In all the programs we wrote till now, we have designed our program around functions i.e. blocks of statements which manipulate data. This is called the procedure-oriented way of programming. There is another way of organizing your program which is to combine data and functionality and wrap it inside something called an object. This is called the object oriented programming paradigm. Most of the time you can use procedural programming, but when writing large programs or have a problem that is better suited to this method, you can use object oriented programming techniques.
Classes and objects are the two main aspects of object oriented programming. A class creates a
new type where objects are instances of the class. An analogy is that you can have variables
of type int which translates to saying that variables that store integers are variables which are
instances (objects) of the int class.
|
Note
|
Note for Static Language Programmers
Note that even integers are treated as objects (of the See C# and Java 1.5 programmers will find this similar to the boxing and unboxing concept. |
Objects can store data using ordinary variables that belong to the object. Variables that belong to an object or class are referred to as fields. Objects can also have functionality by using functions that belong to a class. Such functions are called methods of the class. This terminology is important because it helps us to differentiate between functions and variables which are independent and those which belong to a class or object. Collectively, the fields and methods can be referred to as the attributes of that class.
Fields are of two types - they can belong to each instance/object of the class or they can belong to the class itself. They are called instance variables and class variables respectively.
A class is created using the class keyword. The fields and methods of the class are listed in an
indented block.
Class methods have only one specific difference from ordinary
functions - they must have an extra first name that has to be added to
the beginning of the parameter list, but you do not give a value
for this parameter when you call the method, Python will provide
it. This particular variable refers to the object itself, and by convention, it is given the name
self.
Although, you can give any name for this parameter, it is strongly recommended that you use the
name self - any other name is definitely frowned upon. There are many advantages to using a
standard name - any reader of your program will immediately recognize it and even specialized IDEs
(Integrated Development Environments) can help you if you use self.
|
Note
|
Note for C++/Java/C# Programmers
The self in Python is equivalent to the this pointer in C++ and the this reference in Java
and C#.
|
You must be wondering how Python gives the value for self and why you don’t need to give a value
for it. An example will make this clear. Say you have a class called MyClass and an instance of
this class called myobject. When you call a method of this object as myobject.method(arg1,
arg2), this is automatically converted by Python into MyClass.method(myobject, arg1, arg2) -
this is all the special self is about.
This also means that if you have a method which takes no arguments, then you still have to have one
argument - the self.
The simplest class possible is shown in the following example (save as oop_simplestclass.py).
link:programs/oop_simplestclass.py[role=include]Output:
link:programs/oop_simplestclass.txt[role=include]
We create a new class using the class statement and the name of the class. This is followed by an
indented block of statements which form the body of the class. In this case, we have an empty block
which is indicated using the pass statement.
Next, we create an object/instance of this class using the name of the class followed by a pair of
parentheses. (We will learn more about instantiation in the next section). For our
verification, we confirm the type of the variable by simply printing it. It tells us that we have
an instance of the Person class in the main module.
Notice that the address of the computer memory where your object is stored is also printed. The address will have a different value on your computer since Python can store the object wherever it finds space.
We have already discussed that classes/objects can have methods just like functions except that we
have an extra self variable. We will now see an example (save as oop_method.py).
link:programs/oop_method.py[role=include]Output:
link:programs/oop_method.txt[role=include]
Here we see the self in action. Notice that the say_hi method takes no parameters but still has
the self in the function definition.
There are many method names which have special significance in Python classes. We will see the
significance of the init method now.
The init method is run as soon as an object of a class is instantiated. The method is useful
to do any initialization you want to do with your object. Notice the double underscores both at
the beginning and at the end of the name.
Example (save as oop_init.py):
link:programs/oop_init.py[role=include]Output:
link:programs/oop_init.txt[role=include]
Here, we define the init method as taking a parameter name (along with the usual self).
Here, we just create a new field also called name. Notice these are two different variables even
though they are both called 'name'. There is no problem because the dotted notation self.name
means that there is something called "name" that is part of the object called "self" and the other
name is a local variable. Since we explicitly indicate which name we are referring to, there is
no confusion.
Most importantly, notice that we do not explicitly call the init method but pass the
arguments in the parentheses following the class name when creating a new instance of the
class. This is the special significance of this method.
Now, we are able to use the self.name field in our methods which is demonstrated in the sayHi
method.
We have already discussed the functionality part of classes and objects (i.e. methods), now let us learn about the data part. The data part, i.e. fields, are nothing but ordinary variables that are bound to the namespaces of the classes and objects. This means that these names are valid within the context of these classes and objects only. That’s why they are called name spaces.
There are two types of fields - class variables and object variables which are classified depending on whether the class or the object owns the variables respectively.
Class variables are shared - they can be accessed by all instances of that class. There is only one copy of the class variable and when any one object makes a change to a class variable, that change will be seen by all the other instances.
