import numpy as np
from sklearn.model_selection import train_test_split
from sklearn.metrics import classification_report, accuracy_score
from tensorflow.keras.preprocessing.text import Tokenizer
from tensorflow.keras.preprocessing.sequence import pad_sequences
from BackPropogation import BackPropogation
from sklearn.preprocessing import LabelEncoder
import pickle

# Load the SMS Spam Collection dataset
sms_dataset_path = 'SMSSpamCollection.txt'
sms_data = []
sms_labels = []

with open(sms_dataset_path, 'r', encoding='utf-8') as file:
    for line in file:
        parts = line.strip().split('\t')
        if len(parts) == 2:
            label, message = parts
            sms_labels.append(label)
            sms_data.append(message)
            

# Use LabelEncoder to encode 'spam' and 'ham' into numerical values
label_encoder = LabelEncoder()
sms_labels = label_encoder.fit_transform(sms_labels)

# Assuming your Backpropagation class does not require input_size during initialization
backpropagation = BackPropogation(learning_rate=0.01, epochs=5)

# Split the dataset into training and testing sets
X_train, X_test, y_train, y_test = train_test_split(sms_data, sms_labels, test_size=0.2, random_state=42)

# Tokenize the text data
max_words = 10000
tokenizer = Tokenizer(num_words=max_words, oov_token='<OOV>')
tokenizer.fit_on_texts(X_train)
sequences_train = tokenizer.texts_to_sequences(X_train)
sequences_test = tokenizer.texts_to_sequences(X_test)

# Pad sequences to a fixed length
max_sequence_length = 200
X_train_padded = pad_sequences(sequences_train, maxlen=max_sequence_length, padding='post')
X_test_padded = pad_sequences(sequences_test, maxlen=max_sequence_length, padding='post')

# Flatten the input sequences
X_train_flatten = X_train_padded.reshape((X_train_padded.shape[0], -1))

# Train the Backpropagation model
backpropagation.fit(X_train_flatten, y_train)

# Use the same tokenizer to transform the test data
sequences_test = tokenizer.texts_to_sequences(X_test)
X_test_padded = pad_sequences(sequences_test, maxlen=max_sequence_length, padding='post')

# Make predictions on the test set
predictions = backpropagation.predict(X_test_padded)

# Evaluate and print results
print("Perceptron Classification Report:")
print(classification_report(y_test, predictions))
print("Perceptron Accuracy:", accuracy_score(y_test, predictions))


# Save the trained Backpropagation model using pickle
backpropagation_model_path = 'spam_backpropagation_model.pkl'
with open(backpropagation_model_path, 'wb') as model_file:
    pickle.dump(backpropagation, model_file)

# Save the tokenizer using pickle
tokenizer_path = 'tokenizer_backpropagation.pkl'
with open(tokenizer_path, 'wb') as tokenizer_file:
    pickle.dump(tokenizer, tokenizer_file)