Implement proper clipboard error handling and retry logic

[MartinPTielemans]

Description

The clipboard monitoring functionality lacks robust error handling and could fail silently in various scenarios.

Current issues:

// In command.rs - basic error handling
match self.clipboard_ctx.set_text(translated.clone()) {
    Ok(_) => { /* success */ }
    Err(e) => {
        eprintln!("Failed to update clipboard with translated text: {}. Error: {}", translated.clone(), e);
        // Error is logged but monitoring continues without user notification
    }
}

Potential failure scenarios:

1. Clipboard access permission denied
2. Clipboard locked by another application
3. System clipboard service unavailable
4. Memory issues with large clipboard content
5. Race conditions with rapid clipboard changes

Proposed improvements:

Error Classification

#[derive(Debug)]
enum ClipboardError {
    PermissionDenied,
    ServiceUnavailable,
    ContentTooLarge,
    TemporaryFailure,
}

Retry Logic

• Implement exponential backoff for temporary failures
• Different retry strategies for different error types
• Maximum retry limits to prevent infinite loops

User Feedback

• Notify users when clipboard access fails
• Show status indicators for clipboard health
• Provide troubleshooting guidance

Graceful Degradation

• Continue monitoring even after individual failures
• Fallback mechanisms for critical operations
• Recovery procedures for persistent failures

Benefits:

• More reliable clipboard monitoring
• Better user experience with clear error communication
• Reduced support burden with self-diagnosing issues
• Production-ready robustness for various system states

This would make the core functionality much more reliable across different macOS configurations and usage scenarios.