Please review the two attached patches which address:

The first patch attempts to insure that 'UTF-8' encoding is always used when importing data from an LDIF file by replacing the Java commands which utilize the default Locale and thus potentially would override the use of 'UTF-8' when working with LDIF files.

The following ticket has been issued to address further concerns related to this issue:

The second patch specifically addresses the use of the 'tr_TR.UTF-8' issue reported in this ticket by changing the following problematic LDIF attribute names:

The root of the problem that appears to cause this issue is the use of the 'toLowerCase()' method as utilized in the LDAPJDK.  The following bug has been issued which fully describes this issue:

These patches were tested on a 64-bit Fedora 20 machine by doing the following: