import nltk
import numpy as np
import tensorflow as tf
from keras.models import Sequential, load_model
from keras.layers import Dense
import random
import json
import pickle

###Deu certo###


nltk.download('punkt')


stemmer = nltk.stem.lancaster.LancasterStemmer()

# Carregando a base de dados
with open("/home/mrdev/Ollama_testes/tutorial_IA/intents.json") as file:
    data = json.load(file)

# Separando as informações
try:
    with open("/home/mrdev/Ollama_testes/tutorial_IA/data.pickle", "rb") as f:
        words, labels, training, output = pickle.load(f)
except:
    words = []
    labels = []
    docs_y = []
    docs_x = []

    # Loop para armazenar as informações de forma separada
    for intent in data["intents"]:
        for pattern in intent["pattern"]:
            wrds = nltk.word_tokenize(pattern)
            words.extend(wrds)
            docs_x.append(wrds)
            docs_y.append(intent["tag"])

            if intent["tag"] not in labels:
                labels.append(intent["tag"])

    # Organizando/pre-processamento
    words = [stemmer.stem(w.lower()) for w in words if w != "?"]
    words = sorted(list(set(words)))

    labels = sorted(labels)

    # Iniciando a tokenização e treinamento
    training = []
    output = []

    out_empty = [0 for _ in range(len(labels))]

    for x, doc in enumerate(docs_x):
        bag = []

        wrds = [stemmer.stem(w) for w in doc]

        for w in words:
            if w in wrds:
                bag.append(1)
            else:
                bag.append(0)
        output_row = out_empty[:]
        output_row[labels.index(docs_y[x])] = 1

        training.append(bag)
        output.append(output_row)

    training = np.array(training)
    output = np.array(output)

    with open("/home/mrdev/Ollama_testes/tutorial_IA/data.pickle", "wb") as f:
        pickle.dump((words, labels, training, output), f)

#print(data["intents"])

# Iniciando a criação do modelo
tf.keras.backend.clear_session()  # Limpa a sessão do TensorFlow
model = Sequential()
model.add(Dense(8, input_shape=(len(training[0]),), activation='relu'))
model.add(Dense(8, activation='relu'))
model.add(Dense(len(output[0]), activation='softmax'))

# Compilando o modelo
model.compile(optimizer='adam', loss='categorical_crossentropy', metrics=['accuracy'])

# Tentando carregar o modelo treinado
try:
    model = load_model("/home/mrdev/Ollama_testes/tutorial_IA/model_keras.h5")
except:
    # Treinando o modelo
    model.fit(training, output, epochs=1000, batch_size=8, verbose=1)
    # Salvando o modelo treinado
    model.save("/home/mrdev/Ollama_testes/tutorial_IA/model_keras.h5")


def bag_of_words(s, words):
    bag = [0 for _ in range(len(words))]
    s_words = nltk.word_tokenize(s)
    s_words = [stemmer.stem(word.lower()) for word in s_words]

    for se in s_words:
        for i, w in enumerate(words):
            if w == se:
                bag[i] = 1
    return np.array(bag)


# Iniciando o chat em tempo real
def chat():
    print("__start__")
    while True:
        inp = input("Você: ")
        if inp.lower() == "sair":
            print("Saindo")
            break
        else:
            results = model.predict(np.array([bag_of_words(inp, words)]))
            results_index = np.argmax(results)
            tag = labels[results_index]

            for tg in data["intents"]:
                if tg["tag"] == tag:
                    responses = tg["responses"]

            print(random.choice(responses))


chat()