Leveraging Classes

Software Engineering Principles in Python

Adam Spannbauer

Machine Learning Engineer at Eastman

Extending Document class

class Document:
    def __init__(self, text):
        self.text = text

Tokenize Document

Current document class

class Document:
    def __init__(self, text):
        self.text = text

Tokenize Document init

Revising init

class Document:
    def __init__(self, text):
        self.text = text
        self.tokens = self._tokenize()


doc = Document('test doc')
print(doc.tokens)

['test', 'doc']

Adding _tokenize() method

# Import function to perform tokenization
from .token_utils import tokenize

class Document:
    def __init__(self, text, token_regex=r'[a-zA-Z]+'):
        self.text = text
        self.tokens = self._tokenize()

    def _tokenize(self):
        return tokenize(self.text)

Non-public methods

Tokenize Document init

Private Property Sign

The risks of non-public methods

Lack of documentation
Unpredictability

Private Property Sign

Let's Practice

Software Engineering Principles in Python

Leveraging Classes

Extending Document class

Current document class

Revising __init__

Adding _tokenize() method

Non-public methods

The risks of non-public methods

Let's Practice

Revising init