if 'yes' in text.lower():
return False
elif 'no' in text.lower():
return True
else:
print('Please type either "Yes" or "No"')
return get_feedback()
print("Type something to begin...")
# The following loop will execute each time the user enters input
while True:
try:
input_statement = bot.input.process_input_statement()
statement, response = bot.generate_response(input_statement, CONVERSATION_ID)
bot.output.process_response(response)
print('\n Is "{}" a coherent response to "{}"? \n'.format(response, input_statement))
if get_feedback():
print("please input the correct one")
response1 = bot.input.process_input_statement()
bot.learn_response(response1, input_statement)
bot.storage.add_to_conversation(CONVERSATION_ID, statement, response1)
print("Responses added to bot!")
# Press ctrl-c or ctrl-d on the keyboard to exit
except (KeyboardInterrupt, EOFError, SystemExit):
break
table = str.maketrans('', '', string.punctuation)
stripped = [w.translate(table) for w in words]
for w in stripped:
if w == "":
stripped.remove(w)
elif w == " ":
stripped.remove(w)
filtered_sen_pun = ' '.join(stripped)
print('Punc :', filtered_sen_pun)
# Lowering the case
#filtered_sent=filtered_sen_pun.lower()
input_statement = Statement(filtered_sen_pun)
print('-' * 100)
print('input_statement ==>', input_statement)
response = bot.generate_response(input_statement)
print('=' * 100)
print("Chatbot :", response.text)
print('\n Is "{}" a coherent response to "{}"? \n'.format(
response.text, input_statement.text))
if get_feedback():
print('please input the correct one')
correct_response = Statement(text=input())
bot.learn_response(correct_response, input_statement)
print('Responses added to bot!')
# Press ctrl-c or ctrl-d on the keyboard to exit
except (KeyboardInterrupt, EOFError, SystemExit):
break
if 'yes' in text.lower():
return True
elif 'no' in text.lower():
return False
else:
print('Please type either "Yes" or "No"')
return get_feedback()
print('Type something to begin...')
# The following loop will execute each time the user enters input
while True:
try:
input_statement = bot.input.process_input()
statement, response = bot.generate_response(
input_statement,
CONVERSATION
)
print('\n Is "{}" this a coherent response to "{}"? \n'.format(response, input_statement))
if get_feedback():
bot.learn_response(CONVERSATION, response, input_statement)
bot.output.process_response(response)
# Press ctrl-c or ctrl-d on the keyboard to exit
except (KeyboardInterrupt, EOFError, SystemExit):
break
class NLG_unit(object):
""" This object contains tools to answer in natural language any message non-Fib related
Attributes:
chatterbot_bot(:class:`chatterbot.ChatBot`): chatterbot bot to process the non-fib-Related questions
train_queue(:obj:`dict`): dictionary that act as a queue of training concepts
conversation_id(:obj:`int` or None): Identifier that keeps track of the diferent amounts of conversations learnt
"""
def __init__(self):
self.chatterbot_bot = ChatBot(
"Fibot",
storage_adapter="chatterbot.storage.SQLStorageAdapter",
)
self.train_queue = {}
self.conversation_id = None
"""
This function load the learnt model for chatterbot
"""
def load(self):
self.conversation_id = self.chatterbot_bot.storage.create_conversation()
"""
Parameters:
chat_id (:obj:`str`): chat_id of the person to get feedback from
correct (:obj:`bool`): tells whether the prediction was correct_statement
correct_statement (:obj:`str` or None): Correction if there is
This function enables the chatterbot bot to learn if he predicted good and allows
him to learn new responses.
"""
def give_feedback(self, chat_id, correct = True, correct_statement = None):
if not correct:
message = self.train_queue[chat_id]['message']
correct_statement = Statement(correct_statement)
self.chatterbot_bot.learn_response(correct_statement, message)
self.chatterbot_bot.storage.add_to_conversation(self.conversation_id, message, correct_statement)
self.send_message(chat_id, "Aprendido!. Mándame más si quieres, o desactívame usando /train_off")
else:
self.send_message(chat_id, "Bien. Mándame más si quieres, o desactívame usando /train_off")
"""
Parameters:
chat_id (:obj:`str`): chat_id of the person to process the answer for
message (:obj:`str`): message the user introduced to be processed
This function makes the bot predict the following message, and
asks for checks.
