Return a value from find_format_args instead of using a callback

[Alexendoo]

r? @flip1995

changelog: none

[flip1995]

I haven't reviewed the PR yet. But I don't like the doc comment of Symbol::as_str

in practice it works out ok.

I'll have to look more closely on this PR though, before making a decision.

[Alexendoo]

Assuming I've done it correctly it should be less of a crime than Symbol::as_str, since it's tied to a lifetime that's shorter than the thread local

[Jarcho]

There's nothing that actually stops a &'static LateContext<'_> value from existing. Box::leak is the easy way to do it, but there are others. Symbol::as_ref is unsafe for exactly the same reason as there's nothing stopping &'static Symbol from existing.

Neither of these are really a problem in practice, but there isn't a lifetime you can extend to that isn't technically unsafe.

[Alexendoo]

Yeah that's true, I've updated the comment a bit

[bors]

☔ The latest upstream changes (presumably #11473) made this pull request unmergeable. Please resolve the merge conflicts.

[Alexendoo]

I thought about this again today and realised we could stick the FormatArgs in an Rc, which means it no longer needs any unsafe code since it's not borrowing from the thread local. Updated the PR to use that

[flip1995]

Wow, turns out if you wait long enough, PR reviews resolve themselves. Sorry for taking so long.

Since this PR doesn't add unsafe code now, it was an easy review.

Thanks!

[flip1995]

@bors r+

[bors]

📌 Commit c29de92 has been approved by flip1995

It is now in the queue for this repository.

[bors]

⌛ Testing commit c29de92 with merge b27fc10...

[bors]

☀️ Test successful - checks-action_dev_test, checks-action_remark_test, checks-action_test
Approved by: flip1995
Pushing b27fc10 to master...