Object variables are owned by each individual object/instance of the class. In this case, each
object has its own copy of the field i.e. they are not shared and are not related in any way to the
field by the same name in a different instance. An example will make this easy to understand (save
as oop_objvar.py):
link:programs/oop_objvar.py[role=include]Output:
link:programs/oop_objvar.txt[role=include]
This is a long example but helps demonstrate the nature of class and object variables. Here,
population belongs to the`Robot` class and hence is a class variable. The name variable belongs
to the object (it is assigned using self) and hence is an object variable.
Thus, we refer to the population class variable as Robot.population and not as
self.population. We refer to the object variable name using self.name notation in the methods
of that object. Remember this simple difference between class and object variables. Also note that
an object variable with the same name as a class variable will hide the class variable!
Instead of Robot.population, we could have also used self.class.population because every
object refers to it’s class via the self.class attribute.
The how_many is actually a method that belongs to the class and not to the object. This means we
can define it as either a classmethod or a staticmethod depending on whether we need to know
which class we are part of. Since we refer to a class variable, let’s use classmethod.
We have marked the how_many method as a class method using a decorator.
Decorators can be imagined to be a shortcut to calling a wrapper function, so applying the
@classmethod decorator is same as calling:
how_many = classmethod(how_many)Observe that the init method is used to initialize the Robot instance with a name. In this
method, we increase the population count by 1 since we have one more robot being added. Also
observe that the values of self.name is specific to each object which indicates the nature of
object variables.
Remember, that you must refer to the variables and methods of the same object using the self
only. This is called an attribute reference.
In this program, we also see the use of docstrings for classes as well as methods. We can access
the class docstring at runtime using Robot.doc and the method docstring as
Robot.say_hi.doc
In the die method, we simply decrease the Robot.population count by 1.
All class members are public. One exception: If you use data members with names using the double
underscore prefix such as __privatevar, Python uses name-mangling to effectively make it a
private variable.
Thus, the convention followed is that any variable that is to be used only within the class or object should begin with an underscore and all other names are public and can be used by other classes/objects. Remember that this is only a convention and is not enforced by Python (except for the double underscore prefix).
|
Note
|
Note for C++/Java/C# Programmers
All class members (including the data members) are public and all the methods are virtual in
Python.
|
One of the major benefits of object oriented programming is reuse of code and one of the ways this is achieved is through the inheritance mechanism. Inheritance can be best imagined as implementing a type and subtype relationship between classes.
Suppose you want to write a program which has to keep track of the teachers and students in a college. They have some common characteristics such as name, age and address. They also have specific characteristics such as salary, courses and leaves for teachers and, marks and fees for students.
You can create two independent classes for each type and process them but adding a new common characteristic would mean adding to both of these independent classes. This quickly becomes unwieldy.
A better way would be to create a common class called SchoolMember and then have the teacher and
student classes inherit from this class i.e. they will become sub-types of this type (class) and
then we can add specific characteristics to these sub-types.
There are many advantages to this approach. If we add/change any functionality in SchoolMember,
this is automatically reflected in the subtypes as well. For example, you can add a new ID card
field for both teachers and students by simply adding it to the SchoolMember class. However,
changes in the subtypes do not affect other subtypes. Another advantage is that if you can refer to
a teacher or student object as a SchoolMember object which could be useful in some situations
such as counting of the number of school members. This is called polymorphism where a sub-type
can be substituted in any situation where a parent type is expected i.e. the object can be treated
as an instance of the parent class.
Also observe that we reuse the code of the parent class and we do not need to repeat it in the different classes as we would have had to in case we had used independent classes.
The SchoolMember class in this situation is known as the base class or the superclass. The
Teacher and Student classes are called the derived classes or subclasses.
We will now see this example as a program (save as oop_subclass.py):
link:programs/oop_subclass.py[role=include]Output:
link:programs/oop_subclass.txt[role=include]
To use inheritance, we specify the base class names in a tuple following the class name in the
class definition. Next, we observe that the init method of the base class is explicitly
called using the self variable so that we can initialize the base class part of the object. This
is very important to remember - Python does not automatically call the constructor of the base
class, you have to explicitly call it yourself.
We also observe that we can call methods of the base class by prefixing the class name to the
method call and then pass in the self variable along with any arguments.
Notice that we can treat instances of Teacher or Student as just instances of the
SchoolMember when we use the tell method of the SchoolMember class.
Also, observe that the tell method of the subtype is called and not the tell method of the
SchoolMember class. One way to understand this is that Python always starts looking for methods
in the actual type, which in this case it does. If it could not find the method, it starts looking
at the methods belonging to its base classes one by one in the order they are specified in the
tuple in the class definition.
A note on terminology - if more than one class is listed in the inheritance tuple, then it is called multiple inheritance.
The trailing comma is used at the end of the print statement in the superclass’s tell() method
to print a line and allow the next print to continue on the same line. This is a trick to make
print not print a \n (newline) symbol at the end of the printing.
We have now explored the various aspects of classes and objects as well as the various terminologies associated with it. We have also seen the benefits and pitfalls of object-oriented programming. Python is highly object-oriented and understanding these concepts carefully will help you a lot in the long run.
Next, we will learn how to deal with input/output and how to access files in Python.