app.py

from flask import Flask, render_template, request, jsonify, send_from_directory
from flask_cors import CORS
from flask_limiter import Limiter
from flask_limiter.util import get_remote_address
from deepface import DeepFace
from werkzeug.utils import secure_filename
import os
import tempfile
import shutil
import uuid
import logging
import time
from datetime import datetime
from functools import wraps
import numpy as np
import cv2
from PIL import Image
import io
import threading
import queue
import hashlib

# Configuration du logging
logging.basicConfig(
    filename='app.log',
    level=logging.INFO,
    format='%(asctime)s - %(levelname)s - %(message)s'
)

def timing_decorator(func):
    @wraps(func)
    def wrapper(*args, **kwargs):
        start = time.time()
        result = func(*args, **kwargs)
        end = time.time()
        logging.info(f'{func.__name__} took {end-start:.2f} seconds to execute')
        return result
    return wrapper

class FaceAnalysisApp:
    def __init__(self):
        self.app = Flask(__name__, static_folder='static')
        self.setup_app()
        self.results_cache = {}
        self.task_queue = queue.Queue()
        self.setup_routes()
        self.start_worker_thread()

    def setup_app(self):
        """Configure l'application Flask"""
        # Configuration de base
        self.app.config.update(
            UPLOAD_FOLDER='static/uploads',
            MAX_CONTENT_LENGTH=16 * 1024 * 1024,
            ALLOWED_EXTENSIONS={'png', 'jpg', 'jpeg', 'gif'},
            SECRET_KEY=os.urandom(24)
        )
        
        # Initialisation CORS et Limiter
        CORS(self.app)
        self.limiter = Limiter(
            self.app,
            key_func=get_remote_address,
            default_limits=["200 per day", "50 per hour"]
        )

    def start_worker_thread(self):
        """Démarre le thread de traitement en arrière-plan"""
        def worker():
            while True:
                task = self.task_queue.get()
                if task is None:
                    break
                try:
                    task()
                except Exception as e:
                    logging.error(f"Error in worker thread: {str(e)}")
                finally:
                    self.task_queue.task_done()

        self.worker_thread = threading.Thread(target=worker, daemon=True)
        self.worker_thread.start()

    def validate_image(self, image_stream):
        """Valide et optimise l'image"""
        try:
            img = Image.open(image_stream)
            
            # Vérification des dimensions
            if img.size[0] > 2000 or img.size[1] > 2000:
                img.thumbnail((2000, 2000), Image.LANCZOS)
                
            # Conversion en RGB si nécessaire
            if img.mode not in ('RGB', 'L'):
                img = img.convert('RGB')
                
            # Optimisation
            output = io.BytesIO()
            img.save(output, format='JPEG', quality=85, optimize=True)
            output.seek(0)
            
            return output
        except Exception as e:
            logging.error(f"Image validation error: {str(e)}")
            raise ValueError("Invalid image format")

    def process_face_detection(self, image_path):
        """Détecte les visages avec mise en cache"""
        image_hash = hashlib.md5(open(image_path, 'rb').read()).hexdigest()
        
        if image_hash in self.results_cache:
            return self.results_cache[image_hash]
            
        try:
            result = DeepFace.analyze(
                img_path=image_path,
                actions=['age', 'gender', 'race', 'emotion'],
                enforce_detection=True
            )
            self.results_cache[image_hash] = result
            return result
        except Exception as e:
            logging.error(f"Face detection error: {str(e)}")
            raise

    @timing_decorator
    def verify_faces(self, image1_path, image2_path):
        """Compare deux visages"""
        try:
            # Vérification des images
            face1 = cv2.imread(image1_path)
            face2 = cv2.imread(image2_path)
            if face1 is None or face2 is None:
                raise ValueError("Unable to read one or both images")

            # Comparaison des visages
            result = DeepFace.verify(
                img1_path=image1_path,
                img2_path=image2_path,
                enforce_detection=True,
                model_name="VGG-Face"
            )
            
            # Enrichissement des résultats
            result.update({
                'timestamp': datetime.now().isoformat(),
                'confidence_score': 1 - result.get('distance', 0),
                'processing_time': time.time()
            })
            
            return result
        except Exception as e:
            logging.error(f"Face verification error: {str(e)}")
            raise

    def setup_routes(self):
        """Configure les routes de l'application"""
        
