Update cursor_tracker.py

cursor_tracker.py

@@ -1,16 +1,273 @@
-"""
-Cursor detection and action logging utilities.
-"""
-
 import os
 import json
+import requests
+import subprocess
+import shutil
+import time
+import re
+import threading
+from typing import Dict, List, Set, Optional
+from huggingface_hub import HfApi, list_repo_files
+
 import cv2
 import numpy as np
 from pathlib import Path
-from email.message import EmailMessage
 import smtplib
-from typing import List, Tuple
+from email.message import EmailMessage
+
+# ==== CONFIGURATION ====
+HF_TOKEN = os.getenv("HF_TOKEN", "")
+SOURCE_REPO_ID = os.getenv("SOURCE_REPO", "Fred808/BG1")
+
+# Path Configuration
+DOWNLOAD_FOLDER = "downloads"
+EXTRACT_FOLDER = "extracted"
+FRAMES_OUTPUT_FOLDER = "extracted_frames" # New folder for extracted frames
+CURSOR_TRACKING_OUTPUT_FOLDER = "cursor_tracking_results" # New folder for cursor tracking results
+CURSOR_TEMPLATES_DIR = "cursors"
+
+os.makedirs(DOWNLOAD_FOLDER, exist_ok=True)
+os.makedirs(EXTRACT_FOLDER, exist_ok=True)
+os.makedirs(FRAMES_OUTPUT_FOLDER, exist_ok=True)
+os.makedirs(CURSOR_TRACKING_OUTPUT_FOLDER, exist_ok=True)
+os.makedirs(CURSOR_TEMPLATES_DIR, exist_ok=True) # Ensure cursor templates directory exists
+
+# State Files
+DOWNLOAD_STATE_FILE = "download_progress.json"
+PROCESS_STATE_FILE = "process_progress.json"
+FAILED_FILES_LOG = "failed_files.log"
+
+# Processing Parameters
+CHUNK_SIZE = 1
+PROCESSING_DELAY = 2
+MAX_RETRIES = 3
+MIN_FREE_SPACE_GB = 2  # Minimum free space in GB before processing
+
+# Frame Extraction Parameters
+DEFAULT_FPS = 3 # Default frames per second for extraction
+
+# Cursor Tracking Parameters
+CURSOR_THRESHOLD = 0.8
+
+# Initialize HF API
+hf_api = HfApi(token=HF_TOKEN)
+
+# Global State
+processing_status = {
+    "is_running": False,
+    "current_file": None,
+    "total_files": 0,
+    "processed_files": 0,
+    "failed_files": 0,
+    "extracted_courses": 0,
+    "extracted_videos": 0,
+    "extracted_frames_count": 0,
+    "tracked_cursors_count": 0,
+    "last_update": None,
+    "logs": []
+}
+
+def log_message(message: str):
+    """Log messages with timestamp"""
+    timestamp = time.strftime("%Y-%m-%d %H:%M:%S")
+    log_entry = f"[{timestamp}] {message}"
+    print(log_entry)
+    processing_status["logs"].append(log_entry)
+    processing_status["last_update"] = timestamp
+    if len(processing_status["logs"]) > 100:
+        processing_status["logs"] = processing_status["logs"][-100:]
+
+def log_failed_file(filename: str, error: str):
+    """Log failed files to persistent file"""
+    with open(FAILED_FILES_LOG, "a") as f:
+        f.write(f"{time.strftime("%Y-%m-%d %H:%M:%S")} - {filename}: {error}\n")
+
+def get_disk_usage(path: str) -> Dict[str, float]:
+    """Get disk usage statistics in GB"""
+    statvfs = os.statvfs(path)
+    total = statvfs.f_frsize * statvfs.f_blocks / (1024**3)
+    free = statvfs.f_frsize * statvfs.f_bavail / (1024**3)
+    used = total - free
+    return {"total": total, "free": free, "used": used}
+
+def check_disk_space(path: str = ".") -> bool:
+    """Check if there's enough disk space"""
+    disk_info = get_disk_usage(path)
+    if disk_info["free"] < MIN_FREE_SPACE_GB:
+        log_message(f'⚠️ Low disk space: {disk_info["free"]:.2f}GB free, {disk_info["used"]:.2f}GB used')
+        return False
+    return True
+
+def cleanup_temp_files():
+    """Clean up temporary files to free space"""
+    log_message("🧹 Cleaning up temporary files...")
