Type safe output

What would be the benefit of building a validation library without a type safe output?

Inspired by the Zod parse output type, Regle will also infer your validator types to know which properties are guaranteed to be defined.

$validate

Using r$.$validate will asynchronously run and wait for all your validators to finish, and will return an object containing the status of your form.

const { valid, data } = await r$.$validate();

If your result is true, the data will be type safe.

It will check if the required rule is present, but it will not work with requiredIf or requiredUnless, because we can't know the condition at build time.

type Form = {
  firstName?: string;
  lastName?: string;
}

const form = ref<Form>({ firstName: '', lastName: '' })

const { r$ } = useRegle(form, {
  lastName: { required },
});

async function submit() {
  const { valid, data } = await r$.$validate();

  if (valid) {
    console.log(data);
  }
}

InferSafeOutput

You can also statically infer the safe output from any r$ instance.

import { useRegle, InferSafeOutput } from '@regle/core';
type Form = {
  firstName?: string;
  lastName?: string;
}

const form: Ref<Form> = ref({ firstName: '', lastName: '' })

const { r$ } = useRegle(form, {
  lastName: { required },
});

type FormRequest = InferSafeOutput<typeof r$>;