refactor: Reducing code redundancy

[SimonRastikian]

Tiny cleanup. Creating a helper function and reducing code redundancy.

[claude]

PR title type suggestion: This PR changes only test files, so the type prefix should probably be test: instead of refactor:.

Suggested title: test: reduce code redundancy in request_during_resharing

[claude]

Pull request overview

Consolidates four repeated domain-lookup blocks in test_request_during_resharing into a single helper (create_domain) that finds a domain by its Protocol. Behavior is preserved because each Protocol variant appears at most once in this test's domain configuration (CaitSith, DamgardEtAl, Frost, ConfidentialKeyDerivation), so the previously additional DomainPurpose / Curve filters were redundant.

Changes:

• Introduced create_domain(&RunningContractState, Protocol) -> DomainConfig helper.
• Replaced four inline .iter().find(...) blocks with helper calls.
• Removed now-unused Curve import; added RunningContractState import.

Reviewed changes

Per-file summary

File	Description
crates/e2e-tests/tests/request_during_resharing.rs	Extracted domain lookup helper and replaced repeated find blocks. Adjusted imports.

Findings

Blocking: none.

Non-blocking (nits, follow-ups, suggestions):

• crates/e2e-tests/tests/request_during_resharing.rs:14 — The name create_domain is misleading: the function does not construct or register a domain, it looks one up in the existing contract state. Consider renaming to find_domain or domain_for_protocol so future readers aren't left wondering whether this mutates state.
• crates/e2e-tests/tests/request_during_resharing.rs:10-13 — The doc comment leans toward paraphrasing the code (per engineering-standards.md §Write helpful code comments). The second sentence carries the load-bearing information (why filtering by protocol alone is sufficient); the first sentence is redundant with a well-named function. Consider collapsing to a single line stating the uniqueness invariant, e.g. /// Panics if no domain matches; each protocol appears at most once in this test's domain set.
• crates/e2e-tests/tests/request_during_resharing.rs:64-67 — Minor: since the helper now returns an owned DomainConfig and only .id is used at each call site, you could return DomainId directly (or take &contract_state.domains.domains as input) to avoid the .clone(). Not worth blocking on in a test, but tidier.

✅ Approved

[claude]

PR title type suggestion: This PR changes only test files, so the type prefix should probably be test: instead of refactor:.

Suggested title: test: reduce code redundancy in request_during_resharing test

[barakeinav1]

nit (non-blocking): .find() returns the first match and this only panics on zero matches — not on multiple. The doc says "each protocol appears at most once in this test's set," but nothing enforces it, and production actually has two CaitSith domains (Sign + ForeignTx). If default_for_test ever grows a second same-protocol domain, this would silently pick whichever comes first (the old purpose == Sign filter guarded against that). Cheap to make it fail loudly instead:

let mut matches = contract_state.domains.domains.iter().filter(|d| d.protocol == protocol_type);
let domain = matches.next().unwrap_or_else(|| panic!("no domain with protocol {protocol_type:?}"));
assert!(matches.next().is_none(), "multiple domains with protocol {protocol_type:?}");
domain.id

[barakeinav1]

Thanks, approved — pure test refactor, no logic change. One non-blocking nit on the helper's uniqueness assumption.