- Abstract TxProcessor from StateDB
- Upgrade to last version of go-merkletree for the key-value DB usage

Introduce a constructor parameter for the StateDB called `keep`, which tells
how many checkpoints to keep.  When doing a new checkpoint, if the number of
existing checkpoints exeeds `keep`, the oldest ones will be deleted.

Update usage of `*babyjub.PublicKey` to `babyjub.PublicKeyComp`
- when the key is not defined, internally is used `babyjub.EmptyBJJComp`, which is a `[32]byte` of zeroes of type `babyjub.PublicKeyComp`
- the API continues returning `nil` when the key is not defined

Checked with commonjs version:
c8414bd205

In statedb:
- Store all values using prefixes for keys to allow iteration
- Add methods MTGetRoot, GetAccounts

Implement debugapi, an http server with debugging endpoints:
- debugAPI.GET("sdb/batchnum", a.handleCurrentBatch)
- debugAPI.GET("sdb/mtroot", a.handleMTRoot)
- debugAPI.GET("sdb/accounts", a.handleAccounts)
- debugAPI.GET("sdb/accounts/:Idx", a.handleAccount)

* WIP: rebase

* Make nullable fields use pointers

Add StateDB compute nonces on ProcessTx L2Tx, and update StateDB for
type TypeSynchronizer, TypeTxSelector, TypeBatchBuilder

StateDB intermediate state reset when opening DB to force getting always last
Checkpoint at last BatchNum, avoiding inconsistent intermediate state.

Add linter checks to GHA & Fix code to pass lints.
The linters added are:
- whitespace: Tool for detection of leading and trailing whitespace
- gosec: Inspects source code for security problems
- gci: Gci control golang package import order and make it always deterministic
- misspell: Finds commonly misspelled English words in comments
- gomnd: An analyzer to detect magic numbers

The file utils/utils.go is excluded from the checks of gomnd, as uses magic numbers through the code

Update StateDB to new & ExitInfo struct & BatchNum
Also a small fix at txselector & log packages

- Update GHA lint.yml increasing timeout time to avoid GHA Lint errors
- Update common.BatchNum & common.Idx & common.Nonce usage in StateDB
- Add abstraction method of processTxs to StateDB
    - Which will be used by Synchronizer & BatchBuilder

Previously as the txprocessor.go methods were specific for the BatchBuilder,
the MTCreateAccount & CreateAccount and MTUpdateAccount & UpdateAccount were
also designed to be used by BatchBuilder and TxSelector depending on the
MerkleTree usage calling one kind of method or anotherone.

But now that this methods are being called directly by the StateDB (through the
methods in txprocessors.go), to allow also the methods usage from the
Synchronizer, there can not be the MT and no-MT methods separated, so this
commit unifies MTCreateAccount with CreateAccount, and MTUpdateAccount with
UpdateAccount, which internally will update the MerkleTree depending if the
specific StateDB in usage has the MerkleTree defined or not.