Refactor of batching stores and APIs

[hacheigriega]

Motivation

• Tree entry store refactor
• Merge batch signature store with validator tree entry store
• Batch query now returns its tree entries and signatures as well.
• Remove separate query endpoints for batch signatures and tree entries.
• Querying data result also returns its batch assignment, if exists. (nil if a batch has not been assigned to a data result)
• Fix logic for retrieving latest batch whose signatures have been collected
• Add a flag for querying batches without signatures
• Update Swagger docs

Explanation of Changes

There is no longer a separate query endpoint for batch signatures or tree entries. To obtain them, query the corresponding batch instead:

❯ sedad query batching batch 3 -o json | jq
{
  "batch": {
    "batch_number": "3",
    "block_height": "23",
    "current_data_result_root": "c5d2460186f7233c927e7db2dcc703c0e500b653ca82273b7bfad8045d85a470",
    "data_result_root": "56c4f39b7564ea6a32877fc98743652998753c4cd8a4b455c26dcb3b92774b73",
    "validator_root": "96452cb1ee0deb9c44e992f6b1d807abc00e216fb4b0c4a85d54ec854db024bc",
    "batch_id": "YsZCbHoVhEk/lBg7MxZjyekdUsgMGbYYwQHKd7JPAv4=",
    "proving_metadata": null
  },
  "data_result_entries": {
    "entries": []
  },
  "validator_entries": [
    {
      "validator_address": "sedavaloper14aknletrd5wgys5ypchrp463txvlq27pelzg7j",
      "voting_power_percent": 50000000,
      "eth_address": "W7bYkKmLRmI9wHXcstpF0VCgHkk="
    },
    {
      "validator_address": "sedavaloper1hmkpuzyq8f2nzvwxltne34yk956mft34yd7s6l",
      "voting_power_percent": 50000000,
      "eth_address": "DJ/0RmJT0bSiIOwMCEHIQzU8tJw="
    }
  ],
  "batch_signatures": [
    {
      "validator_address": "sedavaloper15xgmjrur7lr48ye0tfeu30cjs8evrxd77rf8c8",
      "secp256k1_signature": "y4C9kJzRaEZM2e50enCRYUUzfjt3otRcjF9IR5Ma1MEyIRiwgvJ2FzmwuDyd75rj+MYZPNBWdVLL1TsLkM14YQE="
    },
    {
      "validator_address": "sedavaloper14aknletrd5wgys5ypchrp463txvlq27pelzg7j",
      "secp256k1_signature": "Evro8bGJwBWKCEZvT+t5hJhw1SqEJSgfsoht/uzaz2lgh5yUwDLG05HIY7QeHPDVWnM+l+W7C6Lv0B3dKnwUcQE="
    },
    {
      "validator_address": "sedavaloper1hmkpuzyq8f2nzvwxltne34yk956mft34yd7s6l",
      "secp256k1_signature": "nTcVAzt8R7WwjvMn7X5vGWagxOcD2Ft//9d81TScEpc8goL7Pagvw5EvuG0fDChBNUVsoXxAa3889tlLKE+05gE="
    }
  ]
}

To query the latest batch whose signatures have been collected, make the same query without any arguments.

List of batches in the store can now be queried with the flag --with-unsigned to include new batches without signatures.

❯ sedad query batching batches
❯ sedad query batching batches --with-unsigned

Also note that validator tree entries are now stored separately from data result tree entries in a struct:

// ValidatorTreeEntry is an entry in the validator tree.
type ValidatorTreeEntry struct {
	ValidatorAddress   sdk.ValAddress 
	VotingPowerPercent uint32
	EthAddress         []byte
}

// Entries of data result tree of a given batch.
type DataResultTreeEntries struct {
	Entries []byte
}

// BatchSignatures contains basic validator data and its batch signatures
// under various cryptographic schemes.
type BatchSignatures struct {
	ValidatorAddress   sdk.ValAddress
	Secp256K1Signature []byte              
}

Store schema:

0x05 | batch_num | operator    -> BatchSignatures
0x06 | batch_num               -> DataResultEntry
0x07 | batch_num | operator    -> ValidatorEntry

Related PRs and Issues

Closes #423

[gluax]

It doesn't have to be an eth address right? Could be any evm network, and even then I don't think it's limited to only evm networks.

[hacheigriega]

Right but I think we will have to create a new struct every time we add a new signature scheme

[gluax]

Would make more sense to have this labeled with the scheme then? Since different chains could use the same scheme.

[hacheigriega]

eth_address means more like ethereum-style address format. So different chains could still use this if they use this address format.

[Thomasvdam]

Should we not also remove the batching params key?

[hacheigriega]

Should we not also remove the batching params key?

Ah yes I will remove it before I merge this PR