import os
import gradio as gr
from services.aws_service import AwsService
from dotenv import load_dotenv
import cv2
import face_recognition
import numpy as np

load_dotenv()
   
def find_face_encodings(image_path):
  # reading image
  image = cv2.imread(image_path)
  # get face encodings from the image
  face_enc = face_recognition.face_encodings(image)
  # return face encodings
  return face_enc[0]

def face_similarity(img1, img2):
  image_1 = find_face_encodings(img1)
  image_2  = find_face_encodings(img2)

  # checking both images are same
  is_same = face_recognition.compare_faces([image_1], image_2)[0]
  accuracy = 0
  if is_same:
    # finding the distance level between images
    distance = face_recognition.face_distance([image_1], image_2)
    distance = round(distance[0] * 100)
    
    # calcuating accuracy level between images
    accuracy = 100 - round(distance)
  return {"is_same": is_same, "accuracy": accuracy}

def convert_pil_to_open_cv(pil_image):
  open_cv_image = np.array(pil_image) 
  open_cv_image = open_cv_image[:, :, ::-1].copy() 
  return open_cv_image

def check_image_similarity(main_image_folder, compare_image_folder):

  main_image = AwsService.get_image_from_s3(os.environ.get('AWS_S3_BUCKET'), main_image_folder)
  compare_image = AwsService.get_image_from_s3(os.environ.get('AWS_S3_BUCKET'), compare_image_folder)
  
  # download image to local storage
  main_image["pil"].save("main_image.jpg")
  compare_image["pil"].save("compare_image.jpg")

  result = face_similarity("main_image.jpg", "compare_image.jpg")

  os.remove("main_image.jpg")
  os.remove("compare_image.jpg")

  return result

iface = gr.Interface(
  fn=check_image_similarity, 
  inputs=[
    gr.Textbox(lines=1, placeholder="Main Image Key", value="PhotoShoots/653fea0aad75f4caca672886/Croppeds/IMG-1698687849160-2.jpg"),
    gr.Textbox(lines=1, placeholder="Compare Image Key", value="PhotoShoots/653fea0aad75f4caca672886/Croppeds/IMG-1698687855097-1.jpg")
  ], 
  outputs="text"
)

iface.launch()