FastScaff

FastAPI project scaffolding tool - quickly generate standardized FastAPI project structures.

中文文档

Installation

pip install fastscaff

Commands

Create Project

fastscaff new myproject --orm sqlalchemy

Options:

• --orm - ORM choice: sqlalchemy or tortoise (default: tortoise)
• --output - Output directory (default: current directory)
• --with-rbac - Include Casbin RBAC support
• --with-celery - Include Celery task queue support
• --force - Overwrite existing directory

Examples:

# Basic project with SQLAlchemy
fastscaff new myproject --orm sqlalchemy

# Full-featured project
fastscaff new myproject --orm sqlalchemy --with-celery --with-rbac

# Specify output directory
fastscaff new myproject --output /path/to/dir

Generate Models from Database

Generate ORM models by introspecting existing MySQL database tables:

cd myproject
fastscaff models --db-url "mysql://user:pass@localhost:3306/mydb"

Options:

• --db-url - Database connection URL (required)
• --orm - Target ORM: sqlalchemy or tortoise (auto-detected from requirements.txt)
• --tables - Comma-separated table names (default: all tables)
• --output - Output directory (default: current directory)

Examples:

# In project directory - ORM is auto-detected
cd myproject
fastscaff models --db-url "mysql://root:password@localhost:3306/mydb"

# Generate models for specific tables
fastscaff models --db-url "mysql://..." --tables user,order,product

# Explicitly specify ORM
fastscaff models --db-url "mysql://..." --orm tortoise

Generated models include:

• Field type mapping
• Primary keys and auto-increment
• Indexes
• Foreign key relationships
• Table and column comments

Project Structure

myproject/
├── app/
│   ├── main.py              # Application entry point
│   ├── core/
│   │   ├── config.py        # Settings (env-based)
│   │   ├── database.py      # Database connection
│   │   ├── redis.py         # Redis client
│   │   ├── security.py      # Password hashing, JWT
│   │   ├── logger.py        # Structured logging
│   │   └── lifespan.py      # Startup/shutdown events
│   ├── api/v1/
│   │   ├── router.py        # API router
│   │   └── endpoints/       # Route handlers
│   ├── models/              # ORM models
│   ├── schemas/             # Pydantic schemas
│   ├── repositories/        # Data access layer
│   ├── services/            # Business logic layer
│   ├── middleware/          # Request/response middleware
│   ├── exceptions/          # Custom exceptions
│   └── utils/               # Utility functions
├── tests/
├── .env.example
├── Dockerfile
├── docker-compose.yml
├── Makefile
└── requirements.txt

Architecture

The generated project follows a layered architecture:

Services are accessed via a singleton registry pattern:

from app.services import registry

user = await registry.user_service.get_user_by_id(user_id)

Built-in Features

Middleware

• CORS handling
• Request logging with trace ID
• JWT authentication
• Security headers
• Request signing verification

Utilities

• Snowflake ID generator
• Rate limiter (Redis-based)
• Cache decorator
• Password hashing

Database

• SQLite by default (zero configuration)
• MySQL/PostgreSQL ready (just update DATABASE_URL)
• Async database operations
• Request-scoped sessions (SQLAlchemy)

Running the Project

cd myproject
pip install -r requirements.txt
make dev

The project runs immediately with SQLite - no database setup required.

Available make commands:

make dev          # Start development server
make test         # Run tests
make lint         # Run linter
make format       # Format code
make docker-up    # Start all services (Docker)
make docker-down  # Stop all services

If Celery is enabled:

make celery-worker  # Start Celery worker
make celery-beat    # Start Celery beat scheduler

Configuration

Configuration is managed via environment variables. Copy .env.example to .env:

# Application
ENV=dev
DEBUG=true
PORT=8000

# Database
DATABASE_URL=sqlite+aiosqlite:///./app.db
# DATABASE_URL=mysql+aiomysql://user:pass@localhost:3306/mydb

# Redis
REDIS_URL=redis://localhost:6379/0

# JWT
JWT_SECRET_KEY=your-secret-key
JWT_ACCESS_TOKEN_EXPIRE_MINUTES=30

# Celery (if enabled)
CELERY_BROKER_URL=redis://localhost:6379/1
CELERY_RESULT_BACKEND=redis://localhost:6379/1

Development

# Clone the repository
git clone https://github.com/lee-hangzhou/fastscaff.git
cd fastscaff

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

# Run tests
pytest

# Code formatting
ruff check --fix .
ruff format .

License

MIT