Add in ability to encrypt/decrypt msg with keyring

[RoyTimes]

Ref: #1068

• 1 dependency added to 'util-crypto' - ed2curve by the author of tweetnacl.js
• tests included
• covers sr25519 & ed25519, secp256k1 to be added.

[jacogr]

Funky hack to make this operational in sr/ed/eth.

Some small comments.

The only think I feel is missing is possibly explicit tests for various keypair types -

• Ethereum
• sr25519
• ed25519

[RoyTimes]

I have included test cases for cross sr & ed. I'm fairly positive if the tests can be added and work as expected in a similar fashion as the test for cross sr & ed. However, I'm not sure if it's the right cryptographic thing to do when forcibly converting private keys of the same size from x25519 to secp256k1. Two options here: add in an assert to reject usage on secp256k1 curve; OR just do the conversion and compose test cases for them. What do you think?

[jacogr]

As a first-step, rather play safe and assert them away.

We can re-visit if need be or add an alternative approach for them (which would probably be the correct approach here).

[RoyTimes]

Done.

On a better non-hacky approach: I have no doubt that all of ed25519, sr25519 and secp256k1 can do ECDH; on a worse side, the library for secp256k1 supports ECDH but not encryption/decryption and I might have to implement a portion of the cryptographic myself. On a better side, introduce new dependencies that is well trusted to do ed25519 sr25519 ECDH. Tweetnacl does not have it built in. For sr25519, that will involve messing with the Rust implementation.

Long story short, it's a pretty heavy load of work to be done. I'm happy to revisit this when I'm more available.

When this is merged, I'll jump to extension.

[jacogr]

LGTM. That you very much.

Like mentioned, this is an interesting approach, certainly not something I thought about to get over the sr25519-doesn't-have-that-in-the-libs (yet).

[polkadot-js-bot]

This pull request has been automatically locked since there has not been any recent activity after it was closed. Please open a new issue for related bugs.