[Chore] Missing Revealed event on tholos-v2's auto-reveal path

[collinsezedike]

Summary

open_reveal_phase (contracts/tholos-v2/src/lib.rs:1322-1364) auto-reveals the asserter's and disputer's fixed positions the moment Reveal opens: it marks each position.revealed = true, sets position.agrees_with_outcome, and tallies their weight into resolution.agree_weight/disagree_weight (lib.rs:1340-1357). It publishes RevealOpened for the phase transition itself, but never a Revealed event for either of the two positions it just revealed.

By contrast, an external voter's explicit reveal() call does publish Revealed { id, voter, choice } (lib.rs:1562). An off-chain indexer or keeper reconstructing per-voter reveal state purely from Revealed events would silently miss the asserter's and disputer's votes, the two that are guaranteed to exist on every disputed assertion, even though the on-chain state itself is fully consistent.

Scope

• Emit a Revealed event (or equivalent) for each of the asserter's and disputer's fixed positions inside open_reveal_phase, alongside the existing RevealOpened event, so indexers see a complete set of reveals without needing to separately special-case the two protocol-fixed positions.
• No behavior change to phase transition or weight tallying logic itself.
• Add a test asserting both events are published when Reveal opens.

Proposed approach

Reuse the existing Revealed event struct, publishing one for assertion.asserter and one for disputer right after their Position writes in open_reveal_phase's loop, using each one's already-known agrees_with_asserter value as choice.

[ZacLou]

I'd like to work on this — opened a PR at #171.

[xtep103]

Hey Maintainer, I have gone through the details of the issue and i am capable of resolving it, I would love to be assigned to work on it ASAP. Thanks in advance

[ZacLou]

I'd like to take this on. The fix is straightforward: add Revealed event publishing in open_reveal_phase for the asserter and disputer's auto-revealed positions right after each position write in the loop at lib.rs:1340-1357 — publishing one Revealed { id, voter, choice } per fixed position using each one's already-known agrees_with_asserter as choice.

No behavior change to weight tallying or phase logic; this is a pure observability add. Will keep it to 5 lines diff. Ready to open a PR once assigned.