Modify foldBlocks to recurse on ledger events

[Jimbo4350]

Changelog

- description: |
    Modify foldBlocks to recurse on ledger events instead of mapping over them. This allows finer grained control over when `foldBlocks` is stopped.
# uncomment types applicable to the change:
  type:
  # - feature        # introduces a new feature
  # - breaking       # the API has changed in a breaking way
  # - compatible     # the API has changed but is non-breaking
  - optimisation   # measurable performance improvements
  # - improvement    # QoL changes e.g. refactoring
  # - bugfix         # fixes a defect
  # - test           # fixes/modifies tests
  # - maintenance    # not directly related to the code
  # - release        # related to a new release preparation
  # - documentation  # change in code docs, haddocks...

Context

Additional context for the PR goes here. If the PR fixes a particular issue please provide a link to the issue.

How to trust this PR

Highlight important bits of the PR that will make the review faster. If there are commands the reviewer can run to observe the new behavior, describe them.

Checklist

• Commit sequence broadly makes sense and commits have useful messages
• New tests are added if needed and existing tests are updated. See Running tests for more details
• Self-reviewed the diff

[carbolymer]

It feels that we should use an underlying monad in ClientStNext to store state instead of using IORef.

CSP.ClientStNext n BlockInMode ChainPoint ChainTip (StateT s IO) ()

would be cleaner I think.

[Jimbo4350]

Can you open a PR after this one? I'd be interested to see what you have in mind.

[carbolymer]

I took a closer look and it does not seem to be worth the effort. My point was to be able to use state monad in the protocol client to store the state (including an error). The problem with that is the change is quite extensive: we'd have to parameterize types used in the client e.g. LocalNodeClientProtocolsForBlock by an underlying monad additionally.
Moreover, the code in consensus we use has IO hardcoded there in the function argument e.g.:

connectTo :: forall a b.
LocalSnocket
-> NetworkConnectTracers LocalAddress NodeToClientVersion
-> Versions
     NodeToClientVersion
     NodeToClientVersionData
     (OuroborosApplicationWithMinimalCtx
        'InitiatorMode LocalAddress ByteString IO a b)
-> FilePath
-> IO ()

This would require unlifting state into IO and then lifting it again, which is cumbersome.

In the end the semantics would be pretty much the same, but with more boilerplate.