refactor: Made DatabaseStorageProof stateful

[18aaddy]

Resolves #15477

[18aaddy]

@Rjected @mattsse Can you review pls?

[Rjected]

I have some suggestions that should simplify how we use these methods, lmk if you have any questions or are stuck with anything

[Rjected]

do these need to be pub?

[Rjected]

imo we should implement a method on StorageProof that is identical to the previous from_tx implementation, then we should be able to do something like:

StorageProof::from_tx(self.tx()).overlay_storage_multiproof(address, slots, hashed_storage)

[18aaddy]

@Rjected , I have updated my code with the requested changes. Here is an explainer:

• I have added a new trait StorageProofFromTx with a function from_tx() that gives a StorageProof instance.

• The reason for defining a new trait is that StorageProof is used in the trie/db/src/proof.rs file and the DatabaseTrieCursorFactory and DatabaseHashedCursorFactory structs required for implementing from_tx() are also present in trie/db/src. If I were to implement the from_tx() function for StorageProof, it would cause a circular import.

• The trie_cursor_factory and hashed_cursor_factory fields are still public. The reason behind this is that in the with_hashed_cursor_factory() function below, the StorageProof instance is consumed, therefore we need to make a new instance. The StorageProof struct doesn't implement the Copy trait, and therefore a simple .clone() doesn't work.

reth/crates/trie/db/src/proof.rs

Lines 119 to 125 in 97518b3

	Self::new(self.trie_cursor_factory.clone(), self.hashed_cursor_factory.clone(), address)
	.with_hashed_cursor_factory(HashedPostStateCursorFactory::new(
	self.hashed_cursor_factory.clone(),
	&state_sorted,
	))
	.with_prefix_set_mut(prefix_set)
	.storage_proof(slot)

This is the best solution I was able to find, if there was something more obvious possible, my apologies. Would love some feedback.

[18aaddy]

@Rjected @mattsse Can you pls review?

[jenpaff]

@Rjected could you ptal?