[manual_range_patterns]: lint negative values

[y21]

Fixes #11091.

Now also lints negative values in patterns (-1 | -2 | -3)

changelog: [manual_range_patterns]: lint negative values

[rustbot]

r? @giraffate

(rustbot has picked a reviewer for you, use r? to override)

[y21]

Okay, I think this requires some pretty involved changes to handle "large" numbers correctly. For example:

    matches!(
        1u128,
        340282366920938463463374607431768211455
            | 340282366920938463463374607431768211454
            | 340282366920938463463374607431768211453
    );

This currently emits the warning "help: try: -3..=-1", because the integer literal wraps around. This probably needs lots of casting and checks to make all the comparisons etc work. Though maybe there's a nicer way to do this. Will have to think about it.
(or I guess we could also just ignore numbers >i128::MAX for now 🤔 )

[Centri3]

Okay, I think this requires some pretty involved changes to handle "large" numbers correctly. For example:

    matches!(
        1u128,
        340282366920938463463374607431768211455
            | 340282366920938463463374607431768211454
            | 340282366920938463463374607431768211453
    );

This currently emits the warning "help: try: -3..=-1", because the integer literal wraps around. This probably needs lots of casting and checks to make all the comparisons etc work. Though maybe there's a nicer way to do this. Will have to think about it. (or I guess we could also just ignore numbers >i128::MAX for now 🤔 )

The type of the integer is known in LitKind::Int, you can cast it using that (like match int_ty { Isize => val as isize, _ => /* etc */ })

[y21]

yep but what I mean is that, unless I'm overthinking this way too much, the logic will get very complicated very quickly here now that we need these casts and checks basically for every use of min/max (to get additions/less than/greater than/range checks etc. right) and it probably requires very large changes

[Centri3]

That's a good point.

[giraffate]

(or I guess we could also just ignore numbers >i128::MAX for now 🤔 )

IMO it would be better to ignore these numbers or just note it in Known problems section.

[giraffate]

@bors r+

Thanks!

[bors]

📌 Commit c927912 has been approved by giraffate

It is now in the queue for this repository.

[bors]

⌛ Testing commit c927912 with merge 9058b04...

[bors]

☀️ Test successful - checks-action_dev_test, checks-action_remark_test, checks-action_test
Approved by: giraffate
Pushing 9058b04 to master...