Class (computer science)
|
|
In object-oriented programming, a class consists of a collection of types of encapsulated instance variables and types of methods, possibly with implementation of those types together with a constructor function that can be used to create objects of the class. A class is a cohesive package that consists of a particular kind of compile-time metadata. A Class describes the rules by which objects behave; these objects are referred to as "instances" of that class. A class specifies the structure of data which each instance contains as well as the methods (functions) which manipulate the data of the object; such methods are sometimes described as "behavior". A method is function with a special property that it has access to data stored in an object. A class is the most specific type of an object in relation to a specific layer.
Instances of a class will have certain aspects in common. For example, a class Person would describe the properties common to all instances of the Person class. One of the benefits of programming with classes is that all instances of a particular class will follow the defined behaviour of said class. Each person is generally alike; but varies in such properties as "height" and "weight". The class would list types of such instance variables; and also define, via methods, the actions which humans can perform: "run", "jump", "sleep", "throw object", etc.
| Contents |
|
|
Interface to a class
A class implements (or realizes) one or more interfaces. Each interface specifies named operation signatures of some of the methods of the class. The methods of a class can either be used via one of its interfaces or directly. An interface of a class describes the possible behaviours of objects of the class from user's point of view. Multiple classes can implement the same interface. Partial implementations of interfaces are not allowed, that is, every operation has to have an implementation. However usually not every combination of parameter values need to be supported.
Each interface of the class is associated with a type of object references referring to the interface, through which methods of objects can be invoked. Each reference points to a single instance of the class. Each reference has a lifetime, which specifies how long the reference can be used, usually bound to the time when some specific operations are invoked via the interface. It is assumed that there is a mechanism for accessing the object with a valid object reference. However, the object reference does not necessarily point to any single location, since the same object can be located at different times in different places.
The object-oriented programming methodology is designed in such way that the operations of any interface of a class are usually chosen to be independent of each other. This means that an interface places no requirements for clients to invoke the operations of one interface in any particular order. This approach has the benefit that client code can rely that the operations of an interface are available for use whenever the client holds a valid reference to the object. This will also result in a client-server (or layered) design where servers do not depend in any way of the clients.
The methods that are not in any of the interfaces of a class are private to the class, and not intended to be depended on by other classes.
The "set of all interfaces of a class" is sometimes called the interface of the class.
The internal data structures defined as part of a class are not considered to be part of its interface. Rather, public accessor methods can be used to inspect or alter object data. The various object-oriented programming languages enforce this to various degrees. For example, Java does not allow the programmer to access the private data of a class at all, whereas in languages like Objective-C or Perl the programmer can do what they want. In C++, private methods are visible but not accessible in the interface; however they are commonly made invisible by explicitly declaring fully abstract classes that represent the interfaces of the class.
Structure of a class
Oop-uml-class-example.png
A class contains a description of structure of data ("state") stored in the objects of the class. The state of an object is stored in some resource, such as memory or a file. The storage is assumed to be located in a specific location, such that it is possible to access the object via references to the identity of the objects. However, the actual storage location associated with an object may change with time. In such situations, the identity of the object does not change. The state is encapsulated and every access to the state occurs via methods of the class. Specific data items in the state, such as xsize and ysize in the example, are sometimes called class attributes or class properties.
A class implements its interfaces by specifying methods that describe what operations can be performed on the data stored in the objects of the class. Each method specifies only tasks that are related to the stored data. Multimethods can be used when a single task requires access to many objects' data.
A class also describes a set of invariants that are preserved by every method in the class. An invariant is a constraint on the state of an object that should be satisfied by every object of the class. The main purpose of the invariants is to establish what objects belong to the class. An invariant is what distinguishes datatypes and classes from each other, that is, a class does not allow use of all possible values for the state of the object, only those that are well-defined by the semantics of the intended use of the datatype. The set of supported methods often implicitly establishes an invariant. Some programming languages support specification of invariants as part of the definition of the class, and enforce them via the type system. Encapsulation of state is necessary for being able to enforce the invariants of the class.
An implementation of a class specifies constructor and destructor functions that allow creation and destruction of objects of the class. A constructor that takes arguments can be used to create an object from data. A destructor that returns a value can be used to obtain a representation of an object of a class. The main purpose of a constructor is to establish the invariant of the class, failing if the invariant isn't valid. The main purpose of a destructor is to destroy the identity of the object, invalidating any references in the process. Constructors and destructors are also sometimes used to reserve and release resources associated with the object.
A class can also implement a set of auxiliary functions, sometimes called class functions or static methods. Static methods are often used to find, create or destroy objects of the class. Constructors and destructors are sometimes specified as static methods. Often, mechanisms for sending an object to another location or changing the class of an object are specified as static methods.
Associations between classes
An association between two classes is a type of a link between the corresponding objects. A (two-way) association between classes A and B describes a relationship between each object of class A and some objects of class B, and vice versa. Associations are often named with a verb, such as "subscribes-to".
