bpo-44796: Unify TypeVar and ParamSpec substitution#31143
Uh oh!
There was an error while loading. Please reload this page.
Conversation
Add methods __typing_subst__() in TypeVar and ParamSpec.
JelleZijlstra left a comment
JelleZijlstra left a comment There was a problem hiding this comment.
Thanks, this does make things simpler, but we need to be careful that we don't break too many external consumers of typing code. Also, there may be interactions with #31021, which adds another TypeVar variant.
Here's a few comments, but I may need to look at this more to understand it better.
| f"ParamSpec, or Concatenate. Got {t_args}") | ||
| returnsuper().__new__(cls, origin, args) | ||
| @property |
There was a problem hiding this comment.
I think we have to keep __parameters__, it's de facto a public API by now.
There was a problem hiding this comment.
It is inherited from GenericAlias now.
Lib/typing.py Outdated
| ifisinstance(arg, _SpecialForm) orargin (Generic, Protocol): | ||
| raiseTypeError(f"Plain {arg} is not valid as type argument") | ||
| ifisinstance(arg, (type, TypeVar, ForwardRef, types.UnionType, ParamSpec)): | ||
| ifisinstance(arg, (type, TypeVar, ForwardRef, types.UnionType)): |
There was a problem hiding this comment.
Will this affect putting ParamSpec inside Annotated? Looks like we don't have tests for that yet.
There was a problem hiding this comment.
I looked into this.
This check is needed in order to allow instances of ParamSpec (i.e. P in P = ParamSpec("P")) to pass typing._type_check because callable(P) is False. This can show up for instance like Callable[Annotated[P, ""], T]. This line's change would regress the behavior to a runtime error.
n.b. this is moot if #31151 gets merged since this line is removed entirely
There was a problem hiding this comment.
Callable[Annotated[P, ""], T] does not conform the PEP 612 specification.
callable ::= Callable "[" parameters_expression, type_expression "]" parameters_expression ::= | "..." | "[" [ type_expression ("," type_expression)* ] "]" | parameter_specification_variable | concatenate "[" type_expression ("," type_expression)* "," parameter_specification_variable "]" There was a problem hiding this comment.
I'm not convinced it was the intent of the spec here to disallow Annotated, but you can also reach this code here:
fromtypingimportCallable, TypeVar, ParamSpec, get_type_hintsT=TypeVar("T") P=ParamSpec("P") defadd_logging(f: Callable["P", T]): passget_type_hints(add_logging)There was a problem hiding this comment.
Well, I re-added ParamSpec. Now some errors will not be caught at runtime. We will return to this in future.
| and_is_unpacked_typevartuple(old_arg)): | ||
| if_is_unpacked_typevartuple(old_arg): | ||
| original_typevartuple=old_arg.__parameters__[0] | ||
| new_arg=new_arg_by_param[original_typevartuple] |
There was a problem hiding this comment.
This code is not covered by tests.
It perhaps could be simplified too, but since it is not covered by tests I left it as is.
serhiy-storchaka requested review from Fidget-Spinner, gvanrossum and rhettinger as code owners
serhiy-storchaka commented Mar 10, 2022
It now includes PEP 646 changes. |
gvanrossum commented Mar 12, 2022
Did anyone review this? |
serhiy-storchaka commented Mar 17, 2022
Sorry, it was not reviewed after the last changes. But I addressed previous review comments. Is there something wrong with this code? I merged this PR after short delay because it allows the C code for TypeVarTuple substitution (#31828) to be of reasonable complexity. |
6 participants
Add methods typing_subst() in TypeVar and ParamSpec.
This is an alternative to #27511.
https://bugs.python.org/issue44796