Python Pro: Modern Development
This skill defines the gold standard for Python development in the 2024+ era. Move beyond legacy practices (requirements.txt, setup.py, untyped code) to a robust, production-ready workflow.
1. Modern Tooling Stack
| Category | Recommended Tool | Why? |
|---|---|---|
| Package Manager | uv (or Poetry) | uv is 10-100x faster than pip/poetry. Handles python versions + dependencies. |
| Linter / Formatter | ruff | Replaces Black, Isort, Flake8, Pylint. One tool, instant speed. |
| Type Checker | mypy (strict) | Static analysis to catch bugs before runtime. |
| Testing | pytest | The industry standard. See test-expert skill. |
Setup Command (The "Holy Trinity")
bash
# Initialize with uv uv init my-project cd my-project # Add standard dev dependencies uv add --dev ruff mypy pytest pytest-cov
2. Project Structure (The src Layout)
ALWAYS use the src layout. It prevents import errors and ensures you test against the installed package, not local files.
text
my-project/
├── pyproject.toml # Single config file for EVERYTHING
├── uv.lock # Lock file (determinism)
├── README.md
├── src/
│ └── my_package/
│ ├── __init__.py
│ └── main.py
└── tests/
├── __init__.py
└── test_main.py
3. Configuration (pyproject.toml)
Centralize config. Do not use .flake8, pytest.ini, or .coveragerc.
toml
[project]
name = "my-project"
version = "0.1.0"
requires-python = ">=3.10"
dependencies = [
"pydantic>=2.0",
]
[tool.ruff]
line-length = 88
target-version = "py310"
select = ["E", "F", "I", "UP", "B"] # I=Isort, UP=PyUpgrade, B=Bugbear
[tool.mypy]
strict = true
ignore_missing_imports = true
[tool.pytest.ini_options]
testpaths = ["tests"]
pythonpath = ["src"]
addopts = "-ra -q --cov=my_package"
4. Modern Idioms & Best Practices
A. Typing is NOT Optional
Write strict types. It documents your code and catches 40% of bugs.
python
# BAD
def process(data):
return data['items']
# GOOD
from typing import List, Dict
def process(data: Dict[str, List[int]]) -> List[int]:
return data["items"]
B. Use Pydantic for Data Validation
Don't pass raw dictionaries around.
python
from pydantic import BaseModel
class User(BaseModel):
id: int
name: str
email: str
def create_user(user: User) -> None:
print(f"Creating {user.name}")
C. Path Handling (pathlib)
Stop using os.path.
python
# BAD import os path = os.path.join(os.getcwd(), "data", "file.txt") # GOOD from pathlib import Path path = Path.cwd() / "data" / "file.txt" text = path.read_text(encoding="utf-8")
D. AsyncIO
Use async/await for I/O bound tasks.
python
import asyncio
async def fetch_data():
await asyncio.sleep(1)
return "data"
async def main():
# Run concurrently
results = await asyncio.gather(fetch_data(), fetch_data())
Checklist for New Files
- •Imports: Standard lib $\rightarrow$ Third party $\rightarrow$ Local (
ruffhandles this). - •Typing: All function arguments and return values typed.
- •Docstrings: Google or Numpy style for complex logic.
- •No Prints: Use
loggingorstructlog.
code