Thanks for the feedback Petr,
- if this PEP really only affects typing.py and external
projects/tools, it should say so clearly (so e.g. a parser experts can
skip reading the PEP with clear conscience, even though it "introduces
two new syntaxes")
Did propose a new paragraph in the Abstract section that should make
this clear:
"No Python grammar changes are made by this PEP. Correct usage
of required and potentially-missing keys of TypedDicts is intended to be
enforced only by static type checkers and in particular is not enforced by
Python itself at runtime."
- in Specification, clarify what "It is an error" means -- is it a
Python runtime error, or an error type checkers should raise? Same for
"It is valid".
It is a *logical* error. To be explicit:
(1) It's definitely an error that type checkers MUST raise.
(2) The runtime implementations of Required[] and NotRequired[] MAY also
choose to raise errors at runtime on a best effort basis.
In particular the prohibition on nesting Required[] and NotRequired[]
inside each other is likely to be enforced at runtime by the __getitem__
functions on the type form implementations, because it is easy to do so.
All of the preceding clarifications have been proposed in a PR:
https://github.com/python/peps/pull/2388
Best,
--
David Foster | Seattle, WA, USA
Contributor to Python's type system
_______________________________________________
Python-Dev mailing list -- python-dev@python.org
To unsubscribe send an email to python-dev-le...@python.org
https://mail.python.org/mailman3/lists/python-dev.python.org/
Message archived at
https://mail.python.org/archives/list/python-dev@python.org/message/3TWAKPPEORI2VPEAVNCC4GIOR22TAILQ/
Code of Conduct: http://python.org/psf/codeofconduct/
