Apathetic Python Schema 🧩

📘 Roadmap · 📝 Release Notes

Lightweight validation for dict-based configs.
Types can be schema too.

Apathetic Python Schema validates dict-based data structures (configs, usually from JSONC/JSON/TOML) using Python TypedDicts used for mypy/pyright. No need for separate schema definitions — your type annotations are your schema.

Note

Heads up: the AI cooked dinner. It's edible, but watch your step.
Detailed bug reports welcome.

Quick Start

from apathetic_schema import apathetic_schema, ApatheticSchema_ValidationSummary
from apathetic_utils import schema_from_typeddict
from typing import TypedDict
from pathlib import Path
import json

# Define your config schema using TypedDict
class AppConfig(TypedDict):
    name: str
    version: str
    port: int
    debug: bool

# Load config from JSON
with Path("config.json").open() as f:
    config = json.load(f)

# Validate against TypedDict schema
summary = ApatheticSchema_ValidationSummary(valid=True, errors=[], strict_warnings=[], warnings=[], strict=False)
schema = schema_from_typeddict(AppConfig)

is_valid = apathetic_schema.check_schema_conformance(
    config,
    schema,
    "in configuration file",
    strict_config=False,
    summary=summary,
)

if not summary.valid:
    print("Validation errors:", summary.errors)
    print("Warnings:", summary.warnings)

Installation

# Using poetry
poetry add apathetic-schema

# Using pip
pip install apathetic-schema

Documentation

📚 Full Documentation →

For installation guides, API reference, examples, and more, visit our documentation website.

Features

• 🔍 TypedDict validation — Validate dict-based configs using Python TypedDicts
• 📋 Recursive validation — Supports nested TypedDicts and list types
• ⚠️ Error aggregation — Collect and report validation errors and warnings
• 🎯 Strict mode — Optional strict validation that treats warnings as errors
• 🔧 Type-aware — Works with mypy/pyright type annotations
• 🪶 Minimal dependencies — Only requires apathetic-utils
• 📝 Helpful errors — Provides context and suggestions for validation failures

---

⚖️ License

• MIT-a-NOAI License

You're free to use, copy, and modify the library under the standard MIT terms.
The additional rider simply requests that this project not be used to train or fine-tune AI/ML systems until the author deems fair compensation frameworks exist.
Normal use, packaging, and redistribution for human developers are unaffected.