Add file format version and committed links counter to LinksHeader for split memory recovery

[konard]

🤖 AI-Powered Solution

This pull request implements the solution for issue #179 by enhancing the LinksHeader structure to contain sufficient information for split memory implementation recovery.

📋 Issue Reference

Fixes #179

✅ Implementation Details

The solution adds essential fields to the LinksHeader<TLinkAddress> structure to enable index file recovery when only the data file is available:

New Header Fields

1. FormatVersion - Tracks file format compatibility to ensure proper handling of data structure changes
2. CommittedLinks - Counter to support multi-threaded operations as described in the issue comments

Key Changes

LinksHeader.cs:

• Added FormatVersion field for file format tracking
• Added CommittedLinks field for multi-threaded support
• Updated Equals() method to include new fields
• Updated GetHashCode() method to include new fields

SplitMemoryLinksBase.cs:

• Added CurrentFormatVersion constant set to TLinkAddress.One
• Enhanced Init() method to properly initialize new header fields:
  • Sets FormatVersion to current version if not already set
  • Initializes CommittedLinks to current AllocatedLinks value if not set

Recovery Capability

With these changes, the header now contains enough information to recreate the index file if it is lost or dropped. The essential fields (FormatVersion, AllocatedLinks, CommittedLinks) provide the foundation for reconstruction, while other fields (FreeLinks, FirstFreeLink, RootAsSource, RootAsTarget, LastFreeLink) can be calculated through a one-time scan of the data file as intended.

Multi-threaded Support Foundation

The CommittedLinks field provides the groundwork for the multi-threaded link allocation pattern described in the issue:

• Threads can increment AllocatedLinks to reserve address space
• Threads track their lastAllocatedLinks value
• When lastAllocatedLinks == CommittedLinks and insertion is complete, threads can increment CommittedLinks
• This ensures all links are actually committed correctly

🧪 Testing

• ✅ Project builds successfully without errors
• ✅ All existing functionality preserved through backward compatibility
• ✅ New fields are properly initialized in existing and new data stores

🔄 Backward Compatibility

The implementation maintains full backward compatibility:

• Existing data files will have new fields initialized to appropriate defaults
• No breaking changes to existing APIs
• Existing functionality remains unaffected

---

This PR was created automatically by the AI issue solver