我正试着教我的女朋友如何构建聊天机器人。
我们学习了基于规则的聊天机器人基础的代码学院课程，并试图将其集成到Twilio中(在Facebook信使出现问题后(。
from flask import Flask, request
from twilio.twiml.messaging_response import MessagingResponse
app = Flask(__name__)
@app.route("/")
def hello():
return "Hello, World!"
@app.route("/sms", methods=['POST'])
def sms_reply():
"""Respond to incoming calls with a simple text message."""
# Fetch the message
msg = request.form.get('Body')
# Create reply
if msg == "bunny" or "michael":
resp = MessagingResponse()
resp.message("You said: {}".format(msg))
return str(resp)
if __name__ == "__main__":
app.run(debug=True)
在Twilio中使用此代码可以很好地通过WhatsApp解析如上所述的简单内容。
然而，将其集成到这个聊天机器人中是令人困惑的。您如何使用twilio发送和接收消息以使下面的代码正常工作？我玩了一场游戏，但我不能把它发挥作用。
import re
import random
class BunnyBot:
# potential negative responses
negative_responses = ("no", "nope", "nah", "naw", "not a chance", "sorry")
# keywords for exiting the conversation
exit_commands = ("quit", "pause", "exit", "goodbye", "bye", "later", "goodbye", "got to go")
# random starter questions
random_questions = (
"How's your day going Mummy Bunny? ",
"What do you do when you leave the house? ",
"What did you have to eat today? I had some fresh weeds. ",
"Are there other bunnies like me? ",
"Do you want to come hang out with me in the garden? ",
"Would you like to help me dig a hole? ",
"Do you have any carrots today? "
)
def __init__(self):
self.bunnybabble = {'hows_your_day_intent': '.*s*your day.*','carrots_intent': '.s*carrot.*', 'cubed_intent': '.*.*(d+)'}
# Define .greet() below:
def greet(self):
self.name = input("Hi there. Who I am talking to?")
will_chat = input(f"Hey {self.name}! Do you want to chat with me? ")
if will_chat in self.negative_responses:
print ("No worries, see you later! xxx")
return
self.chat()
# Define .make_exit() here:
def make_exit(self, reply):
for word in self.exit_commands:
if word in reply:
print ("Bye for now! I hope we can chat later xxx")
return True
# Define .chat() next:
def chat(self):
reply = input(random.choice(self.random_questions)).lower()
while not self.make_exit(reply):
reply = input(self.match_reply(reply))
# Define .match_reply() below:
def match_reply(self, reply):
found = False
for intent, regPattern in self.bunnybabble.items():
found_match = re.findall(regPattern, reply)
if found_match and intent == 'hows_your_day_intent':
return self.hows_your_day_intent()
elif found_match and intent == 'carrots_intent':
return self.carrots_intent()
elif found_match and intent == 'cubed_intent':
return self.cubed_intent(found_match[0])
if not found:
return self.no_match_intent()
def hows_your_day_intent(self):
responses = ("Pretty relaxing, I've just been napping. ", "Great! I dug a tunnel, do you want to see? ", "Good, but I'm kind of hungry, do you have any snacks? ", "It's been alright, but I'm a bit lonely and would love a pat. ", "So good, just been eating some curtains! ")
return random.choice(responses)
def carrots_intent(self):
responses = ("I love carrots! Can I have one now? ", "Unfortunately I never get a full carrots, only the peel. ", "Carrots are good but so is apple! Do you have a snack for me? ")
return random.choice(responses)
def cubed_intent(self, number):
number = int(number)
cubed_number = number * number * number
return (f"The cube of {number} is {cubed_number}. Who knew bunnies were good at maths? ")
def plans_for_today_intent(self):
responses = ("I might see if my pigeon friends are around! ", "Maybe hang out with Mummy Bunny if she's doing some gardening. ", "Probably watch some TV with Mummy Bunny and Nicole. ", "I'm just going to hang around until some weeds appear for me to eat. ")
return random.choice(responses)
def love_me_intent(self):
responses = ("Of course I do, you're the best Mummy Bunny ever! ", "I love you so much, no one else gets me weeds like you do. ", "Yes I do, you give the best pats and jaw rubs. ", "Absolutely, thanks for always looking after me. ")
return random.choice(responses)
def my_name_intent(self):
responses = ("Honestly, I'm not sure. It's either Bunny, Fluffy Bun or Heather. It's a bit confusing. ", "Well, someone once mentioned Flat Bunny which got me a little worried. ", "Mummy Bunny calls me Bunny Bun a lot? ")
return random.choice(responses)
def favourite_intent(self):
responses = ("Mummy Bunny obviously. She looks after me. ", "Nicole is pretty cool, after all, I am technically her rabbit. She can be pretty loud though. ", "Natalie is nice to me, but I never really see her anymore which is sad. ")
return random.choice(responses)
def no_match_intent(self):
responses = ("Please tell me more. ", "Tell me more! ", "Why do you say that? ", "I see. Can you elaborate? ", "Interesting. Can you tell me more? ", "I see. How do you think? ", "Why? ", "How do you think I feel when you say that? ")
return random.choice(responses)
my_bot = BunnyBot()
my_bot.greet()
我认为这与使用twillio消息响应有关，但我不明白它是如何记住过去说过的话的？
更新
这是新的代码，尝试使用基于这个youtube教程的Twilio在这里：
https://www.youtube.com/watch?v=EeUdel2AJ5g
from flask import Flask, request
from twilio.twiml.messaging_response import MessagingResponse
import re
import random
def send(msg):
resp = MessagingResponse()
resp.message(format(msg))
return str(resp)
def recv():
return request.form.get('Body')
app = Flask(__name__)
@app.route("/")
def hello():
return "Hello, World!"
