@demokit-ai/codegen

Demo data generation and validation for DemoKit - create realistic fixtures from schemas.

Installation

npm install @demokit-ai/codegen @demokit-ai/schema

Features

• Data Generation - Generate realistic demo data from schemas
• ID Generation - UUID, CUID, ULID, and custom ID formats
• Value Generators - Smart generators for emails, names, addresses, dates, etc.
• Validation - Validate generated data against schemas
• Output Formatters - Export to TypeScript, JSON, SQL, or CSV
• Full TypeScript support

Usage

Generate Demo Data

import { generateDemoData } from '@demokit-ai/codegen'

const schema = {
  models: {
    User: {
      properties: {
        id: { type: 'string', format: 'uuid' },
        name: { type: 'string' },
        email: { type: 'string', format: 'email' },
        createdAt: { type: 'string', format: 'date-time' },
      },
    },
    Order: {
      properties: {
        id: { type: 'string', format: 'uuid' },
        userId: { type: 'string', relationshipTo: { model: 'User', field: 'id' } },
        total: { type: 'number', minimum: 0 },
        status: { type: 'string', enum: ['pending', 'completed', 'cancelled'] },
      },
    },
  },
}

const data = generateDemoData(schema, {
  recordCounts: { User: 5, Order: 10 },
  seed: 42, // For reproducible output
})

ID Generation

import {
  generateUUID,
  generateSeededUUID,
  generateCuid,
  generateUlid,
  generatePrefixedId,
} from '@demokit-ai/codegen'

// Random UUID v4
const uuid = generateUUID()
// '550e8400-e29b-41d4-a716-446655440000'

// Deterministic UUID from seed
const seededUuid = generateSeededUUID(12345)

// CUID-like ID
const cuid = generateCuid()
// 'clxxxxxxxxxxxxxxxx'

// ULID
const ulid = generateUlid()
// '01ARZ3NDEKTSV4RRFFQ69G5FAV'

// Prefixed ID
const userId = generatePrefixedId('user', 1)
// 'user_001'

Value Generation

import { generateValue } from '@demokit-ai/codegen'

// Email
const email = generateValue({ type: 'string', name: 'email' }, 0)
// '[email protected]'

// Person name
const name = generateValue({ type: 'string', name: 'name' }, 0)
// 'Emma Smith'

// Price
const price = generateValue({ type: 'number', name: 'price' }, 0)
// 19.99

// Status from enum
const status = generateValue({
  type: 'string',
  name: 'status',
  enum: ['active', 'pending', 'inactive']
}, 0)
// 'active'

Data Validation

import { validateData } from '@demokit-ai/codegen'

const result = validateData(generatedData, { schema })

if (!result.valid) {
  console.log('Validation errors:', result.errors)
  // [{ type: 'type_mismatch', model: 'User', field: 'age', ... }]
}

console.log('Stats:', result.stats)
// { totalRecords: 15, recordsByModel: { User: 5, Order: 10 }, durationMs: 12 }

Output Formatting

TypeScript

import { formatAsTypeScript } from '@demokit-ai/codegen'

const tsCode = formatAsTypeScript(data, {
  asConst: true,
  includeHeader: true,
  narrative: {
    scenario: 'E-commerce demo',
    keyPoints: ['5 users', '10 orders'],
  },
})

// Output:
// /**
//  * Auto-generated demo fixtures
//  * Scenario: E-commerce demo
//  */
// export const DEMO_USER = [...] as const
// export const DEMO_ORDER = [...] as const
// export const DEMO_DATA = { User: DEMO_USER, Order: DEMO_ORDER } as const

JSON

import { formatAsJSON } from '@demokit-ai/codegen'

const json = formatAsJSON(data, { includeMetadata: true })
// { "_metadata": { "generatedAt": "...", "recordCount": 15 }, "data": {...} }

SQL

import { formatAsSQL } from '@demokit-ai/codegen'

const sql = formatAsSQL(data, {
  tableName: (model) => `demo_${model.toLowerCase()}`,
})
// INSERT INTO demo_user (id, name, email) VALUES ('1', 'Alice', '[email protected]');

CSV

import { formatAsCSV } from '@demokit-ai/codegen'

const userCsv = formatAsCSV(data, 'User')
// id,name,email
// 1,Alice,[email protected]

API Reference

Generation

generateDemoData(schema, options)

Generate demo data from a schema.

Options:

• recordCounts - Number of records per model
• seed - Random seed for reproducibility
• baseTimestamp - Base timestamp for date generation

ID Generation

• generateUUID() - Random UUID v4
• generateSeededUUID(seed) - Deterministic UUID from seed
• generateCuid() - CUID-like ID
• generateUlid() - ULID
• generatePrefixedId(prefix, index) - Prefixed ID (e.g., 'user_001')
• generateIdForModel(model, index, format?) - Model-specific ID

Validation

validateData(data, options)

Validate data against a schema.

Options:

• schema - Schema to validate against
• failFast - Stop on first error
• maxErrors - Maximum errors to collect
• collectWarnings - Include warnings in result

Returns:

• valid - Whether data is valid
• errors - Array of validation errors
• warnings - Array of warnings
• stats - Validation statistics

Formatters

• formatAsTypeScript(data, options) - TypeScript module
• formatAsJSON(data, options) - JSON with optional metadata
• formatAsSQL(data, options) - SQL INSERT statements
• formatAsCSV(data, modelName) - CSV for a single model

License

MIT