        @self.app.route('/')
        def index():
            return render_template('index.html')

        @self.app.route('/verify', methods=['POST'])
        @self.limiter.limit("10 per minute")
        def verify_faces_endpoint():
            try:
                # Vérification des fichiers
                if 'image1' not in request.files or 'image2' not in request.files:
                    return jsonify({'error': 'Two images are required'}), 400

                image1 = request.files['image1']
                image2 = request.files['image2']

                # Validation des images
                try:
                    image1_stream = self.validate_image(image1)
                    image2_stream = self.validate_image(image2)
                except ValueError as e:
                    return jsonify({'error': str(e)}), 400

                # Traitement des images
                with tempfile.TemporaryDirectory() as temp_dir:
                    # Sauvegarde temporaire
                    paths = []
                    for img, stream in [(image1, image1_stream), (image2, image2_stream)]:
                        path = os.path.join(temp_dir, secure_filename(img.filename))
                        with open(path, 'wb') as f:
                            f.write(stream.getvalue())
                        paths.append(path)

                    # Vérification des visages
                    result = self.verify_faces(paths[0], paths[1])

                    # Sauvegarde des résultats positifs
                    if result['verified']:
                        permanent_dir = os.path.join(self.app.static_folder, 'verified_faces')
                        os.makedirs(permanent_dir, exist_ok=True)
                        
                        saved_paths = []
                        for i, path in enumerate(paths, 1):
                            name = f"face{i}_{datetime.now().strftime('%Y%m%d_%H%M%S')}_{uuid.uuid4().hex[:8]}.jpg"
                            dest = os.path.join(permanent_dir, name)
                            shutil.copy2(path, dest)
                            saved_paths.append(f'/static/verified_faces/{name}')
                        
                        result['image1_url'] = saved_paths[0]
                        result['image2_url'] = saved_paths[1]

                return jsonify(result)

            except Exception as e:
                logging.error(f"Verification endpoint error: {str(e)}")
                return jsonify({'error': 'An internal error occurred'}), 500

        @self.app.route('/analyze', methods=['POST'])
        @self.limiter.limit("20 per minute")
        def analyze_face_endpoint():
            try:
                if 'image' not in request.files:
                    return jsonify({'error': 'No image provided'}), 400

                image = request.files['image']
                
                # Validation de l'image
                try:
                    image_stream = self.validate_image(image)
                except ValueError as e:
                    return jsonify({'error': str(e)}), 400

                # File d'attente pour les résultats
                result_queue = queue.Queue()
                
                def process_task():
                    try:
                        with tempfile.NamedTemporaryFile(suffix='.jpg', delete=False) as temp_file:
                            temp_file.write(image_stream.getvalue())
                            result = self.process_face_detection(temp_file.name)
                            result_queue.put(('success', result))
                    except Exception as e:
                        result_queue.put(('error', str(e)))
                    finally:
                        try:
                            os.unlink(temp_file.name)
                        except:
                            pass

                # Ajout de la tâche à la file d'attente
                self.task_queue.put(process_task)
                
                # Attente du résultat
                try:
                    status, result = result_queue.get(timeout=30)
                    if status == 'error':
                        return jsonify({'error': result}), 500
                    return jsonify(result)
                except queue.Empty:
                    return jsonify({'error': 'Processing timeout'}), 408

            except Exception as e:
                logging.error(f"Analysis endpoint error: {str(e)}")
                return jsonify({'error': 'An internal error occurred'}), 500

        @self.app.errorhandler(413)
        def request_entity_too_large(error):
            return jsonify({'error': 'File too large'}), 413

        @self.app.errorhandler(429)
        def ratelimit_handler(e):
            return jsonify({'error': 'Rate limit exceeded'}), 429

    def run(self, host='0.0.0.0', port=5000, debug=False):
        """Démarre l'application Flask"""
        self.app.run(host=host, port=port, debug=debug)

if __name__ == '__main__':
    app = FaceAnalysisApp()
    app.run(debug=True)