Writing a procedure can be fun, as I showed you in my article "Let's Build a Procedure." But getting started can be confusing and a little scary. In this short article, I'll show you how to add a procedure to an existing RPG IV program.

First, even though RPG IV's implementation of an ILE procedure is called a "subprocedure," subprocedures are not similar to subroutines. Subroutines have limitations but are faster to call. They are faster to call because there is no call; there's simply a GOTO (under the covers), and the subroutine is entered. A subprocedure, on the other hand, is much more powerful but can be as simple as a subroutine.

Let's consider a very easy subroutine, one line that simply converts a character variable to uppercase. I'm calling it MAKEUPPER.

The MAKEUPPER subroutine is called after a user enters search criteria for a customer name. The customer file is read, and the customer's name is converted to uppercase and then compared to the data the user entered. We use a case-insensitive compare to simplify the search.

Here's an example of that routine:

All RPG programmers have used subroutines, and the subroutine illustrated above should be very easy to understand.

To convert this subroutine to a subprocedure is very easy. You simply create the procedure around the existing subroutine.

So this...

...turns into this:

In this example, the procedure named MAKEUPPER is declared. The first and last lines of the example are P-specs (Procedure specifications). These are similar to the BEGSR and ENDSR opcodes; they isolate the procedure.

Unlike subroutines, subprocedures can have parameters. This makes them more like a program than a subroutine. Our MAKEUPPER subprocedure has two parameters: INPUTVAR and OUTPUTVAR. The attributes of parameters are specified in much the same way data structure subfields are specified. In our example, I use the LIKE keyword to define the parameters the same way I define the customer name field (lines 3 and 4).

As with programs, the values of subprocedure parameters may be modified, and those modified values are sent back to the caller. In fact, as with programs, when you change a parameter in a subprocedure, you are actually changing the memory assigned to the variable you passed in. But you have a few more advanced options when passing parameters to subprocedures. For example, you can declare a parameter in such as way that it cannot be changed by the subprocedure. You can also declare it so that the parameter can be passed with a different size or length than that of the parameter definition.

Since subprocedures are similar to little programs, they can have their own fields, known as "local variables." These variables exist only while the subprocedure's Calc specs are being performed. When the procedure returns to the caller, the variables go away. Again, this is similar to setting on LR and returning from a program.

In our example, no local variables are declared; however, some local named constants (lines 5 and 6) are declared. These local values can be used only from within this subprocedure.

The Calc specs are fairly straightforward: Perform the XLATE opcode and return to the caller. The RETURN opcode is optional in this situation, but it is good style to always include it.

Our original example modified to call a subprocedure would look like this:

One Final Piece

For many RPG programmers, one of the barriers to using subprocedures is something called "prototyping." Prototyping is required and is simply the subprocedure equivalent to a parameter list. Unlike program-to-program calls, which allow an ad hoc parameter list to be specified immediately following the CALL opcode, subprocedures require a named parameter list. This named parameter list is referred to as a prototype. In fact, even if the subprocedure does not have parameters, a prototype is necessary.

In most situations, the prototype is identical to the entry parameter list of the subprocedure. In our example, the entry parameter list is as follows:

To create a prototype for this entry parameter list, all you have to do is copy these three statements and change the letters "PI" (which stand for procedure interface) to the letters "PR" (which stand for prototype). Here's what it will look like:

This three-line prototype needs to be copied up to the D-specs of your program/source member--not the ones inside a subprocedure, but the main D-specs up at the top of the source member. These D-specs are referred to as "global variables or global declarations." Prototypes need to be global in order to use them in the mainline Calcs of your programs. Prototypes also cause no additional overhead, nor do they take up space in the compiled objects. They are there only to help the compiler determine if you specified the proper syntax (that is, the correct parameter list) for the procedure you are calling.

So that's it. You can now convert a subroutine to a subprocedure with or without parameters.

Bob Cozzi is a programmer/consultant, writer/author, and software developer of the RPG xTools, a popular add-on subprocedure library for RPG IV. His book The Modern RPG Language has been the most widely used RPG programming book for nearly two decades. He, along with others, speaks at and runs the highly-popular RPG World conference for RPG programmers.