Converts a JSON schema to a Joi schema for object validation.
enjoi
does not support all of json-schema.
Here is a list of some known missing keyword support still being worked on:
object:patternProperties
- unsupported due to Joi limitations.
Please file issues for other unsupported features.
enjoi.schema(schema [, options])
schema
- a JSON schema or a string type representation (such as'integer'
).options
- an (optional) object of additional options such assubSchemas
and customtypes
.
enjoi.defaults(options)
- configure defaultsoptions
to be used with allenjoi.schema
calls.enjoi.schema
options passed will always override defaults set here.
subSchemas
- an (optional) object with keys representing schema ids, and values representing schemas.refineType(type, format)
- an (optional) function to call to apply to type based on the type and format of the JSON schema.refineSchema(joiSchema, jsonSchema)
- an (optional) function to call to apply to adjust Joi schema base on the original JSON schema. Primary use case is handlingnullable
flag in OpenAPI 3.0extensions
- an array of extensions to pass joi.extend.strictMode
- make schemasstrict(value)
with a default value offalse
.
Example:
const Joi = require('joi');
const Enjoi = require('enjoi');
const schema = Enjoi.schema({
type: 'object',
properties: {
firstName: {
description: 'First name.',
type: 'string'
},
lastName: {
description: 'Last name.',
type: 'string'
},
age: {
description: 'Age in years',
type: 'integer',
minimum: 1
}
},
'required': ['firstName', 'lastName']
});
const { error, value } = schema.validate({firstName: 'John', lastName: 'Doe', age: 45});
Sub-schemas can be provided through the subSchemas
option for $ref
values to lookup against.
Example:
const schema = Enjoi.schema({
type: 'object',
properties: {
a: {
$ref: '#/b' // # is root schema
},
b: {
type: 'string'
},
c: {
$ref: 'sub#/d' // sub# is 'sub' under subSchemas.
}
}
}, {
subSchemas: {
sub: {
d: {
'type': 'string'
}
}
}
});
The above example subSchemas
can be added instead via defaults:
const enjoi = Enjoi.defaults({
subSchemas: {
sub: {
d: {
'type': 'string'
}
}
}
});
const schema = enjoi.schema({
type: 'object',
properties: {
a: {
$ref: '#/b' // # is root schema
},
b: {
type: 'string'
},
c: {
$ref: 'sub#/d' // sub# is 'sub' under subSchemas.
}
}
});
Custom types can be provided through the extensions
option.
const schema = Enjoi.schema({
type: 'thing'
}, {
extensions: [{
type: 'thing',
base: Joi.any()
}]
});
Also with functions.
const schema = Enjoi.schema({
type: 'thing'
}, {
extensions: [{
type: 'thing',
validate(value, helpers) {
if (value !== 'foobar') {
return { value, errors: helpers.error('thing.foobar') };
}
},
messages: {
'thing.foobar': '{#label} must be \'foobar\''
}
}]
});
You can use the refine type function to help refine types based on type
and format
. This will allow transforming a type for lookup.
const schema = Enjoi.schema({
type: 'string',
format: 'email'
}, {
extensions: [{
type: 'email',
base: Joi.string().email()
}],
refineType(type, format) {
if (type === 'string' && format === 'email') {
return 'email';
}
}
});
This can be used in conjunction with function based extensions
for additional logic:
const schemaDesc = {
type: 'string',
format: 'email',
'x-test': true
}
const schema = Enjoi.schema(schemaDesc, {
extensions: [{
type: 'email',
validate(value, helpers) {
const validator = schemaDesc['x-test'] ? Joi.string().email().equal('[email protected]') : Joi.string().email();
const validation = validator.validate(value);
if (validation.error) {
return { value, errors: validation.error };
}
}
}],
refineType(type, format) {
if (type === 'string' && format === 'email') {
return 'email';
}
}
});
Refer to Joi documentation on extensions: https://hapi.dev/module/joi/api/?v=17#extensions