GH:624 - added mode keyword to DataFrame.to_json

[ramvikrams]

• ClosesVersion 2.0 Compatibility Tracker #624
• Tests added: Please use assert_type() to assert the type of any return value

[Dr-Irv]

This is a bit tricky. You have to put the 2 overloads for orient = Literal["records"] first, but without the "= ..." . The first two overloads would then have lines: Literal[True] and mode: Literal["a"] . This may need an * in the right places (see DataFrame.to_dict() for an example). Then you add another pair of overloads with orient: Literal["records"] and lines: Literal[False] = ..., mode: Literal["w"] = .... Then you have the 2 overloads that are there where mode: Literal["w"] = ... is allowed.

You also have to fix the overloads in series.pyi.

For the tests, you should test the valid and invalid combinations, using TYPE_CHECKING_INVALID_USAGE for the invalid ones, with appropriate ignore statements.

[ramvikrams]

Then you add another pair of overloads with orient: Literal["records"] and lines: Literal[False] = ..., mode: Literal["w"] = ....

Were you trying to write mode: Literal["a"] = ... , but it is not valid i think and if it is mode: Literal["w"] then in the last two overloads for mode: Literal["w"] we would have to do lines:[True] in the last two overloads and check lines separately, otherwise the overloads will overlap. As we have added records back to JsonFrameOrient as you had suggested in the review.

[Dr-Irv]

Then you add another pair of overloads with orient: Literal["records"] and lines: Literal[False] = ..., mode: Literal["w"] = ....

Were you trying to write mode: Literal["a"] = ... , but it is not valid i think and if it is mode: Literal["w"] then in the last two overloads for mode: Literal["w"] we would have to do lines:[True] in the last two overloads and check lines separately, otherwise the overloads will overlap. As we have added records back to JsonFrameOrient as you had suggested in the review.

It may be that the second set of overloads I suggested are not needed because of the overlap you mention.

[ramvikrams]

I have done the needfull but the only problem is with the warnings in the test, could you please help with that

[Dr-Irv]

2 small changes, and then it should be good to go

[Dr-Irv]

2 small changes, and then it should be good to go

[Dr-Irv]

Thanks @ramvikrams