unicode_chars.rs should be updated to match the latest confusables document

[est31]

src/libsyntax/parse/lexer/unicode_chars.rs contains a list of confusable characters (see #25957). Its comments mention that it bases on a document. Of this document, a newer revision has been published (link). It contains many more confusables, e.g. ᐸ (canadian syllabics Pa). We should update unicode_chars.rs to include the newly listed confusables as well.

Bug report inspired by this reddit post.

[est31]

As there has been #33128, the person fixing this bug shouldn't just apply a script and regenerate from scratch, but go through the new confusables.txt and add the new characters manually.

[MaloJaffre]

I will try to work on this, because I have a lot of patience 😄