This is not a pydantic question, but to explain why am I asking: pydantic.TypeAdapter() accepts (among many others) all the following type definitions as its argument and can create a working validator for them:

int
int|str
list
list[str|int]
typing.Union[int,str]
typing.Literal[10,20,30]

Example:

>>> validator = pydantic.TypeAdapter(list[str|int]).validate_python

>>> validator([10,20,"stop"])
[10, 20, 'stop']

>>> validator([10,20,None])
(traceback deleted)
pydantic_core._pydantic_core.ValidationError: ...

I want to make a test if an argument is a such type defition. How do I write such test?

• I started with isinstance(arg, type) for simple types like int or list
• then I added isinstance(arg, types.GenericAlias for list[str] etc.
• then I realized it does not recognize int|str (which itself behaves differently than typing.Union[int,str]). Also the Literal[] is not recognized … I’m probably on a wrong track.