ledgerhq DDD Types, State, And Mocks
Apply this pattern when one API resource maps to one entity slice. Use ddd-data-layer-advanced when one response hydrates several entities.
Source: Types, Initial State & Mocks
Place Files
domain/entity/<entity>/src/
├── schema.ts
├── schema.mock.ts
├── schema.test.ts
├── selectors.ts
├── selectors.test.ts
├── slice.ts
└── slice.test.ts
domain/api/<entity>/src/
├── api.ts
└── api.test.ts
Export only the contracts consumers need from each package index.ts.
Keep Ownership Clear
| File | Responsibility |
|---|---|
schema.ts |
Define the runtime schema, infer the TypeScript type, and declare a complete initial state |
schema.mock.ts |
Build valid entity data with sensible defaults and optional overrides |
slice.ts |
Own focused entity reducers and actions |
selectors.ts |
Read or derive entity state without importing an app store type |
api.ts |
Own transport contracts, network calls, and transformations |
Follow the package's runtime-schema precedent (Zod or Typia). Do not maintain a separate handwritten type that can drift from the schema.
Validate The Contract
- Give every entity property an explicit initial value.
- Make every mock pass the runtime schema.
- Keep mock-only helpers out of production behavior.
- Give reducers one clear state transition each.
- Parse untrusted API data before storing it when the API layer does not already guarantee the contract.
- Register reducers and API middleware in each app composition root; never import an app store from a domain package.
Test
- Schema: accept a mock, reject an invalid payload, and cover initial defaults.
- Mock: return valid data and honor supported overrides.
- Slice: cover initial state, replacement/reset, and focused updates.
- Selectors: cover direct and derived values with a minimal local state shape.
- API: cover success, relevant failures, and transformations with a local test store or existing domain test harness.