majiayu000 Python Mypy Type Checking Skill
This skill automatically activates when writing Python code to ensure proper type annotations and compatibility with Mypy static type checking.
Core Principles
- Type Safety: Catch type errors before runtime
- Gradual Typing: Start with critical paths, expand coverage over time
- Strict Mode: Enable strict checks for new code
- CI Integration: Run Mypy in continuous integration
Type Annotation Patterns
Function Signatures
from typing import Optional
from collections.abc import Sequence
# Good: Complete type hints
def process_items(
items: list[str],
max_count: int | None = None,
debug: bool = False,
) -> dict[str, int]:
"""Process items and return counts."""
result: dict[str, int] = {}
# Implementation
return result
# Good: Generic types with TypeVar
from typing import TypeVar
T = TypeVar('T')
def first(items: Sequence[T]) -> T | None:
"""Get first item from sequence."""
return items[0] if items else None
Class Type Hints
from typing import ClassVar
from dataclasses import dataclass
@dataclass
class User:
"""User model with type hints."""
id: int
name: str
email: str | None = None
active: bool = True
# Class variable
_registry: ClassVar[dict[int, 'User']] = {}
def __post_init__(self) -> None:
"""Register user after initialization."""
self._registry[self.id] = self
Protocol for Structural Typing
from typing import Protocol
class Drawable(Protocol):
"""Protocol for drawable objects."""
def draw(self) -> str:
"""Draw the object."""
...
def render(obj: Drawable) -> None:
"""Render any drawable object."""
print(obj.draw())
# Any class with draw() method satisfies this
class Circle:
def draw(self) -> str:
return "○"
render(Circle()) # OK with Mypy
TypedDict for Structured Dictionaries
from typing import TypedDict, NotRequired
class UserDict(TypedDict):
"""Structured user dictionary."""
id: int
name: str
email: NotRequired[str] # Optional key (Python 3.11+)
def create_user(data: UserDict) -> None:
"""Create user from typed dictionary."""
user_id: int = data["id"] # Type-safe access
# Mypy knows 'email' might not exist
Mypy Configuration Best Practices
Recommended mypy.ini
[mypy]
python_version = 3.11
warn_return_any = True
warn_unused_configs = True
disallow_untyped_defs = True
disallow_any_generics = True
disallow_subclassing_any = True
disallow_untyped_calls = True
disallow_incomplete_defs = True
check_untyped_defs = True
no_implicit_optional = True
warn_redundant_casts = True
warn_unused_ignores = True
warn_no_return = True
warn_unreachable = True
strict_equality = True
show_error_codes = True
show_column_numbers = True
# Start strict, relax per-module if needed
[mypy-tests.*]
disallow_untyped_defs = False
[mypy-migrations.*]
ignore_errors = True
# Third-party without stubs
[mypy-some_library.*]
ignore_missing_imports = True
pyproject.toml Configuration
[tool.mypy]
python_version = "3.11"
warn_return_any = true
warn_unused_configs = true
disallow_untyped_defs = true
show_error_codes = true
[[tool.mypy.overrides]]
module = "tests.*"
disallow_untyped_defs = false
[[tool.mypy.overrides]]
module = "migrations.*"
ignore_errors = true
Common Mypy Checks
Strict Optional Checking
# Bad: Implicit Optional
def find_user(id: int) -> User: # Mypy error if can return None
return users.get(id) # dict.get returns User | None
# Good: Explicit Optional
def find_user(id: int) -> User | None:
return users.get(id)
# Good: Narrow type with assertion
def get_user(id: int) -> User:
user = users.get(id)
assert user is not None, f"User {id} not found"
return user # Mypy knows this is User, not None
Type Narrowing
def process(value: str | int) -> str:
"""Process value based on type."""
if isinstance(value, str):
# Mypy knows value is str here
return value.upper()
else:
# Mypy knows value is int here
return str(value * 2)
Generics
from typing import Generic, TypeVar
T = TypeVar('T')
class Stack(Generic[T]):
"""Type-safe stack."""
def __init__(self) -> None:
self._items: list[T] = []
def push(self, item: T) -> None:
self._items.append(item)
def pop(self) -> T:
return self._items.pop()
# Usage
int_stack: Stack[int] = Stack()
int_stack.push(1) # OK
int_stack.push("x") # Mypy error!
