Stream: t-compiler/wg-prioritization/alerts

Topic: #82583 3+ byte wide UTF-8 characters in proc-macro token st…

triagebot (Feb 27 2021 at 15:54, on Zulip):

@WG-prioritization/alerts issue #82583 has been requested for prioritization.


apiraino (Feb 27 2021 at 16:01, on Zulip):

reminds me of #81208 but error is different. I'm not understand if it's because a completely different part of the code of the #[rocket::get("/♥")] macro shields the true error.

Hameer Abbasi (Feb 27 2021 at 16:08, on Zulip):

Goes back to at least 2018-01-01, according to cargo-bisect-rustc, I commented on the issue.

Hameer Abbasi (Feb 27 2021 at 16:09, on Zulip):

Also, code that should work but doesn't, not a regression AFAICT, but a very obscure bug.

Hameer Abbasi (Feb 27 2021 at 16:10, on Zulip):

I'd say P-medium is suitable. I'd be okay with P-high too.

apiraino (Feb 27 2021 at 16:34, on Zulip):

I've tried also compiling with cargo +nightly build and I see the error from #81208
Does your bisection catches the same error?

(this said, I agree on a P-medium)

Hameer Abbasi (Feb 27 2021 at 16:40, on Zulip):

Yes, it's a different error than the one in the title. What's more, the example containing #[rocket::get("/мир")] produces the same error on rocket 0.4.7, which is the latest at the time of writing.

apiraino (Feb 27 2021 at 16:45, on Zulip):

Interesting this comment :-)

Hameer Abbasi (Feb 27 2021 at 16:50, on Zulip):

In that case, the bug is invalid. The proc macro is written weirdly.

Hameer Abbasi (Mar 01 2021 at 10:20, on Zulip):

Ah, I managed to reproduce it, see this comment. But the bisection results don't change.

Last update: Apr 10 2021 at 21:30UTC