Parameter passing modes
Question: Discuss the advantages and disadvantages of ref
parameters, out parameters and parameters that are returned
as a function result.
ref parameters:
| + |
They can be used to implement transient parameters, i.e. parameters that
are passed to a method, changed there and returned from the method to its caller
again. Without ref parameters one would need both a value parameter
and a function return value.
|
| + |
They can be used to implement methods with multiple transient
parameters in which values can be returned to the caller, whereas functions can
only return a single value.
|
| + |
Actual ref parameters are not copied to their formal parameters but
are passed by reference (i.e. only their address is passed). For large
parameters this is more efficient than copying.
|
| - |
ref parameters can lead to side effects, because a formal
ref parameter is an alias of the corresponding actual parameter.
If the method modifies a formal ref parameter the corresponding
actual parameter changes as well.
|
out parameters:
| + |
They can be used to write methods with multiple return values,
whereas a function can have only a single return value.
|
| + |
Like ref parameters, out parameters are passed by value,
i.e. only their address is passed and not their value. For large
parameters this is more efficient than copying.
|
| + |
The compiler makes sure that a method assigns values to all its out
parameters before it returns.
|
| - |
The use of out parameters can lead to side effects, because a formal
out parameter is an alias of its corresponding actual parameter.
If the method modifies its formal out parameter the actual
parameter changes as well.
|
Function return values:
| + |
There are no side effects. |
| + |
The return value can be ignored by the caller, but this can also
be seen as a disadvantage.
|
| - |
A method cannot return multiple function result values. |
|
