schema-dsl

🎯 schema-dsl

Progressive TypeScript schema DSL for concise, serializable field rules — one schema for validation, reuse, export, documentation, and i18n.

📚 Documentation: https://vextjs.github.io/schema-dsl

Quick Start · Documentation · Feature Overview · Examples

npm install schema-dsl

⚡ TL;DR (30-second intro)

What is schema-dsl?

Write field rules like this:

import { s, validate } from 'schema-dsl/pure';

const userSchema = s({
  username: s('string:3-32!').label('Username'),
  email:    s('email!').label('Email'),
  role:     'admin|user|guest',
  contact:  'types:email|phone'
});

const contactEmail = s('email!').label('Email').pattern(/custom/);
const accountEmail = s.email().label('Email').pattern(/custom/).require();

const result = validate(userSchema, req.body);

Then that same set of rules continues to power:

• ✅ Sync / async validation — validate() / validateAsync()
• ✅ Schema derivation — pick / omit / partial to tailor schemas per endpoint
• ✅ Database schemas — export directly to MongoDB / MySQL / PostgreSQL
• ✅ Field documentation — auto-generate Markdown
• ✅ Unified error model — ValidationError + I18nError
• ✅ Internationalization — 5 built-in locales (zh-CN / en-US / ja-JP / es-ES / fr-FR), switchable at runtime

5-minute tutorial: Quick Start | Full docs: Online Documentation

🗺️ Documentation

Getting started:

• Quick Start — up and running in 5 minutes
• DSL Syntax Reference — syntax cheatsheet
• FAQ — common questions

Core features:

• validate() — synchronous validation API
• SchemaUtils — schema reuse
• Conditional Validation API — s.if / s.match
• Async Validation & Framework Integration — Express / Koa / Fastify
• Error Handling & i18n — error model

Export & integration:

• Export Guide — MongoDB / MySQL / PostgreSQL
• TypeScript Guide — type inference and usage
• Extensions and Integration — custom business types, keywords, runtime isolation, and plugins

Full docs: Online Documentation · Chinese Documentation · Feature Index

✨ Why schema-dsl?

🎯 Concise field rules — less boilerplate

❌ Manual JSON Schema — verbose const schema = { type: 'object', properties: { username: { type: 'string', minLength: 3, maxLength: 32 }, email: { type: 'string', format: 'email' }, age: { type: 'number', minimum: 18, maximum: 120 } }, required: ['username', 'email'] };

✅ schema-dsl — concise and clean // just 3 lines const schema = s({ username: 'string:3-32!', email: 'email!', age: 'number:18-120' });

💪 Full-featured

Feature	schema-dsl	Notes
Basic validation	✅	string, number, boolean, date, email, url, phone…
Advanced validation	✅	regex, custom functions, conditional branches, nested objects, arrays…
Cross-type union	✅	types:email|phone — one field accepts multiple types
Error messages	✅	auto-translated + custom messages + field labels
i18n business errors	✅	I18nError with numeric error codes
Database export	✅	MongoDB / MySQL / PostgreSQL schema generation
Documentation generation	✅	Markdown field docs auto-generated
TypeScript	✅	Written in native TypeScript with full type inference
Extension system	✅	One business type definition for pure DSL, s('...'), and s.xxx() entries; plus formats and validators
Schema reuse	✅	pick / omit / partial / extend
Side-effect-controlled entries	✅	root compatibility, schema-dsl/pure for no String.prototype installation, and schema-dsl/runtime for isolated runtime state
Compile-time transform	✅	schema-dsl/transform core and optional schema-dsl/esbuild adapter
Progressive s authoring	✅	Use plain DSL strings, s('email!'), or s.email(); all converge to the same builder implementation

🎨 One schema, many uses (unique capability)

import { s, exporters, SchemaUtils } from 'schema-dsl/pure';

const userSchema = s({
  id:        'uuid!',
  username:  'string:3-32!',
  email:     'email!',
  password:  'string:8-64!',
  age:       'number:18-120',
  createdAt: 'string!'
});

// 📋 derive scenario-specific schemas
const createSchema = SchemaUtils.omit(userSchema, ['id', 'createdAt']);
const updateSchema = SchemaUtils.partial(SchemaUtils.pick(userSchema, ['username', 'email']));
const publicSchema = SchemaUtils.omit(userSchema, ['password']);

// 🗄️ export the same schema to any database
const mongoSchema = new exporters.MongoDBExporter().export(userSchema);
const mysqlDDL    = new exporters.MySQLExporter().export('users', userSchema);
const pgDDL       = new exporters.PostgreSQLExporter().export('users', userSchema);

// 📝 generate field documentation from the same schema
const markdown = exporters.MarkdownExporter.export(userSchema, { title: 'User Field Reference' });

⚠️ SQL exporters only accept anyOf / oneOf when every branch resolves to the same SQL column type (for example ipv4 | ipv6). Ambiguous unions such as string | number now throw an explicit error instead of silently choosing the first branch.

