Schema

Simplified version of JSON Schema

In a scenario, the variables, process result, and actors are defined using JSON Schema. JSON Schema is a standard for defining the structure, format, and validation rules of JSON data.

To simplify usage, LetsFlow supports both standard and simplified schemas.

YAML

vars:
  message: string
  contract:
    properties:
      id: !format uri
      name: !required string
      date: !format date-time
      signed: !default false
  books:
    items:
      properties:
        id: !format uuid
        title: string
        isbn: !pattern: "^\\d{13}$"

JSON

For more information about JSON Schema, please visit:

JSON Schemajson-schema.org

Schema

https://schemas.letsflow.io/v1.0/schema

Normalization

Simplified schemas are converted into their full JSON Schema equivalent before use. This is done when a scenario or a standalone schema is normalized.

Simplified schemas supersede full JSON Schema, thus any valid JSON Schema is also a valid LetsFlow schema.

Simplification features

Simple types

Basic types like string, number, or boolean can be specified directly without wrapping in a type field.

name: string
age: integer

normalizes to

name:
  type: string
age:
  type: integer

Simple References

In JSON Schema, you can reference an external schema or subsection of the current schema using $ref.

In LetsFlow's simplified schema, references can be written directly as strings. Any string starting with http://, https://, or # is seen as a reference, rather than a type.

client: https://schema.example.com/actors/client

normalizes to

client:
  $ref: https://schema.example.com/actors/client

Yaml tags

YAML tags in LetsFlow provide a concise way to define additional metadata or constraints in your schema.

These tags are not available when defining a schema in JSON.

!const

Defines a constant value:

status: !const "active"

!enum

Defines an enumerated set of values:

role: !enum ["admin", "user", "guest"]

The type is determined based on the values of the enum.

!format

Specifies a data format:

email: !format email

The type is always set to string.

!pattern

Defines a regular expression for validation:

phone: !pattern "^\\+?[0-9]{10,15}$"

The type is always set to string.

!default

Provides a default value:

enabled: !default true

The type is determined based on the default value.

!required

Marks a property as required:

properties:
  name: !required string
  age: !required
    type: integer
    minimum: 16
    maximum: 120

The !required tag should only be used on properties, as the property is automatically appended to the required field of the parent object.

Implicit Types

If type is omitted, it may be automatically determined based on the structure:

• The presence of properties implies an object.

• The presence of items implies an array.

• For const or enum the type is determined based on the const or enum value.

Additional properties

Properties that aren't defined don't exist and can't be set unless additionalProperties is set in the definition.

In other words; additionalProperties defaults to false in contrary to the default behavior of JSON Schema.

Immutability

For schema files; LetsFlow assumes that schemas are immutable to be deterministic. It's highly recommended that the URI of the schema ID includes a version number. Modifying a published schema can lead to unexpected behavior.