7.4 — Switch statement basics

Although it is possible to chain many if-else statements together, this is both difficult to read and inefficient. Consider the following program:

#include <iostream>

void printDigitName(int x)
{
    if (x == 1)
        std::cout << "One";
    else if (x == 2)
        std::cout << "Two";
    else if (x == 3)
        std::cout << "Three";
    else
        std::cout << "Unknown";
}

int main()
{
    printDigitName(2);
    std::cout << '\n';

    return 0;
}

While this example isn’t too complex, x is evaluated up to three times (which is inefficient), and the reader has to be sure that it is x being evaluated each time (not some other variable).

Because testing a variable or expression for equality against a set of different values is common, C++ provides an alternative conditional statement called a switch statement that is specialized for this purpose. Here is the same program as above using a switch:

#include <iostream>

void printDigitName(int x)
{
    switch (x)
    {
        case 1:
            std::cout << "One";
            return;
        case 2:
            std::cout << "Two";
            return;
        case 3:
            std::cout << "Three";
            return;
        default:
            std::cout << "Unknown";
            return;
    }
}

int main()
{
    printDigitName(2);
    std::cout << '\n';

    return 0;
}

The idea behind a switch statement is simple: an expression (sometimes called the condition) is evaluated to produce a value. If the expression’s value is equal to the value after any of the case labels, the statements after the matching case label are executed. If no matching value can be found and a default label exists, the statements after the default label are executed instead.

Compared to the original if statement, the switch statement has the advantage of only evaluating the expression once (making it more efficient), and the switch statement also makes it clearer to the reader that it is the same expression being tested for equality in each case.

Let’s examine each of these concepts in more detail.

Starting a switch

We start a switch statement by using the switch keyword, followed by parentheses with the conditional expression that we would like to evaluate inside. Often the expression is just a single variable, but it can be any valid expression.

The one restriction is that the condition must evaluate to an integral type (see lesson 4.1 -- Introduction to fundamental data types if you need a reminder which fundamental types are considered integral types) or an enumerated type (covered in future lesson 10.2 -- Unscoped enumerations), or be convertible to one. Expressions that evaluate to floating point types, strings, and most other non-integral types may not be used here.

Following the conditional expression, we declare a block. Inside the block, we use labels to define all of the values we want to test for equality. There are two kinds of labels.

Case labels

The first kind of label is the case label, which is declared using the case keyword and followed by a constant expression. The constant expression must either match the type of the condition or must be convertible to that type.

If the value of the conditional expression equals the expression after a case label, execution begins at the first statement after that case label and then continues sequentially.

Here’s an example of the condition matching a case label:

#include <iostream>

void printDigitName(int x)
{
    switch (x) // x is evaluated to produce value 2
    {
        case 1:
            std::cout << "One";
            return;
        case 2: // which matches the case statement here
            std::cout << "Two"; // so execution starts here
            return; // and then we return to the caller
        case 3:
            std::cout << "Three";
            return;
        default:
            std::cout << "Unknown";
            return;
    }
}

int main()
{
    printDigitName(2);
    std::cout << '\n';

    return 0;
}

This code prints:

Two

In the above program, x is evaluated to produce value 2. Because there is a case label with value 2, execution jumps to the statement underneath that matching case label. The program prints Two, and then the return statement is executed, which returns back to the caller.

There is no practical limit to the number of case labels you can have, but all case labels in a switch must be unique. That is, you can not do this:

switch (x)
{
    case 54:
    case 54:  // error: already used value 54!
    case '6': // error: '6' converts to integer value 54, which is already used
}

If the conditional expression does not match any of the case labels, no cases are executed. We’ll show an example of this shortly.

The default label

The second kind of label is the default label (often called the default case), which is declared using the default keyword. If the conditional expression does not match any case label and a default label exists, execution begins at the first statement after the default label.

Here’s an example of the condition matching the default label:

#include <iostream>

void printDigitName(int x)
{
    switch (x) // x is evaluated to produce value 5
    {
        case 1:
            std::cout << "One";
            return;
        case 2:
            std::cout << "Two";
            return;
        case 3:
            std::cout << "Three";
            return;
        default: // which does not match to any case labels
            std::cout << "Unknown"; // so execution starts here
            return; // and then we return to the caller
    }
}

int main()
{
    printDigitName(5);
    std::cout << '\n';

    return 0;
}

This code prints:

Unknown

The default label is optional, and there can only be one default label per switch statement. By convention, the default case is placed last in the switch block.

No matching case label and no default case

If the value of the conditional expression does not match any of the case labels, and no default case has been provided, then no cases inside the switch are executed. Execution continues after the end of the switch block.

#include <iostream>

void printDigitName(int x)
{
    switch (x) // x is evaluated to produce value 5
    {
    case 1:
        std::cout << "One";
        return;
    case 2:
        std::cout << "Two";
        return;
    case 3:
        std::cout << "Three";
        return;
    // no matching case exists and there is no default case
    }

    // so execution continues here
    std::cout << "Hello";
}

int main()
{
    printDigitName(5);
    std::cout << '\n';

    return 0;
}

In the above example, x evalutes to 5, but there is no case label matching 5, nor is there a default case. As a result, no cases execute. Execution continues after the switch block, printing Hello.

Taking a break

In the above examples, we used return statements to stop execution of the statements after our labels. However, this also exits the entire function.

A break statement (declared using the break keyword) tells the compiler that we are done executing statements within the switch, and that execution should continue with the statement after the end of the switch block. This allows us to exit a switch statement without exiting the entire function.

Here’s a slightly modified example rewritten using break instead of return:

#include <iostream>

void printDigitName(int x)
{
    switch (x) // x evaluates to 3
    {
        case 1:
            std::cout << "One";
            break;
        case 2:
            std::cout << "Two";
            break;
        case 3:
            std::cout << "Three"; // execution starts here
            break; // jump to the end of the switch block
        default:
            std::cout << "Unknown";
            break;
    }

    // execution continues here
    std::cout << " Ah-Ah-Ah!";
}

int main()
{
    printDigitName(3);
    std::cout << '\n';

    return 0;
}

The above example prints:

Three Ah-Ah-Ah!

So what happens if you don’t end a set of statements under a label with a break or return? We’ll explore that topic, and others, in the next lesson.