import pickle
import numpy as np 
import streamlit as  st 
import cv2
import tensorflow as tf
from tqdm import tqdm 
from PIL import Image
import os 
from tensorflow.keras.preprocessing import sequence
from tensorflow.keras.saving import load_model

st.title("DL-Classifier")

task = st.selectbox('Select One',("Choose any","Sentiment Classification", 'Tumor Detection'))


#choosing tumor detection
#CNN
if task=="Tumor Detection":
     st.subheader("Tumor Detection")
     model_path = os.path.join(os.getcwd(), 'cnn_model.h5')
     cnn_model = load_model(model_path)

     img =st.file_uploader("choose the image",type=('jpg','jpeg','png'))
     def cnn_make_prediction(img,cnn_model):
            img=Image.open(img)
            img=img.resize((128,128))
            img=np.array(img)
            input_img = np.expand_dims(img, axis=0)
            res = cnn_model.predict(input_img)
            if res:
                return"Tumor"
            else:
               return"No Tumor"
     #if img != None:
         #img_f="D:/SEM 3/DL/DL-ALGORITHMS/CNN/tumor_detection/tumordata/"
         #sub_dir=os.listdir(img_f)
         # cel_path=os.path.join(sub_dir,img_f)DL-models/app.py
         #cel_img=os.listdir(cel_path)
         #img_p=cel_img + img.name
         #pred=cnn_make_prediction(img_p,cnn_model)
         #st.write(pred)

     if img is not None:
        
        st.image(img, caption="Uploaded Image.", use_column_width=False, width=200)
        st.write("")

        if st.button("Detect Tumor"):
            result =cnn_make_prediction(img, cnn_model)
            st.subheader("Tumor Detection Result")
            st.write(f"**{result}**")


#choosing sentiment classification 

if task=="Sentiment Classification":
     st.subheader("Sentiment Classification")
     clss_model= st.radio("Select Classification Model:",("RNN","DNN","Backpropagation",'Perceptron','LSTM'))
     select_model=None

     if clss_model=="RNN":
     
          model_path = os.path.join(os.getcwd(), 'rnn_model.h5')
          rnn_model = load_model(model_path)
          with open("rnn_tokeniser.pkl",'rb') as tokeniser_file:
                rnn_tokeniser=pickle.load(tokeniser_file)
                
                st.subheader('RNN Spam Classification')
                
                input=st.text_area("Enter your message here:")
                def rnn_pred(input):
                    max_length=10
                    encoded_test = rnn_tokeniser.texts_to_sequences(input)
                    padded_test = tf.keras.preprocessing.sequence.pad_sequences(encoded_test, maxlen=max_length, padding='post') 
                    predict= (rnn_model.predict(padded_test) > 0.5).astype("int32")
                    if predict:
                         return "Spam "
                    else: 
                         return "Not Spam"        
                if st.button('Check'):
                    pred=rnn_pred([input]) 
                    st.write(pred)

     if clss_model=='Perceptron':
          with open("perceptron_model_saved.pkl",'rb') as model_file:
               percep_model=pickle.load(model_file)
          with open('perceptron_tokeniser_saved.pkl','rb') as model_file:
               percep_token=pickle.load(model_file)
          st.subheader('Perceptron Spam Classification')
          input= st.text_area("Enter your text here")

          def percep_pred(input):
               encoded_test_p = percep_token.texts_to_sequences([input])
               padded_test_p = tf.keras.preprocessing.sequence.pad_sequences(encoded_test_p, maxlen=10) 
               predict_p= percep_model.predict(padded_test_p)
               if predict_p:
                    return "Spam"
               else:
                    return "Not Spam"
          if st.button("Check"):
               pred=percep_pred([input])
               st.write(pred)

     
     if clss_model=="Backpropagation":
          with open('bp_model.pkl','rb') as model_file:
               bp_model=pickle.load(model_file)
          with open('backrpop_tokeniser.pkl','rb') as model_file:
               bp_tokeniser=pickle.load(model_file)
          st.subheader('Movie Review Classification using Backpropagation')
          inp = st.text_area('Enter message')
          def bp_make_predictions(inp, model):
               encoded_inp = bp_tokeniser.texts_to_sequences([inp])
               padded_inp = sequence.pad_sequences(encoded_inp, maxlen=500)
               res = model.predict(padded_inp)
               if res:
                    return "Negative"
               else:
                    return "Positive"        
        
          if st.button('Check'):
               pred = bp_make_predictions([inp], bp_model)
               st.write(pred)

     if clss_model=="DNN":
          
          model_path = os.path.join(os.getcwd(), 'dnn_model.h5')
          dnn_model = load_model(model_path)
          with open("dnn_tokeniser.pkl",'rb') as file:
               dnn_tokeniser = pickle.load(file)

          st.subheader('SMS Spam Classification using DNN')
          inp = st.text_area('Enter message')
        
          def dnn_make_predictions(inp, model):
               encoded_inp = dnn_tokeniser.texts_to_sequences(inp)
               padded_inp = sequence.pad_sequences(encoded_inp, maxlen=10, padding='post')
               res = (model.predict(padded_inp) > 0.5).astype("int32")
               if res:
                    return "Spam"
               else:
                    return "Not Spam"     
        
          if st.button('Check'):
               pred = dnn_make_predictions([inp], dnn_model)
               st.write(pred)
          
     if clss_model=="LSTM":
          model_path = os.path.join(os.getcwd(), 'lstm_model.h5')
          lstm_model = load_model(model_path)

          with open("lstm_tokeniser.pkl",'rb') as file:
               lstm_tokeniser=pickle.load(file)
          st.subheader('Movie Review Classification')
          inp=st.text_area("Enter your review")
          def lstm_make_predictions(inp, model):
               inp = lstm_tokeniser.texts_to_sequences(inp)
               inp = sequence.pad_sequences(inp, maxlen=500)
               res = (model.predict(inp) > 0.5).astype("int32")
               if res:
                    return "Negative"
               else:
                    return "Positive"

          
          if st.button('Check'):
               pred = lstm_make_predictions([inp], lstm_model)
               st.write(pred)