"""
def process_answer_training(self, chat_id, message):
print("Query de %s con mensaje %s"%(chat_id, message))
self.send_chat_action(chat_id)
input_statement = Statement(message)
statement, response = self.chatterbot_bot.generate_response(input_statement, self.conversation_id)
self.send_message(chat_id,'Es "{}" una respuesta coherente a "{}"? (Sí/No)'.format(response, message))
self.train_queue[chat_id] = {
'message': input_statement,
'response': response
}
"""
Parameters:
message (:obj:`str`): message the user introduced to be processed
debug (:obj:`bool`): indicates the level of verbosity
This function returns the predicted responses for message
"""
def get_response(self, message, debug = False):
input_statement = Statement(message)
_, response = self.chatterbot_bot.generate_response(input_statement, self.conversation_id)
if debug:
return "%s, confidence = %d"%(response['text'], response['confidence'])
else:
return str(response)
"""
Parameters:
chat_id (:obj:`str`): chat_id of the person to process the answer for
message (:obj:`str`): content of the messages
This function allows nlg class to send messages for the training duties
"""
def send_message(self, chat_id, message):
params = {
'chat_id': chat_id,
'text': message
}
bot_token = os.getenv('FibotTOKEN')
base_url = 'https://api.telegram.org/bot%s/sendMessage'%bot_token
response = requests.get(base_url, params = params)
"""
Parameters:
chat_id (:obj:`str`): chat_id of the person to process the answer for
action (:obj:`str`): action to send of the messages
This function allows nlg class to send chat_action for the training duties
"""
def send_chat_action(self, chat_id, action = ChatAction.TYPING):
params = {
'chat_id': chat_id,
'action': action
}
bot_token = os.getenv('FibotTOKEN')
base_url = 'https://api.telegram.org/bot%s/sendChatAction'%bot_token
response = requests.get(base_url, params = params)
elif 'maybe' in text.lower():
print('ok')
else:
print('Please type either "Yes" or "No"')
return get_feedback()
print("Type something to begin...")
# The following loop will execute each time the user enters input
while True:
try:
input_statement = bot.input.process_input_statement()
statement, response = bot.generate_response(input_statement,
CONVERSATION_ID)
bot.output.process_response(response)
print('\n Is "{}" a coherent response to "{}"? \n'.format(
response, input_statement))
if get_feedback():
print("please input the correct one")
response1 = bot.input.process_input_statement()
bot.learn_response(response1, input_statement)
bot.storage.add_to_conversation(CONVERSATION_ID, statement,
response1)
print("Responses added to bot!")
# Press ctrl-c or ctrl-d on the keyboard to exit
except (KeyboardInterrupt, EOFError, SystemExit):
break
elif 'No' in text:
return False
else:
print('Please type either "Yes" or "No"')
return get_feedback()
print('Type something to begin...')
# The following loop will execute each time the user enters input
while True:
try:
input_statement = bot.input.process_input_statement()
statement, response, confidence = bot.generate_response(input_statement)
print('\n Is "{}" this a coherent response to "{}"? \n'.format(response, input_statement))
if get_feedback():
bot.learn_response(response)
bot.output.process_response(response, confidence)
# Update the conversation history for the bot
# It is important that this happens last, after the learning step
bot.recent_statements.append(
(statement, response, )
)
# Press ctrl-c or ctrl-d on the keyboard to exit
except (KeyboardInterrupt, EOFError, SystemExit):
break
return get_feedback()
print('Type something to begin...')
