RuntimeTransaction::try_new_from_sanitized_transaction

[apfitzge]

Problem

• Want to make transition of processing functions easier by allowing simpler translation from an already created SanitizedTransaction

Summary of Changes

• add try_from_sanitized_transaction

Fixes #

[tao-stones]

What is the use case of this try_new? RuntimeTransaction<> encapsulates sanitizing transactions within. Wondering what situation makes sense to expose it? I mean if it can't follow lifetime of RuntimeTransaction<SanitizedVersionedTransaction> -> RuntimeTransaction<SanitizedTransaction>, I wonder if there are logical messiness on call site?

[apfitzge]

What is the use case of this try_new? RuntimeTransaction<> encapsulates sanitizing transactions within. Wondering what situation makes sense to expose it? I mean if it can't follow lifetime of RuntimeTransaction -> RuntimeTransaction, I wonder if there are logical messiness on call site?

• Many tests directly create SanitizedTransaction from Transaction (in Use RuntimeTransaction #3041 I add a test-only call to this for easier replacement).
• blockstore_processor calls Bank::verify_transaction which takes VersionedTransaction and returns SanitizedTransaction without ever using SanitizedVersionedTransaction. This will be updated in Use RuntimeTransaction #3041 to create RuntimeTransaction<SanitizedTransaction>.

I'd actually put pressure on the other way.
There's only one place I'm aware we even go to SanitizedVersionedTransaction as an intermediate step, which is in legacy BankingStage.
That is going away already, and at least until new transaction type is widely used I don't think we plan to have this use an intermediate RuntimeTransaction.