Object Oriented Python - Advanced Features

Advertisements

Previous Page
Next Page  

In this we will look into some of the advanced features which Python provide

Core Syntax in our Class design

In this we will look onto, how Python allows us to take advantage of operators in our classes. Python is largely objects and methods call on objects and this even goes on even when its hidden by some convenient syntax.

>>> var1 = 'Hello'
>>> var2 = ' World!'
>>> var1 + var2
'Hello World!'
>>>
>>> var1.__add__(var2)
'Hello World!'
>>> num1 = 45
>>> num2 = 60
>>> num1.__add__(num2)
105
>>> var3 = ['a', 'b']
>>> var4 = ['hello', ' John']
>>> var3.__add__(var4)
['a', 'b', 'hello', ' John']

So if we have to add magic method __add__ to our own classes, could we do that too. Let’s try to do that.

We have a class called Sumlist which has a contructor __init__ which takes list as an argument called my_list.

class SumList(object):
   def __init__(self, my_list):
      self.mylist = my_list
   def __add__(self, other):
     new_list = [ x + y for x, y in zip(self.mylist, other.mylist)]

     return SumList(new_list)
   
   def __repr__(self):
      return str(self.mylist)

aa = SumList([3,6, 9, 12, 15])

bb = SumList([100, 200, 300, 400, 500])
cc = aa + bb # aa.__add__(bb)
print(cc) # should gives us a list ([103, 206, 309, 412, 515])

Output

[103, 206, 309, 412, 515]

But there are many methods which are internally managed by others magic methods. Below are some of them,

'abc' in var # var.__contains__('abc')
var == 'abc' # var.__eq__('abc')
var[1] # var.__getitem__(1)
var[1:3] # var.__getslice__(1, 3)
len(var) # var.__len__()
print(var) # var.__repr__()

Inheriting From built-in types

Classes can also inherit from built-in types this means inherits from any built-in and take advantage of all the functionality found there.

In below example we are inheriting from dictionary but then we are implementing one of its method __setitem__. This (setitem) is invoked when we set key and value in the dictionary. As this is a magic method, this will be called implicitly.

class MyDict(dict):

   def __setitem__(self, key, val):
      print('setting a key and value!')
      dict.__setitem__(self, key, val)

dd = MyDict()
dd['a'] = 10
dd['b'] = 20

for key in dd.keys():
   print('{0} = {1}'.format(key, dd[key]))

Output

setting a key and value!
setting a key and value!
a = 10
b = 20

Let’s extend our previous example, below we have called two magic methods called __getitem__ and __setitem__ better invoked when we deal with list index.

# Mylist inherits from 'list' object but indexes from 1 instead for 0!
class Mylist(list): # inherits from list
   def __getitem__(self, index):
      if index == 0:
         raise IndexError
      if index > 0:
         index = index - 1
         return list.__getitem__(self, index) # this method is called when

# we access a value with subscript like x[1]
   def __setitem__(self, index, value):
      if index == 0:
         raise IndexError
      if index > 0:
      index = index - 1
      list.__setitem__(self, index, value)

x = Mylist(['a', 'b', 'c']) # __init__() inherited from builtin list

print(x) # __repr__() inherited from builtin list

x.append('HELLO'); # append() inherited from builtin list

print(x[1]) # 'a' (Mylist.__getitem__ cutomizes list superclass
               # method. index is 1, but reflects 0!

print (x[4]) # 'HELLO' (index is 4 but reflects 3!

Output

['a', 'b', 'c']
a
HELLO

In above example, we set a three item list in Mylist and implicitly __init__ method is called and when we print the element x, we get the three item list ([‘a’,’b’,’c’]). Then we append another element to this list. Later we ask for index 1 and index 4. But if you see the output, we are getting element from the (index-1) what we have asked for. As we know list indexing start from 0 but here the indexing start from 1 (that’s why we are getting the first item of the list).

Naming Conventions

In this we will look into names we’ll used for variables especially private variables and conventions used by Python programmers worldwide. Although variables are designated as private but there is not privacy in Python and this by design. Like any other well documented languages, Python has naming and style conventions that it promote although it doesn’t enforce them. There is a style guide written by “Guido van Rossum” the originator of Python, that describe the best practices and use of name and is called PEP8. Here is the link for this, https://www.python.org/dev/peps/pep-0008/

PEP stands for Python enhancement proposal and is a series of documentation that distributed among the Python community to discuss proposed changes. For example it is recommended all,

  • Module names − all_lower_case
  • Class names and exception names − CamelCase
  • Global and local names − all_lower_case
  • Functions and method names − all_lower_case
  • Constants − ALL_UPPER_CASE

These are just the recommendation, you can vary if you like. But as most of the developers follows these recommendation so might me your code is less readable.

Why conform to convention?

We can follow the PEP recommendation we it allows us to get,

  • More familiar to the vast majority of developers
  • Clearer to most readers of your code.
  • Will match style of other contributers who work on same code base.
  • Mark of a professional software developers
  • Everyone will accept you.

Variable Naming − ‘Public’ and ‘Private’

In Python, when we are dealing with modules and classes, we designate some variables or attribute as private. In Python, there is no existence of “Private” instance variable which cannot be accessed except inside an object. Private simply means they are simply not intended to be used by the users of the code instead they are intended to be used internally. In general, a convention is being followed by most Python developers i.e. a name prefixed with an underscore for example. _attrval (example below) should be treated as a non-public part of the API or any Python code, whether it is a function, a method or a data member. Below is the naming convention we follow,

  • Public attributes or variables (intended to be used by the importer of this module or user of this class) −regular_lower_case

  • Private attributes or variables (internal use by the module or class) −_single_leading_underscore

  • Private attributes that shouldn’t be subclassed −__double_leading_underscore

  • Magic attributes −__double_underscores__(use them, don’t create them)

class GetSet(object):

   instance_count = 0 # public
   
   __mangled_name = 'no privacy!' # special variable

   def __init__(self, value):
      self._attrval = value # _attrval is for internal use only
      GetSet.instance_count += 1

   @property
   def var(self):
      print('Getting the "var" attribute')
      return self._attrval

   @var.setter
   def var(self, value):
      print('setting the "var" attribute')
      self._attrval = value

   @var.deleter
   def var(self):
      print('deleting the "var" attribute')
      self._attrval = None

cc = GetSet(5)
cc.var = 10 # public name
print(cc._attrval)
print(cc._GetSet__mangled_name)

Output

setting the "var" attribute
10
no privacy!
Advertisements