An association role describes the role of an instance of a class when the instance participates in an association. An association role is related to each end of the association. The role describes an instance of a class from the point of view of a situation in which the instance participates in the association. For example, a "subscriber" role describes instances of the class "Person" when they participate in a "subscribes-to" relationship with the class "Magazine". Also, a "Magazine" has the "subscribed magazine" role when the subscribers subscribe-to it.
Association role multiplicity describes how many instances correspond to each instance of the other class(es) of the association. Common multiplicities are "0..1", "1..1", "1..*" and "0..*", where the "*" specifies any number of instances.
Subclasses and superclasses
Classes are often related in some way. The most popular of these relations is inheritance, which involves subclasses and superclasses, also known respectively as child classes (or derived classes) and parent classes (or base classes). If [car] was a class, then [Jaguar] and [Porsche] might be two sub-classes. If [Button] is a subclass of [Control], then all buttons are controls. Subclasses usually consists of several kinds of modifications to the base class: addition of new instance variables, addition of new methods and overriding of existing methods to support the new instance variables.
Conceptually, a superclass should be considered as a common part of its subclasses. This factoring of commonality is one mechanism for providing reuse. Thus, extending a superclass by modifying the existing class is also likely to narrow its applicability in various situations. In Object-oriented design, careful balance between applicability and functionality of superclasses should be considered. Subclassing is different from subtyping in that subtyping deals with common behaviour whereas subclassing is concerned with common structure.
Some programming languages (for example C++) allow multiple inheritance -- they allow a child class to have more than one parent class. This technique has been criticized by some for its unnecessary complexity and being difficult to implement efficiently, though some projects have certainly benefited from its use. Java, for example has no multiple inheritance, its designers feeling that it would add unnecessary complexity.
Sub- and superclasses are considered to exist within a hierarchy defined by the inheritance relationship. If multiple inheritance is allowed, this hierarchy is a directed acyclic graph (or DAG for short), otherwise it is a tree. The hierarchy has classes as nodes and inheritance relationships as links. The levels of this hierarchy are called layers or levels of abstraction. Classes in the same level are more likely to be associated than classes in different levels.
There are two slightly different points of view as to whether subclasses of the same class are required to be disjoint. Sometimes, subclasses of a particular class are considered to be completely disjoint. That is, every instance of a class has exactly one most-derived class, which is a subclass of every class that the instance has. This view does not allow dynamic change of object's class, as objects are assumed to be created with a fixed most-derived class. The basis for not allowing changes to object's class is that the class is a compile-time type, which does not usually change at runtime, and polymorphism is utilised for any dynamic change to the object's behaviour, so this ability is not necessary. And design that does not need to perform changes to object's type will be more robust and easy-to-use from the point of view of the users of the class.
From another point of view, subclasses are not required to be disjoint. Then there is no concept of a most-derived class, and all types in the inheritance hierarchy that are types of the instance are considered to be equally types of the instance. This view is based on a dynamic classification of objects, such that an object may change its class at runtime. Then object's class is considered to be its current structure, but changes to it are allowed. The basis for allowing object's class to change is performance. It's more efficient to allow changes to object's type, since references to the existing instances do not need to be replaced with references to new instances when the class of the object changes. However, this ability is not readily available in all programming languages.
Reasons for implementing classes
Classes, when used properly, can accelerate development by reducing redundant code entry, testing and bug fixing. If a class has been thoroughly tested and is known to be a solid work, it stands to reason that implementing that class or extending it will reduce if not eliminate the possibility of bugs propagating into the code. In the case of extension new code is being added so it also requires the same level of testing before it can be considered solid.
Another reason for using classes is to simplify the relationships of interrelated data. Rather than writing code to repeatedly draw a GUI window on the terminal screen, it is simpler to represent the window as an object and tell it to draw itself as necessary. With classes, GUI items that are similar to windows (such as dialog boxes) can simply inherit most of their functionality and data structures from the window class. The programmer then need only add code to the dialog class that is unique to its operation. Indeed, GUIs are a very common and useful application of classes, and GUI programming is generally much easier with a good class framework.
Categories of Classes
Abstract and Concrete classes
An abstract class, or abstract base class (ABC), is one that is designed only as a parent class and from which child classes may be derived, and which is not itself suitable for instantiation. Abstract classes are often used to represent abstract concepts or entities. The incomplete features of the abstract class are then shared by a group of sibling sub-classes which add different variations of the missing pieces. In C++, an abstract class is defined as a class having at least one virtual function without an implementation.
Abstract classes are superclasses which contain abstract methods and are defined such that subclasses are to extend them by implementing the methods. The behaviors defined by such a class are "generic" and much of the class will be undefined and unimplemented. Before a class derived from an abstract class can be instantiated, it must implement particular methods for all the abstract methods of its parent classes.
In computing, when specifying an abstract class, the programmer is referring to a class which has elements that are meant to be implemented by inheritance. The abstraction of the class methods to be implemented by the sub-classes is meant to simplify software development.
A concrete class, however, is a class for which entities (instances) may be created. This contrasts with abstract classes which can not be instantiated because it defeats its purpose of being an 'abstract'.
