test(program-builder): session-detail-utils coverage

Context/QuickPlans/2026-04-14-tests-from-backlog.md

@@ -0,0 +1,79 @@
+# Tests from the Backlog
+
+**Task:** Write the tests that are tracked in the backlog (explicit items + coverage gaps surfaced by jcodemunch)
+**Date:** 2026-04-14
+**Branch:** chore+tests
+**Status:** Part A complete
+
+---
+
+## Goal
+
+Close the open test debt on the `chore+tests` branch. Two categories:
+
+1. **Explicit backlog test item** -- E2E display broadcast round-trip (`e2e-display-broadcast-roundtrip.md`)
+2. **Unit test coverage gaps** -- functions with confidence-1.0 `unreached` classification from jcodemunch scan
+
+The E2E test is a multi-hour build-out (requires local Supabase, dual browser contexts, auth fixtures). The unit gaps are tractable in a single session.
+
+---
+
+## Scope
+
+### Part A -- Unit Test Coverage Gaps (this session)
+
+These are all `unreached` at confidence 1.0 -- no test file imports or calls them:
+
+| File                                                     | Functions                                                                                   | Notes                                                                |
+| -------------------------------------------------------- | ------------------------------------------------------------------------------------------- | -------------------------------------------------------------------- |
+| `src/components/program-builder/session-detail-utils.ts` | `formatSetsReps`, `formatLoad`, `formatLoadSpec`, `formatSeconds`, `buildGroupedActivities` | Pure formatting/transformation functions -- high-value, easy to test |
+| `src/domain/types/set-scheme.ts`                         | `parseSetScheme`                                                                            | Domain parser -- complement existing `set-scheme.test.ts`            |
+| `src/components/program-builder/builder-state.ts`        | `buildSessionsPayload`, `buildCreateBlocksPayload`, `buildUpdateBlocksPayload`              | Payload builders used at save time                                   |
+
+Lower-priority gaps (chat/connections/analytics hooks are all unreached but complex to mock; defer unless time allows):
+
+- `src/hooks/use-chat.ts` (11 hooks)
+- `src/hooks/use-connections.ts` (7 hooks)
+- `src/hooks/use-activity-feed.ts`, `use-analytics.ts`, `use-event-items.ts`
+
+### Part B -- E2E Display Broadcast Round-Trip (future session)
+
+Fully spec'd in `Context/Backlog/e2e-display-broadcast-roundtrip.md`. Requires:
+
+- `supabase start` in Playwright webServer config
+- Seeded test user + gym via SQL fixture
+- `e2e/fixtures/` auth storage state
+- `e2e/display-roundtrip.spec.ts` with two browser contexts
+- Separate CI job (slow local Supabase startup)
+
+**Not in scope for this session.**
+
+---
+
+## Approach
+
+### Part A Execution Order
+
+1. **`session-detail-utils.test.ts`** -- create `src/components/program-builder/__tests__/session-detail-utils.test.ts`. All five functions are pure (no adapter, no hooks) -- use representative inputs from the existing `set-scheme` and `session` domain types.
+
+2. **`parseSetScheme` in `set-scheme.test.ts`** -- append cases to existing `src/domain/types/__tests__/set-scheme.test.ts`. Check what `parseSetScheme` does vs. the existing schema tests to avoid duplication.
+
+3. **Builder payload tests** -- append to or create alongside `src/components/program-builder/__tests__/builder-state.test.ts`. The existing test file covers `weeksMatch`; add payload-builder cases. Use factories from `src/test/factories.ts`.
+
+Run `bun run test` after each file to confirm green.
+
+---
+
+## Verification
+
+- `bun run test` passes with 0 failures
+- jcodemunch `get_untested_symbols` re-run shows the targeted functions are no longer `unreached`
+- No new TypeScript errors (`bun run build`)
+
+---
+
+## Risks
+
+- `buildGroupedActivities` and `useSessionTemplatesFull` in session-detail-utils.ts call hooks -- `useSessionTemplatesFull` may need `renderHook` + mock adapter; `buildGroupedActivities` may be pure. Read the file before writing tests.
+- The three builder-state payload functions may depend on complex draft state shapes -- use existing factory patterns to construct inputs rather than hand-rolling fixtures.
+- `parseSetScheme` may duplicate what `setSchemeSchema.parse` already tests -- read the implementation first and only test genuinely distinct behavior.

