Test Cases

Test Strategy

Test type	Target	Tool	Current state
Unit tests (backend)	Pure-function business logic	Vitest	Implemented (2 files)
Integration tests (backend)	API and DB	Vitest	Not implemented (only the `test:integration` script exists)
Unit / E2E (frontend)	Screens	-	Not set up (no test runner)

Coverage is currently limited

Automated tests cover only the backend's services/scoring.ts and services/candidate.ts (parseResponseFilter). Everything else relies on Swagger UI (/docs) and manual verification in the UI. No coverage target has been set.

Running Tests

cd heineken-interview-arrange-backend

bun run test                # vitest run (everything)
bun run test:unit           # __tests__/unit only
bun run test:integration    # __tests__/integration only (currently empty)
bun run test:watch          # Watch mode
bun run test:coverage       # Coverage (v8)

bunx vitest run __tests__/unit/services/scoring.test.ts   # A single file
bunx vitest run -t "should cap score at maxScore"         # A single test

The frontend has no test runner. Its quality gates are:

cd heineken-interview-arrange-frontend
bun run check-types
bun run lint

Test Configuration

vitest.config.ts:

Item	Value
`environment`	`node`
`globals`	`true` (no import needed for `describe` / `it` / `expect`)
`include`	`__tests__/*/.test.ts`
`testTimeout` / `hookTimeout`	10000 ms
Coverage provider	`v8` (`text` / `json` / `html` reporters)

Automated Test Cases

Scoring (`tests/unit/services/scoring.test.ts`)

calculateScore:

ID	Scenario	Expected result	Priority
SC-001	No rules	Returns 0	High
SC-002	`equals` condition matches	Adds the rule's score	High
SC-003	`not_equals` condition	Adds the score when it does not match	High
SC-004	`in` condition	Adds the score when the value is in the list	High
SC-005	`between` condition within range	Adds the score	High
SC-006	`between` condition out of range	Contributes 0	High
SC-007	`contains` condition (array value)	Adds the score when the element is present	Medium
SC-008	`greater_than` condition	Adds the score above the threshold	Medium
SC-009	`less_than` condition	Adds the score below the threshold	Medium
SC-010	Multiple rules	Sums each rule's score	High
SC-011	Total exceeds `maxScore`	Capped at `maxScore`	High

validateScoringRules:

ID	Scenario	Expected result	Priority
SC-012	Well-formed rules	Returns `true`	High
SC-013	Malformed rules	Returns `false`	High
SC-014	Unknown `condition`	Returns `false`	High

Response Filter Parsing (`tests/unit/services/candidate.test.ts`)

parseResponseFilter (the responseFilter query on the candidate list):

ID	Scenario	Expected result	Priority
CF-001	Empty string	Returns `undefined`	High
CF-002	`undefined`	Returns `undefined`	High
CF-003	String that is not valid JSON	Throws `BadRequestError`	High
CF-004	JSON that does not match the schema	Throws `BadRequestError`	High
CF-005	Array form with one condition	Interpreted as AND	High
CF-006	Array form with several conditions	All combined with AND	High
CF-007	Object form with `logic: "and"`	Interpreted as AND	High
CF-008	Object form with `logic: "or"`	Interpreted as OR	High
CF-009	`logic` omitted	Defaults to AND	High
CF-010	`eq` operator	Accepted	Medium
CF-011	`neq` operator	Accepted	Medium
CF-012	`contains` operator	Accepted	Medium
CF-013	`in` operator with an array value	Accepted	Medium
CF-014	Unknown operator	Rejected	High
CF-015	String value	Accepted	Medium
CF-016	Numeric value	Accepted	Medium
CF-017	Array of strings	Accepted	Medium
CF-018	Array of numbers	Accepted	Medium
CF-019	Mixed array of strings and numbers	Accepted	Low
CF-020	Empty conditions array	Handled without error	Medium
CF-021	Field name containing dots	Treated as a nested path	Medium
CF-022	Japanese field names	Handled correctly	High
CF-023	Special characters in values	Handled correctly	Medium

Areas Requiring Manual Verification

With no automated coverage, the following must be checked through Swagger UI (http://localhost:8080/docs) or the UI itself.

