Issue Details (XML | Word | Printable)

Key: DERBY-470
Type: Bug Bug
Status: Closed Closed
Resolution: Fixed
Priority: Major Major
Assignee: Deepa Remesh
Reporter: Deepa Remesh
Votes: 0
Watchers: 0
Operations

If you were logged in you would be able to see more operations.
Derby

Exception when using LOCALIZEDDISPLAY with JSR169

Created: 26/Jul/05 03:55 AM   Updated: 30/Sep/05 02:25 AM
Return to search
Component/s: Localization
Affects Version/s: None
Fix Version/s: 10.1.2.1, 10.2.1.6

Time Tracking:
Issue & Sub-Tasks
Issue Only
Not Specified

File Attachments:
  Size
File Licensed for inclusion in ASF works derby-470.diff 2005-07-28 02:23 AM Deepa Remesh 2 kB
File Licensed for inclusion in ASF works derby-470.status 2005-07-28 02:23 AM Deepa Remesh 0.2 kB
Environment: IBM's J2ME/CDC/Foundation Profile and JSR169

Resolution Date: 30/Sep/05 02:25 AM

Sub-Tasks  All   Open   
No sub-tasks match this view.

 Description  « Hide
As I did not get any response to my mail, I am opening a JIRA issue for this.

When I run the test i18n/I18NImportExport.java with JSR169 using IBM's J2ME/CDC/Foundation Profile, I get the following exception:
"JAVA ERROR: java.lang.NoSuchMethodError: java/sql/ResultSet.getBigDecimal(II)Ljava/math/BigDecimal;"

In the test, LOCALIZEDDISPLAY is on and the table has a DECIMAL column. This causes the exception. On looking at the code, following is done to format DECIMAL types in LocalizedResource:

if (type == Types.NUMERIC || type == Types.DECIMAL) {
      return getNumberAsString(rs.getBigDecimal(columnNumber, rsm.getScale(columnNumber)));
}

where getNumberAsString method does the following:

public String getNumberAsString(Object o){
      if (enableLocalized){
            return formatNumber.format(o, new StringBuffer(), new FieldPosition(0)).toString();
      }
      else {
            return o.toString();
      }
 }

As a solution, I am planning to do the following:
1. Check for BigDecimal class in the JVM by doing a Class.forName
2. Call getNumberAsString for NUMERIC and DECIMAL types only if BigDecimal class is available. Otherwise, return rs.getString.

This solution would mean that NUMERIC and DECIMAL types will not be localized in these JVM environments. Please let me know if any comments on this.

Thanks,
Deepa



 All   Comments   Work Log   Change History   Subversion Commits      Sort Order: Ascending order - Click to sort in descending order
Deepa Remesh added a comment - 28/Jul/05 02:23 AM
I have added the following in LocalizedResource.java to check if BigDecimal is available in the JVM.

private static boolean HAVE_BIG_DECIMAL;

{
boolean haveBigDecimal;
try {
Class.forName("java.math.BigDecimal");
haveBigDecimal = true;
} catch (Throwable t) {
haveBigDecimal = false;
}
HAVE_BIG_DECIMAL = haveBigDecimal;
}

The above code is also there in org.apache.derby.impl.tools.ij.util.java. Will it be okay to make HAVE_BIG_DECIMAL in util.java public and re-use it? This would be good if there are more scenarios which may need a similar check. Please let me know and I can submit a new patch, if needed.

With this patch, I have run derbyall with Sun JDK 1.4.2 and IBM's WCTME 5.7 J2ME/CDC/FP. In J2ME, one testcase failed. The failure is not related to this change.

Kathey Marsden added a comment - 25/Aug/05 03:14 AM
Committed revision 239718 to fix this issue. You can submit centralization of HAVE_BIG_DECIMAL as another patch or with other JS169 changes.

It seems to me that for sharing the function for multiple tools it would be better to put the internal api in a shared utility class under

org.apache.derby.iapi.tools

rather than making the function in ij.util public.

Perhaps the fact that NUMERIC and DECIMAL types will not be localized in these JVM environments should be mentioned in the documentation.

Deepa Remesh added a comment - 26/Aug/05 09:04 AM
I think this patch can be applied to 10.1 branch also.

I am opening a sub-task to update documentation.

Kathey Marsden added a comment - 29/Aug/05 10:28 PM
Committed revision 264135 to the 10.1 branch

Deepa Remesh added a comment - 02/Sep/05 10:02 AM
This patch was committed as revision 239718 to trunk and revision 264135 to the 10.1 branch

Kathey Marsden added a comment - 16/Sep/05 01:40 AM
just reopening to fix up fix version

Deepa Remesh added a comment - 30/Sep/05 02:25 AM
reopening to fix fix version