nonce staking guards, iter close

[xmariachi]

Purpose of Changes and their Description

Includes: nil guards in nonce paths, safer reputer nonce deref, prune-worker no-op consistency, missing iter.Close().

Are these changes tested and documented?

• If tested, please describe how. If not, why tests are not needed. -- unit tests modified accordingly
• If documented, please describe where. If not, describe why docs are not needed. -- no need, no func change
• Added to Unreleased section of CHANGELOG.md? -- no need , no func change

---

Summary by cubic

Modularized the emissions keeper into focused sub-keepers and added robust nonce/staking guards to prevent nil derefs and window inconsistencies. No external message or query API changes; improves safety, clarity, and testability.

• Refactors

  • Split monolithic Keeper into sub-keepers: Params, Topic, Staking, Scores, Nonce, Regrets, ReputerLoss, Weights, Banking, and ActorPenalties.
  • Rewired Msg/Query servers and tests to use injected sub-keepers and simple getters (e.g., GetParamsKeeper()).
  • Moved EMA score logic, weights, staking, topics, regrets, and nonce management into dedicated files with constructor funcs.

• Bug Fixes

  • Added nil input checks and nil-safe iteration across nonce methods (Fulfill/Add/Is*/Prune) to prevent panics; PruneWorkerNonces now no-ops when empty.
  • Closed a missing iterator in staking (GetDelegateStakeRemovalForDelegatorReputerAndTopicId).

^{Written for commit 903d5df. Summary will update on new commits.}

[greptile-apps]

Too many files changed for review. (131 files found, 100 file limit)

[cubic-dev-ai]

No issues found across 3 files