Runtime Environment (build pmz3160sr4-20090219_01(SR4)) IBM J9 VM (build 2.4, J2RE 1.6.0 IBM J9 2.4 z/OS s390-31 jvmmz3160-2009 J9VM - 20090215_029883_bHdSMr JIT - r9_20090213_2028 GC - 20090213_AA) JCL - 20090218_01
        Hide
        Kathey Marsden added a comment -

        Attached is the output for wisconsin for this failure. I wanted to see if it was the same thing or something different than DERBY-4641. I think it is indeed different as for this one the input to ij seems to be garbled instead of the output from runtime statistics as the last line if wisconsin.tmp is:
        ERROR 42X02: Lexical error at line 1, column 1. Encountered: "`" (96), after : "".
        ij>

        Show
        Kathey Marsden added a comment - Attached is the output for wisconsin for this failure. I wanted to see if it was the same thing or something different than DERBY-4641 . I think it is indeed different as for this one the input to ij seems to be garbled instead of the output from runtime statistics as the last line if wisconsin.tmp is: ERROR 42X02: Lexical error at line 1, column 1. Encountered: "`" (96), after : "". ij>
        Hide
        Kathey Marsden added a comment -

        wisconsin seems to be a hybrid test. Half java, half script. The reason for this was to add logic to skip compress for J2ME.
        I don't at least at this moment see where we actually do skip it and am having a bit of trouble with weme to try it out.

        The attached preliminary patch fixes the encoding problem in the old harness by having the test read from the resource rather than the file so it has a known encoding. I will look and see if I can pull it into Junit at least as a JavaHarnessTest.

        Show
        Kathey Marsden added a comment - wisconsin seems to be a hybrid test. Half java, half script. The reason for this was to add logic to skip compress for J2ME. I don't at least at this moment see where we actually do skip it and am having a bit of trouble with weme to try it out. The attached preliminary patch fixes the encoding problem in the old harness by having the test read from the resource rather than the file so it has a known encoding. I will look and see if I can pull it into Junit at least as a JavaHarnessTest.
        Hide
        Kathey Marsden added a comment -

        Bringing wisconsin into LangHarnessJavaTest runs now using the resources instead of files but fails because of lack of support for the sed filtering files, e.g. wisconsin_sed.properties.

        I don't really have time right now to convert the whole test Using RuntimeStatisticsParser.
        Is it worthwhile to bring the sed properties processing into CanonTestCase or would it be better to just leave this one in derbyall until someone gets to converting it? There are actually 21 tests with xxx_sed.properties files. At the slow rate we are getting tests converted, it might be a worthwhile incremental move but would bring some of the arcane logic of the harness into our JUnit tests.

        Thoughts?

        Show
        Kathey Marsden added a comment - Bringing wisconsin into LangHarnessJavaTest runs now using the resources instead of files but fails because of lack of support for the sed filtering files, e.g. wisconsin_sed.properties. I don't really have time right now to convert the whole test Using RuntimeStatisticsParser. Is it worthwhile to bring the sed properties processing into CanonTestCase or would it be better to just leave this one in derbyall until someone gets to converting it? There are actually 21 tests with xxx_sed.properties files. At the slow rate we are getting tests converted, it might be a worthwhile incremental move but would bring some of the arcane logic of the harness into our JUnit tests. Thoughts?
        Hide
        Kathey Marsden added a comment -

        actually, I think not. I realize this would mean bringing jakarta into the set of jars required to run JUnit tests, so I think not a good idea.

        Show
        Kathey Marsden added a comment - actually, I think not. I realize this would mean bringing jakarta into the set of jars required to run JUnit tests, so I think not a good idea.
        Hide
        Kathey Marsden added a comment -

        Here is an updated patch for this issue which removes use of BaseTestCase in wisconsin.java.

        Show
        Kathey Marsden added a comment - Here is an updated patch for this issue which removes use of BaseTestCase in wisconsin.java.
        Hide
        Myrna van Lunteren added a comment -

        Patch was committed to trunk with revision 1355569 (so before 10.10 branch) and 10.9 with revision 1373148.

        Show
        Myrna van Lunteren added a comment - Patch was committed to trunk with revision 1355569 (so before 10.10 branch) and 10.9 with revision 1373148.
        Hide
        Myrna van Lunteren added a comment -

        This did not occur with 10.10.1.1., Closing.

        Show
        Myrna van Lunteren added a comment - This did not occur with 10.10.1.1., Closing.

          People

          • Assignee:
            Kathey Marsden
            Reporter:
            Myrna van Lunteren
          • Votes:
            0 Vote for this issue
            Watchers:
            2 Start watching this issue

            Dates

            • Created:
              Updated:
              Resolved:

              Development