Suppose you wish to write a song with stanzas and chorus. You can write it two ways. At the end of each stanza write the chorus. This means you will have to repeat the chorus several times. Or you can write the chorus at the bottom after all the stanzas are written. You just need to indicate when to sing the chorus. I told my children my need to illustrate this concept. They came up with a song which is used in Program 6-1. I don't know if it meets all the criteria for a song, but it certainly gets my point across.

I called the program Stanza and the procedure Chorus. The chorus is declared before it is called from the main program. Every time the chorus needs to be written, we just call the chorus. This program writes two lines of chorus after each two lines of stanza and then leaves a blank line. I may not have impressed you very much with this program since the procedure is only called four times; you could have just as easily typed it four times instead. But consider that a program may call a procedure thousands of times thereby saving tens of thousands of lines of codes.

A procedure is a subprogram within a program. As you will see a procedure has all the components of a program. In Pascal all subprograms must be declared before the beginning of the main program. After its declaration, a procedure may be called many times from within a program; it is called by its name. When the call is made the subprogram becomes active and completes its assigned task. Then the execution resumes with the next statement in calling program.

MORE ABOUT PROCEDURES

Procedures allow us to modularize programs. A good rule of thumb should be that a module (procedure) should not be any longer than a screen full (20 to 25 lines). This allows for easy reading and debugging. Program 6-2 converts temperature from one unit to the other. The program is modularized in that the Fahrenheit to Celsius conversion is done by one procedure and the Celsius to Fahrenheit conversion is done by another procedure.

This program will ask you whether you want to convert Fahrenheit or Celsius. It will then ask for the temperature. This temperature is then passed to the appropriate procedure to be converted. The result is printed from within the procedure.

Whatever you pass to a procedure is known as arguments. Arguments or parameters are enclosed in parentheses. In this example the temperature is the argument. You could pass, none, one or more arguments to a procedure. Program 6-1 did not pass any arguments to the procedure. In this temperature conversion program one argument is passed to the procedure.

Let me illustrate how a procedure might be applied to everyday life. In this example the procedure is behaving like a toll collector at the tollgate. You pass a certain amount of money to the toll collector, and he opens the gate for you. In this example the money would be the argument. An argument is also known as a parameter.

PROGRAM 6-2

EXPLANATION OF THE PROGRAM

We will be using only two variables in the main program, temperature and whichOne. The variable "temperature" will be used to store your temperature input. Variable "whichOne" will be used to store a character of your choice, F or C. We will be using two procedures as well, ConvertToCelsius and ConvertToFahren.

Let us now look at the main program block. The first few write and writeln statements are used to print an introduction to the program on the screen, and to suggest to enter either C or F for the next input. The first readln statement accepts a character input from the keyboard. This character input is stored in the variable called "whichOne". The next readln statement accepts an integer entered at the keyboard, and this input is stored in the variable called "temperature".

Once the input is completed, the program makes certain decisions using the CASE statement. If variable "whichOne" had a value of 'C', then the program calls ConvertToFahren procedure. If it had a value of 'F', then the program calls ConvertToCelsius. If the input it received was not either 'C' or 'F' then the program will print out an error message, "You did not enter C or F as instructed!".

There are two procedure calls in this program. They are:

As you can see a procedure call is very simple; just call the procedure by its name. If you have anything to pass to the procedure, enclose it in the parentheses. What is in the parentheses in a procedure call is known as the actual parameter list. How does Pascal know that these are procedures? You declared them as procedures.

Next, let us look at a procedure closely. A procedure has a heading like this one:

'Procedure' is a reserved word in Pascal. 'ConvertToCelsius' is the name (identifier) I chose to call this procedure. 'Fahrenheit' is the variable it will use to receive arguments from the procedure call. This variable is declared to be real. You need to pay special attention to this. The procedure call passes the actual parameter 'temperature' and the procedure receives it into 'Fahrenheit'. Please note that 'temperature' also was declared to be real.

