string.Reverse() landmine

[jhudsoncedaron]

"𝄞".Reverse() does something broken.

On investigation it's not string.Reverse() but IEnumerable.Reverse() and appears to be converted back to a string in some contexts.

Clearly trying to reverse a string has a fools trap in it.

Oversight?

[Clockwork-Muse]

You can't generally reverse a string for essentially anything outside the basic ASCII range, which includes the treble-clef character in your example. This is true both for the UTF-16 encoding used natively with strings in C#, and the idealized character concepts present in Unicode (see combining characters).

Thus, there is no Reverse method on strings, deliberately.
(Strictly speaking even the per-character operations are somewhat dangerous)

"Reversing" a string is a non-trivial operation, and is use/context specific.

[jhudsoncedaron]

Thus, there is no Reverse method on strings, deliberately.

Well; Intellisense thinks there is. :(

[Clockwork-Muse]

Intellisense is finding all relevant signatures, which in this case is going to be the IEnumerable<char>, as you've noted.

It has no concept of whether the method is actually (as in, conceptually, not from a compile/runtime perspective) valid or does what the caller "wants". That requires human judgement and understanding.

[tannergooding]

Notably this is just a side effect of extension methods (and with latest C#, extension members).

Anyone can add extensions for any type and when you match against an interface, it can expose functionality that may not be 100% sensible for the concrete type and rather only makes sense for the thing being matched against, which in this case is IEnumerable<char> and so Reverse simply enumerates the individual UTF-16 characters in reverse.

IntelliSense explicitly surfaces and allows filtering between concrete members and extension members.