Mimesis is a Python library for generating high-quality, locale-aware fake data, intended for use in testing, database seeding, API mocking, and data anonymization. Unlike simpler lorem-ipsum generators, it produces realistic names, addresses, dates, financial data, internet-related strings, and more, across 47 locale variants. The architecture is modular: locale-independent and locale-dependent providers (e.g., Person, Address, Internet) load curated JSON datasets, while a suite of schema tools (Field, Fieldset, Schema, SchemaBuilder) allows declarative generation of flat or relational synthetic datasets with lazy evaluation, dependency resolution, and export to CSV/JSON/pickle. A custom Random class with weighted selection and deterministic seeding ensures reproducibility, and optional plugins integrate with factory_boy and pytest-randomly. The codebase is well-structured, extensively typed (mypy strict), and thoroughly tested, with a strong emphasis on performance (~20-30x faster than comparable libraries like Faker). External runtime dependencies are minimal; the only optional ones are pytz and factory-boy.
jsf is a Python library that generates fake data conforming to JSON schemas. It provides programmatic and CLI interfaces for creating realistic test data, supports custom data providers, maintains state for dependent data generation, validates output against schemas, and integrates with FastAPI for automatic API documentation. The architecture follows a modular design where each JSON schema type is handled by a specific class in jsf/schema_types/, with the core parsing logic in jsf/parser.py. The project uses Pants for building, has comprehensive test coverage, and maintains active CI/CD pipelines with automated releases based on conventional commits.