[feat] Implement Utxo set flow separated from Utrexo accumulator updates

[m-kus]

Currently Utreexo state is part of the UtxoSet although it would be better to separate those two to have a cleaner design and solve some issues with batch inclusion proofs, cache integrity, and double spend prevention.

Here are the suggested changes:

• UtxoSet no longer includes Utreexo state
• instead it is extended with two new arrays: leaves_to_delete (similar to leaves_to_add) and cache_keys (to ensure all cached outputs are spent in the end)
• outpoint hashing should be refactored as a trait method
• note that outpoint hash has to include extra data — current block hash (see https://eprint.iacr.org/2019/611.pdf - hardening against collision attacks)
• we need to introduce block hash to the Header struct and change all scripts/tests accordingly (note that we also need to add a constraint to ensure the computed block hash equals the one we provide)

Utreexo accumulator updates / inclusion proof verification is "disabled" for now.

[m-kus]

@Groxan wanna follow up?

[ShantelPeters]

Please @m-kus can I work on this?

[onlydustapp[bot]]

Hey @ShantelPeters! Thanks for showing interest. We've created an application for you to contribute to Raito - Bitcoin ZK Client. Go check it out on OnlyDust!

[Groxan]

@m-kus would be glad :)