Text pre-processing

alp_data_prep.py

import nltk
from nltk.tokenize import word_tokenize
from nltk.text import Text

# read in text data
file = open("crawl-for-parallel-corpora/DataSet/luganda.txt", "r") 
raw = file.read()

# tokenize
tokens = word_tokenize(raw)

# remove punctuation, numbers and make everything lowercase
tokens = [word.lower() for word in tokens if word.isalpha()]

# write output to file
f = open("luganda", "w")
f.write(' '.join(tokens))