# The following loop will execute each time the user enters input
while True:
try:
input_statement = bot.input.process_input_statement()
statement, response, confidence = bot.generate_response(
input_statement)
print('\n Is "{}" this a coherent response to "{}"? \n'.format(
response, input_statement))
if get_feedback():
bot.learn_response(response)
bot.output.process_response(response, confidence)
# Update the conversation history for the bot
# It is important that this happens last, after the learning step
bot.recent_statements.append((
statement,
response,
))
# Press ctrl-c or ctrl-d on the keyboard to exit
except (KeyboardInterrupt, EOFError, SystemExit):
break
],
preprocessors=['chatterbot.preprocessors.clean_whitespace'],
)
trainer = ChatterBotCorpusTrainer(chatbot)
trainer.train('./soporte.yml')
levenshtein_distance = LevenshteinDistance()
#synset_distance = SynsetDistance()
#sentiment_comparison = SentimentComparison()
#jaccard_similarity = JaccardSimilarity()
disparate = Statement(
'Te has equivocado') #convertimos una frase en un tipo statement
entradaDelUsuario = "" #variable que contendrá lo que haya escrito el usuario
while entradaDelUsuario != "adios":
entradaDelUsuario = Statement(text=input()) #leemos la entrada del usuario
respuesta = chatbot.generate_response(entradaDelUsuario)
try:
fecha = datetime_parsing(entradaDelUsuario.text)
fecha = fecha[0]
print(fecha[0])
except:
pass
if levenshtein_distance.compare(entradaDelUsuario, disparate) > 0.51:
print('¿Qué debería haber dicho?')
entradaDelUsuarioCorreccion = Statement(input())
chatbot.learn_response(entradaDelUsuarioCorreccion, entradaDelUsuario)
print("He aprendiendo que cuando digas {} debo responder {}".format(
entradaDelUsuario.text, entradaDelUsuarioCorreccion.text))
print(respuesta)
class Chatterbot:
def __init__(self, bot):
self.bot = bot
self.settings = dataIO.load_json('data/chatterbot/settings.json')
self.chatterbot = ChatBot('Athena',
database='LanLogFour',
storage_adapter='chatterbot.storage.'
'MongoDatabaseAdapter',
input_adapter='chatterbot.input.'
'VariableInputTypeAdapter',
output_adapter='chatterbot.output.'
'OutputAdapter',
output_format='text',
logic_adapters=[
{'import_path': 'chatterbot.logic.'
'BestMatch',
'statement_comparison_function':
'chatterbot.comparisons.'
'levenshtein_distance',
'response_selection_method':
'chatterbot.response_selection.'
'get_most_frequent_response'},
'chatterbot.logic.MathematicalEvaluation'
]
)
self.conversations = {}
self.previous_statement = {}
@commands.command(pass_context=True)
async def chat(self, ctx):
"""Reads messages from chat"""
message = ctx.message.content[6:]
conversation_id = self._get_conversation(ctx.message.channel.id)
response = await self._get_response(message, conversation_id)
await self.bot.say(response)
@commands.command()
@checks.serverowner_or_permissions(manage_server=True)
async def chattertraindocs(self):
"""Runs training on the docs"""
self.chatterbot.set_trainer(ListTrainer)
with open('data/chatterbot/doc1.txt') as f:
doc1 = f.readlines()
with open('data/chatterbot/doc2.txt') as f:
doc2 = f.readlines()
docData = doc1 + doc2
docData = [x.strip(',') for x in docData]
docData = [x.strip('\'') for x in docData]
docData = [x.strip('\n') for x in docData]
self.chatterbot.train(docData)
await self.bot.say("Training complete.")
@commands.command()
@checks.serverowner_or_permissions(manage_server=True)
async def chattertrain(self):
self.chatterbot.set_trainer(ChatterBotCorpusTrainer)
self.chatterbot.train("chatterbot.corpus.english")
await self.bot.say("Training complete")
@commands.command(no_pm=True, pass_context=True)
@checks.serverowner_or_permissions(manage_server=True)
async def chatchannel(self, context, channel: discord.Channel = None):
"""
Set the channel to which the bot will sent its continues updates.
