In computer science, a mutator method is a method used to control changes to a variable.
The mutator method, sometimes called a "setter", is most often used in object-oriented programming, in keeping with the principle of encapsulation. According to this principle, member variables of a class are made private to hide and protect them from other code, and can only be modified by a public member function (the mutator method), which takes the desired new value as a parameter, optionally validates it, and modifies the private member variable.
Often a "setter" is accompanied by a "getter" (also known as an accessor), which returns the value of the private member variable.
Mutator methods may also be used in non-object-oriented environments. In this case, a reference to the variable to be modified is passed to the mutator, along with the new value. In this scenario, the compiler cannot restrict code from bypassing the mutator method and changing the variable directly. The onus falls to the developers to ensure the variable is only modified through the mutator method and not modified directly.
In programming languages that support them, properties offer a convenient alternative without giving up the utility of encapsulation.
In the examples below, a fully implemented mutator method can also validate the input data or take further action such as triggering an event.
Contents |
The alternative to defining mutator and accessor methods, or property blocks, is to give the instance variable some visibility other than private and access it directly from outside the objects. Much finer control of access rights can be defined using mutators and accessors. For example, a parameter may be made read-only simply by defining an accessor but not a mutator. The visibility of the two methods may be different; it is often useful for the accessor to be public while the mutator remains protected, package-private or internal.
The block where the mutator is defined provides an opportunity for validation or preprocessing of incoming data. If all external access is guaranteed to come through the mutator, then these steps cannot be bypassed. For example, if a date is represented by separate private year, month and day variables, then incoming dates can be split by the setDate mutator while for consistency the same private instance variables are accessed by setYear and setMonth. In all cases month values outside of 1 - 12 can be rejected by the same code.
Accessors conversely allow for synthesis of useful data representations from internal variables while keeping their structure encapsulated and hidden from outside modules. A monetary getAmount accessor may build a string from a numeric variable with the number of decimal places defined by a hidden currency parameter.
Modern programming languages often offer the ability to generate the boilerplate for mutators and accessors in a single line—as for example C#'s public string Name { get; set; } and Ruby's attr_accessor :name. In these cases, no code blocks are created for validation, preprocessing or synthesis. These simplified accessors still retain the advantage of encapsulation over simple public instance variables, but it is common that, as system designs progress, the software is maintained and requirements change, the demands on the data become more sophisticated. Many automatic mutators and accessors eventually get replaced by separate blocks of code. The benefit of automatically creating them in the early days of the implementation is that the public interface of the class remains identical whether or not greater sophistication is added, requiring no extensive refactoring if it is.[1]
Manipulation of parameters that have mutators and accessors from inside the class where they are defined often requires some additional thought. In the early days of an implementation, when there is little or no additional code in these blocks, it makes no difference if the private instance variable is accessed directly or not. As validation, cross-validation, data integrity checks, preprocessing or other sophistication is added, subtle bugs may appear where some internal access makes use of the newer code while in other places it is bypassed.
Accessor functions are always less efficient than directly fetching or storing data fields due to the extra steps involved.[2]
Note that it is perfectly possible to do object-oriented programming with guaranteed encapsulation in pure C.
In file student.h:
#ifndef STUDENT_H #define STUDENT_H typedef struct student *student_t; student_t student_new(int age, char *name); void student_delete(student_t s); void student_set_age(student_t s, int age); int student_get_age(student_t s); #endif
In file student.c:
#include "student.h" struct student { int age; char *name; }; student_t student_new(int age, char *name) { student_t s = malloc(sizeof *s); s->age = age; s->name = name; return s; } void student_delete(student_t s) { free(s); } void student_set_age(student_t s, int age) { s->age = age; } int student_get_age(student_t s) { return s->age; }
In file main.c:
#include "student.h" int main(void) { student_t s = student_new(19, "Maurice"); int old_age = student_get_age(s); student_set_age(s, 21); student_delete(s); return 0; }
In file Student.h:
#ifndef STUDENT_H #define STUDENT_H #include <string> class Student { public: Student(const std::string& name); const std::string& name() const; void name(const std::string& name); private: std::string name_; }; #endif
In file Student.cpp:
#include "Student.h" Student::Student(const std::string& name) : name_(name) { } const std::string& Student::name() const { return name_; } void Student::name(const std::string& name) { name_ = name; }
This example illustrates the C# idea of properties, which are a special type of class member. Unlike Java, no explicit methods are defined; a public 'property' contains the logic to handle the actions. Note use of the built-in (undeclared) variable value.
public class Student { private string name; /// <summary> /// Gets or sets student's name /// </summary> public string Name { get { return name; } set { name = value; } } }
In later C# versions (.NET Framework 3.5 and above), this example may be abbreviated as follows, without declaring the private variable name.
public class Student { public string Name { get; set; } }
Using the abbreviated syntax means that the underlying variable is no longer available from inside the class. As a result, the set portion of the property must be present for assignment. Access can be restricted with a set-specific access modifier.
