Think Python/Classes and methods

Object-oriented features
Python is an object-oriented programming language, which means that it provides features that support object-oriented programming.

It is not easy to define object-oriented programming, but we have already seen some of its characteristics:


 * Programs are made up of object definitions and function definitions, and most of the computation is expressed in terms of operations on objects.


 * Each object definition corresponds to some object or concept in the real world, and the functions that operate on that object correspond to the ways real-world objects interact.

For example, the Time class defined in Chapter&#XA0;16 corresponds to the way people record the time of day, and the functions we defined correspond to the kinds of things people do with times. Similarly, the Point and Rectangle classes correspond to the mathematical concepts of a point and a rectangle.

So far, we have not taken advantage of the features Python provides to support object-oriented programming. These features are not strictly necessary; most of them provide alternative syntax for things we have already done. But in many cases, the alternative is more concise and more accurately conveys the structure of the program.

For example, in the Time program, there is no obvious connection between the class definition and the function definitions that follow. With some examination, it is apparent that every function takes at least one Time object as an argument.

This observation is the motivation for methods; a method is a function that is associated with a particular class. We have seen methods for strings, lists, dictionaries and tuples. In this chapter, we will define methods for user-defined types.

Methods are semantically the same as functions, but there are two syntactic differences:


 * Methods are defined inside a class definition in order to make the relationship between the class and the method explicit.
 * The syntax for invoking a method is different from the syntax for calling a function.

In the next few sections, we will take the functions from the previous two chapters and transform them into methods. This transformation is purely mechanical; you can do it simply by following a sequence of steps. If you are comfortable converting from one form to another, you will be able to choose the best form for whatever you are doing.

Printing objects
In Chapter&#XA0;16, we defined a class named Time and in Exercise&#XA0;16.1, you wrote a function named : To call this function, you have to pass a Time object as an argument: To make  a method, all we have to do is move the function definition inside the class definition. Notice the change in indentation.

Now there are two ways to call. The first (and less common) way is to use function syntax:

In this use of dot notation, Time is the name of the class, and  is the name of the method. <TT>start</TT> is passed as a parameter.

The second (and more concise) way is to use method syntax:

In this use of dot notation,  is the name of the method (again), and <TT>start</TT> is the object the method is invoked on, which is called the subject. Just as the subject of a sentence is what the sentence is about, the subject of a method invocation is what the method is about.

Inside the method, the subject is assigned to the first parameter, so in this case <TT>start</TT> is assigned to <TT>time</TT>.

By convention, the first parameter of a method is called <TT>self</TT>, so it would be more common to write like this: The reason for this convention is an implicit metaphor:

suggests that the function is the active agent. It says something like, &#X201C;Hey ! Here&#X2019;s an object for you to print.&#X201D;
 * The syntax for a function call, ,

A method invocation like  says &#X201C;Hey <TT>start</TT>! Please print yourself.&#X201D;
 * In object-oriented programming, the objects are the active agents.

This change in perspective might be more polite, but it is not obvious that it is useful. In the examples we have seen so far, it may not be. But sometimes shifting responsibility from the functions onto the objects makes it possible to write more versatile functions, and makes it easier to maintain and reuse code.

Exercise 1
Rewrite  '' (from Section&#XA0;'16.4') as a method. It is probably not appropriate to rewrite  as a method; it&#X2019;s not clear what object you would invoke it on!''

Another example
Here&#X2019;s a version of <TT>increment</TT> (from Section&#XA0;16.3) rewritten as a method: This version assumes that  is written as a method, as in Exercise&#XA0;17.1. Also, note that it is a pure function, not a modifier.

Here&#X2019;s how you would invoke <TT>increment</TT>: The subject, <TT>start</TT>, gets assigned to the first parameter, <TT>self</TT>. The argument, <TT>1337</TT>, gets assigned to the second parameter, <TT>seconds</TT>.

This mechanism can be confusing, especially if you make an error. For example, if you invoke <TT>increment</TT> with two arguments, you get:

The error message is initially confusing, because there are only two arguments in parentheses. But the subject is also considered an argument, so all together that&#X2019;s three.