📦 Installation

npm install schema-dsl

Runtime requirement: Node.js >= 18.0.0

The entry points below are part of the package contract and are covered by packed-consumer tests for both ESM and CommonJS.

📦 Package Entry Points

schema-dsl keeps the root import compatible with v1-style direct string chaining, and also exposes explicit entries for projects that want tighter control over global side effects.

Entry	Purpose
schema-dsl/pure	Recommended default entry for application code; exports the s / dsl namespace and validation helpers without installing String.prototype extensions.
schema-dsl/runtime	Runtime adapter factory for per-tenant/per-app isolated Locale messages, messageProvider, TypeRegistry scope, PATTERNS, validator instances and I18nError creation.
schema-dsl	Root compatibility entry; imports install the non-enumerable String chain API by default. Prefer schema-dsl/pure for new public examples.
schema-dsl/compat	Explicit compatibility entry that installs String extensions on import.
schema-dsl/register-string	Side-effect entry for explicitly registering String extensions during application startup.
schema-dsl/string-types	Opt-in TypeScript declarations for String-chain authoring; no runtime prototype installation.
schema-dsl/transform	Babel AST transform core that rewrites static string-chain calls into helper calls imported from schema-dsl/pure. Babel AST packages are optional peer dependencies for projects that use this entry.
schema-dsl/esbuild	Optional esbuild plugin adapter around the transform core. esbuild is an optional peer dependency.

import { s, validate } from 'schema-dsl/pure';
import { transformSchemaDsl } from 'schema-dsl/transform';
import { schemaDslEsbuildPlugin } from 'schema-dsl/esbuild';
import { createRuntime } from 'schema-dsl/runtime';

const schema = s({
  email: 'email!',
  username: s('string:3-32!').label('Username'),
  backupEmail: s.email().label('Backup email').require()
});

const transformed = transformSchemaDsl(
  'export const field = "admin|user|guest".label("Role")',
  { filename: 'schema.ts' }
);

const plugins = [schemaDslEsbuildPlugin()];

const tenantRuntime = createRuntime({
  locale: 'tenant-a',
  messages: {
    'tenant.user.missing': { code: 'TENANT_USER_MISSING', message: 'Tenant user {{#id}} is missing' }
  },
  types: {
    tenantId: { type: 'string', pattern: '^tenant_[a-z0-9]+$' }
  },
  messageProvider: ({ key, locale, fallback }) =>
    key === 'number.min' ? `[${locale}] {{#label}} must be >= {{#limit}}` : fallback
});

const tenantSchema = tenantRuntime.s({
  id: 'tenantId!',
  age: 'number:18-120'
});

const tenantEmail = tenantRuntime.s.email().label('Tenant email').require().toSchema();

const tenantResult = tenantRuntime.validate(tenantSchema, { id: 'tenant_demo', age: 16 });

The transform handles static DSL string literals, including naked pipe enums such as "admin|user|guest", and injects imports from schema-dsl/pure. By default it rewrites the complete built-in String-chain API (.label(), .pattern(), .require(), .required(), .toJsonSchema(), and the other methods installed by schema-dsl). Use additionalMethods for user-defined chain methods, and additionalTypes / additionalTypePatterns for registered custom DSL type literals such as "tenant-id!".label("Tenant"); methods remains a legacy replacement set when you intentionally want to override the built-in default list. Dynamic expressions, computed member calls, and already transformed helper calls are left unchanged.

Use schema-dsl/pure for ordinary application code. Use schema-dsl/runtime when a framework needs independent runtime state per app, tenant, worker, or plugin host. createRuntime() keeps message lookup, per-call messageProvider, runtime custom types, namespace factories, pattern overrides, validator caches, custom keyword messages, conditional branches, async custom validators, and createI18nError() inside that runtime instance. Use one runtime for the app/plugin lifecycle, pass request-level locale, messages, messageProvider or { coerce: false } via per-call options, and call configure(..., { mode: 'replace' | 'reset' }), clearCache(), getStats() or dispose() for hot reload and shutdown.

createSchemaDslRuntime() and createSchemaDslAdapter() are equivalent aliases of createRuntime() for adapter-oriented integrations.

