Home > other >  How to calculate hash while saving file to disk using streams?
How to calculate hash while saving file to disk using streams?

Time:08-09

How to calculate hash using streams while saving file to disk?

I don't want to: Save the file first and then load it from disk just to calculate the hash, at any point having to load the entire file into memory, use non async version of methods for which there is async counterpart or use API which is marked as obsolete in .NET 6 or higher.

This is what I have so far always getting "err"

  public async Task HashOnTheFly()
  {
    var path = "/tmp/a.txt";
    await SaveAsync(File.OpenRead(path), "/tmp/b.txt", default);

    async Task SaveAsync(Stream stream, string path, CancellationToken ct)
    {
      var sha512 = SHA512.Create();
      var fileName = Path.GetFileName(path);
      var destinationPath = Path.Combine("/tmp", fileName);
      await using var fileStream = File.Create(destinationPath);

      await using var cryptoStream = new CryptoStream(fileStream, sha512, CryptoStreamMode.Read);
      await stream.CopyToAsync(fileStream, ct);

      if (sha512?.Hash is { } computedHash)
        Console.WriteLine(computedHash);
      else
        Console.WriteLine("err");
    }
  }

CodePudding user response:

You have a few bugs in your code:

  1. You never copy to the CryptoStream, you only copy to the underlying fileStream. So naturally no hash is ever calculated.

  2. You do not close the CryptoStream before attempting to determine the hash. The stream must be closed first to ensure all data is computed and flushed.

  3. Since you are computing your hash as you write, you must use CryptoStreamMode.Write not CryptoStreamMode.Read.

  4. SHA512 implements IDisposable and so should be disposed of via a using statement.

Thus your SaveAsync should be modified as follows:

async Task<byte []> SaveAsync(Stream stream, string path, CancellationToken ct)
{
    var fileName = Path.GetFileName(path);
    var destinationPath = Path.Combine("/tmp", fileName);

    using var sha512 = SHA512.Create();
    await using (var fileStream = File.Create(path, 0, FileOptions.Asynchronous))
    await using (var cryptoStream = new CryptoStream(fileStream, sha512, CryptoStreamMode.Write))
    {
        await stream.CopyToAsync(cryptoStream, ct);
    }
    
    if (sha512?.Hash is { } computedHash)
        Console.WriteLine(Convert.ToBase64String(computedHash));
    else
        Console.WriteLine("err");
    
    return sha512.Hash;
}

Notes:

  • As you are doing asynchronous I/O you may want to open the fileStream with FileOptions.Asynchronous to indicate that the file can be used for asynchronous reading and writing.

  • I modified your code to return the computed hash, as it seems logical to do so.

Demo fiddle here.

  • Related