Functions: Some Basics

When designing and writing a program of non-trivial size and complexity, it is customary to break up the code for that program into bite-size chunks called functions (which, in some languages, are called subroutines or procedures). A typical program is made up of a main function (in which the program begins execution) with calls to one or more user-written and/or library functions. For example, the main function in an application that manages an employee-records systems might present a menu of choices that includes Add a new employee, Remove an existing employee, and Change an existing employee's data, with each such action being handled by a separate function or group of functions. It is also common to put general-purpose algorithms and operations into their own functions. Examples include string-manipulation and math functions.

When a function is called, information may be passed to it by the caller via an argument list, which contains one or more argument expressions, or more simply, arguments. These correspond by position to the parameters in a parameter list in the called function's definition. For example:

function main(): void {
  $pA = create_point(3.5, -6.2);
  $pB = create_point(-2.4, 3.6);

  echo "Distance between the Points is ".distance($pA, $pB)."\n";

function create_point(float $x, float $y): (float, float) {
  return tuple($x, $y);

function distance((float, float) $p1, (float, float) $p2): float {
  $dx = $p1[0] - $p2[0];
  $dy = $p1[1] - $p2[1];
  return \sqrt($dx * $dx + $dy * $dy);
Distance between the Points is 11.438968484964

Function main contains two calls to function create_point. Let's consider the first. Calling a function requires the use of the function-call operator, (). Typically, this operator is preceded by the name of the function being called, in this case, create_point. The comma-separated list of arguments goes inside the parentheses, as in create_point(3.5, -6.2).

The definition of function create_point shows that it has two parameters, $x and $y, each having type float. When that function is called, the first parameter, $x, takes on the value of the first argument, 3.5, and the second parameter, $y, takes on the value of the second argument, -6.2. (The names of the parameters are local to the function definition.) A similar situation exists in the call to function distance except that two float-pair tuple arguments are passed and expected. Function distance, in turn, calls the math library function sqrt passing it a float value.

A function can optionally return one value only, using a return statement; however, as we see with create_point, we return a tuple value, which can contain an arbitrary number of values. The library function sqrt returns a float, which is, in turn, returned by distance. Note that main's return type is declared to be void, which means that function does not return a value.

The final line calls main.

Note carefully, that while tuple($x, $y) looks like a function call, it's actually a tuple literal.

When a function is declared inside a class, interface, or trait, it is called a method; however, the same basic principles are involved as with a top-level function. That is, methods are called using the function-call operator, they take zero or more arguments, and they return zero or one value. For example:

class Point {
  private float $x; // instance property
  private float $y; // instance property

  public function __construct(num $x = 0, num $y = 0) { // instance method
    $this->x = (float)$x; // access instance property
    $this->y = (float)$y; // access instance property

  public function move(num $x = 0, num $y = 0): void { // instance method
    $this->x = (float)$x;
    $this->y = (float)$y;
  // ...

function main(): void {
  $p = new Point();
  $p->move(-2.2, -4);

Here, method move has two parameters of type num, so it can accept ints, floats, or a combination thereof. It returns no value. It is called to move a Point (such as $p in main) to a given location using an expression like $p->move(-2.2, -4), which involves the member-selection operator, ->.

A function can be parameterized; that is, its definition can have one or more placeholder names—called type parameters---that are associated with types via type arguments when that function is called. A function having such placeholder names is called a generic function. For more information, see generic types and functions.

Note, Hack does not support function overloading; that is, multiple functions having the same name provided each has a different signature (that is, a different order and set of parameter types), are not permitted.

Asynchronous functions are discussed in asynchronous operations.