Skip to main content

How To Make Custom Schema

For advanced schema authors, this page shows how to build custom schemas with schemaFactory.

Step 1: define schema input​

import {schemaFactory} from '@livon/schema';

interface SlugSchemaInput {
  name: string;
}

Parameters in this example​

SlugSchemaInput:

  • name (string): schema name used in AST and error metadata.

Step 2: create schema with validate + ast​

import {schemaFactory} from '@livon/schema';

interface SlugSchemaInput {
  name: string;
}

export const slug = ({name}: SlugSchemaInput) =>
  schemaFactory<string>({
    name,
    type: 'slug',
    ast: () => ({type: 'slug', name}),
    validate: (input) => {
      if (typeof input !== 'string') {
        throw {message: 'Expected string', code: 'slug.type'};
      }

      const slugRegex = /^[a-z0-9]+(?:-[a-z0-9]+)*$/;
      if (!slugRegex.test(input)) {
        throw {message: 'Expected slug format', code: 'slug.format'};
      }

      return input;
    },
  });

Parameters in this example​

slug({name}):

  • name (string): custom schema node name.

schemaFactory({...}):

  • name (string): schema name.
  • type (string): schema type id.
  • ast ((ctx) => AstNode): AST builder function.
  • validate ((input, ctx) => value): runtime validator function.

Step 3: use custom schema in object schemas​

import {object, string} from '@livon/schema';
import {slug} from './slug.js';

const Article = object({
  name: 'Article',
  shape: {
    id: string({name: 'ArticleId'}),
    slug: slug({name: 'ArticleSlug'}),
  },
});

Parameters in this example​

object({...}):

  • name (string): object schema name.
  • shape (Record<string, Schema>): field schema map.

string({name}):

  • name (string): field schema node name for AST/error output.

slug({name}):

  • name (string): field-specific custom schema node name.

Rules for custom schemas​

  1. Keep one schema per file.
  2. Export through src/index.ts.
  3. Provide a dedicated docs page with usage examples in Schema APIs.