Implement P3223R2 Making istream::ignore() Less Surprising

[frederick-vs-ja]

As a Defect Report against old standard modes. Fixes #5618.

As a result of DR-backing P3223R2, the new overload is templated, and thus this PR happens to "constrain the new overload to prevent ambiguities" and keeps in.ignore(100, -1L) well-formed.

[Zingam]

@frederick-vs-ja (This is not a review. Just a couple of questions.) Was the paper voted in as DR? I guess not. My reading of the Standard is that resolving the ambiguity is not required is that so?
I made an attempt to implement the paper in libc++ but Windows is giving me troubles llvm/llvm-project#147007

[frederick-vs-ja]

Was the paper voted in as DR? I guess not.

I guess not, either. But according to the paper issue I think it should be treated as a DR.

My reading of the Standard is that resolving the ambiguity is not required is that so?

Yes. In this PR the ambiguity is resolved "by accident", which is a consequence of backporting. I don't think the ambiguity resolution is disallowed, either.

[StephanTLavavej]

I'm mirroring this to the MSVC-internal repo - please notify me if any further changes are pushed.

[StephanTLavavej]

🐱 😸 🐈