Example: [p]chatchannel #talk
"""
if channel:
self.settings['CHANNEL_ID'] = str(channel.id)
dataIO.save_json('data/chatterbot/settings.json', self.settings)
message = 'Channel set to {}'.format(channel.mention)
elif not self.settings['CHANNEL_ID']:
message = 'No Channel set'
else:
channel = discord.utils.get(
self.bot.get_all_channels(), id=self.settings['CHANNEL_ID'])
if channel:
message = 'Current channel is {}'.format(channel.mention)
else:
self.settings['CHANNEL_ID'] = None
message = 'No channel set'
await self.bot.say(message)
@commands.command(pass_context=True)
@checks.serverowner_or_permissions(manage_server=True)
async def chatterignore(self, context, channel: discord.Channel = None):
if channel:
self.settings['BLOCKED_CHANNELS'].append(channel.id)
# TODO: Check if already in list
dataIO.save_json('data/chatterbot/settings.json', self.settings)
message = '{} added to blocked channels.'.format(channel.mention)
elif not self.settings['BLOCKED_CHANNELS']:
message = 'No channels blocked'
# else:
# channel = discord.utils.get(
# self.bot.get_all_channels(),
# id=self.settings['BLOCKED_CHANNELS'])
# if channel:
# message = 'Current channel is {}'.format(channel.mention)
# else:
# self.settings['CHANNEL_ID'] = None
# message = 'No channel set'
await self.bot.say(message)
async def listener(self, message):
if message.author.id == self.bot.user.id:
pass
elif message.content == '':
pass
elif message.content[0] == '!': # Kludge
pass
elif message.mention_everyone is True:
pass
elif message.mentions != []:
pass
elif message.role_mentions != []:
pass
elif message.channel.id == self.settings['CHANNEL_ID']:
conversation_id = self._get_conversation(message.channel.id)
response = await self._get_response(message.content,
conversation_id)
await self.bot.send_message(message.channel, response)
elif message.channel.id not in self.settings['BLOCKED_CHANNELS']:
conversation_id = self._get_conversation(message.channel.id)
input_statement = self.chatterbot.input.process_input_statement(
message.content)
if conversation_id in self.previous_statement:
self.chatterbot.learn_response(
input_statement, self.previous_statement[conversation_id])
self.previous_statement[conversation_id] = input_statement
else:
self.previous_statement[conversation_id] = input_statement
else:
pass
async def _get_response(self, input_item, conversation_id):
input_statement = self.chatterbot.input.process_input_statement(
input_item)
# Preprocess the input statement
for preprocessor in self.chatterbot.preprocessors:
input_statement = preprocessor(self.chatterbot, input_statement)
statement, response = self.chatterbot.generate_response(
input_statement, conversation_id)
# Learn that the user's input was a valid response to the chat bot's
# previous output
if not self.chatterbot.read_only:
if conversation_id in self.previous_statement:
prev = self.previous_statement[conversation_id]
else:
prev = None
self.chatterbot.learn_response(statement, prev)
self.chatterbot.storage.add_to_conversation(conversation_id,
statement, response)
self.previous_statement[conversation_id] = response
# Process the response output with the output adapter
return str(response)
def _get_conversation(self, channel_id):
try:
conversation_id = self.conversations[channel_id]
except KeyError:
conversation_id = self.chatterbot.storage.create_conversation()
self.conversations[channel_id] = conversation_id
return conversation_id
elif 'no' in text.lower():
return False
else:
print('Please type either "Yes" or "No"')
return get_feedback()
print('Type something to begin...')
# The following loop will execute each time the user enters input
while True:
try:
input_statement = Statement(text=input())
response = bot.generate_response(
input_statement
)
print('\n Is "{}" a coherent response to "{}"? \n'.format(
response.text,
input_statement.text
))
if get_feedback():
print('please input the correct one')
correct_response = Statement(text=input())
bot.learn_response(correct_response, input_statement)
print('Responses added to bot!')
# Press ctrl-c or ctrl-d on the keyboard to exit
except (KeyboardInterrupt, EOFError, SystemExit):
break
class AIChatBot:
def __init__(self,
bot_name,
self_training,
low_confidence_threshold,
low_confidence_responses,
random_response=True,
premoderation_callback=None,
db_file=None,
db_uri=None):
"""Create new AIChatBot instance.
Args:
bot_name (str): Bot name.
db_uri (str): Database URI.
db_file (str): Database file (should not be used, left for backward compatibility)
self_training (bool): Will bot train itself.
low_confidence_threshold (float): Confidence threshold, if lower - low_confidence_responses will be used.
low_confidence_responses (list): Responses that will be used if confidence is lower than threshold.
random_response (bool): If true, choose random response instead of the most frequent.
premoderation_callback (function): Premoderation function, set None to disable.
