Type Hint

#python #programming #doc #typing #datamodel

• Types are defined by supported operations
• It is a documentation which can be checked by type checkers like MyPy
• There is no impact on the execution of the code
• How is: var_name: some_type (PEP 484)
• It is not possible to make hints ever on the code, sometimes is quite impossible
• PEP 484
• Duck Typing vs Nominal Typing
• Some options that can be important for mypy tools
  • disallow-untyped-defs: not allow any function without type hint
  • disallow-incomplete-defs: not allow any function with incomplete type hint
• mypy can have global configuration or have mypy.ini for each module
• Give no space between : and variable and one space between the : and type
• = space on both sides
• You can use default value None with two syntax and Optional
  • def show_count(count: int, singular: str, plural: Optional[str] = None)
  • def show_count(count: int, singular: str, plural: str | None = None)
• Any is a magic type which allow any operations and subclasses, therefore accept anything
• Simply type annotation: str, float, int, bytes or classes
• Union[A, B, Union[C, D, E]] = Union[A, B, C, D, E]
• Generic Collections example:
• NoReturn usually used for function that raises exception

def tokenize(text: str) -> list[str]:
    return text.upper().split()

• Code above return a list of string
• Typing Tuple
• Typing Maps
• Sequence or Iterable for parameters and list for return type
• Typing Callable
• Parameterized Generics and TypeVar
• Annotating Positional Only and Variadic Parameters

from collections.abc import Iterable
FromTo = tuple[str, str]
def zip_replace(text: str, changes: Iterable[FromTo]) -> str:
    for from_, to in changes:
        text = text.replace(from_, to)
    return text

References

• Ramalho, 2022, p172-179
• Ramalho, 2022, p253-260
• Ramalho, 2022, p260-265
• Ramalho, 2022, p294