+    
+    # Clean old downloads (keep only current processing file)
+    current_file = processing_status.get("current_file")
+    for file in os.listdir(DOWNLOAD_FOLDER):
+        if file != current_file and file.endswith((".rar", ".zip")):
+            try:
+                os.remove(os.path.join(DOWNLOAD_FOLDER, file))
+                log_message(f"🗑️ Removed old download: {file}")
+            except:
+                pass
+
+def load_json_state(file_path: str, default_value):
+    """Load state from JSON file"""
+    if os.path.exists(file_path):
+        try:
+            with open(file_path, "r") as f:
+                return json.load(f)
+        except json.JSONDecodeError:
+            log_message(f"⚠️ Corrupted state file: {file_path}")
+    return default_value
+
+def save_json_state(file_path: str, data):
+    """Save state to JSON file"""
+    with open(file_path, "w") as f:
+        json.dump(data, f, indent=2)
+
+def download_with_retry(url: str, dest_path: str, max_retries: int = 3) -> bool:
+    """Download file with retry logic and disk space checking"""
+    if not check_disk_space():
+        cleanup_temp_files()
+        if not check_disk_space():
+            log_message("❌ Insufficient disk space even after cleanup")
+            return False
+    
+    headers = {"Authorization": f"Bearer {HF_TOKEN}"}
+    for attempt in range(max_retries):
+        try:
+            with requests.get(url, headers=headers, stream=True) as r:
+                r.raise_for_status()
+                
+                # Check content length if available
+                content_length = r.headers.get("content-length")
+                if content_length:
+                    size_gb = int(content_length) / (1024**3)
+                    disk_info = get_disk_usage(".")
+                    if size_gb > disk_info["free"] - 0.5:  # Leave 0.5GB buffer
+                        log_message(f'❌ File too large: {size_gb:.2f}GB, only {disk_info["free"]:.2f}GB free')
+                        return False
+                
+                with open(dest_path, "wb") as f:
+                    for chunk in r.iter_content(chunk_size=8192):
+                        f.write(chunk)
+            return True
+        except Exception as e:
+            if attempt < max_retries - 1:
+                time.sleep(2 ** attempt)
+                continue
+            log_message(f"❌ Download failed after {max_retries} attempts: {e}")
+            return False
+    return False
+
+def is_multipart_rar(filename: str) -> bool:
+    """Check if this is a multi-part RAR file"""
+    return ".part" in filename.lower() and filename.lower().endswith(".rar")
+
+def get_rar_part_base(filename: str) -> str:
+    """Get the base name for multi-part RAR files"""
+    if ".part" in filename.lower():
+        return filename.split(".part")[0]
+    return filename.replace(".rar", "")
+
+def extract_with_retry(rar_path: str, output_dir: str, max_retries: int = 2) -> bool:
+    """Extract RAR with retry and recovery, handling multi-part archives"""
+    filename = os.path.basename(rar_path)
+    
+    # For multi-part RARs, we need the first part
+    if is_multipart_rar(filename):
+        base_name = get_rar_part_base(filename)
+        first_part = f"{base_name}.part01.rar"
+        first_part_path = os.path.join(os.path.dirname(rar_path), first_part)
+        
+        if not os.path.exists(first_part_path):
+            log_message(f"⚠️ Multi-part RAR detected but first part not found: {first_part}")
+            return False
+        
+        rar_path = first_part_path
+        log_message(f"📦 Processing multi-part RAR starting with: {first_part}")
+    
+    for attempt in max_retries:
+        try:
+            # Test RAR first
+            test_cmd = ["unrar", "t", rar_path]
+            test_result = subprocess.run(test_cmd, capture_output=True, text=True)
+            if test_result.returncode != 0:
+                log_message(f"⚠️ RAR test failed: {test_result.stderr}")
+                if attempt == max_retries - 1:
+                    return False
+                continue
+            
