On 22 Jun 2013 09:50, Booth Martin wrote:
<<SNIP>> I am only really interested in the values being returned.
<<SNIP>>
Others already responded with the changes required for resolution,
both with parameters and an example using an actual /return value/
instead. Hopefully the following will help to further clarify, based on
my presumption of the origin for the apparent confusion...
The parameters are passed by address and are capable of both input
and output [e.g. referred to as an "INOUT" parameter in SQL], such that
upon /return/ from the called-program, a new\changed value for the
variable that was passed as a parameter, will be visible to the caller.
These values are not /being returned/ as effected by a RETURN
statement [the RETURN operation code]. The values are instead, merely
_assigned_ in the called program [i.e. the resolution, as described in
some other replies].
http://pic.dhe.ibm.com/infocenter/iseries/v7r1m0/topic/rzasd/sc092508997.htm#idx3288
http://pic.dhe.ibm.com/infocenter/iseries/v7r1m0/topic/rzasd/sc092508997.htm#zzretrn
IBM i 7.1 Information Center -> Programming -> Programming languages ->
RPG -> ILE RPG Language Reference -> Operations, Expressions, and
Functions -> Chapter 22. Operation Codes
_i RETURN (Return to Caller) i_
"...
The RETURN operation causes a return to the caller. If a value is
returned to the caller, the return value is specified in the expression
operand.
The actions which occur as a result of the RETURN operation differ
depending on whether the operation is in a cycle-main procedure or
subprocedure. ...
..."
Unfortunately the example code snippets for "RetFld" and "RetArr",
presented later in the above doc reference, show three different but
consecutively coded RETURN requests. However only one RETURN operation
would be coded in an actual source, because any RETURN operation that
was coded beyond the first of multiple consecutive unconditional RETURN
operations, would never execute; i.e. the first would effect a "return
to the caller".
In conjunction with the example provided by Paul, the following
documentation hopefully clarifies the distinction of the use and meaning
of a /return value/ as contrasted with the use of a parameter as an
input\output-capable means of obtaining a new\changed value from the
invoked program\procedure.
http://pic.dhe.ibm.com/infocenter/iseries/v7r1m0/topic/rzasd/sc09250869.htm#idx116
IBM i 7.1 Information Center -> Programming -> Programming languages ->
RPG -> ILE RPG Language Reference -> RPG IV Concepts -> Chapter 3.
Procedures and the Program Logic Cycle -> Subprocedure Definition
_i Return Values i_
"A procedure that returns a value is essentially a user-defined
function, similar to a built-in function. To define a return value for a
subprocedure, you must
1. Define the return value on both the prototype and
procedure-interface definitions of the subprocedure.
2. Code a RETURN operation with an expression in the extended-factor
2 field that contains the value to be returned.
You define the length and the type of the return value on the
procedure-interface specification (the definition specification with PI
in positions 24-25). ...
..."
midrange.com