A more complicated example
(from Exercise&#XA0;16.2) is slightly more complicated because it takes two Time objects as parameters. In this case it is conventional to name the first parameter <TT>self</TT> and the second parameter <TT>other</TT>:

To use this method, you have to invoke it on one object and pass the other as an argument: One nice thing about this syntax is that it almost reads like English: &#X201C;end is after start?&#X201D;

The init method
The init method (short for &#X201C;initialization&#X201D;) is a special method that gets invoked when an object is instantiated. Its full name is  (two underscore characters, followed by <TT>init</TT>, and then two more underscores). An init method for the <TT>Time</TT> class might look like this: It is common for the parameters of to have the same names as the attributes. The statement stores the value of the parameter <TT>hour</TT> as an attribute of <TT>self</TT>.

The parameters are optional, so if you call <TT>Time</TT> with no arguments, you get the default values. If you provide one argument, it overrides <TT>hour</TT>: If you provide two arguments, they override <TT>hour</TT> and <TT>minute</TT>. And if you provide three arguments, they override all three default values.

Exercise 2
Write an init method for the <TT>Point</TT> class that takes <TT>x</TT> and <TT>y</TT> as optional parameters and assigns them to the corresponding attributes.

The <TT>__str__</TT> method
is a special method, like , that is supposed to return a string representation of an object.

For example, here is a <TT>str</TT> method for Time objects: When you <TT>print</TT> an object, Python invokes the <TT>str</TT> method:

When I write a new class, I almost always start by writing , which makes it easier to instantiate objects, and , which is useful for debugging.

Exercise 3
Write a <TT>str</TT> method for the <TT>Point</TT> ''class. Create a Point object and print it.''

Operator overloading
By defining other special methods, you can specify the behavior of operators on user-defined types. For example, if you define a method named  for the <TT>Time</TT> class, you can use the <TT>+</TT> operator on Time objects.

Here is what the definition might look like:

And here is how you could use it: When you apply the <TT>+</TT> operator to Time objects, Python invokes . When you print the result, Python invokes . So there is quite a lot happening behind the scenes!

Changing the behavior of an operator so that it works with user-defined types is called operator overloading. For every operator in Python there is a corresponding special method, like . For more details, see <TT>docs.python.org/ref/specialnames.html</TT>.

Exercise 4
Write an '<TT>add</TT>' method for the Point class.

Type-based dispatch
In the previous section we added two Time objects, but you also might want to add an integer to a Time object. The following is a version of that checks the type of <TT>other</TT> and invokes either or <TT>increment</TT>: The built-in function <TT>isinstance</TT> takes a value and a class object, and returns <TT>True</TT> if the value is an instance of the class.

If <TT>other</TT> is a Time object,  invokes . Otherwise it assumes that the parameter is a number and invokes <TT>increment</TT>. This operation is called a type-based dispatch because it dispatches the computation to different methods based on the type of the arguments.

Here are examples that use the <TT>+</TT> operator with different types: Unfortunately, this implementation of addition is not commutative. If the integer is the first operand, you get

The problem is, instead of asking the Time object to add an integer, Python is asking an integer to add a Time object, and it doesn&#X2019;t know how to do that. But there is a clever solution for this problem: the special method, which stands for &#X201C;right-side add.&#X201D; This method is invoked when a Time object appears on the right side of the <TT>+</TT> operator. Here&#X2019;s the definition:

And here&#X2019;s how it&#X2019;s used:

Exercise 5
Write an <TT>add</TT> method for Points that works with either a Point object or a tuple: 


 * If the second operand is a Point, the method should return a new Point whose <I>x</I> coordinate is the sum of the <I>x</I> coordinates of the operands, and likewise for the <I>y</I> coordinates.
 * If the second operand is a tuple, the method should add the first element of the tuple to the <I>x</I> coordinate and the second element to the <I>y</I> coordinate, and return a new Point with the result. 

Polymorphism
Type-based dispatch is useful when it is necessary, but (fortunately) it is not always necessary. Often you can avoid it by writing functions that work correctly for arguments with different types.

Many of the functions we wrote for strings will actually work for any kind of sequence. For example, in Section&#XA0;11.1 we used <TT>histogram</TT> to count the number of times each letter appears in a word. This function also works for lists, tuples, and even dictionaries, as long as the elements of <TT>s</TT> are hashable, so they can be used as keys in <TT>d</TT>. Functions that can work with several types are called polymorphic. Polymorphism can facilitate code reuse. For example, the built-in function <TT>sum</TT>, which adds the elements of a sequence, works as long as the elements of the sequence support addition.

Since Time objects provide an <TT>add</TT> method, they work with <TT>sum</TT>: In general, if all of the operations inside a function work with a given type, then the function works with that type.

