annotations/Azaion.Common/Services/HashExtensions.cs

using System.IO;
using System.IO.Hashing;
using System.Text;

namespace Azaion.Common.Services;

public static class HashExtensions
{
    private const int SampleSize = 1024;
    private const int HashLength = 9;
    private const string VirtualHashPrefix = "V";

    public static string CalcHash(this FileInfo file)
    {
        if (!file.Exists)
            return CalcVirtualHash(file.Name);

        var fileSize = file.Length;
        if (fileSize < SampleSize * 8)
            return CalcHashInner(File.ReadAllBytes(file.FullName));
        
        var combinedData = new byte[8 + SampleSize * 3];
        var offset = 0;

        var fileSizeBytes = BitConverter.GetBytes(fileSize);
        Array.Copy(fileSizeBytes, 0, combinedData, offset, 8);
        offset += 8;

        using var stream = new FileStream(file.FullName, FileMode.Open, FileAccess.Read, FileShare.Read);

        stream.ReadExactly(combinedData, offset, SampleSize);
        offset += SampleSize;

        stream.Position = (fileSize - SampleSize) / 2;
        stream.ReadExactly(combinedData, offset, SampleSize);
        offset += SampleSize;

        stream.Position = fileSize - SampleSize;
        stream.ReadExactly(combinedData, offset, SampleSize);

        return CalcHashInner(combinedData);
    }

    public static string CalcVirtualHash(string filename)
    {
        var bytes = Encoding.UTF8.GetBytes($"VIRTUAL:{filename}");
        var hash = XxHash64.HashToUInt64(bytes);
        var hashBytes = BitConverter.GetBytes(hash);
        var base64 = Convert.ToBase64String(hashBytes);
        return VirtualHashPrefix + base64[..HashLength];
    }

    public static bool IsVirtualHash(this string? hash) => 
        !string.IsNullOrEmpty(hash) && hash.StartsWith(VirtualHashPrefix);

    private static string CalcHashInner(byte[] data)
    {
        var hash = XxHash64.HashToUInt64(data);
        var hashBytes = BitConverter.GetBytes(hash);
        var base64 = Convert.ToBase64String(hashBytes);
        return base64[..HashLength];
    }
}