koz 0.1.0

Python monkeypatch detection and analysis tool

koz

Python monkeypatch detection and analysis tool.

Overview

koz is a static analysis tool that detects and catalogs monkeypatch modifications in Python codebases. It uses AST parsing to identify patches and generates structured output for patch management.

Features

• Detects three types of monkeypatching:
  • @patch_to decorator from fastcore.basic
  • Direct attribute assignment monkeypatching
  • functools.wraps decorator in __init__ methods
• Generates JSON/TOML output with comprehensive metadata
• Integrates with git history for author and timestamp information
• Fast AST-based analysis without code execution

Installation

pip install -e .

Usage

Command Line

# Analyze current directory
koz analyze

# Analyze specific directory
koz analyze /path/to/project

# Output to TOML
koz analyze --format toml --output patches.toml

# Output to JSON
koz analyze --format json --output patches.json

Detected Patterns

1. fastcore.basic.patch_to Decorator

The most common pattern in fastai/fastcore projects:

from fastcore.basic import patch_to

class MyClass:
    def __init__(self):
        self.value = 0

@patch_to(MyClass)
def new_method(self):
    """This method is added to MyClass via patch_to."""
    return self.value * 2

2. Direct Attribute Assignment

def custom_method(self):
    return "custom"

MyClass.method = custom_method

3. functools.wraps Pattern

from functools import wraps

class Wrapper:
    def __init__(self):
        @wraps(TargetClass.method)
        def wrapped(self):
            return "wrapped"
        TargetClass.method = wrapped

Development

# Install development dependencies
pip install -e ".[dev]"

# Run tests
pytest tests/ -v

# Run linter
ruff check src/ tests/

# Format code
black src/ tests/

# Type check
mypy src/

# Install pre-commit hooks
pre-commit install

License

Unlicense