python（Python 开发）

Python Development Expert

Expert-level Python development skill with senior architect capabilities, automated code quality tools, and comprehensive best practices.

Core Capabilities

1. Code Quality Automation

Use bundled scripts to enforce code quality:

Run all quality checks:

uv run python scripts/check_quality.py [target_directory]

This runs:

• Ruff lint: Fast linting for code quality issues
• Ruff format: Code formatting checks
• MyPy: Static type checking
• Pytest: Test execution with coverage
• Bandit: Security vulnerability scanning

Auto-fix common issues:

uv run python scripts/autofix.py [target_directory]

Auto-fixes:

• Code formatting (line length, indentation, quotes)
• Import sorting
• Common linting issues (unused imports, etc.)

2. Project Initialization

Create new Python projects with best practices:

uv run python scripts/init_project.py <project-name> [path]

Creates:

• Modern project structure (src/ layout)
• Configured pyproject.toml with Ruff, MyPy, Pytest
• Git ignore file
• Test structure
• README template

3. Code Quality Standards

See code_quality.md for detailed standards on:

• Linting and formatting configuration
• Type hints and annotations
• Modern Python patterns (match statements, walrus operator, etc.)
• Error handling best practices
• Documentation with comprehensive docstrings
• Testing with pytest
• Dependencies management with uv

4. Architecture Patterns

See architecture_patterns.md for:

• Project organization (feature-based structure)
• Dependency injection with protocols
• Design patterns (Factory, Strategy, Repository, Unit of Work)
• API design with FastAPI
• Async patterns
• Error handling architecture
• Performance optimization

Development Workflow

When Writing New Code

1. Use type hints everywhere:

def process_data(
    items: Sequence[dict[str, Any]],
    max_count: int = 100,
) -> list[ProcessedItem]:
    ...

1. Follow modern Python idioms:

# Use match statements (3.10+)
match status:
    case 200:
        return response.json()
    case 404:
        raise NotFoundError()
    case _:
        raise APIError(status)

1. Add comprehensive docstrings:

def calculate_metrics(data: pd.DataFrame) -> dict[str, float]:
    """Calculate statistical metrics from data.

    Args:
        data: Input DataFrame with numeric columns

    Returns:
        Dictionary mapping metric names to values

    Raises:
        ValueError: If data is empty
    """

Before Committing

1. Run auto-fix: uv run python scripts/autofix.py
2. Run quality checks: uv run python scripts/check_quality.py
3. Ensure all checks pass before committing

Code Review Checklist

• ✅ All functions have type hints
• ✅ Docstrings follow Google style
• ✅ No lint warnings from Ruff
• ✅ MyPy type checking passes
• ✅ No security issues from Bandit
• ✅ Test coverage > 80%
• ✅ Code follows architecture patterns

Tool Configuration

All tools are configured via pyproject.toml:

Ruff: Line length 100, Python 3.12+, comprehensive rule set MyPy: Strict mode with no untyped definitions Pytest: Coverage reporting with missing lines Coverage: Excludes test files and common patterns

Common Patterns

Dependency Injection

from typing import Protocol

class Repository(Protocol):
    def save(self, item: Item) -> None: ...

class Service:
    def __init__(self, repo: Repository) -> None:
        self.repo = repo

Configuration with Pydantic

from pydantic_settings import BaseSettings

class Settings(BaseSettings):
    database_url: str
    api_key: str

    model_config = {"env_file": ".env"}

Error Handling

class DomainError(Exception):
    """Base for all domain errors."""
    pass

class ValidationError(DomainError):
    """Invalid input data."""
    pass

Async Operations

async def fetch_all(urls: list[str]) -> list[Response]:
    async with aiohttp.ClientSession() as session:
        tasks = [fetch_one(session, url) for url in urls]
        return await asyncio.gather(*tasks)

Quick Reference

Install dev dependencies:

uv add --dev ruff mypy pytest pytest-cov bandit

Run single tool:

uv run ruff check .
uv run mypy .
uv run pytest
uv run bandit -r .

Format code:

uv run ruff format .

Type check:

uv run mypy --strict .

Best Practices Summary

1. Always use type hints for function signatures and class attributes
2. Run quality checks before committing code
3. Follow modern Python patterns (match, protocols, dataclasses)
4. Use dependency injection for testability
5. Write comprehensive docstrings with examples
6. Organize by feature not by layer
7. Prefer composition over inheritance
8. Use async for I/O-bound operations
9. Cache expensive computations
10. Test with pytest and maintain >80% coverage