The database supplies several character sets whose attributes are also known to CLIv2. The customer may define additional character sets to the database, and these are known as user-defined character sets. The attributes of such character sets must also be defined to CLIv2. This is done using the CLIv2 TRD2XUT utility, which accepts descriptions of one or more user-defined character sets and creates a load module named TRD2XCI that may then be made available to the CLIv2 runtime routines. All user-defined character sets available to the CLIv2 runtime are defined in one TRD2XCI load module. The description of character sets supplied by the database cannot be altered. If this is done, no error is generated, although the standard definition is always used.
The utility can be run under either z/OS or VOS3. Execution as a CICS or IMS transaction is not supported. For z/OS and VOS3, a sample JCL cataloged procedure is distributed as TRD2XUT SAMPz/OS.
The input to the TRD2XUT utility consists of records of fixed or unspanned varying length. If the records are fixed 80 byte records, columns 73 through 80 are reserved for traditional sequence numbers and are ignored; otherwise, the entire record is used. Records containing only blanks are ignored. If the first non-blank characters of a record are /*, the record is considered a comment and ignored.
Statements may be continued onto multiple records using continuation characters. If the last non-blank character is a minus sign, the minus sign is discarded and the statement continues with the first column of the next record. If the last non-blank character is a plus sign, the plus sign is discarded and the statement continues with the first non-blank column of the next record. A statement may be continued onto any number of records. A semicolon may be added to the end of any statement. The semicolon is discarded before the statement is processed.