This keyword sets the CCSID for
alphanumeric, graphic, and UCS-2 definitions. 
number must be an integer between 0 and 65535. It
must be a valid CCSID value.
- A valid alphanumeric CCSID is
65535, or an EBCDIC CCSID with encoding scheme x'1100' or x'1301',
or an ASCII CCSID with encoding scheme
X'2100', X'3100', X'4100', X'4105', X'5100', X'2300', X'3300',
or the UTF-8 CCSID 1208.
- A valid graphic CCSID is 65535
or a CCSID with the EBCDIC double-byte encoding scheme (X'1200').
- A valid
UCS-2 CCSID has the UCS-2 encoding scheme (x'7200').

For program-described fields, the CCSID keyword
overrides the defaults set
on the control specification or the /SET directive with the
CCSID(*CHAR), CCSID(*GRAPH), or CCSID(*UCS2) keyword. 

Some special values are allowed
- *DFT
- CCSID(*DFT) indicates that the current default CCSID for the module is to be used.
This is useful when the LIKE keyword is used since the new field would otherwise
inherit the CCSID of the source field.
See /SET for more information
about the current default CCSID.
- *{NO}EXACT
- CCSID(*EXACT) and CCSID(*NOEXACT) are
valid for externally-described data structures
and data structures defined with the LIKEREC
keyword.
The CCSID keyword for a data structure controls
the CCSID of alphanumeric subfields.
See CCSID(*EXACT | *NOEXACT) for more information.
- *HEX, or 65535
- CCSID(*HEX) is valid for alphanumeric and graphic
definitions. It indicates that the data is not
considered to have a CCSID.
Items defined with CCSID(*HEX) or CCSID(65535)
cannot be used in CCSID conversions.
- *JOBRUN
- CCSID(*JOBRUN) is valid for alphanumeric and graphic
definitions. It indicates that the data is is in
the job CCSID at runtime.
- *JOBRUNMIX
- CCSID(*JOBRUNMIX) is valid for alphanumeric definitions.
It indicates that the CCSID is the mixed-byte CCSID
related to the job CCSID.
- *UTF8
- CCSID(*UTF8) is valid for alphanumeric definitions.
It indicates that the CCSID is 1208, which
is the CCSID for UTF-8 Unicode data.
- *UTF16
- CCSID(*UTF16) is valid for UCS-2 definitions.
It indicates that the CCSID is 1200, which
is the CCSID for UTF-16 Unicode data.


If the keyword is not specified
- If a data type is specified, the current
default CCSID for the module is assumed.
See /SET for more information
about the current default CCSID.
- If the LIKE keyword is specified, the
new field will have the same CCSID as the LIKE field.

Note:
- If this keyword is not specified for a character definition,
and neither CCSID(*EXACT) nor CCSID(*CHAR) is specified on
a control statement, and CCSID(*CHAR) has not been specified
on a /SET statement that is in effect, the CCSID for
the character field is the mixed CCSID related to the job CCSID.
The mixed CCSID is capable of handling both single-byte
character set (SBCS) data and double-byte character set (DBCS) data.
For example, if the job CCSID is the single-byte CCSID 37,
the mixed CCSID is 937.
The characters X'0E' and X'0F' will be interpreted as shift
characters.
This can lead to incorrect results when this data is converted
to another CCSID, if the data happens to contain X'0E' or
X'0F', and the job CCSID is not a DBCS-capable CCSID.
- This keyword is not allowed for graphic definitions when
CCSID(*GRAPH : *IGNORE)
is specified or assumed.
