Python Coding Standards
Core Principles
- •Simplicity: Simple, understandable code
- •Readability: Readability over cleverness
- •Maintainability: Code that's easy to maintain
- •Testability: Code that's easy to test
- •DRY: Don't Repeat Yourself - but don't overdo it
General Rules
- •Early Returns: Use early returns to avoid nesting
- •Descriptive Names: Meaningful names for variables and functions
- •Minimal Changes: Only change relevant code parts
- •No Over-Engineering: No unnecessary complexity
- •Minimal Comments: Code should be self-explanatory. No redundant comments!
Naming Conventions
| Element | Convention | Example |
|---|---|---|
| Variables/Functions | snake_case | get_user_by_id, is_active |
| Classes | PascalCase | UserService, ApiClient |
| Constants | UPPER_SNAKE_CASE | MAX_RETRY_COUNT |
| Private | Prefix with _ | _internal_method |
| Files/Modules | snake_case | user_service.py |
Project Structure
code
myproject/ ├── src/ │ ├── __init__.py │ ├── main.py # Entry point │ ├── config.py # Settings, env vars │ ├── models.py # Domain models (dataclasses/Pydantic) │ ├── schemas.py # Request/response DTOs │ ├── services/ │ │ ├── __init__.py │ │ └── user_service.py # Business logic │ └── repositories/ │ ├── __init__.py │ └── user_repo.py # Data access ├── tests/ │ ├── __init__.py │ ├── test_services.py │ └── test_repositories.py ├── pyproject.toml └── README.md
Code Style (PEP 8 + PEP 484)
python
from dataclasses import dataclass
@dataclass
class User:
id: str
name: str
email: str
age: int | None = None # Python 3.10+ union syntax
def get_user_by_id(user_id: str) -> User | None:
if not user_id:
raise ValueError("user_id cannot be empty")
# implementation...
Best Practices
python
# Type hints everywhere
def process_items(items: list[str]) -> dict[str, int]:
return {item: len(item) for item in items}
# Pydantic v2 for validation
from pydantic import BaseModel, Field, field_validator, EmailStr
class UserCreate(BaseModel):
name: str = Field(..., min_length=2, max_length=50)
email: EmailStr
age: int | None = Field(None, ge=0, le=150)
@field_validator('name')
@classmethod
def name_must_be_alphanumeric(cls, v: str) -> str:
if not v.replace(' ', '').isalnum():
raise ValueError('Name must be alphanumeric')
return v.strip()
# Context managers
with open('file.txt', 'r') as f:
content = f.read()
# Prefer pathlib over os.path
from pathlib import Path
config_path = Path(__file__).parent / 'config.yaml'
Async/Await
python
# Async function with proper typing
async def fetch_user(user_id: str) -> User | None:
async with httpx.AsyncClient() as client:
response = await client.get(f"/users/{user_id}")
return User(**response.json()) if response.status_code == 200 else None
# Don't block async functions
async def process_data():
# BAD - blocks the event loop
time.sleep(1)
# GOOD - async sleep
await asyncio.sleep(1)
# Gather for concurrent operations
async def fetch_all_users(user_ids: list[str]) -> list[User]:
tasks = [fetch_user(uid) for uid in user_ids]
return await asyncio.gather(*tasks)
Exception Handling
python
# Custom exceptions for domain errors
class UserNotFoundError(Exception):
def __init__(self, user_id: str):
self.user_id = user_id
super().__init__(f"User not found: {user_id}")
# Raise vs Return None
def get_user_strict(user_id: str) -> User:
"""Raises if not found - use when user MUST exist."""
user = repository.get(user_id)
if not user:
raise UserNotFoundError(user_id)
return user
def get_user_optional(user_id: str) -> User | None:
"""Returns None if not found - use when absence is expected."""
return repository.get(user_id)
Comments - Less is More
python
# BAD - redundant comment # Get the user from database user = repository.get_user(user_id) # GOOD - self-explanatory code, no comment needed user = repository.get_user(user_id) # GOOD - comment explains WHY (not obvious) # Rate limit: Azure API allows max 1000 requests/min await rate_limiter.acquire()
Recommended Tooling
| Tool | Purpose |
|---|---|
uv | Package manager (faster than pip/poetry) |
ruff | Linting (replaces flake8, isort, black) |
mypy or pyright | Type checking |
pytest | Testing with pytest-cov, pytest-asyncio |
Production Best Practices
- •Type hints everywhere - Parameters, return types, variables where helpful
- •Pydantic for validation - Don't validate manually, use Pydantic models
- •Async for I/O - Use async/await for network, database, file operations
- •No blocking in async - Never use
time.sleep()or sync I/O in async functions - •Structured logging - Use
loggingmodule with JSON format, not print() - •Environment variables - Use
pydantic-settingsfor config, never hardcode secrets - •Dependency injection - Pass dependencies explicitly, makes testing easier
- •Custom exceptions - Domain-specific errors, not generic Exception
- •Connection pooling - Reuse database/HTTP connections, don't create per request
- •Graceful shutdown - Handle SIGTERM, close connections properly