+            # Extract RAR
+            cmd = ["unrar", "x", "-o+", rar_path, output_dir]
+            if attempt > 0:  # Try recovery on subsequent attempts
+                cmd.insert(2, "-kb")
+            
+            result = subprocess.run(cmd, capture_output=True, text=True)
+            if result.returncode == 0:
+                log_message(f"✅ Successfully extracted: {os.path.basename(rar_path)}")
+                return True
+            else:
+                error_msg = result.stderr or result.stdout
+                log_message(f"⚠️ Extraction attempt {attempt + 1} failed: {error_msg}")
+                
+                if "checksum error" in error_msg.lower() or "CRC failed" in error_msg:
+                    log_message(f"⚠️ Data corruption detected, attempt {attempt + 1}")
+                elif result.returncode == 10:
+                    log_message(f"⚠️ No files to extract (exit code 10)")
+                    return False
+                elif result.returncode == 1:
+                    log_message(f"⚠️ Non-fatal error (exit code 1)")
+                    
+        except Exception as e:
+            log_message(f"❌ Extraction exception: {str(e)}")
+            if attempt == max_retries - 1:
+                return False
+            time.sleep(1)
+    
+    return False
 
+# --- Frame Extraction Utilities ---
+def ensure_dir(path):
+    os.makedirs(path, exist_ok=True)
+
+def extract_frames(video_path, output_dir, fps=DEFAULT_FPS):
+    """Extract frames from video at the specified frames per second (fps)."""
+    log_message(f"[INFO] Extracting frames from {video_path} to {output_dir} at {fps} fps...")
+    ensure_dir(output_dir)
+    cap = cv2.VideoCapture(str(video_path))
+    if not cap.isOpened():
+        log_message(f"[ERROR] Failed to open video file: {video_path}")
+        return 0
+    video_fps = cap.get(cv2.CAP_PROP_FPS)
+    # log_message(f"[DEBUG] Video FPS: {video_fps}")
+    if not video_fps or video_fps <= 0:
+        video_fps = 30  # fallback if FPS is not available
+        log_message(f"[WARN] Using fallback FPS: {video_fps}")
+    frame_interval = int(round(video_fps / fps))
+    # log_message(f"[DEBUG] Frame interval: {frame_interval}")
+    frame_idx = 0
+    saved_idx = 1
+    total_frames = int(cap.get(cv2.CAP_PROP_FRAME_COUNT))
+    log_message(f"[DEBUG] Total frames in video: {total_frames}")
+    while cap.isOpened():
+        ret, frame = cap.read()
+        if not ret:
+            # log_message(f"[DEBUG] No more frames to read at frame_idx {frame_idx}.")
+            break
+        if frame_idx % frame_interval == 0:
+            frame_name = f"{saved_idx:04d}.png"
+            cv2.imwrite(str(Path(output_dir) / frame_name), frame)
+            # log_message(f"[DEBUG] Saved frame {frame_idx} as {frame_name}")
+            saved_idx += 1
+        frame_idx += 1
+    cap.release()
+    log_message(f"Extracted {saved_idx-1} frames from {video_path} to {output_dir}")
+    return saved_idx - 1
+
+# --- Cursor Tracking Utilities ---
 def to_rgb(img):
     if img is None:
         return None
@@ -22,10 +279,12 @@ def to_rgb(img):
 
 def get_mask_from_alpha(template_img):
     if template_img is not None and len(template_img.shape) == 3 and template_img.shape[2] == 4:
+        # Use alpha channel as mask (nonzero alpha = 255)
         return (template_img[:, :, 3] > 0).astype(np.uint8) * 255
     return None
 
-def detect_cursor_in_frame_multi(frame, cursor_templates, threshold=0.8):
+def detect_cursor_in_frame_multi(frame, cursor_templates, threshold=CURSOR_THRESHOLD):
+    """Detect cursor position in a frame using multiple templates. Returns best match above threshold."""
     best_pos = None
     best_conf = -1
     best_template_name = None
@@ -34,14 +293,14 @@ def detect_cursor_in_frame_multi(frame, cursor_templates, threshold=0.8):
         template_rgb = to_rgb(cursor_template)
         mask = get_mask_from_alpha(cursor_template)
         if template_rgb is None or frame_rgb is None or template_rgb.shape[2] != frame_rgb.shape[2]:
-            print(f"[WARN] Skipping template {template_name} due to channel mismatch or load error.")
