PorterStemmer Stemmer没有词干，除非在字符串结尾（在Python中）

Question

我正在创建一个谈论国际象棋的机器人，现在我正在进行预处理。问题是，当Python程序阻塞字符串中的单词时，除非单词是最后一个单词，否则它将不会执行任何操作。那是输出。我所期待的是，词干分析器阻止用户输入的句子中的每个单词。我做错了什么？这是代码：

import re
from nltk.stem import PorterStemmer
port = PorterStemmer()

print("Hello! I am the ChessBot. I will talk about chess, but can only talk     about chess. Nothing else. I can't even understand a greeting! Now, initiate this conversation!")

while True:
    userinput = input('> ')
    userinput = re.sub('(\?|\.|!|,)', '', userinput)
    userinput = str.lower(userinput)
    userinput = userinput.split('\s')
    userinput = " ".join([port.stem(word) for word in userinput])
    print(userinput)