auctiora/wiki/QUICKSTART.md

Quick Start Guide

Get the scraper running in minutes without downloading YOLO models!

Minimal Setup (No Object Detection)

The scraper works perfectly fine without YOLO object detection. You can run it immediately and add object detection later if needed.

Step 1: Run the Scraper

# Using Maven
mvn clean compile exec:java -Dexec.mainClass="com.auction.scraper.TroostwijkScraper"

Or in IntelliJ IDEA:

1. Open TroostwijkScraper.java
2. Right-click on the main method
3. Select "Run 'TroostwijkScraper.main()'"

What You'll See

=== Troostwijk Auction Scraper ===

Initializing scraper...
⚠️  Object detection disabled: YOLO model files not found
   Expected files:
   - models/yolov4.cfg
   - models/yolov4.weights
   - models/coco.names
   Scraper will continue without image analysis.

[1/3] Discovering Dutch auctions...
✓ Found 5 auctions: [12345, 12346, 12347, 12348, 12349]

[2/3] Fetching lot details...
  Processing sale 12345...

[3/3] Starting monitoring service...
✓ Monitoring active. Press Ctrl+C to stop.

Step 2: Test Desktop Notifications

The scraper will automatically send desktop notifications when:

• A new bid is placed on a monitored lot
• An auction is closing within 5 minutes

No setup required - desktop notifications work out of the box!

---

Optional: Add Email Notifications

If you want email notifications in addition to desktop notifications:

# Set environment variable
export NOTIFICATION_CONFIG="smtp:your.email@gmail.com:app_password:your.email@gmail.com"

# Then run the scraper
mvn exec:java -Dexec.mainClass="com.auction.scraper.TroostwijkScraper"

Get Gmail App Password:

1. Enable 2FA in Google Account
2. Go to: Google Account → Security → 2-Step Verification → App passwords
3. Generate password for "Mail"
4. Use that password (not your regular Gmail password)

---

Optional: Add Object Detection Later

If you want AI-powered image analysis to detect objects in auction photos:

1. Create models directory

mkdir models
cd models

2. Download YOLO files

# YOLOv4 config (small)
curl -O https://raw.githubusercontent.com/AlexeyAB/darknet/master/cfg/yolov4.cfg

# YOLOv4 weights (245 MB - takes a few minutes)
curl -LO https://github.com/AlexeyAB/darknet/releases/download/darknet_yolo_v3_optimal/yolov4.weights

# COCO class names
curl -O https://raw.githubusercontent.com/AlexeyAB/darknet/master/data/coco.names

3. Run again

mvn exec:java -Dexec.mainClass="com.auction.scraper.TroostwijkScraper"

Now you'll see:

✓ Object detection enabled with YOLO

The scraper will now analyze auction images and detect objects like:

• Vehicles (cars, trucks, forklifts)
• Equipment (machines, tools)
• Furniture
• Electronics
• And 80+ other object types

---

Features Without Object Detection

Even without YOLO, the scraper provides:

✅ Full auction scraping - Discovers all Dutch auctions ✅ Lot tracking - Monitors bids and closing times ✅ Desktop notifications - Real-time alerts ✅ SQLite database - All data persisted locally ✅ Image downloading - Saves all lot images ✅ Scheduled monitoring - Automatic updates every hour

Object detection simply adds:

• AI-powered image analysis
• Automatic object labeling
• Searchable image database

---

Database Location

The scraper creates troostwijk.db in your current directory with:

• All auction data
• Lot details (title, description, bids, etc.)
• Downloaded image paths
• Object labels (if detection enabled)

View the database with any SQLite browser:

sqlite3 troostwijk.db
.tables
SELECT * FROM lots LIMIT 5;

---

Stopping the Scraper

Press Ctrl+C to stop the monitoring service.

---

Next Steps

1. ✅ Run the scraper without YOLO to test it
2. ✅ Verify desktop notifications work
3. ⚙️ Optional: Add email notifications
4. ⚙️ Optional: Download YOLO models for object detection
5. 🔧 Customize: Edit monitoring frequency, closing alerts, etc.

---

Troubleshooting

Desktop notifications not appearing?

• Windows: Check if Java has notification permissions
• Linux: Ensure desktop environment is running (not headless)
• macOS: Check System Preferences → Notifications

OpenCV warnings?

These are normal and can be ignored:

WARNING: A restricted method in java.lang.System has been called
WARNING: Use --enable-native-access=ALL-UNNAMED to avoid warning

The scraper works fine despite these warnings.

---

Full Documentation

See README.md for complete documentation including:

• Email setup details
• YOLO installation guide
• Configuration options
• Database schema
• API endpoints