Use the ERROR function as the default value when you do not want null values to pass into a transformation. For example, you might want to skip a row when the input value of DEPT_NAME is NULL. You could use the following expression as the default value:
ERROR('Error. DEPT is NULL')
When you use the ERROR function as a default value, the Integration Service skips the row with the null value. The Integration Service writes all rows skipped by the ERROR function into the log file. It does not write these rows to the reject file.
DEPT_NAME RETURN VALUE
NULL 'Error. DEPT is NULL' (Row is skipped)
The following log shows where the Integration Service skips the row with the null value:
TE_11019 Port [DEPT_NAME]: Default value is: ERROR(<<Transformation Error>> [error]: Error. DEPT is NULL
... error('Error. DEPT is NULL')
CMN_1053 EXPTRANS: : ERROR: NULL input column DEPT_NAME: Current Input data:
CMN_1053 Input row from SRCTRANS: Rowdata: ( RowType=4 Src Rowid=2 Targ Rowid=2
DEPT_ID (DEPT_ID:Int:): "2"
DEPT_NAME (DEPT_NAME:Char.25:): "NULL"
MANAGER_ID (MANAGER_ID:Int:): "1"