Test/accrual reconciliation prop

[Glam26]

Here's a summary of all changes made:
Changes Summary
New file: src/test_accrual_reconciliation_prop.rs
A property-based test that asserts the accrual ledger reconciliation invariant: sum(accrued_per_holder) + sum(claimed) == total_revenue at every step.
Structure:

• Custom AccrualOp enum with SetShare , Deposit , and Claim variants
• Proptest strategy generates sequences of 12–25 ops across 4 holders, weighted 3:3:2 (SetShare:Deposit:Claim)
• Initial share balancing distributes 10 000 bps evenly across holders so the invariant holds exactly
• rebalance_shares rebalances after every SetShare to keep total bps at 10 000, maintaining the invariant
• check_invariant wraps the sum(accrued) + sum(claimed) == total_deposited assertion, called after every single operation
• 512 randomized sequences (as required) with deterministic RNG
• Amounts are multiples of 10 000 so amount * share_bps / 10_000 is exact for any bps (no rounding dust)
  Two edge-case tests (standalone):
  ┌──────────────────┬───────────────────────────────────────────────────┐
  │ Test │ What it verifies │
  ├──────────────────┼───────────────────────────────────────────────────┤
  │ edge_no_share_ho │ Revenue deposited when no holder has shares stays │
  │ lder_sequence │ in the accrual index (expected behavior) │
  │ edge_single_hold │ 100% holder can fully deposit → claim → drain │
  │ er_full_drain │ │
  └──────────────────┴───────────────────────────────────────────────────┘
  Key design decisions validated by review:
• saturating_sub on i64 correctly handles edge cases (saturates at i64::MIN , not 0)
• The fast path in rebalance_shares correctly requires a single holder to absorb the full delta
• The fallback loop converges because total headroom >= delta always holds
• Invariant checked after every operation including SetShare
  Modified: src/lib.rs
  Added module declaration:
  // rust
  #[cfg(test)]
  mod test_accrual_reconciliation_prop;
  Build note
  The codebase has 100 pre-existing compilation errors (duplicate contractimpl functions, missing struct fields, custom attribute panics, etc.) — all unrelated to this change. The new test file compiles cleanly with no errors from its code.

closes #532

[drips-wave]

@Glam26 Great news! 🎉 Based on an automated assessment of this PR, the linked Wave issue(s) no longer count against your application limits.

You can now already apply to more issues while waiting for a review of this PR. Keep up the great work! 🚀

Learn more about application limits