file

def text_to_word_list(text, remove_polish_letters):
    ''' Pre process and convert texts to a list of words 
    method inspired by method from eliorc github repo: https://github.com/eliorc/Medium/blob/master/MaLSTM.ipynb'''
    text = remove_polish_letters(text)
    text = str(text)
    text = text.lower()

    # Clean the text
    text = sub(r"[^A-Za-z0-9^,!?.\/'+]", " ", text)
    text = sub(r"\+", " plus ", text)
    text = sub(r",", " ", text)
    text = sub(r"\.", " ", text)
    text = sub(r"!", " ! ", text)
    text = sub(r"\?", " ? ", text)
    text = sub(r"'", " ", text)
    text = sub(r":", " : ", text)
    text = sub(r"\s{2,}", " ", text)

    text = text.split()

    return text