Authentication

ID	Scenario	Expected result	Priority
TC-101	Log in with a Google account	Lands on the top page with `accessToken` in `localStorage`	High
TC-102	Log in with an unregistered account	A `users` row is created with `role` = `member`	High
TC-103	Call a protected API without a token	`401 UNAUTHORIZED`	High
TC-104	Call an API with an expired token	`401` (`Token has expired`)	High
TC-105	Call a write API as `viewer`	`403 FORBIDDEN`	High

Arrange Flow

ID	Scenario	Expected result	Priority
TC-201	Calculate slots in STEP1	Slots come back from the interviewers' Google Calendars	High
TC-202	Place tentative holds in STEP1	Events are created and stored in `arrange_settings.step1.tentativeEventIds`	High
TC-203	Recreate tentative holds	Existing holds are deleted first, then recreated	Medium
TC-204	Define and save segments in STEP2	Persisted to `arrange_settings.step2.segments`	High
TC-205	Mark primary / reserve in STEP2	`project_candidates.selection_type` is updated	High
TC-206	Register additional questions in STEP3	Persisted to `arrange_settings.step3.questions`	Medium
TC-207	Place holds with `PROTOTYPE_MODE=true`	Nothing is written to the calendar	High

Candidate Import

ID	Scenario	Expected result	Priority
TC-301	Upload a UTF-16LE TSV	Columns, sample values and a suggested mapping are returned	High
TC-302	Upload a UTF-8 TSV	`400` from a parse error	Medium
TC-303	Execute an import with a mapping	Returns `201` and progress becomes `processing`	High
TC-304	Import data containing an existing `external_user_id`	Counted as a duplicate, not registered twice	High
TC-305	Run a survey import preview	Returns the filtered count and samples	High
TC-306	Run a survey import	An `import_logs` row is created and counts are filled in on completion	High

Status Management

ID	Scenario	Expected result	Priority
TC-401	Change `not_contacted` → `contacted`	Succeeds and appends one row to `status_histories`	High
TC-402	Change away from `completed`	`400 INVALID_STATUS_TRANSITION`	High
TC-403	Move `bounced` → `contacted`	Succeeds	Medium
TC-404	Move `cancelled` → `not_contacted`	Succeeds	Medium

Scheduling

ID	Scenario	Expected result	Priority
TC-501	Open the public page with a valid token	Shows the request details and the candidate's name	High
TC-502	Open with an expired token	Returns `is_expired: true` or an error	High
TC-503	Submit one preferred slot	Returns `selectedCount: 1` and the status becomes `scheduling`	High
TC-504	Submit 21 preferred slots	`400` (validation error)	Medium
TC-505	Submit preferred slots again	Existing slots are fully replaced	High
TC-506	Submit for a candidate already confirmed	`409 SLOT_NOT_AVAILABLE`	High
TC-507	Confirm the schedule as an admin	A `reservations` row is created and the tentative event becomes confirmed	High
TC-508	Confirm an already-confirmed candidate again	`409 SLOT_NOT_AVAILABLE`	High

Email

ID	Scenario	Expected result	Priority
TC-601	Preview a template	Returns the rendered result with placeholders substituted	High
TC-602	Bulk send to several candidates	Returns `sent_count` / `failed_count` and records `email_logs`	High
TC-603	Send with an empty `candidateIds`	`400` (validation error)	Medium
TC-604	Send with `PROTOTYPE_MODE=true`	No email is actually sent	High
TC-605	Delete a template	Past delivery logs remain	Medium

Databricks Integration

ID	Scenario	Expected result	Priority
TC-701	Run the metadata refresh with `dryRun`	Fetches metadata for 611 surveys and writes nothing	High
TC-702	Run the metadata refresh	`databricks_surveys` is upserted and existing rows survive	High
TC-703	List surveys with Databricks unconfigured	`503 DATABRICKS_NOT_CONFIGURED`	Medium
TC-704	Fetch a survey absent from `databricks_surveys`	`404`, never `502`	High
TC-705	Fetch a survey with a non-numeric `surveyId`	`400 INVALID_SURVEY_ID`; it never reaches a table name	High
TC-706	Import a panel with several answers in one field	Kept as a list rather than truncated to the first value	High
TC-707	Filter responses on a question containing newlines	`200`, never `500`	High
TC-708	Sign in from a disallowed email domain	Redirected back to login as `domain_not_allowed`	High

Adding a Test

Place it at __tests__/unit/services/<service-name>.test.ts
Prioritize pure functions with no external dependencies (DB, Google API, Databricks)
Structure describe in three levels — service name → function name → aspect — matching the existing tests
Running a single file with bunx vitest run <path> is the fastest loop