Add asynchronous archive creation with concurrent file processing

[Copilot]

This PR implements optional asynchronous archive creation for the go-pxar library, allowing concurrent file processing while maintaining full compatibility with the existing synchronous operations and PXAR format requirements.

Features Added

Async Configuration Options

pa := PBSArchive{
    Filename:     "backup.pxar",
    AsyncMode:    true,        // Enable async processing
    AsyncWorkers: 4,           // Use 4 concurrent workers (optional, defaults to CPU count)
}

New Async Methods

• ToBufferAsync() - Asynchronous archive creation to buffer
• ToChannelAsync() - Asynchronous archive creation to channel
• Extended NodeRef interface with async methods for all node types

Usage Example

// Synchronous (existing behavior)
err := pa.ToBuffer(buf)

// Asynchronous (new functionality)  
err := pa.ToBufferAsync(buf)

// Both produce identical results

Technical Implementation

The async implementation uses a worker pool pattern where:

1. Concurrent Processing: Multiple files are read and processed simultaneously using configurable worker goroutines
2. Ordered Assembly: Results are collected and buffered in the correct order to maintain PXAR format requirements
3. Sequential Writing: Final archive writing maintains proper positioning and goodbye record calculations
4. Recursive Async: Nested directories use async methods recursively for maximum performance benefit

Key Benefits

• Performance: Concurrent file I/O processing for improved throughput on multi-file archives
• Format Compliance: Maintains exact PXAR format requirements with correct ordering
• Identical Results: Async and sync methods produce byte-for-byte identical archives (verified by comprehensive tests)
• Backward Compatibility: All existing code continues to work unchanged
• Configurable: Worker count can be tuned based on system capabilities

Testing

Comprehensive test suite added including:

• Sync vs async result validation (identical output verification)
• Multiple worker configuration testing (1, 2, 4, 8 workers)
• Consistency testing across multiple async runs
• Performance benchmarking
• Large file handling validation

All existing tests pass with no regressions.

Documentation

Updated README.md with:

• Usage examples for async functionality
• Configuration options explanation
• Performance characteristics description

Fixes #2.

---

💡 You can make Copilot smarter by setting up custom instructions, customizing its development environment and configuring Model Context Protocol (MCP) servers. Learn more Copilot coding agent tips in the docs.