derive(Subcommand) does not respect attributes on fields of enum variants

[SlayerOfTheBad]

Please complete the following tasks

• I have searched the discussions
• I have searched the open and rejected issues

Rust Version

rustc 1.70.0 (90c541806 2023-05-31)

Clap Version

4.3.11

Minimal reproducible code

#[derive(Args)]
pub struct ListCommand {
    dir: PathBuf,
    limit: u64
}

#[derive(ExternalDerive, Subcommand)]
pub enum Commands {
    #[id(b"LS")]
    List(
        #[id(b"")]
        ListArgs
    )
}

#[derive(Parser)]
#[command(author, version, about, long_about = None)]
struct Cli {
    #[command(subcommand)]
    command: Commands,
}


fn main() {
    Cli::parse();
}

Where ExternalDerive is some derive macro that takes id as an attribute. In a private project currently, so I can not point to a crate with this derive macro.

Steps to reproduce the bug with the above code

cargo run list ~ 100

Actual Behaviour

Compilation fails due to the Subcommand derive macro.

error: expected type, found `#`
  --> audio_lib/src/actions/mod.rs:46:9
   |
46 |         #[id(b"")]
   |         ^ expected type

error: proc-macro derive produced unparsable tokens
  --> audio_lib/src/actions/mod.rs:42:33
   |
42 | #[derive(Debug, Compose, Parse, Subcommand)]
   |                                 ^^^^^^^^^^

error: could not compile `audio_lib` (lib) due to 2 previous errors

Expected Behaviour

I expect the code to compile, and to run, and then exit without doing anything.

Additional Context

No response

Debug Output

No response

[epage]

Could you include a full reproduction case? Without access to ExternalDerive, I can't reproduce this to determine root cause.

[SlayerOfTheBad]

Of course, I have made a minimal reproduction repository here.

It's reproduced simply by running cargo run --bin bin

[epage]

Examples of bad generated code:

                       {
                           return :: std :: result :: Result ::
                           Ok(Self ::
                           List(< #[id(b"")] ListArgs as clap :: FromArgMatches > ::
                           from_arg_matches_mut(__clap_arg_matches) ?))
                       }
// ...
               fn augment_subcommands < 'b > (__clap_app : clap :: Command) -> clap ::
               Command
               {
                   ; let __clap_app = __clap_app ; let __clap_app =
                   __clap_app.subcommand({
                       let __clap_subcommand = clap :: Command :: new("list") ; let
                       __clap_subcommand = __clap_subcommand ; let __clap_subcommand =
                       {
                           < #[id(b"")] ListArgs as clap :: Args > ::
                           augment_args(__clap_subcommand)
                       } ; __clap_subcommand
                   }) ; ; __clap_app
               } fn augment_subcommands_for_update < 'b > (__clap_app : clap :: Command)
               -> clap :: Command
               {
                   ; let __clap_app = __clap_app ; let __clap_app =
                   __clap_app.subcommand({
                       let __clap_subcommand = clap :: Command :: new("list") ; let
                       __clap_subcommand = __clap_subcommand ; let __clap_subcommand =
                       {
                           < #[id(b"")] ListArgs as clap :: Args > ::
                           augment_args_for_update(__clap_subcommand)
                       } ; __clap_subcommand
                   }) ; ; __clap_app
// ...

[epage]

Thanks! That was a big help to see the exact reproduction case you were dealing with

[SlayerOfTheBad]

Of course!
Glad I could help in resolving this issue. I'm surprised it'd never come up before honestly.