suboptimal suggestion for match of checked math

[hkBst]

Description

Given (playground):

fn f1() {
    match 1usize.checked_sub(1) {
        Some(v) => v,
        None => 0,
    };
}

clippy suggests 1usize.checked_sub(1).unwrap_or_default();, but I think it should suggest 1usize.saturating_sub(1); instead for both of those.

Version

Additional Labels

No response

[KisaragiEffective]

side note for searching: 1usize.checked_sub(1).unwrap_or_default(); is suggested by manual_unwrap_or_default

[ada4a]

I think it's fine for manual_unwrap_or_default to suggest what it suggests now; the improvement would instead come from having a separate lint addition to manual_saturating_arithmetic, which suggests replacing x.checked_sub(y).unwrap_or_default() with x.saturating_sub(y) -- with that, cargo clippy --fix would apply one fix after the other, resulting in the desired expression.

(EDIT: apparently there's already a lint for similar patterns, manual_saturating_arithmetic, but it doesn't recognize .unwrap_or_default() as being the same as .unwrap_or(0))

This is the way Clippy lints usually work -- each one concentrates on its specific pattern, and gives a reasonable suggestion within its context, and the suggested version might then unlock further improvements.

@rustbot claim