Default Value

A DefaultValue in programming languages is typically the value of a FunctionParameter?, ClassMember?, GlobalVariable or LocalVariable? when no other value was explicitly provided. For example, in JavaLanguage, every class member of type "int" takes on the default value of 0, and every class member of class type defaults to null.

Default values aren't the only way to handle this situation:

• Languages may require that an explicit value is provided in some instances, and raise a CompilerError? when it is ommitted. For example, it is not valid to leave away parameters in calls to ClassMethod?s in Java. In C++, this is valid if the first function declaration has provided default values to be used for the last one or more parameters.
• They may also not require an explicit value, and neither provide a default value. For example, in C++, the value of an uninitialized variable of type int is undefined, that is, the CeePlusPlusStandard? does not require an implementation to make any guarantees. So:
• An implementation may use an implementation-defined default value. For example, a certain compiler may initialize such variables with 0.
• An implementation may leave the corresponding memory uninitialized, which means that the int may take on a "random" value at RunTime.
• An implementation may cause an exception to be raised if the uninitialized variable is read. (Though it has to permit assignments to it and subsequent reads after an initial assignment.)

The typical behavior of C++ implementations is to emit a CompileTime warning when FlowAnalysis? detected that an uninitialized variable is read from, and just use uninitialized memory at runtime.

The following code in CeePlusPlus defines a default value for a function parameter. It will be used whenever a call to this function omits the last parameter:

 int add_mod(int a, int b, int modulo = 10) {
 return (a + b) % modulo;
 }

In functions like Java and C# that do not allow default parameters, this behavior can be emulated through the use of FunctionOverloading?:

 static int add_mod(int a, int b, int modulo) {
 return (a + b) % modulo;
 }
 static int add_mod(int a, int b) {
 return (a + b) % 10;
 }

or

 static int add_mod(int a, int b, int modulo) {
 return (a + b) % modulo;
 }
 static int add_mod(int a, int b) {
 return add_mod(a, b, 10);
 }

or

 static final int modulo_base = 10;
 static int add_mod(int a, int b, int modulo) {
 return (a + b) % modulo;
 }
 static int add_mod(int a, int b) {
 return add_mod(a, b, modulo_base);
 }

This is sometimes useful in C++ as well, because this approach is more flexible than providing default values. The disadvantage is its verbosity, especially if multiple default-valued parameters are to be emulated.