@app.route("/sms", methods=['POST'])
def sms_reply():
"""Respond to incoming calls with a simple text message."""
# Fetch the message
msg = request.form.get('Body')
# Create reply
if msg == "bunny" or "michael":
resp = MessagingResponse()
resp.message("You said: {}".format(msg))
return str(resp)
class BunnyBot:
# potential negative responses
negative_responses = ("no", "nope", "nah", "naw", "not a chance", "sorry")
# keywords for exiting the conversation
exit_commands = ("quit", "pause", "exit", "goodbye",
"bye", "later", "goodbye", "got to go")
# random starter questions
random_questions = (
"How's your day going Mummy Bunny? ",
"What do you do when you leave the house? ",
"What did you have to eat today? I had some fresh weeds. ",
"Are there other bunnies like me? ",
"Do you want to come hang out with me in the garden? ",
"Would you like to help me dig a hole? ",
"Do you have any carrots today? "
)
def __init__(self):
self.bunnybabble = {'hows_your_day_intent': '.*s*your day.*',
'carrots_intent': '.s*carrot.*', 'cubed_intent': '.*.*(d+)'}
# Define .greet() below:
def greet(self):
self.name = recv("Hi there. Who I am talking to?")
will_chat = recv(f"Hey {self.name}! Do you want to chat with me? ")
if will_chat in self.negative_responses:
send("No worries, see you later! xxx")
return
self.chat()
# Define .make_exit() here:
def make_exit(self, reply):
for word in self.exit_commands:
if word in reply:
send("Bye for now! I hope we can chat later xxx")
return True
# Define .chat() next:
def chat(self):
reply = recv(random.choice(self.random_questions)).lower()
while not self.make_exit(reply):
reply = recv(self.match_reply(reply))
# Define .match_reply() below:
def match_reply(self, reply):
found = False
for intent, regPattern in self.bunnybabble.items():
found_match = re.findall(regPattern, reply)
if found_match and intent == 'hows_your_day_intent':
return self.hows_your_day_intent()
elif found_match and intent == 'carrots_intent':
return self.carrots_intent()
elif found_match and intent == 'cubed_intent':
return self.cubed_intent(found_match[0])
if not found:
return self.no_match_intent()
def hows_your_day_intent(self):
responses = ("Pretty relaxing, I've just been napping. ", "Great! I dug a tunnel, do you want to see? ", "Good, but I'm kind of hungry, do you have any snacks? ",
"It's been alright, but I'm a bit lonely and would love a pat. ", "So good, just been eating some curtains! ")
return random.choice(responses)
def carrots_intent(self):
responses = ("I love carrots! Can I have one now? ", "Unfortunately I never get a full carrots, only the peel. ",
"Carrots are good but so is apple! Do you have a snack for me? ")
return random.choice(responses)
def cubed_intent(self, number):
number = int(number)
cubed_number = number * number * number
return (f"The cube of {number} is {cubed_number}. Who knew bunnies were good at maths? ")
def plans_for_today_intent(self):
responses = ("I might see if my pigeon friends are around! ", "Maybe hang out with Mummy Bunny if she's doing some gardening. ",
"Probably watch some TV with Mummy Bunny and Nicole. ", "I'm just going to hang around until some weeds appear for me to eat. ")
return random.choice(responses)
def love_me_intent(self):
responses = ("Of course I do, you're the best Mummy Bunny ever! ", "I love you so much, no one else gets me weeds like you do. ",
"Yes I do, you give the best pats and jaw rubs. ", "Absolutely, thanks for always looking after me. ")
return random.choice(responses)
def my_name_intent(self):
responses = ("Honestly, I'm not sure. It's either Bunny, Fluffy Bun or Heather. It's a bit confusing. ",
"Well, someone once mentioned Flat Bunny which got me a little worried. ", "Mummy Bunny calls me Bunny Bun a lot? ")
return random.choice(responses)
def favourite_intent(self):
responses = ("Mummy Bunny obviously. She looks after me. ", "Nicole is pretty cool, after all, I am technically her rabbit. She can be pretty loud though. ",
"Natalie is nice to me, but I never really see her anymore which is sad. ")
return random.choice(responses)
def no_match_intent(self):
responses = ("Please tell me more. ", "Tell me more! ", "Why do you say that? ", "I see. Can you elaborate? ",
"Interesting. Can you tell me more? ", "I see. How do you think? ", "Why? ", "How do you think I feel when you say that? ")
return random.choice(responses)
if __name__ == "__main__":
app.run(debug=True)
尝试使用Ngrok，但还没有正常工作-有线索吗？
您就快到了。
TL；DR
你的机器人在命令行上运行良好。唯一的问题是将其连接到twilio
逐步解决方案
- 使用twilio代码创建两个函数。send和recv函数即
def send(msg):
resp = MessagingResponse()
resp.message(format(msg))
return str(resp)
def recv():
return request.form.get('Body')
- 测试这些函数，如果需要，可以修改。因为我没有账户，所以没有测试
- 将打印更改为
send，并输入为send和recv即
输入
msg = input("Hi there. Who I am talking to?")
# change it to
send("Hi there. Who I am talking to?")
msg = recv()
打印
print ("No worries, see you later!")
#change this to
send("No worries, see you later!")
这样你就会有工作聊天机器人
为什么要使用send和recv功能
它将帮助你扩展或将你的机器人移植到其他平台，如telegram或twitter或其他提供商，如gupshup.io