Allow a reason="foo" in lint attributes

[emberian]

No description provided.

[emberian]

r? @brson

[alexcrichton]

It seems a bit odd to have a reason without actually using it anywhere. It seems like the only use case for this would be to have #[<lint level>(no_lint_reason)] to make sure that you justify all your lints. This never actually uses the message, although I suppose it does show up in the NOTE (but not guaranteed).

Also, would you mind adding some tests for lint attributes? They're easy to become wrong over time.

[emberian]

@alexcrichton well when it shows "lint level set here", the reason line will be there.

[emberian]

@alexcrichton is that test sufficient?

[alexcrichton]

Could you add some controls with scope? Just to make sure that it picks up changes in the lint level accordingly. Something like #[allow(...)] and then some without reasons.

[emberian]

I don't understand what you're asking for.

On Fri, Sep 6, 2013 at 2:13 PM, Alex Crichton [email protected]:

Could you add some controls with scope? Just to make sure that it picks up
changes in the lint level accordingly. Something like #[allow(...)] and
then some without reasons.

—
Reply to this email directly or view it on GitHubhttps://github.com//pull/9025#issuecomment-23958899
.

[alexcrichton]

#[deny(no_lint_reason)];

#[allow(no_lint_reason)]
mod foo {
  #[warn(unused_unsafe)];
}

should compile or something like that

[emberian]

#[deny(no_lint_reason, reason="")];

#[allow(no_lint_reason, reason="")]
mod foo {
    #[warn(unused_unsafe)];
}

fn main() { }

compiles; perhaps this isn't quite the desired functionality though.

[emberian]

@alexcrichton is this now acceptable?

[alexcrichton]

The code itself looks fine (a few stray error! statements though), but I'm not entirely convinced that this is necessary. I'd be curious what others thought about this as well.

[catamorphism]

I'll add this to the agenda for the next meeting.

[catamorphism]

We discussed this at the meeting and agreed the pull request is good, but it's not compelling enough of a feature to include. Sorry!