Column names on ResultSet.updateXXX and getXXX methods are handled incorrectly

• Key: DERBY-229
• Type: Bug
• Status: Closed
• Resolution: Fixed
• Priority: Major
• Assignee: Shreyas Kaushik
• Reporter: Daniel John Debrunner
• Component/s: JDBC
• Affects Version/s: 10.0.2.1, 10.1.1.0
• Fix Version/s: 10.1.1.0

Description

Sections 14.2.3 and [14.2.3] of JDBC 3.0 spec say

Column names supplied to getter [updater] methods are case insensitive. If a select list
contains the same column more than once, the first instance of the column will be
returned [updated].

Derby returns or updates the last column in the select list, not the first. With the attached Java class I see

PRE-UPDATE
1,100
POST-UPDATE
1,500
POST-UPDATE getXXX(name)
AB:500,ab500

Comments

[Daniel John Debrunner]

Also present in 10.0.2.1 for getXXX only (since updateXXX not supported)

[Mamta A. Satoor]

The getter [updater] methods with column names (in EmbedResultSet.java) use the protected method findColumnName(String columnName) to determine the column index for the passed column name. And the loop for the lookup starts from the end towards the beginning of the ResultDescription and that seems to be the problem behind the current behavior.
ResultDescription rd = resultDescription;

     // 1 or 0 based? assume 1 (probably wrong)
     for (int i=rd.getColumnCount(); i>=1; i--) {

     String name = rd.getColumnDescriptor(i).getName();
     if (StringUtil.SQLEqualsIgnoreCase(columnName, name)) {
     return i;
     }
     }

[Satheesh Bandaram]

I believe I checked in fix for Shreyas for this bug. Shreyas, can you resolve and close this bug?

[Shreyas Kaushik]

Patched has been submitted which was committed.

[Shreyas Kaushik]

Closing this issue as it was patched