Character conversion

For character conversion, use the ADD (or REPLACE or DELETE) FROM(xxxxx) TO(yyyyy) statement. Duplicate statements are ignored. It is possible that the z/OS Unicode Services uses more than one table to reflect the CONVERSION statement. This might be because an MBCS CCSID is involved or a particular conversion table needed was not found. In the case of MBCS involvement, the system implements a composite conversion with a set of sub-level conversions according to its knowledge base. In the case of missing conversion tables, an indirect conversion – using CCSID 1200 as the intermediate CCSID – is generated.

In general, a direct conversion is supported when:
  • Converting between any combination of SBCS and DBCS
  • Converting between MBCS and DBCS
  • Converting between UTF-8 and UCS-2.
All other conversions will always be indirect conversions.
The parameters that can be specified for character conversion are:
FROM-CCSID
Specifies the FROM-CCSID of the requested conversion. This is the CCSID you are converting from.
TO-CCSID
Specifies the TO-CCSID of the requested conversion. This is the CCSID you are converting to.
TECHNIQUE
Specifies the technique to be used in the conversion.
Possible values for technique-character are:
  • R: Roundtrip
  • E: Enforced Subset
  • C: Customized Subset
  • L: Language Environment Behavior
  • M: Modified Language Environment® Behavior
  • 0 – 9: User-defined conversions
Some special considerations for CCSID 1200: If CCSID 1200 is specified, the CCSID of the most recent UTF-16 version is substituted and all technique-characters are tested. Then the second recent UTF-16 version is substituted and so on. The supported UTF-16 CCSIDs are:
  • 54448 (Unicode 9.0)
  • 42160 (Unicode 6.0)
  • 21680 (Unicode 4.0)
  • 17584 (Unicode 3.0)
  • 13488 (Unicode 2.0)