The best kind of polymorphism is the unintentional kind, where you discover that a function you already wrote can be applied to a type you never planned for.

Debugging
It is legal to add attributes to objects at any point in the execution of a program, but if you are a stickler for type theory, it is a dubious practice to have objects of the same type with different attribute sets. It is usually a good idea to initialize all of an objects attributes in the init method.

If you are not sure whether an object has a particular attribute, you can use the built-in function <TT>hasattr</TT> (see Section&#XA0;15.7).

Another way to access the attributes of an object is through the special attribute, which is a dictionary that maps attribute names (as strings) and values: For purposes of debugging, you might find it useful to keep this function handy: traverses the items in the object&#X2019;s dictionary and prints each attribute name and its corresponding value.

The built-in function <TT>getattr</TT> takes an object and an attribute name (as a string) and returns the attribute&#X2019;s value.

Glossary
<DL CLASS="description"><DT CLASS="dt-description">object-oriented language:</DT><DD CLASS="dd-description"> A language that provides features, such as user-defined classes and method syntax, that facilitate object-oriented programming. </DD><DT CLASS="dt-description">object-oriented programming:</DT><DD CLASS="dd-description"> A style of programming in which data and the operations that manipulate it are organized into classes and methods. </DD><DT CLASS="dt-description">method:</DT><DD CLASS="dd-description"> A function that is defined inside a class definition and is invoked on instances of that class. </DD><DT CLASS="dt-description">subject:</DT><DD CLASS="dd-description"> The object a method is invoked on. </DD><DT CLASS="dt-description">operator overloading:</DT><DD CLASS="dd-description"> Changing the behavior of an operator like <TT>+</TT> so it works with a user-defined type.

</DD><DT CLASS="dt-description">type-based dispatch:</DT><DD CLASS="dd-description"> A programming pattern that checks the type of an operand and invokes different functions for different types. </DD><DT CLASS="dt-description">polymorphic:</DT><DD CLASS="dd-description"> Pertaining to a function that can work with more than one type.

</DD></DL>

Exercise 6
This exercise is a cautionary tale about one of the most common, and difficult to find, errors in Python.


 * Write a definition for a class named <TT>Kangaroo</TT> with the following methods:


 * An  method that initializes an attribute named   to an empty list.


 * A method named  that takes an object of any type and adds it to  .


 * A  method that returns a string representation of the Kangaroo object and the contents of the pouch.

Test your code by creating two <TT>Kangaroo</TT> objects, assigning them to variables named <TT>kanga</TT> and <TT>roo</TT>, and then adding <TT>roo</TT> to the contents of <TT>kanga</TT>&#X2019;s pouch.


 * Download <TT>thinkpython.com/code/BadKangaroo.py</TT>. It contains a solution to the previous problem with one big, nasty bug. Find and fix the bug.

If you get stuck, you can download <TT>thinkpython.com/code/GoodKangaroo.py</TT> , which explains the problem and demonstrates a solution.

Exercise 7
''Visual is a Python module that provides 3-D graphics. It is not always included in a Python installation, so you might have to install it from your software repository or, if it&#X2019;s not there, from '<TT>vpython.org</TT>'.''

''The following example creates a 3-D space that is 256 units wide, long and high, and sets the &#X201C;center&#X201D; to be the point '(128, 128, 128)'. Then it draws a blue sphere.''

<TT>color</TT> is an RGB tuple; that is, the elements are Red-Green-Blue levels between 0.0 and 1.0 (see <TT>wikipedia.org/wiki/RGB_color_model</TT>).

''If you run this code, you should see a window with a black background and a blue sphere. If you drag the middle button up and down, you can zoom in and out. You can also rotate the scene by dragging the right button, but with only one sphere in the world, it is hard to tell the difference.''

The following loop creates a cube of spheres:


 * Put this code in a script and make sure it works for you.


 * Modify the program so that each sphere in the cube has the color that corresponds to its position in RGB space. Notice that the coordinates are in the range 0&#X2013;255, but the RGB tuples are in the range 0.0&#X2013;1.0.


 * Download <TT>thinkpython.com/code/color_list.py</TT> and use the function  to generate a list of the available colors on your system, their names and RGB values. For each named color draw a sphere in the position that corresponds to its RGB values.

You can see my solution at <TT>thinkpython.com/code/color_space.py</TT>.