Positional arg after repeated positional arg - prefer compiler error to runtime error

[nipunn1313]

Please complete the following tasks

• I have searched the discussions
• I have searched the existing issues

Clap Version

3.0.6

Describe your use case

use clap::Parser;

#[derive(clap::Parser, Debug)]
struct Args {
    arg2: Vec<String>,
    arg1: String,
}

fn main() {
    let args = Args::parse();
    dbg!(args);
}

This code is bad. Currently when running it, it fails at runtime

➜  clap_repro cargo run
   Compiling clap-repro v0.1.0 (/Users/nipunn/src/clap_repro)
    Finished dev [unoptimized + debuginfo] target(s) in 0.30s
     Running `target/debug/clap-repro`
thread 'main' panicked at 'Found non-required positional argument with a lower index than a required positional argument: "arg2" index Some(1)', /Users/nipunn/.cargo/registry/src/gitpro.ttaallkk.top-1ecc6299db9ec823/clap-3.0.6/src/parse/parser.rs:210:21
note: run with `RUST_BACKTRACE=1` environment variable to display a backtrace

Describe the solution you'd like

If possible, some kind of compile time error.
Probably would involve a proc-macro to iterate/logic over the struct and bail

Alternatives, if applicable

Status quo runtime error could work, but means you don't catch clear bugs until much later.
If we stick with status quo, a better error message might be nice.

"Found positional argument after a repeated positional argument"

Additional Context

No response

[epage]

#2720 is about supporting all debug asserts in the derive macro. As noted in that issue, there are complications with that. This falls into some of those complications because of flatten making it so we can't fully reason about this case. At best, we can implement a compile error only for when a flatten is not used, leaving it to runtime otherwise.

As noted in #2720, our CHANGELOG and Derive Tutorial's tips, we encourage having a test like below so these will be caught while testing (and for all subcommands, when relevant) as compared to manual testing / using the application.

#[derive(Parser)]
#[clap(author, version, about, long_about = None)]
struct Cli {
    ...
}

#[test]
fn verify_app() {
    use clap::IntoApp;
    Cli::into_app().debug_assert()
}

So like #2720, I lean towards closing this issue out as "won't fix". If there are thoughts or concerns about this, let us know!

[nipunn1313]

Sounds good. May still be nice to have a better runtime error message!

thread 'main' panicked at 'Found non-required positional argument with a lower index than a required positional argument: "arg2" index Some(1)'

was somewhat misleading to us. To my user, it seemed like it was user-error for providing the args in the wrong order (or something like that) - when actually it was a bug in the specification of the CLI by me.

[nipunn1313]

I also don't think debug_assert() catches this issue

Here's a cargo play compatible snippit - drop it into a main.rs. (cargo install cargo-play ; cargo play main.rs)

//# clap = {version="3", features=["derive"]}
use clap::{IntoApp, Parser};

#[derive(clap::Parser, Debug)]
struct Args {
    arg2: Vec<String>,
    arg1: String,
}

fn main() {
    Args::into_app().debug_assert();
    let args = Args::parse();
    dbg!(args);
}

If I run it, the debug_assert doesn't catch the issue, but Args::parse() fails at runtime.

[epage]

Interesting! This lives in a completely different place but doesn't need to. My guess is it predates some of how we do assertions today.