Add MapWindows compile fail check for N=0

[Nydauron]

Related to issue #87155. Adds a constant assertion to MapWindows to check if generic N, representing the size of the window array, is not zero. Before, a zero-length window array would cause MapWindows to panic at runtime. With the const assert, if N=0, the code will not compile.

A map window test that dealt with a zero-width array is moved to the Iterator::map_windows doc-tests in order to check for compilation failure. Before this change, the test was meant to panic.

[rustbot]

Thanks for the pull request, and welcome! The Rust team is excited to review your changes, and you should hear from @Mark-Simulacrum (or someone else) some time within the next two weeks.

Please see the contribution instructions for more information. Namely, in order to ensure the minimum review times lag, PR authors and assigned reviewers should ensure that the review label (S-waiting-on-review and S-waiting-on-author) stays updated, invoking these commands when appropriate:

• @rustbot author: the review is finished, PR author should check the comments and take action accordingly
• @rustbot review: the author is ready for a review, this PR will be queued again in the reviewer's queue

[the8472]

We do want something like this in various places such as slice::array_chunks. The issue is that we currently lack language features to express this properly in the type system or avoiding post-monomorphization errors when users call it from code that is generic over N.

We need some combination of const if, const generic expressions or pattern types to make this work properly.

[Mark-Simulacrum]

r? @the8472 since it sounds like you have more context here. Failing that this is probably T-libs-api (though it's unstable so maybe not strictly required).

[the8472]

Context, previous PRs that tried to do something similar and were rejected because we currently lack the language features to make this ergonomic for users:

• Make <[T]>::array_* methods fail to compile on 0 len arrays #99471
• Make <[T]>::array_* methods fail to compile on 0 len arrays #99471 (comment)

Stabilization of various features is also blocked or at least less likely until we have a better solution:

• Tracking Issue for slice::array_chunks #74985
• Tracking Issue for Iterator::array_chunks #100450
• ...

So for now I'll close this until some progress gets made on the questions in those issues.