DB2 Server for VM: System Administration

Appendix E. Defining Your Own Character Set

If you cannot use one of the IBM-supplied sample character sets, you can create your own. (You may be able to use one of the character sets in the Character Data Representation Architecture Level 1, Registry manual. The CCSIDs in this manual are registered, and already defined.) To create a character set:

1. Identify all characters in your set and their hexadecimal values. For more information, see Step 1: Identify All Characters in Your Character Set.
2. Classify each character.

   The database manager has 12 character classifications that it uses to identify how the character can be used in SQL statements. You must classify those characters in your set that differ from the ENGLISH character set table. For more information, see Step 2: Classify the Characters.

3. Determine the hexadecimal values to which lowercase characters are to be translated. For example, suppose X'6A' is used for lowercase n-tilde, and X'7B' is used for uppercase N-tilde. The database manager does not automatically know that X'6A' should be converted to X'7B'. You must define that relationship. For more information, see Step 3: Determine Translation Characters.
4. Update the SYSTEM.SYSCHARSETS catalog table.

   Having defined character classifications and translations, you must implement them on the database manager. You can code an INSERT command for the DBS utility to process. During this process, you must also choose a name for the character set (for example, PORTUGUESE). For more information, see Step 4: Update the SYSTEM.SYSCHARSETS Catalog Table.

5. Update the SYSTEM.SYSCCSIDS catalog table.

   When the new character set is implemented, you must add a row to the SYSTEM.SYSCCSIDS catalog table to identify the CCSID values to be associated with the new character set. For more information, see Step 5: Update the SYSTEM.SYSCCSIDS Catalog Table.

6. Update the SYSTEM.SYSSTRINGS catalog table.

   After you specify the CCSID values, you must indicate the conversion table information to allow conversions to and from the new CCSID. For more information, see Step 6: Update the SYSTEM.SYSSTRINGS Catalog Table.

7. Update the CCSID-Related CMS Files

   You must now use the CCSID option of the ARISDBMA EXEC to load the CMS files to be able to use the new character set. This EXEC copies the information in SYSTEM.SYSCHARSETS, SYSTEM.SYSCCSIDS and SYSTEM.SYSSTRINGS to three CMS files. See the DB2 Server for VM Program Directory for more information on the ARISDBMA EXEC.

When your character set is loaded into the SYSTEM.SYSCHARSETS catalog table, and the SYSTEM.SYSCCSIDS and SYSTEM.SYSSTRINGS catalog tables have been updated, you can specify the character set by using the CHARNAME initialization parameter.

Each step in defining a character set is discussed below. As an example, a character set (PORTUGUESE) for Brazilian Portuguese is defined.