"""
# Backward compatibility
if db_uri is None and db_file is not None:
db_uri = 'sqlite:///{}'.format(db_file)
warnings.warn(
'use db_uri instead of deprecated db_file, it will be removed in the future',
DeprecationWarning)
self.__chatbot = ChatBot(
bot_name,
logic_adapters=['chatterbot.logic.BestMatch'],
response_selection_method=get_random_response
if random_response else get_most_frequent_response,
filters=['chatterbot.filters.RepetitiveResponseFilter'],
database_uri=db_uri)
self.__conversation_id = self.__chatbot.storage.create_conversation()
self.previous_response = None
self.__self_training = self_training
self.__low_confidence_threshold = low_confidence_threshold
self.__low_confidence_responses = low_confidence_responses
self.__premoderation_callback = premoderation_callback
def get_response(self, question, premoderation_payload=None):
"""Get response to the message.
Args:
question (str): Message.
premoderation_payload: Data to pass to premoderation_callback.
Returns:
str: Response.
"""
# Get supposed response
question = Statement(question)
response = self.__chatbot.generate_response(question,
self.__conversation_id)[1]
""":type : Response"""
# Choose random response if confidence is lower than threshold
if response.confidence < self.__low_confidence_threshold and self.__low_confidence_responses:
response = Statement(choice(self.__low_confidence_responses))
if self.__premoderation_callback:
response = self.__premoderation_callback(question.text,
response.text,
premoderation_payload)
if not response:
return
response = Statement(response)
response.confidence = 1
self.__chatbot.learn_response(response, question)
self.__chatbot.storage.add_to_conversation(self.__conversation_id,
question, response)
if self.__self_training and self.previous_response and '?' in self.previous_response.text and \
self.previous_response.confidence >= self.__low_confidence_threshold:
self.__chatbot.learn_response(question, self.previous_response)
self.__chatbot.storage.add_to_conversation(self.__conversation_id,
self.previous_response,
question)
self.previous_response = response
return response.text
def learn(self, question, response):
"""Learn a response.
Args:
question (str): Question.
response (str): Response
"""
question = Statement(question)
response = Statement(response)
self.__chatbot.learn_response(response, question)
self.__chatbot.storage.add_to_conversation(self.__conversation_id,
question, response)
def statement_exists(self, statement_text):
"""Check if statement exists.
Args:
statement_text (str): Statement text.
Returns:
bool
"""
return True if self.__chatbot.storage.find(statement_text) else False
def remove_statement(self, statement_text):
"""Remove statement from database.
Args:
statement_text (str): Statement text.
"""
self.__chatbot.storage.remove(statement_text)
class Bott:
def quitaNoAlfaNum(self, texto):
import re
tmp = re.compile(r'\W+', re.UNICODE).split(texto)
return ''.join(str(e + " ") for e in tmp)
def __init__(self, nombre_db):
self.nombre_db = self.quitaNoAlfaNum(nombre_db) + "db"
print "*" + self.nombre_db + "*"
self.chatbot = ChatBot(
"Chat",
storage_adapter='chatterbot.storage.MongoDatabaseAdapter',
database_uri='mongodb://localhost:27017/',
database='chat_2',
#TRAINING
trainer='chatterbot.trainers.ListTrainer',
#input_adapter="chatterbot.input.TerminalAdapter",
#output_adapter="chatterbot.output.TerminalAdapter",
logic_adapters=[
{
"import_path":
"chatterbot.logic.BestMatch",
"statement_comparison_function":
"chatterbot.comparisons.levenshtein_distance",
"response_selection_method":
"chatterbot.response_selection.get_most_frequent_response"
},
{
'import_path':
'chatterbot.logic.LowConfidenceAdapter',
'threshold':
0.51,
'default_response':
'Disculpa, no he entendido lo que ha querido decir. Puede replantearlo.'
},
{
'import_path': 'chatterbot.logic.SpecificResponseAdapter',
'input_text': 'Quiero guardar una cita',
'output_text':
'Puede guardar una cita en http://google.com'
},
],
preprocessors=['chatterbot.preprocessors.clean_whitespace'],
#read_only=True,
)
#DEFAULT_SESSION_ID = chatbot.default_conversation_id
from chatterbot.trainers import ChatterBotCorpusTrainer
self.chatbot.set_trainer(ChatterBotCorpusTrainer)
self.chatbot.train("./cruises_es.yml")
# Start by training our bot with the ChatterBot corpus data
self.chatbot.train('chatterbot.corpus.spanish')
'''
conversation = [
"Hello",
"Hi there!",
"How are you doing?",
"I'm doing great.",
"That is good to hear",
"Thank you.",
"You're welcome."