Type Checking Strategies
Gradual Adoption
- Start with New Code: Use strict mode for new modules
- Core Paths First: Type-check critical business logic
- Expand Coverage: Gradually increase
disallow_untyped_defs - Per-Module Configuration: Use mypy overrides for legacy code
[mypy]
# Strict by default
disallow_untyped_defs = True
# Relax for legacy
[mypy-legacy.*]
disallow_untyped_defs = False
check_untyped_defs = True # Still check what we can
Type Ignores (Use Sparingly)
# When third-party library lacks types
import untyped_library # type: ignore[import-untyped]
# When dealing with dynamic code (rare)
def dynamic_call() -> Any:
result = getattr(obj, method_name)() # type: ignore[misc]
return result
Common Patterns
Context Managers
from typing import Generator
from contextlib import contextmanager
@contextmanager
def database_transaction() -> Generator[Connection, None, None]:
"""Type-safe context manager."""
conn = get_connection()
try:
yield conn
conn.commit()
except Exception:
conn.rollback()
raise
finally:
conn.close()
Callable Types
from collections.abc import Callable
def retry(
func: Callable[[int], str],
times: int = 3,
) -> str:
"""Retry a function that takes int and returns str."""
for _ in range(times):
try:
return func(42)
except Exception:
continue
raise RuntimeError("All retries failed")
Overloads for Multiple Signatures
from typing import overload
@overload
def parse(data: str) -> dict[str, str]: ...
@overload
def parse(data: bytes) -> dict[str, bytes]: ...
def parse(data: str | bytes) -> dict[str, str] | dict[str, bytes]:
"""Parse data with type-specific return."""
if isinstance(data, str):
return {"parsed": data}
return {"parsed": data}
Django Integration
Model Type Hints
from django.db import models
from typing import TYPE_CHECKING
if TYPE_CHECKING:
from django.db.models.manager import RelatedManager
class Author(models.Model):
name = models.CharField(max_length=100)
if TYPE_CHECKING:
books: RelatedManager['Book']
class Book(models.Model):
title = models.CharField(max_length=200)
author: models.ForeignKey[Author] = models.ForeignKey(
Author,
related_name='books',
)
Django Mypy Plugin
[mypy]
plugins = mypy_django_plugin.main
[mypy.plugins.django-stubs]
django_settings_module = "myproject.settings"
Running Mypy
Basic Check
# Check all files
mypy .
# Check specific files/directories
mypy src/
# Show error codes
mypy --show-error-codes .
# Generate HTML report
mypy --html-report ./mypy-report .
In CI/CD
# .github/workflows/type-check.yml
- name: Type check with Mypy
run: |
pip install mypy
mypy --strict src/
Anti-Patterns to Avoid
Don't Use Any Unnecessarily
# Bad: Any hides all type errors
def process(data: Any) -> Any:
return data.unknown_method() # No error!
# Good: Use specific types
def process(data: dict[str, int]) -> list[int]:
return list(data.values())
Don't Ignore All Errors
# Bad: Blanket ignore
x = dangerous_call() # type: ignore
# Good: Specific ignore with reason
x = legacy_api_call() # type: ignore[misc] # TODO: Add types to legacy API
Don't Mix str and bytes
# Bad: Mypy will catch this
def process(data: str) -> None:
encoded: bytes = data # Error!
# Good: Explicit conversion
def process(data: str) -> None:
encoded: bytes = data.encode('utf-8')
Coverage Reporting
# Check type coverage
mypy --html-report ./coverage .
# Show coverage stats
mypy --any-exprs-report ./coverage .
Integration with Other Tools
- Pre-commit: Run Mypy before commits
- VS Code: Use Pylance with type checking mode
- Ruff: Complement Mypy with Ruff for runtime checks
Related Skills
| Skill | Purpose |
|---|---|
python-experts:python-style |
Python coding standards |
python-experts:python-code-review |
Code review guidelines |
python-experts:python-testing-expert |
Testing patterns |