Mockdaten-Generator-Experten-Prompt

# Mock Data Generator - TODO ## Context - **Target Database Schema**: Flexible multi-domain schema covering e-commerce, finance, healthcare, and social media - **Data Volume**: Development (small) - 100-1000 records; Testing (medium) - 10,000+ records - **Use Case**: Test fixtures, API mocks, and database seeds - **Output Format**: JSON, SQL, CSV, TypeScript ## Generation Plan ### Entity & Schema Planning - [ ] **MOCK-PLAN-1.1 [Users Entity]**: - **Schema**: id (UUID), firstName, lastName, email (unique), phone, createdAt, updatedAt, status (active/suspended/closed) - **Volume**: 1000 records - **Format**: JSON, and SQL - **Edge Cases**: Min/max name lengths, invalid emails, null optional fields, duplicate prevention - [ ] **MOCK-PLAN-1.2 [Products Entity]**: - **Schema**: id (SKU format PROD-XXXX), name, description, price (2dp), stockQuantity, categoryId (FK), createdAt, updatedAt - **Volume**: 500 records - **Format**: JSON, and SQL - **Edge Cases**: Zero price, negative stock test, max price dec, unicode in descriptions - [ ] **MOCK-PLAN-1.3 [Orders Entity]**: - **Schema**: id (ORDER-XXXXX), userId (FK), orderDate, status (pending, paid, shipped, delivered, cancelled), total (computed) - **Volume**: 2000 records - **Format**: JSON, and SQL - **Edge Cases**: Order before user birthday, null delivery date for pending, totals with extreme values - [ ] **MOCK-PLAN-1.4 [OrderItems Relation]**: - **Schema**: orderId (FK), productId (FK), quantity, priceSnapshot - **Volume**: 5000 records (2-5 items per order) - **Format**: JSON, and SQL - **Edge Cases**: Quantity nullable (checkout removed), product discontinued (zero stock) - [ ] **MOCK-PLAN-1.5 [Reviews Entity]**: - **Schema**: id, productId (FK), userId (FK), rating (1-5), comment, createdAt - **Volume**: 3000 records - **Format**: JSON, and SQL - **Edge Cases**: Rating 0 and 6 for schema-valid, empty comments, emoji and unicode ### Deep-Dive Data Modeling - [ ] **MOCK-PLAN-1.6 [Addresses Entity]**: - **Schema**: id, userId (FK), country, state, city, postalCode, street, isPrimary - **Volume**: 2000 records (multiple user addresses) - **Format**: TypeScript objects - **Edge Cases**: Extremely long street names, invalid postal codes (special), null country (shipping) - [ ] **MOCK-PLAN-1.7 [Product Categories Tree]**: - **Schema**: id, parentId (nullable), level, name, slug, description - **Volume**: 50 records (4 levels deep) - **Format**: CSV and JSON - **Edge Cases**: Root (parent null), circular references prevention, duplicated slugs - [ ] **MOCK-PLAN-1.8 [API Response Models]**: - **Schema**: Standard REST envelope with data, meta, and errors - **Volume**: 5 examples per endpoint (success, error, validation, not found, pagination) - **Format**: JSON - **Edge Cases**: 204 No Content, 500 error (empty errors array), gate delayed (429), network blips ## Generation Items ### Core Data Generation Tasks - [ ] **MOCK-ITEM-1.1 [User Dataset Builder]**: - **Entity**: Users - **Generator**: Faker.js `faker.internet.email`, `faker.name.firstName`, `faker.name.lastName`, custom `randomStatus()` - **Relationships**: None (standalone), to be referenced by Addresses, Orders, Reviews - **Validation**: Email uniqueness enforced by `faker.helpers.unique` - [ ] **MOCK-ITEM-1.2 [Product Catalog Generator]**: - **Entity**: Products - **Generator**: `faker.commerce.productName`, `faker.commerce.price`, custom category assignment from `PRODUCT_CATEGORIES` - **Relationships**: Categories (many-to-one) - **Validation**: SKUs are unique and follow strict regex - [ ] **MOCK-ITEM-1.3 [Order and Line Items Factory]**: - **Entity**: Orders, Order Items - **Generator**: `faker.date.between` for timestamps wisely, `faker.helpers.arrayElement` for status - **Relationships**: Users (FK), Orders (FK) -> orderItems (FK productId) - **Validation**: Generate totals consistently (sum. quantity * priceSnapshot = total) ### Edge Case Implementation Tasks - [ ] **MOCK-ITEM-1.4 [Boundary Value Seeds]**: - **Entity**: All entities - **Generator**: Custom `boundaryCase()` - **Relationships**: N/A - **Validation**: Ensure 100% edge-case coverage by explicit records (e.g., null createdAt, max-length book description) - [ ] **MOCK-ITEM-1.5 [Large Dataset with Index Hints]**: - **Entity**: Orders, Products, Users (load test scenario) - **Generator**: `streamArray()` with seeded RNG - **Relationships**: FK cascade rules (ON DELETE = CASCADE) - **Validation**: Using `EXPLAIN ANALYZE` before/after - [ ] **MOCK-ITEM-1.6 [Audit Trail / Log Seeding]**: - **Entity**: AuditLog - **Generator**: Faker's `faker.internet.ipv4`, `faker.date.between`, JSON object for metadata - **Relationships**: userId (optional, can be null for anonymous actions) - **Validation**: Ensure sequence isn't broken if local timezone changes ## Proposed Code Changes ### Patch-Style Diffs For Generators ``` +---- scripts/ + |-- generate-users.js + |-- generate-products.js + |-- generate-orders.js + +-- generate-reviews.js ``` 1. `[file: scripts/generate-users.js]`: ```diff +import { faker } from '@faker-js/faker'; +faker.seed(12345); // deterministic +const users = Array.from({ length: 1200 }, () => ({ + id: faker.string.uuid(), + firstName: faker.person.firstName(), + lastName: faker.person.lastName(), + email: faker.helpers.unique(faker.internet.email, [faker.person.firstName(), faker.person.lastName()]), + phone: faker.phone.number({ style: 'national' }), + status: faker.helpers.arrayElement(['active', 'suspended', 'archived'], 1), + createdAt: faker.date.past({ years: 5 }),, + updatedAt: faker.date.recent({ days: 90 }), +})); export default users; ``` ### Command-Line Interface - **Local**: ```bash npm run seed:generate # outputs JSON + SQL npm run seed:clean # reverses ``` - **CI**: ```bash npm run test:integration # After seeding ``` ### SQL Seed Template (PostgreSQL-Compliant) ```sql INSERT INTO "public"."Users" ("id", "firstName", "lastName", "email", "createdAt", "updatedAt") VALUES (:id, :firstName, :lastName, :email, :createdAt, :updatedAt); ``` ## Validation Task List (Must-Have in Code) - [ ] `validateUser.js` - Regex for email, sanity check on date ranges - [ ] Return `boolean`/`throw new Error` on failure - [ ] `validateProduct.js` - Price ≥ 0, category exists, unique SKU - [ ] `validateOrders.js` - Sum check using productSeed array, ensure “userId” FK spells correctly - [ ] `validateReferences()` - Ensures all FK (orders.user -> user.id) exist in pool before mutagenesis ## Quality Assurance Task Checklist Before finalizing the generator: - [ ] **Q-1**: Strict schema check on all entities - [ ] **Q-2**: FK satisfied, no orphans after cleanup - [ ] **Q-3**: Deterministic → `seed(12345)` same data on every run - [ ] **Q-4**: Edge cases are deliberately sprinkled (e.g., 0 reviews, post-Feb 29to) - [ ] **Q-5**: Output is valid per target DB (`psql \"\\copy\"` or Stored Procedure) - [ ] **Q-6**: Cleanup scripts in `teardown.sql` are provided and run w/o errors - [ ] **Q-7**: Time to generate < 5 sec for 10K (measure various volumes) - [ ] **Q-8**: Unique email/SKU > 99.99% success rate ## Commands to Execute ### Local Data Generation ```bash # Will generate `/data` JSON and `/sql` inserts npm run db:seed -- --environment dev ``` ### Verify During CI ```bash npm install # ensures Faker version pin npx eslint generated-data/*.ts include "no-null-eslint" - for js lint npm ci npm run seed:all --no-daemon ``` ### Cleanup Script Path ``` cleanup/ destroy-orders.sql destroy-reviews.sql destroy-users.sql (~/scripts/cleanup-all.sh) ``` ## Red Flag Checks - [ ] No hardcoded “John Doe” names in any dataset - [ ] Every table has at least 3 formatting types (2 English, 1 Unicode) - [ ] Every column got <1% Null in an optional case (validate percentage algorithm) - [ ] `generated - faker@9.5.0` is in dependencies list - [ ] Cross-entity consistency: order timestamp ≥ createdAt, ≥ updatedAt - [ ] Run `psql -f seed.sql` and confirm 0 output scripts - [ ] Verify no ad-hoc temp tables left (clean) --- This is the completed **mock-data-plan as TODO** output aimed at LLM/Dev. Use checkboxes and task IDs directly to track implementation in code review via IDE opened report enhance.