Fix/get block transactions count

[tanishqjasoria]

Fixes getBlockTransactionCount when blockParameter is "pending". It did not take in account the pending transactions.

Changes:

Types of changes

What types of changes does your code introduce?
Put an x in the boxes that apply

• Bugfix (non-breaking change which fixes an issue)
• New feature (non-breaking change which adds functionality)
• Breaking change (fix or feature that would cause existing functionality to not work as expected)
• Documentation Update
• Code style update (formatting, renaming)
• Refactoring (no functional changes, no api changes)
• Build related changes
• Other (please describe):

Testing

Requires testing

• Yes
• No

In case you checked yes, did you write tests??

• Yes
• No

[LukaszRozmej]

So this doesn't exactly do what it is stating. It only takes the nonce of the transaction from the pool. Those transactions don't have to be included in the next block. (Unless it should work like that just by giving you what is in the pool instead of what is in potential next block)

The main problem is that Nethermind doesn't have proper support for pending block. It should construct that block and return results based on that. That was never implemented. Also Geth team is keen on deprecating and removing the pending flag altogether.

Taking all that into account I'm more keep on just rejecting this change.

[tanishqjasoria]

So this doesn't exactly do what it is stating. It only takes the nonce of the transaction from the pool. Those transactions don't have to be included in the next block. (Unless it should work like that just by giving you what is in the pool instead of what is in potential next block)

The main problem is that Nethermind doesn't have proper support for pending block. It should construct that block and return results based on that. That was never implemented. Also Geth team is keen on deprecating and removing the pending flag altogether.

Taking all that into account I'm more keep on just rejecting this change.

So should we just tell users to manage nonce locally? I still think there should be a way using which users can query if there are pending transactions from a particular address.
I understand the logic behind the "pending" block, but in most of the places getTransactionCount with pending parameter is used to get the value of next nonce that should be used for the next transaction.

[LukaszRozmej]

How Geth implements it?

[tanishqjasoria]

How Geth implements it?

pool.pendingNonces.get(addr) - they do something similar. They have a dict in txPool where they store pending nonces.

[tanishqjasoria]

https://github.com/ethereum/go-ethereum/blob/master/internal/ethapi/api.go#L1518

[LukaszRozmej]

1. Don't use _nonce, that was added only for ManagedNonce transactions explicitly.
2. If you need you can go through a bucket explicitly in _transactions.
3. What about potential gaps between nonces?

[tanishqjasoria]

The way pending nonce is maintained in geth, they return the next available nonce.
If there is a gap, like there are txns in pool with nonces [2,3,5,6], then geth will return 4 as the next available nonce.

[LukaszRozmej]

ok but it's maintained in just a dictionary? It is able to maintain that there and ignore gaps correctly?
On our side we can implement it by grabbing a bucket with transactions for address and do a simple scan there (maybe first check last item vs the count too to avoid it for non-gap cases)

[tanishqjasoria]

They do is somewhat differently. They maintain two levels in the txPool (pending and all). One where all the transactions are present and other where only the transactions that can be added in any future block. Like the transactions with sequential nonce, if there is a gap in nonce, the rest of the txn are not added in the pending pool. And the pendingNonces dict is only maintained for txns in the pending pool.

Since our system is quite different, we can do it by scanning the txPool.

[LukaszRozmej]

I pushed a fix and optimizations.

1. Current nonce from state was ignored. This resulted in incorrect values being returned.
2. Avoid copying a bucket when doing scan.
3. Avoid full collection scan if no gap.

@tanishqjasoria This needs more tests and a review.
I would prefer fairly straightforward test through RPC and sophisticated tests on TxPool.

[LukaszRozmej]

Please run a node for a while + spam it with request a bit