+            log_message(f"[WARN] Skipping template {template_name} due to channel mismatch or load error.")
             continue
         try:
             result = cv2.matchTemplate(frame_rgb, template_rgb, cv2.TM_CCOEFF_NORMED, mask=mask)
         except Exception as e:
-            print(f"[WARN] matchTemplate failed for {template_name}: {e}")
+            log_message(f"[WARN] matchTemplate failed for {template_name}: {e}")
             continue
-        _, max_val, _, max_loc = cv2.minMaxLoc(result)
+        min_val, max_val, min_loc, max_loc = cv2.minMaxLoc(result)
         if max_val > best_conf:
             best_conf = max_val
             if max_val >= threshold:
@@ -54,111 +313,285 @@ def detect_cursor_in_frame_multi(frame, cursor_templates, threshold=0.8):
         return best_pos, best_conf, best_template_name
     return None, best_conf, None
 
-def ensure_dir(path):
-    os.makedirs(path, exist_ok=True)
-
 def send_email_with_attachment(subject, body, to_email, from_email, app_password, attachment_path):
     msg = EmailMessage()
-    msg['Subject'] = subject
-    msg['From'] = from_email
-    msg['To'] = to_email
+    msg["Subject"] = subject
+    msg["From"] = from_email
+    msg["To"] = to_email
     msg.set_content(body)
-    with open(attachment_path, 'rb') as f:
+    with open(attachment_path, "rb") as f:
         file_data = f.read()
         file_name = Path(attachment_path).name
-    msg.add_attachment(file_data, maintype='application', subtype='octet-stream', filename=file_name)
+    msg.add_attachment(file_data, maintype="application", subtype="octet-stream", filename=file_name)
     try:
-        with smtplib.SMTP_SSL('smtp.gmail.com', 465) as smtp:
+        with smtplib.SMTP_SSL("smtp.gmail.com", 465) as smtp:
             smtp.login(from_email, app_password)
             smtp.send_message(msg)
-        print(f"[SUCCESS] Email sent to {to_email}")
+        log_message(f"[SUCCESS] Email sent to {to_email}")
     except Exception as e:
-        print(f"[ERROR] Failed to send email: {e}")
-
-def track_cursor_from_images(
-    images: List[Tuple[str, np.ndarray]],
-    cursor_templates_dir: str,
-    output_json_path: str = None,
-    threshold=0.8,
-    email_results=False,
-    return_results=False,
-    append=False
-):
-    """Detect cursor in a list of in-memory images and optionally save or return results."""
+        log_message(f"[ERROR] Failed to send email: {e}")
 
+def track_cursor(frames_dir, cursor_templates_dir, output_json_path, threshold=CURSOR_THRESHOLD, start_frame=1, email_results=False):
+    """Detect cursor in each frame using multiple templates, print status, and write positions to a JSON file."""
+    log_message(f"[INFO] Tracking cursors in {frames_dir}...")
