test(events): cancel + refund batch — process_cancel pagination & pro-rata refunds

[0xdevcollins]

Part of #25 (contract test-coverage push).

events

How to test

• Reuse the crate's tests/common.rs harness; add tests/cancel_refund.rs + a one-line mod cancel_refund; in tests/mod.rs.
• Template: contracts/events/src/tests/contributions.rs.
• Run cargo test -p boundless-events, then cargo test --all before the PR.

Acceptance criteria

• New tests/cancel_refund.rs covering every function in scope.
• Happy path + each Error variant + edge cases + auth-rejection (+ idempotency where relevant).
• cargo test --all passes; PR links this issue.

[YfengJ]

Hi maintainers, I would like to apply for this issue if it is available.

If assigned, I will keep the work test-focused: add contracts/events/src/tests/cancel_refund.rs, wire it through tests/mod.rs, reuse the existing tests/common.rs harness and contributions.rs style, and cover the cancel/refund happy paths, auth rejection, pagination/pro-rata edge cases, relevant error variants, and idempotency behavior where applicable.

I will wait for maintainer or GrantFox assignment before starting implementation or opening a PR.

[GBOYEE]

this is test-only — no contract changes, just coverage for the cancel + refund flow in boundless-events.

plan:

• create tests/cancel_refund.rs using the existing common.rs harness
• follow the contributions.rs template for setup pattern
• cover: happy path (full cancel → pro-rata refund to all contributors), partial cancel, double-cancel idempotency, auth rejection (non-admin tries to cancel), each Error variant (milestone_not_funded, already_canceled, etc.), edge case with zero contributors
• one-line mod cancel_refund; in tests/mod.rs
• run cargo test -p boundless-events then cargo test --all before PR

question — the pro-rata refund math: is it proportional to contribution amount, or equal split? that affects how i write the assertion for the refund amounts. also, is there a specific test address pattern i should use for the admin vs contributor roles, or does common.rs export helpers for that?