The procedure can have its own local variables. These variables will be used exclusively by this procedure, not by the main program. There is one local variable declared, celsius. Now we have everything needed to do the conversion. Celsius is derived from Fahrenheit which was passed to the procedure. The procedure displays the results on the screen. Here is a block diagram that shows the relationship between the main program and the procedures.

Look at the above diagram. The program convert has access to temperature, whichOne, ConvertToCelcius and ConvertToFahren. Whatever is placed in temperature will be copied into Fahrenheit if ConvertToCelsius is called or into celsius if ConvertToFahren is called. The call will depend upon the value of 'whichOne'.

APPENDIX 6A

Let us look at the process of writing a procedure. A procedure is constructed just like a program. Instead of the reserved word PROGRAM, we use the reserved word PROCEDURE. A name is given to the procedure (just as the program name), then the parameters passed back and forth are listed in parentheses. Recall that in a program inputs and outputs are listed in the parentheses as well. A procedure has all the other components of a program as well: type, constants, variables, and the program block.

In Pascal, a procedure should be defined first before it can be called. (There is a way around it, but that is for later). Therefore, the main part of the program is written at the bottom of the total program. Calling a procedure is easy, just use the name of the procedure and then in parenthesis list all parameters passed to that procedure.

Parameters listed in the procedure call are the actual parameters. The parameters listed in the procedure heading are referred to as the formal parameters. These names (identifiers) do not have to be same as the actual parameters. They just need to match as far as the number of parameters and the type. The first parameter in the actual parameter list will pass its value to the first parameter in the formal parameter list. The second parameter in the actual parameter list will pass its value to the second parameter in the formal parameter list, and so on. In other words, values are passed by their position rather than their names. When passing variables we must make sure that the receiving variable is the same type as the value it is receiving. For example, you cannot pass a real value to an integer variable. You cannot pass a character to a number nor a number to a character.

ProcedureThree is global to no other procedures.

ProcedureOne is global to ProcedureThree.

ProcedureTwo is global to no other procedures.

Main_Program is global to ProcedureThree, ProcedureOne and

ProcedureTwo.

Main_Program can access:

A, B, C which are declared in the Main_Program.

ProcedureOne and ProcedureTwo.

Main_Program cannot access:

Any of the variables declared locally in the procedures,

i.e. A, B, X, Y, R.

ProcedureOne can access:

A, B, X which are declared in ProcedureOne.

ProcedureThree.

C as global which is declared in the Main_Program.

ProcedureOne cannot access:

Any of the variables that are declared locally in

ProcedureTwo or in ProcedureThree.

i.e. A, Z, X, Y, R.

ProcedureTwo can access:

A, X, Y, R which are declared locally in ProcedureTwo.

B, C as global which are declared in the Main_Program.

ProcedureTwo cannot access:

Any of the variables that are declared locally in

ProcedureOne or in ProcedureThree.

i.e. A, B, X, Z

ProcedureThree can access:

A, Z which are declared locally in ProcedureThree.

B, X as global which are declared in ProcedureOne.

C as global which is declared in the Main_Program.

ProcedureThree cannot access:

Any of the variables that are declared in ProcedureTwo.

Any of the variables in ProcedureOne or in Main_Program

if the same variable name is used locally in

ProcedureThree.

Well, is there an easier way to figure out what can be accessed? Yes. Take for example ProcedureThree. Obviously it can access all the local variables declared within. Because ProcedureThree is nested within ProcedureOne, it can also access any variable declared in ProcedureOne as long the same variable name was not already used as a local variable. Furthermore, since ProcedureOne is nested within the Main_Program, it can also access any variable declared in the Main_Program as long as the same variable name was not already used as a local variable in ProcedureThree or in ProcedureOne. In summary, a procedure can access all its local variables and any global variable if the same identifier was not used within the local declaration.

ASSIGNMENTS FOR CHAPTER 6

1. Refer to Program 6-3. What would be the output of this program.
2. Program 6-3 has several comments with blank spaces. In these blank spaces indicate local
    or global.

PROGRAM 6-3

Program continued..