Topics
Helps to abstract out algorithms with a Strategy Pattern. Instead of calling the algorithm directly, we will wrap it with a class and call a class member function (say algo_1.execute()) to get the output. This allows using different implementations, by simply passing a different class/strategy: algo_2.execute().
The caller function takes input the “strategy” and “executes” it.
import spacy
"""Step 1: move code to a new class"""
class SpacyLemmatizer(BaseLemmatizer):
def __init__(self, modelName):
self.spacyModel = spacy.load(modelName)
def get_lemmas(self, text):
return [token.lemma_ for token in self.spacyModel(text)]
class Process:
"""Step 2: add new class as a member variable"""
def __init__(self, articleLoader: BaseLoader, lemmatizer: BaseLemmatizer):
self.articleLoader = articleLoader
self.lemmatizer = lemmatizer
def run(self):
articles = self.articleLoader.get_articles()
for text in articles['text']:
"""Step 3: refactor usage"""
lemmas = self.lemmatizer.get_lemmas(text)
def main(param: str):
"""Step 4: create a lemmatizer at usage time"""
process = Process(
articleLoader=CSVArticleLoader(filePath='articles.csv')
lemmatizer=SpacyLemmatizer(modelName='en_core_web_sm')
)
process.run()
if __name__ == "__main__":
typer.run(main)If we decide that you would like to use StanfordNLP’s lemmatizer instead of Spacy’s, then Process doesn’t need to be changed. Simply create a new class StanfordNLPLemmatizer with same method get_lemmas, but different implementation.