基于正则表达式 python 规则的 eliza 实现

Question

import random
import re

rules=[["(.*)hello(.*)",["Hi there. Please state your problem"]],
           ["(.*)name(.*)",["Great, good to know","I am not interested 
        in names"]],
  ["(.*)sorry(.*)",["please don't apologize","Apologies are not necessary","What feelings you have when you apologize?"]],
 ["(.*)",["Very interesting","I am not sure I understand you fully","Please continue",
         "Do you feel strongly about discussing such things?","\\2"]]]

grammar = {
"am": "are",
"was": "were",
"i": "you",
"i'd": "you would",
"i've": "you have",
"i'll": "you will",
"my": "your",
"are": "am",
"you've": "I have",
"you'll": "I will",
"your": "my",
"yours": "mine",
"you": "me",
"me": "you"
}

def correction(word):
character=word.lower().split()
for i, j in enumerate(character):
    if j in grammar:
        character[i]=grammar[j]
return " ".join(character)

def test(sentence):
for pattern, message in rules:
    match=re.match(pattern,sentence.rstrip(".!"))
    if match:
        response = random.choice(message)
        temp = " " + correction(match.group())
        response2 = re.sub(r"\\2",temp,response)
        return response2
      else:
        recall=random.choice(message)
        return recall



while True:
sentence =input("You: ")
print("JBot: " + test(sentence))

    if sentence == "quit":
    break

在这个简单的eliza 实现中，有一个名为规则的列表，其中包含一组模式和相应的响应。如果模式匹配或输入的任何其他内容不在规则（最后一条规则）中，则此代码应该得到随机响应。

代码现在只输出，"Hi, there. Please state your problem" 用于所有输入语句。任何帮助为什么会发生这种情况？？

如果你输入了一个在规则中匹配的句子，那么它会回复相应的响应。假设有如下规则：'(.*)are like(.*)', ["What resembelence do you see between {0} and {1}?"]] ，如果输入是 "Cats are like dogs" 响应应该类似于，什么你觉得猫和狗有相似之处吗？因此，它会从匹配中取出一组并放入相应的响应中。

Answer 1

我已经修复了你的代码，现在应该可以正常工作了：

注意事项：

• test 函数循环中的else 会在每次迭代时退出循环，因此您将无法浏览所有规则、语法。我把它放在for 之后，这将强制首先检查每个规则，然后再选择默认的随机选择的答案。

代码：

import random
import re

rules=[["(.*)hello(.*)",["Hi there. Please state your problem"]],
           ["(.*)name(.*)",["Great, good to know","I am not interested in names"]],
  ["(.*)sorry(.*)",["please don't apologize","Apologies are not necessary","What feelings you have when you apologize?"]],
 ["(.*)",["Very interesting","I am not sure I understand you fully","Please continue",
         "Do you feel strongly about discussing such things?","\\2"]]]

grammar = {
"am": "are",
"was": "were",
"i": "you",
"i'd": "you would",
"i've": "you have",
"i'll": "you will",
"my": "your",
"are": "am",
"you've": "I have",
"you'll": "I will",
"your": "my",
"yours": "mine",
"you": "me",
"me": "you"
}

def correction(word):
  character=word.lower().split()
  for i, j in enumerate(character):
      if j in grammar:
          character[i]=grammar[j]
  return " ".join(character)

def test(sentence):
  for pattern, message in rules:
      match=re.match(pattern,sentence.rstrip(".!"))
      if match:
          response = random.choice(message)
          temp = " " + correction(match.group())
          response2 = re.sub(r"\\2",temp,response)
          return response2
  recall=random.choice(random.choice([r[1] for r in rules]))
  return recall



while True:
  sentence =input("You: ")
  print("JBot: " + test(sentence))
  if sentence == "quit":
        break

输出：

You: 'hello'
JBot: Hi there. Please state your problem
You: "i don't have a name"
JBot: Great, good to know
You: "i am so sorry"
JBot: What feelings you have when you apologize?
You: "help me"
JBot: Do you feel strongly about discussing such things?

---

输出非常有趣，真的让我很开心。

---