Keysets V2

[a1denvalu3]

Description

Adding support for keysets v2.

CHANGED

• Replaced From trait implementation for Id with functions v1_from_keys, v2_from_data.
• struct Id, id field:

#[derive(Clone, Copy, PartialEq, Eq, PartialOrd, Ord, Hash)]
pub enum IdBytes {
    V1([u8; 7]),
    V2([u8; 31]),
}

pub struct Id {
    version: KeySetVersion,
    id: IdBytes,
}

• keyset generation functions (generate, generate_from_xpriv, create_new_keyset) take 2 additional arguments: an optional final_expiry and a version to identify the correct version.
• creating new keysets defaults to KeySetVersion::Version01 (creating new keysets defaults to the newer one).

TODO

• ShortIdV2 for short IDs in tokens
• Fix database interactions.

---

Checklist

• I followed the code style guidelines
• I ran just final-check before committing

[ok300]

Looking good.

Some tests would be nice as well, converting to and from byte arrays of different lengths.

[a1denvalu3]

So @ok300 @thesimplekid I thought we could create a new ShortId enum which will be the type of the id field for TokenV4Token. In the case of v1, it will just be the keyset ID itself, while in the case of v2 it will be the first 8 bytes of IDv2.

[thesimplekid]

So @ok300 @thesimplekid I thought we could create a new ShortId enum which will be the type of the id field for TokenV4Token. In the case of v1, it will just be the keyset ID itself, while in the case of v2 it will be the first 8 bytes of IDv2.

Yes I think that makes sense. It's similar to how we handled the transition from TokenV3 to TokenV4.

[ok300]

Sounds good. Maybe ShortKeysetId is clearer.

[thesimplekid]

merging main should fix the current CI issues, I updated the stable rust version.

[a1denvalu3]

I think this is good.

[a1denvalu3]

Do we want to merge this?