Schema Types¶

Define your expected output structure using strutex's schema types.

Basic Types¶

String¶

from strutex import String

# Standard
name = String(description="Customer name")

# Simplified (if no args needed)
middle_name = String  # equivalent to String()

Number¶

For floating-point values:

from strutex import Number

price = Number  # equivalent to Number()

Integer¶

For whole numbers:

from strutex import Integer

quantity = Integer(description="Item count")

Boolean¶

from strutex import Boolean

is_paid = Boolean(description="Payment status")

Specialized Types¶

Enum¶

Restrict output to a specific set of strings.

from strutex import Enum

category = Enum(
    values=["Food", "Transport", "Lodging"],
    description="Expense category"
)

Date¶

Extracts a string formatted as YYYY-MM-DD.

from strutex import Date

invoice_date = Date(description="Invoice date")

DateTime¶

Extracts a string formatted as ISO 8601 (YYYY-MM-DDTHH:MM:SS).

from strutex import DateTime

created_at = DateTime(description="Metadata creation time")

Complex Types¶

Array¶

from strutex import Array, String, Object

# Array of strings
tags = Array(items=String(), description="Item tags")

# Array of objects
items = Array(
    items=Object(
        properties={
            "name": String(),
            "price": Number()
        }
    )
)

Object¶

from strutex import Object, String

address = Object(
    description="Shipping address",
    properties={
        "street": String(),
        "city": String(),
        "zip": String()
    }
)

Required vs Optional¶

By default, all properties are required. To make fields optional:

Explicit RequiredNullable Fields

schema = Object(
    properties={
        "name": String(),
        "email": String()
    },
    required=["name"]  # Only name is required
)

schema = Object(
    properties={
        "name": String(),
        "notes": String(nullable=True)
    }
)

Complete Example¶

from strutex import Object, String, Number, Integer, Array, Boolean

invoice_schema = Object(
    description="Complete invoice",
    properties={
        "invoice_number": String(description="Unique ID"),
        "date": String(description="YYYY-MM-DD"),
        "vendor": Object(
            properties={
                "name": String(),
                "address": String(nullable=True)
            }
        ),
        "items": Array(
            items=Object(
                properties={
                    "description": String(),
                    "quantity": Integer(),
                    "price": Number()
                }
            )
        ),
        "total": Number(),
    }
)

JSON Schema Serialization¶

You can convert any schema object into a standard JSON Schema dictionary, useful for debugging or integrating with other tools.

from strutex import String, Object
import json

schema = Object(
    properties={"name": String(description="User name")},
    required=["name"]
)

# Convert to dict
print(schema.to_dict())
# {
#     "type": "object",
#     "properties": {"name": {"type": "string", "description": "User name"}},
#     "required": ["name"],
#     "additionalProperties": False
# }

# Convert to JSON string
print(str(schema))