serpyco-rs: High-Performance Dataclass Serialization and Validation

Warnings

• breaking As of v1.20.0, `Min` and `Max` validators are now inclusive by default. If your existing code or schemas relied on exclusive behavior without explicitly setting an `inclusive` flag, this change will alter validation logic.
  Fix: For explicit control, use the `inclusive` flag (e.g., `Min(value, inclusive=False)` or `Max(value, inclusive=False)`) to restore exclusive behavior where needed.
• gotcha Prior to v1.17.1, boolean values (`true`/`false`) were incorrectly accepted as integers during validation. If you were implicitly relying on this behavior, validation might now fail for such inputs.
  Fix: Ensure inputs strictly adhere to integer types where integers are expected, and booleans where booleans are expected. Upgrade to v1.17.1 or newer for correct type validation.
• gotcha For JSON Schema generation, simple union types now correctly generate `anyOf` constructs (since v1.17.0) and integer fields explicitly include `"format": "int64"` (since v1.15.0). These changes improve JSON Schema compliance but might cause discrepancies for consumers expecting older schema formats.
  Fix: Review generated JSON schemas after upgrading to ensure compatibility with downstream schema consumers. Adjust expectations or tooling to accommodate the more compliant schema output.

Install

• pip install serpyco-rs Install serpyco-rs

Imports

• Serializer

  from serpyco_rs import Serializer

  The primary class for defining serializers for dataclasses.
• dataclass

  from dataclasses import dataclass

  Used to define the data structures that serpyco-rs serializes.

Quickstart

This example demonstrates how to define a dataclass, create a serializer for it, and then use the serializer to dump a dataclass instance to a dictionary, load a dictionary into a dataclass instance, and generate a JSON Schema for the defined structure.

import dataclasses
from serpyco_rs import Serializer

@dataclasses.dataclass
class User:
    id: int
    name: str
    email: str
    is_active: bool = True

# Create a serializer for the User dataclass
user_serializer = Serializer(User)

# Create a User instance
user_instance = User(id=1, name="Alice", email="alice@example.com")

# Serialize the dataclass instance to a dictionary
dumped_data = user_serializer.dump(user_instance)
print(f"Dumped data: {dumped_data}")
# Expected: {'id': 1, 'name': 'Alice', 'email': 'alice@example.com', 'is_active': True}

# Load data from a dictionary into a dataclass instance
loaded_data = {'id': 2, 'name': 'Bob', 'email': 'bob@example.com', 'is_active': False}
user_from_dict = user_serializer.load(loaded_data)
print(f"Loaded instance: {user_from_dict}")
# Expected: User(id=2, name='Bob', email='bob@example.com', is_active=False)

# Generate JSON Schema for the dataclass
json_schema = user_serializer.json_schema()
print(f"JSON Schema: {json_schema}")