Avoid unnecessary String allocations in MinifyingSugg arithmetic ops

[shulaoda]

Summary

In clippy_lints/src/loops/manual_memcpy.rs, the four Add/Sub trait impls on MinifyingSugg used the pattern:

match (self.to_string().as_str(), rhs.to_string().as_str()) {
    ("0", _) => rhs.clone(),
    (_, "0") => self.clone(),
    ...
}

Every call therefore allocated two Strings purely so they could be matched against the literal "0". The allocations were discarded immediately afterwards.

This PR:

• Adds an inline is_zero() helper that pattern-matches Sugg::NonParen("0") without allocating. In this lint's data flow Sugg::NonParen is the only variant that can carry the textual value "0" (see sugg::ZERO and the literal path in get_offset), so the check is equivalent to the original to_string() == "0".
• Rewrites the four Add/Sub impls to use is_zero() for the zero-side fast paths.
• Keeps self.to_string() == rhs.to_string() for the a - a = 0 collapse inside Sub, but only as the final branch — it now only runs when both operands are confirmed non-zero, so the common case is fully allocation-free.

changelog: none

[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

Why was this reviewer chosen?

The reviewer was selected based on:

• Owners of files modified in this PR: 7 candidates
• 7 candidates expanded to 7 candidates
• Random selection from Jarcho, llogiq, samueltardieu

[samueltardieu]

r? samueltardieu
@rustbot author

View changes since this review

[samueltardieu]

Nit: this should not be needed, as a mod private function it will be inlined if the compiler thinks this is a good idea.

[rustbot]

Reminder, once the PR becomes ready for a review, use @rustbot ready.

[shulaoda]

@rustbot ready

[samueltardieu]

Please squash the commits together (and "u" is hardly an acceptable commit message anyway, typo?).

[shulaoda]

Please squash the commits together (and "u" is hardly an acceptable commit message anyway, typo?).

Sorry about that. I usually use a simple "u" as shorthand for "update", since in previous contributions these commits were always squashed into a single commit named after the PR title, without keeping the individual commit messages.