Add a generic check against "dangerous" Unicode codepoints

It would be nice if spotless could check for (and remove or error out) "dangerous" unicode codepoints.

See today's Rust security advisory: https://blog.rust-lang.org/2021/11/01/cve-2021-42574.html - According to this, the initial list of forbidden codepoints should contain: U+202A, U+202B, U+202C, U+202D, U+202E, U+2066, U+2067, U+2068, U+2069

Additionally, it would be nice to have a similar check against homoglyph attacks, but that is probably a bit more tricky to solve...

Start with issue #976 and the linked Rust security advisory to confirm the nine listed codepoints and the intended remove-or-error behavior. Review Spotless's existing check entry points to determine where a generic Unicode check belongs; done means detecting every listed codepoint and applying the chosen behavior.

Written by the indexing model from the issue text.