# aphoria-code-patterns

## AUDIT (2026-02-06)

### Pattern 1: Unwrap/Expect Isolation
**Finding:** NOT APPLICABLE

- **Total unwrap() calls:** 72
- **Total expect() calls:** 890 (mostly from stemedb crates, not aphoria)
- **In test code:** ALL 72 unwrap() calls are within `#[test]` functions
- **In production code:** 0

Analysis:
- `promotion/version.rs:490` - test function `test_changelog_entry_with_metrics`
- `research/gap_store.rs:365-390` - test functions `test_gap_store_*`
- `research/tests.rs` - all test code
- `types/language.rs:220-230` - test assertions

**Decision:** No fix needed. Clippy's `clippy::unwrap_used` is at `warn` level for crates, but test code is exempt by design. All 72 instances are in test functions where unwrap is acceptable for test assertions.

### Pattern 2: JSON Construction Consistency
**Finding:** 27 instances of `serde_json::json!` macro

**Categories:**

1. **Source metadata construction (5 files):**
   - `bridge.rs:52` - claim_to_assertion
   - `episteme/corpus.rs:191` - corpus building
   - `llm/extractor.rs:431` - LLM extraction
   - `llm/prompt.rs:97` - prompt building
   - `llm/ontology.rs:243` - ontology extraction

2. **Report generation (10 instances):**
   - `report/sarif.rs` - 5 instances (SARIF format requires specific structure)
   - `report/json.rs` - 5 instances (dynamic conflict reports)

3. **Other (7 instances):**
   - `policy_ops.rs:238` - ack payload (recent addition)
   - `report/mod.rs:56` - single value conversion
   - `eval/matcher.rs:328` - test fixture
   - `eval/harness.rs` - 4 test fixtures

**Analysis:**
The `json!` macro is used appropriately for:
- Dynamic JSON construction where struct serialization doesn't apply
- SARIF format which has strict schema requirements
- Test fixtures where convenience matters

This is NOT tech debt - it's appropriate usage. The audit finding was overly aggressive.