I don't know yet if we want to implement "non-ASCII identifiers." In other words, it is still under discussion. However, if we decide to implement it, we need to consider its related facilities. I thought of several potentially useful facilities. May need something else
Macro for transcribing identifiers to Latin letters
I don鈥檛 want this to be the compiler鈥檚 or standard library鈥檚 job since it鈥檚 a hugely complex and open-ended problem. Japanese for example is usually transliterated based on pronunciation, but the same character can have a very different pronunciation when used in different words or even different contexts.
Simplified input tools
This seems to be much more in the realm of a text editor or operating system than a compiler.
What's the use case for such a macro anyway? There seems to already be a very strong consensus that non-ASCII identifiers in public APIs meant for the general Rust community (as opposed to a smaller community that's equipped to easily type such identifiers) are a bad idea.
We can probably close this as unlikely to end up in the standard library.
Most helpful comment
What's the use case for such a macro anyway? There seems to already be a very strong consensus that non-ASCII identifiers in public APIs meant for the general Rust community (as opposed to a smaller community that's equipped to easily type such identifiers) are a bad idea.