invertase/globeLight logoGlobe
PR #305

Validate API Requests with Luthor

Use schema-based validation to protect your Dart APIs from malformed or malicious data.

Manual validation with if statements quickly becomes repetitive and error-prone as your API grows. Schema-based validation provides a cleaner, more maintainable approach by defining the expected shape of your data once and validating against it consistently.

This guide shows you how to use Luthor, a powerful Dart validation library, to validate incoming API requests in a Dart Frog backend. You'll learn to define schemas, validate request bodies, and return structured error responses.

15 min read

Features Covered

  • Installing and setting up Luthor in a Dart Frog project
  • Defining validation schemas for API endpoints
  • Validating request bodies against schemas
  • Returning structured validation error responses
  • Creating reusable validation helpers

Prerequisites

  • Dart SDK Installed: If you have Flutter installed, the Dart SDK is already included. If not, Install Dart.
  • Globe Account: You'll need an account to deploy projects. Sign up or log in to Globe.
  • Globe CLI Installed and Authenticated: Install the CLI by running dart pub global activate globe_cli and log in using globe login.
  • Dart Frog CLI Installed: Install by running dart pub global activate dart_frog_cli.

Step 1: Create Your Project

Bootstrap a new Dart Frog project to work with.

  • In your terminal, run the following commands:

    dart_frog create my_validated_api
cd my_validated_api

Step 2: Add Luthor Dependency

Add the Luthor validation library to your project.

  • Run the following command:

    dart pub add luthor

Step 3: Understand Luthor Basics

Luthor uses a schema-based approach similar to Zod (JavaScript) or Yup. You define the expected structure of your data, then validate incoming data against that schema.

Here's a quick overview of common validators:

import 'package:luthor/luthor.dart';

// String validators
l.string()              // Must be a string
l.string().min(1)       // Minimum 1 character
l.string().max(100)     // Maximum 100 characters
l.string().email()      // Must be valid email format

// Number validators
l.int()                 // Must be an integer
l.double()              // Must be a double
l.int().min(0)          // Minimum value 0
l.int().max(150)        // Maximum value 150

// Other validators
l.boolean()             // Must be a boolean
l.list()                // Must be a list

// Making fields required (fields are optional by default)
l.string().required()   // Required string field
l.int().required()      // Required integer field

// Combining validators
l.string().min(8).max(128).required()  // Required string, 8-128 chars

Step 4: Create a User Registration Schema

Let's create a practical example: validating user registration data.

  • Create a new file at lib/schemas/user_schema.dart:

    import 'package:luthor/luthor.dart';

/// Schema for user registration requests.
/// Expects: { "email": string, "password": string, "name": string, "age": int? }
final registerUserSchema = l.schema({
  'email': l.string().email().required(),
  'password': l.string().min(8).max(128).required(),
  'name': l.string().min(1).max(100).required(),
  'age': l.int().min(0).max(150), // Optional by default
});

/// Schema for user login requests.
/// Expects: { "email": string, "password": string }
final loginUserSchema = l.schema({
  'email': l.string().email().required(),
  'password': l.string().min(8).required(),
});

Fields are optional by default in Luthor. Use .required() to make a field mandatory.

Step 5: Create a Validation Helper

To keep your route handlers clean, create a helper function that validates requests and returns structured errors.

  • Create a new file at lib/helpers/validation_helper.dart:

    import 'dart:io';
import 'package:dart_frog/dart_frog.dart';
import 'package:luthor/luthor.dart';

/// Validates a request body against a Luthor schema.
///
/// Returns the validated data as a Map if successful.
/// Returns a 400 Bad Request Response with structured errors if validation fails.
Future<Object> validateRequest(
  RequestContext context,
  Validator schema,
) async {
  final body = await context.request.json() as Map<String, dynamic>;
  final result = schema.validateSchema<Map<String, dynamic>>(body);

  switch (result) {
    case SchemaValidationSuccess():
      return body;
    case SchemaValidationError(errors: final errors):
      return Response.json(
        statusCode: HttpStatus.badRequest,
        body: {
          'message': 'Validation failed',
          'errors': errors,
        },
      );
  }
}

/// Extension to check if validation returned data or an error response.
extension ValidationResultCheck on Object {
  bool get isValidationError => this is Response;
  Response get asErrorResponse => this as Response;
  Map<String, dynamic> get asValidData => this as Map<String, dynamic>;
}

Step 6: Create Validated API Endpoints

Now use the schema and helper in your route handlers.

