SYDA Schema Reference¶

This document defines the schema format for the SYDA (Synthetic Data) library. It describes the supported field types and constraints.

Basic Schema Structure¶

A schema in SYDA is defined as a JSON or YAML object with field names as keys and field types as values. The simplest form is:

field1:
  type: string
field2:
  type: number
field3:
  type: date

Field Types¶

The following field types are supported:

text¶

Generic text field for storing names, descriptions, and other textual content.

Example:

name:
  type: text

string¶

Same as text, used for storing character data.

Example:

title:
  type: string

number¶

Numeric field that can represent integers or decimal values.

Example:

id:
  type: number

integer¶

Integer value with no decimal places. Can also use the alias int.

Example:

age:
  type: integer

float¶

Floating-point number that can have decimal places.

Example:

price:
  type: float

date¶

Date value in standard format.

Example:

birth_date:
  type: date

datetime¶

Date and time value.

Example:

created_at:
  type: datetime

boolean¶

True/false value. Can also use the alias bool.

Example:

is_active:
  type: boolean

email¶

Email address with appropriate validation.

Example:

contact:
  type: email

phone¶

Phone number field.

Example:

telephone:
  type: phone

address¶

Physical address field.

Example:

location:
  type: address

url¶

URL/website address.

Example:

website:
  type: url

array¶

List of items. Arrays are typically populated using custom generators in the application code.

Example:

items:
  type: array

foreign_key¶

Reference to another table in the database.

Example:

user_id:
  type: foreign_key

Supported Field Properties¶

Fields can be defined with additional properties by using an object instead of a string:

name:
  type: text
  description: Full name of the person
  constraints:
    max_length: 50

The following properties can be used when defining fields:

description¶

Human-readable description of the field's purpose.

Example:

email:
  type: email
  description: "User's email address for notifications"

length¶

Fixed length for the field.

Example:

zip_code:
  type: string
  length: 10

min_length¶

Minimum length for string fields.

Example:

password:
  type: string
  min_length: 8

max_length¶

Maximum length for string fields.

Example:

bio:
  type: text
  max_length: 250

min¶

Minimum value for numeric fields.

Example:

min: 0

max¶

Maximum value for numeric fields.

Example:

max: 1000

decimals¶

Number of decimal places for float fields.

Example:

price:
  type: float
  decimals: 2

format¶

Format string specification (for dates, etc.).

Example:

creation_date:
  type: date
  format: "YYYY-MM-DD"

enum¶

List of possible values for the field.

Example:

status:
  type: string
  enum: ["active", "inactive", "pending"]

pattern¶

Regular expression pattern for validation.

Example:

pattern: "^[A-Z][a-z]+$"

references¶

Foreign key reference to another table and field.

Example:

user_id:
  type: foreign_key
  references: {"schema": "User", "field": "id"}

Note: Field properties are validated during schema validation. For example, using an invalid field type will cause validation to fail with a detailed error message.

Field Constraints¶

Constraints can be specified directly in the field definition or in a separate constraints object:

email:
  type: email
  constraints:
    unique: true
    not_null: true

The following constraints are supported:

not_null¶

Field cannot be null or empty.

Example:

username:
  type: string
  constraints:
    not_null: true

unique¶

Values must be unique within the dataset.

Example:

email:
  type: email
  constraints:
    unique: true

primary_key¶

Field is a primary key for the table.

Example:

id:
  type: integer
  constraints:
    primary_key: true

format¶

Format specification for specialized field types.

Example:

contact:
  type: string
  constraints:
    format: "email"

pattern¶

Regular expression pattern for validation.

Example:

pattern: "^[A-Z][a-z]+$"

min¶

Minimum value for numeric fields.

Example:

min: 0

max¶

Maximum value for numeric fields.

Example:

max: 1000

min_length¶

Minimum string length for text fields.

Example:

password:
  type: string
  constraints:
    min_length: 8

max_length¶

Maximum string length for text fields.

Example:

description:
  type: text
  constraints:
    max_length: 500

length¶

Exact string length for text fields.

Example:

country_code:
  type: string
  constraints:
    length: 2