# Baseline Metrics — `02-baseline-cleanup`

**Date**: 2026-05-16
**Mode**: automatic
**Scope**: missions service production code (post-rename `Azaion.Missions.*`, net10.0)
**Inputs**: `architecture_compliance_baseline.md` (2026-05-14 PASS_WITH_WARNINGS) + `implementation_report_tests.md` (Step 6 outcomes) + Step 7 test results (`test-results/report.csv`)

## Goals

Address the residual Maintainability findings the architecture-baseline scan surfaced, now that the missions/vehicles rename and the test cycle have landed.

| Source | Original finding | Status today |
|--------|------------------|--------------|
| F1 (High Architecture) | `Database/Entities/Aircraft.cs` imports feature-component enums | **Resolved 2026-05-14** by doc retag — enums re-owned by `04_persistence` |
| F2 (High Architecture) | `Database/Entities/Waypoint.cs` imports feature-component enums | **Resolved 2026-05-14** by same doc retag |
| F3 (Low Maintainability) | Dead `using Azaion.Flights.Enums;` in `Database/Entities/Flight.cs` | **Resolved by rename** — `Mission.cs` has no such using; verified |
| F4 (Low Maintainability) | Three empty scaffolding dirs at repo root | **Partial**: `Infrastructure/` is now populated (2 files); `Entities/` and `DTOs/Requests/` remain empty |

**Net actionable scope for this run**: 2 empty directories (`Entities/`, `DTOs/Requests/`).

## Coverage

| Suite | Tests | Pass | Fail | Skip | Source |
|-------|-------|------|------|------|--------|
| E2E (functional + NFT) | 78 | 48 | 0 | 30 | `test-results/report.csv` (2026-05-15 14:03 UTC) |
| Unit | 0 | – | – | – | No unit-test project today (`scripts/run-tests.sh --unit-only` is a no-op) |

All 30 skips are environment-mismatch (`COMPOSE_RESTART_ENABLED!=1` and/or `MissionsContainerHelper.Enabled=false` — the e2e-consumer image deliberately lacks docker-CLI primitives). Each carries an explicit `Skip` reason. AC trace coverage (per implementation report): 56/56 ACs traced.

Line coverage / branch coverage: not measured. The project does not configure `coverlet` or any other coverage collector. **N/A — out of scope for this run.**

## Complexity

| Metric | Value |
|--------|-------|
| Production `.cs` files (excl. `bin/`, `obj/`, `tests/`, `_docs/`) | 37 |
| Production LOC (incl. blank lines & comments) | 1,306 |
| Avg LOC per production file | 35.3 |
| Largest 5 files (LOC) | `Services/VehicleService.cs` 134 · `Program.cs` 120 · `Database/DatabaseMigrator.cs` 119 · `Auth/JwtExtensions.cs` 112 · `Services/MissionService.cs` 107 |
| Test LOC (excl. `bin/`, `obj/`) | 6,511 |

Cyclomatic complexity: not measured. No Roslyn analyzer (`dotnet format analyzers`, `Roslynator`, `SonarAnalyzer.CSharp`) is configured. **N/A — measurement infrastructure absent; out of scope.**

Note on size: 1,306 LOC across 37 files (avg 35 LOC/file, max 134) is well within the simplicity envelope this codebase aims for. There are no hot files calling out for decomposition.

## Code Smells

From `architecture_compliance_baseline.md` only (no static analyzer configured):

| Severity | Count | Open today |
|----------|-------|------------|
| Critical | 0 | 0 |
| High (Architecture) | 2 (F1, F2) | 0 — resolved 2026-05-14 |
| Low (Maintainability) | 2 (F3, F4) | 1 partial (F4: 2 of 3 empty dirs remain); F3 resolved by rename |

## Performance

Per `test-results/report.csv` 2026-05-15 14:03, the 4 NFT-PERF tests (`PerformanceTests.NFT_PERF_01..04`) all passed against thresholds defined in `_docs/02_document/tests/performance-tests.md`. Per-scenario p50/p95/p99 captured by the test harness.

