PortugueseLemmatizer.py

import nltk
from nltk.corpus import stopwords
from nltk.stem import RSLPStemmer
from nltk.stem.wordnet import WordNetLemmatizer

# Download list
!wget https://github.com/michmech/lemmatization-lists/raw/master/lemmatization-pt.txt -q

nltk.download('stopwords')
nltk.download('rslp')

stopwordspt = set(stopwords.words("portuguese"))
stopwordsen = set(stopwords.words("english"))

ptstemmer = RSLPStemmer()
enstemmer = WordNetLemmatizer()

# Convert to dictionary
lmztpt = {}
dic = open("lemmatization-pt.txt")
for line in dic:
  txt = line.split()
  lmztpt[txt[1]] = txt[0]

# Lemmatize wherever possible
def PortugueseMess(word):
  if word in lmztpt.keys():
    return lmztpt.get(word)
  else:
    return ptstemmer.stem(word)