🚀 Quick Start

1. Basic validation

import { s, validate } from 'schema-dsl/pure';

const userSchema = s({
  username: 'string:3-32!',
  email:    'email!',
  age:      'number:18-120',
  role:     'admin|user|guest',
  tags:     'array<string>'
});

// ✅ validation passed
const result = validate(userSchema, {
  username: 'john_doe',
  email:    '[e-mail]',
  age:      25,
  role:     'user',
  tags:     ['verified']
});

console.log(result.valid);   // true
console.log(result.data);    // validated data

// ❌ validation failed
const bad = validate(userSchema, { username: 'ab', email: 'not-email' });
console.log(bad.errors);
// [
//   { path: 'username', message: 'username must be at least 3 characters' },
//   { path: 'email',    message: 'email must be a valid email address' }
// ]

2. Async validation + Express integration

import { s, validateAsync, ValidationError } from 'schema-dsl/pure';

const createUserSchema = s({
  username: 'string:3-32!',
  email:    'email!',
  password: 'string:8-32!'
});

app.post('/api/users', async (req, res, next) => {
  try {
    // throws ValidationError automatically on failure
    const validData = await validateAsync(createUserSchema, req.body);
    const user = await db.users.create(validData);
    res.json({ success: true, data: user });
  } catch (error) {
    next(error);
  }
});

// global error handler
app.use((error, req, res, next) => {
  if (error instanceof ValidationError) {
    return res.status(400).json({ success: false, errors: error.errors });
  }
  next(error);
});

3. Schema reuse (create / update / public)

import { s, SchemaUtils } from 'schema-dsl/pure';

const userSchema = s({
  id:        'uuid!',
  username:  'string:3-32!',
  email:     'email!',
  password:  'string:8-64!',
  createdAt: 'string!'
});

// create endpoint: remove server-generated fields
const createSchema = SchemaUtils.omit(userSchema, ['id', 'createdAt']);

// update endpoint: pick editable fields, all optional
const updateSchema = SchemaUtils.partial(
  SchemaUtils.pick(userSchema, ['username', 'email'])
);

// public response: hide sensitive fields
const publicSchema = SchemaUtils.omit(userSchema, ['password']);

4. Database schema export

import { s, exporters } from 'schema-dsl/pure';

const productSchema = s({
  name:      'string:1-100!',
  price:     'number:>0!',
  stock:     'integer:0-!',
  category:  'string!',
  createdAt: 'datetime!'
});

// MongoDB $jsonSchema (for db.createCollection() document validation; not a Mongoose model schema)
const mongoSchema = new exporters.MongoDBExporter().export(productSchema);
/*
{
  $jsonSchema: {
    bsonType: 'object',
    properties: {
      name:      { bsonType: 'string', minLength: 1, maxLength: 100 },
      price:     { bsonType: 'double', minimum: 0 },
      stock:     { bsonType: 'int',    minimum: 0 },
      category:  { bsonType: 'string' },
      createdAt: { bsonType: 'string' }
    },
    required: ['name', 'price', 'stock', 'category', 'createdAt']
  }
}
*/

// MySQL DDL
const mysqlDDL = new exporters.MySQLExporter().export('products', productSchema);
/*
CREATE TABLE `products` (
  `name`      VARCHAR(100) NOT NULL,
  `price`     DOUBLE NOT NULL,
  `stock`     BIGINT NOT NULL,
  `category`  VARCHAR(255) NOT NULL,
  `createdAt` DATETIME NOT NULL
) ENGINE=InnoDB DEFAULT CHARSET=utf8mb4 COLLATE=utf8mb4_unicode_ci;
*/

// Markdown field documentation
const markdown = exporters.MarkdownExporter.export(productSchema, { title: 'Product Field Reference' });

🗒️ Feature Overview

Common use cases

Use case	API	Docs
API parameter validation	validateAsync + ValidationError	Async Validation
Form / script validation	validate()	validate()
Batch data validation	SchemaUtils.validateBatch()	SchemaUtils
create / update derivation	pick / omit / partial	SchemaUtils
Database table creation	MongoDBExporter / MySQLExporter	Export Guide
Field documentation	MarkdownExporter	Export Guide
Multilingual API errors	I18nError	Error Handling
Conditional / dynamic rules	s.if() / s.match()	Conditional API
Custom DSL types	registerExtensions([...]) / s.registerExtension()	Extensions Overview
No global String extension	schema-dsl/pure	API Reference
Compile-time string-chain transform	transformSchemaDsl() / schemaDslEsbuildPlugin()	API Reference

