from flask import Flask, render_template, request, redirect, url_for

from keras.models import load_model

from keras.preprocessing import image

import numpy as np

import cv2

import io

import base64

from pymongo import MongoClient



app = Flask(__name__)



# Load the trained model

model = load_model('weights.hdf5')

model.compile(loss='binary_crossentropy',

              optimizer='rmsprop',

              metrics=['accuracy'])



# MongoDB connection

client = MongoClient('mongodb://localhost:27017/')

db = client['userfeedback']

feedback_collection = db['feedback']



@app.route('/', methods=['GET', 'POST'])

def index():

    if request.method == 'POST':

        return redirect(url_for('upload'))

    return render_template('index.html')



@app.route('/upload', methods=['GET', 'POST'])

def upload():

    if request.method == 'POST':

        # Get the uploaded image file

        img_file = request.files['file']

        if img_file:

            # Read the image file

            img_bytes = img_file.stream.read()

            # Convert bytes to numpy array

            img_np = np.frombuffer(img_bytes, np.uint8)

            # Decode numpy array to image

            img = cv2.imdecode(img_np, cv2.IMREAD_COLOR)

            # Resize the image to match the input shape expected by the model

            img_resized = cv2.resize(img, (150, 150))

            # Expand the dimensions to match the input shape expected by the model

            x = np.expand_dims(img_resized, axis=0)

            # Normalize the image data

            x = x / 255.0

            # Predict probabilities for each class

            probabilities = model.predict(x)

            # Find the index of the class with the highest probability

            predicted_class_index = np.argmax(probabilities)

            # Determine the class label

            if predicted_class_index == 1:

                prediction = "Cancer"

                # Swap red and violet colors

                img[:,:,0], img[:,:,2] = img[:,:,2], img[:,:,0].copy()

            else:

                prediction = "Normal"

                img[:,:,0], img[:,:,2] = img[:,:,2], img[:,:,0].copy()

                

            # Convert the swapped image to base64 format for HTML rendering

            _, img_encoded = cv2.imencode('.png', img)

            swapped_img_base64 = base64.b64encode(img_encoded).decode()

            # Render the result template with the prediction and swapped image

            return render_template('result.html', prediction=prediction, swapped_img_base64=swapped_img_base64)

    return render_template('upload.html')



@app.route('/submit', methods=['POST'])

def submit_feedback():

    if request.method == 'POST':

        feedback = request.form.get('feedback')

        if feedback:

            feedback_collection.insert_one({'feedback': feedback})

            # Show alert for successful submission

            return '''

                <script>

                    alert('Thank you for submitting your feedback!');

                    window.location.href = '/';

                </script>

            '''

    return redirect(url_for('index'))



if __name__ == '__main__':

    app.run(debug=True)