Update cryptoswift 1.5.1 and fix EthereumPublicKey initializer

[DmitryBespalov]

Updated dependency and fixed bug in the public key initializer

[wainglaister]

It would be good if there was a test to prove the public key recovery fix (the fix is good, just missing the test) :)

[koraykoska]

What does not work with the public key recovery process right now? Why did you change it?

[DmitryBespalov]

@wainglaister ok, will add a test.

@koraykoska it recovers to a wrong address because it uses hash of a wrong preimage: it was hashing a signature instead of a message, which is nonsense to me 😊

I think it's a typo, but without the fix the method produces wrong values

[wainglaister]

Relates to this issue: https://github.com/Boilertalk/Web3.swift/issues/12

[koraykoska]

@DmitryBespalov Can you please add test cases from official specifications so we avoid future regressions

[DmitryBespalov]

@koraykoska sorry, I'm overloaded with the day job tasks at the moment, I won't have time to look at it till the middle of next week

[proggeramlug]

I need this in production, could someone release this please?

[koraykoska]

@proggeramlug CryptoSwift was updated and released. I still fail to find official examples about the recovery of the public key and I think the implementation in this PR is wrong. If you can, please add test cases and link to official docs so I can get it merged.

[podkovyrin]

I would suggest this as a trustworthy source: https://github.com/ethereumjs/ethereumjs-monorepo/blob/master/packages/util/src/signature.ts#L53

Tests: https://github.com/ethereumjs/ethereumjs-monorepo/blob/master/packages/util/test/signature.spec.ts#L101

Reference: Yellow Paper, see page 26, Appendix F, line (314)