controller.py

# controller.py
import cv2
import pandas as pd
import pickle
import os
from datetime import datetime
import csv
import numpy as np
import streamlit as st
import subprocess
from insightface.app import FaceAnalysis
import warnings
warnings.filterwarnings("ignore", category=FutureWarning)

def mark_attendance():
    st.header("Video File Attendance")
    ENCODING_DIR = "split_encodings"
    ATTENDANCE_DIR = "attendance_logs"
    TIMETABLE_CSV = "timetable.csv"
    CAMERA_MAP_CSV = "camera_mapping.csv"
    TOLERANCE = 0.85
    FRAME_SKIP = 8
    RESIZE_SCALE = 0.5

    os.makedirs(ATTENDANCE_DIR, exist_ok=True)

    now = datetime.now()
    current_day = now.strftime("%A")
    current_time = now.strftime("%H:%M")
    st.info(f"Day: {current_day}, Time: {current_time}")

    df = pd.read_csv(TIMETABLE_CSV)
    df['start_time'] = pd.to_datetime(df['start_time'], format='%H:%M').dt.strftime('%H:%M')
    df['end_time'] = pd.to_datetime(df['end_time'], format='%H:%M').dt.strftime('%H:%M')

    current_class = df[
        (df['day'] == current_day) &
        (df['start_time'] <= current_time) &
        (df['end_time'] > current_time)
        ]

    if current_class.empty:
        st.warning("No class is scheduled right now.")
    else:
        camera_df = pd.read_csv(CAMERA_MAP_CSV)
        face_app = FaceAnalysis(name="buffalo_l", providers=["CPUExecutionProvider"])
        face_app.prepare(ctx_id=0)

        for _, row in current_class.iterrows():
            classroom = row['classroom']
            branch = row['branch']
            semester = row['semester']
            section = row['section']
            subject = row['subject']
            start_time = row['start_time']
            end_time = row['end_time']

            encoding_file = os.path.join(ENCODING_DIR, f"{branch}_{semester}.pickle")
            if not os.path.exists(encoding_file):
                st.warning(f"Encoding file not found for {branch}_{semester}")
                continue

            camera_url = camera_df.loc[camera_df['class_id'] == classroom, 'camera_url'].values
            if len(camera_url) == 0:
                st.warning(f"Camera not mapped for {classroom}")
                continue

            camera_url = camera_url[0]

            st.subheader(f"🧑‍🏫 {branch}-{semester}-{section} | 📍 {classroom} | 📚 {subject}")
            st.text(f"🎥 Using camera: {camera_url}")

            with open(encoding_file, 'rb') as f:
                data = pickle.load(f)

            known_encodings = np.array(data['encodings'])
            known_names = data['metadata']
            seen_names = set()
            cap = cv2.VideoCapture(camera_url)
            frame_count = 0

            if not cap.isOpened():
                st.error(f"Could not open camera: {camera_url}")
                continue

            status_box = st.empty()
            while True:
                ret, frame = cap.read()
                if not ret:
                    break

                frame_count += 1
                if frame_count % FRAME_SKIP != 0:
                    continue

                small_frame = cv2.resize(frame, (0, 0), fx=RESIZE_SCALE, fy=RESIZE_SCALE)
                rgb = cv2.cvtColor(small_frame, cv2.COLOR_BGR2RGB)
                faces = face_app.get(rgb)

                for face in faces:
                    embedding = face.normed_embedding
                    dists = np.linalg.norm(known_encodings - embedding, axis=1)
                    min_dist = np.min(dists)
                    min_idx = np.argmin(dists)
                    name = "Unknown"
                    if min_dist < TOLERANCE:
                        name = known_names[min_idx]['name']

                    if name != "Unknown" and name not in seen_names:
                        seen_names.add(name)
                        date_str = now.strftime("%Y-%m-%d")
                        start_safe = start_time.replace(":", "_")
                        end_safe = end_time.replace(":", "_")
                        filename = f"{branch}-{semester}-{section}-{subject}-{start_safe}-{end_safe}-{date_str}.csv"
                        branch_path = os.path.join(ATTENDANCE_DIR, branch)
                        semester_path = os.path.join(branch_path, str(semester))
                        os.makedirs(semester_path, exist_ok=True)
                        attendance_path = os.path.join(semester_path, filename)
                        write_header = not os.path.exists(attendance_path)
                        with open(attendance_path, 'a', newline='') as f:
                            writer = csv.writer(f)
                            if write_header:
                                writer.writerow(["Name", "Time", "Branch", "Semester", "Section", "Classroom"])
                            writer.writerow([name, now.strftime("%H:%M:%S"), branch, semester, section, classroom])
                        status_box.success(f"✓ Marked Present: {name}")

                if len(seen_names) >= 5:
                    break

            cap.release()
            st.success(f"✅ Attendance saved for {branch}-{semester}-{section}.")


print("✅ All active classrooms processed.")