Most object oriented programming languages allow the programmer to specify which classes are considered abstract and will not allow these to be instantiated (in Java, for example, the keyword abstract is used). This also enables the programmer to focus on planning and design. The actual implementation of course is to be done in the derived classes.
In C++, an abstract class is a class having at least one pure virtual function. They can not be instantiated and will generate an error if an attempt is made. They are meant to function as stubs, allowing the programmer to identify what modules of functions (behaviour or methods) are needed without having to actually implement them. This is in line with OOP's philosophy of allowing the programmer to concentrate on how an object should behave without going into the actual detail.
Metaclasses
Metaclasses are classes whose instances are classes. A metaclass describes a common structure of a collection of classes. A metaclass can implement a design pattern or describe a shorthand for particular kinds of classes. Metaclasses are often used to describe frameworks.
In some languages such as SmallTalk and Ruby, a class is also an object; thus each class is an instance of the unique metaclass, which is built in the language. For example, in Objective-C, each object and class is an instance of NSObject. CLOS (Common Lisp Object System) provides metaobject protocols (MOP) to implement those classes and metaclasses.
Non-class-based programming
To the surprise to some familiar with the use of classes for OOP, it has been shown that one can design fully fledged object-oriented languages that do not have builtin supports of classes. Those languages are usually designed with the motive to address the problem of tigh-coupling between implementations and interfaces due to the use of classes. For example, SELF was designed to show that the role of a class can be substituted by using an existing object which serves as a prototype to a new object, and the resulting language is as expressive as SmallTalk with more generality in the creation of objects. See class-based OOP for the criticism of class-based programming and object-based languages for such non-class-based languages.
Run-time representation of classes
As a datatype, a class is usually considered as a compile-time construct. A language may also support prototype or factory metaobjects that represent run-time information about classes, or even represent metadata that provides access to reflection facilities and ability to manipulate data structure formats at run-time. Many languages distinguish this kind of run-time type information about classes from a class on the basis that the information is not needed at run-time. Some dynamic languages do not make strict distinctions between run-time and compile-time constructs, and therefore may not distinguish between metaobjects and classes.
For example: if humans is a metaobject representing the class Person, then instances of class Person can be created by using the facilities of the human metaobject.
Classes without inheritance
Not every language that both supports objects and classes is generally seen as object-oriented. Examples are JavaScript and Visual Basic, which lack the support for inheritance. The lack of inheritance severly impairs the full practice of object-oriented programming. Those languages, sometimes called "object-based languages", do not provide the structural benefits of statically type checked interfaces for objects. This is because in object-based languages it is possible to use and extend data structures and attach methods to them at run-time. This precludes the compiler or interpreter from being able to check the type information specified in the source code as the type is built dynamically and not defined statically. Most of these languages allow for instance behaviour and complex operational polymorphism (see dynamic dispatch and polymorphism).
Instantiation
As explained above, classes can be used to create new objects by instantiating them. In most languages, the structures as defined the class determines how the memory used by its instances will be laid out. This technique is known as the cookie-cutter model.
The alternative to the cookie-cutter model is that of for instance Python, where objects are structured as associative key-value containers. In such models, objects that are instances of the same class could contain different instance variables, as state can be dynamically added to the object. This may resemble Prototype-based languages in some ways, but it is not equivalent.
Examples
C++
Example 1
class example {
// this is a class
};
This example shows how to define a C++ class. It has no data, and performs no functions; it only contains the comment, "this is a class".
Example 2
class Abstract
{
public:
virtual void MyVirtualMethod() = 0;
};
class Concrete : public Abstract
{
public:
void MyVirtualMethod()
{
//do something
}
};
An object of class Abstract can not be created because the function MyVirtualMethod has not been defined (the =0 is C++ syntax for a pure virtual function, a function that must be part of any derived concrete class but is not defined in the abstract base class. The Concrete class is a concrete class because its functions (in this case, only one function) have been declared and implemented.
Example 3
#include <string>
using std::string;
class InetMessage
{
string m_subject, m_to, m_from;
public:
InetMessage (const string& subject,
const string& to,
const string& from);
string subject () const;
string to () const;
string from () const;
};
</pre>
Java
Example 1
public class Example1
{
// This is a Java class, it automatically extends the class Object
}
This example shows the simplest Java class possible.
Example 2
public class Example2 extends Example1
{
// This is a class that extends the class created in Example 1.
protected int data;
public Example2()
{
// This is a constructor for the class. It does not have a return type.
data = 1;
}
public int getData()
{
return data;
}
public void setData(int d)
{
data = d;
}
}
This example shows a class that has a defined constructor, one member data, and two accessor methods for that member data. It extends the previous example's class. Note that in Java all classes automatically extend the class Object. This allows you to write generic code to deal with objects of any type.
See also
ja:クラス fi:Luokka lt:Klasė (programavimas) pl:Klasa abstrakcyjna sv:Klass (programmering) fr:Classe (informatique) zh:类 (计算机科学)