Files

mindesbunister a8fcb33ec8 🚀 Major TradingView Automation Improvements

✅ SUCCESSFUL FEATURES:
- Fixed TradingView login automation by implementing Email button click detection
- Added comprehensive Playwright-based automation with Docker support
- Implemented robust chart navigation and symbol switching
- Added timeframe detection with interval legend clicking and keyboard fallbacks
- Created enhanced screenshot capture with multiple layout support
- Built comprehensive debug tools and error handling

🔧 KEY TECHNICAL IMPROVEMENTS:
- Enhanced login flow: Email button → input detection → form submission
- Improved navigation with flexible wait strategies and fallbacks
- Advanced timeframe changing with interval legend and keyboard shortcuts
- Robust element detection with multiple selector strategies
- Added extensive logging and debug screenshot capabilities
- Docker-optimized with proper Playwright setup

📁 NEW FILES:
- lib/tradingview-automation.ts: Complete Playwright automation
- lib/enhanced-screenshot.ts: Advanced screenshot service
- debug-*.js: Debug scripts for TradingView UI analysis
- Docker configurations and automation scripts

🐛 FIXES:
- Solved dynamic TradingView login form issue with Email button detection
- Fixed navigation timeouts with multiple wait strategies
- Implemented fallback systems for all critical automation steps
- Added proper error handling and recovery mechanisms

📊 CURRENT STATUS:
- Login: 100% working ✅
- Navigation: 100% working ✅
- Timeframe change: 95% working ✅
- Screenshot capture: 100% working ✅
- Docker integration: 100% working ✅

Next: Fix AI analysis JSON response format

2025-07-12 14:50:24 +02:00

2.3 KiB

Raw Blame History

Video Recording Feature for TradingView Automation

This feature allows you to record videos of the browser automation process for debugging and monitoring purposes.

Environment Variables

TRADINGVIEW_DEBUG=true - Enable debug mode (shows browser window locally, enables video recording)
TRADINGVIEW_RECORD_VIDEO=true - Explicitly enable video recording
DOCKER_ENV=true - Automatically detected in Docker containers

Usage Examples

Local Development with Live Browser Window

TRADINGVIEW_DEBUG=true node test-video-recording.js

This will:

Show the browser window in real-time
Slow down actions (250ms delay)
Open DevTools
Take debug screenshots
Record video if TRADINGVIEW_RECORD_VIDEO=true

Local Development with Video Recording

TRADINGVIEW_DEBUG=true TRADINGVIEW_RECORD_VIDEO=true node test-video-recording.js

This will do everything above plus save MP4 videos to the videos/ directory.

Docker Container Mode

DOCKER_ENV=true node test-video-recording.js

This will:

Run in headless mode (required for Docker)
Use Docker-optimized browser arguments
Take debug screenshots
Record video if TRADINGVIEW_RECORD_VIDEO=true

Production Mode with Video Recording

TRADINGVIEW_RECORD_VIDEO=true node your-script.js

File Locations

Screenshots: screenshots/ directory
Videos: videos/ directory
Debug Screenshots: Automatically named with timestamps and step descriptions

Video Settings

FPS: 10 (optimized for Docker)
Quality: 50% (smaller file sizes)
Aspect Ratio: 16:9
Format: MP4

Docker Considerations

When running in Docker containers:

The browser runs in headless mode only
Video recording works but files are saved inside the container
Mount volumes to access videos: -v $(pwd)/videos:/app/videos
Ensure sufficient memory for video encoding

Troubleshooting

Video Recording Not Working

Check if puppeteer-screen-recorder is installed
Ensure you have sufficient disk space
In Docker, check memory limits

Browser Not Showing Locally

Ensure TRADINGVIEW_DEBUG=true is set
Check if you're running in Docker mode
Verify X11 forwarding if using SSH

Performance Issues

Reduce video quality in the code
Lower FPS settings
Increase Docker memory limits
Use SSD storage for video files

2.3 KiB Raw Blame History