The library (and my python implementation) have made the assumption it is measured in bytes, but looking at it, I believe you are right, it should absolutely be in bits.

I will update the test cases and adjust the code as needed to support bits. The test cases were specifically generated on byte alignments, but I see now the intention of CBF.

This will be a major version bump.

Thanks! I'll get to this soon.

What's the status of this?

I haven't had time to get to this yet... It's been a crazy time with Ethereum.

If you create a pull request with the changes and add test cases though, I will verify and commit. Otherwise, it will probably still be a week or so before I can get to this.

I actually realized that this is no longer an issue of mine, as I can use the pkcs7 padding as my input has already been padded.

Any news on this one?

Sorry, haven't had a chance to get to this, and this weekend is big hackathon. I will add something to my calendar to look into this next week.

A fix for this would be much appreciated.

Any progress? I can't use CFB properly - it just returns me a clear random instead of encrypted/decrypted buffer