Objects

The object type validates key-value mappings.

Basic Object Validation

# Schema
type: object

Valid examples:

key: value
another_key: another_value

Sun: 1.9891e30
Jupiter: 1.8986e27

01: cm
m
km

Invalid examples:

"Not an object"

["An", "array", "not", "an", "object"]

Note: Unlike JSON, YAML allows numeric keys, which are treated as strings.

Properties

# Schema
type: object
properties:
  number:
    type: number
  street_name:
    type: string
  street_type:
    enum: [Street, Avenue, Boulevard]

Valid examples:

number: 1600
street_name: Pennsylvania
street_type: Avenue

number: 1600
street_name: Pennsylvania

{}

number: 1600
street_name: Pennsylvania
street_type: Avenue
direction: NW

Invalid examples:

number: "1600"
street_name: Pennsylvania
street_type: Avenue

Note: By default, leaving out properties is valid, and providing additional properties is also valid.

Pattern Properties

# Schema
type: object
patternProperties:
  ^S_:
    type: string
  ^I_:
    type: integer

Valid examples:

S_25: This is a string

I_0: 42

keyword: value

Invalid examples:

S_0: 42

I_42: This is a string

Additional Properties

By default, additional properties are allowed. You can restrict this:

# Schema
type: object
properties:
  number:
    type: number
  street_name:
    type: string
additionalProperties: false

Valid examples:

number: 1600
street_name: Pennsylvania

Invalid examples:

number: 1600
street_name: Pennsylvania
direction: NW

You can also specify a schema for additional properties:

# Schema
type: object
properties:
  number:
    type: number
additionalProperties:
  type: string

Valid examples:

number: 1600
direction: NW

Invalid examples:

number: 1600
office_number: 201

Required Properties

# Schema
type: object
properties:
  name:
    type: string
  email:
    type: string
required:
  - name
  - email

Valid examples:

name: William Shakespeare
email: bill@stratford-upon-avon.co.uk

name: William Shakespeare
email: bill@stratford-upon-avon.co.uk
address: Henley Street, Stratford-upon-Avon, Warwickshire, England
authorship: in question

Invalid examples:

name: William Shakespeare
address: Henley Street, Stratford-upon-Avon, Warwickshire, England

name: William Shakespeare
address: Henley Street, Stratford-upon-Avon, Warwickshire, England
email: null

Note: A property with a null value is considered not being present.

Property Names

propertyNames, by default, uses an implicit type: string and validates a pattern to each key’s string form (JSON Schema behavior).

# Schema
type: object
propertyNames:
  pattern: "^[A-Za-z_][A-Za-z0-9_]*$"

Valid examples:

_a_proper_token_001: "value"

Invalid examples:

-001 invalid: "value"

Error: [1:1] .: Property name '-001 invalid' does not match pattern '^[A-Za-z_][A-Za-z0-9_]*$'

Property Names (YAML-specific)

Unlike JSON, YAML may use non-string scalars as mapping keys (e.g. integers). JSON Schema’s propertyNames is defined in terms of string property names and only supports a pattern on that string projection.

In YAML Schema, propertyNames is extended to support a subschema to validate against each mapping key node as parsed YAML, the same way items validates array elements. Any keywords valid in a subschema apply (type, enum, const, composition, and so on).

The reference implementation is the yaml-schema crate and ys CLI (source on GitHub).

Integer keys

# Schema
type: object
propertyNames:
  type: integer

Valid examples:

1: one
2: two

Invalid examples:

hello: world

Error: [1:1] .hello: Expected a number, but got: "hello" (string)

String keys with enum

# Schema
type: object
propertyNames:
  type: string
  enum:
    - alpha
    - beta

Valid examples:

alpha: 1
beta: 2

Invalid examples:

gamma: 3

Error: [1:1] .gamma: Value "gamma" is not in the enum: ["alpha", "beta"]

Object Size

# Schema
type: object
minProperties: 2
maxProperties: 3

Valid examples:

a: 0
b: 1

a: 0
b: 1
c: 2

Invalid examples:

{}

a: 0

a: 0
b: 1
c: 2
d: 3