+    frames_dir = Path(frames_dir).resolve()
+    output_json_path = Path(output_json_path).resolve()
     cursor_templates_dir = Path(cursor_templates_dir).resolve()
-    if output_json_path:
-        output_json_path = Path(output_json_path).resolve()
-        ensure_dir(output_json_path.parent)
-
+    # log_message(f"[DEBUG] frames_dir: {frames_dir}")
+    # log_message(f"[DEBUG] cursor_templates_dir: {cursor_templates_dir}")
+    # log_message(f"[DEBUG] output_json_path: {output_json_path}")
+    ensure_dir(frames_dir)
+    ensure_dir(output_json_path.parent)
+    # Load all PNG templates from the cursor_templates_dir
     cursor_templates = {}
-    for template_file in cursor_templates_dir.glob('*.png'):
+    for template_file in cursor_templates_dir.glob("*.png"):
         template_img = cv2.imread(str(template_file), cv2.IMREAD_UNCHANGED)
         if template_img is not None:
             cursor_templates[template_file.name] = template_img
         else:
-            print(f"[WARN] Could not load template: {template_file}")
+            log_message(f"[WARN] Could not load template: {template_file}")
     if not cursor_templates:
-        raise FileNotFoundError(f"No cursor templates found in: {cursor_templates_dir}")
-
+        log_message(f"[ERROR] No cursor templates found in: {cursor_templates_dir}")
+        return 0
     results = []
-
-    for frame_filename, frame in images:
+    tracked_count = 0
+    for frame_file in sorted(frames_dir.glob("*.png")):
+        frame_num = int(frame_file.stem)
+        if frame_num < start_frame:
+            continue
+        frame = cv2.imread(str(frame_file), cv2.IMREAD_UNCHANGED)
         if frame is None:
-            print(f"[WARN] Frame {frame_filename} is empty")
+            log_message(f"[WARN] Could not load frame: {frame_file}")
             continue
-
         pos, conf, template_name = detect_cursor_in_frame_multi(frame, cursor_templates, threshold)
         if pos is not None:
-            print(f"{frame_filename}: Cursor at {pos} (template: {template_name})")
+            # log_message(f"{frame_file.name}: Cursor at {pos} (template: {template_name})")
             results.append({
-                'frame': frame_filename,
-                'cursor_active': True,
-                'x': pos[0],
-                'y': pos[1],
-                'confidence': conf,
-                'template': template_name
+                "frame": frame_file.name,
+                "cursor_active": True,
+                "x": pos[0],
+                "y": pos[1],
+                "confidence": conf,
+                "template": template_name
             })
+            tracked_count += 1
         else:
-            print(f"{frame_filename}: Cursor disabled")
+            # log_message(f"{frame_file.name}: Cursor disabled")
             results.append({
-                'frame': frame_filename,
-                'cursor_active': False,
-                'x': None,
-                'y': None,
-                'confidence': conf,
-                'template': None
+                "frame": frame_file.name,
+                "cursor_active": False,
+                "x": None,
+                "y": None,
+                "confidence": conf,
+                "template": None
             })
+    try:
+        with open(output_json_path, "w") as f:
+            json.dump(results, f, indent=2)
+        log_message(f"[SUCCESS] Cursor tracking results saved to {output_json_path}")
+        if email_results:
+            log_message("[INFO] Preparing to email results...")
+            to_email = os.environ.get("TO_EMAIL")
+            from_email = os.environ.get("FROM_EMAIL")
+            app_password = os.environ.get("GMAIL_APP_PASSWORD")
+            if not (to_email and from_email and app_password):
+                log_message("[ERROR] Email environment variables not set. Please set TO_EMAIL, FROM_EMAIL, and GMAIL_APP_PASSWORD.")
+                # return tracked_count # Don't return here, just log error
+            else:
+                send_email_with_attachment(
+                    subject="Cursor Tracking Results",
+                    body="See attached JSON results.",
+                    to_email=to_email,
+                    from_email=from_email,
+                    app_password=app_password,
+                    attachment_path=output_json_path
+                )
+    except Exception as e:
+        log_message(f"[ERROR] Failed to write output JSON: {e}")
+        # raise # Don't raise, just log error
+    return tracked_count
+
+def process_rar_file(rar_path: str) -> bool:
+    """Process a single RAR file - extract, then process videos for frames and cursor tracking"""
+    filename = os.path.basename(rar_path)
+    processing_status["current_file"] = filename
+    
+    # Handle multi-part RAR naming
+    if is_multipart_rar(filename):
+        course_name = get_rar_part_base(filename)
+    else:
