Disallow isolate Unicode characters in comments and strings

Part of #10254.

### Description
As shown in [Security advisory for rustc (CVE-2021-42574)](https://blog.rust-lang.org/2021/11/01/cve-2021-42574.html), the isolate Unicode characters (LRI, RLI, FSI, PDI) can be used to take bit of text out of context and show it at the beginning or end of the line. We should disallow them to prevent this.

We have already disallowed unbalanced BiDi embedding/override characters in #10326. This is unfortunately not effective against the trick described here.

See [How to use Unicode controls for bidi text](https://www.w3.org/International/questions/qa-bidi-unicode-controls) for details on how these characters work.

Start by reviewing the related issues #10254 and #10326, then read the linked Rust security advisory and W3C bidi guidance. Done means isolate Unicode characters LRI, RLI, FSI, and PDI are disallowed in comments and strings without relying on the earlier unbalanced-character check.

Written by the indexing model from the issue text.