Defining Parameter Data

The name of the parameter-data-area that contains data elements which are used as parameters in a subprogram, external subroutine or dialog.

Instead of defining a parameter data area, parameter data can also be defined directly within a program or routine.

See Direct Parameter Data Definition below.

The Natural reserved word END-DEFINE must be used to end the DEFINE DATA statement.

Level number is a 1- or 2-digit number in the range from 01 to 99 (the leading zero is optional) used in conjunction with field grouping. Fields assigned a level number of 02 or greater are considered to be a part of the immediately preceding group which has been assigned a lower level number.

The definition of a group enables reference to a series of fields (may also be only 1 field) by using the group name. With certain statements (CALL, CALLNAT, RESET, WRITE, etc.), you may specify the group name as a shortcut to reference the fields contained in the group.

A group may consist of other groups. When assigning the level numbers for a group, no level numbers may be skipped.

The name of a group. The name must adhere to the rules for defining a Natural variable name.

See also the following sections:

• Naming Conventions for User-Defined Variables in Using Natural.

• Qualifying Data Structures in the Programming Guide.

With an array-definition, you define the lower and upper bounds of dimensions in an array-definition.

For further information, see Array Dimension Definition and Variable Arrays in a Parameter Data Area.

A redefinition may be used to redefine a group or a single field/variable (that is a scalar or an array). See Redefinition.

Note:
In a parameter-data-definition, a redefinition of groups is only permitted within a REDEFINE block.

The name to be assigned to the variable. Rules for Natural variable names apply. For information on naming conventions for user-defined variables.

For further information, see Naming Conventions for User-Defined Variables in Using Natural.

The format and length of the field. For information on format/length definition of user-defined variables.

For further information, see Format and Length of User-Defined Variables in the Programming Guide.

Alphanumeric (A), Unicode (U) or binary (B) for dynamic variable.

A parameter may be defined as DYNAMIC. For further information on processing dynamic variables, see Introduction to Dynamic Variables and Fields in the Programming Guide.

Depending on whether call-by-reference, call-by-value or call-by-value-result is used, the appropriate transfer mechanism is applicable. For further information, see the CALLNAT statement.

Call-by-reference is active by default when you omit the BY VALUE keywords. In this case, a parameter is passed to a subprogram/subroutine by reference (that is, via its address); therefore a field specified as parameter in a CALLNAT/PERFORM statement must have the same format/length as the corresponding field in the invoked subprogram/subroutine.

When you specify BY VALUE, a parameter is passed to a subprogram/subroutine by value; that is, the actual parameter value (instead of its address) is passed. Consequently, the field in the subprogram/subroutine need not have the same format/length as the CALLNAT/PERFORM parameter. The formats/lengths must only be data transfer compatible. For data transfer compatibility, the Rules for Arithmetic Assignment and Data Transfer apply (see Programming Guide).

BY VALUE allows you, for example, to increase the length of a field in a subprogram/subroutine (if this should become necessary due to an enhancement of the subprogram/subroutine) without having to adjust any of the objects that invoke the subprogram/subroutine.

* Program
DEFINE DATA LOCAL
1 #FIELDA (P5)
...
END-DEFINE
...
CALLNAT 'SUBR01' #FIELDA
...

* Subroutine SUBR01
DEFINE DATA PARAMETER
1 #FIELDB (P9) BY VALUE
END-DEFINE
...

While BY VALUE applies to a parameter passed to a subprogram/subroutine, BY VALUE RESULT causes the parameter to be passed by value in both directions; that is, the actual parameter value is passed from the invoking object to the subprogram/subroutine and, on return to the invoking object, the actual parameter value is passed from the subprogram/subroutine back to the invoking object.

With BY VALUE RESULT, the formats/lengths of the fields concerned must be data transfer compatible in both directions.

For a parameter defined without OPTIONAL (default), a value must be passed from the invoking object.

For a parameter defined with OPTIONAL, a value can, but need not be passed from the invoking object to this parameter.

In the invoking object, the notation nX is used to indicate parameters which are skipped, that is, for which no values are passed.

With the SPECIFIED option you can find out at run time whether an optional parameter has been defined or not.

The name to be assigned to the handle; the naming conventions for user-defined variables apply.

For further information, see Naming Conventions for User-Defined Variables in Using Natural.

Is used in conjunction with NaturalX as described in the section NaturalX of the Programming Guide.

With an array-definition, you define the lower and upper bounds of dimensions in an array-definition.

For further information, sse Array Dimension Definition.