]
chatbot.set_trainer(ListTrainer)
chatbot.train(conversation)
'''
#request = raw_input("Escriba su pregunta: ")
'''
response = chatbot.get_response(None)
response = u' '.join((response.text, '')).encode('utf-8').strip()
imprimir 'Bot: '+response
'''
#ORIGINAL
def responder(self, pregunta):
#imprimir("Pregunta..."+pregunta)
#imprimir("Estamos en chatbot")
#pregunta = pregunta.replace('á','ã').replace('é','&etilde;').replace('í','ĩ').replace('ó','õ').replace('ú','ũ').replace('ñ','ñ')
pregunta = u' '.join((pregunta, '')).encode('utf-8').strip()
response = self.chatbot.get_response(pregunta)
response = response.text.replace('ã', 'á').replace(
'&etilde;',
'é').replace('ĩ', 'í').replace('õ', 'ó').replace(
'ũ', 'ú').replace('ñ', 'ñ')
#response = response.encode("utf-8",'replace')
response = u' '.join((response, '')).decode('utf-8').strip()
#imprimir response
return response
#drop database....
#mongo chat_2 --eval "db.dropDatabase()"
#entrenando
#http://chatterbot.readthedocs.io/en/stable/examples.html
#pregunta = raw_input("Ingrese su respuesta y luego su pregunta\n ")
#response = chatbot.get_response("Si, como todos")
#imprimir "Su respuesta:"+response
'''
while True:
pregunta = raw_input("Human: ")
response = responder(pregunta)
imprimir "Bot: "+response
'''
#CONVERSATION_ID = self.chatbot.storage.create_conversation()
#training
#HUMAN:BOT
def training(self, input_):
#FIJARSE EN FUNCIONES->ENVIAR_MENSAJE
tmp = input_.split("lll")
human = tmp[0]
bot = tmp[1]
human_ = Statement(human)
#statement, response = chatbot.generate_response(input_statement, CONVERSATION_ID)
bot_ = Statement(bot)
self.chatbot.learn_response(bot_, human_)
#chatbot.storage.add_to_conversation(CONVERSATION_ID, statement, response1)
exists = self.chatbot.storage.find(bot_.text)
print("Human: " + human_.text)
print("Bot: " + bot_.text)
return exists.text
text = input()
if 'yes' in text.lower():
return True
elif 'no' in text.lower():
return False
else:
print('Please type either "Yes" or "No"')
return get_feedback()
print('Type something to begin...')
# The following loop will execute each time the user enters input
while True:
try:
input_statement = bot.input.process_input()
statement, response = bot.generate_response(input_statement,
CONVERSATION)
print('\n Is "{}" this a coherent response to "{}"? \n'.format(
response, input_statement))
if get_feedback():
bot.learn_response(CONVERSATION, response, input_statement)
bot.output.process_response(response)
# Press ctrl-c or ctrl-d on the keyboard to exit
except (KeyboardInterrupt, EOFError, SystemExit):
break
#get ansewers from lara bot
bot_input = lara.get_response(question)
if (bot_input.text == "I am sorry, but I do not understand."):
print("Lara :", bot_input.text)
print(
"Lara: Could you please suggest a response for your question here")
proposition = input("Your proposition :")
new_response = Statement(proposition, in_response_to=question)
new_response.search_text = lara.storage.tagger.get_bigram_pair_string(
proposition.lower())
new_response.conversation = str('training')
new_response.search_in_response_to = lara.storage.tagger.get_bigram_pair_string(
question.lower())
lara.learn_response(new_response)
#conn = sqlite3.connect("db.sqlite3")
"""
Once we have a Connection object, we can then create a Cursor object.
Cursors allow us to execute SQL queries against a database.
"""
#df = pd.read_sql_query("select * from statement;", conn)
#print(df.shape)
#df.loc[len(df)] = [len(df)+1, question,question,"training","",question,proposition, ""]
#df["conversation"].loc[df.in_response_to == question] = "training"
#df["search_in_response_to"].loc[df.in_response_to == question] = question
#df["persona"].loc[df.in_response_to == question] = ""
#print(df.shape)
#df.to_sql("statement", conn, if_exists="replace",index = False) #if_exists='append'