📖 DSL Syntax Reference

Basic types

s({
  // string
  name:     'string!',         // required
  code:     'string:6',        // exact length 6
  bio:      'string:-500',     // max length 500
  username: 'string:3-32',     // length range 3–32

  // number
  age:   'number:18-120',      // range 18–120
  score: 'integer:0-100',      // integer 0–100
  price: 'number:>0',          // strictly greater than 0
  level: 'number:>=1',         // greater than or equal to 1

  // enum
  status: 'active|inactive|pending',  // string enum
  tier:   'enum:number:1|2|3',        // numeric enum

  // array
  tags:  'array<string>',             // string array
  items: 'array:1-10<number>',        // 1–10 numeric elements

  // boolean
  active: 'boolean!',

  // union type
  contact: 'types:email|phone!',      // email or phone, required
  price2:  'types:number:0-|string',  // number or string
})

Built-in formats

s({
  email:     'email!',          // email address
  website:   'url!',            // URL
  birthday:  'date!',           // YYYY-MM-DD
  createdAt: 'datetime!',       // ISO 8601
  userId:    'uuid!',           // UUID
  phone:     'phone:cn!',       // Chinese mobile number
  idCard:    'idCard:cn!',      // Chinese national ID
  slug:      'slug:3-100!',     // URL-friendly string
})

Fluent chain API (s progressive entries)

import { s } from 'schema-dsl/pure';

const schema = s({
  username: s('string:3-32!')
    .username()
    .label('username')
    .messages({ required: 'Username is required' }),

  email: s('email!').label('email address'),

  phone: s('string:11!')
    .pattern(/^1[3-9]\d{9}$/)
    .label('phone number'),

  recoveryEmail: s.email()
    .label('recovery email')
    .pattern(/@company\.com$/)
    .require(),
});

Conditional validation

// s.match — route to different rules based on a field value
const contactSchema = s({
  type:    'email|phone|wechat',
  contact: s.match('type', {
    email:  'email!',
    phone:  'string:11!',
    wechat: 'string:6-20!',
  })
});

// s.if — simple conditional branch
const orderSchema = s({
  isVip:    'boolean!',
  discount: s.if('isVip', 'number:10-50!', 'number:0-10')
});

// s.if chain assertion
s.if(d => !d.account)
  .message('Account not found')
  .and(d => d.account.balance < amount)
  .message('Insufficient balance')
  .assert(data);

🌍 Internationalization

import { s, validate, Locale, I18nError } from 'schema-dsl/pure';

// built-in locales: zh-CN / en-US / ja-JP / es-ES / fr-FR (auto-loaded, no configuration needed)
const result = validate(schema, data, { locale: 'en-US' });
// error messages automatically use the specified locale

// register a custom locale
Locale.addLocale('zh-CN', {
  'user.notFound':    'User not found',
  'user.forbidden':   { code: 40003, message: 'Access forbidden' },
});

// throw i18n business errors
I18nError.assert(user, 'user.notFound');                    // auto-throw when user is falsy
I18nError.throw('user.forbidden', {}, 403);                 // throw directly
I18nError.assert(ok, 'user.notFound', {}, 404, locale);     // specify locale at runtime

// errors carry a numeric code; frontend can branch on it
try {
  await api.getUser(id);
} catch (error) {
  switch (error.code) {
    case 40003: showForbiddenPage(); break;
  }
}

🔌 Extensions and Plugin Packaging

Use custom DSL types when one business type should work through pure DSL strings, s('...'), and s.xxx() factories. Use PluginManager when you want to package validator formats, keywords, lifecycle hooks, or several extension hooks together.

import { PluginManager, Validator, s } from 'schema-dsl/pure';

const pluginManager = new PluginManager();

// register a custom format plugin (must provide an install function)
pluginManager.register({
  name: 'extra-formats',
  install(core) {
    const validator = core as Validator;
    // register custom formats on the Validator instance via addFormat
    validator.addFormat('hex-color', {
      validate: (v: string) => /^#[0-9A-F]{6}$/i.test(v)
    });
    validator.addFormat('mac-address', {
      validate: (v: string) => /^([0-9A-F]{2}:){5}[0-9A-F]{2}$/i.test(v)
    });
  }
});