+        course_name = filename.replace(".rar", "")
+    
+    extract_dir = os.path.join(EXTRACT_FOLDER, course_name)
+    
+    try:
+        log_message(f"🔄 Processing: {filename}")
+        
+        # Clean up any existing directory
+        if os.path.exists(extract_dir):
+            shutil.rmtree(extract_dir, ignore_errors=True)
+        
+        # Extract RAR
+        os.makedirs(extract_dir, exist_ok=True)
+        if not extract_with_retry(rar_path, extract_dir):
+            raise Exception("RAR extraction failed")
+        
+        # Count extracted files
+        file_count = 0
+        video_files_found = []
+        for root, dirs, files in os.walk(extract_dir):
+            for file in files:
+                file_count += 1
+                if file.lower().endswith((".mp4", ".avi", ".mov", ".mkv")):
+                    video_files_found.append(os.path.join(root, file))
+        
+        processing_status["extracted_courses"] += 1
+        log_message(f"✅ Successfully extracted \'{course_name}\' ({file_count} files, {len(video_files_found)} videos)")
 
-    if return_results:
-        return results
+        # Process video files for frame extraction and cursor tracking
+        for video_path in video_files_found:
+            video_filename = Path(video_path).name
+            # Create a unique output directory for frames for each video
+            frames_output_dir = os.path.join(FRAMES_OUTPUT_FOLDER, f"{course_name}_{video_filename.replace('.', '_')}_frames")
+            ensure_dir(frames_output_dir)
+            
+            extracted_frames_count = extract_frames(video_path, frames_output_dir, fps=DEFAULT_FPS)
+            processing_status["extracted_frames_count"] += extracted_frames_count
+            if extracted_frames_count > 0:
+                processing_status["extracted_videos"] += 1
+                log_message(f"[INFO] Extracted {extracted_frames_count} frames from {video_filename}")
+
+                # Perform cursor tracking on the extracted frames
+                cursor_output_json = os.path.join(CURSOR_TRACKING_OUTPUT_FOLDER, f"{course_name}_{video_filename.replace('.', '_')}_cursor_data.json")
+                tracked_cursors = track_cursor(frames_output_dir, CURSOR_TEMPLATES_DIR, cursor_output_json, threshold=CURSOR_THRESHOLD)
+                processing_status["tracked_cursors_count"] += tracked_cursors
+                log_message(f"[INFO] Tracked {tracked_cursors} cursors in frames from {video_filename}")
+            else:
+                log_message(f"[WARN] No frames extracted from {video_filename}")
+
+        return True
+    
+    except Exception as e:
+        error_msg = str(e)
+        log_message(f"❌ Processing failed: {error_msg}")
+        log_failed_file(filename, error_msg)
+        return False
+    
+    finally:
+        processing_status["current_file"] = None
 
-    if output_json_path:
+def main_processing_loop():
+    """Main processing workflow - extraction, frame extraction, and cursor tracking"""
+    processing_status["is_running"] = True
+    
+    try:
+        # Load state
+        processed_rars = load_json_state(PROCESS_STATE_FILE, {"processed_rars": []})["processed_rars"]
+        download_state = load_json_state(DOWNLOAD_STATE_FILE, {"next_download_index": 0})
+        next_index = download_state["next_download_index"]
+        
+        log_message(f"📊 Starting from index {next_index}")
+        log_message(f"📊 Previously processed: {len(processed_rars)} files")
+        
+        # Get file list
         try:
-            if append and output_json_path.exists():
-                with open(output_json_path, 'r', encoding='utf-8') as f:
-                    old_data = json.load(f)
-                results = old_data + results
-            with open(output_json_path, 'w', encoding='utf-8') as f:
-                json.dump(results, f, indent=2)
-            print(f"[SUCCESS] Cursor tracking results saved to {output_json_path}")
+            files = list(hf_api.list_repo_files(repo_id=SOURCE_REPO_ID, repo_type="dataset"))
+            rar_files = sorted([f for f in files if f.endswith(".rar")])
+            
+            processing_status["total_files"] = len(rar_files)
+            log_message(f"📁 Found {len(rar_files)} RAR files in repository")
+            
+            if next_index >= len(rar_files):
+                log_message("✅ All files have been processed!")
+                return
+                
         except Exception as e:
-            print(f"[ERROR] Failed to write output JSON: {e}")
-            raise
-
-    if email_results:
-        print("[INFO] Preparing to email results...")
-        to_email = os.environ.get("TO_EMAIL")
-        from_email = os.environ.get("FROM_EMAIL")
-        app_password = os.environ.get("GMAIL_APP_PASSWORD")
-        if not (to_email and from_email and app_password):
-            print("[ERROR] Email environment variables not set.")
