Fix meta variable misuse

[ia0]

No description provided.

[KodrAus]

Hi @ia0, are you able to give a quick summary of what this PR is all about and whether it's likely to break anything?

[ia0]

Hi @KodrAus,
This PR is about fixing #![deny(meta_variable_misuse)] lints (from rust-lang/rust#62008). It should not break anything that was not already broken, but one of the change can probably be done differently. I'll add a comment on the given change.

[ia0]

error: unknown macro variable `emeta`
  --> src/impl_error_chain_kind.rs:88:26
   |
88 |             enum [$( $(#[$emeta])* => $eitem $(( $($etyp),* ))* )*
   |                          ^^^^^^

error: unknown macro variable `eitem`
  --> src/impl_error_chain_kind.rs:88:39
   |
88 |             enum [$( $(#[$emeta])* => $eitem $(( $($etyp),* ))* )*
   |                                       ^^^^^^

error: unknown macro variable `etyp`
  --> src/impl_error_chain_kind.rs:88:52
   |
88 |             enum [$( $(#[$emeta])* => $eitem $(( $($etyp),* ))* )*
   |                                                    ^^^^^

Someone familiar with the code should be able to decide how to fix this error:

• Either remove the whole rule because it is never reached (most probable actually).
• Or adapt the right-hand side of the rule such that it does not reference unbound meta-variables (that's what I tried to do).
• Or adapt the left-hand side of the rule such that it binds $emeta, $eitem, and $etyp.

[ia0]

All following errors are similar to this one, so I'm just providing the output for this error.

error: meta-variable repeats with different Kleene operator
   --> src/impl_error_chain_kind.rs:106:55
    |
102 |         queue [($( $qvar:ident: $qtyp:ty ),+) $( $tail:tt )*]
    |                                            - expected repetition
...
106 |             buf [$( #[$bmeta] )* => $bitem: TUPLE [$( $qvar:$qtyp ),*] ]
    |                                                       ^^^^^         - conflicting repetition

[KodrAus]

Thanks for the explanation @ia0!

Ok, this looks good to me. While I'm not deeply familiar with these macros I know we've got a pretty reasonable test suite here so will trust it to have caught any regressions that might've come about.