type `kwargs` in `DataFrame.query` according to `DataFrame.eval`

[MarcoGorelli]

Opening this so I can point to it as a todo. See #1168 for context

[loicdiridollou]

If we look at the docs (https://pandas.pydata.org/docs/reference/api/pandas.eval.html#pandas.eval), it seems like the type would be:
dict[str, str | None | int | list | dict | object | bool]
Is that what you had in mind for that issue? I am a little concerned about the object possibility here since it would pretty much encompass most of the elements so wondering what kind of value we are trying to extract here. There could be a better way to do this to improve the typehinting, let me know.

[MarcoGorelli]

rather than dict, i was thinking that the parameters from DataFrame.query could be added to the DataFrame.eval signature

[loicdiridollou]

Okay that was my other idea, I am worried that:

• It would conflict with the docs so that may be more a pandas lib question
• It would conflict if I actually have a dictionary and I pass it to df.query(..., **dict_of_kwargs) so if we add all the args of eval there we may end up in a weird position where we can't recognize if someone pass a dict of kwargs as it should be from docs

Open to ideas