  • Create a file at routes/auth/register.dart:

    import 'dart:io';
import 'package:dart_frog/dart_frog.dart';
import 'package:my_validated_api/helpers/validation_helper.dart';
import 'package:my_validated_api/schemas/user_schema.dart';

Future<Response> onRequest(RequestContext context) async {
  if (context.request.method != HttpMethod.post) {
    return Response(statusCode: HttpStatus.methodNotAllowed);
  }

  // Validate the request body
  final validationResult = await validateRequest(
    context,
    registerUserSchema,
  );

  // If validation failed, return the error response
  if (validationResult.isValidationError) {
    return validationResult.asErrorResponse;
  }

  // Validation passed - safely access the data
  final data = validationResult.asValidData;
  final email = data['email'] as String;
  final password = data['password'] as String;
  final name = data['name'] as String;
  final age = data['age'] as int?; // Optional field

  // TODO: Add your business logic here
  // - Check if user already exists
  // - Hash the password
  // - Save to database

  return Response.json(
    statusCode: HttpStatus.created,
    body: {
      'message': 'User registered successfully',
      'user': {
        'email': email,
        'name': name,
        if (age != null) 'age': age,
      },
    },
  );
}

  • Create a file at routes/auth/login.dart:

    import 'dart:io';
import 'package:dart_frog/dart_frog.dart';
import 'package:my_validated_api/helpers/validation_helper.dart';
import 'package:my_validated_api/schemas/user_schema.dart';

Future<Response> onRequest(RequestContext context) async {
  if (context.request.method != HttpMethod.post) {
    return Response(statusCode: HttpStatus.methodNotAllowed);
  }

  // Validate the request body
  final validationResult = await validateRequest(context, loginUserSchema);

  if (validationResult.isValidationError) {
    return validationResult.asErrorResponse;
  }

  final data = validationResult.asValidData;
  final email = data['email'] as String;
  final password = data['password'] as String;

  // TODO: Add your authentication logic here
  // - Find user by email
  // - Verify password hash
  // - Generate JWT token

  return Response.json(
    body: {
      'message': 'Login successful',
      'token': 'your-jwt-token-here',
    },
  );
}

Step 7: Test Your Validation Locally

Start the development server and test your validation with various inputs.

  • Start the server:

    dart_frog dev

  • In a new terminal, test with valid data:

    curl -X POST http://localhost:8080/auth/register \
  -H "Content-Type: application/json" \
  -d '{"email": "ada@example.com", "password": "securepass123", "name": "Ada Lovelace"}'

    Expected response:

    {
  "message": "User registered successfully",
  "user": { "email": "ada@example.com", "name": "Ada Lovelace" }
}

  • Test with invalid data (missing fields, wrong types):

    curl -X POST http://localhost:8080/auth/register \
  -H "Content-Type: application/json" \
  -d '{"email": "not-an-email", "password": "short"}'

    Expected response:

    {
  "message": "Validation failed",
  "errors": {
    "email": ["email must be a valid email address"],
    "password": ["password must be at least 8 characters"],
    "name": ["name is required"]
  }
}

Step 8: Create More Complex Schemas

For more advanced use cases, Luthor supports nested schemas and lists.

  • Create a file at lib/schemas/post_schema.dart:

    import 'package:luthor/luthor.dart';

/// Schema for creating a blog post with tags.
final createPostSchema = l.schema({
  'title': l.string().min(1).max(200).required(),
  'content': l.string().min(10).required(),
  'tags': l.list(validators: [l.string().min(1)]),
  'metadata': l.schema({
    'draft': l.boolean(),
  }),
});

This validates:

  • title: Required string (.required()), 1-200 characters
  • content: Required string (.required()), at least 10 characters
  • tags: Optional array of non-empty strings (no .required())
  • metadata: Optional object with a draft boolean (no .required())

Step 9: Named Schemas for Better Debugging

Use named schemas for clearer error messages:

final userSchema = l.withName('User').schema({
  'email': l.string().email().required(),
  'password': l.string().min(8).required(),
});

Step 10: Deploy to Globe

Deploy your validated API to Globe.

globe deploy --prod

Step 11: Test the Live API

After deployment, test your live endpoints:

# Test successful registration
curl -X POST https://<YOUR_GLOBE_URL>/auth/register \
  -H "Content-Type: application/json" \
  -d '{"email": "user@example.com", "password": "securepass123", "name": "Test User"}'

# Test validation errors
curl -X POST https://<YOUR_GLOBE_URL>/auth/register \
  -H "Content-Type: application/json" \
  -d '{"email": "invalid", "password": "123"}'

What's Next

Couldn't find the guide you need? Talk to us in Discord

Edit this page on GitHub
Secure Dart APIs
Auth Email Verification