Address more non-ascii characters

[geoffw0]

Address more non ascii characters in the UncheckedLeapYearAfterModification PR. I ran the check locally this time, since it became apparent the CI check was only reporting the first, not all violations.

It looks like VSCode has removed some unnecessary spacing in the source file alongside the edits.

@bdrodes apologies for this hassle, the check has been quite noisy and isn't flagging anything I'd consider to actually be a problem in 2026. If I had the bandwidth I'd start a discussion about narrowing or disabling the check itself.

[geoffw0]

I'm, er, not reading the above policy. Accept the PR if you can, otherwise you can implement it by search-and-replace of ± with +- in the test.cpp.

[bdrodes]

Sorry @geoffw0 thanks for doing this. This is something you all may want to reconsider given this issue will only become an increasing concern with the fact AI is assisting in writing test code and comments, and that's why these cases are even part of the PR. The AI in vscode makes an inline suggestion that looks right, and adds these characters. It's worse when you fully accept full AI generated test code, or copy any examples from another copilot.

[geoffw0]

I'll start an internal conversation about it.