py-data-cloak

py-data-cloak is a robust, rule-based data anonymization and sanitization engine that works seamlessly on static files (CSV/JSON/SQL dumps) via CLI and integrates natively into the Django ORM for "safe dumps."

Features

• Declarative Ruleset: Define how to mask data using YAML or Python dictionaries.
• Consistency: Maintains referential integrity (e.g., "User 1" is always masked to the same value within a session).
• Streaming Support: Efficiently handles large datasets.
• Django Integration: Drop-in dumpdata_masked command and MaskedModelMixin.
• CLI Tool: Process JSON/CSV files or pipe SQL dumps.

Installation

pip install py-data-cloak
# For Django support
pip install "py-data-cloak[django]"

Usage

CLI

1. Define your rules in privacy_rules.yaml:

   email: "faker:email"
   ssn: "mask_all_but_last_4"
   password: "fixed:secret"

2. Run the tool:

   pycloak process input.json --rules=privacy_rules.yaml --output=clean.json

Django

1. Add pycloak.django to your INSTALLED_APPS.
2. Use the management command:

   python manage.py dumpdata_masked auth.User --output=safe_users.json

Rules Reference

• faker:<provider>: Use any Faker provider (e.g., faker:name, faker:email).
• fixed:<value>: Replace with a static value.
• mask_all_but_last_<n>: Mask characters with * except the last n.
• clear: Set to null or empty string.