Context/Reviews/0016-pr110-session-detail-utils-tests-2026-04-15.md

@@ -0,0 +1,136 @@
+# PR Review: worktree-chore+tests → main
+
+**Date:** 2026-04-15
+**Feature:** N/A (test-debt chore branch)
+**Branch:** worktree-chore+tests
+**Reviewers:** pr-test-analyzer, code-reviewer
+**Status:** ✅ Resolved
+
+## Summary
+
+PR #110 adds 57 unit tests for `src/components/program-builder/session-detail-utils.ts`. The
+code-reviewer found no errors, warnings, or convention violations. The test-coverage analysis
+identified 4 gaps -- all missing test cases -- ranging from important (reachable code paths
+with no coverage) to minor (defensive/edge cases). No architectural concerns or fix-now issues.
+
+## Findings
+
+### Fix-Now
+
+_None._
+
+### Missing Tasks
+
+#### [TASK] P16-001: formatLoad emom and descendingReps without-load paths untested
+
+- **File:** `src/components/program-builder/__tests__/session-detail-utils.test.ts`
+- **Severity:** Medium
+- **Detail:** `formatLoad` handles `fixedSets | emom | descendingReps` in a shared branch with
+  an early `if (!load) return '--'` guard. The `fixedSets` no-load path is exercised indirectly
+  (via an `unspecified` load that routes through `formatLoadSpec`), but the `!load` guard itself
+  is never hit. Both `emomSchema` and `descendingRepsSchema` define `load` as `optional()`, so
+  these are legitimate, reachable states that have no test pinning the `'--'` return.
+
+  Tests to add:
+
+  ```typescript
+  it('returns -- for emom without load', () => {
+    const scheme = { type: 'emom', repsPerMinute: 10, totalMinutes: 20 } as SetScheme
+    expect(formatLoad(scheme, emptyMaxes, 'ex-1')).toBe('--')
+  })
+
+  it('returns -- for descendingReps without load', () => {
+    const scheme = { type: 'descendingReps', repLadder: [21, 15, 9] } as SetScheme
+    expect(formatLoad(scheme, emptyMaxes, 'ex-1')).toBe('--')
+  })
+  ```
+
+- **Status:** ✅ Fixed
+- **Resolution:** Added two tests to `formatLoad` describe block covering the `!load` guard for `emom` and `descendingReps`.
+
+#### [TASK] P16-002: buildGroupedActivities orphan activity behavior undocumented
+
+- **File:** `src/components/program-builder/__tests__/session-detail-utils.test.ts`
+- **Severity:** Medium
+- **Detail:** `buildGroupedActivities` iterates sorted groups and filters activities by group ID.
+  Any activity whose `activityGroupId` references a group not in the `groups` array is silently
+  dropped. This is the current behavior but no test pins it as the intended contract. Without a
+  test, a future refactor could inadvertently change this behavior (e.g., surfacing orphans or
+  throwing) with no failing test to catch it. Data from the adapter could theoretically contain
+  stale group references.
+
+  Test to add:
+
+  ```typescript
+  it('silently drops activities belonging to a non-existent group', () => {
+    const tpl = makeTemplateFull(
+      [{ id: 'g1', ordinal: 1 }],
+      [
+        { activityGroupId: 'g1', ordinal: 1, exerciseId: 'ex-1', setScheme: bwScheme },
+        {
+          activityGroupId: 'nonexistent',
+          ordinal: 2,
+          exerciseId: 'ex-orphan',
+          setScheme: bwScheme,
+        },
+      ],
+    )
+    const result = buildGroupedActivities(tpl)
+    expect(result).toHaveLength(1)
+    expect(result[0].exerciseId).toBe('ex-1')
+  })
+  ```
+
+- **Status:** ✅ Fixed
+- **Resolution:** Added orphan-drop test to `buildGroupedActivities` describe block, pinning the drop-silently contract.
+
+#### [TASK] P16-003: fixedSets mixed number/range combinations not tested
+
+- **File:** `src/components/program-builder/__tests__/session-detail-utils.test.ts`
+- **Severity:** Low
+- **Detail:** `formatSetsReps` for `fixedSets` evaluates `sets` and `reps` independently via
+  separate ternaries. Tests cover both-numeric (`3x5`) and both-range (`3-5x8-12`) but not the
+  two mixed cases: numeric sets + range reps, or range sets + numeric reps. These are distinct
+  code paths and the schema allows them (`sets` and `reps` are each `z.union([z.number(), numberRangeSchema])`).
+- **Status:** ✅ Fixed
+- **Resolution:** Added two tests to `formatSetsReps` describe block: numeric sets + range reps (`3x8-12`) and range sets + numeric reps (`3-5x5`).
+
+#### [TASK] P16-004: Default fallthrough branches not tested
+
+- **File:** `src/components/program-builder/__tests__/session-detail-utils.test.ts`
+- **Severity:** Low
+- **Detail:** Both `formatSetsReps` and `formatLoad` have a final `default`/trailing `return '--'`
+  for unrecognized `scheme.type` values. These are unreachable in strict TypeScript but could be
+  hit if the type system is bypassed (cast or unvalidated API response). The `formatLoadSpec`
+  unknown-type case IS tested (line 100-102). Parity with the other two functions would be
+  complete coverage. Low risk given strict TS enforcement.
+- **Status:** ✅ Fixed
+- **Resolution:** Added `returns -- for unrecognized scheme type` test to both `formatSetsReps` and `formatLoad` describe blocks using `as unknown as SetScheme` cast.
+
+### Architectural Concerns
+
+_None._
+
+### Convention Gaps
+
+_None._
+
+---
+
+## Resolution Checklist
+
+- [x] All [FIX] findings resolved
+- [x] All [TASK] findings added to test file or dismissed
+- [x] All [ADR] findings have ADRs created or dismissed
+- [x] All [RULE] findings applied or dismissed
+- [x] Review verified by review-verify agent
+
+## Resolution Summary
+
+**Resolved at:** 2026-04-15
+**Session:** Implemented all 4 missing test cases directly in the test file (chore branch, no Steps.md)
+
+| Category  | Total | Resolved |
+| --------- | ----- | -------- |
+| [TASK]    | 4     | 4        |
+| **Total** | **4** | **4**    |