public class Student { public string Name { get; private set; } }
In Common Lisp Object System, slot specifications within class definitions may specify any of the :reader, :writer and :accessor options (even multiple times) to define reader methods, setter methods and acessor methods (a reader method and the respective setf method).[3] Slots are always directly accessible through their names with the use of with-slots and slot-value, and the slot accessor options define specialized methods that use slot-value.[4]
CLOS itself has no notion of properties, although the MetaObject Protocol extension specifies means to access a slot's reader and writer function names, including the ones generated with the :accessor option.[5]
The following example shows a definition of a student class using these slot options and direct slot access:
(defclass student () ((name :initarg :name :initform "" :accessor student-name) ; student-name is setf'able (birthdate :initarg :birthdate :initform 0 :reader student-birthdate) (number :initarg :number :initform 0 :reader student-number :writer set-student-number))) ;; Example of a calculated property getter (this is simply a method) (defmethod student-age ((self student)) (- (get-universal-time) (student-birthdate self))) ;; Example of direct slot access within a calculated property setter (defmethod (setf student-age) (new-age (self student)) (with-slots (birthdate) self (setf birthdate (- (get-universal-time) new-age)) new-age)) ;; The slot accessing options generate methods, thus allowing further method definitions (defmethod set-student-number :before (new-number (self student)) ;; You could also check if a student with the new-number already exists. (check-type new-number (integer 1 *)))
This is a simple class in Delphi language that illustrates the concept of public property that access a private field.
interface type TStudent = class private fName: string; procedure SetName(const Value: String); public /// <summary> /// Set or get the name of the student. /// </summary> property Name:String read fName write SetName; end;
In this example of a simple class representing a student with only the name stored, one can see the variable name is private, i.e. only visible from the Student class, and the "setter" and "getter" is public, namely the "getName()" and "setName(name)" methods.
public class Student { private String name; public String getName() { return name; } public void setName(String newName) { name = newName; } }
In this example constructor-function Student is used to create objects representing a student with only the name stored.
function Student(name) { var name_ = name; this.getName = function() { return name_; } this.setName = function(name) { name_ = name; } }
Or:
function Student(name){ var name_ = name; this.__defineGetter__("name", function() { return name_; }); this.__defineSetter__("name", function(name) { name_ = name; }); }
package { public class Student { private var _name : String; public function get name() : String { return _name; } public function set name(value : String) : void { _name = value; } } }
package Student; sub new { bless {}, shift; } sub set_name { my $self = shift; $self->{name} = $_[0]; } sub get_name { my $self = shift; return $self->{name}; } 1;
Or, using Class::Accessor
package Student; use base qw(Class::Accessor); __PACKAGE__->follow_best_practice; Student->mk_accessors(qw(name)); 1;
Or, using Moose Object System:
package Student; use Moose; # Moose uses the attribute name as the setter and getter, the reader and writer properties # allow us to override that and provide our own names, in this case get_name and set_name has 'name' => (is => 'rw', isa => 'Str', reader => 'get_name', writer => 'set_name'); 1;
In this example of a simple class representing a student with only the name stored, one can see the variable name is private, i.e. only visible from the Student class, and the "setter" and "getter" is public, namely the getName() and setName('name') methods.
class Student { private $name; /** * @return the $name */ public function getName() { return $this->name; } /** * @param $newName * the name to set */ public function setName($newName) { $this->name = $newName; } }
This example uses a Python class with one variable, a getter, and a setter.
class Student(object): # Constructor def __init__(self, name): # A variable to hold the students name self._name = name # A function to print the name to stdout def getName(self): return self._name # A function to set a new name def setName(self, newName): self._name = newName name = property(getName, setName)
In Racket, the object system is a way to organize code that comes in addition to modules and units. As in the rest of the language, the object system has first-class values and lexical scope is used to control access to objects and methods.
"Alice"; => "Alice" "Bob"; => "Bob"
Struct definitions are an alternative way to define new types of values, with mutators being present when explicitly required:
"Alice""Bob"; => "Bob"
In Ruby, individual accessor and mutator methods may be defined, or the metaprogramming constructs attr_reader or attr_accessor may be used both to declare a private variable in a class and to provide either read-only or read-write public access to it respectively.
Defining individual accessor and mutator methods creates space for pre-processing or validation of the data
class Student def name @name end def name=(value) @name=value end end
Read-only simple public access to implied @name variable
class Student attr_reader :name end
Read-write simple public access to implied @name variable
class Student attr_accessor :name end
age: aNumber " Set the receiver age to be aNumber if is greater than 0 and less than 150 " (aNumber between: 0 and: 150) ifTrue: [ age := aNumber ]
This example illustrates the VB.NET idea of properties, which are used in classes. Similar to C#, there is an explicit use of the Get and Set methods.
Public Class Student Private _name As String Public Property Name() Get Return _name End Get Set(ByVal value) _name = value End Set End Property End Class
In VB.NET 2010, Auto Implemented properties can be utilized to create a property without having to use the Get and Set syntax. Note that a hidden variable is created by the compiler, called _name, to correspond with the Property name. Using another variable within the class named _name would result in an error. Privileged access to the underlying variable is available from within the class.
Public Class Student Public Property name As String End Class