needless_option_take: add autofix

[samueltardieu]

changelog: [needless_option_take]: add autofix

[rustbot]

r? @llogiq

rustbot has assigned @llogiq.
They will have a look at your PR within the next two weeks and either review your PR or reassign to another reviewer.

Use r? to explicitly pick a reviewer

[llogiq]

The implementation looks good to me, but I'm not sure if I agree with the idea of an autofix here. The problem is that the .take() shows a probable mismatch between the intent of the programmer and the actions of the code, so it is very likely that the code is wrong in other ways that we shouldn't silently "fix".

[samueltardieu]

Your remark opens up to questions for me:

• If you think this signals a programmer's mistake, shouldn't this lint be into suspicious rather than in complexity?
• Should we ever apply autofixes for suspicious lints, or should we require another flag?

In the present case, I'm not sure that someone calling .take() on an option is likely to be a mistake. It might be a remnant of a refactoring, or the programmer not realizing that they own the option (compared to having a &mut) and can use it directly.

Note that if they had used swap, I would have more inclined to see that as a probable error.

[llogiq]

To your first question, you're probably right and we may as well autofix it. To the second question, I would still decide on a lint-by-lint basis, but opt into autofixes rather than opting out (by which I mean there should be good arguments for a fix, otherwise I'd default to not fix).

Thanks for doing this improvement.