Selection Using the Case Selection Structure

When faced with a situation in which a program must select from many processing alternatives based on the value of a single variable, an analyst must expand his or her use of the basic selection structure beyond the standard two processing branches offered by the if statement to allow for multiple branches. One solution to this is to use an approach called nesting in which one (or both) branch(es) of a selection contain another structure (such as selection). This approach is applied to each branch of an algorithm until enough additional branches have been created to handle each alternative. For example, consider the following problem that involves the use of a menu of actions.

Problem Statement

Display a numbered menu involving the use of the terms "single", "double", "triple", and "quadruple" to help young students understand them. Allow the student (user) to select one of the words by entering the number associated with it on the menu, and then display a word indicating the multiplier associated with the term selected.

Sample Softcopy

The numerals on the left of the sample output below are there for the analyst's reference only and will not appear on the screen. Underlined items indicate values entered by the user.

1
2
3
4
5
6
7
8
9
10

MENU OF TERMS

1. Single
2. Double
3. Triple
4. Quadruple

Enter the number of the term you want explained (1-4)? 2 

That word means to multiply by two.

Variable List

Algorithm (Flowchart) - Using Nested if's

The flowchart below (or the alternative [outlined algorithm]) uses multiple if structures (each with only two legs), nested one within another on the false leg (branch) from each test to allow for enough multiple legs to solve the problem. Notice that the final false (else) leg does not contain another if structure, but rather handles the possibility that a user might enter a menu response that is outside of the range of acceptable values by simply displaying the word "ERROR" in place of an appropriate word. Certainly, a more user-friendly approach could have been used, but it would have increased the complexity of this example.

Desk Check (for the Nested if approach)

Because this program has five paths through its algorithm (as illustrated in the flowchart above), an analyst would typically perform and document five desk checks involving carefully selected test data (menu choice values for N) that would result in testing each of those five paths. The Tracing Chart for only one path (resulting form menu choice 2) is provided below to illustrate the documentation involved in tracking multiple tests performed when using nested if structures. The other documentation should be obvious. Note that some of the columns in this chart have no data because the tests that they track were not performed as a result of the test data value being 2.

C++ Source Code (using nested if's)

/* menu.cpp - an example of nested if selection */
/* Written by: Randolph Gibson - Date: Feb. 18, 2011 */

#include <iostream>
using namespace std;

int main ()
{
  int N; /* Menu Choice */

  cout << "MENU OF TERMS\n\n";

  cout << "1. Single\n";
  cout << "2. Double\n";
  cout << "3. Triple\n";
  cout << "4. Quadruple\n\n";

  cout << "Enter the number of the term you want explained (1-4)? ";
  cin >> N;
  cout << "\nThat word means to multiply by ";

  if (N == 1) cout << "one";
  else if (N == 2) cout << "two";
       else if (N == 3) cout << "three";
            else if (N == 4) cout << "four";
                 else cout << "ERROR";

  cout << ".\n";
  return 0;
}

The use of the indentation in the code above is not necessary, but does improve readability.

Algorithm (Flowchart) - Using Case Selection

Solutions using nested if structures like the one above may work, but can become extremely difficult to flowchart and code when they involve many alternatives (paths). Imagine how involved the documentation would be for a menu-oriented program that offered 20 choices. An alternative approach known as Case Selection can be used to document selective situations like the one in the example above where multiple processing paths are required as the result of the value of a single variable. This approach has restrictions. Case Selection can be used only in situations where the branching is based entirely on the value of a single variable and where those values are part of a set of ordinal data. Understand that the use of the Case Selection structure is never necessary, because the situations it solves can always be handled using nested if statements as shown above. Its value is in making the documentation easier to read.

The flowchart below (or the alternative [outlined algorithm]) shows the same solution to the problem in the example above, but using the Case Selection structure. This document is smaller and easier to follow. It more clearly conveys the nature of the selective process as involving a selection of one path as the result of examining the value of a single variable. If you compare it to the first algorithm showing the nested if approach, you will see the weakness of the nested approach. That algorithm makes it appear that a series of tests is required in some specific order to choose the correct path through the program. This is not really the truth. The algorithm for the Case Selection below makes no such implication. It merely shows that a specific value of N will cause a particular processing path to be taken.

Desk Check (for the Case Selection approach)

Because this program has five paths through its algorithm (as illustrated in the flowchart above), an analyst would typically perform and document five desk checks involving carefully selected test data (menu choice values for N) that would result in testing each of those five paths. As in the earlier example, the Tracing Chart for only one path (resulting form menu choice 2) is provided below to illustrate the documentation involved in tracking multiple tests performed when using a Case Selection structures. Note that this chart requires no columns for tests related to N's value, as that value is already being tracked in the column devoted to that variable. The analyst would know which legs were followed based on the labels atop each leg in the case selection on the flowchart. Thus the Tracing Chart is vastly simplified.

That's it; a big improvement!

C++ Source Code (using the Case Selection approach)

/* menu.cpp - an example of case selection */
/* Written by: Randolph Gibson - Date: Feb. 18, 2011 */

#include <iostream>
using namespace std;

int main ()
{
  int N; /* Menu Choice */

  cout << "MENU OF TERMS\n\n";

  cout << "1. Single\n";
  cout << "2. Double\n";
  cout << "3. Triple\n";
  cout << "4. Quadruple\n\n";

  cout << "Enter the number of the term you want explained (1-4)? ";
  cin >> N;
  cout << "\nThat word means to multiply by ";

  switch (N)
  {
   case 1: cout << "one"; break;
   case 2: cout << "two"; break;
   case 3: cout << "three"; break;
   case 4: cout << "four"; break;
   default: cout << "ERROR"; break;
  }

  cout << ".\n";
  return 0;
}

As you can see by comparing the switch statement with the nested if statements in the previous C++ source code, the code for the case selection is easier to read and more clearly communicates the process being performed. The clarity of the Case Selection is the reason that almost all modern programming languages have syntax to implement this structure. Its only weakness is its use of four separate keywords for implementation.

• switch - starts the structure and defines the single variable being tested
• case - starts a single leg and defines the value or values that select it
• break - ends a leg and forces a branching ahead to the closing brace
• default - starts the default leg that will be followed if an unspecified value is tested

Be aware that the syntax and keywords used to implement the Case Selection structure are often quite different in other programming languages.

Click here for more examples of coding the case selection structure.