使用NLTK从停用词算法中删除重音词

Question

I'm trying to develop a simple algorithm in Python to remove stop words from a text, but I'm having problems with words that have accents. I'm using the following code:

import io
from nltk.corpus import stopwords
from nltk.tokenize import word_tokenize
from unicodedata import normalize
import sys

reload(sys)
sys.setdefaultencoding('utf8')

stop_words = set(stopwords.words('portuguese'))
file1 = open("C:\Users\Desktop\Test.txt")
print("File open")
line = file1.read()
words = line.split()
#convert the words to lower case
words = [word.lower() for word in words]
print("Running!")
for r in words:
    if r not in stop_words:
            appendFile = open('finalText.txt','a')
            appendFile.writelines(" "+r)
            appendFile.close()

print("Finished!")

When running the code with the following test file:

E É Á A O Ó U Ú

I have this output:

 É Á Ó Ú

It doesn't seem to recognize accentuated words, and using "setdefaultencoding" for utf-8 does not work, does anyone knows of a solution I can use to solve this problem?

Answer 1

It is not an encoding or accent problem. These are simply words that are not in the list:

from nltk.corpus import stopwords
stop_words = set(stopwords.words('portuguese'))

print(" ".join([w for w in stop_words if len(w) == 1]))
# >>> e à o a
# -> does not contain á é ó ú

print("À".lower() in stop_words)
# >>> True

You can just add words to the set ( stop_words.add("é") ) if you need to.