Type Annotations in Python

Python's dynamic typing makes it very flexible, but it can also make large code bases hard to get into and it limits the effectiveness of static code checkers. Adding type annotations can help with these issues.

For people not yet using type annotations, I'll give a short introduction explaining why it can be useful to add type annotations to your Python programs and go over the basic syntax.

Most of the time will be dedicated to explaining how you figure out the right annotation: first figuring out the expected type and then how to express that in the type system, using a series of examples.

Maarten ter Huurne

Maarten is a Python veteran (20 years and counting) who has added type annotations to thousands of lines of code. He has contributed code to a wide range of open source projects, including Twisted, pydoctor, openMSX and the Linux kernel. When he's not programming, he plays video games and board games.

Links

• Slides
• Specifications:
  • PEP 484 – Type Hints
  • PEP 544 – Protocols (static duck typing)
  • PEP 563 – Postponed Evaluation of Annotations
  • PEP 604 – Allow writing union types as X | Y
  • Full list
• Standard library:
  • typing
  • collections.abc
  • dataclasses
• dataclass-binder – Bind TOML (configuration) files to dataclasses
• Tooling:
  • mypy
  • MonkeyType
  • pytype