improve error message when a cfg-ed out item is resolved

[scrabsha]

the current note for "cannot find function f in this scope" is:

note: found an item that was configured out
 --> src/main.rs:4:4
  |
3 | #[cfg(foo = "bar")]
  |       ----------- the item is gated behind the `bar` feature
4 | fn f() {}
  |    ^

i think mentioning only bar in the label is misleading. in more complicated setups, this leads to obscure notes such as "the item is gated behind the ptr feature".

this PR improves the situation by naming both the name and value in the label (in the example i linked, target_has_atomic and ptr). while i was at it, i also special-cased #[cfg(false)], #[cfg(any())] and #[cfg(feature = "meow")] (aka Cargo features).

i also tried to use #[derive(SubDiagnostic)] on FoundItemConfigureOut, with no success (the resulting diagnostic was showing multiple help snippets, i couldn't figure out how to merge them). any help would be appreciated.

r? you Esteban, as you added the diagnostic initially - feel free to unassign yourself if necessary :)

r? @estebank

nevermind apparently that's not possible --'

[rustbot]

Failed to set assignee to you: invalid assignee

Note: Only org members with at least the repository "read" role, users with write permissions, or people who have commented on the PR may be assigned.

[rustbot]

Failed to set assignee to estebank~~: invalid assignee

Note: Only org members with at least the repository "read" role, users with write permissions, or people who have commented on the PR may be assigned.

[scrabsha]

@rustbot reroll

[Kivooeo]

nevermind apparently that's not possible

r? estebank

[scrabsha]

@Kivooeo if you don't mind me asking, why did it fail when i opened the PR? i remember r?ing people in the past, i'm a bit confused :/

edit: oah i did not suspect this would trigger rustbot again, sorry for the noise

edit: i think this is because rustbot looks for the first occurrence of r question mark anywhere in the message and tries to use that (i thought it had to be at the beginning of a line).

[rustbot]

Failed to set assignee to ing: invalid assignee

Note: Only org members with at least the repository "read" role, users with write permissions, or people who have commented on the PR may be assigned.

[jdonszelmann]

Might be good to also special case any(), the old way of writing false

[scrabsha]

done, thanks!

[Kivooeo]

@Kivooeo if you don't mind me asking, why did it fail when i opened the PR? i remember r?ing people in the past, i'm a bit confused :/

The issue is that the r-? command assigns the reviewer based on the word that follows it. In your first attempt, you wrote "r-? you Esteban" -- but "you" isn't a valid reviewer. Then you tried "estebank~~", which is also incorrect

To assign a reviewer correctly, you need to use their exact GitHub username. In this case, it should be "estebank"