+            log_message(f"❌ Failed to get file list: {str(e)}")
             return
-        send_email_with_attachment(
-            subject="Cursor Tracking Results",
-            body="See attached JSON results.",
-            to_email=to_email,
-            from_email=from_email,
-            app_password=app_password,
-            attachment_path=output_json_path
-        )
+        
+        # Process only one file per run
+        if next_index < len(rar_files):
+            rar_file = rar_files[next_index]
+            filename = os.path.basename(rar_file)
+            
+            if filename in processed_rars:
+                log_message(f"⏭️ Skipping already processed: {filename}")
+                processing_status["processed_files"] += 1
+                # Move to next file
+                next_index += 1
+                save_json_state(DOWNLOAD_STATE_FILE, {"next_download_index": next_index})
+                log_message(f"📊 Moving to next file. Progress: {next_index}/{len(rar_files)}")
+                return
+            
+            log_message(f"📥 Downloading: {filename}")
+            dest_path = os.path.join(DOWNLOAD_FOLDER, filename)
+            
+            # Download file
+            download_url = f"https://huggingface.co/datasets/{SOURCE_REPO_ID}/resolve/main/{rar_file}"
+            if download_with_retry(download_url, dest_path):
+                # Process file
+                if process_rar_file(dest_path):
+                    processed_rars.append(filename)
+                    save_json_state(PROCESS_STATE_FILE, {"processed_rars": processed_rars})
+                    log_message(f"✅ Successfully processed: {filename}")
+                    processing_status["processed_files"] += 1
+                else:
+                    log_message(f"❌ Failed to process: {filename}")
+                    processing_status["failed_files"] += 1
+                
+                # Clean up downloaded file
+                try:
+                    os.remove(dest_path)
+                    log_message(f"🗑️ Cleaned up download: {filename}")
+                except:
+                    pass
+            else:
+                log_message(f"❌ Failed to download: {filename}")
+                processing_status["failed_files"] += 1
+            
+            # Update download state for next run
+            next_index += 1
+            save_json_state(DOWNLOAD_STATE_FILE, {"next_download_index": next_index})
+            
+            # Status update
+            log_message(f"📊 Progress: {next_index}/{len(rar_files)} files processed")
+            log_message(f'📊 Extracted: {processing_status["extracted_courses"]} courses')
+            log_message(f'📊 Videos Processed: {processing_status["extracted_videos"]}')
+            log_message(f'📊 Frames Extracted: {processing_status["extracted_frames_count"]}')
+            log_message(f'📊 Cursors Tracked: {processing_status["tracked_cursors_count"]}')
+            log_message(f'📊 Failed: {processing_status["failed_files"]} files')
+            
+            if next_index < len(rar_files):
+                log_message(f"🔄 Run the script again to process the next file: {os.path.basename(rar_files[next_index])}")
+            else:
+                log_message("🎉 All files have been processed!")
+        else:
+            log_message("✅ All files have been processed!")
+        
+        log_message("🎉 Processing complete!")
+        log_message(f'📊 Final stats: {processing_status["extracted_courses"]} courses extracted, {processing_status["extracted_videos"]} videos processed, {processing_status["extracted_frames_count"]} frames extracted, {processing_status["tracked_cursors_count"]} cursors tracked')
+        
+    except KeyboardInterrupt:
+        log_message("⏹️ Processing interrupted by user")
+    except Exception as e:
+        log_message(f"❌ Fatal error: {str(e)}")
+    finally:
+        processing_status["is_running"] = False
+        cleanup_temp_files()
+
+# Expose necessary functions and variables for download_api.py
+__all__ = [
+    "main_processing_loop",
+    "processing_status",
+    "CURSOR_TRACKING_OUTPUT_FOLDER",
+    "CURSOR_TEMPLATES_DIR",
+    "log_message",
+    "send_email_with_attachment",
+    "track_cursor",
+    "extract_frames",
+    "DEFAULT_FPS",
+    "CURSOR_THRESHOLD",
+    "ensure_dir"
+]
+
+