// create a Validator and install plugins
const validator = new Validator();
pluginManager.install(validator);

// use the custom formats in a schema
const schema = s({ color: 'hex-color!', mac: 'mac-address' });
const result = validator.validate(schema, { color: '#FF5733', mac: '00:1A:2B:3C:4D:5E' });

🔧 Core API Reference

API	Purpose	Returns	Docs
s(schema)	Create a schema	Schema object	DSL Syntax
validate(schema, data)	Synchronous validation	{ valid, errors, data }	validate()
validateAsync(schema, data)	Asynchronous validation	Promise (throws on failure)	Async Validation
SchemaUtils.pick()	Select fields	New schema	SchemaUtils
SchemaUtils.omit()	Exclude fields	New schema	SchemaUtils
SchemaUtils.partial()	Make all fields optional, or only selected fields optional when fields is provided	New schema	SchemaUtils
s.if(condition)	Conditional validation	ConditionalBuilder	Conditional API
s.match(field, map)	Branch validation	ConditionalBuilder	Conditional API
I18nError.throw()	Throw an i18n error	never	Error Handling
I18nError.assert()	Assert then throw	void	Error Handling
schema-dsl/pure	Import the API without installing String extensions	API namespace	API Reference
schema-dsl/string-types	Opt into TypeScript hints for String-chain authoring	Type declarations	TypeScript Usage
transformSchemaDsl()	Rewrite static string-chain DSL calls at compile time	{ code, changed, warnings }	API Reference
schemaDslEsbuildPlugin()	Use the transform in esbuild build/context flows	esbuild plugin	API Reference

📝 TypeScript Usage

import { s, validateAsync, ValidationError } from 'schema-dsl/pure';

// ✅ wrap strings with s() in TypeScript for builder method hints
const userSchema = s({
  username: s('string:3-32!').label('username'),
  email:    s('email!').label('email'),
  age:      s('number:18-100').label('age')
});

try {
  const validData = await validateAsync(userSchema, payload);
  // validData's static type is controlled by the generic passed to validateAsync<T>.
  // Use InferSchema / InferDslDefinition for schema-literal value type extraction.
} catch (error) {
  if (error instanceof ValidationError) {
    error.errors.forEach(e => console.log(`${e.path}: ${e.message}`));
  }
}

Note: In TypeScript projects, use s('...') or s.xxx() to get builder chain hints without adding global String declarations. DSL string literals also support lightweight value-type extraction through InferSchema / InferDslString, but constraints such as length ranges, regexes, defaults, and custom validators remain runtime schema rules. See the TypeScript Guide for details.

🛠️ Development

npm run build      # compile TypeScript
npm run test       # run tests
npm run typecheck  # type check
npm run docs:linkcheck # check docs links

Local documentation preview:

cd website
npm run dev

🤝 Contributing

git clone https://github.com/vextjs/schema-dsl.git
cd schema-dsl
npm install
npm test

See CONTRIBUTING.md for details.

🔗 Links

📖 Core documentation

• Quick Start — up and running in 5 minutes
• DSL Syntax Guide — complete syntax reference
• validate() — synchronous validation API
• API Reference — complete API docs
• TypeScript Guide — required reading for TS users
• Best Practices — avoid common pitfalls
• Troubleshooting — diagnosing issues

🎯 Feature documentation

• SchemaUtils
• Conditional Validation API
• Async Validation
• Error Handling & i18n
• Union Types
• Enum Types

🗄️ Export & integration

• Export Guide
• MongoDB Exporter
• MySQL Exporter
• PostgreSQL Exporter
• Markdown Exporter
• ⚠️ Export Limitations

💻 Examples

• Run all documentation examples with npm run examples:run.
• quick-start.ts — basic usage and registration form
• validate-async.ts — async validation and ValidationError handling
• export-guide.ts — database export overview
• error-handling.ts — field errors and business error handling
• object-dsl-builder.ts — object builder chaining and required-field control
• real-world.ts — combined production-style schemas for users, products, orders, and queries
• plugin-system.ts — plugin system and hooks

📝 Changelog & contributing

• Changelog
• Contributing Guide
• Security Policy

📄 License

Apache-2.0

If this project is useful to you, please consider giving it a Star ⭐

Made with ❤️ by the schema-dsl team