This refactor run does not target performance — **N/A as a baseline-vs-final gate.**

## Dependencies

`Azaion.Missions.csproj` (Sdk.Web, net10.0):

| Package | Version |
|---------|---------|
| linq2db | 6.2.0 |
| Microsoft.AspNetCore.Authentication.JwtBearer | 10.0.5 |
| Npgsql | 10.0.2 |
| Swashbuckle.AspNetCore | 10.1.5 |

Outdated / vulnerable: not measured (would require `dotnet list package --outdated --vulnerable` against a configured NuGet source). Out of scope for this run.

## Build

| Metric | Value | Source |
|--------|-------|--------|
| Test suite wall-clock (last successful run) | ~ minutes (Docker compose up + 78 tests) | `test-results/results.trx` mtime 2026-05-15 14:03 |
| Docker build (cold, prior failed run) | ~42 min ended in CS0246 | terminal log `451778.txt` |
| Docker build (after csproj `Compile Remove="tests/**"` fix) | known-good per prior session | implicit from the green `report.csv` |

## Functionality Inventory

Components and ownership (from `_docs/02_document/module-layout.md` § Per-Component Mapping, post-rename):

| # | Component | Owns | Routes | Tests |
|---|-----------|------|--------|-------|
| 01 | vehicle_catalog | `DTOs/*Vehicle*`, `Database/Entities/Vehicle.cs`, `Enums/{VehicleType,FuelType}.cs`, `Controllers/VehiclesController.cs`, `Services/VehicleService.cs` | `/vehicles`, `/vehicles/{id}/default` | FT-P-01..06, FT-N-01..04, NFT-RES-08 |
| 02 | mission_planning | `DTOs/*Mission*`, `Database/Entities/Mission.cs`, `Controllers/MissionsController.cs`, `Services/MissionService.cs` | `/missions` | FT-P-07..12, FT-N-05..06, NFT-RES-01 |
| 04 | persistence | `Database/{AppDataConnection,DatabaseMigrator}.cs`, all `Database/Entities/*.cs` (excl. domain), `Enums/{ObjectStatus,WaypointSource,WaypointObjective}.cs` | – | NFT-RES-03..04 |
| 05 | authentication | `Auth/JwtExtensions.cs`, JWT-bearer config in `Program.cs` | – | NFT-SEC-* (14 ACs) |
| 06 | infrastructure | `Infrastructure/{ConfigurationResolver,CorsConfigurationValidator}.cs`, `Middleware/ErrorHandlingMiddleware.cs`, `Program.cs` composition | `/health`, `/swagger` | FT-P-13..18, NFT-SEC-13 (CORS), NFT-RES-05..07 |

Empty scaffolding directories (no component owns them): `Entities/`, `DTOs/Requests/`.

## Self-verification

- [x] RUN_DIR created with correct auto-incremented prefix (`02-baseline-cleanup`)
- [x] Coverage measured (E2E only; unit + line coverage marked N/A with reason)
- [x] Complexity measured (file count, LOC, top-5; cyclomatic marked N/A with reason)
- [x] Code smells measured (from baseline scan; static analyzer N/A)
- [x] Performance noted (perf tests green; not a baseline-vs-final gate)
- [x] Dependencies enumerated (outdated/vulnerable scan N/A)
- [x] Build noted (test wall-clock + Docker build status)
- [x] Functionality inventory complete (6 components + 2 empty dirs)
- [x] Measurements reproducible (commands inline in this file or sourced from named artifacts)

## Scope warning for the user (BLOCKING)

The actionable surface is **two empty directories**. Everything else the original baseline scan flagged is already resolved (F1/F2 by doc retag, F3 by rename, F4 partial). Running the full 8-phase refactor for this is heavyweight; quick-assessment (phases 0–2 only) is plausible. See the BLOCKING choice block presented to the user.