Suggest specific exercises that deal with multi-language text handling.

[Insti]

In #428 the majority opinion was that ordinary problems should not be complicated by adding non-ascii characters in the test cases, but everyone believed that there should be exercises that dealt with them.

Here is your chance to suggest a new exercise, or aspect of mutli-language text handling that should be made into an exercise.

[petertseng]

Oh, right, I had an old list lying around. Digging it up, but I admit it probably needs work before we get an exercise out of it.

I think the following points are all variations on a theme.

• In general, the difference between the number of characters in a string versus the number of bytes in a string
  • In Go, iterating over a string yields each codepoint (termed in Go as a rune), but indexing into a string indexes by byte, and taking its length gets its byte length. Consider creating new Unicode exercise go#200 and ledger: handling unicode go#195
  • In Rust, you can iterate over the chars() of a string, whereas again the len() is the byte length. Add test case to Hamming that shows the difference between len() and chars().count() rust#128
• If you are rearranging the characters of a string, don't accidentally separate multiple bytes making up a codepoint. This was originally in the Anagram exercise: If you want to sort then check equality, you can't sort by bytes. Potential new test in anagram: Same bytes different characters. #318
• A character can be represented by codepoints in various ways (there is the possibility of combining characters). Also came up in ledger: handling unicode go#195
• Dealing with half-width versus full-width characters. Tangentially related to ledger: handling unicode go#195, since knowing width was necessary to make the ledger's columns line up.

[petertseng]

finally there is a suggestion at #1507 .